[Kde-scm-interest] Idea: Converting KDESDK incrementally

[Ian Monroe]

2011/3/28 Sebastian Dörner <sebastian at sebastian-doerner.de>:
> Hi,
>
> KDESDK is still using svn up to now and there doesn't seem to be much
> progress to make the transition for the whole repository. Also see the
> brief discussion on December 18th on kde-scm-interest.
> I guess one of the reasons is that KDESDK is rather an ensemble of
> individual sub-projects than one coherent module, so there is nobody who
> knows all the code and its history (correct me if wrong). Also, as
> stated in the a. m. discussion, the current maintainer doesn't have time
> to do it.
>
> I wonder if it would be feasible/sensible to do the conversion
> incrementally. In this model, people feeling responsible for parts of
> KDESDK write rules for their part and move the part to git individually.
> This would require them to know the history only for their part (which
> they certainly do) instead of knowing the history of whole kdesdk (which
> they less likely do).
> E.g. I would feel competent to write rules for kdesdk/dolphin-plugins,
> but don't know anything about the rest.
>
> Is there anything that makes this proposal a bad idea? How does it align
> with release management? Possibly, at the time of 4.7 parts of KDESDK
> would be in git and other parts still in svn.
>
> (Immediate cause for this mail is upcoming GSoC, where I would prefer
> dolphin-plugins in a git repository for consistency with rest of dolphin
> and for easier branching.)

I say no. Kdegraphics went this road and it's an absolute mess. If you
want to officially disband kdesdk forever, and disperse the contents
into extragear, OK. But if you are a module you have to act like a
module.

Ian