openinternet / open-source-assessment Goto Github PK

Per someone's suggestion:

"I’d like to suggest an addition: Testing. A software is only sustainable if it is thoroughly tested. When the tests and the test infrastructure are not conveniently reproducible and self-hostable, it is a very significant roadblock. Red Hat understood that twenty years ago and to this date RHEL is tested with methods and software that have never been released to the public. A competitor would have to re-build the entire test process if they were to compete. They consistently kept their tests a secret and made money out of them. They are not the only one, but they are a good example that demonstrates how vital tests are to the sustainability of a software codebase.

Tests are still perceived as optional and it will probably be years before their importance is acknowledged. Tests are to software what a demonstration is to a theorem. Without them, there is no proof the software actually works."