Mozilla Second Proposal for HTML5 Spec Licence

This message is Mozilla's second proposal for the licence for the HTML5 
specification.

Our understanding is that the goal of the various licensing options 
proposed is to prevent the modification and republishing of the entire 
specification ("forking") but otherwise to meet all of the use cases[0], 
which request permission for a wide variety of types of reuse.

As noted, we believe that the ability to modify and republish in this 
way is an important freedom. This message and proposal is not intended 
as a criticism of the work of the PSIG, because licences which permit it 
were expressly outside its remit.

Those who wish to prevent 'forking' need to be specific - are they 
against the forking of the specification, or the divergence of 
implementations?

If I were to take the HTML5 spec, make some changes to it, put it up on 
my website as GervTML5, and send a message to the browser vendors 
declaring myself as the new canonical source of web standards, I suggest 
that the effect this would have on the W3C and the progression of the 
HTML5 standard would be minor - confined, perhaps, to a raised eyebrow 
and a snicker at my arrogance and hubris. The forking of the 
specification, in itself, as long as the fork is clearly labelled as 
such to avoid confusion, is not something to concern anyone.

The 'problem' arises when a subset of vendors actually decide to follow 
an alternative spec instead of the W3C one - in other words, it is 
divergence of implementations which is the issue. But history tells us 
that attempts to enforce spec compliance legally, by companies who have 
recourse to law far more readily than the W3C, have been doomed to 
misery and failure - not just for them, but for their technology. As 
Tantek points out[1], the effect of the W3C taking such action is highly 
unlikely to be positive for the Web.

So Mozilla wishes to propose the use of the MIT License[2] for the HTML5
spec. This licence is a well-understood, familiar licence with a long
track record in the software industry. Its terms and intent are 
completely clear, and do not suffer from the problem of being 
interpreted different ways by different people. And, examining the 11 
use cases presented[0], we believe this licence meets all of them.

(If the W3C strongly wishes to have extra reassurance that their name 
will not be wrongly associated with derivative works, then perhaps an 
addendum could be considered to remind users of what trademark law forbids.)

The W3C's authority over HTML5 is due to its benevolent stewardship of 
the specification and to the commitment of the vendors concerned to work 
together under its aegis. We do not think that this authority requires 
or benefits from attempts to reinforce it with legal restrictions which, 
if actually enforced, would be a negative rather than positive thing for 
the web.

Gerv

[0] http://www.w3.org/2011/03/html-license-options.html#usecases
[1] http://lists.w3.org/Archives/Public/public-html/2011Apr/0000.html
[2] http://www.opensource.org/licenses/mit-license.php

Received on Wednesday, 27 April 2011 17:57:19 UTC