> Hm. On the other hand, relying on WAL for undo means you cannot drop
> old WAL segments that contain records for any open transaction. We've
> already seen several complaints that the WAL logs grow unmanageably huge
> when there is a long-running transaction, and I think we'll see a lot
> more
Don Baccus <[EMAIL PROTECTED]> writes:
> BTW Oracle has a 1000-element limit on the number of values in an
> "IN" set.
Yeah? What happens when you get past that?
regards, tom lane
---(end of broadcast)---
TIP 5: Have you c
Bruce Momjian <[EMAIL PROTECTED]> writes:
> In fact, I was
> right that you have to call transformTargetEntry() -> transformExpr() ->
> transformIdent() to get isRel set, and I have to do the longer fix.
Yes, I would think that you should do transformTargetEntry() first and
then look to see if yo
mlw <[EMAIL PROTECTED]> writes:
> The above is an example of how to write a function that returns multiple
> results.
One suggestion: you must check not only that fcinfo->resultinfo isn't
NULL, but that it points at the sort of node you're expecting. Say
if (fcinfo->resultinfo == NULL |
Stephan Szabo <[EMAIL PROTECTED]> writes:
> I think what we should probably do is make IN better and use that or then
> support =ANY(=SOME)/=ALL on such things. I think =ANY would be easy
> since IN is defined in terms of it in the spec.
And in our code too ;-). ANY/ALL have been there for awhi
Don Baccus <[EMAIL PROTECTED]> writes:
> What's so hard about writing "IN" rather than "=" ???
Even more to the point, if we did adopt such a (crazy IMHO)
interpretation of '=', what makes anyone think that it'd be
any more efficient than IN?
AFAICT, mlw is hoping that redefining '=' would magic
Bruce Momjian <[EMAIL PROTECTED]> writes:
> The only mention I see of this is in c.h:
> #ifndef __cplusplus
> #ifndef bool
> typedef char bool;
> #endif /* ndef bool */
> #endif /* not C++ */
> If you need more cplusplus stuff, lets figure it ou
Bruce Momjian <[EMAIL PROTECTED]> writes:
> OK, here is another patch that does this:
That seems considerably uglier than your first patch. In particular,
why aren't you looking for isRel being set in the Ident node? It
looks to me like you may have changed the behavior in the case where
the Id
Peter Eisentraut <[EMAIL PROTECTED]> writes:
> No really, I don't see a point of not enforcing the correct syntax, when
> adding '.*' is all it takes to get the alternative behaviour in a standard
> way.
True, although there's a certain inconsistency in allowing a whole row
to be passed to a func
mlw <[EMAIL PROTECTED]> writes:
> If I do:
> create temp table fubar as select ftss_results() as songid;
> select * from cdsongs where songid = fubar.songid;
> That works, but that is slow and a lot of people have emotional difficulties
> with using temporary tables.
If you don't like temp tables
(Machine couldn't find mx record for mohawksoft, replying only
to list)
On Sat, 19 May 2001, mlw wrote:
> Sorry to gripe here. Don't get me wrong, I think Postgres is amazing, and I
> think all you guys do an amazing job.
>
> Is it just me, or do others agree, functions returning sets need to
Bruce Momjian <[EMAIL PROTECTED]> writes:
> Seems a tablename with no FROM clause doesn't get marked as isRel
> because it is not in the range table to be matched.
> What would happen if we added auto-star is that a table name in a target
> list would automatically become tablename.*. Seems it i
mlw <[EMAIL PROTECTED]> writes:
> freedb=# select * from cdsongs where songid = ftss_results() ;
> ERROR: Set-valued function called in context that cannot accept a set
'=' is a scalar operation. Try
select * from cdsongs where songid IN (select ftss_results());
regard
mlw <[EMAIL PROTECTED]> writes:
> Is any support for reworking the postgres headers such that they can be used,
> cleanly, in a C++ program?
You'll get no support for a request for a blank check. What do you have
in mind exactly?
ISTM that making the backend's internal headers C++-clean has alr
> Well, I kind of have that already. I can return a set, but I can't use it in a
> join.
>
> freedb=# select ftss_search('all { pink floyd money }') ;
> ftss_search
> -
> 120
> (1 row)
>
> freedb=# select * from cdsongs where songid = ftss_results() ;
> ERROR: Set-valued
Bruce Momjian <[EMAIL PROTECTED]> writes:
> In fact, multi-query transactions are just a special case of
> subtransactions, where all previous subtransactions are
> committed/visible. We could use the same pg_log-style memory area for
> multi-query transactions, eliminating the command counter a
Peter Eisentraut <[EMAIL PROTECTED]> writes:
> Bruce Momjian writes:
>> This patch fixes the problem by generating an error in the parser:
>>
>> test=> select pg_class from pg_class;
>> ERROR: You can't use a relation alone in a target list.
> Maybe it's the parser that's getting it wrong. Wha
Kaare Rasmussen <[EMAIL PROTECTED]> writes:
>> Second: if VACUUM can run in the background, then there's no reason not
>> to run it fairly frequently. In fact, it could become an automatically
>> scheduled activity like CHECKPOINT is now, or perhaps even a continuously
>> running daemon (which wa
Peter Eisentraut <[EMAIL PROTECTED]> writes:
> createlang tries to be helpful by trying test -f $the_file first, to guard
> against attempts to load PL/Tcl when no Tcl support was configured. This
> approach has a few other subtle flaws: it requires createlang to be run
> on the server machine a
I tried to do a TRUNCATE on a table that was referenced by other table and
the refererential integrity wasn't kept. Seems to me this thing behaves
really differently then DELETE FROM table. I think it should be at least
mentioned in the docs or disabled if the table is referenced by foreign
key.
> Second: if VACUUM can run in the background, then there's no reason not
> to run it fairly frequently. In fact, it could become an automatically
> scheduled activity like CHECKPOINT is now, or perhaps even a continuously
> running daemon (which was the original conception of it at Berkeley, BTW
21 matches
Mail list logo
