< GSoC‎ | 2011
Revision as of 14:37, 15 February 2011 by Trueg (talk | contribs)



Information for Students

These ideas were contributed by our developers and users. They are sometimes vague or incomplete. If you wish to submit a proposal based on these ideas, you may wish to contact the developers and find out more about the particular suggestion you're looking at.

Being accepted as a Google Summer of Code student is quite competitive. Accepted students typically have thoroughly researched the technologies of their proposed project and have been in frequent contact with potential mentors. Simply copying and pasting an idea here will not work. On the other hand, creating a completely new idea without first consulting potential mentors is unlikely to work out.

When writing your proposal or asking for help from the general KDE community don't assume people are familiar with the ideas here. KDE is really big!

If there is no specific contact given you can ask questions on the general KDE development list [email protected] See the KDE mailing lists page for information on available mailing lists and how to subscribe.

Adding a Proposal

Follow the template of other proposals!

When adding an idea to this section, please try to include the following data:

  • if the application is not widely known, a description of what it does and where its code lives
  • a brief explanation
  • the expected results
  • pre-requisites for working on your project
  • if applicable, links to more information or discussions
  • mailing list or IRC channel for your application/library/module
  • your name and email address for contact (if you're willing to be a mentor)

If you are not a developer but have a good idea for a proposal, get in contact with relevant developers first.


How to find ideas? To see previous Project ideas, see: 2010 ideas. Obvious sources of projects are the bugs database, the forum, and your list and IRC channel ideas.


Amarok is a powerful KDE based music player for Linux and Unix, MacOS X and Windows with an intuitive interface. It makes playing the music you love and discovering new music easier than ever before - and it looks good doing it!

Website - Mailing list - IRC channel: #amarok on Freenode.

Project: Playlist sharing

Brief explanation: Amarok playlist sharing will allow you to select a set of tracks and share them with friends over chat. By using Telepathy lots of IM networks are available including jabber, google talk, AIM, MSN and facebook chat.

Expected results: A plugin will add a PlaylistProvider to Amarok that enable the user to share via HTTP streaming (P2P) a playlist with online friends.

Knowledge Prerequisite: Qt/KDE development (includes C++ & git), Telepathy, HTTP streaming and NAT traversal.

Any of these are optional but not all of them.

Skill level: medium to high.

Mentor: Bart Cerneels or Ian Monroe

Project: Self Contained Collection

Brief explanation: sqlite file saved in collections own folder. Tie in with USB media device. More info to follow.

Expected results:

Knowledge Prerequisite: Qt/KDE development (includes C++ & git), SQL.

Skill level: medium.

Mentor: Unless someone else from the Amarok team can be convinced: Bart Cerneels


Photo Management program

digiKam project web site - Mailinglist - IRC channel: #digikam on Freenode.

Project: Camera UI Revamp

Brief explanation: DigiKam features a UI for accessing and downloading pictures from cameras. The code is rather old, using Qt3Support classes for the icon view, the UI code intermangled deeply with backend code, and has not seen very much care and love for some years.

This project would involve taking the old code apart, rewriting a clean code base backend and frontend, but also adding user interface elements to make the most important everyday task as easy as possible.

In more detail: Write a model listing images on a camera (There are two backends, USB mass storage cameras, which are basically files on disk, and gphoto2 cameras, which require access through a library). Take the existing digikam icon view and delegate classes, which are prepared for code reuse, and put together an icon view for the model. Cleanly separate the code that does the actual work (downloading, converting, renaming) from the UI. Wrap that in the main window.

UI design: The current interface is powerful, exposing many options. We want to preserve that. But at the same time, there are three very common actions: a) Download all new files to the last used album b) Download all new files to a new album c) Download all new files to an existing album.
It should be possible to carry out task (a) with one click, task (b) and (c) with two or three clicks, without opening a dialog. Friendly to the new user, preserving access to all options for the poweruser.

Expected results: A camera UI based on Qt model/view and clean code which provides the currently available functionality, offering a quick path to download new pictures.

Knowledge Prerequisite: Qt, C++. Interest in Qt model/view and UI code.

Mentor: Gilles Caulier? (Marcel Wiesweg)

Project: Presentation View

Brief explanation: The presentation view is a full-screen workplace which you can use to present photos to yourself or your audience. At first glance it is looking like the current slide show, but then it is much more flexible: At any moment, you can access UI components like the map view, the metadata tab, the image properties tab, to access information, assign metadata, or show your audience the location of the picture on OpenStreetMap. You can access an icon view component to change the collection your are currently presenting, or a thumbbar component to switch to a different image.

The main view typically shows one image full screen, but you can zoom; You can also change to a layout that shows four or five images at a time, like images put in a paper album. You can change the order of images presented, and store order and layout (preferably in some XML format). You can load these presentations later, playing them automatically, coming back to the traditional slide show.

Technically, it should be future proof as much as possible (Qt Quick. QGraphicsView. scene graph in the future? Will embed QWidgets though) The job is not to develop any of the components mentioned above, but to avoid that, and reuse the existing digikam components and models.

Expected results: Something resembling the vision outlined above

Knowledge Prerequisite: C++, Qt

Mentor: Marcel Wiesweg

Project: MacOSX support

Brief explanation:

digiKam needs to be available under MacOSX in native. Currently, - Macports project is the only way to get last digiKam under Mac. Macports require to compile all depencies and digiKam as well. It's a long and hazardous solution to see digiKam running under Mac.

Also, current digiKam implementation is not optimized for Mac desktop. A lots of point need to be improved to support better this operating system. Graphical interface need to polished too.

See relevant entries in KDE bugzilla about MacOSX support :

257679 257773

Expected results: Provide scripts and configurations to build automatically a DMG archive of digiKam for MacOSX. Improve digiKam GUI everywhere to be more elegant and more optimized for MacOSX.

Knowledge Prerequisite: C++, Qt, MacOSX, scripting, DMG, packaging

Mentor: Julien Narboux ? (Gilles Caulier)

Project: Video metadata support

Brief explanation:

All recent digital-still camera device provide video capture. digiKam must be able to manage these files as images. digiKam can already play video and register files to the database, but it lack important metadata used to catalog and sort item (as date, camera name, and all record condition).

To improve video file support, video metadata management done in background need to be improved, to patch Exiv2 shared library (already used by digiKam)

See relevant entries in KDE bugzilla about video support :

164442 229543

Expected results: Add video files support to Exiv2. Patch digiKam metadata interface to handle video information.

Knowledge Prerequisite: C++, Qt, video format, metadata

Mentor: Gilles Caulier, Andreas Huggel

Project: Clone Tool for Image Editor

Brief explanation:

In digiKam image editor we need a simple clone tool to be able to remove quickly dusts, spots, and other unwanted artefact from an image.

See relevant entry in KDE bugzilla about it :


Expected results: add a new tool (as new image editor plugin) with the clone feature.

Knowledge Prerequisite: C++, Qt

Mentor: Gilles Caulier

Project: Panorama Tool

Brief explanation:

With recent digital still camera, it's possible using a tripod to take images to create panoramic view. We need a tool to assemble these images automatically. Common image corners must be detected and merged without to ask any settings to user. Colors and luminosity of each shot must be adjusted automatically too.

