🚀 go-pugleaf

In article <BzBpxG.D93@zoo.toronto.edu> henry@zoo.toronto.edu (Henry Spencer) writes:

> The one thing that needs careful thought is adapting news to use of
> RFC-1342 for encoding non-ASCII stuff in *headers*.

Agreed.

>  The MIME (RFC-1341) facilities for use in the *body*, for stating
> which character set you are using and encoding >7-bit sets to survive
> 7-bit transmission channels, appear to be directly applicable without
> modification.  I'm debating how much detail to include about them in
> the draft, but the bottom line will be "see MIME".  Almost everything
> in MIME has potential applications to news, and I don't see any
> compelling reason to re-invent or re-specify it.

Sounds good.  

  There might be some legitimate room to address the question of which
character sets are "preferred" for use in articles while being
transported between sites.  There will be more long-term
interoperability if the "preferred" transport character set(s) are
minimised.

  In MIME, the intent was to encourage use of ISO-8859-* and ISO-10646
over all else for longer term interoperability reasons.  In the short
term, it was recognised that other existing practices (e.g. VISCII,
ISO-2022-JP, -KR) would also be used within some communities until 
better alternatives came along.

  The character set used for transport need not be native to the
machine used to display the article.  It only needs to be translatable
to some character set on the system.  X11-based news readers will be
especially able to display multilingual text and images, but even
other devices can probably display correctly using other mechanisms
(e.g. PCs can switch code pages or use MS-Windows or whatever).

Ran
atkinson@itd.nrl.navy.mil