Re: ISO/IEC 10646 as Document Character Set

Gavin Nicol (gtn@ebt.com)
Wed, 3 May 95 19:02:15 EDT

>If we agree that 8859-1 is to be assumed as the default representation
>character set and this goes into the standard, then all we have to do
>is point out to Japanese web server operators (or web server operators
>elsewhere who want to use another character set to represent the
>entity-body), that, if they want their documents to be understood, they
>damned well better start specifying the proper encoding in the HTTP
>response header.

This is the <EM>only</EM> reasonable course. Anything else is bound
to fail.

>The idea of tagging the charset in the document simply won't work as a
>reliable indicator.

Absolutely, and sniffing at data to figure out it's encoding and coded
character set is a long path leading to madness.