Re: [PERFORM] Hash Anti Join performance degradation

On Wed, Jun 1, 2011 at 4:47 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Robert Haas <robertmhaas@gmail.com> writes:
>> I guess the real issue here is that m1.id < m2.id has to be evaluated
>> as a filter condition rather than a join qual.
>
> Well, if you can invent an optimized join technique that works for
> inequalities, go for it ... but I think you should get at least a
> PhD thesis out of that.

Sounds good, except that so far NOT getting a PhD seems like a much
better financial prospect.  :-)

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company