November 2005

On November 16th, 2005, Peter Caday wrote a letter to the Gazette Times editor (scroll to the bottom of the page) saying that funding work on open source software by Oregon State University in the form of the Open Source Lab was a misstep.

I think the response (middle of page) that Conrad Shultz gave was a great one.

As much as I’d like to wade into this discussion I’m a fan of the adage that “seeing is believe”. Peter, if you’re out there and reading this, drop me a line at scott@osuosl.org (spam bots begone!) or give me a call at 541-737-9922. I’d love to give you a tour of the Open Source Lab and hopefully dispell some of the concerns and frustrations that you voiced in your letter. IMHO we’re doing some amazing work here with the most fantastic students the world has ever seen. In any case, I’d love to share with you, or anybody else for that matter, what we’re up to here and why we believe this is such a great thing for OSU and Oregon as a whole.

I got another email a week ago asking some more questions about open source, etc. Instead of penning a response and sending it directly to the person asking, I thought I would blog it and respond that way (in case anybody else cares):

“Hello Scott, My name is ***** ***** and I am a student at Clackamas Community College. I am writing a term paper and the topic I have selected is “The future of open source technology”. I have a few questions in which it would be much appreciated if you have the time to answer.”

1) What do you think the future holds for open source technologies?

The biggest thing I think we’re going to see in the (near) future are tools that are geared for collaboration between non-geeks. For so long we’ve been seeing technology for the sake of technology; geeks writing code to help them write more code. However, as that software developed, it started moving up the social stack. Instead of just email, diff and patch, we now have truly collaborative tools like wiki’s, instant messaging (yes, this covers a lot of ground), blogs, etc that help us disseminate, aggregate and most importantly collaborate.

I see the future bringing these tools to the masses. If I were a novice developer or a lawyer or a doctor, its rather daunting to collaborate. The tools for them just aren’t there yet. However, as these software development technologies mature, they get easier to put in front of joe-shmoe user. Once you can collaborate on documents, talk with the people that you want (either via messaging, voice or video) then we’ll really start to see things take off. The possibilities will be endless.

2) Do you perceive it as “the wave of the future”?

I don’t really view open source as the wave of the future; more like an echo from the past. All we’re seeing with open source is the commodization of an industry. As time goes on, processes and practices are developed to drive the cost of anything down. The fact is, margins on software have always been high, sometimes into the 95% range. Tim Witham from the OSDL said it best, “we’re entering the age of software by the ton”.

Now, if you say “wave of the future” as in we’ll see more collaboration on a global scale, then yes, I totally agree with that. Tools that enable communities to form across the Internet will continue to grow as those tools emerge and then mature. It’s inevitable.

3) What is the current niche of open source operating systems?

I think what this question is asking is where are open source operating systems doing the best right now?

Open source operating systems (and by this I can only assume Linux and some of the BSD’s) are starting to take hold in just about every facet of the marketplace. From servers to desktops down to your cell phone Linux is gaining ground at a phenomenal rate.

Adoption of Linux on the desktop is taking far longer in places with lots of deployed Windows infrastructure. Places like North America and parts of Europe are pretty heavily invested in Windows and the barrier to switch is simply too much (and rightfully so). However, the emerging markets are another story. If you’re starting from square one with your computing infrastructure, something like open source and Linux make perfect sense. The ability to customize these applications for specific locales also helps quite a bit; some companies simply cannot afford to develop localized versions of their software so those regions have to “make do” with what’s available.

On the server side, Linux is quickly gaining ground. Linux is a great play for companies that have deployed VMS or Unix systems already. It is also an excellent choice for organizations that starting from ground zero; if I start a company that needs lots of computing cycles I’m most likely going to leverage Linux just to save on the licensing costs.

The embedded market is also starting to gain ground. Vendors can customize Linux to however they want giving them complete control not only of their platforms but of their own destiny. No upstream vendor is going to change direction or plans on you if you’re using Linux at the core.

4) What are the advantages and disadvantages of open source operating systems at present?

Whoa. Man, I could go on for hours about this but I’ll try to keep it short and sweet:

Pros: Control of destiny, licensing cost savings, ability to customize and greater security (yes, that’s a can of worms but I really believe it).

Cons: Questions about intellectual property rights around open source, the barrier to entry can be difficult, lack of talented developers/users/consumers around open source and little or no marketing.

Like I said, I could elaborate on all of those but that’s what comes to mind for me off the top of my head.

I’m often asked questions about open source via email from people all over the world. They are asking for my opinion but mostly because they don’t know where to start. If you’re new to open source and just looking into either joining or creating a community it can be rather daunting.

I met with Derek Abrams and he is asking these various questions. He distilled them down to two of the most important questions he wanted answered and I’m going to do my best to answer those here:

1) What are the differences in the license offerings?

Three words: can of worms. As of this writing, there are 58 OSI approved licenses that one can use to release their works as “open source”. The sheer volume of licenses makes even the most liberal lawyer (is that an oxymoron?) nervous let alone a person just entering into the open source space.

Most of the time, these open source licenses are much alike. Sometimes it can be as little as one or two sentences that differentiate them. In the case of public agencies, the use of the GPL might be impossible because they can’t indemnify themselves.

My recommendation to people on this subject is to stick with what works and is generally accepted: the GPL. About 70% of “open source” software is currently licensed under the GPL v2.0. Just because everybody else is doing it isn’t the only reason. To be able to point to so many companies and projects that are using this license it helps to put newcomers at ease (especially said newcomers legal department).

The only downer about the GPL is the impending release of v3.0 of the license. Some are saying their may be the “web services” clause included in this version. If that happens, there could be trouble for companies as this “viral” license may “infect” other parts of the organization and “force” the open sourcing of normally internal-only software. Will this happen? That’s what everybody is wondering. However, with an installed base of 70% I feel pretty confident that nothing too drastic is going to happen with v3.0.

My advice? Don’t try to make your own license; what you’re doing is so specific to your organization that you can’t leverage an existing (and hopefully really common) open source license. Go with the GPL.

2) What is the anatomy of an Open Source project?

This is a great question, IMHO.

First and foremost, you need to have buy-in from the players involved. If you have people who are even mildly tepid on the idea of doing an open source project it most likely will not be successful. Buy-in from your hosting organization (say your public institution or company) is critical. It never hurts to run these things up the flag pole. The people you work for never like to be surprised.

Choose the GPL or another well-known open source project if you’re starting from scratch. See 1) above.

At the core of an open source project are the things that make it easier to collaborate; the tools. Wiki’s, forums, mailing lists, IRC, etc are the plumbing required to execute any distributed development. These are almost to the point of being a commodity although some projects still have trouble providing these services. For most open source projects, even deploying these services can detract so much from the projects “main thing” so as to severely hobble a projects progress.

When in doubt, collaborate first. I’m sure the idea you have come up with is a great one. But searching on the Internet will most likely reveal a similar project. Find ways to collaborate and give it a little more than the ole’ College try. Any issues you might have with an existing project will most likely surface with your project down the road. How will you be prepared to collaborate with other people when they come knocking?

Leadership within an open source project is critical. Some people call this the “benevolent dictator” and a great example of this is Linus Torvalds or Jimmy Wales (from Wikipedia). Both Linus and Jimmy are very good at working with the community, delegating and dealing with grievances in an open/transparent fashion. The fact is, if you’re a jerk in the community, you’re going to fail as a leader. Understanding that you’re not always right is a hard lesson (albeit extremely valuable).

You could do this by committee. Some projects do this. However, when conflict comes up or a hard decision needs to be made (and this does happen) you get nothing but death by committee.

Open, honest and transparent communication is very important. Bejamin Mako Hill has written a really good article on why you have to be transparent. This is a great read for anybody considering starting their own project from scratch and looking at either volunteer or paid developers to help make it happen.

The license, the buy-in, tools, transparency and leadership are all about building a sustainable community around your project. The goal of any project leader should be that they can walk away and have the project stand on its own. This is the mark of a trully vibrant and effective community and project and every project leader should strive for this IMHO.

The OSL is participating in a community source FIS system for higher education. The idea of community source is that you collaborate with institutions on the rapid development of some application by committing resources and holding people accountable. This is a great formula for shortening the transaction time for open source. However, community source is the subject of an entirely different posting …