After all, this looks like a compiler bug on CentOS 5 to me. If I compile PyLucene with DEBUG=1, all tests pass - switching back to DEBUG=0 and the bug re-appears. It turned out that some gcc optimizations seem to be wrong, only "-O0" works for me. I will have to check if the problem is present in Fedora 8 using icedtea.
Thank you very much for your help. -- Felix Schwarz software development and Linux system administration main focus: secure database applications http://www.schwarz.eu Gubener Str. 38 10243 Berlin Germany _______________________________________________ pylucene-dev mailing list [email protected] http://lists.osafoundation.org/mailman/listinfo/pylucene-dev
