GSoC Project - Lazy Selection Tool

[Dmitry Kazakov]

Hi, Yang!

To apply to this project I would expect the the proposal to contain 
technical details on how you are going to implement it, and at least a 
draft of the class-level architecture.

The engine is basically implemented in Colorize Mask feature. Ideally it 
should be rewritten to use watershed algorithm, but it might also work 
as it is now. The GUI part is the most difficult one.

If you manage to write a proposal and provide at least one patch by 
Monday evening, we can consider this project.





On 31.03.2017 10:42, Boudewijn Rempt wrote:
> Hi,
>
> It's an _absolute_ requirement for KDE as an org that you can show
> that you have submitted code to a KDE project before you can be
> considered. We've had too many students who basically couldn't code in
> the past. In general, we do want people to start contacting us in
> January latest, integrate with the community, fix some bugs, implement
> some features, show their enthousiasm for Krita.
>
> So, this is a real problem...
>
> If you'd write a proposal and submit it today, and then join us on irc
> over the weekend and show some really impressive hacking skills, we'd
> still be able to consider you.
>
> Fixing this bug:
>
> https://bugs.kde.org/show_bug.cgi?id=349554
>
> Would definitely impress me.
>
> For the proposal itself, Dmitry's input is needed, but I would also
> put in a phase where the existing colorize tool is optimized, and a
> phase before that where you investigate what the best option would be:
> reuse the colorize tool, or implement a new tool.
>
> On Fri, 31 Mar 2017, Yang Fang wrote:
>
>> Hello Krita,
>>
>>
>> My name is Yang Fang, a first year at Stanford University studying computer science. I am applying to do open source work through the GSoC program, and only recently came across the Lazy Selection tool project idea for Krita. As a frequent user of mostly proprietary photo manipulation tools (which do not contain a similar foreground selection tool, or only irritatingly difficult/inaccurate versions of it), I am intrigued and would like to help realize a potentially powerful function that I myself would use often.
>>
>>
>> I chatted with Valorie on the kde-devel IRC channel, and was told that I should've been in contact with my potential team for a substantial amount of time by now - a reasonable request in my opinion. However, I am very much interested in the goals of this particular project and am seeking guidance (especially from DmitryK) on whether there is any way I can still apply to work on it.
>>
>>
>> Thanks for your time,
>>
>> Yang
>>
>>
>>

-- 
Dmitry Kazakov