Binary files

Sometimes it’s useful to describe a file as not-a-text-file.

The term for this is a binary file. This is because all that you know about such a file is that it is not encoded as bytes-representing-numbers (with those numbers mapping to character codes). Instead, the 1s and 0s inside could represent anything, and not necessarily be divided on byte boundaries.

Which is to say… it’s just binary data.

Of course, the text file is really binary data too: the difference is that, in the text file, every number starts on a byte boundary and corresponds to a character. There’s no reason why a non-text file will follow that pattern.

Sometimes the first two bytes of a binary file are used to indicate what kind of file it is, but this is just a convention and often not especially helpful to a human viewer.

What happens if you open a binary file with a text editor?

If you open a binary file in a text editor, it will look scrambled and nonsensical. What you are seeing is the computer trying to break the binary 1s and 0s down into chunks of 8 (a byte) and mapping those to ASCII codes — but that won’t work because that’s not what they are.

Because the alphabet/digit/punctuation characters are only about a quarter of all (255) ASCII codes, most of the characters printed in this way won’t look like normal text. This is why you’ll usually see gibberish instead, with question marks or boxes (unprintable characters) and uncommon symbols.

Opening a binary file in a text editor isn’t inherently dangerous; it just won’t make sense. But… do not edit it (that is: do not change it and then save it). You’ll almost certainly corrupt the file if you do that.