From harttig at mail.dife.de Tue Aug 2 11:42:38 2005
From: harttig at mail.dife.de (Ulrich Harttig)
Date: Tue Aug 2 11:31:45 2005
Subject: [MOBY-l] parameter handling problems
Message-ID: <42EF946E.4090708@mail.dife.de>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
I have some problems with handling a message to a service asking for an input
object and a secondary parameter using the Perl API (and XML::LibXML version
1.58, libxml2 version 2.6.16)
After parsing the message with 'complexServiceInputParser()' extraction of the
the simple article is no problem but accessing the content of the secondary
article with
($secondary) = getNodeContentWithArticle($input, "Parameter", "VarType");
does not find any content although the $input node contains
xml
Is there any reason why this shouldn't work ?
Could someone point me to examples where accessing parameters works ?
Thanks for your help
Best regards
Ulrich
- --
Ulrich Harttig
German Institute of Human Nutrition Potsdam-Rehbruecke
Dept. of Epidemiology
Arthur-Scheunert-Allee 114-116, 14558 Nuthetal, Germany
phone: +49 (0)33200 88-729, fax: +49 (0)33200 88-721
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)
iD8DBQFC75RueYdmlKJoHzkRAuKeAJ9QkwlecCYbquhW6W1u1y9arXzPAwCg0pjG
64tQh4osNp61zpuCQYo5VcI=
=AOm+
-----END PGP SIGNATURE-----
From harttig at mail.dife.de Tue Aug 2 11:42:38 2005
From: harttig at mail.dife.de (Ulrich Harttig)
Date: Tue Aug 2 11:40:25 2005
Subject: [MOBY-l] parameter handling problems
Message-ID: <42EF946E.4090708@mail.dife.de>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
I have some problems with handling a message to a service asking for an input
object and a secondary parameter using the Perl API (and XML::LibXML version
1.58, libxml2 version 2.6.16)
After parsing the message with 'complexServiceInputParser()' extraction of the
the simple article is no problem but accessing the content of the secondary
article with
($secondary) = getNodeContentWithArticle($input, "Parameter", "VarType");
does not find any content although the $input node contains
xml
Is there any reason why this shouldn't work ?
Could someone point me to examples where accessing parameters works ?
Thanks for your help
Best regards
Ulrich
- --
Ulrich Harttig
German Institute of Human Nutrition Potsdam-Rehbruecke
Dept. of Epidemiology
Arthur-Scheunert-Allee 114-116, 14558 Nuthetal, Germany
phone: +49 (0)33200 88-729, fax: +49 (0)33200 88-721
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)
iD8DBQFC75RueYdmlKJoHzkRAuKeAJ9QkwlecCYbquhW6W1u1y9arXzPAwCg0pjG
64tQh4osNp61zpuCQYo5VcI=
=AOm+
-----END PGP SIGNATURE-----
From d.haase at gsf.de Wed Aug 3 02:25:20 2005
From: d.haase at gsf.de (Dirk Haase)
Date: Wed Aug 3 02:16:18 2005
Subject: [MOBY-l] parameter handling problems
In-Reply-To: <42EF946E.4090708@mail.dife.de>
References: <42EF946E.4090708@mail.dife.de>
Message-ID: <200508030825.20164.d.haase@gsf.de>
On Tuesday 02 August 2005 17:42, Ulrich Harttig wrote:
> ($secondary) = getNodeContentWithArticle($input, "Parameter", "VarType");
>
> does not find any content although the $input node contains
>
>
> xml
>
>
> Is there any reason why this shouldn't work ?
If I remember correctly, I tried the same and didn't work for me neither...
> Could someone point me to examples where accessing parameters works ?
Here is what works for me: the service takes several secondaries, so it puts
all parameter values into a %paramHash:
if ( $articleType == SECONDARY ) {
my $param = $input->getAttribute('articleName');
my @valueNodes = $input->getChildrenByTagName('moby:Value');
my $value = $valueNodes[0]->textContent(); # there should be only one value
$paramHash{$param} = $value;
}
Hope this helps.
Regards,
dirk
From harttig at mail.dife.de Wed Aug 3 03:48:34 2005
From: harttig at mail.dife.de (Ulrich Harttig)
Date: Wed Aug 3 03:34:52 2005
Subject: [MOBY-l] parameter handling problems
In-Reply-To: <200508030825.20164.d.haase@gsf.de>
References: <42EF946E.4090708@mail.dife.de> <200508030825.20164.d.haase@gsf.de>
Message-ID: <42F076D2.8060806@mail.dife.de>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Dirk Haase wrote:
> On Tuesday 02 August 2005 17:42, Ulrich Harttig wrote:
>
>
>>($secondary) = getNodeContentWithArticle($input, "Parameter", "VarType");
>>
>>does not find any content although the $input node contains
>>
>>
>> xml
>>
>>
>>Is there any reason why this shouldn't work ?
>
>
> If I remember correctly, I tried the same and didn't work for me neither...
>
>
>>Could someone point me to examples where accessing parameters works ?
>
>
> Here is what works for me: the service takes several secondaries, so it puts
> all parameter values into a %paramHash:
>
> if ( $articleType == SECONDARY ) {
> my $param = $input->getAttribute('articleName');
> my @valueNodes = $input->getChildrenByTagName('moby:Value');
> my $value = $valueNodes[0]->textContent(); # there should be only one value
> $paramHash{$param} = $value;
> }
>
> Hope this helps.
>
> Regards,
> dirk
Thanks a lot for the help, your example did work.
But I still wondered why $input->getChildrenByTagName('moby:Value') worked but
$input->getElementByTagName('moby:Parameter') would not. I was guessing that
this had something to do with the structure/content of the node from the
secondary article.
getElementsByTagName is supposed to return a named descendant of the node.
According to the DOM spec: "A descendant node of any node A is any node below A
in a tree model, where "below" means "away from the root."
But to me it seems that is actually the root of the node ,
judging by the content of the $input parameter, so one would not be able to find
a descendant of the same name.
So if I pass instead the original message using the $input->parentNode to the
CommonSub function 'getNodeContentWithArticle' it then returns the desired
parameter value
($secondary) = getNodeContentWithArticle($input->parentNode, "Parameter",
"VarType");
best regards
Ulrich
- --
Ulrich Harttig
German Institute of Human Nutrition Potsdam-Rehbruecke
Dept. of Epidemiology
Arthur-Scheunert-Allee 114-116, 14558 Nuthetal, Germany
phone: +49 (0)33200 88-729, fax: +49 (0)33200 88-721
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)
iD8DBQFC8HbSeYdmlKJoHzkRAp6ZAJ98gonlRj76yBlZEf8feQYd+fx9QACgtVqI
qDx3gY2ljO29ZcdGqy9QZ14=
=OuRI
-----END PGP SIGNATURE-----
From rebecca.ernst at gsf.de Thu Aug 4 03:11:25 2005
From: rebecca.ernst at gsf.de (Rebecca Ernst)
Date: Thu Aug 4 03:02:16 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
Message-ID: <42F1BF9D.6040401@gsf.de>
Hi there!
It might be boring for you that I start this topic again but the problem
still exists and within Planet we are getting in serious problems now
and I guess other people will also run into them!
Fact is, that old workflows don't work in Taverna 1.2 . According to Tom
this is not how it should be, software should always be downwards
compatible and that's how I see it, too.
Now, due to the fact that it doesn't work but people want to make use of
the new, cool features of 1.2 they start developing services which run
on 1.2. as you can guess they don't work in 1.1. anymore.
It can't be a solution to develop biomoby workflows for every Taverna
version so could we please try to solve this problem now before lots of
people will have to change their services back and forth ??
If noone (except of Heiko Dirk and me) want the collections to be
changed in the BioMoby API we will have to change the default behaviour
of Taverna back to the old one!
Cheers,
Rebecca
--
Rebecca Ernst
MIPS, Inst. for Bioinformatics
GSF Research Center for Environment and Health
Ingolstaedter Landstr. 1
85764 Neuherberg
fon: +49 89 3187 3583
email: Rebecca.Ernst@gsf.de
From senger at ebi.ac.uk Thu Aug 4 03:21:48 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Thu Aug 4 03:12:15 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1BF9D.6040401@gsf.de>
Message-ID:
Rebecca,
I have talked about it with Eddie when I met him in Malaga about
two/three weeks ago. And he promised (!) that he was going to put back the
old CreateMoby... and ParseMoby... classes. And I have never heard from
him since. So it is very good that you started this topic again - perhaps
he is listening :-)
Cheers,
Martin
--
Martin Senger martin.senger@gmail.com
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-580-5600 (ext.2324)
From edward.kawas at gmail.com Thu Aug 4 03:24:25 2005
From: edward.kawas at gmail.com (Edward Kawas)
Date: Thu Aug 4 03:15:23 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To:
Message-ID: <42f1c2c1.6263d1d9.1fa1.67e4@mx.gmail.com>
Hi,
I have put the old functionality back into Taverna.
If you build a fresh copy from the cvs, you will notice
(hopefully) that all is well again.
Eddie
> -----Original Message-----
> From: moby-l-bounces@portal.open-bio.org [mailto:moby-l-
> bounces@portal.open-bio.org] On Behalf Of Martin Senger
> Sent: Thursday, August 04, 2005 12:22 AM
> To: Rebecca Ernst
> Cc: taverna-users@lists.sourceforge.net; mobyl
> Subject: Re: [MOBY-l] Biomoby and Taverna 1.2
>
> Rebecca,
> I have talked about it with Eddie when I met him in
> Malaga about
> two/three weeks ago. And he promised (!) that he was going
> to put back the
> old CreateMoby... and ParseMoby... classes. And I have
> never heard from
> him since. So it is very good that you started this topic
> again - perhaps
> he is listening :-)
>
> Cheers,
> Martin
>
> --
> Martin Senger martin.senger@gmail.com
>
> International Rice Research Institute
> Biometrics and Bioinformatics Unit
> DAPO BOX 7777, Metro Manila
> Philippines, phone: +63-580-5600 (ext.2324)
>
> _______________________________________________
> moby-l mailing list
> moby-l@biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
From markw at illuminae.com Thu Aug 4 03:48:35 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu Aug 4 03:39:00 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1BF9D.6040401@gsf.de>
References: <42F1BF9D.6040401@gsf.de>
Message-ID: <1123141715.42f1c853bd83f@webmail.illuminae.com>
> If noone (except of Heiko Dirk and me) want the collections to be
> changed in the BioMoby API we will have to change the default behaviour
> of Taverna back to the old one!
I think your use-case for wanting the behaviour of collections to change is
unfounded, since it overloads the intended meaning of the Collection element by
putting WAAAAAAY too much semantics in it. The use case of having sets of
multiple alignments represented as Collections of Collections means that the
inner-most Collection now not only represents a "bag" of things, but a "bag" of
things that **DO** have a close semantic relationship to each other - these are
a set of related sequences, and a set of related sequences **is a useful
Object** in the context of MOBY, could easily trigger the discovery of a
particular set of services that can specifically operate on sets of closely
related sequences (versus sets of unrelated sequences), and therefore should be
given a class of its own (IMO).
Moreover, if we allow collections of collections, we then have a problem of
recursion - collections of collections of collections of collections of
collections.... What a nightmare!
The API change is not a significant one, I agree... and in fact, since people
are already mis-using the messaging format in this way much of the software out
there can already handle the situation... but given that particular use-case, I
think you are losing more semantics than you need to be by refusing to create a
new Object Class.
M
From tmo at ebi.ac.uk Thu Aug 4 04:06:45 2005
From: tmo at ebi.ac.uk (Tom Oinn)
Date: Thu Aug 4 03:56:07 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123141715.42f1c853bd83f@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
Message-ID: <42F1CC95.9040803@ebi.ac.uk>
markw@illuminae.com wrote:
>>If noone (except of Heiko Dirk and me) want the collections to be
>>changed in the BioMoby API we will have to change the default behaviour
>>of Taverna back to the old one!
>
>
> I think your use-case for wanting the behaviour of collections to change is
> unfounded, since it overloads the intended meaning of the Collection element by
> putting WAAAAAAY too much semantics in it. The use case of having sets of
> multiple alignments represented as Collections of Collections means that the
> inner-most Collection now not only represents a "bag" of things, but a "bag" of
> things that **DO** have a close semantic relationship to each other - these are
> a set of related sequences, and a set of related sequences **is a useful
> Object** in the context of MOBY, could easily trigger the discovery of a
> particular set of services that can specifically operate on sets of closely
> related sequences (versus sets of unrelated sequences), and therefore should be
> given a class of its own (IMO).
True, but you probably also want to allow, for example, a GC percentage
service to run over that set of sets and produce a set of sets of scores
- if the set of sequences becomes some opaque type without this
information then you can't do this. Strikes me that you need a more
sophisticated typing scheme to do this, perhaps extending the Collection
concept to allow the imposition of semantics to the group of child
objects? I think you do need nested collections though, it's not
impossible to deal with (we have some slightly twisted code in Taverna
to do it but it works well enough)
Tom
From rebecca.ernst at gsf.de Thu Aug 4 04:16:35 2005
From: rebecca.ernst at gsf.de (Rebecca Ernst)
Date: Thu Aug 4 04:07:00 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123141715.42f1c853bd83f@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
Message-ID: <42F1CEE3.6020901@gsf.de>
Hi Mark!
I already understood that noone agrees with us ;-)
Funny thing is that everyone keeps talking about collections of
collections which Dirk never wanted.
snip from Dirk:
Correct. That's why we did not ask for nested collections. We suggest to allow
for multiple Simple output as default - which I think is very appropriate
because especially query services will usually return more than one result.
This way we would save the collection concept for semantically related
entities.
I am still not sure if anyone has really understood why we have problems
with the collections as they are used now but we are happy to leave them
as they are .
Although if I think about it I have to ask one more provoking question
( ;-) ) : Why do we have Simples in BioMoby then? I only have VERY few
services that give back simples according to the API. From my point of
view we could also leave Simples away as they make no sense to me. It is
of no valuable information to me if a service gives back exactly three
or an unknown number of AGI codes....
But as I said already - if everyone is happy with the current use of
collections I will be quiet from now on :-)
The problem is then only shiftet to the clients (As seen in Taverna it
leaves a much bigger job for the clients as they will have to split the
collections)
Cheers,
Rebecca
And you guys in Vancouver would rather go to bed now :-)
markw@illuminae.com wrote:
>>If noone (except of Heiko Dirk and me) want the collections to be
>>changed in the BioMoby API we will have to change the default behaviour
>>of Taverna back to the old one!
>>
>>
>
>I think your use-case for wanting the behaviour of collections to change is
>unfounded, since it overloads the intended meaning of the Collection element by
>putting WAAAAAAY too much semantics in it. The use case of having sets of
>multiple alignments represented as Collections of Collections means that the
>inner-most Collection now not only represents a "bag" of things, but a "bag" of
>things that **DO** have a close semantic relationship to each other - these are
>a set of related sequences, and a set of related sequences **is a useful
>Object** in the context of MOBY, could easily trigger the discovery of a
>particular set of services that can specifically operate on sets of closely
>related sequences (versus sets of unrelated sequences), and therefore should be
>given a class of its own (IMO).
>
>Moreover, if we allow collections of collections, we then have a problem of
>recursion - collections of collections of collections of collections of
>collections.... What a nightmare!
>
>The API change is not a significant one, I agree... and in fact, since people
>are already mis-using the messaging format in this way much of the software out
>there can already handle the situation... but given that particular use-case, I
>think you are losing more semantics than you need to be by refusing to create a
>new Object Class.
>
>M
>
>_______________________________________________
>moby-l mailing list
>moby-l@biomoby.org
>http://biomoby.org/mailman/listinfo/moby-l
>
>
>
--
Rebecca Ernst
MIPS, Inst. for Bioinformatics
GSF Research Center for Environment and Health
Ingolstaedter Landstr. 1
85764 Neuherberg
fon: +49 89 3187 3583
email: Rebecca.Ernst@gsf.de
From markw at illuminae.com Thu Aug 4 04:26:14 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu Aug 4 04:21:37 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1CEE3.6020901@gsf.de>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CEE3.6020901@gsf.de>
Message-ID: <1123143974.42f1d1268d031@webmail.illuminae.com>
Quoting Rebecca Ernst :
> And you guys in Vancouver would rather go to bed now :-)
Granted, I would rather be in bed, but neither Eddie nor I are in Vancouver so
it would be inappropriate :-)
we're both just a bit north of you in Manchester!
M
From markw at illuminae.com Thu Aug 4 04:32:30 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu Aug 4 04:23:19 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1CC95.9040803@ebi.ac.uk>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
Message-ID: <1123144350.42f1d29e7978b@webmail.illuminae.com>
Quoting Tom Oinn :
> - if the set of sequences becomes some opaque type without this
> information then you can't do this.
Nothing becomes opaque by putting it in the MOBY data-typing system - it can
only get richer and more meaningful by assigning it a specific class :-)
Generic services that operate on lower ontological classes will still be
discovered and executed, so this use case still fails the test of need.
> Strikes me that you need a more
> sophisticated typing scheme to do this, perhaps extending the Collection
> concept to allow the imposition of semantics to the group of child
> objects?
Creating a new type of element - limitedCollection or something like that -
would be an alternative... or changing the definition of Collection to "zero or
more" rather than "one or moe" would also allow us to terminate the recursion.
> I think you do need nested collections though, it's not
> impossible to deal with (we have some slightly twisted code in Taverna
> to do it but it works well enough)
Probably we do... but I find it hard to believe that we will have sets of sets
of sets of sets where there is no implicit semantic meaning at any level! As
soon as there is semantic meaning, it should be a fully qualified MOBY Object
rather than an arbitrary collection.... IMO :-)
M
>
> Tom
>
From markw at illuminae.com Thu Aug 4 04:41:19 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu Aug 4 04:31:40 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123143974.42f1d1268d031@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CEE3.6020901@gsf.de>
<1123143974.42f1d1268d031@webmail.illuminae.com>
Message-ID: <1123144879.42f1d4af9ab8e@webmail.illuminae.com>
Quoting markw@illuminae.com:
> Granted, I would rather be in bed, but neither Eddie nor I are in Vancouver
> so
> it would be inappropriate :-)
Gosh... I *am* tired!
I didn't mean to imply that....
:-/
M
From rebecca.ernst at gsf.de Thu Aug 4 04:50:34 2005
From: rebecca.ernst at gsf.de (Rebecca Ernst)
Date: Thu Aug 4 05:00:58 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123144879.42f1d4af9ab8e@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CEE3.6020901@gsf.de>
<1123143974.42f1d1268d031@webmail.illuminae.com>
<1123144879.42f1d4af9ab8e@webmail.illuminae.com>
Message-ID: <42F1D6DA.2060200@gsf.de>
Mark wrote:
> Gosh... I *am* tired!
> I didn't mean to imply that....
> :-/
Hehe - it took me a while to get the meaning 'between the lines' ;-)
As you are in Manchestser we have the whole european day to discuss
about collections ;-) With Tom you can even chat sitting together if you
are so close ...
I was thinking about a way to tell you what my problem with this simple
and collection thing is and I'll give it a very last try with a trivial
example:
My favourite example to explain BioMoby services to non- informatics
people is that I am a service provider running a laundry. Incoming
objects are dirty laundry, out go clean clothes.
In your way of interpreting services the man in the laundry gets the
information that you are either always bringing exactly 3 shirts to wash
or you are bringing a not known number of shirts (one time you bring 2,
the next time 15 shirts but it's *always* only shirts)
What I am saying is that it *can't* be of any use to this guy to know
how many shirts you'll bring - he will have to wash all of them, one by one.
But if someone is bringing three complete outfits belonging together (a
jacket, a trouser, a shirt and socks ) I need to take care about this
relationship and put them back together before giving it back.
Does this make sense to you??
Rebecca
markw@illuminae.com wrote:
>Quoting markw@illuminae.com:
>
>
>
>>Granted, I would rather be in bed, but neither Eddie nor I are in Vancouver
>>so
>>it would be inappropriate :-)
>>
>>
>
>Gosh... I *am* tired!
>
>I didn't mean to imply that....
>
>:-/
>
>M
>
>
>
>
>
--
Rebecca Ernst
MIPS, Inst. for Bioinformatics
GSF Research Center for Environment and Health
Ingolstaedter Landstr. 1
85764 Neuherberg
fon: +49 89 3187 3583
email: Rebecca.Ernst@gsf.de
From tmo at ebi.ac.uk Thu Aug 4 05:11:51 2005
From: tmo at ebi.ac.uk (Tom Oinn)
Date: Thu Aug 4 05:01:11 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123144350.42f1d29e7978b@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
<1123144350.42f1d29e7978b@webmail.illuminae.com>
Message-ID: <42F1DBD7.4000700@ebi.ac.uk>
markw@illuminae.com wrote:
> Quoting Tom Oinn :
>
>
>>- if the set of sequences becomes some opaque type without this
>>information then you can't do this.
>
>
> Nothing becomes opaque by putting it in the MOBY data-typing system - it can
> only get richer and more meaningful by assigning it a specific class :-)
> Generic services that operate on lower ontological classes will still be
> discovered and executed, so this use case still fails the test of need.
Okay, so I might be failing to entirely understand the moby typing
system. In the case I describe I need to define a type such that :
1) It can be discovered as 'set of aligned sequences' as a single entity.
2) It can also be discovered as a collection of 'sequence'
As I understood it the collection / simple wasn't part of the ontology
so this isn't possible? Effectively I believe that 'set of aligned
sequences' cannot be made a restriction of 'collection of 'sequence'' as
the collection concept isn't part of the ontology - is this wrong?
Tom
From rebecca.ernst at gsf.de Thu Aug 4 04:58:00 2005
From: rebecca.ernst at gsf.de (Rebecca Ernst)
Date: Thu Aug 4 05:10:47 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42f1c2c1.6263d1d9.1fa1.67e4@mx.gmail.com>
References: <42f1c2c1.6263d1d9.1fa1.67e4@mx.gmail.com>
Message-ID: <42F1D898.7040006@gsf.de>
Hi Eddie!
that's good news! Thanks a lot for this!
Is it possible to update the packaged one also? I don't have a cvs
checkout and I guess most of the other people also use the packages.
Cheers,
Rebecca
Edward Kawas wrote:
>Hi,
>
>I have put the old functionality back into Taverna.
>
>If you build a fresh copy from the cvs, you will notice
>(hopefully) that all is well again.
>
>Eddie
>
>
>
>>-----Original Message-----
>>From: moby-l-bounces@portal.open-bio.org [mailto:moby-l-
>>bounces@portal.open-bio.org] On Behalf Of Martin Senger
>>Sent: Thursday, August 04, 2005 12:22 AM
>>To: Rebecca Ernst
>>Cc: taverna-users@lists.sourceforge.net; mobyl
>>Subject: Re: [MOBY-l] Biomoby and Taverna 1.2
>>
>>Rebecca,
>> I have talked about it with Eddie when I met him in
>>Malaga about
>>two/three weeks ago. And he promised (!) that he was going
>>to put back the
>>old CreateMoby... and ParseMoby... classes. And I have
>>never heard from
>>him since. So it is very good that you started this topic
>>again - perhaps
>>he is listening :-)
>>
>> Cheers,
>> Martin
>>
>>--
>>Martin Senger martin.senger@gmail.com
>>
>>International Rice Research Institute
>>Biometrics and Bioinformatics Unit
>>DAPO BOX 7777, Metro Manila
>>Philippines, phone: +63-580-5600 (ext.2324)
>>
>>_______________________________________________
>>moby-l mailing list
>>moby-l@biomoby.org
>>http://biomoby.org/mailman/listinfo/moby-l
>>
>>
>
>_______________________________________________
>moby-l mailing list
>moby-l@biomoby.org
>http://biomoby.org/mailman/listinfo/moby-l
>
>
>
--
Rebecca Ernst
MIPS, Inst. for Bioinformatics
GSF Research Center for Environment and Health
Ingolstaedter Landstr. 1
85764 Neuherberg
fon: +49 89 3187 3583
email: Rebecca.Ernst@gsf.de
From markw at illuminae.com Thu Aug 4 05:24:08 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu Aug 4 05:14:42 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1DBD7.4000700@ebi.ac.uk>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
<1123144350.42f1d29e7978b@webmail.illuminae.com>
<42F1DBD7.4000700@ebi.ac.uk>
Message-ID: <1123147448.42f1deb87303f@webmail.illuminae.com>
Quoting Tom Oinn :
> 1) It can be discovered as 'set of aligned sequences' as a single entity.
>
> 2) It can also be discovered as a collection of 'sequence'
>
> As I understood it the collection / simple wasn't part of the ontology
> so this isn't possible?
You're correct, and that is a good point!
I think that the semantics you lose by making them a simple bag of sequences is
more important than the flexibility of service-discovery that you gain. This is
why we (Eddie) is working on MOBY Object degradation widgets for Taverna, such
that you strip away the outer "semantic" tag and are left with a collection of
sequences that could then trigger the latter discovery that you mention above.
I don't think we can achieve both situations in one step, unless we defined a
parent class "genericSequenceCollection"... but then we would have to do that
for all Object classes and our bicycles would explode!
M
From markw at illuminae.com Thu Aug 4 05:25:38 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu Aug 4 05:16:04 2005
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1D6DA.2060200@gsf.de>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CEE3.6020901@gsf.de>
<1123143974.42f1d1268d031@webmail.illuminae.com>
<1123144879.42f1d4af9ab8e@webmail.illuminae.com>
<42F1D6DA.2060200@gsf.de>
Message-ID: <1123147538.42f1df121b200@webmail.illuminae.com>
> With Tom you can even chat sitting together if you
> are so close ...
>
England is small... but not THAT small :-)
I'll think about your dirty laundry later - must get some other things done!
M
From tmo at ebi.ac.uk Thu Aug 4 06:03:19 2005
From: tmo at ebi.ac.uk (Tom Oinn)
Date: Thu Aug 4 05:52:01 2005
Subject: [Taverna-users] Re: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123147448.42f1deb87303f@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
<1123144350.42f1d29e7978b@webmail.illuminae.com>
<42F1DBD7.4000700@ebi.ac.uk>
<1123147448.42f1deb87303f@webmail.illuminae.com>
Message-ID: <42F1E7E7.3060109@ebi.ac.uk>
markw@illuminae.com wrote:
> Quoting Tom Oinn :
>
>
>>1) It can be discovered as 'set of aligned sequences' as a single entity.
>>
>>2) It can also be discovered as a collection of 'sequence'
>>
>>As I understood it the collection / simple wasn't part of the ontology
>>so this isn't possible?
>
> You're correct, and that is a good point!
Nice to know my brain's working again :)
I don't think a decomposition widget is the right way to go - the data
in question _is_ a group of sequences, however that's described. If
you're using RDF to describe the types can't you express the above? I
think there's a misconception that semantic types are the same as
classical object types in e.g. Java.
In the latter case an entity has exactly one type (ignoring things like
interfaces!) and that type then has inheritance and restriction
relationships. This causes issues when you want to use that entity in a
different way - an example of this is in Taverna's code where we have a
type that represents a user interface widget that can bind to a workflow
model but we also know that all such widgets are instances of one of the
core Java ui classes. Because Java doesn't have multiple inheritance or
mixins we can't express this in the Java type system - I believe this is
exactly analogous to the case I described above.
In the case of a semantic description you can make statements describing
the entity, in the current Moby implementation these statements are
pretty much hardcoded as 'entity has type foo', which makes them
directly equivalent to the types used in languages such as Java, with
all the same issues. There's no fundamental reason you need to do this
though, with an extension into generalised descriptions you should be
able to say 'entity is represented by collection of sequences which are
related by...'. The key is to stop thinking of semantic types as being a
lazy way of growing an object heirarchy and actually take advantage of
their full expressive power. I can appreciate a desire to keep things
simple but we have a clear case here where we need something a bit smarter.
Could someone, for example Phil, who actually knows what they're talking
about on these lists perhaps give an example of how this would look in
RDF? I know it's possible to do but a concrete example might be helpful
and I don't have the required knowledge.
Tom
From tmo at ebi.ac.uk Thu Aug 4 06:07:28 2005
From: tmo at ebi.ac.uk (Tom Oinn)
Date: Thu Aug 4 05:56:04 2005
Subject: [Taverna-users] Re: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1E7E7.3060109@ebi.ac.uk>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
<1123144350.42f1d29e7978b@webmail.illuminae.com>
<42F1DBD7.4000700@ebi.ac.uk>
<1123147448.42f1deb87303f@webmail.illuminae.com>
<42F1E7E7.3060109@ebi.ac.uk>
Message-ID: <42F1E8E0.6050102@ebi.ac.uk>
Tom Oinn wrote:
> Could someone, for example Phil, who actually knows what they're talking
> about on these lists perhaps give an example of how this would look in
> RDF?
Um, that wasn't meant to imply that anyone doesn't know what they're
talking about - sounded a bit harsh on re-reading! It would be good to
have some input from the mygrid semantics folks though especially as
this issue is directly related to the description of services and
therefore the mygrid registry activities.
Tom
From p.lord at cs.man.ac.uk Thu Aug 4 12:17:28 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: Thu Aug 4 12:07:39 2005
Subject: [Taverna-users] Re: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1E7E7.3060109@ebi.ac.uk>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
<1123144350.42f1d29e7978b@webmail.illuminae.com>
<42F1DBD7.4000700@ebi.ac.uk>
<1123147448.42f1deb87303f@webmail.illuminae.com>
<42F1E7E7.3060109@ebi.ac.uk>
Message-ID:
>>>>> "Tom" == Tom Oinn writes:
Tom> Could someone, for example Phil, who actually knows what
Tom> they're talking about on these lists perhaps give an example of
Tom> how this would look in RDF? I know it's possible to do but a
Tom> concrete example might be helpful and I don't have the required
Tom> knowledge.
I'm not sure that I would make any claims to be the worlds greatest
expert on RDF. Actually, I would make claims to not be any kind of
expert on RDF.
You can certainly express the notion of a collection in RDF. You can
put semantic types onto any of the nodes underneath this, and there is
very little restriction as to what the types of these things can
be. You can state that for a rdf:bag that the first is a dna sequence,
the next a protein or whatever. The rdf looks something like
a ---- has type ---> bag
a ---- li ----> b
a ---- li ----> c
b ---- hastype -----> dna sequence
b ---- hasvalue -----> "GATC"
c ---- hastype ---> protein sequence
c ---- hasvalue ---> "MAGSYDJ"
If you want this level of complexity in moby-s messages then there is
no particular problem within representing it in the XML, at least as
far as I can see. The problem comes from querying, because people are
likely to want to query for services which consume, for example,
"DNASequence, or some collection which recursively contains a DNA
sequence".
Neither moby-central nor feta are going to be terribly happy with
performing this sort of query.
Phil
From bgulden at purdue.edu Thu Aug 4 14:08:12 2005
From: bgulden at purdue.edu (bgulden@purdue.edu)
Date: Thu Aug 4 13:58:15 2005
Subject: [MOBY-l] Submitting Parameters
Message-ID: <1123178892.42f2598c4e9bf@webmail.purdue.edu>
First, I just wanted to thank you all for helping with my previous problems
with parameters (The ->secondary function worked perfectly). Now on to my
current dilemma.
Is there a special way to include parameters in the execution of Moby
services? According to the perl API services with parameters should be called
in this format where input1 is a primary input and param1 is a secondary input:
$Service->execute(XMLinputlist => [
[
'input1', ' ',
'param1', '0.001 ',
]
]);
However, when I check the actual inputted data to my service I end up with one
tag around both input1 and param1.
Thanks,
Brian
From d.haase at gsf.de Fri Aug 5 04:18:41 2005
From: d.haase at gsf.de (Dirk Haase)
Date: Fri Aug 5 04:12:11 2005
Subject: [MOBY-l] Submitting Parameters
In-Reply-To: <1123178892.42f2598c4e9bf@webmail.purdue.edu>
References: <1123178892.42f2598c4e9bf@webmail.purdue.edu>
Message-ID: <200508051018.41494.d.haase@gsf.de>
On Thursday 04 August 2005 20:08, bgulden@purdue.edu wrote:
> First, I just wanted to thank you all for helping with my previous problems
> with parameters (The ->secondary function worked perfectly). Now on to my
> current dilemma.
>
> Is there a special way to include parameters in the execution of Moby
> services? According to the perl API services with parameters should be
> called in this format where input1 is a primary input and param1 is a
> secondary input: $Service->execute(XMLinputlist => [
> [
> 'input1', ' ',
> 'param1', '0.001 ',
> ]
> ]);
>
> However, when I check the actual inputted data to my service I end up with
> one tag around both input1 and param1.
