--- Original Message ---
From: Joshua D. Drake [EMAIL PROTECTED]
To: Dave Page [EMAIL PROTECTED]
Sent: 15/03/07, 23:51:18
Subject: Re: [GENERAL] pg_dumpall and version confusion
If you want that, create a wrapper program that calls to different
statically compiled versions
Dave Page [EMAIL PROTECTED] writes:
--- Original Message ---
From: Joshua D. Drake [EMAIL PROTECTED]
To: Dave Page [EMAIL PROTECTED]
Sent: 15/03/07, 23:51:18
Subject: Re: [GENERAL] pg_dumpall and version confusion
If you want that, create a wrapper program that calls to different
Jorge Godoy wrote:
Dave Page [EMAIL PROTECTED] writes:
--- Original Message ---
From: Joshua D. Drake [EMAIL PROTECTED]
To: Dave Page [EMAIL PROTECTED]
Sent: 15/03/07, 23:51:18
Subject: Re: [GENERAL] pg_dumpall and version confusion
If you want that, create a wrapper program
Dave Page wrote:
--- Original Message ---
From: Joshua D. Drake [EMAIL PROTECTED]
To: Dave Page [EMAIL PROTECTED]
Sent: 15/03/07, 23:51:18
Subject: Re: [GENERAL] pg_dumpall and version confusion
If you want that, create a wrapper program that calls to different
statically
Joshua D. Drake wrote:
other deficiencies let's not add to complexity by having an
--output-version.
If you want that, create a wrapper program that calls to different
statically compiled versions of pg_dump.
Joshua D. Drake
I am afraid that I have to agree with Tom here. Pg_dump has
Tony Caduto [EMAIL PROTECTED] writes:
I don't think the whole --output-version thing would be a good idea
either, but it would be a simple matter to add some logic for the
GRANT ON SEQUENCE, which seems to be the only thing that is really
causing problems at least between 8.1 and 8.2.
You
I inherited an existing system with no documents on how it was originally
configuerd, so I'll provide as much as I can. OS is CentOS 4.3 Linux
Distribution.
When I search my system for anything postgres I find a boatload of 8.0.4
source, and a boatload of 7.4.8 documentation.
postgres --version
Alan J Batsford [EMAIL PROTECTED] writes:
When I search my system for anything postgres I find a boatload of 8.0.4
source, and a boatload of 7.4.8 documentation.
rpm -qa | grep postgres would probably prove enlightening. It sounds
like you have postgresql-server at 8.1.4 and the other
Tom Lane wrote:
We try to make pg_dump
cope with prior server releases, but since we lack a time machine we
can't go back and teach old versions about subsequent server changes...
Tom,
How come version 8.2 of pg_dump uses GRANT ON SEQUENCE when dumping a
8.1 or older database?
Was it just a
Tony Caduto wrote:
Tom Lane wrote:
We try to make pg_dump
cope with prior server releases, but since we lack a time machine we
can't go back and teach old versions about subsequent server changes...
Tom,
How come version 8.2 of pg_dump uses GRANT ON SEQUENCE when dumping a
8.1 or older
Alvaro Herrera wrote:
What's wrong with that? 8.2 will understand the GRANT ON SEQUENCE
without a problem.
Yes, but 8.1 and earlier wont!
If you dump a 8.1 database (with 8.2 pg_dump) and then use that dump to
restore to another 8.1 server, any permissions on sequences will fail
to
Tony Caduto wrote:
Alvaro Herrera wrote:
What's wrong with that? 8.2 will understand the GRANT ON SEQUENCE
without a problem.
Yes, but 8.1 and earlier wont!
If you dump a 8.1 database (with 8.2 pg_dump) and then use that dump to
restore to another 8.1 server, any permissions on
Dave Page wrote:
This is a problem I've been thinking about on and off recently - and I
am starting to come to the conclusion that shipping each version of
the utilities is the only way things are likely to work unless someone
puts some really significant effort into adding backwards
On Thu, 2007-03-15 at 11:53, Tony Caduto wrote:
Dave Page wrote:
This is a problem I've been thinking about on and off recently - and I
am starting to come to the conclusion that shipping each version of
the utilities is the only way things are likely to work unless someone
puts some
Tony Caduto [EMAIL PROTECTED] writes:
Alvaro Herrera wrote:
What's wrong with that? 8.2 will understand the GRANT ON SEQUENCE
without a problem.
Yes, but 8.1 and earlier wont!
You're missing the point entirely. The versioning contract for pg_dump
is that version M.N can dump from servers
Tom Lane wrote:
I have *no* interest in trying to get it to support an --output-version
switch...
There speaks a man who has the luxury of not having to worry about
multi-version admin tools :-(
/D
---(end of broadcast)---
TIP 4: Have you
Dave Page wrote:
Tom Lane wrote:
I have *no* interest in trying to get it to support an --output-version
switch...
There speaks a man who has the luxury of not having to worry about
multi-version admin tools :-(
Is it more difficult for you to ship a pg_dump-8.0, pg_dump-8.1, etc
along
Alvaro Herrera wrote:
Dave Page wrote:
Tom Lane wrote:
I have *no* interest in trying to get it to support an --output-version
switch...
There speaks a man who has the luxury of not having to worry about
multi-version admin tools :-(
Is it more difficult for you to ship a pg_dump-8.0,
On Thu, Mar 15, 2007 at 09:19:31PM +, Dave Page wrote:
Even if we went through the insane pain (for me at least) of maintaining
multiple branches of the Windows build of PostgreSQL, *and* custom
builds of the utilities built against a single libpq, we still couldn't
ship a single set
Porting it to Windows could be tricky though, given its use of symlinks
and wrapper scripts.
FWIW, as of Vista and Server Longhorn, Windows now supports Symlinks. I
don't know about wrapper scripts though.
--
Paul Lambert
Database Administrator
AutoLedgers
---(end
Dave Page wrote:
Tom Lane wrote:
I have *no* interest in trying to get it to support an --output-version
switch...
There speaks a man who has the luxury of not having to worry about
multi-version admin tools :-(
I am afraid that I have to agree with Tom here. Pg_dump has plenty of
other
21 matches
Mail list logo
