From: Forrest J. Cavalier III (mibsoft@epix.net)
Date: Mon Oct 07 2002 - 13:12:18 CDT
"Charles Lindsey" <chl@clw.cs.man.ac.uk>
> But to insist that arabic/chinese/whatever newsgroups should be in
> perpetuity stuck with tags in ASCII that are largely meaningless to their
> readers, and with an added hierarchy of translators and lookup mechanisms
> on top, is baroqueness for its own sake. And all to avoid a problem that
> does not exist - namely to force a 7bit structure onto a transport that is
> already 8bit clean.
>
OK, I'll play.
To insist the 8th bit should be used in newsgroup names and headers,
sticking other systems with the translation problem (which is
yet to be solved by the drafters, even in theory!), and to insist that
there be no backwards-compatibility in gateways, server software,
and readers is requiring upgraded technology for its own sake.
> And in the few places where interoperability with mail is required,
> notably sending to moderators, it will be no solution, because what
> the moderator will see is the ASCII tag, not the "real" name of the
> group (because you cannot expect mail to understand the machinery for
> translating it into 'prettyname' form).
OK, I'll play again.
And in the few places where interoperability with mail is required,
notably sending to moderators, 5.5.2 is no solution, because what
the moderator will see is the encoded Newsgroup name, not the UTF-8
name of the group (because you cannot expect mail to understand the
machinery for translating 5.5.2 encodings into UTF-8 form.)
Let's be realistic. I'm sure that 5.5.2 encoding or the ASCII tag will cause
the moderator, who knowingly volunteered to moderate the I18N group,
less than 1 second's worth of pause, on the very first message they process.
(For all other messges, it will be immediately obvious what they are
getting and what to do with it.) But if you don't like the ASCII
tag to moderators, you had better draft a different solution in 5.5.2.