On Nov 26, 2007, at 7:00 AM, Peter W. Draper wrote:
> On Fri, 23 Nov 2007, Tim Jenness wrote:
>
>> I've just noticed that hdstrace doesn't trace ".dst" files (I
>> noticed this when looking at some 1992 CGS4 data). Renaming it to
>> be .sdf allows the trace to work fine (and hdsdump works because
>> that simply opens the file).
>>
>> Can anyone remember this ever working? Did I break it with my
>> patches to handle the explicit ".sdf" in filenames?
>
> Tried tracing a DST file using the last Starlink release (2005) and
> that also fails, so not a new issue. Couldn't swear it never worked.
>
Ok. I'll think about it.
>> and CONVERT dst2ndf gets into an infinite loop with 100% CPU.
>
> The problem is that DTA contains some hardcoded HDS error values (in
> HDSCODES), and these numbers have changed during autoconfing. I've
> added a patch and this seems to be working now (need to completely
> rebuild FIGARO and CONVERT since DTA is a static convenience library).
>
>
Thanks. That seems to fix it.
--
Tim Jenness
Joint Astronomy Centre
|