RE: Overlapping ranges

Yves Lafon wrote:
> On Sun, 12 Oct 2008, Stefanos Harhalakis wrote:
> In the same vein, refusing overlapping range request that 
> would be bigger than exporting the whole document (and
> serving the document instead) seems to be a sane
> implementation choice, but the spec shouldn't cover all
> those corner cases, IMHO.

I agree that the specification doesn't need to be ammended to allow the
server to refuse these requests since the server can already refuse them.
However, it should be mentioned in the security considerations.