Conflicts

Overview

Teaching: 15 min
Exercises: 0 min

Questions

• What do I do when my changes conflict with someone else’s?

Objectives

• Explain what conflicts are and when they can occur.

• Resolve conflicts resulting from a merge.

• Understand how to create a pull request.

As soon as people can work in parallel, they’ll likely step on each other’s toes. This will even happen with a single person: if we are working on a piece of software on both our laptop and a server in the lab, we could make different changes to each copy. Version control helps us manage these conflicts by giving us tools to resolve overlapping changes.

We already created a merge conflict during the exercise in the last section. Just in case we want to create another we can do the following. Do the following on a collaborators copy. This could be a second clone of the repository, another person’s computer or via the Github web interface.

$ cat places.csv

name,symbol,creator,comments,lon,lat
Sophie's,cafe,Colin Sauze,good cooked breakfasts,-4.08225,52.415250

Let’s change a line to the collaborator’s copy only:

$ nano places.csv
$ cat places.csv

name,symbol,creator,comments,lon,lat
Sophie's,cafe,Colin Sauze,good cooked breakfasts and burgers,-4.08225,52.415250

and then commit the change:

$ git add places.csv
$ git commit -m "Add a mention of burgers"

[main 5ae9631] Add a mention of burgers
 1 file changed, 1 insertion(+)

and push it to github:

$ git push origin main

Enumerating objects: 5, done.
Counting objects: 100% (5/5), done.
Delta compression using up to 8 threads
Compressing objects: 100% (3/3), done.
Writing objects: 100% (3/3), 331 bytes | 331.00 KiB/s, done.
Total 3 (delta 2), reused 0 (delta 0)
remote: Resolving deltas: 100% (2/2), completed with 2 local objects.
To git@github.com:NOC-OI/favourite-places.git
   29aba7c..dabb4c8  main -> main

Now let’s have the original owner make a different change to their copy without pulling from GitHub:

$ nano places.csv
$ cat places.csv

name,symbol,creator,comments,lon,lat
Sophie's,cafe,Colin Sauze,good cooked breakfasts and milkshakes,-4.08225,52.415250

We can commit the change locally:

$ git add places.csv
$ git commit -m "Mention how good the milkshakes are"

[main 07ebc69] Mention how good the milkshakes are
 1 file changed, 1 insertion(+)

but Git won’t let us push it to GitHub:

$ git push origin main

To git@github.com:NOC-OI/favourite-places.git
 ! [rejected]        main -> main (fetch first)
error: failed to push some refs to 'git@github.com:NOC-OI/favourite-places.git'
hint: Updates were rejected because the remote contains work that you do
hint: not have locally. This is usually caused by another repository pushing
hint: to the same ref. You may want to first integrate the remote changes
hint: (e.g., 'git pull ...') before pushing again.
hint: See the 'Note about fast-forwards' in 'git push --help' for details.

Git rejects the push because it detects that the remote repository has new updates that have not been incorporated into the local branch. What we have to do is pull the changes from GitHub, merge them into the copy we’re currently working in, and then push that. Let’s start by pulling:

$ git pull origin main

remote: Enumerating objects: 5, done.
remote: Counting objects: 100% (5/5), done.
remote: Compressing objects: 100% (1/1), done.
remote: Total 3 (delta 2), reused 3 (delta 2), pack-reused 0
Unpacking objects: 100% (3/3), done.
From git@github.com:NOC-OI/favourite-places.git
 * branch            main     -> FETCH_HEAD
    29aba7c..dabb4c8  main     -> origin/main
Auto-merging places.csv
CONFLICT (content): Merge conflict in places.csv
Automatic merge failed; fix conflicts and then commit the result.

The git pull command updates the local repository to include those changes already included in the remote repository. After the changes from remote branch have been fetched, Git detects that changes made to the local copy overlap with those made to the remote repository, and therefore refuses to merge the two versions to stop us from trampling on our previous work. The conflict is marked in in the affected file:

$ cat places.csv

<<<<<<< HEAD
Sophie's,cafe,Colin Sauze,good cooked breakfasts and milkshakes,-4.08225,52.415250
=======
Sophie's,cafe,Colin Sauze,good cooked breakfasts and burgers,-4.08225,52.415250
>>>>>>> dabb4c8c450e8475aee9b14b4383acc99f42af1d

Our change is preceded by <<<<<<< HEAD. Git has then inserted ======= as a separator between the conflicting changes and marked the end of the content downloaded from GitHub with >>>>>>>. (The string of letters and digits after that marker identifies the commit we’ve just downloaded.)

