Telepocalypse

OPINION://Open and shut case

Having spent too much time in airports and on planes again recently, I’ve been pondering the open-vs.-closed issue for networks.

We often get confused by the words we use, or the lack of them. For example, is the word “Skype” denoting a company, a product, a verb, a network, a protocol, or all of the above? Similar considerations led me earlier to coin the word “Vob” to describe the network of provisionless open SIP-speaking telephony nodes in distinction from the SIP protocol itself. (For instance, just because it’s SIP doesn’t mean it’s telephony or the Vob. SIP is used for other stuff besides the Vob.)

Sometimes the network is merely an emergent phenomenon, such as the Internet. It has been argued that the term “Internet” is really a synonym for the transitive web of social agreements to transit IP packets, as opposed to any concrete thing. We could change the protocol (and with IPv6, we would) whilst retaining the social agreement, and it’s still the Internet. (Thus the definig attribute of “Internetness” can’t be determined by solely technical observations — you have to look at the social and economic agreement to see if something is really “on the Internet”. If your ISP’s terms prevent you from acting as a transit, you’re only an associate member.)

Other protocols, such as SMTP, give rise to similar confusion. Are you really “doing email” when you send a message inside your corporation using a protocol other than SMTP? And what if you aren’t able to relay that message to all the rest of the world? We don’t have a word to describe the “SMTPNet”, so people don’t really notice when they’re only being given partial access to it.

Is there an open network of blogs (such as this one) able to accept trackback pings? The Trackbackosphere?

Ted Shelton spots the idea of a storage-centric network. There is an implicit network of devices that are capable of transmitting and playing tracks purchased through iTunes. This is a closed network. We don’t have a name for it, although we intuitively understand it.

I recently saw similar confusion over IMS. For the uninitiated, this is the carriers getting together to plan a universe of smart network elements upon which you will be induced to become hopelessly dependent and pay eternally with your soul and wallet. But IMS is just a technology; what’s the network? Will there be one IMSNET? Kindasortof, since the spec talks about foreign agents and roaming. But won’t the carriers want to differentiate their service offerings? So will there be lots of little IMS-speaking networks with local and mutually incomprehensible dialects? At least we know whatever the outcome, it’ll be closed.

(As an aside, the option value of an IMS network seems very low, whereas an IP network is practically unbounded except by capacity. Lord knows what the carriers plan to do to make this vision compelling — or compulsory — to their customers.)

BitTorrent is a protocol, although you often see people get confused and think it’s a network. There are as many BitTorrent networks as their are tracker sites. Most are “semi-open” — anyone can become a client, although the content of the server may be restricted. I’m sure there are plenty of darknet torrent nodes too, which are completely closed. Napster was very (too!) open; KaZaA is pretty open still.

So, what determines whether we get an open or closed network architecture? I’m not sure, and don’t have any research, but I do have some hypotheses. I suspect that there is no simple answer, but rather a basket of competing forces.

There are (at least) two forces pushing towards a closed architecture.

Firstly there is the need to preserve the integrity of the network. For Apple, “integrity” means copyright control. For Skype it means controlling telephone spam. For KaZaA it means battling poisoned content distributed by the RIAA and MPAA. The deciding factor seems to be a “before/after” switch: is the damage done as soon as the “bad” message is received, or can the badness be filtered out later? If I can copy the iTunes file, the damage is done; too late to verify if it has remained in (RIAA-)safe hands. The interruption can’t be undone. The poisoned file can maybe be filtered out later, hence the permeable architecture of most file-sharing sites. E-mail can use after-the-event filtering based on message content; telephony cannot.

The second factor is creating a consistency of experience. If you control all the nodes, you can assure senders of the experience that receivers will enjoy. In Skype, for example, you know how the ringing experience will unfold, what sort of “caller ID” they’ll see, and what filtering rules everyone has access to. That didn’t exist on the Vob. [After I wrote this, but before I posted, I saw a similar sentiment on Jeff Pulver’s blog. But fixing compatibility and consistency doesn’t automatically give the Vob parity with Skype in the user’s eyes, IMNVHO.]

Pushing towards openness is the option value of the network — its “stupidity”. SIP is a kinda crazy hybrid. It’s a flexible technology. But it’s hardly what you would call a “stupid network” approach. Just take a look at the abstract from the RFC.

SIP makes use of elements called proxy servers to help route requests to the user’s current location, authenticate and authorize users for services, implement provider call-routing policies, and provide features to users. SIP also provides a registration function that allows users to upload their current locations for use by proxy servers.

Lots of tightly coupled functions in cental nodes, with some stateful sauce sprinkled on top for added flavour. (Has anyone else noticed that you can get 50% of the benefit of SIP for 2% of the cost using dynamic DNS?) Whilst the remaining flexibility has had an unfortunate effect of creating chaos for users, it still makes it quite a useful toolkit to plunder for building your closed network. The temptation to fix a few of the wrinkles and holes will result in a proprietary dialect each time.

Skype’s basic framework is only as flexible as Skype’s internal coding team. But the API enables localised customisation without compromising the integrity of the core network. SMTP enabled the emergence of hundreds of types of clients and dozens of mail servers; if it hadn’t been so flexible and open, it wouldn’t have succeeded.

Completing the picture is the level of market power the owner of a closed network has over its users; the amount of tribute demanded as a consequence; and the availability of viable alternatives.

I’m pretty sure I’ve not captured the whole picture here. I’d be interested to hear reader feedback on what drives networks to be open or closed, or references to academic study on the matter.

