the pixel grows more discord: discord.gg/P74cVatknc TL;DR: Use Multipaper instead to run and develop Minecraft servers, keep using Paper if you are a entry-level Minecraft server developer or owner and you don't have any other options. People have already suggested using Purpur or Fabric with server-side mods as faster alternatives to Paper. If you're a developer, learn maven shading so you can use third-party libraries for connecting to databases like MySQL, MongoDB, Redis. Use MultiLib in your Minecraft server plugin so it's backwards compatible with regular Minecraft servers, and, there will likely be paid work in anyone who sees this and decides to make their plugins Multipaper compatible.
I'm gonna be honest, most server owners probably don't want to manage a kubernetes cluster just to run a Minecraft server for their friends. This is a solution for large scale smp type servers.
@@_xX_me_Xx_ They don't need a kubernetes cluster, they can copy and paste the folder, change the port values, and deploy more individual multipaper instances on one dedicated machine or VPS
Someone has also suggested, gradle is way easier to learn and much faster, so look at the third-party libraries, look up drivers for things like Redis, MongoDB and learn how to use them in your plugin so you can save data across servers and in more stable ways.
The only problem I see is that everyone forks Paper, which ruins game features - starting with spawn rates and ending with TNT duping and other redstone. So currently I run Fabric servers instead
From my experience with playing on a multipaper server and being slightly involved with the development. it is horribly buggy and a big mess. Folia seems to be alot more stable for multithreaded server. But at the end of the day its all still forking from paper which patches a bunch of things and changes the way minecraft works for you without giving control over it. which is why i prefer fabric servers. however they are not scalable, so for big public servers sadly youll have to deal with paper and its forks
I am glad you saw this and decided to comment. Multipaper is in no way 100% stable, also Multipaper is now a Purpur fork, but, that's a fork of what I suspect is a fork of all the other stuff. I made this video so this can at least convince just one or a couple of people to work on the multipaper project so it's actually plausible to make servers with tons of features. Even servers with 100-200 players will need this tech if they decide to add something like ProjectKorra and tons of different plugins, that's why Multipaper has my endorsement and not Folia, the average developer or new developer isn't going to have the resources or energy to test out something like Folia, while with Multipaper you can just simply copy-paste a folder and change a port and run the same bash/bat script
I don’t think this is good advice. Yes multi paper is good for servers who want like 1 million players in one world. I don’t think hypixel uses anything like multipaper because they just make small mini instances for people which is probably better since multipaper doesn’t have much eco system and it is pretty buggy.
Servers who are big enough to need multi paper probably won’t use public spigot plugins they would probably hire their own devs to make their own stuff.
I made this video to encourage people who are developers to make stuff for multipaper, the whales who have hundreds of players WANT to use Multipaper or Folia but can't because there are no plugins for it. Even though the majority of people watching this video aren't going to be making a plugin or running a server, I am hoping this will convince some of them to use Multilib so the plugin can be drag and dropped into Multipaper instances. If people want to make something like MythicMobs, custom enchants, they completely slow down the main server, if some new developers start making cool stuff, more servers will be inclined to use Multipaper, making it more and more stable as it'll receive more funding and devs. In terms of you saying servers can hire their own devs, HERE'S THE PROBLEM! It's hard to find competent Minecraft server plugin developers, especially ones that are experienced in horizontal scaling, by making horizontal scale a default servers might go an give money to those developers to fix up the plugin if there are any issues and will go and give to the Multipaper project. Multilib makes it backwards compatible with single-instance servers. Horizontal scaling should be a default when adding content into the game, servers like originrealms, and other servers that modify the custom game will run into TONS of optimization issues, this is the only real solution when a server wants to add tons of custom content. You are right that they can hire developers but there won't be as many to take the job for a low price, so most servers would give up at that point, I made this video to give the server admin/developer community a push in this direction.
I always wanted performance for when I mod Minecraft sure there's sodium and it's numerous add-ons. However the game is bottlenecked and there are hardly any mods that can permanently resolve that issue. Folia does help with the issues but i would like a permanent solution.
Multipaper is both more difficult and more easier to develop for, because in order to truly test Multipaper you need to have multiple people, but, when testing anything that's multiplayer you need to have more than one person to test it. I made this video to suggest to people to use Multipaper, because, it isn't exactly as easy to horizontal scale but, if for people who actually want to see their work being used and experienced by a ton of players, they'll want to make Multipaper plugins. If enough people do it, more server owners will use Multipaper. It would also be cool if Fabric did give the ability for infinite fabric servers to power one map.
