2012-04-22.log

--- Log opened Sun Apr 22 00:00:46 2012
diginetthe fleshlight forums00:01
diginetprepare to laugh, and cry00:01
joshcryerOh god.00:03
joshcryerYes, yes, indeed.00:03
diginetI particularly love the thread polling how many of them the members owned00:05
diginetonly there would owning more be viewed as an accomplisment00:06
-!- cypan [~cypan@96-41-2-124.dhcp.mtpk.ca.charter.com] has quit [Quit: leaving]00:11
-!- _F7_ [~forrestfl@50.15.211.25] has quit [Ping timeout: 246 seconds]00:14
diginethttp://youtu.be/XP5lz2CYNR4 dear god why00:54
-!- _Sketch_ [~Sketch@71-91-221-175.dhcp.stls.mo.charter.com] has quit [Read error: Connection reset by peer]00:54
diginetI had forgotten about that atrocity00:54
diginetGARUGAMESH00:54
-!- _Sketch_ [~Sketch@71-91-221-175.dhcp.stls.mo.charter.com] has joined ##hplusroadmap00:54
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has joined ##hplusroadmap01:11
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has quit [Ping timeout: 246 seconds]01:30
-!- Cat4D [182bd7e2@gateway/web/freenode/ip.24.43.215.226] has joined ##hplusroadmap01:51
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap02:04
-!- strages_oz [~strages@173-14-214-62-Atlanta.hfc.comcastbusiness.net] has quit [Quit: Lost terminal]02:04
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has joined ##hplusroadmap02:06
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Leaving]03:23
-!- sylph_mako [~mako@118-93-18-64.dsl.dyn.ihug.co.nz] has quit [Ping timeout: 260 seconds]03:26
-!- kvltist [~Kvltist@p5B33FE3E.dip.t-dialin.net] has joined ##hplusroadmap03:33
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has quit [Quit: Leaving]03:44
-!- kvltist [~Kvltist@p5B33FE3E.dip.t-dialin.net] has quit [Ping timeout: 265 seconds]04:57
-!- AdrianG [~amphetami@unaffiliated/amphetamine] has joined ##hplusroadmap05:36
-!- jennicide [jen@173-18-211-34.client.mchsi.com] has joined ##hplusroadmap05:56
-!- jenzebubble [jen@173-18-211-34.client.mchsi.com] has quit [Ping timeout: 276 seconds]05:59
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has joined ##hplusroadmap06:57
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has quit [Ping timeout: 246 seconds]07:19
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has joined ##hplusroadmap07:20
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has joined ##hplusroadmap07:30
-!- jmil [~jmil@c-68-81-252-40.hsd1.pa.comcast.net] has joined ##hplusroadmap07:31
-!- amphetamine [~amphetami@CPE1c6f658faa04-CM001bd713703a.cpe.net.cable.rogers.com] has joined ##hplusroadmap07:48
-!- amphetamine [~amphetami@CPE1c6f658faa04-CM001bd713703a.cpe.net.cable.rogers.com] has quit [Changing host]07:48
-!- amphetamine [~amphetami@unaffiliated/amphetamine] has joined ##hplusroadmap07:48
-!- AdrianG [~amphetami@unaffiliated/amphetamine] has quit [Disconnected by services]07:48
-!- amphetamine [~amphetami@unaffiliated/amphetamine] has quit [Client Quit]07:52
-!- jmil [~jmil@c-68-81-252-40.hsd1.pa.comcast.net] has quit [Quit: jmil]07:57
-!- SDr [SDr@unaffiliated/sdr] has joined ##hplusroadmap08:03
-!- d3nd3 [~dende@cpc10-croy17-2-0-cust245.croy.cable.virginmedia.com] has joined ##hplusroadmap08:06
-!- AdrianG [~amphetami@unaffiliated/amphetamine] has joined ##hplusroadmap08:16
-!- jmil [~jmil@c-68-81-252-40.hsd1.pa.comcast.net] has joined ##hplusroadmap08:27
-!- jmil [~jmil@c-68-81-252-40.hsd1.pa.comcast.net] has quit [Client Quit]08:28
-!- jmil [~jmil@c-68-81-252-40.hsd1.pa.comcast.net] has joined ##hplusroadmap08:29
-!- joshcryer [~g@unaffiliated/joshcryer] has quit []08:34
-!- _F7_ [~forrestfl@50.15.211.25] has joined ##hplusroadmap08:37
ParahSailin_F7_, hey08:39
ParahSailindid that laser cutter for metal have any sort of ventilation?08:39
ParahSailini didnt think of that at the time08:40
-!- jmil [~jmil@c-68-81-252-40.hsd1.pa.comcast.net] has quit [Quit: jmil]08:41
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has quit [Ping timeout: 246 seconds]09:01
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap09:10
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has joined ##hplusroadmap09:13
-!- jmil [~jmil@SEASNet-148-05.seas.upenn.edu] has joined ##hplusroadmap09:27
delinquentmejmil, Y U LOOK FAMIRIAR09:33
-!- d3nd3 [~dende@cpc10-croy17-2-0-cust245.croy.cable.virginmedia.com] has quit [Quit: Leaving.]09:33
jmilwhut?09:34
-!- Cat4D [182bd7e2@gateway/web/freenode/ip.24.43.215.226] has quit [Ping timeout: 245 seconds]09:43
-!- rkos [~chatzilla@a88-113-144-253.elisa-laajakaista.fi] has quit [Read error: Connection reset by peer]09:43
-!- rkos [~chatzilla@a88-113-144-253.elisa-laajakaista.fi] has joined ##hplusroadmap09:43
_F7_Oh, the plasma cutter?09:44
_F7_It's got forced air from the nozzle, but it's not really ventilated09:45
_F7_why?09:45
ParahSailini would want to avoid breathing whatever metal fumes come off of vaporizing metal09:46
_F7_it's slag, it gets blasted downward09:46
_F7_just fat oxides and molten metal09:46
ParahSailinah, so negligible fumes?09:46
_F7_very09:46
-!- nmz787 [~Nathan@cpe-67-242-177-23.rochester.res.rr.com] has joined ##hplusroadmap09:53
kanzurehi nmz78709:53
kanzureroksprok: can i see the protocol site progress sometime? either code or the site itself..09:55
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has quit [Ping timeout: 246 seconds]09:55
kanzurescripted favicons :/ http://news.ycombinator.com/item?id=387362310:05
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has joined ##hplusroadmap10:11
nmz787yo kanzure10:14
delinquentmeso what happens with open source projects in which *only* old versions are open sourced10:14
delinquentmeim trying to model this .. like the dev on the old machines would increase for sure10:14
delinquentmealso I was thinking about a model around getting suppliers to pay in for me to adapt one of their OEM parts to the machine10:15
delinquentmeboth those options contribute to development and also allow for obvious fundability?10:16
delinquentmefundabilteh *10:16
-!- nsh [~nsh@wikipedia/nsh] has joined ##hplusroadmap10:17
delinquentmenmz787, whats on the plate??10:42
delinquentmeyou guys find the lens you needed ?10:42
-!- lichen|2 [~lichen@c-76-105-164-184.hsd1.or.comcast.net] has joined ##hplusroadmap11:00
-!- lichen [~lichen@c-76-105-164-184.hsd1.or.comcast.net] has quit [Ping timeout: 244 seconds]11:01
nmz787havent checked the forum pots11:05
nmz787post11:05
-!- jmil [~jmil@SEASNet-148-05.seas.upenn.edu] has quit [Quit: jmil]11:06
nmz787kanzure: ever hear re chandni?11:15
kanzurenot yet11:16
kanzurewell, yes, but nothing worth reporting11:17
klafka1dammit11:23
klafka1i was supposed togo to legal observers training today11:24
-!- splicer [~ubuntu@c83-255-190-140.bredband.comhem.se] has quit [Ping timeout: 250 seconds]11:27
-!- Netsplit *.net <-> *.split quits: SDr11:37
-!- sylph_mako [~mako@118-93-18-64.dsl.dyn.ihug.co.nz] has joined ##hplusroadmap11:41
klafka1hey kanzure http://groups.google.com/group/julia-dev/browse_thread/thread/61fb4e3847dcc2b9?pli=111:48
klafka1any thoughts on julia the language?11:48
-!- lichen|2 is now known as lichen11:50
kanzureklafka1: are you looking at it because it has JIT?11:54
-!- Netsplit over, joins: SDr11:55
-!- bkero1 [~Ben@li280-127.members.linode.com] has joined ##hplusroadmap11:57
-!- bkero1 [~Ben@li280-127.members.linode.com] has quit [Client Quit]11:58
-!- bkero is now known as bkero211:59
-!- bkero [~Ben@li280-127.members.linode.com] has joined ##hplusroadmap12:00
-!- bkero [~Ben@li280-127.members.linode.com] has quit [Changing host]12:00
-!- bkero [~Ben@osuosl/staff/bkero] has joined ##hplusroadmap12:00
-!- bkero [~Ben@osuosl/staff/bkero] has quit [Client Quit]12:03
-!- bkero [~Ben@osuosl/staff/bkero] has joined ##hplusroadmap12:04
-!- klafka1 [~klafka@c-67-174-253-229.hsd1.ca.comcast.net] has quit [Quit: This computer has gone to sleep]12:11
-!- Cat4D [182bd7e2@gateway/web/freenode/ip.24.43.215.226] has joined ##hplusroadmap12:38
Cat4Di need a dvi/hdmi/wqxga mems gate for beam steering using vid card, prefer both pixel and analogue with 3+ channels gated ... anyone know of any devices? (typical 1-3w each laser channel)12:41
-!- Cat4D [182bd7e2@gateway/web/freenode/ip.24.43.215.226] has quit [Ping timeout: 245 seconds]13:00
