From andrea.preitemartinez at rm.iasf.cnr.it Tue Oct 18 08:52:25 2005 From: andrea.preitemartinez at rm.iasf.cnr.it (Andrea Preite Martinez) Date: Tue, 18 Oct 2005 17:52:25 +0200 Subject: IVOA Note on UCD maintenance Message-ID: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> Dear member, at the recent InterOp meeting in Madrid I got the action to prepare a document desribing in detail the procedure to follow in order to maintain the standard list of UCD words. Together with Sebastien we wrote a draft document that I uploaded in the IVOA document area as a Note: "Maintenance of the list of UCD words. Ver.1.0" ( http://www.ivoa.net/Documents/Notes/ ... I still don't have the link to the note) Before sending it to the TCC I'd appreciate comments on the text !! Regards Andrea ============================================================================== Andrea Preite Martinez andrea.preitemartinez at rm.iasf.cnr.it Istituto di Astrofisica Spaziale Tel.:+39.06.4993.4641 Area di Ricerca di Tor Vergata Fax.:+39.06.2066.0188 Via del Fosso del Cavaliere 100 Cell:+39.339.3817355 00133 Roma CDS :+33.3.90242473 ============================================================================== From anai_gomez at mat.ucm.es Tue Oct 18 10:53:51 2005 From: anai_gomez at mat.ucm.es (Ana Ines Gomez de Castro) Date: Tue, 18 Oct 2005 19:53:51 +0200 (CEST) Subject: IVOA Note on UCD maintenance In-Reply-To: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> Message-ID: Dear Andrea, I've read the document and have some questions (maybe because I didn't attend the InterOp meeting). Here they are! 1. Why do you need two bodies "the scientific board" and the "semantics WG" to approve a new UCD?. Do you envission the SciBoard a "technical body" and the "Semantics WG" a "political-coordinating body"? >From our previous experience, I think that to accept a new UCD we require: - To test that the proposed new UCD cannot be made as a combination of the already existing. - To check that it is scientifically meaningful Is the SciBoard going to be responsible of both duties? 2. I assume that the rigidity in the updating of the UCD list (the list is updated during the InterOp meetings) is caused by the fact that once a new UCD is accepted the data providers have to include it in their "protocols", am I right? 3. Concerning the web based form... don't you think that it would be good to have a Comments field where the "history" of the UCD approval/disapproval is recorded?. It is feasible that a scientist, after receiving the notification of "rejection" answers to the SciBoard providing new reasons for the inclusion of the UCD word, eventually some of these reasons maybe relevant and the SciBoard may reconsider the rejection. I have not the slighlest idea about the degree of activity we might expect from the Community BUT it maybe a good policy to keep all the "rejected UCDs" in data base so, any time someone wish to propose a new UCD, the data base can be cross-checked so some time can be saved to the Chair of the SciBoard. Maybe this list could be accessible to the Community. 4. I do not understand well why on the on hand, it may be necessary to insert rapidly a new UCD (with the "temporary namespace "ivoa:"") and at the same time accept that the list of UCD's is updated in time-scales of InterOp meetings. Hope it helps!. Please, do not waste time answering the questions that have been already discussed at the InterOp meetings, just let me know. Cheers, Ana Prof. Ana I. Gomez de Castro Instituto de Astronomia y Geodesia (CSIC-UCM) Fac. de CC Matematicas Universidad Complutense de Madrid 28040 Madrid, Spain Ph. 34-913944578 (office) 34-659783338 (mobile) Email: aig at mat.ucm.es On Tue, 18 Oct 2005, Andrea Preite Martinez wrote: > Dear member, > > at the recent InterOp meeting in Madrid I got the action to prepare a > document desribing in detail the procedure to follow in order to maintain > the standard list of UCD words. > > Together with Sebastien we wrote a draft document that I uploaded in the > IVOA document area as a Note: "Maintenance of the list of UCD words. > Ver.1.0" > ( http://www.ivoa.net/Documents/Notes/ ... I still don't have the link to > the note) > > Before sending it to the TCC I'd appreciate comments on the text !! > > Regards > > Andrea > > > ============================================================================== > Andrea Preite Martinez andrea.preitemartinez at rm.iasf.cnr.it > Istituto di Astrofisica Spaziale Tel.:+39.06.4993.4641 > Area di Ricerca di Tor Vergata Fax.:+39.06.2066.0188 > Via del Fosso del Cavaliere 100 Cell:+39.339.3817355 > 00133 Roma CDS :+33.3.90242473 > ============================================================================== > > -- Prof. Ana I. Gomez de Castro Instituto de Astronomia y Geodesia (CSIC-UCM) Fac. de CC Matematicas Universidad Complutense de Madrid 28040 Madrid, Spain Ph. 34-913944578 (office) 34-659783338 (mobile) Email: aig at mat.ucm.es From ndelmott at eso.org Tue Oct 18 23:50:17 2005 From: ndelmott at eso.org (Nausicaa Delmotte) Date: Wed, 19 Oct 2005 08:50:17 +0200 Subject: IVOA Note on UCD maintenance In-Reply-To: References: Message-ID: <4355ECA9.1010805@eso.org> Dear all, Ana Ines Gomez de Castro wrote: > 3. Concerning the web based form... don't you think that it would be > good to have a Comments field where the "history" of the UCD > approval/disapproval is recorded?. It is feasible that a scientist, > after receiving the notification of "rejection" answers to the SciBoard > providing new reasons for the inclusion of the UCD word, eventually some > of these reasons maybe relevant and the SciBoard may reconsider the > rejection. > > I have not the slighlest idea about the degree of activity we might > expect from the Community BUT it maybe a good policy to keep all the > "rejected UCDs" in data base so, any time someone wish to propose a new > UCD, the data base can be cross-checked so some time can be saved to > the Chair of the SciBoard. Maybe this list could be accessible to the > Community. Yes, keeping track of the rejected UCD words in a database/file/... would be indeed very nice for future reference. Else, or in addition, maybe we could add a 'status flag' field to the form, so that we could 'sort by' status (rejected, temporary, accepted) and date. Cheers Nausicaa From marco.leoni at eso.org Wed Oct 19 00:39:36 2005 From: marco.leoni at eso.org (Marco C. Leoni) Date: Wed, 19 Oct 2005 09:39:36 +0200 Subject: IVOA Note on UCD maintenance Message-ID: <1129707576.4355f83838e02@webmail.eso.org> The Note can be found here: http://www.ivoa.net/Documents/latest/UCDmaintenance.html Cheers, Marco From Pedro.Osuna at sciops.esa.int Wed Oct 19 02:48:22 2005 From: Pedro.Osuna at sciops.esa.int (Pedro Osuna) Date: Wed, 19 Oct 2005 11:48:22 +0200 Subject: IVOA Note on UCD maintenance In-Reply-To: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> References: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> Message-ID: <1129715301.23228.1769.camel@isol03.vilspa.esa.int> Dear Andrea, I think the procedure is quite thorough and reflects all the steps we suggested in the initial document on UCD. I am worried though about allowing the identifier "ivoa:" as namespace for those UCDs already approved but temporarily out of the UCD standard (as this one will only be updated once or twice a year). What this might end up is with loads of systems using not-yet-standard UCDs, different from one proposal to the other, and all of them starting with the same identifier "ivoa:". Even worse, probably the data provider will write the "ivoa:" ucd, and probably won't change it to the "real" one when the UCD is standardised. A possible solution for this problem would be to allow for specific namespaces for specific needs (let's say, like a "solar:" namespace for solar physics specific UCDs) but this again seems to go in the debate on whether this should be relegated to data models or not, which I think is not worth to retake now. Just my two cents. Cheers, P. On Tue, 2005-10-18 at 17:52, Andrea Preite Martinez wrote: > Dear member, > > at the recent InterOp meeting in Madrid I got the action to prepare a > document desribing in detail the procedure to follow in order to maintain > the standard list of UCD words. > > Together with Sebastien we wrote a draft document that I uploaded in the > IVOA document area as a Note: "Maintenance of the list of UCD words. > Ver.1.0" > ( http://www.ivoa.net/Documents/Notes/ ... I still don't have the link to > the note) > > Before sending it to the TCC I'd appreciate comments on the text !! > > Regards > > Andrea > > > ============================================================================== > Andrea Preite Martinez andrea.preitemartinez at rm.iasf.cnr.it > Istituto di Astrofisica Spaziale Tel.:+39.06.4993.4641 > Area di Ricerca di Tor Vergata Fax.:+39.06.2066.0188 > Via del Fosso del Cavaliere 100 Cell:+39.339.3817355 > 00133 Roma CDS :+33.3.90242473 > ============================================================================== -- Pedro Osuna Alcalaya ESA Science Archives System Engineer Science Archive Team European Space Astronomy Centre (ESAC/ESA) e-mail: Pedro.Osuna at esa.int Tel + 34 91 8131314 --------------------------------- European Space Astronomy Centre European Space Agency P.O. Box 50727 E-28080 Villafranca del Castillo MADRID - SPAIN From rgm at roe.ac.uk Wed Oct 19 03:10:12 2005 From: rgm at roe.ac.uk (Bob Mann) Date: Wed, 19 Oct 2005 11:10:12 +0100 (BST) Subject: IVOA Note on UCD maintenance In-Reply-To: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> References: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> Message-ID: Hi folks, My only comment on this is to ask whether it is better to have each RFM sent out to the Scientific Board as soon as it is submitted or whether they should be collected over a period (three months?) and have the Scientific Board then debate those RFMs which have accumulated during that period. My question is really based on own experience and working habits. If I see an email arriving about a single proposed UCD amendment I'm probably unlikely to take time to consider it if I'm busy, whereas, if I know I have a few hours' work to do every quarter I'm more likely to set aside a period of time for doing that work properly. What do other people think? cheers Bob From francois at vizir.u-strasbg.fr Wed Oct 19 04:55:00 2005 From: francois at vizir.u-strasbg.fr (Francois Ochsenbein) Date: Wed, 19 Oct 2005 13:55:00 +0200 Subject: IVOA Note on UCD maintenance In-Reply-To: Your message of Wed, 19 Oct 2005 11:48:22 +0200 . Message-ID: <200510191155.j9JBt0224839@vizir.u-strasbg.fr> Hi Andrea, I'm sharing exactly the same worries about the ivoa: prefix -- it does not simplify anything, and will just reduce the IVOA-wide interoperability, at least on the mid-term if not on the long term... Cheers, Francois > >Dear Andrea, > > >I think the procedure is quite thorough and reflects all the steps we >suggested in the initial document on UCD. > >I am worried though about allowing the identifier "ivoa:" as namespace >for those UCDs already approved but temporarily out of the UCD standard >(as this one will only be updated once or twice a year). > >What this might end up is with loads of systems using not-yet-standard >UCDs, different from one proposal to the other, and all of them starting >with the same identifier "ivoa:". Even worse, probably the data provider >will write the "ivoa:" ucd, and probably won't change it to the "real" >one when the UCD is standardised. > >A possible solution for this problem would be to allow for specific >namespaces for specific needs (let's say, like a "solar:" namespace for >solar physics specific UCDs) but this again seems to go in the debate on >whether this should be relegated to data models or not, which I think is >not worth to retake now. > >Just my two cents. > >Cheers, >P. > > ================================================================================ Francois Ochsenbein ------ Observatoire Astronomique de Strasbourg 11, rue de l'Universite F-67000 STRASBOURG Phone: +33-(0)390 24 24 29 Email: francois at astro.u-strasbg.fr (France) Fax: +33-(0)390 24 24 32 ================================================================================ From seaman at noao.edu Wed Oct 19 07:03:21 2005 From: seaman at noao.edu (Rob Seaman) Date: Wed, 19 Oct 2005 07:03:21 -0700 Subject: IVOA Note on UCD maintenance In-Reply-To: <1129715301.23228.1769.camel@isol03.vilspa.esa.int> References: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> <1129715301.23228.1769.camel@isol03.vilspa.esa.int> Message-ID: <14DE598C-6B6D-41EB-BDB9-CCAE567AE8B9@noao.edu> Pedro Osuna wrote: > I am worried though about allowing the identifier "ivoa:" as namespace > for those UCDs already approved but temporarily out of the UCD > standard > (as this one will only be updated once or twice a year). Section 2 of the v1.1 UCD standard seems to state that "ivoa;" is equivalent to the default namespace. It is unclear what distinction is being drawn here between "approved" and "not yet standard". On the other hand, I am very heartened that the UCD committee(s) have reached the point of recognizing that at least some alternative namespaces are necessary. > A possible solution for this problem would be to allow for specific > namespaces for specific needs (let's say, like a "solar:" namespace > for > solar physics specific UCDs) The standard already does: "The use of namespaces, indicated by the presence of a colon in the word, is possible..." This is also discussed in section 2, which goes on to attempt to dissuade their use: "but should be avoided as far as possible." I believe this to be a vain hope. The way to control the looming complexity of UCDs is precisely to spread the responsibility (and pain) around. This reminds me of the "image" extension specification for FITS, which tries to convince users that it is both bad - and worse, unnecessary - to create data-less primary header data units. What did the community immediately do with image extensions? I'm tempted to leave that as a rhetorical question - but what we did was to discover an excellent usage for subclassing metadata between the primary headers and the extension headers and for stuffing all the data into the extensions. UCD namespaces are inevitable. Bob Mann wrote: > My only comment on this is to ask whether it is better to have > each RFM sent out to the Scientific Board as soon as it is submitted > or whether they should be collected over a period (three months?) By all means, although I might suggest a monthly cadence. There should also be a mechanism for expedited discussions for UCDs which are needed "yesterday". Note, however, that the pressure for immediate decision-making would also be relieved by the availability of namespaces. Rob Seaman National Optical Astronomy Observatory From tam at lheapop.gsfc.nasa.gov Wed Oct 19 07:08:55 2005 From: tam at lheapop.gsfc.nasa.gov (Thomas McGlynn) Date: Wed, 19 Oct 2005 10:08:55 -0400 Subject: IVOA Note on UCD maintenance In-Reply-To: <1129707576.4355f83838e02@webmail.eso.org> References: <1129707576.4355f83838e02@webmail.eso.org> Message-ID: <43565377.7010107@lheapop.gsfc.nasa.gov> Marco C. Leoni wrote: > The Note can be found here: http://www.ivoa.net/Documents/latest/UCDmaintenance.html > > Cheers, > Marco > > In addition to the procedure for approving changes, I'd suggest one needs to be clearer about the nature of the change that is being contemplated. Here are some of the kinds of changes I can imagine: Deletion: The UCD word is to be removed from the list of standard UCDs. Normally a deletion request should include a statement of what UCD would be used in place of the original UCD. A list of deleted UCDs (e.g., all of UCD1) and their replacements should be maintained. Another example here might be the FITS 'EPOCH' keyword which was originally misdefined and is now deprecated and replaced by EQUINOX. There is a pending request to replace UCDs for mean and variance with the moments of various orders. Modification: The role of the UCD word is being changed (preferably expanded so that existing uses of the word are not made invalid). E.g., this might come up when a concept is originally split into too fine granularity and it's decided that in the future only one of a set of original UCD's should be used. Amendment: The description of the UCD is to be corrected, but its intended usage is not changed. Addition: A new UCD word to be added. This is what most of the fuss has been about. In addition to the general description, I believe that there needs to be some very specific roles about how this word can be combined with other words. Can it be a primary word or is it primarily intended as a modifier? This last point leads to the issue that so far we have discussed the maintenance of the vocabulary only at the level of UCD words. Some management of the use of atoms and maintained guidelines on how given words can be combined into UCDs are probably also needed. E.g., I'd suggest that re-use of atom strings be forbidden or at least be strongly discouraged. A little nit ... Looking at the proposed UCDs I'd like to ask that some standard on the capitalization of UCDs be promulgated. There are suggestions for pos.htm (as the Hierarchical Triangular Mesh index) and phys.IMF (as the Initial Mass Function) Personally I would prefer lower case in both of these, but so long as it's consistent I don't care much. Tom McGlynn From seaman at noao.edu Mon Oct 24 15:51:35 2005 From: seaman at noao.edu (Rob Seaman) Date: Mon, 24 Oct 2005 15:51:35 -0700 Subject: UCD versioning? Message-ID: <6E25F7C3-4800-48FE-8A21-A3CC64C5C09D@noao.edu> Hi, Could somebody comment on the relationship between http:// www.ivoa.net/internal/IVOA/IvoaUCD/UCD-1.9.9b.pdf and http:// www.ivoa.net/Documents/latest/UCD.html? Is the former obsolete? It hasn't been updated in two years. In any event, I don't see any discussion of a versioning strategy. Unfortunately, I also can't seem to get a search to work on the list archive (http://www.ivoa.net/forum/ucd). Surely this topic has been discussed (likely in gory detail)? Or is there some notion of "once UCD, always UCD"? The recent discussion about UCD maintenance seems to me to be orthogonal to the question of explicit versioning. Are new UCDs simply added to the end of a list that preserves no historical memory? Or are new UCDs appended to only the latest version of the list? Are UCDs ever removed from the list? Are their names or locations in the hierarchy never to be changed? And is the interpretation of a document that uses UCDs going to vary as the master list changes over time? The process for adopting UCDs is intended to be very conservative, but even so, one imagines that some UCDs will be adopted that will later be rethought. Note that even if the UCDs associated with a particular document (or image or catalog) don't change, that the nuances of the interpretation of these items might change as new - unselected - UCD options are added to the master list. An assertion versus three options may turn into an assertion versus six options. I gather there is some thought of occupying UCD-space only until we enter the promised land of the ontology-sphere. One might be forgiven, perhaps, for imagining that this may well take a sizable fraction of the Biblical 40 years of wandering, however. That being the case, what provisions are in place to allow UCDs to thrive in the ontology-free wilderness? Could someone summarize how UCD versioning is - or is not - supposed to work? Alternately, feel free to point me toward pertinent discussions. Thanks! Rob Seaman NOAO