From: J.B. Moreno (planb@newsreaders.com)
Date: Wed Mar 05 2003 - 23:51:43 CST
On 3/5/03 11:34 PM, Russ Allbery at <rra@stanford.edu> wrote:
> J B Moreno <planb@newsreaders.com> writes:
>> On 3/5/03 5:16 PM, Russ Allbery at <rra@stanford.edu> wrote:
>
>>> That 15% is not using UTF-8.
>
>>> Both solutions ignore (read: do not standardize the current practices
>>> of) that 15%. No one is proposing making that 15% happy, including
>>> you. So this is a straw man.
>
>> It's a lot closer to their current practice than 2047 is,
>
> No. It's not. This is flatly, completely untrue.
>
> UTF-8 requires that they completely change character sets, possibly give
> up the idea of a standard character width, figure out how to use a Unicode
> font, or at the least link with full-blown character set conversion
> software.
They character set they are changing to will include their current character
set as a subset -- that means it's hardly a hardship.
Standard character width and how to use a unicode font are irrelevant --
when I want to post using UTF8, I go to the menu and chose "Unicode (UTF-8)"
and everything else just happens. If I want to see Russian or Chinese
characters I need to install a unicode font, but for the characters that i
can get by typing, I don't need to do anything special.
Linking with character set conversion software is NOT what most people will
be doing.
> RFC 2047 just requires that they put an encoding and decoding step in
> front of their existing practice, plus the decoding part and possibly the
> encoding part is generally already done in the software that they're using
> right now.
And when someone complains and says that =?ISO-8859-15?Q?erron=E9s?= is
completely unreadable? Sure that won't happen very often, but ten year old
software is still being used. Now it won't display the UTF-8 correctly
either, but it won't mess it up as much either: erronés.
-- J.B. Moreno