Don’t read too much into the PC sales drop just yet

If you’ve been paying any attention at all, you probably know that new PC sales are in the toilet–out of the five biggest vendors, the only one whose sales managed to hold steady in Q1 2013 was Lenovo, while the other four saw a sales decline. So now Slashdot linked to a ZDNet piece stating that Windows is over, and said it must be true because ZDNet always sides with Microsoft.

Let’s not read too much into that. The author of the piece is a longtime open-source advocate. The points he raises are completely valid, but if there’s one person who’s going to take Microsoft to task, it’s Steven Vaughan-Nichols.

Microsoft has a long road ahead, but there is precedent for salvaging the boondoggle known as Windows 8. And I don’t think Windows 8 is the only factor here.

Windows, of course, has a long and checkered history. The first versions of Windows from the 1980s were interesting, but not very usable. It wasn’t until 1990 when Windows 3.0 came out that Windows became usable, and even then, those of us who had Amigas laughed at it. It was a resource hog, the multitasking was choppy, and it wasn’t as stable as an Amiga–and Amigas weren’t exactly the bastion of stability. Windows 3.1 came out the next year and was better, though still far from perfect, but it was cheap enough and good enough that it set the world on fire.

Windows NT 3.1 came out in 1993, and it was good enough that pretty much anyone who saw it acknowledged that it had a future, but device driver support was lacking, so it was exceedingly difficult to get a really good experience with it. Windows NT 3.5 and 3.51 followed in ensuing years and brought better stability and better driver support, but it wasn’t until NT 4.0 came out in 1996 that it was good enough for general use. NT4 was a very good operating system for business, but it fell short as a consumer operating system. More on that in a while.

Yes, for a good while there were two operating systems called Windows. Consumer-oriented versions were based on MS-DOS, though as time went on Microsoft hid those DOS roots more and more. Business-oriented versions were based on OS/2, and though they had compatibility layers bolted on for running DOS and 16-bit Windows applications, they were willing to sacrifice 100% compatibility in order to get better stability.

So, about those consumer versions. Windows 4.0 experienced long delays and was finally released in 1995 as Windows 95. We make fun of it now, but in 1995, most people saw it as a quantum leap forward. At the time, it was as big of a deal as a major Apple release, complete with people standing in line so they could buy it at midnight. And although it had problems, it was much better than Windows 3.1 was. So were Amigas, but I was the only one who cared.

Windows 98 was an incremental improvement over Windows 95 but still managed to be a year late when it came out in 1998. A so-called second edition, Windows 98SE, came out in 1999 and was the best DOS-based version of Windows that Microsoft ever made. It had issues, but there are people who still cling to it today.

Meanwhile, in Windows NT land, Windows NT 5.0 experienced significant delays (see a pattern here?) and finally came out in 2000 with the unsurprising name of Windows 2000. It was better than people remember it being. Microsoft promised five-nines uptime with it, which was laughable, though with a little bit of load balancing it was possible to do. I administered a system built on Windows 2000 that really did deliver better than five-nines uptime, but it was hard enough to do that I’m still pretty proud of that.

Although Windows 2000 was really good, and it had plug and play that worked the way it was supposed to work, there were enough games that didn’t run on it that it didn’t catch on with consumers. So Microsoft tried to remedy that with Windows ME, which was essentially Windows 98SE with better plug and play, but Microsoft rushed it to market and it wasn’t very stable. It wasn’t the first bad version of Windows, but it began the tick-tock pattern of Microsoft alternating good and bad versions of Windows.

Windows XP came out in late 2001, and although it wasn’t much better than Windows 2000 for the things I wanted to do, it was very consumer friendly, and it made consumers forget about Windows ME. Subsequent service packs made it very good and very stable, and it retains a strong following even today, 12 years after it was first released.

Windows Vista came out in 2007. It experienced significant delays and when it finally came out, it was slow and buggy and the changes it foisted on users were a bit jarring. It was a train wreck.

Microsoft realized its mistakes with Vista and released Windows 7 two years later. People throw things at you if you say anything bad about Windows  7. I’ll get ready to duck the cinder blocks and say it isn’t perfect, but it fixed most of the problems that Vista had. It’s a good operating system.

Windows 8 is, of course, worse than Vista. Too many changes, and they’re proving harder to get used to and less productive than Vista’s changes.

But Will Ormeus wrote in Slate that there’s another problem: Older PCs work just fine. I’m typing this on a seven-year-old laptop PC running Windows 7. It’s a little pokey for a couple of things I would like to do on it, but 95% of the time, it’s just fine, and I really don’t see any reason to spend $400 to get something better. My main desktop PC is a couple of years old now, and I built it with the lowest-end components I was willing to trust (an AMD CPU, the cheapest available Asus motherboard, and 8 GB of either Kingston or Crucial memory and a Kingston SSD), and it’s fine. I expect to still be using it in eight years.

I really do think that if Microsoft were to make Windows 8.1 default to an Explorer-like shell on conventional PCs–as rumored–and offer some other new compelling feature (really good encryption is a start, but something else along with that), the public will forgive Microsoft, just like they’ve done in the past. If they lower the price, they’ll do even better.

The days of people buying new versions of Windows just because it’s a new version of Windows are over. Microsoft abused that for 20 years, and now they’re paying the price by not getting the benefit of the doubt.

There’s some speculation that Microsoft is looking at the Apple model, where Apple releases new, incremental upgrades to its operating system every couple of years that have a small number of technical improvements and a small number of user interface or feature improvements, and prices the updates modestly. User adoption of the new versions isn’t 100%, but it’s high enough that it makes Microsoft jealous.

Microsoft needs to reform, and it needs to do it quickly. IBM, Microsoft’s erstwhile business partner and ally, is the precedent Microsoft needs to avoid. IBM got greedy and proprietary in the 1980s, squandered the PC market that it legitimized, and reformed in the early 1990s but it was too late.

I don’t necessarily think it’s too late for Microsoft to reform, but if Windows 8.1 continues to force the issue with the Metro user interface, I don’t think they get a second chance, and even if Windows 8.1 becomes a surprise hit, I don’t think Microsoft can afford to flub the successor.

2 thoughts on “Don’t read too much into the PC sales drop just yet

  • April 17, 2013 at 9:17 am
    Permalink

    I agree that Microsoft will eventually recover from its current self-inflicted wounds. However, I don’t think they’re going to ever be as dominant a force in computing as they were in the past — they’ve lost the opportunity to be an influential force in mobile computing, and as computing becomes more diversified, their role is going to be reduced to that of a niche player in a much bigger field of OS’s and platform types. This will not be a bad thing for anyone except Microsoft’s stockholders.

    Where PC vendors are concerned, I think the drop in sales is not going to reverse — not because mobile devices are the future (everyone in the developed world is still going to want a desktop or laptop for long form typing and such), but because, as you point out, old computers are good enough for most ordinary web surfing/office productivity tasks. PC vendors have become dependent on businesses buying new desktop machines every 3 years, but those businesses are starting to realize that they don’t need to do that anymore except for that tiny fraction of their workforce that actually needs the most powerful hardware available.

    Moore’s law has finally gotten ahead of the software for most basic computing tasks, and PC manufacturers are going to be suffering for a long time now as people and companies start keeping their machines for longer and longer periods of time. Again, I don’t see this as a bad thing for anyone except the shareholders.

    • April 17, 2013 at 3:23 pm
      Permalink

      Agreed on all counts. Microsoft will need to find a little bit of humility, and find something it can sell besides Windows and Office if they want to get out of the doldrums, because forcing new user interfaces onto Windows and Office isn’t making it 1999 again.

      And although PCs aren’t quite as easy to use as a dishwasher or microwave, once SSDs become standard, PCs will last as long as a dishwasher or microwave or other household appliance. I can come up with four companies that manage to remain profitable selling appliances that last 10+ years (let’s see: General Electric, Whirlpool, Electrolux, and Samsung), so it’s possible, but it’s going to be an adjustment.

      Tablets and cell phones are where the upgrade cycle will be, and it may take the rest of the decade for that market segment to stabilize, though it will.

Comments are closed.

%d bloggers like this:
WordPress Appliance - Powered by TurnKey Linux