That is the oddest euphemism for butt that I have ever seen.
Game Discussions Star Citizen Discussion Thread v12
- Thread starter Yaffle
- Start date
-
- Tags
- star citizen
They didn't have that money when they needed a sizzle reel to give the impression that they already had something in development, though.
I've long given up giving two hoots and, as has been seen on numerous other occasions, CIG bizarrely still seems to attract backers, so calling the beginning of the end is as futile as it is risky.
However, in wondering what the end of this mess might be, I've long felt that a slow decline to inevitable ceasing of meaningful work is quite possible. The faithful can applaud themselves for backing a visionary dream. The rest can shrug, mumble, and move on. Others can retire as millionaires.
And on that ending, this would seem the logical start. Cutting jobs, less and less meaningful output, and, if possible, ever more buggy output. But, as I say, with this farce, it's too hard to tell. So I'll just sit quietly on the sidelines, watching the cluster-farce as it crawls ever onwards.
Buy an Idris.
You sir, win the prize of the Internet for that post
If I had a grand in the game (and a semi-pro career hooked to it) I'd perceive ongoing success too
Best thing Tomato ever did was interview that based fan-dev, who laid out some pragmatic thoughts on Server Meshing...
Best thing Tomato ever did was interview that based fan-dev, who laid out some pragmatic thoughts on Server Meshing...
Decent chat here with an MMO code monkey working at Zenimax:
Source: https://www.youtube.com/watch?v=C41vT-f5dSc
TLDR:
- Static Server Meshing is comparable to systems used by current games, such as Destiny. (IE 50 people in Stanton, 50 people in Pyro, and the jump point being a loading screen). [15m25s]
- "With the best will in the world, I don't think we'll ever see a situation where combat can take place over a mesh boundary." [17m52s]
- Capital ship battles will all have to be on one server [18m50s]
- Max player count in one place? He guesses: 100, maybe 200. [26m40s]
- "I know from my own side as a programmer you can sit there and say: This is what the limit's going to be. And there will be stakeholders who go: No, I don't believe that. Until you actually prove to them what happens when those limits get exceeded..." [41m55s]
- "If the Static Meshes come in, and work as planned, you will see absolutely nothing different". Other than things that currently exist performing a bit better, and potentially opening up Pyro, not much will change. Tony Z might have more headroom for the background sim. [And Chris for coffee cup fidelity
Notes:
On his dev background:
- "I'm one of the code monkeys, backend, for Zenimax, on their games." [1m19s]
- Wolfenstein etc got him into netcode, and for a long time that was his thing professionally. Not a graphics programmer etc. Very different world to his server programmer world. Summarises his works and interests as: "Network, backend, trying to control the flow of things." [5m20s+]
- Hasn't had any contact with CIG devs. Just working off own experience here. [41m30s]
On being a space sim fan:
- Backed SC in 2012. Playing semi-regularly since 3.10
- Is playing less of 'another one'. IE Elite Dangerous
- Has multiple SC accounts. Happy to play more villainous ways etc. [32m]
On entity groupings vs older approaches: [9m]
- In the old days bandwidth was the issue. Dial up etc. And even today lots of people don't have the best connections. So devs would send only the information they really needed to synchronise. Great for performance, but restricting. If you wanted to give someone a bag, you needed to work out how many things they could have in the bag. Needed to change all the netcode touching that player, and the bag etc. So anything you wanted to add, got hard.
- By comparison entities are much more "generic", especially when it comes to exchanging them with databases etc. When being moved around, all the nested entities inside don't need to be queried.
- In the old system it was easier to have checks and balances, validation rules, to make sure everything is consistent.
- Part of the reason the inventory UI went up so quickly, he thinks. is that all they're doing is showing you the nested objects inside an entity. (May even have had an internal tool to do that already).
- Recursive imbedding in entity system can become a problem in its own right. [Doesn't say how]
- Entity use far from unique. But scale that SC wishes to deploy them at may be.
On Static Server Meshing: [15m25s]
- A number of other games have, it could be argued, 'Static Server Meshing' in place. If SC was to launch with Static SM, Destiny is already doing the same sort of thing now. (IE 50 people in Stanton, 50 people in Pyro, and the jump point being a loading screen).
On Dynamic Server Meshing: [15m50s]
- Sub-dividing the game world very dynamically, and defining the static meshes much more 'generically' = the more ambitious bit. Games that are based around fixed transition points between servers have to be designed like that from the ground up. Using gameplay elements to hide the seams etc. SC is trying to do that much more dynamically. That's where the real challenge and the ambition is.
On blockading full locations: [16m50s]
- Ninetails blockades (blocking local quantum) would work as a way to prohibit entry to an area temporarily, making player take a longer route during which gameplay unlikely to occur (combat) during which a player can be moved to a different shard / instance etc.
On combat across mesh boundaries: [17m52s]
- "With the best will in the world, I don't think we'll ever see a situation where combat can take place over a mesh boundary.". There are so many edge cases. So it has to be a place that's relatively stable, where you have time for someone playing in Aus, on an EU server, to do the exchange of stuff they need to, from one instance to the next.
On capital ship battles: [18m50s]
- What about large capital ship battles? If they were actively in combat, almost definitely they'll all have to be on one server. If there's no combat going on, the actual carriers themselves could exist in a static 'shard', and have purely interiors existing in dynamic 'shards' [assuming he means server here]. Once you start firing turrets / where you need to start arbitrating decisions between different meshes, that's where there are so many edge cases. Doesn't think they'll try that.
- Lot of load on that single server? Yes. [Heavy sigh]
On being seated lowering demands on server etc: [20m46s]
- So much of SC's combat being seated, means it does limit what you need to communicate to others about you. Where is your turret pointing? Are you firing? What are your health stats. Not a lot else. (Far less, than you walking around. Which has a whole load of other attributes. You could be drunk, you could be leaping, crouching etc). [They joke about exaggerating force reactions to point where nobody can leave seats
- Being seated helps with desync. If you're seated it's much easier for the back end to give authoritative info, compared to all the info related to walking. Moving between meshes, would almost guarantee you'll have to be seated. When seated, you can basically be treated as part of the ship. Handy cheat / short cut. You're no longer responsible for your position. You're not costing the server or the networking anything at that point. (Underlines point about games normally faking a simulation, not actually simulating. It's all smoke and mirrors.)
On everyone in SC all going to one location: [24m45s]
- "Never, ever, ever, ever going to happen that you are all going stand there and see each other."
Possible max player count in one place? [26m40s]
- Guesses: 100, maybe 200. It's not what you can do with server performance per se. Some things have to be metered by the performance of the slowest player. (Min spec computers, bandwidth per region etc).
Possibility of more numbers in combat-free cities? [30m]
- While they're armistice zones, yes. But moment armistice zone goes away, no. (But says has never found an effective crime & punishment system in multiplayer game. Thinks more wallet impact would be an effective stick to guide player behaviour etc. [32m4s])
Base building: [34m50s]
- Still unsure as to exactly how they're going to work. If you can store ships, inventory etc there, he can't see them persisting in an attackable form when you're offline. [Even if it's on the shard that you were on]. What he could see, to tackle 'combat logging' style exploits, that if there were people / hostiles in the area, he could see keeping it around in that shard. But in general has to assume they're no different than bunkers. Even if you drop a bomb on them right now, it may destroy some defensive installations, or local players, but the base itself is indestructible.
Merchantman / Privateer marketplaces: [37m55s]
- Would have the same issues as bases, and as capital ships, on server meshing front. How are they going to exist in the game? Will they be attackable? Will it be that if it's in motion, it'll be attackable. If it's parked, maybe it'll be indestructible? (Similar to ED's Carriers etc).
On large owned items persisting: [38m55s]
- CIG will need to decide whether they allow these things to persist when owner not there etc.
- Items with low friction to ownership can cause issues. Imagine a world where every other person owns a Merchantman. (Not something that he thinks is likely, but something that could happen). Backend guys can't rely on players being sensible, you have to protect against the worst case. IE: What would happen if everyone bought a BMM and decided to park them all together?
On devs versus stakeholders: [41m55s]
- Won't hear more on the above until they've experimented with it and have some results.
- "I know from my own side as a programmer you can sit there and say: This is what the limit's going to be. And there will be stakeholders who go: No, I don't believe that. Until you actually prove to them what happens when those limits get exceeded... Some of the stuff is actually proving to both sides that there is an issue. And then seeing how both sides decide to solve it."
