James O'Neill's Blog

December 23, 2010

On the future of tablets …

Filed under: General musings,Mobility — jamesone111 @ 9:07 pm

After the last post – and conversations with several people, notably Mark Wilson I’ve been thinking about how Microsoft’s mobile offerings might develop now that Windows Phone 7 is out in the market. As I’ve said before Microsoft talk about slates as if the ideal is a more portable PC.  Early in 2011 Intel’s  “Oak Trail” Atom Processors will deliver better battery life than the X86/X64 platform has known to date, but I doubt that we’ll get a sub 750g  slate with X86/X64 CPU AND all-day battery anytime soon. 

The usefulness of a personal computer – whether it runs Windows, MacOs or Linux – stems from its working with all the programs, data  data and devices (collectively “stuff” ) that you want it to.  Tell a user of one OS they would be better on another and the response and reason for staying is “the stuff I want is here, not there” and that is as true of company IT people defending using Windows XP past the end of its support life as it trying to persuade a Mac die-hard to get a Windows PC.

I read something recently to the effect that if usefulness was only a matter carry-ability and battery life the calculator is more useful than an iPad.  It sounds facetious; but taking a iPad for the subset of work that a calculator can handle  would be overkill. But it applies in the other direction, the attraction of a mobile personal computer is that you can take all your stuff with you, but that too is overkill if all you intend to do is the work that can be done on a slate device.

It was 2007 when I first told people to expect  what we’re now calling “Windows 8” to be launched in 2012. More specifically the slides said “Point release” – Windows Server 2008 R2 [And client Windows 7] – after 2 years, major release after 3 more.  The first part of that came true and the second hasn’t been un-said so I laugh when I hear“late 2012” quoted as news. Developers should get the OS a year before shipping so after the Professional Developers conference late in 2011 we’ll know if things are on course. I’ve seen suggestions that Windows 8 will break cover, on a slate, at CES in January 2011 – but anything beyond “let’s show you some ideas we’re working on” would surprise me.  When it appears, if there is nothing to make it a better slate platform, then everyone will be surprised. I’ll be interested to see if there is a break from the past Pen-Based tablets. Handwriting input failed on the Apple Newton, it failed on the Pocket PC (I found it worked OK, but no-one took to it), we may be approaching the moment to call time on handwriting input for PCs as well.  

Just as an aside, I wish Microsoft would do two things to make “instant on” a non issue. First the video below shows unmodified Windows 7 can boot from cold in 10 seconds, when freed from a legacy bios.  I’d tell OEMs now that computers with legacy BIOSes won’t get a Windows 8 logo

Secondly, what is the point of a start menu action  “Reboot, cutting the power at half way”. Because this action is labelled “Shutdown” and millions of people think that Shutting down is the right thing to do. Millions of hours get wasted each year through unnecessary reboots, if Microsoft cut the to options  “Sleep” , “Deep Sleep” and “Reload Windows”, a lot of people will cease to care about boot time.

Other people are lining up a bigger task for the Windows team, getting Windows on a long-life slate by porting it to the ARM processor. That’s the wrong answer. Drivers can’t be CPU independent, applications can in theory but many aren’t are in practice. (The iPad gets criticised for lacking support for flash, but Adobe still don’t have an X64 bit Windows Flash player, so how long would an ARM one take ? ).  If Windows-on-ARM can’t pick up all the stuff from Windows-on-X86 then it is a new platform, and wouldn’t  a new platform for mobile devices be better off not starting with 25 years worth of baggage from Windows (we accept that on X86/X64 PCs because without it we don’t get access to all our stuff).

Microsoft have supported the ARM processors for a over a decade on Mobile devices and both Windows Phone 7 and Zune devices use it. Embedded Compact 7  – the latest incarnation of of Windows CE for Mobile devices was announced earlier this year but devices using it have huge rarity value. There is confusion whether these bits underpin Windows phone 7, but there is a standard .NET way of asking what the underlying OS is: on a Windows PC the easiest way to use it is from Powershell

>  [system.environment]::OSVersion

Platform ServicePack Version     VersionString
-------- ----------- -------     -------------
Win32NT              6.1.7600.0   Microsoft Windows NT 6.1.7600.0

So I wrote a little code to test this on the phone emulator.

OperatingSystem myos = System.Environment.OSVersion;
OSPlatform.Text = myos.Platform.ToString();
OSVersion.Text  = myos.Version.ToString();
OSString.Text   = myos.ToString();
Firmware.Text   = Microsoft.Phone.Info.DeviceExtendedProperties.GetValue("DeviceFirmwareVersion").ToString();

and here is the result.


It doesn’t call itself “Windows Phone” but Windows CE and is quite clear the it is V7 of CE  – it seems that Microsoft producing an OS for ARM based slates running with something as close as possible to Windows Phone 7 and/or Zune HD is both practical and sensible.

Do slates actually matter ?

I mentioned Mark at the start his post started with a view of someone else’s that “the consensus [at a Gartner symposium] is the iPad is either a consumer product or an executive toy” which he disagreed with. Mark takes a more rounded view than just the iPad and quotes Forrester estimating that tablets will account for 20% of PC sales by 2015. It’s not clear if that is saying 1 in 5 new PCs will be tablet PCs (running Windows 8.1 by that point) or if One in 4 PC users will also have a slate device running Windows CE/iOS / Android.

He also quotes a different analyst’s forecast that of 48 Million iPad sales by 2014. Against an installed base of something like a billion PCs it’s tiny.  (It also suggests that there’s not much consensus between analysts, but that’s not news). Apple’s shinny products have made it the darling of Wall Street;  Apples Market value is 20 times its profits, Microsoft’s 12 times , if Microsoft halved that gap, its shareholders would have a heck of a reason to be cheerful.

But will emerging devices have much a role in corporate IT  ? Mark has a wonderfully utopian vision
“What will change (and is changing already) is the type of device that is used to access the desktop. Rather than taking a notebook PC from place to place in the device-centric manner that we do today, enterprises will adapt to human-centric computing models, with end users increasingly accessing their desktop from a variety of devices – perhaps starting out with a smartphone on the way to work; switching to a hosted virtual desktop in the office; using a tablet during meetings; and perhaps using the family PC to finish up some work at home in the evening, with local desktop virtualisation opening up new options for secure computing away from the corporate network”

Truly, I wish I could share this vision. The idea of “Bring your own computer”, which Mark also talks about, is  easy to implement with PCs: acquire a PC connect to the network, install a corporate image, and you have a managed PC in a known state with all your software. How many organizations have got the IT department out of the process ? It’s rare bordering on unheard of. Organisations do allow mail on user phones, but extending that to Bring-your-own-Mac, or Linux PC or even a Windows PC running the newer Windows and Office you use at home (never mind embedded devices) requires a revolution, but talking about that will have to wait for another day.


  1. I think you’re absolutely right about the whole “speed of booting” issue. It’s like selling cars and saying “Now, oil changes take only 90 seconds!” — yes, a seven hour oil change might be bad, but just how often do you change the oil. Reboots should be similar, something that happens once in a blue moon as part of machine maintenance.

    As for the future of tablets, I think it’s really hard to have a crystal ball at this point. Looking back to the past, there have been a number of upstart technologies that have seemed ill-suited to “full-blown” computing (however it was defined at the time) and that have gone on to overtake their more capable progenitors. Minicomputers overtook mainframes in mindshare, and then they were overtaken by desktop micros, and now desktop micros are in many ways overtaken by laptops.

    Ten years ago, most people still wanted or needed desktop computers. Today, in the consumer space laptops are prevailing; a desktop seems clunky and anachronistic, locked away in its own special room of the house, aloof and asocial. And in many areas of business, people expect to be able to bring their computers with them to meetings, etc., so laptops prevail there too.

    In raw numbers, desktops and laptops may hold the edge over “mobile devices” today (depending on whether our count restricts us to “smart” devices or not), but it feels to me like we’re seeing a similar shift in “where it’s at”, and that shouldn’t be a surprise…

    The drum beat of Moore’s law (and technological advances in general) means that we keep getting more and more computing power packed into smaller and smaller devices. It is that process that keeps changing the computing landscape.

    The smaller size of phones and tablets, changes them. They have the compute power of the workstation-class computers of yesteryear, but old-style interaction schemes don’t really work on devices that size, and so new schemes need to be found.

    What made the iPhone work (and lead to Android and Windows Phone 7 copying it) was the idea of radically simplifying the user experience to something much easier to work with than what had gone before, and touch interaction is part of what made that possible. Sure, underneath everything is a lot of computing power, but that power is harnessed in a way that is fundamentally natural.

    In my opinion, it is those same principles that make the iPad as successful as it is. The device itself isn’t sold as a “computer” (computers are, after all, scary and complicated, need to be managed by experts, prone to viruses, get “messed up” — at least in the minds of most people). Like an iPhone, an iPad is a thing, like a TV or a toaster. It’s pretty much impossible to “mess up” its software. There’s no registry you might need to edit, no antivirus to run, and no need to explain the difference between sleep and hibernate.

    And touch interaction turns out to work really well for a lot of tasks. Things people would never have wanted to do on a laptop or desktop, like drawing and painting, like making music, and so on. It has an immediacy and an intimacy that a laptop just doesn’t have. Some people have called it “frictionless” pure software. (Take a look at any touch-based game, or music software like the iMS-20 MS-20 synth emulation, and you’ll see what I mean — you’re not “using an iPad”, you’re using a synth or playing a game. The device becomes whatever the software makes it..)

    Today you can still buy compute-servers you’d keep in a cooled machine room and log into remotely, but those machines don’t dominate the computing scene any longer. You can still buy desktop machines, and plenty of people do, but they too are looking long in the tooth. You can still buy laptops, and lots and lots of people do. But if the balance shifts and some of the people who used to use laptops switch over to using tablets with touch interaction, I won’t be surprised.

    Today, for some “serious work” sessions, I plug my laptop into a larger external display, keyboard, etc. Maybe tablets will go the same way, and use some kinds of add on functionality to support those occasions when touch interaction on a small screen doesn’t cut it. (After all, today, a first generation iPad supports external keyboards and can drive an external screen, (or more esoteric things, like MIDI devices) so it is not such a far fetched idea.)

    Fundamentally, trying to predict how this stuff will work out in the future is difficult. When micros first appeared, it would have been hard to imagine what they would develop into. When laptops first appeared, it would hard to imagine how they would evolve. And I think the same is true of tablets today. We have version 1 of the iPad, and a few Android tablets, but in a year or two, they will at the very least they’ll seem slow and lacking in capability as an first gen iPhone does to an iPhone 4 today.

    One thing I am fairly sure of though, is that if these devices do take off, ten or twenty years later, something else will come along and make these things seem clumsy, awkward, and lacking in intimacy at a fundamental level.

    And with that (sorry this comment wound up so long!), I’m off to play with my iMS-20 app. Thousands of dollars of late 1970s synth and sequencer technology, easier to use, in something I can hold in my hands while walking. I don’t know where the future’s headed, but I’m very much enjoying the present!

    Comment by M.E.O. — December 24, 2010 @ 1:13 am

  2. Hi James,
    Thanks for the links/quotes – and for this post – it makes a good read.

    I’ll leave the distinctions between Microsoft’s various operating systems (and the mess they are in in the mobile space because they have let the market run away) to yourself and others but I’d like to pick up on some of your other points.

    I’m not sure that BYOC is quite as difficult as you might think. We have technologies like NAP/NAC to segment networks and enforce health. We (IT depts) can still enforce organisational policies (and arguably should) for BYOC (just as we do for company car schemes to ensure a suitable vehicle choice). Indeed, I attended a Citrix/Forrester webcast only last week looking at the practicalities of implementing BYOC in the enterprise and it’s not really a technology issue – the hard parts now are people and process (I was pleased to see that it wasn’t just pushing Citrix products!).

    The main point here is abstraction – OS and hardware are pretty much commodities. At TechEd Europe 2009, I heard Microsoft talk about how the desktop delivery method is less important than the management of the desktop (based on a whole bunch of analyst reports) and all we are doing with BYOC is abstracting the apps and data from the device.

    With apps, we might look to enterprise app stores on Windows. VDI/hosted shared desktops are probably also part of the mix at least in the short term (maybe with local client virtualisation on devices that support it) – it shouldn’t really matter if the OS is designed to be device independent (and I almost did that with Windows XP in 2002 – we did need two builds because of the two HALs but we had a much larger variation in our desktop hardware). All of that assumes that the corporate desktop still runs on Windows, perhaps in some form of container – which gets over the legacy app issue – but, where the device supports it, why not use native apps (iOS, Windows Phone 7, Android, whatever) to access core functionality (mail, calendaring, web access, line of business apps) alongside the icon to access a full desktop where required.

    As I see it the main issue right now is that the Windows user experience with touch is horrible (to clarify, Windows Touch is fine for specially designed apps, but awful with the raw operating system).

    For data, whilst I might want all of my data with me on a device, my CIO almost certainly wants me to take as little as possible outside the datacentre and so anything that is on the device needs to be secured somehow in an encrypted container in case the device falls into the wrong hands (again, technology does exist for some platforms – others may be more problematic – which is why we still need IT policies with BYOC).

    I could go on, but this is a blog comment not a BYOC white paper – hopefully you can see that a) this idea of BYOC is not quite as utopian as it first seems; and b) tablets/slates are just another endpoint device – albeit a highly portable one with instant on and all-day battery life.


    Comment by Mark Wilson — December 24, 2010 @ 9:07 am

  3. @ MEON
    The distinction between laptop and desktop is a now a very fine one. Unless you want a high end gaming machine or technical workstation with >2 cores you can ample storage, compute power and graphics in a latop. In fact I’ve bought a laptop and built a desktop in the last few months, and the laptop is more Powerful.

