On Wed Nov 19 02:34:26 2008, bernhard wrote: > On Di. 18. Nov. 2008, 16:45:13, [email protected] wrote: > > Fixed in r32846. > > Chromatic, are you sure about that patch? > > When I just ran 'make fulltest' I got an error in 'testj'. > This halted further execution of 'fulltest'. > IMHO this behavior is no improvement to the old behavior. > > A real solution for this ticket probably needs to be implemented > in t/harness. > > Regards, > Bernhard
The OP complained (rightly so) that failures in one set of tests were hidden if a further set of tests passed. This patch forced the entire test run to abort on failure of a single harness, which doesn't mask the failures. It might not be ideal (as particle suggests in a followup post, having ALL the results reported in a single summary would be nice), but the original issue is resolved. -- Will "Coke" Coleda _______________________________________________ http://lists.parrot.org/mailman/listinfo/parrot-dev
