Responding to Readers: Accounting for Politics in Technology
Apr 29, 2004 5:34 AM PT
In my recent commentary on the role dishonest and incompetent IT textbooks play in education [Paul Murphy, "Eight IT Textbooks, 4,031 Pages, 17 Mentions of Linux," LinuxInsider, April 2, 2004], I used the acronym PHB (pointy-haired boss) to refer to people who just don't get it with respect to systems. Unfortunately, that also demonstrated that I don't get the global nature of the LinuxInsider readership, because I didn't stop to think that not everybody gets the Dilbert cartoons and so a lot of people wouldn't grok the reference.
What makes this doubly embarrassing is that I have a secret vice that, if there were any justice at all, should have made me more sensitive to this kind of idiotism. In my own defense, however, I'd like to point out that making multilingual puns is at least hard, particularly if, like me, you don't actually know the other languages involved.
Inexplicably, people generally don't notice these attempts at humor. For example, Canada is officially bilingual, but I've never gotten so much as a raised eyebrow from pointing out to client suits that cheaper restaurants serve crusty rolls because they see slicing French bread as a real pain. Of course, most of them don't get references to the Liberal Party's (think "democrats") Quebec fund-raising arm as the eminence grease of Canadian politics either.
Mostly, however, the opportunity to make the perfect pre-thought-through pun never comes up. No McLuhan-reading Russian speaker, for example, has yet given me a chance to explain my choice of a blindingly laser-blue sportswagon as intended to let me shout "be seenya" as I wave out the sunroof and roar off. On the other hand, my little boy's grandmother thinks my Unix Guide to Defenestration is some kind of religious pornography, and I know I'll finally have a captive audience for one of these puns when explaining this phenomenon to him as omamatopoeia.
All in all, however, it's a lonely and socially unrewarding vice that doesn't even protect against prostate cancer -- but should have left me more sensitive to the misuse of cultural referents like pointy-haired bosses -- for which I apologize.
I'm from Mars, Thanks. You?
I try to read things like Slashdot commentaries when they apply to either my own columns or other work I find interesting. There are rewards to this because there's almost always a genuinely insightful gem or two hidden among the illiterate fulminations of the furious. One thing that almost always appears, however, is an ascription of bias -- things like someone telling the world not to bother reading LinuxInsider because it's run by and for Microsoft haters.
Obviously, I can only speak for myself on this -- although I happen to know that the senior editor secretly uses Windows at home -- but I do not hate Windows. I often sound that way because I'm comparing Unix to Windows, but the truth is that I simply don't get Windows at all.
Way back when the PC/AT came out, I was running BSD on a Vax and got to evaluate both the AT and MacXL as possible replacements for some IBM DeskWriters. At the time, this seemed such a no-brainer that I assumed the PC would sink without a trace. It didn't; but I still don't get it. If you want something that does desktop jobs efficiently and reliably for a minimal number of dollars, the Mac is still years ahead of the PC. If you want something that does large-scale, multiuser work with a minimum of fuss and dollars, the only decision you should have to make is what hardware to run Unix on, not how many firewalls you need to install with Microsoft's servers.
Pretend for a moment that you're from Mars and know nothing about the PC wars. When you review the options, what you'll find is that there aren't any desktop jobs Windows can do that the Mac isn't better at, and that there aren't any server jobs Windows can do that Unix isn't better at. Try it; the experience will make you wonder how this is even an issue for grown-ups.
Windows, the Behavioral Mystery
Windows isn't more reliable, it isn't cheaper, it isn't better supported, and it was only for a few years in the late 1990s that people could reasonably claim there was more software for it. In 1985, the Mac -- for less money than the PC -- came with a suite of GUI applications that worked out of the box, while the PC/AT had a BASIC interpreter on top of PC-DOS.
Today, a Linux or Solaris machine can run almost all Windows software as well as the Unix stuff, while what software Windows can run depends on which release and service-pack level is installed. To the man from Mars, Windows isn't a decision. It's a behavioral mystery, and one I've never plumbed.
