I have been wondering this for quite some time. Is the collection of entities and the outputs they can send to each other turing complete? The input could be seen as starting value(s) of math_counters and the output as the value(s) of math_counter after the calculation "stopped". I am currently guessing they actually arent turing complete, but that is mostly based on my personal struggle to make them do more useful things. ---- translation guide I am quite sure most of you have no idea what I am talking about. Turing completeness is usually aplied to programming languages and it tells you about how much it can compute. and turing complete basicly means that it can compute as much as a turing machine, wich is a computer model. Nearly all decent programming languages are turing complete. the simplest way to prove this is to make it simulate a turing machine. (turing completeness means that it can compute everything a turing machine can) [ame=http://en.wikipedia.org/wiki/Turing_completeness]wikipedia article[/ame]

If it is (a question which is more for the math-proof-nerds) it's far less practical to use than assembly. You'd need to write a "compiler" to generate the actual logic entitites to get anything useful done. I suppose you could submit carefully tuned "problems" to the physics engine to get around certain issues, recording what ball falls in what basket, etc.

I believe it is Turing complete; it supports looping and branching which should be enough, from my understanding.

I was seriously thinking of making a compiler. the vmf format is quite simple so I would be quite simple once I know how to the essential computations. physics simulation: damn, you solved the problem. if it werent for tf2 low enitity limits this would allow some fine calculations. It definitaly makes it turing complete. Even if you could do it using math and logic entities that stupid limit would disallow anything big. Done, added it as attachement