Lars Willighagen

I understand less about the text/application distinction than I thought, I would have to take some more time to find that out. Anyway, from [what I gather](https://en.wikipedia.org/wiki/Media_type#Registration_trees) both `text/kdl` and...

Looks good to me, did you see this page from the CSS WG being tweeted yesterday? https://wiki.csswg.org/ideas/mistakes > # Incomplete List of Mistakes in the Design of CSS > ......

I think we have a TextMate definition [here](https://github.com/kdl-org/vscode-kdl/blob/main/syntaxes/kdl.tmLanguage.json), but yes, as far as I am aware more usage is required for GitHub to consider adding it.

Or just outright fix the order of matchers in a node filter to something like this ``` node-filter := top | tag? node matcher* | tag? matcher+ | tag ```...

It might not be a problem with the spec, more my interpretation of it I guess (since there isn't a grammar yet).

> ``` > s_expr := > s_term | > selector s_op s_term | > selector s_term > ``` What's `selector` here, `s_expr`? >``` >s_term := > 'top()' | > identifier...

I have a prototype implementation (lacking sibling selectors for now, and tags since my KDL parser does not support them yet): https://github.com/kdl-org/kdljs/commit/ec79f789211df5d25e87b15671c390baae6e928a. I ended up just giving `"top()"` priority over...

I think `top()` is a bit confusing anyway, is this interpretation right? - `top()` *and* `top() > []` select the top-level nodes (this part it says literally) - `top() []`...

> Like, would it be better to have a unary `> foo` to be very specific about "I want the toplevel children with node name `foo`"? I personally do like...

> I just realized that `+` and `~` are both valid KDL identifier characters. *sigh* Is requiring whitespace around combinators (and other operators) and then just saying that you need...