From: Tom Lane <tgl@sss.pgh.pa.us>
To: Michael Paquier <michael@paquier.xyz>
cc: Thomas Munro <thomas.munro@gmail.com>,
        Daniel Gustafsson <dgustafsson@pivotal.io>,
        ANASTACIO Tiago <tiago_anastacio@yahoo.fr>,
        Peter Geoghegan <pg@bowt.ie>,
        PostgreSQL mailing lists <pgsql-bugs@lists.postgresql.org>
Subject: Re: BUG #15726: parallel queries failed ERROR: invalid name syntax
 CONTEXT: parallel worker
In-reply-to: <20190412033250.GE2144@paquier.xyz>
References: 
 <CA+hUKGKJ-amPMuwFhgc5T6UH=Wcn3CG3gmowmkYqTo2WsQJM4A@mail.gmail.com>
 <CA+hUKGLdzBsa_SZajU8J9JDSHzdU6+LVq1g6bUsekiWBpjVDNQ@mail.gmail.com>
 <CAH2-Wzkif1R6imAhgcRv1cXqomYBDez9g1rKAhDixiEi7b2KOg@mail.gmail.com>
 <1831482307.24522245.1554199154256@mail.yahoo.com>
 <CA+hUKGLvwuKEf_dk-zZANRLGfT9eO5hbKB_+KYeb-Uef_Q7G-g@mail.gmail.com>
 <29888.1554218381@sss.pgh.pa.us>
 <CA+hUKGJ=nTwC750vLGtQUGXFHFKxFF31=Qwn=Pd70S4jZanA2g@mail.gmail.com>
 <C6B09DDC-FDDA-466D-AFBD-F8C966CD97C1@pivotal.io>
 <CA+hUKGLftxFoLP6_qH6QnNPogTtOXp7cr+Bv7nV=CABRMWF_PQ@mail.gmail.com>
 <11677.1554995054@sss.pgh.pa.us> <20190412033250.GE2144@paquier.xyz>
Comments: In-reply-to Michael Paquier <michael@paquier.xyz>
	message dated "Fri, 12 Apr 2019 12:32:50 +0900"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1337.1555087669.1@sss.pgh.pa.us>
Date: Fri, 12 Apr 2019 12:47:49 -0400
Message-ID: <1338.1555087669@sss.pgh.pa.us>
Precedence: bulk

Michael Paquier <michael@paquier.xyz> writes:
> On Thu, Apr 11, 2019 at 11:04:14AM -0400, Tom Lane wrote:
>> The extra appearance of "parallel worker" seems a bit redundant,
>> though I don't know if we can get rid of it.
>> 
>> Could we at least be consistent about whether the context is
>> "parallel worker" or "parallel worker process"?

> Indeed.  My vote would be to back-patch that stuff.

After thinking about it some more: can't we just make the new context
message be
	CONTEXT:  while setting parameter "x" to "y"
full stop?  The outer context line about "parallel worker" should
be enough for that.  As a bonus, if we ever decide that such a
context line would be useful for all GUC errors, we wouldn't need
a different spelling of it for that.

I took a quick look through the patch, and I think it's okay
implementation-wise, though personally I'd have used some less
generic name than "pair" for the variables.

			regards, tom lane