Star Citizen’s Server Meshing is Working! What Next? (Ft. Grolo) | Launch Sequence Podcast

[@Caledas]

Yeah Grolo! Great conversation I think those of us that have been following this for some time forget how massive the goal truly was in the beginning and still is, so it’s easy to be like “come on already”. This helps put in the perspective that it has made huge leaps, even if there are more needed to get it where we can play this full time.

[@Why-nr3gn]

I'm just all in. Untill I die or SC servers stop

[@CitizenScott]

I mean CIG are basically the ones saying "come on already" now. That's why the goal of a Global Shard is gone, and they're not even aiming for unified Regional Shards. The entire stripped down 1.0 push is their version of "come on already."

[@InfiniteWatermelon]

Hey Tomato! Thanks for bringing Grolo in. Very cool episode. Hope we can see more!

[@AlbertoMartinez765]

YES!!! Finally these TWO Meet UP

[@kevinm3751]

Grolo is solid, thank you for giving him a bigger platform. Just when I think I am following all the greats for information you seem to find the ones I did not know about!

[@cygnus6623]

Commodore 64...same here. That is where I started and on to PC's in the late eighties/early nineties. Led me to a career in IT after half a life in construction. Star Citizen is it for me now...since 3.9. 4.0 is exciting times for sure. I think we are on the brink of some pretty significant growth in the Verse.

[@stuartriddell2461]

KZbin just told me that I watched more space tomato too than any other channel this year. I wasn't even subscribed. I have subbed now.

[@pallmsunday6667]

Wow my two fav star citizen guys did a collab! great stuff here thanks!

[@CosmicD]

47:30 yes it really caught me by surprise how quickly Jared wrapped this episode up at the end. no "Disco Lando" style motivational speech or glimpse into the future like he did last year. Now there is a huge veil covering what we might see next year, as if CiG aren't quite confident of which aspects of the 1.0 row will be coming to us next year and they have to sort that out. Optimistically, From what I gathered, we shouldn't read the 1.0 row as something that will be held back until 1.0 be released, but the features in that row will be worked on and distributed with 4.x patches, until they are at a point where all these milestones are complete and then they can call it 1.0. If they can implement engineering, life support, actor status 2, the social app, spawning ground vehicles in freight elevators and the first iteration of crafting, that would be nice. But I do understand they first needed to have a stable version of 4.0 / server meshing running, before they can go from there and add new things that heavily rely on this feature.

[@ValentinoDagher]

I really think space tomato has the best star citizen podcasts. I can tell in each one he really researched ahead of time and doesn't have an agenda of pushing his point across but rather allow other perspectives of the game to shine through.

[@fwdcnorac8574]

Mrs. Tomato shadow posting.

[@Bubblegum_Ronin]

Thanks Tomato for this.. 2 of my favorite SC content creators chatting about my favorite space game was perfection.

[@j.d.4697]

What a nice Christmas send-off for the devs.

[@Solice07]

Awesome to see Grolo on here!

[@mikezr1000]

28:28 I think I've been experiencing this in the current build for 4.0 a lot, this explains why I randomly get taken out of QT.

[@VFraunhofer]

It's true that some games made a technology alike server meshing, like did dual universe whose goals are close to Star citizen ones ( kzbin.info/www/bejne/pn3LY4Klic9rrMU ), but I guess every game faces its own challenges, like it's far from Star Citizen on many ways, there's no point in my opinion comparing it, for what we care. It's still impressive seeing it work that smoothly in the case of SC and that's good, especialy for a first phase, that's enthusiasming for future. As for developping time as stated in few comments, this is in my opinion a game that is as ambitious as a GTA VI, with a budget growing closer to it every year, and technologies not developped primarily as were Rockstar Games able to do throught all its history. So honnestly, I'm not surprised or think it's a downside and even less that it's a strategy for making more money as I did read in the comments. I arrived not long ago (3.23) and from this side the game and its state is satisfying for me, even if there's still a lot of work. Heard a lot about it being a scam, and it took me few hours to definitely be stating it's not, it's just a really ambitious one and that's for the best

[@otockian]

Yeah, those tech demos never actually amounted to anything real which should throw a huge red flag to anyone looking at it as a good example.

[@MrRoblcopter]

DU is more like a hybrid peer2peer network, more akin to Elites networking solution. That 30k video is also 30k "simulated" players, that's like saying "hey look I can spawn in 30,000 NPCs."

[@VFraunhofer]

@@MrRoblcopter Well I played it and actually there's a lot of people since very long. Also this test wasn't totaly NPCs (the method isn't the same as just adding NPC), it was simulated to emulate 30k people in the best way outside having 30k real persons that is virtualy unpossible for obvious reasons. Could you explain to me what is the difference ? Because as far as I know, server meshing is the ability to split charge to multiple servers for the same instance, that's actualy what it does, on techno side it's forcibly a bit different from one to another

[@VFraunhofer]

​@@otockian Wut ?... This is a game that exists and is playable, it's been long you have ton of people there playing together... Why everything related to SC as to become a "D*ck's competition" seriously that's crazy. you had giant ships concessions there while you weren't able to take off the station in Sc, doesn't makes it a better or less good game... It's just not the same game, not the same technos, so indeed SC's system is prolly unique as DU solutions (that ARE implemented and working since long) were as well in it's way, for a comparable endgoal

[@MrRoblcopter]

@VFraunhofer that's the thing, in DU there is only one overhead server making players connect with each other. In fact if you go and look for mentions of server meshing, everything is scrubbed, not even their website mentions it anymore, even an NDA has been implemented apparently. The lead dev was even touting that they implemented meshing far before CIG did, that post has also been scrubbed. Also if since you played it then you should have noticed that if there were more than 20-30 player in an area massive desync started showing up, the same exact thing happens in Elite at about 40-50 players. P2P doesn't scale well at all. Also, simulated players will never match an actual real player, that's like doing some back of the napkin carpentry design, then when you try to build it, you find out that that it is possible, only on a piece of paper because the drawing doesn't have physics. CIG can, and have, used simulated players before, but they understand that the simulation is no match for reality.

