#qi-hardware IRC log for Friday, 2011-06-10

wpwrakfirst, a little firmware upgrade (very minor - more cosmetic than anything else): http://downloads.qi-hardware.com/people/werner/wpan/bindist/atusb-256ac01.bin00:01
wpwrakatrf-id: #92 Thu Jun  9 20:56:54 ART 201100:01
wpwrakmd5sum: e3e6c9d230cb32fb98b948cf0a79358600:02
wpwrakheh, this also removes a race condition. nice ;-)00:06
qi-bot[commit] Werner Almesberger: atrf-xmit.c (xfer_one): pulse SLP_TR instead of sending TRX_CMD_TX_START http://qi-hw.com/p/ben-wpan/703ce8100:11
qi-bot[commit] Werner Almesberger: atrf-xmit: use receive interrupt instead of polling http://qi-hw.com/p/ben-wpan/abe923d00:11
qi-bot[commit] Werner Almesberger: tools/atrf-proxy/atrf-proxy.c (cmd_zero): reduce poll delay from 100 ms to 2 ms http://qi-hw.com/p/ben-wpan/78ab01700:11
qi-bot[commit] Werner Almesberger: atusb/fw/board.c (reset_rf): increased 1 us delays to nominally 2 us http://qi-hw.com/p/ben-wpan/256ac0100:11
qi-bot[commit] Werner Almesberger: atrf-xmit: tightened checking of interrupt behaviour http://qi-hw.com/p/ben-wpan/4d4e13200:11
wpwraktime to rebuild the tools. (both host and ben)00:11
wpwrakokay. now the checking is pretty much the same as in the kernel. in fact, it's even tighter. the kernel driver is a bit sloppy on waiting for the PLL lock.00:20
wpwrakah, regarding the atusb firmware, there will be at least one more update in the next days, to add proper interrupt support. so at least the early birds may get a chance to play with dfu-util ;-)00:22
tuxbrainrebuild install done00:22
wpwrakalso stopped and then restarted atrf-proxy ?00:23
tuxbrainwhat now?00:24
wpwraktime atrf-xmit -p 15  usb net:ben  10000:24
wpwrakthen  time atrf-xmit -p 15 net:ben usb  10000:24
wpwrakboth should say 100/10000:25
wpwrakand take 3-5 seconds00:25
wpwrakeven better :)00:25
tuxbraintime atrf-xmit -p 15 net:ben usb 10000:25
wpwrakwell .. it actually seems a bit fast00:26
wpwrakls -l `which atrf-xmit`00:26
tuxbrain-rwxr-xr-x 1 root root 79763 2011-06-10 02:16 /usr/local/bin/atrf-xmit00:26
tuxbrainyes soon it will be early morning instead of late night00:27
wpwrakthe best time of the day ;-)00:27
wpwrakthe ben->usb time look okay. for some reason, usb->ben is a bit slower here. always takes ~4.25 seconds00:29
tuxbrainyes I'm seeing the intro of one of the meeting tomorrow: ladys, gentleman.... go to hell and let me sleep!00:29
wpwrakone last test: download this one  http://downloads.qi-hardware.com/people/werner/tmp/atrf-xmit00:30
wpwrakthen  time ./atrf-xmit -p 15 usb net:ben 10000:30
tuxbrainyour have a 64bits machine?00:30
wpwrakand it's ubuntu 10.04-ish00:30
wpwrakso i hope we're binary compatible :)00:31
tuxbrainsorry dude  no sexual intercurse  thank you :P00:32
wpwrakit didn't work ?00:32
wpwrakhere would be a static binary: http://downloads.qi-hardware.com/people/werner/tmp/atrf-xmit.bz2  (also 64 bit)00:33
tuxbraintime ./atrf-xmit.1 -p 15 usb net:ben 10000:34
tuxbraintime ./atrf-xmit.1 -p 15 net:ben usb 10000:34
tuxbrainI have a better ben/usb/host combination than yours :P00:35
wpwrakheh ;-) may actually be something like that :)00:35
wpwrakthe difference doesn't make a lot of sense, so i'm happier with your result. just wanted to be sure it's for real.00:36
wpwrakdeleting those x86 executables again ...00:36
wpwrakalright. i think i can let you sleep now ;-)00:37
tuxbrainthat's all ? :P00:37
wpwraki'll try to think of an explanation for the MOSI/nSS/etc. weirdness00:37
wpwrakit actually makes a little sense that nSS would now pop up too, because it was not properly configured before00:38
wpwrakand if the problem is truly in the pull-up/down resistors, MOSI, nSS, and SCLK would all be potential troublemakers00:39
wpwrakso what i'll do is measure the pull-up/down strength. maybe this will give me some new ideas.00:40
tuxbrainMy next steps will be flash a bunch of atusb, setup properly the web content of the shop and make a annoucement post (not publish  until I have at least a bunch of checked atben/atusb) 00:41
tuxbrainok just tell me how we should mesure this00:41
wpwrakalso, please review my commercial description00:42
tuxbrainyes, I was also on it... how to test the distance?00:42
wpwrak(measure) ideally with a lab power supply, a few resistors, and a very sensitive voltmeter00:42
wpwrakah yes, that one. lemme look it up ...00:43
wpwrakon the ben, run  atrf-txrx -P -p 300:45
wpwrakon the pc: atrf-txrx -P 1 -p 300:45
wpwrakthen move them apart until you stop seeing a . appear every second00:45
wpwrak(all this is without atrf-proxy and such)00:46
wpwrak(and you either want a very long usb cable, or run on battery ;-)00:46
tuxbrainthe a will apear on ben or on PC?00:47
tuxbrainon ben I supose00:47
wpwrakboth will show similar things00:47
tuxbraintime to sleep00:47
wpwrakthey synchronize with each other. what the characters mean:00:47
wpwrak. = successful exchange (both ways)00:48
wpwrak* = one-way reception (i.e., the other side sent something but didn't acknowledge our last packet)00:49
wpwrak- = a garbled packet00:49
tuxbrainand then             the silence00:49
wpwrak+ = no reception, so we timed out and generated a new packet (you get this one only on the machine that had the timeout value after -P) the other simply waits00:50
tuxbrainok I wil do that test tomorrow, now time to sleep if there were dots inside my rf madness house will be a total success :)00:51
wpwrakyou can also change some more parameters. if you're in wlan hell, maybe first check the spectrum with00:51
wpwrakatrf-rssi -g00:51
wpwrak(either on the pc or on the ben)00:52
tuxbrainoh beautiful :)00:52
wpwrakah, on the ben, if you have jlime. otherwise, no SDL :)00:52
tuxbrainlot of movement on the 11-13 band :) as expected00:53
wpwrakto set the channel in atrf-txrx: -c CHAN00:53
wpwrakthe default channel is 1500:54
tuxbraincool is the most clean one :)00:54
wpwraknote that the channel numbers are not the same as for WiFi. 00:55
tuxbrainwell whatever 15 is almost clean00:55
wpwraki could actually show the wifi channels too. there's enough space ...00:55
wpwrakthe ping test works with the 250 kbps standard data rate. if you want to go faster, you can use  -r 500k  or  -r 1M  or  -r 2M00:57
wpwrakthe faster you go, the shorter the range00:57
tuxbrainok will be also fine to see how much distance we can achive at those rates too00:58
wpwrakyup :)00:58
tuxbrainnow yes I will go to sleep, mmm dreaming on 2Mbit rates :)00:59
wpwrakspeedy dreams ! ;-)00:59
wolfspraulwpwrak: wrt Nokia CTO calling it quits, just wanted to say that in most such large companies, the mode of operation of holders of specific titles is quite standardized, so that people inside the company quickly understand how things can work.01:41
wolfsprauland 'cto' is often misunderstood from the engineering side as the 'super engineer' who has all sorts of powers, but actually not at all.01:42
wolfspraulmost of the time, the CTO has the right to stop any project, but he cannot start any project01:42
wolfspraulit's a very defensive role, like an insurance against something that is technically impossible01:42
wolfspraulso even though I agree with you that Nokia is on a massive self-destruct mission, the CTO stepping down wouldn't worry me from the outside01:43
wpwrakah, i thought there were more strategic elements in that role01:43
wpwrakso it's basically controlling01:43
wolfspraulcan stop anything01:43
wolfspraulthat's a big power01:43
wolfspraulbut cannot start anything :-)01:43
wolfspraulthat's the downside, he he01:44
wolfspraulof course I don't know Nokia specifically, but that's how it is normally01:44
wolfspraulwith that title01:44
wpwraka job for hooligans ;-)01:44
wolfspraulthe CTO will be blamed if something becomes a tech disaster01:44
wolfspraul"why didn't you stop this?"01:44
wpwrakalways good to have some whipping boys around :)01:45
wolfspraulwell, you know engineers are sometimes, what shall we say... a bit too optimistic? :-)01:45
wpwraknot if you ask them whether they can do X until firm deadline Y ;-)01:46
wolfspraulah, Rich Green :-) That was the guy at Sun who always hated Sun's open source strategy..., if I remember correctly01:48
wpwrakhated it for what ?01:49
wolfspraulI forgot the details, could be wrong. RIP Sun...01:50
wolfspraulhe's a long-time Sun guy anyway, 20 years or so01:51
wolfspraulI'm still rooting for Nokia, they don't deserve to go down like this. oh well.01:51
wpwrakseems that they don't get what they deserve01:52
wolfspraulsome strange things happening there. No more forecast for 2011?01:52
wolfspraulthat's unbelievable01:52
wolfspraula company this size without forecast? unbelievable, really01:52
wpwrakwell, the board of directors could still attempt a u-turn. not sure if this would help, though01:52
wolfspraulI cannot begin to imagine what's going on inside. Probably massive infighting everywhere.01:52
wolfsprauloh everybody comes to the office every day to fight, that's for sure01:53
wolfspraulunfortunately it will make it sink faster01:53
wolfspraulno more forecast, aargh01:53
wpwrakDocScrutinizer mentioned that nokia employees are already publicly daydreaming about killing elop. let's see how long until one snaps.01:53
wolfspraultheir credit rating is close to junk already01:54
wolfspraulagain, "no more forecast", you just cannot do this01:54
wolfspraulyou just cannot01:54
wolfspraullet's see how many months they sail without forecast now01:54
wolfsprauleventually they have to come out with one again, he he01:54
wpwrakmaybe the forecast would be worse than all speculation ? :)01:54
wolfspraulyes maybe01:54
wolfspraulbut then it still needs to be on the table IMMEDIATELY01:55
wolfspraulsuch a large company doesn't live in a hole, or another planet01:55
wolfspraulit has a whole ecosystem around it01:55
wolfspraulwith "we have no forecast" they initiate mass defections everywhere that will be hard to turn around01:55
wolfspraultheir partners will draw down investments01:55
wolfspraulall over the place, they won't even know where01:55
wpwrakmaybe it takes a while for these whose heads would be on the chopping blocks to move their assets to places without extradition agreements ?01:55
wolfspraulI have no forecast for Sharism, and that's unprofessional01:56
wolfspraulbut at least there are not many partners that suffer from that01:56
wolfspraulbut Nokia????01:56
wolfspraulI cannot believe it01:56
wolfspraulwe all learn. I don't remember ever watching something like this.01:57
wolfspraulhow many months without forecast? My bet: 601:57
wpwraksharism ought to have a roadmap. when you talk to potential investors, wouldn't they notice the red flag as much as you do in the case of nokia ?01:57
wolfspraulnow I can proudly say I am as clueless about the future as Nokia01:57
wolfspraulwhatever happens happens01:57
wolfspraulwe have the same professional management :-)01:57
wpwrakwell, maybe tepco are setting the new standards. nokia are just very trendy :)01:58
wolfspraulyes, roadmap and forecast same thing here01:58
wolfspraulroadmap + numbers = forecast01:58
wolfspraulnumbers are very important because the busines needs to be confident to be able to deliver that01:58
wolfsprauland who should know if not them?01:58
wolfspraulshould outsiders go around and make a forecast for Nokia????01:59
wpwrakwell, 0 income and all the same old expenses until the windows phone in ... 2013, right ? that's easy to calculate :)02:00
wpwraklayoffs don't really change the equation in that time frame02:00
wolfspraul"It added that it could no longer provide a full-year forecast."02:01
wolfspraulunbelievable to me, I'm amazed.02:02
wolfspraulmaybe they will go out of business faster than I thought02:03
wolfspraulwe should help them with a forecast :-)02:03
wolfspraulnot Werner's brutal one though02:03
wolfspraulmaybe Joerg buys another N900...02:03
wpwrakmaybe you could offer your services as a consultant on how to survive with two-figure monthly sales ? :)02:05
wpwrakthe whole nokia thing makes me wonder if there aren't some major things we haven't been told yet. the current self-destruction does seem to be justified by results like these: http://www.heise.de/newsticker/meldung/Nokia-kommt-besser-ins-neue-Jahr-als-erwartet-1231812.html02:26
wpwrakif that meant an emergency, companies HP ought to be long dead already02:26
wolfspraulwpwrak: "Mass production didn't dent our view of reality."02:47
wpwrakand nice use of "shear"02:52
kristianpaulis that a joke?02:57
wolfspraulkristianpaul: why do you think it's a joke?02:58
kristianpaulwell, i dont see the sense of that kind of annoucement coming from openmoko list02:59
wolfspraulI've long stopped trying to define 'insanity'. Let me see what Wikipedia says :-)03:00
kristianpauli mean shiftd? sharing videos.. well is not bad, just it cant fit on my mind03:00
kristianpaulfrom where it come of course :-)03:00
wolfspraulit's all fine by me :-) joke or not, laugh or not, success or not, whatever...03:00
wolfspraulthey launched FRAME* before, I think. something like that. getting busy now :-)03:00
Action: DocScrutinizer thinks to know some ID-photos that show up on wiki's def of insanity03:02
DocScrutinizer(<wolfspraul> maybe Joerg buys another N900...) I'd actually like to, as soon as I got the money, alas I think they're not on sale anymore03:03
wolfspraulDocScrutinizer: see!? I knew it. Our forecast builds up...03:05
DocScrutinizerwhich in itself is a weird thing, I rarely ever seen a company stopping sale of a model prior to offering a successor ;-)03:05
wolfspraulDocScrutinizer: we can study some exceptional business moves in Nokia...03:07
wolfspraulsomeone's gotta try those too, for the MBA textbooks...03:07
DocScrutinizerwtf is frm.fm?03:07
wolfspraulan expensive TV running Windows03:07
wolfspraulI doubt it's available actually :-)03:08
DocScrutinizerso OM.inc isn't a spinoff of OM ?03:08
wolfspraulno spinoff, same thing03:08
DocScrutinizerbtw  here not even a picture on that site is available03:09
wolfspraul1. FreeRunner 2. Wikireader 3. FRAMED* 4. shiftd03:09
DocScrutinizersee my comment about wikipedia above03:09
DocScrutinizerI'm not sure about the forecast bit you mentioned03:11
wolfspraulyou mean it's ok to give up on it?03:11
DocScrutinizerI don't know what you're refering to. Should I know?03:12
wolfspraulI think with this kind of statement ("unable to provide"), many partners around them will put investments on hold immediately03:12
DocScrutinizer( <wolfspraul> DocScrutinizer: see!? I knew it. Our forecast builds up...)03:12
wolfspraulDocScrutinizer: that Nokia said they are unable to provide a forecast for 201103:12
wolfspraulah sorry03:12
wolfspraulDocScrutinizer: we were chatting about Nokia earlier03:12
wolfspraulI was shocked about Nokia saying they are unable to provide a forecast for 2011 now.03:12
DocScrutinizergot as much03:13
wolfspraulso we were joking that we need to help Nokia with a forecast. Werner suggested 0. But I thought we can add at least one for your N900 purchase :-)03:13
wolfspraulthat's all03:13
wolfsprauljust meaningless chat for fun :-)03:13
wolfspraulmaybe Meego will be spun off in some way, including manufacturing ability03:13
DocScrutinizeraaah, related to my desire for more N900s03:13
wolfspraulthat'd be cool03:13
DocScrutinizerI thought we talked about frm 2 years ago, got puzzled03:14
