April 2005

Mozilla’s external infrastructure has been running on a little more than a shoestring with some of the most heroic efforts by the sysadmins involved. After leaving AOL the Mozilla Foundation essentially had to start from scratch and be able to build an infrastructure capable of handling millions of users a day. It has taken some time but we’re getting ready to deploy the next generation of this infrastructure. The new infrastructure will be much faster, more scalable and redundant. Check out the new gear.

I read an article about Microsoft migrating to 64-bit Windows. My favorite quote was “”the entire Microsoft.com site has been migrated, and we serve 30 million unique visitors every day.” I would absolutely love to know how much gear and most importantly the total cost for doing that for them.

The Mozilla infrastructure easily sees 30 million clients a day across all of its properties and it does it with (although I am in fact just guessing this) much less gear than Microsoft does. Too bad they don’t use Linux on cheap gear. And before anybody flames me, I’m sure I’m comparing apples and oranges here people … this is just Friday morning fodder …

In the wake of several counter updates I have jumped on the bandwagon as well. What the heck. At least there isn’t an RSS feed here … haha …

I took all of the data from late November 9th, 2004 to today from Mozilla’s Bouncer and have created a running graph that now, moving forward, will leverage the download counter on SFX.com for updates.

Enjoy.

We get to have a fascinating view into many different communities here at the OSL. Every community has its bits that make it amazing and at the same time dysfunctional. That said, every community we are lucky enough to host here is always moving forward.

One of the most interesting differences in many of the communities here is the ability to leverage “all comers” to a project. Gentoo has been particularly good at this as has Mozilla. There is something about being able to not only make use of the people streaming into your project but be able to get them to contribute that is critical to the growth of a project.

This comes back to “The Cathedral and the Bazaar” and ESR’s musings on how Linus has been so successful over the years. I really believe that you have to delegate/trust first and ask questions later. I’m not saying you blindly accept patches from *@microsoft.com into the kernel. I’m saying go with your gut, trust people and enable them to make the project better. Linus has said on several occasions he does simply because he is lazy; maybe we should all be so lazy.

One of the most fascinating projects to date I have seen is SpreadFirefox.com that the Mozilla Foundation sponsors and we host. Here we have a community that has formed by people that are excited about using and spreading the word about Firefox. The MoFo loves the project because it is a fantastic way for people excited about Firefox but not developers to be able to contribute to the project. In addition, it is a great way for people to get involved with Mozilla while being able to build their own personal brand. They get ownership because they aren’t necessarily doing some dirty work handed down to them from someone up-on-high. You can direct the community but you simply have to embrace it to succeed. Trying to stifle it by imposing your “vision” for a project will only make it fail.

We have been working with some folks from the OpenOffice project about hosting an instance of www.spreadopenoffice.org. For OO, I see this as a great way to really spread the word about OpenOffice and ramp up for v2.0 to hit the streets. Looking at OO’s value, its even greater than Firefox in the sense that you’re talking about being able to save tangible dollars in an organization with the replacement of competing products.

Discussions have been slow going. Many of the folks in the OpenOffice community see it as splintering their current marketing efforts and possibly taking away from the project as a whole. I view the possibility of SOO.org as a great way to spread the word about what a great product OpenOffice is, to channel the efforts of those who can’t develop and generate a butt-load of press for a project that greatly deserves it.

I am often asked why the OSL does what it does and what it’s main purpose is. When I tell them “we’re here to promote and develop communities around open source projects” they just sort of look at me sideways and say, “seriously, what are you guys in this for?”

Of all the places an open source project could land, IMHO, the best is a University. Universities are here specifically to create knowledge and teach (in that order). There is another component that is often over looked and I believe that is “to build communities” as well. Every successful University is very good at building not only its alumni community but also attracting other successful communities; microbiologists, physicists, etc. In this same vein, the OSL hopes to reach out open source communities in an effort not only to help augment those communities but in the hopes of building the “community of communities” that will help each community get so much further than they would alone.

Yes, we are here to help open source communities. In addition, we want to provide great learning opportunities for our students. One of our students is even going to be doing an internship at the Mozilla Foundation this summer. That’s amazing. Could you imagine being 19 years old again and getting to work for arguably the hottest open source project on the planet? You’d nearly kill for that opportunity.

Let’s face it, I don’t know many people that participate in just one open source project. People are often associated with two or more projects at a time; some of them even compliment each other. By bringing the most talented, energetic and passionate projects here to the OSL, we hope to continue this tradition by creating an environment where these communities can interact and cross-polinate. The “best of breed” ideas will surface and help change the face of the open source world.

Finally, we at the OSL seek to shorten the “open source transaction time” … get innovation from the developer out to the end-user as quickly as possibly. Shortening that time-to-market with ideas and innovation and being able to get it out quickly and reliably is another reason we exist.

For those of you that don’t believe we are doing this for, and in the name of the community, I’m sorry. I’m sorry that you cannot believe that someone understands the importance of open source as you do and more importantly is willing to find resources and dedicate them to these heroic volunteer efforts all for the good of everyone involved.

We re-launched UMO tonight and made it with about 4 minutes to spare with our original deadline (technically we had 4 more hours but we were basing this on PST).

I have sent out a note to developers of extensions telling them that they may need to update. This will also serve as a great way to get people back working on the site agian. Very exciting times.

We still have a long road ahead but I for one am excited we are back on track. Big props to Mike Morgan, Paul Querna, Alan Star, Colin, Mao and the whole UMO team for coming together to knock this out.

Looking forward to it!

I am a Computer Science graduate of Oregon State University. I was lucky to get to go here and am luckier still to have a job promoting open source for the sake of the University and the world.

It was with great disbelief that a I recently learned the that the Computer Science program here at OSU was going to really focus on enhancing its research efforts by cutting the number of instructors they have and hiring more research, tenure track professors for the department.

The department has chosen a strategy that is as follows (bear in mind I’m hearing this second hand from one of the professors in that department so I’m sure some of the details I’m going to lay out are not exact): schools such as the SDSC and MIT have really been focusing on research in Computer Science for the past 40 years. Now they are the premier places for that research to occur; the faculty on staff, those they can attract and the prestige they get in the computer science community has grown immensely over the last 20 years. So, as OSU, we’re going to adopt the same strategy in hopes that in a few years (that’s read 20 years) we’ll be a world leader in computer science research. Prestige, money and people will then follow.

This strategy is flawed in several ways. For one, we’re starting way, way behind the curve; OSU is not even on the radar when it comes to research in this area. In addition, Portland State now has a huge advantage in terms of total faculty over OSU (19 to 35). This means research $$ and prestige will more likely land at PSU and not OSU in the short-term and more likely in the long-term as well.

Now, don’t get me wrong; I love my alma mater and more importantly, I want to see it do well. But decisions like this are not only frustrating they just don’t make any sense to me. If there is one thing I hate more than somebody who complains, its someone who does not offer a possible solution to the problem. So where do I think OSU’s Computer Science Department should focus its energies? I’ll give it to you in three words: take a risk.

40 years ago when these now-leaders-in-computer-science headed down that path, they were in essence taking a chance. They were taking a chance that computers were forever going to change the world they lived in. Fortunately for them, it paid off. The industry-of-research that has grown up around computer science is maturing; OSU is not even in the fold and to try to break into it would be like me trying to build a competitor to the iPod. It would be foolish because I’m too far behind and more importantly, the market is dominated by several well-known, and well-respected competitors.

The next “big thing” in Oregon (and the world for that matter) is open source. With the OSDL, the OTBC, IBM’s Linux Technology Center and now the Open Source Lab in Oregon already, there is no higher-education research, teaching and learning happening anywhere in the state. This is an opportunity. But how do we exploit it?

The Computer Science Department has a raft of interest from students that are wanting to be involved in and learning about open source. These students are motivated and brilliant; they are going to change the world and they want to use open source as their tool to make it happen. If only we had some faculty and more importantly full-time instructors to teach and interact with students on open source, we could attract some of the most talented students in the Northwest and the world.

I will be the first one to agree that the job of the University is to create knowledge; to that I’m not saying we need to stop doing research. However, times have changed. Research dollars are drying up and we need a new way to bring in money to the department.

If you’re going to take on a strategy that will take 20 years, try this one; focus on your students, connect them to the world of open source and then keep that connection back to your department when they move on. When they have changed the world they will remember that the seed was planted at OSU and they will give back.

We have the world at our finger-tips and all we have to do is execute.