Hmm, your code seems correct, should actually work... Are you sure the service is correctly registered?
Regards,
dirk
--
----------------------------------------------------------
Dirk Haase phone +49 89 3187 3583
http://mips.gsf.de/~haase email d.haase@gsf.de
From markw at illuminae.com Fri Aug 5 04:28:18 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Fri Aug 5 04:19:15 2005
Subject: [MOBY-l] Submitting Parameters
In-Reply-To: <200508051018.41494.d.haase@gsf.de>
References: <1123178892.42f2598c4e9bf@webmail.purdue.edu>
<200508051018.41494.d.haase@gsf.de>
Message-ID: <1123230498.42f323223100a@webmail.illuminae.com>
There is a known bug in the current production code relating to the registration
of parameters - if any of them have a default, a min, or a max of "0", the
registration of that parameter will fail, but the service registration will
succeed. This might be what he is seeing.
:-/
It is fixed in the new codebase, but we wont be updating the production server
for a week or two.
M
Quoting Dirk Haase :
> On Thursday 04 August 2005 20:08, bgulden@purdue.edu wrote:
> > First, I just wanted to thank you all for helping with my previous problems
> > with parameters (The ->secondary function worked perfectly). Now on to my
> > current dilemma.
> >
> > Is there a special way to include parameters in the execution of Moby
> > services? According to the perl API services with parameters should be
> > called in this format where input1 is a primary input and param1 is a
> > secondary input: $Service->execute(XMLinputlist => [
> > [
> > 'input1', ' id="123"/>',
> > 'param1', '0.001 ',
> > ]
> > ]);
> >
> > However, when I check the actual inputted data to my service I end up with
> > one tag around both input1 and param1.
>
> Hmm, your code seems correct, should actually work... Are you sure the
> service is correctly registered?
>
> Regards,
> dirk
>
> --
>
> ----------------------------------------------------------
> Dirk Haase phone +49 89 3187 3583
> http://mips.gsf.de/~haase email d.haase@gsf.de
> _______________________________________________
> moby-l mailing list
> moby-l@biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
>
From edward.kawas at gmail.com Mon Aug 8 09:42:55 2005
From: edward.kawas at gmail.com (Edward Kawas)
Date: Mon Aug 8 09:33:30 2005
Subject: [MOBY-l] RE: [MOBY-dev] A bug in Taverna 1.2 perhaps related to
BioMOBY
In-Reply-To: <42F3B988.7020506@cnb.uam.es>
Message-ID: <42f76179.195ac951.0c6c.100f@mx.gmail.com>
Hi,
I think that that has been fixed. I actually noticed that
bug when using the Planet workflows. So, like I said, it
should be fixed.
Out of curiosity, did you cvs build or download a
prepackaged Taverna?
Thanks,
Eddie
> -----Original Message-----
> From: moby-dev-bounces@portal.open-bio.org [mailto:moby-
> dev-bounces@portal.open-bio.org] On Behalf Of Jos? Mar?a
> Fern?ndez Gonz?lez
> Sent: Friday, August 05, 2005 12:10 PM
> To: taverna-users@lists.sourceforge.net; moby-
> l@biomoby.org; mobydev; taverna-
> hackers@lists.sourceforge.net
> Subject: [MOBY-dev] A bug in Taverna 1.2 perhaps related
> to BioMOBY
>
> Hi everybody,
> I don't like crossposting, but I don't know if the
> bug I have found
> today is from the Taverna 1.2 core or the MOBY plugin.
>
> The bug is pretty simple: if you create/load a
> workflow which has a
> service whose name in the workflow is different from the
> service name,
> Taverna does not use the official service name when it is
> invoked.
> Instead, it uses the name given to the service in the
> workflow! The same
> workflow works flawlessly when it is loaded and run in
> Taverna 1.1.
>
> I have discovered the bug with a MOBY service (see
> the sample workflow
> ja5.xml, and the reports taverna-1.1-report.xml and
> taverna-1.2-report.xml), so I don't know which has the
> blame...
>
> This bug should arise when one service is used more
> than once, which is
> very common in real workflows!
>
> Best Regards,
> Jos? Mar?a Fern?ndez
> --
> Jos? Mar?a Fern?ndez Gonz?lez e-mail:
> jmfernandez@cnb.uam.es
> Tlfn: (+34) 91 585 54 50 Fax: (+34) 91 585
45
> 06
> Grupo de Dise?o de Proteinas Protein Design Group
> Centro Nacional de Biotecnolog?a National Center of
> Biotechnology
> C.P.: 28049 Zip Code: 28049
> C/. Darwin n? 3 (Campus Cantoblanco, U. Aut?noma), Madrid
> (Spain)
From fgibbons at hms.harvard.edu Wed Aug 10 11:37:14 2005
From: fgibbons at hms.harvard.edu (Frank Gibbons)
Date: Wed Aug 10 11:21:30 2005
Subject: [MOBY-l] Trouble-ticketing for MOBY; unwanted side effect of
MOBY::Client::Service->execute
Message-ID: <5.2.1.1.2.20050810112754.011101f8@email.med.harvard.edu>
Hi,
Two related items:
I think MOBY::Client::Service->execute has an unwanted side-effect in that
the XMLinputlist argument appears to be modified upon returning from this
function. The code does some editing on it. I don't think that would have
been a problem had XMLinputlist been passed in as a scalar, but it's passed
as an ARRAY ref, and so modifications to it are propagated back to the
calling code. This is a problem if, like me, you want to iterate over
services calling each one with the same XMLinputlist. (At least, I think
that's what's happening - correct me.)
How do we fix something like this? Are there regression tests to make sure
nothing gets broken when this issue is tackled? Who fixes it? (I'll
volunteer) The issue of how to fix this raises the other item: while MOBY
is under CVS control, there's no ticketing system (that I'm aware of). As a
long-time lurker on this list, I know that most issues just get posted to
the list, and then fixed. I think there are a lot advantages to having a
centralized, non-email-based record of what needs to be fixed in the
future, but also of what was fixed in the past. I'm wondering partly why
MOBY isn't on SourceForge (which has a ticketing system for bugs and
feature requests, so that potential users can not only see what issues have
been fixed, but also what features are likely to appear in the future) - if
this has been discussed in the past and resolved, I'd appreciate a pointer.
Thanks,
-Frank
PhD, Computational Biologist,
Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
Tel: 617-432-3555 Fax:
617-432-3557 http://llama.med.harvard.edu/~fgibbons
From markw at illuminae.com Thu Aug 11 12:37:03 2005
From: markw at illuminae.com (mark wilkinson)
Date: Wed Aug 10 11:27:20 2005
Subject: [MOBY-l] Trouble-ticketing for MOBY;
unwanted side effect ofMOBY::Client::Service->execute
Message-ID: <969787678-1123688248-cardhu_blackberry.rim.net-5970-@engine13-cell01>
Hi Frank,
Two quick responses (I'm on the blackberry and it isn't very efficient :-) )
1). The execute method will shortly be significantly changed in light of the upcoming API changes (esp. Insistence on named parameters allows me to use a more sensible "hash" structure in place of the list structure). I don't know if it is worth spending time fixing it.
2). I didn't realize that my code modified the input list. You're right, this is bad. I could post a fix for the such that the input listref is cloned to avoid this (if that is the root of the problem), or you are of course always welcome to fix it yourself and commit.
3). We do have a ticketing system, but nobody ever uses it :-). Click the "bugs" link on the left side of the homepage... I haven't tested it for a while, but it worked last time I tried...
Cheers!
M
-----Original Message-----
From: Frank Gibbons
Date: Wed, 10 Aug 2005 11:37:14
To:moby-l@biomoby.org
Subject: [MOBY-l] Trouble-ticketing for MOBY; unwanted side effect of
MOBY::Client::Service->execute
Hi,
Two related items:
I think MOBY::Client::Service->execute has an unwanted side-effect in that
the XMLinputlist argument appears to be modified upon returning from this
function. The code does some editing on it. I don't think that would have
been a problem had XMLinputlist been passed in as a scalar, but it's passed
as an ARRAY ref, and so modifications to it are propagated back to the
calling code. This is a problem if, like me, you want to iterate over
services calling each one with the same XMLinputlist. (At least, I think
that's what's happening - correct me.)
How do we fix something like this? Are there regression tests to make sure
nothing gets broken when this issue is tackled? Who fixes it? (I'll
volunteer) The issue of how to fix this raises the other item: while MOBY
is under CVS control, there's no ticketing system (that I'm aware of). As a
long-time lurker on this list, I know that most issues just get posted to
the list, and then fixed. I think there are a lot advantages to having a
centralized, non-email-based record of what needs to be fixed in the
future, but also of what was fixed in the past. I'm wondering partly why
MOBY isn't on SourceForge (which has a ticketing system for bugs and
feature requests, so that potential users can not only see what issues have
been fixed, but also what features are likely to appear in the future) - if
this has been discussed in the past and resolved, I'd appreciate a pointer.
Thanks,
-Frank
PhD, Computational Biologist,
Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
Tel: 617-432-3555 Fax:
617-432-3557 http://llama.med.harvard.edu/~fgibbons
_______________________________________________
moby-l mailing list
moby-l@biomoby.org
http://biomoby.org/mailman/listinfo/moby-l
--
Mark Wilkinson
...on the road!
From p.lord at cs.man.ac.uk Mon Aug 15 05:35:58 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: Mon Aug 15 05:27:35 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
Message-ID: <87y873v8xt.fsf@rpc71.cs.man.ac.uk>
>>>>> "Boris" == Boris Steipe writes:
Boris> Why not put the burden of the lease on the agent to combine
Boris> the advantages of both models? I.e. if service is down for
Boris> less then a specific time, it might not get deregistered but
Boris> only flagged as temporarily unavailable ... then un-flagged
Boris> as it comes up again, except if it's down for, say > 1week,
Boris> then it gets deregistered.
Boris> $0.02
Actually, this is just a modification of the agent idea, and not a
lease at all.
The idea of a lease is not just a burden on the service provide, but
it also has a positive advantage. It give the service provide the
ability to control how long they want the service advertised for in
the registry. Of course, they can do the same thing with the agent,
more or less; they just remove the RDF document after a certain time,
although, they don't know exactly when the agent will remove them.
At the end of the day, I am not sure that the two mechanisms are
hugely different in terms of their functionality. It seems to me that
the lease puts slightly more weight on the service provider, but will
provide better garbage collection (just become someone leaves an RDF
document on the web, does not mean that their service is actually
still up).
Phil
From p.lord at cs.man.ac.uk Mon Aug 15 05:38:32 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: Mon Aug 15 05:28:48 2005
Subject: [MOBY-l] Re: [moby] Re: [MOBY-dev] lease versus agent for registry
updating
In-Reply-To: <1123885110.15335.52.camel@bioinfo.icapture.ubc.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<1123879295.15335.30.camel@bioinfo.icapture.ubc.ca>
<42FD1362.8000107@ucalgary.ca>
<1123885110.15335.52.camel@bioinfo.icapture.ubc.ca>
Message-ID: <87slxbv8tj.fsf@rpc71.cs.man.ac.uk>
>>>>> "Mark" == Mark Wilkinson writes:
Mark> We're certainly of like-mind at this end, though certain
Mark> people ;-) at myGrid think that the agent is the wrong way to
Mark> go... I just don't see it. IMO a leasing model would be akin
Mark> to leasing space in Google!
The lease is the more common idiom, I think. I don't think that there
is a huge difference. My main criticism of the agent is that there
does not seem to be a different connection between the agent poll and
the service being available.
Phil
From tmo at ebi.ac.uk Mon Aug 15 05:51:26 2005
From: tmo at ebi.ac.uk (Tom Oinn)
Date: Mon Aug 15 05:38:44 2005
Subject: [MOBY-l] Re: [moby] Re: [MOBY-dev] lease versus agent for registry
updating
In-Reply-To: <87slxbv8tj.fsf@rpc71.cs.man.ac.uk>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca> <1123879295.15335.30.camel@bioinfo.icapture.ubc.ca> <42FD1362.8000107@ucalgary.ca> <1123885110.15335.52.camel@bioinfo.icapture.ubc.ca>
<87slxbv8tj.fsf@rpc71.cs.man.ac.uk>
Message-ID: <4300659E.3050808@ebi.ac.uk>
Phillip Lord wrote:
>>>>>>"Mark" == Mark Wilkinson writes:
>
>
> Mark> We're certainly of like-mind at this end, though certain
> Mark> people ;-) at myGrid think that the agent is the wrong way to
> Mark> go... I just don't see it. IMO a leasing model would be akin
> Mark> to leasing space in Google!
>
> The lease is the more common idiom, I think. I don't think that there
> is a huge difference. My main criticism of the agent is that there
> does not seem to be a different connection between the agent poll and
> the service being available.
Much as I hate* to agree with Phil I, um, do :) Seems to me that the
agent approach is more part of a monitoring framework than a registry, I
don't know offhand whether we're planning to include metrics such as
service liveness and response times as transient metadata in the
registry but that's where I'd see agents fitting in rather than in the
initial registration.
Tom
* it's all an act but hey...
From p.lord at cs.man.ac.uk Mon Aug 15 06:00:53 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: Mon Aug 15 05:50:35 2005
Subject: [MOBY-l] Re: [moby] Re: [MOBY-dev] lease versus agent for
registry updating
In-Reply-To: <4300659E.3050808@ebi.ac.uk>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<1123879295.15335.30.camel@bioinfo.icapture.ubc.ca>
<42FD1362.8000107@ucalgary.ca>
<1123885110.15335.52.camel@bioinfo.icapture.ubc.ca>
<87slxbv8tj.fsf@rpc71.cs.man.ac.uk> <4300659E.3050808@ebi.ac.uk>
Message-ID: <87k6inv7sa.fsf@rpc71.cs.man.ac.uk>
>>>>> "Tom" == Tom Oinn writes:
Tom> Phillip Lord wrote:
>>>>>>> "Mark" == Mark Wilkinson writes:
Mark> We're certainly of like-mind at this end, though certain
Mark> people ;-) at myGrid think that the agent is the wrong way to
Mark> go... I just don't see it. IMO a leasing model would be akin
Mark> to leasing space in Google!
>> The lease is the more common idiom, I think. I don't think that
>> there is a huge difference. My main criticism of the agent is
>> that there does not seem to be a different connection between the
>> agent poll and the service being available.
Tom> Much as I hate* to agree with Phil I, um, do :) Seems to me
Tom> that the agent approach is more part of a monitoring framework
It might be if it monitored the service rather than something else!
Tom> than a registry, I don't know offhand whether we're planning to
Tom> include metrics such as service liveness and response times as
Tom> transient metadata in the registry but that's where I'd see
Tom> agents fitting in rather than in the initial registration.
The current approach that feta takes to the "registry" is that it is a
very dumb thing. It really only uses it for storing files (which is
why it uses a webdav backend by default).
It would be lovely to replace or augment this with something
richer, which understood liveness and response. Or, alternatively,
have a modular search system which could combine responses from feta
and some arbitrary service search system based on information about
liveness.
It's all very feasible. Even a simple minded response time data model
would help a lot.
Phil
From gordonp at ucalgary.ca Mon Aug 15 10:32:26 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Mon Aug 15 10:23:51 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <87y873v8xt.fsf@rpc71.cs.man.ac.uk>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk>
Message-ID: <4300A77A.8030806@ucalgary.ca>
A few advantages of the lease are not really advantages in practice:
1. No one will manually update their lease, they will put it in a cron
job. Therefore you either need to edit your crontab, or remove from the
RDF file in the agent case. Both are just as likely to be ignored by an
administrator when a service stops working.
2. Specifiying that the service is valid for a certain amount of time is
mostly useful for testing purposes. You should really be using a test
registry in that case, not posting dubious services to the production
server.
The chief advantages of the agent are:
1. You can trace the registration and deregistration of a service to a
particular domain name. It's not great security, but at the very least
people require some serious work to pose as the NCBI on purpose (by
hacking their Web domain), and cannot by mistake (e.g. "I registered my
service using the NCBI authority ID because I'm using gi's").
2. The RDF for the services does not have a single point of failure
(i.e. the central registry).
The one agent feature I would l;ike though is that I can call MOBY
Central to tell it that I've changed my RDF, i.e. pushing a
refresh. It's not critical though. If the agent runs once a day, you may
get some latency on bad services, but it's not the end of the world.
Tim Berners-Lee got a lot of flack for his Web idea because it didn't
enforce that what people linked to existed. "People won't use it, they
may end up at dead links!" they said...
>>>>>>"Boris" == Boris Steipe writes:
>>>>>>
>>>>>>
>
> Boris> Why not put the burden of the lease on the agent to combine
> Boris> the advantages of both models? I.e. if service is down for
> Boris> less then a specific time, it might not get deregistered but
> Boris> only flagged as temporarily unavailable ... then un-flagged
> Boris> as it comes up again, except if it's down for, say > 1week,
> Boris> then it gets deregistered.
>
> Boris> $0.02
>
>
>Actually, this is just a modification of the agent idea, and not a
>lease at all.
>
>The idea of a lease is not just a burden on the service provide, but
>it also has a positive advantage. It give the service provide the
>ability to control how long they want the service advertised for in
>the registry. Of course, they can do the same thing with the agent,
>more or less; they just remove the RDF document after a certain time,
>although, they don't know exactly when the agent will remove them.
>
>At the end of the day, I am not sure that the two mechanisms are
>hugely different in terms of their functionality. It seems to me that
>the lease puts slightly more weight on the service provider, but will
>provide better garbage collection (just become someone leaves an RDF
>document on the web, does not mean that their service is actually
>still up).
>
>Phil
>
>_______________________________________________
>moby-l mailing list
>moby-l@biomoby.org
>http://biomoby.org/mailman/listinfo/moby-l
>
>
>
From jmfernandez at cnb.uam.es Fri Aug 5 15:10:00 2005
From: jmfernandez at cnb.uam.es (=?ISO-8859-1?Q?Jos=E9_Mar=EDa_Fern=E1ndez_Gonz=E1lez?=)
Date: Mon Aug 15 11:29:29 2005
Subject: [MOBY-l] A bug in Taverna 1.2 perhaps related to BioMOBY
Message-ID: <42F3B988.7020506@cnb.uam.es>
Hi everybody,
I don't like crossposting, but I don't know if the bug I have found
today is from the Taverna 1.2 core or the MOBY plugin.
The bug is pretty simple: if you create/load a workflow which has a
service whose name in the workflow is different from the service name,
Taverna does not use the official service name when it is invoked.
Instead, it uses the name given to the service in the workflow! The same
workflow works flawlessly when it is loaded and run in Taverna 1.1.
I have discovered the bug with a MOBY service (see the sample workflow
ja5.xml, and the reports taverna-1.1-report.xml and
taverna-1.2-report.xml), so I don't know which has the blame...
This bug should arise when one service is used more than once, which is
very common in real workflows!
Best Regards,
Jos? Mar?a Fern?ndez
--
Jos? Mar?a Fern?ndez Gonz?lez e-mail: jmfernandez@cnb.uam.es
Tlfn: (+34) 91 585 54 50 Fax: (+34) 91 585 45 06
Grupo de Dise?o de Proteinas Protein Design Group
Centro Nacional de Biotecnolog?a National Center of Biotechnology
C.P.: 28049 Zip Code: 28049
C/. Darwin n? 3 (Campus Cantoblanco, U. Aut?noma), Madrid (Spain)
From p.lord at cs.man.ac.uk Tue Aug 16 06:02:52 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: Tue Aug 16 05:52:33 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <4300A77A.8030806@ucalgary.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk> <4300A77A.8030806@ucalgary.ca>
Message-ID: <87mzni8aib.fsf@rpc71.cs.man.ac.uk>
>>>>> "Paul" == Paul Gordon writes:
Paul> A few advantages of the lease are not really advantages in
Paul> practice:
Paul> 1. No one will manually update their lease, they will put it
Paul> in a cron
Paul> job. Therefore you either need to edit your crontab, or
Paul> remove from the RDF file in the agent case. Both are just as
Paul> likely to be ignored by an administrator when a service stops
Paul> working.
This is a necessary problem with neither the agent nor the lease. It
is a current problem with the implementation of the agent. All you do
is add a call back to the service.
Paul> 2. Specifiying that the service is valid for a certain amount
Paul> of time
Paul> is mostly useful for testing purposes. You should really be
Paul> using a test registry in that case, not posting dubious
Paul> services to the production server.
Not really. Take, for example, the SOAPLAB services Martin developed
as part of mygrid. When the toolkit was first produced, it was
deployed on our own servers; now it is supported by EBI's external
services. Clearly, the latter is likely to produce a better uptime
than the former. Hence, EBI might use a month long lease.
Neither of these services were testing. But they did come with
different levels of support.
Paul> The chief advantages of the agent are:
Paul> 1. You can trace the registration and deregistration of a
Paul> service to
Paul> a particular domain name. It's not great security, but at the
Paul> very least people require some serious work to pose as the
Paul> NCBI on purpose (by hacking their Web domain), and cannot by
Paul> mistake (e.g. "I registered my service using the NCBI
Paul> authority ID because I'm using gi's").
Likewise with a lease call back system. Anyway, can you really not
determine the start point of a web services call?
Paul> 2. The RDF for the services does not have a single point of
Paul> failure
Paul> (i.e. the central registry).
This is also untrue. There is nothing to stop a lease percolating
through a set of federated registries. With the agent, you have to
percolate the registered URL's (or the RDF) in the same way.
Ultimately, as two systems are doing similar things. It's just that
one is push and the other pull.
Paul> The one agent feature I would l;ike though is that I can call
Paul> MOBY Central to tell it that I've changed my RDF, i.e. pushing
Paul> a refresh. It's not critical though. If the agent runs once a
Paul> day, you may get some latency on bad services, but it's not
Paul> the end of the world. Tim Berners-Lee got a lot of flack for
Paul> his Web idea because it didn't enforce that what people linked
Paul> to existed. "People won't use it, they may end up at dead
Paul> links!" they said...
You still want people to be able to search on the freshness of
information though. With a lease, you can add queries to moby-central
to say "give me only services with a current lease"--after all the
registry is not required to deregister a service when it's lease runs
out.
Cheers
Phil
From gordonp at ucalgary.ca Tue Aug 16 09:51:08 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Tue Aug 16 09:53:12 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <87mzni8aib.fsf@rpc71.cs.man.ac.uk>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca> <87y873v8xt.fsf@rpc71.cs.man.ac.uk>
<4300A77A.8030806@ucalgary.ca> <87mzni8aib.fsf@rpc71.cs.man.ac.uk>
Message-ID: <4301EF4C.5020506@ucalgary.ca>
My CAN$0.02 :-)
>
> Paul> A few advantages of the lease are not really advantages in
> Paul> practice:
>
> Paul> 1. No one will manually update their lease, they will put it
> Paul> in a cron
> Paul> job. Therefore you either need to edit your crontab, or
> Paul> remove from the RDF file in the agent case. Both are just as
> Paul> likely to be ignored by an administrator when a service stops
> Paul> working.
>
>
>This is a necessary problem with neither the agent nor the lease. It
>is a current problem with the implementation of the agent. All you do
>is add a call back to the service.
>
>
This turns into an issue with the protocol, we would need to require
(when you register a service) test cases that return predictable,
non-null value. This was discussed at the last MOBY meeting, but I
don't think anything came of it. Or am I wrong?
> Paul> The chief advantages of the agent are:
>
> Paul> 1. You can trace the registration and deregistration of a
> Paul> service to
> Paul> a particular domain name. It's not great security, but at the
> Paul> very least people require some serious work to pose as the
> Paul> NCBI on purpose (by hacking their Web domain), and cannot by
> Paul> mistake (e.g. "I registered my service using the NCBI
> Paul> authority ID because I'm using gi's").
>
>Likewise with a lease call back system.
>
Like you said, the two approches aren't so different, because fetching
the RDF from the server is a callback of sorts.
>Anyway, can you really not
>determine the start point of a web services call?
>
>
If there are proxies, multiple host names for a machine (e.g. the host
I'm on right now is www.visualgenomics.ca, moby.ucalgary.ca,
www.gcbioinformatics.ca, and about 6 other names), it can get complicated.
> Paul> 2. The RDF for the services does not have a single point of
> Paul> failure
> Paul> (i.e. the central registry).
>
>This is also untrue. There is nothing to stop a lease percolating
>through a set of federated registries. With the agent, you have to
>percolate the registered URL's (or the RDF) in the same way.
>
>
True, but the mechanism to exchange a URL list could be simpler than
mirroring the entire registry, and is less subject to error propagation
(e.g. don't blindly copy a goof in the main mirror's database system
because it had network connectivity issues, fetch the RDF yourself) .
If we're lucky, people will try to outcompete each other creating the
most robust registry :-)
>Ultimately, as two systems are doing similar things. It's just that
>one is push and the other pull.
>
>
> Paul> The one agent feature I would l;ike though is that I can call
> Paul> MOBY Central to tell it that I've changed my RDF, i.e. pushing
> Paul> a refresh. It's not critical though. If the agent runs once a
> Paul> day, you may get some latency on bad services, but it's not
> Paul> the end of the world. Tim Berners-Lee got a lot of flack for
> Paul> his Web idea because it didn't enforce that what people linked
> Paul> to existed. "People won't use it, they may end up at dead
> Paul> links!" they said...
>
>
>You still want people to be able to search on the freshness of
>information though. With a lease, you can add queries to moby-central
>to say "give me only services with a current lease"--after all the
>registry is not required to deregister a service when it's lease runs
>out.
>
>
Would we set a maximum lease? I'd probably give myself a 1 year lease
so that I wouldn't be bothered again. That doesn't make me check that
the service works throughout the year though. IMHO, having a test case
checked regularly is what really helps here, not the lease. A lease is
someone's word, a test case is action, and actions speak louder than
words :-)
>Cheers
>
>Phil
>
>_______________________________________________
>moby-l mailing list
>moby-l@biomoby.org
>http://biomoby.org/mailman/listinfo/moby-l
>
>
>
From edward.kawas at gmail.com Tue Aug 16 12:26:42 2005
From: edward.kawas at gmail.com (Eddie Kawas)
Date: Tue Aug 16 12:16:35 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <4301EF4C.5020506@ucalgary.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk> <4300A77A.8030806@ucalgary.ca>
<87mzni8aib.fsf@rpc71.cs.man.ac.uk> <4301EF4C.5020506@ucalgary.ca>
Message-ID:
I just wanted to say that I agree with Paul.
I think that when a provider registers a service, the provider enters
a url that causes the service to return something that is commonly
returned by all services that states that the service is running. If
this return value is infact returned, then the service signature is
processed. Otherwise, the offending service is recorded and the
service is not shown to the public until the test passes. If after x
times the test fails, then it is permanently removed.
Eddie
On 8/16/05, Paul Gordon wrote:
> My CAN$0.02 :-)
>
> >
> > Paul> A few advantages of the lease are not really advantages in
> > Paul> practice:
> >
> > Paul> 1. No one will manually update their lease, they will put it
> > Paul> in a cron
> > Paul> job. Therefore you either need to edit your crontab, or
> > Paul> remove from the RDF file in the agent case. Both are just as
> > Paul> likely to be ignored by an administrator when a service stops
> > Paul> working.
> >
> >
> >This is a necessary problem with neither the agent nor the lease. It
> >is a current problem with the implementation of the agent. All you do
> >is add a call back to the service.
> >
> >
> This turns into an issue with the protocol, we would need to require
> (when you register a service) test cases that return predictable,
> non-null value. This was discussed at the last MOBY meeting, but I
> don't think anything came of it. Or am I wrong?
>
> > Paul> The chief advantages of the agent are:
> >
> > Paul> 1. You can trace the registration and deregistration of a
> > Paul> service to
> > Paul> a particular domain name. It's not great security, but at the
> > Paul> very least people require some serious work to pose as the
> > Paul> NCBI on purpose (by hacking their Web domain), and cannot by
> > Paul> mistake (e.g. "I registered my service using the NCBI
> > Paul> authority ID because I'm using gi's").
> >
> >Likewise with a lease call back system.
> >
> Like you said, the two approches aren't so different, because fetching
> the RDF from the server is a callback of sorts.
>
> >Anyway, can you really not
> >determine the start point of a web services call?
> >
> >
> If there are proxies, multiple host names for a machine (e.g. the host
> I'm on right now is www.visualgenomics.ca, moby.ucalgary.ca,
> www.gcbioinformatics.ca, and about 6 other names), it can get complicated.
>
> > Paul> 2. The RDF for the services does not have a single point of
> > Paul> failure
> > Paul> (i.e. the central registry).
> >
> >This is also untrue. There is nothing to stop a lease percolating
> >through a set of federated registries. With the agent, you have to
> >percolate the registered URL's (or the RDF) in the same way.
> >
> >
> True, but the mechanism to exchange a URL list could be simpler than
> mirroring the entire registry, and is less subject to error propagation
> (e.g. don't blindly copy a goof in the main mirror's database system
> because it had network connectivity issues, fetch the RDF yourself) .
> If we're lucky, people will try to outcompete each other creating the
> most robust registry :-)
>
> >Ultimately, as two systems are doing similar things. It's just that
> >one is push and the other pull.
> >
> >
> > Paul> The one agent feature I would l;ike though is that I can call
> > Paul> MOBY Central to tell it that I've changed my RDF, i.e. pushing
> > Paul> a refresh. It's not critical though. If the agent runs once a
> > Paul> day, you may get some latency on bad services, but it's not
> > Paul> the end of the world. Tim Berners-Lee got a lot of flack for
> > Paul> his Web idea because it didn't enforce that what people linked
> > Paul> to existed. "People won't use it, they may end up at dead
> > Paul> links!" they said...
> >
> >
> >You still want people to be able to search on the freshness of
> >information though. With a lease, you can add queries to moby-central
> >to say "give me only services with a current lease"--after all the
> >registry is not required to deregister a service when it's lease runs
> >out.
> >
> >
> Would we set a maximum lease? I'd probably give myself a 1 year lease
> so that I wouldn't be bothered again. That doesn't make me check that
> the service works throughout the year though. IMHO, having a test case
> checked regularly is what really helps here, not the lease. A lease is
> someone's word, a test case is action, and actions speak louder than
> words :-)
>
> >Cheers
> >
> >Phil
> >
> >_______________________________________________
> >moby-l mailing list
> >moby-l@biomoby.org
> >http://biomoby.org/mailman/listinfo/moby-l
> >
> >
> >
>
> _______________________________________________
> moby-l mailing list
> moby-l@biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
>
From p.lord at cs.man.ac.uk Tue Aug 16 13:15:29 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: Tue Aug 16 13:06:36 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <4301EF4C.5020506@ucalgary.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk> <4300A77A.8030806@ucalgary.ca>
<87mzni8aib.fsf@rpc71.cs.man.ac.uk> <4301EF4C.5020506@ucalgary.ca>
Message-ID: <87ek8tpzv2.fsf@rpc71.cs.man.ac.uk>
>>>>> "Paul" == Paul Gordon writes:
>> This is a necessary problem with neither the agent nor the
>> lease. It is a current problem with the implementation of the
>> agent. All you do is add a call back to the service.
Paul> This turns into an issue with the protocol, we would need to
Paul> require (when you register a service) test cases that return
Paul> predictable, non-null value. This was discussed at the last
Paul> MOBY meeting, but I don't think anything came of it. Or am I
Paul> wrong?
I can't comment on this really, because I wasn't there!
Paul> 1. You can trace the registration and deregistration of a
Paul> service to a particular domain name. It's not great security,
Paul> but at the very least people require some serious work to pose
Paul> as the NCBI on purpose (by hacking their Web domain), and
Paul> cannot by mistake (e.g. "I registered my service using the
Paul> NCBI authority ID because I'm using gi's").
>>
>> Likewise with a lease call back system.
Paul> Like you said, the two approches aren't so different, because
Paul> fetching the RDF from the server is a callback of sorts.
In a sense. It's just that you don't know when it is going to
happen. This is the point with a lease. The client requests from the
server a time limited registration. Is it possible for the client to
query MOBY-Central and ask how often the RDF agent will call? Is it
possible to request that it happens more frequently or less?
>> Anyway, can you really not determine the start point of a web
>> services call?
Paul> If there are proxies, multiple host names for a machine
Paul> (e.g. the host I'm on right now is www.visualgenomics.ca,
Paul> moby.ucalgary.ca, www.gcbioinformatics.ca, and about 6 other
Paul> names), it can get complicated.
Or 136.159.169.6 or h6.net169.ucalgary.ca. The IP number is unique
though.
Paul> 2. The RDF for the services does not have a single point of
Paul> failure (i.e. the central registry).
>>
>> This is also untrue. There is nothing to stop a lease percolating
>> through a set of federated registries. With the agent, you have
>> to percolate the registered URL's (or the RDF) in the same way.
Paul> True, but the mechanism to exchange a URL list could be
Paul> simpler than mirroring the entire registry, and is less
Paul> subject to error propagation (e.g. don't blindly copy a goof
Paul> in the main mirror's database system because it had network
Paul> connectivity issues, fetch the RDF yourself) . If we're
Paul> lucky, people will try to outcompete each other creating the
Paul> most robust registry :-)
Who says that the lease is not encoded as a URL? Besides which, the
lease mechanism might federate by passing around data. Or the client
might explicitly register in more than one registry. Again, I don't
think that this is a necessary advantage of either a lease or the
agent.
>> You still want people to be able to search on the freshness of
>> information though. With a lease, you can add queries to
>> moby-central to say "give me only services with a current
>> lease"--after all the registry is not required to deregister a
>> service when it's lease runs out.
Paul> Would we set a maximum lease? I'd probably give myself a 1
Paul> year lease so that I wouldn't be bothered again.
Yes, this is part of the basic idea behind a lease. The registry overs
a specific contract which describes what it is prepared to do for
you.
Paul> That doesn't make me check that the service works throughout
Paul> the year though. IMHO, having a test case checked regularly is
Paul> what really helps here, not the lease. A lease is someone's
Paul> word, a test case is action, and actions speak louder than
Paul> words :-)
As Tom says, service liveness or quality data is not the same thing as
registration, but an orthogonal issue to it. Having service quality
data would be a nice addition to moby-central.
Phil
From goodb at interchange.ubc.ca Tue Aug 16 13:18:47 2005
From: goodb at interchange.ubc.ca (Benjamin Good)
Date: Tue Aug 16 13:16:51 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To:
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk> <4300A77A.8030806@ucalgary.ca>
<87mzni8aib.fsf@rpc71.cs.man.ac.uk> <4301EF4C.5020506@ucalgary.ca>
Message-ID: <4784EB52-25CB-4260-A779-72918248ECB4@interchange.ubc.ca>
Just to add my 2c. The idea of adding a test case to the
registration phase did come up at the meeting, but I'm not sure how
far its gotten?
I would suggest that this test actually be an instance of the moby
datatype registered as the input of the service - instead of a URL.
This would let any moby-capable client test the service whenever
necessary using exactly the same protocol required to run the service
in the usual case. Having this possibility makes it feasible for
client providers to filter out inconsistent services without relying
entirely on the registry to do so. This enables a runtime check on
the service that seems impossible using only the registry. The agent
could then use this same check to make its decision about whether to
keep the service listed or not..
- Ben
On Aug 16, 2005, at 9:26 AM, Eddie Kawas wrote:
> I just wanted to say that I agree with Paul.
>
> I think that when a provider registers a service, the provider enters
> a url that causes the service to return something that is commonly
> returned by all services that states that the service is running. If
> this return value is infact returned, then the service signature is
> processed. Otherwise, the offending service is recorded and the
> service is not shown to the public until the test passes. If after x
> times the test fails, then it is permanently removed.
>
> Eddie
>
>
> On 8/16/05, Paul Gordon wrote:
>
>> My CAN$0.02 :-)
>>
>>
>>>
>>> Paul> A few advantages of the lease are not really advantages in
>>> Paul> practice:
>>>
>>> Paul> 1. No one will manually update their lease, they will put it
>>> Paul> in a cron
>>> Paul> job. Therefore you either need to edit your crontab, or
>>> Paul> remove from the RDF file in the agent case. Both are just as
>>> Paul> likely to be ignored by an administrator when a service stops
>>> Paul> working.
>>>
>>>
>>> This is a necessary problem with neither the agent nor the lease. It
>>> is a current problem with the implementation of the agent. All
>>> you do
>>> is add a call back to the service.
>>>
>>>
>>>
>> This turns into an issue with the protocol, we would need to require
>> (when you register a service) test cases that return predictable,
>> non-null value. This was discussed at the last MOBY meeting, but I
>> don't think anything came of it. Or am I wrong?
>>
>>
>>> Paul> The chief advantages of the agent are:
>>>
>>> Paul> 1. You can trace the registration and deregistration of a
>>> Paul> service to
>>> Paul> a particular domain name. It's not great security, but at
>>> the
>>> Paul> very least people require some serious work to pose as the
>>> Paul> NCBI on purpose (by hacking their Web domain), and cannot by
>>> Paul> mistake (e.g. "I registered my service using the NCBI
>>> Paul> authority ID because I'm using gi's").
>>>
>>> Likewise with a lease call back system.
>>>
>>>
>> Like you said, the two approches aren't so different, because
>> fetching
>> the RDF from the server is a callback of sorts.
>>
>>
>>> Anyway, can you really not
>>> determine the start point of a web services call?
>>>
>>>
>>>
>> If there are proxies, multiple host names for a machine (e.g. the
>> host
>> I'm on right now is www.visualgenomics.ca, moby.ucalgary.ca,
>> www.gcbioinformatics.ca, and about 6 other names), it can get
>> complicated.
>>
>>
>>> Paul> 2. The RDF for the services does not have a single point of
>>> Paul> failure
>>> Paul> (i.e. the central registry).
>>>
>>> This is also untrue. There is nothing to stop a lease percolating
>>> through a set of federated registries. With the agent, you have to
>>> percolate the registered URL's (or the RDF) in the same way.
>>>
>>>
>>>
>> True, but the mechanism to exchange a URL list could be simpler than
>> mirroring the entire registry, and is less subject to error
>> propagation
>> (e.g. don't blindly copy a goof in the main mirror's database system
>> because it had network connectivity issues, fetch the RDF yourself) .
>> If we're lucky, people will try to outcompete each other creating the
>> most robust registry :-)
>>
>>
>>> Ultimately, as two systems are doing similar things. It's just that
>>> one is push and the other pull.
>>>
>>>
>>> Paul> The one agent feature I would l;ike though is that I can call
>>> Paul> MOBY Central to tell it that I've changed my RDF, i.e.
>>> pushing
>>> Paul> a refresh. It's not critical though. If the agent runs once a
>>> Paul> day, you may get some latency on bad services, but it's not
>>> Paul> the end of the world. Tim Berners-Lee got a lot of flack for
>>> Paul> his Web idea because it didn't enforce that what people
>>> linked
>>> Paul> to existed. "People won't use it, they may end up at dead
>>> Paul> links!" they said...
>>>
>>>
>>> You still want people to be able to search on the freshness of
>>> information though. With a lease, you can add queries to moby-
>>> central
>>> to say "give me only services with a current lease"--after all the
>>> registry is not required to deregister a service when it's lease
>>> runs
>>> out.
>>>
>>>
>>>
>> Would we set a maximum lease? I'd probably give myself a 1 year
>> lease
>> so that I wouldn't be bothered again. That doesn't make me check
>> that
>> the service works throughout the year though. IMHO, having a test
>> case
>> checked regularly is what really helps here, not the lease. A
>> lease is
>> someone's word, a test case is action, and actions speak louder than
>> words :-)
>>
>>
>>> Cheers
>>>
>>> Phil
>>>
>>> _______________________________________________
>>> moby-l mailing list
>>> moby-l@biomoby.org
>>> http://biomoby.org/mailman/listinfo/moby-l
>>>
>>>
>>>
>>>
>>
>> _______________________________________________
>> moby-l mailing list
>> moby-l@biomoby.org
>> http://biomoby.org/mailman/listinfo/moby-l
>>
>>
>
> _______________________________________________
> moby-l mailing list
> moby-l@biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
>
>
From markw at illuminae.com Tue Aug 16 13:36:25 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Tue Aug 16 13:45:38 2005
Subject: [moby] Re: [MOBY-l] Re: [MOBY-dev] lease versus agent for
registry updating
In-Reply-To: <4784EB52-25CB-4260-A779-72918248ECB4@interchange.ubc.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk> <4300A77A.8030806@ucalgary.ca>
<87mzni8aib.fsf@rpc71.cs.man.ac.uk> <4301EF4C.5020506@ucalgary.ca>
<4784EB52-25CB-4260-A779-72918248ECB4@interchange.ubc.ca>
Message-ID: <1124213785.18621.138.camel@bioinfo.icapture.ubc.ca>
On Tue, 2005-08-16 at 10:18 -0700, Benjamin Good wrote:
> Just to add my 2c. The idea of adding a test case to the
> registration phase did come up at the meeting, but I'm not sure how
> far its gotten?
It was decided that test cases would be provided as optional additional
metadata attached to the RDF Service Signature on the service-providers
site, identified by a set of (yet to be defined) RDF predicates. Test
cases would *not* be part of the data stored in the registry, but would
be discoverable through e.g. LSID resolution.
Since we are currently in the process of adapting our Service Signature
data model to mimic that of myGrid (and v.v. :-) ) we have not yet
implemented this particular decision. Hopefully this data model will be
firmed-up in the next couple of weeks (that was why Eddie and I were in
Manchester last week), and then we will be able to generate a
"canonical" Service Signature for everyone's existing services, as well
as generate the signature on-the-fly for new registrations.
M
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From senger at ebi.ac.uk Tue Aug 16 20:10:24 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Tue Aug 16 20:01:13 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <4784EB52-25CB-4260-A779-72918248ECB4@interchange.ubc.ca>
Message-ID:
Hi all,
Everybody is talking about lease and agents so if I do not join I would
feel like a pariah...
What was the original problem that started this discussion? I remember
two:
a) How to make a deregistration secure (so nobody can easily remove my
service from a registry), and
b) How I can remove my service registration at any time when I *want*
to do it.
Am I right - is this (at least) the problems we are trying to solve? (I
understand that during this discussion a new problem was born - how to
test that a service is reliable; but I consider it as a separate issue.)
Obviously problem b) requires an action on my side.
In case of agent it is a removal of an RDF document - and hopefully
supported by a new method in the registry API calling an agent to come to
me *now*. This new API call was mentioned and I hope it will happen.
Without it the agent solution is weak (IMHO) - having the similar weakness
as Google has with non existing links/documents (oh, wow, now I really
feel like a pariah by saying that Google has a weakness!)
In case of lease - is there a concept of making a lease to expire
before its planned end? If so, how is it done that nobody else can do it
except me (problem a))?
Can you, Phill perhaps, explain me if the lease can solve my two
problems above? So far I understand that an agent can - assuming tghat
there will be that new API call (Mark, Eddie, will this be - in case the
agent idea is accepted?)
Cheers,
Martin
--
Martin Senger
email: martin.senger@gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From p.lord at cs.man.ac.uk Wed Aug 17 06:23:18 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: Wed Aug 17 06:18:31 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To:
References:
Message-ID: <87vf246ew9.fsf@rpc71.cs.man.ac.uk>
>>>>> "Martin" == Martin Senger writes:
Martin> Hi all,
Martin> Everybody is talking about lease and agents so if I do
Martin> not join I would
Martin> feel like a pariah...
Martin> What was the original problem that started this
Martin> discussion? I remember
Martin> two:
Martin> a) How to make a deregistration secure (so nobody can easily
Martin> remove my
Martin> service from a registry), and
Martin> b) How I can remove my service registration at any time when
Martin> I *want*
Martin> to do it.
Martin> Am I right - is this (at least) the problems we are
Martin> trying to solve? (I
Martin> understand that during this discussion a new problem was
Martin> born - how to test that a service is reliable; but I
Martin> consider it as a separate issue.)
Martin> Obviously problem b) requires an action on my side. In
Martin> case of agent it is a removal of an RDF document - and
Martin> hopefully
Martin> supported by a new method in the registry API calling an
Martin> agent to come to me *now*. This new API call was mentioned
Martin> and I hope it will happen.
Actually, there is a minor security problem here. If I call this
method repeatedly and it really happens *now*, then I can use
moby-central as a proxy to launch a denial of service attack on a
third party. If I did this on NCBI, for example, it might have
interesting repercussions for who ever is hosting the moby-central.
Martin> Without it the agent solution is weak (IMHO) - having the
Martin> similar weakness as Google has with non existing
Martin> links/documents (oh, wow, now I really feel like a pariah by
Martin> saying that Google has a weakness!)
Martin> In case of lease - is there a concept of making a lease
Martin> to expire
Martin> before its planned end? If so, how is it done that nobody
Martin> else can do it except me (problem a))?
Not really sure on this one, to be honest. Clearly, you can just set
up a short lease in the first place. Alternatively, you pass back some
magic cookie when the lease is taken out in the first place, that is
required for deregistration. This enables deregistration to be done
from anywhere, rather than through direct control of the RDF URL, but
still only by the originator.
The ideal would, of course, to have some off the shelf library for
doing leases.
Martin> Can you, Phill perhaps, explain me if the lease can solve
Martin> my two
Martin> problems above? So far I understand that an agent can -
Martin> assuming tghat there will be that new API call (Mark, Eddie,
Martin> will this be - in case the agent idea is accepted?)
Does this make sense?
Phil
From senger at ebi.ac.uk Wed Aug 17 10:57:47 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Wed Aug 17 10:49:15 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <87vf246ew9.fsf@rpc71.cs.man.ac.uk>
Message-ID:
> Actually, there is a minor security problem here. If I call this
> method repeatedly and it really happens *now*
>
You are exaggerating (how to spell this word?). The agent can be clever
enough not to go really *now" but only in a reasonably but short
intervals (once a minute maximum, for example).
> Not really sure on this one, to be honest. Clearly, you can just set
> up a short lease in the first place.
>
Yes, that make sense. Or perhaps, I could re-register again with a
different lease, right?
> Alternatively, you pass back some magic cookie when the lease is taken
> out in the first place, that is required for deregistration.
>
Well, this is exactly what biomoby has now - and nobody seems to like
it. That's why we are talking how to change it.
Thanks, Phill, for claryfying the issue (at least in my mind). I feel
now that a lease gives me more control (if and only if I can re-register
with a different lease anytime). And an agent plays a role of an
independent observer that can stop my registration if it finds that there
is something wrong with it (not running, not found RDF etc.).
But the problem is still with the "if and only if" in the previous
paragraph I think. If I can re-register and change the lease, probably
anybody can do it - and to remove/hijack my registration. And we are back
where we started....
So the above would work only if I can prove who I am. And the only way
how to do it (if we do not want to introduce passwords etc.) is to let
somebody come back to my place (because only I can control such a
place). That means an agent...
So: if I use a lease I am in charge of registering my services. If it
is a new registration (such service does not exist yet), everything is
fine and nobody needs to go to me (but can, and probably will, but there
is no hurry). If I try to re-register (because I wish to change the lease
period), then an agent must come to me to see that my RDF document is
still there (with a new lease perion inside) - and it must come soon,
because it cannot confirm my re-registration.
This could work... (not only lease, not only agent, but both)
Cheers,
Martin
--
Martin Senger
email: martin.senger@gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From gordonp at ucalgary.ca Wed Aug 17 12:16:45 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Wed Aug 17 12:06:35 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To:
References:
Message-ID: <430362ED.1090804@ucalgary.ca>
>>Alternatively, you pass back some magic cookie when the lease is taken
>>out in the first place, that is required for deregistration.
>>
>>
>>
> Well, this is exactly what biomoby has now - and nobody seems to like
>it. That's why we are talking how to change it.
>
>
Absolutely. Keeping a cookie around somewhere for a month, then
manually using it to renew the lease (remember, automating the lease
renewal is just as bad as having an RDF file that you need to change) is
bad. If I lose my cookie, or it is intercepted, I'm toast.
> So the above would work only if I can prove who I am. And the only way
>how to do it (if we do not want to introduce passwords etc.) is to let
>somebody come back to my place (because only I can control such a
>place). That means an agent...
>
>
Exactly my point from before about multiple host names/machine. You
should only be able to advertise a service available in MOBY from the
ucalgary.ca domain if you are able to put an RDF document on a web
server in that domain. It's a security check of sorts. I don't see how
a lease-only system could easily acheive this without some signing
certificates, and in fact I could prevent the NCBI from registering
services by doing it first and getting the cookie they need to fix the
registry. If we can add a "push" like method to the registry saying
"check my RDF ASAP", and maybe an optional expiry for people who like
leases (the default would be turf the service if the test case fails
repeatedly) we should have the best of both worlds...
> So: if I use a lease I am in charge of registering my services. If it
>is a new registration (such service does not exist yet), everything is
>fine and nobody needs to go to me (but can, and probably will, but there
>is no hurry). If I try to re-register (because I wish to change the lease
>period), then an agent must come to me to see that my RDF document is
>still there (with a new lease perion inside) - and it must come soon,
>because it cannot confirm my re-registration.
>
> This could work... (not only lease, not only agent, but both)
>
> Cheers,
> Martin
>
>--
>Martin Senger
> email: martin.senger@gmail.com
> skype: martinsenger
>International Rice Research Institute
>Biometrics and Bioinformatics Unit
>DAPO BOX 7777, Metro Manila
>Philippines, phone: +63-2-580-5600 (ext.2324)
>
>_______________________________________________
>moby-l mailing list
>moby-l@biomoby.org
>http://biomoby.org/mailman/listinfo/moby-l
>
>
>
From markw at illuminae.com Wed Aug 17 12:23:50 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Wed Aug 17 12:14:39 2005
Subject: [moby] Re: [MOBY-l] Re: [MOBY-dev] lease versus agent for
registry updating
In-Reply-To: <430362ED.1090804@ucalgary.ca>
References:
<430362ED.1090804@ucalgary.ca>
Message-ID: <1124295830.21293.24.camel@bioinfo.icapture.ubc.ca>
On Wed, 2005-08-17 at 10:16 -0600, Paul Gordon wrote:
> Keeping a cookie around somewhere for a month, then
> manually using it to renew the lease (remember, automating the lease
> renewal is just as bad as having an RDF file that you need to change) is
> bad. If I lose my cookie, or it is intercepted, I'm toast.
Yes... in fact, that solution makes me lose my cookies! ;-)
> If we can add a "push" like method to the registry saying
> "check my RDF ASAP", and maybe an optional expiry for people who like
> leases (the default would be turf the service if the test case fails
> repeatedly) we should have the best of both worlds...
This is easy enough to do given that we are moving to RDF as our Service
Signature - it just becomes another metadata tag. However, we would
then have to add some code to the registry that stores lease times, and
this then makes some assumptions about what the underling registry has
to know about. Given that we are now trying to support both a MOBY
Central and a myGrid/Feta underlying store we cannot make "arbitrary"
decisions like this anymore... A non-lease mechanism does not require
the registry to "think", only to store the data that it needs and
respond to "come hither agent" calls.
M
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From fgibbons at hms.harvard.edu Thu Aug 18 18:13:02 2005
From: fgibbons at hms.harvard.edu (Frank Gibbons)
Date: Thu Aug 18 18:03:11 2005
Subject: [MOBY-l] Where does articleName belong? How can I set it in a
Simple, using jMoby?
Message-ID: <5.2.1.1.2.20050818172744.011d8e00@email.med.harvard.edu>
Hi,
I'm having trouble understanding where articleName belongs. I thought I
understood it before, but now that I'm trying to use jMoby to access
services that I'm providing from Perl, I'm confused again..... (But I note
from the MOBY digest that just landed in my mailbox that that's OK, since
articleName has several uses/meanings, and may perhaps soon acquire some
more, so perhaps I'm not the only one ;)
My (Perl) services expect input that looks like this:
In jMoby, from reading TestRequest.java, I can see that it's easy to build
an object to contain a message like, using one of the constructors (it
takes two String arguments: namespace and id). But how do I set the
articleName for the enclosing tag? Calling setName() on the
service input MobyDataObject that wraps it, results in that articleName
being handed down to the child element (Object), when it should remain with
the Simple.
Maybe I've got it all wrong. This has worked for me with the Perl API for
over a year, but when trying to interact with those services using jMoby,
it barfs because the response (which looks not unlike the input above)
contains Simples with children, which "is NOT expected" [sic]. But I though
Simples could contain fundamental datatypes like String & friends. Perhaps
this relates to the discussion over the last weeks about what a Collection
is for.
Hopefully, I've overlooked something in the javadocs - if so, I'd
appreciate a pointer to the right class.
Thanks in advance.
-Frank
PhD, Computational Biologist,
Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
Tel: 617-432-3555 Fax:
617-432-3557 http://llama.med.harvard.edu/~fgibbons
From senger at ebi.ac.uk Thu Aug 18 18:59:09 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Thu Aug 18 18:49:04 2005
Subject: [MOBY-l] Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To: <5.2.1.1.2.20050818172744.011d8e00@email.med.harvard.edu>
Message-ID:
Frank,
I guess you will get an reply from Paul explaining your answer.
I want just make a small note: Bear with me for another week (probably
less, just until Monday or Tuesday) when I am going to release a completly
new support/additions to jMoby that surely deals with article names in
both their incarnations (article names in Simples and article names in
members objects (HAS/HASA objects).
Cheers,
Martin
--
Martin Senger
email: martin.senger@gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From gordonp at ucalgary.ca Fri Aug 19 11:21:38 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Fri Aug 19 11:11:41 2005
Subject: [MOBY-l] Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To: <5.2.1.1.2.20050818172744.011d8e00@email.med.harvard.edu>
References: <5.2.1.1.2.20050818172744.011d8e00@email.med.harvard.edu>
Message-ID: <4305F902.7060506@ucalgary.ca>
> Hi,
>
> I'm having trouble understanding where articleName belongs. I thought
> I understood it before, but now that I'm trying to use jMoby to access
> services that I'm providing from Perl, I'm confused again..... (But I
> note from the MOBY digest that just landed in my mailbox that that's
> OK, since articleName has several uses/meanings, and may perhaps soon
> acquire some more, so perhaps I'm not the only one ;)
>
> My (Perl) services expect input that looks like this:
>
>
>
>
>
> In jMoby, from reading TestRequest.java, I can see that it's easy to
> build an object to contain a message like, using one of the
> constructors (it takes two String arguments: namespace and id). But
> how do I set the articleName for the enclosing tag? Calling
> setName() on the service input MobyDataObject that wraps it, results
> in that articleName being handed down to the child element (Object),
> when it should remain with the Simple.
To me it makes sense: calling object.setName() sets the article name in
the Object tag, not the Simple.
The Simple is not a data object of any sort, it is just a wrapper used
to name parameters in calls to services.
I can reuse the Object in another service, with a completely different
parameter name, or even use it as part of a collection. I should not
have to modify the Object name itself to make these calls, as it is
exactly the same Object tag as before. In fact, I never use
object.setName(), it's there because there are other parts of the code
that call it, when you are building composite objects and the
articleName is needed to identify the data member. I'll add a message
to the javadoc stating this.
Because most people are just using one parameter inputs, it's easy to
falsely make the connection object name == simple name.
Was there not some dicussion on the list of changing articleName to
paramName for Simples? If not, we should defiunitely do this!
In any case, the jMOBY code right now doesn't deal nicely with named
parameters. I was striuggling before with how to deal with XML that
both named and unnamed parameters, very confusing. But now that all
parameters have to be named (since API 0.86) its much easier. I'll fix
the code on the weekend. Thanks for pointiung this out.
> Maybe I've got it all wrong. This has worked for me with the Perl API
> for over a year, but when trying to interact with those services using
> jMoby, it barfs because the response (which looks not unlike the input
> above) contains Simples with children, which "is NOT expected" [sic].
> But I though Simples could contain fundamental datatypes like String &
> friends. Perhaps this relates to the discussion over the last weeks
> about what a Collection is for.
Can you send me your exact XML?
> Hopefully, I've overlooked something in the javadocs - if so, I'd
> appreciate a pointer to the right class.
>
> Thanks in advance.
>
> -Frank
>
> PhD, Computational Biologist,
> Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA
> 02115, USA.
> Tel: 617-432-3555 Fax: 617-432-3557
> http://llama.med.harvard.edu/~fgibbons
> _______________________________________________
> moby-l mailing list
> moby-l@biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
>
From gordonp at ucalgary.ca Fri Aug 19 12:06:11 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Fri Aug 19 11:55:44 2005
Subject: [MOBY-l] Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To: <4305F902.7060506@ucalgary.ca>
References: <5.2.1.1.2.20050818172744.011d8e00@email.med.harvard.edu>
<4305F902.7060506@ucalgary.ca>
Message-ID: <43060373.1080304@ucalgary.ca>
Replying to my own :-)
>
> In any case, the jMOBY code right now doesn't deal nicely with named
> parameters. I was striuggling before with how to deal with XML that
> both named and unnamed parameters, very confusing. But now that all
> parameters have to be named (since API 0.86) its much easier. I'll
> fix the code on the weekend. Thanks for pointiung this out.
>
Hmm. A corrolary of requiring all parameters to be named is that there
the returned data should be named too, as MOBY input and output messages
are essentially symmetric. We could get around this by saying that if
you have only one parameter, it does not need to be named. In that way,
the response simple or collection does not need to be named either.
This also makes the vast majority of currently registered services valid
again according to API v.0.86 , as most have one parameter, and it is
not named.
From senger at ebi.ac.uk Fri Aug 19 12:46:07 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Fri Aug 19 12:43:53 2005
Subject: [MOBY-l] Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To: <43060373.1080304@ucalgary.ca>
Message-ID:
Regarding the mandatorness of article name, I have heard the same rumour
as you did, Paul. What I got confirmed is that article name must be usede
for included objects in other objects. What was not yet confirmed is that
also Simples and Collections (the wrapper objects) must have article
name. Obviously not having article names mandatory is a pain in the neck -
I am struggling with it the whole last week...
Also API documentation is very unnerving (is this a right English word to
use if you wish to say a four-letter word but you are too polite or too
afraid to say it?). In one place it says "Articles are required to be
named using the articleName attribute." and a bit further it says: "The
articleName attribute of the Simple and/or Collection elements is
optional...".
Mark, could you say loundly and soonish:
1) How it is really?
2) What to do with currently registered services that do not comply with
the rule from 1) ?
Cheers,
Martin
--
Martin Senger
email: martin.senger@gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From markw at illuminae.com Fri Aug 19 14:42:21 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Fri Aug 19 14:31:57 2005
Subject: [MOBY-l] Re: Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To:
References:
Message-ID: <1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
On Fri, 2005-08-19 at 17:46 +0100, Martin Senger wrote:
> What was not yet confirmed is that
> also Simples and Collections (the wrapper objects) must have article
> name. Obviously not having article names mandatory is a pain in the neck -
> I am struggling with it the whole last week...
It was confirmed a while ago - see below...
> Also API documentation is very unnerving (is this a right English word to
> use if you wish to say a four-letter word but you are too polite or too
> afraid to say it?). In one place it says "Articles are required to be
> named using the articleName attribute." and a bit further it says: "The
> articleName attribute of the Simple and/or Collection elements is
> optional...".
I'll look for this error. I really need someone to take over the task
of documentation, however...
> Mark, could you say loundly and soonish:
> 1) How it is really?
I reported loudly and previously:
: From: markw@illuminae.com
: Reply-To: Core developer announcements 2) What to do with currently registered services that do not comply with
> the rule from 1) ?
I think we can add an arbitratry articleName to their inputs, since they
will no doubt ignore it anyway...
Must run!
M
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From gordonp at ucalgary.ca Fri Aug 19 19:13:47 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Fri Aug 19 19:04:18 2005
Subject: [MOBY-l] Re: Where does articleName belong? How can I set it
in a Simple, using jMoby?
In-Reply-To: <1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
References:
<1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
Message-ID: <430667AB.1070906@ucalgary.ca>
The way I dealt with this was to allow a zero-length string as a name.
There's nothing in the API that forbids this. Anything without an
articleName gets a zero-length name, and if there is more than one, I
raise an exception. This handles currently registered one-parameters
jobs fine without resorting to random names, and enforces uniqueness of
the articleName going forward. Actually, the API doesn't say they need
to be unique, just that they need to be there.
>: 1) *all* parameters going into a service must now be named currently using the
>: articleName attribute of the Simple/Collection/Parameter tag
>
>
>
>
>>2) What to do with currently registered services that do not comply with
>>the rule from 1) ?
>>
>>
>
>I think we can add an arbitratry articleName to their inputs, since they
>will no doubt ignore it anyway...
>
>Must run!
>
>M
>
>
>
>
From senger at ebi.ac.uk Fri Aug 19 20:11:13 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Fri Aug 19 20:01:22 2005
Subject: [MOBY-l] Re: Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To: <1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
Message-ID:
> It was confirmed a while ago - see below...
>
Thank for confirmation. It is a good news. (I knew about the email you
quoted - but I was not sure if it was a definitive announcement or just
your contribution to the discussion of the API changes. Now it's clear. My
confusion just confirms that Biomoby still does not have in place a
formal procedure how and when to do API changes.)
Cheers,
Martin
--
Martin Senger
email: martin.senger@gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From paolo at ist.unige.it Mon Aug 22 07:26:40 2005
From: paolo at ist.unige.it (Paolo Romano)
Date: Mon Aug 22 06:15:21 2005
Subject: [MOBY-l] NETTAB 2005 - Deadlines approaching: early registration
and call for posters
Message-ID: <200508221126.j7MBQeqR019972@ibm43p.biotech.ist.unige.it>
Dear all,
this is a reminder for next deadlines of the NETTAB 2005 Workshop on
"Workflows management: new abilities for the biological information
overflow" that will be held on next October 5-7, 2005, in Naples, Italy.
The Scientific Programme is now available on-line at
http://www.nettab.org/2005/progr.html .
The Opening Lecture will be given by Francis Ouellette.
Francis Ouellette is the Director of the UBiC, the Bioinformatics
Centre of the University of British Columbia, Canada. He is Associate
Professor of the Michael Smith Laboratories and of the Department of
Medical Genetics of UBC. He is also a core faculty member in the new
UBC graduate Training Program in Bioinformatics for Health Research,
associate director of Bioinformatics at Genome British Columbia and
director of the Canadian Genetic Diseases Network (CGDN) bioinformatics
core facility where he helps coordinate the Canadian Bioinformatics
Workshops.
Francis has an exceptional curriculum and his recent research,
training and coordination activities make him one of the most known
and appreciated bioinformaticians.
The title of his opening talk will be
"Workflow management in bioinformatics: the possibilities and the challenges".
----------------------------------------------------------------------
The Call for posters and position papers is closing on next friday
August 26, 2005.
You are all warmly invited to present your recent activity related to the
workshops' topics by submitting a poster abstract (1-2 A4 pages, font size
12 pt, MS Word format) by email to posters2005@nettab.org .
Topics are the following:
Technologies and technological platforms of interest, with emphasis on:
- Web Services (SOAP, WSDL, WSFL, UDDI, ....)
- Web Services Choreography and Orchestration
- Semantic Web (RDF, LSID, OWL, ...)
- comparison of available technologies, limitations, pros and cons
- knowledge representation
- biological data and knowledge modeling tools
- Ontologies, Databases and Applications of Semantics in Bioinformatics
Workflow management systems in bioinformatics
- implementations of web services
- implementations of registries
- reuse and versioning of web services and workflows
- workflow management systems
- web interfaces for accessing and executing workflows
- interactive systems to support work flows
Applications of workflow management systems in bioinformatics
- Methodologies for life sciences analysis, such as:
- gene expression,
- genome annotation,
- mass spec peptide fragment identification,
- Encoding of the above in workflows
- Case studies
- Scenarios and use cases
Check all details at: http://www.nettab.org/2005/call.html .
--------------------------------------------------------------------------------
The early registration deadline is also next Friday August 26, 2005.
The registration form is available on-line at:
http://www.nettab.org/2005/rform.html .
The payment of the fee can either be done on-line, through the Online
Payment Form of the Bioinformatics Italian Society (BITS), or by direct
money transfer.
Participation fees are as follows:
Until August 26, 2005:
- Students: 70,00 Euro
- Academic: 130,00 Euro (reduced fee: 117,00 Euro)
- Non-academic: 270,00 Euro (reduced fee: 243,00 Euro)
After August 26, 2005:
- Students: 70,00 Euro
- Academic: 180,00 Euro (reduced fee: 162,00 Euro)
- Non-academic: 370,00 Euro (reduced fee: 333,00 Euro)
The 10% reduction on fees is applied for members of:
- ISCB (International Society for Computational Biology),
http://www.iscb.org/
- BITS (Bioinformatics Italian Society),
http://www.bioinformatics.it/
- Hormone Responsive Breast Cancer (HRBC) Genomics Network,
http://www.hrbc-genomics.net/
- Oncology over Internet (O2I) project,
http://www.o2i.it/
- Interdisciplinary Laboratory for Technologies in Bioinformatics (LITBIO)
--------------------------------------------------------------------------------
I'm looking forward to meeting many of you in Naples quite soon.
Ciao. Paolo
--
Paolo Romano (paolo.romano@istge.it)
Bioinformatics and Structural Proteomics
National Cancer Research Institute (IST)
Largo Rosanna Benzi, 10, I-16132, Genova, Italy
Tel: +39-010-5737-288 Fax: +39-010-5737-295
Web: http://www.nettab.org/promano/
From p.lord at cs.man.ac.uk Mon Aug 22 08:21:48 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: Mon Aug 22 08:11:34 2005
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <430362ED.1090804@ucalgary.ca>
References:
<430362ED.1090804@ucalgary.ca>
Message-ID: <87fyt2duw3.fsf@rpc71.cs.man.ac.uk>
>>>>> "Paul" == Paul Gordon writes:
>>> Alternatively, you pass back some magic cookie when the lease is
>>> taken out in the first place, that is required for
>>> deregistration.
>>>
>>>
>> Well, this is exactly what biomoby has now - and nobody seems to
>> like it. That's why we are talking how to change it.
>>
Paul> Absolutely. Keeping a cookie around somewhere for a month,
Paul> then manually using it to renew the lease (remember,
Paul> automating the lease renewal is just as bad as having an RDF
Paul> file that you need to change) is bad. If I lose my cookie, or
Paul> it is intercepted, I'm toast.
I don't think that you'd need the cookie to renew your lease, just to
remove it before hand. If you lose it, then you just have to wait for
it to run out naturally.
Of course, if it's built into the service provider libraries, there is
no real reason that you should lose it.
>> So the above would work only if I can prove who I am. And the
>> only way how to do it (if we do not want to introduce passwords
>> etc.) is to let somebody come back to my place (because only I
>> can control such a place). That means an agent...
>>
Paul> Exactly my point from before about multiple host
Paul> names/machine. You should only be able to advertise a service
Paul> available in MOBY from the ucalgary.ca domain if you are able
Paul> to put an RDF document on a web server in that domain. It's a
Paul> security check of sorts. I don't see how a lease-only system
Paul> could easily acheive this without some signing certificates,
Paul> and in fact I could prevent the NCBI from registering services
Paul> by doing it first and getting the cookie they need to fix the
Paul> registry.
I think that the lease is a property of the registration, not of the
service. So, if you register a service, it won't prevent anyone else
from registering the same service again. Again, if you put some sort
of call back into the leasing system, you could prevent anyone from
registering a service that they do not own. Although, personally, I
don't see any major problem with third party registration. Feta
effectively works on third party registration, as most of the semantic
descriptions do not come from the service provider.
Phil
From markw at illuminae.com Mon Aug 22 11:34:49 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Mon Aug 22 11:24:33 2005
Subject: [moby] Re: [MOBY-l] Re: Where does articleName belong? How can
I set it in aSimple, using jMoby?
In-Reply-To: <4309A53B.8080701@cnb.uam.es>
References:
<1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
<4309A53B.8080701@cnb.uam.es>
Message-ID: <1124724889.3068.103.camel@bioinfo.icapture.ubc.ca>
On Mon, 2005-08-22 at 12:13 +0200, David Gonz?lez Pisano wrote:
> Umm, does this mean that also Simples into Collections *must* have an
> articlename too, even an arbitrary one? I was wondering if the lack of
> an articleName in the "Simple inside a Collection" example (see the
> API documentation, section INPUT) means that a Simple has to have
> articleName when is the only input
No, Simples inside a collection do not have an articleName, and at most
they would all have the same articleName (I'm actually not sure how the
registry code handles this at the moment... I suspect that it does
record the articleName of a Simple within a Collection if an articleName
has been provided at registration, but I know for certain that none of
my client libraries pay attention to it even if it is there, since it
was not *intended* to be used, nor meaningful).
> Also, probably the twiki needs an update, because as today still says "The articleName attribute of the Simple and/or Collection elements is
> optional (it may or may not be there, and if there, it may or may not have a non-null value)". Note that the last part of the phrase contradicts Paul's solution.
I'll try to find this and fix it.
again... we really need someone to volunteer to take over the task of
re-writing and maintaining the documentation :-)
M
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From dgpisano at cnb.uam.es Mon Aug 22 06:13:15 2005
From: dgpisano at cnb.uam.es (=?ISO-8859-1?Q?David_Gonz=E1lez_Pisano?=)
Date: Mon Aug 22 11:25:22 2005
Subject: [MOBY-l] Re: Where does articleName belong? How can I set it
in aSimple, using jMoby?
In-Reply-To: <1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
References:
<1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
Message-ID: <4309A53B.8080701@cnb.uam.es>
Mark Wilkinson escribi?:
>On Fri, 2005-08-19 at 17:46 +0100, Martin Senger wrote:
>
>
>>What was not yet confirmed is that
>>also Simples and Collections (the wrapper objects) must have article
>>name. Obviously not having article names mandatory is a pain in the neck -
>>I am struggling with it the whole last week...
>>
>>
>It was confirmed a while ago - see below...
>
>
>
>:
>: A few changes to announce for the 0.86 API (this is not yet running on the
>: production server):
>:
>: 1) *all* parameters going into a service must now be named currently using the
>: articleName attribute of the Simple/Collection/Parameter tag
>
>
Umm, does this mean that also Simples into Collections *must* have an
articlename too, even an arbitrary one? I was wondering if the lack of
an articleName in the "Simple inside a Collection" example (see the API
documentation, section INPUT) means that a Simple has to have
articleName when is the only input
<------- articleName here!
but does not need it when is inside a Collection?
<------- no articleName here?
Also, probably the twiki needs an update, because as today still says "The articleName attribute of the Simple and/or Collection elements is
optional (it may or may not be there, and if there, it may or may not have a non-null value)". Note that the last part of the phrase contradicts Paul's solution.
David (working on the Errors proposal, which heavily depends on the answer to this mail) ;-)
From fgibbons at hms.harvard.edu Mon Aug 22 11:54:11 2005
From: fgibbons at hms.harvard.edu (Frank Gibbons)
Date: Mon Aug 22 11:43:36 2005
Subject: [moby] Re: [MOBY-l] Re: Where does articleName belong? How
can I set it in aSimple, using jMoby?
In-Reply-To: <1124724889.3068.103.camel@bioinfo.icapture.ubc.ca>
References: <4309A53B.8080701@cnb.uam.es>
<1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
<4309A53B.8080701@cnb.uam.es>
Message-ID: <5.2.1.1.2.20050822115221.010ccf88@email.med.harvard.edu>
Mark,
At 11:34 AM 8/22/2005, you wrote:
> > Also, probably the twiki needs an update, because as today still says
> "The articleName attribute of the Simple and/or Collection elements is
> > optional (it may or may not be there, and if there, it may or may not
> have a non-null value)". Note that the last part of the phrase
> contradicts Paul's solution.
>
>I'll try to find this and fix it.
>
>again... we really need someone to volunteer to take over the task of
>re-writing and maintaining the documentation :-)
I'll volunteer for that. You're referring to the Wiki API page, I assume,
or are there other docs (besides the POD and javadoc)?
-Frank
PhD, Computational Biologist,
Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
Tel: 617-432-3555 Fax:
617-432-3557 http://llama.med.harvard.edu/~fgibbons
From senger at ebi.ac.uk Mon Aug 22 12:08:13 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Mon Aug 22 11:57:45 2005
Subject: [moby] Re: [MOBY-l] Re: Where does articleName belong? How can
I set it in aSimple, using jMoby?
In-Reply-To: <5.2.1.1.2.20050822115221.010ccf88@email.med.harvard.edu>
Message-ID:
> I'll volunteer for that. You're referring to the Wiki API page, I assume,
> or are there other docs (besides the POD and javadoc)?
>
That's goooood!
But please be merciless, as much as you can. :-)
The main problem I had with these pages from/at the beginning is that
only the initiated easily understand that the page actually contain two
APIs - one for registry and one for services. The current page mix them
too much (IMHO).
I also volunteer for proof-reading (just send me your draft) - and I
promise I will be chasing Mark with every smallest ambiguity that I could
find.
Cheers,
Martin
--
Martin Senger
email: martin.senger@gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From markw at illuminae.com Mon Aug 22 12:14:49 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Mon Aug 22 12:04:29 2005
Subject: [moby] Re: [MOBY-l] Re: Where does articleName belong? How can
I set it in aSimple, using jMoby?
In-Reply-To:
References:
Message-ID: <1124727289.3068.131.camel@bioinfo.icapture.ubc.ca>
> and I
> promise I will be chasing Mark with every smallest ambiguity that I could
> find.
OOOooohhh yeah... he sure will! I know that from experience ;-)
M
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From markw at illuminae.com Mon Aug 22 12:16:04 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Mon Aug 22 12:05:34 2005
Subject: [moby] Re: [MOBY-l] Re: Where does articleName belong? How can
I set it in aSimple, using jMoby?
In-Reply-To: <5.2.1.1.2.20050822115221.010ccf88@email.med.harvard.edu>
References: <4309A53B.8080701@cnb.uam.es>
<1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
<4309A53B.8080701@cnb.uam.es>
<5.2.1.1.2.20050822115221.010ccf88@email.med.harvard.edu>
Message-ID: <1124727364.3068.134.camel@bioinfo.icapture.ubc.ca>
that's fantastic! Thanks Frank!
I need to set you up with a developers account on the TWiki machine so
that you can edit the main page I think... let's coordinate this off-
list.
Speaking of the Twiki machine... Simon - how's the new website plan
coming?
M
On Mon, 2005-08-22 at 11:54 -0400, Frank Gibbons wrote:
> Mark,
>
> At 11:34 AM 8/22/2005, you wrote:
> > > Also, probably the twiki needs an update, because as today still says
> > "The articleName attribute of the Simple and/or Collection elements is
> > > optional (it may or may not be there, and if there, it may or may not
> > have a non-null value)". Note that the last part of the phrase
> > contradicts Paul's solution.
> >
> >I'll try to find this and fix it.
> >
> >again... we really need someone to volunteer to take over the task of
> >re-writing and maintaining the documentation :-)
>
>
> I'll volunteer for that. You're referring to the Wiki API page, I assume,
> or are there other docs (besides the POD and javadoc)?
>
> -Frank
>
>
> PhD, Computational Biologist,
> Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
> Tel: 617-432-3555 Fax:
> 617-432-3557 http://llama.med.harvard.edu/~fgibbons
>
> _______________________________________________
> moby-l mailing list
> moby-l@biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From silvia77 at dist.unige.it Thu Aug 25 06:12:09 2005
From: silvia77 at dist.unige.it (scaglione)
Date: Thu Aug 25 06:10:47 2005
Subject: [MOBY-l] sth_collection_ins
Message-ID: <1124964729.430d9979a03e9@webmail.unige.it>
hello,
I have followed instructions to install a moby central in my machine.I have
tried the testMOBYClientCentral_v05.pl script on my mobycentral and an error
appeared:
>perl testMOBYCentral_v05.pl
Useless use of hash element in void context
at /usr/lib/perl5/site_perl/5.8.5/RDF/Core/Serializer.pm line 52.
Global symbol "$sth_collection_ins" requires explicit package name
at /usr/lib/perl5/site_perl/5.8.5/MOBY/Central.pm line 3426.
Compilation failed in require at testMOBYCentral_v05.pl line 4.
BEGIN failed--compilation aborted at testMOBYCentral_v05.pl line 4.
what does it mean?
The overall configuration of the moby central seemed to be ok, since the test
of the same script on the main mobycentral works...
many thanks
silvia
--
PhD. Silvia Scaglione
DIST University of Genoa
Viale Causa,13
16145 Genoa
phone +39 010 353 2789
phone CBA +39 010 5737 505
fax +39 010 353 2948
From markw at illuminae.com Thu Aug 25 12:27:21 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Thu Aug 25 12:17:16 2005
Subject: [moby] [MOBY-l] sth_collection_ins
In-Reply-To: <1124964729.430d9979a03e9@webmail.unige.it>
References: <1124964729.430d9979a03e9@webmail.unige.it>
Message-ID: <1124987241.11350.50.camel@bioinfo.icapture.ubc.ca>
Can you first confirm where you got your mobycentral code from? Is it
one of the newer releases? I'm certain that the latest release is
functional (though there are a few small bugs in it that are now fixed
in the CVS, but the CVS code isn't supported yet...)
make/make install this code and see if it solves your problems. You
probably wont need to modify any of your configuration after this
installation.
http://biomoby.org/releases/biomoby.0.8.2a.tar.gz
M
On Thu, 2005-08-25 at 12:12 +0200, scaglione wrote:
> hello,
> I have followed instructions to install a moby central in my machine.I have
> tried the testMOBYClientCentral_v05.pl script on my mobycentral and an error
> appeared:
>
> >perl testMOBYCentral_v05.pl
>
> Useless use of hash element in void context
> at /usr/lib/perl5/site_perl/5.8.5/RDF/Core/Serializer.pm line 52.
> Global symbol "$sth_collection_ins" requires explicit package name
> at /usr/lib/perl5/site_perl/5.8.5/MOBY/Central.pm line 3426.
> Compilation failed in require at testMOBYCentral_v05.pl line 4.
> BEGIN failed--compilation aborted at testMOBYCentral_v05.pl line 4.
>
> what does it mean?
> The overall configuration of the moby central seemed to be ok, since the test
> of the same script on the main mobycentral works...
> many thanks
> silvia
>
>
>
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From jmfernandez at cnb.uam.es Fri Aug 26 12:37:31 2005
From: jmfernandez at cnb.uam.es (=?ISO-8859-1?Q?Jos=E9_Mar=EDa_Fern=E1ndez_Gonz=E1lez?=)
Date: Fri Aug 26 12:26:58 2005
Subject: [MOBY-l] RESOURCES/MOBY-S/Objects not working?
Message-ID: <430F454B.7020603@cnb.uam.es>
Hi people!
Is there any problem with Tomcat server in MOBY Central? Since Tuesday
(or so) I'm playing with latest Taverna CVS, and BioMOBY related stuff
doesn't work because there is a HTTP 500 code related to
http://www.biomoby.org/RESOURCES/MOBY-S/Objects (RDF Object
definitions). As Taverna 1.2 also depends on this URL, MOBY services
tree is not either working.
Best Regards,
Jos? Mar?a
--
Jos? Mar?a Fern?ndez Gonz?lez e-mail: jmfernandez@cnb.uam.es
Tlfn: (+34) 91 585 54 50 Fax: (+34) 91 585 45 06
Grupo de Dise?o de Proteinas Protein Design Group
Centro Nacional de Biotecnolog?a National Center of Biotechnology
C.P.: 28049 Zip Code: 28049
C/. Darwin n? 3 (Campus Cantoblanco, U. Aut?noma), Madrid (Spain)
From edward.kawas at gmail.com Fri Aug 26 13:20:17 2005
From: edward.kawas at gmail.com (Eddie Kawas)
Date: Fri Aug 26 13:10:55 2005
Subject: [MOBY-l] RESOURCES/MOBY-S/Objects not working?
In-Reply-To: <430F454B.7020603@cnb.uam.es>
References: <430F454B.7020603@cnb.uam.es>
Message-ID:
Hi,
I will look into this asap (saturday or sunday). It looks like the
server works because the other RDF documents are being generated so
something must be wrong with either the configuration.
Actually, I just noticed that the types script for objects is
returning stuff that it shouldnt
(http://mobycentral.icapture.ubc.ca:8090/types/Objects) and this
script is utilized by the RESOURCES script. As soon as I am able to
check this out, I will and l will fix it. The problem seems to lie
with the object 'Regex' and its description (no longer limited to one
line per object).
Eddie
On 8/26/05, Jos? Mar?a Fern?ndez Gonz?lez wrote:
> Hi people!
> Is there any problem with Tomcat server in MOBY Central? Since Tuesday
> (or so) I'm playing with latest Taverna CVS, and BioMOBY related stuff
> doesn't work because there is a HTTP 500 code related to
> http://www.biomoby.org/RESOURCES/MOBY-S/Objects (RDF Object
> definitions). As Taverna 1.2 also depends on this URL, MOBY services
> tree is not either working.
>
> Best Regards,
> Jos? Mar?a
>
> --
> Jos? Mar?a Fern?ndez Gonz?lez e-mail: jmfernandez@cnb.uam.es
> Tlfn: (+34) 91 585 54 50 Fax: (+34) 91 585 45 06
> Grupo de Dise?o de Proteinas Protein Design Group
> Centro Nacional de Biotecnolog?a National Center of Biotechnology
> C.P.: 28049 Zip Code: 28049
> C/. Darwin n? 3 (Campus Cantoblanco, U. Aut?noma), Madrid (Spain)
> _______________________________________________
> moby-l mailing list
> moby-l@biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
>
From senger at ebi.ac.uk Fri Aug 26 20:28:31 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Fri Aug 26 20:25:30 2005
Subject: [MOBY-l] RESOURCES/MOBY-S/Objects not working?
In-Reply-To:
Message-ID:
> The problem seems to lie with the object 'Regex' and its description
> (no longer limited to one line per object).
>
Regex is my object! I am always happy to shoot down all servers around
:-)
But seriusly, anything wrong with longer description? Does API preclude
it?
Cheers,
Martin
--
Martin Senger
email: martin.senger@gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From edward.kawas at gmail.com Sat Aug 27 09:55:31 2005
From: edward.kawas at gmail.com (Eddie Kawas)
Date: Sat Aug 27 09:45:34 2005
Subject: [MOBY-l] RESOURCES/MOBY-S/Objects not working?
In-Reply-To:
References:
Message-ID:
I think the description is fine, Mark will need to verify this. I will
just have to update my logic (already done in my head and when i get
home today i will update the codebase).
Eddie
On 8/26/05, Martin Senger wrote:
> > The problem seems to lie with the object 'Regex' and its description
> > (no longer limited to one line per object).
> >
> Regex is my object! I am always happy to shoot down all servers around
> :-)
> But seriusly, anything wrong with longer description? Does API preclude
> it?
>
> Cheers,
> Martin
>
> --
> Martin Senger
> email: martin.senger@gmail.com
> skype: martinsenger
> International Rice Research Institute
> Biometrics and Bioinformatics Unit
> DAPO BOX 7777, Metro Manila
> Philippines, phone: +63-2-580-5600 (ext.2324)
>
>
From senger at ebi.ac.uk Mon Aug 29 10:32:30 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Mon Aug 29 10:25:08 2005
Subject: [MOBY-l] Moses - Moby Services Support
In-Reply-To: <1476.82.66.216.27.1118945157.squirrel@82.66.216.27>
Message-ID:
Hi all,
My first month at Philippines, at IRRI, is almost finished. So it's
time to announce some progress :-)
I have commited code and a lot of documentation helping to write
Biomoby services in Java. The idea is to generate classes for all
registered data types and then generate skeleton for a new service. Its
developer can extend such skeleton , put there the business logic without
worries about the Biomoby XML and about the SOAP itself. These things
should be hidden.
It is similar what Paul Gordon provides for the clients, but here you
can use strongly-type approach (having all data types generated). Or, you
can say, that we can have more ways to do the same, like in Perl :-)
As (almost) a side-effect, the generated code, because it has rich
API/javadoc comments, can be used as a primitive browser of the biomoby
registry. Here you can see the API for all data types and for all services
as they are today (when I find a place where I can run a cronjob, we can
have this API up-to-date anytime you look):
http://www.ebi.ac.uk/~senger/jMoby/APIservices/index.html
The documentation is in jMoby and can be viewed at:
http://biomoby.org/moby-live/Java/docs/Moses.html.
I hope you'll enjoy it and I am happy to hear your suggestions, bug
reports (remember there is also a bugzilla open now, but feel free to send
emails direcly to me) and anything bout Moses.
With regards,
Martin
--
Martin Senger
email: martin.senger@gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From silvia77 at dist.unige.it Thu Aug 25 05:52:50 2005
From: silvia77 at dist.unige.it (scaglione)
Date: Tue Aug 30 18:56:09 2005
Subject: [MOBY-l] sth_collection_ins
Message-ID: <1124963570.430d94f212a32@webmail.unige.it>
hello,
I have followed instructions to install a moby central in my machine.I have
tried the testMOBYClientCentral_v05.pl script on my mobycentral and an error
appeared:
>perl testMOBYCentral_v05.pl
Useless use of hash element in void context
at /usr/lib/perl5/site_perl/5.8.5/RDF/Core/Serializer.pm line 52.
Global symbol "$sth_collection_ins" requires explicit package name
at /usr/lib/perl5/site_perl/5.8.5/MOBY/Central.pm line 3426.
Compilation failed in require at testMOBYCentral_v05.pl line 4.
BEGIN failed--compilation aborted at testMOBYCentral_v05.pl line 4.
what does it mean?
The overall configuration of the moby central seemed to be ok, since the test
of the same script on the main mobycentral works...
many thanks
silvia
--
PhD. Silvia Scaglione
DIST University of Genoa
Viale Causa,13
16145 Genoa
phone +39 010 353 2789
phone CBA +39 010 5737 505
fax +39 010 353 2948
From silvia77 at dist.unige.it Wed Aug 31 09:32:20 2005
From: silvia77 at dist.unige.it (Silvia Scaglione)
Date: Wed Aug 31 09:47:21 2005
Subject: [moby] [MOBY-l] test
In-Reply-To: <1124987241.11350.50.camel@bioinfo.icapture.ubc.ca>
References: <1124964729.430d9979a03e9@webmail.unige.it>
<1124987241.11350.50.camel@bioinfo.icapture.ubc.ca>
Message-ID: <6.2.0.14.0.20050831152820.01cc1998@mail.unige.it>
I have used the newer releases suggested below, but new errors appears
after testing testMOBYClientCentral_v05.pl script on the mobycentral:
...
test 19 [PASS]
test 20 [FAIL] malformed payload invalid character string for
serviceName. Must start with a letter followed by [A-Za-z0-9_]
test 21 [FAIL] Expected to find service; didn't find service
test 22 [FAIL] Expected to find service; didn't find service
test 23 [PASS]
test 24 [PASS]
Use of uninitialized value in print at testMOBYClientCentral_v05.pl line 250.
test 25 [FAIL] WSDL was not retrieved
test 26 [FAIL] The service specified by authority=www.illuminae.com
servicename=myfirstservice2 does not exist in the registry
test 27 [PASS]
...
Moreover, when I tried the testMOBYClientCentral_v05.pl script on my
mobycentral, the error message changes a little bit:
>[www@ml370 scripts]$ perl testMOBYClientCentral_v05.pl
>TESTING MOBY CLIENT with
> URL: http://ml370.biotech.ist.unige.it/cgi-bin/MOBY-Central.pl
> URI: http://ml370.biotech.ist.unige.it/MOBY/Central
> Proxy: No Proxy Server
>
>test 1a [PASS]
>test 1b [PASS]
>test 1 [FAIL] Relationship Type hasa does not exist in the
>biomoby.org Relationship Type system
>
>test 2 [PASS]
>test 3 [FAIL] Relationship Type hasa does not exist in the
>biomoby.org Relationship Type system
>test 4 [FAIL] Relationship type ISA does not exist in the
>relationship ontology
>test 5 [PASS]
>test 6 [FAIL] Object type $term cannot be resolved to an LSID
>test 7 [PASS]
>test 8 [FAIL] Object type $term cannot be resolved to an LSID
>test 9 [FAIL] Service Type HypotheticalService1 does not exist in
>the ontology
>test 10 [PASS]
>test 11 [PASS]
>test 12 [PASS]
>test 13 [PASS]
>test 14 [FAIL] Relationship Type isa does not exist in the
>biomoby.org Relationship Type system
>test 15 [FAIL] Relationship Type isa does not exist in the
>biomoby.org Relationship Type system
>test 16 [PASS]
>test 17 [PASS]
>test 18 [PASS]
>test 19 [PASS]
>test 20 [PASS]
>500 Internal Server Error
> ERROR ERROR ERROR
...any suggestions?
At 18.27 25/08/2005, Mark Wilkinson wrote:
>Can you first confirm where you got your mobycentral code from? Is it
>one of the newer releases? I'm certain that the latest release is
>functional (though there are a few small bugs in it that are now fixed
>in the CVS, but the CVS code isn't supported yet...)
>
>make/make install this code and see if it solves your problems. You
>probably wont need to modify any of your configuration after this
>installation.
>
>http://biomoby.org/releases/biomoby.0.8.2a.tar.gz
>
>
>M
>
>
>
>On Thu, 2005-08-25 at 12:12 +0200, scaglione wrote:
> > hello,
> > I have followed instructions to install a moby central in my machine.I
> have
> > tried the testMOBYClientCentral_v05.pl script on my mobycentral and an
> error
> > appeared:
> >
> > >perl testMOBYCentral_v05.pl
> >
> > Useless use of hash element in void context
> > at /usr/lib/perl5/site_perl/5.8.5/RDF/Core/Serializer.pm line 52.
> > Global symbol "$sth_collection_ins" requires explicit package name
> > at /usr/lib/perl5/site_perl/5.8.5/MOBY/Central.pm line 3426.
> > Compilation failed in require at testMOBYCentral_v05.pl line 4.
> > BEGIN failed--compilation aborted at testMOBYCentral_v05.pl line 4.
> >
> > what does it mean?
> > The overall configuration of the moby central seemed to be ok, since
> the test
> > of the same script on the main mobycentral works...
> > many thanks
> > silvia
> >
> >
> >
>--
>"Ontologists do it with the edges!"
>
>Mark Wilkinson
>Asst. Professor
>Dept. of Medical Genetics
>University of British Columbia
>PI in Bioinformatics
>iCAPTURE Centre
>St. Paul's Hospital
>Rm. 166, 1081 Burrard St.
>Vancouver, BC, V6Z 1Y6
>tel: 604 682 2344 x62129
>fax: 604 806 9274
PhD. Silvia Scaglione
Universita`degli Studi di Genova
Facolta' di Ingegneria, D.I.S.T. BioLab
Viale Causa, 13
16145 GENOVA - Italy
Tel Bio-lab +39 010 353 2789
Tel Centro Biotecnologie Avanzate +39 010 5737 505
Fax +39 010 353 2948
E-mail silvia77@dist.unige.it
From markw at illuminae.com Wed Aug 31 10:03:16 2005
From: markw at illuminae.com (mark wilkinson)
Date: Wed Aug 31 09:59:05 2005
Subject: [moby] [MOBY-l] test
Message-ID: <1390300886-1125497047-cardhu_blackberry.rim.net-15323-@engine08-cell01>
Use "make test" instead - it is up to date and far far far more comprehensive.
M
-----Original Message-----
From: Silvia Scaglione
Date: Wed, 31 Aug 2005 15:32:20
To:moby-l@biomoby.org
Subject: Re: [moby] [MOBY-l] test
I have used the newer releases suggested below, but new errors appears
after testing testMOBYClientCentral_v05.pl script on the mobycentral:
...
test 19 [PASS]
test 20 [FAIL] malformed payload invalid character string for
serviceName. Must start with a letter followed by [A-Za-z0-9_]
test 21 [FAIL] Expected to find service; didn't find service
test 22 [FAIL] Expected to find service; didn't find service
test 23 [PASS]
test 24 [PASS]
Use of uninitialized value in print at testMOBYClientCentral_v05.pl line 250.
test 25 [FAIL] WSDL was not retrieved
test 26 [FAIL] The service specified by authority=www.illuminae.com
servicename=myfirstservice2 does not exist in the registry
test 27 [PASS]
...
Moreover, when I tried the testMOBYClientCentral_v05.pl script on my
mobycentral, the error message changes a little bit:
>[www@ml370 scripts]$ perl testMOBYClientCentral_v05.pl
>TESTING MOBY CLIENT with
> URL: http://ml370.biotech.ist.unige.it/cgi-bin/MOBY-Central.pl
> URI: http://ml370.biotech.ist.unige.it/MOBY/Central
> Proxy: No Proxy Server
>
>test 1a [PASS]
>test 1b [PASS]
>test 1 [FAIL] Relationship Type hasa does not exist in the
>biomoby.org Relationship Type system
>
>test 2 [PASS]
>test 3 [FAIL] Relationship Type hasa does not exist in the
>biomoby.org Relationship Type system
>test 4 [FAIL] Relationship type ISA does not exist in the
>relationship ontology
>test 5 [PASS]
>test 6 [FAIL] Object type $term cannot be resolved to an LSID
>test 7 [PASS]
>test 8 [FAIL] Object type $term cannot be resolved to an LSID
>test 9 [FAIL] Service Type HypotheticalService1 does not exist in
>the ontology
>test 10 [PASS]
>test 11 [PASS]
>test 12 [PASS]
>test 13 [PASS]
>test 14 [FAIL] Relationship Type isa does not exist in the
>biomoby.org Relationship Type system
>test 15 [FAIL] Relationship Type isa does not exist in the
>biomoby.org Relationship Type system
>test 16 [PASS]
>test 17 [PASS]
>test 18 [PASS]
>test 19 [PASS]
>test 20 [PASS]
>500 Internal Server Error
> ERROR ERROR ERROR
...any suggestions?
At 18.27 25/08/2005, Mark Wilkinson wrote:
>Can you first confirm where you got your mobycentral code from? Is it
>one of the newer releases? I'm certain that the latest release is
>functional (though there are a few small bugs in it that are now fixed
>in the CVS, but the CVS code isn't supported yet...)
>
>make/make install this code and see if it solves your problems. You
>probably wont need to modify any of your configuration after this
>installation.
>
>http://biomoby.org/releases/biomoby.0.8.2a.tar.gz
>
>
>M
>
>
>
>On Thu, 2005-08-25 at 12:12 +0200, scaglione wrote:
> > hello,
> > I have followed instructions to install a moby central in my machine.I
> have
> > tried the testMOBYClientCentral_v05.pl script on my mobycentral and an
> error
> > appeared:
> >
> > >perl testMOBYCentral_v05.pl
> >
> > Useless use of hash element in void context
> > at /usr/lib/perl5/site_perl/5.8.5/RDF/Core/Serializer.pm line 52.
> > Global symbol "$sth_collection_ins" requires explicit package name
> > at /usr/lib/perl5/site_perl/5.8.5/MOBY/Central.pm line 3426.
> > Compilation failed in require at testMOBYCentral_v05.pl line 4.
> > BEGIN failed--compilation aborted at testMOBYCentral_v05.pl line 4.
> >
> > what does it mean?
> > The overall configuration of the moby central seemed to be ok, since
> the test
> > of the same script on the main mobycentral works...
> > many thanks
> > silvia
> >
> >
> >
>--
>"Ontologists do it with the edges!"
>
>Mark Wilkinson
>Asst. Professor
>Dept. of Medical Genetics
>University of British Columbia
>PI in Bioinformatics
>iCAPTURE Centre
>St. Paul's Hospital
>Rm. 166, 1081 Burrard St.
>Vancouver, BC, V6Z 1Y6
>tel: 604 682 2344 x62129
>fax: 604 806 9274
PhD. Silvia Scaglione
Universita`degli Studi di Genova
Facolta' di Ingegneria, D.I.S.T. BioLab
Viale Causa, 13
16145 GENOVA - Italy
Tel Bio-lab +39 010 353 2789
Tel Centro Biotecnologie Avanzate +39 010 5737 505
Fax +39 010 353 2948
E-mail silvia77@dist.unige.it
_______________________________________________
moby-l mailing list
moby-l@biomoby.org
http://biomoby.org/mailman/listinfo/moby-l
--
Mark Wilkinson
...on the road!
From harttig at mail.dife.de Tue Aug 2 11:42:38 2005
From: harttig at mail.dife.de (Ulrich Harttig)
Date: Tue, 02 Aug 2005 17:42:38 +0200
Subject: [MOBY-l] parameter handling problems
Message-ID: <42EF946E.4090708@mail.dife.de>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
I have some problems with handling a message to a service asking for an input
object and a secondary parameter using the Perl API (and XML::LibXML version
1.58, libxml2 version 2.6.16)
After parsing the message with 'complexServiceInputParser()' extraction of the
the simple article is no problem but accessing the content of the secondary
article with
($secondary) = getNodeContentWithArticle($input, "Parameter", "VarType");
does not find any content although the $input node contains
xml
Is there any reason why this shouldn't work ?
Could someone point me to examples where accessing parameters works ?
