W3C's 'Moral Majesty'

[Tim Berners-Lee]

Don Park (donpark at docuverse.com) on  11 Sep 1999 19:06:05 -0700
In http://www.lists.ic.ac.uk/hypermail/xml-dev/xml-dev-Sep-1999/0635.html
said:


>1. F2F and weekly telephone conferences
>Without a doubt, both forms of meetings are valuable.  Faces and voices
>encourage people to work closer and reduces extremism through bonding and
>group behaviors.  However, these meetings are not efficient enough to
>justify the cost of participation to individual members.  If the WG member
>is an employee of a large company, then the cost is less and even enjoyable
>to certain extent.  If the WG member is an independent
developer/consultant,
>the cost is high enough to prevent participation unless the member also
>happens to be the Chair.


This a crucial point: whether the cost (and the unfairness that introduces)
is worth the greater understanding and trust which builds in a face to face.
It is not a question of membership in W3C - as there is a way (as you found)
for non-members to come to meetings as invited experts - the criterion
is commitment of effort and acceptance of the charter.

This thread has had both points of the argument made closely and
I am not going to resolve it here.  I understand it very well: I used myself
personally to go to IETF meetings, but I found the three week-long
international trips per year was too much, and as a result I have not
found it easy to keep track of what started life proud to be an electronic
forum for just the reasons you mention.  I even notice that IESG membership
is chosen selected from physical meeting attendees, so one cannot be
part of the steering of the IETF unless one travels. (You can be on the W3C
advisory committee AC or AB without travelling.  But I have to say, the
twice yearly face-face AC participation is very valuable).
When we realized a forum for all chairs would be really interesting, we
started with a face-to-face. The problems of how to run groups well
were so desperately  important and fascinating and frustratingly
deep that it was a really great meeting and everyone resolved to do it
again. But when it came to find a time to do it, no one wanted to travel and
we went to phone conference from then on. There is no easy answer.


>I have been to only one WG F2F meeting and I found it painfully boring,
>tiring, and also disgusting to certain extent.  Creativity is discouraged,
>progress is measured in number of easy decisions and issues raised,
>difficult decisions are deferred without exception, myriad issues are
chewed
>enthusiastically only to be spitted out and left on the floor like cheap
>gum, political complications are raised and usually dealt with sarcastic
>jokes which leaves the conflict hidden and dangerous, dinner conversations
>are laced with petty thoughts, bureaucratic whining, and blatant discussion
>of personal gains.  Discussion of F2F meeting location is not based on
>reason but pleasure without regard to cost.  A meeting at South of France
is
>great if your company is paying for it but not if it is coming out of your
>pocket.

(Or if you happen to be someone who does not live in the USA.
That is another form of exclusion we have to battle with.)

Yuck - that sound horrible.  I hope that you didn't come to any general
conclusions about all meetings of all working
groups. I also hope that you brought this up first with the chair, in an
effort to understand and together improve the situation.  If that didn't
work, I hope that you knew that the staff contact which every working group
has is there as an alternative channel.

The chairs meet regularly by teleconference and have an email list
chairs at w3.org which the general problem of "the art of consensus" (as the
member guidebook is called) are discussed.  Other WG members concerned about
how the groups work are welcome. Everyone is of course trying to make the
process at once open, efficient, fair and of a high technical quality. There
is no short answer, but there is a steady flow of suggestions though from
the chairs forum (and others) via advisory board approval to the process
document and the guide.

There is also a trade-off between hoping that people will use their common
sense of social sense to be fair and work together, or whether bureaucracy
should be introduced to force them to and drive them crazy.  I have seen on
this list general outcries against W3C bureaucracy, along with requests for
more administrative systems to be put in place to monitor working group
accountability.


>Telephone meetings are more efficient than F2F except there is no room for
>in-depth discussions.  It is a breeding ground for hasty decisions without
>proper representation of the issues.

Well, so telephone meetings are rejected, I suppose?  By the way, some
people can't afford those either unless they are called back.

