Do you think the same standard applies on the web? That is, should Google ask PFF for permission before indexing IPCentral?

Tim, there are ways that web sites can prevent search indexing. Obviously, IPcentral has opted not to use this feature, thereby allowing Google and others to index the site.

Your analogy between books and web sites isn't quite right. Many authors who release books do so with no intention of releasing them on the web. Obviously, the case is different with web content.

Google has also given publishers ways to prevent their books from being indexed. The case isn't about whether publishers have a right to exclude their books from Google's index. It's about whether Google has to ask first or not. On the web, you don't have to ask--if the publisher hasn't notified you that indexing is not OK (via a robots.txt file, in most cases) it's presumed to be OK to index (and cache, for that matter) your content. Google's taking the same position with Google Book Search: they assume it's OK to index the books, but if publishers request that they be removed from the index, Google honors the requests. I don't see the difference.

Tim, as long as Google honors the wishes of copyright holders, then fine. Its also important that Google has some copy controls in case copyright holders are not yet aware their books are indexed.

So given that Google meets both of your criteria, does that mean you side with them?

I think MSFT takes the more responsible approach.

Tim, to your first point, IPCentral is on the web, so Google doesn't have to do anything to index it. None of my mother's novels are on the web; Google would have to *copy* them in order to have a digital copy to index, but they *still* wouldn't be on the web unless Google put them there. Apples and oranges.

On the second point, should I be able to drive your car if I need to, unless you tell me explicitly beforehand not to? I know you'd like to see copyrights diluted if not eliminated, and moving from our current opt-in system to an opt-out one ensures that the *rights* no longer have legal meaning. An author would have to track not just Google but any potential copier in the world, including the Show-Me Institute (which many authors might not be aware of).

"IPCentral is on the web, so Google doesn't have to do anything to index it."

This is false. From a computer science perspective, Google's web search engine and its book search engine do exactly the same thing: they retrieve a copy of whatever is being indexed and store it on their servers. The difference, obviously, is that in one case the copy is being retrieved via the Internet, and in the other case it's being retrieved via a scanner. But in both cases, a copy is being made.

And Google is not putting copies of books on the web. They are offering an index that allows users to find books they might want to read--precisely the same thing they do with their search engine for the web. Like its web search engine, Google Book Search displays small excerpts of each book to give the user a context for the search. But if you want to read the whole book, you have to buy it from the publisher or check it out from the library.

So it still sounds to me like your argument applies just as well to Google's search engine for the Web as it does for its search engine for books. Google has a copy of IPCentral stored on its servers, which it created without the permission of PFF. If making copies without permission is always copyright infringement, then Google's search engine for the web is massive copyright infringement.

Tim, neither Patrick nor I accused Google of violating fair use or other copyright doctrines. In fact, I believe Google books may constitute fair use but I have to review to what extent Google is providing digital copies of the books.

Back to the point between web sites and books. Web site owners know that their sites will be indexed, and its easier for them to protect themselves by setting up technical guards against Google or other entities scanning sites. With book owners, its a different story, as they have no "automatic switch" and may incur substantial *cost* in preventing such indexing (if they know about it at all).

Notice I said *cost* which is a concept you, Tim Lee, think should be pushed onto copyright holders rather than Google.