The annual summit brings together users and partners of the Durham, N.C.-based commercial Linux vendor to get everyone
up to speed on the state of open source.
With a new emphasis on standards, open source advocates are looking for the respect they feel they deserve for innovation. It all comes down to customer choice, the executives said. At the end of the day, they said, open source products need to satisfy customer demands before their companies can take market share in the commercial market.
In terms of software innovation, is open source driving or stifling it?
Scott Crenshaw: Look around the summit this week -- session innovation in virtual machines, developer tools, new client products. Looks like innovation is outpacing the world of proprietary software development. Because open source taps creativity and engages the best minds, it doesn't matter what country the company is in or which company a developer works for, if any. The open source model is not the most economic model, but it is the most effective one. We have not been approached by other industry leaders asking how they can take advantage of the collaboration development model we have [developed] using open source.
Are you able to talk about short-term goals with the JBoss/Red Hat merger?
Mike Farris: JBoss and Red Hat Enterprise Linux networking have similar motivations. Look at extending end-to-end management all the way down to the operating system level, including making a system scalable and manageable and then taking that entire process and extending it throughout the entire stack.
JBoss has established a popular open source applications server, but it has also built a suite around Java Enterprise Management Server. What's in that?
Shaun Connolly: [JBoss'] focus has been building out a complete open source platform on service-oriented architecture (SOA). Technologies like Hibernate are a very popular foundation technology for that platform.
The point being SOA is a real wave that is definitely happening. It really solves problems and is all about business ability. The price points of proprietary applications in this area start at $50,000 per CPU, so innovation in that area has been reserved for a select few. Our goal is to expand that innovation out to a broad market and really accelerate the adoption.
What do you say to those who argue lower security quality and TCO with open source?
Farris: If you look at the security landscape this year, it is interesting. A recent Goldman Sachs [Group] report mentions that customers are not looking for security products, but secure products. You can look at things like the inclusion of firewalls, and even the open source development model overall, and it really creates a platform where everything is being built in for the age of the Internet -- by the Internet and for the Internet. Some things we've been doing in the open source community relate to creating a platform that involves the underlying computing on the OS.
Our standpoint is that this takes open standards and will eventually result in ubiquitous standards so that they are available all the time. They include things like single sign-on and directory servers, and we are taking steps to make it all a part of everything we do.
In reference to a quote from Microsoft CEO Steve Ballmer about open source software and it being unreliable from a TCO perspective, how can we drive quality through open source development?
Crenshaw: When people ask this, I can't believe Microsoft is talking about reliability of open source software. Look to the FAA or to Orbitz or most of Wall Street and there is no doubt that open source is reliable.
What we must do now is shift from reliability to availability. Until today, highly available has always been in the domain of companies with the most money and the deepest development benches. But if you think about it, highly available is something that fits everybody. Our role is to provide that availability through virtualization, storage and clustering and then bring that availability to the masses. As far as TCO goes, the FAA [Federal Aviation Administration] saved $15 million and two-thirds time by going with Red Hat. I say we let customer success speak instead of the analyst reports.
Speak to the openness of Linux and open source.
Crenshaw: It is not about ownership at all, it's about community. It's about shared goals, shared risks and shared outcomes. It's not about ownership, it's about giving more than you take. People of all ages and backgrounds contribute great ideas to the open source community. Sometimes I ask myself what I have in common with someone in a coastal community halfway around world developing software, and I realize it really is a community. Ask yourself if developers would warrant working 22-hour days if someone else owned the software? No. But this is a core principal of open source, and the people outside the community just don't seem to get that.
Is it really binary choice between the two or is it about balance?
Connolly: It is really a heterogeneous world out there. It is about customer choice. I think the enterprise open source wave led by Red Hat has really spoken to and broken down barriers. It is not really a religious war any more. It is about achieving goals from a customer standpoint like with our Microsoft alliance last year.
The fact of the matter is I am using a variety of technology. It isn't Java versus .NET, or open source versus commercial. It really is about satisfying the customer who is going to be using this stuff. When the Microsoft alliance came to be, the focus was on interoperability, and I've been to Redmond plenty of times to make sure our products are interoperable.