Semi interesting full circle side note: One of those projects was actually a rudimentary version control system for a report writing app, which used SQLite as the VC data store.
Despite that, the Git command line interface is horribly inconsistent. Commands take separate words, `--options` or one-character flags without a scheme. Lots of synonyms such as staging area, index, or cache all meaning the same thing in the terminology does not help to make learning Git easier.
For new users I would recommend Gitless, which tries to create a better interface to git and to solve the ambiguity in commands. As Gitless is a frontend to libgit2, it works with any git repository and you can also use normal git commands. The downside is that documentation usually only shows how to do stuff with 'git'. If you only learned 'gl', you have no idea how to reproduce it.
Why? There was a time not too long ago where this was true about git. Companies switched to git anyway.
But when you know git and try bzr, it’s mostly the same thing but with different commands. Any gain is small and the pain is instant.