kdenetwork/kmail

[Marc Mutz]

On Thursday 19 September 2002 01:46, Aaron J. Seigo wrote:
> On Wednesday 18 September 2002 05:07, Marc Mutz wrote:
<snip>
> > Suddenly everyone comes 'round the corner and tells you what a
> > fucking bullshit you have produced for the last few releases. Nice.
> > :-(
>
> i don't think that was what he was saying. kmail's devel has been
> conservative to keep stability high while adding some fundamental
> features (such as IMAP). on the up-side, this has created a nice,
> stable piece of software that works well for many.
>
> but life has trade-offs, and in this case the trade-offs include a
> pattern of development that is at times excruciatingly slow and
> overbearingly dominated by a select few individuals. patches have
> been ignored, needed features are slow to arrive, contributors have
> been discouraged, and it has resulted in a body of code that is in
> serious need of refactoring.

What you describe here is IMO not the result of the core developers 
being egomaniac, but simply lacking time. It's a bad fortune that three 
of them are in the process of writing their theses. So instead of 
working on fundamental changes, we concentrate on integration and 
bugfixing. I would _love_ to work on KMime, since that's fun for me, 
but OTOH, I can't stand the thought of releasing KMail as it is. So I 
do the boring stuff instead and you will understand that I'm not as 
productive with that as I could.

If in this situation people start throwing feature patches on you, all 
you can tell them is "come back after 3.1 is out".

I don't want to whine here about my lack of time. The fact remains, 
though, that of the patch contributors of the last release cycle or so 
no-one except Carsten Burghardt, yourself and Zack has stayed committed 
to his patches after they had been comitted.

What I miss is more people who stop complaining and start helping with 
the boring integration and bugfixing. Everyone (including me) wants to 
work on new features. I have tons of ideas that I can't start off or 
even finish because I lack the time and I feel the need to clean up 
after other people, because I'd feel ashamed to have my name on a KMail 
release that looks like KMail did one or two months ago.

I hope you understand that it's quite frustrating to see the rate of 
feature patches increase three or fourfold just after the feature 
freeze has started and you yourself feel compelled to fix the obvious 
shortcomings and bugs that have accumulated over the last months.

> while make_it_cool is atempting to address these challenges, it
> doesn't mean that the kmail in HEAD is a steaming pile of shit, just
> that perhaps it is the right time for development to take a different
> path. in this case that means a slightly more cavalier attitude
> towards short-term stability in order to tune the entire code base
> for current and future needs.

That's what I tried with KMime for the last year or so. I just needed to 
set the priority on out-of-KDE stuff and if I work on KDE/KMail, it 
will be the little things that no-one else bothers with.

Maybe I should focus on what I want to do instead of on what I think I 
need to do...?

Marc

-- 
The illegal we do immediately.
The unconstitutional takes a bit longer.            -- Henry Kissinger
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: signature
URL: <http://mail.kde.org/pipermail/kde-core-devel/attachments/20020919/31ed69f1/attachment.sig>