**** BEGIN LOGGING AT Fri Jul 25 03:00:01 2014
Jul 25 03:54:08 * reckoner133 listens to crickets
Jul 25 08:01:31 <Rotti> hi
Jul 25 11:18:59 <dnivra> Hello everyone! I'm debugging an OS for beagleboard-xM on qemu. I get a floating point exception when trying to write to CM_CLKSEL_DSS(0x48004E40). I read the OMAP35x TRM and I understand that the location is RW and so I don't understand why it faults. I looked around but couldn't find useful info and so any help/insight into what's going wrong would be great.
Jul 25 11:19:32 <woglinde> qemu is not the real hw
Jul 25 11:20:23 <dnivra> true; it could be a qemu bug. But I'm just not certain how I could verify.
Jul 25 11:20:47 <woglinde> try it und the real hw?
Jul 25 11:20:51 <woglinde> args on
Jul 25 11:21:03 <woglinde> damn typos
Jul 25 11:21:36 <dnivra> true; unfortunately I don't have the hardware so I'm kinda stuck with qemu for now.
Jul 25 11:22:03 <woglinde> dont know if one here will try it for you
Jul 25 11:22:12 <woglinde> most have bones now
Jul 25 12:48:27 <leandrosansilva> Hello to all. Is there a way to tell beaglebone black not to use power supplied by usb but only by the external one? I need to keep a BBB communicating with a PC via USB but using an external power supply but, when it's interrupted, beagle must turn off
Jul 25 13:32:37 <Guest98178> Hi Group, what do I need to do to get GL-ES working on BBB ?  thanks!
Jul 25 13:33:32 <XorA> a miracle
Jul 25 13:34:30 <Guest98178> oh, that's easy... :(    I purchased a couple of boards becaust GL-ES was part of the spec....
Jul 25 13:35:18 <XorA> Guest98178: I don't actually know, I beleive it is possible with certain kernels
Jul 25 13:35:34 <XorA> but GL-ES is always a PITA due to closed source bits
Jul 25 13:36:01 <Guest98178> that's what I read, but as I'd like to base something on it that needs (future) support, I'd rather use an official working release of OS and drivers....
Jul 25 13:37:09 <Guest98178> on the slower RPI GL-ES works as advertised, but the BBB's clock speed and number of PIO's won me over
Jul 25 13:37:15 <thurgood> do you need it for console, or X windows?
Jul 25 13:37:39 <Guest98178> mobile device that graphically displays measurement data in realtime...
Jul 25 13:37:59 <Guest98178> so likely an x-windows without elaborate gui
Jul 25 13:38:00 <XorA> for 2d GFX you don't need GL
Jul 25 13:38:19 <Guest98178> it's 3D I'm afraid.
Jul 25 13:38:44 <thurgood> robert nelson has it working for console in a 3.13 kernel
Jul 25 13:38:48 <Guest98178> As it is (or was?) part of the specs, I
Jul 25 13:38:58 <Guest98178> find it odd that nobody really is in rage over it..
Jul 25 13:39:25 <XorA> Guest98178: Imagination Technology have been annoying us for over 10 years now, we ran out of rage
Jul 25 13:39:34 <XorA> they basically hate linux
Jul 25 13:40:05 <Guest98178> ok, and still hanging out around here?  Any decent alternatives, like a faster RPI ?
Jul 25 13:40:24 <XorA> Cubieboard? its got a Mali
Jul 25 13:40:34 <Guest98178> thurgood: that's good to know...
Jul 25 13:40:35 <XorA> not used it personally
Jul 25 13:40:43 <Guest98178> checking now...
Jul 25 13:41:03 <XorA> I would check software state before buying hardware if you need 3d
Jul 25 13:41:38 <XorA> Minnowboard Max probably has the best 3D of any embedded board
Jul 25 13:46:07 <Guest98178> ok, thanks for all your responses - will need some more time to dig into this.  Thanks again
Jul 25 13:46:27 <XorA> a happy customer for once :-D
Jul 25 13:47:37 <thurgood> heh
Jul 25 13:49:27 <thurgood> are there any open source boards or at least boards available in high quantity with specs comparable to say the ODroid-U3
Jul 25 13:51:25 <Decessus> thurgood, I've seen a couple here and there, but that has to be the best price for the hardware I've seen
Jul 25 13:52:20 <thurgood> not as concerned with price as quality/specs and availability
Jul 25 13:53:48 <notserpe> leandrosansilva: in theory you could query pmic status register via i2c and see if AC power is present and if not powerdown.
Jul 25 13:53:51 <Decessus> Quality is not going to be there, in the few I've seen. Most have been low-quality chineese boards
Jul 25 13:54:02 <thurgood> ah
Jul 25 13:54:36 <Decessus> thurgood, what do you plan on using that horse power on?
