6196Re: [decentralization] decentralized recognizability
- Nov 6, 2002
> I'd still like to know how easy it is to spoof the squiggles. SurelyThat is the open question in this kind of visualization. Before you can
> it's not at all hard to generate a new key pair such that the bezier
> transformed version of the new public key is a barely detectable
> variant of the expected public key ... in which case it's surely not
> unlikely that Alice might say "ah, I recognize that, it's Bob's public
> key, I'll encrypt to that" and promptly encrypt to Mallory's.
adequately judge the security of a visualization technique, you need to
run a lot of empirical tests on people to see what the threshold is for
distinguishing two slightly different renderings. There has been
significant work on this by psychologists in the past, so a good way to
jump start your visualization technique would be to look at existing
research on recognition and memorability thresholds.
- << Previous post in topic Next post in topic >>