Desync and why it's fixable

"

iamstryker wrote:

Everytime a thread like this is made GGG responds to it that the poster does not fully understand everything.

I aim to be more successful than my predecessors. Time will tell.

the problem is, if they want to fix desync, they have to remove some of their "wonderful mechanic stuff"

How would you go on about multiplayer sessions? Considering this is an online game this should be the primary aspect to take into consideration.

Which client state will be used as the real gamestate? The host of the instance? Seems like the connected clients will get even bigger desyncs.
Although this would work fine for a singleplayer game (but then there would not be desync anyway), it is problematic for multiplayer sessions.

"

iamstryker wrote:

Everytime a thread like this is made GGG responds to it that the poster does not fully understand everything.

If it was easily fixable then it wouldn't still be a problem.

WOW!! You just hit the nail on the head. Fact is GGG CAN'T fix it. They fucked it all up over and over waaaaaaaaaaaaay too much. They need to sell while they can to a company that knows what the fuck they are doing.

"

danisonxtc wrote:

the problem is, if they want to fix desync, they have to remove some of their "wonderful mechanic stuff"

Doubtful. Can you be more specific?

"

Meistervinny wrote:

Which client state will be used as the real gamestate?

None of them. When I say "the client gamestate ahead of the server," I mean chronologically ahead, not of a higher importance; the server gamestate would still always be the official one.

"

ScrotieMcB wrote:

The client already has have the same answer as the server, using deterministic prediction methods, otherwise it predicts monster movement in a way which disagrees with the server, which would cause more desync; however, this is preventable. Assuming that system worked properly, the problem you mention would solve itself almost automatically.

You are forgetting about player behaviour, until you receive synched player information your client has "no clue" of where the other players are.
Their behaviour affects the mobs behaviour resulting them to be in a different state as well, when you have 4 players in a game you will essentially be running 4 different gamestates!

This is the reason the actual gamestate has to be run on the server and your client runs the simulation of it.
Other players have to send their states to the server first and then the server has to send the state to you which is an even bigger delay than how it works now.
Keep in mind that when running network software waiting for input is usually the most time-consuming thing so you want to avoid having to wait for information as much as possible.

"

Eternallight wrote:

Scrotie, whats your opinion on if desync (fixable or not) is a matter of "too late" at this time? As in, not fixable due to how the system is currently made and would require a re-scripting of the game from the ground up?

I would agree if GGG would not plan to develop the game for a long, coming time.

However how much resources did they spend until this point with giving us the experience we got now? I certainly would value this in when deciding on a new approach which could potentially net more success in the same time. Fixing desync is no short term thing anyway.


@Scrotie, I know that you are not fan of the thread already brought up but care to mention it at least as you was inspired in that thread for the solution you are proposing now? Also mentioning why GGG dismisses suggestions like more frequent resynching could ease away some potential arguing. :P

Scrotie,

Why a new thread now when only recently there was a long thread on the same, with the same info, and by yourself?

https://www.pathofexile.com/forum/view-thread/507419


am I missing some new info in this one - which you hadnt covered in your recent one linked above?


Curious.

"

ScrotieMcB wrote:

None of them. When I say "the client gamestate ahead of the server," I mean chronologically ahead, not of a higher importance; the server gamestate would still always be the official one.

So how will the server decide which clients state to adopt?
Imagine a situation where Player 1 kills a mob on his client, but Player 2 is not synched and the mob is still alive, both send their gamestate to the server, which one will the server adopt?

Player 2 might have gotten a Unique drop, player 1 may have still been standing in melee range while the mob has an on death modifier rendering him death.

From a multiplayer point of view I do not see this being a good solution. Also because the server will have to wait for data from multiple sources, or with the exception of 1 player desync will be increased.

"

MildlyClever wrote:

Scrotie,

Why a new thread now when only recently there was a long thread on the same, with the same info, and by yourself?

https://www.pathofexile.com/forum/view-thread/507419


am I missing some new info in this one - which you hadnt covered in your recent one linked above?


Curious.

The thread you provided was not a suggestion but more an analysis of what is going wrong with a small suggestion at the end to change from TCP to UDP, which GGG dismissed rather quickly.

This thread actually provides another suggestion. It would have been pointless to add this new suggestion to his last thread as the whole discussions in there resolved on his first analysis and the connected proposals.