**** BEGIN LOGGING AT Thu May 31 02:59:58 2012
May 31 03:09:52 <furan> i understand mmap
May 31 03:10:06 <furan> in the beagle case it is necessary to do any sort of fast io
May 31 03:10:13 <furan> I was hoping to talk to someone who had experience with that
May 31 03:10:26 <XMPPwocky> ah
May 31 03:12:49 <furan> it's not roundabout because it gives you access to physical address space, which, you know, is not virtual address space
May 31 04:11:01 <mdp> hey panto, how's your trip?
May 31 04:11:33 <panto> hey mdp
May 31 04:11:37 <panto> pretty good
May 31 04:11:50 <panto> but sleep pattern is all screwed up
May 31 04:11:53 <panto> how you doing?
May 31 04:12:10 <mdp> heh, I only had to go 1 TZ and a 2 hour flight to visit the mothership
May 31 04:12:29 <mdp> pretty good, dinner with jkridner, Russ, Tartarus tonight
May 31 04:12:47 <panto> heh
May 31 04:12:57 <mdp> panto, we might have even talked similar topics to what you are talking there :)
May 31 04:13:01 <panto> got beer? :)
May 31 04:13:04 <mdp> same topics as here :)
May 31 04:13:07 <mdp> beer, yes
May 31 04:13:13 <mru> beer is a good topic
May 31 04:13:13 <panto> mdp, yeah, it's a big thing apparently
May 31 04:13:23 <mdp> it's what's for dinner I hear
May 31 04:13:31 <panto> I'll try to submit something for plumbers
May 31 04:13:40 <panto> will run it through you if you don't mind
May 31 04:13:53 * panto always trusts mdp with his reviews 
May 31 04:14:10 <mdp> I have a family commitment so I won't make LPC this year..too bad…but will review :)
May 31 04:14:19 <ds2> where is LPC this year?
May 31 04:14:32 <panto> ds2, san diego
May 31 04:14:49 <ds2> !?!? when does submissions close?
May 31 04:15:37 * mdp notes that panto apparently just violated the great firewall
May 31 04:15:56 <ds2> Hmmm
May 31 04:16:20 <Russ> "Client Quit"
May 31 04:16:34 <Russ> they must have come and got him
May 31 04:16:40 * mru has not noticed any great firewall in HK
May 31 04:16:49 <mru> which is where that IP comes from
May 31 04:16:50 <ds2> is the mini habob's normal for this time of the year?
May 31 04:17:01 <Russ> in phx?
May 31 04:17:06 <ds2> yeah, in phx
May 31 04:17:08 <mdp> mru, yes
May 31 04:17:15 <Russ> yes, I'm actually in Dallas right now
May 31 04:17:21 <mdp> mru, he's at connect
May 31 04:17:33 <ds2> Russ: living there or moved there?
May 31 04:17:47 <Russ> travel
May 31 04:17:58 <ds2> oh... no wonder last week was so dusty along 10
May 31 04:18:02 <Russ> my outgoing flight was one of 82 out of dfw that was canceled
May 31 04:18:08 <mdp> Russ, travel from which you can't escape :)
May 31 04:18:29 <mru> mdp: oh, that's where I am
May 31 04:18:53 <mdp> mru, everybody that's anybody is there, of course :)
May 31 04:19:15 * _troll_ too
May 31 04:21:07 <KeatonT> what would cause a beaglebone just to power off.
May 31 04:21:27 <ds2> removing power
May 31 04:22:48 <Russ> thanks folks, ds2 will be here all week, be sure to tip your waitress
May 31 04:23:02 <fiola> Can a BeagleBone power off?  Ie. after cleanly shutting Linux down, ask the PMIC to cycle voltage rails to zero and stay there?
May 31 04:23:11 <ds2> here? :)
May 31 04:23:18 <Russ> it has some protection circuits
May 31 04:23:36 <KeatonT> Russ does it power down if it gets too hot?
May 31 04:23:56 <KeatonT> I wouldn't think it would be but heck I'm desperate.
May 31 04:24:54 <Russ> I'm not aware of any of those power circuits
May 31 04:25:10 <fiola> The Linux kernel clearly has the ability to ask the hardware to turn off power, since a shutdown -h on a PC will turn off the ATX PSU rails despite incoming power still being applied.
May 31 04:25:17 <KeatonT> Here's my scenario. I mounted the beaglebone inside a box similar to what you would normally see home security system controllers installed in.
May 31 04:25:28 <KeatonT> A 18" by 8" metal box.
May 31 04:25:33 <Russ> atx psu? do you have the same beagle bone?
May 31 04:26:05 <KeatonT> The bone is mounted on risers so nothing should be shorting out.
May 31 04:26:19 <fiola> Russ: I'm referring to the Linux kernel having that ability.  And the BeagleBone uses Linux, so the kernel hooks are available if the PMIC can do this.
May 31 04:26:21 <KeatonT> If I run the bone without placing the cover back on the box it runs fine.
May 31 04:26:32 <mru> that's one large altoids tin
May 31 04:26:42 <KeatonT> If I run it with the door on it turns off in an hour or so.
May 31 04:26:48 <Russ> mru, almost done filing away on my penguin mints tin
May 31 04:27:04 <Russ> KeatonT, have you tested what voltage it is running on?
May 31 04:27:09 <KeatonT> The metal box does have ventilation holes.
May 31 04:27:18 <aholler_> check logs
May 31 04:27:19 <KeatonT> Russ 5V 3A
May 31 04:27:26 <Russ> and you've tested that?
May 31 04:27:45 <ds2> what's the ripple @ 0.050mA, 1A, 2A, and 3A?
May 31 04:28:00 <Russ> ripple o_O?
May 31 04:28:06 <Russ> I'm just thinking its out of spec
May 31 04:28:27 <ds2> it could be switching from 0 to 5V due to lack of loading
May 31 04:28:29 <KeatonT> I don't have anything to test it with.
May 31 04:28:52 <KeatonT> but it works absolutely fine if the cover is off the box.
May 31 04:29:03 <KeatonT> It ran for over two weeks.
May 31 04:29:09 <fiola> I seem to remember the manual saying that one channel of the BeagleBone ADC is monitoring the supply.
May 31 04:29:11 <Russ> ds2, or all the way up to 143V!
May 31 04:29:22 <KeatonT> Then I placed the cover back on and it died an hour later.
May 31 04:29:33 <KeatonT> I took the cover back off and it stayed running all day today.
May 31 04:29:35 <ds2> Russ: yes... but I don't think it is stuck at 143V... I think it is bouncing hence 'ripple' ;)
May 31 04:29:42 <KeatonT> placed it back on a few hours ago and it turned off.
May 31 04:30:03 <Russ> KeatonT, are there vents in the top?
May 31 04:30:07 <KeatonT> yes
May 31 04:30:25 <KeatonT> I can't image that it could get hot enough to kill it.
May 31 04:30:40 <KeatonT> The box is in an air-conditioned part of the home.
May 31 04:31:37 <Russ> maybe something electrical is different?
May 31 04:31:44 <KeatonT> Nothing in the logs.
May 31 04:32:21 <KeatonT> The logs are just normal operating stuff then all the sudden startup messages (after I reboot it)
May 31 04:32:34 <KeatonT> To get it working again I have to unplug it and plug it back in.
May 31 04:33:29 <ds2> are you sure you aren't pressing on the reset button?
May 31 04:33:53 <KeatonT> Absolutely positive
May 31 04:35:02 <aholler> the bone just thinks he has reached his final destination, a grave.
May 31 04:35:06 <KeatonT> lol
May 31 04:35:23 <KeatonT> I wish there was a way to monitor temp
May 31 04:35:39 <KeatonT> btw just placed the cover back on after opening a few more vent holes on the top of the box.
May 31 04:35:59 <fiola> KeatonT: I suggest you make a little script that gathers info, and stick it in crontab to run rather frquently.  Next time the same problem happens, you'll have something to analyse.
May 31 04:36:14 <KeatonT> but what info would I gather?
May 31 04:36:48 <Russ> isn't there a hwmon driver?
May 31 04:37:54 <fiola> Gah, just discovered that opkg doesn't have the lshw package
May 31 04:38:25 <KeatonT> I'm running debain
May 31 04:38:27 <KeatonT> debian
May 31 04:38:55 <KeatonT> installing lshw
May 31 04:38:58 <KeatonT> if that helps
May 31 04:39:46 <fiola> Well it'll tell you about all your h/w resources, which may provide ideas for things to monitor.  Of course there's everything in /sys
May 31 04:40:36 <KeatonT> :/
May 31 04:41:11 <fiola> Ummm, nothing returned by catting /sys/power/state.
