And another thing 61
by Jon Honeyball
Jon Honeyball has still to be convinced when it comes to Windows 8.1 and Windows RT.
HardCopy Issue: 61 | Published: November 1, 2013
Well, TechNet has gone. I wish I understood the mindset behind this move, but despite its public handwringing I still don’t understand Microsoft’s decision. They say it was costing money, that it was a problem for licensing piracy, and probably made you come out in spots and set fire to your hair. All of which seem to me like good reasons to fix something rather than just walk away.
Microsoft does a fantastically bad job at managing the relationship with its developers, and it cannot complain when we dare to gripe. I accept that TechNet has changed in size, shape and scope over its lifetime. And that maybe it was time to come with something better for the latter half of this decade. But retiring something with so much brand name awareness and good will in the community seems to me a particularly crass thing to do.
I had to decide recently whether to renew my Visual Studio with MSDN subscription. I confess that the cost of nearly £3,000 made me pause and consider what value I have been actually getting from it, and for the first time in very many years, I decided it was time to walk.
Developer support is something incredibly valuable to Microsoft. I can pay a pittance to Apple to get support for its developer tools and OS platforms. I accept, of course, that the breadth and scope of the Apple platform doesn’t begin to scratch the surface of what Microsoft offers, especially in the corporate business server arena, but that’s not the point. Doubtless Microsoft would reply that there are many very low cost developer offerings, with the Express products being free. All of that is true. But I still cannot shake off the thought that Microsoft has lost the plot if it really thinks that it can get developers to cough up nearly three grand a year.
Where do we go with Windows 8.1? This is a topic which is troubling a lot of people at the moment, including a lot of developers. The truth is that 8.1 is a considerable improvement over Windows 8.0. I know we like to make the joke that it takes Microsoft three versions to get something finished: The first is a stab in the dark, the second rights most of the major wrongs, and the third version adds the final polish. Windows 8.1 follows this trend with somewhat disturbing regularity.
The upside of this is that Windows 8.1 on Intel has matured into a product that can be happily deployed to almost any user, whether they be on a desktop machine, a laptop or a touch pad. The ability to boot into the Win32 desktop solves much of the primary concerns for traditional desktop users. The improvements in the touch side is welcome for the more avant-garde operators.
And let’s not forget that the hardware is evolving at a rapid rate too. Intel’s Ivy Bridge technology wasn’t bad, but Haswell is considerably better: better performance, better power management, better battery life. All of these address many of the stark limitations of the hardware from just one year ago. And the roadmap indicates that Intel will continue down this path with ever improving products over the next year or two.
So if you want to deploy Windows 8.1, and to do so on anything other than a standard desktop machine, it makes a great deal of sense to go for the very latest hardware to ensure you get the best experience.
But what of Windows RT? Well, it seems that this is still struggling on, although the market seems to have decided it’s not worth pursuing. At the time of writing, only Microsoft appears to have any RT devices at all, and has announced the second version of its Surface product. Dell has just cancelled its device, which leaves Microsoft in a group of just one. Rumour suggests that Nokia might launch an RT-based product soon, but given that Nokia is fast becoming a division of Microsoft, it’s hard to see how this is any different to being a ‘real’ Microsoft product in all but name.
The RT situation is really quite upsetting. It should have paved the way for very long battery life devices, with serious and significant built in security capabilities designed to keep your computing experience safe, just like an iPad. But the reality is that this touch tablet focused OS ended up needing the old Windows desktop alongside the new Metro UI.
I cannot let the release of Leap Motion pass without comment. This tiny device plugs into a spare USB port and gives your computer a three dimensional space in which you can move your hands, point with your finger, and all sorts of interesting gesture control. With the right software, it is incredibly immersive and fascinating to use, although you have to learn a few techniques like tucking your thumb into the hand unless you want it seen as another finger.
Many have already written off this device, and I think it’s very unfair. For sure, the initial kit of software didn’t give much scope for joy. And trying to use a three dimensional multi-finger cloud-space to control something so resolutely 2D and flat as a Windows desktop seems to be an exercise in futility to me. Windows really has no concept of depth at all, and its z-order is really limited to ‘front’, ‘back’ and a fairly unspecified ‘in-between’.
But in the right context, the Leap Motion controller is utterly compelling and captivating. That’s the whole point – we need to develop new classes of applications where data can be both visualised and manipulated in entirely new ways. This is not an overnight process, but there appears to be a quite vibrant community working with this platform today. And much of what they are bringing out is genuinely thought provoking and interesting. Even Hewlett-Packard has been tempted to bring out a laptop which has the Leap Motion controller embedded into it, alongside a conventional keyboard and trackpad.
It costs very little to get hold of this, and to explore the SDK. If you have an interest in leading edge UI designs and interface possibilities, then it’s definitely worth a look.
The reason for this? Well, nowhere near enough work had been done on the management and system configuration side of Windows 8 to allow you to ignore the historical if powerful configuration and management tools. Worse still, the Office team had failed to deliver on a Metro version of Office for RT, and hence demanded that they should be allowed to ‘simply’ recompile, making it easy for them to port over at least some code. When this wasn’t available to third parties who were just as deserving, is it any surprise that developers shunned the platform?
It’s really no better today with Windows 8.1. Despite 8.1 being much more complete in its Metro based implementation of setup and configuration, the desktop is still there and it’s still only available to the Office team. Well, given that RT is now a sole Microsoft platform, the Office team can at least claim it’s all their game and their rules and their ball.
How well this will fare in the marketplace is far from clear, but if past history is anything to go by, most buyers will simply walk on by. The newer Intel platforms are starting to give ARM a run for its money. For corporate developers, it’s a real game-over situation. I know that some will claim that the Exchange Server based management tools are enough for the sort of light-weight arms-length style of oversight, but for the more seasoned and demanding system administrator, nothing less than full Active Directory support is sufficient. And that immediately rules out Windows RT, because it doesn’t support AD.
Despite my misgivings and some (still highly justified) contempt for Microsoft’s attempts to bring Windows 8.0 to market, it is clear that 8.1 on the latest Intel hardware is starting to offer something of value to the corporate IT world. Security, manageability, and a coding platform that is well known are all compelling advantages. That’s especially true if you decide to write a Win32-based touch application that looks and feels just like Metro. Or go for the WinRT runtime and programming environment if you feel that it is mature enough to support your application needs.
For the broader home and home consumer market? The verdict is still out and I think it will take more than 8.1 to bring them on board in significant numbers. In that space, iOS and Android still rule, and are attracting the developer dollars.