On Sat, Dec 30, 2006 at 07:29:16AM -0500, Dan Collins wrote: > A very interesting and unique idea, though I can't believe that the > change due to a little heat would be detectable? Have a look at the graphs in the paper: http://www.cl.cam.ac.uk/~sjm217/papers/ccs06hotornot.pdf They were collected from tests on real hardware where I put in a temperature probe. Some are from local connections and some transatlantic, and in both cases the quality of the results were good. Without that much analysis, differences of 1--2 C were detectable. The source code is online, so you can also try it yourself. To get accurate results, it does take quite a while. My tests where typically 8--12 hours: http://www.cl.cam.ac.uk/~sjm217/projects/anon/#tempcov Thanks, Steven. -- w: http://www.cl.cam.ac.uk/users/sjm217/
Attachment:
pgpsnoChwCURG.pgp
Description: PGP signature