"Biological Open Source"

Janet Hope janet.hope at anu.edu.au
Fri Nov 17 00:57:23 UTC 2006

Thanks David.  So on that basis maybe my "open source patent licence"  
would look something like the statement below?  Can people see  
problems with this, whether legal or practical, from my "licensor"  
perspective? Can people see problems from the perspective of any  
principle of OS?

"Janet's Biotech Process Patent Non-Assertion Covenant 17 November 2006

"1. Janet irrevocably covenants that, subject solely to the  
reciprocity requirement described below, she will not seek to enforce  
any of her enforceable U.S. or foreign patents against products  
generated through use of Biotech Process or any  improvement to  
Biotech Process.

"2. The foregoing covenant shall not apply and Janet makes no  
assurance, covenant or commitment not to assert or enforce any or all  
of her patent rights against any individual, corporation or other  
entity that asserts, threatens or seeks at any time to enforce its  
own or another party's U.S. or foreign patents or patent rights  
against any products generated through use of Biotech Process [or any  
improvement to Biotech Process].

"3. This statement is not an assurance either (i) that any of Janet's  
issued patents cover products generated through use of Biotech  
Process or are enforceable, or (ii) that products generated through  
use of Biotech Process would not infringe patents or other  
intellectual property rights of any third party.

"4. No other rights except those expressly stated in this Non- 
Assertion Covenant shall be deemed granted, waived, or received by  
implication, or estoppel, or otherwise."

Note that I've put "or any improvement to Biotech Process" in square  
brackets.  That's to indicate that if I left that part out, it would  
be an academic or permissive style of licence, but if I included it,  
it would be a reciprocal or copyleft style of licence.  Is that correct?

I guess that assuming materials etc necessary to practice the  
patented invention are to be treated as source code, I would also  
have to offer to make them available.  How about something like:

"Janet's Biotech Process Materials and Data Transfer Convenant 17  
November 2006"

I, Janet, promise to make any materials or data required to practice  
Biotech Process patents that are in my possession on 17 November 2006  
available to all comers on request at a price no higher than that  
required to cover my expenses for storage, retrieval and postage of  
said materials or data.  This offer remains open for until 18  
November 2009."

Any problems here?

Do these statements go further than I would need to go to be  
consistent with open source principles?  If so, what further  
conditions can I impose without falling foul of OS principles?


On 17/11/2006, at 3:10 AM, David RR Webber (XML) wrote:

> Janet,
> If its patents - then perhaps the solution Sun is using makes sense?
> Sun announced the issuance of a Non-Assertion Covenant (NAC) for UBL.
>  The full text of the Sun UBL NAC can be found at
>   http://www.oasis-open.org/committees/ubl/ipr.php
> and then Jon Bosak's email here:
>  http://lists.oasis-open.org/archives/ubl-dev/200607/msg00005.html
> This could easily be combined with an OSI license then - if some
> combination of technology is occurring -
> e.g. software that explores genonomic permutations based on patented
> genome sequences...?
> DW
>  -------- Original Message --------
> Subject: Re: "Biological Open Source"
> From: Janet Hope <janet.hope at anu.edu.au>
> Date: Thu, November 16, 2006 1:02 am
> To: license-discuss at opensource.org
> Cc: lrosen at rosenlaw.com
> Larry,
> Thanks for your response.  I learned a lot by reading your book, by  
> the
> way!
> I think we talked about this very issue when we met back in 2003.  My
> question is this.  Suppose you have a technology that has nothing  
> to do
> with software whatsoever -- for example, a method for creating a
> genetically engineered plant.  Suppose that the technology is  
> protected
> solely under patent law.  In that case there is no copyright licence,
> because there is no copyright.
> So, I'm talking about a situation in which the assumption that
> Presumably software is involved in Biological Open Source or we
> wouldn't be discussing this here at all. does not hold.  (As I
> mentioned in my earlier post, strictly speaking this discussion is off
> topic for that very reason and may end up getting moved to another
> list.)
> Now suppose that the owner of this patented technology wants to
> distribute it in a way that achieves the same substantive outcomes in
> terms of "technology freedom" and ongoing collaboration as a copyright
> owner seeks to achieve by adopting an open source copyright licence --
> the same *substantive* outcomes, I say, but inevitably through
> different licensing *forms* given that there is no copyright but  
> only a
> patent right.  How would he or she go about it?
> As I understand it, this is what CAMBIA have tried to do with their  
> BiOS
> licence, and they have solved many of the obvious problems
> satisfactorily.  I've identified a few areas where I think the
> practical effect of the licence is more restrictive than would be
> permitted in the case of an open source copyright licence.  What I'd
> like to know is, could CAMBIA have built a better (meaning closer)
> analogy with the open source copyright approach?  Or are these
> differences inevitable given the different legal and/or technical
> context in which they are working?
> If I understand correctly, this is a slightly different problem to  
> that
> addressed by the patent licence grants you mention, because they in a
> sense are merely "getting out of the way" of an open source copyright
> licence, not doing all the work of implementing an open source-style
> approach in the absence of any copyright grant.
> On the other hand, I don't see why my hypothetical patent owner  
> wouldn't
> just use language very similar to that included in OSL 3.0.  The only
> part of that language that wouldn't make sense in the absence of any
> copyrighted work is the reference to an Original Work or Derivative
> Work.  Nevertheless, CAMBIA have chosen to use quite different wording
> in their grant (BiOS PET License Clause 2), and I'm wondering whether
> that wording looks open source to you.  Assuming the patent is a
> process patent, are they granting every possible freedom to licensees,
> or are they holding something back?  If they're holding something  
> back,
> can they still call what they are doing "open source"?
> Does all this make sense? Or am I still missing something?
> Cheers
> Janet
>    Best regards,   /Larry
>  Lawrence Rosen Rosenlaw & Einschlag, a technology law firm
> (www.rosenlaw.com) Stanford University, Lecturer in Law 3001 King  
> Ranch
> Road, Ukiah, CA 95482 707-485-1242  *  fax: 707-485-1243 Author of  
> "Open
> Source Licensing: Software Freedom and                 Intellectual
> Property Law" (Prentice Hall 2004)
>    From: Janet Hope [mailto:janet.hope at anu.edu.au]
> Sent: Wednesday, November 15, 2006 5:58 PM
> To: license-discuss at opensource.org
> Subject: Re: "Biological Open Source"
>  On 15/11/2006, at 11:39 PM, Michael Tiemann wrote:
>  Janet, I for one welcome your participation.  Moreover, while it  
> is not
>  likely to lead to a specific license approval, it is very much on  
> topic
>  for the expertise this list represents.
>  Thanks, Michael, for your positive response.  Having received some
> private as well as public encouragement and no discouragement, I'll go
> ahead and put my question to the list.  (It has been suggested that a
> new list could be started for non-software open source licensing  
> issues
> to cut down on off-topic posts -- I think that would be very helpful.)
>  This post is long. Most of it falls under the headings  
> "Background" and
> "Preliminary licence analysis".  Readers who are interested, but not
> that interested, are encouraged to skip those parts.
>  The licensor, CAMBIA, has issued two "Biological Open  
> Source" ("BiOS")
> licences, one for Plant Enabling Technologies ("PET") and one for
> Genetic Resources Indexing Technologies ("GRIT").  Each must be  
> read in
> conjunction with its own Technology Support Services Subscription
> Agreement ("PET TSSS" and "GRIT TSSS").  All of these instruments are
> currently in version 1.3 and are available at
> http://www.bios.net/daisy.
