This is part two of our exploration of the state of the Windows-versus-Linux rivalry. In part one, Microsoft expert Jeremy Moskowitz presented the view from the Windows side. The experts appearing in this installment are: James Turnbull, consultant, resident expert for SearchEnterpriseLinux.com and author of Hardening Linux; Kenneth Milberg, consultant and resident expert for SearchEnterpriseLinux.com; Frank Hoberg, CEO of Open-Xchange Inc., a Tarrytown, N.Y., provider of open source-based groupware, collaboration and messaging software; Jan Hichert, CEO of Astaro Corp., a Burlington, Mass., provider of network security products for Linux; Paul Froutan, vice president of engineering for Rackspace Managed Hosting, of San Antonio; and Marge Macintyre, program director of Worldwide Linux Strategy at IBM.
Read on, and you'll see that these experts don't always agree.
How do you assess Linux's progress in the battle against Windows in the server operating system arena?
Frank Hoberg: The pace of development, the skills of the community and the stability of Linux really lead me to believe that Windows' days in the data center are numbered. I am not talking tomorrow or next year, but the end game is inevitable: Windows will lose out in the data center.
Marge Macintyre: The momentum of Linux into the mainstream of computing is unstoppable, growing faster than any other operating system. The innovation that results from the collaboration of a worldwide technology community has made Linux an industrial strength, open source operating system. Workloads are moving to larger servers to meet the scalability needed for large enterprise solutions. IBM is participating in this rapid growth and already realizes approximately 40% of its Linux server revenue from the Power and z-architecture platforms.
Functionally, how does Linux on the server stack up against Windows today? Hoberg: Some things are missing with Linux. Windows is still easier to set up and administer, as long as you are doing standard things with it. Linux and open source software in general have an 'if-you-don't-get-it-you-are-not-worth-it' attitude; but we are seeing a lot of progress here, too.
In terms of security, Linux does generally have an edge on Windows. For example, Linux hosts tend not to be as vulnerable to worm and virus attacks in the same way Windows are. Additionally, personal experience tends to suggest Linux hosts are more stable and provide better performance than equivalent Windows hosts.
What is the status of the Linux business desktop?
Hoberg: On the business side, the move to Linux is already happening. Linux is perfect for single application or workstation deployment, and even knowledge workers can take advantage of the flexibility and security of Linux.
What are the chances for widespread adoption of a Linux desktop that uses an office suite other than Microsoft Office?
Kenneth Milberg: I would say this depends on the success, or lack thereof, of a competing open source product.
OpenOffice is a great product. I have used it, and in many ways I prefer its word processor to Microsoft Word. Unfortunately, it has not been publicized as a real alternative to Word in the same way that Linux has been promoted as a real alternative to Windows. Largely because of this, the product is no real threat to Microsoft.
As many of us know, WordPerfect was once a great product, but Novell ended up running it into the ground, and Corel could never get it back on track. I would say that the only way that there can be widespread adoption on the desktop without Office, is for some other company (Are you listening, the "new" Novell?) to take this product or a similar one and spend enough money on it to make it competitive with Office.
While the corporate world may use Linux in the server room, it will not let it near its desktop user base without a desktop product that makes its users feel as comfortable as Office. That is the reality -- as unfortunate as that may sound to us in the industry -- regardless of how good a Linux desktop operating system either Red Hat or Novell come up with.
Some CIOs have told us that they don't think open source applications are mature or robust enough for enterprise usage. Do you agree? If not, could you offer examples of enterprise-ready apps?
Hoberg: If you define software as a stack -- with the operating system at the bottom, middleware like Web and database servers in the middle and applications on top -- open source is doing very well in the first two layers with Linux, the Apache Web Server that runs most Web sites and the MySQL database.
Getting to the actual applications: There's the Firefox Web browser and Thunderbird eMail client, some interesting customer relationship management (CRM) applications and, of course, robust office suites like OpenOffice.
Hichert: The analyst firm Gartner indicated that the Snort Project is mature and fully suitable for use in enterprise environments. Another example is OpenXchange, the groupware and collaboration server that was the engine behind SuSE Linux Openexchange Server. Now, the commercial version is even better and could easily replace Microsoft Exchange Server in a lot of environments.
Our experts have offered several examples of mature, enterprise-ready open source products. So, why do you think chief information officers have the perception that there aren't any available?
Paul Froutan: The challenge of many open source applications is that since they are not backed by a commercial company, there is no marketing and little awareness of their existence. There are hundreds of applications that many CIOs could benefit from if they only knew they existed. Many traditional CIOs are not aware of how to navigate their way through the open source ocean to catch the big fish they are looking for.
Turnbull: My predictions for Linux adoption in the next two years are a steady but not avalanche-like growth in the use of Linux, especially as desktop replacements for Windows and as workgroup and SMB application servers.
I suggest a lot of this growth will be in government and public sector organizations that see the introduction of Linux and other open source software as being more cost effective than commercial alternatives.
I think it will be unlikely that the larger, more conservative organizations, such as those in the banking and finance industries, will adopt Linux on a large scale in the near future. In these environments, there are often too many existing licensing agreements, buying arrangements and outsourcing agreements to allow new products into the environment.