New issue?: vocabulary maintenance [vocabset-5]

[Frederic V. Hessman]

>
>>>> With UCD, there is a normative document of sorts and SKOS alone  
>>>> cannot express the additional info embodied in the "syntax  
>>>> codes", making it impossible to suggest a quick format change  
>>>> (although this problem could, of course, be solved).
> As a technical point, there's nothing stopping anyone adding extra  
> RDF properties to a SKOS vocabulary, so that
> <#poseqra> a skos:Concept;
>    skos:prefLabel "RA";
>    ucd:syntaxCode "Q".
> is perfectly legitimate.
>>>> This said, there is still a very good reason to produce a version  
>>>> of UCD in SKOS exactly so we can pretend it's a normal vocabulary.
> This is still sounding like a `because it's there' argument, and it  
> would be nice if we could have a concrete use-case (or user-story)  
> in the document, if only to illustrate the range of potential  
> vocabulary applications.  However I'm not going to press the point  
> (see below).
>>>> Every time the UCD list is updated, we insure that the SKOS  
>>>> version is too, and everybody is happy.
> But this is problematic.  Who's the `we' who's going to ensure that  
> the SKOS version keeps pace with UCD word-list changes?  Who, if  
> anyone, is going to look after these vocabularies? (but see below).


UCD is widely used in the VO and represents the most important current  
vocabulary to which many other vocabularies are going to want to  
link.  For this reason alone, there MUST be a SKOS version.

However, Norman points out correctly, that this version must be  
maintained, and the only real option is that the UCD working group do  
it.   I can't possibly see any problem with this, once our proposal is  
accepted, but this means that the UCD example in the document does  
represent an emphatic suggestion to the UCD working group that they  
take over this process.  They can then decide whether to do things  
like add attributes to cover the syntax codes and so be able to  
transfer the normative version of the vocabulary to SKOS (a logical  
thing to do, but....).

>>> Are you, then, moving towards the position that the IVOA-T  
>>> shouldn't be included in this vocabularies standard, but should be  
>>> the subject of a parallel standardisation effort?
>> This is what we decided ages ago, in fact during the process of  
>> moving from Andrea, myself, et al.'s original still-born VOcabulary  
>> proposal to the present SKOS proposal.
> We agreed this?  Some agreement on -- hell, some discussion of --  
> this issue is precisely what I've been trying to create this last  
> couple of weeks.

Sorry 'bout that.  I interpreted your enthusiasm as a helpful and  
needed impetus to get on with the process of providing a basis- 
vocabulary soon after the proposal is accepted.   We cannot let the  
IVOAT issues stop the basic proposal.


> How about the following as a resolution of issues [maintenance-4]  
> and [vocabset-5]?
>
> * We include, as part of the vocabularies standard, all six  
> vocabularies listed at <http://www.astro.gla.ac.uk/users/norman/ivoa/vocabularies/issues#vocabset-5 
> >.  Four out of the six of those have `patrons' listed next to them  
> at <http://www.ivoa.net/twiki/bin/view/IVOA/ 
> VocabulariesWorkingArea>.  These people will work off-list to get  
> the vocabularies in a state where they feel they can be made public,  
> and at some point in the next few weeks, they'll be bundled with the  
> document text as a WD.  This will resolve issue [vocabset-5].

FIVE, since A&A, AOIM, UCD1, IAU-93 and constellations will be  
finished and immediately useable.   IVOAT should be mentioned, but as  
work-in-progress and hence different.

> * Those six vocabularies will be public, and citable as part of this  
> standard, but the semantics group will not commit to maintaining  
> them further; specifically, there will be no maintenance process set  
> up as part of this standard.  We can add text to the Good Practices  
> section <http://www.astro.gla.ac.uk/users/norman/ivoa/vocabularies/vocabularies-58.xhtml#practices 
> > about curation standards, but make no requirements.  This will  
> resolve issue [maintenance-4] with the third possible resolution on  
> that page.

Obviously, A&A should be ultimately maintained by A&A, AOIM by the  
informal outreach group, UCD by the UCD working group, and the  
document should make this very clear.  We are simply "donating" the  
first versions in an attempt to boot-strap the process.

We should be able to get IAU-93 into a form which represents ALL of  
the Shobrook&Shobrook content, in which case the vocabulary is - by  
definition - normative and finished.   Should we find some minor  
errors, I'm sure we can find someone to make sure it's corrected, well- 
documented, and put back into the IVOA's (still unknown!!!) vocabulary  
collection.

> * Thinking specifically of the IVOA-T, Rick will lead a process  
> which will develop that IVOA-T, with the expectation that it'll  
> produce some citable document before this Vocabularies document is  
> finished, but not be fully finished then.  In the message I'm  
> responding to, he called for volunteers to take part in that process  
> -- I'm sure he'll have plenty of participation.  Despite that  
> process continuing after the Vocabularies document is completed, I  
> suggest that we include a snapshot of that vocabulary (call it a  
> snapshot, a beta, or an early release) as part of the Vocabualries  
> document, in _whatever_ state it's at at PR time.

No problem.

> * The UCD maintenance group might consider adopting, and  
> subsequently maintaining, the UCD vocabulary.

See the point above that this is an issue not restricted to the UCD  
vocabulary.


Rick

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ivoa.net/pipermail/semantics/attachments/20080214/321ce4ff/attachment-0001.html>