[FinalBeyond]

Any betas? *shot*

[Taco]

Ask Adrian about the math on the whiteboard :P

Ok, so I realize there are more important questions, but maybe if you're short a few questions you can throw that one in there for the hell of it and see if he remembers.

[ICEknight]

Rika Chou, on Nov 1 2005, 02:21 PM, said:

ICEknight, on Nov 1 2005, 11:49 AM, said:

LOst, on Nov 1 2005, 12:46 PM, said:

ICEknight, on Nov 1 2005, 06:49 PM, said:

LOst, on Nov 1 2005, 07:57 AM, said:

We have seen the source code to Sonic and none of the stuff you said were actually possible on the Genesis hardware.

Can you please elaborate on that? What source code do you mean?

Sonic & Knuckles source code of course.

We have seen the source code? If we tell him that, which as far as I know is not true, he won't even try to look for any real source code he might have lying somewhere.

Lost has the "real" source code. Remember?

I thought he said he only had the labels... But either way, since we don't have it we should ask Adrian about it.

Just in case he's willing to share his pieces of code.

[Qjimbo]

LOst, on Oct 4 2004, 12:56 AM, said:

XFox Prower, on Oct 4 2004, 03:43 AM, said:

If I remember correctly, When a ROM is disassembled, all labels are shown as addresses since comments and label names aren't preserved. Did you improvise with the names or did you somehow find the original source document?

Found original :P

So naturally people go on to say...

ICEknight, on Nov 17 2004, 12:08 AM, said:

When will it be released? 2005?

To which LOst replies...

LOst, on Nov 17 2004, 02:07 AM, said:

Between 2006 and 2010 :P

But then he goes on to say...

LOst, on Nov 26 2004, 07:55 PM, said:

ICEknight, on Nov 26 2004, 09:06 PM, said:

</invisible>

Why not releasing it already?

I have released everything I got. I only got the labels.

Which confuses me, why first state that you're going to release the code in 2006-2010, then say you've realsed everything already? It makes no sense.

[ICEknight]

That's like going a bit off-topic, don't you think?

[Qjimbo]

Yeah true, sorry for disrupting the topic.

[Hivebrain]

LOst, on Nov 1 2005, 12:57 PM, said:

3. In the MTV Sonic & Knuckles contest video back in 1994, you explained some of the physics of Knuckles, such as slopes and gravity and stuff. We have seen the source code to Sonic and none of the stuff you said were actually possible on the Genesis hardware. Still, how much of that physics stuff did you do, or helped Naka to do?

Was the stuff he said on MTV wrong? I remember him mentioning something about angles and gravity. The game doesn't actually calculate angles using trigonometry (instead it uses a shortcut table), but gravity is accurately simulated by increasing Sonic's vertical speed.

[LOst]

Hivebrain, on Nov 2 2005, 04:16 AM, said:

LOst, on Nov 1 2005, 12:57 PM, said:

3. In the MTV Sonic & Knuckles contest video back in 1994, you explained some of the physics of Knuckles, such as slopes and gravity and stuff. We have seen the source code to Sonic and none of the stuff you said were actually possible on the Genesis hardware. Still, how much of that physics stuff did you do, or helped Naka to do?

Was the stuff he said on MTV wrong? I remember him mentioning something about angles and gravity. The game doesn't actually calculate angles using trigonometry (instead it uses a shortcut table), but gravity is accurately simulated by increasing Sonic's vertical speed.

He's indeed talking trigonometry, in the beginning. But the stuff he is putting up on the whiteboard is obviously fake. I want him to really answer the question if he was using the real math or just fooled us around to make it funny. We know the stuff he is drawing on the whiteboard is not how Naka wrote it using Assembler (which is why I want to say we have seen the code in a nice way "we have the source code lol"), and we know he didn't program Sonic, or Knuckles for that matter. I just want to make sure that Adrian's work was the real deal, and not just a funny sidekick.

This is what Adrian says anyway (according to my English skills):
So we got Knuckles comming up the slope, an angle of beta degrees say, so he got gravity pulling him down, this is the thrust from the joystick, so then if you stop thrusting, he gets track back, the trick is make sure of - that no gamers know what I am talking about after the first thirty seconds that way I am free to go do what I want.

Is that sounding like how Sonic work? If Hivebrain says so, then it will get interestings. But if not, we should really ask the question anyway. - LOst makes as much sence as Adrian at this moment making two points in the same cut.


So Knuckles is working like the whiteboard says, then you explain to me how to put these formulas into M68k assembler:

Attached File(s)

•  wboard.PNG (5.85K)
  Number of downloads: 2

This post has been edited by LOst: 02 November 2005 - 12:21 AM

[Quickman]

I wish I knew what that big messy mathematician scribble was. It looks like a bad facsimile of some of my working.

On a slope of angle θ with a gravitational force pointing straight downwards (Sonic's world has zero curvature) with a constant acceleration up the slope of a ms^-2 then the drag force is equal to a coefficient of friction μ (which I imagine is a constant for all surfaces) multiplied by the normal reaction force R to the slope. This requires one special function call, to grab the cosine of the angle. The rest is stored in RAM.

The hardware does these checks as iterations rather than as a continuous process, so we can reduce everything to a constant acceleration problem. (Just remember to recalculate the value for a using Newton's second law ready for the next cycle.)

This post has been edited by Quick Man: 02 November 2005 - 01:02 AM

[LOst]

Quick Man, on Nov 2 2005, 08:58 AM, said:

On a slope of angle &theta; with a gravitational force pointing straight downwards (Sonic's world has zero curvature) with a constant acceleration up the slope of a ms^-2 then the drag force is equal to a coefficient of friction &mu; (which I imagine is a constant for all surfaces) multiplied by the normal reaction force R to the slope. This requires one special function call, to grab the cosine of the angle. The rest is stored in RAM.

The hardware does these checks as iterations rather than as a continuous process, so we can reduce everything to a constant acceleration problem. (Just remember to recalculate the value for a using Newton's second law ready for the next cycle.)

This is not an interview for you to answer. Just want to make that clear. I want to know this stuff from Adrian.

[Quickman]

I know that. I was just suggesting how I'd implement it.

[LOst]

Quick Man, on Nov 2 2005, 09:03 AM, said:

I know that. I was just suggesting how I'd implement it.

I just don't want the real answers to be mixed up with yours in the end. You are probably on the right track, just like Adrian's whiteboard... in the beginning. The first part is the obvious part. We know slopes and angles and joysticks are factors of Sonic. But is it his right to take the credits for how Sonic is working when Sonic was written 3 years ago from when he joined the dev team? What if he is just showing off? He is kinda stealing credits, and making it his invention. That's how I get it.

[hxc]

I've sent off the email with most of the questions in..

[hxc]

Double posting but a few pieces of info, before the questions have been answered have come out.

Apparently, according to Peter, Adrian only worked on Sonic Pool.

Chris Senn now works at the same company as both Peter and Adrian, I also have his new email adress.

He said something about Ofer, who created ZBrush, now used in DooM3 etc, but I said about the contact problems, and am awaiting a reply

[LOst]

According to Roger, Adrian worked on S&K. Better ask Adrian himself, I guess.