On Fri, Apr 9, 2010 at 4:37 PM, Andrew Dunstan <andrew@dunslane.net> wrote:
> (Personally I wish some SOC student would do LATERAL. It would be useful and
> is a feature of about the right size, I think).
Actually, I think that requires two rounds of significant executor
refactoring. This is round two:
http://archives.postgresql.org/pgsql-hackers/2009-10/msg00994.php
Round one, which Tom said he was already planning to do for 9.1, is to
eliminate the hack by which "outer tuples" are passed down into
nest-loops-with-inner-indexscan. Once those two things are done the
remaining work might be suitable for a GSoC project.
...Robert