Hello and welcome to 協力
Ryan McDougall
discussions at this year's Metaverse U conference at Stanford, hosted
by the Sirikata folks. There was a feeling that the time was right for
small groups working on independent VW technologies to share
interfaces or even code in areas that are not core to our independent
foci.
We all have our own horses to ride, but there is a lot of overlap
where there are well-known or common solutions; or where we have no
stake, and thus can accept an arbitrary solution for the sake of
momentum and interoperability.
== Ground rules
The pitfall for these kinds of discussions is always that we may
wander off topic, drift in to the atmosphere and lose scope, degrade
into partisanism, or waste effort bikeshedding[1]. In order to guard
against these possibilities, it was felt that the following ground
rules should be held:
- the group consists only of "pigs"[2] -- people actively working to
implement interoperability within a VW system
- the group is like-minded in their belief in openness,
interoperability, cooperation, and collaboration
- the group has dedicated some amount of resources to making a larger
collaboration work through this specific venue
- members should share what what they have, take what they want, but
never to try to argue down another's core principles
- bikeshedding is strictly forbidden -- we are all very busy people,
and must keep the signal-noise ration high, otherwise contributors
will tune out
And the final, but not least requirement,
* all related patents -- or any other conflict of interest -- *must*
be disclosed when making proposals.
== Collaboration
The operation of this group should happen on two levels:
1. RFC-style discussions of potential shared interfaces, whether
expressed in code or protocols
2. collaborative discussions on refactoring code into shareable,
reusable libraries or APIs
== On Openness
While so far the group has been invitation only, it is by no means
closed to anyone who can accept the above constraints. The reason for
the invitation-based nature of the list is to keep the group small,
reactive, and consisting of decision-makers; keep a handle on noise
levels; and prevent a repeat of the MMOX debacle. We are trying to
give this experiment the best chance to succeed, and that chance is
improved if we can first prove to each other that we work together on
the above two low-level points.
== Directions
If we find that we can have common ground, and make common cause, then
the goal is to elevate this working group to something higher, like a
standards body like the IETF, or perhaps even a common open source
Foundation.
While I think membership is visible to all of you, I would like to
recommend that since there's only now 12 of us, we step forward with a
brief introduction and statement of goals.
I will also be following this up with some suggestions for starting
directions, to help get the ice broken.
Sincerely,
[1] http://en.wikipedia.org/wiki/Color_of_the_bikesheda
[2] http://en.wikipedia.org/wiki/The_Chicken_and_the_Pig
== Postscript on who is Here and who is Not
* If you think someone should be here who is not, let me or Ewen know,
and we can add them.
realXtend
- program manager Antti Ilomaki
- lead Architect Ryan McDougall
Sirikata
- manager Henrik Bennetsen
- developers Ewen Cheslack-Postava, Daniel Reiter-Horn, Patrick
Reiter-Horn, Daniel Miller
Intel/Cable Beach/LibOpenMetaverse
- lead developer John Hurliman
MXP (also OpenSim work)
- lead developer Tommi Laukkanen
VastPark
- manager Bruce Joy
- lead developer Craig Presti
IntensityEngine
- lead developer Alon Zakai
Open Invitations
- lead architect Paul Byrne of Sun/Wonderland
- CTO Jon Watte of Forterra
Invited but declined
- 3Di
- OpenSim
(wrt OpenSim: I think they are not very united amongst themselves, and
attempting to include OS at this stage might be premature and break
the bikeshedding/signal-noise constraints. They are however the
largest and most active community, so cooperation with them is very
important. Just perhaps not at this time in this venue.)