Leo White

I've read through the ml workshop abstract and the relevant parts of the related Haskell work. My current thoughts on the proposal are as follows. Firstly, I think that contractiveness...

> separability being optional in the proposal, it may make sense to view it as an independent attribute (it might even give rise to a stronger handling, such as saying...

I think that is a reasonable argument for using a (contextual?) keyword rather than a symbol, but I dislike using attributes and extension nodes for genuine language features. Half the...

I think this PR tries to address a genuine problem. The motivation and technical details are pretty clear in the description. All in all a pretty good first RFC for...

> Do you have a different understanding? I had a slightly different understanding, where the document describes both the solution to be implemented and a summary of the main considerations...

> Looking forward to reading your forthcoming RFC. See #3

I haven't read this in detail yet, but I think it might be interesting to compare it to: https://arxiv.org/abs/1907.07283

> automatically add pure to ground types in interfaces, and ignore it when subtyping IIUC this suggests that you might want to find a different word than "pure" for these...

All I can think of so far is `[@@value]` since it is a bit like you are marking something as a value from the point of view of the value...

So is the idea that `odoc` would extract the code samples and then some other mdx-like tool would call this feature of odoc as part of its processing? That could...