Florian,
> I have completed an early prototype. Since the pencils-down dealine of
> GSoC has already passed (It was 20th August), I'm currently working on
> a patch to lazily assign XIDs.
Hmmm. Would that allow us to use less XIDs on read-mostly databases? I seem
to remember something about that on -hackers.
Whitepages.com will be thrilled ...
--
Josh Berkus
PostgreSQL @ Sun
San Francisco