SuitWatch

Weekly views on Linux in Business
by Doc Searls, Senior Editor of Linux Journal

Thursday, June 27, 2002


KARMIC ADJUSTMENTS

Here at Linux Journal we often half-joke about the years remaining before we finish working off our "Microsoft bashing karma." (Bashing Microsoft is so last millennium, y'know?) That means the trick is to talk about Our Friends From Redmond in an objective way, which is easier these days because Microsoft looms progressively smaller as a dominant force. (See, I didn't bash. We're talking facts here.)

For perspective, let's remember what life with IBM was like when that company's influence was at high ebb -- a period that lasted through most of computing's history. Those of us old enough to remember will recall referring to Big Blue simply as "the environment". In those days you either lived as fleas in the fur of the great blue beast, or you were, so to speak, out in the cold.

What's interesting from today's perspective is not that IBM "controlled" so much of the industry, or that Microsoft later did the same. It's that both dominated through a long period when it was possible for one company to define the whole world for everybody else. For a century or more AT&T did the same in telephone communications. IBM and AT&T provided base infrastructure for everybody else in their industries. Even if they didn't own the iron or the wires, they owned and controlled countless protocols, standards and foundational products. Gradually that kind of hegemony wore away. Few of us remember how Racal-Vadic came out with the first non-"Bell" full duplex 1200bps modem. It may seem like a small thing today, but at the time it looked like a crack across a vast ice shelf. There were a lot of those, mostly forgotten today.

The ice age is over. What dominates now is the Internet, with an infrastructure that consists mostly of protocols: TCP/IP, HTTP, FTP, SMTP, IMAP, LDAP, FTP, Telnet, POP3... "The history of the Net is the history of its protocols", Vint Cerf says. These are the things that nobody owns, everybody can use and anybody can improve -- although once a protocol is established, improvement tends to come in the form of additional protocols supporting additional services. These are established often by quiet adoption by many companies, rather than by loud promotion by one company.

In the last few days I've heard casual mention that two widely different companies were using the Jabber protocol for routing XML streams between here and there. One is Apple, which apparently will use the Jabber protocol (not the server, just the protocol) in its new iChat system. The other is Fidelia <http://www.fidelia.com>, which does performance management for Linux and Solaris systems in enterprises. Neither company mentions this fact anywhere in its public documents (not that I can find, anyway). If this constitutes a trend (and I suspect it does), AOL's and Microsoft's instant messaging (IM) systems will inevitably find themselves adapting to a world where the prevailing IM protocol is one they don't own.

There is no reason either company shouldn't be able to continue dominating the IM service business, whatever that business becomes. They just won't control IM infrastructure, because Jabber (or one of its decedents) will become the Net's primary protocol for instant messaging and presence detection (the protocol's other main virtue).

As the Net grows in the number and variety of services supported by its basic protocols (like, say, wireless networking thanks to 802.11b, and zero configuration networking -- zeroconf <http://www.zeroconf.org/> -- which will make its first major appearance with Apple's forthcoming Rendezvous <http://www.apple.com/macosx/newversion/>), countless new businesses and market categories will be made possible. This is a good thing, even for dominant companies like AOL and Microsoft. The trick for these incumbent dominators is to play a supportive game, not a lock-in game -- to obey the imperative best expressed by John F. Kennedy: "With great power comes great responsibility".

In apparent compliance with that imperative, Microsoft has been expressing a need to take responsibility for security. Not long ago Bill Gates declared that "trustworthy computing" was Job One. This week that commitment bore fruit in the form of "Palladium," a "vision" that would, in the words of the Wall Street Journal, "Make online security and the protection of some Internet content a built-in feature of personal computers."

The rollout was very carefully orchestrated. In Newsweek and MSNBC <http://www.msnbc.com/news/770511.asp>, Steven Levy called Palladium "one of the riskiest ventures the company has ever attempted. Though Microsoft does not claim a panacea, the system is designed to dramatically improve our ability to control and protect personal and corporate information. Even more important, Palladium is intended to become a new platform for a host of yet-unimagined services to enable privacy, commerce and entertainment in the coming decades." He adds, "Because its ultimate success depends on ubiquity, Palladium is either going to be a home run or a mortifying whiff."

Some of the best thinking about Palladium has been coming from Eric Norlin at DigitalID World. In The 282 Billion Dollar Riddle <http://www.digitalidworld.com/modules.php?op=modload&name=News&file=article&sid=72> , Norlin says Microsoft's grand strategy mooshes Palladium together with the .NET development framework and "Trustbridge," a new "federated identity system for the enterprise." He says,

Among other places, here's where Norlin goes:

Then he makes one more point worth passing along:

In fact, selective interoperability appears to be a key strategy. How? Well, get this: Microsoft is going to publish the Palladium source code. "We want to be transparent in all this," Jim Allchin says. For more of the rest of the story, Extreme Tech has a rather deep analysis. <http://www.extremetech.com/article2/0,3973,274309,00.asp>

Meanwhile, right next to the Palladium story in Monday's Wall Street journal was a larger, above-the-fold item that was anything but good news: "New Microsoft Pricing Looms". The subhead said much more: "Some Customers Consider Switching to Rivals' Software". The larger story is about Microsoft's new software pricing policy, which takes effect at the end of July and has been anything but a PR success. Says the Journal,

Meanwhile Ximian is calling Microsoft's July 31 deadline "Y2Pay," and offering customers a special deal:

Any business Ximian gets out of this will be a huge boon, even if most Microsoft customers go ahead and sign new contracts (which is the default assumption). Operating systems are still continuing to commodify at a rapid pace. Linux is simply taking the lead in that much broader movement. The new commodity operating system isn't Linux, but whatever UNIX you feel like using, Linux included. If you're working under the hood in command line mode, the differences between Linux, BSD, Solaris and OS X matter less than the fact that down deep they're all breeds of an OS that's purely infrastructural.

I suspect that Palladium isn't as infrastructural as Microsoft would like it to look. But it still involves a number of concessions that make clear that the company is simply a significant inhabitant of a larger ecosystem, rather than the whole environment.