Armory: armory.thespacecoder.space/ Discord: discord.com/invite/thnaENwaje Armory Patreon: www.patreon.com/scarmory Some merch: www.redbubble.com/people/TheS... My other platforms: www.thespacecoder.space/
Check out our partners at VeeroTech for amazing hosting solutions that power the Armory! affiliates.veerotech.net/idev...
For more of my pictures and many others check out these guys: Atlas Defense Industries: atlasdefenseindustries.com/jo... twitter.com/AtlasDefenseOrg
Star Citizen is currently in Alpha stage of development. This means not all features are impletented yet and there are a few bugs. However this does not make the game any less awesome. Find out more at robertsspaceindustries.com/ My referral code: STAR-LC9S-MFBH
I'll be hiring high lvl NPC's most of the time and let friends take those slots when we're playing together. Unless a real good reputation system comes into play , I think most players think that way. NPC's will get better once servers are more stable. According to CIG NPC's will be very good at what they do if they get xp enough ... and more expensive ofc. I think that s a good thing. (edit: typo)
I like your idea of each method working on a different type of ore, and I think that is how a ship refinery should work in essence. However, to bring in engineering gameplay I would add in the ability to "Tune" your refinery for each ore. I would make it so that ore has more than one element in it so that you have to adjust method for the part you want, as well as the temperature, the pressure, and or the reagent volume. So lets say that you have ore that consists of iron and copper, you can choose to either use the smelting process or electrolysis method. Smelting will give you either Iron or Copper, but only one, and if you choose Copper it will be low quality and lower volume. Electrolysis will give you high purity copper in a larger quantity, but can't be used with iron. The other options would let you dial in on yield and/or purity depending on method.
I'm super late to the party but I think you missed the actual most important issue with the crew problem. OPPORTUNITY COST. realistically it will be nearly impossible to balance AI blades and NPC crew. IF they exist, players will never crew. if they don't get implemented then player crew can be a thing. The problem is human nature. People will always gravitate to the most efficient method available. Metas will always develop. and unfortunately, the truth is that no captain will ever hire a player as a crew member when they can instead hire that player to fly their own ship also fully crewed with ai blades or npc crews. the opportunity cost is too high. by having that player be in a single turret spot. you are giving up the potential to have that player sit in another captain's chair of another ship with all of its turrets also functioning without help. it's a no brainer choice. why have one turret efficiently manned when you can have an additional 4-6 turrets less efficiently manned. im not going to ever hire 2 people to run the two turrets on my connie. when i could instead hire them to bring TWO WHOLE MORE CONNIES.
i for one would like a refining minigame where i can directly impact a refinery's workload and quality. but it would have to be all carrot and no stick so a player's involvement can only ever IMPROVE the outcome to varying degrees so as not to incentivize griefers who would deliberately perform poorly to sabotage the outcome for as many players as possible. instead, if a player gets REALLY GOOD at this skill based refining minigame, their reputation with the refinery goes up and they can take on progressively more difficult refining tasks. if they botch jobs though, it harms their reputation and they are trusted with progressively less until they're basically just playing with a skill assessment simulation level of task that has no actual logistical impact on the refinery output.
Hi, we have "refining version 1.0". I could imagine, they will fine tune this in the future, like use "method x" for the cheap resources and refining method y for the more expensive ones. Also, when we look at the terminal, most times we see they are 1000% over their limit and yes, if you still want to refine your stuff there, it is already more expensive now. So when we have more star systems, maybe they will charge you way higher prices the more they will get to the 100% limit or maybe they even refuse to take any new orders because they storage is full. Then you have to a) wait, b) fly to another refinery/system or c) fly to player driven refineries ( i can imagine we will get something like this ). So long story short, I am pretty sure we will see some major changes there.
That refinery ui is so much better than most of star citizens other ui. The ui/ux team at cig seems to have too many isolated pockets and I hope after the new mfd we see some major ui improvements
One new feature that I would like to see at station refineries is the ability to convert saddlebags into SCU containers of unprocessed materials. Which could then be transported to a different refinery, bringing cargo gameplay into the loop.
