psmith@mozart.uucp (Presley Smith) (10/25/89)
THE SKY is FALLING and Other Myths
In response to various messages on the net:
"Papa Presley" was only reporting the news and giving his view on the
subject. I respect each person's right to their view on this subject. Walt
and others have expressed different views with varying degrees of emotion.
Lets look at few more facts.
There were three people from X3J3 at the X3 meeting: Jerry Wagener, the
Vice Chair, Bill Leonard, and myself. The three of us are the only ones
that saw the debate and who know what was said during the debate. Let me
summarize some additional information that might be helpful:
X3 LB 1235 was sent out on August 16th with the following wording:
"SPARC recommends to X3 approval of the attached revised Project
Proposal for Project 67-D. SPARC noted that this revised Project
Proposal calls for retention of the current ANS X3.9-1978 and the
processing of the new FORTRAN 8X proposal as a separate standard.
SPARC also noted that, in making this recommendation, it is
intended that the current ANS X3.9 not be the subject of future
revision, consistent with X3J3 concerns."
This letter ballot passed at X3 by a vote of 26-10-2-2.
On Wednesday of last week, X3 asked the question "Should this ballot be
re-considered?" The answer was 5 Yes, 23 no, 2 abstain, and 10 not voting.
In fact, one of the yes votes changed to a no at the meeting so the
final vote on LB 1235 was 25-11-2-2. And so the vote stands.
Let's look at what this ballot says:
1. This ballot NEVER mentions ISO or international standards. "this
revised Project Proposal calls for retention of the current ANS
(That's AMERICAN NATIONAL STANDARD!) X3.9-1978 and ..."
The revised project plan for Fortran 8x calls Fortran 8x by the name
FORTRAN Extended and states:
"The standard is a document that specifies the form and establishes
the interpretation of programs expressed in the FORTRAN language.
The document will be a superset of American National Standard
Programming Language FORTRAN X3.9-1978."
It's obvious from this revised project plan that X3 intends for
Fortran 8x (FORTRAN Extended) to be a SUPERSET of FORTRAN 77.
2. X3 specifically discussed the international impacts during this debate.
In particular, in October, 1988, X3 directed X3J3 that: "The U.S.
support a single Fortran standard (domestic/international, i.e., one
document)." The view that was expressed in the X3 meeting was that
X3 was supporting a single Fortran standard, Fortran Extended, as
both the international and domestic standard.
Walt pointed out in one of his articles that "As a member of ISO,
ANSI can adopt any ISO standard; ANSI can also adopt additional
different standards. The recent X3 vote says they want FORTRAN 77
to be an additional standard (it is unlikely that ISO will retain
it {my addition: meaning FORTRAN 77} as a standard."
What Walt has said is what is going to happen. The U.S. is going
to keep FORTRAN 77 on the ANSI books as an active ANSI standard
in the UNITED STATES. Fortran 8x, the ISO standard, or whatever it's
finally called, will be available as an ANSI standard in this country.
It will just have an ANSI number different from X3.9-1978. It is
certainly INTENDED by SPARC and X3 that 8x be a superset of FORTRAN 77
as was stated in the revised project plan.
The various messages on the "international impact" of this vote
are nothing but emotion. X3 is NOT telling any other standards
organization what to do. ISO can decide if they want to keep
FORTRAN 77 or not. What X3 did was a United States only issue.
One comment on the net was: "In other words, ANSI/X3 won't discuss
anything with ISO, and indeed will do the exact opposite of what ISO
wants, and then claim if ISO refuses to be railroaded into changing
its mind, then any problems with different standards are ISO's fault!"
X3 didn't need to discuss anything with ISO. X3 did NOT ask ISO or
anyone else TO DO ANYTHING! X3 only specified the action that
ANSI was going to take. Sounds to me from statements like this that
ISO wants to railroad X3 into doing whatever ISO wants... I don't
believe that the action that X3 took "railroads" anyone.
ISO could certainly go their own way if they want, but if they are
serious about wanting a single standard as soon as possible, then
they are going to ballot on the Fortran 8x that is out for public
review and adopt that as the ISO standard.
One comment was "Unless WG5 as agent for ISO changes its tune there
will be no Fortran Extended Int`l standard. There will simply be
an international Fortran standard." This is exactly what X3 expects.
There will be an international Fortran standard and in this
country, FORTRAN 77 will remain an active U.S. standard.
One of the comments was: "The short answer is that ISO has basically
accepted the draft (modulo edits) and the US has backed away from it."
That's not true at all. The U.S. is planning to have Fortran 8x
as a standard. It will just be a separate standard from FORTRAN 77
and NOT replace the FORTRAN 77 standard. The X3 action does nothing
to make the US back away from Fortran 8x.
3. One comment was "no major updates to that standard are anticipated...
Not anticipated? There is nothing to prevent it, and ..." That's
right, there's nothing to prevent it. But there's also nothing to
prevent some group from going to ANSI and saying they want to
use FORTRAN 77 as a starting point and produce Scientific Language 1
either. Even if Fortran 8x replaces FORTRAN 77, anyone could start
with FORTRAN 77 and produce any updates desired and get that resulting
language standardized. So, what's the point? Either way someone
could take the current FORTRAN 77 and make a new FORTRAN out of it.
This is nothing but a red herring.
4. One comment was "Who will benefit?... Vendors who have dominated
their market segment with their X3.9-1978 extended compilers.