    Now the question on the future of tablets is do they turn into “Personal computers” – that is, things to which you can connect all your devices (Video and DSLR cameras, TV decoders, DVD players/Writers, smart card readers) and run legacy software, new software etc. [In which case WINtel is the way to go] or do you keep a machine which you blow the dust off when you need those things and use the tablet for the 10% of tasks which take up 90% of your time. [In which case Windows CE/Phone / iOS / ANdroid is the way to go.] I don’t think it is clear, so Microsoft should have a bet on both.

    @Mark . BYOC is easy, and a good idea. We don’t have it because we have IT departments. Its very easy to say that’s just a people and process problem, but those are harder to solve than technology problems. If your NAP/NAC system depends on an agent which isn’t available for a device, or on group policy when the device can’t join a domain it’s easier to say no.
    I see more organisations with Line of business apps so dreadfully written that they can’t get onto either of the last two versions of Windows. Do I expect them to have versions of those same apps to run natively on iOS/WP7/Android. Hell will have frozen over and thawed out again first.

    Comment by jamesone111 — December 24, 2010 @ 10:37 am

  4. James, I work in an IT department – and my CIO is only too aware that we have to embrace consumerism in the enterprise – in fact, he sees IT consumerisation as a key theme for 2011 – BYOC (even if it’s only bring your own mobile device) is a part of that. For BYOC, the IT department’s role is one of governance – we shouldn’t prevent people from accessing the systems that they need to be effective in work – we should enable them, but we also need controls in order to prevent total IT anarchy!

    As for policies, don’t just mean Active Directory Group Policy 😉 I mean that I can say “you can bring your own computer as long as it meets or exceeds this specification and you run one of the following approved security suites” – and I need to include HR and legal teams as well as IT as part of the BYOC implementation – after all, if we’re giving people money to buy a device for end user computing, there’s a whole load of employment-related issues to consider.

    As for appcompat – it does remain an issue – but that shouldn’t stop the development of new LoB apps that allow access from modern devices. It’s the appcompat issue that means we’ll be tied to Windows for a while longer (albeit old versions) and that’s why I suggested that VDI/hosted shared desktop solutions are probably part of the mix.

    Signing off for Christmas now – have a good one 🙂


    Comment by Mark Wilson — December 24, 2010 @ 11:38 am

  5. […] James O’Neill found out that WP7 does indeed run Win CE 7 when he ran his test code on the phone emulator. This opens up exciting possibilities for future hardware based on the OS which include the following features from a previous post. […]

    Pingback by Windows Phone 7 is based on Win CE 7 — December 24, 2010 @ 10:01 pm

  6. […] uitgevoerd in de emulator en gaf als resultaat dat het OS zichzelf identificeert als Windows CE 7, schrijft ontwikkelaar James O’Neill in een blogpost over tablets en Windows op […]

    Pingback by ‘Windows Phone 7 is gebaseerd op Windows CE 7′ | Portfolio — December 26, 2010 @ 9:38 am

  7. […] uitgevoerd in de emulator en gaf als resultaat dat het OS zichzelf identificeert als Windows CE 7, schrijft ontwikkelaar James O’Neill in een blogpost over tablets en Windows op […]

    Pingback by ‘Windows Phone 7 is gebaseerd op Windows CE 7′ » Clippy.be — December 26, 2010 @ 9:52 am

  8. […] uitgevoerd in de emulator en gaf als resultaat dat het OS zichzelf identificeert als Windows CE 7, schrijft ontwikkelaar James O’Neill in een blogpost over tablets en Windows op […]

    Pingback by ‘Windows Phone 7 is gebaseerd op Windows CE 7′ | Webtechnologie — December 26, 2010 @ 10:42 am

  9. […] su un emulatore del telefono che il WP7 effettivamente esegue il Win CE 7. Potete leggere sul suo blog quanto è emerso con i test eseguiti da lui e qui di seguito in sintesi le caratteristiche tecniche […]

    Pingback by Windows Phone 7 è basato su Win CE 7 | Windows Phone Magazine Italia — December 27, 2010 @ 11:19 am

  10. […] blog di James O’Neill, si evince che Windows Phone 7 è in realtà basato su Windows Embedded Compact 7, ad oggi non […]

    Pingback by Windows Phone 7 è basato su Windows Embedded Compact 7 - Around .Net — December 30, 2010 @ 12:44 pm

RSS feed for comments on this post.

Create a free website or blog at WordPress.com.

%d bloggers like this: