**** BEGIN LOGGING AT Mon Apr 02 02:59:58 2012 Apr 02 03:58:36 Patrick1238: google it? Apr 02 04:13:16 * mranostay pokes dwery Apr 02 05:06:00 [10:56] Hi, I would like to ask the guide to install TP-Link Wireless Adapter TL-WN723N on Beagleboard-xM Angstrom.. [10:57] I have this wiresless adapter but it couldn't work. The USB port could detect it but couldn't load up.. I guess it is due to the lack of driver. But I have absolutely no idea how to install it Apr 02 05:14:42 Need help, trying to install ubuntu through netinstall on a beaglebone and it asks for ubuntu archive mirror ...which is the one should I be pointing for a Bb Apr 02 06:09:53 Patrick1238, do you have the driver source for the wifi dongle? Apr 02 06:12:05 Patrick1238, the wireless adapter you have seems to have a Realtek 8192SU chipset. afaik, this driver was in 2.6.37 staging Apr 02 06:12:57 you can download the firmware from the vendor and put this into /lib/firmware Apr 02 06:51:59 Hmm Apr 02 06:52:11 How do I change which kernel my beaglebone has booted from? Apr 02 06:53:06 I don't see grub or lilo configuration files in /etc... and I upgraded my system, but it is booted off the old kernel. Apr 02 06:55:18 grub?if its not april fool joke....than beagleboard uses u-boot as boootloader Apr 02 06:55:39 It's not an April Fool's joke - I'm just not familiar with the beaglebone yet. :) Apr 02 06:55:45 Ok... that gives me some place to look. Thanks. Apr 02 06:58:59 Ok so the boot files are on the fat32 partition? Apr 02 06:59:27 Does anyone know if the updated kernel images get downloaded when using opkg? Apr 02 06:59:42 If so, where are they put? What's the "proper" way to upgrade a system and install the new kernel? Apr 02 06:59:51 Or is it black magic right now? :D Apr 02 07:01:16 Ah.. /boot Apr 02 07:04:24 the angstrom homepage should be enough Apr 02 07:04:30 for all info Apr 02 07:06:21 I'll look. Thanks. Apr 02 07:12:01 wiki is also there.....for linaro Apr 02 07:42:53 Hmm my modules.conf got overridden. That explains a few things. Does anyone know what the default modules.conf looks like? I can't find any mention of it online. Apr 02 07:46:41 capttwinky here? Apr 02 07:46:51 if anyone knows where to email him please help ? i much appreciate it Apr 02 07:48:56 hi all Apr 02 07:49:23 anybody online ? Apr 02 07:49:35 i have a question concerning my beaglebone Apr 02 07:49:56 i want to start a program on boot, but i can not find inittab on my beaglebone Apr 02 07:50:04 NAP: Don't know if I can help, but what's your question? Apr 02 07:50:14 Ah Apr 02 07:50:19 NAP: running what distribution you are? Apr 02 07:50:38 angstrom Apr 02 07:50:47 Try putting a script in /etc/init.d and setting it executable. Apr 02 07:50:52 lastest version from the website Apr 02 07:51:07 never mind found it. thanks all Apr 02 07:51:16 NAP: then you want to read up on systemd, because thats the current Way To Do It )TM) Apr 02 07:51:24 NAP: glad you got the help you need Apr 02 07:51:34 i was wondering if it is normal that there is no inittab or if that is a problem of my filesystem Apr 02 07:51:35 NAP: Can you put your stock /etc/modules.conf on pastie.org? Apr 02 07:51:51 I upgraded my system and lost my modules.conf :/ Apr 02 07:52:04 NAP: its perfectly normal that there is no inittab. Apr 02 07:52:28 @LeoThe2nd: aha, ok, thanx Apr 02 07:53:09 NAP: *hint* tabcompletion is cool and usually also works on IRC nicks :) Apr 02 07:53:13 