Max Semenik wrote:
I would wholeheartedly supoport the idea of making tests a mandatory accompaniment to code. One latest example: we made latest security releases without tests.
Actually, we have made security releases that *broke* tests. (ie. tests got outdated)
This is bad not only because bugs not tested for *will* eventually reappear, but also because in my experience the process of writing tests helps to discover more problems with code. If compared with manual tests, writing low-level unit tests allows to look at possible problems more methodicallly, uncovering more problems.
(...)
The problem is that we don't have a proper architecture for testing everything. For instance, I would have liked to add a test for 'stub color in contribution lists' when fixing it. But we have no tests able to check that. Perhaps we could track it with Selenium or storing the html output files, but we would need a more advanced testing fake db.