Kai Henningsen <kaih(_at_)khms(_dot_)westfalen(_dot_)de> writes:
rra(_at_)stanford(_dot_)edu (Russ Allbery) wrote:
Quite to the contrary, that's a summary of the basic foundation of the
current draft,
Don't be silly.
I'm not.
and is exactly what has been argued by multiple people on the USEFOR
list on multiple occasions.
Not where I have seen it, and I've been in a number of those threads.
I suppose we'll have to agree to disagree, then, as I've also been in a
number of those threads. It doesn't really matter whether or not my
characterization is accurate, but I'm not going to withdraw it, as it
remains my honest opinion.
What *has* been argued is that *if* USEFOR says naked UTF-8 in headers
is ok, then interpreting unencoded news headers as UTF-8 is going to be
pretty much universal pretty fast - note the sharply limited domain here
- and that, as the idea has now been around for a few years, some people
already started writing their software along those lines.
That is *very* far from "any untagged 8-bit data is UTF-8".
The same was also argued about article bodies, and if you put headers
together with bodies, you have, precisely, "any untagged 8-bit data."
There is, of course, very little point in not tagging the body data, and I
don't believe anyone was arguing in favor of leaving off the charset tag.
Maybe that's where you're disagreeing with me.
The part of this with which I disagree is the belief that if some standard
is issued stating that unencoded UTF-8 is allowed in headers, software
will immediately start changing to send and interpret that. I don't
believe the level of influence that a new standard would have at this
point is that high, and use of other character sets (without labelling) is
extremely common right now. It is possible to attempt to distinguish
between various untagged character sets, but I'm rather less optimistic
that anyone's going to rush out to write that code because we suggest it
in a standard.
I realize that the Turnpike folks have already done this, so perhaps that
indicates that I'm wrong.
--
Russ Allbery (rra(_at_)stanford(_dot_)edu)
<http://www.eyrie.org/~eagle/>