From kmail-devel Tue Apr 17 08:12:15 2001 From: Andreas Gungl Date: Tue, 17 Apr 2001 08:12:15 +0000 To: kmail-devel Subject: Re: sharing the PGP code X-MARC-Message: https://marc.info/?l=kmail-devel&m=98749494631580 On Monday 16 April 2001 11:32, Mathias Waack wrote: > On Sunday 15 April 2001 22:07, Andreas Gungl wrote: > > I plan to test GPGME then. It's still alpha, but IMO it would be > > good to have an implementation ASAP to be ready when it's getting > > stable. > > I had a look at GPGME some time ago. IIRC there are some problems > accessing PGP keys from within GnuPG, right? > > Regarding to the note of Michael: I think this code should go into > kdelibs. So the size and startup time of KMail should'nt augmented. > And from my understanding, GPGME is not a simple frontend to gpg. Its > a library which implements the OpenPG standard using the same code as > gpg. It should work with data of all applications following this > standard. But I don't know how far from the standard is PGP > currently. Andreas please correct me, if I'm wrong. In the moment we're supporting GnuPG and three PGP versions. So, in the=20 beginning, we could also add GPGME as a kind of built-in OpenPGP support.= =20 Later it might be possible to remove at least the current GnuPG support,=20 when GPGME will provide what it's promising. IMO the greatest problem with the current solutuion is the dependence of=20 the GPG/PGP program output which we have to parse. A library could improv= e=20 this situation significantly, and I do also hope for better response time= s=20 compared to the current situation, where we have to start another process= =20 some times. Yes, there are problems with some PGP keys in GnuPG. But I think, we can=20 either wait, until GPG(ME) is able to handle those keys too, or we could=20 try to use a library for PGP which could give us the possibility to drop=20 the current jungle of three different implementations for PGP 2.x / 5.x /= =20 6.x (or even higher). Personally, I prefer to incorporate GPGME in the hope, that using this=20 library we will be able to remove the current parsing based implementatio= n=20 completely in (let's say) one year. Another direction for investigations are improvements regarding OpenPGP=20 encoded messages. I think, mimelib is the main problem here. Andreas _______________________________________________ Kmail Developers mailing list Kmail@master.kde.org http://master.kde.org/mailman/listinfo/kmail