Wendy Seltzer wrote: > I'm still not clear on what's happening in 1.1. >> 1.1 It must also be noted that licenses associated with >> feeds or entries using these mechanisms are advisory and >> are not, by themselves, legally binding. Section 1.1 contains two sentences that are, I think, at least in part the result of a telephone conversation on this subject that James and I had about a week ago. However, I think that James has written both sentences too strongly and it is quite possible the first sentence is unnecessary. I'll explain the logic behind my reasoning. First, the first sentence and the second, later. The first sentence is concerned with the binding between the feed or entry and the actual license. For one to rely on a license, either as a rights holder or as a grantee, you need to be able to show what the license actually said at the time that you relied upon it. Because of the nature of the tool being used here, a hyperlink, we must accept that the binding between content and license is a weak and fragile one. There is no guarantee that the content of the license associated with a feed at one instance will be the same as it is at some other instance. Also, there is no mechanism provided to ensure that claims made about what the license was at one instance can be proved at a later time. This weakness of the link is going to make it very difficult for either a copyright holder or one who relies on a license found in a license link to rely on being able to make definitive statements about what the license content actually was. Certainly, there are special cases. For instance, those who link to Creative Commons licenses using the appropriate and presumably well managed URLs that Creative Commons provides are going to be able to much more certain when making assertions about license content since these licenses are only modified in very public and dated events. However, Creative Commons is a special case and can't be generalized to all possible licenses. Thus, while we might find that links to Creative Commons licenses (and others which are similarly well managed) might be effective, it is likely that links to at least some other licenses would not be considered effective. Thus, it might be better to word the sentence conditionally. Something like: "licenses associated with feeds or entries using these mechanisms may not be legally binding"... My personal feeling is that given the customs of the net today, it is likely that the most common licenses referred to in these links will, in fact, be Creative Commons licenses which grant rights otherwise restricted by copyright. The problems come in licenses which attempt to restrict rights. One class of such licenses will be those that claim rights that are already reserved under copyright. Such licenses are really redundant and don't accomplish anything useful. I'll ignore those for now. The most interesting cases will be those licenses that attempt to assert limitations to rights which would normally be considered to be granted to consumers of feeds. Such rights would include things like "Fair Use" and "implied licenses." It is *vitally* important to our community that we ensure that such restrictive licenses are not encouraged or facilitated by this rfc. There are those, I among them, who argue that the mere act of encoding data in a syndication format such as RSS or Atom, posting that content on an openly accessible web site, and doing such things as pinging to publicize the presence of that content, creates a limited "implied license" for others to access and copy the data for the purposes of syndication. This is very much like the implied license to copy HTML into system buffers, caches, network routers, local disks etc in the process of reading it. With HTML, courts appear to accept that you have a right to do something that would be prohibited by a strict reading of copyright law. Under limited circumstances, you are allowed to make copies that are technically necessary and facilitative to the achievement of the content creator's assumed purpose of having you read the pages. Similarly, when you publish to a syndication network, one accepts that there is an implied license to do not only the same kind of copying that is permitted for HTML, but also that you accept that your content will be processed by various syndication-related intermediaries -- feed aggregators, feed filters, feed format converters, publish/subscribe systems, etc. However, please note that this is still a *limited* implied license. The fact that syndication is permitted doesn't mean that the general creation of derivative works, repurposing of feed content, etc. is permitted. The license doesn't allow "stealing" -- it only allows moving the stuff around in the process of getting it to readers. The syndication network can only work because we assume that there is an implied right to syndicate. If, for some reason, we were to establish that this implied right could be revoked or modified by licenses associated with a feed or entry, then we'd probably have to shutdown the entire syndication network until such time as every intermediary and aggregator was modified to support license discovery and interpretation. The social and economic cost of this destruction of one of the most exciting areas of innovation on the net today would simply not be accepted by the courts -- at least not in the USA (I believe). Courts aren't going to accept this "poisoning of the stream" simply so that someone can commandeer the syndication network and force it to process data in a manner different from the way it was designed to work. Of course, one could argue that as long as the Atom license link was used by everyone, then it should be easy for folk to make the adjustments and thus the cost might not be too great. However, once we've accepted that there is even one mechanism by which someone can publish content that overrides the implied license to syndicate, we must ask the question: "In how many ways can the override be expressed?" As far as the law is concerned, there is nothing special about IETF standards. Thus, while the IETF might say that you can use a license link to bind content and licenses, the W3C might decide to create a "rights" link instead. ANSI might then create a "copyright-assertion" link. And someone else might decide to use an element instead of a link. The problem here is that operators of syndication components and aggregators would have no means of being sure that they were catching all the licenses associated with the feeds that they were syndicating. Someone who was malicious might, in fact, create a new "standard," publish content with a restrictive license, and then make a bundle by suing everyone downstream. I think courts won't permit a situation in which the consumer of content has no idea how to determine if licenses are associated with content in the same way that they don't let you rely on "Do not enter" signs written in tiny, unreadable type. The courts would probably insist that legislatures do as they did with the copyright symbol -- make a formal, legal determination of how licenses are to be linked to. (Let's hope that we don't get legislatures involved in the definition of Atom. We've got enough issues to deal with already.) In any case, it may be that "Ignorance of the law is no excuse!" however, ignorance of an optional, experimental extension to the Atom format is certainly not a crime. No matter how earnestly content creators may be in wanting to use the license link to restrict rights, they have no means of compelling anyone to pay attention to or even take note of the links... Even in the case of some system that can be shown to understand the license link mechanism, it can be argued that in the absence of a legally accepted syntax for stating restrictions on rights, that the reader can't compelled to "understand" the limitations expressed. (Note: let's not talk about creating a machine readable license format... Just about all useful methods of doing so are already patented...) In summary: 1. It is unwise to induce folk into believing that the license link can be used to override the limited implied license to syndicate, and 2. There is no way to compel anyone to take note of the links anyway. The situation is, of course, different with things like Creative Commons licenses. Courts are likely to say that any publisher who includes the appropriate namespaces in their feeds and includes properly formatted license links that point to Creative Commons licenses is demonstrating a real knowledge and appreciation of the mechanism and is willingly granting licenses to all those who share the same understanding of the mechanism and willingness to use it. Basically, the publisher proves, by making the effort and signaling via the published and controlled syntax, that they do, in fact, wish to grant the rights expressed in the license. Readers of feeds are then free to exploit their understanding of the publisher's signals to exercise the rights granted -- just as they are free to refuse to understand the restrictive licenses published by others. Thus, it would seem that the only effective use of the license link is to grant rights not to restrict them.
The second sentence in 1.1 is: >> Nor can a license associated with a feed or entry >> restrict or forbid access to, redistribution, aggregation, >> caching and display of those items by third party >> intermediaries such as search engines and so-called >> "online aggregators". This second part of 1.1 is stating support for the theory that the act of publishing data in the Atom format creates an "implied license" for the limited purpose of syndication and lists a number of processes which are considered to be part of the syndication process. Hopefully, my discussion of the first sentence explains what this is all about. My only suggestion for this sentence is that it might be less strongly worded. Given that the law in this area is not settled, it might make sense not to say "Nor can a license... restrict..." Rather, it might be more accurate to say something like: "It is believed that a license ... cannot restrict...." My apologies for such a long message... bob wyman