Live data from Hacker News

BMW Australia Refusing to Comply with Terms of GNU Public License

gist.github.com

101–110 of 229 posts

Re: BMW Australia Refusing to Comply with Terms of GNU Public License

#101
post #4

What GPL licensed software are they actually infringing upon? The linked post[1] says they found mentions of systemd, but does the firmware actually contain a linux kernel? I don't see anything that looks like a kernel in the file listing. Also if the firmware is properly signed delivering it over http shouldn't matter, right? [1] https://shkspr.mobi/blog/2016/02/bmw-are-sending-their-softw...

Apparently Mini (also made by BMW) has a similar update structure, and that runs QNX, so it's likely this does too: http://rene.rebe.de/2013-03-31/mini-connected-runs-qnx/

check sum and signing are different methodologies.

Re: BMW Australia Refusing to Comply with Terms of GNU Public License

#102
post #92

It seems to me that they're not distributing any software (i.e., the product is a car, not software), which means licensing doesn't come into play. It would be akin to someone finding out that I'm privately using (and maybe modifying) GPL software, and then demand that I make the source code available to them. Only in GPLv3, the issue of selling a product that uses GPL software internally is addressed. The situation…

If the car has the software then they are distributing the software

Re: BMW Australia Refusing to Comply with Terms of GNU Public License

#103

Google could have helped here: BMW Car IT: Open Source http://www.bmw-carit.com/open-source/ I'll take a guess based on the filename "ConnStarter" in the BMW firmware that ConnMann is the primary GPL'd software. As the BMW site says: "The ConnMan project provides a daemon for managing internet connections within embedded devices running the Linux operating system.... ConnMan is available under the terms of the GPL v2…

The linked site may be useful, but where on that site did you find a download link for the specific software version (including modifications) used? All I see is a generic link to ConnMan's project website. I don't think that suffices?

I think BMW would argue that contributing their modifications directly to the ConnMan Git repository for others to use would suffice. You can find a list of BMW's commits to the ConnMan project here (Daniel Wagner seems to be the main developer from BMW Car IT working on it):

http://git.kernel.org/cgit/network/connman/connman.git/log/?...

Re: BMW Australia Refusing to Comply with Terms of GNU Public License

#104
post #94

Earlier quoted context omitted.

The linked site may be useful, but where on that site did you find a download link for the specific software version (including modifications) used? All I see is a generic link to ConnMan's project website. I don't think that suffices?

This was linked from the project website http://git.kernel.org/cgit/network/connman/connman.git/ edit Never mind, I misunderstood your comment

Yes, but that is just the ConnMan development repository. Correct me if I'm wrong, but I think BMW would need to redistribute the exact version of ConnMan that they used, including any modifications made. At least a deep link to a branch or tag would be nice.

Re: BMW Australia Refusing to Comply with Terms of GNU Public License

#105

I'm confused here. Where is the evidence that they've modified GPL'd software, which they are then distributing, and are refusing to distribute the source code for their modifications?

Not sure if this was the key part of your question, but they don't need to have modified it, they only need to have distributed it.

Re: BMW Australia Refusing to Comply with Terms of GNU Public License

#106
post #43

Earlier quoted context omitted.

Probably not much and that's why I don't put any licenses on my stuff.

So, you release code but don't let anyone use it? Because that's the legal effect of not putting licenses on your stuff.

Yep and I use code without license all the time. Show me one case where that has ever caused a problem.

Re: BMW Australia Refusing to Comply with Terms of GNU Public License

#107

History has proven over and over again that this kind of public shame and blame approach doesn't help in any way, quite the opposite. Hand this over to the Free Software Conservancy or the SFLC. And that should have been the first step. Now that the bridges are burning, BMW will be far less open to civil discussion and cooperative solutions. Well done. #sarcasm

> History has proven over and over again that this kind of public shame and blame approach doesn't help in any way, quite the opposite.

Can you please provide some evidence for this statement?

> Now that the bridges are burning, BMW will be far less open to civil discussion and cooperative solutions. Well done. #sarcasm

Depends on what your goal is. If your goal is to obtain the source code in this specific instance, and if you don't care about the efforts of open source contributors, then I guess it would make sense to privately "discuss" and "cooperate" with BMW.

On the other hand, if you want to scare companies into not stealing other people's work in the future, then I think naming and shaming is a reasonable strategy.

This is a clear case of rationalization about rules not applying to the powerful. Surely you would not expect BMW to care about "burning bridges" and value "cooperation" and "discussion" with you had they caught you stealing their IP.

Re: BMW Australia Refusing to Comply with Terms of GNU Public License

#108
post #20

Earlier quoted context omitted.

I don't really look at GPL too closely, since I do not code for a living, or make commercial products, but I have to say my superficial understanding of opensource and GPL was piqued by the statement from @jackhat above with the 'legal rights of the property creators as owners.' If I take this statement to mean that GPL is the mirror-image of proprietary license agreements in that it is not free to use, since the sti…

The GPL is essentially pay-it-forward; any rights you receive must be passed along to anyone who gets a copy of the work (or of a derivative). That means it's perfectly fine to take a GPL work and modify it without giving the source to anyone, as long as you also don't give binaries to anyone. And if you're distributing binaries privately with just a few people/companies, you also don't need to publish the source pub…

(IANAL)

This brings up an interesting point; which BMW might try to make if this ever goes to court.

There have been times when device manufacturers (and car manufacturers) attempt to prevent device owners from modifying the code on their own devices. I think once folks tried to drag the DMCA into it. There's a lot of muddling up of the ownership waters there.

This may mean that the ownership of the binaries can be similarly challenged. Somehow. A point can be made that a binary living inside a phone (or car) I have bought has not been "given" to me (and thus the source code need not be given to me either).

Granted, it's rather convoluted logic, but then again, this isn't the first time stuff like this has happened.

Would be interesting to see how such an argument actually pans out.

Re: BMW Australia Refusing to Comply with Terms of GNU Public License

#109
post #100

So it is strange for me even if I understand GPL. Does it means all the network routers running custom linux have to provide sources of entire modification? I assume if we want companies to make use of linux on a wide level, there should be the ways to overcome this problem(if exists).

No. There is a boundary between the "kernel" which is strictly GPL and the "user" operating system which is (in its interaction with the kernel) based on LGPL and weaker libraries. So these routers have to distribute changes they make to the kernel, yes, but not any and everything that runs on the router.

Re: BMW Australia Refusing to Comply with Terms of GNU Public License

#110
post #96

Earlier quoted context omitted.

If I sell a computer with Linux on it, and I bundle on my own application that's compiled to run on Linux, am I required to make the source code for that application available? In this case, if they just have an unmodified copy of Linux to provide the run-time environment for their nav system or whatever, and the nav system itself is an application on top of it, I don't know if it follows that they need to release th…

If it's an application, no, but if it's a kernel module (e.g. drivers to talk to their devices), then yes. They still have to offer a copy of the kernel source, even if they haven't modified it.

What makes you think that your distinction is the correct one legally? Has it been tested?
Post reply on HN