From W3C Wiki

TPAC 2007: Session 5: Openness of W3C Working Groups

/!\ This session transcript needs to be edited. Please consider updating this entry

agenda audio


>> Welcome back.

>> I am happy to introduce Daniel Glazman and he is going to be opening up the discussion for openness of W3C.

>> Good afternoon everyone.

I hope you are not sleeping too much after lunch.

>> We are going to open the panel now about the openness of working groups.

>> This has been a can of worms in the consortium in the working days.

>> We all come from IGF where everything is opened to the world, everyone can be a troll for a few minutes, where all of the proceedings and meetings are public.

Almost nothing is confidential.

When the consortium was performed it was all open to the public.

>> It has simply changed and some of the working groups have decide today go open.

>> In the the external groups or -- [ Speaker/Audio unclear due to strong accent ] -- like the working group.

>> Some see the openness of the working groups like the -- [ indiscernible ] solution, but there are pros and cons to openness.

>> Do we want it and why do we want it.

>> We deal with the pattern and the property rights.

How do we do it.

>> We have no affiliation with them only with ourselves.

>> I hope the four panelists today will provide you with answers or a good perspective on the problem.

>> In appearance we are going to have Deborah/Debra Dalh.

>> Then we will have Art Barstow from Nokia.

>> IanHickson from Google and Paul Cotton from Microsoft

>> Hello?

>> Okay.

>> So the things I am going to be talking about today were gathered during, originally during the rechartering process during the mobile and the browser groups when we considered making the groups more open.

We had a lively discussion in both groups and there were a lot of interesting concerns raised.

>> I brought them back to the working groups and there were more issues raised within these groups.

>> There were some things that were raised, where I unfortunately don't have the answers.

>> These are the things to think about, basically, our concerns when you open up the working groups to more public.

I classified these into different areas, the availability of specks and having openness with the public and other organizations, this might save the working group a lot of work if they realize that there is a problem of this speck that is developing in the early stages.

>> It also exposes the working group reasoning and how it is clarified after the reasoning of the publication and how it leads to things like, why did you do this?

>> And then the working groups say, Oh my gosh, why did we do this.

>> There are some concerns that came up in relation to the specks, and that is the need to watch out for the IT R working groups and specks.

>> It may be more of a problem with more of a conduit to the public.

>> There is also a possibility of the public misinterpreting editors draft versus working draft.

My -- favorite antedote on that is the public on the published antedote of that.

It came out a good six months before the editors draft.

>> There is a difference between the working draft and the editors draft.

>> I've also heard that some people are concerned about knowing the people who agree and disagree with people make the consensus hard to contend.

>> So maybe the speck is not as sold as it -- solid as it appears.

>> One of the main things that people kept bringing up over and over again.

>> They are afraid they are speaking to their companies, upcoming products, and people not wanting to take the risk about talking about those kinds of things, and I think this is important with a -- competitive working group.

>> People are very afraid that some of their business confidential information would get out.

>> On the side of the -- relating to W3C as a whole, some of the advantages are working together in making sure what the working group is doing and that is it still active and also the other important advantage is to make sure you are increasing the ability in the standardized organizations as well.

The work in public.

>> There is also a voice in the public work actually belonging to the W3C.

>> The minutes, being a time consuming process.

The working group taking time to respond to noisy and uninformed comments and discussed about coming to a halt when the working group is speaking about something that is confidential.

People mentioned that voting and privacy concerns are, if we go as far as the publishing of the administrative information, like the cell phones on the base to base logistics.

And sometimes people don't want to post that to the list, but for whatever reasons, this can lead to people getting into unarchived public discussions, which is a very dangerous thing, because it is always possible to leave people out of the working group and the discussion.

It makes it difficult to tract rationals for discussions because they are leading through the e-mails.

>> There are a couple of other things: The detail discussion can turn people off, because it is so detailed technical then they think they cannot possibly be concerned in this stuff.

It is the public that becomes uninterested in the work.

It is also a concern that the publishing of the public list can make it volunteer run able to ex -- vulnerable to expand.

This is something you want to think about in making the working group and activities more public.

>> Great.

>> Thank you Deborah/Debra.

Ingly suggest that we keep the questions for the end of the session, so you can hear the full aspects of the discussions.

>> [ Speaker/Audio Faint & Unclear ]

>> Art?

>> Presently audio is unavailable.

>> Okay.

>> Thank you Daniel.

>> This is the third time I had a discussion this year about openness and Frankly, I am sick of it.

I am most certainly hopeful this is the last time we have this discussion.

>> Last may I had a discussion with Nokia.

I think it is blatantly clear from the first panel that we have to be more open and it is pretty [null], nine not to be.

>> I am going to talk about the upsides or what I heard people say or list, and talk about that and then talk getting a little more specific and I have a couple of recommendations of what we can do.

The relevant question is: I was very happy to see that my first bullet was well-covered this morning, and the first issue is who is our customer?

>> Is it the broader web community?

>> I will talk more about that, and I think the other question, that I have playerly heard many -- particularly heard in the advisory committee.

>> There is another issue here about the account ability and how to increase that for the various working groups and the last question that I was asked here is pretty open about the technical information from Daniel.

>> The main upsides that I see, and again, I am delighted that this came up this morning and that is this idea that we have to get more adoption and the community buying into the work that we do, and I assert that we will get that by being more open.

Clearly if the public can see the technical work and what is being done and we can contribute to those decisions, we will -- discussions we will get more feedback and be able to finish our specks more faster.

>> As Deborah/Debra eluded it is a great way to facilitate the liaison, for like Nokia who participates in dozens of consortiums in the W3C's and all of the technical discussions in the W3C's and that is an important value.

We will get a lot more participation from the membership -- from those members who cannot afford the W3C membership fee.

I also represent Nokia as the steering web initiatives and we have talked about the web and the countries.

I think it is realistically for us to know we have all of the knowledge.

How are the communities in the outside world that we are going to be covering here.

>> And the last idea here, increase of the technical work will result in new members.

I have heard the contrary in the W3C meeting last may, and they said if you open it up some of the members are going to work, and that is unfortunately, and they think that is going to happen, but I think if we increase the mobility work and let the public see what is going on here, then we will have more members.

>> There is a lot at risk and I would like to talk about them.

>> A lot of this come up in the hyper tech coordination.

The first problem is we would get too much input.

I obviously disagree with that.

>> This idea that we get too much spam is kind of ridiculous as well.

We can certainly go ahead and use the right access initiatives and prevent them from happening.

>> The third bullet is a huge issue.

And that is the IT R, and that is no license and no license interns.

>> For Nokia, that is a big issue, but I presume what they are working on right now is scale able and can be applied to the working groups as well.

>> If you went to their mail list then you have to agree with the policy and that is a do able solution to that risk.

>> Last, there is supposedly some kind of risk, if it was done more publicly, then we could decrease the value and I couldn't disagree more and that is coming back to that important benefit and that is communicating with the external organizations, and that is important to us.

>> So, more specifically here, with a couple recommendations with the respect of confidentiality, I would like to see is the default mode in and that is perhaps the only mode that charters would have and that is all discussions are done on the public mail list.

I support the member only list, my working group has it, and we use it for the non technical discussion, you know where we are going to eat dinner tonight.

And it is easy as a fair to separate that out and make sure everyone is on board.

>> One thing I noticed the L working group did, and I don't know who chaired that group, but their charter said, for all of the charters who have the nice milestones and the speck and when you are going to get those milestones, but it is right in the charter if there is deviations to that.

They will make that available.

That is important to a company like Nokia.

We recognize what we are doing here, and we need accountability with the respect of schedules.

And what we are looking for here, is the relatively easy way for the working groups to let the world know where they are as compared as to where they said they would be, and I think that would be very helpful.

>> Thank you.

>> I apologize for now because of my voice.

>> My perspective is coming from the working group which is quite dramatically opened, but when we started the working group we had a vision of what it was but we dent know what it -- but we didn't know what it meant.

>> That is what I am going to do here.

>> First of all, why would you want to be open?

>> The answer, as everyone else has said is: The public is large and tactful in a group.

>> When you are in a crowd they are typically nice to you, when you are -- we can use this to our advantage, and we can use it to work out other -- [ Speaker/Audio Not Clear ]

>> By having a group that is truly open, we can actually get continuous feedback from a broad range of people, not just those who have a membership fee or those who have invited them from the group.

You can have a Google alert to mention to you when you have an alert any where and you can have that at any time you want.

>> [ Speaker/Audio Not Clear ]

>> You will find a lot of the people will come in and change of the work.

That is great.

You need to write the tests and the implementation reports, what is wrong with the speck, the editorials and the list goes on.

>> So essentially the open working group is not just a working group, it's a community.

>> Your interactions can be real when you have a community within the group.

>> You don't have a community you have a working group and an audience.

>> Having a community means treating everyone with respect, obviously, you are basically running an openness project, and it is also being tough on those who crawl.

>> Stay polite.

I believe it really means, and being quite radical, and what you really mean is: I believe you should give everyone access to the blogs and you can all go post right now to the working groups, not comment on it, but post on it.

>> [ Speaker/Audio Not Clear ]

>> People should feel free to go ahead and set up a blog or forum or IC channel and you should have all of those.

>> That is part of the channel and have the open meeting for discussions and for the -- [ Speaker/Audio Not Clear ] -- and what would be best here.

>> How to use the speck?

>> Have a fact that anyone can contribute to.

You can contribute questions and answers.

>> You can have a problem of course, and the advantage of having an openness community you will find the people who just happily come along and volunteer to maintain that for you.

I had to do very little work to maintain part of the blog or the working group.

>> [ Speaker/Audio Not Clear ]

>> This can actually be done pretty cheaply, you will be surprised you can find the hosting providers that provide all of the above and the versions and anything else you may need, and for example, that is the working groups and the sites formed with them.

I say out of pocket, 6-dollar as a month, I just covered that.

>> [ Speaker/Audio Not Clear ]

>> It turns out that the service provider is actually not much more unreliable as the [ indiscernible ] hosting.

>> It has just as many problems with the working groups that we do as I have heard people do within the public.

>> There are things that you can't do, you can't have the conferences in my opinion.

You can imagine having 800 people on a call, it just wouldn't work.

>> I would like to say you could have face to face discussions.

>> [ Speaker/Audio Not Clear ]

>> The biggest danger in my opinion is that you will not be open enough.

>> It's easy to create a working group, think you are open, have an open mailing list, or have a blog that you post on, but not actually letting them comment on them.

>> [ Speaker/Audio Not Clear ]

>> I am not sure what working group is doing enough, I'm not sure how -- for example, I am in the west group with our tiers, and I think the west group can do a lot more.

>> I think the danger is if you are not truly open, then you will not get the sense of community.

>> There is also the danger that you mentioned, which is not a lot of danger, but you get a lot of feedback.

>> You will have thousands of e-mails to reflect on that feedback.

>> If you want feedback, that is a great thing.

>> Sometimes if you don't want a quality or immediate owe cree speck, then I would suggest that you don't do it.

>> The last thing I am going to say is: When you are open you are the -- [ Speaker/Audio Not Clear ]

When you are W3C, it's still a pretty small group, we can mesa than begans, [ Speaker/Audio Not Clear ]

>> Don't under estimate the power of your community.

>> In conclusion, I think if your goals are noble and honest, being completely open is the way to go.

But being -- [ Speaker/Audio Not Clear ]

>> Thank you.

>> Hi, my name is Paul Cotton.

I am part of the W3C working group.

>> First the [ indiscernible ] query working group and then the policy.

>> I want to provide some of the practice practical data from my experience.

>> To start off with, I heard some of the people say that the panelist all speak a different jargon.

>> This is my sketch of what they process looks like.

We have communities that tell us we want to do new work, we have a workshop, someone drafts a charter we discuss it, we get working groups, open to discussion, we have consensus, we get candidate feedback, working drafts, recommendations, and hopefully the last step is the smallest and that is -- [ Speaker/Audio Not Clear ]

>> So I thought about this, what is the working group in the XML query.

>> There was a substantial amount of new members to do that work.

It was and always have been a member working group.

There were initially no invited experts in that working group at all.

I had 30 people and cross section representation in the industry, people who wanted to have a query language for XML and the academics as well.

>> And in fact, my most common usage to have was to sustain people when they left the jobs.

We left one working community to another working community, and that is why I invited the experts.

>> We started with a blink piece of paper.

In the nurse couple of -- first couple of years I spoke six or seven times publicly.

>> Not only does the chair or someone in the working group have to get out in the community, but you have to tell the community what you're expectations are.

And many of those are still available on the web today.

If you will go back and look at them, I didn't shy away from the hard questions that we were try to solve.

There is multiple or working force meetings that happened every work.

There were times we were working with the group three or four hours on the phone.

We had face to face meeting every two months, we had well over 350 e-mails a month.

I am a big advocate of publish early and publish off and that is how you tell the -- publish often, and that is how you tell the communities where you are, what you are doing.

>> If you coned the cooperation with the XML working group and yours as well we have more than that.

>> We have a lot of comments.

We had, in fact, 1200 separate technical comments.

I want to reflect on that.

>> It is interesting, not very many of them come from the public, they actually came from the W3C members who were going to implement the specks.

>> But they were not really ready to implement until we went through our last calls.

>> The candidate and the working groups worked on the specks for a long time.

>> I do believe we added significant value to the IT community.

Let's look at the W3C policy.

>> This was started in a blank participation.

>> We brought the complete community who drafted the -- it has a published list from day one and no experts.

We started with the specks and one of the companies submitted a primer as well.

>> We had less need to go out and vaned lize to see what was going on out to happen.

>> We still had face to face meetings every two or three months, weekly face to face, and about 125 e-mails per month.

>> In my role as the chair in the W3C it's my encouragement to publish early and often.

>> We only had two on the wreck track, and we only had a small number of comments and we only added a small amount of features to those specks.

>> We did actually expand the interoperable community during the recommendations Z package we did a lot of -- and we did a lot of that to make sure it occurred.

>> We did a lot of recommendations for the group.

We were able to publish the primmer and guidelines at the end of this month at the end of this meeting.

>> What are some of my summaries and experiences here?

>> Well, to start out with, I think you can do work like query and policy did, without inviting the experts.

We have a large community here at W3C that brings the expertise to the table.

>> I think I benefited from the fact, from the two employers who were pleased to have me go out and -- -- that W3C didn't have to pay, they got that through my participation.

I am a firm believer that starting with a concrete submission that it is a working support for the community.

>> It is very important, and some of the other speakers spoke about this.

>> I have said this, and I will say it again, publish early and often.

>> I think that is one of the important things that the working groups can do in the working communities.

>> Acting as a liaison is also important.

>> Those in the XML query effort will know that we spend a lot of time with our friends in working groups and meetings and discussions.

>> I think publishing a companion primmer.

We have to think ability the audience in the -- about the audience that we are speaking to.

Having the primmer is one way to reach out to the novice in this particular area.

>> They don't want to read the actual speck, but they want to read something where they can read it in a half hour or so and see the differences in the technology.

They may never go and read the speck, but that is something the other working groups can do.

I think the current process actually works, I think it is expansive, it draws the community in and I would like to see us continue using the current process as it is today.

>> Thank you.


>> Okay.

>> Thank you everyone.

>> I would like to review a few notes that I took during the four talks of the presenter.

>> Some of them will be conflict usual.

>> We saw in the XML working group we have 400 or more working groups and it has increased the burden on the W3C staff working conditions.

>> I see levels from your notes.

First is about the communication, feedback and participation, and editing the speck.

It is not only contributing technical things, but you are part of the [ indiscernible ]

>> This provides the number of consortiums on the commitment.

>> [ Speaker/Audio unclear due to strong accent ]

>> What becomes the commitment on the external submission of the working group.

They can disappear or go away.

>> What happens to the charge of the document, what happens to the documents themselves the last thing about the legal issues, and Ian said that wiki's should be open, and this has caused issues in the countries.

They are responsibility for what happens in the blog, they are not everyone.

>> That is just hence per your questions, and you have microphones so please come up and we will distribute those among the panelists.

>> Please don't forget to say your name and where you are from.

>> Use the microphone please?

>> No it doesn't work?

>> It is on now.

>> Ken [ indiscernible ], from the original working groups, we are talking about more openness and more people involved is that any group tends to start with a large group of people and very quickly widows down to a quote unquote of those who get the work done.

>> If you open up the participation whether completely along the way you would have more participation.

Especially if you get into the details of the work, if you are not into the details, you get lost quickly.

I know I have sat down and read e-mails after a weeks worth of work, and then I read e-mail number one and I have no idea what is being said.

