The last Windows largely deployed?
Every now again I find James Gardner's blog posts insightful and provocative. The recent post The last Windows ever is the latest in a running line of doom mongering for Windows.
Since I’ve been working now with Windows Azure, some say is a new Windows Operating System in the cloud, the brand Windows has become more diverse and dare I say stronger. It’s exciting the changing pace of this cloud-scape and arguably Azure, as a Platform as a Service, delivers more features and more guarantees that the alternatives.
So, in context, is the last (desktop) Windows ever to be largely deployed going to be Windows 7?
I’d reason a case for No. Here’s why….
There are certainly more devices now than there ever has been. When I started using early version of Windows, most would agree it wasn't until 3.11 that the average person began to see that standardisation was a good thing. DOS had witnessed this standardisation process and so too did the GUI. It wasn’t about code purity, it was about ease of use, despite hardware differences. This version of Windows commoditised, what was at the time, difficult tasks (such as networking), multi-tasking and later in Windows 95 plug and play.
Its standardisation that's meant that people can get on with this jobs without being hindered with unnecessary road blocks. The likes of shims and workarounds are limited. But I’ve observed that its also competition that drives openness, feature parity and innovation.
These big platforms as James mentions are 10 years shifts, one technology pushing out the other. I don't know if this trend will continue, maybe this period will get shorter with new Cloud supporting services, but I do hope we don’t move back to a world where have to learn the interface for every device we interact with.
If you were to ask if IPhone 4 is to be the last iPhone OS ever widely deployed I would suspect the answer could be the same. No. Its effectiveness as a Phone relies on the ability to widely deploy new software to the device.
My points are therefore three fold, Windows will be even more pervasive through Azure, it will continue to standardise otherwise complicated technology, and device software depends on wide distribution, deployment and use.
Another article sums it up - the desktop in general needs to be compelling again. The encouraging news, and I have this on good authority, Windows 8 will be quite different….
Comments
Anonymous
November 08, 2010
Hello - and forgive me for "doom-mongering" Windows. My intent wasn't to do so, but I was speaking from the experience we're having trying to get the thing deployed for 140k or so desktops. It is too hard and it is too expensive to get Windows 7 out the door for any estate of any size. And then its much too expensive to keep going. Too expensive, in comparison to a nice browser deployment, that is. What I was attempting to say in my post was that other paradigms exist, and they look very, very attractive at present. I just don't think it will be possible to write £200 million pound deals (our most recent one) again for tin+operating system. The value is moving elsewhere. Also note that I said it would be a decade or more before we can unhitch ourselves from heavy desktop stacks, so I wasn't predicting anything very imminent. That we need to unhitch ourselves is very obvious though. We need to be in a position to take advantage of new paradigms, which I'm sure you'll agree are going to appear and be workable in the timeframes I'm talking about. James.Anonymous
November 09, 2010
James, as always I agree with much of what you have said! You write some excellent provactions though ;) Definately new UX paradigms are needed to match new experiences. I love the iPad as a concept but I find all the tablets really lacking (both in hardware and UI). I think you are right about the big deals. Office 365 is an example where costs are coming down, packaging is becoming the norm and subscriptions are the way forward. Lets hope deployment keeps up. My experiences with putting Windows 7 on a Mac Air demonstrates how hard it can be to image a machine.