Patents and the GPL
March 27, 2008 3:37 pm freesoftware, General, workI was talking to someone yesterday (who will remain unnamed) about perhaps providing a modified version of some GPL software for them. Unfortunately, he told me that his hands were tied on the issue since a directive came from the head legal guys that the company was not to distribute any GPL software which might, eventually, be infringing on the company’s patents. Why? Because to do so is to make a promise, on behalf of the company, to provide a royaly-free worldwide irrevocable patent licence grant to users of the software. Once the Pandora’s box is opened, the patents are worthless.
At least one person has told me that the guy was probably just politely telling me that he didn’t want to pay for what I was offering, and that the whole patent thing was just an excuse. That’s certainly possible, but in this case, I don’t believe it to be so. I’ve heard “no thanks, we’re not interested” often enough that I know how to recognise it.
If this is true, I am sure that these guys are not alone – there are companies out there who are consciously not participating in free software projects for fear of losing the opportunity to monetise their patent portfolio.
Am I the only one who finds this state of affairs perverse?
March 27th, 2008 at 4:34 pm
Yeah, but it cut’s both ways… if a company does distribute their stuff under the GPL, then you have an equally powerful assurance that they are *not* going to assert their patents against you in the future (ie Sun releasing Java as GPL). The Free Software world will continue to happily evolve inside of it’s own ecosystem, and let the rest of the world collapse under the increasingly complex patent spaghetti disaster.
March 27th, 2008 at 4:55 pm
Perverse indeed. Even more so than in the physical world, patenting software with an aim of monetizing it means that you want to get something for nothing. If the software is good enough, it will sell itself.
March 27th, 2008 at 5:02 pm
I do find the US patent situation perverse and often baffling, though there doesn’t seem to be much political will to reform it.
March 27th, 2008 at 5:29 pm
No, you are not. GPL really does not encourage investments, never has and never will.
March 27th, 2008 at 5:33 pm
The patent-holder has the option to license the patent only for use by GPL software, and can still go after proprietary software vendors that infringe the patent. So it doesn’t become worthless.
March 27th, 2008 at 7:34 pm
@Joe Buck: The software vendor being pursued for patent infringement has a choice – pay the patent holder, or download a copy of the GPL software and get a royalty free licence. I know which I’d do.
@troll: You chose your name well. GPL is, to my mind, one of the few licences which encourages co-operation, and all innovation comes from co-operation. Investment and innovation are distinct things.
March 27th, 2008 at 11:29 pm
“@Joe Buck: The software vendor being pursued for patent infringement has a choice – pay the patent holder, or download a copy of the GPL software and get a royalty free licence. I know which I’d do.”
ianal but no, they also need to license their software under the gpl, that may not even be possible because of other non-gpl libraries used, and even if possible they still may want to pay to keep their product closed because of added revenue, their patents rights on their code, etc…
March 28th, 2008 at 8:34 am
At the risk of being ostracized: the idea of software patents is not bad. It’s the current implementation that sucks.
Some businesses profit from the confusion to make (short term) profits at the detriment of their customers and environment; but most businesses (and it is probably the case here) simply miss potential opportunities because of that very same confusion.
The GPL does not help. In its latest incarnation it is becoming more confusing and more restrictive (one man’s freedom is another man’s restriction).
It is more cohercitive than encouraging, and as an end result it is probably as divisive as some of the most stringent proprietary licenses.
March 29th, 2008 at 1:16 am
“At the risk of being ostracized: the idea of software patents is not bad. It’s the current implementation that sucks.”
I disagree totally: the idea of software patents is horrific. Even the most trivial program employs a bunch of algorithms which, if patented, would require developers to license patents all over the place and probably still expose themselves to the threat of litigation around yet more undiscovered patents. The main beneficiaries of software patents are to be found in the “intellectual property” litigation business, which sort of explains why it has been precisely that group of people who have been lobbying so hard for them.
“The GPL does not help. In its latest incarnation it is becoming more confusing and more restrictive (one man’s freedom is another man’s restriction).”
The usual nonsense, I see. The GPL and copyleft preserve the freedom for *everyone*, whereas permissively-licensed code can be closed off at the first opportunity and shipped as binaries to the “consumer”. Not much freedom for end-users in such a situation, is there?
March 29th, 2008 at 8:36 am
@The Badger: “Even the most trivial program employs a bunch of algorithms which, if patented, would require developers to license patents all over the place” – it sounds like you have a problem with the current implementation.
If software patents had a higher bar to pass before being accepted, if there was a peer review period pre-acceptance for objections to be listed, and if the burden of proving the worth of a patent was on the patent holder, rather than the inverse, and if the duration of patent protection in software were dramatically shortened, the patent system could be much more manageable – only significant advances could be patented, and those significant advances would come back into the commons at a time in the near future.
Those who argue that patents are purely a means to recuperate R&D costs must surely agree that in IT, where R&D costs are low, reducing the lifetime of a patent is a reasonable thing to do.
March 29th, 2008 at 5:29 pm
“If software patents had a higher bar to pass before being accepted”
I don’t buy into the “quality” argument for software patents. Firstly, progress in algorithms and mathematics is heavily based on prior work, so it seems like a form of theft to let someone patent an algorithm when there have most likely been a large number of people who have brought the domain to its current state before the patent applicant was able to make their filing. Especially in this regard, patents are a form of “intellectual robbery” not “intellectual property”.
“only significant advances could be patented”
Secondly, some of the most significant patents on algorithms are related to public key cryptography, and everyone knows that my first point applies substantially to those algorithms *and* that there wasn’t only one group of people who had made the required “breakthroughs”. Consequently, it seems unfair to grant monopolies to people when other people have reproduced the same work independently. Imagine doing the “hard work”, being unaware of some patent and then being sued for it: again, it’s intellectual robbery.
March 29th, 2008 at 11:06 pm
@Dave: that’s the right direction: shorter duration; better review; more careful awards (is it in Australia that a kid patented the wheel not a long time ago?).
Add to it cheaper, streamlined application process (to level the field between individual inventors and large corporations) and regulated / limited litigation (to keep patent trolls at bay).
Participating in a global, well functioning patent system should not be much different than participating in global open source software, or in the global academic publishing system.
@The Badger: thank you for ostracizing anonymously.
“The GPL and copyleft preserve the freedom for *everyone*” – are developers nobody? isn’t the GPL restricting their freedom?
Most of your arguments shows plenty of indoctrination; little understanding of the issues; and like the above, are plain *wrong*.
March 30th, 2008 at 4:38 pm
“are developers nobody? isn’t the GPL restricting their freedom?”
In fact the GPL does give them freedom, since it permits them to modify and distribute the software. Without a licence, recipients of software have no such freedoms at all, but I don’t suppose you were focusing on that aspect of copyright and licensing.
“Most of your arguments shows plenty of indoctrination; little understanding of the issues; and like the above, are plain *wrong*.”
Wrong, indeed! You wrote that “one man’s freedom is another man’s restriction”, but if one group (the end-users) sees too many restrictions, what the other group (the developers of derived works) enjoys starts to look more like privileges than freedoms. Copyleft is a well-known compromise which ensures that everyone enjoys the same basic set of things which can actually be regarded as freedoms.
You may not think that the GPL lets you do everything you may want as a developer, and that the freedoms offered aren’t rich or generous enough, but that’s precisely because the whole copyleft concept recognises that if you try and offer someone something more, someone else probably loses something in the process. A monarch might complain that he/she is no longer allowed to treat his/her subjects like personal possessions, but do we lament the monarch’s “lack of freedom”?
I’d be interested to hear how “indoctrination” and my supposed “little understanding of the issues” pervade my other arguments, especially given your assertion that “a global, well functioning patent system should not be much different than participating in global open source software”. Last time I checked, it was fairly obvious for someone producing an original work to be sure that they hadn’t infringed someone’s copyright, whereas there are no such guarantees around patents and patent infringements. But I suppose it’s easier to use terms like “indoctrination” than to elucidate on such matters.
April 1st, 2008 at 3:37 am
“the GPL does give them [the developers] freedom, since it permits them to modify and distribute the software.”
Developers own copyright to their own code. They don’t need your “permission”. By law they have *all* rights to the software they write. They only need to license (get permission for) third party code that they use; and when they do license third party code that is GPL’d they are restricted in their freedom – more than when they use non-copylefted licenses. So much for “the usual nonsense” and “preserving the freedom for *everyone*”.
I can refute almost every single statement of yours, but I have no interest to hijack Dave’s blog for an anonymous argument. You are welcome to continue this discussion on my blog at
http://panospace.wordpress.com/2008/03/28/the-genesis-of-software/ or in private.
April 2nd, 2008 at 10:25 am
“Developers own copyright to their own code. They don’t need your “permission”.”
They do if they’re distributing someone else’s code. Keep moving the goalposts!
“They only need to license (get permission for) third party code that they use;”
They need to adhere to the licensing terms of that third party code, yes.
“and when they do license third party code that is GPL’d they are restricted in their freedom”
No they aren’t! They can write whatever code they like and license it in whichever way they please. But when it comes to creating a derived work using someone else’s code, they are not in general allowed to do whatever they like. Still, even code which is only ever seen in the wild in combination with GPL-licensed software – take the LGPL-licensed KDE libraries as an example – may be licensed differently. It’s just the derived work that can’t be licensed in any way one pleases.
Even permissive licences actually stipulate various things that you have to do if you want to incorporate licensed code in derived works. It’s just that they don’t require that the source code be made available throughout the distribution chain. They may stipulate other things, however, including some which are actually related to the topic of this blog posting.
“- more than when they use non-copylefted licenses. So much for “the usual nonsense” and “preserving the freedom for *everyone*”.”
In general, according to most copyright regimes, you have *no* privileges whatsoever when it comes to using other people’s software unless they choose to license it appropriately. Anything they do grant to you is really a privilege that you otherwise wouldn’t have at all. In short, until they grant anything to you, you have *no* freedom. On this basis, claiming that people are “restricted in their freedom” is pure distortion.
You have a different idea of freedom than I do. For you, “freedom” is something the monarch thinks they have when they want to be able to do anything at anyone’s expense. For me, “freedom” is a set of rights that everyone handling the software should have. Your version of “freedom” just happens to have the side-effect of stripping end-users of any rights they may have had if people had behaved considerately. My version grants people rights without taking any away. It’s really as simple as that.
“I can refute almost every single statement of yours”
Well, you haven’t done a very good job so far. All you’ve managed to do is to redefine privileges as “freedoms” and to complain about the copyleft people taking away something (unfettered “freedom” to act however you like with other people’s work) that people actually don’t have in the first place.
“I have no interest to hijack Dave’s blog for an anonymous argument”
What do you expect when you come out with a cheap, peripheral shot at the GPL? It’s not as if the GPL is the only Free Software licence which has explicit language about patents in it. In fact, GPLv2 was more conservative than various permissive licences on this matter. I guess it just doesn’t suit your agenda to take a shot at, say, version 2.0 of the Apache License, but it’s probably just as relevant to the topic of this blog posting.
April 3rd, 2008 at 3:22 pm
The positive point is that the subject is alive : business people are aware of Licenses and Open Source.
It is a business choice which is public and as such impact company’s image and technology adoption.
IONA Open Source: Disbanded After Acquisition ?
http://www.theserverside.com/news/thread.tss?thread_id=48931
(Most likely Software AG doesn’t want the OSS part, but already 2 companies have confirmed their interest in it.)