SCO, Linux and the goose with golden eggs - Embedded.com

SCO, Linux and the goose with golden eggs

Recently lawsuits and threats of legal action have descended on the unlicensed use of hardware and software offerings that were previously considered based on open standards. Given the current downturn in the economy, on the one hand, and the future market opportunities in the new connected computing environment, on the other, such moves are understandable.

But no matter how legally justified the companies who are taking such actions think they may be, these developments are ill considered. They focus on short-term tactical gains without looking at the long-term strategic goals. And while they may benefit the companies who have initiated them over the short term, their ultimate impact could seriously damage the growth of the market in which they wish to participate.

The most spectacular example of the current wave of legal actions is the effort by the SCO Group, which now “owns” Unix, having purchased it from Novell, which bought it from AT&T. Much of the activity involves not “illegal” use of Unix, but the use of code that is supposedly proprietary to Unix in its open source relative, Linux.

The turmoil in Linux/Unix is not an isolated event. Similar efforts are going on in the 802.11 wireless LAN space and in the delivery of multimedia over the Internet — attempts to take what were previously assumed to be open source or at least standards-based mechanisms and establish proprietary and income producing claims.

That SCO, a vendor of operating systems software and operating systems for workstations and servers, is taking such an action at this particular time is a no-brainer in the context of the impact that Linux is having on SCO's base of customers for Unix in these market segments.

And given the widespread use of Linux in small footprint computing and communications appliances as well as in many embedded applications in the network infrastructure, the compulsion to go this route for SCO must have been overwhelming.

Just from vendors of Linux in workstations, servers and even desktops, a significant stream of royalty or licensing income would be forthcoming. And given the potential market for small footprint appliances that are being designed using Linux, winning this case in court would make SCO the Arm or MIPS Technologies of the software world, drawing most of its income from royalties and licenses and spending only such monies as necessary to improve the “product” and adapt it to new markets and applications.

My concern about the impact of such cases has nothing to do with the legal issues involved. It has to do with the precedent they establish both legally and in the market and how this could change the fundamental nature of the World Wide Web and the Internet.

In short, what will happen to the open source, standards-based goose-that-laid-the-golden-eggs that has driven the Internet and World Wide Web into virtually every aspect of our lives and into every embedded device and connected computer? Will such lawsuits such as these, which on their face may or may not have legal validity, kill that goose?

Consider the precedent such cases will have throughout computing. For example, the starting point for most “proprietary” RISC architectures is the work done under Department of Defense grants to Stanford University and the University of California, Berkeley. How much do you want to bet that there is not a lawyer someplace going back and looking at what portions of the RISC architecture preexisted the federal funding, hoping to find a loophole by which a university or a company could gain income from royalties?

And what about TCP/IP itself? Developed at Stanford University by Vinton Cerf and Bob Kahn, it was at the heart of the DoD funded ARPANET, an “inter-network” of proprietary and specialized networks designed to link the nation's universities and research labs together. Who does that belong to: all of us, to the government, to Stanford or some other university or to Cerf and Kahn themselves? If its use or any key element of it preexisted its use in the federally funded inter-network, would that not be a pretext for claims from which income could be derived?

Additionally, there are the many so-called “little languages,” that have emerged out of the Internet such as TCL, Perl and others, the products of which are assumed to be in the public domain. Not only are these scripting languages used in dozens of Internet-oriented products, they are at the core of many products not directly related to networking. TCL, for example, is the mechanism that several embedded tool vendors, such as Wind River Systems, use to tie diverse tools and services together into a common integrated development environment. It is also at the core of at least one business-to-business web services offering.

Nor are the new Web Services frameworks that have been developed all that safe, based upon XML and Java on the one hand, and XML and C-sharp in the case of Microsoft Corp.

XML is pretty much a product of the World Wide Web and IETF standards-driven efforts. But is not just an extension of HTML, which old-timers I have talked to refer to derisively as “hypertext in a straight jacket.” XML is based on the proprietary work done by dozens of hypertext companies in the 70s and 80s. They all had platform, OS and database specific meta-language expressions and meta-protocols that are just now being added back into XML, but which had to be stripped out to make HTML useable on any platform. How similar or dissimilar is the code used in these proprietary fully hypertextual programs to that in XML?

