Mixed Signals
Any sufficiently advanced information is indistinguishable from noise
Monday 9 June 2008, 1:33 PM
xG's unclear communications
I'm finding it increasingly difficult to know how to handle xG, the Florida based but London listed wireless data company with a breakthrough technology it's kept out of public scrutiny for many years.
To date, none of the company's plans has come to fruition. Its first generation of product has come and gone without deployment or any proof of functionality, the flagship roll-out cancelled amid a cloud of ill-feeling and general confusion. Evaluation deals with various third parties have apparently expired without further interest. The tiniest amount of technical data, uncheckable and largely information-free, has been presented on the xG website. What partners it has are contractors, doing xG's bidding while refusing to say anything more.
What there is, is rumour, and in great profusion. And if the technology remains mysterious and uncheckable, then the rumours - mostly promulgated on a couple of independent bulletin boards - are doubly so. There is - or was, it may have been cancelled but I saw the original filing - a court case against the company that may or may not have been brought by the chief engineer, who may or may not have resigned with a few others over what might or might not be arguments over promised stock options that may or may not have been withheld.
That's just one. No aspect of xG's operation has gone unquestioned, and precious few of those questions have been adequately answered. if you have the slightest interest in what's really going on, good luck to you in finding out.
But if you do have that interest, you'll be practically alone. For one constant I can report from first hand experience: nobody without a financial interest in the company has the slightest belief in it.
Since xG invited me and a handful of other hacks over to Florida in 2005 to see a demonstration, I've talked about the company to engineers, network operators, handset manufacturers and anyone else in the wireless business, from independent analysts to billion-dollar CEOs, who are too polite to get away quickly enough.
The replies have ranged on a spectrum from "Who?" to "Ha!".
Doubtless xG's more enthusiastic supporters will claim this is an industry in denial, running frightened from a seismic event that will overturn empires. But I've found that bona fide industry players take real competition seriously and with respect: you get plenty of spin, sure, but where there are real questions they tend to get real discussion.
That's entirely absent here. On the technical and business front, xG is invisible. It's not behaving like any other public company three years into a strategy. It remains impossible for me to match the company's public statements with anything going on outside its walls, and I'll be extremely surprised if anything resembling a rabbit appears out of this particular hat.
But I'll carry on watching, like those under the command of the apocryphal Army officer whose assessment ran: "His men will follow him anywhere, if only out of curiosity".
Monday 9 June 2008, 11:42 AM
BBC threatens iPlayer hackers with mild annoyance
With the tedious inevitability of an unloved season, the BBC's iPlayer has provoked a chain of events with as much predictability as a Bond movie.
The plot is simple: the BBC wants to do its job, which is to provide entertainment, information and education to Britain with as few barriers as possible. The Internet being to television what television was to radio, the BBC knows it has to be there, and be there properly. Hence the iPlayer, which delivers TV to people's computers.
People like this. The BBC likes this. Others in the TV industry, never entirely happy with this 'broadcasting' idea in the first place, insist that every effort is made to prevent the wrong people from doing the wrong thing, for various definitions of wrong, so the iPlayer has various mechanisms designed to stop it working under certain circumstances. You know the drill.
Quite a lot of people, some of them very clever, have different definitions of wrong from some in the industry. Thus, these clever people set themselves the task of defeating iPlayer's mechanisms and, being clever, normally succeed. The BBC, having agreed with the industry that this is bad, then has to change its mechanisms - and the whole cycle repeats. As we've seen this with every 'anti-copying' mechanism ever introduced, this is about as surprising as Bond copping a snog from an exotic woman of questionable alliances, then finding himself in great peril as a result.
But the BBC, being a public service broadcaster, is in two minds about the rights and wrongs of the whole business. This may be most obvious from the latest changes made to prevent suspect people -- who appear to be defined as those not running Windows -- from actually saving content (which you can do in about fifty different ways already, but we'll let that pass). Those clever, naughty people had previously found that if they make their computers pretend to be iPhones, they get an unencrypted stream delivered.
At first, the BBC just tried to check harder that it really was an iPhone on the other end - but this never works.
So, at the the end of last week, it switched in some new encryption. At first, the community of clever, naughty people thought that this was Fairplay, Apple's own DRM system. But over the weekend, the truth came out - it was at heart a simple XOR with a sixteen-bit fixed key.
