I'm still working trying to come up with a way to make this bug reproducible on command. So far, it fails sometimes 10 times out of 10, and then maybe 1 out of 1000. This is *very* frustrating.
Is there any way I can get some statistics on the state of the interpreter at some point? We're require_once()'ing and include_once()'ing a buttload of files and I'd love to get some idea of the memory used during a particular request, including number of lines processed, defines defined, objects defined, etc. Programmatic access to the context would be cool... :-)
Has anyone given any thought to this problem of corruption of defines? Are there any other instances of data corruption that we could maybe draw lines between?
Thanks,
B
--
Brian Lalor
[EMAIL PROTECTED]
(v) 480-333-3196
(f) 480-760-9298