Skip to content
This repository has been archived by the owner on Jan 25, 2023. It is now read-only.

Noto Sans Cuneiform not working #2312

Closed
hreyes82 opened this issue Apr 7, 2022 · 15 comments
Closed

Noto Sans Cuneiform not working #2312

hreyes82 opened this issue Apr 7, 2022 · 15 comments

Comments

@hreyes82
Copy link

hreyes82 commented Apr 7, 2022

NotoSansCuneiform-Regular.ttf
https://fonts.google.com/noto/specimen/Noto+Sans+Cuneiform?noto.query=cunei
2022-04-07
Version Windows 2.000
Issue observed using Microsoft Word

I installed the font as I typically would and when I choose the font to use it, text appears in regular typography not in Cuneiform. I'm using other fonts that do work but this one does not. When I open the downloaded file, the font or typography also displays as regular font not cuneiform. Could this be that the wrong file has been provided?

@Mercury13
Copy link

Mercury13 commented Apr 7, 2022

What characters do you type? This is a Unicode font, so typing in Latin will do nothing.

https://www.fontspace.com/zarathustra-font-f13841
This font supports Persian cuneiform (not Sumerian!) 103A0…D5 and also mapped to Latin. Persian cuneiform is alphasyllabic and only 36 letters long, so it’s possible somehow.

https://fonts.google.com/noto/specimen/Noto+Sans+Cuneiform?noto.query=cunei
This font supports three ranges of Sumerian cuneiform 12000…543, and NOTHING is mapped to Latin. Sumerian cuneiform is hierosyllabic, so there’s no point mapping to Latin. And that’s Unicode font meant to read and write in some language implemented in Unicode, not to make fancy text.

For Unicode font supporting Latin with some strange characters is bad: if we type in Latin with some missing font, and font matcher finds our cuneiform font the most suitable, we’ll get cuneiform instead of Latin.

@hreyes82
Copy link
Author

hreyes82 commented Apr 8, 2022 via email

@Mercury13
Copy link

Mercury13 commented Apr 8, 2022

I regular english words that I want to translate to cuneiform...

If you are e.g. a tattoo master, you need another font. We are serious people who discuss how glyphs combine to words, how to fix bugs in umlauts when e.g. in Tibetan or Gujarati three of them stick to one letter, how to circumvent faulty typography engines, and if there are still tofu remaining (Noto = No tofu).

But first read this post: http://hanzismatter.blogspot.com/2006/08/gibberish-asian-font-mystery-solved.html
And this: http://hanzismatter.blogspot.com/2017/06/from-teresa-b.html
He calls such people “another victim of gibberish font”. As Sumerian cuneiform is hierosyllabic script, you should convert English to cuneiform by syllables, not by letters.
Persian cuneiform is alphasyllabic, but it is even harder for tattoo master: they traded shorter alphabet for harder writing rules.

@simoncozens
Copy link
Collaborator

Closing this; it's not a bug in the font. Fonts draw text, they don't translate it.

@hreyes82
Copy link
Author

hreyes82 commented Apr 8, 2022 via email

@yanone
Copy link

yanone commented Apr 8, 2022

@hreyes82

First of all, please note that this is a public forum and anyone can comment on the issue you created. To our knowledge, the user @Mercury13 is not affiliated with Google Fonts and their response should therefore not be mistaken as originating from Google Fonts.

About your problem:

You seem to expect that you type a Latin character (e.g. A) and that the font in question transcribes it to a Cuneiform caracters (e.g. 𒀀). While certainly fonts exist out there that have this functionality, it is not how fonts on Google Fonts work. Our fonts follow the Unicode standard, a database that is constantly expanding to give every glyph of every script on this planet a code which can be process by machines. The Unicode block for Cuneiforms is described here: https://en.wikipedia.org/wiki/Cuneiform_(Unicode_block)

As you can see, the Cuneiform alphabet is quite different from Latin, and it is therefore not trivial to translate or transcribe Latin into Cuneiform. But it is also not intended.

The font in question, Noto Sans Cuneiform, and all other fonts on Google Fonts, will output the characters in the respective font design only when you put in the correctly encoded characters. It will print 𒀀 only if you put in 𒀀 in your text editor.

For starters, it may help if you copy/paste individual characters from the Wikipedia page I linked and see if they show up in Noto Sans Cuneiform in the app you're using.

@hreyes82
Copy link
Author

hreyes82 commented Apr 8, 2022

Thank you for your response and clarification @yanone I appreciate t as it helps me better understand how it works. I understand what you are trying to say and that Latin word etc. won't be translated or converted to cuneiform, however, not even individual characters do? There is no character in my keyboard that converts using this font?I have installed some other similar fonts and they all type in cuneiform, character by character, converting each character I type. This one does not, and displayed the text as I type it in English, not even a black square or error, so I wondered if there was something wrong I did when I installed it. If I can check a cuneiform transcription and see how "a" for example would be translated in cuneiform, why is this font displaying "a" in English, instead of the cuneiform?

@simoncozens
Copy link
Collaborator

I'm aware that other fonts may work differently, but this is because those fonts have different objectives. Noto follows a standard of encoding called the Unicode Standard, which defines how characters should be represented to the computer. It allows for different scripts to work together regardless of font. Before Unicode was a standard, people would make up their own encodings for characters; "a" might mean "a" in an English font and "က" in a Burmese font and "𒀀" in a Cuneiform font. This was a terrible thing! You changed the font in your document, or you had a different font installed to the person who wrote the document, and the text becomes completely different.

