Should this project be archived?

[elmiko]

This project is targeted at building an OpenShift native console for managing Spark clusters as distinct units from within the OpenShift console. As OpenShift continues to progress forward focusing on the operator pattern and custom resources for extension, these type of console projects seem less and less appropriate.

It appears to me that there is little pressure to bring this project into modernity (if that even applies in the current OpenShift world), and we are starting to accumulate security vulnerabilities against the code here. Given the state of progress on this, i would like to call for consensus regarding whether we should archive this project as a signal to the community that further development is frozen for the moment.

[willb]

@elmiko I guess my initial thought is that if we can team up and address the vulnerabilities now and deprecate for a few months before archiving, that would be ideal. OTOH, the writing is on the wall for this kind of project even if many community users are still not all-in on operators.

[elmiko]

@willb fixing the vulnerabilities is the easy part, imo, it's the testing that is becoming more difficult as time progresses. i think we could just update the vulnerable dependencies and then tag the version, but i'm more concerned about what a release would look like and also what version of openshift we would need to test against.

another concern that leads me to is that what if we find things that are broken in openshift 3.11 that previously worked. i don't think we should invest engineering time to make this work in 4 but do we fix it for other 3 versions? also there is a question about how much testing we can do to confirm that the fixes we are putting in are valid, it is becoming more difficult as knowledge and experience of older openshift versions decays.

[erikerlandson]

Any measure on the number of people who are using this?

[elmiko]

@erikerlandson no idea, but my guess is super low

edit actually, i would wager zero, maybe one if we count Chad ;)

[erikerlandson]

@elmiko, assuming that's true, then there is little to no upside from expending effort on sustaining engineering.

[tmckayus]

I think we can deprecate the console.
oshinko-web is maybe a different story