• [00:04:44] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) Quit ()
  • [00:05:42] <dcordes> mru: is there now way to force host or gadget mode?
  • [00:05:49] <dcordes> s/now/no/
  • [00:06:01] <dcordes> I mean other than using a host cable
  • [00:07:04] <mru> you can set it in the kernel config
  • [00:07:17] <mru> or by writing /sys/devices/platform/musb_hdrc.0/mode
  • [00:07:23] <mru> I think that's the file
  • [00:07:27] * rsalveti (n=salveti@189.70.119.44) Quit ("Bye :)")
  • [00:08:21] * rsalveti (n=salveti@189.70.119.44) has joined #beagle
  • [00:08:23] <dcordes> ok thanks
  • [00:08:37] <mru> I don't know what the valid values are
  • [00:08:40] <mru> try google
  • [00:09:32] <dcramer> I read somewhere that for the host cable to work the id pin has to be grounded
  • [00:09:52] <mru> that's for the auto-selection
  • [00:10:25] <dcramer> meaning ?
  • [00:10:34] <dcramer> selects host mode or gadget mode ?
  • [00:10:38] <mru> yes
  • [00:11:08] <mru> I think grounding the id pin selects host mode
  • [00:11:22] <mru> I haven't messed around with the usb stuff at all
  • [00:11:37] <dcramer> yes, that is how I read it
  • [00:11:38] <mru> ethernet gadget mode is fine for me
  • [00:12:09] <dcramer> ethernet gadget mode has to connect to a Host PC ?
  • [00:12:25] <mru> yes
  • [00:12:44] <dcramer> doesn't work great on a Mac
  • [00:13:03] <mru> if the mac runs linux it works...
  • [00:13:11] <dcramer> hehe
  • [00:13:26] <dcramer> so gadget mode works in linux ?
  • [00:13:47] <mru> yes
  • [00:14:04] <mru> if you have the cdc_ether module
  • [00:14:52] <dcramer> cool, thanks
  • [00:17:02] * ulso (n=Ulf@1-1-2-31a.rny.sth.bostream.se) has joined #beagle
  • [00:17:12] * Olipro__ is now known as Olipro
  • [00:18:33] <dcordes> mru: the fullscreen video playback is quite impressive
  • [00:18:57] <mru> try something with a bit higher resolution
  • [00:19:28] * haggisX (n=haggis@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) has joined #beagle
  • [00:26:41] * ulso (n=Ulf@1-1-2-31a.rny.sth.bostream.se) has left #beagle
  • [00:30:48] * ulso (n=Ulf@1-1-2-31a.rny.sth.bostream.se) has joined #beagle
  • [00:32:38] * ulso (n=Ulf@1-1-2-31a.rny.sth.bostream.se) Quit ("Leaving.")
  • [00:33:04] <dcordes> mru: can omapfbplay play ogg?
  • [00:33:21] <mru> depends on what's inside
  • [00:33:41] <dcordes> ogg theora video?
  • [00:33:45] <mru> should work
  • [00:33:50] <mru> haven't tried
  • [00:34:08] * jconnolly (n=jconnoll@ool-18b872bd.dyn.optonline.net) has joined #beagle
  • [00:34:13] <dcordes> I'm trying http://mirror.bigbuckbunny.de/peach/bigbuckbunny_movies/big_buck_bunny_720p_stereo.ogg then
  • [00:34:30] <mru> the mpeg4 avi plays fine
  • [00:34:37] * ulso (n=ulso@1-1-2-31a.rny.sth.bostream.se) has joined #beagle
  • [00:34:59] <mru> the leftmost column on the download page
  • [00:35:15] <mru> probaby better quality too
  • [00:35:21] <mru> theora isn't very good
  • [00:37:01] <dcordes> ok getting that instead
  • [00:37:06] * jconnolly (n=jconnoll@ool-18b872bd.dyn.optonline.net) Quit (Client Quit)
  • [00:37:36] <mru> that's the one we used at the demo in july
  • [00:37:42] <dcordes> mru: can you hint me on how to alter the resolution of Xfbdev in the demo?
  • [00:37:52] * haggisX (n=haggis@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) Quit ("Leaving")
  • [00:38:02] <dcordes> in that outdoor location? I've seen the photos
  • [00:38:37] <mru> it wasn't quite outdoor
  • [00:38:44] <mru> the entire area was roofed
  • [00:39:22] <dcordes> it looked like it was in some backyard or so
  • [00:39:43] * hagisbasheruk (n=hagisbas@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) has joined #beagle
  • [00:44:42] <dcordes> mru: that's shocking :D people, go throw your htpc boxes out of the window
  • [00:46:12] <vlad_> is it possible to set up uboot to auto-boot after a delay?
  • [00:46:31] <mru> yes
  • [00:46:47] <mru> it should all be documented somewhere
  • [00:46:51] <vlad_> sweet, asking google now
  • [00:47:43] <mru> I don't know how the shipping u-boot is configured
  • [00:48:13] <dcordes> vlad_: mru explained it for me earlier. it should be in the log
  • [00:48:18] <vlad_> ah good, bootcmd can be ;-separated
  • [00:48:21] <vlad_> great, thanks!
  • [00:48:42] <dcordes> yea bootcmd very useful
  • [00:48:54] * ap02 (n=abhijit@c-24-6-175-156.hsd1.ca.comcast.net) Quit ("Leaving")
  • [00:54:19] * [-ip-] (n=ts@mnhm-590f5259.pool.einsundeins.de) Quit ("and thanks for all the fish")
  • [00:59:53] <ulso> anyone using a USB/WiFi dongle?
  • [00:59:55] * cian (n=cian@cian.ws) has joined #beagle
  • [01:00:11] <dcordes> ulso: no but what's your problem/question?
  • [01:01:11] <ulso> I have tried a Netgear and a Linksys USB/WiFi dongle and they don't seem to be supported.
  • [01:01:51] <ulso> I was just wondering if WiFi via USB is supported.
  • [01:01:52] <dcordes> how did you figure?
  • [01:02:01] <dcordes> or try rather
  • [01:02:19] <dcordes> it also depends on the kernel you run
  • [01:02:27] <ulso> dmesg reports that the connected USB device (WiFi dongle) is not supported
  • [01:02:37] <ulso> I'm running angstrom-2008.1
  • [01:03:37] <dcordes> that's just the distribution. it can be ran on all sorts of kernels. where did you get the kernel image?
  • [01:03:42] <dcordes> is it the koen demo?
  • [01:03:47] <ulso> yep
  • [01:04:43] <ulso> 2.6.26-omap1 according to uname -a
  • [01:06:07] <dcordes> you need to find out the exact product names or usb id of the sticks and look up which driver they do require (if any)
  • [01:06:35] <ulso> ok, the VID and PID
  • [01:07:47] <ulso> think I'll check what's supported and try to get suitable hw
  • [01:07:55] <ulso> thanks
  • [01:08:04] <dcordes> no
  • [01:08:19] <dcordes> first see if the kernel you use maybe just lacks the required driver
  • [01:08:48] <ulso> ok
  • [01:08:49] <dcordes> if you get a new one I can suggest ralink rt73 chipset. my favourite
  • [01:11:58] * Olipro_ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [01:11:59] <ulso> that one looks interesting
  • [01:12:33] <dcordes> I have two of them, both broken :(
  • [01:12:42] * hagisbasher_ (n=hagisbas@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) has joined #beagle
  • [01:13:18] <ulso> seems to be use in a number of D-Link devices
  • [01:13:30] * cian (n=cian@cian.ws) Quit ()
  • [01:13:54] <dcordes> there are quite a few vendors
  • [01:14:40] <vlad_> is there any advantage to Xomap instead of Xfbdev? (asked yesterday, but my net died right after)
  • [01:19:25] * FWMiller (n=fwmiller@c-67-162-139-200.hsd1.co.comcast.net) has joined #beagle
  • [01:19:42] <FWMiller> greetings all
  • [01:20:30] <FWMiller> i recently got a beagleboard
  • [01:20:38] <FWMiller> i'm trying to do something with it
  • [01:20:51] <FWMiller> i don't have a serial cable with the ribbon connector
  • [01:20:56] <FWMiller> any advice?
