Rupert's DiarySlightly sanitised for the squeamish

RIM and Vodafone turned up at our front door yesterday, in their Review Bus. We'd had warning: "We'll be outside your office at 16:10. We may have room for you."

My lord, what a way to show off a phone! I had visions of some tour coach appearing, stuffed full of partying reviewers, or perhaps a double decker packed to the straps with keen young PRs. "Why not come on up to our offices and then you can show it off to everyone who's interested?" I asked. "No time!" said the PR. "Can't be a very big bus..." I suggested. "It might be a mini-bus", they conceded.

Hm.

With military precision – if you're Fred Karno's Army – the bus arrived. At 15:05, when I was at my desk thinking about very different things. "There's a woman for you at reception," said the IM. Some of my worst scenarios instantly play out in my head. "She says she's expecting... no, sorry, she's expected. From Voda."

Phew. OK, let's go and see this phenomenon. For fun, we form a massive posse and descend to the street mob handed, preparing to rush the bus.

There is a silver people carrier, just big enough for a small selection of demo bods with a couple of Blackberry Storms squeezed between them, and a rather sheepish PR. "I might have over-hyped it a bit..." she admitted, in the face of six quizzical eyebrows arched heavenwards. "We could always go up to your offices."

Enough of that. How about the Storm? Well, as my colleague David Meyer's more thoughtful conclusion says, it's unfair to judge a novel user interface until you've used it for a while. Things that look immediately clever have hidden niggles, and stuff that seems incomprehensible can snap into reason after a little while in real life.

So when I say "Uhhh....", that's a provisional "Uhhh....". But "Uhhh...." it is. The screen looks fine; it's big, bright and resolute. The phone itself is sleek, rather big and rather heavy – this is, says RIM, taken as implying quality, but I felt the ghosts of chrome bumpers and rocket-fin tail-lights cluster around.

This sense of classic American gas-guzzler only intensified when I laid finger. The Storm's schtick is that the entire screen moves into the body of the phone when you press on it – the idea is that you lightly touch to select something, then click by adding pressure. That by itself is not necessarily a bad thing: it provides positive feedback that's more tactile than any haptics. But two aspects of it are unsettling.

First, you're moving quite a large screen. This takes much more pressure than other systems, and there's a lot more inertia involved. As a result, the sensation is rather wallowy: it's a bit like prodding a tortoise and asking where it hurts. It also drastically limits how quickly you can type, although that may have been exacerbated by the general impression I got that the phone was a bit slow. (There's another update due for the software before release, so for many reasons that impression may prove inaccurate. But it's there now.)

Second, there's a small but perceptible gap around the edge of the screen. On something that's destined to live in the dusty, stringy, bitty world of pockets, that seems a disaster waiting to happen. You know how much crud collects inside the battery compartment and various electronic navels of ordinary phones; imagine that distributed around the action bits of a large component that has to move in order to work, within a crevice that seems impossible to guard or clean. I have enough problems in that department with my own 'built for comfort, not for speed' body - I'm not sure I'm ready for it in my phone.

We shall see. I must admit I don't quite understand the obsessive love so many people have for their Crackberries, so can't predict how popular this not-bad-go-at-Iphone-considering will be. It'll certainly have an easier time penetrating corporates, even if it doesn't arrive in the world's most dramatically shrinking bus.

In the same way that a pillow filled with pebbles is harder to get smooth than one stuffed with sand, chip makers are finding it harder to make transistors behave predictably as they shrink. Here, the stuffing isn't actually lumpier, it's that the pillowcase has got so small that the grains of sand look like pebbles.

And uneven transistors are disasterous. A circuit can only go as fast as its slowest component, and when you're dealing with hundreds of millions of transistors the bottom end of the variability curve is going to have a substantial population. There are tons of things that can vary, too, across speed, temperature, voltage, current and time, and at the level of engineering within a contemporary high performance circuit, there's very little room for imprecision.

Thus, variability leads to low yield - you end up making chips that don't work and can't be sold.

Which is why a recent announcement (pdf) from the University of Southampton is so interesting. Drs Peter Wilson and Reuben Wilcock of that ilk have come up with the CAT - Configurable Analogue Transistor - which is a a complex beastie hiding a simple idea. It's a bit like a an aircraft wing with extensible flaps - at take-off and landing, when you need more lift at lower speed and don't mind (or actually want) drag, you stick the flaps out. When you're actually flying and want low drag at high speeds, you tuck the flaps in and off you go.

