<<Experts agree that while the IPv4 Internet will eventually grow to its limits, IPv4 will probably not disappear within our lifetimes.>>
The point of my original provocation was to show that IPv6 is really not here yet (thanks Pat for the research). If IPv6 were to be used to get QoS, then the whole path from client-to-server would have to be compatible with IPv6.
To get client-to-server QoS from ATM, again the whole path must be compatible with ATM, but since ATM has been around for awhile, and there is a lot of traffic over ATM right now, and various schemes are being developed for bridging the gap (hopefully short and reliable) between client IP and the core ATM networks, ATM is a better bet for carrying Voice, Video, and Data between paying consumers than IPv6.
Also, the standardization and growing adoption of the G-lite last mile, increases the probability of a total ATM compatible path (i.e., variable cost QoS choices) between consumers. |
