Drop the mempool patch and add more data to DSPs instead, to better encapsulate the ZCE system

[A60AB5450353F40E]

Why can't we just extend DSP data with what's needed, rather than patch the mempool code?

From Telegram convo:

Jonas, [10/7/24 1:51 PM]
Without the mempool patch a user could just send a p2pkh as double spend tx. If that confirms he lost nothing.

bitcoincashautist, [10/7/24 1:53 PM]
how so, doesn't that p2pkh provide the 2nd sig needed to sweep the ZCE UTXO?

Jonas, [10/7/24 1:54 PM]
But without mempool replacement the miner might not know about the ZCE at all

bitcoincashautist, [10/7/24 1:55 PM]
won't he get the DSP? like, can't we just mod the DSP scheme to propagate both signatures

Jonas, [10/7/24 2:01 PM]
You can't recreate the full ZCE transaction (as you need to for this) by the information in the DSP.

bitcoincashautist, [10/7/24 2:02 PM]
can't we just add the missing info to DSPs

bitcoincashautist, [10/7/24 2:02 PM]
like, just add the redeem script

Jonas, [10/7/24 2:02 PM]
You also need to know the output to the merchant

bitcoincashautist, [10/7/24 2:02 PM]
add that too

Jonas, [10/7/24 2:03 PM]
Why bother with DSP at all at that point? Why not just broadcast the full tx and let the nodes replace it in the mempool......

bitcoincashautist, [10/7/24 2:04 PM]
so we don't have to modify mempool code

bitcoincashautist, [10/7/24 2:04 PM]
we could cointain the mod to DSPs

Jonas, [10/7/24 2:06 PM]
But if we are replacing a tx in the mempool (as we do with the information in the DSP as you suggest) we are modifying the mempool code.

bitcoincashautist, [10/7/24 2:09 PM]
there is no "the" mempool

with adding it to DSP, the info would propagate no matter the nodes local mempool state

bitcoincashautist, [10/7/24 2:10 PM]
it would not change their mempool state

bitcoincashautist, [10/7/24 2:10 PM]
but miners may act on the info, and kick out one TX from the block template and add the sweep TX

bitcoincashautist, [10/7/24 2:10 PM]
other nodes mempools would get updated after 1conf

Jonas, [10/7/24 2:13 PM]
That's literally changing the mempool state. Miners and other nodes should operate in the same way, otherwise child transactions of the p2pkh could be used for double spending since an attacker know that miners will replace it but nodes won't until a block is mined.

bitcoincashautist, [10/7/24 2:13 PM]
nodes will see the DSP and the DSP score of 0 can propagate to children

bitcoincashautist, [10/7/24 2:14 PM]
DSP= this TX may or may not be replaced, 2 versions exist

mempool patch= I assume this TX will be replaced so I just do it now, as well

bitcoincashautist, [10/7/24 2:15 PM]
could even have something like negative score, like -1 for "likely to get replaced"

bitcoincashautist, [10/7/24 2:16 PM]
imo this replacement stuff naturally belongs in DSP system, it should be encapsulated by it