Hmmm … I wonder what’s the future of openness and mobile operators? I rather liked this quote on the matter, if in a somewhat restricted context:

I believe in the end open distribution models will win. In a way mobile operators are like Michael Gorbachev in the Soviet Union in 1988: they are already preparing to promote perestroika in an effort to control the explosive market forces. Eventually even that will fail because the consumers will vote with their wallets. With some luck we might even see the iTunes phone one day, ‘Carrier willing’.

Ah, the irony of it. The carriers will be forced to open so we can enjoy our closed iTunes and Skype networks.

Et tu, Brute?

Oh, and the “SIP is dead” bit.

The Vob (a provisionless SIP-speaking voice network) is dead. Not even sleeping. SIP will be embedded inside other virtual networks (e.g. IMS, aka “TDM and SS7 on IP”). But they will be overlayed with a ton of proprietary and closed stuff that will make any SIP embedded somewhat irrelevant.

Think of the Skype API (and the Yahoo API, and MSN API, and…) as being like Visual Basic controls. By constraining the set of possible things you could do, you add value to the developer. In VB you don’t call generaly call Win32 APIs directly. The virtual networks will also overlay SIP in ways that make it invisible.

SIP is dead in the same way Latin is dead. The influence is clear, but you won’t hear it spoken on the street.

UPDATE: When people say “SIP” they often mean “The Vob” rather than RFC3261 (or whatever); but because there isn’t an equivalent of “The Web” for SIP they don’t have the language to say it.

UPDATE: The Vob has done what most open source projects do: create an imitation of the proprietary expensive earlier thing. In this case, the PSTN. But DOH! The point of the stupid network was to do crazy new things and send unimagined new messages. As a facsimile of the old thing, the Vob is fairly pointless. And if you don’t have an future equivalent of the “Web” for voice — a single virtual network, then SIP doesn’t have much of an audience. It just becomes a hidden bit in some proprietary virtual network owner’s developer kit.

UPDATE: Or, to ramble even more and eliminate all possibility of succinctness, SIP’s image is irredeemably hitched to the Vob, and the Vob is no more. As a vision of an interoperability standard, it’s extinct. As a convenient short-cut to building a private virtual network, it has its uses, but nobody will know or care.

UPDATE: Looks like Richard has put 1+1 together; if the Vob is dead, and public ENUM is a way to route SIP calls on the Vob, then all SIP and ENUM have become are toolkits to build private proprietary virtual networks with. Now all you have to do is pin the tail on the donkey and decide which private virtual network(s) to put your money on. Cheap-and-nimble Skype; one of the big-three IM networks; the telcos with IMS; or something else from leftfield.

Perfectly and properly proprietary

I like a controversy, because that’s when we learn things.

In the ongoing debate I think we’ve lost sight of the original issue: separation of connectivity from applications. I’m a great proponent of this. Monopolies and markets above this in the stack concern me a lot less.

Skype’s network is proprietary, closed, and — yes — potentially downright dangerous in the long term. But I believe in dealing with reality as we find it, rather than an ideal world we wish we could conjure up. Pedant’s note: the Skype API opens the client UI, not the Skype network.

So far Skype isn’t particularly wedded to connectivity provision (with one exception). That means we don’t have to seek permission to use something else. Skype is a child of the Stupid Network, and self-centered as it may be we should be happy about this precocious toddler.

Enough of the B movie argument. Time of the main billing.

Being proprietary is no sin as long as the user is happy.

I admire Microsoft, and think most people whinge too much about Bill & co’s gazilions. People have forgotten how much word processors and operating systems used to cost, and how painful it was shopping for them. In the old days of the 80s and early 90s, Microsoft took expensive software and made cheaper mass-market versions. That’s a good thing.

When you buy MS Office, you’re not just buying a word processor. You’re buying the assurance that you can exchange editable documents with virtually any other business user. That’s a huge thing. With Windows, you’re buying the value of being able to employ almost anyone and know you don’t have to send them on a training course to understand what left-click and right-click might do. Microsoft’s officers and shareholders have only received a tiny sliver of the (without exaggeration) several trillions of dollars of value their standards have created.

The pattern is quite well-established. Beyond the obvious Word and Windows, SQL Server and Great Plains are more recent examples of how Microsoft has attempted to nibble away at the underbelly of my former-former employer, Oracle, by lowering prices and increasing volume. Microsoft’s products are great value for money.

I also admire Oracle. Proprietary? You betcha. But people haven’t been unloading their treasure onto Larry for nothing. Oracle does something very useful, and people are getting more value out of it than the price they pay, otherwise the revenue flow would stop. Open source alternatives, depite the hope and hype, have only nibbled at the fringes of the business. Oracle provide you with an assurance that your data will continue to be accessible for years to come, through many upgrade cycles of hardware, storage and OS. A vague hope that some voluntary collective (or tiny corporation with an experimental business model) will keep up the good work isn’t very reassuring in comparison.

I admire Skype. Predicable? Yessir - that’s me! It is successful because it solves the user’s problem. And that problem is a lot more than getting someone’s current IP address and creating a session and duplex audio channel. When you access Skype, it just works. (Err… ah. Except I’m currently Skypeless because it refuses to install the latest upgrade. Err. Um. No matter. Ignore the man behind the curtain.)

With Skype, there is no cognitive effort about having to purchase or provision the software. You can recommend it to friends without having to worry about them acquiring an incompatible version. Skype spreads because it does what the users want. It’s a cliche to say that people buy solutions to problems, not technology. Skype’s success suggests that those proffering alternatives failed to understand and solve the user’s problem. Some humility might be in order, not indignation.

That means there was a branding or marketing problem that had to be solved. And probably a usability one. Oh, and a compatibility one. And a nationalisation one. And a commercial one. Get the picture?

