From: Tom Lane <tgl@sss.pgh.pa.us>
To: Fabien COELHO <coelho@cri.ensmp.fr>
cc: PostgreSQL Developers <pgsql-hackers@lists.postgresql.org>
Subject: Re: libpq connection timeout mismanagement
In-reply-to: <alpine.DEB.2.21.1808140909250.27691@lancre>
References: <5735.1533828184@sss.pgh.pa.us>
 <alpine.DEB.2.21.1808111611040.30788@lancre> <20525.1534180589@sss.pgh.pa.us>
 <alpine.DEB.2.21.1808140909250.27691@lancre>
Comments: In-reply-to Fabien COELHO <coelho@cri.ensmp.fr>
	message dated "Tue, 14 Aug 2018 11:03:16 +0200"
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-ID: <5105.1534267032.1@sss.pgh.pa.us>
Content-Transfer-Encoding: 8bit
Date: Tue, 14 Aug 2018 13:17:12 -0400
Message-ID: <5106.1534267032@sss.pgh.pa.us>
Precedence: bulk

Fabien COELHO <coelho@cri.ensmp.fr> writes:
>> Right.  As before, I'm not excited about rejecting trailing junk,
>> considering we never did before.

> My 0.02¤: accepting trailing junk is just a recipee for hiding bugs:
>   sh> psql "connect_timeout=2,port=5433"
>   psql> \conninfo
>   ... port=5432 # port directive was silently ignored
> So erroring out would be a good thing, and it could be done on a new 
> release.

Perhaps.  That should certainly be a separate patch, though, and
it should cover everything not just connection_timeout (I see at
least five parameters that are parsed with atoi() and nothing else).
I'm still not excited, but feel free to send a patch.

			regards, tom lane