#qi-hardware IRC log for Wednesday, 2011-06-29

kyakwpwrak: for Ben's kernel, yeah, i copied the ben-wpan-config-2.6.3805:04
kyakwpwrak: 512 Mb rootfs - i'm not sure if it is "new", it has been 512 Mb for a while now.. at least for the latest several releases05:05
kyakwpwrak: do your explanations about conflicting ieee* kernel modules and testing utils apply to atusb as well?05:06
kyakif so, then i perhaps needed to reboot, because i updated the firmware of atusb and then started atrf-xtal 1000 (all successfully) before i tried to do "iz add .. wpan0" and saw "no such device"05:08
kyakanyway, will be able to check only in the evening..05:08
wolfspraulkyak: you got your boards and they work?05:27
kyakwolfspraul: i got my boards yesterday evening, haven't yet really tried it, just prepared necessary tools05:28
kyakthe software tools it is05:28
wpwrakkyak: (kernel vs. user space) yes, this will apply too10:23
wpwrakkyak: but you're ahead of me there :) having any kernel support for atusb is still brand new. and afaik, it doesn't work yet. stefan is working at it. i'm at a conference this week, so i won't be able to do much (besides daily heavy drinking, of course :)10:25
kyakwpwrak: yeah, i noticed that you were mainly experimenting atben<->atben10:38
kyakalso, if atusb is not yet working, i probably can't do anything but just wait :)10:39
wpwrakkyak: have you tried atrf-rssi with atben yet ? and atrf-txrx -P with a pair of boards (atben/atben, atben/atusb, or atusb10:41
wpwrak /atusb)10:42
kyakwpwrak: no, i haven't. Yesterday i ended up with "izcoordinator... " on Ben (i.e. interface wpan0 was brought up)10:44
kyakbut could'nt do the same on PC. Perhaps due to the kernel vs. user space issue10:45
kyakwpwrak: for my understanding, do lowpan tools count as "userland" or "kernel"?10:46
kyakthe only two things i actually tried on PC were updating of atusb firmware with dfu and runnign atrf-xtal 100010:47
qi-bot[commit] Werner Almesberger: fisl2011/: don't declare a sole objective for M1; fixed dependencies (master) http://qi-hw.com/p/wernermisc/0a58e0510:57
qi-bot[commit] Werner Almesberger: fisl2011/: _really_ fix dependencies (master) http://qi-hw.com/p/wernermisc/450a20a10:57
qi-botThe build has FAILED, see log here: http://fidelio.qi-hardware.com/~xiangfu/compile-log/openwrt-xburst.full_system-06282011-1433/13:21
qi-bot[commit] root: fisl2011/: corrected flow.fig; straightened narrative (master) http://qi-hw.com/p/wernermisc/efacc7613:33
qi-bot[commit] root: Merge branch 'master' of git@projects.qi-hardware.com:wernermisc (master) http://qi-hw.com/p/wernermisc/ca5905a13:33
wpwrakrejon: brrr. the temperature is nasty.15:40
rejonwpwrak, yeah, i haven't left hotel yet15:41
wpwrakrejon: heh, beat you :) i got to the conference already around noon15:42
kristianpaulbtw is rms in FISL this year?15:55
wpwraki don't think so15:59
kristianpaulyou slide #5 looks interesting for him16:00
kristianpaulnice presentation16:01
kristianpaulis short, so i guess you speak a lot ;)16:01
kristianpaulhopefully willbe streaming or recording for the talks?16:02
wpwrakshort ?? it's something like 20-30% too long :)16:03
wpwrakso i'll have to hurry a bit16:03
kristianpaulwell for me :)16:03
kristianpaulyou have 1 hr or so?16:04
kristianpauli forgot to ask that first:)16:04
wpwrak1 h minus time for questions and people changing rooms. so about 45 min16:05
kristianpaulbest whishes then :)16:06
wpwraki have the open-ended slides towards the end. i could talk for hours about sourcing. i have a rich set of horror stories from openmoko :)16:09
kristianpaulopenmoko's un-official history teller 16:14
wpwrakone of them :)16:15
kristianpaulsuposelly osgpsby defalt dont read samples compressed as a pack of 4 samples inside a byte16:18
kristianpaul(very sample is 2 bits)16:18
kristianpaulfrom my sampled data as it is, i ran osgps and i got some tresholds to be triggered16:19
kristianpaulafaik not C/A16:19
kristianpaulbut i had 30s from sampled date, but i dint finshed debug of acquistion at all so i need to check16:19
kristianpaulmay be simulate sige in hw..16:19
kristianpaulbut if iuse the compressed parameter i dont get nothing..16:20
kristianpaulalso diging the code i disover if i did active the pack4 parameter for reading compressed data16:20
kristianpaulit take, yes athe 4 samples16:21
kristianpaulto a LUT16:21
kristianpaulbut the lut is this    const char val[4]={1,3,-1,-3};16:21
kristianpauli have some utilities to uncompress this data aswell, that i was using with octave16:22
kristianpaulbut that LUT never used 316:22
kristianpaulbut used 016:22
kristianpaulman, i'm confused know with this ADC16:22
kristianpaulor i'm ignoring something about dowconvertion inside sige chip16:23
kristianpaulbut those values above make sense16:23
kristianpaul2 bits ADC, etc..16:23
kristianpaulbut why is not zero?16:24
kristianpaulis not needed at all?16:24
kristianpauldatasheet make me guess on that, or require to improve my radio eng skills ;)16:24
qi-bot[commit] root: fisl2011/: changed date; cleaned up flow,fig; improved readability of achieve.fig (master) http://qi-hw.com/p/wernermisc/22ff99b16:49
wpwrakrejon: would you happen to have usbboot installed on your laptop ? my ben has a broken kernel and usbboot had a hellish dependency chain16:54
wpwrakblergh. now it's raining, too. well, maybe it'll get a bit warmer ...16:55
rejonwpwrak, hahhaa17:12
rejonmy ben is broken too!17:12
rejoni just installd17:12
rejonwe can fix our bens17:12
rejonok, i'm heading over to the conference now17:12
rejondone with my conference calls for the day17:12
wpwrakheh. i thought you'd skiip today entirely :)17:16
wpwraki'm at a table/stand roughly in the middle of the exposition area17:17
wpwrakhas KDE-Brasil written on it on one side17:17
wpwraki'm facing roughly the big "telecentro livre"17:18
wpwrakrejon: btw, dress warmly. the hall isn't heated17:19
wolfspraulrejon: Sebastien is MIA, I need confirmation from him for the font change17:21
rejonsounds good woakas 17:21
rejonwolfspraul, 17:21
rejonjust trying to keep on top of outstanding stuff17:21
rejonhopefully he will like it17:21
wolfspraulI like the designs and font style, stupid little 'must not be sold clause', but it doesn't really apply to us...17:21
wolfspraulI wonder what's in peoples minds sometimes :-)17:22
wolfspraulrejon: is your m1 functional?17:22
wolfspraulI demoed to some people today, they loved it. I think I sold 2 (once I have them)17:22
wolfspraulnow it's pretty nice booting into simple mode, and iterate through patches with button press17:23
wolfspraulconnect your camera before you boot, just in case17:23
wolfspraulyour board has a small hardware bug which we can fix ... ahh.. maybe Werner can fix it for you? :-)17:23
wolfspraulwpwrak: if you have a chance, you can remove and short L19 on rejon's m1 :-)17:24
rejonwpwrak, actually you are right, i will not come to the conf. today17:24
wpwrakwolfspraul: there are some folks here who may actually have a soldering iron :)17:24
rejoni will stay here and work, and come out for dinner and drinks17:24
wpwrakrejon: wimp ! :)17:25
rejonwpwrak, the second i enter around people, all my productivity will vanish17:25
rejonwolfspraul, can vouch for that17:25
wolfspraulrejon: ok, then remember this L19 thing. if werner can fix that for you, that'd be great.17:26
