IRC log for #buglabs on 20100916

00:50.56*** join/#buglabs grund (~grund@173.85.82.111)
00:50.56*** mode/#buglabs [+o grund] by ChanServ
00:53.10grundHi bcruskie !
00:53.32bcruskieheya grund
00:54.23grundthat bug is on .190 now if you want to test with it
00:54.59grundI forget what I left on it, I was kind of torqued at it earlier :-D
00:55.14grundbut it has the daily of the customer specific build
00:55.50bcruskiecool I just did some basic SDK connection tests with it last night
00:56.43grundoh, there we go. It has lcd, gps and wondermod
00:57.10bcruskiewas the lcd attached and working last night?
00:57.34grundyes
00:57.56grundbut I'm looking at it now and it only shows lcd in http://bug/modules
00:58.12grundbut dmesg shows them all
00:58.51grundthough jconnolly told me the wondermod has to have the services specifically installed / downloaded in the sdk before it will show
00:58.59grundbut I was hoping gps would show
00:59.10grundoh.
00:59.12grundwait
00:59.14grundI'm a dork
00:59.55bcruskiewell lcd is displaying in sdk tonight... that's good
01:00.04grundwhat I said about the wundermod is true, but gps might not be showing up because I for got to depmod -a
01:00.37grundI've figured out where that packet firmware libertas_tf error comes from
01:02.52grundand I wiped everything off after copying without getting rid of the source and build link from /lib/modules/$(uname -r)/
01:03.46bcruskienice will that keep us from digressing to libertas_sdio?
01:04.20grundyeah, it turns out it's a really poor generic error, that any error condition will print
01:04.57grundand it looks like we pick up a slightly large frame now and then and puke that error before ignoring the packet
01:06.10grundso the modules are all loaded and none of them show in /modules on the web side
01:06.13grund:-(
01:10.06bcruskieyeah just lost the lcd here
01:10.47*** join/#buglabs finsprings_ (~finspring@c-71-61-78-230.hsd1.pa.comcast.net)
01:10.47*** mode/#buglabs [+o finsprings_] by ChanServ
01:10.58bcruskiehey there finsprings
01:11.09finsprings_hey bcruskie
01:11.12bcruskiegrund are you going to try rebooting it?
01:11.37finportgrund, if you haven't lined up a place to stay you're welcome to our house for the weekend of the great race
01:11.55finportonly condition would be that you let the chickens out in the morning and lock them in at night
01:12.11bcruskieand don't bring them into the hot tub
01:12.21finportpreferably not, yes
01:14.08grundfinport: Thanks! but we're staying with friends
01:14.15finportok, no worries
01:14.16grundand we love chickens
01:14.45grundwe go out there pretty often, so we'll have to meet up for a brew some time
01:15.03grundbcruskie: yeah, I guess I better boot it again
01:16.39finportgrund, yeah i'd like that
01:17.01finportthe Penn Brewery reopened after a brief hiatus and is locally owned again
01:17.07finportworth a visit if you've not been
01:17.22finportOktoberfest is this weekend so you'll miss it but still
01:18.00grundYay, that's good news about PB
01:18.13grundYeah, kids have a football laden weekend coming up
01:18.56finportshould get jconnolly over here some time too
01:19.01finportis jconnolly around jconnolly jconnolly jconnolly ?
01:26.00finportThe founder and CTO of dropbox just said Notesy was "quite the slick app" and congratulated me!
01:26.07finportswoons dramatically
01:26.48*** join/#buglabs jconnolly (~jconnolly@firebug.buglabs.net)
01:26.48*** mode/#buglabs [+o jconnolly] by ChanServ
01:27.01*** mode/#buglabs [+o jconnolly] by ChanServ
01:27.55bcruskiewow nice recognition finport
01:28.06finportyeah I know
01:29.16finportApple aren't letting me use the keyword "dropbox" so I was asking the Dropbox folks if they could help. (There are other non-Dropbox,Inc apps that show up for that keyword though so I don't know why I got singled out.
01:29.31finportjconnolly, was that you coming back or just your computer
01:29.47jconnollyi'm back
01:29.56finportJCONNOLLY!
01:30.10finportI'm totally flumoxed with this camera jni man
01:30.11jconnollyforgot to screen out for irssi.  i should've just backgrounded it and forgrounded it into my screen session
01:30.13bcruskiefinport how are the sales on it so far?
01:30.32finportbcruskie, pretty good given how hard it is to find without it showing up under dropbox ;-)
01:36.09jconnollysorry lost connex in the tunnel
01:36.14jconnollyjust heading home on the train now
01:36.53jconnollyfinport: I'll have a tasty lookaroo.  standard stuff for editing the CDT environment vars for BUG configuration?
01:37.07finportmostly
01:37.12finportI had to add the C compiler
01:37.16finportmaybe I should check in what I've got
01:37.17jconnollyi.e. path to includes/libdir
01:37.21finportit shouldn't break anything
01:37.46jconnollyeither way I can back it out of build.  now that accenture's its own branch I'm not as hatefully cautious about daily build breakage
01:37.56finportok I'll put it in svn now
01:37.57finportsec
01:38.45finportjconnolly, don't want .cproject right?
01:38.52jconnollyi like finport too... sounds somehow sleeker than a finsprings
01:38.54finportor orgeclipse.cdt.core.prefs
01:39.04jconnolly.cproject is fine, and orgeclipse.cdt.core.prefs
01:39.05finportor is .cproject where the envs are?
01:39.06jconnollycheck it all in
01:39.26jconnollyeither of those, I forget now.  CDT is magic
01:39.48jconnollyI tried to get under the hood with the project metadata and all that and making it better but it mystified me at some point
01:40.22jconnollywe should just use env vars that correspond to bitbake's but that's for rainy days
01:41.29finportokay jni and modlet committed to trunk in svn
01:42.00jconnollycool.  I'm on my netbook with a 12B/s connection so it'll have to wait
01:42.14finportnic
01:42.15finporte
01:42.19finportwant another to go with it?
01:42.33jconnollygetting rid of the eeepc?
01:42.43grundok, rebooted and still nada in the modules list :-(
01:42.44finportit sits all sad under my bed
01:42.46grundhi jconnolly
01:42.46finportmakes me feel guilty
01:42.58jconnollygrund: can you pastbin /var/log/felix.log
01:43.27jconnollyand for S&Gs /var/log/messages
01:43.38finportjconnolly, http://dl.dropbox.com/u/395371/Screenshot-Properties%20for%20com.buglabs.bug.jni.camera%20.png
01:43.41finportenv settings I'm using
01:43.54finportnote the new CC one
01:43.54CIA-48Bug Labs: dfindlay * r12084 /bug/trunk/com.buglabs.bug.jni.camera/ (17 files in 5 dirs): camera: first non-working pass at most of the API
01:43.54CIA-48Bug Labs: dfindlay * r12085 /bug/trunk/com.buglabs.bug.module.camera/ (8 files in 4 dirs): camera: first non-working pass at new camera API
01:44.07finportohai CIA-48
01:44.17finportgood job I keep my checkin comments to-the-point
01:45.02jconnollythanks finport
01:45.27grundfelix.log http://grund.pastebin.com/9EAFBqS6
01:45.37jconnollygrund: it's definitely felix but any reason you compiled your own kernel?
01:45.51jconnollyjust for the libertas stuffs?
01:46.42jconnollyhm that's not good about the gps jni
01:47.12grundYes, just for the libertas
01:47.35jconnollycool.  i checked in a recipe for the accenture kernel where the defconfig is without all the extra mojo
01:47.37grundI narrowed down where those stupid messages are coming from
01:47.41jconnollynice
01:47.55jconnollyhates coming back into the office having the screen covered in libertas pritnk's
01:48.20grundyeah, I had a "That's it!" moment around 3 today and dove in
01:48.20jconnollygrund: I'll have a look at the gps jni tomorrow too
01:48.58jconnollyyour setup is what exactly?  GPS slot 1, what else?
01:49.19jconnollyhm slot 0
01:49.46jconnollyi wonder if the gps jni changed for the NEW gps module in trunk
01:50.00jconnollywhere there was no ioctl to select an active antenna
01:50.03jconnollyI wouldn't doubt it
01:50.11grundwundermod in there too
01:50.11jconnollyhow can it be this late and we're discovering this...
01:50.35jconnollywell allbets are off for an unsatisfiedlinkerror spawned from bug.bmi
01:50.47grundbecause we haven't had working gps modules with out hardware shenanigans?
01:50.52jconnollyyeah
01:50.56jconnollythat's probably why
01:51.06jconnollypunches the commuter in the seat across
01:51.30jconnollyi think we did a "whoa, it works" thing in June
01:51.39jconnollyand hasn't even been looked at since
01:51.39grundI had a bitch of a time trying to short the pads last night and this morning, I begged it off on reiss and it took him a while too
01:51.56grundyeah, like with the accelerometer on the wunder
01:52.06jconnollyyou guys are getting rusty
01:52.10grund<PROTECTED>
01:52.10jconnollybunch landlubbers
01:52.17jconnollyhey hey hey hey
01:52.36grundmy fault on that one
01:52.46jconnollyi don't see shit on module.accenture in your log
01:52.52jconnollydid you sent it to your BUG?
01:53.08grundno, this was a 3:30 filesystem
01:53.17grunddidn't send anything to it yet
01:53.35jconnollyyeah you won't be seeing much about wundermod without sending the code over to BUG
01:53.46grundyeah, that much I remembered
01:53.52jconnollycool
01:54.00jconnollydon't make me grep my mental logs
01:54.08jconnollywaits for the HR
01:54.26grundgrund>
01:54.26grundthough jconnolly told me the wondermod has to have the services specifically installed / downloaded in the sdk before it will show
01:55.17grundok, not too happy with how colloquy copies and pastes. meh
01:55.32jconnollynot bad
01:55.46jconnollyirssi sticks this italicized I's in there
01:55.58grundwants his bitchx back
01:57.15finportBTW ^R working on the bug across reboots is my fav thing ever
01:59.40grundyeah, major time saver. I should have the bug scp bash_history to my pc on shutdown so I can put it on each new rc I load
02:00.04finportor put sshfs on it
02:00.12finportor the dropbox daemon
02:00.13finport:-)
02:00.18jconnollyamen
02:00.33jconnollyI think I had fuse and sshfs compiled in
02:00.34jconnollyshhh
02:04.04jconnollybcruskie: how does SDK look, I asked kgilmer to wait for feedback about the state of the SDK with only GPS/LCD/WONDERMOD
02:04.15jconnollybefore merging all the other happies in
02:04.43bcruskiejconnolly: will PM you
02:05.13jconnollyk
02:17.13grundha, I searched for if_sdio_read_rx_len in google and the first item was something jconnolly pastebin'd
02:17.31jconnollygrund: happens to me 3 times a day
02:17.51jconnollygrund: don't be bashful, email the guys at OLPC if you need to get under the hood
02:18.02jconnollythey approached gumstix about using their work
02:18.54jconnollyactually i think i will email them tomorrow
02:18.55grundI figured out the basics, we read a max chunk size value from the firmware, and then put up that message if we get a larger chunk
02:19.08grundit's probably arbitrary or reading the wrong value
02:20.26jconnollynice.  I'm going to email them tomorrow though.  I think they'd appreciate hearing back from people pushing it into use.....
02:20.37jconnollywhy not?
02:20.56jconnollyI blogposted about it and I think an actual person anonymously responded about it
02:24.07jconnollyok, pulling into station.  bbiab
02:53.17grundok, I need to sleep before I jump on the train back to the office
02:53.52grundjconnolly|away: I'll hit the libertas_tf folks up on irc tomorrow and ask about this
03:45.09jconnolly|awaynice
10:01.02*** join/#buglabs infobot (~infobot@rikers.org)
10:01.02*** topic/#buglabs is Bug Labs IRC Channel. New with logged, indexed goodness, thanks to Tim Riker. see here: http://ibot.rikers.org/%23buglabs/
11:00.59*** join/#buglabs likewise (~likewise@82-170-243-215.ip.telfort.nl)
11:54.24*** join/#buglabs GNUtoo (~GNUtoo@host55-144-dynamic.54-79-r.retail.telecomitalia.it)
13:03.54*** join/#buglabs TheMammal (~daver@firebug.buglabs.net)
13:09.39*** join/#buglabs ahernandez (~ahernande@firebug.buglabs.net)
13:18.58*** join/#buglabs slashdevnull (~slashdevn@fire.innovationsw.com)
13:19.11*** mode/#buglabs [+o slashdevnull] by ChanServ
13:21.55*** join/#buglabs desenna (~quassel@firebug.buglabs.net)
13:22.00*** join/#buglabs desenna (~quassel@firebug.buglabs.net)
13:29.42grund<PROTECTED>
13:29.51jconnolly|awaynailed
13:29.56grundnarf!
13:30.21*** mode/#buglabs [+o grund] by ChanServ
13:32.23grunddamn, had to look up how to change my password
13:32.26GNUtoolol
13:32.53GNUtoo/msg NickServ identify *******************************
13:33.45GNUtooanyway, I've setup the bugduino + von hippel
13:34.00GNUtooin such maneer:
13:34.05jconnolly28 asterisks?  that's an easy one to guess
13:34.41GNUtoobugduino->ftdi->usb_cable->von-hippel
13:34.43GNUtooand it works
13:34.44GNUtoobut...
13:35.13GNUtoomaybe I should put the von hippel in a slot under the bug device instead of a slot that is upper
13:35.24GNUtoobecause there is no space for displays otherwise
13:35.38GNUtooI should look after my broken touchscreen input...
13:35.52GNUtooisn't there anyway to unbreak it without sending it?
13:39.40GNUtooshould look if bugduino is usable trough java
13:50.28jconnollygrund: can you open a RI about the GPS issue?
13:50.44*** join/#buglabs mpeddi (~mattp@firebug.buglabs.net)
13:51.38GNUtooGPS issue?
13:51.45GNUtoobtw gps works fine for me
13:52.03grundjconnolly: will do
13:52.17GNUtooapart that it need a calibrated touchscreen for navit
13:56.58*** join/#buglabs bugvish (~vish@firebug.buglabs.net)
13:57.57ahernandezslow claps
13:58.05jconnollystops the clock
14:08.37*** join/#buglabs kgilmer (~kgilmer@firebug.buglabs.net)
14:24.59grundhaveahennessy: I figured out what condition causes those "problem fetching packet from firmware"
14:25.07haveahennessyyeah?
14:25.32grundin if_sdio.c on line 393 there is a check if (chunk > size) {
14:26.53grundhere is some debug output
14:26.55GNUtooah libertas_sdio,
14:26.56grundpacket fragment (291 > 249)
14:26.56grundpacket fragment (249 > 242)
14:26.56grundpacket fragment (349 > 237)
14:27.17grundfirst one is chunk, second one is size
14:27.32GNUtooI've a question abuot libertas_sdio, libertas_sdio uses an irq, where the irq should be set , for instance if you want a gpio_to_irq
14:27.35*** join/#buglabs bballantine (~Adium@64.70.116.116)
14:27.55grundsize is pulled from the card and chunk is pulled from the frame
14:29.12haveahennessyhmm
14:29.36haveahennessyyou got ideas why the chunk is bigger than the size?
14:29.49grundNo, didn't get that far last night
14:30.19grundwe use the 8686, right?
14:33.47grundsize comes from if_sdio_read_scratch
14:34.15haveahennessyyes 8686
14:34.24grundreads RX packet length when interrupt (with IF_SDIO_H_INT_UPLD bit set) is received
14:34.57grundcool, it's like I know things
14:36.25grundthen after sdio_readsb chunk is calculated
14:36.28grundchunk = card->buffer[0] | (card->buffer[1] << 8);
14:37.22grundso either the firmware tosses up the wrong value, the calculation is wrong, or the sdio chunk size adjuster is screwing things up
14:37.51*** join/#buglabs kgilmer (~kgilmer@firebug.buglabs.net)
14:37.51haveahennessyyou do know things
14:38.23grundif only the fix was one of those things
14:39.16haveahennessyhaha
14:42.47*** join/#buglabs mgrundy (~buglabs@firebug.buglabs.net)
14:49.55*** join/#buglabs bugvish (~vish@firebug.buglabs.net)
14:52.36*** join/#buglabs GNUtoo|laptop (~gnutoo@host55-144-dynamic.54-79-r.retail.telecomitalia.it)
14:52.58jconnollyhmm no stefan_schmidt yet
14:53.10GNUtoo|laptopjconnolly, what's the issue?
