Travis DePrato
Travis DePrato
Sure, but why is that a concern here? It would be up to the implementing kernel to make sure the resulting messages have the correct `parentId` set in the IOPub...
I'd recommend starting here: https://www.safaribooksonline.com/videos/jupytercon-new-york/9781492025818/9781492025818 (this was a talk given at JupyterCon in 2018). This particular library is just a Python implementation of the low-level Jupyter protocol.
Not entirely sure how to reproduce because it's not easy to create a throwaway mac environment. It seems like `sys.prefix` doesn't match where packages actually get installed. I think it...
This might be fixed now - I was using 4.1. (related: #279)
Yes, but this should still be fixed. :) I'm currently just using the webpack workaround described above.
The fix I'm using for now is actually just setting the `data-slate-editor` attribute on the node that wraps the contenteditable so that the first check above fails (_Is the event...
I just ran into this. A temporary workaround is to `yarn unplug next`.
Update: `git bisect` has narrowed it down to https://github.com/google/gvisor/commit/ce91143bdc77ccdebf7aaae9ee8eb4843aa5f880 (cc @sshambhavi-252).
Incidentally, Julia doesn't try to do the same thing under runc. Not sure if it's detecting something different under gVisor.
> @travigd could you clarify what you mean by "same thing" here? As in it is not deleting /dev/shm/julia-codegen-1-846930886 and then calling ftruncate(2) with a size larger than the tmpfs's...