No subject

intuitive.  It is one of the most common requests for spectral format
support in IRAF.  So I am supportive of expanding paper III, which
introduces the tabular WCS concept, to allow this form of coordinate
representation in at least 1D data.

I haven't seen any difficult obstacle with this.  The sorting issue is
something which I don't think we need do anything about other than provide
strong warnings.  After all, it is also possible to have pixels in an image
be sorted (or more commonly subregions extracted) which would invalidate
the WCS without some further action.  The cuurent concern seems to be how
to distinguish vector and column representations.  This could be addressed
with a related CTYPEia.  Most of the concepts would be the same so it
would be defined in the same section.  For instance CTYPEia = xxxx-COL.

By the way I completely agree that logically there should be no distinction
in the WCS formulism between ASCII and BINARY FITS table formats which
are symmantically the same.

Frank Valdes