Regarding the Patent Policy

There have been many discussions of the W3C Patent Policy and how it  
might apply to this group. I think a lot of people are  
misunderstanding the policy and what it requires. I am not a lawyer,  
so please don't take the following as legal advice, but here is my  
layman's understanding.

First, a link to the full policy: http://www.w3.org/Consortium/Patent- 
Policy-20040205/

Here's a brief summary of what the policy requires:

- The W3C has a goal of making specifications implementable on a  
Royalty-Free basis. This means either no patents, or any applicable  
patents are licensed royalty free. This means the W3C can't approve a  
Recommendation if it knows of patents that are not RF licensed.
- Working group participants must agree to license any patent claims  
they have that may apply to the working group's specs on a royalty  
free basis.
- Working group participants make no promises about whether any other  
organization may hold a patent on anything in the spec.
- There are certain conditions where a participant may exclude  
particular patent claims, but they are required to notify the W3C  
within a limited amount of time of Working Draft publication.

Also, a brief reminder of how patents work:

- Patents cover concepts and ideas, not particular expressions. So if  
a spec covers an idea, who wrote the language or what the details are  
is irrelevant from a patent perspective.
- Writing something down does not necessarily give you a patent claim  
if it is not original, not trivial or already patented.
- Publishing an idea excludes it from future patentability by the  
person who published it, at least in the US.
- Publishing an idea creates prior art against anyone else trying to  
patent it in the future.


So given all this, what is the actual effect of the patent policy? It  
means that anyone who joined the working group (individually if they  
joined as an invited expert, or on behalf of their company if they  
represent a W3C Member) agrees to grant a royalty-free license to any  
patents that may cover parts of the spec. It doesn't matter whether  
they wrote the relevant parts of the spec, or if they originally  
contributed ideas when the spec was not under W3C auspices. It also  
gives *no* protection against patents that are held by non-members of  
the WG, whether or not those people contributed the relevant spec  
language.

So I think the following (suggested by some) are *not* problems:

- It is not a problem to adopt the WHATWG specs - using them gives no  
more or less patent protection than if the HTML WG specced similar  
features from scratch. In both cases, WG Members agree to royalty- 
free licensing, and there is no protection from anyone else. The only  
way to get more protection is to get more companies to join the  
working group.

- It is not a problem to specify features that are covered by a  
granted or pending patent held by a WG member, unless that member  
calls for exclusion within the stated time limit.

- It is not a problem to have a spec based on comments from people  
who have not agreed to the patent policy. Minor ideas are relatively  
unlikely to contain existing patent problems. And W3C Working Groups  
have a longstanding history of accepting comments on drafts from the  
general public without requiring patent agreements.

- It is not a problem to have people subscribed to or posting to the  
list, or participating in the conversation, without having signed the  
policy. This does not reduce or increase the patent-safety of the spec.

Feel free to correct me if I have misunderstood any aspect of the  
patent policy or patent law.

Regards,
Maciej

Received on Thursday, 15 March 2007 06:56:13 UTC