May 31 04:41:43 <ds2> stick a thermistor on the ADC lines
May 31 04:42:27 * Russ *sigh*
May 31 04:42:37 <aholler> I recently heard there is already one somewhere in the hw
May 31 04:42:39 <fiola> A calibrated 1-wire or some other thermometer, thermister is too primitive :P
May 31 04:43:03 <fiola> Yeah,  I expect the SoC to have some sensors
May 31 04:44:01 <aholler> maybe you have to sign a nda or register at some ph0rum to reach that level of knowledge ;)
May 31 04:44:30 <KeatonT> lol
May 31 04:47:02 <fiola> Everything in the AM3358/9 is open except the PowerVR GPU I think.
May 31 04:47:11 <KeatonT> ?
May 31 04:48:55 <aholler> and the rom-code, and the secrets
May 31 04:49:07 <aholler> otherwise they wouldn't be secrets ;)
May 31 04:50:39 <KeatonT> :/
May 31 04:50:52 <KeatonT> well it's still running.
May 31 04:51:03 <KeatonT> hmm looks like my power might be going out soon.
May 31 04:51:10 <KeatonT> Got some thunderstorms
May 31 04:51:15 <KeatonT> and the lights are flickering
May 31 04:51:59 <KeatonT> Good thing the gateway is on battery backup. (Along with the beaglebone.
May 31 04:52:28 <fiola> Now you've done it, alien invasion.  You should NOT have pressed that red button on the BeagleBone.
May 31 04:52:59 <Russ> KeatonT, dfw area/
May 31 04:52:59 <Russ> ?
May 31 04:53:10 <KeatonT> Yep
May 31 04:53:21 <Russ> same here
May 31 04:53:34 <KeatonT> Yea North Dallas for me.
May 31 04:53:44 <KeatonT> Pretty scary thunder going on.
May 31 04:53:47 <Russ> 75+635 whatever that is considered
May 31 04:54:01 <KeatonT> Yea I have a friend who lives over their.
May 31 04:54:22 <KeatonT> I'm actually near Lake Ray Hubbard
May 31 04:54:36 <KeatonT> Never had a power outage here.
May 31 04:55:07 <Russ> nothing going on here right now, but quite the show in the distance
May 31 04:55:15 <KeatonT> Yea I think I'm at the show.
May 31 04:55:23 <KeatonT> The rain just starting coming down
May 31 04:55:44 <KeatonT> heh, so much for troubleshooting power issues with the beaglebone.
May 31 04:55:45 <KeatonT> lol
May 31 04:55:51 <KeatonT> Not really an ideal time for it.
May 31 04:56:56 <ds2> the RTL and the masks are probally still closed :D
May 31 04:57:20 <KeatonT> Sounds almost like hail hitting the house.
May 31 05:04:08 <aholler> ah, next step in the heardbeat series, the patch(es) now got folded and reached -mm ;)
May 31 05:05:03 <aholler> sheard/heart/
May 31 05:13:17 * Russ wonders what county he's in
May 31 05:14:23 <aholler> KeatonT: which kernel do you use? the heartbeat of the old kernels didn't stop on panic, so if the bone still showed a heartbeat it might have been a panic
May 31 05:14:38 <KeatonT> no heartbeat
May 31 05:15:13 <KeatonT> it acts as if it has been powered off.
May 31 05:16:39 <aholler> problem is koens kernels already have that patch ;)
May 31 05:17:46 <aholler> KeatonT: you might enable reboot on panic/watchdog. there is somewhere a switch for that
May 31 05:18:30 <aholler> I think sysctl -a should show it somewhere
May 31 05:20:08 <KeatonT> kernel.panic = 0
May 31 05:21:52 <aholler> give it a positive number (seconds) or <0 (immediately)
May 31 05:22:04 <KeatonT> You know I didn't start having this issue till the storms rolled in.
May 31 05:22:47 <aholler> if the power would have gone, the bone shouldjust have started again when the power would have been restored
May 31 05:23:00 <KeatonT> yea but I have it connected to a battery backup.
May 31 05:23:20 <aholler> and?
May 31 05:23:39 <KeatonT> I wonder if the battery backup when it switches from battery to commercial power is causing some type of weird power flux
May 31 05:24:08 <aholler> must be a bad ups when it acts like that
May 31 05:24:13 <KeatonT> Still it only seems to happen when the cover is on.
May 31 05:24:23 <KeatonT> aholler it's old
May 31 05:24:28 <KeatonT> probably time for replacement
May 31 05:24:38 <KeatonT> I should say oldish
May 31 05:24:44 <KeatonT> 5 years maybe
May 31 05:24:59 <aholler> just set panic, so at least you can be sure it wasn't a panic
May 31 05:26:13 <aholler> you could enable ramoops too
May 31 05:26:22 <KeatonT> Just reproduced the issue.
May 31 05:27:02 <KeatonT> Rapidly plugging and unplugging the A/C adapter will trigger the issue.
May 31 05:27:07 <KeatonT> So it might be my UPS
May 31 05:27:23 <KeatonT> or just a shitty a/c adapter
May 31 05:29:49 <KeatonT> Gunna swap it out with a 5V 2A
May 31 05:46:35 <KeatonT> Yep thinking it's actually a shitty A/C adapter now.
May 31 05:59:08 <Russ> KeatonT, any ping pong sized hail yet?
May 31 05:59:33 <KeatonT> Got some but tiny
May 31 06:00:46 <KeatonT> You?
May 31 06:00:54 <Russ> eh, earlier today, dime sized
May 31 07:08:51 <stuk_gen> koen: i redownload the setup-script, after some days i buil virtual/kernel and systemd-gnome-image. now i put the sdcard in the beagleboard with ulcd7 lite. i adjust the uenv.txt for the screen. BUT the touchscreen is not visibile! there isn't touchscreen0 in /dev ...
May 31 07:14:59 <stuk_gen> maybe the problem is the kernel 3.2.16
May 31 07:33:12 <aholler> stuk_gen: check /proc/config.gz
May 31 07:36:10 <stuk_gen> aholler: mmm tar -zxvf config.gz said invalid tarmagic
May 31 07:36:43 <fiola> I'm running the original 3.2.5+ that came with this BeagleBone, and I installed a driver module (lm75).  Unfortunately the install did something I didn't expect, which is to delete the driver I already had in /lib/modules/3.2.5+/kernel/drivers/hwmon/lm75.ko, install a new 3.2.14 kernel in /boot, and set the /boot/uImage symlink to it.  If I reboot in this state, it will die catastrophically on startup, right?  Because it has
May 31 07:36:43 <fiola> only a single module in /lib/modules/3.2.14 .
May 31 07:36:46 <aholler> that isn't a tar archive, use zcat or zgrep
May 31 07:37:23 <stuk_gen> aholler: a ok...what string i have to check?
May 31 07:38:06 <aholler> check the irc-logs, I once have searched it for you, I don't remember what touchscreen you are using
May 31 07:38:48 <aholler> or just invoke make menuconfig and check it yourself
May 31 07:39:38 <stuk_gen> aholler: a y mean yes but a m?
May 31 07:39:58 <fiola> "gunzip -c /proc/config.gz | grep blah" is what you want
May 31 07:40:15 <aholler> m is a module
May 31 07:40:34 <aholler> fiola: zgrep is more convinient
May 31 07:40:42 <fiola> true
May 31 07:41:03 <stuk_gen> fiola: i have do zcat confg.gz|grep TOUCH
May 31 07:41:12 <stuk_gen> but i have no idea what i have to check :)
May 31 07:42:22 <aholler> your hw, whatever that is
May 31 07:42:29 <stuk_gen> i knoe is a resistive touch
May 31 07:42:31 <stuk_gen> ulcd7 lite
May 31 07:42:55 <aholler> dmesg should show the manufacturer
May 31 07:42:56 <stuk_gen> its said a lot of the resolution and blabla but only 'with a resistive touch'
May 31 07:43:24 <stuk_gen> [    0.045288] Beagle second expansionboard: initializing touchscreen: tsc2007
May 31 07:43:47 <stuk_gen> oooh cool
May 31 07:43:57 <stuk_gen> i found CONFIG_TOUCHSCREEN_TSC2007=m
May 31 07:44:09 <stuk_gen> so its a module...so i can add this module?
May 31 07:44:55 <stuk_gen> kernel-module-tsc2007
May 31 07:45:03 <fiola> Re my long question above .... I know I can just switch the symlink in /boot from 3.2.14 back to 3.2.5+ before next reboot to avoid a major boom, but I'd still be missing the lm75.ko that opkg removed from 3.2.5+
May 31 07:47:47 <fiola> If it's already set to =m, then you should already have the module installed
May 31 07:48:33 <fiola> My A5A has it installed:
May 31 07:48:37 <fiola>     /lib/modules/3.2.5+/kernel/drivers/input/touchscreen/tsc2007.ko
May 31 07:49:41 <fiola> It even loaded on modprobe:
May 31 07:49:45 <fiola> tsc2007                 2718  0
May 31 07:51:16 <stuk_gen> fiola: i stopped to /lib/modules/3.2.16/kernel/drivers/
May 31 07:51:21 <stuk_gen> fiola:  i haven't input
May 31 07:52:25 <fiola> Ah right.  Yeah if you've done a partial upgrade of system then you'll have to install the modules you're missing
May 31 07:52:46 <fiola> My question which nobody has answered is similar. :P
May 31 07:53:04 <stuk_gen> fiola: what do you mean? i build the image whith angstrom oe systemd-gnome-iamge
May 31 07:53:06 <stuk_gen> image*
May 31 07:54:27 <fiola> Well you don't have tsc2007.ko if I understood correctly, so it's not a complete 3.2.16 system to the same extent as this 3.2.5+ one is.
May 31 07:57:28 <stuk_gen> fiola: ok :)
May 31 07:57:36 <stuk_gen> fiola: maybe angstrom do some cleaning
May 31 07:58:46 <koen> stuk_gen: is kernel-module-tsc2007 installed ?
May 31 07:59:10 <stuk_gen> koen: no in config.gz there is  CONFIG_TOUCHSCREEN_TSC2007=m
May 31 07:59:20 <stuk_gen> koen: but i don't find where is it
May 31 07:59:59 <fiola> It should be in   /lib/modules/3.2.16/kernel/drivers/input/touchscreen/tsc2007.ko
May 31 08:00:12 <stuk_gen> no there isn't /input folder
May 31 08:00:17 <koen> stuk_gen: opkg install kernel-module-tsc2007
May 31 08:00:33 <koen> you should have the module in your deploy/ipk/beagleboard directory
May 31 08:00:45 <stuk_gen> koen:  yes i know but i've a problem with dhcp i can't do opkg
May 31 08:01:08 <stuk_gen> koen: mmm...ok i'm looking the deploy
May 31 08:02:20 <fiola> koen: while you're here, is there a way to get my original /lib/modules/3.2.5+/kernel/drivers/hwmon/lm75.ko back?  Ie. a way to ask opkg to install lm75 for the 3.2.5+ kernel?
May 31 08:02:48 <stuk_gen> koen:  yes there is /deploy/ipk/beagleboard/kernel-module-tsc2007_3.2.16-r120d_beagleboard.ipk
May 31 08:03:54 <stuk_gen> now i will try to copy to beagleboard and use modprobe
May 31 08:07:31 <stuk_gen> yes now works :) perfect thanks to all
May 31 08:08:02 <kitsufan> Hi all
May 31 08:08:18 <fiola> Well done stuk :-)
May 31 08:08:40 <av500> so not stuck any more...
May 31 08:08:44 <fiola> Heh
May 31 08:08:49 <kitsufan> Is there a way to include a mw command in the uEnv fille?
May 31 08:09:24 <av500> I think so
May 31 08:10:16 <fiola> Well I am stuck.  Because I can't find a way to ask opkg to re-install the old lm75 module for the 3.2.5+ kernel that it zapped.  opkg --help isn't helping :P
May 31 08:11:33 <kitsufan> Please help
May 31 08:12:03 <LetoThe2nd> kitsufan: you can probably include it scriptwise in the bootcmd.
May 31 08:12:38 <kitsufan> I dont have bootcmd
May 31 08:13:05 <LetoThe2nd> kitsufan: ... or in whatever scripted command runs upon leaving u-boot.
May 31 08:14:17 <aholler> uenvcmd=foo1;foo2;foo3
May 31 08:14:27 <kitsufan> When i paste it into uenv it gets ignored
May 31 08:15:52 <LetoThe2nd> kitsufan: maybe it might make more sense to tell us what you *actually* are trying to archieve.
May 31 08:16:37 <aholler> kitsufan: put "uenvcmd=help" into your uEnv.txt
May 31 08:17:05 <kitsufan> I want to use a memmory write to make my c4 run at 650 mhz but the board has a broken serial
May 31 08:17:22 <aholler> or use "uenvcmd=help;printev"
May 31 08:17:35 <aholler> s/printev/printenv/
May 31 08:18:24 <aholler> broken serial?
May 31 08:19:11 <kitsufan> I cant use the serial console
May 31 08:21:34 <av500> since when does one need mw to run at 650mhz?
May 31 08:21:45 <av500> isnt there a kernel command for that?
May 31 08:22:06 <kitsufan> Setting mpu rate dosent work for me
May 31 08:22:37 <kitsufan> But if you have eny more pointers i would be glad to listen
May 31 08:23:02 <kitsufan> The uenvcmd worked thank you werry much
May 31 08:26:19 <av500> kitsufan: so you have 50mhz more now?
May 31 08:26:45 <LetoThe2nd> *party*
May 31 08:26:52 <aholler> the c4 can run with 720
May 31 08:28:00 <LetoThe2nd> 70MHz more *party* then.
May 31 08:30:25 <aholler> eh, hw-devs, just found in a data sheet: "slave address: read A3h and write A2h"
May 31 08:31:05 <LetoThe2nd> aholler: sounds like modbus.
May 31 08:32:07 <aholler> should be i²c, but I assume they mean addr 0x51 and just don't have a plan
May 31 08:32:52 <LetoThe2nd> plans are usually completely overrated.
May 31 08:41:34 <kitsufan> My c4 bugs the usb at 680
May 31 08:43:16 <aholler> who needs usb? ;)
May 31 08:43:29 <kitsufan> My keyboard :p
May 31 08:43:46 <LetoThe2nd> who needs keyboards since there is siri?
May 31 08:44:23 <aholler> go away with that apple sniffing sw
May 31 08:46:00 <aholler> their next license will require to transfer photos/videos from the camera to apple
May 31 08:46:14 <kitsufan> I use andro :p
May 31 08:46:54 <aholler> because they need to know the environment for the ai for siri
May 31 08:48:26 <LetoThe2nd> http://thedoghousediaries.com/3175
May 31 08:49:09 <stuk_gen> i think i need a new superpowercomputer :) its incredible i spent 3 days to build angstrom image...now i add a package to my recipe and its build all from the beginning O_O another 3 days ç_ç
May 31 08:51:10 <aholler> just go native, doesn't need longer ;)
May 31 08:51:24 <av500> stuk_gen: thats the rm_work thing
May 31 08:51:37 <av500> it removes all sources after buidling
May 31 08:52:01 <mru> but not always immediately after building each package
May 31 08:52:07 <mru> so it can still accumulate quite a lot
May 31 08:52:30 <stuk_gen> av500: ok but why ? isn't a cache?
May 31 08:56:58 <av500> stuk_gen: no idea, I dont use oe
May 31 08:57:09 <stuk_gen> av500: what do you use?
May 31 09:14:00 <guyzmo> hi
May 31 09:15:03 <fiola> I've looked in all 3 mirrors of the BeagleBone demo archives, and none of them have modules for kernel 3.2.5+.  The specific version of  kernel-module-lm75 that opkg deleted was 3.2-r3a+gitr09e9651bcf2ee8d86685f2a8075bc6557b1d3b91 , and lm75_3.2-r3a doesn't even get a single hit in google.  I'm stuck
May 31 09:16:04 <fiola> Surely the SD card image that comes with current A5A beaglebones is archived somewhere ...
May 31 09:21:51 <av500> http://circuitco.com/support/index.php?title=BeagleBone#Image_Files
May 31 09:22:24 <fiola> Thanks, looking
May 31 09:27:46 <fiola> av500: More generally, how does one specify the version number you want on an opkg install?  I tried tacking on the version string in various ways but not successful, and --help shows no examples of it.
May 31 09:28:56 <av500> fiola: I have no idea
May 31 09:29:51 <fiola> Nor does google :P
May 31 10:03:20 <koen> fiola: update to 3.2.18, then you'll get kernel-module-lm75 again
May 31 10:04:01 <koen> fiola: the opkg feeds are internally consistent, so when there's 3.2.18 for bone, older versions are deleted
May 31 10:04:10 <koen> so opkg gives you matching kernel and module
May 31 10:04:11 <koen> s
May 31 10:04:32 <stuk_gen> koen: is there a way to know when new kernel is release in feed?
May 31 10:05:16 <stuk_gen> koen: today there is 3.2.18 two days ago 3.2.16
May 31 10:05:59 <fiola> koen: When you say "update", you mean the opkg command?
May 31 10:08:32 <stuk_gen> fiola: are you using oe bitbake? or an angstrom image?
May 31 10:08:54 <fiola> koen: The situation I have currently is that I'm running the original 3.2.5+, but opkg deleted its corresponding lm75 module.  I don't want to upgrade the system to 3.2.14, as I want to keep this pretty original as rescue.
May 31 10:09:21 <fiola> stuk_gen: this is the original SD card that I got with the new beaglebone last week
May 31 10:09:40 <stuk_gen> fiola: ok...so copy to another sd card :P
May 31 10:09:46 <fiola> So still at 3.2.5+, and that's how I want to keep it until I buy some more sd cards
May 31 10:10:02 <fiola> stuk: Don't have
May 31 10:10:45 <stuk_gen> fiola: mmm so you haven't more way to operate...mmm maybe try to understand how can you run sysroot on eth without using sd but another pc
May 31 10:11:38 * fiola scratches head
May 31 10:12:16 <fiola> I don't think I want anything complicated.  I'm running a production version, and simply want a file that opkg deleted incorrectly :P
May 31 10:12:38 <av500> well, get it from the original image
May 31 10:12:57 <fiola> Yeah, doing that, thanks av500, it's currently unpacking
May 31 10:14:57 <fiola> But if opkg knows how to install something for a kernel that isn't running and break it in the process, then I'd expect it to be able to get the old release back too.  koen said it's consistent for a given release.  Well, it *was* consistent, until opkg broke that consistency ... apparently one-way only, no return :-(
May 31 10:16:20 <fiola> It knows what it did.  Here's the log -->
May 31 10:16:24 <fiola>     Upgrading kernel-module-lm75 on root from 3.2-r3a+gitr09e9651bcf2ee8d86685f2a8075bc6557b1d3b91 to 3.2-r10d+gitre8004dad869568692ca2a45c04e8464ce48cc4d7...
May 31 10:17:04 <fiola> So if it knows how to do it in that direction to *break* consistency, it's odd that it can't do it in the reverse direction to *restore* consistency.
May 31 10:17:37 <fiola> opkg is definitely not a rescue tool :-)
May 31 10:19:36 <cplexx> question: has anyone has any succes using bluetooth and the beaglebone? I am thinking of using a bbone for my RC project, but I need to be sure that  BT can work.
May 31 10:19:46 <aholler> whats so difficult with use tar to make a backup of the sd on the pc?
May 31 10:19:58 <fiola> And opkg really wants its ass kicking for this: --->
May 31 10:20:00 <fiola>     Removing obsolete file /lib/modules/3.2.5+/kernel/drivers/hwmon/lm75.ko.
May 31 10:20:14 <fiola> "obsolete"????   It's the fricking running kernel!
May 31 10:20:49 <aholler> cplexx: if usb works, bluetooth works too
May 31 10:21:34 <aholler> I use a bt-dongle on my beagleboard c4. just don't use an ath3k
May 31 10:21:35 <av500> why is the unix "find" named "find" and not "search"?
May 31 10:22:54 <cplexx> aholler: when I search th groups I only see people complaining it not working. So, I'm curious which dongles have been used succesfully
May 31 10:23:33 <aholler> av500: because a search is always successfull. even when it finds nothing te search still was a success
May 31 10:23:38 <av500> cplexx: usually people that have no issues dont post that much about it
May 31 10:23:54 <cplexx> av500: I know, I just want to be sure, before I spend my money :)
May 31 10:23:59 <av500> aholler: yeah, it always has a great time doing it
May 31 10:24:15 <aholler> cplexx: everyone is able to post whatever he wants in ph0rums
May 31 10:25:18 <cplexx> av500 & aholler, thanks, I guess I'll just buy one, and see. If not I'm sure its not intensic to the hardware.
May 31 10:26:07 <aholler> it all depends if usb works.
May 31 10:26:35 <hitlin37> av500:because of there love for four letter word(u-n-i-x=4)
May 31 10:28:58 <cplexx> aholler: the bt driver / stack should work. I'm seenig evil things on the forum about disabling DMA and such ... brrr...
May 31 10:29:32 <cplexx> if I wanted a slow device I would have bough and arduino
May 31 10:33:22 <koen> fiola: you asked opkg to update to a more recent kernel, so you get what you ask for
May 31 10:34:35 <aholler> cplexx: I don't know how reliable musb on the bone is. on my beagleboard, where I have the bt-dongle to musb too, it isn't.
May 31 10:34:49 <aholler> s/to/on/
May 31 10:36:02 <aholler> ssh is no problem, but fast transfers sometimes will kill it
May 31 10:36:58 <koen> aholler: musb is completely broken on bone
May 31 10:37:00 <fiola> koen: No I did not ask opkg to do that.  I asked it to install kernel-module-lm75
May 31 10:37:03 <cplexx> hmm.. well, I will mainly use it to create a serial connection over bluetooth... it won't be much data
May 31 10:37:10 <koen> aholler: it can't even do mass-storage
May 31 10:37:37 <koen> fiola: since it was already installed it fetched a newer version
May 31 10:37:37 <cplexx> koen> really... hmm
May 31 10:38:43 <cplexx> but it should be reliable
May 31 10:39:07 <cplexx> (as far as wireless tech is...)
May 31 10:40:24 <cplexx> so are there any alternatives to angstrom with more reliable USB/BT support
May 31 10:40:49 <fiola> koen: It deliberately broke the running kernel.  If it decided "Well you can't have that module without a complete upgrade" and installed the entire new kernel and modules, that might be excusable (if a bit heavy-handed).  But just leaving the system in a broken state by deleting something that was consistent before but now no longer it, is simply not acceptable.  There was no need to delete that old driver in the current /li
May 31 10:40:49 <fiola> b/modules/3.2.5+
May 31 10:43:22 <fiola> What opkg did is not consistent with the structure of /boot.  /boot can contain multiple kernels, and just uses a symlink to choose one.  That's why all the modules under /modules *MUST* remain present, to go with their coresponding kernels in /boot.  okpg's behaviour is a BUG.
May 31 10:44:19 <jacekowski> fiola: not only opkg does it
May 31 10:44:46 <jacekowski> iirc apt-get with DKMS did it as well
May 31 10:44:47 <fiola> And opkg knows that full well, because it doesn't try to delete the old kernel in /boot, it just deposits the new kernel in there and changes the symlink.  So why does it delete the old module?  Makes no sense
May 31 10:49:20 <tichku> hey i am beginner with beagle board .....how i can load ubuntu on beagle board..?
May 31 10:52:20 <av500> http://www.google.com/search?q=beagleboard+ubuntu
May 31 10:52:23 <av500> tried that?
May 31 10:58:15 <cplexx> so, anyone did any RC kind of stuff with the beagle bone? Alternative would be to use wifi, but than bbone should be an access point :S not sure about that
May 31 11:03:53 <av500> wifi could have similar usb issues
May 31 11:03:59 <av500> if connected over usb...
May 31 11:04:35 <fiola> av500: I did what you suggested and grabbed the A5 image off circuitco ... except that their site is screwed.  The so-called "A5" image is not the one that came with my A5 BeagleBone last week, because mine is 3.2.5+ while their image is 3.2.16
May 31 11:05:19 <cplexx> av500: Hmm.. it doesn't sound very encouraging. so, why is USB so unreliable? I mean I would expect that to work properly this day and age
May 31 11:05:30 <av500> [12:36:59] <koen> aholler: musb is completely broken on bone
May 31 11:05:41 <av500> cplexx: broken hardware is broken
May 31 11:05:51 <av500> even this day and age...
May 31 11:06:19 <cplexx> so, is the hardware design faulty or is this a software issue?
