[Prev][Next][Index][Thread]

Re: LANG= for character-mapping



David Goldsmith wrote:

>>Do we have to publish it in Unicode instead then? -- ie. let most
>>browsers just break and wait for the *perfect browser* to come along.
>>I don't think so.
>
>Actually, I believe the situation is closer to the reverse: I don't know 
>any browsers which will switch character sets in the middle of a page, 
>but both Netscape and Microsoft's browsers will be supporting Unicode in 
>their next versions, based on what I've heard. Microsoft has already 
>demonstrated this capability in theirs. Other browsers will be doing so 
>too, I believe.
>
>You are more likely to see browsers supporting Unicode before they 
>support charset switching, so I'd recommend publishing your pages that 
>way.

I would even go further: Switching of character encodings with something
like markup inside HTML will never be done. It is very clumsy to handle
and absolutely not necessary. Character sets can already be switched on
the level of a MIME "charset" character encoding by established techniques.
For example, for charset=ISO-2022-JP, you are constantly switching
between ASCII and JIS 208 by using appropriate escape sequences.
But this is done at a level that does not involve HTML or SGML, i.e.
before parsing, thus keeping things neatly separated.

Regards,	Martin.


Follow-Ups: References: