June 26th, 2014

Lessons from the End of XP

Back in April, Microsoft ended support for Windows XP. Sort of. They've released at least one additional fix. They'll continue to support embedded systems for a while. They've worked out some sort of deal with people like the IRS. There were a lot of articles about how to migrate off of XP and whether you really needed to do it before the date support officially ended and really, how much were you risking if you didn't and how should you mitigate it.

I didn't pay much attention. I hate Windows. I've made an entire, highly profitable career out of hating Windows (and I come from Seattle, have a Computer Science degree and worked as a programmer for a little under 10 years, so, you know, at one point there was some effort involved) (and remember that I weaponize understatement. I _really_ _hate_ the Squish and its products). I don't really pay attention to updates or whether they are good or bad or how they handle the end of support or whatever. Nevertheless, events have conspired to bring this Fact of XP Support Ending to my attention repeatedly. My husband points out machines still running XP. My genealogy chapter has members still using XP. Etc.

First of all, there are a fuck ton of people still using XP. Today. Human people and corporation people and institutions as well. There are more using Windows 7, but oddly, about as many people use Mac OSX as use Windows 8, if some of the counter websites are to be believed (not sure I believe them, either). Why? Why haven't they upgraded?

First off, XP was the first NT based unified OS out of Microsoft, so its market potential was enormous (consumer/client and business/server, depending on the terminology you prefer). For a variety of reasons, Microsoft did not release a new OS for a really long time, so the market was unified, huge and more or less unchanging for a really long time. Then, when Vista did come out, people really loathed it. It required top end hardware, presented usability problems and, oh, financial crash! Boom. People really did switch to Windows 7 in droves (improving economy, not a piece of shit), but let's think about a couple of human factors.

(1) XP was viable early because so many people had recently upgraded hardware in the late 1990s, to avoid Y2K issues, and their boxes were still powerful enough to upgrade.

(2) If they were the kind of people who might have upgraded around 2001 (a 5 year old box, say, or a 4 year old one), they were dealing with a not great economic climate, maybe concerned about their job. But even if they did upgrade then, or, say, around 2004 or 2005 when things were going great, their Microsoft choice was still XP.

(4) While many people were recovered enough by 2009 (and desperate enough) to upgrade to Windows 7, the world still sucked really hard, and anyone who needed to borrow money (whether consumer credit cards or home equity or whatever) to upgrade a computer ... couldn't.

Those are the "I can't afford it right now" reasons for why there are still a lot off people running XP. What else stops people from upgrading? The sheer difficulty.

The _normal_ process of upgrading to a new OS/new box involves reinstalling all of your software. If you're still running software from the early 2000s or even the late 1990s, I could see that the prospect of trying to reinstall this stuff stopping people cold. I suspect this accounts for some of my genealogical buddies. I'm a little hazy on the details of transferring data from one Windows box to another, but it doesn't look easy when I read about how to do it in various helpful articles. I am reminded again why I'm a long-time lover of Apple products. They have relentlessly reduced the pain of upgrading, and then I've switched a lot of stuff to the cloud (and my husband set up time machine). At this point, I can joke about taking a hammer to my computer, going out to the mall, buying another one, and being all set up and ready to go in under 2 hours, mostly devoted to the drive to and from the mall. I bought a Chromebook to see whether that could be used as a substitute, and it mostly can, so I now have an upstairs and downstairs laptop (seriously, that stupid) or a desktop replacement laptop and a travel laptop or however you want to characterize me being all obsessed with understanding other people's platforms.

To sum up: Y2K forced a lot of obstreperous people to buy new hardware and a new OS. That hardware supported the upgrade path to XP. They have been reluctant to upgrade since PLUS two recessions and a debt-driven, housing focused, low employment "recovery" in between. Also, huge delay between XP and Vista and Vista sucked and unemployment was still high until quite recently (arguably, still crazy high _now_).

And we wonder why there are still so many people on XP.

A lot of households have an ancient desktop and have resorted to mobile devices (iPads and Android tablets and Kindle Fire) to stay up to date and supply the kiddies. And a lot of schools are running very old systems that have been babied along, to the point where Neverware is an Exciting Startup (Neverware is like the opposite of VMWare. They both involve a server that runs stuff, and a lot of thin clients. VMWare is aimed at businesses that want to run ancient software on new hardware. Neverware is aimed at schools that need to run new software but only have old hardware, some of it broken. The fact that this has to be described in detail to excellent engineers like my husband tells you a lot about how engineers have trouble with business models and are comprehensively oblivious to the world in general).

I think the lesson to be learned from XP is: this is what income inequality and an output gap looks like. If we had less income inequality, then schools and individuals and small businesses would share the same ability to keep up to date on hardware and OSes and software as the richest individuals, schools and businesses. If we didn't have an output gap, people wouldn't have had to delay upgrading their computers because of a massive drop in property taxes, high unemployment and so forth which starves state and local governments and thus schools of money for hardware and software that is, nominally, never mind real dollars, totally cheap. Income inequality means it is hard for the wealthier of us to recognize what it means to not be able to afford a few hundred dollars to buy an all-in-one desktop to replace that dinosaur (possibly with a CRT) running XP that was bought about 10 years ago. Optimistically -- realistically, maybe around 1996.

I don't know if we can be better than this. But it would be nice if we could at least stop blaming people for not upgrading from XP already. They have damn good reasons. A lot of the smart people who upgraded from XP jumped ship to OSX or Linux or whatever. If everyone had been smart and upgraded already, Microsoft might well be out of business.

A Few More Remarks on What I Learned Writing the Previous Post About XP

I've been meaning to write something along those lines for months (it is even on my to-do list, which blog stuff rarely makes it onto because I just sit down and write whatever the hell is on my mind), and the basic argument (people haven't upgraded for a long time because Y2K, it was good enough, two recessions) never budged. I did do a bunch of research to make sure that what I was asserting bore some relationship to reality (names of OSes, dates of release, dates of when individuals/schools/companies might have felt the impact of the crash compared to the release date, etc.), and I also did a bunch of research to understand what would be involved in upgrading from XP to Win7 or Win8. Along the way, I learned that Vista, apparently, really did comprehensively suck -- I hadn't known that, altho obvs I couldn't entirely miss the complaints as they rolled by back in the day.

But fundamentally, I'm trying to understand and explain the behavior of a group (people who have been running Windows boxes for years) that I don't actually have any meaningful contact with. One thing I learned as a result is that they don't understand people who _don't_ use Windows. At All. When people understand both groups, they often switch, and Microsoft isn't benefiting from the switching.

When Google first presented Docs to the world, I wanted to upload all my documents to it and then never have to save anything locally again. I really wanted a Cloud. But that didn't work, and I was so discouraged that I ignored DropBox (and a bunch of related services). Customer behavior is driven by what the customer knows, has experienced, has friends who have experienced. This isn't a new idea. It's kind of a duh thing. But working through the evidence to support my thesis really forced me to recognize just how poorly tech companies and tech-people at all levels understand customers. They aren't as bad as publishers (publishers had zero understanding of their end-user for a really long time, and that hurt them as they were disintermediated), but Microsoft is close. I'm not sure how this is going to end, but it looks like Microsoft is, like many other companies before (IBM, DEC) receding to the high/enterprise end of the market.

Ironic that Jill Lepore's piece about disruption came out in this context.

ETA: Oh, and originally, the thrust of the piece was going to be, Why Are Computers in K-12 So Freaking Old? But when I realized the scope of the problem was so much larger, I re-oriented it.