In cryptographic terms, this is as effective as threatening Bond with a poke from a Hello Kitty pencil. "You expect me to talk?" "No, Mr Bond, I expect you to die. Laughing."
It's hard to imagine why anyone thought this worth the effort. The iPlayer-tweaking community will fix this in microseconds, as anyone capable of writing the latest change will know, and those who aren't tuned into that community will be excluded from the fun already.
The only explanation I can think of is that the BBC has to be able to tell the rest of the industry that it's trying, while simultaneously avoiding committing huge resources to a fight it knows it can't win and doesn't believe in anyway.
But it'll happen again and again anyway. That's the problem with the BBC. Bloody repeats.
Friday 6 June 2008, 6:04 PM
Of Spice Girls, Soho and server marketing
Thursdays are traditionally good nights for an IT evening out. Last night, HP decided to hold a pub quiz - currently in fashion on the PR circuit. Myself, David "Did I mention I'm in a band?" Meyer and Julian Goldsmith from sister site Silicon.com formed a small team and ambled along to the venue near Regent's Park to see what we could do.
A healthy mix of reprobates had the same idea, and as usual one of the best bits was the gossip. The Dennis Publishing lot had plenty to say about Felix, their eponymous boss, who has been entertaining the world by promoting his poetry using red wine and helicopters. New joiners to the company are given a book of Felix' finest rhymes as part of their induction: the stuff itself is "challenging", according to a recent hire.
Rather sweetly, every Dennis person I talked to had managed to memorise the "If you want to talk about that, I'll have to direct you to our PR company" response they've had drilled into them in case someone wants to talk about Felix' (now retracted) admission that he'd disposed of a love rival by pushing him off a cliff - however, no poetry critic appears to have subsequently taken Felix up on his offer of a lift in the helicopter.
Gossip over, pub grub consumed and glasses charged, we repaired to the tables for the highlight of the evening - the quiz itself.
There were four teams : CNet Networks (calling ourselves the Throbbing Cooks. Sorry) and Dennis pulled ahead to an early shared lead, leaving teams Kevin and Indecisive Publishing in the dust. We were a bit worried about Dennis, as they were cheating - they had a ringer on the team who actually knew about football - but with a masterful flourish David Meyer (he's in a band, you know) demonstrated an entirely implausible knowledge of the Spice Girls. We entered the fifth and final round with a useful but not unassailable three point lead.
It was then that the wheels fell off the wagon - or, rather, the gigabit ethernet detached from the server. For the last round proved to require in-depth knowledge of HP - and, more specifically, market share of the server sector as revealed by Gartner's most recent report.
This wasn't a pub quiz: this was a presentation. The (by now considerably inebriated) hacks grew noisily restless. The quiz mistress - who claimed to be channelling Anne Diamond - rose to the challenge and managed to conclude the quiz without an actual riot breaking out. By now, I was channelling Brian Blessed and hereby apologise for some of the RATHER RUDE SHOUT-ING that may have taken place...
There was a rush to the bar as various HP flacks worked out the final results - and all was as it should be. As we graciously received the plaudits of our peers (at least, I think that's what they were) and our star prizes, we did point out the rather self-defeating nature of the final round, which had probably ensured that nobody would write about HP servers for months.
Turns out that the final round should have been about HP-UX - a nicely technical selection of questions about raw, unvarnished Unix. But the HPers had thought that unfair: next time, people, give us the real deal.
All good fun. As for what happened next - when a small but committed exploratory team discovered that legendary Soho drinking establishment Troy's had risen from the dead, and was still capable of exerting its considerable dampening field on any form of common sense - there is some hope that it never comes to light...
Friday 6 June 2008, 2:33 PM
Mobile WiMAX's hidden problem
If you take the industry's word for it, mobile WiMAX -- 802.16e -- is serenely sailing on towards deployment. Networks are being built, laptops sprout adaptors, Powerpoint slides are being churned out in self-assured profusion. Money is being spent, and lots of it.
What nobody's talking about, though, is a technical drawback that has stubbornly refused to go away. Mobile WiMAX may be fast and the devices suitably small and capable - but the transmitters are hideously inefficient. For every milliwatt of radio power radiated, up to nine will be lost as heat, and that has frightening implications for battery life.
