Added to TODO:
> o Have ALTER TABLE rename SERIAL sequences
Seems we at least need this. Doesn't dependency tracking make this
easy to do now?
---
Tom Lane wrote:
> "Christopher Kings-Lynne" <[EMAIL PROTECTED]> writ
"Christopher Kings-Lynne" <[EMAIL PROTECTED]> writes:
>> I think we should stick with the existing naming convention.
^^
> Non-colliding?
No; see above.
> Otherwise, it'd be ludicrous to fail a table rename because
> a sequence with the
> I think we should stick with the existing naming convention. The only
> actual problem that's been pointed out here is that an ALTER TABLE
> (or COLUMN) RENAME on a serial column doesn't update the sequence name
> to match. Seems to me we could fix that with less effort than any of
> these solu
Rod Taylor <[EMAIL PROTECTED]> writes:
> 200N spec proposes 'NEXT VALUE FOR '.
> Tom will shoot me if I submit that though (VALUE as a keyword again). I
> suppose one could make it a variable, and confirm it's value is VALUE?
> Anyway, once again we could extend to include:
> NEXT VALUE ON table(
Yes, our sequence name/column linkage should be much more automatic than
it is now, and if we do that, we can start to think about sequence name
collision avoidance.
---
Rod Taylor wrote:
-- Start of PGP signed section.
> On
On Wed, 2003-03-05 at 14:37, Bruce Momjian wrote:
> Interesting idea. The bigger problem is that apps who use the sequence
> name also would have problems running after the restore. Seems we need
> column.nextval() so you can increment the sequence without knowing the
> sequence name, just the co
Interesting idea. The bigger problem is that apps who use the sequence
name also would have problems running after the restore. Seems we need
column.nextval() so you can increment the sequence without knowing the
sequence name, just the column name. Of course, this related to this
TODO item:
On Wed, 2003-03-05 at 10:36, Bruce Momjian wrote:
> Ewe. There would no longer be a guaranteed name for the serial column
> sequence. Of course, pg_depend has the information, but how do you get
> at that when you create the dump file, and be _sure_ you are going to
> hit the right name, especial
Ewe. There would no longer be a guaranteed name for the serial column
sequence. Of course, pg_depend has the information, but how do you get
at that when you create the dump file, and be _sure_ you are going to
hit the right name, especially if you restore only part of the dump.
Seems this kill
OK,
I have discovered a problem with my auto-naming patch. It's do to with
dumping serial columns with pg_dump, eg:
--
-- TOC entry 2 (OID 1004551)
-- Name: users_users; Type: TABLE; Schema: public; Owner: chriskl
--
CREATE TABLE users_users (
userid serial NOT NULL,
firstname character
10 matches
Mail list logo
