More on CLIM, ILA, etc.

[Scott_Fahlman@SEF1.SLISP.CS.CMU.EDU]

FYI...

------- Forwarded Message

Date: Wed, 26 Aug 92 11:38:36 PDT
From: Bill York <york%oakland-hills@lucid.com>
Message-Id: <9208261838.AA06578@oakland-hills.lucid>
To: Scott_Fahlman@SEF-PMAX.SLISP.CS.CMU.EDU
Cc: rpg@oakland-hills, York@oakland-hills
In-Reply-To: Richard P. Gabriel's message of Tue, 25 Aug 92 10:51:28 PDT <9208251751.AA00998@inferno.lucid>
Subject: [Scott_Fahlman@SEF-PMAX.SLISP.CS.CMU.EDU: CLIM, ILA, etc.]
Reply-To: York@lucid.com

Dick forwarded me your message, so I thought that I'd take a crack at
filling you in on the CLIM/ILA status.  Let me point out that this
message contains my personal opinions as someone involved in the CLIM
effort from day one, and not the official policy of either Lucid or
ILA.

   Return-Path: <Scott_Fahlman@SEF-PMAX.SLISP.CS.CMU.EDU>
   To: rpg@lucid.com
   Subject: CLIM, ILA, etc.
   Date: Tue, 25 Aug 92 08:51:59 -0400
   From: Scott_Fahlman@SEF-PMAX.SLISP.CS.CMU.EDU

   For some time, a subset of current and potential users of CMU Common Lisp
   have been asking about CLIM support.  Our answer had been that we had no
   plans for this because it was a proprietary system, too large and hairy for
   us to implement from scratch.  This seemed unfortunate for all concerned,
   as it threatened a split in the Lisp community between those using
   commercial implementations with CLIM and those using CMU CL who would
   settle on some other package -- it's not clear which of the other
   alternatives would win.

(Let me point out that I don't think that having multiple Lisp UI
packages is a bad thing.  The main point of CLIM was to have at least
ONE package that was endorsed by a large number of Lisp vendors and
was available in a wide variety of Lisp environments so that
programmers could build portable applications.)

   At some conference or other Rob MacLachlan met Mark Son-Bell of ILA, and we
   were surprised to learn that ILA might be willing to release their CLIM 2.0
   code to us for free inclusion in CMU CL.  He said that some of the other
   CLIM vendors knew of this plan and thought it was a good idea -- he may
   have said all, I'm not sure.  After discussing various CMU/ILA licensing
   plans, he just decided to just make their existing code available to
   everyone via anonymous FTP, copyrighted by with some sort of blanket
   permission to use (including commercial use).  We did get a copy of ILA's
   CLIM 2.0 code at that point, but with the old copyright notices on it.  The
   general release with blanket permission was delayed, and suddenly mail to
   ILA started bouncing.  They seem to have disappeared.  So we're not sure
   about the legal status of the ILA code we have.

Basically, ILA is out of the CLIM business.  Mark is continuing his
attempts to make something out of the natural language tools side of
the company, but that is pretty much the only ongoing activity.  None
of CLIM developers remain at ILA, but some of us are still involved in
CLIM activities.  For example, Dennis Doughty (another former ILA
principal) and myself have formed a separate venture to continue
development and distribution of the CLIM version for Macintosh Common
Lisp.  We have licensed the MCL CLIM software from ILA and are
currently working towards a CLIM 1.1 product release for the
finally-final version of MCL 2.0.  We are also currently negotiating a
distribution and support agreement for MCL CLIM with Lucid.  In
addition, I have some CLIM responsibilities in my "day job" here at
Lucid.

So, what is all this to you?  ILA still retains some CLIM assets,
including the CLIM 2 prototype code that Mark sent you a while ago and
the ownership of the CLIM trademark.  ILA still supports the idea the
CLIM should be made available to the widest possible set of Lisp users
in the hope that this will foster Lisp application development.
Toward that end I think that I can assure you access to the ILA CLIM 2
prototype sources under satisfactory copyright, licensing and other
restrictions.  I had been left with the impression that you guys would
be evaluating the software that we sent you a while back and that if
it passed initial evaluation we would work out the copyright/licensing
issues.  If you still want to move forward on this, I can put you back
in touch with Mark to close the loop.