I'm really not sure what the objective is with this video, have you actually ran a large server on multipaper? Multipaper has a lot of issues vs Folia, there's a reason 2b2t are using Folia. There is little benefit to sharding vs regionalised ticking which is what folia has, and sharding is much much more complexity (and leads to the issues). Folia also doesn't NEED 16 threads to run for development, it's telling end users that if they don't have 16 cores there's no point using folia vs paper. Telling people to develop not to develop for PaperAPI is a little weird, Folia (and multipaper but there's is quite a bit more broken) is pretty much the same api with the addition that you need to account for the schedulers and some minor quirks. Also chatgpt is terrible for plugin dev, especially went you need to account for parallel execution. As for the history, Spigot is still updated and they maintain & patch Bukkit, with Paper applying patches onto that and Purpur the same again. Purpur is fine but shouldn't be advertised as a performance fork, it's mostly for customizability. No hate though! Just had someone post this video while asking about server scaling and it was a little concerning
ShreddedPaper I heard might replace or become Multipaper, as it'll eventually be a hybrid between Folia and Multipaper but you won't be able to dynamically launch 20 servers and will need to manually configure things. This video was originally made to encourage more people to make plugins to scale so things would advance further. We are in the middle of getting Pixelgrew converted over to Multipaper. One of the main reasons I recommended Multipaper is a lot of people starting out might use DigitalOcean or Vultr instead of one of the big dedicated server providers, especially since DO and Vultr offer a $200 credit. In general this video was made because there was almost literally no plugins for Multipaper, one of the most important things to a Minecraft server is custom content and I was hoping this video would inspire some people. Sorry for not responding immediately and currently I'm busy covering the MrBeast saga on the channel, and ironically: MrBeast paid a group of people to make their own Multipaper before Multipaper even existed, none of it was adapted into anything, but, this video was to encourage people to learn to develop plugins to scale, because most of the time, if a server owner tries your plugin, it'll just be seen as another plugin that destroys the TPS and they'll move on, designing things to scale may encourage server owners to pay the plugin developers to develop those features, which, there aren't enough developers out there who are capable of putting together plugins with scaling in mind, and it would be a pain for existing plugins to adapt to it.
A few things: Title is a bit.. meh. a tad misleading. 5:39 Gradle is better, much faster, and imo (because it isn't xml), it's much easier to read and extend. Folia is only good for really large single server setups, the only reason to use Multipaper is if you have a large player base, are running a single world and don't mind dealing with bugs. Paper is great for small scale servers where there are not that many people, and where the players are not highly technical (unless you disable patches, sadly limiting still). If you care about the vanilla experience, have a small server for friends or something, use fabric. If you don't want to run just one world with huge player bases, you'd want to look into dynamic servers.
People should stop using PaperMC once Multipaper actually has plugins, the video title is one half of the whole suggestion, people should stop using it once horizontal scaling is achieved. I say in the pinned comment and the first 2-3 minutes, "keep using paper if you are starting out," It is clickbait to a degree but I made it to get the attention of people who develop plugins or want to, because, I wanted to get the word out and I know plugin developers will click on the video. I only suggest maven because, the other top tutorials I skimmed through suggest making a maven project when making a plugin, and I think people are more familiar with it. Gradle is probably better.
@@cubedimensions People should still use paper, even when Multipaper actually has plugins. Horizontal scaling only matters for a very select amount of people. Most people just want to run a server to play on with their friends. (I suppose that may be what you were trying to get at with the just starting out part, but I still think it should be clarified that there are very different use cases for each piece of software)
Hopefully devs will make their plugin support multipaper and for it to be backwards compatible, that's why I say to use multilib in the video and to learn to support external data storage on a plugin, so whales and servers that scale can simply deploy more instances. And I said at the start of the video and in the top comment "if you're just starting out as a dev or you're brand new to running servers, just keep using paper since it's the most supported right now", the title might be a little infuriating to look at but I hope this video gets into the right hands.
Hey! Just thought id jump in here with some of my thoughts and experiences beeing a serveradmin for a big server back in the day and having contacts to many in the serverspace. Multipaper has it merrit, but the biggest issue it will have is error correcting. Unlike docker where it makes its own internal switches multipaper goes over the net. Unless you set it up with dockers wich in that case would require you to have multiple servers running and thus driving up running costs. Ime usually servers either have a couple of people online or thousands. So i only see multipaper beeing usefull for bigger servers, while folia in theory could be usefull everywhere. In either case im glad you didnt go the route of shitting on paper and just explainer urself, big ups my dude
I literally see your videos recommended all the time and briefly skimmed some of the stuff you made before making this video, thanks for providing refreshers and summaries of what went on with Forge/Neoforged! The title is clickbait for a reason: this video was targeted towards Paper people, and I advise developing for Multipaper over Folia, for multiple reasons, Folia has much higher hardware requirements, while Multipaper has minimal hardware requirements, you don't need to build Multipaper as a developer starting out while you need to do so for paper. There isn't enough people who know how to make stuff for Multipaper so I made this video knowing a couple geniuses might see this and will develop some custom content for it. The use I see in Multipaper is the ability to put tons of custom content onto Minecraft servers, as the single-core experience is very limited. Server owners starting out might run a Pebblehost server or something, but, later on they can just simply copy and paste some folders, set up some firewall rules to whitelist the Multipaper Instances, and just keep spinning up more or bigger machines. Thanks for commenting and it's definitely made my afternoon
@@cubedimensions hey man youtube is an ugly beast, clickbait is what you need to do! And im glad ur spreading some positivity instead of shitting all over existing projects. Keep it up!
@@Cygnus_MC I kinda trashed Hytale though :) Also to correct myself: I said you don't need to build for Multipaper, you do need to build for Folia, that's why I prefer people make stuff with multipaper over folia, it's easier to spin up more VPS's/VMs and docker containers for people who have started out, however most people should just get a dedicated machine
@cubedimensions hytale deserves to be trashed so the fanboys stop gushing over every word the devteam spews out lol. Tho yeah that makes sence, the issue i have is not every little timmy has the place/money for a dedi machine
@@Cygnus_MC Well I am glad you get what I am saying, people can pay $1-$2 per month for a powerful VPS in the U.S. on Racknerd or pay $4-$5 per VPS, while it shouldn't be what is used in truly final and heavily optimized production this is what those people can afford and that's why this video was made. Server owners who just start out will either use minehut/aternos, then later on might get tempted in by the free $200 credit DigitalOcean, Linode, Vultr offers. I imagine people would rather go with the $200 free credit over 60-90 days for just spending $5 than to go with any dedicated server. It's more painful but also more rewarding to develop for an infinitely scaling system, that's the only way I see a server with tons of custom content truly be able to handle all what is there. The big and unmanaged dedicated server providers are what they should truly go with, however, sometimes managed hosting is better as maintenance and security can be much better, for example paying for a managed database they automatically bring up to date for you, however to lower costs, unmanaged dedicated servers are truly the most efficient for players in the hundreds. Too bad Hetzner is forced to mistreat actual people, who have a reputation for declining service to people who are even honest.
Fabric is really good too. It really depends on what you are doing and what kind of server you are going for. Just wish more people would recognize that
@@redouble_ What matters is if you can the current run Minecraft server plugins on it, Fabric is probably well maintained, I am suggesting Multipaper in this video to encourage horizontal server sharding and to spread the idea, maybe some genius will make "MultiFabric" or something that can load up Spigot plugins and Fabric mods
Fabric is best for modded/vanilla-like/technical minecraft. If people want a server software for factions, minigame or prison, then the bukkit ecosystem can support that better. It's a matter of recency. Fabric is better architecturally, but the bukkit ecosystem has just been around for longer. Once fabric has enough years behind it, this might change. And also the idea of a multipaper type mod for fabric would be pretty neat.
@@bennyboiii1196 If they do that and also can make some type of Bukkit/Spigot loader that would be awesome, I made this video to try to keep pushing for people to make progress on horizontal scaling, because that's what truly allows for infinite scaling and people who want to expand the player slots of a modded server can just keep launching more instances and deploying more machines
the more people who use it, the more mistakes made publicly, meaning the more demand there are to fix the problems no one knows existed or could be alone if it happens and they are the only ones using multipaper
This comment you're reading isn't related to this video but on something else important about your channel. I have been getting an issue where i haven't been seeing your videos being shown in my sub box or even getting notifications about it, which is weird since i legit have your channel subscribed with notifications turn on. The only thing i see anything related to you are your community posts in my sub box. Im not sure if anyone else is having this issue or if its just me.
This isn't TH-cam, they are known for messing things up but I specifically check the box when uploading my videos to not notify subscribers, instead TH-cam will notify subscribers who are most interested, I learned this when I originally made TH-cam shorts, if you make videos too often they will "wait" sometimes indefinitely before pushing your videos out to any new subscriber. Right now, the traffic on this video is mostly subscribers despite me unchecking notify subscribers, I uncheck the box on purpose to notify subs because, the videos are designed for the algorithm and if the video is a good video but is a flop to my subscribers (including my former short-form subs) it will likely not get pushed out, killing the channel. Most of the subs on this channel are short-form subscribers, so, I have set the videos to not notify any subscribers whenever I upload, because, if any of my low-attention span subscribers from my previous 200 short-forms see this stuff, they will likely not click it or stay through the video at all, and youtube will look at those subs and say "hey if his subscribers aren't going to watch his video a good way through and aren't even clicking on it, let's not even bother suggesting the video" this isn't always the case, but, this is on purpose, and my videos will show up in your feed as long as youtube detects you are likely to watch my videos the full-way through.
@@AngryBirdsFanboy06 The funny thing looking through my comments, only one other person complained I didn't show up in their feed. TH-cam and other social media platforms trial-test you through subs, so turning off follower/subscriber notifications is important for sub-genre people like me who are split out and are all over the place, maybe a celebrity can keep that box checked but I can't because I need to grow.
I have never used telegram before however I know it's massive, the thumbnail is a mix of the Paper logo and the Multipaper logo (I jokingly call it Multipurpur now)
i hate paper cuz they just patch a bunch of vanilla game features for no reason not for optimization just because you can enable them and get the same performance but they still dont work the same when enabled
Nah purple has the same issue, I use it at the moment because it can run well but I would prefer to use fabric but I can’t because it doesn’t have good server optimisation
My Minecraft server, Pixelgrew, runs into lag issues, we want to use Multipaper but there are no plugins for it, we'd love to use ProjectKorra and would love to add more features but it slows the server down significantly. We aren't heavily populated yet. I made this video because I wanted to see more plugins that supports Multipaper. I have spoken to the owner of Multipaper however, the guy hasn't even made more than $100 in donations despite being the only other person supporting a bleeding edge scaling tech. More people will donate to Multipaper when more people have a use for it, more people will be able to use it if there is more plugins.
Pixelgrew which is my server is blowing up but our TPS is dropping to 10 TPS at 20 players and minimal plugins despite us running the server on what I think is a Ryzen dedicated server, modern multiplayer without Multipaper or Folia is a joke now.
@@cubedimensions ik, but what players can do to deal with this plugins, PaperMC advertise himself as "vanilla", while this plugin is completely changing the way that minecraft works. Redstone builds,mob spawning machanics etc. etc. some of this options don't even affect optimization of the game that much, but servers owners still have them turned off, cuz it's in default options of this plugin. I understand the concept of optimization, but for me it's feels like this papermc creators have deep in their ass what ppl think about this plugin.
@@123-w2l9t papermc and purpur doesn't matter that much to me anymore, multipaper and folia are what people need to start developing for, we can't even handle 20 players at full TPS on our server with minimal plugins, imagine 100 players on 1.21 paper? Absolutely pathetic. Developers need to start making multipaper stuff now
Multipaper to my opinion would be the best option out of all of this because, it's easier for developers to deploy 3-4 multipaper instances than it is to have a literal 16 core processor on hand which is the minimum for running Folia. Anyone with an average PC can develop a server plugin, for Folia or Multipaper, they'll either need something beefy and people to help them test the multi-server capability
You don't need to use kubernetes but, yeah you can do that. Multipaper is going to be reworked I've heard back into Folia, where the Folia instances can be linked together, so it will still be both vertical and horizontal scaling at the end of the day. Horizontal scaling is always going to be the way to go because while it has diminishing returns with some instability. It has an infinite performance cap.
the pixel grows more
discord: discord.gg/P74cVatknc
TL;DR: Use Multipaper instead to run and develop Minecraft servers, keep using Paper if you are a entry-level Minecraft server developer or owner and you don't have any other options. People have already suggested using Purpur or Fabric with server-side mods as faster alternatives to Paper. If you're a developer, learn maven shading so you can use third-party libraries for connecting to databases like MySQL, MongoDB, Redis. Use MultiLib in your Minecraft server plugin so it's backwards compatible with regular Minecraft servers, and, there will likely be paid work in anyone who sees this and decides to make their plugins Multipaper compatible.
I'm gonna be honest, most server owners probably don't want to manage a kubernetes cluster just to run a Minecraft server for their friends. This is a solution for large scale smp type servers.
@@_xX_me_Xx_ They don't need a kubernetes cluster, they can copy and paste the folder, change the port values, and deploy more individual multipaper instances on one dedicated machine or VPS
Someone has also suggested, gradle is way easier to learn and much faster, so look at the third-party libraries, look up drivers for things like Redis, MongoDB and learn how to use them in your plugin so you can save data across servers and in more stable ways.
The only problem I see is that everyone forks Paper, which ruins game features - starting with spawn rates and ending with TNT duping and other redstone.
So currently I run Fabric servers instead
From my experience with playing on a multipaper server and being slightly involved with the development. it is horribly buggy and a big mess. Folia seems to be alot more stable for multithreaded server.
But at the end of the day its all still forking from paper which patches a bunch of things and changes the way minecraft works for you without giving control over it. which is why i prefer fabric servers. however they are not scalable, so for big public servers sadly youll have to deal with paper and its forks
I am glad you saw this and decided to comment. Multipaper is in no way 100% stable, also Multipaper is now a Purpur fork, but, that's a fork of what I suspect is a fork of all the other stuff. I made this video so this can at least convince just one or a couple of people to work on the multipaper project so it's actually plausible to make servers with tons of features. Even servers with 100-200 players will need this tech if they decide to add something like ProjectKorra and tons of different plugins, that's why Multipaper has my endorsement and not Folia, the average developer or new developer isn't going to have the resources or energy to test out something like Folia, while with Multipaper you can just simply copy-paste a folder and change a port and run the same bash/bat script
I don’t think this is good advice.
Yes multi paper is good for servers who want like 1 million players in one world. I don’t think hypixel uses anything like multipaper because they just make small mini instances for people which is probably better since multipaper doesn’t have much eco system and it is pretty buggy.
Servers who are big enough to need multi paper probably won’t use public spigot plugins they would probably hire their own devs to make their own stuff.
I made this video to encourage people who are developers to make stuff for multipaper, the whales who have hundreds of players WANT to use Multipaper or Folia but can't because there are no plugins for it. Even though the majority of people watching this video aren't going to be making a plugin or running a server, I am hoping this will convince some of them to use Multilib so the plugin can be drag and dropped into Multipaper instances. If people want to make something like MythicMobs, custom enchants, they completely slow down the main server, if some new developers start making cool stuff, more servers will be inclined to use Multipaper, making it more and more stable as it'll receive more funding and devs.
In terms of you saying servers can hire their own devs, HERE'S THE PROBLEM! It's hard to find competent Minecraft server plugin developers, especially ones that are experienced in horizontal scaling, by making horizontal scale a default servers might go an give money to those developers to fix up the plugin if there are any issues and will go and give to the Multipaper project. Multilib makes it backwards compatible with single-instance servers. Horizontal scaling should be a default when adding content into the game, servers like originrealms, and other servers that modify the custom game will run into TONS of optimization issues, this is the only real solution when a server wants to add tons of custom content.
You are right that they can hire developers but there won't be as many to take the job for a low price, so most servers would give up at that point, I made this video to give the server admin/developer community a push in this direction.
I always wanted performance for when I mod Minecraft sure there's sodium and it's numerous add-ons. However the game is bottlenecked and there are hardly any mods that can permanently resolve that issue. Folia does help with the issues but i would like a permanent solution.
Multipaper is both more difficult and more easier to develop for, because in order to truly test Multipaper you need to have multiple people, but, when testing anything that's multiplayer you need to have more than one person to test it. I made this video to suggest to people to use Multipaper, because, it isn't exactly as easy to horizontal scale but, if for people who actually want to see their work being used and experienced by a ton of players, they'll want to make Multipaper plugins. If enough people do it, more server owners will use Multipaper. It would also be cool if Fabric did give the ability for infinite fabric servers to power one map.
I'm really not sure what the objective is with this video, have you actually ran a large server on multipaper? Multipaper has a lot of issues vs Folia, there's a reason 2b2t are using Folia. There is little benefit to sharding vs regionalised ticking which is what folia has, and sharding is much much more complexity (and leads to the issues). Folia also doesn't NEED 16 threads to run for development, it's telling end users that if they don't have 16 cores there's no point using folia vs paper.
Telling people to develop not to develop for PaperAPI is a little weird, Folia (and multipaper but there's is quite a bit more broken) is pretty much the same api with the addition that you need to account for the schedulers and some minor quirks. Also chatgpt is terrible for plugin dev, especially went you need to account for parallel execution.
As for the history, Spigot is still updated and they maintain & patch Bukkit, with Paper applying patches onto that and Purpur the same again. Purpur is fine but shouldn't be advertised as a performance fork, it's mostly for customizability.