[@sjoer]

I'm also a heavy MMO player, what I like about Grolo is that we agree in arguments where SC chat thinks I'm wrong :D

[@sjoer]

But we all know, gamers are plain stupid!

[@sjoer]

I also have a background in server maintenance (telco/callcenter), programming, electronics and game design :D

[@sjoer]

@19:00 the architecture that Star Citizen is using is not like that! All the entities are stored in the replication layer, you connect to the replication layer and a server behind the replication layer becomes the authority over your entities plus whatever is around you! The replication layer is in essence a database your client gets its data from, but that database is controlled by multiple dedicated game servers behind it.

[@sjoer]

Basically what Grolo said at @26:45

[@sjoer]

What Grolo said about tooling is 100% true, when I was doing game design I found myself making tools for everything imaginable... the problem is also that the person who needs to use the tool is not very capable of using anything but tools.

[@alexvdl1]

This is probably the best launch sequence episode I’ve watched so far. Super interesting!

[@toxicityD]

Star Citizen's server meshing is some HYPE technology. It's very unique in the gaming industry and even unique among web connected applications.

[@ErikLiberty]

Grolo my man! Moving up in the KZbin world! Awesome!

[@sanctred]

I wonder if Grolo has looked into the container system in respect to the dynamic server meshing. With Dynamic server meshing my understanding is that even a single ship (container) can be its own server. An area on a planet can be multiple servers. They demo'd this in 2023 when they showed off Server Meshing for the first time. With Dynamic Server Meshing we are suppose to get to a point where they can allocate servers every container in the game. What Grolo is calling layers IS containers in Star Citizen.

[@grolo-af]

Hey good question! Yes I've looked into it and am aware of it. What I was trying to explain in the video, is that there's very real limits on how many entities/players can be interacting with one another at any given time. It comes all the way down to physical characteristics of the network between the servers running all this. You will eventually overload the network. Partitioning is how you typically overcome this, but when two players are directly affecting one another, partitioning does not save you. I go into this pretty far in some of my other videos. There's a lot of things CIG can do to mitigate this; you're right. For example, imagine you have 100 large ships... 100 players in each. Each one is it's own server and the contents of each are hidden from one another. It's a big 50 on 50 battle and the systems are working fine, because in reality, you have 100 entities affecting one another. Now, what if all 100 ships land on a planet, and all 100 players come pouring out of each and every ship to get into a big fist fight? Uh oh. You now have 10,000 players all in close proximity to one another. They're not in ships. There's no container for them. What dynamic meshing may try to do, is throw multiple servers at this open field and try to balance the players across a bunch of servers as best it can... and it MAY be able to. I'm not claiming it will be impossible to handle 10k players (though my intuition tells me that number would be approaching the upper bounds of what's possible with today's tech). But what I am saying, is if you keep piling players into that field, all in close proximity to one another, eventually you will cross a line and the underlying system WILL fail. Unless they layer before it breaks. If my other videos don't clear up why this is well enough, I'm happy to do another and go very deep on this exact issue. : )

[@sanctred]

@ hey grolo, thanks for responding. I see what you mean here. I did a quick search on top network throughput in AWS between instances and it sounds like if they do it right it can scale up to 25Gbps. That seems like a lot but when you’re talking millions and millions of entities that gets consumed quickly I bet. Not to mention the replication layers and iCache data layer all seem like choke points… it’ll be a fun ride to see how cig handles this and hopefully end with something truly innovative. I’ll do my part and keep buying jpg ships lol!

[@SETHthegodofchaos]

The Container system is in actuality the Zone System. And the Zone System is a spatial partitioning system. There is always onebof those at play of you split the game world. Having zones nested within each other does offer benefits but its not for free. If you have a ship simulated on one server then that server still needs state updates of all other objects around it. For collision checks which is the most expensive part of physic calculation. So if you had 100 ships each on their own server, then they still know about each other and need to collision check. The advantage is that each server only has to check their ship against the other 99. But you pay for that in networking.

[@grolo-af]

@@SETHthegodofchaos Exactly. Which is why I say... eventually something will snap : )

[@SETHthegodofchaos]

@@grolo-af yep. I still see not having to network all the players within the ships as the biggest advantage and where I see the most gains being made. But in most other scenarios I think they only can do go with instancing/layering. They themselves have stated recently a few times now. They mentioned instances for battles (and other self-contained content) at this years 2024 CitizenCon for 1.0. And in the 2021 Q&A they mentioned layering regarding overpopulated solar systems. But then again CIG likes to deploy their own terminology at times so they could mean something entirely different.

[@BobWobbles]

I wonder if we'll ever see player/org generated resource gathering missions. Imagine a situation where 2 orgs are building bases and decide to outsource some of the materials and they end up in a bidding war. As an example, say both orgs need a metric sh*t tonne of titanium or copper or whatever. Org A is willing to pay x UEC per SCU so Org B ups the ante by paying x + 1 UEC. That's the sort of dynamic player driven economy I'd like to see.

[@guyfromdubai]

I believe they have said you will be able to set up "willing to buy x at this location for y price".

[@StarshipCaptainNemo]

If you don't find server boundaries anymore, I think that means it works as intended. Great! :D

[@Immortal_Fish]

this just made me realize that ESO and WoW are 10 years apart, just like there's 10 years between ESO and now

[@raging_naclholic4636]

14:00 I think if it does have a high chance of being another 3.18, it would be entirely star citizens fault for harping that they were focusing on gameplay and experience for this update rather then just another "new engine system test" like the persistent entity update was. IT HAS to be somewhat tangibly more playable than 3.18 otherwise the backlash will be insane

[@Hemebean]

Great guest !

[@IsraelSocial]

Server meshing is impressive, i had a pvp fight yesterday, and it was very responsive, and you can't even notice that it is there. About degradation, they will resolve it eventually, the trch is impressive and I'm excited.

[@Areanyusernamesleft]

Granted, this is a roughly 9 year old video, but it's a pity neither Space Tomato or Grolo have seen the following Gamer Nexus of Chris Roberts on what was then called "zones and instances": kzbin.info/www/bejne/oGHVZ5SFlth5e7Msi=yjtL4S-30jQRQjzx It has a lot of information relative to the fundamental system architecture of what CIG wants to achive, which still should be relevant because one of the things CIG has done a decent job from the start is not making fundamental or significant changes to the game systems without telling the community about them.

[@SETHthegodofchaos]

Yeah those old nexus Interviews were ahead of the time. CR even explains OCS and SM there already even though he doesnt mention them as such.

[@MidnightWolfSDJ]

CIG really needs to focus heavily on getting trash man profession in the game as well as heavily focus on salvage gameplay and dynamic mission spawning based on junk left around the verse. To start CIG needs to make trashcans functionally delete objects. This would be step 1. They also need to add a med gown bin to all medical rooms for people to place their medical gowns for “recycling”. This would basically just be a glorified trashcan but specifically designed to receive medgowns. A small 100auec reward could be given to players who actually take the time to do that (reward scales off economy). M Step 2 would be dynamically spawning trash cleanup missions in areas that have a lot of clutter. If a server detects a lot of entities in on spot, a local mission could be posted to send someone to use the trash cans. Each trash you pick up and throw in the trash could be like 50auec plus whatever the mission complete bonus is. Say 20 cans of soda are reported by the server tech in port Tresslers cafeteria. Someone who is in port Tressler could receive a mission to go put the trash in the garbage can. Step 3 would be doing the exact same thing from step 2 but for ship salvage. If 3-4 ships are left in the same area for longer than x amount of time, a mission could spawn to send a player to that location to salvage what is there. To finish the mission you would have to clean up a minimum of 80% of the junk that is there. I think in game mechanics like this would be a realistic way to get rid of junk. And things would least likely pop out of existence.

[@Stigly]

Grolo!

[@zaeith2669]

I see a "Translate to English" button that translates "Grolo!" to "Sip it!"

[@Billy-bc8pk]

Been waiting for this one!

[@Mithrus]

this was a really great episode.

[@vicsid9606]

Great episode.

[@Vaffi]

Grolo !!! That's it. Haven't watched the episode yet, but I love this guy and wanted to leave something for the algorithm

[@WishIknewwhatthiswas]

A wild Grolo has appeared!!

[@DeviantJones]

Hopefully the shards they speak of are a temporary measure pre-1.0. Dynamic server sizes in theory should be able to scale incredibly well as long as the replication layer and entity graph can scale with it. Long term I see the biggest issue they are facing to be the global latency of the clients and replication layer; keeping the networking in check as the cluster sets grow will be tough. The DGS may actually be the least of their concerns, as if they can shrink/nest/transfer the way they seem to be able to, then they can shrink it small enough that you physically cannot fit enough players in the space to break it. They cannot however make light travel faster through fibre. We need a stretch goal of quantum entangled networking.

[@anthonyclare6750]

2nd, love the podcast

[@jamesschmames6416]

He knew about SC since kickstarter but decided to jump in 3.17? I'm wondering what he was thinking when he said, "now is the time to get in"?

[@Dragon-Believer]

Should have waited 20 more years.

[@grolo-af]

Hey there! Wasn't really thinking about it... a friend convinced me to check it out during 3.17. It had fallen way off my radar years ago. So he walked me through buying a starter pack, logging in, and doing some things. My first session was a disaster in that my friend didn't know how to play all that well either... he didn't know about quantum travel and it was my first time in the game so I knew nothing. We had spawned on Microtech and gotten in one of his ships and he wanted to show me a station... but not knowing how to quantum we just manually flew to the orbital station above New Babbage (Tressler). Using hydrogen fuel. It took like an hour. I was alt-tabbed out thinking this was the most boring game ever. Then when we finally got there, he didn't know how to hail the station for a landing pad. So we tried in vein to gain access to the station. We ended up landing his ship on the outside of the station, and disembarked for a space walk. It was kind of during that space walk that I began to understand the scale of the game and it really grabbed me... but I ended up dying of thirst and logging out. A few days later, we tried again. We just flew to a random place on Microtech to do a bunker mission I think. We landed and started looking around. Several of us were in close proximity and I accidentally dropped a grenade and killed us all. It had taken us almost 2 hours to get to that point? So I logged out again. But it didn't matter. By then I was bitten by the vision / scope / promise of what it could be. Now I'm here for the long haul. Its fortunate that he pulled me in for 3.17. In hindsight, 3.17 was one of the more stable patches in my time here. Had he pulled me in during 3.18 or 3.19, I'm not sure if I would have stayed : ) I'm glad things worked out the way they did. I've had a tremendous amount of fun in the game since. It's not all roses, of course. But like I said in this video... I just take a break when things are in fire. Really helps the sanity. Good question btw.

[@Klefboms]

I'm been living abroad for a while (less than a year i think?) and last time i played, the cap was 50. so thinking it is up to 500!! when i get back feels crazy, and its gone crazy fast

[@eisberg7681]

Nothing in Star Citizen goes crazy fast sadly

[@SupremeKOh]

Grolo the man!!

