There's a certain something to papers whose titles are simple statements, like Bulk Silicon is Susceptible to Fatigue or Bounded Quantification is Undecidable.

This reminds me of the guy who set up a special stand with a fan to rapidly dry his Gillette Mach 3 razor. His claim was that the degradation of the razor's performance over time was due to corrosion of the blade edges, not dulling from cutting hair, and that if the blades were air-dried after use, he could get a year out of a single cartridge.

I'm just amused by the researcher from the University of Extremadura. It made we wonder if April Fool's Day had been suddenly relocated. :)

Back when I was in the streaming media business, we rehabbed a few old Ampex Quad decks and made a little bit of money remastering deteriorating Hee Haw and Roy Acuff shows onto digital Beta tape. We tried selling archival services in our StorageTek robot for a few years, but the guy who was funding the whole thing ran out of cash before he got his remote backup site built.

Ultimately, reliable digital archival is a continual process of copying from one evanescent medium to another, different in degree, not in kind, from a monastery full of scribes. The reliability of the archive lies in social commitment to the copying process, not in the illusory permanence of the media.