Hi,
Quoting Mark Taylor <m.b.taylor-at-bristol.ac.uk>:
> I think you may have misapprehended the point of this feature.
> Paging is not mainly intended as a convenience for users so that
> they get relevant records without looking at all the data, but as
> a protection for software, both at the server and at the client ends.
> Without such a mechanism a client might in principle be delivered
> a multi-megabyte list of nodes which could put strains on its
> resource usage. Similarly, a server might have to stage such a
> response prior to sending it. In fact in many or most situations
> in which this facility is used, the expectation will be that the client
> does retrieve all the nodes in the list rather than just one or a few
> pages, but this can be effected with both ends of the connection able to
> budget their storage in a controlled way during the transfer.
Nonetheless, a filter capability to restrict the response in first place and - if bandwidth is the rationale for paging - then a store-specific MaxRecords-limit can take care of it.
-Markus Received on 2006-08-15Z18:48:19