[@sanctred]

The best improvement for me so far in 4.0 is the NPC performance. FPS combat and Space Combat is a huge improvement.

[@HeimRocker]

I see Grolo - I hit like ❤

[@zeiphie]

Grolo the GOAT!!!!

[@Sylar0n]

Great podcast again! Tomato, try to grab Mad Eye Grumpy for the next podcast.

[@SteveHarvey272]

server meshing is a complete house of cards that falls apart the moment it's even remotely stressed and used as intended, as demonstrated by our PTU testing.

[@meatcaptan]

Holy crap groloooooooooooooo

[@Daggre]

I still want them to release 4.0 before they go on break, simply because we know there's a full wipe coming so there's no reason to do ANYTHING on Live right now except run the Save Stanton missions. Even if there are bugs in 4.0, we could at least start accumulating aUEC, rep, ships, etc. and I don't believe any of the issues specific to 4.0 have been around items or ship loss; at least nothing that's not also in Live now.

[@helline9]

that was cool :)

[@heavymetalsazabi]

Technology is so far advanced that the first/second gen mmos could probably be coded into Minecraft redstone and run in that game at the same framerates and resolution they played at during their launch.

[@augustwest9727]

10:00 😑 Nobody has ever gotten mad at SpaceX for moving to slowly. I guess I should also point out that to build a Rocket to go to outer space takes less then half the time to program a space game. People get angry at CIG because they are as slow as Molasses on a cold Winters Day.

[@VFW-Mayer]

The Big Problem. Everything I can do to other players, I can do between my alt-accounts. If I can loot ships, equipment, resources, or any other thing from other players. I can do the same from my own accounts. EX: I can salvage someones ship for resources. - I can salvage my own ships. If CIG takes away salvaging my own ships across my accounts, then they must take out salvaging enemy ships. EX: If I can Loot Faction Gear from others, I can sell Faction Gear in my shops to anyone else that does not have access to it. This is why Faction Gear in WOW is Account Bound so you can not sell it in your own shop or loot it from others.

[@ChristoffRevan]

You're missing the fact that claiming ships is going to take a lot, LOT longer...so sure, you can salvage the ship of your alt. Then what? You're going to be waiting a great deal of time to claim it back, it's going to be in the DAYS for larger ships. So, you're not going to be making a quick profit loop like you think you'd do. Besides, there will also be insurance fraud detection eventually, and if the AI algorithm detects that your alt is constantly losing ships over and over, it's going to increase the claim times and expense drastically...maybe even give it a crime stat. As for loot, CIG already said that pledged items (armor, guns, and other miscellaneous stuff) will work with a similar system... obviously it's not set in stone with what they'll do, but likely what will happen is that your pledged items either have "free" warranty/insurance, and you can claim them back to then receive at some receptacle. But, it'll take time to arrive, and there will most assuredly be a cool down and perhaps if the game defects that the item is near you, in one of your ships, in your hangar/hab/apartment, or inside your base or just anywhere and anything you own then you won't be able to claim it back as it knows the item isn't lost. Potentially too, the game could perhaps detect an item's ID and if you try to claim another, the previous item will disappear... rendering this fear of duping null and void. Not sure if it'll be as easy as that, but here's the thing: Any worry that us players have, I 100% guarantee that CIG has already thought about it in much more depth and likely much sooner beforehand. The thing is, we don't have the full context because we don't have the game plan in front of us, but THEY do. That's why I really dislike all the whining and complaining on Spectrum because people really seem to miss the mark on so many things due to them only seeing what's in front of them in the alpha and not further on

[@hirofortis]

LOL, Ultima Online was my first MMO. :D

[@Tainted-Soul]

I think the layers idea would work as long at they put friends in the same layer

[@MotherNature26]

43:00 yeah because we are adding in 50 bajillion features and details and feel to these projects now. Its art and tech, and efficiency is not the point. The point is to make the art. Let it take as much time as the artist wants. That's why CR had to go crowd funding route, no stomach or vision from the investor class

[@TheStumafoo]

Good maybe now they learn how to make a space game and not just ships

[@bappo2693]

You guys are still playing starcitizen?

[@Darkknight441]

More shards for pyro would help with lag I’ve experienced there. Stanton feels amazing. Stuttering at the jump gate as well as in the worm hole is annoying.

[@fffx2]

Damage control?

[@Traumglanz]

3.0 did drive me away for like half a decade :D

[@UNBE4R4BLE]

Yes, the PTU is tempting, however i'm not incentivised to do it, and tbh it's equally stimulating to kick back and experience 4.0 through YT..for now anyway

[@btoiscool]

Funny that you say working on things because I'm currently coding up a system for a service for my company lol

[@BernhardMarchhart]

Ha first

[@DennisSheltonjr]

Dual universe is a full single shard server meshed MMO space game.

[@Alundil_55]

lol

[@vi683a]

With the same complexity level of detail?

[@Alundil_55]

@@vi683a nope. And their concurrent player count on steam has never been higher than ~750. Peak player count for the past 6mo is !22!

[@DennisSheltonjr]

@@vi683a no the level of complexity in gameplay is not the same, it's a voxel-based kind of like a multiplayer Minecraft/StarMade/avorian / empyron. Everybody runs around and custom built starships with internal systems all made out of voxels. Plus bases and all that other silliness and some kind of scripting and programming language

[@SETHthegodofchaos]

​@@vi683ahow does that change anything? This is the most common argument I hear but I dont see how that is even relevant. More complexity/fidelity would just mean more data. In RAM and more data having to be computed. Which in turn would mean you need better hardware as well as more servers. More objects in the game means more processing needs to be put ob more CPUs/servers. So if a less conplex game and Star Citizen ran on the same Server Meshing tech, CIG would just have to throw more money at AWS to support the same amount of players? Unless CIG somehow are the only ones that figured out a way to 10x improve how CPUs compute game physics there cant be any difference to what others are doing.

[@augustwest9727]

So before server meshing they had 100 players per server ... Now they have 4 servers with a cap of 500... so 125 per server... I expected it would be a much higher count.

[@SETHthegodofchaos]

Dont they announce 10 for 500 latest?

[@UNBE4R4BLE]

The emergent gameplay = The Churn

[@theDarkside798]

Just remember back to 3.18.. I hope they don’t rush too much.

[@jeremyhenry3126]

I'm not worried about Star Citizen's development time. People love to bitch about everything. They will always bitch and complain. People aren't reasonable. They want everything when they want it. And hate that it takes longer than their unreasonable expectations.

[@Bubblegum_Ronin]

This.

[@behelith]

There isn't even a single player game with the scope and fidelity as star Citizen

[@Bourbon_Gaming]

You just didn't hear about games in development until they were about to go open beta or the game went gold. where as now games go early access almost as soon as they have something that doesn't crash for 10 minutes in a row.

[@Desenrad]

SM simply revealed new tech blockers that will push the release further and further. Can’t forget that’s exactly what Chris wants. Longer dev time so he can keep paying him self lots of money.

[@ChristoffRevan]

Nope... But...man, I almost feel sorry for you, living a life being that miserable. Chris is independently wealthy, he literally made his wealth from other games, movies, and other business ventures...if you actually looked at CIG's financials (which are PUBLIC, if you weren't so lazy you could find them) I don't even think he pays himself anything at all, and even if he does it's not anything more than a normal manager.

[@mercenum5186]

Can you believe Roberts has been collecting an undisclosed salary from SC all this time? Backers can't even know how much hes paying himself lol

[@ChristoffRevan]

@@mercenum5186 False, because CIG's financials are literally public. ANYONE can view them, and by law you can't have things hidden from it.

[@Thor_Asgard_]

If thats called working, CIG is going to be bancrupt before anything gets done. 12 Server for 500 People 😅 and Server performance is basically dead after 2 days again.

[@Qwarzz]

They'll definitely need need to add some carbage collection. There's no need to leave a coffee cup in the forest and come back to it a week later. 12 servers just mean instances in AWS, I wonder what sort of specs they have.

[@IsraelSocial]

For me, i don't really get why they dont have their own servers? They have to have a reason because of the point of money, is a lot of money to paid for that service.

[@Qwarzz]

@@IsraelSocial Plenty of advantages to cloud based servers. It would cost a lot to buy space for servers (And the servers themselves) in multiple continents as well as maintain those facilities. They also don't quite know how much performance they'll need in the end. With cloud they can scale the servers up and down as needed very fast. Having your own server infrastructure is a lot of hassle.

[@ChristoffRevan]

​@@IsraelSocialBecause the whole point of dynamic server meshing is to be able to turn on servers to handle specific areas of a shard, and then be able to turn them off when they're not needed (so, imagine a bunch of players goes to Area 18 and the server there is stressing out, another server spools up to split the city to make it more manageable; once the bulk of those players leave, the other server isn't needed anymore and it can be turned off). This spooling and unspooling basically means that you may end up having servers not used but still costing electricity (which is significant for servers when you combine cooling costs) and maintenance, plus during times of high stress they may still not have enough! With Amazon, they can just have a contract where they only pay them for the servers they're actively using, so when servers are spooled down then they aren't paying for them anymore. When servers need to spool up, they essentially have unlimited servers to do so. That being said, I do think once we have 1.0 they will at least handle 1/4 to maybe 1/2 of the potential servers they need themselves, but it's the alpha...the money is not worth being spent on that right now

[@SETHthegodofchaos]

​@@Qwarzzgarbage collector for C++ when? Should have implemented their game in Java that already comes with a garbage collector! /s

[@teslacuil1437]

your ship stopping because it thinks its out of gas is some loony tunes logic

[@IsraelSocial]

Epic is building something like this on unreal 5, and they want 8k players peer shard. Of course, there is going to be a lot of competition around and is good for the game industry

[@ChristoffRevan]

Entirely different thing, nothing even closely related to what CIG is doing lol

[@SETHthegodofchaos]

​@@ChristoffRevanhow does it differ? Genuine question.

[@jamesschmames6416]

The issue is not "Open Development". Its the length of time in development. I've been a gamer for 30 years. No other game has languished in developer hell this long and not been cancelled.

[@harabecswurveweathers]

They literally mention a game that's been on development longer, Throne & Liberty, and that wasn't cancelled.

[@ChristoffRevan]

Dude, that's literally the thing though...open development means that CIG has to waste time making things polished enough so that it's somewhat playable for actual players to have their hands on. This takes away a lot of time and effort that could be better used to develop the game. Instead of fixing bugs and trying to create some form of stability for patch to patch, they could just continue forwards like any normal game studio without worrying about that until the beta stage of the game. So no, open development IS a "problem" but...that's how the game was funded, to let players get their hands on things as they come out as a "thank you" for our pledges. Also, like they've said...even though whataboutism isn't necessarily the strongest excuse, there's still plenty of games out there that have taken longer. Plus, CIG is making two games at once, and 2017 they basically had to start over with the increased scope when they decided to put in physicalized planetary bodies for 3.0. EVERYTHING changed with that, and everything had to go back to the drawing board.

[@Tsudico]

Duke Nukem Forever....14 years, multiple engine changes (completely different codebase unlike CIG moving from Crytek to Lumberyard), and eventually a different developer took over to actually release. They also weren't doing a live service environment at the same time which limits how you can make changes to code because you can't just break the game for long periods of time to replace key elements.

[@wowyarkan]

you mean they used the original source from UE5.5 meshing and adapted it to their engine but UE can handle 10k players, CiG are a joke

[@SETHthegodofchaos]

Damn how does that work? I want to integrate UE code into my engine too.

[@wraith511802003]

Static meshing is working well, an the normies allowed in are still just as moronic as ever. We need them though cause we need numbers to flesh out issues.

[@UnderdogSMO]

the "seemlessness" regarding other people of SC is not nearly as big a draw for most people as Tomato thinks it is

[@ChristoffRevan]

Huh? You can't even spell it right lmao

[@SETHthegodofchaos]

Minecraft is seamless and its the most sold game. Just kidding. But its true.

[@Qwarzz]

Comparing time with GTA 6 is a bit unfair as well as that is a game by established developer. I'd argue the scope is quite a bit smaller as well.

[@SpaceTomatoToo]

True, the definitely not apples to apples comparison. There are a lot of differences.

[@Qwarzz]

@@SpaceTomatoToo And of course we rarely get to follow the whole development of the game like this. I guess one example could be Dwarf Fortress that I've played for about 16 years but it got released in Steam (and is still in early access) two years ago. That is a game by two guys however...

[@sustainpunch]

My mans still uploading on turkish time lol

[@TheJack4000111]

Tomato said it right. Its Christmas, not Holidays. If people take offence to the term Christmas then the door is open and the people that take offence to Christmas can leave Christian Western Countries. And Good Riddance.

[@rixxy9204]

Bad news - it's still not working.

[@SnowTerebi]

Tomato actually played PSO2 before?!

[@joeynuggetz]

Traditional games use "loading screens' to hide the transition from one server to another. With SC, standing in a hangar shooting someone from inside the space station means that your bullets need to be tracked from the server that you're on to the server that your bullets are landing in. This is why starfield is a crap game with a loading screen between everything.

[@SETHthegodofchaos]

No. Plenty of games already let you move between servers without a loading screen. Most MMOs had this. WoW even. You just dont have interactions across servers. Server Meshing isnt about loading screens. Its about those cross server interactions being possible. And in turn also seamlessly moving from one server into another. There are also plenty of games with loading screens that arent bad...

[@philgriffin8687]

What next? How about developing a game after 13 years instead of focusing purely on ship sales.

[@redsamurai6586]

I thought that when server meshing came in the Desync would stop…for the most part….just yesterday couldn’t even shoot one NPC because they were glitching all over the place 😂

[@IsraelSocial]

That's a bug, you can read it on knowing issues! 🙄

[@Maddock_]

It was never going to stop with static meshing. The fanatics are delusional and love misleading others into thinking "this is the one for real this time"

[@Balilarreds]

Nobody mentioning Mortal Online 2 and server meshing?

[@drunit8122]

2nd

[@daveivory1278]

ESO intial release date April 4 2014 (wiki)

[@Alias_Reign]

I don’t think he realises what server meshing is, in the current patch cycle and 4.0 the desync is directly a cause of server meshing, almost a built in feature of it. It’s why repeating proxies have only historically been used with read only implementation. It will be interesting to see if CIG can cure these “desync” issues or if they will remain with us. This is the part of server meshing (Proxy server architecture) that has never been solved (which CIG told us it solved, although that remains to be seen) and why the industry abandoned this tech in the 90s. I’d argue if it had been solved, we wouldn’t be seeing the levels of desync we are seeing. Although I’ll give them a few updates to prove their case. It’s why if they truly do solve it, it really will be a big deal. With what I’ve seen so far, I remain optimistically sceptical. But to claim that server meshing is currently working is false, and I’d point directly to the desync issues to prove my claim. Server meshing working doesn’t mean the game feels better, it means the inherent problems that make it a viable option to package into a released product that requires two way communication are reduced to the point of them being unnoticeable to the player. That hasn’t happened yet, not even remotely. The problem has never been that Proxy server architecture “server meshing” won’t work at all, no one who matters thinks that. The problem is the inherent limitations and challenges in achieving seamless, two-way communication with acceptable latency and consistency. Proxy server architecture, particularly in a dynamic environment like Star Citizen’s, struggles with maintaining state synchronisation across multiple servers and clients without introducing noticeable lag or desync. As in-game actions and changes increase, the complexity of maintaining accuracy grows exponentially, making it difficult to ensure a smooth and consistent experience (desync). Put simply, so far we are seeing exactly what we would expect to see if server meshing wasn’t working as advertised. Now, I’m not hopeless here. They might know how to solve it but are being blocked by something else. That’s possible at least.

[@CitizenScott]

Dunning-Kruger much?

[@Alias_Reign]

@@CitizenScottexcellent, why don’t you tell me where I’m wrong. Now here you could argue that CIG’s approach is unique and their implementation of a replication layer relieves the server load and solves conflicts and an outcome of this is reduced desync and CIG refining the replication layer could lead to better outcomes. We could then go into more fundamental limitations of the system but you don’t know enough to even have an opinion. You’re just a smartass.

[@SETHthegodofchaos]

I mean we had desync for like 10 years now without SM. How do you evaluate that the current desync is caused by the new SM code and not the existing systems (or overloaded servers).

[@SETHthegodofchaos]

What do you understand under proxy architecture? As far as I understand, they essentially are using a message/event bus with load balancing. Maybe some aspects are stateful, like routing client actions to the same server unless the client is moving into another. Does this not scale well? I assume they do need players to spread out far enough and having too many players in the same area will cause issues. On the clients too. And of course server networking costs will skyrocket.

[@Alias_Reign]

@ With the implementation of server meshing, the desync has increasingly got worse, but server meshing isn’t the sole cause of desync. It does make it worse, though, by introducing additional layers of communication between servers. As regions or instances of the game are split across multiple servers, the handoff process and synchronisation between these servers create added latency and complexity, amplifying the desync issues already present. Which results in what we see in the current patch and 4.0 this is evidenced by the increased desync as more proxy server are introduced. Each new proxy server adds another layer of communication and data handoff between the server and the clients. To be clear I’m not saying server meshing hasn’t had positive effects, I’m only highlighting the negatives because they aren’t really spoke about and they will remain a sticking point until a solution is properly implemented.

[@LucidStrike]

I don't think a smooth 4.0 Live release was ever gonna be possible. Whether next week or next year, it will be rough. But releasing to Live sooner means resilving the problems of the Live release sooner. It was always gonna be up to 4.0.1 Live to save the day. I'm not even bothering with 4.0 Live. I'm going right into 4.0.1 PTU.

[@DeepTitanic]

Planetside manged 1000 player battles

[@DerDrecksack87]

Server meshing is more for stability at this point than anything else, remember when CIG said all players will be on one server in the same instance, across 100 planets and everything without loading screens? Now THAT would have been something nobody does in the business. But after canning those core components, we basically remain with huge maps divided by loading screens(jumpholes), other star systems are on different servers and you get only a fraction of the players per local instance/shard without any direct physical interaction with others, so its a normal multiplayer on a bigger scale, wich aint witchcraft in 2024, wich aint bad but it sure aint ground breaking or "never done before".

[@Recreatorify]

When did they say everyone would be on a single server with no load screens? I came in a little before they started talking about procedural planets, which I was pretty sure was the thing they were talking about with no loading screens (prior to that they were talking about you getting a landing cutscene for planet landing zones which would have been a hidden loading screen), and by the time I came in they already splitting people into different servers based not just on player count but also on region and I don't recall them saying anything about not splitting players by region at least. Not saying you're wrong, necessarily, I just don't remember those two things being a thing at the same time. I can easily be wrong, and I would be interested to know that I am. However, on the end of your comment, where you say it ain't "never been done before" - the way they're doing server meshing, especially the future plans for it, have legitimately never been done before. I have plenty of complaints about what they do, and I'll happily rail against them while laughing, but the server meshing as they've planned is genuinely beyond what any other developers have been wiling to attempt. Whether it's good or not, or will succeed in the ways they want, or if it lives up to players' expectations, I have no idea. It's still an absolutely monumental, first ever task - at least, in the way they're doing it.

[@DerDrecksack87]

@Recreatorify it was early on, around or before 3.0. Back then they took jabs at NMS for having procedual generation, because they were saying until like 2016 that they will hand craft every planet and wont use procedual generation...wich they silently walked back on pretty fast after 2016 or 2018 i cant remember. In the same time frame they also said they can bridge the latency and have everybody internationally on the same server, but of course it didnt work and they walked back on that too, i mean EvE managed to do it by a built in latency that evens out player latency, but they did not want to or could not do that. The landing screen was very early on after they hit the kickstarter goal but they went for what we have now pretty early on right after the first stretch goal, the other two points i listed were during & after that, not that it makes much of a difference because right now we still only have the equivalent of a larger station on a planet we land on, the major cities are still just backdrops because you cant walk, land or do anything with them, but i never really cared about strolling trough Area 18 anyways and i doubt it will ever be "just like Cyberpunk" as some copium addicts fantasised during the years. What i ment by the "never been done before" rhetoric being untrue (the favorite slogan on spectrum besides "its an alpha") is that everything they did or will do has already been done by another game, the 3 pillars they binned (no loading screens, everybody on a same server, 100 hand crafted explorable planets) would have been the "never done before" features no other game has managed to do at the same time. With EvE you got the server meshing, gigantic player numbers on the singular server but no fps interaction or planets to walk on, with NMS and Elite you got planets, fps and multiplayer but they are procedual and not really an MMO, with X4 you got the fps and a lot of much more evolved mechanics but its a singleplayer, so you see that the separate features have already been done (better) in other games, if SC could have delivered what they said(and sold) for a long time, it would have been the game of the century, but how it is and how it is planned to be in the future, its "just" a really big online game as it still lacks the player numbers per server to be an MMO in my opinion(nevermind the loops & mechanics). I dont say its bad, after 10 years i am ok with a modern Freelancer, but its not as big as they like to tell and since many players are on the older side they only played 80/90s games or some old MMOs like WoW and now they play SC, so they dont have a cross reference and just gulp down all the copium. Its a cool game i come back to now & then, but for not developing their own engine and for what we got now, 10+ years is just a harsh time frame to be an exlusive SC player. At least the community is opening up, most of "the old guard" protecting every questionable decision or insta downvoting valid criticisms is mostly gone, but just a few years ago i would have been relentlessly harassed & downvoted on every forum for saying "hey dudes, i dont think you can bridge that global latency in a game like this", so there is hope, new blood might bring new enthusiasm & funding.

[@sanctred]

Jump holes are not loading screens lol…. 😂

[@DerDrecksack87]

@@sanctred oh dude.... Wtf do you think they are? CIG cant get two star systems on the same server, so Stanton & Pyro are on different instances, the wormhole is just an interactive loading screen, even CIG said it.

[@sanctred]

@@DerDrecksack87 Perhaps your definition of a "loading screen" is different than what a loading screen actually is. The only loading screen in Star Citizen is the screen you have when entering the PU from menu. Everything else is streamed in, including pyro.

[@MrSmith123123]

Check Planetside 2 out Grolo

[@lukasambrosy8086]

eyyyy, third

[@zacharydenham3135]

I dont want that many people on a shard/server... its bad enough trying to visit places or run cargo etc wo being attacked as of now... but 1000+ or the whole of NA? Um no thanks...

[@ParaquatSC]

well, the play area of the game just tripled...

[@sanctred]

Griefing is a growing issue in 4.0. I think we’ll be seeing this topic a lot for the next few months.

[@ChristoffRevan]

Ugh, I'm tired of these comments ...you think that certain things improve but not others? CIG already knows this, that's why they're working on better AI tech to make NPCs more capable, and more servers means NPCs can be more numerous. That plus the enhanced reputation and law system means that NPCs will not only be more competent, but exist in such great numbers that violent players will be absolutely destroyed if they try to do things in the wrong places.

[@PaulKlein-m1f]

Yeah, this is a huge CLICK BAIT title. Server meshing works but is extremely unstable and not really trustworthy yet. To say it's figured out is a lie.

[@tenryuta]

and yet servers are barely playable right now, cities and hangars dont have roombas to clean all the possessed trash, ai possessed and bouncing around like paraplegics so bounties take even longer, cargo and loot also bounces around, npcs stand and float in the most questionable places and tpose like crazy, utility tools bug the hell out, some helmets cant be alt h'd for that good cruz quicky. and no ai salvage ships cleaning up after us, you cant sell salvage cans that are full for anything, and the salvage gun goes nuts if you do the unprogrammed thing:attempt to empty or reload(had to toss the whole gun cuz it locked into my hands), ghost roids will destroy your ship willy nilly, and minable roids spawn completely different and late to regular 'set piece' roids... yeah, games doing great with the expansion to 100p servers

[@esmith7194]

Let's stop padding CIG on the back four failure. None of us want the game to fail. But in hindsight everyone has done what they're doing before them. The only problem is they did it at a lower scale and they pick tools that help them along the way. Let's not forget Chris biggest failure is he won't use anything outside of what the team creates. This is always been the core of the problem of the game. Sorry but true is true.

[@PatrickF.Fitzsimmons]

700 million and still in alpha... 🤣🤣🤣🤣🤣🤣🤣🤣

[@sivispacemparabellum5160]

"early" alpha.

[@pallmsunday6667]

Both games remember 🤣🤣

[@karakiri283]

Production of star citizen started in 2011 (pre prod 2010). So no it's not a normal dev time even for a project this size, it's still too much.

[@WishIknewwhatthiswas]

The switch to fully landable planets happened 2016/17. Hard to count anything before then as even relevant.

[@RobotSpartan]

@@WishIknewwhatthiswas Whilst its fair (and accurate) to say that development effectively restarted in 2016, its also unjust to ignore the 5 years that preceded planet tech. perhaps a better, and fairer, way to consider it is that CIG started making a game in 2011, and after 5 years scrapped a large proportion of their work to start a new project with a much grander scale. Or, to use numbers, 8 years of development, with a year or so of work carried over from a scrapped 5 years of prior work.

[@guyfromdubai]

Simultaneous development of 2 AAA levelgames over 12 years sounds fair to me

[@itz_pepe]

@@guyfromdubai I agree, also they're doing Research and Development ... building their own engine + tools which I think results in a lot of backtracking in certain areas over this period. So while the development time is quite a lot, its mostly because they bit a lot more than they can chew ...

[@guyfromdubai]

@@itz_pepe Its kinda crazy the amount of hateposting I see around this game, yet its always from people that spent thousands on it.

[@TheFrostseeker]

server meshing is NOT working. The current EPTU build has extremely bad degradation on pyro and if you attempt to throw or drop an item, that item will just stand still, floating where your character is. throwing grenades is a gamble because of this now

[@LucidStrike]

The fact you used grenades as a benchmark undermines the value of your review: Grenades in SC have been risky almost the entire time, nothing to do with SM. Lol.

[@TheFrostseeker]

@@LucidStrike i did not. I tried to drop a data pack, luminalia gift, weapons, water, you name it. all of them just stood in the air

[@UpDwnChrmStrnge]

That’s not necessarily a server meshing issue. Could also be related to changes they made to the inventory. It happened right from the start and is consistent.

[@icemanA84]

Don’t just read the video title then comment…..actually listen to what has been discussed, everyone clearly knows it needs alot of work. The end goal is dynamic server meshing and they are making positive steps in right direction. It’s still early phase for the technology but it IS working just not to the level we are expecting but the work hasn’t stopped, thankfully patience is still a thing but I fear with time more people will start to suffer from degradation of patience and attention span

[@PsyEm-p1j]

@@icemanA84well said bro 👊. This guy above just venting over one experience and trashing it based on it not being perfect. Of course it’s not perfect but look at the improvements…and watch the whole video before commenting is solid advice 💪

[@thesheep3077]

In 2022 when CIG realized how m,uch money they could make per year, there would have been conversations behind closed doors about if they could extend the dev time by a few more years, then they could make an extra 100m per year. So as a business, they went ahead and stretched it out longer, for more money. It's basic business 101. Unfortunately, the backers all suffer so CIG can get more money. That's why they call him Chris Robber(ts)

[@egth1300]

they make more money if it Releases

[@alexander.thomas.wang.thomsen]

@@egth1300 indeed, people’s insistence that they extend development time to make money that way are missing the forest for the trees.

[@grygaming5519]

Tell me how many games you panned for coming out in a state of broken or half baked. A delayed game will be good. A rushed game will forever be bad.

[@phxbadash]

Of all the moronic takes you idiots keep repeating this is by far the worst. The amount they've taken in throughout all of development is a drop in the bucket compared to what a released game would bring in.

[@guyfromdubai]

Answer this simple question: if their game was 1.0, theyve solved all the problems, theres thousands of players in the same general server thanks to meshing, 5 star systems, all gameplay is in, people are having fun. You think theyd make less money with this released product when they introduce a new gold standard ship for sale?