Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Minecraft Physics (wired.com)
80 points by th0ma5 on Feb 2, 2012 | hide | past | favorite | 23 comments


'Unfortunately, this video is not available in Germany because it may contain music for which GEMA has not granted the respective music rights.' :(


This seems like an error; I'm pretty sure that song is long out of copyright. (GMEA can claim copyright on a particular performance of that work, but how does it know?)

Anyway, the transcript is the only important part; the sound is a computerized British voice reading the transcript with some classical music in the background.


Unfortunately, in Germany GEMA enjoys the privilege of inverted burden of proof. Look for `GEMA-Vermutung' for details.



There's a couple of discussions on Reddit related to this particular video which offer some interesting criticism:

http://www.reddit.com/r/Minecraft/comments/nrxme/minecraft_g...

http://www.reddit.com/r/Minecraft/comments/p7lyz/minecraft_p...


How do they make the jump from 43 blocks to 43 meters? Are they assuming that each block is exactly one meter?

Also why did the video randomly go blurry at times?


Yes, one block is considered to be 1 cubic meter. http://notch.tumblr.com/post/443693773/the-world-is-bigger-n...


Another fun question would be to figure out what the terminal velocity is. It's reached pretty quickly when you jump off of things in Minecraft.


I know it is often the topic of SIGGRAPH demos, but I always get distracted now when game or demo physics don't go the extra mile in calculating all of the factors, or having realistic drag (and things like that)...


What would be the advantage of highly realistic physics though? In some cases it would improve gameplay (I remember the Newtonian physics in the Independence War games resulted in relatively novel gameplay) but in a lot of cases I don't think it would. Minecraft especially is kind of abstract, and it already struggles a bit with performance on some larger multiplayer games. Super realistic physics would just make this worse and not add that much to the gameplay, IMHO.


Well, yes I can see in games it being just eye candy maybe, but even in incredibly flashy renderings (like heck even Target commercials) there's just sort of a basic Verlet integration of speed and acceleration or deceleration but not many things like air drag, friction on the surfaces, little "oops" that happen with uneven surfaces and random errors in trajectories, etc... I have seen SIGGRAPH white papers that contain a lot of these sorts of things and more as a group, and well, lets hope those eventually get read and used. The current state of the art reminds me of when I first learned Photoshop and I couldn't stand any magazine covers I saw (actually I guess I still can't heheheh)


Target commercials aren't really a fair comparison to Minecraft. One is the work of a team backed by large company and gets rendered once on powerful hardware (I think... I don't know what you're specifically referring to but I assume it's some computer animated spot) and the other was largely the work of one programmer and has to run on cheap PCs and Macbooks (sure it doesn't have to but I would argue that's part of Minecraft's success: you don't need a gaming rig to play it, as you do to play something like Skyrim).

I definitely understand where you're coming from but keep in mind that you likely have a much deeper understanding of physics than most game players (and likely most game devs).


Minecraft does have terminal velocity, though (it's obvious from the very first shot showing the pole while falling), therefore it does have some implementation of air resistance. So as soon as the video says "assuming air resistance to be negligible," it's oversimplifying, as Minecraft clearly does not have negligible air resistance.


I'm similar. I'm playing a lot of Terraria recently, and it bugs me that you can have a tower of water in a tube connected to a shallow lake, and the pressure doesn't force the water through.


That bugs me too. I have started a small project to make a more realistic water simulation, and found it to be quite a bit more involved than I had expected. Just adding pressure would fix the problem you mentioned, and would probably be pretty easy to do. Adding velocity so that the water sloshes properly seems to be the hard part.


fluid dynamics is anything but easy. current sandbox games (minecraft, dwarf fortress, terraria) implement water flows as cellular state machines. this is clearly wrong, because pressure is non-local, but computational expense of doing water correctly is prohibitive.


> current sandbox games (minecraft, dwarf fortress, terraria) implement water flows as cellular state machines. this is clearly wrong, because pressure is non-local

Dwarf Fortress's algorithm actually supports global pressure propagation, albeit in an ad-hoc fashion:

http://www.gamasutra.com/view/feature/3549/interview_the_mak...

http://www.gamasutra.com/view/feature/3549/interview_the_mak...


There are some mods that try this. Finite Liquid (http://www.youtube.com/watch?v=XlnUXXzc1Fg) for Minecraft, for example, implements a simple pressure system, along with pipes, pumps and other flow/flow-control tools.


>Similar to Angry Birds, this is a great world to explore physics.

Comparing Angry Birds (a physics-based game) to Minecraft (a game with... gravity) is a bit of a stretch.


Unfortunately GEMA blocks me from seeing that video.


Anyone done this for other popular games? could be interesting as well!


There's the Sonic Physics Guide (http://info.sonicretro.org/Sonic_Physics_Guide), though that's more about implementing the physics of the original Sonic games in code, rather than explaining real-world physics through playing Sonic games.


Fascinating, thanks! Also useful in general for me!




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: