Problems with Manichaean swashes and ligatures #2
Labels
Comments
|
Hi @dscorbett. I was just about to suggest that this problem should be fixed in Unicode upstream, and then I found that you did submit this as an error report to the Consortium! Thank you for that. I wonder what the status of this is. I see that SAH (in L2/19-343) was asking for more evidence (which is a bit irksome given the list of evidence that you compiled above...) but I haven't seen any follow up since. Perhaps I need to prod SAH again. |
|
I think SAH wants me to submit my evidence as a PDF with inline images and an explicit proposal of how to reword that section of the standard. It is a lot of work for a tiny editorial change so I am not in a hurry to get to it. Please, feel free to prod them for me. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Font
NotoSansManichaean-Regular.ttf
Where the font came from, and when
Site: https://github.com/notofonts/manichaean/releases/tag/NotoSansManichaean-v2.005
Date: 2023-10-07
Font version
Version 2.005
Issue
Noto Sans Manichaean implements the two ligatures (cy and cn) mentioned in The Unicode Standard. The standard is wrong.
At the very least, there is a clear bug with the positional forms of cn.First, here’s the bug. In the sequence cnd, where d could be any right- or dual-joining letter, the cn ligature does not join to the d. This is because the ligature has no initial or medial forms, so the font ends up using the isolated or final forms instead. The font should defer ligation of cn till after the topographic features, because the ligature is only formed when the n is isolated.(This has been fixed.)The Manichaean section of The Unicode Standard claims that “Manichaean has two obligatory ligatures for sadhe followed by yodh or nun.” That’s not true: they are not obligatory and it is misleading to think of them as ligatures (not to say you shouldn’t implement them as ligature glyphs). They are essentially swash forms, where the tail of the sadhe wraps around the following letter – sometimes multiple following letters. Sometimes it curves so much it touches the following letter and they ligate; sometimes it curves so little it just looks like kerning.
Currently, the two ligatures are made in 'rclt'. Because the ligation is optional, it should use a feature that users are meant to be able to disable. I suggest 'liga'.
By default, I think the font should use the most common form for these sequences. My research indicates that that is the form where the tail wraps around the next letter, touching it if it is n and not if it is y. If possible, you should consult with an expert in Manichaean. I found some manuscripts at the Digitales Turfan-Archiv and I got the transcriptions from TITUS.
To make this font authentically swashy, you should add (optional?) swash forms to letters like k: “The final stroke of the letter c tends to enclose a following y or n, but without distorting either letter; the final stroke of k and related letters can extend to the end of the word in which the letter occurs” (https://iranicaonline.org/articles/manichean-script).
Character data
𐫝𐫗𐫅𐫝𐫏
U+10ADD MANICHAEAN LETTER SADHE
U+10AD7 MANICHAEAN LETTER NUN
U+10AC5 MANICHAEAN LETTER DALETH
U+10ADD MANICHAEAN LETTER SADHE
U+10ACF MANICHAEAN LETTER YODH
Screenshot
The text was updated successfully, but these errors were encountered: