[R6RS] end game---getting really close

dyb at cs.indiana.edu dyb at cs.indiana.edu
Fri Dec 15 15:59:29 EST 2006

We have just a few outstanding responses to agree to:

7   (Matthew, Anton)
10  (Mike)
28  (Matthew, Anton)
46  (Will)
51  (Will)
54  (Anton)
59  (Matthew, Anton)
73  (see below)
75  (Will, Anton)
80  (Matthew, Mike, Anton)
107 (Anton)
122 (Matthew)
129 (Matthew, Mike, Anton)

For 73, I proposed the following:

      The next draft report will suggest a convention (not requirement)
      that users construct UIDs either from library and record-type names
      or from names constructed using the UUID namespace.

Mike has suggested that we also include something to the effect of:

      The next draft will also make the effective UID a combination of
      the supplied identifier (or symbol in the case of
      make-record-type-descriptor) and the record name.

I would like to hear from the other editors, one of:

  1. keep the status quo: the UID is the identifier or symbol supplied
     by the programmer

  2. go along with the proposed change: the UID is a combination of the
     supplied identifier or symbol and the supplied record name

  3. be noncommittal and say that we will decide this issue before the
     next draft is released

I'm putting Mike down for 2; my vote is 1, but I think we're both okay
with 3 as well.


More information about the R6RS mailing list