GPL 3.0 to Penalize Google, Amazon? 582
Michael Ferris writes "Is this the start of a shakedown by the GNU folks? Michael Singer writes that Eben Moglen and the folks rewriting the GPL are looking at a proposal where companies would be required to pay money if they use GPLed software, even if they don't redistribute the software." From the article: "The current version of the GPL, which was last updated in 1991, fails to trigger the open source license if a company alters the code, but does not distribute its software through a CD or floppy disk...the [current] rule does not apply to companies that distribute software as a service, such as Google and eBay, or even dual-license companies like Sleepycat."
Future versions of the GPL (Score:5, Insightful)
Call me paranoid, but I wouldn't want even the faintest chance that some nasty corporation managed to litigate itself in the position of being able to release a future GPL version, as in bold below:
This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.
That's handing control of the licencing of your code over to whoever is allowed to write GPLv1851, if I'm reading it correctly.They deliver HTML. (Score:5, Insightful)
Google and eBay distribute HTML. That HTML is created by software that uses GPL code. So if I modify a GPL Office Suite, would I have to distribute the code if I email someone a document I made with it? Seems like a bad idea, in general.
I guess people could fork the GPL2.0'd code if the software developers switched to GPL 3.0
This doesn't sound like a good idea (Score:5, Insightful)
hails back to the days of the old BSD style licenses. The GPL works
so well now, precisely because it is unobtrusive. My company runs
GPL'd software because we are able to use it and make modification
without either redistributing the source code and we aren't required
to pay for that ability.
Requiring people to open all their changes or pay for them will put a
lot of businesses off when it comes to dealing with GPL'd software. I
don't think that is a good trade off to make, and I don't think it
will be healthy for the open source community in general.
A move like this will make the newer BSD style licenses and / or
licenses like the Python license much more attractive imo. Now that
open source is finally turning the corner, and solid technologies are
finally moving into the enterprise, why would we even entertain making
changes that will certainly hamper open source adoption?
This isn't a consistent position in my opinion. If you are
developing free software, it should remain free. The idea that you
can get someone hooked on software, and then pull the rug out from
under them and start charging them is ludicrous. If this were to
happen, I can honestly see a major fork in the GPL happening.
Bad, bad, bad! (Score:4, Insightful)
Who what when where? (Score:3, Insightful)
I think this proposal is crazy. If you use software as a part of running your business, that software is benefiting you and indirectly providing services to your customers, even if they never see it. So where do you draw the line?
Let's get more detailed (Score:5, Insightful)
Im speechless..... (Score:5, Insightful)
"If you look at the market, Yahoo, eBay, IBM, Amazon, Google have all sunk millions into the GPL infrastructure," Olson said. "Not only are we changing the rules, we are changing them retroactively. With the new way, it lets the customer pay with either their source code or with their wallet."
Basically, in any other language: 'Now youve had time to build a good infrastructure on the current rules, prepare to be shafted'. If this comes to pass, then in my mind they are no better than Microsoft changing EULA terms with a service pack. Now that there are major companies with an infrastructure built on GPL software, the FSF are looking to essentially move the goalposts and if this is applied retroactively to current code (which from the articles wording I think it will be) then I personally think that its going to do more harm to the GPL community than benefit it in code donations as companies scramble to move away.
Please someone tell me that they cant do this retroactively, that its impossible under the current GPLv2 terms.
Re:They deliver HTML. (Score:5, Insightful)
Hmm. I could swear that the Google appliance in my rack, and the Google toolbar on my desktop weren't just hunks of HTML.
Horrible implications. (Score:5, Insightful)
Regardless of what GPL 3.0 turns out to be, developers are not forced to use it. They can continue to use GPL 2 if they wish, just as they can choose to use a BSD license, Apache license, creative commons, or any other license of their choosing. Furthermore, software that has already been released under GPL 2 cannot be retracted, it remains available under GPL 2 forever.
Where did this mindset come from? (Score:5, Insightful)
Re:This doesn't sound like a good idea (Score:2, Insightful)
Re:This doesn't sound like a good idea (Score:5, Insightful)
If you're not distributing your work, there's no reason why you should be forced to open your code or pay some silly fee.
I hope this whole story is a troll...I really do.
Call it FUD (Score:5, Insightful)
The problem is that it infringes on one of the four freedoms, specifically the freedom to use. If such a provision were to find its way into any license that made it so that companies and individuals were not subject to the same terms, the license would be both discriminatory and non-free.
Simply put, this is somebody making FUD about the GPL. Don't buy into it for a second.
It's not in print yet... it doesn't exist (Score:5, Insightful)
At least, that's what I heard anyway.
When whatever changes come up, they will be reviewed and we can rant and rave about it at that time.
Nothing is founded, no concrete written agendas were tossed out by the article and all we have is a hypothetical situation that would be very different from the current model.
Like many have said and will say, it's usually a good idea to specify the version of the GPL in which you release software. Unless you really don't care what kind of changes are made at a later date.
I'm not saying all of this isn't going to come true, but at this juncture we could very well be required to throw pies to comply with the next redistribution agreement.
Don't get Dramatic (Score:4, Insightful)
And frankly, it's not really a loophole. Web services are not distributing software, they're running a service using software. That's obviously open to interpretation, but I haven't ever heard anyone distribute under the GPL and complain about someone using their software as a web service. There has been at least one derivative license which has addressed this issue.
In the end, GPL 3.0 will likely provide an optional provision which will 'trigger' GPL source distribution requirements for a web service, at the option of the copyright holder; that is really the best choice. Rather than getting into an enormous philosophical debate over whether the idea is "good" or "bad" or "punitive" or whatever, let's simply have two clear licenses and give the option to the copyright holders to decide under what terms they will license their property.
Software for the Rest of Us (Score:5, Insightful)
This will stop many programmers, many of the best programmers, from using source code under that license. We could no longer keep any of the value of the software we created to ourselves. What is a "web service"? Is my email-processing CGI a "web service"? Any software in the same workflow as any other software under this license would have to be released. So many developers won't make small customizations, because that would force us into the source code distribution business, with all its overhead. Or we might just ignore that provision, or the whole license, en masse.
The GPL is successful because it is a fair contract, even though it's revolutionary. Its enforcement teeth are rarely tested, because it's so close to an equitable agreement among peers. Which has resulted in lots of value contributed by profit-driven organizations, despite the claims of many that the license is anticapitalist. Upping the ante, to require private customizations to be published, could stop the rising tide of acceptance that is pushing GPL to be the default, and any proprietary license to be radical. And then the caution it would inspire: investing in GPL'd software might force acceptance of ever-more demanding licenses, like a GPL4.0 that required redistribution of even software that wasn't changed at all, just to get users "to pull their weight".
The GPL2.0 isn't broken. Let's not "fix it" in a way that could destroy its success, and our chances to benefit from one another's work without onerous burdens.
Re:Where did this mindset come from? (Score:5, Insightful)
Like voting? And do corporations get the same negative sides as private citizens, like going to jail? If you won't send a company to jail, and give it a vote, you can't equate them to private citizens
Re:Sensationalism at its finest (Score:3, Insightful)
"If you look at the market, Yahoo, eBay, IBM, Amazon, Google have all sunk millions into the GPL infrastructure," Olson said. "Not only are we changing the rules, we are changing them retroactively. With the new way, it lets the customer pay with either their source code or with their wallet."
Re:This doesn't sound like a good idea (Score:2, Insightful)
Um, you mean they are not?
*You* misunderstand (Score:1, Insightful)
From the first "you", you learn that "you"="code's user". It's called English reading skills.
commercial freeloaders (Score:2, Insightful)
Just becasue something is free as in beer doesnt mean it is devoid of all responsability from the users behalf.
If a company's buisness model is dependent on free software then its in their companies best itnerest to be very generous to the programers who maintain the software they require.
It sounds like the GPL v3.0 is trying to make the commercial world a little bit more responsable.
Re:Future versions of the GPL (Score:2, Insightful)
Re:Submitter mischaracterises the change. (Score:5, Insightful)
This is deeply irresponsible. Any project that ships software under the GPL is going to be spinning their wheels for months over this, and the Microsofts of the world just got a huge weapon to use against OSS usage. After all, now they can say that using GPL software not only costs you in terms of the usual TCO metrics, but there's a potentially hidden and as-yet-unknown cost that can be applied to retroactively!
Grrr!
Re:Bad, bad, bad! (Score:4, Insightful)
How much source code have they released ?
Re:Bad, bad, bad! (Score:1, Insightful)
The point is that other companies can see these successful, high-volume sites running on free software.
Those companies might adopt free software alternatives, and they might contribute their code.
These companies aren't going to be impressed if all they see are weirdy-beardy college students and aging hippies talking about free software.
Re:Future versions of the GPL (Score:5, Insightful)
Programmers can be affected with the new provision (Score:3, Insightful)
It appears that closing this loophole will also close the doors for programmers to freelance in this manner. That is they won't be able to sell their programming service of enhancing a current GPL'd project -- unless, of course, the solicitor agrees to either pay or release the code. The other option is to force the programmer to pay the GPL fee and roll this expense into the contract costs. I think this issue has to be debated and discussed at length, because we can't go about and make a gut decision of saying this provision is a good addition to the GPL just because we want to make big companies like Google, Amazon, etc. pay. After all, it could affect freelance programmers -- this could very well be their bread-n-butter. Let's not get into the question of whether personal enhancements of GPL'd programming tools (e.g. IDE's) are required to be paid or released under the proposed provision, that could be a sticky situation.
Re:This doesn't sound like a good idea (Score:2, Insightful)
Why do they have to benefit? Turn your statement around, how are they harmed? They aren't. I think I understand the good intent of this proposal but I don't agree with it. It's good for the computing industry as a whole if companies can pull from open software. The benefit potentially is to our economy and in keeping us all employed.
So long as companies are making improvements on open software there is the potential that the code could make it back to the public. If companies shy away from an insistance that they open their code (and I believe that is truly what this is about) then no one benefits. Not even the open source developers.
The GPL must change or die. (Score:5, Insightful)
That's not the same as you distributing a document you created. In that case, INPUT occurs on your computer, PROCESSING occurs on your computer, and OUTPUT occurs on your computer.
The question is: What's important in determining if a program has been "distributed"? Clearly if I give you a CD with the program and you run it on your computer, I've distributed the program. Clearly if I take GPL software and modify it and I run it on my computer only, I have not distributed the program.
But if I take GPL'd software, modify it, and then let YOU run it on my computer...
Did I distribute it, or not?
Your immediate answer is probably "Who cares?" But now what if I charge you to run this modified program on my computer, for example, by charging a fee if you use my auction program? Now I'm using software provided by the open source community for my financial gain, but not returning the modifications I made to that software to the community.
The extreme of this problem is that eventually, the internet becomes so fast and clients become so dumb that software is never "distributed" at all. I take an open source office suite and then modify it. According to the GPL, if I then sell that software on CD, or by download, so that people can actually run it on their computers, I must provide the source to it as well. But what if I'd rather just make money off of the GPL'd software I've taken without giving anything back to the community?
Well, then I just put the software on my own server, and instead of selling CDs or downloads, I let people provide input and receive output over a remote connection to the program running on my server.
And wala! People can modify and essentially provide GPL'd software without having to provide source.
GPL 3.0 and "servevices" (Score:4, Insightful)
Currently the GPL forbids redistribution of GPL-covered code linked with other code to form a larger work, when that other code is not also distributed under the GPL: the whole larger work must be so distributed. It is a "derived work", and copyright law is clear on this: one can not redistribute derivatives of copyright works without permission. It's that permission that the GPL grants, when you abide by its terms.
Of course, you are not prevented from redistributing GPL code *aggregated* with non-GPL code.
The problem arises from distinguishing between derived and aggregated works: what if I distribute a GPL app that I write, with a bunch of GPL shared libraries that I didn't write (complete with source), and a nonh-GPL proprietary library, essential for the application, that I did write, without which the application is useless.
Is that an aggregate, or a derived work of all the GPL code I did not write?
On the one hand, there's no (legal) requirement that my app even run, and, since I distribute source, one can implement a version of the non-GPL library. The fact that it works when my non-GPL library is installed in the right directory (LD_LIBRARY_PATH, anyone?) is a happy coincidence. Redistributing my GPL app, with modifications is fine, but you can't redistribute my "essential" non-GPL library, rendering the app rather useless without it. You add value to the app, I license more copies of the essential library.
On the other hand, the proprietary library, the foreign GPL libraries, and my app together, constitute a derived work of the GPL libraries, and must be redistributed en masse under the GPL.
If the libraries are statically linked into a monolithic executable, there is little argument that the whole source must be redistributed. The situation is more controversial if a mere aggregate of files is distributed: some argue that not distributing the proprietary library under the GPL is a violation, other's don't.
I tend to believe that if the GPL-covered peices can be redistributed indivudually, without the proprietary library, regardless of whether the result is useless, there is no violation: the fact that the code is actually only useful when the proprietary library is present is but a happy coincidence. What if I distributed a GPL version of the library, and offered a proprietary replacement, with far better performance?
This controvery gets even muddier when one considers alternate ways of effecting program linkage. In effect, the "functional derived work" exists only at run-time, and, indeed, the maner in which the parts of the aggregate are combined, can, itself, be the subject of restrive licencing, and patent. What about linkage via a remote procedure call mechanism? Y'all remember RPC/XDR over TCP/IP, right? Suddenly, self-assembling functional derived works become a reality. Protocols like SOAP, used to support "Web Services" exacerbate the problem.
This leaves a big, gaping hole in the GPL: socket wrappers were a common "hack" to "get around" the GPL: just layer a RPC mechansism around the proprietary library, and a GPL wrapper to call it from the app, and you were all set. And, that didn't even address the issue of inter-machine communication: 127.0.0.1 and Unix sockets are ubiquitous.
Of course, the minute one's app "links" with proprietary code on a different server in this manner, the GPL loses all force, for one did non distribute the proprietary part.
It is this area that the GPL v. 3.0 tries to address, IMHO, perhaps by more differentiating between linkage and aggregation. This can be done, of course, but then the license starts to lose some of its roots in copyright law.
The problem with such an approach, though, is, what constitutes linkage: does an exchange of HTTP requests and responses? What if there is ultimately GPL code satisfying that request? Surely, if I use a program provided by a vendor
Re:Future versions of the GPL (Score:4, Insightful)
IE... the first version...
Did the first version have a version number in anticipation of future versions?
Re:Future versions of the GPL (Score:5, Insightful)
It's dangerous to leave such power in the hands of people not protected by insanity.
most clever coclusion I've heard in years. Funny also because of the inside pardox. But very true.
Re:Future versions of the GPL (Score:3, Insightful)
FSF must have had their heads in their ass (Score:2, Insightful)
It could be argued that the ASP model of computing is the most natural. The late 80s, 90s and early 00s are a freakish period where computers were cheaper then networks. Think clients are not very powerfull, and with extreemly fast networks, accessing extreemly powerfull and reliable centerlized machines. Duh.
The FSF must have had their heads in their asses if they diddnt see this coming.
How is this legally possible? (Score:4, Insightful)
For this to work as described, the user would have had to sign a contract. Otherwise, it will be just as unenforceable as a EULA (you already had a right to click the Next button on your own computer, etc).
Re:Future versions of the GPL (Score:3, Insightful)
If developers using the GPL want to make money off of companies using the code for commercial purposes, then do what Trolltech did with Qt and dual license.
Personally, I don't think the FSF has any place getting into the economics of the business model - their only say should be in the fate of the code. In fact, v2 essentially states that that won't get involved in how you make money when they basically say "Sell it if you want, for however much you want, all we care about is that if you distribute, you have to give away machine readable source as well, licensed under this license."
If you can't even modify the code without PAYING someone - well, let's just say that it wouldn't be FREE software, now would it? Now I'll concede, they're not demanding money, per se. FTFA:
"If you look at the market, Yahoo, eBay, IBM, Amazon, Google have all sunk millions into the GPL infrastructure," Olson said. "Not only are we changing the rules, we are changing them retroactively. With the new way, it lets the customer pay with either their source code or with their wallet."
I think there may be a place for "paying" with the modified source, but let's not get into the "paying with your wallet" part, OK? If you want money, as I said before, just dual license.
Great, thanks asshats (Score:3, Insightful)
The GPL and the Communist Manifesto (Score:3, Insightful)
The GPL is a single point of control over the vast majority of the FOSS movement (~60-75% of all projects according to Wikipedia). The wording of the GPL impacts not only the fate of the FOSS movement itself, not only the fate of the work product of each individual participating in this movement, but also the fate of all the companies (large and small) that have chosen to assume the risk of depending on this software.
Given the GPL's extreme importance to such a large and growing audience, we should all take a hard look at who really controls it. The GPL is controlled by the FSF. The FSF is a 501(c)(3) non-profit; it has a board of directors who have responsibility for oversight of the President who runs FSF day-to-day (RMS is the president). The board of directors is elected by "Members" (NOTE: If you join the FSF off their website, you are an Associate Member which is a NON-VOTING position). I'm not sure how one gets to be a full Member.
