Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:105875
Comment: DKIM? See http://www.dkim.org
Comment: DomainKeys? See http://antispam.yahoo.com/domainkeys
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
	s=default; d=exussum.co.uk;
	h=Received:To:Subject:MIME-Version:Content-Type:Content-Transfer-Encoding:Date:From:In-Reply-To:References:Message-ID:X-Sender:User-Agent;
	b=OHCb9NO6XJbLHzn9dioYKTu28Bg8uNTMEbAwe/0kL2fE1TxLsUp+EEf/4dWphZ
	dRGjBeN4YbaNmbyf0QsGlo8+l3nwWnlP5019XFyp4YbjjM3XjprIgsojL5ERFEgO
	wy3hKcv0amehsIqwasrhodVpzRGdm++TlTI3b92NcVlj4=;
To: <internals@lists.php.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8;
 format=flowed
Content-Transfer-Encoding: 7bit
Date: Wed, 12 Jun 2019 02:31:55 +0100
In-Reply-To: <CAOwTYAuwwFBaaM-9rY_UTTWn=hz=JfMf9Kt-YVQZ2gQWk5PYhA@mail.gmail.com>
References: <CAOwTYAsLeE8MijU5b8BMTJiJ1r+bKa-g6tAqYyB52o+Zp3VQHw@mail.gmail.com>
 <CADyq6s+XtpjKgd0+DKEpT=AxiY31BFif-qyd7c5BiL1_kcFnRA@mail.gmail.com>
 <CAOwTYAsxrjCqWYEsj6aZkR0YY9a-DoNzXi4Q_x=NVWkyusnqGg@mail.gmail.com>
 <CAOwTYAucSEnOJexgSVsR6MAScHL1Diwh11QeODeWEReH5ZV2XA@mail.gmail.com>
 <CAPg3Xx+--Q+nZBdhiPcx7yuqS8NxQN3W0b7EtiedzE4soEq3TA@mail.gmail.com>
 <CAOwTYAuwwFBaaM-9rY_UTTWn=hz=JfMf9Kt-YVQZ2gQWk5PYhA@mail.gmail.com>
Message-ID: <3f454ccccc7103e5f513cfcd0f4f6e61@exussum.co.uk>
User-Agent: RoundCube Webmail/0.1
Subject: Re: [PHP-DEV] Using [ci skip]
From: scott@exussum.co.uk (Scott Dutton)

Hi Joe

Is there any reason why something like the following isnt run instead ?

runTests=$(git diff origin/master... --name-only --diff-filter=ACMR -- 
':(exclude)docs/*' | wc -l)

if [[ "${runTests}" -gt 0 ]]; then
     ./run-tests.sh
fi
~


Which says if the diff contains files outside of /docs then run the 
tests.

Contributors can then forget about the [ci-skip] and gets the same 
effect

This runs on linux but I cant imagine its a big change to make for this 
to work on windows ?

You can tweak the diff command to only run tests when certain sets of 
files change

Thanks

Scott

On 07.06.2019 15:07, Joe Watkins wrote:
> Hi Peter,
>
> I think it's pretty well known about, but don't think we have it 
> written
> down anywhere. I could make a note of it somewhere.
>
> I think in the case of azure it can be configured but I'm not sure 
> about
> the others. I'll try to find out.
>
> Cheers
> Joe
>
> On Fri, 7 Jun 2019 at 16:00, Peter Cowburn <petercowburn@gmail.com> 
> wrote:
>
>>
>>
>> On Fri, 7 Jun 2019 at 12:09, Joe Watkins <krakjoe@gmail.com> wrote:
>>
>>> Oh to be absolutely clear, I'm talking about commits that *only* 
>>> touch
>>> these non-source files ...
>>>
>>> Cheers
>>> Joe
>>>
>>> On Fri, 7 Jun 2019 at 13:07, Joe Watkins <krakjoe@gmail.com> wrote:
>>>
>>> > Hi Marco,
>>> >
>>> > It wasn't a topic for discussion, it was a request to committers 
>>> in
>>> > php-src.
>>> >
>>> > We do not need to run CI for NEWS changes, and we can definitely 
>>> be sure
>>> > it doesn't effect the build.
>>> >
>>> > The same goes for other files like UPGRADING, UPGRADING.INTERNALS 
>>> ...
>>> >
>>> > Under normal circumstances these files are not changed by 
>>> themslves, but
>>> > occasionally, we have to correct one of these files and omitting 
>>> [ci
>>> ski]
>>> > puts the build behind by up to an hour ...
>>> >
>>> > Cheers
>>> > Joe
>>> >
>>> > On Fri, 7 Jun 2019 at 13:02, Marco Pivetta <ocramius@gmail.com> 
>>> wrote:
>>> >
>>> >> Please avoid doing that:
>>> >>
>>> >>  1. Commit messages are for humans
>>> >>  2. You never know what can break, that's why it's "continuous" 
>>> there
>>> >> (besides religious views around what "continuous integration" 
>>> means)
>>> >>
>>> >> On Fri, Jun 7, 2019, 12:51 Joe Watkins <krakjoe@php.net> wrote:
>>> >>
>>> >> > Hi all,
>>> >> >
>>> >> > Just a friendly reminder that when we're committing changes to 
>>> files
>>> >> that
>>> >> > do not contain source, test code, or build configuration, it's
>>> helpful
>>> >> to
>>> >> > include [ci skip] in the commit message. Omitting it can put 
>>> our CI
>>> >> quite
>>> >> > far behind.
>>>
>>
>> Is this documented somewhere? I'm not seeing it in the docs held in
>> php-src, nor a search of the wiki, for example.
>> Also, is this not something that the CI application(s) can be 
>> configured
>> to do for us?
>>
>>
>>> >> >
>>> >> > Cheers
>>> >> > Joe
>>> >> >
>>> >>
>>> >
>>>
>>