#qi-hardware IRC log for Tuesday, 2011-06-07

tuxbrainhow to check 2a?00:00
wpwrakdmesg -> the pastebin of your choice00:00
wpwrak(if it doesn't say anything about at86rf230, then there's no wpan driver in there)00:00
tuxbraindmesg|grep at86rf230 returns no results00:01
wpwrakif the kernel drivers are active while the user space tools are trying to run, they're giving each other a merry chase00:01
tuxbrainok then previous checks done00:02
wpwraklet's see what the exact message (and case) is ....00:02
wpwrakyeha, "at86rf230" is a good string to search for00:04
DocScrutinizerM1 working on 16bpp? DUH!00:04
DocScrutinizerexciting little board, but honestly, 16bpp for video?00:04
wpwraktuxbrain: did you update the tools to the latest version from git ?00:04
tuxbrainwpwrak: yes00:04
DocScrutinizer24 ok, but 1600:05
wpwraktuxbrain: alright. then run atrf-proxy on the ben00:05
tuxbrainbut I'm having problems on my desktop to find gls dev packages00:05
wpwrakDocScrutinizer: meanwhile, ubb-vga produced quite decent pictures with 16 bits ;-)00:05
tuxbrainubuntu 10.04 here00:05
DocScrutinizerwpwrak: no problem, as long as you don't have video00:06
tuxbrainfuck is gsl00:06
DocScrutinizerI.E. moving pictures00:06
wpwraklibgsl0-dev ?00:06
wpwrakaye, GSL. GNU Scientific Library ;-)00:07
wpwrakDocScrutinizer: i don't think motion makes all that much of a difference00:08
tuxbrainok desktop compilation ok going to compile to ben00:08
DocScrutinizerwpwrak: it does00:08
DocScrutinizeryou get terrible banding00:08
DocScrutinizerwhich you can't usually spot in still pictures00:09
wpwraktuxbrain: and don't forget the  echo jz4740-mmc.0 >/sys/bus/platform/drivers/jz4740-mmc/unbind  on the ben. else atrf-proxy and the mmc driver will have a lot of fun with each other :)00:09
wpwrakDocScrutinizer: hmm. guess someone has to try :)00:09
DocScrutinizerunless of course you got 4:2:2 YUV ow whatever00:09
DocScrutinizerI guess M1 is uing 5:5:5 RGP00:10
wpwrakDocScrutinizer: err ... did i say "16 bits" ? nonsense. it's of course 16 _colors_ !00:10
DocScrutinizersure, for HD00:11
wpwrakDocScrutinizer: this is what you can do with 16 colors plus dithering: http://downloads.qi-hardware.com/people/werner/ubb/vga/ubb-vga-bouquet-dither222.jpg00:11
DocScrutinizerI know what you can do with 3 colors plus dithering00:12
wpwrakgood :)00:12
DocScrutinizermy impact printers all used that technique00:12
DocScrutinizerand Turboprint created some really nice photos00:13
DocScrutinizeron them00:13
tuxbrainok atrf-proxy runnig ..00:13
wpwraktuxbrain: when everything is ready, run on the PC: atrf-id   should report  20b7:1540 ... #79 Sun Jun ...00:13
DocScrutinizeronce you had it calibrated to the 4 color inktape used00:14
DocScrutinizer(4th color = black)00:14
tuxbrainwpwrak: yes it does :)00:14
wpwraktuxbrain: and then  atrf-id -d net:ben   should report  AT86RF231, part 0x03 version 2 manufacturer xxxx001f (Atmel)00:14
tuxbrainyes it does too :)00:15
wpwraknow let's try t send/receive. window A on the PC: atrf-txrx00:15
DocScrutinizerwpwrak: anyway, nice result for a UBB00:15
wpwrakwindow B: atrf-txrx -d net:ben hello00:15
wpwrakDocScrutinizer: i was a bit surprised myself :) the horizontal stripes are timing variations. should be possible to reduce them.00:16
tuxbrain5 bytes payload, CRC OK, LQI 255, ED -53 dBm00:17
DocScrutinizeryou probably have to dig out and re-digest your supertape sources ;-)00:17
DocScrutinizeranyway, o/00:18
wpwraktuxbrain: excellent. now the reverse direction: window A: atrf-txrx -d net:ben00:18
wpwraktuxbrain: then window B: atrf-txrx greetings00:18
tuxbrain9 bytes payload, CRC OK, LQI 255, ED -54 dBm00:19
DocScrutinizerMr Marconi live?00:19
wpwraktuxbrain: kewl. so the basic bits are in good shape.00:20
tuxbrainlet's try spectrum?00:20
wpwraktuxbrain: yeah. atrf-path -g net:ben usb00:20
wpwraktuxbrain: you should see: a blinking green dot in the upper right corner. and some white line00:21
tuxbrainok fanci withe line is dancing on a black box00:22
wpwrakquit with Q00:22
tuxbrainok done00:22
wpwrakthen  atrf-path -g usb net:ben00:22
tuxbrainstuked :(00:23
wpwrakhmm. nothing at all ? not even the green dot ?00:23
tuxbrainno green dot no withe line no reponse to Q00:24
wpwrakinteresting. kill it, then try the other one. does that one still work ?00:24
tuxbrainyes atrf-path -g net:ben usb works00:25
wpwrakokay. exit. atrf-reset -a (will spit out an error, ATUSB_RESET: -84)00:26
wpwrakthen try  atrf-path -g usb net:ben  again00:26
tuxbrainnothing changes still black unresponsible box00:27
wpwrakhmm. searching ...00:27
wpwrakdoes dmesg on the pc say anything interesting ?00:28
DocScrutinizer . . .___ .___ .___ ___ - . - -___ - - -___ . . -___ ___ . - . .___ . -___ -___ .___ . - .___ ___ . . . - . -___00:28
wpwrakDocScrutinizer: whatever :)00:28
tuxbrainthe atrf-resest has thow the error the atusb led blink onces but atrf-path -g usb net:ben still doesn't work00:28
wpwrakgood. atrf-reset does what's expected of it. so far, so good.00:29
tuxbrainno dmesg don't reveal anything00:30
tuxbrainusbfs: USBDEVFS_CONTROL failed cmd atrf-reset rqt 64 rq 2 len 0 ret -8400:30
tuxbrain[705361.792558] usb 2-1: USB disconnect, address 700:30
tuxbrain[705362.070021] usb 2-1: new full speed USB device using uhci_hcd and address 800:30
tuxbrain[705362.237131] usb 2-1: configuration #1 chosen from 1 choice00:30
tuxbrain[705363.790046] usb 2-1: USB disconnect, address 800:31
tuxbrain[705364.580017] usb 2-1: new full speed USB device using uhci_hcd and address 900:31
tuxbrain[705364.755418] usb 2-1: configuration #1 chosen from 1 choice00:31
wpwraklooks good. maybe try the other atusb ?00:32
DocScrutinizerwith all the reconnects?00:32
DocScrutinizeror were those resets?00:32
wpwrakDocScrutinizer: at least one reset, yes00:32
DocScrutinizerI'd try lsusb00:33
wpwrakDocScrutinizer: naw, the tools complain if they don't find the device00:34
wpwraktuxbrain: one possible cause for a hang would be the PLL not locking. atrf-txrx just times out and proceeds in this case, but atrf-path waits.00:39
wpwraktuxbrain: or, alternatively, the interrupt line could malfunction.00:40
wpwraktuxbrain: but let's try the other atusb board first00:40
qi-botThe build was successfull, see images here: http://fidelio.qi-hardware.com/~xiangfu/compile-log/openwrt-xburst.trunk-full_system-06052011-2125/00:40
tuxbrainhe the other atusb doesn't enumerate even reflashing00:41
wpwraktuxbrain: urgh :-(00:41
wpwrakokay, let's debug #1 then ...00:41
wpwraktuxbrain: do yo uhave an oscilloscope that can do at least 10 MHz ?00:42
tuxbrainno :(00:42
wpwraklet's have fun with the full test suite then. cd to ben-wpan/prod/  then  make usb00:43
wpwraksee what happens ;-)00:43
wpwrakah ... first,  touch usb.profile   :)00:44
tuxbrain./atusb net:ben usb00:45
tuxbrain ....  Enumeration\r\033[42;30m PASS \033[0m 00:45
tuxbrain ....  LED00:45
tuxbrainPress [P]ass if the LED blinks, [F]ail if not00:45
tuxbrain\r\033[42;30m PASS \033[0m 00:46
tuxbrain ....  GPIO scan (P_ON)\r\033[41;37m FAIL \033[0m 00:46
tuxbrainmake: *** [usb] Error 100:46
wpwrakhmm, your "echo" sucks :)00:46
tuxbrainany way to improve it?00:46
wpwrakgood. seems that it found something. cat _log   --> pastebin00:46
wpwraktuxbrain: what's your  ls -l /bin/sh  ?00:47
wpwrakah, and my ubuntu gave me dash there. does  /bin/echo -e '\033[7mhi\033[m'  work ?00:50
wpwrakDocScrutinizer: btw, you may want to have a look at this month's community news. there are more pictures there, also from ubb-vga: http://en.qi-hardware.com/wiki/Copyleft_Hardware_News_2011-06-0100:50
tuxbrainyes it gives me a withe bg black fg 'hi'00:51
wpwrak(pastebin) hmm.. MOSI. now that's unusual ...00:51
tuxbrainwhy this sentence on the expert doen't reconforts me....?00:52
wpwrakoh, where there's a hammer, there's a technical solution ....;-)00:54
qi-bot[commit] Werner Almesberger: prod/Common: try to obtain more consistent handling of \033 in echo http://qi-hw.com/p/ben-wpan/6e3edf300:57
wpwraktuxbrain: this should fix the echo problem00:57
tuxbraindudes is late night here , I am about to dislocate my jaw yawning, I will read any posible workarround tomorrow... in two days aprox I will have the production ones (already done, just last payment and shipment left) yes I know I was to be there to do in situ test but a missundertood messages form factory from my part had made them go ahead, so I will do that flashing/tests at home. I'm crosing my fingers to everything works ok , if not ... well let'00:58
tuxbrains see00:58
wpwrakfunny. now i also get that problem with MOSI. maybe i messed up something else :) checking ...00:59
tuxbrainwpwrak: please tellme is all about software....00:59
wpwrak(production) oh dear ... we'll, let's pray to the gods of murphy then ...00:59
wpwrakthe advantage of the production already being done is that you'll either have some ~100 atusb boards you can use as a reference - or none at all to worry about ;-)01:00
wpwrakah, and it would be good if you could find an oscilloscope for tomorrow. we need at least one channel, minimum range 10 MHz.01:02
wpwrakthe scope will be useful for debugging the other board. plus any issues that may pop up with the boards from the factory01:02
wpwrakah, now i see what caused MOSI misbehaviour here. let's try this:01:03
wpwrakatrf-gpio -p HHHHoHxH.HxHxHxHh.ohzoHoHH01:03
wpwrakdoes it complain ?01:03
tuxbrainIt has to?01:04
wpwrakhmm, lemme think about this for a moment :)01:05
wpwrakif you remote the -p, it complains, right ?01:05
tuxbrainif this is  a complain then yes http://pastebin.com/cKwCnmUK01:06
wpwrakyes. it complained about MOSI. very good.01:07
wpwraknow, try again WITH the -p01:07
tuxbrainit complains again01:07
wpwraklooks like a reset problem then01:08
wpwrakdo you have a multimeter ?01:08
wpwrakexcellent. lemme cook up a test sequence ....01:09
wpwrakmeanwhile, you could try this:01:10
wpwrakatrf-gpio 02:8/0x1f reset 02:0/0x1f01:10
tuxbrainnothing apearntly happens01:11
wpwrakhmm. so reset seems to work.01:12
wpwrakwhile atrf-gpio 02:8/0x1f reset 02:0/0x1f; do /bin/echo -n .; done01:13
wpwrakmaybe it's a timing problem01:13
tuxbraina lot of dots fill the shell01:14
wpwrakbut no error ? that would be good then01:14
tuxbrainno no error01:14
wpwrakreset is exonerated then :)01:15
wpwrakdoes   atrf-gpio -p HHHHoHxH.HxHxHxHh.ohzoHoHH01:15
wpwrakstill complain about MOSI ?01:15
wpwrakhow about  atrf-gpio -p delay HHHHoHxH.HxHxHxHh.ohzoHoHH01:16
tuxbrainyes it complains too, 01:17
wpwrakatrf-gpio -p HHHHoHxH.HxHxHxHh.oh0oHoHH01:17
tuxbrainyes it complains too01:18
wpwrakatrf-gpio -p HHHHoHxH.HxHxHxHh.o0zoHoHH01:18
tuxbrainyes it complains too01:18
wpwrak*hmm* very interesting01:19
tuxbrainI'm not kidding :) it complains01:19
wpwrakatrf-gpio -p HHHHoHxH.HxHxHxHh.ohzo0oHH01:20
tuxbrainit complains again01:21
wpwrakatrf-gpio -p HHHHoHxH.HxHxHxHh.ohz0HoHH delay HHHHoHxH.HxHxHxHh.ohzoHoHH '# grr'01:21
tuxbrainagain http://pastebin.com/FpXyLNHi01:23
wpwraksanity check. this doesn't complain, does it ?  atrf-gpio -p HHHHoHxH.HxHxHxHh.ohzhHoHH01:23
wpwrakphew :)01:24
tuxbrainok my fears that you have conenected a chat boot and go to sleep are gone :P01:25
wpwrakand in the last ~5 commands, you always had the -p, right ?01:25
wpwrakyeah, the chat bot could nicely produce some 10^24 patterns for you to try ;-))01:26
tuxbrainman I have c&p whatever you put here not wanna write by hand the HHozhhHH.... thing01:26
wpwrakhehe :)01:26
wpwrakit's still rather mysterious. atrf-gpio 02:8/0x1f reset 02:0/0x1f  still works ?01:27
tuxbrainalso the dots01:28
wpwrakhmm. what these commands did was set all the pins and traces surrounding MOSI to 0. yet it still reads high.01:28
wpwrakafter reset, the transceiver should activate a pull-down, so it should be low (unless the MCU overrides it)01:29
DocScrutinizerwait, the dots still running?01:29
tuxbrainno I have stop them before puting the commands dude :)01:29
wpwraki happen to have a board with a broken reset line. that one fails also with the MOSI, but it also fails the ... reset ... test01:30
wpwraklet's try something a little more aggressive then ...01:30
tuxbrainI have the hammer ready01:31
wpwrakatrf-gpio -p 00000010.00000001.z0zo00001:32
DocScrutinizerthe 'o' look better, indeed01:33
tuxbrainhttp://pastebin.com/0uxiemjX I'm still have the hammer01:33
wpwrakatrf-gpio -p 00000000.00000000.0000000 delay 00000000.00000000.000o000001:35
tuxbraindefinitively MOSI don't go to go to sleep tonight http://pastebin.com/kfC1h8p401:36
DocScrutinizeryour atrf-gpio is a lil turing-engine, huh?01:36
wpwrakatrf-gpio -p 00000000.00000000.0000000 delay 00000000.00000001.00000000 delay 00000000.00000001.000o000001:37
wpwrakDocScrutinizer: it shares the clarity and intuitiveness of the design ;-)01:38
tuxbrainmessages from MOSI:Don't let the party stops!!! http://pastebin.com/6P2AnV6X01:38
wpwraktime for the voltmeter then ...01:39
tuxbraindamn it...01:39
wpwrakMOSI is the line that goes straight from MCU to RF01:39
wpwrak(there's only one)01:39
wpwrakleft and right from it are vias01:40
wpwraktuxbrain: do you see it ?01:40
tuxbraina pic and arrow please?01:40
wpwraki knew i should have guified that mess ...01:41
DocScrutinizerguify vias, traces?01:42
wpwrakDocScrutinizer: that gpio checker utility01:44
wpwrakDocScrutinizer: showing the layout and the (supposed) configuration01:44
DocScrutinizerncurses cuts it01:45
DocScrutinizernot though if even echo sucks01:46
tuxbrainmmmm it marks 0 on tester01:46
DocScrutinizeron ansi esc01:46
tuxbraingnd is the exterior of the usb connector isn't it? so black in usbconnector , red in that line it should be some voltage... isn't it?01:47
DocScrutinizerthe nice thing with digital electronics: you can test easily, as everthing won't mind you shorting sth to gnd01:47
wpwraktuxbrain: yes, sounds good01:48
wpwraktuxbrain: you can test by measuring the voltage between GND and P12 (the 2nd test point from the USB connector)01:49
DocScrutinizernow you hooked me, where's the schematics01:49
DocScrutinizerpdf please01:49
tuxbrainyep last command atrf-gpio -p 00000000.00000000.0000000 delay 00000000.00000001.00000000 delay 00000000.00000001.000o0000 no complains now01:49
wpwrakthat should be ~3.3 (+/- 0.3) V01:50
wpwrakDocScrutinizer: always a reliable source for schematics: http://projects.qi-hardware.com/schhist/01:50
tuxbrain3,25 yes01:50
DocScrutinizerk, too much effort to find the right one01:51
wpwraktuxbrain: maybe this board has a bonding defect. i got a few of these. may be my soldering process01:51
tuxbrainwpwrak: after run the command the MOSI turs on again01:51
wpwraktuxbrain: do you have a USB A to USB A cable ? lemme take a picture of my setup ...01:53
tuxbrainI try to un plug/plug but it that mosi stills on01:53
wolfspraultuxbrain: TUXBRAIN!01:56
wpwraktuxbrain: just a sec ...01:56
tuxbrainhehehe hey wolfy!01:56
kristianpauloh wow !01:56
kristianpaulhi tuxbrain :-)01:56
tuxbraincommon guy you gonna make me red :P01:57
kristianpaulgetting fun with atusb, huh? :-)01:58
kristianpaulhopefully not with hamers and other related hardware..01:59
wpwraktuxbrain: can you fix the board in a way similar to this ?01:59
Action: kristianpaul having some fun trying to sync two clocks domains :/02:00
wpwraktuxbrain: the goal would be to be able to push a bit of paper (the ball on the left) under the antenna, to gently bend the board upward a little02:00
kristianpaulwpwrak: is hard to focus that pic, if is necesary do to so for the scope probe..02:01
kristianpaulah paper..02:01
wpwraktuxbrain: that i've seen in a few of the boards i made is that some contacts fail. if i bend the board, those contacts often come back.02:01
wpwrakkristianpaul: the multimeter probe just happens to be there. it's connected to a very conveniently placed pieced of ground. (much harder to attach it to the shield)02:02
tuxbrainwpwrak: I can do that but not now, my claps are in the deep in the "tools closed" and I will make a lot of noise to pull them out, so I will arrange this setup tomorrow morning and try the bending fix then.02:05
wpwraktuxbrain: okay. the same thing may also help with the other board.02:06
wpwraktuxbrain: most likely, the other board fails to change the clock speed. so the mcu runs at 1 MHz instead of 8 MHz, which is wrong for USB02:06
tuxbrainoh wait! aha clamp find in a shelve02:06
wpwrakif you can "accidentally" drop a box of, say, metal tools from the shelf, you wouldn't have to worry about making noise after that :)02:08
tuxbrainok tellme some comands02:09
tuxbrainhey commands that complains before now not complain02:10
tuxbrainnow it fails in cristal  frequency http://pastebin.com/aw0cMHJ602:14
tuxbrainatrf-path -g usb net:ben  still stucks02:16
wpwraklet's see ...02:17
wpwrakdoes the PC run ntpd ?02:17
tuxbrainops let me check02:18
wpwrak(commands not complaining) that's good :) atrf-path still being stuck is bad, though02:19
wpwrakwe're now running out of excuses for atrf-path to be stuck :)02:20
tuxbrainthe gpio errors on make usb returns :(02:20
wpwraktuxbrain: have you put some paper under the board ?02:22
wpwrak(to bend it)02:22
tuxbrainyes and it work it once to arrive to the cristal check but now it doesn't work :(02:22
wpwraktry removing the paper then :)02:23
wpwrakbending the board such that the defective bonds connect isn't an exact science :)02:23
wpwrakthe good news is that the boards made at the fab are extremely unlikely to have that kind of problem (well, unless i misunderstood completely what's going on)02:24
