Gaining momentum in niche areas
Wagner Rios
Industry Analyst,AMR Research
Linux has demonstrated its worth at the server level and in clustering. In addition to being a robust operating system, its nonproprietary nature allows OEMs to tweak the architecture for resource-intensive applications and environments. It also allows freedom of movement away from Microsoft--something the industry has sought for quite a while. Network administrators report consistent performance advantages of Linux over Windows NT and 2000. It would be a wise move for Microsoft to make its application suites operational on Linux, as the Linux platform continues gaining market share and there remains a shortage of applications for it.
We can expect Linux to continue gaining momentum in niche areas such as finance, telecommunications, Web hosting, applications service provision, and similar environments requiring high resilience and reliability. Linux will continue to attract developers in part because of its memory management capabilities, including process swapping, file buffering, and direct memory access (DMA); virtual file system and the second extended file system; and interprocess communication (IPC).
Unix and Windows fight back
George Weiss
Analyst, Gartner
Linux will threaten but not replace Windows or Unix as the OS that corporations prefer in all server tiers through 2005. Linux will move through three phases of acceptance. In Phase 1, Linux battles for corporate credibility in network services, with early adopters enthused by Linux's robustness and low cost. In Phase 2, Linux gains durable roots--particularly in Web servers and clustered server farms--partly at the expense of Windows 2000 and costly Unix/RISC solutions. In Phase 3, ISV enthusiasm for Linux increases. But Unix's mission-critical scalability and availability, strong Windows 2000 upgrades, and potentially heavy migration costs throttle broad ISV enthusiasm. Increased complexity and insufficient skills will also retard corporate acceptance.
Success will get complicated and expensive for Linux distributors since value will come from an OS/middleware software stack and services. Most ISVs will commit resources only to the top two or three Linux distributions. Some Linux vendors will inevitably fail; others will merge or be acquired.
IBM boldly stated that in 2001 Linux will surmount the chasm separating it from the enterprise. If so, IBM is best positioned to capitalise. However, Linux faces multiple chasms (such as standards, ISV support, skills, scalability, and availability), any one of which could stop the show and leave IBM as the most exposed vendor. IBM will likely continue pushing Linux harder than corporations will accept and may have to adjust its enthusiasm to more rational levels.
The dot-bust has sobered the dot-com speculative frenzy rather than condemned any technology. Linux IPO makers did not suspect the lurking disaster. Today, Linux no longer guarantees a successful start-up and is often buried within the application context (for example, server appliances). That seems about right. Linux is valuable for what it enables you to do and is not an end in itself.
Microsoft's paranoia about Linux is perfectly understandable.
They know they have gone as far as they could and that the future will be stagnant before they lose more and more business to more security consious businesses.