>  Let me say up front that I am not affiliated with CAMBIA and have not
> been involved in the process of drafting the BiOS licences.  As a
> matter of courtesy to the drafters, I am simultaneously posting my
> comments to CAMBIA's own discussion forum.  However, as Michael
> suggests, I suspect that license-discuss is where much of the relevant
> expertise is to be found.
>  The question is: To what extent is CAMBIA's BiOS licence scheme "open
> source" in the software sense?
>  In other words:  If the BiOS licences were software/copyright  
> licences,
> would you recommend them for OSI approval?  If not, why not?
>  In particular (see the rest of this post for more detail on each of
> these points):
>  (1) If a licence is primarily a patent licence and the patent grant
> itself purports to be open source (not just compatible with an open
> source copyright grant elsewhere in the licence), should field of use
> restrictions be permitted? (OSD #6)
>  (2) What should be regarded as the equivalent of "source code" in a
> biotechnology licence? Should it include materials etc in addition to
> that which would be required for disclosure under patent law? If  
> so, on
> what terms should it be made available to users? Specifically, is it
> acceptable to charge a fee that does more than recover the costs for
> storage/retrieval/postage of samples etc (OSD #2)? Is it acceptable  
> for
> the amount of the fee to depend on the size and type of the user's
> organisation (OSD #5) and the length of the user's "subscription" to
> ongoing "technology support services"?
>  (3) What is the nearest patent equivalent of the open source  
> copyright
> licence grant to copy, modify, and distribute copies or derivative
> works (OSD #1)?  Does it depend on the type of patent (eg process or
> product)?  Is the patent grant in the BiOS licences effectively  
> limited
> by the exclusion of "Improvements" from the definition of "BiOS  
> Licensed
> Products"?
>  (4) Is a right to sublicense essential in an open source licence?  
> Is it
> acceptable for this right to be limited or absent? If so, what other
> conditions should be met to ensure that a licensor cannot impose
> non-open source restrictions as a condition of issuing a new licence
> (eg OSD #7)? How can these conditions be met in the case of a
> technology to which the equivalent of a copyright notice cannot be so
> easily attached as it can to software code?
>  (5) Does the grantback in the BiOS licence go beyond a copyleft-style
> provision in restricting users' freedom to fork the innovative  
> process?
> Is there any pressing legal or practical reason why the licensor in a
> biotechnology/patent context could not simply require licensees to  
> make
> defined improvements available on the same terms as the original
> licence, as in the software/copyright context?
>  CAMBIA (formerly an acronym for Center for Application of Molecular
> Biology in International Agriculture) is a private, not-for-profit
> research institute located in Canberra, Australia.  Founded in 1994 by
> its present CEO Dr Richard Jefferson, CAMBIA is a small organisation
> that -- atypically -- combines "wet lab" development of biotechnology
> research tools with intellectual property informatics and policy
> development.  CAMBIA is financed by grants from philanthropic
> organisations including the Rockefeller Foundation, by national and
> international research funding bodies, by official development
> assistance and by licence revenue from its own patented
> technologies.(See http://www.cambia.org/daisy/cambia/589.html.)
>  In 2005, CAMBIA launched an initiative called Biological  
> Innovation for
> Open Society (BIOS: note the upper case "I").  The BIOS initiative
> consists of:
>  (1) Patent Lens: searchable databases containing EPO, US and PCT  
> patent
> documents, together with ancillary IP-related information and  
> tutorials.
>  (2) BioForge (styled after Sourceforge.net): a portal for
> protocol-sharing, comments on patents, and discussion tools in both
> public and secure environments; intended to develop into a
> collaborative technology development platform.
>  (3) Biological Open Source (BiOS with a lower case "i") is the aspect
> of the BIOS initiative that is "intended to extend the metaphor and
> concepts of Open Source to biotechnology and other forms of innovation
> in biology". (See CAMBIA BiOS License for Plant Enabling Technology
> Version 1.3, Recitals, first paragraph.)
>  A description of the materials and methods that researchers can  
> obtain
> from CAMBIA, including an indication of which materials and methods  
> are
> available under which BiOS licence, is at
> http://www.cambia.org/daisy/cambia/materials.html.  These technologies
> are, of course, quite different from those licensed under OSI- 
> certified
> software licences.  Plant enabling technologies and genetic resource
> indexing technologies consist not of software code (though software  
> may
> be one component), but of heterogeneous methods and materials,  
> including
> living biological materials.  Typically, such technologies are legally
> protected by a mixture of intellectual and personal property rights,
> including patents.  By contrast, I understand that most open source
> software licences are primarily copyright licences, though some also
> contain a limited patent grant.
>  As I mentioned in my first post, the fact that the BiOS licences  
> apply
> to plant enabling technologies and genetic resources indexing
> technologies means that they cannot be certified by the Open Source
> Initiative as being "open source" licences because the terms of the
> official Open Source Definition are not broad enough to accommodate  
> the
> legal and technical differences just described. In consequence,  
> there is
> no certification signal to help potential users determine whether  
> or not
> signing up to a BiOS licence will actually give them the nearest
> equivalent to the "software freedom" guaranteed by the principles of
> open source software licensing.  Clearly, the task of translating open
> source licensing principles from one technology setting to another is
> one that creates room for both legitimate adaptations and distortions
> of the kind that would lead the OSI to refuse to certify the relevant
> licence if the technology in question were, in fact, software.  By my
> preliminary analysis, the BiOS licences contain a mixture of the two.
> However, I (and many others) would like to be able to supplement that
> analysis with the expert views of people on this list.
>  Note that I am in no way impugning CAMBIA's good faith or suggesting
> that the BiOS licences are "bad".  Certainly, CAMBIA has made a  
> serious
> effort at a very difficult task.  The licences may well be good
> licences, and good for innovation.  However, the question here is
> whether it is reasonable to describe them as "open source".
>  For simplicity, I refer here only to the terms of the BiOS Plant
> Enabling Technology (PET) agreement. The PET scheme closely resembles
> the GRIT scheme. The main difference is that whereas the PET scheme
> contains a limited sublicensing right, the GRIT scheme does not  
> contain
> any right to sublicense.
>  1. Overall scheme of the licence
>  The BiOS licensing scheme encapsulated in version 1.3 of the PET
> Licence and corresponding TSSS Agreement contemplates an initial
> contribution of IP and technology by CAMBIA that may act as a seed for
> ongoing technology development. ("IP and Technology" is defined in
> Licence cl. 1.8.)  Thus, the subject of the initial licence grant is a
> collection of CAMBIA's own patents, materials and technology data.(See
> definitions of "CAMBIA Patents", "CAMBIA Material" and "Technology
> Data" in Licence cll.1.9, 1.10 and 1.13 respectively.)  Downstream,
> this initial subject-matter may be supplemented by "Improvements",
> "Improvement Patents" and "Improvement Materials" (Licence cll 1.6,  
> 1.7
> and 1.11 respectively) developed by the licensee,  non-exclusively
> granted back to CAMBIA under Licence cl. 3 and then redistributed by
> CAMBIA to all BiOS licensees as part of the grant of "IP and
> Technology".
>  2. Parties
>  Although the licence is described on the BIOS forum as a "template",
> one party (CAMBIA) is explicitly named throughout the licence text as
> the licensor.  I understand that this is discouraged in the OSI
> approval process because it contributes to licence proliferation;
> however, this is a relatively minor point.
>  The sense in which the BiOS licences are templates is that they
> contemplate a number of licensees all receiving CAMBIA's permission to
> use the IP and Technology on "substantially similar" terms. (See
> Licence Recitals, paragraph 3. Why are the terms "substantially
> similar" and not identical?  Two reasons: (1) different licensees may
> agree to different numbered versions of the licence (see Licence cl.
> 7.5 and TSSS Agreement cl. 7.6). (2) The terms on which licensees can
> access any materials needed to practise the licensed inventions vary
> according to the type and size of the licensee organisation (see TSSS
> Agreement cl. 3 and Annex D).)
>  3. Patent grant
>  The licence grant is contained in clause 2 of the PET Licence, the  
> key
> part of which reads:
>  "2.1 CAMBIA hereby grants to BiOS LICENSEE under Licensed Patents in
> the Field of Agriculture a worldwide, non-exclusive, royalty-free  
> right
> and license to make and use the IP & Technology for the purpose of
> developing, making, using, and commercializing BiOS Licensed Products
> without obligation to CAMBIA...."
>  The nature of the grant is crucial to the question whether the BiOS
> licences can be justifiably described as "open source". Note that
> unlike open source software licences, the BiOS licence is primarily a
> patent licence.  While patent grants in open source software licences
> must be compatible with an open source copyright grant, they are not
> the primary means by which these licences seek to protect users'  
> rights
> to access and use the technology.  The question here is not merely
> whether the BiOS patent licence grant is \textit{compatible} with open
> source principles, but whether it is itself "open source" (whatever
> that is taken to mean in the biology context).
>  3a. Grant is restricted to a particular field of use
>  The BiOS patent licence grant is limited to a particular field of  
> use,
> namely Agriculture. ("Agriculture" is defined quite broadly in Licence
> cl. 1.3.)  There is a question whether this conflicts with OSD # 6.
>  While a strict formal interpretation of open source principles would
> presumably prohibit field of use restrictions in open source
> biotechnology licences, the difficulty with such an interpretation is
> that in the patent context, field of use restrictions, together with
> territorial restrictions, may be a useful tool for making technology
> that would otherwise be bound up in exclusive licences more readily
> available for public interest and broader commercial use.  The
> licensing policy of Public Intellectual Property Resource for
> Agriculture (PIPRA), a collective intellectual property management
> regime for agricultural biotechnology recently established by a group
> of land grant universities in the United States, is a case in point.
> One of PIPRA's primary aims is to overcome the fragmentation of public
> sector intellectual property ownership by identifying residual rights
> retained by members who have assigned unnecessarily broad exclusive
> rights in important technologies to major commercial firms. One of the
> architects of the initiative explains the importance of field of use
> and territorial restrictions in this context: "Best practice includes
> partitioning of patents: if you go and license something like your
> agrobacterium technique, license it just for cotton; or better, for
> cotton in the US; or even better... define which varieties, or...
> constrain it to varieties owned by the licensee company in the US.   
> The
> more you can constrain the space of the technology grant, the more is
> still left over [that you may choose] to put into the commons." (Greg
> Graff, personal communication.) Thus, the value of such field of use
> restrictions in terms of achieving wider access to key biotechnologies
> may outweigh the value of keeping to a strict analogy with open source
> software licensing.
>  What do people think about this? I understand that with respect to
> patented code, the F/LOSS community objects to field of use
> restrictions in software patent licences because such restrictions are
> perceived as capable of insidiously undermining freedoms granted in
> relation to the same code under an open source copyright licence.
> However, the approach of open source community leaders to this
> situation appears to have been to campaign for maximum breadth of
> patent licences, while remaining willing to compromise for the sake of
> workability.  Thus, many open source licences, including the Apple,  
> and Mozilla licences, have field of use restrictions in their patent
> grants. The W3C Patent Policy also represents a compromise.  How far
> should this willingness to compromise go in a case where the patent
> grant is the main or even the only grant in the licence?
>  3b. Scope of grant
>  The grant is a "right and licence to make and use the IP and  
> technology
> for the purpose of developing, making, using and commercialising BiOS
> Licensed Products".  Note that the grant does not explicitly include a
> right to sell the IP and Technology itself.  Nor does it permit the
> licensee to make or use the IP and technology for purposes other than
> developing, making, using or commercialising BiOS Licensed Products.
> "BiOS Licensed Product" is defined in Licence cl.1.4.  That definition
> is discussed further below, but significantly, it is restricted to
> assets that are substantially distinct in some respect from the
> licensed IP and technology and are "intended for commercialisation".
> "Commercialisation" is not defined in the BiOS licence.  However, in
> light of recent US case law concerning patent infringement by
> universities (eg Madey v Duke), it may be quite broad.
>  The OS mantra is "anyone, anywhere, for ANY PURPOSE, must be free to
> copy, modify AND DISTRIBUTE the software, for free or for a fee...".
> At first glance, the BiOS patent grant seems considerably narrower  
> than
> this.  However, patent law grants a patent owner a different set of
> rights to those of a copyright owner.  It is therefore not a
> straightforward question what the equivalent of the broad open source
> copyright licence grant should be in the patent context.  It is
> arguable that an open source copyright grant covers all of the
> exclusive rights of the copyright owner that are relevant to the
> licensed technology, and that an open source patent grant should do  
> the
> same.
>  IAAL, but not a patent lawyer.  However, I understand that in  
> Australia
> at least, the precise form of a broad patent licence grant that would
> cover all of the owner's rights-to-exclude would depend on the nature
> of the patent claims.  Section 15 of the Australian Patents Act 1990
> gives the patent holder the right to exploit the invention. The
> definition of "exploit" in Schedule 1 of the Act distinguishes between
> products and processes. For products, it includes the right to make,
> hire, sell or otherwise dispose of the product. For processes, the
> definition includes the right to make, hire, sell or otherwise dispose
> of products resulting from the process. Without looking at the claims
> in CAMBIA's patents, it appears that they are process patents (ie
> methods of doing plant enabling technology and genetic resource
> indexing technology). The licences grant the right to use the methods.
> Any products are supplied separately under the support agreement. The
> licences actually grant the right to sell products using the methods
> under licence. Thus, the licence grant may be broad enough to satisfy
> open source principles with respect to the initial IP and Technology.
>  What do you all think about this?
>  3c. Limited right to sublicense
>  A final point to note regarding the BiOS licence grant is that it
> includes a limited right to sublicense, as follows:
>  "2.1.1 granting limited sublicenses to third parties... to conduct
> research and/or development activities...for BiOS LICENSEE, provided
> that the product and/or other results (including all intellectual
> property rights) resulting from said limited sublicense are owned
> exclusively by BiOS LICENSEE, said limited sublicenses to terminate
> when such activities cease or such ownership terminates, and
>  "2.1.2 granting limited sublicenses to third parties... for the sole
