Definitions and Spaces

Khamon was just blogging out loud. There's no rule against that as long as it's related to SL. I do understand LL's focus on building a product that will sell in the hope that the project might, one day, become profitable. The surprise in my post is expressed at people's ability to adapt to some incarnations of magic, but not others, and how powerful a trump that idea is in the hands of developer.

We can fly, fly, not walk but FLY! Yet there was a very large contingient of residents that did not want point-to-point teleportation reimplemented. We'll take 1/4 cup of magic; don't ask us to swallow 1/2 a cup. That's an old example, but a good one. I still hear people preaching that the world is damaged, no new groups can form, no new businesses can establish, because we don't have a centralized transportation system. And by Heaven they'll fight to Hell and back to keep estate owners from having the ability to key unused grid coords into the tools to place their island anywhere on the map. That's getting dangerously close to choking down 3/4 of a cup.

Now, in all fairness, I understand that their points are usually based on existing business applications that will have to be adapted to the new order. Change is the enemy right? That makes it all the more important that the people who are implementing virtual world software, future metaversal frameworks, consider what they are "preaching" as the norm.

You can fly. Get used to it. It's how this world works. It's not the real world.

Okay, it's not very realistic and not what I'm used to, but I'll fly. Can I log two instances of my character into respective clients and have them performing different tasks.

Well no son, that would be unrealistic and confusing.

Too late, we're only a few sentences into the conversation and I'm already confused.

Dyne, you are correct, my thinking is no better or worse than anyone elses. But I'm not talking about people's ability to comprehend ideas. In fact I expressly disagreed with that patented answer. I'm talking about such verbiage being held up as a reason to not implement features, to not even consider that there may be a market for those features. Many web stores have crashed in recent years because they relied on the cool, flashy stuff and assumed everyone would just latch onto it. I submit that we do actually think, view the world of Second Life, even view each other, differently inworld than we do in real life. And I think that bears exploring through feature introduction and enhancement.

OMG! I see now what Linden Research means when they say "experimentation," "explore," and "compelling." Whoa. This, SuezanneC, is why we don't have choices. It would be too much data to correlate. They *are* exploring.

Yes, it's evidenced in this thread as well. Maybe I would choke on a full cup of virtual mojo. Wonder if we'll ever know?

Hi Khamon!

At work we fax proofs of banners that are, say, 40 feet long and 5 feet tall to customers for approval.

We get people who ask if what they are looking at is actual size.

I guess Second Life has to be designed for them.

The thing is, if you give em the magic stuff without telling them they will accept it because they won't even know it's different than the real world cause they think the earth is flat and goes around the sun and they believe in astrology. And so on.

Actually, yeh, turning a bunch of separate spaces into a single 3d space is a nifty idea, but it's still a 3d space... and it won't do anything about the performance problems of having too many people in a sim... because if I'm there streaming data from 4 sims, I'm putting the same load on each sim as if I was in there sitting on a prim, that's 4 times the lag now.

That's why I specified *opaque* prims.

so who is sl for? visionaries or people trying to copy first life into second life?

I'll answer:

c) LL can't figure out whether to please one or the other.

Masturbating intelligencia.

keep the sl grid as it is...

and license the client/server/software so people can create their own grids

and maybe even create another grid that's not slavishly first life-ish

Hm, what about the outdoors? Pristine natural wilderness is an open-ended space, and most people would like to have more of it right outside their living room window in RL. Certainly in SL people post to express their appreciation of well terraformed new sims, and some posters also lament the crappification of those sims as the land is chopped and built over in the usual way. People care about nature, so it's a good candidate to press into service to support more interesting connectivity between spaces in SL.

This thread also takes aim at opposing attitudes like 'that isn't realistic'. Here's a suggestion: it's fairly realistic to plunge into dense wilderness, and get lost! Sometimes you don't mind, because it's just undeveloped parkland in your RL area, and you know that walking/cycling in any given direction will eventually take you to a recognizable subdivision or golf course (or mall, if you're unlucky). In SL getting lost in certain restricted LOS wilderness areas *could* mean ending up in a different space that is not connected by a shared contiguous edge of a 256X256 height-map having global coordinates.




We've heard Linden musings about a proposed future of SL as a set of parallel private grids with LL hosting the main grid, and charging the private grid owners for software/services (unspecified) that make them an integral part of the aggregate SL universe. In that vision LL is charging for some form of connectivity. Connectivity between grids could be implemented by special portals resembling telehubs, or by a more fine-grained capability that allows the owner to specify a link from his wilderness area to somebody else's wilderness area in another grid.

This would allow content creators to craft an immersive experience for the resident who deliberately plunges into dense wilderness in an attempt to reach a new land. : )



All good stuff, but there's one practical requirement: the interestingly connected spaces need to be represented on a two dimensional map such as the big map of SL.

This is a hard problem. Somebody could axiomatize the proposed geometries, and prove that an information-preserving map to a 2D space cannot be found, but I think that most of the posters in this thread can appreciate the difficulty without seeing a formal result.

This problem is related to one we see in SL right now: crowded subcontinents want to expand in situ, but they're limited by the amount of adjacent empty space available. I'm betting that the current solution is some kind of labor-intensive reshuffling process that can't accomodate all the PI owners' requests. More interesting connectivity might help if it were based on a rich geometry (such as LIT) with a precise mapping to 2D useful for viewing the region of the global map that the user has selected by a focus mechanism such as a mouseclick.

the problem of mapping sl... should be like the problem of mapping internet domains or websites.

i could be seriously misunderstanding the point here... but keeping sl topography the way it is because it would be hard to map otherwise seems backwards.

/me thinks this is like asking Castrovalvians for the quickest way to get there.

but if you're a time lord... should be able to figure it out...

And y'know, we've been hearing a lot of that lately. I first heard that roughly a year and a half ago. I'd like to see a clear path drawn by LL between Point A and Point B, with work toward that end, if I'm to believe it.

As seems so apropos to the situation: show me the money. Otherwise, I'm simply not convinced.

Why?

The mechanisms I described were all based on the assumption that the connections would be defined within the existing space, though not necessarily a 2d one. In another thread there's been some discussion about the possibilities of using large negative offsets (starting beyond max draw distance) for really private private spaces, with the constraint that (1) you couldn't cross parcel-owner boundaries in the private volume, and (2) only a script owned by the parcel owner or an officer of the owning group could llTeleportAgent you into the negative space.

I can't wait for llTeleportAgent, it'll allow for non-linear space *within* a sim if it's done well (not if you have the everything-goes-dark-and-you-get-a-progress-bar), but you really need connected spaces to make it work beyond that... the *sims* need to be involves in the connection so that the scenery in the destination is preloaded as it is in a sim crossing.

Perhaps something like a series of rooms, and one clicks on the door (which teleports the user) to go to the next room. We could even do that now, although having to click would detract from the experience. Something that teleported you instantaneously to the new location on walking into an llVolumeDetect would be much cleaner. Perhaps if the destination location were set up to be identical with the start location within line of sight it could be made almost seamless.

Mapping to a 2d map is essentially impossible once the geometry gets the slightest bit complicated. Even the projection from the spherical surface of the Earth to a flat map has all sorts of distortion problems.

Hm. After getting sort of pseudo-insulted on the sly, perhaps I should clarify my intentions. I value precision in words, but I'm often very bad at it. I'm not belittling Khamon or anyone here at all; it's just a discussion is more interesting with someone prodding the theory a bit rather than saying "gee, that sure does sound great!". Thus, I'm trying to look at the practical upshots.

'course, if Khamon was just looking to use the forum as his personal blog.... well, tough.

For example, take a theoretical non-Elucidean, 4-D, topologically wacky virtual world... and then you have to figure out how to meaningfully reduce it to two dimensions in such a way that people will want to use it. We're still limited by the form of the computer screen and of the human eye.

