If someone used BusyBox in their product, doesn't that normally mean in their ROM image? If so, surely they are obliged under the terms of the GPL to release the sources to their _whole_ ROM image, not just BusyBox...

Just go ahead and honor the license it's far cheaper than a lawsuit there isn't any good reason you have to have be lacking math skills not to be able to see why. Now would they have if they hadn't been called on it no probably not it would have saved them a bit not to but not anymore. Strapped for cash want to get product to market the cheapest way possible thats the name of the game here.

No, not unless the other code is linked with BusyBox.

It's somewhat analogous to what TiVo does. They run custom software on top of a Linux kernel. They provide source code to the kernel, but they're not obligated to provide source code for their app.

The GPL does not require the release of source code for where other binary code on the tape/disk/ROM is a "mere aggregation".

If it is a derived work of a GPL work then you do need to meet the conditions of the GPL. Linking is often used as a proxy for a work being a derived work, but this is not always so (eg, where the API being linked is controlled by a non-GPL-code standard).

Note that there may be GPL code in the device other than Busybox. But the copyright owners of Busybox only have standing to argue violations to their product's licensing.

You take someone to court for breach of license terms. They consult their lawyers who read the license and tell them to agree to your terms out of court. And this makes your case "dead in the water" ?

Clearly it doesn't. The reason we probably won't get this test case is that relevant legal precedents and the license are so clear that no-one is willing to fight the other side of it. If this is so then no test case is needed.

What's the betting that MS paid them to take a dive rather than set a precedent argued by lawyers MS has no control over.

Here's a nutty idea. An open source friendly company intentionally withholds source to some product that's clearly a derivative of a GPL product. Bungle the lawsuit, precedent set, Microsoft loses. Too bad that didn't happen.

> What's the betting that MS paid them to take a dive

If you're the MS board of directors you'd be negligent if you let some small company's small lawyers fight such a test case with such implications for MS. You'd be insane. So of course they'll have been in touch.

>>> You take someone to court for breach of license terms. They consult their lawyers who read the license and tell them to agree to your terms out of court. And this makes your case "dead in the water" ?

Well yes in the context that it was being viewed in.

An out of court settlement means nothing in the big scheme of things - it doesn't set a precedent. So far such a case has never gone to court and so in any future case that does, there is no prior case to bring up as in "in case x vs y, the court held that the licence was valid ..." The bigger the case, the more important previous case lore becomes when you are weighing up the risks if litigation.

It also means there are no previous cases the FSF / SFLC/ etc / etc can hold up in pulic and say "yes, the court HAS examined out licence and agreed that it is valid" - such an ability would be very useful in the "FUD wars" with Microsoft etc who are quite happy to have people believe that the GPL etc aren't valid.