At heart, the CAT is a set of exponentially smaller transistor parts that can be switched in various combinations in parallel across the main transistor. Once you've built your circuit, you test how it works and, if you need to, configure the right combination of extra bits to add to the problematical device to tune the performance so it works in the design. A bonus is that it's then possible to adjust for performance change over the lifetime of the device.

The researchers point out that you don't need to do this to every transistor in the design - part of the trick is identifying which ones are most at risk of affecting yield and concentrating on those, and you won't find there are that many. There are lots of other sensible caveats too, about layout and context - and of course, this is an analogue technique perhaps best suited for tuning transistors which have to operate in the linear part of their performance curves - transistors in digital circuits spend their lives on and off.

This isn't the first technique for adjusting the performance of parts after they've been manufactured - laser trimming, which involves zapping parts of a component with a death ray - has been around for a long time and still sounds more science fiction than a CAT. But this does illustrate a trend that I feel will become more and more important: self-adjusting circuits that don't assume their components are stable or reliable, but actively reconfigure them to operate in their optimal mode.

To return to the aviation analogy - it's like fly-by-wire fighters, where the machine itself looks after the donkey work better than any human can, making it possible to take fundamentally unstable designs and use that instability for phenomenal performance. You can even see a variant of that idea in Google's architecture, where it uses multitudes of cheap low-reliability hard disks and servers and expects the software around them to manage the results.

Expect these ideas to become more and more important as we get closer and closer to areas of physics and engineering where the statistics go against us. Exactly what the implications of this will be - well, we'll have to find out as we go along.

Microsoft has revealed more details of its Azure cloud service, including availability and how – although not how much – it will charge. In an FAQ document on its website, the company says that Azure will be available in the second half of 2009. Prior to that, a Community Technology Preview will be available free of charge 'subject to certain limits'.

When commercial service is started, Microsoft says it will charge for CPU time, bandwidth, storage and transactions, billing users for actual consumption of each service. The company promises "robust service level agreements and guarantees on quality of service", and says that access to the services will be direct, through a Customer Portal, or via third parties who will set up and sell their own services based on Azure. MS, like every cloud company, has happy visions of the world being delighted to provide a constant revenue stream that merrily rises as usage increases. The world's opinion of these visions has yet to be confirmed.

Meanwhile, Dell has been happily telling everyone that it's providing the hardware for the Azure computing centres – and then clamming up when asked for details of how much, what, where and when. Also unclear is how security will work, how reliability will be maintained and – perhaps most interesting of all – how easy it will be for a company that signs up for Azure to move its business elsewhere if it wants.

We'll be returning to that particular can of worms with our very largest tin opener later.

I came across a very interesting explanation of what actually happens in an incident such as the one which felled the LHC, from Vern Paxson, vern@icir.org. He worked on software for the 1980s big science experiment, the Superconducting Super Collider - and has kindly allowed me to quote his description of what the "bad electrical connection" actually was. Over to you, Vern!

"Bad electrical connection" is a layman's explanation of what accelerator physicists refer to as a "quench". This is in fact a routine event when working with superconducting magnets: somewhere something in the winding is imperfect, such that, when under the influence of extremely high magnetic fields, there is physical motion in the magnet's wire, leading to friction.

The friction generates local heat which transitions a small amount of wire from superconducting to "normal". The magnets carry such immense currents that the small resistance of the normal wire turns into a large amount of local heat; this in turn *transitions more wires* from superconducting to normal; and the process spreads.

It all happens fast, with the risk of explosion.

To counter this threat, superconducting magnets have high-speed circuitry designed to detect the onset of quench events. When detected, they very rapidly (and counter-intuitively) heat up the entire magnet, causing it *all* to go normal. The goal of doing so is that while individual wires can't possibly bear (when normal) the heat of the amperage going through them, the entire magnet has enough bulk volume to do so.

When a superconducting magnet is first commissioned, it undergoes "training", in which it is repeatedly (1) powered up to the point of quenching, (2) the quench-protection circuitry kicks in, (3) the whole thing goes normal and thus doesn't explode ... and (4) the imperfect winding that moved due to the high magnetic field is now in a new position where it likely won't move again (since the original motion was driving it towards such a potential-energy minima). Therefore, that particular quench won't recur. However, others might. So during training, you do this again and again until no more quenches occur.