14:53.23jconnollyGNUtoo|laptop: policy-kit from oe-dev...
14:53.36GNUtoo|laptophow does it fails?
14:54.31jconnollyI'm trying to allow root to set NM system connections, I get policykit prompts, but I'm not allowed to modify policy-kit's settings...
14:54.31jconnollyfor example:
14:54.31jconnollypolkit-read-auth-helper: needs to be setgid polkituser
14:54.33jconnollyand
14:54.33jconnolly** (polkit-gnome-authorization:3279): WARNING **: Error: code=3: uid 0 is not authorized to read all authorizations (requires org.freedesktop.policykit.read)
14:54.43GNUtoo|laptophmmm
14:55.08jconnollytried adding root to group polkituser
14:55.58GNUtoo|laptophmmm no
14:56.03GNUtoo|laptopI don't have that on my desktop
14:58.21GNUtoo|laptopI don't know policy-kit at all
14:58.22GNUtoo|laptopbut
14:58.26GNUtoo|laptopis org.freedesktop.policykit.read a dbus thing?
14:58.58jconnollympeddi: sup with those Verizon usb760 dongles, we have any more?
14:59.08jconnollylooks like it GNUtoo|laptop
14:59.17jconnollyi'll have to delve in deep for this one
14:59.22jconnollykisses his weekend goodbye
14:59.33mpeddiWe have one, not provisioned
14:59.37GNUtoo|laptopjconnolly, bitbake mdbus2
14:59.42jconnollympeddi: that'll do
14:59.48GNUtoo|laptopif it's not masked by angstrom
14:59.50GNUtoo|laptopyou'll love it
15:00.21jconnollyGNUtoo|laptop: wow, I will love it
15:00.25jconnollyI'll try it
15:02.19GNUtoo|laptopindeed it seems to use dbus
15:03.12pipixmpeddi - when can we expect those other bugs?
15:03.32mpeddiI see a fresh box over here pipix
15:03.47pipixyah
15:03.58pipixi'm going through that one
15:04.05pipixthe other box is for maurizio
15:04.23mpeddiah
15:08.18*** join/#buglabs stefan_schmidt (~stefan@p5B036A16.dip.t-dialin.net)
15:08.40stefan_schmidthi all
15:09.01grundhi stefan_schmidt
15:09.18stefan_schmidthi grund
15:21.19grundjconnolly: didn't you add an eject or usbmodeswitch for the u760 dongles?
15:21.39jconnollydidn't seem to work grund
15:21.48jconnollystill got hatred about /dev/sr0
15:21.56jconnollyi'll continue working on it
15:21.59jconnollywas just discussing
15:22.20jconnollygrund: if you're feeling bold and want to give it a shot, it's probably just a %k should be a %n or something in the udev rule
15:22.43jconnollygrund: eject should be there.  looks like usb_modeswitch isn't
15:22.46jconnollyI'll add it to default
15:23.15grundok, eject worked manually, I thought you had added a rule already
15:23.42grundif eject is there I don't think we need the modeswitch
15:24.12jconnollyeject did indeed work manually
15:24.34jconnollybut couldn't seem to get it to work on hotplug
15:24.50GNUtooudev rules?
15:24.52jconnollythe way I verified that NM worked with the 760 was with an rc script to eject /dev/sr0
15:25.12jconnollythat seemd to do the trick, but it's hackery and I want udev to do what udev does
15:25.31jconnollyright now I'm going to fix GPS tho
15:26.12grundyeah, I'll see if I can find what I used on the brick
15:30.06jconnollygrund: got a 760 right here
15:35.10CIA-48Bug Labs: jconnolly * r12086 /bug/trunk/com.buglabs.build.oe/meta-bug/packages/tasks/task-bug.bb: com.buglabs.build.oe: trunk, add usb-modeswitch to image
15:37.23grundjconnolly: I have mine attached
15:37.55jconnollygrund: you want the real deal?  so you have the right product/vendor ?
15:38.53grundis the product different on the vz models?
15:39.17jconnollyyou have a sprint dongle, yeah?
15:40.49grundyes
15:44.22kgilmeris there a page somewhere that describes how to setup vnc for bug?
15:44.57bugvishyeah
15:45.23bugvishkgilmer http://wiki.buglabs.net/index.php/Accenture_2.0#Helpful_Hints.2FKnown_Issues
15:45.33kgilmerthx bugvish
15:45.40kgilmeri take back those things i said about you
15:46.23bugvish<PROTECTED>
15:54.34jconnollygrund: found it
15:54.44jconnollystrings lib/libGPS.so _Z68Java_com_buglabs_bug_jni_gps_GPSControl_ioctl_1BMI_1GPS_1ACTIVE_1ANTP10_Jv_JNIEnvP9__jobject
15:54.46jconnolly!=
15:54.50*** join/#buglabs desenna_ (~desenna@firebug.buglabs.net)
15:54.55jconnollycom.buglabs.bug.jni.gps.GPSControl.ioctl_BMI_GPS_ACTIVE_ANT
16:01.42*** join/#buglabs finsprings (~anonymous@c-71-61-78-230.hsd1.pa.comcast.net)
16:01.42*** mode/#buglabs [+o finsprings] by ChanServ
16:07.22*** join/#buglabs finsprings (~finspring@c-71-61-78-230.hsd1.pa.comcast.net)
16:07.27*** mode/#buglabs [+o finsprings] by ChanServ
16:14.21finspringsjconnolly: dunno if you saw emails but progress made last night on camera
16:14.33finspringsso you can spare yourself the tedium of looking at my jni code
16:14.43jconnollyboom
16:14.44jconnollysolutions
16:14.59jconnollyfinsprings: I think I want to peek anyway.  I don't want you to be the Only Camera Guru
16:15.12jconnollyMr. v4l2, Tear Down This Wall!
