Year 4 underway in Watermined.
A tweak to the repeater has the bridge trap (nicknamed the "Ko-boulder") and also the Forest of Spikes working better*. Only new victims of the former are a dwarven caravan guard and two pack animals who attempted to leave the fortress that way during testing.
The corpses of the horse and water buffalo have been left there to add to the items being tossed around.
Only other noticed intruder was a kobold sneaking in while the repeater was down. He was spotted by a guard dog as he passed a door into a chamber in the soil area. Which also happens to be where the main fortress guard elite melee unit trains...
Sword, axe, and spear.
The roof sections of the first two levels of the Keep were completed. These areas are now being set up as barracks for the militia to train in. The third level and higher levels of the tower are building more slowly due to quartzite block shortages and making lead blocks for the third level roof. Other outer defenses have recently been completed including three archery bunkers with underground access that provide coverage of the northern and eastern approaches.
Once the Keep is completed the next construction project will be bridging the river and working on the south part of the embark. A north-south road will be added as well.
Miners have found a fresh vein of gold near the Magma Sea. Small clusters of gold and platinum have also been found up in the mudstone where iron ore is being mined.
* - The repeater is not a compact design. A 86-tile loop with a dozen or so pressure pads on it which trigger bridges and/or spike traps. Initial "working" design had two impulse ramps pushing a cart around. Cart would eventually reach a speed whereas it would run the loop too fast for the pads to reset. After a while, some sort of oscillation behavior would kick in allowing some resets. (My guess is cart "skipping" a pad space and thus allowing it to reset.)
The latest revision was shifting the impulse ramps to roughly opposite halves of the loop; and more importantly ramping up the cart at the end of the loop and smacking it into a wall to slow it and then let it run back down a ramp into the loop at the same point the cart is initially launched. This appears to have the cart running a loop in approximately 180-190 ticks. And the pads appear to be resetting properly each loop.
Still a few questions being raised by this:
1. I looked at the Wiki data on minecarts and the behavior I am seeing (time spent doing loop, acceleration/deceleration effects, etc.) don't seem to jive with the information there. So that will warrant further investigation and observation.
2. Bridge triggered is supposed to take about 200 ticks to reset, and a spike trigger about 140 ticks. Given the <200 tick loop right now I should see spikes operating normally and bridges perhaps having some issues. But I don't think this is happening. Therefore, perhaps a "feature" that a pad rigged to trigger both a bridge and a spike will reset both on the faster spike schedule? This will get some further observation as well, and perhaps some explicit separate testing as well.