Jul 25 13:54:59 <thurgood> 3d graphics mostly
Jul 25 13:56:32 <Decessus> Well, if you want power and availability..nVidia is coming out with a tablet ;p
Jul 25 13:57:11 <thurgood> doing some on a classic beagleboard based board, but we're barely getting by on speed and ram
Jul 25 13:57:29 <thurgood> looking to upgrade to something more modern, probably android based
Jul 25 13:57:50 <Decessus> thurgood, is it a BB white or black?
Jul 25 13:58:32 <thurgood> pre beaglebone
Jul 25 13:58:55 <Decessus> And android is a matter of software, and if it's on android you will loose a *lot* of raw performance
Jul 25 13:58:59 <thurgood> looked at the bb as an upgrade but it's unworkable
Jul 25 13:59:15 <thurgood> that's why we need the big jump in power :P
Jul 25 13:59:26 <XorA> thurgood: something with rk3188
Jul 25 13:59:36 <Decessus> thurgood, are you sure? The black edition's processor is able to execute at twice it's clock speed
Jul 25 14:00:03 <XorA> thurgood: pretty cheap and there are some fairly decent chinese machines made with it
Jul 25 14:00:05 <thurgood> yeah, bb rev c barely runs android
Jul 25 14:00:26 <thurgood> and it's 16 bit color depth is a real kill-joy
Jul 25 14:01:49 <Decessus> thurgood, well, I still think you will get far better results with BBB vs any android board under 2.56GHz quad
Jul 25 14:11:57 <XorA> thurgood: there is always the Arndale Octaboard which is a bit of a monster :-)
Jul 25 14:12:53 <ccbb> hi all
Jul 25 14:13:03 <ccbb> i have a problem
Jul 25 14:13:13 <ccbb> i buy a tplink TL-WN722N
Jul 25 14:13:43 <ccbb> but when i connect it with BBB, it's not work
Jul 25 14:13:59 <ccbb> i used usb hub power already
Jul 25 14:14:33 <ccbb> how can i make it work now?
Jul 25 14:15:11 <ccbb> my pc run ubuntu, i the tplink work well on it
Jul 25 14:15:26 <ccbb> s/i/and/
Jul 25 15:45:42 <jhn> Is thee a reason why the latest images are packed .xz? Some platforms out there ship without a built in .xz decoder.
Jul 25 15:48:01 <Rickta59> https://www.kernel.org/happy-new-year-and-good-bye-bzip2.html jhn
Jul 25 15:50:11 <jhn> I am perfectly happy with .gz - but again, a format was chosen, that has not yet found it’s way into every tar.
Jul 25 16:05:29 <tpw_rules> anybody used the built-in can devices? i'm having trouble with them
Jul 25 16:07:14 <jhn> Oh, let me know of any solution - also awaiting CAN usage RSN.
Jul 25 16:07:50 <tpw_rules> canconfig from TI's guide fails with dcan1
Jul 25 16:08:04 <tpw_rules> you do need transceivers, there's no way to work without that
Jul 25 16:08:54 <jhn> Tranceivers are understood.
Jul 25 16:09:40 <tpw_rules> i don't have my scope handy though >:(
Jul 25 16:15:24 <tpw_rules> bah, canconfig doesn't like the can1 name. it doesn't matter which order the two are registered in
Jul 25 16:16:23 <jhn> U kno that link: http://www.embedded-things.com/bbb/enable-canbus-on-the-beaglebone-black/
Jul 25 16:16:54 <tpw_rules> yes
Jul 25 16:17:00 <tpw_rules> can0 seems to be working fine
Jul 25 16:17:03 <tpw_rules> but can1 won't
Jul 25 16:17:12 <tpw_rules> even if can0 == dcan0 and vice versa
Jul 25 16:17:14 <tpw_rules> or the other eway
Jul 25 16:48:58 <jhn> OK Messieurs, serious trouble ahead with .xz files. D/L the latest Debian image,  tar xf: Unrecognized archive format. D/L again, diff: Identical, same error.
Jul 25 16:49:49 <tpw_rules> use 'xz'
Jul 25 16:49:57 <tpw_rules> they're LZMA, i don't think tar can work with them
Jul 25 16:50:21 <tpw_rules> xz -d -v file | dd of=/dev/card
Jul 25 16:50:30 <XorA> use capital J for xz files
Jul 25 16:50:44 <tpw_rules> or do that apparently
Jul 25 16:50:59 <jhn> Then D/L linux headers fromhttp://www.armhf.com/boards/beaglebone-black/ as .xz: perfect decompression.