> purpose of commercializing BiOS Licensed Products that embody the IP &
> Technology or are generated by use of the IP & Technology, said  
> limited
> sublicenses to terminate when said commercialization ceases.
>  "BiOS LICENSEE shall be responsible to ensure... that any  
> Improvements
> produced by sublicensees are considered to be Improvements
> hereunder.... BiOS LICENSEE shall provide a list of sublicensees to
> CAMBIA in writing at least once a year....
>  "BiOS LICENSEE shall further be responsible to ensure that...third
> parties to whom a sublicense has been granted are notified... that  
> said
> third parties have no right to sublicense absent the execution of a  
> BiOS
> License Agreement with CAMBIA, and that CAMBIA may be approached for a
> BiOS License Agreement.
>  "Other than stipulated under this Article 2.1, no further right to
> sublicense is granted to BiOS LICENSEE hereunder."
>  I am unclear as to whether a right to sublicense is essential for OSI
> certification of new licences.  My understanding from Larry Rosen's
> book is that the right to sublicense is considered desirable, but is
> not a feature of all OSI-approved software licences.  I imagine  
> that in
> the absence of a sublicensing right, OSD #7 becomes very important
> because unless people are sure they are entitled to a licence no  
> matter
> what, the need to go back to the initial licensor for permission to  
> use
> the technology would not be a mere formality and could become too
> restrictive.  However, it is not clear exactly how OSD #7 would  
> operate
> in relation to the exchange of biological materials, methods etc, as
> distinct from code or other content to which a copyright notice can be
> easily attached.
>  Can anyone clarify the OSI requirements re sublicensing?  (Sorry if
> this has been dealt with in previous posts.)
>  4. Source code
>  The freedom to copy and modify open source software programs and to
> distribute copies and modifications is only one important aspect of
> open source licensing.  Another is the practical enablement of that
> freedom through unrestricted access to software source code.  In the
> case of patented biotechnologies, it may be argued that the equivalent
> of source code is automatically publicly available because of the
> disclosure requirement under patent law, and that an open source-style
> licence grant is all that is needed to render a patented technology
> "open source".  On the other hand, biotechnology licensors  
> typically go
> beyond this level of disclosure when attempting a genuine transfer of
> technology.  This practice suggests that patent disclosure many not be
> enough to give full practical effect to the legal freedom to use and
> improve a technology that is intended to be conferred by an open  
> source
> licence grant.
>  In the BiOS PET licensing scheme, separate provision is in fact made
> for access to materials needed to practise the licensed patents.  This
> access is governed by the terms of the TSSS Agreement (see Licence cl.
> 4) and appears to be conditional upon (1) a substantial annual payment
> for at least three years, depending on the licensee's organisation  
> type
> and number of employees and (2) significant reporting requirements
> regarding new materials, technology data and project ideas.  If access
> to materials needed to practise the licensed payments is regarded as
> equivalent to access to source code, these restrictions represent a
> significant departure from established open source principles of
> non-discrimination among licensees, access to source code for free or
> at the minimum charge required to recover the provider's costs, and
> freedom from ongoing obligations to the licensor, including  
> obligations
> that take the form of reporting requirements.
>  5. Grantback
>  The next phase in the scheme established by the BiOS licences and  
> agreements is triggered when a licensee, in the course of exercising
> the licence grant, makes or discovers something new.
>  The BiOS PET licensing scheme distinguishes between "BiOS Licensed
> Products", which may be developed, made, used and commercialised
> "without obligation to CAMBIA", and "Improvements", which are subject
> to a grantback obligation  (see below) that also covers Improvement
> Patents and Improvement Materials (Licence cll. 1.7 and 1.11,
> respectively).  The definitions of these different categories of
> potential follow-on innovations are complicated and tinged with
> uncertainty.  However, the fundamental intention appears to be to  
> allow
> the licensee to appropriate any new asset that has been generated
> through use of the IP and technology and is intended for
> commercialisation.  Such an asset is a "BiOS Licensed  
> Product" (Licence
> cl. 1.4), provided it does not fall within the definition of
> "Improvement" (Licence cl. 1.6).
>  The definition of "Improvement" is thus a critical element of the  
> BiOS
> scheme. This definition is potentially extremely broad.  Note in
> particular that even something as minor as a repeatable observation  
> can
> constitute an Improvement, and that an Improvement need not be a Plant
> Enabling Technology (itself very broadly defined in Licence cl. 1.12)
> because of the catch-all "but for the terms of this License
> Agreement...".
>  5a. Does the definition of "Improvement" limit the licence grant
> itself?
>  A preliminary question is whether the broad definition of  
> "Improvement"
> in Licence cl. 1.6 effectively limits the scope of the main licence
> grant in Licence cl. 2.  As noted earlier, the purpose of the grant is
