Make prims great again …

[Qie Niangao]

The best thing about prims is that a script can turn any prim into any other prim. This was nerfed for mesh because some creators imagined that they could protect their intellectual property by making each mesh object be created as an instance of its model, not merely be assigned that model (as sculpts work).

As illicit copies of mesh are all over the grid, this was obviously pointless, but it prevents creating broad classes of useful original content. It'll never change in Second Life because superstition has won.

Anyway, prims are glorious. Scripts can use prims to efficiently "3D-print" whole regions full of structures, then magically transmogrify them, bit by bit, into wholly new structures using the very same objects but with all-new properties and geometry defined by a remote server. Try doing that with SL's castrated mesh.

[Lindal Kidd]

42 minutes ago, Qie Niangao said:

Anyway, prims are glorious. Scripts can use prims to efficiently "3D-print" whole regions full of structures, then magically transmogrify them, bit by bit, into wholly new structures using the very same objects but with all-new properties and geometry defined by a remote server.

I knew there was a reason I liked prims!

(Not that I would ever attempt something like this. Nah, I like prims because I understand them and can build with them. Blender is a LOT harder to wrap my mind around. Wooden blocks have always been more my speed.)

[Istelathis]

41 minutes ago, Lindal Kidd said:

I knew there was a reason I liked prims!

(Not that I would ever attempt something like this. Nah, I like prims because I understand them and can build with them. Blender is a LOT harder to wrap my mind around. Wooden blocks have always been more my speed.)

Not that it will ever happen, but I would really love it if SL just built in a toolkit where we could actually sculpt inworld rather than rely on Blender to make our builds.

 

This is one of the things that I was really looking forward to in EQ Next , but it never happened.  It was so promising though, and they had the tools in world that many people were using.  From my understanding, SL used to have quite the building scene, I wish I was there at the time to witness others make their builds in world at the time, I imagine if it was like other games, and some regions might have had teams of people building their virtual world together.  

Prims do not offer the same level of flexibility unfortunately, it is better than nothing.  A lot of people get upset at the ugly builds, I tend to appreciate creativity in all of its forms, despite whatever talent the builder may have.  It is an expression of their imagination, and I love to see it in world.  

But yeah, make prims great again.  Alternatively, and fingers crossed, give us sculpting tools in world, that we can sculpt our own creations into existence with.  I could learn blender, but it is just different, I don't have a desire to build something outside of SL, then bring it into the world.. only to go back to blender whenever I feel like changing it up a little bit, to only have to upload it again to see if it works in its new environment.  I would rather have tools in world, where I can let my imagination roam free and work around the landscape and other assets I have arranged in the world.  Not that I am a sophisticated artist, but it really gets the creative juices flowing when I can compare what I am building to the world around it.

Perhaps then, we would start to see others join together in building their projects, heck, perhaps we would even see people uploading videos of them building, it might bring in more people to SL... who knows?  At the very least, it would prompt others to play around in SL more, perhaps build more of a social aspect for those interested in being social, and to learn tips from the masters of sculpting, and so on.  

 

Edited January 19, 2023 by Istelathis

[Fluffy Sharkfin]

7 minutes ago, Istelathis said:

Not that it will ever happen, but I would really love it if SL just built in a toolkit where we could actually sculpt inworld rather than rely on Blender to make our builds.

Sadly, as much as I miss EQ Landmark, converting the entire virtual world of SL to voxels is unfeasible.  Some type of "local voxel editor" that can bake the results into a set of auto-generated LOD models using algorithms like Marching Cubes or Surface Nets for organic models or simple cubes for hard surface modeling would be a slightly more realistic (but still equally unlikely) proposition.

[ChinRey]

21 hours ago, Coffee Pancake said:

Linkset size : 

1prim = 1li

2 prim = 1li

10 prim = 1li

20 prim = 2 li

The bottleneck here is how SL's server software handles linksets, or rather how it doesn't handle them at all. Apart from the physics of course, an object is an object as far as the server is concerned. A simple prim cube is just as hard/easy for it to handle as a 60,000 tri mesh and each and every part of a linkset is treated as a separate object. If LL could upgrade the server software to treat a whole linkset as a single object even 50 prims could be reduced down to one land impact.

 

[ChinRey]

20 hours ago, 2Tessa said:

( No advertisement intended )

https://marketplace.secondlife.com/p/Prim-Generator-32-prims-become-1-sculpted-prim/2124122

https://marketplace.secondlifecom/p/Mesh-Generator-Convert-Prim-into-Mesh/5445477

As others have already said, we definitely do not want prim-to-sculpt converters! They are just a way to cheat the system by taking advantage of the fact that sculpts have much, much lower nominal land impact than they should have had.

Prim-to-mesh converters can be great for prototyping but if you want a reasonably good mesh, you need to clean it and make proper LOD models up in Blender or Maya or such and once you do that, you might as well use Firestorm's built in converter rather than waste money on something else.

[ChinRey]

While we're at it, why not also expand the prim twist options. The current prim system is based on only five basic two dimensional shapes that can be combined in various ways and then modified by 15 simple parameters. Avi Bar-Zeev's original prim generator software only had a few hundred lines of code and it could still generate all the prim shapes we know today and even a few that have been nerfed away from SL. There are so many other basic shapes and modifiers that could easily have been added.

Edited January 19, 2023 by ChinRey

[ChinRey]

1 hour ago, Lindal Kidd said:

I like prims because I understand them and can build with them.

Just like Lindal I like prims because they are more user friendly for builders than mesh and also because they potentially are by far the most resource efficient type of virtual world content anybody has ever managed to come up with.

But most of all, I love them because they are fun to build with.

[Atomic Infinity]

I'm just going to leave these here....

Suzanne Vega's Guitar, by Robbie Dingo

Watch The World, by Robbie Dingo

 

From around 2006 / 2007, and still inspirational.

Edited January 19, 2023 by Atomic Infinity
add dates

[Rick Nightingale]

I'm using prims here to do a cute portal-opening effect by ramping the hollow in them. One of the few uses of planer mapping I've found - the inside of the hollow that is.

[Paul Hexem]

There's no denying that voxel based building would be a huge boon. Even in limited form.

[Rick Nightingale]

One of the killers for me with prims is the way the physics cost jumps through the roof when they switch to mesh accounting (by putting a normal map on there), with hollows and twists etc., if they need 'proper' physics. Like my portals or even a simple arch that you want to walk through. I've no idea if the physics cost is justified compared to anything else, but I know I could make a mesh piece with the same shape and acceptable physics for a fraction of the cost, in most cases.

While you can approximate the physics for a lower overall cost using linked, hidden prims, that isn't really on if you plan on changing shapes. Did I mention like my portals?

So yeah... give us nicer prim LI please!

[Scylla Rhiadra]

I used to really enjoy building -- not that I was particularly good at it. I love the superior look of mesh, and wouldn't want to revert to a world of prims only, but one of the reasons I got into photography here was because I knew I wasn't going to be good at making mesh, or enjoy creating it off the grid as much, so I needed a new creative outlet.

Fortunately, I realized that I could also employ building with photographs. Most of my pics use backdrops I have made myself, with a combination of prims and commercial building parts (as well, obviously, as mesh furnishings). One of the most enjoyable parts, for me, of setting up a pic is building the backdrop.

[Marianne Little]

I wish we could have more tools to use with prims. A tool that smooth out edges and corners, to make rounded corners and rounded edges.

[Rick Nightingale]

Agreed, @Marianne Little. It's one of the reasons I'll often mesh a part that could otherwise be a prim. Those sharp corners, compared to a little edge bevelling with hardened normals are a giveaway of prims vs. mesh. That said I see a lot of mesh that looks like it is prims, with the sharp corners; seems like a waste of mesh to me. I often wonder if some that was made with one of those converters.

Having a 'bevel my edges' setting would be great.

My other pet hate is trying to get prims to line up close enough not to have a flickery gap when looking at just the right (wrong) angle, yet also not overlap and get texture flicker. Sometimes you just can't bed them into each other and need edge-to-edge 'seamlessness'. Can't be done... especially at high location values (silly 32-bit SL).

I want to be able to weld my (bevelled) prims together

[Frionil Fang]

Thirded, being able to bevel the edges would make a world of difference; it's exactly those unnatural sharp edges that makes them look so "primmy" and stand out.

