Update 1.29.4: High Power Update

OCT 19 2025

Hello all,

Today we have updated GunColony with some buffs to underpowered high-caliber weapons like battle rifles and revolvers to make them actually high-powered, competitive options in the battlefield, just in time for them to help you mow down some mobs during spooky season! We have also made some other balance changes, brought back the Halloween playlist, and introduced some major internal changes to the Place game engine powering GunColony in this update.

PLAYLIST UPDATES

• Brought back the Halloween playlist!
• Removed the Summer playlist (PoolDaySummer)
• Removed the October Fools playlist (actually it was removed 1 week after the original update)

BALANCE CHANGES

• Groza: Supply cost reduced by 3
  • The Groza currently trades off too much in stats to earn its 100 RPM additional fire rate and slightly higher mobility compared to the AK-47: its recoil, range, headshot damage, and supply cost are all significantly inferior. To give it a viable niche, we are giving it a large buff to Supply Cost to make it a more cost-effective close-range alternative to battle rifles. It now costs only 2 Supply more than the AK-47.
• G3A3: Supply cost reduced by 1; headshot damage in full-auto increased from 2.1x to 2.2x
  • The G3A3 was designed to be a budget full-auto battle rifle, but in practice it was less efficient than semi-auto battle rifles like the M14, which was cheaper and dealt a higher damage-per-second to hit areas affected by the semi-auto bonus damage multipliers in GunColony. This buff cuts the cost disadvantage by half and gives the G3A3 a slightly more respectable headshot damage to bring its stats more in line to alternatives.
• Rhino: Supply cost increased from 1 -> 2; base percentage of recoil that resets to center increased from 0.8x -> 0.96x (firing rapidly will still reduce it); firing delay reduced from 0.1s to 0.067s (note: due to current rounding behavior the Gunslinger perk will reduce it to 0.05s now)
  • The Rhino should now be much easier to control than before and very well-suited to landing precise shots on target due to the firing delay reduction. In exchange, we have increased the Supply cost of the weapon.
• Colt Python: Supply cost reduced from 2 -> 1; hipfire accuracy reduced by 20%; firing delay reduced from 0.1s to 0.067s
  • The Colt Python has received the same firing delay reduction as the Rhino as well as a Supply Cost decrease! However, we have decided to make it a little more unwieldy to hipfire in return, with the intention of limiting its high close-range firepower to aimed shots. These changes could make the weapon slightly too powerful on net, but we believe that it is important to keep a 1 Supply cost double-action revolver in the game, and will make further changes to the weapon lineup if necessary
• SKS: (Nerf) Is now loaded individually by default (4s total reload speed). The stripper clips is now available as a modification available by default without needing to level up the weapon. The stripper clip modification has a -10 Backup Ammo downside.
  • This is meant to be a targeted change to slightly reduce the power of cheap SKS builds in Competitive mode by reducing the amount of modification slots available in these builds, without affecting the balance of the weapon when using larger magazine modifications. Note: the weapon will actually reload faster by default than before if you have spent 4 or fewer bullets, and gains the ability to fire while reloading to cancel the reload, so this is not exclusively a nerf.

OTHER CHANGES

• 20-Round Magazine (SIG 550, SIG 556): Is now unlocked immediately at level 1, instead of at level 50. This avoids it clashing with the extended magazine unlocks at level 30 and 50.
• There was a bug where people could unlock the Neal SMG very rarely in special vote crates despite it being meant to be only available in scenarios. This was because the Neal SMG was originally coded to be a loadout weapon before we decided to make it scenario-only. We have kept the unlock method because it’s funny, but Neal SMG has been changed to be only usable in Mob-Arena if you do manage to unlock it (same config as RBB-451).
  • The Neal SMG will retain the same Supply Cost in the Scenario mode.
• Updated R8 Revolver description stating that it has a longer charge time than other revolvers. (It still has a 0.1s charge time as before, whereas the other two double-action revolvers are reduced to 0.067s)
• The HTI is now available in loadouts and is no longer an early access weapon
• Added support to Minecraft 1.21.9 and 1.21.10
• Removed October Fools playlist in the previous update

REFACTORED BLOCK BREAKING SYSTEM

We have updated Place Engine with the new Place SDK system now powering block type definitions in the game engine. This means that all of the code that powers the block damage and destruction system in Mob-Arena and PVP has been refactored to use the new block type and stat system. This brings several crucial benefits to the future development of Place and the standalone game, which we will detail in the devblog section below.

If you notice any new bugs with block breaking after the update, please do not hesitate to report it on our Discord server.

During this refactor, we have also discovered a few errors in our previous block stats definitions, which have now been fixed. For example, Glazed Terracotta will take more damage to break compared to before, matching our intended values.

DEVELOPMENT PROGRESS UPDATE

The development of the Place engine is now back in full force in October, and we have been working to implement the Place SDK design originally unveiled in August in a full, end-to-end manner. To recap, the Place SDK is a powerful game object data definition and processing system that will encompass all game objects in the Place engine, such as block, entity, and item types. The Place SDK is what allows the Place engine to be entirely data-driven and moddable using a type-safe Kotlin DSL, while retaining extremely optimized runtime performance. One aspect of the Place SDK that we didn’t mention before is that it is built to support multitenancy, meaning that a single server process supports multiple logical servers with different modlists and can share game object registration data between them to save memory – this is required long-term to get around the “large modpacks using 10+GB of RAM” problem that currently plagues modded Minecraft.

Progress on the Place SDK is coming along steadily. In August, we announced that we wanted to migrate block types and stats to the PlaceSDK system first, and then work on a new mob system in GunColony. We are glad to report that we have finished a working version of the Place SDK DSL system early in the month, and now we have fully integrated it with our block type system, migrating all block definitions from our previous YAML file to the new DSL with the help of LLM-powered tools and a comprehensive unit test to ensure correctness. The full migration elegantly solved several of our previous problems; for example, we are able to load block stats supported by Minecraft from our Minecraft data source in the Place SDK DSL step rather than at runtime, taking advantage of the fact that the DSL is running the full Kotlin programming language rather than being a YAML file. This significantly reduced the “hackiness” of the Place server loading process (at least when it comes to blocks right now). With the successful migration of block types, PlaceSDK is now behind a real gameplay feature in GunColony – every block broken in the Mine-Spades gamemode is already powered by the system that will become the source of much of the technical advantages and uniqueness of Place in the long-term.

What will now commence is the development of future mechanics that go above and beyond what is possible right now in not just the GunColony server, but also the entire Minecraft and voxel gaming communities. Our original mid-term goal was to use the Place SDK to overhaul the mob system in GunColony by Halloween, but due to development delays in September that were covered in the previous update, we will now focus on a smaller subset of features for Halloween, most likely new mob and boss skills powered by the Place SDK engine, with the rest to come later. (Depending on challenges encountered during development, it is possible that we may also launch one larger belated Mob-Arena update after Halloween instead.) By creating the backbone of the system in the near-term, we can reap the gameplay benefits in Mob-Arena, clear technical hurdles to update to the latest Minecraft versions, as well as give us a lot of extra time to refine the system on a real production gamemode before eventually making it available to modders. In the longer term, we retain our ambitious goal to create the best, most customizable (through Place SDK), and most performant mob AI and behavioral system in all of voxel gaming. We want community creators in our future FPS game platform to be able to create any mob they desire and effortlessly spawn thousands of them in the same map to create epic horde-mode, PVE, and PVPVE experiences playable on low-end machines.