No hate though! Just had someone post this video while asking about server scaling and it was a little concerning
ShreddedPaper I heard might replace or become Multipaper, as it'll eventually be a hybrid between Folia and Multipaper but you won't be able to dynamically launch 20 servers and will need to manually configure things. This video was originally made to encourage more people to make plugins to scale so things would advance further. We are in the middle of getting Pixelgrew converted over to Multipaper. One of the main reasons I recommended Multipaper is a lot of people starting out might use DigitalOcean or Vultr instead of one of the big dedicated server providers, especially since DO and Vultr offer a $200 credit.
In general this video was made because there was almost literally no plugins for Multipaper, one of the most important things to a Minecraft server is custom content and I was hoping this video would inspire some people. Sorry for not responding immediately and currently I'm busy covering the MrBeast saga on the channel, and ironically: MrBeast paid a group of people to make their own Multipaper before Multipaper even existed, none of it was adapted into anything, but, this video was to encourage people to learn to develop plugins to scale, because most of the time, if a server owner tries your plugin, it'll just be seen as another plugin that destroys the TPS and they'll move on, designing things to scale may encourage server owners to pay the plugin developers to develop those features, which, there aren't enough developers out there who are capable of putting together plugins with scaling in mind, and it would be a pain for existing plugins to adapt to it.
Let's go, another great advice from you!
A few things:
Title is a bit.. meh. a tad misleading.
5:39 Gradle is better, much faster, and imo (because it isn't xml), it's much easier to read and extend.
Folia is only good for really large single server setups, the only reason to use Multipaper is if you have a large player base, are running a single world and don't mind dealing with bugs.
Paper is great for small scale servers where there are not that many people, and where the players are not highly technical (unless you disable patches, sadly limiting still).
If you care about the vanilla experience, have a small server for friends or something, use fabric.
If you don't want to run just one world with huge player bases, you'd want to look into dynamic servers.
People should stop using PaperMC once Multipaper actually has plugins, the video title is one half of the whole suggestion, people should stop using it once horizontal scaling is achieved. I say in the pinned comment and the first 2-3 minutes, "keep using paper if you are starting out,"
It is clickbait to a degree but I made it to get the attention of people who develop plugins or want to, because, I wanted to get the word out and I know plugin developers will click on the video.
I only suggest maven because, the other top tutorials I skimmed through suggest making a maven project when making a plugin, and I think people are more familiar with it. Gradle is probably better.
@@cubedimensions People should still use paper, even when Multipaper actually has plugins. Horizontal scaling only matters for a very select amount of people. Most people just want to run a server to play on with their friends. (I suppose that may be what you were trying to get at with the just starting out part, but I still think it should be clarified that there are very different use cases for each piece of software)
Hopefully devs will make their plugin support multipaper and for it to be backwards compatible, that's why I say to use multilib in the video and to learn to support external data storage on a plugin, so whales and servers that scale can simply deploy more instances.
And I said at the start of the video and in the top comment "if you're just starting out as a dev or you're brand new to running servers, just keep using paper since it's the most supported right now", the title might be a little infuriating to look at but I hope this video gets into the right hands.
@@cubedimensions Yeah, I get that, and I do hope that the video has the intended effect, because multipaper could be great.
Hey! Just thought id jump in here with some of my thoughts and experiences beeing a serveradmin for a big server back in the day and having contacts to many in the serverspace.
Multipaper has it merrit, but the biggest issue it will have is error correcting. Unlike docker where it makes its own internal switches multipaper goes over the net. Unless you set it up with dockers wich in that case would require you to have multiple servers running and thus driving up running costs.
Ime usually servers either have a couple of people online or thousands. So i only see multipaper beeing usefull for bigger servers, while folia in theory could be usefull everywhere.
In either case im glad you didnt go the route of shitting on paper and just explainer urself, big ups my dude
I literally see your videos recommended all the time and briefly skimmed some of the stuff you made before making this video, thanks for providing refreshers and summaries of what went on with Forge/Neoforged!
The title is clickbait for a reason: this video was targeted towards Paper people, and I advise developing for Multipaper over Folia, for multiple reasons, Folia has much higher hardware requirements, while Multipaper has minimal hardware requirements, you don't need to build Multipaper as a developer starting out while you need to do so for paper. There isn't enough people who know how to make stuff for Multipaper so I made this video knowing a couple geniuses might see this and will develop some custom content for it.
The use I see in Multipaper is the ability to put tons of custom content onto Minecraft servers, as the single-core experience is very limited. Server owners starting out might run a Pebblehost server or something, but, later on they can just simply copy and paste some folders, set up some firewall rules to whitelist the Multipaper Instances, and just keep spinning up more or bigger machines.
