Open Source in Government: Newport News, Va.by Tom Adelstein
What if you ran a city government and had to upgrade your infrastructure, productivity applications, comply with new homeland security standards, and do it in the midst of a budget crunch. If that sounds familiar, you're not the only one facing this dilemma.
In fact, the majority of municipalities in the United States have budget problems. They also urgently need to upgrade their internal information systems while providing Internet eGovernment services such as online document filings, driver's license renewals, school registration, and tax assessments and payments. The costs just continue to rise.
A recent study by a city of 200,000 residents concluded that a computer upgrade would cost $30 million over a three-year period. Multiply that by as many as 20,000 cities and the hit to the economy starts looking significant. As we know, the only way to pay for such an upgrade involves increases in taxes, levies, and bonds. In our system, leaving those funds in the hands of the citizens has a better economic effect than trying to shove them through a bureaucracy.
Of course, a city could hop along, patching things here and there until the costs of repairing and maintaining such an infrastructure would exceed that of replacing it. Then the city would continue with its current computer-related problems and would have spent the same amount of money anyway.
|Linux/Unix System Administration Certification -- Would you like to polish your system administration skills online and receive credit from the University of Illinois? Learn how to administer Linux/Unix systems and gain real experience with a root access account. The four-course series covers the Unix file system, networking, Unix services, and scripting. It's all at the O'Reilly Learning Lab.|
Do any other alternatives exist? Consider Andy Stein's vision of a collaborative development effort by cities throughout the United States and possibly elsewhere. A small standards group called the Open Government Interoperability Project seized Andy's vision and began work in the third quarter of 2003 to make a government collaborative development alliance a reality.
Most of us have recently seen an increase in news reports about open source software showing up in governments. Most of those reports seem to dramatize an upcoming battle between Linux and Microsoft. Rarely do we see information on the pros and cons of collaborative software development.
You may also have noticed a lack of discussions from people in the trenches actually running the information systems for governments. In this report, I caught up with Andy Stein, the CIO of Newport News, Va., whose dream of using the open source software model to create a repository of software for local governments makes much common sense.
In this interview, Andy provides us with insights to what problems local governments face and how they can deploy open source components in the enterprise. I think you'll find he uses sound logic in his approach and has the credibility to influence others to look at the alternatives.
Tom Adelstein: You're currently director of IT at the City of Newport News, Va. Previously you worked for IBM and Capital One. At IBM you had a major role in developing the RISC processor and at Capital One you were the first IT person hired to build that organization. How did you wind up in city government?
Andy Stein: In the private sector, individual accomplishments become "intellectual property" or something called "competitive advantage." After my first interview for a local government position, I saw the potential for "large scale collaboration." That provides quite an incentive. There's only so much one can develop and accomplish working alone. In my case, I find satisfaction in leveraging the knowledge and capability of many while driving toward common goals. In local government, I will accomplish in 2 years more than I could in 25 years in the private sector. That's exciting to me.
TA: Andy, what led you to consider open source software and where do you see it fitting in your city government? Is it an infrastructure play, ERP, eGovernment, or what?
AS: Our researchers came up with open source as a "Best Practice in Software" collaboration and I looked into their findings on that basis.
Our challenge in Newport News involves replacing 80% of our legacy applications. These are mainframe applications written over the last 25 years in a language and utility that has been stable for about 10 years and no longer meets our needs. We want to improve services to our citizens and make it convenient for our citizens and businesses to interact with the city. We want to integrate and modernize our municipal systems, and provide a solid technology foundation that moves us toward eGovernment and allows us to provide modern services online. We also want to integrate well with other municipalities and the state.
As we looked through the list of 155 applications we need to replace, the inevitable questions arose: why are we doing this alone? Do other cities or counties need to provide the same or similar functionality and if so could we do it once and reuse it multiple times? How could we identify the right subset of 20,000 cities and counties that may be interested in collaboratively developing a certain application? What are the Best Practices in Software collaboration?
So, when our researchers said open source was a Best Practice in Software collaboration, I listened. They said that potentially large virtual teams found ways to work together to accomplish a common objective. In local government teams of fragmented developers work on similar applications independently and without knowledge of each other. So, open source looks like a natural fit -- not only for the quality of the products offered, but even more for the collaborative model of development.
I asked myself, why couldn't municipalities share their applications in a style similar to SourceForge.net or FreshMeat.net? The possibilities seem endless and the opportunity exciting.
TA: You have wanted to build an open source alliance with city governments. What problems have you encountered along the way? What will encourage cities to adopt Linux?
AS: I presented my idea of an alliance of local governments a couple of times to groups of CIOs and directors of IT. I received varied responses, from "great idea, long overdue" to "old idea, did not work before, what makes you think we can do this now?" When I asked for potential contributions to a shared repository only 2% of the participants responded. I feel we need to find a means of "jump starting" the repository. So, I decided that we could build it and they will come!
Open source products, including Linux, are very attractive development platforms with a low barrier to entry. Local governments seldom have sufficient resources for their technology needs. We want to provide the most affordable solution and encourage 100% participation. We want our runtime to be platform independent. The applications from our shared repository should run on the most popular operating systems and databases.
TA: Will open standards such as the Global Justice XML Data Model move cities to Linux?
