Re: [Pgbuildfarm-members] Release 4.17 of the PostgreSQL Buildfarm client

From: Andrew Dunstan <andrew(at)dunslane(dot)net>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: PGBuildFarm <pgbuildfarm-members(at)pgfoundry(dot)org>
Subject: Re: [Pgbuildfarm-members] Release 4.17 of the PostgreSQL Buildfarm client
Date: 2016-02-21 14:37:39
Message-ID: 56C9CBB3.8070800@dunslane.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: buildfarm-members pgsql-hackers

On 02/20/2016 07:40 PM, Tom Lane wrote:
> Andrew Dunstan <andrew(at)dunslane(dot)net> writes:
>> Changes of note: ...
>> * fix trigger_exclude pattern in build-farm.conf
>> ... If you don't need any of the other changes (most people probably won't)
> FWIW, I think the fix in the default trigger_exclude pattern is important
> and most buildfarm owners will want to adopt it. The intention of that
> pattern has always been to not run a build cycle if only doc or .po files
> changed, but it was done wrong, resulting in running a build anyway for
> doc-only changes. So unless you are running "build_docs" or "dist" tests,
> you'll want that fix to save useless builds.
>
> (OTOH, if you are running such tests, you'll want to annotate the fact
> that you're intentionally using a nondefault pattern, so you don't break
> it later.)

Yeah, the new default setting is:

trigger_exclude => qr[^doc/|\.po$],

so you could just adjust your config file to use that.

cheers

andrew

In response to

Browse buildfarm-members by date

  From Date Subject
Next Message Christian Ullrich 2016-03-10 20:44:03 [Pgbuildfarm-members] Missing Git commit ID in woodlouse status
Previous Message Tom Lane 2016-02-21 00:40:57 Re: [Pgbuildfarm-members] Release 4.17 of the PostgreSQL Buildfarm client

Browse pgsql-hackers by date

  From Date Subject
Next Message Pavel Stehule 2016-02-21 15:39:46 Re: proposal: enhancing slow query log, and autoexplain log about waiting on lock before query exec time
Previous Message Amit Kapila 2016-02-21 14:15:39 Re: Speed up Clog Access by increasing CLOG buffers