Harry Dalton
Harry Dalton
(hello again! I've updated the issue to include a reproduction and animation showcasing the horizontal shift problem too - thanks for your attention)
This appears to originate at the layout stage as opposed to the rasterisation stage, as the `border` is adopting a fractional value while the negative `margin` applied to overlap the...
(for example, manually setting `margin-left` to `-0.667` instead stops reproducing the issue at that exact zoom level)
The fix for #28940 looks relevant here.
(looks like sub-issues [are still opt-in per-org](https://docs.github.com/en/issues/tracking-your-work-with-issues/using-issues/adding-sub-issues), so I'll just add them to the main issue description here instead)
((all done now))
(creating as a draft PR to track progress, will undraft as I address the testing, accessibility, and documentation checklists above, where appropriate)
Thanks for confirming and the speedy reply :) I have added a commit that makes the same change for the Glyphs 2 docs. At the moment this change is for...
I think they're not ignored but slightly worse, as the default layer of the UFO is re-inserted at their location
The `--update-name-table` argument might do what you need here; it relies on the variable font's `STAT` table being well-constructed, but if it is then will handle a lot of the...