1 2011-11-26 00:29:35 <Diablo-D3> http://www.patrick-wied.at/projects/html-qr/
2 2011-11-26 00:39:26 <graingert> Diablo-D3: definately should have used SVG
3 2011-11-26 00:39:46 <Diablo-D3> I dunno, I find that interesting
4 2011-11-26 00:40:01 <luke-jr> IE supports SVG?
5 2011-11-26 00:40:07 <graingert> yup
6 2011-11-26 00:40:26 <graingert> IE is pretty much HTML5 compliant
7 2011-11-26 00:40:33 <theymos> I think IE7 and later do.
8 2011-11-26 00:40:35 <graingert> apart from SVG webfonts
9 2011-11-26 00:40:48 <graingert> otherwise just shim chrome-frame
10 2011-11-26 00:41:24 <theymos> IE9 gets 100/100 on acid3.
11 2011-11-26 00:41:48 <graingert> bull
12 2011-11-26 00:42:09 <graingert> http://ie.microsoft.com/testdrive/benchmarks/acid3/default.html
13 2011-11-26 00:42:31 <gmaxwell> Firefox is 100/100 with current nightlies, dunno about earlier versions.
14 2011-11-26 00:43:16 <gmaxwell> apparently IE claims 100 now but the output is wrong.
15 2011-11-26 00:43:36 <graingert> I'm getting 100/100
16 2011-11-26 00:43:41 <graingert> on firefox 8
17 2011-11-26 00:43:45 <graingert> but it's not smooth
18 2011-11-26 00:44:07 <graingert> much smooth on refresh (intentional part of the test)
19 2011-11-26 00:44:12 <graingert> smoother*
20 2011-11-26 00:44:30 <gmaxwell> Yea yo uneed to test it with refresh, otherwise network latency goofs it up.
21 2011-11-26 00:44:41 <theymos> I just tried it on IE and got 100/100: http://i.imgur.com/6Q4f5.png I've heard it still technically fails, though.
22 2011-11-26 00:45:11 <gmaxwell> theymos: yea, thats failing, there is no shadow on the text for example.
23 2011-11-26 00:45:19 <graingert> the reference rendering is in html
24 2011-11-26 00:45:35 <graingert> lol I just got 99
25 2011-11-26 00:45:43 <gmaxwell> graingert: partially html.
26 2011-11-26 00:45:46 <gmaxwell> partially an image.
27 2011-11-26 00:45:50 <gmaxwell> e.g. http://acid3.acidtests.org/reference.png
28 2011-11-26 00:46:11 <luke-jr> graingert: is it XHTML5 compliant?
29 2011-11-26 00:46:20 <graingert> is what?
30 2011-11-26 00:46:37 <luke-jr> IE
31 2011-11-26 00:47:19 <graingert> well I would presume so - as that's easier than the lax parsing that html requires
32 2011-11-26 00:47:56 <luke-jr> IE has historically had issues with XML
33 2011-11-26 00:48:11 <graingert> IE9 is satisfactory
34 2011-11-26 00:49:16 <theymos> I don't mind recent IE versions at all. It's standards-compliant, pretty fast, and it has an OK UI. If I didn't care about extensions, I'd probably use IE instead of Firefox or Chrome.
35 2011-11-26 00:50:46 <luke-jr> data:application/xhtml+xml,<!DOCTYPE html5><html><head/><body>foo</body></html>
36 2011-11-26 00:50:48 <luke-jr> try that in IE9
37 2011-11-26 00:51:13 <luke-jr> very simple standards-compliant page
38 2011-11-26 00:53:09 <theymos> No, it doesn't support that. I think it does support the common base64 data: URIs, though.
39 2011-11-26 00:53:25 <graingert> you mean it doesn't work?
40 2011-11-26 00:53:29 <graingert> or it parses like html
41 2011-11-26 00:53:55 <luke-jr> graingert: historically, IE will insist on downloading XHTML
42 2011-11-26 00:54:28 <luke-jr> note that the application/xhtml+xml MIME type is required by XHTML
43 2011-11-26 00:54:41 <luke-jr> ie, it's illegal to serve it as text/html
44 2011-11-26 00:54:52 <graingert> illegal?
45 2011-11-26 00:54:56 <luke-jr> yes
46 2011-11-26 00:54:58 <graingert> or invalid
47 2011-11-26 00:55:09 <luke-jr> same thing
48 2011-11-26 00:55:57 <graingert> seems to be referred to as illegal in the spec
49 2011-11-26 00:56:01 <graingert> seems an odd term to use
50 2011-11-26 00:56:30 <luke-jr> all a spec is, is a set of laws.
51 2011-11-26 00:56:54 <luke-jr> in the context of XHTML, where those laws are strictly enforced, "illegal" seems quite fitting
52 2011-11-26 00:57:08 <graingert> meh
53 2011-11-26 00:59:04 <luke-jr> also, it's illegal to display anything if the XHTML fails to parse strictly
54 2011-11-26 00:59:05 <luke-jr> ;)
55 2011-11-26 00:59:20 <graingert> can't you display an error?
56 2011-11-26 00:59:29 <graingert> also chrome displays up until the error
57 2011-11-26 01:00:18 <roconnor> hi sipa live from room 2^10
58 2011-11-26 01:00:26 <graingert> ?
59 2011-11-26 01:00:38 <luke-jr> graingert: then Chrome is non-compliant
60 2011-11-26 01:01:27 <graingert> chrome must burn
61 2011-11-26 01:01:50 <FellowTraveler> hi all.
62 2011-11-26 01:05:38 <cjdelisle> 21:09 < luke-jr> also, it's illegal to display anything if the XHTML fails to parse strictly <-- which would break like 35% of the web
63 2011-11-26 01:05:43 <roconnor> FellowTraveler: Hi FellowTraveler; I'm quite interested in OT!
64 2011-11-26 01:06:13 <gmaxwell> cjdelisle: thats what distinguishes xhtml from html.
