The unbearable misery of upgrading to Windows 8.1

There are some events that we humans know from experience will be painful: Childbirth, passing a kidney stone, or getting a colonoscopy without anesthesia. After last weekend, I can now add, without hesitation, upgrading Microsoft Windows. I've been doing Windows upgrades since Windows 3, and it's always been relatively straightforward: You either do a straight upgrade that includes everything, or you do a clean install of the new version, restore your files and reinstall your applications.

Things got muddled with Windows 7. I, like many people and companies, decided to skip the abomination against nature that was Windows Vista and stay with Windows XP. That turned out to be the right decision, but when Microsoft introduced Windows 7, even though it really wanted XP users to upgrade, it prohibited direct upgrades from XP. That meant that XP users could move files, folders and some settings over to Windows 7, but they had no choice other than reinstall all of their applications. In my case, I had a huge number of applications that had accumulated over the years that I was running XP. So, I nursed my old system along until last winter, until it was just too slow to be useful any more, and then transferred what I could to a newer PC running Windows 8 and reinstalled my apps.

Then came Windows 8.1, specifically the Windows 8.1 Preview, which I installed when it was released. I hoped that Windows 8.1 would solve some of the worst problems with Windows 8, starting (no pun intended) with the Windows Start menu. Although the Preview never fixed the Start menu issue, it did resolve some of Windows 8's other aggravating design flaws. However, when it came time to upgrade to the final release of Windows 8.1, the Preview turned out to be much more trouble than it was worth.

Here's a step-by-step summary of the update:

1. Bought a new Toshiba Satellite laptop with a current-generation Core i7 processor and 16GB of RAM to replace my old Samsung first-generation i3 laptop.
2. Installed the Windows 8.1 Preview on the new machine to make it simple to transfer everything from my old machine to the new one.
3. Learned that Windows 8.1 Preview cannot be updated to the final release version of 8.1 without having to reinstall all the applications, so I put everything on hold for almost a month until the final version of Windows 8.1 was released.
4. Last weekend, installed the final Windows 8.1 on my new machine from the Windows App Store.
5. Found that I couldn't install the final version on that machine without a serial number, and it would cost more than $100 to get one, so I figured out how to roll the new machine back to Windows 8. That also reinstalled all the crapware I had removed from the machine, but at least it allowed me to install Windows 8.1 for free.
6. Purchased a to-remain-nameless PC transfer program and installed the program on both machines. The software required that I uninstall all Norton software from both PCs, so I got all of the Norton software off my old machine, and all the preinstalled Norton crapware off the new one except for one anti-theft program that failed to uninstall.
7. Ran the PC transfer program, which failed instantly with a cryptic error indicating that the registry on my old machine was corrupt.
8. Sent diagnostic file from old machine to vendor of PC transfer software, and received reply saying that I had to send diagnostic files from BOTH machines, even though there was no way to do so from the old machine using the vendor's program, and oh by the way, I still had some Norton software on the new machine, and I had to remove ALL OF IT.
9. Manually removed every remaining Norton-related file I could find on the new machine and tried the transfer again. It again failed instantly with the same cryptic error.
10. Checked the registry on the old machine and downloaded software to send a diagnostic log from that PC to the transfer software vendor. Sent diagnostic logs from both machines. Received a reply the next day from the same tech support engineer that was word-for-word identical to the previous one, except that he highlighted the note that I HAD TO UNINSTALL ALL THE NORTON SOFTWARE.
11. Reinstalled Norton on the old machine so that I had some protection on the Internet, then purchased and downloaded a copy of Laplink PCMover Professional. Installed PCMover Professional on both machines. Threw the unnamed transfer software into the garbage.
12. PCMover's documentation was much less cryptic than that from the unnamed transfer software, but it required me to disable a fair amount of software (Internet security, backup, defragmenter, scheduled tasks, etc.). Disabled everything relevant I found on both systems and started the transfer.
13. The transfer took about six hours, but it worked. All but a handful of programs from my old machine transferred over, and PCMover was very clear about what it could and couldn't transfer.
14. Once I started working on the new machine, I found out that not everything that transferred worked as expected. For example, Google Chrome worked fine on the new machine, but it couldn't update. The solution was to remove all my Google applications, wait an hour, and then reinstall them, which fixed the problem.

