Re: php with postgres - Mailing list pgsql-hackers

From Andrew Sullivan
Subject Re: php with postgres
Date
Msg-id 20030725122158.GB13791@libertyrms.info
Whole thread Raw
In response to Re: php with postgres  (Bruce Momjian <pgman@candle.pha.pa.us>)
List pgsql-hackers
On Thu, Jul 24, 2003 at 05:52:07PM -0400, Bruce Momjian wrote:

> Well, our behavior has always been to abort any open transctions on
> client exit.  In fact, it is so obvious, we don't even document it.
> 
> There are some SQL databases that commit on client exit, but that sounds
> just wrong to us.  (They might distinguish between client close and

But not so obvious, then, eh?  I hadn't realised it wasn't
documented, but I think there ought to be a mention of it, maybe in
section 3.4, right after it talks about deciding to abort with
ROLLBACK.  You could add something like, "Also, because every
transaction must either completely succeed or completely fail,
PostgreSQL will automatically abort a transaction in process if the
client disconnects before issuing COMMIT."

A
-- 
----
Andrew Sullivan                         204-4141 Yonge Street
Liberty RMS                           Toronto, Ontario Canada
<andrew@libertyrms.info>                              M2P 2A8                                        +1 416 646 3304
x110



pgsql-hackers by date:

Previous
From: Larry Rosenman
Date:
Subject: Re: libpq_r
Next
From: Tom Lane
Date:
Subject: Re: SELECT FOR UPDATE