Launch Week, Our Staff Team, Server Performance!
Diamond is the culmination of months of work; while motivation was sometimes hard to find due to dwindling player activity on our existing servers, we knew that Diamond would be our biggest hit to date. We took everything we've learned over the past three years of hosting Cobblemon Servers and threw it together to make something great. We've planned months ahead for events, feature updates, and scaling options.
Communication and delegation in our staff have been an issue for Roanoke over the past year. Staff being split between Gold & Dusk only made this worse. Diamond would have crashed and burned post-launch without the team working together and delegating tasks incredibly, allowing us to not only patch bugs, but to work on new content at the same time.
Diamond's launch went very well - we broke our concurrent player record, with 138~ players online, and sentiment has been overwhelmingly positive from the community. Cobblemon 1.6 itself being brand new, alongside it's Minecraft version of 1.21.1, meant there were a few teething issues. Lag being the main one. As much as we can test individual features and sidemods, it's unfortunately impossible for us to test performance & lag outside of live servers. Patching performance, tweaking settings, and switching mods to help combat the lag (and crashes it can cause) has taken up most of my time this week.
While the current state of the server is significantly better than at launch, the current level of lag & crashes are still unacceptable to us. We'll go into our concrete plan to solve this issue forever shortly.
Features, Quests, Tournaments, Events!
Over the week, WonderTrade has come and gone due to bugs, but is now live & stable - thanks in part to the WonderTrade Technician that HQ sent down. The WonderTrade Technician itself, and the mini "quest line" it brought, is something that we would not have previously been able to do. It's a small teaser of things to come, with larger "quest lines" coming this weekend alongside daily & weekly tasks. Our plan is to regularly release new quest lines to provide fresh content when Minecraft & Cobblemon get stale. They'll also solve the issue of distributing rare rewards such as Bottlecaps & Legendaries, which were previously only consistently available through competitive tournaments.
On top of regular quest updates, we'll be launching our tournament racket this weekend on the 11th of September with Diamond's first singles tournament. From then on, there'll be weekly tournaments on both Friday's & Saturday's, rotating between singles & doubles. Thanks to merging the Gold & Dusk competitive staff, we have a bigger list of people who can host tournaments, so we'll be able to keep this schedule without any individual staff burnout. Hosting two tournaments a week will allow more players to participate where scheduling would previously stop them.
Our new schedule gives Diamond new content every single week. Even better, thanks to staff restructuring and training, we'll be able to produce that new content without it taking up the entire teams time. That allows us to work on Major Updates & Events.
17th January: Mega Pokemon Release
It's been three years without the best Pokemon gimmick, and that ends on the 17th! The technical challenges are out of the way, and we're currently working on balancing acquiring mega stones.
24th January: Cross-Server Solution for Lag & Scaling
We'll be implementing a technical solution that will stop lag, allow us to re-introduce MC features like mob spawning, allow unlimited amounts of players to join, and let us give players more choice on the environment they'd like to play. Read on for more info!
29th January: Chinese New Year Event
Cobblemon 1.6's launch forced us to push back our planned Christmas/Winter Event - we've been adapting the planned mini-games, textures, and builds to fit in with Chinese New Year.
February Onwards
We won't be releasing concrete dates at this point, but you can expect the Biodome/Safari Zone to release, alongside a Valentines Day mini-event.
Solving lag, our plan to grow Diamond infinitely!
When I hop on online and ask how Diamond is going, the first response is always about server performance. Lag, crashes, low Pokemon counts, no natural mob spawning - these are all related to performance. Minecraft simply isn't designed to have a player count higher than 80~, especially on modded platforms. While we can squeeze out extra performance from a single server, as you're all painfully aware - it has a hard ceiling. Not only does it have a hard ceiling, of around 130~ players, to even get to that point requires sacrificing game-play quality.
That leaves us in a tricky situation, simply having less players isn't an option - we want Diamond to thrive from this week onwards, not slowly dwindle down to player counts seen on Gold & Dusk over the past few months. Adding a new server, calling it Pearl, leaves us in the same situation we had with Gold & Dusk. Maintaining distinct servers & playerbases is hard. It impacts the quality of both servers, because our time has to be split between them.
The solution? We add more servers, but hide it behind technical trickery. Instead of the Spawn World, Overworld, End, Nether, all being on one server, they could be hosted on different servers with player inventories, Pokemon, chat, and tab being sync'd. From the player perspective - there would be no difference between simply loading into a new dimension or loading into a new server. Instead of being capped at 130~ players across all words, we could have 130~ players on each dimension.
