Eclipse TEA is a new Eclipse project, just recently approved and incubated. This talk gives a quick overview of what it is all about, and what to expect in the next few months from the project, as well as hands on demo-ing the strengths with real life examples.
Eclipse TEA is a project that has been developed for the past years at SSI Schaefer in Austria. It has been open sourced now as a result of discussions with the community after presenting what we are doing with this technology at the past EclipseCons.
TEA stands for Tasking Engine Advanced, and that is exactly what it is: It provides a easy to use and easy to understand basis for implementing and integrating arbitrary tasks into the Eclipse IDE. These tasks can be grouped to "TaskChain"s and these can be run in different ways: either from the IDE, or headlessly through a dedicated application.
The initial purpose of TEA was build orchestration, and this will be the first thing to be published on top of the new infrastructure. The build orchestration not only allows to integrate more things (like generators that are really applications in the workspace) into a typical build inside the IDE, but also to run the exact same build headlessly from the command line (or jenkins). No ant, maven, you name it. No duplicate configuration. Also the orchestrated build can boost build performance by factors (compared to normal auto-build), which we can prove easily with our commercial product: ~8M LOC in ~700 projects, built in ~3 minutes on decent hardware - including running external generators, etc.
Also worth mentioning is the integration into Eclipse EASE, which allows to integrate IDE extensions using TEA with EASE scripts in both directions (call EASE from TEA, call TEA from EASE), allowing for even more flexibility
This talk will show the basics of the framework itself, as well as giving some real life examples of what is possible with this infrastructure.