Messages

The size scales assigned to creatures make a lot more sense on an exponential scale. Squirrels have a size of 0.15; 80kg * 0.15 = 12kg, so ridiculously large squirrels, while 80kg^0.15 = 1.93kg, which is much closer to a reasonable value. The other direction doesn't work out quite as well (size 4 elephants should only be around 5,000kg rather than 40 million kilograms) - but in the real world Elephants weigh about twice as much as Rhinoceroses, which are size 3, so it's clearly not a linear scale.

But for huskies and labradors specifically, that's irrelevant. They're exactly the same size as adult humans, so whether the scale is linear, exponential, logarithmic, or whatever is irrelevant. Other animals with a size of 1.0 include Alpacas (real world average weight: 60-85kg) and Wargs. Clearly, substantially larger creatures than your average wolf.

RimWorld Huskies & Labradors are quite a bit larger than their real world counterparts. Their body size is the same value as humans (1.0, compared to wolves at 0.85). You should think of them as equivalent to real world St. Bernard dogs, bred to be strong working dogs and not insignificantly larger and stronger than wild wolves.

Day 1. Colonists that both have the skills enabled to be useful and don't have traits that make them a massive liability are not particularly common; don't let an opportunity pass you by. Around day 30, my standards drop to 'anyone who can do medicine and isn't a pyro'.

One small feature from Zhentar's Vanilla Tweaks that I think should be in Vanilla: the roof overlay with a distinct color for overhead mountain tiles:




edit 1/25:
And a second one, meal meat overlays:

If Psychology detours InteractionWorker_RomanceAttempt.RandomSelectionWeight, then they aren't compatible (and checking the thread, it does). If Fixes is higher in your mod list than Psychology, then Psychology's detour should overwrite the Fixes version (since Fixes uses a standalone detours library, there should not be any conflict detected and Psychology will just "win")

I count 13 extra cells travelled for the given path compared to optimal. That seems like more than 30 ticks.

You must be looking at Mason's screenshot, which is that bad. Rock5's is not.

I don't mind that it's not the most optimal path if it's close enough but what I don't understand is, shouldn't it look primarily for the closest path in a straight line to the target first? I don't understand why it went up in preference to going straight.

That's what the Vanilla pathfinder does. If it worked well, none of us would be in this thread now.

My pathfinder uses a heuristic based on RimWorld's region system, which composes groups of passable cells into a larger unit (each region can be up to 12 cells by 12 cells, so 144 cells or less). Estimating the distance between each linked region lets the heuristic account for large obstructions (like walls and mountains), as well as some chokepoints, like doors. Smaller obstructions that don't completely cross a region are effectively invisible to the heuristic, so it doesn't always make the right choices when it encounters them.

Jump Point Search only works on uniform cost grids, so it's useless for RimWorld.

There's not anything you need to do; A16 is much better about trimming down the worldpawns list but caravans added many new opportunities to create worldpawns so the two have somewhat cancelled out. I've found and reported a bug that's keeping some worldpawns around when they shouldn't be, and I think we'll get some other tweaks in A17 to do a better job of keeping down the worldpawns.

I think that's a vanilla issue; there have been a number of similar reports in the bugs forum.

Slight "hooks" near the end of the path are a known weakness of the current version; I've been experimenting with ways to reduce that tendency without sacrificing too much performance. mason_water's example is a pretty extreme case though; are others seeing it behave that poorly?

This looks like an obvious misspathing. The optimum path looks to be under that little hill in the middle. Instead it goes up, after rounding the mountain, instead of searching in a straight line to the target.

It does look like going under that hill would be faster, but not by very much; only about 30 ticks if I'm counting right; only going a few tiles further than it needs to. That is well within the quality of path I intend for them to pick.

They indeed do not care about getting the fastest path. They only care about getting a "good enough" path, and their standards aren't particularly high.

The 103% already included the bonus from manipulation. I did mess up the calculation a bit though, since that comes out to about a 72% success chance.

To be pedantic, the malariblock sets immunity to 61% and then immunity slowly decays from there. Plague & sleeping sickness decay at 2% per day, and malaria at 3% per day. So malaria acquired three days after taking malariblock starts at ~52% immunity.

Bionic eyes don't improve surgery success chance in A16. So you were at 68.5% success rate, and with brain surgery, a ~75% chance that failures would be fatal.

To explain the actual damage calculations: Up to the first 50% of armor (per individual piece of armor, not the total sum) reduces the damage of every hit. The next 50% gives you a chance of taking no damage from a hit. So  75% blunt armor means each punch has a 25% chance of doing zero damage, and the other 75% of the time each hit will do 50% damage. In the 3:00 test, each punch (that connects) does 2 damage because the power armor absorbs 3 damage and then the shirt absorbs 1 damage. So the damage we see the pawn taking is WAD, and I'm fairly sure it's still in line with Tynan's intent; while conceptually a dozen naked people easily beating down a power armored god makes little sense, you aren't supposed to be able to create invulnerable super soldiers; there's supposed to always be some risk inherent to combat.

The 3:00 pawn goes down because his torso was destroyed, and the 8:00 pawn is downed due to pain, so durability of bionics isn't a factor in either (indeed the 8:00 lasts longer than he would with natural, pain feeling arms).

I did notice one thing mentioned that I'm not sure is an intentional design feature - that low quality armor lasts much longer in battle because it absorbs much less damage.

Bump for fixes update