SaltEMike Reacts to Chris Roberts Giving Dates for Dynamic Server Meshing

The Video - / tangiblesparklycoffeek...
Check out my main channel - / @saltemike
Check out my clips channel -
/ @saltemikeclips
Check Out My Twitch Stream to see this live: / saltemike

[@Gofr5]

Jared's face tells me he was screaming internally the entire time for Chris to stop talking about time frames because he knows he's going to have to deal with that at some point. lol

[@Vioblight]

Nah Jared loves talking time frames. Take his “all of these will release in 2024” comment from Citizen con 2023. lmao 😂 Sure, maybe he is learning from his mistake idk.

[@bluexgreen1]

CR's description of the tech issues being weird/wrong is also going to make Jared's life hard.

[@CitizenScott]

Yeah dude Jared hates having a job for sure

[@Samwiiise]

😂

[@Volf1916]

Those eyes say "Stop shredding my soul."

[@Ehawk_]

I like Jared just nodding in the background and I imagine him thinking "please dont my job harder" while Chris is talking

[@CitizenScott]

Jared's existence relies on engagement. This drives that. He's a corpo.

[@the5starx]

Jared’s face through this whole thing was hilarious 😂

[@CitizenScott]

That's just his face

[@Sentrme]

Chris mentioned "LIMITED VERSION". That is the most important take away.

[@Meister_Dwight]

Yeah probably in an experimental PTU form

[@Yorhosashi]

thank you for this detail!

[@CitizenScott]

Yeah man CR saying that server recovery is dynamic meshing is the key takeaway 100% not coping

[@SETHthegodofchaos]

So 23.12. Preview Channel for 100 people

[@Swatmat]

lets not forget that 3.23 has been the best launch for a patch for many years

[@Netist_]

Is that sarcasm? Genuine question, I don't plan on picking the game up again until we get the cargo changes so I haven't been following the 3.23 launch at all.

[@Kerman_von_Braun]

@@Netist_ Devs took a big fat dump on ship combat and it will likely take years to get it to a serviceable point if at all

[@JamestheKilljoy]

​@@Netist_ it's running surprisingly well. The real test will be when the Xenothreat event drops

[@georgek2574]

@@Netist_ its been very smooth

[@ultrastoat3298]

Watched a few hours of Summit1G's stream of 3.23, very good stuff so far.

[@bobstark4201]

This is a classic case of taking something Chris says and reading way too much into it. "A limited version of dynamic server meshing this year". That doesn't mean full-blown server meshing.

[@JoseJimenez-fc6pu]

Seeing is believing. That should ALWAYS be the mantra whenever Chris puts out anything with dates attached.

[@bluexgreen1]

The limiting factor in static meshing was that so many services were not 'server meshing aware.' Everything worked within each server, but crossing boundaries and planetary rotation (moving borders?) were when meshing had issues at any number of players. Things did get really bad at 600+ players, with significant performance decrease starting at 200 and increasing as they went up. CR here can't even describe what the issues are or mention any of the above. Since this is all pillar tech and he knows so little about it, he is obviously only manning the helm in name now.

[@thomasinlondon2849]

Lol the look on Jared’s face could also be him thinking “he’s doing it again” lol

[@CitizenScott]

Jared should make his own game I agree

[@slipycatfishdog1313]

Bwahahaha, find someone who looks at you the way Jared looks at Chris, you need that in your life… I’m fuggin dying over here!

[@FrazzleCat]

Jared's eyes. Someone get him a life time supply of wine.

[@CitizenScott]

CR already has basically by employing him

[@FrazzleCat]

@@CitizenScott heh fair point.

[@Needlesse]

im excited to see chris talking openly again.

[@RN1441]

Why? He's proven to be an excellent predictor of what WILL NOT be achieved on a given timeline.

[@Gigageorge]

why? every time that man speaks lies come out of his mouth. Him speaking openly doesn't provide any value.

[@Eagle6Airsoft]

Excited for more lies

[@tricktap8502]

Why tho? everytime he talks he just wants more money without actually producing a fun game.

[@CitizenScott]

Me too. All these haters fail to recognize that SC only exists because of CRs words.

[@ArtemisTGM]

Gerad looks so shady.. he's either in the middle of a coke deal or holding in a massive taco bell 🔔 growler 😅

[@tallperson117]

I mean, maybe? The server meshing test a few months ago was much better than I expected it would be. It would be sick if they did get it out this year tho.

[@Brenelael]

From what they've shown working so far this is believable. I never thought I'd ever say that about a Chris Roberts time projection because he's usually absolutely horrible at it but there it is. This is at the very least... Extremely Plausible.

[@HumArchInst]

Maybe they'll get an early internal version working by the end of the year. I don't think they'll be able to change the live deployment steps they told us for while now: 1. system to system meshing 2. static server meshing I think after this they have to put in some in between steps till they reach x. fully dynamic server meshing.

[@mrscsi6472]

by “at the end of this year” i assume he means in the same sort of sense in which we saw server meshing at citizen con, where it’s a bespoke implementation of the system that’s small scale and suitable for demonstrations and not actual gameplay.

[@Aver0xx]

Loved the video! Just something I noticed there is a slight buzzing sound during the video I thought it was on my end from my GOXLR but it stops when I pause the video.

[@Billy-bc8pk]

Gotta agree about the upper-end of concurrency. From the tests, it sounded like the sweet spot for optimisation will be around the 400 player mark. If they can stabilise 400 players per solar system, I think that's more than enough to make that section of the game feel dynamic and lived in, and without becoming too over-crowded, while also opening up a lot of opportunities for unique and emergent encounters and large scale fleet battles. But that's just me spitballing based on the limited tests they ran. Hopefully in the next series of tests we can get a better idea of how well the tech scales, where the stability sweet spot is, and what player traffic is like to scale within a single solar system. Because while 400 might be a sweet spot for Stanton, 800 might be preferable in a place like Pyro.

[@ICircuit64]

My understanding is that with dynamic meshing there should not be a theoretical upper limit on player number per system. When too many players gather in an area, that area will be divided into multiple servers. E.g. a solar system into planets, a planet into areas around points of interests. I may be wrong, but that is what feels logical to me. Of course, even this would break down if too many players gathered at a specific location, like jumptown.

[@dtrjones]

What Chris is saying when he says they kind of have dynamic server meshing already with the server crash recovery, he's saying that they dynamically handoff the player to a server instance that isn't a different server, but a recreated server which they do on the fly because they don't know when a crash can happen. That's a fundamental basis of dynamic server meshing. The other thing to emphasis is the use of the word limited. By implying that they might have a 'limited' dynamic server meshing, it really means that they have a limited capacity to scale the number of servers instead of having fixed server instances which can be done relatively easily with the technology to recreate server instances and handoff like they do with server crash recovery.

[@kristofgyetvai9328]

The word "Limited" is going to hold the weight of the world.....

[@RN1441]

I'll believe it when it's running on my PC and stable. At that point I'll have to ask 'is it fun', which given how many huge question marks there are about teaming up and doing things with friends and what the gameplay systems will even be is a big if.

[@turfish]

He said he "thinks" they can get a "limited version of server meshing in" possibly.

[@unslept_em]

i'd be looking forward to a dynamic test this year, but i'm not getting too excited about it until i see how it goes and to what extent they actually test it

[@vorpalrobot]

Nice thumbnail! Give the editor my compliments.

[@SPAMMAN123456789]

im going by my predictor. It got static server meshing right last time. Chris said this year, I bet next year in the summer. I find CIG tends to be a quarter or 2 late now for things. which is an improvement believe it or not.

[@KildalSC]

If we're really getting 4.0 with meshing before september then I can see dynamic meshing testing by end of year. Servers are expensive so I'm sure they want to move from static to dynamic meshing as quickly as possible.

[@TairnKA]

I'm hopping for a "Live" update, so I get my Falsom jumpsuit back. ;-) I don't know of any place to get them, other than from CIG? ;-D

[@Sheppardsg1]

jared face when u paused imao

[@123Andersonev]

Pretty much hinges on the natural progression of persistent entity streaming, everyone saw how that system works that is if you like the closest to the client in terms of seeing the netcode function, however whilst you can just spin up a fresh shard for PES, persistence hinges on the ability of the replication layer to have solid redundancy which means tracking the state in real time of all the shards (lot of data) and what happens if the replication layer falls over?, you need a backup also in real time (lots of data x lots of data) and you also need to know which server managing the replication layer is the authority at any given time.

