Multiplayer oxygen not included

[Yunru]

28 minutes ago, Heinermann said:

We haven't. I simplified it to outline the issues that need to be tackled without going into detail. I mentioned mod because I agree, the developers won't be doing it. Using my experience I believe such a thing is feasible.

And? If enough people want a multiplayer it's not anyone's business but the devs to tell them they can't have it unless by modding, then people can continue the discussion about what they want or could be cool in the context of modding instead of getting a toxic community reaction by people who feel the need to continuously post their uninformed opinions and efforts to derail the thread. That's why I said it's better to have a serious discussion in a completely separate forum outside the Klei community.

And the reason you're ignoring the fact the reaction tends to be toxic is because it's been discussed to death in the proper subforum is because...?

Also "And?" nothing. You said people should butt out if it didn't concern them. I'm pointing out that it concerns anyone who so much as thought about playing the game.

[metallichydra]

why dont we make a subforum called "suggestions about multiplayer"?

[Nightinggale]

11 hours ago, Heinermann said:

You can ignore most people in this thread because they don't know what they're talking about.

I both agree and disagree with you. It's true that not all people are equally skilled in the art of designing network code for games and yes there are likely people who posted in this thread, who never tried to do so themselves. However openly declaring to ignore anybody not sharing your vision most certainly comes through as very arrogant.

Based on my experience, a complete lack of knowledge during the idea phase can sometimes be a good thing. Some of the most creative and alternative ideas usually comes from people who aren't limited by knowing how the code works and possibly trying to merge their thinking with the existing code.

11 hours ago, Heinermann said:

Once 1.0 is released we can have a serious discussion about a multiplayer mod in a separate forum or discord.

I can declare with 100% certainty that I won't be joining. I do have experience with getting network games to work/stay in sync and I'm not scared of working with games where you have to take network sync into account. However to be completely honest I don't believe in this project if Klei didn't plan multiplayer into the core of the game from the start. Because of that I won't get anywhere near a multiplayer mod project.

11 hours ago, Heinermann said:

Some people just want to play together with their friend or spouse

I don't think we disagree on this point. Nobody wants to prevent other people from playing multiplayer because "I prefer single player". The question isn't if multiplayer is enjoyable for some people or not (it most likely are). The question is if it is technically doable within reasonable development time and possible sideeffects of doing so.

[Hellshound38]

9 hours ago, Nightinggale said:

I both agree and disagree with you. It's true that not all people are equally skilled in the art of designing network code for games and yes there are likely people who posted in this thread, who never tried to do so themselves. However openly declaring to ignore anybody not sharing your vision most certainly comes through as very arrogant.

Based on my experience, a complete lack of knowledge during the idea phase can sometimes be a good thing. Some of the most creative and alternative ideas usually comes from people who aren't limited by knowing how the code works and possibly trying to merge their thinking with the existing code.

I can declare with 100% certainty that I won't be joining. I do have experience with getting network games to work/stay in sync and I'm not scared of working with games where you have to take network sync into account. However to be completely honest I don't believe in this project if Klei didn't plan multiplayer into the core of the game from the start. Because of that I won't get anywhere near a multiplayer mod project.

I don't think we disagree on this point. Nobody wants to prevent other people from playing multiplayer because "I prefer single player". The question isn't if multiplayer is enjoyable for some people or not (it most likely are). The question is if it is technically doable within reasonable development time and possible sideeffects of doing so.

I doubt you have any relevant experience of this scale. Factorio is the perfect example of how much work it would take. The devs had to change all the algorithms to be deterministic. It was a massive undertaking, and the computational algorithms in factorio are multiple orders of magnitude simpler than in oxygen. While most of us would love multiplayer if it existed, we know it won't that is all.

[Oozinator]

Don't Starve players - mimimi want multiplayer - got Don't starve together.
Oxygen not included players - mimimi want multiplayer - Oxygen not included together!

 

 

[Nightinggale]

17 minutes ago, Hellshound38 said:

I doubt you have any relevant experience of this scale. Factorio is the perfect example of how much work it would take. The devs had to change all the algorithms to be deterministic. It was a massive undertaking, and the computational algorithms in factorio are multiple orders of magnitude simpler than in oxygen. While most of us would love multiplayer if it existed, we know it won't that is all.

Huh? I feel like you are trying to tell me that I underestimated the task. That's kind of an odd statement after I declared I wouldn't touch it due to the massive scale. I feel like some people in this thread might be underestimating the challenge, but I'm not one of them.

[Xuhybrid]

Oh look another multiplayer thread yay /sarcasm

[camelot]

 

 

1

16 hours ago, Nightinggale said:

The question is if it is technically doable within reasonable development time and possible sideeffects of doing so.

