Re: [vwrap] Call for a vote on interop BETWEEN independent virtual worlds or not

[Cristina Videira Lopes <[e-mail]>]

That sounds like a good idea, Jonathan, but keep in mind that the 
foundation document suffers from the same problem of the intro document.

I agree that part of the issue is terminology. But part isn't. There's 
the deeper issue on whether VW interop needs anything more than 
client-server architectures powered by HTTP and [mostly] stateless thin 
clients, and then some additional protocols that work on, and with, this 
architecture. The documents leave no doubt that VW interop needs 
something *else* entirely, a new model and a VWRAP protocol suite, that 
is different, and in par, with HTTP(S); one where the client happens to 
play a much more active role than web browsers usually do.

I happen to disagree with that premise, but that's because one of my 
drives for interop is to stay as close to the Web as possible, and that 
includes trying to make the most of web browser clients, which have 
really interesting properties that foster diversity (and which the rest 
of the world uses).

But I understand that the world of technology doesn't stop with 
client-server and HTTP, and that lots of other distributed systems can 
be designed that are quite different from the Web. It seems that the 
charter of this group is to engage in the design of one such system. 
Which is both ambitious and slightly orthogonal to the issue of VW 
interop [on the Web].

WRT how to proceed. First, there should be one intro document only, not 
two. At the very least, if breaking away from the Web is what this group 
wants, that should be made clear somewhere in the intro document. If 
not, that intro document needs to commit a lot less than what the two 
documents do now. There's things in there that simply will not happen on 
the Web, period -- they won't happen in Unity-based web applications, 
and they won't happen in lots of other web-based VWs. They will only 
happen in a class of VWs that uses some version of the Linden Lab viewer 
as client. Maybe.


Jonathan Freedman wrote:
> Hello everyone,
>
> From what I can tell the drafts do support interoperability between 
> the same *class* of virtual world. The catch is that the language 
> needs to be significantly clearer. In fact, I would go so far as to 
> say that the introduction document should drop all references to use 
> cases, focus on describing the context (sandbox virtual worlds) and 
> then be merged with the foundation document.
>
> I have approached editing the Intro document several times and it 
> always ended in despair. I believe the only rational way to move 
> forward is to integrate it into the Foundation document as a simple 
> overview of the context and proceed from there. If others support this 
> direction, I would be honoured to proceed with the first cut of such a 
> merge.
>
> I do want to state that interoperability between the *same class* of 
> virtual worlds seems like the proper end goal. The language needs to 
> be unambiguous and there is no point in distracting the consumer (of 
> the ID)  with discussions of use cases. Use cases, deployment 
> strategies and the like should be left up to interpretation. The 
> formal documentation needs to focus on the protocol rather than 
> implementation.
>
> Regards,
>
> Jonathan Freedman
>
>
> On Mon, Sep 20, 2010 at 11:20 AM, Morgaine 
> <[e-mail] 
> <mailto:[e-mail]>> wrote:
>
>     On Mon, Sep 20, 2010 at 2:51 PM, Mike Dickson <[e-mail]
>     <mailto:[e-mail]>> wrote:
>
>     Put another way we're not specifying a mechanism for
>     interconnection between very different technologies (or more
>     appropriately approaches to virtual worlds).
>
>
>     Unfortunately, no Mike, it's much worse than that.  Even if the
>     technologies of the worlds in question are not only compatible but
>     *IDENTICAL*, Meadhbh claims that we are not creating a protocol
>     for interop BETWEEN those worlds.  At all, whatsoever.
>
>     This cannot be allowed to stand, otherwise the entire purpose of
>     VWRAP as an interop protocol disappears, and instead VWRAP becomes
>     a protocol for building standalone, isolated worlds.  That is not
>     what we're here for, and it has never been --- we have affirmed
>     the goal of interoperation between VWs time and again on this
>     list, repeatedly.
>
>     This issue needs to be cleared up without ambiguity.  We can't
>     have a prolific draft writer writing drafts that do not reflect
>     the goals voiced by almost everyone in this group since OGPX/VWRAP
>     began.  Crista's post is merely the latest expression of concern
>     of many.
>
>     There is a clear disconnect here between the goal of
>     non-interoperating worlds, and the much more useful goal of VW
>     interoperation that virtually everyone else has been discussing
>     and desiring.  We already have non-interoperating worlds, lots of
>     them!  Note that even Joshua mentions interoperation of VWs in his
>     latest post a few weeks ago, in which he welcomed discussion of
>     "/protocols for data transport between virtual world instances/"
>     -- http://www.ietf.org/mail-archive/web/vwrap/current/msg00253.html .
>
>     This needs resolving formally, otherwise our progress on resolving
>     the issues of VW interop is completely blocked.
>
>
>     Morgaine.
>
>
>
>
>
>     ================================
>
>     On Mon, Sep 20, 2010 at 2:51 PM, Mike Dickson <[e-mail]
>     <mailto:[e-mail]>> wrote:
>
>         On 09/19/2010 10:41 PM, Morgaine wrote:
>>         On Sun, Sep 19, 2010 at 5:09 AM, Meadhbh Hamrick
>>         <[e-mail] <mailto:[e-mail]>> wrote:
>>
>>         secondly, VWRAP is not now, nor has it ever been a protocol
>>         to enable
>>         interoperability BETWEEN virtual worlds.
>>         ...
>>         in short, the consensus of this group has generally been to
>>         describe
>>         the mechanisms one could use to build a single virtual world
>>         but does
>>         not dictate that this world be a singleton.
>>
>>
>>         This does not reflect any consensus expressed in this group
>>         whatsoever.
>         I suspect we're getting wrapped around the axle on terminology
>         and what "single virtual world" means.  At least I'd like to
>         interpret it that way as it then matches the discussion over
>         the past months.  Put another way we're not specifying a
>         mechanism for interconnection between very different
>         technologies (or more appropriately approaches to virtual
>         worlds).  It's a single virtual world because it shares a
>         single set of assumptions about how the services that make it
>         up work together to provide services.  If I change in a
>         significant way a service that doesn't match what VWRAP
>         documents then I'm not able to participate in the VWRAP
>         virtual world any longer.
>
>         The comment about a singleton is on target I think with this
>         interpretation.  I can create a walled garden that doesn't
>         interconnect with other "services".  It's using VWRAP and so a
>         part of the VWRAP "Virtual World". 
>
>         If thats not a correct interpretation then yes we have a huge
>         issue.  If it is correct then perhaps we need to refine how we
>         define terms since its caused alot of confusion.
>
>         Mike
>
>
>
>         _______________________________________________
>         vwrap mailing list
>         [e-mail] <mailto:[e-mail]>
>         https://www.ietf.org/mailman/listinfo/vwrap
>
>
>
>     _______________________________________________
>     vwrap mailing list
>     [e-mail] <mailto:[e-mail]>
>     https://www.ietf.org/mailman/listinfo/vwrap
>
>
>
>
> -- 
> Jonathan Freedman
> President
> Open Metaverse Foundation
> [phone]
> ------------------------------------------------------------------------
>
> _______________________________________________
> vwrap mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/vwrap
>