Difficult... How do you get a your 0.2 scu of quartz into a 1 SCU box. Unless you have it an inventory item unto itself, like hand mined gems and you are able to put it into any size box. Maybe it's not difficult after all
Our algorithm, who art in the cloud, liked and shared be thy name; thy content come; thy recommendations be done; on my screen as it is in the cloud. Give us this day our daily content, and forgive us for not liking, as we forgive those who comment against us. And lead us not into buying more ships; but deliver us from Chris Roberts.. Amen. You tube algorithm, for the love of everything digitally holy, please give this content creator the subscribers he deserves!
Off topic and on topic question in regards to mining/refining/selling. What is a fair split in profits? Say I have a full crew mole most of the time I just do an equal share of all profits made but what is an actual equal share when the owner with a kitted mole has spent a couple mil outfitting said mole?
CiG have really grown themselves into a corner, haven't they?from raising the funds to make a game, into an organisation that demands endless funding, just to exist.. and we still don't have an actual product, just lots,and lots, and lots of pretty digital assets, within a build of part baked, constantly changing, systems, that lack proper cohesion. years of CR dreaming away resources, at the expense of backers, and a development that still hasn't dialed into a solid, actionable plan. ..and now they are sending emails to 'select' backers, offering 'limited stock ships' from.previous low volume offerings. this highly targeted 'buy now because you missed out' borders on desperation, and is very alarming.
I really want to love this game, solo playing is amazing but the bugs are frustrating. I want to play to relax not to quite rage, I have enough stress at work, then this shit…
I feel like this video adds so little value and only regurgitates misinformation / speculations. Consider its alpha-lite, that until recently most resources have been spent on sq42, a ton of feature and tech branches are not merged in to the PTU branch etc. SC runs as you would expect from a modular and split codebase where a lot of more polished features are not yet merged in. Some bugs are likely fixed in the next feature iteration branch / SQ42 branch etc. Your video does not in seem to consider these factors at all and seems to continue to misinform. Theres ton of valid criticism for sure, scope creep etc. But this video misses the mark. You do not spend time polishing a MVP feature, especially when you know other teams are touching systems you will adapt for in the future.
All your points amount to "tech debt", they can't pay. EVERY game loop and mechanic has been discussed "too death" for the past 10 years. No joke. And CIG will literally present shit like "we just started thinking about salvage & repair" .... As if they "havent" thought about it at all for the past 10 years... Then release v1 and prove they hadn't given it a lot of thought over the past decade....
@@lordfraybin Technical debt or technical legacy from an ever growing scope and feature creep is indeed a valid criticism of CIG, there is indeed the aspect of CIG having to regularly re-evaluate a gameplay loop because they have new features / tooling to consider to make the experience consistent. E.G ships requiring more passes and polish due to engineering and other complexities being more thought-out and detailed than originally expected (scope creep). - again very valid criticism. This video however (again) does not go into those specifics, instead it talks of development cycles and polish and not really factoring in things that easy account for what CIG is doing and how that results in incomplete / inconsistent MVP / Initial features. To further summarise, if this video was about SQ42 after it's release and if that release is is a buggy cyberpunk-type release... then this video could land itself really well in discussing workflows and development cycles that could've given SQ42 better results. But right now, it doesn't add value and missing the mark (at least to me).
CIG - perpetual Alpha and bugs are ignored for 5+ years. Never in history has there been a company that had a something used as Alpha AND pulled in over $700M without releasing a final game/program. Thats enough money they should have hundreds of programmers that could release the 1.0 game within a year, not in alpha beyond 12 years.
If "history" is an indicator? 60% of the employees work directly with/for - Marketing. I doubt that number has changed, but someone could prove me wrong.
The gameplay is gonna be there the fedelity and quality is what sells it. hey you know what would be EPIC cool An admin office in your hanger with a BUISNESS LINE Kiosk for personal adverts/recrutement and all (not specrum) intergrated so player A can contact player B discuss a contract for a job and draw it up both parties get a fax copy and can record it the gameplay in an admin office is "arange jobs with other players, deligate jobs to other org memebrs online to keep space truckers moving, this gameplay loop is managerial for yourself as a solo player or a party/org" the simple thing make a kiosk that creates Work contracts (when a contract is forfilled you shred it or file it this is done by photocopy and then placing the document (physicalized) in to a File cabnet or a shredder. you can recover the record in your kiosk perminnant ledger or file it in the Supply files ) I'll explain what kind of gameplay you can get out of "physical filed items" imagine a mission for a bounty you found evedence on site (serial number for a ammo clip, made by a person" you find out who created who sold it (emergent bounty mission) when you find it you can request via a call for the player to look up the RECIPTS and fax them over. so they can COPY/SEND a document to you if he document is required for edence. now obviously people can refuse in which case you physically can go to the location and just a hack to enter a VERSION of a hanger with the admin office in, you break and steal the data and take it back to investigate and find out who is behind XYZ crimes. Private investigation and bounty hunting alone would bennfit by proxy of logisics records etc.
Hey SpaceCoder, In my opinion, you are confusing this public Alpha with a finished life service game. CIG is interested in tester feedback for this upcoming game in open development. It isn't intended to be lookt at as a product, fit for delivery. You are asking them to polish systems they possibly intending to replace in the future. A lot is still in flux. Why assign work to something that might get replaced in the future. A Dev once told me once: If they were a scam, they indeed would take the current version, and keep polishing and optimizing to attract as many as possible, before pulling the rug. Instead they are constantly adding new underlying tech, expanding the game and creating new incompatibilities that have to be found and fixed again. That's why they invite people. To help find problems. Keep up the good work!
Your solution is similar to mine. When you finalize the loadout the character steps into an attached changing booth that equips everything then steps out. Similar to Deatg Stranding or Grand Theft Auto. They can put these on ships as well by replacing the lockers.
Some Proect Management Someone who says "No Chris, that's a silly idea" and can make it stick Collective viewing of Jurassic Park, which can be trimmed down to "spent so much time on what they could do they never stopped to think if they should do it" Pre release regression testing with no live till what worked pre patch still works. Ship design around function and use, not what looked cool to the graphic designer Don't release ships without game loops Tattooed on the [select body part] of decision makers "this is a game, not a job". You have a rich Lore, use it for content
I knew a contractor that employed young unskilled builders to do construction work, they used to wing it, terrible workmanship and short cuts, shoddy work was the result. That's probably what goes on over at CIG.
Eve online NMS and Elite clearly demonstrates that CIG now has amateur developers. We were automating complex QA Server load and regression testing over 10 years ago but it requires discipline and professionalism which are both lacking at CIG. I think they knew the code is FUBAR and will never work after the 3.18 disaster and they have decided to milk the final years of SC and get as much money out of the player base until it's dead.
It shows that the number of developers doesn't matter if behind the scenes there's a high turnover or most of them don't have a big experience in making video games. We could notice it when CIG mentioned that most of the designers for the Merchantman all left the company (6 persons apparently), hence stopping the development for the ship. And when we compare this with what No Man's Sky is able to add with each update while they only had 20 employees (40 now but they're working on 2 games!), we can see a huge discrepancy between the 2. I also preferred the approach of Elite Dangerous that was released in the estimated time after the Kickstarter Campaign, and then Frontier added more content through extensions like in the old days. Frontier was also able to release various games since then and they now have 800 employees split on multiple games.
@@space_is_real the past 7 years we've been largely paying for research and development. Which is STILL in R&D... 😶.. And the "original" masterminds behind it, are indeed, gone. They are on their 3rd database engine... and I can tell you, from experience... it will NEVER work the way they need it to work. There just isn't a big enough pipe to push ALL that data, replicate it, save it, update it, sync, rinse and repeat, OVER the internet- to create a GOOD pvp experience. Never going to happen. DBs just can't grow that large, and be easily managed. When ANYTHING goes wrong, it takes TIME, lots of real... heart crunching time, even with the fastest of RAID arrays...to copy that data. More time than fluid, properly functioning AI...etc.. have. The whole notion of persistence/ physicalized "everything" is never going to work well.