See relevant entry in KDE bugzilla about it :


Expected results: add a new tool (as kipi plugin) with auto-panorama feature.

Knowledge Prerequisite: C++, Qt

Mentor: Gilles Caulier


Project: KStars: Decouple computational code from the UI and Database

Brief explanation: KStars has its computational code tightly bound to the rest of the application. The aim of this project is to:

1. Investigate possible replacements for KStars' in-house computation, like libnova. It is important to keep KStars cross-platform. Hence we should look at cross-platform replacements.

2. Decouple KStars' interface and database from its computational code

3. Create a KStars computational engine library which contains all the computational code

4. Write test cases to ensure that

5. Get KStars to use this library

6. If time permits, write an interactive command-line client / python module / D-Bus interface to use this library.

More details will be fleshed out after a week or so.

Expected results: Clean, restructured code for KStars, which now uses a library that can be re-used in user-written scripts etc.

Knowledge Prerequisite: A good grasp over C++ and object-oriented programming, and Qt; writing test cases with Qt; should quickly learn how to write libraries; have a lot of patience; ability to solve software design problems

Mentor: Akarsh Simha <[email protected]>, Victor Carbune <[email protected]>

Project: KStars: Improve the observation planner and logging feature

Brief explanation: KStars already has an observation observation planner and logging feature. It requires a lot of tiny and not-so-tiny improvements though. Here's a list of some of the improvements we'd like to see: 1. "Smart" assignment of observation times to objects (accounting for time taken to observe each object) 2. A time-line view for your observation session plan, on which you can drag objects around to set observing times. 3. Ensure compatibility with OAL 2.0 XML schema is perfect 4. Currently, OAL parsing code is tightly tied to the rest of KStars -- separate it out into a Qt-based OAL library 5. Clean up the Observation Planner interface 6. Clean up the Execute Session interface. Fix bugs. 7. Bring the time-line view of a session into the Execute Session interface 8. Add the ability to export OAL-compliant target lists (removing the time field) 9. Add the ability to read and edit OAL-compliant observation logs within KStars 10. Add the ability to publish observation logs on a server.

More details to be fleshed out later.

Expected results: A better observation planning and logging feature for KStars.

Knowledge Prerequisite: C++ and Qt; XML parsing, in particular with Qt; be comfortable with desiging UIs using Qt

Mentor: Akarsh Simha <[email protected]>, Victor Carbune <[email protected]>

Project: KStars: Printing support

Brief explanation: KStars can currently print rendered images of the sky. But they are not well-annotated.

The aim of this project is to write code that will print star charts from KStars with explanations and annotation, and also print observing lists. Here's a mockup to give you a very rough idea of what we are looking for with regard to observing lists:


The project essentially aims to:

1. Create printed finder charts with Annotations, Multiple Fields-of-View, Legend etc.

2. Create printed observation logger forms with slots for writing down details and a representation of the FOV

3. If time permits, use an OCR library to permit conversion of printed forms into OAL-compliant reports to the best extent possible.

More details will be fleshed out later.

Expected results: Beautiful printed finder charts, printed observation log forms, and possibly digitized observation log reports using OAL XML schema

Knowledge Prerequisite: C++, ability to quickly learn KDE Printing libraries

Mentor: Victor Carbune <[email protected]>, Akarsh Simha <[email protected]>

Project: KStars: Make KStars more usable to the scientifically-inclined user

Brief explanation: KStars is currently good for education and for the hobby sky-observer. For those interested in more scientific aspects of astronomy, KStars does not have features to offer, but definitely does have the right existing framework to support these. The aim of this project is to write those lacking features, such as:

1. Plate solving and mapping of images: Look up astrometry.net. They have a library that can identify and produce the celestial coordinates of all objects in any given photograph of the sky. Use KStars to provide an interface to astrometry.net, and display plate-solved images on the sky-map. Use texture mapping to do this in OpenGL mode.

2. Retrieving scientific data: Make retrieval and visualization of data such as spectra of stars, variable star light curves etc. easy through KStars, via some simple GUI tools that initiate queries on various websites.

3. Astronomy updates: Monitor websites for updates such as comet, asteroid and supernova discoveries and alert the user about such events. (Maybe) keep updated catalogs of currently observable comets, asteroids and supernovae.

More details will be fleshed out later.

Expected results: KStars is more usable for the scientifically inclined user.

Knowledge Prerequisite: C++, Qt, Ability to quickly learn KDE's HTTP API, UI design with Qt

Mentor: Akarsh Simha <[email protected]>, Victor Carbune <[email protected]>


KDE-based Integrated Development Environment, specializing in c++ support, but including a powerful generic framework (definition use chain) which makes it possible to relatively easily support multiple different languages.

Website - Mailing list - IRC channel: #kdevelop on Freenode.

C++0x Support:

Brief explanation: Extend the C++ plugin to support C++0x features

Expected results: All new features that will definitely get into C++0x are supported in the parser. Correct DUChain integration is added and code completion works as expected.

Knowledge Prerequisite: C++, Qt, KDevplatform

Mentor: David Nolden

Test Framework:

Brief explanation: Revive the test framework and make it useable

Expected results: Integrated running of CTest suites with all the features one would expect from an IDE. Create a wizard to add new tests, with CMake integratoin and support for QTestLib for starters. Should be extensible in future.

Knowledge Prerequisite: C++, Qt, KDE, CMake, KDevplatform

Mentor: Aleix Pol

Improved PHP Support:

Brief explanation: Polish XDebug and it's dependencies, improve PHP language support plugin.

Expected results: Make the XDebug plugin work and bring it and its dependenciees into a releasable state. Improve PHP language support plugin by fixing the Namespace support, adding basic code completion for array keys, and hopefully other things.

Knowledge Prerequisite: C++, Qt, KDE, PHP, XDebug, KDevplatform

Mentor: Milian Wolff


Brief explanation:

Expected results:

Knowledge Prerequisite:


KDE Core

KDE Core is the interest group working on the underlying KDE development platform such as kdelibs and kde-runtime. KDE Core provides libraries that are used by all KDE applications and so essentially define what a KDE application is. Working on KDE Core is highly challenging as it requires much forward thinking for maximum utility and flexibility as well as guaranteeing backwards compatibility and cross-platform support.

Project: Support for astronomical calendar systems

Brief explanation: Add support for astronomical calendar systems. KDE is unique in the Linux eco-system for providing support for alternative calendar systems, such as the Hebrew, Islamic Civil, and Japanese calendar systems. Support for such calendar systems is standard in the Windows and Mac worlds. However, KDE does not as yet support calendar systems that require astronomical calculations, such as the Chinese and Islamic Lunar calendars, This project would fill this gap.

Expected results: Documentation, design and production of Chinese, Indian, Islamic and Jalali/Persian calendar systems and their numerous derivatives. The documentation to be of a high enough standard to submit to various standardization bodies. Production of an astronomical library for calculating sunrise, sunset and moon phase (and any other useful calculations) for shared use between the calendar systems and other KDE libraries and applications such as KHolidays, KStars and Marble.

Knowledge Prerequisite: C++, especially an understanding of Binary Compatibility rules and good API design. Some knowledge of Celestial Mechanics and good mathematical literacy. Any experience with regular cultural use of astronomical calendars would be highly useful.

Mentor: John Layt


KDE PIM is the interest group working on applications related to personal information management, e.g. contacts, calendar, mails, etc.

There are interesting projects on all levels of the software stack: libraries, application porting, new applications, access to online resources, etc.

Website - Project Wiki - Mailing list - IRC channel: #kontact and #akonadi on Freenode.

Project: Conversation view in KMail

Brief explanation: The preview pane in KMail should optionally be able to show all messages of a thread at once, which is also known as "conversation view" and popular from the Google Mail web interface. The message list probably needs a special mode that makes it more suitable for using conversation views.

Expected results: A smooth conversation view mode that not only adapts the looks of the preview pane and the message list, but also delivers a good interaction workflow. There needs to be some thought on UI and interaction design behind this to create great usability.

Knowledge Prerequisite: C++ required, Qt recommended

Mentor: Thomas McGuire

Calligra Karbon

Karbon is a vector drawing application with an user interface that is easy to use, highly customizable and extensible.

Web - Mailinglist - IRC channel: #calligra on Freenode.

Project: Variable thickness lines

Brief explanation: One of the most fundamental basics of drawing is varying the width of your lines to show shape, form and perspective. Almost every line tapers at either end, and often gets thicker and thinner in different places as needed. For purely technical and histrorical reasons though, every vector program (Illustrator, Inkscape, Karbon etc) make curves all one hard width.

This proposal is to modify the path a tool that, much like the path tool, would allow drawing curves, but where each node could have its width set so that the line width changed smoothly from node to node.

As Karbon is part of the Calligra suite, this would clearly benefit apps such as Krita, also.

Expected results: Modify path tool is able to scale the width of any path node to an arbitary percentage (say 155%) of the stroke width. See http://bugsfiles.kde.org/attachment.cgi?id=56995 for mockup.

Technologies Used: C++,Qt,SVG?

Mentor: Jan H? jaham @t gmx,net (need to ask :P )

Calligra Kexi

Kexi is an open source data management application, long-awaited competitor for programs like MS Access or Filemaker.

Mailing-list: https://mail.kde.org/mailman/listinfo/kexi/

Project Page: http://kexi-project.org/

Irc channel: #kexi on irc.freenode.net

Forums: http://forum.kde.org/viewforum.php?f=203

Development Wiki: http://community.kde.org/Kexi


Calligra Words

Web - Mailinglist - IRC channel: #calligra on Freenode.

Project: Improve quality of ODF write support

Brief explanation: While a lot of work went into improving the quality of loading ODF text-documents, saving them can still be improved a lot. The goal of the gsoc would be to improve the quality of the by Calligra Words produced ODF. This means fixing the produced ODF text-documents so they a) do validate against the ODF XML-schema and b) are proper loaded with Calligra Words, LibreOffice.org/OpenOffice.org, Microsoft Word, etc. By identifying and fixing bugs and adding unittests for regression-testing we could improve the quality of the core of Calligra significantly.

Mentor: Sebastian Sauer, please contact [email protected]

Project: Implement notes-support

Brief explanation: Implement and extend the notes-support in Calligra Words to view+add+edit+delete+load+save notes/comments/annotations in Calligra Words.

See also bugs 260138 and 260184 and 260102 and 260127.

Mentor: Sebastian Sauer, please contact [email protected]

Project: Fix and extend LaTEX support

Brief explanation: Calligra Words has an import and export filter for LaTEX documents. Those need to be extended to proper import/export and produce better results. We could also implement a new view in Calligra Words to support latex like working.

See also bugs 260063 and 260056.

Mentor: Sebastian Sauer, please contact [email protected]

Project: PDF-Import and/or PDF-Export

Brief explanation: Currently Calligra Words supports exporting the drawn canvas as images using the File=>ExportPDF menu-item. The idea is to create an export and/or import filter to import and/or export to/from PDF (not as images but as text) using poppler.

Mentor: Sebastian Sauer, please contact [email protected]

Project: Integrate with Akonadi

Brief explanation: Integrate with Akonadi to provide access to resources (contacts-variables, mail-merge, ...).

See also bug 260220.

Mentor: Sebastian Sauer, please contact [email protected]

Project: References tool

Brief explanation: Words is tools based and has a write tool, a review tool as well as the beginings of a layout tool. What we don't have yet is a references tool that provide the ui for creating table of contents, citations/bibliography, and end notes. Table of centents has all the underlying engine in place. That same engine can be reused/copied/modified for bibliography (maybe planning for the use of http://www.zotero.org )

Mentor: Casper Boemann, please contact [email protected]

Project: Layout tool

Brief explanation: Words is tools based and has a write tool, a review tool as well as the beginings of a layout tool. What we need is completion of this layout tool. Much of it is just ui allowing the user to set all sorts of layout options using the engine which is already in place.

As it's not such a big project, we expect everything about this tool to be perfect so it can enter directly into the next version of Calligra Words.

Mentor: Casper Boemann, please contact [email protected]

Calligra Krita

Krita is a KDE program for sketching and painting, offering an end–to–end solution for creating digital painting files from scratch by masters.

Mailing-list: https://mail.kde.org/mailman/listinfo/kimageshop/

Project Page: http://www.krita.org/

Irc channel: #krita on irc.freenode.net

Forums: http://forum.kde.org/viewforum.php?f=136.

Wiki: http://community.kde.org/Krita

Project: ABR Brush Support

Brief Explanation: Your task will be implement support for various features from brushes available in Photoshop. You will be extending current brush engines in Krita with set of features like texture painting, shape dynamics, flow, wet edges...The format is not documented and you will work with reverse-engineered information. Your task is to implement missing features, add mapping of the binary format to Krita XML format and possibly improve the reverse-engineering information.

Mentor: Lukáš Tvrdý

Used technologies: C++, Qt

Resource: http://community.kde.org/Krita/ActionPlan2#Photoshop_brush_support_in_Krita

Project: PSD File import/export Support

Brief Explanation: The industry standard for raster graphics is still Photoshop. This file format is closed. This project will entail bringing together the freely available information on the PSD file format and build an import/export filter upon the existing framework in Krita.

Mentor: Boudewijn Rempt

Used technologies: C++, Qt,

Resources: existing open source implementations like http://sourceforge.net/projects/libpsd/ and http://git.gnome.org/browse/gimp/tree/plug-ins/file-psd

Project: Flow and drying: real media support

Brief Explanation: As a painting application, Krita tries to make things possible for digital artists that were previously only available in “real” media. The effects of paint flowing, drying and being absorbed can be used to create interesting effects. Particularly challenging is the flow of undo and redo.

