Making W3C the place for new standards task force

28 Jun 2010


See also: IRC log


IanJ (Chair), Arnaud, Eran, Larry, DanA, Michael


Legal agreements [pre-agenda discussion]

lrosen: we are working on this at OWA...looking for lightweight commitments, both for patent and copyright
... there are some difficult agreements wrt patent grants

[discussion of owf agreement development process]

use cases


* 1 [Core] Develop a new Web standard

* 2 [Sunset] Revise a W3C Recommendation

* 3 [Vocabulary] Develop an industry-specific (vertical) vocabulary

* 4 [Profile] Create a profile of one or more specifications

* 5 [Competition] Develop a competing specification

* 6 [Experiment] Experiment on a specification, test suite, etc. with low start-up costs and ability to work indefinitely

* 7 [Reinvent] Another body is developing a specification that could reuse a W3C Recommendation

* 8 [Promote] Promote an informal standard through some formalization process

IJ: Is this complete? Who wants to take a pass at editorial revision?

dka: One thing that may be missing - coordination with other groups.

see: http://www.w3.org/2010/04/w3c-vision-public/wiki/Newstd#Ideas_and_proposals

dka: I am interested in the liaison part.



<dka> gah

IJ: Yes, liaisons are important but not listed in use cases...questions of discovery, scalability, role of staff

<dka> my phone crashed...

dka: I would have concerns about a model where "w3c does something an periodically checks in with another group."

IJ: What would you like to see?

dka: I don't have a concrete suggestion. But when I've seen liaisons operate, I haven't seen them be terribly effective.

lrosen: One thing that Apache does --- they define this as a mentoring relationship.
... they assume that people bringing work to apache may not yet understand how to find answers, why things are done in a particular way, etc.
... mentoring role may be more helpful than liaising

EHL: Collaboration is valuable, but at least at this point in time, the examples aren't that interesting.
... if you look at some of the venues that were mentioned...OWF doesn't have standards activities...it's focused on process of creating work
... but doesn't have any work going on
... openid foundation is struggling right now...they need to figure things out before they talk to others.
... the open efforts done outside W3C, there aren't many left.
... for the most part, the trend from 2 years ago has died off.
... mostly right now we are seeing proposals from a small number of companies to specify some sort of interop
... in those cases, they don't like to have (too open) a process since they have products to ship.
... we've seen tension between those needs and community needs to keep an open process.
... our focus should be long-term...the _next_ person who wants to start a community-based effort should come to w3c and see if it's a suitable place for them.

Elements of a Successful Standard Community


IJ: So what are big considerations?

EHL: Participation model was the number 1 concern in my case (e.g., oauth)
... I looked at IETF, OASIS, W3C, and keeping it open outside a standards body and my conclusion wrt oauth was that people who were part of original work could continue to be part of the work.
... I also acknowledge that the IETF IPR process was lacking.
... I've learned to live with that.
... I also had issues with the complexity of the IETF process, particularly to newcomers.
... but we've lowered some of those barriers.
... W3C and OASIS ruled themselves out immediately.

IJ: Do you have a gut sense that a two-part model (incubator -> std) model with heavier-weight commitments at transition?

EHL: My impression is that the revenue model has to be decoupled from the product

<dka> +1 +1 +1 +1 +1 +1 +1

<lrosen> +1 +1 +1 +1 +1

scribe: you've failed if revenue model controls the process in any way

EHL: If the lightweight thing succeeds, nobody will go to the heavy thing, and if it fails, it doesn't matter.

dka: I was going to say the same thing as Eran....I feel like we shouldn't talk about revenue models in this task force
... We need to decouple the process from the money question.

lrosen: I wanted to compliment Eran on what he said --- put his finger on a key piece of this.
... membership model of w3c puts them out of the running for some work.
... I think the "pay to play" model puts a burden on the development of certain work.
... maybe we need to move the IP commitments to contributors to standards....you may be able to get enough money from sponsor members who are willing to pay for the free work being done by the standard-setting people.

