(r32) VODataServiceV12RFC < IVOA

IVOA Web>IvoaResReg>ResourceMetadata>VODataServiceV12RFC (revision 32) (raw view)~~EditAttach~~
---+ <span style="color: #0000ff;"><span style="text-decoration: underline;">VODataService 1.2</span></span> Proposed Recommendation: Request for Comments

%TOC{depth="2"}%

<span style="background-color: transparent;">The latest version of <span style="color: #0000ff;"><span style="text-decoration: underline;">VODataService 1.2</span></span></span><span style="background-color: transparent;"> can be found at:</span>
   * https://www.ivoa.net/documents/VODataService/20210223/index.html
A build of the current SVN (which will usually contain some minor fixes in response to RFC comments) is available during RFC from http://docs.g-vo.org/VODataService.pdf

While future versions may be migrated elsewhere, VODataService 1.2 is managed in the legacy volute repository at: http://volute.g-vo.org/viewvc/volute/trunk/projects/registry/VODataService/
---+++ Summary

VODataService 1.2 is an incremental update to the widely-used VODataService Registry extension (98% of all registry records use VODataService resource types). We expect 100% backward compatibility.

We believe there are three main points for TCG review:

   * <p>Post-STC1 (and realistically usable) Space-Time coverage in the registry; sect. 3.2</p>

   * <p>New resource types to support multi-resource services (as per Discovering Data Collections); sect. 3.1</p>

   * <p>Waveband is now governed by a vocabulary of messenger types; promoting this to REC would also make the vocabulary http://www.ivoa.net/rdf/messenger accepted (with Photon and Neutrino included as new terms).</p>
There are also a few minor changes to tableset conventions. This requires review by those familiar with VOSI Tables and TAP.
---++++ Changelog since REC-1.1

<br /> In approximate order of importance:
   * <p>Deprecated !STCResourceProfile and replaced it with spatial, temporal, and spectral children of coverage; spatial has a frame attribute not further defined so far to allow for future extension to e.g., solar system bodies.</p>
   * <p>Dropped vs:Waveband and changed waveband to being controlled by a vocabulary that initially grows a generic Photon and a Neutrino concept over what the previous Waveband had.</p>
   * <p>Introduced new !DataResource and !CatalogResource resource types and wove them into the inheritance hierarchy to !CatalogService; these are to be used for "dependent" resources.</p>
   * <p>New (optional) table/@nrows attribute gives an estimate for the size of a table</p>
   * <p>extendedType is now defined to correspond to VOTable xtypes in the absence of extendedSchema.</p>
   * <p>Required inclusion of quotes for delimited identifiers in a SQL context.</p>
   * <p>Sanctioning the use of footprint/@ivo-id to indicate the footprint standard used (this goes against the semantics of the !ServiceReference tye underlying footprint, because the reference was originally intended to go to a registered service returning footprints &ndash; but no such services ever came around, and instead people used ivo-id as specified now).</p>
   * <p>Now allowing any vs:DataType element to define vs:InputParams.</p>
   * <p>In order to still ensure schema validation of type names, now advising to have an explicit xsi:type in param's dataTypes.</p>
   * <p>Deprecated TAPType.</p>
   * <p>DataType/@arraysize no longer defaults to 1, and the interpretation of arraysize=1 as a scalar is withdrawn. Use empty arraysize for scalars now.</p>
   * <p>Added a SHOULD requirement on !CatalogResources to have a tableset.</p>
   * <p>Deprecated !DataCollection and !StandardSTC (which are no longer needed).</p>
   * <p>Adding a summary of the [[https://www.ivoa.net/documents/discovercollections/20190520/index.html][Discovering Data Collections]] EN.</p>
   * <p>BaseParam's delim attribute no longer defaults to blank. That conflicted with VOTable, where other conventions are in place (e.g., for string arrays).</p>
   * <p>Now discouraging the use of delim outside of !InputParams.</p>
   * <p>No longer requiring unique table names within a tableset; uniqueness is now required within a schema. Many services have been in violation of the old unique-within-tableset rule for a long time without operational difficulties, and this change affects little as fully qualified names are now required for the main uses of tableset.</p>
   * <p>Ported source to ivoatex.</p>
<span style="font-size: x-small;"> </span>
---+++ Reference Interoperable Implementations

