WebMink

The IP Elephant in the IT Kitchen

No idea who wrote this insightful analysis of the Linix Jihad against Sun but I find plenty to agree with. What I think is really happening as the Slashdoterati and Groklawists eviscerate Sun is that they are either terminally confused at the difference between copyright and patents or scared senseless by the problem of software patents. Finding no comfort, they attack Sun instead for daring to expose the problem.

First I'll try to summarise the difference between copyright and patents as I have recently spoken to people who don't understand. Forgive me if you know better, I'm sure you do - skip to the bold text. Copyright exists when you express an idea - the expression of the idea (the code, the text etc.) can't be copied without the agreement of its creator. Copyright can actually be bad for software development. It prevents code re-use if left unchecked. But OSI-approved licenses and Creative Commons licenses (for code and other stuff respectively) allow creators to surrender the right to copy their expressions to those they approve. Clever use of copyright laws by licenses like the GPL and others has allowed a community of software communities to gather in cyberspace and turn the world upside down.

None of this has anything to do with patents. Patents seek to prevent the use not of the expression of an idea but of the idea itself. In previous centuries nations promoted industrial development by allowing ideas to be controlled in this way as part of a social contract: "give us your ideas for posterity and we'll let you exploit them for now". This was mostly OK all the time ideas took effort to express physically, but most nations excluded "pure ideas" from the contract. Today we're facing campaigning to end that exclusion for software - hence 'software patents'. Through unfortunate case law (in the US) and the action of an unaccountable treaty organisation (in Europe), that campaign is nearing completion and is spreading worldwide through trade negotiations under the auspices of the WTO and WIPO.

The problem with software patents is that they apply whether or not you knew they existed. They allow patent terrorists to attack unexpectedly and they allow the cynical taxation of the innovation of others. While I know some industries can make partly reasonable arguments for them, they have no place on the internet or elsewhere in the massively connected society. They are a Bad Thing.

But right now they exist. I know, because when I was at IBM in the UK I was awarded several, US and European. IBM paid handsomely for them because it was very keen to build a huge portfolio under the guidance of its VP, Marshall Phelps (now at Microsoft). It was doing that because raw patents were and are one of IBM's largest sources of revenue. In fact, probably every US IT corporation has them - including even software-patent-opposing companies like MySQL and Red Hat (who have a great patent policy by the way, I'm not knocking here). You have to, because trading them is a crucial (if unsavoury) element of US corporate life. Sun is not a blind advocate of software patents and advocates the reform of patent law.

There's a problem out there, stated in the article as "[open source] in IT has a very nod nod wink wink character to it". No-one so far has ever bothered to deal with the issue of software patents and open source software, perhaps because they hadn't realised that copyrights and licenses were no answer to them. People are using software to which patents may apply. That's not a threat. It's just so. The truth is that IP ownership is the elephant in the kitchen for open source and the community can't fix it. Now Sun has come along and probably solved the problem for the CDDL community and all hell is breaking out in the open source movement between the two. Everyone is pointing towards the elephant and, instead of dealing with it, seeing the people on the other side and pointing at them instead.

Sun is making sure it's not an issue for any project Sun or anyone else starts under CDDL. Sun's patents aren't just for use with OpenSolaris - they are for anyone who will make the same commitment to address that elephant by using an elephant-aware license under an elephant-aware governance. It was clumsily stated, but Sun's not pretending to fix things for everyone. Fixing them involves a license that addresses the problem fairly for all as well as community governance that holds people to account for the code they contribute. Maybe GPL v3 will do it, but only if it creates explicit patent-protected zones the way Sun is doing for CDDL-licensed code.

Of course, saying this in the community draws fire from those who fear that tricksy Sun is after the Precious. So denounce me for "not getting the GPL". The elephant isn't leaving. You can try denying that CDDL is an open source license but doing so implicates the failure of OSI to deal with the elephant. You can try to claim IBM is the saviour but frankly they are just gaming you while they feed the elephant. Or you can perhaps blame Sun for demonstrating a partial solution to the elephant issue and thereby ending the delusional reverie that it's not there.

I've read postings on Groklaw and Slashdot saying all of these and more. But whatever you say, the elephant is here to stay and it needs dealing with before it crushes us all. Personally, I want to be part of the solution, not part of the problem and acknowledging the elephant is the first step.

posted at 2:51 AM (UK) | |

Comments:

Presumably Sun released its patents only to CDDL and not to OSI-approved licenses in general is because of the patent protections in CDDL. However there are other licenses (Apache 2.0, CPL, and EPL come to mind) that have similar patent and IP protections as CDDL. What are the chances that Sun would extend its patent offer to those as well?

# posted by Ed Burnette : 4:19 AM

Right now the view is that all those licenses allow multiple-licensing so any affected code could be licensed as 'CDDL and' but it's not impossible.

# posted by webmink : 4:30 AM

I can't believe the cynicism of IBM with their patent *giveaway*, they are one of the biggest lobbying forces in favour of patents in Europe.

Our Nick Blackledge has written a very nice article on the effects of patents in the software-based economy, check it out. I think Sun have made a few mistakes in marketing exactly how their patents interact with the open source community, in the end however they are doing the right thing. The CDDL is a great license which addresses the current climate in the software industry, I am still struggling to understand the small but vocal group that is ranting and raving against it.

# posted by ChÃ© Kristo : 10:30 AM

Nice. There's a handy list of the companies pushing software patents. As you say, IBM is there, as are HP and Microsoft. Not all EICTA members support software patents though, so their suggestion they do is false.

# posted by webmink : 10:39 AM

Can we leave terms like cynicism and hypocrisy out of it? The sad fact is that under current laws every company with lawyer worth his salt *has* to try to get as many patents as they can, if for no other reason than defense. Even my own company, which for years was strongly opposed to software patents and didn't have *any*, has been forced to get in this game because of a few infringement lawsuits.

# posted by Ed Burnette : 3:29 PM

This is a very insightful commentary.
The main ingredients of a more general solution would appear to be (1) the development of patent regimes that
avoid the threat of what Simon calls "patent terrorism" and encourage, rather than frustrate, technical creativity; (2)
meanwhile, the inclusion of protective measures in free and open source software licences; (3) the voluntary adoption by patent holders of patent policies that
support free and open-source software; (4) the accumulation of defence funds and legal expertise by open source communities; and (5) (regrettably) the acquisition, for defensive purposes only, of software
patents by organizations supporting the communit
Several of these elements are present in Red Hat's patent policy, as Simon points out; and the strong patent-related provisions of the CDDL are commendable. There is another objection developed in an article on groklaw.com related to third-party patents.
So far as I can discern from the article, the essence of the complaint is that Initial Developer, unlike Contributors, doesn't assert that it believes it has the necessary rights to make the grants under section 2.1 of the license, leaving it open for initial developers to distribute software under the license knowing them to be subject to third-party patents.
If this is in fact the concern, then I don't know enough about patent law in general, or the patent regimes operating in different jurisdictions to determine whether this constitutes a genuine weakness. The GPL incompatibility is unfortunate - another problem that might be solved by GPL 3 - the difference in principle between the licenses is that the GPL precludes licensed code from being integrated with non-free code, whereas the CDDL allows it. This fundamental difference of
opinion need not preclude an attempt, in GPL 3, at resolving the "compatibility" issues, however.

# posted by Jason White : 10:14 AM

Jason - The language relating to the original grant is exactly the same as the MPL uses, so if there's a problem it's not one the CDDL authors have introduced. More learned advisors than I tell me that the fact the license is offered at all by the original party means that the right to license is assumed and thus no explicit language is required.

If it's really an issue, it's possible that the Contributor Agreement could spell out the facts, but most people I have asked have said it's unnecessary and have spouted words like "estoppel" at me in their explanations.

# posted by webmink : 2:02 AM

Nice place to get more info about IBM patents.

http://www.freepatentsonline.com/6779049.html

# posted by alex123 : 9:45 PM