Jul 25 16:51:43 <tpw_rules> what
Jul 25 16:52:08 <jhn> Then i file-d the d/L: file tells me for each that it is xz compressed data
Jul 25 16:54:43 <jhn> Then I continued to create some tar files witz gzip, bzip, and xz compression on my system: all were detected fine and automatically uncompressed. There is no need to tell tar what format the file has.
Jul 25 16:55:39 <jhn> And even helping tar with xfJ does not help.
Jul 25 16:59:51 <thurgood> f should come last as far as I can remember
Jul 25 17:04:21 <jhn> xJf doesn’t cut it either.
Jul 25 17:07:43 <XorA> jhn: I was answering tpw_rules when he asked about tar support
Jul 25 17:14:10 <jhn> OK, big question: Could anyone try D/L the http://debian.beagleboard.org/images/bone-debian-7.5-2014-05-14-2gb.img.xz and tell me what his experiences are?
Jul 25 17:15:29 <tpw_rules> it worked for me last night
Jul 25 17:15:39 <tpw_rules> curl <url> | xz -d -v --stdout | dd of=/dev/sdcard
Jul 25 17:15:47 <jhn> F*F*F*F*F*F*FF*F*F
Jul 25 17:16:12 <tpw_rules> excuse me
Jul 25 17:16:42 <jhn> If there is any of the slightest possibility of a rproblem, i will cat it.
Jul 25 17:17:05 <jhn> catch it I mean.           A L W A Y S
Jul 25 17:19:18 <jhn> MD5 is correct as well. So I csn create xz files, decompress xz files at will, but NOT the bone one. Any ideas on how to proceed?
Jul 25 17:19:35 <tpw_rules> does it work with xz?
Jul 25 17:19:49 <tpw_rules> remember it's a blob. not an archive
Jul 25 17:21:55 <jhn> Mgmt lsummary: I can compress and decompress my own xz files.
Jul 25 17:22:46 <jhn> I can decompress http://s3.armhf.com/dist/bone/linux-headers-3.14.4.1-bone-armhf.com.tar.xz
Jul 25 17:23:15 <jhn> I am unable to decompress http://debian.beagleboard.org/images/bone-debian-7.5-2014-05-14-2gb.img.xz
Jul 25 17:23:19 <jhn> Now what?
Jul 25 17:25:22 <Humpelst1lzchen> jhn: xz -d bone-debian-7.5-2014-05-14-2gb.img.xz works for me
Jul 25 17:26:05 <jhn> That is fine for you. For me it unfortunately it does NOT work.
Jul 25 17:26:49 <Humpelst1lzchen> just answering your question from 19:14:10 gmt+2, I don't need the image
Jul 25 17:26:51 <jhn> Are there other sources for the img that are packed with anpother tool? Like with gz?
Jul 25 17:28:44 <tpw_rules> what does not work mean?
Jul 25 17:29:17 <tpw_rules> int sndbuf = 32768; setsockopt(fd, SOL_SOCKET, SO_SNDBUF, (void *)&sndbuf, sizeof(sndbuf)); <-- what in the actual fuck?
Jul 25 17:45:45 <tpw_rules> the best part about open source: it can be buggy as shit because anybody who uses it can fix it themselves >:(
Jul 25 17:48:23 <myself> or at least whine about it
Jul 25 17:48:31 <myself> that's my method ;)
Jul 25 18:21:46 <jhn> tpw_rules, I can not decompress the latest and greatest bbb image. Quite AFU to me.
Jul 25 18:24:01 <jhn> Again the question: Are there other sources from which I can D/L bone-debian-7.5-2014-05-14, preferably as a .gz?
Jul 25 18:30:45 <jhn> What am I to expect of http://s3.armhf.com/dist/bone/debian-wheezy-7.5-rootfs-3.14.4.1-bone-armhf.com.tar.xz? It is only 100 MB whereas the beagle.org image has 492 MB.
Jul 25 18:34:51 <tpw_rules> what does "not decompress" mean?
Jul 25 18:37:30 <jhn> tar: Unrecognized archive format
Jul 25 18:38:47 <tpw_rules> use xz!
Jul 25 18:39:58 <Humpelst1lzchen> jhn: did you try to decompress any other xz?
Jul 25 18:40:21 <jhn> I am able to uncompress http://s3.armhf.com/dist/bone/linux-headers-3.14.4.1-bone-armhf.com.tar.xz
Jul 25 18:40:46 <jhn> With the very same tar xf
Jul 25 18:41:08 <jhn> So logic tells me, it can’t be my tar.
Jul 25 18:44:19 <Humpelst1lzchen> jhn: what does xz -l bone-debian-7.5-2014-05-14-2gb.img.xz say to you?