On the production side, [[http://soft.g-vo.org][DaCHS]] has been producing records with most of these features for quite some time. As an example, see the Heidelberg data center's OAI endpoint at https://dc.g-vo.org/oai.xml <br /> <br /> Also on the production side, VizieR is producing VODataService 1.2-compliant records, including coverage declaration, on http://cdsweb.u-strasbg.fr/registry/beta/ , which is slated to become the primary VizieR publishing registry in Mid-2021.<br /> <br /> On the consumption side, the [[https://dc.g-vo.org/browse/rr/q][Heidelberg RegTAP]] service parses the extra information and exposes it in custom RegTAP extensions rr.stc_spatial, rr.stc_spectral, and rr.stc_temporal. The table's nrows attribute shows up in rr.res_table.

Also on the consumption side, TOPCAT version &gt;=4.8-1 consumes and displays the =nrows= quantity in TAP metadata display where applicable.

---+++ Implementation Validators

The syntax (and a bit of the semantics, too) of VODataService records can be checked with ordinary XML schema validation tools. As setting up the necessary schema locations is somewhat involved, you might want to use DaCHS's dachs val xsd command.

For VOSI tables endpoints of TAP services, the STILTS =taplint= validator version &gt;=3.4-1 will validate against the schema coming with this PR.

---++ Comments from the IVOA Community during RFC/TCG review period: 2021-03-01 to 2021-04-23

The comments from the TCG members during the RFC/TCG review should be included in the next section.

In order to add a comment to the document, please edit this page and add your comment to the list below in the format used for the example (include your Wiki Name so that authors can contact you for further information). When the author(s) of the document have considered the comment, they will provide a response after the comment.

Additional discussion about any of the comments or responses can be conducted on the WG mailing list. However, please be sure to enter your initial comments here for full consideration in any future revisions of this document

   * Sample comment by TWiki.WikiName 
      * Response (by TWiki.WikiName)

---
---

---++ Comments from TCG member during the RFC/TCG Review Period: 2021-03-01 to 2021-04-23

WG chairs or vice chairs must read the Document, provide comments if any (including on topics not directly linked to the Group matters) or indicate that they have no comment.

IG chairs or vice chairs are also encouraged to do the same, althought their inputs are not compulsory.

---+++ TCG Chair & Vice Chair
---+++ [[IvoaApplications][Applications Working Group]]

My primary concern with this standard is regarding the specification of temporal (and to some extent spatial) coverage. With the RFC period for MOC 2.0 now imminent, we will soon have a more robust IVOA standard mechanism for specifying temporal coverage (TMOCs), and indeed for combined spatial/temporal coverage (STMOCs).

It seems clear to me that the goal should be to use MOCs for both temporal and spatial coverage in VODataService. The question is how to reach that goal. Having this interim VODataService standard that uses simple time intervals does provide a path forward that is simpler than taking the extra time to integrate MOC 2.0. However, since MOC 2.0 already has solid code support, implementing simple time ranges for VODataService could end up being an unnecessary diversion from a straighter path to the end goal. However, I'm worried that if there is any take up of VODataService 1.2 time ranges that it could delay implementations with TMOCs or STMOCs, and hinder interoperability once those implementations are created.

I really don't want to delay this standard any longer, and I'm generally in favor of working towards more frequent updates to improve standards instead of taking forever to create one perfect standard. So I'm wondering, does it make sense to work out some brief but fairly specific language on how we will work towards the goal of using TMOCs and STMOCs, particularly addressing how such changes for 1.3 could be both backward compatible and interoperable?

-- IVOA.TomDonaldson - 2021-10-05

   * Hm. While as you said this is a 1.3 discussion, let me already state here why I think the plain intervals are fine for spectrum and time for at least the next few years. First, nobody expects that discovery in the Registry has no false positives. In fact, the Registry is even a win if there's 99% false positives: 1 service matching, 99 services returning nothing with 2000 services not queried is a fine gain as long as a machine does the querying (which is what the VO is mainly about). Then, given that no database so far has support for TMOCs (and SMOCs only made it into upstream pgsphere this year, where there's no implementation of them in sqlserver, the second platform for existing RegTAP services, yet), going for TMOCs in VODataService would be a large implementation hurdle. On the other hand, for most data collections, you have perhaps a few observation campaigns (which would be the intervals), but resolving those into individual days almost certainly buys you nothing for resource discovery, at least not unless you correlate that with the spatial coverage. Which to me means that unless we go all the way for STEMOCs one day (when there's solid software support for them, preferably including indexed database searches) we'd be paying a high price in implementability for almost no gain in functionality. -- IVOA.MarkusDemleitner - 2021-10-05

