LWN: Comments on "The kernel community confronts GPL enforcement"

Yeah, OpenWRT is a great example of ongoing corporate malfesance..

pizza — Thu, 17 Sep 2020 17:06:36 +0000

Sure, I'll be glad to help. I still have one EAP600 deployed, with two more stashed in a box.

I'll be happy to send one of them your way if that will help, but it will take a >2h round trip to retrieve it. (I'll add it to my list for the next time I drive out that way)

I also have an EAP1200H, based on a similar architecture.

I'd prefer to not post my contact info for scrapers to find, but I'm listed as the maintainer of Linux's CW1200 driver, so drop me a direct email and we'll work out the details.

Yeah, OpenWRT is a great example of ongoing corporate malfesance..

mpratt14 — Thu, 17 Sep 2020 15:22:44 +0000

Hi there, hope this is not 4 years too late...

I have done a few PRs for Engenius boards now, we have figured out the 1.5 MB kernel size limit which was a major issue in adding support to the new kernel codebase for a lot of their products (may or may not apply to the EAP600, their newer software is different).

I am working on the EAP350 right now, but I don't have an EAP600, would you be interested in helping?

The kernel community confronts GPL enforcement

gioele — Fri, 16 Sep 2016 09:00:27 +0000

A list of Linux-based cheap electronics that publish all the GPL sources of the code they have used would be much appreciated.

The kernel community confronts GPL enforcement

paulj — Fri, 16 Sep 2016 06:56:46 +0000

The number of vendors of Linux based, cheap electronics, who do this are minuscule, I think.

The kernel community confronts GPL enforcement

Wol — Wed, 14 Sep 2016 17:21:27 +0000

They put their build system on a git server, which they push to a public repository every time they do a release ...

In other words, they do what they should be doing for their internal systems anyway, and one of the steps in "release to manufacturing" is "tag the release and push to the public server".

Once they've fixed their build system (not necessarily a popular job, I will admit ...) it's then just 5 minutes in the release process, which they've probably saved several times over by having a proper build system. But it's the usual PHB syndrome - "you've not got time to fix the problem properly, I need you firefighting!"

Cheers,
Wol

The kernel community confronts GPL enforcement

Wol — Wed, 14 Sep 2016 17:06:26 +0000

> If you don't have such a contract clause, you're in trouble. All you can do is run a 2013 kernel, and you still get to be liable if any of your customers asks for sources. Good luck if it's a consumer product, you'll need it. And get your legal team of their asses writing such contract terms...

At least, in the UK, I'm pretty certain you can conjoin your supplier as defendant. "They didn't give us the source, so we can't pass it on. Please Judge will you put them in the dock with us ...". Then the jury or whatever can acquit or convict the defendants separately ...

Cheers,
Wol

The kernel community confronts GPL enforcement

farnz — Wed, 14 Sep 2016 08:33:13 +0000

Depends in great detail on the right in question; most rights of a purchaser can be signed away, as long as you the consumer are giving informed agreement to the signing away of the right in question. The hard part is getting that informed agreement - to successfully get a consumer to sign away some of the stickier rights, you must spend quite a lot of effort telling the consumer that they're not getting what they want. For example, to sign away the right to a functioning product, you must ensure that the consumer knows that the product is not useful for the purposes they have in mind, and that you don't believe it's possible for the consumer to make it useful (but, on the other hand, this is how you can sell a faulty car to a consumer who just wants it for scrap value - as long as you're honest about what parts you've removed, you can sell a car saying "engine's bust and I don't know why - it started giving out black smoke and not making power. I've not taken any parts off the car, but I don't believe it's economically repairable; at the very least, it needs a new engine").

The kernel community confronts GPL enforcement

farnz — Wed, 14 Sep 2016 07:56:42 +0000

I think that's the best outcome possible - the "procedural grounds" were that Philip Morris had no standing to bring such a case, and that it was an abuse of the process for Philip Morris to attempt to rearrange its affairs specifically to allow one component of the firm to bring such a case.

The point of ISDS arbitration (which this was) is to provide a venue for companies to deal with capricious behaviour by states they've invested in (e.g. sell you permits to drill for oil, the moment you find oil, confiscate the oil wells), not to prevent states from doing anything that might reduce profits. In this case, PM was told that they were abusing the process by filing for compensation, because they had plenty of warning that such behaviour by Australia was expected, and they indeed restructured before plain packaging came in with a view to meeting the requirements the treaty sets out for ISDS - thus proving that this wasn't unexpected or unpredictable, but was in fact a normal business risk.