So, Yes, such failures are expected, but in this case something went wrong with the quench-protection mechanism leading to the explosion. A bit more than just "a poor solder connection" (though clearly still not good, and I believe magnets are generally trained *before* installing them in the beam tunnel).

Vern


(Explanation/disclaimer: this is all from memory rooted in my former life doing software for the Superconducting Super Collider back in the 1980s.)

With the Windows 7 pre-beta preview now out of the bag - and our first review up - we took a little time to flip through the 4MB, 111 page, Reviewers Guide PDF. Microsoft Reviewer's Guides are provided to the scribbling gentry in order to gently guide the perplexed or fatigued into writing what Microsoft would like them to write, and normally run to a few pages of A4. This is the full-on 30,000 word brochure.

But what's most interesting isn't the size, nor the gleaming industry-standard happy people photographed in various conditions of bliss. It's what the company can't quite bring itself to say straight out. For example:

"Windows 7 provides regular progress updates throughout the upgrade process, with those progress updates replacing the “Your upgrade may take several hours to complete” text at the bottom of the screen in Windows Vista."

"Ooops", anyone?

And this: "Microsoft continues to work with closely with it ecosystem partners, beta testers, and early adopters to improve the overall compatibility experience for customers. Given the significant progress the software ecosystem has made in Windows Vista compatibility— and the continuous feedback and telemetry data that
Microsoft expects to receive from beta customers—the Windows 7 engineering team believes that Windows 7 is on the path to deliver a high degree of application compatibility"

Comes close to "Yeah, we messed that up too. Big time. But it's getting better. Really."

And then there's the ever-useful innate belligerence of inanimate objects: "Today, there are dozens of Microsoft Web sites that offer you information about Windows, making it hard to know which one has the information you need." It's the sites that make it hard. Not, you know, anyone in particular

But it's not all trying to say sorry and not quite managing. A lot just wasn't Microsoft's fault: ""When Windows Vista was initially released, not all of Microsoft’s partners had adapted their products to run well on the Windows Vista platform. This diminished many customers’ first experience with Windows Vista."

In fact, did you realise that Vista had been the starting gun for unparalleled innovation in, er, "devices"? Tell 'em, Microsoft! "In the short time since Microsoft launched Vista, the world has seen amazing changes in the nature of devices. They’ve gone from being single-function peripherals to complex, multi-function devices with a large amount of local storage and the ability to run applications"

Any idea what they're on about? It continues: "And they’ve evolved from a single type of connection -- such as USB -- to multi-transport devices that support USB, Bluetooth, and WiFi." Ah. Wonder if they're related to the devices mentioned here - "Although Tablet PCs with touch screens were introduced more than five years ago, new devices have revitalized the role of touch in the user experience." Can you tell what it is yet? Here's another clue:

"Windows 7 also introduces support for new multi-touch technology, enabling you to control what happens on the screen with more than one finger. For example, you can zoom in on an image by moving two fingers closer together, like you’re pinching something, or zoom out by moving two fingers apart"

Yes! It's just like an iPhone (there, I said it). Except - zoom in by pinching? Isn't that the diametric opposite to the way the iPhone multitouch does it? Egad. If only Microsoft had seen an iPhone - instead of one of these mysterious 'devices'…

Lots good in there too, of course, but the most interesting stuff -- application virtualisation, diagnostics, Powershell, security policy management -- is at the back. No happy smiling people. Which is odd, given that if that stuff works, it'll cheer up millions.

A reader points me at an energy story that comes larded about with exciting predictions. MIT Energy Storage Discovery Could Lead To "Unlimited" Solar Power! Illustrated with a large picture of a solar cell array, it has a familiar, breathless style:

"The process, loosely based on plant photosynthesis, uses solar energy to split water into hydrogen and oxygen gases. When needed, the gases can then be re-combined in a fuel cell, creating carbon-free electricity whether the sun is shining or not. According to project leader Prof. Daniel Nocera, “This is the nirvana of what we’ve been talking about for years. Solar power has always been a limited, far-off solution. Now, we can seriously think about solar power as unlimited and soon.”"

Is this as exciting as made out, wonders my writer.