I think there may be an issue here with point of view, and goals. The point of view that CG is the client (the one who has the needs being met) is correct... but viewing each stage as an upgrade or extension of a live service is the problem. The client only cares about bugs and features in a standard gaming methodology: get to feature complete first, with bugs dealt with at a minimal level. Then polish that, for stability. Then work real content, story etc, with another polish pass. And finally another bug hunt, before release. Since they have yet to achieve feature complete, all else is secondary. This is their goal. QOL is irrelevant, until it damages their money-flow.... which it hasn't. No matter how annoyed ple backers are, they keep coming back. keep investing. And new people join all the time. even in the *very* alpha state it is in atm, as a sandbox, it is fairly impressive. Especially when their released competitors are often so much trash. *coughStarfieldcough* The methodology you describe is exactly correct, for someone with a product, being sold as a service to a client. That is not the case here. Their pace is predicated by anting to actually *GET* to release. The pace on that is always borderline crunch. The alternative is polishing a lesser product, one that they are not yet happy with, and know major designs need to change. They are *finally* actually looking at inventory. *For the first time.* Don't kid yourself. Inventory was, previously, a placeholder to enable testing of other systems. And yes, the process of design is to think things up... mess some of it up, and adapt to make new changes. The new ui was linked to how it brought items to the player. Called on functions that necessitated the equivalent of a drawer. The "personal sized freight elevator". that was a bad call. So yes, the placeholder is back until it is redesigned. And this is normal as well. This is not the equivalent of a live service game getting an expansion; this is a feature being added to the foundation of a piece of code. Not a branch higher up. Foundation.
I make use of a simple definition of development as a puzzle. The picture on the box is what your goal is. The pieces must be found and placed correctly to finish it. CIG has a serious issue. It has based its revenue on ship sales. The ships, even the small ones, are complex and take time to develop. when a new ship feature comes out all ships then need to be brought up to the new gold standard, which takes time away from other feature development. For example, the resource network needs to be added to all ships. Even the ships that are very behind in updates like the gladiator, caterpillar and so on. And while all this goes on, they will add more ships. In my opinion CIG needs to do two things. First, they need to get to feature complete and start on beta. Meaning no new features are to be added and work starts on adding content and polish. Secondly, they need to train up a large staff to focus on the ship back log and bring all ships to gold standard.
IF we got a game we could play without mad shit we would never got to play the fkn thing LOL does it suck sometimes they are oblivous Yes ..... are they humans? Yap. shit happens , people like to bitch heh click click
About the drawler in the item bank, I'm pretty sure that they thought more about "optimization" than actual "gameplay". Why? The answer is simple: they tried to switch to a bulk transactional model. In the current implementation every withdraw/deposit operation is atomic. New call to BE for every item movement. With the drawler we were generating a list of items for the operation, and it was processed in a single call. So from the technical point of view - they did it good. The problem is that from UX's point of view it was a disaster. Which leads me to one of two reasons: 1. There are no PO (product oner) on a feature level for some of the teams (that can explain why the passionate team are delivering good things, and why some others doing just tech bare minimum) 2. There is no policy that forces dev team to play their own game. (which multiplies the problem from p1.) And. yes... That's typical hell swamp for the product development, where PM/PO missfunction can create so many issues
@@izraphailzero5610 the worst thing is that if they'd just developed a proper character equipment UI they could've gotten away with doing it that way and just had us get equipped all at once after committing, but instead they tried to lazily double dip with the freight elevator UI and ask it to do things it just isn't designed for, which necessitated an extra step that falls back on the old inventory UI they want to replace. What's worse is it's the second time this year they've tried to replace the inventory panes with something that wasn't designed to do all the same tasks; the loot UI has the same issue and similarly falls back on the old.
Chris is no longer running CIG. That job belongs to the Marketing team It shows in the lack of concern for backers. Only the harvesting of new money matters lately. Older ships make no new money. (unless you make a $50 MKII.)
@@nosir5596 No.. Chris is a slightly imaginative programmer, who borrowed most of his ideas from Star Wars... and became famous for Wing Commander. Sandy - his wife, is the Marketing Major, and SHE runs the Marketing Dept at CIG, which was 65% + of the employees, at one point. Watch Chris and Sandy in an interview together. He does exactly what she asks/tells him to do. Sandy wears the pants at CIG.
@@lordfraybin Chris is an ideas guy. He's the dream guy. He's always been marketing, even if it's not his title, he's the guy selling the dream, he's the guy selling the product everyone else is having to figure out if they can even make.
Very well compiled. I have a thought. I think the reason why CIG relies on feature heavy releases is because of the marketing hype that they can generate around them. I don't think they would be able to market as strongly as they would if smaller features were released incrementally. A big part of how they do development is most likely predicated by the marketing division.
