Category Archives: Opinion

Thoughts on computers and the IT industry.

Opinion

Why “unlimited” broadband must die

1 Comment

In a previous post about Internet Service Providers I hinted that their current business model — where it’s possible to download as much as you like when you like — was unsustainable. Since I’ve had a few people asking me about that I thought I’d dig into the subject.

The first question that we need to ask before we get any further is, What does your ISP do? This may look like a silly question at first glance but exactly what data they pass where is the key to understanding how they make (and lose) money.

Unfortunately the answer gets rather complicated very quickly so I’m going to make some assumptions. First I’m going to assume that you’re in the UK. Second, that you’re using ADSL and not cable or something more exotic. And thirdly I’ll assume that your ISP hasn’t “unbundled the local loop.” If you don’t know what this means ((If you do know what it means any want the hard numbers, have a look on the plus.net blog.)) you can be reasonably sure that the following is, if not correct, then fairly close to the truth.

Let’s look at what happened when you tried to download this web page.

  1. The request went from your computer to your router or modem.
  2. Your phone line is owned by BT, so they (and not your ISP) pick up the request.
  3. BT move your request around the country trying to find the nearest connection to your ISP.
  4. BT hands over the request to your ISP.
  5. Your ISP does not have direct access to every site on the Internet. Instead they probably pass your request onto another supplier. This is called “peering.”
  6. This other supplier actually ends up talking to the machine that this web site is hosted on.

The interesting thing about this is that the bit that your ISP directly controls is actually fairly small. Of course there’s much more to it than I have listed above and it’s clearly an important part, but fairly small nevertheless.

Your ISP pays both BT and the peer for their services. BT get paid both for the number of customers and for the amount of data flowing the over the network. And the peers also charge by the amount of data transferred.

Yet you most likely pay a fixed fee for your Internet connection.

There is nothing intrinsically wrong with a supplier receiving a fixed payment and paying a variable fee. This happens all the time. For example your mortgage may well be fixed rate even though the rate that your bank pays to borrow the same money is highly variable. And the wholesale price for gas and electricity varies on an hour-by-hour basis yet your tariff only changes once or twice a year.

However, two things have been happening in the consumer Internet connection world.

Firstly usage has been rocketing upwards. An increasing number of people are using high-bandwidth sites such as YouTube and the BBC iPlayer, and services such as movies downloads and iTunes ((Some people pin the blame on illegal downloads over peer-to-peer services. I deliberately left that off the list as typical bandwidth usage is drastically increasing even if you exclude infringing activities.)).

On the other side prices have been pushed down, with companies offering free broadband when taken with another service and other ISPs having to respond.

The cost of bandwidth does, of course, drop over time and ISPs are also consolidating which also reduces their overhead, yet I think it’s fair to say that usage is increasing faster than costs are dropping.

Okay, so that’s the problem. What is the solution?

An unrealistic solution, and one I see fairly frequently from the people that tend to download the most, is that the ISPs should just bite the bullet and provide the service that they promised. They said “unlimited” so they should mean unlimited. I have a certain sympathy for this argument, in the sense that it’s a problem that the ISPs have made for themselves. However it should be clear that the numbers just do not add up.

The ISPs are choosing a second option: either selling their users surfing habits to third-parties or billing popular websites for access to their customers. Both seem to be abusing their own customers, and yet there is still no guarantee that these new revenue streams will cover the increasing costs.

It seems to me that the only solution that would work on a large scale is to charge end-users for the bandwidth that they actually use ((Of course the most likely scenario would be “bundles” of bandwidth, much as you buy bundles of texts or minutes for your mobile phone.)). This may not be popular in the short-term, but it’s a whole lot better than them limiting what services you can use and when.

Opinion

So, you got a bad review?

1 Comment

Daniel Jalkut in his recent blog discusses a generally positive review of a useful Mac utility that closes with the suggestion that it “should be free.” The crux of his piece seems to be:

In short, if the product were free as in charity, would the product even exist, and be good enough to mention on MacBreak Weekly, where Leo could wish that it was free?

People have different motivations for making good software ((Jesper notes his reasons for offering his software for free. His argument makes complete sense but does not invalidate Jalkut’s complaint.)) but I think it’s fair to say that the most polished software usually has some form of income stream, whether that’s a licence fee, banner adverts or something less direct.

Of course one problem about selling software is piracy, but fortunately Brad Wardell wrote a great blog entry about just that and the effect that it has on his games company:

It’s irrelevant how many people will play your game (if you’re in the business of selling games that is). It’s only relevant how many people are likely to buy your game.

How, you might ask, is this connected with Jalkut’s argument? Well, the simple truth is that reviewers of your software are not paying customers. Their needs and desires and value judgements are not the same as yours ((I’m reminded of the comparative reviews of word processors that you got before there was little alternative to Microsoft Word. No matter how fully featured the program was, disregarding how user-friendly it was and regardless of the quirky or unique innovations it had, no word processor would ever get an unreserved recommendation without a decent word count feature. How many people even use a word count?)). Of course reviewers can raise the profile of your program but unless it results in more sales and not just more usage of your software then adding features or lowering the price only to please them is a waste of time.

If you want to sell software, your first priority should be keeping your customers happy, not reviewers.

Opinion

Net Neutrality, Privacy and Hypocrisy

One of the big technology debates in the US goes by the thrilling title of “Net Neutrality.” In the UK we seem to have skipped this part of the debate and moved on to the next, all without many consumers even knowing that anything has changed. As we’ll see, this does not work out well for many end-users and exposes hypocrisy and dishonesty on the part of the Internet Service Providers (ISPs).

The first question you might be asking is, What is Net Neutrality?

To cut a long story short, this is all about ISPs wanting to prioritise some types of Internet traffic to the detriment of others. At first glance this makes some sense. The ISPs have a finite amount of capacity with which to connect all their users to to the wider Internet. By design, all traffic on the Internet is traditionally considered to be equal. That’s to say that my movie download from iTunes gets the same priority as your video chat and my neighbours web surfing.

Is that fair? You’ll get delays and jitters in your webchat if the performance of your Internet connection isn’t good enough. My neighbour doesn’t have quite the same requirements as you. She doesn’t want to be waiting too long for that page to download, but the odd outlier isn’t going to cause any significant issue. And me, well, my movie is going to take a while to download anyway and I’m not planning on watching it immediately.

Right, so it makes sense for you to get full speed and my connection should be throttled back?

Let’s not be too hasty.

As I understand it, there are two main arguments given against traffic shaping (as it’s called), one technical, the other political. The technical reason is fairly simple: the Internet was designed with the assumption that all traffic was equal and we don’t really know what effect large scale traffic shaping would have. It could be the butterfly flapping its wings in Japan that causes a tornado in Florida.

The political reason is probably why the ISPs really want an end to Net Neutrality: they are not entirely impartial when it comes to deciding which traffic gets priority. These same companies want to sell music and movies and voice over Internet services, so what’s to say that they wouldn’t give their own voice traffic priority over Vonage or Skype? ((My parents ISP chose to throttle all P2P software, which may not be a bad choice per se. Unfortunately they also included Skype in this category. I’m not sure whether this was purely an accident or because they are mainly a phone company who are providing “free” broadband and are trying to push people away from VoIP.))

This debate never really happened here in the UK. While the big print in the adverts usually says “Unlimited Downloads!” ((Are people really so gullible as to believe that they can reasonably get unlimited anything? ISPs clearly think so as I am only aware of a few that have bandwidth quotas.)) the small print typically hints at traffic shaping. I suspect this already affects many people without them knowing about it; they just blame their PC or Bill Gates without realising that the problem is actually further downstream.

But as I hinted back in the first paragraph, in the UK three of the major ISPs are taking one further step: they are planning on selling our surfing habits to a third-party so that they can send us context-sensitive adverts.

It can’t be just me that thinks that this is creepy.

Advertisers love the idea as they can see exactly the sites that everyone ((They are three of the biggest ISPs in the UK. So while this isn’t actually everyone, they cover around two-thirds of British broadband subscribers.)) is visiting and not just those that, for instance, have DoubleClick or Google banners.

In return, consumers get some nebulous “security” protection, details of all their web surfing sent to a third party without their consent and even more adverts ((One of the most creepy parts is that if you opt-out then there is still the possibility that you’re still going to be tracked anyway.)).

