Game Discussions Star Citizen Discussion Thread v12

[Agony_Aunt]

Interesting topic, although some of the details and history are incorrect

Source: https://www.youtube.com/watch?v=PcJ-EPuzGQE

From that video

"Turbulent are working on all the backend"

I have no idea how true that is, but combined with what Ebob is peddling, its doubly hilarious.

A company that was until a few years ago a reputation management company has taken over the backend stuff from CIG, (a company built from the ground up to develop the game with over 700 people).

That is so funny.

Anyway, funny regarding the comments how the fans are fine with what he's saying and saying they are ok if there are smaller regional services, etc. If a skeptic had said the same things (and i'm sure they have) the faithful would have been all over them screaming they don't understand game development.

Listening to Noob, its good to see him asking things like that, acknowledging that CIG haven't even delivered on the stuff pitched in 2013 yet, although to my mind, he still seems to think some things might be possible which i don't think will happen, but hey, we can have different opinions there, its a large unknown quantity.

 

[Agony_Aunt]

Good arguement!

Not an argument. Just saying i disagree and leaving it there, rather than continuing the debate for another 10 pages with neither side convincing the other they are wrong.

 

[VR Golgot]

Except thats not how you do it. If you're gonna develop a feature after you've gotten your network stuff done and ready, you'll have to change it according to your network stuff. And if said feature encounters technical hurdles that don't work with your network stuff you'll need to rewrite the whole thing. They first got an idea of how such game with interconnected physics grids, container streaming and all that is going to even work and pan out.

Lollll

No…

A) They’re having to rewrite how their features work to fit with the new networking architecture as it is. (That’s why you can see teams from AI & Vehicles, to Game Services & Persistence, to Audio & UI etc in the Server Meshing task).

B) Designing your game first then hoping some miracle networking system can make it come alive is an insane gamble. And one that currently isn’t panning out. CIG’s own announcements suggest that they have no idea how to support various key features. (Single shard, high player counts, capital ship battles, uninterrupted traversal of a solar system etc). Classical game limitations, caused by networking limitations, are now enforcing themselves on the game design.

I’m sure Chris was told by wiser heads back in the day that his dream networking system doesn’t scale. But Chris doesn’t like to hear the word no. So now’s he’s hearing it the hard way…

That's not "downright incompetence", that's iterative development.

You are lost in buzzwords friend. And making stuff up as you go along. It’s very SC

 

[TiberiusDuval]

All of the stepping stones to server meshing like Object Container streaming and such had to be tested on the game first, integrate them, test, move onto the next thing on your road to making Server Meshing. Because really, the utter basics of a multiplayer structure are in SC and have been for years.

Nope. They are doing it backwards. Basically you do networking stuff and estimate what kind communication needs you do need. Over that foundation is then added what ever stuff game has. Not like you do most of the game first, find out that your networking stuff pukes under load, and then try to jury rig network component to get out some functionality. Thats like building house first on crumbling foundation, and then try to repair foundation when it gives way...

 

[Ebob]

A) They’re having to rewrite how their features work to fit with the new networking architecture as it is. (That’s why you can see teams from AI & Vehicles, to Game Services & Persistence, to Audio & UI etc in the Server Meshing task).

Make feature work gameplay wise first before integrating it into 50 shard networking system -> make sure feature works with that and then integrate into new networking system! Simple!

B) Designing your game first then hoping some miracle networking system can make it come alive is an insane gamble. And one that currently isn’t panning out. CIG’s own announcements suggest that they have no idea how to support various key features. (Single shard, high player counts, capital ship battles, uninterrupted traversal of a solar system etc). Classical game limitations, caused by networking limitations, are now enforcing themselves on the game design.

But this is no miracle networking system, this is bog standard destiny networking as you've said yourself!

 

[TiberiusDuval]

Does Elon Musk say this to crowdfunding backers a decade after slapping "Alpha" on his ships and constantly misleading about release dates "coming 2014", "answer the call 2016", "answer the call 2017" etc then? Otherwise I fail to see the relevance?

Elon probably has real rocket engineers doing their work, and he does not micromanage every detail

 

[VR Golgot]

Make feature work gameplay wise first before integrating it into 50 shard networking system -> make sure feature works with that and then integrate into new networking system! Simple!

Armchair dev in a nutshell…

But this is no miracle networking system, this is bog standard destiny networking as you've said yourself!

Yes exactly. They don’t know how to make the miracles happen, so they’re doing something tried and tested, and likely culling all the 'miracle features' which they can’t support. (Such as the $3000 battleship engaging in combat with 80 NPC & player crew running around inside…)

We can only wait and see what other promised features are limited by the hard caps this system will involve.

 

[VR Golgot]

Some lunchtime Spectrum salt…

But at the same time, after 10 years of developing a game that revolves around flight and for which MultiCrew is one of its unique selling points;

• Scanning/ Stealth is dysfunctional: still can't differentiate between EM/ IR/ CS when pinging (no, I'm not always interested in the astroids and ore).
• Distortions are dysfunctional: no interaction with anything other than shields, and even that is limited
• MoM is dysfunctional: can't be controlled by co-pilot even though it was said to be possible (and it should have a major advantage)
• Missiles are dysfunctional: some don't lock, others do but don't fire, flight paths are all over the place
• Flight model is dysfunctional: not only not finished, with atmospheric flight seemingly not even on the roadmap, no... Prox Assist keep switching on etc.
• Physicalized Armor not even in the game yet, not even on roadmap either (unless I missed it).
• MFD's are often dysfunctional, not remembering what MFD is set to what even when you so much as F4 + being hailed by NPCs on target info MFD
• Starter Ships are dysfunctional... STARTER SHIPS

All in a game where it's MOSTLY about flying and players are forced to buy a ship to participate.
All highly essentially required before making ANY meaningful balance changes.
No server meshing required.
No Gen12/ Vulcan required.

These are the things I would like to see finished FIRST, or at least clearly take a top priority this year.
That way, what we have (AC + PU), can be fully enjoyed!

There's a lot of cool stuff on 2022 roadmap! - Star Citizen Spectrum

But at the same time, after 10 years of developing a game that revolves around flight and for which MultiCrew is one of its unique selling points; Scanning/ Stealth is dysfunctional: still can't...

robertsspaceindustries.com

Top comment:

CIG is saying the game is a full service/delivered product (in court).

Can we finally stop using the “its Alpha” excuse? They are paying dividends and citing yearly “profits” .

I just cannot understand how people can be die hard defenders 24/7? Yes, they have some wonderful game play, but its a long way from bug free or finished.

This revolving door of excuses that don’t even come from CIG (unless astroturfing) are really getting old.

/rant over

 

[VR Golgot]

https://www.reddit.com/r/starcitizen/comments/ryqqbb

Source: https://old.reddit.com/r/starcitizen/comments/ryqqbb/i_love_this_game_but_i_just_had_to_do_this/

 

[Ebob]

Armchair dev in a nutshell…

Well Golgot have you worked on video games before?

Yes exactly. They don’t know how to make the miracles happen

I doubt anybody knows exactly how to go faster than light...

so they’re doing something tried and tested, and likely culling all the 'miracle features' which they can’t support. (Such as the $3000 battleship engaging in combat with 80 NPC

Actually 80+ NPCs actually already work in the PU, Ninetails lockdown. So no that is not going to get culled...

We can only wait and see what other promised features are limited by the hard caps this system will involve.

Doom! Features gone! Never to be seen again!

 

[VR Golgot]

Interesting topic, although some of the details and history are incorrect

Source: https://www.youtube.com/watch?v=PcJ-EPuzGQE

Yeah he’s on weak ground claiming a single shard play space was pitched in 2013. They were always claiming they’d use instances at the gameplay level.

The area where he’s stronger is on capital combat ships, which CIG merrily sold from the off with no viable technical plan for delivery (on the networking side alone). Although he does veer into hopium a bit by suggesting capital scale battles alone could be fine if their interiors were instanced. (An approach that the Zenimax dev suspects wouldn’t work for combat, for example. And which would imply some notable running costs even if it did, which the CIG Q&A suggested they’re looking to avoid).

All told though, it’s good to hear a regular calling out their current carrot dangle of a single shard still being possible (when it clearly isn’t, re regional ping at minimum). And to hear him suggest historical claims & sales may be "a tad fraudulent"

 

[Agony_Aunt]

Some lunchtime Spectrum salt…

There's a lot of cool stuff on 2022 roadmap! - Star Citizen Spectrum

But at the same time, after 10 years of developing a game that revolves around flight and for which MultiCrew is one of its unique selling points; Scanning/ Stealth is dysfunctional: still can't...

robertsspaceindustries.com

Top comment:

All in a game where it's MOSTLY about flying and players are forced to buy a ship to participate.

I beg to differ, its a game about making pretty screenshots and on-foot gameplay primarily. Watch most streams and videos and compare how much time they spend flying spaceships vs how much time on foot.

And its no surprise, CIG have put way more effort into the on-foot game than the space game, in this game about space.

I'll admit, in Odyssey i'm 99% now doing ground based activities, but there again, i've been doing space based stuff for 7+ years, on foot stuff is a nice change of pace and scenery.

 

[Agony_Aunt]

https://www.reddit.com/r/starcitizen/comments/ryqqbb

Source: https://old.reddit.com/r/starcitizen/comments/ryqqbb/i_love_this_game_but_i_just_had_to_do_this/

LOL, tagged as drama.

 

[Agony_Aunt]

Yeah he’s on weak ground claiming a single shard play space was pitched in 2013. They were always claiming they’d use instances at the gameplay level.

