How should a notation program handle the extra JSON files?

[tisimst]

Daniel, et al,

I have been thinking about how to make LilyPond natively support SMuFL-compliant fonts. However, I have two burning questions that I can't seem to resolve (or feel good about) that keep me from attempting it (since it's bound to take a fair bit of work):

1. Why are the glyph names (in the font file) more like the unicode point and not the human-friendly names? Switching to the human-friendly names would eliminate the need for "glyphnames.json", no? As a matter of documentation, it's a nice file to have, but as a matter of usage, it seems more efficient to call it by a known name rather than relying on a mapping function. For those of you who are using SMuFL fonts in your application, what do you do?

2. What is the recommendation for storing/using all the JSON files? Is there a specific directory that these are placed in so the notation program knows where to find them? Supposing a system has more than one notation program that can use SMuFL fonts, should there be a single location that they all can refer to?

I'm not sure if these questions have been answered already, but I'd love to get everyone's thoughts on the matter.

- Abraham