Hi, Patrick, Yes, a lot of LIS solutions as well as MPTCP and SCTP can be regarded as session-like solutions.
But, I'd like to see a more firm and precise definition of a 'legal' session layer adopted so that each proposals have to cook their own versions of session-like functionality. If a more rigorous session layer should be introduced, much of the works we see now could be saved and people could concentrate more on more concrete and deeper topics at hand. On Sat, Oct 30, 2010 at 1:22 AM, Patrick Frejborg <[email protected]>wrote: > Hi Dy, > > I think the session layer thing is already happening, at least at the > research level - it seems that separating identifier and locator is > just another wording for adding a session layer to the stack. > RPC is considered to be defined at the session layer - when updating a > mapping database without the end user commanding it to update or when > piggybacking existing transport protocols to update locator, aren't > these typical RPC stylish actions? > > Have a look on all proposals at RRG report, include also HIP - every > proposal have some RPC stylish actions that are executed at a certain > situation. > > It is interesting though, that the lack of a session layer generates > that much stress on the routing system. > > Patrick > > On Fri, Oct 29, 2010 at 3:53 PM, Dae Young KIM <[email protected]> wrote: > > Hi, Patrick, > > > > I'd agree with your session layer argument. The first thing the IETF > should > > acknowledge and have to do is to add the session layer to the current > > Internet architecture. > > > > All other patch works can be stalled before we're done with the session > > layer. > > > > Is there any champion who could bring this issue to the full awareness > and > > acceptance by the IETF? > > > _______________________________________________ > rrg mailing list > [email protected] > http://www.irtf.org/mailman/listinfo/rrg > -- DY
_______________________________________________ rrg mailing list [email protected] http://www.irtf.org/mailman/listinfo/rrg
