Fifty years ago, the internet was born in Room 3420 | Hacker News Reader

Fifty years ago, the internet was born in Room 3420

fastcompany.com

177 pointsvontzy30 comments

> Taylor had an even more practical reason to crave a network. He was regularly getting requests from researchers around the country for funds to buy bigger and better mainframe computers.

> Taylor talked ARPA director Charles Herzfeld into allocating a million dollars for R&D into a new network to connect the computers at MIT, UCLA, SRI, and many other sites. Herzfeld got the money by redirecting it from a ballistic missile research program into the ARPA budget. The cost was justified within DoD circles by saying that ARPA was tasked with building a “survivable” network that wouldn’t go down if any specific part was destroyed, perhaps in a nuclear attack.

This story seems very similar to the Unix story[1]. Ritchie and Thompson wanted to play Space Travel and started building Unix. Eventually, management invested for a different reason: they wanted a good system for text processing of patent applications.

I guess in the early days, no one saw the potential of computers and networking. Therefore there had to be a weird convergence of interests of various stakeholders for projects to get off the ground.

[1] https://en.wikipedia.org/wiki/History_of_Unix

I used to have circuits lab in this room, before an art student had the idea of turning it into a memorial. It used to just have a paper sign on the door marking the event:

http://lh5.ggpht.com/azn.jalapeno/SQf_wBSxnoI/AAAAAAAAArg/nH...

http://lh4.ggpht.com/azn.jalapeno/SQf9KtQGgNI/AAAAAAAAArE/L7...

> we leased this line from AT&T at the blazing speed of 50,000 bits per second

It's amazing that we still have a small number of people connected with connections that are slower than the first one ever, 50 years ago.

Funny how that technological innovation have sped up so much that 50 years is now (actually, today it'd be less) considered to be the time that the average person should have something better than the state-of-the-art at the time.

This is definitely not true for the majority of history. The way the world ran in 1979 would not have been that much notifiably different than 1929, even though people have landed on the moon. Today, the life of the average Joe is completely transformed by internet.

Amusingly, there was a commercial alternative in the same period - Tymnet. Tymshare, a computer time-sharing company ("the cloud", version -1) had machines at various locations and wanted to connect them. So they had their own packet switching network, mostly for terminal to mainframe connections. Control and route setup were centralized (this is now called "software defined networking") but packet forwarding was distributed.

Tymnet was very cost-effective. Its purpose was to share expensive 9600 baud leased lines (hundreds of dollars a month) among multiple Teletype machines running at 110 to 300 baud. Worked fine. Remote echo was rather sluggish, as echoed characters tended to come back in groups of four characters. That's because they used an accumulation timer approach - wait until you have some minimum amount of data or the timer has run out. This is terrible for latency. (The current version of that is the "delayed ACK" timer in TCP, which I consider a mistake.)

There was a real question in the early days over whether networking would be packet-switched like the ARPANET or circuit switched like Tymnet, X.25, and SNA. I thought at the time (1970s) that circuit switching would win out, because packet switching cannot deal well with congestion in the middle of the network. I didn't expect fiber backbone bandwidth to get so cheap that the congestion stayed at the edges.

Let’s not forget the significance of the wording of “request for comment” in the original networking specs. All of the inventors of TCP/IP were academics working on gov’t grants and didn’t have the authority to create specifications for network protocol... but if the spec were present as an invitation for discussion rather than “hey, I figured this out and this is how I’m doing it and you should too” then the project managers and bean counters were happy about it.

See RFC 2555: https://tools.ietf.org/html/rfc2555

I’m actually now listening to the book “where wizards stay up late”, a book about the origins of ARPAnet and the Internet. I’s a bit dry at times but I like it.

Just finished it a week or two ago after learning about it from the podcast "50 things that made the modern economy". I agree it's a little dry, but it is easy to listen to while doing other things.

Good stuff, I recommend it as well.

Wow I actually read that when it first came out as a high school computer nerd. It was so long ago I can't remember anything from it, it's all information that's kinda melted into all the other stuff in my head. I should pick it up again I remember loving it.

That little book is a gem.