Thanks for your help
Best regards
Ulrich
- --
Ulrich Harttig
German Institute of Human Nutrition Potsdam-Rehbruecke
Dept. of Epidemiology
Arthur-Scheunert-Allee 114-116, 14558 Nuthetal, Germany
phone: +49 (0)33200 88-729, fax: +49 (0)33200 88-721
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)
iD8DBQFC75RueYdmlKJoHzkRAuKeAJ9QkwlecCYbquhW6W1u1y9arXzPAwCg0pjG
64tQh4osNp61zpuCQYo5VcI=
=AOm+
-----END PGP SIGNATURE-----
From harttig at mail.dife.de Tue Aug 2 11:42:38 2005
From: harttig at mail.dife.de (Ulrich Harttig)
Date: Tue, 02 Aug 2005 17:42:38 +0200
Subject: [MOBY-l] parameter handling problems
Message-ID: <42EF946E.4090708@mail.dife.de>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
I have some problems with handling a message to a service asking for an input
object and a secondary parameter using the Perl API (and XML::LibXML version
1.58, libxml2 version 2.6.16)
After parsing the message with 'complexServiceInputParser()' extraction of the
the simple article is no problem but accessing the content of the secondary
article with
($secondary) = getNodeContentWithArticle($input, "Parameter", "VarType");
does not find any content although the $input node contains
xml
Is there any reason why this shouldn't work ?
Could someone point me to examples where accessing parameters works ?
Thanks for your help
Best regards
Ulrich
- --
Ulrich Harttig
German Institute of Human Nutrition Potsdam-Rehbruecke
Dept. of Epidemiology
Arthur-Scheunert-Allee 114-116, 14558 Nuthetal, Germany
phone: +49 (0)33200 88-729, fax: +49 (0)33200 88-721
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)
iD8DBQFC75RueYdmlKJoHzkRAuKeAJ9QkwlecCYbquhW6W1u1y9arXzPAwCg0pjG
64tQh4osNp61zpuCQYo5VcI=
=AOm+
-----END PGP SIGNATURE-----
From d.haase at gsf.de Wed Aug 3 02:25:20 2005
From: d.haase at gsf.de (Dirk Haase)
Date: Wed, 3 Aug 2005 08:25:20 +0200
Subject: [MOBY-l] parameter handling problems
In-Reply-To: <42EF946E.4090708@mail.dife.de>
References: <42EF946E.4090708@mail.dife.de>
Message-ID: <200508030825.20164.d.haase@gsf.de>
On Tuesday 02 August 2005 17:42, Ulrich Harttig wrote:
> ($secondary) = getNodeContentWithArticle($input, "Parameter", "VarType");
>
> does not find any content although the $input node contains
>
>
> xml
>
>
> Is there any reason why this shouldn't work ?
If I remember correctly, I tried the same and didn't work for me neither...
> Could someone point me to examples where accessing parameters works ?
Here is what works for me: the service takes several secondaries, so it puts
all parameter values into a %paramHash:
if ( $articleType == SECONDARY ) {
my $param = $input->getAttribute('articleName');
my @valueNodes = $input->getChildrenByTagName('moby:Value');
my $value = $valueNodes[0]->textContent(); # there should be only one value
$paramHash{$param} = $value;
}
Hope this helps.
Regards,
dirk
From harttig at mail.dife.de Wed Aug 3 03:48:34 2005
From: harttig at mail.dife.de (Ulrich Harttig)
Date: Wed, 03 Aug 2005 09:48:34 +0200
Subject: [MOBY-l] parameter handling problems
In-Reply-To: <200508030825.20164.d.haase@gsf.de>
References: <42EF946E.4090708@mail.dife.de> <200508030825.20164.d.haase@gsf.de>
Message-ID: <42F076D2.8060806@mail.dife.de>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Dirk Haase wrote:
> On Tuesday 02 August 2005 17:42, Ulrich Harttig wrote:
>
>
>>($secondary) = getNodeContentWithArticle($input, "Parameter", "VarType");
>>
>>does not find any content although the $input node contains
>>
>>
>> xml
>>
>>
>>Is there any reason why this shouldn't work ?
>
>
> If I remember correctly, I tried the same and didn't work for me neither...
>
>
>>Could someone point me to examples where accessing parameters works ?
>
>
> Here is what works for me: the service takes several secondaries, so it puts
> all parameter values into a %paramHash:
>
> if ( $articleType == SECONDARY ) {
> my $param = $input->getAttribute('articleName');
> my @valueNodes = $input->getChildrenByTagName('moby:Value');
> my $value = $valueNodes[0]->textContent(); # there should be only one value
> $paramHash{$param} = $value;
> }
>
> Hope this helps.
>
> Regards,
> dirk
Thanks a lot for the help, your example did work.
But I still wondered why $input->getChildrenByTagName('moby:Value') worked but
$input->getElementByTagName('moby:Parameter') would not. I was guessing that
this had something to do with the structure/content of the node from the
secondary article.
getElementsByTagName is supposed to return a named descendant of the node.
According to the DOM spec: "A descendant node of any node A is any node below A
in a tree model, where "below" means "away from the root."
But to me it seems that is actually the root of the node ,
judging by the content of the $input parameter, so one would not be able to find
a descendant of the same name.
So if I pass instead the original message using the $input->parentNode to the
CommonSub function 'getNodeContentWithArticle' it then returns the desired
parameter value
($secondary) = getNodeContentWithArticle($input->parentNode, "Parameter",
"VarType");
best regards
Ulrich
- --
Ulrich Harttig
German Institute of Human Nutrition Potsdam-Rehbruecke
Dept. of Epidemiology
Arthur-Scheunert-Allee 114-116, 14558 Nuthetal, Germany
phone: +49 (0)33200 88-729, fax: +49 (0)33200 88-721
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)
iD8DBQFC8HbSeYdmlKJoHzkRAp6ZAJ98gonlRj76yBlZEf8feQYd+fx9QACgtVqI
qDx3gY2ljO29ZcdGqy9QZ14=
=OuRI
-----END PGP SIGNATURE-----
From rebecca.ernst at gsf.de Thu Aug 4 03:11:25 2005
From: rebecca.ernst at gsf.de (Rebecca Ernst)
Date: Thu, 04 Aug 2005 09:11:25 +0200
Subject: [MOBY-l] Biomoby and Taverna 1.2
Message-ID: <42F1BF9D.6040401@gsf.de>
Hi there!
It might be boring for you that I start this topic again but the problem
still exists and within Planet we are getting in serious problems now
and I guess other people will also run into them!
Fact is, that old workflows don't work in Taverna 1.2 . According to Tom
this is not how it should be, software should always be downwards
compatible and that's how I see it, too.
Now, due to the fact that it doesn't work but people want to make use of
the new, cool features of 1.2 they start developing services which run
on 1.2. as you can guess they don't work in 1.1. anymore.
It can't be a solution to develop biomoby workflows for every Taverna
version so could we please try to solve this problem now before lots of
people will have to change their services back and forth ??
If noone (except of Heiko Dirk and me) want the collections to be
changed in the BioMoby API we will have to change the default behaviour
of Taverna back to the old one!
Cheers,
Rebecca
--
Rebecca Ernst
MIPS, Inst. for Bioinformatics
GSF Research Center for Environment and Health
Ingolstaedter Landstr. 1
85764 Neuherberg
fon: +49 89 3187 3583
email: Rebecca.Ernst at gsf.de
From senger at ebi.ac.uk Thu Aug 4 03:21:48 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Thu, 4 Aug 2005 08:21:48 +0100 (BST)
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1BF9D.6040401@gsf.de>
Message-ID:
Rebecca,
I have talked about it with Eddie when I met him in Malaga about
two/three weeks ago. And he promised (!) that he was going to put back the
old CreateMoby... and ParseMoby... classes. And I have never heard from
him since. So it is very good that you started this topic again - perhaps
he is listening :-)
Cheers,
Martin
--
Martin Senger martin.senger at gmail.com
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-580-5600 (ext.2324)
From edward.kawas at gmail.com Thu Aug 4 03:24:25 2005
From: edward.kawas at gmail.com (Edward Kawas)
Date: Thu, 4 Aug 2005 00:24:25 -0700
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To:
Message-ID: <42f1c2c1.6263d1d9.1fa1.67e4@mx.gmail.com>
Hi,
I have put the old functionality back into Taverna.
If you build a fresh copy from the cvs, you will notice
(hopefully) that all is well again.
Eddie
> -----Original Message-----
> From: moby-l-bounces at portal.open-bio.org [mailto:moby-l-
> bounces at portal.open-bio.org] On Behalf Of Martin Senger
> Sent: Thursday, August 04, 2005 12:22 AM
> To: Rebecca Ernst
> Cc: taverna-users at lists.sourceforge.net; mobyl
> Subject: Re: [MOBY-l] Biomoby and Taverna 1.2
>
> Rebecca,
> I have talked about it with Eddie when I met him in
> Malaga about
> two/three weeks ago. And he promised (!) that he was going
> to put back the
> old CreateMoby... and ParseMoby... classes. And I have
> never heard from
> him since. So it is very good that you started this topic
> again - perhaps
> he is listening :-)
>
> Cheers,
> Martin
>
> --
> Martin Senger martin.senger at gmail.com
>
> International Rice Research Institute
> Biometrics and Bioinformatics Unit
> DAPO BOX 7777, Metro Manila
> Philippines, phone: +63-580-5600 (ext.2324)
>
> _______________________________________________
> moby-l mailing list
> moby-l at biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
From markw at illuminae.com Thu Aug 4 03:48:35 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu, 4 Aug 2005 03:48:35 -0400
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1BF9D.6040401@gsf.de>
References: <42F1BF9D.6040401@gsf.de>
Message-ID: <1123141715.42f1c853bd83f@webmail.illuminae.com>
> If noone (except of Heiko Dirk and me) want the collections to be
> changed in the BioMoby API we will have to change the default behaviour
> of Taverna back to the old one!
I think your use-case for wanting the behaviour of collections to change is
unfounded, since it overloads the intended meaning of the Collection element by
putting WAAAAAAY too much semantics in it. The use case of having sets of
multiple alignments represented as Collections of Collections means that the
inner-most Collection now not only represents a "bag" of things, but a "bag" of
things that **DO** have a close semantic relationship to each other - these are
a set of related sequences, and a set of related sequences **is a useful
Object** in the context of MOBY, could easily trigger the discovery of a
particular set of services that can specifically operate on sets of closely
related sequences (versus sets of unrelated sequences), and therefore should be
given a class of its own (IMO).
Moreover, if we allow collections of collections, we then have a problem of
recursion - collections of collections of collections of collections of
collections.... What a nightmare!
The API change is not a significant one, I agree... and in fact, since people
are already mis-using the messaging format in this way much of the software out
there can already handle the situation... but given that particular use-case, I
think you are losing more semantics than you need to be by refusing to create a
new Object Class.
M
From tmo at ebi.ac.uk Thu Aug 4 04:06:45 2005
From: tmo at ebi.ac.uk (Tom Oinn)
Date: Thu, 04 Aug 2005 09:06:45 +0100
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123141715.42f1c853bd83f@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
Message-ID: <42F1CC95.9040803@ebi.ac.uk>
markw at illuminae.com wrote:
>>If noone (except of Heiko Dirk and me) want the collections to be
>>changed in the BioMoby API we will have to change the default behaviour
>>of Taverna back to the old one!
>
>
> I think your use-case for wanting the behaviour of collections to change is
> unfounded, since it overloads the intended meaning of the Collection element by
> putting WAAAAAAY too much semantics in it. The use case of having sets of
> multiple alignments represented as Collections of Collections means that the
> inner-most Collection now not only represents a "bag" of things, but a "bag" of
> things that **DO** have a close semantic relationship to each other - these are
> a set of related sequences, and a set of related sequences **is a useful
> Object** in the context of MOBY, could easily trigger the discovery of a
> particular set of services that can specifically operate on sets of closely
> related sequences (versus sets of unrelated sequences), and therefore should be
> given a class of its own (IMO).
True, but you probably also want to allow, for example, a GC percentage
service to run over that set of sets and produce a set of sets of scores
- if the set of sequences becomes some opaque type without this
information then you can't do this. Strikes me that you need a more
sophisticated typing scheme to do this, perhaps extending the Collection
concept to allow the imposition of semantics to the group of child
objects? I think you do need nested collections though, it's not
impossible to deal with (we have some slightly twisted code in Taverna
to do it but it works well enough)
Tom
From rebecca.ernst at gsf.de Thu Aug 4 04:16:35 2005
From: rebecca.ernst at gsf.de (Rebecca Ernst)
Date: Thu, 04 Aug 2005 10:16:35 +0200
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123141715.42f1c853bd83f@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
Message-ID: <42F1CEE3.6020901@gsf.de>
Hi Mark!
I already understood that noone agrees with us ;-)
Funny thing is that everyone keeps talking about collections of
collections which Dirk never wanted.
snip from Dirk:
Correct. That's why we did not ask for nested collections. We suggest to allow
for multiple Simple output as default - which I think is very appropriate
because especially query services will usually return more than one result.
This way we would save the collection concept for semantically related
entities.
I am still not sure if anyone has really understood why we have problems
with the collections as they are used now but we are happy to leave them
as they are .
Although if I think about it I have to ask one more provoking question
( ;-) ) : Why do we have Simples in BioMoby then? I only have VERY few
services that give back simples according to the API. From my point of
view we could also leave Simples away as they make no sense to me. It is
of no valuable information to me if a service gives back exactly three
or an unknown number of AGI codes....
But as I said already - if everyone is happy with the current use of
collections I will be quiet from now on :-)
The problem is then only shiftet to the clients (As seen in Taverna it
leaves a much bigger job for the clients as they will have to split the
collections)
Cheers,
Rebecca
And you guys in Vancouver would rather go to bed now :-)
markw at illuminae.com wrote:
>>If noone (except of Heiko Dirk and me) want the collections to be
>>changed in the BioMoby API we will have to change the default behaviour
>>of Taverna back to the old one!
>>
>>
>
>I think your use-case for wanting the behaviour of collections to change is
>unfounded, since it overloads the intended meaning of the Collection element by
>putting WAAAAAAY too much semantics in it. The use case of having sets of
>multiple alignments represented as Collections of Collections means that the
>inner-most Collection now not only represents a "bag" of things, but a "bag" of
>things that **DO** have a close semantic relationship to each other - these are
>a set of related sequences, and a set of related sequences **is a useful
>Object** in the context of MOBY, could easily trigger the discovery of a
>particular set of services that can specifically operate on sets of closely
>related sequences (versus sets of unrelated sequences), and therefore should be
>given a class of its own (IMO).
>
>Moreover, if we allow collections of collections, we then have a problem of
>recursion - collections of collections of collections of collections of
>collections.... What a nightmare!
>
>The API change is not a significant one, I agree... and in fact, since people
>are already mis-using the messaging format in this way much of the software out
>there can already handle the situation... but given that particular use-case, I
>think you are losing more semantics than you need to be by refusing to create a
>new Object Class.
>
>M
>
>_______________________________________________
>moby-l mailing list
>moby-l at biomoby.org
>http://biomoby.org/mailman/listinfo/moby-l
>
>
>
--
Rebecca Ernst
MIPS, Inst. for Bioinformatics
GSF Research Center for Environment and Health
Ingolstaedter Landstr. 1
85764 Neuherberg
fon: +49 89 3187 3583
email: Rebecca.Ernst at gsf.de
From markw at illuminae.com Thu Aug 4 04:26:14 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu, 4 Aug 2005 04:26:14 -0400
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1CEE3.6020901@gsf.de>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CEE3.6020901@gsf.de>
Message-ID: <1123143974.42f1d1268d031@webmail.illuminae.com>
Quoting Rebecca Ernst :
> And you guys in Vancouver would rather go to bed now :-)
Granted, I would rather be in bed, but neither Eddie nor I are in Vancouver so
it would be inappropriate :-)
we're both just a bit north of you in Manchester!
M
From markw at illuminae.com Thu Aug 4 04:32:30 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu, 4 Aug 2005 04:32:30 -0400
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1CC95.9040803@ebi.ac.uk>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
Message-ID: <1123144350.42f1d29e7978b@webmail.illuminae.com>
Quoting Tom Oinn :
> - if the set of sequences becomes some opaque type without this
> information then you can't do this.
Nothing becomes opaque by putting it in the MOBY data-typing system - it can
only get richer and more meaningful by assigning it a specific class :-)
Generic services that operate on lower ontological classes will still be
discovered and executed, so this use case still fails the test of need.
> Strikes me that you need a more
> sophisticated typing scheme to do this, perhaps extending the Collection
> concept to allow the imposition of semantics to the group of child
> objects?
Creating a new type of element - limitedCollection or something like that -
would be an alternative... or changing the definition of Collection to "zero or
more" rather than "one or moe" would also allow us to terminate the recursion.
> I think you do need nested collections though, it's not
> impossible to deal with (we have some slightly twisted code in Taverna
> to do it but it works well enough)
Probably we do... but I find it hard to believe that we will have sets of sets
of sets of sets where there is no implicit semantic meaning at any level! As
soon as there is semantic meaning, it should be a fully qualified MOBY Object
rather than an arbitrary collection.... IMO :-)
M
>
> Tom
>
From markw at illuminae.com Thu Aug 4 04:41:19 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu, 4 Aug 2005 04:41:19 -0400
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123143974.42f1d1268d031@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CEE3.6020901@gsf.de>
<1123143974.42f1d1268d031@webmail.illuminae.com>
Message-ID: <1123144879.42f1d4af9ab8e@webmail.illuminae.com>
Quoting markw at illuminae.com:
> Granted, I would rather be in bed, but neither Eddie nor I are in Vancouver
> so
> it would be inappropriate :-)
Gosh... I *am* tired!
I didn't mean to imply that....
:-/
M
From rebecca.ernst at gsf.de Thu Aug 4 04:50:34 2005
From: rebecca.ernst at gsf.de (Rebecca Ernst)
Date: Thu, 04 Aug 2005 10:50:34 +0200
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123144879.42f1d4af9ab8e@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CEE3.6020901@gsf.de>
<1123143974.42f1d1268d031@webmail.illuminae.com>
<1123144879.42f1d4af9ab8e@webmail.illuminae.com>
Message-ID: <42F1D6DA.2060200@gsf.de>
Mark wrote:
> Gosh... I *am* tired!
> I didn't mean to imply that....
> :-/
Hehe - it took me a while to get the meaning 'between the lines' ;-)
As you are in Manchestser we have the whole european day to discuss
about collections ;-) With Tom you can even chat sitting together if you
are so close ...
I was thinking about a way to tell you what my problem with this simple
and collection thing is and I'll give it a very last try with a trivial
example:
My favourite example to explain BioMoby services to non- informatics
people is that I am a service provider running a laundry. Incoming
objects are dirty laundry, out go clean clothes.
In your way of interpreting services the man in the laundry gets the
information that you are either always bringing exactly 3 shirts to wash
or you are bringing a not known number of shirts (one time you bring 2,
the next time 15 shirts but it's *always* only shirts)
What I am saying is that it *can't* be of any use to this guy to know
how many shirts you'll bring - he will have to wash all of them, one by one.
But if someone is bringing three complete outfits belonging together (a
jacket, a trouser, a shirt and socks ) I need to take care about this
relationship and put them back together before giving it back.
Does this make sense to you??
Rebecca
markw at illuminae.com wrote:
>Quoting markw at illuminae.com:
>
>
>
>>Granted, I would rather be in bed, but neither Eddie nor I are in Vancouver
>>so
>>it would be inappropriate :-)
>>
>>
>
>Gosh... I *am* tired!
>
>I didn't mean to imply that....
>
>:-/
>
>M
>
>
>
>
>
--
Rebecca Ernst
MIPS, Inst. for Bioinformatics
GSF Research Center for Environment and Health
Ingolstaedter Landstr. 1
85764 Neuherberg
fon: +49 89 3187 3583
email: Rebecca.Ernst at gsf.de
From tmo at ebi.ac.uk Thu Aug 4 05:11:51 2005
From: tmo at ebi.ac.uk (Tom Oinn)
Date: Thu, 04 Aug 2005 10:11:51 +0100
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123144350.42f1d29e7978b@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
<1123144350.42f1d29e7978b@webmail.illuminae.com>
Message-ID: <42F1DBD7.4000700@ebi.ac.uk>
markw at illuminae.com wrote:
> Quoting Tom Oinn :
>
>
>>- if the set of sequences becomes some opaque type without this
>>information then you can't do this.
>
>
> Nothing becomes opaque by putting it in the MOBY data-typing system - it can
> only get richer and more meaningful by assigning it a specific class :-)
> Generic services that operate on lower ontological classes will still be
> discovered and executed, so this use case still fails the test of need.
Okay, so I might be failing to entirely understand the moby typing
system. In the case I describe I need to define a type such that :
1) It can be discovered as 'set of aligned sequences' as a single entity.
2) It can also be discovered as a collection of 'sequence'
As I understood it the collection / simple wasn't part of the ontology
so this isn't possible? Effectively I believe that 'set of aligned
sequences' cannot be made a restriction of 'collection of 'sequence'' as
the collection concept isn't part of the ontology - is this wrong?
Tom
From rebecca.ernst at gsf.de Thu Aug 4 04:58:00 2005
From: rebecca.ernst at gsf.de (Rebecca Ernst)
Date: Thu, 04 Aug 2005 10:58:00 +0200
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42f1c2c1.6263d1d9.1fa1.67e4@mx.gmail.com>
References: <42f1c2c1.6263d1d9.1fa1.67e4@mx.gmail.com>
Message-ID: <42F1D898.7040006@gsf.de>
Hi Eddie!
that's good news! Thanks a lot for this!
Is it possible to update the packaged one also? I don't have a cvs
checkout and I guess most of the other people also use the packages.
Cheers,
Rebecca
Edward Kawas wrote:
>Hi,
>
>I have put the old functionality back into Taverna.
>
>If you build a fresh copy from the cvs, you will notice
>(hopefully) that all is well again.
>
>Eddie
>
>
>
>>-----Original Message-----
>>From: moby-l-bounces at portal.open-bio.org [mailto:moby-l-
>>bounces at portal.open-bio.org] On Behalf Of Martin Senger
>>Sent: Thursday, August 04, 2005 12:22 AM
>>To: Rebecca Ernst
>>Cc: taverna-users at lists.sourceforge.net; mobyl
>>Subject: Re: [MOBY-l] Biomoby and Taverna 1.2
>>
>>Rebecca,
>> I have talked about it with Eddie when I met him in
>>Malaga about
>>two/three weeks ago. And he promised (!) that he was going
>>to put back the
>>old CreateMoby... and ParseMoby... classes. And I have
>>never heard from
>>him since. So it is very good that you started this topic
>>again - perhaps
>>he is listening :-)
>>
>> Cheers,
>> Martin
>>
>>--
>>Martin Senger martin.senger at gmail.com
>>
>>International Rice Research Institute
>>Biometrics and Bioinformatics Unit
>>DAPO BOX 7777, Metro Manila
>>Philippines, phone: +63-580-5600 (ext.2324)
>>
>>_______________________________________________
>>moby-l mailing list
>>moby-l at biomoby.org
>>http://biomoby.org/mailman/listinfo/moby-l
>>
>>
>
>_______________________________________________
>moby-l mailing list
>moby-l at biomoby.org
>http://biomoby.org/mailman/listinfo/moby-l
>
>
>
--
Rebecca Ernst
MIPS, Inst. for Bioinformatics
GSF Research Center for Environment and Health
Ingolstaedter Landstr. 1
85764 Neuherberg
fon: +49 89 3187 3583
email: Rebecca.Ernst at gsf.de
From markw at illuminae.com Thu Aug 4 05:24:08 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu, 4 Aug 2005 05:24:08 -0400
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1DBD7.4000700@ebi.ac.uk>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
<1123144350.42f1d29e7978b@webmail.illuminae.com>
<42F1DBD7.4000700@ebi.ac.uk>
Message-ID: <1123147448.42f1deb87303f@webmail.illuminae.com>
Quoting Tom Oinn :
> 1) It can be discovered as 'set of aligned sequences' as a single entity.
>
> 2) It can also be discovered as a collection of 'sequence'
>
> As I understood it the collection / simple wasn't part of the ontology
> so this isn't possible?
You're correct, and that is a good point!
I think that the semantics you lose by making them a simple bag of sequences is
more important than the flexibility of service-discovery that you gain. This is
why we (Eddie) is working on MOBY Object degradation widgets for Taverna, such
that you strip away the outer "semantic" tag and are left with a collection of
sequences that could then trigger the latter discovery that you mention above.
I don't think we can achieve both situations in one step, unless we defined a
parent class "genericSequenceCollection"... but then we would have to do that
for all Object classes and our bicycles would explode!
M
From markw at illuminae.com Thu Aug 4 05:25:38 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Thu, 4 Aug 2005 05:25:38 -0400
Subject: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1D6DA.2060200@gsf.de>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CEE3.6020901@gsf.de>
<1123143974.42f1d1268d031@webmail.illuminae.com>
<1123144879.42f1d4af9ab8e@webmail.illuminae.com>
<42F1D6DA.2060200@gsf.de>
Message-ID: <1123147538.42f1df121b200@webmail.illuminae.com>
> With Tom you can even chat sitting together if you
> are so close ...
>
England is small... but not THAT small :-)
I'll think about your dirty laundry later - must get some other things done!
M
From tmo at ebi.ac.uk Thu Aug 4 06:03:19 2005
From: tmo at ebi.ac.uk (Tom Oinn)
Date: Thu, 04 Aug 2005 11:03:19 +0100
Subject: [Taverna-users] Re: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <1123147448.42f1deb87303f@webmail.illuminae.com>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
<1123144350.42f1d29e7978b@webmail.illuminae.com>
<42F1DBD7.4000700@ebi.ac.uk>
<1123147448.42f1deb87303f@webmail.illuminae.com>
Message-ID: <42F1E7E7.3060109@ebi.ac.uk>
markw at illuminae.com wrote:
> Quoting Tom Oinn :
>
>
>>1) It can be discovered as 'set of aligned sequences' as a single entity.
>>
>>2) It can also be discovered as a collection of 'sequence'
>>
>>As I understood it the collection / simple wasn't part of the ontology
>>so this isn't possible?
>
> You're correct, and that is a good point!
Nice to know my brain's working again :)
I don't think a decomposition widget is the right way to go - the data
in question _is_ a group of sequences, however that's described. If
you're using RDF to describe the types can't you express the above? I
think there's a misconception that semantic types are the same as
classical object types in e.g. Java.
In the latter case an entity has exactly one type (ignoring things like
interfaces!) and that type then has inheritance and restriction
relationships. This causes issues when you want to use that entity in a
different way - an example of this is in Taverna's code where we have a
type that represents a user interface widget that can bind to a workflow
model but we also know that all such widgets are instances of one of the
core Java ui classes. Because Java doesn't have multiple inheritance or
mixins we can't express this in the Java type system - I believe this is
exactly analogous to the case I described above.
In the case of a semantic description you can make statements describing
the entity, in the current Moby implementation these statements are
pretty much hardcoded as 'entity has type foo', which makes them
directly equivalent to the types used in languages such as Java, with
all the same issues. There's no fundamental reason you need to do this
though, with an extension into generalised descriptions you should be
able to say 'entity is represented by collection of sequences which are
related by...'. The key is to stop thinking of semantic types as being a
lazy way of growing an object heirarchy and actually take advantage of
their full expressive power. I can appreciate a desire to keep things
simple but we have a clear case here where we need something a bit smarter.
Could someone, for example Phil, who actually knows what they're talking
about on these lists perhaps give an example of how this would look in
RDF? I know it's possible to do but a concrete example might be helpful
and I don't have the required knowledge.
Tom
From tmo at ebi.ac.uk Thu Aug 4 06:07:28 2005
From: tmo at ebi.ac.uk (Tom Oinn)
Date: Thu, 04 Aug 2005 11:07:28 +0100
Subject: [Taverna-users] Re: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1E7E7.3060109@ebi.ac.uk>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
<1123144350.42f1d29e7978b@webmail.illuminae.com>
<42F1DBD7.4000700@ebi.ac.uk>
<1123147448.42f1deb87303f@webmail.illuminae.com>
<42F1E7E7.3060109@ebi.ac.uk>
Message-ID: <42F1E8E0.6050102@ebi.ac.uk>
Tom Oinn wrote:
> Could someone, for example Phil, who actually knows what they're talking
> about on these lists perhaps give an example of how this would look in
> RDF?
Um, that wasn't meant to imply that anyone doesn't know what they're
talking about - sounded a bit harsh on re-reading! It would be good to
have some input from the mygrid semantics folks though especially as
this issue is directly related to the description of services and
therefore the mygrid registry activities.
Tom
From p.lord at cs.man.ac.uk Thu Aug 4 12:17:28 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: 04 Aug 2005 17:17:28 +0100
Subject: [Taverna-users] Re: [MOBY-l] Biomoby and Taverna 1.2
In-Reply-To: <42F1E7E7.3060109@ebi.ac.uk>
References: <42F1BF9D.6040401@gsf.de>
<1123141715.42f1c853bd83f@webmail.illuminae.com>
<42F1CC95.9040803@ebi.ac.uk>
<1123144350.42f1d29e7978b@webmail.illuminae.com>
<42F1DBD7.4000700@ebi.ac.uk>
<1123147448.42f1deb87303f@webmail.illuminae.com>
<42F1E7E7.3060109@ebi.ac.uk>
Message-ID:
>>>>> "Tom" == Tom Oinn writes:
Tom> Could someone, for example Phil, who actually knows what
Tom> they're talking about on these lists perhaps give an example of
Tom> how this would look in RDF? I know it's possible to do but a
Tom> concrete example might be helpful and I don't have the required
Tom> knowledge.
I'm not sure that I would make any claims to be the worlds greatest
expert on RDF. Actually, I would make claims to not be any kind of
expert on RDF.
You can certainly express the notion of a collection in RDF. You can
put semantic types onto any of the nodes underneath this, and there is
very little restriction as to what the types of these things can
be. You can state that for a rdf:bag that the first is a dna sequence,
the next a protein or whatever. The rdf looks something like
a ---- has type ---> bag
a ---- li ----> b
a ---- li ----> c
b ---- hastype -----> dna sequence
b ---- hasvalue -----> "GATC"
c ---- hastype ---> protein sequence
c ---- hasvalue ---> "MAGSYDJ"
If you want this level of complexity in moby-s messages then there is
no particular problem within representing it in the XML, at least as
far as I can see. The problem comes from querying, because people are
likely to want to query for services which consume, for example,
"DNASequence, or some collection which recursively contains a DNA
sequence".
Neither moby-central nor feta are going to be terribly happy with
performing this sort of query.
Phil
From bgulden at purdue.edu Thu Aug 4 14:08:12 2005
From: bgulden at purdue.edu (bgulden@purdue.edu)
Date: Thu, 4 Aug 2005 13:08:12 -0500
Subject: [MOBY-l] Submitting Parameters
Message-ID: <1123178892.42f2598c4e9bf@webmail.purdue.edu>
First, I just wanted to thank you all for helping with my previous problems
with parameters (The ->secondary function worked perfectly). Now on to my
current dilemma.
Is there a special way to include parameters in the execution of Moby
services? According to the perl API services with parameters should be called
in this format where input1 is a primary input and param1 is a secondary input:
$Service->execute(XMLinputlist => [
[
'input1', ' ',
'param1', '0.001 ',
]
]);
However, when I check the actual inputted data to my service I end up with one
tag around both input1 and param1.
Thanks,
Brian
From d.haase at gsf.de Fri Aug 5 04:18:41 2005
From: d.haase at gsf.de (Dirk Haase)
Date: Fri, 5 Aug 2005 10:18:41 +0200
Subject: [MOBY-l] Submitting Parameters
In-Reply-To: <1123178892.42f2598c4e9bf@webmail.purdue.edu>
References: <1123178892.42f2598c4e9bf@webmail.purdue.edu>
Message-ID: <200508051018.41494.d.haase@gsf.de>
On Thursday 04 August 2005 20:08, bgulden at purdue.edu wrote:
> First, I just wanted to thank you all for helping with my previous problems
> with parameters (The ->secondary function worked perfectly). Now on to my
> current dilemma.
>
> Is there a special way to include parameters in the execution of Moby
> services? According to the perl API services with parameters should be
> called in this format where input1 is a primary input and param1 is a
> secondary input: $Service->execute(XMLinputlist => [
> [
> 'input1', ' ',
> 'param1', '0.001 ',
> ]
> ]);
>
> However, when I check the actual inputted data to my service I end up with
> one tag around both input1 and param1.
Hmm, your code seems correct, should actually work... Are you sure the service is correctly registered?
Regards,
dirk
--
----------------------------------------------------------
Dirk Haase phone +49 89 3187 3583
http://mips.gsf.de/~haase email d.haase at gsf.de
From markw at illuminae.com Fri Aug 5 04:28:18 2005
From: markw at illuminae.com (markw@illuminae.com)
Date: Fri, 5 Aug 2005 04:28:18 -0400
Subject: [MOBY-l] Submitting Parameters
In-Reply-To: <200508051018.41494.d.haase@gsf.de>
References: <1123178892.42f2598c4e9bf@webmail.purdue.edu>
<200508051018.41494.d.haase@gsf.de>
Message-ID: <1123230498.42f323223100a@webmail.illuminae.com>
There is a known bug in the current production code relating to the registration
of parameters - if any of them have a default, a min, or a max of "0", the
registration of that parameter will fail, but the service registration will
succeed. This might be what he is seeing.
