My late uncle and I were very different people. Despite being the two ‘fixers’ in the family, the ones who got the busted kettles and the snapped gear cables from the rest of the clan, we were poles apart in one area: our approach to warranties. Even though he would keep his cars going for 20 years, he had a very sharp understanding of what should be his responsibility, and what was down to the vendor.

Actually, that’s an understatement. Woe betide the firm whose slipshod customer handling captured his attention. Once the horn-rimmed specs and the Brylcreemed bonce were aimed in their direction, he would pursue them relentlessly, his measured drawl torturing their receptionists until they actually did put him through to the MD or the Company Secretary (which incidentally is still quite a good one to try, since chancers seldom know enough about company law and structure to try that route).

I am the opposite. I fix (where I can), and like him I take great pleasure in diagnosis. However, I have a low opinion and equally low expectations of what happens when one tries to make a warranty’s promises stick. This is largely because my career in computing has spanned the period during which price of equipment has fallen so spectacularly as to leave me groping for metaphors.

I have a low opinion and equally low expectations of what happens when one tries to make a warranty’s promises stick

I can remember a DEC engineer turning up to put a memory upgrade in our VAX. He marched through the door and waved a plastic briefcase. “I’ve got a Testarossa in here!” he declared – meaning that the contents were worth the £60,000 of a then-hot Ferrari. This last month I’ve received 32 times that much RAM, shipped (and dropped) by the US Postal Service, for £250.

It’s difficult to run a perfect warranty upkeep process when prices are low and margins are tight. Unlike my Uncle, I have a “time is money” attitude (if he was the Obi-Wan Kenobi of our family then I’m more like Iggy Pop). When I start to see signs of undermanning or deliberate sandbagging during a server warranty claim and engineering visit, I will occasionally take the view that throwing money at the problem is worth it to keep the project on track or the service level up to scratch. Those who always take the opposite view are surprised when I sympathise, out of character, because I remember my Uncle and his completely different way of doing things.

However, I don’t think either of us would get very far pursuing warranty or fitness-for-purpose claims in the smartphone and app store business. App stores are supposed to be great, easy gateways for developers to reach new markets, and for users to benefit from an intermediary’s validation and quality control processes. However, there are plenty of opportunities for gaps between the promise and the reality.

One early example from last summer was a first generation Windows Mobile 7 phone. These could lock themselves completely as part of the ActiveSync system update, with a message of “take me to your dealer” for a complete factory reset and retry. I had one on test that duly bricked itself; it had a Vodafone PAYG SIM in it so I went to a Vodafone shop to get it sent away and reset. “Sorry,” they said, “not sold through us. Not our problem”. Despite asking around, I couldn’t find anyone who would actually do the necessary reset.

That was irritating, but not as irritating as Navigon’s Android satnav app. It costs a relatively whopping £60, and with Europe-wide maps it also demands at least an 8GB data card. Nevertheless, it seemed like an excellent deal for me, since I could sell my single-purpose satnav unit (also from Navigon) and come out of the overall deal about £20 in profit.

And a good deal it proved to be for the rest of the year, but then I didn’t travel for a bit so the Android phone got a rest. Next time I got it out, several apps (including Wyse’s excellent Pocket Cloud RDP client) had pending updates. Leaving it on charge and updating, I went to pack, and threw the travel kit in a lightweight laptop bag (pre-checked to remove sharp implements, tools and network cable testers – airport security people simply interpret them as Semtex, so far as I can tell).

So when I sat down in the hire car at Zurich airport, I got a nasty shock: “Activation failure,” said Navigon for Android. “There has been a connectivity failure.” While waiting in the queue to change to a more expensive car with included “Navi”, I proved there jolly well wasn’t a connectivity failure, by surfing the net and looking up the address to fire off a complaint, via the Android Market, to Navigon. I might as well have tucked my complaint in the Schnapps barrel of a passing St Bernard – it vanished.

I could go back to the credit card company and invoke the Sale of Goods Act – except it was an update, several months after the purchase, which interfered with the functionality of my property. I can’t even find a rollback button, which is the kind of thing one might expect after we’ve been through 40 years user interfaces and software delivery.

A truly international marketplace also means there’s little likelihood of a consistent approach to regulation. My Motorola DEFY picks up the central Android Market and the transaction is in sterling, so the actual relevant legal domicile for calling these people to account could be California (for Google), or the UK, or Germany (for Navigon) – except that the app store makes no provision for escalating this kind of failure.

So while app stores show every sign of being the way forward, my experience shows that the current invocations leave much to be desired when it comes to the traditional balance between the rights of the vendor and the rights of the customer. In fact, I’m thinking of proposing a new PC Pro award. I’ll call it the Customer Responsiveness APP Award – or CRAPPA for short.

As you may be able to see in the highest-resolution version of the snapshot above (click to enlarge), it’s not every day one comes across a physically distorted DIMM.

This is one of a set of eight 4GB sticks, originally intended to boost the performance of a Hyper-V host machine at Ratcliffe & Brown Wines & Spirits, the subject of a forthcoming PC Pro Business Clinic. The server upgrade wasn’t part of the subject, but it pretty quickly turned into a source of aggravation – this bendy SIMM is not immediately apparent until it’s placed on a flat surface, and I tend to land DIMMs on a lump of textile, like a mouse mat or a rucksack; anything but a conductive perfectly flat plane like a rack-mounted server lid.

Surprisingly, it sat in the DIMM slot perfectly well. Unsurprisingly, the server (a Dell PowerEdge 2970) spat the dummy the minute power was restored, quite accurately complaining about “unusable memory” in the scrolling front-panel display.

I don’t see many DIMMs that wobble to and fro when put down, or double as a leaf-spring the other side up

There didn’t seem to be a “too much torque on DIMM slot 8″ message, even though this was the source of the problem. I’d felt something a bit peculiar with the memory in my hand, but decided it was worth a try: in this particular case, the server was going from 16GB up to 32GB, and if one stick was faulty then we would have to take out the other half of the matching pair, reducing the available RAM to a comparatively meagre 24GB.

Once back out of the machine and laid on a flat plate, the bend became detectable: I don’t see many DIMMs that wobble to and fro when put down, or double as a leaf-spring the other side up: what amazes me is how it got through the memory vendor’s QA processes. Perhaps their tester has jaws that clamp the edge connectors firmly enough that the sundry distorted open-circuits on the stick are forced shut. But it’s visible even when the DIMM has gone back in the packaging, and doubly so when stacked in a pile with other, perfectly normal DIMMs.

Which leaves two questions: how on Earth do you put a permanent bend in a thing as strong as a plank of fibreglass? And, secondly, how did it manage to leave the factory?

In response to a recent survey telling us that schools are getting the teaching of Information Technology all wrong by not including “computer programs” in the syllabus, the BBC has offered up seven questions about computer programs. I urge you to take the quick quiz and then come back here when you’re done.

I scored five out of seven. I don’t know the correct HTML for inserting an image, and I couldn’t work out which subset of acronyms the question with GNU in it was driving at, mainly because the preceding five questions were not about “computer programs” at all; they were about the history of the people who happened to be involved in the invention of programming, either as a general concept (Jaquard) or as an incredibly early implementation (Hopper and COBOL).

I have to say – despite having earned a decent living for some time as a COBOL developer – I didn’t know Grace Hopper’s nickname and lucked out on that answer too, so really I should have scored a mere four. Knowing Admiral Hopper’s nickname wasn’t necessary to make my programs work or to earn me money, so I was in shameful ignorance, at least if you share the BBC’s perspective on the matter. To my mind, only two of the seven questions actually addressed the subject of the questionnaire.

I don’t ask that the guy who changes the oil on my Mercedes knows who Emil Jellinek was, because it’s not a necessary piece of information for him to do a good job

As you can probably guess, I regard this questionnaire as a terrible example of precisely the problem that the report is alluding to. Outside of the hallowed halls of hackerdom, almost nobody knows what size or shape the job of programming actually has, or how it should be thought about.

Of course, the “historian’s perspective” is one way to do it, but shouldn’t be confused with the artisan’s practical understanding of their tools. I don’t ask that the guy who changes the oil on my Mercedes knows who Emil Jellinek was, because it’s not a necessary piece of information for him to do a good job.

So the question rests with us: with the contributors and readers of PC Pro. Starting with readily available equipment (and personally I’d propose Free Pascal), what would you do to improve the comprehension of “programming”  - not only in schools, but in evidently confused and distant institutions such as the BBC?

I’ve just had a proper argument. My circle of friends and even a few colleagues at Dennis will tell you, this isn’t unusual of itself, so I won’t do the down the pub routine that relies heavily on the phrase “So then I said…”. I’ll give you the helicopter view.

It was an argument about Wi-Fi. I went to a meeting to go through re-wiring a retail shop to accommodate a CCTV system, the sales PCs, the PDQ card-payment setup, and the email workstation. There was also a couple of new ventures, in the shape of kiosks for customers to look through the website and ask about styles, sizes and colours not visible in the shop.

At this meeting were the proprietors, me, and a jobbing interior decorator. The list of snags, water leaks and bits of paint and the like was long and diverse: then we came to the wiring. Just a small shop, but very quickly we arrived at a total of 15 locations. It’s also an old building, which means that it won’t be falling down any time soon; but conversely, drilling holes is going to be a proper rufty-tufty builder’s job, one I am very glad I won’t be undertaking. Looking at the job in hand, the jobbing builder decided to propose a different approach: Why not just put in wireless?

Once the idea occurred to him, it snowballed. With Wi-Fi, customers could just be given an iPad, and wander freely around the whole space, paging through the website. How cool would that be? Wires aren’t needed then. I suggested this might not work out very well, given the background level of theft in that specific shop and the surrounding area too: and that it might not be terribly secure, in an area so full of other shops, offices, homes and restaurants. This is when the conversation kicked up a gear.

We are one of the last great self-taught professions, and from that many difficulties follow

The jobbing builder clearly believed that there is no such thing as a Wi-Fi security problem. To the point where anyone who suggested otherwise was to be cross-examined in an incredulous tone. There is no such thing as a passive Wi-Fi traffic listener Trojan, or those websites that crack WPA2 keys, or people whose credit card numbers or bank details are stolen via Wi-Fi or traffic spoofing. As far as he was concerned, Wi-Fi was the future; the idea that it could be much more expensive and complicated to segregate the network so that the CCTV in the changing cubicles didn’t get re-broadcast across the rest of the planet was, apparently, a stupid thing to suggest. All those videos on YouTube like this are clearly fakes.

I confess: I lost my cool with this tirade of ignorance. At the same time, I was thinking about Part P.

For those who have not come across it, here is the IET’s summary of Part P:  it’s the regulations that attempt to control who is allowed to do electrical wiring work. When introduced, I must say I agreed with the antis, because it seemed to me completely absurd that there could exist anyone who didn’t know how to wire up a wall socket correctly. It only takes a small tickle with 240v AC to entirely convince anyone of the need for proper safety in wiring. I can’t have been the only small boy to have successfully hidden the burns from an incautious poke about in the guts of a radio, surely…

This conviction faded slightly after I saw my first few 13A sockets with bare wires wedged into contact by the earth-pin shutter, and other similar sins, until these days I am pretty much entirely in favour of the concept of Part P. If someone wants to do that kind of work, then go and get the qualification, is now my attitude.

This is a very unusual conclusion to reach if you are a “computer person”. We are one of the last great self-taught professions, and from that many difficulties follow. Assuming that everyone is equally able to teach themselves, and equally able to draw the right conclusions from an individual view of a wider body of evidence, is (I believe) our greatest sin. This hasn’t been that much of a problem while IT and networks in particular has been the province of a priesthood, a charmed circle of übernerds: the problem comes when network technology starts to permeate into the skill levels that gave rise to Part P.

