On 11/29/05, sascha brossmann <brsma.s@gmail.com> wrote:
given the *extremely* irregular frequency of this failure i can well
imagine that it went by undiscovered until too late. and while i am
somewhat happy that it doesn't happen very often, i suspect that
exactly this turns it into a harder problem, as well. :-(
I concur regarding the extreme irregularity! On MY system
pbbuttonsd seems to either make it liklier or trigger it somehow. So
for a couple of days I set up various test situations with pbbuttonsd
activated. Within an hour I had a sudden shutdown, then recreated (as
best as I could recall) exactly the same arrangement of browser and
terminal activities again, and the system did NOT power down the rest
of the day. Literally hours of use without a failure. I can't very well
test it if I can't be sure when it's better!
At this moment I do not have pbbuttonsd running. So far (!!!) I have
only had shutdowns while that daemon is running. I also have only had
shutdowns when actively using a web browser. I almost always have
Firefox going, but I often have Opera running too or instead. Probably
80% of the shutdowns have occurred while I have been composing a
message in gmail. The shutdowns have never occurred when I was away
from the keyboard. I have done some fairly large file transfers without
triggering the shutdown, but like you I suspect some sort of network
interaction....
DRAT this is hard to test.