:-/
It is fixed in the new codebase, but we wont be updating the production server
for a week or two.
M
Quoting Dirk Haase :
> On Thursday 04 August 2005 20:08, bgulden at purdue.edu wrote:
> > First, I just wanted to thank you all for helping with my previous problems
> > with parameters (The ->secondary function worked perfectly). Now on to my
> > current dilemma.
> >
> > Is there a special way to include parameters in the execution of Moby
> > services? According to the perl API services with parameters should be
> > called in this format where input1 is a primary input and param1 is a
> > secondary input: $Service->execute(XMLinputlist => [
> > [
> > 'input1', ' id="123"/>',
> > 'param1', '0.001 ',
> > ]
> > ]);
> >
> > However, when I check the actual inputted data to my service I end up with
> > one tag around both input1 and param1.
>
> Hmm, your code seems correct, should actually work... Are you sure the
> service is correctly registered?
>
> Regards,
> dirk
>
> --
>
> ----------------------------------------------------------
> Dirk Haase phone +49 89 3187 3583
> http://mips.gsf.de/~haase email d.haase at gsf.de
> _______________________________________________
> moby-l mailing list
> moby-l at biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
>
From edward.kawas at gmail.com Mon Aug 8 09:42:55 2005
From: edward.kawas at gmail.com (Edward Kawas)
Date: Mon, 8 Aug 2005 06:42:55 -0700
Subject: [MOBY-l] RE: [MOBY-dev] A bug in Taverna 1.2 perhaps related to
BioMOBY
In-Reply-To: <42F3B988.7020506@cnb.uam.es>
Message-ID: <42f76179.195ac951.0c6c.100f@mx.gmail.com>
Hi,
I think that that has been fixed. I actually noticed that
bug when using the Planet workflows. So, like I said, it
should be fixed.
Out of curiosity, did you cvs build or download a
prepackaged Taverna?
Thanks,
Eddie
> -----Original Message-----
> From: moby-dev-bounces at portal.open-bio.org [mailto:moby-
> dev-bounces at portal.open-bio.org] On Behalf Of Jos? Mar?a
> Fern?ndez Gonz?lez
> Sent: Friday, August 05, 2005 12:10 PM
> To: taverna-users at lists.sourceforge.net; moby-
> l at biomoby.org; mobydev; taverna-
> hackers at lists.sourceforge.net
> Subject: [MOBY-dev] A bug in Taverna 1.2 perhaps related
> to BioMOBY
>
> Hi everybody,
> I don't like crossposting, but I don't know if the
> bug I have found
> today is from the Taverna 1.2 core or the MOBY plugin.
>
> The bug is pretty simple: if you create/load a
> workflow which has a
> service whose name in the workflow is different from the
> service name,
> Taverna does not use the official service name when it is
> invoked.
> Instead, it uses the name given to the service in the
> workflow! The same
> workflow works flawlessly when it is loaded and run in
> Taverna 1.1.
>
> I have discovered the bug with a MOBY service (see
> the sample workflow
> ja5.xml, and the reports taverna-1.1-report.xml and
> taverna-1.2-report.xml), so I don't know which has the
> blame...
>
> This bug should arise when one service is used more
> than once, which is
> very common in real workflows!
>
> Best Regards,
> Jos? Mar?a Fern?ndez
> --
> Jos? Mar?a Fern?ndez Gonz?lez e-mail:
> jmfernandez at cnb.uam.es
> Tlfn: (+34) 91 585 54 50 Fax: (+34) 91 585
45
> 06
> Grupo de Dise?o de Proteinas Protein Design Group
> Centro Nacional de Biotecnolog?a National Center of
> Biotechnology
> C.P.: 28049 Zip Code: 28049
> C/. Darwin n? 3 (Campus Cantoblanco, U. Aut?noma), Madrid
> (Spain)
From fgibbons at hms.harvard.edu Wed Aug 10 11:37:14 2005
From: fgibbons at hms.harvard.edu (Frank Gibbons)
Date: Wed, 10 Aug 2005 11:37:14 -0400
Subject: [MOBY-l] Trouble-ticketing for MOBY; unwanted side effect of
MOBY::Client::Service->execute
Message-ID: <5.2.1.1.2.20050810112754.011101f8@email.med.harvard.edu>
Hi,
Two related items:
I think MOBY::Client::Service->execute has an unwanted side-effect in that
the XMLinputlist argument appears to be modified upon returning from this
function. The code does some editing on it. I don't think that would have
been a problem had XMLinputlist been passed in as a scalar, but it's passed
as an ARRAY ref, and so modifications to it are propagated back to the
calling code. This is a problem if, like me, you want to iterate over
services calling each one with the same XMLinputlist. (At least, I think
that's what's happening - correct me.)
How do we fix something like this? Are there regression tests to make sure
nothing gets broken when this issue is tackled? Who fixes it? (I'll
volunteer) The issue of how to fix this raises the other item: while MOBY
is under CVS control, there's no ticketing system (that I'm aware of). As a
long-time lurker on this list, I know that most issues just get posted to
the list, and then fixed. I think there are a lot advantages to having a
centralized, non-email-based record of what needs to be fixed in the
future, but also of what was fixed in the past. I'm wondering partly why
MOBY isn't on SourceForge (which has a ticketing system for bugs and
feature requests, so that potential users can not only see what issues have
been fixed, but also what features are likely to appear in the future) - if
this has been discussed in the past and resolved, I'd appreciate a pointer.
Thanks,
-Frank
PhD, Computational Biologist,
Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
Tel: 617-432-3555 Fax:
617-432-3557 http://llama.med.harvard.edu/~fgibbons
From markw at illuminae.com Thu Aug 11 12:37:03 2005
From: markw at illuminae.com (mark wilkinson)
Date: Thu, 11 Aug 2005 16:37:03 +0000 GMT
Subject: [MOBY-l] Trouble-ticketing for MOBY;
unwanted side effect ofMOBY::Client::Service->execute
Message-ID: <969787678-1123688248-cardhu_blackberry.rim.net-5970-@engine13-cell01>
Hi Frank,
Two quick responses (I'm on the blackberry and it isn't very efficient :-) )
1). The execute method will shortly be significantly changed in light of the upcoming API changes (esp. Insistence on named parameters allows me to use a more sensible "hash" structure in place of the list structure). I don't know if it is worth spending time fixing it.
2). I didn't realize that my code modified the input list. You're right, this is bad. I could post a fix for the such that the input listref is cloned to avoid this (if that is the root of the problem), or you are of course always welcome to fix it yourself and commit.
3). We do have a ticketing system, but nobody ever uses it :-). Click the "bugs" link on the left side of the homepage... I haven't tested it for a while, but it worked last time I tried...
Cheers!
M
-----Original Message-----
From: Frank Gibbons
Date: Wed, 10 Aug 2005 11:37:14
To:moby-l at biomoby.org
Subject: [MOBY-l] Trouble-ticketing for MOBY; unwanted side effect of
MOBY::Client::Service->execute
Hi,
Two related items:
I think MOBY::Client::Service->execute has an unwanted side-effect in that
the XMLinputlist argument appears to be modified upon returning from this
function. The code does some editing on it. I don't think that would have
been a problem had XMLinputlist been passed in as a scalar, but it's passed
as an ARRAY ref, and so modifications to it are propagated back to the
calling code. This is a problem if, like me, you want to iterate over
services calling each one with the same XMLinputlist. (At least, I think
that's what's happening - correct me.)
How do we fix something like this? Are there regression tests to make sure
nothing gets broken when this issue is tackled? Who fixes it? (I'll
volunteer) The issue of how to fix this raises the other item: while MOBY
is under CVS control, there's no ticketing system (that I'm aware of). As a
long-time lurker on this list, I know that most issues just get posted to
the list, and then fixed. I think there are a lot advantages to having a
centralized, non-email-based record of what needs to be fixed in the
future, but also of what was fixed in the past. I'm wondering partly why
MOBY isn't on SourceForge (which has a ticketing system for bugs and
feature requests, so that potential users can not only see what issues have
been fixed, but also what features are likely to appear in the future) - if
this has been discussed in the past and resolved, I'd appreciate a pointer.
Thanks,
-Frank
PhD, Computational Biologist,
Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
Tel: 617-432-3555 Fax:
617-432-3557 http://llama.med.harvard.edu/~fgibbons
_______________________________________________
moby-l mailing list
moby-l at biomoby.org
http://biomoby.org/mailman/listinfo/moby-l
--
Mark Wilkinson
...on the road!
From p.lord at cs.man.ac.uk Mon Aug 15 05:35:58 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: 15 Aug 2005 10:35:58 +0100
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
Message-ID: <87y873v8xt.fsf@rpc71.cs.man.ac.uk>
>>>>> "Boris" == Boris Steipe writes:
Boris> Why not put the burden of the lease on the agent to combine
Boris> the advantages of both models? I.e. if service is down for
Boris> less then a specific time, it might not get deregistered but
Boris> only flagged as temporarily unavailable ... then un-flagged
Boris> as it comes up again, except if it's down for, say > 1week,
Boris> then it gets deregistered.
Boris> $0.02
Actually, this is just a modification of the agent idea, and not a
lease at all.
The idea of a lease is not just a burden on the service provide, but
it also has a positive advantage. It give the service provide the
ability to control how long they want the service advertised for in
the registry. Of course, they can do the same thing with the agent,
more or less; they just remove the RDF document after a certain time,
although, they don't know exactly when the agent will remove them.
At the end of the day, I am not sure that the two mechanisms are
hugely different in terms of their functionality. It seems to me that
the lease puts slightly more weight on the service provider, but will
provide better garbage collection (just become someone leaves an RDF
document on the web, does not mean that their service is actually
still up).
Phil
From p.lord at cs.man.ac.uk Mon Aug 15 05:38:32 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: 15 Aug 2005 10:38:32 +0100
Subject: [MOBY-l] Re: [moby] Re: [MOBY-dev] lease versus agent for registry
updating
In-Reply-To: <1123885110.15335.52.camel@bioinfo.icapture.ubc.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<1123879295.15335.30.camel@bioinfo.icapture.ubc.ca>
<42FD1362.8000107@ucalgary.ca>
<1123885110.15335.52.camel@bioinfo.icapture.ubc.ca>
Message-ID: <87slxbv8tj.fsf@rpc71.cs.man.ac.uk>
>>>>> "Mark" == Mark Wilkinson writes:
Mark> We're certainly of like-mind at this end, though certain
Mark> people ;-) at myGrid think that the agent is the wrong way to
Mark> go... I just don't see it. IMO a leasing model would be akin
Mark> to leasing space in Google!
The lease is the more common idiom, I think. I don't think that there
is a huge difference. My main criticism of the agent is that there
does not seem to be a different connection between the agent poll and
the service being available.
Phil
From tmo at ebi.ac.uk Mon Aug 15 05:51:26 2005
From: tmo at ebi.ac.uk (Tom Oinn)
Date: Mon, 15 Aug 2005 10:51:26 +0100
Subject: [MOBY-l] Re: [moby] Re: [MOBY-dev] lease versus agent for registry
updating
In-Reply-To: <87slxbv8tj.fsf@rpc71.cs.man.ac.uk>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca> <1123879295.15335.30.camel@bioinfo.icapture.ubc.ca> <42FD1362.8000107@ucalgary.ca> <1123885110.15335.52.camel@bioinfo.icapture.ubc.ca>
<87slxbv8tj.fsf@rpc71.cs.man.ac.uk>
Message-ID: <4300659E.3050808@ebi.ac.uk>
Phillip Lord wrote:
>>>>>>"Mark" == Mark Wilkinson writes:
>
>
> Mark> We're certainly of like-mind at this end, though certain
> Mark> people ;-) at myGrid think that the agent is the wrong way to
> Mark> go... I just don't see it. IMO a leasing model would be akin
> Mark> to leasing space in Google!
>
> The lease is the more common idiom, I think. I don't think that there
> is a huge difference. My main criticism of the agent is that there
> does not seem to be a different connection between the agent poll and
> the service being available.
Much as I hate* to agree with Phil I, um, do :) Seems to me that the
agent approach is more part of a monitoring framework than a registry, I
don't know offhand whether we're planning to include metrics such as
service liveness and response times as transient metadata in the
registry but that's where I'd see agents fitting in rather than in the
initial registration.
Tom
* it's all an act but hey...
From p.lord at cs.man.ac.uk Mon Aug 15 06:00:53 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: 15 Aug 2005 11:00:53 +0100
Subject: [MOBY-l] Re: [moby] Re: [MOBY-dev] lease versus agent for
registry updating
In-Reply-To: <4300659E.3050808@ebi.ac.uk>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<1123879295.15335.30.camel@bioinfo.icapture.ubc.ca>
<42FD1362.8000107@ucalgary.ca>
<1123885110.15335.52.camel@bioinfo.icapture.ubc.ca>
<87slxbv8tj.fsf@rpc71.cs.man.ac.uk> <4300659E.3050808@ebi.ac.uk>
Message-ID: <87k6inv7sa.fsf@rpc71.cs.man.ac.uk>
>>>>> "Tom" == Tom Oinn writes:
Tom> Phillip Lord wrote:
>>>>>>> "Mark" == Mark Wilkinson writes:
Mark> We're certainly of like-mind at this end, though certain
Mark> people ;-) at myGrid think that the agent is the wrong way to
Mark> go... I just don't see it. IMO a leasing model would be akin
Mark> to leasing space in Google!
>> The lease is the more common idiom, I think. I don't think that
>> there is a huge difference. My main criticism of the agent is
>> that there does not seem to be a different connection between the
>> agent poll and the service being available.
Tom> Much as I hate* to agree with Phil I, um, do :) Seems to me
Tom> that the agent approach is more part of a monitoring framework
It might be if it monitored the service rather than something else!
Tom> than a registry, I don't know offhand whether we're planning to
Tom> include metrics such as service liveness and response times as
Tom> transient metadata in the registry but that's where I'd see
Tom> agents fitting in rather than in the initial registration.
The current approach that feta takes to the "registry" is that it is a
very dumb thing. It really only uses it for storing files (which is
why it uses a webdav backend by default).
It would be lovely to replace or augment this with something
richer, which understood liveness and response. Or, alternatively,
have a modular search system which could combine responses from feta
and some arbitrary service search system based on information about
liveness.
It's all very feasible. Even a simple minded response time data model
would help a lot.
Phil
From gordonp at ucalgary.ca Mon Aug 15 10:32:26 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Mon, 15 Aug 2005 08:32:26 -0600
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <87y873v8xt.fsf@rpc71.cs.man.ac.uk>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk>
Message-ID: <4300A77A.8030806@ucalgary.ca>
A few advantages of the lease are not really advantages in practice:
1. No one will manually update their lease, they will put it in a cron
job. Therefore you either need to edit your crontab, or remove from the
RDF file in the agent case. Both are just as likely to be ignored by an
administrator when a service stops working.
2. Specifiying that the service is valid for a certain amount of time is
mostly useful for testing purposes. You should really be using a test
registry in that case, not posting dubious services to the production
server.
The chief advantages of the agent are:
1. You can trace the registration and deregistration of a service to a
particular domain name. It's not great security, but at the very least
people require some serious work to pose as the NCBI on purpose (by
hacking their Web domain), and cannot by mistake (e.g. "I registered my
service using the NCBI authority ID because I'm using gi's").
2. The RDF for the services does not have a single point of failure
(i.e. the central registry).
The one agent feature I would l;ike though is that I can call MOBY
Central to tell it that I've changed my RDF, i.e. pushing a
refresh. It's not critical though. If the agent runs once a day, you may
get some latency on bad services, but it's not the end of the world.
Tim Berners-Lee got a lot of flack for his Web idea because it didn't
enforce that what people linked to existed. "People won't use it, they
may end up at dead links!" they said...
>>>>>>"Boris" == Boris Steipe writes:
>>>>>>
>>>>>>
>
> Boris> Why not put the burden of the lease on the agent to combine
> Boris> the advantages of both models? I.e. if service is down for
> Boris> less then a specific time, it might not get deregistered but
> Boris> only flagged as temporarily unavailable ... then un-flagged
> Boris> as it comes up again, except if it's down for, say > 1week,
> Boris> then it gets deregistered.
>
> Boris> $0.02
>
>
>Actually, this is just a modification of the agent idea, and not a
>lease at all.
>
>The idea of a lease is not just a burden on the service provide, but
>it also has a positive advantage. It give the service provide the
>ability to control how long they want the service advertised for in
>the registry. Of course, they can do the same thing with the agent,
>more or less; they just remove the RDF document after a certain time,
>although, they don't know exactly when the agent will remove them.
>
>At the end of the day, I am not sure that the two mechanisms are
>hugely different in terms of their functionality. It seems to me that
>the lease puts slightly more weight on the service provider, but will
>provide better garbage collection (just become someone leaves an RDF
>document on the web, does not mean that their service is actually
>still up).
>
>Phil
>
>_______________________________________________
>moby-l mailing list
>moby-l at biomoby.org
>http://biomoby.org/mailman/listinfo/moby-l
>
>
>
From jmfernandez at cnb.uam.es Fri Aug 5 15:10:00 2005
From: jmfernandez at cnb.uam.es (=?ISO-8859-1?Q?Jos=E9_Mar=EDa_Fern=E1ndez_Gonz=E1lez?=)
Date: Fri, 05 Aug 2005 21:10:00 +0200
Subject: [MOBY-l] A bug in Taverna 1.2 perhaps related to BioMOBY
Message-ID: <42F3B988.7020506@cnb.uam.es>
Hi everybody,
I don't like crossposting, but I don't know if the bug I have found
today is from the Taverna 1.2 core or the MOBY plugin.
The bug is pretty simple: if you create/load a workflow which has a
service whose name in the workflow is different from the service name,
Taverna does not use the official service name when it is invoked.
Instead, it uses the name given to the service in the workflow! The same
workflow works flawlessly when it is loaded and run in Taverna 1.1.
I have discovered the bug with a MOBY service (see the sample workflow
ja5.xml, and the reports taverna-1.1-report.xml and
taverna-1.2-report.xml), so I don't know which has the blame...
This bug should arise when one service is used more than once, which is
very common in real workflows!
Best Regards,
Jos? Mar?a Fern?ndez
--
Jos? Mar?a Fern?ndez Gonz?lez e-mail: jmfernandez at cnb.uam.es
Tlfn: (+34) 91 585 54 50 Fax: (+34) 91 585 45 06
Grupo de Dise?o de Proteinas Protein Design Group
Centro Nacional de Biotecnolog?a National Center of Biotechnology
C.P.: 28049 Zip Code: 28049
C/. Darwin n? 3 (Campus Cantoblanco, U. Aut?noma), Madrid (Spain)
From p.lord at cs.man.ac.uk Tue Aug 16 06:02:52 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: 16 Aug 2005 11:02:52 +0100
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <4300A77A.8030806@ucalgary.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk> <4300A77A.8030806@ucalgary.ca>
Message-ID: <87mzni8aib.fsf@rpc71.cs.man.ac.uk>
>>>>> "Paul" == Paul Gordon writes:
Paul> A few advantages of the lease are not really advantages in
Paul> practice:
Paul> 1. No one will manually update their lease, they will put it
Paul> in a cron
Paul> job. Therefore you either need to edit your crontab, or
Paul> remove from the RDF file in the agent case. Both are just as
Paul> likely to be ignored by an administrator when a service stops
Paul> working.
This is a necessary problem with neither the agent nor the lease. It
is a current problem with the implementation of the agent. All you do
is add a call back to the service.
Paul> 2. Specifiying that the service is valid for a certain amount
Paul> of time
Paul> is mostly useful for testing purposes. You should really be
Paul> using a test registry in that case, not posting dubious
Paul> services to the production server.
Not really. Take, for example, the SOAPLAB services Martin developed
as part of mygrid. When the toolkit was first produced, it was
deployed on our own servers; now it is supported by EBI's external
services. Clearly, the latter is likely to produce a better uptime
than the former. Hence, EBI might use a month long lease.
Neither of these services were testing. But they did come with
different levels of support.
Paul> The chief advantages of the agent are:
Paul> 1. You can trace the registration and deregistration of a
Paul> service to
Paul> a particular domain name. It's not great security, but at the
Paul> very least people require some serious work to pose as the
Paul> NCBI on purpose (by hacking their Web domain), and cannot by
Paul> mistake (e.g. "I registered my service using the NCBI
Paul> authority ID because I'm using gi's").
Likewise with a lease call back system. Anyway, can you really not
determine the start point of a web services call?
Paul> 2. The RDF for the services does not have a single point of
Paul> failure
Paul> (i.e. the central registry).
This is also untrue. There is nothing to stop a lease percolating
through a set of federated registries. With the agent, you have to
percolate the registered URL's (or the RDF) in the same way.
Ultimately, as two systems are doing similar things. It's just that
one is push and the other pull.
Paul> The one agent feature I would l;ike though is that I can call
Paul> MOBY Central to tell it that I've changed my RDF, i.e. pushing
Paul> a refresh. It's not critical though. If the agent runs once a
Paul> day, you may get some latency on bad services, but it's not
Paul> the end of the world. Tim Berners-Lee got a lot of flack for
Paul> his Web idea because it didn't enforce that what people linked
Paul> to existed. "People won't use it, they may end up at dead
Paul> links!" they said...
You still want people to be able to search on the freshness of
information though. With a lease, you can add queries to moby-central
to say "give me only services with a current lease"--after all the
registry is not required to deregister a service when it's lease runs
out.
Cheers
Phil
From gordonp at ucalgary.ca Tue Aug 16 09:51:08 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Tue, 16 Aug 2005 07:51:08 -0600
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <87mzni8aib.fsf@rpc71.cs.man.ac.uk>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca> <87y873v8xt.fsf@rpc71.cs.man.ac.uk>
<4300A77A.8030806@ucalgary.ca> <87mzni8aib.fsf@rpc71.cs.man.ac.uk>
Message-ID: <4301EF4C.5020506@ucalgary.ca>
My CAN$0.02 :-)
>
> Paul> A few advantages of the lease are not really advantages in
> Paul> practice:
>
> Paul> 1. No one will manually update their lease, they will put it
> Paul> in a cron
> Paul> job. Therefore you either need to edit your crontab, or
> Paul> remove from the RDF file in the agent case. Both are just as
> Paul> likely to be ignored by an administrator when a service stops
> Paul> working.
>
>
>This is a necessary problem with neither the agent nor the lease. It
>is a current problem with the implementation of the agent. All you do
>is add a call back to the service.
>
>
This turns into an issue with the protocol, we would need to require
(when you register a service) test cases that return predictable,
non-null value. This was discussed at the last MOBY meeting, but I
don't think anything came of it. Or am I wrong?
> Paul> The chief advantages of the agent are:
>
> Paul> 1. You can trace the registration and deregistration of a
> Paul> service to
> Paul> a particular domain name. It's not great security, but at the
> Paul> very least people require some serious work to pose as the
> Paul> NCBI on purpose (by hacking their Web domain), and cannot by
> Paul> mistake (e.g. "I registered my service using the NCBI
> Paul> authority ID because I'm using gi's").
>
>Likewise with a lease call back system.
>
Like you said, the two approches aren't so different, because fetching
the RDF from the server is a callback of sorts.
>Anyway, can you really not
>determine the start point of a web services call?
>
>
If there are proxies, multiple host names for a machine (e.g. the host
I'm on right now is www.visualgenomics.ca, moby.ucalgary.ca,
www.gcbioinformatics.ca, and about 6 other names), it can get complicated.
> Paul> 2. The RDF for the services does not have a single point of
> Paul> failure
> Paul> (i.e. the central registry).
>
>This is also untrue. There is nothing to stop a lease percolating
>through a set of federated registries. With the agent, you have to
>percolate the registered URL's (or the RDF) in the same way.
>
>
True, but the mechanism to exchange a URL list could be simpler than
mirroring the entire registry, and is less subject to error propagation
(e.g. don't blindly copy a goof in the main mirror's database system
because it had network connectivity issues, fetch the RDF yourself) .
If we're lucky, people will try to outcompete each other creating the
most robust registry :-)
>Ultimately, as two systems are doing similar things. It's just that
>one is push and the other pull.
>
>
> Paul> The one agent feature I would l;ike though is that I can call
> Paul> MOBY Central to tell it that I've changed my RDF, i.e. pushing
> Paul> a refresh. It's not critical though. If the agent runs once a
> Paul> day, you may get some latency on bad services, but it's not
> Paul> the end of the world. Tim Berners-Lee got a lot of flack for
> Paul> his Web idea because it didn't enforce that what people linked
> Paul> to existed. "People won't use it, they may end up at dead
> Paul> links!" they said...
>
>
>You still want people to be able to search on the freshness of
>information though. With a lease, you can add queries to moby-central
>to say "give me only services with a current lease"--after all the
>registry is not required to deregister a service when it's lease runs
>out.
>
>
Would we set a maximum lease? I'd probably give myself a 1 year lease
so that I wouldn't be bothered again. That doesn't make me check that
the service works throughout the year though. IMHO, having a test case
checked regularly is what really helps here, not the lease. A lease is
someone's word, a test case is action, and actions speak louder than
words :-)
>Cheers
>
>Phil
>
>_______________________________________________
>moby-l mailing list
>moby-l at biomoby.org
>http://biomoby.org/mailman/listinfo/moby-l
>
>
>
From edward.kawas at gmail.com Tue Aug 16 12:26:42 2005
From: edward.kawas at gmail.com (Eddie Kawas)
Date: Tue, 16 Aug 2005 11:26:42 -0500
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <4301EF4C.5020506@ucalgary.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk> <4300A77A.8030806@ucalgary.ca>
<87mzni8aib.fsf@rpc71.cs.man.ac.uk> <4301EF4C.5020506@ucalgary.ca>
Message-ID:
I just wanted to say that I agree with Paul.
I think that when a provider registers a service, the provider enters
a url that causes the service to return something that is commonly
returned by all services that states that the service is running. If
this return value is infact returned, then the service signature is
processed. Otherwise, the offending service is recorded and the
service is not shown to the public until the test passes. If after x
times the test fails, then it is permanently removed.
Eddie
On 8/16/05, Paul Gordon wrote:
> My CAN$0.02 :-)
>
> >
> > Paul> A few advantages of the lease are not really advantages in
> > Paul> practice:
> >
> > Paul> 1. No one will manually update their lease, they will put it
> > Paul> in a cron
> > Paul> job. Therefore you either need to edit your crontab, or
> > Paul> remove from the RDF file in the agent case. Both are just as
> > Paul> likely to be ignored by an administrator when a service stops
> > Paul> working.
> >
> >
> >This is a necessary problem with neither the agent nor the lease. It
> >is a current problem with the implementation of the agent. All you do
> >is add a call back to the service.
> >
> >
> This turns into an issue with the protocol, we would need to require
> (when you register a service) test cases that return predictable,
> non-null value. This was discussed at the last MOBY meeting, but I
> don't think anything came of it. Or am I wrong?
>
> > Paul> The chief advantages of the agent are:
> >
> > Paul> 1. You can trace the registration and deregistration of a
> > Paul> service to
> > Paul> a particular domain name. It's not great security, but at the
> > Paul> very least people require some serious work to pose as the
> > Paul> NCBI on purpose (by hacking their Web domain), and cannot by
> > Paul> mistake (e.g. "I registered my service using the NCBI
> > Paul> authority ID because I'm using gi's").
> >
> >Likewise with a lease call back system.
> >
> Like you said, the two approches aren't so different, because fetching
> the RDF from the server is a callback of sorts.
>
> >Anyway, can you really not
> >determine the start point of a web services call?
> >
> >
> If there are proxies, multiple host names for a machine (e.g. the host
> I'm on right now is www.visualgenomics.ca, moby.ucalgary.ca,
> www.gcbioinformatics.ca, and about 6 other names), it can get complicated.
>
> > Paul> 2. The RDF for the services does not have a single point of
> > Paul> failure
> > Paul> (i.e. the central registry).
> >
> >This is also untrue. There is nothing to stop a lease percolating
> >through a set of federated registries. With the agent, you have to
> >percolate the registered URL's (or the RDF) in the same way.
> >
> >
> True, but the mechanism to exchange a URL list could be simpler than
> mirroring the entire registry, and is less subject to error propagation
> (e.g. don't blindly copy a goof in the main mirror's database system
> because it had network connectivity issues, fetch the RDF yourself) .
> If we're lucky, people will try to outcompete each other creating the
> most robust registry :-)
>
> >Ultimately, as two systems are doing similar things. It's just that
> >one is push and the other pull.
> >
> >
> > Paul> The one agent feature I would l;ike though is that I can call
> > Paul> MOBY Central to tell it that I've changed my RDF, i.e. pushing
> > Paul> a refresh. It's not critical though. If the agent runs once a
> > Paul> day, you may get some latency on bad services, but it's not
> > Paul> the end of the world. Tim Berners-Lee got a lot of flack for
> > Paul> his Web idea because it didn't enforce that what people linked
> > Paul> to existed. "People won't use it, they may end up at dead
> > Paul> links!" they said...
> >
> >
> >You still want people to be able to search on the freshness of
> >information though. With a lease, you can add queries to moby-central
> >to say "give me only services with a current lease"--after all the
> >registry is not required to deregister a service when it's lease runs
> >out.
> >
> >
> Would we set a maximum lease? I'd probably give myself a 1 year lease
> so that I wouldn't be bothered again. That doesn't make me check that
> the service works throughout the year though. IMHO, having a test case
> checked regularly is what really helps here, not the lease. A lease is
> someone's word, a test case is action, and actions speak louder than
> words :-)
>
> >Cheers
> >
> >Phil
> >
> >_______________________________________________
> >moby-l mailing list
> >moby-l at biomoby.org
> >http://biomoby.org/mailman/listinfo/moby-l
> >
> >
> >
>
> _______________________________________________
> moby-l mailing list
> moby-l at biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
>
From p.lord at cs.man.ac.uk Tue Aug 16 13:15:29 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: 16 Aug 2005 18:15:29 +0100
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <4301EF4C.5020506@ucalgary.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk> <4300A77A.8030806@ucalgary.ca>
<87mzni8aib.fsf@rpc71.cs.man.ac.uk> <4301EF4C.5020506@ucalgary.ca>
Message-ID: <87ek8tpzv2.fsf@rpc71.cs.man.ac.uk>
>>>>> "Paul" == Paul Gordon writes:
>> This is a necessary problem with neither the agent nor the
>> lease. It is a current problem with the implementation of the
>> agent. All you do is add a call back to the service.
Paul> This turns into an issue with the protocol, we would need to
Paul> require (when you register a service) test cases that return
Paul> predictable, non-null value. This was discussed at the last
Paul> MOBY meeting, but I don't think anything came of it. Or am I
Paul> wrong?
I can't comment on this really, because I wasn't there!
Paul> 1. You can trace the registration and deregistration of a
Paul> service to a particular domain name. It's not great security,
Paul> but at the very least people require some serious work to pose
Paul> as the NCBI on purpose (by hacking their Web domain), and
Paul> cannot by mistake (e.g. "I registered my service using the
Paul> NCBI authority ID because I'm using gi's").
>>
>> Likewise with a lease call back system.
Paul> Like you said, the two approches aren't so different, because
Paul> fetching the RDF from the server is a callback of sorts.
In a sense. It's just that you don't know when it is going to
happen. This is the point with a lease. The client requests from the
server a time limited registration. Is it possible for the client to
query MOBY-Central and ask how often the RDF agent will call? Is it
possible to request that it happens more frequently or less?
>> Anyway, can you really not determine the start point of a web
>> services call?
Paul> If there are proxies, multiple host names for a machine
Paul> (e.g. the host I'm on right now is www.visualgenomics.ca,
Paul> moby.ucalgary.ca, www.gcbioinformatics.ca, and about 6 other
Paul> names), it can get complicated.
Or 136.159.169.6 or h6.net169.ucalgary.ca. The IP number is unique
though.
Paul> 2. The RDF for the services does not have a single point of
Paul> failure (i.e. the central registry).
>>
>> This is also untrue. There is nothing to stop a lease percolating
>> through a set of federated registries. With the agent, you have
>> to percolate the registered URL's (or the RDF) in the same way.
