It’s interesting to reflect back on the not too distant past and think about how governments have used open source software.
For many state and local governments – as recently as a few years ago – the use of open source software was something of a foreign concept. Many a government IT worker made an impassioned and well reasoned plea to bosses and co-workers to consider using open source software to capitalize on a range of different benefits. (And yes, in case you’re wondering, when I was a state government IT worker I made many such pleas to bosses and co-workers.)
For a long time, those pleas usually went unheeded. How far we’ve now come.
Not only are more and more state and local governments starting to realize the benefits of using open source software, some of them are actually starting to become developers of open source solutions.
The notion of government-sponsored open source software development isn’t necessarily new – it’s how we got SE Linux for example. But it is still a relatively new concept for state and local governments. I can think of two government entities that are leading the charge.
First, the New York State Senate. Under CIO Andrew Hoppin, the New York legislature’s upper chamber has become a leader in the public sector for its use of open source software. Not only does the Senate use Drupal for its public website, they also contribute Drupal modules back to the community.
After becoming the first legislative body in the country to develop iPhone and iPad apps, they not only released the code for these apps on GitHub, they actually did a code walk through for developers at a recent event in Albany.
And now, the State of Washington has released the code for its own mobile apps on GitHub. They are actively encouraging people to submit ideas to help further develop the software and to identify bugs.
Both of these governments (and others who are out there doing the same thing) will realize more benefits by open sourcing the code for their apps than they would have had they kept the source code a secret. By making the code for their apps visible and reusable, they’ll attract more developer interest and help ensure that bugs or security issues get identified quickly.
Other governments will benefit as well – these two organizations are clearly ahead of the curve in developing mobile apps, and other governments will benefit from their experience and expertise. Since’s governments generally don’t compete directly with each other, this type of sharing makes perfect sense.
Improvements or enhancements to these open source applications will in turn provide benefits for the governments that created them – this is one of the driving dynamics of Gov 2.0.
I’m wondering if there are other state and local governments out there doing this same thing. Do you know of a government agency or entity with a GitHub repo or other open source code repository?
Leave a comment below with the details.
Starting about 2005, the city of Newport News, VA started down the open source road, found what appeared to be a healthy and active support community using the Zope/Plone platform for web hosting/development which is based on Python. We trained a team of programmers in Python, Plone, and Zope and formed a team of city leaders to design a new city web site, which the programmers implemented.
The driving force behind the move was our IT director, GovLooper Andy Stein, and had two main purposes:
1) To get us away from mainframe-based, all internally-supported systems to other options such as open-source or off-the-shelf vendor applications, and
2) To build a collaborative community of municipal governments that would work together to not only enhance our web application, but to also develop other common-use applications.
Though we were awarded the 2008 Havlick Award for Innovation in Local Government, the 2008 Digital Government Achievement Award, and gained some international recognition for our efforts, we found minimal interest among other municipalities to collaborate. Therefore, the effort has lost momentum and internal support, though we do continue to use the platform for both our internet and intranet sites. We did succeed in getting one city using the application (we actually host their site) and two other cities are using it for their intranets.
More information about the Open-Egov application can be found at http://www.nngov.com/egov and a repository (with a few other smaller applications) at http://open-egov.nngov.com/products (Note: I just reported to our tech services that the repository site is down, so if you get an error, try again later).
I seem to recall Andy as part of the Government Open Code Collaborative effort with Jim Willis from Rhode Island. http://www.linuxjournal.com/article/7932
I think that was an idea that was ahead of its time. It’s interesting to see the recent excitement around local government open source efforts, but I do wonder why some earlier efforts had such a hard time gaining traction.
Congrats to you guys for being leaders on this.
Mark this post is closely related to yours, worth a look for anyone interested in open source repositories: https://www.govloop.com/group/OSSinGov/forum/topics/the-future-of-the-government
Steve,
I hate to link people outside of govloop unnecessarily, but I think Gunnar Hellekson’s original blog post is where all of the interesting comments are coalescing: http://onepeople.org/node/2159
A lot of us have weighed in, and I for one applaud the use of more Open Source in government. However, I’m a pragmatist, and some parts of Gunnar’s post (and other’s comments) go too far into the ‘free beer, peace & love’ realm. The reality is that we’ll NEVER get 100% of gov software development into the Open Source space, no matter how much we might all want it.
As a taxpayer, and someone involved in helping government try to be more successful at internal collaboration, I think the notion of a limited access ‘forge.gov’ for getting agencies to share across boundaries is an important and worthwhile endeavor, in addition to helping educate them on how to best work with the existing open source community.
