Unicode requirement

List overview All Threads
Download

newer

older

Some updates

Re: [Langcom] Unicode requirement

Amir E. Aharoni

14 Apr 2019 14 Apr '19

6:35 p.m.

Attachments:

attachment.htm (text/html — 480 bytes)

Show replies by thread

Gerard Meijssen

14 Apr 14 Apr

6:41 p.m.

Hoi, In the past the WMF has funded the creation of a Unicode font. Having a Unicode font is essential when we are to support it in MediaWiki. Not having a fully developed font is what hinders the necessary follow up of projects in SignWriting ie all the signed languages. I do agree that a language with a default script not supported in Unicode is hugely problematic. Thanks, GerardM On Sun, 14 Apr 2019 at 18:36, Amir E. Aharoni <amir.aharoni(a)mail.huji.ac.il> wrote:

...

Hi, I have a vague recollection that we require Unicode support to create a new language, but I cannot find it in the policy. Do we indeed require this explicitly, or am I just making things up? Or is it just a de facto practicality—that it's technically difficult to host a language in a script that isn't supposed in Unicode? _______________________________________________ Langcom mailing list Langcom(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/langcom

Amir E. Aharoni

6:49 p.m.

I'm not talking a missing font, even though this would be a problem, too. I'm talking about scripts that aren't encoded at all, for example the Zaghawa alphabet. And I understand, of course, that it's technically problematic. What I'm asking is whether there's an explicit written Language committee policy about it, or is it just a de facto practical matter. בתאריך יום א׳, 14 באפר׳ 2019, 19:41, מאת Gerard Meijssen ‏< gerard.meijssen(a)gmail.com>gt;:

...

_______________________________________________ Langcom mailing list Langcom(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/langcom

Gerard Meijssen

6:58 p.m.

Hoi, Unicode has funding drives where they ask the PUBLIC to support one character.. Why not have the WMF fund the missing characters in scripts we need for our projects? Thanks, GerardM On Sun, 14 Apr 2019 at 18:49, Amir E. Aharoni <amir.aharoni(a)mail.huji.ac.il> wrote:

...

_______________________________________________ Langcom mailing list Langcom(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/langcom

MF-Warburg

7:28 p.m.

For all I know, it is only a de facto matter. Amir E. Aharoni <amir.aharoni(a)mail.huji.ac.il> schrieb am So., 14. Apr. 2019, 18:49:

...

_______________________________________________ Langcom mailing list Langcom(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/langcom

Michael Everson

15 Apr 15 Apr

3:02 a.m.

On 14 Apr 2019, at 17:35, Amir E. Aharoni <amir.aharoni(a)mail.huji.ac.il> wrote:

...

Or is it just a de facto practicality—that it's technically difficult to host a language in a script that isn't supposed in Unicode?

It is impossible to do so reliably and interchangeably. Michael

Michael Everson

3:02 a.m.

On 14 Apr 2019, at 17:41, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

...

In the past the WMF has funded the creation of a Unicode font.

When?

...

Having a Unicode font is essential when we are to support it in MediaWiki. Not having a fully developed font is what hinders the necessary follow up of projects in SignWriting ie all the signed languages.

Actually there’s a reasonable font out there, though there’s lots of difficulties. I worked on the code chart font for SignWriting.

...

I do agree that a language with a default script not supported in Unicode is hugely problematic.

It could only be Private Use Characters. If a script exists, it should be encoded. Encoding isn’t free. Michael

Michael Everson

3:04 a.m.

On 14 Apr 2019, at 17:49, Amir E. Aharoni <amir.aharoni(a)mail.huji.ac.il> wrote:

...

I'm not talking a missing font, even though this would be a problem, too. I'm talking about scripts that aren't encoded at all, for example the Zaghawa alphabet.

We call that Beria. It’s not encoded. If WM needs it encoded, WM should help make that happen.

...

And I understand, of course, that it's technically problematic. What I'm asking is whether there's an explicit written Language committee policy about it, or is it just a de facto practical matter.

A Beria Wiki would use PUA characters and all of that data would have to be transcoded later. Michael

Michael Everson

3:07 a.m.

On 14 Apr 2019, at 17:58, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

...

Unicode has funding drives where they ask the PUBLIC to support one character.. Why not have the WMF fund the missing characters in scripts we need for our projects?

I’ve suggested this for years. At the first Hackathon I attended in Berlin, for example. It’s not as though you guys don’t know some sort of expert. Michael

Michael Everson

3:07 a.m.

On 14 Apr 2019, at 18:28, MF-Warburg <mfwarburg(a)googlemail.com> wrote:

...

For all I know, it is only a de facto matter.

It would not be good for a language to have an encyclopaedia in PUA characters, and their browsers would perpetually be substituting CJK characters encoded in that space too. Michael

Gerard Meijssen

7:28 a.m.

Hoi, The WMF does have the technology to provide the font from its servers. It is not the case that characters would not show properly. Thanks, GerardM On Mon, 15 Apr 2019 at 03:07, Michael Everson <everson(a)evertype.com> wrote:

...

On 14 Apr 2019, at 18:28, MF-Warburg <mfwarburg(a)googlemail.com> wrote:

For all I know, it is only a de facto matter.

It would not be good for a language to have an encyclopaedia in PUA characters, and their browsers would perpetually be substituting CJK characters encoded in that space too. Michael _______________________________________________ Langcom mailing list Langcom(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/langcom

Michael Everson

8:08 p.m.

On 15 Apr 2019, at 06:28, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

...

The WMF does have the technology to provide the font from its servers. It is not the case that characters would not show properly.

If something is worth doing, it is worth doing right. If a language is worthy of a Wikipedia, its characters are worth encoding. I will vehemently oppose any Wikipedia which has to use the PUA to display text. Michael Everson

Phake Nick

9:23 p.m.

Gerard Meijssen <gerard.meijssen(a)gmail.com> 於 2019年4月15日週一下午1:28寫道：

...

Hoi, The WMF does have the technology to provide the font from its servers. It is not the case that characters would not show properly. Thanks, GerardM

How? The topic have been discussed on Chinese Wikipedia for months last year (See: https://zh.wikipedia.org/wiki/Template_talk:%E7%BC%BA%E5%AD%97 ) yet there are still no implementable solution to solve the problem of displaying Chinese characters that are either unencoded by Unicode yet or newly encoded but without sufficient font support yet Michael Everson <everson(a)evertype.com> 於 2019年4月16日週二上午2:08寫道：

...

On 15 Apr 2019, at 06:28, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:

The WMF does have the technology to provide the font from its servers. It is not the case that characters would not show properly.

Unicode's priority is not the same as Wikipedia/WMF. For example, there are quite a few Chinese characters that are used in various different Chinese Wikipedia article but Unicode either haven't encoded them or they have already refused to encode those characters based on various reasons.

1835

days inactive

1836

days old

langcom@lists.wikimedia.org

Manage subscription

12 comments

5 participants

tags (0)

participants (5)

Amir E. Aharoni
Gerard Meijssen
MF-Warburg
Michael Everson
Phake Nick