From: Charles Lindsey (chl@clw.cs.man.ac.uk)
Date: Sun Nov 17 2002 - 09:18:09 CST
In <87y97vck7y.fsf@erlenstar.demon.co.uk> Andrew Gierth <andrew@erlenstar.demon.co.uk> writes:
>>>>>> "Charles" == Charles Lindsey <chl@clw.cs.man.ac.uk> writes:
> Charles> Issue 1 - (8BIT HEADERS)
>[snip alternatives]
>This excludes an extremely important alternative:
>a) UTF-8 is allowed in newsgroup names only, not in any other header
> content
Well nobody asked for that in response to the three times I published the
list of issues. Actually, I think it is an even stricter variant on #1, in
which you actually forbid the use of UTF-8 in places where RFC 2047/2231
could be used, which just leaves the Newgroups in UTF-8. That would indeed
mean that there were fewer thing to go wrong when communicating with
moderators.
>b) newsgroup names are 8bit on the wire but encoded in some form when
> sent by mail
That is essentially what the present draft says, so it is #0.
>c) people will inevitably continue to use 8bit in other headers in a
> variety of charsets, which should not be explicitly permitted in the
> draft due to the known conflicts with mail, but which will remain
> standard practice
I think we are all agreed that practice will not be compliant whatever
route we follow. Officially, they must use UTF-8 or RFC 2047 (whichever we
decide), but of course it will be a long time a'dying, which is why we
pointed out that implementations could test for the presence of UTF-8 in
order to ease the transition. If we go for one of the MUST encode
solutions, then that suggestion will no longer be relevant (but there
would be some other wording to cover the situation, no doubt).
-- Charles H. Lindsey ---------At Home, doing my own thing------------------------ Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl Email: chl@clw.cs.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5