Re: Workaround for PGP send.c bug introduced between 1.5.3 and 1.5.5.1
On Tue, Jan 06, 2004 at 04:09:05PM +0100, Werner Koch wrote:
> On Tue, 6 Jan 2004 09:17:43 -0500, Rudy Taraschi said:
>
> > In version 1.5.5.1, this has been changed to:
>
> > || if ((WithCrypto & APPLICATION_PGP) && !(flags & SENDKEY))
> > || safe_fclose (&tempfp);
>
> > These two code stanzas do *NOT* behave the same way when mutt is
> > compiled with --disable-pgp and --disable-smime (which I do). I'm
>
> Oops. This was my fault. Please change it to
>
> if ((WithCrypto & APPLICATION_PGP)) {
> if (!(flags & SENDKEY))
> safe_fclose (&tempfp);
> }
> else
> safe_fclose (&tempfp);
>
> > There are many more cases where "#ifdef HAVE_PGP" has been replaced
> > by "if ((WithCrypto & APPLICATION_PGP) &&" in send.c, and I don't
> > know if any of them are causing me problems like the one I quoted.
>
> Given that this code is now in Mutt for quite some time, there is some
> chance that are not too many bugs of this kind.
I think there are a number of bugs related to this. Try configuring
mutt (1.5.5.1) with --disable-smime and then compile. Didn't work for
me as there are unresolved smime related symbols (see my previous
e-mail). I think if mutt developers are going to modify mutt features
they need to test configuring mutt both with and without those features.
Also regarding smime, mutt 1.4 would allow me to view a smime message
even if it didn't know how to deal with smime (so no verification). Now
in mutt 1.5 with smime disabled, mutt returns an error stating it can't
decrypt the message (the message isn't encrypted) and there is no easy
way for me to view the base64 encoded text of the message with mutt.
Seems like a step back in function to me.
--
Will Fiveash