Thanks.
Great post! We’ve been having some board level discussions about implementing a code repository for all local/state governments to utilize as part of our organization, the National Association of Government Webmasters (NAGW). Currently, we lack the time and resources to develop our own code repository, but we’ll be looking further into utilizing something like GitHub in a more organized manner.
Thanks Steve, Guy for the comments. I had forgotten about that post from Gunnar. Good to ge back and read that one.
Ron, I can’t say enough good things about GitHub. It has dozens of features that make it an ideal platform for sharing code. I’s interesting to see most developers in the open source space migrate to GitHub over other (older) alternatives like sourceforge and freshmeat.
I’m certainly no Git expert, but if I can provide any insights or assistance on GutHub, please feel free to ping me.
http://github.com/mheadd
I would be remiss for not mentioning the folks at the Office of the Chief Technology Officer in DC. They also are way ahead of the curve in regards to sharing code with other governments (and for lots of other reasons too).
They’re on GitHub at http://github.com/octolabs
I have been working with a number of government customers who have taken open source solutions and enabled a transparent, open data solution. One which is past completion is Edmonton with their http://data.edmonton.ca site. This is based on an open source package called OGDI (Open Government Data Initiative) from Microsoft (http://ogdi.codeplex.com). The data is published in the Microsoft cloud and the front-end web server also runs in the Microsoft cloud – with the OGDI interface being used to query and update the data using REST. The OGDI package comes as a complete “starter kit” that is fully customizable (compare http://data.edmonton.ca with http://ogdisdk.cloudapp.net, the ‘generic’ version). In addition, Microsoft has a number of open source applications, utilities and other materials on the CodePlex site (http://www.codeplex.com).
Stay tuned for an upcoming press release about another well-known city that will be publishing more public data to the cloud using Microsoft’s OGDI solution … coming soon!
Sharing government-developed solutions, particularly solutions that have repeatable value throughout many agencies, is absolutely the right thing to do. It allows for more scalable use of technology, better cross-government collaboration, and fiscal efficiencies through shared costs. In addition, one could argue that code developed in agencies really belongs to the public and needs to be open, and you also have the added benefit of better transparency.
Thus, publishing government-developed solutions as open source can make a lot of sense, as it can be a good model to foster shared use and collaboration.
That said, it’s important to note that publishing code as open source also comes with a great deal of responsibility and potentially a good amount of overhead. Putting code into the public arena and incorporating code from outside resources can introduce issues around IP ownership, use, and licensing, not to mention the need for thorough processes to guarantee code security, quality, and integrity. You can also add the complexity of managing branched projects and complex versioning. These are not easy issues, and if not done correctly, can lead to damaging repercussions that may erode any potential benefits.
Net net: doing open source right can require some strong engineering discipline and legal oversight – particularly when done for an institution (e.g., a government agency) that has significant exposure (and therefore, significant risk). So, while we all agree that sharing government code through open source can be a very good thing, it’s important not to be naïve about what it takes to do right.
Another note: There’s a bit of ambiguity in the article between submitting government-developed code to open source and using products that are open source. While I strongly advocate the sharing of government-created projects via open source as much as possible, I think that agencies (and everyone, really) just need to be pragmatic in the decision of whether to use open source or commercial products in their environment and as part of their solutions, as there are benefits and complexities on both sides. There are some outstanding commercial technologies that provide significant value to Gov 2.0 initiatives – and they shouldn’t be discounted because they aren’t open source.
For example: should an agency choose not to use a really valuable iPhone solution – even if it would provide significant benefit to its constituents – just because it’s not open source? If there’s a great application built on SharePoint that would provide real cost benefits not be considered if it isn’t open source?
In each case: of course not. Products and technologies should be selected based on the capabilities and value they provide to Gov 2.0 solutions, whether they’re open source or commercial technologies. Cutting out either side of the equation severely restricts the options available to agencies, and diminishes the overall quality of Gov 2.0 solution for citizens.
Net net (again): it’s wrong to say that Gov 2.0 is open source (just as it’d be wrong to say that Gov 2.0 is “commercial”), as both open source and commercial software can provide great benefits to Gov 2.0 scenarios. Gov 2.0 is bigger and broader than individual technologies, development, and licensing models – and we’ll miss out on a lot of opportunity if we pigeon-hole it to any one thing.
“There’s a bit of ambiguity in the article between submitting government-developed code to open source and using products that are open source.”
Let me be unambiguous. If part of what defines Gov 2.0 is:
* Enhanced transparency in government operations and processes; and,
* Enhanced opportunities to share solutions both within governments, and between them.
Then open source software is hands down more beneficial. Governments that use open source software better understand the nature of the open source ecosystem, and are more likely to reap the benefits of code reuse.
I’m not suggesting that governments don’t need to be pragmatic, and that there aren’t use cases where closed-source commercial software doesn’t fit. I’d say there is very little danger that the use of closed-source commercial software will wither in government given the many sales reps and lobbyists that the software vendors employ. (I’m not judging, I’m just saying.)
The point here is that governments are uniquely positioned to benefit from the reuse of existing software solutions that are distributed and maintained as open source projects. With very few exceptions governments do not compete directly with each other – the State of Texas is not likely to make a play to license the drivers of the State of Rhode Island.
However, Texas and Rhode Island can both potentially benefit if one of them shares a solution to help license drivers that the other state can make use of. You can’t make this argument with private sector entities because the same dynamic does not apply – many of them are in direct competition with each other, so they have no incentive to reuse code (nor will they gain a benefit from it).
This is what makes the benefit of software reuse in government such a potential game changer. It the same idea that is behind the recent announcement of the Civic Commons (http://civiccommons.com/)
And ultimately, I believe, it is what will help Gov 2.0 realize its full potential.
Apple apps, it is not.
As I said, I think that sharing government developed solutions via open source makes a lot of sense and can provide a great deal of value, but that doesn’t necessarily mean that using only open source programs will be any more or less beneficial to agencies or citizens.
The fact is – there is a LOT of high quality commercial software for which there is no open source equivalent… and for which there will likely be no open source equivalent in the near or far future. The reason commercial software will have a future in governments isn’t because of sales reps and lobbyists – it’s because the products provide real value and address real needs in the area of Gov 2.0 (and other scenarios, of course). There’s a very large community of ISVs out there who feed their families by providing great commercial solutions to governments, and they shouldn’t be disparaged or blackballed because of the revenue model they’ve chosen.
In addition – the vast majority of users of open source products, including government agencies, will never – and probably should never – directly work with the source code and will never be exposed to open source development and licensing processes. Thus, there is almost no practical difference to them whether it’s open source or commercial… and thus, their choice should be driven by the basics: value and capability.
We should all support and foster the sharing of code and best practices among agencies in all channels – and publishing to open source is a great avenue for this. At the same time, we should be advocating that agencies select their software based on value and capability – sometimes this will be open source, sometimes it will be commercial… most of the time, I suspect it will be both. Using open source can be good – but that doesn’t mean using commercial software is bad (as it can also can be very good) – they’re just two paths to the same goal.
And that’s okay – because what’s most important is that the agencies get the best solutions at the best value, and that they spend more time using the technology for the benefit of citizens than parsing the technical choices along subjective lines.
Remembering a previous job life in technical support, I have to remind the community that open source software packages have a wide variety of support arrangements. When you decide to implement any open source solution, it is critical to understand the impact that an extended “issue” might cause. “Open Source” is not necessarily free … and your payments are often in the form of support contracts. Other FLOSS (free/libre open source software) packages often relegate technical support to blogs/message boards/other.
Having internal government agencies sharing source code can be a great option to provide some widespread customization ideas and building on a standard solution. But, we also have to understand that the “support trail” will likely work its way back to a select group of techies who will be called out to help debug problems with the code (maybe not even pieces of code that they authored).
Whenever you implement internally-generated source code … don’t forget to plan for technical support and a “contingency plan” just in case those developers decide to change jobs or otherwise move out of a role that would allow them to provide support.
I am all for using the best solution to address a problem…but I believe the designation of “best solution” has a broad definition that includes technical functionality, efficiency, and sustainability.
Hi Mark, some folks in local government in Oregon are using gethub, i’ll put you in touch. Also expect to see someone on the program at http://www.goscon.org or at the igniteGov discussing same ( http://www.goscon.org/ignitgov ) Government communities, collaborative and forges are on the rapid rise here in the US and in particular overseas where governments are investing in the infrastructure and governance to do so. We’ve seen a particularly sharp increase this past year (a review of GOSCON presentation archives shows us moving from worries about legal and IP issues in 2005 to a program this year full of use case and collaborative projects). Closer to the ground – and fully acknowledging that there are many fine commercial products out there that support the concept of Gov2.0 – open source by its nature is accelerating transparency and (bearing in mind that All Data is Local) information sharing in a way that would be otherwise inaccessible to agencies of modest means to meet critical program needs.