Re: [Evolution-hackers] S/MIME support



On Thu, 2003-07-03 at 15:01, Jacob Perkins wrote:
> > On Thu, 2003-07-03 at 13:44, Jacob Perkins wrote:
> >> Why not use gpgme and gpgsm?  gpgme's abstraction works over gpgsm and
> >> gnupg (and maybe more in the future), so implementing s/mime with gpgme
> >> could also make the pgp code simpler.  There is a working plugin for
> >> KMail
> >> which uses gpgme/gpgsm, and balsa and sylpheed both use gpgme for pgp
> >> support.  The existing interfaces could still be kept, only the backend
> >> code would have to be changed.
> >
> > 2 words:
> >
> > licensing issues :-)
> >
> > we can't use gpgme because it is GPL.
> >
> > Jeff
> 
> Sorry if this is stupid, but I don't understand:
> evolution is GPL, and gnupg/gpgsm is GPL. Is the only reason gpg can be
> used is because the code is not directly linked?

right. see, the problem is that while evolution itself is GPL, connector
is not.

> If gpgme was LGPL, would you use it?

maybe, but I'd probably just leave it use my current gpg code unless
gpgme offered a clear advantage. (if it ain't broke, don't fix it)

>  The gpg people had reason to release
> libgcrypt under LGPL, so I think if it was requested, and a company like
> Ximian could get behind it, and even sponsor[1] development, they might
> consider it for gpgme too.

we've already asked and they refused. plus it'd just be another library
we'd have to link against (which ould be fine if the current gpg code
didn't do what we needed it to do, but it does). gpgme is also not a
stable API.

>  gpgme aims to be, and I believe should be, the
> standard interface for pgp and cms(s/mime) support.

we have some private reasons for wanting to use Mozilla's NSS libs for
S/MIME. Not sure how much I can say about this so I'll leave it at that.

>   Plus, it might allow
> for easier integration of evolution with seahorse[2] in the future.

I don't really know much about seahorse so I can't really comment...

we also want a unified cert/key manager in evolution for gpg/smime/ssl
certs, so unless seahorse adds support for Mozilla NSS for S/MIME certs
and SSL certs, it won't really fit our needs anyway.

Jeff

> 
> [1] https://order.kagi.com/cgi-bin/r1.cgi?P3G&&;
> [2] Bonobo controls for key properties in the address book, global
> (corba/bonobo) pgp/cms contexts, common passphrase entry dialog, more?
> These are just ideas right now, though I would be happy to discuss them
> more with those interested.
-- 
Jeffrey Stedfast
Evolution Hacker - Ximian, Inc.
fejj ximian com  - www.ximian.com




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]