Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  

Author Topic: On the subject of many dogs and FPS  (Read 941 times)

askovdk

  • Bay Watcher
    • View Profile
On the subject of many dogs and FPS
« on: May 09, 2011, 01:47:06 pm »
I'm having frame rate problems on my laptop, and I was thinking the animals were the problem. So I did some testing, and I would like the share this single experiment. The experiment was done on my main computer with a 'unlimited' FPS cap.
So this is just one experiment, - the results may be flawed due to a group of hiding goblins, but the results may still prove interesting for those working with lots of animals.
These experiments happened in 0.31.25, a small world with temperature and weather on, - and animals were hurt very much in the end.

Startup:
To generate a lot of dogs I set up a boring secure fort with 28 female dogs and 2 males: FPS 551.
Spoiler (click to show/hide)

Next spring: Puppies!!!!  FPS 495
Spoiler (click to show/hide)

Experiment with stones:
I wanted to have a long path betwen the holding chamber and the meeting zone, so I dug out a very long path. This did create a lot of idle stone. FPS 242
Spoiler (click to show/hide)

Dumping and smashing the stone: FPS 340
Spoiler (click to show/hide)
I don't know where the extra 100 FPS came from, but smashing the stone didn't help.
Sub-conclusion: I can't support the hypothesis that idle stones hurt the FPS, - at least not when it's 'just' about 2000.

Final test:

Running free: 290
Spoiler (click to show/hide)

Untrained pastured: 408
Not pastured - No path possible: 260
Spoiler (click to show/hide)

Doors unlocked, but not pet passable : 68
Only long path allowed : 42
Spoiler (click to show/hide)

Trained war dogs:
Running free and 'unasigned' (i.e. they follow their trainers) : 242
Pastured with path out : 270
Spoiler (click to show/hide)

No path out: about 300
Free, but assigned to 1 dwaf: 270
Spoiler (click to show/hide)

Atomsmashing all dogs : FPS 400.
Spoiler (click to show/hide)

Conclusion(?)  :-\
The picture is not clear, - most of the deviations may be due to other factors than the dogs.
You should apparently never have animals behind 'unlocked, but not pet-passable' doors. It seems like the animals will find a (dwarf) valid path back to the meeting zone and start walking, - they will then hit the door they can't go through and try to repath, - finding the (dwarf) valid path ect.
Otherwise my only result is that pasturing is generally good and free running is bad. So I plan to have even assigned war dogs put in a pasture, and then put the pasture to inactive whenever I need the dogs 'unleashed'.

Logged
My HMoM forts :
 Kindletours - A flying silver city.
 Boardstrap - Thermal borehole HoMM5 style.

Jimlad11

  • Bay Watcher
  • Insert insightful text here
    • View Profile
Re: On the subject of many dogs and FPS
« Reply #1 on: May 09, 2011, 02:32:54 pm »
deleted
« Last Edit: March 13, 2018, 01:16:30 pm by Jimlad11 »
Logged
☼Cat Tallow Biscuits☼...Mmmm...

Chessrook44

  • Bay Watcher
    • View Profile
Re: On the subject of many dogs and FPS
« Reply #2 on: May 09, 2011, 03:14:12 pm »
New question: What if all the animals get dumped in a single cage?  What's the FPS-change then?
Logged

Lectorog

  • Bay Watcher
    • View Profile
Re: On the subject of many dogs and FPS
« Reply #3 on: May 09, 2011, 03:26:57 pm »
Pets try to path through pet-inaccessible doors. They can't go through, so they wait at the door. When a dwarf goes through, the animal runs through while the door is open. I'm not sure how much this helps the overall clarity, but that's the explanation for the pet-inaccessible doors.
Logged

JmzLost

  • Bay Watcher
    • View Profile
Re: On the subject of many dogs and FPS
« Reply #4 on: May 09, 2011, 05:39:36 pm »
Quote from: Lectorog on May 09, 2011, 03:26:57 pm
Pets try to path through pet-inaccessible doors. They can't go through, so they wait at the door. When a dwarf goes through, the animal runs through while the door is open. I'm not sure how much this helps the overall clarity, but that's the explanation for the pet-inaccessible doors.

The point is setting the door to "tightly closed" won't improve FPS, because the dogs are continuously pathing.  Put them in a pasture or cage to improve FPS.  Pasture will let you unleash them without dwarven slacking intervention.

JMZ
Logged
Quote from: Auto Slaughter on October 10, 2010, 11:58:53 pm
Also, obviously, magma avalanches and tsunamis weren't exactly a contingency covered in the mission briefing.
Quote from: DuckThatQuacks on September 29, 2010, 04:08:11 pm
I can assure you that Ardentdikes is not the first fortress to be flooded with magma. What's unusual is that we actually meant to flood it with magma.

noodle0117

  • Bay Watcher
  • I wonder what would happen if I pull it.
    • View Profile
Re: On the subject of many dogs and FPS
« Reply #5 on: May 09, 2011, 06:08:13 pm »
This thread is of great dwarven science.
I like.
Logged

wuphonsreach

  • Bay Watcher
    • View Profile
Re: On the subject of many dogs and FPS
« Reply #6 on: May 09, 2011, 08:48:23 pm »
Chains are still better for border areas.  (I just wish you could set it so that a war animal could slip its leash if it spots a hostile.)

The problem with war animals in pastures near the borders of safe / not so safe is that when they start chasing a hostile, dwarves will come from far and wide to put it back in the pasture even though its in combat.  If Toady would change the job system so that animals outside of their pasture, who have been in combat recently, do not automatically generate re-pasture jobs (until out of combat for X ticks) then it would work better.

But shoving all of the puppies into a pasture zone in an interior area is a good move.  And war animals are good in pastures at various points in the main corridors.
Logged