.comment: Why Windows Users Should Oppose the Settlement (and Other Notes That Defy Categorization)

By: Dennis E. Powell
Wednesday, January 9, 2002 10:02:18 AM EST
URL: http://www.linuxplanet.com/linuxplanet/opinions/3994/1/

The Effects of a Monopoly

Yes, I'm going this week to hammer again at the proposed settlement in U.S. v. Microsoft. There are a couple of reasons for this. The first is to remind you that the time is running out -- you must file comments you wish to make under the Tunney Act provisions of the case by January 26. The second is to point out something little noted: Even Windows users -- perhaps especially Windows users -- would benefit from a far stronger penalty than the puny one contemplated in the settlement as it stands.

Why address Windows users in a Linux column? Simple. From my email I learn that many readers use Linux at home, but work in Windows shops. Many of those are IT professionals who have sought unsuccessfully to insinuate Linux into at least part of the companies for which they work. I hope to arm those readers with the ordnance necessary to get their companies to take a stand, even in the form of an officially sanctioned comment or, better, in a reconsideration and comment by the CEO, COO, CTO, or whatever three-character character carries weight.

The argument I propose to employ embraces the entire political spectrum -- it's something that fits in nicely with anyone's philosophy of choice. In the vernacular, it's a "no brainer," except that "no brainers" are people who unquestioningly use whatever came on the machine, or who fall victim to the smooth talk of the Microsoft sales department. (If recently published internal Microsoft memorandums are legitimate, then my worst suspicions about the company are overly conservative -- Microsoft is truly reprehensible. I have several excellent books about reptiles here, but the Brian Valentine appears nowhere in any of them, apparently an oversight; then again, I like reptiles, but Valentine's words make my blood run cold. Send the National Geographic to sort it out.)

The argument is this: Competition breeds better products. And without competition, Microsoft products would become as bad as -- well, as bad as they have become.

Here's how it works. A company which has a monopoly has no incentive to improve; in fact, it has no reason even to maintain the quality of its products. Where else are you going to go? (As to Linux, consider the salesman's pitch, "Yeah, there's this other little thing that some guy in Sweden or somewhere did as part of a term paper, but you and I understand each other, don't we? We're both in business.")

The proposed settlement underlines the salesman's words.

Those of us who run Linux exclusively have had the amusement of watching, over the past year, as serious security attacks absolutely ripped machines and companies running Microsoft software. We saw Nimda batter itself against us, with no effect unless we were running web servers, in which case it filled up our logs. We passed around screenshots of the Microsoft update site, beset by Nimda because Microsoft did not itself apply its own patches. We were annoyed because the Web got slower.

Before that there was something called "ILoveYou" and later there were SirCam and its variants, which we deleted, saying a few words of profanity over the waste of bandwidth, and moved on. The New York Times didn't; its Web work was down while it all got sorted out.

We have heard about the ability -- feature? -- of Internet Explorer to allow pretty much anybody to run pretty much any executable they care to on machines running Microsoft software, but we've been immune -- "serves 'em right," we probably said. I sure as hell did. We now know that their plug'n'play extends beyond the local machine and welcomes anyone who wants to do harm.

The list goes on, and extends to things which haven't even happened yet.

Now, you might suppose that any company whose products were so breathtakingly vulnerable to attack might do something about it. And you might be surprised that they haven't.

You shouldn't be. With no competition, Microsoft has no incentive to ship a secure product. I mean -- "where else are you gonna go? That kid from Norway or wherever it is? Yeah, right."

If you are reading this column, you know that in one way Microsoft does have competition. So does Brian Valentine, unless he is hibernating; he certainly did when he wrote his semi-literate memos. But in another way, it doesn't: No matter what you run, you pay for Windows -- the "Microsoft tax" we all know too well. The proposed settlement actually makes this worse, which I suspect is something that none of us thought possible.

But the bottom line is this: Even if you run Windows, if you have any complaints about it, about its reliability or its vulnerability to any damn batch file somebody cares to slap together and set loose on the Internet, and if you have any brains at all, you'll oppose this settlement. Because the more vigorous and threatening the competition that Microsoft faces, the more they'll have to devote resources to development of good software instead of the thinly veiled organized crime that the Valentine marketing memos represent.

A perfect example: I recently bought a hard drive for a notebook machine. It's an IBM drive, and it is tremendously cool -- there are even Linux utilities provided which allow me to configure its power use, to run diagnostics on it, even to make it quieter. And I spent very little money on it. Why? Because IBM has competition. I could have bought competing drives from any of a half-dozen makers. IBM offered value for money, and that's a far cry from its being the only choice, in which case the price would have been higher and performance lower. And Linux utilities? Fat chance!

If you haven't filed your comments with the Department of Justice yet, get to it. The e-mail address is microsoft.atr@usdoj.gov while the fax numbers are 1-202-307-1454 and 1-202-616-9937. Again, be polite and respectful. (Does anyone doubt that Microsoft has people at work sending nasty and insulting messages from "Linux users"? I don't, not after the company's recent Valentines to its sales force.)

KMail Bug #128 -- R. I. P.

What are you running on your machine that hasn't changed since early 1998? Probably very little. The pace of Linux development is such that there isn't much that isn't in whole new versions in the last three years -- more than whole new versions, multiple whole new versions.

But some things, one might have thought, were eternal.

What causes me to think of this is KDE bug #128.

Bug #128 was the perfect bug. It was noticed by everyone who uses KMail, and pretty much everyone was irritated by it. But it did not overwrite your master boot record, or send your private documents to everyone in your address book, or cause the application to crash, or corrupt your folders. It was annoying, but not so annoying that there would be a general effort to wipe it out. Hence its longevity.

What it did was this: Let's say that you received a message and decided to reply to it, or to forward it to someone, but then, before sending the reply or forward you thought better of it and killed the further message. KMail sets a little flag next to messages that have been replied to or forwarded. And even though you'd not actually sent anything, the flag would remain, suggesting you'd done something you hadn't.

That's it. And it was annoying -- not very annoying, just a little.

There had been attempts here and there to fix it, but the fixes often broke something else and were rejected.

"Close 128" was the terse message from Ronen Tzur, who provided the fix that had long eluded the project. That was in late November, though I imagine that most people even now don't know about it; I learned of it when Waldo Bastian dropped me a note on New Year's Eve. (Still, that night there were mass gatherings and celebrations all around the world, so maybe the word did spread.)

And how long had bug #128 been around? It was the dean of its particular nest, having appeared in the KMail in KDE-1.0 beta 3. Bugs are always coming in, including many things that aren't actually bugs; yet it's worth noting that the KDE bug reporting site has logged about 37,000 reports. This one was number 128 in the list.

I'll actually kind of miss it.

A New Holy War

How many angels can dance on the head of a pin? 9mm or .45? Vi or Emacs or something, um, sane? KDE or Gnome?

While Caldera Open Linux no longer exists in that form, and its distribution channels have changed such that new Linux users are unlikely even to hear of it (something which the Caldera people probably see as a blessing), there is still a Caldera and, more important, a Caldera mailing list. On it one finds some of the sharpest Linux people around, many of whom I think of as friends from years gone by. And it is the locus of a new and fiery dispute: The bootloader war.

The venerable, reliable LILO, or the upstart GRUB?

I'm a LILO guy but I've never used GRUB so if it has virtues they are not known to me. While I'll keep using what I know until it no longer meets my needs, I take no position in this one.

The war erupted as a result of complaints over the miserably broken LILO shipped with Caldera Workstation 3.1. Caldera has embraced grub (as, soon after, did Red Hat). Both sides have people who are so competent in the configuration and use of Linux that I, and you, probably, too, would be embarrassed to have them see how we have our machines set up. They know how to do things elegantly and efficiently.

They seem to break down into two categories: Those who have to administer machines remotely (the LILO group) and those who don't and who favor GRUB. The argument, so far, has to do with the ability to reboot remotely versus the ability to get something to boot locally despite a misconfiguration.

It's lively, and reading the comments is enjoyable. The solution, of course, is for distributions -- including Caldera -- to ship the very latest versions of both. If there's room on the CDs for multiple desktops and even Netscape, there's room for current versions of two boot loaders.

My New Year's Resolution

I've undertaken a little project with no idea how it will turn out or even whether it will turn out. But it has a goal worth achieving.

My idea is to cook up how-to documentation for every major distribution, the purpose being to make current distributions friendlier to low-resources machines.

There are loads of people, organizations, even whole countries, who cannot go out and buy the latest whizbang hardware to run the latest whizbang software. And while there is no way in the world that anyone could deny that Linux development has been remarkable, there's also no denying that it has left many users who would by all rights turn to Linux far behind. We can argue endlessly about the meaning of "free" but Linux costs less than the alternatives, unless if you need a new computer to run it, in which case it ain't cheap.

Am I the only one who remembers when a 500mb drive was huge and 16 megs of memory was luxurious?

Dumps fill with 386s and 486s and VGA monitors. And as they do, talented people in poor schools, poor groups, poor nations lose the opportunity to capitalize on their talents because either they have no machines or the machines available to them run, what, DOS and Windows 3.1.

Even if they could afford to go out and get Microsoft's latest, they couldn't use it. The tragic irony is that they can afford to get the latest Linux -- but they mostly can't use that, either.

So what I hope to do is to sort out how to get the most stuff on the least machine, irrespective of the distribution involved, in a way that it can actually be used. Having poked at this for a few weeks, I've discovered that it's a bigger job than I thought.

And after nearly two years of writing this column, I've come to know that the people who read it include some pretty clever folks.

There's a company in Washington state that, in hope of weaseling out of a class-action lawsuit, has offered software to poor schools. I think we can do better than that. And I think our reason would be better: To provide the means for people to achieve useful skills and the dignity that accomplishment brings -- to bring to computing the philosophy expressed by the old line about the difference between giving a hungry man a fish and teaching him to fish.

This whole idea is embryonic, but I think that everyone in the Linux sphere, no matter his or her philosophical stripe, can agree it's worthwhile. I'm thinking that the documentation comes first, then perhaps a little pressure on distributors to better support low-end machines at install time, and perhaps one day a way to actually deliver machines to those who could use them. (See? I told you it's still embryonic.)

It would be good for Linux, because it would enlarge the user and developer base, and because it would remind us whence we came. And it would be, well, just good.

As I said, I've learned that the readers of this column are pretty bright and certainly enterprising. So your suggestions are welcome. If you have an idea about all this, drop me a note.

Right after you've written your comment in U.S. v. Microsoft and sent it to the DoJ.