As a User, I need all examples to always run out of the box.
As a Developer, I would like this to entail is little work to build, maintain, and extend.
Thus, it would be great to prepare functionality that minimizes the work related to examples with automated tests.
Suggestions:
[ ] Define what likely needs to be tested in all examples
[ ] Define where that responsibility lies (gui, or pytrnsys)
[ ] Duplication is acceptable, if it minimizes work for synchronization, AND increases feedback, AND increases robustness
[ ] Should the example be reproducible from scratch (e.g. GUI tutorial)?
[ ] At the very least, the "run simulation" button should run immediately when downloading the new GUI.exe.
[ ] Next, the "process data" button should work on the produced results out of the box.
[ ] The hydraulic export should reproduce the file in the example.
[ ] The examples should have a run.config and process.config.
[ ] Overwriting the run.config from the GUI is problematic. It should thus be possible to choose which one to run.
[ ] Examples of parameters studies should be as simple as possible to minimize computational cost.
[ ] Define a maximum time that examples should be allowed to run (e.g. 2 min. per simulation or 2 min. per example).
[ ] Is it possible to test without touching the file system (have trnsys write to virtual disk)
[ ] Debugging examples requires high quality feedback (e.g. log error shown in CI)
Test specific:
[ ] open json and show diagram <- export pdf and compare?
[ ] massflow solver AND visualization <- export pdf and compare?
[ ] export hydraulic <- compare files
[ ] export hydraulic_control <- do not overwrite example control, compare files
[ ] update run.config <- rename, do not overwrite, compare files
As a User, I need all examples to always run out of the box. As a Developer, I would like this to entail is little work to build, maintain, and extend. Thus, it would be great to prepare functionality that minimizes the work related to examples with automated tests.
Suggestions:
Test specific: