Re: Workshop and meeting requirements

Omnibus reply...

==Introduction
This proposal isn't optimising purely for very fast work - it may easily  
introduce a delay of a few months in the development of a standard. I  
consider this an acceptable price for a model of standardisation that aims  
to privilege building a global consensus over enabling the most powerful  
market players to dictate what we should have.

However, such a model is also not sufficient justification for doing  
everything slowly just to tick off process requirements - those  
requirements should only be there if they serve our overall goals. Being a  
good place to make Web Standards is one of those goals, and while I find  
almost no evidence that a few weeks delay has *ever* been very important  
to doing that, it is clear that a few weeks delay imposed repeatedly for  
what seem purely administrative reasons (rather than a genuine lack of  
resources, or the difficulty of scheduling agreement) is clearly harmful  
to the way we are perceived.

==Glazou's comments
Glazou said we should teach staff what the process says. I agree that  
someone telling him that program committee was a *requirement* is a MAJOR  
FAIL, and that the staff need to know and correctly communicate the  
process. But that struck me as a one-time bug that should be handled by  
the staff, rather than something we could fix by changing some document  
they apparently hadn't read carefully enough…

==WG Meetings - Chris' comments further discussed in particular by Karl,  
SteveZ
Chris argued in favour of allowing shorter lead time because '8 weeks is  
onerous' and because given decent remote participation any event is a  
virtual distributed meeting. He also pointed out that in organising a WG  
meeting he would optimise for making sure a set of people he identified  
would be in the room.

On the latter point, I agree. I organise Webapps meetings in Silicon  
Valley because although it is unfair to the rest of the participants to  
always have to make that trip (which for those outside the "US-friendly  
visa zone" is often a major hardship as well as very expensive) it is  
optimising for the likely presence of people we still consider critical to  
the success of a meeting.

But I am very reluctant to reduce the notice time. I honestly think this  
is a key difference between working in the spirit of W3C as a global  
transparent forum, and asking them to rubber-stamp the work of a few  
people - even allowing for the fact that most of the work in any given  
product of W3C is done by a small handful of people. To borrow from the  
legal world, fair transparent process not only has to be there in  
practice, but has to be seen to be there.

I find the argument that remote participation facilities allow all  
meetings to be virtualised uncompelling. While it may be the case for a  
Working Group who know each other well and trust each other to implement  
good remote participation, in practice there are many critical variables  
that depend on human factors - the chair and and the people in the room,  
as well as on technology that is still pretty fallible. The recent split  
WebRTC meeting at TPAC Shenzhen was reportedly not a success.

In the case of Working Groups, I propose not to change the status quo -  
meeting notice can be shortened unless someone objects. I think this  
provides the right balance between providing flexibility for groups who  
work well (on a "social interaction" level as well as mastering the  
technology available) and trust each other to do so, while allowing  
stakeholders - especially those who are paying W3C membership - to require  
more consideration for the global nature of W3C.

On Mon, 12 May 2014 17:06:55 +0200, GALINDO Virginie  
<Virginie.GALINDO@gemalto.com> wrote:

>> I’m supporting the WG operation clarifications you are offering,  
>> including the asynchronous decision to be clearly stated in WG methods.

That's really a separate issue in the process, but it does have an impact  
on how we deal with people meeting…

>> Sounds good to merge the sections related to WG and Workshop (and to  
>> kill Symposium notion).

I don't mind what events are called really, and as Glazou pointed out,  
sometimes the name actually matters because it sets up expectations around  
the type of event we have.

>> I am more skeptical to the idea to merge Workshop and F2F meeting  
>> requirements. Workshop have their own specificities.

Yes. I didn't mean that we should set the same requirements for all kinds  
of meeting.

>> Workshop on a specific topic happens once every two or three years (if  
>> successful), they are an important milestones in designing the >W3C  
>> action plan, as such we should make sure no-one misses it. So like for  
>> F2F groups 8 weeks advance notice is ok but agenda >should be published  
>> earlier then WG, as we wanna make sure people will be able to book  
>> travel and confirm their interest in this unique >event, I thus  
>> recommends 3 weeks.

