Private Microsim/Instanced Interiors

I'd still like to propose the idea of private microsims, basically chunks of 3 dimensional space that is smaller then the regular islands, accessible either by "doorways" (special prims that allow you to move from the current sim to the microsim by clicking on it) or by direct teleportation if you own the microsim, or granted access to individuals/groups.

Why? This idea has been proposed and shot down tons of times by jaded people here.

1) Privacy of course, which is completely lacking in SL. Anyone claiming otherwise is just daydreaming.
2) Performance - Only your crap is what drags down simulator performance. Your stuff isn't affected by others in the same sim running too many physical prims and resource intensive scripts
3) Cell-style deployment. Private simulators can be passivated when no one is in them, and activated when someone enters them (think of EJB session beans). This way a cluster of servers that would probably support just a few regular sims could support dozens, perhaps even more, microsimulators. Load balancing can be taken advantage of by activating microsims on systems with lower loads.

I'd think that a range of sizes of microsims could be offered, from a minimum to a maximum that makes sense from a resource perspective (a 512m microsim is probably going to eat more resources then it would be worth, while a 32km might exceed reasonable demands for a sim that can be passivated/activated and at that size, competes with private island sims).

A lot can be done with this, where your stuff inside your building is not affected by other residents that you share your outdoor sim with, nor influenced by activities in neighboring sims.

This would work as an Interior region, say of a large building. No windows, cant' see in or out. You walk through the door and you are seperate and away from the regular sim.

This would allow for the Tardis effect. Buildings in SL could be small, but you walk in and you are in a huge structure. Would make for interesting possiblities. And since such neat things that aren't physically possible are supposed to be par for the course in SL, and since a single older server can handle lots of buildings, it's worth a look.

Wow, the poll isn't biased at all.

Someone obviously has no sense of humor.

[Edit following]
Besides, what would you rather see? 3 saying it sucks, with one saying it's okay? I went with 1) Very positive 2) Positive with reservations 3) Neutral 4) Very negative. How is that unbiased?

Allow me to be more obvious...

<sarcasm>Wow, the poll isn't biased at all!!! </sarcasm>

That's better!

This is already possible in a way if LL would allow anyone to own their own openspace sim.

You get 1875 prims for an entire 65536m² and the estate tools certainly allow enough room for privacy in most ways. The reduced perfomance compared to a normal sim should still be more than adequate for any residential use.

The prim usage is that of an 8192m² plot and at $1675/4 sims that comes down to about $400/sim, or a 8192 at L$13/m² which isn't that prohibitively expensive for the extra control. Additionally owning 8192m² would cost $46-50/month compared to $73/month for 1 openspace sim.

It certainly wouldn't be for everyone (I'd prefer it at $50/month ), but there might still be enough demand for it.

(Edited to add that I don't disagree with the OP's proposal, but I don't think it's realistic for LL to start from scratch for "personal" sims and openspaces are here already, it's only policy that keeps a non-sim owner from getting one)

Linden Lab plan on releasing the server code "soon". You'll then be able to host your own sim. But if you want a good connection then you'll have to rent a dedicated server which will still cost you $200 per month.

Basically, you can already get as much privacy as you like in a well deployed skybox, and the people inside taking good care of their personal settings (show on map, etc.)
Though I think making some adjustments on the current ban-tape system would work much nicer.
1) Specifying exact bounderies along X, Y AND Z axis (so it allows for free flight space for example, which might pursuade some estate holders to allow you to actually use these tickertapes)
2) An option to block camerapanning into a banned area

Overall, a parcel owner could leave part of his/her parcel public as well then, without requiring making parcel devisions as well

This is a good idea and its done in a lot of other Online worlds/games (AO and Eq2, Neocron) so obviously can be done.

Not bad feedback so far. Here we go...

2k Suisei Said:

"Linden Lab plan on releasing the server code "soon". You'll then be able to host your own sim. But if you want a good connection then you'll have to rent a dedicated server which will still cost you $200 per month."