@@Kakeyoro it's only recently-- and intermittently-- that they've even really done feature heavy releases. Basically just 3.18 and 3.23; 3.19-22 really would not pass as "major feature patches" in most live service games. 20 was basically an AC patch, 21 only really brought tractor beams which are still pretty broken and undercooked and 22 brought a shallow, slapped together salvage addition that broke the game economy lol
@nosir5596 They've never released ANYTHING in a finished state. Nothing is finished yet. Literally nothing... Not 1 single game feature, mechanic, ship, loop... It boggles my mind. I only follow because I HOPE that some day.... it will be playable, and I can get my $$ worth.
Nicely done @TheSpaceCoder ! I love how you touch upon the game building experience and make it easily accessible for all. Definitely want to see more episodes about various aspects of how coding and game engineering relates to SC. If you had the time and inclination, it'd be really interesting to see the results of your coding for any part of a SC system/feature that peaks your interest. Of course it would be extremely simplistic and only serve as a model but might be interesting to see how you'd tackle certain parts of SC, especially the broken parts 😅. Of course it's probably all too much but thought I'd throw that out there! Keep up these great videos! I like your critical thought while keeping it real.
I suspect that the huge number of kiosks are just there to generate data on the best location. Most likely habs, hangars and near by shops. But, we could be surprised. Also, isn't it better to start with more and progressively remove them later? I do suspect too that is was to train players in a more soft way than removing directly the local inventory for just a few kiosks.
The refinery example highlights how frustrating issues that boil down to individual values in data files are. Why is the reputation earned from a shared mission set to a less than 1.0 modifier? Why are they punishing group play in a game marketed on multicrew? Why do ships that have mass statistics several times higher than other ships have lower hull hp values? Etc. I really do think a team of just 3-4 people who are assigned “fix and polish values and minor bugs for really anything that you can” would do wonders for the game. Doubly so if one member of that team spends a portion of their time documenting it. Imagine the good will and positive sentiment that would generate. Post a video every Friday “this week we fixed the Argo RAFT’s hp, it has Reinforced in the name but less hull hp than some single seat fighters. Bug fixed: “Reinforced” ship was not reinforced. Corrected.” Or “we noticed players never share missions because of reduced reputation and credit earnings. Change: Reputation reward no longer reduces on mission share. Team suggestion: Requested mission team consider having sharing a mission trigger additional enemy spawns / higher cargo value rather than reduced pay out to facilitate group play.” Or “The Heartseeker has hearts again.”
I really like your idea, but I think it's difficult to implement. It is "inefficient" to have Team B review the coding done by Team A, since Team A is more likely to know where the problem lies and how to fix them without breaking other things. The problem is that Team A is no longer there or was already signed in another internal goal. They are constantly launching "tier zero" and moving on. But I'm 100% with you that this should be something they needed to focus on, so they can REALLY show their tech demo working as it should, instead of this hot mess where you clip through the floor and colide with invisible asteroids.
@@matuto1986 Team A was supposed to document and annotate the code as they went. No project is ever supposed to depend on people not leaving the company. Also, without *someone* assigned to going back and checking on bugs, what’s even the point of issue reporting? Besides, I’m pretty confident the marketing boom alone could cover the cost of the 3 employees, even if they do have to chase people down to ask questions sometimes.
Great points here. The only thing I had a disagreement on was something you clearly stated later in the video. Being: before CIG even considers a location/npc clothing etc, they must think about HOW it impacts a a player and the overall game. The lack of forethought in this regard plagues CIG and some of the obvious flops make them look … disjointed
NPCs at outposts with no breathable atmo, but no without space suits on... If they're gonna go systemic with this much stuff they need to plan ahead better. As they turn more of these systems on I think they're gonna be overwhelmed by the amount of unintended consequences they'll have to deal with.
Quality may matter, but not to CIG. Honestly seen asset flip games that care more about the quality of what goes live than CIG seem to. At every stage from evo to live the standard is just too low to be acceptable, and it's becoming increasingly obvious with a number of features that have made it to player hands that there is just no effective internal playtesting or QA. Builds go out to PTU with trivially obvious and immediate problems, features make it to PTU and even to live with glaring "no one ever laid eyes or hands on this as a player" functionality issues (like most of the tractor beam placements on ships, f'ex). Idk what's going on internally but there is absolutely no way to look at what they're doing and conclude that quality of the product is a priority.
