Why does git store file contents as a blob?

Question

This question may seem trivial to answer, but I am struggling to come up with definitive advantages of storing file contents as blobs and not just its original format (e.g. a text file).

Typically, blobs are used in lieu of other storage formats for media: images, videos, audio, etc. git, at least how I see it commonly used, typically tracks revisions to text files and not multimedia.

To summarize formally: what are the advantages of storing file contents as a blob (converts to binary data) rather than the original format of the revision (e.g. leave it as text)?

Answer 1

“Blob” just means a sequence of bytes. A blob in Git will contain the same exact data as a file, it’s just that a blob is stored in the Git object database, and a file is stored on the filesystem.

So there is no difference in the format, the only difference is how they are stored.

For example, if you add an image hello.jpg to your repository, and then commit it, you will have two copies of the same data:

• You will have a file on disk, named hello.jpg, which contains the JPEG data,

• You will have a blob in your Git object database, named with the hash of its contents, which contains the exact same JPEG data in the same format.

The database can use some fancy tricks to store data efficiently, including compression and using deltas, but in the end it is still storing the exact same data that was in the original file.

A text file is no different. “Text” is just a particular type of data that you can store in a binary file.