The day of seperation is here

[Falk Brettschneider]

Hi!

Roland Krause wrote:

>John, I agree with your analysis. You are right on!
>
>There is one last argument I would like to make. KDevelop is quite a
>large codebase and it has a lot of functionality. 
>
>"The devil is in the detail", as we Germans say. Gideon has a much
>better structure but none of the details, hence it has about 1/8th of
>the current functionality of KDevelop (if that). These details have
>been implemented over the course of many years and they will get lost,
>all of them, if we continue with the gideon base. 
>
>I give one example, KDevelop Tools vs. gideon Tools, can you specify a
>file or directory name to the call of a gideon tool? No you can't, not
>because it is impossible, but because it is not implemented. It is a
>detail. No one has bothered to implement it. 
>
>My conclusion is:
>
>I would like to make KDEVELOP_2_BRANCH into HEAD and move gideon into 
>GIDEON_BRANCH. 
>
YES!

I additionally suggest to keep the directory structure of Gideon for 
that ported stuff. This trick would us allow to work without the later 
merging of HEAD and GIDEON_BRANCH, at least partially. I mean the 
GIDEON_BRANCH could use 'cvs update -dPA' in the shared directories. In 
this way development for those code would be on one cvs branch (HEAD).

>
>
>Then, I would like to nominate GiGaFaLk to be the "fearless project
>leader" and to ask Ralf "Kand-ALF" Nolden to be our chief advocate.
>This way we can split between "technical" and "strategical" roles. 
>
>I would like to volunteer myself as "Parts Coordinator", I will make a
>list of available gideon parts and plugins and will try to coordinate
>between the respective maintainers for the port to KDevelop. 
>
>I want to volunteer HarryF as the "head of ViewManagment", i.e. his
>task is to maintain the ViewManager, which he has already started to do
>anyway, unfortunatley on the wrong codebase. 
>
>Everybody else get's to do the fun stuff! Work on parts, plugins,
>gideon, java whatsoever. 
>
>None of the work spent on gideon shall be lost. We can make KDevelop
>into a "multilanguage" IDE, the proof that that is possible is gideon. 
>
>Now, let's make it real.
>
>Roland
>
I agree
(Although I more like the idea of not having any leaders but to decide 
things by rational discussions and clear votes of the _whole_ (core) 
team. I never felt to have a chief for the development of the KDevelop-2 
series. More or less the people, that take the project seriously, care 
for the roadmaps, the publicity and things like that. That can vary from 
time to time. Although I always a friend of perfection and of trying to 
make things professional.)

Ciao
F at lk



_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com