Kevin Bates

hi @jtpio. The format of a kernel_id _should be_ opaque. Since they have always been a UUID (at least in my understanding), it might be difficult to determine if that's...

Hmm - are you expecting things like /api/kernels (to get a list of running kernels) to return different results depending on which handler serviced that request? Or a DELETE /api/kernels/...

Could you please clarify? By "Indeed" are you suggesting that, yes, you do want to have the respective handlers filter on their respective "id types"? Or you are realizing that...

@jtpio - Thanks for the clarification - that all makes sense and seems reasonable to partition based on endpoints and instances. @echarles > There is no issue today if someone...

Sounds good Jeremy. Please note that even with kernel providers, you won't have the isolation that I understand you'd like unless splitting out the handlers *and* implementing your own MKM...

hi @SylvainCorlay. > Subshells open their own shell socket, and appear as a separate kernel in the UI on which you can connect any client. Wouldn't the subshell socket need...

ok. I think we need to understand and define how _similar_ and _different_ these kinds of "kernels" behave relative to standard kernels and feel they might deserve a different name...

> I think that whether it is managed by KernelManager and KernelProvisioner is an implementation detail of jupyter-server. This would be `jupyter_client` and believe these are details that need to...

Thanks @davidbrochart, this is really helpful. Before adding a row or two, I guess I'd like to better understand the user interactions with each option. Option 1 (_Kernel Sub-shells_) It...

Hi @hbcarlos, I'm not sure where this JEP stands but I'd like to try to revive it a bit. I think we should discuss some of the issues I raised...