Re: [BUGS] setting time zone during a transaction causes time warp

2003-02-19 Thread Tom Lane
"Robert Haas" <[EMAIL PROTECTED]> writes: > While using PostgreSQL 7.2.3, I discovered that if I set the time zone > during the transaction, "now" takes on an incorrect value for the > remainder of that transaction. The problem seems to be restricted to the result of 'now'::timestamptz, and not

[BUGS] setting time zone during a transaction causes time warp

2003-02-19 Thread Robert Haas
While using PostgreSQL 7.2.3, I discovered that if I set the time zone during the transaction, "now" takes on an incorrect value for the remainder of that transaction.  Once the transaction is committed, everything goes back to normal.  I've reproduced the problem below using the "psql" client.  T

Re: [BUGS] MacOS X build --with-pam - bad directory for pam_appl.h

2003-02-19 Thread Phillip C . Dibner
Great - thanks for your reply! Will advise other Mac OS X colleagues in the mean time. On Wednesday, February 19, 2003, at 01:50 PM, Bruce Momjian wrote: 7.4 will look in pam/ directory too for OS X. I didn't backpatch this into 7.3.X. ---

Re: [BUGS] MacOS X build --with-pam - bad directory for pam_appl.h

2003-02-19 Thread Bruce Momjian
If I get another report, I may backpatch to 7.3.X. In fact, I probably should anyway. --- Phillip C. Dibner wrote: > Great - thanks for your reply! Will advise other Mac OS X colleagues > in the mean time. > > On Wednes

[BUGS] MacOS X build --with-pam - bad directory for pam_appl.h

2003-02-19 Thread Phillip C . Dibner
Folks, This is minor. Nonetheless: The --with-pam option seeks for pam_appl.h at path , but this directory does not exist in the stock Jaguar installation. Please see configure output snippet below. Of course, --with-libraries can't help here; the security directory is in the argument of t

Re: [BUGS] Trigger Crashes Database Back End.

2003-02-19 Thread Tom Lane
Peter Childs <[EMAIL PROTECTED]> writes: > This only happerns when you have the following trigger (at bottom) on the > database. The trigger is attached to pratically every table. The plpython trigger support fails if the same trigger is attached to more than one table (unless the tables all have

[BUGS] Trigger Crashes Database Back End.

2003-02-19 Thread Peter Childs
I have reported this bug before. It happerns of 7.3, 7.3.1 and 7.3.2. Having done some further research on the bug.. I have discovered that the following tranaction (which can be run from a script) with crash the back end. BEGIN; SET CONSTRAINTS ALL DEFERRED; UPDATE location set phoneid = NUL

Re: [BUGS] my schema + sql function

2003-02-19 Thread Tom Lane
Marek "Wróbel" <[EMAIL PROTECTED]> writes: > In my schema : "my" I create sql function > CREATE OR REPLACE FUNCTION my.my_foo(int4) RETURNS varchar AS ' > select k.k_dupa ||coalesce( '' ['' || k.k_number || '']'' ,) > from kont k > where k_id = $1; > ' LANGUAGE 'sql'; > table kont

[BUGS] Bug #899: pg_dumpall produce erroneous files if the databbase use procedural language.

2003-02-19 Thread pgsql-bugs
Laurent Faillie ([EMAIL PROTECTED]) reports a bug with a severity of 2 The lower the number the more severe it is. Short Description pg_dumpall produce erroneous files if the databbase use procedural language. Long Description Hi all, I'm working to migrate my database from postgresql 7.2.2 to 7

[BUGS] Bug #898: client_encoding can't support GB18030 encoding in postgresql 7.3.1

2003-02-19 Thread pgsql-bugs
zzhb ([EMAIL PROTECTED]) reports a bug with a severity of 3 The lower the number the more severe it is. Short Description client_encoding can't support GB18030 encoding in postgresql 7.3.1 Long Description When I insert records using files, psql -f install.sql if I set PGCLIENTENCODING=GB18030,