From: Erland Sommarskog (sommar-usefor@algonet.se)
Date: Sat Jul 21 2001 - 14:43:21 CDT
Jean-Marc Desperrier <jean-marc.desperrier@certplus.com> writes:
> But the problem is that when USEFOR is standardised, the use of charsets
> will no more be homogenous.
>
> There will be people using utf-8, and other using local encoding for
> headers in the same group.
Yes, there will be a transition period. But we have been through this
before, when we went from 7-bit to 8-bit.
But in difference to that transition, newsreaders can actually apply
some heuristics. A newsreader can easily detect that a string which is
supposed to be UTF-8 is not, and can thus in such case use a local charset
as a fallback.
And RFC2047 only solves this half-way. Readers that don't grasp UTF-8
will need to be rewritten anyway, and while rewriting the authors
can implement the fallback.
> > Not entirely correct. It is certainly true for a popular reader like
> > Free Agent. However, I expect UTF-8 support for my mail reader long
> > before RFC2047 support. To wit, I use mailx in a Telnet window, and all I need is UTF-8
> > support in my Telnet client.
>
> But getting UTF-8 by loosing ISO-8859-1 or ISO-8859-2 support is not very
> great.
>
> I'd be very surprised if mailx is able to convert your message from
> ISO-8859-x to UTF-8, and does not handle RFC 2047.
No, but I will be able to it. To wit, the Telnet client will still
include support for Latin-1, and with a few key presses I can change
the charset. I might even be able to persuade the Telnet author to
implement a shortcut key. If nothing else, there are macro progrocessors
out there that I could use.
When we went from 7-bit to 8-bit, changing my terminal forth and back
from 7-bit to 8-bit was a matter of routine.
-- Erland Sommarskog, Stockholm, sommar@algonet.se