ssrjazz

We all get quests at the same time and then do them all together. Don't see this as a game breaking issue.

Not true. Only the intro quest is solo. The others you get get shared and you can do them cooperatively

My friend Scotti and I each have our own dedicated servers. We both did 10k RWG worlds with completely different seeds. I was playing on his after playing on for a few hours and it dawned on me that the starting town with Rekt in it.....was exactly the same as mine. Is this intentional? Doesn't seem like it should be. Would make starting a new RWG map kinda boring if you know what the layout of the starting town is.

*continues to hit F5 on his SteamDB Watchlist page*

One of the biggest performance hits of large cities were all the windows, especially in the skyscrapers, and caused a bunch of overhead due to unnecessary draw calls. @faatal has posted previously about how windows were rewritten to be much more performant. That and other optimizations that have happened along the way will hopefully add up to less FPS drops in cities in a21 and just better performance overall.

I -really- wish this could get backported to a20, but I know it's too big of a change to be reasonable for that. This is one of the top things I'm looking forward to for a21. Can't wait!

No.... but it seems that 8GB of RAM in a server isn't enough to actually be able to RWG an 8k sized map in a20. I could do a 6k one just fine though. Guess memory requirements went way up.

Did a fresh install of a20 on a linux server via steamcmd (I have command line access to this server) At some point during RWG for an 8k map, the server crashes with this in the output log: 2021-12-06T15:31:40 1095.307 INF WorldGenerator:Wilderness Prefabs Spawned 103 Caught fatal signal - signo:11 code:1 errno:0 addr:0x50 Obtained 9 stack frames. #0 0x007f76eddd1630 in _L_unlock_13 #1 0x007f76eefb27da in void prcore::RemapGenericIntegerToFloat<TexFormatARGB8888, TexFormatRGBAFloat>(prcore::InnerInfo const*) #2 0x007f76eefa050b in prcore::BlitterRemapAny::Blit(prcore::InnerInfo const*) #3 0x007f76eefa8e9b in prcore::BlitImageRemapNoScale(ImageReference const&, ImageReference&) #4 0x007f76ef157c54 in bool GetImagePixelBlock<ColorRGBAf>(unsigned char const*, int, int, GraphicsFormat, int, int, int, int, ColorRGBAf*) #5 0x007f76ef11d6d0 in Texture2D::GetPixels(int, int, int, int, int, ColorRGBAf*, int) const #6 0x007f76ef0ec301 in Texture2DScripting::GetPixels(Texture2D*, int, int, int, int, int, ScriptingExceptionPtr*) #7 0x007f76ef4634fb in Texture2D_CUSTOM_GetPixels(ScriptingBackendNativeObjectPtrOpaque*, int, int, int, int, int) #8 0x00000040b08175 in (wrapper managed-to-native) UnityEngine.Texture2D:GetPixels (UnityEngine.Texture2D,int,int,int,int,int)

Short answer: Yes, shadows are computationally expensive (hard). Longer answer (from 2015, but still valid): In Video Games Why Are Shadows So Hard To Render

I would like to second this request….

I seem to remember a paint/texture discussion somewhere in this A20 diary thread too. I don't recall any discussion of it being taken away.

^^^^ This is what Madmole mentioned several pages back. Honestly, I can't wait for it to be implemented. Having that many block upgrades certainly didn't add to the fun for me. Here's what the man himself had to say to those who think he's ruining the game:

Has Nitrogen been updated (or does it even need to be) to support the latest a19 build that came out last week? If it does need to be updated, any ETA?