___________________________________________________________________ Date: Tue, 17 Jul 2001 16:32:40 EDT To: oai-general@oaisrv.nsdl.cornell.edu From: Terry Kuny Subject: [OAI-general] OAI and intellectual property issues I have a question about OAI and its relationship to intellectual property concerns that I sense might arise - that is, if not beaten back with explicit legal disclaimers on the part of OAI data providers. This is not a hypothetical interest. One of the most important things about OAI (to my mind) is *** that it explicitly promulgates a principle I call "metadata *** transparency" - that organizations who want to make their information accessible to as broad a range of users as possible will make metadata accessible for harvesting by a potentially equally large range of organizations that want to build interfaces to these contents. This is very appealling and has great applicability for many communities. *** The problem I am wrestling with is that what happens *** if an organization builds an OAI repository and then *** makes a claim that they hold the intellectual property *** rights to the repository? Under recent WIPO treaties, this would seem to be a legitimate claim since aggregates of fact or data like this can be considered "original works". If this is so, might it not stop other organizations from aggregating the same data (perhaps to build a competing data service with some prettier interface or better whizbangs)? I know this goes counter to most definitions of "open" and that really a repository built upon metadata made freely accessible SHOULD not do such a thing, but what protection is there for a data provider to ensure that such as situation does not occur? *** Should OAI have some sort of GNU-like license that should *** be signed onto that would preclude service providers from unfairly exploiting open archives to the detriment of others? From the OAI perspective, what IP would a service provider have claims to? I look forward to your comments with interest. Regards, -terry ___________________________________________________________________ Date: Wed, 18 Jul 2001 00:05:08 BST To: "Terry Kuny" , From: "José Luis Borbinha" Subject: RE: [OAI-general] OAI and intellectual property issues You raise a very relevant issue, but it might be not a real problem in the OAI scope... The idea behind OAI is not to consider "the metadata as THE resource", but to raise the awareness about the resource that the metadata describes. In that sense, the value is in the resource, and not in the metadata that it announces. IMHO, the organizations willing to put their metadata available trough an OAI interface must be aware that such metadata will be *** "lost" for the public domain. In fact, they should be happy if *** someone else will produce added value on the top of it and makes money with that, since the target is the resource that the metadata describes... Regards, José Borbinha PS: Maybe this is more like distributing Linux flavors, where the kernel is always the same, and is free, but you can sell something more with it, and everyone wins... ___________________________________________________________________ Date: Wed, 18 Jul 2001 05:48:59 EDT To: OAI-general@oaisrv.nsdl.cornell.edu From: Thomas Krichel Subject: Re: [OAI-general] OAI and intellectual property issues Terry Kuny writes > The problem I am wrestling with is that what happens > if an organization builds an OAI repository and then > makes a claim that they hold the intellectual property > rights to the repository? RePEc does not run an OAI repository as yet, but *** we have the firm intention to set one up AND to *** claim intellectual property rights to the data *** therein. > Under recent WIPO treaties, this would seem to be a > legitimate claim since aggregates of fact or data like this > can be considered "original works". If this is so, > might it not stop other organizations from aggregating > the same data (perhaps to build a competing data service > with some prettier interface or better whizbangs)? Everybody who does not like the usage conditions attached to the RePEc data can build a competing one. > I know this goes counter to most definitions of "open" How come? Our collection has usage conditions, but is freely available within those conditions. > and that really a repository built upon metadata made > freely accessible SHOULD not do such a thing, Why not? Imagine that you built for 8 years a digital library that is arguably the largest distributed academic digital library in the world. You and your mates have been working on it for nothing essentially, at the margin of regular jobs, in overtime spent at the computer in the office rather than with your girlfriend at home. *** Then one day, out of the blue. you find that a commercial company *** have built a web site, have used all your data and claim *** that *they* have collected it, without a single mentioning *** of your efforts. I bet you would be just as outraged as I was when exactly that thing happend to RePEc six weeks ago! It cost me a hell of a lot of nerves and a bit of effort to put things right. > Should OAI have some sort of GNU-like license that should > be signed onto that would preclude service providers > from unfairly exploiting open archives to the detriment > of others? Sure there should be some thought on that. In fact, at the original Santa Fe meeting I presented a framework that would address this issue. > From the OAI perspective, what IP would a > service provider have claims to? I would that think