Am 09.03.2011 18:38, schrieb Pascal Stumpf (via RT):
The sleep.t test lets perl6 sleep for three seconds, but allows for a range of 2-10 seconds when checking the 'real' sleep time.
The spectest suite has a long history of being run by slow compilers/interpreters on slow machines, which is why that test is very liberal. I guess we can experiment with lower tolerances now that the actively developed compilers have better known speed characteristics.
This is way too tolerant and would have almost hidden a nasty bug in parrot's sleep function on OpenBSD when parrot is threaded.
And there's no parrot test catching that? Maybe that would be the first step...
Cheers, Moritz
