Soldato
The sims is polished and has tons of features
******Official Star Citizen / Squadron 42 Thread******
- Thread starter Raumarik
- Start date
It's too specialist to be off the shelf, the only highly-scalable MMO server engine I'm aware of is SpatialOS, they are pretty new though, but did get 500million big ones in investment funding recently. I would have felt a lot more confident about SC if they had announced they were sticking with CryEngine and using SpatialOS for the back-end, rather than cutting some deal with Lumberyard that doesn't appear to have given them anything useful. At least they'd have a company of top-notch network engineers behind them.
The nature of AAA development though is usually to get the performance and fidelity consumers expect you need to custom build all the performance critical components, including the back-end.
What I would really like to see from SC is a proof-of-concept milestone. "We're freezing the codebase for non-essential work and targetting a baseline of one single space station with X small ship, Y large ships running at Z client FPS on our target hardware spec, in doing this we will prove that the fundamentals of Star Citizen are achievable and we have a stable base with which to continue".
It roughly translates to "we don't have a working back-end yet". Don't start asking awkward questions about redundancy, they've got to nail the basics first.
Soldato
Em Amazon have some of the best network engineers in the world with their AWS service and they are tying that together with their Lumberyard system to make a single ecosystem? So far CIG have utilised new coding modules for animations, volumetric fog, gpu physic code and a few other smaller items which have allowed them to push what CryEngine could produce without having to write all the code themselves.
They are also able to provide servers all over the world via AWS that are linked. It means that you can have the one instance. There are not many other options out there for that. I understand that SpatialOS are partnered with Google but see below why it isn't the right choice for at least the next 5 years.
In regards to SpatialOS, it is way too early in development. It has had over 12 updates since mid 2016 and most of them have caused the need to write the majority of code because of changes their end. It isn't ready and likely wouldn't be for another 5 years and they have already been developing it for 5 years to get to this stage of them just getting to a sort of Beta of the systems. Does that not sound familiar to what you are saying. There is no way that CIG or really any other game that is wanting to be released in the next 5 years could use their system.
If it does what they claim it should then yeah in 5 years it would be awesome to see other larger scaled MMO games come out. So far only Worlds Adrift has been released and that is a much more limited concept of a game in comparison.
No and yes, this system is about creating the overarching back end that supports what the game needs to do. So it runs the service, the game servers pick up the service and the client picks up the game server data. There will be multiple copies of the system running and they have direct access to the server at CIG's end unlike the cloud servers from AWS. So redundancy is key here. However, the game should run even if say the economic server drops it just means there wont be live inflation/deflation of commodities till it is back up or for the interdiction system then the AI interdiction would just stop working but the game should still run.
A lot of software runs this way though with specific servers providing specific services. It isn't anything new in terms of utilising software to do it, more just in the game world relative to the game type. Now if the whole server system goes bang at AWS end then that is different of course but then that is just like any online game in that regards.
Not sure why mid gen has decided that utilising separate sever systems to hold specific branches of code means "no working back end" but they are going to remain separate and nothing to do with the current issues that SC has in terms of stability and FPS other than the general gameplay mechanic in itself strains the system. They will still utilise the seperate
Soldato
The latest Sims is not all that tbh
- full of pay walls/DLC content far greater than any other to the point they stripped the base game right back to basics.
- The building tools for creating a house are better, however the content to actually go into it is lacking unless you spend £££
- There was no swimming pools on release because they didn't know how to do them? I mean they have been a pretty essential part of the series previous and the content stripping here really hurt them
- The areas where you buy a plot of land are not really very large and feel stark like stage setting rather than interactive world
- There is/was no customisation clothing options at least without modding it which is stupid. I get as PC gamers we play with mods but the basic functions should still be there
- The construction tools to actually build a house are good
- The graphics are prettier in comparison
Saying that, SimCity is really the one that died and was stripped of any life, gameplay or interest.
