August 2004

I have always thought that Netcraft was a great site and I visit it pretty regularly. I have been reading their articles about Linux usage across the globe and I’m curious that nobody has targeted the Cobalt market.

Cobalt was an early-entry Linux company that was skilled in making appliance-type network boxes for doing all sorts of things like web serving, DNS, DHCP, etc all in one little blue box. They were really great for smaller shops that needed these types of tools in a reliable fashion but without the cost or headaches of having to do it themselves from the ground up. I have worked with these and they are actually quite good pieces of technology.

In 2000, Sun Microsystems acquired them for $2 billion during the height of the dot com craziness. Netcraft recently reported that Sun was discontinuing the Cobalt line. The sad part here is that Sun has not offered up a replacement strategy for the easy-to-use Cobalts.

According to Netcraft as of July 2004 Cobalt still maintains a 20% market share of the Linux server market. That 20% comes in at well over 500,000 machines with only the top 20% of that being by the big name ISP’s. So, somewhere out there are 400,000 Cobalt’s running soon-to-be out-of-date software in need of a replacement. I see two very interesting scenarios here.

1. LiveCD city baby: Most of the Cobalts have CD’s in them. Why couldn’t someone build an easy-to-use LiveCD to either run on or re-install the Cobalts with their OS? Seems like a great opportunity to get some real mindshare as well as machine share for their Linux.

2. Build a better box: With the likes of Pogo Linux and Penquin Computing that specialize in Linux hardware out there, why have they not targeted this market? If you built some custom migration tools and provided pay-for-support for users migrating you just might have a case for getting at some of those 400,000 machines out there. Once you hook them on your products and skills, you can leverage it into more business.

Looking across the Linux landscape even Debian GNU/Linux barely has more machines than there are Cobalts out there. Seems like an excellent opportunity for growth for some lucky Linux distro to me.

I just love my job. I can’t think of a more fun thing to do than to get together with like-minded folks and coordinate on possible projects. Especially when it has to do with open source.

I got a chance to meet with Jay Thompson yesterday from Edge Design about collaborating on a project. Edge has been around for about 20 years doing a mix of graphic design and web work. They have also done some pro-bono work for the likes of the Oregon Food Bank and the Children’s Cancer Association.

Jay has had contacts with several similar groups that are looking for technologies to help enable their organizations to better leverage their volunteers. On a second and third meeting with these different groups, he realized they all really wanted similar variations on the same tool. He immediately thought of open source and set out to find a technology partner to help make that happen.

Many of these non-profit groups have many volunteers and they need to be able to do an assortment of tasks (ideally from the convienience of a web browser). Track time on projects, scheduling of tasks/meetings, portfolio management and reporting. The idea would be to build a modular, open source application that does all of these (essentially an enhanced groupware) that allows users to enable/disable the bits and pieces they want as well as easily develop against the API to quickly extend the application. This is right up our alley.

I am very excited about working on this project. I think it would be an excellent opportunity for the OSL to get some more development work but would also give us a chance to help spearhead the creation of a trully useful open source tool that touches more than just a bunch of geeks (not that I don’t like geeks but they have a lot of fun software already).

Read an excellent article by Brad Wheeler from Indiana University about the possibililities for Open Source in higher education by the year 2007.

This is a great article with excellent talking points about the how’s and where’s of Open Source as well as some thoughts for folks with different perspectives; university administration, foundation, etc. I have been seeing some interesting trends recently and reading Dr. Wheeler’s article really helps put them in perspective.

During my visit to Linux World Expo last week I had a chance to attend a reception held by the OSDL which was given in part to give a chance for the University affiliate members to meet in person. While there I met with Joe Little who is a systems administrator from Stanford. Joe mentioned the fact that a lot of Universities right now are busy trying to capitalize on open source in the sense of each forming labs and trying to create a critical mass at their school. This essentially goes against the nature of open source that it is supposed to be a bit more organic and not so forced. I have to agree completely and unfortunately we’re as much to blame as any other University (see the OSL).

This morning on the univ-linux list Seth Vidal mentioned that it might make sense for the higher education schools to get together:

“Has anyone here ever considered putting together a conference for sysadmins, researchers, etc of linux on college universities and research institutions? Just to get an idea of similar goals, areas of overlap, etc. If something like that were to come about what areas would be of most interest to have talks on? What time of the year would be best?” This is a fantastic idea.

Back to Dr. Wheeler’s article, there is a great quote that comes from a veteren CIO, “With open source software, how do I know that the community cooperation that works today won’t dissipate next year? The Apache Web server has the Apache Software Foundation to ensure its evolution. What is the equivalent in higher education?” Putting this together with the thoughts of Joe Little and Seth Vidal, what do we do?

I am proposing a Higher Education Application Foundation. This foundation would be focused on combining the efforts of higher education across the globe using open source methodologies for sharing and collaboration. I know for one, the OSL would be the first to participate. We need to come together as higher education to help a) learn from each other and b) leverage the tools and experiences we have all learned over the years.

So where do we start?

We finished up Linux World Expo on Thursday and all headed our seperate ways.

I had an excellent time this year. It was good to catch up with everybody, start some new relationships and generally get to know more about the folks that we’re hosting.

I had a few really fruitful conversations that I hope start to really take shape into something more tangible. More on that later.

I would say the overall feeling of LWE this year was that a lot of these companies are left wondering why they were there. Attendence was way up this year (again) with 10,000 and 55 new vendor booths were added for a total of 200. I’m really excited about how big business is embracing open source but I’m left wondering if they know the best way to “give back” outside of letting their people participate.

