December 22nd, 9pm EST. Anarchy Day will start the 21st, 9pm est and will continue as close to the reset as possible.

We've learned that a year-long map is the ideal length to let people accomplish things, but still give a regular breath of fresh air bringing back some of the most long-time community members. This experience comes from many servers in the past- after all this community goes back many years before moving to Vanilla Extract and we have a good feel for how to do healthy resets.

The server was started months earlier than we originally planned to "strike while the iron was hot". Alcohol in voice chat might have been involved. That ended up being the right choice, but we want a December reset for multiple reasons. The biggest one, frankly, is that the guy that is responsible for a lot of the leg work (Syntax) takes most of the month off from work every year and August was a terrible time to make it all happen during the launch. There's still some benefits of December- nearby holiday time off and college students often get a long break as well.

We've experienced resets at many different times on previous servers, and the best experiences were always the end of December. Getting on a December schedule meant either making a very short map or a really long one, so we put it up to a community vote and the vote came out overwhelmingly for a long map- though strangely votes on that post much later started to skew the other way. Even with the later votes that came past when it was decided, it’s still 3:1.

We've always had this info publically available, and it's in the main FAQ on the site. New players starting now will be accepted with a message warning them about the reset date. We know some people don't read everything, but we hope that your experience meeting all of the cool people who will come back around and being there from the start to build everything will be worth it. We find that most people against it change their mind when they see first hand how it really breathes life into the community.

In most places people here played before, owners were scared to announce when the reset was until the last minute to save player count, but every long-time player knew around when it would be. We don't like that approach. We believe that knowing how long a map will last is your right as a player. Knowing how much time you have to invest into projects is important in our busy lives. Our plan is to continue on December resets, probably always later in the month, though the exact date floats a bit based on stuff like holidays and schedule of the server monkey.

If for any unlikely reason we have to change from December, we favor extending a map unexpectedly to shortening it. We will never reset based on a mid-map Minecraft update, we will just expand the world if we need new generation. That said, it is strongly favored to never extend either as a map lasting longer than one year makes for a tough time keeping numbers up! This 16 month map was a nail-biter as a first map, but we did so much better than expected and we continued to get a trickle of new faces. We would only ever do a long map again in the case of an emergency.

You would be surprised how much riff-raff gets accepted and quickly banned or otherwise. A whitelist doesn't fully protect us, just reduces the headache greatly. Even a good meaning player doesn't immediately know how we play this server sometimes. We've got it down to a science, but we still find and fix griefing and theft a lot more behind the scenes than most people know. We want to allow a time for everyone to button up their builds (as well as public places like spawn) without fear of the riff-raff breaking something too late to be caught and rolled back before the final save.

This will only apply to people who are accepted during this safe-time, not just anyone with the new tag. If you're already playing this doesn't apply to you. They will be able to participate in Anarchy Day and the new map immediately. Exact timing for this hasn't been decided yet.

Some of the rewards are...a bit unfair the first day. Imagine this- you get 5 rolls and get some rocks or some shit, and another player starts with a maxed pic. It may not bother you, but we've noticed that some players really enjoy running out naked and being the first to speedrun villagers, or an iron farm or some other goal. We like the competitive fun of the first day and we want to give players the experience of a true naked start- after all we won't have kits yet because we need to build spawn!

The new voting system is built in such a way that it is easy to take part of it down while still counting your votes. Your votes will still be counted and you will get a message telling you that your votes were received. After 24 hours you'll receive any pending vote rewards you racked up. We're also handing out the Extractmas rewards at that time.

Yes. The Dynmap will be up and fully rendered at the same time as the server to help you find a place to live.

The Overworld will start at the same size the current map started at- 20k each direction, 40kx40k total. Like the current map, it will expand for any mid-map Minecraft updates that require new generation. The Nether and End will not have a world border.

Possibly. Some steps are taken to make the experience better- including temporary harsher-than-normal settings to clamp down on lag and pre-generating the entire world to make chunk loading when a ton of people explore more bearable. In places past we've seen smooth reset days and we've seen 4TPS days with chunk loading so slow you might as well be sending the chunks over the postal service. The main factor will be how many players we get- we don't know! This is Vanilla Extract's first reset and if our opening was anything to go by, there will be a lot of players. One mitigating factor after the initial rush is no one has had time to make farms and other sources of lag.

Honestly, even if it is super laggy the experience is still fun- especially if you join us in voice. Just the event of being there with everyone and trying to go find a place to live is worth it if you can make the time work. A lot of people will be there just to be a part of the fanfare then come back later on their more normal hours. Others will grind like maniacs for several days.

There's a channel on Discord called Map 2 Forum. There you'll find threads for each project, and the requirements at the top of these threads (scroll up). Voting on these designs will take place shortly after reset to give time for returning players to make a submission, but will happen much sooner than Map 1.

This community has a tradition of an anything-goes blow up stuff and kill each other event at the end of a map. Don't worry, the map is saved before this. It's a lot of fun, and it's quite cathartic to blow up that build that caused you grief and didn't come out the way you wanted. It's kind of a chaotic way to say goodbye to a map. Info and rules will be posted shortly before, so keep an eye out.

"Rules? On anarchy?"

We're still a community, and this event is supposed to be a combat and destruction day, not channel your inner 4chan day. We still need to follow common decency toward each other and remember it's all in good fun. But feel free to play wild, steal, blow stuff up, kill other players, and even make a lag machine if you want. DO NOT START BEFORE THE ANNOUNCED TIME FOR ANARCHY DAY WHEN IT COMES. This means don't bury a layer of TNT anywhere in anticipation of the destruction (even your own base) and don't steal supplies just because it's the end of the map and "who cares". We will strictly enforce this. There will be a grace period after map saves to go gather supplies and get in position since players might want to go to spawn for the end of the map to celebrate.

Last year we did "The 12 Days of Extractmas", a joke about car dealership-invented holidays like "Rocktober". This was a vote drive where you received tickets for each vote in addition to your normal rewards and you could trade these tickets for cool rewards at spawn. This was a test run of what was always our plan for December resets- giving tickets to let players choose rewards for next map.

A reset brings a lot of people back, and this gives a huge boost to a server on voting sites. The problem is that voting sites reset the vote count on the 1st of every month, so a reset on the 22nd doesn't allow us to take full advantage of the rush of people coming back. An event like this helps us make some of that ground up. A good reset should not just bring people back, it should snowball into getting a bunch of new faces too.

This year Extractmas 2 will start December 1st and go until the end of the map. Just as before you will receive tickets with each vote and you can save them up and turn them into the appropriate chests at spawn for rewards on the next map. These will be given out 24 hours after opening. These rewards won't be as baller as last time and be more tuned to the early game. They'd be useless to any established player on this current map. For example, if we have a maxed pickaxe or something like that, it will be expensive. An example of a cheaper reward might be a shulker box. We'll announce more details and a reward list December 1st.

The voting system was made many years ago by some long time friends and one of the owners of this server for another server. It is built on a Paper plugin that is really not suited for how we use it. It is extremely hard to edit our complex setup, and things like weights and quantities are difficult to make happen- long story short, a lot of copied entries. This led to the situation where Syntax was needed to update it at all by hand, it wasn’t easily readable by the casual observer, and it was not updated enough. We're missing some rewards for new versions and custom items are difficult. In fact, the voting rewards were finished minutes before the first map opened in a desperate panic.

The first step of dealing with this problem was replacing the plugin entirely. The plugin is a good plugin, we just use it outside what it was designed for. Also, because we are moving to Fabric, we couldn't bring forward the old plugin. The system is designed mostly as a datapack with some stuff on top and should be easy to port to any server platform and would even work with a vanilla server jar. There's custom logic to roll the category of reward (regular, lucky, extra lucky, collectible). Each category has a minecraft loot table determining the rewards. We built a tool to one-click convert these voting loot tables to spreadsheets. This way anyone on staff can work on a Google sheet (complete with history and comments) and all Syntax has to do to update it is to export it, drop the new file in, and reload the datapack which can be safely done live.

As for the rewards, the current system has been remade as close to 1:1 as possible to give us a starting point. For example, if there were 2 entries in the old system for 16 oak logs and 32 oak logs, the new loot table entry would be a double weight with 16-32 random quantities. We know no one wants to get odd numbered concrete stacks in their scrap colored blocks chest- but as the rewards are re-tuned this will be adjusted.

Converting the old rewards settings also helps see how well the system is functioning as intended compared to the previous one. Because it's a datapack mostly, it is able to be used on our current Paper server too! We will be applying it before the update, and if all goes well rewards will start to be re-tuned live before the new map starts.

