Print

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

[10ebbor10]

3 questions:

1.What are you going to do about the fact that multiplying the amount of food items will be a major drain on CPU resources. (Especially when things like rotting and such are going to be considered, and food stockpiles aren't failproof anymore. I'd rather have food rotting and such than realistic time, for a variety of reasons)

2. How are you going to handle fast mode? You can't just tell the game to go faster.

3. Why? Are these assumptions correct;
    Pro's
     -Realistic
     - allows for sensible shedules
    Contra
     -Lot's and lots of work
     -Seperates the background (lives of dwarves) from the foreground (live of the fort)
          -Ie, more like seeing highly detailed pictures of dwarven lives, rather than  a single, if somewhat blurry video of a dwarven life.

[GreatWyrmGold]

First off, times aren't that off. If a dwarf takes a week to walk across a fort, he's hauling a boulder or you've got a huge, labyrinthine fort.
Second off, read the OP. Andeerz specifically mentioned food.
10ebbor10: I think the big issue with objects draining FPS is based on the number of STACKS; a stack of 10 plump helmets takes up about as much RAM as a stack of 10,000. The issues of getting fast-mode to run at an acceptable (i.e. faster than slow-mode) speed are important, though.

[Andeerz]

3 questions:

1.What are you going to do about the fact that multiplying the amount of food items will be a major drain on CPU resources. (Especially when things like rotting and such are going to be considered, and food stockpiles aren't failproof anymore. I'd rather have food rotting and such than realistic time, for a variety of reasons)

I cannot give any concrete solution.  I do not have enough knowledge about how the game stores the data to offer any suggestion on improvements.  However, like GWG said, I think that having an increase in number within stacks is not going to harm performance much.  There would be more stuff lying around, but it's hard to say how much this will impact performance without actually trying it.  If eating and drinking frequencies increase and there are more food units per harvest, then I'd imagine the increase in consumption of these items could partially counteract increased production of these items... but I could also see situations where that might not be the case.

2. How are you going to handle fast mode? You can't just tell the game to go faster.

Read the OP.  I am not telling the game to go faster.  I'm telling the game to go the same speed (in fast mode) as it does now in dwarf mode, but have movement speeds changed along with some other things.  I try to address the issues with this in the OP.

3. Why? Are these assumptions correct;
    Pro's
     -Realistic
     - allows for sensible shedules
    Contra
     -Lot's and lots of work
     -Seperates the background (lives of dwarves) from the foreground (live of the fort)
          -Ie, more like seeing highly detailed pictures of dwarven lives, rather than  a single, if somewhat blurry video of a dwarven life.

I think GWG highlighted a huge problem right here:

For example, if you sent an army out to travel the world map, they'd take a month to get to the edge of YOUR map, but then the rest of their (say 4 month journey) would go buy in a flash in fortress time.

These kinds of discrepancies open the game up for huge exploits and holes in the game mechanics when things involving interacting with stuff outside of your fort come along (which are indeed planned!).  My suggestion might be a good way to solve this problem.  Additionally, my suggestion would potentially allow for a concrete distinction between day and night; night raids could happen, forts would be less active at night when most people are asleep, there would be non-trivial reasons for having artificial lighting (like torches and stuff), all of which could allow for some very interesting strategies to be employed and !!FUN!! situations to happen to say the least (i.e. realistic strategic problems that people IRL had to make solutions and accommodate fort designs for... and that would be fun challenges for players).  I could give quite a few examples of what this alone could open up.  I can also give reasons why other suggestions of night time simulation in dwarf mode would not really work in the long run (like having a handful of chunks of time a year being designated as "night").

With regard to the separating background from foreground, with my suggestion, the only thing that would really change in this regard is that the player will see moving entities appear to "teleport" about at normal (fast) game speed instead of seeing them move one tile at a time.  Anything involving moving around would look different but ultimately would behave similarly.  I wouldn't miss being able to see tile-by-tile movement for most of the game.  In slow game speed, one would be able to see the kind of movement one currently sees in dwarf mode.  Sleeping and eating would happen more frequently, along with corresponding pathing.  Hauling would happen faster.  However, crafting, mining,  and other jobs wouldn't necessarily change at all in their time and resource requirements (other than time taken to move raw goods from a stockpile to the workshop).  I don't see any loss in detail here.  In fact, I'd see an increase in detail as stuff like day and night could happen believably.  The only conceivable loss of detail would be during how I propose to change movement during fast game speed, which might change (decrease) the chances of moving entities colliding with things that cross their path.

The lots and lots of work argument is not one I consider valid.  The time discrepancy issue is something that Toady will eventually have to deal with one way or another, and any solution is going to involve a lot of work.

[Rakushun]

So, Toady is currently doing a whole lot of work rewriting old code to improve the game. Hmmmmmmmmmmm. HMMMMMM. But go ahead and keep the "Toady would have to rewrite a bunch of code!" arguments coming, guys.

[10ebbor10]

So, Toady is currently doing a whole lot of work rewriting old code to improve the game. Hmmmmmmmmmmm. HMMMMMM. But go ahead and keep the "Toady would have to rewrite a bunch of code!" arguments coming, guys.

There's a difference between changing code, as happpens now, and replacing the core system of the game. This change would be equivalent to the 2d3d change or the future tile changing tthing.

As for me, I seriously doubt if the benefits would weight against the cost. A simpler system would be to add a timeslowdown mode, that would slow everything except unit mhvement.

Also, to get thex72 change you'd. Need more then just teleporting.

My argument for the foregroud background split still stands. In sslowwmode you'd see yoour dwarves, inHd so to speak and in fast mode you'd seethe fort evolve.

[knutor]

One full grown pig, butchered is supposed to feed 100 humans.  Couldn't it be said that one pig could feed more dwarfs, because of their smallness?  1 pig feeds maybe 200 dwarfs, on a full meal plan, and 400 on a starving emergency plan.  This comment brought to you by the 'DF needs Drunken Weenies society'. 

Drunken Weenies /R

QbertEnhanced, when those alterations to time go into place, don't forget Animal Trap bait spoiling.  Heck don't forget spoiling in and of itself.  Right now I believe, meat rots too fast.  It'll rot really fast when time is changed, I imagine.  If the average cave temperature is 20-40 degrees F, how's food rotting so fast, I wanna know? 

I've given up on setting animal traps manually, I just use the AI.  Haven't seen one miasma doing that, but haven't gotten many cool critters either...  Does the AI even use meatgems as bait.  I'll have to do some research on that and start another thread.  Sorry for changing subject.  Knutor

[Andeerz]

As for me, I seriously doubt if the benefits would weight against the cost. A simpler system would be to add a timeslowdown mode, that would slow everything except unit mhvement.

My suggestion is to add a time slow down mode among other things.  And when you say slow everything except unit movement, do you mean units move just as slow in the slowed-down mode as in normal (fast) mode (the same distance per unit of game time)?  Or do you mean that they would move the same distance per frame regardless of time mode? 

The former would mean that during slowed-down mode, units would move excruciatingly slow and a huge portion of the time discrepancy between adventure and dwarf modes would remain, which would remove a lot of the purpose of having a slowed-down mode.  The latter would open up humongous holes in how the game works.  Any time you wanted units to move faster, all you would have to do is switch to slowed-down mode and magically all units would move super fast (compared to normal speed).  Look at it this way: a lot of what influences an economy is how fast and efficiently resources get around (whether it be raw materials, people who do work, tools, or whatever).  In DF, this still applies.  How efficiently your fort works depends on how well you design the layout, optimizing it for efficient and fast movement of your dwarves.  So, basically, with that kind of slowed-down mode where units move the same distance per frame as normal time mode, any time one switches to slowed-down mode, everything would magically work more efficiently.  I hope I am making sense.     

Also, to get thex72 change you'd. Need more then just teleporting.

