Mon Jan 26 11:46:52 EST 2004
Michael Sperber wrote:
> You asked for a list of issues. However, your list contains issues
> intermingled with statements about *how* you intend to address them.
> While this gives us a useful preview of where you intend to go, it's
> likely to make the discussion in this thread quite messy: On a lot of
> issues, I agree that they should be addressed, but disagree on how
> they should be addressed. (Just like Kent.)
> Can we agree then on discussing the issues only for now? Maybe we
> should make that more precise by stating the issues in the form of
> problems that should be solved.
Well, I think it is sometimes desirable to express issues with a hint
of a solution to give an indication of where the proposer is going (I
may have gone beyond a hint in some cases, it is easy to get carried
away when you have thought about these issues for a long time...).
Note that these "solutions" are not formal proposals at this point.
It is usually difficult to separate issues from solutions (for
example, is the adoption of minus zero and infinities an issue or a
In addition to the list of issues, and arguably of greater importance,
we should elaborate our main goals. In my case it would be
1) Advance the Scheme standard and the major Scheme implementations
to make it possible (in practice) to write modern software that is
portable to the major implementations of Scheme. A module system
would be an obvious advancement, but also "lower-level" stuff like
binary I/O, Unicode characters and strings, and filesystem
2) Make the language cleaner. For example: semantics of internal
definitions closer to toplevel definitions, semantics of
dynamic-wind, semantics of file names, transcript-on /
transcript-off, lexical syntax, continuations, case, etc.
More information about the R6RS