Devblog 121

Before we released the XP system it had universal praise and I was on Cloud 9. Shortly after it was released people started to say that it was a grindfest. I figured it was a vocal minority, as I had played the game for hours and had a blast which mirrored the feedback we had been receiving thus far, but as time went on I heard more and more people complaining through more channels. What started out as a few reddit posts ended up filtering into in-game chat and many other channels. I realized it was time to really figure out what all the fuss was about. Seems like most of the complaining happens after level 20, which few of the testers actually bothered to get through. This is why they were happy with the XP system. The other big problem is the lack of researching. I knew we would add this back in at some point, but I did not realize just how fundamental it is to a lot of people's playstyle. To that end, I’m going to add Researching items back into the game next week. I tried very hard to get it ready for this patch, but late last night I had to concede it wasn’t going to make it. It’s not going to be exactly the same as it was before; I want people to be able to research items to create a stack of one time use blueprints where each use is a crafting session. I’m still ironing out the details, but something like that is what we’re going for. If that doesn’t work we’ll try something else until it does. We're also going to have to take a look at replacing XP earning from hitting trees and rocks with something else. It's literally the most inefficient and unfun way to level, and I think the perception that that is what you’re supposed to do to level up has lead to a lot of bad reviews and unhappy people. Perhaps if we simply removed it people would find the other ways of earning XP more enjoyable and more efficient? Either way, whatever changes we make the goal is to make the game more fun and less of a grind, not the other way around. If you seriously thought we released the XP system so people would click on trees and rocks for 60 hours so they could get a gun, you are sorely mistaken. We’re working on making the XP system preferred by almost everyone. That's our goal.

Yeah yeah, we nerfed it too hard. You can once again sprint with the crossbow and the headshot multiplier has been increased from 1.25 to 1.5. Enjoy!

I’ve gone through and reduced the crafting time for a bunch of items that were far too long and didn’t make any sense: Ladder, Bed, Campfire, Ceiling Light, Tool Cupboard, Bow and Crossbow have all had their crafting times halved. I also reduced the crafting time for Gunpowder and Explosives to 2 sec and 5 sec respectively. I know, right?! The more I go ahead and do this, the more I realize crafting times in general can be annoying and I think we’re going to end up doing a lot more reduction and streamlining in the future.

Arrow raiding was a broken mechanic that had to be removed, however its removal did leave a gap in the game: you virtually could not raid anyone until you had C4. Some changes last week were made to make Beancan Grenades a bit more viable in this regard, but that was just a stop-gap for this weaker version of C4. I give you the Satchel Charge! A new mid-game raiding tool, it's made of a Small Stash and a bunch of Beancan Grenades and is mainly useful on Wood and Sheet Metal Doors. It is available at level 15 and can be rarely found in tool crates.

There was an issue that allowed Explosive Rounds to deal almost double-damage to objects it was hitting, making it a far more viable raiding tool than it was supposed to be. Just so everyone knows, explosive rounds have a very large radius meaning that they score hits a lot easier - in addition they deal radial damage in a small 2m radius from the area of impact. This allows you to hit people behind, cover or windows similar to an airburst round. The bug was that both the radial damage and direct damage was being applied to the object being hit. This is now fixed.

When adding and balancing the Satchel Charge I did a very small pass on the cost of explosives and therefore the cost of raiding. I decreased the cost of C4 by 20% so that rockets are no longer the only way to go, and I increased the cost of Beancan Grenades (and therefore satchel charges) by about 15%. I’ll have to watch this one closely and listen to your feedback, but things should be in a better place now in terms of how much you have to risk to get inside a base. I’ve also lowered the unlock level for the Armored Door to level 17 instead of level 20, which will allow you to defend against the Satchel Charge a bit sooner.

I’ve added XP earning to bandaging and syringing other players, but it is limited to 1XP per person healed to prevent group abuse. This applies to you actually applying the bandage/syringe to another person via right-click, as well as giving them the item and having them use it on themselves or someone else. Consider giving some bandages to newspawns now!

One thing I noticed with the whole sharing mechanism was that when you gave out tools to a newspawn they would likely die quickly or not play for long enough for you to take advantage of the XP gains they were offering you. To this end, I’ve removed the diminishing value from others gathering for you. What this means is instead of someone having to harvest for 20 minutes using your tool for you to earn 3XP total, they now will only have to do it for a few minutes. I’ve also reduced the max XP you can earn from any one person who gathers for you to 2.

The big one is Shelves. These unlock at level 5 and are 80% cheaper to craft. I don’t know why we didn’t do this sooner as they are a functional, required item for base design. Ladders are now unlocked much sooner, now at level 10 I’ve also severely lowered the cost of Wood, Wood+Wire, and Metal Barricades by about half the metal cost, and also lowered their unlock levels to 8, 12, and 14 respectively (about 3 levels sooner each). The Waterpipe Shotgun is available at level 13 (to make room for the new weapon).

A progression from the Waterpipe Shotgun, the Double-Barrelled Shotgun allows you two shots before having to reload. While the damage output is slightly less than the Waterpipe per-shot, as well as its accuracy reduced, landing both shots will yield a higher damage output per second than you could achieve with the Waterpipe even multi-stacking them. Available at level 15 and found in Radtown crates. Enjoy!

You know what sucks? Climbing to the top of the Dome to see those green crates waiting for you, opening one up and seeing... a Torch. Or a Stone Pickaxe, or a Water Bucket. Welp, that’s not going to happen anymore. I’ve hand-crafted the loot tables for the two Radtown crates. This means you’re going to find a lot more useful items and tools, as well as good armor. I’ve also set them up to drop stacks of ammo from 8-20 (depending on type), so you can actually use the weapons you find! You’ll also see more drops of low/mid tier weapons, like Revolvers and the Double-Barrelled Shotgun

The Glass turret is now less Glass... It protects 50% more against explosions from Beancans and Satchels, and its health has been increased to 1,000. I’ve also increased its damage output by 50% and made it slightly more accurate. P.S.: You can very rarely find them in loot crates now :)

A while back Garry did the baseline for the visualization, so you could see how much a tool has been repaired (and thus its max condition) but it remained unimplemented until now. Yay.

I did a full balancing pass on armor to get the damage numbers on geared players back in-line. Armor now adds up from all clothing pieces that cover a certain body part, so wearing a hoodie and a metal chestplate will be better than just wearing a metal chestplate. To compensate I reduced the protection values of various pieces of clothing. Overall this means that hits to the body of a fully geared player will now do 15% more damage than before. In addition to this nearly all clothing now offers more protection against arrows than it does against bullets. For body hits on fully geared players the difference between arrow and bullet protection is 10% which should help make low tier guns more attractive.

I did a fairly significant change to how headshots on Bows and Crossbows work. I reduced the projectile radius of arrows, so getting headshots with them requires a bit more skill. However, the arrow headshot damage is now significantly higher, meaning that crossbows kill naked players at full HP with one shot when taking bleeding damage into account. The projectile radius of Slugshot, Buckshot and Explosive Rounds was also reduced to bring them more in line with the standard rifle and pistol ammo. Speaking of Pistols, the Semi-Auto Pistol now holds 10 shots in its magazine and the Revolver cylinder once again has a capacity of 8 shots. This should boost the usability of those weapons without going too far overboard. And most importantly the Semi-Auto Pistol now deals 40 damage instead of 39.99 - you’re fucking welcome.

I completely rewrote the bleeding damage calculations since they were barely doing any damage in their old state. From now on bleeding is displayed as the total damage that will be applied over the entire bleeding duration, so you will always know exactly how much damage to expect from your wound. The bleeding rate was increased to 20 damage per minute and weapons now do 20% of their applied damage as additional bleeding damage, which gives all weapons a decent damage boost over time.

