Re: TX replacement and nLockTime | Bitcointalksearch.org

mmeijeri

hero member

Activity: 714

Merit: 500

Martijn Meijering

Quote from: gmaxwell on April 07, 2014, 09:43:31 AM

And I can't see how any amount of 'passive mode' could really help there, except perhaps by making sure the sequence number cranks only in one direction, but someone could still fail to include a newer one

Yeah, but then at least you'd know about it in advance, before the transaction was locked.

gmaxwell

staff

Activity: 4284

Merit: 8808

Quote from: mmeijeri on April 07, 2014, 09:28:07 AM

I don't understand how you could ever rely on a newer transaction overriding an older one unless unlocked transactions make it into the blockchain in a "passive mode".

You can't. If replacement was functional in the network you could pay a higher fee and hope that this incentivizes the next miner to do the right thing, but it's not a guarantee. (And I can't see how any amount of 'passive mode' could really help there, except perhaps by making sure the sequence number cranks only in one direction, but someone could still fail to include a newer one).