DocScrutinizer(meego spinoff) that alterego guy didn't come back yet, also didn't answer further oings from me03:16
DocScrutinizerguess it's been just another booze-talk03:16
DocScrutinizerI really wonder how they managed to get that tick Elop in their fur. Obviously he's killing nokia for good, and no doubt what's the plan behind it03:23
Action: DocScrutinizer yawns and waves03:27
wpwraktoo bad it's still way too early for the 100th anniversary of the hindenburg. the crashing and the burning would be quite comparable03:29
Action: DocScrutinizer ponders an idle peek on NOK trend03:30
DocScrutinizermust hit penny stock range meanwhile03:31
tuxbrainanybody understand that shiftd.com ...? 05:12
tuxbrain(framed) wow an i5 for an kiosk setup!!! not much power saver indeed05:19
wpwraktuxbrain: i measured the pull up/pull down resistors. they're quite weak. pull-up about 80 kOhm, pull-down about 150 kOhm. this should still allow them to work, but there's not much margin for error.10:37
wpwraktuxbrain: on atben, all the GPIO tests pass without trouble, correct ?10:38
qi-bot[commit] Werner Almesberger: fab/m1/: quick and dirty adaptation of prettygerbv and drl2gerber to M1 http://qi-hw.com/p/eda-tools/2a6496d13:31
qi-bot[commit] Werner Almesberger: fab/m1/prettygerbv: feeble attempt at showing at least some layer connectivity http://qi-hw.com/p/eda-tools/4e4849113:56
tuxbrainwpwrak: yes atben no GPIO problem at all14:46
tuxbrainwpwrak: (resistors) you mean next batch we will have to use another ones?14:47
rjeffriesOne patent to rule them all: http://c4universe.com/blog/2011/06/09/man-files-patent-to-protect-his-godly-powers/15:08
wpwraktuxbrain: (atben) excellent15:14
wpwraktuxbrain: (next batch) no, it may simply be that the test is too fragile. it doesn't test a particularly critical condition. its main purpose is to spot shorted or broken traces, and that's easiest in this particular condition15:15
wpwrakrjeffries: only in the US ... :)15:16
wpwraktuxbrain: i'm still curious about the voltage levels, though. i haven't quite convinced myself that the whole thing is truly harmless.15:22
qi-bot[commit] Werner Almesberger: eda-tools/fab/gerb2pbm: convert a Gerber file to a PBM image http://qi-hw.com/p/eda-tools/f6b485915:39
wpwraktuxbrain: and i'm also curious what your range test yield :)15:39
rjeffrieswpwrak once tuxbrian starts selling the atben and atusb products, what communication can the buyer do "out of the box??18:34
kristianpaulhehe :-)18:51
kristianpaulSurelly software support can be added in nanonote-openwrt at least for "dirt" pan?18:52
wpwrakrjeffries: right now, very little. but yes, dirtpan isn't so far behind19:00
rjeffries_wpwrak love your branding. sounds mildly p0rn0graphic19:37
wpwrakrjeffries: [...] and all the really simple things (UDP) probably work perfectly19:42
wpwrakrjeffries_: you mean "dirtpan" ?19:42
rjeffries_yes sir19:46
rjeffries_and it is also cheaper than dirt19:47
wpwrakheh, you should have seen one of the things i did at openmoko. there's a script that installs all the firmware and stuff on a device fresh from the factory.19:47
wpwrakguess what i called it ? "devirginator".19:48
wpwrakthat one was actually shipped to the factory, where a bunch of factory workers, women of gentle age, were running it from the command line.19:49
wpwrak(something like one or two thousand times, once per device)19:49
Action: whitequark just received a chinese 1W WiFi dongle20:14
wpwrak1 W ? wow20:14
whitequarkthere is a Ralink RT3070L chip and (as it was found on one site) a custom Chinese-made RF frontend20:14
whitequarkyeah. beyond every certification and illegal to import to Russia (there is a 10mW limit for any unlicensed RF range)20:15
wpwrak"chinese-made" because it says so or because it looks bad ? :)20:15
whitequarkbecause everything on the box is written in Chinese :)20:16
wpwrakokay :)20:16
whitequarkalso, it contains a BackTrack Linux LiveCD and an illustrated instruction for cracking WEP points20:16
whitequarkand there's more!20:17
whitequarkthey've decided to make it a little cheaper... and didn't soldered an RF shield.20:17
whitequark(there is place for it on PCB)20:17
wpwrakif they also included instructions for getting through the chinese firewall, then they'd have made the whole world their enemy ;-)20:17
whitequarkso the components are all exposed (it also has an external antenna, so it definitely can and should be shielded itself)20:18
wpwrakwell, maybe add a picture of Muhammad on the box, just to be sure20:18
wpwrakwith those little wpan devices, we probably get away without shielding. 1 W wifi, though ... :)20:19
whitequarkan FCC nightmare 20:19
whitequarkoh, and the customs declaration was funny too. they've specified, literally, "NETWORKING"20:20
kristianpaulhow do i recover a git stahs?20:20
whitequarkkristianpaul: git stash apply20:20
wpwrakthat's when the FCC inspector asks for the keys to the storage room with the hellfire missiles ...20:20
kristianpaulif i have more that one stash?20:20
whitequarkkristianpaul: git stash apply stash@{1}20:20
whitequarkand so on20:21
