[fitsbits] Re: FITS keyword typing
Steve Allen
sla at ucolick.org
Fri May 13 13:29:53 EDT 2005
On Fri 2005-05-13T19:01:21 +0200, LC's No-Spam Newsreading account hath writ:
> It is true that header keyword are not strongly typed.
> I do not see any easy way of changing this in the "once FITS forever
> FITS" rule.
The IVOA efforts are busy defining precise semantics for everything
that constitutes an "observation". In the syntax of UML/XML/UCD which
they are using to describe the universe everything is strongly typed.
The descriptions of data that they are producing have corresponding
FITS serializations which permit the information to be transmitted
over a wire. Those serializations are defining FITS keywords which
correspond to the entities in the VO data models.
As of last month there are two MIME media types for indicating that
FITS files are being transmitted over the internet.
(Ah, yes, let this serve as a proto announcement. As seen in
http://www.ucolick.org/~sla/fits/mime/
the IETF, IESG, and RFC Editor have acknowledged the publication
of RFC 4047. We await only final action by the IANA.)
The media types reference the FITS standards, and thus the syntax of
the data are nailed down. What remains is to define the semantics of
some FITS files in a more rigorous sense which is described in a
machine-readable format. I see the VO effort laying the groundwork
for exactly that, and this is consistent with the ruminations from IAU
Comm 5 meeting at Sydney
http://fits.gsfc.nasa.gov/iaufwg/sydney_minutes.html
that the FITS and VO communities need to keep aware of each other.
Strong typing for FITS keywords might be achieved as simply as
defining one new standard keyword that indicates the URL of a VO
document giving the detailed semantics of the keywords and table
structures within a FITS file.
--
Steve Allen <sla at ucolick.org> WGS-84 (GPS)
UCO/Lick Observatory Natural Sciences II, Room 165 Lat +36.99858
University of California Voice: +1 831 459 3046 Lng -122.06014
Santa Cruz, CA 95064 http://www.ucolick.org/~sla/ Hgt +250 m
More information about the fitsbits
mailing list