Re: [mmox] MMOX Progress tracking

On Thu, Mar 5, 2009 at 8:02 PM, Jon Watte <jwatte@gmail.com> wrote:

>
> The problem with OGP is that it assumes that there is only one
> object/sim/client platform stack involved in any one scenario. That's why
> it's actually counter to the goals of vendor neutral interoperability. Note
> that it also necessarily involves the client machines, and thus drives
> towards a "universal client" model, which I think is a bad idea for reasons
> I have already described.
>
>

We've addressed the issue of "universal client" many times before, and yet
you keep on bringing it up again.  Shall we examine it in detail once and
for all, so that it doesn't keep getting thrown in to dilute otherwise
logical discussions?

   - VW providers will provide whatever clients they like, and it's not up
   to us to say whether they should be one-service clients or multi-service
   clients.
   - The open-source community has the colossal benefit of unencumbered
   access to multiple libraries and middleware solutions, and frequently
   deploys them in generic clients with multi-service capability.  It is not up
   to us to forbid nor discourage that.
   - The degree to which generic clients can bloom into some theoretical
   "universal" client is severely limited by practical issues such as available
   manpower and machine limits such as memory size.  Raising the specter of
   "universal client" is therefore not even discussing reality.
   - Generic clients are certainly not "bad".  They provide the basis of
   "client-based interop", and empower their users tremendously.
   - We are creating an open *protocol*.  It's not up to us what kind of
   systems people design and implement around that protocol.
   - The web has acquainted users with the web browser, the quintessence of
   a generic client that has a natural tendency towards becoming universal.
   Whether we personally favour it or not, it is not up to us to prevent that
   happening in virtual worlds.
   - It is rather easy to speculate that, in due course, generic open-source
   VW clients will drive the metaverse towards de facto commonality and
   standardization.  Many would regard that as a good thing, and in a group
   formed for the purpose of protocol standardization, that seems a worthy goal
   to support.

Morgaine.

On Thu, Mar 5, 2009 at 8:02 PM, Jon Watte <jwatte@gmail.com> wrote:

> Jesrad wrote:
>
>> Jon, if multiple vendors agree on a specific technology/vendor in the
>> cadre of this interoperability protocol creation, would you consider
>> it to be vendor neutral or not ?
>>
>>
>
> If the protocol allows two vendors with significantly different platform
> stacks (say, OLIVE and Metaverse, or There.com and Second Life) to connect
> to each other and achieve user-visible interoperability, then yes.
>
> The problem with OGP is that it assumes that there is only one
> object/sim/client platform stack involved in any one scenario. That's why
> it's actually counter to the goals of vendor neutral interoperability. Note
> that it also necessarily involves the client machines, and thus drives
> towards a "universal client" model, which I think is a bad idea for reasons
> I have already described.
>
> Sincerely,
>
> jw
>
>
> _______________________________________________
> mmox mailing list
> mmox@ietf.org
> https://www.ietf.org/mailman/listinfo/mmox
>