Types of submissions to encourage

[narock]

One of the ways a preprint service can help disrupt and refactor academic publishing is by including research results that are important to science but which are left under the bus by current publishers: null results, confirmation studies, data, methods. (proposed by Bruce Caron)

[semantic-nomad]

As different type of "type", it would be great to encourage submissions that aren't PDFs.

For example, if I author a "paper" and publish it to the web on my own personal site I should be able to "submit" it to EarthArXiv and used the issued DOI to send potential readers to my personal site (where I have, hopefully, used RASH, RDFa, etc. to markup the entire text).