The area where he’s stronger is on capital combat ships, which CIG merrily sold from the off with no viable technical plan for delivery (on the networking side alone). Although he does veer into hopium a bit by suggesting capital scale battles alone could be fine if their interiors were instanced. (An approach that the Zenimax dev suspects wouldn’t work for combat, for example. And which would imply some notable running costs even if it did, which the CIG Q&A suggested they’re looking to avoid).

All told though, it’s good to hear a regular calling out their current carrot dangle of a single shard still being possible (when it clearly isn’t, re regional ping at minimum). And to hear him suggest historical claims & sales may be "a tad fraudulent"

Yeah, early on it was instances like in ED, but then somewhen, i think around 2014-2015 they started pitching the massive instances thing.

 

[Agony_Aunt]

People still posting on the comments section of the kickstarter

Star Citizen

Reclaim the stars in the exciting new Space Epic from legendary game designer Chris Roberts.

www.kickstarter.com

Mainly salty comments with the odd faithful comment in there.

 

[VR Golgot]

Well Golgot have you worked on video games before?

Have you?

The difference between us is that I’m citing developers, and CIG’s own claims, to support my arguments.

Whereas you’re just pulling claims out of your… magic bag of hopium

I doubt anybody knows exactly how to go faster than light...

Which suggests designing a 'dynamic meshing' system that ignores that rule was foolish. And claiming a single shard is still possible, with the proposed architecture & gameplay, is an incredibly dubious 'carrot on a stick'. Also known as lies

Actually 80+ NPCs actually already work in the PU, Ninetails lockdown. So no that is not going to get culled...

You are forgetting the scaling required:

• No capital ships with interiors have been tested in the PU. (And those without interiors hammer performance as it is).
• How do multiple capital ships engage with 80+ players & NPCs each? Given performance is currently poor with less, and server caps aren’t expected to get beyond 100, they’re going to hit various walls at once.

Only the miracle of 'Dynamic Server Meshing' promises significantly 'denser numbers of AI and player characters', in CIG’s words. You are welcome to hold your breath for it

Doom! Features gone! Never to be seen again!

I never said doom. I did say some of CIG’s big pillar pre-sold features aren’t possible with the proposed networking though, yep.

You realise that introducing dumb memes which nobody said is daft right?

 

[TiberiusDuval]

Well have to say, probably you could get some pretty beefy server to run very demanding stuff on single shard. It will just cost arm and leg. Nothing like el cheap rent a capacity from AWS stuff for parallel loads with basically PC hardware. But something really beefy dedicated stuff. Like what IBM sells even nowadays.

Reminds me of one smaller example, one association I'm member asked me to handle purchase of server hardware, I asked what is the budget..."it can cost whatever it costs". Well after little checking on hardware store I said, yeah I can order anything from complete system under 1000 euro, or system where just CPU costs 20 000 euros. I really need to know how much it is allowed to cost. (Long story short, instead of physical device we ended up on renting vps...)

But that is just single shard, add up thousands of those and price goes to astronomical fast....

 

[Factabulous]

Well have to say, probably you could get some pretty beefy server to run very demanding stuff on single shard. It will just cost arm and leg. Nothing like el cheap rent a capacity from AWS stuff for parallel loads with basically PC hardware. But something really beefy dedicated stuff. Like what IBM sells even nowadays.

You can pay what you like for hardware - the speed of light doesn't care and won't reduce your latency.

 

[TiberiusDuval]

You can pay what you like for hardware - the speed of light doesn't care and won't reduce your latency.

That is true. I just meant that OK you want to simulate say couple of Idrisis banging up each other plus what ever other players. Yeah you can get something that has enough oomph for that. But yes you cannot get away from latency. And anyways you cannot economically get enough those really big beast machines to scale up your network for a big MMO.

 

[VR Golgot]

Yeah, early on it was instances like in ED, but then somewhen, i think around 2014-2015 they started pitching the massive instances thing.

Yeah the earliest quotes I’ve found were from 2015, and were more provisional. IE here and here:

So there wasn't really ever talk of a thousand per server instance there was talk you know, which is a sort of longer range goal, about the possibility of getting multiple servers to work cooperatively together on one sort of big area instance and split it up. So if that was the case, if you had, if you had 10 servers and each server could handle 100 people then potentially get a thousand people, but you know, that's sort of a very long term goal. And you know it does have its own issues because, you know, at some point you run into rendering issues and all the other things that would happen then

By 2016 it was more obviously the plan:

We're going to have this sort of mesh of servers, so we'll be able to have, hopefully, a large amount of players all in the same area, so we won't have to instance it, in the way that originally we were thinking we'd have to instance it. We have a kind of different server design now that could potentially have thousands of players all in the same sort of area at the same time, which would be really cool.