New messenger proposed vocabulary, waveband in VODataService

[Markus Demleitner]

Dear Registry, dear Semantics,

A while ago I asked around if anyone had large objections to abusing
VODataService's waveband to indicate non-electromagnetic messengers
(http://mail.ivoa.net/pipermail/registry/2019-July/005388.html,
http://mail.ivoa.net/pipermail/registry/2020-August/005436.html; for
the bigger picture, see the STC discovery roadmap,
http://ivoa.net/documents/Notes/Regstc/).

Well, since nobody protested, I went ahead and did it; it's still ok
to protest loudly (e.g., at when Semantics Calls[1]).

So:

(a) There's now a draft vocabulary of messengers:

    http://www.ivoa.net/rdf/messenger

Over the terms that already were in VODataService 1.1, there are two
new ones: Photon (gratifyingly a parent of the existing terms), and
Neutrino, because we already have quite a few services that could use
it.  If you have resources with other messengers, please let me know.

In addition, I've made EUV a narrower term than UV.

What's not so terribly nice is that obviously we can't break what's
in existing waveband elements and hence the machine-readable terms
are capitalised, in violation of the recommendations on the term form
from Vocabularies 2.  I don't think we get around that if we go for
the lazy route of re-using waveband (rather than creating a new
VODataService element, messenger, say).

(b) I've dropped the vs:Waveband type (that has contained this
wordlist before) from the VODataService 1.2 draft schema and put
a few words on waveband's meaning into the draft spec (volute rev
5867).  A PDF build is available at
https://docs.g-vo.org/VODataService.pdf.  The new language ended up
on p. 19 (in the first few lines of sect. 3.2) and p. 22 (in the
Element waveband generated documentation).



My personal view: Even after doing it, I still think it's a good idea
to move messengers to a tree-like structure from the flat word list;
for instance, we could introduce soft x-ray and hard x-ray easily,
and machines could work out these resources should be included when
people look for x-rays.  Or we could include Millimeter into Radio
(should we?).  We could even have "SDSS i band" without blowing up
everything, though I'd say we'd need a mighty strong use case to even
think in that direction.

In RegTAP, all that is not so terribly useful yet, because waveband
is one of our hash lists, where it's hard to get to using the
semantic UDFs I have in mind.  But still.  It feels right to me.

Whether it's right to stuff this into waveband I'm less sure about.
But I dislike the thought of introducing an extra element that people
won't take up forever even more.

So... opinions?

         -- Markus

[1] https://wiki.ivoa.net/twiki/bin/view/IVOA/SemanticsCalls-1