Re: Huge speed penalty using <>TRUE instead of =FALSE

On Fri, Jul 17, 2009 at 10:21 AM, Tom Lane<tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Peter Eisentraut <peter_e(at)gmx(dot)net> writes:
>> ... But again, this is data type specific knowledge.
>
> Actually, now that I think about it, the planner already has
> datatype-specific knowledge about boolean equality (see
> simplify_boolean_equality).  It would take just a few more lines of code
> there to recognize "x <> true" and "x <> false" as additional variant
> spellings of the generic "x" or "NOT x" constructs.  Not sure if it's
> worth the trouble though; how many people really write such things?

I don't know, but there's probably somebody. I probably did it myself
a few times, when I was just starting out. If it's easy, it seems
worth doing. The problem with these things is that no matter how lame
it seems to do whatever-it-is, the pain when someone does is really
large... so adding a little bit of code to avoid that seems
worthwhile, at least to me.

> If you really wanted to take it to extremes, you could also reduce
> cases like "x > false", but that's starting to get a bit silly.

Probably that one is beyond even my tolerance.

...Robert