Re: [Gen-art] Gen-ART Telechat Review of draft-ietf-csi-hash-threat-09

[Ana Kukec <[e-mail]>]

Hi all,

Pete, thank you for the comments.

I've changed the draft and took into account all the comments from this 
email. Some comments are below, inline.



Suresh Krishnan wrote:
>
>>  
>>
>> Introduction:
>>    There is a great variaty of hash functions, but only MD5 and SHA-1
>>    are in the wide use, which is also the case for SEND
>> This sentence makes a statement about MD5 and SHA-1 being the only
>> widely
>> used hash functions, but I can't figure out what it is saying about
>> SEND.
>> Is it saying that SEND is widely used?  Or did you mean to say that SEND
>> implementations typically only implement MD5 and SHA-1?
>
> The latter. I propose changing the text to
>
> "There is a great variety of hash functions, but only MD5 and SHA-1
> are widely used. SEND implementations also typically use these two 
> hash algorithms."
>

I've changed the text according to your suggestion Suresh.

>
>
>> But this sentence is just plain
>> incorrect (see below).
>>   Due to
>>    the birthday attack, if the hash function is supplied with a random
>>    input, it returns one of the k equally-likely values, and the number
>>    of operations can be reduced to the number of 1.2*2^(n/2) operations.
>> There is no "birthday attack."  And I think you meant 2^n instead of k.
>> The result you give is due to an equation that is commonly illustrated
>> with
>> a problem known as the "birthday paradox."
>
> Right. A birthday attack is an attack that exploits the mathematics 
> behind the birthday paradox. It is a fairly commonly used term. Would 
> you like me to change something?

That's right -- birthday attack is common term, but only in 
cryptography. I was relying on Bruce Schneier's  book "Applied 
cryptography" where he uses both the term "birthday attack" and the 
equation. Maybe i can make the sentence more clear:

"Due to the birthday attack, if the hash function is supplied with a 
random input, it returns one of the  equally-likely n-bit hash values, 
and the number of operations can be reduced to the number of 1.2*2^(n/2) 
operations."


Other comments are fixed.

Ana