I'm pretty sure I suggested much more than that.  In any case, what is it that I would need to fix this situation?  What am I missing?  Are there problems I failed to mention already?

[Silverionmox]

One full grown pig, butchered is supposed to feed 100 humans.  Couldn't it be said that one pig could feed more dwarfs, because of their smallness?

They may be small, but they are broad and muscular, and they'll need just as much energy if they do just as much work as humans.

As for me, I seriously doubt if the benefits would weight against the cost. A simpler system would be to add a timeslowdown mode, that would slow everything except unit mhvement.

My suggestion is to add a time slow down mode among other things.  And when you say slow everything except unit movement, do you mean units move just as slow in the slowed-down mode as in normal (fast) mode (the same distance per unit of game time)?  Or do you mean that they would move the same distance per frame regardless of time mode? 

The former would mean that during slowed-down mode, units would move excruciatingly slow and a huge portion of the time discrepancy between adventure and dwarf modes would remain, which would remove a lot of the purpose of having a slowed-down mode.  The latter would open up humongous holes in how the game works.  Any time you wanted units to move faster, all you would have to do is switch to slowed-down mode and magically all units would move super fast (compared to normal speed).  Look at it this way: a lot of what influences an economy is how fast and efficiently resources get around (whether it be raw materials, people who do work, tools, or whatever).  In DF, this still applies.  How efficiently your fort works depends on how well you design the layout, optimizing it for efficient and fast movement of your dwarves.  So, basically, with that kind of slowed-down mode where units move the same distance per frame as normal time mode, any time one switches to slowed-down mode, everything would magically work more efficiently.  I hope I am making sense.

That kind of time manipulation tricks will be inevitable when you have different productivity for different modes. Luckily, there's no situation where dwarves are both economically active (and their actions need to be abstracted to stay interesting) and in combat (where minutes matter). So instead of switching the whole game to a different speed one can make a distinction between military operations (and messengers, caravans, etc: whenever timing is important) and domestic affairs (you don't need to witness real-time how every lunch of the whole year is eaten and how every stone block is cut: dwarves taking a few symbolic meals each year and still outputting the appropriate average production in a year is adequate to our needs). This means that you have dwarves in two modes at the same time, sometimes (eg. most of your fortress inside, producing, and a few squads outside battling). So one group of dwarves will move slower in comparison with the other. Since there's an upper limit to speed (a technical one and because you still need to be able to actually see what your military does), the military dwarves will then move at the speeds they do now, while the economic dwarves will move much slower. When combat is over all dwarves are into economy mode and then the game will display the economic dwarves at the speeds they have now in the game.

The advantage is that there is just one system of production. You don't need to code, essentially, another game for the other mode, and you don't need to balance both modes with each other. In addition it will focus the attention of the player on the combat. The time for preparation has passed, it's time for action.

[GreatWyrmGold]

One full grown pig, butchered is supposed to feed 100 humans.  Couldn't it be said that one pig could feed more dwarfs, because of their smallness?  1 pig feeds maybe 200 dwarfs, on a full meal plan, and 400 on a starving emergency plan.  This comment brought to you by the 'DF needs Drunken Weenies society'. 

Feed 100 humans for how long?

QbertEnhanced, when those alterations to time go into place, don't forget Animal Trap bait spoiling.  Heck don't forget spoiling in and of itself.  Right now I believe, meat rots too fast.  It'll rot really fast when time is changed, I imagine.  If the average cave temperature is 20-40 degrees F, how's food rotting so fast, I wanna know? 

What, over weeks or months or longer if properly protected? Given the lack of salting or smoking the meat as of yet, that might be too long. Unless we're currently abstracting that out, in which case it's still OK.

[QbertEnhanced]

So one group of dwarves will move slower in comparison with the other. Since there's an upper limit to speed (a technical one and because you still need to be able to actually see what your military does), the military dwarves will then move at the speeds they do now, while the economic dwarves will move much slower. When combat is over all dwarves are into economy mode and then the game will display the economic dwarves at the speeds they have now in the game.

Problem with this is that it opens HUGE opportunities for exploitation. Shifting dwarves between combat and non combat modes to get hauling done faster. An overall shift in timescale negates practically every single exploit. Time goes by FASTER in one mode than the other, but per unit of game time the same amount is accomplished. It also makes things like thirst calculations and the like make more sense. For example, if a dwarf is fighting and one week goes by in fortress time, but for said dwarf only 10 minutes, how thirsty is this fighting dwarf?
Players could keep their dwarves on fighting time to stretch out food supplies and other tricks like that.

[Silverionmox]

So one group of dwarves will move slower in comparison with the other. Since there's an upper limit to speed (a technical one and because you still need to be able to actually see what your military does), the military dwarves will then move at the speeds they do now, while the economic dwarves will move much slower. When combat is over all dwarves are into economy mode and then the game will display the economic dwarves at the speeds they have now in the game.

Problem with this is that it opens HUGE opportunities for exploitation. Shifting dwarves between combat and non combat modes to get hauling done faster. An overall shift in timescale negates practically every single exploit. Time goes by FASTER in one mode than the other, but per unit of game time the same amount is accomplished. It also makes things like thirst calculations and the like make more sense. For example, if a dwarf is fighting and one week goes by in fortress time, but for said dwarf only 10 minutes, how thirsty is this fighting dwarf?
Players could keep their dwarves on fighting time to stretch out food supplies and other tricks like that.

For completeness: you can't switch at will. Dwarves have to be in combat/threatened to move fast - therefore, a working dwarf is always in economy mode. (Potentially, you could scare haulers at the right time which would then flee with their stuff: that can easily be adressed by making haulers drop their load when they flee (which makes sense anyway)).

Whenever a dwarf is in combat, the economy is at a snail's pace: therefore, dwarves will always eat x units of food per year, whether they are fighting or working. Combat mode just has more ticks between meals etc. to move and fight.

[Manveru Taurënér]

Having 2 speeds to change in between either manually or automatic in the same game mode would just overly complicate and hurt the general game experience in my opinion. I'm all for some minor tweaking of the current speed, but if I had to keep changing speeds back and forth like it's the sims or something I probably wouldn't be interested in playing any more :>

[Silverionmox]

Having 2 speeds to change in between either manually or automatic in the same game mode would just overly complicate and hurt the general game experience in my opinion. I'm all for some minor tweaking of the current speed, but if I had to keep changing speeds back and forth like it's the sims or something I probably wouldn't be interested in playing any more :>

How would you solve the coordination problems between the fortress and the outside world?

[Manveru Taurënér]

How would you solve the coordination problems between the fortress and the outside world?

Well, one could have them get a speedup (and temporary suspension of hunger/thirst/sleeping) when heading off-map as well as for the start of their journey after leaving the edge of the map, to make up for the loss of time travelling to it. There'd also need to bee good ai for organizing all the supplies needed for they journey quickly. I think that way you could get it all down to acceptable levels. It'd also have to be made so you couldn't exploit it by making cancelling these missions hard or even impossible as long as an ambush/force of darkness/megabeast or similar appeared while they were leaving.

Edit: Also along with some minor slowing down of the current fort speed.

[10ebbor10]

Having 2 speeds to change in between either manually or automatic in the same game mode would just overly complicate and hurt the general game experience in my opinion. I'm all for some minor tweaking of the current speed, but if I had to keep changing speeds back and forth like it's the sims or something I probably wouldn't be interested in playing any more :>

How would you solve the coordination problems between the fortress and the outside world?

As said said before, not.

You just stick to the current fortress timescale and stay with it. You could make full moons last longer, have week long day and nights, and it would all come toghether. As for it taking several weeks for armies to move off the map, note that it's an abstraction. What would really happen is that they prepare and stuff for several weeks, and then leave, however, this is not shown in fortress mode