Hi Georg,
Would it be possible for a glyph to take multiple names, same way as it can take multiple unicodes?
I am talking about unencoded glyphs. As an example, in the case of tabular math signs where one would have tf and tosf and they would be identical, the designer can just add one set and double name each glyph with plus.tf, plus.tosf, etc…
Am I making any sense?!
Would this double naming be intended for the automatic feature generation? Or for something else?
Having multiple names for a single glyph would change the data model/app architecture significantly. There is probably a better solution depending on the task to be achieved.
It would generate the glyphs as well as OT features. Example: One glyph with a plus sign, two names (plus.tf, plus.tosf) generates both plus.tf and plus.tosf in the final file and updates the features for tf and tofs.
I would create a new, empty glyph and place the other glyph as a component. That generates the relevant feature code and both glyphs can be exported.
Sure. That is the current way to do it. Hence I proposed a potentially different way.
So you like to have the plus.tf for the plus.tosf, too? You don’t need to do anything to get that. Those number features are a bit funny. You basically always apply two. One for the “height” (cap-height vs old-style) and one for the “width” (tabular vs proportional). So to get .tosf numbers, you apply the “tnum” and the “onum” feature. And the former does “plus” > “plus.tf” and “one” > “one.tf” and the later does “one.tf” > “one.tosf” (but not “one” > “one.tosf”).
I just realize that this might only be true if you have lining or old-style numbers as defaults. But in your case it should be fine. What is your default and what other sets you have?