format=flowed involves putting a single space at the end of line text
which is not the end of a paragraph.  When 'decoding' a message
so-labelled it would be plausible for Mutt to remove such trailing
spaces.  It would also be plausible for it to leave them in.

When is hateful is for it to _double_ such spaces, so that continued
lines now end in two space characters -- and meaning that if I'm quoting
somebody else's message, I'm quoting characters that neither she nor her
mailer inserted.

Moreover, if I attempt to reflow such a paragraph (in the old-fasioned
way, ignoring f=f) by using Vim's gq command words with spurious, and
curous, double spaces between them can end up in the middle of lines.

Vim can cope with ">"s denoting quoted paragraphs, leaving the quote
characters at the start of each line while flowing words around them.
And it handles fine either zero or one space characters at the end of a
line.  But if a line ends in multiple space characters then it presumes,
not unreasonably, that they are there for a purpose, so preserves them
in the reflowed text -- yielding the gaps between words.

I'd spotted that messages which go gappy on quoting were all sent by
Thunderbird, so I was about to hate Thunderbird for inserting the
spurious spaces.  But it appears Thunderbird is following the rules (it
just that only all the f=f mail I receive is from Thunderbird users) --
the hate for the gaps in quoting is definitely with Mutt.

(You can see it by viewing a f=f message then piping it to an external
viewer with pipe_decode=no and then pipe_decode=yes.)
 
Smylers

Reply via email to