#qi-hardware IRC log for Saturday, 2014-04-12

DocScrutinizer05latest hottest news¡ NSA exploited heartbleed since >2 years01:30
DocScrutinizer05It's shocking to see which delay TV news have when it comes to IT stuff01:31
DocScrutinizer05I think I heard first TV news about heartbleed yesterday01:32
DocScrutinizer05hehe http://xkcd.com/1345/01:52
eintopfwpwrak: so now I use some timings from the datasheet... then I need a driver_data for each chip which have the right waiting from the datasheet06:05
eintopfTo minimize this influence in06:07
eintopfExtended Operating Mode RX_AACK, refer to Section 7.2.3 RX_AACK_ON - Receive with06:07
eintopfAutomatic ACK on page 51, the acknowledgment frame response time can be reduced to06:07
eintopf32 ¼s.06:07
eintopfforget it, it's wrong06:08
eintopfwpwrak: http://alles-schallundrauch.blogspot.co.at/2014/04/sind-die-passagier-in-diego-garcia.html08:17
wpwrak(timings) or just pick the worst case. i think they should all be pretty similar.08:19
wpwrak(diego garcia) i like the idea of hiding an iphone in one's behind. that ought to really hurt :)08:21
eintopfthere was a lot of electronic experts on this plane08:27
eintopfit's a job with risk if this is true :/08:27
wpwrakthere's a factual error in the story,though: the date it was posted.08:27
wpwraksome freescale production engineers, it seems08:28
wpwrakthe risk is putting a lot of your people on the same flight. this means that either freescale don't have a policy against this or they didn't care08:29
eintopfbut want they want from the freescale poeple? I think about like iron man movie to built a iron man body :D08:35
eintopfthat's a new theory08:35
wpwrakthere's actually a documentary (film) that describes exactly what happened. look for Millennium (1989)08:40
eintopfah, time travels08:46
wpwraki you don't read the plot, then it's more fun to watch08:46
wpwrakthe story has a few surprises08:47
larscright... the exif location information in an image is "hard to forge"09:13
eintopfI remember some friend on facebook... he said, he is surprised because he copied/rename his photo file local and then upload it and facebook still know the location09:20
eintopfAfter completing the frame transmission, indicated by IRQ_3 (TRX_END), the PLL settles back09:34
eintopfto the receive frequency within tTR11 = 32 ¼s in state PLL_ON.09:35
eintopfand that's not true09:35
eintopfI waiting 32 us if i am still in the TX_BUSY state09:35
eintopfgrml :(09:35
wpwraklarsc: for the vast majority of homo sapiens, that statement would be true ;-)09:37
wpwrakmaybe the TRX_END was here before you started sending ?09:38
wpwrakalso, how do you initiate the sending ? i.e., are you sure it actually has started by the time you wait those 32 us ?09:39
wpwrakah, don't mind about the latter. i was thinking of something else.09:40
eintopfwpwrak: I don't know but I need to figure out what's going on there09:40
eintopfI need some hours09:41
eintopfthen it will be working *I hope*09:41
wpwrakoh, another thing: you can't count on always getting a TRX_END after RX_START. see page 53, that little arrow between "Promiscous Mode" and "Reserved Frames"09:46
wpwrakalso, on FCS (CRC) invalid, the TRX_END should be omitted. though it says differently elsewhere.09:46
eintopfi don't use the RX_START interrupt09:47
eintopfRX_START could bring some performance to read the buffer while filling09:48
eintopfsee page 15209:49
eintopfbut I don't know if I can do that under linux09:49
eintopfi don't think that I can disable the SCLK while async transmit09:49
eintopfehm s/transmit/receiving/09:50
eintopfwpwrak: I can reproduce the step when TX_BUSY doesn't change to PLL_ON. It's when a RTT takes 10 ms and I try a ping with less than 10 ms interval09:51
