So if I understand you, it will happen like this.

Every safecoin has a unique address. A safecoin address is an unsigned long integer which means it’s unique address must be between 0 and 4294967295. There’s where we get the 4.3 billion.

The network attempts to generate a new safecoin, and randomly picks a number between 0 and 4294967295 for it’s address. Presumably it then checks if that address has already been used. If it has not been used, the new owner is given the coin. If it has been used, the potential new owner loses the lottery, and gets no coin.

So we’ll get to a point when there are no more safecoin addresses left. The network, presumably, will keep trying to give people safecoin, but every time it picks a random number, it will find that it’s already been used.

So, the network will try forever, to give people safecoin, but at some point, it will always fail, because all addresses are already spoken for. At least, that’s what would happen if nobody ever spent safecoin, but of course they will. They will spend them to store data on the network. When safecoins are spent, they are recycled by the network and become available again to be awarded to another farmer if their address is selected as stated above. Thus there will always be available safecoin to farm, as safecoin are always being spent to store new data.

If I wanted to map out how many safecoins had been awarded, I could do it by pinging all 4294967295 addresses. However, even if the network responded to each ping in 20msec, that would only be 50 pings per second, which means it would take me 2.72 years to ask about every address. That means we need to select some small subset, such as 1,000,000 addresses, and use that as a statistical sampling. That could be done in 5.6 hours.

That makes sense. I just needed to think it through. Thanks for your response.