wolfspraulit's a nasty little bug especially in quick demo situations. and if werner has access to your m1 and an iron, it's fixed in 1 minute.17:27
wolfsprauljust needs the top acrylic to be off, that's all17:27
rejonwolfspraul, ok, remembering...my unit is functional and on the build from mid-may that has working boot-to-render17:27
rejonok i have my case and m1 and projector17:27
rejonwpwrak, lets fix in the morning at the venue17:27
rejoni don't present until the 2nd17:27
wpwrakwolfspraul, rejon: hmm, i'll see what i can do without my lab17:28
wpwrakrejon: if your M1 works, maybe it's better not to try rework17:28
wolfspraulwpwrak: it's the bug you fixed yourself. just remove and short that big capacitor.17:28
wolfspraulup to you guys, I'm just connecting...17:29
rejonok cool17:29
wpwrakwolfspraul: capacitor ? ;-)17:29
wolfspraulit's a big component, should be quite easy, no?17:29
wolfspraulup to you guys17:29
wolfspraulotherwise I hope Jon doesn't run into the bug when using the camera...17:30
wpwrakwolfspraul: removal may be tricky without proper tools17:30
kristianpaulhum, nice weather in Porto Alegre for this month17:30
kristianpaulwpwrak: no need remove just short it using a blob of esta~no17:30
wpwrakwell, if rejon wants me to give it a try, we can ask the hw guys if they have something. there's one group that recycles electronics into art :)17:31
wpwrakkristianpaul: yeah, that's what i'd try to do17:31
wpwrakkristianpaul: (nice weather) nice = freezing cold with rain ?17:31
wolfspraulrejon: workaround is this: if you want to use the camera in a demo, connect the camera (turned on, i.e. sending a signal) before booting your m117:31
wolfspraulthen you cannot run into the bug17:32
wolfspraulif you first boot your m1, then connect the camera later, it may sometimes not be recognized by m1 until the next reboot...17:32
kristianpaulwpwrak: wikipedia saids 14 °C daily mean for june, here is 30 °C with no wind in the evening, is the hell!17:33
kristianpauleven 32°C i think17:33
wpwrakkristianpaul: you have 30 C ? bastard !17:34
rejonwpwrak, keep me posted about night plans....laying low until....17:34
wpwrakrejon: i guess night plans will arrive by email :)17:34
kyakwpwrak: i've started atrf-txrx -P both on Ben and PC17:35
kyaki see "TX power -4.0 dBm"17:35
kyakis it as expected?17:36
wpwrakmake this atrf-txrx -P 1 on one of them17:36
wpwrakthen they'll start to talk17:36
kyaki see ++++ in PC and **** in Ben17:37
kyakare they talking now?17:37
kyaki gues they are :)17:38
kristianpaulwith evil plans :)17:38
wpwrakhmm. they're talking. but they don't understand each other17:38
wpwrakdoes dmesg say anything on the ben ?17:40
kyaki just had "116 bytes received" message on Ben17:41
kyakand now "112 bytes"17:41
wpwrakkyak: in particular, did you disable the various drivers ? (mmc and at86rf230)17:41
kyaki did disable the at86rf23017:41
kyaki though mmc was already disabled in ben-wpan kernel17:42
kyaklet me double check. A question - should i reboot Ben with atben in it?17:42
wpwrakshouldn't matter unless you have the mmc driver17:43
kyakok, i rebooted it. Is there anything particulare we are looking for in dmesg?17:44
wpwraknote that you have t run the commands by hand, due to the echo bug17:45
wpwrak(unless that's fixed already)17:45
kyaki'm running it by hand17:46
kyakthough you are right, echo looks pretty weird17:46
kyaklike it's redirected to stdout instead of /sys..17:46
kyakso i unbound it. now atrf-txrx?17:47
wpwrakyou should get a series of . on both sides17:47
kyaknope.. it's the same + and * 17:48
kyakdo i need to do anything special on atusb side?17:48
wpwrakah. yes. you need to upgrade the firmware :)17:48
wpwrakfirst, ger dfu-util17:49
kyaki did that yesterday17:49
