Linux Will Succeed Where UNIX Failed

amcpherson's picture

Imagine a world five to ten years from now where an ecosystem of profitable entities competes around a global operating standard for Linux. In this world, application vendors target different platforms at significantly lower cost, whether they are selling scientific applications into clustered environments, desktop applications for end users, or multi-lingual applications in emerging markets. In this world, and only this world, a true alternative to the global domination of a proprietary de facto operating system can be found. The result? Just as with any open system: lower costs for end users, lower risk of technology obsolescence and prevention of vendor lock-in.

This is not a new goal. UNIX tried to achieve this starting over 20 years ago, but met with limited success. Many people in the industry have warned that Linux will experience the same sad fate of .. destined to trade market share for a smaller portion of the market while enviously watching Redmond. But there are significant differences between UNIX and Linux. Most notably there is a global standard that is achieving wide .. the Linux Standard Base. But is it enough? Just what makes Linux different than UNIX? support UNIX

Timing is Everything. Most obviously, the efforts to unify UNIX occurred after fragmentation had already begun and after numerous vendors had vested interests in maintaining that fragmentation. Each vendor was also hedging its .. while publicly expressing their support for UNIX, many vendors had internal Windows NT programs with equal (or greater) funding and human resources. The result was ambivalence at best. bets

At this time for Linux, real fragmentation has not yet occurred, and the best interests of the vast majority of players still lies with maintaining .. there is ..plan .Bno coherence

A Different Technical Approach. Community groups undertook many initiatives to standardize the UNIX operating system: most notably, the Single UNIX Specification (SUS) and Portable Operating System Interface (POSIX). Most UNIX standardization efforts focused on application programming interfaces. Eventually, the POSIX standards (administered by the Open Group) became widely adopted and useful for the industry, even though they did not deliver a unified UNIX. And while these were worthy, state of the art processes, they pre-dated the open source technical and legal models that provide for greater protection against proprietary, closed systems.

At the same time, the LSB builds on earlier efforts that attempted to prevent UNIX fragmentation, such as the POSIX and the Single Unix Specification (SUS). In fact, it uses some of POSIX’s own source code standards and SUS’ interface definitions.

But while the LSB has incorporated the useful aspects of these precursors, the FSG has learned from the UNIX experience and the limitations of POSIX and SUS. Notably, POSIX only defined programming interfaces and therefore could not guarantee binary compatibility. At the other end of the spectrum, standards such as OSF/1, which aimed for binary compatibility, were found to be too restrictive. The LSB aims to strike a balance between the two approaches — it includes a binary compatibility layer that splits the difference between the approaches taken with POSIX and OSF/1.

A New Socioeconomic Environment. Unlike the failed attempts to provide binary interoperability for UNIX, the LSB lives in a very different computing and business environment. A maturing of the systems software market combined with the open source software movement has changed the competitive dynamics of the industry. Instead of the vendor lock-in (and associated high software margins) which characterized the age of UNIX, innovative business models from companies like Red Hat and Novell are redefining the economic dynamic. The greater the number of standardized pieces in a distribution, the more the vendor of that distribution can focus its attention and resources (both technical and marketing) on providing the sort of differentiating value that customers are still willing to pay for: service, support, management software, and so on. Red ..s last blowout earnings report should leave little doubt to this argument.Hat

The Open Source Difference. Another dramatic differentiator between UNIX and Linux lies in the licensing structure of Linux: the GPL. By its very nature, the GPL allows distribution vendors to copy the work of other distributions. This results in much of the same code base between divergent distributions, appreciably narrowing the technical challenge that UNIX faced. The key challenge for open source and Linux is to prevent the difference in distributions that the GPL was not designed to solve: file hierarchy, configuration language, runtime environment compatibility, etc. The Linux Standard Base is designed to solve exactly those problems.

In short, the odds are very good that Linux will not fragment. The GPL, the relative uniformity among Linux distributions, the fact that the FSG and other organizations with similar concerns got off to an early start, the fact that there is a common, proprietary operating system enemy, all make the vision of the LSB easier to achieve. The dream of an ecosystem of profitable entities competing around a global operating standard for Linux is within sight.

Jim Zemlin
Executive Director
Free Standards Group