Print

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

[Slogo]

No the performance gains are going to be unimpressive guarenteed.

The majority of people, that I am aware of, who are really pulling for Multithreading have frame rates often around 10 with forts smaller then the game normally allows.

So to put this in perspective... 2x0 = 0

If the majority of these people were ones who were pulling maybe 100FPS at the start and once they reach 120 Dwarves on a large map with livestocl they hit 30-20FPS then Multithreading would make a difference.

The real solution, other then to get better computers, is just to write more efficiant codes and to take some options out of the Raws/Hardcode and into the game.

That's not really true. I get great framerates even with 100 dwarves but when I look at my cpu usage it's only 1/4th of the available umpf.

I think you underestimate the effects of multithreading and what it can do for DF. DF is in a perfect position for mulithreading as there are some seperate self contained systems. Running the water flow, and temperature calculations are great for multithreading as they aren't something that HAS to be calculated by a specific point in time.

Likewise with pathfinding spawning the pathfinding on different threads should see a pretty good boost in performance. You may have dwarves 'idle' for a moment before finding their path but it's better than your FPS dropping to 10.

To me something like multithreading would be a 2-4 on a top 10 list while full graphics support would be a 6-8.

[Granite26]

I'm pro any sort of graphical improvement, and willing to get behind a text/tiles break at this point in development.  It's something that (according to dev notes) WILL happen.  It's something that will have to be done once and that won't involve extra work later (unlike UI suggestions).

Working under the assumption it would expand the user (sic: donator) base, it makes a lot of sense.

[axus]

Anything without 3-D models and fluid motion isn't "full graphics support".  I'm cool with improving tile graphics support, I wouldn't call it "full graphics support".  "full tile graphics support" maybe?

[LordZorintrhox]

Anything without 3-D models and fluid motion isn't "full graphics support".  I'm cool with improving tile graphics support, I wouldn't call it "full graphics support".  "full tile graphics support" maybe?

That implies that any game without full 3d is incomplete in some fundamental way.  So Super Mario Bros., Myst, Day of the Tentacle, and Starcraft are all incomplete, unfinished games, even though they are considered legendary in the world of games?

[Aqizzar]

To head off yet another graphics-rage shitstorm, talking about three-dimensional graphics for Dwarf Fortress is still valid, since it is theoretically the supposition and extension of the fully 3D Slaves to Armok, and "full tile graphics support" is more specific, especially for people just walking into the vote.

I'm mostly saying this just to keep this tired-ass argument from happening again, which Toady already said not to do earlier in this very thread.

[Zironic]

To head off yet another graphics-rage shitstorm, talking about three-dimensional graphics for Dwarf Fortress is still valid, since it is theoretically the supposition and extension of the fully 3D Slaves to Armok, and "full tile graphics support" is more specific, especially for people just walking into the vote.

I'm mostly saying this just to keep this tired-ass argument from happening again, which Toady already said not to do earlier in this very thread.

I think we should get rid of the voting tool. I think it's arbitrary. Half of things make the game easier in unnecessary ways.

[Dakk]

TL;DR Full graphic support is not a bad idea, but i do believe there are more important things and features i'd like to see first. If giving modders a way to make a graphic for every tile in the game doesn't take much of toady's time, then i'm okay for it, i just don't want it bogging down the development of other, far more FUN stuff.

Regarding the OP, its a valid sugestion, but please don't beg for votes.

[Neonivek]

I think you underestimate the effects of multithreading and what it can do for DF

Well...

If people are running on the smallest forts possible and are getting 10FPS then Multithreading, which would be so it can use two cores, could only at most double this.

It would turn 10FPS on the smallest map into 20FPS.

Add in Normal sized maps and populations and you divide that by 6 which is around 5FPS.

Your kinda overestimating what Multithreading does, it isn't a magical fix.

[Slogo]

I didn't say it was but if I'm running on 4 cores and already get good FPS there's a substantial boost to be had, likewise with 1 core and 10fps. There's a difference between multithreading and multiprocessing, both of which need to be used. Multithreading doesn't hinge purely on your # of cores and can offer performance benefits even on a single core system. http://en.wikipedia.org/wiki/Multithreading_%28computer_hardware%29 offers some good explaination.

What multithreading/processing can really do for DF is decouple the fortress' passage of time with navigation of the interface. So your fortress may slow down to running longer real time per in game time period but you'll see more responsiveness in navigating around the interface and interacting with the world. Basically you would be decoupling the responsiveness of the system and your FPS from the passage of in game time. This would give you a much bigger playability benefit to larger fortresses even if it still took longer for the years to pass.

[Neonivek]

I didn't say it was

No your just expecting more then a 100% improvement from Multithreading. Which while non-magical is close enough for anyone to really care.

Also interesting to note: Getting a response from the interface is unlikely, I believe the Interface is in sync with the game.

if I'm running on 4 cores and already get good FPS there's a substantial boost to be had

No the codes to run on 4 cores is apperantly different then 2 (or so I have been told).

Though I THINK this is very off topic but I don't know if people even want to continue with the topics original discussion.

[Exponent]

Though I THINK this is very off topic but I don't know if people even want to continue with the topics original discussion.

As much as we tend to view death negatively, sometimes it is ethically acceptable to allow a topic die.

In fact, sometimes it is ethically obligatory to allow a topic to die. 

[Enzo]

Though I THINK this is very off topic but I don't know if people even want to continue with the topics original discussion.

As much as we tend to view death negatively, sometimes it is ethically acceptable to allow a topic die.

In fact, sometimes it is ethically obligatory to allow a topic to die. 

But not to talk about multithreading. That has been done to death already.

[Zironic]

The three topics that should never be mentioned again:
Warning, May cause spontaneous flamewar

Spoiler (click to show/hide)

1.Multithreading.
2.Graphics Support.
3."Full Demos".

I still think game play enhancements should come first. The d# have really helped to reduce my lag, so the idea of more processing power is kind of moot for me.

[Footkerchief]

^^^ What's #3 about?  Warez?  That's not really flamebait per se.

[Zironic]

^^^ What's #3 about?  Warez?  That's not really flamebait per se.

It is certainly ban bait.