A few days ago, I wrote about the
shift of wealth in the face of labor migration. Today I happened upon a couple of stories in
linuxtoday that talked about the shift of wealth in the face of "Open Source Software". OSS -- such as
OpenOffice (a free alternative to microsoft Word), MySQL (which you can purchase support for or use free, as an alternative to proprietary databases), Apache (which powers most of the web), and Linux -- is usually cheaper to purchase (e.g., free) and run, particularly since the user has the right and ability to modify the software (you have the source).
Anyway, the point of
one of these stories was that OSS was costing proprietary software companies something like $60 billion dollars a year, but as several
observed, this means that industry, government, education, nonprofits, and consumers
saved $60 billion in software costs. All this assumes that the figure of $60 billions is accurate, but let's just stipulate that and consider what this means.
First, the US government spends a lot of money on IT, on both hardware and software. The GSA looks favorably on OSS, but according to
this story it isn't particularly to save money; it's to get a better software user experience. "Vendors are competing on the quality of their support, and they'll take responsibility alongside your agency, for the maintenance of your application." How can this work? Let me describe briefly the proprietary model, using the vendor I love to hate, Mi¢ro$oft, and then I'll describe how the open software model works using, say,
Apache.
So the stated objective of M$ from the beginning was: a computer in every home and on every desktop, all running mi¢ro$oft software, or something very close. To achieve this, they used a set of illegal practices
Let me interject here that neither Judge Jackson's conclusions of fact nor his conclusions of law were overturned by the appellate court. M$ were, and are, guilty of the charges. Then "W" got elected and decided to drop the government's case. This is why M$ still operate pretty much as they did before the trial.
to leverage their monopoly in
operating systems in order to achieve and hold a monopoly in applications.
Once upon a time, when there was a real market for word processing programs (and data formats), Mi¢ro$oft Word™ had to interoperate with them. If you had a wordperfect document, for example, M$ Word could read that document, and maybe write it back (I'm not sure about that part). If they could write in other formats then, they certainly don't do that now. They don't need to; they're the 800-pound gorilla. As a picture of their market power now, they recently changed Office 2003 (via SP3) so that it
couldn't even open old Mi¢ro$oft office documents. (An uproar ensued, and mi¢ro$oft released a subsequent service pack that re-enabled the old formats. If these guys weren't just testing the waters, then they are even more evil than I'd thought.) You see, the point is to achieve vendor lock-in -- get the customer's entire workforce to "upgrade" from an old version to a new version of MS-Office. Then, when that customer starts sending around documents with the new formats, that tends to force everybody else to "upgrade" to the new MS-Office, and keep the cash flowing.
What happens when there's a software defect (or "bug") in some microsoft product? Can you file a problem report and get it fixed? Ha! If thousands of people report the problem, and columnists write nasty things about how awful it is, then M$ might fix it.
How can they get away with this sort of thing? By controlling the source code to their software. If there is a problem with microsoft word, who can fix it? Only microsoft! What's their incentive to fix any problem? Well, think of it from the customer's point of view. If you've got thousands upon thousands of microsoft word documents in your archives, then things have got to be pretty bad to get you to switch to some other program; Mi¢ro$oft have achieved "vendor lock-in" with your organization. Therefore, the incentive for micro$oft to fix any problem is quite low -- they can't let the product rot so badly that customers will take the trouble to switch, but can invest just enough to keep the existing customers from switching. Meanwhile, the network effect (e.g., many microsoft word documents in emails) will help them preserve their monopoly.
Open source model
Now consider a different model, where the source code is "open" or "free". Say you want to host a web site with some custom capabilities. One natural choice is the Apache web server. How much does it cost to get this software? Zero; you go to their website and click "download." What if there's a bug with this software? You have several ways to go. There are user forums, you can buy support from some company or consultant, you can diagnose and fix it yourself.
That's right, you can have your staff read the code (remember, it's just code!) and make changes to help diagnose the problem, and fix it. Same thing for enhancements -- you want to add some capability, you can ask others how they've done it, you can hire somebody to create the needed module (and to support that module if you like), you can have your staff do it.
Where is the competition here? Well, if you have several support vendors and all of them know Apache, you can pick and choose from among them. (in the case of mi¢ro$oft office, you have ONLY mi¢ro$oft)
What you do not have in this model is the apache foundation saying, "Ah, I have a new data format, resistance is futile, you will be assimilated." Because if they did, anybody else could say, "Ja, shure, enjoy your new format, I'll stick with the old version." Notice: Microsoft introduces a new version and takes the old version off the shelf. This way they can introduce incompatibilities and force "upgrades". By comparison, Apache is now on rev 2.x, but the 1.x versions work just fine and you can still get them -- if not from apache.org directly, you can find sources on old Linux distros.
Winners and Losers
So software like Apache, Linux, MySQL, OpenOffice, FreeBSD, Mozilla/Firefox... it's out there in the world today, and saving governments, industry, education, nonprofits, and consumers quite a lot of money.
Relating this back to my
earlier posting on labor migrations, how does this affect either labor migration (or job migration (outsourcing)?
Well, there are big losers, no doubt about that. Big losers would include the vendor I love to hate. Now if Linux, OpenOffice.org, etc. are successful beyond Mi¢ro$oft's worst nightmares, what would happen? Well, their sales would drop, and there would probably be layoffs. Unlike Gilder, I don't believe layoffs are intrinsically good, but a bunch of the work that goes on at M$ is stuff nobody wants anyway. How many features of microsoft word do you (or most others) actually use? And if 90% of the users don't use 90% of the features, why is all that crap in there? (If you have any further doubts, read Moody's
I Sing the Body Electronic.) So profits would shrink for M$ -- they extort too much money from the populace anyway so I have no tears for them.
For their layoff victims, though: I believe that the software support area will grow. Remember the
GSA story above? They don't necessarily save money by using open source software, but they get more of what they want. They get better service because vendors are working with them and are partners in the application's success. (Much better service than you'll ever from microsoft, as long as M$ believe they can't be fired.)
So if at least some parts of government and industry are spending the same amount of money as on proprietary software, what does that mean? That more of the money is going to people actually doing work people want done! If the people working today at microsoft are developing stuff that almost nobody uses, wouldn't they be happier doing something that will actually help a customer achieve their mission?
By the way, this also means less money going to mi¢ro$oft to be spent on lawyers and lobbyists perpetuating their monopoly.
So I find myself agreeing with Stallman on many points of his
GNU Manifesto....
That's what I think tonight, anyway