See also: IRC log
ArnaudLH, LarryR, CarlC, EduardoG, AndrewU, IanJ, MichaelC, ThomasR.
<trackbot> Date: 09 August 2010
IJ summaries early proposal, to be discussed next week with other task force chairs.
EG: Please mention time-out idea
mchampion: Structure seems ok
Ian: My goal is to work all week on this, send something for comment later this week.
Carl: Point out the benefit here
is "play ground....some structured play"
... some structure:
- way to introduce an idea
IJ: What about peer-moderator?
Carl: Want to ensure that
conversations don't degenerate quickly.
... there needs to be some supervision
Mike: The problem that we (MS)
have is similar...we need to talk to a community of
... take WebSRT for example.
... some pushback in some places on simplifying something, aligning it with user needs.
... that discussion is happening outside W3C
... having a place to brainstorm where there may be civil disagreement, is valuable.
IJ: I can see a W3C with three
things: new discussion forum, community group, WG.
... need to see what values are of current XG process or current IG process
mchampion: Some support in MS for
reusing existing processes where we can.
... some concern about a "community specification" and effect on brand.
... some time ago, IGs were really the open forum where deep technical discussion occurred; WGs did the mechanics of spec writing
IJ: This is the study phase; not suggesting we have 5 things
Arnaud: Agree that adding to what we have may not be a good idea. Phase I is figuring out what we want. Phase II is analyzing what we have and morphing something or renaming if we have to
IJ: see also the two objectives synthesis from discussion
# Simplifying our interface to the world
# Identify and eliminate unnecessary process slowdowns. Provide rationale to the community for remaining timing expectations.
* The vision statements
* The objectives. What high-level messaging is important to
communicate our intent? Are the objectives all in scope? What's missing?
* Whether the comparison table is useful; how to improve it, what's
missing, and so on.
Arnaud: One question of
clarification - what is meant by new idea forum
... "One idea: if a community group is created, discussion SHOULD move there."
is that SHOULD or MUST?
Arnaud: I see this as sort of a
forum for doing a public call for participation.
... perhaps some discussion of scope, etc.
... and then you move on.
[IJ note to self: group creation notification mechanisms]
lrosen: What infrastructure, services are available to newly formed community group?
IJ: Progression along multiple axes (infrastructure, legal, etc.)
lrosen: In Apache there are
various levels of responsibility, e.g., committer privs, v.
members who can vote, ...
... In IETF, anyone can participate and that can sometimes cause problems....
... in Apache, you don't just say you want to be a project...
... there is an incubator project management committee.
... once a project has started:
* anybody can read email
* only trusted people become committers
[Discussion of what hurdles should look like and how much]
[IJ notes that "peer-selected people" is more interesting than "self-selecting"]
Andy: One observation about the
two different systems (open source, open standards) while
acknowledging how they meet in the middle.
... when you are in the standards-creation mode, it tends to be more of a discussion + voting and ultimately more binary
... in the open source world, there is much more granularity - you can make changes line-by-line or module-by-module with multiple authors.
... the opportunities for problems may be greater in the case of code than in the case of standards (which may only be a few pages)
lrosen: I understand that there
are differences here...not my intention to sell the Apache
... trying to highlight that there are these mechanisms that have evolved in Apache to provide for a kind of equitable, peer-directed, reputational value assigned to people who participate.
... it has worked successfully in the software world...not sure if it would work in standards world.
... I agree it is overloaded with process...
... I merely present it as a model to draw from.
carl: The IETF went through a similar process...the reality is that each group, depending on what it's doing, has a set of processes.
carl: w3c is seeking to
distinguish itself (e.g., via its inception then its RF
... agree we should take best from various fora and tie it back together.
... you always compete with organizations that have less structure but that are transient
... W3C will face issues like (1) installed based (2) membership structure (3) may find itself easily conflicted
... much of the stress we are seeing is how to encourage ideas without disturbing the installed base.
IJ: Not reinventing W3C; adding a new offering we've not traditionally offered.
Carl: Management needs to decide
whether this is evolution or revolution.
... need to focus on amount of change management and members are willing to accept
... what are risks of creating a new track? not creating it?
mchampion: There seems to be less dissent on a "new ideas forum"
mchampion: perhaps we can move
quickly on that...and continue to ask people what prevents
people from participating.
... I think we can start the new idea forum quickly; doesn't require a process change. Suggest we make that happen sooner.
IJ: Should we try stuff out and experiment or wait to talk to the AC?
mchampion: I would have a bias towards action. I do think it would rude to modify the process, but something that doesn't require a process change or undermine the business model would be good to do and talk about preliminary results in November.
tlr: On open discussion fora:
Larry cited IETF IPR WG where an open discussion got out of
... if there are two sets of discussions that chronically get out of hand, it's IPR and process discussions....
... individual submission is a valuable piece of the IETF process
... what can W3C learn from this?
... is the lesson that the gating factor towards a lightweight forum....
... might be "put an initial draft on the table and you get a group to discuss it right away"
<mchampion> +1 to Thomas -- quickly create a mechanism to put ideas out in front of W3C community for discussion and review
Arnaud: How far do we want to go
into discussing the process?
... you can define a wide range, from sourceforge to Apache
IJ: Yes, we want to try to come up with consensus on values for various parameters. But not on this call or this week.
Arnaud: There are some strong
beliefs associated with various models.
... I would lean towards the trust position.
lrosen: We need to involve more
public since we are tailoring something to them
... we should engage the public in this discussion
<scribe> ACTION: Ian to put more explicitly in the wiki that public comment is welcome, and also set expectations about public outreach as the process matures. [recorded in http://www.w3.org/2010/08/09-newstd-minutes.html#action01]
<trackbot> Created ACTION-5 - Put more explicitly in the wiki that public comment is welcome, and also set expectations about public outreach as the process matures. [on Ian Jacobs - due 2010-08-16].
23 August, 11:30 ET for 90 minutes
scribe: agenda likely to be about feedback from next week's management meeting