From: | The Hermit Hacker <scrappy(at)hub(dot)org> |
---|---|
To: | pgsql-hackers(at)postgresql(dot)org |
Subject: | i386_solaris port...problem found... |
Date: | 1998-02-12 04:38:39 |
Message-ID: | Pine.BSF.3.96.980212003246.261Z-100000@thelab.hub.org |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Remember that heap_modifytuple error that I was reporting for
i386_solaris? Well, I hit it with my server in Toronto today trying to
get a v6.3 server up for Julie to test ODBC against...which blew my mind,
since I hadn't experienced any problems here at home.
After investigating further, it appears the problem was a versioning
one...when I was running 'initdb', it was taking 'postgres' out of my
path, which is different (v6.2.1 vs v6.3) then the one that I wanted.
Once I set the path correctly, I got a clean initdb from i386_solaris as
well...will try regression tests tomorrow.
With that in mind, would it make more sense to have a --pgbase vs --pglib
option to initdb? So that we'd use:
$pgbase/bin/postgres with $pgbase/lib/..?
--pgdata is still seperate, for obvious reasons. The bin/lib
directories should be central...the data directory can be pretty much
created anywhere.
Comments?
Marc G. Fournier
Systems Administrator @ hub.org
primary: scrappy(at)hub(dot)org secondary: scrappy(at){freebsd|postgresql}.org
From | Date | Subject | |
---|---|---|---|
Next Message | t-ishii | 1998-02-12 07:36:39 | configure fails on PPC/Linux |
Previous Message | Bruce Momjian | 1998-02-12 04:36:24 | Re: [HACKERS] PostGreSQL v6.2.1 for Linux Alpha |