  • [01:21:12] <mru> get a serial cable
  • [01:21:13] <davr> make one?
  • [01:21:21] <mru> you can't do much without one
  • [01:21:27] <FWMiller> ic
  • [01:21:36] <vlad_> (and you'll probably waste more time trying to do something without one than it would take you to go and get one)
  • [01:21:49] <FWMiller> k
  • [01:21:54] <FWMiller> so, another question
  • [01:22:03] <FWMiller> you can boot off an sd card?
  • [01:22:07] <dcordes> ulso: I just looked at the kernel configuration of the demo kernel (/proc/config.gz) and noticed there are not all wireless chipset drivers built.
  • [01:22:12] <mru> FWMiller: yes
  • [01:22:30] <dcordes> ulso: you need to rebuild and include the needed ones
  • [01:23:06] <FWMiller> i'm working on a small kernel
  • [01:23:10] <FWMiller> trying to boot it
  • [01:23:45] <dcordes> you won't boot it without serial cable
  • [01:23:53] <FWMiller> are there docs on the boot process
  • [01:24:04] <FWMiller> i.e. what does uboot do to boot a linux kernel?
  • [01:24:11] <FWMiller> i could follow that process i spose
  • [01:24:15] <FWMiller> off an sd card
  • [01:24:26] <dcordes> if you buy an adapter like they suggest in the doucmentations, for example in the hardware reference manual, be warned! make sure the d-sub is wired correctly
  • [01:24:28] * cian (n=cian@cian.ws) has joined #beagle
  • [01:25:46] <dcordes> FWMiller: http://code.google.com/p/beagleboard/wiki/LinuxBootDiskFormat and here you can see how the demo image is to be booted from sd http://www.angstrom-distribution.org/demo/beagleboard/README.txt the used images are in http://www.angstrom-distribution.org/demo/beagleboard/
  • [01:28:31] * Olipro (n=Olipro@unaffiliated/olipro) Quit (Connection timed out)
  • [01:29:04] * hagisbasheruk (n=hagisbas@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) Quit (Read error: 110 (Connection timed out))
  • [01:33:29] <FWMiller> thanks!
  • [01:34:02] * gulick (n=gulick@rrcs-208-125-50-2.nys.biz.rr.com) Quit (Read error: 104 (Connection reset by peer))
  • [01:39:45] <joshin> Not directly beagle board related but Pandora has a release date - Nov 13th
  • [01:40:35] <dcordes> joshin: nice. are you working with the pandora handheld?
  • [01:41:37] * beagle-dave (n=tux@adsl-69-226-208-117.dsl.pltn13.pacbell.net) has joined #beagle
  • [01:42:02] <beagle-dave> hello, world\n
  • [01:42:39] <joshin> I hope to be. Right now I'm blindly updating my NSLU2 Gentoo tools to build for the 3530
  • [01:43:10] <joshin> I'm just looking for an excuse to buy a beagle board. :)
  • [01:43:12] * Olipro_ is now known as Olipro
  • [01:49:14] * hagisbasher__ (n=hagisbas@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) has joined #beagle
  • [01:54:57] <beagle-dave> So ... Pandora release is 13-Nov. That's hardware to early adopters? Mass availability?
  • [01:55:06] <beagle-dave> A dusting of open source solstice presents?
  • [01:56:10] * cian (n=cian@cian.ws) Quit ()
  • [01:58:55] <hagisbasher__> So whats the Price of Pandora on 13 th Nov ?
  • [02:01:00] <joshin> $330usd depending on the exchange rate and there should be 3000 units available
  • [02:01:09] * cian (n=cian@cian.ws) has joined #beagle
  • [02:01:37] <hagisbasher__> ouch
  • [02:01:46] <joshin> They're doing pre-orders in late September and depending on the demand may kick off another run to have supplies for Christmas.
  • [02:02:30] <joshin> That display, chassis, battery, keyboard, and so on are not free. The price seems reasonably fair... I'd love it for $2.75 but that's not likely. :)
  • [02:02:50] <hagisbasher__> me too :)
  • [02:03:31] <Navi> I'd totally be fine with paying for it if they didn't go with that clamshell design
  • [02:06:04] * hagisbasher_ (n=hagisbas@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) Quit (Read error: 110 (Connection timed out))
  • [02:11:11] * ulso (n=ulso@1-1-2-31a.rny.sth.bostream.se) Quit ("Leaving.")
  • [02:19:52] <dcordes> DISPLAY=:0 omapfbplay -f -b 80 big_buck_bunny_720p_surround.avi
  • [02:20:02] <dcordes> oops
  • [02:26:04] <hagisbasher__> XXX.avi lol
  • [02:26:16] * cian (n=cian@cian.ws) Quit ()
  • [02:26:43] <hagisbasher__> 3.26am , goodnight folks
  • [02:26:50] * hagisbasher__ (n=hagisbas@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) Quit ("Leaving")
  • [02:30:42] <dcordes> did somebody encounter errors with usb mass storage?
  • [02:39:13] * dcordes (n=dcordes@unaffiliated/dcordes) has left #beagle
  • [02:50:31] * cian (n=cianh@cian.ws) has joined #beagle
  • [03:00:34] * ulso (n=ulso@1-1-2-31a.rny.sth.bostream.se) has joined #beagle
  • [03:23:01] <vlad_> installing debian via qemu, thinking that it would be faster than doing it to a SD card on a device, was really a bad move
  • [03:30:41] * harper (n=harper@24-148-50-204.stk-bsr1.chi-stk.il.cable.rcn.com) has joined #beagle
  • [03:31:17] <harper> hello. when i use minicom to connect to a virgin beagleboard - what settings should i use? i am getting a bunch of junk thus far
  • [03:34:02] <vlad_> 115200n81, no flow
  • [03:37:33] * ghaz`` (n=kvirc@e178100135.adsl.alicedsl.de) Quit ("KVIrc 3.2.0 'Realia'")
  • [03:46:26] <harper> ok.
  • [03:47:55] <harper> that is what i have it set to, but its just noise
  • [03:49:09] * JoeBorn (n=jborn@adsl-75-2-247-191.dsl.chcgil.sbcglobal.net) Quit (Read error: 104 (Connection reset by peer))
  • [03:50:48] <harper> ok. whew. the rs232 cable was bad. a new one and now i have shell ;)
  • [03:50:49] <harper> thanks
  • [03:51:25] * bazbell (n=a0192809@nat/ti/x-85a2e9704966bbed) Quit ("Leaving.")
  • [04:00:46] * Olipro_ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [04:13:24] * Olipro (n=Olipro@unaffiliated/olipro) Quit (Nick collision from services.)
  • [04:13:27] * Olipro_ is now known as Olipro
  • [04:21:11] * Beagle4 (n=Beagle4@cpe-76-185-171-53.tx.res.rr.com) has joined #beagle
  • [04:21:12] * harper (n=harper@24-148-50-204.stk-bsr1.chi-stk.il.cable.rcn.com) Quit ("Ex-Chat")
  • [04:22:32] * Beagle4 (n=Beagle4@cpe-76-185-171-53.tx.res.rr.com) Quit (Client Quit)
  • [04:30:45] <vlad_> grr, something on my new root image isn't liked
  • [04:33:20] * cian (n=cianh@cian.ws) Quit ()
  • [05:22:48] * dirk2 (n=dirk@F313d.f.strato-dslnet.de) has joined #beagle
  • [05:34:01] * FWMiller (n=fwmiller@c-67-162-139-200.hsd1.co.comcast.net) Quit ()
  • [05:42:57] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [05:48:37] <dirk2> khasim: hi
  • [05:52:03] <khasim> dirk2: hi
  • [05:55:30] * FWMiller (n=fwmiller@c-67-162-139-200.hsd1.co.comcast.net) has joined #beagle
  • [05:58:52] <dirk2> khasim: Have you found already some time to look into U-Boot V1 NAND?
  • [06:00:16] * FWMiller (n=fwmiller@c-67-162-139-200.hsd1.co.comcast.net) has left #beagle
  • [06:07:09] <khasim> dirk2: nope,
  • [06:15:52] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [06:27:49] * bazbell (n=a0192809@nat/ti/x-7cf09a17a7ee4188) has joined #beagle
  • [06:36:53] <sakoman_> dirk2: not sure if you noticed, but I have a mirror of the u-boot git on gitorious:
  • [06:37:08] <sakoman_> http://gitorious.org/projects/u-boot-omap3/repos/mainline/logs/test
  • [06:41:49] <dirk2> sakoman_: You mentioned it some days ago. Should we use this instead of the one at www.sakoman.net ? I like the gitweb interface at www.sakoman.net, though :)
  • [07:09:14] * haggisX (n=haggis@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) has joined #beagle
  • [07:37:18] * ulso1 (n=ulso@1-1-2-31a.rny.sth.bostream.se) has joined #beagle
  • [07:39:22] * ahu (i=ahu@outpost.ds9a.nl) Quit (brown.freenode.net irc.freenode.net)
  • [07:39:22] * ds2 (i=noinf@netblock-66-245-251-24.dslextreme.com) Quit (brown.freenode.net irc.freenode.net)
  • [07:39:22] * turqy (n=turqy@adsl-75-9-35-139.dsl.rcsntx.sbcglobal.net) Quit (brown.freenode.net irc.freenode.net)
  • [07:40:15] * ahu (i=ahu@outpost.ds9a.nl) has joined #beagle
  • [07:40:15] * turqy (n=turqy@adsl-75-9-35-139.dsl.rcsntx.sbcglobal.net) has joined #beagle
  • [07:40:15] * ds2 (i=noinf@netblock-66-245-251-24.dslextreme.com) has joined #beagle
  • [07:41:46] * ulso (n=ulso@1-1-2-31a.rny.sth.bostream.se) Quit (Read error: 60 (Operation timed out))
  • [07:46:58] * dcordes (n=dcordes@unaffiliated/dcordes) has joined #beagle
  • [07:57:32] * haggisX (n=haggis@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) Quit (Read error: 110 (Connection timed out))
  • [08:04:16] <vlad_> great success.. debian root image works fine with a new kernel, now just gotta rebuild and enable networking drivers (whoops)
  • [08:25:56] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) has joined #beagle
  • [08:29:22] <dirk2> vlad_: What was the issue? Maybe we should add it to the wiki (and close the mail thread with the info)
  • [08:34:26] <vlad_> dirk2: yeah, was gonna update the wiki, but figured I'd do it tomorrow when I was more awake :)
  • [08:34:38] <vlad_> not sure what the actual issue is, though
  • [08:35:01] <vlad_> I just know that a stock omap kernel seems to work fine
  • [08:39:26] <vlad_> tomorrow I get to fight with X
  • [09:01:17] * Olipro_ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [09:17:49] * Olipro (n=Olipro@unaffiliated/olipro) Quit (Read error: 110 (Connection timed out))
  • [09:30:58] * valhalla (n=valhalla@81-174-21-16.dynamic.ngi.it) has joined #beagle
  • [09:32:40] * bazbell (n=a0192809@nat/ti/x-7cf09a17a7ee4188) Quit (Remote closed the connection)
  • [10:22:08] <dirk2> OSGi on the BeagleBoard, success! : http://osgifun.blogspot.com/2008/08/osgi-on-beagleboard-success.html
  • [11:09:12] <mru> morning
  • [11:19:40] <ahu> Aug 17 13:18:44 Too many (15 > 2) bogus answers for 'gaaaaaaa.ds9a.nl.' from 213.156.2.1, assuming spoof attempt.
  • [11:19:43] <ahu> wce, sorry
  • [11:20:54] <mru> kaminsky attack?
  • [11:23:50] <ahu> no, my own work - (this is powerdns output, I work on powerdns)
  • [11:23:57] <ahu> but it is indeed a 'kaminsky style' attack
  • [11:24:06] <ahu> but I really didn't want to paste it here, sorry about that
  • [11:39:45] * cian (n=cian@cian.ws) has joined #beagle
  • [11:55:55] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [12:01:05] * cian (n=cian@cian.ws) Quit ()
  • [12:01:21] * saablover (n=bart@78-21-97-209.access.telenet.be) has joined #beagle
  • [12:01:49] <saablover> hi people, the audio interface on the board is spdif ?
  • [12:02:07] <mru> no
  • [12:02:10] <mru> analogue only
  • [12:02:22] <saablover> damn
  • [12:02:24] <mru> spdif would need an additional chip
  • [12:02:38] <saablover> would be nice board for mpd appliance
  • [12:02:45] <mru> it might be possible to attach one to the expansion port
  • [12:02:57] <saablover> let's have a look
  • [12:03:01] <mru> iirc mcbsp is available there
  • [12:03:26] * cian (n=cianh@cian.ws) has joined #beagle
  • [12:03:35] <saablover> mcbsp ?
  • [12:03:54] <mru> multi-channel buffered serial port
  • [12:04:07] <mru> a serial interface capable of high speeds
  • [12:04:35] <saablover> aha
  • [12:06:04] <mru> yes, it's available on the expansion port
  • [12:06:13] <mru> two of them even
  • [12:07:01] <saablover> do DAC exist that have this interface ?
  • [12:07:12] <mru> for spdif you don't use a dac
  • [12:07:14] <saablover> and linux support ? :)
  • [12:07:37] <saablover> or usb expansion
  • [12:07:40] <saablover> box
  • [12:07:51] <mru> usb audio should work
  • [12:07:57] <mru> insofar usb audio works at all
  • [12:07:59] <saablover> would be cheaper I think
  • [12:08:09] <saablover> why not ?
  • [12:08:17] <saablover> you have great usb thingies
  • [12:08:23] <mru> usb is a terrible interface
  • [12:08:56] <saablover> explain
  • [12:11:07] <mru> you could use something like a TI DIX4192 to get spdif
  • [12:11:12] <mru> http://focus.ti.com/docs/prod/folders/print/dix4192.html
  • [12:11:49] <mru> free samples available ;-)
  • [12:18:06] * dirk2 (n=dirk@F313d.f.strato-dslnet.de) has left #beagle
  • [12:30:17] * root (n=root@rrcs-208-125-50-2.nys.biz.rr.com) has joined #beagle
  • [12:30:38] * root is now known as Guest62082
  • [12:35:53] <saablover> hehe god under us :)
  • [12:36:38] <ahu> ROOOOOOOOT has landed!
  • [12:37:16] <saablover> Guest62082, should we hack you ?
  • [13:21:20] * cian (n=cianh@cian.ws) Quit ()
  • [13:48:48] * bazbell (n=a0192809@nat/ti/x-aef9f8042a125da5) has joined #beagle
  • [14:01:15] * Olipro__ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [14:17:54] * Olipro_ (n=Olipro@unaffiliated/olipro) Quit (Read error: 110 (Connection timed out))
  • [14:36:08] <jkridner> good morning all
  • [14:37:28] <mru> morning jkridner
  • [14:39:34] <jkridner> gmane seems to be getting none of the mailing list messages.
  • [14:59:11] <mru> regarding the sgx drivers, I suppose it would be feasible to RE them enough to use without linux...
  • [15:00:29] <jkridner> with so much of them being binary, I think it'd be difficult, but you could find out whatever external calls were made and satisfy them.
  • [15:00:50] <jkridner> you'd know what was going on in whatever the binaries called, since the rest of the system is open.
  • [15:01:11] <jkridner> if you have to tell someone it is possible, then it probably isn't possible for them.
  • [15:01:21] <mru> yes, I'm talking about interfacing the binaries as a blackbox
  • [15:01:28] <mru> true
  • [15:01:45] <mru> reverse engineering the entire drivers is probably quite tricky
  • [15:01:47] <jkridner> I don't like to encourage people to waste their time.
  • [15:01:50] <mru> how big are they?
  • [15:02:02] <jkridner> still, I agree with you. It should be quite possible.
  • [15:03:58] <jkridner> I haven't had a good look and I don't have a file system with them handy.
  • [15:12:55] <jkridner> I'm still facing my GCC woes.
  • [15:13:34] <jkridner> no older builds seem to exist on the angstrom servers.
  • [15:14:11] <dcordes> I rebuilt the beagle image yesterday. could easily include task-native-sdk
  • [15:14:33] <dcordes> but building for kaiser now
  • [15:14:50] <jkridner> k. I'll try an 'opkg update'.
  • [15:15:24] <dcordes> I didn't updated the feed or anything, I just made a local build of the image
  • [15:15:44] <jkridner> k. how does the feed get updated?
  • [15:15:46] <dcordes> you must bug koen or the openembedded bugtracker about it if it is broken in the repositories
  • [15:16:08] <jkridner> k.
  • [15:16:12] <dcordes> I think the autobuilder does but not to sure
  • [15:16:39] <jkridner> I've been trying to rebuild, but my builds keep failing. just did another mtn update and will see if I can build again.
  • [15:17:18] <sakoman> jkridner: what part of the build is failing?
  • [15:17:30] <dcordes> I also had problems yesterday. package gnome-games did not build so I had to include it from the image install
  • [15:17:44] <dcordes> (in packages/images/beagle-demo.ipk or so)
  • [15:17:45] <jkridner> NOTE: Task failed: /mnt/stuff/tmp/work/armv7a-angstrom-linux-gnueabi/glibc-2.6.1-r10/temp/log.do_compile.20684
  • [15:18:17] <dcordes> do you build the beagle image?
  • [15:19:44] <dcordes> NOTE: package glibc-2.6.1-r10: task do_compile: completed
  • [15:20:11] <jkridner> yes, I build beagleboard-demo-image
  • [15:20:28] <sakoman> my beagle-demo-image builds have been failing on omapfbplay. I'm using the latest linux-omap kernel, it fails because of the header changes: omapfbplay.c:40:38: error: asm-arm/arch-omap/omapfb.h: No such file or directory
  • [15:20:29] <jkridner> is the right name 'beagle-demo'?
  • [15:20:38] <dcordes> I updated OE and built beagleboard-demo-image from scratch yesterday and only gnome-games failed
  • [15:20:38] <sakoman> beagle-demo-image
  • [15:21:20] <dcordes> ls packages/images | grep beagle
  • [15:21:20] <dcordes> beagleboard-demo-image.bb
  • [15:21:50] * jkridner just found that 'opkg install rsync' works. Now I have a way to make nice backups of my running file systems.
  • [15:22:22] <sakoman> dcordes: you are correct of course, beagleboard not beagle
  • [15:26:03] <dcordes> the count of patches required in OE to build linux-omap2 fore beagle looks insane
  • [15:26:15] <dcordes> is there no own branch for beagleboard?
  • [15:26:58] <jkridner> it shouldn't need one.
  • [15:27:10] <jkridner> there was a bit of a queue while Tony was out...
  • [15:27:25] <jkridner> but the Beagle fixes should be able to go into the main tree.
  • [15:27:43] <jkridner> I'm not aware if there is a really bit set of patches that won't (other than the power mangement).
  • [15:28:04] <jkridner> the power management patches are big and probably won't make it mainline too fast.
  • [15:28:10] <ahu> OE?
  • [15:28:16] <jkridner> that's probably what you are seeing.
  • [15:28:23] <jkridner> OE=open embedded
  • [15:29:17] <dcordes> yes there is some power management stuff
  • [15:29:23] <dcordes> org.openembedded.dev/packages/linux/linux-omap2-git/beagleboard
  • [15:32:02] <jkridner> looking at the patches now.
  • [15:32:11] <jkridner> the ASoC stuff still has bugs.
  • [15:32:45] <jkridner> I haven't heard about the oprofile stuff in a while. maybe that is good for going upstream now?
  • [15:33:53] <jkridner> lots of omappm patches in there.
  • [15:34:01] * dirk2 (n=dirk@F32d5.f.strato-dslnet.de) has joined #beagle
  • [15:37:35] <chakie> reading the mailing lists i see that there is no hw accelerated 3d yet. that's fine by me, i don't need that. but it also seems video decoding up to 720p is working fine
  • [15:38:38] <chakie> so one could assume that the ffmpeg port (or similar) is good and that video support is good?
  • [15:40:03] <thomasg> is the DSP for video-decoding supported by open source drivers?
  • [15:40:22] <dcordes> I tried it yesterday and I think it's quite nice
  • [15:40:52] <jkridner> there are now open source drivers for loading code onto the DSP.
  • [15:41:02] <thomasg> great
  • [15:41:08] <jkridner> you have 2 options: DSP/BIOS Link and DSP Bridge.
  • [15:41:34] <jkridner> none have been merged into the Linux kernel tree yet.
  • [15:42:36] * chakie doesn't mind some patch hunting and manual patching if the result is good :)
  • [15:42:40] <jkridner> it is a more complicated picture than I'd like, but there are some starting points.
  • [15:43:07] <jkridner> there aren't any open source codecs for the DSP yet, that I know of.
  • [15:43:16] <jkridner> There is on-going work for Theora.
  • [15:43:27] <mru> theora, what a waste of time
  • [15:44:34] <jkridner> http://wiki.neurostechnology.com/index.php/Summer_of_Code_2008/Ogg_Theora_Codec
  • [15:44:45] <jkridner> mru: content or quality?
  • [15:45:09] <dcordes> does that duck wearing headphones puke?
  • [15:45:45] <dcordes> it looks quite funny :)
  • [15:46:08] <mru> theora is comparable to mpeg1, if even that
  • [15:46:23] <mru> and hardly anything supports it
  • [15:46:30] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [15:47:31] <jkridner> k, so content *and* quality.
  • [15:48:17] <jkridner> so, when do you get started on MPEG4? ;)
  • [15:48:54] <jkridner> sakoman: http://tinderbox.openembedded.net/public/logs/771281.txt
  • [15:49:51] <dcordes> arm-angstrom-linux-gnueabi-gcc: No such file or directory
  • [15:50:32] <jkridner> no tools in my path.... odd.
  • [15:51:02] <jkridner> shouldn't bitbake put them in the path?
  • [15:55:36] <jkridner> ./work/armv7a-angstrom-linux-gnueabi/gcc-cross-4.3.1-r8/staging-pkg/cross/bin/arm-angstrom-linux-gnueabi-gcc
  • [15:56:21] <dcordes> no clue I never had this problem
  • [15:56:57] <dcordes> also I wondering why you get the libc error with uptodate metadata. It build fine for me a few hours ago. maybe you should clean tmp?
  • [15:58:24] <jkridner> what is the cleanest way to add packages to the demo image?
  • [15:59:22] <thomasg> copy the image-reciepe you're using, rename it to demo-image...bb and add the apps in the reciepe
  • [15:59:27] <jkridner> k.
  • [16:08:51] * saablover (n=bart@78-21-97-209.access.telenet.be) Quit ("Ex-Chat")
  • [16:12:54] <dcordes> jkridner: is there a way to play audio in the demo image?
  • [16:13:15] <jkridner> there is, but OE builds a broken ALSA driver.
  • [16:13:31] <dcordes> I get a kernel panic trying to play any sound
  • [16:13:44] <dcordes> even mplayer -ao oss doesn't work
  • [16:13:44] <jkridner> yes, that is the state. :(
  • [16:14:24] <jkridner> the code.google.com has an alsa driver that works, but isn't in upstream shape (and isn't ASoC).
