#qi-hardware IRC log for Friday, 2012-12-07

kristianpaulha same ben issue with macs http://lwn.net/SubscriberLink/528112/9773efa029df260d/01:19
e2580Only a few days left to order an open source hardware encrypted storage device on indiegogo www.indiegogo.com/CryptX2 check it out02:14
wpwrakhmm, not going too well ...02:21
wolfspraulI found this nice project reading through the recent llvm dev meeting... http://llvm.org/devmtg/2012-11/02:23
wolfspraulit's called Shang HLS http://llvm.org/devmtg/2012-11/Hongbin-Generating.pdf02:24
wolfspraulsources are here https://github.com/OpenEDA/Shang/blob/master/README.md02:24
wolfsprauluniversity of illinois open-source license, reading...02:25
wolfspraulalways those licenses :-)02:26
wpwrakhuman stupidity may be incurable :)02:26
wolfspraulah it's the same as the llvm license, bsd-like02:27
wolfspraulalright then02:27
wolfspraulat some point something like that shang (or c-to-verilog or legup) may be useful to me as a higher-layer to feed fpgatools02:27
wpwraklooks very interesting, yes02:27
wpwrakbring down another barrier of entry02:28
wolfspraulI will continue to focus on the bits, and access all the features in the chip02:31
wolfspraulbut then I obviously need to plug something in above02:31
wpwrakyup. a nice translator from a popular language would go a long way towards making these strange fpga critters even more approachable02:33
rohwpwrak: well.. i dont bother too much with the languages.. rather the hostile development env.02:51
rohwhen using fpgas is more like mcu and cpus... use an opensource-compiler like llvm or gcc and be fine, have some loader tool like avrdude... then i guess the userbase will widen dramatically02:51
rohs/is/are/02:51
rohi got literarly piles of develboards in access here, fpga, mcu, some dsp i guess... but most popular is the avr for me (and most others here) .. i guess only because its very easy, streamlined, uncomplicated and you know you can work on your code and hardware, and will not that likely debug your development environment02:53
wolfspraulthat makes perfect sense02:54
wolfspraulbut I do see all sorts of tools slowly improving, high quality open stuff02:54
rohi know some people who love to talk meta-compiler-theory for hours.. but that doesn solve a broken one ;)02:54
wolfspraulthe good stuff will slowly bubble up02:55
wolfspraul:-)02:55
rohwolfspraul: i sure hope, that one day i can just install some packages and be set up to compile verilog or vhdl from a git repo into a bitstream to upload with another tool02:55
wolfspraulyes of course. and that's just about the special case of fpgas, I also think there is a market for nice chips right on digikey and elsewhere that are designed and manufactured in an entirely open way03:11
wolfspraulopen design, open tools, open cell libraries, etc. so that there can be an ecosystem of chips where it's easy to customize or optimize towards use cases.03:12
rohsure. also i find it neccessary to be able to do full CI03:12
wolfspraulgeneral-purpose fpgas of course also being an option, why not03:12
rohcompiling, testing every revision...03:12
wolfspraulat least those are made and readily available in latest-gen processes03:13
wolfspraul(if only from altera and xilinx nowadays)03:13
rohi want to build a test rig for openwrt stuff soon.. we'll see if we can make it work for some series of soc... test every build03:13
wpwrakgrmbl. i hate metal fatigue combined with fragile cables. one separated neatly without any damage visible from the outside.05:05
DocScrutinizer05(tantalum mark) use a decent DVM and see what residual charge you find in the component ;-)10:41
DocScrutinizer05odds are that this is the correct polarization10:41
DocScrutinizer05(I found residual charge in the magnitude of several volts in 'normal' electrolytic capacitors, after they ripe for literally years in a box and bever were used)10:43
DocScrutinizer05never*10:44
DocScrutinizer05and even after shortcircuit that charge builds up again during minutes10:45
DocScrutinizer05so in a certain sense electrolytic capacitors are quite similar to batteries10:45
DocScrutinizer05also: after several years of non-usage a usual electrolytic capacitor can de-format and you need to format it again, which is a process identical to charging a battery10:47
DocScrutinizer05formatting building up AlOx dielectric membrane again, which seems to slowly deteriorate over the years when capacitor never is charged10:48
