Shwart!! wrote:[...]we could use a fairly complex redstone system I devised that automatically detects fires and puts up firebreaks to combat it. This has the drawback of requiring hundreds of sticky pistons per 40*40 area, so it would be over 28 thousand sticky pistons for full coverage, not to mention the similar amounts of redstone and 14 thousand water blocks it would take. And it isn't even 100% effective.
I bet this can be optimised. To the batmob-- uh, creative flatland! (Mind sharing your current design?)
Shwart!! wrote:It seems that deserts cannot have thunderstorms, so if we found one large enough, it would prevent lightning from being a concern.
As fire pretty much exclusively travels upwards, thunderstorms shouldn't become a huge issue either way.
EDIT: I tried a very naïve approach: bands of water sources on the inside of the structure:
In most cases, this is extremely effective at stopping accidental ignition at the foot of the structure. This is what happened on average:
Slightly worse is if there is accidental ignition on a layer without a water strip, glass indicating burnt-down wool (still very manageable in my humble opinion):
The approach even subdues deliberate ignition fairly well. This is what happened when the entire bottom row was ignited:
Worst case is ignition of the entire second-lowest row:
Even with deliberate, knowledge-based griefing like the above, the fire is handled a lot better than without any countermeasures, as seen in this example with no protection:
I still need to test multiple times with a full 40x40 wall, though for now I would say this is at least an "acceptable" level of risk management. I would not be surprised to see an increase in effectiveness when applied on a larger scale.
This approach can most probably be improved further.