[KPhotoAlbum] Improved Image Viewer? [PATCH]

[Shawn Willden]

On Tuesday 19 December 2006 23:41, Jesper K. Pedersen wrote:
> Well there is three problems to branching.
> 1) I don't have enough time to maintain two branches, so I would have to
> let one branch down.

One solution to this is to find some volunteers who are willing to take on the 
task of maintaining the branch.

> 2) History has shown numerous times that as soon as I've accepted a patch
> into the code, people kind of keep interest in maintaining it. this means
> that I need to test patches with new features *before* they go into the
> code. With me focusing on the release, this would just be problems piling
> up for me.

Again, assuming there are folks willing to help, the solution is to get 
another developer or two who can be responsible for vetting incoming patches, 
freeing you up to focus on the release branch, up until the time you decide 
to switch gears and focus on the trunk development.

> 3) If I branches, everyone using SVN would be using the new cool branch,
> and the release would be even less tested before released.

This one is particularly easy to deal with:  don't put much effort into 
keeping the trunk code stable.  If you reduce the set of trunk users to those 
who are willing and able to diagnose and fix bugs, then your trunk user base 
will be small :-)

I think more of the testing comes from downloaders of snapshots, anyway.  
Perhaps one way to make that testing even more effective is to find a 
volunteer or two who will create binary packages for the major distros.

This next bit is going to sound a bit critical, Jesper, so please keep in mind 
that it's written with the best of intentions and that I understand it's only 
my "outsider" perspective, which is almost certainly inaccurate in at least 
some respects.

It has always seemed to me that the pace of KPA development has been slowed by 
your unwillingness to allow anyone else to work on the code.  I think KPA 
would benefit substantially if you were to find another developer or two that 
you can trust, and give them commit privileges and suggest tasks for them.  
Beyond that, your typical response to most feature requests differs from that 
of most other OSS projects.  Where they typically say "You want X?  Send me a 
patch", you tend to say "That's a good idea, I'll look into it later".

Frankly, I think it's amazing that you have the dedication and endurance to 
keep a project of this scale alive and progressing as a one-man effort.  At 
this point, though, I think KPA would benefit from some delegation and some 
expansion of the development team.

Please excuse me if my perception is all wrong, and if it's a lack of 
volunteers, not an unwillingness to delegate that has driven your approach.  
I, for example, am not volunteering.  I'd like to but lack the time to take 
on yet another project.  Actually, I might be willing to create Debian 
binaries of snapshots.  I'll look into that.

Finally, if you simply don't like sharing the work, then that's clearly your 
prerogative, because we all know that there would be much, much less progress 
in KPA without you.

Thanks for a great program,

	Shawn.