Eric Reinhardt’s latest novel, “French Comedies” (Gallimard), offers the opportunity to discuss an interesting question: has France lost a unique opportunity to occupy the very first place in the development of the Internet ?
A novel with the battle between the datagram and X25 as a backdrop
The novel is organized around female encounters with the central character of the story, a major reporter for AFP. Behind romantic sequences, the author, through the investigation of his character, tells us about an industrial battle which took place in the 1970s. The tone is no longer really that of fiction and the battle did take place . Without going into details, it can be summarized as follows. A French engineer, Louis Pouzin, showed within the framework of Cyclades, a program of the IRIA (ex-INRIA), the potential of a new approach in terms of data transmission network based on the datagram. For their part, engineers from the powerful telecoms research center, CNET, while opting for packet switching have defended a different approach to deploy a national telecommunication network, which results in the X25 protocol.
Very schematically, in datagram networks, packets are routed from node to node independently of each other, before being put back in order on arrival. While with X25, a virtual channel is established in the network to route the packets in their order of transmission from point A to point B. X25 was retained for the deployment of Transpac while various arbitrations put an end to the ambitions of Cyclades. We will not discuss here approximations, see confusions (1) of the author (through the words of his character) on the course of this politico-industrial history mixing the (indisputable) cultural differences of researchers in computer science and telecom engineers, rivalries between players in the IT industry of the time and telecom companies as well as the administrations overseeing them. We can no doubt say today that the abandonment of Cyclades was regrettable, that it is established that the TCP / IP protocol, towards which the American project Arpanet switched before being the medium of the Internet, is is partly inspired by the work of Louis Pouzin’s team …
But should we draw the conclusion that if Transpac had been built on the principles of Cyclades, the Internet would today be essentially French?
The thesis, although widely taken up by the media in recent weeks in the wake of Eric Reinhardt’s book, is a little too easy. First, even if the Americans have recognized the interest of French work, other European contributions (especially English) must be retained and reality must not be distorted by minimizing the dynamics that existed around the Arpanet. It is then probable that the North American authors of TCP / IP (2), even in front of a full-scale Cyclades network, would have been keen to advance their own specifications in their work for DARPA. On the other hand, and as it is mentioned in the novel, it is indeed a European invention (3) which is at the base of the Web, the starting point of the extraordinary success of the Internet … without that this is enough to give a preeminence to European applications.
Our great reporter also wishes to underline the handicap that the years of Minitel (operating on the Transpac network) would have constituted for France. This is also a fragile assertion insofar as many Internet players in France cut their teeth in Minitel and where France – without occupying a dominant position – does not appear to be in retreat from comparable European powers.
It seems closer to reality to recall that other reasons explain the American domination of the Internet economy and innovation: the preeminence of their research and their electronics and software industry combined with the size and wealth of the Internet. domestic market, and venture capital financing. These are the characteristics that presided over the birth of GAFA, then nurtured them in a network economy in which “the first to take all” (4) … These are, all things being different, the reasons which convinced us some time ago that the Chinese could one day constitute a danger for the North American giants.
These comments should not be seen as a desire to relativize the work of Louis Pouzin and his team or the advantage of being at the origin of a major invention or innovation, but rather the concern to put the emphasis on the issues associated with the ability of Europeans to progress in cooperation for research and the achievement of a single market.
Finally, read “French comedies”.
1. For example, our “great reporter” suggests that the emergence of time switching in the telephone industry would have been imposed to the detriment of the datagram. An occasion undoubtedly of the author to give all his place to the character of novel which could constitute the boss of the CGE, Ambroise Roux.
2. Vinton G. Cerf and Bob Kahn
3. HTML invented by Tim Berners-Lee and Robert Cailleau.
4. “The winner takes all”