Quoting Bernard Rosset via Dng (dng@lists.dyne.org): > This is what the last line of the abstract claims; however the whole > paper goes beyond my understanding. > > https://eprint.iacr.org/2021/232.pdf
Snakes. Oil. (**COUGH** Theranos **COUGH**) We've been here before with Crown Sterling. https://www.schneier.com/blog/archives/2019/09/crown_sterling_.html https://www.schneier.com/blog/archives/2019/09/the_doghouse_cr_1.html About the video where Crown Sterling CEO Grobert Grant ran a cooked demonstration of Grant's claimed crypto-cracking algorithm (the stuff talked about in the paper): Ars shared the video with Jake Williams, the founder of Rendition Infosec and a former member of the National Security Agency's Tailored Access Operations group. "I'm dumber for having watched that," Williams said. "Bragging that you can factor a 256 bit RSA key in 2019 is like bragging about hacking an unpatched Windows 2000 box. Sure you did it, but nobody should care." The 256-bit key, Williams said, was "absurdly small." (Digital certificates from recognized certificate authorities have used RSA 2048-bit keys for more than seven years.) https://arstechnica.com/information-technology/2019/09/medicine-show-crown-sterling-demos-256-bit-rsa-key-cracking-at-private-event/ > Any way, pushing for ECDSA or even EdDSA, both of which are more and > more supported out there (and have been for a almost a decade > already), is IMHO the most future-proof take. Maybe. A counter-intuitive aspect of crypto is that older (if still not significantly flawed) algorithms and their implementations are often preferable than newer and theoretically more-promising ones -- because the former have withstood determined and expert attacks for longer. Schneier made this point a few years ago about why he felt that Blowfish is still safer than Twofish, even though he felt the latter is technically superior -- because it was too new to be nearly as battle-tested. (And, again, don't forget that weaknesses in the implementation matter as much or more than theoretical weaknesses in the algorithms themselves. Cracking is cracking, no matter whether it was achieved through exploiting unintended side-channels or anything else.) -- Cheers, "My generals are always right about other people's Rick Moen wars and wrong about our own." -- LBJ r...@linuxmafia.com McQ! (4x80) _______________________________________________ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
