This will be updated, but I spent a little while today making a ninepatch system for my UI 
It’s buggy, but it’s almost working.
http://pastebin.java-gaming.org/e18df5e2f101c
CopyableCougar4
Got my new arch install actually working, and got some real linux ASM code going. Learned a bit about shellcoding. Fun stuff; It’s nice to know what’s going on at the lowest level in a system. Paired with the electronics knowledge I’ve been gaining, I almost think I could piece together a real computer from scratch with nothing but a bunch of logic gates… If I wanted to rip my hair out putting all the wires together 
@Slyth
Check this sucker out:
http://www.electronixandmore.com/projects/relaycomputertwo/index.html
Pics and videos at the bottom.
Questions: How often do you guys use for , I use atleast twice every file that has any sort of data.
Also, I have taken a break and am teaching myself electronics.
Yeah. Also Unreal prensetation https://de45xmedrsdbp.cloudfront.net/Resources/files/TemporalAA_small-59732822.pdf and Ubisoft ones http://advances.realtimerendering.com/s2014/drobot/HRAA_notes_final.pdf
In Harldand we have very high triangle count >2-3M and about 50% of that is skinned. So any technique that would need extra pass is out off question. Also I am too busy to make anything too sophisticate that also would run on large set of hardware. So I just decided to ditch all camera jittering at all so and use temporal antialiasing just help with temporal shading aliasing and let the fxaa smooth the geometry aliasing. When I get some free time I will put some effort to actually make some high quality antialiasing instead of just temporally stable.
It’s possible to avoid this. You can create an MSAA G-buffer and render to that. Then only shade the first sample, and only use the extra samples as TSRAA coverage samples. This eliminates the extra pass, at the cost of some memory and bandwidth usage of course. This is relatively easy to implement, as the only thing you need to change in your lighting is to read from an MSAA texture. You don’t need to actually output multiple samples or anything like that. Jittering can be achieved by simply changing the sample you read (for 4xTSRAA, switch between the 1st and 3rd sample each frame).
It’s also worth noting that even with the second pass, it’s not a significant problem. Since you only need to render to a single GL_R16F render target (+ depth of course) and don’t need to sample any textures (except possibly the diffuse one for alpha testing), rendering speeds noticeably.
Dayum… Definitely going to read through this when I have the time. Thanks for sharing it!
seems like a lot of relays, those inductors could easily loose quite a bit of power and heat up. I guess if used in a large design its very obvious what they do however I would most likely use a small capacitor used to trigger a transistor or a dual nor storage gate. Very good article to read through though.
Well obviously transistors are superior, but then you don’t have the clickity-clack!
Check out lambdas. Look for google for the functions map, filter, fold and the like. Also, if you think it’s cool, check out functional programming 
Got my shell program running windows command prompt commands (not really shown in screenshot but it works!):
i followed this guys http://research.michael-schwarz.com/publ/2010/vox/ and wrote simple mesh voxelizer on the cpu :
input mesh :
boolean cubes :
hard edge marching cubes :
smooth marching cubes :
Code for this?! Does it take arbitrary models? How fast is it? SAUCE?!?! It’d be cool if I could use this for global illumination cone tracing, but I never felt like implementing a voxelizer myself.
EDIT: OMG IT RUNS ON THE GPU?!
- hell not on the gpu yet … it’s just a voxelizer, not a sparse-morton-octree
- awesome sauce!
- on the cpu it’s slower than oldschool voxelization (http://web.eecs.utk.edu/~huangj/papers/polygon.pdf), but generates “thinner” results. on the gpu, no idea.
- yes, arbitary models (triangle-list)
- which part of the code are you interested in ?
Bought and started reading a book on regexp’s, and then spent the rest of the day pondering why I hadn’t done this sooner.
My OCD won’t let me give you another medal for this awesome achievement 
_basil you amaze me , you churn out these works of art on a daily basis.
thanks alot lcass! … but it’s not as simple. i’ve been tinkering on the boiler plate, the basics and math for a couple years by now. now things got solid enough to emit actual stuff. o/
After forgetting details one too many times, I finally wrote out full procedure for not getting burned by eclipse when importing random projects off of github: