On 25/02/12 00:27, Roan Kattouw wrote: > You're right that this is the biggest problem with stacked commits: if > you have a dependency chain A-B-C and B is amended or abandoned, you > have to rebase C somehow. A lesser problem is that if A and C are > approved, but B hasn't been reviewed yet, A will be merged but C won't > be (because it can't be merged without also merging B, and B has not > been approved yet). > > So yeah, we want to encourage people to use separate branches for > unrelated commits, so that B doesn't depend on A if A and B are > totally unrelated to each other. I've been trying to work that into > the various documentation pages, and Sumana let me put it in her git > introduction talk script too :) > > Roan
There's no way to treat a set of commits as a bundle? What happens if a developer wants to merge his extension on which he has been working (in Git) for months? I am assuming: * The extension will get a full review. * The author wants to keep the extension versioning (it could even be already published in eg. GihtHub). Will gerrit force it to spawn dozens of commit reviews? _______________________________________________ Wikitech-l mailing list [email protected] https://lists.wikimedia.org/mailman/listinfo/wikitech-l
