* TCP/IP, technology evolution, and telcos
In finishing our commentary – at least for the time being – on TCP/IP’s 25th birthday, there are a few remaining points concerning TCP/IP clones that didn’t do so well. The most impressive technical/political part of the evolution came with the telco’s proposal for Switched Multimegabit Data Service (SMDS) in the early 1990s. At that time, BellCore, which was the technical services and research unit that served the former RBOCs, was proposing this data service as a next step, and there were rather fierce debates among the proponents of SMDS, ATM, and frame relay as to which was “best.”
One of the biggest differences between SMDS and the other two was that SMDS was essentially connectionless, while the other two were connection oriented. Of course, cynical enterprise data folks were not at all surprised that the telcos would promote a service that used more bandwidth.
Within a relatively few years, SMDS lost out in the market. Part of the reason, of course, was that the enterprise would, at that time, be likely to reject anything that was proposed by the telcos. Just part of the nature of the beast.
But in retrospect, there are some interesting thoughts. First, SMDS probably was indeed a technically superior service. Second, if you strip away the labels and look at the underlying technology, SMDS looks very, very similar to TCP/IP. It just didn’t have the right marketing spin.
So we’re left to wonder… Where would we be today had the telcos proposed a TCP/IP-based service instead of SMDS in the early 1990s? Hmm….
