5.8 bullet discussion

[Violet CLM]

Fixed bug of clients sometimes bouncing off bullets without getting hurt (client-side fix).
Please discuss this fix from 5.8, in the normal case when all players are running 5.8 and the game is started, in this thread.

I'm assuming the possible outcomes of this discussion are a) keep the fix, b) revert the fix, or c) make it optional under a new boolean server-wide setting, but who knows really.

[cooba]

Please keep in mind that this fix was implemented after reports of "many players begin to bounce off ammo a hell of a lot more"

[Violet CLM]

As of last night, the consensus seemed to be in favor of "b) revert the fix," in the form of a publicly downloadable dll with all other 5.8 features intact but all bullet-related coded reverted. Some people expressed an interest in testing that for an indefinite period, so, uh, here you go. (https://get.jj2.plus/misc_builds/plus_58_revertbullets.zip) You'll notice there's a lot of build information on display, that's going to make it much easier for us to keep track of testing reports and also serve as a helpful reminder to players that this isn't the real release.

[Zah dud4h]

Ok so i decided to test this locally due to how the servers rolled back to 5.7

we have 2 things that modify the speed of the player (or push the player):
Flips and Bounces

Now from what i understand
Let's separate the terms "Colliding" and "Hurt",thus implying that colliding does not necessarily mean you are hurt

A. Flips are the feedback speed whenever you get HURT by a bullet
B. Bounces are the feedback speed whenever you COLLIDE with a bullet on your screen



What does /bulletbouncing ON do:

On bounce behavior :
- Host can get bounced by ammo
- The rest of the players cannot get bounced by ammo

(Disclaimer: i only tested this with stopped game or hacked the game so that the client thinks friendlyFire is on in a started game so that the client colliding with the bullet does not get hurt but still thinks it as an enemy bullet,both methods share the same result,i tried the friendlyFire hack method too in the bullet-reverted version and well,the allies do bounce on ammo there)

On flip behavior:
- It works as intended for both sides
- It knocks you back and pushes you upwards



What does /bulletbouncing OFF do:

On bounce behavior :
- Host ALSO cannot get bounced by ammo
- The rest of the players cannot get bounced by ammo

On flip behavior:
- It works as intended for both sides
- It just knocks you back without pushing you upwards

I know this was all disclosed but a fresh summary is not a bad thing

I thought about 2 options:
Modify the BulletBouncing command so that the rest of the players actually bounce on ammo if the setting is turned on
OR
Split the command into 2 commands,/bulletbouncing for bouncing on bullets and /bulletflipping for flipping on bullets (probably this was the so-called /miracle idea)

Another issue with this,as stated by Necrolyte,players that join with 5.8 in a server that is hosted in a older version,have bulletbouncing disabled by default (considering how flips behave the same as the command was turned off)

I hope this was helpful in finding a solution to this problem and i hope i was clear enough in my "analysis"
If not,please clarify what was less detailed/correct about my post

[Violet CLM]

Please discuss this fix from 5.8, in the normal case when all players are running 5.8 and the game is started, in this thread.
[Disclaimer: i only tested this with stopped game
players that join with 5.8 in a server that is hosted in a older version
hacked the game
None of these data are helpful.

[master sven]

A7med and me just tested the new 5.8 version in a new 5.8 server and couldn't detect any side-effects. So it seems fine to upgrade to this version. May I add that in the future I would recommend testing more profoundly than we just did and testing each fix one by one before the next one gets stacked so its easier to find out which fix causes a side-effect. Thanks for creating this new 5.8 version!

[Seren]

May I add that in the future I would recommend testing more profoundly than we just did and testing each fix one by one before the next one gets stacked so its easier to find out which fix causes a side-effect.
As far as we know, there are no "side effects" to this bug fix - everything reported so far suggests that it works exactly as intended (aside from some inconsistencies during stopped games - but that's not the source of complaints). The change was tested in multiplayer games for many hours, but I'm afraid no amount of testing on our side could've reached different conclusions, because everyone on our team who had tested the change, has found it to have positive impact. What happened here was simply a difference of opinion. However, in order to collect a wider range of opinions, we are currently recruiting testers (https://www.jazz2online.com/jcf/showthread.php?t=20461) - it's something worth considering if you want to have impact on the direction of development.

[master sven]

Thats why I was referring to the future. I hope many people sign up as testers and will be able to test every fix step by step so there will be less room for unwanted side-effects that are perhaps hard to trace back and satisfaction of the active playerbase about any updates is more or less secured.