- "It's the battle. And why, you know, people always joke about the likes of Peter Molyneux, and other people who will say one thing... Anytime that I ever hear about any game, those things that they say this game is going to be is their vision.... Every game is an iterative process, and whilst you try to stay as close to that vision as you can, sometimes reality means you can't get that. Sometimes that's, y'know, reality brought on by deadlines."
What will Server Meshing look like next year? [45m35s]
- "If the Static Meshes come in, and work as planned, you will see absolutely nothing different."
- Thinks they're quite brave in making all of this detail public, because so much of this stuff goes on in a lot of games, but it's black box info. This stuff, along with load balancing, is normal behind the scenes stuff.
- Potentially will give a little bit more room to the servers. So you might start seeing more responsive AI, things spawning in when they're supposed to, increased item limits in the world.
- Other than things that currently exist performing a bit better, and potentially opening up Pyro, not much will change.
- Should allow Tony Z to do more fancy things in the Quanta background sim.
- The dangling pipes demoed in Pyro, which will have to be networked, will have more headroom. All those coffee makers, with their cups that you can pick up. You can bet that CIG would love for them to be there, working, but currently they don't have the headroom.
Personal Residence Buildings in Landing Zones: [51m25s]
- That could be one of the most trivial cases for a dynamic mesh. If only you, and/or a selection of friends, or even an org, have access. Lots of games do this for habitations. It's an instance. The tricky thing is faking that entrance into the room. Could imagine them using a very sterile L-shaped entrance hall for the transition etc.
- Can't have exact buildings for every potential player. But could have one entrance way behind which is a 'rotating pod' of options. (Could use rental system to make sure only active players were owning a location at any given time, in which case could have bespoke locations).
---
All fits pretty well with what we know from the official Q&A.
Viajero
Volunteer Moderator
Yeah, it is strange to use "continued success" with a seriosuly broken and incomplete product that keeps non delivering after 12+ years on what has been paid for with more than 12 times the funds that were asked for at crowdfund, and even then they had to ask external investors for additional funds (likely to avoid a cash crunch) incurring a debt now estimated in over 100+ millions which would basically shut down the company if recalled, where the developer does it darnedest best to avoid press reviews/scores and has not even dared to publish actual concurrency numbers for all those years.
So I guess I am going to go with Tomato´s made up definition of success being "I am an only SC, dedicated youtuber whose living depends on SC still limping along, please click here".
So I guess I am going to go with Tomato´s made up definition of success being "I am an only SC, dedicated youtuber whose living depends on SC still limping along, please click here".
Last edited:
Viajero
Volunteer Moderator
Here comes our regular reminder that the 100 planets at launch is still stated at the point of sale game description and therefore it is false advertising:
https://robertsspaceindustries.com/funding-goals
https://robertsspaceindustries.com/funding-goals
Success depends on the goal, YMMV.
Rare to see an official YT upload welling with brine, but here's one...
Source: https://youtu.be/UWuPHrDz0f4?si=UXEPApylbdtSmJBB
Etc. Regulars really are tired of event dysfunction
(I'm sure the happybots will be along in a bit...)
save stanton
save server
Etc. Regulars really are tired of event dysfunction
(I'm sure the happybots will be along in a bit...)
Last edited:
That is a release-ready product if ever I encountered one, I tell you what!
But I already have 10!
It's not called Lucky-11 for no reason....
Looks like sync issues to me. Some server is for some reason holding a position in an elevator and it's later asserting its position as the valid one at random times instead of being updated that it's old. Something something server meshing....
But why always Polaris?
(And why is lift asserting its right to orbit that station?
)I dunno, if it's not some dangling transit node in the Polaris this game is even weirder than I thought
This probably sums up the event...
The official event page is also pure salt.
robertsspaceindustries.com
Not unusual for Spectrum, but it does feel like their 'live alpha as a service' shtick is really breaking at the seams.
And the whole thing is made of seams
(It's like watching a solder spaceship firing up for its mission to the sun...)
The official event page is also pure salt.
Save Stanton! - Star Citizen Spectrum
Can You Save Stanton? With Stanton under assault by system-wide pirate attacks and the local authorities flummoxed at every turn, the people need heroes willing to put it all on the line to defend...
robertsspaceindustries.com
Not unusual for Spectrum, but it does feel like their 'live alpha as a service' shtick is really breaking at the seams.
And the whole thing is made of seams
(It's like watching a solder spaceship firing up for its mission to the sun...)
Never been done before, because, much like walking to the moon.......