Being a research project, there are different people working in different areas of the project. There are four basic areas the project can be divided into:The project is basically divided into the Layout, Pragmatics, and Semantics, but we also have a bunch of demonstrators and even a number of projects useful to end users published as part of OpenKieler:
(Note: this image is also available as a PDF file and as the original SVG file and can freely be used in theses of our students. If you are not a student at our group but want to use this image, feel free to get in touch.)
The layout area provides the foundation for pretty much everything else in KIELER. Much effort in the Pragmatics area has gone into freeing the user from so-called "enabling steps": things a developer has to do to prepare for the model change he actually wants to make. This will usually be things like moving nodes around to make space for other nodes, taking care of edge routing etc. Good layout algorithms are necessary to be able to free the user from tasks like this that don't actually contribute anything to the final model.The core component is KIML, the
. KIML provides the connection between graphical editors on the one hand, and layout algorithms on the other hand. Besides integrating various open source graph layout libraries such as OGDF and Graphviz, KIELER also includes custom layout algorithms in KLay, the component. The most important and most advanced algorithm is KLay Layered, a layer-based layout algorithm targeted primarily at the layout of data flow diagrams. However, KIELER also includes KLay Force, which provides implementations of force-based layout algorithms, as well as a KLay Planar, a planarization-based algorithm. Our custom layout algorithms are implemented in pure Java and can easily be adapted to be used outside of Eclipse as well. There's also the KWebS component, which makes our layout algorithms available as a web service. Finally, GrAna, our graph analysis component, allows us to compute a whole lot of different metrics for diagrams, which is handy when it comes to evaluating the performance of our layout algorithms.Our layout infrastructure and algorithms are available under the umbrella of the Eclipse Foundation as the Eclipse Layout Kernel (ELK) project.
With the term 'pragmatics' (don't confuse it with 'pragmatism') we denote the set of all practical aspects that influences the daily work of modelers in the context of Model-Driven Engineering (MDE). This comprises the creation and modification of models (abstract syntax as in the sense of the MVC) as well as the synthesis of diverse views on those models tailored to different stakeholders and different purposes.
Semantics / SCCharts
We also provide an infrastructure to define execution semantics for a meta model. This can be a simulator based on C (see S/SC projects) or Ptolemy (see KlePto project). Simulators can easily be integrated using the KIELER Execution Manager (KIEM). We also support textual simulations (see Esterel and S project). Another recent main contribution are SCCharts. To get an overview about all projects concerning semantics in KIELER follow the link below.
Demonstrators are editors we use to test and demonstrate the technologies developed in the other three areas.
ThinKCharts is KIELER's SyncCharts editor. It is used to test a lot of the ideas in the Pragmatics area. Through our simulation components, SyncCharts can also be executed, with current states highlighted in the editor.
KAOM, the Kieler Actor-Oriented Modeling component, provides a viewer for data flow diagrams. Its primary use is to provide a realistic application for our layer-based layout algorithm. Through KARMA, the Kieler Advanced Rendering of Model Appearance component, we can make sure the data flow diagrams look very similar to what they look like in other tools. We primarily support the import and display of Ptolemy models.
Finally, the Graphs provides a simple editor for diagrams. It is used during the development of layout algorithms.
Since this is a research project, we cannot provide support for all subprojects forever. Those that are no longer considered in active development and future releases are listed on the discontinued projects page. We have still some archived documentation for these old subprojects: