Once you get the conflicts resolved in the outstanding PRs, let’s ensure we have head of 3.x running on atlas-stg and evaluate it. I think you’ve got it pretty close to where we could deploy to production. Deploying to production will take several steps and some coordination with @cintiadr :
- Finishing with changes to 3.x, getting them deployed to atlas-stg, and verifying everything is working as expected.
- Make sure 3.x reflects what we want for the new Atlas master (e.g., make sure README is up to date, etc.)
- Replacing master branch with the 3.x branch. Given that we’re changing from PHP to node, we won’t be merging 3.x into master; rather, we’ll want 3.x to become the new master branch. A little googling suggests something like this is what we’d want:
git checkout 3.x git merge --strategy=ours --no-commit master git commit -m "atlas 3.1 - replacing php with node" git checkout master git merge 3.x git push - At this point, we should probably deploy master to atlas-stg (would expect atlas-stg to behave the same, but this is where atlas-stg should be building from most of the time anyway)
- Assuming all looks good, then you’d tag master as 3.1 and we’d deploy this to production as the new Atlas!