Jump to content

The Netcode is holding up this game's potential


What would you rather see development time spent on  

112 members have voted

  1. 1. What would you rather see development time spent on

    • Back End Netcode Issues
      87
    • More Planes to Fly
      25


Recommended Posts

I would genuinely like to see a slowdown in the aircraft that are being produced and some actual netcode updates being made. Server teams like Combat Box and Finnish are bending over backwards making wine out of water to make fun dynamic servers. The issue is that so much is being held by the netcode. Things that should be possible are not and we are stuck with relatively low player counts (84). 

 

I wouldn't be surprised if a server like Finnish could support 200 players if the netcode was updated. I want this so bad that I would be willing to spend money on a new expansion pack that just focused on back end changes and to see an increase in online player count.

 

 

  • Like 1
Link to post
Share on other sites

You find me a game that can handle 200 players vehicles on a map.

 

 

 

I wouldn't mind seeing the AI dumbed down a hair to allow for more AI vehicles and stuff on the ground to make the war feel more alive in multiplayer, and some simple heavy bombers. But I doubt things are so simple.

  • Sad 1
Link to post
Share on other sites
1 minute ago, 40plus said:

......if you care about multiplayer.......which most don't

 

If iL2 wants to make money, multiplayer is where it's at.

 

 

  • Confused 2
Link to post
Share on other sites
Just now, Denum said:

 

If iL2 wants to make money, multiplayer is where it's at.

 

 

 

They are making money. And since they are, they are in a better position than you to tell where the money is coming from.

  • Upvote 5
Link to post
Share on other sites

I can tel you not if Xsolla is being used as payment option, as for some that payment option does not work.  🤯 

 

But indeed I do not think a 200 players will generate that extra cash for 1C. 

Link to post
Share on other sites

Those of us that fly in multiplayer are the minority of the player base, always have been.  That said, the game engine itself, not the nebulous "netcode", is the bottleneck holding back both online and offline play.  I'm not blaming the developers here either, as there is only so much money floating around in this teeny tiny niche of combat flight simulation, inside the ever shrinking tiny niche of overall flight simulation.

 

If we had the same number of players as one of big name gold level titles we would already have all the planes, maps,  and other content we all lust for.

 

The team, IMHO, is doing an amazing job considering the limitations they have to work within.

  • Upvote 8
Link to post
Share on other sites

There's a difference between making money and MAKING money.

 

 

200? 

 

Try 4000+ at least...

 

Even if they released a multiplayer upgrade for 40$ I'd buy it. 

 

Single player isn't fun nor is it challenging.

 

 

 

 

1 minute ago, BlitzPig_EL said:

The team, IMHO, is doing an amazing job considering the limitations they have to work within.

 I whole heartedly agree 

But I'm not expecting something for nothing.

 

I'm willing to pay a little to get a little. 

 

Even at 50$ for the multiplayer upgrade, it's a small price to pay for the hundreds of hours I'll get out of it.

 

 

 

 

Link to post
Share on other sites
Just now, Denum said:

Ah you're one of those single player don't mess with my experience because the AI might get better then me, aren't you.

 

No. I'm one of the many 'used to fly multiplayer but don't anymore' guys who got fed up with pointless repetitive ahistorical furballs and gaming-the-game behaviour from teenagers. I bought the game to do what I like, when I like, because I enjoy it, not to win imaginary 'I'm a better pretend-pilot than you' points in a competition I'm not even participating in.

 

 

 

 

  • Like 1
  • Haha 3
  • Confused 1
  • Upvote 5
Link to post
Share on other sites
58 minutes ago, 40plus said:

......if you care about multiplayer.......which most don't

It's hard for people to care about multiplayer when it's left so unsupported.

 

The Finnish server is almost locked out during most of the day because there are so many players playing on it. There is a market for this.

1 hour ago, Denum said:

You find me a game that can handle 200 players vehicles on a map.

 

 

 

I wouldn't mind seeing the AI dumbed down a hair to allow for more AI vehicles and stuff on the ground to make the war feel more alive in multiplayer, and some simple heavy bombers. But I doubt things are so simple.

WWIIOL? 

 

I wouldn't mind more dumbed down AI also. 

11 minutes ago, AndyJWest said:

 

No. I'm one of the many 'used to fly multiplayer but don't anymore' guys who got fed up with pointless repetitive ahistorical furballs and gaming-the-game behaviour from teenagers. I bought the game to do what I like, when I like, because I enjoy it, not to win imaginary 'I'm a better pretend-pilot than you' points in a competition I'm not even participating in.

 

 

 

 

People are more than capable of creating historical situations in a server and gatekeeping certain behavior. 

Edited by Enigma89
  • Upvote 1
Link to post
Share on other sites
7 minutes ago, Enigma89 said:

It's hard for people to care about multiplayer when it's left so unsupported.

 

The Finnish server is almost locked out during most of the day because there are so many players playing on it. There is a market for this.

WWIIOL? 

 

I wouldn't mind more dumbed down AI also. 

 

 

I never had the chance to play WWIIOL but isn't it much simpler then iL2? 

Edited by Denum
  • Like 1
Link to post
Share on other sites
1 minute ago, Denum said:

 

 

I never had the chance to play WWIIOL but isn't it much simpler then iL2? 

 

 

In some ways and in other ways WWIIOL is much more complicated. Then again the game was built from the ground up to be a combined arms MP game so it's probably unfair to compare the two.

 

Either way, the sim communities online are shrinking and there seems to be a huge push from developers to cater to single player games. There are a lot of people who do want to play multiplayer but we are getting held back by developer decisions to shun us.

Link to post
Share on other sites
Just now, Enigma89 said:

...the sim communities online are shrinking and there seems to be a huge push from developers to cater to single player games. There are a lot of people who do want to play multiplayer but we are getting held back by developer decisions to shun us.

 

You seem to be suggesting that developers are consciously making decisions which reduce their potential to sell their products. Why would they do that?

Link to post
Share on other sites

Warthunder is making money hand over fist, and I know people suggest that it's because of the joystick aspect but when FS2020 dropped they literally sold out everywhere. 

I think the community is out there, it's just a matter of getting players interested in it again.

 

 

I'm curious in iL2s case if it's utilizing multiple cores or not? 

 

Could they alter the engine to shift the load on the CPU side of things?

This is out of my league knowledge wise but I'm sure it can be improved upon.

Edited by Denum
Link to post
Share on other sites

There are lots of things that could be 'improved upon'. The developers have been making improvements for years - the game is in a very different state than it was in 2013. And they have been funding such improvements by selling new content.

Link to post
Share on other sites

Why should the AI be dumbed down for the multiplayer crowd? Splitting the singleplayer and the multiplayer community by changing features like dumbing down the AI in favor of the multiplayer crowd is never a good idea.

  • Upvote 2
Link to post
Share on other sites
2 minutes ago, Q_Walker said:

Why should the AI be dumbed down for the multiplayer crowd? Splitting the singleplayer and the multiplayer community by changing features like dumbing down the AI in favor of the multiplayer crowd is never a good idea.

Dumbed down in sense of resources

 

Not ability. 

 

Instead of having each gunner be modeled as individual AI, have one AI control all guns and give the player the ability to control one if they choose.

 

It changes nothing in single player ( well it would allow for heavy bombers I guess) and it decreases server load in multiplayer.

 

 

Edited by Denum
Link to post
Share on other sites
1 minute ago, Denum said:

Dumbed down in sense of resources

 

Not ability. 

 

Instead of having each gunner be modeled as individual AI, have one AI control all guns and give the player the ability to control one if they choose.

 

 

Understood, thank you for clarifying.

Link to post
Share on other sites
41 minutes ago, AndyJWest said:

 

You seem to be suggesting that developers are consciously making decisions which reduce their potential to sell their products. Why would they do that?

Any developer making a PC-based simulator is consciously making a decision to reduce potential profit. The money is in mobile not with pc sim gaming. 

 

So it's not impossible for a development company to have a passion project and avoid the big money pie of mobile or console to try to make a quality product. With that said, multiplayer is lagging significantly behind with single player. 

 

Either way we are speculating here if there is or isn't a market with multiplayer. I believe it is, you probably don't. What I wanted to get a sense from here is to see if the developers plan to flesh out MP more.

 

 

Edited by Enigma89
Link to post
Share on other sites

I would love to see netcode updates. I don't think slowing down aircraft development will get you netcode developments. The guys who build the planes don't do the netcode, so you can't just swap them in. Network programmers are a specialized bunch. And turfing your plane makers and hiring network programmers in their stead means you have less money coming in from sales of planes, and if you ever need those guys back they probably won't be waiting around to work for you again. They're gone, along with all their specialized knowledge.
 

31 minutes ago, Denum said:

 

200? 

 

Try 4000+ at least...

 

 


200 is maybe doable with a lot of optimization, some shortcuts, and reasonable mission design. 4000 players on a map is probably impossible with current and near future hardware and networking. Network load increases exponentially to the number of players because of the constant updating of position, and that position needs to be fairly precise (unlike in an MMORPG, where precise location is mostly just for show, and can be fudged with no real impact). You need to do the math to see why it quickly goes from doable to impossible, even if it seems simple.

The server updates clients a certain number of times per second. This is the 'tick rate'. When the server gets overwhelmed with calculations, things slow down because the server can't do all the calculations it needs to do in the time between ticks. The lower the tick rate, the fewer updates you get. Which means the locations you are seeing on your screen are not accurate, your shooting can't be accurate, etc. It can cause rubberbanding or warping as the client catches up with the server, or the server just plain crashes (pardon the pun). You need a relatively high tickrate to support the fast-paced combat and 3D precision needed for the game. Otherwise you just zoom into the ground or the plane warps through your bullet stream.

If you have 10 players in an area, the server sends you info on the location of everyone else, and sends everyone else info of where you are and what your plane is doing. So that is 100 updates per server tick. With 11 players, you have 121 updates per server tick. 1 more player necessitates 21 more updates. It gets worse with every player added. To go from 10 players in a bubble to 20 players in a given bubble (doubling the players), you quadruple the number of updates needed per server tick - from 100 updates to 400 updates. To go from 20 players to 40 players, it quadruples again.  That's just for location of planes. Not including damage modeling, flak, AI, mission logic, etc. When the shooting starts things are going to get insane really quick.

Games with larger number of players have all sorts of tricks and workarounds, from dynamically instancing players into subgroups that are handled by different servers or processes, to literally slowing down time to let the server catch up (EVE online does this, huge fleet battles move at like 1/10th of real time or something like that). Neither of those are really viable for a game concerned with objective based combat with dogfighting. 

There's headroom in the network code I think to expand space - for example, right now apparently flak burst locations are transmitted by the server to every client no matter where they are on the map, making AA fire computationally expensive. 

 

24 minutes ago, Denum said:

Warthunder is making money hand over fist, and I know people suggest that it's because of the joystick aspect but when FS2020 dropped they literally sold out everywhere. 

I think the community is out there, it's just a matter of getting players interested in it again.

 

 

I'm curious in iL2s case if it's utilizing multiple cores or not? 

 

Could they alter the engine to shift the load on the CPU side of things?

This is out of my league knowledge wise but I'm sure it can be improved upon.


Warthunder makes money because of it's tiered competitive system, accessibility, and action gameplay. It doesn't have the barriers to entry of a flight sim. It's an action shooter with a tier grind.

The joysticks all ran out of stock because there was very little demand before so supplies in stock were low, and production capacity was also low. Even a small increase in that demand overwhelms the system, and tooling up to increase production takes too long and is too expensive, so you get backorders. Also, COVID screwed with everyone's supply lines, adding to the headaches.

Il-2 uses multiple cores but the main thread is on one core which jumps around. Compared to other modern games it is actually more CPU heavy. For the Dserver, it's all CPU. 

  • Upvote 3
Link to post
Share on other sites

Was refering to total player counts, I agree. 4000 on a map would be impossible! 

 

COVID completely dumpster fired my supply lines and cost me 30G last year. 

I cannot WAIT for it to be history! 

 

Good info on the server side stuff, I don't have much knowledge of all that so it's nice to learn more. 

 

 

My favorite part about war thunder is how cities would have flak hidden in them. So you'd be attacking a convoy then surprise! 88mm popping around you!

Link to post
Share on other sites
39 minutes ago, Enigma89 said:

Any developer making a PC-based simulator is consciously making a decision to reduce potential profit. The money is in mobile not with pc sim gaming. 

 

So it's not impossible for a development company to have a passion project and avoid the big money pie of mobile or console to try to make a quality product. With that said, multiplayer is lagging significantly behind with single player. 

 

Either way we are speculating here if there is or isn't a market with multiplayer. I believe it is, you probably don't. What I wanted to get a sense from here is to see if the developers plan to flesh out MP more.

 

I don't think there is any question whether there is 'a market for multiplayer' as far as IL-2 GB is concerned. There clearly is one. The question is how to best cater to that market, alongside the larger single-player market. Nothing I've seen in this thread suggests that the participants are in a better position than the developers to answer that question, which makes the poll rather pointless. The developers (and the investors who fund development) aren't going to base decisions on a poll attracting at best a couple of hundred votes. Particularly when the poll is based on a false premise. As RedKestrel says above, the people responsible for new content aren't going to be netcode developers, making any decision regarding further work on multiplayer content contingent on being able to justify it to investors, rather than on reallocating resources currently being used elsewhere. If you want to know what the developers plans are, ask them. And if you want to influence such plans, don't try to second-guess their decisions based on false assumptions. Furthermore, the developers have already decided to go ahead with 'More Planes'. They are accepting advance payment for them. Backtracking on that isn't a remotely rational option, making the poll entirely null and void.

Link to post
Share on other sites
39 minutes ago, AndyJWest said:

 

I don't think there is any question whether there is 'a market for multiplayer' as far as IL-2 GB is concerned. There clearly is one. The question is how to best cater to that market, alongside the larger single-player market. Nothing I've seen in this thread suggests that the participants are in a better position than the developers to answer that question, which makes the poll rather pointless. The developers (and the investors who fund development) aren't going to base decisions on a poll attracting at best a couple of hundred votes. Particularly when the poll is based on a false premise. As RedKestrel says above, the people responsible for new content aren't going to be netcode developers, making any decision regarding further work on multiplayer content contingent on being able to justify it to investors, rather than on reallocating resources currently being used elsewhere. If you want to know what the developers plans are, ask them. And if you want to influence such plans, don't try to second-guess their decisions based on false assumptions. Furthermore, the developers have already decided to go ahead with 'More Planes'. They are accepting advance payment for them. Backtracking on that isn't a remotely rational option, making the poll entirely null and void.

 

 

I like my style better.

 

I want to validate whether I am the only one thinking this. If I find out that my thought is shared by many then I will go to the developers.

 

It's generally a good idea to make sure you don't have a crazy idea before you bother developers. Just some advice for life.

Edited by Enigma89
  • Like 1
Link to post
Share on other sites
12 minutes ago, Enigma89 said:

If I find out that my thought is shared by many then I will go to the developers.

 

The developers are well aware. Variations of this pointless topic appear just about every week, always showing little understanding of the real challenges.

 

Developers can't arbitrarily reassign resources from art to engine programming.

 

Developers have obligations to their overlords, so they must not fall too far behind release schedules. Customers aren't happy either if projects are delayed.

 

The fact that you and your friends would pay $40 for netcode improvements doesn't mean that the product would produce good returns on the broader market.

Edited by Mitthrawnuruodo
  • Upvote 3
Link to post
Share on other sites
19 minutes ago, Enigma89 said:

 

 

I like my style better.

 

I want to validate whether I am the only one thinking this. If I find out that my thought is shared by many then I will go to the developers.

 

It's generally a good idea to make sure you don't have a crazy idea before you bother developers. Just some advice for life.

 

If I want advice for life, I'll look for it elsewhere, thank you.

Edited by AndyJWest
  • Confused 2
Link to post
Share on other sites
1 hour ago, Denum said:

Was refering to total player counts, I agree. 4000 on a map would be impossible! 

 

COVID completely dumpster fired my supply lines and cost me 30G last year. 

I cannot WAIT for it to be history! 

 

Good info on the server side stuff, I don't have much knowledge of all that so it's nice to learn more. 

 

 

My favorite part about war thunder is how cities would have flak hidden in them. So you'd be attacking a convoy then surprise! 88mm popping around you!

Well, there are at least 3500ish active players on Combat Box in any given month. Not concurrent, mind you. 

For flak, you CAN do this in Il-2 if you want - but the aforementioned AA situation means that takes up lots of CPU you want to use elsewhere every time they shoot. If this were changed, you could be a little less discriminating as to where you placed the AA.

 

Link to post
Share on other sites
28 minutes ago, AndyJWest said:

 

If I want advice for life, I'll look for it elsewhere, thank you.

So you want to give out advice and not take it. Gotcha...

 

Sounds like you do belong in single player. Good luck against the AI!

Edited by Enigma89
  • Upvote 2
Link to post
Share on other sites
3 minutes ago, Enigma89 said:

So you want to give out advice and not take it. Gotcha...

 

Sounds like you do belong in single player. Good luck against the AI!

 

What do you want to be when you grow up?

  • Confused 2
Link to post
Share on other sites
7 hours ago, Denum said:

I'm curious in iL2s case if it's utilizing multiple cores or not? 

 

Could they alter the engine to shift the load on the CPU side of things?

This is out of my league knowledge wise but I'm sure it can be improved upon.

This is something that needs to be fixed and improved upon...I have 6 cores/12 threads running at a stable 4 ghz each with only 25-35% usage and every time there is a furball in single player career mode, I get stuttering as well as significant FPS drop. The AI planes, gunners, and ground units consume far more resources than CLoD or 1946, and those games have much more challenging (if imperfect) AI.

 

By the way, I still love this game and live with its shortcomings, but it can be improved significantly. The AI/CPU bottleneck is a serious hindrance to the Normandy landing, Channel, Pacific, and BoB scenarios.

Edited by drewm3i-VR
Link to post
Share on other sites
32 minutes ago, drewm3i-VR said:

This is something that needs to be fixed and improved upon...I have 6 cores/12 threads running at a stable 4 ghz each with only 25-35% usage and every time there is a furball in single player career mode, I get stuttering as well as significant FPS drop. The AI planes, gunners, and ground units consume far more resources than CLoD or 1946, and those games have much more challenging (if imperfect) AI.

 

Yes, there's no doubt that the more modern multicore CPUs could be utilised more efficiently by IL-2 GB, though it isn't likely to be simple: if it was, the developers would surely have done it already. Any such improvements, if and when they come, are likely to benefit both single-and multiplayer environments, making the suggestion that the developers should prioritise unspecified 'netcode' over other work even less sensible - they know where the work is needed, and where it will benefit most.

  • Upvote 1
Link to post
Share on other sites

I recall Jason being quizzed about this on what I believe was his last Q&A stream.  He said that the netcode was optimised in other directions (I don't recall what) and that he always had difficulty getting people to believe that.  Perhaps someone could post a link to that stream.

  • Upvote 1
Link to post
Share on other sites
6 hours ago, AndyJWest said:

 

Yes, there's no doubt that the more modern multicore CPUs could be utilised more efficiently by IL-2 GB, though it isn't likely to be simple: if it was, the developers would surely have done it already. Any such improvements, if and when they come, are likely to benefit both single-and multiplayer environments, making the suggestion that the developers should prioritise unspecified 'netcode' over other work even less sensible - they know where the work is needed, and where it will benefit most.

Just for kicks, I fired up CLoD tonight for a few minutes to test a huge aerial battle and see what would happen. With about 20 ai bombers and 25 or so ai fighters, the framerate was like 140-150 FPS on ultra on a laptop. The AI also shot me down in a few minutes and the bomber gunners were just as believable as GB. I would play it more of if it had VR, even though the FM and interface in GB are much better.

Edited by drewm3i-VR
Link to post
Share on other sites
  • 2 weeks later...

People actually play this game single player? 

The AI breaks single player for me.

 

The netcode dampens multiplayer for me.  

 

My vote is for improving netcode since a viable AI is extremely hard to do without cheating.

Link to post
Share on other sites
On 1/19/2021 at 2:57 PM, Denum said:

You find me a game that can handle 200 players vehicles on a map.

 

 

 

I wouldn't mind seeing the AI dumbed down a hair to allow for more AI vehicles and stuff on the ground to make the war feel more alive in multiplayer, and some simple heavy bombers. But I doubt things are so simple.

 

ACG's campaign missions on Cliffs were pretty much always about 100 players plus 60AI bombers plus literally hundreds of objects on the ground, static and AI.  Broke through that 200 by miles.  Unfortunately in BoX this is not a development item at all, and ACG did beg for it years ago because of the stiff limits.  Now we pack our server to the limit in Kuban and ride on the edge of banding and crashes, with fewer players and a much more staggered and stretched out with shifts campaign mission in order to manage it.  It would be wonderful to actually have the battles and scenes replicated as they were in the war.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...