Thanks for commenting and it's definitely made my afternoon
@@cubedimensions hey man youtube is an ugly beast, clickbait is what you need to do! And im glad ur spreading some positivity instead of shitting all over existing projects. Keep it up!
@@Cygnus_MC I kinda trashed Hytale though :) Also to correct myself: I said you don't need to build for Multipaper, you do need to build for Folia, that's why I prefer people make stuff with multipaper over folia, it's easier to spin up more VPS's/VMs and docker containers for people who have started out, however most people should just get a dedicated machine
@cubedimensions hytale deserves to be trashed so the fanboys stop gushing over every word the devteam spews out lol. Tho yeah that makes sence, the issue i have is not every little timmy has the place/money for a dedi machine
@@Cygnus_MC Well I am glad you get what I am saying, people can pay $1-$2 per month for a powerful VPS in the U.S. on Racknerd or pay $4-$5 per VPS, while it shouldn't be what is used in truly final and heavily optimized production this is what those people can afford and that's why this video was made. Server owners who just start out will either use minehut/aternos, then later on might get tempted in by the free $200 credit DigitalOcean, Linode, Vultr offers. I imagine people would rather go with the $200 free credit over 60-90 days for just spending $5 than to go with any dedicated server.
It's more painful but also more rewarding to develop for an infinitely scaling system, that's the only way I see a server with tons of custom content truly be able to handle all what is there.
The big and unmanaged dedicated server providers are what they should truly go with, however, sometimes managed hosting is better as maintenance and security can be much better, for example paying for a managed database they automatically bring up to date for you, however to lower costs, unmanaged dedicated servers are truly the most efficient for players in the hundreds. Too bad Hetzner is forced to mistreat actual people, who have a reputation for declining service to people who are even honest.
Hot take: fabric is best
Multipaper can infinitely scale, Fabric can't
Fabric is really good too. It really depends on what you are doing and what kind of server you are going for. Just wish more people would recognize that
@@redouble_ What matters is if you can the current run Minecraft server plugins on it, Fabric is probably well maintained, I am suggesting Multipaper in this video to encourage horizontal server sharding and to spread the idea, maybe some genius will make "MultiFabric" or something that can load up Spigot plugins and Fabric mods
Fabric is best for modded/vanilla-like/technical minecraft. If people want a server software for factions, minigame or prison, then the bukkit ecosystem can support that better. It's a matter of recency. Fabric is better architecturally, but the bukkit ecosystem has just been around for longer. Once fabric has enough years behind it, this might change. And also the idea of a multipaper type mod for fabric would be pretty neat.
@@bennyboiii1196 If they do that and also can make some type of Bukkit/Spigot loader that would be awesome, I made this video to try to keep pushing for people to make progress on horizontal scaling, because that's what truly allows for infinite scaling and people who want to expand the player slots of a modded server can just keep launching more instances and deploying more machines
The amount of edge cases and possible race conditions when working with a multithreaded server raise exponentially. For most people, not worth it.
the more people who use it, the more mistakes made publicly, meaning the more demand there are to fix the problems no one knows existed or could be alone if it happens and they are the only ones using multipaper
This comment you're reading isn't related to this video but on something else important about your channel.
I have been getting an issue where i haven't been seeing your videos being shown in my sub box or even getting notifications about it, which is weird since i legit have your channel subscribed with notifications turn on.
The only thing i see anything related to you are your community posts in my sub box.
Im not sure if anyone else is having this issue or if its just me.
This isn't TH-cam, they are known for messing things up but I specifically check the box when uploading my videos to not notify subscribers, instead TH-cam will notify subscribers who are most interested, I learned this when I originally made TH-cam shorts, if you make videos too often they will "wait" sometimes indefinitely before pushing your videos out to any new subscriber. Right now, the traffic on this video is mostly subscribers despite me unchecking notify subscribers, I uncheck the box on purpose to notify subs because, the videos are designed for the algorithm and if the video is a good video but is a flop to my subscribers (including my former short-form subs) it will likely not get pushed out, killing the channel.
Most of the subs on this channel are short-form subscribers, so, I have set the videos to not notify any subscribers whenever I upload, because, if any of my low-attention span subscribers from my previous 200 short-forms see this stuff, they will likely not click it or stay through the video at all, and youtube will look at those subs and say "hey if his subscribers aren't going to watch his video a good way through and aren't even clicking on it, let's not even bother suggesting the video" this isn't always the case, but, this is on purpose, and my videos will show up in your feed as long as youtube detects you are likely to watch my videos the full-way through.
