Bay 12 Games Forum

I do know about the paintbrush that DF provides, but it seems to me to be very clunky.  It won't produce natural looking results like a peoper graphics editor could.  I don't have layers or templates or curve editors or whatever... easier to ask for an import/export mechanism than all that stuff, right?

My suggestion for the next (or near future) release of Dwarf fortress is:

Have rivers not run uphill... not cut through sheer cliff faces because they can't run up-hill.  I know it would take more processor power to more accurately simulate where the rivers are going, but it's really only once that they need to be generated... I've grown very tired of seeing 15z sheer canyon walls on EVERY river, stream, and brook that goes through the high drainage areas on my maps.

It seems to me the algorithm used for river flows is similar to the one used for drainage: Take a 3x3 grid from the region map, Find the high area and the low area, then run a river in the center area that connects those regions.  It works well enough, but when you're dealing with a map that is not just 16x but (16x48=) 768x more complex than that, it just doesn't look right anymore.  And you get the weird rivers that seem to plow right through anything they touch, uphill or not.

Instead, what if DF took a moment and generated a height map for the top Z layer AS IT WOULD BE SEEN IN ADVENTURE/FORTRESS mode, then ran the rivers downhill on that (ONLY ever downhill or straight ahead, never up-hill even one Z, please!).  Then at the bottom of drainage pits, it could form lakes (as it does now).  Once the lake fills up to a certain point, it could spill over and start flowing downhill again.  This might cause rivers to "back up" and widen out, too... which would be very awesome for lakes to do.

Then once the hydrosphere is set up, the erosion calculations could be computed... and then back to running the rivers downhill again.  This could start over from brooks only, or maybe river sources only.  The water would flow downhill again, and this time take a slightly different path due to the terrain being eroded a bit.  The key here is there should be a maximum of erosion possible.  Rivers should NEVER be able to cut into hard rocks.  It should be computationally trivial to figure out how much sediment is there.

Anyways, that's my 2 bits.  But it seems, Oh Toady One... that you are very much into depth of realism for the worlds your program creates, and in my opinion the hydrological systems need quite a bit of work in the world-gen aspect.  Thank you for your time.

What you're missing, Deon, is that Anvils are usually made out of a material that gives them that kind of value.  One made of useless Chert might not have such a high cost.  In addition, the cost jump is minimal from 60 to 100.  Plus the amount of time invested in moving 20 rocks there... I see this as actually somewhat realistic.  I see the mason sitting there, cracking open rock after rock, only to have it split the wrong way due to the natural crystalline faults in the rock, and having to throw that one out.  Only after going through 20 rocks does the mason find one that splits right, and has the right structural integrity to beat out metal works upon.

Tho you might make it use only magma-safe stone.  That way it could stand up to the heat it's going to be subjected to without making your new shiny iron all... magma sticky...

There is a workaround for that.

Open back up Stonesense and close it again... DF will unfreeze