keihaswarrior said:omg PER-POLY HIT detection.
I am so sick of the ****ing coke can. When I shoot someone's leg, I want to hit it, when I shoot over their shoulder I want to miss them...
Well, if it had per-poly detection, there wouldn't be a question of how "tight" it was. Except for lag issues, you would always hit the target exactly how you aimed with per-poly detection because it would be able to determine the exact polygon on the player model that the bullet hit.Logan6 said:And a little tighter to boot.
yurch said:Aiming/shooting
1 Camera 'bob' is extreme. Practically to the point where it's a fix-all. It isn't nessesary. Really.
yurch said:2 A pure weapon vection system. (of which RAv2 had an incomplete version) Yes, that means: No more conefire for movement with or without hipping. The weapon shoots where it points. But before you get all excited, that also means that the vection system is mature enough that the players cannot rely on this for accuracy. This would be something that would require an extreme amount of work to be any good.
yurch said:3 Recoil from an improperly supported firing position should act as such. Possibly altering not on randomization but on the player's input. The conefire described #1 can actually help auto-sprayers achieve a hit - the cone increases when they move, giving thier gun a sort of delayed shotgun-pattern.
yurch said:4 Bonuses from crouch/prone/breathing on accuracy, sensitivity or recoil should take TIME to come into play. Bonus should also eventually occur while standing still.
yurch said:5 Scopes are merely optics attached to rifles. There is no reason the rifle should suddenly handle in a vastly different manner.
yurch said:6 Players move and act in unnatural manners. "Strafing" and jumping in particular, especially the ease of which the player changes directions. This is why bunnyhopping/circlestrafing is so easily abused - the target becomes a series of points the shooter has to interpolate rather than a smoothly moving human. This is both a fault of the animations and the response time of the avatar.
yurch said:9 The crouched player seems to sit lower than his camera. Leaning players shoot out of thier foreheads.
yurch said:10 Most people I try to introduce to the game note that the stanima and run/walk/jog system is pretty akward.
yurch said:11 Movement Sounds! Movement Sounds! Movement Sounds! I assure you, the first thing you will notice if someone else is in the same house/room/hallway as you and is moving, you may be more likely to hear them first than to see them. RvS handles this pretty well in its demo, at least. Moving directly through bushes should make a racket and is often PLAINLY visible.
yurch said:13 Grenades need to stop being treated as nuclear devices. If you want to kill something for certain, shoot it directly.
yurch said:14 I'm under the impression that armour is intended to save your life, not keep you fighting immediately after the fact. Pistol rounds, maybe. PSG rounds, no.
yurch said:15 If no FOV shifting functions exist, reduce ballistics to the percieved distance rather than 'actual'. Would put it at the 55% level or so.
yurch said:16 Some weapons should be inherently inaccurate, even at 'close' ranges. P90 in particular. Many rounds become quite a bit more ineffective when subsonic.
DEFkon said:I'd pretty much agree with yurch on everything. Overall i think what INF really needs is "simplification". At least on the user end. The avatar should feel to the player as an extension of their own body. Going from a kneeling position to a full upright sprint, back to a kneeling position shouldn't be as complicated as opreating a manual transmition. Controls need to be streamlined so they feel intuitive to a stranger to the game.
Mov. 1. double tap is discussed more in the other thread including reasons why not really usefull. The sprint plus strafe ONLY moves you a bit to the left and right... you can do the same with the mouse but the strafe keys are easier to use for certain situations. You do not strafe-sprint or whatever.-RoMe-infers said:Movement:
1. Walk/jog.. double tap forward key for sprinting. When sprinting you should not be able to strafe and can only go forward. (not the case in 2.9).
Aim/Wounds:
3. Slight weapon inertia.
4. When shot, you cannot shoot back during at least 1 or 2 second (because you're hurt!), meaning that the first to hit is most of the time the winner.
6. Local damage, you dont die by being shot in feet, even 100 times.
Coms:
1. Voicecom.
2. A few hand signals is a really good idea.
3. Ability to talk just to the person you are aiming at, if close enough, by pressing a key...yes i'm dreaming .
4. Ability to talk to everybody on the map, no matter the team, by pressing a special key ( for afk and such things).
Can someone tell me what this is? Thx.-RoMe-infers said:5. Yurch's prone system.
It's not? Since when are the weapon positions themselves based on pixels?jayhova said:In addition to this it isn't possible for INF to simulate very small (less that one pixel) changes in sight alignment.
What I'm referring to there is that the 5.7mm 32 grain AP needlebullets that thing fires have a tendancy to scatter every which way in midair to a greater extent than even most pistol ammunition... not even counting brush effects or ricochets there. Controllable recoil or not, that's not a gun I want someone supporting me with.The P90 is not all that inacurate in it's intended application that is as a fully auto weapon. It's inherent inaccuracies are made up in controlabilty.
The real culprits here are the pistols. They are far to accurate IMHO.
There's no reason to 'magnify' it, that error comes naturally from having the sights so close together on a short barrel.OICW said:Pistol sight-misalignment should be magnified more than other weapons IMO.
yurch said:It's not? Since when are the weapon positions themselves based on pixels?
yurch said:What I'm referring to there is that the 5.7mm 32 grain AP needlebullets that thing fires have a tendancy to scatter every which way in midair to a greater extent than even most pistol ammunition... not even counting brush effects or ricochets there. Controllable recoil or not, that's not a gun I want someone supporting me with.
yurch said:Pistols for the most part are 'inaccurate' also because of thier short sighting system and lack of a stock. That should be reflected by the vection handling before the individual ammunition accuracy comes into play.
First of all, a vection system as described is not merely sight misalignment. There are a finite possible number of rotations in UT for any given axis; All that needs to be done is make the weapon's rotations somewhat independant of the player's rotation. 'Misalignment' is the result of simple observation of these rotations. You aren't changing what is already there.jayhova said:Maybe I should rephrase that. I should have said that it would not possible to accuratly represent the missalingments that occur with pistols due to the limitations of pixel based technology. Is that better?
The 5.7mm round already has a 6 inch spread at 50 meters. I'm not too impressed.However the round should be fine if used within 150m.
Greatly increasing weapon bob in some way for 1-2 seconds after getting hit then reducing it to "normal" (considering you're injured, not "healthy normal") does sound like a decently good idea.Hurin said:-RoMe-infers:
4. When shot, you cannot shoot back during at least 1 or 2 second (because you're hurt!), meaning that the first to hit is most of the time the winner.
I cant say i agree, even though it will drastically effect your aim, make it shaky, you should still be able to pull the trigger. So maybe if you are shot increase weapon bob by a certain high amount. This could also apply to severe wounds, the lower your stamina or health the more the weapon should bob in your hand.
-Skillz