asdpc-2024-issuance: add first draft

[aesedepece]

TL;DR: With wit/2 and PoS, halvings are no good. Current block time slows down the oracle. I propose removing halvings, and instead enact a reduction of the block time by half, with an adjustment of the block rewards in the same proportion.

[drcpu-github]

So combined with the epoch period reduction, this essentially preserves the same 175M inflation per year? I'm not exactly sure what the plan is here? If we assume about 300M WIT will be staked, that's an APR of 58% and even if over half of all coins are staked (~700M), that's still a 25% APR. That seems somewhat exaggerated to me.

[aesedepece]

Yes this is one big point of discussion. I'm for introducing a real halving here, to drop annual inflation around ~7%, but I have a conflict of interest here so I'm hesitant to be the one championing that unless the masses are roaring for it.

[aesedepece]

@guidiaz @Tommytrg @parodyBit what do you think?

[drcpu-github]

My proposal would be a 20s block time and 50 wit reward per block.

[guidiaz]

My proposal would be a 20s block time and 50 wit reward per block.

You mean like taking the opportunity of launching PoS as to implement to some extent the halving that it was scheduled whatsoever for Oct. 25th this year on current version? And the reason why you propose to reduce yearly inflation to a 45% current value (78.84M vs 175.2M) instead of 50% comes due to the intention of preserving integer values for both the block time (20") and the block average (50 WIT)?

[drcpu-github]

The halvening on the current version of the network is actually still more than a year away.

The reason why I think a reduction in block reward makes sense is that it will become much more profitable to run a node as soon as you stake a (yet-to-be-seen) amount of WIT. The reason for that is that at the moment you need to run x nodes to utilize the same amount of WIT in an optimal way which will cost significantly more in terms of hardware costs versus running a single node with a high stake amount. Hence, we would (significantly) overpay to achieve the same level of security of the network compared to now. I think we can draw an easy parallel here with how ETH rewards were reduced vastly when they moved to PoS and I don't see why we should not follow the same paradigm.

Above are some variables I used to arrive at the proposed block reward value. I think everything above 15% APR is overpaying for security compared to other networks and it will just weaken the overall economic security: excess tokens get dumped which leads to a lower coin price which leads to lower dollar value staked on validators, etc.

As far as the actual numbers are concerned, yes, both of those are picked for the sake of easy-to-calculate integer values.

[guidiaz]

I agree on the idea that a high yearly inflation rate leads to "excess tokens (that) get dump which leads to a lower coin price which leads to lower dollar value staked on validators".

I don't agree with the reasoning that a reduction in block rewards makes running a node much more profitable, though. On the contrary, the less yearly inflation the less profit (in terms of $WIT, ofc). The improvement that you describe concerning node operators being able to achieve the same level of network security by running less nodes is true, but it comes because of the new PoS mechanism, and not because a reduction in block reward is being proposed.

Nevertheless, I do agree that:

• reducing inflation to anything around 50% would be appropriate and pretty convenient to all stakeholders (at least in the long term);
• it is the Wit/2 upgrade launch the right and only moment to introduce a reduction in the block time (good for developers, supported dapps and data requesters) and an overall reduction of the yearly inflation rate (good for miners/validators, the overall network security and the tokenomics).
• the final block reward should be fine-tuned before release of the Wit/2 upgrade on the Witnet mainnet depending on the minimum block time possible reached during the beta testing of a Wit/2 upgraded testnet.

[drcpu-github]

I don't agree with the reasoning that a reduction in block rewards makes running a node much more profitable, though. On the contrary, the less yearly inflation the less profit (in terms of $WIT, ofc). The improvement that you describe concerning node operators being able to achieve the same level of network security by running less nodes is true, but it comes because of the new PoS mechanism, and not because a reduction in block reward is being proposed.

You misunderstood me, I was not implying that a reduction in rewards will result in an increase in profitability. I was simply saying that the release of wit\2 will make running a node much more profitable because you will need much less hardware to optimally use a fixed amount of WIT compared to the current state. Hence, a reduction makes sense (for many more reasons).

[drcpu-github]

A difficulty bomb like Ethereum PoW? How are you planning to enforce this besides socially?

[aesedepece]

Only socially

[guidiaz]

I guess that "Only socially" obliges that should there be a new WIP intended to reduce the long-term inflation, it should propose ways to accomplish such a thing, but without affecting validators short-term APY.