No. It is, however, curiously misleading.

What has been discovered, as far as I can tell. is a more efficient form of electrolysis using catalytic electrodes, using electricity to turn water into hydrogen and oxygen. This could lead – as with the headline, there's an awful lot wrapped up in 'could' – to a more efficient form of power storage. Which in turn could make domestic solar power more attractive, but that's as far as the solar angle goes. The photosynthesis aspect, implying as it does direct conversion of solar energy to electricity, is a bit of misdirection; the discovery would work just as well – or badly – with any form of generation, from hamster wheels to windmill generators on the chimney.

Whild the rather elderly original press release does mention other original sources of electricity, it doesn't talk about is the efficiency of the new process. There are no details either of how the new process is related to photosynthesis, which is a stupendously complex business (the curious should read Eating the Sun by Oliver Morton for an engaging look at the cogs) that in any case doesn't produce atomic hydrogen but oxygen – it locks the energy recovered from the sun into carbohydrates, which the plant uses for metabolism and we sprinkle on our cornflakes or burn in our cars.

Hydrogen is a much more problematic intermediate storage medium. The paper talks about using a fuel cell to turn stored hydrogen back to electricity, but this glosses over many interesting engineering challenges. And the fuel cell itself remain a technology still heavily laden with "could" in its own right: we are not invited to ponder how far the whole system has to go to compete with existing battery technologies. And "soon" apparently means 'in ten years', which in futurology code means "we have no real idea". At least they avoided invoking nanotechnology.

That there's some interesting science here, I don't doubt. That it may be useful is also true, but only for certain very badly defined values of 'may'. And that the key mechanism for splitting water is in some way related to some aspect of photosynthesis – well, yes, why not. Photosynthesis is extremely clever and does things we'd love to emulate.

None of this leads particularly to making solar power "unlimited and soon", at least not on the evidence in the press release. Perhaps the real energy source that MIT hopes to exploit by the production of such an optimistic story may be best detected in the final paragraph of that press release. See if you can spot it!

"The success of the Nocera lab shows the impact of a mixture of funding sources - governments, philanthropy, and industry. This project was funded by the National Science Foundation and by the Chesonis Family Foundation, which gave MIT $10 million this spring to launch the Solar Revolution Project, with a goal to make the large scale deployment of solar energy within 10 years."

Our advertising team - fine men and women all, and I'm not just saying that because of their role in my drinks bill - forwarded me a fascinating note.

It seems that a bad person had faked himself a convincing identity as a media buyer in Denmark. For those of you lucky enough not to know, media buyers are strange intermediary lifeforms who cut deals with media outlets for advertisers. They can be... challenging to deal with.

This one, however, had a straightforward proposition. He wanted to book adverts for Gillette, which has a big campaign for Fusion/Venus razors going on now. He seemed to check out - his email and ad server domains were very well faked, looking just like those of a genuine agency, and you don't need much persuasion to get ad sales to sell ad space.

Indeed, his adverts were perfectly normal - at first. But his plan was to switch these to spyware/malware bearing versions after a short delay, thus neatly embedding nastiness in perfectly proper publications.

The good news is that his foul scheme has been thwarted, with the community of ad sales bods now thoroughly alerted and all the offending fake campaigns removed. The better news is that you really, really don't want to get on the wrong side of ad sales, who take this all very personally - so anyone else trying it on will find the market a lot harder to fool. Personally, I wouldn't recommend the experiment. These people are best kept sweet...

As usual, the excellent Nature production editors nail a story with a succinct one-liner: "How weird is that?".

Very. It turns out that ordinary sticky tape - Sellotape in the UK, Scotch tape in the US, Durex in Australia (well, once upon a time) - emits high energy X-rays when you pull some off the roll. In fact, enough are given off under the right circumstances to form an X-ray image of an experimenter's finger -- although as the right circumstances involve a high grade vacuum, don't send off for the lead codpiece before wrapping your Christmas presents just yet.

But the phenomenon is very real. It's an example of one of my very favourite bits of quotidian weirdness, triboluminescence. This is the habit of some compounds, including sugar, of emitting light when crushed: if you take a Trebor mint and a pair of pliers into a dark room and let your eyes adjust, you can check this for yourself. (I did as a kid and have been a big fan ever since.)