This excerpt is fascinating: "..Taylor talked ARPA director Charles Herzfeld into allocating a million dollars for R&D into a new network to connect the computers at MIT, UCLA, SRI, and many other sites. Herzfeld got the money by redirecting it from a ballistic missile research program into the ARPA budget. The cost was justified within DoD circles by saying that ARPA was tasked with building a “survivable” network that wouldn’t go down if any specific part was destroyed, perhaps in a nuclear attack."

Research funding is important! I think similar arguments can be made to increase funding for NASA projects (perhaps this already happens).

Semi-related but if you enjoy the history around the internet and computing, I highly recommend watching Halt and Catch Fire (it's on Netflix). I just finished watching it for the first time and what a treat it was.

An extremely well-done tv show!

An extremely well-done 1st season, that's it. Hence why the show stopped in 2017 with 4th season

agreed - season 1 covered all the interesting material quasi-related/inspired by Compaq; then it turned into a dark soap opera that was basically unwatchable

We're celebrating this at King's College London next Wednesday 6 November. PechaKucha talks by the faculty of Digital Humanities followed by drinks. Come along HN!

https://www.eventbrite.co.uk/e/happy-packet-switching-ticket...

If you have Netflix, the segment in “Lo and Behold” with Leonard Kleinrock is worth watching just to see his energy and excitement. The whole documentary is a little overwhelming though.

https://www.netflix.com/title/80097363

In the middle of 1966, Robert Taylor took over the directorship of Information Processing Techniques Office (IPTO)[1]. It was Taylor who launched the project that produced ARPANET. This is a great history series on ARPANET, starting with "ARPANET, Part 1: The Inception."[2]

[1] https://en.wikipedia.org/wiki/Information_Processing_Techniq...

[2] https://technicshistory.com/2019/05/08/arpanet-part-1-the-in...

This day should be an international holiday.

A day where you turn off your phones and computers.

You can come to Turkey. We have this holiday for a different reason: To celebrate a republican independent Turkey.

Quote: "The cost was justified within DoD circles by saying that ARPA was tasked with building a “survivable” network that wouldn’t go down if any specific part was destroyed, perhaps in a nuclear attack."

Oh boy, a white lie that became a truth nowadays. Internet as it stands now it's so resilient that after a full nuclear war to extinct us, a future alien specie just needs to patch a few bits at cables ends and will have Earth coverage. Internet is so resilient that look at China's great firewall effort and still dissidents can communicate with free world.

To me, today's internet was born with the invention of HTTP by Sir Tim Berners-Lee. It's still an interesting article and a great achievement, but a big piece has still been missing. One can definitely credit them for providing the base though.

I was student just before https://en.wikipedia.org/wiki/Eternal_September. I remember using a lot gopher (https://en.wikipedia.org/wiki/Gopher_(protocol) ). There was some hype around IRC. Mosaic has arrived a bit latter. At that time everyone was discussing network applications. We were playing sun mazewar, xrisk and other connected games (like https://en.wikipedia.org/wiki/Internet_Go_server). We were using xarchie to find software. I was using xhtalk to chat with remote friends. Chess games between two remote players (single threaded X application managing two displays), remote administration were common student projects.

IMHO, if Tim Berners-Lee had not invented HTTP, someone else would have build something else to improve over gopher.

That’s the World Wide Web, a different thing from the Internet. There’s plenty of non-HTTP-based usage on the modern Internet, from SSH, to FTP, to chat & videconferencing, websockets, WebRTC, online gaming...

I'd been using the Internet for 10 years before HTTP came along - it was definitely just as useful then as it is now.

Its just prettier now.

I remember a lot of time using archie (xarchie) and ftp before the web became ubiquitous.

Back in the pre-web days, we used to get software via ftp, usually as source code tar balls, from any of several well-known archive sites. One I remember was labrea.stanford.edu. Yes, it was a tar pit...

Standing on the shoulders of giants and all. The internet is a success big enough to handle multiple "inventors." Also, no small feat was accomplished by those that put a computer in every home or so (for their own profit of course). Without them, no need to connect

No, you are showing your ignorance. The World Wide Web is one of many things that can run on the Internet. The World Wide Web is not the Internet, just one aspect. What you are saying would be like claiming that TV was not TV until MTV appeared.

Reply on news.ycombinator.com