**** BEGIN LOGGING AT Mon May 21 02:59:59 2012 May 21 03:51:39 hmmmm May 21 05:42:54 Isn't there a syslog in Angstrom? May 21 05:47:53 there is May 21 05:48:00 use 'journalctl' to view it May 21 05:48:10 or install rsyslog for a more traditional syslog May 21 05:49:02 So by default the file is not there in /var/log huh? Let me try the options you suggested. May 21 05:51:06 I can't seem to locate the 'journalctl' package in the "opkg list". Does it have a different package name? May 21 05:57:23 ah yes beaglebone awesomeness :P May 21 05:57:52 The syslog? May 21 05:58:48 I'm just enjoying playing around with mine. May 21 05:59:02 Makes the arduino look like a toy. May 21 05:59:06 Oh like that.. :) May 21 06:00:04 You using the bone? May 21 06:00:10 KeatonT: koen suggested I use something called 'journalctl' to view my syslog messages in angstrom (BeagleBoard-XM). I can't seem to locate that package though. May 21 06:00:24 I think it is already installed. May 21 06:00:38 So why is it showing command not found? May 21 06:03:44 He also suggested "rsyslog" which I cannot locate using opkg as well. May 21 06:04:02 did you opkg update? May 21 06:04:08 Yup. May 21 06:04:09 yea May 21 06:04:16 opkg install syslog May 21 06:04:20 rsyslog May 21 06:04:42 What does it say when you do "opkg install rsyslog" May 21 06:04:56 Unknown package "rsyslog" May 21 06:06:05 wget http://www.angstrom-distribution.org/feeds/core/ipk/eglibc/armv7a/base/rsyslog_5.8.0-r2.0_armv7a.ipk May 21 06:06:21 That is what opkg should do right? May 21 06:06:33 then do opkg install rsyslog_5.8.0-r2.0_armv7a.ipk May 21 06:06:40 Let me try. May 21 06:06:54 blucalvin: you are running a recent angstrom release, right? May 21 06:07:49 blucalvin: you are running a recent angstrom release, right? May 21 06:09:35 Collected errors: May 21 06:09:36 * wfopen: rsyslog_5.8.0-r2.0_armv7a.ipk: No such file or directory. May 21 06:09:36 * pkg_init_from_file: Failed to extract control file from rsyslog_5.8.0-r2.0_armv7a.ipk. May 21 06:29:23 well goodnight :P May 21 07:50:43 So, using i2cdetect on my beaglebone, I only see two buses, numbered 1 and 3, which I presume correspond to internal buses 0 and 2 May 21 07:51:05 What do I need to do to enable I2C 1, the only one not otherwise in use? May 21 07:54:07 patch the kernel May 21 07:56:29 Is that really the only way? Why does it ship with one of the buses exposed on the expansion headers disabled in the kernel? May 21 08:00:24 nickjohnson: i2c-2 clearly has minor device number 2, just by looking at /dev/i2*. So make the character device yourself and give it a poke. Might crash if there's nothing there though. May 21 08:00:49 fiola: Hm, good idea May 21 08:00:53 Just got my beaglebone a few days ago though, so don't know much about it. Loveloy board :-) May 21 08:01:31 I don't have that bus hooked up, but it seems a reasonble assumption May 21 08:02:01 I haven't backed up my sd card yet, or I'd give it a try :P May 21 08:02:23 Bah, no luck: Error: Could not open file `/dev/i2c-2': No such device May 21 08:02:37 It seems really weird that it's shipped with that bus entirely disabled May 21 08:03:18 On the positive side, the fact that the device entry in table was unused and it detected it instead of crashing is good :-) May 21 08:04:00 There might be an enabling node under /sys May 21 08:05:15 There's stuff under /sys/bus/i2c, just looking May 21 08:05:21 This groups post claims it can be enabled using uEnv.txt: https://groups.google.com/forum/?fromgroups#!topic/beagleboard/Q1pDr1lT7Gk May 21 08:05:46 Yeah, that's stuff for the enabled buses. If there's a kernel driver, you can access the device via there, such as reading and writing cape eeproms. Pretty cool, but not what I need. :) May 21 08:06:54 I was looking for a port enabler, you know, echo 1 > blah May 21 08:08:51 right May 21 08:13:06 I have a question too. Any idea how the beaglebone tries to get itself an IP address when there is no Ethernet cable plugged in and the Ethernet-over-USB gadget is used? Is it expecting the host to pop up a DHCP listener on the gadget interface? May 21 08:14:33 Can't think of any other way to do it, but I really didn't want to run DHCP May 21 08:15:15 By default, it runs its own DHCP server May 21 08:15:26 Which allocates 192.168.7.1 to the host, and 192.168.7.2 to itself May 21 08:15:43 If you want it to use DHCP from the host instead, do `udhcpc -i usb0` on the beaglebone May 21 08:17:12 I think it only allocates that if you give it a physical lead. I'm on the console and it shows no IP address allocated to its Ethernet-over-USB interface May 21 08:17:39 You have to plug in the USB, then eject the mass storage gadget first, yes May 21 08:18:08 If you eject the mass storage then you can't talk to the console to see what's happening May 21 08:19:14 "If you eject the mass storage then you can't talk to the console to see what's happening" ???? May 21 08:19:20 I've been doing that all day May 21 08:19:35 unless with 'eject' you mean 'unplug cable' May 21 08:19:56 Eh? No, ejecting the mass storage device doesn't affect the serial gadget May 21 08:20:00 koen: Really? Well I'm connected to the console right now, but if I eject /dev/sdd then the console session locks May 21 08:20:34 Does another network interface appear? May 21 08:21:13 I agree totally that it shouldn't lock, since the FTDI is on a completely different USB hub port to the two gadgets May 21 08:22:14 Yup, locks up instantly, just tried again May 21 08:22:38 I can't seem to get gcc working on my angstrom. When I type in gcc test.c, the output I get is: May 21 08:22:40 /usr/lib/gcc/arm-angstrom-linux-gnueabi/4.3.3/../../../../arm-angstrom-linux-gnueabi/bin/ld: cannot find -lgcc_s May 21 08:22:40 collect2: ld returned 1 exit status May 21 08:23:05 What could be the trouble? May 21 08:24:26 koen: connmann dhcp in milliseconds, you have that? May 21 08:26:47 koen: The only way I can recover console session after "eject /dev/sdd" is by connecting to the other FTDI port (which in theory is JTAG, lol), hitting return once which causes the mass storage to return, and then reconnect to the console port, which is then back at login. May 21 08:29:57 av500: yes May 21 08:30:10 blucalvin: so install libgcc* May 21 08:32:57 Thanks koen! :) Having used Ubuntu for so long, I can't seem to debug these problems. May 21 08:35:46 or opkg install task-native-sdk to drag in more things you need May 21 08:50:06 Okay, confirmed, using uEnv.txt allows me to bring up I2C2 May 21 08:50:12 Though I haven't tried talking to anything over it May 21 09:06:23 So, the SRM doesn't say anything about it, but is it okay to tie a cape's EEPROM lines high instead of jumpering them when you're building a top cape? May 21 09:06:33 Since it'll always be on top of the stack, you know that address won't be in use May 21 09:19:51 nickjohnson: no May 21 09:20:02 Why not? May 21 09:20:45 because another cape might be just as arrogant, but with with stackable headers May 21 09:20:50 so always add the 2 dips May 21 09:21:35 