An open source advocate chats about security, intellectual property, and what might be ahead.
The whole appeal of open source is that it’s not created in an ivory tower somewhere–it’s truly by the people and for the people. So when it came time to discuss the future of open source, we thought it made the most sense to ask someone who’s spent some time in the trenches with it. Josh Coates is currently principal of Berkeley Data Systems and volunteer with the Internet Archive, but before that he was founder and CTO at Scale Eight, which built the world’s largest online storage system, supported by a network that handled more than a gigabit of WAN traffic. That system, like the Archive and most of the items on Coates’ résumé, was built on an open-source OS. Like most people who’ve spent their careers immersed in open source, Coates feels excited and optimistic about the platform’s future.
Is there a segment of computing that holds more promise for open source than any other?
Open source has penetrated nearly every sector in computing, though some open-source software has been more successful than others. For instance, Photoshop still dominates over Gimp, but Apache dominates over IIS.
Open-source software is strong in operating systems, compilers, languages, development tools, Web servers, and cluster computing tools. You should also note that much of the Internet is basically run on open source programs like Apache, Bind, and Sendmail.
What about segments where open source might face resistance? Would security be one, simply because the code would be inherently hackable?
There is a healthy debate over whether or not open source is more or less secure that closed source. Open-source advocates contend that if the source is open then it will be more widely audited, and thus ultimately stronger against attacks. Detractors contend that open source is more vulnerable to trojans and can be more easily “cased” by attackers.
Google, Amazon, and PayPal are examples of prominent apps that run on Linux, but don’t share their source code. Does the future point toward more proprietary open-source efforts like this, or will there be a rise in collaborative software like ASP .Net?
Linux is the most popular operating system for large-scale Internet applications, though the software used for these installation is typically a mix of open and closed source. Apache, MySQL, Perl, and PHP are common open-source software systems for these installation. Java and Oracle are closed-source systems that are commonly found on Linux systems, whereas ASP and .NET are specific to Windows. The .NET framework is popular among Windows developers, but I believe that Linux-based systems will continue to dominate the server-side of the Web.
Have we reached the point where it’s plausible or advisable for the average computer user to tinker with source code, or even write his own programs?
No, I don’t think so. Many of the most popular open-source software systems are amazingly large and complicated. I don’t believe we will ever come to a point where the average consumer will tinker with source code or write even the most simple programs or scripts. Open-source systems are evolving towards better ease of use and plug-and-play type of systems, and this evolution will continue to make life easier for the non-engineer consumer.
What about the argument that open source isn’t sustainable in the long run because of inevitable intellectual-property issues?
This is a tough one. although I personally believe that open source is fraught with potential intellectual property problems–mostly due to patent infringement or tainted engineers, not “lifted code.” I also believe that inertia has taken over. Open-source software is part of the Internet infrastructure, and at this point I believe it would be harmful and foolish to attempt to do anything about it.
Mozilla’s Web browser, Firefox, has been downloaded by millions, making it one of the first open-source apps to really make a mark in the mainstream. What are some candidates for the next open-source superstar?
Certainly Firefox is making a splash in the consumer side of the market. Wal-Mart has been selling a Linux-based PC and recently started selling a Linux laptop, though I don’t think it’s making waves in the consumer market. In the non-consumer side of things, Linux/Apache will continue to be the open-source darlings most widely used, though the most active open-source applications in development are BitTorrent and instant message apps.
The most potentially disruptive movement in open source is in the form of applications like OpenOffice.org, a clone of the Microsoft Office suite, and OpenGroupware, which is a clone of Microsoft Exchange. I believe these two applications are the key to converting government and business offices into abandoning Microsoft and using open-source alternatives. Why? Because they are free while Office and Exchange are very expensive, and they have the potential to be more open and friendly to more data formats.
Will Microsoft ever embrace the open-source paradigm?
No, I don’t believe they will ever truly embrace this concept. They have flirted with it for many years now, and much of the software that Microsoft research produces is open source. But ultimately, it’s just not their game.
Simple question: How can money be made from a freely-distributed product?
Simple answer: licensing and support. I believe the key is to give away your software, but just not too much of it. Red Hat is the shining example of open-source success on Wall Street. They have a $2 billion market cap, and last year pulled in $120 million in revenue. Of course, these numbers aren’t very impressive compared to software giants like Symantec, Microsoft, and Oracle. But it’s a healthy, growing business based on open source. That’s exciting.
IBM and Novell are larger, more traditional companies that have embraced open source and are generating revenue. And other smaller companies are on their way, like Sendmail Corp. and MySQL.
Are there going to be new distributions of Linux in 2005 that you feel will be especially notable?
An easy, consumer friendly desktop Linux is the current holy grail of Linux distributions, but I don’t think anything earth-shattering will show up in 2005. Linux will continue with incremental improvement, patch by patch, in 2005.
How are open-source languages intertwining, both with each other and with commercial products?
I believe open-source languages have a distinct advantage over proprietary languages in this regard. Proprietary languages respond to an economic demand for integration with commercial products, but open-source languages like PHP, Perl, and Python only need a single developer, somewhere in the world, with a need and means to do the integration. PHP can access Microsoft COM objects, Perl can interface with Oracle’s proprietary interface using the DBD module, and Python can be be seamlessly embedded into Java applications via Jython.
Does open source have the potential to change the business models of not only software companies, but other businesses as well?
Well, without getting too creative, I think that open source has the potential to lower operating costs for business in general. But I don’t think that it is a model-changing element for companies not in the software or technology business.