On Mon, 2008-05-19 at 18:48 -0700, Chris Darroch wrote: > Personally, I'd suggest just removing it from trunk too ... I'm > not sure exactly how a global cache could be made to work across > multiple databases, since they might be of different versions, etc. > It's not likely to be trivial, I would think, and I'm not sure what > the upside is.
I agree. A big complication in order to shave a bit of runtime for small number of users, if any. > So, anyway, here's a patch against trunk ... it should apply cleanly > against 1.3 as well, I believe. If you're not applying this to trunk > as well, please apply the previous patch, which at least moves the > apr_dbd_mutex_unlock() calls into matching #ifdefs. Thanks, I'll let Bill commit this, given that he was the one handling the whole thing. Thanks for the patch Chris! -- Bojan