We follow the money

Web 2.0 Techno utopian types love their earthy metaphors. The web is a new planet that’s being “terraformed” before our eyes, one advertising consultant likes to say. Or the “web is a garden“, if you believe Sun Microsystem’s director of research.

Even my overgrown garden doesn’t have something like this lurking in the corner, and I hope there isn’t such a horror in yours.

But enough with the hot-tub psycho babble. The future of computer networks is both a lot more promising and a lot more ominous than anything you’ll hear at the “Web 2.0” conference in San Francisco this week, where some of the web’s horticulturists will be gathering for an evangelical uplift.

There’s every reason to be optimistic, now in 2005, that computer networks can begin to fulfill their potential. They can even start to be really useful – but it’s only by dispensing with such utopian nonsense – so we can really begin to see what these tools can do for us. Here’s a reality-based guide to what’s happening – and if you hear a futurist omit more than one of these in a presentation, send them to the Exit toot-sweet, with a firm smack on the backside.

1. I for one welcome SkypeBay and GoogleNet, our new corporate overlords!

The Web has become so synonymous with inter-networking that people forget what we really have is a variety of computer networks of different kinds. The most popular messaging software in the world runs on one of these closed networks: it’s called SMS.

In the 1990s, people found it useful to think of One Internet because thinking of a single, unitary computer network appeared to imbue the technology with its own agency and purpose. The One Internet is popular with today’s Californian hive mind set, who view technology as a short-cut around all kinds of messy social and political problems – and that’s the thinking behind Web 2.0 utopianism, too.

But out in the real world, it doesn’t work like this. Limits are invariably imposed on technology by economic or social factors. For example, the reason dark fiber isn’t exploited, and that we don’t have terrific gigabit speed ISPs here in the US (after billions of dollars were invested laying the fibre) is simple. It’s because a) it isn’t economic for a private company to exploit it – there’s no money in it – and b) because there isn’t a consensus for the state to subsidize their operation, either. In South Korea, where government directs capital more explicitly, things are different, and you can see a tentative change in attitude to this here, with the growth of the Muni Wi-Fi movement. Limits are also imposed because the public finds a technology unpalatable: nuclear power and GM food are good examples. The only sure thing is that limits are, eventually, imposed.

In the last month, even the most utopian of Californian technology evangelists have begun to realize that the ugly reality behind the economics can’t be wished away. Google may be embarking on owning a network infrastructure of its own, and eBay splashed out $2bn on VoIP leader Skype.

Fifteen years ago, the rights holders were told that it would soon be possible for secure, and reliable and speedy delivery of content to people’s homes – Video on Demand. Instead the internet came along, which was anything but secure, reliable and speedy. The Wintel lobby jumped to the front of the queue, elbowing aside the consumer electronics manufacturers, and said “Sure! We can build it for you!”

And the impasse has existed ever since. Few people would have objected if a zero-cost box alongside your VCR allowed you to choose one of 50,000 movies tonight. The rights holders still want to this, only by borking your computer. People quite rightfully responded with, “Get the hell out of my PC!”

One is an acceptable balance, the other isn’t. Maybe GoogleNet thinks can succeed where everyone on the open, PC-centric internet has failed. We shall see, but we know a lot of people are going to keep trying.

So last week, even Supernova conference organizer and deregulation advocate Kevin Werbach noted the trend, with some disquiet:

“The threat of vertical integration from the bottom of the stack has been with us since the earliest days of the commercial Internet. Now, surprisingly, it may be coming from the top as well,” wrote Kevin in a post entitled More thoughts on the fragmenting Internet.

“I just can’t help thinking that we’re moving away from the common platform that defined the Internet for the past decade, and we haven’t really examined what that will mean.”

2. What’s value for money in the New Walled Gardens?

But what limits are acceptable? Most of us can agree that when mobile operators block well-established TCP/IP services, such as SMTP and POP3, this isn’t on. And many Reg readers don’t view DRM as acceptable, either. However, the only way to ensure DRM is banished forever for the consumer is either to legislate it out of existence, or to ensure there is no financial incentive for rights holders to pursue those locks and keys – and that attractive alternative compensation paths exist. Given the strength and brutal effectiveness of the rights lobby – and the ineptness and navel-gazing of the technology lobby – the legislation we’ve seen so far has been entirely in one direction. Like many people, I’ve concluded that only a blanket license, or some kind of compensation pool for digital media, can fix this. It’s proved successful for other new technologies, from the player piano, to radio, to loudspeakers in public places. And it’ll eventually work for digital media, too.

So until the rights holders can be convinced of they’re better off under a blanket license – and they’re actually far less opposed to it than they publicly admit – we’re going to be stuck in a DRM dark age.

However, cracks are appearing in the dam. 2006 will see some further fragmentation of what we’ve traditionally accepted at the unitary open internet, as the rights holders give their blessing to experiments such as Mashboxx and PlayLouder MSP.

In these networks, subscribers can merrily swap away, knowing that artists will be compensated. The bargain means that the ISP will try and stop leakage of copyright material leaving the network – which is undoubtedly a technological limitation. But there’s hope that these silos will eventually have enough common purpose to merge – effectively introducing, from the bottom-up what probably won’t be introduced top-down by an order from the Library of Congress’ Copyright Office.

Will these experiments work? It depends on what balance they strike, of course.

Reader Scott Pedigo is vehemently opposed to blanket licenses on principle, but says he might accept a service if these criteria were met: “1) the idiots quit attacking their own customers; 2) the price is right; 3) there is either no DRM, or one which doesn’t get in my way AT ALL.”

I’d agree – that’s a good criteria. For now, no one knows how these new systems will play out. If they’re successful, it’ll be because they’re offering much better value-for-money than either 99 cents a song, or the $14 a month subscription surcharge services such as Rhapsody charge on top of your monthly ISP fee. In other words, the cost of participation will be so much lower and the choice of music (everything on everyone else’s hard drive) much richer. If they go badly, it’ll because they offer the worst of both worlds: restrictive DRM and more blocked ports. We’ll see.

But the Walled Gardens are an example where the cry of “t0tal fr33d0m dUdeZ!” is no longer a destination.

3. Packet-level Balkanization

Ironically, on the day Bubble 2.0 kicks off, there’s a peering spat between two backbone providers, Level3 and Cogent. It threatens to slow down or block off their customers, and anyone trying to reach them. We’ve taken much of this for granted – and no one ever gives issues that concern ISPs much of a thought, even expensive issues such as the internet’s Background Radiation.

Expect more such disputes as the GoogleNets, SkypeBays and Walled Gardens begin to exert their influence. When infrastructure providers also deliver content, packet prioritization and QoS become the next battleground. There’s currently a lobby insisting that no controls be put in place: ostensibly to stop Big Telco borking Skype. Once it becomes an economic imperitive to do so, this will either be circumvented on the sly, or ignored by the New Walled Gardens.

And that’s just in the West (see #5). If you’re in China, ask your local Communist Party representative about the special QoS offers available today!

4. On every Commons, a toxin

When closed networks are unregulated, their owners inhibit technology that threatens their business. When open networks are unregulated, they’re filled with spam. It isn’t pretty: email spam, phishing, blog comment spam, TrackBack litter, fake SEO sites, Wiki wars, and now tag spam. As long as spam pays, it will flourish.

The cumulative effect of careless infrastructure design is now taking its toll. For the first time, Gartner Group warns, security concerns are hitting e-commerce where it hurts: in the wallet, with the analyst company revising predictions downwards.

So how do we strike a balance? There’s no easy answer, but we must first acknowledge that there’s is actually a problem first before we can begin to tackle it.

And not all spam is created by nefarious scamsters. In 2003 a few thousand weblog users managed to knock out Google, thanks to Trackbacks – a hack that allowed bloggers to leave comments with each other. It scored high on ingenuity, and thrilled the Wiki-fiddlers, but made the search results unusable. Sounding rather like Tom Lehrer’s Wernher von Braun, Trackback’s creator Ben Trott acknowledged the problem, but declined to take any responsibility: “This evolution of weblog pages isn’t designed to conform to any one search engine (or journalist),” wrote Ben.

So, “the rockets go up, who cares where they come down?” as Tom Lehrer put it. “It’s not my department”.

It would help if every technologist must share the ecological concerns of the terraformers – – and at least a little sense of social responsibility.

5. The China Syndrome

The US is in one of its periodic bouts of panic – this time about the inexorable rise of the Chinese economy. This panic isn’t shared by the Hive Minders, for whom the sky is always blue, and who assume with that peculiarly Californian sense of entitlement, that this new world is theirs by right.

