On Fri, 30 Apr 2004, Peter W. Draper wrote:
> On Fri, 30 Apr 2004, David Berry wrote:
>
> > > Both of these files would display fine until around
> > > Friday of last week. Both will load into GAIA, but if you look at the
> > > astrometry warnings, you get the message
> > >
> > > >Reading your image produced the following messages:
> > > >Failed, look for error messages on the terminal
> > >
> > > ..but no error is generated at the terminal.
>
> David,
>
> I don't think you're off the hook yet, I've looked at this issue and do
> see this problem in all recent versions of GAIA that I have, i.e. I read
> acfits.sdf into GAIA and it fails badly during an astRead, badly enough
> that I cannot tell anything about what's going wrong (there are also some
> indications that this is bad enough to effect the working of the debugger,
> so I'm worried there's a buffer overrun somewhere). I'll persist looking
> at this until I can point you at something more definite.
David,
took me a while, but I've tracked this one down. It's not as bad as I
feared, it seems that gdb was leading me on a merry trail of false
destruction (refusing to honour any break points or allow me to inspect
the values of any scoped local variables, in the area of the error, while
other scopes within the same file are fine, go figure). Anyway the problem
is the call:
astSet( fitschan, "Warnings=%s", astGetC( fitschan, "AllWarnings" ) );
which now throws an error:
! astSetWarnings(fitschan): Unknown condition 'badctype' specified when
setting the Warnings attribute.
that I was silently ignoring. In fact this causes GAIA to fail to read any
WCS anywhere, not just Jamies.
Peter.
|