It is now up to us to edit this file to remove these markers and reconcile the changes. We can do anything we want: keep the change made in the local repository, keep the change made in the remote repository, write something new to replace both, or get rid of the change entirely. Let’s replace both so that the file looks like this:

$ cat places.csv

Sophie's,cafe,Colin Sauze,good cooked breakfasts, milkshakes and burgers,-4.08225,52.415250

To finish merging, we add places.csv to the changes being made by the merge and then commit:

$ git add places.csv
$ git status

On branch main
All conflicts fixed but you are still merging.
  (use "git commit" to conclude merge)

Changes to be committed:

	modified:   places.csv

$ git commit -m "Merge changes from GitHub"

[main 2abf2b1] Merge changes from GitHub

Now we can push our changes to GitHub:

$ git push origin main

Enumerating objects: 10, done.
Counting objects: 100% (10/10), done.
Delta compression using up to 8 threads
Compressing objects: 100% (6/6), done.
Writing objects: 100% (6/6), 645 bytes | 645.00 KiB/s, done.
Total 6 (delta 4), reused 0 (delta 0)
remote: Resolving deltas: 100% (4/4), completed with 2 local objects.
To git@github.com:NOC-OI/favourite-places.git
   dabb4c8..2abf2b1  main -> main

Git keeps track of what we’ve merged with what, so we don’t have to fix things by hand again when the collaborator who made the first change pulls again:

$ git pull origin main

remote: Enumerating objects: 10, done.
remote: Counting objects: 100% (10/10), done.
remote: Compressing objects: 100% (2/2), done.
remote: Total 6 (delta 4), reused 6 (delta 4), pack-reused 0
Unpacking objects: 100% (6/6), done.
From git@github.com:NOC-OI/favourite-places.git
 * branch            main     -> FETCH_HEAD
    dabb4c8..2abf2b1  main     -> origin/main
Updating dabb4c8..2abf2b1
Fast-forward
 places.csv | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

We get the merged file:

$ cat places.csv

Sophie's,cafe,Colin Sauze,good cooked breakfasts, milkshakes and burgers,-4.08225,52.415250

We don’t need to merge again because Git knows someone has already done that.

Git’s ability to resolve conflicts is very useful, but conflict resolution costs time and effort, and can introduce errors if conflicts are not resolved correctly. If you find yourself resolving a lot of conflicts in a project, consider these technical approaches to reducing them:

• Pull from upstream more frequently, especially before starting new work
• Use topic branches to segregate work, merging to main when complete
• Make smaller more atomic commits
• Where logically appropriate, break large files into smaller ones so that it is less likely that two authors will alter the same file simultaneously

Conflicts can also be minimized with project management strategies:

• Clarify who is responsible for what areas with your collaborators
• Discuss what order tasks should be carried out in with your collaborators so that tasks expected to change the same lines won’t be worked on simultaneously
• If the conflicts are stylistic churn (e.g. tabs vs. spaces), establish a project convention that is governing and use code style tools (e.g. htmltidy, perltidy, rubocop, etc.) to enforce, if necessary

Solving Conflicts that You Created

In the group exercise in the last section you probably created a merge conflict. If you didn’t, make another change to your copy of places.csv to cause one. Rembmer to add, commit and push the change, the push stage should give a merge error. Now attempt to resolve the conflict, you might have to do this multiple times to deal with everyone’s changes. Ask the instructor and helper(s) to help resolve this. It can help if each person takes it in turn to resolve the conflict. Expect total chaos!!

Conflicts on Non-textual files

What does Git do when there is a conflict in an image or some other non-textual file that is stored in version control?

Solution

Let’s try it. Suppose Dracula takes a picture of Martian surface and calls it mars.jpg.

If you do not have an image file of Mars available, you can create a dummy binary file like this:

$ head -c 1024 /dev/urandom > mars.jpg
$ ls -lh mars.jpg

-rw-r--r-- 1 vlad 57095 1.0K Mar  8 20:24 mars.jpg

ls shows us that this created a 1-kilobyte file. It is full of random bytes read from the special file, /dev/urandom.

Now, suppose Dracula adds mars.jpg to his repository:

$ git add mars.jpg
$ git commit -m "Add picture of Martian surface"

[main 8e4115c] Add picture of Martian surface
 1 file changed, 0 insertions(+), 0 deletions(-)
 create mode 100644 mars.jpg

Suppose that Wolfman has added a similar picture in the meantime. His is a picture of the Martian sky, but it is also called mars.jpg. When Dracula tries to push, he gets a familiar message:

$ git push origin main

To https://github.com/vlad/planets.git
 ! [rejected]        main -> main (fetch first)
error: failed to push some refs to 'https://github.com/vlad/planets.git'
hint: Updates were rejected because the remote contains work that you do
hint: not have locally. This is usually caused by another repository pushing
hint: to the same ref. You may want to first integrate the remote changes
hint: (e.g., 'git pull ...') before pushing again.
hint: See the 'Note about fast-forwards' in 'git push --help' for details.