-!- strages_home [~strages@adsl-98-81-13-133.hsv.bellsouth.net] has joined ##hplusroadmap13:16
-!- Mariu [Jimmy98@89.41.57.33] has quit [Quit: leaving in a transit pod]13:27
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has quit [Remote host closed the connection]13:35
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has joined ##hplusroadmap13:38
-!- DrOctothrope [~DrOctothr@173-147-2-229.pools.spcsdns.net] has quit [Ping timeout: 246 seconds]13:41
-!- DrOctothrope [~DrOctothr@50.15.211.25] has joined ##hplusroadmap13:41
-!- augur_ [~augur@208.58.5.87] has joined ##hplusroadmap13:44
-!- _Sketch_ [~Sketch@71-91-221-175.dhcp.stls.mo.charter.com] has quit [Read error: Connection reset by peer]13:44
-!- _Sketch_ [~Sketch@71-91-221-175.dhcp.stls.mo.charter.com] has joined ##hplusroadmap13:45
-!- yashgaroth [~f@cpe-66-27-117-179.san.res.rr.com] has joined ##hplusroadmap13:45
-!- strages_home [~strages@adsl-98-81-13-133.hsv.bellsouth.net] has quit [Quit: Lost terminal]13:46
-!- strages_home [~strages@adsl-98-81-13-133.hsv.bellsouth.net] has joined ##hplusroadmap13:52
-!- DrOctothrope [~DrOctothr@50.15.211.25] has quit [Ping timeout: 240 seconds]13:53
-!- roksprok_ [~Zac@74.83.205.124] has joined ##hplusroadmap13:54
-!- Netsplit *.net <-> *.split quits: augur, roksprok13:54
-!- roksprok_ is now known as roksprok13:54
-!- n_bentha [~lolicon@75.111.75.178] has joined ##hplusroadmap14:00
-!- sylph_mako [~mako@118-93-18-64.dsl.dyn.ihug.co.nz] has quit [Ping timeout: 265 seconds]14:09
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has quit [Quit: Leaving]14:12
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has joined ##hplusroadmap14:12
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Ping timeout: 246 seconds]14:13
-!- nmz787 [~Nathan@cpe-67-242-177-23.rochester.res.rr.com] has quit [Quit: Leaving.]14:15
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap14:16
-!- augur_ [~augur@208.58.5.87] has quit [Remote host closed the connection]14:17
-!- ParahSailin [~parah@unaffiliated/parahsailin] has quit [Read error: Connection reset by peer]14:32
-!- ParahSailin [~parah@unaffiliated/parahsailin] has joined ##hplusroadmap14:32
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has quit [Quit: Leaving]14:44
-!- Guest73883 is now known as poptire14:45
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has quit [Ping timeout: 246 seconds]14:57
-!- superkuh [~superkuh@unaffiliated/superkuh] has quit [Quit: the neuronal action potential is an electrical manipulation of reversible abrupt phase changes in the lipid bilayer]15:01
-!- superkuh [~superkuh@c-24-118-174-49.hsd1.wi.comcast.net] has joined ##hplusroadmap15:01
-!- superkuh [~superkuh@c-24-118-174-49.hsd1.wi.comcast.net] has quit [Changing host]15:01
-!- superkuh [~superkuh@unaffiliated/superkuh] has joined ##hplusroadmap15:01
-!- d3nd3 [~dende@cpc10-croy17-2-0-cust245.croy.cable.virginmedia.com] has joined ##hplusroadmap15:09
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has joined ##hplusroadmap15:10
-!- splicer [~ubuntu@c83-255-190-140.bredband.comhem.se] has joined ##hplusroadmap15:15
-!- Adifex [~Adifex@utdpat241140.utdallas.edu] has joined ##hplusroadmap15:17
-!- klafka1 [~klafka@c-67-174-253-229.hsd1.ca.comcast.net] has joined ##hplusroadmap15:20
klafka1kanzure lookign at it because it's fast15:22
klafka1i guess15:22
klafka1and decent15:22
kanzurefast for what? just in general?15:22
klafka1yeah15:23
klafka1compared to say python/ruby/et al15:23
-!- augur [~augur@c-69-250-19-178.hsd1.md.comcast.net] has joined ##hplusroadmap15:23
klafka1like the language is farly ruby/pythonic but like fast15:23
kanzuredo you need the extra speed? is pypi/cython/cpython not working out?15:24
klafka1i don't currently15:24
klafka1but i mean i have in the past and i will in the future15:25
klafka1the lead of cython in that google groups post is advocating for julia over cython15:25
klafka1in fact15:25
klafka1well not over but as an excellent complement15:25
klafka1like for instance i know right now we are using all ruby but ruby is fucking slow so if i do stuff like build a recommendation engine i'm going to need to do that in another language15:26
kanzuredepends on how you are doing it15:27
kanzureif you put all your recommendation code in each page request, yes, that's going to be slow no matter what you do15:27
kanzureif you pre-compute your recommendations you can just pull those up on each page load15:27
kanzureyour recommendations shouldn't be recomputed on each page load, in general15:27
klafka1kanzure well it depends what sort of algorithm i'm doing15:27
klafka1i would never do that15:27
klafka1but even pre-computing is potentially expensive15:28
kanzureok just checking15:28
klafka1for instance if i do matrix completion via gradient descent15:28
kanzureif you are pre-computing them, 0.1 sec vs. 0.2 sec is only somewhat important15:28
klafka1vs. say a jacquard similarity matrix15:28
kanzure(on a page load, 1sec vs. 3sec is hugely important)15:28
klafka1um but 2-3hrs vs. 30 minutes15:28
klafka1is important15:28
kanzure2-3 hours for all users?15:28
klafka1yeah15:28
kanzurewait, i mean15:28
klafka1i mean i have no idea how long it will take to compute15:29
kanzure2-3 hours per user?15:29
klafka1i was thinking total15:29
kanzurebtw: you can come up with a rough approximation of how long it would take15:29
klafka1but im ean for comparison some of my research, each run to convergence took 7hrs15:29
klafka1kanzure i can't because i don't know what algorithm i'd use15:29
klafka1heh15:29
kanzurei'm really confused15:29
kanzurehow do you know it took 7 hours in that case?15:29
klafka1well i didn't finish! geez hold on, so in comparison in some of my research a run to convergence took 7hrs after i added a bunch of C. it would take 30-40 hrs without the C15:30
klafka1my point is it's in situations like this where using a faster language is important15:30
kanzurei am really skeptical of this. what is your algorithm doing that took it 40 hours?15:31
klafka1? learning a 100^2 variable factor graph from data15:31
-!- Adifex [~Adifex@utdpat241140.utdallas.edu] has quit [Quit: Adifex]15:32
kanzure"learning" a graph?15:32
kanzureand i'm not clear on what the structure of your graph actually is15:32
klafka1i was using stochastic gradient descent wherein I had to estimate the partition function and the gradient using MCMC15:33
klafka1learning a graphical model?15:33
klafka1a factor graph is a generalization of a Markov random field wherein each node has a clique of one15:33
klafka1the structure of the graph represents the conditional probability distribution of the variables15:34
klafka1anyway my point about that was, there are distinct cases when you are doing ML where the speed of the language matters. In my job of building a recommendation engine, I don't know that this will be an issue, but I foresee it may be one, hence my interest in a fast yet syntactically non sucky language.15:35
klafka1algorithms taking 10-100+ hrs are fairly common in computational statistics also kanzure15:36
kanzuredirectededge.com built their own graph database to get their recommendation engine fast enough15:36
kanzure10-100 hours in recommendation engines is not OK in web apps15:36
kanzurebrownies: any thoughts?15:36
klafka1why?15:36
klafka1i'm really really skeptical of graph based recommendation engines btw15:36
klafka1I'm also skeptical as to the point of a graph database unless you're doing lots of traversals15:37
kanzureyou just said you're using a graph of your own15:37
brownieso.O15:37
browniesreading15:37
klafka1a graph is a totally different thing from a graphical model15:37
klafka1and I stated that was something I was doing in my research15:37
browniessounds like some fun research15:39
klafka1it was15:39
browniesanyway it's fairly obvious the approach they're taking and they are doing tons of traversals so it's cool15:39
browniesif you want to go the more "normal" route of piling matrices onto each other then you could do C or Haskell i suppose15:39
brownieshehe, accidental math pun. anyway.15:39
klafka1well the actual part of it that was 'research' was building a system of sample based regularization techniques based on a priori knowledge to bias the model15:39
