Lloyd Fournier
Lloyd Fournier
Another problem: https://github.com/LLFourn/bdk_core_staging/issues/135
I don't think we can say this is really dealt with. Do we have a test case for: > Furthermore, any time you insert a new tx into the chain...
True. This is requires doing CPFP modelling first!
Maybe even esplora tells you the actual feerate with package in mind? I mean mempool.space tells you this I think (at least includes CPFP fee).
see: https://afilini.com/blog/fuzz-your-mother-if-you-want-fuzz/
Would be nice to allow marking utxos as spent in memory even if you haven't really spent them. See https://github.com/bitcoindevkit/bdk_wallet/issues/166 for motivation.
Can you try and re-do this PR on top of master @rajarshimaitra?
Expanding on this. Let's try and create a new txbuilder that is split up into stages: 1. Coin control. Grouping inputs together that *must* be spent together and inputs that...
Update: I've moved the coin selection to its own crate that only depends on `rust-bitcoin` as requested in https://github.com/revault/minisafe/issues/51#issuecomment-1280446716. It's also easier like this to start putting it into `bdk`....
I'll indeed address this in my work coming up. I understand what you are proposing.