So I should build my cape to work around the deficiencies of a broken cape that has stackable headers but always thinks it's on top? May 21 09:22:17 the address has nothing to do with being on top or not May 21 09:22:52 Well, the address is supposed to reflect a board's position in the stack May 21 09:23:24 Which is at most 4 boards tall. My board can't have others stacked on top of it. Therefore, nothing else can have address 4. May 21 09:26:30 nickjohnson: bullshit May 21 09:26:41 Which bit? May 21 09:26:55 11:22 < nickjohnson> Well, the address is supposed to reflect a board's position in the stack May 21 09:26:59 11:23 < nickjohnson> Which is at most 4 boards tall. My board can't have others stacked on top of it. Therefore, nothing else can have address 4. May 21 09:27:05 i2c adresses are also not meant to represent the node number on the bus... May 21 09:27:37 koen: What part of that are you claiming is bullshit? May 21 09:27:50 the parts I quoted May 21 09:28:39 * nickjohnson sighs May 21 09:28:53 SRM 6.2: "Up to four Capes can be stacked onto the BeagleBone." May 21 09:29:01 unless you know more about the beaglebone eeproms that I do May 21 09:29:16 SRM 6.2: "Each board will be equipped with a 2 position dipswitch to set the address of the board based on the stack position." May 21 09:29:35 there isnt actually a limit of 4 though is there? May 21 09:29:39 The only other claim I make in those statements is "My board can't have others stacked on top of it' May 21 09:29:56 XorA: we spitballed the number based on a rough guess of pin loading May 21 09:30:12 koen: aye, but someone will make the bus booster cape :-D May 21 09:30:19 nickjohnson: where does it say address for must be on top? May 21 09:30:21 * XorA looks at prpplague ;-) May 21 09:30:46 koen: Okay, so what's the addressing scheme? May 21 09:30:59 Because it says it's based on stack position May 21 09:31:14 koen: but seriously your cape might be I2C only at which point there are no loading issues and you can keep stacking them up May 21 09:31:53 What situation are you envisaging there's a conflict? Perhaps there's a cape out there that has to be on the bottom of the stack, and thinks the bottom is numbered 4? May 21 09:32:02 Because I find that possibility difficult to credit. May 21 09:34:01 nickjohnson: I repeat: the address has nothing to do with the position May 21 09:34:06 * av500 has a large cape-stcking hammer in his toolbox May 21 09:34:08 +a May 21 09:34:19 So you're saying the SRM is incorrect? May 21 09:34:20 av500: not +i? May 21 09:34:40 nickjohnson: I'm saying that "different depending on the order in which it is stacked onto the main board." is misleading May 21 09:34:56 nickjohnson: and even if it wasn't, nothing says that 0x56 should be the top cape May 21 09:35:27 Who said "different depending on the order in which it is stacked onto the main board"? May 21 09:35:40 the SRM May 21 09:35:48 The SRM _explicitly_ says "Each board will be equipped with a 2 position dipswitch to set the address of the board based on the stack position." May 21 09:35:59 so do that May 21 09:36:03 I don't know how it can be any clearer than "based on the stack position" May 21 09:36:15 * XorA is amused about someone arguing with the manufacturer over what they really meant May 21 09:36:28 XorA: koen is just a student.... May 21 09:36:51 Further, the _only_ way a board can have a guarantee of position is if it's a top board, and there can be at most one of those - so realistically, a top board ought to be able to hardcode _any_ of the four valid IDs without conflict May 21 09:37:26 XorA: When the SRM says one thing and koen says something that directly conflicts with a plain reading of it, yes. May 21 09:37:47 If the documentation is being quoted, and it doesn't match what the manufacturer says, then I guess the manufacturer needs to improve the documentation :P May 21 09:38:47 fiola: where does it say that the top one needs to the the highest ID? May 21 09:38:56 "based on the order in the stack" May 21 09:39:15 top one is 1, the one below that 2, etc May 21 09:39:15 Speaking of which, is the documentation on a wiki so that the community can improve it? May 21 09:39:25 nickjohnson: nothing you have quoted has convinced me that you dont need to obey the other rules for cape design, I know your trying to save your 2cents, but is it really worth it? May 21 09:39:29 koen: In which case, my proposed plan will still be fine May 21 09:39:30 fiola: you can email improvements to gerald May 21 09:39:45 koen: and he will update the wiki? May 21 09:40:00 av500: yes, he uses the ms office wiki suite May 21 09:40:00 XorA: It's more about space than cost. And I don't see any point in adding configurability when I can't see any reason it's needed here May 21 09:40:18 koen: How can we email improvements if the sole authority on what's correct appears to be what's in your head? May 21 09:40:44 nickjohnson: your reading what you want in those quotes, not what is actually written, these are different things May 21 09:41:14 so far no-one has quoted a section which defines which cape gets which address, just that they get different addresses which are unique May 21 09:41:29 XorA: And which are based - in some way - on stack position. May 21 09:42:21 So let me make a simple assertion: There can be at most one top cape per beaglebone. That cape can pick an ID - any ID, but the largest one makes the most sense - statically, and all non-top capes can be configured to use different IDs May 21 09:42:28 Guys and gals, it's not something to argue about. Since it seems to matter, why not just request that the address assignment algorithm be documented? May 21 09:43:02 fiola: the dipswitches are mandated, the docs say the user is responsible for setting it May 21 09:43:06 fiola: what more do you need? May 21 09:43:54 koen: If everyone followed the rule I just proposed, under what circumstances would a conflict occur? May 21 09:43:56 koen: I don't need anything, yet, early days for me. But I see you experts arguing, and that's silly. If there's doubt, improve docs. May 21 09:44:48 nickjohnson: the fact that you're here arguing that only you don't need an eeprom because you're so special May 21 09:45:00 koen: Um, no, I'm equipping it with an eeprom May 21 09:45:06 s/eeprom/dips/ May 21 09:45:24 Okay, so present me with an argument why it's useful to have them, other than "I say so" May 21 09:45:40 I'm not saying _I'm_ special, I'm saying top capes are special, because there's only one of them May 21 09:45:42 the srm says so May 21 09:45:55 11:35 < nickjohnson> The SRM _explicitly_ says "Each board will be equipped with a 2 position dipswitch to set the address of the board based on the stack position." May 21 09:46:03 you even quoted it May 21 09:46:20 Yes, and the premise of my original question was "is this necessary in the case of a top cape" May 21 09:46:32 yes May 21 09:46:32 Does anyone know if SPI_D0 pin is MOSI or MISO on the beaglebone? May 21 09:46:38 lpi: it can be both May 21 09:46:41 