Hacker News new | ask | show | jobs
by HourglassFR 1955 days ago
> If people realized that a sudden drop in price will mean a sudden drop in mining power, which will mean a sudden drop in blocks being generated, which will mean a sudden drop in throughput, they might be even more eager to get out of it.

I agree with your overall argument but as I understand it, this is false. If the mining power of the whole network drops, the proof of work will become simpler so as to have a roughly constant block throughput. Or am I missing something here ?

1 comments

At one point this was approximately true; I'm honestly not sure if it still is.

The problem was how the difficulty was decided -- it looked at the last N blocks that were successfully mined to decide whether it needed to up or down the difficulty. If you suddenly lost 90% of your mining capacity, it would only adjust the difficulty after several more blocks were mined, each of which would take ~100 minutes instead of the usual ~10. The system would eventually right itself, but it wasn't fast.

I can't recall if this ever became a problem in practice, or whether there was a code change to make it resolve faster.