> How about this?
> added datetime default datetime(now()) not null,
Yes, this apparently works by converting from timestamp to datetime.
In general datetime has better dynamic range than does timestamp, but
for calculating 'now' both are equivalent in performance.
- Tom
--
Does this mean that if I performed a pg_dump at 8:00 AM, performed some
updates, inserts and deletes at 10:00AM, destroyed and restored my database
at 12:00 noon using the 8:00AM pg_dump that all of the insert, updates, and
deletes that I did at 10:00 AM would still exist in the newly restored
dat
> I've installed PostgreSQL on a development machine and I'm trying to
> convince my company to migrate applications to it. The biggest question
> I've had is:
>
> "How do you recover the database? After a snapshot is restored from
> archive can you use the transaction log to roll it forward?"
>
I have been trying to re-compile PostgreSQL 6.4.2 on my IRIX 6.5.3
machine so that I can use the readline option. I installed GNU readline
4.0 on my IRIX. It created a new directory called
/usr/local/include/readline which contains the following files:
> ls
---
--BTW, the update below, when done in psql and with [. . .and ssms.autoinv
= 101999;] added (in other words, do it for just one record), it works
fine. When I try to do it for multiple records, bombola! Maybe this is
for the sql group. . .
You know, I have asked this week how to correct som
You know, I have asked this week how to correct some error messages I get
when I vacuum related to TUPLE<>HEAP sizes - I thought I caused the problem
by setting all relname to lower(relname) in pg_class and attname to
lower(attname) in pg_attribute. But I don't think that caused the problem.
I've installed PostgreSQL on a development machine and I'm trying to
convince my company to migrate applications to it. The biggest question
I've had is:
"How do you recover the database? After a snapshot is restored from
archive can you use the transaction log to roll it forward?"
The section o
Version 6.4.2 fails to compile with these errors on my HPUX system with the
HP CC (C++) compiler.
I don't need either of these interfaces (ecpg or libpq++) so I don't really
care; just trying to give feedback to the developers.
Hi.
One problem persists: I can not dump, HELP!
dumpSequence(resume_seq): 0 (!= 1) tuples returned by SELECT
> But not user apache?
Of course no, I have about 20 databases belonging to that
user.
> If you do this same thing, but copy over ALL files, does the problem
> reappear?
I didn't try
