Path: utzoo!mnetor!uunet!nuchat!sugar!ssd
From: ssd@sugar.UUCP (Scott Denham)
Newsgroups: comp.lang.fortran
Subject: Re: FORTRAN 8X discussion
Message-ID: <1499@sugar.UUCP>
Date: 27 Feb 88 20:21:50 GMT
References: <3885@xanth.cs.odu.edu> <50500029@uxe.cso.uiuc.edu>
Organization: Sugar Land UNIX - Houston, TX
Lines: 28
Summary: who's afraid of a Deprecated Feature ??

In article <50500029@uxe.cso.uiuc.edu>, hirchert@uxe.cso.uiuc.edu writes:
> 
> For the record, Fortran 8x does _not_ remove COMMON or EQUIVALENCE or anything
> else in FORTRAN 77.  Furthermore, it provides guarantees that most of these
> features (including COMMON) will also be in Fortran 9x.
> 
> In a nonbinding appendix, Fortran 8x warns that the presence of "superior"
> alternative may cause existing features (including COMMON) to fall into
> relative disuse and eventually be removed (at some time in the distant future).
  
 Furthermore, even the dissapearence of these "deprecated" features from
some future FORTRAN standard is not likely to instantly obsolete the 
remaining body of code using those features. Witness the 66 to 77 change,
in which IBM, Fujitsu, and others continued to support BOTH language 
levels (through a compile-time switch) in their compilers. The fact that
a feature dissapears from the "latest" version of a standard does not 
imply that support for all previous standards will instantly dissapear.
As long as there is a reasonable demand, I believe the vendors will 
continue to provide dsome sort of a path to support code written to the
older standards. As for supporting both the old and new and future 
language constructs in a single source module, if we don't give up 
SOMETHING, we're well on our way to the 1 gigabyte compiler!!:}
  
  
> Kurt W. Hirchert          National Center for Supercomputing Applications

                                     Scott Denham 
                                      Western Atlas International