eintopfthen I don't getting out of TX_BUSY09:51
eintopfbut I can't send two xmit09:52
eintopfI mean the xmit callback should be synchronized09:52
eintopfi think i need some interrupt busy spinlock10:03
eintopfyes, that's a bad word... spinlock10:03
eintopfwpwrak: because I have a chain of spi_async calls10:08
eintopfand at the end I clear the interrupt busy flag10:09
eintopfand if it's set I return IRQ_NONE; insetad IRQ_HANDLED10:09
eintopfI don't clear the irq status register10:10
wpwraknow we're getting somewhere ;-)10:25
wpwrakbeing stuck in TX_BUSY seems odd, though10:25
wpwrak"The following packages will be upgraded: ... apt-transport-https" *grin*10:30
wpwrakwhy could that possibly be, eh ? :)10:30
whitequarknothing in debian-security10:31
wpwrakmaybe they did it earlier10:31
wpwrakhmm, nice storm. the building is shaking and the clouds are racing across the sky. funny that the weather forecast missed that one.10:33
eintopfthe building is shaking?10:40
Action: whitequark read that as "binding"10:40
Action: whitequark is debugging a seemingly random memory corruption. indeed, the binding is shaking!10:40
wpwrakwhitequark: since you languages are safe, it must be a hardware problem :)10:42
whitequarkwpwrak: it's a binding to a C library10:42
whitequarkwhen in doubt, blame C!10:42
wpwrakeintopf: yup. 15th floor. wobbles a bit when there's a lot of wind.10:42
whitequarkI mean seriously, something is broken on the interface between libffi and libsodium. *shrug*10:42
eintopfwpwrak: page 64, I can start to read the buffer at TRX_END but when I am finish and I am still in BUSY_RX_AACK, then I need to wait the full 192 + 384 us?11:55
eintopf192 = the stupid 802.15.4 timing to sending an aack then 384 = sending the ack frame11:55
eintopfokay it's not stupid :(11:56
eintopfwpwrak: maybe I should work with jiffies there11:58
eintopfsave a jiffies at TRX_END11:58
eintopfcheck then wait when I finished the frame_read11:58
eintopfthen save another jiffies11:58
eintopfcalc the us how long it takes11:59
eintopfthen make a substraction of (wait_aack_send + tansmit_aack) - frame_read_takes_us 11:59
eintopfand then wait the rest11:59
eintopfsounds good?11:59
wpwrakwell, first ... wait do to what ? switch to PLL_ON ?12:01
eintopfI getting a TRX_END while BUSY_RX_AACK12:02
eintopfpage 6312:02
eintopfthen TRX_END occurs12:02
eintopfmeans, I can read out the frame buffer12:02
wpwrakyou already have the TRX_END :)12:02
eintopfah yes12:02
eintopfyes, but then I read out the frame buffer12:03
eintopfand wait's on completion12:03
eintopfokay that's a little bit more complex first I read out how long the frame is, then I read the full frame...12:03
eintopfbut simple we say I read out the frame buffer12:03
eintopfbut this takse not long, because on BUSY_RX_AACK after TRX_END the rf231 send outs the ACK Frame12:04
eintopfand then I need to wait at completion time12:04
eintopfand then I clear the interrupt_bussy flag12:05
wpwrakby the way, do you have this ? http://www.atmel.com/images/AVR2022_swpm231-2.0.zip12:06
eintopfthat's the programming exmaple?12:07
wpwrak(interrupt busy) you mean the flag in the rf231 ? or something in the driver ?12:07
wpwrakit's a lot of things :)12:07
wpwrakincluding message sequence diagrams12:07
wpwrakand yes, a bit of code, too. hidden between the charts ;-)12:09
wpwrakbwahaha. NOW they put out a storm warning. more than an hour after everything is calm and nice again.12:14
wpwrakgrmbl. with that storm i'll get seasick ...19:50
wpwraksometimes i wonder if certain companies post job offers specifically looking for an "overengineer" ...20:13
wpwrakhmm. that radio has about 250 registers ...21:01
wpwrakyet they forgot the subspace modulation ...21:02
wpwrakthe radio also has a CMD_PING. alas, without payload21:35
DocScrutinizer05seasick? subspace?22:55
wpwrakseasick: because the building is shaking when there's a strong wind. should reach 80-90 km/h tonight.22:57
wpwraksubspace: http://en.wikipedia.org/wiki/Subspace_(Star_Trek)22:57
DocScrutinizer05dang, you're living on top of a skyscraper?23:02
DocScrutinizer05actually I know such shaking only from really really high skyscrapers23:03
DocScrutinizer05or 60m high light posts that are only a maybe 2m diameter at base23:03
DocScrutinizer05or cranes(?)23:04
wpwrakmy office is on the 15th floor. not quite skyscraper but tallish. and high enough to notice storms.23:04
wpwrak(or earthquakes :)23:04
wpwrakyeah :)23:05
DocScrutinizer05hmmm, strange. I've been in a number of 10+ floors, incl OM-apartment, when there were storms. Hardly ever noticed any movement23:06
DocScrutinizer05maybe your skyscraper is a special type of building23:07
DocScrutinizer05the usual 'joy' with KDE akonadi and nepomuk: http://privatepaste.com/e5edfdde6b23:09
wpwrakit's relatively thin. only some 8 m wide. so it flexes easily.23:09
DocScrutinizer05then yes, I can see  how you might get seasick in 15th floor23:10
wpwrakwell, the base is wider. but it narrows at the ... 4th floor i think23:10
wpwrakand it's very exposed. so the wind has all the attack surface it wants23:11
DocScrutinizer05and that's also where your terrace with your BBQ and generator is located?23:11
wpwrakthat's one floor up. and no, no generator. but my acids are there :)23:11
DocScrutinizer05then I also can understand why all rocks fly away from there23:11
wpwrakwell, i hope the wind didn't blow them away :)23:12
DocScrutinizer05ever considered placing a windmill there instead of a gasoline generator?23:12
DocScrutinizer05I bet you'd be the first one to place a windmill in BUE23:13
DocScrutinizer05and we could spot your place in google satview then23:14
wpwraki can help you with that ...23:14
DocScrutinizer05I wonder why you need blinds in 15th floor23:16
wpwrakthe building in the middle https://www.google.com/maps/@-34.6119828,-58.4124362,231m/data=!3m1!1e323:16
wpwrakhmm, no zoom in those urls23:17
DocScrutinizer05this one? http://wstaw.org/m/2014/04/13/plasma-desktopnG1775.png23:18
wpwrakyes, sun. too bright during daytime (my screens face the window)23:18
wpwrakicon too small :)23:18
wpwrakand no, the one to the right23:19
DocScrutinizer05http://wstaw.org/m/2014/04/13/plasma-desktopnG1775.png or that one?23:19
DocScrutinizer05ugh http://wstaw.org/m/2014/04/13/plasma-desktopoo1775.png23:20
wpwrakthat's the same ?23:20
DocScrutinizer05took a while to upload23:20
DocScrutinizer05I been too fast23:20
wpwrakyes ! correct23:21
wpwrakthere are four apartments per floor. i'm in the northwestern one23:22
wpwrakthe one with the smaller northern "outer" terrace (and the larger southern "inner" terrace)23:22
DocScrutinizer05ooh, so this http://wstaw.org/m/2014/04/13/plasma-desktopgB1775.png is not your table?23:23
wpwrakthe little "house" between the terraces is the head of the stairs. and in the middle of the building is a larger "house" with some of the infrastructure stuff, water tanks, pumps, etc.23:23
DocScrutinizer05aaah even23:25
DocScrutinizer05prolly the ideal place to mount the windmill on top23:25
wpwraknaw, that's on the shared terrace in the middle. (well, or rather "no used except for infrastructure things")23:25
wpwrakthe new google maps may actually have an image from after i moved in23:28
wpwrakthere's a dark spot where my grill is located and a brighter spot where i keep a box (thought the color seems wrong)23:29
DocScrutinizer05the wonders of (image)stitching, or you got really futuristic architects in argentina? http://wstaw.org/m/2014/04/13/plasma-desktopIT1775.png23:29
wpwrakthat's what happens when the storm breaks the barrier of sound :)23:30
wpwrakon bing the building looks even thinner :)23:31
DocScrutinizer05hehe. might be beneficial for you, I wondered if you need the blinds because of this particularly wry building23:31
DocScrutinizer05I haven't counted but seems it has even more than 15 floors23:32
DocScrutinizer0518, now I counted them23:32
wpwrakwell, street level is 0, terrace would be 16. no floors below street level23:33
wpwrakanything else are optical illusions or incrementer bugs :)23:33
DocScrutinizer05anyway seems there are higher buildings in your neighborhood. Thus the blinds?23:34
DocScrutinizer05I hate when other people can look into my windows23:34
DocScrutinizer05that's why I always keep the "curtains" closed - actually it's deco paper pinned to the windows23:36
DocScrutinizer05need to change that stuff for fresh ones eventually23:36
DocScrutinizer05it suffers from my smoking23:36
DocScrutinizer05hmm, dark spot grill?23:38
wpwrakthey're not very close. i don't feel observed. no, the blinds are only for the sun23:38
wpwraki have an "oil drum" grill, black. like this: http://www.hogardesign.com.ar/articulos-chulengo-%C2%A8ne%C2%A8/485-2328/index.html23:40
wpwrakthey're great when there's wind. you can just close them and nothing will disturb 'em.23:41
DocScrutinizer05and you think it's there between the "house" and the white spot?23:42
wpwrakthat's where the real thing is located, yes. dunno if that spot is the grill or some artefact, though23:43
DocScrutinizer05dang, I always forget you're living on the other hemisphere. got puzzled when trying to find out which time of day that photo been taken23:45
wpwrakthe sun is always in the north :)23:45
DocScrutinizer05would prolly kill me23:46
DocScrutinizer05and it moves "from right to left", correct?23:46
wpwrakyou hardly notice23:46
DocScrutinizer05hardly notice? hardly!23:47
DocScrutinizer05I'd probably live in a constant feeling of time running backwards23:47
wpwrakyou're the first person who even mentions that. i think we automatically adjust within a few sunrises and sunsets and aren't even consciously aware of it23:48
wpwrakwell, maybe except for sailors23:48
DocScrutinizer05maybe my mind works in a certain way not exactly standard23:48
DocScrutinizer05I suspect since long time that my filter layer does less abstraction than usually23:49
wpwrakwe probably tell the day mainly by height of the sun and whether it has passed the zenith or not. so rough guess / memory plus exact position23:49
DocScrutinizer05some think this is also the root cause of autism23:49
wpwrakwhat's much more noticeable are changes in day length23:50
wpwrakyeah, different data processing23:50
DocScrutinizer05then OTOH I actually adapt pretty easily to new stuff, like gas and break pedal swapped in one of my former cars23:52
DocScrutinizer05luckily it been an automatic23:53
DocScrutinizer05so no clutch to operate23:53
wpwrakwhat ? pedals swapped ? what sort of madman does that ?23:55
DocScrutinizer05a one legged woman23:55
wpwrakokay, so not a madman ;-)23:55
DocScrutinizer05I can't recall anymore if I used left or right leg to drive this car23:56
DocScrutinizer05which is a bit concerning since I owned it longer than a year23:57
DocScrutinizer05I think I tried both23:57
DocScrutinizer05but actually the gas pedal been too far left for confortable operation with right foot23:58
DocScrutinizer05thrust? ;-P23:59
--- Sun Apr 13 201400:00

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