Settings

Theme

Resolve simple merge conflicts on GitHub

github.com

228 points by moby 9 years ago · 70 comments

Reader

pkamb 9 years ago

On my team I've found that it's incredibly useful to commit the merge conflicts and conflict markers, then immediately resolve the conflicts in the next commit. This gives you one commit that shows exactly how the two branches merged together, followed by a commit that shows exactly how the conflicts were resolved. The resolution commit can then be code reviewed independently for a nice clean view of the conflicts introduced in the merge. It also allows you to easily reset to the merge commit and resolve the conflicts differently.

The standard git workflow (and this github feature) seems to promote resolving the conflicts alongside all of the other changes in the merge working copy. This make me nervous, as there's no way to differentiate the new lines that were introduced to resolve merge conflicts from the thousands of lines of (previously reviewed) code from the feature branch.

If you're not careful, completely unrelated working-copy code and behavior can be introduced in a "merge commit" and neither you or any of your reviewers will notice. "Looks good to me."

  • kovrik 9 years ago

    Yes, that is a problem.

    On the other hand, with your approach you are going to have revisions that won't even build/compile.

    If you have automatic builds or/and unit/integration tests, then you'll have failed builds every time you have a merge conflict.

    Also, you are kind of 'polluting a well': what if meanwhile someone merges that revision into his/her branch? Or what if you have automatic merges configured?

    • phpnode 9 years ago

      Presumably in this model the dev wouldn't push their branch until the merge is actually complete, and there's presumably a convention like prepending `[CONFLICT]` to those commits to discourage people from checking them out directly.

    • sethammons 9 years ago

      And you just lost git bisect

      • paulddraper 9 years ago

        Not really; git bisect skip still works.

        I agree I don't like it either.

      • pkamb 9 years ago

        `git bisect skip`

      • jzwinck 9 years ago

        You can get it back by making your bisect test function return "good" whenever it sees a commit with merge conflicts.

        • robinson7d 9 years ago

          I don't think that would work, but correct me if I'm wrong.

          As far as I know, git bisect does a binary search along the commits; `good` tells it to look at the latter half, `bad` to look at the former.

          So suppose you have five commits (1,2,3,4,5), where 1 is the working state, and 3 is a conflict commit. It will start by asking about the middle commit (3), automatically choose `good`, and determine that 3 was the latest working commit (after checking 4, which says `bad`).

          ----

          EDIT: Obviously this is simplified to explain the issue with marking "good" those commits.

          • michaelmior 9 years ago

            This is resolved if you skip the commits with a conflict instead of marking as good.

        • dantiberian 9 years ago

          I think you'd want to skip, not mark as good?

        • masklinn 9 years ago

          > test function return "good" whenever it sees a commit with merge conflicts

          That's a terrible idea and would completely break bisect. You should signal "skip" (return code 125) if the commit can not be tested at all.

  • eridius 9 years ago

    You actually can distinguish the new lines. For any non-trivial merge conflict resolution committed as part of the merge, `git show $SHA` will actually show you the conflict resolution. More specifically, if the diff contains anything that's not just a line taken from either of the parents, then that thing is shown.

    • pkamb 9 years ago

      Yeah, I have no doubt that you can somehow show this information via the command line. The problem is that it's hidden in GitHub's Pull Request web UI, where all of our code review happens. Committing the conflicts and then resolving in the next commit surfaces the conflict resolutions to the PR where it can be reviewed like all of the other code we write.

      • joatmon-snoo 9 years ago

        If you don't rewrite commit history (it sounds like you don't) you can see it by looking at the diff of the latest commit on the PR.

        Also, I haven't used it extensively since its release, but doesn't the Reviews feature resolve this now?

      • deeplyoptional 9 years ago

        The PR UI does this. After resolving conflicts, the resulting merge commit will show just the resolution.

  • acchow 9 years ago

    Sounds like this would be a nightmare to rebase onto.

  • chrisamanse 9 years ago

    IMO, it is unnecessary to commit the conflicts. Instead, you can see how merges were resolved by diffing the commits.

    • pkamb 9 years ago

      "Diffing the commits" isn't really available in in a GitHub-style Pull Request web UI, which is where 99% of our code review is happening. I'm definitely optimizing for that view of the merge over everything else.

      • squidbidness 9 years ago

        I love how github fosters discovery and remote collaboration, though one of its liabilities is when great git command-line features are effectively lost unless github re-implements or exposes them, because some conventions incentivize only doing what github itself can do.

