• [00:00:47] <raster> mru: thats a classic mistake i've been careful to avoid
  • [00:00:59] <raster> as the config doesnt determine location of button... it doesnt matter
  • [00:01:00] <raster> wel it does
  • [00:01:05] <raster> the theme is a binary blob too
  • [00:01:09] <mru> that was just an example
  • [00:01:11] <raster> its the same lib and same decode path
  • [00:01:26] <raster> iun the end though a "corrupt file" means a corrupt filesystem
  • [00:01:32] <mru> not always
  • [00:01:38] <raster> so how will you fix a binary thats corrupt
  • [00:01:38] <mru> it might mean "buggy app wrote it"
  • [00:01:40] <raster> or a lib?
  • [00:01:42] <raster> its the same thing
  • [00:01:45] <recalcati> raster: what do you mean lock a window down ?
  • [00:01:53] <raster> recalcati: you acked to do it
  • [00:01:56] <raster> its all there
  • [00:02:11] * mrgarlic (n=mrgarlic@pool-70-19-69-81.ny325.east.verizon.net) has joined #beagle
  • [00:02:16] <raster> window icon -> more -> window locks
  • [00:02:21] <raster> or right click on title
  • [00:02:25] <raster> or alt+right click anywhere
  • [00:03:49] <raster> mru: there is a perfectly good cmd-line to tell e to swistch to another5 known good profile if all shit hits the fan
  • [00:04:01] <raster> enlightenment_remote
  • [00:04:20] <mru> have you ever had a windows machine with a messed up registry?
  • [00:04:23] <recalcati> moment, the board is starting
  • [00:04:29] <raster> i dont use windows
  • [00:04:33] <raster> but i do use e
  • [00:04:37] <raster> year in and out
  • [00:04:40] <mru> neither do I
  • [00:04:42] <raster> and my configs dont get messed up
  • [00:04:46] <mru> but I've still suffered from it
  • [00:05:01] <raster> just because they are binary does not mean its designed to be fragile
  • [00:05:02] <mru> I think you're a little over-confident
  • [00:05:16] <raster> well - i've been using this for about 4+ years now
  • [00:05:29] <raster> with no unexpected failures int he config system
  • [00:05:41] <raster> and i am confident
  • [00:06:07] <raster> i was pretty careful in writing it
  • [00:06:15] <mru> I'm sure you were
  • [00:06:24] <mru> but I'm also sure you made a mistake somewhere
  • [00:06:31] <mru> everybody makes mistakes
  • [00:06:48] <recalcati> raster: midori is up
  • [00:07:09] <raster> mru: there's other mechanisms to recocer
  • [00:07:11] <Downix> wait, there are windows machines w/o messed up registries?
  • [00:07:12] <raster> recover
  • [00:07:13] <recalcati> ok more -- windo lock
  • [00:07:23] <raster> rm -rf ~/.e/e/config
  • [00:07:30] <raster> switch to another profile
  • [00:07:57] <mru> but then you lose all the good settings from the old profile
  • [00:08:05] <raster> eet -d ... (to decode and extract to text edit and) eet -e
  • [00:08:16] <raster> thats one thing on my todo
  • [00:08:18] <raster> backups
  • [00:08:26] <raster> keep the last N configs
  • [00:08:31] <recalcati> Remember there Lockd for this windos next time it appears
  • [00:08:38] <raster> (or the last N hours or days)
  • [00:08:54] <raster> and allow you to roll back in time to previous versions... like an scm (simplified)
  • [00:09:11] <mru> for a high-level app it might not matter so much
  • [00:09:26] <raster> recalcati: all you need is there in window locks.
  • [00:09:27] <mru> but please don't tell me to delete /etc and start over if one line in fstab is bad
  • [00:10:07] <mru> I once had to recover a system using only sed, grep, and bash
  • [00:10:17] <raster> mru: as i said - on my todo was keeping N revisions
  • [00:10:20] <mru> I don't remember what was wrong
  • [00:10:29] <mru> but *no* fancy tool would have been able to run
  • [00:10:36] <raster> i'vee never had to do it
  • [00:10:43] <mru> good for you
  • [00:10:50] <raster> so its low on my todo as its not urgent
  • [00:11:06] <raster> my aim is so u DONT have to drop to a terminal and "vi"
  • [00:11:21] <raster> my experience is that there are so many users who simply dont even know what a concole is
  • [00:11:23] <raster> or vi
  • [00:11:27] <raster> so for them the system is broken
  • [00:11:31] <raster> text or binary - they dont care
  • [00:11:34] <raster> they are unable to fix it
  • [00:11:35] <recalcati> checked all locks, good
  • [00:11:57] <recalcati> now I restart and I see if the system remember these settings
  • [00:12:55] <raster> recalcati: e writes them and saves them on change or exit
  • [00:13:05] <raster> tho on change it will queue and efer writes to minimise IO
  • [00:15:15] <recalcati> by now is ok. but I'll have to disable right click for user
  • [00:15:28] <recalcati> so he can't disable locks
  • [00:15:52] <raster> (ie it will defer a write for 5 seconds - if another config change happens int he meantime it defers another 5 secs - until it goes idle enough to get a window of 5 seconds to flush - flushes are forced on exit/restart)
  • [00:16:44] <raster> recalcati: you'll want to delete all key and mouse bindings
  • [00:16:46] <raster> but...
  • [00:16:53] <recalcati> now I'm root on the system, but I'm not root, I could change ~/.e/e/config permission so that only root can change them
  • [00:16:53] <raster> there is no gui for signal bindingsa
  • [00:16:57] <raster> signal bindings
  • [00:17:05] <raster> you will need to change them too
  • [00:17:14] <raster> enlightenment_remote has stuff for that
  • [00:17:40] * mrgarlic__ (n=mrgarlic@pool-70-19-69-81.ny325.east.verizon.net) Quit (Success)
  • [00:17:42] <recalcati> the user try to change ~/.e/e/config but he can't
  • [00:17:57] <raster> e will complain loudly about not being able to save config
  • [00:17:58] <raster> :)
  • [00:18:14] <recalcati> right
  • [00:18:24] <raster> does midori have printing enabled?
  • [00:18:40] <recalcati> no, I'm trying to enable them,
  • [00:18:51] <raster> what will u do about "print command" ?
  • [00:18:52] * mrgarlic_ (n=mrgarlic@pool-70-19-69-81.ny325.east.verizon.net) has joined #beagle
  • [00:18:52] <raster> :)
  • [00:19:11] <recalcati> by now I'm able to use gdb and I have through away some menus
  • [00:19:12] * Openfree` (n=Openfree@222.65.175.96) has joined #beagle
  • [00:19:41] <recalcati> irs easy to hack
  • [00:19:46] <recalcati> its easy to hack
  • [00:20:02] <recalcati> raster: print command ? you do you mean ?
  • [00:20:23] <mru> recalcati: let me tell you how I once broke into a "locked down" windows machine
  • [00:20:34] <mru> you were only supposed to run a single app
  • [00:20:42] <mru> the start menu was disabled
  • [00:21:00] <mru> the app had some online help
  • [00:21:07] <mru> using windows help files
  • [00:21:18] <mru> the windows help viewer has a file->open menu
  • [00:21:37] <mru> from there I browsed to regedit.exe, right-clicked, and selected "run"
  • [00:21:39] <mru> job done
  • [00:22:12] * Thang (i=0cce5fb8@gateway/web/ajax/mibbit.com/x-bb2f618f663a37ba) Quit ("http://www.mibbit.com ajax IRC Client")
  • [00:22:43] <raster> recalcati: 10+ years ago someoen used mwm+netscape on xterminals to provide a kisok browser
  • [00:23:01] <raster> but u still could print.. and the print config allowed you to specify the print command (instead of lpr).
  • [00:23:05] <raster> just specify xterm and bingo
  • [00:23:09] <raster> unlocked.
  • [00:23:14] <mru> by this example I mean to illustrate that there are often paths into the system you might not immediately think of
  • [00:23:16] <raster> (well and print)
  • [00:23:38] <raster> yup
  • [00:23:54] <raster> e itself can be configured in a way to make it useless as a vector
  • [00:23:57] <mru> raster: that attack is similar to the one I described
  • [00:24:02] <raster> it has a gazillion litttle config knobs
  • [00:24:07] <raster> but you have other thngis to worry about too
  • [00:24:19] <raster> mru: yup. this one was just a unix version of it
  • [00:24:20] <raster> :)
  • [00:25:09] <recalcati> tar: dev/.udev/names/tty10: time stamp 2000-01-01 00:00:11 is 0.162048394 s in the future
  • [00:25:21] <recalcati> do you have these at boot ?
  • [00:27:14] <raster> no
  • [00:27:52] <recalcati> I have with 2.6.28 and also 2.6.29
  • [00:29:32] <recalcati> another stange thing. u-boot handles 720p correctly, instead in the kernel is not supported
  • [00:30:23] <recalcati> if I can get the u-boot source I could add the same settings in modedb.c kernel file
  • [00:30:44] <recalcati> but I don't know the right u-boot source
  • [00:32:34] <raster> i havent tried 720p
  • [00:32:41] <raster> i actually dropepd down to 640x480(@32bit)
  • [00:32:45] <recalcati> u-boot is correct
  • [00:32:47] <raster> as i dont much need the high resolutions
  • [00:33:08] <recalcati> LCD normally has that resolution
  • [00:33:30] <recalcati> bitbake u-boot ?
  • [00:34:15] <recalcati> I have compiled u-boot-beagleboard-2008.10+r2+gitr08d7fdcce5dde5c2dc968fac5b2facf17cbabb5-r2.bin .
  • [00:34:18] <recalcati> is ols
  • [00:34:20] <recalcati> old
  • [00:34:25] * mrgarlic (n=mrgarlic@pool-70-19-69-81.ny325.east.verizon.net) Quit (Connection timed out)
  • [00:34:37] <recalcati> maybe from normal denx git tree ?
  • [00:41:37] * ddompe (n=ddompe@ip221-27-10-190.ct.co.cr) has joined #beagle
  • [00:42:12] * mrgarlic (n=mrgarlic@pool-70-19-69-81.ny325.east.verizon.net) has joined #beagle
  • [00:43:21] <recalcati> I set chown root.root /home/user/.e
  • [00:44:10] <recalcati> if a stupid user try to unlock he cen, but the config doesn't change...
  • [00:44:27] <raster> and e will complaqin
  • [00:44:30] <raster> complain
  • [00:44:34] <recalcati> he can close the window, but the system will restart automatically
  • [00:44:38] <recalcati> no complain
  • [00:44:42] <raster> it should
  • [00:45:35] <recalcati> but at restart of the system, after login, it asks me .. languages .. angstrom ... s on
  • [00:46:15] <raster> yup
  • [00:46:23] <raster> it'll be most unhappy
  • [00:46:39] <raster> basically it's failed in loading config
  • [00:46:43] <recalcati> chmod -R 644 /home/recalcati/.e
  • [00:46:49] <recalcati> yes, done
  • [00:46:52] <recalcati> now I restart
  • [00:46:54] <raster> failed in even finding the profile it should use
  • [00:48:24] <recalcati> raster: do you know u-boot correct sources? or recipe?
  • [00:48:41] * fenn (n=fenn@cpe-72-177-52-180.austin.res.rr.com) Quit (Read error: 110 (Connection timed out))
  • [00:48:57] <raster> u havent played with uboot beyond boot params/env
  • [00:49:01] <raster> as it has worked for me
  • [00:51:16] <recalcati> I'd like to see which timings u-boot uses for HD720
  • [00:52:28] <recalcati> because now the kernel try 1280x768, but id not is not a good resolution for a TV
  • [00:53:25] <recalcati> raster: I put my password, but I can0t login
  • [00:53:40] <recalcati> maybe due to chmod -R 644 /home/recalcati/.e
  • [00:55:12] <Downix> ok, trying to find out, what is a "4 wire" LCD?
  • [00:55:27] <recalcati> what a stupid .. I hadn't the x for entering in .e dir
  • [00:57:00] <russ> Downix: are you sure you don't mean, LCD Touchscreen
  • [00:57:33] <Downix> russ: right
  • [00:57:57] <Downix> never studied touchscreens before, I prefer keyboards
  • [00:58:06] * mrgarlic_ (n=mrgarlic@pool-70-19-69-81.ny325.east.verizon.net) Quit (Read error: 110 (Connection timed out))
  • [00:58:57] <russ> recalcati: go look in drivers/video/ps3fb.c and look for 720p
  • [00:59:21] <recalcati> russ: I saw
  • [00:59:37] <recalcati> but I have little difference in pixel clock
  • [00:59:47] <russ> what do you mean?
  • [01:00:25] <recalcati> 1/(74.25*10^6)*10^12 = 13468
  • [01:00:29] <recalcati> in psec
  • [01:01:24] <russ> ok, so maybe the mode is a little odd
  • [01:01:35] <russ> its a mode to compensate for overscan or whatever
  • [01:02:00] <russ> try the 720pf one
  • [01:02:23] <russ> pixclock 13481
  • [01:02:36] <recalcati> "720p", 60, 1124, 644, 13481, 298, 148, 57, 44, 80, 5,
  • [01:03:01] <recalcati> "720pf", 60, 1280, 720, 13481, 220, 70, 19, 6, 80, 5,
  • [01:03:10] <russ> oh, same pixclock
  • [01:03:13] <recalcati> FB_SYNC_BROADCAST, FB_VMODE_NONINTERLACED
  • [01:03:29] <recalcati> I have never FB_SYNC_BROADCAST
  • [01:03:34] <russ> I think they are exactly the same from a signaling standpoint, the first one justs adds a black border
  • [01:04:00] <russ> "broadcast video timings"
  • [01:04:40] <russ> nothing in the video subsystem except ps3fb.c, amifb.c, and pvr2fb.c care about it
  • [01:05:29] <russ> so unless there is something internal in dss2 you want to when its set...
  • [01:05:32] <recalcati> I put 0
  • [01:05:56] <recalcati> and "720pf" become NULL
  • [01:06:16] <recalcati> or not?
  • [01:06:32] <russ> why make it null, you could name it explicitly in the boot args
  • [01:06:34] <recalcati> maybe I can use 720pf from cmdline to fin it
  • [01:06:36] <recalcati> maybe I can use 720pf from cmdline to find it
  • [01:06:42] <recalcati> ok
  • [01:06:43] <russ> too bad tomba isn't here
  • [01:07:27] <recalcati> omapfb.mode=dvi:1280x768MR-24@60 will become omapfb.mode=???
  • [01:07:59] <russ> dvi:720pf maybe?
  • [01:08:12] <russ> the timings I see only are only slightly different
  • [01:08:22] <recalcati> I'm compiling
  • [01:08:32] <russ> 72/216 front /back hsync porch
  • [01:08:48] <russ> 3/22 front/back vertical porch
  • [01:09:07] <russ> same 80, 5 on the hsync vsync
  • [01:09:13] <recalcati> yesterday I read CEA 681-D but I get crazy and didn't understand
  • [01:09:40] <russ> http://www.3dexpress.de/displayconfigx/timings.html <- EDID data from TVs
  • [01:12:20] <recalcati> I'll read it
  • [01:13:19] <recalcati> now I've been restarted with the same resolution, just to test the kernl
  • [01:18:54] <recalcati> not accepted 720pf
  • [01:20:04] <russ> try dvi:1280x768M-24@60
  • [01:23:08] <recalcati> I retry
  • [01:23:19] <recalcati> sometimes I don't see nothing
  • [01:25:48] * lcuk (i=lcuk@cpc3-oldh7-0-0-cust590.manc.cable.ntl.com) Quit ("Leaving")
  • [01:27:43] <recalcati> fbcvt: 1280x768@60: CVT Name - .983M9
  • [01:27:47] <recalcati> not supported
  • [01:28:43] <recalcati> now I try 1280x720??
  • [01:30:51] <recalcati> fbcvt: 1280x720@60: CVT Name - .921M9-R not supported . using omapfb.mode=dvi:1280x720MR-24@60
  • [01:31:38] <recalcati> http://widget.mibbit.com/pb/RZznFd pixel clock too low
  • [01:33:57] <recalcati> fbcvt: 1280x720@60: CVT Name - .921M9-R not supportes . using omapfb.mode=dvi:1280x720MR-24@60
  • [01:35:34] * like2wise (n=chatzill@82-171-51-231.ip.telfort.nl) Quit ("ChatZilla 0.9.84 [Firefox 3.0.8/2009032711]")
  • [01:45:57] * ddompe (n=ddompe@ip221-27-10-190.ct.co.cr) Quit ()
  • [01:46:34] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [01:49:02] * davidcb (n=chatzill@c-76-124-23-143.hsd1.nj.comcast.net) has joined #beagle
  • [01:53:14] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [01:57:08] * bmxr (n=bmxr@S01060018f3b11a22.vf.shawcable.net) Quit (Read error: 104 (Connection reset by peer))
  • [01:58:50] <recalcati> trying to select http://widget.mibbit.com/pb/CXKVNw
  • [02:16:38] * bmxr (n=bmxr@S01060018f3b11a22.vf.shawcable.net) has joined #beagle
  • [02:21:24] * ESC09 (i=4c678e10@gateway/web/ajax/mibbit.com/x-33f81f5de32295cb) Quit ("http://www.mibbit.com ajax IRC Client")
  • [02:31:22] * russ (i=foobar@ip70-176-253-20.ph.ph.cox.net) Quit (Read error: 60 (Operation timed out))
  • [02:32:27] * recalcati (i=5d90c85d@gateway/web/ajax/mibbit.com/x-f439037748db31af) Quit ("http://www.mibbit.com ajax IRC Client")
  • [02:38:19] * brater36 (i=626d5feb@gateway/web/ajax/mibbit.com/x-0c0b1a332c9a2b31) has joined #beagle
  • [02:38:37] <brater36> hey
  • [02:39:01] <brater36> anyone implemented serial UART communication on the beagleboard?
  • [02:47:16] <ds2> yes
  • [02:47:21] <ds2> many months ago
  • [02:48:09] <brater36> oh really, ive been looking for days trying to fine some code for juct sending bytes to the UART on the expansion header
  • [02:48:26] <ds2> what are you doing?
  • [02:48:38] * bavison (n=ben@cpc2-cmbg9-0-0-cust635.cmbg.cable.ntl.com) Quit (Read error: 110 (Connection timed out))
  • [02:48:39] <brater36> im trying to implement a DMX512 signal
  • [02:48:43] <ds2> sending bytes can be as simple as cat foo > /dev/ttyS1
  • [02:48:47] <ds2> or echo FOO > /dev/S1
  • [02:48:52] <brater36> in C code?
  • [02:48:56] <ds2> can you elaborate on DMX512, not familiar with it
  • [02:49:21] <ds2> in C, open( "/dev/ttyS1", ...); write(...); close()
  • [02:50:16] <brater36> its a 250Kbps serial signal, basically, its a repeating serial signal, send 512 bytes one after the other, then cause a break which syncs all te receivers, and then sent all 512 bytes again, forever lol
  • [02:50:52] <brater36> its used in the lighting industry for control of moving concert lights and such
  • [02:50:58] <ds2> I see
  • [02:51:06] <ds2> is this what NXP demos?
  • [02:51:34] <brater36> not sure? where could i find that, ill see
  • [02:52:03] <ds2> I donno... I saw an NXP demo for fancy light control at a trade show
  • [02:52:23] <brater36> oh its possible I guess
  • [02:52:49] <ds2> but getting back to your stuff, you probally have to add an ioctl() to invoke the break
  • [02:53:04] <ds2> and maybe some tcsetattr() calls to set to set the baudrate
  • [02:53:19] <brater36> ok
  • [02:53:36] <brater36> where did you find this info, cuz ive been lost for a while lol
  • [02:53:57] <brater36> im kinda new to embedded linux, spent most of my time writinf C for 8-bit MUCs
  • [02:54:03] <ds2> experience
  • [02:54:07] <ds2> man pages, etc
  • [02:54:11] <brater36> ahh
  • [02:54:19] <ds2> I do Linux stuff for a living :D
  • [02:54:26] <brater36> oh very nice
  • [02:55:26] <watarat> Seem to remember thatthe 2 byte Break is the hard part to organise with some UARTS
  • [02:55:43] <ds2> for the BB it is not an issue; the OMAP emulates a 16550
  • [02:55:52] <ds2> uses the same driver
  • [02:55:59] <watarat> can be a gotcha - at least was that way with some of the early MCU implementtions
  • [02:56:09] <brater36> yeah, most likely, when i did it on an 8bit mcu, i just shut off the transmitter for a time period
  • [02:56:23] <ds2> worse case, wire in a transistor and... :D
  • [02:56:32] <brater36> yeah
  • [02:56:49] <brater36> how do you set the frame format, parity etc?
  • [02:57:12] <watarat> You transmitting or receiving DMX - can be worse problem on Receive - depending on twhat the Break does incoming.
  • [02:57:27] <brater36> transmitting
  • [02:57:35] <watarat> Not so bad!
  • [02:58:05] <watarat> Some old UARTS just gave you F's continuously through the break
  • [02:58:12] <brater36> when i wrote this program for an 8-bit mcu, i did it interrupt driven so that as soon as a byte was sent the next was loaded and sent immediatly, is there a reason to do that for this implementation
  • [02:58:40] <ds2> tcsetattr is the call, IRRC
  • [02:58:44] <ds2> IIRC
  • [03:00:06] * mrgarlic (n=mrgarlic@pool-70-19-69-81.ny325.east.verizon.net) Quit ()
  • [03:00:22] <brater36> ok ill give it a try, any idea how to setup that mux on the BB
  • [03:00:27] <watarat> ds2 - is there a problem going in and setting the hardware registers from a user app - or does that need to be done as part of a 'driver'
  • [03:08:51] <brater36> thank a lot guys
  • [03:09:04] <ds2> waterat:that should only be done from a driver
  • [03:09:09] <ds2> or you will confuse things
  • [03:09:32] <brater36> any idea how to configure the mux?
  • [03:10:03] <ds2> brater36: U boot or the kernel; for the kernel - I have a write up. Note there is a current bug that is breaking the kernel MUX config stuff
  • [03:10:24] <brater36> so Uboot prob best
  • [03:10:41] <ds2> if you have control of it, yes.
  • [03:10:52] <brater36> anyway to do it in code?
  • [03:10:55] <ds2> kernel MUX lets you do other stuff
  • [03:11:04] <ds2> not in userland, just in kernel code
  • [03:11:20] <ds2> well, that's not quite correctly - you can subvert EVERYTHING and go in underneat with /dev/mem
  • [03:11:27] <brater36> gotcha so ill do it in Uboot i guess
  • [03:11:55] <brater36> so just need to configure Uboot b4 the system boots right?
  • [03:12:06] <ds2> no
  • [03:12:11] <ds2> you need to rebuild U-boot from source
  • [03:12:19] <ds2> edit beagle.h
  • [03:12:35] <brater36> ahh, umm well i shoudl be able to do that, i got OE up an running so
  • [03:13:04] <ds2> OE he he he he he OE he he he he
  • [03:13:20] <brater36> oh now what, lol
  • [03:13:26] <brater36> oh no**
  • [03:13:29] <brater36> lol
  • [03:13:42] <ds2> nothing :D
  • [03:13:50] <brater36> lol
  • [03:14:07] * L84Supper (n=ly@c-68-53-183-162.hsd1.il.comcast.net) Quit (Remote closed the connection)
  • [03:15:23] <ds2> I wonder how convoluted it is it to do it in OE...
  • [03:17:00] <brater36> i was just looking for beagle.h and cant seem to find it, there are like 8 of them
  • [03:17:14] <brater36> oe does seem a bit strange lol
  • [03:17:22] * L84Supper (n=ly@c-68-53-183-162.hsd1.il.comcast.net) has joined #beagle
  • [03:20:37] <ds2> one sec
  • [03:20:57] <ds2> board/omap3/beagle/beagle.h
  • [03:22:19] <brater36> wheres that board dir?
  • [03:22:20] <brater36> cant seem to find it
  • [03:24:12] <ds2> in u-boot
  • [03:24:27] <ds2> OE is hiding it if it is not there
  • [03:24:40] <brater36> hmm
  • [03:24:49] <ds2> I did wonder about how convolunted it is with, did I not? :D
  • [03:25:10] <ds2> <-- won't touch OE unless someone is offering money
  • [03:25:18] <ds2> there are better ways to do things then with OE
  • [03:26:03] <brater36> like what, im open to other solutions, OE was a pain to get working too lol
  • [03:26:16] <ds2> I just build things from their GIT trees or tarballs
  • [03:26:57] <brater36> ahh
  • [03:27:41] <ds2> life is too short to spend it fighting the latest build system fad
  • [03:28:23] <brater36> very tru
  • [03:29:26] <djlewis> I just want to write simple s apps directly on the BB
  • [03:29:37] <djlewis> simple C apps
  • [03:29:37] <ds2> s apps?
  • [03:29:39] <ds2> oh
  • [03:29:49] <ds2> thought you want to do asm
  • [03:30:01] <djlewis> used to
  • [03:30:23] <raster> given u use c and need enough libs
  • [03:30:26] <raster> u will want a build system
  • [03:30:31] <raster> u can do it on the bb
  • [03:30:35] <raster> but man is it slow
  • [03:30:43] <ds2> raster: it isn't that bad
  • [03:30:59] <raster> ds2: compared to a cross-compile on a dual core 3ghz box...
  • [03:31:01] <raster> its pain
  • [03:31:09] <ds2> raster: no, I mean building w/o OE
  • [03:31:13] <ds2> not native only compile
  • [03:31:20] <raster> you mean hand-rolled cross-dev?
  • [03:31:30] <ds2> if you want to make it look bad, sure
  • [03:31:44] <raster> given you use enough libs
  • [03:31:48] <raster> you will be in a world of pain
  • [03:32:00] <raster> if the only thign u use is libc
  • [03:32:02] <raster> its easy
  • [03:32:02] <ds2> no, that's a libtool/pkgconf issue
  • [03:32:13] <raster> yes
  • [03:32:13] <ds2> proper libraries do not present many issues
  • [03:32:17] <raster> but everything uses it
  • [03:32:22] <ds2> do not use that peice of crap
  • [03:32:39] <raster> there is nothing else
  • [03:32:45] <raster> other than hand-rolling your -l's
  • [03:32:47] <raster> and -L's
  • [03:32:48] <raster> -I's
  • [03:32:49] <raster> etc.
