Re: PROPOSAL: i74: Encoding for non-ASCII headers

I agree with everything Julian said, except:

Julian Reschke wrote:
> TEXT already allows C1 controls (and always did) 
> (<http://greenbytes.de/tech/webdav/draft-ietf-httpbis-p1-messaging-latest.html#rule.TEXT>):
> 
>   TEXT           = %x20-7E | %x80-FF | LWS
>                  ; any OCTET except CTLs, but including LWS
> 
> That being said, I'd like it to exclude C1 controls.

If C1 controls _in the form of octets %x80-9F_ are excluded, and HTTP
agents begin to reject TEXT containing those octets, it will be harder
to transition to UTF-8 later.  (In case you'd forgotton, UTF-8 uses
those octet values for normal characters).

In other words, rejecting %0x80-9F pretty much commits the high octets
of TEXT to representing iso-8859-1.  Even if nobody uses it (like now
it seems), there may be agents which don't send it, but reject
%0x80-9F if it's recommended by an HTTP RFC.

Excluding C1 controls _encoded as UTF-8_ is quite reasonable.  But
then, there are lots of other controls one might wish to exclude too -
preferably by saying senders SHOULD NOT send them, but perhaps
receivers shouldn't reject them.

-- Jamie