These last weeks there has been an interesting brainstorming in gtk-devel about developing unit tests for Gtk+. A lot of people joined and give insight on this topic, it is great seeing so many people interested in this stuff. Now it is time to reach conclusions and there are some hot topics that still need to be decided:
- Will we use a testing framework like Check or finally develop something specific for Gtk+?
- Should we stop the test suite after the first failed test or should we get a complete list of failed/passed tests?
- Should we fail a test whenever we reach a critical/warning?
- Should we test segmentation faults by forking every single test?
- and more … π
If you are interested too, check it out and help us make the best choices π