The kernel community confronts GPL enforcement

paulj — Tue, 13 Sep 2016 17:03:36 +0000

Go to your favoured electronics retailer, and look at all the stuff they have running Linux (routers, APs, light bulbs, etc.). Look at which of those are made by Vendor A types, and which by Vendor B. I suspect you'll find Vendor A products vastly outnumber Vendor B products.

I.e., Vendor A types are _not_ getting the hell sued of them, and the ratio of geeks:products and geeks:rest-of-population are both too low to make any difference, it would seem.

Note: Vendor B means they supply the full source for the actual product. Not just a dump of the stock upstream code /before/ they modified it. Oh, and they provide a build system that actually produces an image that could run on the product (mod DRM issues - Linus sadly is quite happy to not require that vendors give end-users the ability to install the software, least wrt to the kernel).

The kernel community confronts GPL enforcement

Wol — Tue, 13 Sep 2016 15:24:30 +0000

> Imagine you have vendor A that is a total douchebag. Meanwhile you have vendor B that are dedicated free software types.

> Vendor A gets their product out at a lower price point and faster because they couldn't give a damn about their customers and are willing to ship the first thing that compiles and doesn't crash immediately. They have no interest in supporting or improving the product once it is in the hands of customers.

> Vendor B takes a lot longer, has larger flash, and produces a superior product that is now much more open 'firmware'.

Except that is easily dealt with. Vendor A gets the hell sued out of them, and the geeks all moan blue murder about the crappy products.

Vendor B puts out that they have superior hardware and, while they very much don't make a song-and-dance about it (of course :-) they put out the same sort of crappy drivers as A, but they give pre-release hardware AND SOURCE to any kernel guys who want them.

And yes, B's engineers are under pressure to "just get the code out", but they also have personal pride, and a kernel dev or two to mentor them, and they will produce code that is half-way fit for upstreaming. NEVER play the "all or nothing" game - if there's a bear around, you don't need to outrun the bear, you just need to outrun your companion. B's advertising says "our hardware is better, and we work with the kernel guys". B's code may be just as much rubbish but the end user's experience is better.

Cheers,
Wol

The kernel community confronts GPL enforcement

Wol — Tue, 13 Sep 2016 15:13:35 +0000

> EU laws on liability tend to block the sort of "if you continue..." game that you're describing; if I'm not aware of the hidden fine print, you run foul of our "unfair contract terms" laws.

Even if you make me aware of them, if they're that sort of clause then they're probably illegal. Many rights CAN'T be signed away, and especially if I'm a consumer the law assumes that the supplier has complete knowledge (fit for purpose rules), and disproportionate power (they aren't allowed to do arm-twisting).

A judge is likely to say "are you having a giraffe?", and declare the clauses (and quite possibly the contract) void without giving it much thought at all.

Cheers,
Wol

The kernel community confronts GPL enforcement

Wol — Tue, 13 Sep 2016 15:05:15 +0000

> EU would probably have to wield a bigger hammer and end up having to do a complete product line ban from manufactures that are discovered doing so for *any* of it's products, from iot,mobile devices to tv's, dishwashers, refrigerators etc they all get banned until they comply with the one(s) that did not and arguably the same thing should apply for all devices that do not receive software updates ( open as well as closed ) in timely manner be it iot devices and or mobile devices that can be used for doing payments in one form or another etc.

This is EXACTLY the sort of thing the EU would do! If a manufacturer is shown to be flouting copyright they will be told "don't do it again!". And if they do, the flouting devices will be banned (which probably the manufacturer has discontinued, so they'll laugh at that). Except that next time, they will be told "You want an import licence? *Prove* you've complied, and then we'll get round to doing the paperwork!". OOOPPPSSS. Their competitor has just nicked the market from them ...

And it's stupid! All they've got to do is a "toss the code over the wall" dump and they've complied! And as for shipping compliant devices only to the EU, it's probably more hassle than it's worth to not comply for elsewhere, especially the States, because they'll suddenly find the Americans saying "if you can do it for Europe, why can't you do it for the US?".

