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

List:       kde-core-devel
Subject:    Re: Proposal: package split
From:       Rik Hemsley <rik () kde ! org>
Date:       1999-10-03 16:33:26
[Download RAW message or body]

#if Harri Porten
> Are you suggesting to unite Empath, kdebase (kfind) and koffice (KWord)
> to one big package ? ;) How about other apps that may want to share the
> addressbook ?

It has a CORBA interface :)

> > Users don't understand that having a Maildir based client that doesn't stall
> > the UI is to their advantage. They'll use whatever has pixmap backgrounds
> > and extra 'features'.
> 
> I don't know Magellan but I'd say: the better program has to compete ny
> its merits and not by forcing it on anyone. There are surely users in
> love with e.g. kmail. Why take it away from them as long someone makes
> sure it compiles ?

Right. I wouldn't force anyone to use anything. I like KMail apart from
the fact it doesn't do non-blocking IO and hence stalls, and that it has
lost my mail - mainly because it's using mbox.

> Mmmmh. That's a pitty but proves my point: Empath didn't get enough
> attention cause everyone encountered kmail first as it was part of an
> "official" package. I'd love to give every competitor the same
> treatment.

Quite right too. If Magellan turns out to be a better client than kmail,
then I'll use it. I would still like some kind of integration with the
rest of KDE though, so that's what I'm trying to do.

> > Just my 0.02 euro. I know that programmers will always work on what they want
> > to, and I'm not saying Empath and KabII are the best designed apps on the
> > planet. Programmers also tend to like to start new projects - they know
> > the code better that way. Why another news client is being written, though,
> > I have no idea. Empath is designed to handle news and the hooks to put
> > the support in are already there.
> 
> Perhaps there is some irony in this comment I fail to see. Why did you
> start Empath although pine, elm, kmail, Netscape, XFMail and others were
> already there ? Microsoft is wondering why some Austrian guy is so crazy
> to develop a new word processor although Winword offers every feature
> one can think of.

I started Empath because:

* I was about to lose all net access (and it was for many months).
  KMail didn't work well at all at the time, so I did what
  everyone else does - start afresh.

* KMail doesn't use Maildir as its native format, and I don't like
  losing mail.
  
* I wanted a client that didn't block the UI at _any_ point.

* I didn't like relying on mimelib, considering the license is a little
  strange and the author is not reachable. Also it uses its own string class,
  adding bloat for no gain and adding slowdown as string conversions are
  made.

pine: Er, it's not for KDE.
elm:  Er, it's crap and not for KDE.
mutt: The best mail client in existence, but not for KDE. I investigated
hacking a KDE front end onto mutt, but found that the UI is not separatable
without a huge amount of work, basically amounting to a rewrite.
kmail: Reasons above.
Netscape: Decent GUI, but too big and not for KDE.
XFMail: The worst GUI ever made, but slightly preferable to Netscape as it
doesn't eat half your memory and take an hour to start up. Not for KDE.

Believe me, I've tried every client that I could get a copy of. My personal
favourite is Ishmail, which is now GPL. I looked at porting that to KDE but
decided against it as I was already a long way ahead with Empath.

I would honestly have hacked kmail but the internal design is important. You
can't support asynchronous operation without changing the design beyond
recognition. A new message parser had to be written and the 'mailbox' object
would have to be abstracted to allow for easy handling of different mailbox
types. This basically adds up to a rewrite.

Cheers,
Rik

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