Monopoly and technology

Microsoft has also promised backward compatibility, and for the millions of people who still use older versions of Windows and Office, and who do not need or want to spend their money on a new computer, that promise is a good thing.

But as Henry points out this will probably cause more and more people who are buying a new computer to switch to Apple. This trend will continue until it’s no longer too painful for Microsoft to abandon the old technology.

In fact Microsoft already abandoned DOS with Windows XP. DOS programs are supported but none of the DOS code is in XP. No reason they can’t do the same with later stuff, just add an emulator to the bundled software.

As digitaldigit says, Microsoft did indeed abandon DOS, but after several years in which the successor and DOS had been running hand in hand. And they’ve done the same with the Windows 3.1 legacy, abandoning the pre-NT software several years after NT was introduced. In fact, of course, Windows XP and Vista are built on a foundation that is about 15 years old (NT) but that is not so old really.

I’m no big MS fan, but I think they’ve done this right. They’ve paid attention to compatibility while continuing to move forward. It’s been this balance between pragmatic support for existing applications and continuing to move forward that has probably been the biggest source of their success. And while OS X may be elegant, let’s face it, Apple had got to the stage where they had to do something, and where they didn’t have that many existing customers (in comparison) to annoy.

I’ve got another explanation for you. Backwards compatability isn’t about consumers – it’s about businesses.

There are a ton of businesses, small and large, out there that are run on custom applications that run on Windows. Think databases, customer management, document management, patient records (for HMOs and hospitals), and so on. I’m sure (although I don’t have any data on this) that many billions of dollars were spent on developing these applications. If Microsoft abandons backward compatability, these companies will just not upgrade. MS doesn’t make money on the new operating system and the companies don’t get the advantage of the new OS. Sounds like a lose-lose situation to me. Apple didn’t have to worry about it because companies weren’t running mission-critical applications on Mac at the time. (i.e. some things that are OK when you have 5% of the market aren’t so great when you have 90%)

Anyway, the “start over with a more elegant design” philosophy of programming is debatable. Not everyone thinks it’s a good idea (see, e.g. Joel on Software).

So they do cut their old customers loose, is all I’m saying.

They cut customers loose in the sense of no longer providing technical support and upgrades for very old products. That’s very different from changing their operating system in a way that would break all old applications.

How is showing respect for your old customers/developers by promising that they won’t have to buy new software with each new OS a bad thing?
Because it means inferior software in the long run- more complex, more buggy, slower, more insecure, etc. If they weren’t afraid of losing their monopoly, they could probably cut millions of lines of code from their OS and introduce something lighter, faster, more stable, and more secure. And nothing would stop them from continuing to support those old customers just fine on the operating system those customers already have.
That said, the fact that they keep compatibility, even at such demonstrably high cost, is actually a sign that there is at least some competition in the marketplace. If MS were a true/pure monopoly, they could perform this short-term customer abuse with no worries- the customers would have nothing to defect to. As it is, they are afraid customers will defect to Apple or Linux*, so they must keep compatibility.

* a much bigger threat in abb’s corporate space because of the lower hardware costs and increased competition amongst suppliers

Yes, to all who said this is a business customer issue: exactly.

I’ve calculated the costs of switching to Mac, and it would cost me two or three thousand over-and-above the hardware costs: a fair amount, but amortized over a couple of years for the switch, it’s not bad. As a long-time Mac user (’85-’97) who only switched because Mac sucked so bad compared to a PC/Windows NT Workstation, I’m certainly hankering to switch.

But the business ecosystem has not billions, but tens of billions (or even: hundreds–it’s possible, given the twenty years or so of MS dominance in the marketplace, and there are certainly apps that old that are still in use), invested in proprietary apps built on MS (this is, of course, the true genius of MS vs. Mac: getting tools like VB/Access into hands of amateur/low-level programmers in business environment & supporting them like crazy–MS are the masters of the PRTM gospel & reaped the just rewards). So, there’s not really a choice for Microsoft.

Still, there’s a cost–and it’s increasingly shitty OS’s & the inability to create rapid churn for key apps like OS/Office (which account for massive percentage of MS net profits). I finally left Win2K Pro this spring when I built a new PC for the first time in 5 years and can definitely say: XP Pro is garbage by comparison (much less stable).

@anon at 14:
Not necessarily true. It’s called driver hell for a reason, especially if the driver does weird things when talking to the OS or the hardware (as is won’t to happen when you code for just getting something done, cause hey, who’s gonne be using this in 20 years…)

Jim in post #7 is right. Its about business. Banks, insurance, etc, etc. Moving to a new OS is a huge endeavour for any large company, and one key consideration is will the legacy software work on the new OS. If the legacy software didn’t work, then they wouldn’t move (or would delay).

MS’s biggest competitor is not UNIX, or MacOS – its its own legacy of old Operating systems already paid for and people’s computers all over the world.

And for #13 – Are you sure? My experience (as a pro), is that XP is more stable than Win2000. The problems are largely hardware, rather than S/W – and that’s a problem for any OS.

