compiling 7.1-beta1

"Martin A. Marques" <martin@math.unl.edu.ar> writes:

I'm compiling beta 1 of 7.1 and I have a par of questions.

analyze.c: In function `transformInsertStmt':
analyze.c:425: warning: unused variable `resnode'

Fixed in current sources. (I think there is still one unused-var
complaint left in the XLOG code; I've been waiting on Vadim to do
something about it, because it looks like there is code still to be
written there.)

The other question is if I have to do something special (dump and restore)
when upgrading from 7.1-beta1 to 7.1-final (or any of the other betas)?

You will need an initdb to go from beta1 to beta3. Sorry about that;
we try to avoid forced initdb after beta cycle starts, but sometimes
it's not possible.

You might want to skip testing beta1 and just start with beta3, or even
a current nightly snapshot.

regards, tom lane

Fixed in current sources. (I think there is still one unused-var
complaint left in the XLOG code; I've been waiting on Vadim to do
something about it, because it looks like there is code still to be
written there.)

Just commented out for now.

Vadim

El Jue 18 Ene 2001 15:15, Tom Lane escribi�:

The other question is if I have to do something special (dump and
restore) when upgrading from 7.1-beta1 to 7.1-final (or any of the other
betas)?

You will need an initdb to go from beta1 to beta3. Sorry about that;
we try to avoid forced initdb after beta cycle starts, but sometimes
it's not possible.

You might want to skip testing beta1 and just start with beta3, or even
a current nightly snapshot.

Well, my problem was that a downgrade had to be made to the only Solaris8
SPARC back to Solaris7, and postgres broke. So I recompiled it, in a new
directory, and copied the data directory to the new postgres instaltion
directory, but when I connect with pgaccess to and db, I see the postgres
tables (like pg_aggregate, pg_group, pg_scripts, pg_trigger, etc). Did I do
something wrong? Any way to fix it?

--
System Administration: It's a dirty job,
but someone told I had to do it.
-----------------------------------------------------------------
Mart�n Marqu�s email: martin@math.unl.edu.ar
Santa Fe - Argentina http://math.unl.edu.ar/~martin/
Administrador de sistemas en math.unl.edu.ar
-----------------------------------------------------------------