Progress report! Fortress development nil; advances in !!SCIENCE!!
Turning collapse messages off doesn't help address the playability of the fortress. I think that the lake and the volcano will both be drained before the volcano gets capped. However, an alternative, turning cave-ins themselves off, will cap the volcano. To the !!SCIENCE!! behind this.
The reason why I saw cheating (DF Hack in particular) as a necessity is because of how Dwarf Fortress handles
vertical flows. Everything above the volcano does not move at once. Rather, this happens:
This screenshot was taken T+2 frames after the initial the embark. T+1 frame was the lake rushing to fill the void above the volcano. Where the walkable path around it is.
While cave-ins are disabled, all those tiles are steam and obsidian walls. If I were to start up with it enabled, everything that isn't fastened to a wall will fall. For lack of a better term, a 'tube' will be created along the paths of the falling obsidian walls, filled with steam, lava mist, and other stuff. When these 'tubes' evaporate, the magma inside the volcano will rush to fill it. As these tubes are several Z-levels in height, the volcano will be draining itself to fill these voids. Which, by the way, makes for an incredibly fast way to drain a volcano, should one wish to colonize one.
Overall, this means that the inside of the volcano will rapidly drain while the outermost parts (at the top) will obsidianize. After that, it's chaos. I've let DF run through its course twice to let it 'naturally' solve itself, but both attempts resulted in a crash-to-desktop, I guess because DF couldn't handle the number of processes begging for attention. After only three or four in-game days passing, the volcano's middle area is hollow (across at least
50 Z-levels), while the sides are filled with flowing magma.
Results look like this:
The dilemma of capping the volcano for a playable fort really shines there, I think. A bit of in-game time had elapsed between those two screenshots. That means that without taking immediate action (i.e. disabling caveins), the fort will be slow as cold molasses in January.
An off-shoot about that picture, specifically the tiles that haven't been obsidianized yet. I'm not sure if others have observed flows or not, but I haven't seen anything on the DF Wiki about it.
Anyway, looking at what's been obsidianized and what hasn't, you can see that there's definitely a pattern. Take that snake line's shape, copy it across the map. For a little amount of time, that flow will be moving vertically. It will switch between north-to-south and south-to-north. Then the flow map will rotate 90 degrees after either enough frames passing or enough flow-direction switches. Then the flow will be moving west-to-east or east-to-west. And swap to vertical flowing again. I hypothesize that all movement (that utilizes the flow algorithm) is dependent on these flow maps, which is why miners have such bizarre digging paths, and why giant water flows are so FPS draining (DF has to deal with evaporation and all 8 flow depth maps). Flow movements can be really easy to spot en masse, which this map can demonstrate on the top of the lake and caveins turned off.
This all follows suit with observations I've had with volcanoes and with flows in previous versions/forts in Dwarf Fortress.
!!SCIENCE!! For profit, pleasure, and the sake of !!SCIENCE!!