Kyrah Abattoir

Collision wise you will be limited to the ones of the avatar bounding box. I also suspect that you can't fully override angular motion.

We have a pretty well furnished library of geometry function here: http://wiki.secondlife.com/wiki/Geometric

Oh hey if this isn't one of those "Let me try to force object oriented programming in SL". Yeah this looks quite dated judging from the masses of one liner functions and... the copyright mention being 2005-2006.

Hey Wulfie! Consider it done!

None that I really know @animats I've been going through that transition myself, and with the reomval of blender render, that also means crash course-ing through cycle shaders. If you aren't already i'd recommend you join the Blender Benders group, Graham does classes there, and he will start 2.80 soon I believe. But if you where experienced with 1.79 and cycles, most tools I've seen are basically unchanged, and you can still get the 2.79 keys and mouse controls if you so desire.

I don't, it's a multiple part mesh I have to use "data transfer" on the seam points to get it to be seamless.

Well yes of course, not everyone is bothered by scammers either but does it mean it's ok?

Allright I've poked at it some more, it appears that forcing a triangulation before export and before adding custom normals helps slightly, however i still get banding between "alpha cuts" that should not exist. The rightmost is the original that would have banding on every polygon strip. the two leftmost only appear to have banding at UV and material seams. Another weird thing is that if i tick "recalculate normals" in the uploader, all the material and uv borders become completely visible for some reason.

Thanks for the tip, unfortunately a fully uploaded texture did not solve the problem so it must be something else.

I was making a new set of materials for one of my existing meshs and I noticed this: Weird horizontal bands, they don't follow the object's "sides". Now this is what I know so far: It's probably not the normal map: It is pretty much perfect, this is also a temp texture, so no compression artefacts either. The problem is not due to temp textures it persists when the texture is uploaded too. The exported dae appears fine when re-imported in blender: It's smooth, even with the same texture, it shows nothing out of the ordinary. The bands look like they follow the polygon strips of the leg, but some areas are irregular. Forcing a triangulation before exporting appears to help, but does not solve the problem entirely, for some reason, material and UV seams remain visible. There is no split edges/vertices that I could find. This was exported with the standard blender dae export function, and not Avastar, (the body kit I was using when I made this was not compatible) The UVmap looks flawless to me. The problem only occurs on the leg parts of the model, arms, torso, etc where made at the same time and are unaffected. (What it looks like in blender when re-imported) At this point I'm pretty much out of ideas...

This month it was literally 1 day for me. It depends who you are where you live how often and how much I suspect.

"You have reached your avatar polygon quota."

I'm not cherry picking. You are dead set on the idea that some things are simply not your problem and this irrational "zero sum game" that you won't be able to enjoy SL to its fullest with your fancy hardware if creators where forced to be more reasonable. My position is that SL doesn't incentivize good content, there is no real mechanic in place that forces creators into the balancing act that should be expected from them. I see it pretty much every day on the different content creation groups I am part of: mistakes and bad behaviors that you should never see, perdure because ressource accounting is broken or non-existant, and your advice is typically not wanted if it results in more work. It's also why animesh has such a low penetration rate so far (it's not the only reason obviously): The additional restrictions that come with animesh are a brick wall for many creators who, essentially, don't want to worry about limits. And they wonder why it holds no appeal. I guess at this point SL must be such an engineering nightmare that they have developed PTSD about it . But if I was not able to just "walk" from one side of the mainland to the other to see what's "there" I wouldn't have stuck to SL in 2004. It was magic and simply unlike anything else. I came from there.com, and while this world was fairly large compared to a 2004 SL grid, it was mostly a pre-built experience with gigantic swathes of emptyness. Content creation on there.com was very primitive, but one big difference with SL was that the there.com creators where very aware of the limitations of what they created, and when the ability to create custom models was introduced, it was under very strict guidelines that simply prevented creators to create anything that would be more taxing than the reference content: They told you how many textures to use, how big they could be, down to how many triangles where allowed for each LOD. Despite the horrible exploitative business practices of the company, people stayed because the social elements where huge, and the experience was almost always very smooth. There.com was designed as a place where you could have virtual excursions, adventures, and make friends... And then SL came and literally dwarfed it, it was bigger and better on every front, excepted smoothness. If SL was like Sansar or IMVU, it wouldn't have been anything special. IMVU's popularity kinda soared when SL users began shifting secondlife from a "main focus" application to something you keep running in the background "in case something comes up". And yeah, SL will never run like something created by a game studio could, that's a fair and reasonable expectation there. But before throwing the towel and going "well I guess there is nothing that can be done", why not try to reign content to correspond to what is expected from a real game? It's not going to give a solid 60 fps and smooth region crossings to people using toasters, but regardless of how puny or mighty your computer is, why wouldn't you want SL to run better?

I don't think Sansar is built from lessons learned, everyone who actually founded the bases SL rests upon is gone, and the new people seem either unable to bring Sansar to part, or refuse to go down "that path". In a way it reminds me of Ultima Online, it was ambitious and flawed, and every mmorpg that followed it seem to be doing everything in their power 'not' to follow that path. SL Is flawed obviously, but a lot of those flaws could be corrected if LL didn't have to take in account legacy content. Close all the holes that scripters work around by tying ressource consumption to avatars or parcels rather that the per script/per object naive approach they ended up with. Redesign avatar and land limits to take in account everything, from server load to client impact, nothing should be displayed or tax the servers "for free". Make a new system avatar that actually follows the deform bones rather than the weird hybrid thing we have now. Every single problem SL has is due to the lack of reasonable limits on what is fair ressource consumption. I can put thousands of scripts in a single prim... why?

Another big issue that causes texture multiplication is that SL has shifted towards more of a competitive environment than a collaborative one. I know a circle of people who have a shared pool of textures that they try to use as much as possible to reduce the number of unique textures they use. But your typical content creator is not going to share his assets with anyone else, even if they are highly reuseable. In fact i'd wager that some see baking textures as a form of "soft DRM" since the texture becomes unuseable on its own...

The impact of BAKING textures by the content creator is irrelevant to the discussion, why are you even bringing it up? The impact of those textures is the ram/vram they are hogging: Option A: House with a single 512x512 stone texture repeated on the walls. 6Mb of vram, will look prettier with ALM/Shadow/SSAO. Option B: The same house, the same texture, but it has been baked into 16 1024x1024 pre-lit textures. 144Mb of vram, will probably look just the same if not worse with ALM. And no. Lightmapping is processed at the shader level and is so light and efficient that all cellphone games use it. That's what all 3D games where using since at least as far back as Quake. And quake was doing it without the benefits of hardware acceleration. On the other hand, the impact of pre-lit TEXTURES is that they essentially get rid of the main reason we use textures in the first place instead of a one time "pixel atlas": recycling and repeating, which is one way to get nice sharp details for cheap. You are completely missing the point, what's on the texture doesn't matter since compression is (mostly) irrelevant to vram usage, what DOES matter is how many textures and how big they are. The official viewer allocates up to 768Mb of vram to textures only (the rest is used for other things) before it begins the great "texture juggle". Camming inside a friend's "average" furnished house makes me shoot past that after a few seconds (Firestorm has a higher limit but there will always be a limit). And guess what? This is just one house on a 1700 sqm parcel, regions can contain dozens of those, and much worse ones. And obviously i'm not getting into avatars, those are far, far worse. As for how much of an improvement you can get from texture recycling... I wouldn't be against a region-centric lightmapping solution, but Linden Lab has rejected this option in the past (when I brought it up) and I can imagine many reasons why.

The "peanut" island next if you want long mostly straight roads But personally, i prefer the old continent for roads, and heterocera for rails.

Lightmapping in video games is NOTHING LIKE BAKING SHADOWS INTO TEXTURES. Lightmapping is way more efficient and doesn't use nearly as much memory because the textures themselves are not baked, only the light data, and it's done at a very low resolution. Low end computers can't use realtime shadows because they usually don't have a good GPU, but if they don't have a good GPU they also don't have a lot of ram or vram , so burning shadows into textures solves absolutely nothing, it just makes downloads longer, travel impossible and overwhelm the viewer cache. Imply viewer performance is purely a "personal problem" is neither intelligent, nor cool.

Yeah, I mean SL looks a lot better than it did in 2004, but I'd wager that scene complexity has gone in the 100-1000 times heavier. @Qie Niangao you also get the "self fullfilling prophecy" that as long as people use non-alm non-shadow non-ssao rendering, there will be pressure on creators to bake one-time-use-prelit textures for everything, which makes the experience of using alm/ssao/shadows even worse, thus detracting people from using it... goto 1. Also the "little lie" of passing baked shadows for quality... or work. The reason AAA games look gorgeous is not so much that their game engine is magic, a lot of it is down to efficient content and smart level design that minimizes waste, so that more ressources are available to get that "high fidelity" look. That requires compromises. Compromises that we SL creators struggle to make.

@animatsSounds good as a premium perk

I actually do both.

you can edit the exact influence values from the property panel

I know. It also begins with A, that's why I picked it back then, first in the list.

Secondlife is a game kiddo, you just have a very narrow point of view of what qualifies as a game, as usual.

Still waiting for that oh-so-needed avatar ressource cap.