Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Variants of o with stroke #3

Open
jsbien opened this issue May 12, 2020 · 33 comments
Open

Variants of o with stroke #3

jsbien opened this issue May 12, 2020 · 33 comments

Comments

@jsbien
Copy link

jsbien commented May 12, 2020

o-with-horns
In Polish texts o with stroke was for centuries used for a nasal sound,
but practically never in its present form. The manual for critical
editions suggested using the very first version in the enclosed scan,
but it is not available in the present day fonts (with the relatively
recent exception of TeX Gyre - EC10 and EC11, but for various reasons
this is not a satisfactory solution) . There were even talks
about submitting the "o with horns" as a new character to Unicode, but
I think it is just a variant of U+00F8.

So my feature request is to add at least the recommended form as a variant of
U+00F8. However I would be satisfied also with other solutions, like
creating the shape with some diactitics.

@psb1558
Copy link
Owner

psb1558 commented May 12, 2020

Would these work?
image
I would make them accessible via a cvNN feature.

@jsbien
Copy link
Author

jsbien commented May 12, 2020

Great! Thank you very much!

@psb1558
Copy link
Owner

psb1558 commented May 12, 2020

And italic.
image
These will be accessible via cv16. Fonts will be uploaded late today.

@jsbien
Copy link
Author

jsbien commented May 16, 2020

otfinfo still reports "cv16 unknown feature" for JuniusX-Regular.ttf. Am I doing something wrong?

@psb1558
Copy link
Owner

psb1558 commented May 16, 2020

I don't think you're doing anything wrong. The cvNN features are relatively new, and it looks like otfinfo sees they are there but doesn't know what they are. But if you've got LibreOffice, you can test very straightforwardly:

image

In the font box you need "JuniusX:cv16=1" . . . "JuniusX:cv16=4"

If you've got otfinfo, does that mean you're using TeX? LuaTeX and XeTeX both use the Harfbuzz shaping engine, I think, which is very up to date and will handle cvNN correctly: it's documented in the Fontspec manual. All the major browsers will handle it correctly too, but MS Word gives no access to it.

@jsbien
Copy link
Author

jsbien commented May 16, 2020

My intention is to use it with XeLaTeX (Debian buster), but I have problems - that's why I tried otfinfo. I don't use LibreOffice, I tried to reproduce your example but failed (maybe my fault) I will investigate my XeTeX problem, if necessary I will ask for help on its mailing list.
You haven't mentioned the issue in your commit messages, so I just wanted to check whether the fonts with these characters are really uploaded.

@psb1558
Copy link
Owner

psb1558 commented May 16, 2020

Oh, yes, it's there. I tend to make too many changes per commit, and I don't mention them all. I promise it will work fine with XeLaTeX. I've always used Fontspec with that, and it seems very solid.

@psb1558
Copy link
Owner

psb1558 commented May 16, 2020

For the record, this syntax will replace the slashed o with the fourth of the alternates for ø in XeLaTeX with Fontspec:

\usepackage{fontspec}
\setmainfont{JuniusX}[RawFeature={+cv16=3}]

or anywhere in the document:

\addfontfeature{CharacterVariant={16:3}}

Note that in Fontspec the variants in features like salt and cvNN are indexed from zero instead of from one so use cv16=3 instead of cv16=4.

@psb1558
Copy link
Owner

psb1558 commented May 17, 2020

I guess it's okay to close this now.

@jsbien
Copy link
Author

jsbien commented Dec 17, 2020

I just found yet another variants of the letter:
Vrtel1200orogate
I would appreciate very much if you add them to the font.

@psb1558
Copy link
Owner

psb1558 commented Dec 17, 2020

Can you supply higher resolution images? More like the ones at the top of this thread?

@psb1558 psb1558 reopened this Dec 17, 2020
@jsbien
Copy link
Author

jsbien commented Dec 17, 2020

I can try but not immediately, because I need to use another scanner (tomorrow I hope). Moreover I'm not sure it will help as the quality of the original is low. As for the second glyph, it is just F0011 with the "horns" vertical. The first one is like F0013, but again the "horn" is vertical; moreover is extends upward inside the "counter" (it's Gaskell's term, for me quite strange).

@psb1558
Copy link
Owner

psb1558 commented Dec 17, 2020

I'll rough it in, but finer details can be useful, like: are the horns straight lines or tapered? So I'll wait for the better scan before making the shapes final.

"Counter" is an old typographical term, going back to metal type, when you'd use a counter-punch to make those negative spaces in the punches (which are part of the typefounding process). A useful term, since the spaces in and around letters are as important to the designer as the black strokes.

@jsbien
Copy link
Author

jsbien commented Dec 18, 2020

Vrtel_vi

Vrtel_12
Vrtel_16

@psb1558
Copy link
Owner

psb1558 commented Dec 18, 2020

These are great scans! Many thanks, Janusz! And I love these letters, especially in the middle sample. Will get them done soon.

Out of curiosity, about what are the dates of these prints?

@jsbien
Copy link
Author

jsbien commented Dec 18, 2020

These are the 20th century transcriptions of the manuscripts: 13th or 14th century Holy Cross Sermons and 14th or 15th century Sankt Florian Psalter. The characters appeared also in those early prints which intended to follow the look of manuscripts, but I don't have good examples handy.
They mean a nasal vowel which was long or short, but written identically. Later it was replaced by two clearly distinct vowels, now denoted by ą and ę.

@psb1558
Copy link
Owner

psb1558 commented Dec 18, 2020

Thanks for these! The sermons are interesting, the psalter both interesting and beautiful. The wikipedia scan of the psalter is very good: I can see the ø variant very clearly. The wikipedia article links to what appears to be a normalized or modernized transcript of the psalter text. Do you know of a transcript with original spelling? Or I could fix it up myself--this would answer the need I emailed you about just now.

@jsbien
Copy link
Author

jsbien commented Dec 18, 2020

There has been 7 editions of the transcriptions of the semons, the first one in 1891, the last one in 2009; not all of them are freely available.In almost every edition there is a different convention to render the nasal vowel: from the Greek phi to ą. I had a quick look at them but none seems to be useful for this purpose.
I tried to locate on the scans the following fragment of the 1950 edition but failed:
Vrtel1950p12
The whole text consists of just two pages: https://polona.pl/item/304920. Let's approach the problem other way round: find a nice fragment and I will try to find the transcription.
Reportedly closest to the original is the edition of Paul Diels (https://hdl.handle.net/2027/wu.89014029730 but not accessible for me).
P.S. An English text about the sermons: https://polishlibraries.bn.org.pl/upload/pdf/93724_Wieslaw_Wydra_The_Oldest_Extant_Prose_Text_in_the_Polish_language.The_Phenomenon_of_the_Holy_Cross_Sermons_133-159.pdf

@psb1558
Copy link
Owner

psb1558 commented Dec 18, 2020

I have got access to the Diels, but it looks monstrously difficult, full of lacunae and just very difficult. I also have Psalterii florianensis partem polonicam ad fidem codicis recensuit apparatu critico indice locupletissimo instruxit Wladislaus Nehring (Posnaniae, Żupański, 1883). I'm wondering if this looks like an acceptable text to reproduce for my purposes. A fragment:
image
I was able to download these, by the way, since they're out of copyright. Let me know if you have any use for them.

@jsbien
Copy link
Author

jsbien commented Dec 19, 2020

It demonstrates only the "horned o", but I don't have any better proposal, at least for now. It it's OK for you, then use it.
As for Nehring I can download it myself from Polona, but Hathi Trust seems overcautious and does not allow me to view Diels and I would be happy to have a look at it.

@psb1558
Copy link
Owner

psb1558 commented Dec 21, 2020

Here are all the ø variants:
image
I'm posting them all for comparison because the strokes for the last two are longer than those for the first four. I've noticed that the strokes for those four can be rather subtle, especially at small sizes, and I wonder if it would be a good idea to lengthen them to more or less match the last two.

@jsbien
Copy link
Author

jsbien commented Dec 21, 2020

It's OK for me as it is. I think there is some logic in the vertical strokes being longer. Thank you!

@jsbien
Copy link
Author

jsbien commented Dec 21, 2020

What do you think about these two excerpts from Zaborowski's treatise?
Zaborowski_MBC4orogate
In the first edition the printer was missing the "horned o" and used a replacement, but what this replacement is?
Zaborowski_Polona3
In the second edition the printer had already the "horned o", although not fitting well the rest of the text.

@psb1558
Copy link
Owner

psb1558 commented Dec 21, 2020

I've never seen the top one before. Some kind of symbol, I guess, but there are so many symbols in Unicode, it's a daunting project to figure out what it might be.

The bottom one also has a look of the symbol--or so it seems to me, the o part being more like a circle than an o.

Turns out I can't download the whole of Diels, though it's out of copyright. I should be able to: I'll consult a librarian about it after the Christmas break.

@psb1558
Copy link
Owner

psb1558 commented Dec 21, 2020

In a gothic style like this, how do you tell the difference between a diacritic and a flourish?
image
In the right-hand column, there's the horned o, and plenty of plain o's. But there are also plain y's and y's with horns: are those horns diacritics or flourishes? Likewise with z, some are plain and others have a vertical line on top. In the 1883 edition these horns or lines are not reproduced: did the editors get that right?

@jsbien
Copy link
Author

jsbien commented Dec 21, 2020 via email

@psb1558
Copy link
Owner

psb1558 commented Dec 21, 2020

I love medieval MSS, but they can drive you crazy. Often with Middle and early modern English, you can't tell if a mark at the end of a word is a flourish or indicates a final e. Similar ambiguities, it seems, with this psalter.

I've seen the modernized transcript. Not a lot of help. unfortunately.

@jsbien
Copy link
Author

jsbien commented Dec 21, 2020 via email

@psb1558
Copy link
Owner

psb1558 commented Dec 23, 2020

I don't know if an italic version is needed, but Google's rule is that the italic character set has got to match the roman (except for swashes and the like), so here it is.
image

@jsbien
Copy link
Author

jsbien commented Dec 23, 2020

OK

@jsbien
Copy link
Author

jsbien commented Jan 5, 2021

@jsbien
Copy link
Author

jsbien commented Jul 29, 2021

As "o rogate" has been accepted for Unicode 14.0.0 (to be released September 14), when working on issue #156 it would be perhaps good to reconsider this solution. BTW, personally I use just plane 15 codepoints as I have no need for sophisticated searching.

@psb1558
Copy link
Owner

psb1558 commented Jul 29, 2021

I see that 14.00 will have a good bit in it to make my life interesting. What I'll do, when the release happens, is (as has often happened with MUFI things) move the character (with others in 14.0) from PUA code point to the new Unicode code point.

I may delay re-implementation of the feature currently providing access to the character until 14.00 is released. But the feature will continue to provide access to the variant shapes.

I notice, following a link, that James Kass cricitized "the practice of issuing fonts with glyphs already mapped to proposed characters." Of course, the fonts mentioned in Daniel Bunčić's proposal did nothing of the kind: they were all out there before the proposal was submitted. The arguments against using PUA code points are powerful, and I (mostly) agree with them. But it's hard for a font maker to know what to do when a character is needed by the scholarly community and it is unknown whether it will ever be proposed to the UTC, or whether the UTC will ever take enough time out from processing emojis to approve it.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants