Decide default configuration for MyST-Parser v0.13

I’m nearly done with myst-parser v0.13.2, with lots of changes/additions to the extension system, as explained in https://github.com/executablebooks/MyST-Parser/blob/master/CHANGELOG.md and outlined in https://myst-parser.readthedocs.io/en/latest/using/syntax-optional.html

We should decide what the default configuration will be (see https://myst-parser.readthedocs.io/en/latest/using/intro.html#sphinx-configuration-options) and how it is utilised in _config.yml, e.g. what extensions should be enabled by default and how additional extensions are enabled.

Currently, only dollarmath is enabled by default, then there is myst_extended_syntax to blanket enable all other extensions. Since it is a list now, it might be easier to just list them all and comment out ones disabled by default, something like:

parse:
  enable_extensions:
    # - amsmath
    # - colon_fence
    - deflist
    - dollarmath
    - html_admonition
    - html_image
    # - linkify
    # - replacements
    - smartquotes
    - substitution
  substitutions: {}
  html_meta: {}
  url_schemes: [mailto, http, https]

Note linkify also requires the additional dependency https://github.com/tsutsu3/linkify-it-py

Also need to decide how to deprecate current keys (primarily myst_extended_syntax)

cc @choldgraf @mmcky @AakashGfude

[choldgraf]

Actually, another approach we could take is to enable whatever extensions we need in order to get the default “Jupyter Notebook” markdown experience. Since Jupyter Book is jupyter-specific, that seems to be a reasonable assumption. Another guiding principle could be “things that are particularly useful for large, complex books”.

I think this would mean the following extensions:

• dollarmath (since the notebook supports it)
• linkify (since the notebook auto-links URLs)
• substitution (since this is useful for complex books)
• colon-fence (since it’d improve reading in notebook interfaces)

Anything else? What do @mmcky @jstac @AakashGfude think?

[fperez]

My 1e-2 - even with @chrisjsewell’s caveats, I think I’d still +1 linkify: parity on that point between the “live” notebook experience and the resulting JB html would be great to have. Our current class notes have a ton of links that have been pasted without “markdown-ifying” them over the years (e.g. here), and it’s annoying to see all that rendered as plain text in a web page 😃

So my vote would be for it, though I appreciate the dependency and potential performance impact - your call!