[Kdenlive-devel] [Kdenlive] Kdenlive

[jb]

On Thursday 02 February 2012 09:52:51 you wrote:

> Hi Jean-Baptiste,
> 
> my name is ed rogalsky and I'm following the dev of kdenlive since a year.
> I have tried kdenlive some years ago but could not get warm with it. But it
> changed as I tried this program some month ago for a video project. Its
> great.
> 
> I had some discussions with Dan about integrating unit testing, git and a
> patch level
> in the development but he didn't feel responsible for kdenlive. Ok I
> understand he is
> the maintainer of mlt. I guess you are the project leader of kdenlive.
> I very apriciated the decisions made in randa. In my opinion the version
> 0.8.2.(1) has
> the basic functions that you need for video editing but for daily use its
> very important
> to have robustness and stability. I know how difficult it is to get a big
> program very stable
> (I'm a software developer) :-)))

Hi,

Sorry it took me some days to answer, was sick and no energy. I am the 
Kdenlive maintainer, and for sure I know that we need to work on stability and 
regression testing.
 
For your info, a git / patch infrastructure is provided by KDE, that should 
make it easier for people to provide patches:
https://git.reviewboard.kde.org

But it has currently not been tested yet for Kdenlive, Alberto might have some 
more infos on this part.

> Some words to me:
> I'm a freelancer (Embedded SW/HW, Windows/Linux) drivers) working now for a
> big automotive
> company. I have "some" experience in the Embedded world (C/C++, Assembler,
> Hardware, Realtime,
> DSP Processing, Network/Bus Communication) and PC SW (Dotnet, Windows/Linux
> Drivers (basic),
> Java, UNIX/Linux, Audio and so on).
> 
> I would like to contribute some code to kdenlive and my idea was to do the
> jack integration
> using the mlt jackrack filter. What do you think? I'm playing with it some
> days and the basic
> operations like start/stop transport are working yet. Its just the
> beginning. My biggest
> problem is time. Thats why I will need some weeks or so:--(((.

Great. I know several users are very interested in a Kdenlive - Jack 
integration, and I myself have no real knowledge / interest to work on it 
currently.

I think that all Kdenlive developers have the same problem with time, so it's 
not an issue.

> Before doing a lot of code I wanted to know which refactoring steps are
> planed:
> 1) Only some basic steps to integrated unit testing or
> 2) A complete Redesign of the program(Classes/Models and so on)?
> Just an idea:
> It seems to me that the roadmap in mantis is not maintained. And for me a
> bug tracker is something
> low level very useful for developers, but have a high level roadmap whould
> make the dev process
> more transparent. It dont have to be very complicated. Just a html page with
> functions that are
> planed (like the Roadmap of MLT (Dan)). Perhaps who is working on it and
> present open tasks for
> volunteers.

I forwarded that email on the Kdenlive-devel mailing list so that others might 
answer too. One idea I had for the refactoring was to create a meta-bug called 
"refactoring", with several children bug listing the different tasks planned 
for this refactoring, with a few words describing the task, so that developers 
could discuss how to work on it before starting changing the code.

 
> I think thats it for the first step. If you appreciate my offer I'd love to
> work with the kdenlive
> team.

Great, I think everyone will be glad to have you in the team!


regard
jb