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

List:       kmail-devel
Subject:    Re: Zero-copy parsing of messages achieved
From:       Karl-Heinz Zimmer <khz () kde ! org>
Date:       2002-09-03 4:56:51
[Download RAW message or body]

On Tuesday 03 September 2002 03:32, Don Sanders wrote:
> On Tuesday 03 September 2002 07:26, Michael Häckel wrote:
(...)
> > Also mimelib has quite some other bugs nobody cares to fix or
> > nobody manages to fix there. Instead more and more external code
> > gets written to work around them in libkdenetwork and KMail and
> > less and less code from mimelib is actually used.
> > As far as I remember some time ago you actually agreed that KMime
> > should be written. Now part of the code is there and part of the
> > code is even already used. Therefore you obviously can't assume
> > Marc to be pleased if you tell him now that he shall throw his
> > code away again.
>
> I see.
>
> I think what I asked for was a message handling library that could
> handle large messages efficiently. I wasn't particular about whether
> mimelib should be modified or a new library written.

Since it turned out that lots of other things were sub-optimal too
Marc came to the conclusion that an entirely new project - KMime - was
the best way to get a clean solution.

I remember that when starting to work with the mimelib I too was very
irritated about the way it works and how does (or doesn't) some things
but over the time I simply got used to the mimelib: this does not mean
mimelib is better than it looked first!

I strongly hope your words (*surely* spoken out of enthusiasm, not out
of spite) do not result in Marc becomming disappointed since the work
he has done so far IMNSO _should_ be appreciated and lead to a library
that can be used to finally replace mimelib!
When I asked whether this will be the case for KDE 3.2 or not, I was
looking for information about the _time_ but I was _not_ asking whether
it would happen at all!

> I'm not asking Marc to throw away all the code he has written, in
> fact I use kmime_codecs in my patch.
>
> But I no longer consider replacement parsing code a priority.
     ^^^         ^^^^^^^^^^
Since nobody else did the work that Marc did and he found out that doing
a clean, new KMime would be the solution (and he presented this way to
go right here) I am sure that not /we/ should talk about priorities now
but just be thankfull to get rid of an unpleasant mixture of different
concepts and get the KMime to replace lots of strange sub-optimal
solutions!
If Marc would write about _his_ priorities concerning to KMime I would
be willing to listen more carefully than to what you seem to propose
now: Even if it might be possible to somehow work around the problems
we still have and (based on kmmsgpart.* and the mimelib) we somehow
might get something running that does what it should be for composing
complex MIME messages, I _still_ would be happy to see KMime replace my
temp. code since I am sure that's the way to go for a *clean* solution
usuable also for other KDE projects...

Cheers,
Karl-Heinz
-- 
Karl-Heinz Zimmer, Senior Software Engineer, Klarälvdalens Datakonsult AB
<mailto:khz@klaralvdalens-datakonsult.se>            <mailto:khz@kde.org>
_______________________________________________
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]