i am no linux crack so i gonna google on systemd if thats the way to go Apr 02 07:53:44 i wonder that i found some tuts decribing to autostart a program on beaglebone using inittab ??? Apr 02 07:54:31 LetoThe2nd: you are right :-) tabcompletion really works ;-)) Apr 02 07:54:50 NAP: maybe see http://fedoraproject.org/wiki/Systemd Apr 02 07:55:10 hello Apr 02 07:55:13 NAP: thats why i asked about your "distribution". nobody said that there's no inittab ever. Apr 02 07:55:15 Ahhh success! Apr 02 07:55:22 You have to run update-modules. :D Apr 02 07:55:38 Yay. Magic. Apr 02 07:57:26 WiFiGuy: putting a cript in /etc/inittab does not autostart it or does it? Apr 02 07:57:45 You mean /etc/init.d? Apr 02 07:57:58 If you put a script in /etc/init.d and set +x on it, it should run on boot. Apr 02 07:58:15 e.g. cp /path/to/some/script /etc/init.d Apr 02 07:58:25 chmod 755 /etc/init.d/scriptname Apr 02 07:58:30 Whatever your script is... make sense? Apr 02 07:59:06 WiFiGuy: ugly, relying on the sysv compatibility constructions instead of doing it right first time. Apr 02 07:59:25 I'm oldschool :D I don't know the "right" way on Beaglebone Apr 02 07:59:32 I am new. :P Apr 02 07:59:37 its all ok for existing things, but when you do something new, why start with ways that are already deprected. Apr 02 07:59:46 I didn't know it was deprecated. Apr 02 07:59:47 :) Apr 02 07:59:51 WiFiGuy: i already said - systemd. Apr 02 07:59:57 I wasn't paying attention. :/ Apr 02 08:00:24 I'll read up on that myself. :) Thanks for the pointer. Apr 02 08:01:03 basically it should just be stealing a script for /lib/systemd/something, change it, systemd-enable it (or whatever the command is), done. Apr 02 08:01:52 see: http://patrakov.blogspot.de/2011/01/writing-systemd-service-files.html Apr 02 08:05:18 Thanks. :) Apr 02 08:05:26 I skimmed it, will read more later. Apr 02 08:05:29 Night all. Apr 02 08:27:17 LetoThe2nd: I tried the easy way with a script in init.d but that doesnt work. now i m gonna check the systemd thing Apr 02 08:37:41 still wondering why putting the script into init.d did not execute it on startup Apr 02 08:40:34 mranostay: back Apr 02 08:50:12 LetoThe2nd: are you still there? Apr 02 08:50:55 systemd seems to be a bit complex for me as a linux noob Apr 02 08:51:17 i just want to start a script an everey boot and the script is pritty simple ist only 3 lines Apr 02 08:51:40 NAP: google systemd and rc.local Apr 02 08:51:48 i tried it with init.d but that did not work. i hope you all don't mind if i post the 3 lines Apr 02 08:52:29 #!/bin/sh Apr 02 08:52:38 cd /home/root/proggen/automat Apr 02 08:53:03 thats all Apr 02 08:53:30 its only 2 lines Apr 02 08:53:48 i postet the 3. line 3 times now. it does not show up ? Apr 02 08:54:10 ## /usr/bin/screen -S AUTOMAT /usr/bin/mono /home/root/proggen/automat/automat/bin/automat.exe Apr 02 08:54:19 ok, remove the ## and thats the 3. line Apr 02 08:54:33 / is the irc command prefix Apr 02 08:54:47 yea, it juste realized that , too ;-)) Apr 02 08:55:36 if i start the script from shell it works fine Apr 02 08:56:51 if i put that script into init.d (with chmod +x ) it does not run the program Apr 02 08:57:05 on boot Apr 02 08:57:38 NAP: google systemd and rc.local Apr 02 08:58:39 av500: i am just doing that atm Apr 02 09:08:04 cd in a startup script that basically issues one command. sounds like a windows solution indeed. but then, there's mono involved... Apr 