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

Re: Problems with SRFI document

Hello Michael,

Just a couple of comments on the fix:

1) The stream-low-level-lazy, ...strict, ...delay and ...force
   implementations have a pretty serious bug.  This bug was discovered by
   Alexandro Forero Cuervo and discussed on the SRFI-45
   list (with some spillover to the SRFI-40 list).  I posted a
   fix - please see the patched SRFI-45 document for the details.
   (also, AFC posted a patched version of SRFI-40 incorporating
   a version of the fix to this list).

2) I believe the discussion of stream-unfoldn in the first
   part of the implementation section was obsoleted by the
   SRFI-45-style promises (indeed, the actual implementation
   below it was then changed back to using the naive
   version, which should be correct and safe for space).


On Sat, 11 Sep 2004, Michael Sperber wrote:

"Gordon" == Gordon Weakliem <gweakliem@xxxxxxxxxxx> writes:

Gordon> This is a really nice SRFI, thanks to all who worked on it.
Gordon> I've found a couple of issues with the SRFI document:

Gordon> 1) There's an error in the definition of cutoff1 under the
Gordon>    rationale section, it refers to "cutoff" instead of
Gordon>    recursively calling cutoff1

Gordon> 2) The implementation section is empty.  I realize that the
Gordon>    reference implementation is available in the
Gordon>    pre-finalization archive, but it looks like it was omitted
Gordon>    from the SRFI document.

Good points, both.  Fixed.  Thanks for the note!

Cheers =8-} Mike
Friede, Völkerverständigung und überhaupt blabla