Mike: I do think that we need to focus on creating a model where W3C, broadly defined, is a more attractive venue. And if successful, worry about monetizing it.
... on the priorities of the use cases: a lightweight process that can attract a larger community to pre-standards brainstorming is the highest priority.

[Barriers: http://www.w3.org/2010/04/w3c-vision-public/wiki/Perceived_Barriers]

IJ: Do you agree that the lightweight thing can moot the heavier weight thing?

MS: No; companies such as mine care more about patent policies and participation models.
... process and IPR policy starts to matter to the larger businesses.
... likewise, when you talk about governments.
... there are govt procurement policies that care very much about whether something is "really a standard" or some informal agreement.

<EHL> +q

MS: successful lightweight specs will create a demand for heavier-weight standardization.

EHL: I want to clarify: the stds process is valuable...but the question is whether it will be attractive for some kinds of participants will want to continue there.
... some will want to move on to other (cooler, newer) work.
... in some cases, even big companies may be reluctant to carry something into a standards process.
... so I completely agree about the importance of standardization, but it's a huge struggle to get from lighter to heavier.

IJ: Given social realities - innovation world (cool new light ongoing) -> effort to move from standards....how do you cater to both?

Arnaud: IBM participates in open source projects...we pay engineers to work in projects where others participate at no cost...we see the bigger picture and see something to gain from the investment.
... and I think that standards work is similar...we would probably keep supporting the work (financially) if we think that there's a benefit to us...even if others participating at no cost.
... and you could have granularity at different levels, e.g., participation in a WG could be free, but there might be other benefits at other levels (e.g., Member approval to start work)
... so I don't have a problem with opening up participation in WGs.
... I am a bit concerned with Eran's comment that "W3C and OASIS ruled themselves out"
... W3C has been generous in granting IE status.
... especially people with a history of people who've been involved with work.
... is this a perception issue?

EHL: I think it's a practical matter.
... when I did the review, it looked as though there were 30-40 people in the oauth community...of them, about 8-9 were working for a W3C Member.
... so that would have meant a long list of Inv Experts....and one that we would have wanted to grow indefinitely.

IJ: Do you think individual participant category _has_ to be free?

EHL: likely has to be zero cost
... it's a principle more than the small amount of money.
... I am contributing, why should I have to pay?

<dka> I agree - individual membership needs to be free. I also agree the invited expert mechanism is opaque.

EHL: I want to decouple the value of an organization from participation model.

lrosen: My experience recently has been with apache...we've decoupled membership from contributor base.
... anyone can be a contributor (though people get status within an organization and their rights go up with their status)
... that's distinct from the membership, which is responsible for ensuring that there's a board of directors that does their thing, and ensuring IPR policy exists, and there's PR, and so on.
... +1 to decoupling the financial support from the membership model.

<lrosen> Use case most important: Wide participation by non-members in a WG

Proposed: That lightweight process seems top priority.

[dka has to go]

Arnaud: On lightweight process..I think that there are some things in the current process that we should revisit.
... we added some bits of process on purpose at one point or another
... typically related to quality
... there may be a tradeoff, maybe we don't want to enforce some of them in some cases
... but there may also be ways in which people execute the process.
... everyone wants things to move quickly, but consensus-building adds weight
... different WGs have tried to approach this in different ways

<EHL> +q

Arnaud: so there's flexibility in how you carry out the process.

lrosen: what was heavy? other than the heavy lifting that comes from doing technology, what was heavy?

Arnaud: last call, test suites, creating "disposition of comments"
... you have to answer comments
... you explain the disposition of comments to the Director
... it's heavy but it's there for a reason.

