From: | Paul Ramsey <pramsey(at)refractions(dot)net> |
---|---|
To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, dblasby(at)refractions(dot)net |
Cc: | "Marc G(dot) Fournier" <scrappy(at)postgresql(dot)org>, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: PostGIS dropgeometrycolumn function (Was: Re: [7.4] |
Date: | 2004-02-03 19:13:39 |
Message-ID: | 13173ADF-567D-11D8-B42D-000393D33C2E@refractions.net |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Bitter experience... I am going to cc Dave here, because I could swear
we went through many conniptions trying to make this work.
And yet I just did this:
create view mytables as select relname from pg_class where relam = 0
and relname not like 'pg_%';
And it seems to work fine.
Oh, now I remember. The deal was not views, it was triggers. Since our
geometry_columns contains some information not available via a query on
existing data, a trigger was what we wanted, so we could harvest the
information from a variety of places, and have some spare columns for
things like the geometry selectivity stats.
Paul
On Tuesday, February 3, 2004, at 11:00 AM, Tom Lane wrote:
> Paul Ramsey <pramsey(at)refractions(dot)net> writes:
>> In an idea world though, we would construct the thing as a view, so
>> that when you did a CREATE TABLE that included a geometry type, you
>> would automatically get a row in geometry_columns. That requires a
>> view
>> on system tables though, and that just does not work. :/
>
> Uh, what makes you say it doesn't work?
>
> regards, tom lane
>
Paul Ramsey
Refractions Research
Email: pramsey(at)refractions(dot)net
Phone: (250) 885-0632
From | Date | Subject | |
---|---|---|---|
Next Message | Teodor Sigaev | 2004-02-03 19:18:47 | Re: Multicolumn Indexing using R-Tree |
Previous Message | Tom Lane | 2004-02-03 19:00:39 | Re: PostGIS dropgeometrycolumn function (Was: Re: [7.4] "permissions problem" with pl/pgsql function ) |