[fitsbits] EXPOSURE or OBSTIME?

[Steve Allen]

On Mon 2000-06-05T15:16:28 -0700, Tom Kuiper hath writ:
> Why don't you think it is a good idea?

Because I believe

 1) that FITS header keywords should be viewed as elements of a database
 2) databases should obey the Third Normal Form

> > It should have been avoided, and would have if a data reduction
> > package in use had a means for translating keywords in the fashion
> > akin to that provided by IRAF.
>
> That strikes as cumbersome.  It means that when my program writes a FITS
> file, I have to keep in mind whether the recipient has this capability and,
> if not, what variant of FITS it understands.  I think we're stuck here with
> backward compatibility to a time when there was keyword chaos.

There is no data dictionary which completely describes the
locally-defined keywords found in most archival FITS files from the
1980s.  Although it has subsequently become significantly easier to
find them, it remains the case that FITS files may exist without their
dictionaries.  Writing two keywords which contain the same value does
not help the future data archaeologist to determine whether one can be
ignored.

In the archival sense there will always be data which originated prior
to the establishment of certain conventions, or from agencies who
never bothered to conform to them.  The only way to interpret such
data is for our current favorite data reduction package to have a
keyword translation scheme which permits the existing sub-optimal (and
probably ambiguous) archival keywords to be translated into their
closest modern approximations.

--
Steve Allen          UCO/Lick Observatory       Santa Cruz, CA 95064
sla at ucolick.org      Voice: +1 831 459 3046     http://www.ucolick.org/~sla
PGP: 1024/E46978C5   F6 78 D1 10 62 94 8F 2E    49 89 0E FE 26 B4 14 93