On 12.2.2015 15:24, Arvin Schnell wrote:
Code review, as we understand it now, is here to establish a process that should keep bugs out of our code and to check that the code conforms to given standards (tests, variable names, methods length, ...).
I don't see how the code review done here keep bugs out. Maybe they keep some bugs out. But to generally keep bugs out would require the reviewer to know and understand the existing code which is often not the case.
OK, so what do you suggest to make it even better? Please, consider the current team setup, knowledge and tasks. I'm always open to good ideas.
Fact is, you're the author of that change and you're responsible. So others may comment but not change.
That's on the other hand something else. Ownership and responsibility should not be limited to the code that a developer has created. I myself, for instance, feel the responsibility for all the code that my team develops and maintains, although I have implemented just some parts here and there.
From my understanding the persons who have to fix bugs and implement features have the ownership and responsibility. Often that is just one person.
Yes, and I'm constantly pushing for changing this as one-guy-for-one-module doesn't scale. For instance, see how Ancor has been working for several months now. See where Christopher is heading to. Check out what Ladislav and Josef have been doing in the Travis/RuboCop area. E.g., I've asked for a new Libstorage developer position and I've got it. I'm currently in the process of hiring new developer. The same for AutoYast. My request for Linuxrc position hasn't been approved yet. Anyway, these new people *will not* work only on libstorage/autoyast/linuxrc. I want them take the responsibility for the code that we own as a team. -- Lukas Ocilka, Systems Management (Yast) Team Leader SLE Department, SUSE Linux -- To unsubscribe, e-mail: yast-devel+unsubscribe@opensuse.org To contact the owner, e-mail: yast-devel+owner@opensuse.org