tuxbrainno way, due it works once I'm sure it a fault connect as you said, but it seems to not return to connect and I'm afraid to split the card in two if I still trying to bend it more, time to some sleep, and crossing our finguers that fab atusb/atbens are in good shape.... to at least see the black box with the withe line dancing on it...02:29
tuxbraingood night02:29
wpwraklet's hope for better luck tomorrow :)02:29
wpwraktuxbrain: sweet dreams ! :)02:30
whitequarkwhat's exactly that Theora support for Ben? a SIMD-optimized decoder?06:39
wolfspraulyes, or scaler rather I think06:42
whitequarkhmm, does 4720 have a scaling IPU?06:45
whitequarklike 4750l06:45
DocScrutinizerdoes Ben have a display to make any use of that?07:08
whitequarkwell, there are some cases when you want to squeeze FullHD to 320x24007:11
whitequarkand the IPU is very handy then!07:11
Action: whitequark has printed a quote by Doc and put it on the wall07:14
tuxbrain_awaywolfspraul, wpwrak rejon and other on line of course... I was thinking on change the comercial name of atben/atusb to odino/valhala , this strange inspiration comes as always does to me listening a song :P , this one http://www.jamendo.com/es/track/65580008:59
tuxbrain_awaywhat do you think... i have de promo video in mind :P09:00
tuxbrainof course will not the full song, but the intro , chorus and the "lambada" solo are awesome for a video promo. 09:02
tuxbrainalso the "the wall"+morricone hohoho09:04
wolfspraulI don't care, but if one thing is badly broken in copyleft hardware land, it's brands :-)09:06
wolfspraulso I am starting to focus a bit, for example I removed the "Manufacturing: Sharism" line on the upcoming m1 run09:06
wolfspraulthat's all just Milkymist now09:06
wolfspraultuxbrain is also a cool brand I think09:06
tuxbrainya but Tuxbrain's atusb sound less cooler than Tuxbrain's Valhalla the Might USB 6lowpan dongle09:09
viricwhere I work, 'Valhalla' sounds more like an internal name for a project, never labelled as such outside the company :)09:11
tuxbrainTuxbrain's Odino Empower your NanoNote and cut with wires :P09:11
tuxbrainok ok I need to sleep a little bit more... I was just letting the imagination fly from a lot of booring work09:14
tuxbrainsee you later 09:14
qi-bot[commit] David Kühling: new package libdumb: mod/xm player (dumbplay) plus player library for games http://qi-hw.com/p/openwrt-packages/51707e809:38
DocScrutinizerOdin, valhalla, and the whole stuff has some strange connotations here in Germany09:58
DocScrutinizerprobably in Israel as well09:58
DocScrutinizerjust saying...09:59
DocScrutinizeryou probably don't want those http://koptisch.files.wordpress.com/2009/12/skinheads.jpg as your primary customer target group10:03
DocScrutinizerhttp://www.odin-versand.de  <- probably NOT (work)safe10:06
DocScrutinizeractually take care, this size might be rogue10:07
DocScrutinizersorry for being direct here, but I think it's better you stop that idea before it hurts10:10
wpwraktuxbrain_away: i don't particularly care what you call them, but there's no way in hell that i'm doing a global replace of all the names in code, file names, documentation, etc. ;-)10:28
wpwraktuxbrain_away: so if you feel strongly about those "marketing names", maybe use them as "nicknames". e.g., atben "odino" and atusb "valhala". of course, all this really accomplishes is that customers have to remember both names.10:30
tuxbrain_awayok dude enough and strong arguments againts to dismiss the idea :) atben will be atben and atusb will be atusb  10:35
wpwrakDocScrutinizer: (skinheads) ah well, you have to dismiss pretty much all the northern mythology if you want to avoid things they like :)10:42
wpwraktuxbrain_away: funny song :)10:45
wpwraknice .. this morning looks as if it was foggy. but i think it's not fog - it's volcano ashes. with love from chile.10:47
wpwraktuxbrain_away: we could try to work around the hang and see if the spectrum comes out okay regardless. to do this, you could try:10:56
wpwraktuxbrain_away: 1) in ben-wpan/tools/lib/cwtest.c:start_test_mode_231, change wait_for_interrupt(dsc, IRQ_PLL_LOCK, IRQ_PLL_LOCK, 10, 0); to .... , 10, 10);10:57
wpwraktuxbrain_away: 2) cd ben-wpan/tools && make clean && make install10:58
wpwraktuxbrain_away: 3) atrf-path -g usb net:ben10:59
wpwraktuxbrain_away: now, if you get a reasonably straight line, all is well. if it looks more like this: http://downloads.qi-hardware.com/people/werner/wpan/tmp/jagged.png11:01
wpwraktuxbrain_away: then this would indicate that the transceiver isn't very happy with the change11:02
wpwrakactually, lemme change the atrf-path invocations to always use both bands. that way, we can spot such problems also in the production test ...11:03
wpwrak(background: the test mode can use channel_frequency +/- 0.5 MHz. the spectrum scan in the test scripts only uses one offset, to obtain a faster display update rate. but there are are some communication issues that you don't see that way. you should be able to catch them with other tests, e.g., the gpio test, but may be better to have a bit of redundancy)11:06
qwebirc7110my server has failed miserably -\ this is kyak speaking11:07
qwebirc7110wow, this webchat is pretty good11:09
wpwrakqwebirc7110: (failure) software or hardware ?11:09
qwebirc7110hardware, i assume11:09
qwebirc7110it won't boot11:09
qwebirc7110my assumption is BIOS11:09
wpwrakah, lovely11:09
qwebirc7110i will take it to work tomorrow for a little examination11:10
wpwraki have a server that sometimes shuts down. did so about 3 times per day a couple of weeks ago, just after two visitors arrived. i diagnosed it as a double fan failure. so i bought replacements but didn't install them yet. the interesting things is that, once i had the replacements, the failures stopped :)11:11
