#1k servers if the game becomes popular again

1 messages · Page 1 of 1 (latest)

naive summit
#

500v500, with a big skyscraper map, would be insane and cool, and also unique in the sense no game has 1k people simultaniously

hybrid steppe
#

Ohh the poor server....

paper path
#

Im not sure if that would be possible at all

naive summit
#

or why wouldnt it be?

paper path
#

Well, like ruphir pointed out, the server that has to track thw actions of 1 thousand people would get fucking incenirated

naive summit
#

and luckly the devs have it

#

30 million net

#

and with the big long awaited update, the game will have the player base to support 1 1k server at all time

paper path
#

Money cant exactly buy you a map that can support 1k people at the same time, gameplay that would complement that amount of people, a server that has to track complex actions of 1k people or hell im not sure the game engine can handle it even

hard lily
#

big if

#

never gonna happen at this rate

#

here's to another year of waiting

#

for the never coming update

paper path
# hard lily big if

Well the update will likely come, but as for how successful its gonna be, i am not sure

naive summit
paper path
#

The bigger if is if its good or not

hard lily
naive summit
#

they should release it anytime

#

one day

paper path
# naive summit you just gotta make a map big enough?

Not really cause the map needs to be made somewhat knowing how 1k people will interact with it and to not find it unfun, all while making it so that it doesnt melt ones pc with how much it has to render and proccess at the same time

naive summit
paper path
#

Its never that simple my guy

naive summit
#

but 30 millioon

paper path
#

Throwing money at this problem wont solve the problem

#

The corpos running big game studios think they can do it and make it work, but plot twist, it almost never fucking works

#

This wont be any different

naive summit
paper path
#

Not specifically refering to that when i say "throwing money at a problem"

#

Its not just the map that is the problem

#

Like i said there is a series of issues and hurdles that you have to pull to make it even work, let alone fun to play

#

You need a large, good map to play on, a game engine that could pull it off, code to also make it work, a whole gameplay loop to make it fun and a server that will not explode

#

30 million doesnt cover nearly any of these expenses nor time required for it

#

Not to mention that the current version nor the next one of BBR are likely suitable for a 500v500 match at all

#

Hell, the codebase of BBR has to be fragile in Okis words to make the 254 people in the same match possible

regal copper
#

Pretty sure the netcode alone would be a nightmare, and the strain on servers would be immense.. Not to mention the bandwidth that would be required to play.

sand moat
#

Cant even do 256 rn anyway

oblique jackal
#

try to run larger server sizes and watch in awe as the server crashes

#

where is that threshhold? good question

spare kindle
#

not worth it

covert vault
#

I hope them prepare some small events to celebrate game popular again, or just simply drop it safely and make sure servers wont crash. 1 match 1k compare 4 match 256, what is it for 🤷‍♂️

night onyx
oak kiln
#

More possible to have 256 vs 256

#

Since it seems at 120 tick rate 128 v 128 is very feasible, then the server could handle 60 tick 256 v 256

#

For server performance you'd need to half the tick at this point, so maybe even 512 v 512 is theoretical if it was 30 tick rate, with some optimizations for the game itself to interpolate the physics, maybe even disable physics

#

Its definitely possible though with some tweaks and lowering the tick rate. 🤔 Idk why people are saying its not.

#

You have to consider two things, client interpolation, physics calculations, animations etc and even more so the server. Then the bandwidth of movement/state updates per player.

oblique jackal
oak kiln
# oblique jackal My CPU:

In reality your CPU should do just fine if its a newer CPU. Like for example a 9800x3d would find server sided calculations to be trivial. Even 5800x3d, 7800x3d or intel equal. These kinds of things I do believe are more single core performance, but utilizing computation amongst multiple threads with multithreading would provide a huge uplift. Although making things scale like this without bugs is gonna be a challenge in itself 💀

oblique jackal
#

I also agree that my cpu would do better if it is a better than mine cpu.

oak kiln
# oblique jackal I also agree that my cpu would do better if it is a better than mine cpu.

Yeah, I haven't seen the performance measurements since I've never hosted a battle bit high player high tick server, but I assume the calculations with a high end CPU would be feasible with a super large player count.

I don't understand server hosting entirely though, it's probably better to have these new beefy 512 thread cpus or something with terabytes of ram rather than a cpu designed for gaming

#

But as we've summarized, 512v512 is possible just needs some tweaking. Prob not even much, it'd be better to disable almost all if not all physics calculations and keep a steady 30 tick rate. Hit detection and some other things may bug out at times, but it'd be super fun lol

sand moat
#

Mfw can't even do 128v128

paper path
spare kindle
#

some people already have problems with 256

oak kiln
#

Most of them died, but some still thrive because of player base dying though

sand moat
#

They're 127

paper path
#

Thats correct

#

We cant do the extra 2

#

Its 127v127

#

Not 128v128