  • [03:32:59] <ds2> using them is like driving around with a deployed parachute and insisting you need jet engines to get up to highway speeds
  • [03:33:21] <ds2> hand rolling -l's, -L's and -I's are easy
  • [03:34:19] <ds2> -I's shouldn't be needed if things are installed right
  • [03:34:23] <ds2> same with -L's
  • [03:34:30] <djlewis> Say I want to write a gui GPS display what is simple?
  • [03:34:46] <ds2> ncurses :D
  • [03:34:54] <djlewis> local on BB or outside?
  • [03:35:00] <raster> ds2: i know how it works :) it just becomes uber-painful when you have 30 libs u link against
  • [03:35:16] <ds2> raster: I come from the days before libtools and it was way easier w/o it
  • [03:35:21] <raster> and you may or mayt not know if the lib has built with lib-linking or not
  • [03:35:28] <raster> so u dont know if u have to link to its deps for it or not
  • [03:35:34] <raster> (as some distros insist on doing)
  • [03:35:39] <raster> so did i
  • [03:35:51] <raster> i was writign compile.sh's for cross-compiling years ago
  • [03:35:54] <ds2> did I mention I also dislike most distros
  • [03:35:54] <raster> hand rolling my own
  • [03:36:10] <ds2> distros are a waste of time... start from a rootdisk and go from there
  • [03:36:18] <raster> the problem is - if u want somehting to "work" on multiple distros - with minimum hassle. you end up with autofools
  • [03:36:23] <raster> yes - everyone hates them
  • [03:36:25] <raster> its universal
  • [03:36:43] <ds2> guess if everyone hates them, I am the only willing to do something about them :D
  • [03:37:03] <raster> but it is actually autotools that has meant that at least dozens of libs and apps iat least in E land.. have "just worked" (tm) within cross-build systems
  • [03:37:17] <raster> i hate them too - but i've come to see them as a pain you need :(
  • [03:37:25] <ds2> autoconf is the worse thing ever
  • [03:37:37] <ds2> it is a curse for support, a curse for users, a curse for everything.
  • [03:37:37] <raster> it offends me too
  • [03:37:45] <ds2> config.h works.
  • [03:37:48] <raster> some of my libs - the configure script is 5x the side of the src for the lib
  • [03:37:54] <raster> its atrocious
  • [03:38:06] <raster> the tarball u ship is 300kb
  • [03:38:09] <raster> but the src is 5kb
  • [03:38:11] <raster> (compressed)
  • [03:38:13] <ds2> right so rewrite things to use a config.h
  • [03:38:21] <raster> its not that alone
  • [03:38:35] <raster> i have to say i have come to royally love make distcheck
  • [03:38:42] <ds2> people who use the code needs to have SOME intelligence...it helps keep down the support issues
  • [03:38:51] <raster> it has meant i can deliver deliverables to my clients.. in about 3 minutes
  • [03:38:59] <raster> and i know its qa'd
  • [03:39:16] <ds2> it is not QA'ed til it is really QA'ed
  • [03:39:30] <raster> from a packaging the source up point of view.. it is
  • [03:39:40] <raster> it doesnt include spurious build and temp files
  • [03:39:47] <raster> makes sure everything is actually included thats needed
  • [03:39:56] <raster> and that it builds inside or outside the tree
  • [03:40:06] <ds2> isn't that like git clean?
  • [03:40:08] <raster> with differing build dirs and more
  • [03:40:10] <raster> no
  • [03:40:21] <raster> it keeps develope-rgenerated files
  • [03:40:35] <raster> ie automake does Makefile.am -> Makefile.in
  • [03:40:46] <raster> in git/svn/cvs you dont keep the Makefile.in's
  • [03:40:48] <raster> ever
  • [03:40:54] <raster> as they are generated from the .am's
  • [03:41:03] <raster> but in a disted soruc tarball you do
  • [03:41:23] <ds2> ah but if we avoid autoconf abominiation, this isn't an issue
  • [03:41:33] <raster> i dont need to worry about .git or .svn or CVS scms dirs in the tree
  • [03:41:37] <raster> they wotn be packaged up
  • [03:41:43] <raster> ooh no
  • [03:41:51] <raster> u nstil lneed to weed out the scms files
  • [03:41:53] * satendra (i=dce0e51e@gateway/web/ajax/mibbit.com/x-fe93b240011dfb26) Quit ("http://www.mibbit.com ajax IRC Client")
  • [03:42:03] <raster> it alows me to keep scratch files and dirs in my tree
  • [03:42:06] <raster> that are nmot scsm tracked
  • [03:42:10] <raster> but i know wont be disted
  • [03:42:11] <ds2> w/CVS, I can see theproblem
  • [03:42:22] <ds2> but with .git, it shouldn't be much of a problem
  • [03:42:26] <raster> i just cease worrying
  • [03:42:33] <raster> also it means i know it wil work with
  • [03:42:41] <raster> make DESTDIR=/tmp/whatevere
  • [03:42:54] <raster> so actually building packages (.deb, .rpm etc.) is alwasy clean and correct
  • [03:42:59] <raster> i get "make uninstall" for free
  • [03:43:01] <ds2> ewwwwwwwwwwwwwww packages
  • [03:43:06] <raster> as all instaleld stuff is tracked
  • [03:43:21] <raster> reality is - people use and want packages
  • [03:43:29] <ds2> of course, this isn't going to get resolve... you spend more of your time in userland... I spend it in kernel land
  • [03:43:30] <raster> you end up caterign for them
  • [03:43:46] <ds2> packages are a abomination; for maintainers and for users
  • [03:43:48] <raster> you can try and sit in a bubble and ignore the world of distros and packages.. but its there.. and thats how it works
  • [03:43:57] <raster> and no - packages are good
  • [03:44:08] <ds2> they are bad
  • [03:44:26] <ds2> bad enough to piss off maintainers into locking up code
  • [03:44:29] <raster> comign from, having worked # rh and done rpm, done custom debina distros, doing oe, i've bene in slackware land before and i've rolled my own hand-build embedded rootfs's and kernels before
  • [03:44:33] <raster> i'd take packages any day
  • [03:45:05] <raster> users are too stpuid to compile stuff
  • [03:45:15] <ds2> when a distro takes snapshots and makes packages out of them, bad things happen
  • [03:45:20] <raster> admittedly a lot oof package maintainers are also not a lot better
  • [03:45:24] <raster> they "think they know"
  • [03:45:33] <raster> and thu7s have a false sense of competence
  • [03:45:45] <raster> and as such a distro should take releases
  • [03:45:46] <raster> most do
  • [03:45:52] <raster> but thats their policy
  • [03:46:08] <ds2> packages are fine if they can guarantee their users will only use their stuff and ask them for support
  • [03:46:09] <raster> a snapshot is no different ot users just chekcing out whatever is in git/svn now
  • [03:46:12] <brater36> thanks for all the help!!!
  • [03:46:13] <ds2> but that is not the world
  • [03:46:13] <raster> or a particular date/revision
  • [03:46:24] <ds2> brater36: sorry for going off on a tangent
  • [03:46:33] <brater36> haha its alright
  • [03:46:36] <ds2> raster: no, it is worse cuz a package makes it look official
  • [03:46:43] <brater36> later
  • [03:46:44] <raster> ds2: yeah/ that has annoyed the crap out of me for long enough
  • [03:46:48] <raster> packagers package up
  • [03:46:51] <raster> then they patch it and futz it
  • [03:46:54] <ds2> take pilot-link for example... redhat was taking development snapshots and packaging them up
  • [03:47:02] * brater36 (i=626d5feb@gateway/web/ajax/mibbit.com/x-0c0b1a332c9a2b31) Quit ("http://www.mibbit.com ajax IRC Client")
  • [03:47:03] <raster> and the users come back to upstream devs asking bizarre quesitons on stuff u never released
  • [03:47:08] <ds2> then serious pissed off the pilot-link maintainer cuz of all the questions
  • [03:47:09] <raster> or you never had configured/set up that way
  • [03:47:19] <ds2> exactly
  • [03:47:25] <raster> i have bene pissed off royally by this often enough. thats half the reason i moved to binary configs for e
  • [03:47:30] <raster> to thrawt the damn packagers
  • [03:47:31] <ds2> so I much rather packages not existed
  • [03:47:53] <raster> make it just complicated enough where the level of clue you need puts you in the "well you'll do it right" crowd
  • [03:48:02] <raster> otherwise you take it as we ship it.. or go away
  • [03:48:15] <raster> but reality is people will package
  • [03:48:17] <raster> and so on
  • [03:48:18] <raster> and use them
  • [03:48:29] <ds2> forcing people to goto the tarball enforces a certain level of clue
  • [03:48:35] <raster> so i work to make sure the packages as closely as possible reflect what upstream have/release/do
  • [03:48:43] <raster> anyway
  • [03:48:45] <raster> gotta run
  • [03:48:47] <ds2> :D
  • [03:48:49] <ds2> later
  • [03:48:51] <raster> real life kicking in!