Paul> True, but the mechanism to exchange a URL list could be
Paul> simpler than mirroring the entire registry, and is less
Paul> subject to error propagation (e.g. don't blindly copy a goof
Paul> in the main mirror's database system because it had network
Paul> connectivity issues, fetch the RDF yourself) . If we're
Paul> lucky, people will try to outcompete each other creating the
Paul> most robust registry :-)
Who says that the lease is not encoded as a URL? Besides which, the
lease mechanism might federate by passing around data. Or the client
might explicitly register in more than one registry. Again, I don't
think that this is a necessary advantage of either a lease or the
agent.
>> You still want people to be able to search on the freshness of
>> information though. With a lease, you can add queries to
>> moby-central to say "give me only services with a current
>> lease"--after all the registry is not required to deregister a
>> service when it's lease runs out.
Paul> Would we set a maximum lease? I'd probably give myself a 1
Paul> year lease so that I wouldn't be bothered again.
Yes, this is part of the basic idea behind a lease. The registry overs
a specific contract which describes what it is prepared to do for
you.
Paul> That doesn't make me check that the service works throughout
Paul> the year though. IMHO, having a test case checked regularly is
Paul> what really helps here, not the lease. A lease is someone's
Paul> word, a test case is action, and actions speak louder than
Paul> words :-)
As Tom says, service liveness or quality data is not the same thing as
registration, but an orthogonal issue to it. Having service quality
data would be a nice addition to moby-central.
Phil
From goodb at interchange.ubc.ca Tue Aug 16 13:18:47 2005
From: goodb at interchange.ubc.ca (Benjamin Good)
Date: Tue, 16 Aug 2005 10:18:47 -0700
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To:
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk> <4300A77A.8030806@ucalgary.ca>
<87mzni8aib.fsf@rpc71.cs.man.ac.uk> <4301EF4C.5020506@ucalgary.ca>
Message-ID: <4784EB52-25CB-4260-A779-72918248ECB4@interchange.ubc.ca>
Just to add my 2c. The idea of adding a test case to the
registration phase did come up at the meeting, but I'm not sure how
far its gotten?
I would suggest that this test actually be an instance of the moby
datatype registered as the input of the service - instead of a URL.
This would let any moby-capable client test the service whenever
necessary using exactly the same protocol required to run the service
in the usual case. Having this possibility makes it feasible for
client providers to filter out inconsistent services without relying
entirely on the registry to do so. This enables a runtime check on
the service that seems impossible using only the registry. The agent
could then use this same check to make its decision about whether to
keep the service listed or not..
- Ben
On Aug 16, 2005, at 9:26 AM, Eddie Kawas wrote:
> I just wanted to say that I agree with Paul.
>
> I think that when a provider registers a service, the provider enters
> a url that causes the service to return something that is commonly
> returned by all services that states that the service is running. If
> this return value is infact returned, then the service signature is
> processed. Otherwise, the offending service is recorded and the
> service is not shown to the public until the test passes. If after x
> times the test fails, then it is permanently removed.
>
> Eddie
>
>
> On 8/16/05, Paul Gordon wrote:
>
>> My CAN$0.02 :-)
>>
>>
>>>
>>> Paul> A few advantages of the lease are not really advantages in
>>> Paul> practice:
>>>
>>> Paul> 1. No one will manually update their lease, they will put it
>>> Paul> in a cron
>>> Paul> job. Therefore you either need to edit your crontab, or
>>> Paul> remove from the RDF file in the agent case. Both are just as
>>> Paul> likely to be ignored by an administrator when a service stops
>>> Paul> working.
>>>
>>>
>>> This is a necessary problem with neither the agent nor the lease. It
>>> is a current problem with the implementation of the agent. All
>>> you do
>>> is add a call back to the service.
>>>
>>>
>>>
>> This turns into an issue with the protocol, we would need to require
>> (when you register a service) test cases that return predictable,
>> non-null value. This was discussed at the last MOBY meeting, but I
>> don't think anything came of it. Or am I wrong?
>>
>>
>>> Paul> The chief advantages of the agent are:
>>>
>>> Paul> 1. You can trace the registration and deregistration of a
>>> Paul> service to
>>> Paul> a particular domain name. It's not great security, but at
>>> the
>>> Paul> very least people require some serious work to pose as the
>>> Paul> NCBI on purpose (by hacking their Web domain), and cannot by
>>> Paul> mistake (e.g. "I registered my service using the NCBI
>>> Paul> authority ID because I'm using gi's").
>>>
>>> Likewise with a lease call back system.
>>>
>>>
>> Like you said, the two approches aren't so different, because
>> fetching
>> the RDF from the server is a callback of sorts.
>>
>>
>>> Anyway, can you really not
>>> determine the start point of a web services call?
>>>
>>>
>>>
>> If there are proxies, multiple host names for a machine (e.g. the
>> host
>> I'm on right now is www.visualgenomics.ca, moby.ucalgary.ca,
>> www.gcbioinformatics.ca, and about 6 other names), it can get
>> complicated.
>>
>>
>>> Paul> 2. The RDF for the services does not have a single point of
>>> Paul> failure
>>> Paul> (i.e. the central registry).
>>>
>>> This is also untrue. There is nothing to stop a lease percolating
>>> through a set of federated registries. With the agent, you have to
>>> percolate the registered URL's (or the RDF) in the same way.
>>>
>>>
>>>
>> True, but the mechanism to exchange a URL list could be simpler than
>> mirroring the entire registry, and is less subject to error
>> propagation
>> (e.g. don't blindly copy a goof in the main mirror's database system
>> because it had network connectivity issues, fetch the RDF yourself) .
>> If we're lucky, people will try to outcompete each other creating the
>> most robust registry :-)
>>
>>
>>> Ultimately, as two systems are doing similar things. It's just that
>>> one is push and the other pull.
>>>
>>>
>>> Paul> The one agent feature I would l;ike though is that I can call
>>> Paul> MOBY Central to tell it that I've changed my RDF, i.e.
>>> pushing
>>> Paul> a refresh. It's not critical though. If the agent runs once a
>>> Paul> day, you may get some latency on bad services, but it's not
>>> Paul> the end of the world. Tim Berners-Lee got a lot of flack for
>>> Paul> his Web idea because it didn't enforce that what people
>>> linked
>>> Paul> to existed. "People won't use it, they may end up at dead
>>> Paul> links!" they said...
>>>
>>>
>>> You still want people to be able to search on the freshness of
>>> information though. With a lease, you can add queries to moby-
>>> central
>>> to say "give me only services with a current lease"--after all the
>>> registry is not required to deregister a service when it's lease
>>> runs
>>> out.
>>>
>>>
>>>
>> Would we set a maximum lease? I'd probably give myself a 1 year
>> lease
>> so that I wouldn't be bothered again. That doesn't make me check
>> that
>> the service works throughout the year though. IMHO, having a test
>> case
>> checked regularly is what really helps here, not the lease. A
>> lease is
>> someone's word, a test case is action, and actions speak louder than
>> words :-)
>>
>>
>>> Cheers
>>>
>>> Phil
>>>
>>> _______________________________________________
>>> moby-l mailing list
>>> moby-l at biomoby.org
>>> http://biomoby.org/mailman/listinfo/moby-l
>>>
>>>
>>>
>>>
>>
>> _______________________________________________
>> moby-l mailing list
>> moby-l at biomoby.org
>> http://biomoby.org/mailman/listinfo/moby-l
>>
>>
>
> _______________________________________________
> moby-l mailing list
> moby-l at biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
>
>
From markw at illuminae.com Tue Aug 16 13:36:25 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Tue, 16 Aug 2005 10:36:25 -0700
Subject: [moby] Re: [MOBY-l] Re: [MOBY-dev] lease versus agent for
registry updating
In-Reply-To: <4784EB52-25CB-4260-A779-72918248ECB4@interchange.ubc.ca>
References: <6241454D-0B6A-11DA-A8DF-000A9577512E@utoronto.ca>
<87y873v8xt.fsf@rpc71.cs.man.ac.uk> <4300A77A.8030806@ucalgary.ca>
<87mzni8aib.fsf@rpc71.cs.man.ac.uk> <4301EF4C.5020506@ucalgary.ca>
<4784EB52-25CB-4260-A779-72918248ECB4@interchange.ubc.ca>
Message-ID: <1124213785.18621.138.camel@bioinfo.icapture.ubc.ca>
On Tue, 2005-08-16 at 10:18 -0700, Benjamin Good wrote:
> Just to add my 2c. The idea of adding a test case to the
> registration phase did come up at the meeting, but I'm not sure how
> far its gotten?
It was decided that test cases would be provided as optional additional
metadata attached to the RDF Service Signature on the service-providers
site, identified by a set of (yet to be defined) RDF predicates. Test
cases would *not* be part of the data stored in the registry, but would
be discoverable through e.g. LSID resolution.
Since we are currently in the process of adapting our Service Signature
data model to mimic that of myGrid (and v.v. :-) ) we have not yet
implemented this particular decision. Hopefully this data model will be
firmed-up in the next couple of weeks (that was why Eddie and I were in
Manchester last week), and then we will be able to generate a
"canonical" Service Signature for everyone's existing services, as well
as generate the signature on-the-fly for new registrations.
M
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From senger at ebi.ac.uk Tue Aug 16 20:10:24 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Wed, 17 Aug 2005 01:10:24 +0100 (BST)
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <4784EB52-25CB-4260-A779-72918248ECB4@interchange.ubc.ca>
Message-ID:
Hi all,
Everybody is talking about lease and agents so if I do not join I would
feel like a pariah...
What was the original problem that started this discussion? I remember
two:
a) How to make a deregistration secure (so nobody can easily remove my
service from a registry), and
b) How I can remove my service registration at any time when I *want*
to do it.
Am I right - is this (at least) the problems we are trying to solve? (I
understand that during this discussion a new problem was born - how to
test that a service is reliable; but I consider it as a separate issue.)
Obviously problem b) requires an action on my side.
In case of agent it is a removal of an RDF document - and hopefully
supported by a new method in the registry API calling an agent to come to
me *now*. This new API call was mentioned and I hope it will happen.
Without it the agent solution is weak (IMHO) - having the similar weakness
as Google has with non existing links/documents (oh, wow, now I really
feel like a pariah by saying that Google has a weakness!)
In case of lease - is there a concept of making a lease to expire
before its planned end? If so, how is it done that nobody else can do it
except me (problem a))?
Can you, Phill perhaps, explain me if the lease can solve my two
problems above? So far I understand that an agent can - assuming tghat
there will be that new API call (Mark, Eddie, will this be - in case the
agent idea is accepted?)
Cheers,
Martin
--
Martin Senger
email: martin.senger at gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From p.lord at cs.man.ac.uk Wed Aug 17 06:23:18 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: 17 Aug 2005 11:23:18 +0100
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To:
References:
Message-ID: <87vf246ew9.fsf@rpc71.cs.man.ac.uk>
>>>>> "Martin" == Martin Senger writes:
Martin> Hi all,
Martin> Everybody is talking about lease and agents so if I do
Martin> not join I would
Martin> feel like a pariah...
Martin> What was the original problem that started this
Martin> discussion? I remember
Martin> two:
Martin> a) How to make a deregistration secure (so nobody can easily
Martin> remove my
Martin> service from a registry), and
Martin> b) How I can remove my service registration at any time when
Martin> I *want*
Martin> to do it.
Martin> Am I right - is this (at least) the problems we are
Martin> trying to solve? (I
Martin> understand that during this discussion a new problem was
Martin> born - how to test that a service is reliable; but I
Martin> consider it as a separate issue.)
Martin> Obviously problem b) requires an action on my side. In
Martin> case of agent it is a removal of an RDF document - and
Martin> hopefully
Martin> supported by a new method in the registry API calling an
Martin> agent to come to me *now*. This new API call was mentioned
Martin> and I hope it will happen.
Actually, there is a minor security problem here. If I call this
method repeatedly and it really happens *now*, then I can use
moby-central as a proxy to launch a denial of service attack on a
third party. If I did this on NCBI, for example, it might have
interesting repercussions for who ever is hosting the moby-central.
Martin> Without it the agent solution is weak (IMHO) - having the
Martin> similar weakness as Google has with non existing
Martin> links/documents (oh, wow, now I really feel like a pariah by
Martin> saying that Google has a weakness!)
Martin> In case of lease - is there a concept of making a lease
Martin> to expire
Martin> before its planned end? If so, how is it done that nobody
Martin> else can do it except me (problem a))?
Not really sure on this one, to be honest. Clearly, you can just set
up a short lease in the first place. Alternatively, you pass back some
magic cookie when the lease is taken out in the first place, that is
required for deregistration. This enables deregistration to be done
from anywhere, rather than through direct control of the RDF URL, but
still only by the originator.
The ideal would, of course, to have some off the shelf library for
doing leases.
Martin> Can you, Phill perhaps, explain me if the lease can solve
Martin> my two
Martin> problems above? So far I understand that an agent can -
Martin> assuming tghat there will be that new API call (Mark, Eddie,
Martin> will this be - in case the agent idea is accepted?)
Does this make sense?
Phil
From senger at ebi.ac.uk Wed Aug 17 10:57:47 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Wed, 17 Aug 2005 15:57:47 +0100 (BST)
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <87vf246ew9.fsf@rpc71.cs.man.ac.uk>
Message-ID:
> Actually, there is a minor security problem here. If I call this
> method repeatedly and it really happens *now*
>
You are exaggerating (how to spell this word?). The agent can be clever
enough not to go really *now" but only in a reasonably but short
intervals (once a minute maximum, for example).
> Not really sure on this one, to be honest. Clearly, you can just set
> up a short lease in the first place.
>
Yes, that make sense. Or perhaps, I could re-register again with a
different lease, right?
> Alternatively, you pass back some magic cookie when the lease is taken
> out in the first place, that is required for deregistration.
>
Well, this is exactly what biomoby has now - and nobody seems to like
it. That's why we are talking how to change it.
Thanks, Phill, for claryfying the issue (at least in my mind). I feel
now that a lease gives me more control (if and only if I can re-register
with a different lease anytime). And an agent plays a role of an
independent observer that can stop my registration if it finds that there
is something wrong with it (not running, not found RDF etc.).
But the problem is still with the "if and only if" in the previous
paragraph I think. If I can re-register and change the lease, probably
anybody can do it - and to remove/hijack my registration. And we are back
where we started....
So the above would work only if I can prove who I am. And the only way
how to do it (if we do not want to introduce passwords etc.) is to let
somebody come back to my place (because only I can control such a
place). That means an agent...
So: if I use a lease I am in charge of registering my services. If it
is a new registration (such service does not exist yet), everything is
fine and nobody needs to go to me (but can, and probably will, but there
is no hurry). If I try to re-register (because I wish to change the lease
period), then an agent must come to me to see that my RDF document is
still there (with a new lease perion inside) - and it must come soon,
because it cannot confirm my re-registration.
This could work... (not only lease, not only agent, but both)
Cheers,
Martin
--
Martin Senger
email: martin.senger at gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From gordonp at ucalgary.ca Wed Aug 17 12:16:45 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Wed, 17 Aug 2005 10:16:45 -0600
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To:
References:
Message-ID: <430362ED.1090804@ucalgary.ca>
>>Alternatively, you pass back some magic cookie when the lease is taken
>>out in the first place, that is required for deregistration.
>>
>>
>>
> Well, this is exactly what biomoby has now - and nobody seems to like
>it. That's why we are talking how to change it.
>
>
Absolutely. Keeping a cookie around somewhere for a month, then
manually using it to renew the lease (remember, automating the lease
renewal is just as bad as having an RDF file that you need to change) is
bad. If I lose my cookie, or it is intercepted, I'm toast.
> So the above would work only if I can prove who I am. And the only way
>how to do it (if we do not want to introduce passwords etc.) is to let
>somebody come back to my place (because only I can control such a
>place). That means an agent...
>
>
Exactly my point from before about multiple host names/machine. You
should only be able to advertise a service available in MOBY from the
ucalgary.ca domain if you are able to put an RDF document on a web
server in that domain. It's a security check of sorts. I don't see how
a lease-only system could easily acheive this without some signing
certificates, and in fact I could prevent the NCBI from registering
services by doing it first and getting the cookie they need to fix the
registry. If we can add a "push" like method to the registry saying
"check my RDF ASAP", and maybe an optional expiry for people who like
leases (the default would be turf the service if the test case fails
repeatedly) we should have the best of both worlds...
> So: if I use a lease I am in charge of registering my services. If it
>is a new registration (such service does not exist yet), everything is
>fine and nobody needs to go to me (but can, and probably will, but there
>is no hurry). If I try to re-register (because I wish to change the lease
>period), then an agent must come to me to see that my RDF document is
>still there (with a new lease perion inside) - and it must come soon,
>because it cannot confirm my re-registration.
>
> This could work... (not only lease, not only agent, but both)
>
> Cheers,
> Martin
>
>--
>Martin Senger
> email: martin.senger at gmail.com
> skype: martinsenger
>International Rice Research Institute
>Biometrics and Bioinformatics Unit
>DAPO BOX 7777, Metro Manila
>Philippines, phone: +63-2-580-5600 (ext.2324)
>
>_______________________________________________
>moby-l mailing list
>moby-l at biomoby.org
>http://biomoby.org/mailman/listinfo/moby-l
>
>
>
From markw at illuminae.com Wed Aug 17 12:23:50 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Wed, 17 Aug 2005 09:23:50 -0700
Subject: [moby] Re: [MOBY-l] Re: [MOBY-dev] lease versus agent for
registry updating
In-Reply-To: <430362ED.1090804@ucalgary.ca>
References:
<430362ED.1090804@ucalgary.ca>
Message-ID: <1124295830.21293.24.camel@bioinfo.icapture.ubc.ca>
On Wed, 2005-08-17 at 10:16 -0600, Paul Gordon wrote:
> Keeping a cookie around somewhere for a month, then
> manually using it to renew the lease (remember, automating the lease
> renewal is just as bad as having an RDF file that you need to change) is
> bad. If I lose my cookie, or it is intercepted, I'm toast.
Yes... in fact, that solution makes me lose my cookies! ;-)
> If we can add a "push" like method to the registry saying
> "check my RDF ASAP", and maybe an optional expiry for people who like
> leases (the default would be turf the service if the test case fails
> repeatedly) we should have the best of both worlds...
This is easy enough to do given that we are moving to RDF as our Service
Signature - it just becomes another metadata tag. However, we would
then have to add some code to the registry that stores lease times, and
this then makes some assumptions about what the underling registry has
to know about. Given that we are now trying to support both a MOBY
Central and a myGrid/Feta underlying store we cannot make "arbitrary"
decisions like this anymore... A non-lease mechanism does not require
the registry to "think", only to store the data that it needs and
respond to "come hither agent" calls.
M
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From fgibbons at hms.harvard.edu Thu Aug 18 18:13:02 2005
From: fgibbons at hms.harvard.edu (Frank Gibbons)
Date: Thu, 18 Aug 2005 18:13:02 -0400
Subject: [MOBY-l] Where does articleName belong? How can I set it in a
Simple, using jMoby?
Message-ID: <5.2.1.1.2.20050818172744.011d8e00@email.med.harvard.edu>
Hi,
I'm having trouble understanding where articleName belongs. I thought I
understood it before, but now that I'm trying to use jMoby to access
services that I'm providing from Perl, I'm confused again..... (But I note
from the MOBY digest that just landed in my mailbox that that's OK, since
articleName has several uses/meanings, and may perhaps soon acquire some
more, so perhaps I'm not the only one ;)
My (Perl) services expect input that looks like this:
In jMoby, from reading TestRequest.java, I can see that it's easy to build
an object to contain a message like, using one of the constructors (it
takes two String arguments: namespace and id). But how do I set the
articleName for the enclosing tag? Calling setName() on the
service input MobyDataObject that wraps it, results in that articleName
being handed down to the child element (Object), when it should remain with
the Simple.
Maybe I've got it all wrong. This has worked for me with the Perl API for
over a year, but when trying to interact with those services using jMoby,
it barfs because the response (which looks not unlike the input above)
contains Simples with children, which "is NOT expected" [sic]. But I though
Simples could contain fundamental datatypes like String & friends. Perhaps
this relates to the discussion over the last weeks about what a Collection
is for.
Hopefully, I've overlooked something in the javadocs - if so, I'd
appreciate a pointer to the right class.
Thanks in advance.
-Frank
PhD, Computational Biologist,
Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
Tel: 617-432-3555 Fax:
617-432-3557 http://llama.med.harvard.edu/~fgibbons
From senger at ebi.ac.uk Thu Aug 18 18:59:09 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Thu, 18 Aug 2005 23:59:09 +0100 (BST)
Subject: [MOBY-l] Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To: <5.2.1.1.2.20050818172744.011d8e00@email.med.harvard.edu>
Message-ID:
Frank,
I guess you will get an reply from Paul explaining your answer.
I want just make a small note: Bear with me for another week (probably
less, just until Monday or Tuesday) when I am going to release a completly
new support/additions to jMoby that surely deals with article names in
both their incarnations (article names in Simples and article names in
members objects (HAS/HASA objects).
Cheers,
Martin
--
Martin Senger
email: martin.senger at gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From gordonp at ucalgary.ca Fri Aug 19 11:21:38 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Fri, 19 Aug 2005 09:21:38 -0600
Subject: [MOBY-l] Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To: <5.2.1.1.2.20050818172744.011d8e00@email.med.harvard.edu>
References: <5.2.1.1.2.20050818172744.011d8e00@email.med.harvard.edu>
Message-ID: <4305F902.7060506@ucalgary.ca>
> Hi,
>
> I'm having trouble understanding where articleName belongs. I thought
> I understood it before, but now that I'm trying to use jMoby to access
> services that I'm providing from Perl, I'm confused again..... (But I
> note from the MOBY digest that just landed in my mailbox that that's
> OK, since articleName has several uses/meanings, and may perhaps soon
> acquire some more, so perhaps I'm not the only one ;)
>
> My (Perl) services expect input that looks like this:
>
>
>
>
>
> In jMoby, from reading TestRequest.java, I can see that it's easy to
> build an object to contain a message like, using one of the
> constructors (it takes two String arguments: namespace and id). But
> how do I set the articleName for the enclosing tag? Calling
> setName() on the service input MobyDataObject that wraps it, results
> in that articleName being handed down to the child element (Object),
> when it should remain with the Simple.
To me it makes sense: calling object.setName() sets the article name in
the Object tag, not the Simple.
The Simple is not a data object of any sort, it is just a wrapper used
to name parameters in calls to services.
I can reuse the Object in another service, with a completely different
parameter name, or even use it as part of a collection. I should not
have to modify the Object name itself to make these calls, as it is
exactly the same Object tag as before. In fact, I never use
object.setName(), it's there because there are other parts of the code
that call it, when you are building composite objects and the
articleName is needed to identify the data member. I'll add a message
to the javadoc stating this.
Because most people are just using one parameter inputs, it's easy to
falsely make the connection object name == simple name.
Was there not some dicussion on the list of changing articleName to
paramName for Simples? If not, we should defiunitely do this!
In any case, the jMOBY code right now doesn't deal nicely with named
parameters. I was striuggling before with how to deal with XML that
both named and unnamed parameters, very confusing. But now that all
parameters have to be named (since API 0.86) its much easier. I'll fix
the code on the weekend. Thanks for pointiung this out.
> Maybe I've got it all wrong. This has worked for me with the Perl API
> for over a year, but when trying to interact with those services using
> jMoby, it barfs because the response (which looks not unlike the input
> above) contains Simples with children, which "is NOT expected" [sic].
> But I though Simples could contain fundamental datatypes like String &
> friends. Perhaps this relates to the discussion over the last weeks
> about what a Collection is for.
Can you send me your exact XML?
> Hopefully, I've overlooked something in the javadocs - if so, I'd
> appreciate a pointer to the right class.
>
> Thanks in advance.
>
> -Frank
>
> PhD, Computational Biologist,
> Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA
> 02115, USA.
> Tel: 617-432-3555 Fax: 617-432-3557
> http://llama.med.harvard.edu/~fgibbons
> _______________________________________________
> moby-l mailing list
> moby-l at biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
>
From gordonp at ucalgary.ca Fri Aug 19 12:06:11 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Fri, 19 Aug 2005 10:06:11 -0600
Subject: [MOBY-l] Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To: <4305F902.7060506@ucalgary.ca>
References: <5.2.1.1.2.20050818172744.011d8e00@email.med.harvard.edu>
<4305F902.7060506@ucalgary.ca>
Message-ID: <43060373.1080304@ucalgary.ca>
Replying to my own :-)
>
> In any case, the jMOBY code right now doesn't deal nicely with named
> parameters. I was striuggling before with how to deal with XML that
> both named and unnamed parameters, very confusing. But now that all
> parameters have to be named (since API 0.86) its much easier. I'll
> fix the code on the weekend. Thanks for pointiung this out.
>
Hmm. A corrolary of requiring all parameters to be named is that there
the returned data should be named too, as MOBY input and output messages
are essentially symmetric. We could get around this by saying that if
you have only one parameter, it does not need to be named. In that way,
the response simple or collection does not need to be named either.
This also makes the vast majority of currently registered services valid
again according to API v.0.86 , as most have one parameter, and it is
not named.
From senger at ebi.ac.uk Fri Aug 19 12:46:07 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Fri, 19 Aug 2005 17:46:07 +0100 (BST)
Subject: [MOBY-l] Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To: <43060373.1080304@ucalgary.ca>
Message-ID:
Regarding the mandatorness of article name, I have heard the same rumour
as you did, Paul. What I got confirmed is that article name must be usede
for included objects in other objects. What was not yet confirmed is that
also Simples and Collections (the wrapper objects) must have article
name. Obviously not having article names mandatory is a pain in the neck -
I am struggling with it the whole last week...
Also API documentation is very unnerving (is this a right English word to
use if you wish to say a four-letter word but you are too polite or too
afraid to say it?). In one place it says "Articles are required to be
named using the articleName attribute." and a bit further it says: "The
articleName attribute of the Simple and/or Collection elements is
optional...".
Mark, could you say loundly and soonish:
1) How it is really?
2) What to do with currently registered services that do not comply with
the rule from 1) ?
Cheers,
Martin
--
Martin Senger
email: martin.senger at gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From markw at illuminae.com Fri Aug 19 14:42:21 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Fri, 19 Aug 2005 11:42:21 -0700
Subject: [MOBY-l] Re: Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To:
References:
Message-ID: <1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
On Fri, 2005-08-19 at 17:46 +0100, Martin Senger wrote:
> What was not yet confirmed is that
> also Simples and Collections (the wrapper objects) must have article
> name. Obviously not having article names mandatory is a pain in the neck -
> I am struggling with it the whole last week...
It was confirmed a while ago - see below...
> Also API documentation is very unnerving (is this a right English word to
> use if you wish to say a four-letter word but you are too polite or too
> afraid to say it?). In one place it says "Articles are required to be
> named using the articleName attribute." and a bit further it says: "The
> articleName attribute of the Simple and/or Collection elements is
> optional...".
I'll look for this error. I really need someone to take over the task
of documentation, however...
> Mark, could you say loundly and soonish:
> 1) How it is really?
I reported loudly and previously:
: From: markw at illuminae.com
: Reply-To: Core developer announcements 2) What to do with currently registered services that do not comply with
> the rule from 1) ?
I think we can add an arbitratry articleName to their inputs, since they
will no doubt ignore it anyway...
Must run!
M
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From gordonp at ucalgary.ca Fri Aug 19 19:13:47 2005
From: gordonp at ucalgary.ca (Paul Gordon)
Date: Fri, 19 Aug 2005 17:13:47 -0600
Subject: [MOBY-l] Re: Where does articleName belong? How can I set it
in a Simple, using jMoby?
In-Reply-To: <1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
References:
<1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
Message-ID: <430667AB.1070906@ucalgary.ca>
The way I dealt with this was to allow a zero-length string as a name.
There's nothing in the API that forbids this. Anything without an
articleName gets a zero-length name, and if there is more than one, I
raise an exception. This handles currently registered one-parameters
jobs fine without resorting to random names, and enforces uniqueness of
the articleName going forward. Actually, the API doesn't say they need
to be unique, just that they need to be there.
>: 1) *all* parameters going into a service must now be named currently using the
>: articleName attribute of the Simple/Collection/Parameter tag
>
>
>
>
>>2) What to do with currently registered services that do not comply with
>>the rule from 1) ?
>>
>>
>
>I think we can add an arbitratry articleName to their inputs, since they
>will no doubt ignore it anyway...
>
>Must run!
>
>M
>
>
>
>
From senger at ebi.ac.uk Fri Aug 19 20:11:13 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Sat, 20 Aug 2005 01:11:13 +0100 (BST)
Subject: [MOBY-l] Re: Where does articleName belong? How can I set it in a
Simple, using jMoby?
In-Reply-To: <1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
Message-ID:
> It was confirmed a while ago - see below...
>
Thank for confirmation. It is a good news. (I knew about the email you
quoted - but I was not sure if it was a definitive announcement or just
your contribution to the discussion of the API changes. Now it's clear. My
confusion just confirms that Biomoby still does not have in place a
formal procedure how and when to do API changes.)
Cheers,
Martin
--
Martin Senger
email: martin.senger at gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From paolo at ist.unige.it Mon Aug 22 07:26:40 2005
From: paolo at ist.unige.it (Paolo Romano)
Date: Mon, 22 Aug 2005 13:26:40 +0200 (DFT)
Subject: [MOBY-l] NETTAB 2005 - Deadlines approaching: early registration
and call for posters
Message-ID: <200508221126.j7MBQeqR019972@ibm43p.biotech.ist.unige.it>
Dear all,
this is a reminder for next deadlines of the NETTAB 2005 Workshop on
"Workflows management: new abilities for the biological information
overflow" that will be held on next October 5-7, 2005, in Naples, Italy.
The Scientific Programme is now available on-line at
http://www.nettab.org/2005/progr.html .
The Opening Lecture will be given by Francis Ouellette.
Francis Ouellette is the Director of the UBiC, the Bioinformatics
Centre of the University of British Columbia, Canada. He is Associate
Professor of the Michael Smith Laboratories and of the Department of
Medical Genetics of UBC. He is also a core faculty member in the new
UBC graduate Training Program in Bioinformatics for Health Research,
associate director of Bioinformatics at Genome British Columbia and
director of the Canadian Genetic Diseases Network (CGDN) bioinformatics
core facility where he helps coordinate the Canadian Bioinformatics
Workshops.
Francis has an exceptional curriculum and his recent research,
training and coordination activities make him one of the most known
and appreciated bioinformaticians.
The title of his opening talk will be
"Workflow management in bioinformatics: the possibilities and the challenges".
----------------------------------------------------------------------
The Call for posters and position papers is closing on next friday
August 26, 2005.
You are all warmly invited to present your recent activity related to the
workshops' topics by submitting a poster abstract (1-2 A4 pages, font size
12 pt, MS Word format) by email to posters2005 at nettab.org .
Topics are the following:
Technologies and technological platforms of interest, with emphasis on:
- Web Services (SOAP, WSDL, WSFL, UDDI, ....)
- Web Services Choreography and Orchestration
- Semantic Web (RDF, LSID, OWL, ...)
- comparison of available technologies, limitations, pros and cons
- knowledge representation
- biological data and knowledge modeling tools
- Ontologies, Databases and Applications of Semantics in Bioinformatics
Workflow management systems in bioinformatics
- implementations of web services
- implementations of registries
- reuse and versioning of web services and workflows
- workflow management systems
- web interfaces for accessing and executing workflows
- interactive systems to support work flows
Applications of workflow management systems in bioinformatics
- Methodologies for life sciences analysis, such as:
- gene expression,
- genome annotation,
- mass spec peptide fragment identification,
- Encoding of the above in workflows
- Case studies
- Scenarios and use cases
Check all details at: http://www.nettab.org/2005/call.html .
--------------------------------------------------------------------------------
The early registration deadline is also next Friday August 26, 2005.
The registration form is available on-line at:
http://www.nettab.org/2005/rform.html .
The payment of the fee can either be done on-line, through the Online
Payment Form of the Bioinformatics Italian Society (BITS), or by direct
money transfer.
Participation fees are as follows:
Until August 26, 2005:
- Students: 70,00 Euro
- Academic: 130,00 Euro (reduced fee: 117,00 Euro)
- Non-academic: 270,00 Euro (reduced fee: 243,00 Euro)
After August 26, 2005:
- Students: 70,00 Euro
- Academic: 180,00 Euro (reduced fee: 162,00 Euro)
- Non-academic: 370,00 Euro (reduced fee: 333,00 Euro)
The 10% reduction on fees is applied for members of:
- ISCB (International Society for Computational Biology),
http://www.iscb.org/
- BITS (Bioinformatics Italian Society),
http://www.bioinformatics.it/
- Hormone Responsive Breast Cancer (HRBC) Genomics Network,
http://www.hrbc-genomics.net/
- Oncology over Internet (O2I) project,
http://www.o2i.it/
- Interdisciplinary Laboratory for Technologies in Bioinformatics (LITBIO)
--------------------------------------------------------------------------------
I'm looking forward to meeting many of you in Naples quite soon.
Ciao. Paolo
--
Paolo Romano (paolo.romano at istge.it)
Bioinformatics and Structural Proteomics
National Cancer Research Institute (IST)
Largo Rosanna Benzi, 10, I-16132, Genova, Italy
Tel: +39-010-5737-288 Fax: +39-010-5737-295
Web: http://www.nettab.org/promano/
From p.lord at cs.man.ac.uk Mon Aug 22 08:21:48 2005
From: p.lord at cs.man.ac.uk (Phillip Lord)
Date: 22 Aug 2005 13:21:48 +0100
Subject: [MOBY-l] Re: [MOBY-dev] lease versus agent for registry updating
In-Reply-To: <430362ED.1090804@ucalgary.ca>
References:
<430362ED.1090804@ucalgary.ca>
Message-ID: <87fyt2duw3.fsf@rpc71.cs.man.ac.uk>
>>>>> "Paul" == Paul Gordon writes:
>>> Alternatively, you pass back some magic cookie when the lease is
>>> taken out in the first place, that is required for
>>> deregistration.
>>>
>>>
>> Well, this is exactly what biomoby has now - and nobody seems to
>> like it. That's why we are talking how to change it.
>>
Paul> Absolutely. Keeping a cookie around somewhere for a month,
Paul> then manually using it to renew the lease (remember,
Paul> automating the lease renewal is just as bad as having an RDF
Paul> file that you need to change) is bad. If I lose my cookie, or
Paul> it is intercepted, I'm toast.
I don't think that you'd need the cookie to renew your lease, just to
remove it before hand. If you lose it, then you just have to wait for
it to run out naturally.
Of course, if it's built into the service provider libraries, there is
no real reason that you should lose it.
>> So the above would work only if I can prove who I am. And the
>> only way how to do it (if we do not want to introduce passwords
>> etc.) is to let somebody come back to my place (because only I
>> can control such a place). That means an agent...
>>
Paul> Exactly my point from before about multiple host
Paul> names/machine. You should only be able to advertise a service
Paul> available in MOBY from the ucalgary.ca domain if you are able
Paul> to put an RDF document on a web server in that domain. It's a
Paul> security check of sorts. I don't see how a lease-only system
Paul> could easily acheive this without some signing certificates,
Paul> and in fact I could prevent the NCBI from registering services
Paul> by doing it first and getting the cookie they need to fix the
Paul> registry.
I think that the lease is a property of the registration, not of the
service. So, if you register a service, it won't prevent anyone else
from registering the same service again. Again, if you put some sort
of call back into the leasing system, you could prevent anyone from
registering a service that they do not own. Although, personally, I
don't see any major problem with third party registration. Feta
effectively works on third party registration, as most of the semantic
descriptions do not come from the service provider.
Phil
From markw at illuminae.com Mon Aug 22 11:34:49 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Mon, 22 Aug 2005 08:34:49 -0700
Subject: [moby] Re: [MOBY-l] Re: Where does articleName belong? How can
I set it in aSimple, using jMoby?
In-Reply-To: <4309A53B.8080701@cnb.uam.es>
References:
<1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
<4309A53B.8080701@cnb.uam.es>
Message-ID: <1124724889.3068.103.camel@bioinfo.icapture.ubc.ca>
On Mon, 2005-08-22 at 12:13 +0200, David Gonz?lez Pisano wrote:
> Umm, does this mean that also Simples into Collections *must* have an
> articlename too, even an arbitrary one? I was wondering if the lack of
> an articleName in the "Simple inside a Collection" example (see the
> API documentation, section INPUT) means that a Simple has to have
> articleName when is the only input
No, Simples inside a collection do not have an articleName, and at most
they would all have the same articleName (I'm actually not sure how the
registry code handles this at the moment... I suspect that it does
record the articleName of a Simple within a Collection if an articleName
has been provided at registration, but I know for certain that none of
my client libraries pay attention to it even if it is there, since it
was not *intended* to be used, nor meaningful).
> Also, probably the twiki needs an update, because as today still says "The articleName attribute of the Simple and/or Collection elements is
> optional (it may or may not be there, and if there, it may or may not have a non-null value)". Note that the last part of the phrase contradicts Paul's solution.
I'll try to find this and fix it.
again... we really need someone to volunteer to take over the task of
re-writing and maintaining the documentation :-)
M
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From dgpisano at cnb.uam.es Mon Aug 22 06:13:15 2005
From: dgpisano at cnb.uam.es (=?ISO-8859-1?Q?David_Gonz=E1lez_Pisano?=)
Date: Mon, 22 Aug 2005 12:13:15 +0200
Subject: [MOBY-l] Re: Where does articleName belong? How can I set it
in aSimple, using jMoby?
In-Reply-To: <1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
References:
<1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
Message-ID: <4309A53B.8080701@cnb.uam.es>
Mark Wilkinson escribi?:
>On Fri, 2005-08-19 at 17:46 +0100, Martin Senger wrote:
>
>
>>What was not yet confirmed is that
>>also Simples and Collections (the wrapper objects) must have article
>>name. Obviously not having article names mandatory is a pain in the neck -
>>I am struggling with it the whole last week...
>>
>>
>It was confirmed a while ago - see below...
>
>
>
>:
>: A few changes to announce for the 0.86 API (this is not yet running on the
>: production server):
>:
>: 1) *all* parameters going into a service must now be named currently using the
>: articleName attribute of the Simple/Collection/Parameter tag
>
>
Umm, does this mean that also Simples into Collections *must* have an
articlename too, even an arbitrary one? I was wondering if the lack of
an articleName in the "Simple inside a Collection" example (see the API
documentation, section INPUT) means that a Simple has to have
articleName when is the only input
<------- articleName here!
but does not need it when is inside a Collection?
<------- no articleName here?
Also, probably the twiki needs an update, because as today still says "The articleName attribute of the Simple and/or Collection elements is
optional (it may or may not be there, and if there, it may or may not have a non-null value)". Note that the last part of the phrase contradicts Paul's solution.
David (working on the Errors proposal, which heavily depends on the answer to this mail) ;-)
From fgibbons at hms.harvard.edu Mon Aug 22 11:54:11 2005
From: fgibbons at hms.harvard.edu (Frank Gibbons)
Date: Mon, 22 Aug 2005 11:54:11 -0400
Subject: [moby] Re: [MOBY-l] Re: Where does articleName belong? How
can I set it in aSimple, using jMoby?
In-Reply-To: <1124724889.3068.103.camel@bioinfo.icapture.ubc.ca>
References: <4309A53B.8080701@cnb.uam.es>
<1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
<4309A53B.8080701@cnb.uam.es>
Message-ID: <5.2.1.1.2.20050822115221.010ccf88@email.med.harvard.edu>
Mark,
At 11:34 AM 8/22/2005, you wrote:
> > Also, probably the twiki needs an update, because as today still says
> "The articleName attribute of the Simple and/or Collection elements is
> > optional (it may or may not be there, and if there, it may or may not
> have a non-null value)". Note that the last part of the phrase
> contradicts Paul's solution.
>
>I'll try to find this and fix it.
>
>again... we really need someone to volunteer to take over the task of
>re-writing and maintaining the documentation :-)
I'll volunteer for that. You're referring to the Wiki API page, I assume,
or are there other docs (besides the POD and javadoc)?
-Frank
PhD, Computational Biologist,
Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
Tel: 617-432-3555 Fax:
617-432-3557 http://llama.med.harvard.edu/~fgibbons
From senger at ebi.ac.uk Mon Aug 22 12:08:13 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Mon, 22 Aug 2005 17:08:13 +0100 (BST)
Subject: [moby] Re: [MOBY-l] Re: Where does articleName belong? How can
I set it in aSimple, using jMoby?
In-Reply-To: <5.2.1.1.2.20050822115221.010ccf88@email.med.harvard.edu>
Message-ID:
> I'll volunteer for that. You're referring to the Wiki API page, I assume,
> or are there other docs (besides the POD and javadoc)?
>
That's goooood!
But please be merciless, as much as you can. :-)
The main problem I had with these pages from/at the beginning is that
only the initiated easily understand that the page actually contain two
APIs - one for registry and one for services. The current page mix them
too much (IMHO).
I also volunteer for proof-reading (just send me your draft) - and I
promise I will be chasing Mark with every smallest ambiguity that I could
find.
Cheers,
Martin
--
Martin Senger
email: martin.senger at gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From markw at illuminae.com Mon Aug 22 12:14:49 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Mon, 22 Aug 2005 09:14:49 -0700
Subject: [moby] Re: [MOBY-l] Re: Where does articleName belong? How can
I set it in aSimple, using jMoby?
In-Reply-To:
References:
Message-ID: <1124727289.3068.131.camel@bioinfo.icapture.ubc.ca>
> and I
> promise I will be chasing Mark with every smallest ambiguity that I could
> find.
OOOooohhh yeah... he sure will! I know that from experience ;-)
M
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From markw at illuminae.com Mon Aug 22 12:16:04 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Mon, 22 Aug 2005 09:16:04 -0700
Subject: [moby] Re: [MOBY-l] Re: Where does articleName belong? How can
I set it in aSimple, using jMoby?
In-Reply-To: <5.2.1.1.2.20050822115221.010ccf88@email.med.harvard.edu>
References: <4309A53B.8080701@cnb.uam.es>
<1124476941.26245.39.camel@bioinfo.icapture.ubc.ca>
<4309A53B.8080701@cnb.uam.es>
<5.2.1.1.2.20050822115221.010ccf88@email.med.harvard.edu>
Message-ID: <1124727364.3068.134.camel@bioinfo.icapture.ubc.ca>
that's fantastic! Thanks Frank!
I need to set you up with a developers account on the TWiki machine so
that you can edit the main page I think... let's coordinate this off-
list.
Speaking of the Twiki machine... Simon - how's the new website plan
coming?
M
On Mon, 2005-08-22 at 11:54 -0400, Frank Gibbons wrote:
> Mark,
>
> At 11:34 AM 8/22/2005, you wrote:
> > > Also, probably the twiki needs an update, because as today still says
> > "The articleName attribute of the Simple and/or Collection elements is
> > > optional (it may or may not be there, and if there, it may or may not
> > have a non-null value)". Note that the last part of the phrase
> > contradicts Paul's solution.
> >
> >I'll try to find this and fix it.
> >
> >again... we really need someone to volunteer to take over the task of
> >re-writing and maintaining the documentation :-)
>
>
> I'll volunteer for that. You're referring to the Wiki API page, I assume,
> or are there other docs (besides the POD and javadoc)?
>
> -Frank
>
>
> PhD, Computational Biologist,
> Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA.
> Tel: 617-432-3555 Fax:
> 617-432-3557 http://llama.med.harvard.edu/~fgibbons
>
> _______________________________________________
> moby-l mailing list
> moby-l at biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From silvia77 at dist.unige.it Thu Aug 25 06:12:09 2005
From: silvia77 at dist.unige.it (scaglione)
Date: Thu, 25 Aug 2005 12:12:09 +0200
Subject: [MOBY-l] sth_collection_ins
Message-ID: <1124964729.430d9979a03e9@webmail.unige.it>
hello,
I have followed instructions to install a moby central in my machine.I have
tried the testMOBYClientCentral_v05.pl script on my mobycentral and an error
appeared:
>perl testMOBYCentral_v05.pl
Useless use of hash element in void context
at /usr/lib/perl5/site_perl/5.8.5/RDF/Core/Serializer.pm line 52.
Global symbol "$sth_collection_ins" requires explicit package name
at /usr/lib/perl5/site_perl/5.8.5/MOBY/Central.pm line 3426.
Compilation failed in require at testMOBYCentral_v05.pl line 4.
BEGIN failed--compilation aborted at testMOBYCentral_v05.pl line 4.
what does it mean?
The overall configuration of the moby central seemed to be ok, since the test
of the same script on the main mobycentral works...
many thanks
silvia
--
PhD. Silvia Scaglione
DIST University of Genoa
Viale Causa,13
16145 Genoa
phone +39 010 353 2789
phone CBA +39 010 5737 505
fax +39 010 353 2948
From markw at illuminae.com Thu Aug 25 12:27:21 2005
From: markw at illuminae.com (Mark Wilkinson)
Date: Thu, 25 Aug 2005 09:27:21 -0700
Subject: [moby] [MOBY-l] sth_collection_ins
In-Reply-To: <1124964729.430d9979a03e9@webmail.unige.it>
References: <1124964729.430d9979a03e9@webmail.unige.it>
Message-ID: <1124987241.11350.50.camel@bioinfo.icapture.ubc.ca>
Can you first confirm where you got your mobycentral code from? Is it
one of the newer releases? I'm certain that the latest release is
functional (though there are a few small bugs in it that are now fixed
in the CVS, but the CVS code isn't supported yet...)
make/make install this code and see if it solves your problems. You
probably wont need to modify any of your configuration after this
installation.
http://biomoby.org/releases/biomoby.0.8.2a.tar.gz
M
On Thu, 2005-08-25 at 12:12 +0200, scaglione wrote:
> hello,
> I have followed instructions to install a moby central in my machine.I have
> tried the testMOBYClientCentral_v05.pl script on my mobycentral and an error
> appeared:
>
> >perl testMOBYCentral_v05.pl
>
> Useless use of hash element in void context
> at /usr/lib/perl5/site_perl/5.8.5/RDF/Core/Serializer.pm line 52.
> Global symbol "$sth_collection_ins" requires explicit package name
> at /usr/lib/perl5/site_perl/5.8.5/MOBY/Central.pm line 3426.
> Compilation failed in require at testMOBYCentral_v05.pl line 4.
> BEGIN failed--compilation aborted at testMOBYCentral_v05.pl line 4.
>
> what does it mean?
> The overall configuration of the moby central seemed to be ok, since the test
> of the same script on the main mobycentral works...
> many thanks
> silvia
>
>
>
--
"Ontologists do it with the edges!"
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274
From jmfernandez at cnb.uam.es Fri Aug 26 12:37:31 2005
From: jmfernandez at cnb.uam.es (=?ISO-8859-1?Q?Jos=E9_Mar=EDa_Fern=E1ndez_Gonz=E1lez?=)
Date: Fri, 26 Aug 2005 18:37:31 +0200
Subject: [MOBY-l] RESOURCES/MOBY-S/Objects not working?
Message-ID: <430F454B.7020603@cnb.uam.es>
Hi people!
Is there any problem with Tomcat server in MOBY Central? Since Tuesday
(or so) I'm playing with latest Taverna CVS, and BioMOBY related stuff
doesn't work because there is a HTTP 500 code related to
http://www.biomoby.org/RESOURCES/MOBY-S/Objects (RDF Object
definitions). As Taverna 1.2 also depends on this URL, MOBY services
tree is not either working.
Best Regards,
Jos? Mar?a
--
Jos? Mar?a Fern?ndez Gonz?lez e-mail: jmfernandez at cnb.uam.es
Tlfn: (+34) 91 585 54 50 Fax: (+34) 91 585 45 06
Grupo de Dise?o de Proteinas Protein Design Group
Centro Nacional de Biotecnolog?a National Center of Biotechnology
C.P.: 28049 Zip Code: 28049
C/. Darwin n? 3 (Campus Cantoblanco, U. Aut?noma), Madrid (Spain)
From edward.kawas at gmail.com Fri Aug 26 13:20:17 2005
From: edward.kawas at gmail.com (Eddie Kawas)
Date: Fri, 26 Aug 2005 12:20:17 -0500
Subject: [MOBY-l] RESOURCES/MOBY-S/Objects not working?
In-Reply-To: <430F454B.7020603@cnb.uam.es>
References: <430F454B.7020603@cnb.uam.es>
Message-ID:
Hi,
I will look into this asap (saturday or sunday). It looks like the
server works because the other RDF documents are being generated so
something must be wrong with either the configuration.
Actually, I just noticed that the types script for objects is
returning stuff that it shouldnt
(http://mobycentral.icapture.ubc.ca:8090/types/Objects) and this
script is utilized by the RESOURCES script. As soon as I am able to
check this out, I will and l will fix it. The problem seems to lie
with the object 'Regex' and its description (no longer limited to one
line per object).
Eddie
On 8/26/05, Jos? Mar?a Fern?ndez Gonz?lez wrote:
> Hi people!
> Is there any problem with Tomcat server in MOBY Central? Since Tuesday
> (or so) I'm playing with latest Taverna CVS, and BioMOBY related stuff
> doesn't work because there is a HTTP 500 code related to
> http://www.biomoby.org/RESOURCES/MOBY-S/Objects (RDF Object
> definitions). As Taverna 1.2 also depends on this URL, MOBY services
> tree is not either working.
>
> Best Regards,
> Jos? Mar?a
>
> --
> Jos? Mar?a Fern?ndez Gonz?lez e-mail: jmfernandez at cnb.uam.es
> Tlfn: (+34) 91 585 54 50 Fax: (+34) 91 585 45 06
> Grupo de Dise?o de Proteinas Protein Design Group
> Centro Nacional de Biotecnolog?a National Center of Biotechnology
> C.P.: 28049 Zip Code: 28049
> C/. Darwin n? 3 (Campus Cantoblanco, U. Aut?noma), Madrid (Spain)
> _______________________________________________
> moby-l mailing list
> moby-l at biomoby.org
> http://biomoby.org/mailman/listinfo/moby-l
>
From senger at ebi.ac.uk Fri Aug 26 20:28:31 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Sat, 27 Aug 2005 01:28:31 +0100 (BST)
Subject: [MOBY-l] RESOURCES/MOBY-S/Objects not working?
In-Reply-To:
Message-ID:
> The problem seems to lie with the object 'Regex' and its description
> (no longer limited to one line per object).
>
Regex is my object! I am always happy to shoot down all servers around
:-)
But seriusly, anything wrong with longer description? Does API preclude
it?
Cheers,
Martin
--
Martin Senger
email: martin.senger at gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From edward.kawas at gmail.com Sat Aug 27 09:55:31 2005
From: edward.kawas at gmail.com (Eddie Kawas)
Date: Sat, 27 Aug 2005 08:55:31 -0500
Subject: [MOBY-l] RESOURCES/MOBY-S/Objects not working?
In-Reply-To:
References:
Message-ID:
I think the description is fine, Mark will need to verify this. I will
just have to update my logic (already done in my head and when i get
home today i will update the codebase).
Eddie
On 8/26/05, Martin Senger wrote:
> > The problem seems to lie with the object 'Regex' and its description
> > (no longer limited to one line per object).
> >
> Regex is my object! I am always happy to shoot down all servers around
> :-)
> But seriusly, anything wrong with longer description? Does API preclude
> it?
>
> Cheers,
> Martin
>
> --
> Martin Senger
> email: martin.senger at gmail.com
> skype: martinsenger
> International Rice Research Institute
> Biometrics and Bioinformatics Unit
> DAPO BOX 7777, Metro Manila
> Philippines, phone: +63-2-580-5600 (ext.2324)
>
>
From senger at ebi.ac.uk Mon Aug 29 10:32:30 2005
From: senger at ebi.ac.uk (Martin Senger)
Date: Mon, 29 Aug 2005 15:32:30 +0100 (BST)
Subject: [MOBY-l] Moses - Moby Services Support
In-Reply-To: <1476.82.66.216.27.1118945157.squirrel@82.66.216.27>
Message-ID:
Hi all,
My first month at Philippines, at IRRI, is almost finished. So it's
time to announce some progress :-)
I have commited code and a lot of documentation helping to write
Biomoby services in Java. The idea is to generate classes for all
registered data types and then generate skeleton for a new service. Its
developer can extend such skeleton , put there the business logic without
worries about the Biomoby XML and about the SOAP itself. These things
should be hidden.
It is similar what Paul Gordon provides for the clients, but here you
can use strongly-type approach (having all data types generated). Or, you
can say, that we can have more ways to do the same, like in Perl :-)
As (almost) a side-effect, the generated code, because it has rich
API/javadoc comments, can be used as a primitive browser of the biomoby
registry. Here you can see the API for all data types and for all services
as they are today (when I find a place where I can run a cronjob, we can
have this API up-to-date anytime you look):
http://www.ebi.ac.uk/~senger/jMoby/APIservices/index.html
The documentation is in jMoby and can be viewed at:
http://biomoby.org/moby-live/Java/docs/Moses.html.
I hope you'll enjoy it and I am happy to hear your suggestions, bug
reports (remember there is also a bugzilla open now, but feel free to send
emails direcly to me) and anything bout Moses.
With regards,
Martin
--
Martin Senger
email: martin.senger at gmail.com
skype: martinsenger
International Rice Research Institute
Biometrics and Bioinformatics Unit
DAPO BOX 7777, Metro Manila
Philippines, phone: +63-2-580-5600 (ext.2324)
From silvia77 at dist.unige.it Thu Aug 25 05:52:50 2005
From: silvia77 at dist.unige.it (scaglione)
Date: Thu, 25 Aug 2005 11:52:50 +0200
Subject: [MOBY-l] sth_collection_ins
Message-ID: <1124963570.430d94f212a32@webmail.unige.it>
hello,
I have followed instructions to install a moby central in my machine.I have
tried the testMOBYClientCentral_v05.pl script on my mobycentral and an error
appeared:
>perl testMOBYCentral_v05.pl
Useless use of hash element in void context
at /usr/lib/perl5/site_perl/5.8.5/RDF/Core/Serializer.pm line 52.
Global symbol "$sth_collection_ins" requires explicit package name
at /usr/lib/perl5/site_perl/5.8.5/MOBY/Central.pm line 3426.
Compilation failed in require at testMOBYCentral_v05.pl line 4.
BEGIN failed--compilation aborted at testMOBYCentral_v05.pl line 4.
what does it mean?
The overall configuration of the moby central seemed to be ok, since the test
of the same script on the main mobycentral works...
many thanks
silvia
--
PhD. Silvia Scaglione
DIST University of Genoa
Viale Causa,13
16145 Genoa
phone +39 010 353 2789
phone CBA +39 010 5737 505
fax +39 010 353 2948
From silvia77 at dist.unige.it Wed Aug 31 09:32:20 2005
From: silvia77 at dist.unige.it (Silvia Scaglione)
Date: Wed, 31 Aug 2005 15:32:20 +0200
Subject: [moby] [MOBY-l] test
In-Reply-To: <1124987241.11350.50.camel@bioinfo.icapture.ubc.ca>
References: <1124964729.430d9979a03e9@webmail.unige.it>
<1124987241.11350.50.camel@bioinfo.icapture.ubc.ca>
Message-ID: <6.2.0.14.0.20050831152820.01cc1998@mail.unige.it>
I have used the newer releases suggested below, but new errors appears
after testing testMOBYClientCentral_v05.pl script on the mobycentral:
...
test 19 [PASS]
test 20 [FAIL] malformed payload invalid character string for
serviceName. Must start with a letter followed by [A-Za-z0-9_]
test 21 [FAIL] Expected to find service; didn't find service
test 22 [FAIL] Expected to find service; didn't find service
test 23 [PASS]
test 24 [PASS]
Use of uninitialized value in print at testMOBYClientCentral_v05.pl line 250.
test 25 [FAIL] WSDL was not retrieved
test 26 [FAIL] The service specified by authority=www.illuminae.com
servicename=myfirstservice2 does not exist in the registry
test 27 [PASS]
...
Moreover, when I tried the testMOBYClientCentral_v05.pl script on my
mobycentral, the error message changes a little bit:
>[www at ml370 scripts]$ perl testMOBYClientCentral_v05.pl
>TESTING MOBY CLIENT with
> URL: http://ml370.biotech.ist.unige.it/cgi-bin/MOBY-Central.pl
> URI: http://ml370.biotech.ist.unige.it/MOBY/Central
> Proxy: No Proxy Server
>
>test 1a [PASS]
>test 1b [PASS]
>test 1 [FAIL] Relationship Type hasa does not exist in the
>biomoby.org Relationship Type system
>
>test 2 [PASS]
>test 3 [FAIL] Relationship Type hasa does not exist in the
>biomoby.org Relationship Type system
>test 4 [FAIL] Relationship type ISA does not exist in the
>relationship ontology
>test 5 [PASS]
>test 6 [FAIL] Object type $term cannot be resolved to an LSID
>test 7 [PASS]
>test 8 [FAIL] Object type $term cannot be resolved to an LSID
>test 9 [FAIL] Service Type HypotheticalService1 does not exist in
>the ontology
>test 10 [PASS]
>test 11 [PASS]
>test 12 [PASS]
>test 13 [PASS]
>test 14 [FAIL] Relationship Type isa does not exist in the
>biomoby.org Relationship Type system
>test 15 [FAIL] Relationship Type isa does not exist in the
>biomoby.org Relationship Type system
>test 16 [PASS]
>test 17 [PASS]
>test 18 [PASS]
>test 19 [PASS]
>test 20 [PASS]
>500 Internal Server Error
> ERROR ERROR ERROR
...any suggestions?
At 18.27 25/08/2005, Mark Wilkinson wrote:
>Can you first confirm where you got your mobycentral code from? Is it
>one of the newer releases? I'm certain that the latest release is
>functional (though there are a few small bugs in it that are now fixed
>in the CVS, but the CVS code isn't supported yet...)
>
>make/make install this code and see if it solves your problems. You
>probably wont need to modify any of your configuration after this
>installation.
>
>http://biomoby.org/releases/biomoby.0.8.2a.tar.gz
>
>
>M
>
>
>
>On Thu, 2005-08-25 at 12:12 +0200, scaglione wrote:
> > hello,
> > I have followed instructions to install a moby central in my machine.I
> have
> > tried the testMOBYClientCentral_v05.pl script on my mobycentral and an
> error
> > appeared:
> >
> > >perl testMOBYCentral_v05.pl
> >
> > Useless use of hash element in void context
> > at /usr/lib/perl5/site_perl/5.8.5/RDF/Core/Serializer.pm line 52.
> > Global symbol "$sth_collection_ins" requires explicit package name
> > at /usr/lib/perl5/site_perl/5.8.5/MOBY/Central.pm line 3426.
> > Compilation failed in require at testMOBYCentral_v05.pl line 4.
> > BEGIN failed--compilation aborted at testMOBYCentral_v05.pl line 4.
> >
> > what does it mean?
> > The overall configuration of the moby central seemed to be ok, since
> the test
> > of the same script on the main mobycentral works...
> > many thanks
> > silvia
> >
> >
> >
>--
>"Ontologists do it with the edges!"
>
>Mark Wilkinson
>Asst. Professor
>Dept. of Medical Genetics
>University of British Columbia
>PI in Bioinformatics
>iCAPTURE Centre
>St. Paul's Hospital
>Rm. 166, 1081 Burrard St.
>Vancouver, BC, V6Z 1Y6
>tel: 604 682 2344 x62129
>fax: 604 806 9274
PhD. Silvia Scaglione
Universita`degli Studi di Genova
Facolta' di Ingegneria, D.I.S.T. BioLab
Viale Causa, 13
16145 GENOVA - Italy
Tel Bio-lab +39 010 353 2789
Tel Centro Biotecnologie Avanzate +39 010 5737 505
Fax +39 010 353 2948
E-mail silvia77 at dist.unige.it
From markw at illuminae.com Wed Aug 31 10:03:16 2005
From: markw at illuminae.com (mark wilkinson)
Date: Wed, 31 Aug 2005 14:03:16 +0000 GMT
Subject: [moby] [MOBY-l] test
Message-ID: <1390300886-1125497047-cardhu_blackberry.rim.net-15323-@engine08-cell01>
Use "make test" instead - it is up to date and far far far more comprehensive.
M
-----Original Message-----
From: Silvia Scaglione
Date: Wed, 31 Aug 2005 15:32:20
To:moby-l at biomoby.org
Subject: Re: [moby] [MOBY-l] test
I have used the newer releases suggested below, but new errors appears
after testing testMOBYClientCentral_v05.pl script on the mobycentral:
...
test 19 [PASS]
test 20 [FAIL] malformed payload invalid character string for
serviceName. Must start with a letter followed by [A-Za-z0-9_]
test 21 [FAIL] Expected to find service; didn't find service
test 22 [FAIL] Expected to find service; didn't find service
test 23 [PASS]
test 24 [PASS]
Use of uninitialized value in print at testMOBYClientCentral_v05.pl line 250.
test 25 [FAIL] WSDL was not retrieved
test 26 [FAIL] The service specified by authority=www.illuminae.com
servicename=myfirstservice2 does not exist in the registry
test 27 [PASS]
...
Moreover, when I tried the testMOBYClientCentral_v05.pl script on my
mobycentral, the error message changes a little bit:
>[www at ml370 scripts]$ perl testMOBYClientCentral_v05.pl
>TESTING MOBY CLIENT with
> URL: http://ml370.biotech.ist.unige.it/cgi-bin/MOBY-Central.pl
> URI: http://ml370.biotech.ist.unige.it/MOBY/Central
> Proxy: No Proxy Server
>
>test 1a [PASS]
>test 1b [PASS]
>test 1 [FAIL] Relationship Type hasa does not exist in the
>biomoby.org Relationship Type system
>
>test 2 [PASS]
>test 3 [FAIL] Relationship Type hasa does not exist in the
>biomoby.org Relationship Type system
>test 4 [FAIL] Relationship type ISA does not exist in the
>relationship ontology
>test 5 [PASS]
>test 6 [FAIL] Object type $term cannot be resolved to an LSID
>test 7 [PASS]
>test 8 [FAIL] Object type $term cannot be resolved to an LSID
>test 9 [FAIL] Service Type HypotheticalService1 does not exist in
>the ontology
>test 10 [PASS]
>test 11 [PASS]
>test 12 [PASS]
>test 13 [PASS]
>test 14 [FAIL] Relationship Type isa does not exist in the
>biomoby.org Relationship Type system
>test 15 [FAIL] Relationship Type isa does not exist in the
>biomoby.org Relationship Type system
>test 16 [PASS]
>test 17 [PASS]
>test 18 [PASS]
>test 19 [PASS]
>test 20 [PASS]
>500 Internal Server Error
> ERROR ERROR ERROR
...any suggestions?
At 18.27 25/08/2005, Mark Wilkinson wrote:
>Can you first confirm where you got your mobycentral code from? Is it
>one of the newer releases? I'm certain that the latest release is
>functional (though there are a few small bugs in it that are now fixed
>in the CVS, but the CVS code isn't supported yet...)
>
>make/make install this code and see if it solves your problems. You
>probably wont need to modify any of your configuration after this
>installation.
>
>http://biomoby.org/releases/biomoby.0.8.2a.tar.gz
>
>
>M
>
>
>
>On Thu, 2005-08-25 at 12:12 +0200, scaglione wrote:
> > hello,
> > I have followed instructions to install a moby central in my machine.I
> have
> > tried the testMOBYClientCentral_v05.pl script on my mobycentral and an
> error
> > appeared:
> >
> > >perl testMOBYCentral_v05.pl
> >
> > Useless use of hash element in void context
> > at /usr/lib/perl5/site_perl/5.8.5/RDF/Core/Serializer.pm line 52.
> > Global symbol "$sth_collection_ins" requires explicit package name
> > at /usr/lib/perl5/site_perl/5.8.5/MOBY/Central.pm line 3426.
> > Compilation failed in require at testMOBYCentral_v05.pl line 4.
> > BEGIN failed--compilation aborted at testMOBYCentral_v05.pl line 4.
> >
> > what does it mean?
> > The overall configuration of the moby central seemed to be ok, since
> the test
> > of the same script on the main mobycentral works...
> > many thanks
> > silvia
> >
> >
> >
>--
>"Ontologists do it with the edges!"
>
>Mark Wilkinson
>Asst. Professor
>Dept. of Medical Genetics
>University of British Columbia
>PI in Bioinformatics
>iCAPTURE Centre
>St. Paul's Hospital
>Rm. 166, 1081 Burrard St.
>Vancouver, BC, V6Z 1Y6
>tel: 604 682 2344 x62129
>fax: 604 806 9274
PhD. Silvia Scaglione
Universita`degli Studi di Genova
Facolta' di Ingegneria, D.I.S.T. BioLab
Viale Causa, 13
16145 GENOVA - Italy
Tel Bio-lab +39 010 353 2789
Tel Centro Biotecnologie Avanzate +39 010 5737 505
Fax +39 010 353 2948
E-mail silvia77 at dist.unige.it
_______________________________________________
moby-l mailing list
moby-l at biomoby.org
http://biomoby.org/mailman/listinfo/moby-l
--
Mark Wilkinson
...on the road!