May 31 11:06:47 <av500> seems hw
May 31 11:07:03 <cplexx> so, I should wait for the next revision and hope for the best?
May 31 11:07:06 <av500> or TI unwilling to provide a SW fix
May 31 11:07:30 <cplexx> And there are no such issues with the beagleboard?
May 31 11:09:03 <av500> different issues :)
May 31 11:09:11 <av500> but the EHCI USB seems to work more or less
May 31 11:09:59 <cplexx> so, what are my alternatives for a somewhat powerful embedded, light weight, low-power device which is easy to develop?...
May 31 11:10:23 <cplexx> (with atleast 4 PWM's :) )
May 31 11:10:46 <av500> bone seems fine for that
May 31 11:10:59 <cplexx> yeah, but I also need RC... :S
May 31 11:11:17 <av500> and RC can only be BT?
May 31 11:11:24 <fiola> Not many options if you want Linux.  Tons of options if just an embedded ARM is OK -- zillions of Cortex-M3 or M4's
May 31 11:11:44 <cplexx> well, it would be nice if I could control it with my android phone
May 31 11:12:05 <cplexx> I'm building a quad-copter
May 31 11:12:26 <fiola> Better stick with Linux then if you want decent networking
May 31 11:12:27 <cplexx> but I want to make it somewhat more powerful so I can image processing and such in the future
May 31 11:12:28 <av500> http://www.sparkfun.com/products/10253
May 31 11:12:33 <av500> connects over uart
May 31 11:12:40 <av500> not usb
May 31 11:12:57 <av500> http://www.sparkfun.com/products/10393
May 31 11:13:57 <cplexx> av500 -> thanks. that would be perfect if it was a class 1 bluetooth device
May 31 11:14:53 <cplexx> I've considered these chips, but I guess I should consider them somewhat more :)
May 31 11:16:09 <av500> http://www.sparkfun.com/products/10559
May 31 11:16:14 <av500> class1, no?
May 31 11:17:36 <av500> the actual module does usb, spi, uart...
May 31 11:18:20 <cplexx> yes, it seems so, I've seen these modules, but they are so darn expensive.. the silly USB thing costs $10 ... its almost as expensive as the whole beaglebone
May 31 11:20:09 <av500> google gives me $25 for the module
May 31 11:20:19 <av500> and to solder a few wires you should manage
May 31 11:21:19 * ogra_ also wants google to give him money for modules !
May 31 11:21:38 <cplexx> I guess $25 is managable, but it would be nicer if I could just plugin in the $10 thing in the USB port. Thanks av500 for your help. I will consider this.
May 31 11:24:13 <cplexx> bye everyone
May 31 11:32:54 <aholler> he should have just buy a beagleboard classic or xm. ehci is ok with kernels <3.3
May 31 11:33:34 <aholler> above it got infected by musb
May 31 11:46:15 <ynezz> omap is borken
May 31 11:46:27 <av500> outright fucked
May 31 12:00:20 <mdp> perhaps you're just being encouraged to use the expansion headers instead of usb
May 31 12:00:37 <mdp> using the strong hint method of enouragement
May 31 12:17:10 <aholler> yeah, just use the processor and forget the description SoC ;)
May 31 12:18:35 <koen> hook up bt to a uart
May 31 12:18:37 <koen> wifi to spi
May 31 12:18:41 <koen> no need for (m)usb
May 31 12:18:54 <koen> and uvc webcams work
May 31 12:19:00 <koen> pretty much the only thing that works
May 31 12:20:26 <aholler> bt on an uart isn't that good. at least not if you want to use audio, network and hid too ;)
May 31 12:25:15 <fiola> It's a bit late now, but really USB should never have been invented.  Ethernet already existed at the time and could easily have been made autoconfigurating like IPv6 is today so that it "just works" on plugging.
May 31 12:25:55 <agmlego> You want to use Ethernet in place of USB?
May 31 12:26:23 <aholler> usb is great and IPv6 has lot problems too
May 31 12:26:41 <fiola> It would have made a far superior local connection, just like it is superior for remote.
May 31 12:27:16 <fiola> And having a single link level fabric would have been amazing
May 31 12:27:46 <agmlego> fiola: But no power, so everything would have to have separate poer supplies. And things like flash drives would be a lot more expensive and power-hungry, since Ethernet control hardware is more complex than USB.
May 31 12:28:13 <agmlego> Kind of like the reason Firewire did not fare as well against USB as everyone predicted.
May 31 12:28:15 <fiola> agmlego: There is power over Ethernet too
May 31 12:28:22 <agmlego> The controller issue, not the power issue.
May 31 12:28:29 <agmlego> fiola: Not at the time, there was not.
May 31 12:28:36 <agmlego> And PoE sucks.
May 31 12:28:38 <aholler> and the MACs need an update too, they are too small, just as IPv4 was
May 31 12:28:58 <fiola> agmlego: The effort spent on USB could have added the missing bits to Ethernet back then.
May 31 12:29:42 <agmlego> fiola: But at the fundamental level, Ethernet is not appropriate for the tasks USB is good for any more than USB is good for the tasks Ethernet is for.
May 31 12:29:51 <agmlego> They are different types of connections for a reason.
May 31 12:30:13 <fiola> Anyway, it's academic now.
May 31 12:30:53 <fiola> agmlego: Nah, they're different kinds of connection because people make money from incompatibility.
May 31 12:31:36 <aholler> firewire is insecure, and intel does the same error again with that new foo.
May 31 12:31:37 <agmlego> fiola: Never attribute to malice or deliberate intent what can be assigned to stupidity or reasons you disagree with.
May 31 12:31:55 <agmlego> aholler: How is Firewire any more inecure than USB?
May 31 12:32:00 <agmlego> Or Ethernet, for that matter.
May 31 12:32:06 * mdp suggests using rapidio
May 31 12:32:29 <aholler> because with firewire you have unlimited access to the memory
May 31 12:32:44 <aholler> and usb is master<->slave
May 31 12:32:53 <aholler> so the master has control
May 31 12:32:54 <agmlego> aholler: Oh, so you dislike DMA.
May 31 12:32:55 <LetoThe2nd> RFC 1149 > $RESTOFTHEWORLD.
May 31 12:33:14 <aholler> agmlego: the problem isn't the dma, usb uses dma too
May 31 12:33:15 <fiola> agmlego:  Don't attribute insecurity to connection fabrics when it is more directly achieved through incompetence.
May 31 12:33:18 <agmlego> It is not exactly unlimited.
May 31 12:33:19 * dm8tbr played once with this neat toolkit for firewire
May 31 12:33:36 <dm8tbr> it would disable the logon password for windows and linux by patching it in memory
May 31 12:34:52 <fiola> Security is not a problem of links, it's a problem of people.
May 31 12:35:10 <aholler> ?
May 31 12:35:23 <LetoThe2nd> aholler: i'd suggest just plonking.
May 31 12:37:51 <jannau> firewire is an external interface to the system memory, i.e. inherently insecure
May 31 12:38:32 <fiola> Copper is an external interface to the whole computer, ie. inherently insecure :P
May 31 12:51:32 <jacekowski> dm8tbr: you can do same thing over usb
May 31 12:52:00 <jacekowski> dm8tbr: even though controllers shouldn't do it, most usb controllers will allow unrestricted DMA
May 31 12:54:25 <dm8tbr> jacekowski: cool, is there a g_h4x0r.ko for this yet? :)
May 31 12:55:00 <jacekowski> nope
May 31 12:55:35 <jacekowski> it's much more complex with usb as it relies on bugs in implementations rather than design
May 31 12:55:52 <av500> aholler: connect a 5$ mouse over ethernet?
May 31 12:56:03 <jacekowski> but firewire has an advantage of being a lot faster
May 31 12:56:07 <jacekowski> av500: why not
May 31 12:56:12 <jacekowski> av500: wifi
May 31 12:58:25 <av500> jacekowski: we arew talking about the time when usb was invented
May 31 12:58:29 <av500> no today
May 31 12:58:31 <av500> not
May 31 12:58:47 <av500> so back then, wifi was a good choice for a mouse?
May 31 12:58:56 <jacekowski> ps2 was
May 31 12:59:46 <jacekowski> usb is maybe nice for low bandwidth stuff
May 31 12:59:55 <jacekowski> but it's painfull for anything faster
May 31 13:00:00 <jacekowski> and it's not that fast anyways
May 31 13:01:20 <fiola> Whole lot should have been over Ethernet.  Mice were still mostly serial back when twisted pair Ethernet finally replaced both types of coax Ethernet, could easily have been done at Ethernet MAC-level.
