4 ms·
All the networks you listed are circuit switched, centralized networks; they really don't have anything to do with the Internet at all. The claim that without A
by poet 16y ago
All the networks you listed are circuit switched, centralized networks; they really don't have anything to do with the Internet at all. The claim that without ARPAnet there would be no Internet is almost self evident, considering how fundamentally different the networks you listed are from the Internet. There's no way a commercial entity would have had the same design philosophy [1, Section 3] because a commercial entity has much different motivations.
As an aside, the best example that you could have given would have been Minitel (but still, centralized and circuit switched).
[1] http://www.sigcomm.org/ccr/archive/1995/jan95/ccr-9501-clark.pdf http://www.sigcomm.org/ccr/archive/1995/jan95/ccr-9501-clark...
- WalterBright 16y agoMost of those networks also evolved gateways to interconnect to each other. The history of technology is often one that starts out with proprietary interfaces that evolve under user pressure to standardized, open interfaces. We see that playing out yet again in the iPhone vs Android competition. As soon as there were two computers, people tried to connect them. The more they got connected, the more universal people wanted those connections to be. If commercial companies wouldn't do it, non-commercial entities would (after all, look at linux and the enormous success of free software). Heck, even I worked on hooking computers together back in the 70's using ad-hoc parts and wires and (never having heard of arpanet at the time) even sketched out how a decentralized "web" might work. The idea that a decentralized fault tolerant network is such a novel concept that it required a government to think of it doesn't make a whole lot of sense. Even the telephone system is decentralized. There are decentralized fault tolerant systems all around us as examples.
- poet 16y agoThe idea that a decentralized fault tolerant network is such a novel concept that it required a government to think of it doesn't make a whole lot of sense. Even the telephone system is decentralized. There are decentralized fault tolerant systems all around us as examples. It doesn't require the government to think of it, it requires the government to implement it. It would be a massively stupid decision from a business point of view to implement the Internet as it exists today. There's much more money in centralized networks. I think this should be obvious from the fact that (1) at the time all commercial companies were building centralized networks and (2) today text messages cost upwards of $1k per megabyte. The only reason why the Internet is architected the way it is today is because of government funding. One the government set the Internet in motion towards a decentralized, packet switched network, it could not be overcome by commercial companies.
- WalterBright 16y agoFidoNet is the counterexample that shows the government was not required to implement a global, decentralized, fault tolerant network. Before that, the phone system was a national, decentralized, fault tolerant network built without benefit of government funding, research, or protocols.
- CamperBob 16y agoBefore that, the phone system was a national, decentralized, fault tolerant network built without benefit of government funding, research, or protocols. Which phone network are you talking about? Certainly not any I'm familiar with. I'm no evangelist for government-financed R&D, but I don't think you could possibly have picked a worse example (http://www.corp.att.com/history/history3.html http://www.corp.att.com/history/history3.html) for your argument.