On Tue, Apr 3, 2012 at 5:16 PM, Dale Tronrud <det...@uoxray.uoregon.edu> wrote: > I'm not sure how encryption can solve a problem of "truth or falsity".
AFAIU any given checksum will tell you if a file is corrupted or not. My brain decided to interpret that as true or false. and .... > A person can use their private key to encrypt a lie as well as the truth. > [...] I don't quite follow your prescription, ...I admitted it is a mess - and sorry to mix up the various algorithms. also I must emphasize I do not have a clear picture of how encryption would work here. can I step back - it *seems* that following facts point to a checksum of sorts for a *pdb entry: * random number generator seed * randomly chosen Free R set * integer indices of the Free R set * detector things - serial number, or fingerprint of sorts - known to *pdb only. ... by "checksum of sorts for a *pdb entry", what that means is an easy way to verify if all parts of the entry originated with diffraction images. "detector things" indicates that I am wondering if something besides an SN on a detector would be useful. ... so a scenario that comes to mind is the deposition team runs the checksum (or whatever), and gets the Free R set (for example). they run the battery of tests. they find that refinement is a disaster. they go check the detector specs they have, etc., etc., there were no images used. > The beamline could encrypt all images with their private key, and[...] it > could be > proved that those images came, unmodified, from that beamline. would encryption of images significantly increase the integration time? Also, I am not following the image deposition forum elsewhere. ... anyways, this sounds like it was just an excercise. Thanks anyway. Regards, -Bryan