qwebirc11865heh, i have this habit to close tabs :)11:13
qwebirc11865kyak_: you should ping timeout now11:15
qi-bot[commit] Werner Almesberger: prod/: use positive and negative offset in spectrum test http://qi-hw.com/p/ben-wpan/66fb9d811:16
qi-bot[commit] David Kühling: new package: alex4: Alex the Allegator v4: Game-Boy style jump&n&run game http://qi-hw.com/p/openwrt-packages/51cbdda11:32
qi-bot[commit] David Kühling: alex4: use left shift as fire key (left control missing on Nanonote) http://qi-hw.com/p/openwrt-packages/af0b29d11:49
qi-bot[commit] David Kühling: remove allegro-setup launcher: don't want people to accidentally break sound playback http://qi-hw.com/p/gmenu2x/c61951411:50
qi-bot[commit] David Kühling: add launcher for alex4 http://qi-hw.com/p/gmenu2x/97adb0611:50
tuxbrain_awaywpwrak: I need a "comercial" sort description of atben/atusb to include in the shop, I can add some literature but I need a "specification" items list to put on there,12:01
tuxbrain_awaywpwrak: regarding test, I will retake it this afternoon/night in about 6hours from now12:01
wpwraktuxbrain_away: (test) perfect. i just emerged from my beauty sleep ;-)15:02
wpwraktuxbrain_away: (commercial description) working on it ...15:10
wpwraktuxbrain_away: regarding marketing, if you want a catchy name, then attacking the "wpan" may be a more profitable goal. besides the project name itself, there aren't too many direct references to that name, and "WPAN" per se isn't used only with ieee 802.15.4. so you could give the whole product line a code name. no code change required ;-)15:39
wpwraktuxbrain_away: i'll post the "commercial description" to the list, for comments16:00
tuxbrain_awaymmmm WPAN AtBen and WPAN AtUSB sounds catchy and meaninfull :)17:07
wpwrakhmm. i've grown a bit of a dislike for CamelCase ... it's something you always forget. worse, it keeps on changing. e.g., OpenMoko -> Openmoko. and I think MilkyMist also did a Milkymist metamorphosis17:11
tuxbrain_awayWPAN atben WPAN atusb?17:12
tuxbrain_awayWPAN ATBEN WPAN ATUSB17:12
wpwrakquite a bit of shouting :-)17:12
wpwrakjust "atben" and "atusb" ?17:12
wpwrakor come up with some nice name for the product line17:13
wpwrak"Tuxbrain Foliot" line. whatever ;-)17:13
tuxbrain_awaymmmm atben atusb  has no meaning at first sight , but adding WPAN  the seems to mean Wireles personal network at ben/usb wich is reall meaninful therefor easy to remember, my thought at least17:14
tuxbrain_awaywireles personal area network to be more exact :P17:15
wpwrakmaybe "atben (WPAN)" ?17:15
wpwrakthe names get meaning by their existence ;-)17:16
wpwrakwell, and the deeds they're associated with17:16
wpwrak(why) what did the "why" refer to ? the "atben (WPAN)" or the foliot ? the latter would be some minor magical/mythical creature.17:18
tuxbrain_awayok open voting on channel,17:28
tuxbrain_awaya) WPAN atben17:28
tuxbrain_awayb)atben (WPAN)17:28
Action: tuxbrain_away votes for a17:28
Action: tuxbrain votes for a17:28
tuxbrainok two votes for a option :P17:29
Action: wpwrak votes for b :)17:29
wpwrak(i see "wpan" more as a way of saying "this is not wifi" than of narrowing down what it actually is. unfortunately, the only accurate identifiers are IEEE 802.15.4, which is unwieldy, and 6LoWPAN, which we don't have yet)17:31
tuxbrainbtw how long will take t ohave a 6LoWPAN thing17:33
tuxbrainI was real disconnected this days as you my also notice but for tuxbrain (and I hope for qi-hardware in general) good will17:34
Action: jluis votes for b ( sounds better )17:35
wpwrakhard to tell. all that stuff ough to take a few months of work. alas, i'm running out of time. that's why i made the "dirtpan" as a "quick hack"17:35
wpwraktuxbrain: btw, what's the status with the case/sugru ?17:57
tuxbrainI have not ended a succesfull test without it... so until it happens will be any advance on case to compare if sugru eats dB17:58
wpwrakokay, one item for the to do list then :)18:02
wpwrakdid the smt fab comment on any difficulties ?18:03
zearwpwrak, btw "Note: this is not an SDIO(tm) device and will only work in18:03
zear     products specifically supporting it"18:03
zeardoes that mean that a) you need a special hardware in your 8:10 slot, or18:03
zearb) it simply isn't a plug-and-play sdio, yet it will work in every sdio slot providing there's software for it on the system?18:04
tuxbrainwpwrak: yes two:18:05
wpwrakzear: it means that it doesn't speak the SDIO protocol. you need to be able to bit-bang the individual signals in a way that MMC/SD/SDIO don't do.18:06
tuxbrain-Fiducials are not in the cordinates file list18:06
tuxbrain-They finally have to remove one of the sides of the atusb pcb panel to place correctly the USB connector (as we have adviced)18:06
wpwraktuxbrain: (fiducials) aw. okay, that's fixable.18:06
zearwpwrak, pardon my lack of the proper technical knowledge, but does it mean it will work in any 8:10 slot, or not?18:06
zeari'm mostly concerned about the Dingoo A32018:07
wpwraktuxbrain: (usb) heh, we told the pcb guys about a dozen times ;-) i think the removal should have been easy, though, correct ?18:07
tuxbrainzear: meanwhile you where able to "disconnect" the SDIO driver owning it I suppost that yes18:07
wpwrakzear: no it won't work in any slot with the 8:10 card or uSD form factor18:07
zearzrafa has a dingoo he got from me, so perhaps you two could test ben-wpan on it once meet the next time ;)18:08
wpwrakzear: you need to be able to a) run a specific driver, and b) that driver needs to be able to bit-bang the signals.18:08
zear*once you18:08
wpwrakzear: let me give you a counter-example:18:08
tuxbrainwpwrak: yes, it was easy, they where more annoyed by the fiducial thing, nothing irritating but the have remarked me this a lot18:08
zearwpwrak, so it's all software limitations18:09
zearnot the slot hardware18:09
wpwrakzear: if you have an MMC/SD/SDIO-to-USB controller chip (often found in laptops/netbooks), then that chip would implement operations like "send command with response type R1" or "send command with response type R2 and a data block"18:09
wpwrakzear: the conrtoller would then generate a number of transitions of the signal lines for those commands. atben wouldn't understand any of this.18:10
zearwpwrak, i'm thinking of the miniSD slot in the Dingoo A320, which supports SDIO, so i assume there's no sd-to-usb trickery like in laptops18:10
wpwrakzear: i don't know the dingoo. can you point me to the schematics ? then i could probably tell whether it's possible18:11
zearwpwrak, let me see if i can find anything18:11
zearalso, feel free to bother rafa about showing you the real thing, i already told him he has to show it to you one day ;D18:12
mthwpwrak: I don't think there are public schematics for the Dingoo, but it's a JZ4740 based system18:12
whitequarktuxbrain: any thoughts of when atben/atusb will be available for order?18:12
wpwrakzear: zrafa new lives a long distance (some 1.2 Mm) away from me. currently snowed in below volcanic ashes :)18:12
mthwe're running a kernel on it that is derived from qi-kernel18:12
zearwpwrak, yeah, i know18:12
zearthough you two will eventually meet ;)18:13
wpwrakmth: okay, then the chances of atben working out of the box are probably pretty good. well, unless they did something really stupid ;-)18:13
wpwrakzear: sure. he comes to town every now and then.18:13
tuxbrainwhitequark, even there is no official annoucement yet and the shop descriptions are totally ugly (I will fix it ASAP) you can order them now, I will spend all weekend flashing/testing/praying to have some ready to send on tuesday (monday is holidays here)18:14
zearwpwrak, the thing has a broken screen and afaik it pretty much dusts somewhere on rafa's desk, so would be nice if he actually used it for something ;)18:14
wpwraktuxbrain: just don't start charging credit cards before at least one has been verified to be okay ;-)18:15
zeartuxbrain, any estimated price range of ben-wpan you guys expect it to be?18:15
whitequarktuxbrain: sure. what payment options do you have?18:16
wpwrakzear: ;-) you guys should buy bens instead, not some weird dingoos ;-)18:16
tuxbrainzear: without vat and no shiping 25¬ atben /35¬ atusb /50¬ the combo18:17
zearwpwrak, i do have a ben, from the first batch ;)18:17
zearit's dusting away :P18:17
zeartuxbrain, ouch, a bit pricey18:18
tuxbrainzear: order 1k and will be a lot cheaper :P18:18
zearbut that's an appropriate price for a small batch18:18
wpwraktuxbrain: hmm. so a quick calculation suggests that you're actually losing money on them ? :)18:19
tuxbrainwhitequark: credit card, bank transfer and paypal18:19
tuxbrainwpwrak: not not loosing money but not earning a lot with them :P18:19
zearwpwrak, oh, and the weird dingoos have actually gaming controls, you guys should not buy weird gaming controls-less hardware designs from china ;)18:20
wpwraktuxbrain: (not losing) heh, good.  at least something :)18:20
wpwrakzear: bah, the ben has a full keyboard. that more gaming control could you possibly want ;)18:21
tuxbrainwpwrak: covering fab costs and bit of the hours I will spend flashing leting the sking on my finges with your "confortable" flashing cable :P18:21
zearwpwrak, maybe if you want to play nethack or text adventures18:21
zeartuxbrain, the problem is you have to order two wpans to have any use of them, so the price raises twofold ;P18:22
whitequarktuxbrain: any estimate of how much the vat and shipping will add?18:23
tuxbrainzear , don't look at the money look at the peeled fingers after unpanelize, edge smothing and flashing  hundreds of this beaties just with the joy of hardware freedom lovers in mind :P18:25
tuxbrainwhitequark: where do you live?18:26
whitequarktuxbrain: Russia18:26
zeartuxbrain, my whole interest in wpan limits to the concept of sticking one into a dingoo and a second into a nanonote, and play some duke3d deathmatch ;)18:27
wpwrakzear: to two atben then. well, EUR 50 for the pair. that's not too bad. you'd pay more for a single WLAN card ;)18:29
tuxbrainwhitequark: ufffffff.... I have bad experiece with rusia lately, I have to review all the shiping there looking for better option than I have....18:29
zearan sdio one - yes18:29
zearwpwrak, though 50 EUR is pretty much a price of a new dingoo, so.. ;P18:30
zearon the other hand, having something like atben would be a great opportunity to learn about networking protocols and stuff like that18:31
wpwrakzear: if you can finance the production of as many atben cards as dingoos were made, the prices would come down quite a bit, too ;-) it's unfortunately all about volume in this business ...18:31
zearwpwrak, yeah, i'm not doubting that :P18:31
kristianpaulyeah :(18:31
wpwrakzear: yeah ! no black box :)18:31
wpwrake.g., at EUR 25 for an atben, tuxbrain should have a seriously fat margin. but, given the low volume, one-time costs and higher per-unit costs due to low volume eat all that18:32
zearthat's the problem with small companies. On one hand, your price is dependent on your batch order, on the other hand though, smaller bath == bigger price == smaller public interest in buying your stuff18:33
whitequarktuxbrain: dealextreme had no problem delivering its stuff18:34
whitequarkmaybe you can use the same methods18:34
kristianpaulpostal mail is nice18:34
kristianpauli used ro bring stuff from usa,18:35
kristianpaulsame methof used by dealextreme to my country too18:35
whitequarkkristianpaul: where are you from?18:35
kristianpaulwhitequark: colombia18:36
tuxbrainkristianpaul, whitequark my experiece from other people I have talked about is : USA->South america OK, China->Rusia Ok, EU->USA OK, EU->Southamerica problems, EU->Rusia problems18:37
tuxbrainFedex seems to have a good ratio of avoiding such problems, also DHL seems too, UPS is a totally dissaster, and TNT is like a rolling a dice18:38
kristianpaulwow 50eur are just 130.000 cop, thats not that bad :-)18:38
whitequarkand ca. 2000 rur18:38
whitequarklike 2/3th of a Ben...18:38
wpwrakzear: yeah, it's a vicious circle. kinda like the "poverty tax". if you're poor, you can't buy efficiently, ending up paying more than richer buyers.18:39
kristianpauloh i agree with that whitequark 18:39
wpwraktuxbrain: i had my share of troubles with DHL. fedex are the best. they would still have some room for improvement, but all the others are worse.18:41
tuxbrainso decide fedex for the win on those trouble customs barriers , 18:42
zearwpwrak, with enough cool applications of it and drivers for different systems/architectures, plus a list of devices confirmed to work with it, providing it gets a good advertising (for example videos showing how two systems can play games together), i think wpan has a great chance of winning lots of customers18:43
kristianpaulfidopan ;)18:44
wpwraktuxbrain: regular mail can often also be an option. it's usually very inefficient, but it's generally aimed as "common people", so they may be more lenient on customs restrictions.18:44
wpwrakkristianpaul: ;-))18:45
tuxbrainwpwrak: meanwhile your not live in spain this can apply, I tryied once to oferr it as an option, but you see now any vestige on the tuxbrain website about it? so well you can deduce how it works, until I have at least tree guys(or girls)  working exclusively at shippings that can handle the headaches for me, I will not offer again Spanish postal as shipping option, clear enough?18:48
kristianpaulyes sr18:50
GNUtoohi tuxbrain 18:50
GNUtooyou had oe questions?18:50
whitequarktuxbrain: looks like you're quite annoyed by spanish postal...18:50
GNUtoobut then you disappeared18:51
kristianpaulthat conclude tuxbrain offices should be moved out spain :-)18:51
tuxbrainkristianpaul: or conclude that tuxbrain has to burn out all postal offices and officers to restart from 0 with some efficiency in mind, not much just a bit18:52
kristianpaulthat count too ;)18:53
wpwrakkristianpaul: he should relocate to argentina for a while. i think he'd be quite happy with the spanish post after that ;-)18:56
wpwrakwell, here the bureaucracy actually isn't too bad. works better than i expected. but each trip to the post office means to stand something like 30-60 min in line. sit, if you're lucky and it's a saturday with bad weather.18:57
Action: kristianpaul loves living in his small city18:58
tuxbrainGNUtoo: my apologizes fot my offline period, was needed, there are a lot of OE questions parked that will arise again once I retake that part, but righ now I have another priorities like atben/atusb and a lot of burocratic/marketing papers and comercial chitchat until the end of june, I have partialy solved my OE needs avoiding it and do "on device" compilation (on BIP-6000 latest version QT,PyQt takes about 7 days but at least work is done) to make 19:00
tuxbrainable to work a OpenERP client (Koo) with this I have save my day , and I can develop this part of the businnes quite a lot, but of course this is not the optimal way to work , I "need" an stable way to generate full images form a "stable" and uptodate OE enviroment, but this can be postponed buy now for a couple of weeks19:00
GNUtootuxbrain, np19:01
GNUtootuxbrain, abuot pyqt, it's in oe now19:01
GNUtooand works fine19:01
GNUtoomorphis and mickeyl are making a new program called aurora19:02
GNUtoothat is in QT+QML19:02
tuxbrainGNUtoo: nice to know :)19:04
GNUtooand they use python for it19:05
wpwrakoh, nice. component references can only be 9 characters. or mayhem results.19:30
wpwraker, make that 819:33
qi-bot[commit] Werner Almesberger: fpd2pdf/fpdoc: ignore items in .pos not defined in the schematics http://qi-hw.com/p/eda-tools/0de3b9619:45
qi-bot[commit] Werner Almesberger: atben: include the position of fiducials in atben-pos.csv http://qi-hw.com/p/ben-wpan/9f9e77d19:46
qi-bot[commit] Werner Almesberger: atusb: include the position of fiducials in atusb-pos.csv (for fab) http://qi-hw.com/p/ben-wpan/b41076119:52
kristianpaul"And how many codes must a vendor lock down"20:00
kristianpaulbefore silicon turns to sand?20:00
kristianpaulfrom http://openbsd.org/lyrics.html#4920:00
dvdkwhitequark: tuxbrain: saw you talking about atben/atusb and shipping20:22
dvdkmaybe we could handle this like the UBB? i order a few and forward them using german mail (insured worldwide shipment should be about eur 6.50)20:23
dvdkthe nice thing is: in case they are lost, then we get money back and tuxbrain can sell even more :)20:23
whitequarkdvdk: that will be very good20:24
whitequarkbut they are quite costly. you're going to buy a not so small quantity, right?20:24
dvdkwhitequark: maybe this time i will only buy as much as i can re-send.   maybe not more than 3-5?20:25
dvdkif i do another thing like ubb (ordering/reselling 50), pulster.de is going to be angry again :)20:25
dvdkwhitequark: you might have to pay customs tax, though.20:27
dvdkok, it's 3,45+2,05+1,5020:28
dvdk7 euro, insured for up to 100 euro20:29
dvdkpretty expensive20:29
dvdkplus about 25 cent for a proper shiping envelope20:29
dvdkwhitequark: btw you also alsked about mplayer acceleration.20:30
dvdkwhitequark: it's scaling + color conversion in hardware.  actual decoding up to Y+U+V image planes is done in software20:30
dvdkthe best way to get video on the ben is to transcode it via ffmpeg2theora or use ffmpeg to generate webm/vp8 video at the correct resolution (down-scaled).20:31
whitequarkdvdk: hm, same for 4750l20:33
whitequarktoo bad they're putting different ipus and codecs in each next chip20:33
whitequarkdvdk: (pulster.de) how that's related to you?20:34
dvdkwhitequark: (I was somewhat kidding), pulster.de commecrially sells UBBs, and he wasn't to happy last time i announced on the mailinglist that i could ship ubbs worldwide for 0.80 cent shipment cost 20:38
dvdkwhitequark: (wrt video codecs): having the codec in software is better anyways, hackability-wise20:39
whitequarkdvdk: sure, the only thing you should do in hardware is SIMD20:40
whitequarkthe good part is, SIMDs are somewhat compatible, at least upwards (ben 4720 -> my 4750l)20:41
whitequarkthe bad part, through, is absence of documentation and tools20:41
whitequarkdvdk: (pulster) I just noticed they sell SD-WiFi cards, and apparently they work with Linux20:45
whitequarkdvdk: do you know (just in the case) if there are cheaper variants? 89 euro is a lot20:46
kristianpaulwpan variant you mean? ;)20:47
kristianpauljust kidding..20:47
qi-botThe build has FAILED, see log here: http://fidelio.qi-hardware.com/~xiangfu/compile-log/openwrt-xburst.full_system-06072011-0240/20:48
whitequarkkristianpaul: the whole point of using wifi in the project I want to make is to reuse an existing access point20:48
wpwrakdvdk: pulster's been giving you a hard time about ubb ? funny, he himself got rather late into the game ;-)20:48
whitequarkkristianpaul: and, in that particular case, I don't have any USB ports for atusb on it, nor an open firmware20:49
kristianpaulsigh, will be nice haVh i could too20:50
qi-botThe build was successfull, see images here: http://fidelio.qi-hardware.com/~xiangfu/compile-log/openwrt-xburst.full_system-06072011-0240/20:51
kristianpaulwhitequark: sure,re-use is good, i miss that too20:52
kristianpaulbut thinking about offline is good from time to time20:52
kristianpaulhum, no usb ports, oh well, hard situation20:53
whitequarkkristianpaul: there's no sense in usb ports when you only have stock firmware20:54
whitequarkeven with root20:54
whitequarktalking about this, tp-link does some rather good (and cheap, for its features) routers20:54
whitequarkwith EHCI and perfectly functional OpenWRT installable20:55
kristianpauli was about to point a fonera, i think are kinda cheap too, for those wifi to usb-host bridges.20:58
dvdkwpwrak: maybe pulster had to wait some time with UBB shipments for market to become desaturated :)21:00
whitequarkkristianpaul: my one (gigabit router with draft-n wifi and 400mhz mips cpu) was about $100. don't know how fonera compares to that21:00
dvdkkristianpaul: fonera?  i have been reflashing fonera routers with openwrt backfire.  but no usb on them.21:00
dvdkmy fonera was 13 eur on ebay :)  bought 3 of them.21:01
whitequarkdvdk: are there any uses for UBB except plugging it to Ben for some weird task?21:01
dvdkwhitequark: plugging it into other devices for weird tasks? :)21:02
whitequarkdvdk: (fonera) that's cheapier than the cheapiest of most crappy d-links here :)21:02
dvdkwhitequark: no uses I currently know of.  but i also gave a few ubb to people who don't own nanonotes, so maybe there is something to them.21:03
dvdkwhitequark: yeah and with backfire, they _rule_21:03
whitequarkdvdk: I don't know any other devices with a SD slot accessible as GPIOs, would be interesting to hear about these21:03
dvdkwhitequark: i'd guess almost all embedded systems can reporgram their sd card pins.21:03
dvdkyou might look at the efika mx devices sold by tuxbrain.  they use a freescale imx chip, that directly feeds the (micro) SD slots.  I'd guess the imx can reuse the SD pins as GPIOs21:05
whitequarkdvdk: ah yes, that may be applicable to some developer boards as well. apart from that, and maybe dingoo's and clones, I don't know any devices with SD card slot and a firmware which allows you to use pins as gpios21:05
dvdkwhitequark: as long as you have linux an can put software on the devices, reprogramming should be simple.  not even a kernel module needed (user-space sw only)21:06
whitequarkdvdk: you'll need docs for chipset too21:07
whitequarkuserspace bitbanging without mmap is going to be slow as hell21:08
dvdkwhitequark: you don't have a chipset, just a single chip (SOC) in most embedded devices nowadays.  and most of these are pretty well documented.21:09
whitequarkdvdk: sure, a soc21:13
wpwrakdvdk: and there must be enough space for UBB+cable (or whatever you connect). e.g., this excludes the openmoko phones :-(21:13
kristianpauldvdk: http://wiki.fon.com/wiki/Fonera_2.0g21:14
wpwrakwhitequark: you could it use also to make an uSD->SD adapter21:16
whitequarkwpwrak: would a flex something (dunno how is it called. that thing used to connect LCD displays to boards) work with openmoko? it's not very hard to solder those to plain board I think21:17
whitequarkI mean, you can solder it to UBB and then route under the battery maybe21:17
whitequarknot a very clean solution, through21:18
wpwrakwhitequark: yes, that may work. but with difficulty. the uSD adapter there is quite inaccessible.21:18
wpwrakwhitequark: the main problem would be that the FPC would come out near the PCB. then you'd have to bend it upwards in very little space. this puts mechanical strain on the FPC and it will in turn have a spring force working against the uSD card holder, possibly forcing it open.21:19
dvdkkristianpaul: i wonder whether fonera 2.0g can still be reflashed via ethernet using the bootloader bug.21:19
wpwraktuxbrain: if one were to generate a set of fab files now, they would include the fiducials in *-pos.csv :-)21:42
wpwraktuxbrain: did you see my suggestions for trying to make the spectrum scan work, even if the board isn't entirely healthy ?21:55
--- Wed Jun 8 201100:00

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