[R6RS] my notes on today's conference call (10 May 2007)

[William D Clinger]

Telephone conference call May 10 2007 8:00am-10:00am
Will, Matthew, Mike, Anton, Kent present by 8:02am

0. finalize agenda (1 minute)

1. action items (1 minute)
   - update r6rs to reflect additional decisions:
     - probably use "should" for programmer requirements that implementations
       are not required to enforce
     - change the specifications of char-alphabetic?, char-whitespace?, and
       string-titlecase to conform with Unicode 5?
       (http://lists.r6rs.org/pipermail/r6rs-discuss/2007-April/002250.html)
     - adopt Posix semantics for set-port-position! when underlying
       object is a file or bytevector output port, with unwritten bytes
       having unspecified values
     - move formal semantics to a nonnormative appendix
     - help Mike fix "uniquely determined" wording for syntactic datums

2. add port-length and port-has-port-length? procedures?
   - and corresponding arguments to custom port makers
        include them?
        no: unanimously
    should reconsider (file-options) defaults by email

3. allow port positions to be "magic cookies"?
    yes: Will, Kent, Matthew, Mike, Anton

4. combine get-string and get-string-n?
    yes: Will, Kent, Matthew, Mike, Anton

5. final ticket 210 resolution
   - http://lists.r6rs.org/pipermail/r6rs-discuss/2007-April/002256.html
    that proposal is a little too vague;
    we should all read it after it has been revised a bit,
    and communicate by email

6. library names
    Mike moved we replace (r6rs ...) by (rnrs ... 6); Anton seconded.
    yes: Matthew, Mike, Anton
    no: 
    abstain: Will, Kent

7. eqv? coverage
    doesn't have to be specified on transcoders
    but should be (partially) specified on records

8. bitwise-bit-count specification for negative inputs
    shall we accept this change?
    yes: Will, Kent, Matthew, Mike
    abstain: Anton

9. direction for Robby and Jacob on the semantics
    sense of the committee is that we would like for them
    to focus on the use of the semantics for specifying
    meaning as opposed to its use as a tool for stepping
    through example programs;
    for example, we would appreciate an accurate semantics
    for the library top level and other problematic issues,
    but we do not need them a formal semantics for most
    other features that are removed by macro expansion;
    the complicating optimizations could be split off into
    a separate paper that could be cited by the R6RS

10. non-normative appendix document
    how library source code is related to the filesystem (ticket 83)

11. reference implementations

12. (file-options) defaults
    these apply to output files only
    all three standard defaults should be reversed
    Kent moved we use the following names for non-defaults
    (Mike seconded):
        no-create
        no-fail
        no-truncate
    yes: Kent, Matthew, Mike, Anton
    abstain: Will

13. procedure going forward
    Mike is done with most tickets; expects to finish this weekend
    proofreading next week
    release 5.93 to public on 22 May
    more proofreading by public
    public releases every two weeks or so
    Kent will post a note to the R6RS discussion list

14. adjourned at 9:49am

Will