Thoughts about the state of the Internet, circa January 2003.
Great changes are the result of small choices. MANY small choices.
The world, like the Internet, is a social contract. Enforcement is not part of the contract.
Be careful what you ask for. You might get it.
20.000 messages arrived in my mailbox last month.
5.600 of them were certainly spam.
I'm told that I can get herbal viagra for free, view persons doing improbable acts, or find eternal bliss. All online.
At the same time, I have....
a friendly message from a South African cafe owner talking about the project where he and I are part of a team of fifty or so pepole running a registration site for Linux....
a message from a colleague in Seattle telling me that he's forwarded my comments of last night to his colleague on the US East Coast....
a message from someone I've never heard of before asking me for help with something I wrote something sensible about five years ago....
a message from my mother telling me how much she enjoyed a birthday party she attended....
Time to surf the Web.
My favourite newssite contains a link to a website for people who want to put pictures on buildings by manipulating room lights, and one to a discussion on how Nike is defending its rights to mislead people by claiming free speech rights...
The stock price of my favourite stock is down slightly.
My bank account is in no better shape than usual.
A Norwegian came in second in a ski competition in Finland, all of 2 hours ago.
I didn't pay anyone for any of this information; I've ignored perhaps a dozen banner ads.
the Internet went through what was probably its last flag day ever; all 1000 computers switched from using NCP to using TCP/IP on January 1.
In order to check my bank balance, I have to pull out last month's statement and subtract the amount of the cheques I've written.
In order to figure out who won the ski race, I have to wait for the TV news or tomorrow's newspaper.
All the people I work closely with live in Trondheim.
And I've never heard of the blinkenlights project.
it's Sunday morning. In the past hours I've answered a dozen emails, written parts of an article, fixed a few software bugs (with feedback from the guy in the Netherlands who encountered them), cleaned up my garden, and spoken (briefly) with my kids. Not doing anything for an hour would seem like a Strange Thing.
Who's telling me I have to hurry this fast?
We didn't have one.
There were lots of ideals around. "Freedom" figured high among them - meaning the ability to share ideas, get information, figure out what was really going on, and making up your own mind. Other terms bandied about was "survivability" and "no central point of control" - partly because of the military influence of the funding source; partly just because the buzzwords sounded cool...
Back in the early 90s, people decided that this Internet thing (being used mostly for email and file sharing) looked cool. And that people who weren't eligible for government grants would also want to get in on the deal. So they decided to offer Internet as a sevice, "gatewaying" into the academic-only "real Internet". Thus was born the commercial Internet.
And on April 12, 1994, Canter and Siegel sent the first spam message.
In late 1990, Tim Berners-Lee was challenged to make a better phone book. He
had read about Ted Nelson, and thought hypertext was a cool idea. So he started
the Web.
In 1993, Marc Andreesen released the first Mosaic.
At the time of release, the first porn Web site ("Amateur Action", descended from a BBS by the same name) was already in operation.
In many ways, society is built around an unwritten contract.
Consider as you walk down the street the number of unbroken windows.
This is not so because the police catch those who break windows. It's because the vast majority of the people who are able to break them would like to live in a society where windows are not routinely broken.
A very similar social contract existed on the early Internet: it was a cooperative venture, where everyone knew that everyone had to cooperate to keep it running - so everyone did. Offenders were tracked down and chastised quickly - usually by operators at the sites where they had their Internet access; the ultimate threat was "behave or lose access". But there was nothing legalistic about it - people did what they thought was the Right Thing, most of the time.
An often overlooked aspect of the current Internet is that the contract still exists.
The Internet works because when congestion occurs, applications back off. This is enshrined in Sally Floyd's "Congestion Control Principles" (RFC 2914, published September 2000). But long before that, we all knew about the term "congestion collapse" (first seen in RFC 896, January 1984): If major applications do not behave within certain VERY strict limits, the Internet will collapse.
People have obeyed the rules. And continue to do so. Because they know it's required to make the Internet work.
Most people don't think of themselves as bringing down the world.
They just want to do things a little differently - garner a slightly bigger slice of the pie for themselves; hog some more bandwidth; listen to some music they haven't bought; do a few things where they don't see any good reason not to. "What's the harm?" is probably the most common excuse for this behaviour.
And in many cases they are right. There is no harm.
But in other cases, the aggregate effect of hundreds or thousands or millions of people doing the same "harmless" thing has great effect. Spam consumes 1/4 of my email - and I get a lot of legitimate email.
Perhaps equally important is the effct on the perception of the Internet; a porn site that I do not see hurts me not at all - but the impression that "the Internet is for porn" hurt my ability to convince friends to bother about the Internet until it became so obviously useful that "everyone" connected anyway.
And of course, lots of things are done out of sheer ignorance - Tim Berners-Lee didn't set out to damage congestion control on the Internet with the asinine HTTP/0.9 protocol - he just didn't know any better.
It is most tempting to bend the rules when a seemingly innocous "getting advantage" gives real financial benefit - when Netscape came out with a version of their browser that opened four sessions to a server rather than one, the howls of outrage from the people fearing what this would do to the network was drowned out by the gleeful shouts of the users seeing their page load speeds go up, and the sound of money raining into the pockets of Netscape's owners from the ever increasing stock price.
And we were rescued by the advent of fiber and multigigabit transmission systems.... if we are lucky, not so many mistakes will be made when IP telephony over the normal Internet gets into everyday usage.
In the OSI wars, part of the reason why the Internet flat-out won was that it started with usable applications. Primitive, clunky and not well designed - but you could do something with them the day you installed!
On the other hand, the OSI protocols had a data connection. You could install a network (X.25) - and have nothing. Except that you can easily install an Internet IP service to run on top of it......
Once the naming, addressing and functional habits have been established, it is VERY hard to change - what you want to change to has to be MUCH better (or forced upon you - see Microsoft Exchange).
Moral: Partial solutions invite stopgap solutions. And once they stop the gap, you are well and truly stopped.
In the early 80s, "everyone" (apart from a few strange people in California) was convinced that the future lay with OSI and the internationally standardized solutions for communications. It just took a little while to build them, so while we were waiting, we would roll out some TCP/IP to "fill the gap".
Nowadays, it's obvious what happened. We ended up (as far as anything's ever ended) with a much leaner, meaner protocol suite, and an Internet that's become a global force for change, not "just another networking technology".
There's no truth to the statement that "if we don't invent it, someone else will".
The technologies we invent influence the world greatly - because of what they make possible, what they make easy, and what they make hard.
The omissions we do are as important as the commissions; if we fail to find a way to do something, that something will not be done.
We need to think. And we need to know what's required to make the things we want to happen, easy. Because just making it possible is not enough.
What we make possible may happen. But what we make easy, WILL happen.
We can't not act. So let's act responsibly.
We have to live with the results anyway.