From Tony Dicenzo:

1. In paragraph 2, the definition is further complicated by the possibility of the administrative domaings being 
heterogeneous. That raises the issue/requirement of interoperability.


2.In paragraph 3....I think this unintentionally suggests Seti-at-home was the first type of grid. It wasn't of course. 
I suggest reversing the examples, and show Seti as a later phenomenon. Cycle stealing came well after the early grids.


3. I see Qual of Service as a characteristic, but Idon't think Highest is a common requirement. Lots of academic use 
that doesn't demand highest.


4. That list of characteristics does not seem sufficient to me to define grid. i'm not sure what is missing, but test it
 yourself....what is the minimum set of characteristics to define 'grid'? 

5. I think Collaboration grids is '...OSF's long term vision'. But not necessarily that of Commercial Enterprises.


6. Every IT center is a Data Center. I do not undestand how they might be different that Collab Grids. I expect a case 
can be made there are Data Grid. Certainly SNIA things so.

By data center grids, do you mean grids under the control of a single Enterprise? I see you go to lengths to explain 
that might be misleading to say they are under the control of one administration (which I agree with).


7. Not clear what a Cluster Grid is. In fact, I think this is redundant with Collab grids. The definition proposed could
 apply to any grid.



Tony