1 2015-06-30 01:44:41 <Laab> hey all
  2 2015-06-30 01:45:13 <Laab> i have nothing related to bitcoin but i'm stuck on a python problem, hoping that anyone can help!
  3 2015-06-30 01:47:11 <teward> Laab: ##python
  4 2015-06-30 01:47:18 <teward> (this is not a python help channel)
  5 2015-06-30 01:47:25 <Laab> i tried already :((
  6 2015-06-30 01:47:26 <teward> erm... #python
  7 2015-06-30 01:47:49 <Laab> i tried stackexchange, tried #scrapy, tried, #python, tried google T__T
  8 2015-06-30 01:49:00 <teward> well it's not here
  9 2015-06-30 02:55:12 <cfields> wumpus: for backlog: after #6354, I'm good to go for rc3 whenever you are. Only build related issue left on my radar is #6248 which i have a non-trivial fix for. If you want that in for rc3, i can fast-track a PR for it tomorrow.
 10 2015-06-30 02:56:28 <cfields> for reference, it involves detecting the "relocations" config in qt, and using -fPIC for qt objects rather than fPIC. The result is still a pie object, just minus the fPIE optimizations for the qt objects.
 11 2015-06-30 02:56:45 <cfields> that doesn't affect our release binaries though.
 12 2015-06-30 02:57:02 <cfields> *rather than fPIE
 13 2015-06-30 02:57:21 <Luke-Jr> thanks, I was getting confused there :P
 14 2015-06-30 02:58:31 <cfields> after researching a good bit, -fPIE is just an optimization that you can take advantage of on top of -fPIC when building a pie binary
 15 2015-06-30 02:59:49 <cfields> docs say to use -fPIE for pie binaries, but gcc mailing list has gcc devs quoted saying that -fPIC objects in pie binaries will continue to be a supported config
 16 2015-06-30 04:50:16 <leakypat> petertodd: you mentioned there is a DNS seed for testnet RBF nodes?
 17 2015-06-30 04:50:56 <leakypat> Do you have the address again? Is it just a case of adding it to Bitcoin.conf?
 18 2015-06-30 04:51:33 <leakypat> Also phantomcircuit , you are mining on RBF correct?
 19 2015-06-30 04:51:42 <leakypat> Can I connect directly to your node?
 20 2015-06-30 05:09:17 <petertodd> leakypat: rbf-seed.tbtc.petertodd.org
 21 2015-06-30 05:10:04 <petertodd> leakypat: though the full-rbf trees at https://github.com/petertodd/bitcoin/tree/replace-by-fee-v0.10.2 and https://github.com/petertodd/bitcoin/tree/replace-by-fee-v0.11.0rc2 should both find other RBF nodes automatically, though it may take awhile
 22 2015-06-30 05:38:57 <leakypat> Ok, yeah it's finding one other node currently
 23 2015-06-30 05:39:43 <leakypat> And is relaying the RBF transactions ok, but always the original one gets mined
 24 2015-06-30 05:51:36 <petertodd> leakypat: right, probably not any RBF hashing power then
 25 2015-06-30 05:54:20 <leakypat> petertodd: phantomcircuit said he had upgraded his testnet miners to support it
 26 2015-06-30 05:54:30 <leakypat> Maybe i should get me a miner
 27 2015-06-30 05:55:08 <petertodd> leakypat: yeah, he did - maybe he tripped over the power cord? :P
 28 2015-06-30 05:55:17 <petertodd> leakypat: I need to get some hashing power myself
 29 2015-06-30 05:55:55 <leakypat> Testnet looks like it runs at 1-2 th/sec
 30 2015-06-30 05:56:28 <petertodd> leakypat: that's about $200 worth of hashing power IIRC
 31 2015-06-30 05:56:33 <petertodd> leakypat: (fixed cost)
 32 2015-06-30 05:56:51 <leakypat> Yeah, I'm sure I can find someone to offload their old miners to me
 33 2015-06-30 05:57:28 <Luke-Jr> assuming they still work
 34 2015-06-30 05:57:41 <Luke-Jr> I'm finding a lot of this stuff isn't built to last (no surprise)
 35 2015-06-30 05:57:55 <leakypat> I'll make sure I buy locally :)
 36 2015-06-30 05:59:55 <leakypat> If anyone in Tokyo has any of course
 37 2015-06-30 06:00:29 <Luke-Jr> MagicalTux? <.<
 38 2015-06-30 06:00:39 <CodeShark> lol
 39 2015-06-30 06:00:56 <leakypat> Haha, he probably does
 40 2015-06-30 06:01:21 <Luke-Jr> considering that class action lawyer made a big annoyance of himself with Eligius because MT was mining there… yeah, probably
 41 2015-06-30 06:01:54 <Luke-Jr> it was long enough ago that it probably doesn't make a profit anymore either
 42 2015-06-30 06:02:14 <leakypat> I asked him
 43 2015-06-30 06:03:11 <petertodd> leakypat: you can buy hashing power via nicehash.com too
 44 2015-06-30 06:03:37 <Luke-Jr> eh, I don't advise supporting hashing-for-sale services
 45 2015-06-30 06:04:11 <Luke-Jr> if your goal is bitcoin long-term especially
 46 2015-06-30 06:06:11 <leakypat> This ant miner s5 looks cool though :)
 47 2015-06-30 06:07:46 <Luke-Jr> yeah, get one of those and help me reveng its code :P
 48 2015-06-30 06:15:59 <leakypat> Urg an RBF reddit thread
 49 2015-06-30 06:18:44 <Luke-Jr> another?
 50 2015-06-30 06:24:51 <leakypat> Just saw this one , I've been busy all day - working on RBF lol
 51 2015-06-30 06:35:05 <petertodd> leakypat: what are you making?
 52 2015-06-30 06:40:57 <leakypat> Double spend detection, "push stuck transaction" (switchable between FSS and full)
 53 2015-06-30 06:41:16 <leakypat> And a "recall transaction" button for testnet only
 54 2015-06-30 06:42:04 <leakypat> Also warnings on receiving low fee / weird transactions
 55 2015-06-30 06:42:43 <leakypat> Also looking at fee recommendation
 56 2015-06-30 06:43:04 <leakypat> And mem pool / last few block fee data feeds
 57 2015-06-30 06:44:25 <leakypat> It's already throwing up a lot of interesting UI/UX questions :)
 58 2015-06-30 07:00:42 <jonasschnelli> leakypat: what wallet are you working on?
 59 2015-06-30 07:00:50 <leakypat> Ninki wallet
 60 2015-06-30 07:03:15 <leakypat> Backup your keys! :)
 61 2015-06-30 07:03:57 <bitcoin-dev358> hi
 62 2015-06-30 07:04:20 <bitcoin-dev358> any one tell me which is be best thing to start wallet development
 63 2015-06-30 07:05:06 <jonasschnelli> bitcoin-dev358: consider helping on of the existing wallet. Fullnode bitcoin-core wallet needs support. Bitcoinj SPV java library also.
 64 2015-06-30 07:07:10 <jonasschnelli> leakypat: looks nice. Nice UI.
 65 2015-06-30 07:07:43 <leakypat> Thanks
 66 2015-06-30 07:07:54 <leakypat> What is your username?
 67 2015-06-30 07:08:03 <jonasschnelli> leakypat: major issue: don't send the master public key by email.
 68 2015-06-30 07:09:34 <jonasschnelli> leakypat: what i never got (same for Darkwallet) why using a Chrome as app sandbox?
 69 2015-06-30 07:09:56 <leakypat> Yeah, emailing pub key I'm going to change
 70 2015-06-30 07:10:35 <leakypat> Have been trying to find a way to make it easy for users to sign up for multisig wallets, but pretty much resigned to the fact that- it isn't easy
 71 2015-06-30 07:12:19 <jonasschnelli> leakypat: You don't need the master pub key for a hdm wallet IMO.
 72 2015-06-30 07:12:34 <leakypat> It's a multisig wallet
 73 2015-06-30 07:12:40 <leakypat> So they need it for recovery
 74 2015-06-30 07:12:47 <leakypat> If we go offline
 75 2015-06-30 07:13:06 <leakypat> I think dark wallet was a chrome extension rather than an app
 76 2015-06-30 07:13:07 <jonasschnelli> leakypat: how would you recover with a master pub key?
 77 2015-06-30 07:13:34 <leakypat> You need your two private key seeds + the master public key of the key that we hold
 78 2015-06-30 07:14:04 <leakypat> So the bip32 * 3 address chain can be derived
 79 2015-06-30 07:14:40 <jonasschnelli> leakypat: So you support recovery over a chain your company owns?
