On 10 February 2014 19:21, Andreas Färber <[email protected]> wrote: > Please remember to label the pull request [PULL 0/m] and to thread the > actual commits as [PULL n/m]. We had to adopt our scripts, too. > > Peter, please either enforce those rules or drop them for all of us!
My workflow for applying pull requests doesn't technically require this or check for it, and so I'm not going to "enforce" this, since I might well not notice (and in this case didn't). The rule is not for my benefit when applying pulls, but for people who might be CC'd on patches forming part of a pull request and who would otherwise not be able to tell whether they could reasonably ignore them. As usual we should: (a) document our best practice: http://wiki.qemu.org/Contribute/SubmitAPullRequest (b) politely nudge people if they have not followed it (and anybody who is adversely affected should feel free to do that; I don't have any particular greater standing here than you do); since this is pretty much always simply an occasional contributor who wasn't aware of something, a gentle nudge should be sufficient (c) be open to reconsidering our practices if on balance the burden turns out to be greater than the benefit. (In this case personally I think the rule is a useful one.) It's inevitably the case that some of our process isn't enforced by technical means and so there will be occasional accidental violations; we should in general assume good faith. I'm not going to punitively refuse to apply pull requests simply because they don't meet every detail of what we prefer to do (though of course I will and already have refused pulls which fail on significant points like lack of signoff or review). thanks -- PMM
