[fitsbits] SIMPLE? (was Re: Bintable proposals)

[Rob Seaman]

Mark Calabretta writes:

> Some people suggested setting SIMPLE = F.  However, when you think about
> it, really the only thing that software can do with SIMPLE = F is either
> ignore it or reject the file outright.  While I'm sure this is quite
> appropriate for RPFITS (which contradicts the standard), using it in the
> other case is tantamount to saying that that usage will never be sanctioned.
> That is, if NAXIS2 = -1 ever did become recognised usage, future readers
> might still reject these old files on the basis of SIMPLE = F.

As your internal format, RPFITS is not the business of the FITS community
or this newsgroup.

The same is not precisely true of SIMPLE = F data.  Note that the keyword
is not "FITS = F".  I'd also have to say that the FITS standard has found
a whole heck of a lot of complexity to squeeze into overtly SIMPLE data :-)

It seems to me that there are a small number of rules implicitly embodied
in the SIMPLE = T declaration.  Number one would be the division of the
objects into distinct header-data units.  Number two is the 2880 blocking.
Number three is likely precisely the point that each header contains
sufficient information to describe the size of the associated data records.
This isn't simply a question of SIMPLE = T - it's also a question of the
GCOUNT/PCOUNT mechanism underlying FITS extensions.

There may be a few other such meta-rules.  Sounds like an interesting
issue for the community to clear up.

But whatever the rules, it isn't obvious that SIMPLE = F data is as
unmanageable as Mark suggests.  It seems to me the most obvious use for
SIMPLE = F is precisely the question of data acquisition.  Certainly there
must be many data acquisition systems that construct incomplete FITS objects
as temporary byproducts of the particular instrument and telescope in
question.  Incomplete FITS is non-conforming FITS.  Wouldn't it be
appropriate to label these temporary files accurately?

And surely to the extent that a SIMPLE = F file is itself an internal
format for some organization, the question of whether general purpose
FITS readers handle the format correctly is moot.  Nothing stops an
observatory from fielding their own readers to support their own
internal formats.  In fact - SIMPLE = F is really likely a label that
the data objects are conforming to FITS in virtually all ways.  This is
a way of asserting partial agreement to the standard.  That is quite
a lot of information for a piece of software to receive.

Rob Seaman
NOAO/IRAF