But Asia has a history of doing things differently, and it has a distinctly pragmatic approach to computer networks. Asians doesn’t mistake the Web for being the only computer system in the world, and many Register readers in the West appear to agree.

China has vowed to control its own technology destiny. It’s ruthless about “openness”, and unlike Japan, is developing through innovation not imitation. In addition to its own CPU, DSP and 3G air interface, Chinese engineers have already proposed a successor to IPv6, IPv8. One wonders whether the PRC needs a next generation TCP/IP when it has already created a Walled Garden using today’s IPv4.

You can argue that the China factor really belongs with #4 – but it’s more economic than technical. A nation which isn’t afraid to tell Bill Gates’ Microsoft where to get off, feels no obligation to listen to HTML jocks.

6. The Rise of the Anti-Social Software Movement

You’ll have noticed that some of the most utopian technology advocates here are also the biggest control freaks. That’s no coincidence. For example, one reason that many (but not all) of the earliest weblog evangelists liked the tool so much was because they could turn off criticism. Some have gone onto new technologies, and tried to control what’s said there, too.

It’s a good job they didn’t invent, say, the water cooler.

So over to Charles Eicher, who has the following suggestion: Don’t fight it – Embrace it!

“Any Anti-Social Software application is defined not by what it includes, but by what it excludes. Anti-Social Software users want control over what they exclude, they want to filter out the noise from the signal, and most of the signal too. Until Memeorandum can clearly demonstrate how content is excluded from its pages, under what criterion, and until users have control over who and what is excluded, it will be unsuccessful in the new Anti-Social Web,” he writes.

Single digit growth? Web 2.0 organizer and Google hagiographer John Battelle illustrates his ‘database of intentions’ pic: John C Dvorak

“I would invite people to discuss these principles at my upcoming conference on Anti-Social Software,” continues Charles, “however I cannot since nobody is invited.”

Wonderful stuff. Surely some new Honor Tags must be minted for the burgeoning Anti-Social Software movement?

This we might call “User Level Balkanization” – the fact that people use computer communications to be anti-social. And if you think we’re making this up – look at the first use that Google’s NOFOLLOW tag was put to. In fact, er … look at Google itself.

The future isn’t what it used to be

So fine, you say. But why the relentless emphasis on infrastructure issues? Won’t all this messy economics go away? Please?

Not if you want to be taken seriously.

Let’s acknowledge what the Web has been successful at: as a presentation layer. But the Web 2.0 kids desperately want to write system apps on their “global operating system” – only they don’t have the cojones to do system level thinking. Real engineers look at where systems (and humans) fail – their priority isn’t a cool demo. They’re pessimistic. And there’s no place for pessimism at a Web 2.0 conference.

For example, look at how many of these new services will depend on Flickr. Flickr is a joy indeed, but it spends more time in a paralytic state than a Tenderloin wino. If as much effort went into keeping the Flickr system running as goes into writing cute messages explaining why it’s not working , we’d feel a lot more confident in trusting it.

“Let me just lean against this angle bracket” – a web service takes a breather

Just as you can’t build a house on sand, you can’t build “a global operating system” based on a presentation layer and a few scripting kludges. Or maybe you can, but you’ll need a different kind of engineer to build it, one with a different set of values. The Bubble 2.0 crew have some nice ideas, but by that yardstick, they don’t measure up.

In addition, technology types are often in such a hurry to replace one evil overlord, they simply usher in another.

The future isn’t too hard to guess, so long as you don’t ignore the things that made the past and the present so crap. Utopians are notoriously bad at this, but I guess that’s the starting point for defining a utopian.

So why be so optimistic, then? Partly because of the composition of Web 2.0 itself. With keynote speakers such as dotcom analyst Mary Meeker, it fits Disraeli’s description of the Whig opposition as a “range of exhausted volcanos”. And what else would you expect from presentation layer people Thinking Big but the superficial and pretentious? Cascading Style Sheets as a transmission protocol. Really?

The same tune has been heard before, from the same place, and it was found wanting the first time round. Without even articulating the feeling explicitly much of the time (although it’s often expressed as inchoate computer rage) – people want different problems to be solved. We no longer look to presentation layer people to fix infrastructure problems.

Even more refreshingly, people don’t even look to technology to fix problems that don’t exist.

So how good and bad do you think it will get?