| 00:34:05 | EasyAt|RollyChai: | EasyAt|RollyChai is now known as EasyAt|sofa |
| 00:46:15 | EasyAt|sofa: | EasyAt|sofa is now known as EasyAt|kitchen |
| 01:20:35 | EasyAt|kitchen: | EasyAt|kitchen is now known as EasyAt|sofa |
| 02:56:27 | zacm_: | zacm_ is now known as zacm |
| 05:57:33 | firepacket: | firepacket is now known as Guest2211 |
| 07:03:47 | mike4: | mike4 is now known as c--O-O |
| 07:09:48 | roidster: | roidster is now known as Guest49608 |
| 10:21:06 | ielo: | ielo is now known as YOURWORSTENEMY |
| 11:15:48 | jarpiain: | jarpiain is now known as Guest982 |
| 15:37:28 | austin: | austin is now known as Guest62102 |
| 15:40:12 | EasyAt|sofa: | EasyAt|sofa is now known as EasyAt |
| 15:45:11 | gmaxwell: | petertodd: I really can't handle any argument stress right now, so I've bowed out of that BIP39 thing at the moment; but you might want to link to sipa past strenghtening proposal |
| 15:45:48 | gmaxwell: | (the one where it it iterates until it finds a distinguished point) |
| 15:45:57 | petertodd: | gmaxwell: cool - I'm very happy to have come up with another use of the term "soft-fork" :P |
| 15:46:46 | petertodd: | sipa: https://github.com/bitcoin/bips/pull/17#issuecomment-34451762 <- how to make bip39 better and less prone to user error |
| 15:47:21 | gmaxwell: | LOL. WOW SUCH INSULT "Apple approves Dogecoin iOS app" |
| 15:47:33 | petertodd: | haha |
| 15:47:53 | petertodd: | granted an electrum read-only wallet got approved too |
| 15:48:38 | jgarzik: | heh |
| 15:48:51 | sipa: | gmaxwell, petertodd: https://bitcointalk.org/index.php?topic=102349.0 |
| 15:48:53 | sipa: | that one? |
| 15:49:49 | petertodd: | sipa: this is to solve the problem that for arbitrary wordlists they have to allow any utf8 string, so I'm suggesting use distinguished points to fix the problem |
| 15:50:11 | sipa: | have you read my proposal? |
| 15:50:13 | petertodd: | sipa: much simplier version of what you propose there :) |
| 15:50:20 | petertodd: | sipa: yes, I remember it well from ages ago |
| 15:50:26 | gmaxwell: | sipa: no, on the BIPS pull req. Slush opened a pull req on the BIP39 bit (the mnemonic encoding one) that changed it to approved, reduced the PBKDF2 iteration count from 4096 to 2048, and inserted some text about brain wallets (apparently some _prior_ changes that I'd missed also removed all the constraints so that it's now just a poorly hardened brainwallet system with a recommended key generation procedure which cannot be enforced. |
| 15:50:43 | gmaxwell: | s/BIP39 bit/BIP39 bip/ |
| 15:51:04 | gmaxwell: | (the new text explicitly promotes using it as a brain wallet too. :-/) |
| 16:33:28 | jgarzik: | jgarzik is now known as home_jg |
| 17:47:56 | gmaxwell: | its interesting there appears to be a 1PH/s user on eligius which is concealing their hashrate by splitting between addresses: |
| 17:48:08 | gmaxwell: | http://eligius.st/~wizkid057/newstats/userstats.php/1A73ExsM2doRwTLp82rv5U36QHbBFmHD1X and http://eligius.st/~wizkid057/newstats/userstats.php/1Nbq2XZaRsKknf5fcT2wTXvBS31PaUWSeX |
| 17:48:57 | gmaxwell: | (note that the latter loses almost all of its hashrate at precisely the same moment that the former pulls another several hundred TH/s out of a hat) |
| 17:53:21 | jgarzik: | gmaxwell, concealing, or just moving among buckets? |
| 17:53:49 | jgarzik: | gmaxwell, and is it confirmed that KNC is mining with their customer's equipment, versus just buying some of their own stock? |
| 17:54:18 | tromp: | in scrypt as used by Litecoin, are the two calls to PBKDF2 negligable in runtime compared to the single call to scruptROMix? |
| 18:00:37 | Luke-Jr: | jgarzik: I don't think anyone has suggested they are mining with customer equipment.. |
| 18:01:25 | gmaxwell: | yea, I don't think they're mining with customer equipment, I think they just do like bitfury: mark products >20x over cost, and for every unit sold buy 10 units for yourself. |
| 18:02:19 | gmaxwell: | (actual numbers may differ) |
| 18:32:00 | jgarzik: | gmaxwell, that I can believe |
| 18:32:12 | Guest62102: | Guest62102 has left #bitcoin-wizards |
| 18:36:55 | jgarzik: | http://maidsafe.net just appeared |
| 18:37:52 | jgarzik: | open source platform for distributed applications. I'll be interested to see if it goes anywhere |
| 18:38:53 | jgarzik: | "secure DHT" is one of 11 components |
| 18:42:30 | Emcy: | [X] DHT |
| 18:45:59 | Emcy: | how can you secure a DHT anyway. The whole thing is wide open |
| 18:47:08 | Emcy: | project tox claims to have done "dht hardning" recently too, apparently after criticism that it was essentially an open book of metadata |
| 18:47:08 | gmaxwell: | you can secure it, for some defintion of secure its inherently secure. That definition may not be very useful however. |
| 18:47:34 | Emcy: | hm ok |
| 18:47:59 | jgarzik: | you can make life harder for attackers. nothing is ever perfect. require each message be signed, or proof of work, or proof of funds/burn/sacrifice/... |
| 18:48:46 | gmaxwell: | also if you only depend on it very weakly then you may not need a strong notion of security. |
| 18:48:53 | gmaxwell: | e.g. if only 1 in 100 uses has to be successful. |
| 18:49:43 | Emcy: | i think tox just went with running every dht lookup thru onion routing |
| 18:49:51 | jgarzik: | or if the DHT is not crucial to the success of an operation, but rather just a cache |
| 18:49:51 | Emcy: | thier own onion |
| 18:52:05 | gmaxwell: | yea, thats helpful if your concern is privacy rather than availablity. |
| 18:52:18 | jcrubino: | jcrubino has left #bitcoin-wizards |
| 18:58:42 | TD: | jgarzik: i found it quite hard to grok what maidsafe actually was |
| 18:58:47 | TD: | jgarzik: though they have a big pile of c++ |
| 18:58:51 | TD: | so i guess there's some useful stuff in there |
| 18:59:34 | jgarzik: | TD, sounds like they are trying to be a decentralized AWS. But looking at the pieces they have, I'm not sure how they get there. Especially when I see zero about working out payments, IMO the biggest piece of the entire decentralized-AWS puzzle. |
| 19:00:48 | gmaxwell: | "big pile of c++" — so, open transactions? |
| 19:01:13 | jgarzik: | except without the transactions |
| 19:01:20 | TD: | it's open though! |
| 19:01:26 | TD: | did anything ever happen with opentransactions? |
| 19:01:37 | jgarzik: | gthe creator sent himself to multiple conferences |
| 19:01:39 | TD: | the few times i met chris he seemed pretty reasonable and clued in, but i never hear anything about his project |
| 19:01:53 | jgarzik: | it pops up now and again |
| 19:01:59 | nsh: | like the plague |
| 19:02:13 | nsh: | (jk) |
| 19:05:17 | TD: | man, i hope slashdot doesn't die over this beta crap |
| 19:05:59 | Emcy: | new coke |
| 19:06:12 | jgarzik: | TD, ? |
| 19:06:34 | TD: | the owners of /. want to replace the site with a total redesign that manages to suck much harder than the current design, hardly a paragon of great layout and programming |
| 19:06:40 | jgarzik: | TD, I got a note about how I'm staying with Classic Slashdot "for the moment" |
| 19:06:56 | jgarzik: | * jgarzik hopes they don't take away Classic |
| 19:06:58 | TD: | the new design sucks so badly that a lot of people will likely leave if the don't make huge changes |
| 19:07:01 | TD: | well, that's their plan |
| 19:07:10 | TD: | they want to remove classic and make everyone use the new site |
| 19:07:14 | nsh: | i vaguely remember there was a site that did something similar whose userbase was eventually subsumed into reddit's. i can no longer recall its name :) |
| 19:07:26 | TD: | * TD ponders whether it'd be possible to make a decentralised slashdot-like discussion service |
| 19:07:41 | TD: | a la usenet or something |
| 19:08:02 | nsh: | you could already throw some localstorage html/css/js on top of usenet and convert it into webthreads |
| 19:08:26 | TD: | usenet doesn't have the moderation system, which is what distinguishes slashdot from other forums |
| 19:08:48 | nsh: | * nsh nods |
| 19:11:31 | TD: | the problem is there's lots of random stuff needed to make a /. competitor which isn't easily decentralised. like, banning users who are too abusive, paying the editors, etc |
| 19:13:04 | petertodd: | in an ideal world it'd be easy to have paid mods use broadcast encryption to encrypt their moderation decisions to paying readers |
| 19:13:12 | petertodd: | like hell that'll ever actually be popular though... |
| 19:15:23 | jgarzik: | nsh, digg? |
| 19:15:38 | nsh: | that was it :) |
| 19:16:14 | jgarzik: | It always amused me that reddit beat digg. Seemed like the two innovations of reddit were: (a) down vote, as well as up vote, and (b) make the UI simple and ugly |
| 19:16:32 | jgarzik: | well, ok, reddit has a "build your own community" thing too |
| 19:16:40 | petertodd: | jgarzik: the reddit UI is beautiful in its simplicity |
| 19:16:46 | petertodd: | jgarzik: yes, subreddits are brilliant |
| 19:20:17 | petertodd: | ^ decentralized IRC just won't be the same without join/quit floods |
| 19:26:03 | Emcy: | reddit never pandered to The Man, thats why |
| 19:26:05 | andytoshi: | ..maybe it could have onion circuits that sometimes go in circles? |
| 19:26:19 | Emcy: | they even had a child lover subred until surprisingly recently |
| 19:26:32 | Emcy: | meanwhile digg tried and failed to censor the AACS key |
| 19:26:55 | petertodd: | andytoshi: lol! actually that's a really good question; much easier to envision without explicit bitmessage-style attempts at privacy |
| 19:27:18 | Emcy: | people never let them live it down |
| 19:58:33 | andytoshi: | alex_fun just said on -dev re the bitcoin source that 'it was made simple and elegant, simple for many to understand and implement in any coin in any shape they like'. |
| 19:58:53 | andytoshi: | i've been working on an essay which talks about this "bitcoin is easy/safe to understand and modify" meme, git://wpsoftware.net/bitcoin/alt-essay.git |
| 19:59:30 | andytoshi: | i'll give push access to anybody who msgs me to ask. progress has been very slow and it'd be cool to have some wizard coauthors and a 2014 release date |
| 20:02:25 | jgarzik: | andytoshi: simple to modify, sure |
| 20:02:54 | jgarzik: | andytoshi: simple to understand? Even really smart people continue to miss significant details (though our knowledge base is slowly getting better over time) |
| 20:03:02 | jgarzik: | Look at all the buggy reimplementations |
| 20:03:51 | andytoshi: | jgarzik: yeah, that's what i'm trying to say. gmaxwell summed it up quite well as "the entirety of bitcoin is a cryptosystem", every modification is rolling your own crypto" |
| 20:03:57 | sipa: | jgarzik: we're not exactly doing a good job of documenting it either |
| 20:05:41 | andytoshi: | ..and the point of this essay is to (a) highlight this point and (b) talk about why 'rolling your own crypto' is an awful dangerous idea. because the bitcoin community seems largely unaware of the crypto community or its norms |
| 20:08:38 | jgarzik: | andytoshi: one of my theses RE bitcoin... bitcoin makes financial software /too easy/ to write |
| 20:09:00 | jgarzik: | It is seductively easy to use and develop for... but that teaches you none of the lessons of crypto, or open source, or financial security, or... |
| 20:10:40 | andytoshi: | jgarzik: definitely agreed. but at the same time it still leaves this software /too hard/ to write for people who need to do it properly :). but one thing we can do to mitigate the 'too easy' problem is to have a single document explaining "homebrew crypto is garbage, don't use it and for god's sake don't put money into it" |
| 20:10:59 | jgarzik: | yep |
| 20:11:08 | jgarzik: | brainwallets are another example of "cool, easy, don't do it" |
| 20:11:37 | andytoshi: | and if we can really get that idea into the community that should go a long way toward killing the traction that these kind of ideas get today |
| 20:12:12 | gmaxwell: | Brainwallets have a kind of unique risk where it puts the end user directly in competition with all attackers for who has the most imagination. |
| 20:12:41 | gmaxwell: | and they look like passwords, which have an entirely different risk surface. |
| 20:15:29 | sipa: | i like this phrase: everyone is smart enough to come up with a scheme they are unable to break themselves |
| 20:16:59 | midnightmagic: | i've heard that one many times now |
| 20:17:25 | gmaxwell: | Usually when we build software we bake the security intelligence into the software— and use peer review to boost the security beyond "unable to break themselves" levels. Brainwallets put JoeUser directly in the battle and structurally cannot have peer review. |
| 20:25:29 | jgarzik: | "Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it." --Brian Kernighan |
| 20:25:59 | jgarzik: | It's almost a paraphrase of what was written above ;p |
| 20:26:52 | nsh_: | * nsh_ smiles |
| 20:27:01 | nsh_: | fortunately smartness is superadditive |
| 20:27:13 | nsh_: | at least when it comes to spotting mistakes |
| 20:27:55 | gmaxwell: | sometimes. sometimes the opposite, e.g. all the reviewers fail the same way. |
| 20:28:43 | Emcy: | jgarzik isnt losing millions of dollars ont he regular a pretty good tutor |
| 20:28:54 | jgarzik: | Emcy, looking at bitcoin's history? |
| 20:28:56 | jgarzik: | Emcy, no :) |
| 20:29:06 | Emcy: | you know, people used to use radium for all sort of trivial shit too until the knowledge percolated thru society |
| 20:29:22 | nsh_: | heh.. |
| 20:29:39 | nsh_: | /topic 21st century radium factory |
| 20:29:42 | nsh_: | :) |
| 20:30:08 | Emcy: | http://www.thingamababy.com/baby/2006/05/fun_with_radiat.html |
| 20:30:15 | Emcy: | irradiate your kid, its fun! |
| 20:49:25 | nsh_: | nsh_ is now known as nsh |
| 21:37:04 | spinza: | spinza is now known as spin123456 |
| 21:37:04 | spin123456: | spin123456 is now known as spinza |