--- Day changed Thu Jan 31 2019 01:21 <@provoostenator> OMG OMG OMG! achow101 congrats! 01:24 <@provoostenator> I updated the links from my PR. I expect to rebase that again next week or so, hopefully when the PSBT refactor and one or more hww branch PR's is merged. 07:17 < achow101> provoostenator: can you update the topic of this channel? 07:23 -!- provoostenator changed the topic of ##hwi to: Hardware Wallet Interaction scripts | https://github.com/bitcoin-core/HWI 07:24 -!- mode/##hwi [+oooo achow101 instagibbs jnewbery meshcollider] by provoostenator 07:24 -!- mode/##hwi [+o moneyball] by provoostenator 07:25 <@achow101> provoostenator: can you set up chanserv here so people can take ops if they need it? 07:26 <@provoostenator> @achow101 sure, how? 07:27 <@achow101> provoostenator: /msg chanserv flags ##hwi +o 07:27 <@achow101> that lets op themselves here 07:29 <@provoostenator> Done, op yourself if you want op. 07:32 <@provoostenator> achow101 can you rebase https://github.com/bitcoin/bitcoin/pull/15226 since your upstream thing just got merged 07:32 <@provoostenator> I'll give it a review after 07:32 <@achow101> provoostenator: yes, doing that now 08:39 <@provoostenator> achow101: your "--usecli, unify RPC arg" PR would also benefit from a trivial rebase, currently breaks p2p_invalid_messages.py on my machine and at least once on Travis 08:40 <@provoostenator> (whereas when I rebase it myself it works fine) 09:39 -!- harding [~quassel@li1228-87.members.linode.com] has quit [Remote host closed the connection] 09:56 -!- harding [~quassel@li1258-130.members.linode.com] has joined ##hwi 11:51 <@provoostenator> achow101: your "Import watch only pubkeys" PR doesn't have the GUI "Payment Request" fix yet, right? 11:52 <@achow101> provoostenator: it should 11:52 <@provoostenator> Wait nvm, I messed up the import 11:53 <@provoostenator> Did we make it safer? It no longer adds watch-only keys unless I set watchonly true. 11:54 <@achow101> i rebased on 14491 at some point which I think changed that. not sure 11:57 <@provoostenator> Yeah, it actually works (I forgot the keypool arg) 12:03 -!- harding [~quassel@li1258-130.members.linode.com] has quit [Remote host closed the connection] 12:03 -!- harding [~quassel@li1258-130.members.linode.com] has joined ##hwi 12:04 <@provoostenator> achow101 should there be any difference in the output of dumpwallet before and after you call generateaddress? 12:04 <@provoostenator> In the case of watch-only wallets, which only have scripts, no key entries in the dump. 12:04 <@achow101> i don't think so 12:05 <@provoostenator> I'm still a bit confused about the actual nature of the keypool. 12:05 <@achow101> keypool state doesn't get dumped since that isn't something that should be importable 12:06 <@provoostenator> And that's because when you import, it's going to rescan anyway, which is what really decides what the keypool needs to look like? 12:06 <@achow101> importwallet doesn't add things to the keypool 12:06 <@provoostenator> (although it could lead to you generating an address twice, if it was unused before) 12:07 <@provoostenator> Right, normally the keypool is only populated upon wallet creation, and whenever it runs low and there's an HD seed. 12:07 <@provoostenator> And your PR also populates it during import. 12:09 <@provoostenator> Ideally I'd like TopUpKeyPool() to handle these imported keys, but I guess that's not possible, because we only import the script, which isn't enough info. 12:11 <@provoostenator> That also means makes dumpwallet useless for these watch-only wallets. 12:12 <@provoostenator> It would be more useful if we added the public keys to the wallet somehow, but perhaps that's too invasive? 12:13 <@provoostenator> The current solution is just enough to make hardware wallets work, and then we can do things properly in a descriptor-wallet. 12:13 <@achow101> i really don't think that we should have import things in the keypool unless they are explicilty added 12:14 <@achow101> I also think that we should deprecate all of the other import commands and just have everything through importmulti 12:15 <@provoostenator> That just means having to move "keypool" logic from importmulti into the Descriptor class later on. 12:16 <@provoostenator> Because in a descriptor based wallet there seems no need for import commands to touch the keypool. 12:17 <@provoostenator> It would just import the descriptor itself, and then the wallet can handle the rest, generating addresses on demand, watching for relevant stuff based on a cache of keys/scripts, etc