Jumping techniques. Deeper explanation.

Started by Juice
03/10/2013 17:30
Page 2 of 2
ackE05/10/2013 07:29(edited 05/10/2013 07:31)

The code is taken from the half-life sdk and Valve moved it to Guthub these days.
https://github.com/ValveSoftware/halflife

And here is the file this post is refering to:
https://github.com/ValveSoftware/halflife/blob/master/pm_shared/pm_shared.c

Arion05/10/2013 10:23

Thank you guys :) Keep it going.

Juice05/10/2013 12:49(edited 05/10/2013 12:50)

Originally posted by ackE
The code is taken from the half-life sdk and Valve moved it to Guthub these days.
https://github.com/ValveSoftware/halflife

And here is the file this post is refering to:
https://github.com/ValveSoftware/halflife/blob/master/pm_shared/pm_shared.c

If you look up closer you may notice some changes between HLSDK and what I posted ;) There's CSSDK project on AlliedMods: https://amxmodx.org/showthread.php?s=c16187bba6500802d82db53217494900&t=123097 The project has started from scratch some time ago so there's no CS code left on github. I got pm_shared.c with help of wayback machine :D

spr1n mapper05/10/2013 17:36

Good job there explaining those things! I have read all you wrote about jumpbugs very carefully. =)

Nonetheless I have to point out that you made several mistakes yourself making those assumptions in jumpbug section.
First of all only one paragraph is not quite correct (about CS engine), following remarks and FPS explanation are certain and very important.
Secondly the jump after performing successful jb is not as high as jump with normal bhop, but as regular jump (45/63ducked)
Then slowdown that kicks in while jumpbug works in a different pattern rather as exceeding weapon_max_speed*1.2 . It works with very low velocity and in a different scale comparing to regular bhop from what I can tell.
And last, your luck reference and according calculations are quite subjective as you ignore height difference and fps consistency, thus its misleading in that context. However that calculation worth to be a part of "inconstant fps" explanation in KZ-Tutorial.

cheers!

Juice05/10/2013 17:56(edited 05/10/2013 18:04)

Originally posted by spr1n
jump after performing successful jb is not as high as jump with normal bhop, but as regular jump (45/63ducked)

As I mentioned in the very beginning, the text is not mine(and it needs some correction, I am working on it)
actually jump height depens if you have fuser2 parameter or not, so if you jump from height enough to clear that parameter, then the height will be 45/63.
Thanks for pointing out that mistake!. Updated first post :)

Originally posted by spr1n
Then slowdown that kicks in while jumpbug works in a different pattern rather as exceeding weapon_max_speed*1.2 . It works with very low velocity and in a different scale comparing to regular bhop from what I can tell.

This is wrong. In PM_CatacorizePosition we can see that PM_AddToTouched is called if there's something 2 units below us - that modifies our velocity, forcing our speed to ~150. I'm not sure but if you fly to the ground at very acute angle, that slowdown is skipped and our if speed is more than weapon_max_speed*1.2, then it crops down by PM_PreventMegaBunnyJumping. (Happened to me on awp_rooftops, getting a speed boost from redroof then performing a jb at the bottom of the map :D)

spr1n mapper05/10/2013 19:11

Originally posted by Juice ._.
Originally posted by spr1n
Then slowdown that kicks in while jumpbug works in a different pattern rather as exceeding weapon_max_speed*1.2 . It works with very low velocity and in a different scale comparing to regular bhop from what I can tell.

This is wrong. In PM_CatacorizePosition we can see that PM_AddToTouched is called if there's something 2 units below us - that modifies our velocity, forcing our speed to ~150. I'm not sure but if you fly to the ground at very acute angle, that slowdown is skipped and our if speed is more than weapon_max_speed*1.2, then it crops down by PM_PreventMegaBunnyJumping. (Happened to me on awp_rooftops, getting a speed boost from redroof then performing a jb at the bottom of the map :D)

You got me wrong. Im saying slowdown is much stronger with jumpbug compared to bhop and it even occurs with lower speed than weapon_max_speed*1.2 . So there must be something else that affects this slowdown additionally. By the way I assume such speed boosts that you refer to, are just anomalies caused by server lags or connection, happened to me few times, but never in a specific pattern.

ackE05/10/2013 20:38

Originally posted by Juice ._.
Originally posted by ackE
The code is taken from the half-life sdk and Valve moved it to Guthub these days.
https://github.com/ValveSoftware/halflife

And here is the file this post is refering to:
https://github.com/ValveSoftware/halflife/blob/master/pm_shared/pm_shared.c

If you look up closer you may notice some changes between HLSDK and what I posted ;) There's CSSDK project on AlliedMods: https://amxmodx.org/showthread.php?s=c16187bba6500802d82db53217494900&t=123097 The project has started from scratch some time ago so there's no CS code left on github. I got pm_shared.c with help of wayback machine :D


Yes but all of Half-Life was recompiled a few months back bringing it to steam pipe and mac/linux support so they probably used the newer files for that, which should be in the latest builds. But that is just me guessing.
Great job on this none the less!

WeMeRA11/10/2013 02:57

intresting, thanks for sharing

rawerecord holder Tournament #111/10/2013 03:03

Very interesting! I read it all, but there are things that my mind couldn't understand!

surfrock28/12/2013 19:58

now take kz_ep_gigablock_b01 ^^

volte30/01/2014 15:58

Hey juice, intresting read indeed!

Can you also go into some deeper explanation of slidebugs? I mean I can imagine it works in somewhat a similar way as EB and JB, but I can't quite wrap my head around it :P

Cobrex mapper Tournament #220/05/2014 04:08

NumB used to say something about a bug that could destroy the jumping scene. His example was something like : kzno_xtremeblock could be done in 3 seconds. I can't quite remember what he said 100% but it was something like this.

Juice, any idea on what he was talking about?
How come NumB and spr1n were jumpbug gods and masters of understanding the game. Coincidence, hah? They found a way to do them easier?

iNsprate moviemaker20/05/2014 04:48

interesting indeed.

Juice21/05/2014 01:30

Originally posted by Cobrex
NumB used to say something about a bug that could destroy the jumping scene. His example was something like : kzno_xtremeblock could be done in 3 seconds. I can't quite remember what he said 100% but it was something like this.

Juice, any idea on what he was talking about?

please don't start it again

many years have passed and still there's no proof about this bug actually exist. I believe that edj#2 is a myth as well as 18-units land bug :E

the only technique that provides 'unlimited' speed is that you unduck on slanted walk-able surfaces, watch video here: http://www.youtube.com/watch?v=6At8CoNpWWY

Page 2 of 2

Please log in to post replies.