I've tried using the Postgres module a bit lately, and now I've stumbled across an odd behaviour: db->list_fields(table_name) throws an error "Error in frontend-backend communications.". This seems to be caused by the list_fields() function assuming that the "pg_class" table always has a "relexpires" columns, which is not the case in the Postgres database I'm connecting to. Should there always be a "relexpires" column in the "pg_class" table in a Postgres database (that is, is the one I'm connecting to buggy?), or is Pike just making in incorrect assumption?
(Sorry if this resolved already)
Modern PostgreSQL:s don't have the relexpires field. I only get hits on Postgres95 when I search for it. Perhaps it was dropped because it wasn't used:
(From Postgres95 docs)
The relexpires entry is the amount of history that should be saved for this relation. This is not used in the current system.
/ Björn Wingman
Previous text:
2003-02-14 14:53: Subject: Postgres support bug?
I've tried using the Postgres module a bit lately, and now I've stumbled across an odd behaviour: db->list_fields(table_name) throws an error "Error in frontend-backend communications.". This seems to be caused by the list_fields() function assuming that the "pg_class" table always has a "relexpires" columns, which is not the case in the Postgres database I'm connecting to. Should there always be a "relexpires" column in the "pg_class" table in a Postgres database (that is, is the one I'm connecting to buggy?), or is Pike just making in incorrect assumption?
/ Leif Stensson, Lysator
Ah, thanks for the information. Fix committed into Pike 7.5 now.
/ Leif Stensson, Lysator
Previous text:
2003-02-20 11:56: Subject: Postgres support bug?
(Sorry if this resolved already)
Modern PostgreSQL:s don't have the relexpires field. I only get hits on Postgres95 when I search for it. Perhaps it was dropped because it wasn't used:
(From Postgres95 docs)
The relexpires entry is the amount of history that should be saved for this relation. This is not used in the current system.
/ Björn Wingman
"Leif Stensson, Lysator @ Pike developers forum" 10353@lyskom.lysator.liu.se writes:
Ah, thanks for the information. Fix committed into Pike 7.5 now.
[...]
Ah, and here I was thinking to make a triumphant (?) reentry with a patch :)
Hello everyone.
You can make a triumpant backport. :)
/ Peter Bortas
Previous text:
2003-02-21 13:08: Subject: Re: Postgres support bug?
"Leif Stensson, Lysator @ Pike developers forum" 10353@lyskom.lysator.liu.se writes:
Ah, thanks for the information. Fix committed into Pike 7.5 now.
[...]
Ah, and here I was thinking to make a triumphant (?) reentry with a patch :)
Hello everyone.
-- kinkie (kinkie-pike-devel [at] kinkie [dot] it) Random fortune, unrelated to the message: Birthdays are like busses, never the number you want.
/ Brevbäraren
Welcome.
/ Martin Nilsson (Åskblod)
Previous text:
2003-02-21 13:08: Subject: Re: Postgres support bug?
"Leif Stensson, Lysator @ Pike developers forum" 10353@lyskom.lysator.liu.se writes:
Ah, thanks for the information. Fix committed into Pike 7.5 now.
[...]
Ah, and here I was thinking to make a triumphant (?) reentry with a patch :)
Hello everyone.
-- kinkie (kinkie-pike-devel [at] kinkie [dot] it) Random fortune, unrelated to the message: Birthdays are like busses, never the number you want.
/ Brevbäraren
pike-devel@lists.lysator.liu.se
-
Bj�rn Wingman ＠ Pike developers forum -
Kinkie -
Leif Stensson, Lysator ＠ Pike developers forum
-
Martin Nilsson (�skblod) ＠ Pike (-) developers forum
-
Peter Bortas ＠ Pike developers forum