A chat with internet pioneer Charley Kline
On Wednesday October 29th 1969, a UCLA graduate student named Charley Kline sent the first ever transmissions over the ARPANET – the research network that evolved into the internet. Ahead of the 42nd anniversary of that momentous day, the AOL Mail Blog spoke to Charley (who is an AOL Mail user) about the origins of the internet, the role of email and the online future.
AOL: How did the networking project start?
Charley Kline: The Defense Department's ARPANET [Advanced Research Projects Agency Network] was the first packet switch, or data transmission, network. There are all kinds of rumors that it was developed to build a network that would survive nuclear war but that's garbage. The project was partly about researching network technologies but mostly the goal was just to connect all of ARPA's different sites so their people could work together.
How did you get involved?
In the late 60s ARPA was funding about half of all computer science research in the world, mostly at universities. I was a graduate student at UCLA and my professor Len Kleinrock had done his doctoral research on packet switching so being part of the ARPANET project was an opportunity for him to test out his ideas.
The problem was that different computers were incompatible and couldn't talk to one another. A guy named Wes Clark had the idea of building a switch so each computer would only have to work out how to exchange information with this one switch and not all the other computers. This was called an IMP or Information Message Processor – today you would call it a very primitive router.
UCLA received the first switch and Stanford Research Institute (SRI) got the second. Of course we had to build our own hardware to connect the IMP to our computer system. There were no standard interfaces back then, you couldn't just plug it in.
Once we had it connected, I developed software for our operating system to transmit stuff but all I could do was send data back and forth to myself as a test until SRI got their IMP set up.
What happened on October 29th 1969?
Once the IMPs were working, we decided the simplest idea was to use a terminal on one computer to log in to the system on another computer. The SRI system was a little easier to access than mine so the first test was for me to log in to their computer, which involved me remotely typing the command "LOGIN" into the SRI machine.
So I'm on the phone to SRI and I type the L and say, "OK I typed in L, you got that?" Bill Duvall, the guy at SRI, is watching his monitor and he has the L. I type the O. Got the O. Typed the G. "Wait a minute", Bill says, "my system crashed. I'll call you back".
Bill's computer was smart and would finish known commands. It knew that the letters L-O-G could only stand for "login", so it finished the command and sent the I-N at the same time as the G. However, his system only had a one-character buffer so it had a buffer overflow and crashed.
He fixed that and an hour later it all worked, so I could type commands and use his system remotely. And that was the beginning. That was October 29th 1969.
Were you aware of how significant this day would become?
I had no sense of how big a step this was. I was a 21-year old graduate student having fun hacking on computers. We all thought it was neat that the packet switch worked but we had no idea it would become a big thing. After all there were only two computers in the world that could do this.
The funniest thing was that about 10 years ago I was watching Hollywood Squares and one of the questions was "What was the first thing sent on the internet?" What I did became a trivia question. I certainly never expected that.
When did you begin to realize the possibilities of networking?
Probably in the 80s when there was much more commercial interest in networks. But even then I didn't foresee the drop in communication costs or the advent of the PC, so I didn't see the mass-market appeal. At UCLA we had a big mainframe – one of only two that had 4MB of memory. It cost $4m and took up an entire room with huge air conditioners. That's what I grew up with so it was hard to imagine affordable and powerful PCs. I did think that people would have dumb terminals from which you could access services like AOL but of course now I have a phone with far more processing power and memory than any of the giant machines I worked with at UCLA.
What helped drive the commercial interest in networks?
Email quickly became the killer app because it was simple, easy to transmit, reliable, robust and cheap.
I've been an email user from way back. I probably have 20 accounts including AOL. I've always had email accounts at the companies I worked for but I liked AOL for the convenience – I could get to it anywhere. Even though I was early on AOL it still wasn't early enough to get the name I wanted. Charley went very quickly.
How could email be improved?
My PhD is in computer and network security so I've always felt that emails should be signed and encrypted. I pushed for that and better authentication, which would help us catch the spammers.
Spam is a big problem. Spam filters are pretty good – they filter most of it but I still have to look at my spam folder to make sure something important didn't end up there.
Encryption wasn't done originally because it took time and computers were slow, but today there is no excuse to not have it.
What is the future of online communication?
I've been talking about online video chatting for years. You had picture phones in the 60s but they were really expensive and pretty much didn't work. Now you have Skype and AV by AOL, but I think the next step is video mail where you can record a message and send it as an email.
In a world of social networking, is email still relevant?
On social networks everybody sees what you post. Unless it becomes more convenient to change this each time you post a message, email will remain useful when you want to talk directly to people.
Official messages like statements from your bank will still come via email especially as improved security and DomainKeys Identified Mail lets you know that it is a genuine message.
Then there are commercial services like daily deals newsletters. People won't want them cluttering up their Facebook page but email lets you filter them into folders until you're ready to read them.
How will you be celebrating the 42nd anniversary of the first internet transmission?
I'll be at an event at UCLA this Saturday. Len Klinerock is trying to restore the room where we had the IMP and make it part of an Internet History Center.
Did you ever think those computers would become museum pieces?
No, but what better place to celebrate the internet than where we first started using ARPANET and developing the protocols that would evolve into the internet.
Photos courtesy of Charley Kline