I loved the brief period where global chat was a thing, and the wisps floating around the plaza from other plaza servers was awesome.
It would be even more awesome to see greater player counts on plaza servers, it would make them feel so much more alive. Obviously server performance is a concern here, but I think I speak for most of us when I say that even if there can only be one or two plaza servers per region but with increased max players, it would be worth it.
They’ve tried increasing it before. It went really bad. Performance drops considerably. I remember I was getting like 10 fps when they attempted 128 players. The reason they settled on 64 is because it’s the biggest they could get without completely destroying everyone’s FPS.
Dang, that sucks. Hopefully eventually some optimization or distance culling could help to bump that number up even a little. It’s just a shame because the plaza can feel so empty sometimes even when it’s full. At the very least I think we need global chat back
The network traffic of that many players is the real issue here. The more players, the more the server has to network and then also the more players the more games that are actively going on comparatively.
We do plan to separate the Casino into its own set of sub-servers which should help make the Plazas feel a little more busy and also the Casino more busy as well.
I feel like this would be an unnecessary separation as well as making the Casino too crowded altogether, since some days on more populated servers almost everyone is in the Casino. Perhaps what could be done instead is maybe a “Global” Casino akin to the Global Gameworld Ports. You could make an elevator in the center between the slot machines that can bring players “upstairs” to the Global Casino.
That’s what this plan is. Each Plaza would connect to the same Casino server (until it’s full and then they’d go to the overflow Casino). But they wouldn’t be able to access the Casino server without joining Plaza first.
On discord a response I got was that it’s still on the table on low priority because tests on if it would actually improve performance significantly enough to be worth the nightmare of implementing it were inconclusive.