I came across this while reading about GIT. As I understand so far, Git generates a hash for a file and stores the data per that hash. The data is written to a hard drive or maybe a DVD. What controls exactly where that data is written? Does this work within the directory/file structure or is there something different?
Does the hash become the file name?
Edit:
So the contents of the data to be saved is hashed and that value becomes the file name. The essence is that once the file is writ it is never changed. If it were changed, the name would be proof that the contents were changed. A good thing for a version manager.

It's stored in a hidden folder at the root of the git repository's directory tree. The hidden folder is named ".git".

In a Git tutorial, when writing about blobs, it stated:

Ok, I think I follow that. But every file must have a name. If the hash of a file becomes its name, the file still has a name. I don't yet have a Git repository to look at, but, I presume that if I found the right directory there will be a collection of files whose names are the hash functions, meaning the names are fundamentally useless to me if I attempt to use them directly. But obviously Git can access the data with those names. Am I completely off track here?

http://gitready.com/advanced/2009/03...directory.html

They're in .git/objects.

The hashes are in the filenames.

I think I follow. It makes reading about Git a bit more comfortable. Thank you.