klafka1heh15:39
browniesi think you are going to want something higher level than C15:40
kanzureklafka1: maybe i am misunderstanding. is it 10-100 hours per recommendation?15:40
brownies*per* recommendation? -_- no15:41
klafka1well 10-100 hrs to learn to convergence15:41
klafka1so that'd be for _all_ recommendations15:41
kanzurewell ok. a 100 hour algorithm that you run once or twice a year.. that sounds ok15:41
klafka1a 100 hr algorithm isn't really ok either15:41
browniesnaw, i believe in his approach he'd have to update and re-run things as new data came in15:41
klafka1yeah15:41
klafka1unless i did something online15:41
browniesthere's something to be said for the pagerank-esque approch man15:41
browniesbecause you just update one edge and let it hang out15:42
browniesnext time the edge is needed, stuff will be computed on the fly as always and it'll all Just Work(tm)15:42
klafka1idk i've really only dug a little into IR stuff15:42
browniesIR?15:42
klafka1information retrieval15:42
browniesoh15:42
klafka1i mean recommendation engines / collaborative filtering are a subset of IR iirc15:42
browniesi strongly doubt you'll need to optimize things on that front15:42
klafka1yeah15:42
browniesjust throw it all into an existing SQL database with a sane schema15:43
browniesthat should last you a good long while.15:43
klafka1i'm going to start with just a jacquard similarity matrix15:43
klafka1which has the advantage of being retardedly fast and I can update online15:43
browniesboring, but fine.15:44
klafka1yep15:44
browniesincidetnally, i need to build a recommendation engine soon too.15:44
brownieslet me know how it goes!15:44
klafka1well a couple things i've been learning is that for recommendation engines since all models suck pretty bad, it is often better to focus on variable selection, as well as non-algorithm stuff like placement and presentation15:44
klafka1because I have basically pieces of content, as well as how people interact with that content15:45
klafka1so for instance i have 'bob's blog article 1' and a person may have 'visited/read' the article they may have 'liked' the article they may hvae 'shared' they may have 'commented' etc...15:45
ParahSailinwhat are you guys talking about15:46
klafka1recommendation engines15:46
kanzurerecommendation engines, algorithms, web apps15:46
klafka1it's kind of sad though, i'd like to have problems that let me do devilishly cool algorithms15:46
klafka1i'm really itching  to implement ensemble MCMC methods15:47
klafka1i really really really want to15:47
klafka1argh i think i drank too much caffeine to the point where i'm having a hard time writing boring stuff15:48
-!- nmz787 [~Nathan@cpe-67-242-177-23.rochester.res.rr.com] has joined ##hplusroadmap15:53
-!- augur [~augur@c-69-250-19-178.hsd1.md.comcast.net] has quit [Read error: Connection reset by peer]15:55
-!- augur [~augur@c-69-250-19-178.hsd1.md.comcast.net] has joined ##hplusroadmap15:56
delinquentmeklafka1, you should check out the recommendation engine they're working on for reddit16:06
delinquentmelast I heard they were in the process of defining the solution space16:06
delinquentmelooking @ a number of diff algos16:06
kanzureugh reddit again16:06
strangewarpgoddamnit reddit16:07
strangewarpneeds fewer upvote-based economies; more overlapping social-vetting economies. also link them to physical goods maybe eventually if the governments are ok with it [puppy dog eyes]16:09
klafka1what is an 'overlapping social vetting' economy?16:11
klafka1and how is that different from an 'upvote based economy'16:11
klafka1isn't one a subset of the other16:11
klafka1also delinquentme link plz16:12
strangewarpBasically I'd like to see whether many multiple overlapping reputation economies, bundled up by aggregators, could be more successful than single reputation economies, which end up biased if they have a monopoly on the service16:15
strangewarpAnd I think such a system could be extended to verify that people are experts in given skills, and whatnot16:16
klafka1so are you looking at determining the importance of content explicitly based on the known knowledge of people in the system and their actions within the system?16:16
klafka1e.g. a new bio paper comes out it coudl be tagged as interesting or good or what have you because a lot of known biology experts are commenting on it ?16:17
delinquentmeklafka1,  http://groups.google.com/group/rrecommender16:28
-!- nathaniel [~nathaniel@reddit/operator/nathaniel] has quit [Quit: Leaving]16:32
klafka1ah16:32
klafka1this thread is hilarious http://news.ycombinator.com/item?id=387585516:39
strangewarpsorry, was reading something there; I forgot that I'd been saying anything in IRC16:40
strangewarpklafka1: I imagine there would be feedback between concrete skill-verification, and the weight given to a person's evaluations of the worth of people and objects explicitly linked to their areas of expertise16:41
n_benthaklafka1: those guys are strange16:42
klafka1so you are still working on the premise of a voting economy but votes non-uniformly distributed based on a priori knowledge of their skill sets16:42
n_benthaceo's are supposed to earn a bit more than the rest of the people16:42
strangewarpTempered, of course, by systems to ensure that people with opposing political/religious opinions can't explicitly gang up on one another, except when the issue is concrete and testable..16:43
klafka1I kind of agree with the guy who said startup CEOs should make about as much as the most senior engineers16:43
kanzuren_bentha: or more realistically, you want to pay market rate for someone to achieve market results16:43
strangewarpklafka1: I suppose I'm saying for all the veneration of Reddit, there's still a long way to go16:43
klafka1strangewarp agreed16:43
kanzureno reddit veneration allowed in here16:43
kanzureshoo16:43
klafka1I've thought about this a lot too with regards to verifying expert skills16:44
klafka1it's hard16:44
strangewarpyeah :\16:44
n_benthathen why is the president paid more than senators?16:44
strangewarpI'm only a casual, so far as thinking about most things, but it's good to get some external validation on the issue16:44
n_benthabut yea, a 5x difference in pay rate is a little much.16:45
klafka1the whole point is to not pay a startup CEO a ton of money because you want them to have most of their investment be in stock16:45
kanzurenot really16:45
kanzuresome startup founders are being paid a substantial payout in series fundraising16:46
kanzureso that they can focus on the startup16:46
-!- Cat4D [182bd7e2@gateway/web/freenode/ip.24.43.215.226] has joined ##hplusroadmap16:46
kanzureinstead of e.g. waiting four years to get their "millions" out16:46
klafka1I've also heard that called a fairly classic warning sign in a startup16:46
kanzureklafka1: if you're an investor, you want a $100 million exit not a $10 million exit16:47
kanzureto a founder making $50k/year, a $10 million exit is fantastic16:47
kanzurean investor needs to incentive the founder to the point that a $10 million exit is not appealing16:47
klafka1mmmm16:48
klafka1interesting16:48
klafka1it sounds like in general then that what to pay a CEO depends on a number of factors, however, I think it can be a  fairly large warning sign when one is payed a very large salary particularly in an early stage startup16:52
delinquentmecorn based filament for repraps16:52
klafka1what is fizz buzz?16:56
klafka1um ok i just found it16:56
klafka1wow really people fail this that call themselves coders?16:56
klafka1i won't even call myeslf a developer but that is instantly easy16:57
diginetI think we need more. . .developers developers developers developers16:57
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Leaving]17:05
-!- Jora [~Jora@unaffiliated/jora] has joined ##hplusroadmap17:09
kanzurehi Jora17:11
Jorahi kanzure17:11
Jorathis channel active a lot or no?17:11
kanzureyep pretty active17:11
Joracool17:12
kanzurewhy?17:12
Jorathe neuroscience room is pretty dead usually and i wanted to find somewhere that discussed transhumanism a bit that was more active17:12
Joraive never been in here before17:13
kanzureah alright17:13
