Coolthulhu
-
Content Count
1706 -
Joined
-
Last visited
3 Followers
-
Important note: dynamically changing the rate of simulation based on performance is not "optimization", it's an ugly hack that tries to hide bad performance by pushing it into later game, reducing reproducibility of bugs. Looking (and clapping) at you, slicksters starving in 5kg/tile CO2. -
Ecosystems Plants growing wild Wild critters reproducing for as long as there is excess food, burying eggs and starving to death if there isn't Plants reproducing without help of dupes and pips Weed plants to liven up the place, cause allergies, indicate temperature Wild plants having preference for other plants being/not being nearby, reproducing or dying if met fully or not at all Hungry critters pathing through soft terrain, digging up tiles (get rid of that /2 mass on dig), shoving items out of the way, and aiming for their food Soil tiles forming from big soil/sand/etc. items outside compactors if big enough and left for a while Pufts growing slime tiles Dirt near water or mud sweating mud items or reinforcing mud tiles (at the cost of water and own mass) Basically, a living world, as opposed to current "critters are input-output machines that waste more CPU cycles doing things that don't matter". Would require optimizing the AI heavily, but that is needed anyway. As is well known by anyone who tried to raise pufts or slicksters.
-
Output and Intake Fittings delete resources
Coolthulhu commented on fritt's bug in Oxygen Not Included: Spaced Out!
I'm pretty sure it gets even worse if you use the external loaders. I haven't seen the bug happen without them, but with them it was as much as 50% of all in-out. -
When exuberant plants drop their 4kg of rot pile, they produce a tooltip saying that food has decayed. No actual food is rotting, it seems that the warning is just tied to rot piles being produced, regardless of source. This is misleading, since "food rotten" suggests improper storage and/or overproduction.
-
My game crashed. After the crash, I checked Player.log and it ended with this: src/steamnetworkingsockets/clientlib/steamnetworkingsockets_lowlevel.cpp (1852) : Assertion Failed: SteamnetworkingSockets service thread waited 1665ms for lock! This directly adds to network latency! It could be a bug, but it's usually caused by general performance problem such as thread starvation or a debug output handler taking too long. I'm not sure if it's the actual cause, but it has a failing assertion so it might be it. It could have been a memory leak instead, since the game certainly gets more RAM-hungry as the time goes on without restarting it.
-
Trailblazers sometimes create a weird invisible "teleport" that allows dupes to get on/off rocket at the point the trailblazer was supposed to land. On the screenshot, dupe tries to store food in a fridge inside the rocket, taking a path through this teleport. The weirdness seems to clear after a save+load.
-
Same here. I noticed that starmap seems to have acceptable performance right after loading a save, but then it degrades really fast. Leaking resources?
-
The UI for rockets can sometimes display data for one rocket when another is selected. On the screenshot, you can see the right rocket (selected one) with an empty liquid cargo module. It is empty as you can see both from the mouseover tooltip and from the bar rendered on it in game world. The rocket UI shows it with a partly filled liquid cargo module and with an artifact module - those two belong to the rocket on the left. The selected cargo module is one from the right one. This can be confirmed by looking at rocket's name: both mouseover tooltip and rocket UI show the name of the rocket on the right. The rocket UI shows correct data for selected module - the (not shown) storage data below will show the module as empty. It's just the rocket summary thing that is linked to the wrong rocket.
-
Seems that rocket data is not properly cleaned up on delete. On the screenshot, you can see a rocket with no Artifact Module, but the UI for this rocket mentions the non-existent Artifact Module. This rocket was built on a spot where a previous rocket was completely deconstructed. Had to do it in "two saves+loads" because of rocket deconstruction crash, but after the last load, only a single solar panel remained from the old rocket. This solar panel somehow "stored" data from the deconstructed rocket - probably by referring to the old rocket that wasn't cleaned up properly. Regardless, the solar panel was removed before any part of the new rocket was built. Almost certainly related to
-
Some of my mealwood seeds are spawning with two "(Original)" notes below them. This isn't two two tooltips or anything - only one seed is spawned and it seems to happen to all the mealwood farmed on this planet. Mealwood isn't native to this planet, it was sent there with a launcher.
-
Correction for point 5: it will plant some seed of the same type, which may be the original or a completely different mutated variant of the seed. It doesn't even need to be the same variant for the whole copy block, you can copy an original seed and get mixes of original and all nearby mutants. Pretty sure it's related to proximity, either seed to farm or seed to dupe.
-
Pretty sure it must be a bug, since artifact missions start a cooldown on artifacts on that PoI.
-
I'm getting common crashes when fully deconstructing rockets. After loading a save, it's usually possible to deconstruct it without a crash. Log file shows nullptrs in rocket animator.
-
Expected: sending a rocket that can't mine a PoI, but can harvest an artifact, on a round trip, will make it go back as soon as it grabs the artifact. Got: the rocket fails to turn around, doesn't even show "Flight status" in tooltip.
-
Beetas kept in hydrogen atmosphere seem to be inserting hydrogen "bits" into their hives once in a while. Could they be trying to grab uranium that was completely exhausted by some other beeta a moment ago? The hydrogen doesn't do anything, just sits there in the hive forever.