"Honestly seen asset flip games that care more about the quality of what goes live than CIG seem to." I don't know what definition of "quality" would make that true... Not to mention, Star Citizen's "Live" branch is not a release branch. Say it with me now... _it's alpha!_ No seriously, I wish I didn't have to mention it, but when you're talking about the quality of what "goes live" in such an ambiguous way, I kinda have to. At most, you could say quality isn't a priority _right now..._ but with the definition of "quality" we seem to be using, it's really not supposed to be a priority right now. The thing very few people seem to understand about SC in particular is that NO amount of internal testing would suffice. Everything has to be tested at full scale in an online environment, more than any other game. Because of what it's trying to do, and the ways it's trying to do it. That's why we're even able to play it right now, they're not just giving us "early access." Especially for the really big things like PES, RL, server meshing, you need everybody. Not evocati, not PTU... you need as many people as possible in the shit. We are the bug testers, and to some degree we all know this... but everyone has a different idea of how many issues are acceptable at each stage. Yeah, lots of dumb things make it into the Live build, and sometimes they even stay there for a long time. Sometimes there is a totally understandable reason for it, sometimes less so. But ALWAYS, 100% of the time, the answer is more complicated than armchair devs in the comments imagine. Most people have no real concept of why SC is considered to be so "ambitious," and why it's so difficult to make, but everyone thinks they know better. When I look at SC, I don't see a product that places a low priority on quality. I see an insanely ambitious product (with a nearly impossible scope) that is still WIP. In the early/just crossing into mid stage of development. If quality wasn't a priority, they would have shipped it years ago. But they don't just want to release something functional, they want to deliver the highest "quality" we've ever seen. You don't put that much focus on "immersion and fidelity" unless you intend to give it the polish those things require. There is absolutely no way to look at what they're doing and conclude anything else.
Not defending the constant broken state of SC. But as a QA professional in the same field. The amount of issues we raise and are subsequently ignored by the production team for people to then turn around and say “this was never even tested.” Is staggeringly high. There are bugs we miss, but not to the degree people think. QA doesn’t budget the project. They don’t put the time costs for fixing bugs to developing features. It’s rare we even get a say in what should cancel or delay a release. And in star citizens case I doubt that is any different. Stop blaming the little guy when it comes to big problems. It doesn’t fix the issue.
@@nosir5596 ..what? You literally said: "features that have made it to player hands that there is just no effective internal playtesting or QA" In what way is saying QA is ineffective, not blaming them?
They can say pre-alpha all they like, but they run “playable now” and “live now” advertisements. They’re using our backer money to run those adds to attract new players. New players whom they immediately offer a monthly sub for otherwise gated gear, cosmetics, and hanger items. So, it’s a live service. A live service I dearly love, but they have chosen to make it a live service, so the “pre-alpha” excuse is a bit thin.
Server meshing is the new Jesus tech buzz word. When that fails and we realize good old proper Netcoding is what is required and CIG are minimizing server infrastructure costs then we'll wait for the next Icache/ssm//replication layer Jesus tech is going to improve the gameplay experience soon TM. Coming to you soon from Cotizencon
I disagree. Alpha is exactly what this is. Things overall work about 75% of the time, and they are in a constant adjust and release model…very typical /public/ alpha. The real issue is the length of time known bugs are allowed to persist in the code. Their ‘get 10 votes’ approach to verify a bug before they work on it is understandable to a point. They can’t respond to every bug, but a lower threshold would help players feel like their input matters…provided they actually work on and publicize the player-reported-bug-fixes better. I don’t know their company internals, but I have doubts they have any dedicated bug hunters. If there really are hundreds of devs out of the 1,000+ employees I keep hearing they have, it shouldn’t be a problem to create and run such a bug hunting team to work those lingering issues.
