Vietnamese misdisplay (was: Re: Vietnamese)

[Clytie Siddall <clytie@riverland.net.au>]

On 18/02/2006, at 11:18 PM, Davide Viti wrote:

> > The other possibility is one of us shooting a few screenshots and
> > putting them on a web page.
>
> just uploaded such an image (see [2])

Thanks, Davide, that saves me a lot of time and messing around. :)

But the news is not good. :(

It's very odd: most of the Vietnamese characters are displaying  
correctly, except for some vowels, and their tonal varieties (each  
vowel has six forms, one for each tone marked).

These are simply omitted.

Which results in text like:
____
It's vry odd: all th Vitnms chrctrs ar displying corrctly, xcpt for  
th two vowls with a hook on th sid, and all thir tonl vritis.
____

which is much harder to guess in Vietnamese, since all our words are  
very short, monosyllabic, and composed mostly of vowels. Vowels are  
much more important than consonants in our language (in contrast to  
English).

I have not seen this type of display problem before (which is  
strange, because Vietnamese is a test case for Unicode support, so  
I've become a de facto Unicode beta tester).

Usually, if there is a Unicode input/display problem, the text will  
not display (or misdisplay) _all_ the non-Latin- 1 characters, or  
_all_ the combined-diacritics characters.

One interesting problem with Mac Carbon-based apps. separated the  
vowel and the accent. In one app., the accents would occasionally  
even separate entirely from the parent vowel, and follow the cursor  
all over the page! They were not deletable. I've seen a lot of  
Vietnamese input/display problems, but this one in the Debian  
Installer is new. (just what we need right now... :( )

Could it be a font problem? You need a full Unicode font to display  
Vietnamese. Many so-called "Unicode fonts" don't cover the whole  
range, and we have some uncommon characters and combinations. Full  
Unicode fonts include Lucida Grande, Arial, Everson Mono, Gentium,  
Junicode and Titus Cyberbit Basic, plus the Vietnamese-designed  
Unicode fonts: http://www.vn.net/sections.php? 
op=listarticles&secid=4

I can attach any of those fonts to a message, if that is useful.

Just to check if this is specific to someone's installation,  
Christian, do you get the same text displayed? Anyone who is willing  
to check, please do.

Here is a comparison:

The first line should be:

√ Trình đơn cài đặt chính của Debian. (The Debian  
Installer main menu.)

but is:

X Trình đn cài đặt chính ca Debian ([meaningless] install main  
[meaningless] Debian)

which omits two vowels: ơ and ủ

The second line should be:

√ Hãy chọn bước tiếp theo trong quá trình cài đặt:  
(Choose the next step in the installation process:)

but is:

X Hãy chọn bc tiếp theo trong quá trình cài đặt: (Choose  
[meaningless] following in the installation process:)

omitting two vowels, ư and ớ

Other vowels are omitted in the rest of the text.

I hope we can work this out. :(

from Clytie (vi-VN, Vietnamese free-software translation team / nhóm  
Việt hóa phần mềm tự do)
http://groups-beta.google.com/group/vi-VN

Attachment: PGP.sig
Description: This is a digitally signed message part