From andrea.preitemartinez at rm.iasf.cnr.it Tue Oct 18 08:52:25 2005 From: andrea.preitemartinez at rm.iasf.cnr.it (Andrea Preite Martinez) Date: Tue, 18 Oct 2005 17:52:25 +0200 Subject: IVOA Note on UCD maintenance Message-ID: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> Dear member, at the recent InterOp meeting in Madrid I got the action to prepare a document desribing in detail the procedure to follow in order to maintain the standard list of UCD words. Together with Sebastien we wrote a draft document that I uploaded in the IVOA document area as a Note: "Maintenance of the list of UCD words. Ver.1.0" ( http://www.ivoa.net/Documents/Notes/ ... I still don't have the link to the note) Before sending it to the TCC I'd appreciate comments on the text !! Regards Andrea ============================================================================== Andrea Preite Martinez andrea.preitemartinez at rm.iasf.cnr.it Istituto di Astrofisica Spaziale Tel.:+39.06.4993.4641 Area di Ricerca di Tor Vergata Fax.:+39.06.2066.0188 Via del Fosso del Cavaliere 100 Cell:+39.339.3817355 00133 Roma CDS :+33.3.90242473 ============================================================================== From anai_gomez at mat.ucm.es Tue Oct 18 10:53:51 2005 From: anai_gomez at mat.ucm.es (Ana Ines Gomez de Castro) Date: Tue, 18 Oct 2005 19:53:51 +0200 (CEST) Subject: IVOA Note on UCD maintenance In-Reply-To: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> Message-ID: Dear Andrea, I've read the document and have some questions (maybe because I didn't attend the InterOp meeting). Here they are! 1. Why do you need two bodies "the scientific board" and the "semantics WG" to approve a new UCD?. Do you envission the SciBoard a "technical body" and the "Semantics WG" a "political-coordinating body"? >From our previous experience, I think that to accept a new UCD we require: - To test that the proposed new UCD cannot be made as a combination of the already existing. - To check that it is scientifically meaningful Is the SciBoard going to be responsible of both duties? 2. I assume that the rigidity in the updating of the UCD list (the list is updated during the InterOp meetings) is caused by the fact that once a new UCD is accepted the data providers have to include it in their "protocols", am I right? 3. Concerning the web based form... don't you think that it would be good to have a Comments field where the "history" of the UCD approval/disapproval is recorded?. It is feasible that a scientist, after receiving the notification of "rejection" answers to the SciBoard providing new reasons for the inclusion of the UCD word, eventually some of these reasons maybe relevant and the SciBoard may reconsider the rejection. I have not the slighlest idea about the degree of activity we might expect from the Community BUT it maybe a good policy to keep all the "rejected UCDs" in data base so, any time someone wish to propose a new UCD, the data base can be cross-checked so some time can be saved to the Chair of the SciBoard. Maybe this list could be accessible to the Community. 4. I do not understand well why on the on hand, it may be necessary to insert rapidly a new UCD (with the "temporary namespace "ivoa:"") and at the same time accept that the list of UCD's is updated in time-scales of InterOp meetings. Hope it helps!. Please, do not waste time answering the questions that have been already discussed at the InterOp meetings, just let me know. Cheers, Ana Prof. Ana I. Gomez de Castro Instituto de Astronomia y Geodesia (CSIC-UCM) Fac. de CC Matematicas Universidad Complutense de Madrid 28040 Madrid, Spain Ph. 34-913944578 (office) 34-659783338 (mobile) Email: aig at mat.ucm.es On Tue, 18 Oct 2005, Andrea Preite Martinez wrote: > Dear member, > > at the recent InterOp meeting in Madrid I got the action to prepare a > document desribing in detail the procedure to follow in order to maintain > the standard list of UCD words. > > Together with Sebastien we wrote a draft document that I uploaded in the > IVOA document area as a Note: "Maintenance of the list of UCD words. > Ver.1.0" > ( http://www.ivoa.net/Documents/Notes/ ... I still don't have the link to > the note) > > Before sending it to the TCC I'd appreciate comments on the text !! > > Regards > > Andrea > > > ============================================================================== > Andrea Preite Martinez andrea.preitemartinez at rm.iasf.cnr.it > Istituto di Astrofisica Spaziale Tel.:+39.06.4993.4641 > Area di Ricerca di Tor Vergata Fax.:+39.06.2066.0188 > Via del Fosso del Cavaliere 100 Cell:+39.339.3817355 > 00133 Roma CDS :+33.3.90242473 > ============================================================================== > > -- Prof. Ana I. Gomez de Castro Instituto de Astronomia y Geodesia (CSIC-UCM) Fac. de CC Matematicas Universidad Complutense de Madrid 28040 Madrid, Spain Ph. 34-913944578 (office) 34-659783338 (mobile) Email: aig at mat.ucm.es From ndelmott at eso.org Tue Oct 18 23:50:17 2005 From: ndelmott at eso.org (Nausicaa Delmotte) Date: Wed, 19 Oct 2005 08:50:17 +0200 Subject: IVOA Note on UCD maintenance In-Reply-To: References: Message-ID: <4355ECA9.1010805@eso.org> Dear all, Ana Ines Gomez de Castro wrote: > 3. Concerning the web based form... don't you think that it would be > good to have a Comments field where the "history" of the UCD > approval/disapproval is recorded?. It is feasible that a scientist, > after receiving the notification of "rejection" answers to the SciBoard > providing new reasons for the inclusion of the UCD word, eventually some > of these reasons maybe relevant and the SciBoard may reconsider the > rejection. > > I have not the slighlest idea about the degree of activity we might > expect from the Community BUT it maybe a good policy to keep all the > "rejected UCDs" in data base so, any time someone wish to propose a new > UCD, the data base can be cross-checked so some time can be saved to > the Chair of the SciBoard. Maybe this list could be accessible to the > Community. Yes, keeping track of the rejected UCD words in a database/file/... would be indeed very nice for future reference. Else, or in addition, maybe we could add a 'status flag' field to the form, so that we could 'sort by' status (rejected, temporary, accepted) and date. Cheers Nausicaa From marco.leoni at eso.org Wed Oct 19 00:39:36 2005 From: marco.leoni at eso.org (Marco C. Leoni) Date: Wed, 19 Oct 2005 09:39:36 +0200 Subject: IVOA Note on UCD maintenance Message-ID: <1129707576.4355f83838e02@webmail.eso.org> The Note can be found here: http://www.ivoa.net/Documents/latest/UCDmaintenance.html Cheers, Marco From Pedro.Osuna at sciops.esa.int Wed Oct 19 02:48:22 2005 From: Pedro.Osuna at sciops.esa.int (Pedro Osuna) Date: Wed, 19 Oct 2005 11:48:22 +0200 Subject: IVOA Note on UCD maintenance In-Reply-To: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> References: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> Message-ID: <1129715301.23228.1769.camel@isol03.vilspa.esa.int> Dear Andrea, I think the procedure is quite thorough and reflects all the steps we suggested in the initial document on UCD. I am worried though about allowing the identifier "ivoa:" as namespace for those UCDs already approved but temporarily out of the UCD standard (as this one will only be updated once or twice a year). What this might end up is with loads of systems using not-yet-standard UCDs, different from one proposal to the other, and all of them starting with the same identifier "ivoa:". Even worse, probably the data provider will write the "ivoa:" ucd, and probably won't change it to the "real" one when the UCD is standardised. A possible solution for this problem would be to allow for specific namespaces for specific needs (let's say, like a "solar:" namespace for solar physics specific UCDs) but this again seems to go in the debate on whether this should be relegated to data models or not, which I think is not worth to retake now. Just my two cents. Cheers, P. On Tue, 2005-10-18 at 17:52, Andrea Preite Martinez wrote: > Dear member, > > at the recent InterOp meeting in Madrid I got the action to prepare a > document desribing in detail the procedure to follow in order to maintain > the standard list of UCD words. > > Together with Sebastien we wrote a draft document that I uploaded in the > IVOA document area as a Note: "Maintenance of the list of UCD words. > Ver.1.0" > ( http://www.ivoa.net/Documents/Notes/ ... I still don't have the link to > the note) > > Before sending it to the TCC I'd appreciate comments on the text !! > > Regards > > Andrea > > > ============================================================================== > Andrea Preite Martinez andrea.preitemartinez at rm.iasf.cnr.it > Istituto di Astrofisica Spaziale Tel.:+39.06.4993.4641 > Area di Ricerca di Tor Vergata Fax.:+39.06.2066.0188 > Via del Fosso del Cavaliere 100 Cell:+39.339.3817355 > 00133 Roma CDS :+33.3.90242473 > ============================================================================== -- Pedro Osuna Alcalaya ESA Science Archives System Engineer Science Archive Team European Space Astronomy Centre (ESAC/ESA) e-mail: Pedro.Osuna at esa.int Tel + 34 91 8131314 --------------------------------- European Space Astronomy Centre European Space Agency P.O. Box 50727 E-28080 Villafranca del Castillo MADRID - SPAIN From rgm at roe.ac.uk Wed Oct 19 03:10:12 2005 From: rgm at roe.ac.uk (Bob Mann) Date: Wed, 19 Oct 2005 11:10:12 +0100 (BST) Subject: IVOA Note on UCD maintenance In-Reply-To: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> References: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> Message-ID: Hi folks, My only comment on this is to ask whether it is better to have each RFM sent out to the Scientific Board as soon as it is submitted or whether they should be collected