That light is not due to heat or sparks of static electricity, at least not as we normally know it. In fact, nobody's quite sure what happens: the leading theory is that the disruption of molecular bonds results in a sudden localised distribution of positive and negative charges, which snap back into neutrality with a burst of photons. And this makes some sense, given that charges changing speed are the way light comes about.

But X-rays? To make them normally, we have to use tons of energy to accelerate electrons in vacuum to very high velocities and smash them into carefully constructed targets. There are no known low-energy producers, and only the most exotic solid-state emitters.

So finding X-rays streaming out of a 99p roll of sticky tape by the application of thumb and finger is akin to discovering a nest of dragons behind the Pick'n'Mix counter in Woolworths. The researchers were "a little bit scared" when they first realised what was happening - they were checking a Russian report from 1953, which first noted the effect - but have now decided that there's no danger in everyday tape use.

They're now decidedly excited, but comprehensively baffled. Given that nobody quite understands the long-observed sugar crushing experience, this new variation - far removed from anything else in a physicist's experience - is utterly bewildering. But investigations are underway, with an intriguing codicil that the mechanism may be good enough to trigger nuclear fusion.

Gotta love that physics.

Always a pleasure to report on Microsoft doing good stuff - and this is good stuff. The company's researchers have studded the sides of a mobile phone with infra-red sensors, which can detect the presence and movement of a finger held some way away from actually touching the device. A bit of gesture recognition software, and you can control a surprising variety of things on the phone without having to slap your bananas on the thing. The technology is called SideSight, and while it's still very very prototype the researchers say that it works better than they expected. You can do things like rotate your hands to rotate the objects on screen, or shift the 'paper' you're writing on while you scrawl away with a stylus.

So what? Well, says Microsoft, even on things as big as phones the existing touch interfaces have their problems - if you touch the screen, you're going to obscure the very thing you're trying to manipulate and this can make selecting or moving items difficult (something I find frustrating on Opera, for example, where it's hard to know when you've tapped on a link, or even if you've hit the right one).

And the current batch of touch interfaces need quite a lot of real estate on devices - something Microsoft says is harder to arrange as things get smaller. SideSight raises the possibility of adding quite capable touch interfaces to things as small as watches (let's not mention Spot) or even smaller, and embedding them into a wide variety of things that don't even look like electronic gadgets.

As always, there's no guarantee that this will pan out into something that's really useful: history is littered with clever input ideas that look promising in the labs and even make it into products, but which just don't mesh with the way people actually want to interact with their gizmos.

But if you don't try, you don't win. This is just the sort of innovation that keeps the industry alive and its inhabitants thinking - so top marks to Microsoft for making it happen and for sharing the results.

And yes, it really does feel good to say that. More, please.

It was a salutary experience. I'd just sent the iPhone 3G back to Apple after two weeks of playing around with it, when the nice chaps from Research In Motion came in with their latest Blackberry. It was... well, a Blackberry. Perfectly good at doing what it does, but – "when are you getting an application store?", I asked. In the final analysis, that's the one thing on the iPhone which sets it apart from everything else. "Well, we might think about it", they said.

They've done more than that. Today, RIM announced its Online Application Storefront, which looks awfully like the Apple idea. Developers set their own rate, and get 80 percent of revenue. However, as you might expect from a company that tends more towards partnerships than Apple does, there are differences: mobile operators can have their own stores, for example, and presumably get to set their own rules about revenue splits. Whether this means you can't then go to the main RIM store, RIM couldn't tell me at the time of writing – but this and other questions are lodged and with luck, we'll know more tomorrow.

This is, of course, essential to RIM's survival. The company is in a good position to manage that, as its corporate bias gives it some insulation from the still-defiantly-consumerist iPhone, and it should be able to get developers on board if the storefront works properly. It's also thought about how to manage corporate apps via that (let's hope it allows someone to write a decent Exchange integrator), and as Andrew Lim, CNet UK's rabidly enthusiastic mobile phone bod said to me earlier – the more recent Blackberry hardware doesn't suck.

And a microsecond after the RIM announcement, Android opened its box and let everyone in to see, hear, smell, touch and taste the open source goodness. All rather thrilling, even if T-Mobile hasn't sent us a G1 ahead of the European launch: it feels rather like the early days of PCs, all over again. Which is nice. Geezers need excitement...