On
Eternal Suggestion Voting.
This is one of those ideas that I recognize is a very big change that could break a few systems, and probably saves. Nevertheless, I think that, in a game that wants to tout its realism, this would be a very good way of making a major leap towards that goal.
People frequently make fun of how we currently deal with items in small integer units - one malachite smelts to one copper crafts into one earring or a pair of low boots or other metal object. One wood makes one bed or barrel or bracelet. One stone can make a door or a table or an entire workshop or a couple stone mugs.
There have been several suggestions that are intermediate solutions, but I think that what we really need is to simply go the whole distance, and make the realistic answer - a real system of volume and mass. (Also, by extension, density. In fact, I would expect density to be the prime number recorded for each material, and have individual materials requirements call for either volume or mass - so that wooden furniture calls for a certain volume of wood, regardless of the density of the wood, while others, like charcoal, may call for a certain mass of wood.)
This means that we do not deal with whole units of wood, but rather with X number of kilograms of wood. It takes 10 kgs of wood to make 4 kgs of charcoal. It takes one cubic meter of wood to make a desk (still abstracted to just volume alone.) This would mean that a copper earring, at only .1 kg of mass, would only use up .1 kg of copper (possibly more, if you want to include waste and a sort of entropic decay of matter), while a 6 kg helmet takes up 6 kg of copper.
This also has some major stacking ramifications - you can make it so that when you dump 12 kg of iron onto a pile of 38 kg of iron, it creates a single 50 kg iron pile.
Even better, you can start making items from more than one material - a pick or a spear, for example, is currently entirely made of metal. Now, you could make picks with a wooden handle (that only takes a small fraction of the wood you get from chopping down a tree), and make only the pick itself from metal.
Now, when we are talking about volumes, however, we can finally start addressing the volume of a single cubic tile. I would estimate that a single cubic tile is 10' tall, since that is the distance between floors of a standard building for us humans. Dwarves may make them shorter theoretically, but when we are dealing with stone caves, a little extra ceiling room for buttressing may be a good idea. Since everything in the game seems to presume we are dealing with cubes, that would make it a 10' x 10' x 10' cube, for 1000 cubic feet. If, however, we are going with metric, that is fairly close to 3m x 3m x 3m for 27 cubic meters of volume to fill.
This can also have other ramifications - currently, mining has recieved complaints because you essentially just hit a wall with a pick to vaporize a wall, possibly leaving behind a single stone. Now, however, we could simply use that to fracture a wall, and leave behind 27 cubic meters of mullock/rubble material that needs to be hauled off, or, if the mining check was successful, would leave behind X amount of usable stone, and some leftover cubic meters of rubble/mullock.
This would also mean that once any tile hits 27 cubic meters of material, it would become impassible. (Something like 20 cubic meters of material might also just become a ramp.) For the purposes of pathing, it might be OK to let creatures push aside loose material that blocks their path, but basically, there has to be enough empty volume in a tile for a creature to move through it.
This could also mean that, instead of leaving exactly one random item in a stockpile, you could actually start stacking those items, like in a warehouse, potentially to the point where you might create "walls" of material, and have stairs climbing up on top of the stored materials in your warehouse.
This could potentially be combined with other suggestions, like the
sand suggestion, and even the fluid models - rather than a 0-7 level of fluid, we would have 0-27 cubic meters (1 cubic meter = 1,000 liters, by the way) of fluid. This fluid, rather than flowing in random directions, could simply divide its flow out evenly across all adjacent tiles with a lower total fluid level than itself. Solid objects, of course, would be capable of displacing fluids (provided it was not melting into magma, in which case, it should probably increase the total volume of magma when it melts, as well), so a 5 cubic meter stone would mean there were 5 less cubic meters of space for a fluid to occupy.
Finally, as if there wasn't enough to tie all this into, this could finally be used to create that realistic cave-in system that the game has been lacking, making it the butt of several jokes, such as the "support an entire fortress on a single pillar made of soap" joke. Although it would take much more work than simply implimenting the overall mass and volume suggestion, one can create a system where every non-wall tile that has a floor or wall above it must have a ceiling above it to bear the weight of the tiles above (presumably, this may include some smaller supports, if need be), with the mass of the areas above requiring enough supporting material below to avoid a sort of warning "sagging" or an outright cave-in. When a floor is supported off of a wall, that wall must be able to bear the mass of the floors, plus all the walls (and the mass it supports above it), or will start to fail. Walls should theoretically be able to pass their burden diagonally downward, so as to mean that simply building thicker walls eventually solves the problem.
This will probably require some form of interface change to be able to enable such a system. A seperate look-menu option on walls or the like could give tile-specific information, but it would probably be best to have a special view mode where the colors were changed so that you could see the amount of stress each part of the overall structure was bearing. (So that walls and floors would turn colors from dark blue to bright red to show the amount of stress they are bearing.)
To combat this, we could start getting into serious engineering solutions to the problems of cave-ins, from simply digging narrow tunnels that do not stack multiple floors overtop one another, to the classical masonry solutions of simply making thicker walls near the base of structures, growing narrower at the top (possibly using flying buttresses or the like, if it were a free-standing stone structure), to using more presumably modern (but still theoretically technologically available for creatures of that scientific era) concrete with steel rebar supports to buttress very large structures. This would require a reworking of the way that dwarves are able to address constructions or the like, as they would need a way to create "supports" that are not actually built structures that take up a tile, but simply slapped on as an addition to a wall.
Again, I recognize these suggestions, especially some of the last parts, are fairly massive, but I think that for the long-term life of DF, a full volume and mass system would help to solve quite a few of the niggling inaccuracies of the game.