And as for "we keep changing the code in the devices we ship", then just make your code change procedure including dumping the build environment every time you make a release. If that means a load of code ends up over the wall that never actually makes it into a device, so what.

Cheers,
Wol

The kernel community confronts GPL enforcement

mathstuf — Mon, 12 Sep 2016 21:17:19 +0000

Ah, was easy to find. http://www.bbc.com/news/world-asia-15815311

Seems to have ended OK though (not the best outcome: tossed on procedural grounds rather than actual arguments, but better than the reverse) http://www.mccabecentre.org/focus-areas/tobacco/philip-mo...

The kernel community confronts GPL enforcement

mathstuf — Mon, 12 Sep 2016 21:13:04 +0000

> These three between them mean that if you can get the EU to take action, business will listen.

Or lobby for a trade agreement where they can sue for lost profits based on laws passed in countries (cf. cigarette manufacturers suing Australia over the generic packaging laws).

The kernel community confronts GPL enforcement

johannbg — Mon, 12 Sep 2016 19:18:40 +0000

Product are being produced for different markets all the time and you can see it by the share number of existing standards.
If you need an recent simpler sample the non availability of LG V20 in EU at an time they could overtake Samsung in that same region since Samsung is in serious damage control with note 7.

In anycase we live in a world driven by greed and as such the solution is something that either increases or decreases the profit margin of the individual or corporate in question so if license B is more profitable than license A they will use that instead. It's as simple as that. In the end of the day what is effectively being discussed is how much enforcement can be put on the gpl before it becomes a negative value for companies in which the answer to that is none. . .

The kernel community confronts GPL enforcement

farnz — Mon, 12 Sep 2016 19:01:27 +0000

Three things come into play:

The EU is a huge market, covering several of the world's largest economies. Refusing to deal with the EU is like refusing to deal with the USA; you can't do it and stay big.
EU laws on liability tend to block the sort of "if you continue..." game that you're describing; if I'm not aware of the hidden fine print, you run foul of our "unfair contract terms" laws. In general, it takes active and clear intent by the end user to disclaim their statutory rights; I'm aware of one case where a consumer was willing to take this to court (Microsoft Windows NT 3.51 retail edition), and Microsoft chose to settle at the point where the consumer had the case moved to a precedent-setting court.
Other countries in geographic proximity to the EU (Turkey, UAE) tend to prefer European models of goods, as they're more convenient to buy than US models. If you're doing an EU model anyway, several non-EU countries will buy the EU model in preference to your "rest-of-world" model.

These three between them mean that if you can get the EU to take action, business will listen.

The kernel community confronts GPL enforcement

tialaramex — Mon, 12 Sep 2016 18:28:25 +0000

The EU is too big a market to just walk away, in GDP terms it's the biggest market in the world. And once you commit to NOT walking away your R&D people are going to constantly pester you, why build product A (for the Rest of the World) and product B (for the EU) rather than their proposed product C (for both) ? You're throwing money away.

The kernel community confronts GPL enforcement

johannbg — Mon, 12 Sep 2016 17:17:40 +0000

I'm not so sure if that would have any real effect if EU banned non-compliant devices since manufactures would just target other markets with said device in which they would get away with it or simply use something else then.

EU would probably have to wield a bigger hammer and end up having to do a complete product line ban from manufactures that are discovered doing so for *any* of it's products, from iot,mobile devices to tv's, dishwashers, refrigerators etc they all get banned until they comply with the one(s) that did not and arguably the same thing should apply for all devices that do not receive software updates ( open as well as closed ) in timely manner be it iot devices and or mobile devices that can be used for doing payments in one form or another etc.

And then there is that evolution that the consumer no longer actually owns any devices he purchases, which btw he is completely unaware of since no sales person is mentioning that fact or he might have relinquished his ownership through a simple gesture on that devices you know the classic "If you continue to use <insert product or application web sites, changes to terms in your bank, on loans, insurance etc> you acknowledge an apply our new means of fubar-ing you" which may include an hidden fine print in which the consumer relinquishes all rights in claiming any code or make the manufacturer liable in any shape or form...

The kernel community confronts GPL enforcement

Wol — Mon, 12 Sep 2016 15:04:04 +0000

Ask the EU to ban the import of non-compliant devices.

The US typically asks for damages for past transgressions. The EU typically seeks to enforce future compliance. Which jurisdiction is most likely to get the source for you?

