Both the original suggestion and this follow-up article stem from the fact that Microsoft has changed the way it treats users. The trend started about two years ago, coincident with top-management shuffles and the distraction of the antitrust suit. Since that time, Microsoft has become increasingly callous toward its ultimate customers--you and me.
For example, Windows Millennium Edition is perhaps the worst version of Windows in more than a decade. At most, it should have been a free update to Windows 98SE. Instead, Microsoft gussied it up and shoved it out the door as if it were an entirely new version of Windows. The only reason I can see for doing this is that Microsoft needed to release something new to keep the cash flow going. Lacking a real product, it foisted a half-baked Windows ME on the world. (See last year's 10 Ways To Make Windows ME Run Better.)
There are even broader problems, too, such as the corporate blind spots that have caused exactly the same kind of security flaw to show up repeatedly in an amazing range of Microsoft products. We described a recent instance of this problem in January:
This last year, we saw an astonishing stream of "unchecked buffer" security problems emerge in numerous Microsoft products. (A malicious hacker can use an unchecked buffer to insert dangerous commands into your software.) This systemic problem with unchecked buffers in Microsoft products reached its culmination late last year, starting in November, with the first reports of security problems in the "universal plug and play" system built into XP and ME; and into the Internet Connection Sharing service in Win98. Microsoft initially downplayed these problems (assigning them a threat rating of "low"), and only changed their tune after non-Microsoft researchers dug deeper and uncovered another unchecked buffer that could allow malicious hackers completely to take over a system at its deepest levels. The problem was so severe that the FBI actually issued a dire warning in December. After that, Microsoft belatedly acknowledged that the problem actually was "critical" and released a patch.
I think it says something very bad about Microsoft when third-party researchers and government agencies have to uncover and publicize glaring, extreme security problems in Microsoft products, especially when it's the same kind of problem that's cropped up again and again in other Microsoft products. Something's seriously amiss in Redmond.
This kind of security problem is bad enough as an isolated thing, but it's not isolated at all. Rather, it's part of a long-standing pattern with Microsoft, one that recurs again and again.
Problems like these cannot be ignored. Something really is amiss at Microsoft. So we all need to start thinking about Windows alternatives, if only in self-defense.
The trouble is, it's not a simple choice.
If I had to pick a single operating-system alternative to explore, it would be Linux. But, despite what the Linux partisans say, it's neither easy nor cheap to change from Windows to Linux. Yes, I've seen many articles and white papers that "prove" that Linux costs less to set up and administer than Windows, but almost all those studies ignore retraining costs.
Those costs are real. Take any Windows-based enterprise, switch everyone to Linux, and you can count on lost productivity as people learn the new software, work out file conversions, and so on. Linux itself may be cheap or even free to obtain, but the conversion will force you to take major hit in productivity.
I know--some of you are rolling your eyes. "A major hit in productivity? Langa, you're nuts!"
But if you're saying that, I'll wager that you have current or previous Linux experience. For people who already know Linux and who already know what they're doing with the operating system, I'll agree that the productivity loss can be smaller.
Once you know what you're doing, you probably can save a system's Windows-based user files, wipe the system, install and configure Linux, reinstall and configure all necessary drivers and networking, install and configure the productivity software, reload and convert the data files as needed, and resume work on a Windows-free system more or less where you left off, in as little as maybe three hours--call it half a day or so total time. (The time could be spread out over several days.) But even a half-day is a lot of lost productivity when you're talking about dozens or hundreds or thousands of seats.
And remember, that's the best-case scenario, where the people making the change know Linux and merely have to make the actual switch, with no steep learning curve.
In most real-life situations, the conversion will take a lot longer. Most people have no experience with anything other than Windows or with anything other than Microsoft's Office suite. For these people who've never seen other ways of doing things, even minor changes are difficult. For example, I recently helped some people upgrade from an ancient Windows 95-based system--the only operating system they'd ever known--to Windows 98, and I saw them struggle with even that small step.
