Game Discussions Star Citizen Discussion Thread v12

Source: https://www.reddit.com/r/starcitizen/comments/1blj83n/going_from_one_server_to_another_is_super_smooth/

Video shows interactions across servers, just like what was demo'd at citcon

Reddit OP in comment said:
The physics grid of microtec is a sphere that rotates at the speed of microtecs own rotation, so if you go out of it you move away at a couple of thousand m/s. I prevented this by going to the north pole (OM-1) of microtech and flying to the north pole of the microtec physics grid sphere. This transition is only possible at the north and south poles of planets, because there is no rotation there!
 
No reason to dunk - I'm actually happy to finally see some solid, core-tech progress towards the game I kickstarted almost 12 years ago; I've been waiting 6 years since Chris and Erin began touting the "server meshing coming next year" bovine excrement.

tq7GcdZ.png

AUAzgbG.png
 
lol, that is not really an issue.

To each their own. If i'm doing a lot of Odyssey stuff in ED those large hangers take time to cross and it would be more annoying if you had a sidewinder spawn on a large pad.

The size difference between a sidewinder and an anaconda is significant, and i think the size difference between the smallest and largest ships (which can go in hangars) is even bigger.
 
Feature Incomplete:

Lol, so they're still trying to morph body shape. And it's a whole world of pain :unsure:

Source: https://www.youtube.com/watch?v=XsEKbJmtuW4&t=2400s

If CIG didn't plan for armours/clothes to adjust to different body shapes they will be in a world of pain redoing them all.

Also, if they try and balance different body shapes/sizes by affecting their characteristics, then everyone is just going to max out the sliders for max benefits. Dumb idea.

I smell another refactor!

Oh, and of course, ships and animations, climbing stairs and ladders.... i smell death in the air for being too big/small when climbing ladders.
 
Ok, a variable changes according to player's position... sorry, dishonest troll.

So theoretically, let's say a mesh handles 50 players max, 100 players can see eachother in the vicinity of this area.

What about performance client side?
What if a ship crewed with 10 people goes from one mesh to the other?
What if one mesh crashes, what the other sees? What happens to crashed mesh's players? What if players from the sane mesh go to the area handled by the other? Or even just shoot at?

So yeah, dishonest troll but there's so much questions about edge cases and I know CIG's history well enough to doubt they actually ask themselves about edge cases beforehand.
 
Last edited:
Ok, a variable changes according to player's position... sorry, dishonest troll.

So theoretically, let's say a mesh handles 50 players max, 100 players can see eachother in the vicinity of this area.

What about performance client side?
What if a ship crewed with 10 people goes from one mesh to the other?
What if one mesh crashes, what the other sees? What happens to crashed mesh's players? What if players from the sane mesh go to the area handled by the other? Or even just shoot at?

So yeah, dishonest troll but there's so much questions about edge cases and I know CIG's history well enough to doubt they actually ask themselves about edge cases beforehand.
With news of suitable test locations - at the poles of worlds - spreading among the SC community, the streamers will certainly be rallying large numbers of people to test together at such server borders. Moving, crossing, interactions, fighting, etc, seeing what works, what breaks, what people see and don't see, etc.

And with all the crashing, instabilities and 30k's happening, CIG have a lot of fixing to do too.
 

Viajero

Volunteer Moderator
Looks like a bit of the CitCon demo reenacted at one of the poles of that planet. Seems that sweet spot at the north pole was due the fact most physics grids transitions have significant motion and speeds difference. Which was probably always the case, it is the same in Elite.

But that begs the question: Will the majority of boundaries be hostage to that relative grid motion issue? Because if that is the case it means there will be little or no effective interaction among players or entities across server boundaries (unless you want to go to the planet poles).

That renders "seamless transitions" pretty much useless in the firs place I suspect. Unless you can use it dynamically eventually for large capital ship battles or something where servers and grids are assigned on the fly, made to match across borders, as ships or vehicles spawn, get destroyed/damaged or move around etc.

Thing is even this version of static, it all looks tremendously unstable, by the looks of it. Perhaps too early to tell, but even in this reddit post where the physic grids aligned perfectly the 2 or 3 participants crashed a few times etc. That was even less impressive than the CitCon demo, which was somewhat broken anyways.

Something else that is not clear yet is the unloading/loading bit. Between Stanton and Pyro I think it was quite clear the client has to unload all assets / entities and load new ones when changing star system. Within a system though, in Stanton, what assets / entities etc, need to be unloaded from the original server, and which ones need to be loaded from the new one, if at all?
 
Last edited:

Viajero

Volunteer Moderator
Berks summary vid is a pretty good representation of how it went (though his viewpoint is on the sunny side as ever)

Source: https://youtu.be/kanNz9qTDQA

That reflects well the significant instability we saw yesterday.

Also, the player list changed when transitioning servers it seems. Even though they are all supposed to be in the same shard / replication layer. That would suggest interaction between them would be somewhat limited or even not possible? Whether at boundaries or not?
 
Last edited:
Back
Top Bottom