#4427611 - 06/25/18 01:15 AM
Re: Is ED deliberately misleading potential customers?
[Re: - Ice]
|
Joined: Dec 2003
Posts: 6,893
GrayGhost
Hotshot
|
Hotshot
Joined: Dec 2003
Posts: 6,893
|
I think that you don't read unless it fits your narrative. That the missile on the two different clients may not be displayed in the same position is a known thing, and has been known for a very long time. Heck, the mechanism for how/why this happens has been known for a long time, how What do you think, GG? Is that close enough, do you think?
Last edited by GrayGhost; 06/25/18 01:17 AM.
-- 44th VFW
|
|
#4427713 - 06/25/18 07:41 PM
Re: Is ED deliberately misleading potential customers?
[Re: GrayGhost]
|
Joined: Jun 2005
Posts: 16,082
- Ice
Veteran
|
Veteran
Joined: Jun 2005
Posts: 16,082
Philippines / North East UK
|
I think that you don't read unless it fits your narrative. That the missile on the two different clients may not be displayed in the same position is a known thing, and has been known for a very long time. Heck, the mechanism for how/why this happens has been known for a long time, how You've just confessed that the MP sync is broken and has been known to be broken for a very long time and yet nothing has been done about it. I didn't know you supported my narrative, GG! For a game that is focused on COMBAT (or at least claims to be), isn't accuracy of unit and weapon position paramount? I guess ED thinks it's okay to show one guy that his missle went through the target while showing the other guy that the missile missed by miles? In a high fidelity COMBAT SIMULATOR?
- Ice
|
|
#4427893 - 06/27/18 01:52 AM
Re: Is ED deliberately misleading potential customers?
[Re: - Ice]
|
Joined: Dec 2003
Posts: 6,893
GrayGhost
Hotshot
|
Hotshot
Joined: Dec 2003
Posts: 6,893
|
You've just confessed that the MP sync is broken and has been known to be broken for a very long time and yet nothing has been done about it. I didn't know you supported my narrative, GG! Oh a confession! I haven't confessed anything - like I said, this has been known for a very long time. The missile 'sync' consists of sending the launch signal and then each client simulates the missile independently. As for fuzes of that era, they were as reliable as the rest of the electronics. A lot misses at that time were due to lack of training and good DLZ display/knowledge as well.
Last edited by GrayGhost; 06/27/18 01:54 AM.
-- 44th VFW
|
|
#4427903 - 06/27/18 05:26 AM
Re: Is ED deliberately misleading potential customers?
[Re: GrayGhost]
|
Joined: Mar 2003
Posts: 3,922
Paradaz
Senior Member
|
Senior Member
Joined: Mar 2003
Posts: 3,922
UK
|
The missile 'sync' consists of sending the launch signal and then each client simulates the missile independently.
Doesn’t that mean that the more clients that are present, the more likely it is that different sets of data with lag/latency and jitter will be formed and interpreted? It sounds like it would be impossible for one missile to guarantee the same coordinates are output as each client has different network properties and behaviours!
On the Eighth day God created Paratroopers and the Devil stood to attention.
|
|
#4427918 - 06/27/18 12:46 PM
Re: Is ED deliberately misleading potential customers?
[Re: Paradaz]
|
Joined: Dec 2003
Posts: 6,893
GrayGhost
Hotshot
|
Hotshot
Joined: Dec 2003
Posts: 6,893
|
I don't actually know the answer to to that because I don't know the exact mechanism of transferring the launch command, but that sounds like it would make sense. You are right, it is impossible to guarantee the same coordinates - this method depends on the two missiles behaving the same way on both clients. It all works fine _most_ of the time, but it would be nice if guided missiles (I suspect unguided rockets aren't as big a deal and given that MLRSs and other entities can put a LOT of those in the virtual air ... ) would be synced every second or five to ensure they're on track in both simulations. I have seen significant desync even with a low number of players - it's rare, but very obvious when it happens. Guidance improvements can actually alleviate a bunch of it because it will smooth out missile reactions at longer ranges, but this is only one part of the puzzle as things happen as shorter ranges also. Doesn’t that mean that the more clients that are present, the more likely it is that different sets of data with lag/latency and jitter will be formed and interpreted? It sounds like it would be impossible for one missile to guarantee the same coordinates are output as each client has different network properties and behaviours!
Last edited by GrayGhost; 06/27/18 12:48 PM.
-- 44th VFW
|
|
#4427933 - 06/27/18 03:10 PM
Re: Is ED deliberately misleading potential customers?