Mentor: Boudewijn Rempt

Used technologies: C++, Qt

Resources: http://www.billbaxter.com , http://www.levien.com/gimp/wetdream.html , http://research.edm.uhasselt.be/~tvanlaerhoven

Project: 3D Sketching tool

Brief Explanation: When drawing comics, it can be particularly challenging to draw in the correct perspective. With a similar interface to Manga Studio, this project entails extending the guided painting interface of Krita into the third dimension, making sure parallel lines drawn by hand will have the correct perspective. This project can be extended by making it possible to import 3D objects created by, e.g. Google Sketchup

Mentor: Lukáš Tvrdý or Cyrille Berger

Used technologies: C++, Qt,

Project: Text Balloon support for Comics work

Brief Explanation: Comic books artists are one of the target user groups for Krita. A challenge when working on comic books is the placement and lettering of the text balloons -- and the internationalization! This project will entail creating vector-based support for translatable balloons that contain text that looks hand-written.

Mentor: Boudewijn Rempt

Used technologies: C++, Qt, SVG

Expected results: a layer type that contains vector-shaped balloons of various types and textual content that can easily be internationalized.

Resources: http://en.wikipedia.org/wiki/Speech_balloon

Project: Painting and Separation of 3D Textures

Brief Explanation: As one of it’s use cases, Krita’s vision statement includes painting textures for 3D. 3D textures are typically comprised of a number of separate black and white, RGB or RGBA images. Thus painting for textures typically requires painting on more than one single layer / channel at a time. For example painting a scratch into a metal surface may require painting black onto the bump channel, another colour on the diffuse channel, and another to the specularity channel (as well as possibly some others such as the displacement channel). All of these are affected simultaneously.

Currently Krita’s painting system is only able to paint onto single layers at a time and brushes have not been designed in such a way as to allow adjusting multiple channels simultaneously as would be needed. This topic would require looking at how Krita’s current painting system could be extended to paint the necessary adjustments to the channels used in 3D rendering, show the textures created in OpenGL and then export those channels for use in 3D applications.

Mentor: Lukáš Tvrdý

Used technologies: C++, Qt

Resources: http://www.krita.org

Project: Advanced selection using SIOX

Brief explanation: SIOX stands for Simple Interactive Object Extraction and is a solution for extracting foreground from still images with very little user interaction.

Possible mentor: Cyrille Berger (or Sven Langkamp)

Used Technologies: C++, Qt


Project: Tagging and management for Krita resources

Brief explanation: Krita comes with a rich selection of resources: patterns, gradients, brushes, brush presets, soon materials for texture painting. These resources need to be managed: added, deleted, changes and tagged. Existing tagging specifications exist for Gimp and for Viaduct, and Krita should be compatible here. Integration with Get Hot New Stuff and Nepomuk are important aspects of this project. Especially interesting here is to make this system fit in the workflow of artists working on textures. There will be data structures and gui work.

Expected results: A functioning implementation of resource management and tagging.

Used Technologies: C++, Qt, digital art.

Mentor: Boudewijn Rempt

Project: GPU-acceleration for Krita

Brief explanation: Although Krita was one of the first painting applications with OpenGL and shader support, Krita uses the CPU for all its calculations. Until now the OpenGL and shader support have been used solely for display purposes, like the 3D brush model, real-time preview of gradients on the canvas. We would like to improve on this by using the GPU to improve the performance of blending layers, painting and transforming. Expected results: optimized composition operators, transformation code Used Technologies: C++, Qt, OpenGL, GLSL, OpenCL, General-Purspose GPU coding

Mentor: Lukáš Tvrdý

Project: GPU-backend for OpenGTL

Brief explanation: Krita uses the OpenGTL family of languages for easy creation of filters, colorspaces and possibly brush engines. OpenGTL currently uses LLVM to compile these scripts to code that runs on the CPU. Alternatively, compiling to the GPU would mean considerably performance improvements. There are two possibilities to implement such a change:

  • Convert from the OpenGTL languages to OpenCL/GLSL and then run the converted shader on the GPU, this can be done by writing an AST output backend
  • The mesa library uses llvm for compiling to the GPU from OpenCL/GLSL, so it should be possible to bypass the conversion and plug OpenGTL directly on the mesa library

The first possibility has the advantage to be more portable, the second solution might offer performance gain. The conversion approach should be implemented first, and then, the students could investigate the use of the mesa library when available.

This is a very challenging task!

Resources: http://www.opengtl.org

Expected results: OpenCL AST generator, report on the possibility to interface OpenGTL with mesa

Used Technologies: C++, GLSL, OpenCL (possibly LLVM)

Mentor: Cyrille Berger

Project: Substrate simulation

Brief explanation: Substrates for painting and drawing have a direct influence on the result of the art. The goal of this project is bringing this richness of these effects to Krita. There is an existing body of literature and academic projects on this topic, with Bill Baxter and Tom van Laerhoven being among the best known researchers. For the right effect, we need to take care of the 3d structure of the substrate, it’s effect on paint tools -- smoothness, absorbency and other parameters. An interesting option is to make it possible to apply different substrates to existing paintings. Expected results: an optional substrate simulation that works with all current Krita tools

Used Technologies: C++, Qt, OpenGL

Mentor: Boudewijn Rempt

Project: Color Mixing

Brief explanation: Building on Emanuele Tamponi’s ground-breaking work on spectral colorspaces and color mixing, the topic of this project is to extend and finish this work and make is usable in Krita. Spectral colorspaces have the advantage that mixing colors produces the same result as in when mixing real-world paints. Problems are transparency (alpha channel) and composition.

Expected results: a color mixer that works, layer composition and painting.

Used Technologies: C++, Qt, color theory, mathematics

Mentor: Boudewijn Rempt

Project: 3D Material Image Maps

Brief explanation: 3D materials are made up of a bunch of images called image maps. If the user could associate layers as image maps in Krita, and paint on all of them at the same time, artists could paint whole 3D materials - something currently only available in high end 3d apps like zBrush (not even Photoshop / Corel Painter). The trick is that the position of what's painted needs to match on every map/layer, but the colours vary. For example, a scratch on a human characters skin would have a red colour map, a white (=raised) bump map, a light grey (=semi-shiny) specularity map etc, all in the exact same location on the each image map. Traditional methods of trying to create each image from scratch or by manipulating the colour map are very, very slow and painful. A simple version of this could be done as follows:

1. Each layer has a toggle in the layers docker called "texture map" or similar. This is turned off by default. When active, the brush paints on *all* layers that currently have "texture map" active.

2. When picking a colour, a dropdown lets the user pick "Default" or any active texture map layer. "Default" is just the current behaviour. If the user selects a layer in the dropdown, then the selected colour will be applied to that layer when painting on *any* layer.

3. In the file or layer menu is an option "Export texture maps" which saves each texture map layer as an image. The layer name and extension appended automatically to the file name. For example, on a file called character.kra, the layer titled "colour" would be saved as "character-colour.jpg" (or whatever format was selected).

For step 3, a simple, one click / shortcut, method is vital, as artists often have to check their material in their 3d app every few minutes, and wading through saving 10 layers individually, each with manual file naming and confirming file format settings each time is unacceptably slow. For any artist who requires this level of control, they can use Layers menu -> "Save Layer as Image" already in krita.

