Postgres Pro
Downloads
Home   >   mailing lists

Thread: Indexam API changes

Indexam API changes

From
Heikki Linnakangas
Date:
There's a bunch of mails in the patch queue about the indexam API, so we
need to discuss that.

The first question is: do we want to refactor the bitmap index scan API,
if we don't have any immediate use for it? Namely, since we don't have
anyone actively working on the bitmap index patch nor the git patch.

There was also discussion on adding support for "candidate matches",
mainly for GIT, but GiST could possibly also take advantage of that.

If people think it's worth it, I can fix the bit-rot in the patch and
work on it, but personally I don't think it is.

--
   Heikki Linnakangas
   EnterpriseDB   http://www.enterprisedb.com

Re: Indexam API changes

From
Oleg Bartunov
Date:
On Tue, 8 Apr 2008, Heikki Linnakangas wrote:

> There's a bunch of mails in the patch queue about the indexam API, so we need
> to discuss that.
>
> The first question is: do we want to refactor the bitmap index scan API, if
> we don't have any immediate use for it? Namely, since we don't have anyone
> actively working on the bitmap index patch nor the git patch.
>
> There was also discussion on adding support for "candidate matches", mainly
> for GIT, but GiST could possibly also take advantage of that.

we talked about GIN. It'd be great if we eliminate @@@ operator in 8.4 !

>
> If people think it's worth it, I can fix the bit-rot in the patch and work on
> it, but personally I don't think it is.


     Regards,
         Oleg
_____________________________________________________________
Oleg Bartunov, Research Scientist, Head of AstroNet (www.astronet.ru),
Sternberg Astronomical Institute, Moscow University, Russia
Internet: oleg@sai.msu.su, http://www.sai.msu.su/~megera/
phone: +007(495)939-16-83, +007(495)939-23-83

Re: Indexam API changes

From
Gregory Stark
Date:
"Heikki Linnakangas" <heikki@enterprisedb.com> writes:

> There's a bunch of mails in the patch queue about the indexam API, so we need
> to discuss that.
>
> The first question is: do we want to refactor the bitmap index scan API, if we
> don't have any immediate use for it? Namely, since we don't have anyone
> actively working on the bitmap index patch nor the git patch.

I haven't read the patch. My understanding from the discussion is that it
would allow callers of the indexam to receive a hunk of index pointers and
process them rather than have to wait for the complete index scan to finish
before processing any. Is that it?

In general I think we need to be more open to incremental improvements. I
think there are several fronts on which we refuse patches to do X because it's
useless without Y and have nobody working on Y because they would have to
solve X first.

--
  Gregory Stark
  EnterpriseDB          http://www.enterprisedb.com
  Ask me about EnterpriseDB's On-Demand Production Tuning