  • [03:48:52] <raster> :)
  • [03:48:59] <ds2> <--- finishing dinner
  • [03:52:34] * _gm__ (n=_gm_@190.173.66.170) has joined #beagle
  • [03:58:04] * _gm___ (n=_gm_@190.173.83.184) Quit (Read error: 110 (Connection timed out))
  • [04:12:40] * xndr (n=alex@cpe-74-66-14-187.nyc.res.rr.com) has joined #beagle
  • [04:26:06] * _gm___ (n=_gm_@190.173.71.132) has joined #beagle
  • [04:26:14] * davidcb (n=chatzill@c-76-124-23-143.hsd1.nj.comcast.net) has left #beagle
  • [04:27:51] * MostAwesomeDude (n=simpson@c-98-232-178-218.hsd1.or.comcast.net) has joined #beagle
  • [04:34:30] * _gm__ (n=_gm_@190.173.66.170) Quit (Read error: 110 (Connection timed out))
  • [04:41:44] * montamer (n=vijay@203.199.213.3) has joined #beagle
  • [04:46:55] * mib_jn81mk (i=8e3bb839@gateway/web/ajax/mibbit.com/x-09f030c6405a66dd) has joined #beagle
  • [04:47:07] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [04:48:07] <mib_jn81mk> Is the serial port output on this TTL, or level shifted to RS232? I seem to recall hearing there was no RS232 translator IC on this so it must just be low voltage out. HOwever, they are recommending using a rs232 to usb converter...
  • [04:49:13] <ds2> the Console header is RS-232
  • [04:49:24] <ds2> the UART on the expansion header is CMOS 1.8V
  • [04:50:08] <mib_jn81mk> So its only outputting 1.8V signals... not rs232 actually...
  • [04:50:23] <ds2> which one?
  • [04:50:31] <ds2> there are 2 exposed ports on the board
  • [04:51:16] <mib_jn81mk> the header
  • [04:51:28] <mib_jn81mk> that they used at esc to demo
  • [04:51:50] <ds2> they are both headers, you will have to be more specific
  • [04:52:06] <ds2> I had a board that uses both of them and I was at ESC... some peopledid see it so... :D
  • [04:52:24] * szricky (i=74061432@gateway/web/ajax/mibbit.com/x-28532d632e9d2e7a) has joined #beagle
  • [04:52:29] <mib_jn81mk> in the demo / class
  • [04:52:38] <ds2> is it a header near the center of the board or on the edge?
  • [04:53:00] <ds2> the BYOE sessions? they serial port wasn't used unless you had an issue... but if I were to guess, it is the center header with 10 pins
  • [04:53:01] * mrc3 (n=ddiaz@189.157.115.100) has joined #beagle
  • [04:53:09] <ds2> if so, that port is indeed at RS-232 levels.
  • [04:53:31] <mib_jn81mk> No, Jason explained how to connect to it so you could do development.
  • [04:53:42] <ds2> Oh during the Beagleboard 101 session.
  • [04:53:44] <mib_jn81mk> but he said there was no level shifter?
  • [04:53:46] <mib_jn81mk> rs232 has a typical voltage standard of a space 5 to 15V and space -5 to -15V,
  • [04:53:51] <ds2> That is the 10pin header
  • [04:54:05] <ds2> it is at RS232 levels and there is a RS-232 transceiver for that
  • [04:54:24] <ds2> I don't think he did anything with the expansion header in Beagle board 101
  • [04:55:09] <ds2> it is a TI part equiv. to the common MAX232 style chips
  • [04:55:18] <mib_jn81mk> OK.
  • [04:55:26] <mib_jn81mk> Yes, he didnt connect the other port.
  • [04:55:57] <ds2> let me find a picture
  • [04:56:09] <mib_jn81mk> no one has usb to rs232 converters so we could only listen to how to do it but now I'm home... so i want to try it
  • [04:56:42] <ds2> actually, the instructors all had one to fix problems
  • [04:56:50] <ds2> I had mine, so did klass, etc
  • [04:56:58] <mib_jn81mk> yeah, there were 3 he brought too :)
  • [04:57:04] <ds2> okay - refering to the picture on www.beagleboard.org
  • [04:57:25] <mib_jn81mk> But we didn't try the demo ... now I forget what he was shoing with it ! via the serial port I thought there were some cool things one could do
  • [04:57:31] <ds2> the white circles on the right edge of the board is the expansion header. It is not normally populated
  • [04:57:42] <ds2> that contains the pins for UART2
  • [04:57:57] * khasi1 (n=a0393720@192.163.20.231) has joined #beagle
  • [04:58:14] <ds2> the 10 pin header on the left next to the power jack and MMC slot. That is UART3. That port _IS_ at RS-232 levels
  • [04:58:19] <ds2> does this help?
  • [04:58:34] <mib_jn81mk> yup - that makes sense.
  • [04:58:39] <ds2> 'k
  • [04:59:02] <ds2> I had the expansion port populated so...
  • [04:59:21] * _gm___ (n=_gm_@190.173.71.132) Quit (Read error: 60 (Operation timed out))
  • [04:59:26] * ESC09student (i=4c678e10@gateway/web/ajax/mibbit.com/x-ac1da66ceb9d39b1) has joined #beagle
  • [04:59:27] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [04:59:33] <mib_jn81mk> Weren't there some neat things to do with the serial port? Is that what was used to switch the mouse from your PC to the Linux PC etc
  • [04:59:53] <ds2> well, it is the console port and the only way to talk to U-boot
  • [05:00:02] * GregorR (n=gregor@65.183.185.209) Quit (Remote closed the connection)
  • [05:00:57] <ds2> you could also hook it up to a modem and do PPP with it
  • [05:01:11] <ds2> basically the same thing you would use a serial port on a PC
  • [05:01:20] <ESC09student> linux n00b here. LOL I couldn't figure out why some dirs and files had a little square block before their names. it was the letter 'm' !!!! I remember from the lectures that the terminal font size isn't easy to change, any tips?
  • [05:02:05] <ds2> AFAIK, the image lacks fants
  • [05:02:06] <ds2> fonts
  • [05:02:12] <mib_jn81mk> Yeah, maybe it was just to have access to u-boot... I thought there were some other reason to have access to it though via your pc - while developing
  • [05:02:25] <ds2> you could try CTRL-right click to bring up a menu
  • [05:02:50] <ds2> mib_jn81mk: U-boot access is needed if you want to setup your board to boot with the onboard NAND
  • [05:04:02] <mib_jn81mk> yeah - but there was more... I'm looking at the slides, but the demo's that he didn't get to dont ring a bell... I thought there was some cool demo that needed rs232 access....
  • [05:06:43] <ds2> don't think there is anything remarkable but then that might be because I am used to having console access
  • [05:07:38] <mib_jn81mk> hmmm... maybe it was just for having access to make a boot script then...
  • [05:08:05] <ds2> that could be or changing kernel config
  • [05:09:20] <mib_jn81mk> unfortunately the slides are quite sparse so I'm forgetting what some of them meant... Like there is one "access monitor EDID" then gives four points... not sure what it was intending to show.
  • [05:10:13] <ds2> there are commands in U-boot to send i2c commands
  • [05:10:29] <ds2> I think he is staying - you can use that to query EDID in u-boot
  • [05:11:12] * GregorR (n=gregor@65.183.185.209) has joined #beagle
  • [05:11:33] <mib_jn81mk> I'll have to google EDID ;)
  • [05:11:54] <ds2> EDID is a protocol for a monitor to id itself
  • [05:12:00] <ds2> it is built on top of I2C
  • [05:13:07] <mib_jn81mk> thansk
  • [05:13:41] * _gm___ (n=_gm_@190.173.72.116) has joined #beagle
  • [05:13:42] <mib_jn81mk> Were you there when he showed setting up the server so that you could move your mouse from the beagle to your PC?
  • [05:14:24] * dirk (n=dirk@p5B040840.dip0.t-ipconnect.de) has joined #beagle
  • [05:14:33] * BThompsonD (n=bernie@cpe-72-190-75-99.tx.res.rr.com) Quit (Remote closed the connection)
  • [05:14:52] <ds2> no, I got there later
  • [05:15:04] <ds2> I believe he uses synegy to do that
  • [05:15:12] <mib_jn81mk> I think he did it around lunch so most were gone...
  • [05:15:20] <mib_jn81mk> you were a ta right? Or just a speaker?
  • [05:15:23] <ds2> Oh wait, I was there
  • [05:15:32] <ds2> All the speakers were TA's
  • [05:15:39] <ds2> we all help each other
  • [05:15:47] <mib_jn81mk> that's right
  • [05:15:56] <mib_jn81mk> Liked your talk I remember
  • [05:16:53] <ds2> don't remember much from early monday... too many things happening
  • [05:17:09] <mib_jn81mk> Yes it was quite the week.
