[wfc] Result of WFC Vote on New FITS Standard

[Arnold Rots]

I have the pleasure of announcing the result of the vote of the WGAS
(North American) FITS Committee on the proposal for a new version of
the Statndards document:

    Do you vote in favor of accepting version 3a of FITS Standard
    document as the authoritative standard, and, by implication, do you
    endorse the changes proposed in version 3a?

Yes:            14
No:              0
Abstain:         0
Total votes:    14
Non-voting:      2
WFC membership: 16

Comments submitted by WFC members are appended.

Since 2/3 of the committee membership cast a vote and 2/3 of the votes
cast were in the affirmative, I declare that the proposal has been
approved by the WGAS FITS Committee.

Respectfully submitted,

  - Arnold Rots
    Chair WFC


The following comments were received:

------------------------
The committee is to be commended for an excellent job.  The "response  
to comments" is a textbook example of how such should be handled and  
will prove a useful reference the next time the standard is tweaked.

That said, "once FITS, always FITS" is a very severe standard to meet  
and a number of the textual changes are technically unenforceable  
should challenges arise.  I suspect that no important challenges will.
------------------------

------------------------
The new standard is obviously an improvement,  but I
do have a few questions/comments after reading over the standard:

1) 2.2 Defn of a FITS file - To be consistent with section 3.7, should
a FITS file be defined as "a format that conforms to the
specifications in this OR EARLIER VERSIONS OF THIS standard"?

2) 2.2 Defn of Primary Header - A picky point, but it may not be clear 
from the sentence structure whether the added  ",if present"  refers
to the first header or the primary data array. 
It could be interpreted to mean the primary header is optional.
 
3) 4.3.2 - Last sentence - If a specific unit is "recommended" by
the standard, how can you say it "must" be used.

4) 4.4.2.1 Date Keyword - remove second "a" in first sentence last paragraph
(verbatim copy of a another HDU?).

5) 4.4.2.2 Equinox - Why should an EQUINOX value of 2000 imply ICRS?
How does one indicate FK4 or FK5 coordinates that were precessed to year 
2000?
And what  does "exactly 2000" mean for a floating point value? Maybe I 
misunderstand, but this sounds like a major change that will cause an
incorrect interpretation of at least some archived FITS data (even
though the numerical difference is probably negligible). It would make
more sense to define a RADECSYS keyword (as done in the WCS papers and
the VO SSAP data model paper) but even then, defining a default
reference frame could effect the interpretation of archived FITS files.
------------------------

------------------------
Let me add to my earlier comments regarding the changes to the EQUINOX
reserved keyword definition some more confusion. It is defined in 
section 4.4.2.2 as:

EQUINOX Keyword The value field shall contain a floating-point number
giving the equinox in years for the celestial coordinate system in
which positions are expressed. The special case where EQUINOX is
exactly 2000 is taken to be a reference to the International Celestial
Reference System (ICRS).

I see in section 8.2 that EQUINOXa (where a may be a blank?)
is also listed as a reserved keyword (along with RADESYS) as:

RADESYSa ? [character; default: FK4, FK5, or ICRS: see below]
Name of the reference frame of equatorial or ecliptic coordinates,
whose value must be one of those speci?ed in Table 8.4. The default
value is FK4 if the value of EQUINOXa < 1984.0, FK5 if EQUINOXa >=
1984.0, or ICRS if EQUINOXa is not given.

EQUINOXa ? [floating; default: see below]
Epoch of the mean equator and equinox in years, whose value must be non-
negative. The interpretation of epoch depends upon the value of RADESYSa if
present: Besselian if the value is FK4 or FK4-NO-E, Julian if the value 
is FK5; not applicable if the value is ICRS or GAPPT.

This seems to imply the assumed reference frame should vary with the 
value of EQUINOX, and one should assume FK5 for EQUINOX=2000 and ICRS
if EQUINOX is not specified.
Does this contradict the earlier definition in 4.4.2.2 or am I missing 
something?
------------------------


--------------------------------------------------------------------------
Arnold H. Rots                                Chandra X-ray Science Center
Smithsonian Astrophysical Observatory                tel:  +1 617 496 7701
60 Garden Street, MS 67                              fax:  +1 617 495 7356
Cambridge, MA 02138                             arots at head.cfa.harvard.edu
USA                                     http://hea-www.harvard.edu/~arots/
--------------------------------------------------------------------------