If the mempool gets filled above the 250 MB threshold, small fee transactions get dropped, low-powered full nodes have their memory overloaded and go offline, and third world users get priced out of opening Lightning channels.
But if I insert a 50kb picture into the blockchain using Ordinals, none of the above happens.
Both use cases pay fees to the miners, both can benefit from the SegWit discount.
Which one is worse?
The default mempool size is 300 MB. Transactions that were dropped can still be rebroadcast later. Nodes don’t usually go offline from having a full mempool, because the mempool limit is so small.
Either of the two may increase the minimum feerate for getting into the next block and price out less valuable transactions. Both filling the mempool or submitting a transaction with an inscription to the network are just ways to bid on blockspace. It seems to me that both Inscriptions and low-value transactions will be priced out in the long run, but while demand for blockspace is low, it follows that it’s cheap to acquire the undemanded blockspace. There isn’t much that one can do against it except to make more competitive bids for the blockspace.—If someone wants to make sure that the blockspace gets utilized fully, they just need to ensure that there is always another transaction bidding for blockspace at the minimum feerate.
reply
I'm impressed by your fairly reasonable take, when I saw that you replied I honestly expected you to roast me or imply that I'm debating in bad faith with the examples I'm presenting.
However, it's concerning to me that there are still devs and influencers who are trying to police the way that others use their own bitcoin.
The fact that we still refer to it as "permissionless", advocate for building a market for block space and suggest that "Bitcoin is for enemies" while we agree that some legitimate use cases that the network enables are attacks or spam is one sweet piece of irony. These people have no issue still calling themselves libertarians.
reply
Thanks. Yes, the irony of the “not-in-my-blockchain” folk was not lost on me either.
reply
The fact that we still refer to it as "permissionless", advocate for building a market for block space and suggest that "Bitcoin is for enemies" while we agree that some legitimate use cases that the network enables are attacks or spam is one sweet piece of irony.
Well said, couldn't agree more.
reply
The size that were drop can still be rebroadcast later as said Nodes won't go offline from having a full mempool, because the mempool limit is small
reply
I don't know man, maybe they're using Raspberry Pi 3 systems with 1GB of RAM. As far as I know, the mempool data is stored in the RAM.
reply
I dont think it goes offline, it just don't save new transactions. We have a diversity of nodes some may have 32GB of RAM so it's OK. Next time I make a node i will get it 16gb RAM and reserve 8GB for the mempool on clearnet+tor+i2p , problem solved guys
reply
low-powered full nodes have their memory overloaded and go offline
If you're running Bitcoin Core on a computer with so little ram you can't do the default 300MB mempool it's totally ok to just reduce the mempool size limit. There's no significant consequences to doing that; the mempool is not consensus. Equally, it's totally ok to increase that limit too.
reply
people are assholes, it's called life.
reply