 80 2015-06-30 07:15:25 <leakypat> It is a 2 of 3 multi sig wallet, the user holds two keys (the ones you wrote down) and we hold one
 81 2015-06-30 07:15:36 <leakypat> Any 2 of the 3 can spend the funds
 82 2015-06-30 07:16:07 <jonasschnelli> Okay. I see.
 83 2015-06-30 07:16:12 <leakypat> So you can recover the funds anytime external to our wallet
 84 2015-06-30 07:17:02 <leakypat> So the xpub for the key that Ninki holds
 85 2015-06-30 07:18:38 <leakypat> The chrome app is for (assuming you are delivering html/Js code) a sandboxed env where no extensions or other sites get loads, signed delivery if code, a content security policy and lockdown from xss attacks
 86 2015-06-30 07:18:52 <leakypat> Of course it isn't perfect
 87 2015-06-30 07:19:21 <jonasschnelli> But with a chrome app you heavily relay on google? Not?
 88 2015-06-30 07:19:27 <leakypat> Of code*
 89 2015-06-30 07:20:11 <leakypat> you rely on chromium and their delivery mechanism if you use it , yes
 90 2015-06-30 07:20:57 <leakypat> It's similar to deploying on android or iOS via play/iTunes
 91 2015-06-30 07:24:45 <leakypat> This is probably the coolest feature:
 92 2015-06-30 07:24:48 <leakypat> http://ninki.ghost.io/2015/03/17/ninki-bitcoin-air-wallet/
 93 2015-06-30 07:25:29 <leakypat> You can "add contact" with other Ninki users and exchange xpubs nodes on your hd chains
 94 2015-06-30 07:25:55 <leakypat> Then you can generate addressees for each
 95 2015-06-30 07:26:01 <leakypat> So just click and send
 96 2015-06-30 07:26:11 <leakypat> *other
 97 2015-06-30 07:32:53 <jonasschnelli> leakypat: Yes. The address book integration and address sharing is nice!
 98 2015-06-30 07:33:37 <leakypat> jonasschnelli: you created 2 accounts?
 99 2015-06-30 07:33:51 <jonasschnelli> leakypat: no. One.
100 2015-06-30 07:33:58 <leakypat> You can add me as NinkiBen
101 2015-06-30 07:41:35 <petertodd> leakypat: all sounds like useful features!
102 2015-06-30 07:50:44 <leakypat> jonasschnelli: 4f709f91742a70c601c791c351fe26723f5f620a
103 2015-06-30 07:50:51 <leakypat> My validation code
104 2015-06-30 07:50:55 <leakypat> Send me yours
105 2015-06-30 07:54:12 <michagogo> cfields: I am now
106 2015-06-30 10:38:42 <michagogo> wumpus: sorry, was going to do that once I got to my computer, but didn't get around to it.
107 2015-06-30 10:53:48 <jonasschnelli> Request for review/merging https://github.com/bitcoin/bitcoin/pull/5994
108 2015-06-30 10:55:14 <jonasschnelli> It would be a required feature to allow rpc testing of my core-wallet implementation. And i'd like to not add another core-chaning PR on top.
109 2015-06-30 10:55:38 <wumpus> jonasschnelli: I'll have a look
110 2015-06-30 11:00:19 <jonasschnelli> thanks
111 2015-06-30 14:04:38 <sdaftuar> jonasschnelli: been meaning to check with you on #6148, were you able to successfully test that pull?
112 2015-06-30 14:05:49 <jonasschnelli> sdaftuar: i need to dive in there again.
113 2015-06-30 14:06:14 <jonasschnelli> but certainly it is something for 0.12!
114 2015-06-30 14:06:21 <sdaftuar> ok no worries, just thought i'd check in.
115 2015-06-30 14:06:51 <jonasschnelli> but currently totally absorbed with the wallet redesign... :)
116 2015-06-30 14:06:55 <sdaftuar> yeah i'd like to get that merged, and then i'm also working on wrapping up a pull to announce blocks via headers, which would nicely solve the relaying of reorgs by pruning nodes too
117 2015-06-30 14:07:06 <sdaftuar> yep i know :)
118 2015-06-30 14:07:14 <jonasschnelli> but will test #6148 soon. Thanks for the remindder.
