I see... that's too bad. Maybe if we had a clobber builder that only did compilation? -Darin
On Wed, Feb 11, 2009 at 9:00 PM, Evan Martin <[email protected]> wrote: > In this case, I believe some commits went through and got a green > build, but then a later commit was red in such a way that a clobber > revealed the older commit also should have caused a red build. The > clobber builder is so slow we would've needed to revert hours of > commits to go to the last green build from there. > > On Wed, Feb 11, 2009 at 8:48 PM, Darin Fisher <[email protected]> wrote: > > i'd be happy with a blind policy of reverting everything since the last > > green build whenever there is bustage. heck, we could probably automate > > that. then, the sheriff could coordinate re-landings perhaps? > > -darin > > > > On Wed, Feb 11, 2009 at 6:26 PM, Evan Martin <[email protected]> wrote: > >> > >> As best as we can figure out, I believe what happened was: > >> 1) Lei committed a change that passed the try servers. > >> 2) I committed a change that passed the try servers. > >> 3) Both of those changes were wrong, but only after a clobber build. > >> > >> It took a bit of time to piece out who to blame and revert etc. > >> > >> > >> > > > > > --~--~---------~--~----~------------~-------~--~----~ Chromium Developers mailing list: [email protected] View archives, change email options, or unsubscribe: http://groups.google.com/group/chromium-dev -~----------~----~----~----~------~----~------~--~---