EHL: process can be simpler but standards can rarely be faster
... people have to realize that ... it's hard for newcomers to appreciate.
... standards take time (1-3 years)...you don't know if you'll even need it in 2 years

lrosen: the adoption of the standard has probably started to take place long before the formal process has completed.
... the one that I've been focused on is: at some point companies have to make an IP commitment

<EHL> most standards *require* implementation as part of the process

lrosen: so it becomes important to be able to point to the stable thing
... let the std live...I'd rather have heavy-duty on software QA end

dka: one thing I was going to talk about is IPR

-> http://www.w3.org/2010/03/outposts-proposal-snapshot.html Outposts proposal

dka: need IPR policy for incubator-type work
... in may cases, these experimental standards are being developed in concert with open source implementations
... there ought to be a way to link those within an experimental process
... we should link the OS development and the standards

<Zakim> Ian, you wanted to talk about conversation

IJ: Comment on larry's idea of lighter weight standards process and shifting burden to software?

dka: I agree that standards-making is hard. I don't think we need a lighterweight process for the standards part a W3C....it's one of the values of W3C
... i would simply augment it with connections to OS development

IJ will add "dev outreach to the wiki"

scribe: backporting implementation experience

<EHL> +q

IJ: One issue that has been raised related to implementation - RF commitments not backed until REC.
... Some bits of process aren't going through the process completely (IETF or W3C)

EHL: Indeed, in the IETF, some specs are published, but then not carried through the process....
... in some cases, people just lose interest...don't care what some other body (e.g., IESG or TAG)..."we're done"
... because IETF and W3C have these sanity checks, you are adding a layer of complexity that most people don't want to deal with...I don't want to get rid of it, but the question is how do you get people to go through it (e.g., simplify? pay somebody to carry it through?)

lrosen: I think that IETF has it mostly right, but for the IPR policy
... there are peer review mechanisms that are alright
... IPR uncertainty at the end of the process
... W3C can add some regularity to this...figure out how to make an IPR policy (not necessarily changing the patent policy)
... licensing policies and commitment levels, when contributors give what....

Arnaud: what about reconciling the two models? The reason the IPR policy today is tied to Rec is that companies are willing to commit IPR to things people agree to.
... I'm not willing to commit to any and everything.


<Arnaud> was the call scheduled for 1h 1/2?

IJ: Anybody want to call any potential customers to learn more about their needs?


<dka> I think I am doing this already, Ian...

larry: I can forward survey to the apache list


I'm going to send survey to XG chairs tomorrow

Feel free to send comments before then

IJ will also be drafting and email that larry and others can use to talk to other orgs

strawman proposals

<scribe> ACTION: DanA to write up some notes on what he thinks are important for liaisons [recorded in http://www.w3.org/2010/06/28-newstd-minutes.html#action01]

IJ: areas of study:

- IPR staircase

- quality / accountability/ burden balance

- document license if you have incubator then rec track



lrosen: I'm not volunteering to do an action, but someone from W3C should join the OWF discussion list


<scribe> ACTION: Arnaud will write down a few ideas on simplification of process and operations (to start by fleshing out comments on this call) [recorded in http://www.w3.org/2010/06/28-newstd-minutes.html#action02]

lrosen: can Eran say more about what works for him at the IETF?

EHL: I no longer share Larry's concern for the IPR issues.
... I'm not aware of significant issues due to the lack of tighter IPR policies at the IETF.
... in one case, community decided it was better to pay royalties than to abandon the technology.
... so I've not seen the problem materialize in the IETF.
... they other part of the IETF...I took the time to work that system and I'm productive in it.
... several people held my hand through the process.
... I am now comfortable, and like to help other people figure it out now.

next meeting

<dka> time good

Likely 12 July, same starting time

Summary of Action Items

[NEW] ACTION: Arnaud will write down a few ideas on simplification of process and operations (to start by fleshing out comments on this call) [recorded in http://www.w3.org/2010/06/28-newstd-minutes.html#action02]
[NEW] ACTION: DanA to write up some notes on what he thinks are important for liaisons [recorded in http://www.w3.org/2010/06/28-newstd-minutes.html#action01]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2010/06/28 17:40:50 $