Jul 25 18:45:15 <jhn> Nothing , as I do not have xz. It’s that _other_ U*X OS.
Jul 25 18:46:47 <Humpelst1lzchen> uuh?
Jul 25 18:47:08 <Humpelst1lzchen> jhn: and how do you try to uncompress bone-debian-7.5-2014-05-14-2gb.img.xz
Jul 25 18:48:06 <jhn> tar xf like any other xz file.
Jul 25 18:49:03 <Humpelst1lzchen> jhn: but you did notice that bone-debian-7.5-2014-05-14-2gb.img.xz does not contain a tar?
Jul 25 18:49:29 <jhn> See my 20:30 and 20:40 entries - both tar xf fine
Jul 25 18:49:59 <Humpelst1lzchen> jhn: but both files are tars
Jul 25 18:50:41 <jhn> True. My understanding is: XZcompresses a single file. Be it a tar or an img.
Jul 25 18:50:57 <jhn> So I expect to be left with the img after tar xf.
Jul 25 18:51:36 <jhn> U R right - this is the visible and hence obvious difference.
Jul 25 18:51:37 <Humpelst1lzchen> jhn: first line: yes
Jul 25 18:51:44 <jhn> What a F*!
Jul 25 18:52:06 <Humpelst1lzchen> but tar will throw an error as soon as it doesn't see a tar after uncompressing
Jul 25 18:58:23 <tpw_rules> (i told you!)
Jul 25 18:59:29 <tpw_rules> if it's another unix os, The Unarchiver is a brilliant gui unarchiving utility
Jul 25 18:59:33 <tpw_rules> or just use brew to install xz
Jul 25 18:59:58 <tpw_rules> incidentally i got canbus working :D
Jul 25 19:00:05 <Humpelst1lzchen> what is brew?
Jul 25 19:00:15 <tpw_rules> osx package manager
Jul 25 19:00:29 <Humpelst1lzchen> aah apple user. That explains some bits
Jul 25 19:00:46 <tpw_rules> well i dunno. also i'm an apple user thank you very much :<
Jul 25 19:00:55 <Humpelst1lzchen> always ;)
Jul 25 19:01:17 <tpw_rules> now i need to go steal a car and plug my beaglebone in
Jul 25 19:08:14 <jhn> I have a car where I can guarantee the BB _WILL_ fail!
Jul 25 19:09:20 <tpw_rules> define fail
Jul 25 19:10:09 <jhn> Will not work in the sense of not delivering data. Not a single bit.
Jul 25 19:10:51 <tpw_rules> well it has two interfaces so it can talk to itself while inside the car
Jul 25 19:11:11 <jhn> Isn’t brew a multiplatform package mgr?
Jul 25 19:11:32 <tpw_rules> not as far as i know
Jul 25 19:11:44 <jhn> In that case the conversaion will be very deteministic and monotone.
Jul 25 19:12:40 <jhn> Are you a woman?
Jul 25 19:13:04 <tpw_rules> i don't think that's relevant?
Jul 25 19:13:54 <jhn> There is an old rule by mother nature, taht women are always right. And you are exposing that behaviour… :-)
Jul 25 19:17:01 <Cueball> LoL
Jul 25 19:18:45 <jhn> tpw_rules Humpelst1lzchen, thx for tracking that down.
Jul 25 19:23:51 <jhn> tpw_rules didn’t get the blob thing as I didn’t know that tar will exit if it does not see a tar inside. I installed Unarchiver by now, and was able to extract and mount the image.
Jul 25 19:24:13 <tpw_rules> what are you trying to do with it? use disk utility to write it to an sd card or dd
Jul 25 19:25:43 <jhn> Wanted to see whats in there and hopefully use it for my cross compiler toolchain.
Jul 25 19:25:46 <tpw_rules> oh
Jul 25 19:25:52 <tpw_rules> all the binaries are going to be arm
Jul 25 19:26:23 <jhn> I am more interested in the headers and libs.
Jul 25 19:26:57 <jhn> So I can cross compile on the host and just copy over the image.
Jul 25 19:27:28 <jhn> I can then even cross compile static IIUC.
Jul 25 19:31:36 <jacekowski> cross compiling is pita
Jul 25 19:31:39 <jacekowski> but possible
Jul 25 19:37:18 <jhn> Waiting 6 minutes for a compile on the BBB wheras the host needs 30 seconds is _more_ PITA
Jul 25 20:04:54 <ds2> cross compiling is easy
Jul 25 20:05:04 <ds2> ...as long as you avoid autoconf and libtool crap
Jul 25 20:25:02 <notserpe> two montreal mayors and one from the 'burbs resigned and all kinds of other incredible stuff but Toronto gets the spotlight for horrendous Canadian city :D
Jul 25 20:25:18 <notserpe> doh, wrong channel.