And Sun's Java and Microsoft's C-sharp could be in trouble here also. Pre-existing both were several other comparable efforts that look awfully similar. For example, there is AT&T's long forgotten C@+, a language which combined some elements of C, C++ and Smalltalk. The language was designed to be platform independent and it allowed the creation of downloadable virtual machines or run time executables in ways very similar to what Java and C-sharp do. And what about whoever has some legal rights to any or all of the Smalltalk language, many aspects of which served as the conceptual underpinning of both languages?

Certainly, the claims of the inventors and originators of a proprietary language, operating system, architecture or protocol should be considered. But what about the claims of the company to which they sold the rights, or the company to which that second firm eventually sold out to. Should all those claims be considered equally? Or should it be something like what has emerged in real estate law in states like California?

The owner of land over which there is a path by which there is access to public lands on the other side only has rights if it is established at the time of purchase or if the access is the result of work performed by the owner. If the original owner makes no effort to block access, a later purchaser does not have rights equal in magnitude to the first owner. They cannot, after the fact and after its use in the public domain, retroactively establish ownership, or claim misuse or illegal use.

The only company that I see as acting like an adult about this is IBM. This single company has thrown its corporate weight behind standards and open source and against proprietary offerings, while the rest of the industry has reverted to childhood: throwing things at one another, arguing over “who started it,” and coming up with ways to get a bigger share of the pie than the other kids.

After a 50- or 60-year history of developing proprietary hardware, software and communications protocols — and charging an arm and a leg for their use — the company has gone the other direction, opting across the board for standards. Not only does it use standard and open source software throughout its products and services, it has moved some of its proprietary offerings, such as the internally developed platform and language independent integrated development environment, which is now in the open source domain as part of eclipse.org.

Its general philosophy seems to be not to try to make money on proprietary technology, but rather take standards and open source technology and make its profits building products that are faster, better, and easier to use. Long term, that seems to me to be the only way to ensure the long-term health of the connected computing environment upon which we all depend.

Bernard Cole is site leader and contributing editor of iApplianceweb and an independent editorial services consultant working with high technology companies. He welcomes your feedback. Call him at 928-525-9087 or send an email to .


Reader Feedback

A few points about SCO and Linux:

1. SCO is suing IBM over alledged breach of contract. The suit involves noother parties, including Linus Torvalds, or any other Linux kerneldeveloper.

2. SCO was still distributing *their* version of Linux for months after theybecame aware of alledged “code lifting,” which means, under the GPL, anysuch “lifted” source code is now in the public domain.

3. Even if there is any such lifted source codes, it is extremely likelythat it is very small, which would be rewritten in hours if not days.

Linux is safe, and people who rely on Linux have absolutely nothing to worryabout.

Mike Wagner


If a person copies from one source without acknowledgement, the copied structure is a product of plagiarism.

But if a product is made up by altering the structure of another and combining it with bits and pieces from various other sources, the final product is “new” in nature and it belongs to the entity that put it together, if that entity chooses to lay claim to it. If that entity chooses not to lay claim to the product, the product now belongs to the public domain.

SCO may have a leg to stand on in terms of its claim that IBM leaked source code that was contractually licensed to them.

SCO on the other hand does not have a claim as to the ownership of the code in the Linux Kernel, as alterations were made to the code prior to the release and distribution.

If SCO do go after various companies that distribute versions of the Linux Kernel in question, they might open themselves up to law suits of libel where they will have to pay out more in damages to the companies that lay a claim against them, than what they will receive from the IBM lawsuit.

I hope Mr Mike Wagner is right in saying that Linux is safe. Open-source code in the public domain is a great resource for us developers, especially those of us who have startup companies and we charge customers for our time and not for intellectual property, as we don't have the resources to pay for patents and copyrights.

Sumoj Simon

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.