Sorry your browser is not supported!

You are using an outdated browser that does not support modern web technologies, in order to use this site please update to a new browser.

Browsers supported include Chrome, FireFox, Safari, Opera, Internet Explorer 10+ or Microsoft Edge.

Work in Progress / Fastest .X sliding collision ever. d-load & screenies inside.

Author
Message
FoxBlitzz
20
Years of Service
User Offline
Joined: 19th Nov 2003
Location: United States
Posted: 17th Jan 2004 05:43
Both rar files are missing the required map files to run. You better fix that. Also, I prefer that you put number 2 in a zip because Winzip is more straight forward. It automaticly unzips to a designated folder in C:\Unzipped AND it opens the contents when finished.

Other than that, nice job. I'm getting around 460 to 500 fps. What 3D cards do you guys use? XD

HP Pavilion | Intel Pentium 4 CPU 2.40 GHz
512 MB Ram | NVIDIA GeForceFX 5600 AGP, with 256 MB Ram
Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 17th Jan 2004 16:16
Quote: "Both rar files are missing the required map files to run."


thats strange, worked for the rest of the people. I won't rezip anything cuz I'm working on a new demo which will be available in .zip and .rar .

Quote: "I thought you said you were earlier on but it would be a low price. I guess you're undecided then?"


I said that IF i was going to sell it, it would be for a very low price.

I'm first going to finish my system and the I'm going to make a level, I'll send it over to you then .



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
Sam Cameron
20
Years of Service
User Offline
Joined: 26th Dec 2003
Location: 0,0,0 I\'m the center of the world !!!
Posted: 17th Jan 2004 17:06
Mussi, choose a very very very low price, because you have a lot of
competition for just a code (look nuclearglory,Exeat collisions, etc... and I'm sure exist more codes) , I will never pay for a code, sorry, maybe for a DLL,or something awesome, and even I don't think so
Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 17th Jan 2004 17:25
I wouldn't call them competition just joking



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
Chris K
20
Years of Service
User Offline
Joined: 7th Oct 2003
Location: Lake Hylia
Posted: 17th Jan 2004 18:11
Surely you can just dump the DBPro code into a DLL somehow.

walaber
20
Years of Service
User Offline
Joined: 22nd Oct 2003
Location: Los Angeles, CA
Posted: 17th Jan 2004 18:18
if it's just DBPro code, WTF is the point of a DLL?!?

Go Go Gadget DBPRO!
Chris K
20
Years of Service
User Offline
Joined: 7th Oct 2003
Location: Lake Hylia
Posted: 17th Jan 2004 18:37
So no one can see your code.

Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 17th Jan 2004 19:27
Quote: "Surely you can just dump the DBPro code into a DLL somehow.
"


dunno if it's possible, maybe easier for the users that way



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 18th Jan 2004 02:47
Quote: "I'm first going to finish my system and the I'm going to make a level, I'll send it over to you then "


Okay, I have a strange feeling you're going to make this really wierd level my collision probably can't cope with lol.


Quote: "Surely you can just dump the DBPro code into a DLL somehow."


It would require a lot of rewriting. Fortunately due to IanM's DBP to C++ Interface Library you can use the same commands but you'd have to rewrite all variables, arrays, types and basically anything else to C++ format. Not to mention adding line termination characters to everything... would just be a waste of time IMO.


"Computers are useless they can only give you answers."
Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 18th Jan 2004 13:27
Quote: "Okay, I have a strange feeling you're going to make this really wierd level my collision probably can't cope with lol."


... you'll see, it's not to make your life hard but it's because the users would want it that way



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 18th Jan 2004 13:34 Edited at: 18th Jan 2004 14:37
I'm fine with that.

Edit: Urghh, it's going to take me a little longer to get a demo out. RobK's new IDE update won't load my project and in addition I've decided to create a Technical Document/Tutorial on how to use and recreate my collision system. Explains everything behind the collision from how it works, why it does everything and how to program it.


"Computers are useless they can only give you answers."
Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 18th Jan 2004 16:24
just wanted to tell that I almost doubled the FPS, now I'm going to start on tweaking up some code



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 18th Jan 2004 20:18 Edited at: 18th Jan 2004 20:36
hey, about turning it into a .dll, would that mean it might be useable with DBC ?

EDIT
just wanted to add that if I turn collsion off I almost get the same FPS as when it was turned on




Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
QuothTheRaven
21
Years of Service
User Offline
Joined: 2nd Oct 2002
Location: United States
Posted: 18th Jan 2004 22:01
I was able to break 1000, very impressive. But it's not sliding friction, it's static. You don't slide against the objects.

Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 18th Jan 2004 22:30
oke, I don't know what you downloaded but I'm absolutely sure it is sliding collision. maybe you downloaded the first test, the second one has more of a sliding collision. hope to finish my final code soon so I can get rid of any other confusion



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 19th Jan 2004 02:05
Stop bragging


"Computers are useless they can only give you answers."
feiting
20
Years of Service
User Offline
Joined: 19th Oct 2003
Location:
Posted: 19th Jan 2004 04:46
I've been watching this thread patiently and must say I'm pretty thrilled/excited/IMPATIENT well I am patient but I'm more excited now.... please get something! I'll test it with a Bryce level! Exported to 3ds then eventually .X Static tree collision!

Well I'm working on fixing my carpal T, I think it's working (vitamin B6) so I could just upload the .x level for you guys to download/test earlier than me.

He who does not take the assassin job, will get killed by one who does.
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 19th Jan 2004 07:38
We should have Rich do a 'stress test' with mine, Mussi's and the NuclearGlory collision systems for the next newsletter. For instance, he could make odd shaped levels to see who has the most accurate, he could make really polygon and object extensive levels to see who's runs the fastest and even value who's is the easiest to use. It would make for an interesting and useful featurette although it looks as if Mussi's going to be the hands-down winner though.


"Computers are useless they can only give you answers."
walaber
20
Years of Service
User Offline
Joined: 22nd Oct 2003
Location: Los Angeles, CA
Posted: 19th Jan 2004 09:50 Edited at: 19th Jan 2004 09:51
I'm sure both exeat and mussi are actually working on their systems (seeing as mussi has released a few demos), but personally I think it would be really funny if this was just a big prank and really there is no "magic-super-fast-sliding-.x-collision-using-only-intersect-object-like-once-per-loop" system

p.s. this little "rivalry" has inspired me to finally break out the ol' intersect object command for myself (first time), and I must say I have my own little sliding collision system up and running, and it was a great learning experience... I can feel my code skillz (purposeful "hip" mispelling) getting better

looking forward to both demos!

Go Go Gadget DBPRO!
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 19th Jan 2004 12:06
Yup, it's actually really, really simple to do the collision, isn't it? To be honest, I haven't touched my code in about 3 days because I've been too busy modelling instead. heh, I guess I'll have to do a little optmizing later on.


"Computers are useless they can only give you answers."
walaber
20
Years of Service
User Offline
Joined: 22nd Oct 2003
Location: Los Angeles, CA
Posted: 19th Jan 2004 13:48
yep, that's the stage I'm at... although mine is probably nowhere near as optimized as either of yours' is. mine is sort of specific to the current project I'm working on (NOT a fps!), but I'm really happy with the results, and it can only get faster, right?

Go Go Gadget DBPRO!
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 19th Jan 2004 13:54
Yup, as soon as you learn all the little tricks to get it to skip checks then it will run very fast. Again, my Tutorial will clear up a lot of techniques and ways to get it run very fast, it will be by no means complete though and quite often I've asked people to email me solutions to problems I had.


"Computers are useless they can only give you answers."
nuclear glory
20
Years of Service
User Offline
Joined: 2nd Oct 2003
Location:
Posted: 19th Jan 2004 22:18


Nice continued work. Please keep at it.

I did get a chance to download the free-flight example, looks good. I went a step further and swapped out the first X model with a level of my own and noticed some things.

Let me know if you want me to post them. I don't want to be the 'hater' who's trying to shoot down your progress, lol.

Lead Programmer/Director
Powerful Collision DLL for DBPro and DBC: http://www.nuclearglory.com
Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 19th Jan 2004 22:28
well you can post them, better yet... I want you to pst them

I'm adding a very nice thing to the engine, no more multiple files!!! just one big .X but with lots of limbs you can still do multiple files but this is just so you won't have to export 200 meshes by hand . and it also makes sure you don't lose any textures in the progres .

I'm just getting warm tho, gonna add a function which will speed things even up more and when I've done al that I'm going to tweak my code for even a HIGHER FPS!!

oké, sorry for the bragging... I'll stop when you guys can see it for yourselfs



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
nuclear glory
20
Years of Service
User Offline
Joined: 2nd Oct 2003
Location:
Posted: 19th Jan 2004 22:44 Edited at: 19th Jan 2004 22:45
Okay, here's what I did.

I swapped out the first .X level with the demo level from our own collision system.

Here's some thing I noticed:

1) With the large majority of the mesh, collision wouldn't work. I'm not sure what was causing it, but I would backup and try to hit the wall differently with no success. It acted like the wall wasn't there. When I traveled back to the center of the room and move upward to the ceiling center piece then collision was working as it's suppose to. Collision seemed to only work in the center of the level.