Where does the hypocrisy come in? Well, recently the UK government suggested that ISPs should police the Internet for copyright infringement. Three offences and your connection is shut-down. These offences, incidentally, seem not to require any legal process or evidence — it’s pretty much the suspicion that seals your fate.

In response to this, the ISPs say:

“ISPs cannot ‘monitor or record’ the nature of the data flowing over their network, argues [the ISPA]. UK data protection laws make deep packet inspection illegal … and even if it wasn’t, complete monitoring is impossible. ‘ISPs are no more able to inspect and filter every single packet passing across their network than the Post Office is able to open every envelope,’ says the ISPA.”

Hang on!

When they can make money selling those records it’s just fine. And when this transaction hinders their own users experience by flashing up adverts without their permission, that’s also dandy. But when it comes to scanning the traffic for law enforcement purposes then it’s both illegal and impossible?

Being an Internet Service Provider has gone from being an almost glamorous, high margin business in the mid-nineties to one barely above commodity level today. It is, perhaps, no wonder that they are trying every last trick to eke out every last penny, but is selling out their own customers really the best strategy?

Opinion

Is Microsoft-Yahoo the next HPaq?

So Microsoft is trying to buy Yahoo. I’ll leave the detailed analysis to people better qualified than myself but I thought that I could add a little perspective simply by looking back and remembering something that happened less than ten years ago.

MSFT, HPQ and YHOO stockAs you can no doubt guess from the title, the event that springs to my mind is the merger of HP and Compaq. The main problem with HPaq at the time was that merging HP’s loss-making PC business with Compaq’s loss-making PC business just wasn’t a good idea. Fiorina pushed the whole MBA line of thinking: being the biggest player will allow greater economies of scale, lower prices and more profit. Unfortunately, two big losses merged tends to make a big loss also, albeit perhaps smaller than the old combined total ((HP obviously have other areas that do make money, most notably their ink cartridge printer division.)).

The saddest things from my perspective are that HP had a great engineering tradition and Compaq had some great technologies (Alpha for one), all of which were rapidly divested. HP traded a future for short term benefits. And, as can be seen on the above stock chart, the 2001 merger didn’t really do much for the market price.

Compare this to Microsoft and Yahoo. Yahoo have been in the doldrums for a while now. Back in ’94 they were the Google of the day, but they fairly quickly lost their way. Their indexing method (actual people as I recall) just didn’t scale as quickly as the Internet as a whole and they were quickly outpaced by up-and-coming new, automated systems like Altavista, Hotbot and Google. Yahoo retaliated by buying wholesale into the portal phase of the dot com boom and just outright buying interesting looking technologies, everything from music to photography sites.

And that’s pretty much where they are now. Of course they’ve changed management team a few times, but the share price has been fairly flat since 2001 (not sure what that little peak in 2004 was).

Microsoft are the company that brought you Windows and Office, and don’t really need much more of an introduction. Pretty much everything else they do loses money ((I like this Cringely article that explains why they like to do this.)). Or at least, any profits they do make in other divisions is line-noise compared with Windows and Office.

So if you combine Microsoft’s barely known Windows Live service ((If anyone uses this it’s because it’s the default in Windows. It’s undoubtedly getting better, but then so is the Zune.)) and Yahoo’s fairly popular if not hugely profitable web sites, what do you get?

Well, like the HP/Compaq merger, you get a lot of duplication. Microsoft and Yahoo both have search, instant messenger, photo publishing and, most famously, webmail. By some accounts, if they merge they will have over 80% of the webmail market. But where do you go from there? Do you migrate all Yahoo’s Mail users over to Hotmail?

And, most importantly, do a large number of regular visitors automatically mean big bucks? Is Microsoft buying a future — by saving time developing code and attracting users — or simply buying the past?

In the end, I just can’t see what Microsoft think they’re buying. Yahoo seems not to have the technology required to take on Google and their combined page impressions have no obvious quick routes to profits. I hope it works out — a world where Google is as dominant as Microsoft is now is not much of an improvement — but I have doubts.

Opinion

Check-boxes

When was the last time that you came across a printer with only a parallel port? Have you ever used one of those two serial ports on the back of your PC?

Personally I don’t think I’ve used a parallel port for over five years and I don’t recall ever having used a serial port despite using computers for over twenty years. So why do many PC manufacturers still include them on new PCs? Would they not be better using the same space for more USB ports, a Firewire socket or even to save a few pounds by simply not including it?