Jul 25 20:31:50 <xiqy> is it possible to do audio out over hdmi on the bbb *without* a connected display? I.e. bbb -> receiver -> speakers?
Jul 25 20:55:47 <tpw_rules> jhn: https://github.com/jsnyder/arm-eabi-toolchain
Jul 25 20:55:55 <tpw_rules> i don't know anything else, don't ask
Jul 25 20:57:25 <jhn> You are using exact this on which Mac OS release?
Jul 25 20:58:34 <tpw_rules> i'm not using it
Jul 25 20:58:37 <tpw_rules> it looked relevant for you
Jul 25 20:58:45 <tpw_rules> what do you even need to build?
Jul 25 20:59:03 <jhn> Ah Ok, got it.
Jul 25 20:59:11 <tpw_rules> switch to debian if you haven't, it's repos seem better than ångstrom
Jul 25 21:00:21 <jhn> Building a library around can and eth and an application that uses this.
Jul 25 21:00:51 <tpw_rules> how have you got it to 6 minutes build time? all the can stuff i've been doing couldn't've taken more than 10 seconds
Jul 25 21:01:33 <jhn> The library dealing with the eth took 6 minutes.
Jul 25 21:01:41 <tpw_rules> if you're using make properly, that's only necessary once
Jul 25 21:01:53 <tpw_rules> even if you change it, it will only recompile a small portion
Jul 25 21:07:26 <jhn> True. Having succeeded at cross compiling Atmel AVR I want to master ARM as well. It is confusing and irritating, but in the end it is just _just_ the right combination of a dozen configure parameters.
Jul 25 21:10:35 <tpw_rules> but the arm already has a toolchain. oh well, it's your call
Jul 25 21:10:43 <notserpe> jhn: welcome to embedded, here's your merit badge!
Jul 25 21:11:11 <tpw_rules> i got cross compiling working for rpi but i don't remember how
Jul 25 21:11:46 <tpw_rules> it was for no os though, so i don't think it would apply
Jul 25 21:12:36 <jhn> Yup, AVR is bare metal as well - this is a _bit_ easier, as you do not need a sysroot.
Jul 25 21:14:02 <notserpe> jhn: you're familiar with linaro's prebuilt arm toolchains?
Jul 25 21:14:42 <jhn> Having fallen on my nose and bleeded severely, I want to stay away from prebuilt.
Jul 25 21:56:42 <jhn> There is loads of support for BBB running Angström, but zilch for Debian. I have a Rev C in front of me with an uptime of 5 days. I can’t wait until *someone* gets a toolchain *someday* out of the house.
Jul 25 21:58:58 <rcn-ee> jhn: https://github.com/RobertCNelson/cross-compiler
Jul 25 21:59:59 <jhn> Youp, you have been referenced multiple time, you are next on my radar! :-)
Jul 25 22:00:41 <rcn-ee> it's built with the exact same libs as what's in the image.. the only catch, you have to run it debian wheezy..
Jul 25 22:01:22 <jhn> Darwin != Debian :-(
Jul 25 22:01:50 <rcn-ee> jhn, nfs/samba + vm + ssh.. if that's too hard, then i can't help. ;)
Jul 25 22:07:15 <jhn> When cross compiling and having the bbb debian root on the host, do I only have to set sysroot to the bbb root?
Jul 25 22:08:17 <rcn-ee> sure..  but it's just faster/easier to build native. ;)
Jul 25 22:10:34 <jhn> I am pretty sure my 4 core intel will outperform the brave ARM.
Jul 25 22:14:16 <rcn-ee> Fire up qemu, boot an "armhf" image, watch it build slower then the BBB. ; )
Jul 25 22:15:53 <jhn> I agree, I never embraced emulators. I don’t trust debuggers either… :-)
Jul 25 22:22:27 <tpw_rules> whee, beaglebone can interfaces are go
Jul 25 22:23:27 <tpw_rules> btw angström vs debian: the beaglebone project is migrating to debian and all new beaglebones ship with it. plus, it's more supported in general
Jul 25 22:27:53 <jhn> No contradiction to my saying. The switch happened and everybody must adjust now.
Jul 25 22:30:51 <rcn-ee> what switch. ;) Some of had been using debian since the beginning...
Jul 25 22:38:43 <BeagleGithub> [kernel] RobertCNelson pushed 1 new commit to 3.8: http://git.io/2PgF5w
Jul 25 22:38:43 <BeagleGithub> kernel/3.8 64a7d6a Robert Nelson: 3.8: merge in all cape changes up to bone62...
