"Daiki Ueno" <ueno@unixuser.org> writes:

> 2007/8/27, Michaël Cadilhac <michael@cadilhac.name>:
>> > However, if NotDashEscaped is removed, we should also treat the
>> > original text as *not* dash-escaped.  Otherwise, the contents you will
>> > see is not what signed in the clearsigned message.  I personally think
>> > this is not a big problem though.
>>
>> I don't really understand, the NotDashEscaped header is not used to
>> compute the signature, is it?
>
> No, but suppose the following case: the original text contains a line
> led by "- " (0x2D 0x20) as it is, and NotDashEscaped is specified by
> the sender.   You will see the "- " prefix removed by Gnus.

Naaahaaaa! I just completely misinterpreted what `NotDashEscaped' was!
You're utterly right, we should just remove it from the output of GnuPG.

Sorry for the noise :-)

-- 
 |   Michaël `Micha' Cadilhac       |  ... KVim is cited in the talk.        |
 |   http://michael.cadilhac.name   |   "I can't tell if I am more sorry     |
 |   JID/MSN:                       |        for vim or for KDE."            |
 `----  michael.cadilhac@gmail.com  |          -- RMS                   -  --'