The OSL just might be the answer. We started this thing sort of on a whim; we had a chance to do some good things for open source projects and we jumped at the chance. We found some good matches in Gentoo and Freenode and have been able to promote the use of open source both here at OSU but also in the community. LWE was really the exclamation point for me; the OSL is all about helping the community help itself. Our focus has been on the community distributions as well as projects that have a community focus. Don’t get me wrong; I have nothing wrong with Red Hat and SuSE. They have great business models and more importantly they are successful at what they do. However, I think there is a big missing piece here.

The community-based distros (Debian, Gentoo and now cAos (CentOS)) are fantastic platforms for users that just want something that works and that they have complete control over. Not only that, these distributions are released with many of the main themes in Free/Open Source Software as their underpinnings. We want to help these projects succeed by providing an economy of scale by co-locating them here at OSU.

There is much more to this than just an economy of scale. There are people and businesses out there clamoring for pay-for-support options for these distributions. How do you provide that and not slight the distributions and developers who work so hard with the right-thing-to-do attitude?

I love Ben and Jerry’s Ice Cream. Call me a fatty, call me un-healthy I just don’t care. The ice cream is dang good. You know what I really like about it? Ben and Jerry and what they have done to build a fantastic company with a conscience. Not only that, but the people that work there are happy to be a part of a team making a difference “just makin’ ice cream”.

Why can’t we do this with open source?

Here is what I am proposing. We start an open source co-op. Here is how it would work:

* Company X wants support for community distro Y.
* Company X pays for support to OS Co-op
* Company X employee calls phone number/gets on IRC and asks for help
* OS Co-op contracts with people in community distro Y to provide support
* OS Co-op takes a fair percentage of the contract and donates it to the OSL
* The OSL and its board of trustees works to determine priorities in the community distros for use of the donations.

Why not have an single group, acting as a community with rules and regulations, help drive some of the community distros? I feel like this could be an extremely successful business model as long as it is done responsibly and soley with the community in mind. Take care of the community and help it continue to flourish.

Linux World Expo 2004 started off with a bang this year with Matthew Szulik giving a rousing keynote concering choice and the world that we live in. I was impressed that a man in his position would wear his beliefs so openly on his sleeve; I would say I was even refreshed by it. The Red Hat theme for this years LWE is choice. Choice for the customer to have flexibility in choosing solutions. He ended his speech with “First they ignore you. Then they laugh at you. Then they fight you.” You could not walk out of their without goosebumps. Then the exposition opened up and we all hit the floor in our custom tie-dye shirts. We were a sight to behold.

We headed over to the .org pavillion to meet with the folks that we do hosting for. This would be your Gentoo, Debian and cAos Linux as well as a few others. It was good to see everybody, hand out some of our shwag and connect in person.

One of our big goals at the OSL is to help out community-based projects that are in need of hosting. So when I visited with Leon Shiman at the X.org booth I was a little taken aback that he saw our efforts as a bit selfish and detrimental to open source as a whole. “Of course you want these projects co-located at your facility; you’ll have better access to grants and $$ because of it”. Let’s just say it didn’t leave me with a good taste in my mouth about the future of X.org.

I also spent a fair amount of the day talking with folks in the Gentoo community. I’m really curious what makes that beast tick and so I cornered everybody I could find. More on that later (most likely the topic of a future talk).

The vendor booths were just insane; big booths, booth babes, shwag, the whole nine yards. In talking with folks at the conference it almost seems like the vendors are restless; like they aren’t sure what they are at this conference for. It is, afterall, a pretty tough sell. How the heck do you sell something that is free? (I know, I know … its not really free).

Spent the afternoon walking the floor and talking with folks. At 5:30pm we headed over to the OSDL reception at the W hotel. It was a quite nice and got a chance to catch up with folks from Intel, Red Hat and IBM as well as the OSDL folks.

At 6:30pm Alex and I were off to the Gentoo community meeting at the convention center. I took a few minutes to talk to the group (about 30 or so people) and tell them how excited we were with their success and handed out t-shirts and our custom leathermans. Quite fun and good to connect with everybody. To me, these are the folks in the trenches doing the brunt of the open source work.

Finally, we ended up the evening going to “Gaylords” which was a fantastic Indian restaurant. We hooked up with the Gentoo folks so there were like 14 or so of at this restaurant. Very fun. Then off to bed as we have another big day ahead of us.

Ever since Google launched Google News I have been in heaven. I can get my news quickly and easily as well as use their fantastic search technologies to drill down where I care most when I see something of interest.

When the Brooke Wilberger abduction happened, we all here in Corvallis were pretty interested in finding out as much as we could when we could. Enter Google News and their notifications. I could be notified when new stories hit or updated information becomes available. It didn’t matter where or who, it was all about getting me what I wanted when I wanted it.

Microsoft announced that it was launching its own news aggregator service in hopes of stealing some of Google’s thunder. In the Washington Post today there is an article about how the news site gives favorable placement to articles from Microsoft’s own MSNBC.com news site. Here we see again Microsoft thinking like a big company that has to leverage its position with one service to help another. That’s all well and good but I predict the service will flop. Why do I think that? Value.

The value of the Google news service is obvious; it aggregates news with no bias. As a reader of news, the last thing I want is bias. Knowing there is no bias means I’m more likely to use the service (and more importantly shun Microsoft’s). The value created is two-fold; not only do I get value as a consumer but Google gets my eyeballs to advertise to. To me, that’s a small price to pay for unbiased reporting.

Microsoft needs to learn that value moving forward for them will be created in the communities they can foster and help. Not in how much traffic they can push to each of their sisters companies.