On Saturday 25 August 2012 19:03:31 Slávek Banko wrote:
On Saturday 25 of August 2012 18:09:55 Martin Gräßlin wrote:
No you are wrong. KBugBuster used to be part of KDE-sdk till 4.5 - which is two years ago. Much younger than the 3.5 version Trinity forked. Also there had been some work which has not been finished going on by some students to be found in [1].
Did you look in the log of revisions? http://websvn.kde.org/tags/KDE/4.5.5/kdesdk/kbugbuster/?view=log
"Much younger" according to him does not seem like something substantial.
Qt4 vs Qt3 - yes that is quite a difference. (You can address me as "you" - no need to switch to "him")
I do not know the condition of the referenced student works, but apparently it is a complete rewrite. Not to diminish their value, but to modify the current version KBugBuster are not very usable.
think about it. One of the reasons (to my knowledge) to restart the work is that KBugBuster builds up on a very old way to interact with Bugzilla. Maybe that is in the time of a JSON interface not the smartest thing to do. Look for example at [1] how handy it can be to interact with Bugzilla through the JSON API in combination with QJson.
In summary, it seems that we have two practically identical versions KBugBuster - one abandoned in KDE4, and second active in Trinity. I did not find interesting reason why I should prefer to work on a version for KDE4. When they are practically the same, so it will probably be easy to reflect changes from Trinity version to KDE4 version.
Well I cannot tell you what to do but do you really want to work on a Qt 3 version?
The other thing is that you probably want users, right? How large is the target audience for a Qt 3 app (remember my distribution - Debian testing - no longer shipps Qt 3) compared to a Qt 4 app? With a nice Qt 4 based Model/View approach its even possible to write apps for smartphones (I would love to be able to manage bugs from my smartphone - have to think about that).
Last but not least think about the impact it could have if someone from the Trinity project approaches the KDE community with some work done on KBugBuster and asking for re-inclusion into the KDE-sdk. That could change the complete view on the Trinity project. Would anyone notice that work went into the forked KBugBuster? Unlikely.
Of course it's your decision on what to work. But if I had the chance to use the code which has been ported over to Qt 4 already, I would not think twice. Especially not if there are no other differences.
Best Regards Martin
[1] http://quickgit.kde.org/index.php?p=scratch%2Fgraesslin%2Fchangelog- generator.git&a=summary