kanzureyes we have some neuroscience people17:14
klafka1oh that reminds me17:14
klafka1http://www.youtube.com/watch?v=q4KrhDZQ08817:15
klafka1Connectomics: Sebastian Seung vs. Tony Movshon, Columbia 201217:15
kanzurewe had someone in here claiming to be a friend of sebastien's17:15
kanzureabout a week ago17:16
Jorathanks17:16
kanzure*sebastian's17:16
Joralol , maybe true maybe not17:16
n_benthathx 4 da link17:17
klafka1interesting17:18
klafka1he's pretty cool17:19
klafka1or the stuff he's doing is17:19
klafka1i don't know anything about the person17:19
Jorathe best lay17:19
Joraare you all mainly biologists17:26
kanzuremix of biologists, coders, machinists17:27
n_bentha<3 machinists <317:27
delinquentmeJora,  you're in the right place :D17:30
delinquentmelots of discussion on tons of cool projects17:30
delinquentmethe join message covers it pretty well =]17:30
Jorayes looks like it :)17:30
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has quit [Quit: Leaving]17:30
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has joined ##hplusroadmap17:30
Jorai added this to my favorites & im watching the video that was posted earlier17:30
kanzureJora: are you a biologist?17:38
* delinquentme prays for electrical engineer17:39
Jorano im a information systems student with degree in psychology17:39
delinquentmenice!17:39
delinquentme<3 me some shrinkery17:39
delinquentmeJora, list you cognitive biases17:39
delinquentmejk jk17:39
Joralol17:39
Joraare you an EE?17:40
delinquentmeha!17:40
delinquentmewellll17:40
Jorastudent?17:40
delinquentmeim a learning EE17:40
delinquentmebusiness marketer on paper web programmer by day and aspiring biohacker17:41
delinquentmelike the rest of the wild-eyed optimists in here :D17:41
delinquentmeJora, you into carbon copies?17:41
kanzuredelinquentme: i don't think it's fair to say the others in here match your description17:42
delinquentmei meant the biohacker part :D17:42
kanzuresome of the users here are "accomplished biohackers, actual web programmers"17:42
delinquentmekinda17:42
delinquentmeto an extent17:42
Joranot familiar with the lingo of carbon copy unless you're talking about email17:42
delinquentmeJora, its a project to copy brains :D17:42
Jorai just have had a couple programming courses, im not a web programmer yet17:42
-!- Lucas_ [44c1135b@gateway/web/freenode/ip.68.193.19.91] has joined ##hplusroadmap17:43
Joradelinquentme sounds cool17:43
delinquentmeyeah im not sure if that was a cut @ me17:43
kanzurehi Lucas_17:43
Lucas_hi17:43
kanzuredelinquentme: wasn't. like i said, what you said wasn't fair17:43
Lucas_that was frighteningly fast17:43
klafka1biologists / coders / statisticians or some mix thereof17:44
Joraid like to do a masters in comp sci and then get some bio classes and go to grad school for computational neuroscience with some emphasis on neural engineering in the distant future17:44
klafka1i discourage people from comp neuroscience17:44
klafka1heh17:44
kanzurewhaat klafka1 why17:44
Jorabut i want to finish this info systems degree and work and get my maths17:44
klafka1they are the string theorists of the neuroscience wordl17:44
kanzurehardly17:44
kanzuremany of them have simulations based on physiological data17:45
klafka1well i suppose it depends what you mean by 'comp neuroscience'.17:45
klafka1link?17:45
klafka1and what sort of physiological data17:45
Joraid want to collect physiological data and make programs and interfaces to link people to computers17:45
klafka1how do they verify the 'wiring and firing' of their models?17:45
kanzurethe type extracted by neurophysiologists :P17:45
kanzureklafka1: by scanning tissues17:45
kanzureklafka1: like http://3scan.com/17:45
klafka1link?17:45
klafka1:P17:45
kanzurewiring is a huge part of the task, yes, but also the set of ion channels and receptors17:46
klafka1most comp neurosci people i've seen build neural nets or 'neural nets' to model stuff and verify it only at a very high level17:46
kanzureno that's your "AI" people or "machine learning" peeps17:46
klafka1no that's not at all17:46
klafka1well maybe AI people17:47
klafka1but not ML people17:47
kanzureok, it's unfair of me to place ML people there17:47
klafka1heh17:47
kanzurehowever! ANNs are very popular in ML17:47
Joradoes machine learning/ai have a place in neuroscience/neuro engineering stuff?17:47
klafka1um so yes17:47
kanzureyes, but not in the way that klafka is talking about17:47
klafka1yeah17:47
klafka1for instance17:47
klafka1connectomics is a great example17:47
klafka1connectomics is basically a huge computer vision/image problem17:47
kanzureklafka1: this doesn't feel like ANN-style research to me.. http://www.neuron.yale.edu/neuron/17:48
kanzure"NEURON: for empirically-based simulations of neurons and networks of neurons"17:48
delinquentmeklafka1, 3scan (theil backed) is working on it >> http://www.3scan.com/17:48
klafka13scan seems very interesting17:49
Lucas_kanzure: How do you know people at WFPF? (world federation parkour and freerunning )17:55
kanzureoh fooey.. let me think17:56
delinquentmeklafka1, there were a number of cool startups who got funded: http://www.kurzweilai.net/breakout-labs-announces-first-grants-to-support-radical-scientific-innovation17:56
kanzurei know someone there; just trying to think of the right person17:56
klafka1ah delinquentme17:56
klafka1excellent17:56
kanzureziyadb: was it you?17:56
delinquentmeklafka1, keep working on that17:56
delinquentmeyoull need the evil in that "excellent"17:56
kanzureit might be zacharycohn17:56
-!- yashgaroth [~f@cpe-66-27-117-179.san.res.rr.com] has quit [Ping timeout: 244 seconds]17:57
kanzureLucas_: as an aspiring superhero, parkour is important17:57
-!- Cat4D [182bd7e2@gateway/web/freenode/ip.24.43.215.226] has quit [Ping timeout: 245 seconds]17:57
Lucas_kanzure: wow, how don't you know? lol.  I know a few guys over there and have had the pleasure to train with them17:57
kanzureit's someone i know for other reasons17:58
Jorahow old is everyone here?18:00
Joraim 2818:00
kanzurei'm 1218:00
Lucas_2018:00
delinquentme27!18:00
Joralol18:00
-!- superkuh [~superkuh@unaffiliated/superkuh] has quit [Quit: the neuronal action potential is an electrical manipulation of reversible abrupt phase changes in the lipid bilayer]18:00
-!- superkuh [~superkuh@unaffiliated/superkuh] has joined ##hplusroadmap18:01
JoraParkour just seems like an early 20s thing thats why I asked18:02
delinquentmeJora, its for anyone with health insurance18:03
Jorayeah18:04
klafka12618:06
-!- _F7_ [~forrestfl@50.15.211.25] has quit [Quit: Computer has gone to sleep]18:08
Joranew game of thrones tonight :)18:09
delinquentmea rapman 3d printer?18:11
delinquentmeat like 3x the cost18:11
delinquentmeNICE18:11
-!- yashgaroth [~f@cpe-66-27-117-179.san.res.rr.com] has joined ##hplusroadmap18:14
yashgarothso uh my computer losing power every few hours and then starting right back up, I'm guessing is a PSU issue?18:15
delinquentmeyashgaroth, when was the last time you cleaned out the dust18:16
delinquentmesounds like a CPU overheating issue18:16
yashgarotha long time ago18:16
delinquentmedo that.18:16
delinquentmecompressed air if you have em18:16
delinquentmebetter yet ... air compressor to the insides18:16
yashgarothbut doesn't that usually give an error or something, and it wouldn't restart til it cools down?18:16
delinquentme+ gas mask18:16
delinquentmewell as soon as it kicks out into reboot mode its going to start cooling as the CPU load is much less no?18:17
delinquentmealso you computer =[18:17
n_benthayou don't need compressed air, yashgaroth18:17
delinquentmeit probs has a cat growing inside it18:17
yashgarothidk speedfan says it's not running hot18:17
n_benthajust take it apart, and take out the dust bunnies with your hands and paper towels18:17
n_benthaand reseat the memory18:17
delinquentmepay specific attention to the fuzzies in the CPU heat sink18:18
delinquentmereset the memory?18:18
n_benthareseating the memory can do wonders--trust me18:18
delinquentmeive never reset memory18:18
yashgarothram's seated, and even then shouldn't it throw an error?18:18
delinquentmedoesnt that happen when you reboot ?18:18
delinquentmesooo usually you start with simple fixes18:18
delinquentmeand then go up the line of severity in treatments18:19
delinquentmecleaning out the tower should be # 118:19
n_benthano. serious shit can happen to ram. even when u think it's seated fine18:19
n_benthajust clean it out and reconnect everything.18:19
delinquentmen_bentha, how do you reset it?18:19
delinquentmeoic18:19
n_benthareseat*18:19
yashgarothI had a previous problem where it wouldn't start unless I reseated the ram, but this is new18:19
n_benthadisconnect and reconnect :)18:19
delinquentmeyeah thats harmless enough18:19
delinquentmejust make sure that if your ram sticks are different that they both go back into their corresponding slots18:19
n_benthaa defrag and would be good too :)18:20
delinquentmeyashgaroth, windows pc?18:21
yashgarothwell I'll try dusting next time it shuts off, which should be in a few hours18:21
yashgarothyeah 718:21
yashgarothI know I know, but I like vidya games18:21
delinquentmeif you want a super quick fix you could also just remove the side panel18:21
delinquentmefor a little extra air flow18:21
n_bentha0.o18:21
yashgarothit's actually had the panel off for a while, my replacement cpu fan is too big for it to close :/18:22
n_benthahow old is the system?18:22
delinquentmeyour mobo might also have a display out to tell you whats going wrong18:22
delinquentmeLOL18:22
delinquentmeawesome18:22
n_benthawhat r your specs?18:22
delinquentmeyashgaroth, you playing d3?18:22
-!- Thorbinator_ [~Thorbinat@c-67-166-146-169.hsd1.ca.comcast.net] has quit [Read error: Connection reset by peer]18:22
yashgarothyup18:22
delinquentmeNICE18:22
yashgarothuhh it's pretty fast, I keep switching out parts18:22
yashgaroththe mobo/case is like 5 years old maybe18:23
yashgarothfriend's sister works at blizzard so I've been in the beta since it opened :D18:23
n_benthacool18:24
delinquentmeyashgaroth, die.18:24
delinquentmedie.18:24
delinquentmealso shes probs dead18:24
delinquentmelol i heard they work em harder than EA18:24
yashgarothnah she doesn't do coding18:24
delinquentmeoic18:24
yashgarothI don't think they have them on meth like MS does18:25
Joralol18:25
n_benthalololol]18:26
yashgarothyou laugh, but18:26
delinquentmeyashgaroth, company insurance plan provides adderall18:26
Joradoes MS having them working 80 plus hours per week?18:26
klafka1MS has them on meth eh?18:26
yashgarothyes18:26
Joralol.....18:27
yashgarothit's fairly well known, at least if you're from seattle I guess18:27
Joraor do the programmers take meth to code at max capacity18:27
JoraLOL!18:27
yashgarothI think they just hand it out for free and encourage working while tweaking18:27
Jorameth is the dumbest illegal drug to do18:27
JoraI honestly doubt that.18:28
yashgarothnot if it's high quality18:28
yashgarothI can think of several worse drugs18:28
delinquentmecoffee 2.0!18:28
n_benthaadderall is probably what they giv e them18:28
Joraif they have add18:28
n_benthaeven if they don't?18:30
-!- Lucas_ [44c1135b@gateway/web/freenode/ip.68.193.19.91] has quit [Quit: Page closed]18:30
n_benthaoh i'm supposed to get some of that cephalexin stuff next month18:30
n_bentha:)18:30
delinquentmelololol18:30
yashgaroththen I guess every single pre-med has ADD huh18:30
n_benthawait no18:30
n_benthashit18:30
klafka1i'm not convinced amphetamine based stimulants are useful for creative work18:30
delinquentmen_bentha, you've met the hiring criteria18:30
delinquentmeklafka1, INTERESTING18:30
delinquentmeagreed.18:30
yashgarothyeah but a lot of coding work isn't creative per se18:31
klafka1mmm fair enough, that may be why i dislike it generally.18:31
n_benthaprovigil* --made by the company cephalon18:31
delinquentmeagreed w 80% confidence18:31
klafka1oh provigil18:31
klafka1yeah18:31
klafka1modafinil18:31
n_benthayeah!18:31
yashgarothyeah I got some of that, it's pretty nice18:31
n_benthanuvigil18:31
klafka1modafinil is incredibly expensive18:31
n_benthamy insurance is 20 dollars co pay18:32
n_benthaso i get whaetver i want for 20 bucks18:32
yashgarothit's like a dollar a pill from india18:32
kanzuremy insurance doesn't cover my adderall18:32
kanzureso i pay out of pocket18:32
klafka1did your insurance approve your modafinil?18:32
n_benthai'll just tell my doc to write me a big 90-day scrips for 2x what i need18:32
klafka1what do you have?18:32
n_benthayes my insurance does18:32
klafka1they are very not keen on covering it for grey label uses18:32
n_benthablue cross blue shield18:32
n_benthaI've got a "legitimate" use for it18:32
klafka1it's white label uses fyi are narcolepsy18:32
yashgarothfor narcolepsy or do you just have a cool GP?18:32
klafka1are you narcoleptic?18:33
n_bentha-note the "'s18:33
n_benthaI've got a really awesome GP18:33
klafka1GP ?18:33
yashgarothgeneral practitioner18:33
yashgarothidk what the american term is18:33
klafka1so you know that even if a dr writes a script for it and even if the insurance covers it they can still reject it for you18:33
n_benthaGP is the american term, yashy18:34
klafka1i know because i went through this with my GF18:34
klafka1as she has MS, and provigil is sometimes prescribed grey-label for that but we ran into a huge hassle where most insurance companies won't cover it18:34
n_benthawell, I'm not going to pay more than 20 bucks for it, so we'll see how it goes18:34
n_benthaklafka1, just get your doc to change what the dx18:34
n_benthaso much can be done in this world if one monkey signs off for X18:35
klafka1um so you're saying 'tell the dr to diagnose someone with narcolepsy' ?18:35
klafka1that's illegal you know18:35
n_benthajust a couple buttons need to be pressed by X monkey and you can do Y18:35
n_benthahaha legal.18:35
yashgaroth$1.40 per pill off mymodafinil, they're pretty reliable18:36
-!- Cat4D [182bd7e2@gateway/web/freenode/ip.24.43.215.226] has joined ##hplusroadmap18:37
n_benthathat's 40+ dollars for a 30 day script18:37
klafka1yashgaroth that's pretty nice18:37
-!- aristarchus [~aristarch@unaffiliated/aristarchus] has joined ##hplusroadmap18:38
yashgarothstill cheap, especially if you need it18:38
delinquentmeso a vendor strolls into town touting "smart pills"18:38
delinquentmepeople swarm yadda ya ... " How do they work " ... " Is this really a smart pill ? " Yes they work. yes it is infact a smart pill18:39
delinquentmeone guy walks up18:39
delinquentmeTAKE MY MONEY18:39
delinquentmevendor sells him a jar ...18:39
delinquentmedude continutes to come back week after week stocking up on these smart pills18:39
delinquentmeone day he walks up to the vendor ... says " These smart pills taste a little bit like rabbit shit "18:40
delinquentmevendor: " Youre getting smarter already ! "18:40
delinquentme</share>18:40
n_benthaheh18:40
delinquentme1 ! score!18:42
klafka1lol18:42
kanzurei think i am paying $12/pill18:44
n_benthadon't do it bro18:44
katsmeow-afkfor what??18:44
kanzureadderall.18:44
-!- jmil [~jmil@c-68-81-252-40.hsd1.pa.comcast.net] has joined ##hplusroadmap18:44
katsmeow-afkouch18:45
kanzure*shrug*18:45
n_benthadayyyum18:45
katsmeow-afkcharge it to insurance18:45
kanzureinsurance doesn't cover it18:45
klafka1kanzure is it generic?18:45
kanzurealso, i'd be willing to pay up to $500/mo in pills18:45
klafka1that seems really high for generic18:45
n_benthai'll sell it to you for 5 dollars a pill?18:45
katsmeow-afkwill if a doc writes it?18:45
kanzureklafka1: no, the generic is actually a different formulation; but generic costs the same anyway18:45
n_benthastill make lots of profit18:45
n_benthake ke ke18:45
n_bentha(j/k)18:45
delinquentmedoes anyone have advice on how to estimate the market cap of a private company?18:46
delinquentmewebsite18:46
kanzurekatsmeow-afk: a doc does! heh.18:46
delinquentmenumber of employees18:46
kanzuredelinquentme: market cap only applies to publicly traded companies18:46
klafka1delinquentme the amount of money the last round of investors paid per stock share * # of shares of stock18:46
klafka1is a decent lower bound as long as the company is doing better18:46
delinquentmeyeah i mean like18:46
delinquentme" how can I estimate what they're worth "18:47
delinquentmeprivate companies18:47
klafka1 not having that specific amount of information?18:47
klafka1do you know what their revenue is?18:47
delinquentmenah18:47
klafka1it depends on what you know baout the company18:47
delinquentmelike i need to go off publically available information18:47
-!- jmil [~jmil@c-68-81-252-40.hsd1.pa.comcast.net] has quit [Client Quit]18:47
kanzureyou should stick with public biotech companies, and read their SEC filings18:47
-!- Thorbinator [~Thorbinat@c-67-166-146-169.hsd1.ca.comcast.net] has joined ##hplusroadmap18:52
kanzuregmail has screwed up all the links in my emails18:55
kanzurehttp://goog_1445275241/18:55
kanzurewhat the hell is this?18:55
katsmeow-afka google_browser shortcut,preaps18:56
kanzureanyway.. simon linked to this:18:57
kanzure<$0.50 16bit chip http://avnetexpress.avnet.com/store/em/EMController/Microcontroller/Texas-Instruments/MSP430F2011IN/_/R-4243501/A-4243501/An-0?action=part&catalogId=500201&langId=-1&storeId=500201&listIndex=-118:57
kanzureand <$5 programmer for it18:57
kanzurehttp://e2e.ti.com/group/msp430launchpad/w/default.aspx18:57
katsmeow-afkwow18:57
katsmeow-afkalso wow because i haven't heard from avnet or hamilton-avnet in forever18:58
kanzurewhen i was first looking into microcontroller programming,18:58
kanzureall the dev boards looked hiddeously expensive18:58
kanzureand i didn't want to make my own flasher at the time18:58
katsmeow-afkmany still are18:59
kanzurei was really suspicious of these >$100 dev boards18:59
kanzureso anyway. <$5 dev board makes me happy. i wonder if it works.18:59
poptirekanzure: have you really never heard of the msp430?19:01
poptireand you call yourself a hacker19:01
kanzurehaha19:01
kanzurenope i haven't used it/heard of it19:02
poptirethey've been around for a long time19:02
poptirethey're good19:02
poptirethey're cheap19:02
katsmeow-afki heard of it last year, ut prices have dropped19:02
-!- augur [~augur@c-69-250-19-178.hsd1.md.comcast.net] has quit [Ping timeout: 260 seconds]19:02
poptirethe linux toolchain is a bit crappy19:02
kanzureahh. that sucks.19:02
poptirein terms of setting it up, at least19:02
poptireonce it's set up, you're good19:02
poptirekatsmeow-afk: so far as i know, the dev board's always been $4.3019:03
katsmeow-afkany MSP430's with hardware multipliers?19:04
-!- Cat4D [182bd7e2@gateway/web/freenode/ip.24.43.215.226] has quit [Quit: Page closed]19:05
* katsmeow-afk searches around19:05
katsmeow-afkit has an external coprocessor for math19:07
kanzurewell, the z80 doesn't even do multiplication19:08
kanzurea math coprocessor would have been nice :|19:08
klafka1wow UF cut their CS dept19:08
katsmeow-afki wasn't holding it against the chip, i just wondered if i had a compelling reason to use it yet19:08
katsmeow-afkok, just chasing down one MSP430 with a hardware multiplier, the MSP430F161119:12
katsmeow-afk, it's over $12 per chip19:13
Jorayou guys building your own cpus19:13
katsmeow-afkit *is* a microcontroller, has a cpu on it19:14
Jorai meant desktops sorry19:14
AdrianGklafka1: entirely?19:14
klafka1looks like it19:15
klafka1http://www.forbes.com/sites/stevensalzberg/2012/04/22/university-of-florida-eliminates-computer-science-department-increases-athletic-budgets-hmm/19:15
katsmeow-afkno money in smart computers, no one wants to spend money on them, they pay for the roman arena model of entertainment19:16
kanzureJora: http://code.google.com/p/homecmos19:17
Joranyan cat, gotta love it19:18
-!- n_bentha [~lolicon@75.111.75.178] has left ##hplusroadmap []19:20
poptireklafka1: there's a gif for that19:23
poptirehttp://gifs.gifbin.com/florida.gif19:23
poptireand, honestly, to be competitive in the global marketplace we need big, strong dudes, not dweeby engineers19:25
poptire:P19:25
klafka1haha19:25
klafka1true19:25
klafka1it was good enough for sampson19:25
klafka1dammit19:25
poptirewho else is gonna beat up the chinese?19:25
-!- Jora [~Jora@unaffiliated/jora] has quit [Quit: Leaving]19:27
delinquentmeyashgaroth, http://sphotos.xx.fbcdn.net/hphotos-ash4/319726_306004566140259_126894987384552_702980_622646966_n.jpg19:28
delinquentmefound ur comp19:28
yashgarothI might do that actually, this micro case is really annoying19:28
klafka1Perl is ?some assembly required?. Python is ?batteries included?. PHP is ?kitchen sink, but it?s from Canada and both faucets are labeled C?.19:29
klafka1lol19:29
poptireklafka1: what would the C programming language fall under in that analogy?19:30
klafka1'completely DIY"19:31
louipchow bout lisp19:31
poptire"some assembly required"  :P19:32
poptireRe: C19:32
delinquentmeno.19:33
delinquentmephp is >> http://g-cdn.apartmenttherapy.com/234991/Useless01_rect540.jpg19:33
louipcphp slices your avocado?19:34
delinquentme:D19:34
delinquentmesorry to break the silly mode but WTF:19:35
delinquentmehttp://www.satimagingcorp.com/galleryimages/worldview-2-qatar-doha-the-pearl.jpg19:35
delinquentmethat is BEAUTIFUL and I've neber heard of / seen it before19:35
delinquentmebigger: http://hqwallpapers.eu/wallpaper/pearlqatar1_en_wikipedia_org_wiki_the_pearl_qatar_high_resolution_desktop_7000x3851_wallpaper-1007101.jpg19:35
louipcthat dubai?19:36
delinquentmequatar19:36
louipchaha they're all doing that stuff now?19:36
diginet*qatar19:36
diginetLisp is the one true language19:37
yashgarothperfect storm of oil money, absolute monarchy, and slave labor19:37
delinquentmeyashgaroth, +119:38
yashgarothI mean, ffs dubai, the point of skyscrapers is that it's cheaper to build up than out19:39
louipcit's funny how they'll have to constantly battle erosion19:39
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has quit [Ping timeout: 246 seconds]19:43
-!- ENKI-][1 [~john@c-71-232-69-59.hsd1.ma.comcast.net] has joined ##hplusroadmap19:46
-!- eudoxia [~eudoxia@r186-52-158-52.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap19:52
-!- Netsplit *.net <-> *.split quits: ENKI-][19:52
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has joined ##hplusroadmap19:55
delinquentmehttp://forums.reprap.org/read.php?178,12658419:58
delinquentmedelta printer!19:58
-!- eudoxia [~eudoxia@r186-52-158-52.dialup.adsl.anteldata.net.uy] has quit [Quit: quit]20:01
delinquentmedamn!!!!!! http://www.marginallyclever.com/2012/02/building-a-delta-robot-alternate-design/20:01
kanzureit's just a picture of a video.. what is it actually20:02
-!- _F7_ [~forrestfl@50.15.211.25] has joined ##hplusroadmap20:03
delinquentmekanzure, you cant run the video?20:08
delinquentmeits a delta 3d printer all in metal an acrylic20:08
delinquentmeits quite beautiful20:08
-!- Mariu [Jimmy98@89.41.57.33] has joined ##hplusroadmap20:09
-!- Cat4D [182bd7e2@gateway/web/freenode/ip.24.43.215.226] has joined ##hplusroadmap20:21
kanzurei'd rather not watch videos on the interwebs, yes20:26
delinquentme?? no comprendo20:31
-!- _Sketch_ [~Sketch@71-91-221-175.dhcp.stls.mo.charter.com] has quit [Read error: Connection reset by peer]20:31
delinquentmeOH MAN20:32
delinquentmejust ate so much delish health fewd20:32
-!- _Sketch_ [~Sketch@71-91-221-175.dhcp.stls.mo.charter.com] has joined ##hplusroadmap20:32
delinquentmebeans and avocado and tomatoes and chiggn'20:32
delinquentmejar20:32
kanzuredelinquentme: signal to noise ratio, man20:32
delinquentmetheres nothing going on right now20:32
delinquentmehttp://www.festo.com/net/SupportPortal/Downloads/26890/NHB_2012_en.pdf  << amazing design20:33
delinquentme... http://www.festo.com/cms/nl-be_be/16306_16336.htm#id_1633620:35
delinquentmethat is almost not real20:35
-!- nmz787 [~Nathan@cpe-67-242-177-23.rochester.res.rr.com] has quit [Quit: Leaving.]20:44
delinquentmeGOD i love robots http://www.makerbot.com/blog/2012/04/13/robot-makes-chairs-out-of-refrigerators-really/20:51
-!- diginet [~diginet@adsl-69-153-135-2.dsl.hstntx.swbell.net] has left ##hplusroadmap ["Leaving"]20:53
-!- Cat4D [182bd7e2@gateway/web/freenode/ip.24.43.215.226] has quit [Ping timeout: 245 seconds]20:57
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has joined ##hplusroadmap21:06
delinquentmekanzure, so steppers21:16
kanzuredelinquentme: reprap is not precise enough21:16
delinquentmebut what I mean is if you scale it21:16
delinquentmecould it be ?21:16
kanzureno21:16
kanzurefenn: please dump your original reasons21:16
delinquentmethere are settings for movement / revolution21:16
delinquentmein the code21:16
kanzuresoftware isn't the problem21:17
delinquentmewell yeah if i know whats the issue ill keep an eye out21:17
kanzurewhat?21:17
kanzurethat's not how you do design :P21:18
kanzureanyway, the xy stage and gantry is mostly done21:18
kanzurejust not ordered/assembled yet21:18
kanzurealso, i need to whip up a parts list for photolithography work21:18
delinquentmeI mean if its simply a matter of scale right?21:19
delinquentme( and so I can explain myself ... lets assume it is )21:19
delinquentme... deal?21:19
louipcreprap make crappy parts21:19
delinquentmeso if the normal machine is spec'd out to use acme threaded rods which are say 10mm / turn21:20
-!- diginet [~diginet@adsl-69-153-135-2.dsl.hstntx.swbell.net] has joined ##hplusroadmap21:20
delinquentmeif you leave the software as it is .. and put in 5mm/ rot you've effectively scaled that axis by .521:20
kanzurelouipc: one of our projects is a submicron resolution laser cutter for rapid prototyping of microfluidic chips21:20
delinquentmeit just seems like a really simple solution21:20
delinquentmealso louipc i think that depends on how well the printer is setup21:21
kanzurei'm not really interested in reprap's software21:21
louipckanzure: how bout laser sintering?21:21
kanzurei think linuxcnc is fine21:21
-!- d3nd3 [~dende@cpc10-croy17-2-0-cust245.croy.cable.virginmedia.com] has quit [Quit: Leaving.]21:21
kanzurelouipc: what material?21:21
kanzurebtw, if you're unfamiliar with this technology.. http://diyhpl.us/~bryan/papers2/microfluidics/21:22
delinquentmeso its a software issue?21:22
kanzureit's not a software issue21:22
diginetso, I was going some research on plant-based proteinaceous fibers21:22
-!- _F7_ [~forrestfl@50.15.211.25] has quit [Ping timeout: 252 seconds]21:22
diginetand I came across this abomination: http://heartsleevesblog.com/2011/11/15/so-silky-its-milky/ [-]21:22
diginet"According to the Daily Mail, QMilch is the first man-made fibre produced entirely without chemicals."21:22
kanzure"the daily mail" is not reputable21:23
diginetI don't know what's worse, using hte Daily Mail as a source, or not understanding that CHEMICALS ENCOMPASSES ANYTHING MADE OF ATOMS YOU DOLT21:23
kanzureand people should stop reading that site21:23
louipckanzure: I dunno, steel, aluminum, tungsten, titanium21:23
diginetoh I know21:23
kanzurelouipc: ah, no.21:23
kanzurelouipc: we're doing pdms and acrylic and maybe glass21:23
louipcfrom powdered medium21:23
diginet"But wait, there’s more.  Apparently the amino acids in the protein are antibacterial, anti-ageing and can help regulate blood circulation and body temperature."21:24
diginetwhat. . .21:24
kanzurewell, i guess i have seen sls of plastic before21:24
diginetthe stupid. . .it hurts21:24
kanzurelouipc: i haven't considered it for this case, but i'm familiar with sls21:24
delinquentmediginet, im with you21:24
delinquentmeits right up there with people not eating GMO crops21:24
diginetI'm all for sustainability, but finding information of the topic is so hard21:24
diginet*on21:25
diginetthat isn't this kind of woo21:25
delinquentmegive me a threshold of GMO and well talk21:25
diginetI'm sick ZOMG CHEMICUHLS R BAD21:25
louipckanzure: ah right21:25
diginetanyway21:26
louipcdelinquentme: I think the problem with the rep rap is the process itself.. it's extrusion21:26
diginetI had sort of a cool idea, as a side project to spider silk.  Apparently, there are fibers made from reconsituted soy protein extract21:26
diginetso they sort of mirror the properties of animal fibers21:26
-!- _0bitcount [~ulises11@213.37.254.87.dyn.user.ono.com] has quit [Remote host closed the connection]21:26
delinquentmelouipc, i dont follow21:27
delinquentmeyou're saying by definition the process has issues?21:27
diginetSince duckweed has similar protein content as soy does, it should be possible to make fibers from duckweed protein extract by a similar process21:27
louipcyeah you're just dragging about a long noodle21:27
delinquentmelouipc, I still dont follow21:27
delinquentmeyou're saying that that cant work structurally?21:28
delinquentmeand for kanzure application im simply saying to build upon the gantry controls21:28
louipcwell not if you want to build layers the way it's being done21:28
kanzure10 micron 3D printing isn't really useful.. just use photolithography like any other sane person21:28
louipcif you were pushing the plastic through a die to form an extruded shape that would be fine21:28
diginetIMO, the best 3d printer would work like a laser printer21:29
kanzurethe reprap gantry is not useful for this21:29
louipcsame is done with aluminum in industry21:29
diginetreplace the toner with the target material21:29
diginetprint in layers21:29
delinquentmelouipc, I think the real question here is .. what tolerances can this print tech meet21:29
delinquentmesure im not making planes out of it21:29
delinquentmebut can it serve to make chess pieces? sure21:29
louipcdelinquentme: yeah but look at the surface finish21:29
kanzurechess pieces are thousands of times larger than what we're building21:30
delinquentmehow about something more complex / structurally testing21:30
delinquentmekanzure, that was @ louipc21:30
kanzureyes i know21:30
delinquentmewhat about it? its got streaks21:30
delinquentmecan that work reliably as a coat hanger?21:30
delinquentmepeople make reliable whistles out of it21:31
delinquentmeso the trick is just finding the limits or the threshold of application its useful for21:31
delinquentmedont build monster truck roll cages21:31
delinquentmeinstead build toy car bumpers21:31
delinquentmekanzure, i just wanted to make sure you knew that scaling could be done in software21:31
louipcok.. but are you gonna want one machine for cheap stuff, then one machine to do better stuff?21:31
kanzurehaha don't worry about me and software :321:32
delinquentmeso long as you got that thats all I really needed to convey :D21:32
delinquentmecheck21:32
louipcor just one machine period21:32
delinquentmelouipc, what am I building?21:32
delinquentmeare there cost efficiencies in 1 machine over the other?21:32
delinquentme( probably )21:32
delinquentmecost or other efficiencies21:32
louipcisn't the point of reprap to print whatever you want/need at the moment?21:32
louipcit costs less to have one machine than two machines21:33
delinquentmerepraps you can get in smalle resolution .. but the same volume takes longer to print21:33
louipcunless you really need two to increase your production21:33
delinquentmehigher resolution ... with a smaller filament21:33
delinquentmeso then yeah its about the application right21:33
louipcsure, but what if you can use a machine for both applications21:34
louipcand whatever tolerance/rigidity requirements you need you set for that application21:35
louipcthat ribbed finish is a structural weakness, and it'll always be there with extrusion21:36
-!- Cat4D [182bd7e2@gateway/web/freenode/ip.24.43.215.226] has joined ##hplusroadmap21:39
-!- _F7_ [~forrestfl@50.15.211.25] has joined ##hplusroadmap21:39
kanzurehi _F7_21:43
-!- srangewarp [~strangewa@c-76-25-200-47.hsd1.co.comcast.net] has joined ##hplusroadmap21:43
delinquentmelouipc, true @ the finish21:43
-!- rkos_ [~chatzilla@a88-113-144-253.elisa-laajakaista.fi] has joined ##hplusroadmap21:43
delinquentmebut thats like saying ... dont use concrete .. it cant handle extension forces21:43
delinquentmeim saying " Pick an appropriate application "21:44
louipcdelinquentme: but yeah I guess it's for fun stuff for now...21:44
delinquentme=]21:44
delinquentmeyeah21:44
delinquentmeso far21:44
louipcguess it'll help people work out other aspects of the machine21:44
-!- strangewarp [~strangewa@c-76-25-200-47.hsd1.co.comcast.net] has quit [Read error: Connection reset by peer]21:44
-!- rkos [~chatzilla@a88-113-144-253.elisa-laajakaista.fi] has quit [Read error: Connection reset by peer]21:44
-!- srangewarp is now known as strangewarp21:44
louipcuntil they figure out a better additive method21:44
louipc:P21:44
-!- rkos_ is now known as rkos21:44
delinquentmelouipc, there is actually a funded startup whos working on cheap accurate 3d printing21:47
-!- nuba_ [~nuba@pauleira.com] has joined ##hplusroadmap21:53
-!- Netsplit *.net <-> *.split quits: azonenberg, yashgaroth, kristianpaul, kanzure, strages_home, nuba, qnm21:57
louipcdelinquentme: who's that?21:58
-!- Netsplit over, joins: kanzure21:58
delinquentmeform something21:59
delinquentmeALRIGHT22:02
delinquentmelater all22:02
-!- delinquentme [~asdfasdf@c-24-3-85-154.hsd1.pa.comcast.net] has quit [Quit: Leaving]22:02
kanzurehttp://phdcomics.com/movie/ but sadly they are charging for it22:02
-!- AdrianG [~amphetami@unaffiliated/amphetamine] has quit [Ping timeout: 246 seconds]22:08
kanzure"anatomy of a professor's homepage" http://www.phdcomics.com/comics/archive.php?comicid=148722:10
-!- strages_home [~strages@adsl-98-81-13-133.hsv.bellsouth.net] has joined ##hplusroadmap22:12
kanzure"reported sightings of our professor" is also strangely accurate http://www.phdcomics.com/comics/archive.php?comicid=148222:12
-!- yashgaroth [~f@cpe-66-27-117-179.san.res.rr.com] has joined ##hplusroadmap22:14
-!- kristianpaul [~kristianp@unaffiliated/kristianpaul] has joined ##hplusroadmap22:14
-!- qnm [~qnm@2001:44b8:3110:f300:208:9bff:fec0:179a] has joined ##hplusroadmap22:14
-!- azonenberg [~azonenber@2001:470:888b:2:206:70ff:fe01:46] has joined ##hplusroadmap22:16
-!- sylph_mako [~mako@118-93-18-64.dsl.dyn.ihug.co.nz] has joined ##hplusroadmap22:19
Mokbortolan_http://www.eurekalert.org/pub_releases/2012-04/uos-ntm041912.php22:36
Mokbortolan_the team describes a new method for making medical devices called nerve guidance conduits or NGCs.22:36
-!- Mariu [Jimmy98@89.41.57.33] has quit [Ping timeout: 265 seconds]22:38
Mokbortolan_so, could you just "hijack" a nerve with this?22:39
Mokbortolan_and make it go somewhere else entirely22:39
Mokbortolan_ooh22:43
Mokbortolan_I wonder if you could a channel where they'd have to grow through fine channels with some kind of monitoring ability22:43
Mokbortolan_^make22:44
Mokbortolan_you could just make a cut-out22:44
Mokbortolan_though, you'd have to be reeaaally sure everything gets back together properly...22:45
diginetis there any way to attack the side group in amino acid residues to convert them to glycine?22:48
diginet(that's practical)22:48
-!- augur [~augur@208.58.5.87] has joined ##hplusroadmap22:54
kanzurehttp://www.sciencescape.org/news/launchstory23:23
kanzure"21 million papers in 25,000 journals"23:23
kanzuremeh just a landing page23:23
kanzurenevermind23:23
kanzurehttp://www.pythonanywhere.com/ random python shells23:24
-!- _F7_ [~forrestfl@50.15.211.25] has quit [Ping timeout: 260 seconds]23:25
-!- _F7_ [~forrestfl@50.15.211.25] has joined ##hplusroadmap23:26
diginetkanzure, I had an idea I want to try on spider silk, that, if it were to work, would make the production of it super simple23:29
diginetbasically, I've been reading a little more, and apparently, according to several papers, there is a large constituency of people who argue that its properties lie more in the spinning/extrusion process than the exact sequence23:30
diginetanyhow, if the sequence only has to be approximate, I could put glycine and trace amino acids in a vat, and then alanine in another vat, and follow protocols for peptide synthesis, but, since the make up, not the sequence is important, I can just let it happen, then after glycine rich chains and alanine chains have been built up independently, combine the two23:32
diginetlet it happen, without worrying about protection for ensuring a specific sequence23:32
diginetif this works, make the spider silk would be as simple as buying whey extract at a health food store23:33
Mokbortolan_or you could glue it to a spider's abdomen and use its ductwork as a print head23:33
Mokbortolan_a tube23:33
diginethahahahaahah23:33
diginetsomething tells me that wouldn't work so well :P23:33
Mokbortolan_it would probably resent it23:33
diginethah23:34
diginetanyhow, I23:34
diginetI'm trying to find a way to chemically induce hydrolysis of the peptide bonds so I can get individual amino acids23:35
yashgarothoh god there's so many things wrong here I don't know where to begin23:37
diginet?23:38
yashgarothsequence is still important even if exact sequence isn't23:39
-!- _F7_ [~forrestfl@50.15.211.25] has quit [Quit: Computer has gone to sleep]23:39
diginetof course23:39
diginetbut I'm talking on the scale of the individual "sentences"23:39
yashgarothso am I23:39
kanzureyashgaroth: i'm about to sleep, but let me know if i should /kick diginet out of the channel for being too wrong23:39
yashgarothno no at least he's not a conspiracy nutjob23:40
diginetam I disliked now?23:40
kanzurediginet: i hate everyone equally23:40
diginetconfused23:40
yashgarothit's just you seem to completely change your approach every three hours23:40
kanzure'night23:41
diginetyashgaroth, oh I know, it's my OCD kicking in (I hate using that, because the term has been so marginalized, but it's true). I always second guess myself23:41
diginetI know I probably sound horribly harebrained and nutty23:42
yashgarothlike you're going down the list of options and deciding the easier ones won't work, but based on maybe a paper you read or a website blurb23:42
diginetthat's a bit of an exagerration, I've only been reading published papers, but I see what you mean23:43
yashgarothtry insect cell expression first, or even just dissolving spider webs and trying to reconstitute them23:44
yashgaroththere's a reason no one uses insect expression or peptide synthesis unless absolutely necessary23:45
yashgarotherr, in vitro expression not insect*23:45
klafka1diginet i obviously have no idea but i suspect even if hte sequence of peptides isn't that important the folding and conformation of them is, as well as the general distribution of various sidechains/charges/etc...23:46
diginetI guess my problem is obsession with having every detail thoroughly described in a process. Culturing cells is obviously not like performing a chemical synthesis with stoichiometric amounts23:46
diginetklafka1, oh yeah definitely23:47
yashgarothhahaha if you want every detail, biology is deffo not for you23:47
diginethahaha yep23:47
kanzureyour first plan will absolutely fail23:47
kanzureand you will debug it about 100 times before it works23:48
kanzurethis is how modern biology works :/23:48
diginetyeah :(23:48
yashgarothwe're reverse engineering an immensely complex system, and it's barely even robust23:48
diginetI was reading obsessively about aeroponics a couple months ago, I just couldn't deal with the fact that the parameter space is not even vaguelly fully fleshed out23:49
klafka1modern biology in a nutshell - a fuckton of assumptions to find a correlation that will go in a pile that will hopefully be big enough to one day produce reasonable evidence that there is some kind of causation23:49
diginethahah23:49
diginetvery true23:50
* kanzure sleepz23:50
diginetI guess I'm just a weirdo, lol23:50
diginetanyway, regardless of whether it turns out as I would want, I still think regenerated protein fibers are interesting23:52
klafka1what does regenerated mean?23:52
diginetwell for example the protein from soybeans in extracted and spun into a fiber that's called "soybean silk"23:53
yashgarothyou mean native soybean proteins, or transgenic spider silk components?23:54
diginetno, native soybean proteins23:54
yashgarothwell sure if you denature any proteins and extrude them, they'll stick together in a string23:55
diginetthey do more than that, but yeah, that's the general idea23:55
diginetunfortunately they kind of suck23:55
yashgarothno they suck terribly, compared to a keratin or collagen or any other protein that's evolved to be useful in that shape23:56
diginetyeah23:58
yashgaroththe secondary structure is the important part - even if the interstitial portions can vary in size and composition, you still need defined segments that stick between the proteins23:58
diginetwhat I'm saying is that, the basic structure of spider silk is strings of alanines interspersed among amorphous glycine rich regions, so shouldn't it be possible to randomly grow the glycine rich chains and alanine chains in seperate reactors, then combine the mixture so that the two peptide varities will attach to one another to form longer peptides?23:59
--- Log closed Mon Apr 23 00:00:10 2012

Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!