Rationally, I know the answer and understand the sociology of acceptance, account control, and the organizational stupification that dragged the hearts and minds of captive users along the PC path. Emotionally, however, my bottom line on the PC is that I just don't get it, particularly now that client-server is so obviously turning into its mother -- that is, evolving via server centralization, desktop lockdowns and the increasing distancing between systems people and business users -- to replicate the priest-ridden glass room its adherents vilify as a 1970s mainframe architecture.
But that's not the same as hating either Windows or Microsoft. I don't, although, come to think of it, there is rather a lot to despise about Microsoft's business practices and the fools who support it, isn't there? Umm...
A Solaris-Powered Helicopter
I'd buy a Solaris-powered helicopter if I could [Kirk L. Kroeker, "DOJ Bans Linux from US in Wake of iWidget Brouhaha," TechNewsWorld, April 1, 2004], but the point is that a lot of readers seem to be outraged that I'm so clearly biased in Sun's favor. My bottom line on Sun is that I like stuff that works, and their stuff generally does -- even if Sun Press did turn down my latest book.
The more general response to a charge of bias is that I'm for what Unix can do and particularly interested in services to larger user communities. Take a long, hard look at computing science today, and I think you'll agree that most of the really leading-edge research for single systems is taking place in the BSD community, that the Linux people are wreaking miracles of application delivery support across an enormous range, and that both Solaris and Sparc are evolving rapidly toward the Plan 9 view of network integrated computing.
To me, it's the bigger stuff that's most exciting, and I make no secret of the resulting pro-Sun bias because they're the people making it happen.
The Power of Politics
I get e-mail from readers who object to my tendency to inject politics into some of my columns. Politics, they tell me, isn't technology and has no place in a technology-oriented journal like LinuxInsider.
In most cases, I think that's right. A writer explaining how to make a piece of software work, reviewing a technology product or reporting a piece of technology news has no business injecting politics into the issues -- but those aren't usually the things I write about. My focus is mostly on the management and business decisions around the use of technology, not the technology itself.
In many cases, that means political issues have direct consequences -- for example, the president's recent decision to have Dick Cheney read the riot act to Beijing is going to affect your technology job because it is part of a strategy to reduce the imbalance in the flow of technology and manufacturing jobs from the United States to China.
Don't kid yourself about this; the high-level management decisions that ultimately determine whether you keep your job or whether your favorite technology makes it in the market can be much more heavily dependent on politics than on technology.
Politics and Tech
That's usually easy to see at the operational level, where the personalities and preferences of individual players are known, but a kind of increased "nebulosity factor" manifests as you trace things upward in terms of both players and scale. As a result, it's generally not possible to know with any degree of certainty what the actual impact of national politics is on technology decisions; only that there is an impact.
Sometimes the impact is pretty clear. For example, I think you should be out there supporting any elected, or wannabe elected, representative who promises to support making it illegal to export personal data on Americans for processing outside the United States. That's fundamental to national economic security, important to national security in the military sense and valuable in terms of keeping your job.
On the other hand, much of the speculation on the impact politics has on technology sounds a lot like conspiracy theory. For example, I have a theory -- which I have no hope of confirming or disproving anytime soon -- that Novell's takeover of SuSE was motivated by IBM in a last-ditch effort to get a deal to have IBM Global Services support SuSE on every desktop and server owned by DaimlerCrysler past opponents in Detroit.
If so, what got in the way of what would have been a genuinely big deal for Linux was adroit manipulation of national economic agendas in Washington. So, did that happen? Maybe, maybe not. I not only don't know, I don't know how to find out. But I do know that taking discussion of technologies past reportage and how-tos means taking politics into account.
Paul Murphy, a LinuxInsider columnist, wrote and published The Unix Guide to Defenestration. Murphy is a 20-year veteran of the IT consulting industry, specializing in Unix and Unix-related management issues.