See also: IRC log
Arnaud, Eduardo, Larry, Harry, Ian.
Regrets: Dom, tlr, Chris Messina, karl
<trackbot> Date: 12 July 2010
<Arnaud> Ian, I can't join yet, I'll be a bit late
Survey results -> http://www.w3.org/2002/09/wbs/1/newstd2/results
IJ observations on survey results -> http://lists.w3.org/Archives/Public/public-vision-newstd/2010Jul/0016.html
IJ: Anybody look at results?
<hhalpin> I have looked at them.
[Just IJ for now]
<hhalpin> Interesting, a mix of W3C sort of folks and web developers it seemed.
activities needing a host:
http://www.w3.org/2002/09/wbs/1/newstd2/results#xq6
EG: There may be an OASIS TC working on BML (not sure)
IJ: Anyone know http://openrecommender.org?
<hhalpin> Strong +1 on not having "two-tiered" process, but on seeing how these things work.
<hhalpin> together
IJ: I liked Kimberly's comment.
EG: Manu's comment about less team involvement seems consistent as well
hhalpin: One thing I'd like to discuss....strong emphasis about taking grassroots efforts with energy and helping them with pat pol and infrastructure.
<hhalpin> the idea is to clarify how to move from a lightweight process to a more formal track.
<hhalpin> almost no-one seems to disagree there.
<hhalpin> it WOULD be useful.
<hhalpin> of course, moving to a more formal process needs to an option.
<hhalpin> that some groups may not need to take.
<hhalpin> or want to even.
IJ: I am hearing some people say that if we get the lightweight thing right, we may not need the heavier weight thing. What do you see as the value of the classical track?
EG: It's important to remember that there are multiple communities, who want different things (and different degrees of formality)
<hhalpin> +1
EG: there wil be people want nothing to do with W3C, but there are other communities who will be happy if we have something to offer.
+1
<hhalpin> It seems to me the main reason for the heavy-weight track is the legal agreements and the fact that it can work in very politically fraught technological spaces.
<hhalpin> but I'm not a lawyer, would like to hear what larry has to say...
IJ: Who values what aspects of the classical track?
lrosen: I can't imagine Linux
being developed by a lightweight mechanism...it's too big to be
done in someone's garage.
... what I understand is that there are some things that may be
just between a few companies, or just a few developers, or it's
new and not yet provent, or maybe a couple of anarchists who
want to do something out of the ordinary...that's what the
lightweight process is for.
... but for _more meaningful agreement_ you move to a more
formal process....a different way of doing business.
Eduardo: it would be a good
exercise to separate answers/comments to know where they came
from.
... for example, if someone says "zero-fee is important, but I
would never participate in W3C' the other answers may not be
that relevant.
hhalpin: we need to figure out
why other groups using the XG process.
... or how to transform it into a new lightweight process
(requiring an in-depth evaluation, comparision to other similar
processes)
... what W3C needs to communicate is how the process
evolved...
... one characterization I've heard about w3c process is that
it's good when things go wrong.
... even if not crafted today for small communities that get
along and are working closely
lrosen: I've seen mature products
come to apache and be required to go through the incubator
process.
... it's not that the mechanism of the project is wrong, it's
that they need to learn how apache works.
... those projects are likely to get promoted to top-level
projects since there are mature communities that go with them,
but there's a learning curve.
IJ: Reasking Arnaud - what is the value and to whom of the second track?
Arnaud: On larry's comment -we are trying to make it easier for people to come to w3c..larry's comment suggests a bit the opposite --- forcing people to go through incubator group
<hhalpin> agree with Arnaud, but Incubator should be an option.
IJ: Another way to view live is "everything starts as a community process" that is, as an XG.
<hhalpin> some things are already mature, thus the "Public Submission" idea
Arnaud: To Ian's question "what's the value add of the classic process," I think the answer involves (1) IPR (2) Rigor
<hhalpin> i.e. the community may already exist and have draft specs etc., and so could want to go straight to a heavier-weight process due to its value add, IP, maturity, etc.
Arnaud: The rigor can be painful
at times, but that's the value.
... it is an open question whether people will move from casual
track to more formal one...hard to know.
larry: I think Arnaud has it
exactly right - I didn't mean to imply that graduating from
incubator to top level project is a function of how they learn
about us. There is a checklist (e.g., related to licenses,
etc.).
