Hi There,
I work for a software company based in Australia and I’m currently in charge of managing releases for one of the software offerings at this company. In terms of Git, we have two permanent branches: master and develop and our documented release process is as follows: 1. One week before the date of a release we take a snapshot of the develop branch and this snapshot is deployed to a test environment for final testing before the release. The branch is named according to internal conventions as e.g. release/3.5.0. 2. Fixes for any bugs found by the test team while testing this snapshot are committed directly to the release branch (e.g. release/3.5.0). 3. Once the testers have signed-off on the release, two PRs are created in Azure DevOps to merge the release branch to both master and develop. So, for example, PR #1 has a source branch of release/3.5.0 and a target branch of master, and PR #2 has a source branch of release/3.5.0 and a target branch of develop. The idea is that any changes committed to the release branch are merged into both master and develop with the aim of maintaining parity between develop and master. 4. master is then deployed to our production environment. The issue is that sometimes there are code differences between master and develop post-release, even though there are no file changes detected when test PRs are created with the following setup: 1. Source branch: master, target branch: develop. - No file differences 2. Source branch: release/3.5.0 target branch: master. - No file differences 3. Source branch: release/3.5.0, target branch: develop. - No file differences If, however, I create a PR with a source branch: develop and target branch: master, the file differences are there (these are file differences related to changes made In the release, not new changes submitted by our development team post-release). This happens when a PR is reverted in the release branch during the week of testing, but there seem to be some other non-revert-related situations where this occurs as well. I’d like to know why this occurs and how I can prevent it from happening in the future. I’d also like to know if you have any best practice recommendations as our process loosely follows the GitFlow model which has fallen into disuse, but our test automation also doesn’t have sufficient coverage at this point in time for us to simply convert to a fully CI/CD model. Any advice anyone can give would be *hugely* appreciated. -- You received this message because you are subscribed to the Google Groups "Git for human beings" group. To unsubscribe from this group and stop receiving emails from it, send an email to git-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/git-users/8830ceab-64cb-49e9-84b7-c26a923d4907n%40googlegroups.com.
