2
$\begingroup$

I just sent an email to NIST commenting and enquiring about NIST-SP-800-90Ar1.

The email has 3 parts: 1) option proliferation, 2) deriving seedlen for DRBGs based on hash functions, and 3) inconsistency between notes in test vector and statements in the publication.

I want to generate some discussion outside NIST and I'm considering posting my email here, but I'm not sure if it's appropriate.

Maybe I should break it up into actual questions for the main site?

$\endgroup$

2 Answers 2

Reset to default
6
$\begingroup$

It seems to me that it would be better to post separate specific technical questions here, out of which technical criticism might arise. This doesn't strike me as a good forum for an open letter airing grievances with NIST or petitioning NIST for changes.

$\endgroup$
1
  • 1
    $\begingroup$ ... but feel free to create some discussion somewhere else, such as a mail list or chat room (ask for it in the side channel, our chat if you cannot create one) and advertise it in your meta post here. $\endgroup$
    – Maarten Bodewes Mod
    Oct 17, 2019 at 18:56
-5
$\begingroup$

Yes, if you have three points to raise then ask three questions on the site. Absolutely. We can get into the underpinnings below the mathy stuff that way. You've clearly realised that maths is just a straw man. Neither Barker nor Kelsey inspire confidence either professionally or personally. Their methods seem, unsound. And running to the stats library to find something that 'fits' isn't good science.

However, be warned. This is a pro US/NIST site so criticisms are not welcomed. Posing incisive questions of NIST has proved unfruitful personally. This raises broader issues that are probably better addressed in less constrained forums. Whatever answers you find should be tempered with the NSA warning on page iii, NOBUS policy, the US $B82 intelligence budget, and scope of 44 U.S.C. § 3541. Also look up Price's work on compartmentalisation and the Asch effect to understand why we answer as we do.

You could also ask NIST to correct that spelling mistake on the front page of the 90B implementation. Rolls-Royce is always correct.

$\endgroup$
1
  • 1
    $\begingroup$ This idea that this is a pro NIST site is unwarranted. The site consists of many people, some of which may be pro-NIST, others may be against NIST and most will probably not take such a black-and-white stance at all. Please do not use this site to promote political ideas in questions and answers. Tthe main site and Meta is a Q/A site, and not a forum. Objective reasoning against ideas is of course allowed, and we allow more free speech in comments and especially in chat. $\endgroup$
    – Maarten Bodewes Mod
    Nov 10, 2019 at 15:18

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .