**** BEGIN LOGGING AT Wed Mar 06 02:59:58 2013 Mar 06 04:00:02 is there a reason a BBB should be hotter? Mar 06 04:00:55 sharp looking tuxedo? Mar 06 04:01:15 it's the unzipped thing... Mar 06 04:17:18 * mranostay spams list Mar 06 04:17:50 that's not a nice thing to doo Mar 06 04:17:54 -o Mar 06 04:19:19 who said i'm nice? Mar 06 04:19:29 * mru certainly didn't Mar 06 05:26:08 mranostay, holy crap, your bbb is glowing red hot! Mar 06 05:26:43 Russ: no @$@# Mar 06 05:27:01 * Russ crops the image Mar 06 05:27:59 Russ: he's probally not running your PM kernels ;) Mar 06 05:54:34 Boo Mar 06 06:00:40 moaning Mar 06 06:02:21 hey Crofton|work Mar 06 06:14:08 yawn Mar 06 06:17:13 hey panto + davest Mar 06 06:17:16 in that order Mar 06 06:20:38 boing Mar 06 06:20:53 is that an antenna crashing down? Mar 06 06:22:56 bouncing back up... Mar 06 06:24:30 must be a big helical Mar 06 06:32:14 if I'm making a debug shield for the bbb, does anyone have feature requests? Mar 06 06:32:28 Russ: debug as in? Mar 06 06:32:46 we call them capes btw k thx Mar 06 06:32:52 cape Mar 06 06:32:52 cape Mar 06 06:33:16 lure! Mar 06 06:33:22 like, it'll have usb to serial with the 6 pin connector on the bottom Mar 06 06:33:43 someone has been interviewing with Atmel :P Mar 06 06:34:00 seriously, I just don't have any capes yet Mar 06 06:35:06 how is this better than just plugging in an FTDI cable? Mar 06 06:35:38 it'd also give you software access to the power button and the reset button Mar 06 06:36:16 and let you cut power to the board, so long as you use the debug board's barrel connector instead of the beagle's Mar 06 06:37:38 too bad the jtag signals aren't available too. Mar 06 06:38:04 what did i miss. Mar 06 06:38:06 hmm...could always include a connector for a cable Mar 06 06:43:02 Russ: what rev you have? Mar 06 06:48:11 A4 Mar 06 06:57:53 Russ: not that hot i assume? Mar 06 07:03:48 never measured Mar 06 07:04:53 does it hurt to touch? Mar 06 07:05:08 on the bear where does it hurt? Mar 06 07:08:48 on board hub? Mar 06 07:09:01 SPI to USB interface so USB host works reliably Mar 06 07:09:02 :D Mar 06 07:16:18 mranostay, did you figure out the position of the 6 pin connector, or did you guess? Mar 06 07:17:26 looks like 0.125 to me Mar 06 08:22:31 * av500 sets out to make a cape prototype Mar 06 08:24:11 trollcape? Mar 06 08:25:37 yeah Mar 06 08:40:47 mp3cape Mar 06 08:43:00 I could do that Mar 06 08:43:06 solder a MAS3507 Mar 06 08:43:22 in fact, I have MAS+DAC modules here Mar 06 08:43:25 just need a glue cape Mar 06 08:43:30 and boom, MP3 player Mar 06 08:43:45 hmm, I could solder a MP3 cape to a launchpad Mar 06 08:43:52 and play 1sec of mp3 Mar 06 08:44:54 http://fab.fritzing.org/fritzing-fab Mar 06 08:45:46 I think they have added support for capes too Mar 06 08:48:14 fädeldraht ftw Mar 06 08:48:40 duh, of course uart0 is not on the expansion connector :( Mar 06 08:49:29 get in line with mrpackethead ;) Mar 06 08:49:46 what? Mar 06 08:50:21 for the expansion connector whish list Mar 06 08:59:24 the parts for fritzing are done by well known people ;) Mar 06 08:59:45 s/parts/bb-parts/ Mar 06 09:15:40 creeper-tech: http://www.youtube.com/watch?v=6yY9UieIcjM Mar 06 09:16:46 what is creepy? Mar 06 09:18:33 I don't know, I only aproximating, for details you have to ask the creeper card inventors Mar 06 09:19:28 anyone know much about backups Mar 06 09:19:46 you need a tar tutorial? ;) Mar 06 09:20:01 lol. Mar 06 09:20:18 mrpackethead: I heard they are good, I plan to order one soon Mar 06 09:20:24 havent deciced on the color yet... Mar 06 09:20:24 i'm in the process of moving all my old machiens into a virtual server Mar 06 09:20:40 and was hoping to tidy up backups Mar 06 09:21:02 I use tar for that Mar 06 09:21:08 so black Mar 06 09:21:13 and sticky Mar 06 09:21:16 feathers too? Mar 06 09:21:34 black and secrect and deleted, subject to NDA Mar 06 09:24:10 cape done, the rest is SW Mar 06 09:24:36 as usual the part which needs most time Mar 06 09:26:03 what is the cape? Mar 06 09:26:07 another something-killer: http://www.cnx-software.com/2013/03/06/49-90-marsboard-is-yet-another-allwinner-a10-development-board/ Mar 06 09:26:18 KotH: ^^^^ SATA!!!1!1!eleven Mar 06 09:26:42 hello any1 thr? Mar 06 09:26:51 no Mar 06 09:27:08 wat? Mar 06 09:27:42 please turn on your spell checker Mar 06 09:29:22 any one from beaglebord.org plz we want your help ??? Mar 06 09:30:14 just ask Mar 06 09:30:17 dont ask to ask Mar 06 09:31:08 here we go: https://plus.google.com/112266164281670850856/posts/d8bS2D64RSq Mar 06 09:31:12 its the reset-cape :) Mar 06 09:31:28 maybe I will add a few leds though :) Mar 06 09:32:06 big-finger-board ;) Mar 06 09:32:12 s/board/vape/ Mar 06 09:32:17 s/v/c/ Mar 06 09:32:51 an entire cape to rest soemthign? Mar 06 09:32:53 wow. Mar 06 09:33:16 yes, the onboard reset switch is too tiny for me Mar 06 09:33:48 why do you ahve an FTDI break out Mar 06 09:34:00 because the other one is under the cape Mar 06 09:34:00 because he doesn't have a rn-42 ;) Mar 06 09:34:13 shru: whats your prblem.. JUST ASK Mar 06 09:34:19 if we dont' like it, we'll tell you Mar 06 09:34:29 too bad there is no uart0 on the expansion Mar 06 09:34:35 so I need to teach uboot uart1... Mar 06 09:34:40 well, its SW, should be doable = Mar 06 09:34:41 :) Mar 06 09:35:26 beagleboard xm supports camera MT9V034 ? Mar 06 09:36:48 more or less yes Mar 06 09:36:49 how to interface it ... i am using android in my board Mar 06 09:36:52 av500: sata but only 100mbit doesn't match. so it's just another fail-board Mar 06 09:37:02 XM has a parallel camera interface Mar 06 09:37:08 and that sensor has parallel out Mar 06 09:37:13 as in? Mar 06 09:37:28 so, with some glue and some more SW it should work Mar 06 09:37:33 but.... Mar 06 09:37:41 you wont get exact steps here Mar 06 09:37:47 shru: is this a school proejct? Mar 06 09:37:50 you will need to dep dive into that yourself Mar 06 09:37:55 deep dive* Mar 06 09:38:06 mini project Mar 06 09:38:12 but you are lucky: http://labs.isee.biz/index.php/Aptina_mt9v034_sensor_V4L2_based_linux_driver Mar 06 09:38:29 it has already been done Mar 06 09:39:35 https://github.com/Aptina/BeagleBoard-xM/tree/master/MT9V034 Mar 06 09:39:41 even aptina supports it Mar 06 09:40:27 even https://github.com/Aptina/BeagleBoard-xM/blob/master/beagleboard_xm_adapter_ap21023_rev0.pdf Mar 06 09:40:29 schametics Mar 06 09:40:40 oh thank u .... Mar 06 09:40:48 search engines ... the unknown land Mar 06 09:41:31 and the glue pcb even uses all TI parts, very good Mar 06 09:42:59 shru. for htat Mar 06 09:43:09 you must return with treasure for the trolls Mar 06 09:43:21 what? Mar 06 09:43:31 unless you want the fleas of 1000 camels to infest your armpits Mar 06 09:43:33 sorry didnt get you Mar 06 09:43:37 its the rules of #bealge Mar 06 09:43:57 the trolls have imparted much wisdom to you Mar 06 09:44:24 if you want to survive the dark forest, bring treasure. Mar 06 09:44:24 exact steps: https://github.com/Aptina/BeagleBoard-xM/blob/master/MT9V034/Angstrom/README_Beagleboard-xM_mt9v034.txt Mar 06 09:44:31 av500: cool! thanks for the link Mar 06 09:44:39 EXACT STEPS!!!111!! Mar 06 09:45:05 shru: some of the trolls like cheese. but most like beer Mar 06 09:45:17 maybe we should rename the channel to #wegoogleforyou Mar 06 09:45:38 #beagoogle? Mar 06 09:45:47 #boogle Mar 06 09:45:53 its like one of those new designer dogs Mar 06 09:46:04 wasnt lycos the dog that searches? Mar 06 09:46:17 a lab x poodle is a labradoodle Mar 06 09:46:46 im sure shru will be back soon with treasure. Mar 06 09:46:50 http://www.youtube.com/watch?v=M6x7Xdy8BtY Mar 06 09:47:08 fine print: bitcoins or nigerian scames not accepted. Mar 06 09:47:47 but scones are Mar 06 09:48:14 and only cherry ones Mar 06 09:48:33 non of that raisin nonsensw Mar 06 09:48:51 devon scones of course Mar 06 09:49:13 devon Mar 06 09:49:18 nah.. Mar 06 09:49:36 * KotH hands everyone some cookies Mar 06 09:50:02 * dm8tbr noms some of the cookies Mar 06 09:50:05 down with the english Mar 06 09:50:10 hello...how can i get an accerlerometer for my beagle xm Mar 06 09:50:24 shru: did you bring scones? Mar 06 09:50:29 * jackmitchell facepalms Mar 06 09:50:35 email av500 some scones... Mar 06 09:50:45 mrpackethead: sorry, i native english speaking not am Mar 06 09:51:11 how to write drivers for beaglebord xm with android os Mar 06 09:51:51 æ‹–é’“ Mar 06 09:51:52 shru: learn Mar 06 09:52:00 there is not Howto Mar 06 09:52:01 no* Mar 06 09:52:22 teach Mar 06 09:52:22 shru: do you know what the internet is? Mar 06 09:52:47 shru: pay us Mar 06 09:52:59 sure Mar 06 09:53:35 shru: do you have enough money to pay us 500-1000chf/hour ? Mar 06 09:53:45 s Mar 06 09:53:47 shru: sorry, this is not a teaching channel Mar 06 09:53:50 KotH: your not worth 1000 Francs an hour Mar 06 09:54:02 mrpackethead: i might not be, but i can still charge it :) Mar 06 09:54:11 your worth at least 5000 Mar 06 09:54:12 mrpackethead: ~800 go into the chocolate he eats per hour Mar 06 09:54:18 mrpackethead: lol Mar 06 09:54:25 mrpackethead: that would be extortion Mar 06 09:54:38 KotH: danger money Mar 06 09:54:45 mrpackethead: i have a conscience! Mar 06 09:54:48 is anyone from beagleboard xm technical plz we need ur help? Mar 06 09:54:56 http://norvig.com/21-days.html Mar 06 09:55:01 i dont want to talk non sence Mar 06 09:55:16 shru: the best brains on all thigns beagle are here Mar 06 09:55:19 shru: this is as technical as it gets Mar 06 09:55:23 shru: call TI they might have an advisor to help you Mar 06 09:55:30 aholler: thanks! :) Mar 06 09:55:46 shru: look, teaching someone something is a long and tedious process Mar 06 09:56:05 shru: nothing something someone does willingly, when you can instead develop cool new gadgets Mar 06 09:56:19 shru: so you have to give some very good incentive to do so Mar 06 09:56:40 and tell funny jokes Mar 06 09:56:44 shru: oh.. and as av500 already said, this is not a teaching channel. we dont teach here. we troll people Mar 06 09:56:58 unless you are a hobbit like me Mar 06 09:57:10 i try to run away from the trolls Mar 06 09:57:18 i know its not teaching channel but can help us know.... Mar 06 09:57:22 shru: we can answer questions, but we do not hold lessons Mar 06 09:57:32 okay Mar 06 09:57:40 apropos holding lessons Mar 06 09:57:53 only if you have the ring.... you might rule them all! Mar 06 09:57:53 CLASS IS IN SESSION. Mar 06 09:58:07 KotH, stop eating chocolates! Mar 06 09:58:11 i have been asked to give an introductory talk on electronics at one of the LUG meetings Mar 06 09:58:26 okay... :( Mar 06 09:58:31 as we have a few non-EEs here: what are the things you usually want to know about electronics? Mar 06 09:58:48 how to avoid them Mar 06 09:58:58 mrpackethead: i'm not eating chocolate, but chocolate cokies :) Mar 06 09:59:10 dont' talk back Mar 06 09:59:26 ok.. i talk to my cookies instead Mar 06 09:59:42 shru: take small steps, make the camera work with std linux first Mar 06 09:59:47 then worry about Android Mar 06 09:59:53 "surviving the errata" would be a good talk Mar 06 10:00:05 electronics.. everyone wants to know how the flux capacitor works Mar 06 10:00:13 by experience, by the time step one is done, your project is already over Mar 06 10:00:23 aholler: what do you mean by that? Mar 06 10:01:35 shru: ask your question ... Mar 06 10:02:06 KotH: it depends what type of electronics you mean, but I would suggest going over pull-ups/downs, clocks, oscillators and interrupts but that is more geared to understanding silicon I suppose... Mar 06 10:02:29 jackmitchell: well.. that's my problem... i dont know what electronics i should mean Mar 06 10:02:41 nothing now i'll try and ask you tommorow Mar 06 10:02:51 thank you everyone..:) Mar 06 10:03:00 glad we were able to entertain Mar 06 10:03:03 jackmitchell: the people at the LUG are mostly sysadmins and coders. no experience with embeded stuff... i've i'm lucky there are one or two electriciancs in the crowd Mar 06 10:03:17 KotH: http://hardwarebug.org/ Mar 06 10:03:26 will someone pelase show mr shru out, and bring the next contestent into the room Mar 06 10:03:38 jackmitchell: i know i cannot teach all of electronics, so i try to select something that people might be interested in Mar 06 10:03:52 make a LED blink Mar 06 10:03:53 aholler: i know mrus blog Mar 06 10:04:00 KotH: while not strictly electronics, maybe some low level protocols would be quite interesting, i2c, spi etc... Mar 06 10:04:15 aholler: do you mean hw errata that hunt you when writing your software? Mar 06 10:04:26 there coudl be a years talking in that topic alone. Mar 06 10:04:46 "how many ways can you make a LED blink" Mar 06 10:04:47 but i'll be back to eat your heads again Mar 06 10:04:51 exactly Mar 06 10:05:01 braiiiins! Mar 06 10:05:11 shru: are you a troll Mar 06 10:05:15 i think you are Mar 06 10:05:29 mrpackethead: getting a LED into the lasing regime would be a nice exercize in physics :) Mar 06 10:05:30 shru: do you work for IT? Mar 06 10:05:35 TI? Mar 06 10:05:50 no ia a student Mar 06 10:05:58 *i am Mar 06 10:06:07 who has a homework assignmetn due on Friday? Mar 06 10:06:15 o/ Mar 06 10:06:33 KotH does assignmetns for chocolate Mar 06 10:06:49 actually for credit points ^' Mar 06 10:07:04 (1) Led Blinker.. Mechaincal timer method Mar 06 10:07:09 wat chocolate story is all bot? Mar 06 10:07:27 (2) Led Blinker, Two Transistors, a couple of resistors and a cap Mar 06 10:07:38 (3) D Flip Flop Mar 06 10:07:43 (4) PRU Mar 06 10:07:48 ite endless. Mar 06 10:07:53 ok, i need sleep Mar 06 10:07:59 sleep well Mar 06 10:08:10 dont led the little tiny DIL bugs byte you Mar 06 10:08:12 sleep well................... Mar 06 10:08:19 shru: we only take good swiss chocolate Mar 06 10:08:38 swiss woud'nt knwo chocolate if it hit them in the head. Mar 06 10:08:47 hehehe ... even i like chocolate.... hmmm yummy..:p Mar 06 10:08:58 shru: for plain chocolate läderach is our favorite, otherwise you may choose between Lindth, Caillers, Teuscher and other small swiss chocolatiers Mar 06 10:09:22 shru: we do not accept belgian chcolate Mar 06 10:09:26 Russ: yesterday you've asked for features for your debug-cape: add a rn-42 or some ti-bt-solution to the cape Mar 06 10:09:54 okay ..cya Mar 06 10:09:56 jackmitchell: hmm.. actually the interfaces might be a good idea Mar 06 10:10:21 jackmitchell: maybe i should model the talk around an arduino or a simple uC board design Mar 06 10:10:39 digital stuff is much easier to teach to software guys than analog.. Mar 06 10:11:01 isn't it all ohm? Mar 06 10:11:28 aholler, how would it be used? Mar 06 10:11:47 I think it would be much more interesting to know how that accelerometer got it's data into the uC than how I can use different size resistors to change how bright an led is Mar 06 10:11:51 Russ: to get rid of cables (bt-serial) Mar 06 10:12:08 mrpackethead: now i realy wonder what it would take to make a laser out of a LED ^^' Mar 06 10:12:22 laser diode Mar 06 10:12:23 ? Mar 06 10:12:38 not a laser diode directly Mar 06 10:12:53 but is it possible, and if it is, how would you make a laser out of a standard LED Mar 06 10:13:41 KotH: well, you should start from the atom model and electron flow and P and N junctions Mar 06 10:13:45 no? Mar 06 10:13:54 well..that's the same for a normal LED and a laser diode Mar 06 10:13:55 jeez ups, '1:36 A.M. Out For Delivery' Mar 06 10:14:21 both work on the recombination light within a p-n junction Mar 06 10:14:42 a laser is even a "LED" when driven below threshold Mar 06 10:14:53 so, the question is, what makes the laser diode a coherent source? Mar 06 10:17:28 to convert a LED to a laser, you do like for a shutgun, you saw off the end :) Mar 06 10:17:33 shotgun* Mar 06 10:17:35 lol Mar 06 10:20:52 Russ: combined with the reset using cts such a bt-solution would be nice. I've recently bought me a rn-42 and like it a lot Mar 06 10:21:47 I still have to add the cts-reset but that's already on my homework list ;) Mar 06 10:22:34 or was it dtr? I forgot Mar 06 10:24:44 ah, there it is: https://github.com/russdill/dtr/wiki Mar 06 10:24:56 yup Mar 06 10:25:32 hmmm Mar 06 10:25:36 dunno, a bluetooth module might be hard to combine in with what I was thinking, but there is so much pcb realestate that its certainly possible Mar 06 10:25:46 maybe I should connect the reset to DTR on the FTDI Mar 06 10:25:58 av500, use a gpio Mar 06 10:26:06 together with a switch to have the bt always on the perfect wireless debug solution would be ready Mar 06 10:26:19 Russ: but then I can do it from the terminal Mar 06 10:26:38 av500, I'm spoiled with conmux Mar 06 10:26:39 CTRL-A H Mar 06 10:27:13 to use the rn-42 only 4 connections are needed Mar 06 10:27:13 Russ Mar 06 10:27:16 nothing else Mar 06 10:27:33 Russ well I can also hook a custom upload script to minicom :) Mar 06 10:31:21 have to look at the rn-42-datasheet, maybe it's possible to switch some gpio using remote commands. Mar 06 10:32:09 an advanced project would be to use btle for the serial ;) Mar 06 10:34:22 ti could offer such a firmware for the cc2541 ;) Mar 06 10:35:17 then all what would be needed is a serial-btle-driver for linux Mar 06 10:35:35 shouldn't be that hard Mar 06 10:35:43 aholler: write one Mar 06 10:36:06 aholler: are there free GPIOs on the sensor tag? Mar 06 10:36:15 av500: I would have to register to get the btle-stack do build my own fw Mar 06 10:36:26 av500 have you heard of google Mar 06 10:36:36 and off course, time is needed Mar 06 10:36:41 -f Mar 06 10:36:45 apt-get install time Mar 06 10:37:00 apt-get install payed-time Mar 06 10:37:09 s/y/i/ Mar 06 10:37:24 thyme? Mar 06 10:37:36 can be nice on pizza Mar 06 10:37:59 thyme is always nice :) Mar 06 10:38:04 hmm, that idea sounds like killer-app for btle ;) Mar 06 10:38:23 apt-get install tardus Mar 06 10:38:45 or apt-get install dellorian Mar 06 10:39:41 mv #beagle #think-tank Mar 06 10:39:55 think different pleae Mar 06 10:39:58 +s Mar 06 10:40:28 * KotH wonders why a think-tank is called a think-tank, when one should think out of the box Mar 06 10:40:40 no rotten fruits here Mar 06 10:41:15 KotH: we need a freiluft think event Mar 06 10:41:35 av500: is that similar to "luftbaden" ? :) Mar 06 10:42:32 yeah Mar 06 10:43:47 Russ: so add a cc2541 and hire some ti-btle-staff to write the serial-sw ;) Mar 06 10:45:56 av500: any idea why luftbäder went out of fashion in germany? they have been very poplular not even 100y ago Mar 06 10:46:07 it doesn't work out of the box? Mar 06 10:46:24 serial over btle? no Mar 06 10:47:02 btle has much reduced functionality Mar 06 10:47:56 but should be enough to use it as a (low-speed) serial Mar 06 10:51:17 the ti-stack has ota (over the air) firmware updates, so almost everything should be already there Mar 06 10:51:32 russ: ble only has some profiles Mar 06 10:52:50 s/ota/oad/ Mar 06 10:53:44 and because this is #think-tank: oad-u-boot Mar 06 10:53:54 ;) Mar 06 10:54:04 o-boot Mar 06 10:54:11 yes, adding uboot always makes a project better Mar 06 10:54:55 * av500 watches the RAID leds blink blue and red Mar 06 10:55:03 red? Mar 06 10:55:23 yes, its syncing Mar 06 10:55:40 im upgrading drives Mar 06 11:00:10 <_av500_> https://lh5.googleusercontent.com/-v_vhewaxpVs/UTbvGAp7iLI/AAAAAAAAhZw/IS5iv6s8cFA/s250/zPan3gJ.gif Mar 06 11:07:37 oha, one of my laptops usb ports shuts down the bone Mar 06 11:07:41 the other works Mar 06 11:07:58 its the always-on chanrging one that fails... Mar 06 11:12:34 hmm, a cc2541 costs around 2€ for 2.500, 3€ for 1.000, 5€ for one Mar 06 11:14:33 buy for 2, sell for 5 Mar 06 11:15:44 I should start a kickstarter to build a btle-ftdi replacement Mar 06 11:17:33 but I live in a conservative underdeveloped and de-innovative country where it needs years to do so. Mar 06 11:17:59 aholler, you're complaining? what you think I should do? Mar 06 11:19:02 panto: you've already complained, now I do ;) Mar 06 11:27:51 * av500 wonders who designed u-boot? David Lynch? Mar 06 11:30:53 didn't you met him at ew? Mar 06 11:31:37 no Mar 06 11:31:53 I also like how particularly useless uEnv.txt is Mar 06 11:32:36 u-boot is designed for storing the environment on a raw flash medium Mar 06 11:32:42 yes Mar 06 11:32:46 the mmc/fat filesystem is a hack Mar 06 11:33:05 as already mentioned, use a eeprom Mar 06 11:33:27 panto: well, since it can read uEnv.txt from mmc/fat, it could read the env from there too Mar 06 11:33:45 but I guess that was too hard Mar 06 11:34:03 ? Mar 06 11:34:22 you want to write binary environments? Mar 06 11:34:33 why binary? Mar 06 11:34:34 av500, it's more complicated than that Mar 06 11:34:41 panto: I am all ears Mar 06 11:35:05 I would guess a better way would be to have a partition set aside just for storing the environment without a filesystem Mar 06 11:35:19 av500: just remove the -t from the config-script and you are done Mar 06 11:35:24 av500, cause the problem is not reading the environment, it's writting it Mar 06 11:35:27 *writing it Mar 06 11:35:32 panto: no need for that Mar 06 11:35:38 I can write the SD card Mar 06 11:35:41 from the PC Mar 06 11:35:43 av500, I kinda disagree with that Mar 06 11:35:48 and uboot could read that Mar 06 11:35:48 it's a copout Mar 06 11:36:39 panto: the only thing needed was to not require boot.scr to have that stupid header Mar 06 11:36:42 hmm, I think I don't understand the problem.could someone explain? Mar 06 11:36:56 aholler: I want a txt file with the env Mar 06 11:37:01 that I write to the sdcard Mar 06 11:37:05 how hard is that? Mar 06 11:37:17 that's uEnv.txt, not? Mar 06 11:37:21 no Mar 06 11:37:31 why? Mar 06 11:37:35 av500, err, I think boot.scr is gone Mar 06 11:37:35 uenv sets some variable that work with the compiled in env Mar 06 11:37:59 so I might as wll just recompile uboot Mar 06 11:38:10 and write that to SD Mar 06 11:38:12 so your problem is that you want to replace the environment? Mar 06 11:38:17 I don't have a boot.scr at my u-boot Mar 06 11:38:20 aholler: even add to it Mar 06 11:38:22 but yes Mar 06 11:38:25 let me replace it Mar 06 11:38:31 with a text file Mar 06 11:38:31 that should be a 5min job Mar 06 11:38:37 pray tell Mar 06 11:38:37 err, guys Mar 06 11:38:43 it's already like this Mar 06 11:38:45 Or I still don't understand Mar 06 11:38:55 the bone's u-boot doesn't use boot.scr at all Mar 06 11:38:56 aholler: ok, how do I replace the env? Mar 06 11:39:00 with mine Mar 06 11:39:01 there's only uEnv.txt Mar 06 11:39:05 which is just text Mar 06 11:39:08 yes Mar 06 11:39:11 but its not the env Mar 06 11:39:14 its variables Mar 06 11:39:19 like optargs Mar 06 11:39:22 env is variables Mar 06 11:39:25 that are used as $(optargs) Mar 06 11:39:41 in the compiled-in env Mar 06 11:39:42 are you talking about overriding the built-in environment? Mar 06 11:39:49 yes Mar 06 11:39:57 av500, preEnv.txt is what you look for i think Mar 06 11:40:12 any more envs? Mar 06 11:40:18 heh, no, just that two Mar 06 11:40:33 not in mainline uboot Mar 06 11:40:40 pre overrides the env (i might be wrog, its quite a while i touched that stuff) Mar 06 11:40:42 you can cut down the builtin enviroment to almost nothing Mar 06 11:40:45 *wrong Mar 06 11:40:53 panto: but see, I do not want that Mar 06 11:41:00 panto: I want to write a simple text file Mar 06 11:41:04 yes, I can recompile uboot Mar 06 11:41:11 but then, what is the point? Mar 06 11:41:11 av500, you can't just do that Mar 06 11:41:28 hmm, i thought linaro u-boot was plain upstream nowadays Mar 06 11:41:29 that's why I was talking about having the raw block partition as env storage Mar 06 11:41:31 panto: so answer me please, how do I set my own enc Mar 06 11:41:41 from the sd card Mar 06 11:41:51 I cant Mar 06 11:41:56 av500, I'm sure you can Mar 06 11:42:06 just override everything in the built-in environment Mar 06 11:42:19 you could delete uEnv.txt and provide a boot.scr i bet ;) Mar 06 11:42:34 ogra_: but boot.scr is not a text file Mar 06 11:42:40 yeah, indeed Mar 06 11:42:40 av500, ok, ok, let's take it from the top Mar 06 11:42:48 because reading text files is *hard* Mar 06 11:42:48 but a workaround at least Mar 06 11:42:55 heh Mar 06 11:43:11 ogra_: soon we will sell >$50 BBBs to the arduino crowd Mar 06 11:43:14 think them, not me Mar 06 11:43:16 you _need_ some kind of environment before switching u-boot to accessing the full RAM Mar 06 11:43:18 er, < Mar 06 11:43:24 yeah Mar 06 11:43:52 panto: ok Mar 06 11:44:06 after you've switched, you can ditch all the environment and load the uEnv.txt from the medium (which has a filesystem) Mar 06 11:45:54 panto: what I want in fact, is that some kind of env is read before I get to uboot cmd line Mar 06 11:46:44 so how can I do that? Mar 06 11:46:49 without recompiling uboot? Mar 06 11:47:21 let another one recompile it for you Mar 06 11:47:27 sigh Mar 06 11:48:18 * ogra_ really doesnt get that preEnv.txt stuff isnt upstream, i think its essential for uEnv.txt support Mar 06 11:48:19 you can't access the filesystem in the very reduced memory setup Mar 06 11:48:41 you load u-boot0bin from it Mar 06 11:48:53 indeed Mar 06 11:48:53 u-boot.bin Mar 06 11:48:57 yes Mar 06 11:49:27 it's very risky Mar 06 11:49:31 ? Mar 06 11:49:37 ? Mar 06 11:49:50 ok, ok, it is possible Mar 06 11:50:14 but you need to do parsing and all that crap Mar 06 11:50:19 panto: how did boot.scr do it? Mar 06 11:50:27 who loaded that MLO or uboot? Mar 06 11:50:41 $diety Mar 06 11:51:02 MLO is just a cutdown u-boot Mar 06 11:51:08 so you could say it was u-boot Mar 06 11:51:13 I still don't get the problem Mar 06 11:51:34 aholler: I - as a user - want to supply my own env Mar 06 11:51:40 with out recompiling uboot Mar 06 11:51:52 env import -d -t Mar 06 11:51:53 and Mar 06 11:52:07 [12:51:39] with out recompiling uboot Mar 06 11:53:04 moaning Mar 06 11:53:07 +1 Mar 06 11:53:13 moaning is free Mar 06 11:53:14 so let someone else add the -d to the config and let him recompile u-boot for you Mar 06 11:53:32 * av500 gives up Mar 06 11:54:38 it was by intention that uEnv.txt doesn't replace the environment completly Mar 06 11:54:38 ogra_: url to your uboot? Mar 06 11:55:25 aholler: ok, let me give you a use case: Mar 06 11:55:28 if you need docuentation for env: http://git.denx.de/cgi-bin/gitweb.cgi?p=u-boot.git;a=blob;f=common/cmd_nvedit.c;h=3a05e6010306f6053624514c48843c375e43b9f6;hb=HEAD#l763 Mar 06 11:55:41 av500, https://code.launchpad.net/u-boot-linaro Mar 06 11:56:02 I have this BBB that boots from MMC just fine. now for some dev work, I want to tftp boot it, so yes I can add some foo to uenv to do that Mar 06 11:56:05 av500: I understand your problem. but if you want a change you have to complain at the maintainers Mar 06 11:56:13 but, if I dont want that, I have to edit uenv again Mar 06 11:56:20 what I want is to stop the autoboot Mar 06 11:56:22 av500, talk to jcrigby in #linaro about details (i really havent touched u-boot for a wile) Mar 06 11:56:26 and be able to say tftp Mar 06 11:56:37 or mmcboot Mar 06 11:56:39 or whatever Mar 06 11:56:49 and uenv holds my serverip, and ipaddr etc Mar 06 11:56:55 right now, its all or nothing Mar 06 11:57:06 at line 907 the documentation for env import can be seen Mar 06 11:57:12 yes Mar 06 11:57:15 I get that Mar 06 11:57:23 but somebody has to call import Mar 06 11:58:15 aholler: and I understand who I have to complain to, but then stop telling me that everything is fine Mar 06 11:58:49 av500: I don't understand that too, otherwise env import would have -r Mar 06 11:59:18 -r? Mar 06 11:59:18 to get rid of carriage returns Mar 06 11:59:24 ah Mar 06 11:59:59 solution is that you to got in fight ;) Mar 06 12:00:06 +have Mar 06 12:00:43 everyone loves maintainers ;) Mar 06 12:01:54 does anyone knows where the heck can i get kernel sources for an LCD7 cape? Mar 06 12:02:09 it doesn't seem supported by TI PPS kernel :( Mar 06 12:03:47 #wegoogleforyou Mar 06 12:05:16 http://lmgtfy.com/?q=lcd7+cape+kernel+source Mar 06 12:07:03 ogra_: https://launchpad.net/u-boot-linaro Mar 06 12:07:13 ogra_: but I am to stupid, where is the code? Mar 06 12:07:25 I bzr'ed somerthing, but its mostly empty Mar 06 12:11:28 ah http://git.linaro.org/gitweb?p=boot/u-boot-linaro-stable.git;a=shortlog Mar 06 12:17:10 av500: why you need that? Mar 06 12:17:33 what? Mar 06 12:18:13 just add something to the config which loads an environment otherwise than by using CONFIG_BOOTCOMMAND Mar 06 12:18:33 aholler: very funny... Mar 06 12:18:56 ezequiel: LCD7 cape is supported by the angstrom kernel Mar 06 12:19:11 i spent the whole day googling and found lots of patches and discussion, but can't seem to find where the heck are the damn sources Mar 06 12:19:12 for TI kernels you need to ask TI Mar 06 12:19:45 yes, i know angstrom supports LCD7 cape. i'm building angstrom now, i guess the kernel sources will eventually appear. Mar 06 12:19:51 yes Mar 06 12:20:03 just make sure to comment the rm_work stuff Mar 06 12:20:11 otherwise they will disappear as fast... Mar 06 12:20:36 mmm... good to know, i guessed there was something like that Mar 06 12:21:02 btw, angstrom sources allow to pick beagleboard, but not beaglebone... Mar 06 12:21:10 i guess it's the same kernel Mar 06 12:22:13 ezequiel: patches against 3.8 for the cape should be in https://github.com/beagleboard/kernel/tree/3.9/patches somewhere; but they may have been merged to mainline between 3.2 and 3.8 Mar 06 12:23:00 da8xx-fb should hold some hints Mar 06 12:23:01 ezequiel: yes, same kernel Mar 06 12:23:07 cool Mar 06 12:23:12 if not have a look at koens git-repos: https://github.com/koenkooi Mar 06 12:23:57 aholler: but this "something" has to be called by somebody, no? Mar 06 12:24:37 aholler: thanks Mar 06 12:24:38 av500: isn't there some config var which does this? Mar 06 12:25:08 I dont know :) Mar 06 12:25:26 soon you will tell me to read source code! Mar 06 12:25:53 rtfs ;) Mar 06 12:27:18 aholler: yes, its called preboot :) Mar 06 12:27:36 http://git.linaro.org/gitweb?p=boot/u-boot-linaro-stable.git;a=commitdiff;h=88da8539c1805bf6d19c8c9d2a0e34938958ce38 Mar 06 12:27:47 defined by CONFIG_PREBOOT Mar 06 12:27:53 std-u-boot knows that too Mar 06 12:28:01 and lo and behold, it loads an env Mar 06 12:28:25 aholler: of course, all the linaro patch does is to: [13:18:14] just add something to the config which loads an environment otherwise than by using CONFIG_BOOTCOMMAND Mar 06 12:28:27 :) Mar 06 12:29:09 u-boot needs a Linus Mar 06 12:29:24 just copy CONFIG_BOOTCOMMAND to CONFIG_PREBOOT and replace uEnv.txt with preEnv.txt Mar 06 12:29:48 and add -d to env import Mar 06 12:29:58 yes, that patch does that more or ledd Mar 06 12:29:59 less Mar 06 12:30:00 and my patch for -r ;) Mar 06 12:30:29 maybe I will call it replaceEnv.txt :) Mar 06 12:30:39 and trollEnv.txt Mar 06 12:32:31 hmm, but it's a good idea, I will make such a patch for me too ;) Mar 06 12:32:56 if I ever will compile u-boot again Mar 06 12:34:43 adding some magic like uEnvCmd to that preEnv.txt should be good too. don't know if linaro does that Mar 06 12:37:12 in fact, I would just set bootenv in preboot, call the stuff to load it and afterwards change bootenv back to uEnv.txt Mar 06 12:40:41 someone's used Simon Glass' patches for secure u-boot ? Mar 06 12:40:51 av500: that is something nice too http://dpaste.de/s6Ccn/ Mar 06 12:41:19 if you're already starting u-boot hacking ;) Mar 06 12:42:06 just use env import instead of load Mar 06 12:42:36 that snippets still loads cmd's Mar 06 12:43:55 but enough free u-boot consulting … Mar 06 12:44:18 oh make it my turn next ;) Mar 06 12:45:14 aholler: as I understand it, my lacie NAS has all the boot/recovery magic in uboot scripts Mar 06 12:45:38 time for u-bootOS Mar 06 12:45:39 sure, that's how it's usually done Mar 06 12:46:29 I'd prefer to do minimal stuff in uboot and the rest in a recovery kernel Mar 06 12:46:33 but well Mar 06 12:47:14 some should add lua to both, u-boot and the kernel Mar 06 12:47:19 yes Mar 06 12:47:25 and a way for them to talk to each other Mar 06 12:47:32 via luaEnv.txt :) Mar 06 12:48:16 at least we shouldn't have to switch brain between different shell-dialects Mar 06 12:48:29 hush/ash/whatever Mar 06 12:48:32 as long ast it is not split brain Mar 06 12:48:37 aholler: wait to uboot lua and kernel lua diverge Mar 06 12:48:42 till* Mar 06 12:49:52 replacing the init for initrd's with lua is next on my never-ending-stupid-idea list ;) Mar 06 12:50:28 I just don't see any reason to use a shell to do the stuff which initrd's are doing Mar 06 12:50:55 usually they just load some modules and a splash Mar 06 12:51:30 but ever initrd reinvents that Mar 06 12:51:39 +y Mar 06 12:51:43 and they mount /usr, very important! Mar 06 12:51:55 ah no, they dont Mar 06 12:52:00 that was the problem Mar 06 12:52:26 yes, thats done by init Mar 06 12:52:56 the real init Mar 06 12:54:11 yes Mar 06 12:54:20 will the real init please stand up Mar 06 12:55:21 but some funding is needed to implement lua for initrd's ;) Mar 06 12:56:10 would that be in-kernel lua? Mar 06 12:56:14 or a standalone process? Mar 06 12:56:26 thats my idea, in-kernel (as module or static) Mar 06 12:56:39 well, then initrd is 2nd step Mar 06 12:56:43 so no userland with libc would be necessary Mar 06 12:56:48 right Mar 06 12:56:48 for initrd's Mar 06 12:56:56 for other stuff too Mar 06 12:57:02 like arduino like linux Mar 06 12:57:12 inkernel lua and load user scripts from mmc Mar 06 12:57:14 done Mar 06 12:57:16 blink all the leds Mar 06 12:57:29 yes, lua is easy to extend Mar 06 12:58:01 but starting with loadmodule pivot-root and such stuff which is needed for initrd's might be easier to get accepted Mar 06 12:58:05 (mainline) Mar 06 12:59:12 so, if someone has a funding pot, mail me ;) Mar 06 13:00:37 I have a jar with pennies Mar 06 13:00:44 it can be yours! Mar 06 13:01:13 I use a asbach-uralt-5l-bottle for that. it's full with DM ;) Mar 06 13:03:19 aholler: you must be old... not even have you heard of that old currency, but you even own some of it! :) Mar 06 13:03:27 maybe google will realize it, at least I've suggested that on the u-boot-ml to some chrome-people Mar 06 13:04:07 old and crumpy Mar 06 13:04:38 av500 has a nice graph about that Mar 06 13:04:47 ? Mar 06 13:05:16 uh, would have to look at your blog Mar 06 13:06:15 and you post too much to find it in seconds ;) Mar 06 13:06:48 sorry Mar 06 13:15:24 https://plus.google.com/112266164281670850856/posts/5hne3K3eAgm Mar 06 13:16:20 http://www.h-online.com/open/news/item/Chinese-government-criticises-Android-dominance-1817471.html Mar 06 13:16:22 lol Mar 06 13:16:32 I loled too Mar 06 13:17:05 oh the irony, toys out of the pram comes to mind Mar 06 13:17:23 those chinese Mar 06 13:18:07 somebody should explain them how a "fork" works Mar 06 13:18:16 they don't understand that Mar 06 13:18:29 that would mean do it yourself Mar 06 13:19:24 yes Mar 06 13:19:28 not just copy Mar 06 13:19:31 without exact steps Mar 06 13:21:14 aholler: there are exact steps: "The definition of open: 'mkdir android ; cd android ; repo init -u git://android.git.kernel.org/platform/manifest.git ; repo sync ; make'" Mar 06 13:21:21 ;) Mar 06 13:21:46 that's why it's dorminant Mar 06 13:21:59 dormant or dominant? Mar 06 13:22:05 whatever Mar 06 13:22:27 it's all going according to plan I'm told Mar 06 13:22:29 backup just ended, now I can work again ;) Mar 06 13:22:43 sword fight is over..back to work Mar 06 13:23:40 uups, forgot to backup stupid windows, still time ;) Mar 06 13:24:10 it just doesn't leave me alone :( Mar 06 13:25:43 i know, "just screw it" ;) Mar 06 13:29:53 so, you're a screw driver? Mar 06 13:31:21 upps, the url above was wrong, I meant https://plus.google.com/112266164281670850856/posts/eE5xowrY7z5 Mar 06 13:32:27 aholler: yep Mar 06 13:34:56 "just screw it" is the age maintainers often have ;) Mar 06 13:35:23 yep Mar 06 13:36:14 especially the crumpy ones Mar 06 13:37:37 therefor fighting with them isn't easy ;) Mar 06 13:38:55 hello, does anyone has some experience how fast we can toggle the gpios? Mar 06 13:39:33 this is getting boring :) Mar 06 13:39:36 not 6MHz Mar 06 13:40:57 wtf? is the beagleboard irc the same as the pandaboard? Mar 06 13:41:03 no Mar 06 13:41:12 but some people are in both Mar 06 13:42:01 you two timing troll av500 :P Mar 06 13:42:40 I see him creeping in edev to; a troll of many bridges Mar 06 13:43:04 * av500 crawls back under his rock Mar 06 13:44:02 probably i try to ask in a different way. Do anyone know a good microcontroller like pandaboard, beagleboard which can toggle gpios with 6MHz or faster? Mar 06 13:44:16 but why? Mar 06 13:44:19 why gpios Mar 06 13:44:25 why not a dedicated clock output Mar 06 13:44:27 or SPI? Mar 06 13:44:31 gerolsprudel_: anything with pwm should be able to do that no? Mar 06 13:45:00 and, why not make a fixed 6MHz clock? Mar 06 13:45:09 if that is all you need Mar 06 13:45:23 go fpga Mar 06 13:45:34 http://uk.farnell.com/iqd-frequency-products/lf-x335b/crystal-oscillator-6mhz/dp/9712666 Mar 06 13:45:37 there, fixed Mar 06 13:46:35 gerolsprudel_: try a SAM7S256 Mar 06 13:47:09 eh, old tech Mar 06 13:47:12 gerolsprudel_: even my venerable Ti-85 can togle pins faster than 6MHz (if you overclock it) Mar 06 13:47:34 aholler: i rather call it "proven technology" :) Mar 06 13:48:00 pru him Mar 06 13:49:20 I have a NI sbRIO-9606. But I am looking for a "low-cost" version to listen to minimum 8 channels auf audiodata Mar 06 13:51:56 I'm confused, how to listen with toggling pins? Mar 06 13:52:10 gerolsprudel_: then tell us what you want to di Mar 06 13:52:12 do Mar 06 13:56:06 probably a physics lab, that uses some damn expensive NI equipment that needs "one more" and wants to build it cheap Mar 06 13:56:16 gerolsprudel_: if in doubt: use tin foil Mar 06 13:56:52 gerolsprudel_: unless you are using high power pulsed lasers, then it's a bad idea Mar 06 14:14:23 well, as you said i am now wearing tin foil on my head. but i dont know how it helped me to find a "low-cost" controller Mar 06 14:16:01 av500, nobody said it...why? to bitbang Mar 06 14:16:33 gerolsprudel_: as long as you do not state a problem, we cannot help you Mar 06 14:20:25 hmm ti goes moonshot Mar 06 14:21:14 omg Mar 06 14:21:45 http://www.theinquirer.net/inquirer/news/2252493/texas-instruments-supplies-hps-project-moonshot-with-quadcore-arm-chips Mar 06 14:22:22 I've just read something similar at the reg Mar 06 14:22:33 yes, they need a way to unload their DSPs somewhere Mar 06 14:22:54 those things still have dsps? Mar 06 14:23:02 yes Mar 06 14:23:04 it's TI Mar 06 14:23:08 i MUST have a DSP Mar 06 14:23:10 it Mar 06 14:23:31 they cannot put all DSPs to be JTAG adapters for something Mar 06 14:24:07 I wonder what they will be used for in servers Mar 06 14:24:48 doing packet inspection? Mar 06 14:25:12 awesome multichannel startup sound Mar 06 14:25:25 the great fw done by dsps Mar 06 14:25:36 (firewall not firmware) Mar 06 14:25:44 think Canyon.mid, but played by Karajan Mar 06 14:26:25 dsping the ethernet-stack Mar 06 14:26:51 istreamer instead of gstreamer Mar 06 14:28:09 a filter here, some packet forwarding there ... Mar 06 14:28:58 a loadbalancer inside dsps Mar 06 14:29:10 great stuff Mar 06 14:30:18 especially when dealing with large numbers like in ipv6 ;) Mar 06 14:31:42 you can easily overload many ciscos when adding v6-rules to the existing v4-rules ;) Mar 06 14:33:32 just that floating point doesn't make much sense for that Mar 06 14:35:15 if you don't go down to layer1 Mar 06 14:35:41 well, you always have to do load balancing and fair queueing and stuff Mar 06 14:35:51 which is application of control theory Mar 06 14:36:03 and a dsp makes a lot of sense there ;) Mar 06 14:36:11 ? Mar 06 14:37:00 you don't use highly complicated algorithms for such Mar 06 14:37:13 but you can Mar 06 14:37:17 av500, DSP is our hammer Mar 06 14:37:18 and you should Mar 06 14:37:27 otherwise a dsp doesnt make sense Mar 06 14:37:36 and managment doesnt know the solution is too complicated Mar 06 14:39:50 mdp: is it the hammer of all Ti or just a few of them? ;) Mar 06 14:40:35 KotH: all but the calculator folks Mar 06 14:40:39 KotH: some TIs..the others ones...my TI doesn't even make silicon Mar 06 14:41:00 mdp: of course, myti is a website :) Mar 06 14:41:09 myTI? isn't that the apache business? Mar 06 14:41:16 mytiborad.org Mar 06 14:41:32 oh how the myTI fall Mar 06 14:41:49 mdp: serious question, I'm trying to boot the BBB from USB from my laptop Mar 06 14:41:50 those are the evil registration collectors Mar 06 14:41:55 very often it power off Mar 06 14:42:06 dmesg does not tell me that the usb port overcurrented Mar 06 14:42:20 koen: ^^^^ Mar 06 14:42:21 hrm..I've never tried my BBB yet Mar 06 14:42:56 av500, Tartarus seems to talk a lot in "usb boot *should* work" type statements though Mar 06 14:43:07 well it works Mar 06 14:43:12 av500, have you tried usb boot with DC Mar 06 14:43:12 but its kinda erratic Mar 06 14:43:23 well, I dont boot from usb Mar 06 14:43:29 just power Mar 06 14:43:30 ? Mar 06 14:43:32 yes Mar 06 14:43:35 oh, ok Mar 06 14:43:38 av500: does it work with a usb power supply? Mar 06 14:43:48 KotH: no Mar 06 14:43:49 So, to me, the hard part of usb booting is the network manager killing, or alternate mojo Mar 06 14:43:52 hmm... Mar 06 14:44:03 av500: it's the pmic that shuts down Mar 06 14:44:05 my guess whould have been that the laptop cannot hanlde the power spikes Mar 06 14:44:07 I had intermittent issues with beagle white on usb power so I just went to DC all the time Mar 06 14:44:10 toss the right lines in /etc/network/interfaces, /etc/dhcp/dhcpd.conf and the right files in /tftpboot Mar 06 14:44:21 Tartarus, wrong question it turns out Mar 06 14:44:24 av500: the work around is to up the trip point in the pmic to above 500mA Mar 06 14:44:26 but yes, I'm DC power all the time Mar 06 14:44:36 Tartarus, he first mentioned usb boot..but didn't mean it the way I read it Mar 06 14:44:42 yeah sorry Mar 06 14:44:51 Tartarus: btw, I looked into fastboot and DFU Mar 06 14:44:51 I get it now ;) Mar 06 14:45:05 the fastboot patches apply now, but the am335x binding is not there Mar 06 14:45:18 DFU binding is also not there Mar 06 14:45:19 av500: ok Mar 06 14:45:27 Yeah, DFU is still in progress Mar 06 14:45:35 so, its not just about adding the right license headers Mar 06 14:45:38 (as in, I'm actively doing that this week/last week) Mar 06 14:45:43 ok Mar 06 14:45:59 av500: re fastboot, yes, there's some glue to be done, those patches were for I don't know what platform Mar 06 14:46:11 at91 Mar 06 14:46:13 or so Mar 06 14:46:14 but the "core" part of fastboot is there and license compatible which seems to be a big hurdle, or was in the past Mar 06 14:46:22 but they talk about the new vs old driver model Mar 06 14:46:29 and that's when they lost me Mar 06 14:47:17 so, I get the BBB to run off USB with no ETH plugged Mar 06 14:47:29 with ETH plugged, it power off very often Mar 06 14:47:36 damn, even without Mar 06 14:50:20 koen: how do I up the trip point? Mar 06 14:54:21 lsd Mar 06 14:57:27 ok, laptop usb port to dc in works Mar 06 14:57:30 never tried it, but I've read robert crumb when I was young ;) Mar 06 14:57:31 so its not the laptop Mar 06 14:57:44 but we knew that Mar 06 14:58:21 hmm, maybe I can jumper UBS 5V to DCIN Mar 06 14:59:05 av500: you don't have any of those fancy usb-to-barrel-plug cables? :) Mar 06 14:59:11 now I have Mar 06 14:59:16 but it blocks one USB port Mar 06 14:59:20 one is for serial console Mar 06 14:59:29 dont want to use a 3rd, its on the other side Mar 06 14:59:33 ah Mar 06 14:59:42 steal from the ftdi board? Mar 06 14:59:43 I am looking for a nice and small couch hacking setup Mar 06 14:59:49 yes Mar 06 15:00:00 or I jumper usb 5V to DCIN Mar 06 15:00:18 * dm8tbr votes for whatever is easier Mar 06 15:00:25 pretty is for products Mar 06 15:00:54 * av500 is sad that koen abandonned beagle for minnow Mar 06 15:00:55 av500: check kipisz' message on the alpha list Mar 06 15:01:14 koen: pls undo last line I wrote :) Mar 06 15:01:23 and damn, where is that list, why do I dont get mails? Mar 06 15:01:28 couch hacking sounds like a use case for the bt-serial-reset-debug cape ;) Mar 06 15:01:57 To change the current limit setting from 500 mA to 1300mA, I entered at the u-boot prompt Mar 06 15:01:58 i2c mw 0x24 1 0x3e. Mar 06 15:02:03 aholler: right, since I can tftp/nfs boot from anywhere Mar 06 15:02:18 aholler: in that case, ssh into the box with the uart is also ok Mar 06 15:02:29 av500, koen cited the quiet nature of pet minnows versus the endless howling by beagles Mar 06 15:02:47 aholler: in fact, I dont need BT there Mar 06 15:02:51 just ssh Mar 06 15:03:10 hello all. i was windering if i can use ds5 to debug a bealgleboard xm Mar 06 15:03:11 and remote reset Mar 06 15:03:24 * av500 prefers ds2 Mar 06 15:03:40 i want to build an image processing robot with motor control and checkpoint navigation Mar 06 15:04:03 preferably using the linux kermel Mar 06 15:04:21 koen is going fishing? Mar 06 15:04:28 the kernel has not robotics code inside Mar 06 15:04:34 also no image processing Mar 06 15:05:10 yeah, but can i not import the opencv library into my project in ds5? Mar 06 15:06:28 sure Mar 06 15:07:45 so will it work? i have found using the linux distro images gives me a lot of things i dont need. i want as minimal an install as possible and run the program Mar 06 15:08:45 omg, the minnowboard use angstrom Mar 06 15:09:17 surprise Mar 06 15:09:36 hack0r distro Mar 06 15:10:13 i dont want a desktop environment, i just want the file system and possible a camera driver Mar 06 15:10:17 possibly Mar 06 15:10:23 aholler: well, its easy, just install any linux distro and be done Mar 06 15:10:30 assuming there is an x86 port.... Mar 06 15:10:34 install openembedded and create you r own image with what you need Mar 06 15:10:45 prasad: worry about minimal later Mar 06 15:10:52 you have so many other things to worry about Mar 06 15:10:58 tackle these first Mar 06 15:11:02 can yocto help here? Mar 06 15:11:04 what av500 said Mar 06 15:11:06 av500: that angstrom just gave me clue about who is behind that community project ;) Mar 06 15:11:07 then worry about things you dont need Mar 06 15:11:17 aholler: in this case its intel Mar 06 15:11:28 i have already written the program and it runs well on my desktop but lags a lot on the baord itself Mar 06 15:11:28 the things you do not need are just charge on floating gtes Mar 06 15:11:32 they want to be loved badly Mar 06 15:11:39 prasad: then find out why Mar 06 15:11:49 hint: its not the "other" stuff in the distro Mar 06 15:11:57 aholler: you are missing out not being on #minnowboard Mar 06 15:12:17 mdp: I fear they will setup a password there soon Mar 06 15:12:43 hehe Mar 06 15:12:51 av500: yeah, im just a noob. My programming skills might not be optimal. but i have experience on cortex m processors and i feel ds5 might be an easier option for me Mar 06 15:13:03 than to profile my code on full blown linux Mar 06 15:13:12 fullHD Mar 06 15:13:13 what is ds5? Mar 06 15:13:21 ide from ARM Mar 06 15:13:38 prasad: how will ds5 help you? Mar 06 15:14:37 im assuming, the bare mettal linux image wont have the things i dont need. I hope to import the opencv library and the programs i have already written. and burn them directly onto the beagleboard Mar 06 15:14:39 panto, do you remember why you added a check for reduced blanking mode to the TI-LDCD driver? Mar 06 15:14:49 yes Mar 06 15:15:02 when i disable the check, i have the LDC7 cape working on 3.8 Mar 06 15:15:03 prasad: you are chasing after the wrong thing Mar 06 15:15:06 because non-reduced blanking modes did not work Mar 06 15:15:19 av500: what would you suggest? Mar 06 15:15:22 you have to optimize your image processing code Mar 06 15:15:35 shoragan, I bet the resolution is low enough Mar 06 15:16:08 av500: i will do that too. So you think ds5 wont help me too much? Mar 06 15:16:12 I think I also put a dt knob to enable the check there Mar 06 15:16:25 so no need to hack the code Mar 06 15:17:28 mdp: what do they talk about there? how to install distro xy? Mar 06 15:17:36 Lures Mar 06 15:18:09 av500: Would you suggest writing my own algorithms? Im currently using functions provided by the opencv library Mar 06 15:18:40 I'm actually curious about how to deal with a specific issue of creating platform devices and defining them in SSDTs..but they're not quite at that point on the s/w side I guess Mar 06 15:18:57 use perf top to look for hot spots Mar 06 15:19:11 aholler, it was kinda silent...I'm trying to decide what, if anything my st7735fb driver will need to be compatible... Mar 06 15:19:59 prasad: hint, openCV on the beagle is not as fast as on your core I7 desktop.... Mar 06 15:20:42 hmm, that might be the first x86-borad with usb-client support Mar 06 15:21:19 aholler: I had that since ages in my desktop Mar 06 15:21:23 prasad: talking about bare metal and linux in the same sentence is kind of an oxymoron Mar 06 15:21:35 av500: how that? Mar 06 15:21:46 aholler: PCI card Mar 06 15:21:52 prasad: a kernel always works on the bare metal hardware, but if you use a kernel, you are not bare metal anymore Mar 06 15:22:11 morning Mar 06 15:22:15 mourning even Mar 06 15:22:15 damn Mar 06 15:22:22 mdp, I bet no-one thought about those kind of problems Mar 06 15:22:26 howdy mranostay Mar 06 15:22:28 they're in for a surprise Mar 06 15:22:35 I dunno Mar 06 15:22:41 they have smart people there Mar 06 15:22:52 it's probably resource constrained like everywhere else Mar 06 15:22:54 dt for x86? Mar 06 15:22:54 reality sucks Mar 06 15:23:02 the world is filled with mistakes that smart people made Mar 06 15:23:05 av500: i understand that and i reduced the size of the camera frame to compensate slightly. I would just like to know if ds5 would give me any sort of advantage. Mar 06 15:23:14 * KotH doesnt make mistakes Mar 06 15:23:15 panto, google doesn't make mistakes...true story. ;) Mar 06 15:23:20 but then, i'm not smart either Mar 06 15:23:28 smart is overated Mar 06 15:23:32 *overrated Mar 06 15:23:43 clever is where it's at Mar 06 15:23:47 prasad: learn linux, not ds5 Mar 06 15:23:51 and *evil* Mar 06 15:24:10 mdp: it's mad and evil that go together, not clever and evil Mar 06 15:24:18 prasad: what do you imagine ds5 would do for you? Mar 06 15:24:19 koth: sorry, im not very clear about the terminology. Mar 06 15:24:21 mdp: especially when it comes to scientists Mar 06 15:24:24 guide your hand when writing code? Mar 06 15:24:30 KotH: quit watching movies Mar 06 15:24:36 :P Mar 06 15:24:51 prasad: bare metal = writing application code directly for the hardware with no OS or any other abstraction inbetween Mar 06 15:25:06 mdp: i don't watch movies, i watch animes Mar 06 15:25:11 :-) Mar 06 15:25:24 mdp: has more cute girls ;-) Mar 06 15:25:36 av500: i imagine the beagleboard would use less resource if i used the linux image provided on ds5 Mar 06 15:25:42 <_av500_> koen: what is the value to change it back? Mar 06 15:25:49 prasad, in a nutshell no Mar 06 15:25:57 mdp: I guess you have to learn acpi when dealing the minnow Mar 06 15:26:04 aholler: yes Mar 06 15:26:05 +with Mar 06 15:26:05 prasad: why? Mar 06 15:26:11 what resources do you see used? Mar 06 15:26:15 aholler, acpi... lol. Mar 06 15:26:20 aholler: their vague plan is to load SSDTs via initramfs Mar 06 15:26:27 what is ds5 anyways? Mar 06 15:26:33 what are SSDTS? Mar 06 15:26:34 Koth: thank you. but is the linux kernel an OS? Mar 06 15:26:42 oh boy, if people think DT is hard, wait until they have to use ACPI Mar 06 15:26:53 av500, a standardized data description ;) Mar 06 15:27:00 prasad: i guess you want to go back to college and take a course CS101 Mar 06 15:27:01 it's like DT, only a standard Mar 06 15:27:02 mdp: you already can load acpi patches Mar 06 15:27:19 mdp: cant I boot the minnow like any other PC? Mar 06 15:27:20 aholler: right..they're just talking about what their canon approach will be Mar 06 15:27:20 at least i've such kernel options Mar 06 15:27:21 av500: i imagine all the distributions have services running in the background Mar 06 15:27:29 prasad: and? Mar 06 15:27:42 +seen Mar 06 15:27:44 does that take away your CPU time? Mar 06 15:27:47 tbh, there's no much public talk yet..but it's new Mar 06 15:27:50 do you profile your app? Mar 06 15:28:05 sorry, i got disconnected. did panto reply to my question? :) Mar 06 15:28:17 I just hope minnowboard will sign fedora and ubuntu bootloader shims! Mar 06 15:28:33 av500: services i do not want. services which take up memory. Cycle time i do not think is too much of an issue Mar 06 15:28:35 sure they do, it has uefi Mar 06 15:28:42 aholler, definitely endless resources on how to patch your DSDT that's always broken..and loading SSDTs..lots of people do various things on broken x86 h/w Mar 06 15:28:52 btw. just got a bios update for my vulnerable laptop ;) Mar 06 15:29:01 shoragan, high resolution non reduced blanking modes did not work Mar 06 15:29:19 there's also dt knob to disable the check, so just turn it on Mar 06 15:29:31 prasad: it might be outdated, but you want to read tannenbaums book on operating systems Mar 06 15:29:37 prasad: and then you want to read a book on linux Mar 06 15:29:48 prasad: and then you might have a chance to answer your questions yourself Mar 06 15:30:10 prasad: i hate to say that, but you are asking questions like someone who has no understanding of how a computer works at all Mar 06 15:30:39 panto, the reduced blanking check uses hardcoded values. where do the come from? Mar 06 15:30:40 prasad, I would bet euros to donuts that the problem is in your code, and not in the distro/Linux taking away resources Mar 06 15:30:44 prasad: and doing a resource limited project with that kind of knowledge is a sure receipt for desaster Mar 06 15:30:58 shoragan, they come from the vesa standard Mar 06 15:31:11 ok Mar 06 15:31:56 ok. will persist with trying to optimise the code in distro. thank you Mar 06 15:32:04 so, it should be possible to calculate a matching RB mode for the LCD7, right? Mar 06 15:32:14 * KotH shakes his head Mar 06 15:32:32 shoragan, if you're using a fixed lcd (like LCD7) there's no need to use EDID parsing Mar 06 15:32:34 * av500 joins KotH Mar 06 15:32:37 do coders today learn only a programming language with no fucking background? Mar 06 15:32:48 KotH: yes Mar 06 15:32:53 just plug in the values to the panel node Mar 06 15:33:02 panto, the mode comes from the DT Mar 06 15:33:05 they learn java Mar 06 15:33:15 they learn something? Mar 06 15:33:18 KotH, they learn a programming language? are you sure? Mar 06 15:33:20 but tilcdc_crtc_mode_valid declares it MODE_BAD Mar 06 15:33:23 mru, hehe Mar 06 15:33:29 shoragan, then that's a bug :) Mar 06 15:33:36 aholler: even with java... you have to have some knowlege how the vm works and how it interacts with the OS Mar 06 15:33:39 k, let me look Mar 06 15:33:54 mru: sadly, that sounds true... Mar 06 15:34:05 mru: are you back from hk? Mar 06 15:34:05 KotH: not true, reality proves it's doable without such knowledge Mar 06 15:34:46 aholler: hmm.. i wonder how they manage Mar 06 15:34:57 KotH: no Mar 06 15:34:58 also the lcd7dts overlay is missing fragments to configure the PWM, taking the ones from lcd3-00a2 works Mar 06 15:35:02 i had to learn java and i had to get some understanding how the vm works to do anything non-trivial with that Mar 06 15:35:12 it's unbelievable with how less background knowledge you can becoma an expert Mar 06 15:35:44 shoragan, put 'ti,allow-non-reduced-blanking-modes;' in the DT node for now Mar 06 15:35:51 ok Mar 06 15:35:57 aholler: a big mouth is enough Mar 06 15:36:07 av500: you mean you missed trolling time in the ph0rum? :) Mar 06 15:36:20 aholler: the number of people who call me expert in various fields i have no clue about is a good proof of that Mar 06 15:36:32 I should take a look at your DT definitions too Mar 06 15:37:04 KotH, so you're not really an expert in middle age poetry? Mar 06 15:37:39 heh, the fritzing designer ; Mar 06 15:37:57 panto, in my studies, I found that the acpi code doesn't yet have provisions for creating platform devices from an SSDT arbitrarily. just specific ACPI classes have support for PC things. there's only one M$ proprietary CSRT parser that creates regular platform devices Mar 06 15:38:23 panto: nope... but i can speak fluently middle high german Mar 06 15:39:37 KotH: what is the difference in dialects? Mar 06 15:39:40 mdp, I haven't had the time to take a close look at SSDT, but my gut feeling is that the holes in functionality that embedded people tend to expect is going to be huge Mar 06 15:40:23 mranostay: the speaker Mar 06 15:40:30 they will struggle when it comes to certificates ;) Mar 06 15:40:51 mranostay: sometimes also the morphosyntactic usage of the word "err" and "ähm" Mar 06 15:41:10 panto, my gut sings the same thoughts Mar 06 15:41:24 i wonder if the minnow has an rtc ;) Mar 06 15:41:32 panto: but we can patch that in the board file, no? Mar 06 15:41:36 panto, somebody reminded me that this board just isn't intended for the same sorts of things Mar 06 15:41:48 * panto shrugs Mar 06 15:41:53 panto, probably best to stick with things interfaced via pci-e that will just work Mar 06 15:42:49 it's ironic really Mar 06 15:43:08 for ages the embedded people were looked down upon the x86 Linux devs Mar 06 15:43:24 and now x86 turns to embedded as an avenue for growth Mar 06 15:43:24 insulted Mar 06 15:43:25 panto: and it is our turn? :) Mar 06 15:43:57 panto: av500 looks down on everyone Mar 06 15:44:03 it is something... those guys will find out they're not in Kansas anymore Mar 06 15:44:06 from his ivory tower Mar 06 15:44:16 KotH, guarded by bears Mar 06 15:44:36 * mdp read that as "guarded by beers" Mar 06 15:44:39 awesome Mar 06 15:44:43 mdp: me too Mar 06 15:44:46 freud Mar 06 15:44:57 mranostay, s/w is hard, let's go drinking Mar 06 15:44:57 beers are awesome but terrible guards Mar 06 15:45:19 it is 7:45a :) Mar 06 15:45:52 not the 17th Mar 06 15:46:02 panto, take a look at https://github.com/jluebbe/beagle-kernel/commit/ea69fa6bca3e64cf365c862ddff612669959a2bc Mar 06 15:46:14 to the code mocking Mar 06 15:46:17 * mranostay clicks Mar 06 15:46:28 that should work, although i need to recompile and test in the final form Mar 06 15:46:32 mranostay, you're right, I'll get through the day with my "angry playlist" ;) Mar 06 15:47:07 shoragan: where are you setting the pinmux? Mar 06 15:47:13 shoragan, yeah, that's a bug Mar 06 15:47:30 in case of hardcoded timings the check should be disabled Mar 06 15:47:46 in my defense, I haven't got an LCD7 yet Mar 06 15:47:56 me neither Mar 06 15:47:59 koen: ^^^ Mar 06 15:48:01 but yeah, it should work Mar 06 15:49:12 mdp: what is the angry playlist today? Mar 06 15:49:47 some older green day atm Mar 06 15:49:58 when they were angry Mar 06 15:50:42 mranostay, the pinmux was already in that dt overlay Mar 06 15:51:34 although some color shown by fbtest flicker badly Mar 06 15:52:40 shoragan, hmm, I'll have to check when I have mine Mar 06 15:53:22 mranostay, over the wall Mar 06 15:55:03 koen: nvm, got the PDF Mar 06 15:55:03 mdp: engineering? Mar 06 15:55:42 hrmm, forgot my S-o-B :/ Mar 06 15:58:34 system on a breadboard? Mar 06 15:58:47 +1 Mar 06 15:58:55 mdp: breadboard is all you need Mar 06 15:59:09 +1 Mar 06 16:34:10 did anyone try the PVR-driver on 3.8 yet? Mar 06 16:36:09 doesn't work for am335x Mar 06 16:36:30 fbdev lacks vsync infrastructure Mar 06 16:36:53 and drm lacks dmabuff Mar 06 16:40:58 can mali help here? Mar 06 16:43:52 prpplague: hey purple Mar 06 16:46:22 mranostay: hey Mar 06 17:27:31 mdp: since I am stupid, what is "OF"? Mar 06 17:27:44 OpenFirmware Mar 06 17:28:08 open like democratic? Mar 06 17:28:12 or free? Mar 06 17:28:22 troll troll troll Mar 06 17:28:24 av500, before DT, DT was OF Mar 06 17:28:32 Open like Sun open Mar 06 17:28:36 like buzz Mar 06 17:29:23 panto, um, more to the point OF used/uses DT Mar 06 17:29:52 mdp, I know, I was around then :) Mar 06 17:30:04 just correcting any confusion..I know you know ;) Mar 06 17:30:12 DT is only the data storage model Mar 06 17:30:21 OF is a full blown bootloader/environment Mar 06 17:30:45 forth interpreter included Mar 06 17:30:56 IMO the problem with OF was the same as ACPI Mar 06 17:31:25 when you put in general computing capabilities in your firmware, you're just asking for trouble Mar 06 17:32:06 it has to be turing complere Mar 06 17:32:12 complete Mar 06 17:32:18 mru: no trip the china? Mar 06 17:32:57 av500, btw, it's ieee1275, just to give it more clout ;) Mar 06 17:34:01 mdp: I heard about clout storage.... Mar 06 17:34:11 I use personal clout Mar 06 17:34:24 storez warez Mar 06 17:38:45 i stick to old fashioned usb3-hds Mar 06 17:48:59 someone cries for nor-flash: http://www.theregister.co.uk/2013/03/06/in_memory_desktop_computing/ Mar 06 17:50:01 i should stop reading the interwebs :/ Mar 06 17:50:37 aholler: there are some CPUs now that run from spi flash Mar 06 17:50:43 so you dont need nor Mar 06 17:50:47 can be nand too Mar 06 17:51:00 sounds slow Mar 06 17:51:02 lots of them Mar 06 17:51:10 aholler, you haven't heard of quad spi? Mar 06 17:51:26 i live in a cloud Mar 06 17:51:30 no clout! Mar 06 17:52:14 av500: isn't what atmel offers since some time? Mar 06 17:52:21 +that Mar 06 17:52:37 54MB/s read on those parts Mar 06 17:52:40 maybe Mar 06 17:53:03 54mb/s isn't that fast compared with memory speed Mar 06 17:53:46 it's a NOR Mar 06 17:54:05 iirc x16 typical nor is 100ishMB/s Mar 06 17:54:16 I wonder how such a system would feel like if all (executed) code would be in flash Mar 06 17:54:19 trade-off for much less wires Mar 06 17:54:42 with 100mb/s Mar 06 17:54:51 aholler: it's designed for XIP too Mar 06 17:54:56 most modern spi flash designed spi controllers that do quad spi also do xip Mar 06 17:56:01 depends on what your use case is...lots of people have done XIP in nor Mar 06 17:56:17 yeah, but it might be interesting to see a (kde-)desktop living in xip Mar 06 17:56:27 why not xbmc? Mar 06 17:56:31 that's the real test of value Mar 06 17:56:44 for the crowd Mar 06 17:56:57 xmbc fits in my cache Mar 06 17:57:15 I kinda don't see the need for XIP nowadays, but what do I know.. Mar 06 17:57:40 panto, I wonder if we have at on our IP just because it's reused Mar 06 17:57:56 s/at/it/ Mar 06 17:58:22 but then looking at it..they did take care to make the addressable space much larger..so dunno Mar 06 17:58:56 it made sense 10 years ago Mar 06 17:59:05 now? not for linux certainly Mar 06 17:59:12 yeah Mar 06 17:59:14 maybe for uCLinux jobs Mar 06 17:59:31 don't disregard hack value Mar 06 17:59:42 mdp, right, 6502s for everyone! Mar 06 17:59:47 yes. get 40gb ram, move the distro into that on startup, done with still 30gb free for data Mar 06 17:59:49 * panto throws chips out like candy Mar 06 18:01:24 what would be needed is some statistics to what data cache first Mar 06 18:01:34 s/data/code/ Mar 06 18:01:42 I think win already does such Mar 06 18:01:59 you mean boot time access optimizations? Mar 06 18:02:06 macosx does too Mar 06 18:02:09 no, afterwards Mar 06 18:02:17 apllication caching Mar 06 18:02:33 aholler, isn't that what fusion drive of mac os x does? Mar 06 18:02:35 preloading the stuff you use Mar 06 18:02:48 no rotten fruits here Mar 06 18:03:32 I still managed to live without jesus-hw Mar 06 18:04:27 anyway, it's just another stupid article Mar 06 18:06:05 imho Mar 06 18:06:50 panto: this chips aren't edible... Mar 06 18:07:13 now there is a kickstarter idea.. edbile IC Mar 06 18:07:36 senile IC Mar 06 18:07:56 support the right to be forgotten Mar 06 18:08:06 s/be/get/ Mar 06 18:08:20 no right to party? Mar 06 18:14:42 one could use dsps to calculate the statistics used for refreshing dram to forget values ;) Mar 06 18:15:39 meh, one penguin too much ;) Mar 06 18:27:29 Jalapeno flavoured ICs? Mar 06 18:36:54 * mdp waits for pics of jalapeno ICs from shenzhen. Mar 06 18:37:09 howdy robclark Mar 06 19:12:07 I am designing a custom cape right now, that uses the PIN19/20 of P9 Mar 06 19:12:51 since the beaglebone kernel uses these pins to probe for official capes, I need to bypass this procedure Mar 06 19:13:18 how to bypass the cape probing and load a custom device-tree? Mar 06 19:13:46 why not just other pins? Mar 06 19:13:56 +use Mar 06 19:14:17 I need the dcan0 device Mar 06 19:14:51 can you give me a hint, where to look within the kernel sources Mar 06 19:15:03 I do not have a hardware-debugger Mar 06 19:15:18 thus it will be more or less printk debugging Mar 06 19:16:29 and you can't mux dcan0 to somewhere else? Mar 06 19:17:13 koen: does the beaglebone hae a shunt resistor somewhere inline with the 5V power supply? Mar 06 19:17:56 well, all other pins are used as well :-) Mar 06 19:19:35 mranostay: yes, hooked up to ADC 8 Mar 06 19:19:46 mranostay: but you have to tell the PMIC to mux that Mar 06 19:19:52 (yes, the pmic has a mux as well) Mar 06 19:20:02 initialization of the beaglebone seems to be here: http://arago-project.org/git/projects/?p=linux-am33x.git;a=blob;f=arch/arm/mach-omap2/board-am335xevm.c;h=02c8f9ff68a38a9ae64481ea124aeda19f75aebd;hb=HEAD Mar 06 19:20:33 koen: have exact steps? :) Mar 06 19:21:19 I have, but I can't find them Mar 06 19:21:27 ask jkridner Mar 06 19:24:10 koen: do you know where the eeprom probing starts in the arago-kernel? Mar 06 19:39:24 I got it, the file I mentioned is patched using the meta-ti tree, see http://arago-project.org/git/?p=meta-ti.git;a=blob;f=recipes-kernel/linux/linux-ti33x-psp-3.2/beaglebone/0003-beaglebone-rebase-everything-onto-3.2-WARNING-MEGAPA.patch;h=53f7355892b5276b8b74c74b84f60fdc753d2bd8;hb=master Mar 06 19:39:59 so I just need to remove i2c2_init and add all other init's that are needed by my hardware Mar 06 20:05:40 http://arstechnica.com/information-technology/2013/03/how-two-volunteers-built-the-raspberry-pis-operating-system/ Mar 06 20:05:43 ffs Mar 06 20:08:32 "a home-built cluster of ARM computers" ... effing Beowulf Eleven! Mar 06 20:09:00 I mean WTF? and they rely on this thing for a distro? Mar 06 20:12:21 somebody should port "TI-RTOS" to it Mar 06 20:13:59 did none of the rpi people come to the yocto talks? Mar 06 20:15:26 I am not sure what R-pi people look like Mar 06 20:15:26 and did they not see this either? http://wiki.debian.org/ArmHardFloatPort Mar 06 20:16:01 I think that is armv7 only though Mar 06 20:16:17 armv6 + vfp is "odd" Mar 06 20:16:33 wtf? rpi is armv6? Mar 06 20:16:40 er, yes Mar 06 20:16:53 Russ, that's a large part of its suckiness Mar 06 20:16:54 Russ, yes Mar 06 20:17:06 rpi marketing is not based in reality Mar 06 20:17:26 so its like an M0 Mar 06 20:17:32 apparently opencv runs better on rpi than omap3 Mar 06 20:17:44 wat Mar 06 20:18:05 ? Mar 06 20:18:43 I'd imagine it's more of a configuration/optimization issue Mar 06 20:19:06 well apparently a guy gave a talk at a robotics conf and made this claim Mar 06 20:19:22 https://upload.wikimedia.org/wikipedia/commons/0/0a/Acorn-ARM-Evaluation-System.jpg there it is, the core of the rpi, in all it's glory Mar 06 20:19:33 isn't the OMAP2 a v6 w/vfp? Mar 06 20:19:40 yep Mar 06 20:19:52 Russ: LOL Mar 06 20:24:04 Crofton|work: armv5 + vfp is odd, armv6 + vfp isn't odd Mar 06 20:24:32 odd in the sense that are not a lot of them around for hacking Mar 06 20:25:42 and to make one now, is definitely odd Mar 06 20:30:32 N800's and N810's! Mar 06 20:31:10 'night Mar 06 20:44:05 Crofton|work: claimed based on what? Mar 06 20:45:15 opencv? Mar 06 20:45:17 not sure Mar 06 20:45:29 I can see if the slides are available Mar 06 20:46:34 anyone know of a good thermally conductive potting resin Mar 06 20:47:16 molten lead Mar 06 20:47:18 maybe he used some vfp compiled stuff on the rpis and compared that with non-vfp on an armv7. or something similiar. Mar 06 20:48:15 hooray, I already have connectors to make my own lure Mar 06 20:48:24 who knows Mar 06 20:48:28 from a project 12 years ago Mar 06 20:48:42 I can ask for more details if people are really interested Mar 06 20:48:56 i have a copy of lure in the darkness Mar 06 20:50:28 no, lure of the temptress is it named Mar 06 20:51:18 http://en.wikipedia.org/wiki/Lure_of_the_Temptress Mar 06 20:51:41 jupp, that one Mar 06 20:51:43 is that a new Minnow accessory board? Mar 06 20:52:47 Russ: lead work take the heat away i'm sure, but there woud'nt be much heat there Mar 06 20:52:48 :-) Mar 06 20:54:28 mrpackethead, ah, so it needs to not be electrically conductive :p Mar 06 20:55:54 i have a small smps i need to bond to the case Mar 06 20:57:13 mrpackethead: there are tons of thermal conductive resins and glues Mar 06 20:57:27 mrpackethead: which one is the right one depends highly on your need Mar 06 20:57:50 and whether you can backe the stuff or not :) Mar 06 20:57:51 does it also need to do potting? Mar 06 20:58:30 jkridner: You on campus now.. ? Mar 06 20:58:37 jo Mar 06 20:58:40 Crofton|work: not really Mar 06 20:58:42 yes, theres many many to choose from thats for ure Mar 06 20:58:46 mrpackethead: oh.. and how fast do you need it ? :) Mar 06 20:59:46 * KotH had recently the problem of buying a elecitrically conductive epoxy that worked at room temperature, _RIGHT_NOW_... turned out that the only suppliers that had it, wouldnt ship it to switzerland Mar 06 20:59:48 if curing takes hours, its ok Mar 06 21:00:30 mrpackethead: i'm sure you have some supliers for mechanical stuff or series production Mar 06 21:00:34 mrpackethead: ask them Mar 06 21:00:43 mrpackethead: the usually have such stuff on stock Mar 06 21:00:49 yeah, i do.. and they do. Mar 06 21:01:03 so? where is the problem? Mar 06 21:01:06 there was the odd chance that someone like you had a good experience on a product Mar 06 21:01:28 a recommednation from someone who is not selling it, is a good thing Mar 06 21:01:32 chances are high that even if we would name one, you wouldnt be able to buy it Mar 06 21:01:49 that's why i said go to your assembly company Mar 06 21:01:59 they dont sell that stuff, they use it Mar 06 21:02:03 and know how well it works Mar 06 21:02:08 and for what kind of stuff Mar 06 21:02:28 Jacmet: thanks for assisting me in trolling the unresponsive DT folks Mar 06 21:02:31 and usually they are quite happy to give you a small sample or glue something for you Mar 06 21:03:05 mdp: ;) Mar 06 21:04:06 now that akpm is copied, he should be able to observe the silence too Mar 06 21:04:30 akpm? Mar 06 21:04:52 morton Mar 06 21:05:36 bed time for little KotH Mar 06 21:05:37 copied? Mar 06 21:05:39 night boys and trolls Mar 06 21:05:52 sleep tight and dont work too much! Mar 06 21:08:07 s/copied/cced Mar 06 21:09:15 later cheese Mar 06 21:09:29 I thought intel found some replacement ;) Mar 06 21:10:35 nite koth Mar 06 21:11:42 now I understamd the engagment for lures ;) Mar 06 21:15:01 mdp: I was at a TI presentation about the keystone stuff yesterday, with a nice 'full mainline support' slide ;) Mar 06 21:15:21 yeah, they have a separate team for those parts Mar 06 21:15:53 mdp: but ofcause, talk is cheap (I do remember seeing some patches posted last year) Mar 06 21:16:05 yeah, results are everything Mar 06 21:16:34 they have an influential person involved with them that has been able to steer things correctly Mar 06 21:16:46 everything is subject to resource availability of course Mar 06 21:16:46 mdp: that's great Mar 06 21:17:36 mdp: the devices are not really a good fit for what we want though, but it was nice to see Mar 06 21:17:45 this channel is a funny source of rumors ;) Mar 06 21:18:10 Jacmet: they're pretty targeted SoCs Mar 06 21:18:29 * Crofton|work wants a keystone board Mar 06 21:18:41 is there already one? Mar 06 21:18:41 Jacmet: I suppose they give the "everything's an XBMC client" crowd fits Mar 06 21:19:15 I want a $35 Keystone board ;) Mar 06 21:19:44 jupp, quad core with usb3 and sata Mar 06 21:20:05 besides gbit Mar 06 21:20:07 mdp: well, if you need to do video encoding, and it takes N expensive/power hungy keystone chips to do the same as a Netra then it becomes a hard sell Mar 06 21:20:18 sure Mar 06 21:20:25 but then they each have their own niche Mar 06 21:21:15 mdp: sure Mar 06 21:21:21 Jacmet: we should put out a slide about Netra's "full mainline support" Mar 06 21:21:25 that's easy to do Mar 06 21:21:37 * mdp waves hands Mar 06 21:21:43 heh Mar 06 21:22:28 mdp: yeah, I mentioned something about the less than perfect am335x mainline situation so long after it's introduction Mar 06 21:22:48 av500: Wow, that looks like a nice clean construction! Mar 06 21:23:24 Jacmet: hehe, nobody wants to hear about reality...especially sales Mar 06 21:23:56 ok, early exit today..ttyl...bbl Mar 06 21:24:10 * Jacmet likes asking the difficult questions in a room full of potential buyers ;) Mar 06 21:24:16 mdp: see ya Mar 06 21:25:04 I encourage it...the only way companies get better is if people call bullshit at the buyer end and it filters back to the factory Mar 06 21:25:12 ok..cya Mar 06 21:32:35 Jayneil: yeah, I'm at TI Forest Lane Mar 06 21:33:36 I'd just like them to publish the details of the PRU nicely. Mar 06 21:33:47 that doe'snt seem to be a big ask. Mar 06 21:33:49 jkridner: Ok... available for a meet up..? I am in the E wing.. Mar 06 21:35:31 mdp: before you go, let me give another use case for lua in kernel, unut tests ;) Mar 06 21:35:46 unit Mar 06 21:36:48 scripting at kernel level would make that possible Mar 06 21:37:59 I like brain storming ;) Mar 06 21:40:19 I've just remembered you searched about how to test in-kernel stuff Mar 06 21:45:47 mdp: are you snowed in? Mar 06 22:12:43 <_av500_> gmulak: ? Mar 06 22:13:16 jo av500 Mar 06 22:13:20 <_av500_> jo Mar 06 22:13:31 <_av500_> back from elternabend :) Mar 06 22:20:07 hehe Mar 06 22:20:10 we had last week Mar 06 22:20:55 _av500_: troll Mar 06 22:21:13 mranostay: beer Mar 06 22:22:03 Can someone help me with ARM SMP? Mar 06 22:25:39 no Mar 06 22:25:54 beware of the RCU stall Mar 06 22:25:56 So say I have two cores running a multithreaded kernel with user/kernel separation. Now say, core #1 needs to relocate (expand) the L1 table for a userland process. If, at the same time, core #2 is executing another thread from the process, it will still continue using the old translation table. How would I get around that? Mar 06 22:26:53 woglinde: beer trolls! Mar 06 22:29:47 :( Mar 06 22:32:04 krsitina maybee mru knows Mar 06 22:32:14 mranostay yes Mar 06 22:32:21 IPI and resync? Mar 06 22:32:41 IIRC, there should be code that does that sort of thing already Mar 06 22:33:13 ds2 he is not coding linux Mar 06 22:33:25 Oh Mar 06 22:33:44 well... IPI's are not linux specific Mar 06 22:33:47 they are arch specific Mar 06 22:34:24 what stands IPI for? Mar 06 22:34:35 Interprocessor Interrupt Mar 06 22:34:38 ah Mar 06 22:34:46 on x86, that should be part of the APIC Mar 06 22:34:58 on ARM, I think it is a CP (been a while) Mar 06 22:35:13 you need that to do stuff like signal the other core or even boot it Mar 06 22:35:22 boot it? Mar 06 22:35:32 yes boot it Mar 06 22:35:47 usually, only the main core boots up. it is up to the main core to bring up the other cores Mar 06 22:35:57 yes Mar 06 22:36:07 but its context switch Mar 06 22:36:09 would be more specific but it has been 5+ years since I dived in that part Mar 06 22:36:10 not booting Mar 06 22:36:27 context switch from reset? Mar 06 22:36:34 I call releasing the other cores as booting it Mar 06 22:36:41 or do you shutdown core btw context switch Mar 06 22:36:59 I dont know so I ask Mar 06 22:37:18 depends in implementation Mar 06 22:37:29 usually the other cores are held in reset til the main one releases it Mar 06 22:37:41 and then those cores jump to a known address (which is what i am calling boot) Mar 06 22:38:11 take a look at arch/arm/kernel/smp.c Mar 06 22:38:49 * woglinde wonders why kristina did not say anything anymore Mar 06 22:39:04 maybe reclusing to avoid GPL infection? ;) Mar 06 22:39:31 lol Mar 06 22:40:39 okay good nite Mar 06 22:40:44 ds2: In my scenario, it was implied that both cores are online and running. Mar 06 22:41:08 Kristina: yes, you still send an IPI to signal the other core. The other core can look at the message and do the required flushes Mar 06 22:48:03 Ah, seems reasonable. Mar 06 22:50:30 Relocating the TT is actually quite a tricky task with SMP. Mar 06 22:50:59 yes but usually the IPI is a very high priority interrupt Mar 06 22:51:38 But then that would require syncronizing everything so the other cores wait for the core that does the relocation. Mar 06 22:52:44 And in the meantime, the task should be marked as non-scheduelable. Mar 06 22:54:07 yes, it can be expensive Mar 06 22:55:54 Maybe allocating the whole TT at pmap creation would make more sense. Mar 06 23:24:53 * _av500_ wonders what user.txt is about Mar 06 23:25:06 <_av500_> since there is already uEnv.txt Mar 06 23:29:58 <_av500_> Tartarus: shouldnt tftpboot be called tftpload? Mar 06 23:30:03 <_av500_> since it does not boot Mar 06 23:30:51 _av500_: For consistency yes, but it's age-old named Mar 06 23:30:57 <_av500_> yes Mar 06 23:31:12 I just noticed today ext4write has a different syntax from fatwrite and cursed it's been like that too long to fix Mar 06 23:31:28 I suspect most people just call it as tftp Mar 06 23:31:37 <_av500_> u-boot-ng anybody? :) Mar 06 23:32:48 <_av500_> Tartarus: btw, I patched the linaor preEnv.txt in Mar 06 23:33:08 what now? Mar 06 23:34:09 <_av500_> ? Mar 06 23:34:12 I think user.txt is used when a button is pressed Mar 06 23:34:16 <_av500_> aholler: yes Mar 06 23:34:19 <_av500_> I know that Mar 06 23:34:28 it was a thing on older beagleboards, yes Mar 06 23:34:36 and Linaro has their own uEnv.txt file? Mar 06 23:34:52 <_av500_> Tartarus: no, a "pre" one Mar 06 23:34:57 <_av500_> under CONFIF_PREBOOT Mar 06 23:35:09 <_av500_> that allows me to overwrite internal bootcmd Mar 06 23:35:13 just a config setting, ok Mar 06 23:35:18 <_av500_> without recompiling uboot Mar 06 23:36:52 <_av500_> Tartarus: http://git.linaro.org/gitweb?p=boot/u-boot-linaro-stable.git;a=commitdiff;h=88da8539c1805bf6d19c8c9d2a0e34938958ce38 Mar 06 23:36:58 <_av500_> this one is exactly what I needed Mar 06 23:37:11 <_av500_> sinbce it runs *before* the bootcms Mar 06 23:37:20 <_av500_> bootcmd Mar 06 23:37:24 yeah Mar 06 23:37:38 I was going to say who do I have to yell at in Linaro to get them to try and upstream anything Mar 06 23:37:42 Then I recalled I know people there Mar 06 23:37:43 <_av500_> now I can use my preEnv.txt to setup tftpboot Mar 06 23:37:45 I should start yelling at them Mar 06 23:38:15 <_av500_> Tartarus: what I dont like is that this is not in 3 board files Mar 06 23:38:22 tftpboot would work with uEnv.txt too Mar 06 23:38:23 Indeed Mar 06 23:38:25 <_av500_> omap3, omap4 and am335 Mar 06 23:38:45 I've wanted to get a common "tries many things" file for ti parts for a while Mar 06 23:38:58 So everyone can opt in to uEnv.txt and preEnv.txt and all that Mar 06 23:39:03 <_av500_> yeah Mar 06 23:39:03 and not have N copies Mar 06 23:39:22 <_av500_> using grep for stuff in uboot reveals many copies of many things :) Mar 06 23:40:19 why you just don't add tftpboot to uEnv.txt? Mar 06 23:40:31 if that's all you want? Mar 06 23:40:39 <_av500_> its not all I want Mar 06 23:41:23 <_av500_> aholler: the thing is, uEnv.txt only gets read when the boot starts Mar 06 23:41:29 <_av500_> what if I want to change bootdelay itself? Mar 06 23:41:56 you recompile ;) Mar 06 23:42:06 <_av500_> yes, you keep saying that Mar 06 23:42:12 na, just a joke. ;) Mar 06 23:42:18 <_av500_> but I am joe the tinkerer that just built his bootdelay cape Mar 06 23:42:27 preEnv.txt makes sense Mar 06 23:42:31 <_av500_> so he wants to check a GPIO for that Mar 06 23:43:09 I was already happy to even get uEnv.txt upstream Mar 06 23:43:26 <_av500_> aholler: I understsnd that Mar 06 23:43:37 <_av500_> but in a way, its a regression vs boot.scr Mar 06 23:43:45 u-boot isn't a very friendly place to place patches Mar 06 23:44:10 my first version was downwards compatible Mar 06 23:44:25 <_av500_> since it's tightly coupled to what is compiled in Mar 06 23:46:33 boot.scr was called in bootcmd too Mar 06 23:46:43 s/called/loaded/ Mar 06 23:47:49 anyway, the preEnv.txt is a good thing to have. Mar 06 23:47:55 <_av500_> yes Mar 06 23:48:04 should be combined with the snipped I've posted yesterday Mar 06 23:48:17 if someone already touches the config Mar 06 23:49:30 here is it again: http://dpaste.de/s6Ccn/ Mar 06 23:49:54 did forget where I have that from, but it's nice Mar 06 23:50:08 <_av500_> aholler: http://pastebin.com/Aav0zN4v Mar 06 23:50:11 if that load will be replaced by env import -t Mar 06 23:50:24 <_av500_> the bootdelay 5 is only for demonstration :) Mar 06 23:50:48 oh, the trip upper is there ;) Mar 06 23:52:01 ide in the above snipped might be replaced by tftp and whatever else is supported by the board Mar 06 23:52:55 I'm doing here something similiar to test several partitions for the existence of uEnv.txt