Google already acts as a publisher of tons of copyrighted information without expressed author consent (just click on the “cache” button). I believe, tin-foil hat applies, that the current method of implied-consent of indexing is not sustainable. The time will come, perhaps sooner than expected, that search engines will be legally forced to transition to opt-in technologies. Why not use robots.txt to explicitly allow, rather than disallow? It is a remarkably easy concept.

I am sure plenty of people do as I do: register a bunch of gmail accounts and when you reach your viewing limit of a particular book, switch to another account and continue. I used all my gmail invites for myself for this reason.

But even if an individual user doesn’t circumvent the restrictions this way, google is still copying ALL the content and disseminating (nearly) all the content – they’re just disseminating to different users in different sessions.

You might argue that google is just doing what libraries already do, but the point is publishers already have a working relationship with the libraries, and they often charge more for libraries precisely because they are disseminating to all (eg, journal publishers charge a lot more for library subscriptions than they do for individual subscriptions).

All google need to do to avoid this whole storm is just ask the publishers for permission. That is exactly what Amazon did for search-inside-the-book. Google’s bleating “it’s too hard” just doesn’t wash: if you can index the library of congress you can get on the phone to a few thousand publishers, Given their reluctance to do this, one can only assume that they know most publishers will say no. So they carry on regardless, even though it is not their content.

I find it remarkable how far google apologists are prepared to go to defend google’s arrogant “what’s good for google is good for everyone” attitude, and how they overlook the way google treats the public like we’re morons. Very redolent of another famous software company, only everyone loves to hate them.

What I think the legal outcome must boil down to is the format in which the data is stored and how strictly access to it is controlled.

The problem is, that, no matter how copyrighted material is stored, it is not impossible to engineer a protocol to reassemble the original work in its entirety, given that search results are available in the first place…

Unless the search interface is designed very, very cleverly.

But these mechanics are beyond the comprehension of most of the players in the legal scene, I would guess. It is scary on how little knowledge some people base their convictions and claims.

I research very old technical handbooks. These works–and their authors—have been largely ignored by scholars. Terms like “theatrum machinarum,

I also don’t know of any clause in fair use that stipulates an either-or situation of copying/disseminating like you’re describing – I think that’s a logical leap and the law isn’t that precise in defining those (IANAL as well though). Certainly copy ALL disseminate ALL isn’t allowed, but copy ALL disseminate NONE is allowed, as is copy SOME disseminate SOME (depending on how much). In the case of copy ALL disseminate SOME, which is Gprint, I don’t think there’s precident, other than the above similar cases which are allowed. That’s why it comes down to implementation. Here’s a hypothetical – say you were writing a research report, and you went to the library and copied an article and took it home (allowed under fair use). Does that act of copying the entire article mean that you can’t then quote the article (dissemenating a piece of it) in your research report? I would imagine this is still under fair use.

Frederic – you should totally be entitled to compensation for your work, provided that people want it. This debate about Gprint involves a debate about ensuring that the system of copyright maintains (or restores, or whatever) its place as a system to encourage the creation of new ideas – encouragement in the form of monetary compensation for valuable ideas. I disagree with the notion that you shouldn’t worry about getting paid and that the world should be one big happy place where you write for free – we’ve seen in history that without the possibility for proper compensation from creating new ideas, no new ideas flow.

As an outside observer not in either industry, I can only see it from the perspective of ‘things are changing fast, but the spirit of the law is maintained with these changes’. I do however think that O’Reilly’s really got a good understanding of what is happening, and as someone who is both a forward-thinker who understand the technology AND a publisher who profits from content sales, his insights are especially important in this debate.

Isn’t that just plain infringement? They are storing and reproducing the text for the entire world. Just because a single user only gets to see a portion of the book, doesn’t mean they’re not reproducing the entire thing.

I don’t know of any fair use that allows both public dissemination and full reproduction; you can do one but not the other (eg, you can make full backups for personal use, or you can publically disseminate small portions of a work).

IANAL, but I can’t see how they have any case at all.

Analogy alert! First, the PageRank algorithm itself IS public and has been since 1998 when it was published by Page & Brin, and countless intellectually hungry CS students have poured over it since. But as for application code / trade secrets, they are *not* the same as books – the content of trade secrets ideally is meant to be kept *secret* to maintain competitive advantage to generate income, while the content of books is ideally meant to be seen by as many people as possible in order to generates income. Hypothetically if G wanted to make public the private writings and notes of authors (e.g. their ‘trade secrets’) that they used to then go and write books, then sure G’s trade secrets are up for grabs, but really what you’re saying is totally different so the analogy doesn’t work. You’re right – apples to oranges – and in the case of ‘fair use’ not all IP is created equal, because some IP such as trade secrets are not covered by fair use.

Two things I’ve seen mentioned in discussions about G’s indexing:
1. Once they have the books indexed they can change their TOS and give out the books and anything goes!, and
2. the slippery slope of how much to give out (your 20, 50, 100 pages example – a valid point).

If this goes to court and is held in G’s favor, that doesn’t give them the right to do anything but what is the specific implementation they’re proposing, supported by a nice amount of reasoning by the judges which could be used to guide future cases. The legality of both of these scenarios (esp 1) is in doubt and copyright holders can challenge any changes in court. The slippery slope is an example of where copyright law is not hard and fast, and where careful consideration of intentions and outcomes must take place. Outcomes are important in this case, because that means the courts would have to look at how incentives would change and whether the spirit of the copyright law as a means to incentivize people for their work is being upheld. Hence, yes, you need to recognize the ‘they’ll make money from it’ argument.

As for the burden of copyright holders to enforce – I don’t see the difference between search companies now versus anybody else. Copyright holders have always had to enforce their copyrights throughout history, and new technologies have always brought some new challenges to this (xerox copiers anyone?). To answer your question about authors: What if an author wanted to release a book into the world, but didn’t want libraries to stock the books, or people to quote their books? Where’s the opt-out or opt-in there? The fact is there is no legal requirement for an ‘opt-out’ policy – it’s there for politeness but not for any legal reason. Part of the nature of copyright is that in exchange for protection by the law (for the kind of wholesale copying that does not incentivize the author for their work), authors give up certain control over their content since it has entered the public sphere. The amount of that control is what is being debated, but not the existence or absence of it as I think the ‘opt-in’ argument stems from.

I think it’s a mischaracterization to say that Google is just saying ‘we’re doing whatever we want with other peoples property and we think we can get away with it’. Google is doing what any forward thinking company should be doing – creating new ideas and with them, pushing boundaries and continuing to be a driver of change. There really are no ‘legal niceties’ here, the fact is there’s only the spirit of the law and a mishmash of laws regarding fair use that don’t point to anything specific that says they can’t pursue these ideas.

O’reilly’s wrote a very good op-ed article from NYT on this which is worth a read:
http://radar.oreilly.com/archives/2005/09/ny_times_op_ed_on_authors_guil.html

Let me also add that – - from my teaching days – - I personally know many intellectually hungry CS students in other parts of the world who share the conviction that computer technology is their route to a prosperous future. Their personal benefit will be incalculable once they are able to learn directly from the best CS minds in the world by being allowed to read and review Google’s Pagerank algorithm and other application source code. I eagerly await Mr. Schmidt’s announcement of its publication online (incremental chunks are fine)

Or… wait. Is this just another case like the news.com incident where Google’s principled stand only applies to other people’s data?

Feh. Rank hypocrisy is fast replacing PageRank as Google’s defining feature.

Probable objections to my statement:

• “That’s apples and oranges