> limited to "developing, making, using and commercialising BiOS  
> Licensed
> Products".  "BiOS Licensed Products" is defined in cl. 1.4 to mean  
> "any
> tangible or intangible asset of BiOS LICENSEE (including without
> limitation any material or method, but excluding Improvements)".  This
> suggests that anything that falls within the definition of Improvement
> automatically falls outside the definition of BiOS Licensed Product  
> and
> therefore is not covered by the grant even for purposes other than
> commercialisation.
>  Similarly, any new development that is neither a BiOS Licensed  
> Product
> nor an Improvement does not appear to be covered by the terms of the
> licence.  In either case, further permissions may be required in order
> to avoid infringing the licensor's proprietary rights.  CAMBIA appears
> to be free to make such permissions conditional upon payment of
> royalties or other conventional licensing terms -- or to refuse them
> altogether.
>  5b. Grantback: copyleft equivalent or licensing "club"?
>  From an open source perspective, the breadth of the definition of
> "Improvements" is not necessarily a problem, although it could be  
> if it
> extended the range of follow-on innovations that would be caught by  
> the
> grantback well beyond those that would be caught by a copyleft-style
> provision relating to "derivative works".  As I understand it, the
> question of what constitutes an appropriate incentive for licensees to
> innovate in a copyleft-style licence is a separate issue from
> compatibility with open source principles.  The broadest examples of
> copyleft "hooks" in open source software licences are tailored to  
> catch
> all derivative works that are distributed outside the boundaries of  
> the
> licensee's own organisation; other open source licences permit varying
> degrees of freedom to operate with derivative works. The BiOS  
> grantback
> appears to fall somewhere in the middle of this spectrum.  In a 2004
> article on open source patent licensing, Sara Boettiger and Dan Burk
> point out that copyright confers exclusive rights only against
> unauthorised copying or other violations of the specifically  
> enumerated
> rights of the copyright owner arising out of contact with the
> copyrighted work itself.  By contrast, independent creation is not a
> defense to claims of patent infringement. The BiOS licence appears
> designed to replicate the copyright situation by excluding from the
> definition of Improvement improvements that are "developed without the
> use of the IP and Technology".  Similarly, the BiOS arrangement  
> appears
> intended to mirror open source software licences by permitting in- 
> house
> use of improvements that are actively protected as trade secrets:  
> these
> are not caught by the BiOS grantback. (Note, however, that this escape
> clause may be somewhat narrower than the positive requirement in open
> source software licensing for a derivative work to be externally
> deployed before it triggers any copyleft obligation: clearly, there is
> a difference between active dissemination of a technology and merely
> failing to  adequately protect a trade secret.  Further, note that
> in-house use of an improvement under conditions of trade secrecy may
> not be permitted at all under the BiOS scheme unless interpreted as a
> form of "commercialisation": see Licence Grant in cl. 2.2 and
> definition of BiOS Licensed Product in cl. 1.4.)
>  The point at which the BiOS licence provisions relating to follow-on
> innovations appear to me to depart from open source principles is not
> in the breadth of what may be caught by the grantback, but in the fact
> of the grantback itself.  While the copyleft or "reciprocal"  
> obligation
> to be found in some open source licences has been characterised in
> academic literature as a type of grantback, this does not appear to me
> to be an accurate description.  A copyleft licence does not  
> establish a
> tit-for-tat relationship between the licensor and licensee.  Rather, a
> copyleft licensor says to the licensee: "Do as you have been done by".
> The copyleft obligation is not to the licensor per se -- though he or
> she may be responsible for enforcing it -- but to the whole community
> of potential users of the follow-on innovation.  The licensor may or
> may not be a member of this community, depending on the scope of  
> his or
> her activities and interests.  For this reason, I prefer to  
> describe the
> copyleft obligation as a "grantforward" (as in "passing it forward")
> rather than a "grantback".
>  The BiOS grantback, by contrast, appears to be just that: whether or
> not the intention is noble, it still constitutes a type of privilege
> granted by the licensee to the licensor in partial consideration of  
> the
> licence grant.  Under cll. 3 and 4 of the Licence and relevant
> provisions of the TSSS Agreement, a BiOS licensee is obliged to grant
> to CAMBIA a worldwide, non-exclusive, royalty-free licence (with the
> right to sublicense to other BiOS licensees) to Improvement Patents,
> Improvements, Technology Data, and any  Improvement Material provided
> by the licensee to CAMBIA and necessary to practise Improvements.
> (Definitions are in Licence cll. 1.7, 1.6, 1.13 and 1.11  
> respectively.)
>  Thus, the form of the BiOS licence differs from a copyleft open  
> source
> licence: instead of requiring licensees to license a subset of
> follow-on innovations on the same terms as those of the original
> licence, it builds the follow-on licence into the original licence and
