Instanced Insides/Pocket Dimensions

I haven't proposed this in the voting machine, as I haven't done a thurough look to see if something like this hasn't been requested before, but I'd like to see a feature which allows us to have "instanced insides/maps/pocket dimensions", where we can basically get our own micro-sim, which could be as small as a 512m area or as large as LL feels is feasibly possible with current technology constraints. It would be something that would be paid for separately like an island cost.

Why? There would be a number of benefits:

1) Privacy. I don't think anyone would argue that the current environment does not provide even a shred of privacy, even with the access/ban tools that land provides, due to game client limitations. I know plenty of people who buy private islands just for this purpose -- and wind up having to rent parcels out to friends just because of the cost and the fact that the resources offered by an island is a bit of an overkill for an individual.

2) Performance. Everyone who owns land on a particular sim shares the same computational resource available to that sim. Scripts, prims, etc. There is no mechanism that I am aware of that fairly distributes computational resources by % of land owned by the sim. So a person who owns 80% of a sim could experience poor script performance because of a person owning a 16m plot running several thousand scripts on that plot (extreme case, just to illustrate that fact.) Graphical demands of surrounding locations, including neighboring sims, are also a consideration that affects game performance and framerate.

By allowing people to instance the insides of their properties, it removes their activities from the mainland/private island sim, restricting it to their microsim. Performance in the microsim is not affected by events in the regular sim, and, importantly, the regular mainland/private island is not affected by events in the microsim. This should be a big win to improving the game experience overall.

This would also have the side-benefit of reducing the need for people to buy private islands, a significant hardware investment cost even if as many as 4 can be put on a single system. I would think that if properly implemented, multiple microsims (whose performance requirements could be purchased by expected demands by calculating space, max # of occupants, max # of prims, and max # of scripts/instruction execution rate in the monthly fee that would be billed to the customer by Linden Labs - these could be options selected by the customer just like ordering which size of french fries you want with your meal) should be executable on a single server. In fact, it would have the added benefit that if a microsim is left unoccupied for a period of time, it could be suspended, freeing up CPU and memory resources on the server. Regular sims probably do not do that since there's the risk of anyone in a neighboring sim crossing over at any time, and the load/startup time would not be insignificant. A microsim would probably be quicker to awaken too, certainly not any different then other MMORPGs that used instanced maps.

Will it make owning land outside of the microsim obsolete? No. Not if it is a requirement for a "doorway" to exist that links the microsim to the mainland/private island. By keeping in the spirit of using microsims as a "building interrior", requiring access points to microsims to be on either land owned by the microsim owner, or owned by a group the microsim owner belngs to (and would have permission to deploy such objects on group land, which would be another group privilage option), the need to own at least some land is not eliminated.

How will it affect the land market? Hard to say. While would be possible to have an object like a mushroom be an access point to a microsim, many may find it desirable to build/deploy a house structure, whose doors provide access to the microsim. It would eliminate the prim count required for the building contents, but not the building itself, landscaping, and outdoor objects (such as pools, outdoor furniture, etc.) It would increase the significance of small plots. A person could then do well with a 512 or a 1024m plot, instead of needing a huge 8000m plot if they want to have more then simple furniture in the house. Demand for land would relax, bringing prices back down to more realistic levels.

Other features could become available down the line, such as cameras that would make windows inside of the microsim functional, etc.

As the major performance limit in SL is not the (processor distributed) sims but the central asset server, such portals would not improve most performance difficulties.

If you are using a portal to get to a space where you control all the sim resources, LL would probably tell you to buy an island where you do control them.

Instances from other MMOs are more a way of partioning players than a radical shift in virtual geography. Put another way, each time you cross a sim boundary, you have in fact entered a portal to the next sim.

I wouldn't expect this feature to ever get into the top 100 of LL's plans (such as they are).

If they would implement llTeleportAgent this would be a more practical way to simulate "insides" of things.

This idea looks a lot like the Parcel Basement idea.