Jul 25 22:38:59 <tpw_rules> argh debian armhf doesn't have python 3.4
Jul 25 22:39:41 <rcn-ee> jessie does: https://packages.debian.org/jessie/python3.4
Jul 25 22:40:13 <tpw_rules> since i'm not too experienced with debian, how do i get that into my system?
Jul 25 22:40:42 <rcn-ee> http://elinux.org/BeagleBoardDebian#Debian_Testing_.28jessie.29
Jul 25 22:40:49 <ryanneufeld> tpw_rules: you're correct, tar is unable to open that compression
Jul 25 22:40:56 <ryanneufeld> oh
Jul 25 22:41:05 <tpw_rules> aww really? i thought that's the one i downloaded
Jul 25 22:41:08 <tpw_rules> blast
Jul 25 22:41:21 <tpw_rules> wait a second why's it a tar
Jul 25 22:41:28 <jhn> apt-get helped me install boost , maybe py is avail as well
Jul 25 22:41:53 * tpw_rules backs up his configuration changes
Jul 25 22:41:54 <rcn-ee> it's not an *.img but a multi device *.tar with an install script..
Jul 25 22:42:01 <tpw_rules> darn
Jul 25 22:42:07 <tpw_rules> yay third format's the charm?
Jul 25 22:42:46 <rcn-ee> otherwise there's always the "netinstall" https://github.com/RobertCNelson/netinstall  If you've ever used Debian's net.iso (debian installer)
Jul 25 22:43:43 <tpw_rules> which is better? i haven't
Jul 25 22:47:23 <tpw_rules> i'm going to follow the wikipedia guide
Jul 25 22:47:53 <tpw_rules> wget is complaining about certs, you may want to check that
Jul 25 22:48:10 <rcn-ee> install ca-certificates
Jul 25 22:48:19 <tpw_rules> this is on osx
Jul 25 22:49:00 <tpw_rules> hm, firefox is okay. must be on my end then
Jul 25 22:49:18 <rcn-ee> that script won't run on osx, you need the *.img's..
Jul 25 22:49:27 <tpw_rules> why not?
Jul 25 22:49:59 <rcn-ee> here's a jessie *img i did for drew: http://rcn-ee.net/deb/testing/2014-07-03/bone-debian-jessie-console-2014-07-03-4gb.img.xz
Jul 25 22:50:07 <tpw_rules> same one as before?
Jul 25 22:50:19 <rcn-ee> un-released.. it was  test..
Jul 25 22:50:31 <tpw_rules> well as the tar would have made
Jul 25 22:51:12 <rcn-ee> except the *.tar has a script you need to run to create the microSD.. Other then building a hackintosh and finding out how useless their unix tools are... Yeah.. i haven't bothered to add support..
Jul 25 22:51:30 <tpw_rules> well yeah but what can that script possibly do that's broken
Jul 25 22:51:31 * veremit chuckles to himself
Jul 25 22:51:59 <rcn-ee> mac os has a really crappy old version of fdisk..
Jul 25 22:52:27 <tpw_rules> now you've made me make it work
Jul 25 22:53:24 <tpw_rules> also curl works great with https so i guess wget is having a bad day
Jul 25 22:55:19 <rcn-ee> i think on mac osx you need gnutls installed for wget to work with https
Jul 25 22:55:35 <tpw_rules> i'll look into it
Jul 25 22:55:54 <tpw_rules> do you have some sort of speed = 1/progress script on your fileserver
Jul 25 22:55:59 <rcn-ee> http://coolestguidesontheplanet.com/install-and-configure-wget-on-os-x/
Jul 25 22:57:12 <jhn> dunno if it is a redirection problem, but wget —trust-server-names might help
Jul 25 22:57:23 <tpw_rules> curl works
Jul 25 22:57:28 <tpw_rules> or just removing the s
Jul 25 22:57:50 <jhn> I replacd wget by curl -L
Jul 25 22:57:55 <tpw_rules> the one osx command line feature i miss daily on linux is 'open'
Jul 25 22:58:01 <tpw_rules> open <file> and it opens the relevant program
Jul 25 22:58:03 <tpw_rules> or director
Jul 25 23:00:50 <tpw_rules> first hiccup: mktemp template is mandatory
Jul 25 23:05:56 <BeagleGithub> [kernel] RobertCNelson closed pull request #98: cape: add BB-BONE-SERL-01-00A2 (3.8...3.8) http://git.io/_fnczg
Jul 25 23:57:20 <tpw_rules> rcn-ee: script seems to work
Jul 25 23:57:22 <tpw_rules> also, are you german?
Jul 25 23:57:41 <rcn-ee> part of me is.
