-
Notifications
You must be signed in to change notification settings - Fork 8
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Versioning Proposal #398
Versioning Proposal #398
Conversation
Drive by comment: Thanks for adding these actions!
Doesn't this cause problems with PRs? e.g. opening a PR will default to being against
Propose having an action to bump the version (from |
Ok, so I think I got the auto-version bumper working (hard to tell without merging this, but it was successfully bumping the version every commit before I switched that to only happen on It's also set to auto-generate a tag for minor and major releases, and to do minor releases on any commit on |
I guess it depends on if we're prioritizing user-experience, so that by default people get a more stable version, or dev experience, so we don't have to do some git history magic if we mess up. Doesn't look like github has a way to set a PR-only default branch, unfortunately. Just a thought while I was working on things that distinguish a "release" before we're on CRAN. |
We can get a good dev experience and a good user experience by keeping |
@nmdefries @dsweber2 this is addressed now right (we're only bumping on merge to dev)? |
Can we / should we automate merging back to dev from main? That will probably trigger a version bump in dev, but I'm fine with always starting on |
Is addressed by the fact that the version bump applies at merge time on the
Oh, yeah this is probably a good idea; I guess we can just copy the action from a different repo? |
There's this. Having to manually merge the sync PR is minorly annoying, but figuring out how to get GH Actions to auto merge from main to dev might be even more annoying. |
56f9517
to
f4cce8f
Compare
title: "chore: sync main->dev" | ||
labels: chore | ||
body: | | ||
Syncing Main->Dev. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
issue: should auto-tag someone so the PR doesn't get lost. Example.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
suggestion (nonblocking): could be useful to have a "Release" workflow (and button) to merge dev
to main
.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
oh hey, that's handy. Given that all that happens is a version bump I think auto-merge should be the better option.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Not sure the release workflow is worth the effort to actually get it set up, given that that one is specific to python code, but good to know it exists
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Agree about maybe not worth the effort, but that release workflow is for R -- python is just used to do the bumping. (forecast-eval
doesn't have any python code.)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good!
This implements the github actions to make
0.<increment on release>.<increment on PR>
usable by making the dev docs on thedev
branch, and the release docs on themain
branch. Along with this, I suggest that we makemain
the default branch, so that people have to opt-in to the dev version, rather than opt-out of it.It also introduces a semver check to make sure that any PR includes a bump to the semantic version. Hypothetically this should fail on this PR for the first commit, to be fixed after confirming this (if only there were a way to test gh actions).