If your needs are a whole outdoor sim, then you're better off just getting a private island. The idea here is to have a microsim that can represent the interrior of a structure, which would not have the shere demands that a region-type sim would have. The insides of most buildings are NOT the size of a region sim. A 900m sim reprenting a 30x30 grid (with a height of, say, 30m too) could emulate the inside of a structure just fine. And you would not be impacted by your neighbors running scripts that slowdown the region grid's performance down to 1 sim fps. The microsim is a sim all of its own which is not attached directly to another sim. The only attachment to a region sim would be a doorway/warp link/access node/insert your favorite "wisk me there means". Other then that, teleport would be the only other way inside. A person who wants just enough of a personal sim to model the inside of their house wouldn't need a $200/month setup. I would expect a microsim would be a far cheaper option.

And given that microsims would probably take off, the fact that they can be passivated when the last person in the microsim leaves, and activated when someone returns means that a server's resources can be allocated to service other microsims that are active.

Microsims can passivate/activate on demand. Region sims can't, because there's always a chance someone could wander in (or hell, just look in) from a neighboring sim. THerefore a region sim is always active and requires a dedicated server. Microsims can be passivated and their state stored away, so in case they're not used for a month or so, they're not eating a month's worth of server time the way a region sim would be.

Margarita Nemeth said:

"Basically, you can already get as much privacy as you like in a well deployed skybox, and the people inside taking good care of their personal settings (show on map, etc.)
Though I think making some adjustments on the current ban-tape system would work much nicer.
1) Specifying exact bounderies along X, Y AND Z axis (so it allows for free flight space for example, which might pursuade some estate holders to allow you to actually use these tickertapes)
2) An option to block camerapanning into a banned area

Overall, a parcel owner could leave part of his/her parcel public as well then, without requiring making parcel devisions as well"

Second Life does squat for privacy, and you know that. Ban tapes only go up to 100m or so, so any skybox up there is going to need death orbs to keep people out. And while it works, it is always annoying to hear "So and so got zapped off your property" when a more desirable way would be not to have intruders around at all.

I will also stress again performance benefits, not just privacy.

Colette Meiji said:

"This is a good idea and its done in a lot of other Online worlds/games (AO and Eq2, Neocron) so obviously can be done."

Thank you for the reference, Colette. If non-direct competitors are doing it, then I don't see why LL doesn't do it. It's money they're leaving at the table. Once the R&D is done, all they need to do is make it a sale option, let the customer pay their one time setup fee, and get their monthly recurring fee set up, and then automatically generate the microsim data, which would be added to the "pool of microsims". Since there is no dedicated server for the microsim, I think Linden involvement in the establishment of new microsims would be at the most a minimum. All they need to do is manage the cluster and ensure that load balancing and server stability runs smoothly.

There's other things this could do too. Imagine this: You get a microsim to simulate the inside of a submarine. The interrior could be super detailed, using maybe even a couple thousand prims. Performance in the microsim is affected only by what you run in it (assuming the server the microsim runs on is properly load-balanced -- maybe Quality of Service could be a scaled-payment option?) The gateway to the sub interror sim is linked to a simple 30 prim object submarine in the regular grid. From inside the microsim, you can control the submarine on the outside using regular vehicle controls (obviously remote object messing will be needed). The windows from the sub in the main grid provide a camera system that can be fed back into the microsim.

So you can have a vehicle that also serves as a home, that can interact with the world in very novel ways. (Of course, if your submarine in the main grid gets returned, you'd have to teleport out to the main grid and reset it, unless objects can have 'home locations' set where they'd get returned to if kicked out of someone else's parcel.)

Just an idea. Tell me that that isn't a cool idea.

BTW, everyone who votes, I'd really like to hear a post on the forum stating your thoughts and reasons for why you voted, good or bad.

It would be most interesting if there were some way of doing this on a parcel level, removing the idea of simulators entirely.
Except of course for private island owners, who would be guaranteed a whole sim. Though if they owned multiple then the same resource sharing stuff would work well, you'd have your own five-sim "grid" within which the micro-sim parcels would operate.

But eh...that's fanciful thinking. The main thought on this is that unlike above this would be working WITHIN the regular simulator system, which raises issues of how treat the two differently. Also with passing avatars around you've got complications as that's not always the nicest of processes (sometimes it takes ages or outright fails).

For privacy I'm not 100% sure as a result, there may be simpler options that allow the same or greater flexibility. But for performance and such it is certainly a much better concept, especially if it could be applied well. As currently we have a ton of simulators which may only have one or two (if any) people in them while others may be labouring to handle loads. We could have half the main-grid unutilised that could be doing more work for the parts that need it.

Well how about hosting a private sim on your home computer?. Obviously the connection wouldn't be good enough to have lots of guests around. But it might be okay for just you and maybe the odd friend.

1) Terms of Service, many ISPs, including cable companies, forbid running servers off their network. You risk losing your internet connectivity if you get caught.

2) Would LL allow external sims to use their asset servers? What kind of risks and exposures would allowing that have?

3) Sometimes it's nice just to pay someone to handle the responsibility of keeping the sim running and the machine up to date and what not.

Okay, let me think...

Well how about going to live in a little house in 3DS max?. The sky is a little gray. But other than that, it's very private.

Don't have much to say besides that it is a good idea. Privacy is always good and there should be an option for more privacy than is currently available. I think it would sell and LL wouldn't regret it.

But..but...it's not Second Life!

I see a lot of votes, particularly negative votes, with no reasons specified. Talk, people. Tell me why this is such a horrid idea.

-- topic reported (biased poll) --

If it's so biased Kyrah then why have so many (comparatively) voted against? Besides, I read the last option being several options grouped together, some of which were tongue-in-cheek. ie; "It's a bad idea" or "Privacy is bad" or "Performance is bad" etc.

Interesting - Since Polls dont actually mean anything around here, why does it matter if they are biased or not?

7 votes against, can't say I'm not disappointed to see so many against it, but I am disappointed to see votes without anyone putting out a real good reason why it's such a bad idea. Comments people, just saying "No" doesn't have a lot of value-add unless you give a reason. That's why I had voter names assigned to each choice -- accountability. So that you can post your reason for voting as you did (positive or negative) for discussion. This is a forum afterall.

As for the forum being biased, how? And how is that a reportable event? Cite the forum rules that I supposedly violated with the poll. I felt that the poll entries were reasonable:

1) Positive, I like the idea very well as it stands

2) Positive with reservations, I like the idea, but only with certain changes to the model (discussed in forum)

3) Neutral, I'd rather see other features go in before this one. Bugfixes don't count because every single one of us (me, included, and notice I haven't placed a vote) would vote for this one if that was the case.

4) Negative, I hate the idea and this i why (discussed in forum).

Reviewed the forum guidelines, and polls aren't even mentioned in it. Nothing has been said that warrants a moderator to review this thread.

And yes, I'll admit a little sarcasm with the last poll entry, I've proposed this before with vehement opposition with very little few reasons (good or bad) for the opposing opinions. That's why I'm trying to get people to explain their positions, and be held accountable for their poll decision. If people think this is a really bad idea, then please try to convince me. Just stamping "No" doesn't get anything done.

/13/fc/181052/1.html#post1491397

Privacy is possible today, microsims are possible today (if you don't want 65536m² of land, then just make a 4096m² island and leave the rest as water), only policy keeps it out of reach from anyone without a private sim.

It would also effectively split the sim team in half because you're proposing variable sized sims with a seemingly infinite prim count and minimal cost on top of that, which is nothing like the current sim model so they'd have to start back at square one.