In the admittedly unlikely scenario that another protocol stack replaces TCP/IP, would you say that it’s no longer the internet?
The defining technical requirement for the internet is connecting multiple dissimilar networks across a large geographic area. This was accomplished by 1970.
The defining social requirement is that anyone be allowed to use it, and this was in place by 1984. Eric Raymond agrees with me:
I don’t think low level details like protocols should enter into it, except for the basic distinction between circuit and packet switched networks, but that distinction was known and in use long before the internet was established.
Hah, semantics! Are we lucking out that, while it is being replaced, that the replacement is called “IPv6”?
For the purposes of your question, though, no. Using the term “the internet” generically, it’s when a worldwide network became available, regardless of the name or technical details of the network-layer protocol. It also wouldn’t matter to me whether it was circuit-switched or packet-switched.
I mentioned IP simply as the first solution that got adopted in a sufficiently large network. I think it’s also significant that it was a network where central management was minimal.
It’s interesting to view the Homer Goes To College episode of The Simpsons from 1993. As well as being a stonecold classic it shows that while the internet existed it was then seen (by the writers of the show at least) as something only nerds engaged with.
Cite? ARPAnet wasn’t an internet, as it was a homogenous network of IMPs and TIPs, connecting terminals and computers. That is, it connected terminals and computers, not subnetworks, as far as I’m aware.
I wonder when that was. About 1990 I had a UNIX PC in my basement. Back then there was only one, low volume, alt.sex newsgroup - no spam, and not even very interesting.
PLATO had a somewhat lewd lesson by 1977, and there have been lewd printout files as far back as I remember.
Well, a network of terminals + computer counts as a network in this sense. But you’re right, I misspoke, or oversimplified, or was just plain sloppy before. The ARPAnet kicked off in October 69 with only 4 computers online.
There’s also the existence of gateways, which I glossed over earlier. Even if every LAN used the same protocol like Ethernet, if they were connected over large distances an internet would be required. Because Ethernet is not routable and therefore a higher level routing protocol is needed.
So it’s not that the networks are dissimilar so much as they’re physically separate and autonomous. The gateway/routers are important because it means the network is decentralized (as opposed to every message being required to go through a single master computer or something). This was the key importance of ARPAnet in my opinion (that and packet switching, which amounts to the same thing), and it shouldn’t matter whether you consider the 4 original nodes to be separate networks or just 4 single computers. They were autonomous, physically separate, packet switched machines that used routers to communicate.
The first broadcast network was ALOHAnet in Hawaii, for communicating with data acquisition gear on volcanos. Transmitted one character at a time. Not what you’d call a LAN, of course, but it inspired Xerox PARC to create Ethernet. In 78, I was working on a competitor to Ethernet, very similar technology but worked way better with heavy loads (broadcast Ethernet starts to suck over 70% capacity). Of course, Ethernet won; it was simpler. I was happy to have a 1Mb/s network to use in the lab, even after I moved on to other projects. Others were still using 300 baud acoustic-coupled modems to dial up to the DEC mainframe.
One thing that makes ARPAnet not The Internet is that nodes were identified by 6-bit node number and 2-bit host number. It wasn’t scalable. But a bigger distinction is that the message format and protocol between IMPs was proprietary to BBN – they were free to change it at any time; no other vendors were allowed. In that it was more like X.25, which is a network ACCESS protocol, not a protocol to actually provide the service. The standardized part was between the network users (computers and terminals) and IMPs / TIPs.
It was also reliable delivery, unlike IP and more like X.25, but that’s probably not important here. X.25 scaled up pretty well despite being a reliable delivery service. Not to present day Internet, but enough to serve industry in the 80’s and 90’s.
In contrast, IP was open, so a router by any manufacturer could participate.
In the popular mindset, the internet was born on August 9th, 1995 (the same day Jerry Garcia died) when the Netscape IPO occurred. After this, it was all “internet, internet, internet” on the news, the stock market took off, and the net truly came upon the public consciousness.
See, I can’t put AOL coming into existence as the start. Those of us who were on-line before AOL teased those who could only use AOL as being Almost On-Line. So 1994 is definitely too late.
If there were ISP’s as early as 1989, the internet had to exist, or their couldn’t be a service provider for it.
So I have to agree with 1983 as well, for being the birth of the internet. In 1989 we can say it made it’s first baby step as ISP’s started to crop up. We can call AOL/Mosaic in 1994/1995 it’s debutante ball when it was introduced to the general public.
But really this whole thing was a slow roll. I remember geting my first SLIP dial up account in 1993 with a static IP!!! For me that was when the “Internet” was accessible because I wasn’t limited to telnet sessions or random local applications.
You are right about that. No spam. Almost everyone on it kind of knew what they were doing. Flame wars, but higher quality flame wars.
Internal Bell Labs newsgroups had an annual flame war about hanging up Christmas decorations in public areas. It had an agreed upon start time, and was judged by the volume of posts.
Mosaic users were newbies. I don’t know what to call Netscape users.
grumble grumble These people think you need something more than a text editor to write html.
I first heard of computer networks in the mid 1980s, when I saw a news report about white supremacists who were using them to communicate with each other, and it showed a swastika in green X’s on a IIe computer. The next day, a co-worker who also saw it was talking about it and was horrified, as was I, but said, “But it’s part of free speech.”
My first recollection of the Internet by name was around 1993, when the Paul Bernardo/Karla Homolka murder trial was going on in Canada, and there was a media blackout and also enough people who were interested in this case to get the Internet, just to keep updated on the story.
You missed the animated .plan files! those were the best but it is not like anyone on the internet will let you finger them anyway.
Oddly enough I know lots of devs who still use IRC…and gopher was awesome for finding actual information in a textual fashion. And it was accessible to anyone with a connection if you had a simple telnet client.
But really you had chat, email, newsgroups, primitive hypertext and file sharing before the “web” took off.
I did start using netscape the day it came out…because it let me read the text before the pictures downloaded
Heh. I had all those things on PLATO in 1975.
Chat - term-talk, and multi-user chat rooms.
Mail of course.
Newsgroups - pad was the first, which got expanded into many topical ones.
The Tutor language supported hypertext
File sharing was no problem since we were all on one machine, even if PLATO terminals were scattered throughout the world.
And we also had multiplayer interactive games, and the first MUDs.
All this and touch screens too.
I remember 1994 as the year where it really started exploding into the popular culture. It had been somewhat known before then, but that’s when Mosaic and Netscape started hitting their stride and making it accessible to the masses. I don’t know when I would place its birthdate myself, but I suspect it would be some time in the 1980s.
I joined Prodigy in 1994, a couple days before the OJ Simpson murder case coverage began (I think the big story when I joined was Woodstock 2). Shortly thereafter there was a TIME Magazine story about how internet users had topped one million (or maybe 2 or 5 million?). Although I’m pathetically behind the tech curve, I’ll go with 1994.
