On 03/08/16 18:34, David Woodhouse wrote: > On Tue, 2016-03-08 at 18:24 +0100, Laszlo Ersek wrote: >> Here again I can only point to people who I consider my betters -- are >> you suggesting that the QEMU workflow and the Linux workflow are utterly >> wrong? > > It is not "the Linux workflow". Linus will *eat* you if you rebase > trees which you ask him to pull. > > See http://www.mail-archive.com/[email protected]/msg39091.html
I don't know how to formulate this any clearer than I did before. It is not about the branch that linus pulls from the subsystem maintainer. It is about the patches that the subsystem maintainer picks up from emails of individual contributors. Let me quote Linus's email back at you: > Notice that this really is about other peoples _history_, not about > other peoples _code_. If they sent stuff to you as an emailed patch, > and you applied it with "git am -s", then it's their code, but it's > _your_ history. > > So you can go wild on the "git rebase" thing on it, even though you > didn't write the code, as long as the commit itself is your private > one. It is not surprising that Linus will satisfy his appetite for person-meat with a pull-requestor if said requestor rebases a branch pending a pull. The commit to be pulled could even completely disappear from the repo to pull from, if an aggressive git garbage collection occurs after the rebase. The branch ref moves to a different commit hash (due to the rebase), the old tip commit becomes dangling, and the gc reaps it. Understandable. But we're not talking about the handling of pull requests. We're talking about patches that contributors send in email. Most individual contributors send patches in email *only*. This covers Linux (to my knowledge), QEMU (definitely), libvirt (I believe), and edk2. Is your recommendation to *require* contirbutors to email pull requests, alongside their patches (which is "my" requirement)? I think that would go too far. Or do you recommend that contributors be *allowed* to email pull requests (alongside their patches), and if they do, their pull requests be merged correctly? I'm fine with this if the Intel leadership / edk2 maintainers reaches a consensus that we don't insist on a linear history any longer. Thanks Laszlo _______________________________________________ edk2-devel mailing list [email protected] https://lists.01.org/mailman/listinfo/edk2-devel
