Copyright and go game records

It has occurred to me that this blog lacks a certain laser-like focus on any one topic of discussion. This probably has a limiting function on my ability to attract regualr readership. Just when I start thinking that I need to set up at least three seperate single topic blogs to serve my need to comment on a diversity of subjects, a story like the following comes along. For the first time ever, I have tagged a post with "go" and "copyright"!

Roy Laird has published this interesting article in the American Go Association e-Journal on recent controversy on the application of copyright to the digital records of go games. (Reproduced in full in accordance with the terms in the AGEJ.)

"How did Kitani play against the variation of the san-ren-sei opening you've been studying? Was there ever a game that used the exact same first ten moves as your last match? Who likes the "avalanche" more, Takemiya or Cho Chikun? Thanks to searchable databases, answers to questions like these are now just a mouseclick away with software like GoGoD, GoBase, BiGo, Smart Go, and MasterGo. The latest entry in the expanding go software industry, Frank de Groot's Moyo Go Studio, has reignited the controversy about whether game records can be copyrighted ("A World of Game Records," here), an intellectual property debate that now rages worldwide as Google proposes to put libraries online and cheap Chinese DVD knock-offs show up on American street corners.

After reportedly paying CyberKiwon $600 for the use of their games, DeGroot, a Norwegian software engineer, is now openly and systematically harvesting game records from the collections of his competitors for his own commercial use, without their permission and against their wishes. On his blog (here), he describes the process of siphoning data from other programs in detail: "As I write this, the games on the latest GoGoD CD are importing into Moyo Go Studio and it looks good - thousands of new games!" And later, "I have calculated that it will take me about three months to export all of (SmartGo's) 30,000 games."

The creators of those co llections are outraged at what they consider DeGroot's blatant theft of their work, having invested thousands of hours (and dollars) in the laborious game-by-game manual entry of game records into their collections. While there's a general consensus that no one has exclusive rights to a game record -- many well-known games appear in all the major collections - the question of whether a specific collection can be copyrighted is still being hotly debated. And beyond the legal issues, there's a more fundamental question of the ethics of taking work without permission or compensation.

De Groot's position on the legal issue is that "There is nothing in a set of SGFs (games recorded in the widely used Smart Go Format) that makes them copyrightable, when there are no added comments." However, according to US law, "A (copyrightable) 'compilation' is a work formed by the selection and assembly of pre-existing materials (e.g. uncopyrightab le facts) or of data that are selected, coordinated or arranged in such a way that the resulting work as a whole constitutes a work of authorship."

Phone numbers, for instance, cannot be copyrighted, but phone books can be, as long as the collector exercises a "minimal degree of creative judgment," beyond mere "industrious collecting." Other types of legal protection are also available; for instance, some programs contain a so-called "shrink-wrap" contract agreement in which the consumer agrees not to reproduce the compilation. (see here for a fuller discussion of the law involved) A directive enacted by the European Union in 1996 also explicitly prohibits "unauthorized extraction of all or a substantial part of the data from a database for commercial purposes" and "unauthorized re-utilization of all or part of th e contents of a database for commercial purposes."

Although the legal issues of the use of game records may be unresolved, there seems to be no disagreement on the ethical question. In a response to criticism of his behavior on rec.games.go, de Groot wrote, "I agree. Still, I am going to do it. It is wrong ethically, I fully agree. But not legally." (To view the entire thread go here.)

Interestingly, despite seeming to take an "information must be free" position, de Groot is encrypting the games he has taken, rather than making them freely available in SGF format, as have the creators of more established programs like GoGoD, GoBase, BiG o, Smart Go, and MasterGo.

Game collection developers who have invested significant resources over the years to build and maintain their collections are worried that de Groot's actions threaten the usability and existence of such collections. The obvious response to sticky digital fingers is for programs like GoGoD and SmartGo to remove the handy feature that allows the user to export game records as sgf files. And if someone can simply take such work product without permission or cost, the go software market - which is fairly limited to begin with -- is undermined and may well force out developers, an obvious loss for the go consumer.

Beyond the legal and ethical issues, the reality is that the go community is close-knit and thus far, the general response to de Groot's (who is not a go player) actions has been fairly negative. Major information sources like Sensei's Library and Gobas e contain no references to Moyo Go; go software link pages (e.g. here) don't mention it and distributors don't sell it. The AGA, committed to the free flow of information, does provide a link to MoyoGo on its Computer Go page -- here -- along with dozens of other go software programs, and will include the program in an upcoming series of reviews of such software. Any references to Moyo Go will note the controversy; as informed citizens of the world go community, we each must decide how to live in that world."

Tags: go copyright