GPL patent grant for 19 patents

Posted 11 May 2000 at 20:55 UTC by raph Share This

I finally got around to doing a patent grant for my archive of patents. Probably the most immediately interesting patents are the screening algorithms, which can be used to make top-quality prints on inkjets.

At the same time, I've released a testbed implementation, called rinkj, that drives some Epson printers. I'm working with the Ghostscript and Gimp-print people on integrating these algorithms for real work.

Some of the other patents may be of interest too. If you're interested in playing with them, let me know, and I may be able to release some of my prototype code to a GPL release.

The patent grant itself is a very simple document. I've had my lawyer look at it, so hopefully it's legally sound. I would encourage all other holders of patents that may be relevant for GPL software, and who wish to help the Gnu project, to adopt this grant for their own patents. Here is the text:

Whereas, Raph Levien (hereinafter "Inventor") has obtained patent protection for related technology (hereinafter "Patented Technology"), Inventor wishes to aid the the GNU free software project in achieving its goals, and Inventor also wishes to increase public awareness of Patented Technology, Inventor hereby grants a fully paid up, nonexclusive, royalty free license to practice the patents listed below ("the Patents") if and only if practiced in conjunction with software distributed under the terms of any version of the GNU General Public License as published by the Free Software Foundation, 59 Temple Place, Suite 330, Boston, MA 02111. Inventor reserves all other rights, including without limitation, licensing for software not distributed under the GNU General Public License.

YAAAAAY! love it, raph, posted 11 May 2000 at 22:00 UTC by lkcl » (Master)

Hmmm, posted 12 May 2000 at 00:12 UTC by tjl » (Journeyer)

Is this really legally binding? I mean, someone could just put a notice like that on a web page somewhere and then sue the people who used his patents.

If I were to use such a grant, I'd really want to see an actual real binding signature on such a document. Maybe some body like SPI could collect actual documentation from people making such grants.

This is not to imply that you would ever do something like that but someone might, at some point. I'm considering getting a patent myself that I would then give a similar grant for and would like to see some way to make sure this works.

Further Question, posted 12 May 2000 at 01:14 UTC by alan » (Master)

Have people given thought to giving more favourable non GPL licenses to people who also GPL some of their patents in return. I'm just wondering how you build an incentive to spread this practice beyond the ethical one, which tends to carry limited weight in the western world today.

this is still evil, posted 12 May 2000 at 02:05 UTC by splork » (Master)

Patents are evil even if infected with the GPL license. They still stifles the use of a good idea by making it so that someone can't use it (in this case because they didn't release GPLed source code of their implementation of the idea).

I must be missing something in the GPL here, posted 12 May 2000 at 02:42 UTC by decklin » (Master)

My first reaction was that this must be violating the GPL (aka "the sky is falling!" ;-). c.f. the following in the preamble:

Finally, any free program is threatened constantly by software patents. We wish to avoid the danger that redistributors of a free program will individually obtain patent licenses, in effect making the program proprietary. To prevent this, we have made it clear that any patent must be licensed for everyone's free use or not licensed at all.

However, I can find nothing in the actual terms of the GPL which says anything to this effect. I'm going to mail GNU about this; I'm not sure if they really want to prevent you from licensing your patents as you did.

If I was in charge, though, I would. I think that ``open'' patents should be freely licensed to everyone, not just those who use the GPL. There's nothing in these specific patents that would really need to go into the core of, say, FreeBSD, but many hackers still prefer some other DFSG-free liscense to the ``GNU Public Virus''.

I have to agree with splork; patents are evil. I don't think they should be used as a wedge to get people to GPL their software, because we don't need the patent system to make the GPL work. Opening them up to everybody sends a better message: the whole system is a botch.

But to put things back in perspective, thanks for at least doing this. Only in the free software communitity can you do something good and still get criticized for it. ;-)

A little followup, posted 12 May 2000 at 04:53 UTC by raph » (Master)

tjl: Your point is a good one. I will be happy to send a signed copy of the grant to anyone who reasonably requests one (ie no spam, please). Also, I'm looking into publishing the grant more officially, for example by including a notice in the Federal Register. It's my understanding that IBM did that with the Coppersmith patent on DES.

splork: I understand your reservations. Obviously, I'm trying to balance a number of different things here, among them ability to use the ideas in free software and ability for me to make money through licensing. Obviously, if I released the patent under a very nonrestrictive license, say X, I'd run the serious hazard that a potential licensee could just create (or .a) and link it to all their proprietary stuff. That's just not what I'm trying to do here.

I know a number of other people are working on ideas for patent pools, licenses, and so on, intended to more broadly help free software. I'm generally in favor of such things, but don't have the time or resources to push it forward myself. But if the work gets done, and there's a nice framework for free software licensing with reasonable protections against the stuff being simply ripped off for use in proprietary software, I'd be eager to participate.

Lastly, this might be analogous to enumerating the circles of hell, but I believe there are more evil and less evil patents, and do my best to cluster mine around the latter end of the spectrum. In particular, the screening patents represent some truly original ideas and a lot of work. They are not on the critical path of implementing any standards. They cover a specific implementation of high-quality error diffusion rather than the general idea, so anyone else is free to come up with different ways of achieving the same goals (for example, I believe Eschbach's work at Xerox on using templates in error diffusion was designed to produce similar results while staying clear of the patent - feel free to ask legal counsel at Xerox whether they're willing to grant use of any resulting patents for free software under broader terms than these). And finally, they are free for use in any GPL software, not just stuff derived from what I write (contrast, for example, with the patent grant in the MPL.

decklin: I don't have a thin skin about these issues, and do appreciate your acknowledgement that this grant is at least a step in the right direction. I agree the situation with patents is far from perfect, and as I say would be happy to work with people who are constructively trying to make things better.

Hm, are patents evil in this case?, posted 12 May 2000 at 07:39 UTC by Radagast » (Journeyer)

I agree with the general principle of patents being evil, of course (those who know me know that I'm over at the zealous end of the free software spectrum), but imagine what effects a few reasonably broad, useful and defendable patents licensed in this way could have on the future of free vs. proprietary software... I might be getting lured by the dark side here, but it's definitely an attractive thought. "No, Microsoft, you can't do that, only GPLed software can".

Is that a red right hand I see hidden in your coat?

Gratuitous license?, posted 12 May 2000 at 15:21 UTC by schoen » (Master)

Some people would probably argue that, because this license isn't part of a contract, it is revocable at any time. (These would probably be the same people who argue the same thing about gratuitous copyright licenses.)

We've seen that this could be an issue even with well-intentioned and trustworthy people, because, if they get sued for something, their copyrights and patents might be seen as assets, which could be taken away and given to someone else -- who might then try to revoke the license.

Can you make this license part of a contract somehow, so that it would be harder for somebody to revoke in the future?

Be careful, posted 13 May 2000 at 02:24 UTC by crj » (Journeyer)

A possible problem you may find with your grant is that it does not prevent your patents from being used for proprietary use. I can think of 2 scenarios:

1: My companys develops some internal-use app. which uses your patents. The app is "licensed" under the GPL, so there is no infrigement. Of course this application is never distributed outside the company - so it is defacto proprietary.

2: My company manufactures printers and uses your algorithms in the embedded software of the printer. "Oh, you want my source?. Sure, heres the code..."(which is a mix of c and assembly - but not the development environment, compiler, or documentation for some proprietary chip)

If your goal is to promote free software, I think that it does that.

But if you have another purpose in protecting your right to be paid for proprietary implementations, it may not work.

Issues -- LPF, revocability, internal use, embeddes systems, posted 13 May 2000 at 22:32 UTC by kmself » (Journeyer)

The grant looks reasonable to me. It's specific to the GNU GPL (though allows use under any version). One thought is that Raph might choose to restrict use to GNU GPL v. 2 or later. The FSF saw reason to cease using version 1 of the GPL, Raph might wish to discover why before allowing patent use under its terms. There's also the issue of use under the GNU LGPL, which can be linked with non-GPLd and non-LGPLd code. The way the grant reads, such use would be not be allowed. I'd simply like to clarify whether or not this is Raph's intent.

The grant itself isn't dissimilar to the patent grant under the IBM Public License.

Answering a number of people's concerns regarding whether or not patents could be utilized to the advantage of free software, this is the approach which has been suggested by a number of people, including RMS in creating the League for Programming Freedom, and its call for a mutual defense policy, including patent pooling (this is an idea which is reinvented frequently -- often several times a year). I'd say that Raph's actions here are in accord with generally accepted free software principles.

Addressing what I believe to be Decklin Foster's (decklin) question about the patent being available to all -- the license is freely granted to all persons and parties without prejudice, the only condition is that use be under the terms of the GPL.

Seth Schoen's (schoen) point about this being a revocable grant might be solved by the addition of a single word "irrevocable" to the grant.

Cliff Johnson (crj) is concerned about internal use and embedded products. The first may well have to be accepted as a reality of the GPL -- internal use of code doesn't trigger release obligations of GPLd software, it may not trigger a license-invalidating use either. I think this may be a good thing. Another potential use is as server software in which code is not actually distributed. Though current versions of the GPL do not address this, copyright law includes provisions for performance and public display, which may be covered in future versions of the GPL. In a world of web-based applications and application service providers (ASPs), this may be a required future change.

Use within embedded products is addressable in two ways as well. First, distributing hardware which includes GPLd code distributes the code as well, and is, IMO, a triggering event requiring availability of source. IBM addresses this issue in its PSL by withholding specific grant of hardware patents. This, IMO, would be well considered by Raph as a possible revision. While software should be freely distributable, non-triggering use of free software in embedded hardware systems might be considered a special case, and subject to patent royalties or other alternative licensing terms.

Good Move :), posted 15 May 2000 at 17:44 UTC by idcmp » (Journeyer)

Way to go Raph! While I can understand that the details might be messy, the idea is a great one and I hope others follow suit.

GPL is OK with this grant as far as I can tell, posted 15 May 2000 at 18:02 UTC by BrucePerens » (Master)

Decklin: RMS and I discussed this a long time ago. If a patent grant allows free use of the patented principle in GPL software, that is sufficient to satisfy the restriction that the patent must be licensed for everyone's free use. GPL code is free for everyone to use.



Four magic words, posted 15 May 2000 at 19:02 UTC by taral » (Journeyer)

I seem to remember that you want the four magic words:

nonexclusive, irrevocable, world-wide, royalty-free

These pretty much ensure that there are no restrictions on use. You can then tack on whatever you like (like "under the GPL").

compatibility with GPL == relicensability under GPL, posted 16 May 2000 at 00:51 UTC by clahey » (Master)

The whole idea of licenses being compatibility with the GPL is that you can relicense the given code as GPL and then link with them. It seems like this would be the case with this patent as well.

In other words, by granting the use of his patents with GPL software, Raph has granted use of his patents with all GPL compatible software. This still fulfills Raph's desire to keep his patent restricted from proprietary users. Even if the code was once GPL compatible, once they link it with proprietary software or redistribute revisions without source code, it is no longer GPL compatible.

It seems like this would fix everyone's complaint about patent use not being allowed in BSD software.

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!

Share this page