The Towers Era

Late 2019 <---> Early 2020

In early 2020, I (Idra) played on a Slimefun-oriented server known as CC. This was the first server in a while that I clicked with and really enjoyed playing on. Throughout my time there, I came to know several other members of the community, and we formed a friendship group, which we now call the 'Towers'. Out of the 10 or so initial Towers, only one (Nethereasy) remains on the staff team and active, two (redstonepowder and _Flintlock) are part of the staff team and on hiatus. One tower (__Strykertanks__) still plays on the server occasionally, but has little involvement in any behind-the-scenes work.

Anyway - unfortunately (or fortunately, depending on how you look at it), we didn't last long on CC. As we quickly discovered, the owner was rather power-hungry and abusive. In a disappointing turn of events, one of the Towers - redstonepowder - lost a large number of items due to a plugin being removed with no prior notice. After complaining about this, some items were refunded, but the owner refused to refund the majority of the items. We discussed this in global chat for some time afterwards and I expressed my disapproval of the owner's actions. I was promptly muted.

This is the point where we were set on the path that eventually culminated in MetaMechanists. Within a week, almost all the Towers had quit CC.

We briefly tried playing on a whitelisted vanilla server, but got bored without Slimefun. A few months passed before I had the idea that we could start our own server. With most of the Towers on board, I rented a server from BisectHosting, and we started tinkering.

The Construction Era

December 2020 <---> March 2021

It took us a while to figure out how to run the server properly. Initially, we threw a bunch of plugins on and called it a day, but it quickly became apparent that this just isn't how to build a quality server. We spent quite a while finding new plugins and trying to build a coherent rank system. After copying and pasting some rules from another server, making a banner on a website, creating a Discord server and adding a few channels, we were all set. We posted our server to vote lists and waited for players.

Things were very different then. We used Towny, McMMO, an economy, McMMO-based ranks, a market, and a lot more stuff that we don't have now. The server was also a lot more incoherent - some things didn't work properly, balance was thrown off in many places, and performance issues were ever-present since we barely knew what view distance did at the time. Back then, I did most of the technical work with Nethereasy and _Flintlock, but I ended up taking over most of the maintenance since I was good at it and enjoyed it.

It took a very long time to gain momentum. For a long period of time we had no players online for the most part. The Towers mostly lived in one town, and for a while that was the only place where there was much to see. Eventually many of the Towers burnt out or got bored, and the server withered and was dormant for a while. One day though, __Strykertanks__ invited a large group of friends onto the server. This kick-started growth, and we were suddenly getting players!

The Expansion Era

March 2021 <---> June 2021

Things then shifted into high gear. For the first time it really felt like we were running a community, not just playing by ourselves. I even ended up creating an addon to try and counteract the lag that cargo networks cause, although it eventually broke and was removed.

Having so many new players at once inevitably led to drama though, and we decided it was time to hire a new staff member. We created an Enjin application page and waited for the applications to come in. The first staff member that we hired was neutronstar01, who is still with us today. We decided to hire another staff member three days later as we had another strong application.

Initially, our strongest application was from a player known as Androninja. However, 40 minutes before the deadline, we got an application from Aria_Night which we considered to be stronger, and we chose Aria_Night as our new staff member. Later, we discovered that Androninja was a somewhat toxic individual, and had we hired him, the direction of the server would be very different. Moreover, several lives and relationships would be drastically changed today.

With neutronstar01 and Aria_Night on the team, we continued full steam ahead, learning as we went. After that, we continued to face regular stability and performance issues because we had basically no idea what we were doing. The server was improving every day though, and everything was going relatively well. That is, until we met a player known as Hynlerin.

Hynlerin and myself were messing around one day with fireballs when we had an idea for a minigame, which we called '2 idiots 1 fireball' or more aptly, 'Fireball'. We went to Hynlerin's island home and starting building. Eventually the minigame became a server-sanctioned project and the entire staff team was in on the project.

Unfortunately, this resulted in a lot of drama over control of the minigame. Hynlerin wanted to be the person to run the minigame, and even suggested hosting the minigame on his own infrastructure and bridging our servers together. We convened our first real staff meeting, and over a series of negotatiations, we ended up in a stalemate, with neither side happy. Hynlerin also offered to help fund the server by sending me money. Initially, we were going to accept this offer. I believe that Hynlerin did have good intentions and wanted to help the server, but ultimately, Aria suspected that this wasn't a good idea, and suggested an alternative. That was when she introdcued me to Void.

Void had a spare high-spec server sitting around that had previously hosted a small Minecraft server for a community that Aria was part of. He offered to let us use this server for free. We accepted immediately. To this day, all of our servers and other infrastructure is hosted on Void's machine, and we are hugely grateful to him for this. The server would not be in its current state without him.

Spoiler: it was not '8-12 hours'. Transferring the server was no small ordeal. I had neglected to set world borders when the world was first created, and as a result, the world size had ballooned to over 300 GB. Transferring this amount of data takes a long time, and we were down for 3-4 days doing the transfer. Finally though, the migration was complete and the server came back up.

The Post-Migration Era

June 2021 <---> October 2021

Sometimes after the migration, Aria started building for the server, and before long we hired two new builders, Sky and Kimchi, who are both still with us today. We ran our first event, a halloween-themed set of puzzles that would be open for 3 weeks.

There were plenty more technical challenges along the way, of course.

Due to the ad-hoc way the server was built, some systems just didn't function as intended. Many were broken very subtle ways, like incoherent color schemes, permissions that players shouldn't have, performance issues, and so on. The pressure started mounting to pay off this technical debt, but it was simply too difficult. So many issues were ingrained in the way that the server was set up. Maintenance started becoming a nightmare, and it eventually culminated in burnout.

Around this time, we started considering the possibility of a 1.18 reset. Initially our reasoning was that a reset may be necessary because of the new terrain generation that 1.18 introduced. However, as the issues with the server became more and more apparent, the focus shifted to paying off the massive amount of technical debt that we had accumulated. Even though it became clear eventually that a reset was not necessary, the high maintenance requirement of the server combined with me burning out culminated in a staff meeting, where we laid out a 7-point plan.

• There would be no poll to reset for 1.18 - the reset would happen regardless of community opinion
• Towny was to be replaced with GriefPrevention
• Custom Enchants were to be removed due to compatibility issues
• ChatGames were to be removed as several individuals had memorised all the answers
• Ranks were to be removed
• The server was to move to a whitelist/application model
• The economy was to be removed

Later, some ammendations were made. The whitelist model was determined to be redundant. We decided to keep ranks after coming up with a new ranking system. We also split the server into a Lobby and Survival server, since there was a continual stream of exploits that allowed for items to be transferred into the Lobby server, giving an unfair advantage in minigames and events. Most, if not all, of these decisions are still represented on the server today.

The Reconstruction Era

October 2021 <---> January 2022

After this, I spent several weeks building a new server completely from scratch. Every single config file was meticulously combed several times, everything was checked and double-checked, and redundant plugins were removed or replaced. The intention was to get the technical backend right from the beginning, and then continue in a low-maintenance state after launch. This was completely successful, and it's what has allowed us to keep running this long with minimal maintenance.

However, an unexpected problem got in the way. Even though the server was set up and 1.18 had been released, Paper had still not updated to 1.18. The waiting time was far longer than we expected, and the server was mostly inactive for 2 months, with worryingly low player counts every day. Finally, in early January 2022, paper was released, and we were ready to go - or so we thought.

The new server made use of databases, something I hadn't really dealt with before. The databases had to be cloned and renamed to set up the new server. In theory this should just be a few terminal commands. Easy, right?

Yeah, that didn't go so well. In particular, we lost the LuckPerms database, which held data on hundreds of permission nodes. I was able to reconstruct the databases within a week though, and we finally launched on 15/01/2022. This marked the start of a new era for the server.

The Golden Era

January 2022 <---> September 2022

After launch, we hit an all-time high of 30 concurrent players. This was a huge milestone for us, and it gave me the motivation I needed to continue maintaining and improving the server for a while. The staff team shared this motivation, and for a while the server blossomed.

