No, that would certainly not always be the case. In this case, it’s used for the tests that treat Bitcoin Core as a black box, especially the functional tests, and benchmarks that call RPC commands of the binary and expect specific outcomes. It’s just simpler to use a scripting language for these tests.

A lot of coding around testing... in python, which is interesting choice given the code base is C++

Is this always the case for large C++ projects?

jp305

No. it would be a good next step.

Additionally, i am debating whether it make sense to break commits into components like wallet, consensus, networking, GUI and others. it would add insights on the kind of work that is being done. What could those components be? and, then I need to figure out a way to group it from the commit log.