MS is betting the house on Windows 8, but they know there will be growing pains. This $40 upgrade strategy is one way to mitigate the risk of alienating corporations by driving adoption in the existing consumer market.
Yes, when you combine a non-PC (the iPad) with the PC market there is growth. When you subtract the non-PC from the "PC market" there is negative growth. Trying to make the iPad's success somehow a sign that the PC market is growing is the funniest thing I've read on HN to date.
And sure, the Surface might invigorate Microsoft, and prove a worthy competitor to the iPad, both in terms of sales, as well as in terms of capability and usability. But regardless of whether it succeeds, it has no bearing on growing the PC market, nor on whether current users of PCs will upgrade to Windows 8.
The important difference is 'internet/media consumption device' (iPad) vs. 'general computing and productivity device' (PC, Surface?). I believe the growth and shrinkage of their respective markets indicate the typical consumer's desires.
Microsoft failed in previous tablet efforts by trying to graft the desktop environment onto a touchscreen. Failed miserably, and not because of hardware immaturity. In contrast, Apple has succeeded immensely by discarding the desktop metaphor and embracing the touch environment exclusively. Microsoft is trying to straddle the two environments solely to prop up Windows, not because it's a better UX. This to me, dooms it to failure, just as Microsoft failed in its previous tablet ventures.
The reason tablets are separated from the PC market is because tablets are a different and competing product! And make no mistake, this OS price is not for tablets. What Metro tablet am I going to buy that doesn't... have metro on it?
Do you have a source for this? I'm pretty sure it's growing. Just not as fast as tablets/smart phones.
One of the initial blockers was hardware - we were updating our computers in cycles rather than all at once, so many of them were simply not powerful enough to run Windows XP. You could easily forsee a delay of several years due to this alone.
Then you have the domain controllers. We actually stuck with Windows 2000 Server on them for a while, but really they should have been upgraded to server 2003 to be properly managing Windows XP clients. Upgrading the server OS means training your network admin staff, moving over any configuration or scripting that relies on deprecated things, re-creating your software deployment chain to use the latest features available to you, and so on. For us this was maybe a couple of man months of work.
Upgrading the server OS in the MS stack typically means you're simultaneously upgrading a lot of the other software too, such as the email server, so there's all the testing you would expect from such an initiative.
Next you need to re-create your base image that will get installed on all the computers. This could require months of testing because it's quite hard to get a base image running on all of your computers if they aren't identical hardware. This part also covers the sort of things you would expect in doing an upgrade on your personal computer, such as finding new device drivers (probably not an issue going from Windows 7 to Windows 8). I think this part took us 5 or 6 iterations, which worked out as about one man month of work.
Also on the desktop side your admins have to know how to use the new OS - a lot of the control panel changed in XP for example so we had to learn all that. (Of course by this time XP had been out a while and we knew most of it quite well already)
Once you have a working base image you need to update your group policy, as a new OS brings new settings, so you have to ensure everything is suitably locked down.
Then you have to actually test doing mass roll-outs of that image and applying the policy and make sure it works across your entire spectrum of hardware (multiplied by number of policies you have if they differ across some hardware, multiplied again by differing user policies).
Next up is making sure all your software actually deploys correctly on your new images. This is a great opportunity to upgrade to the latest versions of any software that's lagging behind a little (which may well be necessary if any of your software fails to run on the new OS). This was probably the most tedious and frustrating part, especially when applications aren't available as .msi installers as that makes the process for deploying them far lengthier.
Then of course you need to test all the software you deploy to make sure it still works.
Finally you can train users, decide on a good downtime window to do the deployment (and make sure you have a rollback plan for when it inevitably goes wrong), and then roll it out!
Second, most of our software is custom/old-as-dirt. Upgrades require direct participation of vendors who have us over a barrel, and for that reason, no incentive to move quickly. It's an extremely painful/slow/expensive process.
IMHO - Big Law is due for disruption in this area. NetDocuments appears to be making some headway. Unfortunately, lawyers and, thus, law firms, are extraordinarily risk averse. Startups will have difficulty capturing this market.
This is the killer. I'm part of a large (6,000+; 1000+ in my state alone) engineering firm; our IS team are currently about to roll out Win7 at the end of the month. We have a lot of software—particularly drafting and modelling software—that we rely on (with few alternatives), and all of their related plugins.
It's not that we don't want to move forward, but the cost of buying new tools (that may not be compatible with our clients'), re-training staff, etc - is possibly greater in terms of lost productivity than just dealing with XP.
Hardware isn't a problem though, as any PC from the last 5-6 years can run Win7 in an "office" environment; a bit less if you need to run modelling software. We usually upgrade on a 2-3 year cycle and a lot of staff are now moving from C2D/4GB/HDD machines to i5/8GB/SSD machines. Everyone loves the SSD's.
7 in terms of stability perception is like XP - I still have XP on my Mac Book dual-camped partition.
And we had real problems with Vista - rendering, certain apps not working, slowness overall.
How can something like that have room for improvement? Copying a file from one place to another is and has been a solved problem for decades now.
Nothing is straightforward in software. It takes a lot of work to make it look easy.