1 2013-03-24 00:05:23 <owowo> OMG! 50btc only 1199!! I GOT BUY!
2 2013-03-24 00:06:36 <SwedFTP> Has there ever been any discussion on associating domains with an address possibly via a TXT record? I've heard that in the earlier days you were able to send bitcoin to IP's running a client, however I haven't heard whether or not anything has ever been implemented or talked about like I asked above.
3 2013-03-24 00:07:20 <gavinandresen> SwedFTP: https://gist.github.com/gavinandresen/4120476 <-- I'm near the end of implementing this
4 2013-03-24 00:07:42 <gmaxwell> owowo: are you kidding or are you a shill for the scammer? :P
5 2013-03-24 00:08:32 <BlueMatt> gavinandresen: I wrote you a memory-limited mempool, and am working on a better worst-case algo for receiver pays
6 2013-03-24 00:08:47 <gavinandresen> BlueMatt: awesome!
7 2013-03-24 00:09:10 <gmaxwell> BlueMatt: using union-find for set coloring?
8 2013-03-24 00:09:42 <sipa> gavinandresen: hmm, refund_to is an Output?
9 2013-03-24 00:09:55 <sipa> gavinandresen: sounds like it should be bytes script
10 2013-03-24 00:10:19 <BlueMatt> gmaxwell: havnt decided exactly how Im gonna do it yet...just started working on it
11 2013-03-24 00:10:58 <gavinandresen> sipa: repeated bytes scripts? TD wants multiple everything....
12 2013-03-24 00:10:59 <gmaxwell> I still feel a little uneasy about memory-limited mempools. I wonder if past 2-3 x the current blocksize if it shouldn't use priority weighed random eviction to make flooding out a transaction less reliable.
13 2013-03-24 00:11:39 <BlueMatt> gmaxwell: hmm?
14 2013-03-24 00:12:44 <gavinandresen> yeah, hmm? mempool eviction will be least-fee or least-priority....
15 2013-03-24 00:12:48 <gmaxwell> BlueMatt: I pay you with a low priority transaction, then I flood out the mempool to make that txn get forgotten and replace it with a double spend.
16 2013-03-24 00:12:52 <gavinandresen> (I assume)
17 2013-03-24 00:13:10 <gavinandresen> gmaxwell: flood it out with high-priority transactions?
18 2013-03-24 00:13:18 <gmaxwell> gavinandresen: yes.
19 2013-03-24 00:13:19 <BlueMatt> gmaxwell: meh, why should you not be allowed to do that?
20 2013-03-24 00:13:34 <gavinandresen> ??? and anyway, better double-spend detection/reporting should be done, too. That's orthogonal
21 2013-03-24 00:13:43 <gmaxwell> BlueMatt: because it dramatically reduces the security of unconfirmed transactions.
22 2013-03-24 00:14:01 <gmaxwell> gavinandresen: the double spend would not be detected??? after all the flood out makes everyone forget the transaction.
23 2013-03-24 00:14:15 <BlueMatt> well if you have the ability to create that many high-prio txn...
24 2013-03-24 00:14:17 <gmaxwell> Thats why I was suggesting randomizing it somewhat, that way its likely _some_ node would still remember it.
25 2013-03-24 00:14:23 <BlueMatt> and anyway, you can detect that attack as the receiver
26 2013-03-24 00:14:33 <gmaxwell> BlueMatt: How?
27 2013-03-24 00:14:46 <gavinandresen> double-spend detection will have a separate big list of (txid,nOut). At least, that's what I imagine
28 2013-03-24 00:15:02 <gavinandresen> well, (txid,nOut,count). And that could be randomly evicted
29 2013-03-24 00:15:16 <BlueMatt> well, mempool size is configurable, so its not const across network, and my impl isnt entirely deterministic, though Im not sure if its enough to break such an attack
30 2013-03-24 00:15:18 <gmaxwell> BlueMatt: and besides, in any case it's too late at that point. E.g. you'd exclude accepting unconfirmed for lower value transactions than the current system does.
31 2013-03-24 00:15:23 <gavinandresen> ??? no reason to tie the double-spend detection to the mempool
32 2013-03-24 00:15:57 <comboy> these payment requests will be stored in the blockchain?
33 2013-03-24 00:16:00 <gmaxwell> gavinandresen: hm. That sounds okay to me.
34 2013-03-24 00:16:37 <BlueMatt> or you'd just have that one node on the network with 10GB mempool that will ALWAYS find double-spends and alert people
35 2013-03-24 00:16:57 <sipa> gavinandresen: so, multiple bytes? i just mean that a refund_to doesn't need an amount
36 2013-03-24 00:17:19 <sipa> comboy: absolutely not
37 2013-03-24 00:17:51 <gavinandresen> sipa: ask TD when he is on, there might have been some reason to have a value??? (e.g. "please refund 11 of the BTC to HERE, and the other 6 HERE")
38 2013-03-24 00:18:21 <gavinandresen> sipa: common case would be one Output with no value
39 2013-03-24 00:18:29 <gavinandresen> ??? no amount....
40 2013-03-24 00:19:55 <gmaxwell> gavinandresen: if me and alice want to pay you to throw a pie at sipa, we each may have ponyed up part of it.. and so if you fail and refund you should split it. Also, privacy practices might encourage keeping your wallet in certian sized coins. .. dunno if thats worth the complexity.
41 2013-03-24 00:20:46 <gavinandresen> gmaxwell: yes, I think those were the reasons for making refund multiple Outputs. And I agree, not sure it is worth the added complexity
42 2013-03-24 00:21:19 <gmaxwell> I don't know how you handle things like partial refunds.
43 2013-03-24 00:22:06 <comboy> sorry I'm jumping in with zero information, but I got so curious about it, end user experience for this is that I buy something and then in my bitcoin client I see some popup "paying for pizza press OK"? or am I getting it completely wrong?
44 2013-03-24 00:23:35 <gruez> quick question:
45 2013-03-24 00:23:41 <gavinandresen> comboy: You click on a "Pay now" link, your bitcoin client opens and yeah, it opens up the Send dialog with who you're paying (e.g. "merchant.com"), how much you're being asked to pay, and a message from the merchant.
46 2013-03-24 00:23:45 <gruez> if a pool is using GBT or stratum
47 2013-03-24 00:23:52 <gruez> does it mean the transactions are generated locally?
48 2013-03-24 00:24:13 <sipa> gruez: GBT allows the client to choose which transactions end up in the block
49 2013-03-24 00:24:23 <sipa> gruez: stratum, afaik, only allows control of the coinbase
50 2013-03-24 00:24:59 <gruez> sipa: how does the miner get transactions?
51 2013-03-24 00:25:03 <gruez> from the local bitcoind?
52 2013-03-24 00:25:06 <sipa> for example
53 2013-03-24 00:25:39 <comboy> gavinandresen: but these details are propagated through bitcoin network right? so the link only contains some hash to identify which payment request is for me?
54 2013-03-24 00:26:07 <gavinandresen> comboy: no, you get the payment request from a web page
55 2013-03-24 00:26:17 <sipa> comboy: why would this information need to be on the bitcoin network? it just goes directly from receiver to sender via HTTP(S)/...
56 2013-03-24 00:26:24 <gavinandresen> (or email or however, a payment request is just a file)
57 2013-03-24 00:27:09 <comboy> ok ok, Imagined something different, thanks
58 2013-03-24 00:29:48 <comboy> btw do you guys believe that if we go big scale merchant payments will actually be handled directly by bitcoin network and not some abstractions on top of iit (more centralized but being able to confirm payments quickly)?
59 2013-03-24 00:30:36 <sipa> comboy: that's a good question :)
60 2013-03-24 00:30:54 <sipa> comboy: i'd say opinions differ, but imho, yes, at some point transactions will need to happen off-chain
61 2013-03-24 00:31:16 <sipa> there are nice solutions though that aren't necessarily fully centralized, though
62 2013-03-24 00:32:27 <comboy> any interesting links for long afternoon? I'd love to read about these
63 2013-03-24 00:41:34 <helo> "subver" : "/Gangnam Style -minimal:0.8.0/"
64 2013-03-24 00:42:11 <sipa> 5.9.24.81
65 2013-03-24 00:42:31 <helo> ze germans
66 2013-03-24 01:32:00 <SwedFTP> gavinandresen, That is an ideal implementation of that idea, good job.
67 2013-03-24 01:47:08 <Luke-Jr> FYI: between blocks 227716 and 227717, bitcoind 0.6.0 was rejecting proposals generated by 0.8.1
68 2013-03-24 01:47:39 <gmaxwell> Luke-Jr: any idea why?
69 2013-03-24 01:48:06 <Luke-Jr> gmaxwell: it *might* be the bug in 0.6 proposals code, but I wouldn't have expected it to affect the case of two nodes on localhost
70 2013-03-24 01:48:26 <Luke-Jr> 03/24/13 02:41:18 ERROR: FetchInputs() : d396f63b5c mapTransactions prev not found b20a46a95d
71 2013-03-24 01:48:45 <gmaxwell> GAH@#$@$#@ go patch your nodes to log the full w#@$@#$*@)( ( hashes
72 2013-03-24 01:50:23 <gmaxwell> d396f63b5cee28a1f836524bfc76cc5cebf32f4c5774bb87c8c4a56ae1e0f8d4
73 2013-03-24 01:51:26 <lianj> ack
74 2013-03-24 01:51:27 <gmaxwell> b20a46a95debdfb4ebe39cc48c906000b71022d3e0f9365b0a038103c995c6e8 < so it sounds like perhaps your 0.6 node has a corrupted database that lacks that txn.
75 2013-03-24 01:51:56 <gmaxwell> Luke-Jr: is it forked as of 227717?
76 2013-03-24 01:52:17 <Luke-Jr> gmaxwell: nope
77 2013-03-24 01:52:26 <Luke-Jr> just confirmed it is an instance of the 0.6 proposal bug
78 2013-03-24 01:52:31 <Luke-Jr> sorry for the false alarm
79 2013-03-24 01:52:33 <gmaxwell> ah, b20a46a95 was also mined in that block.
80 2013-03-24 01:52:41 <gmaxwell> No, it was a good excercise.
81 2013-03-24 01:53:00 <gmaxwell> Your proposal code must be failing to check the proposal for newly created outputs.
82 2013-03-24 01:53:05 <Luke-Jr> correct
83 2013-03-24 01:53:31 <Luke-Jr> it's fixed in the 0.8 port, but 0.6.0 is dead soon enough that I didn't fix it there
84 2013-03-24 01:53:40 <gmaxwell> in any case, please fix all your code to log full transaction IDs. It was only easy to figure out the full ID here because the txn in question was just mined.
85 2013-03-24 01:59:15 <user_corrupt> hello wonderful bitcoin devs
86 2013-03-24 01:59:47 <user_corrupt> just a run of the mill web developer who is intent on participating in this historic new market that is being developed
87 2013-03-24 02:13:20 <EvilPete> block #227720 - are deepbit's clocks drifting? looks like its creation time is 9 minutes after the network received it
88 2013-03-24 02:17:20 <fishfish> hhi everyone, out of curiosity, if I want to store and use a wallet locally , just confirming I have no choice but to d/l the whole blockchain at least once?
89 2013-03-24 02:18:46 <jaakkos> fishfish: you can use an SPV client
90 2013-03-24 02:18:55 <jaakkos> (simple payment verification)
91 2013-03-24 02:19:00 <jaakkos> fishfish: such as the Android client
92 2013-03-24 02:19:27 <EvilPete> fishfish: there are lighter weight clients, eg: electrum, multibit
93 2013-03-24 02:19:55 <fishfish> thanks guys, a bit of background, trying to setup a service online similar to instawallet
94 2013-03-24 02:20:03 <fishfish> so the android client wouldnt' work
95 2013-03-24 02:20:50 <kjj> ugh. if you need to make an online wallet, you pretty much need to be running the full client, and that means maintaining the full block chain locally
96 2013-03-24 02:21:03 <fishfish> thanks kjj
97 2013-03-24 02:23:50 <user_corrupt> someone just told me that there is a limit to how many transactions per second bitcoin network can handle.... if bitcoin becomes widely accepted as a currency, can this be increased?
98 2013-03-24 02:24:16 <kjj> user_corrupt: a billion threads on that topic on the forums
99 2013-03-24 02:25:48 <gmaxwell> user_corrupt: anything can be changed with enough consensus. The harder question is what is the transaction rate it can handle before that compromises decenteralization, and how much decenteralization is enough.
100 2013-03-24 02:26:10 <user_corrupt> gotcha
101 2013-03-24 02:27:07 <petertodd> user_corrupt: The key thing is transactions don't have to happen on the blockchain - you can have someone else keep the ledger for you. More importantly, there exist protocols so those leger holders don't need to be trusted to any significant degree.
102 2013-03-24 02:27:41 <EvilPete> user_corrupt: super short version: yes it can be changed, but it is also possible to do something about the spammy transactions. the bitcoin model is people pay fees to get better processing. opinions vary widely on what the right solution is.
103 2013-03-24 02:28:05 <user_corrupt> very interesting
104 2013-03-24 02:28:55 <user_corrupt> well, i would support some kind of bitcoin.inc with servers in the caiman islands, but then that is when the rockefellers send some goons after you I guess
105 2013-03-24 02:29:54 <EvilPete> user_corrupt: I think right now the feeling that fees are not a good way to increase adoption. right now miners generally only do 250k-ish blocks and the protocol already allows for 1MB blocks. that'll buy some time but the issue has to be dealt with
106 2013-03-24 02:30:11 <gmaxwell> user_corrupt: Yea, as petertodd mentions.. transactions can be external too. Bitcoin is both a currency and a payment network. Unfortunately the goals of a trustworthy currency and scalable payment network are somewhat at odds, but you could always trade the bitcoin currency via external payment networks.
107 2013-03-24 02:31:10 <gmaxwell> (To make the currency trustworthy its supply and history must be super highly public to the whole world, but thats kinda lame from a scaling perspective, also lame for privacy)
108 2013-03-24 02:32:23 <warren> Didn't the USD supply go secret?
109 2013-03-24 02:32:29 <user_corrupt> as a soon to be bitcoin consumer, i have much more patience and understanding for the sincere technical issues that you are all facing, as opposed to the myriad monkey business that central banks are involved with in their efforts to "regulate" the conventional money supply
110 2013-03-24 02:32:44 <kjj> warren: ?
111 2013-03-24 02:33:23 <petertodd> user_corrupt: That's a very good attitude to take - know what you have invested in.
112 2013-03-24 02:34:08 <user_corrupt> also, I think that public confidence in these institutions will continue to fall like a rock
113 2013-03-24 02:34:52 <jgarzik> woo! new register allocator in gcc 4.8
114 2013-03-24 02:34:57 <jgarzik> kjj: around
115 2013-03-24 02:34:57 <petertodd> I have an analogy I like for what Bitcoin as a currency/store of value is: imagine if the banks of the world all had vaults made out of bullet proof glass, and anyone was allowed to wander the vaults counting the gold bars. (although with crypto, you can be sure the gold bars are solid too)
116 2013-03-24 02:35:15 <kjj> jgarzik: do you have a node running your UDP patch?
117 2013-03-24 02:35:21 <petertodd> jgarzik: You must be a kernel developer... :P
118 2013-03-24 02:36:18 <jgarzik> kjj: not ATM
119 2013-03-24 02:36:43 <gmaxwell> warren: fed bank loans are secret, yes. I'm not sure what the ratio of secrecy to computationally intractably obfscuated the problem is though.
120 2013-03-24 02:37:02 <kjj> I probably won't be able to compile one until Monday, but I'd be willing to test communications with you if you want to set one up. or is the code not ready for that yet?
121 2013-03-24 02:37:06 <petertodd> jgarzik: re: udp, I was also thinking it'd be good to start thinking about how we can have a "headers only" P2P network - UDP strikes me as ideal for that.
122 2013-03-24 02:37:39 <warren> I missed this, what is the purpose of a header only P2P network?
123 2013-03-24 02:38:06 <gmaxwell> warren: headers let you figure out if you're on the right chain or not??? for example.
124 2013-03-24 02:38:23 <jgarzik> petertodd: yep
125 2013-03-24 02:38:35 <petertodd> warren: SPV nodes can in many cases operate with purely the block headers. timestamping is your canonical example, but with payment protocols and UTXO proofs there are many others.
126 2013-03-24 02:41:21 <user_corrupt> im really excited to be in this room, you guys may really be changing the world here
127 2013-03-24 02:41:26 <gmaxwell> warren: e.g. I pay you. I tell you that you've been paid and give you the SPV fragment. But you're not just going to believe me??? I could be mining my own chain. So you need an independant way to confirm the identity of the best chain.
128 2013-03-24 02:41:38 <user_corrupt> this could be a quantum leap forward for humanity
129 2013-03-24 02:42:21 <hmmmstrange> shhh, dont tell anyone yet
130 2013-03-24 02:42:50 <user_corrupt> i was recently in dc watching history unfold with a congressman that I know, a protege of Ron Paul, I get the same feeling in here
131 2013-03-24 02:43:36 <petertodd> user_corrupt: Bitcoin is just money; a real quantum leap for humanity would be if we were working on cold fusion tech in here. But anyway, this is getting off topic for -dev
132 2013-03-24 02:43:41 <Luke-Jr> [03:34:12] <jgarzik> woo! new register allocator in gcc 4.8 <-- does this mean if I upgrade GCC, stupid antivirus sw won't think BFGMiner is a virus?
133 2013-03-24 02:44:12 <jgarzik> heh
134 2013-03-24 02:44:15 <warren> user_corrupt: I kind of get the sense that the developers are most concerned with getting things technically right. The trolltalk forum is where you can find the people who care about the political stuff.
135 2013-03-24 02:44:38 <user_corrupt> heh, sorry guys, just a bit excited here
136 2013-03-24 02:45:52 <petertodd> user_corrupt: Talking about politics on IRC is a particularly effective DoS attack against developer wrist bandwidth.
137 2013-03-24 02:46:41 <user_corrupt> i apologize, I'll keep my sentimentality in check then
138 2013-03-24 02:46:47 <petertodd> user_corrupt: Thanks
139 2013-03-24 02:52:55 <daemon> Hey all has anyone here got an Avalon, multiverse, Bflylabs or other ASIC miner
140 2013-03-24 02:53:07 <jgarzik> What is an ASIC miner?
141 2013-03-24 02:53:18 <daemon> I was led to understand that some of you devs had managed to purchase one of the avalon(s) or be gifted one when there initial batch was created
142 2013-03-24 02:53:33 <jgarzik> Who led you on like that?
143 2013-03-24 02:53:37 <daemon> jgarzik, A bitcoin miner based around an ASIC
144 2013-03-24 02:53:40 <daemon> jgarzik, SomeoneWeird
145 2013-03-24 02:53:56 <BlueMatt> jgarzik has an Avalon
146 2013-03-24 02:54:05 <BlueMatt> (sorry jgarzik)
147 2013-03-24 02:54:47 <daemon> I am just trying to prove a little something to my self; my original question to #mtgox: http://pastebin.com/raw.php?i=R38YLHHR
148 2013-03-24 02:55:28 <gmaxwell> daemon: avalon has shipped their initial batch to a bunch of people, ??? now, thats on promise that they'll deliver in the future, of course.
149 2013-03-24 02:55:35 <gmaxwell> s/on/no/
150 2013-03-24 02:56:05 <jgarzik> ACTION has never heard of multiverse (outside of Snowcrash)
151 2013-03-24 02:56:12 <Luke-Jr> ^ same
152 2013-03-24 02:56:18 <daemon> gmaxwell, yes that is my suspicion; see I cannot see why anhyone who developed such a system would sell it on; it would be like selling the goose that ays the golden egg
153 2013-03-24 02:56:38 <Luke-Jr> I'm pretty confident BFL isn't a scam, and there is zero evidence suggesting they are
154 2013-03-24 02:56:50 <Luke-Jr> Avalon, on the other hand.. I *hope* isn't a scam.
155 2013-03-24 02:56:57 <daemon> Luke-Jr, gmatteson__ http://www.multiversecomputers.com/Mining-Products.html
156 2013-03-24 02:57:06 <daemon> Luke-Jr, google would suggest otherwise
157 2013-03-24 02:57:19 <Luke-Jr> daemon: that website for multiverse certainly looks scammy
158 2013-03-24 02:57:24 <Luke-Jr> daemon: Google isn't God
159 2013-03-24 02:57:35 <gmaxwell> daemon: Oh lots of reasons, mining is quite risky. And your hardware compets with itself. Also if all the hashpower ends up in one parties hands it basically defeats the purpose of bitcoin, undermines trust, and makes the coins you mined worthless.
160 2013-03-24 02:58:04 <warren> "We are too backed up on US shipping, to accept any more US purchases. We should be accepting US purchases in 24 to 48hrs. International Orders are not affected at this time. Terminals for US Orders are closed during this time." Sounds legit.
161 2013-03-24 02:58:14 <daemon> gmaxwell, my worry was that if all is well in ASIC land and everyone is telling the truth there will be far to much supply vs demand
162 2013-03-24 02:58:32 <jgarzik> daemon: supply is highly constrained at the moment
163 2013-03-24 02:58:55 <gmaxwell> You cannot just purchase an asic device, you can get in a long preorder line right now with uncertian delivery dates.
164 2013-03-24 02:59:16 <petertodd> daemon: That said, because of the hashpower ending up in one hand issue, frankly if you can take the financial risk I'd encourage you to buy mining equipment. Just don't invest more than you can afford to lose, and remember the risk is pretty high that you will lose it.
165 2013-03-24 02:59:41 <daemon> gmaxwell, I have a few friends who know alot more than me when it comes to electronics and at least FPGA development, my backup plan if the date was to long was to pay them 2000 dollars to develop my own
166 2013-03-24 02:59:46 <daemon> oops petertodd ^ erven
167 2013-03-24 02:59:50 <petertodd> daemon: FWIW I've put the funds I can afford to lose to mining in BFL equipment
168 2013-03-24 03:00:06 <warren> daemon: Even if you do receive mining hardware after months of waiting, there's no guarantee that you'll make your money back if too many people mine.
169 2013-03-24 03:00:10 <petertodd> daemon: If any ASIC isn't a scam, FPGA's won't be very viable.
170 2013-03-24 03:00:13 <Luke-Jr> daemon: you can't develop your own ASIC for $2000 or anything close to it
171 2013-03-24 03:00:39 <daemon> Luke-Jr, you can develop a basic FPGA for very little
172 2013-03-24 03:00:40 <jgarzik> hashweb.org
173 2013-03-24 03:00:41 <jgarzik> eh?
174 2013-03-24 03:00:46 <daemon> ACTION blinks
175 2013-03-24 03:00:52 <daemon> oh im through my znc
176 2013-03-24 03:00:54 <daemon> #web
177 2013-03-24 03:02:13 <gmaxwell> daemon: sure, and there are lots of people mining on FPGAs right now??? though $2000 doesn't buy you very much in fpgas.
178 2013-03-24 03:02:22 <daemon> warren, let me pose you an interesting dillema, I have for the next 3 years free power, quite literally the BAR is 300A, 3 100A lines. some is userd for other purposes so probably 3*80. I could purchase from india basic mobos and psus capable of running a single PCIe lane and gpus with a high mash/s vs there price the best combination comes to a hardware break even of 119 days
179 2013-03-24 03:02:36 <daemon> I want to exploit the fact I have free power as fast as humanly possible
180 2013-03-24 03:02:40 <daemon> I have 5000 GBP to invest
181 2013-03-24 03:02:53 <petertodd> daemon: Mining w/ ASICs is capital intensive, not power intensive.
182 2013-03-24 03:03:09 <petertodd> daemon: This will be true for at least another year or two.
183 2013-03-24 03:03:18 <gwillen> petertodd: do you really think it will last a year
184 2013-03-24 03:03:20 <warren> "free" has a cost to someone.
185 2013-03-24 03:03:25 <daemon> petertodd, I did a very long detailed spreadsheet for x86+gpu+shipping etc.... ASIC is my next stop
186 2013-03-24 03:03:40 <gwillen> petertodd: I am expecting that power costs will reach a double-digit percentage of revenue within 6 months after BFL starts shipping
187 2013-03-24 03:03:43 <gwillen> probably sooner
188 2013-03-24 03:03:58 <daemon> warren, I am a hotel manager by trade, the chain of hotels I work for has an odd deal with a contract with the power supplier basically they pay a flat rate to them for around 300 hotels
189 2013-03-24 03:04:00 <gwillen> that's very handwavy but a year would surprise me
190 2013-03-24 03:04:02 <daemon> I have a 3 year contract at this post
191 2013-03-24 03:04:04 <warren> gwillen: 6 months? I expect 1-2 months
192 2013-03-24 03:04:09 <petertodd> gwillen: Given how crazy limited the supply is, yes, besides, double-digit percentage is still mostly capital intensive.
193 2013-03-24 03:04:11 <gwillen> warren: *nods*
194 2013-03-24 03:04:49 <petertodd> gwillen: I will say though, you're totally right and I could be totally wrong.
195 2013-03-24 03:04:57 <daemon> petertodd, would you say a traditional GPU based setup with a 122 day break-even on hardware is more sane, given that if it all goes wrong and ASICS fail I can still sell the systems as systems in there own right
196 2013-03-24 03:04:58 <gwillen> petertodd: I am handwaving wildly
197 2013-03-24 03:05:08 <daemon> petertodd, or should I join the line for ASICS,
198 2013-03-24 03:05:11 <gwillen> petertodd: and I deliberately hedged with "double-digit percentage"
199 2013-03-24 03:05:29 <petertodd> gwillen: We all are. :P Though remember the next step is to do 30nm and 20nm bitcoin ASICs, which again is a huge pile of capital.
200 2013-03-24 03:05:32 <gwillen> ACTION nods
201 2013-03-24 03:05:47 <gwillen> petertodd: you don't need smaller process to totally wipe out the profit margin though
202 2013-03-24 03:05:51 <gwillen> petertodd: you just need a lot of units shipped
203 2013-03-24 03:05:54 <petertodd> daemon: Selling used GPU's while everyone else is selling used GPU's sounds crazy to me.
204 2013-03-24 03:05:57 <jgarzik> gwillen: bingo
205 2013-03-24 03:05:57 <Luke-Jr> daemon: I'd be surprised if you could obtain an ASIC befoe June
206 2013-03-24 03:06:15 <jgarzik> Luke-Jr: He could always buy a preorder on eBay for $20,000
207 2013-03-24 03:06:23 <warren> pete79: I'm surprised how much used GPU prices have been holding up.
208 2013-03-24 03:06:26 <warren> oops
209 2013-03-24 03:06:29 <warren> petertodd: ^^
210 2013-03-24 03:06:44 <daemon> Luke-Jr, my priamry concern with ASIC's is there existance in whole; I keep finding links to proof of them with broken youtube links, and except for jgarzik I do not know a single poerson in the virtual world or real world who has ever seen one
211 2013-03-24 03:06:44 <gwillen> jgarzik: that sounds ... contraindicated.
212 2013-03-24 03:06:53 <petertodd> warren: heh, yeah me too, although then again I have a co-worker who wants to buy a few thousand worth of GPU's so...
213 2013-03-24 03:07:08 <warren> petertodd: why!?
214 2013-03-24 03:07:26 <daemon> petertodd, GPU's as long as there not overclocked (like cpu's) still can perform quite reliably
215 2013-03-24 03:07:28 <petertodd> warren: He's got some crypto stuff he wants to work on and needs the computer power.
216 2013-03-24 03:07:47 <Luke-Jr> daemon: I saw one in person, but it wasn't running.
217 2013-03-24 03:07:52 <Luke-Jr> last week
218 2013-03-24 03:08:11 <petertodd> Luke-Jr: Oh yeah, your BFL trip. So they never had a working one for you to try?
219 2013-03-24 03:08:19 <Luke-Jr> petertodd: nope
220 2013-03-24 03:08:27 <MKCoin> Let's just hope ASICs don't break sha-256 too quickly :)
221 2013-03-24 03:08:28 <daemon> Luke-Jr, once again .. not a running version; from my research it seems to point to the fact that ASIC development is highly volatile it can cost millions in investment to get them right
222 2013-03-24 03:08:32 <MKCoin> or .. hope?
223 2013-03-24 03:08:52 <daemon> Luke-Jr, my concern is that if they are real; they will flood the market and BTC price will fall out of the sky
224 2013-03-24 03:08:57 <petertodd> Luke-Jr: Sucks. I noticed Josh's last update was March 15th
225 2013-03-24 03:09:02 <MKCoin> Depends on how fast they would obsolete themselves
226 2013-03-24 03:09:07 <gwillen> daemon: there's no reason for BTC price to drop with ASICs
227 2013-03-24 03:09:09 <daemon> Luke-Jr, or the other concern is that they are all a big scam or there was very few made for a extremely large price and no more will be
228 2013-03-24 03:09:12 <gwillen> daemon: increasing hashrate does not increase supply
229 2013-03-24 03:09:20 <gwillen> daemon: supply is a constant 25BTC / 10 minutes
230 2013-03-24 03:09:21 <warren> daemon: how much have you studied how bitcoin works? Asics have nothing to do with more supply.
231 2013-03-24 03:09:41 <petertodd> daemon: The economics of ASIC production mean making just a few is crazy. That however is orthogonal to selling them to others...
232 2013-03-24 03:09:42 <lianj> gwillen: totally right
233 2013-03-24 03:09:47 <gmaxwell> daemon: as mentioned, avalon has shipped their first batch. They're not currently a scam.
234 2013-03-24 03:10:26 <Luke-Jr> daemon: there's no reason to think Bitcoin price would fall much
235 2013-03-24 03:10:30 <warren> daemon: most people here are worried that too many ASIC's will be quickly sold, driving average revenue down quickly.
236 2013-03-24 03:10:33 <daemon> gwillen, warren I believed I had studied everything I seem to have skipped the rate limitation; but that does mean that as each block is released the asic users are more likely to crypt them right?
237 2013-03-24 03:10:43 <gwillen> daemon: correct
238 2013-03-24 03:10:47 <Luke-Jr> daemon: you *can't* make "very few" ASICs :p
239 2013-03-24 03:10:58 <gwillen> daemon: that comes at the expense of other miners
240 2013-03-24 03:11:01 <petertodd> gmaxwell: Granted, it is possible that Avalon is having problems getting their second batch of ASIC's made - they're a lots of low-cost silicon design after all.
241 2013-03-24 03:11:02 <daemon> I see
242 2013-03-24 03:11:03 <gwillen> daemon: but the supply will not rise
243 2013-03-24 03:11:14 <daemon> that makes a little more sence; how did I miss that -_-"
244 2013-03-24 03:11:17 <Luke-Jr> gmaxwell: my concern with Avalon is that they seem to be paying for batch 1 with batch 2 preorders etc
245 2013-03-24 03:11:25 <gmaxwell> Luke-Jr: seems like it.
246 2013-03-24 03:11:36 <lianj> haha
247 2013-03-24 03:11:40 <warren> I haven't seen where this "88 BTC" cost of batch #3 is coming from, but that'd make sense if they want to pay for batch #2...
248 2013-03-24 03:11:44 <gmaxwell> Luke-Jr: but it's not clear how much of 1 was paid with 2's preorders. Clearly not the whole amount.
249 2013-03-24 03:12:01 <petertodd> Does anyone know what actual fab Avalon contracted with?
250 2013-03-24 03:12:03 <gmaxwell> and they increased #2s price.
251 2013-03-24 03:12:06 <gmaxwell> petertodd: TSMC.
252 2013-03-24 03:12:18 <gmaxwell> petertodd: they posted their contracts with the prices blacked out.
253 2013-03-24 03:12:31 <gmaxwell> well, mostly blacked out. :)
254 2013-03-24 03:12:31 <Luke-Jr> gmaxwell: sure, just makes it fishy IMO
255 2013-03-24 03:12:41 <daemon> out of curiosity can anyone validate that this was real at one point: http://fpgamining.com/
256 2013-03-24 03:12:50 <Luke-Jr> daemon: yes
257 2013-03-24 03:12:56 <Luke-Jr> daemon: I have their X6500 product
258 2013-03-24 03:13:06 <Luke-Jr> and we all know TheSeven
259 2013-03-24 03:13:07 <daemon> Luke-Jr, how does it perform at present
260 2013-03-24 03:13:33 <gmaxwell> Luke-Jr: I suspect what they did is under compute their costs, and run short when it came to assembly/passives/etc. I noticed the assembly of my units was not terribly uniform, and that there was a parts change.
261 2013-03-24 03:13:56 <petertodd> gmaxwell: Hand assembled?
262 2013-03-24 03:14:08 <Luke-Jr> daemon: yep, hashing away
263 2013-03-24 03:14:15 <Luke-Jr> XBS 0: 43.0C/44.9C | 382.0/389.5/381.0Mh/s | A:19257 R: 77 HW: 221 U: 5.32/m
264 2013-03-24 03:14:24 <daemon> pretty nice
265 2013-03-24 03:15:06 <gmaxwell> petertodd: the pcbs are clearly done in a nice oven... must have taken them a while to work out the profiles on these things. But the mechnicals are all hand assembled, though it seems they did a reasonable job designing to simplify that.
266 2013-03-24 03:15:53 <petertodd> gmaxwell: Ah. I noticed the picture on their website almost looks like the IC's were hand-soldered.
267 2013-03-24 03:16:03 <petertodd> gmaxwell: Heh, I wonder if it's RoHS compliant...
268 2013-03-24 03:24:45 <daemon> petertodd, gmaxwell, gwillen: Would you at the present point buy a GPU based miner (hardware break even at current rate: 100 days) given that if ASIC's take over and make them pointless you can re sell them on ebay as reasonable systems and given that you have free power, ... would you?
269 2013-03-24 03:25:13 <daemon> the buy in per system would be .. =SUM(G11:K11) dollars
270 2013-03-24 03:25:14 <daemon> err
271 2013-03-24 03:25:18 <petertodd> daemon: Heck no. My time is worth too much to dick around with selling on ebay.
272 2013-03-24 03:25:19 <daemon> 510 dollars
273 2013-03-24 03:25:32 <gwillen> yeah, == petertodd
274 2013-03-24 03:25:45 <daemon> so the systems are worthless to buy and I would have to immediatly sell them on ebay
275 2013-03-24 03:26:23 <petertodd> Speaking of, I should stop wasting time on IRC. :P
276 2013-03-24 03:26:24 <petertodd> later
277 2013-03-24 03:26:29 <daemon> later petertodd
278 2013-03-24 04:08:23 <EvilPete> 91.6% version 2
279 2013-03-24 04:08:47 <jgarzik> maybe under 24 hours
280 2013-03-24 04:12:10 <gmaxwell> The right question is probably "how many consecutive v2 blocks do we need now to trigger it".. and "what are the chances that 95% of the hashpower gets N in a row"
281 2013-03-24 04:13:10 <Luke-Jr> I reserve the right to be in a grumpy mood if v2 blocks somehow hardforks
282 2013-03-24 04:20:01 <MC1984> some sortof countdown going on in here?
283 2013-03-24 04:20:25 <Graet> yes, some sort ;)
284 2013-03-24 04:20:39 <Graet> https://bitcointalk.org/index.php?topic=154521.0
285 2013-03-24 04:26:54 <MC1984> oh thats exciting
286 2013-03-24 04:27:23 <MC1984> i hope it doesnt blow up in our faces somehow when the threshold is reached
287 2013-03-24 04:28:17 <MC1984> lol deepbit is heading to be orphaned
288 2013-03-24 04:28:21 <digitalmagus> Question: So with the ASIC mining rigs coming online now and over the next year or so, those people are going to make lots of $$, and good for them; however, further down the timeline, as the number of bitcoin rewards deminish, there comes a point where even running an ASIC mining rig, or racks and racks of these becomes unprofitable and they quit. Given ASIC is the end of the technology
289 2013-03-24 04:28:22 <digitalmagus> improvements possible (as far as I know), and given P2P transactions rely on miners to process the transactions, if there are no miners, how will transactions occur??
290 2013-03-24 04:28:26 <Luke-Jr> no, Deepbit is v2
291 2013-03-24 04:28:41 <MC1984> hang on ie got a teeny tiny violin around here somewhere...
292 2013-03-24 04:30:13 <CodeShark> digitalmagus: "there comes a point where even running an ASIC mining rig, or racks and racks of these becomes unprofitable and they quit." - if they all quit then guess what? suddenly mining is super profitable again because the difficulty drops!
293 2013-03-24 04:30:31 <digitalmagus> ahhh ok got it. that makes sense. thanks :)
294 2013-03-24 04:30:42 <MC1984> i think mining is desined to eentually be baely profitable
295 2013-03-24 04:30:57 <MC1984> which means mining is free not counting the cost of hardware
296 2013-03-24 04:31:10 <warren> digitalmagus: if you have tons of hardware that is useless for anything but mining, lots of people will mine at break-even cost (or even less)
297 2013-03-24 04:31:39 <EvilPete> mining is exactly that.. steady state is just profitable enough to keep the dedicated working. there is a huge advantage to be had if you can get ahead of the curve though, eg: an early ASIC online
298 2013-03-24 04:31:57 <digitalmagus> warren... or less, I doubt it. I think this is all leading to a handful of data centers jam packed with huge ASIC mining rigs that only a few millionaires can afford.
299 2013-03-24 04:32:16 <MC1984> i thought that but ie been told its not as likely as it seems
300 2013-03-24 04:32:32 <digitalmagus> MC1984: how so?
301 2013-03-24 04:32:37 <MC1984> not entirely convinced though
302 2013-03-24 04:33:06 <MC1984> datacentre efficiency gains dont apply to bitcoin as mining as much as other computing i think
303 2013-03-24 04:33:11 <MC1984> i cant remember how though
304 2013-03-24 04:33:50 <EvilPete> Its far more cost effective to park a few asic miners in your office at work, assuming they're quiet enough.
305 2013-03-24 04:33:58 <digitalmagus> I only say data center because most people don't have enough dedicated power outlets to run 1,000 x 1500 watt boxes
306 2013-03-24 04:33:59 <MC1984> even so, if eficiancy gains wont do it, the fincen thing will
307 2013-03-24 04:34:46 <MC1984> i seriously want a miner providing resistive heating in every house in the land/world
308 2013-03-24 04:35:28 <digitalmagus> EvilPete: Yeah but the thing is... your $1200 ASIC box at 60Gh/h will get 1x 25 block every 10 years if you are competing against rich guys that have a data center filled with asic rigs doing 500 TeraHash/s ... EACH
309 2013-03-24 04:36:20 <MC1984> like, if/when asic bicoin chips are pennies, mining heates should turn you a nice profit considering mining profit cancels out expenditure at best, and then you dont hae to pay to heat your house otherwise
310 2013-03-24 04:36:25 <MC1984> = free electric heating maybe
311 2013-03-24 04:36:47 <MC1984> gas companies gonna be butthurt
312 2013-03-24 04:36:57 <SomeoneWeird> yehlol
313 2013-03-24 04:37:00 <EvilPete> digitalmagus: yeah, but so will the millionaires' machines. there's no way they're going to overspend on that.
314 2013-03-24 04:37:18 <EvilPete> digitalmagus: the same thing happened on the cpu->gpu transition. the world didn't end.
315 2013-03-24 04:37:32 <EvilPete> and the early adopters made a killing.
316 2013-03-24 04:37:35 <digitalmagus> No the world didn't end because everybody has a video card in their PC
317 2013-03-24 04:37:50 <Graet> not really
318 2013-03-24 04:37:59 <user_corrupt> if bitcoin ends up getting used much more people as a regular currency, could it be possible to increase the speed of transactions but also keep it decentralized so that there is no single server that any government can audit to keep track of who is buying what?
319 2013-03-24 04:38:00 <Graet> and lots had underperforming nvidia
320 2013-03-24 04:38:08 <MC1984> what really scares me is the stories about asic miner
321 2013-03-24 04:38:16 <jgarzik> user_corrupt: sure. Use off-chain transactions.
322 2013-03-24 04:38:24 <MC1984> i didnt consider miners outsourcing the actual equipment and labour
323 2013-03-24 04:38:31 <MC1984> to a single coercible firm
324 2013-03-24 04:38:33 <MC1984> jesus fuck
325 2013-03-24 04:38:35 <EvilPete> I'm not sure nvidia gpus really count as a miner.. their microarchitecture really isn't suited
326 2013-03-24 04:38:55 <warren> MC1984: they will be the minority soon enough
327 2013-03-24 04:38:56 <user_corrupt> jgarzik: cool, thanks
328 2013-03-24 04:39:04 <MC1984> ?
329 2013-03-24 04:39:05 <digitalmagus> people... which may dwindle down to a few dozen or few hundred people.
330 2013-03-24 04:39:05 <digitalmagus> The world isn't going to end in the next 2 years because some can still afford $1500 ASIC mining rig that will be profitable... but after say 5 years, you need a $50,000 ASIC rig to be profitable... and now we're talking about a couple dozen thousand people... then 2 or 3 years later you need $500,000 mining rigs... at which point, as I was saying, all mining will be done only by very rich
331 2013-03-24 04:39:16 <MC1984> i read earlier asicminer has 50th ready to go
332 2013-03-24 04:39:24 <MC1984> from purchased shares
333 2013-03-24 04:39:46 <warren> digitalmagus: you're assuming the "$1500 ASIC mining rig that will be profitable" is true.
334 2013-03-24 04:40:12 <EvilPete> digitalmagus: people won't spend $500,000 on a mining rig unless it can pay for itself in an acceptable timeframe.
335 2013-03-24 04:40:40 <EvilPete> and the longer it takes, the bigger the risk it'll be a loss. mining is far from a "sure thing"
336 2013-03-24 04:40:59 <warren> digitalmagus: if people pre-order too many, average revenue per ASIC will go down significantly. Then future ASIC's will be cheaper, further driving down revenue per ASIC.
337 2013-03-24 04:41:01 <MC1984> LOL ask bfl investors
338 2013-03-24 04:41:04 <MC1984> imean customers
339 2013-03-24 04:41:41 <digitalmagus> EvilPete... Oh but they will... it's an escalating thing... those who can have the most Thashes/sec first, will be able to recoup their money faster, thus it is inevitable that many years down the road, we'll see gigantic ASIC mining rigs.. look BFL vapourwear already claims 1.5 Thashes/sec for $30,000 ... you don't think they are going to make a $60,000 box next?
340 2013-03-24 04:42:35 <CodeShark> digitalmagus: are you saying that for every doubling in price of an ASIC miner, the hashrate more than doubles?
341 2013-03-24 04:43:17 <digitalmagus> CodeShark, let me do some quick math
342 2013-03-24 04:43:55 <MC1984> Graet what pool do you op?
343 2013-03-24 04:44:16 <CodeShark> or actually, the correct metric is hashrate per unit of power consumption
344 2013-03-24 04:44:28 <Graet> Ozcoin MC1984
345 2013-03-24 04:45:34 <CodeShark> seems hard to believe - since price correlates with size of chip
346 2013-03-24 04:45:46 <CodeShark> many small chips are easier to manufacture than a few large ones
347 2013-03-24 04:45:46 <MC1984> ar you giving up running a pool or am i reading that thread wrong
348 2013-03-24 04:46:10 <Graet> i guess you are reading a thread wrong
349 2013-03-24 04:46:37 <user_corrupt> what would happen if someone creates quantum computing? (naive question from non computer scientist)
350 2013-03-24 04:46:49 <CodeShark> large chips are much more likely to run into defects in the wafers
351 2013-03-24 04:46:56 <MC1984> yeah
352 2013-03-24 04:47:12 <warren> user_corrupt: Google "lamport signature"
353 2013-03-24 04:47:31 <Graet> so many threads ...
354 2013-03-24 04:47:32 <MC1984> "ozcoin is disabling getwork entirely."
355 2013-03-24 04:47:35 <MC1984> confused me
356 2013-03-24 04:47:52 <MC1984> probably because i dont understand what getwork is properly, or its being replaced
357 2013-03-24 04:47:59 <Graet> ok, because we mainly use stratum now'
358 2013-03-24 04:48:21 <Graet> the way the miner talks to the pool
359 2013-03-24 04:48:45 <MC1984> i thouht stratum was just for electrum servers and stuff
360 2013-03-24 04:48:46 <Graet> our old setup couldnt handle v2 blocks without erxtensive work
361 2013-03-24 04:49:09 <Graet> not just
362 2013-03-24 04:49:22 <digitalmagus> CodeShark: Well in theory even if the size of the chips cost the same, economics would suggest that scales of economies are reached. The more chips get produced the cheaper they become. Also if you analize BFL's pricing structure, you'll realize that the Jalapeno 4.5Gh/s costs $33.11 per GH, where as their top of the line SC $30,000 box costs $19.93 per Gh/s.... and they are both using
363 2013-03-24 04:49:22 <digitalmagus> the exact same chip. So while BFL gets lower margin profits on the high end boxes, they get many more total dollars in revenue. But yeah, eventually they will go to smaller nanometer dies, and pack even more chips etc.
364 2013-03-24 04:49:38 <Graet> btcguild, slushpool ozcoin and others use it for mining
365 2013-03-24 04:49:44 <EvilPete> digitalmagus: the bottom line is that the price that makes sense to spend on hardware varies with the value of the BTC it generates. If BTC goes through the roof then 25BTC/block suddenly is worth a lot more.
366 2013-03-24 04:50:31 <digitalmagus> EvilPete: Yes that's true, but awards are given to whomever solves the hash first... and only the rich with huge rigs will have the highest probability BY FAR.
367 2013-03-24 04:50:36 <doublec> MC1984: http://mining.bitcoin.cz/stratum-mining describes it
368 2013-03-24 04:51:17 <CodeShark> digitalmagus: that's why pools exist :)
369 2013-03-24 04:51:31 <EvilPete> pools exist to even out bad luck
370 2013-03-24 04:51:55 <digitalmagus> EvilPete: That's a good point actually.
371 2013-03-24 04:52:58 <digitalmagus> So let's just all hope some millionaire rich kid who got in early doesn't buy a quantum computer that can solve at 5 TRILLION terahashes/sec and control 99.99% of the hashing network including all pools. :P
372 2013-03-24 04:53:23 <EvilPete> digitalmagus: its not a race to "solve" a hash, you have to find one. If somebody has 100x the hash power of me, then they'll statistically find 100x more blocks than me. But they won't drive me to zero.
373 2013-03-24 04:53:54 <EvilPete> digitalmagus: if he can do 99.99% he's killed bitcoin outright.
374 2013-03-24 04:54:27 <CodeShark> if someone had a quantum computer capable of something like that, its effects on bitcoin would be the least of my concerns :p
375 2013-03-24 04:54:48 <digitalmagus> EvilPete: They won't bring your probability to zero, but they will bring your probability of winning a bitcoin reward to once in 50 years, at which point you make no profit with your small rig.
376 2013-03-24 04:55:05 <EvilPete> this is getting way off topic though. I'm going to back out of this. I am confident that market forces will resolve this satisfactorily.
377 2013-03-24 04:55:29 <digitalmagus> EvilPete: thanks for the discussion :)
378 2013-03-24 04:55:34 <CodeShark> digitalmagus: again, *pools*
379 2013-03-24 04:55:53 <digitalmagus> ah right pools.. ok gotta remember that :)
380 2013-03-24 04:56:11 <MC1984> the great chain
381 2013-03-24 04:56:17 <MC1984> wait i mean the invisible hand
382 2013-03-24 04:56:25 <CodeShark> pools mean you get a steady payout that is pretty close to proportional to your hashrate
383 2013-03-24 04:57:13 <CodeShark> someone with 10x the hashing power will get 10x as much as you...but you'll get paid your share immediately
384 2013-03-24 04:57:44 <digitalmagus> right. Thanks CodeShark
385 2013-03-24 04:59:38 <MC1984> god dammit p2pool is 12th
386 2013-03-24 05:00:15 <MC1984> and btc guild is rediculously strong
387 2013-03-24 05:25:15 <midnightmagic> p2pool is only 350GH right now.
388 2013-03-24 05:33:09 <MC1984> graph shows users going up but rate going way down
389 2013-03-24 05:33:10 <MC1984> s
390 2013-03-24 05:33:13 <MC1984> strange
391 2013-03-24 05:35:57 <OneMiner> People jumping into the game mining with whatever while miners that can math get out with their GPUs.
392 2013-03-24 05:44:49 <muhoo> gpus? i thought it was all asics these days?
393 2013-03-24 05:54:36 <EvilPete> fyi: block 227930 (roughly 192 from now, roughly 32 hours if 10 minute average) is the magic number. +/- a couple to allow for extra v1's to get in
394 2013-03-24 05:56:18 <vazakl> bitcoin rules
395 2013-03-24 05:56:20 <EvilPete> so probably 227940 to allow those solo or unknown groups to spit out a few more non version=2's
396 2013-03-24 05:57:37 <EvilPete> 95% supermajority should have happened in time for breakfast in the US, monday morning.
397 2013-03-24 06:03:10 <aba_> how to create a bitcoin wallet service ?
398 2013-03-24 06:03:22 <aba_> does bitcoin.org provide any api?
399 2013-03-24 06:04:00 <aba_> anybody?
400 2013-03-24 06:04:27 <aba_> anybody?
401 2013-03-24 06:05:38 <aba_> anybody?
402 2013-03-24 06:06:30 <Scrat> aba_: if you're asking how to create a wallet service you probably shouldn't do it
403 2013-03-24 06:06:44 <Scrat> and no, bitcoin.org does not provide an api. why would it?
404 2013-03-24 06:07:23 <Scrat> a centralized api goes against the principles of bitcoin
405 2013-03-24 06:08:16 <EvilPete> 12:05 ## Left #bitcoin-: aba_ (75cb7a68@gateway/web/freenode/ip.117.203.122.104)
406 2013-03-24 06:08:32 <EvilPete> gotta love those driveby types
407 2013-03-24 06:08:36 <Scrat> ACTION turns on join/parts
408 2013-03-24 06:08:41 <Scrat> thanks EvilPete :p
409 2013-03-24 06:08:54 <peawormsworth> Scrat: you can send questions like aba_ over to the #bitcoin-tech channel.
410 2013-03-24 06:09:12 <EvilPete> should proably have pointed him to coinbase.. his impatience would go well there :)
411 2013-03-24 06:09:12 <peawormsworth> although... i dont think he will find his answers in chat
412 2013-03-24 06:34:41 <afadfsfs> how to create a bitcoin service?
413 2013-03-24 06:35:18 <afadfsfs> like a wallet service? is it required to downnload the blockchain on to my server?
414 2013-03-24 06:37:55 <afadfsfs> anybody?
415 2013-03-24 06:40:08 <weex> afadfsfs: pretty much
416 2013-03-24 06:41:03 <afadfsfs> thanks :) isn't there any api which can be used to build upon a service like that of a wallet?
417 2013-03-24 06:41:43 <weex> bitcoind has a json-rpc api
418 2013-03-24 06:49:05 <afadfsfs> weex: thanks a lot!
419 2013-03-24 06:49:43 <weex> anytime
420 2013-03-24 07:00:50 <EvilPete> 178 blocks till v2 supermajority
421 2013-03-24 07:05:55 <Appendage> hello
422 2013-03-24 07:06:02 <Appendage> how do I remove encryption from my Bitcoin-Qt wallet?
423 2013-03-24 07:06:10 <Appendage> (I am on OS X)
424 2013-03-24 07:06:34 <gmaxwell> Appendage: you cannot, you can however set the password to a trivial string (like a space or empty)
425 2013-03-24 07:06:40 <Appendage> oh dear
426 2013-03-24 07:06:42 <Appendage> wel ok, thanks
427 2013-03-24 07:06:57 <EvilPete> How about export ?
428 2013-03-24 07:08:06 <EvilPete> actually, I can't count. 163 blocks till v2 supermajority
429 2013-03-24 07:17:11 <CodeShark> 921 out of the last 1000?
430 2013-03-24 07:19:53 <EvilPete> yeah, cointing back puts the 50th v1 at 226930, so supermajority happens at 227930 if there's no more v1's
431 2013-03-24 07:20:48 <EvilPete> there's still a slow trickle of v1's coming in from a few places so there might be a few more
432 2013-03-24 07:38:44 <dhrbebre> http://i.imgur.com/PfS2BZU.png ???????????????????????????????????????????????????1??????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????? ?????????????????? 109.99 ?????????????????? 209.99???????????????????????????????????????????????? ????????????????????? 599.99???????????????????????????????????????????????? ???????????????????????????????????????????????
433 2013-03-24 07:40:06 <SomeoneWeird> oh wow they're getting smarter
434 2013-03-24 07:41:47 <hserhreshr> http://redd.it/1awmxb buy 50btc free20btconly$1199
435 2013-03-24 07:41:48 <hserhreshr> http://redd.it/1awmxb buy 50btc free20btconly$1199
436 2013-03-24 07:41:50 <hserhreshr> http://redd.it/1awmxb buy 50btc free20btconly$1199
437 2013-03-24 07:41:51 <hserhreshr> http://redd.it/1awmxb buy 50btc free20btconly$1199
438 2013-03-24 07:41:55 <hserhreshr> http://redd.it/1awmxb buy 50btc free20btconly$1199
439 2013-03-24 07:41:56 <hserhreshr> http://redd.it/1awmxb buy 50btc free20btconly$1199
440 2013-03-24 07:41:58 <hserhreshr> http://redd.it/1awmxb buy 50btc free20btconly$1199
441 2013-03-24 07:42:49 <SomeoneWeird> goawai
442 2013-03-24 07:43:10 <Scrat> SomeoneWeird: y u hatin on freebitcoins.exe?
443 2013-03-24 07:43:27 <warren> hmm, Bitcon Wallet for Android seems really buggy
444 2013-03-24 08:08:21 <kapiteined> hi, what mechanism prevents the generation of the same pub/priv keypair in different clients ?
445 2013-03-24 08:08:57 <Scrat> kapiteined: statistics
446 2013-03-24 08:09:22 <Scrat> 2^256 is a massive number
447 2013-03-24 08:09:27 <kapiteined> right, the change of that happening is just to remote, right?
448 2013-03-24 08:09:33 <sipa> actually, only 2^160 matters
449 2013-03-24 08:09:41 <Scrat> or well 2^128, birthday paradox
450 2013-03-24 08:09:43 <sipa> but that is still a ridiculously large number
451 2013-03-24 08:09:50 <sipa> Scrat: collisions are of no interest
452 2013-03-24 08:09:55 <sipa> you need a preimage
453 2013-03-24 08:10:10 <Scrat> sipa: was talking more of an accidental privkey collision
454 2013-03-24 08:11:03 <madmac2501> i am looking to implement a new sha256 on cpu, but i don't fully understand the data structures
455 2013-03-24 08:11:09 <madmac2501> where can i get some documentation?
456 2013-03-24 08:11:20 <sipa> madmac2501: perhaps you should try implementing it in software first :)
457 2013-03-24 08:12:14 <madmac2501> sipa, yes, it will be in software
458 2013-03-24 08:12:38 <madmac2501> i mean using a cpu, but in software
459 2013-03-24 08:13:01 <sipa> madmac2501: http://csrc.nist.gov/publications/fips/fips180-4/fips-180-4.pdf section 6.2
460 2013-03-24 08:13:18 <sipa> madmac2501: but don't implement your own crypto, unless you know what you're doing
461 2013-03-24 08:13:23 <sipa> (and it seems you don't)]
462 2013-03-24 08:13:42 <madmac2501> sipa, well, i was asking more about this function call "sha256_func func = sha256_funcs[opt_algo];"
463 2013-03-24 08:14:24 <sipa> that sounds like code for selecting which implementation to use
464 2013-03-24 08:14:30 <sipa> where do you see that?
465 2013-03-24 08:14:34 <madmac2501> sipa, i will reuse some assembly code, just need to made the inteface call
466 2013-03-24 08:14:51 <madmac2501> in driver-cpu.c line 824
467 2013-03-24 08:15:09 <sipa> i have no idea what you're talking about
468 2013-03-24 08:15:16 <sipa> i have no driver-cpu.c on my system
469 2013-03-24 08:16:33 <madmac2501> oh, sorry, i am looking at bfgminer, not bitcoin code
470 2013-03-24 08:17:05 <kapiteined> thanks for clearing that up for me, have to go, bye
471 2013-03-24 08:17:12 <madmac2501> sipa, i just wake up, and don't think clearly :)
472 2013-03-24 08:17:18 <sipa> madmac2501: technically, that isn't a function call at all
473 2013-03-24 08:17:38 <madmac2501> sipa, yeah, i know, it is selection a function from a array
474 2013-03-24 08:17:40 <sipa> it just picks a sha256 implementation from an array of function points
475 2013-03-24 08:17:51 <sipa> *pointers
476 2013-03-24 08:18:14 <sipa> so what is your question?
477 2013-03-24 08:18:25 <madmac2501> i didn't want to paste the full call
478 2013-03-24 08:18:34 <sipa> (and why do you want sha256 in CPU?)
479 2013-03-24 08:18:55 <madmac2501> the meaning of the parameters and the meaning of "struct work"
480 2013-03-24 08:19:38 <madmac2501> it is just an exercise, i want to try new avx2 intel extension
481 2013-03-24 08:19:49 <sipa> let me look
482 2013-03-24 08:20:49 <sipa> i guess thr means threshold, a maximum value to report back
483 2013-03-24 08:21:17 <sipa> midstate, data, hash1 sound like the datastructures in getwork
484 2013-03-24 08:21:21 <madmac2501> sipa, that is info about the thread, i think
485 2013-03-24 08:21:37 <sipa> could be!
486 2013-03-24 08:21:47 <sipa> yeah, i guess
487 2013-03-24 08:21:57 <madmac2501> sipa, why data is "unsigned char data[128]" ?
488 2013-03-24 08:22:09 <sipa> the data being hashed, i assume
489 2013-03-24 08:22:15 <sipa> padded version of the block header
490 2013-03-24 08:22:18 <sipa> you're probably best off asking jgarzik or Luke-Jr
491 2013-03-24 08:23:05 <madmac2501> sipa, yes, do you know their time zone?
492 2013-03-24 08:23:18 <sipa> both are in the US
493 2013-03-24 08:23:47 <madmac2501> sipa, oh, well, then i will wait, thanks for the help
494 2013-03-24 08:27:10 <madmac2501> sipa, btw, do you know how to submit patches to bfgminer?
495 2013-03-24 08:27:32 <sipa> pull request on github, i suppose
496 2013-03-24 08:30:15 <madmac2501> sipa, thanks
497 2013-03-24 08:53:20 <EvilPete> 145 blocks to go till v2 supermajority.. not long now
498 2013-03-24 08:57:25 <madmac2501> i am looking for a description of the fields in a block and its sizes, someone can help me?
499 2013-03-24 09:00:50 <madmac2501> ok, found
500 2013-03-24 09:06:36 <htserhre4h> http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 -
501 2013-03-24 09:06:37 <htserhre4h> http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 -
502 2013-03-24 09:06:40 <htserhre4h> http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 -
503 2013-03-24 09:07:00 <htserhre4h> http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 -
504 2013-03-24 09:07:01 <htserhre4h> http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 - http://redd.it/1awp1k buy50btc free2btc only$1199 -
505 2013-03-24 09:14:29 <warren> anyone have OP access in #p2pool? spammer is going nuts there.
506 2013-03-24 09:30:58 <[7]> warren: what specifically seems buggy about it?
507 2013-03-24 09:31:12 <sipa> anyone know spanish? www.deltamonkeyforce.org lists a donation address that was used in many non-canonical sigs lately
508 2013-03-24 09:31:31 <warren> [7]: works fine if you use it, but closing the app or background behavior is odd
509 2013-03-24 09:32:29 <warren> Also rotating the phone can easily make the entire app crash if you're viewing certain screens.
510 2013-03-24 09:33:24 <Ogig> im spanish sipa, what do you need from that url?
511 2013-03-24 09:33:38 <sipa> Ogig: a contact address
512 2013-03-24 09:33:56 <sipa> Ogig: assuming they understand english, i'll contact them myself
513 2013-03-24 09:37:30 <Ogig> sipa, i can't find any contact info, not a mail or anything. There is a forum with no activity tho.
514 2013-03-24 09:37:55 <[7]> warren: rotating often reinstantiates the wholr activity, whivh might behave mostly like a restart of thr app. what does it do in the background for you? i haven't noticed problems with that yet
515 2013-03-24 09:38:43 <Ogig> sipa, it's just a joomla template with almost no real content.
516 2013-03-24 09:38:54 <warren> [7]: after you Disconnect, the connections don't go away from the pulldown
517 2013-03-24 09:39:08 <sipa> Ogig: never mind; i found an address on the whois info for that domain
518 2013-03-24 09:39:11 <warren> I want to kill the app entirely most of the time
519 2013-03-24 09:45:11 <[7]> well it tries to stay synced in the background
520 2013-03-24 09:45:32 <[7]> but you can tell it to only download blocks while charging
521 2013-03-24 09:46:11 <warren> ACTION sleep
522 2013-03-24 10:17:05 <[7]> is jgarzik awake?
523 2013-03-24 10:17:16 <[7]> or maybe nanotube?
524 2013-03-24 10:17:55 <sipa> [7]: i think think that will take a few more hours
525 2013-03-24 10:19:12 <[7]> we need a solution for the link spam problem in #bitcoin-fpga
526 2013-03-24 10:20:03 <[7]> I've talked to freenode staff, and it seems easiest to just let the bitcoin freenode group take control of that channel as well
527 2013-03-24 10:20:13 <[7]> (currently it isn't registered)
528 2013-03-24 10:25:01 <Tykling> is it safe to delete the blk000x.dat and blkindex.dat that are not inside the blocks/ folder ?
529 2013-03-24 10:25:20 <Tykling> it looks like they haven't been updated for a while
530 2013-03-24 10:25:46 <Nothing4You> is a bitcoin address that only has numbers in it possible?
531 2013-03-24 10:25:52 <[7]> Tykling: depends on whether you might want to downgrade at some point
532 2013-03-24 10:26:20 <[7]> it has paid off for several pool ops to keep blkindex.dat around during the recent fork
533 2013-03-24 10:26:22 <sipa> Nothing4You: in theory, yes
534 2013-03-24 10:26:30 <Nothing4You> ok
535 2013-03-24 10:26:37 <sipa> Nothing4You: in practice, don't know
536 2013-03-24 10:26:49 <Tykling> [7]: right, ok, but for a regular user that is less likely I guess (I kind of need the diskspace on this laptop tbh)
537 2013-03-24 10:27:01 <[7]> Nothing4You: should be fairly easy to calculate how much time you'll need to find one on average
538 2013-03-24 10:27:03 <sipa> Nothing4You: pretty sure it's possible to construct
539 2013-03-24 10:27:23 <[7]> Tykling: I know that 0.8.x doesn't use blkindex.dat anymore, not sure about the other file
540 2013-03-24 10:27:37 <sipa> Tykling: it will not gain you much diskspace
541 2013-03-24 10:27:48 <sipa> Tykling: only blkindex.dat, basically
542 2013-03-24 10:27:48 <Tykling> sipa: they are 5 gigabytes
543 2013-03-24 10:27:53 <Nothing4You> how would i calculate that?
544 2013-03-24 10:28:01 <sipa> yes, but the block files are hard linked
545 2013-03-24 10:28:09 <Nothing4You> i mean the time it needs to find one
546 2013-03-24 10:28:14 <Tykling> sipa: oh, I see
547 2013-03-24 10:28:21 <sipa> Tykling: so even they appear twice, they only take space once
548 2013-03-24 10:28:28 <sipa> but yes, you can delete them
549 2013-03-24 10:28:38 <Tykling> right, cool thanks
550 2013-03-24 10:33:42 <[7]> Nothing4You: well you can assume that if you generate random private keys, they will have randomly distributed public keys
551 2013-03-24 10:33:53 <Nothing4You> i know
552 2013-03-24 10:34:09 <[7]> now looks at the conversion of random binary data to base58 and calculate the chance to get all digits
553 2013-03-24 10:34:41 <[7]> which is 9/58 per digit
554 2013-03-24 10:35:18 <[7]> for simplicity we can assume that the checksum will behave random as well
555 2013-03-24 10:35:29 <[7]> now how long is a bitcoin address?
556 2013-03-24 10:35:34 <Nothing4You> 34 chars
557 2013-03-24 10:35:41 <Nothing4You> or shorter
558 2013-03-24 10:36:01 <Nothing4You> 34 in vanitygen regex
559 2013-03-24 10:36:05 <[7]> the first 8 bits are the version number and thus don't count
560 2013-03-24 10:36:50 <[7]> so let's say 32 chars for an overly optimistic estimation
561 2013-03-24 10:36:58 <Nothing4You> k
562 2013-03-24 10:37:21 <sipa> ;;calc (10/58)**33
563 2013-03-24 10:37:22 <gribble> 0
564 2013-03-24 10:37:29 <Nothing4You> lol
565 2013-03-24 10:37:46 <sipa> ;;calc 5.8**33
566 2013-03-24 10:37:47 <[7]> so we have one all-digits address in 1/((9/58)**32) addresses on average
567 2013-03-24 10:37:47 <gribble> 15599970876632756333314048
568 2013-03-24 10:38:12 <Nothing4You> 10/58 != 58/10
569 2013-03-24 10:38:33 <sipa> oh, 9, indeef
570 2013-03-24 10:38:46 <gribble> 504801894443496583089618944
571 2013-03-24 10:38:46 <sipa> ;;calc (58/9)**33
572 2013-03-24 10:39:28 <Nothing4You> wtf
573 2013-03-24 10:39:32 <[7]> Nothing4You: so I think we don't need a more accurate estimate to realize that we'll probably never run across one :)
574 2013-03-24 10:39:32 <sipa> Nothing4You: it's the inverse, obviously
575 2013-03-24 10:39:38 <Nothing4You> i get a different result than gribble
576 2013-03-24 10:39:45 <Nothing4You> yeah i believe so
577 2013-03-24 10:40:02 <sipa> ;;calc 33*log(58/9)/log(2)
578 2013-03-24 10:40:04 <gribble> 88.7058477916
579 2013-03-24 10:40:42 <sipa> it would require almost a million times more work to find one than has been used to build the current bitcoin chain
580 2013-03-24 10:40:47 <Nothing4You> my calculation is about 479370097175 less than gribbles result
581 2013-03-24 10:40:56 <sipa> rounding
582 2013-03-24 10:41:00 <sipa> expected
583 2013-03-24 10:41:07 <Nothing4You> rounding is bad :/
584 2013-03-24 10:41:12 <[7]> the myths of floating point numbers :(
585 2013-03-24 10:42:03 <sipa> a programmer had a problem, he thought "i'll use floating point to solve it!", now he has 1.9999997 problems
586 2013-03-24 10:42:13 <Nothing4You> lol
587 2013-03-24 10:47:32 <area> If you're having floating point problems, I feel bad for you, son. I've got 0.999998 problems but that just isn't 1.
588 2013-03-24 10:58:13 <fishfish> area: lol]
589 2013-03-24 11:29:43 <pegu> is there a site like blockchain.info which will dump the block header in hex?
590 2013-03-24 11:31:16 <asadab> how to create a bitcoin wallet service?
591 2013-03-24 11:31:29 <asadab> anybody?
592 2013-03-24 11:35:39 <sturles> asadab: You buy Bitcoin Wallet Service 1.0 from your local Wallmart, insert the CD and press Install, next, next, next, next, next, next, next, next, next, done.
593 2013-03-24 11:37:13 <starsoccer> really?
594 2013-03-24 11:37:14 <starsoccer> lol
595 2013-03-24 11:37:55 <sipa> asadab: that's like asking "how to build a rocket engine?" - it will always require a nontrivial amount of work even with the best advise possible (which will likely need to come from people who have done so already, and probably with less help)
596 2013-03-24 11:40:14 <[7]> asadab: if you don't even want to run a bitcoind, you probanly just shouldn't do it at all
597 2013-03-24 11:41:27 <[7]> actually you should only ever do such a thing if you know every single part that's involved well enough to be able to do it without any help
598 2013-03-24 11:41:46 <[7]> because otherwise you WILL screw up, especially on security aspects
599 2013-03-24 11:42:48 <[7]> I, even though I would be perfectly capable of implementing such a thing myself, wouldn't ever dare to run a public eWallet, just because of security and liability concerns
600 2013-03-24 11:46:45 <sipa> +1 [7]
601 2013-03-24 11:47:53 <[7]> asadab: if you want to get a general grasp of how this kind of thing might work, you could try out my bitcoin WebUI project
602 2013-03-24 11:47:54 <TD> sigh. my bitcoind seems to leak memory like a sieve but i can't figure out why
603 2013-03-24 11:48:08 <[7]> that is basically the a very simple eWallet minus multi-user capability
604 2013-03-24 11:48:15 <[7]> offloading all the hard work to bitcoind
605 2013-03-24 11:48:25 <[7]> TD: which version?
606 2013-03-24 11:48:30 <TD> current
607 2013-03-24 11:48:42 <[7]> what order of magnitude of a leak?
608 2013-03-24 11:49:42 <[7]> mine is using up 387MB
609 2013-03-24 11:50:00 <[7]> running since a few days, but next to no RPC accesses
610 2013-03-24 11:50:24 <TD> i've been trying it with tcmalloc to see if it makes any difference
611 2013-03-24 11:50:38 <TD> it usually bloats up to 1gig and then gets OOM killed within about 24 hours
612 2013-03-24 11:50:57 <TD> it's possible that tcmalloc is the thing that's killing it, but i was having trouble before that also
613 2013-03-24 11:51:01 <doublec> I find it gets to 1 gig-ish then sits there
614 2013-03-24 11:51:03 <[7]> how many p2p connections? is it getting accessed over RPC a lot?
615 2013-03-24 11:51:22 <TD> no RPC access. i've just set it running again, will track p2p connection count this time
616 2013-03-24 11:51:46 <[7]> TD: if it isn't a public node then p2p connection count will be low enough to not matter
617 2013-03-24 11:51:53 <TD> it's a public node
618 2013-03-24 11:51:58 <TD> basically i want to optimize it so it uses less memory
619 2013-03-24 11:52:50 <Scrat> 1 month uptime, 360MB here
620 2013-03-24 11:53:01 <TD> Scrat: is that accepting connections?
621 2013-03-24 11:53:02 <[7]> that can't be a 0.8.x then :)
622 2013-03-24 11:53:14 <Scrat> TD: nop
623 2013-03-24 11:53:17 <[7]> or at least not 0.8.1
624 2013-03-24 11:53:21 <sipa> TD: i've been wanting to merge jeff's patches, so we get rid of those huge per-connection buffers
625 2013-03-24 11:53:38 <TD> i tried adding a vector().swap() call but it didn't seem to do much
626 2013-03-24 11:53:48 <[7]> but seriously... how many connections do you need for 600MB of additional RAM usage=
627 2013-03-24 11:53:52 <[7]> s/=/?
628 2013-03-24 11:53:54 <sipa> TD: but it causes segfaults ~1 per day
629 2013-03-24 11:55:11 <TD> yes, well i'm hoping for a simpler change
630 2013-03-24 11:58:13 <Jere_Jones> sipa: What do you mean by "non-canonical signature"?
631 2013-03-24 12:01:59 <Jere_Jones> sipa: Nevermind. I found your pull request so I get it now.
632 2013-03-24 12:04:00 <Luke-Jr> madmac2501: the midstate/data/hash1 are as originally used in the getwork protocol: https://en.bitcoin.it/wiki/Getwork
633 2013-03-24 12:05:24 <Luke-Jr> Joric: ping
634 2013-03-24 12:05:34 <Joric> hey!
635 2013-03-24 12:05:43 <Luke-Jr> Joric: 04:43:38 Withdraw error. Out of reserve.
636 2013-03-24 12:05:44 <Luke-Jr> :/
637 2013-03-24 12:06:03 <Joric> lose a few coins
638 2013-03-24 12:06:14 <Luke-Jr> I don't want to lose XD
639 2013-03-24 12:07:15 <Luke-Jr> header sez Reserve: 2.07 BTC and Balance: 1173 1 coin = 0.01 BTC
640 2013-03-24 12:08:05 <Joric> well sorry i'm not an american federal reserve bank
641 2013-03-24 12:08:12 <Joric> i wish
642 2013-03-24 12:08:58 <Joric> did you crack it or just got lucky?
643 2013-03-24 12:09:02 <Luke-Jr> lucky :p
644 2013-03-24 12:09:17 <Luke-Jr> how about a feature to withdraw only part of my balance? <.<
645 2013-03-24 12:09:38 <madmac2501> Luke-Jr, oh, thanks, as an exercise i am trying to use avx intel extensions, the first aproach was totally slow like 0.03 M/s, i am using some intel asm plus some c code from other guy, without using midstate
646 2013-03-24 12:10:02 <Joric> i'd implement any feature before i got a job :(
647 2013-03-24 12:10:08 <Luke-Jr> madmac2501: yeah, without midstate you won't be able to get nearly as fast
648 2013-03-24 12:10:27 <sipa> still, the effect of midstate shouldn't be more than a 2x speedup
649 2013-03-24 12:10:51 <Joric> btw good job with 0.8!
650 2013-03-24 12:11:15 <Joric> always knew the lack of codereview will destroy civilization
651 2013-03-24 12:11:21 <madmac2501> Luke-Jr, anyway it was just a very bad implementation, this would need to be adapted to full asm
652 2013-03-24 12:11:28 <Luke-Jr> Joric: how about letting Withdraw do as much as it can, rather than just failing entirely? :p
653 2013-03-24 12:11:29 <Joric> was gavin on vacation
654 2013-03-24 12:11:41 <sipa> Joric: what are you talking about?
655 2013-03-24 12:12:11 <Joric> sipa, the blockchain forking in 0.8
656 2013-03-24 12:12:26 <sipa> well, 0.8's fault was not copying a bug...
657 2013-03-24 12:12:33 <sipa> no code review would have exposed that