And I'm simply asking, why would it be necessary? May 21 09:46:47 And nobody's provided an explanation why it would be May 21 09:46:53 lpi: wait for mdp to wake up to tell you which one to use :) May 21 09:47:28 koen: lol why him? There is not default config for these ports? May 21 09:47:34 pins? May 21 09:47:56 nickjohnson: have a look at http://boardzoo.com/product_info.php?products_id=87 May 21 09:48:08 nickjohnson: and then define "top cape" for me May 21 09:48:20 lpi: there is, but I forget what the default is :) May 21 09:49:25 koen: I found that, but I don't like to follow advice from forums when it's about I/O ^^ May 21 09:49:55 koen: Thank you. Is there a reason you didn't simply mention the existence of something like that initially? May 21 09:50:46 Sorry, I forget to give the link ^^ http://www.digipedia.pl/usenet/thread/13017/25151/ May 21 09:50:54 (monday morning) May 21 09:51:28 lpi: that's a copy of the mailinglist disguised as a forum :) May 21 09:52:02 Ok, so should be good advice then May 21 09:52:21 Cute breakout cape. I think I'll get one of those. May 21 10:02:51 jsabeaudry_: you'll like 11m: it can configure mux based on eeprom contents May 21 10:03:09 jsabeaudry_: it's not automatic yet, we need to check the eeproms in the wild first May 21 10:09:49 oh joy, more musb patches May 21 10:17:12 Is /etc/network/interfaces the actual network config file on the sd card, or is it somewhere else? Because the netmask in that file doesn't match the netmask on the usb0 interface. May 21 10:18:05 File says /24, while interface is /30 May 21 10:19:38 fiola: I think its not used any more May 21 10:19:53 The file not used anymore? May 21 10:19:54 depends on what your are running May 21 10:20:20 fiola: the new systemd images use connmann and that does not use /etc/network/interfaces May 21 10:20:25 since that would be to easy May 21 10:20:37 the new images don't have /etc/network/interfaces either :) May 21 10:20:48 I haven't changed anything yet from what came with the beaglebone a few days ago May 21 10:21:08 kk thanks, I'll look for this connman thing May 21 10:22:44 fiola: learn python... May 21 10:23:27 Already know it. :-) May 21 10:23:53 good May 21 10:23:54 then learn the dbus-api for python together with the connman-stuff May 21 10:24:06 it's as ugly as the stuff frombluez May 21 10:24:17 Aha, found /etc/init.d/connman, I can drill down from there, cool May 21 10:25:26 Well, I've seen enough *nix systems in 40 years to find them all ugly, or all lovely, depending on what I've had for breakfast :P May 21 10:27:05 some people use linux since 20a May 21 10:27:21 Oh, only 33 years, bah. Feels more like 50 :P May 21 10:52:00 Before I go mad - to set GPIO1_6 high/low on a beaglebone should I just be able to clear the relevant bit in GPIO_OE (reg_value &= ~(1 << 6);) and then set the relevant bit in GPIO_SETDATAOUT and GPIO_CLEARDATAOUT (reg_value |= 1 << 6;) or am I doing something dumb? May 21 10:52:47 also pinmux May 21 10:56:54 av500, Is that in the TRM? May 21 10:57:04 sure May 21 10:57:23 PRCM section or so May 21 10:58:13 gpio_set_value()? May 21 10:59:02 koen, I'm using that in somewhere different but want to cover mmapping things as well May 21 11:04:09 Is it obligated to recompile the kernel to activate the SPI? May 21 11:04:33 fierman, are you still having USB disconnect issues with your xM? May 21 11:06:53 * mdp wakes up May 21 11:07:00 lpi,did you figure it out? May 21 11:07:25 mdp: yes thank you :) May 21 11:07:35 ok :) May 21 11:07:57 for those who don't know: SPI_D0 is MISO and SPI_D1 is MOSI May 21 11:08:05 av500, Couldn't find anything relevant under the PRCM section, there's something under Control Module but that seems pretty dirty - I would've thought there was a mode register if anything for it? May 21 11:08:19 lpi, the situation with the mcspi driver configuring the opposite of the use case example in the TRM has made this a FAQ May 21 11:12:43 Sorry, I didn't understand your last sentence, thanks to my poor english :s May 21 11:13:15 Could you rephrase it please? May 21 11:15:26 mdp: ping! May 21 11:15:58 lpi, there's a diagram in the TRM showing D0 used as MOSI…it's misleading when people don't understand that it's programmable May 21 11:16:06 in any case, you've got it now :) May 21 11:17:34 mdp: ok thx :) May 21 11:19:35 Need to recompile the kernel, any link to download latest kernel sources? May 21 11:20:23 lpi: github.com/beagleboard/kernel May 21 11:20:33 thank you very much May 21 11:24:18 koen: it's a script that connect to GIT, and I can't use git from my laptop here: my university block the GIT port on routers -_- May 21 11:26:21 lpi: try https://github.com/koenkooi/linux/tree/linux-ti33x-psp-3.2.16-r11m+gitr720e07b4c1f687b61b147b31c698cb6816d72f01 May 21 11:27:00 lpi, can you just edit the script to force git to use https? May 21 11:28:36 you could, but the http transport on the arago server is broken May 21 11:28:55 I tried it an hour ago, no luck May 21 11:31:59 The MUX settings for the beaglebone should be set as per the expansion header P8 pinout right? So I shouldn't need to mess with the mux settings to get GPIO1_6 to drive an output? May 21 11:32:39 try and find out May 21 12:00:07 Does everyone run Angstrom on their BeagleB*'s, or is there a pile of ARM distros being used? May 21 12:00:37 oO( pilebuntu ) May 21 12:00:44 Hahaha May 21 12:01:29 Seriously? A lot of people use Ubuntu? It's not the most lightweight, aweful load of bloat there May 21 12:02:05 there are people who use it. May 21 12:02:11 kk May 21 12:05:01 All my Intel machines run Gentoo, but no way can I run that on the beaglebone. It would take weeks to update, and the SD card probably wouldn't last a month from all the rewrites. May 21 12:05:20 fiola: aholler and mru run gentoo on beagle* May 21 12:05:27 Wow! May 21 12:05:33 fiola: gentoo-prefix should be able to crossbuild stuff May 21 12:05:36 cross-compile May 21 12:05:43 no one in their right mind does native builds May 21 12:05:45 gentoo-prefix is something else May 21 12:05:46 Ah yeah, that's different May 21 12:05:47 and its not like angstrom is *that* far away from gentoo May 21 12:06:34 i build native, takes a week or such if you want kde. without you just have to unpack a stage3 and you are done May 21 12:06:35 ogra_: Angstrom newbie here, just had the beaglebone for a few days May 21 12:06:55 aholler: interesting May 21 12:07:48 aholler: I assume you try to limit the emerge --sync to once a fortnight or so, and NEVER do a world update? May 21 12:08:34 i just sync when I think I should do and never do a world update. May 21 12:08:40 * fiola nods May 21 12:09:33 Well I won't be doing anything like that any time soon. Need to get my feet on this board first. Loving it so far, very cute :-) May 21 12:10:19 Haven't even done blinkenlights yet, haha May 21 12:10:30 well, being the ubuntu guy here ... if you want quick results ubuntu-server is surely a good start ... for constant use (unless you want to develop for ubuntu) i would rather