What projections has IDC made for Linux this year? We are predicting that it will actually grow market share slightly...
and become the No. 2 server operating environment in the market. And we expect it may grow from No. 3 on the desktop to No. 2, possibly by the end of 2003. Can you put into perspective Linux's growth since its early days? Linux has seen dramatic growth from when we first started watching it in 1994 when it had a tiny percentage of market share, to the end of 2001 when worldwide paid shipments [free downloads don't count here], Linux had 26% of market share. This is not server shipments, but software.
Linux is also getting a lot of attention due to the moves Microsoft is making by changing its licensing rules and their talking about shared-source offerings, which is in no way open-source computing. But it is a clear response to the open-source community.
We're starting to see Linux considered more, now that doesn't mean Linux is being selected, but it is being considered and that's something. Is application availability the biggest roadblock to enterprise adoption?
That's part of what is holding Linux back in several areas of potential growth. In some cases, an application like what an enterprise may want may be available on Linux, but maybe a specific application is not. Quicken, for example, GNU Cash does same thing. Can't get Microsoft Office, but you can get StarOffice.
Major vendors have been announcing products for some time. IBM has Domino, Notes and other development tools on Linux; Oracle has made quite a bit about its RAC 9i on Linux. They've announced good benchmarks there, some better than midrange Unix systems.
Linux is on the way to becoming a mainstream OS in most markets. What we projected back in 1995-1996 was that by 2005, it would be a mainstream OS, and clearly, Linux is on that path. What Linux issues are enterprises hoping to resolve soon?
They are on the lookout for enterprise management capabilities. How will vendors fit Linux into their offerings?
They'll also be looking for greater scalability and performance. Already, we've seen SGI produce a 64-processor Linux system. They've succeeded scaling Linux farther than Windows 2000 does, only not as much as Unix systems can scale [2,000 processors].
What a lot of enterprises will be looking for is whether their favorite applications will be available on Linux. Some won't necessarily be looking for performance or reliability, because those are generally good. But, is their favorite application or tool available on Linux? What arguments can CIO make to upper management to make the move to Linux in the enterprise?
It depends on what is going on in an organization and what architecture is in place. There are a number of reasons people use to bring Linux into the enterprise.
Either a budget has been reduced and the requirements of the services IT provides remain the same or increase or a budget is flat and IT is required to do more with less or more with the same. How an organization responds depends on where the cost structure is in a particular organization. If the cost structure is biased toward administrative costs, this would lead you a fewer number of servers and this usually leans an organization toward Unix or Windows, not Linux.
If what an enterprise is trying to do is lower the cost of hardware or software, or move from a RISC system to Intel, there are so many vendors available and so many options that the prices are lower even if the administrative costs are higher.
Another issue is Microsoft's business licensing agreement. If those are overrun to an organization, it may consider Linux strongly.
FOR MORE INFORMATION:
Share your thoughts on this story and other enterprise Linux issues inside the SearchEnterpriseLinux.com discussion forum What are some common misconceptions enterprises may still have about Linux?
A common one is that Linux is free. It is not. You can download the source code if you choose to, but it is important to consider that a complete IT solution is not free. There are staff and equipment total cost of ownership considerations. The TCO of an IT solution over five years, staff-related expenses total 50% to 70% in some cases. When you're done, Windows is cheaper in some cases. You need management tools and staff must be trained on the OS. Staff-related expenses some times make free software more expensive. We urge companies to consider carefully what you have to go through.