Re: Rethinking Licensing (was Re: [ba-ohs-talk] open source +patents = puzzled)
Yes, I think something along these lines makes sense. (01)
Automating the process is the hard part, but a "software ecology"
is a sound concept, imho. (02)
I suspect that the system would need to work something like this:
* You have access to this huge library, with lots of APIs.
* Anything you build that uses those APIs would be sold within
the ecological framework.
* Any API you use would receive a fractional percent of the
revenue, based on the contribution of that API to the whole.
* The API's "contribution" would have to be factored with
respect to relative size and the total count of APIs in the
system. (It's a tough one. The code for a database API might
be huge, and yet only used once in a trivial little application.
On the other hand, that application might access a database
that people access a lot, like a TV listing. So 50% of a tiny
$10 app might consist of that API -- but the value primarily
comes from the data -- not the API.
* That's where negotiation comes in -- to argue the relative
value of the pieces. That's something the mechanism needs
to take into account.
* Interestingly, a large API or body of code (whichever metric
is used) might be more expensive, but smaller, lighter APIs
would see more use -- making an interesting economy that
favored smaller APIs! (03)
I think the negotiation / determination of relative value is the
trickiest proposition -- as is the need to move revenues through
the system, so they are automatically distributed to contributors. (04)
Then there is the need to know in advance how much of the
revenue will be diverted, so you can set the pricing properly.
But those are mostly mechanical details. It's the relative value
problem that makes a good solution hard to find. (05)
Jack Park wrote: (06)
> Johannes asks great questions, and Paul, as always, has great responses
> (notice I didn't say "answers"). Paul's response (included below) has
> prompted me to go public with some of my thinking.
> But first, I'm not a lawyer, and never even played one on TV, but I do
> recall that a paper discussing the design and intentions of the Mozilla
> Public License (which I cannot find with google anymore) spent a lot of
> verbiage talking about the very issues Johannes raises. A quick read of
> MPL 1.1 tells me I have no idea what they are saying, however.
> So, I'm going to go public with some of my thinking along these
> lines. Whilst I do not much care for the attitudes that reside behind the
> GNU GPL, I think they arose out of some honest problems that exist in the
> software industry. I also think that the GNU GPL goes a long way toward
> resolving those issues, even if MS doesn't like it.
> I've been wondering if it isn't time to rethink everything one more
> time. Here is where I am going with that. Ok. I don't much care for the
> viral nature of the GNU GPL, but, in reality, that's likely because I don't
> much care for the idea of wrecking the software industry. What, I wonder,
> would happen if the viral nature were harnessed, instead, to help the
> software industry. How might that happen?
> To think about that, we go all the way back to some of the earliest ideas
> sprouted by the likes of Engelbart and Nelson, hyperlinking and
> micropayment streams. If you're going to connect with everyone, why not
> turn the whole web of connections into a financial ecology, one that
> doesn't impact people in the same way that paying hundreds of dollars for
> an operating system that crashes all the time does.
> Is this being done already? Actually, yes. Buy Jason Hunter's Servlets
> book and you get the rights to use his source code in your projects. You
> can even use that code in a commercial project, so long as everyone on the
> team owns the latest copy of the book. (I confess, at this writing, I'm a
> bit confused after reading the license--it's clear you can use the software
> in commercial projects, but it's not clear you can actually sell the
> product without further discussions with Hunter). Micropayments, indeed.
> Also, you are not granted the right to distribute the source code. To get
> the source code, you must buy the book. In some sense, paying, say $50 for
> access to code that will save you far more than that amount if you use it
> makes sense.
> Suppose, then, that we look at inheriting directly from GNU GPL, by, say,
> adding some verbiage to it that takes advantage of the viral nature and
> essentially does precisely what the license already does, with the added
> proviso that distribution of derivative works implies the equivalent of
> distribution of the "latest copy of the book" with it. Essentially, a
> royalty stream. It's also known as MLM -- multilevel marketing (which has,
> imho, a bad name owing to the observation that many folks involved in MLM
> don't know anything about the products they are hustling; they're
> apparently in it for the ride).
> Is this a disruptive idea? Probably. Is it an original idea? I already said
> it isn't original. Who else is doing it? Google micropayments.
> You'll likely land at Brad Cox's site, in which he discusses the whole
> idea. Could it work? Maybe. How might it work?
> Consider this: I have this WikiWiki that serves as the backbone for plug in
> Wikis. One of them does IBIS, one of them tries to mimic the Lucid Fried
> Eggs program (see that program at http://www.memes.net), one of them serves
> as a development and presentation platform for online courses, one of them
> serves as an authoring and presentation platform for manuscripts (which I
> happen to need in order to put my book _XML Topic Maps_ on the web). Two
> others in the pipeline serve users directly ("MyWiki") and groups, and
> enable construction of Web portals. Sounds like a lot of stuff. It is,
> but, given the Wiki nature of things, it turns out that it's not too hard
> to add features. What's unique? To start with, the entire Wiki universe
> is version controlled at the "purple number" level, meaning paragraphs,
> etc. Also, each "purple number" is really the URL of a "home page" for the
> specific unit of information identified by the number. And, the homepage
> is also a meeting grounds where anyone can go (click the "purple number)
> and launch an IBIS discussion about that specific information object. The
> homepage could also serve as a repository for backlinks that keep track of
> where that object has been transcluded. Homepage also serves as access to
> previous versions. Homepage is where new versions are created. Also, and
> not the least, that page could be the launching ground for that which
> ScholOnto is all about: building links between information
> objects. Essentially, the project is information-object centric, with a
> user interface that tries to make things easy. <note>NexistWiki is
> functional on my local net at home. It's now in a box to be taken online at
> http://www.nexist.org as soon as I finish making apache and tomcat talk to
> each other, but it is, by no means, completed, yet</note>
> Having said all that, I am wondering how this project might be
> published. My first thought was to write a book about it, include the
> source code with the book, and give it a license similar to the one Jason
> Hunter uses. More recent thinking involves generalizing that
> approach. For that, I am reminded of the OpenDoc initiative started by
> Apple, IBM, and others. In that space, an "object bus" somewhat akin to
> Corba, was created such that anyone could publish widgets (e.g.
> spreadsheets, word processors, etc) that seamlessly drop onto the
> bus. Apple even went so far as to drop the "File" menu in favor of one
> called "Document": click new and you got a new document, a canvas on which
> you could drag and drop widgets that automatically settled into the
> document; start with a word processor, drag out a spreadsheet when you need
> it, drag in a graphics engine, whatever. For me, the appealing aspect was
> that OpenDoc opened an entire financial ecosystem in which anyone could play.
> I think we can resurrect that ecosystem. That's the point of going public
> with my thinking now. I think that something along the lines of the GNU
> GPL could be harnessed to revive (not wreck) the software industry. Lofty
> thought, that. To do so, I am thinking, in relation to my Wiki project,
> that I could publish it with or without a book, but subject to some
> micropayment system. The micropayment system is, in fact, a publishing
> scheme by which authors that add value to the Wiki can put their "jar"
> files into the catalog of widgets available. The beginnings of a small
> financial ecosystem. Suppose, then, that this publishing mechanism sent a
> bucket brigade of micropayments backward to those who contribute stuff
> that's already free, like, to some consortium that improves Java, to the
> Apache foundation, even to individuals who make great things (e.g.
> TouchGraph). Beer money at first, perhaps a career eventually. Who knows.
> Summarizing, I think that I am doing nothing less than revising
> OpenSourceThink. You buy (at some nominal cost) some software and you
> still get the source code. You are free to revise it, fix it,
> whatever. You are also responsible for playing within the constraints of
> an ecosystem in which that software resides. You are also free to enhance
> that ecosystem by dropping your additions back into the pool, laughing, I
> suppose, all the way to the bank.
> Finally, I must tie this into the OHS thread, for that is what this list is
> all about. Since I'm rethinking things, why not rethink Bootstrap Alliance
> while I'm at it. What is the BA value proposition? I think the value
> proposition resides in the possibility of taking Dr. Engelbart's vision to
> the masses, with great potential to benefit humankind. Under that,
> however, what is BA? Here, I am rethinkiing again. How many people
> reading this are aware that Mcdonalds Corp is not a hamburger
> company? Mcdonalds is a real estate investment company, but they have
> "positioned" themselves, that is, their value proposition, resides in
> hamburgers. The same kind of thinking could apply to BA. The value
> proposition stands as is, but the entity, BA, could be a software
> development firm, one that is the center of an entire financial ecosystem
> comprised of people all over the world contributing enhancements to that
> ecosystem, and thus, to humankind. Disruptive thought? Yup.
> Beginnings of a FAQ:
> Q: Won't people ignore this and just steal the source code anyway?
> A: Yup
> 'nough said for now.
> At 10:25 AM 5/13/2002 -0400, you wrote:
> >Johannes Ernst wrote:
> > > Am I off-base here, or is there quite a can of worms hidden here
> > > somewhere? Anyone ever looked at the patent portfolio of any of the
> > > (particularly larger) companies that often release some of their own
> > > code into the public domain?
> >I am not a lawyer, but I am a twenty+ year software developer with an
> >interest in copyright and patent legal issues, and it is my
> >understanding that:
> >1. You can release code under an open source license such as BSD (which
> >does not reference patents) and then sue users of the code for patent
> >infringement even if they are only using exactly the code you wrote.
> >This is clearly unfriendly, and users might fight in court saying you
> >gave them an implied license, but I think the author (patent holder)
> >might win. Look at it this way, if you own a patent on an algorithm,
> >like many "non-profits" do these days,
> > http://www.theatlantic.com/issues/2000/03/press.htm
> > http://www.mp3licensing.com/royalty/software.html
> > http://www.mpeg.org/MPEG/mp3-licensing.html
> >why not just release an explicitly public domain implementation and then
> >say if someone wants to use the code, they must also pay to license the
> >The unfortunate thing about software patents is they imply having a
> >copyright license to copy code in no way ensures you have all the
> >licenses needed to run the code. Note that some especially greedy
> >non-profits (or for-profits) try to double dip by charging separately
> >both for patents and access to copyrighted implementations, but that's
> >another story (they often try to also claim trade secrets in the
> >copyrighted code as well to prevent discussing its shortcomings).
> >2. Some licenses like the GPL explicitly say any referred patent will be
> >licensed freely.
> > http://www.gnu.org/copyleft/gpl.html
> > > Finally, any free program is threatened constantly by software patents.
> > > We wish to avoid the danger that redistributors of a free program will
> > > individually obtain patent licenses, in effect making the program
> > > proprietary. To prevent this, we have made it clear that any patent
> > > must be licensed for everyone's free use or not licensed at all.
> >[Note: IIRC Richard Stallman says this can be read to say a patent only
> >needs to cover free GPL implementations and can be otherwise enforced
> >against others.]
> >While one may or may not like the GPL's political stance on enforced
> >sharing, this is one gotcha it it confronts up front -- contributors
> >embedding patented algorithms to trap the unwary. Note that patents (and
> >copyrights) unlike trademarks can be enforced at any time -- even after
> >letting users slide by for years infringing the use of them (witness the
> >current British Telecom effort to enforce a hyperlinking patent).
> > http://www.wired.com/news/print/0,1294,50398,00.html
> >3. Note though, that if GPL'd code is found to unknowingly reference a
> >patent that isn't freely licensed, the GPL says the code can't be
> >redistributed by anyone.
> > >From the GPL license text:
> > > If, as a consequence of a court judgment or allegation of patent
> > infringement
> > > or for any other reason (not limited to patent issues), conditions are
> > > imposed on you (whether by court order, agreement or otherwise) that
> > > contradict the conditions of this License, they do not excuse you from
> > > the conditions of this License. If you cannot distribute so as to satisfy
> > > simultaneously your obligations under this License and any other pertinent
> > > obligations, then as a consequence you may not distribute the Program
> > at all.
> > > For example, if a patent license would not permit royalty-free
> > redistribution
> > > of the Program by all those who receive copies directly or
> > > indirectly through you, then the only way you could satisfy both it and
> > > this License would be to refrain entirely
> > > from distribution of the Program.
> >Contrast this with "permission to use".
> > http://www.bootstrap.org/colloquium/permission.html
> >One reading of "permission to use" would imply if you contribute any
> >code to any Bootstrap effort you grant a license to Bootstrap and
> >Stanford to use any software patents somehow referenced by the code
> >however they see fit in any projects (including sublicensing to
> >Microsoft). However, this license does not extend to any end users of
> >your code or content (since "permission to use" just grants rights to
> >Stanford and BI, not the end users). Note that if you don't own all
> >patents referenced by your contributed code (say you put in hyperlinking
> >code but don't have a worldwide license to BT's hyperlinking patent)
> >then the indemnification clause of "permission to use" might be read to
> >imply you need to pay Stanford or BI whatever it takes to make the
> >patent holder happy (could be $billions for worldwide licensing rights,
> >including all legal fees etc.). This is why I harp on the revision of
> >"Permission to use" so that the Bootstrap OHS coding effort can move
> >forward as "open source" or "free software", where contributors in the
> >U.S.A. feel comfortable contributing given software patents. [Just
> >rattling around as another Thursday CPC meetign has presumably came and
> >went... ]
> >Proposal: regarding patents, repeal "permission to use" and use the GPL
> >license for the OHS project. Insist (only) that all contributors say
> >their work is original, and that they have permission to contribute it
> >from their employer if such is needed, and that they contribute the code
> >under the GPL license.
