Version 100-2.0 of NASA FITS Standard released

[Bill Cotton]

Steve Allen writes:
 > On Mon 1999-04-26T10:36:23 -0400, Bill Cotton hath writ:
 > >    This is yet a further redefinition of "deprecated".  My
 > > understanding was that the weasel words were intended to apply to new
 > > applications (projects), not new data.  Many new random groups FITS
 > > files are generated daily and will for the foreseeable future.
 > > The BLOCKED keyword is largely irrelevant and EPOCH was a mistake, at
 > > least in its intended usage.  Deprecation of random groups is a
 > > denial of present reality.
 > 
 > Dr. Cotton,
 > 
 > The discussion could be more focussed it you would make it clear
 > whether you object to the word or the concept of deprecation.
 > 
 > For example, I think it would help to know whether you would object to
 > a longwinded "Prithee do not employ this for any purpose which does
 > not already use it".
 > 

   I object only to the term "deprecate" which I interprete as meaning
"likely to be disallowed".  I am certainly no fan of random groups and
was part of an effort to get it replaced in the radio community by a
tables-based scheme. Unfortunately, this effort largely failed.  The
radio community isn't in a position to make this sort of change for
the foreseeable future as most working software is on minimal (or
less) life support.  On the other hand, discouraging use of random
groups in new applications is entirely appropriate; I would have no
problem with your wording or something of this nature. 

   A common current practice is for radio telescopes to produce raw
data in binary tables format but processed interferometer data is
almost always archived and interchanged between packages in random
groups format.  I object to even potentially disallowing a working
format while it's still being actively used.

   The question is how to express disaproval of the usage of random
groups in future applications, if that is indeed the intent.  Don
Wells' tome on "deprecate" illustrates the ambiguity of the term in
common usage.  A simple phrase like "use discouraged in future
applications" (or even the suggestion above) uses only words whose
meaning is relatively unambiguous.  

-Bill Cotton