This depends on the Workshop style. I think the prose should suggest that  
long agenda lead is helpful in many cases, but don't know that we should  
make a requirement.

The examples of the games workshop or the recent TAG event are different.  
The events that Alex wrote about last year with position papers, and the  
payments workshop with a committee, papers and the whole nine yards  
actually did allow people to say "I work for an organisation that is  
obviously relevant and we want to attend", and get an invitation.

I think we clearly want to allow such events to take place. We have done  
so in the past. A few years ago people were doing this as "Barcamps". A  
decade ago as W3C staff I was organising W3C-sponsored open-ended  
exploratory workshops outside the remit of any working group as part of  
the SWAD-E project, in places like Hungary (the year it joined the EU) and  
Argentina (still not a mmber… ;) ).

I understand we want everything now, and that we have attention spans that  
prefer a 6-second video clip, but it seems that most people really are  
quite capable of deciding to talk in depth about complex issues in 9-12  
weeks instead of right now or never.

Indeed, while I find it personally hard to discern a scheduling pattern in  
the WHATWG (although I know Hixie publishes some information on this) it  
seems that quite often discussions there take place months after issues  
are raised, and nobody seems to have a problem with it since they are  
aware that this is how things work.

>> Workshop may fit with a large number of attendees – up to now, have  
>> seem from 50 to 120 and have sometimes social events, funded >by  
>> sponsors. This makes the logistic heavier, so confirmation of  
>> attendance is expected one week in advance.

More or less any face to face meeting can have this requirement, but I  
think it should be explicitly left to the organisers. If there is such a  
requirement, it should be published when notice is provided.

>> Reports are hard to write for workshop, impacts on W3C action plan is  
>> huge and requires approval of program committee, which is not >the same  
>> as getting approval/review of chairs and staff, so I suggest to give  
>> more time to issue report.

This depends on the type of event, but yes.

Reports are also important, IMHO.

>> Workshop as you mentioned may require position paper to attend, lets  
>> keep that clear in the future chapter

I certainly want to allow the possibility of events organised like a  
traditional academic conference, complete with real papers.

I just want to be clear that there are other ways to organise workshops,  
for example the "extensible Web" thing the TAG did…
>
>> I understand that you want to give more freedom to organizer, but on  
>> the other hand, it would be a pity to ignore experiments from  
>> successful >workshop (such as the recent payment one). What do you  
>> think about suggesting workshop constraints in the future process  
>> chapter with something >like.
> >  >Group Face-to-face >meetings >Group Distributed meetings >Workshop
> Meeting announcement >(before) >eight weeks* >one week* >eight weeks*
> Agenda available (before) >two weeks >24 hours (or longer if a meeting  
> is scheduled after a weekend >or holiday) >three weeks
> Participation confirmed >(before) >three days >24 hours >one week
> Action items available (after) >three days >24 hours >one week >days
> Minutes available (after) >two weeks >48 hours >three weeks

>

Yes, I expect to have something like this. (This table is why the email is  
in HTML - and the mess that created is why I want to find more time to get  
the HTML email CG running and moving the HTML email world forward…)

>> You did not really address the conference question. I think a dedicated  
>> section would help, stating that W3C can also organize conferences,  
>> which >specificities will be discussed on a case by case basis.

I agree, despite the fact that it explicitly competes with members, and  
with potential members, that we should accept that W3C *does* run  
conferences of various kinds. W3Conf, and the MLW series are the examples  
I know best, but there have been other kinds of event. Including  
conferences where there isn't an open selection process and speakers are  
simply invited. 'm not sure how far we should go in putting this into the  
process, but it's an issue I will try not to lose because the current  
process says it doesn't happen, which is plain wrong.

>> Hope it helps,

Yes. Thank you (and everyone) for the comments.

cheers

Chaals

-- 
Charles McCathie Nevile - Consultant (web standards) CTO Office, Yandex
chaals@yandex-team.ru Find more at http://yandex.com