Issue 11277 was closed upon the generally accepted conclusion that the failure is that mmap(2) on Mac OS X has a long time known bug (see msg129107, or directly http://lists.apple.com/archives/darwin-development/2003/Jun/msg00141.html).

Issue 11277 fixed the problem in the test suite, but it does give end-users no hint that such a problem exists (on the documentation side).  And it's my believe that a beginner or even a non-low-level programmer may be lost in an equal situation.  (Though one could argue that such a HE should not use mmap(2), but everyone has a first-time somewhen, say.)

In addition, there are many people who don't have the chance to work/test on multiple operating systems and/or hardware, and who need the hand of people who are experienced and can give hints on stupidiness or however this word is correctly spelled.  (Am i preaching already?)

Because Python is expected to be a user-friendly language i do request a documentation patch which gives a hint.  I'll apply a patch, but i'm somewhat new to Python and rST etc., so it surely needs a hand.  (Maybe even its speech is too aggressive??)