[fitsbits] Bintable proposals

[Lucio Chiappetti]

On Fri, 16 Nov 2001, Don Wells wrote:

> Lucio Chiappetti writes:
>  > ..On the other hand concerning "preallocation" one snag of FITS
>  > used for acquisition.. is that one has to preallocate space for the
>  > header to store history keywords, or shift down the data part of
>  > the HDU. Probably the easiest thing would be to add all history,
>  > administration and documentation keywords to an "header only" HDU
>  > postponed after the data.
>
> This is not a problem with the multiple-file real-time disk solution
> which I described in an earlier posting:

Of course.

I realize now that I've badly worded my statements. My main concern was
not for an acquisition environment (of course one can do whatever one
likes until the files are "closed and ready").

I should have said "starting from the discussion about acquisition, I
would like to point out a general preallocation problem".  The kind of
environment I had in mind was more a pipeline one (not necessarily an
acquisition one), in which one can also do in-place modification of data
keeping all old headers and adding history information.

For me the possible solutions are two : use a non-FITS format as working
format and convert to FITS at the end for transport/archiving, or place
all administrative info in a "dataless, header-only" HDU *at the end*.

More a matter of practice in using the standard, than a comment about the
standard itself.

----------------------------------------------------------------------------
Lucio Chiappetti - IFCTR/CNR - via Bassini 15 - I-20133 Milano (Italy)
For more info : http://www.ifctr.mi.cnr.it/~lucio/personal.html
----------------------------------------------------------------------------