Another one: how do you make any money off it yet keep it as a useful tool, presuming you're doing this commercially as Linden Lab is doing?

I don't think these are unimportant things to consider, even when making a simple comment of "why do we limit ourselves to spaces we could see in reality"? I think "because if they don't seem like reality, they don't seem like spaces - like a place as opposed to a computer display" isn't a bad theory. Saying "well, people can damn well ADJUST!" like Jerod does is senseless if people don't WANT to adjust. We live in a world where, when given even the subset of total freedom from reality SL does, most people make... McMansions and casinos. We live in a world where people brag about their inability to press two buttons and set a clock on a VCR. I'm not saying I like this state of affairs, 'cause I don't... just that I try to work from the starting principle that Linden Lab is not actually run by idiots. They might be wrong, and time will tell, but I don't think they're stupid.

This kind of thing is a big problem, one for which not much progrss has been made in demo form, let alone something you'd want to throw all the rest of SL's complexity at. This is the best I've seen for computer graphics in non-Euclidean 3d geometries: http://www.geometrygames.org/CurvedSpaces/index.html. I've not seen anything in 4d beyond wireframe hypercubes.

It is very hard to make these kind of things intuitive - our brains are wired to cope with 3d, Euclidean space.

while it's usually hard to do things outside of euclidean space for conscious processing... it's fairly easy to make even non-euclidean processes automatic.

I'd kill (or at least maim) for a multi-levelled world to explore.


Seconded... as long as "magic" includes sci-fi elements too.

You've lost me, what is a non-euclidean process?

processes involving trying to do (think about) things in non-euclidean spaces.

Ah ok, I think I know what you mean. Like drawing triangles and so on in hyperbolic space can become natural?

I'd love to explore a hyperbolic Second Life, I just doubt that it's feasible. What coordinate system do we script in? How would people manage to not get lost?

Sticking to a piecewise euclidean space would be easier, and maybe not that far from being at least simulated if we were to get portals.

That's the kind of thing I was thinking of. Of course, right now you have to right-click and select "Open Door" on the "sit" slot. With the next release we're supposed to be getting the ability to make "click" perform a sit or buy action instead of just "touch", which will allow you to do things like move the user to a new room where the "same door" is open, but it'll be hard to keep their position the same and it wouldn't look the same to other people. llTeleportAgent would let you do the same thing on volume detect as you walk down a corridoor.

True, but anyone who's worked on skin textures has dealt with that. (speaking of which I wish I had a paint program that let me paint stuff on a sphere or other shape and have it map that to a two-d UV map)

Actually, I don't think that's what we're talking about here. It's more like "why are we limited to such a small subset of the spaces we could see in reality". Reality includes space, deep oceans, and other places where gravity doesn't have the same effect. Reality includes worlds with more or less curvature.

And one huge reason for non-linear space is to allow us to use it to simulate things in the real world that SL can't really manage now, but the ancestors of SL could.

Think of the Star Trek "holodeck". In a limited space they simulate huge areas, with moving floors and short-range transporters and tractor beams and other buzzphrases. That's something people have little difficulty understanding... but it's an example of a non-linear space... it's bigger inside than outside.

Doctor Who's Tardis, that's another one people have little difficulty dealing with.

The wardrobe in "The Lion, The Witch, and the Wardrobe".

Mister Rogers' Neighborhood, and every other puppet show, mixes scales with wild abandon. Fraggle Rock, the Muppets, ...

Then there's cartoons. Where would the Roadrunner be without the ability to run into a painting?

Non-linear space is really so commonplace on the big screen and the small that yit doesn't even raise an eyebrow in kid's shows. It's not scary, it's a useful tool.

Deep Paint or Tattoo might help you. The painting capabilities of Tattoo (the only one of the two I've tried) are pretty simplistic compared to full-blown Photoshop, but I know several people who swear by it as a means of blocking out the basic areas of a skin or clothing.