Updated VOTable 1.5 Working Draft

[Markus Demleitner]

Hi François,

On Tue, Nov 07, 2023 at 06:02:38PM +0100, Francois Ochsenbein wrote:
> * section 3.5 (on TIMESYS): shouldn't the list of allowed timescale
>   values be listed, as are the system COOSYS values? The same remark

In principle, I'm not overly wild about in-document vocabularies;
vocabularies are designed to grow (and to have labels and
descriptions, without which they're not terribly useful).  But you're
right, if we do it in one place, we should do it in the other, too.

I'll leave it to the editor to decide whether or not to put it

  At the time or writing, this vocabulary includes the terms
  % GENERATED: !vocterms timescale
  % /GENERATED

near the current line 809 and running make generate.  For
TIMESYS/@refposition, I'd say there current cross reference to COOSYS
should be enough (or perhaps add "see there for a snapshot of the
defined values as of the release of this document.

>   applies on the list of refposition, listed for COOSYS, but not for
>   TIMESYS (both share the same list). It looks also strange to me that
>   the system attribute in COOSYS is optional, while the timescale
>   attribute in TIMESYS is required — shouldn't this be harmonized ?

Yes, it should.  The question is how to do that.  Making
COOSYS/@system mandatory will make some VOTables invalid, and we're
officially not supposed to do that in a minor version.

I suppose it would be a reasonable and defendable policy if we said
now that we reserve the right not make it mandatory two minor
versions down the road or so.

> And last, for the comfort in reading the document, some additional
> space between the paragraphs (the \parskip value) would be nice.

Since VOTable sets

  \parindent=0pt

(which, frankly, I'd probably drop if I were the editor) you are
right that discerning paragraphs is not always simple.  Hence, having
a bit of parskip is indeed wise, and I'd totally go for

  \parskip=0.5ex

-- that's also a lot less invasive than re-introducing the parskip
(which will probably result in overfull hboxes).

             -- Markus