[Re: bisher]
|
Joined: Jun 2005
Posts: 16,082
- Ice
Veteran
|
Veteran
Joined: Jun 2005
Posts: 16,082
Philippines / North East UK
|
Oh a confession! I haven't confessed anything - like I said, this has been known for a very long time. The missile 'sync' consists of sending the launch signal and then each client simulates the missile independently. You must be confused as to what "confession" means. Would "admitted" be a better term? "each client simulates the missile independently" -- I'm no coder but that doesn't sound like the best way to do it. What if you fire a missile at me and in your client, your missile hits me but in my client, the missile doesn't? What happens then? I suppose each client also simulates the FM of aircraft independently? I don't actually know the answer to to that because I don't know the exact mechanism of transferring the launch command, but that sounds like it would make sense. You are right, it is impossible to guarantee the same coordinates - this method depends on the two missiles behaving the same way on both clients. It all works fine _most_ of the time, but it would be nice if guided missiles (I suspect unguided rockets aren't as big a deal and given that MLRSs and other entities can put a LOT of those in the virtual air ... ) would be synced every second or five to ensure they're on track in both simulations. I wonder if it works this way for shooter games? Each individual client calculates ballistics and penetration values of each projectile individually? Must deal with a lot of inaccuracies and no matter how small, it will all build up to some messed up results! 'The MP sync is broken and has been a known issue for a very long time and nothing has been done about it'. Is this what you are trying to say Ice? Good point. A good starting point for interesting debate I lose the goodness of your point when you make judgement statements re posters, that are a disconnect to the conversation about the flight sim Not sure what you're trying to say here, bisher.
- Ice
|
|
#4428177 - 06/29/18 04:28 AM
Re: Is ED deliberately misleading potential customers?
[Re: - Ice]
|
Joined: Dec 2003
Posts: 6,893
GrayGhost
Hotshot
|
Hotshot
Joined: Dec 2003
Posts: 6,893
|
You must be confused as to what "confession" means. Would "admitted" be a better term? I'm not. 'Confession/Admitting' has a negative connotation, as in someone was doing something bad or trying to hide something. Neither of those is true. "each client simulates the missile independently" -- I'm no coder but that doesn't sound like the best way to do it. What if you fire a missile at me and in your client, your missile hits me but in my client, the missile doesn't? What happens then? In the case of THIS game, the missile on the shooting client's machine is taken as correct. In your scenario, you watch the missile not track or barely miss or explode far from you, but you take damage assigned from the detonation distance on my client. I suppose each client also simulates the FM of aircraft independently? They do ... kind of a strange question. FMs might not be super-heavy things to simulate (not simple either) but if you can avoid running them all in one place, you probably should. I wonder if it works this way for shooter games? Each individual client calculates ballistics and penetration values of each projectile individually? Must deal with a lot of inaccuracies and no matter how small, it will all build up to some messed up results! In ARMA3 the standard way of dealing with someone close up was to run circles around them while reloading your gun. I haven't played for a while, but it was an effective way of avoiding hits. It always struck me as strange that I could put a stream of bullets in someone's way, they'd run through that stream and not take a hit. There is no game that isn't affected by these MP issues. It's just that some are affected more, others less, IMHO. You can say what you want about a given solution being good or bad ... there's no perfect solution. In the case of this game, we appear to be using an older solution that reduces network traffic but results in the aforementioned artifacts some times.
-- 44th VFW
|
|
#4428244 - 06/29/18 04:11 PM
Re: Is ED deliberately misleading potential customers?
[Re: GrayGhost]
|
Joined: Jun 2005
Posts: 16,082
- Ice
Veteran
|
Veteran
Joined: Jun 2005
Posts: 16,082
Philippines / North East UK
|
You must be confused as to what "confession" means. Would "admitted" be a better term? I'm not. 'Confession/Admitting' has a negative connotation, as in someone was doing something bad or trying to hide something. Neither of those is true. Sure. Suggest a better, more fitting term? Either way, you've already helped me prove my point. Thanks! In the case of THIS game, the missile on the shooting client's machine is taken as correct. In your scenario, you watch the missile not track or barely miss or explode far from you, but you take damage assigned from the detonation distance on my client. So in THIS game, it's possible for me to watch me do a successful missile avoidance maneuver but still blow up from a direct hit (your client) even though I just saw it fly past me (my client). Wow. Such advanced MP netcode. Why not simply use the firing client's calculations to extrapolate missile coordinates and use those coordinates to show the defensive client where the missile is? One calculation, done. Why do two only to ignore the other and when the other is also totally unneeded? I suppose each client also simulates the FM of aircraft independently? They do ... kind of a strange question. FMs might not be super-heavy things to simulate (not simple either) but if you can avoid running them all in one place, you probably should. No, you claimed each client simulates each missile independently.... so does each client simulate the FM of each aircraft independently as well? ie, in a dogfight, your client simulates the FM of the missile and the FM of both our aircraft and my client separately simulates the FM of the missile and the FM of both our aircraft? It was kind of a rhetorical question, obviously if each client calculates the FM of each asset in the air --- aircraft, weapons, missiles, etc, then the game is doing a lot of work it does not need to. In the case of this game, we appear to be using an older solution that reduces network traffic but results in the aforementioned artifacts some times. I agree! Like you said, it's been broken for some time, ED knows it's been broken for some time, and has simply been accepted as the norm rather than be a priority fix. Who wants accurate positioning of missiles or correct simulation of fuzes if the community has already made up excuses for this behavior? Now let's go back to ensuring accurate wing vapors and correct water droplet behavior on the canopy!!