02 09:09:25 NAP: as av500 said, rc.local is another start point. but i personally guess that the thing that you try to run has some side-dependencies that you don't know/realize and there fore does not work on startup the way you expect it. Apr 02 09:10:08 NAP: so - try running a simple command on startup first, like "touch /tmp/dummy" and see if that works. then insert your thing. make things work one after the other. Apr 02 09:10:46 * av500 needs to have a coffee to be awake enough to make coffee Apr 02 09:11:26 av500: you need emergency bootstrap rations for that case Apr 02 09:11:52 statically linked[brewed], of course. Apr 02 09:11:56 dm8tbr: yeah, in fact I only have those, they are called nespresso capsules :) Apr 02 09:12:21 just tear of the lid, and chew away Apr 02 09:12:40 :D Apr 02 09:16:38 av500: i need 10 of them Apr 02 09:18:11 they come in 10-packs Apr 02 09:19:21 nice. thats what i need every monday Apr 02 09:24:05 systemctl enable rc-local.service returns >> Failed to issue method call: No such file or directory << Apr 02 09:26:18 NAP: full path to the file, maybe... Apr 02 09:27:06 LetoThe2nd: yes, you are right. i am a windows guy and a linux noob. the cd is really windows like ;-) and yes, its a mono program as i mostly program in c# on windows ;-) Apr 02 09:27:39 NAP: then lesson #1: try methodic debugging :) Apr 02 09:29:10 oh, it's bacuase i was really wondering thet autostarting a programm on linux is that complex ;-) not just a autostart.bat or a registry entry ;-) Apr 02 09:29:45 NAP: http://superuser.com/questions/278396/systemd-does-not-run-etc-rc-local Apr 02 09:30:08 NAP: 1) there is no registy, but if lennart has his way, we will have one soon Apr 02 09:30:29 NAP: well, thats the price we pay for having cool features like... true multiuser? true headless? true multiprocess? ;) Apr 02 09:30:33 2) there is no autostart.bat because lennart took it away Apr 02 09:30:42 LetoThe2nd: true bullshit Apr 02 09:31:12 av500: i've never heard of a multiprocess-enabled autoexec.bat ;) Apr 02 09:31:33 damn, I am in dire need of one Apr 02 09:32:00 NAP: in the past, it was easy to drop your stuff into rc.local Apr 02 09:32:11 then some people came along an deemed that un-linuxy Apr 02 09:32:19 so they are taking it away Apr 02 09:32:42 just come back in 2 ys, it will be like windows Apr 02 09:33:26 av500: yeah, thats exactly what i mean - you gain shiny new features, but add complexity. if you are fine with a 90s singleprocess sysv-init-one-after-the-other-startup, nobody keeps you from using that. but it's not the recommmended thing anymore. Apr 02 09:34:02 let'S go settle that "outside" Apr 02 09:34:05 all you'd have to do is switch back to busybox init instead of systemd ;) Apr 02 09:34:16 av500: drinking contest? ;) Apr 02 09:34:31 oh, don't get me wrong. linux is fine for me. and my mono program is working really well. so everithing is ok. its just that i was wondering why somthing simple like starting a program has to be so complex ;-) Apr 02 09:35:04 NAP: because you can same 100ms at boot doing it the new-shiny-complex way Apr 02 09:35:09 save* Apr 02 09:35:24 you could always pass your tool as init=, then you'd have the total autostart thing. Apr 02 09:35:43 and for all linux users take together, the total boot time can go down several seconds, that a huga gain..... Apr 02 09:35:57 damn, i cant tzpe today Apr 02 09:35:57 ;-) Apr 02 09:36:20 av500: hehe, i tkae it you don't like systemd. Apr 02 09:37:06 LetoThe2nd: I dont like the fact that NAP does not find a way to "enter your start script here" Apr 02 09:37:22 its all "the upstream will take care of that" Apr 02 09:38:57 av500: point taken. Apr 02 09:42:57 I want to use the DSP of beagleboard for Video compression. What is the best way to do it.... gst-dsp or gstreamer-ti..? Apr 02 09:44:16 anand_: good question Apr 02 09:45:52 gst-ti uses link, gst-dsp uses bridge Apr 02 09:46:15 if you are planning on using angstrom, then link is all there already Apr 02 09:46:57 i am using angstrom. Apr 02 09:47:22 thanks Apr 02 09:48:23 depending on your use case, you might also just use plain dsplink and CE Apr 02 09:48:34 and forego DMAI and gst-ti and gstreamer Apr 02 09:50:27 i want to stream the video over wireless ... so i think gstreamer is need. Apr 02 10:00:56 hello all, I am new to BB, just upgraded the Angstrom and now cloud9 scripts return error when I try to execute them (EAFNOSUPPORT, Address family not supported by protocol). Looked around but found nothing more than others having the issue and that it disappeared after a while... any idea? Apr 02 10:08:54 trust you tried a reboot after 'upgrading'? I'm not quite sure how an upgrade works, since I've only used debian Apr 02 10:09:11 but if libc changed, things like that can happen Apr 02 10:14:07 sure I did reboot Apr 02 10:14:49 cloud9 service came up again, but scripts now return errors, even the blinkled one. the bone101 is not up on port 80 either Apr 02 10:36:54 Hi everyone! Can anyone please tell me how to modify UARTFIFOConfig() function for UART1 instead of UART0 in AM335x Starterware Apr 02 10:39:25 Is there anybody alive?? Apr 02 10:40:24 Hi eric, can u answer my question? Apr 02 10:41:08 i guess am not dead Apr 02 10:41:47 OK...thank God, sundar is alive :) so plz say smth abt my ques? Apr 02 10:42:15 Can anyone please tell me how to modify UARTFIFOConfig() function for UART1 instead of UART0 in AM335x Starterware Apr 02 10:42:30 but sorry i don't have knowledge on that Apr 02 10:43:55 do not ask some particular programming question. better ask what you want to do and what you have tried so far Apr 02 10:43:58 I have modied uartEcho.c and related headers/Macros/other c files in /drivers, /includes for UART1. The project is successfully built but stucking in UARTFIFOConfig function when I run Apr 02 10:43:59 that might help Apr 02 10:45:15 OK Sundar, putting it simply, I just wanna run UART1 of beaglebone using starterware APIs Apr 02 10:46:42 haven't got idea on starterware. somebody else might know. so wait Apr 02 10:47:23 OK...no probs. Apr 02 10:48:31 Hi Anand! can u help me out Apr 02 10:50:16 The person "Anand_" has just joined the channel. he might not have the slightest idea about your problem. be patient Apr 02 10:51:11 Gloria: talking to every one joining the channel is just annoying and even a bit rude. ask again in a few hours if nobody answers, or write to the mailing list in the mean time. son't make our irc clients scroll pointlessly. Apr 02 11:18:34 still no sucess with my autostart stuff. i even tried a two liner in /etc/init.d with 1.line #!