SIP does (almost) exactly what it says on the tin: it initiates (and tears down) sessions. No more, no less. The standard says nothing about the semantics of those sessions, or about stuff outside of the session protocol.

For example, one essential ingredient of a personal communications system is a means of limiting inbound calls on your attention. For this we have buddy lists and protocols for asking to join other peoples’ lists. Skype unifies the semantics of this: you know exactly what the other person’s experience will be, and you know it will work. (Anyone responding “XMPP/Jabber” will be given a good slap and asked to re-read this section: the absence of a unifying client means the semantics are not well-defined at the user level because you don’t know how the message will be consumed and presented at the other end; only the syntax and semantics of the machine-to-machine protocol. Machines != people.)

Skype has merely embraced and extended SIP inside a proprietary wrapper in order to solve a wider bunch of user problems. So does being a Skypehead make you the new Bellhead? Yes, but with the vital consideration that the end-to-end principle isn’t violated by Skype. Will be be getting the bill for Skype in five or ten years from now, just like we pay $60 to Bill G. when we buy a $300 PC in Wal-Mart. Possibly. But you’ll have banked a lot more value in the interim.

I wish I’d bought Microsoft stock early on, but my mind was poisoned against it by the horrors of FAR PASCAL pointers and the ugliness of Windows compared to the elegance of Unix. I’d now be a richer man if I’d seen the bigger picture.

I’m glad I worked for Oracle and got plenty of stock grants. I did very nicely out of it, thank you.

If Skype does an IPO, I’ll be calling my broker.

Edging forwards

As a little diversion, I’d like to just expand on the “moving beyond the old phone network” theme. Is Skype and its ilk really making use of the stupid network, or does the power of VoIP just derive from arbitrage of telco pricing?

Purely as an example, let’s take Phil Wolff’s recent wishlist for Skype voicemail features. This is not meant to be a critique or discussion of the feature requests. I am solely using this as fodder to gauge how “stupid” our feature requests are, and the list is conveniently long enough to make a good sample size.

Below is the list, somewhat edited and abridged. In each case I’ve classified the feature into one of four buckets:

• This can be done on the PSTN/AIN without any major compromises (“Equivalence”)
• This feature can be implemented on the PSTN using a voice/touchtone UI. The experience may be significantly inferior to a multimodal one. (“Inferior”)
• This is a feature that can be done on the PSTN, but requires or benefits significantly from an off-line Web UI. Vonage’s control center is an example of this: you don’t use it during calls, but use it to configure and provision service. (“Hybrid”)
• This feature can’t be implemented on the PSTN in any fashion and requires an integrated multimodal client. This is because it requires a data display or input not feasible via a voice UI. (“Impossible”)

I use the abbreviation VM for voicemail. You may disagree with some of the individual classifications. It doesn’t make much difference to the overall picture.

So, what are the scores on the doors?

Now, if we were really serious about moving beyond the PSTN, there would be a lot more “impossible [for the PSTN]” feature requests. Methinks there’s still a lot of room for the re-invention of this space, and thinking is very much constrained by how we’ve done things in the past.

On the other hand...

It’s election time here in the UK, and the media are required to give a “balanced picture” by offering equal time to the main parties.

So in case you think I’ve gone completely Skype nuts, here’s a few balancing news items.

I had a Skype conference call with a client a few days ago. It was embarassingly bad, mainly because one of the participants was on a flaky hotel Wi-Fi link. In the vertically integrated telco model, this would be the telcos problem. Skype’s layered model tends to push them towards a “not our problem” approach. Mistake! The Skype client needs to be aware of the connectivity quality underneath it. If it can’t deliver, and is disrupting a conf call, then it needs to inform the participant, end the call, or somehow manage the situation. Skype needs to manage the user experience better, even the bits they don’t have control over.

I’m also having too many Skype calls where I need to re-dial to get a better connection. Plus I’ve had several failed SkypeOut calls (what is a 10040 error, or whatever it was, I don’t know — but that’s not my idea of a great user experience).

Skype also has some blind spots and weaknesses. It’s a winner so far at consumer-to-consumer chat, both for close buddies and strangers. Yet there’s space for Google to build a C2B VoIP service (“click here to talk to Acme Rotavtor Supplies”), Microsoft to federate their RTC/Outlook universe (B2B2B2B^n), and the PSTN to soldier on as the B2C preferred means of contact. Skype’s assumption that all calling is free doesn’t jive well with what we’re seeing elsewhere.

That said, some of Skype’s competition don’t get that “great user experience” idea. Yahoo have some dreadful advertising pop-up in their IM client, and you have to hunt for a configuration setting to turn it off. They litter their IM client with irrelevant crap. They force-feed users on upgrades to try to sell them BT’s VoIP services. And the webcam experience in Yahoo — and its ability to deal with firewalls and NAT — sucks like grandma’s lost her dentures. Last time I tried their voice client, it wasn’t good. Skype does things like automatic volume control very well.

In our world of toothpase, Yahoo is the cap that rolls off into the sink, the tube that splits in your toiletries bag, and the dispenser that squirts a bit too much out when you squeeze the end. Skype has the perfect flip-cap, a tube that stands on its end, and even tastes good.

I’d also like to clarify one thing… I said:

Anyone who thinks they can roll a VoIP strategy without taking Skype into account has lost the plot.

What I mean is that you need to take it into account, not necessarily automatically embrace it. If you’re a telco, it means either fighting or co-opting Skype in some way. (What to know how - click here.)

