In <ilun0at1172(_dot_)fsf(_at_)latte(_dot_)josefsson(_dot_)org> Simon Josefsson
<simon+ietf-822(_at_)josefsson(_dot_)org> writes:
"Charles Lindsey" <chl(_at_)clerew(_dot_)man(_dot_)ac(_dot_)uk> writes:
But I still think that MUST is too strong. It is going to happen whether
you like it or not, so you have to warn of the dangers and give the aware
clients a mechanism to get around it.
What mechanism? All mechanisms that has been proposed so far are
flawed in one way or another, as far as I can tell.
The essential mechanism you need is an ability for the user to ask to see
the text as sent (maybe after CTE decoding, but before flowing). If he
can at least see that, then he can arrange to check inline PGP signatures.
Without that, he is stuck.
Suggesting one of
them doesn't do much good. Although I lean towards MUST NOT, a SHOULD
NOT would work.
I could live with SHOULD.
The main point is that the document don't say
implementations SHOULD use a known broken mechanism.
Indeed.
No, Opera's format=flowed handling is broken. Look at the raw
messages and you will probably find that each format=flowed
"paragraph" is just one long (>76 octets) line with a terminating SPC.
Perhaps it has been fixed in later releases.
The latest Opera versions (anything later than version 7) use their m2
mail client, which is a complete rewrite. It seems to make a reasonable
attempt at format=flowed, but I have not used it sufficiently to have
spotted all the niggles. As I pointed out yesterday, its compose window
seems not to do the right things always, but its display windows seem OK.
--
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl
Email: chl(_at_)clerew(_dot_)man(_dot_)ac(_dot_)uk Snail: 5 Clerewood Ave,
CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5