Tuesday, January 29, 2013

Why the Internet works and how to break it



ernet was a person, it would be beginning to feel its age this year as it gets into its 30s, with a mid-life crisis looming. As it happens, the internet has never looked better: it's faster, bigger, better and richer than it was in its 20s.
But there are people having a midlife crisis on the internet's behalf. Governments want to change how it is governed, how it works and, most disturbingly, its openness. So, it is worth taking a moment to outline the first principles of the internet that have made it successful, why they are worth preserving and what we can expect if they are preserved.
Bob Kahn and I began work on the design that became the internet in the 1970s, motivated by the spectacular success of the Arpanet project, funded by the US Defense Department , in which small computers sent data 'packets' across dedicated telephone circuits. It was a homogeneous network connecting inhomogeneous computers: different operating systems, different word sizes, different computational capacities.

We met in 1973 at Stanford and started working on a design to allow up to 256 networks to be connected in such a way that the host computers would not need to know anything about the layout of this super-network . At the same time, every host computer would be able to talk to every other one despite their different operating systems and other differences. We also worked on a detailed design of the Transmission Control Protocol (TCP) and began implementing and testing it in 1975.

We were sure that this was powerful . The packets we were using to transport data were remarkably adaptable : they could be transmitted over any digital communication channel , bringing with them any information that could be digitised. The network was not designed for a particular application and this has allowed it to support applications that weren't predicted in the early formulation of the internet's design.

We didn't , for instance, anticipate the hand-held mobile. We did anticipate an 'Internet of Things' — more on that in a moment — and personal computing. We even foresaw notebook computing, whereby a computer that isn't powerful can perform tricky tasks by drawing on the internet. We didn't have to imagine it. Alan Kay had shown a notebook-computer concept around 1968 he called FLEX and Xerox Parc built the Alto personal computer along with the Ethernet in the early 1970s. They were living in a world that others would not experience for 20 years.

The system Bob
 and I designed, alongside collaborators from Europe and Asia who visited our lab in the mid-1970 s, has since grown by factors of a million or more on all dimensions : a many million times more users , a million times more hosts, a million times more networks, all connected a million times faster.

But the numbers aren't the only difference . The internet era is different from the telephone era for at least two reasons: it allows groups to communicate , coordinate, collaborate and share information, and it supports every medium of communication invented, all in one network. People can discover each other without knowing who they are and they can find groups with common interests.

The institutions spawned by the internet and which regulate and build the internet are similarly meritocratic and diverse. The Internet Society , Internet Architecture Board, Internet Engineering and Research Working Groups, Internet Governance Forum and Internet Corporation for Assigned Names and Numbers : all of them are run by many 
stakeholders who together decide policies and standardisation. It is a meritocracy that respects ideas more than institutions. It values openness and sharing of information, freedom of choice and expression.

Of course, the internet can be abused and people harmed from that abuse . Protection of personal information should be a high priority for all internet application providers. We also need to educate people about what can happen when they share information on the internet: once it is available to anyone, it is possible for someone to upload to other sites or to capture and store the information . Any country that gets the internet soon finds out that some harm comes from people who are in other national jurisdictions. We will need to find ways for international cooperation to deal with abuse.

But as we do figure out better ways to make cyberspace safe to use, we must preserve the very properties that have made it so successful: freedom of expression, transparency and openness, participatory policy and technology development.

No comments:

Post a Comment