Step 5 started last Friday evening; it took me to Monday night to get through Step 14. So, what are the lessons to be learned from this experience?

1. Never install a Windows Preview or Beta version on your primary PC. Even if you're a developer, load it on a PC that you can afford to wipe when the Gold Master or Final Release is available.
2. Use Laplink to do the system transfer when you're not doing an in-place upgrade, and avoid transfer software from companies you've never heard about before, even if they have good reviews. You don't know if those reviews were paid for.

If you're upgrading to Windows 8, read this

I bit the bullet last weekend and retired my six-year-old notebook PC running Windows XP. To replace it, I switched to an under-used Samsung notebook that I'd installed the consumer preview version of Windows 8 on earlier in the year. Unfortunately, the only option I had was to do a clean install of the released version of Windows 8 Pro. I reinstalled the software I need and transferred my data files from my XP system, but I found that writing emails and documents on the new computer was an exercise in frustration. My cursor bounced around the screen, randomly jumping back into the text, and highlighting blocks of text that were deleted as I typed.

I did some research online and learned that the problem I had wasn't uncommon, starting with Windows 7. When I did the clean install of Windows 8, my vendor-specific drivers were replaced with generic Windows drivers, and my trackpad driver was replaced with Microsoft's old PS/2 mouse driver. I downloaded and installed the latest drivers for my PC from Samsung's website, and all of the problems I had were fixed. So, after you upgrade to Windows 8, especially if you did a clean install, go to your PC vendor's website and install the current drivers. Signed Windows 7 drivers seem to work fine under Windows 8. Don't assume that Microsoft will install the latest, or even the correct, drivers for you.

Seton Hall University to equip students with Windows 8 tablets and Ultrabooks

TabTimes reports that Seton Hall University will give out Samsung Windows 8 tablets to all new science and honor degree juniors after the operating system ships this Fall; other new students will get Samsung Ultrabooks. The University will also give out Nokia Lumia smartphones to all incoming freshmen starting this fall.

iPad's market share increased in Q2, Android's share flat on higher shipments

TechCrunch reports on Strategy Analytics' Q2 tablet shipment report. iPad shipments increased to 17 million units in Q2 2012 from 9.3 million units in Q2 2011, while Android tablet shipments were 7.3 million units in Q2 2012 vs. 4.4 million in Q2 2011. Microsoft-based tablet shipments fell from 600,000 units in Q2 2011 to 300,000 in Q2 2012, as did "Others" (primarily the RIM PlayBook,) which fell from 700,000 in Q2 2011 to 300,000 in Q2 2012.

In terms of market share, Apple's share increased from 62% to 68.3%, while Android's share stayed flat at 29.3% despite the increase in units shipped. Microsoft's share dropped from 4% to 1.2%, and "Others" dropped from 4.7% to 1.2% Total tablet shipments grew 68.8% year-over-year.

A race to the top, or to the bottom?

In the last few weeks, announcements and rumors have been flying about technology leaders launching new mobile computing products:

• Microsoft jumped into the tablet business with its Surface tablets for Windows RT and Windows 8.
• Google announced its Nexus 7 tablet, built by Asus.
• Amazon is rumored to be announcing a replacement for the current Kindle Fire in late July or early August.
• That was followed a couple of days ago by rumors from reliable sources that Apple plans to announce  a 7" to 8" iPad in September or October.
• Yesterday, Bloomberg broke a story that Amazon is working on a smartphone to be built by Foxconn, and is trying to acquire patents in order to protect itself from the widespread litigation between smartphone and software vendors.

Assuming that all the rumors are true, it means that Amazon, Apple, Google and Microsoft will be competing directly against each other with tablets and (with the exception of Microsoft) smartphones. The margins on mobile devices are very low--only Apple has figured out how to make solid margins on its hardware. Samsung is making money because it manufactures so much of its own products, and it sells so many of them. No one other than Apple and Samsung makes money on smartphones. As for tablets, Apple makes excellent margins, but according to the latest hardware cost breakdowns, Google is just about breaking even on the Nexus 7, while Amazon is making a small gross margin on its Kindle Fire.

If Apple jumps into the 7" tablet market, will it compete with Amazon and Google on price, or will it add features that they don't have, such as 4G support, in order to command a higher price? How does Amazon expect to differentiate its smartphones from the pack of Android models? When Microsoft's Surface tablets make it to market, will they be competitive?

With everyone cloning everyone else and making incremental improvements, are we nearing the point where there are no clear lines of differentiation for anyone? If:

• Everyone has access to the same components, 
• Everyone's trying take advantage of whatever gaps have been left by their competitors until they get filled, and
• Android Jelly Bean's user experience is finally competitive with iOS and Windows RT,

Doesn't that mean that competition will inevitably come down to the number of apps available, customers' investments in libraries of apps (which affects switching costs) and price? That's a market that looks very much like personal computers.

HP to focus on Intel-based Windows 8 tablets, skip ARM-based Windows RT tablets

Bloomberg reports that HP has decided to focus on building Intel-based Windows 8 tablets and not release an ARM-based tablet running Windows RT. HP was one of the vendors given early access to the Windows RT platform by Microsoft. As fewer manufacturers commit to Windows RT, more pressure is being put on Microsoft's Surface tablet to be successful. If Microsoft ends up being the only company with a first-tier Windows RT tablet, it will make any chance of successfully competing against Apple almost impossible.

DigiTimes reports Microsoft is outsourcing Surface tablets to Pegatron; expected prices above US$599

DigiTimes reports that its sources say that Pegatron, which is one of Apple's iPad manufacturers, has won the contract to manufacture Microsoft's Surface tablets. These sources also estimate that the price of the Surface for Windows RT, which will compete directly with the iPad, will be priced above $599, while the Intel-based Surface for Windows 8 Pro will be priced above $799. In both cases, the high prices may outweigh the appeal of the tablets' industrial design. In addition, Microsoft has said that the tablets will only be sold in its retail stores (of which there are 21, plus four more under construction,) and online. Without other retail outlets, it's highly unlikely that Microsoft can get the volume it needs to drop the tablets' prices enough to be competitive with Apple and other manufacturers.

Microsoft surfaces Surface tablets, but why now?

Yesterday, Microsoft announced two new tablets in Los Angeles: The Surface for Windows RT, which uses an Nvidia Tegra ARM-based processor and runs Windows RT, and the Surface for Windows 8 Pro, which uses an Intel Ivy Bridge Core i5 processor. Everyone's been focusing on the "what" of the new tablets, but I want to focus on the "why." Microsoft's event yesterday seemed to be thrown together at the last minute--the company only gave a few days of lead time to the press, the site it chose for the announcement was a little "no-name" studio in Los Angeles, and the press that covered the event couldn't even get in until more than a half-hour after the scheduled start time.

Once they got in, Microsoft showed them two tablets, both largely clones of the iPad with a bit bigger screen (but even the more sophisticated Intel-based model's display's resolution is far below that of Apple's Retina Display.) The most innovative features were the thin covers with integrated keyboards and trackpads. More importantly, Microsoft announced neither prices nor availability dates for the two tablets. All we know is that the Windows RT version will ship around the same time as Windows 8, and that the Windows 8 Pro version will ship three months later.

Why throw together a last-minute announcement of tablets that Microsoft has obviously put a lot of work into, and not even bother announcing prices and availability? Here's my totally uninformed speculation: Last night's event was a "Hail Mary" pass aimed not at tablet buyers but at app developers. Microsoft knows that it has to have a solid library of apps if it wants to have any chance of competing with Apple. Interest in the Windows RT platform by both tablet builders and app developers has been lukewarm at best: Very few Windows RT tablets were shown publicly before last night, and developers aren't going to write apps for tablets that don't exist. In addition, Microsoft knows, through take-up of its development tools, how many developers are working on Windows RT apps, and the number has probably been very disappointing. The company saw what happened to HP's TouchPad and RIM's PlayBook, and it didn't want the same thing to happen to its tablet business.

Microsoft needed to show app developers that Windows RT is a viable platform, and the best way to do that was to show it running on an attractive tablet. That's what Microsoft did last night. In addition, Microsoft's plan to release the Windows RT tablet at the same time as Windows 8, while delaying the Windows 8 Pro version of the tablet for three months, tells developers Microsoft's priorities: Write tablet apps for Windows RT first, because we can always run existing Windows PC software on Windows 8.

In short, Microsoft is very worried that it's not going to have enough Windows RT apps to compete with Apple. That's why it announced its Surface tablets well before it wanted to, in order to encourage app developers to get on board.

The tablet makers' dilemma

Last week, DigiTimes wrote that "Android 4.0-based tablet PCs will be the dominant products in the segment in the third quarter of 2012 as most brand vendors will not rush to launch Windows 8-based models until the fourth quarter." The reasons for holding off on introducing Windows 8 tablets are fairly clear:

• Microsoft had been expected to release Windows 8 in Q2, but it's now looking like it won't be released until Q3 or possibly even Q4. 
• It will take time for vendors to test Windows 8 and related products for their tablets. 
• Vendors are suspicious of Microsoft's intentions with the Barnes & Noble deal--is B&N going to become Microsoft's "preferred" tablet vendor in the same way that Nokia is favored for Windows Phone 7 smartphones?
• Vendors see no reason to rush Windows 8 tablets to market. 

But, will Android 4.0 products be "dominant" in Q3? Not likely. According to Google's Android Developers site, all versions of Android 4 constitute less than 5% of the active Android installed base. That number overstates Android 4's true share because Google bases its installed base numbers on usage of the Google Play website, which only "authorized" devices can access. If the Kindle Fire and Barnes & Noble Nook Tablet and Nook Color were included, Android 2.3.3 to 2.3.7 would have even more market share than the 63.9% reported by Google.

That leaves tablet vendors other than Apple, Amazon and Barnes & Noble in a quandry. Should they continue to invest in Android tablets when, to date, Google hasn't been able to provide a version of Android for tablets that's compelling to consumers? How much hope should they place in Google getting it right with Android 5? Should they follow Amazon's model and fork a stable version of Android with their own user interface, perhaps even going back to 2.3.X?

Is Windows 8 the solution? Possibly--the Windows Phone user experience is superior to that of Android, and both Windows Phone and Windows 8 uses Microsoft's Metro design language  However, Windows 8 is the "floor wax/dessert topping" of operating systems. It's designed to be used for both touch-oriented tablets and keyboard-oriented personal computers, and as a result, it has compromises that iOS and Android don't have. The "tablet" portion of Windows 8 looks like its usability will be competitive with iOS, but there are likely to be very few apps available for it at launch, compared to more than 600,000 for iOS. The "personal computer" portion of Windows 8 looks and works almost identically to Windows 7, which means that it has nothing to do with the tablet apps.

One the one hand, computer companies can't afford to concede the tablet market to Apple. On the other, it's clear that Android, at least the stock version offered by Google, isn't competitive with iOS. Windows 8 is likely to be better, but as a tablet operating system, it's starting where Android was more than a year ago. So what should tablet vendors do? The safest approach is to wait and see: Develop prototypes of Android 5 and Windows 8 tablets that can be taken into production quickly, but let competitors test the waters first. Don't commit the resources to production or distribution until one of the operating systems proves that can compete with iOS. Another option is to fork and skin Android in order to improve its user interface--riskier, but possibly  the only way to make Android a viable alternative to iOS.

Windows 8: Destination unknown

Earlier this week, Microsoft formally introduced Windows 8 to its developer community at its Build conference. The new operating system was well received--so much so that Microsoft CEO Steve Ballmer had to caution people who downloaded the developer preview version that it's very early software and should in no way be used for production environments. Windows 8 is, in a way, Microsoft's answer to both iOS and OSX from Apple: It will run on PCs, notebooks and netbooks, like Windows 7, but it will also run on tablets, like iOS. It'll run on Intel processors like Windows 7, but it will also run on ARM processors like iOS. It runs all the Win32 desktop-style applications that Windows 7 does, but it also has a touch-oriented user interface and supports touch-based applications like iOS.

Windows 8's "secret sauce" is a new touch-oriented user interface called Metro. Metro is based on the Windows Mobile 7 operating system's user interface for smartphones, which is in turn based on the now-defunct Zune user interface. It does the sliding, zooming and selecting kinds of things that you'd expect from iOS and Android, but it does it on top of Windows. That represents both an opportunity and a problem.

The opportunity is that one operating system, Windows 8, could service both the legacy PC market and the new tablet market. Most observers believe that Apple will eventually unify iOS and OSX; Microsoft has done it with Windows 8. It's much easier for developers to build and support applications for one operating system instead of two.

Windows 8 supports two different kinds of applications: Win32 and Metro-style. Win32 applications are the programs and user interface elements that we've seen on Windows since Windows 3.1. They've been refined and updated over the years, but a Windows XP user shouldn't have any difficulty using a Win32-style program on Windows 8. Win32 apps are designed for a keyboard and mouse, and will work poorly, or not at all, with a touch interface. Metro-style applications, on the other hand, are all full-screen, use a radically different user interface, and are optimized for touch. You can use a mouse and a keyboard with Metro-style apps, but you probably won't want to do so.

Windows 8 systems will open into the Metro user interface, using a touchscreen. If you want to run a Win32 app, you'll switch back to the old user interface, using a keyboard and mouse. You'll then go back and forth between user interfaces as you switch from program to program. You can configure Windows 8 to always use a Win32-style GUI, but then you'll lose access to the Metro-style apps. That's the problem with running both Win32 and Metro-style applications on the same device.

Apple knew that tablets were going to be used very differently from desktop computers, so it optimized iOS for touch on smartphones and tablets. iOS doesn't try to run OSX-style keyboard-and-mouse applications; it offers similar functionality, but the user interfaces are very different. Windows 8, on the other hand, wants to be equally good at desktop and tablet applications, even though the use cases are very different.

For existing Windows users who want to continue using Win32-style apps, from what Microsoft showed this week, there's not much value in upgrading to Windows 8,  other than cutting down on boot times. Customers who are primarily interested in Metro-style tablet apps will find that app developers are starting from scratch, and it may take years for Microsoft to reach parity with the selection in Apple's iOS App Store. Customers who want to use both types of apps, and switch back and forth, are likely to get frustrated very quickly with dealing with two dramatically different user interfaces.

My suspicion, sitting a year or more away from when Windows 8 formally launches, is that Metro-style apps will be used almost exclusively on tablets, and that desktops, notebooks and netbooks will almost exclusively be used for Win32-style apps, despite Microsoft's goal of getting touch-enabled displays on every kind of computing device. This approach will allow customers to avoid switching between user interfaces, but it won't do much for Microsoft's growth prospects. Windows 8 on conventional PCs and Windows 8 on tablets will represent two different markets, and on the tablet side, Microsoft won't be able to leverage its huge PC installed base. It'll be starting from zero. Given that Apple will probably be on the iPad 4 and Google will be on Android 5 by the time that Windows 8 ships, that's not a good place for Microsoft to be.