If there are people around who don't know much about coding but would
like to help lisp, here's an idea (having ~200 lispers on the list
might make some of this simpler, too):

The ANSI draft specification dpANS2 and it's unofficial revision dpANS3
have essentially been the bible for free Lisp implementations since
Common Lisp became an ANSI spec - they contain virtually all the
material present in the actual ANSI spec and are freely available
(although the tex style is apparently rather old - not surprising given
it was written in the early 1990s.)  It is available from here:
ftp://parcftp.xerox.com/pub/cl/

(IIRC it's been on that particular server since it was written because
that server has some quasi-official status related to the review
process for the draft - I think it's in the readme files somewhere. 
There is nothing prohibiting distributing it in the text, IIRC, but you
do have to be careful not to claim these drafts are any kind of
official ANSI anything if you do.  It's in there somewhere.)

The problem with these draft specs, and the primary reason more hasn't
been done with them, is because their legal status is - um - murky. 
Basically the upshot, based on what I have been able to find out, is
that there was a tangle of contracts, contributors, and source material
that merged to become the ANSI spec and a lot of people and companies
were involved.  Permission was given to use the material for creating
the spec, but of course the public draft spec is a different matter. 
As a practical matter I doubt if anyone is worried about it any more
(after all it's all been freely downloadable for 10 years) but because
there is no explicit permission from whoever the copyright holders may
be to use it without restriction its put a damper on doing anything
creative based on these drafts.

There has been some excellent work done on things like the ANSI random
tester and test collections, and a few flaws have been discovered in
the original ANSI Common Lisp spec.  Unfortunately, the J13 committee
no longer has enough members to undertake to update the spec any more,
based on the latest (informal) report available from here: 
http://www.franz.com/~smh/J13/

As Lisp continues to progress, I think the hope would be that minor
spec flaws could be fixed, and more parts added on as they reach a
maturity level which warrants standardization.  Of course, another ANSI
effort is probably beyond the available resources, but the free lisp
community itself might create its own spec, with no "official"
standing, as a guide and aid for the future.  The obvious perfect
starting point for such an effort is dpANS3, but while its status is
murky any efforts to build on it are going to be tentative at best.

Of course, any Lisp definition which emerges from this cannot be terms
ANSI anything, and probably shouldn't be Common Lisp in order to avoid
confusion.  I would recommend calling it Community Lisp - similar
enough for people to expect (correctly) its close historical relation
to Common Lisp, but clearly a distinct name.  Also, since the Lisp
community itself would drive the innovations which could be documented
in the spec, I think Community Lisp would be a very fitting name for
the result. But even if this vision appeals, there is still the problem
of the status of the drafts.

A complete list of contributors to the draft ANSI spec probably isn't
available anywhere, but I think most of the relevant parties are listed
in the edit history included with the drafts.  They might be able to
indicate other interested parties.  I would like to propose a community
effort to locate the contributors to the ANSI draft specification, and
ask formal permission to place the work under the Modified BSD license.
 We obviously can't go the route of the Hyperspec since ANSI has no
incentive to release the official version publicly even if they could,
but the draft will serve - dpANS3 is by all accounts I have heard very
very close to the ANSI final in content.

If we are able to do this, Lisp would become possibly the most liberal
major language in existance - not only would implementations of the
spec be freely available, but the spec itself would be freely
available.  Indeed, the original motivation for looking at the dpANS3
document was to see if perhaps Literate Programming could weave code
and text together to form something completely new - a specification
which is in itself runnable!

Anyway, does the idea of a grass roots effort to contact the copyright
holders of the dpANS content sound like a worthwhile project?  It would
need to be coordinated and organized, of course, and people would have
to be very polite and patient when contacting people.  Any thoughts on
how to organize something like that?

Cheers,
CY

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 
_______________________________________________
Gardeners mailing list
[email protected]
http://www.lispniks.com/mailman/listinfo/gardeners

Reply via email to