kristianpaulthe point is i need to create another branch20:21
wpwrakgit stash list  ... then pick the one you need20:21
kristianpaulbut from the commit and uncommtied changes i had done20:21
wpwrakhmm, currently 24 stashes on ben-wpan. not that i'd really intend to use any of them ...20:21
whitequarkwpwrak: I'm going to take some photos of the device, probably20:21
kristianpaulwpwrak: he :-)20:22
kristianpaula hurry up when changing branches i guess?20:22
wpwrakwhitequark: be careful. that sort of pornography may be illegal even in russia20:22
whitequarkwpwrak: anyway, the thingie works really great. my internal Atheros card reports an RSSI of -35 to -40 on my network (AP in ~2 meters), and the chinese one says it is -15 dBm. and also it sees twice as lot of neighboor APs...20:24
wpwrakwow, so they also improved the receiver sensitivity. that's good.20:26
whitequarkthey say it can work up to 500m indoors and 1000-2000m outdoors (with _omnidirectional_ antenna!)20:26
wpwrakyou and that card, ruler of the city !20:27
whitequarkMoscow is somewhat bigger than 2km in radius... it's more like 18km20:28
whitequarkbut hey, it was $20, and there are 18dB yagi's of roughly the same cost20:29
kristianpaulah, just WEP, i tought WPA2 was already brokable20:29
whitequarkkristianpaul: you can capture a 4-way handshake and then run a dictionary attach20:30
whitequarkgiven how stupid people sometimes are, that works in surprisingly many cases20:30
kristianpaulyou'll wonder, all my neighboords use wpa2 :/20:31
whitequarkwpwrak: check http://www.dealextreme.com/p/35405 and http://www.dealextreme.com/p/3475920:31
wpwrakwhitequark: you have to pick a strategically important place for your lair with the RF to rule them all. no point in domineering the outskirts. peasant revolutions are so last millennium.20:31
Action: kristianpaul needs a good antenna for satellite listening20:32
wpwrakoh, quite large. thought it was just a USB dongle.20:32
whitequarkwpwrak: the dongle itself is much smaller than it looks like20:33
Action: whitequark searches for his camera20:33
kristianpaulreceivr is almost done, actually migrate a code made for lpt is not so hard when having a ubb :-)20:33
whitequarkkristianpaul: actually, I didn't buy that for using someone others' wifi illegally. there is simply no point in that when you have a cheap symmetrical 20Mbit connection. it's more for connecting to free public networks in strange places, or in some other areas where AP is far away20:34
wpwrakkristianpaul: so you're #3 to actually do something with UBB, after tuxbrain and myself ? welcome to the club !20:35
whitequarkwpwrak: the intention of buying ben was somewhat damaged after I've seen that the only option is FedEx, for 2/3 of Ben's cost. is this really a sole option?20:36
kristianpaulwhitequark: (20Mbit) i hope i could said that20:36
wpwrakwhitequark: wolfgang can probably ship it anyway you want. or if not wolfgang, then tuxbrain20:37
whitequarkwpwrak: if the tuxbrain can ship the whole package (ben, atben and atusb), that would be really great20:37
whitequark*without "the"20:37
kristianpaulwhitequark: (illegally.) oh no, i'm happy with my conection, just wondering the world wide implication of brokable security layes in all kind of wireless networks/infrastucures20:37
wpwrakwhitequark: the big advantage of fedex is of course that there's less risk of endless customs delays and such20:38
whitequarkwpwrak: somehow DealExtreme has no problem getting through the customs with all sorts of weird and definitely illegal stuff20:38
whitequarkand they just use tracked airmail20:38
wpwrakwhitequark: maybe the magic is in the "NETWORKING" ;-)20:39
Action: kristianpaul loves cheap tax-free airmail/postmail20:39
whitequarkwpwrak: probably. I've once received a package from them with declated contents, literally, "WEIRD STUFFS"20:39
wpwrakhopefully in chinese ;-)20:41
whitequarkin plain English, of course20:41
whitequark(otherwise I wouldn't be able to read that...)20:41
wpwrakyou could have google translated it :)20:42
wpwrakfot atben/atusb, i used 847330. Parts and Accessories of the Automatic Data Processing Machines20:42
wpwrak(i very carefully left out any mention of RF)20:42
Action: whitequark wonders what number "weird stuffs" correspond to... 20:43
whitequark42 maybe20:43
whitequarkwpwrak: so, is that customs stuff more strict in Europe?20:44
wpwrakthey may be about RF. europe and us.20:45
wpwrakin argentina, they usually hunt for things like power cables and adapters. easier to find.20:45
whitequarkhuh? power cables? what's the point for restricting import of these?20:46
whitequark(customs stuff) I was asking because I've seen declarations from China, Britain and US, and none of them had any catalogue numbers like an example you've shown20:47
wpwrakif ther'es a power cable in the box, it probably connects to something. this something will almost certainly not have the argentina seal of approval. hence a reason to stop the shipment and see how much it is worth to you to still have it ...20:47
