Small bug: Some Greek letters showing up in "Other" category

I have a limited number of Greek letters… Xi and pi show up in “Letter, Greek” category of the main font view, but Delta, Omega, mu show up in the “Other” category.

From the “Greek → Basic” section, this isn’t so bad, but if I am looking at “All” glyphs, it’s easy to lose those characters way down in the “Other” category

It’s possible there’s some logic to this I’m not considering, I don’t know a whole lot about Greek…

They also have a math usage

Hmm, that’s a good point. I would expect math to show up under “Symbol → Math”, not “Other” though. also arent Delta/increment and mu/micro different unicodes?

Interestingly, it looks like Delta, Omega, and mu are all lacking unicode points by default (and when assigned, the category doesn’t change)

Can you send me the .glyphs file?

as an update: saving as a new file and restarting glyphs fixed the problem. can’t say I understand why, but I’ll take it :slight_smile:

For a Latin-only font, what’s the best practice for including micro/mu, increment/Delta, pi/piSymbol? Which glyphnames should we use, and which unicodes? Include component versions? Double-encode glyphs?

that’s a great question question. for pi, at least, you would probably only include pi/U+03C0, as piSymbol/U+03D6 is a different design. I could be wrong on this (probably am!), but I was under the impression U+03D6 is there in greek specifically to differentiate the math symbol and the lowercase letter

I was thinking it might make sense to orient around keyboard shortcuts for Latin-only fonts, but I’m not sure if they target the same unicodes on different operating systems. For example, on a Mac, Option-J brings up increment rather than Delta, and Option-M brings up micro, rather than mu. this would lead me to believe Delta and mu could be omitted in a Latin-only font, but I still have that lingering fear about some different operating system, or some weird edge case out there looking for the Greek unicode and not finding it, so I have been including both :man_shrugging:

I wonder what the others think / if there is a definitive best practice

1 Like

Codes: take the ones in the current MacRoman and Win1252 listing.

Design: the ones with Greek codes should be designed properly Greek. If you don’t support Greek otherwise, it doesn’t matter, and you can do the scientific pseudo-Greek or whatever looks right to you in the Latin context.

Usage problem: in websites, unfortunately, very often the fallback font takes precedence because of bad CSS. A user error, but so common that Delta and pi in Greek are often overridden with the Latinized glyphs of the primary font in the CSS definition. Similar scenarios in office software, where Greek will fall back to a proper Greek font except for the two pseudo-Greek letters in the otherwise Greek-less font chosen by the user.

So, in webfont exports at least, it may be a good idea to choose an appropriate subsetting and cut out Greek codes completely unless you fully support Greek in your glyph set. And avoid using Greek codes unless you have to. Unfortunately some Keyboard Layouts still produce the Greek codes (IIRC for Delta and pi, but need to go through them again) when they should have given you the symbol code. Double encoding is only okay if you have full Greek support. Don’t do it for micro, increment, product, summation or Ohm just to be ‘on the safe side.’