Author Archives: Gunnar

OSFA responds to draft “Shared First” policy

[If you agree with this position, please head over to the IT Shared Service Strategy IdeaScale site and upvote! Thanks!]

Open Source for America applauds OMB’s effort to increase the efficiency of the Federal IT budget through the principles of commoditization, reuse, sharing, and collaboration described in the draft IT Shared Services Policy distributed on December 8th, 2011. These principles are also the hallmarks of open source software, and while there is no explicit mention of open source in the plan at this time, we believe that there should be. We see a unique opportunity for open source to improve the effectiveness of Shared First.

“A number of barriers exist which have prevented the broader adoption of shared IT services. Lack of information sharing among the Federal agencies, budgetary restrictions, acquisition issues, and other factors have all contributed to a culture in which proprietary, specialized systems are the norm.”

We could not agree more. We believe that OMB should explicitly mention open source as a recommended method for overcoming these barriers. Open source software itself, of course, can reduce costs and ease acquisition challenges. We can assume that many agencies will naturally use open source software as part of their Shared First implementation.

Embracing the open source approach, though, and using it to encourage sharing between agencies, departments, other governments, and the general public is the purest expression of the goals of the Shared First policy. Open source excels at the Shared First Design Goals, including visibility, commoditization, reusability, extensibility, and standardization, and we believe it should be actively and explicitly encouraged by the policy.

Much of the draft is concerned with the sharing of existing infrastructure, and is in this way very similar to the Federal Data Center Consolidation Initiative. As the scope of the Shared First policy is much broader than the FDCCI, and concerns itself with LoB information systems and applications, we would like OMB to consider expanding the scope of the mandate to include the sharing of development resources, and not just their ongoing operations and maintenance.

As the policy encourages agencies to “move up the stack,” and share ever-more complex layers of their information systems, agencies will need to perform more customization to meet their specific mission needs. If they were to use proprietary commercial offerings or systems developed for just one agency, this can become very difficult and very expensive.

If agencies were to instead employ open source software, or better still: share their taxpayer-funded software under an open source license, customization would become easier, and the need for customization would be reduced, owing to the natural modularity of open source projects.

In the process, agencies would be making themselves available to contributions and improvements from their partner agencies as well as state and local governments who also have a use for that same software. Certainly, it is possible to share complex application software amongst agencies without open source software. We believe, however, that releasing software under an open source license can simplify this process, and simultaneously encourage sharing among other state and local governments, and the private sector as well. This approach has already been successful at NASA and the National Institutes for Health. We see no reason why other agencies cannot realize the same benefits.

“Open source is… the most concrete form of civic participation.”

— Macon Phillips, White House New Media Director

An excellent example of this is the effort currently underway at the EPA to develop a shared management system for Freedom of Information Act (FOIA) responses, one of the goals of the President’s National Action Plan under the Open Government Initiative. Almost as soon as the FOIA project was announced, it became clear that many citizens, state and local governments were interested in developing this platform alongside the EPA.

Should this project succeed, there would be a single, common platform for handling FOIA responses that would be freely available to the 93 agencies required to perform this task. Rather than 93 purpose-built systems, agencies could take advantage of each other’s innovation, the innovation of the private sector, and the innovation of thousands of state, local, municipal, and tribal governments who have their own Freedom of Information requirements.

Unlike working with proprietary FOI software, the number of agency-specific customizations would be drastically reduced, and since customizations can be contributed back to the main project, the ongoing maintenance burden of these customizations would be reduced, as well.

This kind of inter-, intra-, and extra-agency collaboration using the open source model is already finding success throughout the government, including the Veterans’ Administration Open Source VistA program, the National Security Agency’s SELinux project, and the OMB’s own data.gov.

It’s not difficult to imagine that these projects can encourage the growth of small businesses specializing in the implementation and maintenance of this software. Incorporating these stakeholders would bring to bear on the EPA’s FOIA platform far more development and testing resources than the EPA could muster on its own. Because the software assets have been commoditized, the EPA and other agencies would no longer be beholden to a single vendor for its FOIA software. This is the kind of relationship that OMB would like with its software assets, and this is precisely the kind of collaboration and commoditization that the current Shared First policy encourages. This is why we believe that the policy should make the role of open source explicit.

Open source software is particularly useful when the Federal government mandates action by state governments. The New York State Office of Temporary Disability Assistance has had great success in sharing open source eligibility logic with its counterparts in other states, allowing them all to share the burden of implementing CMS eligibility requirements. The Office of the National Coordinator for Health IT has employed open source software to encourage the adoption of its CONNECT standards for electronic health records. Without open source, these efforts would rely on potentially expensive proprietary software, which would have to be either unfunded or subsidized by the Federal partner to encourage adoption of their mandates. Instead, these organizations can now work collaboratively and transparently to solve their common missions.

Releasing agency software as open source has other benefits, as well. One of the primary concerns of the draft policy is to encourage visibility: agencies cannot share with each other if they do not know what’s being shared. The open source community has largely solved this discoverability problem through tools like sourceforge.net and github.com, both of which may be used, free of cost, by Federal staff and the public.

Open Source for America is excited by this renewed focus on agency collaboration and transparency. We strongly support the Shared First policy, and believe that it is one of the best tools available to meet agencies’ current challenges. We look forward to working together with OMB and the implementing agencies to make the policy successful. We also believe that by encouraging agencies to open source their software, and to share that software with each other, we can together ensure that agencies are putting their existing budgets to their best and highest use.

OSFA Responds to the US Open Government National Action Plan

As part of the Open Government Partnership, the US Government’s National Action Plan was released and the White House asked for input from the public on how citizen participation could be improved, facilitated, and measured. Here is OSFA’s response, which we submitted on January 3rd, 2012.

Open Source for America (OSFA) appreciates this opportunity to comment on the Open Government National Action Plan. OSFA was formed in 2009 to advocate for the use of open source software and the open source development process in the US Government. OSFA sees the Open Government Initiative as a unique opportunity to institutionalize the use of these tools to further to goals of transparency, collaboration, and participation &emdash; values which are inherent to the open source process itself. More information on OSFA can be found at http://www.opensourceforamerica.org.

Below, we focus our attention on one of the questions posed in the blog notice: What are the most effective forms of technology and web tools to encourage public participation, engage with the private sector/non-profit and academic communities, and provide the public with greater and more meaningful opportunities to influence agencies’ plans?

OSFA stands ready via its leadership and membership to work collaboratively to achieve the goals of the Action Plan and help identify Open Source tools that can assist in the work of the Action Plan.

Participation in Open Government Plans can take many forms, and be useful at different stages of the consultative process. This should be reflected in any system of measurement.

As a general matter, however, the forms of technology and web tools should be rooted in the open source development model and rely on open standards to encourage the widest participation and broadest utilization. In many respects, the work and progress of the Open Innovation agenda is relevant to achieving the goals of the National Action Plan.

This approach has a number of important collateral benefits for government: it encourages an on-going, cohesive community that wants to improve the technology and integrate user experience. It fosters a highly engaged constituency (vs a specific regimented approach) which becomes part of the process by developing new tools, modifying tools and the reuse of content and data which is produced through the consultation.

Based on our assessment of open government initiatives around the world to date, open government thrives where citizen interest is met with public sector support, and accelerates when open source innovation is a central core tenet of the policy. Open government initiatives and open source innovation share many core values , including: transparency , meritocracy of ideas , focus on the public interest. The tools generated from open innovation are also, based on our experience, highly available, affordable, and interoperable.

Too often, participation is narrowly understood as a written response to a government proposal. Instead, focus should be on encouraging participation through useful work by developing tools on a collaborative basis from relevant stakeholders which involve elements ranging from collecting and verifying data, bench marking its implementation, and the monitoring of its execution. This kind of participation is not well accounted for under the current consultative processes, which understand formal participation only as a persuasive document. Fortunately, we are no longer confined to the strictures of simply faxing in (or even mailing in) comments; the Web has allowed broader and more vigorous participation.

There are some useful examples that the Action Plan could draw upon. “Apps” contests, like Apps for Democracy and Apps for America, have been used largely as a tool for improving or supplementing the services delivered by government. We believe that there is also an opportunity to engage the software development community as policy is being developed, through contests and other means, to ensure their effective administration and help prevent unintended consequences.

Such tools would also increase the type and nature of feedback an agency might receive. We can imagine agencies being more than a convening body for input on the immediate matter, but a collection point for related data sets and tools that could improve the quality of the conversation more generally.

The challenge, in our view, is the ‘next step’ in realizing the full potential of Web participation. Having benefited from the enormous potential of the Internet, using open source tools and open standards, there is a need to focus on tools and processes that assist agencies in assessing the data and making actionable responses. As a step toward this end, the National Action Plan should include a recognition of this core element and seek community input on the state of such tools (including business analytic tools), and how they can be reused and broadly populated.

Plinket Collaborative

The Plinket Collaborative community generates and supports web authoring environment that libraries in partnership with states/regional organizations use to create new web sites for their patrons. It also creates shared content and other generic information that can be provided through the web sites that are valuable to library patrons.

The Collaborative provides identifies requirements unique to library needs in a highly collaborative manner through a formal and inclusive governance model. The actual technical development of the templates and tools based upon the Plone and Zope platforms is done by a third party contractor, so the code generated is done by a directed effort rather than an open source style development effort. Plone and Zope are both traditional open source projects.

Formal governance; a central fiscal agent to manage expenses; sharing of documentation and practices; encouraging outreach and promotion by its members are key factors in the projects stability and growth.

The complete Plinket case study can be found on the CENATIC oberservatory web site and is provided courtesy of the author, Deb Bryant.

NASA World Wind

Originally created as an educational program in 2004, in 2006 the Department of Energy (DOE), impressed by World Wind.NET, asked NASA to make World Wind into a platform-neutral technology. Two significant decisions were made: re-factor World Wind using cross-platform Java and re-architect World Wind from an application to an API centric SDK that could be more easily used by other applications.

The re-factored and re-designed version of World Wind has greatly improved the versatility of this technology and the degree of acceptance in the software application development community, as witnessed by the significant government and commercial use of this technology. Since World Wind provides the infrastructure for information exchange, efforts to provide value can concentrate on solutions. In this way NASA stimulates entrepreneurial enterprise, provides the government with control over the information exchange medium, and enhances the ability for the world to communicate and share information.

Moving the community portal off government servers also served to improve the projects ability to grow the project as well as leave agency technical resources to focus on core development. A strategy of supporting a fully open community allows actors from outside of government to participate in the project, furthering the adoption of the software and broadening its use in a range of applications

The complete World Wind case study can be found at the CENATIC web site and is provided courtesy of the author, Deb Bryant.

PloneGov

CommunesPlone, now in its fifth year of operation, has strong adoption within the Belgium local governments and also includes several French local governments. The software, well-supported by an open source- educated IT work force and an supportive open source community and subject matter experts, started small and grew through a dedicated IT ecosystem which included individuals with expertise in open source collaborative models. Early investments in education and coaching enabled the public sector IT departments to engage directly in the development of the software and the community itself.

You can read the complete PloneGov case study at the CENATIC web site and is provided courtesy of the author, Deb Bryant.

Plone, the basis for the applications within the project, included foreign language support earlier than its equivalent solutions, making internationalization of the project a natural progression.

Distinguishing characteristics of this project include:

  • A cohesive local public sector effort supported through open source coaching at project inception.
  • Has developed strong and direct relationship with the open source community.
  • A broader collaborative initiative designed (and based on the Plone platform) to support activities not core to government IT operations such as writing case studies, creating conferences and workshops, developing community partners.

CommunesPlone is exemplary in the degree of collaboration within the public sector and amongst the open source community.

There are many lessons to learn from the project including best practices closely modeled after open source development, and also from the manner in which resources are dedicated to outreach, education, and promotion.

Project aspects to watch with interest: Funding of centralized or coordinated efforts, and scalability as more agencies outside of the current community adopt the software and require deployment support.

Sahana

Sahana came into existence through the collaborative effort of IT professionals in Sri Lanka and around the world to provide relief to victims of the December 26, 2004 Tsunami in Indonesia. In October 2009, the governance and management of this community based, open source disaster management software program, through sponsorship from the Swedish International Development Cooperation Agency (SIDA), IBM and the National Science Foundation (NSF), was given to the Sahana Software Foundation, a 501(c)(3) non-profit organization. Sahana has been successfully deployed in 14 disaster relief efforts as well as 14 pre-deployment preparedness and mitigation strategies around the world.

Sahana is primarily supported through the donation of funds, infrastructure and resources from organizations in industry (e.g., Google, IBM, Sri Lanka Telecom IDC), government (e.g., NSF, SIDA) and the non-profit sector (e.g., World Food Programme, Lanka Software Foundation). By establishing onsite teams that train IT professionals in the disaster area, Sahana has successful integrated industry workers, government agents, volunteers and resources. This effort not only alleviates suffering for victims, but streamlines interagency communication and improves the robustness of Sahana’s software.

Drawing heavily from the Apache Foundation’s community model of a meritocracy, the developer and user community is directed by the Sahana Software Foundation’s Board and Project Management Committees. This governance directs Sahana’s three primary internal projects: Sahana-Agasti (PHP-based, resource management), Sahana-Eden (Emergency Management Environment) and Sahana-Mobile (code base for mobile platforms)

In general, governments around the world have benefited or may benefit from the Sahana software with no financial investment. As the project has matured, it has gained the confidence and interest of government agencies for permanent deployment within their Emergency Management Operations. Most recently in the US, government has begun contributing to the development effort.

The Sahana project has called to the open source industry to help develop at cooperative business model, one which supports the viable commercialization of a product with a service industry to support and sustain the project while maintaining its humanitarian principles.

You can read the complete Sahana case study at the CENATIC oberservatory web site and is provided courtesy of the author, Deb Bryant.