DKIM TTPs (was Re: [ietf-dkim] editorials and nits)

[Douglas Otis]

On Jul 5, 2006, at 6:35 PM, Stephen Farrell wrote:

> Douglas Otis wrote:
>>  From your reference:
>> ---
>> In cryptography, a trusted third party (TTP) is an entity which  
>> facilitates interactions between two parties who both trust the  
>> third party; they use this trust to secure their own interactions.  
>> TTPs are common in cryptographic protocols, for example, a  
>> certificate authority (CA).
>> ---
>> While DNS associates a key with a domain name, there should be no  
>> expectation this domain name represents a tangible entity or  
>> offers meaningful recourse.
>
> Irrelevant.

Setting expectations within the base draft remains important, where  
the current language seems balanced.  General expectations related to  
typical trusted third-party services usually resolve an identifier to  
something tangible.  This is not a property of DNS.  Describing DNS  
as a trusted third-party and equivalent to the usual trusted  
certificate authorities produces inappropriately high expectations.


>> There are thousands of entities involved in these associations,  
>> where the basis is often limited to just the domain name itself.    
>> It is difficult to consider an amalgam of often anonymous entities  
>> a "trusted third party" for "securing" email interactions.  Use of  
>> DNS by DKIM certainly falls short of the expectations of a TTP as  
>> set by Certificate Authorities or the example given of a notary  
>> public.  For DKIM to offer security, a separate assessment of the  
>> DKIM domain name should be made (likely by a TTP).  In that sense  
>> of trust or "securing" interactions, DNS fails this definition of  
>> TTP for email in my view.
>
> Perhaps you haven't been involved in the fairly tedious exercise of  
> engineering one of these, e.g. when I worked for a large German  
> company we made an internal PKI where every building supervisor was  
> part of the TTP (mostly as RAs, as it happened). At the time there  
> were O(400,000) employees, and about O(1) private keys, so, like  
> everyone else, we kind of got used to the idea that a TTP needn't  
> be monolithic.

While a service need not monolithic, the criteria established for the  
RA in your example likely was based upon tangible entities.  If there  
was a problem, there would also be meaningful recourse.  DNS domain  
delegations and email interactions are orthogonal from a trust  
standpoint.  While DNS could be called at TTP with respect to domain  
delegations (to and by often anonymous entities), with respect to  
email interactions, trust is missing.  When DKIM is based upon DNS  
keys, either pre-arranged acceptance (white-listing), or some other  
trusted third-party remains essential for secure email interactions.


> Main point is that there *is* an accepted way to use this term, and  
> yours is not it.
>
> For DKIM, the DNS is a TTP.

The use of this TTP term with respect to DKIM and DNS creates false  
expectations in regard to what DKIM signature verification might  
imply.  While technically this term could be applied, it remains  
misleading.

-Doug