To give you some idea of how that measures up to alternatives, a WiMAX power amplifier - the part of the transmitter which boosts the synthesised signal to a level that can make it to the base station - is between ten and twenty percent efficient. 3G amplifiers are between 40 and 45 percent efficient, and good old-fashioned GSM can nudge 55 percent. That puts WiMAX at up to a 5:1 disadvantage; considering that the maximum power output is around 250mW, up to 2.5 watts will be consumed.
Given that an Eee, for example, takes around 7 or 8 watts in total – and that's without the low-power technology that's coming up – that means getting on for a third less battery life. Not a very palatable idea.
What's more, there's not much that can be done. While there are plenty of ways to reduce power consumption with digital systems, the power amplifier in a WiMAX transmitter is resolutely analogue. It has to reproduce the complex waveform of WiMAX's OFDM (Orthogonal Frequency Division Multiplexing) modulation scheme, which has multiple simultaneous carrier frequencies that rely on a high degree of accuracy and very low distortion in the transmitter.
Moreover, if modulation peaks correspond on many different carriers, the OFDM signal can suddenly demand much more power from the transmitter than the average – and the only way to make an linear amplifier capable of providing a lot of power very quickly is to have it running with a large overhead all the time. That's inefficient. It's also basic physics. If you can't meet that need, you'll distort the output signal – leading to signal degradation, interference with adjacent channels and generally poor performance for you and others in the area.
This issue has been known about for a long time, and came up frequently during the 802.16e standardisation process – but in the end, the importance of getting the standard finished quickly led to the decision that the problem would best be solved by manufacturers. Unfortunately, nobody's found a way of doing this yet: various companies have promised new techniques that could cope, but so far there's no real evidence that they work. The short-term solutions may include reducing the output power from the specified maximum, but that will affect range, speed and reliability.
Mobile WiMAX's closest future competitor, LTE, has decided not to use ODFM for the uplink for precisely these reasons. And, while it'll turn up later than 802.16e, it may have an easier time of it if WiMAX spends its early years in working out this fundamental problem.
Tuesday 3 June 2008, 9:57 PM
Is Google's new VR toy really, virtually all its own work?
You've probably heard of Photosynth by now, the rather fabulous technology from MIcrosoft Research that automatically stitches together overlapping pictures into a virtual world. You won't have seen it in action yet - although parts of it are probably in Worldwide Telescope - because, well, Microsoft has other things on its plate at the moment.
Today -- oops, how did that happen? -- Google has gone and quietly launched a much more basic, yet still tremendous fun, piece of software that does something rather similar but has the distinct advantage of being attached to a real application. It doesn't really have a name - there's a "Look around" tag that invokes it -- but you get to it through Panoramio, a recent Google acquisition which couples user-uploaded pictures to Google Earth.
Here's the blog where the feature is introduced. It's far easier to play with than to describe, but here goes.
When a photo's got a 'Look around' tag beneath it, click on that tag. That'll take you to a Flash application with the photo in it. Mousing over the photo will cause various quadrilaterals to appear, each of which describe the borders of other photos that share part of the one you're looking at. Click within one of those and you'll be smoothly mixed to the chosen photo - which has its own set of quadrilaterals, and so on. It's very addictive, and the frustration you feel when you've explored an entire photo set is real.
So - good work, Google. But there's one twist to the story. A Moscow chap called Vladimir Slepnev has previous created and GPL'd an apparently identical application called OpenPhotoVR.
As he points out, the coincidences are visible, technical and rather far-fetched. He mentions "the Jacobian trick for transitions (don't ask)" - so I had to go and look it up. He's right. Don't ask. Unless talk of first-order partial derivatives of a vector-valued function makes you warm, in which case ask someone else. But it's a very obscure mathematical function: I wouldn't like to speculate how likely it is that two people would independently choose it out of many more common transition functions, but I do see his point.
He's philosophical about the possibility -- probability -- of being an unrewarded, uncredited contributor to the Googleplex, and considers the honour of potentially being upstream of Google's source reward enough. Although he'd like his next gig to be somewhat more material in its compensation.
Reading this, Google? (of course you are.) Any comment?
Previous