For asynchronous multiplayer and interactions mostly are trading resources/dupes then yes (based on my experience developing a similar game in 2011, and my team at that moment is very small and inexperienced). The biggest side effect is you need servers to record and sync user's data, which required a monthly maintenance fee. So expect to see some kind of micropayment in-game (purchase resources/skins...) or pay a small monthly fee for these servers or Klei could figure out some other way to cover that cost (sponsorship, partnership...)

[Glassyfo]

24 minutes ago, camelot said:

So expect to see some kind of micropayment in-game (purchase resources/skins...) or pay a small monthly fee for these servers or Klei could figure out some other way to cover that cost (sponsorship, partnership...)

Haaaaaaaaard no on that one. Very, very hard no.

[camelot]

1 minute ago, bleeter6 said:

Haaaaaaaaard no on that one. Very, very hard no.

Uhm, so you never played DST?

[Nightinggale]

4 hours ago, camelot said:

The biggest side effect is you need servers to record and sync user's data, which required a monthly maintenance fee.

I find it odd that people claim a paid online server is needed. Before such servers were available, all games would use a player's computer free of charge. It worked fine and it worked for offline LAN games. ONI would be expensive if a server has to run to keep track of the game itself because the server CPU load would be the same for each game as it is for single player right now. If the game needs a "central server" for keeping track of a game, use the computer of one of the players. This can be done without monthly payments or anything of that nature.

Besides the talk about a server doesn't really matter because that's not where the problem is. Adding a server into the game to host multiplayer isn't the problem. The problem isn't network communication. The core of the problem is lack of network communication. If computer A does this, then computer B has to do the same. This is either done by network communication (has to be added for each case in the code) or by ensuring predictable output of the code (same input, same output).

Let's look at an example of what can happen. Say we have a single space scanner, which controls bunker doors. It will close the doors at a random time before the meteors start to strike. The two computers don't agree on which random number is used and on one computer the doors are closed well in advance, which then blocks solar power production. On the other computer the doors fail to close in time and a meteor destroys the solar powerplant. The two computers are no longer playing the same game.

That's the problem. Every single instance where the game can branch off in more than one direction, the code will need to be updated to make sure all computers in the game will pick the same branch. When you start looking at all the cases of possible branching, then you realize that fixing this will take ages.

Another problem is that making the code predictable like this means it's even harder to use more than one CPU core because if we have two cores, each running a task and the output depends on which one executes first, then we have to make sure they execute in the same order on all computers. In singleplayer it will be like "who cares if the dupe farts on the left or right cell". This means adding multiplayer is restricting the abilities to use more CPU cores.

[camelot]

 

 

1

16 minutes ago, Nightinggale said:

I find it odd that people claim a paid online server is needed.

Are we talking about synchronous or asynchronous? Because in the asynchronous case, the minimum concurrent players needed so trading can occur normally would be hundreds. Pretty sure that hosting data on player's computer like DST case with 6-8 concurrent players is not gonna work for most people. So 24/7 dedicated servers is better

And synchronous multiplayer for ONI is not a very viable option

[Nightinggale]

I can't see gas/liquid movements and temperature updates realistically done in any way other than synchronous. The alternative is a server, which transmits element, mass and temperature for each cell for each update. With 256x384 cells and 5 updates each second, that's almost half a million cells to transmit each second. If we say 32 bit for each variable, then that's almost 6 MB/s. That's for each player and we haven't even added overhead. Don't bother considering compression because we don't have enough CPU time even without compression running in the background.

This is assuming normal speed. Increasing the game speed will increase the bandwidth requirements.

Sure we can reduce to once a second and we can skip vacuum cells (they didn't change) and other stuff like that, but we are still very far from anything, which is reasonable for a network game, particularly if played online.

This is part of why I said either Klei planned for multiplayer from the start, or I don't believe it will happen. Either the physics model is prepared for concurrency for synchronous execution or it will never stop desyncing games.

[suxkar]

13 hours ago, Oozinator said:

Don't Starve players - mimimi want multiplayer - got Don't starve together.
Oxygen not included players - mimimi want multiplayer - Oxygen not included together!

 

 

It's not the first time I see people not appreciating your contribution.
Unacceptable.
Thank you sir

[Glassyfo]

8 hours ago, camelot said:

Uhm, so you never played DST?

Nope. And not gonna.

[Gracefulmuse]

5 minutes ago, bleeter6 said:

Nope. And not gonna.

Its not bad. I prefer normal Don't Starve, but I do play DST sometimes with my husband. It's micro transactions aren't really a problem. Everything you can buy is purely cosmetic, and can potentially be earned by playing the game.***

 

***This newest character that was added did require a micro transaction to unlock, though I think you could still unlock with game play, just hard. I am NOT a fan of that, and I hope they do not repeat that in the future. 

[phod]

1 hour ago, bleeter6 said:

Nope. And not gonna.

Some of the most enjoyable moments I've ever had in a game were in DST, playing with my daughter. Granted, I have no interest in playing with random people, I can't imagine that would work out well. 

[Oozinator]

8 hours ago, suxkar said:

It's not the first time

 

[Tytan]

as it is a game of micro-management, I do not even see how to apply a multiplayer that would not be different from the final game, the only way I see is that multplayer would work just like trips to planets work, that is, you go to the planet of his friend, but without any interaction, and can come back with resources from there, there he could leave a point of "selling" as if it were a store where he could leave certain resources where he would pre-define value in another resource
ex:
100 kg of thermium per 1 t of wolframite

is the only kind of interaction I see without changing the game

[Glassyfo]

5 hours ago, Tytan said:

as it is a game of micro-management, I do not even see how to apply a multiplayer that would not be different from the final game, the only way I see is that multplayer would work just like trips to planets work, that is, you go to the planet of his friend, but without any interaction, and can come back with resources from there, there he could leave a point of "selling" as if it were a store where he could leave certain resources where he would pre-define value in another resource
ex:
100 kg of thermium per 1 t of wolframite

is the only kind of interaction I see without changing the game

Bartering would have been a possible multiplayer option of it weren't for the fact that rockets already get most things you need. Those that rockets don't, well, why would you give those away?

[Tytan]

1 hour ago, bleeter6 said:

Bartering would have been a possible multiplayer option of it weren't for the fact that rockets already get most things you need. Those that rockets don't, well, why would you give those away?

because in this way you could pick up the specific object, and in large quantities

[camelot]

5 hours ago, bleeter6 said:

Bartering would have been a possible multiplayer option of it weren't for the fact that rockets already get most things you need. Those that rockets don't, well, why would you give those away?

That is boring as heck. If bartering is the main theme, then your world in multiplayer mode should be missing some key resources. For example, you spawned in ocean asteroid, with a lot of algae, pacus, water but have no ice biomes and only a very far ice planet. Bartering water/pacus for wolframite/wheezewort via rockets make sense now. Or make some jobs/buildings are gated behind blueprints, and you don't have all blueprints in your world, so you need to bartering Abyssalite mining perk for Doctor perk (transmitting digitally via Printer or Super Computer) or 1 Miner dupe for 8 pacus. I actually wrote a very lengthy post about bartering idea and no one cares but since Klei is going to release ONI officially and making DLC after that, now it's good time to discuss this more seriously

 

 

 

20 hours ago, Nightinggale said:

I can't see gas/liquid movements and temperature updates realistically done in any way other than synchronous

Read my above post. Everyone has their own world, the servers only need to record data about available resources/blueprints/dupes name and stats for bartering. This way, 1 server might serve around 10,000 players. The cost is not very big so skins sale/customized packages could cover that

[Glassyfo]

4 hours ago, Tytan said:

because in this way you could pick up the specific object, and in large quantities

That would require someone to want to give away the finite object, which is what I'm questioning.

[Angpaur]

For a long time I was totally rejecting thought that ONI could work as a multiplayer game, because not only problems with synchronizing all data and calculation across all players but also I didn't see a good way to just let many players manage base at same time.

But I just wondered for a moment how multiplayer in ONI would work if it was like in Don't Starve - you have your own dupe, it has its own inventory, equipment, stats. You move on the map by yourself and build things using stuff in your inventory/equipment. All machines that now needs a dupe to operate should be changed to automatic, so you just need to deliver ingredients and it works by itself, like crock pot in Don't Starve. This will avoid need to do boring stuff like operating a grill for a whole cycle. Manual generators could use hatches instead of dupes and the critters will have to be feed from time to time to make them going.

People who find DST a fun probably would find this to be fun too. I usually don't play multiplayer games but this could change things for me

[Nightinggale]

5 hours ago, camelot said:

Everyone has their own world

But is it really multiplayer if people aren't on the same map?

Besides the trading idea is flawed. The core concept of trading is that one person has a lot of A, but no B. Another person is missing A, but can produce a lot of B and then they exchange goods to make both have enough of both A and B. All players will have the same stuff and missing the same stuff. For a trading idea to work, you would need to make maps without say oil biomes and then you instead get plenty of water, which you then trade with somebody who has plenty of oil, but very little water. I don't like that concept because it forces you to trade or you are screwed. Also I don't consider that to be "playing together".