Tcsc: oh right cool , yeah.

 
Defusco: Zomg: yeah like http://sendvid.com/ph0lzog1 10 seconds on that?

Elvis: Baxx: side note: i think you’ll find if the radius is way bigger than the collision grid size, things wont work right

Elvis: The collision grid needs to be at least maxDiameter+maxVelocity

Reineck: Et09: this is what it seems to do http://pastebin.com/HYAEh2NA

Defusco: Tcsc: yeah they don’t like it! I forgot about limiting them to within grid size actually whilst doing this

Stohl: Is it eval’ed or something

Reineck: Et09: yep, I pasted the thing in node, the error stack trace had it from function dl, so I did dl.toString and got that :

Reineck: Seems that it was eval’d

Champaco: Doesn’thave eval in the code.

Followell: Is this the complete code?

Kucha: There’s no w function declared

Cullin: Function wnhfj { u = nhfj; };

Elvis: Baxx: and i think i use maxVelocity in the code to mean max component velocity, when but when referring to the grid size i mean the actual maximum magnitude of the velocity, which will be sqrt2 * maximumCOmponentVelocity

Dinardo: Well, that’s weird, the beautifier delets the function

Barbee: I just threw it into vim and did :%s/;/;r/g

Altig: It looks like it downloads an .exe file

Defusco: Tcsc: sometimes the collision output on the debugger will be more than 1 like 18 or something when it’s just one entity colliding with another – is this just because it creates a little feedback loop briefly?

Defusco: I’m not sure if this is even important but that’s what I’ve just been looking at

Belfanti: It does an eval on the last line

Elvis: Yes, that’s possible,

Elvis: That would be because of my dodgy contact handling

Briant: All you got to do is delete last line, print olc, and that’s what it’s evaluating

Elvis: Really i should be generating a manifold for each contact and handling it separately

Elvis: I did say this is not a production worthy system.

Collymore: Lol this is hilarious

Hendershott: Did someone really write an h.264 decoder in javascript? and if so, why?

Defusco: Tcsc: yeah no worries, was just curious, I don’t mean to deliberately poke holes or whatever ;

Harvison: This is the ofending script:

Bader: Http://pastebin.com/SZxJmFru

Hage: What is olc MinusFour

Boatman: Yeah, ‘zomg’ pasted it

Elvis: Baxx: yeah. its def. possible that e.g. one entity pushes the other into the wall, which pushes it back into the entity, which pushes it back into the wall, etc.

Wippert: Why would microsoft make it so you can just download and run an arbitrary program from javascript

Elvis: I think that shouldn’t really happen even though i’m not handing contact properly, but there are probably other small bugs

Elvis: Yeah theres a bug somewhere in there that’s causing it

Elvis: Maybe in fp correction code at the end of collide

Elvis: If i were serious about this i would want to record the collision position, normal, and resolution steps for each collision and display them with the debug renderer

Elvis: That would make this easy to track down

Defusco: Tcsc: ah think you’re right – as when I was mashing f11 it seemed to go into that condition even though the distSQ wasn’t less than radSq

Gleaves: I am trying to concat 15 strings together, is there a limit to the number of string concat can take? it seems like it just doesnt work?

Defusco: Tcsc: yeah this is fine for now honestly, there’s still a lot of work for me to even remake this

Elvis: Baxx: wait is this the first version of that code?

Elvis: That bug is probably fixed in the 2nd if it is

Defusco: Tcsc: oh right cool , yeah this is still mark1