Last video took about an hour and a half to upload, this one took 6 minutes
In that hour and a half, I managed to implemented jump plates and better looking boost pads using beams, as well as better particles for the engines.
1 Like
Oh cool, thanks.
Got textures to render! Materials, Textures & Bump maps soon to come!
Special thanks to @sircfenner
Edit: Coincidentally thought the colors were a bit washed out and found a bug that caused colors to be incorrectly detailed. Here’s the fixed version!
Edit 2: Transparency from decals is now transferred (look @ ScriptOn)
9 Likes
Double post because nobody responded in 5 hours and I’ve still been working lol
Pretty excited about this.
So I began rendering images and noticed that the textures would always come out a little grainy… I could solve this through more samples but sometimes that’s not reasonably applicable (ex: looking @ textures from a far distance)
Notice how the black outline around my legs is kinda spotty and the text on my shirt is difficult to read.
Basically what was happening is I would draw a ray, convert the local space that the ray hit into a pixel on the block with a relative %, then multiply that % by the X and Y sizes of the texture map (ex 512x512 image times 10% is 51.2). The problem is that .2 at the end. I can’t get 20% of a pixel so I needed some way to get blend nearby pixels.
Result is me searching in a + formation around the selected pixel. Center of the + is the nearest pixel, the up/down/left/right are each one pixel over. I then multiply these colors by their inverse distances from the ray hit, and get weight. I average these weighted values together and now I get smooth image stretching!
Result, before/after of an up-close image:
Here’s the original compared to the others. The middle image is one of the things I tried but abandoned- I averaged the 4 nearest pixels. This works mostly well but small details are destroyed because it’s essentially just blurring the texture. Take a look at the text on my shirt to see this (compare taking the nearest pixel (left) to blurring the nearest 4 pixels (center) to using weighted values (right).
Click to expand and view detail. (Image is 3000x1000, scaled the entire thing up by 40% to help show the individual pixels a bit better)
Look at the text on my shirt, it shows the strengths/weaknesses of each format. I ended up going with the one on the right.
11 Likes
If it doesnt just fling off to infinity on a slow computer, is it really roblox?
1 Like
I did try making some more cars
I have no idea if they’re any good though. I like just using the LShift drift for every corner.
1 Like
That’s really good! Nice to meet you again. 
On Black Friday while everyone was shopping I had some free time, so instead of working on one of my projects I wrote a basic Mortal Coil solver.
What is Mortal Coil? It’s a programming optimization problem on hacker.org.
8 Likes
I recently wrote a basic brute forcing Runaway Robot solver. I’ll do Mortal Coil eventually…!
Make a gif, I want to see how the shadow looks in action 
1 Like
I have spent too much time on this (and also not nearly enough.) here is a thread about it https://devforum.roblox.com/t/mortal-coil-puzzle/33796
2 Likes
I remember something about Shedletsky once having said “get in the top 10 for mortal coil and you’ll get a job at roblox” long ago. Had some good fun with a few fellow robloxian programmers about that game back then.
It was a tweet from builderman inviting anyone who could beat level 1000
https://twitter.com/DavidBaszucki/status/667173653574914050
The puzzle is hard because it increases exponentially in complexity, where any solver will take about twice as long every 10 levels.
2 Likes
If this offer still stands I’d gladly waste the rest of my life trying
I think it would be much easier to just go to college, get a degree, move to california, get a few years of experience and then apply normally.
1 Like
looks really good!
1 Like