  • [16:14:52] <dcordes> ok
  • [16:16:02] <dcordes> jkridner: does your OE build now?
  • [16:16:36] <jkridner> I was adding all the things I wanted in my new demo image based on what I have installed on my current running image.
  • [16:18:12] <dcordes> useful
  • [16:18:22] <dcordes> make sure you add task-native-sdk
  • [16:18:29] <jkridner> got it. :)
  • [16:18:43] <jkridner> also, jamvm (despite the boos and hisses regarding Java)
  • [16:19:01] <jkridner> it is a lot of fun to host beagleboard.org on a Beagle. :)
  • [16:19:16] <chakie> is that so?
  • [16:19:36] <jkridner> Helma runs rather nicely with 'jamvm'.
  • [16:19:58] <dcordes> jkridner: beagleboard.org runs on the beagle with demo image??
  • [16:20:18] <jkridner> not demo image... need to do 'opkg install jamvm'.
  • [16:21:00] <jkridner> then you can follow the README to install the beagleboard.org sources (except that I haven't gotten git to compile yet, so I had to manually copy).
  • [16:21:26] <jkridner> since I have root access, I can rsync.
  • [16:22:17] <dcordes> which beagleboard.org sources?
  • [16:22:36] * Olipro__ is now known as Olipro
  • [16:22:51] <jkridner> http://www.beagleboard.org/gitweb/?p=beagleboard.org.git;a=summary
  • [16:24:52] <dcordes> oh
  • [16:30:17] <jkridner> it is pretty neat to browse on my home network the same server app running on my beagle board.
  • [17:14:46] <Crofton> What's up with C70?
  • [17:15:21] <mru> it's up to no good, apparently
  • [17:15:30] * mru wonders why they added it in the first place
  • [17:15:44] <mru> and why it doesn't affect my board
  • [17:23:35] <dirk2> sakoman mentioned that overo and evm have the same capacitor and it doesn't affect these boards, too
  • [17:24:33] <dirk2> Crofton: If you like to remove it: http://www.flickr.com/photos/25691331@N04/2766671437/in/pool-beagleboard
  • [17:24:48] <mru> I think I'll leave mine
  • [17:27:03] <dirk2> I'm fine with MPU timer for the time being
  • [17:28:40] <dirk2> jkridner: What do you think about moving Beagle FAQ to wiki, e.g. http://elinux.org/BeagleBoardFAQ ? It would be easier to edit and for people to contribute.
  • [17:30:09] <mru> I'm using gptimer 12 or whatever that hack that was posted did
  • [17:39:31] * cian (n=cianh@cian.ws) has joined #beagle
  • [17:44:14] * cian_ (n=cian@cian.ws) has joined #beagle
  • [17:45:26] * cian_ (n=cian@cian.ws) Quit (Client Quit)
  • [17:48:13] * cian_ (n=cian@cian.ws) has joined #beagle
  • [17:58:12] * cian_ (n=cian@cian.ws) Quit ()
  • [18:11:56] * cian (n=cianh@cian.ws) Quit ()
  • [18:13:18] * t_s_o (n=tso@173.84-49-129.nextgentel.com) has joined #beagle
  • [18:18:14] <uberfry> hi all
  • [18:18:20] <uberfry> anyone know how to access the framebuffer?
  • [18:20:09] * JoeBorn (n=jborn@adsl-75-2-247-191.dsl.chcgil.sbcglobal.net) has joined #beagle
  • [18:25:08] <mru> uberfry: /dev/fb0
  • [18:25:13] <mru> what do you want to do with it?
  • [18:25:43] <uberfry> well I meant more like physical memory...
  • [18:25:58] <uberfry> I'd like to avoid linux for my application
  • [18:26:07] <zalassi> oO
  • [18:26:25] <uberfry> I know...you're thinking - why reinvent the wheel?
  • [18:26:34] <uberfry> linux doesn't really suit this application
  • [18:26:38] <zalassi> no, I'm thinking, you don't mean framebuffer
  • [18:26:45] <uberfry> I don't?
  • [18:26:46] <banderson> LOL
  • [18:26:50] <uberfry> the videoram
  • [18:26:57] <uberfry> on 360 I access the framebuffer
  • [18:27:02] <uberfry> to display stuff
  • [18:27:16] <zalassi> you mean... like how to access it directly using assembly, etc?
  • [18:27:20] <uberfry> yeah
  • [18:27:23] <zalassi> ah
  • [18:27:29] <mru> the video memory is normal system ram on the omap
  • [18:27:43] <uberfry> mru: yeah but which address?
  • [18:27:53] <mru> whichever you tell the display controller
  • [18:27:54] <uberfry> what's the resolution?
  • [18:27:59] <mru> whatever you configure
  • [18:28:00] <uberfry> how do you setup the resolution?
  • [18:28:11] <mru> write the right values to the right registers
  • [18:28:19] <mru> linux does all that for you ;-)
  • [18:28:31] <mru> you also need to configure some clocks
  • [18:28:33] <uberfry> do you by any chance have a memory map? :)
  • [18:28:45] <mru> everything you need is in the TRM
  • [18:28:50] <uberfry> TRM?
  • [18:28:57] <mru> technical reference manual
  • [18:29:00] <uberfry> ah ic
  • [18:29:02] <uberfry> sec
  • [18:29:09] <zalassi> there's also the kernel driver sources
  • [18:29:14] <mru> look for a document called spruf98a.pdf
  • [18:29:25] <mru> it's on the TI website
  • [18:29:32] <uberfry> thanks mru and zalassi
  • [18:29:43] <uberfry> but zalassi, I don't like reading linux kernel parts...
  • [18:29:45] <uberfry> for one reason
  • [18:29:51] <uberfry> they make it conform to something
  • [18:30:00] <mru> I'd still recommend using linux, at least initially
  • [18:30:03] <uberfry> stuff that can be done in 5 lines is extended into 25
  • [18:30:19] <uberfry> mru: sure, but later on I might want to market it
  • [18:30:24] <mru> what are you doing that makes linux unsuitable?
  • [18:30:34] <mru> everybody uses linux these days
  • [18:30:38] <uberfry> something not open source :/
  • [18:30:49] <mru> you can still use the linux kernel
  • [18:30:50] <zalassi> does linux have rules about binary blobs or something?
  • [18:30:54] <uberfry> I don't mean to be a dick, but I have tuition fees
  • [18:31:09] <uberfry> they don't pay themselves :(
  • [18:31:12] <mru> uberfry: that's ok, I write close source for a living
  • [18:31:13] <uberfry> so I need to dev something
  • [18:31:17] <uberfry> ah ok
  • [18:31:24] <uberfry> good then, you understand me :)
  • [18:31:49] <mru> it's worse than that, I work for a company that makes pay-tv systems
  • [18:31:53] <uberfry> wow, 37 megs?
  • [18:31:59] <uberfry> oh hehe
  • [18:32:03] <uberfry> which, if I may ask?
  • [18:32:07] <mru> sure, nds
  • [18:32:13] <uberfry> cool
  • [18:32:16] <uberfry> which country?
  • [18:32:20] <mru> uk
  • [18:32:27] <mru> you've heard of them?
  • [18:32:27] <uberfry> icic
  • [18:32:31] <mru> that's unusual
  • [18:32:42] <uberfry> nds yes, but never really looked into it
  • [18:32:51] <mru> nobody cares to read the fine print on the back of the smartcard ;-)
  • [18:32:57] <uberfry> lol
  • [18:33:06] <uberfry> anyway
  • [18:33:11] <uberfry> I studied the x360 system
  • [18:33:22] <mru> x360?
  • [18:33:24] <mru> xbox?
  • [18:33:26] <uberfry> so if someone tries to hack my device to run on other systems...tough luck...