Skype isn’t perfect. It hasn’t taken the feature set very far past the PSTN. (What can you do in Skype but can’t do with an all-you-can-eat PSTN plan and an IM client?) But the bar anyone else needs to cross is raising out of reach real fast. You’re not only going to have to outflank Skype in product features and usability, but you’re going to have to outdistribute Skype when Skype has an army of tens of millions of pushers already in its grasp.

No matter how much you love SIP and ENUM, there’s no “Perfect for the VoB” sticker on a box in Costco that assures you your SIP phone will interoperate with the rest of the universe, and that it’ll fulfill your full set of needs (voicemail, call control, etc.). The Vob is a marketing failure, not a technology one. The “promise” isn’t clear and the brand non-existent. Compare to WiFi, whose ascent is as much an accident of cute naming as meeting customer need. Also score another similarity between Skype and Microsoft — the genius of both is not the technology, it’s the sales and marketing.

As far as I’m concerned, Skype’s well past the Mum Test — my mother is now getting her friends onto Skype. If you dismiss or ignore Skype, flawed as it may be, you do it at your peril.

Well, I'll be darned

The twitch. The urge. The craving. You can’t resist. No hope.

I got home from my first term at university at Christmas in 1989, and I immediately set to work. Two days later and a marathon coding session, I had my own hand-crafted version of Tetris. Pure machine code, full colour, absolute addiction. I had become my own drug supplier, no longer dependent on the erratic availability of monochrome PCs in the college library.

So it’s nice to know that the mobile operators have spent hundreds of millions of dollars on virtual vending machines and hundreds of billions of dollars on networks to deliver that content, just to discover this.

OPINION://The telecom earthquake

I use the “subscribe to a search” feature in Bloglines to turn up nuggets of interest.

Searching for “disintermediation” comes up with an article or two every day. Many of them are mine ;)

Searching for “Skype” hits the 200-article limit every couple of hours.

Anyone who thinks they can roll a VoIP strategy without taking Skype into account has lost the plot.

Some more Skype musings, since I’ve got Skype on the brain at the moment…

Skype reminds me of a consumer packaged good company. It’s more like Coca-Cola or Procter & Gamble than a telco. Their strategic differentiator is their sales and distribution method. Glugging another coke from the machine by the elevators is easier than hunting down a clean mug and boiling a kettle for coffee. P&G work hard to get their toothpaste on just the right shelf at Wal-Mart and 7-Eleven. Anyone can make sugar water and peppermint-flavoured microabrasives. Few people understand how to market and sell them.

Skype is about convenience, and putting new features within easy reach; leading users through small incremental advances in how they communicate.

You can also view Skype in two very different ways. A recent Analysys report talks about the class of “Private Voice Applications”, of which Skype is an example. This is a node-centric view of the world — Skype the PC software application. But looking through the other end of the telescope, you can view Skype as a virtual network — just at the application layer, divorced from transport.

This takes us to a more familiar realm. Suddenly we start to see all those familiar network-centric terms and issues crop up — interconnect, termination charges, roaming, vertical integration — and have some more clues as to how Skype might gain market power and their business model evolve. The word “Skype” has this fuzzy definition that blurs the corporate, product and network identities. Just because we don’t have a snappy buzzword for the Skype Network doesn’t mean it doesn’t exist.

In a similar vein, we lack another word in our vocabulary. A decade ago the Web burst upon us. HTTP as the protocol; Netscape as the user interface; the Web as the network of public HTTP-speaking nodes. SIP is an analog to HTTP. Softphones are similar to Netscape. But we don’t have a word to describe the virtual network of public SIP-speaking nodes. If Mosaic and Netscape spawned the hypertext Web, the equivalent today would be the voice “Vob”.

But the Vob is a failure to date. The number of open SIP nodes addressable via ENUM (or otherwise) is miniscule compared to the proprietary Skype virtual network. SIP has been absorbed by the telecom borg. At Sprint I watched Dynamicsoft struggle for a year to get a SIP push-to-talk engine to work. (Flarion demo’d a PTT app they whipped together in 2 days using standard Microsoft development tools, because they didn’t need to bleed low-level optimisation across various layers of the stack.) Sprint is still probably working on pointless faux-smart network reincarnations on IP and SIP. Next-Generation Voice Network? You must be kidding. Last Generation Voice Network, maybe.

SIP is history as far as the future of voice is concerned. Get over it. The Vob is dead. Shuffled off to join the big pile of dead over-complex standards. DCE, CORBA — please make some space between you, we’ve got company tonight.

Skype isn’t this decade’s Netscape either, because it isn’t substitutable; they own their private Vob — they’ve achieved what Microsoft failed to do with MSN and the Web. Yet they retain the strategic power Netscape had to reverse themselves from the client into various centralised server functions. (Yes my dears, even in P2P voice there’s a pile of trust, directory and routing stuff that some people are going to want to do behind private barricades.)

It would be a tragic mistake to underestimate the potential market power Skype is accumulating. According to Skype’s own figures from VON Canada, they’re sustaining a growth rate of 1000% a year. Just another 2 years of this growth and they would have over 200 million concurrent users online. This is not beyond plausibility given how Skype and broadband are symbiotically driving adoption of one-another; the addressable market is exploding too.

That means even if you’re a mega-telco — a Verizon or a Vodafone — you’re screwed. You can create your own Private Voice Application, and start marketing it to your early-adopter users, but who ya gonna call? Ain’t nobody but Skypers out there. Want some Skype presence in your Vodafone-branded VoIP app? Gonna cost ya!

I almost can’t believe what I’m seeing unfold in front of my eyes. This time last year I was being rapped over the knuckles by corporate security for running Skype inside the Sprint corporate network. Somehow, I don’t think they’ve got the joke yet.

Feeling viral

More contagious than avian flu, less lethal than ebola. Viral Skype marketing.

Rant of the day

Telcos today make most of their money from voice-centric communications, and various VoIP reincarnations are in play too. Will the telcos be able to shift to presence-centric systems that take advantage of the flexibility of the stupid network?

Well, I’m currently experiencing a very sad non-telco example of how even the best of old companies can fail to transition to a digital world. I’ll share it with you, if you don’t mind.

For many years I’ve been a loyal Economist subscriber. When they launched the web site they gave free access to print subscribers. Over time I’ve found myself reading ever more on-line, and having less time for the print version. Apart from anything I’m often away from home on work or pleasure, and they just pile up and become too much to catch up with.

So as my subscription was up for renewal I thought I’d stick their RSS feed into Bloglines and give up the print edition in favour of the online version.

Oh, woe is me.

There’s just lots of details that are wrong. Today is publication day when the paper presses roll. So I get splatted with 80 articles in my RSS feed in one go. There’s no categorisation. No personalisation. The summaries are often a bit too skimpy to know whether I should read the article. No full feeds of articles — not even one ones published in full for free on the Web. No embedded links to related articles — in-house or otherwise.

So the Economist’s editorial role isn’t making as good a transition to the Web as I’d like. Newsgathering and comment isn’t enough on its own to meet the customer’s need. (Hah! SIP and ENUM on their own aren’t enough…)

But holy cow they’ve screwed up one thing. I’m looking at a serious article and there’s a Flash-based ad for Xerox on the right with a graphic of a sweeping sonar. And get this — it pings aloud three times about every ten seconds. Repeatedly. Forever. All I can say to the Economist is

WHAT THE **** ARE YOU THINKING?!?!?!

How dare you treat a paying subscriber this way. It’s just totally offensive. It tells me that I’m being treated as eyeball fodder for some crappy marketing schtick, and you don’t see excellence in journalism as your reason for being.

Your values have been corrupted. Fix them.

Now I hope they can pull this round. I don’t see dead trees being the medium of choice for anyone much in 5-10 years from now. I’d like to see the Economist prosper.

Poor distribution technology still leaves hard reporage and editorial tasks. Someone’s got to do them, ‘cos I’m willing to pay for them.

But all traditional telephony has to offer is distribution of your voice over a distance. There isn’t anything else behind the curtain. Get your distribution elsewhere and there’s nothing left to disintermediate. And the values of the telco — wring the customer dry — don’t fit the new world.

I reckon the last ever phone call will be from a telemarketer.

And it’ll be to the wrong number.

Prosciutto business models

I’ve several times toyed with the idea of setting up a “Disintermediation Daily” e-mail newsletter to gather a massive list of email addresses to pimp my consulting services to keep people informed of vertically integrated business models unraveling and turning into focused horizontal layers.

Anyhow, I’m too lazy. But here’s a good example of disintermediation anyway — today’s thin and spicy business model. A lot of ideas I get about telecom come from outside of the industry. This one goes to the heart of the matchmaking relationship in banking between lenders and borrowers.

Zopa is a UK start-up that could be considered an “unbank”. They aggregate prospective borrowers into buckets of people with similar credit ratings. A typical grouping might have 50 people. Lenders then directly fund these mini-syndicates, accepting the diluted risk of non-payment, and receiving a pre-negotiated interest rate. Zopa deal with collections and enforcement. The middleman also preserves the privacy of lenders and borrowers.

Zopa lives off a 1% commission on the amount the borrower takes. Unlike a bank, they don’t have an interest rate spread. They also don’t have an expensive distribution network; it’s all on-line.

What’s really interesting is how they’ll sponsor larger lenders who exceed the legal limits for “personal” lending to get a credit license. Thus the barrier to entry from regulation the banks have erected becomes much less of an issue.

You can imagine very similar structures emerging in telecom with new intermediaries brokering those offering and requiring connectivity, and dealing with the regulatorium. Some of the Wi-Fi aggregators have headed down this route, but they didn’t create the right mix of costs and incentives. They burdened themselves with network build-outs (vertical integrated business model) and high access fees, rather than seeding a marketplace where many supply and retail models could compete. But a Zopa for connectivity will emerge somewhere, it’s just a matter of time.

Power glitches

The most optimally superior OfcomWatch has a pair of articles on powerline broadband. (You US-based readers really need an FCC watch blog as good as this … over in Britain our telecom industry might be nothing special, and the regulator pretty average, but we’re world-class at analysing and bitching about it.)

It occurs to me that “broadband” over powerline misses the point somewhat. Reach always beats speed in telecom. Any signal is better than no signal. The point about powerline-based IP communications should be pervasiveness, not speed. You could do amazing things with a service of even a few hundred bits per second that was guaranteeably absolutely everywhere in a city. Think of it as being the signal rather than media bearer channel, if you’re inclined that way. Most devices attached to a power socket aren’t Net-enabled today. There’s an order of magnitude more of them than there are PCs. Lower the power, drop the bitrate, and your electromagnetic pollution problem eases up. Solve an unsolved problem, rather than flail at a solved one.

Vandalism for fun and profit

Two weeks ago I flew back from the USA. One of the privileges of a nomadic lifestyle is sitting at the end of the runway at Boston Logan airport for half an hour for mysterious reasons not explained by the captain, and then taking 90 minutes to fly the 191 miles to Newark. I’m trying to decide whether the pilot was lost, or just wanted to point out his mum’s house in deepest New Jersey to the co-pilot. We certainly circled one field and awful lot of times.

