Why has IBM embraced Linux, along with some of the Open Source religion? As a profit-seeking entity- which has a track record for innovation (at least in basic physics and semiconductors), why are they defending Open Source and- indirectly- the GPL?

Linux helps codify and stabilize the standards we see the Internet and e-Business resting upon. Linux, freeBSD and their brethren, with their accessible source code, provide the implementation model for many- if not all- of the communications protocol standards we see in use today. NFS still works with ancient equipment. So does TCP/IP. Newer standards, riding on top of TCP/IP provide the "World Wide Web" as we see it today. These standards are under attack, for they are secured- and stabilized- by the GPL and various Open Source licenses derived from the GPL.

Who owns the railroad gage? This is an open standard- albeit an old one. While there are multiple gages for railroad trackage, there has never been anyone who had final control over whether or not a railroad company could use it for their own trackage. Standards allowed more companies to manufacture the railroad cars.

Open standards are interesting- look at BetaMax versus VHS. Sony charged fairly hefty license fees for their intellectual property while JVC followed Phillips' (of the "compact cassette") approach and licensed their IP cheaply and widely. (I recall a story claiming that both Phillips and JVC gave "free" licenses to any manufacturer who followed their specifications. I don't know the truth of this.) So, JVC's VHS had the more open standards; whose format defines the standard?

If the standards become hidden within the code of various proprietary environments, the standards that define the Internet itself are at risk. If, indeed, Microsoft Operating Systems underlie the largest "tenant" of cyberspace- at least if they think themselves so- Microsoft can change the standards without needing to share the critical information with any of their competitors. It has been said that Microsoft's own compliance- with their own standard (SMB)- is poor at best. If we see the de-standardization (at least for open and testable standards compliance), we enter the realm where a standard is "whatever Microsoft thinks it is- at least for this afternoon". And, by the pressure of the Microsoft client systems, the servers would need to be traded for Windows-based servers since your customer won't be able to communicate reliably with your non-Microsoft servers.

This tactic is nothing new. The "Embrace and Extend" (Engulf and Devour?) tactic the Halloween documents from within Microsoft have been around for a long time. The "drift" of Java (with their now-dead Visual Java) was about undermining standards; an attempt to make Microsoft's Java Runtime the standard by using the weight of their customer base to ensure their monopoly.

Each effort to chip away at our standards does all of us a disservice. Monopolies are not known for competent service (have you been to a DMV office?) since there's no one else for you to buy from. Standards like the NTSC and FM radio have enabled communications mediums; once the FCC had locked them down; it was a little harder to close the medium from competitors. HDTV is supposed to be the next big thing- and there are folks trying to undermine it.

Without the "open source" pre-dating the GPL (implementing the ARPANET between universities and the US DoD) which placed implementations where all can see them, the semantics of the interface laid out for all to see, the Internet, as we know it, would have been impossible. Many of the standards grew from the Universities where the licensing had to be wide open (being government-supported bodies) and where having the source was assumed. Unix grew in the Universities because AT&T's Bell Labs wasn't allowed to sell it for more than the cost of the tapes. This was source- and helped define Unix's APIs as standards- all because people could look over the code to be sure of the implementation.

It works like this in research (except for genomics where many researchers are more concerned with profits than answers): we have learned more about the physical realm all because researchers shared the knowledge with their "competitors", forming a cooperative organization where the sum is greater than the parts. As in open source, one individual may help another jump ahead, but the human race as a whole profits far more than any individual. It didn't hurt that most of the researchers from centuries ago didn't depend on their scientific research as their "day job"; perhaps we need more "amateurs".

Wither Standards? Only if Microsoft has their way.