Open Stack

On Fri, Jul 25, 2014 at 07:35:52AM -0400, Sean Dague wrote:
> On 07/25/2014 07:17 AM, Daniel P. Berrange wrote:
> > On Fri, Jul 25, 2014 at 07:09:56AM -0400, Sean Dague wrote:
> >> On 07/25/2014 06:53 AM, Daniel P. Berrange wrote:
> >>> On Fri, Jul 25, 2014 at 06:38:29AM -0400, Sean Dague wrote:
> >>>> On 07/25/2014 01:18 AM, Ian Wienand wrote:
> >>>>> On 07/16/2014 11:15 PM, Alexis Lee wrote:
> >>>>>> What do you think about allowing some text after the words "recheck no
> >>>>>> bug"?
> >>>>>
> >>>>> I think this is a good idea; I am often away from a change for a bit,
> >>>>> something happens in-between and Jenkins fails it, but chasing it down
> >>>>> days later is fairly pointless given how fast things move.
> >>>>>
> >>>>> It would be nice if I could indicate "I thought about this".  In fact,
> >>>>> there might be an argument for *requiring* a reason
> >>>>>
> >>>>> I proposed [1] to allow this
> >>>>>
> >>>>> -i
> >>>>>
> >>>>> [1] https://review.openstack.org/#/c/109492/
> >>>>
> >>>> At the QA / Infra meetup we actually talked about the recheck syntax,
> >>>> and to change the way elastic recheck is interacting with the user.
> >>>>
> >>>> https://review.openstack.org/#/q/status:open+project:openstack-infra/elastic-recheck+branch:master+topic:erchanges,n,z
> >>>>
> >>>> and
> >>>>
> >>>> https://review.openstack.org/#/q/status:open+project:openstack-infra/config+branch:master+topic:er,n,z
> >>>>
> >>>> Are the result of that. Basically going forward we'll just support
> >>>>
> >>>> 'recheck.*'
> >>>
> >>> I'm not sure I understand what you mean by that ? Are we going to
> >>> use the literal string 'recheck.*' or do you mean we'll use 'recheck'
> >>> and the user can put arbitrary text after it ?
> >>
> >> Sorry, I think in regex. recheck + arbitrary string.
> > 
> > Would that still allow us to only trigger 3rd party CI ? eg if we do
> > 'recheck xenserver' I don't want to trigger the main CI, only the Xen
> > CI.
> 
> No, the 3rd party folks went off and created a grammar without
> discussing it with the infra team (also against specific objections to
> doing so). Such it is.

Whether or not we agree with the current syntax, it is *critical* to
maintain this ability to trigger only 3rd party CI systems, otherwise
the odds of being able to get a pass from all CI go down the toilet
even further than they already are. 

We must resolve this before introducing the new syntax

Regards,
Daniel
-- 
|: http://berrange.com      -o-    http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org              -o-             http://virt-manager.org :|
|: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org       -o-       http://live.gnome.org/gtk-vnc :|