  • sjrd 9 years ago

    Or much simpler: do not allow merge conflicts to happen in the first place! Any Pull Request that shows with merge conflicts must be rebased on top of the target branch. Problem solved.

  • nothrabannosir 9 years ago

    Great idea! Although this does break cherry-pick, doesn't it?

    • btown 9 years ago

      I'd imagine this also breaks bisect (and might make your CI system very confused), since you have a non-good commit.

      • azernik 9 years ago

        Bisect knows how to deal with this; you can tell it to ignore a commit that you know is broken for reasons unrelated to the issue upper investigating (and try adjacent ones instead).

    • aargh_aargh 9 years ago

      Why would it? You can get a conflict on a cherry-pick as well.

      • nothrabannosir 9 years ago

        I meant to write bisect---I was in the middle of an actual git workflow and accidentally wrote this instead. :)

  • fogleman 9 years ago

    That's a really good idea! I'm going to start doing that.

messutied 9 years ago

So simple, so useful, I wonder if this feature wasn't already in Gitlab since it seems to be more full featured

orivej 9 years ago

diff3 conflict style display would be considerably more useful.

rosstex 9 years ago

Finally! This is excellent news.

jklein11 9 years ago

To me this feels like making a commit without unit testing first. When I find a conflict I like to be able to resolve it and then do some unit testing to make sure that my revision didn't miss anything.

  • Ph0X 9 years ago

    I think they explicitly say "simple merge conflicts" in the title. At the end of the day, you should use your own best judgement for when this is useful, and for when you need to go back to your workspace. It's most definitely not meant to be used for every merge conflict.

    But not everyone is working on big projects with tests, and not every merge conflict is actually complex code modification.

    Sometimes it's just two commits adding something at the end of the file and there's not real conflict, or maybe you modified the same line twice and forgot to pull before doing your 2nd edit.

  • swampthing 9 years ago

    I suspect the use-case they have in mind are folks who have CI hooked up to Github (so using this feature will automatically trigger tests).

    • euyyn 9 years ago

      Sure, but that's a much slower cycle than running your unit tests locally.

mojuba 9 years ago

I didn't know I could merge on github.com in the first place... where is their merge function, by the way?

  • aargh_aargh 9 years ago

    It's literally the hallmark of GitHub... How else would pull requests work? (well, there _is_ the rebase option now...)

    • mojuba 9 years ago

      Hmm. I usually do merges locally as serious stuff should be built and tested before pushing anyway, so probably why never used GitHub's hosted functions.

      • bjacobel 9 years ago

        > serious stuff should be built and tested before pushing anyway

        Yes and no. Build in your CI server that's set up to mirror your prod environment after pushing, but before merging. That's what the whole industry of CI providers and integrations built into and around GitHub and GitLab is for.

        • stormbrew 9 years ago

          To be fair, one of the annoying things about how PRs work is that they don't test the merge, they test the commit relative to its original base. Your tests may pass in the PR, but fail once applied to later changes in the main line.

          • pavel_lishin 9 years ago

            Most CI solutions will take care of that; they'll pull in the PR, merge it locally with master, and run tests on that.

            • x1798DE 9 years ago

              I can't speak to the "merge-and-run" behavior of CI, but in my experience, most of these merge conflicts arise because of a time delay between making the PR (at which point CI is run) and merging the PR. It would be quite resource intensive to re-build every PR against every new branch in master.

              • pavel_lishin 9 years ago

                > It would be quite resource intensive to re-build every PR against every new branch in master.

                Depends on just how resource intensive a build is, and how often commits are made to master.

                Here at work, we don't do that, but it wouldn't be a complete clusterfuck if we were to.

          • Zarel 9 years ago

            In my experience, Travis CI tests the merge at the time the pullreq is created. It also has a "re-run" button that lets you merge the latest master and run the tests on top of that.

        • steveklabnik 9 years ago

          And depending on setups, testing locally is testing far fewer environments than CI. I only have one laptop, but many of my projects need at least OS X, Linux, and Windows, and often multiple versions of those.

      • aargh_aargh 9 years ago

        If you write tests diligently, you can integrate CI with GitHub. Travis CI might be the most popular option.

Keyboard Shortcuts

j
Next item
k
Previous item
o / Enter
Open selected item
?
Show this help
Esc
Close modal / clear selection