Jul 25 23:57:49 <tpw_rules> do you know how to speak it?
Jul 25 23:58:05 <rcn-ee> nope. ;)
Jul 25 23:58:16 <tpw_rules> i can't figure out why you are arbitrarily capitalizing words in your script
Jul 25 23:58:52 <rcn-ee> yeah, i'm slowly factoring those out. ;)
Jul 25 23:59:20 <tpw_rules> what do i need to do to the boot partition on eMMC to get it to prefer the sd card?
Jul 26 00:00:10 <rcn-ee> Either.. no bootloader in the eMMC, or use the "bbb-uEnv.txt" (rename as uEnv.txt) that it dumped to the boot partition..
Jul 26 00:00:26 <ds2> Or change the sysboot resistors
Jul 26 00:00:39 <tpw_rules> just copy from sd to emmc? it worked earlier until i flashed a new debian onto the emmc
Jul 26 00:00:46 <rcn-ee> i added a "--bbb-old-bootloader-in-emmc" flat to setup_sdcard.sh in git. ;)
Jul 26 00:01:15 <ds2> rcn-ee: old == preDT?
Jul 26 00:01:48 <tpw_rules> i see no such file
Jul 26 00:01:55 <rcn-ee> ds2 pre v2014.07: got kinda drunk over july 4th: http://elinux.org/Beagleboard:U-boot_partitioning_layout_2.0
Jul 26 00:02:39 <rcn-ee> tpw_rules, dump this into uEnv.txt : https://github.com/RobertCNelson/omap-image-builder/blob/master/tools/setup_sdcard.sh#L514
Jul 26 00:02:40 <tpw_rules> ah well, just renamed the bootloader
Jul 26 00:03:05 <tpw_rules> why don't you use the fancy echo <<EOF syntax
Jul 26 00:04:10 <ds2> aka the change that diverges even more from Angstrom
Jul 26 00:06:03 <rcn-ee> ds2, correct.. but you can run a kernel: make install (or debian/ubuntu apt-get install linux-image), (or fedora yum install linux) yada yda.. and reboot...
Jul 26 00:07:12 <ds2> rcn-ee: I personally think screwing around with u-boot like that is a bad idea
Jul 26 00:07:57 <rcn-ee> actually it's the way u-boot is headed.. dennis @ fedora has been working on it...  i got tired of waiting..
Jul 26 00:08:17 <ds2> but in practice, what irks me is this change has wasted a lot of time cuz it is looking at a different place then I installed my kernel (I manually install as I do kernel dev)
Jul 26 00:09:08 <ds2> until Fedora shows up in many embedded products, that is largely, IMO irrelevant for the BB's
Jul 26 00:09:11 <tpw_rules> rcn-ee: wow that image has nothing
Jul 26 00:10:01 <rcn-ee> it's searching /uEnv.txt (looking for uenvcmd to be set) then /boot.scr then /boot/boot.scr until finally "/boot/uEnv.txt" on all 7 possible parittions..
Jul 26 00:12:04 <rcn-ee> tpw_rules, i was just testing egl/weston with that image.. so it's pretty empty
Jul 26 00:12:14 <tpw_rules> i burned it with the script
Jul 26 00:12:17 <rcn-ee> or wait, you grabbed the minimal. ;)
Jul 26 00:12:46 <tpw_rules> to be fair i really didn't want a gui
Jul 26 00:12:58 <tpw_rules> tmux is all the gui i need
Jul 26 00:13:06 <tpw_rules> rs232 master race
Jul 26 00:14:59 <tpw_rules> what is the point of autogen.sh to generate configure.sh to generate make to generate a program
Jul 26 00:15:04 <tpw_rules> how many levels do we need
Jul 26 00:15:26 <rcn-ee> that is unix's fault..
Jul 26 00:15:56 <ds2> as many as it takes to defeat any cross building system
Jul 26 00:16:48 <tpw_rules> i assume i'll be installing an entire development environment
Jul 26 00:17:20 <tpw_rules> actually first i need to expand rootfs
Jul 26 00:17:32 <rcn-ee> start with build-essential
Jul 26 00:17:44 <rcn-ee> if you ran the script, it should be 100%...
Jul 26 00:17:58 <tpw_rules> autoconf isn't installed, does that cover it
Jul 26 00:18:05 <tpw_rules> oh is it? i never looked
Jul 26 00:18:10 <vagrantc> ds2: i think the u-boot improvements rcn-ee are great ... debian kernels can now work essentially out of the box...
Jul 26 00:18:15 <tpw_rules> so it is
Jul 26 00:18:29 <tpw_rules> i guess it would be
Jul 26 00:18:29 <ds2> rcn-ee: all these details are handled automatically by the flasherimage, right?
Jul 26 00:19:13 <rcn-ee> ds2, hell no ;) i looked with grub does: https://github.com/RobertCNelson/omap-image-builder/blob/master/target/other/zz-uenv_txt
Jul 26 00:19:58 <rcn-ee> i just pass uname_r to the u-boot and it finds.. or if it's an external device, i pass uuid
Jul 26 00:20:05 <mrpackethead> hey rcn-ee
Jul 26 00:20:06 <mrpackethead> :-)
Jul 26 00:22:26 <vagrantc> rcn-ee: doesn't it also work with flash-kernel's bootscript, independent of uname_r ?
Jul 26 00:23:34 <rcn-ee> vagrantc, it 'should'...  i kinda neutered flash-kernel https://github.com/RobertCNelson/netinstall/blob/master/mk_mmc.sh#L833 (those ubuntu guys like to add bugs that lock me up..)
Jul 26 00:25:24 <rcn-ee> vagrantc, i kinda wish the ubuntu d-i guys would atleast copy more of your work, they somehow broke: "d-i apt-setup/local0/repository string"
Jul 26 00:25:28 <ds2> rcn-ee: no, I mean for someone just doing a update of...say... an brand new but old board (old stock). they still make a flasher uSD, press button, power on and it will flash it like before...right?
Jul 26 00:25:57 <rcn-ee> ds2, the "flasher" images work with everything after Angstrom's 2013.07.xy
Jul 26 00:26:08 <rcn-ee> with no button pushes..
Jul 26 00:26:32 <ds2> rcn-ee: hmmmm...makes it hard to put together a tutorial
Jul 26 00:26:47 <ds2> I suppose button pushes don't hurtwith these newer flasher images
Jul 26 00:27:02 <rcn-ee> it doesn't hurt... specially if someone had fun with the eMMC's bootloader..
Jul 26 00:27:32 <ds2> 'k
Jul 26 00:28:21 <rcn-ee> My mistake was pushing out a "bone" image thinking well there's bone people and boneblack people... they'll know to push the button... opps!!!!
Jul 26 00:29:09 <rcn-ee> i'll resolve that in my next monthly snapshot, around the first of august
Jul 26 00:29:13 <vagrantc> rcn-ee: i can't take much credit for d-i, but yeah, i've often been surprised by brokenness in ubuntu's d-i as well.
Jul 26 00:30:46 <tpw_rules> aw crap. what built before won't now and it's not even my code
Jul 26 00:31:46 <rcn-ee> it's like swiss cheese.  i should create a bug on launchpad, please rebase with debian testing.. but they like to ignore my bugs. ;)
Jul 26 00:36:29 <rcn-ee> vagrantc, btw: for flash-kernel here's my list of boards: https://github.com/RobertCNelson/netinstall/blob/master/lib/flash_kernel/all.db all the i.mx6's should boot with debian v3.14.x minus the dl-udoo which i need to submit upstream..
Jul 26 00:40:41 <tpw_rules> where does ldconfig hide? i need to change /etc/ld.so.conf
Jul 26 00:42:50 <tpw_rules> rcn-ee: you built the thing, i'm blaming you
Jul 26 00:43:09 <rcn-ee> tpw_rules, /sbin/ldconfig
Jul 26 00:43:31 <tpw_rules> oh, apparently that's not in $path
Jul 26 00:43:56 <rcn-ee> (/sbin) isn't traditionally in a user path in debian..
Jul 26 00:44:09 <tpw_rules> it was on wheezy?
Jul 26 00:44:44 <rcn-ee> tpw_rules, i patched that for bb.org: https://github.com/RobertCNelson/boot-scripts/blob/master/mods/debian-add-sbin-usr-sbin-to-default-path.diff
Jul 26 00:45:02 <tpw_rules> i'm stealing that patch
Jul 26 00:45:18 <rcn-ee> that's why it's there. ;)
Jul 26 00:45:38 <tpw_rules> ooooh, it runs if i say sudo. that's clever
Jul 26 00:46:12 <rcn-ee> with a lot of programs they'll run just fine as a normal user.. just add /sbin/...
Jul 26 00:46:45 <tpw_rules> i usually need the "you're not root, stupid" prompt to remind me to run them correctly
Jul 26 00:53:29 <tpw_rules> can i set the version of a library in pkg-config's mind? i've built one from source but the stupid configure script of another doesn't consider trunk > 0.0.8
Jul 26 01:22:34 <tpw_rules> > shutdown: command not found
Jul 26 01:23:14 <tpw_rules> oh it's an sbin again. i just never remember to use sudo
**** ENDING LOGGING AT Sat Jul 26 03:00:00 2014