Working groups are largely free to chose the form in which they will meet
simply because each working group contains a different mixture of people,
financial resources, technical complexity, social pressures, time pressure
and   so on.  This is not something which W3C process puts a lot of
constraints on - apart from for example forbidding going for more than 3
months at a time without publishing their results.

>2. Mailing lists
>Both WG and IG mailing lists are great but unsearchable and private.

That depends on the mailing list.  The XML Signature list, for example,
is open.

>Confidentially is too broad and laced with political issues which ends up
>reducing representation of the outsiders.  Issues are raised, dropped by
>inattention, and killed by time.  Field of vision in mailing list
>discussions is one dimensional (even with hyperlinks) which restricts
>awareness and thus limits the level of complexity the WG can deal with.


A call for open working group lists.  Maybe we should move in that direction
with new charters.   If you open a group to the public, then you cut the set
of things anyone will be prepared to say. For example, an employee of a
large company will be concerned lest his personal engineering opinions be
taken by the press and used out of context to stir up scandal.  As you point
out, making the group visible isn't much use unless it also responds to
input - and opening up input with no filter can drive a group into the
ground.  What happens in the W3C - and needs to happen in the community - is
that a common problem is identified, that the solution fits


>All this is encouraged indirectly by W3C and its policies.

W3C is a lot of people trying to get things
right.  It started with very few policies and was after a while slammed for
not having any. Those people who were concerned spent a lot of time
trying to improve it.

>If W3C can change itself, it should first relax its policy to reflect
public
>opinions more formally rather than at mercy of WG members' kindness.

Good point.  Though to do something formally isn't to "relax" anything:
it means a lot of hard work. More meetings. More rules - i.e. more
bureaucracy.

> It should also install policy of independence from W3C bureaucracy.

What on earth do you mean by this?  Install a policy of W3C
(the good side of people working together) being independent
of W3C (the bad side of people working together).
You are asking for more bureaucracy: rules about WG accountability
to public mailing lists. Who is going to be independent from this?


>The
>director should be stripped of his unspoken right to interfere with his
>unfairly heavy hand.

You have the shoe on the other foot, I think.  It is a spoken right - nay,
written.
According to the way the original membership contracts were drawn up,
formally the Director decides what becomes a Recommendation and
what does not.  In fact there is a process
http://www.w3.org/Consortium/Process
by which reviews are done, and the director as a person has very little
part.
There is also an ethos - discussed
largely by chairs and at the AC meeting - that we do not count votes,
but ideas: we do not simply override a minority opinion, but if it has
not been addressed, it is recorded and passed up a level.
At the final AC review, a comment in just one review can stop the
whole process until it has been resolved.  Only occasionally have
the staff had to point out that a negative review comment was
not technically sound - normally an analysis leads to either a change
in the document or a withdrawal of the comment.

Complaints about the heavy hand of the director (HHD) or
for that matter the staff (who, no matter that they are giving up
perfectly good and much more lucrative jobs in industry to
be there and have very little reason to be arbitrary or
capricious) being unreasonable have historically occurred when
the staff have had the indeed heavy burden of ensuring that
all the bits fit together in the long run.    It is all to easy to
design a wonderful and beautiful part of the machine in a working group,
and there are documents to read and write, and meetings to go to, a-plenty
without having to constantly review everyone else's work.

Here we are for example with differences about how namespaces are going
to be used.  It's not just a question of designing a namespace spec, but
it being one which the other XML groups and SMIL and RDF and P3P
will find meets their expectations.  This is hard work.  It involves
crossing boundaries of the use of vocabulary.  It involves the same person
being
aware of a lot of projects at the same time.  It tends as a result to be a
full
time job.


When the full time person (a staff contact in a WG) has to try to bring
two WGs together, it may look to the WG as though that person -- a minority
in the WG -- is claiming some sort of divine priority.  In fact the poor
soul is bearing all the stress between two lively groups of independent
creative thinkers.


We really need  common underlying philosophy about how the future web
will work.  I have joined in on this list from the technical point of view
because I was
staggered to find there seemed to be big holes in that common understanding.
On my postings I have maintained the right to state my own personal opinion.