Allowing artists to paint a single material rather than creating multiple separate image maps by hand, would make Krita formidable for painting 3D textures, and the most advanced open source application for 3D texturing.

Used Technologies: C++, Qt, color theory, mathematics

Mentor: Boudewijn Rempt

Project: Shift drag sensors

Brief explanation: Currently shift+dragging the left mouse button horizontally can be used to alter the brush size. This is awesome as it minimises time wasted going back and forward between where you're painting and the brush editor a million times. It would be even better if brush softness, hue, value, page zoom etc could be changed in a similar way. Shift + drag could be used with any othe three mouse/stylus buttons (left - LMB, middle - MMB, and right - RMB click) both horizontally and vertically, giving us up to 6 things that could be quickly changed, right at the users cursor. This proposal is to make a system to link any of these shift+drag options to modify brush properties or page zooming in a fast, intuitive manner. For example shift + vertical LMB drag could modify brush softness, shift+rmb horizontal drag could be assigned to changing opacity, shift + mmb horizontal could be assigned to modifying brush hue, shift + mmb + vertical drag could be assigned to adjusting brush value and so on. Possibly these could be implemented as sensors and thus assigned to anything in a brush that currently has a curve.

Used Technologies: C++, Qt, color theory, mathematics

Mentor: Boudewijn Rempt

Calligra Stage

Stage is an easy to use yet still flexible KDE-based presentation application.

Mailing-list: https://mail.kde.org/mailman/listinfo/calligra-devel

Project Page: http://www.calligra-suite.org/stage/

Irc channel: #calligra on irc.freenode.net

Forums: http://forum.kde.org/viewforum.php?f=203

Wiki: http://community.kde.org/Calligra/Stage

Project: Plasma-based Presentation View

Brief explanation: The presentation view in presentation applications is used in multi-monitor displays to display information about the presentation on one screen while the full slides appear on another screen. This provides information to the presenter that is not visible to the audience. This information often includes a preview of the current and next slide, timing information, notes, and a list of slides for navigation.

Stage already has a basic presentation view, but it is fairly limited in its capabilities and rigid in its layout. This project would be to design a new presentation view for Stage using plasma, ideally using QML-based plasma widgets. It would involve developing a plasma containment for the presentation view and plasma widgets for the individual elements needed by the view, as well as an interface for setting up the view. These elements would, at a minimum, include: a preview of the current slide, a preview of the next slide, a list of slides, notes for the current slide, next and previous slide buttons, and a timer. Users would need to be able to place, position, and resize the elements. The presentation view would also need to support multiple different views that could be easily switched between during a presentation.

Used Technologies: C++, Qt, QML, Plasma

Mentor: see mailing list: https://mail.kde.org/mailman/listinfo/calligra-devel

Project: Single-canvas Presentation Mode (like Prezi or Zippl)

Brief explanation: Unlike normal slide-based presentation tools, single-canvas presentation tools have a single infinite canvas. Parts of that canvas are set of as frames, and the presentation progresses by moving from one of these frames to another in a pre-defined or arbitrary manner. Alternate routes between frame scan be defined, as well as branch points where the presenter can select between multiple different ways to proceed from a given frame. This allows for a non-linear presentation.

This project would involve implementing a single-canvas presentation mode in Stage. Users would need to be able to set up a canvas, set aside parts of that canvas as frames of arbitrary size, set paths between those frames (including a default path and alternate paths selected by clicking on elements in the presentation), zoom out to view the entire canvas, and zoom in on arbitrary parts of the canvas.

See here for examples of other implementations:

Used Technologies: C++, Qt

Mentor: see mailing list: https://mail.kde.org/mailman/listinfo/calligra-devel

Project: Unified Slide Manager

Brief explanation: Stage currently has three different, independent tools for managing slides: the Document Panel, the Slide Sorter, and the Custom Slideshow Editor. These use completely different interfaces and have different capabilities (for instance only the Slide Sorter has drag-and-drop). However, they are all used to manage slides in your presentation.

This project would be to combine them into a single interface. The interface would normally reside in a docable panel similar to the existing Document panel. By pressing a button, it will expand to take up most of the window, behaving as the Slide Sorter. In this mode, there should be an additional, normally hidden panel at the bottom that the user can use to create add and remove slides from custom slideshows.

In the process, the existing Document Panel and Slide Sorter should be modified to take better advantage of features provided by KDE and to provide better consistency with other KDE programs. Changes would likely include: overlay buttons when you hover on a slide, drag-and-drop interfaces, menus on drag-and-drop, and selection effects similar to those seen in KDE-based file managers.

See here for descriptions and mock-ups:

Used Technologies: C++, Qt

Mentor: see mailing list: https://mail.kde.org/mailman/listinfo/calligra-devel


KDE's window manager

Techbase page - Mailinglist - IRC channel: #kwin on Freenode.

Project: Modularization of Workspace

Brief explanation: The Workspace class is the monolithic core of KWin. Nevertheless many parts of it do not depend on each other and can be split out of Workspace into own classes. The modularization is an important prerequisite to a port of KWin to Wayland and providing a small window manager for mobile devices.

Expected results: A concept for what can be moved into own classes exists and several classes have been split out of Workspace. The X11 dependend code is moved into an own plugin. Plus if tests for the new classes are written.

Knowledge Prerequisite: Qt/C++ and general understanding of window management. Plus for X11 knowledge

Mentor: Martin Gräßlin (mgraesslin)

Project: Unit Testing Framework

Brief explanation: Testing a window manager is rather difficult as it requires a running instance and windows need to be created and interacted with. This project should set up a test framework based on Xephyr, KWin's scripting interface and QTest.

Expected results: An infrastructure to test the core of KWin with simple KWin scripts should exist and several unit tests for existing functionality should be implemented

Knowledge Prerequisite: Qt/C++, JavaScript, QTest

Mentor: Martin Gräßlin (mgraesslin)

Project: Initial Support for Wayland Clients

Brief explanation: Wayland is the next iteration for composited window management. The task of this project is to add support for managing Wayland clients on X11 and integrate them into the existing compositing rendering code for GLES.

Expected results: KWin is able to manage Wayland clients and can render the clients. Plus for support to interact with them (move/activate/restack).

Knowledge Prerequisite: Qt/C++ and general understanding of Window management. Plus for OpenGL knowledge

Mentor: Martin Gräßlin (mgraesslin)


Website- Documentation/Howtos - Ontologies - Mailing list - IRC channel: #nepomuk-kde on Freenode.

(Also see the Nepomuk techbase page for a long list of Nepomuk-related ToDos and ideas.)

Project: A "real" query parser

Brief explanation: Nepomuk has a query parser which is used to handle input from the user in such situations as the KRunner interface or the Dolphin search bar. Currently this parser is one big HACK and in dire need of reimplementation.

Expected results: A desktop query parser which is based on an actual grammar including support for nested terms (important when it comes to querying a graph), quotes (used to group keywords into one search term), resource URIs (it should be able to detect that a string is a resource URI and not a full text search term), namespace abbreviations ("nao:hasTag" should match exactly the one property while "hastag" can match multiple properties), as well as property and type expansion (this means that in "hastag:xyz" "hastag" should be expanded to an actual property URI.)

