0

A developed piece of code goes these phases in our team:

  1. Developers write the code
  2. Then they send it to testers
  3. If tests pass, then testers send the change/new feature/bugfix to operations team for deployment.

Currently, we've adopted the following approach: We have a master branch from which we create one or more custom branches to make our new commits. Then we send a pull request to the release branch along with the task to test the newly developed code. If tests pass the testers merge the commits to release branch, create a pull request from release to master and open a deployment task. After the deployment, the changes are merged to master.

But there a problem here. We're never sure that the production runs the code that's in the master branch, because a human factor is involved. So next time when I want to create a branch from master I'll have to check if all the previous changes are reflected in it.

Can someone suggest a better flow? What are the best practices for this kind of situations?

P.S. I know automating the whole software development-deployment lifecycle would probably solve many problems, including this one, but currently we're a bit away from full automation.

1

Not sure if Feature Flags, Toggles, Controls is something that you have looked into. But it might be good to know about them, as they solve the branching / merging issue in a complete different way!

There is also a nice article Feature Toggles (aka Feature Flags), written by Martin Fowler.

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.