[fitswcs] WCS -TAB complexities

[William Pence]

Eric Greisen wrote:
> There was a suggestion that we add a paragraph on implementation
> difficulties.  I have been traveling a lot and it is taking my mental
> accuity to new lows - any suggestions on what to say where in the
> paper?  And will this move us forward to consider the thing finally?

I'd suggest adding a few sentences to the end of section 6.1.3 modeled 
roughly on a similar cautionary statement in the description of the 
variable-length array convention in binary tables:

"The -TAB implementation is more complicated than most other WCS conventions 
because the coordinate system is not completely defined by keywords in a 
single FITS header.  Software that supports -TAB must be able to gather all 
the necessary WCS parameters that are in general distributed over 2 FITS 
headers and in  the body of the WCS extension table.  The producers of FITS 
data products should consider the capabilities of the likely recipients of 
their files when deciding whether or not to use the -TAB convention, and in 
general should use it only in cases where other simpler WCS conventions are 
not adequate."

Regarding whether to proceed with the approval of Paper III, it is formally 
up to the executive committee of the IAU FITS Working Group to ask the 
regional FITS committees to review and vote on it.  Personally, I am open to 
the idea of proceeding with this now if there is a general consensus to do 
so.  I had hoped to first do more actual software implementation tests with 
FITS files that use -TAB, but that does not seem practical until 
higher-level support is provided by a user-friendly WCS interface library. 
When Mark Calabretta recently implemented some low-level support for -TAB in 
his WCS library,  he discovered several things which resulted in 
modifications to the text in Paper III.  It is possible that once the higher 
level support for -TAB is implemented this will also produce some more 
suggested changes to Paper III.   On the other hand, I'd be reluctant to 
hold up the approval of the paper for this reason alone.

Does anyone here on the fitswcs list not support having the regional FITS 
committees review and vote on Paper III now?  (Reply to me personally if you 
would rather keep your views private).

Bill
-- 
____________________________________________________________________
Dr. William Pence                          William.D.Pence at nasa.gov
NASA/GSFC Code 662         HEASARC         +1-301-286-4599 (voice)
Greenbelt MD 20771                         +1-301-286-1684 (fax)