From: Holger W. <ho...@co...> - 2003-03-28 16:25:05
|
Ian Romanick wrote: > Brian Paul wrote: > >> >> OK, I'll just have to disable the bad code for now. > > > Is there a reasonable stand-alone test that exhibits the problem? I > have access to a headless K6-2 box. I could use that to try and track > the problem down. Josh and me implemented regression tests for the transform functions, these will run at library initialisation time. Just enable them to get an idea which functions fail and why. Holger |