Star Citizen: Can 3.23 actually happen?
Вставка
- Опубліковано 23 сер 2024
- A short overview of the mechanics allowing CIG a chance to actually get 3.23 out successfully. NOT a prediction of said successful deployment!
Referenced Video: • Star Citizen: Around t... - Ігри
If CIG has been doing things in the background like you believe, this would put them head and shoulders above the rest of the games industry. This is Amazon and Google levels of technical skill.
Yep, I'm trying to reserve judgement here....but this is a big test for CIG, and I think they know it. You don't "Call your shot" unless you are either sure you can hit it, or need the boost getting it right gives you. Either way, the next 6 months is going to be make or break time I fear.
why? You have insight into what other people are doing in the industry? Do any of you heard about SpatialOS? They literally had this meshing solution in 2015.
Your independent commentary on the state of the game is a much needed and much appreciated breath of fresh air that helps re-energize my enthusiasm for the project.
Looking forward to your future response videos as we draw nearer to patch 3.23’s release.
Catch’ya in the Black…
Awesome commentary on 3.23 i hope we can see you on more SC podcasts and shows, networking infrastructure is something the community has been lacking in understanding so its nice to see that role filled
Thank you for taking the time to analyse and write your understanding of the game development process for 3.23
You're so nice to listen in background.
Also, I think we should expect something like early 3.18 experience, to be completely honest.
I expect some issues, but mostly where a feature has had to widen its scope for multiplayer reasons.
Thanks! I just read a post where someone called me "Sensationalist"....so I'm getting all sorts of odd input on this video! Something tells me a drinking game is in order :)
I'm trying to reserve my excitement but hopeful they can pull it off. Keep up the vids though man, you give great insight!
Thanks for the calm, rational, informed (or at least thought out) and easy delivery of your information.
With the addition of over 200 Dev's back on the PU from SQ42 will make leaps & bounds this year for SC. I've been here since 2012 & this is probably the most excited I've been since OCS was first dropped on us. More eyes on that prize will push this baby through. Let's go!!!!! o7
I have to admit, I'm actually optimistic, but a lot of that is dependent on them having called this Babe Ruth home run back when they did the pivot to SQ42 and them having executed the plan......if this was a bolt-on later, then we are all in big trouble here!
You are a refreshing voice in a gaming community that is either full of naive hype or indiscriminate hate. Thank you for this type of video. Subbed.
I appreciate the clarity you're bringing to technical speculation about this project. It's time I subscribed.
Thanks for calling out the meaningless expression "netcode" that is often thrown around in ignorant gamer discussions. It's a useful expression only insofar as anyone using it is raising a flag to say "I have no idea what I'm talking about" :)
Hey look I just shared your video a few times on a couple of my different pages, I wish you the best of luck with this channel. I also hope that a lot of people are paying attention to this channel especially people that play the game, because it helps to understand the magnitude for Why itz taking so long for it to get completed... Sorry for typoz im at work most of this was Voice to Text.
Man I love your insight on these things. It's quite interesting to get more behind the scenes processes with the paint peeled back
Your videos are breath of fresh air. Love the content keep it flowing my man.
I was blown away by them calling that big of a shot. I don’t expect all of that to make it in (I would be **shocked** if master modes actually release in a good state), but it wouldn’t take the whole list to be something approaching SC’s biggest patch in forever. Looking forward to seeing more from them this month for sure.
Yeah it has been said by CIG many times before that Squadron 42 runs as a client and server based model. It's just that the server potion of the game is running localy on your device.
So I would imagin both the Game client, infrastructure and server codebases are shared between the games and its just the networking side of it and variance bettween the S42 and SC feature set that need to be addressed when they are porting code between the two projects.
I'd also think that the client version is stripped down a bit. No need for as much code to work with sync issues, or other such problems. But I get your point...I'm pondering a "Deeper Dive" for the uber nerds or coders who are fans. I'm struggling with generalizing enough for the general population to understand a concept without sounding like I'm a fool to people who actually know the topic. One of the burdens of a communicator ;). I mean, even I'm not happy with how I explained compile time options...as you said "BUILD_LOCAL_INSTALLER" vs "BUILD_MMO_CLIENT" With the proper flags in the code to include the needed stuff for networking and exclude client code not needed for online...etc.
