[dev-servo] Should commits landed in servo individually pass all tests?

2017-11-02 Thread Manish Goregaokar
So I and emilio were discussing whether or not to squash https://github.com/servo/servo/pull/18750 and it seemed like we have different ideas of how "atomic" commits should be before landing. When we land pull requests in servo/servo we like individual commits to build so that git bisect works. D

Re: [dev-servo] Should commits landed in servo individually pass all tests?

2017-11-02 Thread Gregory Szorc
On Thu, Nov 2, 2017 at 5:00 PM, Manish Goregaokar wrote: > So I and emilio were discussing whether or not to squash > https://github.com/servo/servo/pull/18750 and it seemed like we have > different ideas of how "atomic" commits should be before landing. > > When we land pull requests in servo/se

Re: [dev-servo] Should commits landed in servo individually pass all tests?

2017-11-02 Thread Emilio Cobos Álvarez
Hi Manish, On 11/03/2017 01:00 AM, Manish Goregaokar wrote: > So I and emilio were discussing whether or not to squash > https://github.com/servo/servo/pull/18750 and it seemed like we have > different ideas of how "atomic" commits should be before landing. First of all thanks for landing that PR

Re: [dev-servo] Should commits landed in servo individually pass all tests?

2017-11-03 Thread Anthony Ramine
> Le 3 nov. 2017 à 01:00, Manish Goregaokar a écrit : > > So I and emilio were discussing whether or not to squash > https://github.com/servo/servo/pull/18750 and it seemed like we have > different ideas of how "atomic" commits should be before landing. It should definitely have been squashed,

Re: [dev-servo] Should commits landed in servo individually pass all tests?

2017-11-03 Thread Anthony Ramine
> Le 3 nov. 2017 à 02:10, Gregory Szorc a écrit : > > On Thu, Nov 2, 2017 at 5:00 PM, Manish Goregaokar > wrote: > >> > > The Firefox repo has the same dilemma. Ideally history is good down to the > commit level. In reality, we only test on "push heads" (commits that were > heads at time of

Re: [dev-servo] Should commits landed in servo individually pass all tests?

2017-11-03 Thread Anthony Ramine
Let me add some necessary additions to this quite rude email... I am sorry, Manish. > Le 3 nov. 2017 à 10:33, Anthony Ramine a écrit : > > >> Le 3 nov. 2017 à 01:00, Manish Goregaokar a écrit : >> >> So I and emilio were discussing whether or not to squash >> https://github.com/servo/servo/p

Re: [dev-servo] Should commits landed in servo individually pass all tests?

2017-11-03 Thread Manish Goregaokar
> "Always been the case" = I remember back in 2015 that reviewers often asked me whether intermediate commits were building properly. Maybe that was just out of curiosity, but to me it sounded like something we should strive for. To be clear; I'm talking about *passing all tests*, not building. It