Englor said:
The variants of the direct resources;
hammers
commerce/gold
food
Resource abstraction is somewhat of a pet peeve of mine. How is it you can build STONE pyramids faster by chopping a WOOD forest? It makes no sense to me. I say you can't work on a stone resource at all unless you have stone, but then, stone should be more widely available than one tile per continent. Same with a lot of other buildings: most should require wood, not that you need to chop a forest down to make them, but a forest has to BE there, somewhere in the fat cross, to make wood buildings. For marble buildings, you need marble. No marble, no building. For more modern buildings you just need concrete which is a generic enough resource that nothing special should be required in the game.
I also like the direction Civ's gone in when requiring certain resources for certain units, say, iron for swordsmen, but I think even that can be improved upon by splitting out the production of arms FOR units, from the manpower levees to fill the units, and the time required to train them. For example, your city should be able to produce, say, a unit's worth of swords, shields, etc., fairly quickly, and have them in stock somewhere. But to man the unit should be a hit on the population similar to slave-whipping, regardless of what government civic you're in. Just because you have seven iron tiles in your fat cross doesn't mean your size 7 city can blast out unlimited numbers of swordsmen!!! It takes, not just the iron for the swords, but also the people wielding them, which in history has often been the most tricky consideration to plan for.
Most countries will have the iron, etc., required for arming an army, but not all countries will have the population needed to MAN a large army. It's this population depletion that often crippled countries, in historic conflicts, or fed into their ability to expand militarily: in the U.S., for example, high population allowed the American cavalry to constantly refill with new recruits, due to agriculture supporting that high population, while the hunter-gatherer or low-tech farming of the native tribes didn't support that level of population, so while each side had rifles at a certain point, the Americans had more people pointing the rifles, and won by numbers.
That's another point to consider: acquisition of weapons by means other than discovering its tech. If a high-tech unit gets defeated (see "spear defeats a tank" hehe), not all of its weaponry will be unusable, and that's often how weapons got acquired in historic conflicts: by capture rather than manufacture.
Bottom line, weapons should be a game entity separate from units, built in city production, retained in a city armory, assigned to soldiers recruited into units (trained or untrained by the barracks--require time if trained, or require no time if untrained), and quantities of these weapons get acquired by the enemy if a unit loses in combat.
This would allow for some interesting strategies: if you produce a lot of weapons and warehouse them, your people can be productive on working tiles in the fat cross, but if there's an invasion, you could call up a levee of untrained militia to augment the "professional" units for which there was a training time investment and a maintenance cost paid, turn by turn. High population cities can field large numbers of this augmentee type of unit, while low population cities are unable to--so here we'd have an incentive to build up city size beyond just "how many tiles can they work". And after a militia disbands, they go back to the city to work tiles (if they survived the combat--which brings up another thing, "healing" a unit isn't as realistic as it would be to "reinforce" a unit with new recruits, from a city capable of training the new members of the unit lost through combat. You can't do that at all in foreign terrain! Nor in a newly-captured city!)