How complete is to-vega?

[domoritz]

I'd like to see whether to-vega could be a recommended way to write Vega-Lite in JS. To recommend it fully, we need to make sure that it is a) stable and b) feature complete. I was wondering what your thoughts on completeness are. Are there things expressible in Vega-Lite but not in to-vega? How do you plan to ensure feature compatibility in the future? Do you think it would make sense to generate the API automatically from our Typescript typings (like Altair is generated from the JSON schema).

See discussion in vega/vega-lite#2280

[gjmcn]

There are no major issues with completeness, but ensuring feature compatibility in the future and using the typings/schema needs to be addressed. Also, the current version is a first attempt and I am considering a number of other improvements that are relevant to the points you raise. I am on holiday until the start of January, but will add more details when I get back ...

[domoritz]

Sounds great. Do you want to join the Vega slack and chat with us Vega-Lite devs about your plans?

[gjmcn]

Great, I have joined.

[RandomFractals]

re. vega lite json shcema

[gjmcn]

I started updating and improving to-vega, but so many major changes were required that I decided to start a new project: Vizsla. This has various advantages over to-vega; in particular, specs can be combined more easily.

The improvements in Vizsla do not address the points raised above — feature completeness, using the schema etc. I have opened issue 3 in Vizsla to discuss these.