Laurenz

> Adding to the confusion, even when citations are created directly with the cite syntax (as in case 4 above), the resultant element in the PDF is a link to...

I think they are semantically the same thing (jumps to a destination) and adding a different link type just to help with styling would be a workaround.

Explicit v spacing has higher priority than block spacing. Otherwise, it would be impossible to override normal block spacing locally.

> The intended behavior in the example above should then be achieved by replacing the heading show rule with: The canonical way is `#show heading: set block(spacing: 2em)`, reusing the...

Ah, now I get it. Yeah, the example in the outline docs is a bit problematic then.

I'm not 100% following. Is the proposal that lengths should be multipliable with lengths? Because afaict that would require us to introduce a new type that is basically length^2.

This is controlled by OpenType MATH font metrics. Not sure whether I would want to expose overriding those. It's a slippery slope with overriding other things in fonts. I'll reword...

Yeah, that's indeed a problem. We discussed this a bit on Discord, but I'm also posting this here for posterity: - The `Content::query` function shouldn't be used. It's more of...

I think docs can wait for now

Hey @schrottkatze, are you still planning to work on this? If not, no worries, I would then just close the PR.