[fitswcs] WCS Paper III regional voting results

[William Pence]

I am pleased to report that all 4 regional FITS committees have completed
the voting on the WCS Paper III.  In all cases, the number of votes in favor
of approving the paper greatly exceeded the minimum 2/3 majority required to
pass.  The actual vote tally is summarized in the following table:

Committee              Yes   No  Abstain  Not Voting
=========              ===   ==  =======  ==========
Japan                   10    0     1        0
North America           15    0     1        0
Australia/New Zealand    6    0     1        0
Europe                  11    0     0        1
                        =============================
              TOTAL      42    0     3        1


In addition to casting a vote, the committee members were encouraged to
offer any comments or suggestions about the paper.  All the comments that
were received are reproduced below.

The final step in the formal approval process of this paper will be
for the IAU FITS Working Group to vote on whether to accept it as
part of the FITS Standard.  This vote will likely take place within
the next couple months, once any remaining issues (in particular
any issues raised by the attached comments) have been resolved.

Bill Pence
Chairman, IAU FITS Working Group

____________________________________________________________________
      Comments Submitted by the Regional FITS Committee Members

The Australian committee chairman sent the following comment:

- The abstainee said "I don't see a problem in voting 'yes', but
   since I'm not properly up with the play, I had better vote
   'abstain'".    Several respondees expressed a similar sentiment
   to a lesser degree - that they had not analysed the paper in
   detail but had sufficient confidence in it to vote 'yes'.  None
   reported any specific shortcomings or misgivings.

The following 3 comments came from the North American committee:

- With some hesitation, I vote YES on WCS Paper 3.
   I continue to have some reservations about the maturity of the
   -TAB part of the proposal and hope it will be more widely deployed
   in software by the time the IAU votes. In particular, I find -TAB's
   use with coordinate systems in binary tables rather confusing still.

- There is no question that this is a major resource for helping people
   to understand spectral WCS.  One could wish the interface were better
   structured so that the simple cases are simpler, but so long as it
   works we should probably declare victory.

- We need this proposal, but I keep my fingers crossed that the -TAB
   component will be up to the task it is supposed to perform.  It
   probably will, but the case has not been made to the degree that I
   would have liked to see.

The following comments came from the European committee:

- for celestial coordinates (Paper I) keywords for expressing errors
   and/or uncertainties were introduced.  This seems to be missing for
   spectral coordinate although it may be of interest.

- the -TAB provides a table lookup scheme.  It was not fully clear if
   it in some way would be possible to express discontinueties.  Due
   to the usage of masks in CCD production there sometimes can
   be a slightly different gap between some rows/columns.  Such
   discontinueties are assumed not to exist but I did not find it
   explicitly said.

- Having read the latest version over the last day or so, I have found a
   number of typographical and punctuation errors, and the odd place where
   the wording might be improved to clarify the meaning.  I'll send the
   list to Eric.  These don't change the material content.

- In terms of omission my biggest disappointment is that no provision has
   been made for echelle and objective-prism data.  In these cases I'd have
   liked the paper to offer some advice on the additional metadata required
   to extract and determine the WCS of tabulated spectra supported by Paper
   III (given that such data are very common and other cases are covered in
   great detail).  Since we're dealing with reduced data, as opposed to raw
   data, we're placing an onus on non-FITS package writers and individual
   astronomers to get their metadata to conform to this paper.  OK there is
   a trend towards data centres offering the reduced data, but it's far
   from comprehensive.  Having trying to hack some astronomer code for the
   64-bit era recently, inclusion of correct WCS is likely to be the
   exception rather than the rule.

- Also an example of the WCS headers for tabulated spectra (e.g. echelle
   or objective-prism or multi-object), perhaps including the Green Bank
   convention, would be welcome.  It is a complicated paper and if we want
   institutions to support it, offering more examples and templates will
   help bring this about and generate accurate headers.

- What's the provision for the extent or widths of pixels, e.g. as needed
   for binned time-tagged photon events?  In the Starlink NDF data
   structure an AXIS component---the equivalent to the -TAB
   facility---could have an array of widths too.

- I did also wonder if there were any special formulae and headers needed
   for X-ray spectra akin to those presented in Section 5.

- Perhaps I missed them, but are there recommendations when to use -TAB
   rather than some model-based headers like those in Section 5?  Can there
   be both in cases where the model doesn't offer a particularly good fit
   to the empirical wavelength axis?  I would think so, given that one is
   allowed multiple WCS representations via the `a' (= A--Z) symbol in
   keywords.

-- 
____________________________________________________________________
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)