recommend angstrom though May 21 12:12:01 Interesting. Don't know ubuntu server. I ran ubuntu desktop for 9 months back when the fuss started, just to see what everyone was on about. Didn't enjoy the experience. It's hard to use anything else after Gentoo. May 21 12:12:50 You get used to the extreme flexibility May 21 12:13:02 well, ubuntu desktop feels a lot better with a dual core 1GHz and at least 1G ram May 21 12:13:12 Haha, I bet :P May 21 12:13:20 its really not made for a 512M single core machine May 21 12:13:25 as every desktop May 21 12:13:31 it runs but isnt fun May 21 12:13:34 yeah May 21 12:14:20 bone has 256MB ram May 21 12:14:24 I don't use "desktops" anyway. I used a good window manager and an iconic launcher. Everything else just gets in the way of getting stuff done, for me. May 21 12:14:42 but native compile times are overrated. e.g. u-boot builds in 3 minutes on a beagleboard c4 (720mhz) May 21 12:14:49 ++ May 21 12:16:20 yes, I rebuild uboot all day long May 21 12:16:29 * fiola chuckles May 21 12:16:32 as everything May 21 12:16:32 right after hello_world.c May 21 12:16:57 just cross-compilers are building stuff all day long May 21 12:17:01 to have an advantage May 21 12:17:53 It's actually all a big bad joke we're playing on ourselves, and we don't even realize it. :-) May 21 12:18:20 what is? May 21 12:18:50 It used to have a name: The Software Crisis. But people stopped referring to that when the crisis became the norm in computing :P May 21 12:21:35 now we just call it "fucked up" May 21 12:21:51 And "cost overrrun" May 21 12:22:02 And "oh it's just a bug" May 21 12:22:12 and systemd May 21 12:22:13 av500: snafu May 21 12:22:19 mru: hahahah May 21 12:23:19 ordering some level-shifters needs longer than native compiling a system on a beagle ;) May 21 12:24:01 use cross-shifters May 21 12:25:29 -> level compilers. May 21 12:25:52 Well time to go to bed. Not sure what I'm doing going to bed just after midday, but I guess it's the Beagle effect. :P 'night night :-) May 21 12:57:33 koen: your script return an error when trying to ls a folder: /home/lpi/fprint-pro/kernel/patches/dirs/*.patch: May 21 12:57:41 is it important? May 21 13:08:13 I'm lost concerning SPI with beagle bone, is someone can give me a link that explain how to activate SPI? I saw something like 4 how-to, all different May 21 13:17:50 lpi: spi or spidev? May 21 13:17:50 lpi: spidev2.0 is already active by default May 21 13:18:28 koen: I have nothing in /dev, no spidev device :s May 21 13:20:19 maybe there is a kernel module to plug?? May 21 13:22:18 modprobe spidev May 21 13:22:29 provided you have a recent kernel May 21 13:27:34 koen: ok, module loaded, checked with lsmod, but no spi device into /dev always May 21 13:33:24 and you have a recent 3.2.16 kernel running? May 21 13:35:45 koen: Linux beaglebone 3.2.16 #1 Wed May 9 11:28:21 CEST 2012 armv7l GNU/Linux May 21 13:36:55 branded with cloud9 image May 21 13:37:19 what does 'dmesg | grep -i cape' say? May 21 13:37:26 (pastebin it, don't paste it) May 21 13:37:35 cape of good hope May 21 13:39:28 http://pastebin.com/yekzCkDY May 21 13:41:03 lpi: and 'opkg update ; opkg install kernel-image-3.2.16' ? May 21 13:54:17 koen: no internet on the board, trying to make my host as gateway, but seems complex in my situation May 21 13:58:48 have you an alternative? May 21 14:00:25 copy http://feeds.angstrom-distribution.org/feeds/v2012.05/ipk/eglibc/armv7a/machine/beaglebone/kernel-image-3.2.16_3.2.16-r11n+gitr720e07b4c1f687b61b147b31c698cb6816d72f01_beaglebone.ipk to you bone and install it May 21 14:07:47 koen, you're the men May 21 14:07:54 I can see spidev 2.0 now May 21 14:08:12 but there are 2 spi on beaglebone aren't they? May 21 14:35:00 I need to know if there is a device out there that can be used with the BeagleBoard xM that can send data through cellular communication like CDMA May 21 14:41:47 koen, all is working fine, thank you very much May 21 14:45:12 koen: was the version number configuration for fb-test acceptable? May 21 14:45:43 koen: as you know i don't do a lot of user space stuff, so that method was what i had notes on May 21 14:45:57 prpplague: it works for me May 21 14:46:04 dandy May 21 14:46:51 just fyi, for anyone interested, i am still looking for a maintainer/developer for the arduino-for-userspace-linux, ping me at danders@elinux.org if you are interested May 21 14:51:36 prpplague: 1) in tears because of the 40 pin IDE cable in the box May 21 14:51:48 prpplague: 2) for non-ES, do I need to solder the resistors? May 21 14:51:55 3) is there an out of box SW? May 21 14:52:08 ah, tct display thing? May 21 14:52:33 av500: 1) hehe whats wrong with the 40 pin ide? May 21 14:52:42 ugly May 21 14:52:48 and ugly May 21 14:52:51 und HUGE May 21 14:52:52 av500: 2) yes you need to solder bridge the resistors, hehe May 21 14:53:01 np, I have ES May 21 14:53:05 and I'm wrong channel :) May 21 14:53:17 av500: it's for beagle as well, so it is fine May 21 14:53:32 yes, the IDE cable is ugly for both May 21 14:53:38 av500: 3) ubuntu works with it no problems, just need to pass a boot arg May 21 14:53:49 12.04? May 21 14:53:53 or 11.04? May 21 14:53:56 er, 11.10 May 21 14:54:02 av500: i haven't tested with 12.04 yet May 21 14:54:06 ok, so 11.10 May 21 14:54:07 av500: 10 and 11 work fine May 21 14:54:36 av500: there were a number of reasons we went with the 40-pin May 21 14:55:28 prpplague: you should hire me as a consultant :) May 21 14:55:34 av500: we wanted people to be able to solder their own connections, we wanted to be able to use multiple specific lvds connectors, and we wanted to be able to be able to easily interface with debug or receive boards with lvds receiver May 21 14:55:44 av500: hehe May 21 14:56:31 av500: just fyi, we designed the board to meet a specific customer specifications for a P.O. just so happens after we were done we decided to sell it retail May 21 14:57:03 av500: we were hope'n it would sell better than it has May 21 14:57:10 av500: on the retail end May 21 14:59:26 av500: suggestions are welcome May 21 15:00:36 mini USB ! May 21 15:00:49 yesterdays technology of tomorrow :) May 21 15:01:25 the future of retro computing! May 21 15:01:54 hehe May 21 15:03:09 or retribution of feature competing May 21 15:03:38 the rapture of featuritis May 21 15:04:01 hehe May 21 15:20:26 prpplague: ping May 21 15:21:43 is it possible to use interrupt with gpio? there is no real tutorial about it for the beagle May 21 15:22:00 sure May 21 15:22:07 gpios can generate interrupts May 21 15:31:12 av500: yeah, but using userspace fs? May 21 15:32:05 as I understand it, userspace can sleep on an IRQ May 21 15:33:39 userspace is overrated May 21 17:55:12 Hello everybody!! May 21 17:55:54 I'm having terrible days struggling to change or take rid of the "beagleboard.org" logo. May 21 17:56:12 I have a beagleboard xm rev C running Angstrom. May 21 17:56:23 psplash just doesn't work............ May 21 17:56:32 Any idea?? May 21 17:58:19 I need help, I'm going crazy! May 21 17:59:13 we're all crazy here May 21 18:00:32 Crazy enough to help me to solve this SH*T!?! May 21 18:09:40 Nobody knows? May 21 18:10:30 steak: try drivers/video/logo May 21 18:10:32 build a kernel without logo May 21 18:10:38 or that :) May 21 18:11:12 * mdp is strangely thinking about ribeyes right now May 21 18:11:28 heh, me too May 21 18:11:48 drivers/video/logo ? May 21 18:12:22 its in the kernel src but go with av500 's suggestion, just disable it (make config) May 21 18:12:31 mag, it is a powerful suggestion, I'm going to lobby for a change to the dinner menu :) May 21 18:12:54 LOL (bye bye brats :) May 21 18:13:36 You know what? I can't build the kernel for the beagleboard. May 21 18:13:43 How do I do this? May 21 18:14:04 * mag looks for a webpage May 21 18:14:21 I have Beagleboard XM / Angstrom May 21 18:15:06 when I am looking at patches the filepaths starts with "a/arch..." or "b/arch..." where do the a and the b come from? May 21 18:15:19 git May 21 18:16:50 serekson: well a/ is the root of the git project May 21 18:16:58 git repository May 21 18:17:03 linux in your case May 21 18:17:26 so short for linux_original/arch... and linux_modified/arch... May 21 18:17:27 steak: just google "how to build a linux kernel" or the angstrom site probably has instructions which would be better for you May 21 18:17:42 you'll need to download angstrom's toolchain as well May 21 18:17:50 those instructions are unclear May 21 18:17:51 there should be instructions on their site May 21 18:17:59 I have the toolchain working perfectly May 21 18:18:16 angstrom ARM toolchain? May 21 18:18:20 yes May 21 18:18:23 k May 21 18:18:32 let find their instructions and see if i can help May 21 18:18:42 I have a development machine with ubuntu May 21 18:18:53 wuth angstrom arm toolchain May 21 18:19:09 mag: Lovely May 21 18:19:39 steak: have you cloned the kernel? May 21 18:19:41 do you know git? May 21 18:19:48 I know git May 21 18:20:05 steak… read this and try… http://www.angstrom-distribution.org/building-angstrom May 21 18:20:08 have you cloned the kernel? May 21 18:20:16 I haven't cloned the kernel but I have a comlete image of the SD card on my HDD May 21 18:20:51 k. i don't recall what's on the image but it probably doesn't have the kernel src May 21 18:20:57 Do I really have to rebuild all the kernel for one freaking tiny image?? :((( May 21 18:21:05 i'm the wrong person to be helping. i haven't played w/ angstrom much May 21 18:21:10 It doesn't have the src. May 21 18:21:35 oh right, everything is OE. heehee May 21 18:21:44 steak… see the bottom of this page to get src, clone tree, and build: http://downloads.angstrom-distribution.org/demo/beaglebone/ May 21 18:21:47 better follow the instructions on http://www.angstrom-distribution.org/building-angstrom May 21 18:22:11 opps… looking for beagleboard May 21 18:22:11 kkeller: thanks, I tried that and it ends with a makeinfo not found and when I try to install I have no repos for this. May 21 18:22:34 yez :) May 21 18:22:39 then you need to ask the angstrom irc channel May 21 18:23:38 that's a good idea May 21 18:23:57 steak: what kkeller said May 21 18:24:16 I found someone saying that installing texinfo solves the makeinfo issue May 21 18:24:19 I'll try May 21 18:25:12 steak there are a lot of beagleboard google group messages on how to build the angstrom kernel… maybe one of those issues are the same as yours May 21 18:26:08 steak, I don't know the packaging split here, but yes, makeinfo/texinfo are inherently tied together May 21 18:26:44 considering as they are in the same upstream project May 21 18:27:30 I saw all of the Google groups things but they are mostly outdated and not aligned with the XM edition of the beagleboard May 21 18:27:43 I had to cut them off my searches. May 21 18:28:15 did you see the one titled: Building Angstrom with alternative Linux boot logo May 21 18:29:13 No I didn't!! Thank yoU! this may help May 21 18:29:31 Right now I succeded launching a bitbake May 21 18:30:01 gave this command: MACHINE=beagleboard ./oebb.sh config beagleboard May 21 18:30:12 then this: MACHINE=beagleboard ./oebb.sh update May 21 18:30:40 and then this: MACHINE=beagleboard ./oebb.sh bitbake u-boot -c configure May 21 18:31:54 But... once it's built??? May 21 18:33:42 is it safe to branch off of someone elses git repository? May 21 18:33:52 as in am I going to somehow annoy them if I do May 21 18:33:59 It's compiling loads of packages. May 21 18:35:39 * mranostay pops in May 21 18:36:07 * mranostay bangs head May 21 18:36:42 * mranostay digs around for his troll hat May 21 18:37:03 steak: it will go for a long time, likely. it may have behooved you to add something like: May 21 18:37:19 BB_NUMBER_THREADS = "n" May 21 18:37:19 PARALLEL_MAKE = "-j n" May 21 18:37:29 where 'n' is > than the number of cores on your build machine May 21 18:38:11 add that to your conf/local.conf file, I mean May 21 18:39:24 serekson: You mean making a branch in a local clone of an upstream git repo, right? May 21 18:39:32 (AOT, pushing a branch upstream) May 21 18:39:48 if so, nobody will ever know about your branch except you (unless you push it) May 21 18:40:06 mranostay: http://i1.cpcache.com/product/95602871/troll_pridebr_trucker_hat.jpg May 21 18:40:56 mag: thanks, I was pretty sure that was the case but wanted to make sure May 21 18:41:03 np May 21 18:41:06 google image search for "troll hat" also returned this: http://farm3.staticflickr.com/2565/3913377492_8b17aecf48_z.jpg May 21 18:42:04 mru: ok nfc what that is May 21 18:42:14 trollhatten nice May 21 18:42:27 that's a road sign in sweden May 21 18:42:33 Trollhättan is a city May 21 18:42:40 it literally means "the troll hat" May 21 18:43:09 awesome++ May 21 18:43:15 heh May 21 18:43:37 * mranostay knows where he is going next summer May 21 18:45:26 bb/oe May 21 18:45:35 oops May 21 18:45:42 mru: we need a troll conference May 21 19:12:14 (how) can I convince the sgx demos to use /dev/fb1 instead of /dev/fb0? May 21 19:12:25 * tasslehoff struggles with framebuffers these days May 21 19:26:01 mranostay, you should start a technology company with troll in the name... May 21 19:26:38 er i think that has been done May 21 19:26:40 Hello everyone May 21 19:27:16 mranostay, what?!? May 21 19:27:27 * mdp goes back to visionary creative mode May 21 19:27:38 someone knows the needed magic to start usb (ehci) inside a 0map-kernel 3.4? May 21 19:28:05 0map? May 21 19:28:20 does someone know the right way to gracefully halt a beaglebone (running on Angstrom) through a pin (maybe a GPIO pin): driver/other? May 21 19:29:48 guanyu, gpio in interrupt mode…use poll() on the userspace gpio api for that pin…execute shutdown on that event May 21 19:30:12 guanyu, that assumes you meant from an external interrupt/event on the gpio May 21 19:31:20 guanyu: just do it like the user-button. make your gpio a key May 21 19:32:20 mdp: yes, it's from an external event. I had thought about such things, but thought there were already mechanisms (or tools?) properly handling