[KoMiKoZa]

Originally Posted by Xutax_Kamay Sorry I didn't even see your post, and yea sure, would be awesome Thank you

Oi,

Our conversation was recently cleaned up by moderators as it had been derailing (large ad hominems by the guy and others joining in).

Basically, you won't find anything useful in it but I'll post the version of what remains and what I remember.

Spoiler

It goes like this:

- I post the plugin, say it's pretty interesting and others should check it out.

- dr.lex joins in and asks "Why would you think it to be a bug? Source engine has such things as 'bullet spread', it obviously influences your acuracy as you move. I personally think it's more of a cheat than a fix, and it's basically useless."

- I reply: "It's not about bullet spread, please, read the description [I translated it in Russian], the reason why this happens has been thoroughly explained by the author".

- dr.lex goes: "Always wondered how people can have such a high self-esteem thinking they know better than the devs who made the engine". He also reminded me that there's "such thing as VAC and I'm not hinting at anything, but you've been warned".

- dr.lex adds: "I've run some tests on a test server. I'm writing this so people could decide on their own if they need this or not. Source engines have such cvars as:

sv_unlag - Lagcompensation (1 on /0 off), if you disable it, it'll be the same as this plugin, even better.
sv_maxunlag - Amount of time for storing players' positons (max. 1.0).

For example, 0.2 = maximum lagcompensation will be 200ms, if 0.6 - 600ms etc.

"Case closed" (c).

- I'm telling him to elaborate on what tests he ran and if he could show the results to public before advising anything and "closing cases". I said that his statements and explanations don't explain me anything, and I find it hard to believe what he says based on his already made comments about spread, self-esteem and lagcomp.

- He stops responding in this particular topic.

- Some other guy obviously with some experience joins in and says: "That's interesting, although this plugin wouldn't be very effective in a game like L4D2, maybe only on lower ticks.", then they started arguing with each other.

- I post in the topic again saying he should address my request for proofs of his claims instead of arguing with another person about basically the same topic. I remind him that:

1. He didn't understand what the plugin was for.
2. He told me pointless stuff like "spread and movement", and how obvious it was. I asked him "Do you really think the author would make such 'cheat' for himself not knowing how spread works in Source?".
3. He "ran some tests" and came to a conclusion that disabling lagcomp will yield same results where as the author had explicitly stated "it's not just about lagcomp" (I even gave him your translated reply).
4. He decides to not reply to a lot of what I tell him and is basically evading it, only replying to what he wants.

- dr.lex makes such comments as "It's elementary knowledge, I'm tired of explaining such stuff to dummies. The only thing that really helps with hit registration is higher tickrate, the rest is "dumb stories made by schkids". He says he's not gonna post anything and will stop posting completely, "smart people will understand it and make right conclusions".

At this point I get pretty worked up and tell him that he's given zero useful information and is unable to hold a discussion at all. Moderator steps in and deletes a lot of info, so I have to give you this recap.

I wouldn't take the guy seriously. I even pointed out he behaves pretty inadequately and I couldn't take him seriously because of that.

As to your question, I haven't tested this in particular because I'd hoped the guy would bring proofs, but I guess I'll have to do that now. I'll run tests on my own and add to that topic and this one as well. Not sure when I'll have time to test this specifically.

[Xutax_Kamay]

Originally Posted by KoMiKoZa Oi, Our conversation was recently cleaned up by moderators as it had been derailing (large ad hominems by the guy and others joining in). Basically, you won't find anything useful in it but I'll post the version of what remains and what I remember. Spoiler It goes like this: - I post the plugin, say it's pretty interesting and others should check it out. - dr.lex joins in and asks "Why would you think it to be a bug? Source engine has such things as 'bullet spread', it obviously influences your acuracy as you move. I personally think it's more of a cheat than a fix, and it's basically useless." - I reply: "It's not about bullet spread, please, read the description [I translated it in Russian], the reason why this happens has been thoroughly explained by the author". - dr.lex goes: "Always wondered how people can have such a high self-esteem thinking they know better than the devs who made the engine". He also reminded me that there's "such thing as VAC and I'm not hinting at anything, but you've been warned". - dr.lex adds: "I've run some tests on a test server. I'm writing this so people could decide on their own if they need this or not. Source engines have such cvars as: sv_unlag - Lagcompensation (1 on /0 off), if you disable it, it'll be the same as this plugin, even better. sv_maxunlag - Amount of time for storing players' positons (max. 1.0). For example, 0.2 = maximum lagcompensation will be 200ms, if 0.6 - 600ms etc. "Case closed" (c). - I'm telling him to elaborate on what tests he ran and if he could show the results to public before advising anything and "closing cases". I said that his statements and explanations don't explain me anything, and I find it hard to believe what he says based on his already made comments about spread, self-esteem and lagcomp. - He stops responding in this particular topic. - Some other guy obviously with some experience joins in and says: "That's interesting, although this plugin wouldn't be very effective in a game like L4D2, maybe only on lower ticks.", then they started arguing with each other. - I post in the topic again saying he should address my request for proofs of his claims instead of arguing with another person about basically the same topic. I remind him that: 1. He didn't understand what the plugin was for. 2. He told me pointless stuff like "spread and movement", and how obvious it was. I asked him "Do you really think the author would make such 'cheat' for himself not knowing how spread works in Source?". 3. He "ran some tests" and came to a conclusion that disabling lagcomp will yield same results where as the author had explicitly stated "it's not just about lagcomp" (I even gave him your translated reply). 4. He decides to not reply to a lot of what I tell him and is basically evading it, only replying to what he wants. - dr.lex makes such comments as "It's elementary knowledge, I'm tired of explaining such stuff to dummies. The only thing that really helps with hit registration is higher tickrate, the rest is "dumb stories made by schkids". He says he's not gonna post anything and will stop posting completely, "smart people will understand it and make right conclusions". At this point I get pretty worked up and tell him that he's given zero useful information and is unable to hold a discussion at all. Moderator steps in and deletes a lot of info, so I have to give you this recap. I wouldn't take the guy seriously. I even pointed out he behaves pretty inadequately and I couldn't take him seriously because of that. As to your question, I haven't tested this in particular because I'd hoped the guy would bring proofs, but I guess I'll have to do that now. I'll run tests on my own and add to that topic and this one as well. Not sure when I'll have time to test this specifically.

