problem is that we have no current method to gather this information.
talkback data is gathered for only those that opt in to the program;
but more importantly it gathers reports for only thoese that have crashed.
we never see reports from users that are crash free for a given period,
so talkback present a bit more negetive view of the number.
also not everyone downloads and enters the test on the same day.
each day we receive reports from folks that might have been running
builds for several months, or who have just downloaded that day.
trying to control all the variables in a MTBF experiement is a difficult
task.
chris h.
Sebastian Späth wrote:
[EMAIL PROTECTED]">Ben Bucksch schrieb:I'm not sure that overall MTBF is all that useful. It might be better
to say that x% of users should have at least n hours MTBF.As for concrete values, I'd suggest:
90% users: MTBF > 250 h
99% users: MTBF > 50 h
MTBF average: 500 h
Complicate to test. Just make it an accetable MTBF and standard
deviation value. By this you capture that as well.
Sebastian