Anyone with a modicum of technical knowledge can through see this techno-babble they're putting out. "AWS can provide servers all over the world that are linked"
They didn't need to switch to Lumberyard to use AWS. By switching to Lumberyard though, they are now tied into AWS and can't use any other cloud provider. Oh and they've opened themselves up to this Crytek legal action, and spent a load of effort doing things that *aren't* proving that Star Citizen is even possible.
Still. 5+ years down the line and the basic tech still isn't proven.
Soldato
Anyone with a modicum of technical knowledge can through see this techno-babble they're putting out. "AWS can provide servers all over the world that are linked"
They didn't need to switch to Lumberyard to use AWS. By switching to Lumberyard though, they are now tied into AWS and can't use any other cloud provider. Oh and they've opened themselves up to this Crytek legal action, and spent a load of effort doing things that *aren't* proving that Star Citizen is even possible.
Still. 5+ years down the line and the basic tech still isn't proven.
No they didn't need too but it means they are getting the base concept of server coverage with the use of the engine for free. They can actually still use another service but have to pay a standard licence agreement for the engine just like they would have with CryEngine so it doesn't lock them to anything.
I am not going to go into the **** with CryTek but put it this way the two studios I have been doing a few bits of work for at moment have dropped CryTek and moved to Lumberyard following that rubbish. They don't believe in CryTek management or integrity anymore bearing in mind I know one that has pretty much the same GLA with the same clauses and they have always utilised two engines for their game to test out which is viable for what they want.
And I wont go over again why this 'basic' tech takes 5 years to build from scratch with a team you are putting together and trying to hire accordingly. If you had the person in networking day one then yeah 5 years, they didn't employee the total team network side till mid 2015. They are still looking for network software engineers today and have been advertising for 9 months with positions for that team but not found the right people at least that are actively looking. That can stop people completing work too.
The whole point is reality dictates they have only been working on this for around 2 years and the initial concept design and systems the previous 2 years. The 5 years is rubbish as pointed out so many times. When the company you work for has a single person have an initial concept how long does it take to go from that first thought to even having a working tech demo. It was around 2 years for us to get through that stage with the initial team and go to the publisher to prove our concept before we even got into the meat of development.
CIG started at the idea in the head stage bar what he paid CryTek to produce for advertising.
Associate
It doesn't matter how good Amazon's network engineers are if they aren't actually working on the features that CIG require for CryEngine. There's this weird belief within the SC community that because CIG are using Lumberyard, Amazon are doing all the work for them. That's not how it works.
CIG are currently on their second iteration of their networking stack. This iteration is going to be thrown away when their actual, it's the real netcode this time honest guv, version comes out at some undefined point in the future. Doing this work over and over is incredibly wasteful. What really annoys me is that they throw around technical jargon like "culling" and "object container streaming" and "variable serialisation" knowing full well that it's enough to keep the majority of the core backers happy, and it's treated as though these things have never been done before. They *have* been done before. Lots of multiplayer games use similar technologies, but they don't feel the need to crow about it -- they just get on and implement it.
So a company that's developing an MMO didn't think it important to get their network team in place from the beginning? They thought that it was sensible to do three years of development before ramping that team up? The network model is *fundamental* to a multiplayer game; it's the foundation that you build your game mechanics on top of. You can't just say "I'm sure that we'll figure it out", go off and scope your entire game and start developing it without knowing if what you want to do is even possible! Well, you can if you're Chris Roberts I guess.
And using "It's hard to recruit people" is a rubbish excuse. It's easy to recruit people, especially when you've got hundreds of millions of dollars in funding. Money talks, you see. Even if you can't recruit, you go out and find a consultancy who do that sort of thing and you pay them to do the work. Or you find expensive contractors who would be happy to provide their services in exchange for wodges of cash. It's not like CIG don't know this -- they've used consultancies all over the place for other bits of work. They just didn't think that the network part of the system was important, and that's a huge and fundamental mistake.