[@phxbadash]

Hes not wrong about the server recovery being like dynamic meshing, the only thing missing is the logic that determines when to spin up a new server

[@rolinthor]

And where/how-large-an area-in-which to spin up a server.

[@makimera]

and after that we got the updated roadmap for 4.0 with server meshing v1.

[@SpaceBound-1]

Don't take this date seriously. Dynamic server meshing will be finished when it is finished.

[@Sentrme]

Best comment. Also any large complex system has to be dynamic to even function.

[@michaeltrager5992]

Very best comment!

[@SpaceBound-1]

I think Na Mike has said this in one of his videos before.

[@gkenny12]

Lol great content very entertaining

[@PatriotLaLiLuLeLo]

Tried 3.23 yesterday and was at least happy they got some actual changes out after so long , had trouble saving my character and thought I had infinite loading screen glitch till I finally got in then landed at micro and shot dog aliens that were stuck in place but i was at least happy they finally added stuff after so long it felt like it's been so long since a good update...

[@Stryker2079]

He said “Limited version” lol

[@RicoZaid_]

A LIMITED version.

[@UNiTEDDKSilent21]

It's important not to get lost in the words, he said "I think we'll have", so it's just what he thinks, and who is "we".

[@talonodyssey]

Regardless of how long it takes, if they get this right it will mark a new age of online MMO gaming. So many games current performance would benefit from having multiple servers running in the background.

[@Ogata123]

The largest takeaway is that people still do believe the timetables somehow lol

[@Skyrat12]

We're so far away I still feel. I mean, I can't even buy/sell shit at CenterMass at The Commons in NB! Inventory is totally the biggest challenge right now I feel.

[@TheWorldsprayer]

"From a certain point of view"

[@andurian01]

Dates? Come on Mike. He said I "Think" and "limited" so in his opinion they could get a very early version this year probably internally.

[@epav2396]

Ya mike was reaching on this one lol

[@RomanorumVita]

Knowing Chris, he very much means it'll be EOY in a "Pyro" sense. Somebody will probably be playing it, maybe in Evocati, by EOY.

[@criticalchai]

I get Chris' point. technically the server recovery is dynamic server meshing based. the replication layer does move everything to a new server seamlessly. of course instead of a single player and his ship over its moving the entire server, but I get his point. and so far it has kept most of the missions and other data intact. So he must be pretty confident it can be put in. But the wording of his its partially in means he could call a better recovery on server mesh servers dynamic.... i am cautiously optimistic

[@ZbigniewZiggyCzachor]

"LIMITED" version of Dynamic Server Meshing, not FULL version

[@theninedivides6851]

3.23 has been a kick ass patch been getting great performance and so far things are very responsive even the recovery from crash is not too bad

[@JB-gb7uu]

Well, static server meshing is supposed to be coming this year.

[@nickstinger4709]

I think it's great they will give us this server mesh and replication. However, it's ALSO important to reduce server crashing and improve server performance. Is there much hope to start improving server reliability or will we forever be looking for bandaid solutions?

[@angepano8591]

I think that he's just trying to blur the line with Static Server Mesh, which is coming and will be a great achievement for CIG with more credit than is needed. If he said that SSM would be fully working and STABLE by the end of this year with at least Stanton and Pyro running then MOST people would rejoice. But he just can't let it go....

[@LordAikido]

It's really just a show of how willing they are to invest in server architecture, if they up the power of the servers... this issue goes away, but they need to fork out the cash for cluster servers that can handle the concurrent player load. Or are they gonna take that money and put it towards more artists, and development budgets. (This goes back to the 'are they a live service or an alpha' debate).

[@cidhomerimperialpneulemen2107]

Do note, he said a "limited version" of dynamic server meshing.

[@saltemikereacts]

Yep, in the end…static will do the job for me

[@JaysHobbys]

He did say limited. I will say that so far this is a good patch even with the issues with crashing and loosing cargo that I have been experiencing. The patch is fairly good.

[@BoatLoadsofDope]

3.23 is damn good for a live patch so far. I even threw a bit of money at the pulse because of it.

[@RYCOPZ]

Jared's face man.

