--- Day changed Mon Jan 07 2019
06:48 < instagibbs> achow101, if it's a lot of burden to maintin it just remove support or stub it out, tbh
06:49 < achow101> i figured it out eventually. it's mostly just a copy paste from the trezor. the issue i was having was installing their library from source. it said it was installing version 4.0.2 when the tag I checked out was very clearly 6.0. turns out they don't bump their version numbers in setup.py with new tags
06:50 < achow101> the version on pypi is 4.0.2 which is like 2 years old so things i looked up in the master branch of their repo were not working
06:57 < achow101> instagibbs: thoughts on the ledger nano x?
06:58 < instagibbs> I'll work on support, though I imagine it's basically the same thing
06:58 < instagibbs> might be identical
06:58 < instagibbs> larger screen + bluetooth seems to be the delta
06:58 < achow101> the way that the ledger has structured bolos seems like the implementation should be exactly identical
06:59 < instagibbs> yep     
06:59 < achow101> also bluetooth....
06:59 < instagibbs> wonder how hard it would be to intentionally disable it :)
07:01 < instagibbs> achow101, yeah confirmed, no API changes at all
07:01 < instagibbs> for x
07:02 < achow101> cool
07:03 < instagibbs> think it's worth it to at some point strip out the btchip-python stuff? Seems like they're very infrequently changing stuff anyways.
07:03 < achow101> and replace it with what?
07:03 < instagibbs> well I mean just take the parts we're using, slim it down
07:04 < instagibbs> removing a dependency
07:04 < instagibbs> the python API is pretty bizarre and supports all this backwards compat stuff we don't care about
07:05 < achow101> if you want to do that, go ahead
07:06 < achow101> i'm just not keen on implementing their usb protocols directly when someone else has already done that for us in the library
07:07 < instagibbs> I'll see about doing it upstream instead
08:23 < instagibbs> achow101, test gap: https://github.com/achow101/HWI/pull/94
08:23 < instagibbs> ledger tests passing when they shouldn't :)
08:52 < instagibbs> manual ledger tests: https://github.com/achow101/HWI/pull/95
08:53 < instagibbs> better than nothing
08:55 < gmaxwell> bluetooth sounds like a bad idea for a hardware wallet. :P big attack surface!
09:05 < instagibbs> I just wanted the slightly larger screen to make address verification easier
09:09 < instagibbs> apparently the secure element is also faster at signing, no numbers on hand tho
10:23 < achow101> still a closed source secure element though
10:53 < instagibbs> that's expected for basically forever
11:17 < gmaxwell> I dunno that it actually matters much.
11:17 < gmaxwell> It's not like the st microcontroller in the trezor is open either.
11:17 < gmaxwell> And even if these things were using an open chip (e.g. riscv) there would be basically no pratical way to know if the code running on it were the code you expected.
11:18 < gmaxwell> thats why it would be really good to start using a sidechannel busting signing protocol on all these things.
11:21 < arubi> +1 ^
11:21 < arubi> bring in the s2c's :)
14:21 -!- arubi [~ese168@gateway/tor-sasl/ese168] has quit [Remote host closed the connection]     
14:21 -!- arubi_ [~ese168@gateway/tor-sasl/ese168] has joined ##hwi