/bin/sh and 2. line /bin/touch /home/root/proggen/automat/testtest and no file is created. Apr 02 11:22:56 NAP: probably because /etc/init.d does not automatically get executed, at least not on systems that i know. Apr 02 11:24:08 NAP: FWIW, it's just a container for the scripts that get linked to in the respctive runlevel folders /etc/rcX.d, but thats debian knowledge, so no idea about android there. Apr 02 11:24:19 didnd av500 say axactly the oposite? Apr 02 11:24:54 NAP: don't think so, it was the wifi guy who suggested that. Apr 02 11:24:55 its not android. its the latest angstrom image for my beaglebone Apr 02 11:25:06 sorry, meant angstrom, not android. Apr 02 11:25:44 ok, maybe it was the wifi guy ;-) Apr 02 11:25:54 are the /etc/rcX.d folders there? Apr 02 11:26:32 yes rc0.d - rcS.d Apr 02 11:26:54 i mean 0 1 2 3 4 5 6 S Apr 02 11:26:57 NAP: then type "runlevel" to find out what runlevel you are in Apr 02 11:27:36 1 Apr 02 11:27:59 NAP: then place a link to your /etc/init.d/ script inside /etc/rc1.d Apr 02 11:28:50 NAP: and stick to the nomenclatore, e.g. make its name start with S99 for example if you want it to be started pretty late. Apr 02 11:29:06 i'll try that Apr 02 11:36:50 doesnt work Apr 02 11:37:50 but i wonder as all other stuff in rc1.d starts with a K Apr 02 11:39:26 that is such a primitive system Apr 02 11:40:00 NAP: S is Start, K is Kill Apr 02 11:42:17 i already suggested to go back to busybox init with its inittab and one /etc/rc.S directory Apr 02 11:42:29 that's the simple solution Apr 02 11:42:33 good for simple systems Apr 02 11:42:47 LetoThe2nd: i knew that. i was just wondering as rc1.d has K20samba but samba is running Apr 02 11:42:54 for more complex systems, something like openrc is much better Apr 02 11:43:09 NAP: check rc2.d or rc3.d Apr 02 11:43:27 NAP: honestly, i don't know where you might have gone wrong, but at the moment i can't spare the time to dive into the details. Apr 02 11:43:47 NAP: i think you already know the necessary starters, good luck Apr 02 11:43:51 LetoThe2nd: no problem. Apr 02 11:44:01 LetoThe2nd: thanx for your help anyway Apr 02 11:44:02 should i2c_smbus_read_word_data return data in host byte order? Apr 02 11:46:43 NAP: on a systemd system the stuff in init.d might not be used Apr 02 11:46:55 you have samba because there is a systemd module for it Apr 02 11:47:09 there is compat stuff like rc.local Apr 02 11:47:15 so go and make that work Apr 02 11:47:58 mru: looking at ne10 code, for e.g. a simple _add(), I see they long 2xQ regs, add and write back, so there is no benefit using more regs? Apr 02 11:48:04 they load* Apr 02 11:48:06 av500: i tried that without success so i went back to the init.d which did not work eigther. i guess i give the systemd stuff another try Apr 02 11:48:18 or write a simple systemd module Apr 02 11:48:56 * LetoThe2nd feels the wind of history passing by, repeating itself Apr 02 11:49:27 av500: simple and systemd are mutually exclusive Apr 02 11:49:41 av500: i followed the steps written here http://pavelmialko.livejournal.com/927.html Apr 02 11:49:57 NAP: tbh, I know nothing about systemd :) Apr 02 11:51:15 av500: what code are you looking at? Apr 02 11:52:15 mru: sec Apr 02 11:52:41 https://github.com/projectNe10/Ne10/blob/master/source/NE10_add.asm.s#L48 Apr 02 11:53:02 omg that's ugly Apr 02 11:53:47 of course Apr 02 11:53:56 corporate code... Apr 02 11:54:29 I wish they'd put as much effort into the code as into the header comments Apr 02 11:54:33 :) Apr 02 11:55:28 this is their another effort into neon after falied attempt with omx Apr 02 11:56:34 actually, that function looks like something gcc would output Apr 02 11:58:03 atleast