When I generate legacy diacritics (from Ctrl-clicking the appropriate category list), I get glyphs that inherit the left sidebearing from my existing combining diacritics. But instead of also taking the right sidebearing from there, they all have advance widths of 1200. Is there a reason why they wouldn’t just take the same width and sidebearings as the combining marks? It’s especially puzzling because if I place the components of the accents in an empty slot, the sidebearings to change to match the source glyph.
They just get the default width. You seem to have a UPM of 2000 and 1200 is the default width.
It just seems odd to me that they wouldn’t inherit the width from the source of the component (in the same way that, for example, a generated /Dcroat gets the correct width from an existing /Eth; or, as I mentioned, as they would if I manually added the components).
I get that sidebearings are sort of provisional for combining marks. And default advance width is a sensible width for the glyph.
But maybe the better solution would be automatically centering the legacy marks on the default width upon generation. At any rate, the present behavior requires redoing the metrics one way or the other (changing the advance width if you want the sidebearings to match the combining marks, or changing the sidebearings if you want them centered on the default width).
Obviously either change is easy enough to accomplish; I’m just suggesting a place where the application gave results that were unintuitive to me.
I think I improved that a bit.