[fitsbits] Celsius unit in FITS keyword comments

[Norman Gray]

Lucio, hello.

On 3 Mar 2016, at 14:31, Lucio Chiappetti wrote:

> I cannot talk for IVOA, but somebody more familiar with that 
> environment should (after all there is no sense in FITS and IVOA 
> diverging, nor in FITS ruling over this matter).

Mark quoted one of the few mentions of 'Celsius' in the IVOA VOUnits 
document [1].  Re-reading it now, in the light of this discussion, I'm 
afraid to say (and I wrote it) that it sheds much less light than it 
should.

My excuse is that when we inserted this text, we were trying hard to 
avoid triggering a long online discussion about what units should and 
shouldn't be permitted in this context.  That section is entitled 'What 
this document will not do'.  I suspect that we subconsciously avoided 
even thinking about degrees celsius, on the grounds that 'celsius' is 
(effectively) an adjective rather than a unit noun, it has latinate word 
order (not 'celsius degrees') and there isn't even a standard unit 
abbreviation ('deg', 'degC', 'oC',...).  And it's not standardised other 
than grudgingly.

The VOUnits document includes a list of 'known units' which is very 
closely related to the list of units mentioned in the FITS document.  
The attitude taken towards these units is that:

> [...]a VOUnits processor must interpret the symbols of [the list of 
> known units] compatibly with the indicated units: a m is always a 
> metre of one type or another, and may not be interpreted as, for 
> example, a minute.
>
> Unrecognised units should be accepted by parsers, as long as they are 
> parsed giving preference to the syntaxes and prefixes described here. 
> Thus, for example, the string furlong/week should parse successfully 
> (though per- haps with suitably prominent warnings) as the 
> femto-‘urlong’ per week.

We also included a discussion (Sect. 2.10) of why we thought it was 
necessary to permit _any_ unknown units, as long as they were 
syntactically conformant with this specification.

It surprised us, when writing this, how many little deviations there 
were between the BIPM, ISO and IAU specifications for units.

----

I think we should all switch to either degrees Réaumur or degrees 
Celsius (not the silly modern one, but Mr Celsius's original with 0C at 
the boiling point and 100C at the freezing point of water).  Degrees 
Rankine, anyone? (I confess to having a soft spot for the Rankine scale, 
since he, like Kelvin, was a Glasgow man).

All the best,

Norman


[1] http://www.ivoa.net/documents/VOUnits/


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