that these would be rather small and difficult to defend in court. But I am by no means a legal expert. Cheers, Thomas Krichel mailto:krichel@openlib.org http://openlib.org/home/krichel RePEc:per:1965-06-05:thomas_krichel ___________________________________________________________________ Date: Wed, 18 Jul 2001 06:09:17 EDT To: OAI-general@oaisrv.nsdl.cornell.edu From: Thomas Krichel Subject: Re: [OAI-general] OAI and intellectual property issues José Luis Borbinha writes > The idea behind OAI is not to consider "the metadata as THE > resource", but to raise the awareness about the resource that the > metadata describes. In that sense, the value is in the resource, > and not in the metadata that it announces. Some collections made available through OAI may want to mix metadata about resources with data on non-resources. That is definitely the case in RePEc. > IMHO, the organizations willing to put their metadata available > trough an OAI interface must be aware that such metadata will be > "lost" for the public domain. There is no hope that my community will agree to that. See my previous mail. > In fact, they should be happy if someone else will produce added > value on the top of it and makes money with that, since the target > is the resource that the metadata describes... *** We are not happy. We are happy to give the data away *** for free only if it is not sold or included in a product *** for sale. Cheers, Thomas Krichel mailto:krichel@openlib.org http://openlib.org/home/krichel RePEc:per:1965-06-05:thomas_krichel ___________________________________________________________________ Date: Wed, 18 Jul 2001 10:05:23 EDT To: Terry Kuny cc: oai-general@oaisrv.nsdl.cornell.edu From: "Michael L. Nelson" Subject: Re: [OAI-general] OAI and intellectual property issues Terry said: >Should OAI have some sort of GNU-like license that should >be signed onto that would preclude service providers >from unfairly exploiting open archives to the detriment >of others? From the OAI perspective, what IP would a >service provider have claims to? *** actually, the rights relationship that SPs may have to DPs is (in a *** sense) outside the scope of the protocol. take a look at: http://www.openarchives.org/OAI/openarchivesprotocol.htm#Identify *** specifically, the example at the end of that section. it is the *** responsibility of the DP to advertise its metdata use policy through *** the Identify verb. the eprints.org folks have introduced an "eprints" namespace to help delineate some of the issues. note that this namespace is optional -- this is just a convenient way to provide pointers to a policy that will be defined "outside" the protocol. *** its up to the SP to respect the policies that the DP advertises. Jose said: >The idea behind OAI is not to consider "the metadata as THE >resource", but to raise the awareness about the resource that the >metadata describes. In that sense, the value is in the resource, >and not in the metadata that it announces. that's not *necessarily* true (though so far, has been true in practice). without sliding into the "data" vs "metdata" issue, its entirely possible and appropriate that a DP might wish to restrict the use of their metadata (as per Thomas' example). "Identify" gives DPs/SPs a way to announce and discover, respectively, this policy. How all of this actually plays out is hoped to be uncovered in the initial 12-16 months (starting Jan 01) of the official unveiling of the OAI protocol. regards, Michael ___________________________________________________________________ Date: Thu, 19 Jul 2001 01:33:17 BST To: From: "José Luis Borbinha" Subject: RE: [OAI-general] OAI and intellectual property issues If your data is available for free, even under terms and conditions for usage, how is a commercial company going to make money just exploiting a web site with it? (well, in fact AltaVista and Google have been doing that, as other metacrawlers have been doing with them, and so on... so I'm afraid that we are all contradicting ourselves...). On the other side if you've built a digital library, that should mean that you have collected one or more collection of resources, and that you developed your metadata to describe these resources. And it is this metadata that you put available through OAI, not the resources (which can be movies, texts, or more and richer metadata...) A good thing that I see in OAI is its simplicity and button-up *** approach. It'd be sad to see it intoxicated by its possible *** success and jump in complicated processes and visions like we had for Z39.50, for the top levels of the ISO model, etc... I was aware that I was entering in dangerous areas with my previous email (I'm not so naive neither libertarian...), but I think that this is in fact a very interesting issue to discuss, since it raises the core problem of the scope of the OAI!!! Very often nice initiatives have been failing nowadays because of the lack of clear and understandable objectives! They start as exploratory activities, in the hope that the light will come in one of the next coners, but in the end they simply get lost in messy and confuse purposes... So, for the future, I'd like to post here my position about this: *** please lets try to keep the model simple, and assume that what we *** put available through a public OAI interface is just structured *** data for public usage, available for usages in meta services just like AltaVista and Google have been doing with our sites. This doesn't excludes the usage of private OAI networks for more restricted materials/purposes/data, but lets not afect the OAI work with these requirements for now, or at leat until we have it really imposed (which can be the case, but it we are not there yet...). But of course that you all can disagree of me... :-) Regards, Jose Borbinha ___________________________________________________________________ Date: Wed, 18 Jul 2001 20:04:41 MDT To: oai-general@oaisrv.nsdl.cornell.edu From: Thorsten Schwander Subject: RE: [OAI-general] OAI and intellectual property issues Hi, for an actual example how a wonderful open project, a metadata collection that has been created by a community effort, a shared public resource, can be turned into a quagmire of heavy contracts, licensing fees, forced user registration and anti-competition clauses I want to point out the fate of the once free cddb, now Gracenote http://www.gracenote.com/terms.html . There are threats, cease and desist letters to developers of alternatives, and even lawsuits. Some quotes from news-services Basically cddb is a online lookup service for music CD titles and songlists that provides this information to CD playing software. The database was build by individual users contributing their manually generated datasets to the database -- i.e. the users typed in the songlists. "Today when I attempted to rip a new CD I bought, I received this message. "Your CD player application is either not licensed to use the Gracenote(tm) CDDB(tm) service or its license has expired. If you are unsure what this means, please see our web site at http://www.cddb.com/lic/Grip. If you are a developer and feel you have received this message in error or wish to get your application licensed, please contact support for assistance."" "CDDB is refusing connections from Media Jukebox until the Media Jukebox guys sign an "exclusive agreement" to use CDDB's database." and http://www.gracenote.com/press/2001051000.html BERKELEY, CA (May 10, 2001) -- Gracenote (formerly CDDB) a content delivery platform provider specializing in music recognition services, today announced a lawsuit against Roxio, Inc., at present, a wholly-owned subsidiary of Adaptec, Inc. (Nasdaq: ADPT) for breach of contract, patent infringement, trademark infringement and other violations of Federal law. While not entirely the same I think there are some obvious analogies here, and it goes to show that as soon as commercial interest are involved almost nothing is inconceivable. *** Could a OAI service provider, who has built a rich collection of metadata via *** OAI, claim exclusive rights to that collection. Could they prevent others to *** build a similar collection or provide similar services and functionality? Cheers Thorsten ___________________________________________________________________ Date: Wed, 18 Jul 2001 23:13:45 EDT To: oai-general@oaisrv.nsdl.cornell.edu From: "Michael L. Nelson" Subject: RE: [OAI-general] OAI and intellectual property issues actually, to continue the analogy... assuming that a SP attempting to create a Gracenote/CDDB situation, we can probably count on another SP to come along, harvest the same DPs, and provide similar services -- http://www.freecddb.org/ *** within the OAI context, we can handle rogue SPs. now if a DP collects *** eprints from users, and then signs an exclusive license with a SP... well *** that's a whole different ball game. regards, Michael ___________________________________________________________________ Date: Thu, 19 Jul 2001 07:35:21 EDT To: Thorsten Schwander cc: oai-general@oaisrv.nsdl.cornell.edu From: Thomas Krichel Subject: Re: [OAI-general] OAI and intellectual property issues Thorsten Schwander writes > While not entirely the same I think there are some obvious analogies here, yes! > and it goes to show that as soon as commercial interest are involved almost > nothing is inconceivable. Sure. But had the RePEc rip-off really ended up in a court case, I would have turned to this community saying that I need $X to fight it. And I am convinced that I would have gotten a response. After all, we can not be sure whom it will hit next. Cheers, Thomas Krichel mailto:krichel@openlib.org http://openlib.org/home/krichel RePEc:per:1965-06-05:thomas_krichel ___________________________________________________________________ Date: Thu, 19 Jul 2001 06:24:21 PDT To: Thomas Krichel cc: Thorsten Schwander , oai-general@oaisrv.nsdl.cornell.edu From: billn Subject: Re: [OAI-general] OAI and intellectual property issues I believe that the IP issue here is real and will cause problems unless OAI sets a standard disclaimer that says (essentially): *** "You are permitted to harvest metadata from this site for the purpose of *** access and/or providing a service for others. You may build services *** with metadata and claim IP rights for the metadata *plus* service, but *** you may not prevent or interfere with others who harvest metadata even *** for the same purpose. Only the created service can be protected by IP, *** not the metadata or access rights." While this probably leaves loopholes a lawyer could drive a semi through, OAI needs something that establishes its *intention* in the legal arena. I imagine a call to the EFF could provide some help with the wording. This style should cover the use of public and free access sites, but some additional restrictions may be added where a private company who invested large sums to develop the data requires a license and/or payment for access to the underlying data. Charging for the metadata would be counterproductive, although they may require registration first. *** This issue will not go away if ignored. It will simply create a quagmire *** of different legal verbiage and restrictions that could cripple the *** potential of a great concept. Please give serious consideration to *** establishing your IP standards. Bill Nicholls Advanced Software and Technology, Byte.com ___________________________________________________________________ Date: Thu, 19 Jul 2001 10:31:02 EDT To: billn cc: Thomas Krichel , Thorsten Schwander , oai-general@oaisrv.nsdl.cornell.edu From: Herbert Van de Sompel Subject: Re: [OAI-general] OAI and intellectual property issues I have been following this interesting discussion from a distance (while packing our stuff for our move to the UK). A lot of really important issues have been raised. In this mail, I would like to add a few elements to the discussion: * First, the "open" in Open Archives Initiative doesn't refer to the concept of "open access". It refers to the fact that repositories have a machine interface that allows third parties to harvest metadata from the repositories. see the OAI FAQ at http://www.openarchives.org/faq.htm under the heading "What do you mean by 'Open'". Even at the time of the Santa Fe convention of the OAI (the predecessor of the current metadata harvesting protocol, which focused on open access preprint literature), the notion of "open" meant "open in an architectural sense". see for instance the paper on the Convention (http://www.dlib.org/february00/vandesompel-oai/02vandesompel-oai.html) and the Convention itself *** (http://www.openarchives.org/sfc/sfc.htm#openarchives). As such, it is *** fair to say that the OAI protocol doesn't make any statement regarding the *** business models under which the full-content described by metadata in *** repositories is available. *** * As already mentioned in this discussion, the OAI protocol (as opposed to *** the Santa Fe Convention, I guess) does not address issues of acceptable *** use of harvested metadata: which service provider can do what with *** harvested metadata?". At the Cornell meeting (september 2000) where the *** foundations for the OAI protocol were agreed upon, an explicit choice has *** been made to hand over acceptable use issues to communities implementing *** the OAI protocol. Having said that, the protocol does explicitely allow *** for the inclusion of an "about" container attached to each harvested *** metadata record. Typically such an "about" container could be used to *** specify the terms and conditions of the usage of a metadata record. The *** very first example in the protocol document actually shows an example *** where this is done. As such, it allows individual communities to express *** terms and conditions regarding metadata use at the level of individual *** records. In addition to that, at the level of a repository, the response *** to the Identify verb allows for the inclusion of an open-ended *** "description" container. Again, communities could use this container to *** include terms and condition information for all metadata records in the *** repository. *** So, I think that from a technical perspective, the hooks are there to *** allow communities to specify terms and conditions re the usage of metadata *** harvested from their repositories. Then again, the issue raised by this *** discussion (in my opinion) is whether it would be a good idea for the OAI *** to provide a "default" for terms and conditions at the level of individual *** records and at the level of repositories, which could be overwritten by *** communities with diverging needs. The problem I see with that is that it *** would probably mean that the OAI would have to take some kind of position *** in the IP debate, which I think the OAI has been avoiding since the moment *** of the move from the Santa Fe Convention to the OAI Protocol. The latter remark is by all means an expression of a personal perspective of the situation, and it should not be regarded an official statement of the OAI. many greetings herbert van de sompel ___________________________________________________________________ Date: Thu, 19 Jul 2001 10:53:47 EDT To: Herbert Van de Sompel cc: billn@yelmtel.com, krichel@openlib.org, oai-general@oaisrv.nsdl.cornell.edu, schwander@horse.lanl.gov From: zubair@cs.odu.edu Subject: Re: [OAI-general] OAI and intellectual property issues I wholeheartedly agree with the view Herbert is taking on the IP issue. *** Taking any stand on the IP issue for OAI would be a mess. Once the data provider has the "technical" capability to communicate the terms and conditions, then it is up to the data provider and the service provider to see that the terms and conditions are followed. The OAI should not be in the picture. Zubair ___________________________________________________________________ Date: Thu, 19 Jul 2001 08:50:06 PDT To: oai-general@oaisrv.nsdl.cornell.edu From: billn Subject: Re: [OAI-general] OAI and intellectual property issues *** My suggestion was to create a standard disclaimer in order to provide *** some basic simple structure to OAI metadata harvesting. It does not *** preclude anyone who creates metadata from their own raw data from adding *** restrictions or completely rewriting the IP restrictions for that *** particular database. *** My reasons for recommending this course of action are: *** 1. For those metadata providers who do not have laywers on staff, a *** working legal disclaimer, or a place to start. *** 2. A base for *consistent* IP standards for OAI metadata. Get it right *** and many