In your message of 24 Aug (09:24 hrs) to Andrew MacIntyre, you wrote:

 AM> FWIW, my understanding is that all "soft" CRs should be 
 AM> stripped from the message when it is stored.  They are only 
 AM> a convenience for a message editor, and whenever someone 
 AM> has different margins the message editor should reformat 
 AM> the paragraph to suit anyway.

 PE> The trouble is that 0x8d has meaning to the Japanese, and I can't 
 PE> tell whether it's a Japanese character or a soft CR.  This is 
 PE> what happens when you use non-control characters for control 
 PE> characters.  :-(

David Nugent's suggestion of ignoring them (ie treat them as part of the
data stream) seems a good idea.  If your software has to make no
interpretive decisions so much the better.  Is there some other sequence
that you might be able to make an informed guess as to whether the data is
Japanese (or whatever...) rather than ASCII?

Regards, Andrew.

--- msgedsq 2.0.5