Finally, it's just more flexible. This new system was handmade by us, for us and is more modular, ready to accept modifications for things like voting events and Holidays. We also have a lot more control with how the random chance works, things like distribution of rolls inside the table, and more. So much more. We can even make conditional entries. We're all really excited to start making voting better for everyone and we're just waiting on Syntax to get off his ass and put it on the main server to get started.

As for the experience using it, all of the functionality of the plugin has been replicated. There will still be weekly, monthly, and all time leaderboards as well as milestone rewards. The /vote command will tell you the time before each vote and provide links just like before. The chat spam is the same, however if you have more than 5 votes queued when you join we've eliminated the wall of votes in favor of just printing how many times a player voted a total of each reward. Please note when it's first rolled out you will only be able to see your own vote counts and may not be able to access a leaderboard and the GUI will be in chat. This is temporary, as while it is possible to do an inventory GUI in a datapack, it's very jank so a separate GUI to plug into the stats will be made for Fabric. We will still have the data for leaders. We find making it outside the datapack is both easier and acceptable because the core functionality still works if for some emergency we have to migrate away from Fabric in the future for some stupid reason.

The map will be saved a few hours before Anarchy day, allowing some grace time for everyone to prepare before starting the fun. The raw download will be provided immediately. It is going to be in the neighborhood of 130GB zipped, 180GB unzipped. This is because the overworld was entirely pre-generated for performance, in addition to the nether and end being generated up to 5k and 10k respectively. There will also be a cut-down download that will remove any chunks players didn't build in. We don't know what size that will be but expect the savings to be significant. The cut down download will allow you to see and save any build you want, but for truly exploring the map it'll be a bit messy. Part of the overworld was generated on 1.19.2, part of it 1.20, and bits of the nether and end in various other versions, as well as some mid update Paper bug fixes that affected generation. Chunks players didn't build in won't generate on your local copy exactly how they were in the map in this version, and there may be seams.

For the spectatable version, unfortunately that will come later. Syntax needs to set up some infrastructure first, this includes a new backup system and a new server just for this. Don't ask why it’s all linked, it's a bit dumb but that's how it is. We will start by hosting the 2 versions on Google Drive before later self-hosting them and the live copy along with a Dynmap and individual region file downloads. All the custom stuff will allow us to provide all of this kinda stuff in many versions indefinitely much more cheaply than standard cloud storage. Have you seen storage prices? These changes will also allow us to take better, more frequent offsite backups and spend less than we currently do.

The easiest way to get a build you need will be joining and using Litematica, but for those that are not familiar with that mod we will have a guide on how to take advantage of downloading individual region files and placing them in your local world to just get your base when the additional download options are released. Remember, even though this will eventually be full of awesome options, we are starting with just the big downloads, so if you need something like redstone from your builds make sure to take a copy while Map 1 is still active if you don't have the space or internet connection to download a huge map and extract it. If you don't use Litematica, ask someone on discord to take a schematic for you and place it on a superflat or something.

Dynmap will stay and have at least as many views as it does now and have the web interface replaced with a customized one. The resolution of some of these views will increase because we now have more disk space on our server to work with. The Dynmap is often the same file size as the map itself! In addition, we will be adding a second live map that is less simple to use but lets you fly around as if you were in the game.

Paper (a Spigot fork (a Bukkit fork)) is excellent and made my truly skilled people, but the purpose of Paper isn't really compatible with the type of server we want to run. Paper's goal is to retain the general minecraft experience while fixing things that Mojang won't and making the game more customizable and more palatable for large servers. The general experience is the point- all of the content is there, but it works a bit differently under the hood. Think of it this way- Fabric is true vanilla with some things on top if you want to add them, and Paper is more of a re-coding of a lot of the server-side stuff out of the gate. For most servers, this is excellent and they see large performance gains and have access to a huge library of Bukkit, Spigot, and Paper plugins that are often forward-compatible.

For a server that wants to maintain the true vanilla experience, this causes a lot of problems and doesn't provide a lot of advantages. A lot of things that do save a ton of performance have to be disabled or even jankily worked around to make sure things like complex redstone work as close as possible to vanilla. We often run into the case where a player tries to build redstone they saw on YouTube for it to just not work without modification. This can be frustrating to someone that’s not both redstone experienced and experienced with the differences in Paper.