2) I began to test out the collision against the ceiling center piece a bit. I found that when I rode the camera across the edges of this center piece it would hang on the edges and/or suddenly pop the camera out from the face once you jiggled the camera a bit.

I have a very good idea of what your problem is with edges and places where models come to a point. The problem comes from the nature of the "intersect object" command which I commonly call "ray-cast". The issue is that when you shoot your "ray" outwards to detect collision there is no way to determine where the edge is coming from and there is a very high chance that the ray will miss the model edge. And by the time this is detected your container (a sphere I assume) has already penetrated the edge and you get those weird effects. This will cause a lot of trouble with stairs also.

Those were the biggest things I noticed.

Do you have any way of increasing the size of the camera collision shape?

Lead Programmer/Director
Powerful Collision DLL for DBPro and DBC: http://www.nuclearglory.com
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 20th Jan 2004 06:26
Quote: "I'm adding a very nice thing to the engine, no more multiple files!!! just one big .X but with lots of limbs you can still do multiple files but this is just so you won't have to export 200 meshes by hand . and it also makes sure you don't lose any textures in the progres ."


I already have that working incredibly smoothly in mine


"Computers are useless they can only give you answers."
Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 20th Jan 2004 20:09 Edited at: 20th Jan 2004 21:10
@ nuclearglory

I've seen it, I've seen this before... that's why I'm just sitting back cause I already know that I have the solutions .(matrix humor )

I know what ya mean I already have a solution for it .

Quote: "Do you have any way of increasing the size of the camera collision shape?"


japs, you can increase the radius and the height.

@ exeat

I tought that it wouldn't be a problem for you .

@ everyone who's interrested in my bragging

Tought of some new optimizations which are gonna ge super duper fast .



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
John H
Retired Moderator
21
Years of Service
User Offline
Joined: 14th Oct 2002
Location: Burlington, VT
Posted: 21st Jan 2004 05:03
Personally Im a NG Fan, but I wanna see how this comes along


We need help! Email us! join@eternaldestinyonline.com
walaber
20
Years of Service
User Offline
Joined: 22nd Oct 2003
Location: Los Angeles, CA
Posted: 21st Jan 2004 06:27
yeah... my current system using intersect object is working fine, but my game is a 3rd-person game with 2 characters onscreen at all times, who need to be sliding against the walls/props/etc. I'm currently struggling with the tradeoff between fast collision, and accurate collision.

once I've proven to myself that I can write a relatively nice collision engine, and it's implemented into my game, I'm gonna slap in the Nuclearglory engine, and compare the FPS/accuracy.

but I can't let myself just use a DLL unless I can actually code an alternative myself, and am certain that the DLL is better than my own solution.


p.s. mussi and exeat, release a demo already, the suspense is killing me

Go Go Gadget DBPRO!
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 21st Jan 2004 09:12
Yeah, that's the main problem - deciding wether to make it more accurate or faster. Of course there will always be times where the collision won't work perfectly but that's a sacrifice that has to be made.

I'm waiting for Mussi to make and send me a level before releasing a demo.


"Computers are useless they can only give you answers."
nuclear glory
20
Years of Service
User Offline
Joined: 2nd Oct 2003
Location:
Posted: 21st Jan 2004 15:49 Edited at: 21st Jan 2004 16:33
Quote: "Yeah, that's the main problem - deciding wether to make it more accurate or faster. "


That's the truth. I strove for pinpoint accuracy in all occasions without hurting the performance. I have a level... it was released with the 2.00 package and as a secondary example for the current release in the commercial package (once somebody buys it).

If you wouldn't mind, I'd like to see both of you're collision codes tested against it. It has some very interesting geometry, that really tests the limits of your collision accuracy. Be sure to increase your sphere radius/height too (to make it harder to get around). There are stairs, very sharp points coming from the wall, and wedge shaped corners.

lol, I'll make a package with our latest demo version so you can test it and see what I'm talking about. It'll make for good accuracy competition

I'll post it in this spot once I get it ready...

EDIT:
Okay, here it is: http://www.nuclearglory.com/developer/NGColl/friendly_test.zip (202k)

Let me know if you have any trouble with it.

Also, BE SURE to increase the speed of the camera. Make it go as fast as you possibly want and try to escape from either of the rooms.

Do either of your systems have limits on speed?

In the demo above, be sure to wedge the camera in all sorts of tight spots (even at high speed). In short, try to screw it up.

Full DBP source code is included. If you experience jitters anyplace just increase the "concave mode" at the top of the source. You can also set the size of the ellipsoid there.