119 2015-06-30 14:07:44 <sdaftuar> cool, let me know if you run into any issues
120 2015-06-30 14:07:54 <hearn_> jonasschnelli: btw what editor/ide do you use?
121 2015-06-30 14:08:04 <jonasschnelli> hearn: XCode. :)
122 2015-06-30 14:08:12 <hearn> ah ha
123 2015-06-30 14:08:12 <jonasschnelli> Theres no better IDE on mac.
124 2015-06-30 14:08:15 <hearn> i am setting up CLion at the moment
125 2015-06-30 14:08:31 <hearn> IntelliJ is the best java IDE by far so I have high hopes for CLion. It needs a CMake build system so I'm patching a basic onetogether for OSX
126 2015-06-30 14:08:36 <hearn> i can share it with you when i'm done if you like
127 2015-06-30 14:08:54 <jonasschnelli> Intellij was much better than Eclipse. But nothing is more welded to your mac then XCode.
128 2015-06-30 14:09:16 <jonasschnelli> And i used XCode daily during the last 8 years. So it' s kind of my home. :)
129 2015-06-30 14:11:13 <hearn> ok, cool, fair enough :
130 2015-06-30 14:11:14 <hearn> :)
131 2015-06-30 14:20:59 <leakypat> damn dossers
132 2015-06-30 14:45:33 <wumpus> jonasschnelli: any opinion on https://github.com/bitcoin/bitcoin/pull/5994#discussion_r33565065 ? do you have a specific reason for moving away from the CReserveKey concept?
133 2015-06-30 15:58:51 <michagogo> cfields: TIL about the `rename` command
134 2015-06-30 16:15:04 <wumpus> * [new tag]         v0.11.0rc3 -> v0.11.0rc3
135 2015-06-30 17:05:45 <cfields> michagogo: glad i could help :)
136 2015-06-30 17:06:36 <cfields> ping gitian builders, notice that the dir for win signatures has changed: 0.11.0rc3-win -> 0.11.0rc3-win-unsigned
137 2015-06-30 17:10:36 <michagogo> cfields: you know that doesn't actually ping them, right? :P
138 2015-06-30 17:10:45 <koobs> it pings me
139 2015-06-30 17:10:55 <koobs> *summons*
140 2015-06-30 17:11:05 <cfields> michagogo: heh, yea
141 2015-06-30 17:11:07 <michagogo> koobs: really?
142 2015-06-30 17:11:20 <koobs> michagogo: only if also by cfields
143 2015-06-30 17:45:33 <michagogo> cfields: Hm, I wonder if anything bad would happen if you were to upload the sigs (but not the binaries) before we got 3 builders, meaning that they'd be broken for anyone else who tried to apply them if there was something wrong with your build
144 2015-06-30 17:46:23 <michagogo> Wait, never mind, if there were an attacker they would know how to make the same malicious binaries
145 2015-06-30 18:00:44 <aliakbar> Hi everyone!
146 2015-06-30 18:01:21 <Donster> hi
147 2015-06-30 18:01:26 <Donster> aliakbar
148 2015-06-30 18:01:29 <cfields> michagogo: i considered that, and I think it'd generally be safe. Only reason I'm not doing that is because it's possible for us to get some kind of 50/50 gitian build, where there are 2 sets of matching binaries
149 2015-06-30 18:01:58 <aliakbar> How can I reactively broadcast blocks, i.e. after having forked the blockchain and being one ahead, another miner publishes a block and catches up with my height..how can I react to this and publish my secret block
150 2015-06-30 18:01:59 <aliakbar> ?
151 2015-06-30 18:02:03 <cfields> michagogo: for ex, a result that accidentally depends on the date, where eu builders get a different result from us builders
152 2015-06-30 18:02:28 <cfields> in that case, we'd likely want to fix that problem and re-tag, rather than having signed binaries floating around
153 2015-06-30 18:03:29 <aliakbar> I tried following lines:
154 2015-06-30 18:03:34 <aliakbar> LOCK(cs_main);
155 2015-06-30 18:03:37 <aliakbar> pfrom->PushMessage("inv", pfrom->vPrivateInv);
156 2015-06-30 18:03:54 <aliakbar> inside ProcessMessage-function in main.cpp
157 2015-06-30 18:04:01 <aliakbar> didn't make any difference :/
158 2015-06-30 18:07:18 <aliakbar> btw...vPrivateInv is the vectory including all private blocks
159 2015-06-30 18:22:24 <cfields> michagogo: fyi, there's a scammer calling himself `michagogo who pm'd me so that we could "make some money, bro". Seems legit, because "Profit is guaranty". I'm hearing him out for a min to see what he's after. Figured you should be aware, though.
160 2015-06-30 18:22:35 <michagogo> o_O
161 2015-06-30 18:54:33 <teward> cfields: FWIW that's the serb
162 2015-06-30 18:54:40 <teward> michagogo: ^
163 2015-06-30 18:54:48 <teward> Michail1: ^ for reference of the ping elsewhere
164 2015-06-30 18:55:16 <michagogo> the serb?
165 2015-06-30 19:00:42 <teward> a well known scammer
166 2015-06-30 19:00:43 <jonasschnelli> <wumpus:#bitcoin-dev> jonasschnelli: any opinion on https://github.com/bitcoin/bitcoin/pull/5994#discussion_r33565065: will try your approach soon. Indeed it should respect CReserveKey/keepKey.
167 2015-06-30 19:01:13 <michagogo> teward: from that same ident?
168 2015-06-30 19:01:45 <teward> michagogo: irccloud is one of the source id methods, but the realname gives it away
169 2015-06-30 19:01:56 <michagogo> If you have logs, james (IRCCloud admin) probably would like to have them
170 2015-06-30 19:02:27 <michagogo> The [us]id##### ident is unique and unchangable per account
171 2015-06-30 19:02:43 <teward> talk to Michail1, we know they constantly shift among many things
172 2015-06-30 19:02:46 <michagogo> (unchangable except for the first letter, which is u for free accounts and s for subscribers)
173 2015-06-30 20:08:11 <morcos> sipa: you around?
174 2015-06-30 20:09:09 <morcos> in DisconnectTip(), why do we call UpdateTip() after we resurrect the mempool transactions?
175 2015-06-30 20:10:30 <morcos> it means that if block 101 is disconnected, then all the transactions in old block 101 are added back to the mempool with height 101, instead of height 100, which seems counterintuitive to me.
176 2015-06-30 20:17:11 <Michail1> Sorry.  Been busy all day.   Strange that serb would be trying any scams for people in here.  No one buys/sells bitcoins, so it would have seemed pointless that he would squat michagogo.
177 2015-06-30 20:19:59 <cfields> * [new tag]         v0.11.0rc3 -> v0.11.0rc3
178 2015-06-30 20:20:02 <cfields> To git@github.com:bitcoin/bitcoin-detached-sigs.git
179 2015-06-30 20:21:50 <michagogo> Michail1: do you have logs of previous occurrences?
180 2015-06-30 20:23:22 <Michail1> :)
181 2015-06-30 21:15:42 <Luke-Jr> petertodd: have you considered a RBF variant that enables full only for txs with (eg) all seq numbers 0x12345678 or something?
182 2015-06-30 21:18:17 <petertodd> Luke-Jr: yup, I mentioned that on the dev list today actually
183 2015-06-30 21:18:50 <petertodd> Luke-Jr: see bottom of http://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015-June/009276.html
184 2015-06-30 21:19:31 <Luke-Jr> ah, still catching up on my emails..
185 2015-06-30 22:06:20 <paveljanik> leav second over...
186 2015-06-30 22:06:24 <paveljanik> leap
187 2015-06-30 22:06:41 <paveljanik> 14M mempool
188 2015-06-30 22:11:35 <paveljanik> not over, but before us.
189 2015-06-30 22:12:12 <phantomcircuit> it's an event that lasts for 0 seconds
190 2015-06-30 22:12:15 <phantomcircuit> ironic no?
191 2015-06-30 22:21:39 <leakypat> dos still ongoing then...
192 2015-06-30 22:22:41 <leakypat> 9 hours I'm still at 0 confs
193 2015-06-30 22:24:22 <leakypat> Actually in this case I have a withdrawal from a brokerage holding things up so wouldn't be able to change the fee anyway unless they did an ANYONECANPAY trans
194 2015-06-30 22:25:56 <leakypat> I tried a CPFP with a massive fee but I guess no miners are picking it up
195 2015-06-30 22:37:28 <Luke-Jr> leakypat: you need to make sure the original tx gets to the mempool still
196 2015-06-30 22:43:04 <leakypat> Ah ok
197 2015-06-30 22:43:19 <leakypat> Good point
198 2015-06-30 23:01:19 <leakypat> Is is possible to run testnet and mainnet at the same time?
199 2015-06-30 23:01:48 <leakypat> I was thinking this might increase the number of mainnet nodes
200 2015-06-30 23:02:05 <leakypat> I am always switching on my dev machines, would be good if by default it ran both somehow
201 2015-06-30 23:02:41 <leakypat> Ie. Testnet=1 ran mainnet and testnet
202 2015-06-30 23:17:29 <leakypat> petertodd: you around ?
203 2015-06-30 23:28:34 <wallet42> hey, I got a weird crashed bitcoind
204 2015-06-30 23:28:48 <wallet42> exit code 137
205 2015-06-30 23:32:41 <petertodd> leakypat: hey
206 2015-06-30 23:34:58 <phantomcircuit> wallet42, linux?
207 2015-06-30 23:35:08 <phantomcircuit> check dmesg and then debug.log
208 2015-06-30 23:36:44 <leakypat> petertodd: hey, had a question about mem pool and RBF, so say I have broadcast a bunch of full RBF transactions adjusting the change amount each time to bump the fee, there will be n transactions in the mem pool with dependency on the same inputs- when a block is found I assume all these transactions will be removed from the pools
209 2015-06-30 23:37:11 <leakypat> ie when any one of those transactions is included in a block rather
210 2015-06-30 23:37:13 <petertodd> leakypat: each RBF replacement completely removes the replaced transactions from the mempool
211 2015-06-30 23:37:27 <CodeShark> and dependencies, presumably
212 2015-06-30 23:37:36 <CodeShark> :)
213 2015-06-30 23:37:39 <petertodd> leakypat: as for conflicts in blocks, it works exactly the way it always has: conflicting tx and deps get removed
214 2015-06-30 23:37:58 <petertodd> CodeShark: ha, yes, and dependencies, for that matter, fees and size of dependencies is added up and evaluated
215 2015-06-30 23:39:20 <CodeShark> could we gain anything by keeping several versions around and using that for fee estimation?
216 2015-06-30 23:39:29 <CodeShark> or for double-spend detection
217 2015-06-30 23:39:42 <petertodd> CodeShark: sure! like job security :P
218 2015-06-30 23:39:44 <leakypat> Ok, so if one miner has transaction RBF 3 as but another miner didn't hear it and still has transaction RBF 2, if RBF 3 is included in a block RBF 2 will be removed from the other miners mem pool
219 2015-06-30 23:40:04 <petertodd> CodeShark: it's a good idea, but that kind of complexity will need a redesign of the mempool I think
220 2015-06-30 23:40:07 <leakypat> If the RBF3 block is then orphaned
221 2015-06-30 23:40:27 <petertodd> leakypat: yeah, remember this isn't any different from how txs get treated normally when different miners heard about different txs
222 2015-06-30 23:40:30 <leakypat> RBF 3 will be returned to everyone's mempool
223 2015-06-30 23:41:23 <petertodd> leakypat: you'd need to double check, but IIRC dead blocks do return their txs to the mempool right now (I know we broke that for awhile...)
224 2015-06-30 23:44:46 <leakypat> It's interesting to try and deal with this in the UI, as i display an unconfirmed amount which will be based on whatever the higher fee candidate is
225 2015-06-30 23:45:08 <leakypat> But anyone of the candidates could be mined in theory
226 2015-06-30 23:46:31 <CodeShark> petertodd: a new structure for the mempool that supports this would be fairly straightforward...but the harvesting of useful data isn't completely trivial
227 2015-06-30 23:46:33 <leakypat> So if my 100 bit fee transaction is mines instead of the 200 bit one by total unconfirmed + confirmed balance will increase by 100bits :)
228 2015-06-30 23:46:46 <leakypat> my
229 2015-06-30 23:47:17 <CodeShark> it's also a fairly large structure
230 2015-06-30 23:47:34 <CodeShark> so probably not the most helpful thing for lightweight clients with poor connections
231 2015-06-30 23:47:54 <phantomcircuit> petertodd, Disconnect* does return the transactions to the mempool now
232 2015-06-30 23:48:26 <leakypat> So it's almost like the unconfirmed balance has a range now
233 2015-06-30 23:49:41 <phantomcircuit> petertodd, oh btw i went back and looked at the mempool limit patch and realized that even with a very full mempool the cost of walking the entire thing is relatively small
234 2015-06-30 23:49:43 <CodeShark> but I suppose there could be incentives for larger nodes to host this information...
235 2015-06-30 23:50:15 <phantomcircuit> (iterating over 11 million integers is pretty fast)
236 2015-06-30 23:50:48 <phantomcircuit> generating 288MB of testnet transactions however was not
237 2015-06-30 23:50:50 <Diablo-D3> phantomcircuit: linux completes an infinite loop in less than 10 seconds
238 2015-06-30 23:51:06 <phantomcircuit> Diablo-D3, :P
239 2015-06-30 23:51:15 <leakypat> Also, spending a zero conf RBF candidate output is a no no
240 2015-06-30 23:51:30 <Diablo-D3> RBF?
241 2015-06-30 23:51:31 <phantomcircuit> blargh
242 2015-06-30 23:51:37 <phantomcircuit> Diablo-D3, replace by fee
243 2015-06-30 23:51:43 <phantomcircuit> the lights keep flickering
244 2015-06-30 23:51:53 <phantomcircuit> i should probably turn off the miner while im using my gpu...
245 2015-06-30 23:52:29 <Diablo-D3> phantomcircuit: or add a new circuit
246 2015-06-30 23:52:33 <petertodd> leakypat: yeah, I'm sure this stuff if handled in the most sophisticated way possible ends up with some crazy transaction graph of possibilities and outcomes
247 2015-06-30 23:52:42 <Diablo-D3> a friend of mine had... well, not BAD wiring in his home office
248 2015-06-30 23:52:48 <Diablo-D3> but not really the bestest either
249 2015-06-30 23:52:55 <petertodd> CodeShark: yeah, a from-scratch rewrite should result in fairly clean code
250 2015-06-30 23:53:00 <phantomcircuit> Diablo-D3, box is already full
251 2015-06-30 23:53:04 <phantomcircuit> i'd have to add another box
252 2015-06-30 23:53:09 <phantomcircuit> *effort*
253 2015-06-30 23:53:33 <petertodd> phantomcircuit: indeed! I'm banking on that in my full-rbf patch actually, at least eventually, as it's best if it can walk all dependencies, which is fine if there's a hard upper limit
254 2015-06-30 23:53:43 <Diablo-D3> he had an electrician run a new 20 amp line
255 2015-06-30 23:53:43 <Diablo-D3> phantomcircuit: oh, meh, fuck
256 2015-06-30 23:53:43 <Diablo-D3> single junction box, its own breaker
257 2015-06-30 23:53:44 <petertodd> phantomcircuit: it's just a bunch of pointer following after all (at least now it is)
258 2015-06-30 23:53:46 <Diablo-D3> phantomcircuit: anyone who builds a house with an already full box is an asshole
259 2015-06-30 23:54:13 <phantomcircuit> Diablo-D3, ancient house with a bunch of circuits added after the fact
260 2015-06-30 23:54:26 <phantomcircuit> at least one of which powers a single outlet
261 2015-06-30 23:54:37 <phantomcircuit> er this is ot for this channel
262 2015-06-30 23:54:38 <phantomcircuit> my bad
263 2015-06-30 23:55:09 <Diablo-D3> burning your house down is never ot for #bitcoin-dev.
264 2015-06-30 23:55:20 <Diablo-D3> *especially* if you do it via mining
265 2015-06-30 23:57:47 <CodeShark> petertodd: the more useful information we have for fee estimation the better, I think :)
266 2015-06-30 23:58:10 <CodeShark> from a usability perspective nothing would be better than just guessing a near optimal fee
267 2015-06-30 23:58:37 <petertodd> CodeShark: indeed
268 2015-06-30 23:59:09 <petertodd> CodeShark: well, even better, is to have lightning where fees are predictable :P
269 2015-06-30 23:59:24 <CodeShark> well, yes...of course...in due time
270 2015-06-30 23:59:35 <CodeShark> it's inevitable and you know it ;)