Open source licensing being questioned by anti-copyrighttypes
in [Java Programming]
|
Prev: XWSSecurity and WebSphere
Next: Enum Idiom Question
From: Arne Vajhøj on 29 May 2010 21:50 On 29-05-2010 18:22, ClassCastException wrote: > On Sat, 29 May 2010 21:54:24 +0000, Stefan Ram wrote: >> Thomas Pornin<pornin(a)bolet.org> writes: >>> So the GPL is, in practice (not in theory), incompatible with for-profit >> >> If you put a box in a shop with some software for $329, people who >> think that it addresses their needs will buy it. It might contain a >> sheet of paper with the GPL and directory with the source code. Most >> consumers will not read either of them. There are bought many goods in >> our world that one also might get for free by some other means. > > Perhaps. Though at that price point they might feel ripped off if they > later learned they could have gotten it for free elsewhere. > > On the other hand I can see someone spending even $40 on a CD that's nice > and easy to install and considering that okay even if they find they > could have had it for free as a downloadable .iso file to figure out what > to do with to install it. People will willingly pay, sometimes quite a > lot, for convenience. Maybe, but you will not get rich by selling $40 software. Arne
From: Arne Vajhøj on 29 May 2010 22:10 On 29-05-2010 04:17, ClassCastException wrote: > On Fri, 28 May 2010 20:31:14 -0400, Arne Vajhøj wrote: >> On 27-05-2010 22:58, ClassCastException wrote: >>> In dealing with Java, OpenJDK, and Clojure stuff in recent months I'd >>> come to suspect that open source licensing is itself a source of >>> trouble. Notably, Clojure's license is incompatible with the GPL, under >>> which large chunks of other open source software is licensed. >> >> And so what? >> >> This means that you can not modify the Clojure compiler using GPL code. >> >> Very few libraries are GPL. LGPL and GPL with linking exception was >> invented for libraries. >> >> And it has no impact on people writing apps in Clojure. > > Actually, it does; if you GPL Clojure code and distribute it it will > violate at least one license because the GPL'd code will link against > (and thus try to force the "viral" GPL upon) code that comes with Clojure > and has a GPL-incompatible license. If you want to make it work, then you can make it work. Use a linking exception. Use GPL V3. Distribute the source and ask the users to get Clojure themselves and build. Make a very narrow definition of how GPL should be interpreted. The point is that the two copyright holders you and Clojure will not bring you to court. If you want to see a problem, then you can definitely find problems. > You can add a linking exception to the GPL on your Clojure code that > permits linking against Clojure libraries and the like without requiring > the GPL attach to those, but this has two problems: > > 1. If existing, say, Java code is GPL, and you want to use it in your > Clojure project, you still can't, since the GPL without linking > exception will "contaminate" your own code and then the base Clojure > code that's not GPLable. > 2. This, and similar situations, will lead to the proliferation of > hundreds of almost-GPLs with different and incompatible linking > exceptions. Not much Java code intended for inclusion in other programs are GPL. And "similar" is not specific. > The GPL v3 apparently tries. It has something in it about an automatic > linking exception for the code's programming language's "base libraries", > but someone said this was vague enough or otherwise had loopholes that > made it impossible to confidently apply to Clojure's GPL-incompatible > libraries. "someone", "vague enough" and "otherwise had loopholes" is just empty words. I would not be concerned over those. >>> So I decided to do a little reading on copyright in general. Why does >>> it even exist? The nominal purpose, it turns out, is to "promote the >>> progress of science and the useful arts" by providing a way for the >>> creators of any popular or important work to ensure remuneration, >>> basically. Which smells suspiciously like a grant of monopoly -- which, >>> barring the notion of "fair use", it basically is. >> >> If everybody could copy software exactly as they wanted then I am pretty >> sure that the software industry would be in a very poor shape. > > I actually doubt this; I think it would work rather differently from now > in some ways, but that people would have found a way to make it work and > to profit in it. Plenty of businesses profit from open source software in > various ways, including by selling support or simply by funding > development of open source software that they use in-house and get > productivity gains from, and, by funding it, get more influence to have > features they'd find useful added and the bugs that particularly harm > their productivity prioritized. Since open source requires copyright then no copyright means no open source. So the existence of open source does not prove that copyright is not necessary. >> Sounds like blogs from teenagers that wants to be able to download >> everything for free and have parents to pay the bills. > > I have my doubts whether Against Monopoly (run by a pair of degreed > economists) and Techdirt (run by a successful dot-com entrepreneur) are > "from teenagers that want to be able to download everything for free and > have parents to pay the bills". :) Given that you do not provide links or names or anything, then the teenager guess is as good as any. >>> It thus seems that copyright was twisted away from its original >>> purpose, to which it might have been poorly suited to begin with, and >>> open source licenses try to twist it back toward that purpose. >> >> Not really. >> >> Open source use copyright the exact same way as closed source. The only >> way. >> >> Open source has different license terms than closed source, but that >> does not change the copyright as such. > > I'm sorry you don't seem to get what I'm driving at. > > Copyright was born from the theory that letting authors of "writings and > discoveries" close off access and control the use of their work, e.g. to > set up a tollbooth, would promote progress. > > At least in the case of software, this turned out to be wrong (in some > opinions), and the GPL's "copyleft" was specifically designed to force > copyright to do the reverse: force access open as widely as possible, by > requiring publication of the source code and disallowing monopoly. The > theory this time being that maximizing access and minimizing any one > vendor's control over program code would promote progress. > > Judging by the stellar progress made in improving Linux since its > inception, copyleft is at least as viable as traditional exclusive > copyright in promoting progress in software. > > However, both have proven capable of getting in the way in various > (separate sets of) situations. There may be many different ways to use copyright. But copyright works identical for all - open source or closed source. In fact open source are some of the fiercest enforcer of copyright, because: * they need it even more when they hand out source code * a large fraction of open source people do not believe in patents, so they only have copyright Arne
