Fergus Cassidy

Hark the Harald

The Internet Engineering Task Force (IETF) is a self-organised group of people who contribute to the engineering and evolution of internet technologies. It is the principal body engaged in the development of new internet standard specifications.

The IETF is unusual in that it exists as a collection of meetings or 'happenings', but is not a corporation and has no board of directors, no members, and no dues.

But it does have a chairman. Harald Alvestrand shared his thoughts with me on a number of general issues to do with the net. While speaking in a personal capacity, I nonetheless think that his views are well worth relaying to you:

At the moment the net is perceived by many to be going through the doldrums. Much comment is pessimistic and hand-wringing, with issues like spam and viruses to the fore. The death of email has even been predicted. What's your view on this 'doom and gloom' scenario?
There's nothing new about doom and gloom. There's a common expression on the internet: 'Death of the Net predicted: News at 11'. It's usually used to point out that doom and gloom is old hat, and that the net has survived many crises before this; before we believe that things are going to crash and burn any time soon, we need the engineering evidence.
   What we see when we look at the real world is that email and the internet are enormously useful tools, and that millions of people have it as an important part of their daily lives. And its usage is still growing. This is another crisis. We're used to those.

What stage do you think the net is at right now?
It depends totally on what part of the net you're looking at. Clearly, parts of the net - the IPv4 infrastructure, the pure provision of bandwidth - is out of the 'experimentation' phase and making the difficult transition into the 'commodity' phase - it has to be there, it has to work, and apart from that, the public cares no more about that than about the people who build highways or power grids.
   It's still a challenge, but it's a challenge where what we want to achieve is more or less defined.    Other parts - the commercial web arena, for instance - are still in the 'sorting out' phase, where different business models are being tested, accepted and rejected in the market, and while it's already deeply embedded in people's daily lives, the changes have not stopped.
   And of course, there is the 'new stuff', with peer-to-peer applications and direct internet telephony being some of the ones that currently have public mindshare, but more 'esoteric' stuff like web services and computing resource grids might turn out to be more important five years down the road - in that space, we still don't know the names of the applications we will take for granted in a few years' time.
   The net isn't at a single stage. It's too diverse for that.

You remarked recently that a social contract existed on the early internet - where everyone had to cooperate to keep it running. Could you explain more on this?
This is a big concept to fit into a few words. In the early internet, connection was established through personal contact - when MIT or Norway connected these weren't faceless - it was "Jeff's connection" and "Paal's connection".
   These people wanted the internet to work - and knew that bad behaviour on the parts of their users would reflect on them. And they passed this understanding on to the people who they allowed to connect to the net. The chain of personal responsibility was enough to keep the net running.

Will that idea of a social contract be workable in the future?
It has to be. Just like the real world, it is not the presence of policemen that keep windows unbroken - it's the fact that most people understand that their neighbourhood is a better place if they keep them whole.
   Of the millions of internet users, there are only a few who do not understand this. We have to deal with them - but without the social contract, there's no way the net will continue to work.

You have said that enforcement is not part of the social contract. Yet right now that is the direction which some governments and legislators seem to be heading in. Do they have a point?
They have a point - the social contract needs to be backed up by mechanisms to deal with egregious violators. But I'm very wary of the tendency to push aside old and valuable concepts like due process and "innocent until found guilty" when dealing with "internet crime" (whatever that is).
   Current debate sometimes seems a pell-mell rush to circumscribe all people's liberty and privacy in the name of the pursuit of a few sociopaths. And when it's a genuine disagreement, like the file-sharing vs RIAA debates, or the DVD decryption debate, attempting to shortcut the debate by passing laws is dangerous indeed.

A couple of years ago, you suggested that "the internet is an idea, a network and a society". Have your views changed and what do you mean by a society?
In the years since that statement, the 'society' of the internet has grown manifold. In parts of the world, it's hard to see the difference between the society of the internet and the society of people.
   Still, I think the concept is valuable - the society of internet users still shares characteristics like the belief in the free exchange of ideas and the idea that there's nothing magical about borders - things that its denizens often think of as "too obvious to mention". So I think it's still a good line.

From your vantage point, has the commercialisation of the internet changed the direction and purpose of the IETF?
It has certainly influenced it. What were million-dollar decisions ten years ago may now be billion-dollar decisions. But a lot of the people are still the same, and a lot of the arguments are too.
   I'm sometimes surprised at the degree to which the IETF remains faithful to its focus on technology, openness and rational argument; the organisation is still chiefly focused on making technology that is useful to the internet users.

What do you see as the biggest challenges for you and the IETF?
As Mike O'Dell is fond of saying, "Scaling is The Problem". We're still working in the IETF in many of the same ways we did ten years ago, and the strains are showing. Some things have to change.
   But the task has been likened to "changing the engines while the plane's in flight" - it requires careful thinking, and a LOT of time.

Any idea on what the net will look like in ten years?
I think that 10 years from now, people will take it for granted to such a degree that they don't even think about it - the ability to communicate is simply something that is there; a platform on which new and interesting things are built at need.
   We're seeing the beginning of that now - a few years ago, people asked "do you have email?"; nowadays, it's "what's your email address?". Ten years from now, I expect to no longer have to worry about "how can I connect to the Internet from here"; I'll simply be connected.
   But the future tends to surprise me - I'm looking forward to seeing it!

Lastly, If you were granted three wishes for the internet, what would you ask for?
Three wishes, no holds barred? :-)
- That people would listen as well as speak when they debate.
- That there were effective ways to deal with spammers, and that the spammers who are clearly criminals would get arrested for their crimes.
- That there would be peace on earth, and good net for all mankind :-)


LINKS
The Internet Engineering Task Force (IETF)

Harald Alvestrand's web site which includes his "thoughts on the state of the internet, circa January 2003".