COMMUNITY - FORUMS - GENERAL DISCUSSION
I don't think the servers are going to be good enough

the game is not going to be able to handle 1000 plus players on the battlefield

they should have stayed with improbable because of the redundancy system they had with their server Network would have been useful to them but no they rather build that from scratch which means less money going into the game and more money going to hardware.

on EVE Online the largest battle I have ever participated in was 5600 + players in one solar system but took 14 hours don't expect a battle even at 25% of that being possible in Chronicles of Elyria without massive failure to be able to handle something like that you probably will have to spend 13mill on Hardware which you don't have

improbable also has a deal with Amazon and Google with their cloud services you don't have to pay for them till you start making money on the game itself so you would have unlimited time to test

now you have to pay for the maintenance and replacement of parts for your own Hardware it would have been cheaper to go through them cuz if there was any problems less than 15 minutes they would have updated the server Network and gave you a new server to use.

the redundancy that technology offered was the reason why you guys should have stuck with it

i expect to be disliked off this planet


2/8/2018 6:59:31 PM #31

We don't even know what Improbable can accomplish. Honestly, I'm not even worried about their servers...I'm worried about the average player's ability to even walk into a city. If they do accomplish being able to sustain the amount of interactions in a capital of 250+ people who all have the capability of interacting with individual potatoes on the ground...how the HECK is the average non-fiber internet going to be able to take in that much information going on around them without a REALLY elaborate priority system.

Also, Improbable dropped Java support...which is the entire code language they are using. So it's improbable's fault. They had a test build and seemed to be moving to support Java and they dropped it, which forced CoE to have to make their own.

I don't think CoE wanted to do this, but had to make do with what they could.

2/8/2018 7:30:50 PM #32

Posted By radkid at

they should have stayed with improbable because of the redundancy system they had with their server Network would have been useful to them but no they rather build that from scratch which means less money going into the game and more money going to hardware.

Chronicles of Elyria without massive failure to be able to handle something like that you probably will have to spend 13mill on Hardware which you don't have

improbable also has a deal with Amazon and Google with their cloud services you don't have to pay for them till you start making money on the game itself so you would have unlimited time to test

now you have to pay for the maintenance and replacement of parts for your own Hardware it would have been cheaper to go through them cuz if there was any problems less than 15 minutes they would have updated the server Network and gave you a new server to use.

the redundancy that technology offered was the reason why you guys should have stuck with it

I don't know where you got the idea that the fact they are not using Spatial OS anymore means that they have to buy and create their own hardware network? That's why cloud services have been created for. Amazon, Google, Azure are there exactly to the purpose of not needing to buy hardware. They have the hardware, they maintain it and you only have to pay for its use. The same way they had to pay to improbable. Today it is way more cheaper to use cloud services than own your own hardware. Not only because it saves yourself the need to buy hardware and maintained it. But these systems already have tools in place that allows you to add instances on demand the moment you need it making it ideal for growing your computing power as you need it (ie. in the area you have a battle). So it is more likely that they use a cloud service than buying their own hardware.

the redundancy that technology offered was the reason why you guys should have stuck with it

All this is provided by Google Cloud (or any other cloud service) not by Improbable. So they still have this, without using Spatial OS.

improbable also has a deal with Amazon and Google with their cloud services you don't have to pay for them till you start making money on the game itself so you would have unlimited time to test

First, the deal is only with Google and not with Amazon. Second, how do you know that CoE applies to that deal? Do you work for either SBS, Improbable or Google to assure that CoE applies to it?

Making assumptions are fine, but if you are going to make statements just make sure you have the right facts.


2/9/2018 9:28:35 AM #33

Posted By Rhaegys at 8:30 PM - Thu Feb 08 2018

(...)how do you know that CoE applies to that deal? (...)make sure you have the right facts(...)

Please do not misunderstand me, I disagree with the original poster on a fundamental level, but that one point is actually the only fact in his posting that he got right ^^ Would Improbable count as a reliable source here? '(...)members of the SpatialOS Games Innovation Program will receive subsidies that will substantially or completely cover the costs of cloud-based online development. The program awards SpatialOS credits that will cover SpatialOS and Google Cloud usage costs, including underlying server fees, during development(...) Right now, five studios have been accepted onto the program (...)Chronicles of Elyria by Soulbound Studios(...)'

Since most of the points in the original posting have already been refuted (namely large scale battles being difficult for the GPU of the local machine and the network protocol, not the server; hardware maintenance not being an issue with cloud based services regardless; redundancy not being actually part of SpatialOS), I'm only going to point out that while you do not have to pay to make your voice count, not having contributed at a minimal level after two years shows that he never had any faith in the project to begin with, hence counting for the credibility of a 'sudden change' in trust.


Sage willing to help with Purity (if you spot me on Discord and have some Plague on your account that could be nullified with a trade, drop me a message on Discord)

2/9/2018 10:12:14 AM #34

Nowhere in any of the places I have read has there ever been said that this is a battlefield game, so why would you think that there would be everyone in a battle? It is more of a life in Elyria simulation than a giant battle. So, unless you are saying that every waking moment, you step out of your front door and start fighting, then you have a bigger problem than if the servers will handle playing this game. This is not a FPS, but a MEOW (I love that term!)