> places the initial IP owner, CAMBIA, at the centre of the network as a
> kind of gatekeeper. While CAMBIA does have an obligation to make these
> follow-on innovations available on the same terms as the initial
> licence, that obligation extends only to existing BiOS licensees.  In
> essence, the arrangement is a licensing "club".  Given the  
> restrictions
> on sublicensing by BiOS licensees (Licence cl. 2), new members can  
> join
> this club only by executing a new licence with CAMBIA; CAMBIA is under
> no obligation to grant such a licence.
>  Club-type licensing arrangements are actually quite familiar in the
> biotechnology context.  Depending on the circumstances, such an
> arrangement may be good for innovation, but it seems to me that it is
> not open source. Why?  Because an arrangement in which the initial
> licensor retains a central position linked to his or her ownership of
> the seed IP and Technology is inconsistent with the freedom to fork  
> the
> innovative process.
>  The freedom to create a ‘code fork’ is regarded by some as a defining
> characteristic of open source -- recall Eric Raymond's open letter to
> Sun a little while back. According to Steve Weber (author of The
> Success of Open Source), under the terms of an open source licence,
> anyone who is dissatisfied with the conduct of a project leader -- on
> technical, administrative, political or even purely personal  
> grounds --
> is free to take the collaborative effort in a new direction. Even  
> though
> in practice, forking is rare, the ever-present possibility makes  
> project
> leaders responsible to their co-developers and ensures that no
> individual or group unduly dominates the process of technology
> development. At the same time, it ensures that a promising technology
> need not be left on the shelf because of waning interest or incapacity
> on the part of an initial innovator.  This is a concern often  
> expressed
> on this list in the form of: "What happens to the licence if So-and- 
> so,
> the licensor, goes bankrupt or disappears in ten years' time?"
>  Of course, all else being equal, it is natural for an initial  
> innovator
> to remain in charge of ongoing development and to act as a champion of
> the technology. However, it seems key to the open source approach that
> the initial innovator not use his or her ownership of the intellectual
> property in the initial seed technology to retain control over its
> ongoing development.
>  It appears to me that this freedom to fork the innovative process  
> with
> or without the approval of the initial licensor is lacking in the BiOS
> scheme.  Why?
>  (a) The grantback incorporates stringent reporting requirements to
> CAMBIA (Licence cll. 3.2 and 3.4) and obliges the licensee to provide
> CAMBIA with any materials necessary to practise Improvements (Licence
> cl. 3.2).
>  (b) Discussions of technology data between BiOS licensees must be
> channeled through CAMBIA (TSSS Agreement, cl. 2.3).
>  (c) CAMBIA retains the right to impose an obligation on licensees to
> include CAMBIA’s trademarks in downstream technologies (Licence cl.
> 7.9).  (This provision closely resembles one that I thought was  
> removed
> from an early version of the open source BSD software licence  
> because it
> imposed too much of a restriction on the freedom of downstream
> developers -- is this correct?)
>  (d) Under Licence cl. 7.5 and a similar provision in the TSSS
> Agreement, CAMBIA "may publish revised and/or new versions of the BiOS
> License for Plant Enabling Technologies from time to time. Such new
> versions will be similar in spirit to the present version, but may
> differ in detail to address new or newly identified issues. Each
> version will be given a distinguishing version number. ... Where there
> is... any controversy between the parties respecting the  
> interpretation
> or application of the terms of this Agreement, the latest...  
> version of
> the Agreement published on the BIOS website shall be controlling."
>  (e) Perhaps the most extreme instance of CAMBIA's apparent attempt to
> keep its hand on the reins of future development is in the TSSS
> Agreement, cl. 6, which reads in part:
>  "From time to time BiOS LICENSEE may have interest in exploring
> solutions to certain problems in its field of business. In such case
> BiOS LICENSEE shall [emphasis added] send a written project  
> proposal to
> CAMBIA. To the extent that such project proposal is consistent with  
> the
> terms and intent of the BIOS License, which consistency shall be
> determined by CAMBIA in its sole discretion, CAMBIA shall use its best
> efforts to... bring the proposal to the attention of other parties
> within the BIOS Initiative...".
>  In other words, the licensee is obliged to send a written proposal to
> CAMBIA, but CAMBIA is not obliged to do anything with it.  The effect
> (whether or not intended) is that CAMBIA gets a private preview of any
> new project.
>  Thank you for your input.
>  Janet
> Dr Janet Hope
> Centre for Governance of Knowledge and Development
> Research School of Social Sciences
> Australian National University
> T: +61 2 6125 0172
> F: +61 2 6125 1507
> janet.hope at anu.edu.au
> http://rsss.anu.edu.au/~janeth

Dr Janet Hope
Centre for Governance of Knowledge and Development
Research School of Social Sciences
Australian National University
T: +61 2 6125 0172
F: +61 2 6125 1507
janet.hope at anu.edu.au

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensource.org/pipermail/license-discuss_lists.opensource.org/attachments/20061117/8684820c/attachment.html>

More information about the License-discuss mailing list