kyakdfu-util -d 20b7:1540 -D atusb.bin17:49
kyakwhereas atusb.bin is the latest (by date) from your directory17:50
wpwrakthen try http://downloads.qi-hardware.com/people/werner/wpan/bindist/atusb-b0b04ed.bin17:50
wpwrakmaybe i need to update it. mompls17:51
kyakyah, i think it was b0b0 :)17:51
kyaki like the crc or whatever it is17:51
kyaki'll to update the firmware again17:51
wpwrakit's the shor commit number17:51
kyakheh.. i forgot to stop atrf-txrx during update17:53
wpwraklooks good17:53
kyakand it started to use 100% CPU17:54
wpwraklatest version: http://downloads.qi-hardware.com/people/werner/wpan/bindist/atusb-b3b037e.bin17:54
wpwrakwhatdid ?17:54
kyaknevermind, i just killed it.. and reflashed the atusb-b3b037e.bin17:55
kyaknow we have dots!17:55
kyakrunning fast17:55
wpwrakyeah ! :)17:56
wpwrakas it should17:56
kyakgues b0b0b was not so good after all :)17:56
kyakok, so we know it's working. I guess i can take it from here. Do you need me to run any tests?17:56
wpwrakit may be from befoee bthe latest interrupt changes17:57
wpwraki'll be interested in collecting a few serial numbers. but i'll di rhat when i'm back home17:57
wpwrakthe boards should leave tuxbrain i good working order17:58
kyakok.. i'll travel around and see when these dots turn to something else, thus detecting the wireless "bubble" :)18:00
wpwrakkyak: ah yes, that would be good to know18:01
kyakit works with interruptions (-*) when i'm in another room, not sure how it would affect the real TCP transmission18:05
wpwrakhehe, recruited my first listener :)18:10
wpwrakkyak: you could try with dirtpan18:11
kyakwpwrak: i'm failing at "iz add.." step18:13
kyakor is it not required for dirtpan?18:13
wpwrakshould be required18:13
kyaki get "Receive failed: Netlink Error (errno = No such device)" on atusb side18:13
kyakatben seems fine18:13
kyaki guess i'm lacking drivers for at86rf23018:14
kyakthe above mentioned error is from "iz add wpan-phy0"18:15
kristianpaulwhet i do a ftp trasnfer it uses tcp isnt?18:15
kristianpaultoo slow then ,i'll switch to udp, i guess than handshake is adding cpu load to rtems..18:17
wpwrakkyak: ah, yes, the atusb side isn't ready yet18:18
wpwrakkyak: you'd need two atbens to play with it for now18:18
wpwrakgrmbl. how do i leave gmu ?!?18:21
wpwrakremote login, then killall -1 gmu.bin did it :)18:22
kyakwpwrak: alt+q18:27
kyakwpwrak: ok, so i'll be following stefan's progress for atusb.. Will wait till it becomes usable18:29
wpwrakwe should chain him to his computer so he works harder :)18:32
kyaknot that chaining someone to something ever made their work better :)18:35
kristianpauli dotn get it, why i should use fdopen with sockets?18:36
wpwraki think galley owners would disagree with you disagreeing :)18:36
wpwrakalso provides great motivation for maintaining strict workplace safety :)18:36
kristianpaulokay, i'll just use it 18:38
wpwrakkristianpaul: should or shouldn't ?18:39
kristianpaulmust? :-D18:40
kristianpaul"fdopen function can be used to initialize a FILE structure from a file descriptor; however, file descriptors are a purely Unix concept not present in standard C."18:40
kristianpauli fear about that word "initialize"18:41
kristianpaulbut i guess it mean tell how to do something18:41
kristianpaulor read something18:41
wpwrakrejon: fixed my ben :) on another try, it actually booted okay. now t has the latest kernel18:42
rejongreat wpwrak 18:42
kristianpaul"The fdopen() function associates a stream with a file descriptor. " ah, thats different18:43
wpwrakkristianpaul: most sources indicate that you shouldn't use FILEs for sockets18:43
wpwrakkristianpaul: appears to be fragile and unportab;e18:43
kristianpauli dint coded osgps btw :-)18:43