Perhaps I should do a main episode, then a "Deeper Dive" when needed to get into the more nitty gritty for those interested...or just split the video into two sections....hurm, worth a thought!
Thanks for the video and for joining Space Tomato's podcast. Exciting year ahead.
Huge love, keep talking you are providing some excellent perspective and insight
2024 is the biggest year for SC, people don't quite realize this but the *Alpha phase is coming to an end!* We'll be in Soft-Beta sooner than y'all think.
Ya know it won't stay an Alpha forever right? They want this game Gold more than us
This is the most honest title I've seen about 3.23 pre-patch so far. Kudos. CIG should have to actually hit their patch projections for a year or two before we start touting it as, "BIGGEST PATCH YET!"
Cheers, homie.
@tybio thank you for using your super power to help explain to the community what is going on. I feel your insights have been invaluable.
Thanks! Keep up the good work!
Thanks, I very much appreciate this!
Thanks Tybio for your feedback. Use to work in a server IT lab with Unix servers, curious what Amazon AWS environment like.
Love it, Tybio! Keep it up!
Even if they only get half of what they're shooting for in it's gonna be a great patch.
this is exactly what i feel too
hoping for the best :)
Love your videos, pro tip for viewers: set the playback speed to 1.25 :)
😉
What they have presented yesterday, is in the line of thinking in my predictions (3.22 will be a long lasting patch / 4 months), albeit I upped the patch one number higher as I expect this import of squadron gui tested features in Star Citizen will likely take a longer time. 3.22 is bound to stay a longer time with us, replication layer testing stays for the forseable time in this release branch, and instead of 3.24, 3.23 will receive the SQ42 ui updates. But I think it will take longer than 4 months to get this released.
Great video. Although I am not sure I fully understood it.
3:24 hmm, how did you conclude that? 🤔 What do you mean by separating and infrastructure? The networking/replication code?
I understood it as them abstracting networking code away under the Seralized Variables code API feature? I mean, before the game and networking code was already in separate functions already. The thing they changed seems to have been mainly the ease of how networking (not the network itself) could be setup in the code as well as optimizing bandwidth. This functionality might now be redone with gRPC instead.
3:30 I can understand why you wouldnt want to use "netcode", but doesnt networking or state replication make more sense here, rather than infrastructure? When I think of infrastructure I think more in the realms of Kubernetes and AWS. And less so about my software.
3:35 That might be true (I am not as deep into networks as you are) but netcode is (also) an established and widely used term for multiplayer videogames. I personally like to define it as everything that comes after the basic networking and state replication systems. Mainly the code that is responsible for latency compensation. Mainly by reducing the effects of perceived latency on the player clients. Keywords here being client side prediction with server reconciliation/rollback and server lag compensation (the server predicting what the clients saw). Together, it makes the game more smooth and reduces the impact of latency. There are entire YT channels that compared different videogame netcode implementations and their pros and cons (e.g. still dying while already behind cover). Netcode could even be the bandwidth optimizations from Seralized Variables and OCS.
Therefore, I dont agree with your definition and refusal to use the term. I think it has to be more differentiated given the context of videogames. Especially because using terms like infrastructure makes it more confusing for gamers again.
4:35 Shouldnt Game Code be Engine Code and Infrastructure be Game Code/Game Logic? And Network would be Networking/Replication/Netcode. And underneath that the network itself? Maybe I misunderstood but this seems incorrect to me.
5:45 But the thing is that even the singleplayer game uses Seralized Variables and OCS for persisting game state. So that we can have savefiles to losd from (although its more like an entire DB). So even parts of the bottom layer (which are networking code) is being used locally for the singleplayer games too.
I know this may sound like "you dont know game development" and it is kind of a meme in the critics camp. But while I do think you are highly competent in your domain, but since the field of computer science is just insanely large, I recommend having another look at networking and netcode for videogames. In just fear info in this video is going to confuse people more than it is actually helping. Which would be a shame as I dont think it is its primary goal.
Thanks for taking the time to post such a thoughtful question! I'm assembling all of these from the comments and going to ponder how best to address them. I think I may record a "Deeper Dive" tomorrow that gets much further into how I believe they have this setup with a framework (Rather than Infrastructure, I agree with your point there) that allows them to build stand-alone or MMO client from the same general code base...so they just have to do the work to tie in these systems to the functions exposed by the network team. But I have to build more detailed flow charts before I have any chance of making a logical statement here :)
Pro tip, set playback speed to 1.25x and he talks normal speed
It does have a chance if the networking team has a DevOps representative on each team. They went to DevOps many years ago, with more developers on the PU. I know the teams can make at least 70% output if the teams are tight. The other 30% are the most complex systems.
I wonder how “containers” and next gen nvidia data center type tech will improve the PU
But... I think in this last SCL, they didn't talk about server meshing, nor the replication layer, nor pyro and jump points or even Vulkan. That is critical, key and a turning point. We are facing a new and WE HOPEFULLY definitive 3.18 that lays the most basic foundations in the quality of life of the players.
We are assuming that meshing, replication layer are going to be included and I don't believe it. Rather a general facelift carried over from the work already done in SQ42.
They talk about having adapted all the SQ42 functions ported to the PU with all the work that this entails.
They will carry substantial modifications of elements that for the most part are already found in the current legacy and original code of the PU.
The list of features they have announced for 3.23 would be this, I think I left something out, but basically:
-Improved EVA function (even without the push and pull system to move using nearby surfaces)
-Distribution centers (concept similar to a city or station but on the surface of the planets, all services concentrated)
-MasterModes for each ship. (I suppose in summary they will be simple speed modifiers in certain situations such as combat, atmospheric or space flight)
-New Quantum Booster to jump between points in the same solar system. (It replaces the current original system that only allows you to jump from point to point in a straight line like a train on rails)
-Replacement of the entire legacy user interface with the new and redesigned one in building blocks (Mobiglas/Hud/Starmap/Radar/action sphere/etc) They have commented that the radar will eventually reach all the areas of the mapping as they add them all radar areas. The replacement by building blocks has already been carried out with all the ASOP screens and from different stores/services should not be a problem) (Hopefully the mfd of each ship will begin to arrive progressively as well)
-Compass and other player status (finally, they will implement the player's vital statistics and orientation)
-Character interaction system with the interactable environment, replaces the current and legacy interaction system including ragdolls that accompany certain interactions (in part we have already seen including certain changes in the gladius ship, where the character interacts with the buttons on the cabin)
-New dynamic sights and crosshair system (The sights are implemented as a shader technique and the new crosshair simply enables third-person aiming)
-Manipulation of magazine bullets (the Tarkov fashion)
-Replacement of the character designer (They have already done this before and now they are going to leave it as seen in other MMOs)
-Personal hangars (Controlling the loading of objects in the ships, surely instantiated spaces for each player or group of players in the party)
-New xeno threat... (Generation of dynamic Raid-type missions like the current assault on Orison)
-Modification and readjustment of the prices of the entire current market (Has Tony ZzzZ been fired? Will there no longer be a quanta system managing the trade?)
And they have warned that some of them will arrive after 3.23. 3.23.x 3.24?? or they will arrive early.
They also haven't talked about pyro and jump points. I think that both server meshing and pyro will arrive in the long-awaited version 4.0 or close to it like PTU or Preview Channel. This summer? Late summer Q3?
They talk about quite a few weeks, but I think they do not exceed Q1-Q2 for the topics covered in the SCL.
I don't see it as titanic that they deliver these things in 3-4 months, taking into account all the time they have worked on it and so vehemently.
The worrying part is the division of spatial zones into servers and the communication between them (Server meshing), that is still behind all this facelift that is coming in the short-medium term.
Until then, until meshing, rep layer and vulkan (local performance) are implemented and debugged, servers and services will continue to behave as before. But everything is more beautiful and comfortable (which is already a lot). (And with less 30k, since they are fixing the faults that cause it PLUS the new "hot recovery system" (Rep Layer)
It is a progressive change that will culminate in version 4.0, 2024? together with SQ42 or near it?? maybe...
I don't think any such thing, I believe that 3.23 is going to move a ton of low hanging fruit over to the PU, then the rest of the year will be about the Replication Layer split, meshing etc...etc.
Good video again, my only concern is that they may be trying to bit off too much in one patch, while also seeing the possibility that for this to be done right a lot of it needs to go in at the same time.
I do expect 3.23 to initially not be great as far as stability goes, I think it will be breaking a lot of things which are not expected and we'll see quite a few fixes before it gets better and that is fine as long as CIG can go in and fix these issues in quick succession. I do agree that if CIG pull off what they say the intent to in the end with 3.23, they will make a massive, massive step forward and we will be so much closer to the next big step, which will follow much quicker this time ..
Nah it's not "too much", all of these features were already ready for the single player...it was just a matter of tweaking them for the multiplayer (yes I know it's not just a simple click of a button, but is HAS to be easier and quicker then making completely new features).
As long as they don't pull of a 3.18, we're good. They crammed in as many features as they possibly could, all the while bringing probably the biggest overhaul to their underlying network infrastructure at that time in the same patch, which completely broke the game for months. They had to do multiple hotfix patches just to get the game in a somewhat playable state. Even post 3.19, it was bad.
That being said, I just hope they do intense playtests of Replication Layer in the 3.23 Tech Preview & then push it to live
@@trinityx3o522 They said months before 3.18 that PES was likely going to be a dumpster fire...it was such a massive change to the backend, anyone that knows anything about the development knew it was going to take a lot of tweaking once it was in. So, they didn't "pull anything", they warned people beforehand, but like people prove time and time again...gamers don't read, and thus they (you) were shocked that things were unstable. What else did you expect for such a major piece of tech?
@@ChristoffRevan In my comment, I haven't mentioned once of being "shocked" by the outcome of PES' implementation. I'm just re-iterating on what happened & what CIG said. That's why they created the Tech Preview channel in the first place. I'm just stating my opinion, with a bit of speculation that if CIG were to implement the RL, it would most likely have to be in a 3.23.x patch rather than the main patch itself as this patch is significantly bigger just in terms of features. Who knows how many things will break if the Replication Layer was also to be crammed in the same build? But considering their approach to this, I'm pretty sure they won't take that risk.
There's no need to be skeptical, the work has been held back for a long time now, stuff has been done for many months already, it's all about that NetCode and adapting stuff to multiplayer. I have a strong inkling that CIG wanted to release all of this in one huge big 4.0 patch with both Pyro system and all these new features, but that clearly didn't work so now they'll release everything before Pyro instead.
PS: No glasses? :O
Funny, I just posted a reply above saying this exact thing! I see 3.23 as our reward for the wait, clearing the deck for the rest of the year to be mostly focused on replication layer split and meshing....you know, get some good will before things may get a little bumpy :)
Yea, I actually put in my contacts because I was sick of fighting with the lights so they wouldn't cause reflections. I mean, it clearly wasn't because I Wanted to look sexy on camera...for rather brutally obvious reasons! ;)
Amazing video and cool seeing you on the tomato podcast.
I’m less mad about Dev time being diverted to Squadron… however, I still won’t forgive CIG for depriving the PU for so long.
Keep the videos coming!
They didn't, though. Not really. Like, think about it. There's almost no part of S42's development that won't benefit the PU in some way. Ships, locations, sounds, gameplay, music, the only thing we don't get is the frigging cutscenes.
@@jasonmack760 fair. I’m going to complain until we get the features in the PU.
I think they had to do this. I know people want to have input, but speaking as someone who has to translate customer desires to our developers....you can't create something new when you are constantly having to go back and tweak the work you just finished based on feedback...and deal with the community having random meltdowns about stuff that is a place holder :). I was angry about it because I'm inherently selfish, but I don't /blame/ them for it.....I'd have done the same thing, then released a set of features at a time when the full plan is in place so people can give feedback that's actually valuable.
Ehh, close enough of an explanation, I suppose. A bit of a net-centric view though. Plus, it will a big mess for a little while. But not the same type of mess as 3.18. With 3.18 it was more the code dealing with the underlying structure that needed a lot of ironing out and more underlying structure to make it work better that was likely not foreseen until 3.18 was out and could not be realized 3.19 and later.
