If you can represent that configuration, I'm all for it. I had left it out because from where I stand it seemed hard or impossible to support. If it;s possible, so much the better.
+1 Melanie Diva Canto wrote: > OK, I think I'm coming close to having the sharp definitions that I > think are necessary. They're slight variations of Melanie's, because > Melanie's are still coercing OSGrid-like grids to become something that > they aren't. > > - "Simulator trust domain": a set of simulators that trust each other, > operated by one single authority > > - "Grid": a set of simulators that share resources. Resources include: > assets, user accounts, and assorted services like lookup services, > forums, etc. [And this is the thing] A Grid includes one *or more* > simulator trust domains. > > - "Hypergrid": the interconnection of grids. > > In most cases (I would expect) a grid contains exactly 1 trust domain. > But in some cases, it may contain more. If you think about it, the > Linden Lab grid is a "grid" according to the above definition, and so is > OSGrid. The mechanisms by which grid operators establish sim trust > domains are entirely up to the grid operators. They may apply formal > TOSs or they may put the grid login password on the web. They may issue > formal trust certificates (one or more sets) or they may not do such a > thing. > > > For the purposes of security, it's clear that the client can't talk > directly to a central grid-wide server, but it has to do it with the > specific simulators, which then can redirect to their authorities. > _______________________________________________ > Opensim-dev mailing list > [email protected] > https://lists.berlios.de/mailman/listinfo/opensim-dev > > _______________________________________________ Opensim-dev mailing list [email protected] https://lists.berlios.de/mailman/listinfo/opensim-dev
