[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kmail-devel
Subject:    Re: Crypto plugins
From:       Michael =?iso-8859-1?q?H=E4ckel?= <haeckel () kde ! org>
Date:       2001-10-31 12:23:20
[Download RAW message or body]

Hi,

On Wednesday 31 October 2001 09:45, Jan-Oliver Wagner wrote:
>
> The BSI didn't provide any specification except that KMail and Mutt
> are to be Sphinx-enabled. The design you criticize, is the one
> of the Ägypten team. I am glad for your critics for it clearly
> shows and stores the overall way to go.
> Can you make a small web page
> describing the optimal design, which we can always refer to from the
> Ägypten project to explain that only a first step is made?

The actual problem here, is the the MIME implementation of KMail is currently 
in progress of being rewritten since the library we currently use has a few 
design limitations.

Mainly it is memory and CPU inefficient, when it comes to huge attachments. 
This results in bug reports like "KMail crashed, when I tried to send KDE 2.2 
(80 MB) via e-mail to a friend".
Also it doesn't natively support nested MIME parts. Access to nested parts is 
only possible by first parsing the message into the top level parts, where a 
multipart is also a single part and then parsing this multipart again into 
its subparts. This works, but is very inefficient.

It was also planned, to implement things like PGP/MIME directely into this 
library. This means of course, that this library needs finishing first, 
before PGP/MIME can be implemented, which for your plans doesn't seem to be 
an option. Therefore some double work has to be done.

Regards,
Michael Häckel
_______________________________________________
kmail Developers mailing list
kmail@mail.kde.org
http://mail.kde.org/mailman/listinfo/kmail

[prev in list] [next in list] [prev in thread] [next in thread]