Re: review: More frame options in window functions
From:
Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To:
Dimitri Fontaine <dfontaine(at)hi-media(dot)com>
Cc:
Robert Haas <robertmhaas(at)gmail(dot)com>,
Martijn van Oosterhout <kleptog(at)svana(dot)org>,
Hitoshi Harada <umi(dot)tanuki(at)gmail(dot)com>, Erik Rijkers <er(at)xs4all(dot)nl>,
PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>
Subject:
Re: review: More frame options in window functions
Dimitri Fontaine <dfontaine(at)hi-media(dot)com> writes:
> Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> writes:
>> However, what it *is* associated with is a sort ordering, and the notion
>> that btree opclasses are what define orderings is sufficiently deeply
>> wired into the system that undoing it would be a huge PITA. So unless
>> we can see a pretty clear future need for more information in this
>> category, I'm not really inclined to invent some new structure
>> altogether. I'm just wondering if anyone does see that...
> I think there's the associativity property of operators that we might
> want to have someday, in order for the planner to know some more about
> joins on A = B then on B = C, or replace with < if you will.
We already do know about that, at least in the case of =. The reason it
doesn't do transitive < deductions is not lack of information but doubt
that it's worth the cycles to try.
regards, tom lane