Countess of Tarnham

County Tarnham, Rhynelands Duchy, Vornair. Luna Server (NA-E)

2/9/2018 12:47:04 PM #35

Posted By Scheneighnay at 02:34 AM - Fri Feb 09 2018

However, I feel like the servers won't be as big an issue as end computers will be.

UE4 is extremely graphically intensive, even when rigged down to its lowest levels. I feel like getting a large amount of players together for a battle will will reduce the game to a slideshow and even start causing crashes before the server goes down.

So the question lies in how battles can be broken down to avoid the massive strain a large battle will put on PCs and server hardware.

This is the more important question. The server > client traffic will probably be pretty high compared to other games if they can pull off the backend (unless they do some magic with the rendering mechanics as well), but ultimately how feasible larger battles will be probably comes down to how well a computer can handle the load. One of the things that I most look forward to for curiosity's sake is the alpha/beta testing of battlefield mechanics.

It'll give me an idea of how graphics-intensive the game is going to be, and while I'm probably going to need to do some hardware upgrades anyway the improvement from alpha to beta should give me an idea of how high peoples' specs are generally going to need to be to run the game - something that will help me work out how many of my friends I'm going to be able to convince to play the game with me once we get to Exposition.


To touch Divinity, one must be prepared to brave Reality.

2/10/2018 6:39:50 AM #36

I'll second that opinion, Wolfguarde.


2/10/2018 12:13:22 PM #37

As far as I could understand SpatialOS is optimized for simulating a large game world, by splitting it into independent areas on the fly. But I haven't seen anybody claim that SpatialOS is optimized for simulating a huge battle with lots of players in the same area.


2/10/2018 5:08:48 PM #38

Posted By Halvgrim at 06:13 AM - Sat Feb 10 2018

As far as I could understand SpatialOS is optimized for simulating a large game world, by splitting it into independent areas on the fly. But I haven't seen anybody claim that SpatialOS is optimized for simulating a huge battle with lots of players in the same area.

That is exactly what it does. It tracks actors and their actions and handles the load balancing on the servers. It doesn't matter what the actors are doing.


2/10/2018 8:04:29 PM #39

the game is not going to be able to handle 1000 plus players on the battlefield

We've never said that it would. In addition to which, a distributed OS like SpatialOS doesn't solve that problem. Whether we used SpatialOS or our own spatial partitioning system it's important to understand the differences between horizontal and vertical scale-out.

Horizontal scale-out is the ability to support a bigger world. With a bigger world you can also support more players. So long as they're spread out. As soon as players start to cluster in one location you're talking about vertical scale-out, which is a bigger problem to solve as you're fighting physics.

As players gather in the world, the load on the individual servers as well as your clients at home increases. We can procedurally sub-divide the server around that location in order to lessen the load, but at some point, there's so many processes running on individual servers that the network traffic and latency actually starts causing the simulation to perform more poorly. That's why it's easier to get 100,000 players in a single world than 1,000 players in a single neighborhood.

And - even if we could subdivide the servers enough to solve the back-end problem - could your PC process the client-side load of updating and rendering 1,000 players on-screen?

In the end, there's a lot of work to do to optimize CoE, but the hardest optimization problems aren't solved through horizontal scale-out or even sub-dividing the world into smaller and smaller chunks. It's solved through understanding of the game mechanics and experience optimizing client engine code like UE4 - something we have experience with.

they should have stayed with improbable because of the redundancy system they had with their server Network would have been useful to them but no they rather build that from scratch which means less money going into the game and more money going to hardware.

We don't pay for hardware. Like most online games, we lease our hardware from server farms. They also have guaranteed up-times with Service Level Agreements (SLAs). They also have redundancy in their hardware to ensure we don't lose data, etc.

improbable also has a deal with Amazon and Google with their cloud services you don't have to pay for them till you start making money on the game itself so you would have unlimited time to test

I could be wrong, but I believe improbable only has a deal with Google, not Amazon. As well, I recommend you read the details of the program more carefully. It could be that you missed a few things.

now you have to pay for the maintenance and replacement of parts for your own Hardware

We don't, actually.

i expect to be disliked off this planet

I'm not going to dislike your post. But I do wonder what benefit there is to coming to a game's website and posting your negative feedback about a decision that's already been made. While people may commiserate their concern, all that really does is stir up concern again for an issue that's already been laid to bed.

You're clearly not going to change our minds with your post, so I don't entirely understand what the purpose of the post was. And if you suspect to be down-voted going in? Why waste the time on the post in the first place? A downvote means people feel you're not contributing to the community.

In any case, I'm closing this thread. Your questions/concerns have been addressed, and there's no reason for people to further come to our defense. We made a decision we're happy with, and continue to feel it was the best long-term decision for us and for the players.