3 ms·
> Most importantly: we've moved away as a field not only from FFDH itself, but also from asymmetric systems with opaque parameters. For another example of asym
by weinzierl 9y ago
> Most importantly: we've moved away as a field not only from FFDH itself, but also from asymmetric systems with opaque parameters.
For another example of asymmetric systems with opaque parameters: The BADA55 paper [1] and the related slides [2] shine a light on the parameters of elliptic curves.
[1] https://bada55.cr.yp.to/bada55-20150927.pdf https://bada55.cr.yp.to/bada55-20150927.pdf
[2] https://cr.yp.to/talks/2014.10.08/slides-djb-20141008-4x3.pdf https://cr.yp.to/talks/2014.10.08/slides-djb-20141008-4x3.pd...
- tptacek 9y agoBADA55 sort of has to be understood in context. There's an academic rivalry between Bernstein and the team behind the Brainpool curves. Both teams reject fully opaque parameters, for the same reason the authors of this SNFS paper do. But they use different approaches to mitigate the risk. The Brainpool approach is to start parameter generation from mathematical constants. We assume that the ultimate creator of the universe didn't embed an NSA backdoor in the digits of pi, for instance, so basing a parameter on some permutation of pi is unlikely to result in compromise. Bernstein's approach is to optimize parameters for size, performance, and security, and to provide a detailed systems rationale for those parameters. The Curve25519 paper is a good example of what that approach looks like. Bernstein's argument in BADA55 is that the Brainpool approach has a weakness, which is that there are so many different mathematical constants to work from and so many different permutations of them that a backdoor designer still has enough flexibility to find a weakened parameter. Bernstein's argument is also interesting scientifically. But unlike the SNFS paper, it's not so interesting from a Kremlinology standpoint: it is, let's face it, really unlike that there's a backdoor expressed in the permutation of universal mathematical constants used in any cryptographic standard. (I'm doing this from memory and may be missing an aspect of that paper).
- JoachimSchipper 9y ago> it is, let's face it, really unlike that there's a backdoor expressed in the permutation of universal mathematical constants used in any cryptographic standard. Why? I mean, there's plenty of reasons to be skeptical of backdoors in standards in the first place, but why do you consider it especially unlikely for any hypothetical backdoor to hide in the details of "nothing up my sleeve" numbers?
- goalieca 9y agoNot a mathematician but have a background in engineering. Some damn constants are everywhere and have too many useful properties. Take "e" for instance. I'll be damned if that isn't a terribly versatile number.