MJD - not acceptable according to IAU (?)

[Peter Bunclark]

Barry M. Schlesinger wrote:
> > 
> I've seen it elsewhere.  The misuse of the term "Julian Day" to
> represent day-of-year has not been at all uncommon in the area of
> tracking satellites and processing the data.

Did these guys write the Ariane-5 navigation software?

No matter how much you wish for it, some old FITS readers cannot be updated -
don't have source, don't have library, can't remember how to use a line editor...
Software which *is* maintainable is not a problem.
	The ISO-DATE idea seems great to me, perhaps UTC-DATE being a bit more
definite.  I still think it's best to drop DATE-OBS;  then old software can't 
possibly misinterpret it, but old software is much more likely to handle its
absence gracefully.
	As for MJD-OBS, it serves a different purpose;  you don't read it yourself, you let
your period-finding software loose on it.  And as has been pointed out,
the definition is watertight,  if you can't understand the 0.5 day thing then 
you're not clever enough to do astrophysics.

I want to do this:
1) leave MJD-OBS alone, but recommend it is UT of start of observation; comments
         to clarify;
2) Remove DATE-OBS from FITS writers, certainly after 2000.
3) Adopt UTC-DATE in ISO format, allowing full time-resolution - ie new software
    has to be prepared to parse any allowable case.  A reader encountering 
    DATE-OBS and UTC-DATE takes the newer one as true, or, does a consistency
    check between the two.
	UTC-DATE will refer to the start of observation unless comments or other
     descriptors indicate otherwise.

Pete.