> >The old approach over the past two years isn't working to produce code.
> >To paraprase an old TV commercial -- "Where the code?"
> >Why isn't there any significant OHS code released through this
> >colloquium? Perhaps, realistically, nobody who insists on a more
> >permissive license (like BSD or Apache) because they have proprietary
> >commercialization plans for OHS is likely to be going to do any work of
> >this magnitude gratis anyway, and there are too many smooth talking
> >(university) competitors for grant funds to pay people to do it
> > http://www.dli2.nsf.gov/
> >otherwise significant funds would have been here by now -- given the
> >effort Doug (and others) have expended in beating the bushes. [Feel free
> >to surprise me.] Yet, the "permission to use" policy has driven away the
> >very peopel who might build a libre OHS gratis. If someone finally gets
> >significant grants funds for an OHS, and it can be proprietary, the
> >effort will probably just vanish in the noise of all the other ongoing
> >proprietary KM efforts brought up by list participants.
> >With Doug's visibility coupled to a GPL'd OHS implementation, the
> >project might get somewhere faster. We could even have two competing
> >projects -- an Apache version and a GPL'd version. I'd work on the GPL
> >version. Sadly, "permission to use" directly blocks my own efforts in
> >any public OHS direction involving the community of this list. It is
> >very frustrating -- because I can't just say to this list "here is a
> >GPL'd OHS project" without also writing a blank check to Stanford and
> >BI. (I would if I could just to see the project go forward and Doug's
> >vision realized, but I can't - and even if I could, I doubt my wife
> >would approve. :-)
> >Besides, I could not in good conscience convince any other person to
> >contribute even one line of code under "permission to use" -- since even
> >if I were willing to risk my entire financial future myself, I could not
> >ask others to be so foolish given all the thousands of other open source
> >or free software projects out there (many I first saw here) that touch
> >on OHS issues and do not have such a problem.
> >That is mainly why I stopped posting on this list -- to disassociate my
> >efforts from any twisting of "permission to use". I post now in hopes
> >this "permission to use" and patent license etc. issues can finally be
> >resolved. Alternatively, can some lawyer or judge just declare
> >"permission to use" invalid because it is a very broad agreement to hold
> >if not signed (or at least the "extended" part) -- essentially becoming
> >in effect like an overly broad non-compete agreement for any prudent
> >Anyway, better things to do than be a broken record on this topic. If
> >the "permission to use" issue as regards software patents and other
> >things gets resolved someday (and Mei Lin wrote 1 May 2002 a discusion
> >is in the works), fantastic -- and please let me know so I can
> >contribute then. If not, this is the last I'll post on this "permission
> >to use" indemnification issue, to avoid dragging down the other very
> >positive aspects of this discussion forum (and myself in negativity). I
> >learn much by following this list. I'm unfortunately still left in a
> >quandry about whether posting on other issues here or unrev creates too
> >much of an association of my other efforts (like the Pointrel Data
> >Repository System) and "permission to use" and so the prudent thing on
> >that basis is not to post (such as during most of the past year) until
> >the issue is resolved. Whether I can contain myself given all the great
> >discussions going on here remains to be seen.
> >It's also saddening to see great contributors like Chris Dent and
> >Kathryn La Barre
> > http://ella.slis.indiana.edu./~klabarre/unrev_firstpage.html
> > http://ella.slis.indiana.edu/~cjdent/unrev/index.cgi
> >left twisting in the wind so long regarding a firm answer from BI or BA
> >or Stanford on copyright issues for the email archives -- when the whole
> >point of "permission to use" should be to at the very least make work
> >such as theirs possible! So "permission to use" prevents my work here on
> >code, but yet has produced no real benefit when it could, thus making it
> >in practice the worst of both worlds. As I wrote before (April 30, 2002
> >on the unrev list), if the work can't be done pro-bono, I'll pledge
> >US$300 towards legal fees for a lawyer to unravel this mess of
> >"permission to use" to strike a better compromise for OHS coding
> >volunteers and get Stanford and BI (or whoever) to sign off on it.
> >-Paul Fernhout
> >Kurtz-Fernhout Software
> >Developers of custom software and educational simulations
> >Creators of the Garden with Insight(TM) garden simulator
> >http://www.kurtz-fernhout.com (07)