2016-08-21.log

--- Log opened Sun Aug 21 00:00:31 2016
-!- ihaq [anoncicada@gateway/shell/elitebnc/x-xpjtemxzglwqtjvu] has quit [Quit: faggory daggory doo!]00:01
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]00:41
-!- ebowden [~ebowden@2001:8003:100e:c500:99c:56c1:3eb9:7b97] has joined ##hplusroadmap00:42
-!- ebowden [~ebowden@2001:8003:100e:c500:99c:56c1:3eb9:7b97] has quit [Changing host]00:42
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap00:42
-!- zeysmo [iM64ovm2tn@columba.uberspace.de] has quit [Ping timeout: 252 seconds]00:47
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]00:56
-!- ebowden [~ebowden@2001:8003:100e:c500:7537:f4f0:36fb:bb7b] has joined ##hplusroadmap01:00
-!- ebowden [~ebowden@2001:8003:100e:c500:7537:f4f0:36fb:bb7b] has quit [Changing host]01:00
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap01:00
-!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-ccltzrcypydvyimu] has quit [Quit: Connection closed for inactivity]01:02
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]01:22
-!- cynsia [cyn@2600:3c03::f03c:91ff:fe79:d5d] has quit [Quit: Leaving]01:22
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap01:39
-!- augur [~augur@2602:304:cdac:e260:bc11:e1c7:d01a:384f] has quit [Remote host closed the connection]02:14
nmz787_op-amp topology cheatsheet: http://chemelec.com/Projects/Op-Amp/Op-Amp.png02:30
-!- Aurelius_Laptop [~cpopell@c-73-200-185-48.hsd1.va.comcast.net] has quit [Ping timeout: 244 seconds]03:08
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]03:43
-!- ebowden [~ebowden@2001:8003:100e:c500:a17c:2dd4:7592:ad53] has joined ##hplusroadmap03:44
-!- ebowden [~ebowden@2001:8003:100e:c500:a17c:2dd4:7592:ad53] has quit [Changing host]03:44
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap03:44
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap03:50
chris_99http://sensl.com/ is interesting, i didn't know you could get silicon photomultipliers commercially03:53
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]04:17
-!- ebowden [~ebowden@2001:8003:100e:c500:a17c:2dd4:7592:ad53] has joined ##hplusroadmap04:17
-!- ebowden [~ebowden@2001:8003:100e:c500:a17c:2dd4:7592:ad53] has quit [Changing host]04:17
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap04:17
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-lsnurnkrkdgohpok] has joined ##hplusroadmap04:28
-!- Gurkenglas [Gurkenglas@dslb-178-005-223-183.178.005.pools.vodafone-ip.de] has quit [Ping timeout: 265 seconds]05:47
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]06:31
-!- ebowden [~ebowden@2001:8003:100e:c500:39c4:9cb3:c3c9:8ff] has joined ##hplusroadmap06:32
-!- ebowden [~ebowden@2001:8003:100e:c500:39c4:9cb3:c3c9:8ff] has quit [Changing host]06:32
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap06:32
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]06:36
-!- ebowden [~ebowden@2001:8003:100e:c500:71bb:63d3:9318:61f1] has joined ##hplusroadmap06:38
-!- ebowden [~ebowden@2001:8003:100e:c500:71bb:63d3:9318:61f1] has quit [Changing host]06:38
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap06:38
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]06:50
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap06:51
-!- Gurkenglas [Gurkenglas@dslb-178-005-223-183.178.005.pools.vodafone-ip.de] has joined ##hplusroadmap06:56
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]07:10
-!- ebowden [~ebowden@2001:8003:100e:c500:dda2:985:3d53:d791] has joined ##hplusroadmap07:11
-!- ebowden [~ebowden@2001:8003:100e:c500:dda2:985:3d53:d791] has quit [Changing host]07:11
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap07:11
-!- Gurkenglas [Gurkenglas@dslb-178-005-223-183.178.005.pools.vodafone-ip.de] has quit [Ping timeout: 260 seconds]07:41
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-lsnurnkrkdgohpok] has quit [Quit: faggory daggory doo!]07:53
-!- Gurkenglas [Gurkenglas@dslb-178-005-223-183.178.005.pools.vodafone-ip.de] has joined ##hplusroadmap07:57
kanzurehmm.08:05
-!- Jawmare [~Jawmare@unaffiliated/jawmare] has quit [Ping timeout: 250 seconds]08:08
-!- Filosofem [~Jawmare@unaffiliated/jawmare] has joined ##hplusroadmap08:09
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]08:09
-!- Sooka [~Malvolio@unaffiliated/malvolio] has quit [Ping timeout: 265 seconds]08:14
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap08:24
-!- irseeyou [~irseeyou@c-67-168-101-20.hsd1.wa.comcast.net] has joined ##hplusroadmap08:35
-!- ebowden [~ebowden@CPE-101-180-249-107.lnse3.cha.bigpond.net.au] has joined ##hplusroadmap08:45
-!- ebowden [~ebowden@CPE-101-180-249-107.lnse3.cha.bigpond.net.au] has quit [Changing host]08:45
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap08:45
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Ping timeout: 260 seconds]08:50
-!- ebowden [~ebowden@2001:8003:100e:c500:dda2:985:3d53:d791] has joined ##hplusroadmap08:54
-!- ebowden [~ebowden@2001:8003:100e:c500:dda2:985:3d53:d791] has quit [Changing host]08:54
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap08:54
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-pttpxjgbaunsdbri] has joined ##hplusroadmap08:56
-!- tft27 [~textual@237.107.195.178.dynamic.wline.res.cust.swisscom.ch] has joined ##hplusroadmap09:09
-!- tft27 [~textual@237.107.195.178.dynamic.wline.res.cust.swisscom.ch] has left ##hplusroadmap []09:09
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]09:12
-!- zeysmo [hYmt4bgiz8@columba.uberspace.de] has joined ##hplusroadmap09:41
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-pttpxjgbaunsdbri] has quit [Changing host]09:44
-!- anoncicada [anoncicada@unaffiliated/anarcanon] has joined ##hplusroadmap09:44
-!- anoncicada [anoncicada@unaffiliated/anarcanon] has quit [Changing host]09:44
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-pttpxjgbaunsdbri] has joined ##hplusroadmap09:44
-!- Aurelius_Laptop [~cpopell@static-108-28-104-83.washdc.fios.verizon.net] has joined ##hplusroadmap09:54
-!- ebowden [~ebowden@CPE-101-180-249-107.lnse3.cha.bigpond.net.au] has joined ##hplusroadmap10:06
-!- ebowden [~ebowden@CPE-101-180-249-107.lnse3.cha.bigpond.net.au] has quit [Changing host]10:06
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap10:06
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Ping timeout: 260 seconds]10:11
-!- winsen [~winsen@unaffiliated/winsen] has joined ##hplusroadmap10:12
-!- irseeyou [~irseeyou@c-67-168-101-20.hsd1.wa.comcast.net] has quit [Quit: Leaving.]10:16
-!- ArturShaik [~ArturShai@31.29.9.62] has quit [Ping timeout: 265 seconds]10:24
-!- Aurelius_Laptop [~cpopell@static-108-28-104-83.washdc.fios.verizon.net] has quit [Ping timeout: 252 seconds]10:40
-!- Orpheon [~Orpheon@27.160.62.81.dynamic.wline.res.cust.swisscom.ch] has joined ##hplusroadmap10:51
-!- yashgaroth [~yashgarot@2602:306:35fa:d500:f5e0:f867:a11d:8d52] has joined ##hplusroadmap10:51
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Ping timeout: 250 seconds]10:53
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-pttpxjgbaunsdbri] has quit [Ping timeout: 250 seconds]10:58
-!- anonciada [anoncicada@gateway/shell/elitebnc/x-udbgtidsefpkkkrk] has joined ##hplusroadmap10:58
-!- anonciada is now known as anoncicada10:58
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-udbgtidsefpkkkrk] has quit [Changing host]10:58
-!- anoncicada [anoncicada@unaffiliated/anarcanon] has joined ##hplusroadmap10:58
-!- anoncicada [anoncicada@unaffiliated/anarcanon] has quit [Changing host]10:58
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-udbgtidsefpkkkrk] has joined ##hplusroadmap10:58
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap11:06
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap11:10
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Ping timeout: 260 seconds]11:11
maakumust assume tor is now compromised :(11:39
-!- nildicit_ [~nildicit@unaffiliated/nildicit] has joined ##hplusroadmap11:40
-!- nildicit_ [~nildicit@unaffiliated/nildicit] has left ##hplusroadmap []11:40
superkuhI don't think so. But I think the type of naturally occurring cultural infection causing the drama will leave it open to outside compromise.11:46
maakueh, character assasination in the court of public opinion, followed by a cabal of the accursors taking over displaced positions of power?11:48
maakuit's classic strategy for organization compromise11:48
-!- irseeyou [~irseeyou@c-67-168-101-20.hsd1.wa.comcast.net] has joined ##hplusroadmap11:51
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Quit: text1]11:53
kanzureyep11:57
-!- ebowden [~ebowden@2001:8003:100e:c500:f03a:fc32:3a73:da6] has joined ##hplusroadmap12:01
-!- ebowden [~ebowden@2001:8003:100e:c500:f03a:fc32:3a73:da6] has quit [Changing host]12:01
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap12:01
nmz787_well if tor is so affected by politics or bitchfests (I have no idea what you guys are talking about aside from the history of comments I've seen in here)... then seems like that software wasn't very 'secure' to begin with.12:03
maakunmz787_: we're talking about the organization, not the software12:05
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Ping timeout: 260 seconds]12:05
maakubut tor was never so trustless... you trust the organization to gate new relay and exit nodes12:06
maakuso an organizational compromise IS a compromise of tor the software12:06
chris_99what kind of changes to the software could they make, that wouldn't be obvious12:08
maakunone. they don't need to. you're anonymity of tor is due to the presumed independence and honesty of the relay network, which is hand crafted by the tor organization (your node downloads a signed list of relay nodes)12:08
chris_99hmm, anyone can be a relay node right12:09
chris_99so it's not just the tor people you trust12:10
maakuchris_99: tor clients download a list of reliable relay nodes from master nodes controlled by the tor organization12:10
maakuit is not p2p12:10
maaku(Today you learned!)12:10
chris_99but i thought anyone can be a relay node, like anyone can get added to it?12:11
maakuif that were the case it'd be trivial to attack tor by sybil nodes12:11
saturn2the process is mostly automated but the list is still controlled by the tor project12:12
chris_99ah, gotcha12:12
nmz787_got at least one reasonable looking image from my BRL-CAD ray-tracer interrogation method for rastering model slices12:14
nmz787_now to refactor it into the python-BRLCAD-tcl library, and then work on g-code or maybe FIB output actually12:17
nmz787_I wanted to try that weird metalense thing12:18
nmz787_but also wanted to try the photolithography for a basic colored fluid pump... so that would require g-code12:19
kanzurewhat makes the pump colored?12:29
streetyisn't the fluid colored?12:31
kanzureoh.12:34
kanzurebut..12:34
nmz787_yeah12:37
nmz787_I figured I'd have one of those 'start it up and [insert slogan of choice] becomes apparent' kind of proof of concept things12:37
nmz787_huh, you could use that principle with drinking water or something, water filter output... and when algae/biofilm starts to develop your slogan could become obvious/apparent12:38
streetyideas on logo?12:39
streetythat seems quite drawn out12:39
nmz787_idk, DIYbio or hplusroadmap or something12:39
streetywhat sort of pump were you planning on?12:39
nmz787_tiny peristaltic12:39
nmz787_(the water filter/biofilm idea I mentioned wouldn't have pumps... it would just be some tubes in a logo fashion in a clear material that would obviously change if any biofilm grew)12:40
streetyhave you seen anything on hplc with the pump part of the microfluidic device? Everything I've seen uses a standard hplc pump and just connects it up to a microfluidic device12:41
nmz787_hmm, well HPLC is pretty high pressure, I am pretty sure at least 10X if not 50X higher than most microfluidics... I bet there are some people experimenting with super-high-pressures in tiny environs12:43
nmz787_i'd look up 'supercritical microfluidic' or 'subcritical microfluidic pump' or something12:43
nmz787_this is a rendering of the pump model, with some tube 'leads' attached to each input/output: https://github.com/nmz787/python-brlcad-tcl/blob/master/examples/output/microfluidic_pump.stl12:44
chris_99neat, github displays the stl12:45
streetyI was thinking the same12:45
streetythe microfluidic hplc has been something I've been thinking about recently. Not sure I will do anything with it but I'll add those keywords to my research12:46
-!- Guest85738 [~augur@76-218-206-38.lightspeed.sntcca.sbcglobal.net] has joined ##hplusroadmap12:47
kanzureyou can also use centrifugal force with microfluidics for separation12:49
kanzureand flow fractionation and size fractionation and a few other insane things12:49
kanzurehttp://diyhpl.us/~bryan/papers2/microfluidics/12:49
nmz787_streety: I guess something to consider is hydraulic advantage... you can have a microns-scale pump, and as long as the material you neck-down (reduce the cross-section of the fluidic) can handle the pressure, I think that might work to give you higher pressures in the smaller regions (I could be totally wrong about this, I don't really have much experience with hydraulics other than bleeding brakes or clutches in automobiles)12:49
-!- cynsia [cyn@2600:3c03::f03c:91ff:fe79:d5d] has joined ##hplusroadmap12:54
-!- sandeepkr_ [~sandeepkr@103.49.155.114] has joined ##hplusroadmap12:54
-!- sandeepkr__ [~sandeepkr@182.75.187.110] has quit [Ping timeout: 265 seconds]12:55
streetyI'm probably not too far off then as I was thinking along the same lines. A piston with a large area on the low pressure side and a small area on the high pressure side. Trading volume for pressure. Presumably you could have multiple stages progressively raising the pressure.12:56
streetyI've seen a couple of articles using a photoactivated polymer to cast the piston in situ12:57
nmz787_I'd guess that model I showed might work, with low efficiency or something13:03
streetywould you not be limited to the pressure applied to the three cross channels?13:03
nmz787_uhh, hmm... maybe you have parallelized first-stages, which then act as pump-pressure for the next stage13:05
nmz787_fluidic amplifiers? you need more 'current' at lower pressure, to get high pressure but with lower 'current'13:05
nmz787_or some MEMS thing13:06
-!- sandeepkr__ [~sandeepkr@182.75.187.110] has joined ##hplusroadmap13:08
-!- sandeepkr_ [~sandeepkr@103.49.155.114] has quit [Ping timeout: 264 seconds]13:12
kanzuregenehacker was in here a few days ago wanting to do a fluidic radio thingy13:19
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap13:23
chris_99fluidic radio thing?13:23
kanzureyeah at least some have been done with liquid metal stuff13:28
chris_99ahh13:29
kanzurei don't seem to have any references on this. how embarrassing.13:29
kanzure13:30 < genehacker> so we can build a probe that has some fluidic logic and sensors and stuff and it can do science13:33
kanzure13:30 < genehacker> problem is getting data back13:33
kanzure13:31 < genehacker> so I'm thinking maybe I can use said fluidics to modulate the radar cross section of said probe13:33
kanzure13:32 < genehacker> so we can hit it with radar, look at how the cross section changes and get data back13:33
kanzurehmm that's not usually what he's on about13:33
nmz787_I responded to him too13:34
nmz787_didn't find too much on the terms I searched, but didn't spend too much and it was probably late13:34
paskykanzure: sorry behind polar circle now, but never really worked with multiple gpus13:34
chris_99oh interesting, so the RF would tell you the parts of the chip with fluid in?13:34
-!- FourFire [~fourfire@81.4.122.176] has joined ##hplusroadmap13:36
kanzurehttp://diyhpl.us/~bryan/papers2/microfluidics/Stretchable%20microfluidic%20radiofrequency%20antennas.pdf13:36
kanzurechris_99: for genehacker's radar idea, yes it seems to be something like "cause large-scale structural changes that can be detected by [orbital?] radar"13:36
chris_99gotcha. Ive seen those mercury antennas before, they sound really interesting13:37
kanzure"A reconfigurable liquid metal antenna driven by electrochemically controlled capillarity" https://www.researchgate.net/profile/Mohammad_Khan157/publication/277929374_A_reconfigurable_liquid_metal_antenna_driven_by_electrochemically_controlled_capillarity/links/55e6374208aec74dbe74e535.pdf13:38
jcorganyou probably don't need to go to such great lengths13:40
jcorganRFIDs signal data bits by changing their impedance, causing a reflected wave to be modulated13:41
jcorganhttp://www.eetimes.com/document.asp?doc_id=127630613:41
jcorganso you could probably adapt such a technique for this, but with either this or traditional radar the signal power required would be enormous13:42
nmz787_chris_99: have you heard of RF retroreflectors?13:45
chris_99i don't think so13:45
nmz787_.title https://www.youtube.com/watch?v=mAai6dRAtFo13:46
yoleauxDEF CON 22 - Michael Ossmann - The NSA Playset: RF Retroreflectors - YouTube13:46
kanzurejcorgan: genehacker was thinking about venus in particular13:46
chris_99nmz787_, cheers, will have a looky13:46
kanzurepasky: looks like, for multi-gpu data parallelism (multi-gpu training), folks are just averaging all the weights together.... i feel highly uncomfortable about this.13:47
nmz787_where do I get this? http://dl.acm.org/citation.cfm?id=254048513:48
nmz787_.title13:48
yoleaux403 - Forbidden Access to The Digital Library13:48
nmz787_"13:48
nmz787_Accelerated robust point cloud registration in natural environments through positive and unlabeled learning"13:48
kanzurenmz787_: http://www2.ift.ulaval.ca/~pgiguere/papers/IJCAI2013.pdf13:48
nmz787_oh, http://www.aaai.org/ocs/index.php/IJCAI/IJCAI13/paper/download/6848/707613:48
jcorganor just add .sci-hub.bz to url domain13:49
chris_99nmz787_, oh interesting, so you energise a circuit, and it emits RF back?13:49
chris_99is that kind of what RFID does?13:49
jcorganvery similar13:49
nmz787_man, I really need to get a wall-poster or math symbols to be able to translate these kind of papers to python, etc13:51
nmz787_wall-poster OF math symbols13:51
jcorganthe signal received by an antenna will be re-radiated back out at an amplitude and phase dependent on the impedance of the circuit it is attached to13:51
nmz787_huh, very cool13:53
chris_99oh, this is the thing, where they use a high power CW transmitter13:53
chris_99right?13:53
jcorganyes13:53
chris_99ah yeah i think i recall it now13:53
jcorganthink of the RFID tag you use for highway tolls collection13:54
nmz787_if I made some kind of microfluidic, I think mossmann is local enough that I could find him/someone he works with to test out i.e. saltwater vs normal water13:54
chris_99nmz787_, check this http://arxiv.org/abs/1606.01196 'The Desktop Muon Detector' uses SiPM apparently13:54
nmz787_like a pulse train of salinity13:55
nmz787_hmm, what is the sensor?13:56
nmz787_from sensl.com ?13:56
chris_99i think so13:57
-!- Guest85738 is now known as augur13:58
nmz787_huh13:58
chris_99it's a SiPM13:58
nmz787_it might be worth trying one of those with some phosphor to make a monochromator or something13:59
nmz787_(or for my xray detection project)13:59
chris_99oh cool, do you have an x-ray tube?13:59
nmz787_nah, but FIB beams induce bremstralung x-ray formation14:00
chris_99ah interesting, do FIB machines, detect x-rays?, i thought they used an electron gun and detected electrons14:02
nmz787_if you have an x-ray detector they both can (FIB or SEM)14:03
chris_99and you mean you can get an image of what you're scanning with that?14:03
nmz787_SEM shoots electrons which interact and cause photons to be emitted, FIB shoots atoms which posses electrons which upon impact interact with the electrons of the impacted material and then photons are emitted14:04
chris_99yeah, i mean the x-rays, you can obtain images with that?14:04
nmz787_well image is usually detected with electrons... but there are definitely direct-ion imagers/detectors (since ions are charged)14:05
nmz787_you can make an x-ray emission map, sure14:05
chris_99ah neat14:05
nmz787_and call it am image14:05
nmz787_.wik x-ray microprobe14:05
yoleauxnmz787_: Sorry, that command (.wik) crashed.14:05
nmz787_.wik Electron microprobe14:06
yoleaux"An electron microprobe (EMP), also known as an electron probe microanalyzer (EPMA) or electron micro probe analyzer (EMPA), is an analytical tool used to non-destructively determine the chemical composition of small volumes of solid materials." — https://en.wikipedia.org/wiki/Electron_microprobe14:06
chris_99cool, never heard of that before14:06
nmz787_slower but more resolution: Energy-dispersive X-ray spectroscopy14:07
nmz787_.wik Energy-dispersive X-ray spectroscopy14:07
yoleaux"Energy-dispersive X-ray spectroscopy (EDS, EDX, or XEDS), sometimes called energy dispersive X-ray analysis (EDXA) or energy dispersive X-ray microanalysis (EDXMA), is an analytical technique used for the elemental analysis or chemical characterization of a sample. It relies on an interaction of some source of X-ray excitation and a  …" — https://en.wikipedia.org/wiki/Energy-dispersive_X-ray_spectroscopy14:07
nmz787_basically uses a monochromator I think14:07
nmz787_so scans to get res, but is slower... also beam-dwell of electrons/ions will cause sample destruction eventually14:07
chris_99ahh14:08
nmz787_so then you can try lowering the voltage of your e-beam to do electron energy loss spectroscopy14:08
nmz787_.wik electron energy loss spectroscopy14:08
yoleaux"In electron energy loss spectroscopy (EELS) a material is exposed to a beam of electrons with a known, narrow range of kinetic energies. Some of the electrons will undergo inelastic scattering, which means that they lose energy and have their paths slightly and randomly deflected." — https://en.wikipedia.org/wiki/Electron_energy_loss_spectroscopy14:08
nmz787_I think you can also do something with charge conduction... where at each spot you ramp the voltage and record the point at which the rate of conduction is saturated14:09
nmz787_different materials will have different properties14:09
nmz787_something like that14:10
chris_99https://en.wikipedia.org/wiki/Electron_spectrometer thats intriguing, so you measure how much the electrons bend14:12
chris_99does anyone know if you can actually commercially buy a single photon source off the shelf, or are they either ridiculously expensive or..14:14
nmz787_I saw this http://hackaday.com/2016/05/07/single-photon-source-for-quantum-computing-and-experimentation/14:21
chris_99i think cambridge created a single photon led, not sure what temperature theirs requires though.  I'm just reading a bit about quantum dots and how they relate to single photon sources14:23
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Remote host closed the connection]14:40
paskykanzure: it sounds a bit crazy; do they at least do that & sync after each N batches? based on https://github.com/BVLC/caffe/blob/master/docs/multigpu.md caffe just averages the gradient, not the weights14:42
kanzurehmm.14:43
kanzurei saw one that syncs after each training round14:43
pasky(i.e. it is like if you had k-times larger batch)14:43
kanzurei'ts possible that i'm wrong... maybe averaging weights, averaging the models, etc. is somehow the right thing to do..... i don't know.14:44
kanzureit's certainly a cool result if it works.......14:44
paskythat sounds unlikely to me, the local optima can be very different weight configurations14:44
paskyon the other hand if you average often enough, it may not disrupt the convergence too much14:45
kanzure"stochastic gradient descent"14:45
kanzure"One weird trick for parallelizing convolutional neural networks" http://arxiv.org/pdf/1404.5997.pdf14:45
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap14:50
kanzure"GeePS: Scalable deep learning on distributed GPUs with a GPU-specialized parameter server" http://www.istc-cc.cmu.edu/publications/papers/2016/GeePS-cui-eurosys16.pdf14:52
-!- gene_hacker [~chatzilla@c-50-188-129-64.hsd1.or.comcast.net] has joined ##hplusroadmap14:53
-!- augur [~augur@76-218-206-38.lightspeed.sntcca.sbcglobal.net] has quit [Remote host closed the connection]14:59
kanzuregene_hacker: streety and nmz787_ were thinking about some microfluidic liquid chromatography stuff, pump stuff and photoactivated polymers14:59
nmz787_hi gene_hacker15:00
streetyaveraging the gradient would seem to make more sense than the weights. Maybe it works out to be the same in the end. The parallelization I've done has been cross-validating a model architecture using multiple GPUs.15:09
streetyEasy approach for quick gains15:10
-!- cynsia [cyn@2600:3c03::f03c:91ff:fe79:d5d] has quit [Quit: Leaving]15:14
kanzure.wik bulk synchronous parallel15:14
yoleaux"The Bulk Synchronous Parallel (BSP) abstract computer is a bridging model for designing parallel algorithms. It serves a purpose similar to the Parallel Random Access Machine (PRAM) model. BSP differs from PRAM by not taking communication and synchronization for granted." — https://en.wikipedia.org/wiki/Bulk_synchronous_parallel15:14
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Leaving]15:15
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap15:16
kanzurewtf. researchgate emails me after they saw me loading that last pdf link in here.15:18
kanzureand they specifically reference that document15:18
kanzurethis is the most insanely draconian science site i have ever seen15:19
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Ping timeout: 260 seconds]15:21
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap15:34
nmz787_whoo, I think I have rastering!15:34
nmz787_http://gifmaker.cc/PlayGIFAnimation.php?folder=2016082109xmUPfh351sTmeLbHTBEyG9&file=output_Wqx5Ua.gif15:41
-!- ldlework [~ldlework@unaffiliated/ldlework] has quit [Quit: co'o ro pendo be mi]15:43
-!- ldlework [~ldlework@104.131.139.45] has joined ##hplusroadmap15:43
-!- ldlework [~ldlework@104.131.139.45] has quit [Changing host]15:44
-!- ldlework [~ldlework@unaffiliated/ldlework] has joined ##hplusroadmap15:44
-!- ldleworker [~ldlework@104.131.139.45] has joined ##hplusroadmap15:47
-!- ldleworker [~ldlework@104.131.139.45] has quit [Remote host closed the connection]15:48
-!- jenca [~lajenca@104.131.139.45] has joined ##hplusroadmap15:50
nmz787_http://i.imgur.com/Kb20d7u.gif15:52
nmz787_if the last one didn't work15:52
chris_99what is it?15:53
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Quit: text1.text]15:53
nmz787_same model, the microfluidic pump example15:53
nmz787_just each slice has been turned into a rastered bitmap15:54
nmz787_looking down in the Z axis at the model15:54
nmz787_in this case I have 6 layers for the model... but in practice I think this model needs fabbed with only 2 layers15:54
-!- jenca [~lajenca@104.131.139.45] has left ##hplusroadmap ["Textual IRC Client: www.textualapp.com"]15:56
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap15:56
nmz787_or maybe I need another layer between the top pneumatic channels and the lower pumped-channel15:56
chris_99i'm not sure what a microfluidic pump is, is there a video of one, you'd recommend15:56
nmz787_yielding 2 patterned layers, a spacer layer, and the top layer15:56
nmz787_err, the top layer being one of the patterened layers15:57
chris_99https://www.youtube.com/watch?v=VryS7LvjWaA is it something like this?15:57
nmz787_it is just like your 3 fingers pressing a tube of tooth-paste15:57
nmz787_one after the other15:57
-!- ldlework [~ldlework@unaffiliated/ldlework] has left ##hplusroadmap ["0.4.2"]15:57
nmz787_causes flow in one direction15:57
chris_99ah ok, so what are you using, some kind of solenoid?15:58
nmz787_https://en.wikipedia.org/wiki/Peristaltic_pump15:58
chris_99oh, ive seen those15:58
nmz787_not sure why the linear pump they have isn't animated, sorry15:58
chris_99oh neat, i hadn't seeen the linear style15:59
chris_99one thouse15:59
chris_99*though15:59
nmz787_the top 3 'finger' channels are pneumatic, they expand downward and cut off flow in the lower tube (perpedicular)16:00
chris_99so it literally pushes different parts of the microfluidic chip?16:00
nmz787_if you expand them one after another, it causes flow16:00
nmz787_yeah16:01
chris_99cool16:01
nmz787_the channels can be solid, and the interface either silicone or very-thin glass which can flex enough to cut off a very shallow channel below it16:01
nmz787_or everything can be silicone16:02
chris_99ahh16:02
chris_99so i'm a bit confused, what's the advantage of a peristaltic style pump, over just forcing the liquid through with a syringe?16:03
chris_99oh is it the fact you can build it all as part of the single chip, rather than needing external hardware16:05
-!- augur [~augur@noisebridge130.static.monkeybrains.net] has joined ##hplusroadmap16:10
-!- ebowden [~ebowden@2001:8003:100e:c500:f03a:fc32:3a73:da6] has joined ##hplusroadmap16:11
-!- ebowden [~ebowden@2001:8003:100e:c500:f03a:fc32:3a73:da6] has quit [Changing host]16:11
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap16:11
streetyI think you still need external hardware to pressurize the three fingers16:11
chris_99hmm true16:12
streetyprobably get some metering ability16:12
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Ping timeout: 260 seconds]16:15
nmz787_well they are also composed of 3 valves... and valves are useful in general to be able to make.16:21
nmz787_but a macro air valve can probably control lots of micro pumps16:21
nmz787_so you can probably drive 10s or 100s of parallelized stuff with the same macro stuff16:21
chris_99ahh16:21
chris_99cool16:21
chris_99are you planning on fabricating your chip, yourself?16:23
nmz787_and yeah metering, if you can control the size of the pump, and number of 'fingers', you can control output flow and ripple more or less finely16:23
nmz787_yeah, I have a spin coater and a laser etcher CNC16:23
nmz787_(unfortunately they are not mated to each other)16:23
streetydon't you need the line to each finger to be individually addressed?16:23
chris_99what's the spin coater used for in this context? to lay down a very clean amount of the PDMS or whatnot?16:25
-!- augur [~augur@noisebridge130.static.monkeybrains.net] has quit [Read error: Connection reset by peer]16:26
streetyI think it may also be used to apply a layer of photoresist16:26
streetynot sure what nmz787_ has planned though16:26
nmz787_streety: yeah, but you can hook line A up to many 'needs actuated finger A's16:26
nmz787_chris_99: yeah photoresist, then that gets exposed with laser, then use as-is possibly or cast PDMS on, then peel off and sandwich16:27
streetygood point16:28
nmz787_chris_99: also though, I want to convert exported rasters to the FIB tool format, so I can etch things with that... and maybe possibly do deposition16:28
streetycan you then block individual pumps with another line?16:28
chris_99ahh, could you use it to directly spin the PDMS, and etch directly? or would that not work16:28
nmz787_streety: yeah I guess if it was called for, sure16:28
nmz787_I've read of multiplexers16:28
-!- augur [~augur@noisebridge130.static.monkeybrains.net] has joined ##hplusroadmap16:29
nmz787_but I think there is probably some kind of max-levels of valving/muxing you can do, before you lose strength16:29
nmz787_I want to translate this to use my BRL-CAD python lib: https://gist.github.com/nmz787/83a826ca4d40011262e7382b46c2257016:32
nmz787_which outputted this http://imgur.com/PKFXtCV16:32
nmz787_based on this http://www.seas.harvard.edu/capasso/wp-content/uploads/publications/sciencejune3-2016.pdf16:32
nmz787_Metalenses at visible wavelengths: Diffraction-limited focusing and subwavelength resolution imaging16:33
-!- augur [~augur@noisebridge130.static.monkeybrains.net] has quit [Ping timeout: 244 seconds]16:33
nmz787_I want to try simply FIBbing some silicon then coating it with gold or maybe imprinting it into some plexiglass16:34
-!- augur [~augur@noisebridge130.static.monkeybrains.net] has joined ##hplusroadmap16:35
chris_99neat!  Does your spin coater (or spin coaters in general) have a way of measuring the thickness of what you spin16:36
chris_99so you could spin for a specific duration etc16:36
chris_99until it hits a certain thickness16:37
nmz787_nah, you have to do that with another machine, I guess in some high-volume setups they may be integrated16:41
chris_99ah gotcha16:41
chris_99i was trying to find which came first spin art, or spin coating ;) but can't seem to find a decent history of spin coating16:42
nmz787_either go through curing then cut and image it, or you use an interferometer on the gel/thin-film, or I think another way too, time of flight or something maybe16:42
chris_99mmm16:43
chris_99i was thinking if you had a laser or something point at it, during the spining16:43
nmz787_yeah maybe it is just deflection16:43
nmz787_the last tech I was thinking of16:43
chris_99would it be fairly easy to diy a spin coater, or is there some complex aspects im not thinking of, like precise motor control16:44
streetythere is a video on the applied science channel16:45
streetyhttps://www.youtube.com/watch?v=321tptQ8PrU16:46
chris_99cheers16:46
kanzure.title16:46
yoleauxDesign and build a spin coater - YouTube16:46
-!- Malvolio is now known as Sooka16:46
chris_99i love krasnows stuff16:46
chris_99so for making the microfluidic slide, you say you spin the photoresist, then you pour the PDMS into it, i guess? does that mean you then have to spin a layer of PDMS too, to cover the bottom part of the slide?16:48
-!- Sooka is now known as Malvolio16:48
streetyI think you peel the PDMS layer off after it has set and then apply that16:52
nmz787_spin resist, pattern resist, develop resist, pour on silicone and spin (I think, or maybe people usually just pour and let it settle/self-level with gravity) then peel off.16:55
nmz787_for the interface layer, I think people just spin silicone, then cure it, then peel and use it for the in-between layer16:56
-!- augur [~augur@noisebridge130.static.monkeybrains.net] has quit [Remote host closed the connection]17:00
chris_99sorry im confused what covers the slide? so the liquid doesn't fall out i mean.  is the interface layer the thing that does that?17:03
nmz787_that or a glass microscope slide17:04
nmz787_or a combination depending on your sandwich of layers17:05
chris_99gotcha17:05
chris_99do you have to use a vacuum chamber for the resin to degas it, prior to spinning17:05
nmz787_ideally yeah17:13
nmz787_here's a grey-scaled output, based on the thickness of the region in the slice: http://imgur.com/a/HKjKW17:30
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Remote host closed the connection]17:50
-!- duper [~super@ca4.blackcatz.org] has joined ##hplusroadmap17:58
-!- Orpheon [~Orpheon@27.160.62.81.dynamic.wline.res.cust.swisscom.ch] has quit [Read error: Connection reset by peer]18:03
nmz787_even finer slices, with grey-scale: https://giphy.com/gifs/iARXASzJvXR3W18:12
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-udbgtidsefpkkkrk] has quit [Quit: faggory daggory doo!]19:11
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-sfldulfrkatdqige] has joined ##hplusroadmap19:17
-!- irseeyou [~irseeyou@c-67-168-101-20.hsd1.wa.comcast.net] has quit [Ping timeout: 240 seconds]19:20
-!- irseeyou [~irseeyou@c-67-168-101-20.hsd1.wa.comcast.net] has joined ##hplusroadmap19:20
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-sfldulfrkatdqige] has quit [Changing host]19:23
-!- anoncicada [anoncicada@unaffiliated/anarcanon] has joined ##hplusroadmap19:23
-!- anoncicada [anoncicada@unaffiliated/anarcanon] has quit [Changing host]19:23
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-sfldulfrkatdqige] has joined ##hplusroadmap19:23
-!- c0rw1n- [~c0rw1n@223.100-136-217.adsl-dyn.isp.belgacom.be] has joined ##hplusroadmap19:25
-!- c0rw1n [~c0rw1n@193.47-244-81.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 265 seconds]19:27
-!- Gurkenglas [Gurkenglas@dslb-178-005-223-183.178.005.pools.vodafone-ip.de] has quit [Ping timeout: 260 seconds]19:37
-!- ebowden [~ebowden@147.69.145.152] has joined ##hplusroadmap19:41
-!- ebowden [~ebowden@147.69.145.152] has quit [Changing host]19:41
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap19:41
-!- augur [~augur@noisebridge130.static.monkeybrains.net] has joined ##hplusroadmap19:48
-!- Gurkenglas [Gurkenglas@dslb-178-005-223-183.178.005.pools.vodafone-ip.de] has joined ##hplusroadmap19:59
-!- winsen [~winsen@unaffiliated/winsen] has quit [Quit: Leaving]20:00
-!- ArturShaik [~ArturShai@31.29.9.62] has joined ##hplusroadmap20:14
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Quit: time to haxor]20:22
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap20:24
-!- irseeyou [~irseeyou@c-67-168-101-20.hsd1.wa.comcast.net] has quit [Quit: Leaving.]20:56
-!- maaku [~quassel@173-228-107-141.dsl.static.fusionbroadband.com] has quit [Remote host closed the connection]21:27
-!- maaku [~quassel@173.228.107.141] has joined ##hplusroadmap21:28
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Quit: text1]21:34
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap21:36
-!- andytoshi [~andytoshi@unaffiliated/andytoshi] has joined ##hplusroadmap21:41
kanzurebloop21:50
andytoshiblop21:52
kanzurehm.21:53
kanzurewhere is eudoxia22:19
kanzurenevermind. seems that he was alive at least 8 days ago.22:19
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Ping timeout: 244 seconds]22:33
-!- jcorgan [~jcorgan@unaffiliated/jcorgan] has quit [Ping timeout: 244 seconds]22:39
-!- jcorgan [~jcorgan@64-142-68-61.dsl.static.sonic.net] has joined ##hplusroadmap22:39
-!- jcorgan [~jcorgan@64-142-68-61.dsl.static.sonic.net] has quit [Changing host]22:40
-!- jcorgan [~jcorgan@unaffiliated/jcorgan] has joined ##hplusroadmap22:40
-!- yashgaroth [~yashgarot@2602:306:35fa:d500:f5e0:f867:a11d:8d52] has quit [Quit: Leaving]22:43
-!- justanot1eruser [~Justan@unaffiliated/justanotheruser] has quit [Ping timeout: 240 seconds]22:47
-!- drewbot [~cinch@ec2-23-20-26-181.compute-1.amazonaws.com] has quit [Remote host closed the connection]22:52
-!- drewbot [~cinch@ec2-54-196-87-168.compute-1.amazonaws.com] has joined ##hplusroadmap22:53
-!- justanotheruser [~Justan@unaffiliated/justanotheruser] has joined ##hplusroadmap23:00
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]23:06
-!- ebowden [~ebowden@147.69.145.152] has joined ##hplusroadmap23:07
-!- ebowden [~ebowden@147.69.145.152] has quit [Changing host]23:07
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap23:07
nmz787_welp, got the raster export committed!23:21
-!- ebowden [~ebowden@unaffiliated/ebowden] has quit [Remote host closed the connection]23:27
-!- ebowden [~ebowden@147.69.145.152] has joined ##hplusroadmap23:29
-!- ebowden [~ebowden@147.69.145.152] has quit [Changing host]23:29
-!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap23:29
nmz787_.tell abetusk I implemented raster generation from BRL-CAD model slices... would love some tips for getting the points exported to G-code: https://github.com/nmz787/python-brlcad-tcl23:32
yoleauxnmz787_: I'll pass your message to abetusk.23:32
--- Log closed Mon Aug 22 00:00:32 2016

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