... there's a maturity level that needs to be met in order to
fit in.
... they need to certify to the rest of the community that
they've met certain maturity standards to qualify them as
fitting in and to satisfying the customers.
<lrosen> i've never heard anyone in Apache say "I'd rather stay in the Incubator"!
<hhalpin> I think the value add is mostly in RF status....
<lrosen> I've heard lots of people say, "This project ought not to graduate to a top level project"!
IJ: I am hearing people say "we are implementing anyway; don't need your formal track for more rigor"
EG: In OASIS, one pattern was to
get to "committee specification" then get implementation
experience, then some time later, bring it for public review,
redo it, and then go for OASIS standard.
... that almost never happened
... In many cases, people went straight for the heavyweight
process.
lrosen: I'm not sure why you are
so shy, IJ about the value proposition of the classic track. I
expect you'll have more in incubator status and a smaller
number in classic track.
... I could see graduating to the formal track additional
efforts like: monthly reports, more promotion by W3C, more
serious outreach to the industry
... I don't think the problem is "no value perceived in the
classic track" it's that there may not be a need for major
value in an incubator project until it earns it.
<lrosen> By the way, Apache also has an "Attic".
Arnaud: Are we missing the main
point here? We are worried that we are making it easier to
start in incubator land and stay there, but that may not be the
right question to look at. Remember, the premise was that we
want to increase the chance that people bring more work to W3C.
I don't think that opening up an incubator activity still
increases the chance that more will come to w3c and then move
to the rec track.
... so I don't think we will be making things worse. We do need
to be careful about not diluting the value associated with
labels for specs.
<dom> [but who monitors the monitor? :) ]
Arnaud: the other thing is that
ad-hoc groups all claim that they are not doing
standards....they get together and start working on specs
(e.g,. under OWF agreement) and they don't claim that they are
doing standards, and they say "eventually we will submit it to
ietf or w3c" but we haven't seen it happen yet.
... making it easier for them to do the ad-hoc work in w3c
increases the chances.
hhalpin: I think that the RF patent policy is a big value of the classic track.
IJ: Larry, would you be able to sketch a series of steps among 2 tiers.
EG: I've discussed with Ian idea of a time-limited patent commitment.
Larry: Need to be able to
distinguish "contribution" from "participation"
... right...might be a distinction useful between incubator and
classic status
lrosen: I want discussion
(participation, contribution) to take place on the OWF list.
:)
... if there are reasons why companies may not to agree to
OWFa
... if there are questions about application of OWFa to W3C,
that discussion should happen here (or in PSIG)
IJ: Another bit to consider - starting with non-assert first, then license.
lrosen: We are doing that in OWF...non assert for field of use...to implement those parts of the spec whose function is defined in detail (not merely referenced)
EG: I don't really agree with the hierarchical approach you seem to be suggesting ... between non-assert and license. It's not clear that it's easier to get a non-assert than a license.
lrosen: OWF lets you issue a RF license if you prefer.
<Arnaud> Ian, I have to admit to be surprised we are spending so much time on this
<Arnaud> was this on the agenda?
part of agenda 2
lrosen: What I think ought to be
done, is that W3C ought to rethink its patent policy.
... if we come up with an incubator policy that looks something
like OWFa...then graduating to the w3c patent policy would be a
slight step backwards.
... it almost causes the PSIG and W3C to say to itself, is it
time to broaden our policy.
<scribe> ACTION: Larry will write down thoughts on what lightweight commitments for incubator might look like and then what it would mean to graduate to w3c rec track (and relation to those commitments) [recorded in http://www.w3.org/2010/07/12-newstd-minutes.html#action01]
<trackbot> Sorry, couldn't find user - Larry
Arnaud: My first impression based
on my own view of the process and people's responses to my
overview is that "there doesn't seem to be much to gain to
streamline the current process."
... I would venture that it's probably best to leave it
alone...but would be happy to hear from the group on that.
IJ: +1 to priority on incubator
design
... I think worth looking at both "operational improvements"
and "some process tweaks"
Arnaud: +1 to looking at those somewhat, but those aren't the main goal.
IJ: I will take to the list.
IJ: Feedback very welcome.
EG: I'll take another look.