OBJECT names (was Re: Draft of FITS standard revision 1.2..)

[Don Wells]

Dear NOST Technical Panel (and FITS community),

I recommend that some additional text be appended to the single
sentence associated with the 'OBJECT Keyword' entry on page 23 in
Section 5.4.2.2 (Keywords Describing Observations).

This text should recommend that new software implementations which
read and write FITS files should support OBJECT strings with lengths
of at least 26 characters, and preferably 32 (maybe 36?) characters.

The text should also recommend that data systems which originate
OBJECT strings (e.g., telescope/instrument data systems) encourage
observers to use strings which conform to the recommendations of the
Designations Task Group of IAU Commission 5 (Astronomical Data). The
text should cite "IAU Recommendations for Nomenclature"
<http://cdsweb.u-strasbg.fr/iau-spec.html>.

-Don Wells

	  -=-=- Background information on the problem -=-=-

1) Archivists at the Data Centers routinely encounter ambiguous
designations, which are due in part to the habit among astronomers of
truncating existing designations. It is likely that often designations
are truncated in order to fit into the 8-character effective limit to
FITS "OBJECT" strings. The decision by the NOST Technical Panel that
in Section 5.2.1 (page 16) they should drop the '..previous
requirement that reading the data values should not require decoding
any more than the first eight characters of a character string
value..' will make it possible to eliminate the FITS excuse for
truncating designations.

2) Astronomers are using source surveys across many different
wavelength regimes so the need for the acronym part of the designation
is becoming essential and thus 8 characters is insufficient to
uniquely identify sources whose designation involves coordinates.

3) Astronomers who generate surveys and catalogs are now voluntarily
submitting their acronyms to the IAU designations registry and thus
having their designations screened by the Designations TG for
non-conformity before publication. There is a growing recognition that
this process eliminates a significant source of ambiguity and
confusion in the literature and archives.

		-=-=- OBJECT length requirements -=-=-

The IAU Acronym Registry <http://vizier.u-strasbg.fr/cgi-bin/Dic>
already contains a registered entry for the Sloan Digital Sky Survey
which needs 24 characters, and at least two additional suffix
characters should be allowed for ambiguity resolution and slightly
longer acronyms. This is the basis for the recommendation of 26
character minimum length for software implementations of OBJECT.
Probably it will be prudent to support 32- or maybe even 36-character
OBJECT lengths in new datasystems. Examples:

    0        1         2         3
    123456789012345678901234567890123456  Length
    ------------------------------------
    UCAC 39100201                           13
    NGC 598:HBW 1282                        16
    ICRF JHHMMSS.s+DDMMSS                   21
    FIRST J094044.5+363328                  22 (already in the literature)
    SDSS J125740.44+333827.9                24 (longest current real case)
    Lynds 1427:JHSMS JHHMMSS.ss+DDMMSS.s    36 (extreme hypothetical case)