The New York City Council is expected to vote on a far-reaching open data bill on Wednesday that would codify many of the principles articulated by open government advocates in recent years.

If made law, the bill would go further than San Francisco's pioneering 2010 open data law in depth and scope, obliging agencies to provide data online in machine-readable format though a single, citywide portal. But perhaps in a nod to the amount of work involved in working through large volumes of existing data, city agencies won't have to make theirs available through the city's portal until the end of 2018.

The city's move is the latest step taken in the United States as part of a wider movement by open government advocates to remake government services in the Web 2.0 age. While the specifics can be esoteric, the impact of the changes are expected to be profound. The goal of publishing machine-readable information using common technical standards is to enable both the public and government employees to make cities better, whether that's through the new raw material for a civic-oriented business to more strategically and efficiently making services available to citizens.

This bill is the result of "many, many hours spent between the lawyers of the Bloomberg administration, and the lawyers on the city council," said New York City Councilwoman Gale Brewer, its chief sponsor.

In an interview on Monday, she said that while the Mayor's office has made some city data available online, this law would ensure that data across all city agencies would be made public and searchable.

"This data would be social services, and police, and things that you didn't know that you wanted to know about," she said.

Addressing a key complaint by open government advocates about unsearchable PDFs, the law specifies that public data sets have to be "accessible to external search capabilities." It also gives the city's Department of Information Technology and Telecommunications (DoITT) six months to publish a technical standards manual that will guide agencies' process of publishing the information that they generate on the web, and make it machine-readable. DoITT will have to explain the reasoning behind its choices, and "shall include a plan to adopt or utilize a web application programming interface that permits application programs to request and receive public data sets directly from the web portal." DoITT will also be obliged to consult with web standards bodies to keep current with open standards.

While the law gives agencies six years to get all their information online and searchable, they'll have to start reporting to the mayor and the public about what data sets they do have within one and a half years after the passage of the law, and submit a compliance report within two years. The new law also says that the data is to be made available in one central location at NYC.gov, and that DoITT should actively solicit feedback from its community of users about its utility.

There's no penalty for the agencies for not complying, but Brewer expects the public feedback mechanism to exert a significant amount of force.

"It's a little bit of an honor system, but New Yorkers are going to scream and yell if something isn't right, and then the Mayor's office will have to take the appropriate action," she said.

At the same time as the law tries to make the city's data open and interpretable, the city's soon-to-be new policy makes clear it's not a free-for-all. Information about agencies' internal deliberative processes, procurement and negotiating positions, for example, are exempted.

The New York City Council has been mulling over the legislation for years, holding hearings about open data and open tech standards in 2009 and 2010. And they haven't been alone.

In California, for example, open government advocate Adriel Hampton is pushing the state and city to define open data too. On Monday, he praised the New York City legislation, but bemoaned the timeline.

"The New York open data law has some great provisions - a centralized home for data sets, a requirement that data be accessible to search engines, and a provision for public input on desired data sets," he said after taking a look. "One drawback that reflects governments general inability to move at the speed of technology is that the law has an implementation timeline that stretches out seven years, through the end of 2018."

Stephen Romalewski, an open data advocate who is also the director of the CUNY Mapping Service at the Graduate Center of the City University of New York, said the bill largely accepts what the administration is already doing — but gives them years to do it.

"They've been negotiating this for almost three years now," Romalewski said. "They used to have a bill that actually pushed the city to have a more visionary approach, and now they've stepped back and said, all right, the administration is doing this and we're going to accept that and write that into the bill."

Romalewski said that during public comment on the bill, advisory groups had called for the naming of a "chief data officer," someone who would take responsibility for "ensuring that agencies develop datasets, publicize them, make them accessible in meaningful ways." That requirement didn't make it into the bill.

The city has a chief digital officer, Rachel Sterne, but her portfolio has been more focused on social media and on improving the experience someone has when trying to interact with New York City agencies online.

Meanwhile, the City of Chicago has a chief data officer who works with data users across agencies and shepherds the city's ambitious data releases.

"I'm looking at this from a distance, but in Chicago it seems like that's what happening," Romalewski said. "They have a chief technology officer, they have a chief data officer, and they say that's the direction they're moving in."

On the other hand, the bill does call for a comprehensive inventory of the data the city already has — and that, he said, was a step forward.

The New York City bill, Int 0029-2010, was voted out of the council's technology committee Tuesday and is expected come before the entire City Council at a stated meeting Wednesday.

This post has been updated.
This post has been corrected: An earlier version incorrectly attributed to Stephen Romalewski the idea that legislation had called for the naming of a "chief data officer." He said instead that it came up during discussion of the legislation.
With Nick Judd