Re: Generating Prepaid Card Numbers
Available news archives: comp.lang.tcl - comp.lang.python - comp.security.firewalls - sci.crypt - comp.lang.php - comp.lang.javascript
Google
 
Web news.hping.org


sci.crypt archive

Re: Generating Prepaid Card Numbers

From: Andrew Swallow <am.swallow@btopenworld.com>
Date: Tue Dec 20 2005 - 19:37:55 CET

Michael Fork wrote:
>>Assuming you have 14 digits available to define the sample space, you
>>should be able to generate a billion card numbers with less than 1 in
>>10000 chance of a collision. The suggestion of using AES over the
>>activation numbers is OK assuming you can keep the key you use a
>>secret. I don't think you need CTR mode.
>
>
> Is there an advantage to using AES over a hardware based RNG, which has
> no key to keep secret?
>

There are two advantages in using AES.
1. You have to buy and use the hardware RNG.
2. The second box can be heavy to lug around hotel rooms when working
away from the office.

These disadvantages with a hardware RNG may not be a problem to you.

Andrew Swallow
Received on Fri Dec 23 20:11:17 2005