I have just read the charter and am concerned about two points.
I know I had chance to comment earlier, but didn't notice at that time.

A) I think that the goal of the group the BoF sets up should also be to clarify the use cases that are in scope.
B) I think that deciding a priori to say nothing about the API is wrong.

For the former, there are several functionalities in other standards that should be able to map to this and use it as a common implementation, and there are several issues we have had in OGSA-DAI with practical problems it ought to be able to help with.  We should enumerate those carefully.  Then decide which we will address.

For the latter, as a producer of "upper" M/W I am fed up with recoding because underlying technology changes.  Agreeing APIs might reduce this churn.  At the very least it should finish the clarification of the use case.

I am not at all convinced that naming removes the need for open, though we might make open an implicit operation.

I think the scope needs clarifying in one respect.  I believe DB only thinks of doing pull protocols, i.e. the destination of the data initiates each step in the stream's operation.  I think we are ruling out push protocols, which I believe may be being considered within the context of InfoD, even down to this level?

Malcolm