Re: [ogpx] OGPX Charter+Intro ambiguity in Virtual World vs Virtual Worlds

[Morgaine <[e-mail]>]

On Sat, Jul 25, 2009 at 3:13 PM, David W Levine <[e-mail]> wrote:

>
> Since Linden runs a multi-thousand server deployment, one of the earliest
> design points was the ability to move stepwise from the current legacy model
> to the next generation architecture. OpenSim is structured as a base
> platform with a set of extensions, such that one can create multiple diverse
> systems from the core parts.



David, that's an interesting observation, and very pertinent to the issue I
raised about multiple interoperating virtual worlds versus a single world.
There is more to this than just a choice of phrase, and Opensim provides a
good example.

Opensim-based grids and Hypergrid-connected Opensim worlds have the ability
to run a multiplicity of interop protocols, with OGP and SL-legacy being
just a subset.  All such interworking code is being progressively refactored
into alternative modules that sit side by side in Opensim, with new modules
expected to appear in the future, exactly as you say.  It is a modular and
extensible framework by design.

What this means for the topology and character of future Opensim-based
worlds is that we should expect much diversity.  In particular, an Opensim
world that communicates with SL through OGP may be
*simultaneously*connected to some other world via MXP, Hypergrid, or
some other protocol.
This diversity in protocols is very likely to result in world features in
Opensim worlds that are quite foreign to SL for example, as has already
happened with realXtend.

>From this it seems a safe conclusion that two virtual worlds that employ OGP
for interop between them are often going to be very different to each
other.  The only commonality necessary is that they can both speak the same
language of OGP, which after all is just a communications protocol and not a
world specification.  To speak of commonality of worlds is just not
appropriate in this context, since it would be purely coincidental.  We
don't have one world in any real sense of the phrase -- what we have is one
protocol, and multiple worlds will interoperate by speaking it, no matter
how else they differ.

Many factors encourage and *require* a model of multiple separate but
interoperating worlds:


   - product differentiation as a part of business,
   - strong separation of worlds as a part of security,
   - world diversity as a part of normal cultural variety,
   - worlds uniqueness stemming from human creativity and sheer
   eccentricity,
   - sovereignty of worlds as a part of politics, and relatedly,
   - the essential demarcation of legal jurisdictions.


Given the above, it is quite wrong that the concept of multiple
interoperating virtual worlds has been dropped in favour of conjuring up a
fictitious "single world" spanning the endpoints.  This is particularly so
when that cute phrase serves no useful purpose in the protocol and
confusingly conjures up a protocol entity that does not actually exist.

Indeed, it is a dangerous turn of phrase, and it is likely to result in
legal issues when differing legal requirements at two communicating OGP
endpoints find themselves applying to the "one world" of the protocol.  I
believe that the "one world" language is an unnecessary complication with no
benefit to protocol users nor to understanding, and should be omitted.

In answer to my earlier post on this subject, "one world" has been portrayed
by some as just a literary element with no strong or binding meaning.  Well
if that is so then the case for removal from the drafts is strengthened even
further, since an Internet standard does not benefit from superfluous
literary adornments.


Morgaine.









On Sat, Jul 25, 2009 at 3:13 PM, David W Levine <[e-mail]> wrote:

>
> >
> > > i think i can speak for John Hurliman and David Lavine when i say, "we
> > > are all about incremental enhancement."
> >
> > Did you start out being all about it, or did you come to it...
> incrementally?
> >
> >
> > d/
> >
> > --
> >
> >    Dave Crocker
> >    Brandenburg InternetWorking
> >    bbiw.net
> >
> >
> The epigenesis of the OGPX work was Linden Lab's desire to re-structure
> their internal architecture to allow a scalable architecture. They realized
> that what they were going to do would give them a chance to actually define
> the interoperability points at the same time. Since Linden runs a
> multi-thousand server deployment, one of the earliest design points was the
> ability to move stepwise from the current legacy model to the next
> generation architecture. OpenSim is structured as a base platform with a set
> of extensions, such that one can create multiple diverse systems from the
> core parts. Factored interfaces, extensability and incremental change is
> pretty deep in the DNA of the work.
>
> - David W. Levine
> ~ Zha Ewry (In Second Life)
>
> _______________________________________________
> ogpx mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/ogpx
>
>