over a period (three months?) and have the Scientific Board then debate those RFMs which have accumulated during that period. My question is really based on own experience and working habits. If I see an email arriving about a single proposed UCD amendment I'm probably unlikely to take time to consider it if I'm busy, whereas, if I know I have a few hours' work to do every quarter I'm more likely to set aside a period of time for doing that work properly. What do other people think? cheers Bob From francois at vizir.u-strasbg.fr Wed Oct 19 04:55:00 2005 From: francois at vizir.u-strasbg.fr (Francois Ochsenbein) Date: Wed, 19 Oct 2005 13:55:00 +0200 Subject: IVOA Note on UCD maintenance In-Reply-To: Your message of Wed, 19 Oct 2005 11:48:22 +0200 . Message-ID: <200510191155.j9JBt0224839@vizir.u-strasbg.fr> Hi Andrea, I'm sharing exactly the same worries about the ivoa: prefix -- it does not simplify anything, and will just reduce the IVOA-wide interoperability, at least on the mid-term if not on the long term... Cheers, Francois > >Dear Andrea, > > >I think the procedure is quite thorough and reflects all the steps we >suggested in the initial document on UCD. > >I am worried though about allowing the identifier "ivoa:" as namespace >for those UCDs already approved but temporarily out of the UCD standard >(as this one will only be updated once or twice a year). > >What this might end up is with loads of systems using not-yet-standard >UCDs, different from one proposal to the other, and all of them starting >with the same identifier "ivoa:". Even worse, probably the data provider >will write the "ivoa:" ucd, and probably won't change it to the "real" >one when the UCD is standardised. > >A possible solution for this problem would be to allow for specific >namespaces for specific needs (let's say, like a "solar:" namespace for >solar physics specific UCDs) but this again seems to go in the debate on >whether this should be relegated to data models or not, which I think is >not worth to retake now. > >Just my two cents. > >Cheers, >P. > > ================================================================================ Francois Ochsenbein ------ Observatoire Astronomique de Strasbourg 11, rue de l'Universite F-67000 STRASBOURG Phone: +33-(0)390 24 24 29 Email: francois at astro.u-strasbg.fr (France) Fax: +33-(0)390 24 24 32 ================================================================================ From seaman at noao.edu Wed Oct 19 07:03:21 2005 From: seaman at noao.edu (Rob Seaman) Date: Wed, 19 Oct 2005 07:03:21 -0700 Subject: IVOA Note on UCD maintenance In-Reply-To: <1129715301.23228.1769.camel@isol03.vilspa.esa.int> References: <20051018175225.oye1p5on6kq8s84s@webmail.sic.rm.cnr.it> <1129715301.23228.1769.camel@isol03.vilspa.esa.int> Message-ID: <14DE598C-6B6D-41EB-BDB9-CCAE567AE8B9@noao.edu> Pedro Osuna wrote: > I am worried though about allowing the identifier "ivoa:" as namespace > for those UCDs already approved but temporarily out of the UCD > standard > (as this one will only be updated once or twice a year). Section 2 of the v1.1 UCD standard seems to state that "ivoa;" is equivalent to the default namespace. It is unclear what distinction is being drawn here between "approved" and "not yet standard". On the other hand, I am very heartened that the UCD committee(s) have reached the point of recognizing that at least some alternative namespaces are necessary. > A possible solution for this problem would be to allow for specific > namespaces for specific needs (let's say, like a "solar:" namespace > for > solar physics specific UCDs) The standard already does: "The use of namespaces, indicated by the presence of a colon in the word, is possible..." This is also discussed in section 2, which goes on to attempt to dissuade their use: "but should be avoided as far as possible." I believe this to be a vain hope. The way to control the looming complexity of UCDs is precisely to spread the responsibility (and pain) around. This reminds me of the "image" extension specification for FITS, which tries to convince users that it is both bad - and worse, unnecessary - to create data-less primary header data units. What did the community immediately do with image extensions? I'm tempted to leave that as a rhetorical question - but what we did was to discover an excellent usage for subclassing metadata between the primary headers and the extension headers and for stuffing all the data into the extensions. UCD namespaces are inevitable. Bob Mann wrote: > My only comment on this is to ask whether it is better to have > each RFM sent out to the Scientific Board as soon as it is submitted > or whether they should be collected over a period (three months?) By all means, although I might suggest a monthly cadence. There should also be a mechanism for expedited discussions for UCDs which are needed "yesterday". Note, however, that the pressure for immediate