Right now non-standard transactions compete for block space only within mining pools that allow out-of-band payments, which naturally makes them less reliable and more expensive. This makes non-standard transactions economically disincentivized compared to standard transactions.
So filters don't "work" in the sense that they prevent all spam. But they work in the sense that they disincentivize them.
Well summarized 💯
Incentives all the way down.
1wk timeframe under Pools Ranking.
Those that enable OOB are profiting more.
Not even OOB, it's paid in fees.
Non-standard**
Yes, for those sending spam. Plus they risk being orphaned in the future.
Who will orphan them? Foundry? Antpool?
Only very very very very marginally. Ultimately if they don’t get mined this block that just means there’s more of them (and fewer normal transactions) for the next miner that isn’t filtering. Over the course of a few blocks it averages out.
It only actually works if you have more miners blocking spam than the portion of spam transactions.
I don't think double the regular fee is marginal. And delaying non-standard transactions means that the fees for the non-standard transactions goes up. Again, creating more disincentive. I don't understand how that would "average out." Enough delayed non-standard transactions mean more competition for the same block space creating more fee pressure.
It’s double the regular fee if you use a bot run by someone taking a cut before MARA takes a cut with Slipstream….Libre Relay should get you mined with 1 sat/vB no problem (used to be F2Pool ran that, no idea what they do now but generally they’ll mine anything).
Your assumption about delays driving fees up only holds if those transactors are incredibly time-sensitive. If they aren’t (which is true of most of the inscription nonsense people) then it’s all one market - There is N weight in block space created in a day and only maybe 3% of it is restricted to non-inscriptions (or whatever). The inscription people will just get into the remaining 97% of block space that is now slightly less full because the filtooring hashpower took somewhat lower fee non-inscriptions.
There’s some edge cases around if the 3% of filtooring miners find a few blocks in a row so the prevailing feerate is somewhat lower in their blocks than others, but with so little hashpower it won’t happen.
And, more generally, I didn’t claim that filters don’t work period. Indeed, as long as there is very little revenue to be had from the spam there is some cost to working around them. They’re maybe good for like 0.1 sat/vB in additional cost from effort and whatever. But it’s very very little, and if there starts to be lots of transactions that are non-standard then it will get driven down further (while driving small miners out of business!)
nostr:npub10vlhsqm4qar0g42p8g3plqyktmktd8hnprew45w638xzezgja95qapsp42 which app did you post this link with?
Oh come on Jimmy. Lots of transactions in general pay too high a fee. With such a small sample size that’s not evidence of anything.
And, again, my point above was about what happens when people *regularly* want to create non-standard transactions. As long as it’s incredibly rare, there’s not a lot of demand and only one or two pools might offer it (indeed, F2Pool and MARA today). The whole start of this drama was because a potentially-large transactor wanted to fill the UTXO set with garbage because they didn’t want to rely on OP_RETURNs. Obviously that changes the calculus here.
Potentially large transactor? Please tell me you’re not referring to Citrea?
Come on, Jimmy, slipstream bad, let's not incentivize that behavior! Add to that the potential harm to the utxo set, all for inconsequential storage costs of op return data? What if someone discovers a way for anchors in op return to facilitate or enhance a legitimate scaling solution, should core be standing in the way? Core is on the right side of this one, and wonderfully, anyone that disagrees is free to not upgrade or switch to knots. All this said, I very much appreciate all you do and have done for so long!