May 31 13:01:53 <av500> but ethernet has no dedicated master
May 31 13:02:02 <av500> so devices could talk with each other besides a PC
May 31 13:02:07 <av500> Intel could not allow that
May 31 13:02:08 <jacekowski> hdmi should have been done over ethernet
May 31 13:02:15 <fiola> Which is perfect
May 31 13:02:27 <jacekowski> rather than decompressing video on sending device - send compressed stream to display
May 31 13:02:54 <fiola> Oh I know Intel could not allow it.  They make money every thing they create a new standard and make all previous systems obsolete
May 31 13:02:57 <av500> jacekowski: it would have taken them 10ys to decide on a compression format :)
May 31 13:03:10 <fiola> every time*
May 31 13:03:20 <jacekowski> av500: well, same format they use for blurays
May 31 13:03:25 <jacekowski> h26
May 31 13:03:27 <jacekowski> h264
May 31 13:03:50 <av500> jacekowski: H264? but then the free/open/patent haters people could not use it....
May 31 13:04:05 <jacekowski> then let them die
May 31 13:04:38 <av500> at least let them go blind watching VGA flickering....
May 31 13:04:46 <jacekowski> anyways
May 31 13:04:52 <jacekowski> i've got serious problem here
May 31 13:04:56 <jacekowski> pizza hut or dominos
May 31 13:05:29 <fiola> Money is made from increasing incompatibility instead of compatibility, just as money is made from restricting access instead of enabling it.  It's how this stupid world works.
May 31 13:06:21 <av500> yes, money is being made by selling stuff
May 31 13:06:22 <fiola> jacekowski: With sufficient jalopenos on top, you can't tell the difference anyway :P
May 31 13:06:33 <av500> or by armed robbery
May 31 13:06:56 <jacekowski> i was thinking about making a career in organised crime
May 31 13:07:02 <fiola> Aye, these days principlally by robbery, performed by lawyers
May 31 13:07:11 <av500> jacekowski: work at a bank?
May 31 13:07:26 <jacekowski> no
May 31 13:07:46 <fiola> Biggest organized crime offfice, after the Fed, is the USPTO
May 31 13:07:59 <jacekowski> just some extortions and stuff
May 31 13:08:24 <jacekowski> but i suppose HMRC wouldn't like the competition
May 31 13:08:43 <av500> fiola: not really, because they are not really making money from that
May 31 13:09:10 <jacekowski> av500: that's very bad business model then
May 31 13:11:25 <fiola> Of course they're making money from it.  The USPTO make money from every patent filing, which is why they accept rubbish with open arms.  And the megacorps make money from it by keeping the upstarts off their turf.  And the patent trolls make money by leeching off everyone.
May 31 13:12:26 <fiola> And the politicians make money off it because the megacorps pay them to keep the status quo.
May 31 13:12:51 <fiola> And the lawyers make money off it from all the lititation.
May 31 13:13:07 <fiola> So everybody wins, except those who are trying to make something new.
May 31 13:13:09 <av500> it's a mega conspiracy
May 31 13:13:26 <av500> fiola: just buy apple, they make nice *new* things....
May 31 13:13:50 <fiola> They don't make anything, China makes it all
May 31 13:14:13 <av500> sudo make me an ipad
May 31 13:15:03 <jannau> make: *** No rule to make target `me'.  Stop.
May 31 13:16:24 <ogra_> missing the Makefile, eh ?
May 31 13:16:54 <av500> ogra_: xcode uses makefile?
May 31 13:18:08 <ogra_> xcode builds HW ?
May 31 13:20:08 * aholler fears the time when people have to configure routes to use their mouse
May 31 13:20:39 <av500> the mouse in my house know more routes than I do....
May 31 13:20:42 <av500> +s
May 31 13:21:06 <aholler> they even know your house much better than you ;)
May 31 13:21:27 <av500> that's autorouting at it's best...
May 31 15:39:31 <recroot> hey, guys.. sorry to bother you with a silly question..but how o you get the mic working on the beagleboard?
May 31 15:39:49 <av500> it has no mic
May 31 15:40:17 <Russ> did you try talking louder?
May 31 15:40:24 <recroot> i connected my headphones to it..
May 31 15:40:47 <Russ> so you just want audio *out*?
May 31 15:41:28 <recroot> i want to record a short clip and play it back
May 31 15:41:32 <recroot> but all i get is static
May 31 15:41:45 <recroot> as response
May 31 15:42:02 <Russ> when you connect your headphones...
May 31 15:42:48 <recroot> yeah?
May 31 15:44:17 <Russ> these things
May 31 15:44:18 <Russ> https://www.google.com/search?q=headphones&tbm=isch
May 31 15:46:12 <lpi> Russ: You can use headphones as mic
May 31 15:46:16 <lpi> true story
May 31 15:46:42 <recroot> well, thanks for trying to help..in line headphones worked..
May 31 15:46:54 <Russ> I know, you can also use a solar panel as an led
May 31 15:47:09 <Russ> doesn't mean it will be "ideal"
May 31 15:47:28 <recroot> showing me the google result did clear out what an headphone is..
May 31 15:48:06 <Russ> :)
May 31 15:52:41 <aholler> it has a pmic
May 31 15:55:00 <thealphanerd> anyone have experience with gps?
May 31 15:55:19 <thealphanerd> would you rather directly connect via uart or go with a module that has usb integrated?
May 31 15:55:31 <av500> what you prefer
May 31 15:55:43 <aholler> bt ;)
May 31 15:55:54 <av500> aholler: ethernet!
May 31 15:56:29 <thealphanerd> av500: lol… I'll rephrase… is it non trivial to use uart rather than usb… or is it going to be susbstantially more complex.  Never worked with gps before
May 31 15:56:48 <av500> its ttyUSBx vs ttyOx
May 31 15:56:57 <av500> assuming you are able to plug in the usb plug
May 31 15:57:43 <thealphanerd> I guess I mean more with having the system recognize it…
May 31 15:57:44 <aholler> or rfcomm0 ;)
May 31 15:58:00 <Russ> gspd is pretty smart
May 31 15:58:03 <aholler> thealphanerd: it's an serial, regardless how you connect it
May 31 15:58:27 <thealphanerd> good to know… just connecting using bare wires vs a usb module
May 31 15:59:49 <aholler> bare wires have their pitfalls
May 31 16:01:05 <thealphanerd> not as encapsulated for sure
May 31 16:02:37 <av500> aholler: bare wire across the street can be a pitfall
May 31 16:03:32 <aholler> hmm, that reminds me at the cables acrosse the streets I've recently have seen when I was in rumania.
May 31 16:04:04 <av500> or the US
May 31 16:04:42 <aholler> maybe they have exported them ;)
May 31 16:06:26 <thealphanerd> av500: it is a valid thing to consider… this is for a wearable technology… so have something modular / easy to replace / encapsulated might be a good idea
May 31 16:07:54 <av500> well, soldering a 3 pin plug is not magic
May 31 16:07:58 <av500> VCC, GND, TX
May 31 16:08:07 <av500> RX if you want to command the GPS
May 31 16:08:10 <av500> or need to...
May 31 16:08:13 <thealphanerd> probably use socket anyways
May 31 16:09:49 <aholler> just use bt and the phone
May 31 16:11:32 <thealphanerd> not really a solution for this project unfortunately
May 31 16:13:02 <av500> use ... a map
May 31 16:13:13 <av500> or just ask people
May 31 16:13:48 <ogra_> uuuh, scary
May 31 16:14:07 <ogra_> that involves VoA protocols
May 31 18:27:22 <gerardo1> hi
May 31 18:28:28 <ds2> ;2A
May 31 18:29:02 <gerardo1> hi anyone has been able to add the camera driver li-5m03 to begleboard, I need help}
May 31 18:34:01 <gerardo1> no
May 31 18:34:02 <gerardo1> ?
May 31 18:38:03 <spacecolonyone> anyone in here familiar with writing to the framebuffer in step with vsync from userland?
May 31 18:41:37 <spacecolonyone> I'm looking at the waitforvsync and waitforgo ioctl atm, but my understanding is if I'm trying to write a simple userland app I should (must?) avoid using ioctls.
May 31 18:46:51 <aholler> why that?
May 31 18:48:42 <spacecolonyone> aholler: why which part?
