There's been some interesting debate in the Open Source community regarding Open Source applications that run both on Linux and Windows. One camp feels most users select an operating system based on its available applications. If the applications people want are on Windows, they will tend to stick with Windows. Conversely, if the applications they want are only on Linux, they will eventually end up using Linux. By porting free software to Windows one increases the valuable applications on that platform. If Windows has Microsoft applications plus a stable of free software apps while desktop Linux has only the free software apps, why would anyone switch to Linux (and incur the training and data migration costs) when they already have all the software they need and want right? And as long as Microsoft can keep people on Windows Microsoft will gain the time needed to improve its applications and, most importantly, the supporting software stack.

The second camp feels that Open Source applications that run on both Windows and Linux is an important step in "mainstreaming" Open Source and Linux. Since most desktops run Windows, why not infiltrate (and infuriate) the Empire with Open Source applications such as Firefox and MySQL to "seed" the masses with the concept that Open Source is ready to be used beyond the intimidating world of the techies. Giving users the chance to use an Open Source application on Windows lets them get comfortable before migrating to Linux. Anyway, a transitional desktop (that runs Open Source applications on a Windows platform) is an important first step in migrating to a Linux desktop.

I was interested in posing questions on this topic to various people that work with, contribute to, or provide customer support and consulting for Open Source applications that run on Windows and Linux.

Featured Panelists:
Marten Mickos, CEO, MySQL AB
Andy Astor: CEO, EnterpriseDB
David Boswell: Member, MozDev Community Organization

Q. Do Open Source applications on Windows hurt or help Linux adoption and why?

Marten Mickos: For the most part, this question is irrelevant. Open Source products don't exist to hurt or help other Open Source products - they exist to deliver value that hasn't been available before. Linux is likely to grow significantly in the market irrespective of whether Open Source applications make it or don't make it on Windows. But I think it's also good to offer users a wide selection of platform choices and let them decide which they want to use.

Andy Astor: They help, but only a little bit. Once Windows people see the quality of Open Source software, as well as products based on Open Source, they may consider looking for more Open Source in their infrastructure stack. But since the total cost of ownership for commercial Linux and Windows is relatively comparable, Windows lovers aren't going to be particularly motivated to switch.

David Boswell: Anything that helps bring attention to Open Source software should help the adoption of Linux and other applications. Although Linux is well known in technical circles, computer users as a whole are largely unaware of alternatives to Microsoft and other popular proprietary applications. Introducing people to Open Source software by installing Firefox on their computer won't necessarily make them Linux converts, but it's a start.

Q. From an enterprise perspective, should you adopt Open Source applications on Windows before considering moving to Linux?

Andy Astor: I would say no. People new to Open Source will find much more education in a Linux environment than they will in Windows. Experimentation is probably best in Linux.

David Boswell: Introducing Open Source applications on existing Windows systems before switching to Linux makes a lot of sense. On Windows, OpenOffice and Firefox can live side by side with Microsoft Office and Internet Explorer. This lets people get familiar with these new applications without having to learn a whole new operating system at the same time. If an organization does make a switch to Linux later, the time spent with Open Source applications on Windows isn't wasted since those applications, like Firefox and OpenOffice, are essentially the same on both platforms.

Q. Drilling down on those using Open Source apps on Windows - are these new users or are they migration users?

Marten Mickos: The majority of MySQL users (on any platform) started fresh on MySQL and didn't migrate. But now, with MySQL's stronger enterprise presence, we see a growing number of customers who are migrating from legacy databases over to ours.

David Boswell: My dad is a recent Firefox convert and he hasn't used any Open Source software on his computer before. He made the switch out of concerns with Internet Explorer security. I think he's fairly typical of users who are being introduced to Open Source initially through desktop apps like Firefox or OpenOffice. People will use the browser and productivity applications that came with their Windows machine and only install new software if they are looking for an alternative.

Q. What percentage of development efforts is used to create the Windows version of Open Source applications as opposed to the Linux version?

Marten Mickos: When it comes to platform support, our first priority is to write source code that runs unaltered on any operating system from the start. Thereafter we spend a fairly reasonable amount of time and money on tuning and trimming for a specific operating system. We devote those efforts in proportion to the size of the business opportunity on the operating system in question. We are not a non-profit. We want to make money!

David Boswell: For Open Source applications that run on multiple platforms there's a huge incentive to reduce the amount of platform-specific code. Ideally all development would be devoted to creating one piece of software that ran on a variety of different operating systems. It isn't usually realistic to make 100% of the code platform-independent, but communities will move in this direction. One example of this is Mozilla's creation of XUL (the XML-based User Interface Language). When Netscape first opened its browser code, the front-end was built with platform-specific code. Making one change meant making that same change up to a dozen different times. XUL fixed that situation by moving all front-end code into an XML format that let a developer make one change that would work across all platforms.

Q. Why do people continue to pay $350 dollars for Microsoft Office when they can get OpenOffice for free? The power of the Microsoft brand and reach? The lack of OpenOffice education? Is Microsoft Office superior? Uncertainty with the interoperability of the files?

Marten Mickos: Brand is key. Interoperability is another major reason (or, more specifically, the perception of differences in interoperability). Generally speaking, to break into an old market with a new product, it's not enough to have a slightly superior product. You need some vastly compelling reasons for the customer to switch (such as a new business model, new pricing, faster performance, etc.).

Andy Astor: Individuals have no motivation to switch. Most of us work for corporations that standardized on Windows and Microsoft Office. So, until CIOs demand that all applications default to Open Source (five years from now), Microsoft Office will prevail - but not forever.

David Boswell: I think it's a combination of many reasons. People are familiar with Office and are reluctant to change and many people haven't even heard of OpenOffice. You could also add distribution to this list. Most consumers probably get Microsoft Office when they buy their computer and the extra cost of the software is less noticeable than if they were buying Office separately. If Dell offered a configuration option to have OpenOffice installed on your computer for $5 instead of paying for Office, I bet a lot of people would give it a try.