[iaufwg] Registering the FOREIGN extension name

[Thierry Forveille]

Mark Calabretta writes:
 > On Thu 2006/10/05 21:47:59 -1000, forveill at cfht.hawaii.edu wrote
 > in a message to: "Mark Calabretta" <mcalabre at atnf.csiro.au>
 > and copied to: "William Pence" <pence at milkyway.gsfc.nasa.gov>,
 >       "IAU-FWG" <iaufwg at nrao.edu>
 > 
 > >and must not be reused, while that from plain conventions should
 > >just be avoided as much as reasonable. We don't want one brain-dead
 > >historical convention defined by one guy in an obscure corner to
 > >block FITS evolution, if that case ever occurs :-)
 > 
 > You could argue that "once FITS always FITS" means that conventions,
 > having a kind of semi-legal status, can't formally be expired,
 > (archived data might linger for centuries, etc.) but in practice it's
 > hard to imagine the namespace ever becoming so cluttered that it would
 > be necessary.
 > 
Well, the "once FITS always FITS" rule is precisely what I'd like to 
avoid for something that underwent as little review as a convention.
Sure it's a (presumably) valid FITS file and should not cause readers
to crash, but any (mis)interpretation of its content is on a best
effort basis only. I agree that the problem may never occur in practice,
but let's not paint ourselves into the corner of having to live with
every convention forever.