Re: [vwrap] Relationship between Avatar/Agent and Asset Service?

[Meadhbh Hamrick <[e-mail]>]

Hey Izzy,

FWIW... despite what morgaine says, the original members of this group
were actually interested in "service level interop." and here's the
use case.

part of the reason Linden was interested in open protocols was so we
could attempt to become the center of a much larger virtual ecosystem.
the vision we shared with some of our partners was a world where you
could log into Second Life or OpenSIm instance with your facebook or
twitter credentials. or google. or yahoo! or OSGrid or Linden / Second
Life. we also wanted to allow individual organizations (either your
corporate IT department or a commercial online service like the
Wikimedia commons) to act as an asset server for your inventory. we
wanted to give individual grid deployers (or even individual users)
the ability to pick who routed their voice and audio packets.

these are just a few examples of the use cases we wanted to support
that we felt justified "service level interop."

people who wanted to be able to walk from one world to the other,
could still do that, but only if both virtual worlds supported enough
services to move avatars between them.

so "virtual worlds level interop" is less interesting, IMHO, since it
requires participants to effectively implement all services if they
want to participate in the virtual world.

now... Linden, IBM and Intel seem to have de-prioritized their
participation in this group, so it's certainly a valid thing to say
"hey, let's focus on full-on VW interop." but ultimately, this will be
limiting your participation to a much smaller community of deployers.
and at the end of the day, you're going to have to specify services
as... well... services, so the VW interop thing has always seemed like
a red herring at worst and an artificial constraint at best.

the current charter was developed at a time when "service level
interop" was the assumption, which is why we split off each service
into a different document. the charter was also too small a document
to capture the assumptions of the types of systems we wanted to work
with, which is why in 2009 Dave Crocker recommended we put a bare
minimum in the charter with the understanding that the intro doc would
expand that understanding.

so... if peeps here want to continue with service level interop,
that's fine. we _may_ be able to keep the same charter and even some
parts of the intro doc. if we want to go with full-on interop between
OpenSim instances, that will probably necessitate a change in the
charter and definitely a change in the intro.

just my 2 cents.

-cheers
-meadhbh
--
meadhbh hamrick * it's pronounced "maeve"
@OhMeadhbh * http://meadhbh.org/ * [e-mail]



On Tue, Mar 29, 2011 at 7:58 AM, Izzy Alanis <[e-mail]> wrote:
> Starting new thread as the previous one is becoming too unwieldily.
>
> Morgaine said:
>> We either have interoperability between worlds, in which an inhabitant can travel from one world to another and take their avatar and/or possessions with them, or else we don't have that.  It's black and white, and no amount of fudging about "service level interoperability" is going to overcome the lack of VW interoperability as a user would understand it.
>
> The world is full of shades of grey.
>
> Is there an assumption that an Avatar has/uses one (and only one?)
> "Asset Service" -- that I have a single repository of possessions to
> take? Can I have and use different asset services? Can I use multiple
> asset services? Can people share?
>
>  - Izzy
> _______________________________________________
> vwrap mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/vwrap
>