Newsgroups: comp.lang.lisp
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!news.kei.com!sol.ctr.columbia.edu!howland.reston.ans.net!pipex!dircon!rheged!simon
From: simon@rheged.dircon.co.uk (Simon Brooke)
Subject: Bias in X3J13? (long and possibly boring)
Message-ID: <Cy5F1E.4xq@rheged.dircon.co.uk>
Organization: none. Disorganization: total.
References: <37eb4h$k4f@vertex.tor.hookup.net> <hbakerCxquDG.LEF@netcom.com> <Cxxwx0.1nC@rheged.dircon.co.uk> <3854ul$r5r@cantaloupe.srv.cs.cmu.edu>
Date: Sun, 23 Oct 1994 23:10:24 GMT
Lines: 165

In article <3854ul$r5r@cantaloupe.srv.cs.cmu.edu>,
Scott Fahlman <sef@CS.CMU.EDU> wrote:
>
>Simon,
>
>This pile of conspiracy theories about Lisp (or "LisP" as you call it)
>is truly amazing.  

Scott

You were very close to the centre of these events, whereas I was the
other side of the Atlantic ocean and *very* much a bit player -- not
involved at all until after the gunsmoke had started to clear.
Consequently, your recall and your information is bound to be better
than mine. I have been interested to learn from many of the things
which you write, both in your response now and in other work of yours
which I have read.

However, I believe you overstate some parts of the case you make in
your response to me.

>But seriously...  I won't try to refute all of the points in your
>message that I disagree with -- just about everything -- but will
>address a couple of them.
>
>   At the time of the formation of X3J13, overwhelmingly the largest
>   company seriously involved in the LisP business was Xerox.

>Well, the largest company doing anything with Lisp implementation was
>probably IBM.

I'm not aware -- perhaps I'm inadequately informed -- of any LisP
machine marketed by IBM; certainly not of one which was ever to any
extent widely used. Similarly I'm not aware of any major widely used
LisP implementation developed by IBM. That IBM is/was a large company
is true; that it was in the mid eighties a significant player in the
LisP market is a claim which is new to me.

If you say that Symbolics, TI and Lucid all had a larger share of the
LisP market than Xerox, I am sure that you must be right. Certainly,
you have better information on this point than I. In this country,
during the middle eighties, I was aware of probably 70 Xerox machines
in active use in academia and industry, two Symbolics, and no
Explorers (I always wanted a Symbolics, but could never pursuade
anyone to buy me one). Were the ratios greatly different in the States?

>The Xerox people
>declined to participate in the Common Lisp effort early on, so the
>language naturally reflected the established tastes of the Maclisp
>clan.

A number of people have kindly written to set me right on this point.
I had not remembered, or had not appreciated, the fact that Xerox had
been invited early, and had declined. It is certainly true that Danny
Bobrow and Mark Stefik were involved in the CLOS specification
process. The impression we got this side of the pond (and very far
from the protagonists) was that this was not entirely a happy
business.

>
>So the suggestion that X3J13
>was a conspiracy to wound Xerox is ridiculous.
>
But I made no such allegation. I was careful to say:

...the essential aim of a substantial number of the participants in
X3J13 was to make Common LISP as different as possible from
InterLISP...

>   This allegation explains both the comments system and the choice of
>   LISP2, two decisions each of which are otherwise inexplicable. I am
>   prepared to believe the claim that the case-insensitive reader was a
>   requirement of the United States Department of Defense.
>
>Any technical decision you disagree with is inexplicable, unless it's
>part of a political conspiracy?  You must be lots of fun to work with.
>

That isn't really fair. There are lots of technical points I disagree
with in Common LISP. These two are particularly peculiar. Can you name
one person currently writing on the design of functional languages who
would defend the double name space? See for example Gabriel & Pitman
in L&SC 1,1, in whose acknowlegements you are yourself credited.

Similarly, a comment system which effectively prevents in-core working
*must* have been deliberately provocative. So many other potential
solutions would have allowed both in-core and file-based development
styles to co-exist.  This one did not. It was not chosen by tossing
coins, surely?

>The case-insensitive reader has nothing to do with DoD requirments.

If you are sure of this, you must be right. I have to say that this
explanation has been given to me in good faith by a number of people
whose good sense and judgement I trust, not as something they believed
or surmised but simply as a fact.

>It was pretty much the norm throughout all of
>computerdom until Unix and the Saucer People started rotting people's
>brains.  :-)

So were core storage and magnetic drums. Which is to say, by the time
the Common LISP process started, it was industrial archaeology.

>   I do not believe your claim that '...the slump in the LisP
>   market...'  was '...well underway...' in 1984, when the aluminium book
>   was published.
>
>The claim was that the slump in Lisp was well underway when X3J13 was
>formed.  The publication of the first edition of CLtL was years
>earlier, early in the AI boom.  Actually, I don't think the Lisp
>market slumped until some years after X3J13 began its work.

Again, I think you exagerate. I certainly have been guilty of using
'X3J13' loosely to refer to the Common LISP standards process,
but still... my first edition copy of CLtL, on whose title page you
are credited as first contributor after Steele, was published in 1984.
The earliest X3J13 papers which I've kept are dated 1987, but clearly
refer to earlier work; I haven't anything to hand which states
definitely when X3J13 was formed, but it can't have been that many
years after CLtL. In any case, the actual claim was:

  'I don't think the X3J13 work in any way contributed to the slump in the 
  Lisp market, which was well underway before the result of their efforts 
  became widely available.'

In so far as X3J13 was primarily a clean up process on CLtL, its
results 'became widely available' with the publication of the book.
However, in saying this, I'm aware how much I was, in my earlier
posting, conflating X3J13 with the 'gang of five'. I apologise for
this, although I'm not sure that it will make me any more acceptable
to you.

I hope you will accept this posting as a reasoned and reasonable
response. It is not ridiculous or paranoid to observe that commercial
axes are commonly ground in standardisation processes. You yourself
wrote, in an email message dated Sunday 12 September 1982

	However, Common Lisp is politics, not art.

I concede that in not being aware that Xerox had been invited early in
the process, I may have overstated a case and made some unfair
inferences, but the case is still there. To quote your own words
again:

	When my grandchildren, if any, ask me why certain things
	turned out in somewhat ugly ways, I will tell them that it is
	for the same reason that slaves count as 3/5 of a person in
	the U.S.  Constitution -- that is the price you pay for
	keeping the South on board (or the North, depending).

Some things did turn out in somewhat ugly ways. Your explanation --
that standardisation is an art of compromise -- undoubtedly explains
much of this. But it does not contradict my assertion that, for some
members, there was a commercial imperative to make the standard more
like their own existing product, and less like their competitors.

Final note: I use the form LisP because the word is a contraction of
List Processing. Thus L and P are initials, i and s are not. I
appreciate that this is an idiosyncracy or fad. If it irritates you, I
apologise.

Simon
-- 
---------simon@rheged.dircon.co.uk