Knowledge Prerequisite: Previous experience with a parser generator would be good but is not required. A bit of Qt experience would be nice but again, is not required.

Mentor: Sebastian Trueg <[email protected]>

Project: Query builder widget

Brief explanation: The query builder widget is tightly coupled with the above project of the query parser. They might even be merged into one single project. The idea is fancy auto-completion in the search line-edit that proposes types, properties, and other resources. It should take usage frequency of properties, types, and property values into account (when that does not mean too big a performance downer). It could even try to use some built-in thesaurus allowing to match "nfo:Audio" from "audio", "music", "song", and others.

Expected results: A fancy looking search line-edit with auto-completion that helps the user build a query.

Knowledge Prerequisite: Qt and KDE development experience helps.

Mentor: Sebastian Trueg


Brief explanation:

Expected results:

Knowledge Prerequisite:



Website - Mailing list - IRC channel: #plasma on Freenode.

Project: QML QtComponents set

Brief explanation: The QtComponents project is aiming to provide an api and a series of widget sets completely based upon QML. the actual implementation is platform-dependent, so KDE needs its own platform specific set made with the plasma theming mechanism,for both the desktop and the mobile.

Expected results: A complete set of QtComponents for the use in QML based plasmoid for the desktop, if there is enough time, the start of a mobile version (with as less as possible changed from the desktop case)

Knowledge Prerequisite: both QML and Qt C++ programming are useful (and preferred), most of it can be learned on the way

Mentor: either Marco Martin and/or Artur De Souza

Project: Educational desktop

Brief explanation: The flexibility of the Plasma Desktop shell can bring desktops fine tuned to a particular use case: With the KDE edu project, KDE has a strong offering of educational software. To further leverage the stength of the KDE platform, a customized Plasma desktop for the educational environment can make KDE the obvious choice.

Expected results: A set of plasmoids and a custom plasma layout (with eventual modifications to the desktop shell itself) to do a strong customization. In designing those plasmoids, the mobile use case has to be taken into account as well.

Knowledge Prerequisite: Qt/C++, probably QML


Project: Plasma media center first release

Brief explanation: The Plasma media center project needs some work to get to a first alpha release quality: mostly a reingeneering of the components of the main gui and porting the graphical elements to qml

Expected results: a basically working media center at least with an index of local media files (videos, pictures and music)

Knowledge Prerequisite: Qt C++, QML and some ideas on the Plasma architecture, QML can be learned on the way

Mentor: Marco Martin

Project: QML-ify Plasma widgets

Brief explanation: The Plasma2 library will be heavily based on QML, probably dropping qgraphicswidget support altogether. All the default plasmoid set will heve to be ported to be either pure QML/Javascript or QML/C++ in the meantime.

Expected results: At least 6-7 minor plasmoids ported to QML or some (to define) of the main ones (like kickoff,taskbar, systray, whatever)

Knowledge Prerequisite: Qt C++ some idea on Plasma and kdelibs arch, already knowing something of QML helps

Mentor: Marco MArtin

Project: Move Plasma Functionality in Compositor

Brief explanation: Plasma uses lots of functionality which are better served in the window and compositing manager (KWin). For example Plasma uses an own glow animation before showing a hidden panel. This could be OpenGL accelerated when moved into KWin. Other examples involve Wallpaper rendering or Tooltips.

It would be the task of the project to identify these areas and to discuss strategies with the Plasma and KWin community how to better handle these areas and to implement the solution.

Expected results: Rendering functionality moved from Plasma to KWin with fallback for legacy (non-composited) and other window managers.

Knowledge Prerequisite: C++/Qt, basic understanding of X, OpenGL useful but not required.


Project: Helping in libplasma2 development

Brief explanation: There are several things that needs to be done in both designing and preparing libplasma for the next major redesign.

Expected results: A thing can be building a working branch on git that doesn't have any derecated class of functions and is completely without widgets (so will be unable to load classic c++ plasmoids) but that can still load qml based ones.

Knowledge Prerequisite: C++/Qt


Project: Plasma mobile components

Brief explanation: Develop (and polish the ones already started) components for a classic mobile ui, such as application laucher grid, plasmoids container, notifications and systemtray plasmoid and so on.

Expected results: A quite classic mobile shell at a near production quality, while alternate containments and plasmoids will make posible to experiment new and more radical user interfaces.

Knowledge Prerequisite: C++/Qt, QML

Mentor: Marco Martin


Abstraction library for sound and video support. Used by KDE notifications, Amarok, Dragon Player and Qt Software.

Website - Mailing list - IRC channel: #phonon on Freenode.

Community wiki with TODOs and outstanding issues.


Brief explanation:

Expected results:

Knowledge Prerequisite:



Kate is a powerful programmer's editor.

Website - Mailing list - IRC channel: #kate on Freenode.

Project: Modeline Editor

Brief explanation: Kate supports modelines, also called document variables. Through this, users can configure individual settings for specific documents. The task of this project is to add a modeline editor, that can write/change the document variables in the document.

Expected results: A modeline editor that is able to edit the current modeliens. This editor could be reused in the "Modes" tab in Kate's config page.

Knowledge Prerequisite: C++/Qt, Qt-Designer, high motivation

Mentor: Dominik Haumann / Christoph Cullmann

Project: Further improve the Vi Input Mode

Brief explanation: Kate supports a modal, Vi[m]-like editing mode. The support for Vim features is already quite good in some areas, but there is more to be done:

  • A “jump list” should be maintained of the last cursor positions, making it possible to jump back to earlier positions and forward again (ctrl+i/ctrl+o in vim)
  • More command line mode commands should be supported, and it should be possible to use marks in ranges, i.e. “:'a,'bs/foo/bar/”
  • A thorough test suite
  • [your favourite feature]

Expected results: An improved vi input mode for kate

Knowledge Prerequisite: C++/Qt, high motivation

Mentor: Erlend Hamberg

Project: Elastic Tabstop support

Brief explanation: Elastic tabstops are a way to visually align code automatically. Current ways of indentation are a relic from the typewriter era and it is time to try new concepts. Elastic tabstops

  • Put an end to the “tabs vs spaces” debate
  • Save the programmer from tinkering with alignment
  • Allow the use of proportional fonts
  • Degrade gracefully for non-supporting editors

Apart from implementing the feature itself, one should care about the following:

  • the option “align dynamically wrapped lines to the indentation level” has to be extended to allow the setting “align dynamically wrapped lines to the last tabstop”
  • A way has to be found to integrate existing options such as tab width and indentation step width.

Expected results: A working option to use elastic tabstops instead of traditional alignment.

Knowledge Prerequisite: C++, Qt

Mentor: Joseph Wenninger


Rekonq is a web browser for KDE based on WebKit. It first focuses on being a light, fast & clean way to access to net. Its development is doubly based on using the new amazing features offered by the WebKit rendering engine and on the rock solid network KDE technologies.

