Mediatype

Mediatypes are the classification of type of items on Internet Archive to indicate how the system should present and produce derivations from them. Created and added to in an ad-hoc fashion over the course of decades, these are the currently relevant mediatypes: texts, audio (and etree), movies, software, data, image, and web.

The identifiers with the same name as the mediatypes appear to have magic that makes them act like virtual collections containing all the items of that mediatype. So texts contains all the texts, image contains all the images, etc.

Texts

Presents printed materials or compilations of images. Usually is rendered into the book reading software, as well as receiving OCR (optical character recognition) and alternate formats such as PDF, DJVU and EPUB. Most affiliated with the Book Scanning Operation.

Audio (and etree)

Music, sound, or any number of audio files. Originally created.

There is also a specialized etree mediatype affiliated with audio, that allows for specific treatment of those files.

Account

Accounts for the patrons of the Archive are also considered items, and are given an "account" mediatype. It is currently only possible to create an account through the account creation system. All accounts on the archive start with a @ character, e.g. https://archive.org/details/@username.

It is possible to do a bunch of weird and horrible things to your own account item. Please don't.

Other Considerations

There are occasionally other mediatypes in the mediatype metadata field, due to errors, old tests, and other factors, but they are not specifically valid. They are often trapped inside dark or disused items.