#qi-hardware IRC log for Thursday, 2014-03-06

DocScrutinizer05oh dear! I *hate* TV science for the plebs. "the earth needs exactly one year to complete once circle around the sun" What a fluke! *cough* NO you *idiots*! one year is exactly the time the earn needs to complete a circle around the earth01:18
DocScrutinizer05s/exactly the/defined as the/01:19
wpwrakyou mean the sun around the earth ? or did you watch a program of those worthless aristotle-doubters ?01:24
DocScrutinizer05s/around the earth/around the sun/ damnit01:25
wpwrakwell, what a lucky coincidence, isn't it :)01:26
DocScrutinizer05Harald Lesch: "man sollte sich nicht wundern dass die Katze da wo die Augen sind Loecher im Fell hat"01:27
wpwrakwell, evolution explains that. the cats who had the holes elsewhere couldn't see and were run over by chariots. (they didn't have cars back then)01:28
wpwrakDocScrutinizer05: quoted "one shouldn't be surprised that cats have their eyes where there are holes in the coat"01:30
whitequarkyou mean like01:30
whitequarkoh, coat as in fur.01:31
wpwrakthese evolved a little later ;-)01:31
DocScrutinizer05the puddle thinks: "what a miracle that the gound is so comfortably formed to accomodate each single one of my bumps and dents and bays"01:31
wpwrakand yes, coat as in fur :)01:31
wpwrakeffect is rather pleased that cause took care of matching it so neatly01:32
DocScrutinizer05go figure: the earth would only need 0.8 years to turn around the sun once01:35
wpwrakwell, if something bumped into it, it would01:36
wpwrakor, rather, it could01:36
wpwrakalso, a calendar year isn't strictly the time the earth takes for one rotation around the sun01:37
DocScrutinizer05no, actually it wouldn't. Just Greenwich had a lot of trouble with their leap seconds then01:37
wpwrakso in a way, they're quite correct :)01:37
wpwrakmake something big enough bump into it. well, preferably several things, as one hit big enough to change the orbit like that wouldn't be much fun01:38
wpwrakat 0.8 AU, closeness to venus may also become a bit of a problem01:39
DocScrutinizer05yeah, and since the moon is constantly decelerating the earth, in a few billion years it will take 3 days from one sunrise to the next, right?01:40
DocScrutinizer05(btw I'm already accomodating to that effect, so I'm ahead of my time)01:42
wpwrakwell, that depends a lot on what time reference the alien visitors who'll observe that will use 01:42
DocScrutinizer05that TV crap for sure did use the SOL3/earth/human definition of "year"01:43
whitequarkare you guys doing drugs again? :]01:44
wpwrakbecause by the time you're at 3 present days, the sun will have done some unpleasant thing ...01:44
wpwrakwhitequark: naw, we quit all those downers :)01:45
DocScrutinizer05hmm, dunno. Not sure how long it will take. But I'm sure another myth is nonsense: that the moon eventually will vanish in empty space since it separated from earth too much01:45
whitequarkDocScrutinizer05: well, it's not going to exactly vanish, it'll still make circles around the sun01:46
DocScrutinizer05you're aware that the efect that increases the moon's orbit distance is diminuishing with orbit distance?01:47
DocScrutinizer05and with earth own rotation decelerating, too01:48
whitequarkoh, you're totally right01:48
wpwrakwell, it is presently escaping at about 3.8 cm/year01:48
whitequarkit'll just continue until moon and earth are tidally locked.01:48
whitequarkjust like a set of others satellites01:49
DocScrutinizer05I bet it's rather easy to calculate moon orbital distance when eart rotation came to a halt, like moon's rotation did long ago01:49
whitequarkyup, just add rotational energy of earth to the energy of gravitational field01:50
DocScrutinizer05"caqme to a halt" -> "... in relation to moon"01:50
DocScrutinizer05so, about maybe 50 times what a day takes now01:51
DocScrutinizer05we can make a little bet, who's closest wins the pot01:52
DocScrutinizer05I say: current rotation speed / 57 is where moon gets in sync with earth01:53
DocScrutinizer05funny that moon seems to already have stopped completely. I wonder how that happened, and if the moon went back and forth in the end, like a pendulum01:55
DocScrutinizer05or maybe it still does?01:56
DocScrutinizer05with a period of a few dozen million years?01:57
wpwrakWiki Pedia, PhD, says the period will be some 47 days: http://en.wikipedia.org/wiki/Orbit_of_the_Moon#Tidal_evolution02:01
DocScrutinizer05dang, 10 days off. Still the best guess of all02:12
DocScrutinizer05wait, does my initial guess count?02:13
DocScrutinizer05hmmpf, 50 billion years02:22
DocScrutinizer05>>about 2.3 billion years from now, the increase of the Sun's radiation will have caused the Earth's oceans to vaporize,<<  that's unfair ;-P02:23
wpwrakyou'll be really good at "flappy bird" by then :)02:23
DocScrutinizer05what's "flappy bird"?02:24
wpwrakand 2.3 Gyr should be enough to build a decent radiation shelter with air conditioning ;)02:24
wpwrakgoogle for it :)02:24
DocScrutinizer05I'm not concerned about the shelter. It's the lack of tide acceleration that spoils the bet02:25
DocScrutinizer05actually that 47 factor never will happen02:26
DocScrutinizer05or maybe in a few 100 billion years02:26
DocScrutinizer05and for sure not as a steady endpoint state02:27
whitequarkDocScrutinizer05: in about 600 My C3 photosynthesis will become impossible02:27
wpwraki knew it ! us carnivores will outlive the vegetarians :)02:29
whitequarkno, plants will outlive us all02:31
whitequarkbut not by long02:31
whitequarkwell, I suppose 600 My is rather enough to gtfo out of here02:35
eintopfDocScrutinizer05: "flappy bird" is a boring popular fancy android game. But it's a remake of an very old game.07:12
CYB3Reintopf: what is the name of the original game07:48
eintopfmhh, I can play it on my rockbox device but that's not a bird... it's a helicopter08:03
eintopfbut it must be a cover, too08:04
eintopfCYB3R: http://www.rgcd.co.uk/2011/05/chopper-drop-zx-spectrum.html08:10
eintopfthat looks like the original one 08:10
eintopfand it's better than flappy bird08:10
CYB3Reintopf: Chopper Drop looks a lot more complex08:11
eintopfCYB3R: ehm yes08:19
eintopflarsc: thanks :-)08:20
eintopfnow I have something to fill my freetime08:20
CYB3RA pity, i don't have flash plugin installed08:21
zrafawpwrak: hey13:12
zrafawpwrak: testing ubbctl with SIE. I have some strange behaviour I do not understand yet13:13
zrafaI did the unbind of mmc driver13:13
zrafaI run ubbctl and it runs and show me the pin status13:14
zrafaright now I have this :13:14
zrafa(after some high-low of some pins) :13:14
zrafaJlime$ ./ubbctl 13:14
zrafanPWR=1 DAT2=1 DAT3=0 CMD=0 CLK=Z1 DAT0=0 DAT1=013:14
zrafaBut, when I check ubb pins with the tester I see that DAT3, DAT1 and VDD have 3V13:15
zrafawpwrak: now I set all in 0s13:18
zrafaJlime$ ./ubbctl 13:18
zrafanPWR=0 DAT2=0 DAT3=0 CMD=0 CLK=0 DAT0=0 DAT1=013:18
zrafabut I still have 3V in DAT3, DAT1 and VDD13:18
zrafamaybe I do not understand what ubbctl is for? :P13:19
wpwrakhmm, that seems a little odd indeed. lemme check what i did there ...13:28
wpwrakVDD = 3 is normal, it's nPWR, i.e., active-low. so nPWR = 0 means power is on13:30
wpwrakand in any case, you can have all sorts of values on VDD if nPWR = 1, since it connects via pull-ups to the data lines13:31
wpwrakso you must load VDD if you want to know if it really can supply power13:31
wpwrakseeing 3 V on DAT1/3 is odd. with DAT1=0 and DAT3=0, they should be driven low13:33
wpwrakif you run  ubbctl dat0=0 dat3=0  do they change ?13:33
wpwrakoh, and regarding drivers, is there an at86rf230 driver in your kernel ? if yes, that one could be in control of the port13:34
zrafawpwrak: yes, the pins changed in software, but I still see 3V in those13:41
qi-bot[commit] Werner Almesberger: ubbctl/README: fix typo in example (master) http://qi-hw.com/p/ben-blinkenlights/77a7a3813:42
zrafawpwrak: I am using a jlime installed a long ago13:43
zrafawpwrak: Jlime$ find /sys | grep at8613:43
zrafaJlime$ lsmod | grep at13:43
zrafaJlime$ dmesg | grep at8613:43
zrafaJlime$ 13:43
wpwrakalright, nothing there13:44
wpwrakvery weird that you'd get 3 V13:44
zrafawpwrak: maybe SIE board is too different in those pins in comparison with ben13:44
wpwraklemme see what else could get in the way ...13:44
zrafaI need to go out for two hours, but I will keep in the channel. THanks a lot for your help . I will continue testing the software and will keep you informed :)13:45
wpwrakthat would be an explanation. but it shouldn't be.13:45
wpwrakif yuo can try with your ben, that would help13:45
wpwrakdo you have an UBB ?13:46
zrafaaround ten from you from tuxbrain :)  So it is great to test the software13:46
zrafaI will bring myben to uni. I am doing those tests here13:46
wpwrakperfect (2x) :)13:47
wpwrakone possibility would be that the pins are configured as interrupt. that may disable gpio output. ubbctl doesn't consider this possibility yet. lemme try it out ...13:48
qi-bot[commit] Werner Almesberger: libubb/include/ubb/regbase.h (PDTRGx): add trigger mode registers (master) http://qi-hw.com/p/ben-blinkenlights/3f5c05214:42
qi-bot[commit] Werner Almesberger: ubbctl/: also support decoding and setting of interrupts and alternate functions (master) http://qi-hw.com/p/ben-blinkenlights/9e223f714:42
zrafaJlime$ ./ubbctl 15:43
zrafanPWR=1 DAT2=1 DAT3=1 CMD=1 CLK=1 DAT0=1 DAT1=115:43
zrafawith this status all the pins have 3V, but no DAT2.15:43
zrafawith this status : 15:44
zrafaJlime$ ./ubbctl 15:45
zrafanPWR=0 DAT2=1 DAT3=1 CMD=1 CLK=1 DAT0=1 DAT1=115:45
zrafathe same, all the pins have 3v, but no DAT215:45
zrafawith this status : 15:46
zrafaJlime$ ./ubbctl 15:46
zrafanPWR=0 DAT2=0 DAT3=0 CMD=0 CLK=0 DAT0=0 DAT1=015:46
zrafathe same behaviour as the first tests, vdd, dat3 and dat1 have 3v. THe others 0v15:47
zrafawith this status : 15:49
zrafaJlime$ ./ubbctl 15:49
zrafanPWR=1 DAT2=0 DAT3=0 CMD=0 CLK=0 DAT0=0 DAT1=015:49
zrafamore odd: just dat2 and CLK does not have 3v. THe others all have 3v15:50
zrafawpwrak: ^15:50
zrafawpwrak: btw, with jlime toolchain I needed to build ubbctl with -lgcc_s in order to link without a horrible error gcc+ld showed15:53
wpwrak-lgcc_s seems odd :) if i have jlime vs. owrt issues, i usually just use -static ;-)16:40
wpwrakyour results are all very odd. that's without anything collected on the outside ? on a SIE, and using the latest ubbctl ? (i.e., with the interrupt configuration i just added now)16:41
zrafawpwrak: no latest with the interrupt configuration :)16:57
zrafawpwrak: I will try16:57
zrafain a while16:57
zrafawpwrak: okey, testing latest ubbctl17:14
zrafathe only pins which ubbctl looks to control is CMD and DAT017:15
zrafait means that :17:15
zrafaJlime$ ./ubbctl 17:16
zrafanPWR=0 DAT2=0 DAT3=0 CMD=1 CLK=0 DAT0=0 DAT1=017:16
zrafadoes pin CMD has 3V17:16
zrafawith :17:16
zrafaJlime$ ./ubbctl 17:16
zrafanPWR=0 DAT2=0 DAT3=0 CMD=0 CLK=0 DAT0=0 DAT1=017:16
zrafapin CMD has 0v17:16
zrafaTHe same with pin DAT0 :17:16
zrafaJlime$ ./ubbctl 17:17
zrafanPWR=0 DAT2=0 DAT3=0 CMD=1 CLK=0 DAT0=1  DAT1=017:17
zrafaping DAT0 has 3v17:17
zrafaand with :17:17
zrafaJlime$ ./ubbctl 17:17
zrafanPWR=0 DAT2=0 DAT3=0 CMD=1 CLK=0 DAT0=0 DAT1=017:17
zrafapin DAT0 has 0v.17:18
zrafabut the others looks like always 3v (DAT3, DAT1 and vdd)17:18
wpwraklemme check the SIE schematics. maybe the port is different17:20
wpwrakdid you get your ben ?17:20
zrafaDAT2 always is 0, still if I set it to 1 with ubbctl.17:20
zrafawpwrak: no yet, at uni now. Maybe I check from home :)17:20
zrafawpwrak: short status:17:20
zrafaWith nPWR=0   DAT3, DAT1, and VDD are always having 3V. NO matter if I set them 1 or 017:21
zrafaWith nPWR=0   DAT2 is always having 0v. No matter if I set it 1 or 017:22
wpwrakon the ben, ou get weird levels if nPWR = 1 because the inputs then float and everything but CLK is connected via the pull-ups (which float as well in this case)17:22
zrafaWIth nPWR=0   I can set CMD and DAT0, 1 or 0, and it gives me 3v or 0v after set by software with ubbctl17:22
wpwrakyour SIE is v2 ?17:22
zrafaSo CMD and DAT0 looks like have proper behaviour set by ubbctl17:23
zrafalet me check sie version17:23
zrafaboard says : "SACK_BOARD RC1"   2010031717:24
wpwrakSIE only connects CMD, CLK, and DAT0 ;-)17:24
wpwrakthat's actually enough to make SWD work. you just need to tweak ben.c a little17:25
wpwrakSIE also doesn17:25
wpwrak't have the nPWR switch, so power is always on17:25
zrafawpwrak: that explains why I can control DAT0 and CMD like I want (on/off 3v/0v) via ubbctl right? :)17:26
wpwrakyes, and CLK should be good, too17:26
zrafawpwrak: ahh.. right. Let me check17:27
wpwrakthey're on the same port and pins as on the ben17:27
wpwrakin anelok/libswd/ben.c you see that i already use CLK and DAT0. just move SWD_DIO over to CMD and you should be fine17:28
zrafawpwrak: I can control CLK as well :)17:28
wpwrakwell, unless .... lemme check what happens with nPWR ...17:28
wpwraknPWR may not be even bonded out17:31
wpwrakso it's safe to let libswd do whatever it's doing with it now17:32
zrafawpwrak: cool. THanks a lot !!! I had not checked sie schematics before tests, because I do not understand them mainly :)17:33
zrafawpwrak: I guess that Rodo did today or yesterday, but I have not seen him since. anyway I will give him the news about we can continue with the tests of libswd using sie as well, still if he already knows :)17:34
wpwrakand once you control your little kinetis chips, you can even make your own SWD programmer ;-)17:36
zrafaI was hoping that your swd programmer would be useful for those chips as well ;-)))17:39
wpwrakit shuoldn't need much change. try loading a led blinker and see how it goes17:41
wpwrakthere's a blinker in ybox/fw/ybox.c revision b0b84100ca03c094509c105a994edc1e0e08cb7317:44
wpwrakblinks pin 30 on port E. note the setting of clock gates17:45
wpwraklater revisions use my gpio library. but that one is a little harder to understand17:46
whitequarkooooh, this is neat: http://hackaday.com/2014/03/06/make-a-plastic-bender-using-stuff-you-already-have/20:03
whitequarkextremely simple plastic bending machine20:03
wpwraknot very well-defined bend angle, though. but certainly an idea that could be developed20:09
wpwrakDocScrutinizer05: you should like this: http://scanlime.org/2008/09/using-an-avr-as-an-rfid-tag/20:54
wpwrakespecially the "Power filtering using the AVR’s die capacitance" bit20:54
DocScrutinizer05sorry, I'm currently in a mood where I rather like this: http://www.youtube.com/watch?v=cCBwLJjzDJQ   and this: http://www.youtube.com/watch?v=lVgTTAukMOU20:58
wpwrakah yes, tax declaration time ...21:01
DocScrutinizer05nah, intriguer's galore, and I have to defend myself and the democratic majority of maemo21:02
wpwrakyou seem to have a talent for getting drawn into these things ...21:03
DocScrutinizer05nah, it's teh age old BS of accountants trying to tell techstaff how to do their job21:04
DocScrutinizer05this time they even gone really busy, doing that admin business by themselves and without informing or coordinating with sysops21:05
DocScrutinizer05neither with maemo council which is highest authority in maemo21:06
wpwrakwhen i was a little kid, it was all the rage to form "gangs". so i thought i'd make mine. designed a hierarchy (me on top, naturally), made insignia from shiny promotional stickers, even defined gang rules along with a penal code for those violating them. of course this thing didn't even last a day. but man, was i serious about it.21:17
