Vaska, you¨re still mixing those:

I think you are mixing two things which should be separated.

The first problem is the language of the page (defined in the header)
The second problem is how to create a non-ascii character

He is right.

It is a tricky business because for a French typist I can use entities and change an é into é

It's wise to use codepage that contain this character, or better UTF.

but with Chinese everything comes up unreadable (as you've mentioned)

Even when using Unicode?

There will be a situation where one page will have the header encoding in ZH and an input/text field as EN-US. I'm pretty sure that the field itself won't establish the language parameters that go into the field - the operating system will.

No, the browser will. It will send the characters in the encoding (charset, not language!) of the page.

One thing I don't understand though, is at what point does the computer actually use the xml:lang attribute? At the input (client-side)? When it gets to the server/table (server-side)? I can type any language I want into the textarea, but what comes out can vary...

The 'lang' attrib is mostly for screen readers, CSS language tools and some processing applications. It doesn't determine the way how characters are inputed/printed/transfered. That's a part for charset.

What, where, which formats do I use and stick with if the idea is to support just about any lanugage that's out there (theoretically)?

Some Unicode - I don't know how it works with Asian/Arabic/Hebrew - whether UTF8, 16 or 32, what about the Endians etc. ...

--
Jan Brasna aka JohnyB :: www.alphanumeric.cz | www.janbrasna.com
******************************************************
The discussion list for  http://webstandardsgroup.org/

See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list & getting help
******************************************************

Reply via email to