[Kde-scm-interest] Fwd: [Kde-games-devel] Re: RFC: git move proposal; "duolithic" kdegames

Stefan Majewsky stefan.majewsky at googlemail.com
Thu Mar 3 11:05:57 CET 2011


---------- Forwarded message ----------
From: Stefan Majewsky <stefan.majewsky at googlemail.com>
Date: Thu, Mar 3, 2011 at 11:05 AM
Subject: Re: [Kde-games-devel] Re: [Kde-scm-interest] RFC: git move
proposal; "duolithic" kdegames
To: KDE games development <kde-games-devel at kde.org>


On Thu, Mar 3, 2011 at 3:53 AM, Arno Rehn <arno at arnorehn.de> wrote:
> I don't see why ripping out the data files would nullify all other history. It
> probably requires some effort, but the data files can be removed from history
> after the git conversion. Nicolás has proven that nothing is impossible with
> git-filter-branch ;)
> Seriously, I'd try to keep as much history as possible.

Interesting. I'm by no means an svn2git expert and did not want to
impose in the proposal that this is possible in a reasonable
timeframe.

Re the concerns over how long SVN is going to stay, when will the
lifetime of SVN be decided upon? This issue has been touched in the
k-c-d discussion on the creation of the kde-wallpapers module, but as
far as I can see, nothing definite has been decided.

On Thu, Mar 3, 2011 at 4:35 AM, Ian Wadham <iandw.au at gmail.com> wrote:
> I am not so keen on 1. above.  If there is no dependency from games to data
> (not even a warning?), how is a newcomer or a non-games person building
> from source supposed to know that data exists and where to find it?
> Without such data, the compiled games could be a big turnoff IMO.

Apart from printing a notification in the CMake run for kdegames, one
could show a standard warning message box that kdegames-data is
missing. Note however that the usual method for users to obtain the
games is still through packages, where this problem does not exist in
the first place. I expect most users who compile the games by hand for
the first time had already installed them from packages previously, so
they should notice very fast that the error is on their side.

Greetings
Stefan


More information about the Kde-scm-interest mailing list