On Friday, May 16, 2014 6:45:44 PM UTC+3, Ryan VanderMeulen wrote:
> While working to track down various job backlogs and busted pushes in our CI 
> infrastructure, our team has observed some common anti-patterns in people's 
> TryServer usage that contribute to these problems. In order to try to help 
> developers find a balance between over-using and under-using Try, we have 
> collected these patterns into a wiki page that will hopefully clarify what is 
> recommended to ensure that a patch is less-likely to cause bustage while not 
> needlessly wasting machine resources:
> 
> 
> 
> https://wiki.mozilla.org/Sheriffing/How:To:Recommended_Try_Practices
> 
> 
> 
> Note that while the page above is designed to give general recommendations 
> for what to do, it is *NOT* a substitute for using your best judgement as a 
> developer. We just ask that you be conscientious about the impact your pushes 
> have on others, both from a resource utilization standpoint and a bustage 
> (i.e. tree closure) standpoint, and factor that into deciding what jobs to 
> run.
> 
> 
> 
> As always, if you have any questions, you can always feel free to ping myself 
> (RyanVM) or anybody else from our sheriffing team (edmorley, philor, KWierso, 
> and Tomcat) on IRC or via email with any questions and we'd be happy to 
> assist :)
> 
> 
> 
> Thanks!
> 
> 
> 
> -Ryan

I filed bug 1013415 (simpler try chooser with clearer implications). I think 
it's an important discussion.

- avih
_______________________________________________
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform

Reply via email to