We have run into an issue where automatically fixing issues in the main repository will incorrectly resolve issues in all forked repositories when they are next synced.
Info on automatically resolving issues on Bitbucket: https://confluence.atlassian.com/bitbucket/resolve-issues-automatically-when-users-push-code-221451126.html
An easy way to test for this issue is to:
- Create a new repository with some files.
- Fork it.
- Now in the main repository...
- Create an issue #1 with some random details.
- Make some change to a file and add the message 'Fixed #1' to the commit.
- Stage the file and commit the change, then push back up to Bitbucket.
- In the forked repository...
- Create an issue #1 with some other random details.
- Sync the repository to pull in the changes from main.
- Note that I'm actually doing this via command line client side using fetch and merge, but it should be the same and you have no choice if it's going to run into a conflict.
- Issue #1 is now resolved in both the main repo and the fork, even though issue #1 in the fork could actually be something completely different!
So onto my question, is there a way around this issue OR do we just not use the automatic resolve feature in commit messages on the main repository? (meaning that we always manually resolve issues with the resolve button)