65 2011-11-26 01:06:30 <gmaxwell> cjdelisle: HTML is fails best effort, xhtml is always validated.
66 2011-11-26 01:06:41 <cjdelisle> ahh, very cool
67 2011-11-26 01:07:07 <cjdelisle> the whole "browsers get more clever so web devs get more stupid" arms race kinda sucked
68 2011-11-26 01:07:19 <luke-jr> cjdelisle: there's no XHTML on the web
69 2011-11-26 01:07:37 <luke-jr> because IE won't do anything with it
70 2011-11-26 01:07:39 <luke-jr> even if it's valid
71 2011-11-26 01:07:57 <cjdelisle> oh right, indeed
72 2011-11-26 01:08:27 <cjdelisle> actually IIRC XWiki sends XHTML *unless* the useragent says IE
73 2011-11-26 01:09:19 <gmaxwell> The browser vendors don't like xhtml because they don't believe people will ever use it.
74 2011-11-26 01:09:33 <gmaxwell> And .. it introduces a whole new class of failure modes: permitting things you should have rejected.
75 2011-11-26 01:09:54 <gmaxwell> s/ever/never/
76 2011-11-26 01:10:04 <cjdelisle> yea
77 2011-11-26 01:10:25 <gmaxwell> thats where the whole WHATWG and HTML5 thing came from the W3C was bent on XHTML and the browser vendors revolted and forked off.
78 2011-11-26 01:10:53 <gmaxwell> it's been healing.. and everyone except IE has workable XHTML support though it often lags HTML in features.
79 2011-11-26 01:11:06 <cjdelisle> it's too bad since browser makers have a responsibilty to teach manors to the unwashed webdev masses..
80 2011-11-26 01:12:22 <gmaxwell> Thats true, but its much harder than it seems. E.g. with xhtml if you create a comment field.. for example, and you validate it enough to keep people from inserting evil stuff, you're fine for html... but w/ xhtml if someone fails to close a tag and your validator isn't a full XML parser you'll possibly miss it and the page won't render.
81 2011-11-26 01:12:34 <gmaxwell> Strictness has some tough consequences.
82 2011-11-26 01:13:22 <cjdelisle> /nod
83 2011-11-26 01:14:13 <cjdelisle> reminds me that something I wanted to do was try to build the xwiki rendering engine with gcj so it took stdin and printed to stdout
84 2011-11-26 01:20:29 <FellowTraveler> roconnor, sorry I got pulled away there for a sec.
85 2011-11-26 01:21:00 <FellowTraveler> coincidentally I am actually the author of OT so if you are interested in doing some related side project, I'll be happy to answer any questions or give you support.
86 2011-11-26 01:25:34 <roconnor> FellowTraveler: I know you are the author of OT :D I've been reading about it, but haven't got around to installing it yet; I probably will soon. I do have a question though if you have a sec.
87 2011-11-26 01:26:18 <FellowTraveler> roconnor no prob, feel free to ask anytime or initiate a side channel chat.
88 2011-11-26 01:26:33 <FellowTraveler> also email: fellowtraveler@rayservers.net
89 2011-11-26 01:28:27 <gmaxwell> eldentyrell is reporting in #bitcoin that the osx build of .5 doesn't include bitcoind
90 2011-11-26 01:28:35 <roconnor> FellowTraveler: I think I understand how if you have a server Ivan who holds deposits for Alice and Bob how Alice and Bob exchange funds. But I don't quite see how this can extend to when Alice "banks" with server Ivan and Bob "banks" with sever Jay. and they want to exchange funds (denominated in the same currency). I hope my question makes sense.
91 2011-11-26 01:29:05 <roconnor> FellowTraveler: or maybe you have this written up somewhere that I missed.
92 2011-11-26 01:31:09 <FellowTraveler> roconnor: the answer is, that a server MUST redeem any instrument that it issued. Some other server cannot redeem those instruments.
93 2011-11-26 01:31:14 <FellowTraveler> However, keep in mind that the issuer can issue funds onto multiple servers, and they can be transferred between the servers in various ways pretty easily (using BITCOIN, for example, as a "glue" between them&)
94 2011-11-26 01:31:16 <FellowTraveler> also the servers cannot forge transactions against the users, and if the server disappears, the issuer can just re-issue those funds onto another server.
95 2011-11-26 01:31:50 <FellowTraveler> Finally, of course, the client software automates the process of querying whichever server it needs to, when doing any specific transaction. You might have a list of a dozen servers configured inside your wallet.
96 2011-11-26 01:31:56 <FellowTraveler> (Meaning you have a dozen server contracts imported.)
97 2011-11-26 01:31:57 <FellowTraveler> .
98 2011-11-26 01:33:31 <roconnor> So the banks "Ivan" and "Jay" have to transfer funds via the server who issues the currency?
99 2011-11-26 01:33:53 <FellowTraveler> In your example, Bob's client would receive a cheque from Alice drawn from her account at server Ivan. A well-designed client would just connect to server Ivan and redeem it there, and move it to server Jay if that's what he has configured as his default.
100 2011-11-26 01:34:04 <roconnor> oh
101 2011-11-26 01:34:08 <roconnor> that makes sense
102 2011-11-26 01:34:28 <roconnor> thanks, that helps a lot
103 2011-11-26 01:34:32 <FellowTraveler> As for moving funds between servers, there are many ways: (1) any server-to-server transfer can go through the issuer, who is on all servers already. (2) Use OT market to convert to Bitcoin, then bailout, move to any other server, bail back in, convert back on market again.
104 2011-11-26 01:34:48 <FellowTraveler> I see a future where people can store their money in gold, then use Bitcoins to move it onto a debit card, and spend as dollars.
105 2011-11-26 01:35:06 <roconnor> intresting vision
106 2011-11-26 01:35:18 <FellowTraveler> So even people who distrust bitcoin, and prefer gold, and who prefer to spend dollars, will STILL find that Bitcoin is a useful tool that solves unique problems.
107 2011-11-26 01:35:28 <FellowTraveler> Because it makes that possible.
108 2011-11-26 01:35:45 <FellowTraveler> another server-to-server transfer method is to use Ripple (have a p2p protocol between users.)
109 2011-11-26 01:36:04 <FellowTraveler> because ANYTIME someone will give you "acceptable value" on some other system, then your client software would presumably be happy to transfer on-server units to them in return.
110 2011-11-26 01:36:17 <FellowTraveler> So you can actually get your money on other systems without ever having to bail in or out of the server itself
111 2011-11-26 01:36:31 <FellowTraveler> You just get the funds transferred from an existing user based on some other transaction in another system.
112 2011-11-26 01:36:57 <FellowTraveler> At some point I think clients will even employ screen-scraping in order to integrate with their current-day slave banks.
113 2011-11-26 01:37:17 <roconnor> thanks
114 2011-11-26 01:37:23 <FellowTraveler> no prob.
115 2011-11-26 01:41:16 <FellowTraveler> One more thing: my goal with Bitcoin is for voting pools on the blockchain to serve as the "issuer", on OT servers.
116 2011-11-26 01:41:25 <FellowTraveler> That way you don't have to trust some asshole with your coins.
117 2011-11-26 01:41:42 <roconnor> FellowTraveler: do you have anything written up about this?
118 2011-11-26 01:41:44 <FellowTraveler> (in order to still have access to the servers, and therefore to the untraceable cash, the instant finality of settlement, etc.)
119 2011-11-26 01:42:00 <FellowTraveler> roconnor just read my wiki, lots of articles and other media.
120 2011-11-26 01:42:04 <roconnor> ok
121 2011-11-26 01:42:10 <FellowTraveler> feel free to join mailing list if any other questions.
122 2011-11-26 01:42:16 <graingert> OT servers?
123 2011-11-26 01:43:06 <FellowTraveler> graingert Normally if you want to have BTC units on a server, like at MtGox, or at MyBitcoin, then you have to trust that server with your coins, until you bail them back off the server again.
124 2011-11-26 01:43:16 <FellowTraveler> My goal with OT is to eliminate that need, so you can have Low-Trust servers.
125 2011-11-26 01:43:41 <FellowTraveler> One way of doing that is to store the Bitcoins in a voting pool while they are bailed in.
126 2011-11-26 01:43:49 <FellowTraveler> So the server cannot steal them, or hackers cannot break into it and steal them.
127 2011-11-26 01:43:56 <FellowTraveler> Even if the server disappears, they are still recoverable.
128 2011-11-26 01:44:15 <gmaxwell> sounds fine, bitcoin facilitates that.
129 2011-11-26 01:44:24 <FellowTraveler> In that case, the "Bitcoin voting pool", which is already made possible by your existing bitcoin scripting language, serves the role of the "issuer"
130 2011-11-26 01:44:43 <FellowTraveler> whereas with Gold, or some other physical asset, the issuer would normally be one of the USERS (and presumably has a gold storage unit somewhere)
131 2011-11-26 01:44:49 <graingert> ah
132 2011-11-26 01:44:57 <graingert> so you're going to be using op_eval for this?
133 2011-11-26 01:44:59 <FellowTraveler> But with Bitcoin, since voting pools are possible, I want to go the extra step and reduce the need to trust that server.
134 2011-11-26 01:45:02 <theymos> Are you talking about having outputs require multiple signatures?
135 2011-11-26 01:45:18 <FellowTraveler> No, one of you will code that for me and I will end up using it in my own protocol, because I have to do things on my side as well. But yes op_eval.
136 2011-11-26 01:45:23 <FellowTraveler> theymos yes.
137 2011-11-26 01:46:36 <theymos> I was thinking before that it would be cool to be able to deposit your coins into several "banks" and specify exactly how many signatures are required out of the total.
138 2011-11-26 01:47:33 <theymos> Kind of reminds me of how Mixmaster allows you to specify the mixes you use to relay or email. You can do the same with BTC banks.
139 2011-11-26 01:47:56 <FellowTraveler> Let's say there are 50 servers in the pool. So once I bail my Bitcoins onto a specific server, they are now safe in the pool. The server gives me BTC units inside.
140 2011-11-26 01:48:01 <FellowTraveler> Now I have access to all the things a server can do (untraceable cash, markets, recurring payments, scriptable contracts, instant finality of settlement&)
141 2011-11-26 01:48:02 <FellowTraveler> and whenever I want to bail back out again, I just send my counter-signed request to the pool, and then the 50 servers vote to transfer my BTC back to me at a normal BTC address.
142 2011-11-26 01:48:22 <FellowTraveler> If my server disappears, I can still submit a recovery request to the other servers, and get a 3/4ths vote and get my coins back.
143 2011-11-26 01:48:48 <FellowTraveler> After all, I still have my account (my last signed receipt) which my server was incapable of forging, and the other OT servers (the pool members) have access to the same audit data that any issuer would.
144 2011-11-26 01:48:51 <FellowTraveler> That's the basic plan.
145 2011-11-26 01:48:52 <FellowTraveler> .
146 2011-11-26 01:48:59 <graingert> IC
147 2011-11-26 01:49:20 <FellowTraveler> also should enable microtransactions so you don't have to do them on the blockchain itself.
148 2011-11-26 01:51:08 <gmaxwell> yup.
149 2011-11-26 01:51:08 <theymos> That scheme is pretty good, but I'd like it better if you could choose exactly which of those 50 servers you trust and which you don't. I suppose that would make the assets less fungible, though.
150 2011-11-26 01:51:27 <FellowTraveler> theymos the idea is, users should be able to configure their clients with the list of pools they trust.
151 2011-11-26 01:51:39 <FellowTraveler> A server should be able to join multiple pools, and the pools can discriminate about which servers join.
152 2011-11-26 01:51:46 <gmaxwell> 50 is a bit high in any case.
153 2011-11-26 01:52:06 <FellowTraveler> gmaxwell it was only an example. The actual number of members in any pool will vary in the real world.
154 2011-11-26 01:52:11 <FellowTraveler> and Buyer beware.
155 2011-11-26 01:52:18 <FellowTraveler> because if all 50 servers are actually the same owner....
156 2011-11-26 01:52:21 <FellowTraveler> .
157 2011-11-26 01:52:38 <gmaxwell> Indeed, there are real costs to having large numbers. .. e.g. more risk that they aren't independant, and larger transactions to move funds in and out of bitcoin.
158 2011-11-26 01:52:52 <gmaxwell> But yea, people will be able to figure out the right sizes in practice.
159 2011-11-26 01:53:16 <FellowTraveler> I just like to say "50" because people think to themselves, "NO ONE could ever hack FIFTY servers."
160 2011-11-26 01:53:37 <FellowTraveler> of course you can never be paranoid enough. We should have all crypto-cards in our cell phones right now if we were doing it right.
161 2011-11-26 01:54:05 <FellowTraveler> What is the average number of members in your mining pools ?
162 2011-11-26 01:55:18 <theymos> One thing you'd want to make sure of is that every server in the pool is signing things so that some servers don't get destroyed without anyone noticing.
163 2011-11-26 01:56:10 <FellowTraveler> Yeah I have high-level notes on the protocol but all those details still need to be worked out.
164 2011-11-26 01:56:27 <FellowTraveler> In fact I would love if you guys would work it out. You want an open standard anyway, right? For all your own servers.
165 2011-11-26 01:56:45 <FellowTraveler> You can rely on OT being able to do whatever hashes or signed receipts or whatever that you need.
166 2011-11-26 01:57:03 <FellowTraveler> This will require auditing FYI.
167 2011-11-26 01:57:27 <FellowTraveler> to prevent inflation. Servers can't forge your transactions but they can still inflate currencies using dummy accounts, unless auditing is in place.
168 2011-11-26 01:57:33 <FellowTraveler> Necessary piece for low-trust servers.
169 2011-11-26 01:57:52 <FellowTraveler> And that means I need a common distributed hash table, so the place where users get their receipts, is the SAME PLACE where issuers do their audits
170 2011-11-26 01:58:08 <FellowTraveler> (in the case of bitcoin, that means voting pool members will play the role of "issuer" and vote whenever they have to decide something)
171 2011-11-26 01:58:09 <FellowTraveler> .
172 2011-11-26 01:58:33 <FellowTraveler> It doesn't have to store the receipts forever, but it DOES need to be censorship-resistant, and store them long enough to be useful.
173 2011-11-26 01:58:52 <FellowTraveler> And of course since we have digital cash, we can use that as part of the protocol in order to pay for resources like receipt storage
174 2011-11-26 01:59:46 <FellowTraveler> I think the next generation client will include some kind of anonymous network client with DHT, just to finish out my needs with auditing and low-trust.
175 2011-11-26 02:00:07 <FellowTraveler> And therefore it seems smart to design the anon-net-node to allocate resources based on digital cash.
176 2011-11-26 02:00:18 <FellowTraveler> Anyway just thinking out loud.
177 2011-11-26 02:16:23 <gmaxwell> er. are you aware of an attack resistant dht? basically everything out there is seriously vulnerable. Except perhaps freenet and only CHK and SSK are not vulnerable there (e.g. it doesn't provide attack resistant human friendly nameing)
178 2011-11-26 02:24:03 <coventry> What is the purpose of OP_CODESEPARATOR
179 2011-11-26 02:29:08 <coventry> This is really confusing. There is no instance of OP_CODESEPARATOR in etotheipi's list of non-standard transactions. Is it intended for some future purpose?
180 2011-11-26 02:38:32 <theymos> I think it's necessary in cases where multiple checksig/checkmultisig commands are executed in the same script.
181 2011-11-26 02:46:16 <coventry> Thansk, theymos. Do you know of any examples? I am having trouble wrapping my head around it.
182 2011-11-26 02:50:16 <theymos> If you have a script like "<sigA> <pubkeyA> OP_CHECKSIGVERIFY <sigB> <pubkeyB> OP_CHECKSIG" then each signature will have to sign the other signature, which is impossible. So you put a code separator after the first part, and then the second signature doesn't have to sign the first (but the first does have to sign the second). I don't think it's ever been used in a real transaction. It's also possible I don't understand it correctly.
183 2011-11-26 02:58:54 <FellowTraveler> gmaxwell I don't have much else in mind besides maybe Namecoin, maybe Freenet,, maybe Tahoe-LAFS over i2p.
184 2011-11-26 02:59:01 <FellowTraveler> Does anyone else have any suggestions for solving that problem?
185 2011-11-26 02:59:17 <FellowTraveler> Or the best way to integrate those pieces?
186 2011-11-26 03:00:30 <cjdelisle> you can make dhts attack resistant by forcing the node to add some proof of work to their id -- a "how did you generate that" packet
187 2011-11-26 03:01:06 <theymos> You're trying to prevent servers from issuing more currency than is actually backed by BTC?
188 2011-11-26 03:01:15 <cjdelisle> but if it takes 1 second to generate an id and you have 1000 nodes, with a replication factor of 8, you can still pwn that dht with around 8000 seconds of processing
189 2011-11-26 03:01:15 <FellowTraveler> I don't want to have to make a dht attack resistant. Instead I just want to click a button and have the ideal DHT running on my machine so I can focus on coding OT.
190 2011-11-26 03:01:36 <cjdelisle> hahaha
191 2011-11-26 03:01:39 <cjdelisle> :)
192 2011-11-26 03:01:54 <FellowTraveler> theymos I would like a common place for receipts to be exchanged, in order to facilitate audits (and therefore low-trust servers)
193 2011-11-26 03:02:16 <FellowTraveler> I would like for the users and auditors to be able to retrieve the receipts from a common, reliable, censorship-resistant place.
194 2011-11-26 03:02:32 <FellowTraveler> and since this is for digital cash, obviously that can be figured in for solving the problem, like if resources need to be paid for.
195 2011-11-26 03:02:46 <cjdelisle> what is the risk? modification or just "feigning forgetfullness"?
196 2011-11-26 03:03:12 <FellowTraveler> not modification, since everything is signed.
197 2011-11-26 03:03:27 <FellowTraveler> On freenet or anywhere else, there's a specific time to live
198 2011-11-26 03:03:34 <theymos> What are receipts in this context and why are they important?
199 2011-11-26 03:03:40 <FellowTraveler> receipts don't have to be stored forever, just a few weeks probably, or even less if real-time auditing.
200 2011-11-26 03:03:59 <FellowTraveler> cheques being deposited, cash being withdrawn, market offers being initiated...
201 2011-11-26 03:04:03 <FellowTraveler> payment plans processing.
202 2011-11-26 03:04:10 <FellowTraveler> Bitcoins being bailed in.
203 2011-11-26 03:04:12 <FellowTraveler> and out.
204 2011-11-26 03:04:14 <FellowTraveler> etc.
205 2011-11-26 03:04:17 <FellowTraveler> transactions of all sorts.
206 2011-11-26 03:04:55 <cjdelisle> ok so the attack model is that you get a bunch of nodes which are close to eachother in key space and then you have them all forget about a certain receipt in concert -- then the dht is unable to load that receipt because it's nolonger there.
207 2011-11-26 03:05:14 <FellowTraveler> The idea is, if the USER has to get his receipt from the same place as the AUDITOR, therefore the auditor has access to the same receipts.
208 2011-11-26 03:05:40 <FellowTraveler> Any illicit funds (inflated from a dummy acct, for example) would not be able to be spent without flowing into a normal account and thus showing up on audits.
209 2011-11-26 03:06:08 <theymos> Hash the receipts and put the hashes into the Bitcoin block chain for timestamping.
210 2011-11-26 03:06:11 <FellowTraveler> in real-time auditing you'd catch it within the first receipt. Otherwise if you do 24-hour auditing, then you have to put a 36 hour delay (or whatever) on bailouts.
211 2011-11-26 03:06:25 <FellowTraveler> Yeah but I don't want to have to store them forever. A few weeks is fine.
212 2011-11-26 03:06:50 <theymos> Put the hashes in transactions and spend the transactions after a few weeks. The network will forget about it.
213 2011-11-26 03:07:20 <FellowTraveler> network forgets spends eventually?
214 2011-11-26 03:07:26 <FellowTraveler> reduces the past to a hash at some point?
215 2011-11-26 03:07:59 <coventry> Thanks, theymos. I will take a look at multisig.
216 2011-11-26 03:08:04 <theymos> It doesn't do that now, but it will. Spent outputs will eventually be forgotten.
217 2011-11-26 03:08:11 <FellowTraveler> cool thanks for that.
218 2011-11-26 03:08:42 <theymos> Outputs can also have a value of 0, so you don't need to have any money tied up.
219 2011-11-26 03:08:55 <theymos> Bitcoin was pretty much designed for stuff like this.
220 2011-11-26 03:09:34 <cjdelisle> heh what *I* would do is scrap the servers and rely on bitcoin to handle transfer of property by making an issuer hand out like 1 satoshi which the issuer treats as a certificate and whoever has that satoshi is eligable to buy back the ton of grain.
221 2011-11-26 03:09:52 <cjdelisle> But I realize that reads like "you're doing it wrong" and I won't get in your way ;)
222 2011-11-26 03:10:12 <FellowTraveler> every technology is useful for its own piece.
223 2011-11-26 03:10:28 <FellowTraveler> I like to find angles where combining pieces results in a "whole greater than its parts"
224 2011-11-26 03:10:36 <cjdelisle> indeed
225 2011-11-26 03:10:40 <FellowTraveler> like using Bitcoin for transferring gold, as discussed above.
226 2011-11-26 03:10:55 <cjdelisle> ahh, I didn't see it
227 2011-11-26 03:11:19 <theymos> It feels really wrong to me to tie assets to BTC, since bitcoins are not really separate units and you can "irreversibly" combine them with others. Blind signing also has advantages.
228 2011-11-26 03:11:58 <cjdelisle> So many parts of the problem are so well solved -- by different software which doesn't interface well...
229 2011-11-26 03:12:58 <FellowTraveler> theymos Bitcoins are covertible now on markets, into yen, dollars, euros, etc.
230 2011-11-26 03:13:24 <FellowTraveler> Therefore if a gold issuer had issued gold onto an OT server, I could store my money there in gold, and convert it to Bitcoins on the OT market when I want to move it onto a debit card and spend as dollars.
231 2011-11-26 03:13:47 <FellowTraveler> Or I could take those bitcoins to another server and bail-in there, and convert back to gold again for storage.
232 2011-11-26 03:13:49 <cjdelisle> :D
233 2011-11-26 03:13:59 <FellowTraveler> I'm not saying you should do that, I'm just saying that "people will do that".
234 2011-11-26 03:14:12 <FellowTraveler> Therefore Bitcoin is useful as a gold and dollar transfer system.
235 2011-11-26 03:15:00 <theymos> That sounds like trading, which is fine. It bothers me only when individual coins (outputs) are used to represent things, since you can't always track them accurately.
236 2011-11-26 03:15:27 <FellowTraveler> In fact, if you think of the total overall value that is stored on the earth, most of it isn't even in monetary form. It's in land, or stocks, or equipment, etc.
237 2011-11-26 03:15:42 <FellowTraveler> Only some percentage is stored in money, and only some percentage of that will ever be stored in Bitcoin.
