The CTO should ensure that agencies create websites that use open source software and distribute data in open formats that are accessible to all search engines. OMB should direct agencies to actively make all their online resources searchable by major public search engines and available in open formats.
Similarly, agencies should have a policy to exercise a preference for open-source software for government activities as a means to improve stability, transparency, metadata quality, and cost-efficiency.
Additionally, agencies should strongly consider supplementing increased searchability with proactive efforts to promote and advertise data to potential users who may not know the proper search terms to use to find the data. Agencies’ responsibility does not end with making data easier to find. Agencies also need to ensure that the information inside these databases is presented in a simple, straightforward manner that allows average citizens to understand and use the data.
- From the 21st Century RTK Agenda
Comments (8)
I am not a techie, so the following may only echo the original point.
I enjoyed a short piece in the Jan/Feb issue of The Atlantic on iGov -- which focused, in part, on "Application Programming Interface" as a key element to be sure datasets are easily used by programmers for useful interfaces or mash-ups of various data.
source for the article: http://www.theatlantic.com/doc/200901/technology-government
I mean not to insult your intelligence 21stCenturyRTK but I think that you should familiarize yourself more with open source software and the security vulnerabilities open source exposes the federal government to.
It is true that the open source content management system Drupal has been used by the Obama administration in one of their most recent projects, but you cannot expect to have the rest of the government follow this blunt move. As Vivec Kundra pointed out on his 80/20 vision for IT implementations, you can expect 80% of the government IT implementations to be open source, cloud implementations, even crowdsourcing ideas as is the very site that we are collaborating in at this very moment. While dedicating the other 20% to secured infrastructure that resides withing well layered and secured servers running on enterprise software products that hosts the more valuable and sensitive data.
The worst offense is the ubiquitous use of the Microsoft word processing format which produces documents in the .doc format. Adequately viewing such a document requires purchase of the program from Microsoft at a substantial price.
See Open Source Agency.
Fredbaud,
I would like to clarify your post – you are correct the .doc format is nearly ubiquitous and the .doc format is arguably the most widely used format and is supported by nearly every word processor in existence today including Google Docs, IBM Lotus Symphony, OpenOffice and many others (some that are 'free' and some that require purchase).
However, you are correct, to take advantage of the rich feature set within Microsoft Office and support, security updates, and collaboration, you should use an appropriately licensed copy.
21stCenturyRTK -
A government preference for a particular business model does not seem appropriate (when you say 'open source' do you mean the business, licensing, or development model for software?). You do hit a fantastic goal of 'ensure that online government data are accessible to the widest possible audience' - why would you need a procurement preference to do so? Regardless of the goal, shouldn't government be able to set requirements and procure systems/software/solutions based on merit? Procurement and solution considerations should include the widest choice - and based on merit the best alternative should be used.
Furthermore, what about hardware? A far bigger challenge for access to government data lies in the proprietary architectures inherent in the legacy systems in place today. 'open formats' for hardware is an interesting discussion and perhaps 'open architectures' is the right conversation to achieve your goals. Software (of all kinds) is the solution to unlock this data.
Agencies should, whenever possible, minimize the domains/urls they use to publish information. Better disclosure of data (as many mention here) is worthless if it can not be found easily because it is spread across many domains needlessly. Data spread across many sites when it should be all on one site muddies the information architecture waters.
Agencies should be required to disclose and justify how many domains - urls - they use on the OMB 300.
Example: an agency chooses to purchase and use a url for a specific project, or for a specific one time event. The project and event are wholly government owned and funded. Often this information is duplicated: published under the agency's own url AND the one-time-use only url.
Managers within an agency often believe their information deserves its own domain/url. But there is often no benefit to the agency or to the taxpayer to publish this information under a different url. The specific project and event can often be easily and quickly published on the agency's central website at a lower cost - which also enhances its searchability. Limiting agency urls to only justifiable agency usages provides better searchability, transparency and accountability for taxpayers.
Some agencies use several domains - even .com urls - to publish data which incurs significant taxpayer expense. Each additional domain incurs a fee for the domain, extra costs for hosting, and extra costs for specialized web and or application development - when a less expensive and centralized website option is available.
Data that is spread across several websites is difficult, if not impossible, to gather metrics on (# visits, visitor satisfaction, visitors' concerns/comments). And often these specialized websites do not meet federally mandated regulations and requirements such as FISMA, No Fear Act, Section 508, FOIA and other OMB regulations.
I agree that data should be available in open, standard formats. I am also an Open Source advocate. I do not understand why you would stipulate that the government MUST use open source.
The government should choose software on its merits, not whether it is open source or proprietary.
I strongly agree with open internet based data formats for information that is provided to and published by the government. I am also favor open source but agree with Russell, the requirement for open source expressed here is overly strong. The best solution for the government and people should be used. Developing open source assets to do so would be a good idea (see http://opengov.ideascale.com/akira/dtd/3887-4049).