--- Log opened Fri Jun 19 00:00:46 2015 | ||
-!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 255 seconds] | 00:21 | |
-!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 01:17 | |
-!- wrldpc1 [~ben@228.22.150.119.ap.yournet.ne.jp] has quit [Quit: wrldpc1] | 01:18 | |
-!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 272 seconds] | 01:22 | |
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap | 01:48 | |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 02:15 | |
-!- ebowden_ [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 02:15 | |
-!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 03:06 | |
-!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 264 seconds] | 03:11 | |
-!- wrldpc1 [~ben@228.22.150.119.ap.yournet.ne.jp] has joined ##hplusroadmap | 03:31 | |
-!- wrldpc1 [~ben@228.22.150.119.ap.yournet.ne.jp] has quit [Quit: wrldpc1] | 03:44 | |
-!- HEx1 [~HEx@hexwab.plus.com] has quit [Ping timeout: 264 seconds] | 03:47 | |
-!- Madplatypus [uid19957@gateway/web/irccloud.com/x-cdyrjmhfcmdgftwf] has quit [Quit: Connection closed for inactivity] | 04:12 | |
kanzure | .title https://www.youtube.com/watch?v=eqavcH3DKKc&list=TLdFdq0gVQhiw | 04:38 |
---|---|---|
yoleaux | Tubing in FreeCAD - YouTube | 04:38 |
fenn | "Nanotechnology-Inspired Grand Challenges for the Next Decade" how about actually fucking implementing a set of reactions to perform diamondoid mechanosynthesis | 04:39 |
kanzure | but that would be haaaaaarrrd | 04:41 |
-!- poppingtonic [brian@unaffiliated/poppingtonic] has joined ##hplusroadmap | 04:50 | |
kanzure | http://dunveganspace.com/assets/bitsat-infosheet.pdf | 04:51 |
-!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 04:55 | |
-!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 265 seconds] | 05:00 | |
fenn | that's more like "parametric extrusions" than tubing; i was expecting something like the BRL-CAD "pipe" object http://www.abclinuxu.cz/images/clanky/jansa/brlcad3-pipe1.png http://brlcad.org/~starseeker/vol3/book/tutorial_series/volume_III.xhtml#id907997 | 05:10 |
kanzure | http://diyhpl.us/wiki/transcripts/bitcoin-adam3us-fungibility-privacy/ | 05:11 |
fenn | the merchant shouldn't have expected to get his bank notes back after being stolen! | 05:14 |
fenn | he should have taken out an insurance policy against theft | 05:14 |
fenn | and/or implemented better security | 05:14 |
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 255 seconds] | 05:17 | |
-!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap | 05:25 | |
kanzure | perhaps insurance wasn't invented yet | 05:30 |
fenn | anyway i don't think the answer to his problem is "make money untraceable" instead it should be "don't have stupid expectations that the bank will take care of your mistakes" | 05:31 |
fenn | also, a point that seems to never be adequately addressed in these untraceable money proposals is that most people don't want to support criminal enterprises and would like to know if they are receiving stolen/extorted/bloody money | 05:33 |
fenn | not only for the reason that they don't want the police to seize their money, but because they don't want to do business with criminals | 05:34 |
kanzure | having the ability to filter money (for political or criminal reasons) means that you are essentially bifurcating the money and making it not fungible | 05:35 |
kanzure | it would be trivial to pass every dollar through that taint function.. oops your whole money supply is now tainted. | 05:35 |
fenn | don't use a boolean taint value | 05:35 |
fenn | some people would be fine doing business with drug dealers but not murderers, or would be ok with doing business with people who do business with murderers, etc. | 05:36 |
kanzure | over time your money will touch the hands of everyone, with enough circulation, so ... things are just going to stop moving.. | 05:37 |
fenn | i don't know how to actually implement this | 05:37 |
fenn | but for example intel doesn't buy "blood tantalum" | 05:37 |
fenn | it seems like it should be possible to make reasonable assumptions and observations without actually being legally compelled to | 05:38 |
fenn | it's probably better for each participant in the economy to have their own "taint function" so as to not enable someone from gaming a fixed set of rules | 05:38 |
kanzure | and what would the input data be? | 05:40 |
fenn | i think what would happen is "clean money" would become more valuable and only be spent where it is required, and "dirty money" would be circulated in the dirty money economy, but there would still be significant overlap | 05:40 |
fenn | you couldn't trade clean money for dirty money because mumble mumble | 05:41 |
kanzure | as you increase the number of economy participants that will reject transactions because of some taint function output, you decrease the number of transactions that anyone is going to be able to make | 05:41 |
kanzure | also because the whole thing is a push system and not a pull system | 05:41 |
kanzure | also what's so bad about threaded screw leads? CaptHindsight was complaining. | 05:41 |
fenn | yeah he seemed stuck on the build quality thing even though nobody said anything about it | 05:42 |
fenn | i dont know what you mean "push system" | 05:43 |
kanzure | anyone can make a bitcoin transaction to any known bitcoin address | 05:43 |
fenn | oh | 05:43 |
fenn | well then they just gave you some money | 05:43 |
kanzure | and now your balance is tainted | 05:44 |
fenn | no, those particular bitcoins are tainted | 05:44 |
kanzure | you taint everything that appears in the transaction's outputs, and possibly all past/future transactions from that address even if those other transactions are unrelated to that utxo | 05:45 |
fenn | ok well you shouldnt reuse addresses | 05:46 |
kanzure | true dat | 05:46 |
fenn | i mean, this is a quirk of how bitcoin handles transcations, there are no actual bitcoins | 05:46 |
fenn | like there are no serial numbers | 05:46 |
kanzure | still, there is a utxo taint spreading factor there, if oytu are a hapless user and happen to make a transaction with multiple inputs (including at least one tainted) and multiple outputs, those outputs are now tainted. | 05:46 |
fenn | ok so imagine you decreased the taint factor by 0.1 each transaction, then people would just make pointless transactions back and forth until their money is clean again... but simple analysis of the transaction graph would show that somethign fishy is going on, so you hire a company to do analytics on the graph to tell you if the money is being laundered and if so how long ago, but then you have | 05:49 |
fenn | to trust this company not to be a pawn of $badguy | 05:49 |
kanzure | you cannot guarantee that others will ever include a taint reduction factor in their personal taint calculations | 05:49 |
fenn | eventually it will all have a nonzero taint value | 05:50 |
fenn | but that value will be different for every observer | 05:50 |
kanzure | right | 05:50 |
fenn | and i'm fine with that | 05:51 |
kanzure | so you don't think that will slow things down or destroy your ability to spend? | 05:51 |
kanzure | and what about users that don't run a taint function initially? are they now left out of your economy? | 05:51 |
fenn | users that don't filter their money would be expected to have an average taint score until criminals figure out they can take advantage of them to launder their tainted money | 05:52 |
fenn | if you take money from criminals you're the sort of person being filtered anyway | 05:53 |
kanzure | what if you (correctly!) believe that some law is unjust and the person is not really a criminal | 05:53 |
fenn | then you're (correctly?) taking a political stand by putting your money where your beliefs are | 05:54 |
fenn | it'd be more productive to try to convince others that the law is unjust and to set their taint filters accordingly | 05:54 |
kanzure | something something then only outlaws will.. er.. only criminals believe in.. uh.. well, i'll just wait for maaku to fix this. | 05:54 |
kanzure | maaku: halp | 05:54 |
fenn | there's something similar in the legal system called jury nullification... the jury can just say "meh we don't believe in this law so the defendant is considered innocent" | 05:55 |
fenn | someone who has their taint filter set too high won't be able to accept money from anyone | 05:56 |
fenn | i never liked marc fawzi's "energy accounting" because it reduced everything to one dimension | 05:57 |
kanzure | and then satoshi strolls through and is like "that's an excellent idea, marc" aaaaaaaaaaaaaaaaa | 05:57 |
fenn | i want to know what the web of inputs to this transaction looks like so i can decide if i want to participate in {rainforest clearcutting, strip mining, patent trolling, etc} | 05:58 |
fenn | not just "150GJ were wasted in this production process" | 05:58 |
fenn | so i don't think the system would ever completely lock up or converge to a specific value because people have different ideas about what is right and wrong | 06:00 |
kanzure | i wonder if i could blame marc fawzi for turning me off from those topics or something, thus why in january 2009 i was like "bitcoin is shit" | 06:00 |
fenn | yes, you certainly could | 06:00 |
fenn | sue, sue! | 06:01 |
fenn | for intellectual damages, by having damaged your intellect | 06:01 |
-!- ebowden [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 06:01 | |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 06:01 | |
-!- ebowden [~ebowden@128.250.233.154] has quit [Remote host closed the connection] | 06:01 | |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 06:02 | |
kanzure | i have 148 emails from him | 06:02 |
kanzure | did you know vinay gupta is involved these days | 06:03 |
kanzure | i think he's working with ralph merkle | 06:03 |
-!- ebowden [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 06:03 | |
fenn | in minecraft there is an emergent concept of "semantic compression" where you can make dirt out of composted tree leaves and then sift the dirt to get iron or silicon or grass seeds because it's all just "dirt" and not "dirt made from compost" | 06:04 |
kanzure | that's not fair | 06:05 |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 06:06 | |
fenn | i think maybe it's impossible to get anything done in the real world without semantic compression because it's too hard to operate if you never decide if X is a chair or not | 06:06 |
fenn | there's just an explosion of simulations and lower level decision making | 06:07 |
fenn | if you never put X in a category with extreme prejudice | 06:07 |
kanzure | heh apparently i sent marc fawzi some private emails where i attempted to convince him to give up his economics and instead focus on building tech | 06:08 |
fenn | well he was just posting to the wrong mailing list like a fucktard | 06:09 |
fenn | it would have been fine if he had just posted to p2presearch in the first place | 06:09 |
-!- Beatzebub_ [~beatzebub@d75-155-236-222.bchsia.telus.net] has quit [Quit: Beatzebub_] | 06:09 | |
kanzure | yeah how did he get there anyway | 06:10 |
fenn | i dunno.. factor E farm -> lifecycle analysis -> energy accounting? | 06:11 |
kanzure | no i mean to openmanufacturing | 06:12 |
archels | here's a new one along the range of "biologically inspired", "biologically plausible", "biologically realistic" and so on: "[...] a spike-timing-dependent learning rule which is reminiscent of experimental results" | 06:12 |
kanzure | .title http://p2pfoundation.net/backups/p2p_research-archives/2009-January/001097.html | 06:12 |
yoleaux | [p2p-research] Sustainable Abundance: Distributed vs Decentralized | 06:12 |
fenn | weird he lives in oakland and is tweeting about michael o church and lisp | 06:13 |
fenn | i don't get what this has to do with manufacturing or why he thinks it does | 06:16 |
kanzure | "manufacturing uses energy"? | 06:17 |
fenn | well i guess it's related in that manufacturing is part of the economy... but so is everything else | 06:17 |
kanzure | "Hello all, I was invited to this group by a friend just recently and I've been listening to the discussions with a good deal of curiosity and interest." | 06:18 |
-!- JayDugger [~jwdugger@108.19.186.58] has joined ##hplusroadmap | 06:18 | |
fenn | the openmanufacturing list did have a lot of discussions about decentralized manufacturing, like the "3d printer in every garage" manifesto | 06:18 |
fenn | i don't remember if there was an actual manifesto or not | 06:19 |
fenn | why does gmaxwell spell it "decenteralized"? | 06:19 |
kanzure | he has a spelling problem in general | 06:19 |
midnightmagic | it's a very, very good way to recognise it's actually him, fwiw | 06:19 |
kanzure | he spells "mechanism" like "mechenism" | 06:19 |
kanzure | and he does this weird grammar thing where he forgets to decompress an entire part of his sentences | 06:20 |
midnightmagic | and it's contagious. | 06:20 |
fenn | yeah that's common with thinking ahead of your typing speed ability | 06:20 |
kanzure | like: "but the constants its useless as a concrete explanation for things" -> "but the constants it uses are useless as a concrete explanation for things" | 06:20 |
midnightmagic | sometimes it won't be a compression-like thing but more like switching sentences midstream. | 06:21 |
kanzure | "you see what you have to do while reading my text is switch into the universe where i had already said the other sentence i thought i had been saying" | 06:21 |
midnightmagic | like he wrote half of it but because IRC input fields are short, he forgot what he wrote fifteen words ago and doesn't check it again (since he's composing partial responses in multiple windows, is my guess) | 06:21 |
fenn | ok i was just wondering if it was some political statement about geometry or something | 06:22 |
kanzure | not to my knowledge | 06:23 |
midnightmagic | lol | 06:23 |
midnightmagic | he would be amused i think, on some level, to discover people are assuming he's doing it on purpose because $smart_person_thing | 06:24 |
kanzure | midnightmagic: have you ever read the marc fawzi stuff that satoshi nakamoto was referencing? | 06:24 |
kanzure | midnightmagic: the backstory here is that i had banned marc fawzi because i hated him and his ideas :-) | 06:24 |
midnightmagic | kanzure: Not only have I probably not read it, but I wasn't even aware Satoshi mentioned anyone else ever. | 06:24 |
-!- ebowden [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 06:24 | |
midnightmagic | except wei dai in his paper. and finney to thank him a bunch of times. | 06:25 |
kanzure | http://diyhpl.us/~bryan/irc/bitcoin-satoshi/email-p2presearch-2009-02-13-023120.txt | 06:25 |
