You are browsing the archive for jordanhatcher.
Some things I’ve been reading that should be of interest. There is a great deal more that I’ll put up soon.
Read the history of open data in Richard Poydner’s “Peter Murray-Rust and the data-mining robots“. Indeed, without Peter Murray-Rust, we wouldn’t have the Open Data Commons project.
Report on Ordnance Survey licensing practices. Free Our Data covers in this post the report from the Commons Select Committee on Communities and Local Government looking into Ordnance Survey’s licensing and business model.
Ordnance Survey is Britain’s national mapping agency. Due to UK government works being able to have IP rights over their work, if OS or some other agency wants to make their work public domain, they’d need some tool such as the PDDL to do it.
Also, Sophie L. Rovner “It’s All About Access” Chemical and Engineering News.
“It won’t be long before open access is old hat, taken for granted by a new generation of tools and services that depend on unrestricted access to research literature and data. As those tools and services come along, they will be the hot story. But historians will note that they all depend on open access and that open access was not easily won.” Peter Suber
Free and libre/open source software (F/LOSS) movements have spawned similar solutions in many other contexts, each at differing stages of development. As F/LOSS enters the routine and familiarity of middle age, the open content movement–open source for non-software copyright and best embodied by the work of Creative Commons –has just graduated university and is getting a feel for the world. Even younger is the open data movement, whose legal tools have just started to come online.
And with that, starts a new article I’ve written for the Open Source Business Resource, which is now available in the February issue. Implementing Open Data: The Open Data Commons Project
A comment via personal email and via the discuss list for the Open Data Commons set of legal tools. This is about the Public Domain Dedication & Licence:
Would it be possible to include an indemnification clause for misuse of data, for example when the data is passed on to a third person not bound by the license? The reason is that the disclaimer only binds the person downloading the data from the data source. If that person passes the data on to a third party, all warranty disclaimers become unenforceable because that third party did not agree to them (and the license is not passed on to the third party because it is not a “copyleft”-type license due to the public domain dedication). Are there any other legal instruments to protect the data provider against liability and lawsuits stemming from the data (imagine that the data has been modified down the road and is not accurate anymore)?
This is an interesting idea, and one that I suspect that attorneys reviewing the PDDL for use by a particular data provider will be interested in hearing more about. But just to be clear for all the readers, I’d like to give a few illustrations.
An indemnification clause states that you will pay someone you have a relationship with for the claims of third parties. In the context of the PDDL and the above suggestion, an indemnification clause would be saying that anyone that got the information from the data provider under the PDDL would pay them if someone that the data recipient gave the database to turned around and sued the original provider.
A=Database/data provider, offering data under the PDDL.
B=Recipient of database/data from A under the PDDL.
C=Any third party receiving the database/data from B.
D=Another third party receiving the database/data from C (and so on from D to E to F…)
A gives database/data to B under PDDL. The Section 5 disclaimer and limitation of liability applies because B agreed to it (assuming that this disclaimer is valid in the relevant jurisdiction).
B gives it to C without the text of the PDDL attached, which is permitted because the database/data is in the public domain. B hasn’t agreed with A to the disclaimer in Section 5, and so it doesn’t apply.
C relies on the data, suffers damage because the data is inaccurate. C wants to sue A for providing bad data. An indemnification clause would mean that B pays A for costs/fees/damages/etc associated with the C vs A suit.
There seem to be two primary questions here:
- Can C sue A and consequently what is the liability for A?
- Would an indemnification clause be a suitable means for protecting A?
1. C vs A and A’s liability
This question requires a bit more research and thought on my end, so I won’t comment too much. But from a practical standpoint,
- It seems like if you were presented with public domain data that you’d be pretty aware that the product is “AS IS” and that anyone could have modified it, and that it would be unreasonable to think otherwise.
- If C was wanting to make sure that they had the most up-to-date and clean data, then they’d come to A (assuming A is that provider) who would present it to them with the PDDL and its Section 5 disclaimer.
2. Would an indemnification clause be a suitable means for protecting A?You can in part avoid the first question by putting in an indemnification clause that says whatever liability is there for A (if any), B pays for it. While I think from a practical legal standpoint that something like this might be desirable, I’m not sure how this fits in with the norms of the Free and Open Source Software, and open content movements. If the PDDL is too overreaching then many might reject it on philosophical grounds. I’m also not sure that an indemnification fits into the spirit, if not the letter, of the Science Commons protocol.
It will be interesting to see the discussion develop around the idea of liability. Please post your comments or join the discuss list to contribute.
They have some pretty neat visualizations on here, including the US State of the Union 2008 speech as a tag cloud, a map of the underground economy as a percentage of GDP, a word tree of Alberto Gonzales’s testimony in the Senate, and annual visitors to UK railway stations.
Two issues come up in terms of rights — as this is supposed to be a social site around data, making clear the copyright status of the visualizations and giving users the rights to embed or remix these images would be important. They tangentially mention this in the podcast when the two principal researchers at Many Eyes discuss getting a request to place a visualization up on a website. They should think about making that clear up front.
The second issue is naturally enough, given the nature of the Open Data Commons, the rights associated with the data. Since users upload data, they could ask them to describe what rights the data comes under and then advertise that to the users. Users could also make clear the rights associated with data for datasets that they upload. For public domain data, they could use the Public Domain Dedication and Licence that we have produced. Or, once CCZero has gone live, they could use a CCZero Assertion, which would state that the user believes there to be no IP rights over the data.
You also agree not to submit anyone else’s copyrightable material to alphaWorks Services unless You obtain written permission of the copyright holder to license the copyrightable material to IBM, consistent with the terms of this Agreement.
But the upload page, in the section “Tell us about your data” doesn’t have a rights area. Though it does ask uploaders to give the source and a URL for the source, which could be useful in double checking the origin of data if you wanted to clear the rights.
Just some food for thought. And while you are there, be sure and check out the Size of Scottish Islands by group visualization.
Creative Commons has released the beta version of CCZero (or CC ∅). CCZero implements the Science Commons Protocol for implementing Open Access Data by waiving related intellectual property rights, including copyright and unfair competition. Open Data Commons also implements this protocol in the Public Domain Dedication and Licence with the accompanying Community Norms statement.
The homepage for the licensing tool is at:
There are actually two underlying CCZero legal tools: one waives copyright and related rights and the other asserts that the work has no copyright in the United States.
CC ∅ Waiver 1.0 United States
This legal code is for authors and rightsholder to waive copyright and related rights to the covered work and thus place it into the public domain. It is geared towards the law of the United States and
does not mention, for example, database rightsUPDATE see below. It has a back-up license (like the Public Domain Dedication and Licence) in case a court finds the wiaver invalid.
CC ∅ assertion 1.0 United States
The assertion is for third parties (not authors or rightsholders) to say that they’ve looked into the copyright status of the work and believe it to be in the public domain — in other words out of copyright. The text limits the assertion to US law and includes a clause recognizing that the asserter may be liable for making this assertion.
I’ll be posting more about CCZero and its relationship to the Open Data Commons but initially I’d like to point out that:
CCZero is based on US law and doesn’t include a specific waiver of database rights, though further internationalisation will presumably result in versions that do waive database rights;
- UPDATE — Apologies I initially missed it, but there is language waiving database rights thought they aren’t mentioned by name. The waiver states ” including but not limited to … and any rights protecting the extraction, dissemination and reuse of data ,..”
- The assertion only covers copyright law and not other areas of law, including those covered by the waiver (privacy, moral rights, unfair competition, and so on), though this makes sense in the context of these other rights; and
- The CCZero assertion only covers copyright law in the United States.
Both the Open Data Commons Public Domain Dedication and Licence and the CCZero texts are drafts, and so your comments would be most welcome.
I’ve update the theme for the site — I don’t claim to be a web designer by trade, and I’m aware of some needed tweaks that we will work on in the near future. However the original site was put up quickly and didn’t have trackbacks and comments working on the legal tools (and didn’t take advantage of the latest WordPress features). Any suggestions or offers of help most welcome.
After a short break for the holidays, we’ll be posting and promoting the new drafts more regularly. I hope that everyone had a pleasant time over the holiday break. Happy new year!
2008 is looking like it will be the year of open data. With the release of the Science Commons protocol, the announcement of CCZero, and of course our project, it looks like there will be quite a few options on the table for licensing data in an open way this year. This is after a long time where there were no good options for those looking at licensing data.
Hopefully we will soon release the draft Public Domain Dedication & Licence for use and then we can start getting some feedback from projects making use of the licence and their experiences. With some early adopters, we can quickly start to see some of the benefits of the public domain approach, and maybe some variations on the Community Norms (you are after all free to roll your own).
We will be doing some speaking and presentations on open data throughout the year — details as soon as they are available. One of many will be at this year’s OKCON on 15th of March at the London School of Economics. Details here. If you know of an event you think we should speak at about open data, please contact us with the details.