Cheers,
Wol

The kernel community confronts GPL enforcement

CycoJ — Fri, 09 Sep 2016 21:41:34 +0000

Linus in their eagerness to attack SFC (BTW can Linus really not make an argument without ad hominem attacks?) and others really overlook the real danger that the situation is creating. I predict that we will see the emergence if copyright trolls in the next 5 years or so (one could argue this has already happened), some companies will acquire copyright in Linux code one way or the other and use that copyright for to extract money through trolling. I think this will be a much bigger danger to the community than "GPL communists". It's also funny how Linus accuses the SFC of ulterior motives, but then brings up IBM or other companies as examples of good court action. Since when do companies not have ulterior motives?!

The kernel community confronts GPL enforcement

rfontana — Wed, 07 Sep 2016 15:19:16 +0000

Here was the original rationale for the 'copyleft sunset' provision:
https://lists.fedorahosted.org/pipermail/copyleft-next/20...

The kernel community confronts GPL enforcement

flussence — Tue, 06 Sep 2016 17:28:25 +0000

> What are you going to do with 15-year old Linux?
Well, assuming my ISP-supplied DSL router isn't replaced with a differently awful piece of hardware within the next 5 years...

The kernel community confronts GPL enforcement

ballombe — Tue, 06 Sep 2016 17:27:21 +0000

The fact is there are very very few linux kernel related GPL-enforcement lawsuits.
So Linus claim that people are over-eager to sue falls a bit flat.

The kernel community confronts GPL enforcement

patrick_g — Tue, 06 Sep 2016 13:52:33 +0000

> So there's definitely value in 15 year old Linux

More than in a current and modern version of FreeBSD/OpenBSD/NetBSD ?
I doubt it.

The kernel community confronts GPL enforcement

paulj — Tue, 06 Sep 2016 09:46:49 +0000

Linux 2.2 and 2.4 worked well for a lot of people. As a bit of core kernel software to start with, to add your own drivers and what not, it'd be a good start. Certainly, it'd save one a _lot_ of work from starting a kernel from scratch (and note there are people doing exactly that to build a permissive licence kernel). So there's definitely value in 15 year old Linux. There's software I've worked on where the 15 year old code similarly would be a good start.

As "the rest" I mean non-copyleft, permissive open-source software and proprietary particularly. Until such time as there's a *general* sunset clause that makes _all_ X-years-old software permissively licensed, if I choose copyleft for some software instead of permissive then I did so for a reason, and that reason would not be invalidated by a relatively short amount of time if the copyright system still gives others 90+ years (and even then, doesn't require source release).

It doesn't seem a universally useful clause anyway.

The kernel community confronts GPL enforcement

seyman — Tue, 06 Sep 2016 08:35:50 +0000

> The problem is, so long as copyright generally doesn't have an (effective) sunset clause, then putting in one a copyleft licence would put copyleft sofware at a disadvantage to the rest.

I'm not sure what "the rest" refers to in the above sentence but I've always viewed copyleft as giving rights to people that copyright normally denies them. Putting software in the public domain early just goes one step further.

> Imagine if Linux code became permissive after 15 years?

The Linux code becoming permissive today would be early 2.4.x (2.4.0 was released in january 2001). It doesn't do much compared to a modern kernel, it would probably not compile with modern C compilers and would contain security holes whose fix are not under a permissive license. I don't see much of a problem.

The kernel community confronts GPL enforcement

Cyberax — Tue, 06 Sep 2016 08:06:14 +0000

What are you going to do with 15-year old Linux? It'll be pretty much useless these days without much hardware support (back then 2.4 kernel has not even been released!). And if your project has not improved significantly in 15 years then it probably makes no harm to relicense it permissively.

I kinda like that idea, and 15 years is a good enough interval.

The kernel community confronts GPL enforcement

paulj — Tue, 06 Sep 2016 07:44:03 +0000

The problem is, so long as copyright generally doesn't have an (effective) sunset clause, then putting in one a copyleft licence would put copyleft sofware at a disadvantage to the rest. If I copyleft something, I want it to stay copylefted so long as other software stays copyrighted under their terms. Imagine if Linux code became permissive after 15 years?

I would never put my code under a licence with such a clause. Either permissive is the correct choice or copyleft is, and I'd pick that from the start.

The kernel community confronts GPL enforcement

pabs — Tue, 06 Sep 2016 07:32:08 +0000

I suspect that is a clause that aims to kill off the "copyright is however damn long Disney wants it to be" problem within the small space of FLOSS.

The kernel community confronts GPL enforcement

paulj — Mon, 05 Sep 2016 12:39:08 +0000

Interesting, but some questionable clauses in there. Particularly the automatic sunsetting of copyleft after 15 years.

The kernel community confronts GPL enforcement

pabs — Mon, 05 Sep 2016 10:13:27 +0000

The experimental copyleft-next license attempts that:

https://github.com/copyleft-next/copyleft-next

The kernel community confronts GPL enforcement

pabs — Mon, 05 Sep 2016 10:12:58 +0000

The experimental copyleft-next license attempts that:

https://github.com/copyleft-next/copyleft-next

The kernel community confronts GPL enforcement

cas — Sun, 04 Sep 2016 13:04:29 +0000

> I know there's the tinkerer argument that Matthew likes to make: if I
> have a device, I should have the source code; it's useful to me. But
> in practice, is it actually that useful? That device is going to have
> a pretty short lifespan,

a device's lifespan is a lot longer than just the (possibly short) period of time that the manufacturer happens to sell it.

manufacturers may want you to throw away last year's model and buy a new one at least every year, but being a mindless consumer is not a legal obligation.

> because the manufacturer is unlikely to make enough margin on the
> product to maintain the hairball of code out of tree indefinitely.
> It's likely based on a 3 year old kernel version anyway, and good luck
> trying to get the patches forward-ported and then merged. The vendor
> won't live long enough to have made it worth the effort.

that's precisely why they should release the code, so that interested persons (users, tinkerers) can maintain it for themselves. fix broken shit, add features, disable anti-features, even patch it to work with newer kernels.

also....the manufacturer wants to make a profit. that's fine, i have no problem with that. however, it's not my problem - I don't owe them that, I'm not obliged to keep buying stuff from them just to contribute to their profit. in fact, once burnt, forever boycott - i might get caught out once by a manufacturer, but i won't ever buy anything again from manufacturers who screw me over on their GPL obligations. GPL code is not just a shortcut to profit where you don't have to pay the developers of the majority of your code...it comes with an obligation to provide complete source code with the binaries (or, failing that, the obligation to provide that source code on request to **ANY** third-party).

I strongly recommend my harsh and unforgiving boycott policy to all consumers of products with embedded GPL code. Manufacturers would be less likely to infringe if more people did that and resisted any temptation to sell out to the shiny.

> Do you really care about a device that much?

yes, I do. I don't want to have to buy a new phone or tablet or wifi router or whatever every year (or even every two or three or four years) just because there's a new model out and the vendor has abandoned their previous models (i.e. the one that I bought).

I buy new devices either when my current device has broken or when there has been sufficient increase in functionality that it's worth spending the money on a new one. e.g. my phone is an HTC Desire HD (2010?), and my tablet is an original 2012 Nexus 7. The phone's running CyanogenMod 7, and the tablet's running whatever google's latest release was (they stopped releasing updates for it a year or so back - my primary use for it is to run fbreader, so that doesn't bother me much...if/when it does, I'll eventually get around to installing CM or some AOSP-based "ROM" on it).

My ADSL modem is a Billion 7401VP that I've had for at least 10 years, running in bridged mode with pppoe (my linux gateway handles the firewall and routing)...i see no reason to replace it since nothing better than ADSL2 is available to me at the moment (and when NBN Fibre is available, I'll only need an ethernet port to plug into). If I used an openWRT style device then I'd want the complete source code for it, including all drivers for unusual components.

My wifi "access point" is a $15 ath9k-based TP-Link USB dongle plugged into the gateway box running hostapd, with a longish USB cable so it can hang off a plastic coat-hanger hooked over the picture-rail (fancy, glamorous, and ever-so-stylish) - it used to be an original eeepc 701 but that eventually died, so I had to quickly replace it.

