They have been around from the beginning, and I never got around to the local variations in the landscape that I'd need to get a better distribution of differently-sized small features. We might see something on this with underbrush (which is up in the first category on the dev page with the other things I've been working on), since I wanted to mix tree/brush density up within a given world map square, though soil information might have to come first for that to be satisfying. Another obstacle is having new moist tiles or tiles with some water without having them having floors on a different Z level. Brooks are currently handled that way. I don't like the idea of having 3/7 water a Z level down, although a lot of the streams should probably connect up with the aquifer. This would make fewer places inhabitable by digging creatures unless the stream were smoothly lowered a bit, but I'm not for turning them into rivers with less water in them since that introduces problems with connecting rivers (would all 3/7 rivers connect to 7/7 rivers in waterfalls? Otherwise you'd have reverse waterfalls, or you'd have to introduce a 4/7 ground square, which feels like a can of worms) and people walking across a brook should not be in a separate Z level with respect to projectiles/LOS, etc. People swimming on the surface of a river should probably also be vulnerable to archers, but having an ankle-high brook hide you is worse.
Hey Toady, wouldn't it be physically accurate to keep the brook surface tile and make the "body" of the brook a localized aquifer rather than 7/7 water? Units can still walk on it, it's still a source of water, nothing can swim or live in it, and the best part is that there'd be no flow calculations so brook sites would be great for FPS.
Random revisiting, perhaps, but upon reflection, I really like this solution.
Digging into a "local aquifer" would produce a tile that constantly fills up with water (that can be diverted and fed into cisterns), but perhaps not at a terrifically fast rate. Digging through a brook tile that is "only enough tile to get your boot wet" should still create a pit that EVENTUALLY will fill up with water.
It might need some tweaking of the aquifer code, however, so that there is only so much water that the entire "aquifer" can produce per unit time even if you dig into it, and that it still has an "upriver" section that has priority for filling with water before the "downriver" sections fill up with water.
Silverionmox's comment about ramps only allowing 4/7 or 3/7 water also makes perfect sense, of course.
Perhaps having coding with brooks (and rivers) that make areas have Spring and Summer floods when brooks are based off of mountains, when the snowpack melts annually would also make for some lovely verisimilartude.