238 2011-11-26 03:16:10 <FellowTraveler> But I bet you, even if Bitcoin only ends up as, say, 1% of the overall money supply, you can see that 1% as representing the ACTUAL VALUE that Bitcoin brings to the table
239 2011-11-26 03:16:18 <FellowTraveler> in terms of its being able to function as a lubricant for all the others.
240 2011-11-26 03:16:21 <FellowTraveler> Due to its unique properties.
241 2011-11-26 03:17:13 <FellowTraveler> And then if there are crackdowns, all that will do is cause MORE allocation in bitcoin form instead of dollar or gold form.
242 2011-11-26 03:17:18 <FellowTraveler> Because it's more censorship resistant.
243 2011-11-26 03:17:28 <luke-jr> no
244 2011-11-26 03:17:41 <luke-jr> you assume everyone is antiauthoritarian
245 2011-11-26 03:17:49 <luke-jr> and that censorship is abd
246 2011-11-26 03:17:50 <luke-jr> bad*
247 2011-11-26 03:17:53 <luke-jr> both of which are false.
248 2011-11-26 03:18:24 <FellowTraveler> I simply mean that money routes around obstacles, just like networks.
249 2011-11-26 03:18:33 <luke-jr> if Bitcoin is made illegal, probably most developers will cease.
250 2011-11-26 03:19:02 <FellowTraveler> If filesharing is made illegal, probably Bittorrent will cease also.
251 2011-11-26 03:19:29 <luke-jr> probably.
252 2011-11-26 03:19:36 <FellowTraveler> :P
253 2011-11-26 03:19:42 <luke-jr> let's hope it's never made illegal
254 2011-11-26 03:19:57 <FellowTraveler> I don't know look at SOPA.
255 2011-11-26 03:20:01 <luke-jr> hint: filesharing is 100% legal pretty much everywhere
256 2011-11-26 03:20:12 <FellowTraveler> Perhaps authorities cannot move to overtly ban Bitcoin because it will cause a Streisand effect.
257 2011-11-26 03:20:17 <luke-jr> SOPA has nothing to do with filesharing
258 2011-11-26 03:20:21 <theymos> You can see a similar occurance with Limewire. The main developers all had to quit, but a fork was quickly created.
259 2011-11-26 03:20:33 <FellowTraveler> I just mean that SOPA resulted in 10K+ darknet members on reddit
260 2011-11-26 03:20:42 <FellowTraveler> Maybe banning Bitcoin wouldn't kill it, but cause it to explode.
261 2011-11-26 03:21:01 <luke-jr> only among criminals
262 2011-11-26 03:21:17 <cjdelisle> luke-jr thinks the OWS people are a criminal gang
263 2011-11-26 03:21:24 <FellowTraveler> criminals == people who don't use slave money.
264 2011-11-26 03:21:25 <cjdelisle> so take that with a grain of salt
265 2011-11-26 03:21:39 <luke-jr> wtf is OWS?
266 2011-11-26 03:21:52 <cjdelisle> occupy wall street
267 2011-11-26 03:22:03 <luke-jr> sounds like it from what I've heard on IRC and such
268 2011-11-26 03:22:19 <FellowTraveler> OWS people are the unwashed masses. They are right to oppose corporate welfare and banks, but they are also a bunch of anti-voluntarists.
269 2011-11-26 03:22:49 <cjdelisle> my only point is some people use the word 'criminal' more loosely than others
270 2011-11-26 03:22:54 <FellowTraveler> Just like in Europe, obviously the ruling classes deserve to have riots thrown at them, but the rioters are still basically ignorant and criminal.
271 2011-11-26 03:23:15 <FellowTraveler> Seems like everything is always a false dichotomy in politics. That's the best way to manipulate people.
272 2011-11-26 03:23:48 <FellowTraveler> Everyone is "party right".
273 2011-11-26 03:23:50 <luke-jr> the "ruling class" is to be disciplined by THEIR authorities
274 2011-11-26 03:23:52 <luke-jr> not their subjects
275 2011-11-26 03:24:12 <FellowTraveler> I think in Europe they are "cogs".
276 2011-11-26 03:24:26 <[Tycho]> luke-jr: are you going to implement "green address" payments ?
277 2011-11-26 03:24:29 <FellowTraveler> They line up in queues. They just follow orders.
278 2011-11-26 03:24:34 <luke-jr> [Tycho]: no, it's stupid
279 2011-11-26 03:24:59 <luke-jr> [Tycho]: more importantly, if I did implement it, people "trusting" me would be fools
280 2011-11-26 03:25:10 <luke-jr> since an invalid block could very well reverse them
281 2011-11-26 03:25:45 <[Tycho]> So people trusting mtgox (if any) are fools too ?
282 2011-11-26 03:25:57 <luke-jr> no, MtGox doesn't let you withdraw until you have confirmed funds
283 2011-11-26 03:25:58 <FellowTraveler> Yes.
284 2011-11-26 03:26:05 <luke-jr> Eligius pays out long before then
285 2011-11-26 03:26:40 <luke-jr> because if it turns out the miner didn't actually earn those funds, the block paying them is invalid anyway, and their reward is reversed
286 2011-11-26 03:27:28 <[Tycho]> Oh, I forgot about your special method.
287 2011-11-26 03:27:34 <luke-jr> in other words, there's a relatively high chance an Eligius payout *will* be reversed, and if you take it with 0 confirms, you *are* taking a big risk
288 2011-11-26 03:27:56 <luke-jr> well, maybe not big since payouts are tiny, but still
289 2011-11-26 03:28:12 <cjdelisle> hey what are your thoughts on using & rather than && if you're checking that a bunch of numbers are 0? Shouldn't it be faster to do a bunch of calculations and 1 branch rather than a bunch of branches?
290 2011-11-26 03:28:17 <[Tycho]> I was thinking about normal TXes.
291 2011-11-26 03:28:52 <cjdelisle> err I guess that would be |
292 2011-11-26 03:28:57 <FellowTraveler> off topic but FYI since I just got this running. Here's a couple client-script OT test scripts (we now have scripting): http://pastebin.com/Curs5WTL
293 2011-11-26 03:29:07 <FellowTraveler> testing/debugging server side scripting now (smart contracts).
294 2011-11-26 03:30:49 <cjdelisle> if (a == 0 && b == 0 && c == 0 ...) {
295 2011-11-26 03:31:11 <cjdelisle> vs. if ((a | b | c | d) == 0)
296 2011-11-26 03:31:54 <cjdelisle> seems like it should run faster because it promises that the right side of the && won't run unless the left side is true
297 2011-11-26 03:32:10 <luke-jr> so does &&
298 2011-11-26 03:32:21 <luke-jr> actually
299 2011-11-26 03:32:23 <luke-jr> only && does
300 2011-11-26 03:32:26 <cjdelisle> that's what I mean && has to branch to make that promise
301 2011-11-26 03:32:35 <luke-jr> so?
302 2011-11-26 03:32:44 <cjdelisle> I understand branches are slow
303 2011-11-26 03:32:47 <luke-jr> &
304 2011-11-26 03:32:52 <luke-jr> not slower than verifying a signature
305 2011-11-26 03:33:08 <cjdelisle> no, I'm just checking for zeros atm
306 2011-11-26 03:33:34 <cjdelisle> meh, rare occurance.. readability > *
307 2011-11-26 05:14:28 <skittixch> hey all, I'm a beginner trying to make a very simple ticker website with javascript, and I'm having a little trouble understanding what's going on in the wiki. http://codepad.org/TaMCfOiQ any ideas?
308 2011-11-26 05:59:44 <vrs> skittixch: wrong channel
309 2011-11-26 06:01:02 <vrs> oh, mtgox stuff. anyway, this is probably a JS problem and you should ask there (and provide actually useful info like error messages and the like)
310 2011-11-26 06:10:24 <skittixch> cool, thanks!
311 2011-11-26 06:32:57 <diki> Wow
312 2011-11-26 06:33:07 <diki> considering that the doom 3 sourcecode is 9 megabytes
313 2011-11-26 06:33:16 <diki> how on earth is the game 1.75 gigs?
314 2011-11-26 06:33:24 <Diablo-D3> are you retarded?
315 2011-11-26 06:33:38 <Diablo-D3> did you forget the 1.75 gigs of game media?
316 2011-11-26 06:33:50 <bd_> diki: source code doesn't include game assets (textures, sounds, models, maps...)
317 2011-11-26 06:34:01 <diki> What??
318 2011-11-26 06:34:03 <bd_> that 9mb is just the game engine itself
319 2011-11-26 06:34:04 <diki> Why?
320 2011-11-26 06:34:14 <bd_> diki: because they didn't want to release the game assets?
321 2011-11-26 06:34:28 <bd_> because open-sourcing the game assets would make it basically impossible to sell doom 3?
322 2011-11-26 06:34:37 <Diablo-D3> diki: because you touch yourself at night
323 2011-11-26 06:34:39 <batouzo> diki: do you have any idea how computers are working?
324 2011-11-26 06:34:59 <diki> batouzo:do not speak
325 2011-11-26 06:35:05 <diki> Diablo-D3:same goes for you
326 2011-11-26 06:35:24 <Diablo-D3> an ant does not ask a god not to speak, it is considered rude.
327 2011-11-26 06:35:37 <diki> You're forgetting who's the god ;)
328 2011-11-26 06:36:08 <luke-jr> Diablo-D3 thinks he is Gilgamesh
329 2011-11-26 06:37:00 <Diablo-D3> I said god, not demigod.
330 2011-11-26 06:37:15 <diki> :trollface:
331 2011-11-26 07:40:12 <wumpus> diki: in games it's the assets that are big, not the code, usually.. (and it has some pleasant advantages for the "hot code" to fit in the cache). that said, I was pleasantly surprised by the readabilty and cleanness of the doom3 code
332 2011-11-26 11:45:09 <denisx> why is the target in getwork not also deprecated? it is part of the data already
333 2011-11-26 11:45:51 <sipa> it is not
334 2011-11-26 11:46:29 <sipa> well, the real target is, but maybe not the target requested by the pool
335 2011-11-26 11:46:55 <denisx> sipa: I'm talking about bitcoind and not a pool
336 2011-11-26 11:47:01 <denisx> the pool rewrites anyway
337 2011-11-26 11:47:20 <sipa> right, but target field is part of the getwork protocol, if you want it to work with pools
338 2011-11-26 11:47:57 <denisx> sipa: ah ok, that makes sense
339 2011-11-26 11:48:36 <sipa> you could make it implicit, saying that the field is optional, and if it's not present, use the nbits field in the data
340 2011-11-26 11:48:43 <sipa> but that makes it harder for implementors
341 2011-11-26 12:58:17 <luke-jr> ;;bc,gen 50000000
342 2011-11-26 12:58:18 <gribble> The expected generation output, at 50000000 Khps, given current difficulty of 1192497.7500895 , is 42.1731772873 BTC per day and 1.7572157203 BTC per hour.
343 2011-11-26 15:01:35 <denisx> anyone knows how to hash the transactions from getmemorypool?
344 2011-11-26 15:05:39 <makomk> Hash the for what - to build a block?
345 2011-11-26 15:05:51 <denisx> makomk: no, for the merkletree
346 2011-11-26 15:06:15 <denisx> so, yeah, for the block
347 2011-11-26 15:06:58 <sipa> Hdouble sha256
348 2011-11-26 15:07:04 <sipa> -H
349 2011-11-26 15:07:18 <diki> and what is -H?
350 2011-11-26 15:08:20 <denisx> diki: a typo
351 2011-11-26 15:08:46 <diki> no i really meant what hdouble means
352 2011-11-26 15:08:48 <diki> or -h
353 2011-11-26 15:09:06 <denisx> that he only meant double
354 2011-11-26 15:09:12 <denisx> minus h
355 2011-11-26 15:10:56 <luke-jr> denisx: I'm obviously doing it :P
356 2011-11-26 15:11:11 <denisx> luke-jr: yeah, but its python
357 2011-11-26 15:11:20 <luke-jr> shrug
358 2011-11-26 15:18:04 <justmoon> da2ce7 speaking at prague bitcoin conference: http://www.ustream.tv/channel/justmoon
359 2011-11-26 15:18:15 <justmoon> stream is low quality and contingent on the flaky wifi
360 2011-11-26 15:18:41 <luke-jr> justmoon: do they have a local bitcoin node to seed the blockchain? :p
361 2011-11-26 15:19:18 <justmoon> luke-jr: no, that may very well be the problem -_-
362 2011-11-26 15:20:07 <luke-jr> fail
363 2011-11-26 15:20:36 <justmoon> luke-jr: to be fair, we are here for the talks, not for the wifi :)
364 2011-11-26 15:21:40 <luke-jr> justmoon: but you need the block chain to trade
365 2011-11-26 15:21:50 <luke-jr> you know, lunch and stuff
366 2011-11-26 15:22:45 <justmoon> amir added an extra talk and let the money panel run overtime, so there wasn't much in the way of breaks lol
367 2011-11-26 15:23:27 <davout> amir probably codes a lot better than he manages time ;)
368 2011-11-26 15:24:12 <justmoon> I say he did a fine job!
369 2011-11-26 15:24:23 <justmoon> ;)
370 2011-11-26 15:24:50 <[Tycho]> Hello, justmoon
371 2011-11-26 15:24:58 <davout> we'll see when we're having the next coffee break at 10pm hah! :D
372 2011-11-26 15:25:06 <justmoon> lol
373 2011-11-26 15:25:10 <justmoon> hey tycho
374 2011-11-26 15:25:51 <diki> ;;seen zenith
375 2011-11-26 15:25:51 <gribble> I have not seen zenith.
376 2011-11-26 15:27:13 <diki> davout:yup
377 2011-11-26 15:27:15 <diki> that is correct
378 2011-11-26 15:28:12 <[Tycho]> I see bitcoin people !
379 2011-11-26 15:33:51 <batouzo> meh the conference cuted Open Transactions talk in half
380 2011-11-26 15:34:06 <batouzo> organized missunderstooded the questions with end of speach apparently?
381 2011-11-26 15:49:27 <justmoon> just spoke with amir - if you guys have questions for any of the speakers, put it in the chat in #bitcoin
382 2011-11-26 15:50:15 <luke-jr> [11:50:12] [479] luke-jr #bitcoin
383 2011-11-26 15:51:05 <justmoon> o_O
384 2011-11-26 15:51:31 <luke-jr> justmoon: is that a pi at the end?
385 2011-11-26 15:51:55 <justmoon> I'm on a borrowed macbook
386 2011-11-26 15:52:07 <justmoon> so don't ask me what kind of awesome unicode it adds to messages
387 2011-11-26 15:52:11 <luke-jr> lol
388 2011-11-26 15:56:14 <batouzo> justmoon: yea I have questions
389 2011-11-26 15:56:25 <justmoon> batouzo: put them in #bitcoin
390 2011-11-26 15:56:32 <justmoon> (just the normal bitcoin channel)
391 2011-11-26 15:56:42 <batouzo> amir's irc nick?
392 2011-11-26 15:57:16 <justmoon> batouzo: genjix
393 2011-11-26 15:57:21 <justmoon> send them to me
394 2011-11-26 15:57:24 <justmoon> I'll read #bitcoin
395 2011-11-26 15:57:27 <justmoon> and give them to amir
396 2011-11-26 15:57:56 <[Tycho]> Who is amir ?
397 2011-11-26 15:58:01 <justmoon> genjix
398 2011-11-26 15:58:12 <justmoon> from bitcoin consultancy / libbitcoin
399 2011-11-26 16:01:49 <batouzo> genjix: question to Cameron(OT) - with OT can we quickly buy gold, then trade it for silver on exchange, and then buy some good in shops for that?
400 2011-11-26 16:02:10 <batouzo> justmoon: btw genjix is not on channels here, so you can forward
401 2011-11-26 16:03:30 <justmoon> batouzo: genjix is the guy on the far left - he asked me to collect questions from IRC
402 2011-11-26 16:03:49 <justmoon> batouzo: if he turns around and sees me waving I'll try and get your question in :)
403 2011-11-26 16:04:08 <batouzo> justmoon: ok. so mine #1 is above.
404 2011-11-26 16:05:27 <batouzo> justmoon: #2 with OT can I easly borrow some bitcoins to some person, and have him and his friends promise he will pay back in 2 weeks - and can I easly prove if it would be not done as agreed
405 2011-11-26 16:06:32 <davout> batouzo: you can do this with just pgp
406 2011-11-26 16:07:07 <batouzo> davout: sure. And with 1000 customers?
407 2011-11-26 16:07:43 <batouzo> you must code some system, and get everyone to use it. and that is what ot is
408 2011-11-26 16:08:15 <davout> batouzo: i'll continue on #bitcoin
409 2011-11-26 17:10:07 <jeremias> hi everyone, created this example application for django-bitcoin: http://kangasbros.fi:9999/
410 2011-11-26 17:10:12 <jeremias> https://github.com/kangasbros/django-assmarket
411 2011-11-26 17:10:19 <jeremias> (death prediction market)
412 2011-11-26 17:10:25 <jeremias> or so called assasination market
413 2011-11-26 17:10:46 <jeremias> I'm at the euro bitcoin conference, so if anyone haves questions, you are welcome
414 2011-11-26 17:11:00 <jeremias> to ask
415 2011-11-26 20:30:39 <Lolcust_Backup> batouzo isn't the dark exchange more or less that - a peer to peer system for exchanging coins for "stuff" with PGP-based pseudonymous reputation system ?