Re: Patent Policy issues...

On 3/12/2012 1:24 PM, Charles McCathieNevile wrote:
> After a bit of noodling with Art, he motivated the following thoughts.
>
> On Tue, 06 Mar 2012 13:16:35 +0100, Charles McCathieNevile 
> <chaals@opera.com> wrote:
>
>> On Tue, 06 Mar 2012 13:02:12 +0100, Arthur Barstow 
>> <art.barstow@nokia.com> wrote:
>> (in another thread - http://www.w3.org/mid/4F55FCC4.3090601@nokia.com 
>> ff)
>>
>>> So among the problems I see are:
>>
>>> PAGs suck (time, resources, joy from the WG, etc.);
>>> the PP takes too many resources to implement for me as an AC rep and 
>>> our IP department;
>>> the totality of the PP for WGs plus the CG's two patent policies are 
>>> at least one patent policy too many.
>>
>> Raised ISSUE-4 on this.
> ...
>>> Proposed solution #2 -> drop the PP for WGs and move WGs to use the 
>>> CG patent policies
>>
>> If we don't have agreement to get to "finished", that still causes 
>> problems... the CG process has some benefits, but also some drawbacks.
>
> Under the current WG policy, members participating are bound to make a 
> commitment to the final product. Under the CG policy, the commitment 
> only exends to contributions from the member, with there being a 
> seperate step allowing them to commit to the finished spec. A plus for 
> this is that it could allow a member who has useful contributions to 
> be in the group, without having to hand over IP they don't want to 
> talk about.
>
> My concern (because I think getting as much RF commitment as possible 
> is a really really important goal) is that it is fairly easy to make 
> sure the spec is not finished, and easier to watch quietly while 
> someone else includes material that a troll knows they cover with IP, 
> then walk away leaving the spec effectively unimplementable for anyone 
> who doesn't license the IP - and without anyone even knowing, since 
> there isn't the requirement to identify or exclude anything.

One could look at the strengths and weaknesses of the CG process in 
isolation.  I agree that this is a weakness.  But CGs are known to have 
weaknesses: they are designed to allow to start quickly without a lot of 
overhead.

One could also look at strengths and weaknesses of CGs together with 
WGs.  For specs that we want to make into normative specs, the goal 
should be to transition them into WGs at an appropriate time.  In that 
case, the more robust WG processes take over.  But by starting in a CG, 
we got started much faster.

Do we stop trolls entirely?  No, that probably is impossible.  But that 
deficiency is true today as well, for a sufficiently subtle troll who 
inserted their IP into something that feeds W3C specs.

>
> There are nice features about the CG appraoch, like a conditional 
> license to implement when the spec is in development (which is 
> important is you want to get tests developed and passed). I think this 
> should be explored further...
>
> But it depends on what the W3C membership at large want to do, and 
> without a lot of motivation they might just let sleeping dogs lie...
>
> cheers
>

Received on Monday, 12 March 2012 20:23:46 UTC