ꭒ - u with left hook

[felileg]

Hello,

I may be out of line, as Junicode is more for medievalists than romanists, but I use this font for its faithful handling of diacritics when digitising Gallo-Romanic dialectal atlases. That is why I would find it useful to add support for the letter ꭒ, used in particular in dialectology in various phonetic alphabets.

Weird rendering of ꭒ with diacritics

Would this be possible? Thank you!

[psb1558]

Junicode supports lots of different academic fields. It's just that for most fields other than medieval I have to rely on requests like this one rather than my own knowledge of the relevant scripts. So very glad to get this issue, and glad to add the character.

[felileg]

Thank you very much!

[psb1558]

This will be included next time I build a new version. It is U+AB52.

[felileg]

Nice!

I forgot to mention that the most common shape is the following Perhaps for the italics?

[psb1558]

I think I misinterpreted what I was seeing when I looked over the (very small) collection of fonts supplying this character: the top of the left hook is about the same height as the top of lowercase o, but as to the shape of the hook there is disagreement between, e.g. Arimo (left) and Symbola (right):

For the roman I could try something more like this:

Or, if you think it should be more like your italic example, could you post an image?

Also, could you advise about the correct position of the diacritic below? At present my anchor is where it is for other flavors of u:

But in your image it is somewhat to the right of that.

[psb1558]

This would be better, I think:

Edit: A further improvement for the hook:

[psb1558]

Variants. Bold, Light, Condensed, Expanded:

[felileg]

The original fonts (Rousselot-Gilliéron script) are always in italics and rounded, as follows: The continuations are mainly hand writed, but remain in italics and rounded: The few modern fonts that support this letter show it as rounded or even curled:

[felileg]

As for the diacritic, I think the first example was not well printed. The second one shows a more centered stroke.

[felileg]

I want to make it clear that these are just leads, I think you’ve done an wonderful job!

[psb1558]

Here's another go:

Walther Beta (a font that I've never seen or heard of) seems most interested in sticking close to the early prints.

[felileg]

Sounds perfect! Walther indeed seems to be more concerned with having a uniform look and respecting the original glyphs It is used to digitise the Französisches Etymologisches Wörterbuch (the font is named after its author Walther von Wartburg). I have access to it through my studies, I don't know if it is published anywhere.

[kenmcd]

A couple more examples:

Note: the TNR version could easily be confused. Not good.

[psb1558]

Mmm. The Noto design is fantastic.

[psb1558]

One more. I think maybe this will do it.

[kenmcd]

Are the readers/users of this character going to know what it is? Your hooks are definitely prettier than Noto Serif (and some of the other examples). But even on the italic the Noto Serif hooks are unambiguously hooks, not swashes. As an ignorant-of-the-usage observer I am just wondering. Really like the looks of the curly hooks tho.

[psb1558]

My first impulse was simply to take a hook from one of my IPA characters and use that—see my first attempt. But this is (despite the Unicode name) evidently not supposed to be an IPA-style hook (the Rousselot-Gilliéron phonetic script is pre-IPA). It will be used in highly specialized environments where the likelihood of ambiguity is slender. And then, the roman shape resembles that in the "Walther Beta" font used in the Französisches Etymologisches Wörterbuch, familiar to many working in this field (even I know that reference); and the italic shape, resembling that in early Rousselot-Gilliéron fonts, will also be familiar to many.

So I'm not worried about the few readers who will ever see this character being confused by it. They'll be fine, I think.

[psb1558]

Fonts containing this glyph are now in the repository.

[felileg]

I think the same, it's a niche use. The most important thing is that ꭒ and u are sufficiently differentiated.

I've tested the font in use and it works perfectly for transcriptions! (Perhaps I'll suggest a few other less crucial characters in the future if there's room left in the font).

[psb1558]

There's plenty of room in the font.