wpwrak"maemo council which is highest authority in maemo" sounds very much like that ;-)21:18
DocScrutinizer05wpwrak: I don't mind what you think it sounds, it's an elected entity to manage all business that the ~30k people in maemo community can't manage on their own21:31
DocScrutinizer05it's up to council to e.g. decide about packages that have to get deleted from repos for one reason or another (dangerous, rogue, copyright...), or about general decisions regarding management of the maemo infra 21:33
wpwrakwow, it's bigger than i thought. so not you're into serious politics. shall i congratulate or pity you ? ;-)21:34
DocScrutinizer05and unlike your little gang, it's pretty hard to find somebody foolish enough to waste his time for that thing, and it lasts since 4 or 5 years now21:34
DocScrutinizer05and thanks, a litle pity is appreciated. It's a PITA21:45
wpwrakah well, politics. i guess you either have what it takes to be good at them, or you're a decent human being :)21:48
DocScrutinizer05kinda to the point, yes. Just that council never been about politics21:49
DocScrutinizer05at least not meant to21:49
DocScrutinizer05it's as much about politics as is steward of a multi-apartment house21:50
DocScrutinizer05when you consider painting the stairway with new color politics, then yes, council is all about politics21:52
wpwrak(steawrd) bwah. if THAT sort of snake pit is no politics for you ...21:52
DocScrutinizer05but in the end the steward is just the idiot to get his ass kicked by everybody 21:53
wpwrakthat's what it does indeed seem to boil down to21:54
DocScrutinizer05and now the accountant has decided that they rather paint the stairway themselves and wants to kick the steward21:54
wpwrakwhich would be you, i guess21:55
DocScrutinizer05but here similarities end, since maemo accountant has no saying in steward's tasks21:55
DocScrutinizer05well, actually no, I'm also maemo infra manager, which basically means I'm the steward of the sysops. And the accountant wants to manage sysops now and discard me from that position21:57
wpwrakthank them and be happy to have more time for more productive tasks ?21:57
DocScrutinizer05but accountant has no saying in that, it's council's domain to decide about it21:58
wpwrakwell, they decided to bypass that already. so the coup seems perfect. hey, sysop isn't much fun unless you get to play BOFH every day.21:59
DocScrutinizer05sorry, nope. Neither the sysops nor the council nor I think that maemo will survive when the management gets done by those guys21:59
DocScrutinizer05and again, it's not the cashier to decide how system management gets done, or by whom22:00
wpwrakthe best way to prove their inability is to let them run the show for a bit. then, when the villagers come for them with torches and pitchforks, you can raise up from the ashes like a phoenix. or maybe you'll have lost interest in such silly games by then and someone else does it.22:01
DocScrutinizer05we been there, seen the chaos22:01
DocScrutinizer05roundabout 12 months ago22:01
wpwrak;-) then alert the troops and let them battle it out22:01
DocScrutinizer05one of our sysops in a chat with me: >>i supposes that you should ask which will be the new planned situation after you leave... techstaff is not a bitch. We are not supposed to accept any director or coordinator<<22:04
wpwrakif all else fails, just move all the stuff you care about to github and let them manage their little playground22:04
DocScrutinizer05we're sooooo || close to doing exactly that22:05
wpwraknaw, seriously, this sounds like exactly the sort of job you want to get rid of. you may not think of it like that now, but you may consider it the highlight of all of 2014 in hindsight, even eclipsing the start of neo900 production and your invitation to give the opening keynote at MWC 2015 ;-)22:06
DocScrutinizer05believe me, no hindsight needed. It's just my sense of responsibility that makes me engage in this bullshit dogfight22:08
DocScrutinizer05and of course the guys who ask me to NOT quit22:08
DocScrutinizer05first of all the sysops22:09
wpwrakmaybe take a little walk through the city, consider the monuments honoring famous sysops. that should clear your head :)22:09
DocScrutinizer05if they'd think they are happy without me around, I'd leave with speed of light22:09
DocScrutinizer05I don't care if those sysops are famous22:10
DocScrutinizer05I see them doing a brilliant job and all I can do to help them I will do. They asked me to not quit and continue holding the keys and the accountability for what's going on inside our infra22:11
DocScrutinizer05they are volunteers like I am22:11
wpwrakbtw, how is neo900 coming along. it's been a long time that you haven't mentioned anything technical. has it become boring, with all issues solved ? of waiting for another batch of pcbs ?22:11
DocScrutinizer05our common interest is to keep maemo_the_OS alive22:11
DocScrutinizer05I received a mail from my notary today, that we can make an appointment to start founding a Limited22:12
wpwrakwhee ! so all technical work rests until this is sorted out ?22:13
DocScrutinizer05yesterday I ahd a envelope in my letterbox with a bare PCB from Nik22:13
wpwrakah, nice :)22:13
wpwrakdoes it look good ?22:13
DocScrutinizer05a few days ago I found a RGB LED driver that does common anode22:14
whitequarkthat was hard?22:14
DocScrutinizer05not really22:14
DocScrutinizer05err dang, sorry, common cathode?22:15
DocScrutinizer05we couldn't find RGB LEDs for the lp552322:15
DocScrutinizer05I need a electrical-current sensor on hall basis - i.e. with a EST of <50mR, and capable of detecting <10mA. Any suggestions?22:17
DocScrutinizer05ESR even22:17
DocScrutinizer05alternative concepts welcome22:18
DocScrutinizer05e.g a opamp working on 50µV would also be fine22:19
wpwrakfor lab use ? part of the device ?22:19
DocScrutinizer05part of device22:19
DocScrutinizer05sensing supply current of modem22:20
DocScrutinizer05so it shall have low ESR, detect <10mA reliably (or as small a current as it gets, the 10mA are a ballpark nr), and withstand >3A bursts22:22
wpwrakwhat bandwidth ?22:27
DocScrutinizer05doesn't matter much22:29
DocScrutinizer05shouldn't miss bursts in the range of several milliseconds22:29
DocScrutinizer05precision also doesn't matter22:29
wpwrakok, DC ;-)22:29
DocScrutinizer05yes, basically DC22:29
DocScrutinizer05if the sensor averages over a window, the window time may even be in the range of seconds22:30
whitequarkDocScrutinizer05: use a battery monitor? let me remember how is it properly named22:31
DocScrutinizer05I recall to have seen a few SMD hall sensors made for current sensing22:31
whitequarkgas gauge or something22:31
DocScrutinizer05gas gauge22:31
wpwrakyou need enormous gain, though. that's a bit tricky. there are these, but they only go to 100x: http://datasheets.maximintegrated.com/en/ds/MAX4376-MAX4378.pdf22:31
whitequarkeg http://www.ti.com/product/bq2700022:31
wpwrakah, sweet memories :)22:32
DocScrutinizer05but actually yeah, why not simply use a dedicated bq2720022:32
DocScrutinizer05whether there's one or two 10mR in series to battery, as seen from modem, shouldn't matter much22:33
DocScrutinizer05oops, 20mR?22:33
DocScrutinizer05whitequark: tthanks, how silly of me22:33
whitequarkDocScrutinizer05: you're welcome22:33
DocScrutinizer05ok, any suggestion for a dirt simple yes/no detector for >20mW@50R in the frequency range of 500..2800MHz?22:42
DocScrutinizer05probably abusing any arbitrary opamp will already do, the clamp diodes should suffice to rectify the RF22:43
whitequarkDocScrutinizer05: you know those cellphone straps with LEDs blinking when there's RF traffic?22:44
whitequarkI doubt you can go simpler than that22:44
DocScrutinizer05yep, I know them. Did you know that they are outlawed since they introduce massive harmonics?22:44
whitequarkany articles on that? never heard of22:45
whitequarkperhaps not outlawed in RU22:45
DocScrutinizer05one of them even reliably caused my Nokia 2110 to lose connection 22:45
DocScrutinizer05you can't find them anymore here. I had a damn hard time to find one a maybe 5 years ago22:46
DocScrutinizer05thus I don't have any of those critters. A pity22:46
DocScrutinizer05but yeah, there are powered ones, with a transistor inside22:46
whitequarkwell, I mean, they're not exactly complex inside22:46
whitequarkno, I don't think so22:47
whitequarkI recall a dude in my highschool recreating one with a LED, a RF diode and two pieces of wire of precise length22:47
DocScrutinizer05I have a powered one, a key ring22:47
whitequarkit worked like a charm22:47
DocScrutinizer05yes, that's the simple wariant. LED + schottky diode + wire22:47
DocScrutinizer05variant even22:47
wpwrakhmm, i'd try to keep my RF signal path clean22:52
wpwrakif i want to honeypot my phone, i'll fire up real RF equipment and do it with that, but for just making sure it doens't talk to strangers, simply grounding its power supply ought to be sufficient22:53
DocScrutinizer05wpwrak: the concept is different. We don't want to shut down the phone since that's not what it's made for. We want to monitor and analyze its activities22:54
DocScrutinizer05grounding power supply would also disable the receiver22:55
wpwrakwell, a phone goes on the air all the time. any malware worth its silver would simply piggy-back on that22:57
wpwrakso unless you actually build a full receiver into it, you still won't see what's really going on22:57
DocScrutinizer05no, a phone goes "on the air" only on area handover (not even cell handover, when both cells are same area code), and on expiry of T3212 timer22:59
DocScrutinizer05and that location update is always looking pretty identical and short22:59
DocScrutinizer05and any malware would need to store and compress data for several hours and then send it via high bandwidth hidden channel when any such even makes modem go on the air23:01
DocScrutinizer05which is pretty hard to establish in modem firmware23:01
DocScrutinizer05NB that we consider linux APE safe per definition23:02
DocScrutinizer05it's only about modem firmware23:02
DocScrutinizer05in the end it's a feature for the tinfoil hat fraction23:04
wpwrakhmm, may be more useful to leave some room for SDR that actually has a chance to decode what the modem sends23:04
DocScrutinizer05or we ship your set of a dozen private BTS23:05
DocScrutinizer05with the phone23:05
wpwrakwell, talk to harald about that ;-)23:05
DocScrutinizer05sorry, I'm short on time for that fun23:06
wpwrakbut seriously, merely detecting that the phone goes on the air seems more like a placebo to me. besides, anyone who really cares can carry a separate detection device. actually, they probably already own one :)23:06
DocScrutinizer05again, I don't care23:06
wpwrakand it seems that "serious" implants already hide in those regular transmissions. i didn't follow the TAO revelations in detail but some of them go clearly in this direction23:07
DocScrutinizer05we don't employ a low-ESR FET (or 3) to switch off the modem hard, we use the money for detectors rather23:08
wpwrakhmm, ground the TX antenna ? :)23:08
DocScrutinizer05THE FSCK there is no TX antenna23:08
wpwrakwell, TX path23:09
wpwrakhowever it is implemented23:09
DocScrutinizer05why should I?23:09
DocScrutinizer05that's even more silly than anything else23:09
DocScrutinizer05and more expensive and hard to do23:09
wpwrakwell, it's an alternative. and maybe you could divert TX into a detector.23:10
wpwrakdunno about cost, though23:10
--- Fri Mar 7 201400:00

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