One particular thorn in Syntax's side was a YouTuber called "ianxofour", who often makes farms with needless alternate versions of stuff, like a no-slime zero tick tnt duper on a tree farm, simply because they are cool and he likes to do something different just to be different. Ianxofour is good about noting the differences in Paper in his descriptions, but his farms often become the meta and people just make hundreds of youtube farm tutorials that are copies or partial copies of his design. Suddenly that tree farm doesn't work how it's made on YouTube, and it's defeating for the builder. We want players to see something, build it, and it to work. A lot of time was spent helping players fix their farms to make them work on our server. That is an experience you will have on most servers and really most are much more brutal than ours after the careful changes we've made, but it reflects on us. Players don't think This build doesn't work on Paper, they think This build doesn't work on Vanilla Extract.

The particular issue with zero tick redstone was able to be fixed by disabling an alternate redstone algorithm that gave a lot of performance for the awful flashing dust and is frankly a change that Mojang should make. It retained all behavior except some update order stuff. If this change were in vanilla, it would cause issues for existing farms, but we'd all be on the same page and that would be talked about by everyone, and it would probably be received positively. This example, while fixed inside Paper by disabling it, is the core of the problem- even if they are good changes, the problem is they are changes at all.

Another great example of the performance tools not helping is Paper lets you customize the tick frequency of individual behaviors of certain mobs like villagers. This can save a ton of performance and works with most iron farms and trading halls, but it breaks almost all modern raid farms. It was found out that even if you told it to tick every behavior every tick, it still caused issues because simply having Paper control the tickrate replaced some of the villager code with Paper's code. It's not like Mojank made a bug free game by a longshot, but we want our bugs to be fresh, free-range bugs from the source.

These problems don't just present themselves in redstone. A good example is there was an awesome huge dolphin highway that some folks made which suddenly no longer worked on Paper after updating to 1.20. Testing revealed it was Paper itself and not a setting, and it still worked in Vanilla. There are tons of little issues that are in the backlog for the Paper team and more gets added all the time. There's an issue with bees killing themselves by suffocation even more than they already do in Vanilla, and I have commented on that issue years ago. It still remains unresolved. Anyone who had tried to make an overworld wither cage obsidian farm before the end platform worked for obsidian farms can tell you withers don't behave the same. It's just death by a million cuts and there are countless examples.

There’s also little quality of life things, like how Fabric will make Litematica work so much better and not mess up placement of stairs and directional things like that, and the ability to change some settings without reloading to allow more performance-aware configuration.

Running Fabric isn't a magic fix-all, and it carries its own issues. The most prominent is that Fabric mods tend to need to be updated for each version, and you can't drop in an old one if it needs no obvious changes for the new version. For a time we had a plugin that was 6 years old! This limits us to our own mods or open source mods we can fix and re-compile ourselves. For more complex mods like block logging and permissions that are above our pay-grade to maintain, we need to rely only on mods that are popular and have a history of being well-maintained by the community. Fabric is just now starting to be used to run larger and non-private servers- it was usually used for small and technical servers previously. There wasn't a lot of demand for some of the Paper plugin alternatives we needed.

The block logging in particular was a huge reason we couldn't launch with Fabric- at the time we didn't feel there was a mature one, and the prominent solution had changed hands many times and hadn't had a committed maintainer for long. Core Protect, despite being extremely widely used and around for many years, has a lot of quirks and things it does not properly log that we already have to work around. It even caused huge database issues that required manual rescue by editing the database directly. Block logging is complex. Syntax has been following block logging for fabric for a long time, and he believes it's finally in a state we can work with, but there will still be holes we have to plug ourselves. There are still things in Paper that do help without interfering with the core gameplay (or couldn't be disabled) and we will see a performance hit from losing those. There's a huge hardware upgrade coming in the future and we'll make up some of it with that eventually and we have some plans for more in-house performance optimizations and even some things we can do for performance that were simply not possible in Paper.

Any plugin we used, script we ran on top of a scripting plugin, and even some core Paper settings have to be replaced or recreated by us. We had already put a lot of work into building this server, but once the decision was made to transition to Fabric a long time ago, a lot of needed changes were put behind this because it's just more tech debt and double the work to make it then remake it later. Because we're now ready to launch on Fabric, a lot of things that were previously on the backburner to make us better will now be able to move forward. It was a huge wrench we threw into our own process, but we feel it is absolutely worth it to give everyone the experience they should have and by doing this we have learned how to do a lot ourselves and to make stuff that’s more portable.

We've had some people ask if you need a Fabric client to connect to a Fabric server. You don't and you can continue to use what you currently use.