The client chunks BKS outputs into 1 BKS chunks for voting. As a result, moving BKS often takes a lot of space on the blockchain. The current 0.01 BKS fee makes this process very expensive. Do we accept this and say that we are not going to cater to people moving BKS around a lot, or should we lower the fee?

I am in favor of a 0.001 BKS fee, which is equal to the minimum dust limit I believe.

If BKC-BKS are to become a dual coin pair whose value will flow back and forth between the two coins often to support each other, just like the NBT-NSR pairs have evolved to be, then the fee should be lowered greately to allow low friction value transfer.

However marketcap and liquidity have a life-blood significance for a currency such as Nubits. For Blockcredits liquidity isn’t as important. So the similarity between the two pairs can be debated.

I agree, we don’t necessarily have a reason to expect people will use BKS as a trading token like we do NSR; it could be considered as primarily voting tokens and open market trading is not encouraged like it is with NSR. Also, BKC would be primarily a consumable resource for the exchange, so open market trading of that is somewhat discouraged as well when compared with NBT. Therefore, B&C can afford to have much higher fees than Nu.

In terms of auctions, if B&C wants to do a share buyback it can do so in big events rather than the continuous burn gateways of Nu. That way, the volume is high and Tx fees won’t become a burden.

So, how much friction do we want to impose upon BKS Txs? As the chunking creates fixed-size outputs, we can estimate the blockchain size/BKS, and thus the % cost. If 1 input costs 200 B, there are ~5 inputs per 1 KB. So, the fee gets charged once every 5 BKS. A 0.01 fee translates into ~0.2% fee.

This calculation for Nu:
10,000 NSR/input
fee gets charged once every 50,000 NSR
If fee is 1 NSR, Cost = 0.002%