Over the last six years, I've become better at using Git for version control. But my conceptions of the index, the working copy, the object graph and remotes have just grown fuzzier.
Sometimes, I can only understand something by implementing it. So, I wrote Gitlet, my own version of Git. I pored over tutorials. I read articles about internals. I tried to understand how API commands work by reading the docs, then gave up and ran hundreds of experiments on repositories and rummaged throught the .git directory to figure out the results.
I discovered that, if approached from the inside out, Git is easy to understand. It is the product of simple ideas that, when combined, produce something very deep and beautiful.
If you would like to understand what happens when you run the basic Git commands, you can read Git in six hundred words.
Afterwards, read the heavily annotated Gitlet source. 1000 lines of code sounds intimidating. But it's OK. The annotations explain both Git and the code in great detail. The code mirrors the terminology of the Git command line interface, so it should be approachable. And the implementation of the main Git commands is just 350 lines.
I wrote an article, Introducing Gitlet, about the process of writing the code.