Re: DRAFT Agenda for the Process Call Wednesday 12th February 7am PDT

On 2/11/20 3:23 PM, Michael Champion wrote:
 >
 > -  It’s being rushed.  “Move fast and break things”  has been a disaster for
 > the web, and it’s a particularly bad approach to foundational documents
 > such as the patent policy.

I agree with "move fast and break things" being the wrong approach, which is 
why the updates here are incremental and build on the existing Process and 
Patent Policy instead of creating a new one from scratch or tossing out 
significant aspects of our current Process.

 > - It’s too complex.  One can argue, as I still do in Issue
 > https://github.com/w3c/w3process/issues/346 that the proposed process itself
 > is too complex.  But an even bigger red flag is that the dependencies
 > between the proposed patent policy  and process are so complex that they
 > must move in lockstep.   Complex dependencies moving targets are a portent
 > of project doom in my experience.

This isn't actually true, Process 2020 can be deployed without updates to the 
Patent Policy, and is currently drafted on the basis of the current Patent 
Policy. We just don't get some of the benefits that groups want  (particularly 
the ability to have *patent-protected* "living standards" as perpetual CRs) 
without an update to the Patent Policy.

 > - Little solid stakeholder buy-in.  W3M and the AB are taking the lack of
 > strong opposition at TPAC  and afterwards as evidence of support.  I’m not
 > sure that’s a valid inference even about the W3C community.  The Service
 > Workers WG -- the main target of the “living standards” features -- has
 > given a lukewarm at best endorsement in
 >   https://github.com/w3c/w3process/issues/346#issuecomment-554701435 .
 > As for the broader community, I can easily see some interpreting this as yet
 > another skirmish in a turf war (“no more bleeding of W3C work to WHATWG!!!”)
 > rather than a respectful move to learn from WHATWG’s success and allow
 > groups that are outside WHATWG’s scope to develop W3C-flavored Living
 > Standards.

There were multiple people from different groups who came up to me and were 
enthusiastic about the proposed updates to the Patent Policy. They're just not 
interested in getting involved with the Process updating discussions 
themselves, so aren't commenting on the issues.

 > What’s the alternative?  I’d suggest:
 >
 > 1. Restart the PSIG conversation on the basis of the existing Process,
 >    aiming to get locked-in patent commitments after the CR exclusion period
 >    expires.  Don't add complicated interdependences between a draft PP and a
 >    draft Process.
 >    Give PSIG time to fix some of the worst bugs, archaic references, and
 >    ambiguities while they are at it. Such a patent policy would  allow WGs
 >    to call CR’s “living standards” because they have solid patent
 >    commitments, and CRs can get updated fairly quickly if wide review finds
 >    problems.

Let's let PSIG determine the scope and schedule of its changes. They're having 
an F2F meeting on Thursday to figure it out, so I think before making claims 
about it, we should let them discuss. I will point out a couple things:
   1. We're revising an existing policy, not creating a new one from scratch.
      It is not going to be as difficult or complicated as drafting the
      original policy.
   2. PSIG's members are very careful, and I do expect they will push back if
      they're not happy with the draft they have.

As for interdependencies between the PP and the Process, because we were 
careful to draft the Process this way, there's only one point where they lock 
together: the definition of “Patent Review Draft” in the Process draft. The 
only change that needs to be made when we update the Patent Policy is this one:
       https://github.com/w3c/w3process/pull/363/files

The rest of the draft has been adjusted to either to Patent Review Draft or, 
in the cases where it references other concepts, to reference Patent Policy 
section title rather than the section number, so that the Patent Policy can be 
swapped out without any further changes to the Process.

This decoupling was very intentional as we were drafting the EverTeal text; 
even if no one else was paying attention to this detail, Florian and I were.

 > 2.Move ahead with Process 2020 by removing any features that depend on an
 > updated patent policy.  The original Everteal proposal addressed real problems
 > the CSS WG has with the current Director review overhead.  The fixes were
 > proposed by the CSS editors, so I am confident they fix the problem. If it
 > takes longer than a couple months to back out the new PP dependencies and
 > various flavors of CRs, take the time to do it cleanly.

As stated above, there are no features in Process 2020 that depend on an 
update of the Patent Policy. There are benefits that won't be realized until 
we have both, but that's not the same as a dependency.

~fantasai

Received on Wednesday, 12 February 2020 01:32:23 UTC