decision-making would also be relieved by the availability of namespaces. Rob Seaman National Optical Astronomy Observatory From tam at lheapop.gsfc.nasa.gov Wed Oct 19 07:08:55 2005 From: tam at lheapop.gsfc.nasa.gov (Thomas McGlynn) Date: Wed, 19 Oct 2005 10:08:55 -0400 Subject: IVOA Note on UCD maintenance In-Reply-To: <1129707576.4355f83838e02@webmail.eso.org> References: <1129707576.4355f83838e02@webmail.eso.org> Message-ID: <43565377.7010107@lheapop.gsfc.nasa.gov> Marco C. Leoni wrote: > The Note can be found here: http://www.ivoa.net/Documents/latest/UCDmaintenance.html > > Cheers, > Marco > > In addition to the procedure for approving changes, I'd suggest one needs to be clearer about the nature of the change that is being contemplated. Here are some of the kinds of changes I can imagine: Deletion: The UCD word is to be removed from the list of standard UCDs. Normally a deletion request should include a statement of what UCD would be used in place of the original UCD. A list of deleted UCDs (e.g., all of UCD1) and their replacements should be maintained. Another example here might be the FITS 'EPOCH' keyword which was originally misdefined and is now deprecated and replaced by EQUINOX. There is a pending request to replace UCDs for mean and variance with the moments of various orders. Modification: The role of the UCD word is being changed (preferably expanded so that existing uses of the word are not made invalid). E.g., this might come up when a concept is originally split into too fine granularity and it's decided that in the future only one of a set of original UCD's should be used. Amendment: The description of the UCD is to be corrected, but its intended usage is not changed. Addition: A new UCD word to be added. This is what most of the fuss has been about. In addition to the general description, I believe that there needs to be some very specific roles about how this word can be combined with other words. Can it be a primary word or is it primarily intended as a modifier? This last point leads to the issue that so far we have discussed the maintenance of the vocabulary only at the level of UCD words. Some management of the use of atoms and maintained guidelines on how given words can be combined into UCDs are probably also needed. E.g., I'd suggest that re-use of atom strings be forbidden or at least be strongly discouraged. A little nit ... Looking at the proposed UCDs I'd like to ask that some standard on the capitalization of UCDs be promulgated. There are suggestions for pos.htm (as the Hierarchical Triangular Mesh index) and phys.IMF (as the Initial Mass Function) Personally I would prefer lower case in both of these, but so long as it's consistent I don't care much. Tom McGlynn From seaman at noao.edu Mon Oct 24 15:51:35 2005 From: seaman at noao.edu (Rob Seaman) Date: Mon, 24 Oct 2005 15:51:35 -0700 Subject: UCD versioning? Message-ID: <6E25F7C3-4800-48FE-8A21-A3CC64C5C09D@noao.edu> Hi, Could somebody comment on the relationship between http:// www.ivoa.net/internal/IVOA/IvoaUCD/UCD-1.9.9b.pdf and http:// www.ivoa.net/Documents/latest/UCD.html? Is the former obsolete? It hasn't been updated in two years. In any event, I don't see any discussion of a versioning strategy. Unfortunately, I also can't seem to get a search to work on the list archive (http://www.ivoa.net/forum/ucd). Surely this topic has been discussed (likely in gory detail)? Or is there some notion of "once UCD, always UCD"? The recent discussion about UCD maintenance seems to me to be orthogonal to the question of explicit versioning. Are new UCDs simply added to the end of a list that preserves no historical memory? Or are new UCDs appended to only the latest version of the list? Are UCDs ever removed from the list? Are their names or locations in the hierarchy never to be changed? And is the interpretation of a document that uses UCDs going to vary as the master list changes over time? The process for adopting UCDs is intended to be very conservative, but even so, one imagines that some UCDs will be adopted that will later be rethought. Note that even if the UCDs associated with a particular document (or image or catalog) don't change, that the nuances of the interpretation of these items might change as new - unselected - UCD options are added to the master list. An assertion versus three options may turn into an assertion versus six options. I gather there is some thought of occupying UCD-space only until we enter the promised land of the ontology-sphere. One might be forgiven, perhaps, for imagining that this may well take a sizable fraction of the Biblical 40 years of wandering, however. That being the case, what provisions are in place to allow UCDs to thrive in the ontology-free wilderness? Could someone summarize how UCD versioning is - or is not - supposed to work? Alternately, feel free to point me toward pertinent discussions. Thanks! Rob Seaman NOAO