[Gnuk-users] Testing NeuG

[NIIBE Yutaka]

Hello,

I'm preparing new release of NeuG standalone device.  I don't change
any implementation of the RNG itself, but there are some changes for
USB driver, thread library, and (optional) Fraucheky.

Besides, thanks to the improvement of GCC, the computation of SHA256
is faster.

Today, I'd like to ask your comments about RNG Testing.

I wrote in NeuG's README:

    Test results
    ============

    See files under the directory test-results, for test result of
    "rngtest" in rng-tools, NIST STS 2.1.1, and Dieharder.

    I collect 110 files of 125MB (= 13750MB), and use scripts to invoke
    dieharder and rngtest.  Collecting 110 files, it took three days.

    For NIST STS 2.1.1, I used only 10 files of size 125MB.

Now, I don't think those tests make much sense.  I keep doing that for
every release (to find regression), as I did that for the first
release (when I didn't have enough knowledge).

I think that testing by TestU01 [0] or PractRand [1] is better.

I contacted the person of TestU01, and I knew that it's now under
GPLv3+, but the copyright notice in the source code was not changed
yet [2].  Thus, I haven't package it by myself.

For PractRand, I found that it includes binary in its distribution
when it was 0.90.  And I waited for better source code distribution.
Now, it's 0.92, but no change for my concern.  So, I hesitate to
package PractRand.  (I only want the randomness test part of it.)

Besides, modern test suites like TestU01 and PractRand tend to require
modern computer, but I don't have a powerful computer these days.

Perhaps, I will just do same test for 1.0.1 as 1.0.  And I will
postpone doing better test.


[0] TestU01:
http://simul.iro.umontreal.ca/testu01/tu01.html

[1] PractRand:
http://pracrand.sourceforge.net/

[2] Debian wnpp: RFP TestU01
https://bugs.debian.org/596005

[3] Randomness test tools (in Japanese):
http://no-passwd.net/fst-01-neug-handbook/randomness-tests.html
--