Hacker News Reader

Comment by CWuestefeld | Hacker News Reader

Apparently the claim isn't that anyone copied their code, misrepresented it, or anything like that. The claim is that the defendants didn't make the original source code available. As I read the OP, the source code they're talking about isn't a modified version of the original package, but simply a direct usage of the exact package.

"You have to provide the source code, whether or not you modified the program," Williamson said. "Just distributing the program, even if you haven't made any changes yourself, you still have an obligation to provide the source code."

Sheesh. It's so important that an additional copy of the original source is published, that there needs to be a lawsuit over it.

I thought that the open-source movement was supposed to be about freeing information and code. But every day, it seems more and more like a passive-aggressive means of control.

View on HN

Not entirely silly. They also have to publish any modifications they make and any part of the software that is directly based on busybox - which is the part that is of interest to the authors and foss community. Anyways... it's not passive-aggressive: companies used the software, so they should respect the license - noone forced them to use this particular application. Busybox people probably sued the largest number of users over GPL so far - you'd think that big companies would know better than to repeat the mistake.

Why is that good? If someone ports busybox, or libc, or the compiler, or some other software to a new architecture / environment / ..., everyone gets that port / patch back for free. You're also more likely to get a stable custom firmware for that device once it's known what is running on the original box.

They also have to publish any modifications...everyone gets that port

While the article doesn't quite make it explicit, it strongly implies that there are no modifications. It seems to be purely a question of providing an additional archive of the source.

You're also more likely to get a stable custom firmware for that device once it's known what is running on the original box

No more so than simply documenting what packages at what revisions were used. How does the source itself help? (Yes, I can think of edge cases, like the project goes away, but in such a case there's no one left to defend the licensing terms anyway)

companies used the software, so they should respect the license

Yes. There's no question that they should. The license is the license, they chose to use it, and they are now obligated to live by its terms. If they're not doing so, the court must penalize them. That's life in the real world.

My complaint is with the terms themselves, and the enforcement of them. First, I don't see how an additional archive somewhere out there, that no one is ever going to look at anyway. is of any benefit. Second, I don't see how a lawsuit to enforce this is of any benefit (yes, they have the right to do so, but that doesn't mean they're not jerks).

Well... there are two ways to ensure that there are no modifications to the original software. One is to ask politely. The other is to enforce the license. I get the impression that if it was never enforced, there would be some companies who simply say - "yeah, we're using the original source, go away" whether they modify it or not. Since publishing the unmodified source is neither hard, nor time consuming, I think it's perfectly reasonable to enforce it on everyone.

What about the case where the company modifies the code and then publishes the original unmodified source as their work? How is "yeah, look at our unmodified source repo," any better than "yeah, we're using the original source, go away"

The license states that source must be provided "on a medium customarily used for software interchange". Honestly I would think that in this day and age, a link to a git repo or whatnot would qualify. I wonder if such issues would arise if there was simply a set of links to the source for the unmodified versions.

Of course, setting up a github fork for your company's branch of a project is relatively straightforward, and also solves such issues.

The GPL school is definitely in the passive-aggressive means of control camp. They don't want freedom unless it's their brand of free.

Disclaimer: IANAL. That said, I think part of the reason to enforce cases like this is that if this clause in the license goes unenforced for long enough with the IP holders not suing, then it becomes an implicit rewriting of the "must provide source" part of the license, and then weakens any later legal case.

I agree, the provision to provide a tarball of source is silly. It probably made more sense in the days of semi-disconnected FTP sites before http and Google made everything so much more discoverable.

GPLv3 explicitly says that you can link to the source being available elsewhere ("may be on a different server (operated by you or a third party)"), but if that server goes down, you have to come up with a new link. And of course, that only works if you're using the package unmodified; the source you link to has to be the source you compiled.

So it's not as if the FSF isn't trying to make things reasonable.

Reply on news.ycombinator.com