DocScrutinizer05sorry if last statement maybe is not entirely precise, it's knowledge I acquired 1969, I still recall my dad explaining it to me. Never checked back10:51
DocScrutinizer05it's just funny as for 99.999% of my factoids I don't recall the source they came from10:56
wpwrakDocScrutinizer05: must have been a memorable factoid :)11:57
DocScrutinizer05yeah11:57
wpwrakcaps degenerating in such a way after long disuse could explain some equipment failures11:58
DocScrutinizer05I asked him why he's keeping his "Metz mecablitz" electron flasher plugged in to mains and active for whole night11:58
DocScrutinizer05dang, that been a monster11:59
DocScrutinizer05like 5kg suicase with shoulder strap11:59
DocScrutinizer05with electromechanical step-up circuit12:00
DocScrutinizer05"wagnerscher hammer"12:00
DocScrutinizer05basically a small relay buzzing all the time12:00
DocScrutinizer05and alead acid battery12:01
wpwrakwow. did its brightness correspond to the effort that went into making it ?12:02
DocScrutinizer05yep12:11
DocScrutinizer05seems impossible to find a photo of that thing nowadays12:12
DocScrutinizer05anyway it had a dish reflector made of alu, with a ring shaped electron flash tube12:13
DocScrutinizer05diameter of the dish some 20+ cm12:13
DocScrutinizer05I loved to do all kinds of weird experiments with that flash, that's why I was very interested in what my dad was doing with it when he 'reformatted the capacitors'12:17
DocScrutinizer05wpwrak: I guess I already told you about the "Balsen Keksdosen Effekt"?12:25
DocScrutinizer05you can hear the photons bouncing off a thin tin lid, like found on cookie tin boxes12:26
wpwrak(flash) fun :) single-use bulbs ?12:27
DocScrutinizer05nah, no way12:27
wpwrak(balsen) photons ? why do i sense a sudden pull on my leg ? :)12:28
DocScrutinizer05>>...with a ring shaped electron flash tube12:29
DocScrutinizer05I'd guess some 150Ws, on hindsight12:30
wpwrakpretty powerful12:30
DocScrutinizer05what's with your leg?12:37
wpwrakdo you know the expression "pulling someone's leg" ?12:38
DocScrutinizer05nope12:38
wpwrakah :) i means that you're trying to fool someone.12:38
DocScrutinizer05hah!12:38
DocScrutinizer05do your math! a 1m^2 mirror 'feels' ~1pond force from sunlight's photons bouncing off12:39
DocScrutinizer05iirc12:40
DocScrutinizer05also energy on 1m^2 is ~1000W12:40
DocScrutinizer05a 150Ws electron flash has an effective power of 150/10^-3 .. 150/10^-6 roughly12:42
wpwrakah, you mean the whole output of the flash. i see. yeah, that should work12:44
lekernel1pond?12:48
DocScrutinizer05aka 1g12:48
DocScrutinizer05though gram is not a proper force unit12:48
wpwraklekernel: a unit old people use :)12:49
lekernelbtw, if you have a directional light source like a LED, does it serve as a (very weak) propulsor?12:51
wpwrakwhy not ? anything else your little spaceship radiates could be made such that it doesn't counteract the led.12:55
DocScrutinizer05photon propulsion is a long considered interstellar drive principle12:59
wpwraklekernel: maybe you can hack this one up for EHSM: http://en.wikipedia.org/wiki/Nuclear_photonic_rocket13:00
DocScrutinizer05http://www.daviddarling.info/encyclopedia/P/photon_propulsion.html13:00
wpwrakDocScrutinizer05: e.g., stationary lasers pushing an "unpowered" craft, as in the novel "accelerando"13:00
DocScrutinizer05yeah13:01
wpwrakof course, as long as our incapable scientists still fail to solve something as basic as FTL, interstellar travel will still suck ...13:03
Action: DocScrutinizer05 idly wonders how an externally laser propulsed(?) spaceship will *brake* when it reached its destination :-o13:04
DocScrutinizer05hard impact? X-P13:05
wpwraki think in the novel they did it by ejecting the screen and have it act as a mirror13:06
DocScrutinizer05haha, nice one13:07
DocScrutinizer05did they consider the doppler effect though?13:08
wpwrak(on detaching the screen/sail) http://en.wikipedia.org/wiki/Solar_sail#Interstellar_flight13:10
DocScrutinizer05a considerable amount of the laser's energy will accelerate the mirror even further, while the light bouncing back from mirror to ship is shifted far into the red13:10
wpwrak(doppler) dunno13:10
wpwrakwell yes, it'll be pretty inefficient13:10
wpwrakyou could also pick destinations with a very high escape velocity.. then slow down just enough to get caught, and gradually bleed off speed over the next few centuries.13:13
--- Sat Dec 8 201200:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!