For most people, change is hard; that's the real world. Unfortunately, many Linux partisans forget that. Their learning curve is way in the past and may even have been spread out over several years. They have no recollection of what it's like to be new to the operating system and to have to face the entire learning curve at once.
And sadly, some of the more rabid Linuxophiles add a needless social tension to the already challenging technological task of switching operating systems. These Linux "fanatics" (as opposed to the sane enthusiasts) assume that anyone with lesser Linux skills, or anyone who forms an opinion contrary to theirs, must be an idiot. Worse, they're often all too eager to express these negative feelings. This kind of negative interaction creates a nonmonetary human cost that may not show up in the return-on-investment estimates of Windows-to-Linux conversions, but it's there and very real.
Getting back to the technology: There also are issues with Linux itself. Although driver support is improving rapidly, it's still nowhere near as complete as that of Windows. Although you can get your peripherals to work in Linux, many times they'll be in a more or less generic state that lacks special features available on the same hardware when run under Windows.
Linux utility and application software is also evolving fast and even leads the pack in a few key areas (most notably in server software). But it still lags way behind Windows in most general productivity applications.
And then there's the huge, fundamental issue of which Linux to use. There are about 170 distributions available--many more if you count the various builds within each distribution. Deciding which one to deploy is no minor matter.
We'll come back to Linux choices in a moment, but for now, let's look at some other alternatives.
Let me say up front: The Apple Macintosh is a fine platform, technically. And the Mac OS is once again current and competitive, having emerged from a long period of near-stagnation lasting the better part of a decade.
Good hardware running a good operating system--what's the catch? It's this: Apple's business model continues to ensure that the Mac remains overpriced and undersupported.
You'll pay more--often significantly more--for Mac hardware than for the equivalent Intel or AMD-based hardware. And although you can find Mac software for just about every purpose, you'll have far fewer software choices and options than in the supposedly choice-less Wintel world. To add insult to injury, you'll often also pay at least a few dollars more for Mac-based software than for exactly the same software in a PC version.
All the training and conversion issues mentioned above (in the Linux section) also pertain to the Mac. Although you'll hear Mac partisans say how "intuitive" the Mac is, no human is born with any intuition about any computer operating system. It's all training; whatever operating system you learn first is the one that will seem "intuitive" to you, and all other operating systems will seem clumsier in comparison.
Here's a classic example. Apple once ran ads touting its one-button mouse as proof of how easy it is to use a Mac. But when proficient Windows users first sit at a Mac, you can watch as they temporarily grind to a halt the first time they try to "right click" on something--there's no right mouse button to click with. Of course, there's an equivalent function on the Mac, but it's something that must be learned--there's no intuition involved at all.
An operating-system switch from Windows to the Mac will involve major hassles, just as with any change of operating systems. But the Mac carries the additional burden of relatively high costs, and so (to me) it seems poorly suited as a general alternative to Windows.
Hundreds Of Choices
There are, of course, many other operating-system choices. For example, there's BSD. Like the many flavors of Linux out there, BSD is an evolutionary descendant of classic Unix, but one that remains closer to its roots than does Linux. BSD exists in several variants: FreeBSD, OpenBSD,
BSD/OS, and NetBSD.
In fact, with so many operating systems it would be hard to try them all. But collectively, if we pool our knowledge, surely we can hit all the major ones.
So, let me ask you: Which non-Windows operating systems have you tried? Which brand/distribution and versions gave you the best results? What worked and what didn't? How long did the operating-system conversion take? What did it cost? Have you migrated entirely away from Windows, or do you run your alternative operating system in a dual-boot setup? What did you just plain like, and what was disappointing?
5 Top Federal Initiatives For 2015As InformationWeek Government readers were busy firming up their fiscal year 2015 budgets, we asked them to rate more than 30 IT initiatives in terms of importance and current leadership focus. No surprise, among more than 30 options, security is No. 1. After that, things get less predictable.
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?