Disclosure is what drives innovation. Free software is the ultimate form of disclosure.

Free software is the scaffolding for innovation as well. When people were rushing to build out Web sites in the 1990s, the fastest way to do it was on Linux and 386. Today, the equivalent is free software in the cloud.

How can you drive innovation even faster? First, architect for it. Make your platform extensible and embeddable. If you have a cool idea to change the web, you can write an entire browser but that’s not likely to succeed. You can license Internet Explorer, which would be a painful conversation. Or you can write an extension for Firefox.

Platform tolerance: It’s critical to work with Windows. Free software that works with Linux and Windows is critical because it offers inclusion to those who continue on with Windows.

Fresh blood: The real innovation comes from people around the core of a project — rather than in the core of an open source project. So you need tools to make sure people can get in, innovate on your platform, then move on easily.

The purpose of methodologies is to harness talent.

Architect for collaboration and contribution across community

Don’t create environments that demand permission. Example: Nobody should have to ask for permission and participate in the project.

Introduce a rhythm or cadence to your projects. Have a pulse across open source, where multiple projects move forward together.

Success requires innovation on the business and technology side.

The demand side is obvious for free software

But the supply side is trickier: Who pays for the innovation?

We’re already better than the proprietary world, in that we have multiple suppliers.

But we need to do better. We need an ecosystem where every component is sustainable.

He doesn’t see how advertising will fund web-based apps or free software apps over the long haul.

The emphasis on online services could justify the continued investment in free software.

Can we not emulate, but blow past Apple, in terms of delivering a user experience?

Linux desktop has to be art.

It’s not a mad crusade. We need beautiful, free software that is art.

To survive in a Web world, you need something that’s attractive.

It has to be profoundly usable.

That’s the challenge for the Linux desktop for the next 2 years.