Thanks for putting your time and thought into this! Ian Hickson wrote:
> On Tue, 2 Oct 2001, Robert O'Callahan wrote: >>I hear that you made some comments on IRC about the W3C patent policy. >>Unfortunately I wasn't around ... I and I'm sure a lot of others would >>like to hear what you have to say. Could you make a newsgroup post or >>something? Thanks (hopefully)! > > There isn't much to say. Most of it is pretty clearly stated in > > http://www.w3.org/2001/10/patent-response > > To summarise: > > 1. The current situation is *not* that of royalty free licenses > automatically applying to standards. The W3C currently has *no* > patent policy, and this has bitten the community in the past, > e.g. with P3P. But there seems to have been an *expectation* of royalty free licensing even if the W3C had no specific policy. The P3P example itself seems to demonstrate that when that expectation is violated, people jump up and down and the whole process grinds to a halt until the patent issues are resolved (in this case, in favour of non-encumberance). The new proposal appears to demolish this expectation. If P3P had been developed with the new proposal in place, would it still be freely implementable? Or would it just have been converted to RAND rules? > 2. The proposal would guarentee that all standards are available to > everyone in the form of Reasonable and Non-Discriminatory licenses, > i.e. free software groups would not be excluded (assuming they can > pool together resources to afford the license). There is currently > NO guarentee of this kind. I cannot see how this could work for free software projects if someone demands a per-unit royalty (as permitted under the new proposal). > 3. Without these rules, there is no guarentee that some standards will > not get sabottaged, e.g. P3P, XPointer. With or without these rules, standards can still be sabotaged by third parties. > 4. If the rules demand Royalty Free licensing throughout, some > standards will simply not get developed at the W3C (and thus > in an environment reasonably open to the public) -- e.g. XPointer > has a patent covering a part of it (IIRC), and there is no sign that > the company owning this patent will in any way accept to license it > for free -- they would rather that part of the standard was > dropped, as far as I can tell. (There is work on creating a subset > of XPointer that isn't covered by this patent.) > > This is an important point to emphasise: Get real, welcome to the > world, companies are not going to give away their "Intellectual > Property" when they don't have to. Just like the W3C will become > irrelevant if they have no freely implementable standards, they will > also be irrelevant if every standard is developed somewhere else > simply because companies don't want to give away their "Intellectual > Property". Some standards would be developed elsewhere. OTOH more standards overall might be freely implementable because of this policy. Even in the real world patents can often be worked around, freely licensed under pressure, or shown to be irrevelant or invalid. > 5. Other standard organisations (IETF, ANSI, etc) do not have any rules > requiring Royalty Free licenses for patents covering their standards. That's a problem. > 6. Many companies (Microsoft is a particular famous example of this) > get patents left right and centre for one reason and one reason > alone: defense in case someone else sues them (the idea is that if > company A sues company B for breaking patent W, company B can turn > around and counter-sue company A for breaking patents X, Y and Z. It > is then cheaper for company A to drop the subject than to run both > battles through the courts). They could offer a royalty-free license which automatically ends if the licensee sues the patent holder. This kind of license has been used extensively before. [Offtopic: defensive patent portfolios are rapidly becoming obsolete in the face of "pure parasite" companies like Rambus and Techsearch which only license their "intellectual property" to others and do not actually produce anything themselves, and are therefore immune to patent countersuits.] > 7. Given 6, I would strongly recommend that free software contributors > start getting patents of their own. (This is similar in spirit to > the idea given in [1]). Indeed, I would strongly recommend that free > software proponents patent every idea under the sun, and then make > them publically available to all free software projects (and maybe > _only_ these -- thus turning the tables on the proprietary world). > An example of someone who did something like this is at [2]. Good idea, if we had the money to obtain patents, and deep enough pockets to plausibly litigate patent infringement. > 8. My personal idea of patents is somewhat similar to RMS'. [3] > > 9. If you want patent-free protocols, then join [4]. > > 10. If you worry about patented web standards, remember that Microsoft > own a patent on CSS [5] and have never given anyone an explicit > license for its use. Do you trust Microsoft not to sue Mozilla > through the floor because of this? (I do; see 6.) If they can be trusted this way, then they should have no problem issuing a royalty-free license with a countersuit exception. Thanks! Rob