We’ve learned that we must pull first and resolve any conflicts:

$ git pull origin main

When there is a conflict on an image or other binary file, git prints a message like this:

$ git pull origin main
remote: Counting objects: 3, done.
remote: Compressing objects: 100% (3/3), done.
remote: Total 3 (delta 0), reused 0 (delta 0)
Unpacking objects: 100% (3/3), done.
From https://github.com/vlad/planets.git
 * branch            main     -> FETCH_HEAD
   6a67967..439dc8c  main     -> origin/main
warning: Cannot merge binary files: mars.jpg (HEAD vs. 439dc8c08869c342438f6dc4a2b615b05b93c76e)
Auto-merging mars.jpg
CONFLICT (add/add): Merge conflict in mars.jpg
Automatic merge failed; fix conflicts and then commit the result.

The conflict message here is mostly the same as it was for places.csv, but there is one key additional line:

warning: Cannot merge binary files: mars.jpg (HEAD vs. 439dc8c08869c342438f6dc4a2b615b05b93c76e)

Git cannot automatically insert conflict markers into an image as it does for text files. So, instead of editing the image file, we must check out the version we want to keep. Then we can add and commit this version.

On the key line above, Git has conveniently given us commit identifiers for the two versions of mars.jpg. Our version is HEAD, and Wolfman’s version is 439dc8c0.... If we want to use our version, we can use git checkout:

$ git checkout HEAD mars.jpg
$ git add mars.jpg
$ git commit -m "Use image of surface instead of sky"

[main 21032c3] Use image of surface instead of sky

If instead we want to use Wolfman’s version, we can use git checkout with Wolfman’s commit identifier, 439dc8c0:

$ git checkout 439dc8c0 mars.jpg
$ git add mars.jpg
$ git commit -m "Use image of sky instead of surface"

[main da21b34] Use image of sky instead of surface

We can also keep both images. The catch is that we cannot keep them under the same name. But, we can check out each version in succession and rename it, then add the renamed versions. First, check out each image and rename it:

$ git checkout HEAD mars.jpg
$ git mv mars.jpg mars-surface.jpg
$ git checkout 439dc8c0 mars.jpg
$ mv mars.jpg mars-sky.jpg

Then, remove the old mars.jpg and add the two new files:

$ git rm mars.jpg
$ git add mars-surface.jpg
$ git add mars-sky.jpg
$ git commit -m "Use two images: surface and sky"

[main 94ae08c] Use two images: surface and sky
 2 files changed, 0 insertions(+), 0 deletions(-)
 create mode 100644 mars-sky.jpg
 rename mars.jpg => mars-surface.jpg (100%)

Now both images of Mars are checked into the repository, and mars.jpg no longer exists.

Pull Requests

As you have seen, multiple people working on the same repository can cause a lot of problems. Github provides another way to collaborate called pull requests. In these each person makes their own copy of the repository on Github, this is known as a fork. Instead of working in the same repository as everyone else, each person works on their own fork. When they are ready to send the changes back to the main repository they create a pull request (literally asking the owner of the repository to do a git pull from theirs). Github wraps this in a nice interface, which allows you to write a comment about what you changed, for the owner to review your changes (and possibly request you make more before they accept it) and for them to finally accept or reject them. Once they are accepted the changes from your fork are merged into the upstream repository. If there are any conflicts then it will be up to the owner of the upstream repository to resolve them.

Pull Requesting Changes

• Create a fork of the favourite-places repository by visiting https://github.com/NOC-OI/favourite-places and clicking on the fork link near the top right hand corner.
• Create an additional change to the places.csv file in the repository.
• Github should now tell you that your fork is “1 commit ahead” of the upstream repository and offer a “Contribute” button to start the pull request.
• Click this and choose “Open Pull Request”.
• The next screen will highlight the differences between your version and the upstream one. Go ahead and click “Create pull request”.
• The repository owner should now get an alert about your pull request and can choose whether to merge it or not.

When do you think it is best to use pull requests and when is it best to allow multiple people to write to the same repository?

Solution

Pull requests are best for larger projects where you want to control or at least review any changes. This does require the project’s owners to spend time reviewing any pull requests. Smaller projects where all the developers trust each other might be better suited to allowing multiple people direct write access.

Making additional changes to a Pull Request

• Create another pull request using the same method as above.
• After submitting the pull request, you decide to make another change either because you found a mistake or the upstream repository owner asked you to fix something.
• Make this change to your fork and push them to github.
• What happens to the pull request?

Solution

The pull request should automatically update with any additional changes you make.

Pull Requests in Big Projects

In many large projects all work will be done via pull requests and merging changes back into the main repository will involve a large scale code review process. There may also be autoamted checks involved to stop code which doesn’t pass tests (or doesn’t have tests) from being merged into a production system.

Key Points

• Conflicts occur when two or more people change the same lines of the same file.

• The version control system does not allow people to overwrite each other’s changes blindly, but highlights conflicts so that they can be resolved.

• Pull requests are a better way to manage several people working on the same files at the same time.