[fitsbits] Associating ancillary data with primary HDU

[Thierry.Forveille at ujf-grenoble.fr]

Quoting Rob Seaman <seaman at noao.edu>:
> There is nothing magic about IEEE floating point, and one could  
> argue that the quantum nature of reality makes integer  
> representations more "intrinsic" than mantissa+exponent :-)
>
Well, concerning time, you'll need significantly more than 16
bits to count practically useful values in units of Planck's
time ;-)

>  The nature of the exposure time map will vary, but one imagines  
> most instruments will be constructed such that exposure is  
> relatively flat, varying by just a factor of a few over the focal  
> plane (or focal plane equivalent), perhaps with isolated areas of  
> particularly low or high exposure (similar to FPN)?
I don't know what William's setup is, but one common situation
with non-destructively readable integrating detectors is that
they are read out at pre-defined intervals, and the value is
stored when it looks like the next one would be saturated
(Fowler sampling is much preferable if adequate processing
power is available, of course). The exposure time can then
vary by an order of magnitude or somewhat more, but takes one of
a usually small (i.e. <255) number of possible values.