Re: Fixing GIN for empty/null/full-scan cases

On Jan 18, 2011, at 1:58 PM, Tom Lane wrote:

> I'm noticing also that I get different rowcounts than you do, although
> possibly that has something to do with the partial-index conditions,
> which I'm not trying to duplicate here (all rows in my table pass those
> two tests).

Shall I send you data with the other two columns?:

>> * Why does it take 3-4x longer to create the GIN than the GiST index
>> on tsvector?
> 
> Perhaps more maintenance_work_mem would help with that; although the
> fine manual says specifically that GIN text search indexes take about
> three times longer to build than equivalent GiST indexes, so maybe that
> behavior is as designed.

Okay then, thanks.

David