The 5 years is only rubbish if you're so invested in the game that you can't bear for it to be criticised. Like it or not, the game has been in development since *at the latest* 2012. Your example of your company taking time to build a concept for the publisher is moot because that's effectively what CIG were doing in 2011-12 in their crowdfunding push. Since the beginning of 2013 they've been extremely well funded, and should have had no trouble building prototypes and iterating over the gameplay and networking fundamentals. Basically, at this stage they should have the majority of gameplay blocked out and it should look really ugly and be full of bugs. They'd at least know the concept is sound and they can begin the process of tearing it down and building it back up again, but shiny this time. But that's not how they work. They don't make millions of dollars that way. They make their money by making shiny tech demos and selling pictures of ships or land or tanks. They make their money by piling up technical debt in the hope that one day they can pay it off, but with no real plan for how to do it.
They've only been working on it for two years? That's not what I was told when I backed the game in 2014. Were they lying? Where is Squadron 42, I was supposed to be playing that by now. An MMO requires an experienced networking team? Why didn't somebody tell them sooner?!
There's a reason why they're struggling to fill positions btw....why the hell would a decent qualified engineer go to such a shambolic outfit when more reputable, less risky studios will be lining up to take them on. The ambitions that CR has for SC are so far detached from what they have thus far been able to achieve, who the hell wants to be "that guy that failed to get Star Citizen's back end to work".
There's a reason why they're struggling to fill positions btw....why the hell would a decent qualified engineer go to such a shambolic outfit when more reputable, less risky studios will be lining up to take them on. The ambitions that CR has for SC are so far detached from what they have thus far been able to achieve, who the hell wants to be "that guy that failed to get Star Citizen's back end to work".
Soldato
I didn't say that. Stop taking things out context to make a stupid point.
Soldato
Not suggesting they are doing anything direct for CIG, I am suggesting they are working hard on their AWS service overall that should offer one of the top cloud based server providers in the world. Very different thing but still critical to what CIG require. The partner providing this really needs to be at that level, Google & Amazon are generally the go too at moment for that.
I also didn't say that what they wanted for the MMO meant they didn't think of it but the engineers needed are few and far between. Regardless of money, people often haven't got the luxury to move hundreds or thousands of miles and they haven't got unlimited monies to head hunt as that costs. They are picking up the team to make it work. They had some network guys in 2013 but they were the start of a team. You can't click fingers and just go oh hey just what we need. They got supper lucky with CryTek messing up tbh and Amazon getting Lumberyard sorted in honesty.
They have never suggested they didn't think it was important for the networking either. That is an assumption you have made to suit your narrative.
I haven't been invested. I waited two years watching before buying-in in September 2014. And yes of course development goes that way. If you go on a studio tour you would see the base mechanics in real rough buggy mess for the mechanics. They just show the public the shiny. The same happens when you show a publisher. The difference is they also see the game design doc or feature design doc.
The networking fundamentals have been worked on for two years. There is no way their aim is to get the system up and running in September and not have a few years behind them already.
Both you and Mig have taken a stance because you can't see them doing stuff although both seem know the business. That doesn't make sense. The obvious reason we see what we do and the level we do, for instance stuff that is visually polished is answered in that they have to get funding from backers seeing nice things. Would it be good to have it more open and see the rubbish that woudl be scrapped yeah. I never understood the NDA that you have to sign when going to studio. There loads more there they have never shown but they seem content with only showing what gets to tier 0 to make it somewhat playable.
Associate
- Joined
- 30 Oct 2017
- Posts
- 29
Curlyriff, I’m going on a tour of the Manchester studio next Friday hopefully, really can’t wait.
I'm still amazed by how much most people are missing the point, missing the procedurally-generated forest for the high-fidelity trees.
It doesn't matter that CIG released the game in this state to the Live environment. The frame-rate doesn't matter. The patch cycle doesn't matter. The so called "netcode", which every network engineer will tell you is not even a thing, doesn't matter. The point is that you can't treat CIG like a game development company because CIG is not a game development company. They are a company creating digital assets for sale and they're attempting to cobble together an online world in which to use those assets. The company is led by a has-been game developer who is not qualified, and in fact uniquely unqualified to do his job. The company has produced nothing, has no history, and no path forward. Holding them to the standards of a game development company is unbelievably naïve.
That's one part of it.
The next part is the assumption that CIG gives a flying about what they release to the PTU or whatnot as if it's all part of normal game testing and deployment. It isn't. Every single byte of code that lands on the PTU, or anywhere else, is done for purely optic reasons - not to do anything for backers or the people who have funded this. It's to create gaming journal headlines like "Long-Awaited 3.0 Patch Goes Live" just in time for backers to spend their Christmas money. It's so they can create a false narrative of progress (like how 2.0, 2.1, 2.2, now 3.0, are all numbers going up hence progress is really happening guys, no really!). So when I see people arguing about the relative quality of what's released to "Test" vs. "Live" it actually makes me sad. Going into this with the belief that the "Test" and "Live" environments are being used as if Roberts understands them, how they're used, why they're used, etc. It literally hurts to listen to.
Once you get past the fact that SC has nothing to do with the game promised during its kickstarter anymore, instead of banging your head against the wall trying to understand this project, you will actually be able to predict its course. Soon there will be a concept ship sale. No matter the state of the game at the time, no matter the bug count, no matter the lawsuit status - there will be something new to spend your money on, maybe a sandworm pet? Anything and everything CIG can do to move dollars from your bank account to theirs is their #1 priority all day and all night, 365 days of the year and that is the sad truth.
It doesn't matter that CIG released the game in this state to the Live environment. The frame-rate doesn't matter. The patch cycle doesn't matter. The so called "netcode", which every network engineer will tell you is not even a thing, doesn't matter. The point is that you can't treat CIG like a game development company because CIG is not a game development company. They are a company creating digital assets for sale and they're attempting to cobble together an online world in which to use those assets. The company is led by a has-been game developer who is not qualified, and in fact uniquely unqualified to do his job. The company has produced nothing, has no history, and no path forward. Holding them to the standards of a game development company is unbelievably naïve.
That's one part of it.
The next part is the assumption that CIG gives a flying about what they release to the PTU or whatnot as if it's all part of normal game testing and deployment. It isn't. Every single byte of code that lands on the PTU, or anywhere else, is done for purely optic reasons - not to do anything for backers or the people who have funded this. It's to create gaming journal headlines like "Long-Awaited 3.0 Patch Goes Live" just in time for backers to spend their Christmas money. It's so they can create a false narrative of progress (like how 2.0, 2.1, 2.2, now 3.0, are all numbers going up hence progress is really happening guys, no really!). So when I see people arguing about the relative quality of what's released to "Test" vs. "Live" it actually makes me sad. Going into this with the belief that the "Test" and "Live" environments are being used as if Roberts understands them, how they're used, why they're used, etc. It literally hurts to listen to.
Once you get past the fact that SC has nothing to do with the game promised during its kickstarter anymore, instead of banging your head against the wall trying to understand this project, you will actually be able to predict its course. Soon there will be a concept ship sale. No matter the state of the game at the time, no matter the bug count, no matter the lawsuit status - there will be something new to spend your money on, maybe a sandworm pet? Anything and everything CIG can do to move dollars from your bank account to theirs is their #1 priority all day and all night, 365 days of the year and that is the sad truth.
Associate
I've not tried this since I got a copy free with an AMD gfx card
I take it from the last few comments that it's still not in a playable state?
I take it from the last few comments that it's still not in a playable state?
Associate
Of course, but the issue is that even with a solution like AWS the server architecture for an MMO -- and especially for one with twitch-physics -- requires a bespoke design. What using AWS or any other large cloud provider does is take some of the effort out of the infrastructure: dynamic instancing of servers, data replication, etc.
