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

what should RnRS be all about?

As we have been thinking about the agenda for the upcoming R5RS
meeting on June 25, we have come to believe that we need to devote at
least some of the time at the meeting to discussing where we (the RnRS
authors) are going, on a much higher level than (for example) debating
whether open-input-file should return #f or invoke an alternate
continuation when the file is not found.

In some ways, the RnRS process is the victim of its own success.  The
original goal, as we recall it, was to try to bring some unity to all
the Scheme variants that were springing up at Indiana, MIT, Yale, and
elsewhere.  The theory was that this would be good for the Scheme
community and the resulting consensus would also help the whole Scheme
universe expand.  Now we have an IEEE Scheme standard and a growing
body of Scheme users, and we dare say the original goals have mostly
been accomplished.

This leads to the obvious question:

  * Where is the RnRS process going?  Where should it go?  Which of
    the following are the right models (and what other ones should be
    added to the list)?

      - Continuing to polish and regularize the current Scheme
        definition, adding useful procedures and data types, cleaning
        up interfaces, improving the organization of the Scheme
        report, etc.

      - Debating and working out ways to add major new blocks of
        functionality to Scheme, such as macros, exception/condition
        handling, modules, object-oriented programming constructs,
        "industrial-strength" I/O, parallelism, etc.

      - Watching the community of Scheme implementors for new,
        valuable, and tested ideas that might merit "promotion" to
        RnRS status, so that they would eventually become available in
        "all" Schemes.

      - Considering whether and how Scheme should change to have the
        maximum impact on real-world programming.

      - Serving as a forum for discussion of ongoing research projects
        in the Scheme community, and as a medium for

      - Declaring RnRS a finished project and disbanding the authors

A second important organizational issue is

  * Have the ground rules changed since Brandeis?  Should they?

We would like to see some e-mail discussion of these organizational
issues before the upcoming R5RS meeting, as well as some discussion at
the meeting itself.
						-Bert Halstead
						-Norman Adams