Erik,

I'm curious about what you believe to be "non-economic" txs. As far as I can tell, any transaction included in the blockchain is economically motivated by the very evidence of fees paid. That said, for the sake of argument if we assume that there exists a category of information that constitutes "non-economic" information, then so long as there is any variance in the way to express a single economic intention, there exists a vector for including "non-economic" information. I'll add beyond this that there must always be variance in the way to express the same intent because the signature data must be indistinguishable from entropy for Bitcoin's security to hold.

Even if we eliminate small UTXOs, OP_RETURN, or whatever other vector of the day that is currently being used to propagate such "non-economic" information, we will always have the potential variance in the signature data to do so. The best you can hope for is to make such means so inefficient that the real cost-per-bit is expensive enough that there are fewer distinct use cases. However, this isn't enough to actually prevent the "spam". By increasing the cost-per-bit, it may limit it to only "non-economic" information of extremely high value (note the contradiction), it limits the number of use cases while also increasing the impact of the use cases that make it past that threshold. Thus, it isn't the impact of spam that is being reduced so much as it is reducing the number of distinct use cases that result in "spam". Perhaps this is enough to make spam more intermittent, and maybe on those grounds alone it could be worth it, but I doubt it.

IMO the proper way to handle things like this isn't to introduce consensus or relay policy to incentivize the expansion of the chain weight these "non-economic" use cases require, but rather to reduce the necessary chain footprint of supposed "economically motivated" transactions, which incidentally is the entire point of all layered scaling tech. The current fees we are experiencing are still significantly lower than they need to be if Bitcoin is going to survive in a post-subsidy era. If our layered protocols can't survive the current fee environment, the answer is to fix the layered protocols.

Food for thought.

Stay Inspired,
Keags

On Tue, May 9, 2023 at 12:38 PM Erik Aronesty via bitcoin-dev <bitcoin-dev@lists.linuxfoundation.org> wrote:

> no data at all

exactly, which is why a relationship between "cpfp-inclusive outputs" and "fees" makes sense.   it's clear that's a good definition of dust, and not too hard to get a working pr up for the network-layer.   i get that your node will still route.   i get that it would break timestamps, indeed, it would break all non-economic use cases if we made it a consensus change.

but that's the point of the discussion.   

the question is whether breaking all non-economic use cases is the right move, given the game-theory of what underpins bitcoin

i'm sad (honestly) to say that it might be

it may very well be that bitcoin *cannot* be a "global ledger of all things" in order to remain useful and decentralized, and instead the monetary use case must be it's only goal
 
also, i'm not really advocating for this solution so much as i would like a 

- rational conversation about the incentives 
- whether this solution would be an effective enough barrier to keep most non-economic tx off bitcoin

obviously it's easy enough to evade if every non-economic user simply keeps enough bitcoin around and sends it back to himself

so maybe it's a useless idea?   but maybe that's enough of a hassle to stop people (it certainly breaks ordinals, since it can never be 1 sat)

_______________________________________________
bitcoin-dev mailing list
bitcoin-dev@lists.linuxfoundation.org
https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev