Re: UTF-8 and RFC 2047

New Message Reply About this list Date view Thread view Subject view Author view

From: Russ Allbery (rra@stanford.edu)
Date: Sun Jun 30 2002 - 16:25:24 CDT


Erland Sommarskog <sommar-usefor@algonet.se> writes:
> Charles Lindsey <chl@clw.cs.man.ac.uk> writes:

>> and they MAY, when it is detected that neither has been used,
>> attempt to interpet the header according to whatever other character
>> set can be deduced, or has been configued as a default by the
>> reader.

> This is something I like to be strengthened, but I don't know really
> how. SHOULD is definitely not right. "should as a matter of good
> practice" maybe.

> What I'm trying to get at is that the standard should deprecate all
> behaviour like the one in Cyrus. Thus an agent should accept the illegal
> text and try to make much sense of it as possible. The only exception is
> when the agent knows that real harm can occur. For instance, a reading
> agent that knows that its writes to a VT320 tty, should probably filter
> CSI (char 155), as this is a control character for the tty. But, as far
> as I know, a Windows newsreader, never have to make such considerations.

Something like:

   It is recommended, as a last recourse, that characters in unknown
   character sets be passed unaltered and displayed in the default
   character set so long as they are not control characters in that
   character set. This is better than altering or rejecting the
   characters since the user will at least have some chance of making
   sense of the text.

Perhaps?

-- 
Russ Allbery (rra@stanford.edu)             <http://www.eyrie.org/~eagle/>


New Message Reply About this list Date view Thread view Subject view Author view


This archive was generated by hypermail 2b29.