May Contain Nuts

It’s countdown to Mysterium ‘06 time already! Here’s a few things I’m working on right now…

• DPWR.NET’s software upgrade from Invision Board 2.0 to 2.1.7
• Archiver Issue 3 (loooooong overdue now, and probably going to be a post-Mysterium mega-issue)
• A certain island for The Writing Project, which I’d like to actually have finished by Mysterium… and since we’re leaving on the 26th, I need to finish it in 9 days.
• Yearly audit of DPWR’s archive.
  • Status: 271 articles of 856 revised, less than 100 of those 271 actually uploaded.
• Still with the web dev contract from Hell…
• Packing
• Work

Oy.

The gallery here at the Realm is currently experiencing technical difficulties, so I’ll just point you all to my portfolio site, where I’ve uploaded a few WIP shots .

Hooray for low-poly work.

First, a bit of link-age to one of the kookiest and coolest webcomics around, Panda Express!

Now, on with the point . I was discussing my desire to build one of the islands we were developing at The Writing Project before everybody lost the time and/or interest in working on it with my long-time partner-in-crime, SR388. When I mentioned building it in MAX and trying to keep the polygon count as low as possible (I’ve budgeted myself to 100,000 polys for the entire scene), he was naturally curious if I planned to do it in realtime. Of course, I am, and that’s been the goal of this project since its conception: learn to make realtime-friendly environments. He asked me if I planned on importing it into Uru, since even he’s apparently heard that such a thing is indeed now possible, and I responded that while I’d love to, I didn’t want to think about how hard it would be to import the thing into Blender and then export it from there to Uru.

So, he sent me a link to a freeware realtime engine originally developed by Disney in 2002 for one of their online games. The engine is called – of all things – Panda3D. It appears from the screenshots of projects built with it that it’s a bit basic in many areas, and probably doesn’t support pixel shaders or anything fancy like that, but as a basic 3D engine to get started with, I think it’ll be great. The screenshots also largely appear to be from projects created by students at an educational institution, so time constraints undoubtedly have an effect on the quality of the images. Even better is that it ships with an exporter to take data straight from 3DS MAX into the engine, which will make transitioning the scene into the engine much easier (I’d still LOVE to get my hands on Cyan’s Plasma tools, though…;)).

All of that, though, isn’t quite as spiffy as the fact that it, like Plasma and a few other engines out there, uses Python to interface with the C/C++ code of the engine itself, so by teaching myself how to use this engine and script with it, I’ll be leaning an important skill set. How cool’s that? .

In the meantime, though, I may be requesting the assistance of someone who is a bit more Python- and programming-savvy than myself to help get me started with this. I won’t have the opportunity to get started on moving things into this engine for a while yet, but anyone who’s willing to volunteer their time in advance is welcome to do so, and I appreciate any help you guys can give.

I have just two words for 3DS MAX 6’s UVW Unwrap modifier:

Jesus Christ.

I don’t think the thing could be less productivity-inducing if they tried, really. It’s tedious, rather shoddy about breaking up meshes for mapping, and can be incredibly flakey. I just hope they’ve made improvements to it in MAX 7 and 8, because in 6 and undoubtedly any version before that, it’s rather torturous. Took me 3 hours today trying to figure out the best way to map the terrain using said mapping tool, and let me tell you, it’s a bitch. I finally decided the best way to do it would be to start with the terrain mesh’s default mapping coordinates, then grab the areas that were near-vertical in the scene and move their corresponding UVW coordinates farther out so that basically what I’m doing is trying to melt the mesh’s coordinates into a plane. I’m not super concerned with there being a little bit of warping going on, partly because it’s a terrain, and rock textures tend to stretch pretty well without being terribly noticeable, and also partly because most of the areas where warping would be a problem are on the rim of the island, not very visible to the player anyway.

Unfortunately, I’m probably going to lose a lot of the work I did learning how to use the UVW system because the terrain is still missing a rather large cavern for some power generators to be stashed in, plus it needs a couple more holes drilled for drive shafts to be run through. I probably have about two hours to go in trying to get the mesh all prettified for textures, and then I have to figure out how to get Cyan’s method of doing texture decals to blend between surface materials working in MAX. That should be loads of fun.

I also figured out how to do texture baking in MAX today… I knew the process existed and what it was for, I just didn’t know where to access it and what steps it entailed. As it turns out, it’s rather simple to do, you just pick the object(s) you want to bake, and hit the button. Piece of cake! And for those who don’t know what texture baking is, here you go: it’s a method of adding lighting and shadow data to an object’s texture map so that the light no longer has to emit shadows (or even exist at all) in order for the object to appear lit by said light in the scene. It’s how Cyan did pretty much ALL of the shadows in Uru (save for the select few that were done in realtime by the engine, such as the avatars, a few other objects, and most of Eder Gira), and it enables you to get the appearance of some rather high-end lighting (including soft shadows and whatnot) without having to compute that stuff in realtime, thus drastically improving your game’s framerate. It’s amazing what old-school tricks still work to create a convincing effect in modern games…

Also, slightly off the subject of UVW maps, I’d like to say that my current terrain mesh weighs in at under 1200 polygons, and it looks pretty damn good for such a low-res mesh. It’d probably work pretty well in Uru without much effort at all (save the nightmare of importing it to Blender…). This is, I should note, in stark contrast to the terrain mesh for Pahket, which weighs in at almost 30,000 polygons. Yes, I did leave my brain somewhere else during the three days I was building that terrain, and I promise it will never happen ever again. I’m rather proud of the terrain I’ve created, and I’m aiming at keeping it below 2,000 polys total after the remaining holes are punched into it. Since I’ve now figured out how to make UVW maps that will work for the terrain, I want to focus on getting the terrain finished and then move on to building the final versions of all of the island’s various machines and other objects. Once all of the machines, buildings, objects, etc. are placed in the scene, I’ll finalize the texture on the terrain, because any terrain editing may disrupt the UVW map, which would suck.

I’m also slightly terrified of the thought of having to UVW map a 30,000-poly mesh *shudder*. Maybe I should stay out of the film industry for my own good .