Vint Cerf on why TCP/IP was so long in coming | Vint Cerf responds to our discussion on 25 years of TCP/IP | Wide Area Networking Alert Newsletter By Steve Taylor and Jim Metzler, Network World, 01/29/08 In the past couple of newsletters, we've been celebrating the 25th anniversary of TCP/IP, and, as such, we've been taking a look back at the path that TCP/IP has taken over the past quarter century. In response to the first newsletter in the series, Vint Cerf pointed out that there was a long development cycle for both TCP/IP and for X.25.
Vint wrote: “Keep in mind that TCP/IP development began in 1973. X.25 started almost concurrently. Larry Roberts was running Telenet at the time and when he asked me what protocol to use for his service I recommended TCP/IP, but he rejected this on the grounds that he could not sell a datagram service. He thought he could only sell ‘virtual circuits’ to replace real circuits at a lower price. (Based on statistical multiplexing, you could offer the same burst rate service at a lower price because the statistical sharing allows more users on the same capacity.) Ultimately, we just ran TCP/IP over X.25, ATM, Frame Relay, MPLS, PON, etc etc. That was the whole point of the design of TCP/IP.”
Thanks to Vint, first of all, for pointing out that there was a long development cycle for both TCP/IP and for X.25. (...) Vint also brings up an excellent point in terms of marketability vs. technology. Over the years, we can come up with many examples both of where the best technology did (or did not) win and of how marketing has defined a service. For example, many of the “best” features of frame relay, such as the ability to use Switched Virtual Circuits (SVC) in addition to Permanent Virtual Circuits (PVC) were never widely marketed because the pricing was too complex. Rather, the PVC was a simple replacement for a leased line at a fraction of the cost with better performance. This pricing model, for which we’ll give primary credit to Christine Heckart, was carried forward to ATM, and helped define roughly 10 years of “state-of-the-art” networking.