I’m trying to be polite and it might not work in my current mood, so I’ll settle for blunt: thick people think differently from nerds. It’s not a matter of less of something, like an IQ score, instead it has many aspects and parts. There’s emotion, there’s ego, there’s ownership of the topic, there’s animal cunning versus lofty and mistaken intellect: it’s a rich minefield of disasters, at least if your tempter works like mine.

The clever thing to do with this type of problem is to avoid getting dragged into Meldrew-like expressions of exasperation, but I will say that the inception and history of Part P makes me worry about the take-up of IP networking in the wider population of trades. Part P protects against a simple phenomenon – a pretty immediate and intensely memorable electric shock; good small network design protects against a rather more subtle, long-term and generally less physically painful series of mishaps.

But the underlying point to Part P remains that incidents arising from electrical wiring put in blithely by workers and DIYers, quite convinced they were doing it right, were prominent enough that Health and Safety decided to get involved. There is no equivalent body for network data security – unless, that is, you count the loss adjusters who now turn up when your bank account is emptied by an online data theft incident, and seek to prove that you were negligent in your use of the bank’s website to get them out of reinstating the contents of your bank account.

Happily for me, this particular client had been dealing with that type of mishap already, and were also better diplomats: they pointed out to the builder that he couldn’t very well remark on the superior strength and thickness of the Victorian buttresses and brickwork, and then recommend Wi-Fi. This contradiction provided a way out of the contretemps without too much loss of face all round – something that, as a classic nerd, I never have been very good at ensuring.

It’s properly, seriously hot here at Monza. This is, many would say, the most theatrical of the Formula 1 weekends and in the 30-plus degree heat, there’s a vast amount of technology on show. Most of it’s related to making cars go round at over 200mph, and this is the province of items like a solid tungsten nose-weight, or a £200,000 steering wheel — and that’s what you’ll hear about when the big TV stations walk around the pit lanes or chat with the drivers and managers.

I didn’t do that. I went to the quieter truck back in the pit lane and met Anthony, who is the sole IT support man for the whole Team Lotus presence here at Monza. He showed me the calm, unassuming and cool-running short rack that is the heart of the telemetry and on-track analysis system.

It’s 22 VMs and a couple of iSCSI SAN boxes, one SSD and the other spinning disk, and there’s a little rack with colour coded RJ45 cables in it. Such is the pervasive nature of Team Lotus’ WAN that the colour codes aren’t your usual subdivisions between floors or departments: some of them (red, as I recall) lead straight to the car. There is of course a wireless radio somewhere back there, shared antennas servicing all the teams and all the cars — but data flows only inwards. They can’t remap the car out on the back straight on the fly.

What really struck me looking around all the kit on display (when I wasn’t being annoyed by the tyrannosaur snorts of those little car things) was just how ordinary it was. You might imagine that everything is cast from spun unobtanium, suspended on the wings of angels or something, but in reality what is inside the rack and even in the broiling-hot trackside seating for the team owner and team manager (this has a special name but I want Anthony to have a good appraisal this year, so…)  are all standard regular everyday bits of Dell hardware.

Part of the reason for this is that the maintenance of the kit fits into Dell’s mildly elevated service platform, no matter where in the world the team goes. The other part of the reason is that the Optiplexes driving the screens, read by the multimillionaire owner of the Formula 1 team, manage perfectly well in 50-degree heat.

And to prove my point, here’s a screen-snap I wangled out of Anthony, probably against his better judgement, of the entirely regular and normal IP-based network monitoring utility he uses to keep track of all the things on his WAN with an IP address.

The red blob, top right? That’s a Formula 1 car that happens to be turned off at the moment.

I was in a meeting last week where both of the staff from my client had a strong affinity for the word “layman”. Whenever I strayed into territory they preferred to find too technical, they would say “well, I’m just a layman…”. I’m thinking of a particular conversation about their server hard disk running out of space. “What” they asked “you mean the memory? We bought some more of that, didn’t we?”

I despair of the whole concept of the “layman”  - they seem to stop being laymen and turn into the copyright-smashers from hell when it comes to downloading the illegal copies of movies that make up the bulk of the space consumed on their file server, after all.

Now, I’m sure we all have our stories about wilful ignorance in pursuit of a bit of nerd-baiting, but this particular BBC article caught my eye, because it implies that the “layman” state of mind is doing a good deal more damage.

Yet, even when you’re roaming in a foreign country, it’s not as if the voicemail storage is moved from your home phone company onto a different server

In short, Jude Law is alleging his phone was hacked while he was in New York. Apparently this means that because the phone was passing through a US cellular network at the time, it makes the illegal activity subject to US law.

Astute readers of the PC Pro site may well be ahead of me here, but let me summarise our mobile expert’s take on the matter. Paul Ockenden quite rightly pointed out that the hacking consisted of taking a guess that a phone user hadn’t changed their default voicemail PIN, as used for picking up messages when not on one’s home cellular network, or when calling in from a landline.

Yet, even when you’re roaming in a foreign country, it’s not as if the voicemail storage is moved from your home phone company onto a different server. Nor is it the case that the phone is where the voicemail is stored, or recorded – I’m sure that when Mr Law was flying over to New York, to take the case in point, any voicemails left for him while his phone was turned off in the plane wouldn’t have been lost or rejected.

So when the News of the World “hacked his phone” all they were doing – as has been established already – was to make a UK domestic link to a UK domestic voicemail server. The country the phone was in is immaterial.

The only incident of real “phone hacking” I’ve personally handled was that of a chap whose phone was cloned while he was at a business convention in New York. His bill hit £1,200 in three days – a charge that his network provider and the forces of law and order in the Big Apple seemed to be indifferent to pursuing.

I guess this particular kind of technically blind justice depends on whether you hit the mood of the moment. In the case of Mr Law (what a great pun) and News International, I would personally prefer that they make a case that is less dependent on a basic misunderstanding of a technical cloud-computing matter.

There’s a very special pleasure to getting out of town on a working day: a pleasure that’s doubled when you can borrow an open-topped car to do it in, and even more Brownie points attached to going somewhere with a fascinating story to tell.

This combination of plus points made me very happy to kick off PC Pro’s new Business Clinic feature – where Real World Computing contributors such as myself pay a free visit to a business for a spot of IT troubleshooting –  with an inaugural visit to see Dan Rooke, down at the Sandy Balls Holiday Centre in the northern extreme of the New Forest.

You have to get over the name. It’s a very cheap joke, and one that I found produced intriguing results: one colleague turns out to have actually worked at the camp; a client had some of his staff staying there at the time, and another had never heard of the place but immediately vowed to give it a try – not the kind of result I was expecting at all.

Of course, at PC Pro we were expecting some incredibly high-powered, preferably secret or privately funded City consulting firm to be our first volunteer for a visit from the Real World Computing squad, but, as you can read in issue 203, Sandy Balls had some chewy and not instantly fixable technical problems to confront.

Credit is due to Dan and the rest of the team for being prepared to go first. Especially at a time when trust in reporters and journalism in general is taking a bit of a battering, they were welcoming and quite straightforward about the things they would rather not share, but which made up part of Sandy Balls’ reason for asking for an outsider’s view.

We have been talking about how to balance people’s need for help with a need for suitable confidentiality – something we’ve been balancing in the Real World columns for years simply by anonymising the companies and individuals we write about. In Business Clinic we don’t want anonymity, we want to show that real people in businesses just like yours – and utterly different from yours – share your problems and your frustrations.

Of course, all of this depends on a suitably persistent stream of volunteers. I’d like everyone to take a look at the Sandy Balls piece and let us know if you’d like the treatment – it won’t always be me doing it, so you have nothing to fear there!

The first Business Clinic will appear in the new-look issue 203 of PC Pro, on sale 14 July. If you would like to volunteer your business for a free visit from our Real World Computing experts, email us at businessclinic@pcpro.co.uk explaining why you need our help.

Relax – when I say Microsoft’s “connecting you to the cloud” this isn’t the long-predicted release of a brain implant chip worthy of Cronenberg at his finest. It’s a bit more straightforward, though describing it to those not already in daily contact with the cloud produces furrowed brows – not because it’s hard to understand, but because systems designers considering a cloud rollout in their business can’t believe it’s not included already.

It is now. Robert Wahbe was the main presenter in the TechEd keynote, and he laid out a toolkit for linking your internal servers to your Windows Azure cloud instances. There’s Azure Connect, which is all about the TCP/IP pipeline between the inside of your organisation, and the inside of your cloud presence: and there’s Concero (not Concerto, before the subeditors shoot me), which is a data synchroniser, so you can have an internal server and a cloud server and keep the two in step.

Before the hawks jump on me, I haven’t found out at this early stage what granularity Concero has – whether it’s de-duplicating, or data atoms, or block sized, or any of the other options in what’s becoming an amazingly versatile and interesting field. All you really need to know at this point, is that Microsoft has realised that most of its faithful are faithful for good reasons, and there’s every reason to go through a fleshing-out process with cloud toolkits.

Microsoft has realised that most of its faithful are faithful for good reasons, and there’s every reason to go through a fleshing-out process with cloud toolkits

It seems a little disappointing to talk about “fleshing out” in the context of a show with lots of announcements – all of us at PC Pro love a big headline or the chance to use the term “game-changing” but the fact is, most people want to see long-term improvements in the tools they are using in projects they can’t just flip about the marketplace at whim.

TechEd 2011 is definitely for them, with a slew of announcements as broad as this: clever code tricks to flip your status updates to Windows Phone 7, Hyper-V support for CentOS as a guest, Exchange Clustering features added to Hyper-V, 3rd-party SAN management interfaces included in System Center Virtual Machine Manager.  It’s pretty clear that the challenge for Wahbe and friends isn’t about stealing market share from anyone else: it’s about stealing market share from previous versions.

There were two products that bucked the trend. The first is the early signs of legitimate, non-cracked and hacked uses for Kinect. Wahbe introduced a video from a Canadian liver cancer surgeon who was able to do wobbly, bloodstained T’ai-chi inside the operating room to run an Xbox with Kinect as a cursor manipulator for a CAT scan data set. No illuminated fingertips or Tron-style neon armpits required – Kinect watched his hands and the middleware on the Xbox translated the movements into rotations and (pardon the gory analogy) drill-downs through the patient. That struck me as a lot more wow-worthy than a lightsaber emulator.

The other product, which I will be covering in a lot more depth at a later date, is Windows Multipoint server. This is a resource-sharing server, aimed squarely at education, charities and the Third World, where there are more eyeballs and screens than there are computers to power them. Dean Paron demonstrated the wide variety of ways that Multipoint can give small groups of people Windows 7 desktops without running multiple machines.

Only as the demo progressed and he started using phrases like “I wanted to include, but”… or “I’m going to see if we can…” that the penny dropped: TechEd was his first demonstration of the product to a non-internal audience, and he wasn’t a sales professional: he was the head of the development team. Contact like that is what makes these events worth the jet-lag.

As you can see from the picture above, I have a deep dark secret to share with you. I am an irrational Motorola fanboy. Here in the picture is the Motorola Xoom tablet and my recently purchased, thoroughly huge and dorky looking Motorola S805 A2DP headphones.

The A2DP bit is what makes them so much fun. Like the HT820, which I also own, these headphones will stream music wirelessly, for a sensible amount of time, over Bluetooth, all the way to the outside world while the Xoom is sitting on my basement desk. That’s 25 feet away in a concrete-walled building.

I think they’re huge fun, and a seriously accompaniment for the Xoom, because you surely cannot be carrying that around in your top pocket when you’re working. The S805 takes the concepts found on the HT820 a bit further, with rings on the unfeasibly huge ear-pads to change the sound level (left ear) and to jump back and forward a track (so that would be the right ear then – I am getting the hang of this consumer electronics review game you know!).

Not only that – you can use the big chrome buttons to stop and start the stream and – here’s the best part – both the 820 and the 805 have an inbuilt microphone, so they qualify as hands-free headsets too. I can see the advert campaign – starring, no doubt, Dom Joly – in my mind’s eye already. Take those weedy white earbud things and plug them where the sun don’t shine, Apple fanboy!

You would never imagine in a million years that the two headsets came from the same manufacturer, let alone the manufacturer of the device that they won’t link with

Except despite having the Moto DEFY as my backup contract phone, and still having my RIZR as the phone-left-in-car on a PAYG SIM, I also have an iPhone. This is mainly because it doesn’t matter how hard it tries, Motorola seems to be the Citroen of the personal electronics business. Each device has at least one barking mad, senselessly perverse trick it’s going to pull on you.

The RIZR has a very loud “signal lost” tone. In fact, it’s not only loud: I can’t find where to turn it off, at all. The software is one generation too early to be sensibly friendly, and “signal lost” sounds almost exactly like the shop doorbell in every fly-blown hardware store and bike shop on the planet.

In fact, it’s not only “signal lost”, it’s “network roaming”, so you can be sitting still in a room in Central London and as a double-decker bus goes past outside, the RIZR will go “BOOODLE-OOODLE-OOO” at the top of its voice, just so you know it flipped off Vodafone and piggybacked through Virgin for a few seconds. I loved the way it let me use an MP3 as a ringtone without having to sign up to a million services and buy the wrong track by mistake half a dozen times, so I used Les Rhythmes Digitales’ “Jacques yo body” – which would only ever come out a fifth the volume level of that incredibly loud “BOODLE-OODLE-OOO”.

The Defy (sorry, DEFY) has a Motorola-specific super-social-networking thing layered on top of plain Android. It won’t start up without attempting to log into whatever that is – something the XOOM doesn’t try on you, thankfully. On the other hand, the DEFY can be charged through the USB connector (in fact, it’s the only option) whereas the Xoom actually has an annoying, separately coded software pop-up message that scolds you if you put in a certain subtype of USB lead. It says “do not try to charge through the USB port” – a state of affairs that it clearly anticipated, and specifically wrote some unhelpful bloody code for!

Pardon my Clarksonisms here, but in the midst of case design and some software features that deeply appeal to me, like the way that the DEFY and Xoom both just cope with my essentially random collection of music without bugging me about genres, Gracenote, or signing in to any online music store, Motorola can’t resist dropping in some woeful clangers, that could so easily be resolved.

The DEFY happily uses both the HT820 and the S805 as headsets and music-playback headphones; the Xoom won’t touch the HT820, representing it solely as a hands-free and continuing to play through its internal speakers – but it is happy with the S805. You would never imagine in a million years that the two headsets came from the same manufacturer, let alone the manufacturer of the device that they won’t link with.

I was, incidentally, borrowing the XOOM as a testbed for Citrix Receiver: a brilliant idea because it would have shown a Windows desktop on a touch tablet device. Less brilliant (as you can see from the picture with the phones) is that I can’t get a good shot off the Xoom’s screen, without appalling reflections of the testbed room here; and if I do a software snap of the screen via Android’s utilities, then all you see is a Windows Desktop. Hardly convincing…

Even with all this driving me nuts; I think I still want a Xoom…

Dell brought together a rich mixture of hacks and industry faces for a big announcement recently. It wasn’t a single product – no new laptop, no box to kick – but rather a whole slew of announcements that boil down to the simple statement that Dell wants to be a cloud provider in its own right. It has a huge preconfigured stack of servers, storage, switching and power which it can wheel into your datacenter on demand, called vStart, which takes care of the private cloud.

If that wasn’t quite enough for you then how about splashing out a billion dollars on cloud hosting centres? In the spirit of one of my all-time favourite books (the Tiger That Isn’t), I was minded to ask, almost the instant the announcement came from the lips of Brian Jones, head of public and large enterprise from Dell USA: a billion dollars, wow, but how much is that, really?

The money is going on datacenters. There are at least 10 candidate locations, though the Japanese earthquake has made the location of at least one the subject of one of those rushed sentences you can tell the speaker would rather you didn’t notice.  And some of the decision-making on locations for these centres has been about things that hardcore datacenter techies don’t like to see themselves bothered by, such as the legal position over government or company data and whether it can be moved outside the country it came from. A billion dollars spread across ten centres is $100m a centre (I worked that out on my own you know!) and that set me thinking: how big is a big datacenter these days?

I won’t give you all the research wedged into a few lines in a single blog entry. It’s a minefield out there, because almost nobody can see a clear reason to talk publically about just how big their datacenters are. Then there are the obviously fantastical made-up numbers for the amount of storage used by “them” to keep track of “us”: I heard the term “exabyte” used seriously as a measure of storage just this week, in connection with this kind of paranoid estimation.

Nothing in this world ages faster than investment in a datacenter

Staying away from things that aren’t finished yet, or things that are estimated but undeclared, I have found references from the reasonably recent past that suggest the top ten publicly visible data centres all across the world, are running at a rough average of $500m per location.

I’m sure there are more recent figures, or bigger deals, but the point I want to follow here has two aspects. The first is that nothing in this world ages faster than investment in a datacenter. According to Intel’s Alan Priestley, the new-generation Xeon CPUs coming on stream for rollout within vStart and other equivalent Dell platforms (including servers you can buy in the usual way) are almost 50% more efficient in power draw than their preceding generation, and 93% more efficient than kit from 2006. A funny year to pick, I hear you say – except that 2006 was the last pre-recession year, and represents the last hardware refresh for a hell of a lot of businesses out there.

Power draw

Ninety three per cent is about power draw, not about instructions-per-second or other measures of performance, and I bet it’s more to do with idle power draw than 100% loaded – but it’s also a startling figure. Work done in 8% of the time previously required, or 8% of the electricity needed, produces such a short payback on money spent that it tends to bedazzle finance directors, used to thinking about multiyear waits for the real benefit to emerge. Ninety three per cent, looked at another way, is a fourteen-times saving.

Now, I doubt that the entire of computing has shrunk at the same rate since 2006 as the power demands of an enterprise or cloud-ready server platform. Of course, disks have been getting bigger and memory has been getting cheaper per byte, but look at it like this: let’s meet in the middle and say that the scale-up advantage of starting to build a datacenter made out of all-new kit, in 2011, is seven times – that Dell’s money goes seven times further in 2011 than it would have done in 2006. This means, in terms of the kit they say we are still using in our server rooms – the quad-core, partly-virtualised architectures that the mid-field and trailing edge of the server farms out there are still using – Dell’s spend on cloud datacenters doesn’t look like a billion dollars. It’s more like seven billion.

There were parts of what Dell said which I found jarring: for example, that virtualisation was hard and lots of IT specialists were fighting shy of it. My initial reaction was if you are fighting shy of this technology then possibly you shouldn’t tick the box in the customer survey that says “IT specialist”.

I can find lots of people from each of the great virtualisation tribes who will tell you how easy it all is. Other speakers asserted that small businesses had no case to make for keeping physical servers because there was “no way they could meet the regulatory burden of keeping systems up to date”. Every small business I visit has no problem with this issue because they are frankly, utterly unexposed to whatever “regulatory burden” this speaker had in mind, and are very happy running their servers, sometimes for years without a break. But that’s a diversion.

What woke up my old banker’s mathematical instincts was the size of that investment number (a billion in today’s money: seven billion in 2006 money), and the track record Dell has for responding to customer demands, rather than taking bold and unsupported moves out into the unknown.

Surely, rather in the same way that a giant shopping centre development will advertise itself by the anchor tenants who sign up to being involved, even before the digger’s shifted the first cowpat, Dell must be propping up the $1bn with some big customer partners? Yes, it agreed cautiously, this is all based on existing customer demand. No, Dell’s spokesman said, he couldn’t say who any of those customers were.

All I know is if I had issued a ten-year loan back in 2005 or so to support one of the existing mega datacenters, then this announcement, based on all-new kit, would have me just a little bit worried.