that. May 21 19:32:35 i'm sure google have some suggestions :) May 21 19:32:37 *has May 21 19:32:57 guanyu: aholler points out a better way May 21 19:33:28 The trick is that we should notify the beaglebone to stop, then wait for *some* time and cut down the power of the board. May 21 19:33:38 wait? May 21 19:34:06 aholler: using gpio_keys driver, and then polling a /dev/input/by_path/*, right? May 21 19:34:51 guanyu: no need for polling, just assign it the right button May 21 19:35:50 aholler: like a power-off event? Interesting... May 21 19:36:49 actually cutting power is another problem May 21 19:37:42 I don't know if it's possible to instruct the tps to do so. It should, but ... May 21 19:38:18 well, from your application perspective in userspace, you don't have to "poll"…you can blocking poll/read May 21 19:38:29 I thought he said graceful halt? May 21 19:39:12 yes, he did May 21 19:39:18 if you can instruct the tps to do so, you could do that in the kernels power down stuff May 21 19:39:45 hrm? May 21 19:39:59 umounting filesystems…etc. etc.? May 21 19:40:15 just like every pc does it May 21 19:40:19 aholler: I see. Looking into the TPS is maybe a good idea. May 21 19:40:33 yes, well, if the kernel doesn't have the power down hooks implemented then it's a broken kernel May 21 19:40:43 not sure why you would whack the tps directly May 21 19:40:56 because he is the master for the power? May 21 19:41:12 isn't that the p in tps? May 21 19:41:19 should only be hitting that through the generic hooks May 21 19:41:41 * koen hassles aholler over the cover of his tps report May 21 19:41:51 heh May 21 19:42:35 but I think I once read through the stuff if it's possible, and I think I've to the conclusion "not on the beagle" ;) May 21 19:42:39 mdp/aholler: the trick is just to do a sort of "shutdown -h now" on a falling edge signal (on a gpio). Then, I can arrange some time and externally cut the power supply with a relay... May 21 19:43:04 guanyu: that should be easy May 21 19:43:26 It's a bit dirty (what if the beaglebone puts more time to halt as usually??), but well... May 21 19:44:09 just change the gpio from inside the kernel. the kernel knows when he wants a power down May 21 19:44:18 no delay necessary May 21 19:47:03 aholler: I didn't know (I'm just a beginner). How can I know that the kernel is ready to power down? Where should I look? May 21 19:48:38 the kernel knows itself. I'm talking about modifying the kernel. May 21 19:49:05 heh, nm, nothing in omap-land implements pm_power_off(), sadly enough May 21 19:49:14 aholler, you're correct…omap sucks May 21 19:50:17 anyway, back to ehci, someone has a patch for 3.4 to enable it? ;) May 21 19:51:20 Running Android on Rowboat beagle board and cannot find any documentation on how to use USB Devices with it - I have installed android ICS and trying to use the Host API, but it never shows my device (USB is wired so begal is for sure host) - Works fine on my XOOM May 21 19:52:23 aholler, I keep making the mistake of assuming anything works correctly :( also back to edma hacking May 21 19:55:27 mdp: at least the kernel boots ;) May 21 19:55:48 http://fpaste.org/voH1/ May 21 19:56:24 aholller, in a past employment life we used to say, "boots, serial, ethernet….ship it!" ;) May 21 19:56:57 now ethernet got replaced by mmc ;) May 21 19:57:13 hello there! could you please recommend a precompiled and free operating system that includes a modern web browser like firefox? all i need to do is play some simple media files in full screen - using the beagle. angstrom doesn't do the job May 21 19:57:51 <_av500_> why a browser to play media files? May 21 19:57:51 fedora, ubuntu, debian, arch, … May 21 19:58:42 aholler, from my understanding from those around me focused on fixing xM upstream..there's a heavy pile of usb/ehci releated fixes waiting to be pulled into 3.5..perhaps something will address 3.4 in there :) May 21 19:59:09 av500: it has to be an html page to simulate an infotainment system May 21 20:00:00 mdp: tell them about including cc: stable@... in their patches May 21 20:00:02 <_av500_> i have no idea which browser has full video support May 21 20:00:04 aholler, I find I can't adjust to mmc as the first root device to bringup..nfsroot is insanely more useful to me :) May 21 20:00:23 aholler, I think they all know…but probably should be doublechecked May 21 20:00:43 some stuff is from other folks and may need to be proxied to stable May 21 20:01:17 gregkh is happy if he gets patches ;) May 21 20:01:23 yep May 21 20:01:34 aholler: should really any of those distributions do it as long as they are made for the board? May 21 20:01:53 <_av500_> none is May 21 20:02:02 aholler, not ones with html in them…then…unhappy gregkh May 21 20:02:02 <_av500_> angstrom is closest to that May 21 20:03:20 JohnCope: JohnCope they all have support for arm, just getting a kernel might be a problem May 21 20:03:38 mhm May 21 20:04:12 but the beagle is supported by most of them out of the box May 21 20:07:26 just read that welte uses poky ;) May 21 20:09:46 aholler: his buildsystem guys bitches constantly about it May 21 20:09:55 aholler: they went with poky for the corporate appeal May 21 20:10:18 ah, I thought about that May 21 20:11:28 anyone able to point me in right direction? May 21 20:12:36 Greg__: what device? May 21 20:12:45 rowboat May 21 20:15:00 issue is: Running Android on Rowboat beagle board and cannot find any documentation on how to use USB Devices with it - I have installed android ICS and trying to use the Host API, but it never shows my device (USB is wired so begal is for sure host) - Works fine on my XOOM May 21 20:16:45 if someone needs a patch to become a kernel-hacker: git grep X-Loader ;) May 21 20:17:39 * fiola peeks happily at the beaglebone again for another day's fun May 21 20:18:14 aholler: re: omap usb, arm-soc/for-next may work better (its queued for 3.5 iiuc). May 21 20:19:00 mdp: I will stick to 3.2, that one works ;) May 21 20:19:20 * mdp points at mag and runs away May 21 20:19:21 s/mdp/mag/ May 21 20:19:21 All Ethernet-overUSB issues sorted out yesterday. Today's it's native Ethernet config day, and trying to find authoritative info in systemd etc. A lot of previous Angstrom info on the net seems to be obsolete. May 21 20:19:30 heh May 21 20:20:11 upps, sorry, higlighted (yellow) on white background isn't that readable ;) May 21 20:22:20 depends on the phase of moon which background color I use ;) May 21 20:27:50 Running Android on Rowboat beagle board and cannot find any documentation on how to use USB Devices with it - I have installed android ICS and trying to use the Host API, but it never shows my device (USB is wired so begal is for sure host) - Works fine on my XOOM May 21 20:27:57 any1 can help GREATLY appreciated May 21 20:29:50 Greg__: This isn't help (sorry), but a question. Isn't Android just Java/Dalvik middleware, whereas USB devices are just plain Linux so not really related? May 21 20:31:19 I have an Android tablet, but never messed with it at