Placing Blocks in Planes is Bugged
So this is going to be a couple of issues wrapped into one. I believe it is caused by placing blocks in planes. I was attempting to build a cube of blocks when I realized this problem, everytime I placed a plane of blocks it would get laggier and laggier. I assume this is an interative looping problem. I think each block that is about to be placed is doing a check for if it can be placed against every block already in the grid. I could be wrong, but to reproduce this issue just try to build a cube/brick out of planes with 0.5m armor blocks on the order of; 170m x 20m x 80m. Each plane can only be 1840 blocks or whatever, which is fine, I suppose, but even with that limit it gets slower and slower at placing them each time.
Additionally, while attempting to debug this issue, I bundled the partially built cube into a blueprint to see if I could just build a cube out of the blueprint instead of manually building it. Boy was I wrong, that file structure is a fucking disaster, no offense, but what struck me as odd was for only a few thousand blocks the size was 160mb. Redship is only 15mb. Redship is a larger grid, and so thus uses fewer blocks, but this was strange to me either way. I think something is actually being caused by the planes placement. We didn't have this problem in SE1. So I wonder if there is an optimized solution not being utilized. I wish I could attach my .json. But you probably don't want 160mb anyways
I have the same bug
I suppose I should have posted specs. I'm sorry.
9700x 32gb of ddr5 @8ghz, 2080 super, 980 evo ssd.
Given this thread; https://support.keenswh.com/spaceengineers2/pc/topic/45870-high-network-usage-when-placing-blocks
I decided to try to turn off my ethernet adapter to see if I could notice any improvements, hoping it was related to the network spikes. It is not, and in fact not at all correlated with the network spikes either as the initial spikes do rise at the same time, the fall back down comes at a different time from when I regain control of the application. (note I the system isn't lagging while this is happening. Also note that the application ram usage spikes significantly (to near system max, so I assume it would take more if I had more) for an indefinite amount of time AFTER the block placement (minutes, and seems to randomly return to original allocation at an unknown pattern) and that the cpu spikes to about 8% so a short time. (seconds)
I suppose I should have posted specs. I'm sorry.
9700x 32gb of ddr5 @8ghz, 2080 super, 980 evo ssd.
Given this thread; https://support.keenswh.com/spaceengineers2/pc/topic/45870-high-network-usage-when-placing-blocks
I decided to try to turn off my ethernet adapter to see if I could notice any improvements, hoping it was related to the network spikes. It is not, and in fact not at all correlated with the network spikes either as the initial spikes do rise at the same time, the fall back down comes at a different time from when I regain control of the application. (note I the system isn't lagging while this is happening. Also note that the application ram usage spikes significantly (to near system max, so I assume it would take more if I had more) for an indefinite amount of time AFTER the block placement (minutes, and seems to randomly return to original allocation at an unknown pattern) and that the cpu spikes to about 8% so a short time. (seconds)
Update, that cpu spike seems to go to 32% if I've left the application running in the background for 30 minutes and then attempt another plane of armor blocks.
Update, that cpu spike seems to go to 32% if I've left the application running in the background for 30 minutes and then attempt another plane of armor blocks.
Update: If I use mirror mode it takes nearly 20 times as long to place the blocks, but then I do not immediately regain control of the application, the cpu spike holds at 40% for an additional duration of about 2 times as long as a normal full 2000 block placement would have normally taken before I regain control and my cpu returns to "idle" (in space engineers 2) levels.
Update: If I use mirror mode it takes nearly 20 times as long to place the blocks, but then I do not immediately regain control of the application, the cpu spike holds at 40% for an additional duration of about 2 times as long as a normal full 2000 block placement would have normally taken before I regain control and my cpu returns to "idle" (in space engineers 2) levels.
Please close the topic. The analytics memory allocation error patch seems to have fixed this issue.
Please close the topic. The analytics memory allocation error patch seems to have fixed this issue.
Replies have been locked on this page!