RuneStorm http://www.runestorm.com/forums/ |
|
need to test. "proto type" http://www.runestorm.com/forums/viewtopic.php?f=69&t=76874 |
Page 1 of 1 |
Author: | mzoltan22 [ Wed Mar 16, 2011 3:01 pm ] |
Post subject: | need to test. "proto type" |
ok, so i made a bullet handler, so now bullets have a curved path, and will require some time to get to the target. this bullets also decelerate, and upon penetration, they change direction. If the angle is within a certain bound, bullet will ricochet. I dont use projectiles. On more details, see Bullethandler.uc, and contact me, if you have found something bad or don't understand something (regarding features i mentioned). You should test it out, if you have a ddr2 system you should have no problem. If it works as i woild like it to work, and it works you well in offline, or on your lan, it should perform almost identical to the stock BallisticMinigun, on network. the test weapon is a replica of the minigun... There is no tracer for bullets (justa commmented out Spawn Ammo pack to see trajectory after shot). For some reason, exit holes don't have decals, but its not a priority (or they got removed pretty fast?). link: http://qshare.com/get/951802/BallisticTactical.zip.html |
Author: | Sgt. Kelly [ Wed Mar 16, 2011 4:39 pm ] |
Post subject: | Re: need to test. "proto type" |
If it's accommodating for bullet drop and travel time, then it's not really hitscan. Hitscan is when you shoot and it hits the wall immediately. These sound more like projectiles. I would recommend looking at the code of the Tank Master V tank made by Black Eagle, it uses the bullet style you're talking about and works with all the proper effects. When I get some free time I'll check this out though, it sounds like an interesting idea. ![]() |
Author: | mzoltan22 [ Thu Mar 17, 2011 4:28 am ] |
Post subject: | Re: need to test. "proto type" |
precisely, it is a flying hitscan wich changes course according to deceleration and outer forces (now mainly just static values). first, i would like to make this so, that the array wich contain bullets does not get syncronised between client and server, otherwise if it does geet syncronised it would lag the network. since i don't exectly know network behaviour on ut 2004 and also i don't have the means to test this online (yet) i would like someone to test it as described, so if that lags, but otherwise ballsitic without that addition wouldn't lag, that means that probably those arrays get syncronised. after i could make it not to syncronise, i would like to make a similar system based on projectiles and compare performance. the projectile one is actually quite easy to do. afterwards i would pick the better performancing. i expect that the projectile based would be less efficient in the end, thats why i started with this "flying" hitscan. |
Author: | mzoltan22 [ Fri Apr 08, 2011 2:17 pm ] |
Post subject: | Re: need to test. "proto type" |
so, mostly penetration and ricochet is done from some time. Tough, need some way to manually move an emitter or something like that along the path of bullet, so now, the path is not realy visible. Please test it online and otherwise. Btw, i gonna abandon this project on unreal script, since its slow, and i guess will do try to make it happen on Ogre3d and Bullet for rendering and Physics. There is a link: http://qshare.com/get/951612/BallisticTactical.zip.html Use This code if you think its usefull, for somtehing, but its incomplete, and due to script slow on pcs with ddr1 ram (tough, these are special cases), tough other things should be straightforward (its commented up i think good enough). There is one case wich is not handled: when bullet hits surface under a very small angle, bullet is removed. It could be handled so, that, it will be a ricochet most likely, but after setting the new direction vector, need to check if outerforce vector gona turn this direction vector back into the previously hit surface. If yes, then it should go along the surface and untill it flies off of this surface it should not use gravity as a vector (so it won't cause collisions over and over again and again, without movement), but rahter as an additional decelerative component. like you could check sometimes, with the reverse of the prevous hitnormal vector, if there is still a surface near the "glideing" bullet path. If you can only try it offline, or on lan, do it please, and report the results. |
Author: | Bjossi [ Fri Apr 08, 2011 5:58 pm ] |
Post subject: | Re: need to test. "proto type" |
mzoltan22 wrote: precisely, it is a flying hitscan wich changes course according to deceleration and outer forces (now mainly just static values). You don't understand. Hitscan is a vector, it does not have velocity or de/acceleration, it is only a straight line that cuts off at a defined distance. You can also delay its effect by a specified length of time, which BW weapons use to imitate bullets having velocity. The idea of turning bullets into real projectiles is not new but has always interested me. A good while ago Xavious made a test CYLO that fired bullet projectiles. It worked pretty well for the most part (only tested offline) except it was far, FAR too good at headshots. We never figured that one out and ended up forgetting about the test project. |
Author: | Black Eagle [ Sat Apr 09, 2011 3:16 am ] |
Post subject: | Re: need to test. "proto type" |
Its possible, and perhaps more civilized to not use projectiles, but instead do several progressive traces to mimic the path, this could be done using a minimum of 2 traces, as opposed to the usual 1, unless the HitLocation is close enough that it doesn't need to do anymore... ![]() |
Author: | mzoltan22 [ Sat Apr 09, 2011 3:41 am ] |
Post subject: | Re: need to test. "proto type" |
Bjossi wrote: mzoltan22 wrote: precisely, it is a flying hitscan wich changes course according to deceleration and outer forces (now mainly just static values). You don't understand. Hitscan is a vector, it does not have velocity or de/acceleration, it is only a straight line that cuts off at a defined distance. You can also delay its effect by a specified length of time, which BW weapons use to imitate bullets having velocity. I soon will EDIT the first post to avoid any further confusion. *edited. |
Author: | Blade sword [ Sat Apr 09, 2011 1:34 pm ] |
Post subject: | Re: need to test. "proto type" |
I will try your mod soon, since I'm interested of that kind of ballistics simulation. First thank you for the mutator and I will try this, by hoping some other people does too. |
Author: | mzoltan22 [ Sat Apr 09, 2011 3:05 pm ] |
Post subject: | Re: need to test. "proto type" |
Blade sword wrote: I will try your mod soon, since I'm interested of that kind of ballistics simulation. First thank you for the mutator and I will try this, by hoping some other people does too. Thank you. However, its not finished, the ballistic calculations are finished, but accessory calculations for appliyng damage changing some bullet parameters upon impact (so to simulate bullet deformation/tumbleing), and calculating the angle thresholds for those things are just constant numbers now (they should be no much trouble). There is one thing i would like to add, but i don't know how to yet: give spin to bullets, and so depending on that, when they impact under an angle, they should not just bend towards (or away from, in case of exit) hitnormal streight, this new direction vector should also roll around hitnormal (in the moment of entering, or exiting material). Well, i could use sin and cosine but i would like to make it without them (if possible). |
Author: | Blade sword [ Mon Apr 11, 2011 4:11 pm ] |
Post subject: | Re: need to test. "proto type" |
I wanted to try your mutator except I don't figure how to use it. I moved the u file to my system folder, but I think there is something else needed. |
Author: | Bjossi [ Mon Apr 11, 2011 5:22 pm ] |
Post subject: | Re: need to test. "proto type" |
An ucl file needs to be there as well. |
Author: | Blade sword [ Mon Apr 11, 2011 5:46 pm ] |
Post subject: | Re: need to test. "proto type" |
Yeah |
Author: | Sgt. Kelly [ Mon Apr 11, 2011 7:39 pm ] |
Post subject: | Re: need to test. "proto type" |
Blade sword, this isn't a chat room. One word responses like 'yeah' don't add anything to the conversation. ![]() |
Author: | {ABA}Worlock [ Mon Apr 11, 2011 8:53 pm ] |
Post subject: | Re: need to test. "proto type" |
Will this allow the user to skip/ricochet a projectile to hit the target and do damage, depending on the range. |
Author: | mzoltan22 [ Tue Apr 12, 2011 1:33 am ] |
Post subject: | Re: need to test. "proto type" |
oops, i forgot about that file. I will reupload with the .ucl (as soon as i got home). To answer Warlock-s question. In some cases when bullet can ricochet and it hits people it might hurt them. The plan is, that bullets will have no such thing as damage defined precisely. damage to material (be it armor or living things) will be calculated, from energy given off by bullet, in particular, energy given off under a sertain amount of time when it is moving inside the target, and also, mass loss will positively influence damage. But this things will be calculated from 7 things: bullet mass, bullet speed, bullet integrity (how prone it is to fracturing, loosing weight), bullet slowdown coefficient, angle of impact (this can change slowdown coefficient), volume integrity, and volume friction. Additionally, there is armor for for pawns, wich are not a discrete material, but only an assigned value to that pawn, so armor will need to be modified to have thickness, friction and integrity values for areas of body, so the hit can calculate more realistic armors. So this way, for example, if you have a high integrity bullet, and it hits you without armor, and flesh and bone has a quite low integrity coefficient, bullet will lose not much energy on impact, and will pass right through the body without doing much damage (if we are talking about genetically modified sci-fi humans, who, have the potential falws of humans fixed), or it can hit an artery, so eventually you will bleed out, but not instantly (if we have simple "stock" humans or things like that). On the other hand, if the same bullet hits a high integrity armor, it will deform/tumble on impact, and if still was able to penetrate armor it will damage tissue more. With a good enough processor (meaning good frequency) and a ddr2 system it will still run just fine with those calculations, But unreal script being very slow, it will be crap for ddr1 systems below 1.8 ghz processors when 5 guy with miniguns keep shooting on 3600 rpm. tough, if minigun bullets would be set up so, that they won't be able to penetrate much (so avoiding too much traces and penetration/ricochet calculations to run in one tick cycle for one bullet), it still should be ok. Problem is that recently, i don't have much time for this. EDIT* I've reuploaded with the .u and .ucl file, link in the first post. Please give some feedbacks. If functions well for you, i continue finishing the features of this bullet thing. |
Page 1 of 1 | All times are UTC - 6 hours |
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group http://www.phpbb.com/ |