VOUnits 1.1 WD posted

[Norman Gray]

Greetings, all.

The WD for version 1.1 of the VOUnits specification is now available: <https://ivoa.net/documents/VOUnits/20220525/>

This version is intended to resolve a few minor errors in the specification, and make some changes, again minor, in the specification, which have emerged as useful since version 1.0.

You might be interested in two related talks at the recent Interop:

  * Sébastien Derriere and colleagues at CDS surveyed the unit strings in VizieR metadata to validate them against the VOUnits spec: <https://wiki.ivoa.net/internal/IVOA/InterOpApr2022Ops/VizieR_VOUnits.pdf>
  * Mark Taylor surveyed UCD and VOUnit usage in the VO: <https://wiki.ivoa.net/internal/IVOA/InterOpNov2021Ops/uuc.pdf>

Changes:

  * Clarified: that Besselian year is not in FITS.

  * In FITS 4.0 (FITS Working Group, 2018), the preferred symbol for the year is a, not yr (see Table 2).

  * The VOUnits syntax now accepts '1' as positively indicating a dimensionless unit string, and the text no longer recommends that the empty string `positively indicates that the corresponding quantity is dimensionless'. An application is free to interpret the empty string as indicating a dimensionless quantity, that the units are unknown, or that this is a parsing error (see Sect. 2.2, GitHub issue 9).

  * Added brief rationale for the syntax of the VOUFLOAT terminal (see Appx. C.4, GitHub issue 8).

There are also some outstanding questions:

  * The current grammars support a 'scale factor' in several places, accommodating a unit string of, eg, '10**6W' (which is merely an odd way of writing a MW).  Should this be additionally permitted inside function calls, eg in a unit of 'log{10**6W)'?  Tentative answer: probably, yes, for consistency.  But supposing that that is OK, should it be only in the VOUnits syntax or in the deduced formal grammars for FITS, OGIP and CDS as well?

  * The document mentions 'Sun' as a unit, glossed as 'relative to the Sun'.  This isn't currently one of the VOUnits 'known units' -- should it be?

  * The unit string '%' was omitted from the VOUnits 'preferred units' list.  That was probably just a mistake, so reinstating it should be uncontroversial.



The authors of this revised VOUnit version welcome your comments.

There is a collection of issues on the standard's github page: <https://github.com/ivoa-std/VOUnits/issues>.  Feel free to comment on those issues, or add to the list.

The specification is also being discussed on the Semantics list.  See

  * Semantics WG wiki page: https://wiki.ivoa.net/twiki/bin/view/IVOA/IvoaSemantics
  * WG mailing list archive: http://mail.ivoa.net/pipermail/semantics/

Markus Demleitner
Norman Gray
Sébastien Derriere


-- 
Norman Gray  :  https://nxg.me.uk
SUPA School of Physics and Astronomy, University of Glasgow, UK