>> If we open it up more, do we receive more useful input?

>> Well, it depends on which group, in the XML case we found that the participation was large.

Different people have expertise in different areas and you end up with different commenting on different areas.

>> The total number of people is different, but this is with the different areas.

>> Would you say within the working group that the percentage of people who remain active are the same as with the smaller group?

>> Yeah, that is pretty well correct.

>> I would say in the groups I have seen you quickly get down to 10% of the people.

>> The working group has about 800 people, and the [ indiscernible ] list had five or 600 people and in both cases there is a core group and that is an active group everywhere and that is five or 10 people and then you have the extra 50 or 80 people who have their additions on everything.

>> It is a working group among all of the people.

>> You have 350 e-mails every two months, so it's more than that.

So to be able to fulfill it and to keep doing things on the mailing list, you have to have a selected few people on the mailing list and there is a few others that don't control it too much and that is it.

>> I want to follow up with a antidotal evidence in the that chair.

>> If we were not doing that, we wouldn't get the occasional comments or the really good feedback for us.

If we were doing everything we wouldn't get that at all.

>> Another thing is: When the working group is started as a two year charter, there could be people that want to participate in something that they can't.

>> That is for various reasons, maybe they want to change the employers, and then if you close that and they change the jobs and responsibilities they won't see that.

>> Steve?

>> So HTML5 is clearly seemed to be the [ indiscernible ] communication, and the XML is the business to business or [ indiscernible ] expectations.

These seem to describe the standards that Paul Cotton were talking about versus what Ian and Art was talking about.

>> Is there room for the two different models regarding the particular audience and what the standards is.

>> I understand what you are talking about.

>> The -- [ Speaker/Audio unclear due to strong accent ]

>> And you have everything behind.

>> And these are two different communities.

>> We had had this on the web panel this morning with Molly.

>> And industries and banks and [ indiscernible ] who are interested in the web services and signatures and things like that, so yes the question that comes to the conclusion of your point, can we have a consortium that is particularly opened in default closed.

>> I just want to make is comment that the openness factor may vary for a particular substitution even in its stage.

>> If you have the working gyps or the published working group, I think when they have those representations and they are -- recommendations and they are being used out in the community, I think without a doubt you have to think about the errata process as much as possible.

We don't know how many people are vested in the sense that users are adopting them and users implementing them.

We owe the community a strong service to make that as open as possible.

I believe that part of the W3C process works as well.

Many working groups are using that as well.

I guess I would respond to Steve's point, that there may be vary ability within the working group as it moves through the different stages of the development and the maybe of the specks.

>> This is Tex [ indiscernible ]

>> I work for [ indiscernible ] working groups.

I like Art's approach, the [ indiscernible ] is obvious and everything else is assign.

>> We want transparency and we want the information available.

>> Well there is another part to this.

>> The answer to me is: You don't need one list where everyone is in the same pool, but certain lists towards the certain aspects that you are developing.

It can be toward the same dialogue that is open, but it is engaged on the particular [ indiscernible ] that makes sense.

>> Some want to talk about usage, authoring, and so on and so forth.

I recognize the teams that need to be implicit.

>> You may not be reaching your user community, you may be reaching the mob roll from the particular perspective.

>> Since most of them are in the English language you are not acting throughout the community in the same way.

We may be overlooking the requirements or overruling the requirements because we are paying attention to communities in this.

>> I am sure that other folks have the similar concerns.

How do we address this when we are doing everything in English which is not speaking to the entire community.

>> Ian?

>> First of all, you will have more.

Some people have volunteered to translate.

>> But if you look at the blog you will see the [ indiscernible ] which is the translator into the English.

>> Unfortunately this part of the community hasn't read it as much adds I would like it to, but we haven't had much feedback the other way, but yeah, that is an area that I would like to develop.

>> [ Speaker/Audio unclear due to strong accent ]

>> I am actually a company spokes person, and I can actually speak on behalf of a company.

When I am doing it under their side, I have a PR person with me, I am doing it in a special way.

>> I have to get the statements cleared the six lawyers and marketing.

It is hard for me to function when I am speaking on their side all of the time.

>> I have been in cases where I have put things out in the close forum that are visible.

>> Your journalists have picked up on them.

