> On Aug 8, 2015, at 4:36 AMPDT, Juliusz Chroboczek > <[email protected]> wrote: > >> What you are suggesting here is heresy. You are saying that the basic >> assumption made 35 years ago, Layering, doesn't work. What a surprise. > > No. What this says is that strict layering has a performance cost, and > a way to recover some of this performance is to organise carefully > controlled leaks between the layers. (You're already doing that -- your > layer 3 routing protocol is probably aware of physical carrier sense.)
Not that I ever heard of. First of all, "carrier sense" was quite intermittent and indicative of nothing more than whether or not you could transmit in the CSMA/CD Ethernet that was dominant until the mid 90's. but that information doesn't carry across layer 2 bridges/switches. After the mid 90's Ethernet repeaters were displaced by switches in layer 2 networks, many of which are fairly large. Carrier sense at the edge of the network where it interfaces with a layer 3 device is meaningless with respect to the layer 2 network as a whole. > > The success of the (reasonably well layered) TCP/IP suite would indicate > that the market has decided that this is a cost well worth paying. Precisely my point, except that it is not true. The datagram service that was provided for with such "success" by TCP/IP does not provide the same service over all physical layers. In fact the now predominant physical layers do not provide sufficiently low-jitter, low loss service for all legacy services to work well. > > -- Juliusz _______________________________________________ homenet mailing list [email protected] https://www.ietf.org/mailman/listinfo/homenet
