Re: HTML Character Representation/Transmission Model

Gavin Nicol (gtn@ebt.com)
Tue, 11 Apr 95 09:16:08 EDT

>> The default character set must remain 8859-1 and numeric character
>> references must continue to map to 8859-1, this has been the
>> default since the creation of HTML and must continue to be the default.
>
..
>Yes; once clearly explained, the concept of a single document character
>set seems a clean and workable solution.

I am sorry for my lack of clarity: perhaps living in Japan so long has
robbed me of the ability to communicate in English well...

>Lou, I think I hear you saying there must be no nasty surprises and
>numeric character references must continue to give the same printed
>characters they did before?

For ISO-8859-1 documents, this is not a problem. However, my proposal
may break Mosaic-L10N, and some of the other non-ISO-8859-1
browsers. The behaviour of such browsers is undefined by the current
HTML specification anyway, but...

>So, a change to the wording which had precisely no effect on current
>documents but set us up better for the future would be a worthwhile
>change, no?

I wouldn't say "no effect". The overall effect should be minimal
from a practical point of view, but profound in the long term.