I don't feel like Git is "the one true way" to DVCS though. For example using SHA1 as hashing algorithm doesn't seem like a good idea. Or calculating "packfiles

I don't feel like Git is "the one true way" to DVCS though. For example using SHA1 as hashing algorithm doesn't seem like a good idea. Or calculating "packfiles" ad-hoc on clone operations, that's wasteful and makes cloning big repos on unstable connections a PITA. But it's certainly true that Git is the "current best practice" in the DVCS landscape.