News for hackers. Stuff that matters.
Re:Multihash – self-describing hashes by advisedwang on Sunday February 26, 2017 @02:54PM (#13739410)
Looks awesome, so long as its easy to ensure attackers can't quietly make you use a weak hash. Some JWT implementations quietly allowed the "none" algorithm which allowed an attacker to forge creds [1]. Any similar "pluggable" system needs to learn fro this and be careful to avoid letting attackers pick weak algorithms.

[1] https://auth0.com/blog/critical-vulnerabilities-in-json-web-...

Re:Multihash – self-describing hashes by _prometheus on Sunday February 26, 2017 @03:05PM (#13739481)
That's right. It's really important to make sure there is restrictions on what hashes to use if your system is receiving hashes and only checking them for self-consistency.

Particularly relevant is "Crypto Extensibility" (formats and protocols to be able to extend a protocol), vs "Crypto Agility" (the use of Crypto Extensibility to use simultaneously a large variety of algorithms, with the key feature that one can be downgraded to an old/possibly broken hash.

AGL describes it well here: https://www.imperialviolet.org/2016/05/16/agility.html

---

I've filed https://github.com/multiformats/multihash/issues/70 to track this.