Making an Open Source Project Work: The 12-Step Program

Open source changes everything. That's what the pundits will tell you. What they mean is that open source isn't just about free or better software. It's a whole new way of thinking about how to solve problems and manage projects. When we launched the project that I'm working on -- the Linux Professional Institute -- a year ago, we knew that we would be able to leverage this open source methodology to make the project stronger and better, even though we weren't developing software.

Trenches/ L P I Blocks

Open source changes everything. That’s what the pundits will tell you. What they mean is that open source isn’t just about free or better software. It’s a whole new way of thinking about how to solve problems and manage projects. When we launched the project that I’m working on — the Linux Professional Institute — a year ago, we knew that we would be able to leverage this open source methodology to make the project stronger and better, even though we weren’t developing software.

To step back, the Linux Professional Institute (LPI) has been around since the fall of 1998. Its goal is to develop a certification program for Linux professionals similar to those available to Microsoft (MCSE) and Novell (CNE) users. We have developed a series of exams that people can take to demonstrate their knowledge of and proficiency with Linux.

Of course Linux is different from NT or NetWare. And one of the challenges we face at the LPI is that no one entity controls the Linux operating system, so no one can dictate the standards for certification. Anyone can create their own Linux certification programs — in fact several have. In this environment, our efforts have been focused on building a consensus and making sure that we are as inclusive as possible — something that the open source model is ideally suited to achieve.

In our first year of operation, there have been a lot of lessons learned. So, without further ado, here are the 12 most important:

1. Democracy is Critical and Inefficient. Most of the LPI’s major decisions have been made through public mailing lists where anyone can participate. They have been “consensus points” that were hammered out in (sometimes lively) discussions on the lists. This process has allowed many different viewpoints to be recognized and debated. The end result has been a much stronger and better-defined product. This democratic process can drag on, however. It has sometimes taken great effort to ensure that discussions stay on track and do eventually reach a conclusion.

2. Living in a Fishbowl Can Be Challenging. What do you do when all your major discussions occur in a public forum, with competitors and members of the media watching your every move? On the one hand, it’s important to be as open as possible. On the other hand, we have been very aware that anything posted to our public lists can very easily turn up in the media — or be used by competing programs. Day by day, this remains one of our toughest challenges.

3. Building Coalitions is Critical. Early on, we recognized that our initial group of volunteers didn’t have all the expertise necessary to truly pull off a world-class certification program. So we built an advisory council of other companies and individuals who could help move us forward. This council has put us in touch with organizations that could help us, and provided the public backing for LPI that has attracted others to our project. This public support also gave us the legitimacy we needed to go out and attract financial investment.

4. Be Inclusive. We use public mailing lists that are open to everyone because we want to hear what others have to say and also because we need the help of many people to make this project a success. We want to see arguments against some part of our project discussed, debated, and resolved before the project goes forward and arguments flare up in the larger public media. We also have sought to include as many different viewpoints and people so that others would not go off to create yet another competing project. Finally, by being inclusive, we have fostered a community of volunteerism. Our project will not succeed without the many long hours put in by volunteers, and the forum where they participate is the public mailing list.

5. Recruiting New Volunteers is Crucial. The fact of the matter is that volunteers come and go. People’s lives change, and sometimes they find themselves with less time to volunteer on projects. People burn out too. At LPI we have learned to constantly grow our pool of volunteers — we advertise and promote ways that people can become involved and are constantly working to identify and recruit new people to bring new energy to the project.

6. Give Credit Where Credit is Due. Part of recruiting and retaining volunteers is to make sure we publicly recognize the many contributions that have been made to our project. We have worked to make sure that special Web pages, articles in magazines, and our online newsletters all recognize the people who help out in some way.

7. Delegate. Once volunteers are found, it is extremely important to give them real tasks to do. Being caught up in the management of the project, it is easy for us to just do something ourselves because doing things ourselves often takes less time than explaining them to others. Everyone is guilty of this. But sometimes taking the quick and easy way means losing an opportunity to get someone else involved. And all those quick and easy tasks can add up too, and without delegation you can soon find yourself in a state of complete and utter overwork.

8. Accountability is a Must. With every action delegated or assigned, there must be deadlines and also someone designated to make sure the deadlines are met. We’re all volunteers and it’s easy for things to fall through the cracks. Within the LPI board, we’ve implemented a list of weekly action items that outline what each of us has committed to, and each week we are asked the status on each item. We’ve also worked to make sure there are committee chairpersons responsible for the actions of a committee.

9. Do Not Rush in Response to Market Pressure. Market pressure, of course, is very real. Our project has been in a race to get our program to market as quickly as possible. There is great pressure for us to push our program out as soon as possible, but the challenge has been to balance that pressure against the need for a top-quality program. We will live and die by the quality of our certification exams and how they are perceived by people within the Linux and larger information-technology worlds. We have focused on making sure our program is of the highest quality and have made a conscious decision to deliver it when it is right to do so.

10. Don’t Ignore Market Pressure. Having said that market pressure cannot be the end-all and be-all of your project, what good would it be if we delivered a high-quality program after the Linux and larger IT markets had already crowned someone else the Linux certification program? Market pressures count. We have had to set a timeline for the project that reflects this.

11. Communicate, Communicate, Communicate. If there is one thing we have learned above all, it is that our project must be in constant communication with the larger world. There must be news releases to the media, updates on the Web site, newsletters sent out in e-mail, and meetings with sponsors and supporters. Every little piece helps. We have spoken at conferences, written for magazines, and scheduled BOF (Birds of a Feather) sessions at conferences. The key has been to let people know that we exist, what we are doing, and how they can become involved. Without big-league resources like PR firms, we have to work doubly hard to make sure that our message gets out there. Our supporters and sponsors (from the coalitions we have built) have also been of great assistance in spreading the word.

12. Keep the Faith. It’s a challenge at times to remain solid in our belief that the project will become reality. Faced with inevitable roadblocks and challenges, and with naysayers claiming we could never do it, the struggle has been to remain positive and moving forward. We have learned to rely on one another to get through worrying times, and to continue to project this positive image publicly to motivate and inspire others to help us.

It has been a crazy first year for the LPI. We have accomplished an amazing amount in a short time period, and could not have done it without the many volunteers and supporters who have helped us along the way. We have encountered and triumphed over many challenges and learned a great many lessons along the way. In the end, probably the greatest thing we learned was that you must gather as many supporters as possible and move as quickly as possible to bring your project to reality. And in the end, you must simply believe in your project.

Dan York is the chair of the board of directors of the Linux Professional Institute (http://www.lpi.org), a community project to develop professional certification for Linux. He can be reached at dan@lpi.org.

Comments are closed.