On Fri, 11 Oct 2002, Andi Gutmans wrote: > At 02:52 PM 10/11/2002 +0200, Derick Rethans wrote: > >On Fri, 11 Oct 2002, Andi Gutmans wrote: > > > > > We really shouldn't change PHP just so that tests end up being better ;) > > > >We're trying to the point that there may be no failed tests when we > >release, and this is needed to make this test pass. The non-bundled GD > >crashes when this test is run. > > You shouldn't clutter PHP just so that the make test prints out a nice > "passed". If the non-bundled GD crashes there's no reason the test > shouldn't crash IMO.
I agree. The test harness is not just for our own QA. It is also a series of tests for people to run on their systems to see what works and what doesn't. If they are using the non-bundled GD then this function is broken and we shouldn't hide this fact from them. -Rasmus -- PHP CVS Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