Reviving a player now requires you to hold down E for 6 seconds without moving. Revive attempts that are interrupted will still increase the time it takes for the player to die from his wounds and players will never die from their wounds while someone is trying to revive them. To balance this, bandages and syringes can no longer be used to revive but players will be in the wounded state for 40 to 50 seconds (instead of 15 to 30) before succumbing to their wounds.

Yes, the #1 item on the reddit changelog wishlist actually made it into the game! You can now rotate the placement guide of deployables and building blocks by pressing the R key. For most objects this rotates in increments of 90 degrees, but building blocks rotate by varying amounts to fit their sockets.

I think this one has been annoying literally everyone for ages. Projectiles used to be locked for 10 seconds before you could pick them up when they got stuck in another entity. This is now no longer the case - you can pick up arrows and throwables the second they stop moving.

This was my personal passion project over the weekend. Projectile movement is now verified on the server using a checkpoint system. This makes the server side projectile movement verification much more robust, especially against false positives. Let’s see if it fucks everything up.

Kept working on rendering this week, shaving off another ~2 ms from our frame time:

• Optimized HUD UI rendering
• Optimized motion blur effect; around 20% faster
• Optimized bloom effect again; around 25% faster
• Optimized sharpen and vignetting; around 30% faster
• Optimized terrain rendering; around 40% faster

Terrain rendering is one of the main contributors for our high frame times. I've disabled far splats for regular/default rendering (PVT off), reducing bandwidth drastically. A full featured terrain now requires VT to be enabled. While on the topic of virtual texturing: while it's one of the most powerful optimizations we have - improving performance not only on the terrain surface but also on all terrain-blended rocks/cliffs - it's been fairly unstable. I'm planning to take some time to work on it again soon, to address general instability and micro-stutter, as well as blurriness issues when walking on cliffs, among others. I've identified a bunch of areas in need of optimization work and I'll be looking into CPU-side graphics performance as well this week.

Door traps! I'd made some pretty good progress on one of them already before Helk and I decided that we really wanted to make this one (sketch 3). So as Paul finishes up on fleshing out the concept for that one, I decided to work my way to completing the Single Shot Door Trap. You can check it out below in both 2D and 3D! It’s a single shot trap, which means you would have to go back to reload whenever it’s triggered. Probably best for low level/tier players to add some sort of basic security.

Helk and I discussed adding some more real-life military guns as rare loot drops (similar to the M249 Machine Gun). A couple of guns he suggested are the LR-300 Assault Rifle (basically a tricked out M4) and an MP5. I started work on the LR-300, so here’s some progress on it so far. After this is done, I'll make the high-poly version and then bring it into Mudbox to apply dents scratches to it.

We could have YOLO merged the graphics update in today, but we decided that it’s best to give it some time on pre-release, so expect to see it up and running there within a few days. Also we have been neglecting to fix Savas Island for a long time now, so you’ll see that too when the graphics update goes live. More meds, more guns, more good stuff. (Pictured: Procmap.)

The progress from the dungeon branch will not merge into main this update. Some of the implementations still need to be made for the art, and rather than releasing half working features I’ll keep them on hold there until working.

The military tunnels represent one of the last parts of the new dungeons that is still in a greyboxed state. I started to transform this into its final form this week. First by completing the dungeon that is in the game, and then adding more pieces to the set so we can craft more interesting underground places, including mergers with natural caves.

I made a start on the low-poly for the updated bone armour this week. Generally speaking, the simpler the shape of an object, the easier it is to recreate in 3D and arrays of small, complicated shapes tend to be a pain in the butt. I initially wanted each bone as a separate element but the polycount started ballooning pretty quickly, and it wasn't even looking that great due to triangles being wasted in areas no one would ever see. I've ended up going for a bit of a hybrid approach and grouping clusters of bones into elements which looks pretty good and is way, way more efficient. Here's a shot of one of my test bakes, hopefully you shouldn't be able to tell I've used any workarounds!