this is gcc style.....omx refused to compile with gcc as that was written for armcc Apr 02 11:58:22 assuming you could even find it Apr 02 12:03:21 mru: so, looking at that code, would it benefit from using more regs? Apr 02 12:03:39 it would benefit from being written by a human rather than a compiler Apr 02 12:03:50 would a human use more regs? Apr 02 12:03:57 a human might actually vectorise it Apr 02 12:04:46 yep.intrinsics/api always confuses everyone.better to go with hand. Apr 02 12:05:24 that looks like they compiled some C code and added comments by hand Apr 02 12:06:20 mru: but aren't *you* somehow part of *them* :) Apr 02 12:06:43 no Apr 02 12:07:03 freemasons? stonecutters? Apr 02 12:07:43 can't comment on that Apr 02 12:08:30 cookie cutters Apr 02 12:08:41 pastegrammers Apr 02 12:12:25 g+ has issues with tagging :/ Apr 02 12:13:32 so you might not be a sandworm god emperor after all? Apr 02 12:14:26 mru: nah, whenever i want to tag you in the stonecutters pic, it says "Oops". Apr 02 12:17:04 hm, i think its better now. Apr 02 12:31:38 Hi, could someone help me? I have a beglebone here and I need to write some code to use serial ports, does anyone know how can I do? Please. Apr 02 12:33:15 douglasmsi: google does: http://www.gigamegablog.com/2012/01/22/beaglebone-coding-101-using-the-serial-and-analog-pins/ Apr 02 12:36:08 LetoThe2nd thanks! Apr 02 12:37:06 douglasmsi: here is another link http://www.nathandumont.com/node/250 Apr 02 12:39:27 NAP thanks Apr 02 13:55:49 hi Apr 02 13:56:23 what about real-time ability of the beagleboard? i.e. when ubuntu is used as OS.. Apr 02 13:56:50 ubuntu is not an rtos Apr 02 13:57:07 and "real-time ability" is meaning less Apr 02 13:57:21 mh ok sry i aksed the wrong question ;) Apr 02 13:57:39 dunno the right yet Apr 02 13:58:08 av500: not entirely meaningless, but not exactly well-defined Apr 02 13:59:04 ubuntu = desktop os Apr 02 13:59:37 you can't expect real time performance with ubuntu Apr 02 13:59:56 sundar: define "real time performance" Apr 02 14:00:02 of course you can do "real time" in ubuntu Apr 02 14:00:12 can you? Apr 02 14:00:25 it depends on what you do Apr 02 14:00:28 depends on the chip performance... Apr 02 14:00:35 yes, you can use it for a "real time tectonic shift monitor" Apr 02 14:00:37 not just the hardware Apr 02 14:00:48 mhm Apr 02 14:00:54 i'd say it rather depends on task versus performance. Apr 02 14:00:57 m0erk: define "real time" Apr 02 14:00:57 is there any rtos for the beagleboard? Apr 02 14:01:12 av500: I think you'll have a hard time giving upper bounds for any operation in ubuntu Apr 02 14:01:14 real time != false time Apr 02 14:01:22 av500 yes thats the point.. at the moment iam not quite shure about the definition :) Apr 02 14:01:24 m0erk: qnx comes to mind Apr 02 14:01:47 LetoThe2nd: real time says nothing about task vs performance Apr 02 14:01:57 you can have a fully loaded system that is still realtime Apr 02 14:02:46 av500: i know ;-) Apr 02 14:03:03 real time does not mean fast. Apr 02 14:03:03 but yes, if you order "real time" from dealextreme, you might get fake time.... Apr 02 14:03:18 i want to analyize an audio signal and react on that signal Apr 02 14:03:26 cheap chinese cloned time Apr 02 14:03:54 m0erk, what is the frequency of your input sound? Apr 02 14:03:54 m0erk: when I play loud rock music, my wife reacts immediately Apr 02 14:04:06 i dont know yet how fast we need do react