  • [05:17:29] <ds2> indeed...
  • [05:17:41] <ds2> it was tiring but fun
  • [05:18:16] <mib_jn81mk> I agree
  • [05:18:57] <ds2> do you have embedded for projects in mind?
  • [05:21:17] * DiBruno (n=blaine@d216-232-231-113.bchsia.telus.net) has joined #beagle
  • [05:21:40] <mib_jn81mk> Not really a specific project at the moment...
  • [05:22:25] <ds2> I am curious if most people are interested in using it as an embedded thing or as a portable PC
  • [05:23:06] <mib_jn81mk> For networking what do most people do, put on a usb/wifi dongle?
  • [05:24:02] <ds2> that is common
  • [05:24:25] <ds2> or for a few of us, wire in a Ethernet via the expansion port
  • [05:24:52] <mib_jn81mk> Via the serial expansion port you were talking about before?
  • [05:25:13] <ds2> the one on the right
  • [05:25:30] <ds2> there are provisions for a second MMC slot, GPIOs, etc
  • [05:26:42] <GregorR> ds2: Is a wearable computer an embedded project or a portable PC? :)
  • [05:28:03] <ds2> GregorR: are you running a desktop like Gnome or KDE on it? :D
  • [05:29:19] <GregorR> Wasn't planning on it, the wearable display is 640x480, so I'm not sure what WM I want to use yet.
  • [05:29:34] <mib_jn81mk> yes, that's a definition thing... depends if you use the old definition of what an 'embedded system' is.
  • [05:29:38] <GregorR> I'm actually leaning towards Matchbox. With that little resolution, fullscreen generally makes more sense.
  • [05:30:35] <GregorR> (Although 640x480 is sort of right on the edge; at 320x240 I'd say full-screen is absolutely necessary, and on 800x600 I'd say a proper windowing system is the obvious choice, but 640x480 could go either way)
  • [05:32:13] * bmxr (n=bmxr@S01060018f3b11a22.vf.shawcable.net) Quit ("ChatZilla 0.9.84 [Firefox 3.0.8/2009032712]")
  • [05:33:15] <ds2> what kind of input devices will you have?
  • [05:33:42] <mib_jn81mk> ...eeek... .I need a new laptop battery.... (sorry - said that outloud)
  • [05:33:50] <GregorR> I'm keeping it simple for price/sanity sake, a small bluetooth keyboard and this strange little ring-mounted mouse.
  • [05:34:01] <mib_jn81mk> cool
  • [05:34:31] <GregorR> I may get other things when I have the full system running, depends on how things go with it.
  • [05:35:07] <ds2> I see... is it a HMD or just screen strapped onto your arm?
  • [05:35:52] <GregorR> HMD. A Myvu Crystal modified to be unintrusive and one-eye.
  • [05:36:51] <GregorR> (I saw some brief mention on a blog about the Myvu Crystal being suitable for that, and my desire to see if I could do the same started this whole project X-D
  • [05:36:53] <GregorR> )
  • [05:37:03] * _gm___ (n=_gm_@190.173.72.116) Quit (Read error: 110 (Connection timed out))
  • [05:37:37] * _gm__ (n=_gm_@190.173.65.199) has joined #beagle
  • [05:37:43] <mib_jn81mk> HMD
  • [05:39:57] <ds2> Hmmmm
  • [05:40:36] <GregorR> Mmhm.
  • [05:40:45] <GregorR> (See, I can type content-free one-word messages too! :) )
  • [05:41:00] <ds2> aren't they S-Video in?
  • [05:41:09] * szricky (i=74061432@gateway/web/ajax/mibbit.com/x-28532d632e9d2e7a) Quit ("http://www.mibbit.com ajax IRC Client")
  • [05:41:17] <GregorR> Worse, composite in :(
  • [05:41:37] <GregorR> (Of course, an S-video-to-composite adapter costs $1, but there's a degredation in quality)
  • [05:41:42] <ds2> the good thing is the BB can do that
  • [05:41:57] <ds2> but that really means 320x200 is more realistic then 640x480
  • [05:42:21] <GregorR> Weeeelll, 640x480 without making poor choices w.r.t. fonts an such.
  • [05:42:46] <GregorR> If the icons are big enough and the fonts are big enough, 640x480 will still be better than 320x240, although it won't be a perfectly crisp 640x480 either *shrugs*
  • [05:42:53] <ds2> can the Myvu really show all 640 pixels?
  • [05:43:24] <GregorR> It is a true 640x480 display, yes, I only don't trust the transport technology :P
  • [05:43:51] <GregorR> (The Crystal is the only display they make that's 640x480, all the others they make are 320x240 or worse)
  • [05:44:30] <ds2> ah
  • [05:44:45] <ds2> rarely do the S-Video in displays do all 640 per line
  • [05:44:58] <ds2> let alone composite
  • [05:45:47] <GregorR> Well, that's what I mean; the actual screen in the Myvu Crystal is 640x480, but using composite to drive that, in spite of the fact that composite NTSC is <bleh>x480, is never ideal.
  • [05:46:10] <GregorR> So I trust the screen, but not the transport, but I still feel 640x480 will be a better choice so long as I'm careful about fonts etc.
  • [05:46:49] * Grackle (i=826fefdf@gateway/web/ajax/mibbit.com/x-ea561907cfd70169) has joined #beagle
  • [05:48:04] <ds2> IIRC, NTSC is 480 only if interlaced
  • [05:48:22] <GregorR> NTSC is NTSC only if interlaced :P
  • [05:48:24] <ds2> the best way is to build it and see
  • [05:48:29] <GregorR> Yup
  • [05:49:15] <Grackle> I recently got a Revision C2 board, and I installed the angstrom demo system. Unfortunately, I'm getting no video output. The beagle logo shows up while u-boot is running, but no output is shown when the kernel boots and I try to run mplayer with the sample video.
  • [05:49:36] <Grackle> I'm using the HDMI output.
  • [05:50:20] <Grackle> Any thoughts on what could be going wrong here?
  • [05:51:36] <mib_jn81mk> Do you use a module for your ethernet setup ds2?
  • [05:53:05] * garren (n=garren@dsl-246-33-211.telkomadsl.co.za) has joined #beagle
  • [05:53:20] <GregorR> Grackle: There's some complicated trickery involved in telling the omapfb driver which video output device to use, I don't recall what the default is on Angstrom. It's documented mostly in a few mailing list posts, e.g. http://markmail.org/message/wcyq56ivlytksc7m
  • [05:53:49] <ds2> mib_jn81mk: Yeah... EN28J eval module
  • [05:56:51] <GregorR> Grackle: Hm, looking at my board, it seems like DVI is default though ...
  • [05:57:17] <mib_jn81mk> not showing up on google... EN28J
  • [05:58:19] <ds2> let me find the full number... 1 sec
  • [05:58:34] <Grackle> GregorR: yeah, I was following along here and it seems as if I'm doing everything right: http://code.google.com/p/beagleboard/wiki/BeagleboardRevCValidation
  • [05:58:39] <Grackle> except it doesn't display anything
  • [05:58:51] <djlewis> I installed the angstrom demo a few days ago and it worked fine on my new C2
  • [05:59:04] <ds2> ENC28J60 is the full part name
  • [06:00:28] <mib_jn81mk> Thanks
  • [06:00:46] <djlewis> Grackle: did you follow the http://elinux.org/BeagleBoardBeginners to the letter?
  • [06:02:17] <djlewis> Grackle: I believe the default startup is 1024x768x60Hz
  • [06:03:33] * russ (i=foobar@ip70-176-253-20.ph.ph.cox.net) has joined #beagle
  • [06:04:09] * russ is now known as Russ
  • [06:04:20] * Crofton (n=balister@wsip-70-164-137-164.lv.lv.cox.net) has joined #beagle
  • [06:04:29] * _gm__ (n=_gm_@190.173.65.199) Quit (Read error: 104 (Connection reset by peer))
  • [06:08:14] * kozak (n=subbu@117.192.15.35) has joined #beagle
  • [06:10:57] <mib_jn81mk> will any dvi to vga adaptor work on the beagle board?
  • [06:11:23] <Grackle> Trying again, djlewis
  • [06:11:34] <Russ> mib_jn81mk: no, there is no analog signal
  • [06:11:40] * Grackle pretty sure he followed it properly before, but trying again just to be sure
  • [06:11:41] <ds2> only if it is active and has DACs on it
  • [06:12:19] <Russ> ds2: true true
  • [06:13:54] <djlewis> Grackle: you verified your monitor is capable?
  • [06:14:01] <mib_jn81mk> Is this a dvi-i dvi-d dvi-A thing? So I can't get a cheap dongle to do this... ah... you're just talking about drive strength...
  • [06:14:20] <Russ> no, there is no analog signal
  • [06:14:22] <ds2> yes
  • [06:14:27] <ds2> Beagle is a DVI-D
  • [06:14:36] <djlewis> grackle: also, I had a usb hub with webcam and mouse on EHCI and added a keyboard
  • [06:14:38] * ArteK (n=Artur@81.15.241.96) has joined #beagle
  • [06:14:42] <Russ> er, yes,there is no alalog signal
  • [06:14:43] <ds2> you need DVI-A or DVI-I for VGA w/cheap converters
  • [06:14:52] <geckosenator> analog on svideo right?
  • [06:14:56] <djlewis> and did a reboot and lost video after post uboot.
  • [06:16:01] <Russ> well, yes
  • [06:16:19] <Russ> but I don't know what kind of hardware would be required to covert that to vga
  • [06:17:00] <garren> morning all
  • [06:17:08] <ds2> V-sync doubler basically
  • [06:17:10] <mib_jn81mk> If the beagle is DVI-D don't I just get a DVI-D to VGA? (Saw a cheap dongle with those specs). But you said it needs to be active so I need to get one that has a power supply>
  • [06:17:23] <Grackle> djlewis: yeah it should be capable, u-boot can show graphics, and at one point I tried a kernel that did output something, but the TV complained that it was out of range.
  • [06:17:43] <Grackle> but with the standard kernel it doesn't try to show anything
  • [06:17:44] <ds2> I suspect the specs for "DVI-D to VGA" to be suspect
  • [06:18:04] <GregorR> Unless it's not a dongle, and not cheap.
  • [06:18:09] <ds2> if it is indeed a DVI-D to VGA (i.e. there is electronics in there), then yeys, it'll work
  • [06:18:41] <djlewis> Grackle: what all do you have on USB
  • [06:18:49] <Grackle> nothing
  • [06:19:13] <mib_jn81mk> Yeah, I'll have to look into it... posts on the sight seem to indicate that it is dvi-d because someone wanted dvi-i and it was not dvi-i when they tried it... (china stuff)
  • [06:19:14] <Grackle> I'm powering the device with a USB cord on the gadget port
  • [06:19:16] <djlewis> Grackle: Ionly had a blank screenonce so far. USB was my cause.
  • [06:19:17] <Grackle> connected to my laptop
  • [06:19:34] <Grackle> and the host port is not connected to anything
  • [06:19:44] <djlewis> you are using a LCD monitor?
  • [06:20:14] <Grackle> I am using an HDTV
  • [06:20:23] <djlewis> That might be the problem.
  • [06:20:46] <Grackle> The HDMI cable is connected to a sony digital audio/video control center, which has multiple HDMI connections that it can connect to the TV
  • [06:20:51] <Grackle> You think the problem is there? hmm
  • [06:21:06] <djlewis> HDTV and LCD DVI are not one and the same
  • [06:21:14] <GregorR> If it displayed the logo from U-Boot, the problem must be on the board, no?
  • [06:21:32] <djlewis> I dont think there is a problem with your BB.
  • [06:21:44] <geckosenator> ds2: I have a DVI-D to VGA adapter and it didn't work
  • [06:22:01] <ds2> geckosenator: what's wrong with it?
  • [06:22:01] <djlewis> I suspect the signal resolution is changing after Uboot that is not compatible with the HDTV
  • [06:22:09] <ds2> HDTV is rather strict on the mode timing
  • [06:22:11] <geckosenator> ds2: it worked with a different computer not the beagle
  • [06:22:14] * _gm__ (n=_gm_@190.173.65.250) has joined #beagle
  • [06:22:15] <mib_jn81mk> WOW - seems like dvd-d to vga converters cost a lot... looks like the cheap ones are dvd-i to vga (the small dongles)
  • [06:22:31] <geckosenator> ds2: I had an lcd that took either vga or hdmi, and hdmi was working with the beagle at 640x480, but when I converted to vga.. no luck
  • [06:22:32] <ds2> the U-boot mode is hand crafted to work on HDTV, LCD monitors, and the Pico
  • [06:22:46] <ds2> geckosenator: I suspect it can be made to work
  • [06:22:53] <geckosenator> ds2: probably tweaking timings though
  • [06:23:04] <Russ> mib_jn81mk: right, because dvi-i has an analog signal
  • [06:23:26] <Russ> ds2: someone really needs to get the I2C bus reading the timings from the connected display
  • [06:23:37] <mib_jn81mk> So was Jason using a powered device at the conference for the display?
  • [06:23:50] <Grackle> hahhh
  • [06:24:00] <Grackle> when I connect it directly to the TV it works!
  • [06:24:07] <koen> Russ: tomba had a patch for dss2 that could read edid and choose mode accordingly
  • [06:24:09] <djlewis> cool
  • [06:24:15] <ds2> Russ: someone one did... Tomba threw together a quick patch
  • [06:24:24] <Russ> did it work well?
  • [06:24:31] <djlewis> Graackle: then you found a problem that you can get around.
  • [06:24:43] <ds2> mib_jn81mk: the thing Jason used is the a digital input provided by the ESC folks
  • [06:24:45] <djlewis> sorry I am tired and my typing is suffering.
  • [06:24:47] <koen> Russ: http://groups.google.com/group/beagleboard/msg/bd988bdaa65b7d58
  • [06:25:00] <Grackle> djlewis: yeah, good to know it was with my setup and not the board :)
  • [06:25:12] <GregorR> Dern, if I had a S-Video male-male cable, I'd have all the components I need, found out I already have an S-Video-to-composite adapter, but it's S-Video female. What a stupid cable to be the one thing that's missing X_X
  • [06:25:22] <mib_jn81mk> No - I recall when he was having all those problems he said something about I only have it going through this VGA converter box....
  • [06:26:05] <djlewis> I gotta get some sleeep here in CDT zone guys...
  • [06:26:09] <ds2> that box is from the ESC folks
  • [06:26:37] <ds2> the ESC folks knew in advance they needed to provide a digital input to their equipment (projector and I think a recorder of theirs)
  • [06:29:02] <mib_jn81mk> hmmmm might be cheaper then to get an s-video to vga adapter...
  • [06:29:55] * montamer_ (n=vijay@203.199.213.3) has joined #beagle
  • [06:30:32] <Russ> or a new DVI panel
  • [06:32:49] * Grackle (i=826fefdf@gateway/web/ajax/mibbit.com/x-ea561907cfd70169) Quit ("http://www.mibbit.com ajax IRC Client")
  • [06:33:23] <Russ> http://www.newegg.com/Product/Product.aspx?Item=N82E16824009166 <- 1366x768
  • [06:33:51] <geckosenator> Russ: are you replacing the oscillator?
  • [06:34:07] <Russ> I want to
  • [06:34:12] <GregorR> mib_jn81mk: S-Video can only do NTSC resolution ~= 640x480
  • [06:34:14] <Russ> If I can get the right part
  • [06:34:28] <Russ> hmmm..that panel claims DVI,but I don't see a connection
  • [06:34:56] <ds2> Russ: we should gather people together to make an improve BB
  • [06:35:58] <geckosenator> ds2: improve BB or add expansion boards?
  • [06:36:08] <ds2> geckosenator: both
  • [06:36:14] <geckosenator> ok
  • [06:36:18] <ds2> a BB with the PM improvements
  • [06:36:32] <Russ> right, maybe add a battery and backup battery connection
  • [06:36:41] <ESC09student> I'm using the BB to learn Linux, which Linux distro is most like Angstrom? should I just get the OReily Linux book?
  • [06:36:44] <ds2> maybe with an option to not stuff S-Video
  • [06:37:03] <Russ> ESC09student: depends on at what level you mean
  • [06:37:07] <ESC09student> n00b
  • [06:37:09] <koen> hmmm, big ARM story on slashdot
  • [06:37:13] <GregorR> ESC09student: Desktop distribution that's most like Angstrom? Probably Debian.
  • [06:37:18] <ds2> ESC09student: there aren't any distro that is like Angstrom... the closests to desktop stuff is gentoo or ubuntu (probally)
  • [06:37:20] <Russ> ESC09student: a lot of the config stuff matches debian (or Ubuntu) but the destop and gpe
  • [06:37:21] <ESC09student> thanks!
  • [06:38:09] <ESC09student> for a battery update, couldn't you tie batteries into the barrel connector input?
  • [06:38:20] <GregorR> ds2: opkg is extremely similar to dpkg, all the networking configuration etc is done like it's done in Debian (and not Ubuntu until both Angstrom and Debian get network-manager), so I'd say Debian, but :P
  • [06:38:32] <ds2> no, the helper chip on boardd has an internal RTC
  • [06:38:37] <Russ> koen: heh, right next to "no more openmoko phone"
  • [06:38:47] <tomba> good morning. lots of dss discussions during the night, I see =)
  • [06:38:49] <ds2> the battery input for the RTC is not wired up right now
  • [06:39:09] <ESC09student> rev D!