I've only just started thinking that it's maybe time to start seriously researching options for replacing the phone - it still works for voice and SMS and the very small number of apps installed work OK. and a faster, newer tablet might be nice...but it's hard to see that upgrading either of them is worth the price (at least $300 for a significantly better tablet, and $400+ for a significantly better phone. or $600+ for a 5.5" or larger phablet to combine the two devices into one).

yes, i know...for a geek I don't have much of an "ooh shiny" gadget fetish. I'm keeping my geek card, though - gadget fetishism isn't a requirement. nor is wasteful consumerism.

money isn't worthless, and neither is the time I worked to earn the money to buy these things. I'm not going to throw them out and buy new ones just because some manufacturers want to sell me new shit. nor do i want to contribute to the ecological disaster of disposable consumption and e-waste any more than I absolutely have to.

which is, of course, part of the reason why the scumbag manufacturers don't want to release source. they want their products to die when they say they do, they don't want their customers keeping them alive and not buying new shit all the time. planned obsolescence isn't necessary if you can sucker people to keep on buying the same shit (maybe slightly better...or maybe slightly worse) in shiny new packaging.

The kernel community confronts GPL enforcement

cas — Sun, 04 Sep 2016 11:52:25 +0000

> Unfortunately those "bloody end users" have no legal standing with respect to the
> copyright on the code which may be being violated. An end user may have standing
> with regard to fitness-for-purpose or behaving-as-advertised, but not for some
> copyright claim between the vendor and their supplier.

one possible solution is for a GPL update, e.g. to e.g. GPLv2.1 and GPLv3.1 which adds a clause explicitly granting end-users of the code such standing, so that they *have* a court-enforcable right to get the source code for the hardware they've purchased.

I am not a lawyer so I have no idea how such a clause could or should be worded, but AFAICS there's no reason why it couldn't work (for the same reason that the GPL itself works: copyright law - if you don't agree to the license, then nothing else grants you permission to distribute source or binary derivatives)

OTH, such a clause may have unintended consequences...probably does, I've spent a grand total of two minutes thinking about it since I read your post.

Yeah, OpenWRT is a great example of ongoing corporate malfesance..

Darkmere — Sun, 04 Sep 2016 11:38:37 +0000

Ubi are.. Not doing well with their GPL part. And I'd love to see improvement there.

The kernel community confronts GPL enforcement

karath — Sat, 03 Sep 2016 08:40:02 +0000

Thank-you for the article!

The kernel community confronts GPL enforcement

raven667 — Fri, 02 Sep 2016 15:11:56 +0000

> companies doing mistakes go titsup to make space for more savvy companies

That's only the case for very small companies where they are providing a commodity in a competitive market where the customer has sufficient information to make a rational decision, the first rule of business is to get leverage over the market so that it is not competitive and the consumer does not have information to make a rational decision so that mistakes do not have the same penalties. In addition the timeline between a grievous error and a market correction can be so long that there effectively is no feedback loop, no learning or adjustment happens.

The kernel community confronts GPL enforcement

nhippi — Fri, 02 Sep 2016 13:38:17 +0000

True, the people involved may end up taking their company getting sued personally. Companies are not rational machines. People do incredibly bad decisions for their companies based on bad ideas ("sunk cost" fallacy comes up often...). But that's business - companies doing mistakes go titsup to make space for more savvy companies (or just companies that make different mistakes). But so far the companies that have stopped using Linux due to GPL threats is negligible. Cisco, sued for linksys violations, is platinum member of Linux foundation.

Yeah, OpenWRT is a great example of ongoing corporate malfesance..

pizza — Fri, 02 Sep 2016 12:07:47 +0000

Engenius (aka Senao).

I'm specifically referring to their EAP350 and EAP600 devices. They don't seem to have GPL sources posted at all for their newer EAP family members, though several other current models do.

Here's the thread I started on the gpl-violations list: http://legal.gpl-violations.narkive.com/tnqhGjxS/probable...

When I bought this stuff, it was because Engenius was the only game in town for high-powered devices. Ubiquiti may be an option today, but I think they're just as bad as Engenius when it comes to GPL code releases.

Going back to the EAP600, some sort of bug in the (proprietary madwifi) wifi stack has rendered 5GHz operation nearly unusable in my environment, and their ebtables binary is simply broken due to a mismatch between their running kernel and the headers used to compile it. To add insult to injury, I can't even build a toolchain to recompile this stuff for myself because uclibc needs the correct headers too.

The kernel community confronts GPL enforcement

paulj — Fri, 02 Sep 2016 09:39:01 +0000

It's a shame the GPL didn't also pass on some kind of sub-licence to sue over copyright violations to recipients of the code (in binary or source form).