Re: [Nmh-workers] Handling non-ASCII

[Earl Hood]

On Mon, Apr 16, 2012 at 12:27 PM, Lyndon Nerenberg wrote:

>>    Updated Estimate: .3,141.00
>>    Updated Estimate: ?3,141.00
>
> Presumably there is a loud warning being printed to say "***danger
> corrupt message - unprintable characters replaced  with '.'"

Such warnings can be easily overlooked depending on how the warning is
issued.  With that said, a warning should be printed that invalid
character sequences were encountered.

IMO, I prefer '?' or U+FFFD, since this is how other applications behave
when receiving invalid character sequences.  U+FFFD is the best if your
output device support unicode, but '?' can be used if dealing with a
non-unicode-aware devices.

Side Note: In software I maintain, I provide the option to specify what
the default character encoding is.  The default is US-ASCII, but there
are environments that use other default encodings in their email, even
though such encodings are not MIME encoded.  Unsure if nmh should
provide something similar; and mh-profile setting maybe.

--ewh