Re: libpq should append auth failures, not overwrite

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Fabien COELHO <coelho(at)cri(dot)ensmp(dot)fr>
Cc: Michael Paquier <michael(at)paquier(dot)xyz>, PostgreSQL Developers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: libpq should append auth failures, not overwrite
Date: 2018-08-13 14:54:39
Message-ID: 29073.1534172079@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Fabien COELHO <coelho(at)cri(dot)ensmp(dot)fr> writes:
> ISTM that both the hostname and ip should be shown to avoid confusion
> about hosts with multiple ips, esp. as ips are given in any order by the
> dns.
> ...
> Also for homogeneity, I'd suggest to always add the server line. If the
> server introduction is inserted in all cases, including when only one host
> is used, hints become partially redundant:
> server "local.coelho.net" port 5434:
> could not connect to server: Connection refused
> Is the server running on host "local.coelho.net" (127.0.0.1) and accepting
> TCP/IP connections on port 5434?
> This would allow to simplify more hints, which you seem to have done on
> "open read-write session" and "SHOW transaction_read_only" but not others.

As I explained in my comments, the reason I did not do these things
is that I didn't want to change the output for cases in which just a
single host name is given. I still don't. People will think it's
change for the sake of change, and they tend not to like that.
The multi-host feature is new enough that I think we can still get away
with changing how errors are reported in those cases ... but what you're
proposing here is to mess with error-reporting behavior that was settled
on decades ago. I'm not really interested in taking the flak that will
come with that.

regards, tom lane

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2018-08-13 15:16:23 Re: NetBSD vs libxml2
Previous Message Andres Freund 2018-08-13 14:54:05 Re: logical decoding / rewrite map vs. maxAllocatedDescs