Re: [vwrap] Status and future of the VWRAP working group

[Morgaine <[e-mail]>]

Indeed, server-side inventories are one of the worst design decisions of the
existing implementations.

They give us no end of problems, such as the impossibility of high-speed
client-side searching and regular lag issues, not to mention being
non-portable because each world provider would implement them differently.
They're also non-scalable because they're centralized, and non supportive of
VW tourism because a given world provider would control what can be put into
them, instead of the traveling user being in control.

Inventories need to be client-side (I refer to the trees of asset metadata,
not the asset data, in case that's not clear), with an auto sync of deltas
to a configured asset service for resilience/backup.  That way only those
people who do not look after their local inventory resources will suffer
client-server transport lag to recreate them, and the full power of
client-side CPUs and client-side databases and scripting can be deployed to
make inventories a real powerhouse.

All we really need are shared asset services that are independent of the
world providers.  That alone is enough to virtually ensure portability,
interoperability, and scalability, as long as a lightweight protocol can be
found that is also extensible to support tomorrow's worlds.  Making a
standard to support only what we're doing today isn't particularly
interesting.


Morgaine.






==================================

On Sun, Mar 27, 2011 at 2:41 AM, Dzonatas Sol <[e-mail]> wrote:

> If there was focus only on the login (agent domain) and inventory services,
> then I think we will get some where.
>
> Yes, we know there are issues with inventory that are already in the
> simulator side, so this is the suggestion to build apart from those issues.
> Start from there.
>
>
>
> Izzy Alanis wrote:
>
>> Are you suggesting a re-charter around shared services only?
>> Or that we shift the deliverables to push that to the forefront? (In
>> which case, we still really do need an intro doc and to address the
>> messaging semantics)
>>
>> I'm not quite convinced that the charter is unsalvageable. I think the
>> last rev of the intro has a lot of good things in it too -- lots I
>> would like to see changed, but good stuff too.
>>
>>
>>
>> On Sat, Mar 26, 2011 at 8:09 PM, Morgaine
>> <[e-mail]> wrote:
>>
>>
>>> I'm glad to see some renewed participation here!� Perhaps the threat of
>>> death sharpens the mind. ;-)
>>>
>>> Since there seems to be some fresh thinking in the air, I am going to add
>>> two short points to the discussion.� The first is a matter of procedure,
>>> and
>>> the second is related to our technical direction:
>>>
>>> Notwithstanding that the IETF places certain duties on Barry and others
>>> to
>>> ensure that there is visible progress in the form of documents, I must
>>> say
>>> that "documents at all costs" is not a particularly good way of achieving
>>> technical progress.� It's the "documents at all costs" push that gave us
>>> several documents previously, only one of which turned out to be usable
>>> for
>>> interoperation between VWs.� Documents churned out before there is
>>> agreement
>>> on goals and direction are a hindrance to the process, not an indicator
>>> of
>>> progress, and they waste everyone's time.� Progress is certainly not a
>>> matter of just putting pen to paper, as has been suggested.� Far from it.
>>> First we must agree as a group on how a given protocol is going to meet
>>> our
>>> goals, and drafts then present that formally with hard technical details
>>> added.� Done the other way around just results in much angst and wasted
>>> effort, as happened here.
>>>
>>> Given the almost unanimous agreement that crystallized around Crista's
>>> thread of a few months ago which could be paraphrased as "The VWRAP
>>> documents do nothing for interop between virtual worlds", I would like to
>>> suggest that instead of continuing to beat the dead horse of OGP that we
>>> still have on our hands, why don't we focus on delivering something that
>>> is
>>> actually usable by compatible groups like Opensim and realXtend and iED?
>>> There is a nugget of gold at the heart of the VWRAP concept which can
>>> provide exactly that:� the idea of shared asset services, and a protocol
>>> for
>>> accessing them.
>>>
>>> There is a huge amount of activity in our sector of the virtual worlds
>>> community.� There is also no end of interest in interoperation, but the
>>> trouble seems to be that each group is rather narrowly focused on their
>>> own
>>> particular code base.� Where I think a group such as ours can contribute
>>> is
>>> by providing a lightweight protocol which is easily used by all, without
>>> the
>>> previous baggage.� Simple problems demand simple solutions, and while a
>>> massively scalable shared asset service is not exactly simple, it is
>>> nevertheless a lot simpler than the much larger task that we had set
>>> ourselves previously.
>>>
>>> Perhaps that would be a good place to start, afresh.
>>>
>>>
>>> Morgaine.
>>>
>>>
>>>
>>>
>>>
>>>
>>> =====================================
>>>
>>> On Wed, Mar 23, 2011 at 8:10 PM, Barry Leiba <[e-mail]>
>>> wrote:
>>>
>>>
>>>> Reminder: If anyone's done anything related to what's below, please
>>>> post here and get some discussion going. �There's still about two and
>>>> a half weeks to get something ready.
>>>>
>>>> Barry, as chair
>>>>
>>>> On Sun, Feb 27, 2011 at 1:52 PM, Barry Leiba <[e-mail]>
>>>> wrote:
>>>>
>>>>
>>>>> On Wed, Jan 19, 2011 at 3:15 PM, Barry Leiba <[e-mail]>
>>>>> wrote:
>>>>>
>>>>>
>>>>>> As for timescales, we already started work on a new Intro in October
>>>>>>> and
>>>>>>> November, as I described in my first email in this thread.� It was
>>>>>>> being
>>>>>>> done informally, not as an official draft but as input to a totally
>>>>>>> new
>>>>>>> draft.� It was not being done as a revision because the previous
>>>>>>> Intro
>>>>>>> simply did not meet key requirements for many contributors, as was
>>>>>>> clear
>>>>>>> from the group's very intense discussions of September.
>>>>>>>
>>>>>>>
>>>>>> Can you see if you can get it into reasonable shape to introduce
>>>>>> publicly, and then submit it as draft-morgaine-vwrap-intro-00 ? �That
>>>>>> would give people something concrete to work from.
>>>>>>
>>>>>>
>>>>> I haven't seen any activity on this, so let me repeat this with a
>>>>> deadline:
>>>>>
>>>>> The chairs ask the proponents to please get a new intro document into
>>>>> reasonable (not final) shape to introduce publicly, and to submit it
>>>>> as an Internet Draft with a name like "draft-SOMEONE-vwrap-intro-00"
>>>>> by 10 April (the significance of which will be left for the reader to
>>>>> research, should s/he care to). �There may, of course, be any
>>>>> (reasonable) number of authors listed on the draft, and any one may be
>>>>> the name chosen to live in the draft name.
>>>>>
>>>>> If we're not able to do that, I think we need to seriously question
>>>>> whether there's enough real energy to continue.
>>>>>
>>>>> Barry, as chair
>>>>>
>>>>>
>>>>>
>>>> _______________________________________________
>>>> vwrap mailing list
>>>> [e-mail]
>>>> https://www.ietf.org/mailman/listinfo/vwrap
>>>>
>>>>
>>> _______________________________________________
>>> vwrap mailing list
>>> [e-mail]
>>> https://www.ietf.org/mailman/listinfo/vwrap
>>>
>>>
>>>
>>>
>> _______________________________________________
>> vwrap mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/vwrap
>>
>>
>>
>
>
> --
> --- https://twitter.com/Dzonatas_Sol ---
> Web Development, Software Engineering, Virtual Reality, Consultant
>
>