Project: Adblock improvements & elements manipulation

Brief explanation:

Rekonq currently has an initial implementation of an ads blocking mechanism. This project aims to expand it providing the feature parity with AdBlockPlus and implement a sort of HTML elements manipulation system (enable it, and then click on the page on the elements you want to hide. When you're ok, save your changes, reset them or leave them just for this time)

Expected results:

An improvements in the adblock rules handling, some (new) tests to check adblock behavior and performance, the elements manipulation feature with (possibly) the ability to save and remember applied changes.

Knowledge Prerequisite:

C++ programming and hopefully Qt

Mentor: Andrea Diamantini

Project: Plasma KPart for Rekonq

Brief explanation:

Rekonq provides a specific webpage displayed when the user clicks on the "new tab" button. This webpage has features like: previews of favorite pages, previews of closed tabs, an history list, a bookmark list, and a download list. The goal of this project is develop the same features as this webpage but using plasmoids displayed in a Plasma KPart. Additional plasmoids, eg. RSS feeder, could be integrated.

Expected results:

  • Customizable new tab page.
  • Plasma theme integration
  • Same plasmoids inside Rekonq and on the desktop

Mentor: Lionel Chauvin


An open personal cloud which runs on your personal server. It enables accessing your data from all of your devices. Sharing with other people is also possible. It support automatic backups, versioning and encryption.

Website - Mailing list - IRC channel: #owncloud on Freenode.

Project: Local sync client

Brief explanation: Build a client to sync your ownCloud files with your local disc to enable offline use.

Expected results: Development of an application you run on your local PC. This applications syncs local folders with folders on your personal ownCloud server everytime you are online. This applications needs a GUI written in KDE/Qt to configure the ownCloud url, folders, login and password. The applications sits in the systray and informs the user about the syncing progress or sync conflicts. The idea is that the clients mounts the ownCloud folders via WebDAV into a "hidden" directory and syncs the folders via rsync or an own syncing script. Errors should be reported to the user. To quote Frank from the mailing list:

"The client should:

1. read a configuration file so see which folders to sync and which owncloud server to use.
2. mount the owncloud server via fuse or a different tool into a hidden directory.
3. call a syncing tool like csync to sync between the local directory and the mounted directory
4. unmount owncloud
5. write a log-file about the synced files and potential conflicts which the user has to solve.

The client should be as portable as possible of course."

Current inactive projects such as Cloudsync or owncloud-sync-client (both on Gitorious) may possibly be starting places worth investigating.

Knowledge Prerequisite: Python, Ruby, PHP or C++ and KDE/Qt depending on the technology you choose for the desktop syncing client. Enthusiasm for Cloud/Desktop integration and trying new things.

Mentor: Kyle Cunningham? Robin Appleman? Frank K?

KDE Usability

Project: Usability Survey Framework

Brief explanation: Surveys are one of many methods used in creating more usable software. Surveys allow designers to collect information about user's experiences and usability problems with software. A Usability Survey Framework would allow designers and developers to create small, custom surveys that can be attached to events or services. Users could subscribe to the Usability Survey Service and opt in to current studies. Survey studies could be one-time surveys, or several survey entries over a period of time.

For example, if we wanted to learn more about how users interact with workspaces, a survey could occasionally open when a user accesses or configures the workspace. If we wanted to learn more about the Plasmoid installation process, we could open a survey when a user installs the next plasmoid.

A Usability Survey Framework would help users become more involved in KDE through direct participation. Developers would be able to easily design and launch surveys to collect important usability feedback. Designers would be able to easily conduct usability studies.

Expected results: The Usability Survey Framework could be implemented as a web service, downloadable Plasma widget, or mini application. It should be an easily configurable survey that collects data based on an event.

Possible features:

  • Uses XML for survey design to make it easy to create and launch new surveys
  • Uploading or emailing data at end of study
  • Can be a single survey, or a survey study over a period of time
  • Configured to open on a specific event

Knowledge Prerequisite: JavaScript, C++. Qt/KDE/Plasma knowledge would help greatly.

Mentor: Celeste Lyn Paul, someone else?


Umbrello UML Modeller QGraphicsView Port:

Brief explanation: Umbrello UML Modeller uses the old and limited Q3Canvas class. There is an unfinished port to QGraphicsView that should be completed.

Expected results: A stable Umbrello using QGraphicsView.

Knowledge Prerequisite: Programming Qt. Basic understanding of UML.

Mentor: Jonathan Riddell


Project: One-click Spammer removal

Brief explanation: Currently Wiki administrators have to go to the spammer's user page and block him for a specified time. Then any pages added by him must be deleted and finally any images added must be found and deleted.

Expected results: It would be very useful if administrators could call up the Ban page and be able to remove pages and images with a single click from there. Ideally if spam had been added to an existing page that would be reverted, but this is a less common scenario. The work should be presented as a mediawiki extension.

Knowledge Prerequisite: ?

Mentor: ?


Project: Advanced text layout recognition engine

Brief explanation: Okular has a very naive algorithm to detect the layout in text in which basically everything is considered to be layouted in a line. You will need to know some text layouting algorithms to improve the detection of columns in text so that text selection works better.

See relevant entry in KDE bugzilla about it: 161324

Expected results A better text selection experience for the user.

Knowledge Prerequisite C++, Qt

Mentor Albert Astals Cid


Knights is a chess program for KDE, it resides in Extragear/Games. It supports local plays, playing against a computer engine, an opponent on a chess server, and also watching two computers. It uses the KDE technologies to provide a consistent look-and-feel with Oxygen colors and Plasma clocks.

Knights web site

Project: Saving, loading and analyzing games

Brief explanation: Knights currently does not support any form of storing and analyzing games, expect for simple undo and redo. This suffices for casual play, but for serious play something more powerful is needed.

One part of the project is to enable saving board positions to a file, and reading from it at a different time. This should be done using a standard format (PGN), so that external games can be analyzed with Knights and vice-versa.

Another part is an "analysis" mode, where a playerc can freely move both his and the opponent's pieces. The interface should rate the current board position, provide hints, maintain alternative timelines for comparison, etc. The student should be in contact with chess players (could be on a forum or IRC) to know the other important features of such a mode.

Expected results: Ability to save move history to a file, and read from it at a later time. An advanced 'analysis' mode for direct manipulation and interaction with a chess engine, with tools to help the player.

Knowledge Prerequisite: Qt, C++. Knowledge of Chess rules is not needed, but some contact with chess players is preferred.

Mentor: Miha Čančula


Gluon is a Free and Open Source framework for creating and distributing games - supporting the flow of the idea all the way from the author to the player of the finished game, and back.

Gluon Website

Contacting the Gluon team (irc, email etc)

Project: Achievements/statistics system

Brief explanation: Many digital distribution platforms, such as XBox Live, PlayStation Home and Steam, have a system by which players are granted certain trophies by performing certain tasks in the games they play. Even though they can be, these tasks are not necessarily a part of the normal game-play: For example it might be that you have played a certain level a specific number of times. The Open Collaboration Services have, since version 1.7, contained a generalised system for storing achievements and the progress a user has made on them.

This project is to integrate the OCS Achievements module into Gluon, and to create a general system by which statistics required for tracking progress in a game can be generated.

Expected results:

  • A set of Components and Assets for storing and tracking statistics of game interactions
  • Same for handling the Achievements themselves

Knowledge Prerequisite: Qt/KDE development (includes C++ & git). Knowledge of achievements systems a plus.

Skill level: medium to high.

Mentor: Arjen Hiemstra and Dan Leinir Turthra Jensen

Project: Dynamic Shader Generator

Brief explanation: Gluon these days features a completely shader-based graphics library. One of the issues with this is that you need to create shaders, which is not trivial for most people. This project would alleviate that problem by creating a system similar to many professional graphics applications, where shaders are created by linking small elements - nodes - together. If time permits, this would also be exposed in the interface properly through a node-based editor.

Expected results: A system that is able to create basic shaders and can be extended to create more advanced shaders.

Knowledge Prerequisite: Qt/KDE development (includes C++ & git), OpenGL/GLSL and a general knowledge of graphics programming is a definite must.

Skill level: medium to high.

Mentor: Arjen Hiemstra and Dan Leinir Turthra Jensen

Project: Integrating SMARTS Game AI system

Brief explanation: In the fall of 2009 and spring of 2010, three students at Aalborg University created a game AI system based on the simple Behavior Tree concept for use with Gluon. As this was an educational project, however, while the project was completed, the code was never merged into Gluon. This project is to take the project code, clean it up and adapt to new Gluon APIs, and if time permits to create a more pleasantly built and better integrated editor for the behavior trees.

Expected results:

  • A new sub-module in Gluon containing the standalone parts of the SMARTS AI system (gitorious)
  • Cleaned up integration of the SMARTS components and assets
  • Optional: An editor KPart for integration directly in Gluon Creator

Knowledge prerequisite: C++ development experience, Qt/KDE development experience an advantage but can be disregarded if knowledgeable of other modern OOP frameworks.

Knowledge of the Behavior Tree concept in general an advantage, but this can be gained in a reasonably short amount of time (as both video tutorials on the concept as well as the reports from the university projects SMARTS resulted from are available, containing descriptions of this).

Skill level: Novice to medium.

Mentor: Dan Leinir Turthra Jensen and Arjen Hiemstra


Telepathy is a cross-desktop framework for real-time communication and collaboration - think IM, Voice/Video Conferencing and Collaborative document editing/gaming/etc.

More information:

Project: Distributed Shared-State system based on DBus tubes for KDE apps

Brief explanation: Telepathy DBus tubes provide an easy way for collaborative applications such as KWhiteboard to be built. However, a common issue is the need of applications wishing to use DBus Tubes is the need for a way for participants to keep their application states synchronised in the absence of a single "server". This is an intellectually challenging project as well as requiring coding, so please ensure you join us on IRC and get to know us and the project before submitting a proposal for this project.

Expected results:

  • A distributed-state mechanism generic enough to be used by any KDE application wishing to build a serverless collaborative system on top of DBus tubes.
  • A sample implementation using this state layer, either by adding basic functionality to KWhiteboard or another application of your choice.

Knowledge Prerequisite:

  • Qt/KDE development (includes C++ & git).
  • Telepathy development
  • As a minimum, a basic understanding of how DBus works
  • A desire to take on a challenging project

Skill level: high.

Mentor: George Goldberg

Project: Amarok Playlist Sharing

See [1]

Project: Innovative new UI/Interaction Methods

Brief explanation: Telepathy allows real-time communication to be integrated far deeper into the computing experience. Come up with some interesting novel ideas for how to make this integration happen (new ways of displaying information, novel plasma integration methods, you name it), and implement them.

Expected results:

  • This project is very open ended, but we would expect to see a series of deliverables outlined in the proposal along with some justification for why they are interesting ideas. Sufficient research should be undertaken at the proposal stage to ensure the deliverables are suitable and achievable within the context of the rest of the KDE Software Compilation and Workspaces.

Knowledge Prerequisite:

  • Qt/KDE development (includes C++ & git).
  • At least basic Telepathy development
  • A good imagination and a keen interest in innovative interaction design.

Skill level: medium.

Mentor: George Goldberg

Project: Multi-Player gaming in KDE games

Brief explanation: Telepathy tubes allow collaborative features to easily be added to any application. This project is to develop multiplayer support on top of telepathy for the KDEgames suite of games. If you wish to propose this idea, it is important to discuss your proposal both with the KDE games team and the KDE Telepathy team.

Expected results:

  • Infrastructure for Telepathy Tubes based multiplayer gaming in KDE games.
  • Support for one or more games being played multiplayer over Telepathy Tubes.

Knowledge Prerequisite:

  • Qt/KDE development (includes C++ & git).
  • At least basic Telepathy development

Skill level: medium.

Mentor: Someone from KDE Telepathy and someone from KDE Games.


simon is an open source speech recognition system using many parts of the KDE infrastructure. You can find out more about us on our homepage, blog or the development site.

There is also a demonstration video on youtube.

Project: simon workspace integration (Voice control the KDE desktop)

Brief explanation The speech recognition system simon is currently based around the concept of being individually configured for each application that is going to be voice controlled. While this works fine for smaller installations it gets tedious in the many cases where no special configuration is needed and the user "just" wants to browse application menus, select buttons, etc.

The protocol AT-SPI (version 2) exports information about control elements of applications on the DBus interface at a toolkit level (no effort needed from application developers). This system is already used for screen readers like Orca.

simon could parse this information of running programs, generate the necessary vocabulary, grammar and commands to trigger any on-screen buttons, menus, etc.

This would give simon control of running Qt and GTK applications without any specific configuration and greatly enhance the use of simon in a full-fledged desktop.

Expected results A simon plugin that listens to AT-SPI messages on DBus and exports the necessary vocabulary / grammar / commands.

Knowledge Prerequisite C++, Qt, Motivation

Mentor Peter Grasch <[email protected]>

Contact Besides myself personally ([email protected]) you can also contact us at [email protected] and through the forum on our developer page.

Resources The command infrastructure is heavily documented (an older snapshot can be found online).

Project: Context dependent speech recognition

Brief explanation As speech recognition is at its core a statistical analysis, recognition accuracy decreases as more "candidates" need to be considered. Therefore a system that only distinguishes between the numbers from 1-9 will always work better than the same system trying to differentiate between hundreds of words. In short: It's easier to separate "One" and "Nine" than "One", "Con", "Oh", etc. In a speech control scenario only a couple of commands are relevant in each context. simons scenario system could be used and combined with a newly developed context recognition system (probably based on running applications or even the current activity as exposed by KDEs activity system) to deactivate words when they are not needed at the moment.

Expected results Context recognition system and an extended simon / simond protocol to relay context information. The scenario format and UI would also need to be adapted to allow the user to define in which contexts the scenario should become active. The result would be a greatly improved recognition performance in almost any practical use case.

Knowledge Prerequisite C++, Qt, a lot of motivation :)

Mentor Peter Grasch <[email protected]>

Contact Besides myself personally ([email protected]) you can also contact us at [email protected] and through the forum on our developer page.

This page was last edited on 30 March 2013, at 09:21. Content is available under Creative Commons License SA 4.0 unless otherwise noted.