--- Day changed Wed Jan 23 2019 08:13 < instagibbs> achow101, probably derping on trezor pin, https://github.com/achow101/HWI/pull/109#issuecomment-456860509 08:14 < instagibbs> really really like being able to init without the web wallet 08:19 <@provoostenator> instagibss: indeed, especially because their web wallet doesn't even work with Tor, so it's guaranteed self-doxxing. 08:20 <@provoostenator> (well, a VPN might help, if you can trust the VPN provider) 08:21 <@provoostenator> instagibbs: by the way, bumpfee would indeed be a nice addition, but from what I recall that method needs to be improved in general, because it won't work if there's not room in the change output. 08:22 < instagibbs> hmmm updating their firmware is the next challenge :D 08:22 <@provoostenator> I should really add a new output in that case. Also it'd be nice if I could just give it a factor to increase the fee with, to prevent any fat finger issues. 08:23 < instagibbs> bumpfee improval honestly is going to take work 08:23 < instagibbs> ive done implementations a couple of times 08:23 <@provoostenator> Reset device to throwaway keys, update firmware, restore? 08:23 < instagibbs> best idea, I think, is to directly use CreateTransaction, and allow coin selection to choose already-spent coins 08:23 < instagibbs> to double spend the original tx 08:23 <@provoostenator> Part of the problem is that coin selection is still not entirely cleanup up. 08:24 < instagibbs> use coin control as the "intent" for the rbf intention conveyance 08:24 < instagibbs> right 08:24 < instagibbs> if you want to push for it, I'm willing to lend a hand 08:24 < instagibbs> previously I got no review traction or interest and stopped 08:24 < instagibbs> you can always manually rbf 08:24 <@provoostenator> It would be nice if you feed an existing transaction hash and turn that into a mutable transaction object, where you can bump the fee and (strictly) add inputs / outputs. 08:25 <@provoostenator> My motivation to work on RBF depends on the fee level, currently not very high. 08:25 <@provoostenator> That said, reviewing and testing I can do. 08:26 < instagibbs> right, i need to go on sabbatical to fix fee bumping in Core 08:26 < instagibbs> lol 08:30 < instagibbs> oh no, i installed their web bridge to install firmware update, and now enumerate is @!#$!@ed 08:31 <@provoostenator> Doesn't Liquid have a block size limit? You wouldn't want transactions to freeze... that would evaporate confidence! 09:52 <@provoostenator> I updated my bitcoind branch to once again match the HWI branch it works with (no longer using named --psbt and --desc in some calls) 09:52 <@provoostenator> Also added an account argument to fetch keys. 10:03 < instagibbs> what is the account option anyways 10:06 <@provoostenator> BIP32 account 10:07 <@provoostenator> m/purpose'/coin'/account'/[0,1]/* 10:07 <@provoostenator> Where afaik "account" is the only thing someone might want to change. 10:23 < achow101> instagibbs: you can update firmware using trezorctl which comes with the python-trezor library 10:24 < instagibbs> too late now :) good to know tho 10:25 < achow101> A list of firmware versions can be found at https://wallet.trezor.io/data/firmware/1/releases.json with the firmware binaries at https://wallet.trezor.io/data/firmware/1/trezor-.bin 10:25 < achow101> trezorctl will call back to trezor.io if you don't give it a firmware binary to use 10:29 < achow101> should we add an updatefirmware command? 10:38 < instagibbs> something to ponder if there's any downsides 10:42 < instagibbs> I'd like it so I don't feel compelled to use these centralized wallet solutions just to update 10:42 < instagibbs> esp ledger(I sheepishly don't know how to update it without their apps) 12:02 < achow101> any ideas why the latest hww branch is causing tests to fail? 12:02 < achow101> bitcoind tests run fine 12:03 < instagibbs> weird 19:53 < achow101> I fixed that problem. i missed a thing in the rebase