whooohi, does anyone have a working version of fw_env.config for the nanonote, found in a git repo, but it doesn't work, fw_printenv just tells me Bad CRC and some default values20:48
kristianpaulhum interesting device http://www.dealextreme.com/p/qh-2-high-power-gps-l1-l2-signal-jammer-blocker-with-ac-car-charger-1217-1237-mhz-1565-1575-mhz-5587120:48
wpwrakkristianpaul: wanna start a fight with whitequark ? his 1 W wifi vs. your jammer ? it's the wrong frequency, though. but maybe the harmonics are strong enough anyway ;-)20:49
kristianpaulwpwrak: no, actually i need one of those for gsm signal blocking 20:50
whitequarkwpwrak: there are all sorts of jammers on DX. GPS, GSM, WiFi, 3G, anything you could imagine20:50
wpwrakkristianpaul: girlfriend on the phone all the time ? ;-)20:50
whitequarkI think that maybe half of products they're selling are (illegal, of course) transceivers, amplifiers and jammers20:51
whitequarkwpwrak: (power cable) do they open the boxes then? or use X-rays?20:51
wpwrakwhitequark: dunno if they xray them before opening. they look at pretty much everything.20:52
wpwrakwhitequark: they even find cheques in mail. and steal them, of course.20:53
Action: whitequark now knows a postal system worse than russian20:54
kristianpaulhere is http://www.dealextreme.com/p/portable-rechargeable-cdma-gsm-dcs-phs-3g-cell-phone-signal-jammer-blocker-4717520:56
kristianpaulprice is good :-)20:57
whitequarkwpwrak: the photos: http://rghost.net/10316841 http://rghost.net/10316851 http://rghost.net/10316861 http://rghost.net/10316871 http://rghost.net/1031688121:11
whitequarkmaybe you'll spot something interesting in RF pathway...21:11
whitequarkah, I forgot the dimensions. it is 59x75x13mm, and with a 17cm antenna21:17
tuxbrainwpwrak: anything do you want to improve on the testing thing? I can take it as is as valid?21:56
tuxbrainanything more you want I try?21:56
jluistuxbrain, can I help in any way du ring the testing?22:01
tuxbrainjluis I go to girona this weekend and I will bring the thigs there to perform as much as posible but if there's is work to do on the tuesday I will be happy to accept help :)22:02
jluisIt's a deal dude!22:03
jluisjust point me to the docs ;)  22:04
jluishere or by mail. It's your chice? Links ! Links ! links!22:05
tuxbrainjluis: http://downloads.qi-hardware.com/people/werner/wpan/prod/index.html22:11
tuxbrainok first range test , awesome, 12meters in my 2.4GHz messy enviroment with one wall in the middle :)22:12
tuxbrainafter that the number of the packages lost increase exponentialy22:13
jluistuxbrain, thans at least I will have the soft ready 22:15
tuxbrainabout 7 meter with two thick walls in the middle (walls= 15cm of pure no holes brick) 22:16
wpwraktuxbrain: i'd love to know the voltage of the pin the gpio test complains about22:17
tuxbrainwhat pin it is ?22:17
wpwraktuxbrain: (range) wow ;-)22:18
tuxbrainI think we will be save to say 10m indoor :)22:18
wpwraktuxbrain: i think the last time it was nSS, no ?22:19
tuxbrainI will test in outdors tomorrow or on sunday.22:19
tuxbrainyes but where is it :)22:19
wpwraktuxbrain: (10 m) sounds good, yes. very nice. so proper SMT improved the RF performance. that's what i hoped :-)22:19
wpwrakfiring up gimp ...22:20
tuxbrainwow about same range with 500k22:23
tuxbrainthings start not not be just dots at 11-12 meters22:24
wpwraknice :)22:25
wpwrak10 m is the intended range for WPAN (presumably free from obstructions). so all this is very good.22:25
tuxbrainvoltage of that pin is 3,25V after the test run22:28
tuxbrainalso on a fresh plug is 3,25V22:29
wpwrakand during this ?  atrf-gpio -p HHHHoH1H.HxHxHxHh.ohzoHoHH delay=1000022:29
wpwraklet's make it   atrf-gpio -p HHHHoH1H.HxHxHxHh.ohzoHoHH delay=10000022:30
wpwrakthen you don't have to rush ;-)22:30
wpwrakah, does it complain ?22:30
tuxbrainit goes down to 3,23V during the 10s and then return to 3,2522:32
kristianpaul10m, cool !22:32
tuxbrainno complains22:32
tuxbrainstarting 1Mbit range test22:33
Action: kristianpaul :D22:33
wpwrakhmm. but it should complain ;-)22:34
wpwrak"make usb" still fails then ?22:34
kristianpaul(should complain) i dont get this guys..22:34
wpwrakmompls. afk for 5 min22:35
tuxbrainI have to buy a bigger flat :) 1M still maitains sending packages after some orientation at 12 meters :) , with 250-500k orientation was not so meaningful but with 1M yes, changing NN position can turn off the signal almost completly at 12M but at right position almost all packages are sended recieved (2% lose)22:40
tuxbrain12meters not 12Mbits :P22:40
LunaFrizzleHi all22:41
LunaFrizzleHard times22:41
tuxbrainhi luna22:41
LunaFrizzleStudying for the finals22:41
LunaFrizzleHi tuxbrain :)22:42
LunaFrizzleAre you... the whole tuxbrain into one person ?22:42
tuxbrainalmost :)22:42
tuxbrainwe are 1 working full time there (me) and Victor my partner working at partial time22:43
LunaFrizzleOh :o22:45
LunaFrizzleI knew the tuxbrain website but I though so far that it was a bigger team22:45
tuxbrainI have to be glad then :)22:46
LunaFrizzleI was really impressed back with the Arduino + Ben Nanonote video22:47
LunaFrizzleAs well as the other SDIO one22:47
