Neil Hodgson wrote:
April White:
Having a program like autoit bang on the keyboard does not seem much
better, though it can do it faster but I don't really think that'll
prove anything.
The only thing testing ever proves is that software has bugs.
Testing thread correctness is really hard as failures are often
random. If you can detect a failure, you may be able to provoke a bug
by running many repetitions with either naturally occurring or
programmed timing changes.
AND, to find even more [threading-related] bugs, do the testing on a
multi-CPU box (dual-proc, dual-core, and maybe even HT) - since on a
single CPU box, there is no actual simultaneous execution of multiple
threads... which means some of these bugs can still slip through, only
to be found by those users with the shiny new dual-core they got over
the holidays. ;)
Robert Roessler
[EMAIL PROTECTED]
http://www.rftp.com
_______________________________________________
Scite-interest mailing list
Scite-interest@lyra.org
http://mailman.lyra.org/mailman/listinfo/scite-interest