Of course, this is just a simple explanation of the concept. Just splitting up the dimensions doesn't help much, as the majority of time spent is in the overworld. We would add multiple overworld dimensions, and push new players towards fresh worlds with the least amount of players online using RTP. Similar to a Pokemon Map when choosing where to fly, you could choose which dimension to teleport to (randomly or not). We can effectively add unlimited dimensions, letting us have infinite amounts of players online.
Less players per server means less lag, and also lets us re-enable (and buff!) features like mob spawning, Pokemon per player. Less lag means less crashes, and when they do occur, it'll only affect that individual server. If a random bug occurs at the moment, the entirety of Diamond goes down - everyone is kicked off. With the new system, only that individual world would go down, and the players would just get kicked into spawn while the server restarts.
Tournaments can be hosted on their very own world, letting us have large amounts of participants without lag from other players exploring the world causing issues. Upcoming projects like The Biodome/Safari Zone can be hosted on their own world, too, letting us go crazy on the amount of Pokemon around.
We can give players more choice on the type of world they'd like to live in - we can have vanilla overworlds, Terralith ones, peaceful worlds, and potentially even Skyblock worlds! All under one server, one set of Pokemon, one inventory, one chat.
That sounds too good to be true, what's the catch?
It's a lot of work to setup a system like this. Luckily, Roanoke is in a unique position to pull this off. Almost all of our sidemods & features are developed in-house, and a number of them already have support for a "cross-server" solution like this baked in.
This system will require more maintenance than running a single server, on the technical side, but to a manageable extent.
The main issue is, as always, Money! Roanoke currently rents a dedicated server in Germany, on which we host Diamond, Green, and a large amount of development servers, both for technical work (testing sidemods), and for building. Between these servers, our box is full, and we don't have room to adequately test & release this cross-server solution for Diamond.
We will need a new dedicated server alongside this one over the next 2~ months for testing & production.
Monetisation, Money, P2W
Roanoke is expensive. For the past year our costs have averaged out at around $400/month. Whether we have 0 or 200 players online, Roanoke costs that much a month. That's great when we have closer to 200 online, because we're making that money - not so much when we have 0 online.
Roanoke has maintained a strict NOP2W store since it's launch. It was one of the founding principles I created it under. The reality is that there's a limit to how much money we've been able to bring in consistently under this model. Without more funding, we can't add new features, maintain existing ones, or grow. We've stagnated at our current level for the past year, partly due to this reason.
As you'll have noticed, Roanoke Diamond launched with a F2P (Free-To-Play) style store. We sell what we'd previously consider advantages - i.e. access to /pokeheal, tiered ranks, extra sethomes/claim blocks. All benefits that we sell can be earned in-game via our Pokedex Rank system. We didn't make this change lightly, it was discussed for months and only released with a unanimous vote from our management team (Myself (Comp), Novoro, Ivy, Squillis, Mango).
For Roanoke to grow into the most popular Cobblemon Server, and hopefully eventually the most popular server full stop, we need more money, and we need it consistently. We have pushed monetisation hard on Diamond so far, with frequent announcements, coupons, and new cosmetics. I don't love to do that, but it is necessary.
In the short term, we need to raise enough money to pay for 3~ months of expenses, including the new dedicated server(s) to enable our cross-server solution and make Diamond the best it can be. The more money we raise, the faster we can bring out the solution & new features on top of it.
Like with houses in real life, location is everything. Our current server is in Germany, because to host it in the US would double our costs for the same hardware. The majority of our players are in the US, meaning the majority of our players have bad ping, thus a worse game-play experience.
I would like to move Roanoke's hosting back to the US. Preferably, in Ashburn Virginia, which is ironically very close to the real life Roanoke. Hosting on the East Coast means EU players (such as myself) can still have a quality connection with reasonable ping (120-200ms).
We're looking at roughly $300~ per dedicated server, per month. With one dedicated server, which would be fine to start with, that brings our monthly costs to around $500-600/month. In the first month or two, our cost would really be $800~ as we'd be paying for two servers at once to avoid downtime.
Diamond's launch has been great for us financially, and while we could just about afford to make this move right now, we do not want to overstretch - leaving myself personally liable for costs without money (guaranteed to be) coming in.
If you can afford to give us money, that's great! If not, that's completely fine - and there are other ways to help. Tell your friends about Roanoke, link people to it on Reddit when they ask about cool servers, even just playing helps us grow. The more we grow, the more money we make, which means we can grow more, which means we make more money... and so and so on, til Hypixel is forgotten and Roanoke is the top server.
Comments