Upgrade to JupyterLab 3.0

[GeorgianaElena]

Fixes #81

TODOs

• Figure out what to do about qgrid
• Build image
• Deploy and test on staging

Unfortunately I don't think qgrid is compatible with JupyterLab3, but there's an open PR for this: quantopian/qgrid#356
@yuvipanda, do you know if it's actively used by users? I found two references about it:

• Add qgrid, jupyter themes to image #42
• Investigate snippets nbextension #41

[yuvipanda]

@GeorgianaElena if qgrid will still work on classic notebook, let's just document that and go ahead.

[damianavila]

AFAIK, Quantopian does not exist anymore and I am not sure if qgrid is being maintained recently by others... so I agree with @yuvipanda.

[GeorgianaElena]

Thanks @yuvipanda and @damianavila! It all seems to be working (qgrid only with classical notebook).

I will deploy this to prod tomorrow so that users don't get interrupted and presented with a new interface on a Friday :D
Also, hopefully, we will get some feedback over the weekend if I missed anything and there's something not working ok so that everything it's working by Monday.

[damianavila]

I will deploy this to prod tomorrow so that users don't get interrupted and presented with a new interface on a Friday :D
Also, hopefully, we will get some feedback over the weekend if I missed anything and there's something not working ok so that everything it's working by Monday.

Maybe, in the future, we should think about ANN these sort of big changes? IIRC, there was something written in the ecosystem to make wide ANN in a Jupyterhub sort of setup... let me find it... ok, here it is: https://github.com/rcthomas/jupyterhub-announcement

Again, just something to think about for the future (not asking to use it now 😉 ).

[GeorgianaElena]

Maybe, in the future, we should think about ANN these sort of big changes? IIRC, there was something written in the ecosystem to make wide ANN in a Jupyterhub sort of setup... let me find it... ok, here it is: https://github.com/rcthomas/jupyterhub-announcement

Thanks for sharing @damianavila! This seems like a pretty cool service to have.

Unfortunately, I failed to deploy this over the weekend :( However, @damianavila's message got me thinking if we should instead coordinate with the UToronto folks before and schedule this deployment, similar to what happened when the hub version was bumped. @yuvipanda, what do you think?