Unicode assigns a single meaning to a single character. If you type the letter "a" you get the letter "a", whatever font you use. And this makes sense - by typing the letter "a", you've just told the computer you want the letter "a", not "𒀀"! In Unicode, Cuneiform letters are encoded in a completely separate "block" to the Latin letters, so there's no confusion. Other fonts which use a "hacked" encoding might make it easier for you to type, but they have a fundamental problem in that they turn a document written Latin letters into something that it is not (Cuneiform letters). So that's why we follow Unicode in this project.

But this means you do have to work a bit harder to input Cuneiform text. To access the Unicode Cuneiform block, you can either install a keyboard which supports those characters, or you can use a character picker such as UniView to choose the characters you need.

@r12a
Copy link

r12a commented Apr 8, 2022

But this means you do have to work a bit harder to input Cuneiform text. To access the Unicode Cuneiform block, you can either install a keyboard which supports those characters, or you can use a character picker such as UniView to choose the characters you need.

Hint: To quickly transfer cuneiform from the left panel to the text area, double-click on a character in the left panel, or change the direction of the red arrow in the middle by clicking on it. (I recommend reading the Help page for tips like that.)

Actually, there's little chance it's relevant but there's a Ugaritic character app which makes input much easier still, but is only focused on use of cuneiform for Ugaritic. There is no more general cuneiform character app yet.

hth

@hreyes82
Copy link
Author

hreyes82 commented Apr 8, 2022

Thank you @r12a @simoncozens for your explanations and input, it truly helps too. I guess I made the mistake of thinking that I would just install this font like any other and it would turn characters I was typing into cuneiform. This definitely helps me understand what is going on so thanks.

How would I be able to get to Sumerian cuneiform? Is there anything out there already put together?

@twardoch
Copy link

twardoch commented Apr 8, 2022

To further clarify: if you have a physical keyboard, it doesn’t really matter what letters are printed on it. It matters what keyboard layout you have active in your operating system. Mostly, you choose just one keyboard layout when you install your computer, or the manufacturer has picked one for you.

The default (and often only) keyboard layout is typically associated with the UI language. If your operating system is in English, then your default keyboard layout is possibly U.S. English.

But you can install (in your System Preferences or Control Panel) additional keyboard layouts. Then in the taskbar you’ll see a switcher for the keyboard layouts.

If you have the U.S. English and Greek keyboard layouts installed, and the U.S. English layout is active and you press the A key, that key press translates to the Unicode Latin letter »a«, so that letter gets inserted into your text editor. But if you activate the Greek keyboard layout and press A, that gets translated into the Unicode Greek letter alpha, so »α« gets inserted into the text.

Each Unicode writing system and language typically needs its dedicated keyboard layout. Operating systems don’t have have dedicated keyboard layouts for all writing systems (scripts). So many scripts require some additional software which will produce the right Unicodes — this may be an app that installs additional keyboard layouts (or even lets you create your own key mappings), or this can be the Character Map / Character Picker app that comes with your system, where you can click the letters with your mouse, or this can be a special website that includes an HTML character picker or visual keyboard dedicated to the given writing system.

@twardoch
Copy link

twardoch commented Apr 8, 2022

Because this is quite complicated, some people have created shortcut solutions, for example fonts where the letters look like Cuneiform or Greek, but actually are mapped to the Unicodes of the Latin script. This is typically called »hack encoding«.

Hack encoding may look like a simpler solution in the short run, but ultimately, it isn’t. Unicode (where each writing system has its own codes) is much better long-term, and Unicode fonts often have built-in mechanisms that allow automatic formation of ligatures or selection of the right letter forms depending on the surrounding letters (this is important for many writing systems).

All Noto fonts follow Unicode, so in order to use them, you need some typing solution (keyboard layout, visual kayboard, character map / picker app etc.) that supports Unicode.

@Mercury13
Copy link

Mercury13 commented Apr 8, 2022

@hreyes82, to copy cuneiform chars, you need to open some program that shows them all and copy one by one via clipboard. They have separate codes differing from Latin — unlike all those Wingdings and Webdings.
You may use https://unicode-table.com, it shows lots of character samples — not all but most, including 90% of cuneiform (Ugaritic, Persian, Sumerian — except early dynastic).
Or I’m writing https://mercury13.github.io/unicodia/ — it has ALL character samples of Unicode 14, but right now it’s Russian only. I’m going to globalize it, listen to my updates. The Big Task is to host Russian in separate language resource, not in EXE, and then I’ll start to translate.
If you install the font, you can use this: https://babelstone.co.uk/Software/BabelMap.html Unlike my Unicodia, it does not bundle all fonts to prevent tofu. Well, Noto is what my Unicodia bundles for cuneiform.
Windows 10 also has all cuneiform except early dynastic from the same vendor (Monotype).

@Mercury13
Copy link

Two updates.

  1. Unicode 14.
  2. https://r12a.github.io/uniview/ is also a cool thing.

@Mercury13
Copy link

@hreyes82, wait for updates of my Unicodia. In ≈2 months if I do not perish (there’s a war in Ukraine) I’ll globalize it.

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

No branches or pull requests

6 participants