VODataService waveband: multi-messenger vocabulary?

[Markus Demleitner]

Dear Registry folks,

A year ago, I mused about ways to make the VO a bit more open to
non-EM messengers:

On Mon, Jul 29, 2019 at 11:11:18AM +0200, Markus Demleitner wrote:
> When I thought about what the Semantics chair might have to say to
> the authors of the VODataService 1.2 PR, it occured to me that, as we
> did with VOResource 1.1, we might want to take some word lists out of
> the schema and into a vocabulary.
> 
> The prime candidate I'd see there is waveband.  This is currently an
> exhaustive list of the electromagnetic spectrum:
> 
>   "Radio" | "Millimeter" | "Infrared" | "Optical" | "UV" | "EUV" |
>   "X-ray" | "Gamma-ray"
> 
> -- the documentation gives more or less precise limits for where they
> are.
> 
> As this is exhaustive, there'd be no real reason to put it into a
> vocabulary, were it not for non-electromagnetic messengers; we
> already have several neutrino services, gravitation and possibly
> diverse charged particles (in particular in the solar system) can
> reasonably expected to come.  And in the solar system, even dust
> particles might count as messengers one day.
> 
> VODataService 1.2's coverage could cope with that on the spectral
> axis, now that we're writing this in energy.  But there's no
> information on the messenger particle(s) yet.  You can ask "give me
> services with messengers between 1 GeV and 1 TeV" but not "give mie
> services with neutrinos between 1 GeV and 1 TeV".
> 
> We *could* add a messenger attribute on spectral (defaulting to
> photon), which probably would be the tidiest thing to do.  That would
> be for VODataService 1.3, though, unless someone had a very urgent
> use case.
> 
> A much lighter alternative would be to abuse waveband.  I give you
> that <waveband>neutrino</waveband>, <waveband>gravitation</waveband>,
> or even <waveband>He</waveband> would look somewhat odd, but other
> than that I think the reasonable uses cases (see above) would be
> covered with very little extra effort (like a new attribute).

One year later, I'm still waiting for a bit more takeup of
VODataService before starting RFC (Registry operators: hint, hint!),
and I've grown fond of the idea of having a messenger vocabulary.

So, I think I'd like to go for the lazy, waveband-abusing solution.
And so here's what I'll do if you don't stop me: 

(1) In VODataService 1.2's schema, I'll make waveband a token with
the proviso that terms must be taken from
http://www.ivoa.net/rdf/messenger

(2) Create a draft vocabulary there containing the VODataService 1.1
terms above, plus "Neutrino", because we already have several
services that publish neutrino data.

So... does anyone feel really bad about <waveband>Neutrino</waveband>
(of course, UIs don't have to present that as "waveband", although it
will probably still be resource.waveband in RegTAP)?

       -- Markus