Comments on RFC3161

[Alfonso De Gregorio <agregorio@xxxxxxx>]

Hi Everyone,

In my opinion, in a future rfc3161 update, the following parts should
be corrected .

Comment A
There is an inconsistency between these statements present in section 1:
 - "It also establishes several security-relevant requirements for TSA
   operation, with regards to processing requests to generate responses"
 - "This standard does not establish overall security requirements for
   TSA operation, just like other PKIX standards do not establish such
   requirements for CA operation"


Comment B
There should be no ambiguity (neither minor) in the scope of TSP. 

In my opinion, it's preferable to say that "The TSA is a TTP that 
creates time-stamp tokens in order to indicate that a datum existed 
_at_ a particular point in time", according to section 2.
And should be avoided to say that "A time-stamping service supports 
assertions of proof that a datum existed _before_ a particular time",
as stated several times in section 1.   
It's part of the verification phase to understand if a token is subsequent 
or antecedent to another one.


Comment C
The part of section 2.4.2:

rfc3161> In such a case, the ordering of time-stamp tokens issued by 
rfc3161> the same TSA or different TSAs is only possible when the 
rfc3161> difference between the genTime of the first time-stamp token 
rfc3161> and the genTime of the second time-stamp token is greater than 
rfc3161> the sum of the accuracies of the genTime for each time-stamp token.

should be replaced with: 
In such a case, the ordering of time-stamp tokens issued by 
the same TSA or different TSAs is only possible when the absolute value
difference between the genTime of the first time-stamp token 
and the genTime of the second time-stamp token is greater than 
the sum of the accuracies of the genTime for each time-stamp token.

In fact, when a TSA clock drift outside the declared accuracy it's
possible to observe time-stamp tokens that have inconsistent genTime
associated with them (eg. the genTime value of a _first_ time-stamp 
token is grater either than the genTime value of a second token, or
the sum of the accuracies of the two tokens); please refer to my 
previous message for an example.

Thanks,
alfonso