kubeflow / pipelines

Machine Learning Pipelines for Kubeflow
https://www.kubeflow.org/docs/components/pipelines/
Apache License 2.0
3.6k stars 1.62k forks source link

[feature] remove long-deprecated python based visualizations #11133

Open gregsheremeta opened 2 months ago

gregsheremeta commented 2 months ago

What feature would you like to see?

As far as I can tell, there are two primary categories of visualization in Kubeflow Pipelines:

Output viewer visualizations are the original visualizations that were included in Kubeflow Pipelines long ago.

In 2019, Kirin Patel introduced "Python based visualizations" to Kubeflow Pipelines (commits, blog post).

Note the introduction of the "visualization server".

In 2020, Kirin had left the project, and the Kubeflow Pipelines community decided to deprecate Python based visualizations. issue PR

The recommendation became to use notebooks or the HTML output viewer to accomplish what you could have using the "Python based visualizations" feature.

Post deprecation, there was a lot of language left over in the output viewer documentation that described "Python based visualizations" as the "new" way to do visualizations. This was confusing. I've previously posted a PR to clean some of that up, but much lingers.

It's been 4 years since it was deprecated. It probably should have been removed when Kubeflow Pipelines 2.0 was released.

Let's remove it now.

gregsheremeta commented 2 months ago

One thing I'm not sure how to do: how does Kubeflow / KFP handle cleaning up deprecated and deleted things? Should we include some kind of upgrade scripts of manifests that are capable of cleaning up Deployments, Roles / RBAC, SAs, etc.?

github-actions[bot] commented 1 week ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

gregsheremeta commented 1 week ago

/lifecycle frozen