However this distribution is still fragmented and piecemeal. The
preprints distributed in this fashion are mostly unrefereed, and the
archival value is often dubious. The computer system itself that you
have stored your magnum opus on may be gone in a year or less---what
happens to your paper, then? Worse, how do you refer to the wonderful
preprints that you just read? And how reliable were they, anyway,
being unrefereed?