On Jan 27, 2010, at 11:13 AM, Adam Heath wrote: > Hans Bakker wrote: >> Scott, >> >> If you want people to co-operate with you, you have to avoid doing the >> following actions: >> >> 1. revert a commit without discussion with the original author. > > Actually, this is a perfectly good response. If some commit breaks, > and it's not immediately clear what/why it is broken, and the original > author is not immediately available, and you can definately verify > that reverting just one commit makes things work again, then reverting > and fixing later is the better approach. > > You seem to be thinking that reverting a commit is an attack against > you; again, as I have said, don't take things personal. We all want > the code to be the best it can be, and if someone discovers a bug with > an implementation, that is a *good* thing.
I agree with this. Breaking existing stuff is a big no-no for various reasons (technical, professional, and karmic). Sometimes people do it by mistake, and then they shouldn't mind the revert. If people broke something on purpose, then the revert will hopefully lead to some lively, and well needed discussion about the should-have-been-proposed change. -David
