positive/negative prims?

me, draconic and some of my other buider freinds have been noticing a few things that differ from alot of the basic 3D programs we're useto, and the most signifigant thing we've all noticed is the lack of postive and negative prims..

having the ablity to "subtract" one prim shape from annother would be a huge help to builders for just about any sort of object. a negative prim would essentially be invisible, but the positive shape its grouped with would be hollowed out in the place of the new shape.
the simplest example that comes to mind is any of the Bryce programs, they alow you to group positive and negative shapes together to make seemingly more complex looking shapes easily.

this could solve things like complaints about hollow shape, offset and other similar complaints ive seen recently. it also makes it alot easyer to make things like octagons and hexagons and the like, with fewer prims. (one for the basic shape, and the 4 negatives cutting the corners, vs using 8 triangles)

I'd like to see this though i'd like to know how it will affect objects that aren't owned by the person. I'd hate to see some walk around with this massive huge negitive prim hiding pretty much or cutting through everyone else's prims.

Yes Mordre -- good idea. I've seen this suggested here before, so it has been awhile. A "hacky" implemention of this currently is using invisibility prims with the alpha bug. It's not true as what you're suggesting.

I think this would open up great avenues for creativity, and make it easier to make windows in houses and portholes in submarines... among a great many other things.

I know this as a boolean subtraction, and yes, it would indeed be great. My group hears me complain about this every time I try to build something in SL.

It wouldn't require streaming much more information than any two prims do already, though it would give the renderer some extra work. If LL felt they needed to balance it out, they could have the operation itself count as a prim. So you would use three prims for each subtraction: the prim to be drawn, the prim subtracted from it, and the boolean subtraction.

In any event, this would do much for creating much more interesting shapes in SL.

Fully agree. A prim-based system feels...primitive. Boolean subtraction would help, particularly with arches, windows and doors. I would accept the prim-count being incremented when the op was used.

I think i had the rendering engine actually do this to me when i was ghosting, my tail specifically.

I believe the prims i had attached never got their texture and acted like a sort of cutting tool. Perhaps they could offer simply a null texture as a way for negetive prims?

It would be nice, but how would this affect the physics and texture systems? Since prims are built to only handle varied length of texture keys within a given range (as I know them, anyway), what would we do here to prevent overflows of data?

It's a neat idea on paper (and has been suggested before), but the implications for being buggy and otherwise uncooperative are there.

yes they should consider it... think of it this way..
a wall with two holes cut out for windows would take up atleast 5 primes.
but using a negative then you could create the same effect with 3 primes.
thats just a example but really a whole world would open up for builders

Let's keep it simple for Non-3D modeler and Computer idiots. Not all of us who build on Sl are in the tech biz.
I have no idea what anyone has said here lol

I mean add all the cool things but keep the interface as EZ as click and point. as us can see I haven't even mastered the keyboard yet

There is potential that extends well beyond modelling for this concept.

I don't think that it is an expensive operation for the renderer. I actually think that it would improve overall system performance. I also don't think that it could be abused by someone walking around with a giant negative prim or that it would cause texturing problems.

The underlying system already understands linked objects, so only objects that are linked would be eligible for the boolean operation. The boolean operation also wouldn't have to be realtime. The mesh could be computed at load time and cached for the duration of the render window. (NOTE: It would have to be smart enough to distinguish between eligible static meshes and those that are affected by animation through script or other means.)

This solves the texturing problem as the mesh is then static. It also, on average, reduces the total number of polygons on screen because the renderer no longer deals with five separate meshes for the single wall with two windows; it deals with a single mesh, no "almost" touching opposing normals, and doesn't have to do nearly as much z-sorting.

Also, if the system understood boolean operations, then an obligatory "add" operation could be performed for any arbitrary set of linked prims when they are loaded by the client. A corresponding UV-map is generated by the client so the textures line up. I would also add an operation that determines if the addition is more efficient than the purely "prim" based structure on the client, so it can choose the easier of the two.

Just my three and half cents.

I'm not on the higher end of tech-understanding in SL, but I think that one of the problems is that the SL system doesn't imagine our prims as "solid". Would it understand how to subtract one shape from another where they interpenetrate? A cube is really just 6 flat faces, there's no "inside" to it....if you rotate your view inside of a cube and look outward, your view won't be blocked by all the "wood" that you're trying to look through, you'll just see straight out into the world....matter of fact, unless the cube has been made as a hollowed shape, you won't even see the "inside" of the walls that make the cube....because they're really just one-sided faces, and they all face outwards.

I'd love to see something like this too, but when I think about it, I don't see how the system would logically understand the idea of boolean subraction for prims. Am I mistaken?


If it IS possible, however, I'd like to see it implemented, and if Boolean subtraction can be done with interpenetrating prims, I'd like to see negative prims be used to punch holes in terrain surfaces in order to go underground.

See this thread for more with regards to terrain surfaces and prim subtraction:
/13/a9/39675/1.html

A boolean operation takes into account the volumetric space of an object, and generates the necessary faces to approximate the resulting form. So, if you used one cube to cut a smaller cube into another, it wouldn't necessarily leave the first cube with a square cut out of the side (i.e., you see "into" the cube and observe its hollowness). The system would understand the original cube was philisophically "solid" and end up reworking the faces so that this illusion is maintained - one cube cut from a larger cube.

This is a good idea, I would like to imagine ways to create shaped holes in things by using subtract or minus prims which were phantim and created space inside otherwise solid objects.. incredible useful.

I also would love to see ' melding' where shapes can be fused together into a seamless prim ( which might cost several prims in value...) with a smooth and unique shape. This would create amazing forms. I would also like to see' extruded' shapes and rounded corners, so u can make spiked objects, merged shapes and others. Prim making i hope will get way more complex one day. The present system is quite dated now in some ways and i hope we get more such ideas and features.

This would be great also for tunnels, and to place holes thru things in a different offset place than just in the middle of things
.

NOT 5 prims... _2_!

make a cube

make a hollowed out larger cube

rotate hollow cube 45 degrees and overlap first cube to cut off the 4 corners...

Similarly, a hexegon could be done with two prisms, right?

Racer P.

PS... ME WANTS THIS TOO!