Mar 06 23:53:12 aholler: how will you prioritize them? Mar 06 23:53:27 the partitions? from first to last Mar 06 23:54:11 <_av500_> Tartarus: btw, has anybody looked into make uboot shut up a bit, like: http://pastebin.com/VWnRW103 Mar 06 23:54:12 but combined with preEnv.txt it makes more sense as you can change bootdevices and such there Mar 06 23:54:33 but where to get preEnv.txt from? catch-22 ;) Mar 06 23:54:38 interesting. Mar 06 23:55:08 I've kinda thought that user.txt should override uEnv.txt such that you can raise the priority of one above the others, but I really have nothing thought out. Mar 06 23:55:20 I'm using such especially to test several fs-types (fat, ext) Mar 06 23:55:56 because I'm regulary switching hd's on some devices Mar 06 23:56:23 or use sticks or similiar Mar 06 23:56:32 <_av500_> jkridner: well, using uEnvt.txt already mean you are overriding what is compiled in Mar 06 23:56:44 just write it to the onboard eeprom Mar 06 23:56:49 <_av500_> but then you could argue, uEnv.txt is "as shipped" and user.txt is for the tinkerer Mar 06 23:57:25 yeah, because we already override the compiled in defaults with uEnv.txt for "as shipped" defaults. :( Mar 06 23:57:38 you could -d for preEnv.txt Mar 06 23:57:43 +use Mar 06 23:57:53 to get rid of the compiled in stuff Mar 06 23:58:13 _av500_: you mean make -s ? Mar 06 23:58:17 Or using MAKEALL ? Mar 06 23:58:57 or just looking for the kernel-style CC/LD/etc ? Mar 07 00:00:00 I would just remove user.txt Mar 07 00:00:21 if someone wants to check the button, he could do that through using uEnv.txt Mar 07 00:00:32 by adding the check there Mar 07 00:01:35 <_av500_> Tartarus: kernel style Mar 07 00:01:58 nope, since most of the active folks use MAKEALL Mar 07 00:02:06 <_av500_> ah Mar 07 00:02:20 <_av500_> Tartarus: btw, is it be desing that make build everything all the time? Mar 07 00:02:25 <_av500_> by desing* Mar 07 00:02:40 No, you're the second person to report that rebuild is apparently broken Mar 07 00:02:46 <_av500_> ok Mar 07 00:02:48 or just massively crap Mar 07 00:02:51 must be something new Mar 07 00:02:54 not sure which Mar 07 00:03:06 it didn't do that a year before ;) Mar 07 00:03:16 <_av500_> well, not everything since I needed to make clean to get rid of some linker errors after my DFU tests Mar 07 00:03:30 <_av500_> so its neither nor Mar 07 00:03:56 <_av500_> neither nice nor thorough Mar 07 00:04:21 yeah, I'm a sucker for separate objdirs Mar 07 00:26:20 <_av500_> aholler: I think it would be better if there was a magic line in pre/uEnv that say "delete the env" instead of -d Mar 07 00:26:38 <_av500_> that way you can delete or append the env as you wish Mar 07 00:27:34 <_av500_> btw, I finally found a use case for my rpi Mar 07 00:27:45 <_av500_> I usedthe box to transport my BBB and cables :) Mar 07 00:27:51 <_av500_> use* Mar 07 00:35:28 prpplague: hey purple-o Mar 07 00:38:27 av500: that would imply to read the env twice Mar 07 00:38:59 if such a deleteEnvCommand is there Mar 07 00:39:06 but good idea Mar 07 00:40:30 env import uEnv.txt && test deleteEnvCommand && env import -d -t uEnv.txt Mar 07 00:40:45 something like that in working hush ;) Mar 07 00:41:13 device tree for u-boot configuration? :) Mar 07 00:41:37 since when is u-boot a device? Mar 07 00:41:49 greetings all Mar 07 00:43:12 at least I wouldn't want to configure those which are a device with dt ;) Mar 07 00:44:47 <_av500_> aholler: not read twice Mar 07 00:45:00 <_av500_> just make the first line in uEnv a DELETE_ENV enttry Mar 07 00:45:17 <_av500_> so anythingthat follows is from scratch Mar 07 00:45:32 that would need ugly code changes Mar 07 00:45:37 <_av500_> it would be implemented in env import Mar 07 00:45:40 <_av500_> i know Mar 07 00:45:43 <_av500_> so what Mar 07 00:45:45 <_av500_> uboot is ugly Mar 07 00:45:50 <_av500_> thats the "u" Mar 07 00:46:38 I don't see a problem just reading the env twice Mar 07 00:47:06 it comes from memory, so it's fast Mar 07 00:47:35 * mranostay blames Tartarus Mar 07 00:48:13 Does anyone know if beagleBone supports meteor? Mar 07 00:48:57 <_av500_> aholler: you are right Mar 07 00:49:19 Marz_: What is meteor? It seems too generic to turn up a useful Google hit. Mar 07 00:49:36 <_av500_> a russian thing Mar 07 00:49:44 Here is a link to their examples: http://meteor.com/examples/todos Mar 07 00:50:08 It's basically a framework for nodejs Mar 07 00:50:27 <_av500_> oh Mar 07 00:50:37 <_av500_> jkridner will take it from here Mar 07 00:50:54 * _av500_ slows backs away Mar 07 00:51:10 <_av500_> hmm, "hot code pushes" Mar 07 00:51:15 Lol Mar 07 00:51:20 <_av500_> mranostay: creep it Mar 07 00:51:34 * mranostay hurls a box of cards in here Mar 07 00:53:04 troll troll troll Mar 07 00:53:18 <_av500_> mru: did you go on the Shenzhen trip? Mar 07 00:53:22 hehe Mar 07 00:53:42 troll troll troll Mar 07 00:53:43 _av500_: yes Mar 07 00:53:45 Marz_: I'm not sure what it is built on, but if it just requires a C compiler and node.js, then you'd pretty well be good to go. Mar 07 00:53:45 bag full with a* and r* Mar 07 00:53:59 <_av500_> mru: what did you buy for me? Mar 07 00:54:05 I'm behind a firewall right now or I'd just try the install script. Mar 07 00:54:21 you did not place an order Mar 07 00:54:35 jkridner: enough with the javascript! Mar 07 00:54:37 <_av500_> little did I know you were going Mar 07 00:54:56 I think I'll have to create #beagle-javascript :-) Mar 07 00:55:12 nor did I until a few minutes prior Mar 07 00:55:20 mranostay: You like "Telemetry"... can't we use JavaScript for the "Internet of Things" :) Mar 07 00:55:37 Sweet, all I know is that it supports this: Linux: x86 or x86_64 Debian- and Red Hat-based systems Mar 07 00:56:03 do you know how big it is? Mar 07 00:56:07 java(script) runs everyhwere is fast and error free ;) Mar 07 00:56:07 And uses MongoDB Mar 07 00:56:16 Nope Mar 07 00:56:18 k. Mar 07 00:56:21 <_av500_> jkridner: about this big <-----------------------------------------> Mar 07 00:56:29 <_av500_> (not to scale) Mar 07 00:56:43 She must've been disappointed, lol Mar 07 00:56:45 I think the whole idea of doing installation as "curl http://xyz.com | sh" is really, really, really messed up Mar 07 00:56:53 * mranostay hands Marz_ a hard Mar 07 00:56:55 *card Mar 07 00:57:04 dammit Mar 07 00:57:05 * _av500_ hands mranostay a card Mar 07 00:57:10 * mranostay hands himself a card Mar 07 00:57:14 How come? Mar 07 00:57:17 lol jkridner Mar 07 00:57:20 i swear that was a typo Mar 07 00:57:25 <_av500_> right Mar 07 00:57:33 I'm a newbie at terminal/console stuff Mar 07 00:58:32 Marz_: well, the install script will certainly fail... Mar 07 00:58:39 hot code pushes? is that where you push GPLv3 code into other projects w/o notifying them? Mar 07 00:58:41 it does tests for x86_64 Mar 07 00:59:18 so, quick answer: no. Mar 07 00:59:54 :/ there would need to be a specific install script for beaglebone then? Mar 07 01:00:13 <_av500_> x86_64 and aarch64 is almost the same Mar 07 01:00:14 mranostay: keep your hard card in your pocket plesase Mar 07 01:00:19 Marz_: the "How come?" is about the sh thing? Mar 07 01:00:37 running shell scripts as root that you haven't read is insane. Mar 07 01:01:04 <_av500_> jkridner: its like clicking OK on website TOS, you dont read them either Mar 07 01:01:13 the blackbeagle is working. Mar 07 01:01:16 Lol Mar 07 01:01:20 and so is the red one Mar 07 01:01:22 and the green one. Mar 07 01:01:46 the TOS can only do theoretical harm in some hypothetical future. The sh script can do 'rm -rf /' :) Mar 07 01:01:51 _av500_: like the SGX agreement? ;) Mar 07 01:02:09 <_av500_> ds2: what agreement? :) Mar 07 01:02:11 jkridner: or even worse 'opkg remove nodejs' :) Mar 07 01:02:11 * jkridner heads to dinner. Mar 07 01:02:18 * _av500_ heads to bed Mar 07 01:02:19 mranostay: boo! :( Mar 07 01:02:23 first thing i do :) Mar 07 01:02:54 So a quick install of Meteor on a BeagleBone would break the beaglebone? Mar 07 01:03:06 <_av500_> totally Mar 07 01:03:07 no, the script would just complain and fail. Mar 07 01:03:08 yeah it burst into flames Mar 07 01:03:11 <_av500_> it would explode Mar 07 01:03:18 Boom crash Mar 07 01:03:24 <_av500_> big badaboom Mar 07 01:03:37 banging bits along the way Mar 07 01:03:39 remember the meteor landing in russia a few weeks ago? Mar 07 01:03:43 Marz_: if you can't tell, these guys aren't a big fan of the fact I'm a fan of JavaScript and node.js. :) Mar 07 01:03:52 <_av500_> ds2: that wasa failed install? Mar 07 01:03:59 that's from a quick "installing meteor" attempt Mar 07 01:04:04 They're missing out, nodejs is the future Mar 07 01:04:05 lol Mar 07 01:04:26 <_av500_> jkridner: can you gather your diciples and go somewhere else? Mar 07 01:04:42 <_av500_> like #beagle_nodejs_lalaland Mar 07 01:04:47 +1 Mar 07 01:04:54 <_av500_> we have uboot to fix Mar 07 01:05:01 <_av500_> and device trees to cut Mar 07 01:05:11 How could you create a web-based interface for the beaglebone? Mar 07 01:05:11 _av500_: you may want to repeat that :P Mar 07 01:05:20 That's responsive Mar 07 01:05:27 you open a socket, listen and react Mar 07 01:05:31 that's how! Mar 07 01:05:31 Marz_: don't use a Beaglebone :) Mar 07 01:05:39 opkg install apache2 Mar 07 01:05:41 mranostay, -1 Mar 07 01:05:49 Lol Mar 07 01:05:52 mine works just fine... :P Mar 07 01:06:07 * mranostay +1's himself Mar 07 01:06:32 nopkg install foo ;) Mar 07 01:06:34 <_av500_> mranostay: repeat what? Mar 07 01:06:50 <_av500_> jkridner: btw, are the BBB conf calls regular? Mar 07 01:06:57 Well, here's the deal, I'm setting up a storefront and I'm currently using an Arduino to light up my products via a webapp I installed, the problem I'm having is that the requests are being processed really REALLY slowly Mar 07 01:06:58 * _av500_ misses his TI calls Mar 07 01:07:17 Like 6 seconds to turn the lights on Mar 07 01:07:22 * ds2 sets _av500_ for bi weekly NV calls Mar 07 01:07:23 ;) Mar 07 01:07:39 <_av500_> ds2: great Mar 07 01:07:42 So I thought nodejs could make things work faster Mar 07 01:07:48 <_av500_> beats daily calls with china :) Mar 07 01:08:00 I donno about that Mar 07 01:08:07 lot of people to talk to in china Mar 07 01:08:25 <_av500_> bazillions Mar 07 01:08:26 I believe my team mate just opened up a socket, and that's where I am sending my requests Mar 07 01:08:33 i much rather talk to china. Mar 07 01:08:52 So what's better? Nodejs or just opening up a socket? Mar 07 01:08:55 <_av500_> ds2: how is T5 btw? Mar 07 01:09:06 how would I know? :D Mar 07 01:09:07 you an and your teammates shared sockets? ew sick... Mar 07 01:09:12 davidh_: does china listen? Mar 07 01:09:14 Lol Mar 07 01:09:32 you may want to get tested Mar 07 01:09:34 uups, ds2 was meant Mar 07 01:10:29 you expect people to listen?! Mar 07 01:10:33 demanding Mar 07 01:10:48 Do you? Mar 07 01:11:25 <_av500_> ok, im out, to much philosophy now... Mar 07 01:11:44 L8rz Mar 07 01:12:16 ds2: 我将讨论中国 Mar 07 01:12:44 eh, graphiti Mar 07 01:12:50 wtf? Mar 07 01:14:03 哈哈 Mar 07 01:14:07 哈哈 = lol Mar 07 01:14:19 first thing you'll learn in chinese IRC Mar 07 01:14:28 did i wander in #beagle-cn ? Mar 07 01:14:48 哈哈 sounds like ha ha Mar 07 01:16:05 im not sure if you can do hostnames in chiense yet Mar 07 01:16:31 it's utf Mar 07 01:16:35 ah i think you can Mar 07 01:18:06 inverted ? doesn't sound like anything Mar 07 01:18:21 whats inverted Mar 07 02:28:34 any scooop on beagle bone black? Mar 07 02:30:00 sssh, don't say that colour Mar 07 02:31:19 lol why is that? Mar 07 02:32:50 sssh, it's a secret **** ENDING LOGGING AT Thu Mar 07 02:59:58 2013