Gartner: Linux Not About to Do Damage to Windows

[saracen]

Microsoft Windows will not suffer irreparable damage on the server side at the hands of the Linux operating system over the next five years, Gartner analyst George Weiss told attendees at the Gartner Open Source Summit. In fact, in terms of worldwide server operating system revenue, Linux would come in below both Windows and Unix by 2011 in spite of its enormous growth, he told attendees in a session entitled "Enterprise Linux: Has it Arrived?"

http://osnews.com/story.php/16021/Gartner-Linux-Not-About-to-Do-Damage-to-Windows

Thoughts?

[kripkenstein]

In fact, in terms of worldwide server operating system revenue, Linux would come in below both Windows and Unix by 2011 in spite of its enormous growth [emphasis mine]

The key issue here is 'revenue'. Linux is very cheap (Red Hat, Suse), or even free (Fedora, openSuse, CentOS, Debian, etc. etc.). Windows is not cheap; neither is Unix.

So just by comparing the revenue, you don't learn very much. Linux might have twice the market share of Unix but only a tenth the amount of revenue.

The interesting question is market share, in my opinion. I expect the current trend to continue, i.e. Linux replacing Unix, but Windows remaining fairly solid - mostly due to Microsoft leveraging their desktop monopoly, etc.

[BoneKracker]

I'm an IT Strategy consultant, and I think Gartner may be wrong on this (except in that "significant damage" is quite subjective -- it's somewhat a matter of perspective and spin).

The key decision-maker (CIO, CFO) perspectives on this however are enterprise architecture and return on investment. The burning need is effective integration at reduced life-cycle costs.

When considering these needs in terms of enterprise architecture, desktop and server operating systems become much less salient and are simply one of hundreds of choices that must be made to establish an interoperable and adaptive "system of systems".

Much more thought is being given to things like "service-oriented architecture", middleware, identity management, virtualization, and other subjects. Server platforms, one of which is much the same as another, drop to commodity status. In that context, Linux has a huge cost advantage.

Secondly, as GNU/Linux and BSD become more usable and personal use consequently expands, more and more people are comfortable working with it as a server or development platform. This erodes a major resistance factor coming from mid-level IT managers who have historically been polarized with respect to platform choice.

Thirdly, while changing at an almost geologic time-scale, governmental choices carry enormous economic and social weight. More and more, the open source and free software models make undeniable sense for their use, and they are beginning to realize this. As a separate dynamic, widespread government adoption would have a domino effect. As an example, over the past decade, the U.S. IT services and products industries have benefitted immensely from regulatory changes favoring the use of "Commercial Off-the-Shelf" (COTS) IT solutions over custom-developed ones. Similar legislation favoring the use of open source or free software would have comparable impact.

Governments actively seek opportunities for their works to serve "dual-purpose", where they are useful not only in satisfaction of their original governmental need, but in satisfaction of similar needs throughout academia, industry, and the general public. This is a very, very close fit (an increasingly, undeniably close fit) with the reusability model of FOSS.