[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: CFP: Scheme Implementors Workshop



First, I should say that I think this whole thing is being blown way
out of proportion.  The CFP says:

+ The purpose of this workshop is to bring together the major
+ implementors of Scheme in order to review the state of various
+ Scheme implementations and to assess alternatives for the continued
+ development of Scheme. 

It doesn't say "we're going to form a secret cabal who will henceforth
control all future development of Scheme in order to further our
nefarious ends and who will bend all users to our iron will, crushing
all dissenters beneath the heels of our boots."  Which seems to be the
way that people are taking it.  If some people can't get together to
talk about Scheme without others raising a hue and a cry, something's
really wrong.


On Fri, 1 Mar 1996 20:02:48 -0500, Alan Bawden <Alan@lcs.mit.edu> said:

> It is my belief that the best thing for the future of Scheme would
> be to freeze the definition of "Scheme" at the contents of R4RS (or
> perhaps R5RS if that document can be produced without any
> controversy).

Makes sense.

> Further work on the evolution of Scheme-family languages should then
> take place under the auspices of some more traditional standards
> organization, such as ANSI or IEEE or ISO.

Why?  Scheme's a very pretty, coherent language.  It seems plausible
that further pretty, coherent Scheme-family languages could arise by
people having good ideas, going off and implementing them, making them
available to others, and having those others say "hey, that's a good
idea, let's use it."  (Dylan is a good example of the kind of
mechanism that I'm talking about.  I don't particularly like some of
the choices that were made, but at least it hangs together nicely, and
I wouldn't complain if it were to catch on.)  That's evolution in
action, it's a natural way of growth, and it's the way that many
programming languages have arisen.  I'd be so bold as to claim that
it's the way that almost all programming languages have arisen, and
that the exceptions aren't anything to be particularly proud of.  (Are
Fortran 90 and Common Lisp really our models?)

But you're saying that we should shut out all sorts of organic
development, that we should restrict ourselves to the confines of
standards organizations.  Maybe that approach is one that should be
followed; it's not the only one that should be followed, and even if
you disagree with me on that, it's certainly not the only one that
will be followed.  Trying to prevent people from developing new
dialects without going through the standards organization simply won't
work, and complaining too much about it will only result in their
being less open about their new dialects while in development, not
more so.

> Those Scheme-family languages would necessarily have names like
> "ANSI Scheme" or "IEEE Scheme" to distinguish them from the classic
> "Scheme" defined in the "Final Report".

I don't think that that would do a good job of distinguishing
languages.  How about names like "T" or "Dylan" or "Guile" or
"EuLisp"?

> 2.  Clarify that this workshop has no actual power over the future of
>     Scheme -- you just want to get a bunch of implementors together
>     to talk.

There's not some Scheme God out there doling out power to certain
approved groups.  If the workshop ends up influencing Scheme
implementations that people use, it will have power.  If it doesn't,
it won't.  It's not something that anybody can declare by fiat.

david carlton
carlton@math.mit.edu