[comp.lang.fortran] Update on 8x and IBM

psmith@convex.UUCP (01/25/88)

                     Update on FORTRAN 8x Status

The following letter has been sent from IBM to all current FORTRAN licensees...
which number in the thousands.  The letters are part of a package that also
contains copies of the IBM, DEC, UNISYS, Boeing, and Peritus ballot comments
to the X3J3 Committee.  The package also contains information on where to
write for additional information, a table of X3J3 votes, and a summary of
major additions to FORTRAN specified in the proposed FORTRAN 8x standard. 
------------------------------------------------------------------------


November 4, 1987

To:  Users of IBM Fortran Products

Subject: Fortran Standardization 

The proposed new Fortran standard is available now through February 23, 1988
for public review.  The purpose of public review is to learn whether the addi-
tions proposed for Fortran are acceptable to its user community.

IBM believes that the addition of many new features and the replacement of some
current features would result in a language that is significantly larger and
more complex than FORTRAN 77 (replaced features have been kept for compatibil-
ity).  IBM believes this increased complexity will impair Fortran's traditional
strengths of simplicity, high performance and universal availability.  We are
not persuaded that the potential benefit of many of the proposed new features
outweighs the risk.

I am concerned about how this proposed standard may affect your programming 
productivity.  These concerns are elaborated in the attachments.  I strongly
urge you to review the attachments and obtain and review the proposed standard.
In this way you may independently assess the potential benefits and costs.

Your review comments should then be sent to the address in the attachment during
the public review period.  Your comments do not obligate the committee to make
changes; rather it is the specificity and the number of comments from individ-
uals and companies that are persuasive.

Thank you for your interest in this matter.

Morris Taradalsky
IBM General Products Division
Vice President of Programming

----------------------------------------------------------------------------

                          My Comments

IBM is taking an active stand against the standard.  I have talked to two
installations that have received these letters and each has written a response
to the committee.

As IBM's Taradalsky stated in his letter, "...it is the specificity and the
number of comments from individuals and companies that are persuasive."  There
need to be lots of letters written and each must have a specific message
of what the user would like to see changed in the proposed standard.

The FORTRAN 8x issue is currently a political one and and not technical one. 
Many negative comments will redirect the committee to solve the technical 
issues.  Remember the deadline for comments and letters is February 23, 1988.

Letters should be written to:

		X3 Secretariat
		311 First Street NW, Suite 500
		Washington, DC   20001-2178

There is no requirement that you purchase the proposed standard or even 
have seen or read it in order to write.  A sample letter might be:

"I am opposed to the adoption of the proposed FORTRAN 8x standard because:

   1.  I think the dependent compilation feature (MODULE/USE) will increase
       compile times to the point of being unacceptable.

   2.  I am not happy that the committee has failed to standardize the:

        - DO WHILE statement
        - INCLUDE statement
        - POINTER data type
        - BIT data type

   3.  It is unthinkable to even consider possibly removing the COMMON, 
       EQUIVALENT, or DIMENSION statements in ANY future standard.

   4.  The language is too complex.  If the user wants to use Ada, then
       let him use Ada.  I'm opposed to Adatran.

   5.  I believe that numerical precision control is a poorly designed and
       unnecessary modification to the language.

   6.  I think that the abstract data typing is a needlessly complex change
       to FORTRAN.  It will not allow for the most efficient code generation
       for data types such as BIT.

   7.  I believe the committee should have followed the charter and standardized
       existing practice instead of inventing a new language. 

   Please revise this proposed standard to fix these and other problems that 
   may have been encountered during the public review period.

                                            Sincerely,

					    XXX

This is a selected set of sample complaints;  there are certainly many others
that could be used in a letter.    Remember the deadline is February 23, 1988
for letters to reach the ANSI committee in Washington, DC. 

cdb@hpclcdb.HP.COM (Carl Burch) (01/29/88)

Like Pres Smith, I'd like to promote the writing of as many public review
letters as possible from the Fortran community.  Actually more, as I'd
like to see all concerned Fortran programmers reviewing the draft Standard
and writing their thoughts - not particularly restricted to the negative 
comments he solicits.  Actually, with IBM, UNISYS and DEC opposing the draft,
I recommend that you not assume that features you want that are in the current 
draft are safe and assured - they are not.

The part I am responding to is the following :

> There is no requirement that you purchase the proposed standard or even 
> have seen or read it in order to write.  A sample letter might be:

Form letters written on the basis of someone else's opinion such as those 
suggested in the base note are hardly a professional response to a request 
for comments on a proposed ANSI standard.  If you are going to sign
comments on anything more complicated than a box of detergent, they
should be your own.  The address to get the draft Standard has been
posted several times in this notes group - use it and then send in YOUR
comments, not mine or Pres's or anyone else's.

To paraphrase Ollie North, we want all your comments - the good, the bad,
and the ugly!  
						Carl Burch
						HP Fortran