- Our Linode instance, where the Minecraft server lives, costs $60/month.
- Our web and email server instance on Digital Ocean costs $10/month.
- Google Drive storage for backups costs $2.99/month.
Linode is more than twice as expensive as managed Minecraft hosting, but the performance is much better. We have far more tools to control memory usage than we did when we were on managed hosting. You benefit from the performance improvement whether you notice it or not.
Our Patreon takes in $35/month. This is fantastic! Thank you! We have had a few one-time contributions, and they’ve been a surprise and a delight every time.
Please keep in mind that we do not get to keep all $35/month. There are fees. To keep the fees as low as possible, I’ve been withdrawing funds only every three months. I can get away with this, so it’s worth the wait. In February I withdrew $72.56, which gives us about $24.18/month, which covers about 1/3 of our operating expenses.
Public Minecraft Servers Do Not Make Money
Well, OK, maybe some do. I wouldn’t know. As you can see from the numbers above, VoidTree is about $40 in the hole every month it runs.
Where does this money come from? My pocket.
I’m OK with this. Working on VoidTree is wonderful, complicated, occasionally frustrating and positively rewarding experience which has probably put off my becoming senile for at least a couple of years. It’s a fantastic way for a retired IT person to stay in the game. I’ve made many sustaining friendships through VoidTree during some stressful times indeed. It would be hard to get such a rich experience otherwise out of a mere $40/month.
I do very much need to keep the burn rate from creeping upwards, though.
Minecraft is a resource-intensive game–that means it takes a lot of RAM and CPU power to run lag-free. As a Minecraft server becomes more popular and gains players, it need more resources to keep players from lagging. But that’s not the biggest issue.
Every version of Minecraft has used more resources than the previous version. For a while this resource creep has been partially offset by the increasing power and lower prices of new computer hardware.
Such was not the case in Minecraft 1.18. The expansion of the world on the Y axis resulted in a tremendous increase in the resources used. All of the guides for “How to tune your Minecraft server” are now woefully out of date. Some large public Minecraft servers are really struggling.
What are you supposed to do if you suddenly need twice as much resources to run your Minecraft server and contributions aren’t increasing? You avoid having that happen to you is what.
Staff have all played on other Minecraft servers, and some of us keep in touch with friends there. We’ve been watching what other servers do and how that works out for them.
Minecraft servers often monetize by selling items to players for cash. It’s not pay to play if anyone can get the same item for a voting reward if they wait long enough, right? But items in Minecraft can break or get lost. How would you feel if you died and lost an item you just paid $5 for? Yeah, me too. So let’s not go there, hm?
You will never see “loot crates” for sale to support VoidTree. Taking money in exchange for a random chance for a good item is gambling. Gambling has a place, but the people who spend money on loot crates are often the ones who have less money and fewer opportunities for fun–young people and students, and adults who are house-bound for some reason. It feels yucky, and I don’t think it’s sustainable.
Resource usage is not the only problem. More users will mean more support requests, more bugs found and fixed, more potential interpersonal drama and more providing a shoulder to cry on for someone who has died and lost all of their stuff.
I’ve completely shelved some plans to attract more players through alliances with streamers and similar. Maybe some day, but not now.
Likewise, we initially wanted to bring up a creative server alongside survival, but paying the additional $30/month is not in the cards. We’ll know we’re ready to expand when we have enough regular players contributing regularly that we can cover the cost.
How to Help Out
We cheerfully accept support contributions towards running VoidTree, both one time and recurring. But there are many ways players can support us that do not require cash money.
- Play! The server is just a bunch of inert code unless people are on it having fun. You make VoidTree come alive.
- Vote! Voting gives you rewards and helps us reach new players.
- Bring friends! Most of our growth has been via word of mouth.
- Share! Tell people about VoidTree on Twitter. Make an account on Planet Minecraft and post pictures of your builds or discuss VoidTree in the forums. When we announce posts on social media, like them.
When the World Ends
We reset the world when 1.18 came out, and we will do so again when 1.19 arrives. There are a lot of good reasons to do this. It’s our opportunity to leave behind decisions we made that didn’t work out and to reset a few things behind the scenes to work better. A reset will happen approximately four weeks after Minecraft 1.19.0 arrives. It takes that long for all of the plugins we use to update and for us to test our custom code.
We have no information about when 1.19 will arrive. I frequently do web searches looking for clues. The only thing we know so far is “some time in 2022.” I’ve been hoping that this wouldn’t happen until December, but I’m seeing rumors that we’re looking at June. No official word from Microsoft yet, so please don’t quote me on any of this.
We know that saying goodbye to all you built is sad. I’d like to offer you at least one thing to look forward to.
For VoidTree Season 2, we followed standard advice to pregenerate our world to improve load performance when a player enters a chunk for the first time. Pregeneration has downsides. A pregenerated world must have boundaries that you cannot cross. It also takes up a lot of disk space to store areas that have never been visited, and we have to manage ours carefully to avoid out of disk space crashes. At the same time we’ve noticed that our premium hosting on Linode may be good enough that nobody will notice load lag on a new chunk.
Resetting VoidTree would allow us to choose a non-pregenerated, infinite world and might present more flexible options about how to respond to future version changes.
Beyond these esoteric technical issues, a reset tends to be fun. People who wandered off come back for the excitement of something new.
Closed vs. Open Beta
Our initial plan was to get our custom code stable, gain experience, drop the allowlist and declare Open Beta. Some things have changed.
We will not be dropping the allowlist. Sure, getting spontaneous connections from new players is great, but some new players cause new problems. VoidTree has become a warm, friendly, grief-free hangout, and nobody wants to see that change.
In December 2021, a very serious exploit threatened Minecraft servers and, indeed, the entire internet. The main method of combating exploits is to keep everything up to date with the latest versions, but preventing expoiters from joining our server is also important.
The upshot is that without a plan to drop the allowlist, we’re no longer sure what Open Beta will mean for our server. Maybe it doesn’t matter. Maybe we’ll declare Open Beta when Solly clears his bug list and feels comfortable with the stability of our custom code base.