Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[FleshForge]

Oh sure, job calculation of all kinds has an impact, and the more complex it is the bigger the impact - make 100 guys engravers and then designate a large area to be engraved e.g.  Not that I think engraving SHOULD be that complex but evidently it is.

[oso]

In my current fort I have a hell fortress,
 The underworld part is still locked up, but there are dozens of undeads in the upper part that started moving around immediately, and are still there. 

DAMNIT MAN! Wheres my spoiler alert... I didn't know about that yet...

[FleshForge]

I went back and fixed that but, you know, you quoted it...

[Axecleaver]

[bump]
I will admit this makes for something of a necro post. But what's a few months for an important issue? And this thread has so may useful suggestions and valid points that it'd be a shame for it to get buried.

Anyway, I added this issue as a suggestion in the DF Eternal Suggestion Voting suggestion list. Specifically, I called it "FPS Improvement: Blood spatter = steady FPS loss". If any of you can suggest a better title or description, share it and I'll edit it.

Also, do vote for this if you think it's more important than most of the other 300+ suggestions. It has one vote so far (one out of my three votes).

PS: I noticed that "Improved (Speed Up) Pathfinder" is currently ranked #3 with 983 votes! This suggestion claims:

People with slow computers can run the game better. People with fast computers can have more Dwarfes.

This shows that a LOT of players consider FPS an important issue. However, this also shows how poorly informed many players are since, apparently, pathfinding has little actual impact on FPS.

[rephikul]

how poorly informed many players are since, apparently, pathfinding has little actual impact on FPS.

I agree with the bolded part.

[Axecleaver]

I'm not sure if I understand you correctly.
Are you agreeing with me that many players are misinformed about pathfinding having a large impact in FPS? Or, are you implying that I must be poorly informed if I believe this? If it is the latter, then may I point to what Khift wrote on page one of this thread?

With regards to "pathfinding friendly", no such thing exists. It's all bullshit people have come up with to try and explain why some forts run faster than others. No test has ever successfully proven that any style or form of fortress is more "pathfinding friendly" than others. The only verifiable fact ever found is that crowded one tile hallways are poor for both dwarf moving speed and FPS, and I have very few of those. Ramps vs. stairs? Some tests show one, others show the other. Many show no difference at all. Vertical vs. horizontal? Same thing. Wide open, or with defined traffic pathways? Same thing. The pathfinding algorithm isn't even understood fully. We can't possibly begin to design our fortresses around it and anyone who claims to know how is full of it. We don't even know if the pathfinding algorithm treats diagonals as one step or 1.4 steps, let alone anything more obscure.

Also, I think there's ample evidence to prove that the poorly designed system of blood spatter and other contaminants is the primary culprit for FPS death in late-game maps.

[Jacob/Lee]

Oh yeah, containments have an effect on FPS easy. I ran DFcleanmap, no FPS gain, but I found a random animal and looked at it's inventory. It has 20 pages of various waters, bloods, illithid toxins, pus and other vile stuff. Multiply that by the amount of animals I have (450) and that's a lot of pages.

[mrbaggins]

If contaminants on clothes is an issue, maybe a working clothiers industry and replacing peoples clothes would work?

How do you even clothe a dwarf anyway?

[rephikul]

may I point to what Khift wrote on page one of this thread?

With regards to "pathfinding friendly", no such thing exists. It's all bullshit people have come up with to try and explain why some forts run faster than others. No test has ever successfully proven that any style or form of fortress is more "pathfinding friendly" than others. The only verifiable fact ever found is that crowded one tile hallways are poor for both dwarf moving speed and FPS, and I have very few of those. Ramps vs. stairs? Some tests show one, others show the other. Many show no difference at all. Vertical vs. horizontal? Same thing. Wide open, or with defined traffic pathways? Same thing. The pathfinding algorithm isn't even understood fully. We can't possibly begin to design our fortresses around it and anyone who claims to know how is full of it. We don't even know if the pathfinding algorithm treats diagonals as one step or 1.4 steps, let alone anything more obscure.

Also, I think there's ample evidence to prove that the poorly designed system of blood spatter and other contaminants is the primary culprit for FPS death in late-game maps.

Really? That's all you have to show as *proof* that path finding has no impact on fps? Have you read anything on path finding outside the scope of this thread, or done some experiments yourself? While it's not easy to exclusively test pathing because increase pathing always bring many other variables into the equation, but since it's always show up as candidate for FPS loss in any observable situations it's pretty close.

EDIT: Decided to test for the hell of it: 2 herds of unicorns, 60 pages of unicorns total with unicorn being the only common creature type existing give 46-47 fps when they cannot move and 39-42 fps when they are moving. Noticable FPS fluctuation with observable pauses when they can move, which's expected. This was done on my mod, so vanilla DF will give roughly half of that FPS in case you want to confirm yourself.

i3 @ 2.27GHZ per core