[sldev] AWG - Scope (Future SL Architecture versus Multiworld
zha.ewry at gmail.com
Fri Oct 26 06:33:55 PDT 2007
>From my perspective, primarily technical, but, yes, I work for IBM, and yes
interoperability is something we think is very important, the goal of the
AWG is to help Linden with it's next generation architecture. Linden kicked
off the AWG process with scaling requirements based not just on the next
year or two of growth of SecondLife, but growing the approach to world wide
web scale.The broader vision of getting to web scale, and getting a broad
approach which permits that was placed on the table from the very begining.
That said, I don't think there's a lot of technical conflict between
getting the architecture to scale, and getting it to interoperate.
The desire to allow non Linden Labs hosts to host regions and assets, is, in
fact inherently about interoperability. From the point of view of software
design, a cluster of sims hosted by a star trek roleplaying community, and a
second life like virtual world run by another vendor, pretty much look the
same. Make it possible to do the one, and you've largely made it possible to
do the other.
Enabling a cluster of interoperable virtual worlds centered around the
secondLife approach would be one step to a much broader interoperability
story. Growing beyond that, to standards which would allow very disparate
virtual worlds to connect and interoperate is a longer term, broader effort,
one which would both inform the AWG and be informed by the AWG's work.
When Zero talked about goals, and schedule, and some of these issues, he
made it very clear that Linden is going to be balancing short term needs to
evolve SecondLife and keep it growing, and stable, as well as the longer
term goals. I fully expect that at times, as a community, this will lead to
Linden pushing ahead on some features they need before there is a complete
community centered AWG story, and sometimes, the AWG is going to lay out
something that's needed by the larger community, which Linden won't
immediately implement. This may lead to some interesting short term
challenges, but long term, I don't see it as a major challenge
- Zha .
On 10/26/07, Tao Takashi <tao.takashi at googlemail.com> wrote:
> Matthew Dowd wrote:
> > > Originally, I thought the AWG was attempting to develop a new
> > architecture for SL focused on better scalability and opening up the
> > possibility of people running their own SL sims.
> > It is. When Zha talks about interoperability, he's speaking of
> > alternate implementations of the SL protocol inherently.
> > Other virtual worlds... unless they want to completely start over using
> > the protocols we make, they are irrelevant.
> Well, most of the existing VWs probably wouldn't even want their worlds to
> connect I would guess (it's about business models) or it might be difficult
> because of different models behind them (it's about technology).
> So if they start to use the protocol we define (or work with us to define
> maybe a better one) they can interconnect.
> I guess we are also more talking about VWs to come which may or may not
> use the protocol or even some of the components which might be open source.
> Still IMHo the goal should also to be able to provide as many models as
> possible not just e.g. the prim model in existance right now (same for
> agent data etc.). This does not mean that we have to define all these models
> (we actually can't anyway) but to provide some sort of extension mechanism
> and if it's just one field in some protocol which defines the model to use.
> So in the end some might change their protocol to interconnect or some new
> ones will use it as a basis once it is ready and will be able to
> (potentially) interconnect.
> (when talking about it publically to non-SL people I nevertheless talk
> about interconnecting VWs because this simply drives more people than
> talking about something SL related. Some people seem to stop listening as
> soon as SL is mentioned and of course we don't want this ;-) The details can
> then be explained later in the discussion).
> -- Tao
> taotakashi at gmail.com
> RL: Christian Scholz, cs at comlounge.net
> IRC: MrTopf/Tao_T
> Click here to unsubscribe or manage your list subscription:
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the SLDev