Jake Vanderplas

If the team is open to it, I'd like to work on adding `code-cell-output` to the myst notebook spec. Some semi-organized thoughts: Jupyter notebooks can have multiple outputs per cell;...

I was imagining this as a more general thing than asserting outputs. Currently `ipynb` can store code cell outputs within the file structure, while `myst-nb` has no mechanism to do...

OK, fair enough. If you’re not open to storing outputs in markdown, are you open to potentially supporting a different way to validate code output in-situ, since the testing tools...

Thanks for the context, @choldgraf. I was aware of the code at jupytext, but I thought that this repo was the source of truth for the myst-nb spec. If it...

Done in https://github.com/altair-viz/altair_widgets/commit/921e6ba50ea945be4485b09d9e8b29fa73f3c3a1. Thanks

Related question: are there any situations in which a valid Vega/Vegalite JSON string would be broken through doing `spec = json.dumps(json.loads(spec))`?

Another idea: I think it would make sense to let the Vegalite class take an optional data argument, that it would then roll into the dict/string spec (rather than having...

Is this still an issue? Do you have example code that reproduces the error? Thanks!

If you don't use ``additionalProperties: false`` on the ``anyOf`` schema, but instead use it in each of the components of the anyOf, does that work? Agreed on the lack of...

Maybe I'm not understanding... here's a concrete example where instances pass even with ``additionalProperties: false``: ```python import jsonschema schema = { 'anyOf': [ {'$ref': '#/definitions/A'}, {'$ref': '#/definitions/B'} ], 'definitions': {...