[daip] Re: [forwarded from Jessica Krick] aips help

[Eric Greisen]

I have taken a look at the code in debug with images masquerading as
yours.  The ---CAR is confusing aips no end.  I am suprised actually
since it usually defaults to linear with unknowns, but OHGEO and HGEOM
really do demand coordinates and neither can really believe that an
ra-dec pair would occur on a linear "projection".  HGEOM just refuses
to handle these coordinates, while OHGEO confuses them with ecliptic
coordinates and, having done the conversion, finds that none of your
pixels end up in the image.

I may be able to fix this poor behavior, but those fixes will appear
only in the 31DEC99 aips.  Where did this -CAR coordinate come from?
It is not natural to any telescope system.

Eric