Anyhow, it gave me plenty of time to read the Continental in-flight magazine, and as a courtesy to following passengers I ripped out a few pages and stuffed them in my pocket. Here goes.

I just love this Ye Olde Telecom approach. We’ve looked at every possible type of message we think you ought to be able to send over the network and priced the bits accordingly. Sorry if your desired protocol isn’t one of the ones we’ve created a gateway for.

Also, has it occurred to Verizon that offering a, say, “$20 and call anywhere in the world as much as you want” plan might generate a lot more business in a fixed-cost environment? Would anyone dare spend ten bucks just to find their contact’s phone is going to voicemail?

And why on earth vary the rate depending on which bit of land or sea the plane happens to be over (no pun intended)? You might as well double the price for people with surnames beginning with vowels — it has just as much correlation with the value received by the customer. It isn’t like there’s a substitute product that varies with geographical position you’re competing against.

Now for some more fun. This is a really messed-up industry. Exhibit A, Gorilla Mobile:

Now termination and origination fee avoidance has been a preoccupation for competing telcos for decades. It’s just now we’re seeing it at the retail level. Pay us a small fee, and we’ll create an audio channel to your handset via the least-cost manner, and then interconnect you on cheaply to where you really want to go.

Exhibit B, Talk Parade:

Nothing exciting, I grant you. Just a softphone and VoIP-to-PSTN bridging service. But look at the positioning — “Laptop … $4.99 … FREE … Stop paying those high cell phone bills”. Ever get the feeling that they’re competing on price?

Now what’s totally screwy is that we’re seeing full-page ads for services whose sole purpose is disintermediation of the pricing structure of incumbent industry players. No new value, no new features, no unique convenience, no innovative distribution or pricing or sales method. Just arbitrage plays.

I’m glad they’re doing it: as David Isenberg said at F2C, it’s the public’s trillion dollar dividend from the dissolutin of the smart network. Yet these ads are all symptoms of a deep sickness in the teleconomy. Meeting customer needs plays a weak second to pricing games. That’s unsustainable in the logn run. Just enjoy the dividend cheques while they last.

If in doubt, plagiarise

Edited email transcript between me and my brother today about his journey to work at a client site…

—BEGIN—

New thing for today: Skype from a train at 100 mph. It’s a good job that the networking world is built on QoS, otherwise, sharing the same link with everyone else on the train, the sound could have been terrible. Oh, hold it a sec, …

WiFi on intercity trains is great. Now, the reason Skype rules on trains is because of the following two scenarios:

(A) You call your colleague on your mobile phone. Train enters tunnel, you and colleague can no longer talk. You come out of tunnel and seconds afterwards, you lose your connection. WHAT! I’ve just come out of the tunnel. You then have to wait for 15 seconds before the signal indicator on your phone seems to indicate it’s back on the network. You call back … until the next tunnel. Then you get pissed off and text “I’ll call you when I’m in the office”.

or…

(B) You remembered to bring your laptop and headset, so you can use Skype. You call your colleague. You enter a tunnel and keep talking because the WiFi signal is still there. Even when the connection rate becomes poor, you don’t get cut off; the sound quality just degrades a bit, but picks up again immediately when the throughput is better.

All this and simultaneously you can hold a chat session with your ex-colleague who is maintaining some databases in Perth, Australia on the other side of the planet and (slowly) replicating your e-mail and reading your Bloglines RSS feeds.

I like scenario B better. Makes the first class ticket worth it. Because of my extra 3 hours of work on the train from Newcastle to London, I’ll go away at 5pm today. Travelling didn’t suck after all. As British Rail said in the 1980’s, “We’re getting there”.

—END—

Scaling up

I was reading Andrew Odlyzko and Benjamin Tilly’s new paper on network scaling on the plane to the US last week. In brief, it challenges the widely-repeated (even my me!) laws-of-thumb for network scaling:

• Sarnoff’s Law (linear increase in network value for broadcast networks with # of nodes);
• Metcalf’s Law (n-squared rate for point-to-point networks); and
• Reed’s Law (exponential rate for group-forming subnetworks).

In their place they propose a nlog(n) scaling property, and have some empirical and anecdotal evidence to support it. It also starts to capture the idea that most possible connections (or options to connect) have negligible or negative value (as noted previously).

The crux of the paper is the implication that this has for peering, as greatly more benefit accrues to the smaller partner when two dissimilarly sized networks join together. This is an important observation that holds up under many circumstances even if the nlog(n) formula is wrong. It means that the big guys get to charge the small guys for interconnect. Yes, we sort-of knew this already; but the insight is that you can’t have asymmetric charging and simultaneously have the above 3 laws hold true. Something has to give.

Much as I admire the contribution of this paper, it somehow feels incomplete.

Firstly, it doesn’t account for differences between broadcast, P2P, and group forming networks. Surely there should be some influence from the nature of the network on its scaling properties?

In a related fashion, it says nothing about the rate of change of adoption or growth of networks (as opposed to absolute, static value). Do group-forming networks grow faster? Both the traditional “laws” and the proposed new one just feel like they’re operating at too simplistic a level. I suspect that the effect of eliminating “membranes” in networks that constrain connectivity is a dynamic process that is better modelled in the first or second order derivative. Networks mostly grow via little local explosions of adoption. Maybe different rules apply for organic vs. orchestrated growth?

Lastly, and more importantly, it continues to confuse the value of a network to its owners with that of its users. It there’s one lesson of the Stupid Network, it’s that there’s a massive increase in consumer surplus. The value to users diverges from that to owners. You can’t measure user value by looking at industry revenue.

Consider when two networks merge under some roaming, interconnect or collaboration agreement. Now the customer has a choice of two places to buy service for the merged whole. (The case of fixed network access and no choice is the exception, not the rule; most networks are mobile or virtual.) This erodes pricing power. Whether or when that pricing power erodes quicker than new chargeable value is created, I don’t know. This supplier choice aspect isn’t in the model.

I have to admit I’m pretty unread in the economics of networks and peering, but it looks like there’s a lot of work still to be done. I look forward to seeing a more definitive paper on the matter.

Game theory

Bob Frankston worries about content and connectivity bundling by Comcast and Disney.

I wonder… what if the various commons movements acted to create and enforce reciprocal responsibilities as part of their condition of use? As a Comcast subscriber, you would simply see your attempts to access Creative Commons on GPL licensed content blocked.

“We note you have attempted to access this content from a parasitical ISP. Due to our policy of reciprocal obligation for open access to connectivity and service, you will be required to contribute $10 to the Electronic Frontier Foundation before proceeding.”

Various other socially-minded web sites like BoingBoing could do likewise. “As a Verizon customer, you are now required to view 10 pages of Flash-ridden adverts before viewing the article. If you do not like this requirement, please call Verizon and discuss their e-mail blocking policies with them.”

The danger, of course, is that this could create a mutually-reinforcing fragmentation of the Internet, rather than the converse that we seek. I’m not a good enough game theorist to construct the right incentives to escape such a Prisoners’ Dilemma. BTW, where does the apostrophe go? Google doesn’t tell.

One of the weaknesses of the Internet and IP was that nobody thought through the social and economic sides of a standard means of interconnecting networks. Where’s the General Peering Agreement, a viral means of ensuring open access?

Absolutely relative

James Enck points to a DSL broadband deal offering 8Mbps download speed and a 500Gb monthly cap. Compared to the competition, this is generous. But let’s put these scarcity business models in perspective, shall we? A 500Gb cap on an 8Mbps download means you can still only operate your DSL line at solid full-tilt for less than 6 full days a month. Would you buy a house where the heating only worked on Tuesdays and Thursdays?

Perhaps these services would be better advertised using two numbers, peak and mean speed. For BT’s basic broadband package with a 1Gb cap, would you be so keen on a “broadband” product than only allowed you less than ½ a kilobyte per second average transfer through the month - a mere fraction of an unlimited dial-up account?

UPDATE: On the other hand, the wireline operators don’t hold a candle to the wireless ones at spreading scarcity. If I wanted to use my pre-pay handset as a Bluetooth modem, that 1Gb would only cost me, um, £7600 (about US$14000). Bargain!

OPINION://A connected freedom

It’s over a week since the F2C conference, and things have been pretty busy round here. But I’d still like to share a few of my observations.

Although we may not realise it yet, I think that identifying a “Freedom to Connect” as distinct from other freedoms is a major philosophical step forward. In a world prior to telecommunications and digital computers it was hard to conceive of information as being distinct from the medium in which it was carried. For instance, we don’t talk about freedom of “speech” by accident. By various legal contortions we’ve re-interpreted the word “speech” to include a wide array of things that are not speech at all under its usual colloquial definition.

Likewise publishing was attached to dead trees, religion to churches, and assembly to public places. Regulation and social norm referred to these physical artifacts rather than the underlying information concepts. You ban and burn books, not the words in them; you encourage chuch-going as a proxy for piety; and you outlaw or incite gatherings in named classes of locations.

Freedom to Connect re-frames this debate. Freedom of speech is cleaved into two; freedom to make a noise, and freedom to make particular noises expressing certain ideas. Playing a musical instrument, singing, talking — they’re all in the latter category, and none and be freely performed without the former freedom to make a noise.

Like any freedom, there are duties and responsibilities. Some speech, such as child pornography, is an accessory to a very physical and non-digital crime. We outlaw it as a form of speech. We also make regulations about noise as a separate rulemaking sphere. Crank up your stereo high enough and the neighbours will complain. But we only regulate noise to the extent that a third party’s ability to make and receive noise is unduly affected.

So as a guiding principle, Freedom to Connect decouples the noisemaking from the application-specific nature of that noise. This has very practical consequences. For instance, should underlay transmissions (in the frequency gaps) be allowed in UHF frequencies currently used by TV? Absolutely, if no third party is adversely affected; and this should be done without any regard to the current usage of that “noise medium”.

Only where third parties do suffer loss from the noisemaking per se does that application begin to matter and we contaminate our horizontal thinking with vertical reasoning. A town crier disturbing your lie-in to relay distant news might be judged a reasonable balance of public good and private loss; a call to buy goods in the market might not; and a rowdy street brawl definitely wouldn’t.

This Freedom to Connect — qualified by the absence of a noise-damaged third party — should be regarded as an inalienable right. Non-genotiable. This isn’t a matter of polite debate. Be angry. Rebel. Who the f*ck in some office hundreds or thousands of miles away has any business telling you what electromagnetic or sonic emissions you may emit when nobody else cares? I simply don’t accede to your authority.

This same standard doesn’t — indeed can not — apply at the application layer. This is where our abstract ideas, rather than physical bits, are expressed. Ideas can be recounted, relayed and spread. Bits CANNOT — networks create an ilusion of path continuity which has no physical embodiment. Each bit transmission is a separate event invoking its own Freedom to Connect. However, no expression of ideas is “value-neutral”, even if initially performed in a private space. Society has an interest in all idea expression, since some ideas are toxic and others deserve public support in their spread. Every idea expression has a potential for future third party harm or good, whereas once a bit has been sent and received without interfering with someone else’s bit, it’s all over.