Over the next seven months, much new content was released. This included a new spawn from the build team, as well as weekly events. BraydenLeungPro and myself even created our own Slimefun addon (which we still use today). The huge progress, though, was in the marginal gains. Many tiny changes over time - quality of life plugins, new colour schemes and graphics, balancing changes, bug fixes - constituted the majority of progress during this time, and it resulted in a significantly higher-quality server with fewer bugs and a far better player experience. Everything felt much more smooth and streamlined.

This doesn't mean that everything went without hiccups. There were still performance issues, crashes, plugin incompatibilities, dumb decisions that created more problems than they solved. In particular, we faced more disagreements within the staff/building team. The period of stability and growth had motivated all of us and brought us closer together. We were more of a tight-knit group, but inevitably this caused us to experience the less savoury sides of each staff and build member, and this made infighting, disagreements, and drama in general somewhat more common.

Thankfully we have been able to keep a handle on the drama (largely thanks to the server's therapist Aria_Night), and it has not torn us apart like it has many other communities. Despite our differences and disagreements we are still able to work together, and as I've learnt, that is what really makes or breaks tight-knit teams and communities.

In early September, I'd finished my first year of A-levels and had just returned to school. It was still looking bright for the future. We'd just experienced a major downturn in players, but chalked it down to a combination of the chat reporting controversy, school starting again, and too little emphasis on voting. We had just hired a new staff member - Justin - who had experience in development and quickly became an admin, working with me to develop new plugins and implement new features. Big things were planned.

Unfortunately, fate had other ideas.

In mid-September, all of our Slimefun data was lost, and as a result, the server was almost unplayable. At least we had backups, right? Nope, those had failed to save properly. Needless to say, the backup system received a major overhaul after this incident.

After a long series of painful decisions and attempts to restore the data, we put out a poll to figure out if it would be viable to just reset entirely. The vote was overwhelmingly in favour of a reset. This marked the start of another era.

The Modpack Era

September 2022 <---> November 2022

While we worked on reconstructing the server for a second time, we created two spin-off servers. One server was for a popular game Factorio, but ultimately lost momentum and died quickly. The other server hosted the "Create: Above And Beyond modpack".

I also discovered a major problem with reconstruction. It wasn't a technical problem. It was the problem of me just not having the motivation to make significant progress. At the time I had just started my second year of A-levels, and had far less time to work on the server. I had lots of external programming to do, including managing a huge technical project. This meant I was already heavily occupied and by the time I got to working on the server, I was tired and stressed from everything else that was going on. That's why it took over a month to get the server back up, even though in theory it should be relatively simple to do.

The modpack server was somewhat popular for a few weeks... until a suspected problem with the modpack caused the server data to be irrecoverably lost (unbelievably, even backups suffered from the same issue). My inexperience with Forge and the lack of people who had the same issue caused me to quickly hit a dead end in debugging. To this day, we have no idea what caused the issue. This whole ordeal caused us to set up a new modpack server hosting Vault Hunters, which ran for a month or so.

In terms of reconstruction, both me and Justin had a sudden spike of motivation. Progress rapidly accelerated, and we decided to compromise on some new features that were ultimately considered too time-consuming to set up. We were thus able to launch the server soon after that.

This is where we are now. It has been an incredible journey running this server from beginning to end. I've met hundreds of new people, and the skills I've learnt will stay with me for the rest of my life. Despite all the ups and downs of the server, we've succeeded, we continue to succeed, and we'll keep on succeeding for the foreseeable future.

The Golden Era 2

January 2022 <---> September 2022

When we launched, we beat our previous record of 30 concurrent players, hitting 45 players, and consistently maintaining peaks of 40 players over about two weeks. However, after this, school and work started for many people, so the server died down a lot. This, combined with some other issues (players were getting burnt out since we'd made progression too easy), meant that we dropped to 10 player peaks, which dropped even further over the coming months due to a lack of updates. Myself and Justin were quite burnt out over this period, and despite having plenty of ideas for next reset, we didn't get much done for several months. Eventually, a combination of events meant we started getting far more done on the server: a new developer (iTwins) joined our team, as well as a new moderator (Sin), and myself and Justin started regaining motivation to work on the server. We've now in the process of developing quite a lot of content for next reset, including some fairly mind-blowing changes to Slimefun, far better progression, and an entire storyline. A lot of this is already developed and hopefully it shouldn't be too long before next reset.