If there are some accounts with substancial amounts of NXT which are nobody's property and have weaker passwords than maybe we could create special thread to organise group cracking.
I'm sure a lot of people would like to use script/program to make such group brute force treasure hunting
It could be constructed in similar way as altcoins pool mining with system of distribution f.e. lucky one would get half of found NXTs and the rest of could get proportional part of other half. Also it could be a simple solo treasure hunting.
This would be great for increasing NXT community and would be simply fun. In future because of growing amounts of NXT's lost in void it would become even more successful. What do you think about it?
I'm sure a lot of people would like to use script/program to make such group brute force treasure hunting
It could be constructed in similar way as altcoins pool mining with system of distribution f.e. lucky one would get half of found NXTs and the rest of could get proportional part of other half. Also it could be a simple solo treasure hunting.
This would be great for increasing NXT community and would be simply fun. In future because of growing amounts of NXT's lost in void it would become even more successful. What do you think about it?
I thought about this "darkNXT mining" recently. The problem is that to be successful, a somewhat large number of "miners" needed, and they need to be committed, i.e. no to abandon mining after a week or so without payouts.
Let's do some math. Complexity to "break" an account without a public key is 2^64, or 2^63 on average. According to earlier replies in this thread, there are about 1'000 accounts without public key, or about 2^10. Thus, on 2^53 work on average until at least one account is "broken".
The speed limiting factor is generating of public key from a (random) private key, the reasonable (yet optimistic) estimate is about 8'000-10'000 per second per CPU core, or 2^13.2 per second per CPU core. For 2^53 work this translates to 2^39.8 CPUcore-seconds, or about 110 million CPUcore-days / 300'000 CPU-core-years.
If we are very optimistic and assume that all miners are quad-core and that it is possible to 8x accelerate code by using e.g. AVX, then we'll be looking at "only" about a 3'500'000 Quad-core CPU day / 9'400 Quad-core CPU-years. GPUs won't reduce that figure more than tenfold (and that is *very* optimistic assumption).
So I am not convinced we have enough computing power to reliably mine darkNXT at this moment... To unlock one account per month (with current software) we'll need about 1 million "miners". For highly-optimized (and yet-to-be-seen) software that number drops to about 120'000, which is still pretty high number.