I was talking about a player-owned shop. If a faction has a single shop served and stocked by 10 members and used by other players in the server, the stock of credits in that shop can grow quickly. It's an unlikely event, but a possibility (near spawn to accommodate newb traffic).
I had excluded player shops from the equation, because I was trying to construct a worst-case scenario, which would mean ALL players on a server are working towards getting the oldest player to reach the cap. Player shops would have no benefit in that environment.
Yes, my calculations were for a single player, but multiple players working together can be simulated by just dividing the result accordingly.
Custom configs were excluded because their effects are easily inferable, once the result for default configs is known. (e.g. multiplying item prices by 10 will make players hit the cap 10 times faster, same for mining multipliers)
Voting boni are a constant non-changing source of income, and thus the graph can easily be adjusted for them.
Unless you can point out a believable set of variables for the guess, that would make the estimated time drop rather low, I'd say the most foolproof way to determine whether or not a 64bit integer is enough would be an experiment similar to this:
- Empty-server/single-player, default configs, brand new universe
- Note time of the beginning of the experiment
- Reach the current limit as quickly as possible as a single player[not total credit+item value, only the credit counter counts]
- Note time
- Clear credits[set counter to 0]
- Repeat from step 3 on, until the day is over, or the tester is too bored to continue. (however, at least 10 loops should be done)
- Repeat from step 1 on, once the tester recovered, or use a similar enough other tester
- Plot a chart, and determine growth behaviour and growth rate