--- Log opened Fri Aug 02 00:00:43 2013 --- Day changed Fri Aug 02 2013 00:00 -!- poppingtonic [~poppingto@212.49.88.100] has joined ##hplusroadmap 00:03 -!- klafka [~klafka@c-24-6-18-31.hsd1.ca.comcast.net] has joined ##hplusroadmap 00:08 -!- klafka [~klafka@c-24-6-18-31.hsd1.ca.comcast.net] has quit [Ping timeout: 268 seconds] 00:46 -!- spresser [uid6132@gateway/web/irccloud.com/x-frvalrhtthqcgxee] has quit [Ping timeout: 245 seconds] 00:58 -!- poppingtonic [~poppingto@212.49.88.100] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 00:59 <@nmz787> superkuh: what comment got removed from http://travisgoodspeed.blogspot.com/2013/07/hillbilly-tracking-of-low-earth-orbit.html 01:20 -!- upgrayeddd [uid2969@gateway/web/irccloud.com/x-kplvglgyzjeqihoy] has joined ##hplusroadmap 01:34 -!- ThomasEgi [~thomas@panda3d/ThomasEgi] has joined ##hplusroadmap 01:55 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has quit [Read error: Connection reset by peer] 01:55 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has joined ##hplusroadmap 01:55 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has quit [Changing host] 01:55 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 02:14 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has quit [Read error: Connection reset by peer] 02:15 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has joined ##hplusroadmap 02:15 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has quit [Changing host] 02:15 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 02:20 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has quit [Ping timeout: 268 seconds] 02:32 -!- rigel [~yourmom@c-76-105-237-98.hsd1.or.comcast.net] has quit [Ping timeout: 268 seconds] 02:33 -!- rigel [~yourmom@c-76-105-237-98.hsd1.or.comcast.net] has joined ##hplusroadmap 02:44 -!- spresser [uid6132@gateway/web/irccloud.com/x-labqogdflsupuiuh] has joined ##hplusroadmap 02:46 -!- chido [chidori@pasky.or.cz] has quit [Ping timeout: 245 seconds] 02:58 -!- Jaakko97 [~Jaakko@cpc13-newc15-2-0-cust64.16-2.cable.virginmedia.com] has joined ##hplusroadmap 02:59 -!- Jaakko97 [~Jaakko@cpc13-newc15-2-0-cust64.16-2.cable.virginmedia.com] has quit [Client Quit] 03:05 < superkuh> I made a mistake. I misread and thought he was using a DVB-S card to send Diseqc commands. So I asked which it was. On a second read through I realized that he was talking about someone else, and he did not use Diseqc motors. 03:05 < superkuh> So I removed my comments. 03:05 < superkuh> -s 03:06 < superkuh> Doesn't matter anymore. I have bought a DVB-S card (skystar 2 HD) and it seems to work for sending positioning commands to my Diseqc motors. 03:09 -!- EnLilaSko [EnLilaSko@unaffiliated/enlilasko] has joined ##hplusroadmap 03:17 -!- Adillian [~Adillian@124.43.125.14] has quit [Ping timeout: 240 seconds] 03:38 -!- poppingtonic [~poppingto@212.49.88.101] has joined ##hplusroadmap 03:54 -!- weles [~mariusz@c-71-234-3-169.hsd1.ct.comcast.net] has quit [Ping timeout: 260 seconds] 03:58 -!- weles [~mariusz@c-71-234-3-169.hsd1.ct.comcast.net] has joined ##hplusroadmap 04:03 -!- anannie [~chatzilla@unaffiliated/anannie] has quit [Ping timeout: 245 seconds] 04:08 -!- makoLime [~mako@103-9-42-133.flip.co.nz] has quit [Ping timeout: 264 seconds] 04:31 -!- poppingtonic [~poppingto@212.49.88.101] has quit [Ping timeout: 264 seconds] 05:01 -!- weles [~mariusz@c-71-234-3-169.hsd1.ct.comcast.net] has quit [Ping timeout: 264 seconds] 05:20 -!- weles [~mariusz@wsip-70-183-164-169.ri.ri.cox.net] has joined ##hplusroadmap 05:21 -!- yorick [~yorick@oftn/member/yorick] has joined ##hplusroadmap 05:29 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has joined ##hplusroadmap 05:29 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has quit [Changing host] 05:29 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 05:33 -!- poppingtonic [~poppingto@212.49.88.104] has joined ##hplusroadmap 06:00 -!- phillyj [~Thunderbi@pool-108-36-3-19.phlapa.east.verizon.net] has joined ##hplusroadmap 06:03 -!- phillyj [~Thunderbi@pool-108-36-3-19.phlapa.east.verizon.net] has quit [Client Quit] 06:50 -!- pads [~not@100.43.114.90] has joined ##hplusroadmap 06:51 -!- pads is now known as Guest10542 06:53 -!- padz [~not@100.43.114.90] has quit [Ping timeout: 248 seconds] 07:07 -!- nsh is now known as MISTARU_ANATORMU 07:07 -!- MISTARU_ANATORMU is now known as nsh 07:33 -!- klafka [~klafka@c-24-6-18-31.hsd1.ca.comcast.net] has joined ##hplusroadmap 07:35 <@kanzure> fredox: august 8 is http://www.meetup.com/emergence-24/events/120263392/ (yes the topic might seem boring, but the people aren't) 07:37 < fredox> very vague that group 07:37 <@kanzure> they are a bunch of old perl hackers turned scifi authors 07:38 < gradstudentbot> Yeah, there's a clear trend. 07:46 -!- ThomasEgi [~thomas@panda3d/ThomasEgi] has quit [Remote host closed the connection] 07:50 < klafka> what? 07:50 <@kanzure> huh? 07:50 < klafka> like who? 07:50 < klafka> old perl hackers turned scifi authors 07:51 <@kanzure> what do you mean who? check the page. 07:51 <@kanzure> it's nobody you know 07:58 < klafka> oh 07:58 < klafka> well i mean they could be well known scifi authors! 07:58 <@kanzure> they could be.. but they aren't. hah. 07:59 < klafka> yeah they aren't 08:35 -!- nicedice [~nicedice@unaffiliated/nicedice] has joined ##hplusroadmap 08:39 -!- cpopell_ [47fff18b@gateway/web/freenode/ip.71.255.241.139] has joined ##hplusroadmap 09:25 -!- tomkinsc [~tomkinsc@24.4.11.202] has joined ##hplusroadmap 09:42 -!- weles [~mariusz@wsip-70-183-164-169.ri.ri.cox.net] has quit [Quit: Leaving] 10:26 -!- ThomasEgi [~thomas@panda3d/ThomasEgi] has joined ##hplusroadmap 10:32 < klafka> btw 'in our dna' is a new techy term i hate 10:33 <@kanzure> why? we have sequenced your dna and now we know things about it. 10:34 < fredox> the 'our' part bothers especially bothers me 10:35 <@kanzure> "we have sequenced 1000 people and we're pretty sure that we all have dna polymerase variants in our dna".. sounds right to me. 10:36 < fredox> its time to revive phrenology 10:39 -!- AshleyWaffle [~quassel@unaffiliated/anastasiawyatt] has quit [Read error: Connection reset by peer] 10:40 < klafka> um 10:40 < klafka> my problem is like 10:40 < klafka> 'oh we have sales in our dna' 10:40 < klafka> 'ad networks are really in our dna' 10:40 < klafka> 'big data is in our dna' 10:40 < klafka> 'in our dna' is the new 'wheelhouse' 10:41 <@kanzure> ah i see. yeah, that's dumb. 10:44 < klafka> btw it seems like more and more poeple are using golang 10:45 <@kanzure> sure. it seems fun. 10:46 < klafka> wow i did not know trulia ipo'd 10:47 < klafka> and has a market cap of 1.4b 10:47 < klafka> damn 10:47 < ParahSailin> golang is probably a step forward from using c for things and probably a step backward from high level languages 10:48 < klafka> but it's not the 10 steps backward c++ is 10:49 < klafka> also it seems to be really good at concurrency 10:49 < cpopell_> kanzure: new emotiv revealed today on kickstarter-know if it can follow anything more than 4 'thoughts' at a time? 10:49 < cpopell_> I haven't been staying up to date on what they do and their market releases are always cagey 10:49 <@kanzure> emotiv has been spamming me for weeks about it. let me know if you want their firmware to double check their claims. 10:49 <@kanzure> imho it's not worth your attention 10:50 < cpopell_> figured as much. 10:50 < cpopell_> Alas 10:52 <@kanzure> i wonder if emokit works for it though 10:52 <@kanzure> i should probably go crack their keys 10:52 <@kanzure> blah why does it have to be me 11:03 -!- tomkinsc [~tomkinsc@24.4.11.202] has quit [Quit: My computer has gone to sleep. ZZZzzz…] 11:09 < klafka> what is the cheapest single sensor eeg ? 11:10 < klafka> i've been thinking about just using cheap eegs to interject random 'crowd noise' into reactive art 11:11 <@kanzure> openeeg, neurosky, things like that 11:11 <@kanzure> they probably have links to cheap sens0rs 11:17 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has quit [Read error: Connection reset by peer] 11:17 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has joined ##hplusroadmap 11:17 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has quit [Changing host] 11:17 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 11:26 < cpopell_> god the patent office is some stupid shit 11:26 < cpopell_> some chinese firm patent DLP/DMD projection. in 2008 >_> 11:27 < cpopell_> (you know, the thing TI had had patented since ~94 or so) 11:29 <@kanzure> only the claims matter, so maybe their specific combination of claims is different 11:30 <@kanzure> nmz787: we can probably just take a dna synthesizer patent. surely there might be one that is expired by now. i wonder if patents from the 80s had anything useful in them. 11:31 < cpopell_> I have a friend that recycles patents from the 80s that were never followed through on due to lack of processing power (espec. declass cold war stuff) to be 'innovative' at work 11:32 <@kanzure> makes sense to me 11:37 -!- tomkinsc [~tomkinsc@c-24-4-11-202.hsd1.ca.comcast.net] has joined ##hplusroadmap 11:38 * heath waves hello to everyone 11:47 <@heath> cpopell_: it takes patent examinars ~30 minutes for one query 11:48 < cpopell_> am aware. was working on figuring out tools to help with that, but it's backburnered atm 11:48 <@heath> and it isn't going to be fixed, we (open source connections and isotope11) were subcontracted out to fix this and then the budget cuts came in 11:48 <@heath> the project had ~2 years of work 11:48 <@heath> was coming to a close... aaaand scratched 11:52 -!- Urchin [~urchin@unaffiliated/urchin] has joined ##hplusroadmap 12:05 <@nmz787> kanzure: maybe, but the academic articles seem to cover things pretty well, I'm not one to ignore info though if some folks in here dig up some 12:05 <@kanzure> i bet there's even some expired patents that had some crazy designs that nobody commercialized 12:05 -!- pasky_ is now known as pasky 12:05 <@kanzure> for synthesizers. 12:05 <@kanzure> "Nobody would want to make that much DNA" 12:16 <@nmz787> lol 12:19 <@kanzure> "Nobody needs more than 128 kb bp" 12:19 <@kanzure> erm, i mean 128 kbp 12:20 -!- Netsplit *.net <-> *.split quits: upgrayeddd 12:20 < gradstudentbot> Yeah, I don't know. 12:23 <@nmz787> i worked on cleaning code with jslint yesterday, kinda need a break 12:23 <@nmz787> so working on synthesis a bit 12:23 <@nmz787> also backing up files 12:24 <@kanzure> also try pyflake or pylint or flake8 sometime for python things. 12:25 <@nmz787> for some reason my js code seems to break the jQuery uploader template stuff (which it uses to add new uploads to the displayed queue) 13:02 <@nmz787> fenn: so what do you know about beta agonists (since you mentioned blockers yesterday)? https://en.wikipedia.org/wiki/Ractopamine 13:10 -!- tomkinsc [~tomkinsc@c-24-4-11-202.hsd1.ca.comcast.net] has quit [Quit: http://takeitapart.com http://cht-t.com] 13:11 <@nmz787> kanzure: so macro or meso synthesizers need mixers or shakers or something, otherwise diffusion will take a long long time 13:11 <@nmz787> paperbot: http://nar.oxfordjournals.org/content/23/6/982.full.pdf 13:11 < paperbot> http://libgen.org/scimag/get.php?doi=10.1093%2Fnar%2F23.6.982 13:12 <@heath> nmz787: what's your js project? 13:13 <@nmz787> just an uploader page right now 13:16 <@nmz787> something in my code is broken though 13:17 <@nmz787> heath: this is the demo page, if you add a file an entry is created in the queue http://blueimp.github.io/jQuery-File-Upload/ 13:17 <@nmz787> for me though, it doesn't happen http://dev.takeitapart.com:8002/media/jQuery-File-Upload/index_TIA.html 13:20 <@heath> nmz787: are you receiving any error in your console? 13:21 <@nmz787> no :( 13:22 <@heath> what about in the network tab, are there some requests not returning 200? 13:24 <@nmz787> there's one, but it seems unrelated and happens before i add the file 13:24 <@nmz787> i see the file show up in the add callback for the uploader 13:24 <@nmz787> i'm logging the name 13:31 -!- tomkinsc [~tomkinsc@c-24-4-11-202.hsd1.ca.comcast.net] has joined ##hplusroadmap 13:46 < ParahSailin> is there a more concise python way to manipulate DOM than etree? 13:47 < ParahSailin> this is quite a bit more verbose than im used to in jquery 13:49 < tomkinsc> xml.dom.minidom may be a bit more concise, but etree is arguably a better solution 13:49 < tomkinsc> google also pointed to this: https://pypi.python.org/pypi/pyquery 13:50 < ParahSailin> this makes me smile 13:54 < tomkinsc> if you are concerned at all about memory use, you might take a look at xml.dom.pulldom as well 13:58 < ParahSailin> i just dont want to type very much generating simple reports 13:58 < ParahSailin> thanks, pyquery is nice 14:07 -!- ParaSa1lin [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has joined ##hplusroadmap 14:07 -!- ParaSa1lin [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has quit [Changing host] 14:07 -!- ParaSa1lin [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 14:10 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has quit [Ping timeout: 276 seconds] 14:10 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has joined ##hplusroadmap 14:10 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has quit [Changing host] 14:10 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 14:12 <@kanzure> ParahSail1n: from gi.repository import webkit... the bulldozer of dom manipulation. 14:12 <@kanzure> nmz787: can we just use a laser to heat it 14:12 -!- ParaSa1lin [~parahsail@unaffiliated/parahsailin] has quit [Ping timeout: 245 seconds] 14:13 -!- ParaSa1lin [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 14:15 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has quit [Ping timeout: 245 seconds] 14:16 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 14:18 <@nmz787> kanzure: heat cycling isn't really a part of this AFAIK 14:18 <@nmz787> for PCR sure! 14:18 <@kanzure> shrug, okay 14:19 -!- ParaSa1lin [~parahsail@unaffiliated/parahsailin] has quit [Ping timeout: 264 seconds] 14:20 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has quit [Ping timeout: 240 seconds] 14:20 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has joined ##hplusroadmap 14:20 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has quit [Changing host] 14:20 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 14:22 -!- upgrayeddd [uid2969@gateway/web/irccloud.com/x-pdylasqhgkjcusds] has joined ##hplusroadmap 14:23 -!- ParaSa1lin [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has joined ##hplusroadmap 14:23 -!- ParaSa1lin [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has quit [Changing host] 14:23 -!- ParaSa1lin [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 14:26 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has quit [Ping timeout: 245 seconds] 14:28 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has joined ##hplusroadmap 14:28 -!- ParahSail1n [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has quit [Changing host] 14:28 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 14:29 -!- ParaSa1lin [~parahsail@unaffiliated/parahsailin] has quit [Ping timeout: 268 seconds] 14:31 -!- ParaSa1lin [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has joined ##hplusroadmap 14:31 -!- ParaSa1lin [~parahsail@99-25-202-211.lightspeed.hstntx.sbcglobal.net] has quit [Changing host] 14:31 -!- ParaSa1lin [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 14:33 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has quit [Ping timeout: 256 seconds] 14:35 -!- nicedice [~nicedice@unaffiliated/nicedice] has quit [Quit: Leaving] 14:41 -!- ParaSa1lin [~parahsail@unaffiliated/parahsailin] has quit [Ping timeout: 276 seconds] 14:47 -!- ParahSail1n [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap 15:03 -!- cpopell_ [47fff18b@gateway/web/freenode/ip.71.255.241.139] has quit [Ping timeout: 250 seconds] 15:18 <@nmz787> paperbot: http://pubs.rsc.org/en/Content/ArticleLanding/2012/LC/C2LC40098G 15:18 < paperbot> http://diyhpl.us/~bryan/papers2/paperbot/Cell-free%20protein%20synthesis%20from%20a%20single%20copy%20of%20DNA%20in%20a%20glass%20microchamber.pdf 16:15 -!- augur_ [~augur@c-71-57-182-133.hsd1.fl.comcast.net] has joined ##hplusroadmap 16:15 -!- tomkinsc [~tomkinsc@c-24-4-11-202.hsd1.ca.comcast.net] has quit [Quit: http://takeitapart.com http://cht-t.com] 16:16 -!- augur [~augur@c-71-57-182-133.hsd1.fl.comcast.net] has quit [Ping timeout: 264 seconds] 16:34 -!- valyap [~emankcin@c-76-23-254-105.hsd1.ct.comcast.net] has joined ##hplusroadmap 16:43 <@kanzure> "Glowing Plant Project is safe but Kickstarter bans future project creators from giving away genetically-modified organisms:" 16:43 <@kanzure> http://www.theverge.com/2013/8/2/4583562/kickstarter-bans-project-creators-from-giving-GMO-rewards 16:43 <@kanzure> guess we need another crowd funding site for genetic engineering (ugh) 16:43 <@kanzure> and not microryza, they haven't really been that successful 16:47 < ParahSail1n> lol, man that deloitte scumbag made out with all the loot 16:47 < cpopell> hm? 16:47 < ParahSail1n> .title 16:47 < yoleaux> Kickstarter bans project creators from giving away genetically-modified organisms | The Verge 16:48 < cpopell> no, I mean the deloitte part 16:48 <@kanzure> just read the logs 16:48 < ParahSail1n> oh the guy who pulled off that scam was from bain or mckinsey, some such 16:49 < cpopell> oh the whole thing was a scam? 16:49 <@kanzure> i'm p. sure it was just a genome compiler marketing stunt.. i mean.. they're the ones that keep scratching each other's backs. 16:51 < ParahSail1n> i used genome compiler on the grc build 37 with -O2 and it built a lisp interpreter 16:51 <@kanzure> what? 16:56 <@kanzure> nmz787: so, i don't think we need $2000 for a 4-channel oscilloscope. that seems like a ripoff. 16:57 <@kanzure> nmz787: i also think $600/syringe pump seems a little steep.. 16:57 <@kanzure> and if an hplc really costs $20k then we should work on building a cheaper one 16:57 <@kanzure> instead of this other stuff 16:58 <@kanzure> and why are we paying $350/hour x 50 hours to a random FIB shop 16:58 <@kanzure> the balance is a sane thing to want, but why on earth would anyone want it to be RS232? 16:59 < ParahSail1n> god, hplc... our lab's one broke down so much 16:59 <@kanzure> i think "enzmyes, $500" is disingenous. it seems a little random. what enzymes are we playing with? 17:00 <@kanzure> taq? and we'd only need $500 of taq for.. how many runs? 17:00 <@kanzure> https://docs.google.com/spreadsheet/ccc?key=0AkTsdtdxo56DdDFCUmNqMzVyYi04eGw4cUJ6bG1KMlE#gid=0 17:01 < ParahSail1n> depending on what you're doing, the enzymes can really add up 17:01 <@kanzure> sure 17:01 <@kanzure> i'd expect more than $500 17:02 < ParahSail1n> damn 50 hours of fib? 17:02 <@kanzure> yeah i have no idea where that's coming from 17:03 < ParahSail1n> if you have to use fib, you're probably not coming up with a very scalable fabrication protocol 17:04 < ParahSail1n> amidites: itemized; enzymes: not itemized 17:04 < ParahSail1n> heh 17:04 <@kanzure> right. there's probably some metalwork that has to be considered. or maybe we can just 3d print some junk to hold everything together. i'm not sure. 17:04 <@kanzure> also tubing.. 17:06 < ParahSail1n> save on the balance cost; put that into better thermocycler 17:06 <@kanzure> i really like the oligomaker's design. 192 channels is a good intermediate between "a pippetor on an arm moving around" and "no moving parts" and "a microarray with a bajillion wells that you can't easily clean or sequence". 17:07 <@kanzure> i mean, the way the oligomaker lays out the different channels in a circular design 17:07 < ParahSail1n> thermocycler block temperature consistency avoids frustration 17:08 <@kanzure> yes, it is less frustrating when your shit works :) 17:08 <@kanzure> that's a universal truth 17:08 < ParahSail1n> decent gradient thermocycler will really save you time 17:10 <@kanzure> if you can't get a decent hplc for less than 20k then i think we should solve that problem too 17:16 <@nmz787> $2k is cheap for a 4 channel scope 17:16 <@nmz787> chinese cheap 17:16 <@nmz787> new usa models are easily $4k-5k 17:17 <@kanzure> what about just getting a bunch of sound cards 17:17 <@nmz787> ha ha ha 17:17 <@nmz787> funny 17:17 <@nmz787> (not) 17:17 <@nmz787> you want an oscope to be 'the right answer' 'the truth' 17:17 <@nmz787> not some hobbled together thing 17:19 <@nmz787> 50 hours of fib is cheaper than buying a fib 17:20 <@kanzure> what are we using a FIB for again? 17:20 <@nmz787> honestly though it seems like from their milling rate spec, it'd be quite cheap to fab things with them in production 17:21 <@nmz787> microfluidics 17:21 <@nmz787> or meso fluidics 17:21 < ParahSail1n> fib is for making nano holes in things 17:21 <@nmz787> nah these dudes use plasma ion source 17:21 < ParahSail1n> for micro and meso, you do photolithography 17:21 <@nmz787> not like most 17:21 < ParahSail1n> because fib is 300/hr 17:22 <@nmz787> http://www.oregon-physics.com/lab_services/fib_micromachining_milling.html 17:23 <@nmz787> so at 1 micron spot, they do 100 microns^3 per second 17:23 < ParahSail1n> or you can do photolithography and wet/dry etch 17:23 <@nmz787> sure 17:24 <@nmz787> i have those capabilities at a local lab for my time plus $20.hr 17:24 <@kanzure> what do you estimate making an hplc would be like 17:24 <@kanzure> because 20k is sutpid and it's an important thing 17:24 < gradstudentbot> Hey, I got 100% yield! Oh wait, no. 17:24 <@kanzure> at 20k it might be cheaper to just ship out samples to some other guy 17:25 <@kanzure> we could use science exchange hah 17:25 < ParahSail1n> that might be better than "amigad, a salt bubble in the hplc and it sploded' 17:25 -!- yorick [~yorick@oftn/member/yorick] has quit [Remote host closed the connection] 17:26 <@kanzure> ParahSail1n: maybe i should make you do the BOM 17:27 < ParahSail1n> well i think you should just stack up a whole bunch of chairs 17:27 < ParahSail1n> this is for a space program, right? 17:28 <@kanzure> sort of 17:29 < ParahSail1n> oligo synthesis? 17:30 < ParahSail1n> oligomaker clone? 17:30 <@kanzure> oligo synthesis 17:30 <@kanzure> design not determined yet. "not micron". 17:31 < ParahSail1n> is this intended to be part of a long sequence synthesis system? 17:32 <@kanzure> undecided. possibly no. my original thought was that we just need something to prove that it can be done for less than 50k or 100k or whatever in parts. 17:33 <@kanzure> ideally the answer is yes.. hell yeah i want oligos as long as fucking possible. but you can't just start at a bajillion megabases. 17:34 <@kanzure> and plus, you don't need a thousand megabases to prove that your bill of materials (well, parts) was less than $50k for a single unit. 17:35 <@kanzure> i figure that if a single unit is cheap enough you can just iterate on the design to possibly get longer oligos, or refactor the design later to try for something a little more expensive but synthesizes longer stuff. 17:35 <@kanzure> it's entirely possible that i'm wrong, though 17:35 < ParahSail1n> cambrian was doing synthesis on microbeads or something? 17:36 <@kanzure> indeterminable. they are hush hush plus i don't have a copy of their slide where they showed a picture of their setup. i think they are probably using beads somewhere, and then lasers to sample the beads either for fluorescence or for popping the beads out of wells (maybe) or something else.. 17:37 <@kanzure> i don't want to start with something extremely complex that wont work 17:39 <@kanzure> i don't entirely care about what their design is. i don't think it should influence what choices i make. 17:40 -!- EnLilaSko [EnLilaSko@unaffiliated/enlilasko] has quit [Remote host closed the connection] 17:42 < ParahSail1n> should probably have in mind where you want to go with it eventually, as that will inform initial design considerations 17:47 <@kanzure> in the past that has overly constrained the problem space 17:47 <@kanzure> "a system that will be infinitely scalable to all future desirable amounts of DNA".. yeah right. 17:55 <@nmz787> yeah they are doing some emulsion bead shit 17:55 <@nmz787> the reagent savings and ease of separation at the microscale is so appealing though 17:56 < gradstudentbot> I think I have ebola. 17:57 <@nmz787> as for enzymes that is probably a low estimate, since my goal is to get the oligos generated into long-mers and shoved into a cell 17:57 <@nmz787> well it's reasonable estimate for that 17:57 <@nmz787> low for directed evolution (which is the next reasonable step after micro synthesis and transformation) 17:58 <@nmz787> synthego said they're doing the same thing basically, and they just got 8.3 mil, so this BOM seems totally reasonable 17:58 <@nmz787> also enzymes like tdt 17:58 < ParahSail1n> well, the fib shit is totally insane 17:58 <@nmz787> for doing 'green chem' 17:59 <@nmz787> why? 17:59 <@nmz787> it's easy 17:59 <@nmz787> and i need wires 17:59 <@nmz787> so they can mill then lay traces 17:59 < ParahSail1n> you're choosing like the most expensive possible way to do microfabrication 18:00 <@kanzure> i'm not convinced you can build longmers using this sort of system design (in general) without giant arrays or lots of parallel strands. 18:00 <@nmz787> they said they'd do a lot of the work up for free, if i wrote up a short white paper on the process 18:00 <@kanzure> from now on they shall henceforth be called longmers 18:00 <@nmz787> (of fabbing microlfuidics with fib in general) 18:00 <@nmz787> kanzure: of course there will be parallelization 18:00 <@nmz787> that's another point after low diffusion time 18:01 <@kanzure> we haven't really decided on parallelization though 18:01 <@kanzure> are we doing an array-based system 18:01 <@nmz787> i've been thinking a fluid phase synthesis 18:01 <@kanzure> or one of the more traditional n-channel systems (n=32,64,192,392,etc.) 18:01 <@nmz787> balloon-like chamber that just gets bigger 18:01 <@nmz787> silicon bottom and pdms or teflon top 18:02 <@nmz787> add stoichiometric amounts of reagents 18:02 <@nmz787> or just under 18:02 <@nmz787> then filter 18:03 <@nmz787> and start again with the filtrate 18:03 <@nmz787> like every 10 additions, skim off the 10*nRound size fragment 18:03 <@kanzure> filter through an hplc column? 18:04 <@nmz787> that or gel or some micro are media free 18:04 <@nmz787> the walls do all the work 18:04 <@nmz787> that might have been much smaller, but that's the beauty of using a fib, it can do the dynamic range 18:04 < gradstudentbot> Where did you put the revisions to the paper? 18:05 <@kanzure> how is a balloon-chamber the same as parallelization? 18:06 <@nmz787> not, just you asked are we doing array 18:09 <@kanzure> are we? i am still confused. 18:11 <@nmz787> array means? 18:14 -!- ThomasEgi [~thomas@panda3d/ThomasEgi] has quit [Remote host closed the connection] 18:19 <@kanzure> it means either microarray or an array of mesoscopic wells or an array of containers laid out in a grid 18:19 <@kanzure> array always implies grid 18:20 <@kanzure> (in this context) 18:30 -!- yashgaroth [~ffffff@cpe-66-27-118-94.san.res.rr.com] has joined ##hplusroadmap 19:02 < fredox> do you have any strand lengths in mind for particular things you want to be able to synth? say X kbp 19:12 <@kanzure> well, the minimum is a primer at least 16 bp. 19:13 <@kanzure> i am trying to keep expectations low 19:13 <@kanzure> because it's not reasonable to assume you can build a machine that can make 100mers flawlessly every time 19:14 < fredox> my limited research indicates 50mers is about as long as you'd practically go for a single well 19:15 <@kanzure> there are some machines that are claimed to do 100-250, but i've never heard of more than 250. 19:17 < fredox> time factors i've seen are around 10mins per cycle 19:19 < gradstudentbot> What do you mean this isn't going to work? 19:22 <@kanzure> i've seen 2 minutes per cycle 19:24 <@kanzure> if you make a machine that has a moving pipettor and a 100x100 array of wells then you have to wait around forever while it performs the reaction in each step. even if you have n pipettors it's still worse than dmd-based synthesis. but the advantage is that it's easier to debug than using a DMD and squinting really hard. 19:24 <@kanzure> erm by n pippetors i mean size-of-your-row number of tips acting at once 19:24 <@kanzure> i don't know what those setups are called 19:25 <@kanzure> multipipettor claw things 19:25 < gradstudentbot> Did you see that hack? 19:25 -!- cpopell [47fff18b@gateway/web/freenode/ip.71.255.241.139] has quit [Ping timeout: 250 seconds] 19:30 <@nmz787> no i was planning all silicon(e) parallelization 19:30 <@nmz787> they would be an array of sorts 19:30 <@nmz787> more like a binary tree 19:31 -!- augur_ is now known as augur 19:32 < fredox> another thing i'm pondering is what would be the cheapest simplest way to verify your machine was producing the intended sequence 19:32 <@heath> """ Short oligonucleotides 15–25-nt in length can be synthesized without any capping step. Oligonucleotides >40 nt in length required a capping step to achieve high yield of full-length product. Each synthesis cycle contained seven reaction steps and four washing steps. The step-sequence was:""" 19:32 <@heath> nmz787: source? 19:33 <@nmz787> just and old sequencing will verify 19:33 <@nmz787> any old 19:33 <@nmz787> * 19:33 <@nmz787> that was a direct copy past 19:33 <@heath> https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860119/ 19:33 <@heath> kk 19:33 * heath decides to try google before scholar.google.com next time 19:33 <@nmz787> heath google gave this http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860119/ 19:34 <@nmz787> which is probably in bryans repo 19:34 <@heath> you don't have to protect thymine 19:34 <@heath> you have dT in there 19:34 <@kanzure> fredox: the cheapest way would be to send it off for sequencing because sequencing is extremely cheap. 19:35 < fredox> no open sequencer in the works? 19:36 <@heath> maybe it's needed using the protocol from that paper though.. 19:36 <@nmz787> d just emeans deoxy 19:36 < ParahSail1n> http://www.dezeen.com/2013/07/25/farm-432-insect-breeding-kitchen-appliance-by-katharina-unger/ 19:37 <@nmz787> fredox: there is an open source synthesizer 19:37 <@kanzure> pogam.. 19:37 <@nmz787> fredox: http://genomebiology.com/content/5/8/R58 19:37 <@nmz787> posam 19:38 <@kanzure> oh look, tito posted a teardown from biocurious of a sequencer 19:38 <@kanzure> http://titojankowski.com/the-500000-dna-sequencer-tear-down/ 19:38 <@kanzure> this is literally the only thing he has ever done right 19:38 <@kanzure> now where's the firmware dump 19:38 <@nmz787> yeah 19:39 <@heath> nmz787: i was wrong for thinking d meant "protected", but i thought d stood for "dimethoxytrityl" 19:39 <@nmz787> ahh i only saw the flickr page 19:39 <@nmz787> a few days ago 19:39 <@nmz787> they didn't seem to get good closeups of the boards 19:39 <@nmz787> to see what chips were going on 19:39 <@kanzure> nope 19:39 <@heath> kkkkkkk 19:39 <@kanzure> what's really fucking annoying is that i specifically asked for that 19:39 <@nmz787> nah that's dmt heath 19:39 <@kanzure> i sent them emails and even told them in here that i wanted that shit 19:39 <@kanzure> plus firmware dumps 19:39 < gradstudentbot> Yeah, that's a reasonable explanation. 19:39 <@kanzure> but nooo they had to go piss the opportunity away 19:39 <@kanzure> the fuckers. :( 19:40 * heath hasn't fully read the paper on manual oligy synth, i became distracted with d3's source 19:40 <@kanzure> i guess they just think circuits are magical or something. "details? why would anyone want those?" 19:40 <@nmz787> kanzure: and on aliexpres the cheapest syringe pumps are $300 19:40 <@nmz787> which is kinda weird 19:41 <@kanzure> syringe pumps vary dramatically in performance 19:41 <@kanzure> but the upside is that they are usually very explicit about what they are capable of 19:41 <@nmz787> the pneumatic valves are cheaper, but aliexpress search sucks and i couldn't search "high pressure" 19:41 <@nmz787> so i was finding things that maxed at 115 psi 19:41 <@nmz787> which i think is right around or under the quake style valve psi 19:42 <@kanzure> where was that church paper about doing a micropipettor plus microarray 19:44 <@kanzure> "Owing to its open-top architecture, accessibility of the beads and the bead size, this platform is well suited for a pick-and-place approach using micropipettes to retrieve specific beads from the 454-Picotiterplate (PTP) and transfer them into conventional multi-well plates for further processing." 19:45 <@kanzure> http://arep.med.harvard.edu/pdf/Matzas_10.pdf 19:46 < fredox> nmz787: that design was essentially what i had in mind 19:47 <@kanzure> supplementary files for that last paper, http://www.nature.com/nbt/journal/v28/n12/extref/nbt.1710-S1.pdf 19:48 <@kanzure> oh look they have a pic of their setup 19:48 <@kanzure> that array is way larger than i imagined 19:49 <@kanzure> wtf "This configuration allowed semi-automated bead extraction, but requiring verification of bead locations and fine adjustments for pipette placement by a skilled operator." 19:50 <@nmz787> fredox: which? 19:51 <@kanzure> ParahSail1n: how much would a picotiter plate cost 19:52 < fredox> the inkjet 19:53 -!- randallagordon [~randall@71-34-73-76.ptld.qwest.net] has quit [Ping timeout: 264 seconds] 19:53 <@heath> ah, "A synthesis starts with an appropriately base-protected [N^6-benzoyl adine, N^2-isobutryl guanine, N^4-benzoyl cytosine(thymine is usually not protected)]" 19:53 < ParahSail1n> picotiter? 19:53 * heath was mistaking that line for what i was reading in the google doc 20:00 <@kanzure> hah their 454 GS FLX run was $15k in reagents 20:00 <@kanzure> crazy 20:00 <@kanzure> i think the church lab is like a flaming oil well 20:02 < gradstudentbot> Hey, let's write a paper about that. 20:08 <@kanzure> rofl 20:18 <@kanzure> so anyway, that might be a good way to store a library of beads. 4096x4096 means an entire 24mer library. 20:37 < ParahSail1n> 454 is a pretty expensive way to get data 20:37 <@kanzure> i just mean their plates 20:59 -!- tomkinsc [~tomkinsc@c-24-4-11-202.hsd1.ca.comcast.net] has joined ##hplusroadmap 21:00 -!- fredox [~chatzilla@c27-253-21-32.brodm4.vic.optusnet.com.au] has quit [Read error: Connection reset by peer] 21:04 -!- fredox [~chatzilla@c27-253-21-32.brodm4.vic.optusnet.com.au] has joined ##hplusroadmap 21:30 -!- randallagordon [~randall@75-164-241-15.ptld.qwest.net] has joined ##hplusroadmap 21:33 <@kanzure> /win 4 21:39 -!- heath [quassel@unaffiliated/ybit] has quit [Quit: No Ping reply in 180 seconds.] 21:44 -!- gnusha [~gnusha@131.252.130.248] has joined ##hplusroadmap 21:44 -!- Topic for ##hplusroadmap: biohacking, nootropics, transhumanism, open hardware | sponsored by george church and the NRA | http://gnusha.org/logs http://diyhpl.us/wiki http://groups.google.com/group/diybio | banned by the Federal Death Administration | official paperbot fan club 21:44 -!- Topic set by kanzure [~kanzure@131.252.130.248] [Sat Mar 23 20:40:45 2013] 21:44 [Users ##hplusroadmap] 21:44 [@bkero ] [ Burninate ] [ jrayhawk_ ] [ paperbot ] [ sivoais ] [ upgrayeddd] 21:44 [@kanzure ] [ cogitokat ] [ juri_ ] [ ParahSail1n ] [ smeaaagle ] [ Urchin ] 21:44 [@nmz787 ] [ Coornail ] [ juul ] [ ParahSailin ] [ spresser ] [ valyap ] 21:44 [ _sol_ ] [ devrandom ] [ klafka ] [ pasky ] [ strangewarp] [ yashgaroth] 21:44 [ abetusk ] [ fenn ] [ lichen ] [ phryk ] [ streety ] [ yoleaux ] 21:44 [ AlonzoTG] [ fredox ] [ lupfantomo ] [ poppingtonic ] [ superkuh ] [ Zhwazi ] 21:44 [ archels ] [ gnusha ] [ monkeynipples] [ randallagordon] [ Thomas42_ ] [ zubaz ] 21:44 [ audy ] [ gradstudentbot] [ nsh ] [ rigel ] [ Thorbinator] 21:44 [ augur ] [ Guest10542 ] [ nuba_ ] [ ryankarason ] [ tomkinsc ] 21:44 [ balrog ] [ HEx1 ] [ nully ] [ saurik ] [ ua ] 21:44 [ brownies] [ ivan` ] [ oblique ] [ Shehrazad_ ] [ underscor ] 21:44 -!- Irssi: ##hplusroadmap: Total of 62 nicks [3 ops, 0 halfops, 0 voices, 59 normal] 21:44 -!- Channel ##hplusroadmap created Thu Feb 25 23:40:30 2010 21:44 -!- Irssi: Join to ##hplusroadmap was synced in 6 secs 21:58 <@kanzure> https://www.lookout.com/resources/reports/dragon-lady "Mobile security company Lookout released a report today at DefCon that reveals the amazing size, scope, and complexity of Android malware operations in Russia. The report found the bulk of this Russian malware wasn't coming from lone individuals in basements, but well-oiled malware producing machines." 22:06 < fredox> well-oiled malware producing machines ...like google 22:06 <@kanzure> "Mobile Ad Networks: Lookout recently reported on a new malware, BadNews, which was found to be a new technique to drive mobile traffic to SMS fraud campaigns. BadNews was designed to look like an advertising library in legitimate Android applications, but the advertisements that it displayed linked directly to SMS fraud malware hosted by top HQs." 22:06 -!- poppingtonic [~poppingto@212.49.88.104] has quit [Ping timeout: 248 seconds] 22:06 -!- poppingtonic [~poppingto@212.49.88.104] has joined ##hplusroadmap 22:44 -!- tomkinsc [~tomkinsc@c-24-4-11-202.hsd1.ca.comcast.net] has quit [Quit: My computer has gone to sleep. ZZZzzz…] 22:48 -!- heath [quassel@2600:3c02::f03c:91ff:feae:6e5b] has joined ##hplusroadmap 22:48 -!- heath [quassel@2600:3c02::f03c:91ff:feae:6e5b] has quit [Changing host] 22:48 -!- heath [quassel@unaffiliated/ybit] has joined ##hplusroadmap 23:01 <@nmz787> kanzure: ok added tubing and assorted connectors as a cost 23:01 <@kanzure> i don't think fib is a good idea 23:01 <@kanzure> and if hplc is really 20k then we should just work on that first 23:01 <@nmz787> if you remove the hplc the total is 23k 23:02 < ParahSail1n> why is there hplc in this? 23:02 <@kanzure> and 18k fib? 23:03 <@nmz787> the fib is a good idea i think, it makes nano and micro fluidics very simple to prototype rapidly. yes it's costly, but this company /is/ willing to work with me and I'd like to take advantage of that. They were using COMSOL for their ion optics modelling so I might be able to learn some of that through them too, or get them to crunch my designs 23:03 < ParahSail1n> i only used desalted oligos 23:03 <@nmz787> how long though? 23:04 <@nmz787> i think they do hplc on all bathces 23:04 <@nmz787> batches 23:05 <@nmz787> hplc would allow moving into directed evolution work once synthesis was acquired, and could serve useful as a crude sequencer during development 23:05 < ParahSail1n> hplc is an option which i've never used with idt 23:05 <@nmz787> it sounds like i should check sci exchange 23:05 <@nmz787> but kanzure this is something the eugene lab ppl said they could help with 23:05 <@kanzure> doesn't matter to me. that's an hour away from you. fuck that. 23:05 <@nmz787> equip that is 23:06 <@nmz787> well maybe they could get it cheaper? 23:06 < yashgaroth> how are you doing sequencing via hplc? 23:06 <@nmz787> or some indefinite loan 23:06 <@kanzure> oh yeah, indefinite loan payments, that's totally what i want 23:06 <@nmz787> no 23:06 <@nmz787> like $1/year 23:07 < ParahSail1n> i think you're writing this BOM as if the the NSF is gonna foot the bill 23:07 < gradstudentbot> I am completely satisfied with the size of my bench space. 23:08 <@nmz787> i kinda just assumed it was a standard thing in any decent lab 23:08 <@nmz787> http://www.invitrogen.com/site/us/en/home/Products-and-Services/Product-Types/Primers-Oligos-Nucleotides/invitrogen-custom-dna-oligos/Oligo-Ordering-Details/Oligo-Purity-Selection-Guide.html 23:09 <@nmz787> HPLC 23:09 <@nmz787> (50 nm+, 10-55 bp) 23:09 < ParahSail1n> not just the hplc, but using FIB instead of normal microfabrication protocols 23:09 <@nmz787> thats nanomolar i believe 23:09 <@nmz787> but FIB is pretty normal around here 23:09 <@nmz787> they make them in the shop i referred to earlier 23:09 <@nmz787> and the company that makes a shitload more is local too 23:09 <@nmz787> FEI 23:10 <@nmz787> I've toured their place 23:10 <@kanzure> ParahSail1n: would you be willing to do a BOM too? 23:10 <@nmz787> it seems like a solid clean direct way to do things 23:10 <@nmz787> mill and add circuitry 23:10 < ParahSail1n> im not clear what exactly is do be done 23:11 <@kanzure> that's still being explored 23:11 <@nmz787> ParahSail1n: do you do wafer fab? 23:11 < ParahSail1n> i have in the past 23:11 <@nmz787> would you like to do it for these designs? 23:11 <@nmz787> for less than the quoted $375/hr 23:11 < ParahSail1n> i have a former boss who'd probably do it for money 23:12 < gradstudentbot> That's not really surprising since they did it ex vivo. 23:12 <@nmz787> (not including their free overhead shit that they've already done some of) 23:12 <@kanzure> i don't get it. what FIB things are involved? 23:12 <@nmz787> mill and add circuitry 23:12 <@kanzure> why would he have to waferr fab this? 23:12 <@nmz787> mill and add circuitry 23:12 <@kanzure> circuitry can be done extremely cheap, even if we need custom pcbs (which we don't. use a fucking breadboard.) 23:12 <@nmz787> not pcbs 23:12 < ParahSail1n> im unclear what is to be fabricated, but if you want a wafer etched with some sort of profile to use as a template for pdms soft lithography 23:13 <@nmz787> i could give you cad files or something 23:13 <@kanzure> i think nmz787 is talking about doing microfluidic things 23:13 <@nmz787> gerbers 23:13 <@nmz787> i dunno what you'd want 23:13 < ParahSail1n> that would be about an order of a magnitude cheaper than 50 hours of fib 23:13 <@kanzure> and he's being intentionally unclear about it 23:13 <@kanzure> because i have no other reasonable explanation of this 23:13 <@nmz787> some square channels, roughly square crosssection, some parabolic bottom channels 23:14 <@kanzure> nmz787: so hey, i'm still the same person. i'm always going to be a truly pedantic asshole, even when you're sending me potential BOMs. 23:14 <@nmz787> some traces to the ends and sides of some of the channels 23:14 <@kanzure> what channels 23:15 <@nmz787> the fluidics channels for a microfluidic synthesizer 23:15 <@kanzure> but we weren't going to do a microfluidic synthesizer 23:15 <@kanzure> it was going to be a cheap copy of some other junk 23:15 <@nmz787> you said you wouldn't mind if I did it in parallel 23:15 <@nmz787> i don't have any ideas for copies, you mentioned browsing patents 23:16 <@nmz787> i looked a bit earlier but didn't find anything 23:16 <@nmz787> that's really not interesting or promising though 23:16 <@kanzure> you don't have any ideas about how conventional synthesizers work? that sounds really bad yo. 23:16 <@kanzure> it's interesting because they work 23:16 <@nmz787> once a design is found, some wafer people could definitely come up with cheaper world available processes 23:16 <@kanzure> found? 23:16 <@nmz787> but for now it makes things easy 23:17 <@nmz787> no it's not interesting because they take long as hell and eat costly reagents 23:17 -!- AshleyWaffle [~quassel@unaffiliated/anastasiawyatt] has joined ##hplusroadmap 23:17 < ParahSail1n> maybe a cheap open hplc would be a better project 23:17 <@kanzure> the costly reagents aren't the point 23:17 <@nmz787> they are to me 23:18 <@nmz787> that's a huge point 23:18 <@kanzure> you don't have any machine whatsoever, why would you care 23:18 < ParahSail1n> i dont have an idea on how cheap it could be, or how many people would benefit from such a thing 23:18 <@nmz787> because i don't want a crappy copy 23:18 <@kanzure> you don't have anything at all, though 23:18 <@nmz787> exactly 23:18 <@kanzure> ...? 23:18 <@nmz787> i need to build it 23:18 <@kanzure> yeah, a standard synthesizer 23:19 <@nmz787> no man 23:19 <@nmz787> they're different physics to make them spin 23:19 <@kanzure> the reason you don't have a synthesizer isn't because the reagents are expensive 23:19 <@nmz787> less is more 23:19 < ParahSail1n> fundamentally all you are doing is pumping fluid from a reservoir through a column that you buy at high pressure 23:19 < ParahSail1n> the detectors on the other side of the column are an independent unit 23:20 <@nmz787> the hplc would be useful if you want me to do pipettor synthesis 23:21 <@nmz787> since it can detect 10-55 bp 23:22 <@nmz787> ParahSail1n: Desalt: Oligos are processed through normal phase chromatography column which removes salts but not failure sequences 23:23 < ParahSail1n> works plenty fine unless you are doing supercool stuff like mlpa reactions with the oligos 23:24 <@nmz787> yeah but not for writing clonal genes 23:25 < ParahSail1n> well you're not really writing genes, you're writing oligos to stitch together 23:25 < ParahSail1n> and you can go over the sequence later and correct introduced snps with the usual techniques 23:26 <@nmz787> and the higher the purity between stiching operations, the higher the yield overall 23:26 < ParahSail1n> you can always improve yield after you have a working prototype 23:26 <@kanzure> and oligo length 23:26 <@kanzure> and reagent volumes 23:27 < ParahSail1n> i suspect page purification would be easier to hack together than hplc 23:27 < ParahSail1n> casting page in capillaries is pretty easy 23:28 < ParahSail1n> and you only need electric potential and buffer to push stuff through that 23:28 <@kanzure> ParahSail1n: also we don't have to do just one thing at a time. if you think we need to tackle a column then we can do that. 23:28 < ParahSail1n> the massive fluid handling and air compressor of an hplc 23:29 < ParahSail1n> kanzure, column is not essential 23:29 <@kanzure> well it would be nice for protein purification reasons 23:29 <@kanzure> dunno 23:29 < ParahSail1n> ah, well page will only do oligos 23:29 <@nmz787> yeah thats the plan for the fib, page 23:30 <@nmz787> there are some examples of just using the microchannel to do gel free separation 23:30 <@kanzure> each of those were separate research projects that took a lot of time to build up 23:30 <@kanzure> ParahSail1n: can't you chime in with your experience in microfluidics here 23:31 < ParahSail1n> ive been 23:31 <@kanzure> ParahSail1n: because nmz787 doesn't know about that 23:31 <@nmz787> page is anyways done in microchannels 23:31 <@kanzure> no i mean.. you have actual physical hands-on experience with it 23:31 <@nmz787> all the ABI sequencers use 50 micron tubes 23:31 < ParahSail1n> fib is insane for this, in microfabrication you avoid beams wherever possible 23:31 <@nmz787> why? 23:31 <@nmz787> that sounds stupid 23:32 <@nmz787> it's a sweet instrument 23:32 <@nmz787> it makes sense 23:32 < ParahSail1n> because you are drawing nano features with one tiny head 23:32 <@nmz787> and realistically if you use it, it's not that expensive to run 23:32 <@nmz787> electricity wise 23:32 < ParahSail1n> the only thing you use beams in production for is e-beam lithography of masks for normal lithography 23:32 <@nmz787> huh? 23:32 <@nmz787> you can make the beam micron sized to nano size 23:32 -!- yashgaroth [~ffffff@cpe-66-27-118-94.san.res.rr.com] has quit [Quit: Leaving] 23:32 < ParahSail1n> because it takes you days to print a billion features in a die with an e-beam 23:33 < ParahSail1n> and you're not even talking about features that require such an expensive proces 23:33 <@nmz787> why do ebeam on a chrome plate then do lithography 23:33 <@nmz787> just beam on the silicon to mill in the first place 23:33 <@nmz787> it's just like CNC 23:33 <@nmz787> this isn't e tho 23:33 < ParahSail1n> this is simple stuff you could pattern with a photolithography mask that you printed with an inkjet printer 23:33 <@nmz787> its gallium 23:34 <@nmz787> no way, printers are at least an order of magnitude too coarse 23:34 < ParahSail1n> ok then send the cad drawing to one of the shops and theyll make it high res 23:35 <@nmz787> that's what i'm saying to do 23:35 <@nmz787> i mentioned the milling rate eariler 23:35 <@nmz787> earlier 23:35 < ParahSail1n> the "shops" dont do fib shit to make those original masks 23:35 <@nmz787> and kanzure we long ago calculated some sequencer math length 23:35 <@nmz787> it was 60 cm i believe 23:36 <@nmz787> ParahSail1n: right 23:36 <@nmz787> round here they do 23:36 < ParahSail1n> do you need <3 um features? 23:36 <@nmz787> http://www.norsam.com/ 23:36 <@nmz787> yes 23:37 <@nmz787> all my reagent volumes have been based on 1 micron cube 23:37 < gradstudentbot> What do you mean this isn't going to work? 23:37 < ParahSail1n> so it would break if you made it 9x bigger volume? 23:38 <@nmz787> no 23:38 <@nmz787> but i don't want to do silicon etching 23:38 <@nmz787> this thing is CNC 23:38 <@nmz787> it's easy 23:40 <@nmz787> ok so it says 60 cm path would take 100 minutes to mill 23:41 < valyap> this is one of the busiest channels on freenode 23:41 <@nmz787> that's a column that gets 300-500 bp read length 23:41 < valyap> has their been a breakthrought? 23:41 <@kanzure> valyap: no, go away 23:41 < valyap> hi kanzure 23:42 <@nmz787> the other option is getting some folks together to seriously hack this bluray writer as a polar stereolithography writer 23:42 < ParahSail1n> so you want to use a vector scanning fabrication technique in a mass produced device? 23:42 <@nmz787> but that isn't a sure thing 23:43 <@nmz787> and wouldn't get electric traces 23:43 <@kanzure> ParahSail1n: not necessarily mass produced, not sure 23:43 <@nmz787> raster 23:43 < ParahSail1n> this thing is valuable enough that people will pay 18k for each one? 23:43 <@kanzure> ParahSail1n: also, there's nothing wrong with vector-based designs. vector writing *shrug* i'd like to try it sometime. 23:43 <@kanzure> ParahSail1n: 18k for what? 23:43 < ParahSail1n> no, its just low throughput 23:44 <@kanzure> oh, the time on the fib 23:44 <@kanzure> yeah, probably not 23:44 < ParahSail1n> 50 hours of fib time, tracing a single print head across every single feature 23:44 <@nmz787> since bluray track spacing is 300nm 23:44 <@kanzure> also, i think it's going to take more than 50 hours/100 minutes =~ 30 iterations 23:44 <@kanzure> ParahSail1n: how many iterations on a design of this magnitude/complexity do you think it would take to get a synthesizer to work? 23:44 <@kanzure> microfluidic synthesizer. 23:44 <@nmz787> i just said 100 mins 23:45 <@nmz787> not 18k? 23:45 <@kanzure> you said 100 min per 60cm (per run) 23:45 <@nmz787> and i said fib was just for prototyping 23:45 <@kanzure> so 50 hours divided by 100 minute runs 23:45 <@kanzure> is 30 iterations 23:45 <@nmz787> 50 hours? 23:45 <@kanzure> you said 50 hours 23:45 <@kanzure> sigh 23:45 < ParahSail1n> photolithography smaller than 3um feature is a thing 23:45 <@nmz787> not for a device in production 23:45 <@kanzure> no you said design time 23:45 < ParahSail1n> tell that to intel 23:45 <@nmz787> (to kanzure ) 23:45 <@kanzure> please try to follow 23:45 <@nmz787> not you 23:46 < ParahSail1n> or tell that to stm micro 23:46 <@nmz787> no i def know smaller than 3 um lith 23:46 <@nmz787> i'm saying per device wont be $18k in fib 23:47 <@nmz787> that is for research and dev 23:47 <@nmz787> and worst case 23:47 <@kanzure> 30 iterations is not worst case -_- 23:47 < gradstudentbot> I don't know whether I am Turing dreaming that I am a machine, or a machine dreaming that I am Turing! 23:47 < ParahSail1n> maybe you want to do your iterations on cheap 3um photomasks 23:48 < ParahSail1n> and then you can buy a fancy ebeam mask later after you've gotten it working on easy mode 23:48 <@kanzure> "easy" 23:49 <@nmz787> i was saying that FIBing silicon to mill and add circuitry was easier and cleaner than wafer fab IMO and IME 23:50 <@nmz787> i've only made few hundred micron sized transistors once 23:50 <@nmz787> and seen some stuff FIBbed a few times 23:50 <@nmz787> like 3 or 4 23:50 <@nmz787> :/ 23:50 < ParahSail1n> 100 minutes of fib time is not terrible 23:50 <@nmz787> FIB was pretty much like a laser cutter and welder in one 23:52 <@nmz787> I'm totally open to how to make the thing 23:52 <@nmz787> but i've been looking at techniques for a while, and FIB is damn attractive as someone who wants it to just work 23:53 <@nmz787> plus with nano capabilities we can start to think about using enzyme traps 23:53 <@nmz787> !!! 23:55 <@nmz787> tDt!? 23:56 <@nmz787> the circuit laying would be considerably less than 100 minutes of FIB time 23:57 <@nmz787> so even if there was no better way to do that later 23:57 <@nmz787> but I'm pretty sure there are automated fabs that could easily make at least the silcon part 23:57 <@nmz787> or glass 23:59 <@nmz787> for the $500 a pop in R&D, i'd rather spend it locally than send to that stanford foundry --- Log closed Sat Aug 03 00:00:44 2013