LunaFrizzleAs a matter of fact, I plan to dig up this when my finals will be finished to make some things22:47
LunaFrizzleOr at least try22:47
LunaFrizzleSDIO and other bitbanging stuffs make me realize that computers (regular ones)  can be used as "very big arduinos" in some way22:48
tuxbrainTest at 2M 7meters, with good orientation, after that even with orientation signal start loose Not bad :)22:48
LunaFrizzleI'll have to dig up also how to code for the parallel port :)22:48
tuxbrainLunaFrizzle: Great I will be very happy enrol you in the ArduNote thing :)22:49
LunaFrizzleI just need to become rich so I'll be able to buy lot of electronic components22:49
LunaFrizzleAnd a decent soldering iron :/22:50
Fusinhi qiots ;)22:50
tuxbrainThat was my excuse to create tuxbrain :P22:50
LunaFrizzleOh becoming rich haha22:52
tuxbrainOk I have to retry this test with more atben/atusb pairs but, I think I will be save to put in the comercial description22:52
tuxbrain250k to 500k 10m22:52
LunaFrizzlehi Fusin 22:52
tuxbrain1M-2M 5 meters22:53
LunaFrizzletuxbrain, well so far the Efika (at last ! it's available !) looks very interesting but I don't want to sold one of my kidneys to buy one ;)22:53
tuxbraincommon you sell your kidneys very cheap dude :)22:54
tuxbrainLunaFrizzle:  where are you from?22:55
LunaFrizzletuxbrain, France22:57
LunaFrizzletuxbrain, I've just spent too much money buying an Amstrad CPC 464, an Amiga 500 and a C6422:57
LunaFrizzle(this month)22:57
LunaFrizzleSo I need to seriously calm down22:58
tuxbrainHAHAHA ok ok don't worry, Efika's will still there next month :)22:58
LunaFrizzleHehe tuxbrain I wish that was the only thing I wanted to buy23:00
tuxbrainmeanwhile you put me on the first item on "next month shoppinglist"  for me is enough , you al so want a pair of atben/atusb? they are just out of the fab oven, they are still hot }:P 23:02
LunaFrizzleI wonder if it's possible to do Serial prot programming when you are using an USB to RS232 cable, or do you need to have documentation from the manufacturer23:02
LunaFrizzleHaha non I already have an arduino and a nanonote tuxbrain but thanks :)23:03
LunaFrizzleBut I'll definitely buy a UBB for sure23:03
tuxbrainwell dudes, I need to sleep badly.23:04
LunaFrizzleEven though I found that 4+ euros may be a bit expensive for a small piece of pcb but I'm sure that the more reasonnable price you could get23:04
tuxbrainnext range test atben vs atben two NN talking each other wirelessly for first time , good night23:06
wpwraksorry, took a bit longer23:07
wpwrakLunaFrizzle: (ubb) it's all a function of the quantity. pcb-making has a high setup cost. the per-unit cost is almost nothing. so if tuxbrain could sell thousands of units, the price would be maybe 50 cents.23:10
wpwrakah, tuxbrain escaped. well, then i have more time to mess up the firmware :-)23:12
qi-bot[commit] Werner Almesberger: atusb/fw/: SET_INTERFACE can now select among interfaces (changes user_setup) http://qi-hw.com/p/ben-wpan/89d13ce23:13
qi-bot[commit] Werner Almesberger: atusb/fw/: define USB bus current (40 mA) in board.h based on measured value http://qi-hw.com/p/ben-wpan/3cc92cc23:13
qi-bot[commit] Werner Almesberger: atusb/fw/usb/dfu.h: define DFU interface descriptor, for later sharing http://qi-hw.com/p/ben-wpan/7a2c78523:13
wpwraki realized that i hadn't done current measurements yet. so i did them today. all is as expected. cleaned up the values indicated on USB23:13
wpwrakLunaFrizzle: tuxbrain's fabs in spain are a bit expensive, but they also help if there are small problems. you basically get a review as part of the package. considering that these things were my first designs that were industrially produced and i think that this was also tuxbrain's first board, having more eyes look at things is a good idea.23:19
whitequarkwpwrak: may I ask, what were you doing previously?23:20
wpwrakLunaFrizzle: if you go with a half-baked design to china, chances are that they'll just do what you told them, or improvise some correction that may or may not be successful.23:20
wpwrakwhitequark: before qi-hw ? openmoko (system level stuff). before that, i had a project with philips research (real-time file i/o in the kernel). my general background is low-level system sw, kernel and such. you may have heard of one of my earlier projects: the lilo boot loader.23:22
wpwrakwhitequark: and i also have a background in networking. layer 2 and above. i ran the linux-atm project as part of that work.23:23
wpwrakwhitequark: my current foray into electronics is something where i can relive childhood memories of being a clueless newbie again ;-)23:24
LunaFrizzlewpwrak, ok I understand :)23:25
LunaFrizzlewpwrak, I would like to learn more about "bitbanging" (understanding "programming the parallel por/serial port") but so far as you know, modern computers (and laptops since I'm using a netbook 95% of the time) doesnot have either of theses outputs. However, USB to Serial cable does exist. As USB to Parallel port are only designed for printer and nothing else, are USB to RS232 usable when it comes to serial port 23:28
