Re: [Nmh-workers] Handling non-ASCII

[Ken Hornstein]

>What about non-utf8 multi-byte encodings.  Can 2022 style encodings
>have embedded printable ASCII characters?  (Thankfully I've been away
>from that stuff long enough to have forgotten.)

I believe all of the non-spam 2022 stuff you'd see would be properly
encoded, right?  So we shouldn't care unless it's common for 2022
users to send messages that don't have RFC 2047-encoding.  But to
answer your original question .... I think that the answer is "yes".

>Also, I would convert the unprintable characters to '.' for display.
>The \xyz escapes are visually cluttering and obscure the viable text.

Also, in a scan listing it would mess up the alignment of the scan output
unless you put that smarts in the format engine.  And THAT is not something
I'm personally interested in doing.

I'm not so crazy about "." as a replacement character, though.  How about
this: if your locale is UTF-8 we use the standard UTF-8 replacement
character (U+FFFD, rendered as “�”), and any other locale we use a regular
old "?".  That seems more intuitive to me.

--Ken