>> i can't get my head around this qos stuff, at least as it applies to
>> backbone or isp networks. our customers pay us not to drop packets.
>> so adding complexity in order to choose which packets to drop seems a
>> bit strange.
> I think "End-to-end Arguments in System Design" addresses this issue
> very well. I don't think any of us can get our head around on this QoS
> stuff. I have been "eating my brain" to get my PhD to find a solution
> for the complexity. I believe any approach to tackle the dilemma will
> satisfy as long as we feel happy about the choice we make ranging from
> zero to infinite complexity to satisfy QoS demanding applications
> because Quality x Efficiency = Choice :) QxE=C, where C is a
> constant. In economy, this fact is known as "demand-supply" theory. I
> mean the choice also depends on the economy :)
let me see if i understand
o people are paying me to carry their packets today
o they expect me not to drop any packets
o and there is an untapped market out there of folk who want to pay
me more money to actually drop their packets?
let me guess. they gonna be picky about which of their packets i drop,
so picky that it's gonna cost me a lot to drop just those? and to do it
'right', my network will be less reliable for all customers, the vast
majority of which i presume will still want me to drop no packets?
if they will pay me X/packet-dropped, how about if i just refund them
X/2/packet-dropped if they'll drop the packets themselves.
randy
This archive was generated by hypermail 2b29 : Mon Aug 04 2003 - 04:10:03 EDT