DATE-OBS='31/12/99'

[Steve Allen]

In article <31D0FC80.BB8 at ast.cam.ac.uk>, Guy Rixon  <gtr at ast.cam.ac.uk> wrote:
>I strongly support Pete's suggestion of a new keyword to take over from
>DATE-OBS in new files and agree that yyyymmdd is the best format.

Take note, however, of the WCS draft paper at
http://fits.cv.nrao.edu/documents/wcs/wcs.html
which proposes to standardize the usage of MJD-OBS.  This will work
fine if the proposal also solves the backwards compatibility problem
of existing FITS headers using different timescales.

>How about DATEFULL, or DATELONG, or even, since the yyyymmdd format is thought
>to be an ISO standard, DATE-ISO?

A date format which does not also include the time of the observation
increases the risk of "midnight" coding errors where the DATE and TIME
are sampled on different "days".

DATE-ISO does not distinguish between FITS 'DATE' and 'DATE-OBS'; i.e.,
between the date of the construction of the FITS header and the date
when the data were acquired.  However, it is a much more human-readable
form than MJD.
--
Steve Allen          UCO/Lick Observatory       Santa Cruz, CA 95064
sla at ucolick.org      Voice: +1 408 459 3046     FAX: +1 408 454 9863
WWW: http://www.ucolick.org/~sla                PGP public keys:  see WWW