Apr 02 14:04:11 that is real time! yay Apr 02 14:04:16 I need to duck pretty fast Apr 02 14:04:22 its like 80hz-44khz Apr 02 14:05:01 you can do that without an rtos Apr 02 14:05:08 yes probably Apr 02 14:05:21 very certainly. Apr 02 14:05:21 depends on what the requirements are Apr 02 14:05:27 yes Apr 02 14:05:29 depends Apr 02 14:05:33 if you can live with few underruns/overruns Apr 02 14:05:40 we want an active audio comensation :P Apr 02 14:05:42 and home runs Apr 02 14:05:48 mru: he did not say "react in a certain amount of time." - he just said "react". Apr 02 14:05:49 if you to guarantee a latency of less than 10ms, I would not use linux Apr 02 14:06:17 thats a statement, ok.. so what else? Apr 02 14:06:52 if missing the target once in a while is acceptable, linux will work fine Apr 02 14:06:56 maybe we can use the DSP to speed this up? Apr 02 14:07:02 mru: because the jiffies precision isn't good enough? Apr 02 14:07:15 because the linux scheduler isn't deterministic Apr 02 14:07:22 jiffies can be tuned Apr 02 14:07:27 jiffies has nothing to do with it Apr 02 14:07:32 you can have a sched every 1ms if needed Apr 02 14:07:49 you can make a jiffy a full second and still have low latency Apr 02 14:08:09 if you have to poll from userspace? Apr 02 14:08:20 just move audio processing into the I2S DMA IRQ handler Apr 02 14:08:44 and pray nothing blocks that irq for too long Apr 02 14:08:51 hook in on the FIQ Apr 02 14:09:24 whats with TI's SDK? Apr 02 14:09:34 is it suitable for beagle? Apr 02 14:09:39 last time I checked it was on ISO file or a DVD Apr 02 14:09:45 no audio inputs on that Apr 02 14:10:08 ok Apr 02 14:10:18 m0erk: if it claims to support the beagle, its suitable Apr 02 14:10:27 but it will be linux after all Apr 02 14:10:38 yes, i also think so Apr 02 14:10:53 so it wont get you closer to real time Apr 02 14:11:32 is there any way? Apr 02 14:11:44 there are many ways Apr 02 14:12:06 but again, it all depends Apr 02 14:12:27 yes, i get some more information about the task Apr 02 14:12:49 m0erk: is that some kind of school project? Apr 02 14:13:13 it is actually a project from an university Apr 02 14:13:30 m0erk, how long can be the latency for your active compensation? Apr 02 14:14:02 av500: is fiq available in non-secure mode on omap3? Apr 02 14:14:18 sundar: i dont know yet Apr 02 14:14:37 if it's just a school project, linux will be fine Apr 02 14:14:37 maybe thats what you have to figure out first Apr 02 14:15:22 mru: no idea Apr 02 14:15:31 last time we used it was on Davinci Apr 02 14:15:49 the aim is to get practical know-how on an own active noise compensaition system Apr 02 14:16:21 *cancellation Apr 02 14:16:50 there is a running ANC system running on an rapidcontrolboard Apr 02 14:16:57 ~9000$ Apr 02 14:17:15 if the same is possible with 150$ would be kind of nice :P Apr 02 14:17:24 without doing HW dev Apr 02 14:17:54 you are still in "depends" country Apr 02 14:18:15 but moving quickly towards buzzword ocean. (IMHO) Apr 02 14:18:24 in future we see the DM/OMAP SoCs also in Cars and Automotive, its a very nice plattform from TI Apr 02 14:18:39 you've seen the future? Apr 02 14:18:53 av500: yes iam still at the beginning right now Apr 02 14:19:02 beginning of the future? Apr 02 14:19:16 or end of the future Apr 02 14:19:24 mru: in the past, everything used to be better. inclunding the future. Apr 02 14:19:37 like iPad3 in ur car :D Apr 02 14:19:38 sundar: where it overflows and wraps back to the past? Apr 02 14:19:47 nice thought :) Apr 02 14:20:03 or an android tabloid Apr 02 14:20:03 and again we'll start with stones Apr 02 14:20:33 thats the magic of the human thoughts Apr 02 14:20:36 * av500 throws the 1st stone Apr 02 14:20:45 av500, ouch! Apr 02 14:20:52 according to clinton the [rolling] stones and cockroaches are the only forms of life that could survive nuclear apocalypse. Apr 02 14:21:07 cockroaches? Apr 02 14:21:12 not sure about cockroaches Apr 02 14:21:15 sundar: remember the z transform? Apr 02 14:21:17 rofl Apr 02 14:21:42 z transform? slightly. Apr 02 14:22:54 in the beginning i will start with the beagle as simulink target und get the algorhythim on it, maybe thats all i need to do and the system is just fast enough to handle it :P Apr 02 14:23:46 hope it all-goes-rhythmic Apr 02 14:24:31 google: "rush the rhythm method" Apr 02 14:30:10 and sorry for my bad english :\ Apr 02 14:30:26 kein problem Apr 02 14:30:30 :) Apr 02 14:35:18 good morning all Apr 02 14:35:36 gm Apr 02 14:35:46 gm Apr 02 14:37:46 hi av500 Apr 02 14:37:56 sundar: nice to meet you. Apr 02 14:38:03 :) my pleasure Apr 02 14:45:39 Hey, I found a great video to work with rxtx using Java. www.youtube.com/watch?v=KQ4NtRAETp4 Apr 02 14:47:05 isnt rxtx intended to be used with java? Apr 02 14:56:25 exclusively Apr 02 14:56:35 it's also a flaming pile of turd Apr 02 14:56:45 Has anybody run into buffer underrun problems on the BBxm using ALSA? Any known fixes or bugs? Apr 02 15:13:48 SilicaGel: "it's also a flaming pile of turd" <-- isn't that a requirement to be used with java? Apr 02 15:14:00 jno33: use a larger buffer Apr 02 15:14:01 its enforced at runtime, yes Apr 02 15:14:17 mru: but is that still "real time" then? Apr 02 15:14:59 real time is a very vague thing Apr 02 15:15:07 the human brain does a great deal of buffering Apr 02 15:16:34 mru: we've upped the buffer to 8000 samples at 8k mono 16bit, still suffering from underruns; should we crank that higher? Apr 02 15:17:23 thats 1s Apr 02 15:18:28 yes, it's already quite extreme, that's why I'm not thinking the buffer is the problem Apr 02 15:18:46 just checking to see if there is any other known issues on the BBxm Apr 02 15:18:56 mru: no it's not :( :( :( :( :( Apr 02 15:31:41 have to leave now. have a nice day. Apr 02 16:23:14 jkridner_: I just ordered a raspi Apr 02 16:23:32 where? Apr 02 16:23:46 rs? e14? Apr 02 16:24:10 e14 Apr 02 16:24:12 more importantly, why? Apr 02 16:24:27 how much did you get paid for doing it? Apr 02 16:24:46 mru: professional interest Apr 02 16:24:49 koen: how was the trip Apr 02 16:25:00 prpplague: at the mentor graphics office in fremont right now Apr 02 16:25:07 koen: ahh dandy Apr 02 16:26:57 koen: the next thing to (back-)order is the vavaldi-pi http://aseigo.blogspot.de/2012/04/vivaldi-ordering-delay.html Apr 02 16:27:04 vivaldi-pi Apr 02 16:28:12 av500: cant I just buy the non branded version? Apr 02 16:28:31 that would be cheating Apr 02 16:28:45 av500: that one needs to be bach-ordered Apr 02 16:55:36 what is the atheros firmware package name? Apr 02 17:35:51 I think I am reading the documentation right, but am just checking. The micro-usb port on the beagleboard-xm can be used to power the board and can take 1Amp? Apr 02 17:43:12 kvarley, i read it a bit ago and i think it might need a bit more than that if you plug in some usb devices Apr 02 17:43:26