   * You make a good point that the 1.2 features are a useful step forward for data discovery, likely pruning the search space significantly. Indeed any implementation could be a useful testbed for informing the evolution of these features and whether the limitations are important. Since the time and spectrum intervals are relatively simple, implementations may be a useful first step towards handling more complex implementations. I will be interested to watch the implementation evolve for sqlserver-based regtap. My intuition is that handling the spatial constraints within sqlserver will not necessarily be the best way to go, and that some extension of MOC server may be an effective pre or post filter to the regtap sql queries. We'll see. On the interoperability question if we want to support STMOC, I suppose this would end up being another "support both" scenario which is somewhat disappointing. Though maybe it's not such a big (or immediate) concern since given the inevitable lag time in populating this metadata for a substantial number of resources, it may take a while to really see how these features are being used and how to evolve. Apps approves. -- IVOA.TomDonaldson - 2021-10-06

---+++ [[IvoaDAL][Data Access Layer Working Group]]

Overall we are happy with this update and it is good to see the split into spatial, spectral and temporal coverage and the use of MOCs

Issues/Concerns
   * 2. In the example, change the `xmlns:vs` value to point toward the XSD of VODataService 1.2. Currently it points toward 1.1 which does not include some node elements used in the example, such as `spatial`, `temporal`, ... 
      * No, the v1.1 is regrettably correct. The minor version should never have made it into the namespace URIs, but for now we must keep them in there or we break everything and need a new major version (cf. the schema versioning EN and the explanation in sect. 2.1). -- IVOA.MarkusDemleitner - 2021-09-14
   * 3.1 From a data provider perspective, after reading 3.1 I'm somewhat confused as to how I should describe the CASDA services. As we have a table of fits images and catalogues stored as tables and accessible via TAP and SIA2/SODA I am guessing these should be vs:CatalogResource ? 
      * No, it would likely be CatalogService; I've added a back reference to 2.2.2 in the opening paragraph of the section. Does that help? -- IVOA.MarkusDemleitner - 2021-09-14
   * 3.2 https://www.ivoa.net/rdf/messenger/2020-08-26/messenger.html seems to be missing any gravitational wave messengers - should these be included now? 
      * I don't think so. Our vocabularies are designed to it's easy to add things as they're needed, and so far nobody has asked for Gravitation. And, for starters, I don't want to make the call between graviton (would be a particle in parallel to the others, but who knows it it exists) or gravitational-wave (is what we're actually detecting, but it doesn't quite fit with the others, and at least will need some convention for spectral limits). Things like these tend to be a lot less arbitrary when you have the domain experts on board. -- IVOA.MarkusDemleitner - 2021-09-14
   * 3.2 spectral - joules are required for spectral limits but there are no joules mentioned in the messenger descriptions- not required but it would be useful to provide a translation 
      * I'll put them in if there's more requests for that, but frankly I think the translation from whatever representation people prefer in the respective bands to particle energy is going to be performed "during serialisation". I do not expect humands in any spectral band to actually handle Joule ever (though it would be cool if we convered to aJ in the optical). The messenger descriptions are aimed at humans, not machines, and hence they use the conventional units per band. But as I said: poke me again and I'll put in the particle energies. -- IVOA.MarkusDemleitner - 2021-09-14
   * 3.2 stc:STCResourceProfile: "This element is imported from another schema" can you refer to the schema? 
      * Regrettably, not easily. You see, this text is produced through an XSLT (ivoatex/schemadoc.xslt, around line 70). To get to the schema location, I'd have to figure out the prefix mapping (and, for full generality, perhaps even the schema locations). I doubt I can do that from within xslt1 without a lot of apparatus. But I've added some prose above explitily mentioning the schema in Volute rev. 6030 -- IVOA.MarkusDemleitner - 2021-09-14
   * 3.2 vs:SpatialCoverage Attributes: is the "an external standard" in the comment still needed? 
      * Nah, that was a typo. Thanks! -- IVOA.MarkusDemleitner - 2021-09-14
   * 3.2 Where are vs:FloatInterval and vs:ServiceReference defined? 
      * Good point; they've been missing in the text so far. I've added them in volute rev. 6031 -- IVOA.MarkusDemleitner - 2021-09-14 
         * The description of frame seems a bit out of place now - would it be better above this new block for the types? -- IVOA.JamesDempsey - 2021-09-15 
            * I agree it would, and hence I've done the move in volute rev. 6036. -- IVOA.MarkusDemleitner - 2021-09-20
   * 3.3 vs:TableSet, vs:TableSchema, vs:Table, vs:ForeignKey (etc) Type Schema Definitions - recommend either all or none of the elements (e.g. `title`, `description`, &hellip;) specify the default of maxOccurs="1" 
      * Fair enough. In in Volute rev. 6032. I've left some minOccurs="1", as I think they're a useful cue that something is indeed mandatory. You *could* say that cue should then be verywhere. And I'd have to oblige. But I'll try first to get away with what we have now. -- IVOA.MarkusDemleitner - 2021-09-14
   * 3.3.3 What is &ldquo;an XML type legally extended from the types associated with these elements&rdquo;? - an example might help here 
      * Frankly, this is not my text, and I can't really say if the "legally" was intended to carry actual meaning. And very frankly: before conjuring up an example for something that nobody in the past 10 years bothered to do (all extensions, used trick (2)), I'd rather drop all of 3.3.3 (which is somewhat misplaced in the spec anyway) -- IVOA.MarkusDemleitner - 2021-09-14
   * 3.3.3 Is the namespace changing from http://www.ivoa.net/xml/VODataService/v1.1 ? 
      * No (see above) -- IVOA.MarkusDemleitner - 2021-09-14
   * 3.5 - Std defaults to true in inputParam but not in tableParam - is that intended? 
      * Interesting. Again, this is inherited material, and so I can't *really* be sure. The std="false" in tableParam certainly makes a lot of sense as most tables do not follow a standard. Having std true in inputParam by default is, ummm, daring. Now that you told me, I suppose I'll have to do something about this in DaCHS. Be that as it may, while I don't think anyone does anything with inputParam/@std, I'd be very cautious touching this -- there are more than 25000 VODataService records out there, and even just changing a default for a forgotten attribute may blow a few of them up. I'd need a stronger reason for risking that than just internal symmetry. -- IVOA.MarkusDemleitner - 2021-09-14
   * 3.5.2 vs:TableParam Extension Metadata Elements : &ldquo;See the specification document for definitions of recognized keywords.&rdquo; =&gt; which specification? 
      * Good catch; these definitions used to be in the schema in ways that did not port to TeX, and I forgot to put them in the text while porting the document. I've put them back into the schema in a somewhat terse way in Volute rev. 6032 -- IVOA.MarkusDemleitner - 2021-09-14
   * Appendix A - adding sections affected by each change would help review and use by implementers 
      * Did that for the edits since REC-1.1 in volute rev. 6033.
Edits:

   * 3.1.2 typo - &ldquo;instead of vs:DataResource when the resource&rsquo;s capabilties&rdquo; =&gt; &ldquo;capabilities&rdquo;
   * 3.1.3 suggest a comma after coverage in "has metadata on coverage and the facilities"
   * 3.1.3 remove &ldquo;of&rdquo; in "SHOULD declare of the metadata"
   * 3.5 - extendedType - add &ldquo;the&rdquo; to &ldquo;attempt to handle value&rdquo;
-- IVOA.JamesDempsey - 2021-09-10

A build with all the fixes in is at http://docs.g-vo.org/VODataService.pdf. Thanks! -- IVOA.MarkusDemleitner - 2021-09-14

Thanks for the fixes and addressing the concerns above. I agree leaving inclusion of gravitational wave support until we have a domain expert involved makes sense. -- IVOA.JamesDempsey - 2021-09-15
---+++ [[IvoaDataModel][Data Model Working Group]]

Being not a great user of such a model, my comments are limited to point out one typo:
   * section 2.2.2 page 13: That a data collection<span style="color: orangered;" class="WYSIWYG_COLOR"> *ist* </span> published through
Approved

-- IVOA.LaurentMichel - 2021-09-15

Fixed in volute rev. 5949. Thanks! -- IVOA.MarkusDemleitner - 2021-09-20

---+++ [[IvoaGridAndWebServices][Grid & Web Services Working Group]]

GWS is not really involved in VODataService types but I agree on the standard and I approve it. The document is well written but there are still some typos, e.g. capabilties insted of capabilities, collection *ist* as mentioned by DM WG.
   * Aw, yet another one. Fixed, thanks. -- IVOA.MarkusDemleitner - 2021-10-06
Approved

-- IVOA.GiulianoTaffoni - 2021-10-05
---+++ [[IvoaResReg][Registry Working Group]]

Approved with the hope of adding STMOC in 1.3 given sufficient takeup.<br />I am specifically not digging into the VOSI and TAPTypes bit and will take the lead from DAL there.

-- IVOA.TheresaDower - 2021-08-20
---+++ [[IvoaSemantics][Semantics Working Group]]

This document refers to the messanger vocabulary (http://www.ivoa.net/rdf/messenger), which is still not approuved by the IVOA. Fix this dictionary before the next step toward the REC procedure.

   * Which is to mean: review this vocabulary as part of this RFC, as accepting VODataService 1.2 will also make messenger an IVOA vocabulary. -- IVOA.MarkusDemleitner - 2021-05-06

---+++ [[IvoaDCP][Data Curation & Preservation Interest Group]]

---+++ [[IvoaEducation][Education Interest Group]]

---+++ [[IvoaKDD][Knowledge Discovery Interest Group]]

---+++ [[IvoaTheory][Education Interest Group]]

---+++ [[IvoaRadio][Radio Astronomy Interest Group]]

---+++ [[IvoaSS][Solar System Interest Group]]

I have a few minor questions that might simply be typos, or me not having the schema details at my fingertips:

   * Page 21, the "Type" patterns for elements _temporal_ and <em>spectral </em>allow negative values. Is that intentional? 
      * No. There are just generic float patterns. I'm fairly neutral on whether to specialise them to "non-negative floating point", but I'd need some extra encouragement to do that since I don't think there's a large benefit to catching this kind of problem in the schema, where there's always the possibility one day someone has reason to have something negative here (fun fact: about 15% of the parallaxes in Gaia eDR3 are negative...) -- IVOA.MarkusDemleitner - 2021-09-20
   * Page 22, Element _regionOfRegard_, _Type_: This is allowed to be negative. Is that appropriate, given the definition? 
      * Probably not. But then just saying it's a float lets us use an XSD simple type, whereas there is no XSD simple type positiveFloat or something like this. Be that as it may: this is stuff inherited from VODataService 1.1, and I don't think any client uses it; there are also only eight resources declaring it (all mine, btw). So, if I were to touch this at all, I'd deprecate it as not useful at this point. -- IVOA.MarkusDemleitner - 2021-09-20
   * Page 35, arraysize, _Type_: The _arraysize_ string, according to this pattern, can contain the number '0' and nothing else, in addition to being empty. The interpretation of '0' is not stated, as far as I can see. * Interesting point. Again, this is legacy from VODataService 1.1, and one may rightfully ask why the authors back then felt the need to constrain arraysize more than VOTable itself (which in the schema just says arraysize is xs:string). I'd again say that the few errors we might catch by being more restrictive with the patterns probably don't justify the extra complexity in the pattern &ndash; and the curses we'd get from people that actually have zero-length arrays for some reason or other. -- IVOA.MarkusDemleitner - 2021-09-20
   * Page 41, A.3, last bullet: Regarding _@arraysize_, this bullet says "Use empty arraysize for scalars now." That should be stated in the body of the document, not in an appendix. As far I could tell, it is not. * It's in the schema's annotation for the arraysize attribute, taken over into the generated documentation on (current) p. 44 (&ldquo;Leave arraysize empty for scalar values&rdquo;). -- IVOA.MarkusDemleitner - 2021-09-20
   * Page 42, first bullet on page: This item states "BaseParam's delim attribute no longer defaults to blank." From what I can see in the code supplied in this document, _BaseParam_ does not have a _@delim_ attribute. I did not analyse the XSD or compare it to any of the code samples in this document. Does that need to be done systematically? 
      * Oops, yes, that's an attribute of DataType, which then is a child of all kinds of xParams. Fixed that in the changelog entries. As to auto-validating samples: the ipac-resource.xml that's shown at the beginning I indeed XSD-validate occasionally, and the schema file get exercised as part of DaCHS. Going beyond that in this document is probably hard, as I suppose it would involve processing of natural language. Do you have any specific changes in mind? -- IVOA.MarkusDemleitner - 2021-09-20
I also note a few typos that I don't think have been mentioned elsewhere:

   * Page 14, paragraph that begins "With the advent of services...": The comma should be deleted from the "&ndash;," pair. 
      * Um, I don't think so &ndash; I don't have a proper reference to a style guide now, but I'm fairly certain the presence of a parenthetical construction doesn't change the need for a comma. Or am I confusing German and English here? -- IVOA.MarkusDemleitner - 2021-09-20
      * That punctuation looks weird to me too. However, I'd say it's OK to leave this stylistic stuff to author preference, since there's little danger of ambiguity, and there are clearly different norms of how to do these things, even between English and US English, so effort arguing about it or making changes is probably not well spent. -- IVOA.MarkTaylor - 2021-09-27
   * Page 14, paragraph that begins "The solution eventually adopted...": "for these problems are auxiliary" should be "for these problems is auxiliary" (the subject is "solution")
   * Page 22, Element _waveband_, _Comment_: "...when it is now also coverage non-electromagnetic messengers." I suggest this should be "...when it now also covers non-electromagnetic messengers." or something similar.
   * Page 31, paragraph that begins "Normally, a VOResource...": "with _capability_ element" should be "with a _capability_ element"; "to specific URI" should be "to a specific URI".
   * Page 31, paragraph that beings "Listing required parameters...": Either the comma following "allowed" should be deleted or a comma should be added after "interface".
Irrespective of the answers to my questions and assuming minor editing as indicated here and elsewhere - Accepted. -- IVOA.AnneRaugh - 2021-09-17

Fixed everything I've not commented to the contrary in volute rev. 6037 (also visibile on the pre-built PDF). Thanks a lot! -- IVOA.MarkusDemleitner - 2021-09-20

---+++ [[IvoaTheory][Theory Interest Group]]

---+++ [[IvoaVOEvent][Time Domain Interest Group]]

---+++ [[IvoaOps][Operations]]

Changes from this review in Volute rev. 5949 -- IVOA.MarkusDemleitner - 2021-05-06

This mostly looks good, but a few comments:

   * Sec 3.3: The =nrows= attribute gives the approximate number of rows. I'm quite happy that that's a reasonable thing to supply, but I wonder if the name "nrows" might be confusing, and conceivably clash with a future requirement for recording the exact number of rows. I would suggest at least that the nrows _Meaning_ entry should read _"The approximate size of the table in rows"_ rather than _"The size of the table in rows"_. And maybe rename the attribute =nrows_approx= or similar? Not sure. 
      * I've added the "approximate" to the meaning. On indicating the nature of nrows in the attribute name: I think that's overdoing things. You see, in a transactional system, "number of rows" is a tricky concept (which, by the way, is the reason why a "select count(*)" generally is slow). True, most of our tables are static, and then it won't matter. But given the complication in general I consider it highly unlikely that we'll have nrows-exact one of these days (and what would it be good for over nrows-approx?) -- IVOA.MarkusDemleitner - 2021-05-06 
         * It's user-facing information, and users without a transactional theory background might think that "<code>nrows</code>" means, well, number of rows. Why care? If I'm looking at a data release I'm very familiar with (e.g. Gaia EDR3) and find a declared number of rows different from the public value, I might suspect problems with the ingestion. On the other hand, if its 1470000000 I can probably guess it's a guess. I think it's OK as now with "approximate" in the documentation, but I wouldn't be surprised to get the odd confused user reaction in the future. -- IVOA.MarkTaylor - 2021-05-07
   * Sec 3.1.3: The _Comment_ for element =tableset= still says _"Each schema name and each table name must be unique within this tableset."_ My understanding from the change log and (I think?) the !CatalogResource type definition is that this is no longer exactly the case; table names must be unique only within their host schema. 
      * Right. Changed the Comment to "Each schema name must be unique within a tableset." -- IVOA.MarkusDemleitner - 2021-05-06 
         * That's not quite the text I was referring to: see line 680 of the XSD ( =CatalogResource= definition) not line 126 ( =DataCollection= definition). -- IVOA.MarkTaylor - 2021-05-07 
            * Ah, dang. That text clearly shouldn't have been in the attributes but rather in the TableSet type. I've fixed it at the other location, too. Thanks for not being lulled by my assurances. -- IVOA.MarkusDemleitner - 2021-05-10
   * Sec 3.2: _"By default, the MOCs are to be interpreted in the ICRS. Future extensions to non-celestial frames (e.g., on planet surfaces) will use the frame attribute."_ This is reasonable, but MOC currently says things like _"the SMOC *must* be based on the HEALPix tessellation of the sphere, expressed in the ICRS coordinate reference system"_, which apparently excludes non-celestial use of MOCs. I have [[http://mail.ivoa.net/pipermail/apps/2021-March/001510.html][lobbied the MOC authors]] to relax this language a bit. 
      * So will others, including me. Let's get http://www.ivoa.net/rdf/refframe cleaned up to make it easier for them! -- IVOA.MarkusDemleitner - 2021-05-06
   * Also: what about STMOC? Does the approach to !STC coverage need to be reviewed in the light of the introduction of space-time MOCs in MOC 2.0? 
      * Given that there is no database support for STMOCs in sight, this would delay this quite a bit, and given that time coverage has only be sparingly declared so far, it would seem that correlated space/time coverage would be of low priority. Taken together: If we want STMOC, let's do it in VODataService 1.3 &ndash; and until then, let's get time coverage into the Registry at all. -- IVOA.MarkusDemleitner - 2021-05-06
   * Sec 3.5.3: The =delim= attribute is defined as _'&lt;xs:attribute name="delim" type="xs:string" default=" "/&gt;'_. Should that be _'fixed=" "'_ instead? 
      * No, I don't think so. For one, this might break records (I don't think it will, as delim, as far as I can make out, is one of those "let's add it just in case" features that then nobody uses and are just trouble). And then, outside of VOTable, delim might one day come in handy after all ("just in case"). -- IVOA.MarkusDemleitner - 2021-05-06
   * Sec 2.2.2: _"ist published"_ -&gt; _"is published"_ * Entschuldigung -- IVOA.MarkusDemleitner - 2021-05-06
-- IVOA.MarkTaylor - 2021-03-12

Thanks for responses, I'm happy with it now. Recommend Accept. -- IVOA.MarkTaylor - 2021-05-10

Late addition for some UCD-related things I've just noticed:
   * The example in Section 2 includes two incorrect/made-up UCDs; =meta.id;name= and =src.spect.dopplerVeloc=.  They should be replaced by legal UCDs (or dropped).
   * Documentation of the =ucd= attribute in section 3.5 says _"There are no requirements for compliance with any particular UCD standard. The format of the UCD can be used to distinguish between UCD1, UCD1+, and SIA-UCD."_  However it is not true that the format of the UCD can be used to distinguish between UCD1 and UCD1+.  In fact =PHYS=, =POS=, =SPECT=, =STAT= and =TIME= are legal as both UCD1 and UCD1+ terms, since UCD 1.10 Section 2 says: _"The UCD syntax is case-insensitive -- all uppercase characters should be converted to lowercase before parsing."_  So the sentence beginning _"The format of the UCD can be used to distinguish..."_ should be removed.  If you can think of anything to replace it that would be helpful to people trying to distinguish between the two, so much the better...
-- IVOA.MarkTaylor - 2021-10-11

   * Oh, legacy!  I've put in modern UCDs in the sample record, and I've dropped the over-bold claim in the schema file.  Thanks!  -- IVOA.MarkusDemleitner - 2021-10-12
      * Good thanks! -- IVOA.MarkTaylor - 2021-10-12

---+++ [[IvoaStdsDocsProc][Standards and Processes Committee]]

---++ TCG Vote : 2021-03-01 to 2021-04-23

If you have minor comments (typos) on the last version of the document please indicate it in the Comments column of the table and post them in the TCG comments section above with the date.

| Group | Yes | No | Abstain | Comments |
| TCG | | | | |
| Apps | * | | | |
| DAL | * | | | |
| DM | * | | | |
| GWS | * | | | |
| Registry | * | | | |
| Semantics | | | | |
| DCP | | | | |
| KDIG | | | | |
| SSIG | | | | |
| Theory | | | | |
| TD | | | | |
| Ops | * | | | |
| !StdProc | | | | |
Topic revision: r32 - 2021-10-12 - MarkTaylor
IVOA
Log in or Register
IVOA.net
Wiki Home
WebChanges
WebTopicList
WebStatistics
Twiki Meta & Help
IVOA
Know
Main
Sandbox
TWiki
TWiki intro
TWiki tutorial
User registration
Notify me
Working Groups
Interest Groups
Committees
Stds&Procs
www.ivoa.net
Documents
Events
Members
XML Schema