>> clear
and published process for community contribution that's advertised at every
opportunity
I agree 100%. We can definitely improve here.
In fact this has been a recent topic at the Board - where we try
to understand what has been the default process in the past - and what
we should do going forward. The fact that Board struggles to find
clarity itself shows that there is *indeed* room for improvement :)
Here is my perspective on this - I think we have oscillated
from one extreme to another - with the ideal somewhere in the middle..
In the past we have tried pretty open input on this list - but you
are right - you can't drive consensus and a coherent design from many email
threads. So - the pendulum swung back past the ideal to an editors
list - which is *too* small. We are rather confused however as draft
specs get thrown over the wall to the public - exposing us to IP risks
and often with no clarity as to the big picture.
One key issue is that we are doing more than open
source - we are creating specifications - and that adds the need for more
discussion about the refinement of requirements to specifications - and
we have to protect the specification from taking on contributions that
are IP encumbered.
That's why the current proposal is to have two groups: a) an internal list - open to ALL members - which
conducts ALL spec-related business. Members are signed up to the
IP framework - which protects the spec b) an external list - open to everyone - which discusses
deployment and implementation issues after the spec is publicly released
But - right now - many people wanting to provide input
are not members - so this kinda breaks down. Note - it is not that
the Consortium is trying to encourage individuals to become members for
the money - its costs us more than the $100 to process a membership throughout
the year - it is that all members are committed to the IP framework is
the key.
If we can't persuade those who care to become members
then we could implement a regular review cycle for those willing to sign
a reviewers agreement for IP protection of the spec. This is nowhere
near as good as having people's ongoing active involvement however - and
undermines the need for companies to join the Consortium - which means
we lack the membership dues to do anything very cohesive for outreach and
education.
>> Efforts seem to be scattered over web3d.org,
xj3d.org, yumatech, www-vrml and various x3d mailing lists.
I agree - some rationalization would be good. web3d.org
should be the place at which you can find everything you need. We
have established the X3D forums for that purpose - and I hope it will be
a useful resource for building some community history as you suggest.
So - I guess my "ideal" path forward would
be:
- we continue this dialog - and if necessary modify
the upcoming 3.0 Participation Framework - so the community members that
wish to participate in defining the future of Consortium standards feel
comfortable joining the Consortium - so they come under the IP umbrella.
- we conduct an all-member X3D strategy session -
where we improve and refine the process, focus and milestones for X3D development
- we work together to make it happen :)
Brian - how close is this ideal to actually happening
in your view? Be honest! :)
Neil
Brian Hay <bhay@construct3d.com>
10/02/2004 06:02 PM
To
Neil TREVETT <Neil.Trevett@3dlabs.com>
cc
miriam@werple.net.au, www-vrml@web3d.org
Subject
Re: [www-vrml] Tim BL Is
Baffled, Too
Neil TREVETT wrote:
> >> I wrote a wishlist of features I'd like to see in VRML
many years
> ago. Most of the list still stands -- only a few have been implemented
> in current X3D.
>
> In the end - decisions get made by those that show up. It pains
me that
> X3D is not getting the benefit of your insights and experience.
I feel there are two equally valid points here, which contribute to the
sense of community disenfranchisement in some.
1) As Neil implied, I don't feel we (non consortium members) have the
right to complain after the fact if we don't make the effort to become
DIRECTLY, ACTIVELY and REGULARLY involved in published consortium
working group processes.
2) As Miriam states however, long-standing community participants
(consortium members or not) HAVE provided input and feedback through
historically valid channels (www-vrml mailing list) only to have that
input seemingly disregarded.
From the core spec developers point of view it must be a nightmare to
try to review and include all input from a mailing list.
But obviously there must be some well-defined process for development of
such a comprehensive spec but that process needs to be open, transparent
and inclusive.
Maybe all that's required is a clear and published process for community
contibution that's advertised at every opportunity (e.g. web3d.org home
page, all mailing list correspondence, regular reminders on different
forums). People need to feel this is a community effort.
Part of the problem is that there's no one central place we can go to
look in on debate, issues and contibute ideas and code. Efforts seem to
be scattered over web3d.org, xj3d.org, yumatech, www-vrml and various
x3d mailing lists. This doesn't make it easy for new community members.
Contrast this with many other open-source efforts. For example, the PHP
Drupal CMS at http://www.drupal.org. Here you have WEB access to the CVS
(concurrent versioning system) so you don't need to be a CVS dweeb to
see what's going on. CVS also provides a history and rollback capability
for developers. Also there's a web-based project issue tracker (trouble
ticket system) where anyone can report bugs, suggest features, take on
a
development task etc and the rest of the community can provide comment.
No code gets committed to Drupal until it has been reviewed.
Drupal is typical of many open-source efforts. There's no reason why
such a system can't be used for an ISO specification also. The
technology is free, robust and relatively easy to setup.
Such measures I feel would greatly enhance community participation,
transparency and accountability, plus provide a "living history"
of
X3D's evolution for future community members. As it stands right now,
newbies see this monumental spec but (apart from mailing list archives)
there's precious little information on how it came to be.