Linux level so this question may well be out of plain ignorance. May 21 20:38:38 <_av500_> Greg__: you might have to fill in some low level glue for the host api to work May 21 20:39:17 <_av500_> normally host means an otg host, so musb May 21 20:39:27 <_av500_> not sure what android does with the EHCI host at all May 21 20:45:05 ya i assume there should be SOMETHING low level I can do... I just can't find ANY documentation from begal or otherwise May 21 20:45:39 Yes, fiola... you are right, however I want to use the HOST API from android, so i do not have to rewrite all my drivers data for my multiple devices May 21 21:10:12 <_av500_> Greg__: there is no documentation about that May 21 21:10:21 <_av500_> the android source code is the doc May 21 21:14:06 it's not like rtfs is a new concept May 21 21:25:51 pata interface for the pandaboard? is this 2000? :) May 21 21:37:14 mranostay, um, I think that was a joke…uses pata ribbon cable as a cheap way to extend it May 21 21:38:16 mranostay, ieee1284 interface would be cooler anyway ;) May 21 22:10:43 So in the beaglebone I typed in init 1 and it said going into safe mode or boot mode or somthing and now when I connect through the usb with linux ubuntu 11.10 terminal/shell it just gives me a blank screen .. what shall i do to get my angstrom login screen back? May 21 22:12:53 <_av500_> reboot? May 21 22:13:35 init 1 usually is without network, it's unlikely you want that. you could use the serial to change back to init-level 3 May 21 22:16:45 I've tried rebooting, and un plugging the usb and pressing the reset button(not all at the same time of course) let me see if typing any commands do anything May 21 22:17:02 Oh now it says resource busy when i try to connect May 21 22:31:41 hey guys, I'm using a clone board similar to the beaglebord-xm and I was wondering if it is possible to shortcircuit a board if I connected the touchscreen while the board was running since after I plugged in the screen, the board will not boot with any sd card sticks. May 21 22:32:10 indeed very possible May 21 22:32:39 damn it May 21 22:34:37 beagle clones? May 21 22:35:06 yeah May 21 22:35:06 arduinos May 21 22:35:11 or similar to beagle May 21 22:35:30 i wouldn't call that a clone May 21 22:46:02 clonebones May 21 22:47:10 <_av500_> clownbones May 21 22:47:42 funnybones May 21 22:47:50 cloned borgs May 21 22:49:04 it could be a wishbone May 21 22:49:37 assimilated bones May 21 22:50:21 are we done? May 21 22:50:50 mranostay: done pronounced to rhyme with bone? May 21 22:51:56 There was a similar discussion in Raspberry Pi forums, on "what is a clone". Basically there are no exact clones of anything, it's a fuzzy cloud of boards with overlapping niches. May 21 22:52:59 so you bought a bone instead of waiting for a r-pi? May 21 22:53:50 <_av500_> r-pi clones are verboten May 21 22:54:38 * djlewis prefers b-pi May 21 22:54:51 blackberries, yum May 21 22:55:39 Well when the Pi was announced, I actually expected TI to come back with a BeagleBite, even more cut-down version of BeagleBone. They can't be happy with Broadcom stealing the party after all. May 21 22:56:04 And the other ARM licensees too May 21 22:56:17 so are r-pi's shipping regularly now? May 21 22:56:37 fiola: bcm is stealing no party May 21 22:56:49 1) only a handful of those things have even shipped May 21 22:56:51 2) they are useless May 21 22:56:58 as people will soon realise May 21 22:56:59 Yeah, it's in continuous production now. But not ramped up too highly yet, under 100k/month May 21 22:57:44 I'm an engineer, not a fanboi/grrl of anything. So all boards have their good points and their bad points. May 21 22:58:14 binary blob boot loaders are always bad May 21 22:58:29 and excessive hyping too May 21 22:59:08 Aye, Pi is closed hardware. That doesn't sit well with me either. May 21 22:59:11 <_av500_> fiola: the goal of beagle and bone is to provide something that other people can base their work on May 21 22:59:24 <_av500_> as in buy the parts and build it May 21 22:59:25 fiola: it's worse than closed hardware May 21 22:59:43 it's closed hardware with closed software on top May 21 22:59:50 <_av500_> the goal for r-pi seems to get rid of old stock for Broadcom.... May 21 23:00:05 and it's not even doing that very well May 21 23:00:06 <_av500_> which you even cannot buy May 21 23:00:10 _av500_: I know, but the BGA doesn't really allow small outfits to clone it. All boards have their plus and their minus. May 21 23:00:13 <_av500_> and the PCB design is "closed" May 21 23:00:25 <_av500_> fiola: BGA is no issue these days May 21 23:00:37 you can do it in a toaster May 21 23:00:40 i can't solder them May 21 23:00:49 <_av500_> aholler: you are not a toaster May 21 23:00:54 think oven flow May 21 23:01:28 * djlewis is thinking on heading home :) May 21 23:01:37 I'll just repeat, all boards (and SoCs) have their plus and their minus. Engineers don't get into the fanboi thing, it's dull. May 21 23:03:17 oh, what a good believe in engineers May 21 23:04:42 aholler: you mistake the meaning of that statement. It means that alleged engineers who act as fanbois are not engineers, just fanbois. May 21 23:05:10 or paid May 21 23:05:14 heh May 21 23:05:27 It's certainly not an expression of faith in humanity. I lost that level of innocence decades ago :P May 21 23:05:43 i do like stuff that works maybe i'm not a real engineer then :) May 21 23:06:33 stuff that can be fixed is good enough for me May 21 23:06:49 No, if you like stuff that works, and are not swayed out of that by fluff and peergroup reinforcement, then I think you are a very good engineer. May 21 23:06:49 so you don't buy hw? May 21 23:07:00 ;) May 21 23:08:16 mru: well broadcom stuff is beyond that :) May 21 23:08:26 mranostay: I know all about that May 21 23:09:03 If IRC had moderation, I'd give aholler a +1 Insightful. It's true, we can't deny it ... hardware does not work 100%. It's our (hard) task to make it work most of the time though. May 21 23:11:31 somebody can help me with u-boot? Bought beaglebone recently and wonder how to boot it manualy... May 21 23:11:59 press space when it prints that May 21 23:12:09 or the anykey May 21 23:12:19 It never ceases to amaze (or maybe amuse) me that software CAN work 100% reliably, and yet software engineers seem incapable of achieving it for non-trivial projects. (Note that I lectured on software engineering, so I think I'm entitled to kick my own discipline :P) May 21 23:12:58 so you are responsible for those out there? May 21 23:13:03 well, software systems are some of the most complex things humans have ever built May 21 23:13:10 mru: yes May 21 23:13:17 far more complicated than, say, a car May 21 23:13:25 or an airplane (sans software) May 21 23:13:43 aholler: this one I found. But if I play with set env - and then fatload - it does not boot automatically. I tried to inspect omap3_beagle.h for default options, but chance... May 21 23:14:02 fatload only loads, use bootm May 21 23:14:42 N4ND: I'm a new beaglebone user too. What