16:15.19jconnolly<PROTECTED>
16:15.26finspringsRemind me to run the obfuscator on that code before I submit next time
16:16.36ahernandezle 5guys train!
16:16.47grundreally?
16:16.52kgilmerahernandez for reals?
16:17.05grund<PROTECTED>
16:17.15grundI'm in
16:17.23jconnollyfor realz
16:17.24jconnolly+1
16:17.25ahernandezrollll ouuuut
16:17.26jconnollylet's go
16:45.06*** join/#buglabs bug_pip (~agibb@firebug.buglabs.net)
17:04.46stefan_schmidtjconnolly: ping
17:08.04jconnollystefan_schmidt: pong
17:08.28stefan_schmidtjconnolly: I might have something for you wrt the polkit problem
17:08.39stefan_schmidtjconnolly: give some more minutes to test it out
17:09.44stefan_schmidtjconnolly: One thing is for sure. The ownerships for /usr/libexec/plokit* are busted
17:11.44jconnollysweeeeeet
17:12.03jconnollyi did a rather frustrated find / -name \*pol\* to check that kind of thing
17:12.12jconnollyI saw /usr/share and tried that.
17:12.24stefan_schmidtjconnolly: yup, there we go. Being able to save the connection now for all users :)
17:12.46jconnollystefan_schmidt: my hero
17:12.54stefan_schmidtjconnolly: I had at least 1 hour headache about finding out about it before spotting that we fixed that already in oe.dev :D
17:13.02stefan_schmidtjconnolly: commit is on the way
17:13.19jconnollyspectacular.  i'll worry about it for the accenture branch
17:13.38jconnollyi really thought it was going to take much longer to track this down
17:13.39jconnollywhat a relief
17:14.41stefan_schmidtjconnolly: After an hour I was thinking the same, but then it was only half an hour fetching the fix and getting it tested :)
17:15.06stefan_schmidtjconnolly: You also need to worry about the 3G config stuff as I have no usb 3G stick around here
17:15.31stefan_schmidtjconnolly: and btw, is there any magic needed for wifi to work? My board does not show it up at all. Never did.
17:16.41CIA-48Bug Labs: sschmidt * r12087 /bug/trunk/com.buglabs.build.oe/meta-oe/recipes/not-mergeable/ (5 files in 3 dirs):
17:16.42CIA-48Bug Labs: com.buglabs.build.oe: policykit: Import in our overlay and add ownership fixes
17:16.42CIA-48Bug Labs: for RI1409
17:18.13jconnollystefan_schmidt: shouldn't need magic but the /var/log/messages may provide clues if you have problems probing libertas_tf_sdio, etc
17:18.39stefan_schmidtjconnolly: yeah, need to have a look at this. So far it was no bummer though
17:18.57stefan_schmidtjconnolly: next one are the missing icons for me. Does that fit your prios?
17:23.12stefan_schmidthmm, none of the libertas kernel modules has been loaded. Doing both by hand and nothing shows up either
17:23.31stefan_schmidtjconnolly: do all prototype boards have a working wifi?
17:23.47jconnollyyes, they should
17:24.07stefan_schmidtstrange
17:24.14stefan_schmidtthat is all I get from dmesg
17:24.15stefan_schmidthttp://pastebin.com/z7gDd80X
17:24.23grundjconnolly: maybe the usbmodeswitch would be cleaner, eject works but it pukes a lot on the way
17:24.31jconnollystefan_schmidt: icons is a good next prio.
17:24.38stefan_schmidtjconnolly: ok
17:24.39jconnollygrund: ok sounds, good.  grund I'll add it to repo for now if you want to test
17:24.54grundjconnolly: yeah, that will work
17:25.16jconnollyTheMammal haveahennessy, are you aware of any places where uboot might effect libertas_tf_sdio probing?
17:25.28jconnollygrund: ok, i'll ping when it's up
17:26.23jconnollygrund: it's there already
17:26.23jconnollyhttp://repo.buglabs.net/2.0/armv7a-extras/usb-modeswitch_1.1.2-r0.1_armv7a.ipk
17:26.36grundjconnolly: I noticed if you tried to get networking working in u-boot libertas wouldn't load, but we don't have it configured that way
17:27.06jconnollyah yes grund is uboot guru
17:27.09grundand thanks for the package
17:27.54stefan_schmidtgrund: but even when loading the modules by hand no wifi device shows up
17:28.17jconnollystefan_schmidt: you have /lib/firmware/* ?
17:28.30jconnollystefan_schmidt: next message should be
17:28.30jconnolly/var/log/syslog:Jan  1 02:16:28 bug20 libertas_tf_sdio mmc2:0001:1: firmware: requesting sd8686_helper.bin
17:28.31stefan_schmidtjconnolly: yup, standanrd image
17:28.33jconnolly/var/log/syslog:Jan  1 02:16:28 bug20 libertas_tf_sdio mmc2:0001:1: firmware: requesting sd8686tf.bin
17:28.36jconnollywtf
17:28.41CIA-48Bug Labs: sschmidt * r12087 /bug/trunk/com.buglabs.build.oe/meta-oe/recipes/not-mergeable/ (5 files in 3 dirs):
17:28.41CIA-48Bug Labs: com.buglabs.build.oe: policykit: Import in our overlay and add ownership fixes
17:28.41CIA-48Bug Labs: for RI1409
17:28.53stefan_schmidtroot@bug20:~# ll /lib/firmware/
17:28.53stefan_schmidttotal 128
17:28.53stefan_schmidtdrwxr-xr-x 2 root root   1024 Sep 16  2010 .
17:28.53stefan_schmidtdrwxr-xr-x 5 root root   2048 Jan  1 00:00 ..
17:28.53stefan_schmidt-rw-r--r-- 1 root root   2085 Sep 16  2010 Marvell-Licence.txt
17:28.54stefan_schmidt-rw-r--r-- 1 root root   2140 Sep 16  2010 sd8686_helper.bin
17:28.56stefan_schmidt-rw-r--r-- 1 root root 120676 Sep 16  2010 sd8686tf.bin
17:29.32stefan_schmidtjconnolly: no requests in my dmesg, looks to me like libertas fails already during driver probe
17:29.34jconnollylooks right
17:29.38jconnollyyeah
17:29.51stefan_schmidtjconnolly: it should, its the same image we are all using :)
17:31.46grundstefan_schmidt: rev a or b?
17:32.01stefan_schmidtgrund: a
17:32.15stefan_schmidtgrund: nobody is sending me new hardware ;)
17:32.41grundWe're working on it :-/
17:32.43stefan_schmidtMore serious, I can life without wifi for now. But it might make some testing more difficult
17:32.51stefan_schmidtgrund: yeah, just kidding
17:33.12stefan_schmidtgrund: doesn't make sense to ship stuff to europe when still not all problems are shaked out
17:33.32grundstefan_schmidt: it's ok, it's driving me crazy too :-)
17:33.40stefan_schmidtgrund: :)
17:34.54stefan_schmidtjconnolly: hmm, could you give me some hints about what icons you are missing again?
17:35.09kgilmerjconnolly, fyi from yesterday's bug20 build opkg is restarting the world everytime i install a new package.
17:35.29stefan_schmidtjconnolly: So far I see all icons I expect in about, connection-editor and applet
17:37.55jconnollylet me have a look re: icons
17:38.07jconnollykgilmer: restarting the world?
17:39.53jconnollystefan_schmidt: one that i see is broken is password-protected networks
17:40.05jconnollybut you don't have wifi =(
17:41.00jconnollystefan_schmidt: I'll try to gather more and document them on the RI
17:41.10grundstefan_schmidt: I fired up my rev a to see what happens with this morning's customer build and it worked
17:41.12grundmeh
17:41.47grundmaybe pop the sd and fsck it?
17:42.19kgilmeri'll make an RI jconnolly
17:42.32jconnollythx kgilmer
17:42.42jconnollygrund: never worked for him.. =(
17:42.46stefan_schmidtjconnolly: ah, that might indeed be a problem. No wifi hides some icons from me
17:43.13stefan_schmidtgrund: hmm, could do it, but I would really wonder if that helps.
17:43.29stefan_schmidtgrund: doing that often enough when testing new images. AND it never worked for me
17:43.35jconnollystefan_schmidt: maybe worth it to update to a newest uboot.
17:43.46stefan_schmidtjconnolly: yup, will do this
17:43.55jconnollylmk if you need a uImage
17:43.56stefan_schmidtneed to test the build from the git repo anyway
17:44.02jconnollyerr, uboot-image
17:45.43stefan_schmidtjconnolly: a tested one should at least rule it out. A link to yours would be great
17:45.59jconnollyk, let me find it
17:50.50jconnollystefan_schmidt: it's a bit of a mess because you've got a revA 2.0 base
17:50.53jconnollystefan_schmidt: try http://192.168.0.104/u-boot_reva.bin
17:51.13stefan_schmidtjconnolly: forbidden
17:52.14jconnollyeek sec
17:52.41jconnollytry again
17:53.14stefan_schmidtworking
17:53.37jconnolly<PROTECTED>
17:53.45jconnollysufjan stevens makes me sleepy
17:56.00ahernandeztopic
17:56.39jconnollyi want that preserved in time
18:04.30stefan_schmidtok, gf is calling, will have a look at the u-boot and icons things tomorrow
18:04.34stefan_schmidtnight all
18:13.26*** join/#buglabs wwward (~wwward@216.27.144.194)
18:16.42grundjconnolly: adding the following line in 00-bug20.rules worked best
18:16.44grundATTRS{idVendor}=="1410", ATTRS{idProduct}=="5030", RUN+="/usr/bin/eject %k"
18:17.12grundthe modeswitch stuff just ganked up constantly
18:17.46jconnollyok grund.  I'll add it to the build and test it out tomorrow
18:17.56jconnollygrund: can you update RI?
18:17.59jconnollysec for link
18:19.03jconnollyhttp://redmine.buglabs.net/issues/show/1310
18:20.00grundcrap.