fenn | i am pretty convinced satoshi is hal finney | 06:26 |
kanzure | not wei dai? | 06:26 |
kanzure | although wei dai liked hal finney enough that he probably tried to emulate hal anyway | 06:26 |
midnightmagic | :-( satoshi was not hal. | 06:27 |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 06:27 | |
midnightmagic | wei dai is not satoshi either. or, if wei dai is, then I have the dubious honour of correcting his code. :) | 06:27 |
kanzure | http://extropians.weidai.com/extropians.3Q97/4356.html | 06:28 |
fenn | not sure if timelock cryptography really counts as "tunneling through the singularity" if there's nothing left to execute your code on the other side | 06:29 |
midnightmagic | Or no one left to read it. | 06:30 |
-!- GAit [~lnahum@2-230-161-158.ip202.fastwebnet.it] has quit [Read error: Connection reset by peer] | 06:30 | |
kanzure | http://diyhpl.us/~bryan/irc/bitcoin-satoshi/sourceforge/users.yaml | 06:31 |
fenn | heh that's a very short list | 06:32 |
-!- GAit [~lnahum@2-230-161-158.ip202.fastwebnet.it] has joined ##hplusroadmap | 06:32 | |
fenn | did people bitch and moan about you publishing their email addresses | 06:33 |
kanzure | yes, i had to .tar a file -_- | 06:34 |
-!- erasmus [~esb@unaffiliated/erasmus] has joined ##hplusroadmap | 06:35 | |
-!- eudoxia [~eudoxia@r167-57-172-92.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap | 06:39 | |
kanzure | weird seeing my name and email address in these lists | 06:39 |
kanzure | hah and jojack | 06:39 |
kanzure | ( http://diyhpl.us/~bryan/irc/bitcoin-satoshi/ning-users-p2pfoundation-2013-05-19.pdf ) | 06:39 |
-!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 06:40 | |
-!- NixCipher [~NixCipher@46.182.28.172] has quit [Ping timeout: 245 seconds] | 06:41 | |
fenn | why is it a pdf? | 06:42 |
eudoxia | kanzure: you should send the NNI a grand challenge proposal that's just a copy of the minimal toolset paper | 06:42 |
-!- poppingtonic [brian@unaffiliated/poppingtonic] has quit [Ping timeout: 256 seconds] | 06:43 | |
kanzure | fenn: that's how i received or found the file | 06:43 |
fenn | i recognize most of these names | 06:44 |
kanzure | yep | 06:44 |
-!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 250 seconds] | 06:45 | |
fenn | ok so maybe finney is not nakamoto, but then it's pretty amazing that finney never made bitcoin | 06:46 |
fenn | heh satoshi on scalability concerns http://diyhpl.us/~bryan/irc/bitcoin-satoshi/email-cryptography-2008-11-03-013743.txt | 06:48 |
-!- ebowden [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 06:50 | |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 06:51 | |
kanzure | many people have been linking to an old satoshi post where he was like "yep here's how you would roll out a fork, <example is something about increasing the max block size>" as evidence that "our glorious leader wants us to increase the block size, duh" | 06:51 |
-!- ebowden [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 06:52 | |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 06:52 | |
fenn | yes but this particular email is evidence that he had not really considered the risk of systemic decentralization | 06:53 |
fenn | derp. s/decentralization/centralization/ | 06:53 |
-!- NixCipher [~NixCipher@37.221.161.72] has joined ##hplusroadmap | 06:55 | |
-!- ebowden [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 06:55 | |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 06:56 | |
-!- ebowden_ [~ebowden@128.250.233.154] has joined ##hplusroadmap | 07:00 | |
-!- ebowden [~ebowden@128.250.233.154] has quit [Ping timeout: 256 seconds] | 07:00 | |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 07:04 | |
-!- ebowden_ [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 07:04 | |
-!- eudoxia_ [~eudoxia@r167-57-119-114.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap | 07:10 | |
-!- ebowden [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 07:11 | |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 07:11 | |
kanzure | hmph | 07:11 |
-!- eudoxia [~eudoxia@r167-57-172-92.dialup.adsl.anteldata.net.uy] has quit [Ping timeout: 246 seconds] | 07:13 | |
-!- ebowden [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 07:13 | |
-!- ebowden_ [~ebowden@128.250.233.154] has joined ##hplusroadmap | 07:13 | |
-!- eudoxia_ [~eudoxia@r167-57-119-114.dialup.adsl.anteldata.net.uy] has quit [Client Quit] | 07:13 | |
-!- eudoxia [~eudoxia@r167-57-119-114.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap | 07:14 | |
-!- flamoot [~yaaic@209-112-38-110.dedicated.allstream.net] has joined ##hplusroadmap | 07:16 | |
fenn | MKULTRA | 07:16 |
CaptHindsight | kanzure: I was trying to determine the amount of drama required for the project. I wouldn't use a ball screw or lead screw for the design since there is no contact or high loads. | 07:18 |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 07:19 | |
-!- ebowden_ [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 07:19 | |
CaptHindsight | printers that size are usually done with timing belts and optical encoder strips | 07:19 |
fenn | CaptHindsight: what don't you like about epson print heads specifically? | 07:20 |
CaptHindsight | fenn: I work with just about every printhead on the market. They all have their ups and downs | 07:21 |
CaptHindsight | Epson are low cost, slow, grey scale, don't circulate the fluid, no heaters, optimal viscosity 4-6cPs, awkward shape | 07:23 |
-!- ebowden_ [~ebowden@128.250.233.154] has joined ##hplusroadmap | 07:23 | |
CaptHindsight | they are made for aqueous inks that have the colorants well stabilized | 07:24 |
-!- ebowden [~ebowden@128.250.233.154] has quit [Ping timeout: 252 seconds] | 07:24 | |
kanzure | CaptHindsight: drama is up to you, i'm more interested in a working thing | 07:24 |
kanzure | and actuation can be trivially changed switched out by anyone that wants anyway | 07:25 |
fenn | "trivial" | 07:26 |
-!- ebowden_ [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 07:26 | |
kanzure | well it's trivial compared to the other stuff going on there | 07:26 |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 07:26 | |
fenn | it abstracts well at least | 07:26 |
kanzure | everyone does actuation, nobody does phosphoramidite chemistry.. stuff. | 07:27 |
fenn | hey how come you have such a hard-on for optical DLP chemistry, but no love for e-beam etching/sintering | 07:27 |
kanzure | doesn't ebeam stuff require lots of vacuum and vacuum pumping | 07:28 |
fenn | also why does nobody use ebeam for resin hardening at sub micron feature sizes | 07:28 |
kanzure | fenn: you should brain dump things to CaptHindsight because it sounds like he's going to start designing a dna synthesizer | 07:29 |
fenn | yes ebeam requires hard vacuum | 07:29 |
kanzure | CaptHindsight: we should try to go all out on this.. e.g. extremely high parallelism. | 07:29 |
fenn | is this just a repeat of POSAM? | 07:29 |
fenn | .title http://www.ncbi.nlm.nih.gov/pmc/articles/PMC507883/ | 07:30 |
kanzure | so far i think the answer is yes...? | 07:30 |
yoleaux | POSaM: a fast, flexible, open-source, inkjet oligonucleotide synthesizer and microarrayer | 07:30 |
CaptHindsight | kanzure: take a step back and think about what the samples need to go through after it's deposited | 07:30 |
kanzure | yes i'm not very clear about sample recovery after synthesis with posam | 07:31 |
fenn | it's just a piece of paper i think | 07:31 |
CaptHindsight | if drops are being deposited in a argon atmosphere | 07:31 |
fenn | reprap did atmosphere control by putting an airtight bag around the work table and the deposition head | 07:31 |
fenn | it would be straightforward to add a sample recovery airlock that didn't waste very much gas | 07:32 |
fenn | especially if you could just add more gas to make up for the lost dead volume | 07:32 |
eleitl | argon? what for? | 07:32 |
CaptHindsight | or do we just build a whole system? | 07:32 |
eleitl | are we talking MNT here? | 07:33 |
kanzure | eleitl: http://bioinformatics.org/pogo/ | 07:33 |
kanzure | eleitl: https://www.takeitapart.com/guide/94 | 07:33 |
CaptHindsight | nitrogen/argon | 07:33 |
eleitl | Oh, gas blanket. I see. | 07:33 |
eudoxia | oh hi eleitl great to see you're back | 07:34 |
kanzure | manuals and maintenance instructions for the abi391 are at http://diyhpl.us/~bryan/papers2/DNA/abi391/ | 07:34 |
eleitl | Howdy eudoxia. | 07:34 |
eleitl | Hope to stay around longer this time. | 07:35 |
-!- ebowden [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 07:35 | |
-!- ebowden_ [~ebowden@128.250.233.154] has joined ##hplusroadmap | 07:35 | |
kanzure | fenn: so wouldn't there be damage to the samples if the paper is exposed to regular atmosphere | 07:36 |
kanzure | i mean this isn't the same thing as the paper-dried dna samples that igem distributes | 07:36 |
eleitl | Are you doing reactions on paper? | 07:36 |
fenn | the reactants that you care about have already reacted together; anything else is just noise | 07:36 |
kanzure | eleitl: undecided | 07:36 |
eleitl | Cellulose is hygroscopic, so it will have some water absorbed. | 07:37 |
eleitl | Even if you bake it out, it will reabsorb in no time at all. | 07:37 |
kanzure | eleitl: CaptHindsight wants to use an inkjet head to deposit reactants, although i would still like to convince him to consider a pneumatic system | 07:37 |
-!- ebowden_ [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 07:37 | |
kanzure | (but i'm fine with inkjet stuff) | 07:37 |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 07:37 | |
eleitl | Piezo inkjet? Will it gunk up? | 07:37 |
fenn | i do wonder how you're supposed to get the dot size to be exactly the same every time. if not the same size, how are side reactions prevented | 07:37 |
kanzure | there are wash cycles | 07:38 |
kanzure | or.. something.. | 07:38 |
kanzure | i believe that answer is for both of you | 07:38 |
eleitl | Piezo has repeatable droplet size, but it will be likely different for different reagents. | 07:38 |
fenn | hmm probably not repeatable on a molecular scale | 07:38 |
CaptHindsight | eleitl: well define repeatable | 07:39 |
eleitl | Probably 1%, or thereabouts. I would need to search to know for sure. | 07:39 |
kanzure | you could use a microarray with wells (indentations) so that liquids are more likely to collect in the same location | 07:39 |
fenn | but i could see for example adding a 1.1mm diameter circle of A-base reactant and then adding a 0.9mm diameter circle of make-it-so reactant | 07:40 |
kanzure | also with wells you could use beads which could survive wash cycles | 07:40 |
fenn | (make-it-so is the deprotection reaction) | 07:40 |
kanzure | .title https://www.youtube.com/watch?v=0Qrm1ZjIuIU | 07:40 |
yoleaux | POSaM color test - YouTube | 07:40 |
CaptHindsight | I was just asked by the NIST about designing an inkjet with a microbalance to insure total uncertainty to <1% from drop to drop | 07:40 |
CaptHindsight | but not using Epson | 07:41 |
eleitl | http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2762308/ <-- another technique for repeatable droplets | 07:41 |
fenn | why is it peeing on the microscope slides | 07:41 |
-!- ebowden [~ebowden@128.250.233.154] has quit [Remote host closed the connection] | 07:43 | |
CaptHindsight | if you're trying to print organic semiconductors like LED's then drop size variation and placement are very important | 07:44 |
CaptHindsight | how critical is it to DNA synthesis? | 07:44 |
kanzure | i like how entire sections of this document are missing ("System performance" in particular) http://bioinformatics.org/pogo/POSAM_Man_Ch2_User_v1-0_040414.pdf | 07:44 |
CaptHindsight | how do drop size variations effect the outcome of the reactions? | 07:44 |
CaptHindsight | if obe drop from one step is off center from the previous ~75um dia drops by 25um how does that effect the outcome? | 07:45 |
CaptHindsight | obe/one | 07:45 |
CaptHindsight | the inkjet industry is really paranoid and controlling | 07:46 |
CaptHindsight | printheads specs are always under tight NDA's | 07:47 |
CaptHindsight | the Epsons are cheap and they sort of worked for the application | 07:47 |
CaptHindsight | thats why it was used in that design | 07:47 |
eleitl | how long did your Epson piezos live? | 07:48 |
CaptHindsight | depends on what you run through them and how you treat them (how often you flush them) | 07:48 |
eleitl | what liquids are especially corrosive for Epsons? | 07:49 |
CaptHindsight | they don't circulate fluid in the reservoir, what goes in is tough to get back out the way it came in, thats the main problem with Epson heads | 07:49 |
fenn | the posam thing used rather expensive (~$5k) danaher motion control components, so i doubt print head cost was a big concern | 07:50 |
fenn | oops it was parker, not danaher | 07:50 |
fenn | "The three-axis positioning system utilizes Parker-Daedal 506-series ball screw linear actuators driven by servo motors" | 07:50 |
CaptHindsight | you can get a Epson head for the cost of the printer, they start at under $100 and most of the heads are very similar | 07:50 |
eleitl | so you consider them disposables | 07:51 |
CaptHindsight | on the edge, the next piezo heads up the tree have 126 nozzles and cost ~$300ea | 07:52 |
CaptHindsight | one ink chamber | 07:52 |
CaptHindsight | and fire ~8k/s top speed | 07:52 |
kanzure | well you could always do a chamber wash cycle by moving the head to some waste area, then flushing | 07:53 |
CaptHindsight | 40pl-60pl drop size | 07:53 |
CaptHindsight | if you don't mind flushing a few mL down the drain each time | 07:53 |
CaptHindsight | whats the substrate going to be? | 07:54 |
CaptHindsight | synthetic paper like teslin? | 07:55 |
-!- flamoot [~yaaic@209-112-38-110.dedicated.allstream.net] has quit [Quit: ilugiysh <3 tomalt - Yaaic - Yet another Android IRC client] | 07:55 | |
fenn | i'd expect a glass microscope slide or a 96 well plate | 07:55 |
fenn | (96 well plates are usually styrene but come in a variety of materials) | 07:56 |
kanzure | nah this was like 10k drops/slide, i also think it would be annoying to calibrate with multiple microarray plates | 07:56 |
fenn | array plates are good for detecting the presence/absence of one specific sequence out of thousands of targets, but it's a pain to combine 10,000 oligos into one test tube for gene synthesis | 07:58 |
fenn | an array plate is just a glass slide | 07:58 |
CaptHindsight | I have to call Germany and it's already 5pm there ,, back in a bit | 08:00 |
CaptHindsight | but keep going :) | 08:00 |
kanzure | er i thought array plates had wells | 08:00 |
fenn | i think they are just flat plates but i could be wrong | 08:01 |
fenn | there are "microcapillary array plates" too, but i'm not sure what they are used for | 08:01 |
kanzure | those 1 million pixel array plates usually have (micro) beads inside each well, and then you use a micropipette to extract each bead from each well | 08:02 |
kanzure | http://longnow.org/revive/wp-content/uploads/2013/02/img4921a.jpg | 08:03 |
kanzure | gah why does nobody make good pictures of equipment | 08:03 |
kanzure | or video >:( | 08:03 |
fenn | our uploaded blurry .jpg files must last 10,000 years | 08:04 |
CaptHindsight | what do you prefer the samples to printed on | 08:04 |
kanzure | the thing that works best | 08:05 |
CaptHindsight | we can just make that as well | 08:05 |
-!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 08:05 | |
CaptHindsight | inert substrate with 100K wells/indentations to isolate each well | 08:05 |
CaptHindsight | 4pl is the drop volume of the epsons | 08:06 |
CaptHindsight | a drop of that volume on a porous substrate ends up making a 50-75um splotch | 08:07 |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 08:07 | |
CaptHindsight | how much of each reagent is active component and how much is just vehicle? | 08:08 |
fenn | 96 well plates are pretty standard for doing synthetic biology reactions in, so i'd suggest that to be a good target to aim for. however there is a distance of about 4mm between the top of each well and the bottom that you'd have to shoot through | 08:08 |
fenn | some of these microarray plates being advertised look like they use a flat plate with a set of 96 wells that can be attached on top | 08:09 |
kanzure | there was some good images of george church's setup, with micropipetting.. there was even a video. where did it go? :-( | 08:10 |
CaptHindsight | fenn: this type? http://www.taq-dna.com/rich_files/bilder/96-plate_PCR_ABI_and_Sequencer.png | 08:10 |
kanzure | way too large | 08:10 |
fenn | CaptHindsight: no, that has rounded bottoms which is not what you want | 08:11 |
CaptHindsight | what would be ideal? | 08:11 |
kanzure | the one that works | 08:11 |
CaptHindsight | an array of 1nL wells? | 08:11 |
kanzure | (we're still digging, trying to find evidence of this thing existing) | 08:11 |
kanzure | maybe this was just from the other timeline, blah | 08:12 |
fenn | http://www.usascientific.com/96-well-plate-flat-bottom-non-sterile-5sleeve.aspx | 08:12 |
CaptHindsight | maybe think about it this way..... | 08:12 |
CaptHindsight | how do you really want it to work? | 08:13 |
fenn | i might just be dumb and a flat plate is actually more versatile | 08:14 |
-!- ebowden [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 08:14 | |
CaptHindsight | what size wells, number of wells, number of different fluids in a single synthesizer | 08:14 |
-!- ebowden_ [~ebowden@128.250.233.154] has joined ##hplusroadmap | 08:14 | |
-!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 244 seconds] | 08:14 | |
kanzure | here are some pics of their super awesome setup http://www.nature.com/nbt/journal/v28/n12/extref/nbt.1710-S1.pdf | 08:15 |
kanzure | (from their "High-fidelity gene synthesis by retrieval of sequence verified DNA identified using high-throughput pyrosequencing" paper) | 08:15 |
kanzure | "picotiter plate" | 08:15 |
CaptHindsight | it's funny I've been building very similar equipment but not for DNA | 08:16 |
kanzure | picotiter is a good idea | 08:16 |
fenn | kanzure: hey if you transferred it with a laser beam you'd be cambrian genomics... | 08:16 |
kanzure | these are 4096x4096 arrays | 08:17 |
-!- drewbot [~cinch@ec2-54-91-83-253.compute-1.amazonaws.com] has quit [Remote host closed the connection] | 08:17 | |
-!- drewbot [~cinch@ec2-54-158-222-200.compute-1.amazonaws.com] has joined ##hplusroadmap | 08:17 | |
kanzure | "Finally, in a proof-of-principle synthesis with particles in microwells, an unprecedented density of 500,000 spots/cm2 was demonstrated." | 08:17 |
kanzure | "synthesis in microcavities (500,000 cavities per cm2, 10-µm cavity depth)" | 08:18 |
kanzure | (text is quoted from http://www.mdpi.com/2076-3905/3/4/245/htm ) | 08:18 |
-!- ebowden_ [~ebowden@128.250.233.154] has quit [Ping timeout: 252 seconds] | 08:18 | |
CaptHindsight | http://imagebin.ca/v/25mmIFYFS9l2 this one is 5 axis with 5um repeatability | 08:18 |
CaptHindsight | holds inkjet and laser and microscopes for doing <10um geometry deposition | 08:19 |
CaptHindsight | parts are up to 250mm^3 | 08:19 |
fenn | kanzure so each well is approx ~14 micron diameter | 08:19 |
fenn | i think that is too small for inkjets | 08:20 |
kanzure | cool, you could use a 10 micron bead and get something sorta snug | 08:20 |
kanzure | yeah probably | 08:20 |
kanzure | not the end of the world tho. there's a lot you can do with just 10k drops/slide. | 08:20 |
CaptHindsight | there are printheads with <1pl drop volumes or you just make one | 08:20 |
fenn | what's the diameter of a 1pl drop? | 08:21 |
CaptHindsight | in flight or? | 08:21 |
fenn | .wa diameter of a 1 picoliter sphere | 08:21 |
yoleaux | sphere: volume 1 pL (picoliter): diameter: (6/pi)^(1/3) pL^(1/3) (cube root picoliters)~~1.2407 pL^(1/3) (cube root picoliters); Visual representation: http://is.gd/TtzOKP; Unit conversions: 4.885×10⁻⁴ inches; 12.41 µm (micrometers); 0.01241 mm (millimeters); Comparisons as diameter: ~0.6 × wool fiber diameter (~20 µm); ~typical cotton fiber diameter (~10 µm); ~typical fog or cloud droplet diameter (~10 µm) | 08:21 |
fenn | hum ok | 08:21 |
fenn | 12 micron | 08:21 |
kanzure | looks like we could relax the requirements and still have large-scale things going on | 08:22 |
kanzure | (instead of trying to hit all system limits, hehe) | 08:22 |
CaptHindsight | it's a teardrop that wiggles around into a near sphere then it hits the surface at 1-10m/s | 08:22 |
CaptHindsight | if it hits glass it splatters | 08:23 |
fenn | what if it hits a 14 micron diameter well etched into the glass? | 08:23 |
fenn | possibly containing a 10 micron diameter bead | 08:23 |
CaptHindsight | thats the thing | 08:24 |
CaptHindsight | is it the first material in the well or is it hitting the previous layers/material | 08:24 |
fenn | i have no intuition about this sort of velocities and viscosities | 08:24 |
fenn | presumably the previous drop would be evaporated? | 08:25 |
CaptHindsight | https://www.youtube.com/watch?v=hq-r0vAMURw tij drops in flight | 08:25 |
fenn | ooo aaah | 08:26 |
CaptHindsight | https://www.youtube.com/watch?v=PDq5qPbVt4M piezo drops | 08:26 |
CaptHindsight | https://www.youtube.com/watch?v=XZRq1JzRxB8 Ink Jet - Drop Formation and Surface Interaction | 08:27 |
kanzure | splattering could be mitigated by spacing the wells out | 08:27 |
kanzure | depending on splat trajectory | 08:27 |
CaptHindsight | https://www.youtube.com/watch?v=E47t4wHSW84 Drop-Substrate Interaction: Dimatix Printhead on Glass | 08:28 |
kanzure | also the beads may not be necessary? or they might be helpful? uh | 08:28 |
kanzure | also you can develop fluids that are less likely to splash | 08:28 |
CaptHindsight | kanzure: if they still have the chemical properties required | 08:29 |
CaptHindsight | inkjet heads are happy in a very narrow range of operating conditions | 08:30 |
CaptHindsight | i consider them failed lab experiments since you can just look and then wrong and they clog or act up | 08:30 |
fenn | splattering could be mitigated by adding forests of pillars to gradually slow the droplet down and increase the wetted surface area so that it doesn't want to leave the surface as esily | 08:31 |
CaptHindsight | but if I know what we want to jet and what it's going to hit I'll figure out what makes the most sense to use | 08:31 |
fenn | instead of a sharp impediance transition from empty space to solid glass you'd have a low density of pillars in the center of the well gradually increasing to a high density near the edges of the well | 08:32 |
CaptHindsight | the next thing is the environment that the samples need to be in until they are done being processed? | 08:32 |
kanzure | CaptHindsight: if there are 15 micron diameter holes (wells) on a plate under the inkjet head, can you get an inkjet to reliably deposit liquid into the correct holes? | 08:32 |
kanzure | also do you think an inkjet could deposit 10 micron diameter (glass/gold/plastic) beads by any chance | 08:33 |
CaptHindsight | kanzure: that printer in the pic I posted looks for <10um valleys and puts fluid into them | 08:33 |
CaptHindsight | kanzure: yes, but it would not use an Epson head | 08:34 |
kanzure | if the system uses beads then you gotta prep it by inserting beads- this can be aciheved by washing the surface with a liquid + beads, but you have no guarantees about the beads getting into every hole | 08:34 |
fenn | what about scraping the surface with a rubber blade to remove extra beads? | 08:35 |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 08:35 | |
CaptHindsight | we can also make patterns in aluminum and keep the pores isolated | 08:36 |
-!- ebowden [~ebowden@128.250.233.154] has quit [Read error: Connection reset by peer] | 08:36 | |
-!- ebowden_ [~ebowden@128.250.233.154] has joined ##hplusroadmap | 08:37 | |
CaptHindsight | 40nm wells is very dense :) | 08:37 |
fenn | too dense | 08:37 |
fenn | can't even see things that small | 08:38 |
CaptHindsight | how small can the pores be and still have room for a bead of DNA? | 08:38 |
CaptHindsight | https://engineering.purdue.edu/MSE/Research/REU/StudentPages/2003/bBowser/BowserFig1.jpg | 08:39 |
fenn | umm... well a dna helix is 2nm in diameter so it could probably fit in a 40nm well, but a bead would just be getting in the way at that scale | 08:39 |
-!- ebowden_ [~ebowden@128.250.233.154] has quit [Ping timeout: 245 seconds] | 08:41 | |
fenn | 0.2nm additional length per base pair | 08:41 |
CaptHindsight | we can also form or print few um wells | 08:42 |
CaptHindsight | lets go through the process from end to end | 08:43 |
fenn | i'm still not clear what the purpose of this machine is | 08:44 |
fenn | what is the DNA to be used for? | 08:44 |
-!- ebowden [~ebowden@128.250.233.154] has joined ##hplusroadmap | 08:46 | |
CaptHindsight | I just make the machines :) | 08:46 |
-!- HEx1 [~HEx@hexwab.plus.com] has joined ##hplusroadmap | 08:50 | |
CaptHindsight | fenn: http://www.pemed.com/lab/synthes/abipcr01_5.jpg what DNA do you make with this? | 08:50 |
-!- ebowden [~ebowden@128.250.233.154] has quit [Ping timeout: 245 seconds] | 08:51 | |
fenn | large quantities of 80-mers | 08:51 |
fenn | relatively low error rate, but they're all the same sequence | 08:51 |
fenn | this is useful for PCR primers and not much else. it's possible to stitch them together to synthesize entire genes, but it's a pain | 08:52 |
CaptHindsight | http://bioinformatics.org/pogo/POSAM_Man_Ch1_Assembly_v1-2_040601.pdf | 08:52 |
CaptHindsight | Epson printhead based synthesizer | 08:53 |
CaptHindsight | these are the two systems I was asked about yesterday | 08:54 |
fenn | "if you didn't think the technology was important you wouldn't be reading this, so we'll just tell you what you need to know to build and run the instrument" | 08:54 |
fenn | heh | 08:54 |
kanzure | fenn: i wasn't worried about removing beads, just about getting a bead into every hole (which is what the head would be used for) | 08:56 |
kanzure | oh i guess you could also just use some leds to flash light and check for bead presence, i'm sure | 08:56 |
fenn | the head is depositing beads? | 08:56 |
kanzure | well i was asking | 08:56 |
kanzure | if the answer is yes then that would be useful for setup | 08:56 |
kanzure | why would all the 80mers be the same sequence if you have a large number of pores? | 08:57 |
CaptHindsight | I have to look over how they are processed | 08:57 |
kanzure | if you have 10 million pores with 10 million different 80mers then you could do a lot of interesting things | 08:57 |
CaptHindsight | there are flow charts but they assume that I'm already assuming something about temperature and atmosphere | 08:58 |
fenn | all the 80mers are the same sequence in a bulk liquid synthesis machine, which is what he was asking about (the pictured device) | 08:58 |
kanzure | oh ok | 08:58 |
kanzure | yea it's true that stitching sucks, dunno what to tell you, gibson assembly magic perhaps | 08:58 |
fenn | yeah but you still have to do a zillion synthesis runs | 08:59 |
kanzure | in parallel, at least | 08:59 |
fenn | not in parallel if you only have one machine | 08:59 |
kanzure | CaptHindsight: what's the deposition rate per drop.... | 08:59 |
kanzure | like let's run it on continuous motion and get the drops flying with horizontal motion to get to the correct destination :P | 09:00 |
CaptHindsight | kanzure: with Epson ~14Khz max for 4pL | 09:00 |
fenn | i dont get why affymetrix chips don't just deal with the high error rate and make long error-riddled hybridization sequences anyway | 09:03 |
kanzure | .wa 1 million / (14000 / second) | 09:03 |
yoleaux | 1000000/(14000/(second)): 71.43 seconds; Unit conversions: 1 minute 11.43 seconds; 1.19 minutes; Comparison as half‐life: ~(0.24 ~1/4) × half-life of uranium-241 (~300 s); Corresponding quantities: Distance x traveled by light in a vacuum from x = ct:: 2.141×10⁷ km (kilometers): 0.14 au (astronomical units): 2.141×10¹⁰ meters: 13.31 million miles; Frequency nu from nu = 1/T:: 0.014 Hz (hertz) | 09:03 |
fenn | DNA hybridization doesn't care if there is a high error rate | 09:04 |
-!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 09:04 | |
fenn | (1cm/s)/14kHz = 0.7 micron | 09:05 |
CaptHindsight | the Epson head in that microarray paper was out of an Epson 700 | 09:05 |
fenn | if you're moving at a constant speed, the horizontal velocity is just a linear offset | 09:05 |
CaptHindsight | the replacement part head was sold for ~$70ea | 09:05 |
CaptHindsight | http://service-manuals.waraxe.us/en/epson/printer/stylus-color-700 this was very hacked back in the day | 09:07 |
CaptHindsight | http://www.directcolorsystems.com/ these are all made from hacked Epson desktop printers | 09:09 |
kanzure | fenn: not sure what you were calculating | 09:09 |
kanzure | pore spacing? | 09:09 |
fenn | the horizontal velocity component of a bead | 09:10 |
CaptHindsight | same for most of those t-shirt inkjets | 09:10 |
kanzure | i think you mean "a bed or a drop" | 09:10 |
kanzure | *bead | 09:11 |
fenn | right | 09:11 |
CaptHindsight | http://easydtg.com/ and similar | 09:11 |
fenn | this POSAM paper assumes you're going to be using the array for hybridization, which seems not so useful to me | 09:11 |
fenn | however if you could put 1000 overlapping 80mer sequences in each well of a 96 well plate, that would make something like a 1000*40*96 length sequence | 09:13 |
kanzure | why are you talking about a 96 well plate there | 09:14 |
fenn | because i like them | 09:14 |
fenn | because i dont want to move and fucking beads around | 09:14 |
fenn | any* | 09:14 |
fenn | yeah i realize there will be a high error rate | 09:15 |
kanzure | no i mean... 1000 overlapping 80mer sequences, with 40 bp unique content per 80 bp molecule, then that should be 1000 * 40 | 09:15 |
fenn | yes and 96 of them | 09:15 |
fenn | because it's in 96 separate wells | 09:15 |
kanzure | the number of wells does not seem to modulate how many overlapping sequences though | 09:15 |
fenn | correct | 09:15 |
kanzure | ok. heh. | 09:15 |
fenn | oh, so you would combine all of them in a later step | 09:16 |
fenn | each well would self-assemble into a 1000*40 length sequence | 09:16 |
kanzure | yes we should think gibson assembly protocol thoughts for the machine design | 09:16 |
fenn | hacking the gibson... | 09:17 |
CaptHindsight | https://en.wikipedia.org/wiki/Gibson_assembly | 09:20 |
CaptHindsight | do the fragments automatically anneal after trimming? | 09:20 |
fenn | each well in a 96 well plate has surface are of .3165cm^2 so there's maximum 316500 possible 10 micron^2 spots assuming zero spacing | 09:21 |
fenn | however i think this is too many sequences to deal with reliably in one self-assembly reaction | 09:22 |
fenn | "The method can simultaneously combine numerous (>10) DNA fragments based on sequence identity." | 09:23 |
kanzure | "numerous" ">10" | 09:24 |
fenn | wow what a crap article | 09:24 |
fenn | god dammit | 09:28 |
fenn | searching for "diy gibson assembly" gives me diyhpl.us results | 09:28 |
fenn | http://diyhpl.us/~bryan/papers2/bio/dna-assembly/ | 09:29 |
fenn | hmm should i read Synthetic_assembly_overview.pdf | 09:30 |
ParahSailin | how do you get tmux to automatically have the process names in the panel name on the bottom | 09:31 |
fenn | oh i had forgotten about combinatorial assembly for different permutations | 09:31 |
ParahSailin | it seems centos has some magic in bashrc that ubuntu does not | 09:32 |
JayDugger | I think Ubuntu has a preference for byobu vice tmux. | 09:34 |
fenn | wtf this assembly overview is talking about a _4_ base pair overlap | 09:36 |
fenn | that's 256 sequences max | 09:36 |
CaptHindsight | how much oxygen in the atmosphere is allowed during processing? | 09:37 |
CaptHindsight | "The system is flushed with nitrogen at a 40 lpm flow rate for 40 minutes prior to oligo synthesis. Assuming good mixing of inert gas with air, the time constant (τ) for the flushing process would be approximate 7.9 minutes." | 09:41 |
CaptHindsight | internal volume of approximately 510 liters | 09:41 |
ParahSailin | is it oxygen or water vapor that is the poison? | 09:42 |
CaptHindsight | so they don't pull a vacuum they just flush with nitrogen | 09:42 |
CaptHindsight | it mentions both water and oxygen | 09:42 |
fenn | Excluding waterfrom solvents. | 09:45 |
fenn | The most critical factorin any synthesis is how the reagents are handled to excluded water fromthe system. From the moment a bottle is opened, it is in contact with waterin the air, and all the solvents used are hygroscopic and will absorb watervapor, which reduces yields. This problem is so severe that it is advisableto avoid large-scale, lengthy, or important runs on rainy or high humiditydays. | 09:45 |
fenn | i don't see anything about oxygen in here (except for the oxidation step with iodine) | 09:46 |
fenn | reading http://molbio.mgh.harvard.edu/szostakweb/protocols/oligo/ | 09:46 |
fenn | not sure why the posam instruction manual mentions oxygen removal | 09:47 |
CaptHindsight | fenn: from what i can see the POSaM arrayer just places a mix of 6 drops on top of one another | 09:48 |
CaptHindsight | fenn: am I missing something here? | 09:48 |
fenn | it's possible whoever wrote the POSAM manual didn't understand the chemistry and thought that the nitrogen was to remove oxygen, but it was actually used to remove water vapor | 09:48 |
CaptHindsight | ok | 09:49 |
fenn | CaptHindsight: that sounds about right... it might also do a bulk liquid wash step | 09:49 |
fenn | i've never seen it in action | 09:49 |
CaptHindsight | what happens between each drop? | 09:50 |
CaptHindsight | is there some time to wait for the reaction to occur? | 09:50 |
CaptHindsight | before adding the next drop? | 09:50 |
fenn | i imagine the drops are something like A G C T deprotect wash | 09:52 |
CaptHindsight | I see the program now: wash with acentonitrile, print, wash, double print, wash, oxidize, wash, etc | 09:52 |
fenn | double print? | 09:53 |
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has joined ##hplusroadmap | 09:53 | |
fenn | is "wash" just "blow dry with nitrogen"? | 09:53 |
CaptHindsight | page 23 in the Posam manual chapter 1 | 09:53 |
CaptHindsight | “WASH,” dispenses acetonitrile from nozzle 1 for 0.7 seconds. | 09:54 |
CaptHindsight | “DRY” uses the blower to remove the liquid from the same slides. | 09:54 |
CaptHindsight | the prinhead fluid lineup: BANK1=Unused | 09:55 |
CaptHindsight | BANK2=XTetrazole | 09:55 |
CaptHindsight | BANK3=Adenine | 09:55 |
CaptHindsight | BANK4=Cytosine | 09:55 |
CaptHindsight | BANK5=Guanine | 09:55 |
CaptHindsight | BANK6=Thymine | 09:55 |
fenn | 2. ACTIVATION-COUPLING)Following deblocking of the 5 hydroxyl group, the next protected phosphoramiditeis delivered to the reaction column along with the weakly acidic activatortetrazole | 09:56 |
CaptHindsight | “XTetrazole” is the catalyst | 09:56 |
fenn | ok i think there is slightly different chemistry being used here (unsurprisingly) | 09:57 |
CaptHindsight | heh | 09:57 |
CaptHindsight | give a few days to catch up | 09:57 |
kanzure | the abi 391 manuals have some details on their chemistry too http://diyhpl.us/~bryan/papers2/DNA/abi391/ | 09:58 |
kanzure | yeah i'm not sure how the wash works in posam- can't be a wash, because you'll lose your work | 09:59 |
kanzure | nitrogen gas blast might just blow away the reaction material anyway | 09:59 |
CaptHindsight | says “WASH,” dispenses acetonitrile | 09:59 |
CaptHindsight | a nice solvent | 10:00 |
CaptHindsight | not really a blast for removal | 10:00 |
CaptHindsight | just to dry like a hair dryer vs hair removal by high pressure | 10:01 |
CaptHindsight | a few orders of magnitude difference in pressure :) | 10:01 |
fenn | .title http://youtu.be/0Qrm1ZjIuIU | 10:02 |
yoleaux | POSaM color test - YouTube | 10:02 |
CaptHindsight | what vehicle is the ACGT in? | 10:02 |
fenn | around 1m25s would seem to suggest it's a violent blow dry | 10:02 |
kanzure | but you'll just blow everything away :/ | 10:03 |
fenn | no, the dna is covalently bound to the substrate | 10:03 |
kanzure | perhaps it's synthesized on to a linker attached to the glass | 10:03 |
kanzure | and removal? | 10:03 |
kanzure | oh perhaps no removal, just pcr | 10:04 |
kanzure | no, wait... | 10:04 |
fenn | "To cleave the DNA fromthe support matrix and remove protecting groups completely, the supportbound product must be treated with concentrated ammonia at 55° to 60°Covernight" | 10:04 |
kanzure | but then... how do you individually transfer different drops. | 10:04 |
kanzure | oh i think they just dunk the whole slide | 10:04 |
kanzure | submerge | 10:05 |
kanzure | bleh | 10:05 |
CaptHindsight | https://youtu.be/0Qrm1ZjIuIU?t=1m21s vs https://www.youtube.com/watch?v=0Qrm1ZjIuIU&feature=youtu.be | 10:05 |
fenn | POSAM is designed for hybridization with mRNA extracted from cells, so they never remove the DNA from the slide anyway | 10:05 |
kanzure | ah so perhaps we should put some design thought into this. wells would be a good start. | 10:06 |
kanzure | so conflicted. who has time to wait for manually handling a few million beads? | 10:06 |
CaptHindsight | give me the list of problems vs possible solutions | 10:06 |
fenn | so i like the 96 well plate format, but it's a long distance from the top of the plate to the bottom of the well. however you can just add the sides of the well later on, after you've synthesized the DNA stuck to the bottom of the well | 10:06 |
fenn | so i'd say synthesize dna in 96 circular spots on a flat plate | 10:07 |
fenn | or however many wells you want | 10:07 |
fenn | i think 384 is the other common number | 10:08 |
kanzure | that's not enough to do anything interesting | 10:08 |
kanzure | 96 or 384 is enough for making some primers i guess | 10:08 |
fenn | you could make 384 sets of primers :P | 10:09 |
kanzure | with a million wells you could possibly assemble multi-million bp molecules, including multiple separate genes | 10:09 |
kanzure | or permutations on various genes | 10:09 |
fenn | no dumbass you do self-assembly in each well | 10:09 |
kanzure | you said synthesize in the 96 well format | 10:09 |
kanzure | or.. something | 10:10 |
fenn | yes so you have 96 sets of 1000 oligos | 10:10 |
fenn | the parallel assembly is tricky and will probably fail though | 10:10 |
fenn | so maybe 1000 is optimistic. it doesn't seem too bad; with 40bp overlap that's 4^40 possible strand combinations which is way more than 1000 | 10:11 |
fenn | i don't really know why gibson assembly fails in practice | 10:11 |
fenn | kanzure: cambrian genomics is using a million beads per slide because 99% of their beads are completely empty in order to ensure that each bead only has one dna molecule on it | 10:14 |
fenn | you don't actually have to move a million beads one by one | 10:14 |
fenn | you only have to move 1%*1million = 10,000 beads | 10:14 |
fenn | anyway i don't know why we're even talking about beads | 10:16 |
fenn | at some point you clone a single molecule of DNA and either it has errors or it doesn't | 10:18 |
fenn | if it has errors you throw it away | 10:19 |
fenn | if it doesn't have errors you keep it | 10:19 |
fenn | you'd like to do this as soon as possible in the assembly process, but you also don't want to move around a million beads | 10:20 |
fenn | so i say self-assemble a large number of short strands that are not likely to contain errors, because they are short. do this in the same well you synthesized them in | 10:21 |
fenn | then you only have to deal with 96 "things" instead of 10,000 "things" | 10:22 |
fenn | i'd rather throw away 91 "things" with a 95% fail rate than throw away 500 "things" with a 5% fail rate | 10:24 |
fenn | i need some help with the math regarding the expected error rate vs a given oligo length | 10:25 |
CaptHindsight | making slides with small wells is not a problem | 10:26 |
kanzure | beads because it is easier to move beads | 10:26 |
fenn | szostak says: P (n,m,x)= [m!/(m - n)!n!][x]n[1 - x]m-n | 10:26 |
fenn | 10:26 | |
kanzure | not sure i understand why they had so many unused beads...? | 10:26 |
fenn | where P is theprobability of finding n errors in an oligonucleotide m inlength with x level of misincorporation (fraction "wrong" nucleotides delivered) | 10:26 |
kanzure | oh, one dna molecule per bead. not sure we have that requirement though. | 10:26 |
CaptHindsight | whats in the well after it's done printing and reacting? what is that bit called? | 10:29 |
CaptHindsight | a short strand? | 10:29 |
fenn | P (n,m,x)= [m!/(m - n)!n!][x]^n[1 - x]^(m-n) | 10:29 |
fenn | this equation is totally wacky if you're trying to get n=0 | 10:30 |
-!- GAit [~lnahum@2-230-161-158.ip202.fastwebnet.it] has quit [Read error: Connection reset by peer] | 10:30 | |
fenn | CaptHindsight: an oligomer | 10:30 |
fenn | .wik oligomer | 10:31 |
yoleaux | "In chemistry, an oligomer (i/əˈlɪɡəmər/) (oligo-, "a few" + -mer, "parts") is a molecular complex that consists of a few monomer units, in contrast to a polymer, where the number of monomers is, in principle, not limited. Dimers, trimers, and tetramers are, for instance, oligomers composed of two, three and four monomers, respectively." — https://en.wikipedia.org/wiki/Oligomer | 10:31 |
CaptHindsight | ok so same term in polymer chemistry | 10:31 |
CaptHindsight | when do they become beads? | 10:32 |
CaptHindsight | at what length? | 10:32 |
-!- GAit [~lnahum@2-230-161-158.ip202.fastwebnet.it] has joined ##hplusroadmap | 10:32 | |
fenn | a bead is made of glass or iron or polystyrene | 10:32 |
CaptHindsight | oh so those are just the carriers | 10:33 |
fenn | dna is chemically bound to the bead to make it easier to move around | 10:33 |
fenn | dna does bunch up into blobs once it gets over 1000 base pairs or so | 10:33 |
CaptHindsight | what I thought they were, but they aren't :) | 10:34 |
CaptHindsight | http://www.sumanasinc.com/webcontent/animations/content/highthroughput.html | 10:35 |
fenn | gah they're talking | 10:38 |
CaptHindsight | what causes a oligo to fail? impurities? | 10:45 |
fenn | reaction kinetics i think... it takes a certain amount of time for each step to complete, and the longer you wait the more complete it becomes but it's never 100% (?) | 10:49 |
CaptHindsight | depositing ACGT and some catalysts layer by layer is a no brainer | 10:50 |
fenn | if you want an 80mer and each step is 99% efficient your probability of success is .99^80 = 44% | 10:50 |
fenn | er, overall efficiency | 10:51 |
fenn | 56% of the molecules will have one or more errors | 10:51 |
CaptHindsight | how long is the reaction time per step? does it vary based on the volume of the material? | 10:51 |
fenn | i have no idea | 10:52 |
fenn | 30 seconds? | 10:52 |
CaptHindsight | whats the concentration of the ACTG in the vehicle? | 10:54 |
CaptHindsight | how many oligomer strands are being formed in each well if you are placing 4pL drops in each layer? | 10:55 |
CaptHindsight | I work with monomers and oligomers every day, just not ACTG types | 10:55 |
fenn | the posam example takes 160 seconds per cycle | 10:56 |
CaptHindsight | I'm just thinking about synthesizing and filtering out bad strands on the fly | 10:56 |
fenn | szostak says 25mg/ml or maybe 0.1 to 1 M | 10:59 |
fenn | per ml of what i don't know yet | 10:59 |
fenn | oh anhydrous acetonitrile is the solvent | 11:00 |
fenn | "CAUTION: Acetonitrile vapor is poisonous." | 11:00 |
fenn | "it may be useful to producesmall amounts of dry acetonitrile immediately prior to synthesis sincethe acetonitrile will accumulate water after the bottle has been opened.Moisture greatly diminishes synthetic yield. In this case acetonitrilecan be prepared in any laboratory equipped for routine distillations. However,the time and effort involved in setting up and maintaining a still shouldbe | 11:02 |
kanzure | re: impurities, see http://diyhpl.us/~bryan/papers2/DNA/Origin%20of%20impurities%20in%20oligonucleotides.pdf | 11:02 |
fenn | balanced against the cost of obtaining dry acetonitrile." | 11:02 |
fenn | Oligonucleotide impurities originate from: | 11:06 |
fenn | x incomplete addition of nucleotides forming so-called truncated fragments, whose synthesis has been prematurely terminated by a subsequent capping step; | 11:07 |
fenn | x inefficient capping of sequences that have failed to incorporate a nucleotide forming oligonucleotides with internal deletion | 11:07 |
fenn | s known as (n – x) fragments | 11:07 |
fenn | x inefficient detritylation, also resulting in (n – x) fragments | 11:07 |
fenn | x premature detritylation during coupling resulting in so-called (n + x) fragments which have duplicated nucleotides in the sequence | 11:07 |
fenn | x depurination during the detritylation step, forming oligonucleotides with abasic sites which are cleaved later by ammonia during the deprotection stage | 11:08 |
fenn | x incomplete deprotection; | 11:08 |
fenn | x oligonucleotide derivatization with reagents used or formed in the synthesis and deprotection steps | 11:08 |
fenn | Generally, impurities resulting from failed couplings and depurination are present in the greatest amounts in synthetic oligonucleotides. | 11:08 |
kanzure | so basically...... yield problems everywhere. | 11:09 |
fenn | (x's above are bullet points) | 11:09 |
kanzure | specifically, which step if any is *not* listed in that list heh | 11:09 |
fenn | well that's the point, every step introduces error | 11:11 |
CaptHindsight | so theoretically if you had a nano-dispenser of each monomer you could cut down on those errors | 11:13 |
fenn | um, no? | 11:13 |
fenn | what do you mean nano-dispenser? | 11:13 |
kanzure | he probably means single molecule dispenser | 11:14 |
CaptHindsight | lets say you could dispense ACTG a molecule at a time | 11:14 |
kanzure | and the answer is no; you can't shoot molecules that precisely, really. | 11:14 |
fenn | even polymerases make errors when they are guaranteed to have a dNTP ready | 11:14 |
CaptHindsight | lets just say that you could | 11:14 |
CaptHindsight | I'm trying to understand what causes the errors | 11:15 |
kanzure | more theoretically the origin of errors is that even with a 99.9% error rate you can't get perfect results because if you have 500 steps each at 99.9% yield rate then you end up with a bunch of contamination and wrong results | 11:16 |
CaptHindsight | yeah you need 99.99999% | 11:16 |
fenn | .c .9999999^500 | 11:17 |
yoleaux | 0.9999999⁵⁰⁰ = 0.999950001247479291757300560074181538120812312868853770936326... | 11:17 |
CaptHindsight | or some quality control at each step on each stand of oligomer | 11:17 |
kanzure | .c 0.999^50 | 11:17 |
yoleaux | 0.999⁵⁰ = 0.951205628197031349983451345476926181140585219353852365239790... | 11:17 |
fenn | yes ideally you'd have quality control at each step | 11:18 |
fenn | but i have no idea how to do that | 11:18 |
kanzure | fluorophores, probably | 11:18 |
CaptHindsight | something similar to a nanopore | 11:19 |
CaptHindsight | synthesize and sequence at the same time | 11:19 |
fenn | maybe something with mass spectrometry where you have a dna molecule on the end of a carbon nanotube and the period of its vibration is proportional to the mass of the molecule being synthesized | 11:19 |
kanzure | the ultimate solution is a form of polymerase that can be electronically controlled | 11:19 |
fenn | kanzure that doesn't close the loop | 11:20 |
fenn | you'd still get errors | 11:20 |
kanzure | i have previously proposed a solution where you use different enzymes introduced into a ribosome reaction, but this is only for protein synthesis and not for dna synthesis sadly.... https://groups.google.com/d/msg/enzymaticsynthesis/3YEEv0OULo0/zJZPETWDbMIJ | 11:20 |
kanzure | fenn: polymerase error rate is like 0.0001% | 11:20 |
fenn | with spell checking | 11:21 |
kanzure | hmm i don't remember | 11:21 |
fenn | or maybe not.. you need super duper error correction to copy a genome though | 11:21 |
CaptHindsight | my monomer just arrived a week late, now I must formulate! | 11:21 |
fenn | good luck formulating sir | 11:22 |
CaptHindsight | lets get back to this soon | 11:22 |
kanzure | was there anything decided | 11:22 |
CaptHindsight | still learning | 11:22 |
fenn | i decided i don't know squat about dna synthesis | 11:23 |
CaptHindsight | a inkjet synthesizer is trivial | 11:23 |
CaptHindsight | lets see what we can come up with | 11:23 |
CaptHindsight | I see a lot of crossover from my other work | 11:24 |
fenn | do "single molecule sequencers" really detect the light from a single molecule? | 11:24 |
CaptHindsight | but like I said I have some catching up to do on the bio side | 11:25 |
CaptHindsight | you can | 11:25 |
CaptHindsight | it's amazing how sensitive detectors have become | 11:25 |
CaptHindsight | single photons | 11:26 |
CaptHindsight | 20um diameter avalanche photodiode | 11:27 |
fenn | but dont you have to repeat the read 500 times to get a good signal to noise ratio? | 11:27 |
CaptHindsight | depends on the environment | 11:28 |
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 245 seconds] | 11:28 | |
CaptHindsight | what wavelength do they emit at? | 11:28 |
CaptHindsight | not exactly a good fit for this app but some specs are there https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=5255 | 11:30 |
fenn | a variety of visible wavelengths i think | 11:31 |
CaptHindsight | http://www.nature.com/nnano/journal/v8/n10/full/nnano.2013.150.html | 11:33 |
fenn | .title | 11:33 |
yoleaux | Quantum interference in plasmonic circuits : Nature Nanotechnology : Nature Publishing Group | 11:33 |
CaptHindsight | we build the nano machine shop first | 11:35 |
CaptHindsight | people ask me what are the applications for micro and nano fabrication, this is a good example | 11:36 |
-!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap | 11:36 | |
CaptHindsight | I've also been thinking about grafting semiconductors onto cellular structures | 11:37 |
fenn | looks like a cheaper way to make a quantum computer at least | 11:37 |
CaptHindsight | don't stem cells have the ability to modify their DNA? | 11:38 |
CaptHindsight | what we could get cells to just start making custom DNA? | 11:39 |
fenn | no, stem cells don't have the ability to modify their dna | 11:39 |
CaptHindsight | heh, grammar block there | 11:39 |
-!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 276 seconds] | 11:39 | |
CaptHindsight | how do they become other cells? | 11:39 |
kanzure | there's probably a way to make neurons learn things (the horror!) and influence dna methylation, but i doubt it can be used to insert novel sequences | 11:39 |
kanzure | they become other cells through factors like oct8 (a protein) that causes them to activate some other cascade of gene expression | 11:40 |
fenn | stem cells have different patterns of chromatin binding to their dna than differentiated cells | 11:40 |
CaptHindsight | are you trying to generate sequences too novel for them? | 11:41 |
fenn | dna is wrapped around chromatin tightly which prevents other molecules from accessing it and being able to transcribe it into mRNA | 11:41 |
fenn | .title http://www.nature.com/nrm/journal/v7/n7/fig_tab/nrm1938_F2.html | 11:42 |
yoleaux | Figure 2 : Chromatin in pluripotent embryonic stem cells and differentiation : Nature Reviews Molecular Cell Biology | 11:42 |
fenn | this diagram shows what i am trying to say | 11:42 |
fenn | the cell on the left is a stem cell | 11:43 |
CaptHindsight | yeah, so is the goal to make DNA that is too far outside the box for a cell to make? | 11:44 |
fenn | cells only copy DNA that already exists | 11:45 |
fenn | nothing nature that i am aware of actively makes new DNA | 11:45 |
fenn | some weird things like immune cells will do random permutations of short sequences, but that's about it | 11:45 |
fenn | .wik terminal dideoxy transferase | 11:46 |
yoleaux | "TA cloning is a subcloning technique that avoids the use of restriction enzymes and is easier and quicker than traditional subcloning. The technique relies on the ability of adenine (A) and thymine (T) (complementary basepairs) on different DNA fragments to hybridize and, in the presence of ligase, become ligated together." — https://en.wikipedia.org/wiki/TA_cloning | 11:46 |
fenn | gah | 11:46 |
fenn | .wik tdt | 11:46 |
yoleaux | "Disambiguation: TDT" — https://en.wikipedia.org/wiki/TDT | 11:46 |
fenn | .wik Terminal_deoxynucleotidyl_transferase | 11:46 |
yoleaux | "Terminal deoxynucleotidyl transferase (TdT), also known as DNA nucleotidylexotransferase (DNTT) or terminal transferase, is a specialized DNA polymerase expressed in immature, pre-B, pre-T lymphoid cells, and acute lymphoblastic leukemia/lymphoma cells." — https://en.wikipedia.org/wiki/Terminal_deoxynucleotidyl_transferase | 11:46 |
CaptHindsight | ah so all the DNA is already there in stem cells | 11:47 |
fenn | "TdT catalyses the addition of nucleotides to the 3' terminus of a DNA molecule. Unlike most DNA polymerases, it does not require a template. | 11:47 |
CaptHindsight | they just expose or make available different sections | 11:47 |
CaptHindsight | bbl | 11:48 |
fenn | right. actually they hide sections, but the overall effect is that different behaviors/metabolisms are expressed | 11:48 |
CaptHindsight | thanks for the input! | 11:49 |
CaptHindsight | I follow | 11:49 |
-!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 11:57 | |
kanzure | someone should compile those notes into some machine design goals | 12:06 |
eleitl | Interesting search engine https://ebooks.wtf/ | 12:16 |
-!- justanotheruser is now known as justanotherusr | 12:17 | |
eleitl | By the way, it may be that the days of LibGen are counted. Wiley is hunting them. | 12:17 |
kanzure | yeah i'm trying to purchase some plane tickets | 12:17 |
kanzure | i haven't been able to figure out where they are located though | 12:17 |
kanzure | they claim to have moved out of russia but who the hell knows | 12:17 |
eleitl | Kasachstan, imo. | 12:17 |
kanzure | no, that's the other person | 12:17 |
eleitl | OIC. | 12:17 |
eleitl | Actually, all you need is a bunch of hard drives and quality time with a good connection. | 12:18 |
kanzure | eleitl: lg situation is under control (sorta) | 12:18 |
kanzure | no, their upload rate is 1 kbps | 12:18 |
kanzure | thus the plane tickets | 12:18 |
-!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-wrjwafncfgtvwewb] has quit [Quit: Connection closed for inactivity] | 12:18 | |
eleitl | Good luck, kanzure. You're doing $deity's work. | 12:19 |
-!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 264 seconds] | 12:22 | |
fenn | wouldn't it be easier to send hard drives in the mail | 12:22 |
eudoxia | kanzure: you're going on a trip to back libgen up and take the backup to a place with better bandwidth? | 12:23 |
eleitl | He's going to Equador. | 12:25 |
eleitl | Friendly llamas there. | 12:25 |
eudoxia | cool kanz pet a llama for me | 12:25 |
ParahSailin | is libgen going to have to retreat to an onion? | 12:26 |
chris_99 | it's blocked 'ere now apparently | 12:26 |
ParahSailin | http://gen.lib.rus.ec/ works | 12:26 |
chris_99 | oh so it does :) | 12:27 |
fenn | wtf is rus.ec | 12:27 |
ParahSailin | ecuador i bet | 12:27 |
fenn | cloudflare host registered to a russian sounding name in portugal via ecuador namespace | 12:28 |
fenn | i think it's funny | 12:29 |
eudoxia | it's a sign we finally live in the cyberpunk future | 12:31 |
fenn | oh wait, the tech email is in poland, but lives on portugal street | 12:32 |
eleitl | Confusticate and bebother these dwarves. | 12:33 |
fenn | wanted for international criminal librarianship | 12:34 |
justanotherusr | kanzure: So I saw this guy who was pretending to be you on reddit. Not everyone here knows that you wouldn't dare go on reddit so look out for people mistaking your imposter for you. | 12:34 |
-!- Beatzebub [~beatzebub@d75-155-236-222.bchsia.telus.net] has joined ##hplusroadmap | 12:36 | |
cluckj | imitation is the sincerest form of flattery | 12:43 |
jrayhawk | haha | 12:43 |
kanzure | justanotherusr: i realized that i couldn't possibly hate myself any more than i already do, so what's the worst that could happen by increasingly becoming even more of a hypocrite? | 12:43 |
cluckj | I also find that going on reddit is an exercise in self-loathing | 12:44 |
* eleitl is in /top | 12:44 | |
eleitl | Not /centuryclub though. | 12:44 |
kanzure | i drowned in an ocean of blood http://www.reddit.com/r/Bitcoin/comments/3aeow8/blockstream_has_a_very_serious_conflict_of/csc05yc | 12:44 |
eleitl | I don't hate myself sufficiently for that. | 12:44 |
kanzure | and i'm finally back to my old ways of "here's a million bookmarks" http://www.reddit.com/r/Bitcoin/comments/3aeow8/blockstream_has_a_very_serious_conflict_of/csbxxdu | 12:46 |
kanzure | fenn: to send hard drives in the mail you would first have to know where to send them, and also convince someone to manually handle hard drives | 12:48 |
kanzure | my plan was to go to their network hub and just place my hand on the oc-148 fiber and try to sense where their hq is located | 12:48 |
justanotherusr | "That is not true. We hear this every day. "If all people in the world will use the blockchain, then it will become a failure!"" | 12:50 |
justanotherusr | what, of course its a failure, the only way to support that is by making bitcoin paypal | 12:50 |
kanzure | right... | 12:51 |
kanzure | but also, he was sort of saying "100 MB blockchain working would mean it's working!!!!" and uh.. yes that's true... | 12:52 |
-!- erasmus [~esb@unaffiliated/erasmus] has quit [Read error: Connection reset by peer] | 12:53 | |
-!- Viper168_ [~Viper@unaffiliated/viper168] has joined ##hplusroadmap | 12:53 | |
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 250 seconds] | 12:54 | |
justanotherusr | "The point is, people aren't stupid. Just like you, they won't send their money if it isn't safe. And the blocks will get smaller. But if they do, it's because they see Bitcoin as safe." | 12:54 |
justanotherusr | thats actually an interesting argument, if bitcoin is made useless by centralization will the central authority fix it so bitcoin can be profitable for them again? | 12:55 |
justanotherusr | seems like a silly risk to take, but I wonder how that would play out. I can guess abandonment and lack of faith the most likely way though | 12:55 |
kanzure | er, profitability was probably not the centralization goal in the first place :-) | 12:55 |
justanotherusr | ah right | 12:55 |
kanzure | anyway, if people want centralized bitcoin then they should make better proposals for that... there are many ways to centralize bitcoin, and so far all of the proposals for doing so are quite bad proposals. very inefficient. | 12:55 |
-!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 12:57 | |
-!- Viper168_ is now known as Viper168 | 12:58 | |
fenn | "The high rate of usage implies that Bitcoin is safe and decentralized, because people using it have decided that it's safe for them. It implies BTC costs $100k (high demand) and the mining is hugely profitable and therefore decentralized." there are so many misconceptions and wrong statements in this one sentence | 12:58 |
eleitl | You still need a single person or two making decisions over which code clients run. | 12:59 |
eleitl | Somebody dig up Satoshi. We have a job for him. | 12:59 |
fenn | thaw him out | 12:59 |
fenn | get working on that ALS cure | 12:59 |
kanzure | i thought robots were the cure? | 12:59 |
eleitl | Don't think it's Hal Finney. | 13:00 |
eudoxia | i confess it was me all along | 13:00 |
eleitl | Hah! | 13:00 |
eudoxia | the right block size is *spins wheel* | 13:00 |
eudoxia | 12.6 megabytes | 13:00 |
fenn | 8888888888888 bibibits | 13:01 |
fenn | has anyone proposed a quantitative measure of "decentralization" yet? besides just "number of nodes" | 13:02 |
eudoxia | full client users/thin client users | 13:02 |
fenn | why are people saying the internet is so much faster in 2015 than it was in 2010? it doesn't seem any faster to me | 13:03 |
justanotherusr | fenn: this is coming from the same person who has claimed that mining centralization can't happen because the large miners will "inspire" lots of competition to grow as big as them | 13:03 |
kanzure | lots of redditors have been saying "we'll be saved by nielsen's law and moore's law" | 13:05 |
* fenn looks at http://www.nngroup.com/articles/law-of-bandwidth/ | 13:05 | |
fenn | "bandwidth increases by 50% each year" and "blocksize doubles every 2 years" doesn't add up | 13:06 |
fenn | anyway it seems pretty arbitrary | 13:07 |
kanzure | yes, it's arbitrary and broken | 13:08 |
kanzure | if anything the max block size should be reduced | 13:09 |
eudoxia | question: the reason mining stops in 2104 is because you run out of zeroes to pad, no? | 13:09 |
fenn | wait, is this "nielsen's law" just "jacob neilsen's personal internet speed"? | 13:09 |
kanzure | eudoxia: was arbitrary decision by satoshi nakamoto | 13:10 |
kanzure | fenn: that would be hilarious | 13:10 |
eudoxia | kanzure: thanks | 13:10 |
kanzure | fenn: many of their bandwidth arguments are coming from gavinandresen blog posts | 13:12 |
kanzure | like http://gavinandresen.ninja/does-more-transactions-necessarily-mean-more-centralized | 13:13 |
fenn | with current mining fees at ~$2.50/block and the mining subsidy at ~$6k/block you'd expect the miners to say wait a minute, we need to raise fees before the subsidy runs out | 13:17 |
fenn | but for some reason they think more transactions would let them make more money | 13:18 |
fenn | are people really that dumb? | 13:18 |
fenn | even if you manage to fill up the entire 20MB block with paying transactions that's only $90/block in fees | 13:21 |
cluckj | someone has a secret plan to make money off of the change | 13:24 |
fenn | hmmm bc.i claims transaction fees were 26.5BTC which doesn't make sense if the minimum fee is 10 uBTC and there are a maximum of 1800 transactions per block | 13:26 |
fenn | oh that's per day | 13:27 |
kanzure | fenn: some people are concerned that there hasn't been enough time to get enough users, e.g. such that if you continue to allow zero-fee transactions for a while longer that when people do start paying fees that there will be far more transactoins and fees available, or something.... | 13:28 |
kanzure | fenn: and then there's an entirely different set of people that seem to think that fees should never be mandatory, or that the network should self-hard-fork in the event that only transactions with fees get into the blockchain or something | 13:29 |
kanzure | (thankfully nobody has made an actual "official" proposal about that self-hard-forking concept, heh) | 13:29 |
fenn | they are perfectly free to include zero-fee transactions in their blocks | 13:29 |
fenn | mine away | 13:29 |
kanzure | also, there's some chance that average transaction fee will increase to some absurdly high level if there's lots of transaction competition, such that the minimum transaction fee likely to get int othe blockchain is greater than the value of anyone's savings on the blockchain | 13:31 |
kanzure | however, for that to be true there would have to be a massive increase in transaction volume, a massive increase in average or median transaction fees, full blocks, transaction competition around fees, and also the exchange rate of bitcoin would have to go down(?) because otherwise if it went up then your transaction fees would actually be lower not greater | 13:32 |
fenn | .title https://gist.github.com/gavinandresen/5044482 | 13:33 |
yoleaux | Back-of-the-envelope calculations for marginal cost of transactions | 13:33 |
fenn | "Andresen calculated that a miner should demand a fee of at least 0.0008 BTC, which at the time was worth $0.41, to include an average-sized transaction." | 13:34 |
fenn | oh this was based on the price of the block subsidy, not the cost of cpu time to validate blocks | 13:39 |
fenn | i guess it's a really complicated question | 13:40 |
-!- erasmus [~esb@unaffiliated/erasmus] has joined ##hplusroadmap | 13:50 | |
-!- Acty [uid89656@gateway/web/irccloud.com/x-rnahoprkqiffdtdr] has quit [Quit: Connection closed for inactivity] | 13:51 | |
-!- eudoxia [~eudoxia@r167-57-119-114.dialup.adsl.anteldata.net.uy] has quit [Quit: Leaving] | 13:52 | |
-!- Madplatypus [uid19957@gateway/web/irccloud.com/x-fdoiarrahrlfbroc] has joined ##hplusroadmap | 13:52 | |
kanzure | yes it's a complex question | 13:59 |
-!- juri_ [~juri@pool-72-83-63-219.washdc.east.verizon.net] has joined ##hplusroadmap | 14:08 | |
-!- juri_ [~juri@pool-72-83-63-219.washdc.east.verizon.net] has quit [Read error: Connection reset by peer] | 14:11 | |
erasmus | truly. | 14:12 |
-!- juri_ [~juri@pool-72-83-63-219.washdc.east.verizon.net] has joined ##hplusroadmap | 14:16 | |
-!- augur [~augur@c-73-46-94-9.hsd1.fl.comcast.net] has quit [Quit: Leaving...] | 14:27 | |
-!- GAit [~lnahum@2-230-161-158.ip202.fastwebnet.it] has quit [Read error: Connection reset by peer] | 14:31 | |
-!- GAit [~lnahum@2-230-161-158.ip202.fastwebnet.it] has joined ##hplusroadmap | 14:31 | |
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has quit [Ping timeout: 265 seconds] | 14:37 | |
-!- Zinglon [~Zinglon@ip565f6f48.direct-adsl.nl] has joined ##hplusroadmap | 14:39 | |
-!- erasmus [~esb@unaffiliated/erasmus] has quit [Read error: Connection reset by peer] | 14:41 | |
-!- juri_ [~juri@pool-72-83-63-219.washdc.east.verizon.net] has quit [Ping timeout: 256 seconds] | 14:45 | |
-!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 256 seconds] | 14:46 | |
-!- juri_ [~juri@vpn166.sdf.org] has joined ##hplusroadmap | 14:47 | |
-!- Zinglon [~Zinglon@ip565f6f48.direct-adsl.nl] has quit [Ping timeout: 264 seconds] | 14:49 | |
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-vfpkfvcxqmwovowd] has joined ##hplusroadmap | 15:02 | |
-!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 15:05 | |
-!- delinquentme [~delinquen@74.61.157.78] has quit [Client Quit] | 15:06 | |
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Remote host closed the connection] | 15:13 | |
-!- erasmus [~esb@unaffiliated/erasmus] has joined ##hplusroadmap | 15:24 | |
archels | ,--8<- | 15:36 |
archels | |20/00:17:52 < RandIter> Yeah see if you can hook up a blood line with a poor chinese kid for a few hours to get his stem cells into yours | 15:36 |
archels | |20/00:18:10 < AdrianG> i am planning on doing that. | 15:36 |
archels | |20/00:26:56 < RandIter> I wonder if any rich people today do it already on a scheduled basis, considering it is practically guaranteed to produce anti-aging results. An adequate financial compensation along with good nutrition provided to the poor kid should more than make up for the strain. | 15:36 |
archels | `-->8- | 15:36 |
archels | because why bother vitrifying your umbilical cord when you can just have a living, walking incubator? :/ | 15:37 |
kanzure | nbote that at least three of those users are already muted in here | 15:39 |
kanzure | for unrelated reasons | 15:39 |
eleitl | Blood tranfusion rejuvenation? | 15:39 |
kanzure | eleitl: surely you remember adriang.. he's the one from extropy-chat that is afraid that everyone is going to upload him and then torture him forever. | 15:40 |
archels | hey eleitl | 15:40 |
eleitl | There are several nutbags on exi-chat. | 15:40 |
archels | yeah it came up some time ago in the context of storing a newborn's umbilical cord long-term | 15:40 |
eleitl | Hey archels. | 15:40 |
eleitl | Exi-chat has been dead a long time, so who cares. | 15:40 |
archels | oh, incidentally, watched a pretty good German sci-fi movie earlier: Cargo | 15:41 |
eleitl | Extropy Institute has been glorious, in the early days. But, records are not public. | 15:41 |
archels | (it involves some uploading) | 15:41 |
eleitl | Interesting reviewer distribution on Amazon. | 15:42 |
archels | wasn't there a recent effort here to resurrect old Extropy magazine issues? | 15:42 |
kanzure | shhh don't let anyone know | 15:42 |
eleitl | actually, the recent Alcor magazines are surprisingly good | 15:43 |
archels | oh, wouldn't that be a good thing? | 15:43 |
archels | some of the stuff there was pure gold as well | 15:43 |
kanzure | archels: no; they were "private" and there were various threats about publicizing | 15:43 |
eleitl | I have not seen them, since I was not willing to be a paying member. But people who were tell me it was great. | 15:43 |
archels | well, I don't want to start a whole discussion about "reasonable expectation of privacy", but, be serious | 15:44 |
kanzure | they are approximately the same quality as the email content from the time | 15:44 |
eleitl | For the time, right now you would be probably party-pooping all over that. | 15:44 |
eleitl | Do you have access to these, kanzure? | 15:44 |
kanzure | archels: i mean specifically that many people have made stern comments to me in the distant past about the content not being public | 15:44 |
kanzure | eleitl: http://fennetic.net/irc/extropy/ | 15:44 |
archels | :( | 15:45 |
kanzure | eleitl: what's your theory about why email content quality has dropped everywhere so dramatically? | 15:45 |
eleitl | But this is not the old, closed ExI list. | 15:45 |
kanzure | ah, for old archives all i have is http://diyhpl.us/~bryan/irc/extropians/ | 15:46 |
eleitl | Email is a dying medium, and the new kids just don't care. | 15:46 |
kanzure | that doesn't explain why the old people aren't writing email (not that i want them to write email--- i want them to take neuroscience classes, and stuff) | 15:46 |
eleitl | Assuming, there are any new kids at all. | 15:46 |
eleitl | Old people are now mostly offline, or just exchange catty emails in clandestine mailservs. | 15:47 |
eleitl | See silent-tristero. | 15:47 |
kanzure | perhaps a better question would be why there seemed to be a high signal-noise ratio | 15:47 |
eleitl | Echo chamber. | 15:47 |
eleitl | High bareer to entry. If you were online in the 1980s, there was a very good reason. | 15:47 |
kanzure | did they have to cull lots of people out of the group...? was there someone editing everyone's emails to make them look better? | 15:47 |
kanzure | hmm. | 15:48 |
eleitl | s-t did have some mod action, but not in several years. | 15:48 |
eleitl | Manual digests was all there was to it. | 15:48 |
kanzure | ah, did everything start as digest mode? | 15:48 |
kanzure | or was it always transactional..? | 15:48 |
eleitl | I was always feeling like an impostor on s-t. | 15:49 |
kanzure | perhaps long delays are beneficial for convincing people to write worthy content | 15:49 |
eleitl | They had interactions, and digest was extra. I might misremember. | 15:49 |
kanzure | ok | 15:49 |
kanzure | i was thinking something like, | 15:49 |
kanzure | "if you have something worth saying, then it is worth saying it slowly" | 15:49 |
eleitl | In any case, email is dead. Usenet probably, too. | 15:49 |
kanzure | (and also "it is worth having someone else yell at you until you fix your line width") | 15:49 |
eleitl | If you think about Victorian letters, Jesus Christ. | 15:50 |
eleitl | Email has nothing on these. | 15:50 |
kanzure | "Selections from the Silent Tristero mailing list" http://www.midnightbeach.com/jon/st/ | 15:50 |
kanzure | the other problem i have observed is that it is extremely hard to convince everyone to show up at the same place | 15:51 |
eleitl | FoRK the same. Mostly dead. | 15:51 |
eleitl | Asynchronous global communications is really neat. | 15:51 |
kanzure | on the other hand, i hate the "let's just write lots of emails and pretend we're transhumanists" model of doing things :-) | 15:52 |
eleitl | There always was action. It was just not mirrored in emails. | 15:52 |
eleitl | Mutuall exclusive, in fact. | 15:52 |
eleitl | But, look up Steven B. Harris on Usenet. | 15:52 |
kanzure | i see some cryonet posts..? | 15:53 |
eleitl | CryoNet was at times good, but the laissez-faire approach has killed it | 15:53 |
eleitl | CryoNet too, for obvious reasons | 15:54 |
eleitl | But a lot is there in Usenet in 1990s. Maybe it's not retained. | 15:54 |
eleitl | I'm not going to say more on a public channel, but he's one of the good guys. | 15:55 |
eleitl | Interesting dynamics, when you get plenty of very smart people in one room. | 15:55 |
kanzure | did you happen to read today's backlog regarding oligonucleotide synthesis | 15:55 |
eleitl | Sorry, no. Too drunk now. Will do tomorrow. Is beyond my expertise in any case. | 15:56 |
kanzure | ( http://gnusha.org/logs/2015-06-19.log ) | 15:56 |
kanzure | bah | 15:56 |
eleitl | I've only seen macroscale stuff. | 15:56 |
eleitl | By the way, our shop is still likely to go broke next year or year after. | 15:57 |
eleitl | Politics is the bane of any larger enterprise. | 15:57 |
eleitl | It doesn't matter what you do, it's all entirely capricious. | 15:58 |
kanzure | have i shown you http://fennetic.net/irc/human_chimpanzee_brain_differences.png | 15:59 |
eleitl | No. Do you have the references somewhere? | 16:00 |
archels | eleitl: didn't realise you guys were still running | 16:04 |
archels | what've you been up to? | 16:04 |
-!- vivi [~vivi@inhu.me] has quit [Read error: Connection reset by peer] | 16:04 | |
-!- desayuno [~brekfast@unaffiliated/brekfast] has joined ##hplusroadmap | 16:04 | |
eleitl | We're running? | 16:04 |
eleitl | Silently, I guess. | 16:04 |
kanzure | diagram is from http://diyhpl.us/~bryan/papers2/neuro/Transcriptomic%20insights%20into%20human%20brain%20evolution:%20acceleration,%20neutrality,%20heterochrony.pdf | 16:04 |
eleitl | I've had good success in failures. | 16:04 |
eleitl | Our cryonics project tanked, our nonprofit got taken over by vampires (ToV, that is) | 16:04 |
-!- vivi [~vivi@inhu.me] has joined ##hplusroadmap | 16:04 | |
eleitl | Greate fun, all in all. | 16:05 |
-!- vivi is now known as Guest8333 | 16:05 | |
eleitl | There might be something coming up in Switzerland, but it's too early to tell. | 16:05 |
kanzure | at this point i'm like 100% convinced that you can't find better friends than in here | 16:06 |
eleitl | Not holding my breath. | 16:06 |
kanzure | or better partners for ventures | 16:06 |
eleitl | Physical space is thin. And you have less energy as you age. | 16:06 |
eleitl | I think the slow burn approach a la US in 1970s no longer works. | 16:07 |
eleitl | Especially, in Switzerland.# | 16:07 |
eleitl | So you have to bank on the brand. | 16:07 |
kanzure | so.. strike like lightning? | 16:07 |
eleitl | One tiem effort, once burned, that's it. | 16:07 |
eleitl | My spelling has gone all to shit. | 16:08 |
kanzure | better pick the battles well then | 16:08 |
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 246 seconds] | 16:08 | |
eleitl | I'm going to write a few emails, and then write it all off, in case it doesn't work. | 16:08 |
eleitl | Somebody needs to make a full brain dump of Mike Darwin, while he's still on the premises. | 16:09 |
kanzure | i'm still working on aubrey, although i'm in no rush at the moment | 16:09 |
kanzure | yeah i can volunteer for that, although i will need the assistance of at least one additional interrogator (i can type everything but i have reduced question-asking ability) | 16:09 |
eleitl | Aubrey has a very good story, but it is a lot of money down the drain. | 16:09 |
kanzure | he had some good concerns, they were fair | 16:09 |
eleitl | In order to corner Mike you need a quality time of weeks. | 16:09 |
kanzure | to corner him, or do you mean to interrogate? | 16:10 |
eleitl | It should be more of a continuous project, and that takes a lot of effort. | 16:10 |
kanzure | do you think you could convince him to spend 2 hours/week on the phone for this? | 16:10 |
eleitl | Total media recording and transcripts. And extremely well structured interrogation. | 16:11 |
eleitl | No, that won't do. | 16:11 |
kanzure | hmm. | 16:11 |
kanzure | describe | 16:11 |
eleitl | He tends to go off, and you need massive force to get him back on track. | 16:11 |
-!- augur [~augur@c-73-46-94-9.hsd1.fl.comcast.net] has joined ##hplusroadmap | 16:12 | |
eleitl | Which is very tiring. I can't do it, if I have to record. | 16:12 |
kanzure | i am good at the "remembering why we are talking about a certain thing, and the thing before this and up the stack, so as to convince others to get back to the original topic". this is one of my specialties. | 16:12 |
eleitl | In general, we have a number of people who know things, and won't talk, and about one person who knows things and is willing to talk. | 16:12 |
eleitl | We should actually do it as a team, but do it in person. And that is prohibitive. | 16:13 |
kanzure | eh i can pay for it | 16:13 |
kanzure | might work once they axe your division? | 16:13 |
eleitl | I have no idea how long Mike is still around. | 16:14 |
eleitl | I've had big fears when his place burned down | 16:14 |
eleitl | A lot of irreplaceable information was lost | 16:14 |
kanzure | i wonder how much docs he lost | 16:14 |
kanzure | shit | 16:14 |
eleitl | Some 80% to 90% of it. | 16:14 |
eleitl | But, he is still available, and he is willing to talk. | 16:15 |
eleitl | Nobody else does. | 16:15 |
eleitl | Aschwin is too busy, and imo too cagey. | 16:15 |
eleitl | 21CM/CCR folks are busy, and have IP issues. | 16:15 |
eleitl | We're clearly SOL here. | 16:15 |
archels | all this talk about fast burning, one shot—what about persistence? Might not get your KickStarter funded, but more likely to actually fundamentally get you places | 16:16 |
eleitl | The one-shot is about burning the Switzerland brand. | 16:16 |
eleitl | Danila is also poaching there at the moment, though I don't know anything about it. | 16:17 |
eleitl | The people who are involved are competent and well-meaning, but they're not aware about the one-time shot brand. | 16:17 |
eleitl | See talking to press, and shutdown. | 16:17 |
eleitl | Both no-no. | 16:18 |
kanzure | what is danila medvedev poaching for? | 16:18 |
eleitl | Krauts have a bad history of talking to press. They seem to think that no publicicity can be bad publicity. | 16:18 |
archels | nod, it would need to be a pretty tight group of individuals | 16:18 |
eleitl | Danila is trying to set up something in Switzerland for wealthy clients. I have no idea what, and whether it's still ongoing. | 16:19 |
eleitl | The problem with Switzerland and wealthy Russians is that it just doesn't mix. | 16:19 |
eleitl | archels: exactly right | 16:20 |
eleitl | Very small group, figuring out everything in advance before going public (for very small values of public) | 16:20 |
eleitl | I don't feel this is a given here. So high probability of failure. | 16:21 |
eleitl | The first rule of assistend suicide club... | 16:21 |
eleitl | The natives *hate* it. See Sir Terry. | 16:22 |
eleitl | And unlike elsewhere, the native can change the laws easily enough. | 16:22 |
eleitl | the natives | 16:22 |
eleitl | So keep it quiet, talk FOAF. | 16:23 |
kanzure | i'm also pretty sure that money would resolve a bunch of your pessimism | 16:23 |
kanzure | easy to be pessimistic when there's very few resources :-) | 16:24 |
eleitl | Money by itself is useless. Cryonics has a great history of burning multiple MUSD with literally nothing to show for it. | 16:24 |
eleitl | Which is why I probably sound like a nazi when talking about control. | 16:25 |
kanzure | control is good.. for certain project types. cryonics is probably one of those. | 16:25 |
eleitl | The problem is getting people with clue with money, and that has been a very rare occurance so far. | 16:25 |
kanzure | if you had no clue, what stpes could have been taken to avoid burning millions on cryofail? | 16:26 |
eleitl | All the successes involve self-bootstrap. All the people with money coming in were no effect. | 16:26 |
eleitl | The problem with non-experts is that you don't know who the experts are. Unless you're an expert. It's a very harsh filter. | 16:27 |
eleitl | You don't know who to listen to. And all the people with money have own ideas. Very strong ideas. | 16:27 |
eleitl | Which happen to be wrong, but remember the Golden Rule. | 16:27 |
eleitl | The man with the gold makes the rules. | 16:27 |
eleitl | I've had 3 data points personally which were all negative. | 16:28 |
eleitl | Mike has plenty more data points. I believe him there. | 16:28 |
eleitl | Even so, there's considerable conflict within the experts. | 16:29 |
kanzure | yeah, i will figure out how to organize a proper interrogation session | 16:29 |
eleitl | And I have no fucking clue who is right. | 16:29 |
-!- erasmus [~esb@unaffiliated/erasmus] has quit [Quit: fencing] | 16:29 | |
eleitl | Because I'm not an expert. | 16:29 |
kanzure | expertise is overrated; being correct is more important. | 16:29 |
eleitl | It seems the only way to figure out is evidence-based. You have to run the lab and do the experiments. | 16:30 |
eleitl | No validation, no voice. | 16:30 |
eleitl | Which makes a success very expensive, especially in a highly regulated environment. | 16:30 |
-!- eudoxia [~eudoxia@r186-50-158-185.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap | 16:31 | |
eleitl | E.g. need for animal experiments, to validate reversibility. | 16:32 |
eleitl | Where do you think you can murder dogs for fun an profit? Asia, probably. | 16:32 |
kanzure | i wouldn't call it murder, that's a misrepresentation | 16:33 |
eleitl | I was just being dramatic, for argumentative reasons. | 16:33 |
kanzure | i'm also surprised that you would go straight to dog ? any reason? | 16:33 |
eleitl | Dogs are docile and have human-assessable personalities. | 16:34 |
eleitl | They're also larger than rabbits. | 16:34 |
kanzure | you mean like.... because you see a benefit regarding... personality assay? | 16:34 |
eleitl | Because if you want to assess survivors, you need to know them before, and have a relationship with them. | 16:34 |
kanzure | so i have been anticipating <10% survival rate initially, per just *one* organ type | 16:35 |
eleitl | Pigs are not trusting, and very hard to work with if uncooperative. Sheep are just too fucking stupid. | 16:35 |
kanzure | but i suppose it's possible there might be >1% whole animal survival initially ? | 16:35 |
eleitl | Standard run with dogs is just arrest, then resuscitation after delay. | 16:35 |
kanzure | including vitrification? | 16:35 |
eleitl | It is surprisingly hard, if you start from scratch. | 16:35 |
eleitl | No, just arrest, then resuscitation. Normothermic ischemia, at first. | 16:36 |
kanzure | yes i suppose that's important too :-) | 16:36 |
eleitl | Then you progress to premedication, hypothermia, the whole works. | 16:36 |
kanzure | yep | 16:36 |
eleitl | When I left we were still killing dogs. | 16:36 |
kanzure | also: there is already some % cell survival for vitrification and thaw. | 16:37 |
eleitl | Medicine is hard, let's go shopping. | 16:37 |
kanzure | positive % cell survival | 16:37 |
eleitl | Organ viabilit or tissue viability is different from resuscitated animal. | 16:37 |
kanzure | true, but it's necessary | 16:37 |
eleitl | You have to look at reversible stages to assess what you're doing, or your're flying blind. | 16:37 |
eleitl | You can't get into orbit if you're flying blind. | 16:38 |
eleitl | It is very hard to come back where you used to be if you're starting from scratch. | 16:38 |
eleitl | Even with the same team, in the same site. | 16:38 |
eleitl | This is how medicine is different from software development. | 16:39 |
eleitl | If anything, it's closer to greefielding a cutting-edge gab. | 16:39 |
eleitl | fab | 16:39 |
eleitl | You document everything and you replicate everything documented, and you've still got 100% dead wafers. | 16:39 |
eleitl | Or 100% dead dogs. | 16:40 |
kanzure | yes, lots of people should be working on reducing the differences between different lab configurations, reducing the costs of lab setup, and increasing the repeatability and reliability of complex procedures | 16:40 |
eleitl | The easiest way to contain the magic is not trying to replicate the magic. | 16:40 |
eleitl | A lot of suggestions I hear is that you mothball the facility, and use hirelings for on-demand cases. | 16:40 |
eleitl | This Just Doesn't Work(tm). | 16:41 |
kanzure | that does not seem like a good strategy for containing magic | 16:41 |
kanzure | magic must be thoroughly destroyed at every opportunity | 16:41 |
eleitl | Advanced technology is based on magic, unfortunately. | 16:41 |
eleitl | And that makes it a very fickle beast. | 16:41 |
kanzure | the transhumanist approach to dealing with 100s of millions of pages of documentation is not to consider it all magic; it should be to automate as much as possible | 16:42 |
eleitl | The procedures are contained in people. You can't replicate people, at least not yet. | 16:42 |
eleitl | Your magic is not documented in the first place. | 16:42 |
eleitl | that's your problem | 16:42 |
kanzure | hm | 16:42 |
kanzure | hmm | 16:42 |
eleitl | You should work in the area for a year or two. | 16:43 |
eleitl | Then you will believe me. Or not. | 16:43 |
kanzure | no, i believe you that it is hard to get repeatability | 16:43 |
kanzure | i have read the alcor patient reports-- it's insane | 16:43 |
eleitl | Or, talk to a competent intensive care team. Tell them what it takes to greenfield a competent Intensive Care unit. | 16:43 |
kanzure | on the other hand, checklists seem to have helped many surgery teams | 16:44 |
kanzure | even just basic things on a checklist- like washing hands- seems to help | 16:44 |
eleitl | you should look at patient report team trend over time, starting with BPI tech briefs | 16:44 |
eleitl | you will see something that you won't like | 16:44 |
eleitl | Now look at CI "patient reports". | 16:44 |
eleitl | Now that's not much to report, innit. | 16:45 |
eudoxia | this backlog is quality stuff, keep it up | 16:45 |
eleitl | The problem is I know I can talk to you that way, because you're likely getting it, and there's nothing at stake here. | 16:46 |
eleitl | When you have a working, living process already in place you no longer have a chance with rants like these. | 16:46 |
eleitl | Medicine has a procedure and institutions to deal with these, cryonics has not. | 16:47 |
kanzure | why are rants not allowed? | 16:47 |
eleitl | Rants are allowed, but dismissed. | 16:47 |
eleitl | >/dev/null | 16:47 |
dingo | hey kanzure: my salt of deploying openstack cluster as a graphviz, if you're interested https://teamcity-master.pexpect.org/tmp/ | 16:47 |
dingo | i found it quite useful | 16:47 |
dingo | your words about "I never found graphviz output to be useful" stuck with me for a while, maybe helped keep me from doing it :) | 16:48 |
dingo | since deep down i knew it would be true, but it wasn't | 16:48 |
dingo | it became a sort of "static analysis", and i discovered 4 classes of bugs which were resolved | 16:48 |
kanzure | are these just application dependencies? | 16:49 |
catern | graphviz | 16:49 |
dingo | and now i see some of these "orphan" nodes should be a require of a dependent node, and we're just "lucky" that the run-order balances to ensure it | 16:49 |
catern | sooooo beauuuuuuutiful | 16:49 |
dingo | its salt states | 16:49 |
catern | kanzure: you don't find graphviz to be useful?! | 16:49 |
dingo | 'label': ('name: {self.sls_name} [{self.run_order}]\n' | 16:49 |
dingo | 'func: {self.salt_module}.{self.salt_function}\n' | 16:49 |
dingo | 'path: {self.sls_path}' | 16:49 |
kanzure | catern: i have never seen a useful graphviz graph | 16:49 |
dingo | ouch bad paste, my bad | 16:49 |
dingo | but that's the contents of the boxes | 16:49 |
kanzure | dingo: i think you probably already knew that your salt states were all over the place | 16:49 |
catern | kanzure: I mean, I use it to produce graphs for graph theory stuff | 16:49 |
kanzure | catern: yep, i used to work in a graph theory lab | 16:50 |
dingo | i plan to use networkx, thanks for the tip | 16:50 |
dingo | i needed to resolve the code first, and i knew how graphviz worked, i didn't want to wrench the process | 16:50 |
kanzure | i think that this data would be better represented by a list of "path" values sorted that way. e.g. yaml file of {path: {name: name, func: func}} or something | 16:51 |
dingo | i'd like to make the "run order" matter more than it does, the number inside [] | 16:51 |
dingo | in its rendered output, to order as a time-series, as that's what it is | 16:51 |
dingo | i have 'duration' in millesconds float to use | 16:51 |
dingo | it would be nice to place start-time with a duration width horizontally | 16:52 |
kanzure | dingo: https://en.wikipedia.org/wiki/Systemantics | 16:52 |
eleitl | Another analogy, your team are practiced musicians, who're tuned in to each other. | 16:55 |
eleitl | You need time to get even a practiced team to work with each other. | 16:55 |
eleitl | It is muscle memory, coordination, nonverbal stuff. Can't document that. | 16:56 |
kanzure | eleitl: what do you estimate the complexity would be of robotic-only cryo surgery | 16:56 |
eleitl | Want to see it watch a surgical team during a complex procedure, be part of it. | 16:56 |
eleitl | Can't be done. Not for many decades. | 16:56 |
kanzure | yes, i have seen large choreographed surgery teams >20 people | 16:56 |
eleitl | Sirgical robots are assclowns. | 16:56 |
eleitl | Surgical. | 16:56 |
eleitl | kanzure, try being on the team while in a procedure | 16:57 |
eleitl | Just look what happens. | 16:57 |
eleitl | Very instructive. | 16:57 |
kanzure | watching? sure | 16:57 |
eleitl | Yeah, just gown up, and be a good observer. | 16:57 |
eleitl | If you don't interfere they'll tolerate you there. | 16:58 |
eleitl | Don't get me wrong, you can reduce manpower with simple automation. | 16:58 |
eleitl | Just logging can reduce the team by maybe 2-5 people. | 16:58 |
eleitl | If you can do intubation manually, you can use the cooldown machinery do the rest. No need to control each valve. | 16:59 |
eleitl | You can't, in any case. | 16:59 |
kanzure | and each valve is usually... manually controlled? | 16:59 |
eleitl | But please don't try a robot try an intubation. At least just yet. | 16:59 |
kanzure | agreed | 16:59 |
eleitl | We killed enough dogs by just being people. | 17:00 |
kanzure | learn anything? | 17:00 |
eleitl | I've tried telling what I learned. | 17:00 |
eleitl | Certainly, a lot of humility. | 17:00 |
kanzure | you missed a great opportunity for "I've learned that we're all doomed" | 17:01 |
eleitl | No, it's just hard. | 17:01 |
eleitl | Harder than what most people are used to, but that's not that big of a problem. | 17:01 |
eleitl | Medicine is doing it, so why can't we? | 17:01 |
eleitl | But if you ever hear someone, just train the morticians, and let the embalmer handle the perfusion. | 17:02 |
eleitl | Just do me a favor, and shoot them dead while you have a chance? | 17:02 |
kanzure | uh, what about anesthesia? | 17:02 |
eleitl | In case of animal experiments, sure, you obviously need an anaesthesiologist. And a licencense veterinarian. And an ethics board. | 17:03 |
cluckj | "tacit knowledge" is what you're looking for | 17:03 |
eleitl | In case of a normal patient, there are drugs that make sure you don't pull a Lazarus in the ice bath | 17:04 |
eleitl | For active shutdown, you need the whole monty, because you're starting with a fully conscious human patient | 17:04 |
eleitl | I'm not aware that such a procedure has been done. | 17:05 |
eudoxia | you know, alcor might not be perfect, but at least they fought the legal battle for the right to cryopreserve | 17:05 |
eudoxia | CI was like "morticians? sure, whatever" | 17:05 |
eudoxia | "an embalming pump will do" | 17:05 |
eleitl | This is not to say it has not been done, but I'm not aware of anyone ever talking about that. It's just what makes most sense technically. | 17:05 |
eleitl | Exactly, eudoxia. | 17:06 |
eleitl | And most cryonics is that way. | 17:06 |
eudoxia | mhm | 17:06 |
eleitl | Which is why I'm not part of any of that. | 17:06 |
eleitl | People who have tried to compromise initially just were stuck there. | 17:07 |
eleitl | You shouldn't go intere in the assumption that you can cut corners, and improve later on. | 17:07 |
eleitl | Because there is never a later on. | 17:07 |
eleitl | The best way to know is your experimental animals stop dying. | 17:08 |
kanzure | i wonder whether alcor had excellent surgeons from th ebeginning | 17:08 |
-!- Guest8333 is now known as vivi | 17:08 | |
eleitl | How do you know you're still doing well? | 17:08 |
eleitl | No cryonics patient have ever sued. | 17:08 |
eleitl | Dead men tell no tales, and all that. | 17:08 |
eleitl | Tell me what you think about this tirade. | 17:10 |
eleitl | If you think I'm full of shit, I want to hear it. | 17:10 |
kanzure | short-term you have never been full of shit :-) | 17:11 |
kanzure | long-term, we need to get our act together :-) | 17:11 |
-!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap | 17:11 | |
kanzure | i have been thinking of a "backup plan" in case nobody succeeds in the next 10-20 years | 17:11 |
kanzure | i think that some projects can begin now that in 200 years will see results, maximum | 17:11 |
eudoxia | kanzure: what are you success criteria for that timeframe? | 17:12 |
kanzure | blowing up the galaxy | 17:12 |
eudoxia | the speed of light forbids | 17:12 |
kanzure | well, i'm exaggerating...... slightly. | 17:12 |
eleitl | Pity be on the fools that are in our light cone. | 17:12 |
kanzure | many of these important problems could have been solved >100 years ago with the right constraints and right approach | 17:13 |
kanzure | i don't want to leave shit at the same state for the next hapless idiots that show up | 17:13 |
eleitl | No, modern medicine is barely there 1970s. And cryogenics is also about a century old. | 17:13 |
eleitl | And any patient would not have survived WWII, because supply disruption. | 17:14 |
eleitl | This thing is barely 50 years old, and it might not well last another 50 years old. | 17:14 |
eleitl | A single month of LN supply missed: total patient loss. | 17:15 |
kanzure | active supply is a dumb idea | 17:15 |
kanzure | if you are on a 500k year cryoship journey then you don't have no active supply | 17:15 |
eleitl | A patient fund is not based on a steady state nevermind contractiving economy: total patient losss. | 17:15 |
eleitl | I agree with you on need to make cryogens on site, but this not done by anyone | 17:15 |
eleitl | Because, no need, and too expensive, and it has worked since the 1970s, so why should it break now? | 17:16 |
eleitl | Ditto financials. | 17:16 |
kanzure | heh not a particularly long-term perspective, but whatever | 17:16 |
eleitl | Had zero success there. | 17:16 |
eleitl | A century would do. But you won't get a century. | 17:16 |
eudoxia | it's funny how a project started by libertarians depends on the continued existence of fractional-reserve banking | 17:17 |
eleitl | Or just a constantly growing economy. | 17:17 |
eudoxia | mike darwin said, "a cryonics org should be able to weather a war that last centuries". so far none come close. | 17:17 |
kanzure | the *org*? hah | 17:17 |
eleitl | Building institutions that last while the world changes is very hard. | 17:18 |
kanzure | what good would an org do when you are 100 kilolightyears out? | 17:18 |
eleitl | Storage is irrelevant if you can't be resuscitated. | 17:18 |
kanzure | yes, i'm aware | 17:18 |
eudoxia | kanzure: well, organizations are hard, but so are autonomous eternal cryobunkers / space vehicles | 17:19 |
eleitl | So just barely hangin on wouldn't do. | 17:19 |
kanzure | eudoxia: eternal organizations seem harder :-) | 17:19 |
eudoxia | probably | 17:19 |
eleitl | If after a century you're still hanging on, then probably you have no chances. | 17:20 |
kanzure | eleitl: i was commenting in here the other day that all cryo trips into distant future should always involve a female and either a male or fertilization material, because it would suck in 10 million years to find a human coming out of storage but it's the last of its kind | 17:20 |
eleitl | If the world hasn't largely shrugged off our current problems by 2080, the patients are fucked. | 17:20 |
eleitl | The resuscitation procedure doesn't involve a warmup, so it doesn't matter what kind of primat in which shape is in there. | 17:21 |
eleitl | primate | 17:21 |
kanzure | perhaps | 17:22 |
eleitl | If cryonicists knew what kind of world awaits them in case of successful resuscitation, they likely be scared shitless. | 17:22 |
eleitl | Anything less can't possibly "revive" them. | 17:23 |
eleitl | I don't mind, I like to be scared. | 17:23 |
kanzure | hmm, i would rather have something that works on its own, rather than relying on future entities | 17:24 |
eleitl | Too bad. We currently can barely make you fit to take that journey. | 17:24 |
eleitl | It sucks to rely on future people, but if you fail it won't hurt one bit. | 17:24 |
kanzure | my plans are slightly different | 17:24 |
eleitl | Not dying? | 17:25 |
kanzure | no, i mean my plans regarding whether i want to rely on future magic people to fix my mistakes | 17:26 |
eleitl | There is a tendency to be arbitrarily sloppy today, because our friends in the future | 17:27 |
eleitl | But a low-tech society of the future is entirely useless, even if your environment is self-contained. | 17:28 |
eleitl | So you have to do due diligence, but also rely on people on the future to be there, and having enough clue. | 17:28 |
eleitl | If not, you can just stick with embalming. Or mummification. | 17:29 |
eleitl | Right now I think the people of the future will suck, but what do I know. | 17:29 |
eudoxia | regarding embalming, some years ago darwin wrote a new cryonet post about embalming, and said he was working on a longer blog post about the embalming process used on eva peron. that didn't happen though. | 17:29 |
eleitl | Last time he worked on sugar-based preservation, I think. | 17:31 |
eleitl | In case anything came out of it it has been lost in the fire. | 17:31 |
eudoxia | i didn't know his place caught fire | 17:31 |
eudoxia | that's terrible | 17:32 |
eleitl | I have problems with alternative to cryopreservation, though it kinda works for desiccation-resistant anhydrobiosis. | 17:32 |
eudoxia | from the few pictures i saw of the inside of his home it was like a cryonics museum | 17:32 |
eleitl | Pretty much total loss. | 17:32 |
eleitl | Yes, very terrible. | 17:32 |
eleitl | He worked on digitizing, but it wasn't much. | 17:33 |
eleitl | If ever you think you're too paranoid in regards to backups: you're probably not paranoid enough. | 17:33 |
eleitl | Anyways, time to hit the matress. | 17:34 |
eleitl | Good night, you bright shiny people. | 17:34 |
* eudoxia has been slacking on his backups | 17:34 | |
eudoxia | good nihgt | 17:34 |
eudoxia | night* | 17:34 |
dingo | ahh i can sole the time-series easily, just a variable-gradient legend of the same color hues, with markers indicating time | 17:39 |
dingo | using the exact same colors, determining the horizontal width, stacked side to side in order | 17:39 |
kanzure | "An experimental System-on-Chip with a custom compiler toolchain" https://github.com/combinatorylogic/soc | 17:45 |
kanzure | https://github.com/combinatorylogic/soc/blob/master/backends/small1/hw/rtl/cpu.v | 17:47 |
-!- mindout [054331f6@gateway/web/freenode/ip.5.67.49.246] has joined ##hplusroadmap | 18:07 | |
-!- mindout [054331f6@gateway/web/freenode/ip.5.67.49.246] has left ##hplusroadmap [] | 18:08 | |
maaku | kanzure: if you like that, you might be interested in the more production-ready moxie SoC | 18:11 |
-!- eudoxia [~eudoxia@r186-50-158-185.dialup.adsl.anteldata.net.uy] has quit [Quit: Leaving] | 18:16 | |
kanzure | show? | 18:17 |
maaku | show? | 18:33 |
-!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-zroddtkfnharibct] has joined ##hplusroadmap | 18:36 | |
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-vfpkfvcxqmwovowd] has quit [Quit: Connection closed for inactivity] | 19:21 | |
kanzure | i mean send link | 19:39 |
JayDugger | Good evening, everyone. | 19:55 |
kanzure | yessss? | 20:06 |
-!- wrldpc1 [~ben@hccd37dcd12.bai.ne.jp] has joined ##hplusroadmap | 20:07 | |
-!- ant4t [~s3an@2602:306:cd87:2e0:9dae:12dc:d5b5:c725] has joined ##hplusroadmap | 20:16 | |
-!- justanotherusr [~Justan@unaffiliated/justanotheruser] has quit [Ping timeout: 265 seconds] | 20:26 | |
kanzure | hi ant4t | 20:36 |
-!- wrldpc1 [~ben@hccd37dcd12.bai.ne.jp] has quit [Quit: wrldpc1] | 20:53 | |
-!- wrldpc1 [~ben@hccd37dcd12.bai.ne.jp] has joined ##hplusroadmap | 20:58 | |
-!- wrldpc1 [~ben@hccd37dcd12.bai.ne.jp] has quit [Quit: wrldpc1] | 22:02 | |
-!- sheena2 [~home@S0106c8be196316d1.ok.shawcable.net] has joined ##hplusroadmap | 22:20 | |
-!- sheena [~home@S0106c8be196316d1.ok.shawcable.net] has quit [Read error: Connection reset by peer] | 22:20 | |
-!- erasmus [~esb@unaffiliated/erasmus] has joined ##hplusroadmap | 22:28 | |
-!- Quashie__ [~boingredd@50.14.92.17] has joined ##hplusroadmap | 23:12 | |
-!- Quashie_ [~boingredd@50.14.92.17] has quit [Ping timeout: 276 seconds] | 23:13 | |
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 256 seconds] | 23:14 | |
-!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap | 23:17 | |
-!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-zroddtkfnharibct] has quit [Quit: Connection closed for inactivity] | 23:18 | |
-!- ant4t [~s3an@2602:306:cd87:2e0:9dae:12dc:d5b5:c725] has quit [Ping timeout: 256 seconds] | 23:21 | |
-!- wrldpc1 [~ben@hccd37dcd12.bai.ne.jp] has joined ##hplusroadmap | 23:34 | |
-!- Quashie_ [~boingredd@50.14.92.17] has joined ##hplusroadmap | 23:39 | |
-!- Quashie__ [~boingredd@50.14.92.17] has quit [Ping timeout: 255 seconds] | 23:41 | |
-!- wrldpc1 [~ben@hccd37dcd12.bai.ne.jp] has quit [Quit: wrldpc1] | 23:44 | |
-!- wrldpc1 [~ben@hccd37dcd12.bai.ne.jp] has joined ##hplusroadmap | 23:49 | |
maaku | kanzure: https://github.com/atgreen/moxie-cores | 23:54 |
maaku | http://moxielogic.org/blog/pages/architecture.html | 23:54 |
maaku | .title | 23:54 |
yoleaux | The Moxie Blog – Architecture | 23:54 |
--- Log closed Sat Jun 20 00:00:46 2015 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!