Adopting a single (new) standard would have promoted competition
in each niche." Well, I can tell you that LOTs of FORTRAN users
are going to want their favorite vendor extensions in their new
Fortran 8x compilers. Either that or they are going to get to
rewrite a lot of their code to remove those vendor extension that
they have used in the past.
There is also nothing preventing users from switching languages
either. Plessey, a British company, sent a letter to X3 expressing
their desire to see FORTRAN 77 retained as a separate standard because
the did NOT expect to write any new code in FORTRAN and wanted to
be sure that their cost of maintaining the old code was as minimal
as possible.
The success of Fortran 8x does not depend on hardware vendors
implementing it. A whole group of language vendors have emerged to
produce Ada systems for a variety of machines. Verdix, Telesoft,
Softec, etc. all produce Ada systems for various machines and MULTIPLE
hardware vendors have purchased those products and ported them to
their machines.
Free enterprise will open opportunities for companies to cash in on
producing Fortran 8x products IF the demand is great enough to justify it.
The hardware vendors will provide the users with what they demand when
the users bring the vendors the purchase orders for the products. If
the hardware vendors don't have the product and compiler house has the
product, then that compiler house gets the money and the business.
You should also note that the X3 committee is made up of 1/3 producers,
1/3 consumers, 1/3 general interest members. Since both votes were
carried by over a 2/3 margin, this is certainly NOT just a vendor
action within X3. The X3 representatives for Share, Guide, and DECUS
all voted to NOT reconsider the X3 ballot.
5. One person reminded me that I had said "Then your favorite vendor
will be in a position of which standard to implement and FORTRAN
could go the way of other languages that have gotten into similar
situations." In the next paragraph that person noted that I also
said "stay tuned for more information from the WG5 meeting in Paris...
the future of FORTRAN may be being decided this week..." I made that
statement prior to the WG5 meeting in Paris when it was unclear to
me that WG5 and ANSI would not part ways and each develop their own
standard. The X3 resolution just says that the U.S. will maintain
FORTRAN 77 as a defined subset. That's not the same issue as really
having two different standards.
Since the U.S. is maintaining FORTRAN 77 as a PROPER subset and
NOT just as a part of FORTRAN 8x, it gives the compiler writer an
option to really retain the implementation of the FORTRAN 77 that
is currently available and enhance that implementation into an
8x implementation over time.
6. The resolutions from the SC22 plenary are NOT out of sync with what
X3 did:
AH. SC22 stated that it was the intent that the second DP 1539
(the current public review document) be the basis for the
international Fortran standard.
The X3 vote did not change that. X3 and ANSI are now referring to
DP 1539 as FORTRAN Extended and expect ISO to pass the document as
the international Fortran standard and for it to be a US standard also.
AS. SC22 notes interest in the US member body in retaining
FORTRAN 77 as a national standard...
Also notes that WG5 doesn't recommend this.
Also requests that WG5 work with X3J3 to reconcile these conflicting
positions.
X3 and SPARC were both aware of the X3J3 position and the WG5
position. There was discussion that this action might keep pressure
on X3J3 and WG5 to work out the conflicting positions.
I'm sure that this message will inspire more flames from all sides.
I now want to express one additional opinion of mine... having been at
the X3 meeting. This is JUST MY OPINION:
I believe that X3 recognizes that certain people and certain groups
have NOT been willing to work in the spirit of compromise on this
standard. There's been a lot of claims of compromise...if you'll
just do it my way. If you look at the responses on the net, you will
see the old, unchanging positions that have gotten us to this point
such as the following examples from the various messages:
If the US doesn't do what ISO wants... Another symptom of the economic
decline of the US... Most of the "compromises" were put in to appease the
"reactionaries"... etc. etc.
Those FOR Fortran 8x continue to vote YES and those AGAINST it continue
vote NO. Compromise and consensus is achieved by working together to get
the NO votes to a YES vote. At the meeting, X3 discussed the rigidity of
the positions that each side has taken and lack of any progress toward
obtaining a real consensus. This deadlock was obvious from the fact that
the number of NO votes was remaining fairly constant on each round and this
indicates a problem. If the X3J3 committee is really working on obtaining
consensus, the number of NO votes should be decreasing.
Concern was expressed in the X3 meeting that removing FORTRAN 77 as an
active standard was not in the interest of a significant subset of the
FORTRAN community. This was emphasised by letters such as the one
from Plessey. One of the goals of the standards process is to produce
standards that address the needs of as many in the community as possible.
This was discussed at the X3 meeting.
X3 has now forced compromise within the U.S. by this action. In doing so,
I now expect some of the strong NO votes to become YES votes for Fortran
Extended. There was a lot of concern expressed at the X3 meeting about
appeals similar to what has gone on in the ANSI C standards efforts.
It's my belief that X3 is trying to eliminate as many confrontations
as they can in an attempt to INSURE that Fortran 8x will be a standard in
the U.S. I believe this action was taken to force some of the compromises
that X3J3 and WG5 were not willing to make...
The blasts on the net have been the old SKY is FALLING... but in reality
all X3 did was say that FORTRAN 77 would remain a proper subset of
Fortran 8x and in the US only. The sun is still shinning here today
and the sky is still blue... and Fortran 8x is still on track to become
both a domestic and an international Fortran standard.
End of MY OPINION... and this LONG article...