Friday, December 07, 2007

my learning path leads to a theory scheme

My learning path coincided with major Networking advances like the formation of RIPE and the EBONE and the withering of the ISO protocols like X.25 and X.400.
Also, the transition to Autonomous routing and later to CIDR, the rising of Multimedia, the creation of the World Wide Web just after the Archie and Gopher network services initiated the Networked Informational Resources track.

As my path started from the rather general consideration "what is the Net ?" like the guys at SUN Microsystems started at "the Net is the computer" it unavoidedly set going a radar on net innovation. This is why it bleeped on notable signals that I was lucky to receive from developments like RIPE, EBONE, CIDR, MIME, ARCHIE, WWW. In fact, some time later O'Reilly media also thought alike.

For example to OSI or not to OSI was a major Net issue in early 80s where I got heavily involved. How heavy ? well in the order of turning a National Research Network to adopt the Internet tools and methods (aka TCPIP, and IETF, RFC) rather than those of COSINE (aka X.25, X.400, FTAM and ISO). COSINE was the name of a EU project to advance networking. The name stands for "co-operation for open systems in europe", a grouping of experts (like IETF perhaps). The project was based on the idea of creating local organizations, the so called NRNs (national research networks) which would meet in a federated union like RARE (like NANOG perhaps).

Another example is my path hit was the problem 'what is networked information' as I was a member of the working group that pondered the issue. I guess it is no coincidene that XML FAQ maintainance is done by a colleague in the group Peter, peer from in Ireland as National Reseach Network developer.

This kind of investigation about the big picture prompted the radar to detect the previous stages, to seek a Darwinean like logic of developments. How come ?

Now, this was not just a random thread, strung by the famous resident myth of Ariadne (the project I was leading techinically in greece) but driven by the background framework of a general evolution theory called Metasystem Transition Theory by V. Turchin. A physisist that turned to “meta-computing”, his customised term for Functional Programming, a field of Computing associated with paradigm shifting the so called Von Neumann/Turing model in early 80s.

A relevant event is also my involvement in this line of work with David Turner, who fixed at the time the so problem of software crisis as the ultimate motivation for functional programming.

Symbol system evolution was critical for Turchin aiming to form a possible picture that models the emergence of Science, the most advanced point in his evolutionary scheme. Turchin creted the term “meta-science” to mark the next major advance to follow Science. In this course, Turchin assigns software a major role and
arrived at the idea to emulate Nature’s progress that evolved constructs such as Life, Brain, Language, Symbol systems, Society and Science.

Emulation technically means for him implementing in software the fundamental mechanism of MST (see earlier posts). He has obtained and published quite a few results (the supercompiler, A proof of ZF axioms completeness in constructive mathematics).

He is a great believer in the importance of the programming language REFAL (belongs to functional programming category) and he published its possible use as an alternative to XML the much awaited development on the WEB which does induce a further level of automation into processing networked resources.

I shifted the locus of attention and said "what if I think metasystem transitions in Networking" . The result was to extract from observation a fundamental concept modeling network advances. I was lucky to be involved with some of the occuring changes and the practise of the learn-by-doing was feasible. I generalized the innovation path that I followed on the radar or traced in retrospect as "a thread of network automation steps" accumulated by the global Network entity

The thread started with the Strowger switch, automating the manually operated switchboard connecting people in La Port-Indiana, the first network automation quantum gradually propagating everywhere until caught by the next Tofflerian wave of the digital switch giving rise to the global telekom network. And then ARPANET was overlaid, the mother of all computer networks inducing a new kind of automation by the NCP nodes.

If I am to be strict, the really first automation quantum on the Net must be that associated with the system of fryctories in ancient Greece. It is said that it was used to transfer at almost the speed of light (it did use light from fires) the news to Mycenae that the Trojan horse had succeeded in Troia.