Reflections on the Evolution of GOSCON and Open Government

As part of our run-up to GOSCON, we're bringing you interviews with our speakers and members of our conference committee. These interviews provide a taste of what attendees will hear and learn at the conference, as well as insights into the evolution of the Open Source in Government movement.

For our first interview, we've talked with Andy Stein, long-time member of the GOSCON Conference Committee. Mr. Stein is the Director of Information Technology at the City of Newport News, Virginia. Using Open Source as a best practice model for collaborative software development, the City of Newport News is engaged in shared projects with other localities. The City of Newport News has developed a strategy to replace legacy applications through a collaborative ecosystem with public entities and via public-private partnerships.

Mr. Stein, you’ve been a member of the GOSCON conference steering committee since 2005. Can you tell us about how you initially got involved with the conference and your thoughts on its evolution in the past five years?

It was GOCC, the Government Open Code Collaborative, where Deb [Bryant, GOSCON Conference Director] and I met in 2004. This was our initial attempt at the state and local level to share software through collaborative development. Our intent was to build a software repository similar to SourceForge or FreshMeat, but specific to government, where contributions could be managed and controlled. When Deb shared with me her intent to start a conference on this topic, I gladly participated. The difficulties with getting GOCC off the ground showed that more conversations had to take place before we found a system of sharing that is practical for governments.

GOSCON evolved over five years in step with industry and is a reflection of what is currently being done in government. The first two or three years were dominated by collaborative software projects similar to Sakai and Kuali. We wanted to learn from them and copy their successful models. Year three and four were influenced by the use of Open Source in the Department of Defense and by the two visionaries Vivek Kundra (then CTO of Washington, D.C.) and Aneesh Chopra (then Virginia Secretary of Technology).

This year, we have representatives from Civic Commons, Open311 and DataSF at the state and local level, or data.gov, apps.gov, forge.mil at the Federal level. The evolution continues and the number of players increases, as does their size and scope. We need large organizations with resources to be involved as active participants and contributors.

The Plone CMS-based E-Government platform you developed for the City of Newport News, Va., was widely celebrated and earned the 2008 Digital Government Achievement Award in the Government to Government category. Can you tell us more about the history of that project?

Open eGov has been in productive use since the spring of 2007. It is still in use by the City of Newport News, as well as Franklin County, Virginia. It has been available as a free download under an Open Source license from our city website or as a hosted service for Virginia localities.

It was our proof of concept for larger collaborative projects we were hoping to build in order to share expenses and get the advantages of semi-custom development. We knew it was difficult to share the cost of new software development. We did not know how open governments are to collaborating and reusing each other’s intellectual property and did not know how much interest there was in reusing existing software that is proven and stable. The community we envisioned would use that software base and layer additional functionality on it that would benefit all while sharing the costs. It seemed like an attractive model. If it had worked, it could have been extended to other applications of mutual benefit and greater value. This was the intent.

But we could not get the community to grow beyond two, although we provided the software to Franklin County as a service at a very affordable cost-sharing price and without expecting them to develop technical skills. All technical administration and support was provided by Newport News, including hosting the website at a rate of $260 per month. We offered a similar arrangement to other local governments in Virginia.

Related government communities have been formed on the Plone technology, but they are independent and largely uncoordinated. You can take a look at a list by country at http://plonegov.org/references/view. While several attempts have been made, this larger community does not share much and does not leverage each others' assets.  The reason seems to be language barriers and lack of administrative support. Members did not have the will, motivation or resources to develop an effective method of sharing. Newport News and the Belgium sub-communities met twice, one time at each location. We even organized a "Sprint”: one week dedicated to developing new software while together at the same location.

Open eGov was built on top of Plone 2.5. [The] current version of Plone is 4. Migrating to the current version would be expensive, since we would have to cover the entire bill. The other factor is that SharePoint is now available from Microsoft; the cost is affordable, and as a technology it is more strategic to us as a foundational software product

However, other Plone-based e-Gov solutions have met with great success. One such project and community is Plinkit, for public libraries. They are better organized, have a funding source and are active in six states.

What lessons learned can you share about collaboration within the government IT community? What would you do differently now?

Don’t assume that 'they' will come if you build it. Collaboration in government is not very effective as an ad-hoc, spontaneous activity. It requires structure, administration, organization, planning, coordination, common purpose, peer relationships of trust, etc. The community must gel first before the code repository will be of use. Otherwise, sharing will likely not occur.

You were tapped by Aneesh Chopra as his expert in collaboration before he became the United States’ first CTO. What insights into the collaborative process can you share?

Aneesh had I had a common purpose: to learn from the experience of other successful communities and duplicate their method of sharing and collaboration. Aneesh had very large projects like ERP and Medicaid, each estimated at the $100 million dollar implementation cost. Newport News had ERP and taxation, each estimated at about $6 million dollars - different scale, but similar problems. We concluded that collaboration is difficult but [has]* real opportunities for significant cost savings.

One of the common factors in all successful attempts to collaborate was the objective third-party facilitator and coordinator [which is] typically an NGO, non-profit organization with a funding source focused on managing the administrative tasks of sharing, including the project, when applicable. The independent organizational entity is stable and constant as members join and leave the community, focuses on serving the members and facilitating their successful leveraging of common assets and resources.

Where do you see open government going next? Any particular trends or hot topics that have piqued your interest of late?

The time may be right for open technologies to be served to government as cloud services. Increasingly more Open Source consultants can focus on providing cloud services within their area of expertise. NGOs could be formed by retired government employees (as an example) to provide the same. An “Open Cloud Provider”, whether for-profit or not, could leverage open source technologies and the associated technical community to provide generic services of interest to users. For governments, it could be an Open eGov type service provided in a similar manner as Newport News is hosting the website for Franklin County. It is similar to a hosted application under the previous “Application Service Provider,” except that joining or leaving the cloud could be as simple as an online registration and form of payment.

Plinkit, described above, is this type of model. It targets rural, small public libraries. The assumption is that they don’t have any budget for a web presence. Plinkit contracts with the state, which subsidizes the service for their libraries at a nominal fee, or some may charge back the libraries. There are currently over 300 libraries using the Plinkit service, sharing the same technology. Most likely they don’t know and don’t care that their website is powered by Open Source. All they know is that they have available, scalable and feature-rich technology to serve the needs of their citizens, which could have not otherwise been possible if they had to implement it on their own.

With Open Source as a back-end, one level removed from governments and infrastructure type technology while often providing citizen-facing services, it has better chances for adoption. It is no longer governments collaborating at the technology level, but private sector and NGOs collaborating and leveraging the Open Source communities.

Many thanks to Mr. Stein for sharing his time and thoughts with us. For our next interview, we'll be bringing you voices from the City of Portland.