Peter and Tim,
On Wednesday, March 17, 2004, at 04:23 PM, Peter W. Draper wrote:
>>
>> So HDS should install a cut down version of PRM_PAR itself for bad
>> values
>> (hds_par?) and PRM_PAR should then include that for bad value
>> definitions?
>
> Probably, by parsing the output from the hds_datestamp file (which
> PRIMDAT
> could probably do for now). The only question I have is whether
> unformatting these decimal strings to the right bit pattern is as
> reliable
> as the hexadecimal scheme we have now. Must have done that for some
> good
> reason.
Or a partner to the hds_machine.f program could call the appropriate
routines and write the include files directly, which is probably more
reliable than parsing the datestamp file.
If I recall correctly, printf is required to format floats in such a
way that they can be reread without losing precision, but we probably
shouldn't depend on that. However, the writer of include files could
either write the bad values out as integer-equivalents as well as
formatted as floats, or else we could skip the intermediary and have
other components read the HDS-generated include file directly.
It could also write out the Fortran include file using the hex notation
if it determined that the Fortran compiler supported that.
Norman
--
----------------------------------------------------------------------
Norman Gray http://www.astro.gla.ac.uk/users/norman/
Physics and Astronomy, University of Glasgow [log in to unmask]
|