Can a node drop a transaction.

[LoyceV]

could this lead to other miners rejecting the new block, and continuing to try to build on the previous block?

That's possible, but those miners risk losing their work when the rest of the network continues without them.

[pooya87]

I think you may have misunderstood me. I appreciate that once a transaction is confirmed on the blockchain, then it is there forever. My point is that nodes are in control of the mempool, and not the blockchain, and so if a large number of nodes run software that excludes certain transactions, then it will make it harder for those transactions to make it into a confirmed block. Of course a rogue miner who finds a block can add that transaction onto the chain. He is then dependant on other miners building on top of his block.

Well that is exactly how Bitcoin has been working for as long as the "standard rules" existed. Nodes don't relay valid but non-standard transactions which makes propagating such transactions nearly impossible and only a miner that is willing to manually include it in their candid block can mine them.

Here is an example of a valid but non-standard tx no node wanted to relay and no miner wanted to mine: https://bitcointalk.org/index.php?topic=5192454.0

Bitcoin knots is just adding more non-standard rules and if a miner mines any of such transactions, it still won't be invalid chain and other miners should build on top of it regardless of its non-standardness.

BTW this is not censorship, and in fact it is a very positive thing that has kept the network clean for years. The only thing people don't seem to agree on is what should be non-standard after the Ordinals Attack began!

[Cricktor]

From what I understand about the service at https://slipstream.mara.com, they also accept non-standard transactions to be mined by their non-public mempool at that service. I haven't tried Slipstream by Mara pool yet myself because I don't have a need to do so.

The non-public mempool of Slipstream service has been used with success to withdraw vulnerable low-bit-count private key puzzles, namely puzzles #67 and #68, IIRC. See https://privatekeys.pw/puzzles/bitcoin-puzzle-tx and https://bitcointalk.org/index.php?topic=1306983.msg18765941#msg18765941

Now the bank could build its own mempool, and only include chosen transactions. These get added to the chain when it finds a block. It may not matter if this takes a week or more in some cases. This possibility could become significant as the block reward decreases. could this lead to other miners rejecting the new block, and continuing to try to build on the previous block?

Every mining pool has somewhat an individual mempool. But frankly what is the economical incentive to cherry pick only certain transactions for your block template when this doesn't give the maximum transaction fees of a possible next block to be mined?

As long as the newly found block is valid other miners should accept this new block and numbercrunch on the next block based on that last new block. Anything else doesn't make sense from the game theory of Bitcoin.