No, you can't just click your fingers and hire the best people. However, with a good remuneration package you should be able to get a decent enough team in place in under a year. If they've had positions open for 3 years, that rings big warning bells for me: they simply aren't willing to pay for top talent, or top talent isn't willing to come work for them.
No, it's an inference I've made based on what I've seen so far during development. I see CIG sponsoring an open source game networking library (libyojimbo) and I wonder what that's all about. Were they planning on using that in their product? I hear them talk about problems like the server FPS (?!) influencing client FPS and that they know how they can fix it. I hear them talk about how they've been working on their new network stack for two years, but they haven't demonstrated *anything* working with it. Given how much they like to show off what they can do, I would wholly expect them to be able to create a fancy tech demo that shows thousands of clients interacting with no obvious instancing, which is what they claim they'll be able to do. I look at all this and *infer* that they're struggling with the network aspect of the game, and my *guess* is that it's because they had no idea how complex it is to design and that it's not something you can just bodge into a game that's already halfway through development.
But for a project that claims to be THE MOST OPEN DEVELOPMENT EVAR they really ought not to be scared about hiding things away.
What worries me is that a) they've been working on the new network stack for two years and we've not seen hide nor hair of it; and b) such a large piece of work is not going to be trivially integrated with a target that's been moving for the duration of its development. It's why we say that the fundamentals need to be in place *before* you start work on the game proper, because if you don't you're opening yourself up to a world of pain when you want to integrate it.
I completely get that some backers need to see the shiny. But what we're finding, particularly of late, is that even some of the more stalwart defenders of the game are starting to get a bit itchy about the lack of gameplay, and the sheer amount of bugs that are plaguing even a polished build of 3.0. Yes, I know that it's in Alpha and bugs are to be expected but this build went through a lengthy Evocati process.
I want to make it clear that I'm a backer of the game, and I want to see it succeed because I'm a fan of space games and SC as pitched appeals to me. But I'm disappointed in how the project has progressed, particularly regarding the huge amount of scope creep that's evident and the state of 3.0 makes me extremely nervous as to its viability going forward. I sincerely hope that I'm proven wrong, but as an outsider looking in, all I see is a project that's being hopelessly mismanaged and that's desperately in need of having its scope frozen.
Associate
It's playable, in that you can play it. It's graphically reasonably impressive and theoretically there's a bunch of missions that you can do, but most of them are bugged. If your client crashes, for instance, your ship will remain where it was when you crashed but you'll wake up back in the starting location and you'll have to wait a few arbitrary minutes while your game decides to give you a new ship. As far as I know, CIG aren't planning on hotfixing any of the known bugs so it'll remain this way until 3.1 drops, which is slated for end of Q1.I've not tried this since I got a copy free with an AMD gfx card
I take it from the last few comments that it's still not in a playable state?
Oh, and if you get on a fresh server you'll get decent (~30) frames per second but it doesn't take long for that to tank to < 15 fps. There are certain things that can happen (e.g. a ship containing a lot of cargo blowing up) that causes frame rates to tank even further, sometimes below 5 fps. It's a frustrating experience but still worth giving a go.
Associate
- Joined
- 30 Oct 2017
- Posts
- 29
Where did you see that there aren’t plans for hotfixes? CIG have rolled out hotfixes to 3.0 since it’s release.
Associate
Thanks @xyphic I may give it a download, but those framerates sound terrible
Soldato
Em no they haven't actually :/
They have done some server side hot fixes to the background applications like interdiction but they are not looking to do much hotfix other than sorting out the major bugs like the player sphere issue. I think at best we will get maybe two hotfixes before 3.1 tbh.
They have already said they have moved almost everyone from 3.0 branch back to main dev branch. Then they will pull 3.1 branch off the dev branch end of February which would suggest 4 weeks of Evocati and bug fixing accordingly to get end of March 3.1 out with the optimisations and the new UI etc.