Reverting kaddressbook back to previous version?

[Matt McCutchen]

On Wed, 2010-04-14 at 15:53 +0200, Thomas Janssen wrote:
> On Tue, Apr 13, 2010 at 10:46 AM, Juha Tuomala <Juha.Tuomala at iki.fi> wrote:
> > On Mon, 12 Apr 2010, Ryan Rix wrote:
> >> On Mon 12 April 2010 6:40:59 am Juha Tuomala wrote:
> >>> I recall, that the earlier version had some level of Akonadi support
> >>> as well, so in theory, would it be possible to revert the codebase
> >>> back to the one that can actually be used?
> >>
> >> Sure, try `man yum`.
> >
> > You mean that we're here to solve our own problems, not to make a
> > good distribution for great public?
> 
> We *have* a good distribution for great public. Kaddressbook works as
> expected for me.

Please take the request seriously.  If Tuju is right that most users
would be better off with the older version, then that's what Fedora
should ship.  Tuju, if you can possibly be bothered to list some of the
regressions you consider most severe, that might help the discussion.

I have no experience with kaddressbook, but I had a similar experience
in October 2008 with Evolution 2.24.  There, the merging of the disk
summary code before it was anything near release quality caused many
regressions, including breaking threaded search folders, which I rely on
heavily.  Unfortunately, Evolution 2.22 had many equally severe bugs
(notably a crash when editing a sorted task list), so by pursuing
disk-summary in 2.24 rather than just fixing bugs, upstream left Fedora
between a rock and a hard place.  I filed a bug requesting a reversion
to 2.22, which may have been a bad idea on the whole but IMO deserved
more consideration than the knee-jerk WONTFIX it got:

https://bugzilla.redhat.com/show_bug.cgi?id=468950

-- 
Matt