- Ice
|
|
#4428298 - 06/30/18 12:27 AM
Re: Is ED deliberately misleading potential customers?
[Re: - Ice]
|
Joined: Dec 2003
Posts: 6,893
GrayGhost
Hotshot
|
Hotshot
Joined: Dec 2003
Posts: 6,893
|
Sure. Suggest a better, more fitting term? Either way, you've already helped me prove my point. Thanks! What have you proven that we didn't know? As for a more fitting term? You've been informed of the facts I guess? So in THIS game, it's possible for me to watch me do a successful missile avoidance maneuver but still blow up from a direct hit (your client) even though I just saw it fly past me (my client). Wow. Such advanced MP netcode. Why not simply use the firing client's calculations to extrapolate missile coordinates and use those coordinates to show the defensive client where the missile is? One calculation, done. Why do two only to ignore the other and when the other is also totally unneeded? It's more likely you'll never perform a missile avoidance maneuver Calculations are done at all times - there's just no coordination. No, you claimed each client simulates each missile independently.... so does each client simulate the FM of each aircraft independently as well? ie, in a dogfight, your client simulates the FM of the missile and the FM of both our aircraft and my client separately simulates the FM of the missile and the FM of both our aircraft? It was kind of a rhetorical question, obviously if each client calculates the FM of each asset in the air --- aircraft, weapons, missiles, etc, then the game is doing a lot of work it does not need to. Right. I agree! Like you said, it's been broken for some time, ED knows it's been broken for some time, and has simply been accepted as the norm rather than be a priority fix. Who wants accurate positioning of missiles or correct simulation of fuzes if the community has already made up excuses for this behavior? Now let's go back to ensuring accurate wing vapors and correct water droplet behavior on the canopy!! It isn't broken ... what it is, is old.
-- 44th VFW
|
|
#4428308 - 06/30/18 04:24 AM
Re: Is ED deliberately misleading potential customers?
[Re: - Ice]
|
Joined: Jan 2006
Posts: 1,506
Brit44 'Aldo'
Every Human is Unique
|
Every Human is Unique
Member
Joined: Jan 2006
Posts: 1,506
|
"each client simulates the missile independently" I do not have a dog in this fight and I will not be following at any regular bases, but what you have described is the client / client protocols of the Y2K era ( ie DX6 and DirrectPlay). With server / client protocols, the firing unit should send the time / action to the server. The server then updates the other clients and adds the entity to the servers 'world'. There are two types of MP code. Client/Client and Server/Client. I would hope that a modern PC simulation is not using Client/Client. a quick google for client/client found this explanation. https://gafferongames.com/post/what_every_programmer_needs_to_know_about_game_networking/
Last edited by Brit44 'Aldo'; 06/30/18 04:31 AM.
TPA who TWI "The 10th Amendment simply says that any powers that aren’t mentioned in the Constitution as belonging to the government belong to the states themselves."
|
|
#4428335 - 06/30/18 01:59 PM
Re: Is ED deliberately misleading potential customers?
[Re: Brit44 'Aldo']
|
Joined: Dec 2003
Posts: 6,893
GrayGhost
Hotshot
|
Hotshot
Joined: Dec 2003
Posts: 6,893
|
I know, the concept is oldish. I don't know the guts of the implementation, but the server is definitely involved. I don't think this thing has been client/client since Flaming Cliffs came to be. "each client simulates the missile independently" I do not have a dog in this fight and I will not be following at any regular bases, but what you have described is the client / client protocols of the Y2K era ( ie DX6 and DirrectPlay). With server / client protocols, the firing unit should send the time / action to the server. The server then updates the other clients and adds the entity to the servers 'world'.
-- 44th VFW
|
|
|
|
Exodus
by RedOneAlpha. 04/18/24 05:46 PM
|
|
|
|
|
|
|
|
|
|