Now let's not kid each other here: We all know how sketchy the oversight of a Board of Directors can be; we all know that groups of theoretical "equals" can be strongly influenced by a small number of strong personalities; and we all know that "Strong Personality" is a very accurate two word description of RMS (and I doubt Eben Moglen is far behind). I think those among us who are objective (and especially those among us who have personally interacted with RMS) can agree that RMS is on a personal jihad and he's using the force of law to achieve his ends. The question is, do we all want to participate in that same jihad?
Of course, one can respond by saying: "There is a competitive market for licenses - no one is forcing developers to use the GPL." This might be true on paper but it's not true in practice. We're software developers, not lawyers. I'm willing to bet that the vast majority of developers release their work under the GPL simply because "that's the one I've heard the most about and that's what everybody else does", and NOT because they researched other available licenses and arrived at an understanding of their implications, and especially an understanding of who controls these licenses.
So where does that leave us? We have a single pseudo-fanatic who has substantial control over the single most important component of our movement. This should scare the shit out of all of you.
Me? I'm going to either change to a non-GPL license or stay with GPLv2.
It's funny how the FOSS movement parallels Russian history: RMS (Lenin) started a revolution to overthrow proprietary software (the Czars). But the new regime became a dictatorship (Communism). It's now time for the proletarians (FOSS developers) to revolt again and achieve self-determination (free, accountable, and transparent democracy).
Re:Future versions of the GPL (Score:4, Insightful)
Seriously, if something like this happens, it is simply going to make more people say that people like Jonathan Schwartz and Bill Gates were right all along not just about the GPL (few people make distinctions about which version they are speaking of when it comes to the GPL) but also in regards to RMS.
I haven't written -much- under the GPL, but what I have I never intended to force the user to be required to distribute -any- change so long as they never polluted the world with bastardized
Instead of modifying the "GPL" for this purpose, create the sGPL (strict GPL). I'm sure it will be useful for people, but it inherently changes the GPL too much.
Re:They deliver HTML. (Score:3, Insightful)
Re:Google and Amazon will be unaffected (Score:2, Insightful)
but regardless of any actual litigation, the water has been poisoned.
here's the thing, or actually several things...
a. companies are very risk averse - litigation is expensive and potentially deadly.
b. The GPL has never been formally defined.
b1. The GPL reads like the Port Huron statement - it's not a standard legal document.
b2. This makes it nearly impossible to render a confident legal opinion on exactly what is and is not a valid condition/restriction of the GPL.
b3. the FSF is now claiming that activation of the GPL doesn't actually require redistribution.
c. The FSF is apparently declaring their intent to impose this new interpretation RETROACTIVELY.
d. as you ask, IS THIS EVEN LEGAL - probably not, but no one really knows at this point.
e. If it's not binding the GPL is invalid.
f. If it is binding the GPL is a HUGE LIABILITY.
g. The GPL is either a joke or a HUGE LIABILITY.
h. companies don't want to invest in jokes a/o HUGE LIABILITIES - though they've been known to
i. this begs the question - IS THE FSF A HUGE LIABILITY ?
j. a-i = HUGE LIABILTY = RISKY = DONT USE FREE SOFTWARE.
What I'm getting at is that this event will signal to those interests who currently invest in software that GPL software can't be trusted, there are too many legal issues, the governing authority is capricious and likely to act in bad faith, they resent commercial applications of the license, and are litigious. Whether or not this results in a court case, the use of GPL sofware will almost certainly decline.
Re:Who what when where? (Score:2, Insightful)
So typical of Stallman. It's so clear that he's a megalomaniac where it's all about control and not about free code. Look at how he treated Ulrich Drepper and how he compared proprietary software developers to murderers and perjuring cops.
http://www.gnu.org/philosophy/my_doom.html [gnu.org]
http://slashdot.org/article.pl?sid=01/08/19/20392
Re:Stop your FUD (Score:5, Insightful)
It's all very well to trust the current board, but when you say "any future license", you are trusting all future boards, and 20 years from now, every single member will be someone you don't now know.
Re:Who what when where? (Score:2, Insightful)
Re:Future versions of the GPL (Score:3, Insightful)
Re:FSF? (Score:3, Insightful)
Free software is quite simple to clarify: "free speech, not free beer." Anyone who hears that will get it. Look how short and simple the Free Software definition [gnu.org] is; just four bullet points. Compare that to the Open Source definition [opensource.org]. It may not make as good a buzzword, but "free software" ultimately does the best job getting the point across.
Re:The GPL must change or die. (Score:3, Insightful)
Re:Future versions of the GPL (Score:2, Insightful)
Remind me again, what the road to hell is paved with.