On Wed, May 4, 2016 at 12:26 PM, zimbatm <[email protected]> wrote: > Each contributor has his own motivations and every round were we provide > feedback is another one where we might lose the contributor. He might run > out of energy, or have moved onto other things. > > Even after improving the CONTRIBUTING.md, naming of commits is still a > really big friction to getting valid code into nixpkgs. I'm talking of how > commits should be named after a patter like "package: init at x.y.z" or > "package: a.b.c -> x.y.z". > > I must admit I don't really know the motivations behind this rule. All I can > think of it that we could theoretically build some tooling and get pretty > cool stats out of it. And that spelunking git history becomes a tiny bit > easier. > > Given all that I think we should reconsider that rule. In my opinion if a > contributor submits valid nix code that is useful to the project we should > just be able to merge it and move forward. For me it's more important than > the commit naming rule.
I don't think it is necessary to throw out that naming convention to keep new contributors happy. The person with commit rights who is performing the merge can --amend the commit to keep it in line with naming conventions just before they merge. Commit messages are easy to change, so it doesn't need to be a back-and-forth review process if there is a prescribed format. Louis _______________________________________________ nix-dev mailing list [email protected] http://lists.science.uu.nl/mailman/listinfo/nix-dev