@@cubedimensions oh, i didn't know that before, thanks!
@@AngryBirdsFanboy06 The funny thing looking through my comments, only one other person complained I didn't show up in their feed.
TH-cam and other social media platforms trial-test you through subs, so turning off follower/subscriber notifications is important for sub-genre people like me who are split out and are all over the place, maybe a celebrity can keep that box checked but I can't because I need to grow.
Not a dev but I gotta say the yap session was just so interesting I had to watchall the way.
that's why I ask people to 2x speed my videos and I thought I went on for quite a while but decided to leave it in cause why not
@@cubedimensions didn't even use 2x
PaperMc logo looks like telegram X logo
I have never used telegram before however I know it's massive, the thumbnail is a mix of the Paper logo and the Multipaper logo (I jokingly call it Multipurpur now)
WOW it does, I just looked it up
i hate paper cuz they just patch a bunch of vanilla game features for no reason not for optimization just because you can enable them and get the same performance but they still dont work the same when enabled
That's why people prefer Purpur I believe
Nah purple has the same issue, I use it at the moment because it can run well but I would prefer to use fabric but I can’t because it doesn’t have good server optimisation
Then paper is clearly not for you. Not sure why you'd hate it, just don't use it.
nah mate paper is a gem iam a server owner with 100's players I know how good it is
@@0269_m cope u give your players an incomplete experience
Instructions unclear, paper printed my servers
looks like a multipaper ad lol
My Minecraft server, Pixelgrew, runs into lag issues, we want to use Multipaper but there are no plugins for it, we'd love to use ProjectKorra and would love to add more features but it slows the server down significantly. We aren't heavily populated yet. I made this video because I wanted to see more plugins that supports Multipaper. I have spoken to the owner of Multipaper however, the guy hasn't even made more than $100 in donations despite being the only other person supporting a bleeding edge scaling tech. More people will donate to Multipaper when more people have a use for it, more people will be able to use it if there is more plugins.
with not multi-folia?
"my sequel"
i have an attention span :>
then watch the whole video so it gets recommend more :)
Idea for next video: How to deal with PaperMC as a player on server.
Pixelgrew which is my server is blowing up but our TPS is dropping to 10 TPS at 20 players and minimal plugins despite us running the server on what I think is a Ryzen dedicated server, modern multiplayer without Multipaper or Folia is a joke now.
@@cubedimensions ik, but what players can do to deal with this plugins, PaperMC advertise himself as "vanilla", while this plugin is completely changing the way that minecraft works. Redstone builds,mob spawning machanics etc. etc. some of this options don't even affect optimization of the game that much, but servers owners still have them turned off, cuz it's in default options of this plugin. I understand the concept of optimization, but for me it's feels like this papermc creators have deep in their ass what ppl think about this plugin.
@@123-w2l9t papermc and purpur doesn't matter that much to me anymore, multipaper and folia are what people need to start developing for, we can't even handle 20 players at full TPS on our server with minimal plugins, imagine 100 players on 1.21 paper? Absolutely pathetic. Developers need to start making multipaper stuff now
Pumpkin might solve everything in a few years. I'm currently trying to contribute daily to it :D
Meh, complete rewrite is bs, will make big delays for every time Minecraft receives an update.
ngl i just want something that supports plugin and at the same time not break farms whilst having great performance.
Multipaper to my opinion would be the best option out of all of this because, it's easier for developers to deploy 3-4 multipaper instances than it is to have a literal 16 core processor on hand which is the minimum for running Folia. Anyone with an average PC can develop a server plugin, for Folia or Multipaper, they'll either need something beefy and people to help them test the multi-server capability
You still missed the main point. Multi paper is still paper which means it breaks most farms
@@walksanator actually not anymore if I recall I mentioned in this video: multipaper is actually a purpur fork now
@@cubedimensions really so purpur straps out all the vanilla dupe fixes. Switches the configuration to match vanilla?
@@walksanator I have no idea but Purpur is considered more customizable and optimal.
Don’t mind me, just helping with the algorithm.
watching the video fully on 1x speed is some of the best help too!
Well nowww you tell me xD
broment
Basically learn k8s.
You don't need to use kubernetes but, yeah you can do that. Multipaper is going to be reworked I've heard back into Folia, where the Folia instances can be linked together, so it will still be both vertical and horizontal scaling at the end of the day.
Horizontal scaling is always going to be the way to go because while it has diminishing returns with some instability. It has an infinite performance cap.