The problems are largely hardware, rather than S/W – and that’s a problem for any OS.

But much less of a problem for Apple, as they make all the hardware, and go to some lengths to standise it. That doesn’t explain every difference between Mac OS and Windows, but a lot of people seem to agree that WinXP is very stable on a good bit of hardware: no easy feat for microsoft to pull off, given the vast variety of hardware they have to support.

While in contrast, Microsoft customers can look forward to a piece of bloatware that will be extraordinarily obese even by its previous standards.

I really don’t think that this is very fair– does, for example, Windows require much faster and more expensive hardware to run than linux or Mac OSX? No. If you want a really fast OS, go and run BeOS: it’s not supported anymore, but it’s beautifully fast (mainly due to design decisions taken from the ground up).

Yes, it’s all about legacy business apps. Consumers are very easily converted to a new OS or software package, by comparison. But if you have an office or agency system that holds a decade’s worth of data, and whose workings are familiar to staff, then migration is a nightmare.

cian –

Well, I kept a Win2K server running for more than a year once, so: I guess it depends on your apps & configuration (as well as, as you say, hardware).

In the meantime, seems like there’s (CT surprise!) near unanimous consent: Macs are cool, but legacy business apps are going to hold people to their current platform.

Of course, MS still has all those AS400/RPG users to convert while waiting for Vista–over their customers’ dead bodies (as I learned when working with legacy customers at ERP shop in 90s).

The really, really frightening thing for Microsoft is the trend where the business sector go for commoditized solutions and slow upgrade paths (a twofer with low margins and stagnation), and the consumer sector goes toward high differentiation/ tight integration. Today, growth is in the consumer space. As pointed out in # 22, consumers more easily converts. Microsoft’s only genuine consumer space presence is the x-box. Methinks the iPod/iTunes/OS X combo is a stronger proposal than x-box/vista. Although the business sector lock-in has served Microsoft well in the consumer space til now, it is in the future better to have a media lock-in than a business sector lock-in. Microsoft is of course going to have a commanding market share for the foreseeable future. However, that is true for GM as well.

abb1 —

Well, in this instance, it was Win2K Pro (which I was using as a VSS server). In any case–I’m certainly willing to say that my experience w/XP Pro may be uniquely bad… I certainly don’t have anything at stake in the matter.

Dan K–

It’s not clear to me what this means: “…the iPod/iTunes/OS X combo is a stronger proposal than x-box/vista.” Is this from a top-line corporate perspective or a consumer choice perspective? In the Apple instance, it’s all they have (damned good, at that); but for MS, there’s no stopping a Windows user from buying an iPod (I have three) and nothing prevents a Mac user from buying an Xbox (is there?)–and certainly the MS revenue stream is much more balanced than X-box/Windows?

For all that, just did a quick check of comparative revenues for last year & wow, is Apple catching up. Apple had $1.34B net on $13.9B rev, while MS had $12.25B on $39.8B–who’d have imagined Apple would ever approach even 1/3 MS revenue again? Of course, the MS margins are, as usual, nasty. Thirty-percent net margins… it’s enough to make you hate them just for that.

I’ll add–you’re spot on with this, Dan: “The really, really frightening thing for Microsoft is the trend where the business sector go for commoditized solutions and slow upgrade paths (a twofer with low margins and stagnation)…”

MS makes 90% of its profits from OS and Office. If sales of these two take a significant hit, they’re hosed (X-box still loses them hundreds of millions a year, they don’t make much on big server stuff, and Great Plains acquisition is all but a bust).

joel – a lot of the WinXP drivers that were written for old hardware were pretty bad, so its quite likely that was the problem. Of course anybody running any serious server app on Windows is a fool…:) (run it on AS/400. Like a rock those things).

As for revenue. Well Apple is mostly a hardware company, so their costs per unit will be far higher that Microsofts. And the MS margins sort of reflect the fact that in software your main costs are fixed, no matter how many licenses you sell.

The future may be the consumer – but I doubt that’s windows, or MacOS. The trouble with computers is the things are too complex, and too prone to breaking.

“The trouble with computers is the things are too complex, and too prone to breaking.”

Yeah, when I saw the NYT piece’s details about 50M lines of code and the delay I immediately thought three things:

1) I hope the project team remembers reading Brooks.
2) Didn’t Bill Joy once say that you couldn’t write a program larger than 10K lines without bugs (much to the chagrin, no doubt, of Watts Humphrey)? Is it even possible to write a commercial product with 50M lines of code that isn’t a complete wreck? Isn’t this why fighter jets and space shuttles are so ridiculously expensive?
3) Douglas Hofstadter’s hilarious Hofstadter’s Law: “It always takes longer than you think, even accounting for Hofstadter’s Law.”

agm@#15:

That’s much of the point of using a virtual machine: to avoid driver, library, and other such compatibility errors. Software under VMWare uses a virtual VMWare graphics card, network card, etc. instead of accessing the physical hardware.