1 2018-02-22 00:04:48 <roasbeef> bsm117532: I see, off the top, I know of no such publicly published analysis
2 2018-02-22 00:05:30 <bsm117532> I would have thought someone would have done it for the CSV fork, or for lightning timeouts...
3 2018-02-22 00:27:08 <roasbeef> all our timeouts are expressed in blocks atm
4 2018-02-22 00:27:12 <roasbeef> relative and absolute
5 2018-02-22 00:27:26 <roasbeef> pre-spec, lnd used time in some places, but then switched over to blocks post-spec
6 2018-02-22 00:27:39 <roasbeef> idk blocks just feels "cripser"
7 2018-02-22 00:27:47 <roasbeef> all in the system
8 2018-02-22 00:27:50 <roasbeef> crisper*
9 2018-02-22 00:28:36 <bsm117532> i agree on that. Business processes tend to have this 4pm problem though
10 2018-02-22 00:29:05 <bsm117532> Let's make all the transactions ever at 4pm. It's a bad idea, but the way things work now.
11 2018-02-22 00:30:50 <bsm117532> I'm interested in timelocks to enforce business processes
12 2018-02-22 00:31:07 <bsm117532> Working on updating this idea https://blog.sldx.com/re-imagining-cold-storage-with-timelocks-1f293bfe421f
13 2018-02-22 00:39:23 <bsm117532> I never mentioned it in that article, but segwit is a critical component of the idea, because offline signed transactions can become invalid if malleation occurs. So now that we have that...
14 2018-02-22 00:49:03 <bsm117532> roasbeef: How do you handle timelocks across two blockchains, where the block times are different? I imagine you'd have to use absolute timestamps and Median Time Past?
15 2018-02-22 03:10:19 <roasbeef> bsm117532: either you tell me what the height is on say ltc, or I just know it (atm), then I can do the route as normal, for the border node (straddles both chains) i need to scale up or down depending on the diff in block
16 2018-02-22 03:18:20 <roasbeef> or reandevous style, you just give the the encrypted latter half, i handle getting to the border node, dunzo
17 2018-02-22 08:33:05 <FraJah> meows
18 2018-02-22 10:13:29 <BaranByk> hi all.
19 2018-02-22 14:46:43 <denis2342> every time when my bitcoind is stopped with SIGINT because the system is shutting down or rebooting all the blocks from that session get replayed with ââ¬ÅRolling Forwardââ¬Â¦.ââ¬Â. This does not happen when I stop the deamon with bitcoin-cli stop
20 2018-02-22 14:46:52 <denis2342> can someone try to confirm this?
21 2018-02-22 14:47:12 <denis2342> running 0.16.0rc4 btw
22 2018-02-22 14:53:34 <wumpus> that's expected behavior, to make sure the in-memory work is entirely written to disk, you need to wait for bitcoind to exit. SIGINT does the same as bitcoin-cli stop, the only destinction is whether you wait for it to quit.
23 2018-02-22 14:53:51 <wumpus> (and not, say, send SIGKILL a second after...)
24 2018-02-22 14:56:41 <denis2342> wumpus: not convinced, because the timeout is 60 seconds before it gets killed and I saw that it didnt take that long
25 2018-02-22 14:57:03 <denis2342> I try to send the SIGINT by hand without timeout and see what happensââ¬Â¦
26 2018-02-22 14:57:04 <wumpus> check the debug.log then and see if the shutdown sequence finished
27 2018-02-22 14:58:16 <wumpus> if it says there that it shut down completely (the "Shutdown: done" message) and it didn't flush everything to disk, then there might be a bug...
28 2018-02-22 15:10:05 <denis2342> wumpus: youââ¬â¢re right, after a system shutdown the message ââ¬ÅShutdown: doneââ¬Â is missing
29 2018-02-22 15:16:10 <wumpus> ok
30 2018-02-22 20:12:41 <armoredShawarma_> Hello
31 2018-02-22 21:03:04 <armoredShawarma_> Does anyone know of any good open source projects where I can learn and contribute?
32 2018-02-22 22:46:29 <FraJah> nite nite everyone. ^_^