Chief executive Marcin Kleczynski said: "Backup is an important pillar in security. Anti-malware software can only get you so far..."

Although backups can be used to rescue a device that has been compromised by malware, it's a problem if the malware is also in the backup. And since online files can be shared by sending someone a link, Kleczynski argues that the scanning process also helps users protect their reputations.

Malwarebytes Secure Backup can be used to back up files either to an external hard drive or by uploading the files to a cloud service. Backups can be made manually or completely automatically, and can keep unlimited versions of files that change. Deleted files are archived.

The backup software can be set to scan a PC for all the documents, photos, music and videos on a PC and upload them, or files and folders can be selected manually.

Both PC and mobile files can be encrypted using an AES standard system, in which case, the cloud service provider cannot read them. Nor can any government authorities who subpoena data. However, Malwarebytes warns that: "If you lose/forget your username/password combination, your data will be unrecoverable — there is no way to de-crypt data without that information."

The cloud service is provided by a third party, SOS Online Backup, which was founded in 2001. SOS's website says that it has "11 datacenters across multiple continents" and uses "military-grade security". It has offices in California in the USA, in India, and Ukraine, and uses datacentres in the UK, Australia, and South Africa, among other places.

SOS charges $99.99 for 100GB of storage in its Home and Home Office service, which is much more than Malwarebytes is charging. Also, SOS charges $20 per device for up to five devices, whereas Malwarebytes lets you "back up as many different devices and drives as you want".

Kleczynski said that Malwarebytes looked at several cloud services for encryption and security, including Dropbox, and SOS was "the fastest service we looked at".

Malwarebytes has been beta-testing the backup service before today's launch, and is targeting its services at individuals rather than companies.

Malwarebytes entered the business market last September with the Malwarebytes Enterprise Edition of its anti-virus product (see: Malwarebytes steps up to protect large enterprises), and Kleczynski is well aware of the high level of service required to satisfy that market. He aims to get the Secure Backup service running extremely reliably before thinking about a move to the next level.

However, SOS Online Backup already offers a business service with a centralised management console to manage remote backups.

There's a lot of competition at the low end of the online backup market, with Google's Gdrive and Microsoft's SkyDrive, as well as services such as Dropbox, Mozy, Carbonite, and many more. But Kleczynski reckons Malwarebytes Secure Backup is unique in combining anti-virus scanning with cloud storage, and will also appeal because it supports iOS and Android devices, as well as Windows PCs from XP SP3 on.

Malwarebytes Secure Backup has another advantage, in that the company can try to sign up millions of people who already use Malwarebytes Anti-Malware software.

The University of Cambridge's Computer Lab has celebrated its first 75 years with, among other things, a readable coffee-table book by Professor Haroon Ahmed. Although packed with historical information, it's more of a souvenir than an academic work, and the numerous photographs will certainly remind the thousands of staff and former students of the Lab's contribution to computing's rapid progress from EDSAC's valves and mercury delay lines to the diminutive Raspberry Pi.

The book is not too constrained by the Computer Lab's actual existence. It starts with two of Cambridge's greatest pioneers, Charles Babbage and Alan Turing, who are both seminal figures in computing. After this pre-history, it takes the 75 years in the title from the founding of the Mathematical Laboratory. Its two staff members helped university researchers tackle numerical problems using mechanical calculators and differential analysers.

The Mathematical Laboratory's greatest era began when Sir Maurice Wilkes returned from war service in 1945. He visited the USA to attend lectures on building digital computers, learned about ENIAC and EDVAC, and decided that Cambridge would build its own digital computer. The result, EDSAC, was the first programmable computer in general service, and several EDSAC users won Nobel prizes. It also led to the world's first business computer, LEO (Lyons Electronic Office). This was built for the tea-shop company, though Leo Computers sold systems commercially.

It wasn't until 1970 that the university set up a separate organisation to provide computing services, but by that time, the Mathematical Laboratory was no longer developing its own machines, such as EDSAC, EDSAC 2 and Titan. The university could buy them from IBM and similar suppliers. Of course, this didn't stop the lab from building useful and interesting research machines like CAP, the "Capabilities computer" used by Bjarne Stroustrup, of C++ fame.

More recently, the Computer Lab was transformed by Wilkes's right-hand man and successor, the late Sir Roger Needham. He supported a deal under which Microsoft contributed to the construction of today's purpose-built lab in the William Gates building, named after the Microsoft co-founder's father. More controversially, Needham also set up Microsoft Research in Cambridge.

The university's computer activities stimulated the development of a "Silicon Fen", and the book includes brief sketches of a few Cambridge-based technology companies that have some connection with the lab. This includes Acorn and its ARM spin-off, Sophos, RealVNC, Jagex (online games), blinkx, and Raspberry Pi. However, there doesn't appear to be any mention of the multi-award-winning software pioneer and entrepreneur Mike Lynch OBE, who did both his degrees at Cambridge and became a fellow of Christ's College. This is odd because Lynch gave a talk during last week's 75th anniversary celebrations.

Lynch built Cambridge's second most successful technology company, and is much more closely associated with the university than the most successful, which is ARM. Of course, Lynch sold his company -- Autonomy -- to Hewlett-Packard for $11 billion, which has resulted in some negative publicity, but he hasn't ceased to exist.

But the main problem with the book is that it's a bit of a patchwork. There isn't even a timeline to give readers not already familiar with Cambridge's history something to hang on to. Dip into it at random and you'll quickly find something entertaining to read, but I found that trying to use it for reference could be somewhat frustrating.

The publisher, TMI, has produced similar glossy tomes about Exeter College: The First 700 Years, The London Oratory School (150 years), UCLA (100 years), Charterhouse, Grosvenor House, Cowdray Park Polo Club and many more. Pre-order a book and you get your name listed in the back, as a Subscriber, and several hundred people are listed in Cambridge Computing. It's not what Cambridge University Press would do, perhaps, but it seems to work.

Cambridge Computing: The First 75 Years

by Haroon Ahmed

TMI

176 pages

ISBN: 9781906507831

40 plus P&P

During Friday's trading, the shares even hit $386.50. This is a long way from last September's peak at $705.07.

The fall has a knock-on effect. Companies are used to seeing their stock price increase when they become Apple suppliers, but they are now more likely to see it decrease. This may make them less keen to deal with Apple. It will certainly discourage them from devoting all or most of their production to the American company (see: Apple's dimming luster roils suppliers, investors).

All this means that Apple's next financial results, to be announced tomorrow (Tuesday, April 23), will be watched even more closely than usual.

Analysts already expect Apple's results to be weak, by its own extraordinary standards. According to Reuters, this could mean revenues increase by 8 percent and profits by just 2 percent. A much better result could change the direction of the share price, which is still trending down.

Either way, sentiment has changed over the past six months, thanks mainly to Samsung's success with Android phones. As Harvard Business School professor Clayton Christensen -- author of The Innovator’s Dilemma and the defining expert on disruptive innovation -- said earlier this year, Apple's proprietary devices are being disrupted by the open Android operating system. This is now good enough to serve buyers who were not being served by high-priced iPhones and iPads.

This represents a threat to Apple's future profits, on which its share price depends.

On Friday, IBM's shares also tumbled, registering their biggest one-day drop in eight years, following disappointing financial results. IBM's shares fell by $17.15 (8.3 percent) to $190.00.

The American company is reportedly negotiating to sell its $5 billion x86 server business to China's Lenovo, which has already taken over IBM's PC division.

According to the Wall Street Journal, "IBM pointed to contracts slipping from the first quarter into the second quarter, possibly because of the early Easter; a slowdown in business spending in China; and the weakening yen" to explain its weak results. In which case, there should be a rebound in the next quarter.

Nonetheless, when two of the leading tech companies see their shares fall, this casts a shadow over the whole industry. It is bound to raise doubts about claims that Dell is worth much more than its buy-out price, for example.

The price of Apple shares tumbled overnight and hit a new 52-week low of $400.78 during todaqy's trading: the lowest price since December 2011. The shares closed at $426.24 last night, Tuesday April 16, and are 5.31 percent down at the time of writing.

Apple's market capitalisation has fallen from a peak of $658 billion last September, when the shares hit $705.07, to $379 billion today. The loss of $280 billion is roughly equivalent to a dozen Dells or 30 Red Hats.

Analysts were already fretting about Apple's next round of financial results, which are due next week. See yesterday's post by Larry Dignam: Analysts start to sweat Apple's Q2, outlook

The new information that moved the market, according to Bloomberg, came from Apple's audio-chip supplier, Cirrus Logic Inc. Cirrus said its sales would miss earnings, and that it would record a net inventory reserve of $23.3 million. Bloomberg reported that "Most of that -- $20.7 million -- is from a high-volume product from one customer, Cirrus said, without naming the client."

It added: "Apple accounts for more than 90 percent of Cirrus’s revenue, according to supply chain estimates compiled by Bloomberg."

The supposition is that the pile of unsold chips would have gone into iPhones and iPads.

We are also entering the period when some buyers may be expecting Apple to introduce new products, and therefore defer purchases of these inessential items. Some expected a new iPad this month, and a new iPhone 5S could appear in June or possibly September.

Also, as Larry Dignam noted yesterday, if Apple launches a $300 (unsubsidised) iPhone in September, this could hurt sales of more expensive and more profitable products.

Finally, Taiwan's DigiTimes claimed today that "Upstream sources in the iPad mini supply chain expect a 20-30 percent decline in shipments for the [iPad mini] during the second quarter of 2013 due to lacking demand in the market."

Either way, Apple shares have long been a "momentum play", where people buy stock in the expectation that it will keep going up. This is what creates bubbles. Since September, however, the momentum has been in the opposite direction, with the big profits being made by short sellers.

Meanwhile Apple's real business, which nowadays is based on selling truckloads of iPhones and iPads, continues to do well. If not, we'll find out next week.

Previously: What you need to know about Apple's tanking share-price and tumbling market cap

There has already been concern about the massive amounts of electricity consumed by the data centres required to support cloud computing, but new calculations show that this concern may be misplaced. CEET's report, The Power of Wireless Cloud (PDF), reckons that these data centres will only account for 9 percent of the energy used for wireless cloud computing by 2015. The much bigger problem is the 90 percent required to provide wireless cloud access using Wi-Fi, 3G and LTE services.

Dr Kerry Hinton, CEET's Deputy Director, said in a statement: "When Greenpeace analysed cloud efficiency it hit a nerve with the likes of Google, Microsoft and Apple by suggesting that data centres are to blame for a 'dirty cloud'. In fact, the problem is much worse, data centres aren't the biggest issue. The trend towards wireless is the real problem, and the networks are to blame. By 2015, the energy consumption of data centres will be a drop in the ocean compared to wireless networks in delivering cloud services."

CEET's report projects that the "wireless cloud will consume up to 43TWh [in 2015], compared to only 9.2TWh in 2012, an increase of 460%. This is an increase in carbon footprint from 6 megatonnes of CO

The calculation is based on figures for the number of micro-Joules required to send one bit of information via LTE, typical power consumption levels for ADSL routers and so on, and independent projections of the growth in the use of cloud services. It calculates that cloud users will be transferring 23 exabytes per month by 2015.

It seems unlikely that the result is any more than a ball-park figure, but it does make a significant point: Cloud computing is promoted as being "clean" because the data centres are relatively efficient in their use of energy. However, the reality is that it's dirty. Very dirty.

Historically, most users accessed the internet and cloud services via cable connections to desktop PCs. Today, these wires have often been replaced by wireless connections that are slower, less reliable, less secure, and more expensive to run. And while it's sometimes possible to switch to a cable, an increasing number of devices such as smartphones and tablets don't even have ports for wired internet use.

The main benefits are, of course, increased freedom and mobility, including the ability to access the cloud from places that wired internet connections cannot reach.

Another benefit is the ability to access cloud data from any device. But while this may appear to be free, it carries a considerable social cost. It consumes far more energy to keep streaming or downloading and uploading files wirelessly than it does to carry the same files around and access them from the local device.

One of the social aspects that CEET does not cover is the short life span of wireless devices such as smartphones and tablets, which are often very difficult to repair. Smartphones that cost $600/600 may be junked after 24 months or less.

Although smartphone fans may be pleased that these devices could soon be selling 2 billion units a year, it's inevitable that we'll soon have to dispose of 2 billion mobiles a year. This is a waste of resources on a colossal scale.

CEET is a partnership between the University of Melbourne, Alcatel-Lucent (and its research arm, Bell Labs) and the Victorian State Government in Australia. It is dedicated to improving the energy-efficiency of telecommunications. Success also saves money, and should make telecommunications cheaper.

While the legislation is to be applauded, it's two decades too late to capture the early history of web development in the UK. Massive amounts of valuable data have presumably been lost forever, and there will always be a digital black hole in British history. The consolation is that the Internet Archive, founded by American digital activist Brewster Kahle in 1996, scooped up and preserved some of it in its Wayback Machine.

The British Library has been one of the UK's copyright libraries since 1662, which means publishers have been legally obliged to give it free copy of everything they print. This has resulted in a priceless archive, albeit one that takes up 500 miles of shelf space.

It would have been logical to make the BL similarly responsible for storing copies of web-based publications as well. If it didn't feel it had the legal right, or the money, the British government should speedily have provided both.

On a personal note, I challenged librarians to start archiving the web when I opened the Libtech 96 exhibition in 1996, but my pleas fell on deaf ears.

The government eventually gave the archive its blessing in 2003. Unfortunately, the project was then delayed for another decade because, in the words of the Financial Times: "To win support for the project, the British Library had to convince the UK’s biggest publishers that the archive would not damage their commercial interests."

This has resulted in restricted access to the archive. People will only be able to read the UK's digital archive while on the premises of one of the UK's six legal depositories, and only one person will be allowed to read a document at a time. This mirrors the fact that readers had to visit a copyright library to access printed publications, but is it a sensible use of current technology?

It's perhaps surprising that digital articles are not delivered to each user on a floppy disk, preceded by a flunky carrying a red flag.

As well as the British Library, the six legal depositories comprise the Oxford and Cambridge university libraries, the National Libraries of Scotland and Wales, and the Trinity College library in Dublin.

The digital system should save money because publishers will be able to supply digital files to these copyright libraries instead of paper publications, and the digital files will also be cheaper to store and retrieve.

The BL has selected a "Curator's 100" list of websites that shows an emphasis on cultural artefacts. The list includes the Beano comic, the Unst bus shelter in Scotland, Street Art London, Rights of Women, Mumsnet, The Dracula Society, and Daily Mash, as well as more obvious government websites.

The 100 list also includes shopping and advertising sites such as Amazon, Argos, eBay, Gumtree and Trip Adviser. It also includes Facebook and Twitter.

This has prompted some alarm, because people who make stupid remarks online -- there's no shortage -- may well find their comments preserved for the amusement of future generations.

The archive of e-books and publications should be accessible in the summer, and web content from the start of 2014.

British Library goes digital

Thousands of London bus users already pay their fares using contactless bank cards instead of TfL Oyster cards, which have been widely used over the past decade. Users pay different charges for different London Underground zones, so that's a trickier proposition. It's even harder to handle train travel, where the various different companies charge a confusingly wide range of fares, depending on the time of travel, the train used, and who is travelling.

Almost everyone who lives in London, and millions of annual visitors, use London transport, so TfL's choices are extremely influential. Once TfL's systems come into common use, other businesses will try to exploit the same payment mechanisms. TfL could therefore drive the widespread adoption of systems such as contactless EMV (Europay, MasterCard and Visa) bank cards, Barclaycard's PayTag stickers, and the NFC (Near Field Communication) system already built into many smartphones.

According to Dave Birch, a digital money expert from Consult Hyperion, TfL's gates (or barriers) can already work with NFC phones, but the system is not currently switched on.

Consult Hyperion has been advising TfL on its strategy and implementation. It also acts as a consultancy for many of the leading suppliers of payment systems, including Barclaycard, MasterCard and Visa.

Birch says travellers will probably be able to use bank cards for tube travel by the end of this year. This will be possible thanks to back-end processing that limits each user's total spending on TfL tickets to the cost of the applicable one-day TravelCard. Oyster cards already do this, but they only work in London.

"Ideally, what you want is one national card like they have in the Netherlands," says Birch. "That's very difficult to do in the UK because you have so many [rail] franchises and so many ticket prices…."

With complex train fares, which often extend beyond London, users will be able to buy tickets using their smartphones. In fact, a week ago, O2 launched a TrainTravel smartphone app that does this on Apple iOS, Android and some BlackBerry mobile phones, regardless of network. The payment is made via the O2 Wallet.

Birch says: "The idea is terrific, but if I buy the ticket on my phone, I still have to go to the station to collect my [paper] ticket." This is needed to get through the station barriers and onto the train.

He envisages a system where people can use their mobile phones as tickets, without needing paper versions.

Barclaycard's PayTag is another option. Basically, PayTag is a Visa card that comes in the form of a sticker that you stick to your phone. It works without PINs with today's contactless card readers, including the ones used on London buses, for amounts up to 20. It solves the problem of not all smartphones supporting NFC (and Apple's iPhone is a laggard in this respect).

Tags also enable tourists to use the system. Birch says they could buy PayTags or TfL tags at the airport, stick them to their phones, and use them as contactless transport tickets. If they choose a tag supplied by a bank or credit card company, they should also be able to use them for other purchases. PayTags are already supported by McDonald's, Subway, Boots and other outlets.

Since they are actually bank cards, tags don't need battery power, and will continue to work even if the phone is dead.

It remains to be seen whether Oyster cards will be withdrawn, because they do have advantages. The Oyster card is a "stored value card", which means TfL passengers pre-load it with cash which they can then spend. TfL has their money before they even travel.

Most importantly, Oyster cards can be read very quickly: in 300 to 350 milliseconds, according to Consult Hyperion. According to a slide shown by Cubic at the Transport Ticketing 2013 Conference, bank card transactions peak at 545ms, with smaller peaks at 420ms and 490ms.

TfL needs to be able to get 26 users per minute through barriers at peak times, and that's a challenge for NFC. Last May, Shashi Verma, TfL's director of customer experience, told NFC Times: "I’d love it to work. But it doesn’t work [for Transport for London] right now."

Also, Oyster cards are anonymous, and many users will be aware that using a bank card will mean their movements can be tracked.

On March 27 2013, TfL announced that it had issued an OJEU (Official Journal of the European Union) notice for the 1 billion Electra contract, which will "take the Capital's transport ticketing systems into the next decade". Whoever wins it will be responsible for "the provision and maintenance of front and back office revenue collection systems for all of TfL's services, including approximately 4,000 retail outlets, and at National Rail stations where Oyster is accepted".

The Electra contract will be awarded in October 2014, with a 10-month transition period to take over from Cubic Transportation Systems (unless Cubic wins the bid). Whichever company is appointed to collect TfL's fares is in for some interesting times.

Or maybe Microsoft could release Blue as Windows 8 SP1, or Windows 8.1. That really would be confusing because the version name — as distinct from the marketing name — is Windows 6.3.

The one thing we know for certain is that it will continue Microsoft's record of picking confusing names for its NT (New Technology) line of operating systems. Sometimes deliberately.

Historically, Windows NT was developed by Dave Cutler, former star programmer with Digital Equipment Corporation (DEC) and developer of its VAX VMS minicomputer operating system. Indeed, wags duly noted that incrementing VMS by one letter resulted in WNT.

Dave's new operating system was targeted as a 32-bit replacement for the floundering 16-bit OS/2 being co-developed by Microsoft and IBM, in which case it might have been OS/2 3.0 NT. After IBM and Microsoft fell out, it was quickly rejigged as a Windows-replacement system instead.

Microsoft made a hash of NT's version naming right from the start by launching it as Windows NT 3.1. This gave it parity with what was then the current DOS-based (old technology) Windows 3.1. It skipped versions 1.0. 2.0, and 3.0 before progressing to Windows NT 3.5 and NT 4.0.

While NT was matching Windows' numbering, the Windows team decided to switch from version numbers to years, so instead of Windows 4, we got Windows 95 in 1995. This was followed by Windows 98 and 98SE [correction] (Second Edition).

The NT team also switched to date numbering and grabbed Windows 2000. This left the DOS/Windows team with a quandary that it solved by releasing Windows ME (Millennium Edition), or Me. Happily, this was the end of the line. Smart companies were already migrating to NT4 and/or Windows 2000, and the rest weren't dim-witted enough to persist with DOS-based Windows forever.

Microsoft now needed to make Windows 2000 look a bit jollier and more like something consumers would buy, but Windows 2001 was a bit too close to Windows 2000. It therefore came up with Windows XP, which was short for eXPerience.

Windows 2000 followed NT 4.0, so it was really NT 5.0, and Windows XP became version 5.1 to reflect its new UI.

Windows XP united the market, replacing both Windows 2000 and Me. However, it was widely derided in the press and proved absurdly vulnerable to waves of malware, so it was as good as dead by 2003. Or maybe not. Microsoft patched it up with Service Pack 2 in 2004, and smart companies stuck with it until 2009-10.

The naming problem was now acute. Fill in the next term in this series: Windows NT4, Windows 2000, Windows XP. With no logic to follow except inconsistency, someone picked Vista. This was released as NT 6.0, reflecting the fact that Microsoft had redone the internal plumbing.

Enter Steve Sinofsky. He had run the Microsoft Office team, which had stuck with year names (Office 2003, Office 2007, Office 2010, etc). He decided to go back to simple version numbering, and since he was developing the version of Windows after version 6 (Vista), he called it Windows 7. Just don't look at the version number, because it's actually version 6.1.

Windows 7 was a huge success, so the next version had to be called Windows 8. The introduction of a new touch-oriented interface (Metro, as was), programming system (WinRT), new software distribution system (Windows Store), etc, would have justified a new version number, so Windows 8 could have been version 7... .

But it wasn't. Microsoft simply incremented version 6.0 (Vista) and 6.1 (Windows 7) to 6.2 (Windows 8). This explains why Windows 8 Blue might appear as version 6.3.

All this makes perfect sense, but only on the planet Zog.

If Microsoft had called Windows NT 3.1 by its (more or less) correct name, NT 1.0, then everything would be fine. After v1 (NT3.1), v2 (NT 3.5), v3 (NT 4.0), v4 (Windows 2000), v5 (Windows XP), v6 (Vista), and v7 (Windows 7), then Windows 8 would be version 8.

Still, I won't mind having a Windows Blue version. It's no worse than OS X Ocelot, Podgy Penguin, or Hungry Horace.

Correction, 10.13am PST: "98SE (Special Edition)" corrected to "(Second Edition)".

Update, 3.47pm PST: In a post today at Microsoft's Official Blog, PR supremo Frank Shaw commented on "a set of plans referred to internally as 'Blue'. NB: Chances of products being named thusly are slim to none".

IDC's Loren Loverde said in a statement: "Based on our latest quarterly figures, global PC shipments were expected to decline by 7.7 percent in the first quarter, as vendors and the supply chain work through the Windows 8 transition. However, our February monthly data suggest that we could see a drop touching double digits in the first quarter and a mid-single-digit decline in the second quarter before we see any recovery in the second half of the year."

China's slower-than-expected growth was "partially due to the timing of Chinese New Year", said IDC. There were also government budget cuts and anti-corruption measures.

IDC said: "March should recover somewhat in China, but not enough to offset the weak February results." However, PC market sales in Europe, the Middle East, and Africa (EMEA), Latin America, and Asia-Pacific (excluding Japan) were close to forecast figures.

IDC is currently forecasting that sales for the whole of 2013 will be just 1.3 percent down on last year's, and that sales will be up by 1.9 percent in 2017. In other words, PC sales will essentially remain flat.

The PC industry is now a replacement market, and most users see little reason to replace their current machines. Microsoft's major programs no longer need more powerful PCs to run them, and almost any PC bought in the past seven years can run Windows 7, Windows 8, and Microsoft Office perfectly well.

With Windows 8, Microsoft made an effort to provide the incentive for users to buy new PCs with touchscreens, and it also tried to expand into new markets by including a free tablet interface and software along with the established desktop interface. Intel also has an attractive new processor, code-named Haswell, on the way.

However, Loverde said: "Even getting to positive growth in the second half of 2013 will take some attractive new PC designs and more competitive pricing relative to tablets and other products."

But one thing that does bode well for future PC sales is the large installed base of business systems running the 11-year-old Windows XP. With the ending of Window's XP support, companies will have to tackle the malware assault without the benefit of security patches. This will increase the cost and reduce the benefits of running XP.

It already costs more to run 32-bit Windows XP than 64-bit Windows 7, because of its relative lack of security and stability and its inferior real-world performance (more delays in standard operations such as web browsing, more reboots, etc) even on the same hardware. Given the additional time savings from installing new PCs that have more modern processors with much integrated graphics, and the potential savings from reduced power usage, the business case is easily made.

Of course, not all businesses behave rationally even when it's in their own interests, and some are so cash strapped they're thinking only of survival. Following a triple- or quadruple-dip recession, the rump of XP users may be bigger than expected, but there should be enough upgraders to sustain the PC market into 2017.

"Even so, IT professionals everywhere are inclined to focus on patching Microsoft programs, operating systems and just a few other programs," says Secunia. "Ignoring the threat that vulnerabilities represent in non-Microsoft programs is both reckless and unnecessary."

Not only are third-party software suppliers responsible for the vast majority of security vulnerabilities, they are responsible for an increasing share. The 14 percent of vulnerabilities found in Microsoft programs and Windows operating systems in 2012 is a dramatic improvement on the 22 percent found in 2011. The number was as high as 43 percent in 2007.

The Top 50 most popular programs comprised 29 Microsoft programs and 21 third-party programs, giving Microsoft a 58 percent share of the software under consideration. Microsoft's 14 percent share of the vulnerabilities was made up of 5.5 percent found in operating systems and 8.5 percent found in other Microsoft programs.

Of the 21 third-party programs, 1,137 vulnerabilities were found in 18 products from eight vendors.

The bulk of the vulnerabilities were found in web browsers and software from the usual suspects: Apple and Adobe. Google Chrome led the way with 291 vulnerabilities, followed by Mozilla Firefox (257), Apple iTunes (243), Adobe Flash (67), Oracle Java JRE SE (66), Adobe Air (56), Adobe Reader (43), and Apple QuickTime (29).

Windows 7 had the most vulnerabilities among the Microsoft products (50), followed by Internet Explorer (41) and the .Net Framework (14).

All this warns against relying too much on numbers. Google Chrome is sandboxed to provide a high level of security, and it's updated very frequently. It's much safer than its score might suggest. By contrast, Oracle Java JRE SE has been a security disaster to the point where the most rational approach is to uninstall it, regardless of its lower score.

Secunia points out that Windows 7 had a high number of vulnerabilities last year, as "a result of the work of one security researcher, who decided to dig into one specific component, win32k.sys. By doing so, he discovered 22 vulnerabilities in 2010 and 59 vulnerabilities in 2011." Only four had been found in 2009.

No doubt all software contains bugs, if someone is prepared to dig deep enough to find them. Chrome's high number of vulnerabilities may therefore indicate that it's more secure because more bugs have been found and remediated.

On the positive side, Secunia reports that in 2012, 84 percent of vulnerabilities had a patch available on the day they were disclosed, as compared with 72 percent in 2011. Morten R Stengaard, Secunia's Director of Product Management, said: "This means that it is possible to remediate the majority of vulnerabilities. There is no excuse for not patching. To take advantage of this improvement in patch availability, organizations must know which programs are present on their systems and which of these programs are insecure, and then take an intelligent and prioritized approach to remediating them."

In all, Secunia reported a total of 9,776 vulnerabilities in 2,503 vulnerable products from 421 vendors.

Needless to say, it may only take one unpatched vulnerability in one program to compromise a company's security.

The Copenhagen company gets the bulk of its data from its free Personal Software Inspector (PSI) program, which is installed on millions of Windows PCs (including mine). These PCs have, on average, 72 programs installed. Secunia says these programs vary "from country to country and region to region" so it's simpler to focus on the 50 most common ones.

PSI makes regular checks to see if a PC contains any programs that do not have the latest patches installed, and makes it easy for users to patch them. This is important since not all vendors provide scheduled updates, and they may not notify users when patched versions are released.

Secunia also sells a Corporate Software Inspector (CSI) and is currently beta-testing a small business version of its product.

Individual links on the OneTab page can be restored by clicking on them, or they can all be restored by clicking "Restore All". Sets of links can also be saved to disk or loaded using the Export/Import commands, or they can be shared by publishing them as a web page. Since the Export command produces a plain text file, it provides a simple way of copying a set of tabs from one PC to another.

OneTab thus provides a way for creating and sharing a set of websites that can all be opened together. To do this, open the websites you want to save, then click the OneTab button to create a page of bookmarks. You can now reload all the sites at once by holding down the Control key (or the Command key on Macs) while restoring them. In this case, the bookmarks will be kept on OneTab instead of being deleted.

You can drag-and-drop the bookmarks on OneTab to change the order in which the sites are loaded.

It doesn't matter if you close the OneTab tab or close the browser: OneTab keeps a consolidated stack of tabs, and the latest set is loaded when you restart the browser.

OneTab only takes a second or two to install, and it's also easy to remove: just right-click the icon and select "Remove from Chrome".

OneTab may not appeal to people who have modern PCs with 8GB or 16GB of memory. However, many business users are still running 32-bit Windows XP on PCs that are limited to about 3.5GB, and they may only have 2GB installed. In these cases, 1GB or 2GB of Chrome tabs can cause real performance problems, and OneTab offers a quick solution. It's certainly quicker and easier than bookmarking each site separately.

I tested OneTab on an antique Windows XP desktop PC with 3GB of memory installed. I had 32 tabs loaded in Chrome, which came to about 1.3GB, with Chrome running 37 separate processes, including extensions. OneTab reduced this to five processes consuming about 120MB of memory.

To see what Chrome is running, press Shift-Esc to bring up its Task Manager. For more detail, click the link in the bottom left: "Stats for nerds". This produces the same display as typing "about:memory" in the address bar (no spaces).

Chrome has a built-in "Bookmark All Tabs" feature, but it's clunky in comparison. You do it by right-clicking on any tab and then selecting that option, or by pressing Ctrl+Shift+D. However, you may have to open a new folder to save the bookmarks. It's tedious to retrieve them, and you probably didn't want to add 30 or 50 permanent bookmarks. anyway. Also, closing and then re-opening tabs separately is very tedious. OneTab is quicker and easier.

Of course, you can also save and reload sets of Chrome tabs using other add-ons, such as Session Manager, Session Buddy, and FreshStart from the Chrome Web Store. Many of these offer more features. However, OneTab does what most people need with only two clicks (one to create a tab, one to restore), and it's both free and ad-free.

In answer to the obvious question — "How do you make money?" — the site says: "OneTab is free of charge and is not designed to make money. It was created because we badly 'needed' it for our own use, and we wanted to share it with the world."

SketchInsight is, essentially, an intelligent whiteboard that responds to drawn commands. For example, if you want a pie chart, you draw a circle, and if you want to change it to a bar chart, you draw a bar. Drawing an L shape will create a graph with X and Y axes, while handwriting "ye" or "pop" will label the axes by finding "years" and "population" in the data, as Lee's demo shows. In effect, it looks like a graphical version of Autocomplete.

Microsoft said: "Creating personalized, expressive data charts becomes quick and easy. The presenter simply sketches an example, and SketchInsight automatically completes the chart by synthesizing from example sketches."

Of course, it would be possible to produce the same sort of illustrations using PowerPoint. However, SketchInsight enables presenters to create illustrations as and when required, and change them with a single touch. For example, in a presentation about regional sales, the presented can eliminate areas by touching them on a map or create new illustrations in response to questions.

Although the big-screen demo shows SketchInsight being used for a presentation, the same ideas could presumably be used to explore and analyse datasets on a tablet PC such as Microsoft Surface Pro. I'm sure anyone who has struggled with Excel charts (eg, me) would welcome something that is actually easy to use. However, the demonstration video doesn't specify a particular system, nor does it explain how datasets are prepared and loaded.

The video can be viewed at or downloaded from the Microsoft Research page: Telling Stories with Data via Freeform Sketching. It's also available on YouTube.

Klein derived his "8 percent" slogan and #demand8 hashtag from Boston Consulting Group figures for 2010, "the latest we have". These showed that the UK's internet economy accounted for 8.3 percent of GDP, which was the highest in the world. The UK was followed by South Korea (7.3 percent), China (5.5 percent), Japan, and the US (both 4.7 percent).

Klein said the internet economy is already bigger than education, health, and construction as a percentage of GDP, and not far behind the financial services industry.

This was good, but "there's something wrong in the city", said Klein. The UK's financial industry ("the city" for short), managed funds worth 658 billion, but it isn't being used to back the UK's digital industries. "Why is this money sitting on the sidelines of the internet economy?" he wondered.

Klein pointed out that US giants such as Apple, Amazon, eBay, Google, and Facebook are generating 15bn worth of revenues in the UK.

Some established British companies are successful in the digital economy, said Klein, and he cited Pearson publishing (33 percent digital) and WPP advertising and marketing (30 percent) as success stories. Tesco, by contrast, is making only 6 percent of its sales online, "but at least they could answer the question", Klein said. Not all supermarkets could or would.

To build a "balanced internet economy", Klein reckons that all the UK's quoted businesses should get to at least the "8 percent" level, and be prepared to explain why they have not reached it. However, he would prefer quarterly monitoring, so that investors can act accordingly: "celebrate the champions and short the laggards". He said: "No one wants to be the next Kodak or HMV."

Of course, the "8 percent" slogan is already out of date. Klein said the internet economy should contribute 12.4 percent of the UK's GDP in 2016, when it would be worth 225 billion to the economy.

Early in his keynote, Klein had pointed out that Europe's economy is doing very badly, and suffers from very high youth unemployment. The unemployment level for under 25s is about 18 percent in the UK, 57.6 percent in Greece, and 56.5 percent in Spain. Young people need to be given the skills and encouraged to contribute.

Failure to engage with and invest in the new internet economy is not just an immediate problem, but also one with long-term consequences.

Klein is a London-based venture capitalist at Index Ventures and co-founder of Seedcamp, a startup accelerator. He has been involved with investing in a lot of early stage internet companies including MOO, Last.fm, Tweetdeck, AlertMe, Chartbeat, GlassesDirect, MyHeritage, Soluto, and Songkick. Earlier, he worked for Skype, and he founded what is now Lovefilm International.

While Klein must be applauded for putting his money (and other people's money) where his mouth is, the British are a cynical lot. The UK has been a world leader in other high technology industries. These include computing — where LEO (Lyons Electronic Office) was the first business computer — and aviation, where the De Havilland Comet was the first commercial jetliner in production. Converting innovation into large-scale industry is a different matter, but if the UK fails again, it won't be Saul Klein's fault.

Update: Saul Klein has now posted a video of his keynote speech at the London Web Summit on his blog at localglo.be, along with his slidedeck at Slideshare.

BT plans to launch its own TV Sports package this summer, based on its rights to the Premier League, FA Cup, Scottish Premier League, UEFA Europa League, and German Bundesliga football (for American readers: soccer). BT's sports package will also include some live Premiership Rugby and WTA tennis matches.

The acquisition means that BT will also be able to offer sports provided in the UK on the ESPN America channel. These include NCAA College Basketball, NCAA College Football, and NASCAR racing. BT may run these on an ESPN-branded channel.

The channels will be broadcast from a new BT Sport studio in the Olympic Park, London.

The deal does not include ESPN's web properties ESPN.co.uk, ESPNcricinfo (cricket), ESPNFC (football), ESPNscrum (rugby), and ESPNF1 (Formula 1), or its video-on-demand service, ESPN Player. It also excludes TV channel ESPN Classic, which looks unlikely to survive.

Although ESPN has a strong position in sports broadcasting globally, it has struggled during its four-year attempt to break into the UK market. It arrived perhaps a decade too late to compete effectively with Rupert Murdoch's BSkyB (Sky), which has been broadcasting since 1990. While they always looked doomed, both ESPN and the failed Setanta were trying to capitalize on the European Commission's attempt to break Sky's monopoly by preventing it from having exclusive rights to Premier League games. (Obviously, the European Commission will not cover any losses incurred by suppliers or consumers.)

When buying Premier League rights, BT said in a blog post at BT Vision: "BT will succeed where Setanta and ESPN failed because it is a much bigger business with millions of existing customers, and because the communications giant will have top quality games to sell."

The BT Vision business unit only has about 700,000 customers, compared to Sky's 10 million, but there is a significant market in broadcasting live sports in British pubs and clubs.

For BT Group, sport is simply part of a larger play for the broadband internet market. BT is currently rolling out FTTC (fibre-to-the-cabinet) broadband services in competition with Virgin's less widespread cable TV network, which carries Sky's channels at a substantial extra charge, and FTTC broadband.

Virgin Media is based on a conglomeration of numerous local cable companies. It was formed via the merger of two large competitors, NTL and Telewest, which then merged with the phone company Virgin Mobile UK. Earlier this month, US-based Liberty Global, owned by John Malone, said it had agreed to buy Virgin Media in a cash and stock deal worth $23.3 billion (15 billion).

BT, the privatised phone giant that used to be part of the British government-owned postal service, will now be competing against two risk-taking multi-billionaires: Murdoch and Malone. Both of them control global empires. BT, by contrast, is less powerful than local European rivals such as Deutsche Telekom (which owns Orange and T-Mobile) and Spain's Telefonica (which owns O2 and operates across South America). However, Ofcom — the UK telecoms regulator that limits BT's prices — seems unlikely to pay any attention to BT's global position, rather than its local one.

The service will help people who use mobile phones or tablets for banking, though they could simply use it for news or to catch up on their Twitter or Facebook feeds while waiting in a queue.

Last year, BT and Avaya commissioned some Youbiquity finance research, conducted by Davies Hickman Partners Ltd with 2,000 consumers. This showed that "free in-store Wi-Fi access was high on the most-wanted list for consumers in the UK, Germany, Spain and USA." The result: "In the branch, 41 percent want access to free Wi-Fi." The UK figure was 38 percent.

The research found that, despite the growth in online banking, 64% of customers still wanted to use bank branches. In particular, they want them to solve problems (as opposed to using call centres) and to make purchases. These visits are characterised either by their "high value" (the purchase of financial products) or "high complexity" (problem solving).

Another factor is that "people don't have a high level of trust in financial services providers, particularly in the UK (31 percent)."

The research also found that customers needed help with online banking, and would like to get it via web-based chat (28 percent), click-to-call (21 percent), co-browsing (13 percent) or video chat (12 percent). At the moment, bank branches don't help with online banking problems, but they could.

In a statement, Ashok Vaswani, CEO of Barclays Retail and Business Banking, said:

"We’ve been listening to our customers about how we can make banking easier for them, and they tell us they like the idea of mobile banking but would often value being shown how it works. We are always keen to show our customers how easy our mobile apps are to use, so we are now rolling out free Wi-Fi across our Barclays branches, and our front line staff say this will help them give product demos using our customers own mobiles and tablets. We hope by providing this free service our customers will be able to instantly download and use all our mobile apps, and we are on hand to show them how."

BT says it ran trials in three branches of Barclays, and has now rolled out the equipment to more than half of the 1,500 branches. The rest should go live later this month or by the end of March 2013.

In February 2012, Barclays launched Pingit, as "Europe’s first person-to-person service for sending and receiving money using mobile phone numbers". A free service, it works without sharing bank details. Pingit requires a UK current account and UK mobile phone number.

This is a major limitation if you share pages from magazines or brochures, or tweet infographics. Text may be unreadable at 600px.

Twitpic support told me: "we've found that the majority of our users do not use the full size image and to conserve resources and keep Twitpic a free service, all images are being resized to 600px."

This doesn't make much sense. Users have to click the "View full size" button to see an image full size, and if they don't, few extra resources are used. (And if they do it, Twitpic gets to show them ads.) In any case, the smart way to conserve resources would be to limit the size of the file, not the size of the image. Twitpic allows images up to 10MB, so multi-megabyte gifs and videos are likely to consume a lot more resources than still images.

I have suggested that Twitpic limit still images to a megabyte in size, and resize larger pictures to 1280 x 768 pixels. This would fill the average user's screen, and be better than some competitors. However, I've also been looking for an alternative service.

Twitpic's key features are:

(1) Twitter shows thumbnails of Twitpic images under tweets on its website, so they are more likely to be viewed than images that open a new page. This eliminates Instagram and many other photo services.

(2) Twitpic keeps a running total of how many times an image has been viewed, which provides feedback. Twitter's own photo-sharing service, based on PhotoBucket, does not.

(3) Your Twitpic profile shows the last 20 images you have uploaded. Someone who comes to look at one image may look at several.

(4) It's easy to use.

(5) You can sign in using your Twitter account, so there's no password to remember.

There are two obvious alternatives, the first of which is Yfrog. This is not quite as easy to use, and in my experience, images don't download quite as quickly as they do from Twitpic. However, Yfrog images show up in Twitter, and it does keep a running total of views.

The main problem with Yfrog is that there doesn't seem to be a way to save a photo without tweeting it. This means you can't upload a handful of photos at once and tweet them later, as needed.

The second obvious alternative is our old friend Flickr, the popular photo-sharing website taken over by Yahoo. It may be a disadvantage that you cannot log in with a Twitter account, but you can log in with a Yahoo account, and surely almost every IT person has one of those.

Flickr photos show up inside Twitter, which is the main requirement. It also records the number of views, though not in the most obvious place. Visitors can view or download a larger image, if they can figure out how to select "View all sizes", and they can also see previous images in your photostream. Flickr's main drawback is that you can't download original images (not even your own) without paying for a pro account. However, the images are large enough for most purposes.

You can share Flickr photos to Twitter, Facebook, Pinterest and Tumblr by selecting the service you want from a drop-down menu. This is a quick way to get a shortened link to tweet, even if you elect to do that separately.

Anyone who runs company accounts on several social networks should probably be managing a Flickr account as well.

So why don't more people tweet pictures from Flickr?

I suspect that this is due to Flickr's origins as a site for storing and sharing personal photographs. Everything you upload is marked as your copyright, and you can apply whatever rights you want. This doesn't appeal to people who are simply sharing photos they have found online, and are making no claim to ownership. Of course, the reverse applies to companies sharing their own copyright images.

Tweeting from Microsoft's SkyDrive is another option, though Twitter users may not know how to access the full-sized image. Clicking on the photo in Twitter pops up a larger version but in degraded quality. The solution is to click on the photo's filename (in this case, W_Hotel-DSC_0296.JPG), which takes you from Twitter to SkyDrive, where you can select either View Original or Download.

Most people have Hotmail accounts -- or use the Hotmail-replacement, Outlook.com -- and this provides instant access to SkyDrive storage (and web-based Microsoft Office applications). The only drawback is that SkyDrive doesn't track the number of views, as far as I can see, so there's no way of telling whether an image is popular.

I uploaded the same photo to all four sites and then downloaded it to compare results. The original was 4288 x 2848 pixels (2.90MB). Twitpic reduced this to 600 x 398 pixels (73K), which is inadequate. Flickr reduced it to 1024 x 680 (360K), which is acceptable for viewing on screen. Both Yfrog and SkyDrive downloaded full-sized originals, though only SkyDrive kept the original filename.

I also uploaded my photo directly to Twitter. There isn't a download button, but I saved the large image and deleted Twitter's broken file extension (.jpg-large). It turned out to be 1024 x 680 (143K).

To sum up, Twitter provides a quick way of sharing photos that you don't own, don't want to track, don't need to keep online, and will never use on other social networks ... as long as you don't care too much about the image quality. Twitpic is still very good for sharing photos that are up to 600px wide, but not larger. Yfrog is a good substitute for Twitpic, but Flickr may be better, depending on your needs.

If you actually want to share original images via Twitter, SkyDrive is the best option. However, Yfrog works well enough, and counts views.

Her Majesty's Revenue & Customs (HMRC) reckons the switch to monthly reporting will save money. The problem is that, according to a DTE Business Advisers' survey of 126 SMEs, "96 percent of companies or their clients had no knowledge of the changes, or those who do, had no idea of its implementation".

The new system, called RTI (Real-Time Information), comes into force in the next tax year, which starts on 6 April, 2013. It is needed to support the switch to another UK government idea, Universal Credit.

HMRC has posted an online guide, Preparing to operate PAYE in real time (RTI), on its website. It includes a page on Payroll options and software packages. Unusually, it has also published a colourful infographic showing RTI stats & figures (I've posted the large version online).

In this case, "real time" means "within seven days".

Companies have had advance warning, and ZDNet published a substantial story more than two years ago: The future of PAYE: How the 66-year-old tax system is changing. HMRC has also been running a pilot project with 500 employers.

But only three months ago, the Institute of Chartered Accountants in England and Wales (ICAEW) said the new PAYE requirements were "at best unrealistic and at worst impossible".

Currently, HMRC is writing to all the UK's employers to remind them of the impending change. Firms that do not comply will face penalties.

The system should mean that employees pay the right amount of tax throughout the year, instead of adjusting the total once a year. It should also ensure that HMRC gets its cut, though this may make life difficult for companies with cash-flow problems. (You can underpay HMRC and catch up at the end of the year — if you have the cash. We may therefore see more companies go bust quicker.)

HMRC reckons it will collect an extra 3bn during the year, rather than having to wait for the end of the year.

HMRC also says that "RTI will increase the accuracy of HMRC's information about tax credits claimants, enabling better detection of fraud and error in the system and potentially saving the UK hundreds of millions of pounds."

Finally, HMRC reckons that employers will save "300m in reduced administration costs once the system is up and running". Since there are roughly 5 million private businesses operating in the UK, this represents a saving of 60 each. Given the hourly rates charged by chartered accountants, it's hard to see this as a good deal… except for chartered accountants and software houses.

Going private will mean Dell and his management teams won't have to worry about Wall Street's quarterly scrutiny, with its concern for short-term numbers and lack of interest in long-term strategies. In theory, Dell will be able to restructure and re-orientate the business in private, without it turning into the sort of media-driven horror-show exemplified by companies such as Hewlett-Packard and Yahoo.

Going private probably means one thing, and definitely means another. It probably means that Mike and his backers feel that Dell's shares are undervalued. It definitely means that Dell will be back on the stock market when it can sell shares for significantly more than the $13.65 (plus costs) that it is paying for them now.

This might happen in two or three years, but I suspect five years is more likely. It could be 10 years. But it has to happen because Silver Lake and the banks will want their money back, plus profits.

One of the problems with Dell shares -- and this also applies to Microsoft shares -- is that they were massively overvalued in the first tech bubble at the end of the 1990s. Dell shares increased in value by 16,000 percent, as shown in the graph below. You think Apple shares boomed when they increased in value by 4,000 percent? The graph below shows the difference.

The problem is that Dell's 2000 stock price was based on a fantastic boom in PC sales in the 1990s, and the anticipation that Dell would rake in profits for the next decade. No sane person is looking forward to a similar boom in PC sales over the next 10 years. Dell's PC profits will be dramatically lower, and this results in the lower share price.

So why is Mike making such a big bet on the future?

The answer is that Dell is moving into enterprise software and services, which generate better profits than PCs. To speed up this transition, Dell has been buying other companies. These include EqualLogic in 2008, Perot Systems in 2009, KACE Networks and Boomi in 2010, and Compellent in 2011. Last year, Dell picked up AppAssure Software, SonicWall, Clerity Solutions, Quest Software, Gale Technologies and Credant Technologies.

In sum, Dell is building a company to compete against IBM, HP and Oracle in the enterprise market. Mike will presumably try to avoid competing against the Asian tigers -- Samsung, Lenovo, Acer and Asus -- in the PC market. However, this does not mean Dell will stop making PCs. PCs still sell roughly a million units a day, and businesses depend on them. Dell will aim to sell fleets of thousands of PCs and servers that it can pre-configure, install, and maintain as a service. This provides less money up front, but more profit in the long run.

Dell's last 10-K regulatory filing gives us a good idea of the state of the business a year ago. In fiscal 2012, Dell sold $8.3bn worth of servers and network products, $1.9bn of storage, $8.3bn of services, and $10.2bn of software and peripherals. That's around $28bn. It also sold $14.1bn worth of desktop PCs, and $19.1bn worth of "mobility" products (notebooks, mobile workstations, smartphones, and tablets). The total came to $62.1bn.

For comparison, Dell's turnover is roughly the size of HP's Enterprise business unit, which brings in about $57bn. Dell can grow if it can profit from HP's woes, and vice versa, of course.

It remains to be seen whether the company founder can do it. He certainly knows the PC business, and there was a time when he was a hero. In my Computer Weekly column in March 1999, quoting someone I couldn't name, I wrote: "How can I do what Dell does? I wanna be like Mike! That's what most computer company bosses say every morning, if only to themselves." That was before the crash. Nobody wants to be like Mike today.

The fact is that Mike's not doing so well. In 2004, he stepped back from the business, and Kevin Rollins was CEO until the start of 2007, when Mike took back the reins. Maybe he didn't think Dell was doing well enough, but look at the graph of Dell's annual turnover, below.

Dell started from nothing and zoomed past Apple, particularly after the mega-launch of Windows 95. There was a glitch after the millennium financial debacle, but on the whole, Dell did pretty well to grow its annual turnover from $0.5bn in 1990 to $31.89bn in 2000 and then to $57.42bn in 2007. After that, growth disappeared, and Dell increased turnover by just $5bn in five years.

It's true that Mike has had to battle against a flat-lining PC market and a huge global recession, but it's also possible to imagine that Rollins would have done a better job.

Dell's growth obviously doesn't compare with Apple's, but then, nobody's does. In any case, Apple didn't get big by selling computers, and its actions show no real interest in corporate and professional computing.

Should we blame Mike because Dell hasn't made any visible progress in selling smartphones and tablets? Possibly, but HP didn't do any better by buying Palm and having an embarrassing flop in the tablet market. IBM and Oracle didn't even try.

Can Mike transform Dell into an enterprise powerhouse in the next five years?

Maybe, but I don't think the shareholders would bet on it. And now that Dell has raised enough cash to take the company off the market, they won't have to.

Almost all Samsung's products are used in business, including in-house training, and in higher education colleges. However, it's making a push into the schools market by developing software that enables many of its devices to work together. Smart Education is being launched in the UK this month, and the company ran hands-on demos of its Smart Classroom Solution (sic) for teachers who attended last week's BETT educational technology exhibition at London's ExCel.

Samsung said it was piloting Smart School in schools in 24 countries to "fine-tune our software." The most advanced projects include schools in its home city of Seoul, South Korea, in Sydney, Australia, and in Memphis, in the United States. (Last year, Samsung USA made a YouTube video about its school.) It has now started a pilot in the UK but says it is not ready to name the school.

This is just the start of Samsung's "big vision for 2020."

In its Smart School pilot projects, Samsung has been supplying teachers and some whole classes with their own personal devices. These include the multi-touch Series 7 Slate running Windows 7, and Galaxy Tab or Galaxy Note tablets running Google's Android. The Slate also works with a digital pen, as do Samsung's Note devices.

BC Cho, from Samsung's headquarters in Seoul, said "pen is a real differentiator for us," and that when devices have handwriting support, "note-taking is way easier."

The school software suite includes Samsung's Interactive Management Solution (IMS), a mobile Learning Management System (m-LMS), and a Student Information System. The overall aim is to provide efficiently managed, technology-based personalised learning (and sell lots of Samsung devices).

The Smart Classroom Solution "is a complete digital education package consisting of Samsung tablets, a server and software", says Samsung. Using this Android-based system, teachers can watch their students' screens remotely and lock them if necessary. They can also display students' screens on an interactive whiteboard. (Windows 8 client support will be added later.)

Although Samsung is not ready to talk about its pilot school in the UK, it already has other partnerships.  One is with Birmingham Metropolitan College, which has over 40,000 students -- mostly adults -- spread across a number of sites. The college's Mustafa Shevket said it had invested more than 10,000 in Samsung equipment (including a Samsung Surface) for a learning centre in Kidderminster. "All the staff have been issued with tablets, and there are eBoards in all the curriculum areas," he told me.

Samsung will not be providing content for Smart Schools, but it has partnered with Houghton Mifflin Harcourt, which Tim Cannon, HMH's executive vice president for strategy and alliances, described as "the world's leading educational content company."

Cannon said: "It's only through technology that we're going to change education".

The goal was "a personalised educational experience" so that students could follow their own courses and sit customised examinations. Education would be student-centric instead of teacher-centric.

"Android and Windows 8 are going to be the prime operating systems in use," said Cannon, but content should be "operating system independent, with a minimum form factor." (Phones have small screens, most Android tablets in education have 7-inch screens, and Windows machines typically have 10-inch to 30-inch screens.)

Read this

IT skills in 2013: What's really in demand?

• Read more

Of Samsung, Cannon said: "We've been working together closely for 18 months. You ain't seen nothing yet… That's an Americanism."

Graham Long, vice president of Samsung UK and Ireland, said: "We've been in the education market for a very long time, and we bring great products to market. But increasingly what you will hear from us is: it's about solutions."

It remains to be seen whether UK schools will adopt the system. However, pricing deals, and reducing the number of different suppliers, could make it attractive. Taking a global approach will certainly give Samsung huge economies of scale, and an advantage in dealing with large American partners such as Houghton Mifflin Harcourt and Microsoft.

Samsung wouldn't put any financial numbers on its ambitions, but it aims to get more than 20 percent of the global education market by 2015.

Schools normally turn to local suppliers for localised equipment and content, but that approach is threatened by globalisation. British schools no longer use mostly British computers (originally from Acorn, Amstrad, Sinclair and Research Machines), and the internet is now providing a global market for online courses. It seems that purely local UK suppliers can look forward to being squeezed...

Chrome was trendy. It was relatively light, fast and very stable, while Firefox had grown somewhat old and bloated. (Perhaps my copy was running too many badly written extensions.) Chrome's market share increased rapidly from 11.41 percent in February 2011 to 19.11 percent in December, on NetMarketShare's numbers, while Firefox slipped a fraction from 22.69 percent to 21.83 percent.

As for Internet Explorer, its market share had plunged from 59.22 percent in February 2011 to 51.87 percent in December. The gap between Chrome and IE had closed by 15 percentage points in only 10 months! It only had 30-odd points to go!

IE was in trouble. It was carrying the combined weight of Silicon Valley and open source hatred, and even if some of those are idiots, mud tends to stick. Further, the "browser choice" screens introduced in March 2010 at the behest of the European Union's Department for Bashing Microsoft could persuade some users to try other browsers. (Later, we learned that Microsoft had sometimes -- whoops! -- accidentally left browser ballots out.)

Even worse, Microsoft seemed to have shot itself in the foot by refusing to provide a "modern" browser to run on its horribly insecure, hopelessly outdated and yet still hugely popular Windows XP. Of course, users could upgrade from IE6 (which was a very modern browser in 2001) to IE8, but that was still unable to handle full HTML5 or run JavaScript-bloated websites at speed. (Not surprisingly, Google's stuff runs best in Google's browser.)

With Google running persistent adverts for Chrome on its high-traffic search page, as well as expensive TV adverts, XP users would surely switch in droves….

But the January 2013 numbers (graph below), which NetMarketShare published on February 1, show that I would have lost both bets.

Chrome had climbed impressively to Firefox's level at around 20 percent market share, but then more or less stopped growing. Worse, it had suffered a slight decline to 17.48 percent in January 2013, behind Firefox's 19.94 percent. (I certainly don't think NetMarketShare's numbers are accurate to two decimal places, but the long-term trends are probably close enough.)

Also, while IE had declined steadily from 66 percent in 2009 to 51.87 percent in December 2011, it did not continue on the path to around 40 percent or so today. Its market share stopped falling, and it has even gained a few percentage points. It's at 55.14 percent today.

IE and Firefox have certainly improved, and perhaps Chrome is less impressive than it used to be. In my experience -- where it's usually overloaded with 35 to 50 tabs -- Chrome's built-in Flash crashes often, and the whole thing falls over perhaps once a day. I'll live with that because I have more faith in Chrome's security than in any other browser, but I can understand people with different priorities switching away.

Whatever the reason, predictions that any of us could have made based on "obvious trends" have, so far, failed to materialise. Chrome has not humbled Firefox; it has not overtaken IE.

As physicist Niels Bohr observed: "Prediction is very difficult, especially if it's about the future." I think I'll stay on safer ground: predicting the past.

Footnote: As Larry Deneberg has pointed out, that Bohr quote has also been credited to Confucius, Yogi Berra, Will Rogers, Mark Twain, Groucho Marx, Woody Allen, Victor Borge, Albert Einstein, Enrico Fermi, Vint Cerf, Freeman Dyson, George Bernard Shaw, Sam Goldwyn, Cecil B DeMille, Winston Churchill, Dan Quayle and many others, possibly because they are easier to spell than Niels Bohr. Or possibly because most people don’t know nuffink about nuffink, they just make stuff up. I just wish more of them would admit it. You can quote me on that.