  • [18:33:27] <uberfry> yes
  • [18:33:42] <mru> anything can be hacked
  • [18:33:50] <mru> that's the first thing they teach us at nds
  • [18:33:50] <uberfry> well yeah, but not easily
  • [18:33:58] <uberfry> you go ahead and try hacking the 360
  • [18:34:01] <uberfry> without an exploit
  • [18:34:05] <mru> already been done
  • [18:34:16] <uberfry> you have?
  • [18:34:19] <mru> not me
  • [18:34:30] <uberfry> I know of no way to decrypt the memory without the per-session key
  • [18:34:43] <mru> well, then you obtain the key
  • [18:34:48] <mru> it's in there somewhere
  • [18:34:55] <uberfry> it's generated at each bootup
  • [18:35:00] <uberfry> it's based on the rtc
  • [18:35:19] <mru> so set the rtc to a known value
  • [18:35:27] <uberfry> you can't
  • [18:35:35] <zalassi> any system that gives the user the key and the ciphertext is theoretically hackable, isn't it?
  • [18:35:37] <uberfry> or do you know how to access the die?
  • [18:35:39] <mru> sniff the key then
  • [18:35:46] <uberfry> no way...all in die
  • [18:35:53] <mru> yes, any self-contained system is hackable
  • [18:36:08] <uberfry> mru: you have the right tools to decap and bond
  • [18:36:09] <uberfry> ?
  • [18:36:20] <mru> did you read about the philips/nxp mifare chip?
  • [18:36:28] <uberfry> of course
  • [18:36:34] <uberfry> but still requires bruteforce
  • [18:36:41] <uberfry> but look...this is a per-session 128 bit key
  • [18:36:54] <mru> yes, but the key is in there
  • [18:37:08] <Crofton|work> brute force gets easier every day :)
  • [18:37:16] <uberfry> yes but does everyone want to do decapsulation?
  • [18:37:26] <mru> not everybody has to
  • [18:37:59] <uberfry> but anyway, I know how MS packs their system
  • [18:38:00] <zalassi> worst case you get some unscrupulous company in China/Russia to open up the chip and put it under a microscope <_<;
  • [18:38:08] <uberfry> lol
  • [18:38:14] <uberfry> yeah but I doubt that'll work for 360 ;)
  • [18:38:24] <zalassi> does the chip catch fire when opened or something?
  • [18:39:01] <banderson> This chip will self destruct in 10...9...8
  • [18:39:07] <uberfry> well, let's say there's stuff that, when you destroy it, the system won't power up again
  • [18:39:17] <zalassi> doesn't matter
  • [18:39:17] <uberfry> you need to get beyond 2 layers to get to the per-box key
  • [18:39:19] <uberfry> called fuses
  • [18:39:30] <zalassi> all they need to do is figure out the chip schematic
  • [18:39:31] <uberfry> you destroy the upper layers, you get to the fuses
  • [18:39:37] <uberfry> but you don't power up again
  • [18:39:38] <zalassi> doesn't need to keep that particular one still working
  • [18:40:25] <uberfry> anyway, I know how x360 handles security and I know their flaws
  • [18:40:33] <uberfry> which helps me build a stronger system
  • [18:40:40] <mru> sony made a smart move with the ps3
  • [18:40:54] <uberfry> you think?
  • [18:40:58] <uberfry> we have the bootrom
  • [18:40:59] <mru> open up enough of it to remove most of the incentive to hack the rest
  • [18:41:12] <banderson> Am I missing something ...hasn't the 360 been hacked and running Linux already?
  • [18:41:21] <uberfry> banderson: with an exploit, yes
  • [18:41:59] <mru> so it only counts for you if you hack it without hacking it?
  • [18:42:05] <banderson> LOL
  • [18:42:21] <uberfry> mru: someone got their hands on a very early box
  • [18:42:24] <uberfry> very early
  • [18:42:33] <uberfry> dated may 2004 or so
  • [18:42:45] <uberfry> fortunately, it was unencrypted
  • [18:43:02] <uberfry> and so they got the algo to unpack the bootloaders and kernel
  • [18:43:10] <Crofton|work> anyone want to make me a 28 pin ribbon cable connector for the beagle expansion connector?
  • [18:43:40] <uberfry> Crofton: what for?
  • [18:43:52] <uberfry> btw, who's the person that makes the pcb design?
  • [18:44:40] <Crofton|work> note sure if he is here
  • [18:45:01] <uberfry> it would be nice to have a chat with him/her
  • [18:45:02] <Crofton|work> I'm having a rough time finding the needed pieces in the Digikey catalog
  • [18:46:26] <banderson> Crofton|work: You putting same connector on both ends?
  • [18:46:49] <Crofton|work> I'd like to
  • [18:47:07] <Crofton|work> I want to run some ribbon cable to a FPGA board at first
  • [18:47:14] <banderson> Crofton|work: I will do a quick look to see what I have here...then I should be able to get you part number.
  • [18:47:54] <Crofton|work> good luck
  • [18:48:02] <Crofton|work> looks like a 28 pin connector
  • [18:48:08] <Crofton|work> what appears uncommon?
  • [18:48:20] <Crofton|work> maybe I need a second set of eyes
  • [18:49:15] <uberfry> Crofton: I had an idea about a year ago
  • [18:49:24] <uberfry> take IDE wire
  • [18:49:31] <uberfry> and you know those lcd connectors?
  • [18:49:39] <uberfry> the ones that are usually on fpga boards
  • [18:49:59] <Crofton|work> I am getting a proto board to do the FPGA termination
  • [18:50:11] <Crofton|work> we'll have to dedictate one bank to 1.9 v io
  • [18:50:21] <uberfry> 1.9V?
  • [18:50:28] <uberfry> since when does that exist?
  • [18:50:31] <uberfry> you mean 1.8V?
  • [18:51:08] * cian (n=cian@cian.ws) has joined #beagle
  • [18:53:27] <Crofton|work> right
  • [18:54:13] <uberfry> WOW
  • [18:54:21] <uberfry> omap 3 can have 1920x1080?!
  • [18:54:28] <mru> should be possible
  • [18:54:33] <mru> I have yet to try it
  • [18:54:39] <uberfry> quite impressive
  • [18:54:46] <uberfry> eventho my current lcd doesnt display it
  • [18:55:30] * dirk2 (n=dirk@F32d5.f.strato-dslnet.de) has left #beagle
  • [18:56:55] <banderson> Crofton|work: Ya it looks like they only make 26 or 30 ...don't see any 28 yet
  • [18:57:08] <Crofton|work> solved the cable problem, Floppy is more than 30 pins :)
  • [18:57:16] <banderson> nice
  • [18:57:16] <Crofton|work> I think a 30 will fot
  • [18:57:42] <Crofton|work> I can make the cable from a floppy cable
  • [18:57:44] <banderson> Crofton: will have to cut out the "fliped" part but should work...
  • [18:57:48] <Crofton|work> progress
  • [18:58:09] <Crofton|work> I guess this connector should be examined :)
  • [18:59:16] <Crofton|work> I'll order 30 pin connector
  • [18:59:31] <Crofton|work> and cut out the wire from the floppy cab;e
  • [18:59:51] * cian (n=cian@cian.ws) Quit ()
  • [19:01:58] * Olipro_ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [19:02:06] * Olipro (n=Olipro@unaffiliated/olipro) Quit (Nick collision from services.)
  • [19:02:08] * Olipro_ is now known as Olipro
  • [19:02:32] <banderson> Crofton|work: you could to two 14's
  • [19:02:42] * charliex (n=chalrie@mail.finitemonkeys.com) has joined #beagle
  • [19:03:18] <Crofton|work> mechanically, I'm not sure that will work
  • [19:03:30] * ulso1 (n=ulso@1-1-2-31a.rny.sth.bostream.se) has left #beagle
  • [19:03:35] <Crofton|work> I'd have to find exactly the right part
  • [19:03:39] <banderson> Crofton|work: doh! ya probably not
  • [19:04:12] <charliex> hi, i just found the beagleboard, looks very interesting, but digikey are out of stock, anyone know the lead time ?
  • [19:08:07] <dcordes> charliex: abuse digikey live support
  • [19:08:55] * cian (n=cianh@cian.ws) has joined #beagle
  • [19:10:11] <charliex> ok dokey
  • [19:12:11] <charliex> blergh 7-9 weeks
  • [19:12:45] <dcordes> I should have ordere 2 or 3 :(
  • [19:13:00] <charliex> wonder if a new rev will be available before then ?
  • [19:13:14] <Crofton|work> charliex, ask here tomorrow
  • [19:13:23] <Crofton|work> I don't think it will be that long
  • [19:14:23] <charliex> will do. shame i just found this board, i've been struggling with a PC 104 arm 200Mhz with FPGA video and i gave up yesterday and just stick in a mini pc instead. all i need to do is add CAN bus to it and some ADC's and it'll be almost perfect..
  • [19:14:56] <dcordes> Crofton|work: how many units were sold with the first batch?
  • [19:15:17] <Crofton|work> I'm not sure
  • [19:17:01] * b00berfry (i=nooberfr@vodsl-8706.vo.lu) has joined #beagle
  • [19:17:30] * uberfry (i=nooberfr@vodsl-10903.vo.lu) Quit (Read error: 60 (Operation timed out))
  • [19:26:58] * ulso (n=ulso@1-1-2-31a.rny.sth.bostream.se) has joined #beagle
  • [19:30:14] * lgentili (i=c82be332@gateway/web/ajax/mibbit.com/x-80fc24d1096878c9) has joined #beagle
  • [19:30:57] <dcordes> does somebody have experience with debian arm on the beagle?
  • [19:31:53] <mru> some say debian isn't very well optimised on arm
  • [19:31:57] <mru> to put it mildly
  • [19:32:18] <banderson> dcordes: There was a guy on beagle list that was doing some thing along those lines...
  • [19:32:44] <mru> I now for a fact that most debian packages are built for armv4
  • [19:32:49] * _Ade_ (n=Ade@78-86-195-216.zone2.bethere.co.uk) has joined #beagle
  • [19:32:54] <mru> with softfloat
  • [19:33:17] <mru> imho you're better off with openembedded or gentoo
  • [19:33:24] <dcordes> hm. I'm not sure about ubuntu on it
  • [19:33:30] <suihkulokki> -funroll-loops !
  • [19:33:34] <dcordes> maybe gentoo is the choice
  • [19:33:58] <mru> suihkulokki: too muck loop unrolling can be bad
  • [19:34:00] <banderson> http://elinux.org/BeagleBoardDebian
  • [19:34:10] <mru> the cortex-a8 branch predictor is pretty good
  • [19:38:42] <charliex> i use debian arm on a 200mhz ep9302 its ok, gcc used to be terrible for arm, but arm hired someone to tidy it up so its a lot better
  • [19:39:18] <suihkulokki> uh, that doc is bad, people should guide lenny/armel instead of etch/arm :(
  • [19:39:48] <mru> charliex: on a low-end arm debian is probably ok, since that's what it's built for
  • [19:40:00] <vlad_> ugh, I didn't even notice that it's not armel
  • [19:40:11] <vlad_> well crap
  • [19:40:15] * vlad_ starts over
  • [19:41:43] <dcordes> suihkulokki: that's what I thought
  • [19:41:55] <dcordes> but it's a nice explanation
  • [19:41:58] <dcordes> banderson: thanks
  • [19:42:11] <banderson> np
  • [19:42:41] <calculus> Crofton|work: what do you use to program the FPGA?
  • [19:42:52] <vlad_> hrm, so if debian's optimized for armv4, is there another good userspace with lots of binary packages available? OE kinda threw me off (via angstrom); opkg was crashing all over the place
  • [19:42:54] <calculus> Xilinx ISE?
  • [19:43:07] <Crofton|work> calculus, yes
  • [19:43:51] <calculus> ok, then it is the same that I have used in the past
  • [19:43:53] * ulso (n=ulso@1-1-2-31a.rny.sth.bostream.se) Quit ("Leaving.")
  • [19:43:55] <calculus> thanks
  • [19:44:20] <mru> banderson: the rtc on the beagle works fine, but obviously resets if you remove power
  • [19:44:35] <dcordes> vlad_: opkg is not so mature it seems. I wonder why they switched from ipkg. anyway, it always works best to include the packages needed on image build time
  • [19:44:51] <dcordes> always depends on how well the machine is maintained
  • [19:44:55] <vlad_> dcordes: yeah, though in this case i'm not building an image, I'm doing development
  • [19:45:09] <Crofton> opkg is basically a big fixed ipkg
  • [19:45:23] <vlad_> and would like to treat the device as a normal box, where I can pull in packages as needed :)
  • [19:45:24] <dcordes> instead of package not found, it gives me a segfault
  • [19:45:35] <suihkulokki> vlad_: you could try armel with those instructions just replacing the versatile kernel and initrd from armel/lenny
  • [19:45:57] <vlad_> suihkulokki: nod; just found an email from jkridner pointing at http://mojo.handhelds.org/hasty-armv5el
  • [19:46:04] <vlad_> which seems like a better way to go
  • [19:46:58] <dcordes> why?
  • [19:47:35] <vlad_> or the armv6 variants, though they're based on ubuntu 7.10 and not 8.04
  • [19:47:48] <vlad_> dcordes: because of what mru as saying earlier, about debian being optimized for armv4
  • [19:47:58] <dcordes> ok
  • [19:48:13] <mru> omap3530 is armv7
  • [19:48:46] <vlad_> yeah, but I don't really want to rebuild the entire userspace.. so v6 is closer :)
  • [19:49:05] <mru> v6 is certainly closer
  • [19:50:48] <mru> but there are significant differences
  • [19:51:36] <mru> most notably, of course, a simd unit
  • [19:52:05] <suihkulokki> vlad_: despite of what mru will claim, for most use cases you'll only need to recompile few selected apps and libraries to get enough performance
  • [19:52:07] * uberfry (i=nooberfr@vodsl-11238.vo.lu) has joined #beagle
  • [19:52:54] <mru> armv7 has no new arm instructions of significance to normal apps
  • [19:53:00] <vlad_> suihkulokki: yeah, that's true
  • [19:54:11] <vlad_> mru: gcc won't use simd by default though, even when optimizing for v7, will it? (will it even with -ftree-vectorize?)
  • [19:54:12] <dcordes> mru: so in most cases armv7 will not provide much extra performance as opposed to armv6?
  • [19:54:13] <mru> what's more important is much better support unaligned memory access
  • [19:54:42] <mru> if compiling for armv6 or earlier, it's generally better to mask and shift
  • [19:54:53] <mru> on armv7 doing unaligned accesses is faster
  • [19:55:00] <mru> only one cycle penalty
  • [19:55:18] <mru> some instructions also have different cycle timing compared to armv6
  • [19:55:35] <mru> so to get optimal performance, instructions need to be reordered
  • [19:56:00] <mru> now I don't know how much of this gcc takes into account in reality
  • [19:57:19] * Navi (n=Navi@unaffiliated/navi) Quit (Read error: 104 (Connection reset by peer))
  • [19:59:19] * felipec (n=felipec@a88-113-6-167.elisa-laajakaista.fi) has joined #beagle
  • [20:06:14] * b00berfry (i=nooberfr@vodsl-8706.vo.lu) Quit (Read error: 110 (Connection timed out))
  • [20:08:28] * t_s_o (n=tso@173.84-49-129.nextgentel.com) Quit ("Konversation terminated!")
  • [20:18:49] * cian (n=cianh@cian.ws) Quit ()
  • [20:19:40] * cian (n=cianh@cian.ws) has joined #beagle
  • [20:24:05] <vlad_> well, gonna try doing a lenny armel base install and then pulling in the handhelds mojo packages
  • [20:26:42] <vlad_> ah, I bet my problem yesterday using the anstrom kernel was that the kernel didn't have support for old-ABI binaries
  • [20:31:03] * cian_ (n=cian@cian.ws) has joined #beagle
  • [20:31:28] <dcordes> # CONFIG_OABI_COMPAT is not set
  • [20:31:38] <dcordes> vlad_: that is in the demo kernel from OE
  • [20:31:39] <mru> see what I mean about debian being a bit old?
