- From: Thomas Wrobel <darkflame@gmail.com>
- Date: Mon, 15 Nov 2010 17:33:43 +0100
- To: cperey@perey.com
- Cc: jacques lemordant <jacques.lemordant@inria.fr>, "Public POI @ W3C" <public-poiwg@w3.org>
Also, regarding timing, my 2 cents on that is we certainly need at least 2 specifications of time update; * For the linked data changing. (ie, 3d mesh last updated time) * For the updates of the trigger/content connection specification. (ie, the 3d model was moved, but the mesh stayed the same, thus the client knows to check and reposition, but doesn't have to redownload the whole thing again). On 15 November 2010 17:29, Thomas Wrobel <darkflame@gmail.com> wrote: > +1 for beyond the scope, but as someone working in > collaborative/realtime AR, it does influence some aspects of the > flexibility of the design; > For example, it reaffirms that we shouldn't focus on any one > distribution protocol, as "single document" models like html are > unsuitable for simultaneous editing by multiple user's, and not too > convient for real-time manipulation either. > > This would also, for example, influence the choice (discused earlier) > of having objects positioned relative to other objects optionaly by > ID, rather then requiring nested data structures. > > Also, real-time AR works best with smaller nuggets of information, so > you'd have individual objects describing and positioned separately, > rather then a single file describing a large scene. > > So while I agree no specific use case should be designed or is within > the scope, it should be at least kept somewhat in mind :) > > -Thomas > > On 15 November 2010 16:39, Christine Perey <cperey@perey.com> wrote: >> Hi Jacques, >> >> Yes, as long as it is for testing the data model. I got distracted by the >> systems/tools :-) >> >> I see your point! >> >> Thank you, >> >> Christine >> >> Spime Wrangler >> >> cperey@perey.com >> mobile (until Dec 7) +86 132 6171 6195 >> VoIP (rings in Beijing) +1 (617) 848-8159 >> Skype Christine_Perey >> >> On 11/15/2010 4:35 PM, jacques lemordant wrote: >>> >>> Hi Christine, >>>> >>>> I think that collaborative AR use cases are interesting, but unless it is >>>> revised, are beyond the scope of the POI WG charter. >>> >>> I read carefully the scope of the POI WG charter and could not understand >>> while collaborative AR will be not allowed as a USE case? >>> >>> Collaborative AR wil help us to correctly define time attributes for a POI >>> data model. >>> >>> It implies that the POI data model to have some attributes to specify the >>> frequency at which pooling must be done by the AR browser >>> (because a remote expert is changing the POIs data at some known >>> frequency) >>> >>> As we are concerned only by the POIs data model and not the tools for >>> collaborative AR, this USE Case will not drive the POI WG off the tracks >>> >>> Commercial AR advertising is very similar to collaborative AR with a >>> marketing guy changing the POIs data in realtime >>> >>> They are both an example of Realtime AR .... >>> >>> jacques >>> >>> >>> >>> >> >> >
Received on Monday, 15 November 2010 16:34:15 UTC