[R6RS] my notes on yesterday's conference call (13 December 2006)

William D Clinger will at ccs.neu.edu
Thu Dec 14 14:20:04 EST 2006


Conference call December 13 2006 8:00am-9:00am EDT
Present by 8:04am:
Kent, Matthew, Anton, Will;
Mike was not present because of a time conflict,
but had expressed some of his views electonically

0. finalize agenda (5 minutes)

1. action items (1 minute)
   - record "vague response" for ticket 87 re: attempting to address
     concerns of the comment (Kent)
        not done
   - record "defer" handling until 1/15 pending handling of formal
     comment #87 responses to tickets 26, 36, 41, 42, 48 (Kent)
        not done
   - rethink "syntax-rules issues" (ticket 111, 114) responses (Kent)
        done
   - remove mention of "an error" wording from ticket 61, 62 responses (Kent)
        done
   - draft response for ticket 88 (Kent)
        done
   - revise ticket 130 response to maintain prohibition of multiple
     versions of one library (Matthew)
        not done
   - flush hash-table-values and change hash-table->vector (possibily
     renamed hash-table-entries) to return two values: a vector of the keys
     and a vector of the corresponding values (Anton)
        done
   - follow up with steering committee re: front page (Kent)
        can wait

NOTE: tickets whose responses are not yet accepted and by whom:

   6:	Will			       83:  Anton
  -7:	All			       84:  Anton
   8:	Anton			       85:  Anton
   10:	Mike			      -86:  All
   12:	Anton			      +87:  All
  -18:	All			       88:  Will, Matthew, Mike, Anton
   23:	Anton			       89:  Will, Anton
   25:	Will, Anton		      -90:  Will, Kent, Matthew, Anton
  +26:	All			       91:  Anton
  &35:	All			       92:  Anton
  +36:	All			      &96:  All
   39:	Will			       97:  Will, Anton
   40:	Anton			       98:  Matthew, Anton
  +41:	All			      -99:  All
  +42:	All			       106: Will
   45:	Will, Matthew, Mike, Anton     107: Anton
  -46:	All			       108: Will, Matthew, Mike, Anton
  +48:	All			      -109: All
   51:	Will, Matthew		      -111: All
   52:	Matthew, Anton		       113: Will, Matthew
   54:	Anton			      -114: All
  -59:	All			      -115: All
   61:	Will, Matthew, Anton	     &+117: Will, Kent, Matthew, Anton
   62:	Will, Matthew, Mike, Anton    -118: All
   65:	Anton			       120: Matthew, Anton
   66:	Anton			       121: Matthew
   67:	Anton			      -122: All
  -68:	All			       124: All
  -69:	All			       125: Matthew
   70:	Anton			       127: Will
   71:	Anton			       129: All
   72:	Anton			      +130: Kent, Matthew, Mike
   73:	Will, Anton		       131: Will
   75:	Will, Anton		      -132: All
  -76:	All			      *135: All
  -78:	All			      *136: Will, Matthew
   79:	Will, Matthew, Mike, Anton    *137: All
  -80:	All			      *138: All
   81:	Anton			      *139: All
   82:	Anton

   - no decision yet

   + response does not yet reflect technical decision made:
     26, 36, 41, 42, 48: should reflect decision to defer until 1/15
     87: should be softened to "attempt to address concerns"
     117: should reflect decision to reject allow bodies more widely
     130: needs to say multiple versions of a library remain prohibited

   & questions/comments:
     35: can we list as withdrawn?
     96: response may need to be updated to reflect response to 78
     117: response should either not mention declarations or
          mention that they have been flushed

   * came in after 11/15; no response necessary

2. rationalize various iteration procedures
   - ticket 78
        vote on current draft w/clarification: 5-0
        (Mike had voted electronically)

3. syntax-rules issues
   - tickets 111: syntax-rules should evaluate to a transformer
   - tickets 114: identifier macros with syntax-rules
   - summary of proposed response:
     - make syntax-rules evaluate to a transformer
     - allow _ (unparenthesized) pattern for syntax-rules
     - add make-variable-transformer to base library
     - add identifier-syntax to base library
     - base library exports only three things for expand: syntax-rules,
       make-variable-transformer, and identifier-syntax
   - alternative response:
     - make syntax-rules evaluate to a transformer
     - add identifier-syntax to base library
     - base library exports only two things for expand: syntax-rules
       and identifier-syntax
    vote on alternative response: 4-0

