Re: [ogpx] revised draft charter for OGPX working group

@dave crocker . mmm... good feedback, but i think a number of us
thought this was such a large design space that it was appropriate to
move some details out of the charter and into the "intro and
requirements" doc.

@kajikawa . right, the role of the agent domain and the region domain
are crucial to the protocol. and the LLIDL resource descriptions +
documentation of their semantics are supposed to define service
interfaces. strictly speaking, we're not developing a software
architecture here, but processing expectations _are_ in scope, so
understanding what the software looks like would be a "good idea"
(tm).

@david . right . it CRITICAL that we not create a deployment design
space that is so large that we find ourselves working through use
cases 10 years from now. the linden deployment pattern is well
documented. OpenSIm has a couple deployment patterns: standalone,
grid, and probably both with 3rd party cable beach service. we should
probably also chat with ryan @ realXtend to see if realXtend and 3di
deployment patterns are different from OpenSim.

-cheers

On Sat, Jul 18, 2009 at 7:42 AM, David W Levine<dwl@us.ibm.com> wrote:
>
> If you look at:
> http://www.ietf.org/id/draft-levine-ogp-layering-00.txt
>
> It begins to address this at the requirements and design pattern level. I'd
> strongly argue, that if its not clear that we want to permit the protocols
> to work properly in *all* of those deployments patterns, we need to update
> the charter to make that clear.
>
> My personal perspective is that the OGPX specifications are going to end up
> defining several hundred "REST" and "RESTish" resources/services, and that
> grouping and we need to describe how to group and deploy them as deployers
> see fit. There re some natural clusterings which will be very hard to
> decouple, but, beyond those, the specifications should strive to give
> deployers and implementers as much flexibility as possible.
>
> -  David W. Levine
> ~ Zha Ewry, In Second Life
>
>
>
>
> Kajikawa Jeremy <belxjander@gmail.com>
> Sent by: ogpx-bounces@ietf.org
>
> 07/18/2009 05:22 AM
>
> To
> ogpx@ietf.org
> cc
> Subject
> Re: [ogpx] revised draft charter for OGPX working group
>
>
>
>
> Well a sketch I recently worked up has
>  AgentDomain and RegionDomain
>
> these can be instanced...
>
>  but both need "SI" or Service Interfaces
>  basically abstracting U G A I M into classes for re-use
>  and standardizing the protocol chatter
>
> On Sat, 2009-07-18 at 11:57 +0100, Alexey Melnikov wrote:
>> Hi Dave,
>>
>> Dave CROCKER wrote:
>>
>> > Also:  A topic that's been discussed frequently is the difference
>> > between having a client able to access multiple servers, versus having
>> > independent servers directly interact.  From the draft charter, I
>> > cannot tell which of these will be covered or how.
>>
>> I think this is an important question. There are 3 groups of entities
>> involved in OGP and the charter doesn't make this clear.
>>
>> _______________________________________________
>> ogpx mailing list
>> ogpx@ietf.org
>> https://www.ietf.org/mailman/listinfo/ogpx
>
> _______________________________________________
> ogpx mailing list
> ogpx@ietf.org
> https://www.ietf.org/mailman/listinfo/ogpx
>
>
> _______________________________________________
> ogpx mailing list
> ogpx@ietf.org
> https://www.ietf.org/mailman/listinfo/ogpx
>
>