In response to the objection I made on behalf of Mozilla to the previous proposed charter of the HTML Working Group, the W3C has proposed a new charter that allows the group to experiment with an open document license for extension specifications to HTML. If you work for a member company of W3C, I'd encourage you to encourage your AC representative to support the open licensing experiment. Here's why:

W3C publishes specifications, generally written by members of the W3C's working groups, under W3C's copyright ownership [1] and licensed under the W3C Document License, which does not allow other parties to modify the documents. This means that if, for some reason, W3C is no longer the right place to standardize something, those who want to work on that standard elsewhere have to start over rather than starting from the W3C document.

I'm one of several people who would prefer that the W3C license its specifications in a way that allows others to modify and republish the documents (a license that allows forking). I want this to be possible so that if the W3C stops being the right venue for standardizing something, the work can continue elsewhere. I don't want this to be a regular occurrence, nor do I want those who do it to do it without appropriate credit to the specification's original authors. But I want it to be possible to do when it's necessary, so I don't want legal restrictions that prevent forking, only social ones.

I also want it to be possible so that the W3C can collaborate effectively with other venues (such as the WHATWG) that do publish documents under a forking license. Currently this sort of collaboration isn't possible because the W3C can and does take documents from elsewhere that allow modification (often with little credit), but the movement can't happen in the other direction.

For example, when Ian Hickson set out to write what has become HTML5, he had to start over from scratch because previous HTML specifications were not licensed in a way that they can be modified. Ian wrote a new HTML specification at the WHATWG under a license that does allow modifications, and the W3C is using this specification as the bulk of their HTML5 specification (often with little credit to its original author).

Having licenses that allow modification and forking is also a useful governance mechanism. Without such licenses, once W3C has become the place for a particular specification, it is under relatively little pressure to continue to be a good forum. Allowing modification puts a little more competitive pressure on W3C to remain a good forum for standardization. This will reduce the chance of the history of HTML standardization from around 1999-2006 repeating itself.

Some might counter that this lessens the point of standards. I disagree. Standards are useful where there is benefit from a common agreement on how things work. If that agreement still has more value than changing it, then the party modifying the specification won't change it. When the value of sharing a common standard isn't large enough to motivate people to make and keep an agreement, that probably means it isn't worthwhile to do so.

For these reasons, I (as Mozilla's representative to the W3C's Advisory Committee) objected to the previous draft charter of the HTML Working Group. While we'd like to see all the documents of the group published under a liberal license, our objection didn't ask for this. We asked only that the charter allow the group to experiment with liberal licensing on extension specifications. We hope this experiment will give other members of the W3C more experience in working in an environment with liberal licensing and make them more comfortable with extending the experiment to other documents in the future.

The new proposed charter is under review by the W3C's Advisory Committee (the committee consisting of one representative per W3C member company) through May 29. If you are a member of the Advisory Committee or know your company's representative on the Advisory Committee, I'd encourage you to support liberal licensing.