4. add negative meta levels?
   - ticket 109
    vote on adding them: 4-0

5. add "lazy" procedure?  retain (r6rs promises) library?
   - ticket 7
    vote to remove the library (and let someone submit a SRFI),
    leaving delay and force (but not lazy) in the R5RS compatibility
    library: 4-0
    (Mike had expressed an opposing view differently, but was not
    present for this particular motion)
    Anton will draft response

6. add file-exists? and delete-file
   - ticket 124
    vote to add: 4-0

7. add "parent-descriptors" clause?
   - ticket 90
    Kent moved that the response explain why the limitation exists
    instead of saying we will remove it; Anton seconded
    vote: 2-0-2 (Will and Matthew abstained; Mike presumably would
        have voted no)

8. provide UTF-16 codec?
   - ticket 68
    vote to accept comment's proposal
        (add UTF-16, drop UTF-16BE and UTF-16LE): 4-0
    (Mike presumably would have voted no)

9. add lib syntax?
   - ticket 86
    Will moved that we drop the single-identifier shorthand; Anton seconded
    vote: 4-0

10. rename bytes to ???
   - ticket 99
   - sentiment seems to prefer non-plural, non-hyphenated name
    Will moved to rename to a non-plural, non-hyphenated name
        to be decided upon later; Anton seconded
    vote: 4-0

11. add "...-vector->bytes" and "...-bytes->vector" procedures?
   - ticket 115
    The editors agree to study this issue.
    (There are several different ways to interpret the comment,
    all with their own advantages and disadvantages.)
    Anton will draft a response of that sort.

12. allow set!, etc., to return unspecified number of unspecified values
   - ticket 118
    Kent asked whether (unspecified) could return an unspecified
    number of unspecified values.
    Will moved that returning zero values to a one-value-expecting
    continuation automatically coerce the zero values to a single
    unspecified value, and set!, vector-set!, write, etc be allowed
    to return zero values; Anton seconded.
    vote: 1-3 (Will voted yes)
    no resolution for ticket 118
    Matthew moved that we not change the report, but continue to
    discuss the possibility of changing it; Kent seconded.
    vote: 4-0

Matthew had to leave soon, so two agenda items were moved forward.

17. allow additional line terminators?
   - ticket 46
    requires more reading of the Unicode specification
    probably could handle by email (action item: Matthew)

18. allow #<n>(---) vector syntax?
   - ticket 59
    Matthew moved to forbid the #4(0 1 2 3) notation (etc);
    Will seconded
    vote: 4-0 (electronically, Mike would have voted yes)

Matthew had to leave, so only three editors were left.

13. add ->single and ->double
   - ticket 122
   - perhaps as number->single, number->double
     (then should ->exact become number->exact and ->inexact become
      number->inexact?)
    Will moved we flush real->single and real->double (since they
    can be defined in terms of bytes objects); Kent seconded.
    vote: 3-0

Idea for subsequent discussion: rename ->exact and ->inexact to
number->exact and number->inexact

14. allow (nongenerative #t)
   - ticket 129
   - or (nongenerative)
     (then should (sealed #t) change to (sealed) and (opaque #t) change
      to (opaque)?)
    Will doesn't understand the draft response.
    Let's discuss this one by email.
    Straw poll on formal comment's general proposal: 3-0

15. allow (define-record-type point (fields x y)) abbreviation
   - ticket 132
    vote to accept formal comment's proposal: 3-0

16. allow string exit codes
   - ticket 18
   - how is this affected by script -> program change?
    We should punt on exit codes; move them to non-binding appendix,
    or get rid of them altogether.

19. change "find" to raise exception when returned element would have been #f?
   - ticket 69
    Will moved that our response be limited to the first two sentences
    of the current draft response; Anton seconded.
    vote: 3-0 (Mike might have voted no)

20. clean up make-transcoder to accept three symbol arguments?
   - ticket 76
    We need a draft response for this.
    Mike asked Anton to do it, but Mike should probably draft it.

21. fix i/o interface inconsistencies?
   - ticket 80
    We need a draft response for this, and it depends on tickets
    #88 and #89.
    Mike asked Anton to do it.
    Let's discuss this on the mailing list.

22. adjourned around 9:58am



More information about the R6RS mailing list