3 ms·
I did this! I killed it and I didn't mean to. Ten (10) days ago, I filed an issue in the iodide project: "Compatibility with 'percent' notebook format" https:
by westurner 6y ago
I did this! I killed it and I didn't mean to.
Ten (10) days ago, I filed an issue in the iodide project: "Compatibility with 'percent' notebook format"
https://github.com/iodide-project/iodide/issues/2942 https://github.com/iodide-project/iodide/issues/2942
And then six (6) days ago, I added this comment to that issue: https://github.com/iodide-project/iodide/issues/2942#issuecomment-697656505 https://github.com/iodide-project/iodide/issues/2942#issueco...
And now it's almost dead, and I didn't mean to kill it.
But I also suggested that it would be great if conda-forge had a WASM build target:
- "Consider moving CPython patches upstream" https://github.com/iodide-project/pyodide/issues/635#issuecomment-699271731 https://github.com/iodide-project/pyodide/issues/635#issueco...
For students, being able to go to a URL and have a notebook interface with the SciPy stack preinstalled without needing to have an organization manage shell accounts and/or e.g. JupyterHub for every student should be worth the necessary budget allocation. Their local machines have plenty of CPU, storage, and memory for all but big data workloads.
Iodide is/was really cool. Pyiodide (much of the SciPy stack compiled to WASM) is also a great idea.
Jyve with latest JupyterLab, nbgrader, and configurable cloud storage could also solve.