[Gnuk-users] Security of NeuG?

[NIIBE Yutaka]

These kind of considerations from me is not scientific at all, but
somehow religious.  Please just ignore if you don't like.

Since it's free software, you can modify it as you like.  It's
possible to add device side encryption for NeuG standalone device.  If
you want to do that, please go ahead.  Even I will be able to include
your enhancement to the NeuG distribution.

Just, I won't implement that by myself, as I believe I understand the
philosophy of the gambler.

On 02/22/2015 09:34 AM, Daniel Kahn Gillmor wrote:
> Using a symmetric-key cipher that provides integrity protection as well
> (e.g. any of the common AEAD modes) would provide not only
> confidentiality but also assurance that the stream hasn't been tampered
> with.

Good point.  I understand someone wants this feature.

Then, next question would be:

    Is it really enough when you get the exact random bytes from the device?

I feel it's not enough.

Even if it's only the encrypted random bytes which are tapped, I won't
use the bytes sequence (decrypted) for my system.


Here, I write something, it would be red herring for you, but it's
important for me.


Engineers in Japanese Industry have tendency not to purchase.  Or they
have a practice of saving cost by minimum procurement.  Even if it's
deadly needed, or rather fair to pay.

In this situation, some would invent the structure like:

                                  V--[SECRET 0]
    [TRNG] --+--><integrity check + conditioning>-->  [the machine]
             |
             |                    V--[SECRET 1]
             +--><integrity check + conditioning>-->  [another machine]
             :
             :
             |                    V--[SECRET N]
             \--><integrity check + conditioning>-->  [the other machine]


... to save money of their employer.


In my interpretation of Chuji (the gambler) 's view, this structure
avoids a game completely (except on the TRNG device).  It avoids the
important concept of "try one's luck".

"Chicken!", I'd say.

I know the risk that once this message will be available on the web
(by the mailman), this could be interpreted as my curse against
Japanese Industry.  But since I'm not a chicken, I send this message
unencrypted.  Rather, I am brave enough to show my faith.  So, this
message will be signed.

By this message, I'd like to show you how I use technology.
Computation-wise, you know, public key encryption and digital signing
share same technology.

Give me best luck, Chuji (or NeuG)!
-- 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: OpenPGP digital signature
URL: <http://lists.alioth.debian.org/pipermail/gnuk-users/attachments/20150223/ce020302/attachment.sig>