The layer compatibility is broken, I can’t generate fonts. Glyphs started to compare default layer with alternate layers and says they are not compatible.
Glyphs even compares the default layer to any other layers.
The layer compatibility is broken, I can’t generate fonts. Glyphs started to compare default layer with alternate layers and says they are not compatible.
Glyphs even compares the default layer to any other layers.
Which output is this?
It will compare with alternate layers if the switch is not exactly the same everywhere. Are you sure it’s about that glyph? You can filter for special layers, the go through them and see if there’s one not exactly switching on 400. (Glyphs should indicate those as incompatible though.) Perhaps consider the alternate glyph method.
The output is from Simon’s Why You Not Compatible script, but I used it just to reflect the fact that Glyphs marks the glyphs incompatible.
In any event this is a new bug in Glyphs. The fact that Glyphs compares the default layer with some random other layers should be a clue to that fact.
I emailed the Glyphs file to support@
I fixed the issue by removing one of the cap components, then selecting another layer, then run propagate smart components. Weird. Then weirdest is the Glyphs was comparing the default layer with some random layers.
There were some layers that had duplicated caps (_cap.LC-beak).
It is visible (but I admit, not strong enough).
See the blue circle. That is the extra, unused corner.
I re-build the glyph from scratch. The general issue of cap & corner components out of sync stands. Many times I can’t just do a propagate from some random layer because each layer has individual scaling factors from intermediate layers in the cap or corner component.
Look out for those circles. And I’ll try to improve the compatibility warnings.
I always look out for these circles, in this case I apparently didn’t see it.
Like you already admitted, they are really pale. I have good eyes (just recently checked again) and find them hard to see
An indicator similar to the one for duplicate nodes would be very helpful. Likewise, a more prominent indicator for detached corner components (instead of just a small pale blue circle at (0,0) would help a lot.
Also a built-in method to propagate these to only selected layers, same alternate layers, same intermediate layers, with or without background layers, and observing or overwriting the any scaling of the cap and corner components that might already exit in the layers.
The use of intermediate layers to create scalable C&C components is a bit of a hack. It does work, but the whole system of smart components, C&C components, and anchors need an overhaul.
There is a script for finding detached corner components.