kristianpaulbut fragile make me thing..18:44
kristianpaulwpwrak: what you suguest instead, use send and receive?18:44
wpwrakyup. they're the best18:45
wpwrak(the whole send* and recv* family. depends a bit on what exactly you need)18:45
kristianpauli just want to stream data from mm1 to PC18:46
kristianpaulwell to osgps actually18:46
kristianpaulwich implemented the FILE for socket18:46
kristianpaulmay be they never tested it..18:47
kristianpaulmay be because i'm new to this, but a FILE for sockets looks usefull :D18:47
wpwrakit does. and it often works, apparently18:48
wpwrakbut that's noot really by design18:48
wpwrakso you can have surprises :)18:48
kristianpaulshame, when i tought all stuff in *nix was about files :)18:54
kristianpaulpassing messages..18:55
kristianpaulwpwrak: "Not selling hardware " fro FSIL slides,you still mean sell is the point but not hardware it self, instead what it does, the value it generates, that we should sell? 19:02
wpwrakkristianpaul: naw, i wasn't so philosophical ;-) it's just a way of saying that, if you intend to make (and sell) hw, you're at risk of patent attacks19:04
wpwrakkristianpaul: (while the risk is much lower for sw)19:04
wpwrakkristianpaul: (especially free sw)19:05
kristianpauli still confuse, man i should asssisted to you talk, or get mad reading the slides :S19:05
wpwrakso that strategy is not an option. and you bhave to kepe listening for therest i say :)19:06
kristianpaulbut when you mean hardware, you dont meant a product19:07
wpwrakyes, i mean hw as a product19:07
Action: kristianpaul wonder how many patent attacks19:09
Action: kristianpaul sharism/qi-hardware had have19:09
wpwrakwell, openmoko suffered one19:09
wpwrak(via pulster)19:10
wpwrakqi-hw 0 so far19:10
kristianpaulthe same who sell frerunner?19:10
kristianpaulor i'm confused?19:10
wpwraka distributpr19:10
wpwrakthat one19:10
kristianpaulindeed,youhave interesting histories to tell from OM :)19:11
wpwrakit was something stupid. we installed mp3 codecs on the freerunners without even thinking there could be a problem19:11
wpwrakwell, there was19:12
kristianpaulah, i undertand wolfgang know :)19:12
wpwrakthat's why wolfgang is so careful about codecs now. once bitten, twice shy19:12
wpwrakyup :)19:12
wpwrakrejon: does the hotel have good heating ? i don't feel the cold so much anymore, which is probably a bad sign :)19:14
rejonthat's why i always pack a jumprope19:14
rejonon trips19:14
rejonyes good heating19:14
wpwrakkewl. or, rather, hot :)19:20
kristianpaulwait wait wait,NFS seems to implement fdopen hum19:21
rejonwpwrak, forgot to intro you to carol from google last night19:21
rejonshe runs google summer of code19:21
wpwrakat, google recruiting :)19:22
kristianpaulmay be not19:22
wpwrakshe was seated atthe corner, on your side ?19:22
kristianpauldamn hot here,i need a fan19:22
wpwrakkristianpaul: GRRRR19:23
kristianpaulsh*t fdopen is not in K&R, hum..19:33
kristianpaulah, posix,...19:34
kristianpaulokay, let use berkely sockets as wikipedia saids :)19:35
wpwrakrejon: does your ttavel kit by any chance contain a usb hub ?19:52
rejonwpwrak, nope19:53
rejonthere is a hw store next to this hotel eko btw19:53
wpwrakoh, kewl. they thought of everything :)19:54
wpwrakhmm, forecast predicts 19 C for tomorrow. then down to 13. wow. sunday minimum will be -1 !20:01
wpwrakrejon: heading towards the hotel20:11
rejonwpwrak, cool...i'm not coming out yet ;)20:12
wpwrakcya in a bit20:12
wpwrakrejon: arrrived at hotel :)21:53
wpwrakrejon: are you still here ?21:53
viricgrh somehow I don't get into this channel automatically.22:10
--- Thu Jun 30 201100:00

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