May 31 18:48:56 <spacecolonyone> gerardo1: I was just looking into that earlier today
May 31 18:49:21 <aholler> why you need to avoid ioctls
May 31 18:49:31 <spacecolonyone> my impression is that it is already there, depending on which kernel you are using and from where
May 31 18:49:47 <aholler> true, they aren't very portable, but some stuff never is
May 31 18:49:55 <spacecolonyone> aholler: aren't they typically kernel space?
May 31 18:50:00 <aholler> no
May 31 18:52:18 <spacecolonyone> ok, well, I was doing more digging and I think I amy have been making it more complex than needed. I'm leaning toward a simple OpenGL app (a basic bbxm image with Enlightenment shoudl handle that right?)  that just uses some of the opengl calls for constraining when buffers are swapped.
May 31 19:26:40 <aholler> heartbeat, the final: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commit;h=49dca5aebfdeadd4bf27b6cb4c60392147dc35a4 ;)
May 31 19:36:08 <aholler> 5 weeks from commit to inclusion, not that bad.
May 31 20:29:52 <lautriv> hello, need some ideas to track down an error, startup of system hangs at "g_ether gadget: full speed config #1:  CDC Ethernet (EEM)" (cdc built into kernel, no module) i see the device on the host but can't connect probably because the userlabd did not assign any IP. beside the hang, any suggestions to track this down ?
May 31 20:30:19 <lautriv> s/labd/land/
May 31 21:23:19 <_av500_> spacecolonyone: EGLSwapBuffers or so
May 31 21:23:29 <_av500_> that should take care of vsync
May 31 22:01:01 <thurbad> does Angstrom support syncfs in any package, or is there another way to sync only specific partitions?
May 31 22:02:45 <aholler> just write 1 oneliner
May 31 22:06:28 <zhivko> Hi everybody- is it possible to uselibrxtx-java library to communicate to i2c? Can somebody show me some sample of this? I don't get any port enumerated on beaglebone with CommPortIdentifier.getPortIdentifiers()...
May 31 22:07:04 <zhivko> Should I connect i2c device before running sample or it shoul enumerate port in anycase ?
May 31 22:11:06 <zhivko> hello anybody out here ?
May 31 22:13:56 <aholler> you know that i2c != rs232?
May 31 22:27:52 <zhivko> aholer - yes I do
May 31 22:28:43 <zhivko> aholler: but I saw that java classes from librxtx support i2c from this class: CommPortIdentifier.PORT_I2C
May 31 22:29:24 <zhivko> aholler: So I thought it should support i2c - and I could use /dev/i2c-3 as ordinary tty com port...
May 31 22:30:24 <zhivko> aholler: I also made C++ sample to communicate to /dev/i2c-3 but I would communicate to i2c with java - plan to use it within tomcat and open-jdk
May 31 22:32:06 <zhivko> aholler: ? still following ?
May 31 22:33:00 <aholler> yes, so just read the source how rxtx tries to enumerate i2c-buses. maybe the names in /dev changed
May 31 22:33:10 <_av500_> 2012 and java cannot do rs232, was it not meant to run everywhere? on toasters even
May 31 22:33:53 <zhivko> 2012?
May 31 22:33:54 <aholler> java is patented and dead ;)
May 31 22:34:04 <zhivko> openjdk patented ?
May 31 22:34:11 <_av500_> zhivko: in my timezone its 2012
May 31 22:34:44 <_av500_> aholler: its not much patented, oracle lost that one
May 31 22:35:02 <zhivko> so this librxtx and i2c is dead end? Has anybody tried it ?
May 31 22:35:13 <_av500_> zhivko: no
May 31 22:35:17 <_av500_> its widely used
May 31 22:35:28 <_av500_> guessing by the amount of cursing about it
May 31 22:35:29 <aholler> but they have a lot more patents and not everyone is as big as google and could fight against oracle
May 31 22:35:43 <_av500_> who else still cares for java?
May 31 22:35:44 <thurbad> oracle lost the idea that 9 lines of code = millions in damages
May 31 22:36:04 <fiola> Oracle holds patents which they claim cover Java.  The fact that the jury rejected patents '104 and '520 in the Google case doesn't stop Oracle from making the same or other patent claims against othe Java users.
May 31 22:36:08 <zhivko> can we stick to topic?
May 31 22:36:19 <_av500_> its friday
May 31 22:36:27 <zhivko> so java and i2c - no go? or it can be done ?
May 31 22:36:33 <_av500_> it can be donbe
May 31 22:36:37 <_av500_> -b
May 31 22:36:45 <zhivko> av500: yes friday is good news
May 31 22:36:56 <aholler> trollday
May 31 22:37:08 <aholler> so be prepared ;)
May 31 22:37:11 <thurbad> every day is a trollday :P
May 31 22:37:13 <zhivko> av500 - can you please point me to some sample out there ?
May 31 22:37:24 <_av500_> zhivko: so use rxtx
May 31 22:37:28 <_av500_> i have no idea
May 31 22:37:31 <_av500_> i dont use java
May 31 22:37:34 <kalem> hi all
May 31 22:37:41 <_av500_> but i am sure google will find samples
May 31 22:37:43 <kalem> just one
May 31 22:37:52 <_av500_> since its THE java lib for serial
May 31 22:38:30 <_av500_> kalem: ask
May 31 22:38:54 <zhivko> So problem is that Enumeration<CommPortIdentifier> ports = CommPortIdentifier.getPortIdentifiers() gets me no port - but it should give /dev/i2c-1 and /dev/i2c-3 isn't it right ?
May 31 22:39:02 <kalem> waiting for DM365 leopardboard
May 31 22:39:33 <kalem> I need gstreamer support and TI integration, like cmem and dsplink
May 31 22:39:35 <_av500_> zhivko: probably
May 31 22:39:40 <_av500_> add some printfs and debug it
May 31 22:40:02 <_av500_> kalem: TI offers all that
May 31 22:40:05 <zhivko> you mean in java code or to librxtx-java ?
May 31 22:40:13 <_av500_> as do 3rd parties like ridgerun
May 31 22:40:23 <_av500_> zhivko: in rxtx i guess
May 31 22:40:47 <kalem> _av500_: so ridgerun and DVSDK 3 or 4 is the answer
May 31 22:41:00 <kalem> _av500_: and what about arago ?
May 31 22:41:06 <mranostay> trollday!
May 31 22:41:10 <_av500_> kalem: this is the beagle channel
May 31 22:41:31 <kalem> _av500_: yes, but beagle without OS is nothing :-(
May 31 22:41:42 <_av500_> kalem: the leopard is not a beagle
May 31 22:41:46 <_av500_> cat vs dog thing
May 31 22:41:50 <kalem> ah
May 31 22:41:54 <zhivko> so I need source and cross compile and everything? could I use librxtx-java-dbg ?
May 31 22:41:58 * _av500_ also never got a free leopardboard from TI
May 31 22:42:04 <aholler> all animals (or borgs)
May 31 22:42:09 <_av500_> zhivko: I have no idea
May 31 22:42:44 <_av500_> mdp: customer ^^^^
May 31 22:42:50 <aholler> hmm, I would like one of the robots and/or clocks to play with ;)
May 31 22:43:05 <_av500_> robot clocks?
May 31 22:43:14 <zhivko> _Av500_: problem is that also serial ports are no enumerated - and there are no such examples with librxtx and beglebone :(
May 31 22:43:38 <_av500_> zhivko: yes
May 31 22:43:53 <_av500_> but dont expect an example for rxtx and EVERY hardware out there
May 31 22:44:04 <_av500_> go by analogy
May 31 22:44:11 <_av500_> find similar cases
May 31 22:44:15 <_av500_> be creative
May 31 22:44:36 <aholler> zhivko: linux doesn't offer something to enumerate those stuff. so check the source of rxtx how it does that
May 31 22:45:54 <_av500_> yep
May 31 22:45:57 <aholler> I would think it searches through /dev /proc or /sys and some names have changed
May 31 22:45:58 <zhivko> av500 aholler: thanks guys ! will try that ughh....
May 31 22:46:23 <zhivko> could I compile this librxtx.so on beaglebone ?
May 31 22:46:27 <_av500_> yes, some source is ughhly
May 31 22:46:56 <_av500_> zhivko: no idea, try it
May 31 22:47:57 <zhivko> av500: oh yes I tried to install older version of librxtx - how can I find what versions are there from repository ?
