[sugar] Re: [OLPC-devel] pygtk performance issue

[Ivan Krstić]

Johan Dahlin wrote:
> I increased that timeout to 1000 today, so it should be less than an issue
> from now on.

Not for us. When the machine is being used passively, i.e. for
low-screen-activity purposes such as reading, we suspend the hardware
and just power the display. Coming out of that mode unnecessarily, even
if just once every second, is to be avoided if at all possible.

I've seen some quick-fix patches (that don't add the API required to
solve this somewhat more gracefully) floating around. Gustavo had one on
python-dev, and there's one on the Red Hat bugzilla. Python upstream
won't integrate them due to POSIX compliance concerns, but I have far
less scruples about straying from POSIX in exotic situations if it means
I can resolve this problem. Have you looked at the patches, and can you
comment on them?

-- 
Ivan Krsti? <krstic at solarsail.hcs.harvard.edu> | GPG: 0x147C722D