Yesterday I noticed yet another web 15th birthday story.
The usual Tim Berners-Lee / CERN story proposing and then releasing URL / HTML / HTTP freely. The precise birth of that “web” depends on which point in that process you consider significant – the proposal to do it (1989), the agreement to do it, the doing of it, or the agreement to let it go free (1993).
The point that always confuses me is the DARPA TCP/IP story – I’m guessing that’s the invention of the internet – network of interconnected communications – (as opposed to the web of information on the internet).
From memory that packet-based redundant / multi-route connectivity was invented for reasons of secure (US) military communications so that messages broken into packets on multiple, random network routes could never be (easily) intercepted, and a receiver could always know if a packet had been lost, since the message could not be rebuilt without it – secure as in reliable.
Let me check. Yep, that’s it – ARPANet in 1967/68. I guess the perspective that agitates W3C people is the “free” collaborative standard aspect as opposed to the earlier military need aspect of ISoc. 20 years between the internet and the web, but it “took off” when the web information standards were set free, since the important internet comms standards were already free to use.
[Post Note : Even spam pre-dates the web; almost as old as the Arpanet itself, 30 years.]
It wasn’t security that inspired the internet protocol, but reliability. What mattered was making sure that messages got through, even if a particular router got bombed. Packetization means that a fixed channel did not need to remain connected for the entire message — if a bomb took out a router in the middle of transmission, the receiving computer would know just what was missing, and request a resend of just those packets, and the sender would send them by a different route.
Also, the TCP/IP protocols were always “free”. That is, anybody could use them to make an internet. “The” Internet is just that one with public name servers, run originally by universities, to which anyone could connect to by getting an IP number. On top of the TCP/IP protocols were free applications, mainly email, telnet, and ftp. Nothing was stopping anybody from writing other applications, which is just what Berners-Lee did. But, as the artcle said, it is up to the author/ institution whether or not they put their application into the public domain. And it wasn’t just that CERN published the HTTP/URL protocols — what made them take off was also that they were so much more useful than gopher et al.
Hi Scott, that kind of “reliabilty” is what I mean by “security” … secure system – end-to-end reliability.
(The word security has been devalued since 9/11)
Useful and free. Agreed.