Well I guess it took you time to write it all that up so thank you.

I'll test it myself, but I'm pretty sure that it doesn't do anything except disabling lag compensation for others players that you're shooting against, don't waste much too your time on it.

I can agree that in the past I was really immature sometimes, but you know everything change with time

[backwards]

After reviewing the source code I think you may have messed up the logic.

https://github.com/XutaxKamay/firebu...letsfix.sp#L84

You are obtaining the position from the most recent user command processed on the server. This will be the real time position the player currently is at. When using lag compensation all the targets along with the attacker should be re-winded back in time to the positions they were when the shoot command was issued. The usercmd packet includes a field "tick_count" which is the data used to sync the client-server for prediction.

The position you really want to teleport the player to would be the position the player was at in the issued tick_count usercmd + 1. (If the firebullets pos is really using old player pos) I might be missing something though so correct me if i'm wrong.

[MAMAC]

Ya plugin mess even more my server... so or i have a problem on server or plugin have a problem.

[Xutax_Kamay]

Originally Posted by MAMAC Ya plugin mess even more my server... so or i have a problem on server or plugin have a problem.

Can you elaborate more?

What game?

A video would be cool to see.

Originally Posted by 1337norway After reviewing the source code I think you may have messed up the logic. https://github.com/XutaxKamay/firebu...letsfix.sp#L84 You are obtaining the position from the most recent user command processed on the server. This will be the real time position the player currently is at. When using lag compensation all the targets along with the attacker should be re-winded back in time to the positions they were when the shoot command was issued. The usercmd packet includes a field "tick_count" which is the data used to sync the client-server for prediction. The position you really want to teleport the player to would be the position the player was at in the issued tick_count usercmd + 1. (If the firebullets pos is really using old player pos) I might be missing something though so correct me if i'm wrong.

We discussed about it on steam I guess it's not anymore needed to talk about it here o:

[luidgi]

Can you help in setting up the plugin?
The plugin has no errors, but everyone says that when shooting at a character’s model, the bullet flies by

[VerMon]

Why don't you keep the old eye angle as well?
Because now there is used old eye location but current direction of view.
It also plays a role, and a big one, especially at long distances.
Also, maybe will be better not to shift the tick when you attacked using knife?

Hook FireBullets function and pass old arguments instead current one by changing m_vecDirShooting and m_vecSrc.
Because m_vecSrc itself taken from Weapon_ShootPosition, and m_vecDirShooting is calculated in place.
So the is only way.

[VerMon]

Originally Posted by 1337norway After reviewing the source code I think you may have messed up the logic. https://github.com/XutaxKamay/firebu...letsfix.sp#L84 You are obtaining the position from the most recent user command processed on the server. This will be the real time position the player currently is at. When using lag compensation all the targets along with the attacker should be re-winded back in time to the positions they were when the shoot command was issued. The usercmd packet includes a field "tick_count" which is the data used to sync the client-server for prediction. The position you really want to teleport the player to would be the position the player was at in the issued tick_count usercmd + 1. (If the firebullets pos is really using old player pos) I might be missing something though so correct me if i'm wrong.

void CLagCompensationManager::StartLagCompensation should not happen for the attacker, It is real time for the player but not for others, we should only rewind them
// Don't lag compensate yourself you loser...
if ( player == pPlayer )
{
continue;
}

[Xutax_Kamay]

Originally Posted by VerMon Why don't you keep the old eye angle as well? Because now there is used old eye location but current direction of view. It also plays a role, and a big one, especially at long distances. Also, maybe will be better not to shift the tick when you attacked using knife? Hook FireBullets function and pass old arguments instead current one by changing m_vecDirShooting and m_vecSrc. Because m_vecSrc itself taken from Weapon_ShootPosition, and m_vecDirShooting is calculated in place. So the is only way.

I do not suggest to anyone doing this.

This bug occurs only for moving data (localplayer's data, the most noticeable, position one) that are processed by

CGameMovement::ProcessMovement

and used in events in PostThink +/-.


I'll try to make this more clear, for example:

- Once

CGameMovement::ProcessMovement

has processed the computation of the player's origin, your position is shifted by 1 tick:

1) Because of the rendering process (in order to smooth out the game) there is 1 tick of interpolation in the client's code (https://github.com/ValveSoftware/sou...tity.cpp#L2822) from predicted data.

2) But even if you remove interpolation, the bug will still occurs and still take the next position to shoot because the predicted data hasn't been saved yet for rendering since it is still in computation (stuck in RunCommand).

Which results to shoots from your next position and not current rendering one.

3) Since the viewangles are not modified in any ways by ProcessMovement (excepted teleportation) and sent by the user directly computed in CreateMove (ExtraMouseSample), it is not delayed.

[DevilGames21]

Hello!
Can this plugin fix the knife hit registration ? When someone knife an other player he hears the sound from the knife but the other player doesnt take damage.

Thanks.