18:20.08grundon boot, forgot to test
18:24.03grundphew, works
18:26.14jconnolly*phew*
18:41.15grundlet me qualify that with I get the proper usb id
18:50.16grundok, now I get usbserial
18:56.58grundjconnolly: so then the problem is just NM being able to start it, because something is up with migrating the config from another machine? (part question part statement)
19:01.43jconnollygrund: yes, that's being worked on right now
19:01.51jconnollyhopefully it works ootb shortly
19:02.04grundsweet
19:06.39CIA-48Bug Labs: kgilmer * r12088 /dragonfly/trunk/com.buglabs.bug.simulator/com/buglabs/bug/ (simulator/Activator.java sysfs/LEDDevice.java): com.buglabs.bug.simulator: fix for BS led API NPE.
19:07.30*** join/#buglabs bballantine1 (~Adium@64.70.116.116)
19:09.36grundjconnolly: netm-cli sweet, thanks
19:13.51jconnollygrund: works well
19:40.08wwwardjconnolly - what happens when I create two bundles that use the same Von Hippel module - how is it arbitrated if I call the same ADC channel, for example, from the two different bundles?  I have that condition now, it's slightly difficult to test.
19:47.59jconnollywwward: you'll have to implement a messaging framework.  ADCEventListener or somesuch framework.
19:48.24jconnollywhere you can lock around ADC write/reads and unlock when it's free...
19:48.32jconnollyit should be part of the API, but it isn't
19:48.55wwwardjconnolly Yeah, but it works without that today - is it blocking or luck of the draw?
19:49.15wwwardjconnolly I'm familiar with using synchronize or something of the sort in Java to ensure only one thread gets in there at a time.
19:50.07wwward(but I assume that's not implemented in the VH API today?)
19:50.19jconnollyI'll have to look at the driver, but I don't see any reason you *couldn't* have writeADC in one bundle and writeADC in another bundle both being executed one-after-the-other without a read in between
19:50.32jconnollyno, those individual write and read operations aren't atomic
19:50.37jconnollyi mean they are
19:50.48jconnollybut you can write/write/write/write read, for example
19:51.05jconnollyif you want to write/read only, you need to wrap those two methods
19:51.14jconnollyI could write an OSGi service to do it
19:51.25wwwardNah, save it for the class project. :P
19:51.34jconnollythat was my next suggestion ;D
19:51.51wwward"And here's how we make it work when a flock of geese come to peck at it."
19:52.18jconnollytim gunn style, I like it.  "Make it work!"
19:52.33wwward"must work when 4chan comes to query your sensor"
19:53.02wwwardThanks for the insights.
19:54.26wwwardHrm
19:54.29jconnollyno prob.  I was hoping that condition would never arise, leave it to Bill Ward to dig it up
19:54.35wwwardI wonder if we could convince Gunn to show up to a hackathon and judge it.
19:54.47wwwardwhen I grow up, I want to be a scalability engineer.
19:55.10jconnollyscalability engineers are known by their less flattering titles
19:55.11wwwardif the past has been any indication, my progress to adulthood will be longer than my natural lifespan.
19:55.15jconnolly"No engineers"
19:55.23jconnolly"great idea, it doesn't scale.  next!"
19:55.42wwward?
19:56.01jconnollyit is the role of a scalability engineer to only say "no"
19:56.51wwwardI was unfamiliar with that aspect of it.
19:56.58wwwardThat explains twitter
19:58.09jconnollyi think 1 in 10 news.ycombinator.com posts deal with "i had a great idea.  it's in the toilet because 'it doesn't scale'"
19:58.16wwwardWell
19:59.19wwwardthere's business ideas that don't scale, and there's technology that doesn't scale.  Getting a Bug 1.3 to accept thousands of GETs a second to find out what the value of a single ADC channel is at any particular moment is a scaling problem.  Figuring out how you're going to make a profit off of some obscure concept of personalized widget making is a real challenge.
20:00.44grundahh, the ... of every business plan
20:00.45grund1) build awesome widget thingie
20:00.45grund2) ...
20:00.45grund3) profit!
20:01.13wwwardno
20:01.13wwward3) get VC funding
20:01.13wwward4) Sell firm or vest+bounce
20:01.54grund<PROTECTED>
20:02.14wwwardi should know the song you're thinking about, but I don't.
20:02.24jconnollygo on
20:02.25grund<PROTECTED>
20:02.27jconnollytake the $$$
20:02.36wwwardyeha
20:02.36jconnollywoo woo
20:03.10wwwardeh, it's a valid scheme for an individual to profit from an idea and execution, I suppose.
20:03.17grundthey never put the woo woos and ah-ooooh lawds in the lyric listing
20:03.28wwwardhahahahah
20:03.41wwwardthere's probably a music notation for "woo woo"
20:04.02grundvish?
20:04.21grundbugvish: know that one? ^^^ heh heh
20:06.03bugvishSteve Miller's too advanced for western music notation
20:06.48wwwardoh, snap!
20:06.55bugvishvirtuosic guitar and vocal talent
20:07.34bugvishcant cross the barrier from sonic waves to ink - just wouldnt be right
20:08.42wwwardmust be handed down orally, from one generation to the next. :)
20:09.38bugvishheh
20:10.17jconnollymust be handed down via long car rides with family, to associate with car-sickness and instill a deep negative association for Americana
20:10.51wwwardI missed out on the sickness part.  Of which I'm extremely glad - as boats on the ocean do me in.
20:21.16bugvishkgilmer: what's the best way to figure out what to replace with deprecated functions in the new SDK
20:21.57bugvishI'm lookin for some PublicWSAdmin lovin
20:23.23kgilmerum
20:24.28kgilmerbugvish, you can continue to use publicwsadmin, or you can register your PublicWSProvider implementation with the service registry.
20:24.41kgilmerthey are equivalent, the latter is less code.
20:25.06kgilmercontext.registerService(PublicWSProvider.class.getName(), myimpl, null)  .. etc
20:26.07wwwardIs PublicWSProvider new?
20:26.57kgilmerit's old
20:27.07kgilmerancient
20:27.07bugvishso if i register provider with the registry...
20:27.16kgilmerit's been with us since the very beginning.
20:27.40kgilmer*poof* your WS will be exported in http://asdfadsfadsfa/service/...
20:27.41bugvishhow do i consume the functionality it provides
20:27.58kgilmerwha?  huh?
20:28.03kgilmerin syrup form?
20:28.29bugvishhuh?
20:28.43bugvishsyrup?
20:28.43bugvishin?
20:28.46ahernandezcorn syrup
20:28.51wwward"corn sugars"
20:29.45bugvishif i need to reference the provider in my client class
20:29.45jconnollyyeah, PublicWSProvider is nice until you need to register a service that responds to POST requests
20:29.46kgilmerif you want to converse in meatspace bugvish be my guest.
20:29.49jconnollythen it hates life... shhhhh
20:30.10bugvishi grab a reference from context?
20:30.59jconnollybugvish: http://www.buglabs.net/applications/OSGiLogWS#download
20:31.04jconnollydirt simple example
20:39.13bugvishnice jconnolly that's got five stars
20:39.21jconnollyi rated it
20:39.21bugvishfive star app
20:39.24jconnolly;D
20:39.34jconnollypats self on back
20:39.39wwwarddon't hurt yourself there
20:39.47jconnollyi also created a bot to download it every 25s
20:39.53jconnollyso the download count is high
20:40.38bugvishjavas so much fun isnt it guys?
20:40.47bugvishso straightforward
20:40.56jconnollybugvish: i wrote that 4 weeks after I started here
20:41.01kgilmeran enlightening read: http://en.wikipedia.org/wiki/ARM_architecture
20:41.02jconnollywhere are you with that, cowboy?
20:41.29jconnollykgilmer: nothing you read on wikipedia is true
20:41.58jconnollyrush limbaugh taught me that
20:42.19jconnollyhttp://news.gather.com/viewArticle.action?articleId=281474978523843
20:43.54wwwardjconnolly hey, rush is right, wikipedia is wrong. ;)
20:44.01jconnollyremember the days before BUG had ctrl+R?
20:44.02wwwardit's a left wing conspiracy
20:44.08jconnollyman, what a throwback
20:44.23kgilmerone thing is for sure, the naming conventions for arm cpus is freakishly confusing
20:46.27kgilmerhmmpfh
20:46.31kgilmeroe git repo seems to be down
21:02.02*** join/#buglabs GNUtoo|laptop (~gnutoo@host55-144-dynamic.54-79-r.retail.telecomitalia.it)
21:07.31jconnollyyeah saw that kgilmer
21:13.33*** join/#buglabs bcruskie (~chatzilla@cpe-72-226-49-176.nycap.res.rr.com)
21:14.09*** mode/#buglabs [+o bcruskie] by ChanServ
21:22.04grundis there a easy way to bring up x11vnc at a different resolution than the lcd?
21:22.34jconnollygrund: remove your LCD
21:22.41jconnollydefaults to 1200x1024
21:22.54jconnollyif you boot with LCD attached, enjoy your 320x230
21:22.55jconnolly*40
21:23.52*** part/#buglabs mpeddi (~mattp@firebug.buglabs.net)
21:24.47wwwardjconnolly would be nice to run vnc against a disconnected display.  If you've ever used Windows Remote Desktop, it's a joy to be able to set resolutions independent of the native hardware on the remote device.  Not sure if this can be accomplished with various Linux distributions, but it would be a win.
21:24.54grundhmmm, that worked, just wondering if I tried after I brought it up without the lcd
21:26.05jconnollythe problem is omapfb and xrandr
21:26.10jconnollythere's a RI for it
21:31.49wwwardgrab an umbrella.
21:32.39grundafter that lightning strike? mbn
21:32.53wwwardheh
21:33.04wwwardone landed fairly close by Lincoln Center.
21:33.12wwward(my locale)
21:33.26grundWe just had one hit near the office
21:33.31wwwardwonder how those folks in the Fashion Week tents are getting on?
21:33.40grundkarma for the trees
21:33.46wwwarda tornado would be rather remarkable at this location.
21:34.01wwwardregardless,
21:34.11wwwardthis wave will be over shortly.
21:34.23wwwardhttp://radar.weather.gov/radar.php?rid=dix&product=N0R&overlay=11101111&loop=no
21:55.20*** join/#buglabs kgilmer (~kgilmer@dsl254-120-154.nyc1.dsl.speakeasy.net)
21:59.48grundkgilmer: alive? dry? roof on house?
22:01.11kgilmeryeah i made it in the nick of tie
22:01.12kgilmertime
22:01.15kgilmerfreakish storm
22:01.40kgilmercould see the rain front like a wall
22:01.40kgilmertrees are broken here
22:01.53grundhouse lost roof somewhere in brooklyn, no details yet
22:02.30kgilmeric
22:04.39grund<PROTECTED>
22:19.49*** part/#buglabs slashdevnull (~slashdevn@fire.innovationsw.com)
22:27.01*** part/#buglabs bballantine1 (~Adium@64.70.116.116)
22:40.55*** join/#buglabs kgilmer (~kgilmer@dsl254-120-154.nyc1.dsl.speakeasy.net)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.