May 31 22:48:19 <zhivko> av500: with apt-get install package=version
May 31 22:48:37 <denix> kalem: arago does not support leopard
May 31 22:49:02 <denix> kalem: it used to support dm3xx, but that support is ancient now
May 31 22:49:04 <_av500_> denix: ah yes, you lurk here too
May 31 22:49:05 <kalem> denix: ehe, I supposed it
May 31 22:49:28 <denix> _av500_: yes, I'm everywhere! :-P
May 31 22:49:30 <kalem> denix: so the only way is ridgerun ?
May 31 22:49:34 <_av500_> denix: can has leopardboard?
May 31 22:49:39 <kalem> _av500_: http://beagleboard.org/leopard
May 31 22:49:52 <denix> kalem: gst guys that provided support for that are long since gone - try #gst_ti
May 31 22:50:01 <zhivko> maybe I need to use something like: System.setProperty("gnu.io.rxtx.SerialPorts", "/dev/ttyACM1")
May 31 22:50:58 <_av500_> denix: when I left gst_ti it was all tumbleweed
May 31 22:51:03 <denix> _av500_: I have one somewhere in my drawer - was never able to do anything useful with it...
May 31 22:51:36 <_av500_> zhivko: maybe, find out
May 31 22:51:56 <denix> _av500_: I know. I try to /part #gst_ti, but my znc keeps opening it up... :)
May 31 22:52:54 <_av500_> denix: I could have earned good money there
May 31 22:53:07 <denix> how so?
May 31 22:53:18 <_av500_> well, by replying to people
May 31 22:53:24 <denix> LOL
May 31 22:53:32 <_av500_> most of the time I had a pm asking me to help them for $$$
May 31 22:53:46 <denix> sweet!
May 31 22:53:59 <zhivko> How can I get soirces for this librxtx package? Anybody knows that ?
May 31 22:54:13 <_av500_> some peopleI helped for free: "remove ffmpegcolorspace, add a few queue"
May 31 22:54:14 <denix> google?
May 31 22:54:15 <zhivko> I mean sources :)
May 31 22:54:35 <_av500_> still google
May 31 22:54:57 <_av500_> google even fixes your typos...
May 31 22:55:37 <kalem> denix: sorry, really
May 31 22:55:45 <kalem> denix: but I cannot understand
May 31 22:56:08 <kalem> denix: ridgerun or DVSDK 3.0, no other way ?
May 31 22:57:54 <denix> kalem: if you need everything pre-integrated, then yes. otherwise you can try to assemble it yourself - most if not all should be open, but not maintained
May 31 22:58:38 <kalem> denix: ok, thank you very much
May 31 22:59:23 <kalem> denix: my experience is on DM6446 with a terrify MVL 5.0
May 31 22:59:33 <_av500_> 2.6.10 ftw
May 31 22:59:39 <denix> yay!
May 31 22:59:56 <_av500_> at least that stuff worked!!
May 31 23:00:00 <denix> no, 2.6.18 was much better!
May 31 23:00:18 <_av500_> pah, we never upgraded to that
May 31 23:00:26 <_av500_> too risky
May 31 23:00:30 <denix> LOL
May 31 23:00:43 <kalem> 2.6.18 on MVL 5.0
May 31 23:00:52 <KeatonT> Are you guys talking about kernels?
May 31 23:00:59 <_av500_> yep
May 31 23:01:06 <kalem> yes
May 31 23:01:07 <KeatonT> what
May 31 23:01:08 <_av500_> kalem: maybe it was mvl 4.0 then
May 31 23:01:11 <KeatonT> what's the latest on the bone?
May 31 23:01:17 <kalem> _av500_: no 5.0
May 31 23:01:32 <_av500_> we used 2.6.10 on dm6441, I am sure
May 31 23:01:54 <_av500_> but that was like 5ys ago
May 31 23:01:58 <denix> KeatonT: 3.2.18
May 31 23:02:15 <KeatonT> looks like I need to update
May 31 23:02:27 <kalem> _av500_: with DVSDK 2.0 and no compatibility with comunity kernel
May 31 23:02:40 <kalem> _av500_: 2.6.10 on MVL 4.0
May 31 23:02:47 <_av500_> see, 4.0
May 31 23:02:48 <kalem> _av500_: 2.6.18 on MVL 5.0
May 31 23:03:06 <denix> kalem: you may want to check 2.6.32 davinci kernel on arago - that's probably the most dm3xx support you can get...
May 31 23:03:13 <_av500_> kalem: when I used it, the term DVSDK was not invented yet
May 31 23:03:26 <KeatonT> what is the difference between 3.2.17 and 3.2.18?
May 31 23:03:29 <kalem> denix: I will try
May 31 23:03:36 <_av500_> KeatonT: 0.0.01
May 31 23:03:47 <denix> :)
May 31 23:03:57 <KeatonT> I lol
May 31 23:04:00 <KeatonT> 'ed for real.
May 31 23:04:04 <KeatonT> but really.
May 31 23:04:14 <aholler> ask lkml
May 31 23:04:14 <kalem> denix: I tried 2.6.32 davinci on dm6446, and I got a lot of drivers problems
May 31 23:04:19 <aholler> or git
May 31 23:04:32 <aholler> or gregkh
May 31 23:04:34 <_av500_> kalem: I canimagine
May 31 23:04:38 <_av500_> kalem: I can imagine
May 31 23:05:14 <kalem> _av500_: I hope to not repeat on DM365
May 31 23:05:16 <denix> kalem: mainline or PSP one on arago?
May 31 23:05:41 <kalem> denix: arago, build by myself
May 31 23:06:07 <kalem> denix: i don't know... should i need to upgrade u-boot on dm6446 ?
May 31 23:06:26 <denix> kalem: ubl?
May 31 23:06:40 <kalem> denix: what is ubl ?
May 31 23:06:49 <_av500_> a swiss bank
May 31 23:06:54 <_av500_> ah no, ubs
May 31 23:08:03 <kalem> umh
May 31 23:09:06 <denix> http://processors.wiki.ti.com/index.php/RBL_UBL_and_host_program
May 31 23:09:35 <KeatonT> So no change log for the kernel
May 31 23:10:11 <denix> anyway, like _av500_ correctly noted, it's #beagle, not #davinci :)
May 31 23:10:36 <aholler> as if the world doesn't already have enough three letter akronyms
May 31 23:11:10 <denix> no, _av500_ is 5 letters and 2 punctuations...
May 31 23:11:37 <kalem> denix: ok
May 31 23:12:04 <aholler> i ment psp, rbl, ubl, spl, mlo, foo, ...
May 31 23:12:25 <_av500_> KeatonT: git log is the kernel changelog
May 31 23:12:46 <KeatonT> yea trying to find it for the bone
May 31 23:12:54 <_av500_> and .18 release notes
May 31 23:13:24 <_av500_> for bone its .18 with bone patches on top
May 31 23:13:38 <_av500_> should be all in angstrom git
May 31 23:13:41 <KeatonT> _av500_ where are the release notes.
May 31 23:13:54 <_av500_> for linux itself?
May 31 23:14:11 <KeatonT> for the .18 kernel
May 31 23:14:38 <_av500_> no idea, I thought Linus writes a few words
May 31 23:15:11 <KeatonT> o_O
May 31 23:16:03 <aholler> https://lkml.org/lkml/2012/5/21/205
May 31 23:16:22 <aholler> years of experience in using google :/
May 31 23:17:23 <aholler> and years with altavista
May 31 23:17:51 <KeatonT> lol
May 31 23:18:02 <KeatonT> seems faster with the new kernel but it could all be in my head
May 31 23:19:13 <denix> _av500_: it's not in angstrom git!
May 31 23:20:11 <denix> aholler: you forgot "bar"
May 31 23:20:26 <aholler> ti has a bar?
May 31 23:21:31 <denix> aholler: I wish... no, it's another 3-letter acronym :) foo-bar
May 31 23:23:48 <ds2> hmmm
May 31 23:43:22 <mdp> _av500, we're at the pub..can't even attempt to parse your comment :)
May 31 23:44:28 <mranostay> why you have your computer at the pub?
Jun 01 00:04:30 <zhivko> _av500_: librxtx find only lp as vali port prefix ??? just compiled librxtx from source and turned debug on
Jun 01 00:05:05 <zhivko> _av500_: no tty no i2c prefix as valid??? wtf
Jun 01 00:23:30 <mdp> mranostay, phone. :P
**** ENDING LOGGING AT Fri Jun 01 02:59:58 2012