Open Stack

Having lint checks as non-voting seems like a good compromise to me here.

2015-08-12 17:09 GMT+02:00 Yolanda Robla Mota <yolanda.robla-mota at hp.com>:

> Hi
> So my point here, is that i don't want to do a discrimination between
> rules on puppet
> modules and system-config. If we enforce lint rules, we should do it
> everywhere.
>
> For lint in system-config, once concern was that it affects history. But
> as long as we are moving
> more functionality to the modules, we have this problem there as well.
>
> I'd like that we have common lint rules, the same for system-config and
> for puppet modules.
> Maybe a first step is to add lint rules to system-config as non-voting and
> go gradually iterating
> on it?
>
> That's my take.
>
> El 12/08/15 a las 16:56, Paul Belanger escribió:
>
> Greetings,
>>
>> This email comes from some personal frustrations regarding our code review
>> policy for system-config. Specifically, lint or style checks for puppet
>> code.
>>
>> Back in Nov. 30 2012 (yes I checked) I enabled voting for
>> gate-ci-puppet-lint[1]. It was one of my first efforts for infra :) Since
>> then
>> we've grown to a large family of puppet contributors.
>>
>> However, system-config underwent a change, maybe a year ago, to remove
>> lint
>> checks.  My understanding of the reason to disable the check, was since
>> system-config was not being uploaded into the puppet forge, there was no
>> need to
>> have lint checks running.
>>
>> About 4-5 months ago, I asked to re-enable the lint check, but we denied.
>> Comments revolved around git blame / history issue and wasted effort. I
>> don't
>> have issue with this reasoning, if people don't want to do it, I don't
>> want to
>> force it.
>>
>> However, recently. I got my hand smacked in 2 different code reviews for
>> arrow
>> alignment issues. Honestly, I wasn't even mad about the -1 for the
>> alignment.
>> However, I'm concerned about the wasted effort the -1 caused me.
>> Basically, I
>> had to wait a few days to get the -1, since it was a human doing the
>> review, not
>> the gate. Additionally, if I was getting a -1 for style checks, why didn't
>> jenkins do it?
>>
>> So, my question is simple.  What is our policy on style checks for
>> system-config.  From what I understand, it goes both ways.  People don't
>> want
>> gate checksi (wasted effort), however people are doing human code review
>> for
>> style checks (because they like unified puppet modules). Needless to say,
>> this
>> is slightly confusing.
>>
>> All and all, I would rather jenkins give me a -1 if my code does not pass
>> style
>> over a human. Since, I can quickly run my tests locally before uploading
>> into
>> the gate.
>>
>> For the record, I want to re-enable the lint gate for system-config.
>> This keeps
>> it inline with 99% of our other openstack / openstack-infra puppet
>> modules.
>>
>> [1]
>> https://github.com/openstack-infra/project-config/commit/bee9131dce447d8dd53f246438fd3363a88da426
>>
>> _______________________________________________
>> OpenStack-Infra mailing list
>> OpenStack-Infra at lists.openstack.org
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-infra
>>
>
> --
> Yolanda Robla Mota
> Cloud Automation and Distribution Engineer
> +34 605641639
> yolanda.robla-mota at hp.com
>
>
>
> _______________________________________________
> OpenStack-Infra mailing list
> OpenStack-Infra at lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-infra
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-infra/attachments/20150812/63acc0c4/attachment-0001.html>