Finn Brunton, U-M Assistant Professor of Information, will present "The Accidental Archive." Brunton's talk, with the alternative title "Researching criminals, otherkin, cipher anarchists, spammers and the online history that doesn't want to be kept," kicks off a new series of monthly events addressing emergent research. Light refreshments will be served.
Notes to follow. There's a hashtag, #MLibRes, on which a half dozen people are tracking. This is an impressionistic transcript, not a summary or tidy narrative, more notes drawn from what I remember from the time and some of my own thoughts mixed in.
"Preservation is constant toil, entropy requires no maintenance."
Links are not permanent, and the scholar who depends on them persisting risks losing their research materials. Finn gave out bits of a manifesto, one of which was "embrace wget".
The Law of the Good Neighbor, a theory of library collection shelving from Aby Warberg. An essay of the same name by Michael Steinberg.
A discussion of Rodona Garst, a notorious spammer, and the ephemeral collections of data about her operations collected under the name Behind Enemy Lines. One such collection is on archive.org so it has some hope of permanence. (but if you really want a copy, make one now)
Deep Blue is an institutional repository run by the University of Michigan. One notable project archived there is the Michigan Terminal System, which has original distribution tapes saved into non-ephemeral bits. Finn has been collecting tarballs of references for the works he's done, and one idea (new to him and recommended by the audience) was to put his tarballs in Deep Blue, with the right kind of dark-archive permanence that was suitable to the contents.
Reference to "shearing layers" from Stewart Brand's "How buildings learn". Different parts of the system change at different speeds. From Quiet Babylon, a discussion of shearing layers looks interesting.
Heresies expounded: "we should save everything", primarily because the Internet is all significant but to different populations. The example given is the Enron corpus, which is also a very useful spam corpus; if you discarded the spam you'd have been left with something less useful. Saving everything is problematic to archivists who are used to assessing collections and picking the good parts and compiling a finding aid, rather than taking all of the unfiled materials and collecting them willy-nilly.
There's more, but this captures the flavor of the event well; there will be a recording released.
Future events are Fourth Mondays. Follow @UMLibRes for details.