roadmap
Tony Linde
ael at star.le.ac.uk
Thu Apr 1 02:48:36 PST 2004
Hi Bob,
I guess we see v1.0 in different lights. For me, when we release a v1.0
standard we are saying it is solid, we have tested it every which way, we've
got it working in several apps and whatever we do in the future will build
on that, not scrap it and start again. And I don't think we can get to that
level of certainty in less than the timeframe I've proposed.
It is easier to junk a v0.7 standard and come up with a radically different
v0.8 standard than it is to do the same from v1.0 to v1.1 or v1.0 to v2.0.
But if you want to rewrite my proposal changing some of the v0.9x numbers to
v1.0 and the v1.0 to v2.0, please do so and we can take a look at it. But we
must make it clear that the v1.0 should not be taken as fixed and that we
reserve the right to junk it and rebuild v2.0 in the light of experience.
> Can't we make use of ADQL? Why is
> a registry so fundamentally different than another VO
> resource? A registry is all metadata, yes, but the
> mechanisms for querying it should not need to be any
> different, should they?
The registry is fundamentally different from a data source. Any data source
has an unknown schema but is stored in relational tables (or we assume it is
since ADQL is based on SQL and in most cases it is). The registry has a
known schema but completely unknown storage mechanism. So any RQL (registry
query language) needs to be based on the schema, not on storage mechanism.
Cheers,
Tony.
> -----Original Message-----
> From: Robert Hanisch [mailto:rjhanisch at worldnet.att.net]
> Sent: 01 April 2004 04:59
> To: Tony Linde; registry at ivoa.net
> Subject: Re: roadmap
>
> Tony et al.,
>
> I start with this part of Tony's message:
>
> > > I know that this seems like a long time but the fact is that
> > > standards which are widely agreed and which have been
> proved to work
> > > do take many years to develop.
>
> Yes and no. In the astronomy community, FITS is the
> archetype of doing things slowly and deliberately. FITS is
> loved and hated equally (well, maybe not equally!); the
> slowness means that FITS is not technologically very
> state-of-the-art. But it works, and as a community we are
> extremely well-served by having it. On the other hand, it
> has taken us over 10 years to reach agreements on how to
> express coordinates. 10 years. And we are not finished yet!
>
> The VO is young and still volatile. This is a plus and a
> minus. The plus is we should be free to experiment, the
> minus is that we need to build Something, Now, so that the
> astronomy community knows what we are up to and buys into
> what we are doing.
>
> We need to reach agreements quickly, build to them, and
> iterate to improve them. We need to be willing to build and
> revise, or even in some cases throw away and start again. We
> have set up a standards process that recognizes change, and
> is intended to be responsive to change. I do not believe we
> can afford to set 2-3 year schedules for reaching consensus.
> If it takes this long, the community upon which we will
> utimately depend for support will see us as purveyors of
> snake-oil, and will blow us off.
>
> > > 2004
> > > ====
> > > 1. Publish RM V1.0 to REC status
> > > 31-Mar-2004
> > >
> > > 2. Agree modified Resource Metadata Schema (RMS) draft v0.91
> Whatever the number, this should be ready to promote to V1.0
> this summer.
> > > 30-May-2004
> > >
> > > 3. Agree modified Registry Harvesting draft (RH) v0.2
> > > 30-May-2004
> Ditto.
> > >
> > > 4. Demonstrate viability of RMS v0.91 and RH v0.2
> between projects
> > > 1-Jun-2004 to 30-Sep-2004
> ok. We validate RMS and RH over the summer, and bring them
> forward to RECs by 1 Sep 04 (bring forward by a month).
> > >
> > > 5. Create draft Registry Interface spec (RI) v0.1 (incl
> harvesting
> > > & query)
> > > 30-Nov-2004
> What is this? A query specification? We already have
> harvesting given above. And we need a query spec in parallel
> with the above if we are to make use of the registry,
> otherwise it is a blackhole that we pour info into and have
> no way of getting it out. Can't we make use of ADQL? Why is
> a registry so fundamentally different than another VO
> resource? A registry is all metadata, yes, but the
> mechanisms for querying it should not need to be any
> different, should they?
>
> So I am not sure what to say about the rest, except that the
> event horizon is way too far away. We need to push ourselves
> harder than this, accepting incompleteness and revisions, but
> developing and demonstrating registry capabilties --
> metadata, schema, harvesting, querying -- in ~6 months.
>
> We have some other very big challenges to deal with, not at
> all mentioned here. Curation, revision control,
> synchronization of distributed registries. We have an NVO
> registry with harvested information from a number of
> publishing registries, and it has 4000 or 5000 entries. Most
> entries are next to useless because the key metadata fields
> have either not been populated at all, or have been populated
> blindly with inappropriate information. In some ways even
> worse, well-intentioned content providers have assigned
> resource Titles and Shortnames and Descriptions that make no
> sense, that conflict with astronomy community understanding/
> expectation, and generally defeat the purpose of the
> registry. Our best-made plans for metadata elements (RM) and
> their encoding (RMS) are worthless if we have no way to
> review, endorse, validate, and update such information.
>
> Bob
>
> > > 6. Agree draft RMS v0.92
> > > 30-Nov-2004
> > >
> > > 2005
> > > ====
> > > 7. Demonstrate viability of RMS v0.92 and RI v0.1
> > > 1-Dec-2004 to 31-Mar-2005
> > >
> > > 8. Develop draft RM v1.1 (incorporating schema)
> > > 31-May-2005
> > >
> > > 9. Agree modified RI v0.2
> > > 31-May-2005
> > >
> > > 10. Demonstrate RM v1.1 and RI v0.2
> > > 1-Jun-2005 to 30-Sep-2005
> > >
> > > 11. Publish RM v1.1 to PR/REC
> > > 31-Dec-2005
> > >
> > > 12. Publish RI v1.0 to PR/REC
> > > 31-Dec-2005
> > >
> > > I know that this seems like a long time but the fact is that
> > > standards which are widely agreed and which have been
> proved to work
> > > do take many years to develop.
> > >
> > > The key to keeping this effort moving forwards is that we
> continue
> > > to develop working versions of the registry schema and the
> > > harvesting interface. This will allow us to prove that these
> > > standards work and to find the problem areas.
> > >
> > > Feel free to publish an alternative timetable if you think we can
> > > deliver the standards in less time than I've indicated.
> > > And I look forward to other comments as well.
> > >
> > > Cheers,
> > > Tony.
> > >
> > > __
> > > Tony Linde
> > > Phone: +44 (0)116 223 1292 Mobile: +44 (0)7753 603356
> > > Fax: +44 (0)116 252 3311 Email: ael at star.le.ac.uk
> > > Post: Department of Physics & Astronomy,
> > > University of Leicester
> > > Leicester, UK LE1 7RH
> > >
> > > Project Manager, Director,
> > > AstroGrid Leicester e-Science Centre
> > > http://www.astrogrid.org http://www.e-science.le.ac.uk/
> > >
> >
> >
>
More information about the registry
mailing list