On Wed, Aug 03, 2005 at 12:12:17PM -0700, Stas Bekman wrote: > Dean Arnold wrote: > > >I think Stas mentioned he'd only tested with DBD::mysql; have > >any other DBD's been tested with this ? > > No, I haven't tried. I want to make it working with mysql dbd first, so I > will know that DBI gives what we need (Tim, take_imp_data still needs some > work when you get a chance :). Once this is done, we can move on to other > dbds.
I'm working on it again now. Finally. The next release has the new ChildHandles attribute (using weakrefs) so I'm planning to have take_imp_data effectively destroy all the child statement handles before returning. That should address the issues. Tim. > Basically DBI::Pool started as a very complicated code, where it was > trying to figure out how to reap the heart out of each thread. But then > the magician Tim came in and provided an API to do it really easy. So now > DBI::Pool or whatever it'll be called should be a piece of cake: a cold > storage of the reaped hearts, served on demand and overriding some of the > API (connect/disconnect) ala Apache::DBI to make things working > transparently.
