Gregory P. Smith added the comment: I saw a small regression over 4k when using a 64k buffer on one of my machines (dual core amd64 linux). With 32k everything (amd64 linux, armv7l 32-bit linux, 64-bit os x 10.6) showed a dramatic improvement on the microbenchmark. approaching 50% less cpu use in many cases.
i doubt applications will notice as much as they're likely to be dominated by their own application code rather than the subprocess internals. re: 3.3 or not, true, but since it doesn't change any APIs and is minor I did it anyways. If you think it doesn't belong there, leave it to the release manager to back out. This and the #19506 change should be invisible to users. ---------- resolution: -> fixed stage: needs patch -> commit review status: open -> closed _______________________________________ Python tracker <rep...@bugs.python.org> <http://bugs.python.org/issue19929> _______________________________________ _______________________________________________ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