I finally realised why they are there when I read “Heavier Than Air.” It, of course, has nothing to do with their utility — almost no-one has used either port for years — and everything to do with feature matrices.

Imagine the scene. The purchasing department ((I am reminded of a story I heard while working at a large telecoms company. My department ordered an upgraded CPU for one of the Macs used for DTP. The card they asked for would have nearly doubled the performance. The purchasing department, in an attempt to save money, instead ordered a cache card. Unfortunately this cost nearly as much as the replacement CPU but only increased performance by about 10%. It proved to be a waste of money rather than a wise investment.)) of a big company is defining what the corporate standard for PCs for the next year is going to be. How do they go about it? Do they sit down and exhaustively analyse the requirements for new PCs across the whole company? Do they talk to users, book meetings with the IT guys and negotiate with the facilities team?

Or do they just copy last years spec sheet and hope for the best?

It doesn’t take too many years of this to get back to the point where dedicated printer ports and serial lines were kind of useful for many users. But now they’re just left there on the check-list not because anyone needs them but because no-one has thought to remove them.

You’ll see this in action at the various manufacturer web sites. The business oriented machines, those intended to be purchased in their thousands by blue chips, have those superfluous ports but the consumer machines are more likely to have only USB.

In the grand scheme of things maybe it doesn’t matter why you get a few extra ports on every PC you buy but I do think that it’s interesting to discover why they keep giving us stuff that we don’t want.

Opinion

The Perfect Thing

My first reaction was to hate it.

Actually, no. My first reaction was what a cute, almost-square book. My second was to hate it. It’s only a few pages into it and Levy is already discussing his great taste in music and disparaging mine (“… a pathetic Pet Shop Boys tune, the sort of thing that Nick Hornby would listen to on a bad day.”)

It wasn’t exactly what I was expecting. But — step back — what was I expecting? Well, the blurb on the back tells you that you can learn how the iPod became the defining object of the 21st century. Having read most of Steven Levy‘s previous books, the obvious point of comparison was “Insanely Great,” the story of the Macintosh. But while the Mac took several years to put together, the iPod took only around six months. In hindsight maybe it’s pretty obvious that it’s not all the trials and tribulations of putting the machine together.

So what is it? In some ways this is the technology version of Alain de Botton’s “The Art of Travel.” It doesn’t so much talk about the making of the iPod as discuss the experience of owning and using one.

For example, one chapter is on the idea of “shuffle” and how people use it and attach super-human abilities to its ability to select the most appropriate mixes while theoretically being entirely random. Even when he talks about the design it’s more along the lines of how it makes people feel. One interesting observation was that when people speak about the iPod, they invariably reach into their bag or pocket to retrieve their machine and start touching it, pressing the controls, polishing the screen. How many other gadgets evoke this kind of sentimental reaction?

Certainly much mirrors my own experience. I remember sitting on the tube on the way home from Tottenham Court Road after buying a first generation iPod, the tiny cube ((I suppose 2001 was when the G4 cube still held the promise of being successful.)) that it was packaged in sat on my knee, and thinking, “I can’t believe I just spent ?350 on a Walkman.” ((In fact I was lucky to get away with that. A number of shops, having just one machine left, wanted to charge me more that list price. I knew I wanted one but not that much.)) Yet when I got home and started using it, any doubt vanished and quickly it became my most prized gadget.

In fact, as I type this my fifth generation iPod is on the blink. The emotional impact of this neat pile of components being on the verge of failing is quite unlike almost any other gadget I own.

Anyway, I was writing about the book “The Perfect Thing” and not my own experience, so let’s get back to the point. If you’ve ever read any other Levy books (or his column in Newsweek) you’ll know that he has an easy, engaging style and this book is no different. He covers all the bases, touching on everything from the design to downloading music from the Internet (and the fun that Apple had making iTunes) to Podcasts.

Given that it’s more about the experience of owning an iPod the fact that it’s already dated — there’s no mention of the iPhone, the Touch or the current range of Nano’s — is less of an issue than I initially expected.

Overall it’s much more of a “fluff” piece than most of his other books, but that does not make it any less entertaining or complete. Worth a read, just not as highly recommended as “Crypto” or “Insanely Great.”