-
Notifications
You must be signed in to change notification settings - Fork 697
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
Cut 1.2.0 tag and release a stable version of master #1321
Comments
/cc @kubeflow/wg-training-leads |
SGTM. |
SGTM. |
Should we maintain a branch for it? |
That's a good idea. How about rebase all master changes to v1.1-branch? @gaocegege Seems v1.1-branch was created long time ago to align with Kubeflow release. I think we have two options.
@gaocegege Please help on that if it sounds good to you. |
If we plan to cut a new branch. we don't need to maintain two test pipelines. That save me a lot of time. I will file a PR to enable test and images build for v1.1-branch later once it's created |
SGTM |
Update: this is not valid anymore. we plan to use v1.2-branch since there're some features inside |
My understanding is that, you've got some features changes after v1.1. Then following the complexities for merging to v1.1 branch, I think the easiest solution is just to release v1.2 and create a new v1.2-branch. WDUT? |
Sounds good. Let me cut branch |
We have cut v1.2-branch to host legacy tensorflow controllers and enable test for any new coming bug fixes. alpha version of universal training operator has been released https://github.com/kubeflow/tf-operator/releases/tag/v1.3.0-alpha.2. We can close this issue. |
I think it's a good time to consider merging all-in-one-operator back to master branch and keep evolving the development.
I would suggest to cut a 1.1.1 tag or 1.2.0 tag from master. This would be the final version of original tf operator. Just for backup purpose.
v1.1.0...master
Seems not too many feature changes since 1.1.0. I think 1.1.1 is better than 1.2.0?We determine to use 1.2.0 because there're few new features
rebase merge all-in-one-operator to master. I cut one for review
All in one operator rebase merge to master #1320
prepare for alpha operator release from master.
Manifest, test pipelines, etc Universal Operator Phase III: Graduate operator to production grade #1318
What do you think? @kubeflow/wg-training-leads
I am personally ok with any plans. It's fine for us to have more work done in dev branch but it may takes more time and it will also block changes to master (if there's no conflicts, It's ok to merge)
The text was updated successfully, but these errors were encountered: