On Saturday, October 17, 2015 at 7:09:01 AM UTC-4, jean-pierre.muench wrote:
>
>
>
> Am 17.10.2015 um 02:39 schrieb Jeffrey Walton:
>
> Hi Everyone,
>
> There's a new wiki page on RDRAND at https://cryptopp.com/wiki/RDRAND.
>
> The generator was cut-in recently due to some off-list discussions, and it 
> was approved by Wei. The download is not yet available, but it should be in 
> the next couple of days.
>
> Yippie, RDRAND support :)
>

Yeah, at least two others were interested in the class. I'm not sure how I 
missed the demand.

Sorry about that.
 

> Open question: should the class file respond to DiscardBytes? I kinda feel 
> like it should be a nop, but there could be folks who want to call it to 
> increase their comfort level.
>
> The promise of DiscardBytes() is, well, to discard bytes. If this actually 
> makes sense from a security standpoint or efficiency stand point is not at 
> us to judge, or we wouldn't have created this function in the first place. 
> If the user wants to discard 500 bytes, let him do this via DiscardBytes() 
> or he'll just write DiscardBytes() himself. 
>

OK, so what is the conclusion? Yes, DiscardByte should actually consume 
bytes; or no leave it as a nop (and the user can discard the bytes himself)?

(I'm not committed to anything. I just need to know what to do).
 

> Comments or suggestions are welcomed.
>
> My suggestion is to also add RDSEED while we're working at it. 
> We'd then have to add a note that the user shouldn't call RDSEED for large 
> amounts of data but rather to seed PRNGs (along with OSGenerateRandom() ).
>

Lol... I _knew_ that was going to come up. 

>
> I'd volunteer to add RDSEED classes to this patch (and maybe rename it to 
> RDRNGs).
>

This could present a design wrinkle. Check out 
https://en.wikipedia.org/wiki/Comparison_of_hardware_random_number_generators.

In the bigger picture, we might have opportunities for improvement *if* we 
add two or three of these. I see you followed up with the same (re: random 
numbers and TPMs).

I personally have an EntroypKey that I use when I need a good seed 
(http://www.entropykey.co.uk/). 

So, do we:

  (1) just provide them by deriving from RandomNumberGenerator
  (2) add an interface that adds a standard method like Available() to set 
up the contract
  (3) something else...

Note: we already _had_ to provide Available() to help avoid exceptions.

Whatever we decide, it needs to occur quickly :)

Jeff

-- 
-- 
You received this message because you are subscribed to the "Crypto++ Users" 
Google Group.
To unsubscribe, send an email to [email protected].
More information about Crypto++ and this group is available at 
http://www.cryptopp.com.
--- 
You received this message because you are subscribed to the Google Groups 
"Crypto++ Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to