Don't forget to ride the camera along the stairs in the demo too!

Some friendly competition never hurt eh?

P.S. This uses the DEMO VERSION of the current DLL release. A notice is placed on screen during gameplay. Make sure to read everything on screen!

Lead Programmer/Director
Powerful Collision DLL for DBPro and DBC: http://www.nuclearglory.com
Sam Cameron
20
Years of Service
User Offline
Joined: 26th Dec 2003
Location: 0,0,0 I\'m the center of the world !!!
Posted: 21st Jan 2004 17:39
I get a 172 FPS max in you test nuclearglory (with a rem in sync limit of course)
nuclear glory
20
Years of Service
User Offline
Joined: 2nd Oct 2003
Location:
Posted: 21st Jan 2004 18:44 Edited at: 21st Jan 2004 18:46
Okay, thanks.

There's one more thing I should make note of: In the demo, the 0/9 keys are used to activate/deactivate the mesh objects, which has the effect of almost disabling the entire collision system in that demo.

All who test it. Please press the key to de-activate the meshes for collision and then report how much it increased your FPS. That should give us a better idea of the impact the collision system is having on performance. It's hard to say how much the drawing operations are affecting performance, so that's the reason.

Lead Programmer/Director
Powerful Collision DLL for DBPro and DBC: http://www.nuclearglory.com
Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 21st Jan 2004 20:47 Edited at: 21st Jan 2004 20:48
hey you guys,

I'm struggeling with the speed/accuracy 3 .
I'm working on this system now which has a high accuracy, but which I think wil be fast aswell. I keep finding new things so that's why it takes so long, I hope to finish it soon .

when were all ready we really need to check for which one is the fastest,accurate and easyiest to use .



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
qwe
20
Years of Service
User Offline
Joined: 3rd Sep 2003
Location: place
Posted: 21st Jan 2004 20:56
how about corners?
sometimes i dont slide off some of the objects at certain angles
its nice

Dell Dimension 4550, Windows XP Home
2.52 ghz, 256 Mb RAM, Radeon 9500 Pro / 9700
Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 21st Jan 2004 20:59 Edited at: 21st Jan 2004 21:04
it will be much nicer, thats for sure. tell me about your corner problems so I can look it up if it still exists and if it did or does see what the couse of it was or is.

EDIT

hey nuclearglory I found a jittering bug in the old_demo.exe . it's when you collide with the wall right above one of that sharp points.



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
nuclear glory
20
Years of Service
User Offline
Joined: 2nd Oct 2003
Location:
Posted: 22nd Jan 2004 04:03
@Mussi
At the top of the source it says to increase the concave mode if you experience jitters. Try setting it to 5 then running the demo. Various concave settings were designed to control how strict the system is with multiple collisions. For some games the settings can be too strict and the mode should be lowered. So that's the reason

Lead Programmer/Director
Powerful Collision DLL for DBPro and DBC: http://www.nuclearglory.com
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 22nd Jan 2004 06:05
Yep, I'm all for giving the collision compo a try.

Will be a little while before I give my collision a test run on it though because I just realised I have a bit of work I need to do for school first.

It should be interesting to see the results


"Computers are useless they can only give you answers."
Chris K
20
Years of Service
User Offline
Joined: 7th Oct 2003
Location: Lake Hylia
Posted: 27th Jan 2004 10:34
How are they going?

Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 27th Jan 2004 16:09
I'm rewriting almost everything... why? cuz I want it to be highly accurate , I beliefe my old system would beat exeat's and nuclearglory's system in speed but would stay a bit behind at accuracy. so I tought up this new system which is gonne be very accurate, dunno how fast it's gonna be but I have a feeling that it will pretty fast to



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
qwe
20
Years of Service
User Offline
Joined: 3rd Sep 2003
Location: place
Posted: 3rd Feb 2004 23:49
im waiting
is it a subroutine like kensupen's code or is it a dll?

when i said what about corners.. i meant there were no corners in the test level that i downloaded.. so i could see if corners work good

Dell Dimension 4550, Windows XP Home
2.52 ghz, 256 Mb RAM, Radeon 9500 Pro / 9700
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 4th Feb 2004 07:32
Mine's a set of functions in a seperate include, so you basically insert the file into the project then use functions like:

ColEnable Object, CollisionType

etc.

The code Mussi posted here was also to be a set of functions as well I think but last I heard he decided to go with making a DLL because the Intersect Object command is too inaccurate.


"Computers are useless they can only give you answers."
walaber
20
Years of Service
User Offline
Joined: 22nd Oct 2003
Location: Los Angeles, CA
Posted: 4th Feb 2004 07:56
I wonder what he means by "innacurate"? It seems perfectly accurate to me.

