Understanding the evolution of the two “dominant design” LAN protocols of the 1970’s and 1980’s, TCP/IP and OSI, dictated the rationale for whom I interviewed and the subsequent reconstruction of the early history of computer communications. There were other protocols that played important roles, particularly the protocols of Xerox that influenced the evolution of DARPA’s TCP/IP. (I discuss the Xerox Network Systems (XNS) suite of protocols and its predecessor PARC Universal Packet (PUP, or Pup) protocols with Yogen Dalal in our interview. DECnet was innovated within DEC, and first released in 1975. Herein I discuss is innovation with its author Stuart Wecker. The integration of DECnet into DEC’s VAX line of computers would transform DEC into a networking market leader in the 1980’s. IBM released its Systems Network Architecture (SNA) in the early 1970’s, and while undoubtedly important, evolved independently until the late 1980’s when market dynamics forced alignment with OSI. The decisions within IBM are given cursory review within the Robert Donnan and Daniel Warmenhoven interviews.
In this interview, Stuart Wecker describes his arrival at Digital Equipment Corporation (DEC) and his work with a variety of networking products: including the Digital Data Communications Message Protocol (DDCMP) and the DECnet networking architecture. Wecker recalls the strong influence of Donald Davies, Derek Barber, and Bob Metcalfe on his thinking, as well as the precedents of the American Arpanet and French Cyclades networks. Wecker contrasts the peer-to-peer design of DECnet to the hierarchical design of IBM’s Systems Network Architecture, and establishes how this peer-to-peer strategy was defined before the design and marketing of DEC’s VAX line in the mid-1970s. Wecker also reflects on related technical issues, including DEC’s approach to routing and congestion control, integration with Ethernet technologies, and the combination of datagram and virtual circuit approaches within DECnet.
I was unaware of the importance of Stuart Wecker until my interview with Gordon Bell. When I reached Stu, he had left DEC and had become President of Technology Concepts in Sudbury, MA. I will always remember the red brick walls typical of a cotton-weaving mill and the unusual number of interruptions that I attribute to his new role in a new organization. This interview became a technical discussion of his networking philosophies and the uniqueness of DECnet. I felt his generous gift of time was our early realization that we had both gone to: Rensselaer Polytechnic Institute (RPI).
Keywords: Digital Equipment Corporation (DEC), communication protocols, DECnet, Arpanet, VAX, peer-to-peer architecture