[@Accuracy158]

Jared was there as security but not to protect CR... Just to make sure he didn't get too out of hand. 😆

[@reamoinmcdonachadh9519]

They are sort of overdue another Tech Preview Test for Pyro and Server Meshing

[@atlantikjcx7449]

they will likely do another after Invictus

[@Jack-vv7zb]

I mean they have already tested server meshing with 800 ppl at the start of the year so doesn't seem impossible

[@BazyG]

Just 10 minutes ago " 3.23.1 build is now on the PTU channel for all backers!"

[@andyandreson3989]

They have to get Invictus patch tested. It was a huge patch as well. Though I haven’t jumped in yet.

[@blackbird9962]

Terada talking to Chris Roberts

[@stuartthurstan]

In an interview in 2016 Chris said "by the end of this year backers will have everything they pledged for and much more." Forgive me if I assume that anything Chris says is complete bull.

[@Volf1916]

As far as stability. No the servers are not better. Example, two restart/replication layer move in an hour and 10 min stops while if your in a location that entails any data transfer such as selling or buying cargo, blink it is gone. Landing or taking off, hope your not at the threshold because yeah doors start closed. So lots of new bugs and all the old ones.

[@RayTieRom]

"limited version"

[@VoltaikEntertainment]

He did say he "thinks", which really means he is not 100% sure.

[@NightNord]

I mean it's mostly true. They basically skipped the whole iterative development and went straight for the whole thing. Which is why they went for PES first. That said, their "Dynamic Meshing" is area-based. So its not the dynamic meshing that will support unlimited players. Also its probably the dynamic meshing as tech only - not support for gameplay. So missions, etc will probably break

[@hectordrake2882]

He needs a gag order. ;)

[@Pa1ad1no]

Does Chris Roberts knows in which year are we? Does Chris Roberts knows that we know in which year are we?

[@DaringDan]

If Chris says it add YEARS to it.

[@blazemonger1]

I have always said that I believe CIG has broken through the hurdles that held them back last year and are on track to deliver what they said/showed during Citcon last year during 2024. And so far, they have. 3.23.x will take another 2-3 months and two, maybe three patches to cook before they move to 4.0PTU early August. I can see them already having a number of mechanics for 4.0 in the existing codebase where the main pillars will be Pyro and Dynamic Server Meshing. My expectation is that Citcon 2024 will coincide with the release of 4.0, will bring a releasedate for SQ42 and we will see one or two more 4.0.x patched in 2024. 2025 will see the start of bringing it all together, the arrival of base building and one or two more systems and then.. Citcon 2025 I could see the announcement of being feature complete (or at least close to), followed by the beta announcement in first half of 2026. That beta may well start end 2026 for a possible release 2028 with 6-8 systems

[@518UN4]

Jared standing there all exhausted from keeping CR from promising way too much to random people all day long KEKW

[@D-rizzy31]

thumbnail gave me anxiety not gonna lie... will edit with update after video lol Edit: It didn't fill me with hopium. He did not sounds super convincing, but I was pleasantly surprised with hearing it "could" be this year. Like with all things SC.... time will tell.

[@Xenon676]

Frankly I would not be surprised.. If all CIG employees have another job. That is considered their main job. This game is just something they work on. When they feel like it.

[@mylesmarciniak1623]

I think keyword he said is "limited" he is not saying full server meshing.

[@THENEONGRID]

He is just beeing uncomfortable standing there and not having anything to say. If he is a bit socially akward, that is all the explanation.

[@StellaAlamarr]

I don't think the difference between static and dynamic is a huge difference. Its literally just a ai system that auto dtermines how and where to allocate resources. Pyro is coming this year with server meshing. Dynamic should be at the same time or soon after. They could already be working in it now.

[@GhostofZuuzGaming]

so he meant static server meshing this year, dynamic SM wont be this year

[@GeorgeNoX]

Its coming soon, surely Chris wouldn't lie, right? lol

[@fwg1994]

Nothing CR said sounded wrong, but I think he's severely underestimating how complicated creating a system to balance the load will be. There's a lot of questions that need answering in how it will handle the load and I think it's really easy to overlook some of those questions until it's mid development and you don't have an answer for it. Also I think it's ability dynamically migrate people depends heavily on how quickly it can migrate people. Because if it is going to take two minutes to migrate everyone over, that's going to severely limit how it can be dynamic. You'd probably need to anticipate where there's likely to be high load and spin up servers preemptively because you can't be having significant interruptions like that in the middle of combat because another 10 people jumped in and suddenly you need to split off a new server to handle capacity. It'd also be difficult to spin down servers with low activity. Unless a server is completely empty, you'll end up causing disruptions for someone. If the migration is basically instant, and it's just the start up and shut down that takes time, that's a lot more manageable.

[@WarlockSRB]

@SaltEMike Cpt_FoxyLoxy has tested Tali modularity in 3.23.1 PTU, has video on YT :)

[@thomasknight5465]

Limited version basically = pyro and 4.0

[@reamoinmcdonachadh9519]

Jared is looking kind of "yeah, right" in a disbelieving manner

[@ThomasD66]

Reminds me of Jill Biden standing near Joe lately.

[@Rikalonius]

I'd like to get excited, but the ability login to a server has gone from 1 in 3 to 0 since it went live. So, maybe when they can work out the client/server connection, I'll start believing in future tech.

[@Haegemon]

What feels is Chris wants that to be, but he isn't who is developing, so I would prefer these were the words of the chief manager, creative director, lead programmer...

[@solventob]

Anything either of them say timewise means nothing. Given the last 11 years, server meshing could as likely come in another 11 years, or be completely scrapped for the "next best thing"

[@o-super2744]

I love Terada.

[@BGIANAKy]

“Limited” is the keyword here.

[@saltemikereacts]

No the fact that he gave any date at all is

[@BGIANAKy]

@@saltemikereacts haha. May be 😂😂

[@MrSulen-bi5nu]

They will have a "limited version". That could mean anything.. it does not mean it will be in game, EPTU, PTU or PU.

[@Eagle6Airsoft]

Limited because it’s a sketch on the back of a napkin. Here’s your limited meshing. He puts two napkins together.

[@MrSulen-bi5nu]

@@Eagle6Airsoft hehe.. yeah.. yupp, could mean anything

[@ericweeks8386]

Have we ever seen Chris Roberts and Elon Musk in the same room at the same time?

[@royzevisionneur2045]

Chris 🤔Jarred 🤨 Terrada 😁

[@RosscoAW]

The fact that the game already is operating using a graph database (Amazon's Neptune), which is the fundament big change about Persistent Entity Streaming, is far, far more impressive from both a simple technical (software engineering) perspective and, more importantly, an at-scale deployment perspective. By all indications, they just need to modify existing game systems to work with the new persistent environment (making missions and trespass and other state machines persist correctly across servers), which is not a "server meshing" issue and is simply a consequence of sheer tech debt, not any kind of foundational lack of tech or deliverable bottleneck. Bluntly, all of the hard work has already been done. What we have now is already "server meshing," in every meaningful sense except for one: tech debt and polish. What has yet to be completed is not a design, production, or engineering bottleneck, it's simply a little bit more bug fixing, optimization, and NOC scaling.

[@thomasinlondon2849]

Amazon will suck most of the backer cash away, sad true, and not many other options.

[@justinnixon5823]

Chris looks really tired, poor guy that is a face of a hard worker.

[@SpaceDad42]

I don’t think he works at all. He sits back and lets the money roll in. This is why the game is going downhill and woke. No one is at the helm and the idiots are in control.

[@nerdglider]

Hasn’t Chris been making games since the 80s how does he look so young?

[@Diogenes76]

Shit static meshing is all I care about right now (ok and some cargo sprinkles!)

[@rixxy9204]

I don't believe a word of it. In fact I would expect the _opposite_ of whatever CR says these days... so it's definitely *not* going to be in this year.

[@thesheep3077]

Do a video on the community response to 3.23 on LIVE, it's a mess

[@IdleWorker]

Wait, does CR speak french?

[@FinNapalm]

It can be dynamic, but it might not be good for gaming.

[@sirbonobo3907]

SoonTM as usual

[@canadianrabbit4838]

Bad Chris robert bad, you know you cant give a proper release date.

[@brbrd102]

Yeah sure Chris. Lets get static meshing flawless first.