Proposal for a section talking about MIVOT

[lmichel]

Section 3.6 RESOURCE Element of the actual document presents MIVOT as a possible use of type=meta RESOURCE.
However, the situation has evolved since MIVOT is a VO standard and MANGO is close to being set as another one.

I suggest adding a section focused on data annotations.
As this section would not be normative, it could be located in an appendix or elsewhere.
I believe it is the editor's responsibility to answer this question.

This MIVOT/MANGO section should explain what MIVOT is and what it is used for. It should provide a brief overview of MANGO and mention implementations.
I can write it and submit a pull request if the authors agree with this proposal.

[mbtaylor]

@lmichel, On the whole I'm wary of introducing linkage between standards where it's not essential, since this kind of dependency tends to complicate the evolution of the standards involved. This problem gets worse the more detailed such dependencies are.

Especially for VOTable, which is implicated in so many other IVOA standards, making reference to all the other relevant documents would create a lot of unhelpful noise. Generally the direction of reference is one-way - for instance VOTable relies on UCD, VOUnits, DALI (for xtypes) so it provides some explanation of them, but does not talk in any detail about TAP, ObsCore, DataLink which rely on VOTable. MIVOT relies on VOTable and not the other way round, though since MIVOT annotations are something that you might find in any VOTable perhaps MIVOT qualifies as a special case here (the same could be said about DataLink resource descriptors).

So if this is done it should restrict itself to an overview of what role MIVOT (and maybe MANGO, though referencing a standard in development is another thing I'm not too keen on) can play in a VOTable document without getting into the details of the MIVOT specification. It also should not repeat content from MIVOT itself. In particular I don't think an example MIVOTised VOTable would be a good idea here, for details and examples it should be sufficient to point at the MIVOT document. I also discourage the idea to "mention implementations", since such references inevitably go out of date; the VOTable document doesn't even mention VOTable implementations.

As you say, there is already a short discussion of MIVOT in section 3.6, which I'd say does this job of signposting the MIVOT standard. Do you think this is out of date, out of place, or needs expanding?

Maybe the best thing would be to move the discussion from Section 3.6 into section 1.4 "VOTable in the VO Architecture", which could contain an updated discussion of MIVOT's (any maybe MANGO's) use in VOTable documents. Section 1.4 could use an update anyway.

What do you think? Opinions from other VO members are also welcome.

[lmichel]

I totally agree with the requirement of not coupling standards as much as possible

I'll open a PR:

• move MIVOT mention from 3.6 (RESOURCE) to 1.4 as suggested by @mbtaylor
• update section 3.4 ( COOSYS) (STC mention)

[lmichel]

see #88

[mbtaylor]

Closed by #88. Thanks @lmichel.