Xonotic Forums

(09-30-2012, 06:19 PM)C.Brutail Wrote: [ -> ]Explanations coming soon

Explanation:

Saw that video. Really, really amazing stuff! Congratulations to everyone who did that

One question: Was the Mario scene video on a brush? Does that mean Darkplaces and Xonotic support ogv videos in shaders at last?

It might as well have been a camera entity, though. I remember C.Brutail asking me about cameras while he was mapping this (but he didn't tell me what he was doing at the moment so I had no idea)

Hehe. It's an ugly camera entity abuse (though one of my favorite in the demo ). I'll give proper explanation asap (when I finally get time to write huge walls of text with screenshots).

Awesome!!

I really want that physics stuff explained!

Truly awesome! :-)

(10-09-2012, 05:12 AM)Maddin Wrote: [ -> ]Awesome!!

I really want that physics stuff explained!

That's "just" ODE integration in DarkPlaces. I don't think we can use it for more than demos like that, though. It's not really useable at all in an actual game.

Wow, that was just awesome! I love the Beethoven dog picture. I wonder how many people got that.

What's the Klein Bottle like? Will it be in game?

Iirc the Klein Bottle is already merged in master, it's a 3rd slot for the Tuba.
I'm writing the first article about the stage, but it's pretty long. Please be patient

Awesome. I not going to smoke again during a match. Go rockets!
Take your time to write the article.

epic ....

also I sure do hope we get ode in the actual game ....

I want to make rope bridges and other stuff hanging from strings ....


if physics and collisions just won't work ... then it will still be useful for client side decorational clutter (like empty bottels and such)

Getting offtopic here, but I have to tell it before reaching that part in the articles: ODE is useless for us, mostly because it's totally random. Even if you use the same settings, the simulation won't run the same way twice.

(10-10-2012, 12:25 AM)C.Brutail Wrote: [ -> ]Getting offtopic here, but I have to tell it before reaching that part in the articles: ODE is useless for us, mostly because it's totally random. Even if you use the same settings, the simulation won't run the same way twice.

Aren't physics meant to be random? Or rather, aren't random physics a bit more realistic or at least fun? My only issue with ODE is that sometimes it lags, and objects fall through the floor or don't collide properly (last time I checked at least). BulletX might be better, but I heard that's even harder to integrate in DarkPlaces.

Quote:Aren't physics meant to be random?

what

(10-10-2012, 04:22 AM)asyyy Wrote: [ -> ]what

If you drop a coin on the table, it's not always going to bounce in the same direction. In real life that's because of clear physical reasons... physics in 3D software are of course simpler. Still, physics acting randomly could be considered realistic for this reason. And either way it doesn't bother me (if eg: I shoot a barrel and it jumps toward a different direction each time).

Well physics in real life work always the same if you can manage to get always the same starting conditions and the same forces working at the same points (that's not taking quantum mechanics into account).

A barrel that would fly in a random direction when being shot would possibly fly straight into your face on some occasions - that would be just plain weird, don't you agree? :-)

(10-10-2012, 05:58 AM)Cyber Killer Wrote: [ -> ]Well physics in real life work always the same if you can manage to get always the same starting conditions and the same forces working at the same points (that's not taking quantum mechanics into account).

A barrel that would fly in a random direction when being shot would possibly fly straight into your face on some occasions - that would be just plain weird, don't you agree? :-)

Normally, it flies opposite of where it's shot from. So this can be used to throw a barrel in someone's face with a well placed laser shot It should however always fly in the opposite direction, but with how random physics are it might not be very noticeable. I tested this with a crate some time ago using the sandbox... apart from glitches it behaved very correctly IMO.

Well duh, that's exactly the case - on simulated physics you will get always the same result when you set the same "experiment", hence - computer physics are not random too, just as real life physics (they might appear like that because you can never have exactly the same conditions, but hypothetically you should get always the same result in reaction to the same circumstances in real life too). :-)

I believe ODE is badly behaved because of floating point errors and timing fluctuations that make the whole thing rather chaotic (literally). C.Brutail and divVerent experienced this while recording the orchestra demo: they had to restart it a few times because the ODE parts would not work the right way. And the bots have no influence on the physical stuff in the background on that map as far as I know, so ODE in DP is pretty bad at what it's doing.

EDIT: Taoki: look up determinism on Wikipedia.

although for small non-player collision decorational clutter it should be fine ... right ...

Sure, but the bad performance stays. This thing sucks and it's not worth the effort to integrate a whole physics engine into releases just to have the possibility of having a couple of objects that can fall on the ground in a map.

details:

Get rid of this patch mesh terrain, it looks weird. Create a model with less polygons thus optimising performance AND look.

patchmeshs are easier to handle than models
and they look nicer than brushes ...

Like the mesh, dislike the yellow and black warning colours.