COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002				       May 6, 1981
C00017 ENDMK
C⊗;
			       May 6, 1981






Dr. George H. Heilmeier
Vice President, Corporate Research Development, and Engineering
Texas Instruments, Inc.
M/S 400, P.O. Box 225474
Dallas TX 75265



Dr. Ralph Eugene Kromer
Director, Computer Science Laboratory
Texas Instruments, Inc.
M/S 371, P.O. Box 225621
Dallas TX 75265






Dear Drs. Heilmeier and Kromer:

I am writing to  follow up on my  conversations with you both  a few weeks  ago.
Gene had the unfortunate  position of fielding the  brunt of my displeasure,  my
delivery has mellowed somewhat in the meantime, but the message is the same:   I
am not at all pleased with the way TI does business.  Unfortunately, indications
are  that  this  kind  of  thing  has  happened  before;  unfortunately,   those
individuals did not react  as I did; so  it remains for me  to expose the  sore.
This I do without pleasure, but I would hope that this exercise will assure that
such a situation will not occur again;  I don't happen to believe that any  good
can come from remaining silent when one is involved in an unpleasant situation.

It would be more encouraging  if I could point to  an individual or a  condition
and say  "that's  your difficulty";  unfortunately  it's not  that  easy:   I've
personally enjoyed the people I've dealt  with at TI; professionally, I  applaud
your plans to utilize AI  technology.  Perhaps it's a  size problem --a lack  of
corporate direction; whatever it is, it can be fatal to small companies that try
to deal with TI, and will be fatal to TI, professionally, if not financially.

So let me relate my  TI experiences over the  last eighteen months, and  perhaps
you can put them in your perspective.

Things began  on a  positive note  in 1979,  with net  correspondence from  Mark
Miller about TI's plans for LISP on their machines.  At that time I was  putting
the finishing touches on TLC-LISP for the Z-80 and gearing up for the 1980  LISP
Conference that my wife and I were running. In early 1980 it seemed that TI  was
ready to begin serious negotiations for  a "TI-LISP" (as I understood it  later,
"George Heilmeier said `Why don't we have  LISP on TI's machines?'" and since  I
was doing quality LISP on smaller machines, I was approached). In April of  1980
I came to  Dallas and  spent two  days talking with  people about  what kind  of
features TI-LISP should have, and what architectural features would be good  for
LISP.

A second motivation for  my trip to  Dallas was to pursue  funding for the  LISP
Conference. Since all funds were coming  out of my pocket, and all  preparations
were being done at home  by myself and my wife,  the time and financial  burdens
were severe.   Again,  indications  through  the  net  were  that  TI  would  be
interested in trading  some funds for  some exposure in  the LISP community.   I
didn't ask for a lot:  $1500 for the Conference reception, meaning that I  could
free up money  to pay  someone to  help with  the postage,  and processing,  and
phoning, and ....

A few weeks  later, I was  quite disappointed to  hear that funds  would not  be
available to help support  the Conference.  Doubts about  TI began to appear  at
this early stage: here was a corporation, making great claims about committments
to AI and LISP, asking TLC to  define the definitive LISP for TI, yet  unwilling
to help  relieve  my burden  in  handling that  Conference.  Unfortunately  TI's
reaction was not singular; pleas for financial aid went universally  unanswered.
One might argue that anyone who tries to put on an international conference with
no funds  or  facilities, is  clearly  suffering from  delusions  and/or  faulty
judgement and therefore not to be trusted, but that's another issue. In fact, we
did do it; without funds;  and with only a few  volunteers in the last days;  it
screwed up eight months of our lives, but it got done. Unfortunately, one  thing
that didn't get done was TLC's business; I  had to call Mark and Truman to  tell
them that there was no way I  could get the Conference business attended to  and
pursue TI's interests simultaneously.  TLC went on "hold" for the next 6 months.

In early September,  I called Mark  and Truman again  to resume our  discussions
about TI-LISP. In  the interim it  had occurred to  me that there  was a  larger
issue than TI-LISP that  TI could address; in  particular, the problems  finding
LISP programmers, and finding  markets for LISP-based products.   TI could do  a
larger service,  while building  their  own talent-base.   This  was part  of  a
long-range plan of mine in education of mathematics and computer science people.
We discussed  these  issues  and decided  that  a  meeting with  Gene  would  be
appropriate. At that meeting, it was clear that my long range goals were not  to
be supported. With more than a little irony I was told the "gee, six months  ago
we'd have no trouble funding this effort, but we're committed to different goals
now"; I had to bite real hard not  to retort that "six-months prior I was up  to
my ass  in  LISP Conference  details,  and if  TI  would have  helped  with  the
Conference only slightly ..."

Furthermore, my doubts continued to be raised about the short-term TI-LISP plans
as well. For in that  meeting, several guidelines were  laid down by Gene:  

 "(a)George Heilmeier wants TI-LISP defined by December 1980; 
  (b) it must run on  the LISP machine 
  (c) it must run on the DEC-20;
  (d) it must be portable to the new TI machines; 
  (e)  it must  be a  professional  (non-toy) LISP;  
  (f) it  must  cost substantially  less   than  $400K".    

These are tight constraints,  indeed; particularly since  this meeting was  late
October 1980.

With a sinking feeling, my reaction in that meeting was to specify  LISP-machine
LISP; the attendees concurred. I was not at all pleased when I left Dallas  that
time, because I sensed that the constraints, in fact, had no solution.   However
I was assured that that the money for the TI LISP effort was in the budget,  and
that a  short proposal  addressing the  constraints would  suffice to  keep  the
dialog open to free some of the conditions; I'm afraid I didn't believe it,  but
I tried.

Over the  next two  months I  examined the  options for  a LISP  satisfying  the
conditions: indeed, one would be hard-pressed  to turn LISP-machine LISP into  a
portable system  without  substantial  rewriting. The  only  viable  alternative
seemed to be  InterLISP. I  sent a  despairing draft  of the  proposal to  Mark,
finally.  In this time  period I communicated with  Mark several times about  my
growing uneasiness about the whole situation.  His responses were not "glowingly
optimistic":  budget cuts were  frequent, questions of how  to supply me with  a
machine came  up  from  Gene,  displeasure over  the  InterLISP  suggestion  was
substantial.

I concurred on the InterLISP issue, but reiterated the constraints as laid  down
up Gene.  Mark  suggested  that  December  wasn't  that  solid  after  all.   An
Alice-in-Wonderland atmosphere  was  seeping  through  the  cracks.   

I naively  suggested that  we could  finesse  the machine  problem since  I  was
getting a  68000-based  machine for  another  LISP  project.  I  was  told  that
wouldn't do:  can't use the competition's hardware --seems damn childish to  me,
particularly since  TI  wouldn't  even  supply  me  with  9900's  and  the  LISP
development was  to be  portable, not  68000-based, anyway.   We compromised  on
trying to lease a LISP-machine from LMI.

With growing frustration, since it is now February-March of 1980, I suggest that
if TI's is serious about TLC then they  should be willing to write a "letter  of
intent to do business", because large amounts  of money are at stake; I have  to
line up  personnel, space,  and equipment.   By mid-March,  no letter  has  been
generated: Stratton's signature is required; but, press on ...

I'm pissed  off totally  by now,  and  conversations with  Mark and  Truman  are
sympathetic, but  not encouraging.  So,  Gene, when  I  asked you  in  mid-April
"what's going on with TI's  LISP effort" and you said  "six months to two  years
before even issuing a request for proposal" (implying --among other things--  to
me open bidding and that my efforts counted for nothing), and later said "TI had
paid for my  services both times  I'd come  to Dallas" (implying  that I'd  done
nothing else), you should  understand why I raised  my voice and suggested  that
TI's conduct in this matter had been "discourteous and unprofessional!!"

I do hope that this letter will accomplish some good; I hope you will reflect on
the sorry business of  TI-LISP. I hope  that you will  resist the temptation  to
just "close  ranks",  thinking "gee,  we're  sure lucky  we  didn't have  to  do
business with him; he's irrational!"


						Yours sincerely,




						John R. Allen
						18215 Bayview Dr.
						Los Gatos Ca, 95030
						(408) 353-3857

CC: Dr. Mark Miller
    Dr. Truman Blocker