From: Arne Vajhøj on 29 May 2010 22:24 On 28-05-2010 17:11, Lew wrote: > The GPL as cited only says you have to provide the source code. Actually it does a lot more. Open source means a lot more than access to the source code. > It does > not preclude incentives to pay for it. Correct. > For example, source code is > really only useful if you have a developer handy. Users that don't have > one handy will have to pay to have one handy in order to use the source > code. They can pay a third party to provide the GPLed code and make it > work, or they can hire an employee to do so. Either way, they're paying > for the software. For custom software the difference between open source and closed source are not that important. But for consumer software it is bigger, because they will never want it customized. > And why would they redistribute it? It's open source; there's no onus on > them to distribute what any user can obtain on their own. So that > provides no disincentive to purchase the software. Bullshit. If the open source is distributed for free, then there are no incentive for users to redistribute. But if it is distributed for a cost (above pocket change), then there sure is an incentive. Let us say that your favorite Linux distro was sold for $1000 and you bought a copy. I am pretty sure that you would be willing to provide your brother and your best friend with a copy. And they would also be willing to provide it to a few people. > Sure, they can obtain it for free, and theoretically even use it for > free. But what if they want more than the minimum functionality? Someone > still has to apply expertise to the use of the software, expertise that > must be purchased regardless of licensing fees. This is again custom software not consumer software. Arne
From: Lew on 30 May 2010 11:15 Arne Vajhøj wrote: >> In fact open source are some of the fiercest enforcer of copyright, >> because: >> * they need it even more when they hand out source code >> * a large fraction of open source people do not believe in >> patents, so they only have copyright Bent C Dalager wrote: > This doesn't make much sense. Open source doesn't use copyright to > protect its source from being copied and so they do not "have False. They do use copyright to give them legally assertable authority to determine the rules under which source can be copied, and they therefore use copyright to protect source from being copied outside the boundaries of the license. > copyright" in the sense that copyright is protecting their business > model. To the extent copyright is used as a tool at all in open source > licenses it tend to be used to force the source to remain open but > this then is an ideological goal, not a means with which to protect > one's business in any way. > > Neither patents nor copyright are needed for software development to > thrive so it seems to me your conclusion is based on false premises. Arne never proffered the premise that patents or copyright are needed for software development, so that does not apply. He only said they need it more when they give away source code, clearly because of the need for all the legal ammunition at your disposal when you give things away freely. What Arne did suggest was that, in fact, open-source licenses rely on copyright for their enforceability, and that copyright holders do, in fact, rely on that legal theory to enforce their rights, and they are, in fact, fierce enforcers of their copyright rights. All these are evident in the world as a matter of record. Whether it's true or relevant that a "large fraction of open source people do not believe in patents, so they only have copyright [to protect their interests]" is debatable, and you'd be justified to challenge Arne on both the evidence of the assertion and the applicability to the use of copyright law to enforce open-source licenses. The statement itself is not very far-fetched, so I'm willing to accept that many "open-source people" favor copyright protection over patent protection. They probably have very well thought-out reasaons for this preference, no doubt well founded in legal and tactical concerns. I believe Arne could also make a very strong case for relevance. His stated premises seem at least plausible, certainly not worthy of dismissal out of hand, his conclusions are strongly explained and his asserted facts are mostly in evidence. There's a difference between an assertion that copyright law is used to enforce ownership rights, which is quite clearly so, and an assertion that there is no better way to enforce those rights. Perhaps you mean to assert that there is such a better way, which is an interesting proposition in itself depending on how one proposes to go about it. -- Lew
From: Arne Vajhøj on 30 May 2010 22:56
On 30-05-2010 08:45, Bent C Dalager wrote: > On 2010-05-30, Arne Vajhøj<arne(a)vajhoej.dk> wrote: >> Since open source requires copyright then no copyright means >> no open source. > > If open source "requires" copyright then it is only because in the > current legal regime copyright is mandatory. Copyright is not mandatory. It is perfectly legal to put code in public domain. Open source chose not to do so. Because they want to have some license conditions. No copyright => no license conditions => no open source. > It does help, of course, that copyright is such a huge and > indiscriminate sledgehammer you can use it to kill just about > /anything/, including copyright. Open source will tend to use the best > tool for the job even if that tool is copyright. Open source does not kill copyright. They just use copyright for a different purpose than MS, IBM, Oracle etc. (or should I say for those companies close source offerings - they do have open source offerings as well). >> In fact open source are some of the fiercest enforcer of copyright, >> because: >> * they need it even more when they hand out source code >> * a large fraction of open source people do not believe in >> patents, so they only have copyright > > This doesn't make much sense. Open source doesn't use copyright to > protect its source from being copied and so they do not "have > copyright" in the sense that copyright is protecting their business > model. To the extent copyright is used as a tool at all in open source > licenses it tend to be used to force the source to remain open but > this then is an ideological goal, not a means with which to protect > one's business in any way. Copyright is a tool. That tool is used for different purposes: - traditional closed source - to make money - open source - give end users certain rights and limit some responsibilities - Roedy - make sure that the code is not used for military purposes etc. Copyright may have been invented for the first purpose, but copyright is copyright no matter what purpose it is used for. > Neither patents nor copyright are needed for software development to > thrive so it seems to me your conclusion is based on false premises. Neither traditional closed source nor open source would be possible without copyright. I am not aware of any company making a living from public domain software. Based on that I am very skeptical about software industry in a world with no copyright. I don't think anyone would miss software patents if they disappeared. That is anyone in software. The lawyers would definitely miss it. Arne |