Enivex
Enivex
By the way: if you plan to add more selectors it would be useful to have non-inclusive before and after selectors. Especially if the IDs get implemented. That way I...
Point 1: I don't think this is viable due to incremental compilation, but someone correct if I'm wrong. Point 2: Using tabs instead of spaces for indentation would save some...
> @Enivex iirc the IDs were indeed introduced because of incremental compilation (I want to say that they were introduced in one of the gradient PRs but frankly it's been...
> We pass everything down to the CSL style, the rest is up to the style in use. I think there's not much we can do about it, unfortunately. :/...
> Try `Renderer=HarfBuzz` and LuaTeX, it should give bettter results here (it basically applies the solution I suggested in my comment on #526). I honestly thought this was the default...
Has this been resolved now @LaurenzV ?
The only real way to solve this would be to merge the parts of the glyph in order to remove the overlap. We'll see a lot more of this once...
> The slight overlap is intended by the OpenType MATH spec. The overlap is intended, but it's not taken into account when anti-aliasing (it's done on each part individually), which...
Closely related issue https://github.com/typst/typst/issues/4624
While the runt cost is high, the cost of the resulting extremely awful word spacing on the first line may be bad enough to offset that. (I'm only speculating.)