Full text Indexing -out of contrib and into main..

From: "John Huttley" <John(at)mwk(dot)co(dot)nz>
To: <pgsql-hackers(at)postgresql(dot)org>
Subject: Full text Indexing -out of contrib and into main..
Date: 2000-11-28 01:51:43
Message-ID: 004001c058dd$c2482aa0$1401a8c0@MWK.co.nz
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

>
> Maybe asking 'Why isn't the contrib full-text-indexer not in the main
> tree?' would be more productive on that front.

Well, yes. Why isn't it?

Full text indexing should be just as much a feature as any other key feature in
PG.
With the advent of unlimited file and record lengths in 7.1, this would be a good
time to
include it.

FTI is particularly useful in the context of web content engines.

> How did you attempt to build it under the RPM install? I assume you had
> the postgresql-devel package installed, and the include paths set
> properly.... Of course, most of what is in contrib assumes a full source
> tree is lying around (argh)....ie, it's wanting to include
> Makefile.global in its Makefile. And, on the RPM dist, Makefile.global
> isn't (yet) packaged.

Yes, I have the devel RPM, but FTI couldn't find its include files. Or its
libraries. Or something.

Building from a source tree has always been better for me. The catch is mixing
that with the RPMS,
which put things in unholy locations.
Its very hard (for me at least) to update an RPM version with a version compiled
from the source.

I recently updated my PG system from 6.5.3 to 7.0.3, still RPMs, (another fun
job) and have not tried to compile FTI
subsequently.

However if I tried hard enough I'm sure I could fix it. For the moment, I'm on
another job so I'm not worrying.

Regards

John

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Nathan Myers 2000-11-28 01:56:37 Re: Indexing for geographic objects?
Previous Message Nathan Myers 2000-11-28 01:49:46 Re: 8192 BLCKSZ ?