Deprecation of Random Groups

[Lucio Chiappetti]

On Wed, 29 Apr 1998, Stephen Walton wrote:
> With regard to the draft standard's statement that random groups are
> deprecated, Steve Allen comments:
> 
> > Would another term with definition "should not be applied to any new
> > applications" be more acceptable?

  Probably yes, specially if handled as a "recommendation".

> As someone with no vested interest:  I believe that, in context, "deprecated"
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

  Me too. I never used random groups nor ascii tables.
  I will never plan to use the former in X-ray astronomy (and I understood
  they were going to be replaced by binary tables in radioastronomy, but I
  have just learned that was not the case) and I believe the latter can
  also be replaced in a more compact form by binary tables (at least that is
  what I will use for any tabular material).
  
> means that new code which can read the old files can be written, but that new
> applications should not _write_ random groups.  In the same way, several

  One could perhaps state "outside of specific discipline fields where 
  actively used" ?

> classic Fortran-77 constructs are "deprecated" in the Fortran-90 standard. 
> This means that some future standard will delete them, and that a code which

  That's a different story.
  I do not regard the "deprecations" in f77->f90 as always fortunate, and
  anyhow the "once FITS forever FITS" principle does not allow cancellation.

  (Of course if one is writing a SPECIFIC reader, and not a GENERAL PURPOSE
  reader, one can always assume to be more restrictive and RECOGNISE some
  extension and just say "I do not support them")

----------------------------------------------------------------------------
Lucio Chiappetti - IFCTR/CNR - via Bassini 15 - I-20133 Milano (Italy)      
For more info : http://www.ifctr.mi.cnr.it/~lucio/personal.html             
----------------------------------------------------------------------------