8 replies to this topic

[DeadlyNerd]

Ghost heat does squat in a game of whack-a-mole. It's easily handled by doing a high heat alpha strike and backing away into the confines of one's cave. Pinpoint hit is still there, and it's still ruining the game.

I propose "ghost" convergence instead.

• each weapon gets a "targeting computer load factor"
  
• when weapons are fired, their load factors are summarized and, if they surpass the load threshold of the mech targeting computer, weapons are fired with decreased convergence.
  
• load resets every 0.5sec
  
• convergence loss is differently calculated for arms and torsos/head
  
• convergence loss amounts for arms - rest of the body are different for each mech
  
• convergence loss is applied separately for each body part
  
• all mechs have the same targeting computer load threshold

-there could be a bar in the mechlab displaying the load for each weapon group
-load a computer can take could be increased through mech/pilot skills

Why different amounts? Some mechs are built with fixated arms(jager), some mechs are built for torso action(banshee), some only have a certain part specialized(hunchback) etc. They're supposed to excel in those areas, and this way they're allowed to.

Why applied separately? This means that mechs meant to boat in a certain body part will still be able to, yet you won't be able to shoot your left+right arm+left torso without suffering unpredictable convergence loss.

[meteorol]

There have been countless ideas on how to fix the high pinpoint damage issue.
If i recall correctly, your idea (or something very similar in regards to target computer load) has been here in a huge thread (with eveything calculated and whatnot) like a year ago.

The thing is: there have been many great ideas how to solve the pinpoint issue. The sad part is, they are WAY out of PGIs reach. They need months to do even the smallest things. Recoding the whole system of how weapons work will not happen. Never. What we will see are minor balance tweaks.

If anything, they will make ACs burst fire weapons and PPCs splash damage, but this is also unlikely. All those ideas on recoil, cone of fire, target computer overload, convergence (which was in the game in CB) are nice but will never be reality.
PGI ran waaaaay to deep into their current system. They neither have the ability nor the time to change it again.

[DeadlyNerd]

meteorol, on 09 April 2014 - 12:17 AM, said:

Recoding the whole system of how weapons work will not happen.

Not disagreeing with you but there's certain things people need to know about today's programming. Everything is done object oriented. I pity those that still think programming in C will change the world.
Well structured object oriented programming rarely needs "recoding", because it's highly modular, if done right. This is, after all, what they boasted about with UI2.0 .

All weapons in MWO are, hopefully, classes with a certain superclass. That superclass should contain attributes like heat, rate of fire, damage etc. because they're general attributes across all weapons.
I presume same is, or at least should be, with mech parts. Therefore, implementing a simple attribute like convergence(which is likely an array containing values for each level of load), is child's play.
Algorithm that calculates convergence already exists. It's simplified yes, but it exists. It just needs expansion.

My point is that technology for this already exists within MWO code. PGI either has no knowledge of how the code fully works(black magic), or they're just a lazy bunch of twats. I'd go for both.

Edited by DeadlyNerd, 09 April 2014 - 01:38 AM.

[Haji1096]

The idea your guys are talking about is here: http://www.qqmercs.com/?p=2780

From what I understand, PGI can't implement convergence because the engine doesn't support it / server side authority / host state rewind related issues.

[Funkadelic Mayhem]

if PGI took a tenth of the time finding a better engine for this game they would have saved a nindyth of the QQ! And had a much more solid game.

[DeadlyNerd]

Haji1096, on 09 April 2014 - 02:53 AM, said:

From what I understand, PGI can't implement convergence because the engine doesn't support it / server side authority / host state rewind related issues.

Convergence already exists, or that's what they made us think with the whole "convergence speed" skill. I wouldn't be surprised if it's the latter.
Server side authority is bs. Convergence(or where the weapon hits) is calculated server side, there's no lack in authority.
HSR is a failed system that still doesn't work as intended on lasers. Also, if they've any brains, they have the HSR and convergence decoupled and probably HSR taking in values from convergence before sending everything to the client. This means that any modification to existing convergence can be done without touching HSR.

Everything PGI spews as "reasons" is pretty much technical-term bullcrap meant to scare the uneducated masses.

[meteorol]

DeadlyNerd, on 09 April 2014 - 07:08 AM, said:

Convergence already exists, or that's what they made us think with the whole "convergence speed" skill. I wouldn't be surprised if it's the latter.

No, convergence was in this game during closed beta. They removed it because it was really hard for high ping players to hit anything. The "convergence speed" skill does not work at the moment. It is a 2 year old placeholder waiting for its replacement. It does literally nothing at the moment.
They removed convergence and implemented HSR to give highping players (which seems to be like half of the playerbase) the ability to hit anything. Apparently it was outside of their capabilities to keep convergence and implement HSR.

[Haji1096]

DeadlyNerd, on 09 April 2014 - 07:08 AM, said:

Convergence already exists, or that's what they made us think with the whole "convergence speed" skill. I wouldn't be surprised if it's the latter.
Server side authority is bs. Convergence(or where the weapon hits) is calculated server side, there's no lack in authority.
HSR is a failed system that still doesn't work as intended on lasers. Also, if they've any brains, they have the HSR and convergence decoupled and probably HSR taking in values from convergence before sending everything to the client. This means that any modification to existing convergence can be done without touching HSR.

Everything PGI spews as "reasons" is pretty much technical-term bullcrap meant to scare the uneducated masses.

I hope you are right. I'm not an expert in networking or programming, even though I know more than an average person in those areas..I can't speak to the validity of their technical reasons for not implementing convergence.

IMO, Convergence would solve the problem of high alpha FLD weapons. Seems like that is pretty important. I have a good ping. I know its selfish, but I don't care if other people can't hit anything. HSR is still problematic anyway. I haven't seen any stats on how good HSR is relative to what how hits were occuring in closed beta.

[DeadlyNerd]

Haji1096, on 09 April 2014 - 09:59 AM, said:

IMO, Convergence would solve the problem of high alpha FLD weapons. Seems like that is pretty important. I have a good ping. I know its selfish, but I don't care if other people can't hit anything. HSR is still problematic anyway. I haven't seen any stats on how good HSR is relative to what how hits were occuring in closed beta.

Actually early open beta had pretty good hit detection. Then they managed to **** something up, along with not updating their servers for the amount of players that were playing(I'm guessing Russ needed a new boat). HSR was never supposed to be a solution if they just opened up european servers and/or upgraded the US servers.
Now, their netcode is so patched up, even the creator doesn't know how it does what it does.

This convergence PGI is so afraid of is basically randomizing the trajectory of a 3D object according to certain input values. Valve had it done in CS 1.0 and I believe that was 15 years ago.

Edited by DeadlyNerd, 10 April 2014 - 03:57 AM.