Things get a bit grey when we move into “store and forward” applications. Are we invoking Freedom to Connect (F2C), or Right to Say (R2S)? What about IP networks (storage for nonseconds), P2P networks (hours), book warehouses (months)? Are these “conduits” for which F2C rights apply?

To me the test is whether the bits are forwarded “with due haste and without regard to their content or nature”. In other words, stupid linking of stupid physical conduits transitively results in stupid virtual conduit, and F2C applies. The Grokster case is classic F2C territory; the transmission technology can’t know in advance of its use. But BitTorrent node operators who knowingly offer copyright-infringing content are involved in the expression of ideas, and can’t hide behind F2C firewalls.

DRM, on the other hand, looks a lot more like a R2S issue. Certain ideas are encumbered by their originators by rules on their expression. As long as DRM only interposes on the use of their ideas, it’s not a F2C issue. On the other hand, if we’re no longer able to source devices to originate or consume our personal bit streams, we have a F2C problem. Forcing every screen, speaker and microphone to have DRM and imposing a huge upgrade cost is not reasonable. It’s criminal.

Let’s take this template and apply it to another sphere, municipal or community networks. Should they be allowed, banned, ignored, or what? Well, a pure bitpipe (wireless or fixed) is just a conduit for electrophotonic bitnoise. Deploying one may financially damage an incumbent of rival bitpipes, but that’s just tough luck. Inherent in networks is that they connect multiple points, and if a multitude of points secede from you en block and connect themselves autonomously, so be it.

Offering telephony service however, should by default be outlawed. (Exceptions to prove the rule: demonstrated persistent failure of the market to offer a service, and maybe in extremis rules for very small and isolated communities.) PSTN-style telephony is not a content-neutral form of bit transport. It is an expression of an idea of how certain types of communication should be performed, and it is clearly not the only way of connecting talking humans. By perpetuating PSTN telephony the municipality is doing direct harm to its citizens who might have an interest in newer forms of communication that require message-passing the PSTN cannot support, such as presence data, integrated chat, wideband audio codecs, etc. Same arguments apply to video and so on. (Concept hat tip: Bob Frankston).

A fairer approach would be to offer non-discriminatory wholesale access to the network, and let incumbent service providers roll their existing user base onto it. Of course, the fact that the users can source their video streams and voice services from anyone on the planet…

Municipalities can also be dangerous suppliers of connectivity. Beware of Freedom to Disconnect. What if a corrupt local official who objects to your speech decides to cut you off? Without a robust Freedom to Connect right, you’re in trouble. If there are essential application-layer resources they own (like everyone else only owns a PSTN phone), you’re in even more trouble. R2S depends on F2C. Beware of central control over the latter if you want to avoid central control over the former.

What if there is “noise damage” to a third party? Then we’re into a more subjective area of costs and benefits, and the political and judicial process has been dealing with these trade-offs for an eternity. But rather than competing markets for vertical integration, as one F2C panellist called for, I’d prefer competing markets for horizontal integration. In other words, a diversity of mechanisms for letting “noise” co-exist. Merely creating a monolithic property right to spectrum or street rights of way is just one of many such models.

I rather liked Andrew Schwartzman’s observation that anti-trust is not enough; democracy, citizenship, free expression are the bedrocks of society. That means market concentration metrics alone are not an appropriate measure of whether we are achieving our societal objectives in F2C.

Likewise, Bob Frankston noted that the hotizontal regulation vs anti-trust issue was badly framed. His thesis is that market competition can never work for bit pipes. It is the wrong model — we need utility model. The problems of the Louisiana folks suggested there were issues in financing and operating such a model. My suspicion is that reflects more on fuzziness of our map rather than whether there is a utility “there” there.

Freedom to Connect is important because it moves the debate on beyond economics and horizontal vs. vertical regulation disputes. You wouldn’t make up a foreign policy based only on impacts to your arms industry. So Freedom to Connect is about more than telco profitability. It has more in common with the civil rights movement than free market economics. When laws are passed saying you must seek permission of telcos before laying a glass strand down your street, you should disobey. Martin Luther King didn’t make his case on the basis that ending equal rights would be a great fillip to the economy. There are certain barriers and impediments that are simply unacceptable in a free society.

Freedom to Connect identifies, names and defines the critical social and political framework that enables an information-based society to work. It’s an idea that is going to be with us for a long time to come, one way or another.

Welcome to planet Zog

A most valued reader forwards the following link to me:

It might be big, but it will never be dumb

There is a myth in telecommunications companies today – partly promoted by the newly integrated IT guys – that the most economical way forward for networks is for the whole network to evolve into a “big dumb network”. […]

Well, just as there never is a free lunch, there will never be a big dumb network – at least not unless there are big dumb operators.

You don’t say! It only gets better…

A critical step for the immediate introduction of NGN services will be an open discussion on how much intelligence will actually be in the network – a big clever network – and where it should be located. Then, how this intelligence can be used to seamlessly support issues like NGN roaming, consistent inter-domain SIP handling and (inter-domain) QoS must be considered. These are time critical issues, as purchasing decisions for NGN equipment are now under way.

Ah! NGN roaming, what a beautiful concept. I’m sure I’ll create a great deal of end-user value. I guess all those Skype calls I’ve made around the world on weedy DSL and WiFi links — while running a BitTorrent client in the background and no QoS — were just an illusion.

Ever get the feeling these guys are having an oh shit moment, and realising that their smart network business just reached its use-by date?

PS — I’ll write about Freedom to Connect real soon, but I’ve had a bad head cold and awful jetlag for a few days, and real-world business needs to be sorted out first.