  • [20:31:52] <vlad_> dcordes: yep, that's it then
  • [20:31:57] <vlad_> boy, that page needs to be revamped :)
  • [20:32:26] <vlad_> maybe I'll do that when I get back from the airport
  • [20:32:45] <vlad_> so much work just so that I can get to profiling x/cairo/firefox =/
  • [20:34:19] <mru> profiling firefox?
  • [20:34:29] <mru> I think you'll find it's spending most of its time wasting time
  • [20:35:16] <dcordes> webkit ftw
  • [20:36:48] * cian_ (n=cian@cian.ws) Quit ()
  • [20:37:09] * vlad_ rolls his eyes
  • [20:39:39] * cian_ (n=cian@cian.ws) has joined #beagle
  • [20:48:32] * cian_ (n=cian@cian.ws) Quit ()
  • [20:49:54] * cian_ (n=cian@cian.ws) has joined #beagle
  • [20:55:59] * felipec (n=felipec@a88-113-6-167.elisa-laajakaista.fi) Quit ("Leaving")
  • [21:17:57] * cian_ (n=cian@cian.ws) Quit ()
  • [21:49:16] * b00berfry (n=fubar@vodsl-9991.vo.lu) has joined #beagle
  • [22:05:36] * uberfry (i=nooberfr@vodsl-11238.vo.lu) Quit (Read error: 113 (No route to host))
  • [22:14:47] * valhalla (n=valhalla@81-174-21-16.dynamic.ngi.it) Quit ("Leaving")
  • [22:29:40] * Olipro_ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [22:29:52] * Olipro (n=Olipro@unaffiliated/olipro) Quit (Nick collision from services.)
  • [22:29:54] * Olipro_ is now known as Olipro
  • [22:38:05] * ap02 (n=abhijit@c-24-6-175-156.hsd1.ca.comcast.net) has joined #beagle
  • [22:43:15] * ulso (n=ulso@1-1-2-31a.rny.sth.bostream.se) has joined #beagle
  • [22:45:28] * ulso (n=ulso@1-1-2-31a.rny.sth.bostream.se) has left #beagle
  • [22:52:08] * haggisX (n=haggis@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) has joined #beagle
  • [22:55:37] * ap02 (n=abhijit@c-24-6-175-156.hsd1.ca.comcast.net) Quit ("Leaving")
  • [22:56:16] * JoeBorn (n=jborn@adsl-75-2-247-191.dsl.chcgil.sbcglobal.net) Quit (Read error: 104 (Connection reset by peer))
  • [22:59:23] * haggisX (n=haggis@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) Quit ("Leaving")
  • [23:17:48] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) Quit ()
  • [23:20:19] * bobbarker (n=bob@adsl-69-109-116-44.dsl.pltn13.pacbell.net) has joined #beagle
  • [23:20:23] * lgentili (i=c82be332@gateway/web/ajax/mibbit.com/x-80fc24d1096878c9) Quit ("http://www.mibbit.com ajax IRC Client")
  • [23:21:46] <bobbarker> Can I bend the 10th pin on RS232 to fit an "idiot proof" cable?
  • [23:22:33] <vlad_> I'd snip it off instead of bending, but I'd also just get the right cable :)
  • [23:22:45] <mru> I de-proofed my cable
  • [23:22:56] <bobbarker> mru, how?
  • [23:23:04] <mru> with a small knife
  • [23:23:12] <mru> it's just a thin layer of plastic
  • [23:23:14] <mru> usually
  • [23:23:33] <bobbarker> So it still doesn't connect to anything?
  • [23:23:56] <mru> the pin isn't connected on the board
  • [23:24:08] <mru> doesn't matter what happens to it
  • [23:24:25] <mru> it won't be connected anywhere in the cable either
  • [23:25:51] <bobbarker> I'm doing "My Beagle Board Out of the box experience
  • [23:26:41] <bobbarker> " and the hyperterminal thing isn't working, how do I troubleshoot?
  • [23:26:58] <mru> get a different terminal program
  • [23:27:06] <vlad_> (ckermit's great)
  • [23:27:48] <vlad_> are you sure you have the right kind of cable, though? I had the wrong kind of pinout (the "other" one) for idc10 serial originally
  • [23:27:50] <bobbarker> Anybody else have windows?
  • [23:28:34] <bobbarker> vlad, what'
  • [23:28:42] <bobbarker> s the "other" one
  • [23:28:44] <bobbarker> ?
  • [23:28:45] <mru> s/have/hate/
  • [23:29:01] <dcordes> bobbarker: you should open your d-sub connector and look whether or not it has the appropriate wiring if possible
  • [23:29:18] <vlad_> bobbarker: one way to tell is to remove your null modem adapter
  • [23:29:26] <vlad_> if you end up getting output, but can't type, then you have the wrong kind
  • [23:30:00] <dcordes> you get output when you remove the cable??
  • [23:30:01] <vlad_> (you need AT/Everex style, straight through pin assignments.. the wrong one does something funky with the pins, I forget what it's called)
  • [23:30:22] <bobbarker> vlad, thanks.
  • [23:30:27] <mru> the everex goes sideways, the "other one" goes up and down
  • [23:30:32] <bobbarker> dcordes I don't have any output.
  • [23:30:40] <bobbarker> Input or anthing...
  • [23:30:40] <dcordes> vlad_: the idc10->d-sub I had, had one wire upper row, one wire lower, one upper, etc..
  • [23:30:47] <vlad_> ahh
  • [23:31:00] <mru> dcordes: that's useful with ribbon cables
  • [23:31:16] <dcordes> and not 12345 top row, 6789 bottom row as it is supposed to be
  • [23:31:35] <bobbarker> I've got one that looks like this http://www.pccables.com/07120.htm
  • [23:31:40] <vlad_> grumble, debian armel install was almost done and then I had to go and do something stupid and crash my laptop.. which crashed vmware which crashed qemu. fail.
  • [23:31:50] <dcordes> bobbarker: also from the inside?
  • [23:32:01] <mru> debian seems such a pain
  • [23:32:04] <dcordes> bobbarker: mine looked 1:1 like that one, but d-sub was wired wrong.
  • [23:32:12] <vlad_> mru: why?
  • [23:32:21] <vlad_> mru: i'm not a huge fan, fwiw, but I don't see many better options
  • [23:32:26] <mru> mucking about with qemu...
  • [23:32:29] <vlad_> well, ubuntu I guess, though they ahve their own problems
  • [23:32:30] <mru> no eabi
  • [23:32:34] <vlad_> no, this is eabi
  • [23:32:40] <mru> there's gentoo
  • [23:32:47] <mru> and openembedded/angstrom
  • [23:32:54] <vlad_> heh.. see, gentoo looks like such a pain to me :)
  • [23:33:07] <Pavlov> but gentoo is faster
  • [23:33:13] <vlad_> oe/angstrom would maybe work, but crashy package tools threw me off
  • [23:33:38] <mru> cross-compiling gentoo is easy on a gentoo machine
  • [23:33:46] <vlad_> yes, but it involves compiling
  • [23:33:54] <vlad_> I don't like to compile the world
  • [23:34:02] <mru> you should
  • [23:34:07] <bobbarker> I guess my best bet would be to try the null modem cable with another computer?
  • [23:34:07] <vlad_> (granted, I will probably end up recompiling a pile of stuff anyway, but still)
  • [23:34:37] <vlad_> bobbarker: you have a null modem cable and not a null modem adapter?
  • [23:34:51] <bobbarker> yes
  • [23:35:01] <vlad_> if you have a non-null-modem cable you can try that and see if you have the wrong idc10-db9 cable
  • [23:35:54] <bobbarker> k thanks