If the punishment for blatantly violating the GPL or LGPL for years at a time while raking in millions of dollars in profits is getting bitched at on the internet, it seems likely to me that a lot of companies will do the cost-benefit analysis and decide that GPL compliance isn't really worth the hassle.
It was definitely made clear to the offenders that they would be taken to court, and at that point most of them opted to settle. For those that didn't, the court stuff was mostly uneventful. The judge would tell them they had to comply with the license, they would either stonewall or give us a (usually non-working) dump of their code & build system, ad infinitum, until we told them we were going to ask the judge for an injunction. That was usually sufficient motivation.
The reason there are no court cases is because we had them dead to rights and their lawyers knew it. Their lawyers told them there was no point in fighting, and to just give us what we wanted.
The lawyers are on staff at those organizations, but I'm not familiar with the details of how their funding works. They are both non-profits, though, so if you're curious you can pull their Form 990s.
I don't know what the monetary damages were. I was never purview to those conversations (as I said, I was only handling the technical side of things). I heard through the grape vine that some of them requested damages as well as compliance, but I don't know for sure.
Source: I run PortableApps.com and have had my GPL software used illegally by competing projects in other countries as well as at least one Fortune 500 here in the states.
For Linux and OSX, it would be cool to see similar things.. though I've seen some interesting things using Dropbox that may yeild a portable *nix app solution one day.
It would be great if the code of that library (hardware acceleration for audio/video decoding/encoding) was released with an open license. I hope it happens.
What is sad, is that the OP in the Google Group thread is using the "bitch about it hard and loud" strategy. Trying to be negative and so on. It's not a constructive attitude.
Working with companies like this is often like working with young children. Usually, calmly saying "please don't do that" doesn't work (and, indeed, doesn't appear to be working with Allwinner after 4 years of pointing out GPL violations); you have to make it plain and clear that what the child / company is doing is bad behavior, and that it's making you or someone else upset.
I'd be more inclined to take your side on this and criticize Luc if Allwinner actually had a good track record of responding immediately and adequately to such (L)GPL violations. From what I can tell, however, neither have been the case, and I therefore completely understand (and even agree with) Luc's visible hostility.
That's why we see as few lawsuits as we do.
I think an author with a proper lawyer would be able to use the DMCA to shut them down rather quickly - meaning it won't drag out for ages.
That guy who started the thread on Google Groups (Luc), is prone to polarizing the discussion. I suppose that Allwinner can fix the issue and still keep their library closed. That would be bad for everyone.
What you want is code that you can commit back into your reposority but what you get is some code that is quickly hacked together and you don't even want to look at.
Freescale does an excellent job of putting all of their stuff up on Github. The iMX6 family is a joy to work with thanks to their efforts.
They also go above and beyond with documentation. No NDA/MDA needed. Tens of thousands of pages of freely available technical reference including the entire TRMs for the SoCs:
For instance, the vpu kernel driver (GPL license) is now open:
And the libstagefright (includes libvpu) as Apache license:
The Mali GPU driver is still stuck in proprietary limbo though, AFAIK. Rockchip can't release the code without ARM's permission, and ARM has only released partial support as of now. They claim that it's working, though I am still trying to get a complete X11/Wayland driver running myself (mostly due to all the android dependencies which I don't need under pure Linux)
So things are getting better, though there is still some wide gaping holes in development.
That depends on the end manufacturer. There are lots of Mediatek source code dumps on Github, for example. The problem is that the chipset manufacturers don't actively tell their customers what they are obliged to do under the license - and so they assume that the chipset vendor has given the code and docs under "gongkai" terms.
It's a real shame because their SoCs are cheap, powerful and very popular amongst Linux/ARM hobbyists - see http://linux-sunxi.org/Category:Devices
* Well, AllWinner cut and paste ARM cores together. I wonder if ARM have this much trouble with licensing too?
As far as I know, no. To actually produce chips Allwinner and others have to go through one of a small number of fabs. TSMC, Globalfoundries, UMC and Samsung will not make chips with pirated blocks.
What are some alternatives that have better track records with the GPL?
Developing for mainline might be slower but in the long-run it's better and even cheaper. I think that they are seeing this now.
You can check the slides at https://fosdem.org/2015/schedule/event/allwinner_upstream/ for the current level of support. This is about the community work for mainline support.
Though it is possible. I got mainline Linux running on an A10.
The GPL is "we require that you share and provide freedoms to others"
In other words, copyright is usually used to block people from being good sharing neighbors. GPL requires sharing. Not sharing is wrong. So, copyright for music is against the natural tendency to be a good neighbor and people rightly blame the copyright restrictions for being awful. GPL violations means you are both infringing copyright and being a bad citizen, bad neighbor.
How does not paying for works not hurt sharing? If return on investment on works is being hurt then that's a demotivator for future works. Monument Valley had something like a 95% pirating rate on Android. That affects the expectations of indie devs and investors. Brushing this under the carpet doesn't change the facts.
If pirating was more under control, we'd see more innovation and top notch applications in the Android market. Instead, we see all the top notch applications go to iOS either exclusively or only to Android after they've made money on iOS and then publishers decide to push out a port because its just going to get pirated anyway. The difference between these two ecosystems is proof enough that pirating games hurts innovation, companies, and customers.
You certainly can ask people if they want to share their own work. You can't take it for granted that people have a moral obligation to share the products of their labor. People might say if you don't agree with the license "then stop using GPL code", similar to how someone might say "then stop using copyrighted music/movies/games". It seems to me, that these statements are equivalent, but the attitudes of people are not the same.
I personally choose to not live in such bubbles.
And the general lack of compatibility between western and eastern models of imaginary property.
In this example allwinner is applying the eastern "share, modify, pass around" model, which isn't compatible with the western model of strict license adherence.
In Bunnie's case they're trying to adapt hardware/software developed with no apparent western-style license to a functional project.
I don't think your comment matches reality. Eastern "share, modify, pass around"? How does that even remotely match what is discussed in the google group thread? What All Winner did was to take free and open source GPL code paid for by the community (including competitors), modify said code, make binaries, ship defective non-compliant hardware and software BSP to customers and expect everyone to be silent about it. Nope. That's not an "Eastern model" at all. No need to hide their behavior behind shadowy cultural claims. AllWinner knows what they're doing is illegal.
As for not sharing the source at all, I agree, this is why I think the analogy doesn't hold, but just for reference, the Chinese companies in Bunnie's case don't officially share share anything either, and yet in practice everything is shared. It's piracy, but no one minds because they all benefit from it.
It does mean that if you crack open any product built around Allwinner (or other similar SOC's), you'll find PCB's that are all close variations of each other, and if you look at the code accompanying it, you'll find plenty of "sharing" going on even if they're not sharing the source with us.
How is that different than the gist of the license. And isn't the whole problem here that they didn't share?
GPL violations as far as I understand it are always a matter to be settled between the original authors and the offenders. I am not aware that the GPL gives bystanders and angry fanboys'n girls the right to sue anyone.
The author(A) allows the distributor(D) to distribute their(A) code so long as they(D) provide the source plus any modifications to the recipient.
The recipient has no right to demand the code due to there being no agreement to provide it between the distributor and the recipient.
So the author, on behalf of the recipient, must be the one to enforce the conditions of the agreement they made with the distributor.
In cases involving projects like Linux or BusyBox this can be anyone who has committed code, as these projects do not require copyright assignment.
In cases involving projects like GNU and Ubuntu all committed code is owned by the parent organization, the FSF and Canonical respectively. So they are the only ones who can enforce the GPL.
That said, the end user can report violations to groups like The Software Freedom Law Center and GPL-Violations.org. These groups have contacts to various copyright holders and will assist with enforcing them on a pro bono basis.