This update is more of plugging feature sets into the underlying code that is made for it. With this many feature sets coming in it will be bumpy just due to the number and there will be little issues that will need to be ironed out to make a given feature work smoothly with its underlying structure that supports it in the MMO version.
This is not the issue.. and really has not been for years. Their ideas are sound and were very likely to be made possible as technology had come along far enough to make it happen. The real question going forward is how effectively they will be able to realize all their plans and how much will they have to scale back features to make it all work on current technology.
Always hard to generalize enough for the vast majority of the community to understand without sounding like an idiot to those who know the deeper details. I didn't think getting into compile flags, or includes, or other means of structuring building a "Client" or a "Standalone" on the same codebase would make sense to most...
I agree on the effectiveness, that's why I ended with the "I'm pulling for you, but you called a /hell/ of a shot here....I hope you make it!"
The answer will be how many players they can have in a given area against the effectiveness of server meshing without having to reduce the quality of everything else too much. The good thing is that the tech now exists and it can only improve. And, since it is pretty evident that we have also been paying for CiG to create a game engine that can be licensed to other companies, that will improve many games in the future, hopefully@@MrTybio
What Citcon showed was the long ago promised payoff for delaying things to focus on the engine platform first that could service all the intended titles with shared assets and data interfaces. We have seen the new EVA many times over the years in demos or for SQ42 teasers as well as the newer UIs. Those are all Star Engine client features regardless of local or network play so they will be relatively easy to port over because It's mostly eye candy and animations. The MMO server issues will still be there until that side changes. Now that we have an engine coming together making the SQ42 sequels will be easier and not need most of the devs. Content made for the MMO or SQ can serve both titles as well as Arena Commander. They are essentially building 3 game experiences in parallel with shared client assets consuming different back-end services and game config rules. Like you said they have decoupled the service layers of the game. Modular distributed systems are easier to build and iterate on but harder to design without painting yourself into a corner sometimes.
Yep, I think of 3.23 as our pay-off for the waiting...rather than gating all of this with 4.0 and Pyro once the hard work with replication layer and meshing is done, they are taking a chance to move all of the "low hanging fruit" (Relative term here people) into the PU so they get some breathing room for the rest of the year on the server infrastructure which is likely to be much less...er, cleanly deployed.
Like, shared, and commented… I was already subscribed. Also, I agree.
Going to watch it, brb …lol
Edit: So I watched it …holy fuck man, and ya 100% achievable, I’m sure most of it is already done as well, I just started a month ago, and really think I came in on the best time. Can’t wait 🤙🏻
Greeeat Video Keep it up
Hey Tybio. Still would like to have a robust discussion with you. I approach forming my opinions about Star Citizen' 'Development Saga' from a similarly nerdy-technical viewpoint. As you did, I find myself with a strong desire to inform people to promote an accurate knowledge base in the community. In my case I hope to address certain issues such as: A) aspects of the development which are subject to massive lacks of understanding by most, B) rampant misinformation in the larger community, and C) knee-jerk responses which are understandable while also being without actual merit-in-fact.
I hope we can connect and talk soon.
There is a discord linked in the channel information, feel free to drop by. I'm not on it all the time, but I'm around at least a bit each day during the week and randomly on the weekends.
30+ updates are smacking their virtual heads at this title. 🤭
It's hard for me to get on board with this idea that 3.23 has "all this stuff" and that 3-4 months is the time frame of delivery. Since 2016 S42 has been progressively but systematically absorbing more and more resource. They have continually pushed the narrative that this is justifiable as all the work will benefit the PU.
With CIG saying they're now shifting focus to the PU, there should be a complete dump of stuff backed up and waiting to be pushed to the PU. At it's best, any perceived acceleration now is merely fulfilling the role of playing catch up.
That's not how I see it, but I totally understand where you are coming from. I've seen amazing things when a quality dev team has the time and space to work....so I think it is possible...but a big big ask. I'm not sure I'd have gone out this aggressively setting expectations, but if they can live up to them it will be a base-line change in...well, everything I suppose.
@@MrTybio given the history of the project, I do feel that the way they have laid their cards out can only be driven by a level of confidence that is quite high. There are always issues with SC and I'm sure some things will slip, but Ive a feeling they might be on top of this one. My reasoning is that a lot of stuff is probably just a matter of porting content and that they have significantly more people than usual working on this. I also feel like a lot of what they're aiming to push out has been in dev for quite some time, probably longer than people are actually thinking.
This has a vibe of things coming together about it. We know big projects can hide the progress for a long time, only for it to come together quickly "at the end" from the end user perspective. I feel like the probably have a good 2 years worth of effort that is coming together, which ties in with my original replies thinking. I'm not as concerned as you seem to be about a repeat of the 3.18 patch personally, but you never can actually tell.
Thanks!
Thank you!
here is something to ponder..
This patch is the ONLY patch we are getting this year.. ie, all the other patches we get will be based on THIS one. Yep let that sink in .. alll the other patches bah the last one are ALL going to be based on this 23 patch.
:) .. hope that helps.
little tip for people who solo cargo .. it takes over 1 hour to load a M2.. no kidding. ( I test it for ya ) .. solo ^_^
That's not how I read things at all. I think 3.23 will land before June and the rest of the year will be the server side work. But I suppose time will tell!
@@MrTybio To be fair Ty i dont see it as a bad thing, i mean the amount of stuff we are getting is amazing for me tbh.. a lot of the patch as we know is still in half mode.. missions eva map stuff all will come in other patches as updates .. and i hope just like you we get server meshing ( i think that will be a major patch on its own tbh ) and rep stuff.. / pyro end of year! ( if all goes well )
as somenone who did game dev.. im pretty sure thats how i c it being planned and why they are so bullish to give us all the mas content drop from the get go.. most of it is done .. rest will come in later patches at updated.. ( and ships ofc ) :D
keep up the great vids :)
I dont think 3.23 and SQ42 will be out before Armaggedon...just sayin
Armageddon came out in 1998...so.... ;)
3.18 felt like everything came crashing down, but i feel like it was for the better we probably wouldnt be here with the state of
the game today if 3.18 didnt happend
I am a little surprised, that so many people are hyped this hard about the patch. It seemed to me like just a bunch of new lipstick for the pig.
This was especially obvious in the economy bit, which sounded more like ship grind planning than a complex system for the players to interact with.
Maybe it is the glacial development speed so far, that make rather superficial changes look more meaningful, while the big questions behind „how much does the ship cost in game compared to the $ price in the pledge store?“ are avoided.
For example, how do they want to resolve the problem of item based (ships/components/fps equipment) progression in a full loot game with a pledge store and insurance?
The way I'm thinking is that if they get this done, then they have hundreds of people who will move on to the in-game stuff as the network team fights with replication layer split, meshing etc....then when Pyro is out and a little stable, late this year or early 25...they will have another batch to dump of progression/mission/reputation updates. But that's just me having a theory, all of it depends on 3.23 going well...so I'm trying to keep that in the back of my mind for a "What comes next" video if I ever get more confidence in things :)
Their versioning system is ridiculous, the amount of features they are going to release is at least should sound like 3.3, if not 4.0, this is definitely major release for them.
The numbers really don't mean much...people put too much stock in them
I don't think they care much. The only one that matters is Alpha 4.0 with the full game structure in place....from there I'd expect a Beta to be the next major version, not Alpha 5.0...but honestly, who the heck knows.
I think you are leaning severely optimistic for sensationalism, from past experiences any change or bug "fix" results in another disruption somewhere else and gets designated as a fix later issue. 3.23 could be a larger example of throwing shit at a wall and seeing what sticks.
Thats exactly what it is..
but in this case this shit smells reallllly guuud :P
You honestly think this was a sensationalist commentary? I mean, I'm just shocked, I've never EVER had anyone accuse me of being sensationalist....
Contemplative, boring, dull, slurring drunkenly...I've had all those...but not sensationalist, so thanks! A first for me!!!!!!
no more money till space whale! the current dev cycle is way too slow for me to remain confident in the project
Understandable.
@@MrTybio I really appreciate your takes on it, as well. Its definitely valuable.
Hope you all the best in the UA-cam journey!
love playing siege of orison and going to jail for no reason for 15 minutes, then it never working for the rest of the night because the shuttle never shows up even region hopping