On 7/7/15 6:21 PM, Tankred Hase wrote: > Hi, > > Some more context here: > > https://blog.whiteout.io/2015/07/06/standardizing-secure-pgp-private-key-sync/ > > Using a 24 char backup code works fine for Whiteout users. We tested this in > our UX labs and people had no problem writing it down or copy/pasting it into > their password manager. So I can't confirm your hypothesis.
Ok, but the previous statement assume that the only use-case for server-side storage of pgp keys is with/by using password manager in desktop browsers, that's a constraint that should be specified. Use-cases without password manager and/or on mobile browsers that are not integrated with password managers would then lead to a less secure environment because end-users would not likely use a 24 char password. > We discussed using key stretching mechanism other than S2K but Werner Koch > and others were strongly in favor of reusing existing ciphers and mechanism > already available in current OpenPGP implementations. I understand the need to re-use existing encryption code/standard, also already in WebCrypto specs, but we shall then define the boundaries/constraint of use. Maybe, it could make sense to define multiple schema for password-hashing in order to address lower-entropy password typed by end-users each time they use the software implementing this specs? Fabio _______________________________________________ http://openpgpjs.org Subscribe/unsubscribe: http://list.openpgpjs.org
