PC V2.3 Experimental
- Thread starter Hated
- Start date
Roadkill115
Refugee
just reposting (as it got moved) the 2.3 relevant thought i had; yay! linear biome progression is dead?
While I was not a super fan of that system, I hope they add a optional toggle because my friend did enjoyed it playing his solo game so I would not be against them adding Biome Progression back as a optional toggle setting.
Roland
Community Moderator
The option is built in. You simply choose to unlock them in the same order as before. Your friend can still do burnt --> desert --> snow --> wasteland without any toggle setting.
You’re right I am thinking of the loot stage cap.
khzmusik
Hunter
I like the notion that buildings give you some protection from the biome hazards, but not 100%. Like, maybe 50% or something.
Having said that, one of the things TFP said they were going to add back in are temperature-related debuffs (being to hot or too cold). Those could also be used as a kind of "gate" for the later biomes. The debuffs probably would not happen indoors (they didn't used to happen in earlier alphas), so I could see that mechanic being the driver for entering POIs in a biome, rather than protection from the biome-specific hazards.
Old Crow
Hunter
Loot Stage Cap is gone now anyway.
EvilPolygons
Survivor
Loot stage caps never bothered me. Most games (even sandboxes) keep the best loot in harder zones. The push back on this particular issue was a little strange, I think.
KirillWright
Scavenger
I liked this mechanic too. Settings would be nice!
KirillWright
Scavenger
I would be happy to watch you search for errors in such a large project. But there are few like me. And the language barrier has not been abolished.
Post automatically merged:
This also stopped bothering me much since it became possible to create items with quality 6.
Whiskey EzCo
Refugee
Just made an account for this idea.
What if for people who play without biome progression off but enjoy the storms, you made these items actual clothing pieces. I saw people on pancakes video annoyed by the idea of having this be the complete alternative but I like the idea of having clothing options when biome progression is turned off. Just a thought.
What if for people who play without biome progression off but enjoy the storms, you made these items actual clothing pieces. I saw people on pancakes video annoyed by the idea of having this be the complete alternative but I like the idea of having clothing options when biome progression is turned off. Just a thought.
Sapient6
Scavenger
The issue for me was doing t6 missions. The new POIs brought about some decent variety, but in the RWG world I spun up for 2.x, two of the new POIs available for t6 missions (Area 7 and Hybrid Mining) are only available in the forest.
Loot stage (and, sadly, difficulty) is already significantly lower in the earlier biomes. It didn't really strike me as necessary to exacerbate that further by capping the loot stage.
Roland
Community Moderator
If your friend moves to each biome to start doing the next tier of quests then they’ll naturally loosely follow what the loot caps enforced.
Rabbitslovecactus
Scavenger
Personally I say ditch all the biome challenges. The only challenge should be to have to gather resources to make the PPE for the biome you plan on going into.
The loot cap appeared to be penalizing players who choose to stay in the same zone, rather than rewarding players for trying harder biomes.
I was spending a week / horde in each biome and kept getting loot capped.
Obsessive Compulsive
Refugee
Your ram allocations can be improved dramatically. That is not a micro optimization. As entity levels and chunk cache builds you need to keep allocations reduced. This in turn helps with garbage removal and the build up. You have a lot of loops with declarations inside of the loop. They should be declared as either static to the class or outside the loop at minimum. Some of it is well optimized but there is a lot of entity updates making new declarations every tick that are not needed. I was reducing the average server load by up to 30% by improving those declarations. A lot of your code is running single thread so it makes little to no sense to create new declarations. Please fix the declarations inside of loops. That is simply poor code.
You could have created an entity pool to reuse a base entity class to avoid declarations, garbage collection and clean up. As people move in and out of chunks, you are fully unloading the entity and then respawning it when they come back to the same chunk. That is where pooling could have saved ram and processing. You take a hit every time an entity spawns.
Particle system could use a pool with a basic struct.
Falling blocks are a type of entity. An entity pool would dramatically improve performance with larger collapses.
You really need more multi threading to avoid the main thread getting bottle necked.
Clients are performing decent but they suffer with bloat. They run processes that the server runs and the server runs processes that the client runs. If you are going to run a sync system, there should be corrections server side else there is no point in a sync. You would be better off with a data oriented system which is what the Unity developers recommend.
You have the burst compiler not being used. You are not using the IJob system which would allow you to safely multi thread. You are not using the Math library provided by Unity to work along side the Burst Compiler. You can use a lot more simd operations if you use the math library which ironically you have installed but do not use.
There are optimizations that could be done to the load process and the shutdown.
Garbage clean up can be run on a secondary thread, instead of main. Allowing you to run it across the update tick. It can also be run with ijob with a callback on the next frame.
You have a thread manager class that has some of the strangest coroutine code I have ever seen. I was able to fix it and reduced allocations 3 years ago. You guys do some weird stuff. Some of it is great and other parts are... well yikes. The reason people keep poking you is weirdos like me keep making optimizers that prove something is being over looked. We are very limited with harmony injections so many are expecting more from you with direct access to the unity profiler.
Perhaps the issues do not become as apparent until purposely over loading a server. Spawn too many z and simulate 30 players just to see what happens in the profiler. You might be surprised
You could have created an entity pool to reuse a base entity class to avoid declarations, garbage collection and clean up. As people move in and out of chunks, you are fully unloading the entity and then respawning it when they come back to the same chunk. That is where pooling could have saved ram and processing. You take a hit every time an entity spawns.
Particle system could use a pool with a basic struct.
Falling blocks are a type of entity. An entity pool would dramatically improve performance with larger collapses.
You really need more multi threading to avoid the main thread getting bottle necked.
Clients are performing decent but they suffer with bloat. They run processes that the server runs and the server runs processes that the client runs. If you are going to run a sync system, there should be corrections server side else there is no point in a sync. You would be better off with a data oriented system which is what the Unity developers recommend.
You have the burst compiler not being used. You are not using the IJob system which would allow you to safely multi thread. You are not using the Math library provided by Unity to work along side the Burst Compiler. You can use a lot more simd operations if you use the math library which ironically you have installed but do not use.
There are optimizations that could be done to the load process and the shutdown.
Garbage clean up can be run on a secondary thread, instead of main. Allowing you to run it across the update tick. It can also be run with ijob with a callback on the next frame.
You have a thread manager class that has some of the strangest coroutine code I have ever seen. I was able to fix it and reduced allocations 3 years ago. You guys do some weird stuff. Some of it is great and other parts are... well yikes. The reason people keep poking you is weirdos like me keep making optimizers that prove something is being over looked. We are very limited with harmony injections so many are expecting more from you with direct access to the unity profiler.
Perhaps the issues do not become as apparent until purposely over loading a server. Spawn too many z and simulate 30 players just to see what happens in the profiler. You might be surprised
Space4Ace
Survivor
- Right click on the game in steam
- Click on properties
- Click the “betas” tab
- The drop down menu will have latest_experimental available <— you want THAT
- Select that and wait for the game to download
EvilPolygons
Survivor
Now that's a good point.
Exempting tier 5+ from the loot cap might have made sense, then. Generally speaking, by the time I'm ready to do t5 POIs, I've long since left the forest behind and have to go back to do any t5 POIs that may have spawned there. What's the point of backtracking and doing an infested t5 when you know the loot is going to suck? So yeah, keep the biome loot caps but make t5 POIs a special exception.
Another idea for the "would make a good mod" list.