Help With Non-English Character Encoding

Discussion started on Other Topics

There are 50 characters in my alphabet 26 of which are okay. But for the remaining 24 I want to be able to type by pressing Caps Lock + Shift key + from A-X but I have no idea how to do this- any ideas?
All considerations much appreciated!
#1 - 2014-11-02, 12:51
« Last Edit: 2014-11-03, 11:02 by Thomas Phinney (FontLab) »

What is your alphabet? What platform are you on?

Probably, what you need is a custom keyboard layout.

Mac OS:
#2 - 2014-11-03, 10:52

The alphabet is not in use. It has been developed from spoken language in Ethiopia but at present there is not a written alphabet for our language. We use Latin. I’ve started from scratch and have developed an alphabet but I have no idea how to create the computerised version of this alphabet and then encode it to be able to use it on the internet and make it accessible globally.

Many thanks

#3 - 2014-11-08, 14:19

Ah, that is harder then!

So, this is a multi-part answer.

- for the long term, the writing system needs to be encoded in Unicode. There probably needs to be some evidence of demand and usage (or desired usage) by a number of people to help that happen. After that, somebody will need to put together a proposal to the Unicode consortium to encode the characters.

- in the short term, you'll need to either use Private Use Area codepoints, or map the letters to Latin. Both approaches have pluses and minuses, but I recommend the PUA approach. If you use PUA codepoints, it will be easier to convert text encoded that way to “real” codepoints later, because there will be no ambiguity. If you have your language's text using “false encoded” Latin, it will have to be checked carefully to make sure that any real English (or other western language) words are in your text are preserved in re-coding.

What is the name of the language, btw? Why do you think it needs a newly invented alphabet, and that such an alphabet will become popular? I am not arguing against it, but I will point out that when there is already a written form of a language, new proposals are rarely successful unless imposed by a central authority.

Good luck,

#4 - 2014-11-08, 18:10

Thank you for your response. It is very helpful. I am currently writing a business plan to prove the need.
The alphabet I have developed is for one of the Ethiopian languages that currently uses Latin for its written version. However Latin cannot successfully reproduce all the sounds required by the language.

If you can help me further how do I go about contacting the Unicode consortium? Also currently the characters I have designed are in Illustrator. Do you know how I would begin to work with PUA code points?

I appreciate any help you can give or if possible sign post me.
#5 - 2014-11-11, 15:24

Of course, Latin has nearly infinite diacritical (accent) marks and the like available, so I am not at all convinced that there are sounds you can’t represent in Latin, with combining accents. But it’s your language, so I will leave you to determine that sort of thing.

You should go to the Unicode site and start reading a bit about Unicode. Here are some particular pages I suggest:


These folks might help you with an encoding proposal:

If not them, perhaps Michael Everson.

After you've done some reading, I can point you at email addresses and the like, if necessary.
#6 - 2014-11-11, 16:59

Thank you for your help that’s great.

I have been on the Unicode site a lot but I have found it difficult to find specific information to guide me through this process.

I have started to follow up on the two other leads and that is very helpful thank you.
I will be in touch with you soon.

I am confident about my language and the need for the system I have produced.

I look forward to sharing with you.
#7 - 2014-11-15, 06:16


0 Members and 1 Guest are viewing this topic.