   We are also aware of the stormy relations that had developed between ILA
   and various allies, on the one hand, and Symbolics and Franz on the other.
   One story, I think from Scott McKay at Symbolics, had it that ILA was
   willing to release their CLIM 2.0 implementation because they were going to
   abandon it anyway.  The claim was that the Symbolics/Franz implementation
   was much better, and that ILA was going to start building on that code
   instead of their own.  A project by ILA to port this code to Macintosh CL
   was mentioned.

I will leave aside any discussions of the technical merits of the two
bodies of software.  The overriding issue was that ILA did not have
the resources to continue to hold the coalition together, nor did it
have strong allies who were committed to pushing ahead with a
development program based on the original body of software.  Thus, the
only real CLIM 2 product effort left was the one being conducted by
Symbolics and Franz.  One of the last significant actions ILA took
regarding CLIM was to broaden participation in a Franz/Symbolics CLIM
2 code-sharing agreement to include Lucid and ILA.  Lucid has decided
to base its CLIM 2 efforts on this prototype, as have we in the MCL
CLIM camp.  I believe that Franz and Symbolics plan to offer this
software to other parties for a fee; you should perhaps contact Bob
Laddaga at Symbolics (Laddaga@Symbolics.COM) if you are interested in
pursuing this.

   Now I see that Lucid is going to do CLIM 2.0 for Macintosh.  I have also
   seen a couple of posts signed by Bill York of Lucid, though I don't know if
   this is the same Bill York who used to be at ILA.  It's unclear to me
   whether Lucid acquired ILA, or just won the Apple contract away from them,
   or whether ILA self-destructed and Lucid picked up some of the pieces.  Can
   you enlighten me?

Yes, I am the same Bill York.  I was finally forced in June to go out
and find a real job; I had far exceeded my ability to continue to
invest in ILA/CLIM activities.  As I mentioned above, I negotiated the
rights to MCL CLIM as part of my parting with ILA and will continue to
work on that product with Doughty under the name "Illudium Software".
As Lucid has become or is about to become a VAR for Apple's Macintosh
Common Lisp product itself, it seemed natural to negotiate a
distribution and support agreement with Lucid (although life gets
complicated for me sometimes if I forget which hat I'm wearing), and
that is in progress.

   Is there now some agreement on CLIM 2.0, or can we look forward to a lot of
   turbulence, with differning versions from Lucid and Franz?  If there is
   some single CLIM standard, we would still be interested in finding a free
   or not-too-expensive way of making this available to our users.  Of coruse,
   it's up to the vendors who control the code to determine whether to use CMU
   CL as a unifying force in this way, or whethr it is more important to
   preserve CLIM as something that the commercial Lisp have and the free Lisps
   don't.

Things are certainly not copacetic.  There are still some unresolved
spec issues and no real mechanism in place to resolve them.  The
various parties are now also disagreeing about when the code sharing
agreements are/were in effect and what code was covered.  Still, I
think that you will see all the current parties (Lucid, Symbolics,
Franz and Illudium) release CLIM 2 products for their various
platforms over the next two quarters, all based on the same basic
shared software "backbone", which I certainly view as a kind of
success, even if ILA isn't there to see it.

As for CLIM and CMU CL, I think that you have two basic options.  The
first is to attempt to license the commercial CLIM 2 software from
whomever claims to have the rights to grant the license.  I doubt that
they will give it away, but you should check.  The second is to base
CMU CL CLIM on the ILA software prototype.  This software is N%
compatible with the actual API of the product version for reasonably
high N.  The number has been decreasing somewhat of late, as changes
are made to the spec and the product software, but the basic
capabilities remain pretty similar, and a modest development team
should be able to track the changes or mount a one-time catch up
effort once the final version of CLIM 2 is released by the vendors.

   Cheers,
   Scott

I hope that this clears things up for you.  Sorry for not keeping you
informed as things progressed, but it's been a bit hectic.  Feel free
to contact me for more information or further CLIM discussions.  Buy
me a beer sometime and I'll tell you the whole story if you like!

------- End of Forwarded Message