The detailed discussion of the issues involved in establishing trust relationships between entities on the Grid is well 
worth publishing as an informational document. It points out at a detailed level the problems with the existing 
mechanisms which are limited to authorities stating their obligations, usually in long confusing documents, but no 
corresponding type of documentation for the subscribers or relying parties to acknowledge their responsibilities. 
However, I do think the reason for including the QIK model should be mentioned in the abstract and introduction. I'm not
 sure if QIK is a solution that is specified and/or being implemented within the WS-trust area, or if it is merely 
intended to be a model for a possible solution.  If the authors would like to encourage further work on the model in 
order to produce a GGF standard, they should make this clear.

There were a few small points that I found confusing while reading the document:

In the line on page 3 that states "According to this definition, the essence of trust lies in the disappointment of a 
trusting parties expectation�

Do you mean mistrust? (or not disappointing)

In paragraph 2.4.1 it might be useful to identify the second bullet as the policy authority and the third bullet as the 
certificate authority.

In Figure 6, the numbers on the figure don't quite match the numbers in the itemized steps.