LunaFrizzleprogramming ? I know that mini pcie card does exist for adding both parallel and serial port on my netbook but they are a bit pricey (60euros min).23:28
wpwrakLunaFrizzle: it all depends a lot on what you want to do. the usual usb-serial cables have the problem of using voltages regular logic doesn't like (+/- 12 V). there are 3.3 V cables, though, but you won't find them at the shop down the street.23:32
wpwrakLunaFrizzle: also, if you take a usb-to-"dump chip" solution, you have high overhead and latency 23:33
wpwrakLunaFrizzle: often, it's preferable to have "local intelligence" right next to the circuit you want to talk to. that way, you can send higher-level commands to the local cpu, and let that one worry about low-level operations.23:34
LunaFrizzlewpwrak, afair I have an 3.3v USB to RS232 cable at home (i bought it to use with an arduino mini)23:34
wpwrakLunaFrizzle: examples of such an arrangement would be ben+UBB or the MCU+RF combination in atusb23:34
LunaFrizzlewpwrak, hum ok23:34
wpwrakFTDI+arduino would be another example23:34
wpwrakactually, i think they're now switching to atmel usb chip. but it's the same concept.23:35
LunaFrizzlewpwrak, I have that23:35
whitequarkwpwrak: just found that DX has also 1.5W and 2W versions of the dongle23:35
LunaFrizzlewpwrak, so buying a mini pcie card for parallel port is a good idea ?23:36
wpwrakwhat's nice about the ben is that it's a real linux system. so you have all the usual system features (networking, file system, etc.) and a regular development environment, with all the usual libraries and such23:36
LunaFrizzlewpwrak, here's the only one I've found, it's name is the PEX1S1PMINI23:37
wpwrakLunaFrizzle: depends on how many I/Os you need. i'd rather use something like the ben. it's also cheaper if something goes wrong and you end up sending, say, 220 V into the computer ;-)23:37
LunaFrizzleNo no wpwrak I'm planning to use that on a netbook powered by a 11v battery23:37
LunaFrizzle14v PSU at worst23:38
wpwrakwhitequark: i suppose that, once you decide that standards compliance is not your thing, your just take it as far as you can ;-)23:38
LunaFrizzle(19,5v actually, my bad)23:38
wpwrak11-14 V should already make a nice cloud of smoke ;-)23:38
LunaFrizzlewpwrak, I'll start with the Ben, but the question was for "what when i'll need more I/O and make bigger stuffs ?"23:39
LunaFrizzleand take over the world23:39
wpwrakLunaFrizzle: if you don't need the LCD, you can reuse those signals. they're available at the bottom of the PCB: http://downloads.qi-hardware.com/people/werner/tmp/tp-map-ba.png23:40
LunaFrizzlewpwrak, oh nice !23:44
LunaFrizzlewpwrak, the more I use my ben, the more I think I should buy another one that would stay in the box... "just in case"23:44
LunaFrizzlewpwrak, so far the chargin led broke a little when I teared it appart one time23:45
wpwrak"broke a little" ;-))23:45
LunaFrizzlewpwrak, maybe the soldering was too weak I don't know23:45
LunaFrizzlewpwrak, yeah yeah sorry ._.23:45
LunaFrizzlewpwrak, it still works but it have to be in the "good position"23:46
LunaFrizzlewpwrak, led pins, very interesting23:46
wpwrakthe lead-free solder is a pain. it's very brittle. the old leaded ones were flexible and would also provide mechanical support. with lead-free, that's almost completely gone.23:46
qi-bot[commit] Sergey Kukunin: Fix segmentation fault when try show Manual on Explorer or other service apps http://qi-hw.com/p/gmenu2x/28fe39c23:48
LunaFrizzlewpwrak, so far that means that if I use the lcd pins i'll have to access the terminal via the TXD/RXD pins ? Also that my Ben will be... or won't be like before ever (but anyway)23:48
wpwrakLunaFrizzle: well, that's for the case that you want to dedicate a board at that purpose, yes. you also still have USB, which is probably what you would use most of the time.23:49
LunaFrizzlewpwrak, oh yeah I ironically forgot about it23:50
wpwrakLunaFrizzle: if you want to optimize your process, you can also get a dedicated prototyping board, such as the new SIE. that should be cheaper than a ben. but i don't know how well the software support is. http://linuxencaja.com/images/9/9d/SIE_KICAD.png23:50
LunaFrizzlewpwrak, now we're talking :)23:50
Fusinhi LunaFrizzle  & all who are awake here ;)23:51
wpwrakLunaFrizzle: but if you just need a small number of boards for experiements, you're probably better off with a single platform. the few euros you safe with a cheaper board you waste maintaining parallel development environments (which in the end means that you'll abandon most of them)23:52
wpwrake.g., i have a while pile of openmoko phones with debug board and so on. i could use them also for embedded control. but it's just too much work. so i do everything with ben+ubb now.23:54
wpwrakalso platform dependencies get worse if you share code with others. e.g., i can't just write things that work only with gcc 1.something from '92. if i'd do things only for my personal use, i could do this, and never have to worry about updates.23:56
--- Sat Jun 11 201100:00

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