>> And the other case is, sometimes I will say things, that is not on behalf of my company or me, but one of the things to get past the conversion is to put ideas out there to posiest interactional position.

>> [ Speaker/Audio unclear due to strong accent ]

>> There has to be an allowance where the conversation is off of the record and that is where the discussion takes place.

There are some things that can't take place in the bright light.

>> All right.

I agree.

>> I personally agree, because, let's say your company is seven representative in the consortium environmental groups, but only one, like you, is a speaker for the company.

>> Everything is visible and the e-mail address is visible and the older public and press is going to attach the name of the company to the individual, that person becomes defact to spokes person for the company.

>> It is a real problem and issue to solve, I agree that we dismiss it.

>> First thing I want to say, I think they have the drugs for the schizophrenia and more seriously, if you agree to do all of the technical work for the public and maintain a list for the discussions and I suppose you can broaden that to having dinner than the sensitive issues that are raised.

>> Ann [ indiscernible ] Boeing.

>> I am one who has spoke and it's uncomfortable to speak about the openness, personally an open person.

>> One of the main reasons I say to my company to say is that we have privacy in your conversations.

>> I would be crease to here privately from the other issues if that is an issue.

>> This is a serious question for the consortium for the whole.

>> Paul's claim on the open or closed for the different stages is something to look at.

You know, some sort of middle ground.

>> Thanks a lot for having this panel.

It was interesting and useful.

>> Jerry Carter for [ indiscernible ]

I want to be careful on the subject of I PR, we cannot sweep that under the respecting.

Under the occurrence of the policy, that is a level of the corporate entity rather than the individuals and the commitments to join the working groups? I am not quite sure whether the specification can be relied upon.

>> Okay.

>> If you can't follow the model that is being used by the HTML working groups then take that back to the XML voice farm or where it came from.

>> I work with the two groups privately.

>> There is one aspect of it that I have discomfort with.

There is no negative feedback.

>> There is none in the feedback cycle.

I have a funny feeling in the pit of my stomach when I submit anything to the list that is read, because I know there is no bound on the comments that will come.

>> In the tag, however, I haven't seen it being a lot of problems, most of the comments are good and worthwhile.

I have seep working groups spend a lot of time implementing a process that is basically, if we have 1500 comments we have to write 1500 answers.

I am not sure that is a good use of our time.

>> There is nothing that is going to prevent that from going 10 to 15,000.

>> I think you have hit the nail on the head.

No matter what process the W3C has it has to set expectations for people and it has to be explicit what those expectations are.

>> We have to respond to every public comment, but we were not going to respond to the working groups and their comments, because they are part of it.

>> And the tagging that you use, if you go to the home page, there is an explicit protocol on how to get the tagged tag's attention.

I think the participants and public.

>> Last question.

>> I would like to share a little bit of practice that we pinched from the device description working group and we have tried to get this announced public ally and privately.

>> So what we do for the sake of 10 or 15 minutes of my time I write up a draft log entry of what we discussed in the meeting and then I post a round list.

They are able to come in and correct that on anything that is wrong, and then it gives a public digest that is readable, and foilable of what we are doing.

>> [ Speaker/Audio unclear due to strong accent ]

>> Just one additional comment, what you just said is exactly what I said in the beginning openness increases the burden on the numbers.

It's more work to increase openness.

>> In my opinion it's more work not to increase openness.

>> I think you have heard owl the stay -- you have all heard the comment that people would like to die rather than speaking in public.

>> I would like to introduce a new set of members saying they won't bite you.

>> This particular organization concentrates a lot of stress on the chairs.

I know my co-chair is stressed about a lot of things.

He is a person with a family and kids and tough like that, so if you pray, say a prayer for the XML working group for the co-chairs.

>> I just want to make a plead to the A3C reps out there.

I know they have been wrestling about this for four or five years and if you think that openness is important area your membership value, please consider that when you go to your AB.

>> Thank you very much.

We don't have time today for anymore questions.

>> Thank you very much.


>> Thank you that was an excellent panel.

We don't have the remedy on this issue, so we will continue to talk about it.

>> You will look at the average group now it is more open more than it was 10 or 12 years ago.

It's generally the way it is going.

>> I am wasting time why we railroad waiting -- while we are waiting for David to come up.

>> Okay.