There’s been a fair bit of buzz, both on some edublogs I respect and in the Sakai listserv, about Thomson Reuters suing George Mason University, alleging that the Zotero team illegally reverse engineered EndNote. My initial reaction to Thomson’s move was very negative and my final reaction may be equally negative. Even casual readers of this blog know that I am not a big fan of using intellectual property as an anti-competitive bludgeon in an already thin educational software market. I’m also not a big fan of the DMCA, a draconian law that goes way too far in an effort to stop piracy.
However, I am withholding judgment at the moment because, after reading Thomson Reuters’ complaint, I think that it actually may have significant merit.
The dispute is not around Zotero’s functionality in general but a specific feature that allows Zotero to import EndNote’s proprietary file format. More importantly, it is about the ability to import the style files that enables EndNote to format references from particular journals properly. If George Mason University had gone about its merry way building Zotero with every feature it has today except the ability to import this information from EndNote, there would be no suit.
What’s the big deal about this import feature? In their complaint, Thomson Reuters alleges that they have created over 3,500 publication import styles for EndNote while the Zotero team has only created 15 styles. So it’s not just an issue of the Zotero team copying a feature of EndNote. The specific feature they copied enables users to steal the labor that Thomson Reuters invested in creating an importer for each one of those many journals. That seems wrong to me.
I’m going to think about it some more, but my reaction at the moment is that the Zotero team should remove that importer feature, negotiate a settlement with Thomson Reuters, and compete honestly by finding a way to crowdsource the creation of their own import style files for various journals. If a company like Thomson Reuters is investing many person/hours into creating features that benefit academics, it is in nobody’s interest to discourage them—and, more importantly, other companies that may follow them into the market—by simply stealing the product of their labor. I’m not talking about “stealing” an idea about what features would be beneficial in a software category. I’m talking about stealing the actual work product in the form of 3,500+ laboriously created import format files. I may be opposed to educational software patents, and I may support Creative Commons and OER and all that good stuff, but I still believe in copyright laws. The last time I checked, even icons of openness like Richard Stallman and Lawrence Lessig do too.
Sakai maintains a number of lists. Can you say which specific list has discussed this?
I don’t know if I buy that–Thomson Reuters has sent them other notices to cease behavior, such as to not use the “EndNote” trademark to describe files with the application/x-endnote-refer mimetype. I am sure that TR would have found something to sue over.
This is a false claim, though. Many EndNote styles are duplicates with separate names. Many are created by users or publishers. Zotero has a lot of styles that were created independently of .ens files. Zotero had the ability to read and use .ens style files, but the Zotero devs avoided making it generate stand-alone .CSL files (despite what the complaint alleges).
In addition to the styles distributed with EndNote (many of which were not made by Thomson Reuters, users can create their own style files. If style files can have a copyright (and that’s a big if–they are not novel & are trivial descriptions)), I should have the copyright on style files that I create and be able to do what I want with them.
Interoperability is a big deal. RMS would certainly want it. The OpenOffice.org developers and users would certainly want it. Even the DMCA grants exceptions to the “no reverse engineering” laws for compatibility.
If there was no site license agreement that may place extra restrictions on GMU, Thomson Reuters would have absolutely no case (rather than a thin one).
The feature has already been disabled. I hope that it is temporary.
Never kowtow to a bully. This isn’t the first skirmish TR has created. It probably wouldn’t be the last. And TR probably has a weak case.
As above, this is already happening.
As above, neither of these has really happened.
Michael,
You’re right. There is a line between open source and piracy. Open source is a collective work done by the users of a system to serve the users of a system. If it’s using work from another system to cut the corners without consent, and that it might jeopardize the work of hundreds of employees because of a fraudulent tactic, it should be treated as piracy.
Hey Michael, I’ll admit to having no patience left for any of these lawsuits, period, so it may well turn out that I overreacted. But the argument doesn’t hold much water for me – the claim isn’t that Zotero reverse engineered the 3500 publication import styles – that would seem to be an obvious case of copyright infringement. As far as I understand, the claim is that Zotero allows for the importing of citation records into Zotero from Endnotes proprietary .ens files. Maybe this will be seen as also infringing on their IP. But they are not creating functionality for themselves on the backs for Thompsons’ investment; instead, as the lawsuit states, they are threatening Endnotes user base by helping them to not be locked in any more to a proprietary format. Thompson will claim, and the courts will likely back them, that creating a proprietary format for storing data isn’t illegal. And it’s not, as far as I know. But adopting products that use them is clearly not a great move, in the long run, and if the courts continue to uphold private companies’ rights to bully open source projects into not providing ways to transition out of these formats, we’ll be stuck with the current situation of ad hoc proprietary standards ruling the day, or innovation dragged down by attempts to derive open standards in advance of the market existing (witness the mess of the instructional standards world). I don’t have the answer, but this *is* “intellectual property as an anti-competitive bludgeon.”
Thomson Reuters is not claiming that anyone illegally stole or distributed the styles, since that didn’t occur. In fact, my understanding is that many or most of the styles included with EndNote were created not by Thomson Reuters, but instead by enthusiastic users. The goal was to allow Zotero users switching from EndNote to use the styles they had already bought from Thomson Reuters or created themselves. But, Thomson Reuters wants to protect their proprietary file format with a license agreement, so that even if I were to make my own style with EndNote, it would be illegal to use it with any other application, although it supposedly belongs to me. If a court were to rule in Thomson Reuters’ favor, it would mean that Microsoft could put a provision in their license agreement that could prohibit users from opening their Word files in another application, thus ensuring that those who had already invested significant time in creating Word documents could never switch to another word processing application.
BTW, the “15 styles” number is not even correct to an order of magnitude…there are 1000+ Zotero styles, and I believe that there have been 50-100 available for the last six months. Thomson Reuters seems to have fudged things a little bit.
@Robert,
Thomson Reuters’s filing does make the claim that Zotero has distributed styles derived from EndNote. But you are correct that this never occurred.
Scott, I think your interpretation of the allegation is a little different than mine. My understanding is not that Zotero reverse-engineered the 3500 import files, but that the reverse engineering of the format facilitated the copying of those files by users in violation of copyright. If this is correct interpretation, then Thomson Reuters would be using the anti-reverse engineering clause in their license contract to go after GMU because that’s a lot easier than going after the million billion end users for copyright infringement. I don’t like proprietary formats either, but the ends don’t necessarily justify the means. If you want an open format, then create an open format and do the work necessary to get the import files in that format.
Now, there’s a question of fact that’s separate from the questions of law and ethics. Robert claims that Zotero does not, in fact, enable users who have not licensed EndNote to make use of EndNote import templates. If he is correct, then it starts to look more like Thomson is just using the lawsuit to scare GMU out of continuing to develop Zotero. Which we all should strenuously oppose as consumers, even if Thomson’s actions are within their legal rights.
I don’t think this is straightforward, which is why I’m very cautious in my language right now. I’m trying to reserve judgment.
@Michael
I had a fairly lengthy post here that refuted some of your claims. Perhaps it is awaiting moderation, as I included HTML links and blockquotes in it? Or perhaps there was some other problem. I will avoid HTML in this post.
Yes, Thomson Reuters has argued a breach of contract, not of a copyright violation. Reverse engineering may be allowed for compatibility under copyright law, but GMU may have given up this right in their site license agreement.
Zotero made a feature available in a beta version of their software to use the undocumented EndNote style files as-is, without creating a stand-alone CSL file (n.b. Thomson Reuters got the facts wrong in their filing). This feature has since been disabled. This is somewhat disappointing: many EndNote users have created their own EndNote style files. It would be difficult for Thomson Reuters to claim that the company, rather than end users owned such user-created content. However, this (hopefully temporary) discontinuance of supporting EnsNote style files may be beneficial: more people are generating more CSL files from scratch.
Sorry about the posting problems, Rick; your comment was indeed swallowed by my moderation system (a system with which I am not particularly happy in general, but that’s another story).
