From psmith@mozart.convex.com Fri Mar 20 18:23:35 1992
Received: from convex.convex.com by dkuug.dk via EUnet with SMTP (5.64+/8+bit/IDA-1.2.8)
	id AA09378; Fri, 20 Mar 92 18:23:35 +0100
Received: from mozart.convex.com by convex.convex.com (5.64/1.35)
	id AA27980; Fri, 20 Mar 92 11:23:14 -0600
Received: by mozart.convex.com (5.64/1.28)
	id AA00847; Fri, 20 Mar 92 11:23:12 -0600
Date: Fri, 20 Mar 92 11:23:12 -0600
From: psmith@mozart.convex.com (Presley Smith)
Message-Id: <9203201723.AA00847@mozart.convex.com>
To: JLS@liverpool.ac.uk, SC22WG5@dkuug.dk
Subject: Re:  (SC22WG5.82)  Re: Procedures comment by Psmith
Cc: psmith@mozart.convex.com
X-Charset: ASCII
X-Char-Esc: 29

I was wondering when Laurie would jump in... 

>The X3J3 committee has gotten a bad reputation due to some of the members
>and the way they behaved.   I can relate my personal experience of one
>of our company's sales people calling me and saying that a member of
>X3J3 had attacked him on a sales call because I didn't know what I
>was talking about and was un-professional (the words were stronger
>than that...),etc.  Why would anyone jump a salesperson about a technical
>committee issue or member?   That's like me jumping the Sun repairman
>because I don't like what Sun representive said on the committee.   Dumb.
>But it happened.  In fact, on two different occasions that I know about.

>I feel I must reply to this since Presley has claimed the high moral
>ground and has sited the above as an indication of the low. Presley did not
>name me as the person involved in one of the above cases, but I am quite
>prepared to own up.

I never claimed "high moral ground."  And I would have never named names.

>I did refuse to deal with a Convex salesman and I indicated that the
>reason was what I perceived as unprofessional behaviour of a senior
>member of the company's software development team. What had incensed
>me was the note that Presley had put out on the net and distributed
>in wordprocessed form inviting public commentors to "tick here"
>as a means of providing informed commentary on the draft at its first
>public airing. Presley was not a member of J3 at the time and he had
>put out a summary of the contents of F8x which was factually extremely
>inaccurate, including a number of straight errors. He then suggested
>that the public commentors need not read the actual draft but should simply
>send in the comment as suplied by him. I considered this then and still
>do to have been an unprincipled and unprofessional attempt to subvert
>the comment process. If the salesman was upset personally by my refusal to deal
>with Convex as a result I am sorry but I would react in the same way again
>were the circumstances to repeat themselves.

How soon we forget the facts... In fact, the presentation was put together
by the CONVEX member of X3J3 at the time and in fact, there was one 
error in it.  If I remember correctly, it said the IMPLICIT NONE was
not in the draft when it was.  (I could dig out the presentation and 
we could review what the error really was, but that's pointless now...)
It was in fact a mistake, and was not made on purpose.

The second issue that Lawrie was unhappy about was that I supposedly
had some sort of "form" letter that was to be used to respond to the
public review...

What really happened was that we had a User Group meeting and put together
a presentation on the current state of Fortran 8x.  It, in fact, had the 
error discussed above.   We handed out a copy of that presentation,
information on where to obtain a copy of the proposed standard for review,
a copy of the address of where to write review letters, a couple of sample 
letters, and a copy of the committee ballots and responses from the 
X3J3 vote to forward this proposed standard for public review.  That
included the information from IBM, DEC etc.  both postive and negative.
We did not distribute anything for a "tick here" response.    

We, in fact, encouraged people to obtain copies and read them.  We
also tried to educate them to both the positive and negative sides of
each issue.  I think what Lawrie thinks is that our presentation was
all negative.  In fact, each item presented had a page defining the 
item (array notation for an example) and then had one page pointing out 
the positive aspects of it and one page pointing out the negative aspects
of it.  It was a balanced perspective.  And it was educational.

We can fight all day as to whether everyone who commented read and understood
the standard.   I find no place in any of the procedures where a person must
read and understand the standard in order to comment on it.  There were 
Fortran Forums being done to educate people on Fortran 8x... and letters
were generated from people who attended those meetings.  I can show you
letters that say "I attended the Fortran Forum at ... and I want to 
comment on ..."  I'd bet that not all of those letters were from people
who had read and digested the proposed standard.   Did the Fortran Forum
presentations point out any negative aspects to anything?  Or were they
just "sell" jobs?  (Let's be fair here! Our presentation had both sides!)
 
I'll bet Lawrie doesn't read all the laws made by Parliment there in 
England.  But, I'll bet that some of those laws get discussed on the 
TV and Radio...and I'll be Lawrie has an opinion on those laws that are
discussed and what he'd like to see changed about them.  Okay Lawrie...
from now on you cannot write your MP about a law you want changed until
you've gotten a copy, read it and understood it, and then you have the
right to write a letter!!!  Just listening to a discussion of both 
sides of the issue on the TV by knowledgable people does not give you
the right to voice your opinion by a letter!!!

Somehow we sometimes don't see the practical side of what we are saying...

>The comment process was not designed to be a referendum, nor was it meant to
>be a GALLOP pole; it is far too unscientific in its sampling technology for
>that. The comment process is an attempt to obtain final concidered technical
>judgement from as wide a section of the interested public as possible.
>The aim being to ensure that flaws overlooked by the technical committee
>do not if humanly possible continue into the final standard. The committee
>is supposed to take comments seriously and it should expect the commentors to
>given the draft serious reading and consideration before commenting.

I still maintain that if the committee decided to do something a certain
way and the majority of the comments were negative on that, it indicates
something other than all the commentors are wrong because the committee
knows what's best for them...   

The only way to "ensure that flaws overlooked by the technical committee"
is to prototype it first and see if it really works.  Most of what was
added to Fortran 77 was already working and in use in some FORTRAN compiler.

>I do not wish to get drawn into what is largely a discussion about US
>due process, but I would like to remark that one of the things that upset
>a number of non-US participants was that for a very long period scant
>attention was paid by a significant section of J3 to the "due processes"
>required by WG5 and the international community for whom J3 was acting as
>the drafting body. From '85 onwards, through the grand compromise, the 1st
>public comment and on until the Paris'88 meeting J3 had largely gone in
>precisely the opposite direction to that required by WG5 and by the National
>body comments on the CD ballot. When J3 voted not to allow the only variant
>at Jackson to be remotely acceptable to WG5 to be even presented in Paris
>WG5 was not very pleased with the apparant "due process".
>Strange as it may seem I am now going to agree with Presley. The time
>for looking back at the problems of the past is over. A cooperative
>relationship needs to be rebuilt and the rules of the relationship
>between J3 and WG5 need to be spelt out so we do not get into the mess
>again. If that means US procedures might need bringing more into line
>with international ones then so be it.

There's nothing wrong with the US procedures.  They work with all the
other committees.   X3J3 needs to work within those procedures to get
the job done in a more efficient manner.  

Thanks for the comments Lawrie.  This allows us to get the fact out
we really did and how "terrible" it was...

FYI.  Presley