AS: I don't know. I see open standards as something which allows for more diversity of operating systems while supporting integration. All platforms should benefit from standards, including Linux. The best software collaboration allows for platform choice. Independent of standards, Linux holds an attraction to local governments since it has a low barrier to entry.
TA: In an earlier conversation, you said you wanted to migrate from a mainframe foundation in the enterprise to an Internet foundation. What makes Internet protocols your preference?
AS: The foundation of eGovernment and eCommerce is inherently an Internet model. We have to end up with an Internet capability to provide the services citizens and businesses desire. Internally, we also prefer the same model because it's the most efficient way to manage and it reduces the number of skills our developers need to learn. The Internet model solves problems, such as how to distribute software to the desktop and leads to thin or thinner clients, which would reduce desktop management costs.
TA: How do you regard Linux as ready for the enterprise today? What about the desktop? Are you considering Sun's Java Desktop System?
AS: Linux has shown itself as a stable and a solid performer in the server area.
We have decided to test the Sun Java Desktop System and thin clients using Sun technology. We're also looking at OpenOffice and StarOffice 7 across various platforms. We're looking at our options, but we're not ready to commit.
We have a significantly large group of light or casual users who are good candidates for Linux thin-client technology. Others have a business need to run applications that don't run on Linux. They only run on Microsoft.
I'm interested in seeing the results of the tests. I believe we will end up supporting both desktop environments.
TA: What barriers do cities face in adopting open source software? Are the acquisition-cost-savings compelling enough?
AS: In general, public and private institutions will implement software applications that have a trusted support model. Software will break and the process of bringing it back to operation needs to be reliable. With open source software, the support structure is not always clear. Red Hat's behind Linux, and Sun's support of OpenOffice and StarOffice are examples of steps in the right direction. I'm also interested in seeing what Novell offers.
TA: What issues in government procurement practices keep open source out of the purchasing equation?
AS: In Virginia, our procurement rules and guidelines encourage competition and choice. No preference or restriction is shown to open source. However, it is not very often that we receive open source solutions as responses to bid requests. I understand that other states have more restrictive procurement processes regarding open source. I think procurement practices should allow people a level of flexibility.
TA: Does the SCO lawsuit and subsequent threats concern you?
AS: Of course it does. It further complicates the decision-making process and promotes uncertainty. Architecturally what we want is platform independence. That way our applications can be ported to any platform of choice. So yes, I am concerned.
TA: When you finish modernizing your enterprise, what do you think it will look like?
AS: I believe we will end up with an ERP for our core systems of finance, budgeting, human resources, payroll, and purchasing. We may find additional off-the-shelf solutions that will meet our needs at a reasonable cost, in areas such as criminal justice systems, permits and codes, and GIS.
We will attempt to build the rest of our applications with reusability in mind. We should collaboratively develop a shared repository of local government software applications that are flexible and configurable. The repository would become a public asset. Any municipality will be free to download, enhance, and deploy the applications. There will be as few barriers as possible. Enhancements should be incorporated into the base software and all municipalities should be able to benefit. If we build it once and reuse it 20,000 times, the economy will benefit as well.
There will still be cases in which, for practical considerations, we must custom build software. We will only do this as a last resort.
TA: Do you see a role for open source companies in developing the shared repository?
AS: We need to create an echo system that is mutually beneficial to local governments as well as open source companies. The most difficult part will be to jumpstart the repository and get a critical mass of municipalities in the practice to download and contribute enhancements. At that point open source companies will have an opportunity to provide consulting services and support. If an open source company has built an application and contributes it, they will probably see more success because more entities will want and need their services.
TA: You were quoted as saying the modernization effort in Newport News will cost $30 million. With 20,000 cities facing similar problems, how much can taxpayers save by going to open source?
AS: In Newport News we could avoid having to spend about $10 million dollars on custom built software. The less custom code we have to manage and maintain, [it] reduces our yearly operating costs as well. Our Application Development group has 27 employees who maintain our existing portfolio of 155 applications. With a collaborative process in place, the bulk of the maintenance and enhancement work may occur with a larger virtual team. A small number of developers should be reserved for changes mandated by the local government. I estimate that about 20% to 30% of my development staff could be reduced through normal, retirement-triggered attrition. In a few years this may result in another $200,000 savings per year.
I don't know if Newport News is a "typical" model for potential savings. But even if you cut my numbers in half or more, the multiplying factor is significant. The way local governments run Information Systems today, millions of dollars could be saved each year thorough collaborative development.
TA: Are there additional benefits for citizens in creating a shared repository?
AS: When localities start using "standard" software from a shared repository, the functionality of our web sites will become more consistent. eGovernment will become more meaningful with consistency. Many citizens need to interact with multiple municipalities in their geographic area. From parks and recreation, festivals and special events, to moving across boundary lines, it will be much more user-friendly to transact online.
TA: How can local citizens create a voice to encourage their local governments to participate?
AS: Citizens can call their City Council representatives and voice their opinion. It is their tax money at stake. The promise is more and better services for less.
Tom Adelstein became an author in 1985 and has published and written non-fiction books, journalistic investigative reports, novels and screen plays prolifically ever since.
Return to the LinuxDevCenter.com.