On Fri, Oct 1, 2010 at 2:29 PM, David Burstin <[email protected]> wrote: > This article: > > http://www.newscientist.com/article/dn19520-random-numbers-created-out-of-nothing.html > sounds interesting, but being Friday, I was wanting to hear the thoughts of > those who know much more about this kind of stuff than I do. (That means > you, Silky et al)
Yeah, I don't know much about this specifically. All I really know[2] is this implementation of a similar technology: http://www.idquantique.com/true-random-number-generator/products-overview.html But it's interesting to note[1], as with almost any cryptographic system, the biggest issue is implementation-based attacks (i.e. side channel attacks and the recent padding oracle). It's cool though, and I don't know enough about the physics of it to really have any useful opinions. But it's probably interesting, from a cryptography point of view, to wonder what is better: numbers from an unknown source, or numbers from "known" sources with "known" seeds that you can just combine in an unknown way. I'd probably only feel comfortable with it once I knew how it works, so it's something to look into I suppose. Few relevant links: - http://arxiv.org/abs/1006.3512 - http://www.technologyreview.com/blog/arxiv/25355/?nlid=3170 > Cheers > Dave -- silky http://dnoondt.wordpress.com/ "Every morning when I wake up, I experience an exquisite joy — the joy of being this signature." [1] http://www.physorg.com/news202139419.html [2] From my presentation on QKD: http://www.programmingbranch.com/presentations/
