There was some discussion about Krita's roadmap yesterday on IRC. Here's what I've got in mind (all this has already been discussed on the mailing lists):

2.9: the twelve kickstarter features + MVC branch + resource manager + dirty presets and preset locking. (Note that some features might not be ready...) Release date: end of December, maybe January.
3.0: Port to Qt5. Parallel: work on the animation plugin. A crowd-funding campaign to increase the animation project funding. Release end of March.
3.1 and later: More kickstarter campaigns, with these features as potential kickstarter topics:

[*]LOD Performance optimization for big images
[*]Python Scripting
[*]OSX Port
[*]New Text Tool (probably SVG based)
[*]Improved Vector tools

Basically, 2015 is going to be a bit weird, because the port to Qt5 means we shouldn't do feature development, just porting, and I assume that doing the port will take about a month, plus a month to stabilize and a month for unforeseen things. I also would like that to be the moment the Calligra repository splits up, too.

Yay, a Krita christmas.

Hm... The text-tool will be tricky. Multiple Lines have only been introduced in SVG 2, and that is nowhere near a recommendation. We might need to contact the Inkscape guys on advice on tackling this, when the time comes.
On the good side though, it's not like Krita is alien to writing to XML. Majority of the work will be inside Krita itself.

Of course, first finish all existing projects and get a stable version of Krita out

thanks for posting that Boud! I was wondering about this direction. When the time comes to communicate about the 3.0 and Qt5 port, I would also emphasize WHY this is happening. As a Krita user, I am asking these questions (these don't have to be answered now).

1. Will this speed up Krita?
2. Will this make it easier to fix bugs and do new development?
3. Will I notice any visual difference?
4. What types of things change from Qt4 to Qt5

I think exaplaining these things help bring more excitement to the 3.0 release. I am sure it is a great thing, but technical improvements like this can be difficult to explain to non-technical people.

The QT5 port is kinda difficult to explain.

Main reasons to do it include:
* Getting access to new QT features.
* Having old QT bugs fixed... hopefully.
* QT5 is much more cross platform than QT4.8(And a requirement for an android port)
* New interface baubles
* Cleaning up the code.

The qt4->qt5 port won't have many direct benefits to users, but we don't have a choice -- Qt4 is going to be unmaintained at one point. Debian will stop packaging Qt4-based applications. But there are also some benefits down the road: better OpenGL performance, an Android port, easier OSX port. The dependency on KDE libraries will get less.

But on the whole, I wish it wasn't necessary. Some people like new tech for the sake of it, but I'd rather bring bug fixes and features to the users than doing ports. I remember the Qt3->Qt4 port of 2006 in my nightmares. This will be easier, plus, we've waited long enough so Qt5 is more stable that Qt4 was.

Hi I was wondering if something can be explained to user with no programming experience. Lod and qt5 aren't anything i'm familiar with.

What are the steps for improving the ui performance? Anything listed so far going to help that? example, moving layers is a bit laggy, and "sticky" at the moment, its actually faster to just to use the arrow keys in the layer docker to move layers. Gmic previews slow (i'm aware thats a brand new feature, just mentioning), and large brush performance lag.

Will the future kickstarter be aiming for any of those things? What about some of the features that didn't get in from the last kickstarter (stretch goals), plans for those?
I'm getting the impression that with some of the ideas, the users aren't going to see any "useable" results directly, but it will help with future development, is that a correct assumption? I'd still pledge.

thanks in advance.

For UI improvements, make a wishbug in Krita's usability component of this bugtracker: bugs.kde.org

LOD strokes is the brush-speed up technology that'll need funding.

QT is the framework or fundament that Krita is build on. It needs updating once in a while due to bugfixes and support on various OSes. This updating can take forever if you don't set apart a few months for it, so hence why it's going to be the main goal for 3.0.

thanks for quick response.

--- edit ---

Krita development needs (according to first thread on roadmap) between 85,000 - 120,000 euro's a year to keep up with last years development.

on top of that, the last KS feature from last time (layer styles) still needs to be completed. I don't know how it will be done. I think it can be, I just think it will a take long time, and some of the cool stuff we all want as users probably won't come until 2016.

Well, fortunately we have a semi-regular income from the development fund, donations and dvd sales as well. The remaining layer styles kickstarter feature should take about a month.

boudewijn wrote:Well, fortunately we have a semi-regular income from the development fund, donations and dvd sales as well. The remaining layer styles kickstarter feature should take about a month.

oops sorry, I meant I don't know how the major funding will happen for other features. I know money has already been set aside for layerstyles.

--- edit ---

--- edit ---

We used to use kjs back in the 1.x days, and in the early 2.x days we moved to kross for scripting support. Unfortunately, kross is practically unmaintained. It's been ported to frameworks 5, but I honestly don't dare depend on it. Last year I did an experimental python binding using sip and pyqt, similar to what Kate uses, and that seemed promising.