sites will follow that base, making access and presentation of *** the data *much less complex* from the IP POV. *** 3. It is much easier to set up a consistent IP approach in the beginning *** than try and straighten out the mess that will happen if every provider *** goes their own way with IP protection. Can you imagine the complexity of *** dealing with hundreds of different IP restrictions? As a programmer, I'd *** run screaming in the other direction. *** 4. Nothing I proposed prevents any site from using a different IP setup, *** and I'm sure commercial sites sharing data want better protection for *** the underlying raw data. *** 5. IP is guaranteed to be an issue. If we avoid dealing with it now, *** even in less than perfect form, we will reap the results of our dodge *** later with much pain. I believe that IP chaos could stifle the the *** potential benefits of the whole OAI concept. *** 6. OAI set standards to make access to and the use of metadata uniform *** and easy. While IP is not a programming issue, it *will* control how we *** access, present and use both metadata and raw data. Without some *** coherent IP structure, the only people who will benefit are the IP *** lawyers. *** There is no doubt overlap in those items as I am writing this 'off the *** top'. Even though IP does not appear to be essential to the standard, it *** has such potential to add complexity to both metadata and code that we *** should not ignore it. Bill Nicholls ___________________________________________________________________ Date: Thu, 19 Jul 2001 08:55:05 PDT To: oai-general@oaisrv.nsdl.cornell.edu From: Garey Mills Subject: [OAI-general] Re: OAI and intellectual property issues Hi - *** I agree with Bill Nichols on all of his points about the *** advisability of setting up a default IP policy, but I think the point is *** that this is a job for the community of implementors and potential users. Garey Mills Library Systems Office UC Berkeley ___________________________________________________________________ Date: Thu, 19 Jul 2001 17:28:13 BST To: billn , oai-general@oaisrv.nsdl.cornell.edu cc: c.oppenheim@lboro.ac.uk From: Steve Hitchcock Subject: Re: [OAI-general] OAI and intellectual property issues Expert interpretations seem to be hard to find. Charles Oppenheim has written about this, but I can't find anything online to link to. I guess this directive would give Thomas' RePEc database some protection. The point about preventing others from building similar services using OAi metadata would seem to be trickier. Herbert has covered the data provider issues. From the legal point of view I'm not sure there is anything exceptional in the relationship between OAi DPs and service providers, compared with a non-OAi equivalent. The *** difficulty DPs have in setting terms of use at this stage is that maybe *** they can't anticipate what SPs will want to use their data or, more *** significantly, what their business models will be. Steve ___________________________________________________________________ Date: Thu, 19 Jul 2001 09:42:33 PDT cc: oai-general@oaisrv.nsdl.cornell.edu From: billn Subject: Re: [OAI-general] Re: OAI and intellectual property issues *** The subgroup might even consider both free and commercial versions of *** the IP statement. It needs the status as coming from OAI as a *** recommendation or base IP statement. Bill Nicholls ___________________________________________________________________ Date: Thu, 19 Jul 2001 13:01:31 EDT To: oai-general@oaisrv.nsdl.cornell.edu From: Hussein Suleman Subject: Re: [OAI-general] Re: OAI and intellectual property issues i agree to some degree with many people in this discussion ... the reason why OAI needs to stay out of IP recommendations is that the Initiative needs to adopt a non-partisan stance so as to promote widespread adoption of the underlying principles behind the technology. that said, not supporting a single baseline IP statement is one approach to achieve this ... another would be to specify a range of possible IP statements as baselines and i would like for us to investigate the feasibility of doing this ... *** would it be possible for us to specify a small set of standard *** statements that are strong enough to serve as a superset statement list? for example, if an IP statement said "xyz" then it does not necessarily preclude the possibility of some or all of the metadata having looser restrictions ... any data provider that conforms to "xyz" may harvest the the metadata ... but data providers that are willing to read the individual statements and weigh the legalities are also welcome at their own peril to implement "looser-than-xyz" policies ... *** i also recommend a simple controlled vocabulary with correspondences to *** rights statements that are listed as a supplemental tech document on the *** OAI site ... something along the lines of: "completely free" = "All metadata and associated data may be used without restriction by any commercial or non-commercial entity for free access or resale" "metadata publicly free" = "All metadata may be used without restriction by any commercial or non-commercial entity, but may not be further propagated in a non-value-added form for profit" ... *** now i know people are going to say this is too complex ... i dont think *** it is ... with a little careful wording we can cover most cases with a *** few statements ... and a controlled vocabulary will make it so much *** easier for service providers to abide by the data providers' policies ... ttfn ----hussein