Why did Fdev choose to go the way they have with the netcode?

Status
Thread Closed: Not open for further replies.
example?
last time someone wrote that, the examples he gave, all have been running on a engine that doesn't do p2p at all. they all just got a hidden server/client netcode that was able to promote clients dynamically to servers if the initial server closed the session.

Sure, the most recent one I know.

Care to tell me about these other examples? Or explain what exactly you mean by that "hidden netcode" ...?

(what engine "doesn't do p2p"?)
 
Same with a central server. And mind you, most of the calculations would have to be done on this single server while in P2P connections the calculation would be shared between the clients. Think crowd computing like the Seti program. There's a reason why they don't use a mighty workstation instead.

thats probably wishfull thinking. the calculations are done simultaniously on all machines, and all peers have to send "Their" result to all peers - and all peers have to aknowledge that back to the all other.
and ontop of that, all clients have to send the result to the server anyway for cheat protection. (which AFAIK lacks a bit...)

the only thing that is probably spread across all peers, is the CPU power required for NPC ships... since they are not handledy by the server either.

and i think many here thinking that a client/server structure would need to have one persistant servers hosting all clients logged in at the same time... thats a missconception to.
they only need that for instances with more then one player in the same place. 99% of the playerbase would still be playing solo with just connecting the server for the BGS data.
 
Sure, the most recent one I know.

Care to tell me about these other examples? Or explain what exactly you mean by that "hidden netcode" ...?

(what engine "doesn't do p2p"?)

ok, guesss what i see in most recent review on that game:
[FONT=&quot]Pros:[/FONT]
[FONT=&quot]- Fun V-Trigger mechanic[/FONT]

[FONT=&quot]Cons:[/FONT]
[FONT=&quot]- Over simplified gameplay[/FONT]
[FONT=&quot]- Limited offline content[/FONT]
[FONT=&quot]- Poor online netcode[/FONT]
[FONT=&quot]- Small amount of characters (even with DLC)[/FONT]
[FONT=&quot]- Lazily written and animated story mode[/FONT]

and unreal engine's netcode is client2server, and i forgot what examples it was, but mostly shooter.
 
ok, guesss what i see in most recent review on that game:

You always get that with p2p games because most people either have awful connections or happen to try and play with friends on the other side of the world. They essentially complain about their own connection quality.

and unreal engine's netcode is client2server, and i forgot what examples it was, but mostly shooter.

Shooters are not fighting games. I was very specific about that.

Also an engine having an inbuild network component doesn't stop anyone from implementing and using their own.
 
Another example. Some years ago I've played a game called Vendetta, flight model not unlike Elite but on a single server in the US. I'm sitting in Europa and while most players didn't have any issues I was shot at by ships that I even could not see. No problems with ED though and the few occasions where I actually saw rubber bending or ghosting was very likely a long physical distance to these players or their connections were just bad. Jumpgate was very similar, almost impossible for me to do any sort of PVP. Whenever I now die in Elite in a PVP fight it feels at least that I can't blame the connection. :p

That very much just depends on your location. I'm in S/E Australia and I either have empty instances or I use a VPN and deal with the 200ms lag.
 
P2P + Server hybrid model is where the gaming industry is heading. The technology hasn't been fully bake yet an there still lots of issues to sort out but games with client to server models I suspect will become an increasing rarity over the next decade and will at some point disappear altogether.
 
Client-server networking would have meant subscriptions and a distinctly substandard multiplayer experience for players further away from the servers (or more expensive subscriptions to cover having servers in far-flung locations).

P2P has downsides of course, but they aren't significant for most of ED's intended gameplay.
 
Peripherally related; here is a presentation made by Frontier staff, on their back end information systems:

[video=youtube;EvJPyjmfdz0]https://www.youtube.com/watch?v=EvJPyjmfdz0[/video]

...if you're interested in that sort of thing. :)
 
Believe it or not but response time between a few players can be actually better with P2P, because it's a direct connection between the player's machines. Otherwise each manoeuvre, each shot and each touch to another ship would require an additional stage - the central server.

true but only half of the story. these peers don't need to exchange only in pairs, but need to share every event with every other peer. this traffic is actually a huge redundancy (for the client) compared to a client server model where each peer just transmits it's own actions and receives aggregated status updates. the thing to note here is that this redundancy is taxing every peer's connection which is the limiting factor, that's bad, plus the problem is exponential with number of peers. sure, you get away with a raspberry pi server, but there is a much lower hard cap on the global experience.

the other disadvantage is that you have no opportunity to balance and ease latency differences since peers talk directly between them. and a mp game is not a file sharing network, it has players fighting for their lives so it has to maintain a consistent state in real time properly processing events, hopefully in order, and it also has to give a fair opportunity to each peer. it can become dauntingly complex to do that just across peers without an arbiter, not to mention that you have to assume from the start that every peer might be compromised, and a portion of them actually will be.

so there are no pure p2p games and ed definitely has servers, it's just about knowing how much of the action goes through them. i'm curious, really. i guess they went for the minimum barely necessary, offloading everything possible to the peers. this is something they tweaked in the beta (although we don't know what, exactly, i assume it was optimization of peer traffic) and i saw some crazy bugs being reported that evidenced loss of state info in many situations, which was worrying, but seems they fixed it on release (or rolled back?).

there's much fuzz about p2p, it's just a technique, never used pure and it depends a lot on the use case and implementation. the biggest advantages of p2p are resillience, decentralization and anonymity. none of them are particular relevant for a game like ed, actually. so i'd say the reasons are economical, or at least were at the moment (maybe back then they were too scared or uncertain about necessary provisions). i think it is actually a handicap for the game, wonder if they would like to change that if they could.
 
Client-server networking would have meant subscriptions and a distinctly substandard multiplayer experience for players further away from the servers (or more expensive subscriptions to cover having servers in far-flung locations).

P2P has downsides of course, but they aren't significant for most of ED's intended gameplay.

Aren't significant?
Outside of the USA or Europe, it can be basically impossible to actually have a multi-player experience unless you use third party software.
 
Aren't significant?
Outside of the USA or Europe, it can be basically impossible to actually have a multi-player experience unless you use third party software.

I see plenty of other players while playing from New Zealand without using a VPN. But there does seem to be a healthy community of Aus/NZ players, which may not be true in other regions.
 
I see plenty of other players while playing from New Zealand without using a VPN. But there does seem to be a healthy community of Aus/NZ players, which may not be true in other regions.

Compared to the populations in the same systems if you use a VPN, the Aus/NZ population is tiny. I can fly through the bubble and never see a single player in open, then I can turn the VPN on and connect to the West Coast of the USA and there are hundreds.
 
No need to guess and if you're really curious as you say watch the video right above your post. If you feel you've understood at least the essence of it come back and tell me your conclusions. :D

they don't really talk about combat or network code in that video. there was another more specific one (can't remember where) about instancing and how peers got wired together as soon as they were in sight. but it was mainly about the handshake, no details about how it went from there. would watch again if someone hears a bell!
 
Believe it or not but response time between a few players can be actually better with P2P, because it's a direct connection between the player's machines. Otherwise each manoeuvre, each shot and each touch to another ship would require an additional stage - the central server. What the servers in ED do - in simplified terms - is to broker between decent connections. It's getting tricky with oversee connections and greater number of players in one instance, but that would be even much worse if everything (mostly collision detection is the delicate factor here) would have to pass a central server. The limitation of light speed is actually a thing when it comes to connections that are further away than 1000km with signal that have to go back an force and even have to be calculated in the middle.

Just saw my previous speaker actually mentioned EVE and, yeah, Jumpgate and pretty much answered the question from another angle.

In theory at least. Problem is that a considerable amount of people don't have a beefy connection to support multiple connected peers. Nothing Frontier can do about, but they should have considered it. I'm still not convinced that P2P was the right thing to do. Counter Strike relies far more on low latency than Elite for example and they have a client/server model. In Elite it's just annoying. Rubber banding everywhere, increased jump times and over all poor stability. Things that drive me away from Open a lot of times :-D

Sure the player base splits if you introduce local servers, but why would I want to play with somebody from America? It's not that I don't like them, but physics and network topology will result in a poor gaming experience for both of us. And there could always be an option to do exactly that: International districts like in Guild Wars (I think ArenaNet has done some incredible network stuff back in the days).

Client-server networking would have meant subscriptions and a distinctly substandard multiplayer experience for players further away from the servers (or more expensive subscriptions to cover having servers in far-flung locations).

P2P has downsides of course, but they aren't significant for most of ED's intended gameplay.

Server based games don't necessarily result in subscriptions. Guild Wars for example runs (still) without one and they had to support a far larger player base. The cost was covered by selling new standalone expansions and if I recall correctly they added a shop years later. Elite had one from the start and it sells pretty good, seeing all the swag in Open. ;-)
 
Last edited:
[video=youtube;02KS4C263AQ]https://www.youtube.com/watch?v=02KS4C263AQ[/video]

Interesting too see this when it goes live.
 
We see two general trends in the gaming industry:

1) Subscription based games are dying out.

2) Games with networking are moving towards P2P or hybrid rather than full on C/S.

FD are simply a bit ahead of the current trend. This means also they are on the front end in the industry of having to make P2P work for a (well, i wanted to say twitch based, but don't want to get into that semantic argument) FPS type game (please don't argue the definition, i already apparently can't call it a twitch game, leave me a definition that i can use!).

We can only speculate how much cash FD would have neeed to go full C/S for ED and whether they could have covered it with cosmetic sales. But i think it would be likely that the current model would not have satisfied the costs, and they might have had to focus less on gameplay features and more on customization... which would have led to the inevitable complaints that the game has even slower releases and the devs are focused on selling cosmetics.

If they had gone subscription based, I think the game would have been . I've never subscribed to a game and never will. I have enough monthly bills to pay, i'm not paying monthly for games on top. I have a PS4, but no PSN subscription, i can live without the "community" benefits of games. Any game that requires PSN, i simply don't buy.

So, overall, despite issues with P2P, i think they went the right direction, and in doing so, have future proofed themselves as well.
 
So, overall, despite issues with P2P, i think they went the right direction, and in doing so, have future proofed themselves as well.

Also Agony, theoretically, as time goes by and people have better and more stable internet connections courtesy of their ISPs (theoretical, I know) the p2p experience should improve, of course that is rather dependent on ISPs being able to deliver. My 50mbps line sounds great on paper but the connection is woeful and leads to a bad experience sometimes.
 
Here a response from Frontier during the kickstarter : "The game will be played across many servers, augmented by peer-to-peer traffic for fast responses."
 
Last edited:
Status
Thread Closed: Not open for further replies.
Back
Top Bottom