does "reboot manually" mean? Obviously I've power cycled, and I've "shutdown -r now" too. Anything else? May 21 23:14:45 aholler: yes, but then "Starting kernel" and "Uncompressing linux" appear and then nothing May 21 23:15:02 fiola: I mean to hit the key during boot May 21 23:15:28 I think there I have something wrong: setenv bootargs console=ttyS2,115200n8 mputrate=auto buddy=none vram=12M omapfb.mode=dvi:1024x768MR-16@60 omapdss.def_disp=dvi root=/dev/mmcblk0p2 rootfstype=ext3 rootwait run_hardware_tests quiet May 21 23:15:42 s/ttyS2/ttyO2/ May 21 23:15:42 tried many of these May 21 23:15:52 ah, ok. May 21 23:15:57 I will try May 21 23:16:28 kernels >.34 use 0map-serials May 21 23:16:34 s/usr/are using/ May 21 23:17:34 aholler: still the same - Uncompressing Linux... done, booting the kernel. May 21 23:17:55 and then nothing... May 21 23:18:02 quiet isn't that good to if you want to see output May 21 23:18:13 +o May 21 23:18:27 ah, ok. ;-) sounds logic... May 21 23:18:52 just - I saw it in uEnv.txt - so tried to be precise May 21 23:20:25 aholler: even without quiet - there is nothing new/added in output May 21 23:20:52 then check your kernel May 21 23:21:08 its original one May 21 23:21:59 aholler: as I failed to boot my distro, I am trying to boot original, shipped OS May 21 23:22:19 but even this one I can not manage May 21 23:23:14 via u-boot, typing commands... probably something I missed. Sometimes I am not sure if pages at the internet are related to beagleboard or beaglebone. May 21 23:28:57 fiola: did you manage to run different than original distro? May 21 23:30:11 N4ND: Not tried yet. I expect to stick with Angstrom all the way until (i) the end, or (ii) until it pisses me off for doing something really stupid. I'm still on track (i) at present. May 21 23:31:01 I hate it when people can't set up mail servers right ;_; May 21 23:31:22 seems you never used sendmail May 21 23:31:28 Mail servers are a nightmare. May 21 23:31:30 Hahahaha May 21 23:31:58 No they setup open relay May 21 23:32:32 fiola: hm, OK. I need to set up my own mediatomb server... This is why... May 21 23:32:57 Using a beaglebone? May 21 23:33:05 yes May 21 23:33:07 I spent many years configuring Sendmail MTAs, and actually love regular expressions and rule rewriting. But that doesn't mean that I considered Sendmail good. It's not inaccurately described as the MTA with an encrypted config file. May 21 23:33:19 https://www.youtube.com/watch?v=W8Q0PiWHF34 May 21 23:33:49 With a seperate usb drive i hope May 21 23:33:57 yes May 21 23:36:02 the needed tabs are inspiring May 21 23:36:10 my idea is to replace Nero Mediahome which is only supported now by new LG TV firmware (they branded it). Mean, my NAS is not visible anymore - and I resign to compile right mediatom on it (ZyNOS) and trying beagle May 21 23:38:06 but I stuck on the beginning - not knowing how to boot original mmc card ;-) May 21 23:38:11 my beaglebone isn't playing well with any wifi adapters I throw at it. dmesg makes it look okay http://pastie.org/3947323 (as far as I know -- linux networking is not my strength) but ifconfig and iwconfig show no interfaces May 21 23:38:39 I've tried this with an rt2800usb-based adapter and a belkin that's supposed to be pretty linux-friendly May 21 23:39:01 did you install the firmware? May 21 23:39:05 sbma44: you need a driver May 21 23:39:33 aholler: the firmware? afaik that one is installed & in place, but perhaps I have it in the wrong location May 21 23:39:47 firmware is needed by the driver May 21 23:40:03 but the driver doesn't come up, so you need a driver May 21 23:40:13 based on google, that is a zd1211 driver, so make sure you the modules/firmware installed for it.. May 21 23:40:14 ah May 21 23:41:09 sbma44: that output is just from the usb-subsystem. so at least that works ;) May 21 23:41:40 or appears like it works ;) May 21 23:42:23 I did have this working on a pre-PWM version of the kernel! but of course I needed PWM (and eventually had to give up on directly messing w/ mmap). but yeah, I think it works :) May 21 23:49:16 getting back to speed with where I left this -- I was at 3.2.14, looks like I had to downgrade to get PWM http://www.gigamegablog.com/2012/03/16/beaglebone-coding-101-buttons-and-pwm/ -- rebooting now May 21 23:49:46 this thread says to disable NetworkManager and this wifi adapter should magically start behaving. is that advice crazy? https://groups.google.com/d/msg/beagleboard/skMwVjaoZIo/Q6gnW8Dk-K8J May 21 23:51:22 actually nm, that's pretty old May 21 23:51:45 so when downgraded the kernel, you installed the downgraded zd1211 modules? May 21 23:51:58 (lsmod would show if they are loaded) May 21 23:52:55 rcn-ee: actually, this is another bit of weirdness: lsmod shows nothing. didn't with 3.2.14 either. May 21 23:53:27 rcn-ee: http://pastie.org/3947374 May 21 23:53:49 wonder if r-pi's are rational or not May 21 23:53:57 (apologies if I'm being really dumb about this) May 21 23:54:25 sbma44, https://github.com/Angstrom-distribution/meta-ti/blob/master/recipes-kernel/linux/linux-ti33x-psp-3.2/beaglebone/defconfig#L1288 May 21 23:55:15 the zd1211 driver is setup as a module, so if they aren't loaded (weither you don't have matching kernel modules to your boot kernel, or something else) its not going to work.. May 21 23:55:32 find /lib/modules -name 'zd*' May 21 23:56:18 i'm guessing he just copied the uImage, so it's looking at empty /lib/modules/`uname -r` May 21 23:56:39 aholler: thanks! looks like rcn-ee is right, I have modules for this adapter under 3.1.0+ but not the specific kernel May 21 23:56:59 rcn-ee: you're right, I just did an opkg update and then copied the uImage per the gigamegablog recipe May 21 23:57:09 for the downgrade, just copied the uImage May 21 23:58:40 yeah, read what they wrote.. other then getting pwm working, that instruction isn't to useful for anything else. ;) May 22 00:00:32 rcn-ee: got it. now opkg is misbehaving :P maybe I should just wipe out this SD card, it's not like there's anything that needs saving May 22 00:01:05 you could just find that image they pulled the uImage out of and copy the modules directory, then it would work.. May 22 00:05:17 rcn-ee: yeah, let me see if they've got an SD adapter around here -- I don't like this opkg error ("Failed to open //var/lib/opkg/info/perl-module-pluggable.pm.list"), I'm guessing it's coming from an interrupted opkg upgrade and it's going to be hard to get rid of May 22 00:05:27 rcn-ee: I think I should just nuke this thing if I can May 22 00:05:38 then come back and install the modules properly May 22 00:09:18 eh, maybe not. I guess the full-system angstrom images aren't really organized to make that easy. May 22 00:42:49 Just about done with the basic code for the alarm system. May 22 00:45:09 opkg upgrade = famous last command line of many a working system May 22 00:47:28 lol May 22 02:58:52 Anyone have any recommendations for running the bone on battery when the power goes out. **** ENDING LOGGING AT Tue May 22 02:59:58 2012