I appreciate the additional information you’re providing and would like to elevate this to top-level attention in a follow-up post. Are you a Zotero developer? Do you have direct experience with the project that gives you first-hand knowledge of the situation?
Sorry, I forgot to mention that the conversation I referred to in the original post is in the Sakai advocacy list.
I am not a Zotero developer. If I was, I would probably not be making public comment. I am an occasional Zotero user who has submitted a patch or two and is fairly active on their forums. I had been an occasional EndNote user as well, but my license has expired in the past year. I had planned to renew it, but no longer will because of this case. I am also one of the lead developers of a free/open source web-based reference manager. I follow the development of reference managers (including both Zotero and EndNote) fairly closely.
Thank you for the heads up on the Sakai list.
Given the open questions here and on the Sakai list, I thought I’d give you a couple of links.
Zotero development takes place in public, so it is trivial to see that at least some of the Thomson Reuters claims are false.
This changeset add support for the undocumented EndNote style files. After a long discussion, it was decided
Sorry–broke the quote in the above post.
The decision was:
The discussion seems to be developing well enough without me inserting myself gratuitously, though I do want to reflect on one aspect that strikes me as more than tangential. I also would like to note that I agree with Robert’s and Rick’s observations, particularly Robert’s Microsoft analogy. My point draws on Michael’s earlier comment: “a company like Thomson Reuters is investing many person/hours into creating features that benefit academics.” Sorry; TR clearly isn’t interested in the benefit of anyone but TR. Moreover, they appear to be aggressively trying to ensure that only TR benefits. Thanks for an insightful discussion, everyone.
I’m the creator of CSL, the open format that TR alleges Zotero converts Endnote style files into. So I have knowledge about the background here.
There are two issues here. First, there is the question of the legal (and, as you wish to explore, ethical) issue around the ability to read Endnote style files. I really don’t think there is any basis to argue this either legally or ethically wrong, any more than it would be to say its wrong for any word-processor other that Microsoft Word to read .doc and .dot files. To allow a company to claim ownership in this way would be clearly anti-competitive, and anti-consumer. TR has no leg to stand on.
The second is about the copyright issues that relate to particular file instances. I really don’t want to get into a huge discussion about this because it’s irrelevant here. TR has made claims here which are demonstrably false; Zotero does NOT distribute Endnote-derived style files. All CSL files in existence have been created entirely independently of Endnote.
So my point here is these two issues are completely orthogonal, both legally and ethically, but that you seem to be collapsing them in an unhelpful way.
Re your phrase “steal the labor”:
Investment of labor is not a guarantee of return. The fact that somebody put in a lot of hard work does not mean they should be unquestioningly handed a monopoly on the results of that work — especially when the results can be shared at zero cost, as is possible on the Internet.
The traditional justification for handing out that monopoly is not a moral one (i.e., not “Someone worked hard so they should get paid!”) but rather a systemic one: if we don’t hand them that monopoly now, then others may be less motivated to do similar work in the future.
That’s an interesting idea, but it’s also testable, and so far it’s not looking like it really holds up. Which makes sense, considering that these monopolies (copyrights, that is) were originally invented not because of the systemic motivation described above, but for an entirely different reason: to subsidize the up-front costs of distribution in the pre-Internet era. We’ve romanticized the history a lot since then, which is probably one reason the word “steal” is often used even though “copy” or “compatibly match an interface” would be more correct.
Thomson Reuter needs to find a business model that doesn’t rely on restricting other parties from constructing compatibility interfaces, that’s all. If they don’t find such a business model, then they’ll go out of business. This is what I would expect to happen to me if I ran a business based on faulty assumptions about my ability to restrict what others do, and there’s no reason that shouldn’t hold true for TR or anyone else as well.
(They might argue that the rules have been changed on them in the middle of the game, and to some degree that’s true… But the Internet is hardly news at this point, and the rules changed for everyone else too. You can’t sue the future.)
I hadn’t checked up on Endnote in many years. Took one look at their pricing — oh my God, two hundred and fifty dollars or more! — and said, fuck them. If they think they can charge rapacious prices today for work they did years ago, they have another thing coming.