> Second, W3C should start using and investing in
>groupware tools that allows efficient and focused online meetings,
>bookkeeping of issues and decisions so that status, factors, and
>justifications can be seen at a glance.


Oh boy, are you preaching to the choir!  There is not a staff meeting which
can't be derailed into a three hour rathole by talking about tools we need
and how they should work.  A breakout group in the face-face chairs
meeting worked on "Tools" and filled a flipchart with a small print list
of tools which we clearly essential.  A tools group was formed but everyone
is so busy. Volunteers didn't seem to do it.

Actually, paid staff have gotten quite a way. We do have tools to
- allow HTTP 1.1 write back so minutes and issues list can be edited
directly during meeting; (Amaya/Jigsaw)
- track changes so we can take chances, get messy make mistakes - and
recover. (Jigsaw/CVS)
- diff changes, validate, check links, etc
- control access flexibly - so we can be flexible about group membership
- create new documents from templates  - to save time
- bookkeeping issues: event tracking agent (ETA) This is a PHP3/mysql tool
which allows issues to be created, edited, tracked. In its early days.
- meeting registration;
- phone bridge monitoring, caller identification for faster roll call
- mailing list creation, management, audit,
- mailing list archives we are adding functionality to the open source
hypermail right now;
- the mIRC author kindly added URI tracking for us;
- and so on

Of course we have limited staff effort - meeting organization and group
coordination and running the website and stuff takes time, and there seems
an
infinite wish-list. And it is all funded from membership fees which even
those who are committing many times that amount in their own effort
seem to question.


> We must change W3C to fit the
>changing needs.

Absolutely.  That is why the process provides for its own evolution.

> W3C must allow us to change it because it can not change
>itself.

If you   change W3C then you add to W3C so you are W3C.

>Most of all, W3C must stop being arrogant bureacratic fools with
>grandeur self image of being THE innovative leader with THE right vision.
>To me, W3C is Tim Berners-Lee.

Thanks. I will keep both of those remarks for when my head gets too big
next. Let's face it I have this title of being The Inventor Of The Web  for
some of the media and their chief concern is often why I don't want to drive
around in big cars and tell everyone what to do.  Let me tell you that W3C
is not me
any more than Australia is the Queen of England. We are all innovative
leaders
and none of us have the right vision. But we all try. And you and I reserve
the
right to tear each other's vision apart.

>  Tim, you must stop thinking that you have a
>monopoly on the future vision.

[BTW When you address me please include me in the To: line of the message
too!]

Actually I was pretty happy doing techy stuff and trying to build a good
process. But in the Tokyo AC meeting among other places, people stated very
clearly that one of the things they wanted was a consistent plan for the
future, a vision of where we were going, coupled with a derivation from that
of the work and relationship of the various groups - and that that was my
job.
I am philsophically very against the "One True Vision" idea and am very
happy for others to
write theirs. In fact we all refine each others ramblings into the best
consistent framework we can make at one point.  I try to deliver the result
to not just the W3C
fee-investing people but anyone who can grab a web browser or a Intl. WWW
conference.

And don't think that my personal vision thing is all that is going out of
W3C.

There are plenty of examples where I would have done it differently.
Everyone knows that I think the web loses 10 points every time we solve
something using a
processing instruction -- but the community wants it that way and we have
http://www.w3.org/TR/xml-stylesheet/
I still think an RDF property would have been better but there we are.

> Instead of pushing your vision, try building
>a place where visions of others from far abroad can come together and
>interact so that the right vision can be born out of them.

That is just what I would love to do.  You put it so well.
But --- do you mean come together ...by plane?
or telephone?  or mailing list?  From dream to reality,  from concept to
deployment plan -- that is the difficult bit with social syetems as well as
technical ones.

> Be a mother, not a father.

If you mean, nurture, don't dictate, that sounds like sound advice,
though put in a rather gender linked way ;-)

Regards,

Tim Berners-Lee



> Best,Don Park    -   mailto:donpark at docuverse.com
>Docuverse   -   http://www.docuverse.com
>



xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)