Re: Internationalization (was [w3c-wai-ig] <none>)

> the page should of course contain unicode-encoded text, not graphics.

Unfortunately you still have to consider the target audience.  Unless there
is a very large percentage penetration of Window XP, in many countries
most users will be set up to handle their own language only as images
and will do most serious work in English.

I'm not even sure how complete the Windows XP fonts' coverage of 
Indian languages is.  (Note that, although some web sites suggest
downloading the Word 2000 Unicode font and using it with Windows 98/IE,
that's actually iilegal, as the font is only licensed for use with
MS Office; most people fail to read the click through licence agreement.)

If I wanted to target the Chinese community in England, I would estimate
only 5 to 10% support for Unicoded material and only marginally more for
misrepresenting 8 bit fonts.  In practice, for those with web access,
English would be the language of choice, but if the content was, say,
about the weekend school where I'm learning Chinese (they don't actually
have a web site), they might want to produce Chinese for effect (they
use hardcopy English and Chinese at the moment - most of the students
are from Chinese families).  I would think any Chinese web page they
produced would have to be either entirely images++ or in an image version
and a proper text (Big5 on the wire, not Unicode) version in parallel.
Attempts to get people in my class to install fonts to give them access
to Chinese resources on the web have met with rather limited success (most
in the adult class are more interested in conversation than reading).

(Chinese in Chinese speaking coutries is sent in text format, but
maybe half of it has an undeclared character set, so assumes a browser
that is misrepresenting a font's encoding, or that the browser's heuristic
for this particular standards violation is to assume GB2312 or Big5, 
depending on the country.  The origins of this, like for most other
non-Latin 1 fonts, is improvising ways of using early, non-character
set aware, browsers; for users within a closed community, it is
no problem, so there is no great pressure to generate valid pages.)

++ Given the dominance of Internet Explorer, one could use embedded
fonts and reach a lot of people, but most authors would not 
appreciate that option.

Received on Friday, 3 October 2003 02:28:02 UTC