@@rupert456 Addressing this because these all of these misconceptions are very common. (TL;DR warning: Brandolini's Law) "Jesus tech" is the only buzz word I see here... Server meshing is an actual thing that exists. Good old proper netcode and beefy server infrastructure will only take you so far. And it will not be enough to have thousands of players or dozens of star systems. For that, we will need all of them. Good netcode, good server infrastructure, and static/dynamic server meshing. If dynamic never works, we can do something with static, but it would be far from ideal. With no server meshing at all, it won't even get past 100 players. SC doesn't perform badly because it has "bad netcode" or "cheap servers," but those are the first things everyone points to when they have no clue. And that's fair, it's a natural assumption to make..... (when you don't have a clue.) Want to know why SC's performance is rough? Compare it to something like Hell Let Loose. Similar player count (100 players), and a similarly high level of graphics/fidelity, but SC's "map" is... just a little bit bigger? And there are just a few more vehicles with just a bit more detail, POI's, NPC's, and you can see all of it without ever seeing a loading screen... Seems obvious when you look at it that way. It's not the infrastructure, it's the load its carrying. Actually the infrastructure must be pretty decent to even hold up at all. None of the "buzz words" you listed (icache/PES, static server meshing, replication layer) was supposed to be a "Jesus tech" silver bullet to "improve the gameplay experience." They're all just parts of the machine. They're not even "buzz words" either, they're just... the names of those parts. If you don't care how it all works that's cool, but that doesn't mean they're just buzz words with no meaning. Not to mention, there actually have been some pretty major improvements since RL was implemented. 30k's are literally a thing of the past, CTD's are extremely rare. Server errors were common for a bit, but pretty rare now (more common in PTU of course). It's becoming more and more normal to have long, uninterrupted play sessions where the biggest problem you run into is lag/desync. And the lag _is_ pretty bad, can cause a hangar door to eat your ship once in awhile, but it's better than getting a 30k or CTD every 30-90 minutes. Tbh, it seems like it's about ready for some hot, meshy, server on server action. But yes, TL;DR, it's possible that dynamic server meshing never works, and the game never gets where it's trying to go. And if that happens, I hope you say "I told you so" and it feels really cool 👍
@@rupert456 You forgot to mention the future jesus tech announcement: "AI meshing tech". This is to give us "life like" NPC's in the verse, to interact with. These will require their own dedicated servers, mixed into our glorious meshing soup.
You really relaly really need to be regular with Comm Relay. Maybe start doing it on twitch as well? And audio levels really need to be adjusted. Your mic volume is definitely low as to what it should be on normal levels. You can observe yourself by comparing your Mic volume for this stream with your normal YT video upload. I think its almost half of what it should be.
As a sidenote for feedback: the discrepancy between mic volumes makes it really difficult to follow the conversation - your voice is barely heard if one keeps volume at a level where your guest is not too loud. It would be great if future talks can have normalised volume levels.
You're right, I noticed this too.. I kept adjusting the sound. As suggested, it might be good to have a third party sound check before the feed starts to normalize sound levels. ☺️
Hey @TheSpaceCoder I appreciate your discussion here and your guests. Thank you! Not sure how you feel about this, but here's a request, I'd be very interested, and I suspect others would be too, in hearing a conversation between yourself and another coder on the topic of server meshing. I did some very basic math based on a rubics cube model of meshing, combining current server/net code models (of roughly around 100 players per server), focusing on the CONNECTIONS to the center cube/server out of the 27 cubes/servers and found that each server might only be able to support a max of 3 to 4 players (connections) per server when all 27 cubes/servers are "loaded". Keep in mind this is with a total of 100 connections to any one cube/server which is why I chose the center. This seems like a lot servers dynamically spinning up to support 100 players in a possibly small space.. ? I did a bit of coding in the past but I don't consider myself a coder anymore. I'm hoping I'm wrong with these numbers/thoughts. I'd love to hear your thoughts one day on server meshing if you decided to share them!
Hi! Glad you liked the show! I have a couple videos on server meshing on the channel but definitely would love to get a few devs on the show to share their thoughts as well. And your math is actually very interesting, you assume that servers are meshed based on physical space/volume in the Verse. But fortunately that's not the case. Verse is organized into object containers, these can have pretty much any size and can be nested. This creates a graph system. Not all nodes of the graph are streamed in at all times and you can split the graph between whatever number of servers you choose. The problem then becomes moving player and entities between the different servers but on the same graph. IIRC one of the CitizenCon panels from last year had this visualized
Thanks for the reply! It was pretty basic math 😅, but I did take into account that the servers would spin up dynamically and split player load up until it reached 27 connected spaces and continue to add server/spaces as required, even if the same exact world space was being sub divided - I'm mainly focused on the connections between so as to only transmit the data required and further broken down by the virtual space between the players. At any rate, I know very little of it all so I'll stop there, and would definitely love to hear an episode on some thoughts from game coders. I think it would be very interesting to hear experts postulating on the tech and perhaps even looking at other games that are implementing server meshing. Thanks again, keep up the great work!