[Love Zhaoying]

1 hour ago, Rick Daylight said:

I want to be able to weld my (bevelled) prims together

Heck, if you can get them beveled, that's only one step from ...bedeviled!

[ChinRey]

5 hours ago, Marianne Little said:

I wish we could have more tools to use with prims. A tool that smooth out edges and corners, to make rounded corners and rounded edges.

A long and rather techy reply. Sorry about that, I'll get there in the end but I thought it was time to look at what a prim really is and maybe what it could become.

Any prim is made from a combination of two simple two dimensional mathematical cruves. It starts with a profile. SL prims support four of them, a triangle, a square, a circle and a semicircle:

To add an extra dimension to the shape, we extend it along a path. SL prims officially support two paths, a straight line and a circle. (There are also two weird paths that aren't officially supported but I think we should leave them out for now.)

If we extend the triangle along a straight line, we get a prism, the square becomes a cube and the circle becomes a cylinder:

 

If we extend the profiles along a circle path, the triangle becomes what SL confusingly calls a "ring":

The square becomes a "tube":

The circle becomes a torus:

and the semicircle becomes a sphere:

 

Those are the seven basic prim shapes we have, all based on just five (or really only two with variations) very simple mathematical curves. What if we added a few more curves? How about a squircle:

Or a rounded square:

There's your smooth edge prim!

The mathematical formulas for these shapes (and many other potentially useful ones) aren't that much more complicated for a computer to handle than the ones it has to deal with for the current prim shapes and in theory it's dead easy to add them to the system.

Unfortunately, reality isn't always as simple as theory. LL made a lot of changes to the prim code during the first few years before they lost interest, sometimes as misguided attempts to make it more "user friendly", sometimes for... for reasons. Beq once called it "loose code" and she didn't mean that in a nice way. TheBlackBox had to reverse engineer the whole thing when he made MeshStudio and he once told me in detail how hard it had been to figure on what on earth was going on. Back in 2008 Avi Bar-Zeev, the inventor of the prim, tried to write an article about How SL Primitives [Really] Work but he gave up when he saw what LL had done to his code after he left the Lab. It would probably have been easy to go back to Bar-Zeev's original code and take it from there but that might well break some existing prim builds that are based on the current standard.

---

While I'm on this rant, take a look at this: https://drajmarsh.bitbucket.io/supershapes.html

This is an online shape generator based on the same fundamental principles as prims but using much more complicated curves. Something like this would not be very useful to us of course, the shapes it generates are way too high poly and they are more fun than practical anyway. But there's a vast range of possibilities between this and SL's basic prim system and I so wish we had the chance to explore at least a few of them.

[Love Zhaoying]

25 minutes ago, ChinRey said:

A long and rather techy reply.

Those always get me..excited.

26 minutes ago, ChinRey said:

How about a squircle:

I'm running in squircles!!

Edited January 20, 2023 by Love Zhaoying

[ChinRey]

On 1/18/2023 at 11:11 PM, Kimmi Zehetbauer said:

Faster loading on slower machines then Mesh.

Every single possible shape in the current prim system are generated from a dataset of less than 60 bytes. This is less than half of what a single mesh triangle with its vertices needs. So even the simplest mesh will always take longer to load from the assets server or from cache than the most complex prim.

The viewer's render pipeline also handles prim data more efficiently than mesh. Some tests I've done indicate that when it comes to actual render weight, one mesh tri as roughly equal to 20-50 prim tris. I'm not sure if this is because the viewer is better optimized to handle prims or if they are genuinely easier to handle for our cpus and gpus than polylist meshes but I believe it's a little bit of both.

 

[EliseAnne85]

Love the rounded prim addition idea!  Makes a great shadow shape when flattened.  Most of the shapes are great to make shadows when flattened and then make that shadow your root prim for any full perm item.  IOW, the shadow made from a flattened prim that's textured becomes your root prim and the item changes into your name that way.  I usually toss the mesh full perm shadow out and make my own shadow with prims is also what I'm saying.  The mesh one is not always needed and it's cumbersome.  I make the shadow texture smaller too. 

Edited January 20, 2023 by EliseAnne85