KDevelop3 was: nasty bug, CKDevelop::edit_widget

[Ralf Nolden]

Sandy Meier wrote:
> 
> > What's happening to the project ? Is it because nobody is interested or/and
> > has no time anymore ?
> The "problem"  is that some people of the team want to develop on the 1.x
> branch and other on the 3.x branch. As this is a sparetime project everyone
> is free to choose where to put his effort. Unfortunately this has splitted
> the manpower in the past.
Yes, agreed to a certain extend. It did. But we needed that and we don't
disagree on that. 

> An example: From the developer point of view MDI in KDevelop1.4 will be
> "waste of time" as KDevelop3 already contains MDI and KDevelop1.4 is a dead
> branch. It is only important for the user. On what is more important we
> disagree sometimes. :-)
Yes :-) Well, Sandy, you know why John and I decided to go that way for
some  month. It was and is important to provide the user - for which, in
the end, we're providing kdevelop for productive work *now* - a
constantly working version. And 1.3 wasn't the end unfortunately as we
saw on the result that the kdestudio shipping/download numbers increast
vastly more than our 1.3. That was an alarm sign. Developers wanting to
go with KDE 2.x technology didn't want to stick with a KDevelop based on
KDE 1, even if we provided that "binary" version.

It is not that John and I didn't like to work on 3.0. In fact, the work
I'm doing right now is to get developers (which is hard currently and
it's always newbies I can get - what I really lack is a *descriptive and
long, explanatory* book of the architecture where one can read himself
into within a day and then start working on a plugin. Next is the docs.
I'm working with Lauri to update the docs of 3.0, plus the fortune of
the work Falk is doing is that the handling of 2.0 and 3.0 from the
user's point is very similar. You know my time will be limited in autumn
again because of my exams, not to speak of summer due to the linux days
and expos I have to go to. So this time it's work for both versions at
the same time and we're doing it in HEAD first for testing, then copy
the docs over into 2.0. This will also hopefully include the KDE 2.0
development book. I think that is more maintainable than the programming
manual as, as I said, I won't have that much time like in 99 when I
wrote the books, to redo the work for the new API and organization
completely.

I wouldn't say or fear 3.0 would never get out. It's a matter of time. I
know it's a bit diappointing if it takes us so long to get it going. But
Sandy, please remember and compare: in 98, we had to deal with the
change from the KDE 1.0 to 1.1 API. That was slowing us down as well
sometimes. But it was a completely different architecture. Now the goals
are much higher and the design matters more to some certain extend than
the actual code on the long run. And remember, Qt 3.0 in summer will
require more changes. Also it's unsure what to do with the new qt
plugins. I haven't looked at it yet, so there may be other possibilities
to regard as well. It's just that for the short-sight, the focus is on
getting 2.0 shipped and the following versions staying usable and
shipping with KDE releases. Without that, 3.0 chould be released a few
month earlier, yes. But at the cost that we have no user base and have
to start getting users and developers from scratch again. I'm also
constantly working on keeping in touch with KDE and the distributions so
we get more support and hopefully sometimes fulltime developers paid by
distributors/companies which is hard this year as well but we could get
some if we keep up showing we have something and have something in
store.


-- 
Finally, even I have to admit that being myself was the best thing
that ever could have happened to me. - Le Grand Charmeur

**********************************
Ralf Nolden

The KDevelop Project
http://www.kdevelop.org

nolden at kde.org
rnolden at kdevelop.org
**********************************



-
to unsubscribe from this list send an email to kdevelop-devel-request at kdevelop.org with the following body:
unsubscribe »your-email-address«