Perhaps he means that he wishes it provided more information? Like the normal of the polygon intersected? Now THAT would be nice. I heard some rumors that this will be included in update 6... if it did, sliding collision and lots of other stuff could be possible...

Go Go Gadget DBPRO!

Athlon XP 2400+ || DDR-SDRAM 1GB || Nvidia Ti4200 AGP 8x 128MB
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 4th Feb 2004 08:24
The only real way for using Intersect Object for collision is to check from several points on a model to several points to the space they want to move to. If you don't use enough points then there will be times when you'll miss a collision and the object will go right through you but if you have too many then it slows the system down a lot. You need to get an even balance and this isn't easy. He obviously doesn't want to sacrifice any accuracy at all.

Returning the normals of the polygon is hit and I do it in my collision routine. You need to shoot 2 rays out and then get the world coordinates of where they've hit. You can then grab the normals of the polygon (provided they both hit the same polygon) by working out the gradient between the 2 points. Not hard and a more in-depth explanation is included in my documentation.

Returning the exact polygon that has been hit is already possible in DirectX with a command that works the same way as Intersect Object but it returns the polygon rather then the distance. Obviously this isn't implemented into DBP yet, but it is possible. And also, with the polygon, then you can easily get the normals from the mesh data.


"Computers are useless they can only give you answers."
walaber
20
Years of Service
User Offline
Joined: 22nd Oct 2003
Location: Los Angeles, CA
Posted: 4th Feb 2004 08:31 Edited at: 4th Feb 2004 08:33
exeat@ yeah I know... I do the exact same thing in my engine too... find a collision hit, then use separate rays to determine the normal. works just fine, but if the command returned the normal, I could dut down on the amount of times I call intersect object... that's all.

Go Go Gadget DBPRO!

Athlon XP 2400+ || DDR-SDRAM 1GB || Nvidia Ti4200 AGP 8x 128MB
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 4th Feb 2004 08:41
Yeah, the problem is that you can't return 2 values, so it would either be normal or distance. I guess you could do something like the Checklist commands and then store them all internally and we could access them with Checklist Value A, etc or more descriptively "Intersect Normal()" and "Intersect Distance()"... but I never heard of them implementing that in the first place anyway. o_O


"Computers are useless they can only give you answers."
walaber
20
Years of Service
User Offline
Joined: 22nd Oct 2003
Location: Los Angeles, CA
Posted: 4th Feb 2004 10:11
somewhere in another thread... i think someone mentioned one of the TGC staff saying it would be updated to include more feedback... can't remember which thread it was... ah well. it works for now.

Go Go Gadget DBPRO!

Athlon XP 2400+ || DDR-SDRAM 1GB || Nvidia Ti4200 AGP 8x 128MB
Dave J
Retired Moderator
21
Years of Service
User Offline
Joined: 11th Feb 2003
Location: Secret Military Pub, Down Under
Posted: 4th Feb 2004 11:31
It could well be true, it's just that's the first I've heard of it lol. In fact, I hope it is because as you said, it would make things a lot easier (and faster).


"Computers are useless they can only give you answers."
Mussi
21
Years of Service
User Offline
Joined: 27th Jan 2003
Location: Netherlands
Posted: 4th Feb 2004 19:05
Quote: "I wonder what he means by "innacurate"? It seems perfectly accurate to me."


what I meant was something like this situation:



ofourse you could do more intersects but that would realy slow everything down. In my .DLL I'm making now this doesn't happen. The .Dll I'm working on retuns which object you hit plus it returns the poly normalangles and as an extra it returns the world coordinates of where the hit occured . As an extra in my .DLL I made my own intersect command which calculates: distance, normal angles, world coordinates of hit position.

it should be ready for a demo soon, so stay tuned



Specs: AMD Athlon 1800, 256 DDRRam 266mhz, 80GB HD 7200rmp U133, Geforce 4 Ti4400 128mb
nuclear glory
20
Years of Service
User Offline
Joined: 2nd Oct 2003
Location:
Posted: 4th Feb 2004 23:41 Edited at: 4th Feb 2004 23:43
You are very persistent Mussi. I highly respect that quality.

I am going to enjoy battling it out with you in our "biggest and baddest" collision system compo.

Here're some rules:
- no hits below the belt (IE: character slander)
- no hard feelings in advance

I was getting bored with no competition

Lead Programmer/Director
Powerful Collision DLL for DBPro and DBC: http://www.nuclearglory.com

Login to post a reply

Server time is: 2024-05-20 04:31:48
Your offset time is: 2024-05-20 04:31:48