Re: [HACKERS] ordering RH6.1

On Thu, 16 Dec 1999, Lamar Owen wrote:
> On Thu, 16 Dec 1999, Lamar Owen wrote:
>
> > [Cristian, Jeff: we have a problem here. RedHat 6.1 Install versus
> > RedHat 6.0 upgraded to 6.1 behaves differently. Ideas of where to start
> > looking?]
> > I'm going to retry this exact set of queries again at home -- I wasn't
> > able to reproduce the last set of results -- but we'll see what happens
> > here.
>
> Ok, confirmation. On my home machine, which was upgraded to RedHat 6.1 from
> RedHat 6.0, I get the correct results:

More information: it seems that the i18n support is the cause of this. If you
remove or rename the file /etc/sysconfig/i18n and restart, then even the fresh
RedHat 6.1 install provides the correct results for this query. This file, I
think, is created during a fresh installation of RH 6.1 -- it doesn't seem to
belong to any RPM. An upgrade wouldn't create this file..... (Jeff? Cristian?
am I right on this?)

Running regression, I get the float8 and geometry failures, but I now get an
opr_sanity failure -- but the other collation failures are gone. The opr_sanity
failure diff:

*** expected/opr_sanity.out Wed May 12 11:02:34 1999
--- results/opr_sanity.out Thu Dec 16 22:39:45 1999
***************
*** 48,56 ****
(p1.proargtypes[0] < p2.proargtypes[0]);
proargtypes|proargtypes
-----------+-----------
- 25| 1043
1042| 1043
! (2 rows)

QUERY: SELECT DISTINCT p1.proargtypes[1], p2.proargtypes[1]
FROM pg_proc AS p1, pg_proc AS p2
--- 48,55 ----
(p1.proargtypes[0] < p2.proargtypes[0]);
proargtypes|proargtypes
-----------+-----------
1042| 1043
! (1 row)

QUERY: SELECT DISTINCT p1.proargtypes[1], p2.proargtypes[1]
FROM pg_proc AS p1, pg_proc AS p2

HOWEVER, after doing an initdb and rerunning regression, this test no longer
fails. FWIW.

I seems that charmap (i18n) rears its ugly head even if locale doesn't.

--
Lamar Owen
WGCR Internet Radio