Of the dozens of design decisions that TBL made during 1989, all of which continue to shape the way we build and use the web twenty-five years later, the most important was not requiring permission to link. Seems obvious now – a non-feature in fact – but it’s the reason you’re not reading this on Xanadu (or Prestel or on an X500 terminal or something). The logic of the times – embedded in those other systems – was that documents and data sources had owners and that you couldn’t just link to them without some kind of formal permission. Permission was defined as a system-to-system, technical handshake kind of thing or a person-to-person interaction, like a phone call, or, God forbid, a contract and some kind of payment. TBL’s judgement was that the community he was building his system for – the academics and engineers – wouldn’t want that and that the spontaneity of the hyperlink would trump the formality of permission. And, of course, he was right. It’s the spontaneously-created hyperlink that triggered the marvellous, unstoppable promiscuity of the World Wide Web. It explains the web’s insane rate of growth and the profusion of web services. It’s the root of all this.
1984. I’m well into my fourth ‘gap year’ (in fact, I’m redefining the term ‘gap’). I’m working at a telesales place in Queen’s Park. This telesales place is different from the other dumps I worked at during the slack years, though – it’s run by a Californian cult whose working practices involve shouty ritual humiliation, enforced separation from family and the loud singing of Mustang Sally at 8am. One of my tasks during my few months in North West London – which included cleaning the cult toilets, walking the cult dogs and photocopying cult ‘training’ materials – involved cold-calling organisations of every kind and grilling them about their IT. Later in life computers became my thing – look, I’m using one now! – but I hadn’t encountered one then (isn’t that a crazy thought? 21 years old and never met a computer!). I might as well have been asking questions about, you know, particle accelerators.
I had a script. My goal was to acquire ‘qualified leads’ for my client to contact later. “Do you use computers in your business?” “What kind of computers do you own?” “How many?” “Do they run MS-DOS, CP-M, GEM, DR-DOS or [long list of defunct operating systems]…” One of the organisations I phoned was called CERN. Yes, CERN – essentially the coolest place on earth. Of course, in 1984 I knew even less about CERN than I did about computers.
My list said CERN was in Switzerland, though, which made it a more interesting call than the others I’d make that day so I was going to enjoy it. The cult had taught me that persistence was vital, so, after ten minutes bothering various receptionists, supervisors and under-managers, I was put through to ‘someone in IT’. He was impatient, seemed a bit absent-minded and spoke very quickly – but he spoke English and, at some point, he obviously resolved to help the clueless drone from London on the phone. He told me he was a developer, in information systems. He didn’t buy computers for CERN and he didn’t know who did. He was obviously a million miles from being a lead, qualified or otherwise.
The cult I brushed with (they never bothered to try to recruit me) was called ‘Exegesis‘ and the telesales firm, which was called ‘Programmes’ emerged after Exegesis was investigated by police. They were a pretty scary bunch – intense, contemptuous of the unconverted, messianic.
But I had a form to complete and knew that a miserable stand-up humiliation awaited me if I didn’t finish it after spending so long on the phone, so I pushed on, obtaining a list of computer technology so exotic, so science fiction, as to render the whole call pointless. I don’t remember the detail but by the time I’d finished I’d used three extra sheets of paper and pretty much the whole of my new friend’s Geneva lunch hour. The cheer from the other telesales drones when I hung up raised the roof. I was ecstatic: on my entirely useless list there were PCs and workstations, minicomputers, embedded systems, mainframes and supercomputers… hundreds of them.
It was many years before it occurred to me that this harried developer with a rather posh voice might just have been the inventor of the World Wide Web. It’s difficult, of course, to know exactly how much influence I had on the final shape of the world-changing technology that was even then forming in that patient man’s head but it’s gratifying to know that I was there at the beginning. Something to tell my grandchildren.
The semantic web is a powerful thing but it’s… well… semantic. Trying to imagine the net in the future, it becomes obvious that we’re going to need a temporal web too. Living, as we do, in the first moments of the web’s existence, we haven’t needed to think much about time. It’s as if everything that’s taken place so far all happened in a single, cataclysmic moment.
Once the web’s lifespan starts to stretch – across generations and centuries – we’re going to need an accessible historic record. Something that’s ‘online’ (as in ‘not offline in a tape library’) and preferably ‘inline’ (continuous with the current content). In this article for The Guardian I visualise this as a ‘giant rewind knob for the web’.
My example is the war in Iraq. Imagine the benefits to humanity in the future of being able to rewind to any point in the rolling popular history we call blogging and take a snapshot of the state of the war and opinion about it. More to the point, with so much information, conversation and collaboration moving onto the net, imagine a future without it.
In the article I also wonder if we, in the UK, shouldn’t be pressing the BBC to take on this task. Lots of people think the BBC’s proper role on the net should be to boost connection and participation (and there is some ambitious work going on already). Perhaps, as well as promoting communication, the Beeb ought also to be promoting recollection.
(Maybe the techies out there can tell me if this kind of work is already going on. I’m pretty sure Kahle’s Way Back Machine is going in the right direction but it’s a long way from being fine-grained enough and it certainly can’t present historic content ‘inline’)
A temporal web?