  • [06:39:12] <ds2> GregorR: if you look at it that way, sure
  • [06:39:25] <GregorR> ds2: All things are a matter of perspective :P
  • [06:39:41] <ds2> I think the BB board could be useful if it is available with certain parts not stuffedd
  • [06:40:02] <ds2> the console and JTAGheaders are another 2 things to add to the not stuffed list
  • [06:40:12] * khasi1 (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [06:40:12] <ds2> 0.1" pitched stuff is easy to add afterwards
  • [06:40:14] <raster> koen: arm netbooks .. nothing too new :) but about damn bloody time
  • [06:40:48] <Russ> I think JTAG gets stuffed so that the board is easier to test and diagnose during manufacturing
  • [06:40:51] <Russ> same with console
  • [06:41:56] <ESC09student> I talked to a vendor at ESC, he went from schematic to boards in 7 weeks, very cool. apparently the uC is very easy to wire up to peripherals.
  • [06:42:41] <ds2> Russ: it can be done with spring pin jigs
  • [06:43:00] <Russ> true
  • [06:43:03] <ds2> ESC09student: compare to building up something in 2 weeks using the expansion port
  • [06:43:55] <ESC09student> nice. ummm which port is the expansion port? :/
  • [06:43:58] <geckosenator> ds2: it would be nice to have a battery clip for the rtc
  • [06:45:08] <koen> raster: lately I've been getting mails asking for advice with designing ARM netbooks because my name's on the touchbook website
  • [06:45:19] <raster> hahahaha
  • [06:45:31] * koen would really love an arm netbook with >10 hours battery life
  • [06:45:33] <raster> i'm not surprised
  • [06:45:50] <raster> i want a smartphone...
  • [06:45:57] <raster> wit5h a decent soc
  • [06:46:37] * koen stills wants a hands on experience with mx51 and snapdragon devices running linux
  • [06:46:39] <raster> that you can, without hassle, relfash with linux of your choice
  • [06:46:58] <raster> have qualcomm emerged from their closedness?
  • [06:47:13] <raster> speaking of....
  • [06:47:24] <ESC09student> got it, J3, the one labeled expansion port...
  • [06:49:26] * montamer_ is now known as montamer|away
  • [06:49:35] <raster> koen: u know if haret is happy with simple uImages (as opposed to zImage) ?
  • [06:50:34] <ds2> >10hr is easy if you don't spec the size of the battery!
  • [06:50:52] <raster> it very much also hinges on your screen
  • [06:50:55] * DiBruno (n=blaine@d216-232-231-113.bchsia.telus.net) Quit (Read error: 60 (Operation timed out))
  • [06:50:59] <raster> the lcd backlight munches through power
  • [06:51:08] <ds2> ESC09student: refering to the image on www.beagleboard.org - the expansion port is the pads on the right edge of the board.. it is not stuffed from teh factory
  • [06:51:13] <ds2> indeed
  • [06:51:36] <raster> i suspect on idle systems the backlight is your biggest single problem
  • [06:51:40] <ds2> on my display, the backlight draws 15mA@30V or almost 0.5W by itself
  • [06:51:47] <Russ> raster: the color epaper hasn't caught up yet
  • [06:51:57] <raster> also if you have other subsystems (wifi, 3g, etc.) these can eat quite a bit of power themselves
  • [06:52:22] <raster> ds2: led backlight?
  • [06:52:27] <ds2> raster: yep
  • [06:52:32] <raster> i do say the backlight technology is improving
  • [06:52:35] <raster> aaah ok
  • [06:52:36] <raster> how big?
  • [06:52:39] <raster> 4"?
  • [06:52:40] <ds2> I think it is either 8 or 7 LEDs in series
  • [06:52:44] <ds2> 4.3" WQVGA
  • [06:52:47] <raster> aaha ok
  • [06:52:54] <raster> 3" and below will save on that
  • [06:53:06] <ds2> wonder if EL is less power hungry
  • [06:53:13] <raster> 480x272 - right? "psp lcd"?
  • [06:53:22] <ds2> yep. the PSP LCD, sharp flavor
  • [06:53:26] <raster> oled is something i have been eying
  • [06:53:34] <raster> in theory it should be significantly less
  • [06:53:40] <raster> especially for a "black screen"
  • [06:53:41] <raster> :)
  • [06:53:53] <ds2> yes, except one only has 2 arms and 2 legs :D
  • [06:54:04] <raster> hhaha
  • [06:54:36] <ds2> the drooling at the ESC OLED vendors stops as soon as they mention the prices
  • [06:54:48] <raster> ooooh
  • [06:54:51] <raster> that bad eh?
  • [06:54:55] <ds2> yep
  • [06:54:57] <raster> what are the prices?
  • [06:55:06] <ds2> of course, I wanted something larger like 7"
  • [06:55:07] <raster> like $80+ for a 3" qvga?
  • [06:55:17] <ds2> > $100, IIRC
  • [06:55:20] <raster> as best i saw oled still is fairly lo-res
  • [06:55:26] <raster> oooh. ouch
  • [06:55:32] <ds2> I wanted at least 640x480 pixels to deal with
  • [06:55:53] <ds2> my next fun project is going to be a "pad" (tablet pc)
  • [06:56:10] <geckosenator> I was offered a 7" oled lcd
  • [06:56:16] <geckosenator> I mean, for $500
  • [06:56:18] <geckosenator> I didn't take it
  • [06:56:20] <ds2> hahah
  • [06:56:24] <geckosenator> heh
  • [06:56:26] <raster> oled lcd
  • [06:56:30] <raster> thats a .. nice combo
  • [06:56:30] <raster> :)
  • [06:56:31] <geckosenator> but it would be nice, it uses much less power
  • [06:56:41] <Russ> maybe they should have given some away as raffle prizes
  • [06:56:47] <geckosenator> yeah
  • [06:56:51] <Russ> were there any big raffle prizes at ESC?
  • [06:56:59] <ds2> Russ: yeah, a scope
  • [06:57:09] <ds2> and some scholarships for UCSC extension
  • [06:57:17] <Russ> linux world has the big prizes
  • [06:57:56] <geckosenator> an oscope?
  • [06:57:57] <koen> raster: the haret I used (like 4-5 years ago) just jumped to the memory location and exec'd it
  • [06:58:02] <ds2> yes, an Oscope
  • [06:58:18] * k3v|n (n=ikevin@adsl-69-108-110-199.dsl.irvnca.pacbell.net) has joined #beagle
  • [06:58:18] <geckosenator> I would like a small portable oscope that plugs in the beagle over usb
  • [06:58:21] <koen> raster: a zImage works when used like that, uImages don't because of the uboot header
  • [06:58:28] * mckoan|away is now known as mckoan
  • [06:58:37] <geckosenator> I just got a portable butane soldering iron today
  • [06:58:40] <ds2> I wish there were more cheap 800x480 LCDs
  • [06:58:46] <geckosenator> what is cheap?
  • [06:58:51] <ds2> <$50
  • [06:58:53] <k3v|n> hi
  • [06:58:54] <geckosenator> I got one for $70
  • [06:59:16] <ds2> I really want the one in the N810 but they seem to go for around $90-$100
  • [06:59:18] <Russ> koen: someone needs to modify u-boot so that when it reads kernels from NAND, it only reads the kernel,not the whole 4MB partition
  • [06:59:19] <geckosenator> ds2: are you going to make it in quantities?
  • [06:59:29] <geckosenator> ds2: or just make one and post the plans?
  • [06:59:35] <koen> Russ: ah, that's why it's so slow
  • [06:59:36] <ds2> geckosenator: make what in quantities?
  • [06:59:40] * mckoan is now known as mckoan|away
  • [07:00:11] <geckosenator> ds2: the netbook with cheap lcds in it
  • [07:00:12] <Russ> koen: thats my thought, maybe u-boot lies and it really does just read the kernel
  • [07:00:41] <ds2> geckosenator: I am thinking of making a "Build your own netbook/nettop tablet kit"
  • [07:00:44] <geckosenator> ds2: what about buying eeepcs and replace the hardware with a beagle board?
  • [07:01:00] * bar_foo (n=falseadd@ufc.paivola.fi) has joined #beagle
  • [07:01:03] <geckosenator> keep the screen keyboard, batteries connectors..
  • [07:01:13] <ds2> geckosenator: why? that makes no sense.. the eeePC KB + "mice" sucks
  • [07:01:21] <geckosenator> I haven't tried one
  • [07:01:27] <ds2> note all the typos I am making... i am typing on one of them
  • [07:01:30] <raster> koen: crapper! thats probably the problem
  • [07:01:33] <geckosenator> ds2: heh
  • [07:01:34] <raster> i have a uImage
  • [07:01:36] <raster> poopie
  • [07:01:45] * eFfeM (n=frans@195-241-226-180.ip.telfort.nl) has joined #beagle
  • [07:01:54] <raster> koen: tried u-boot with haret?
  • [07:02:13] <ds2> geckosenator: I prefer more of a pure tablet anyways
  • [07:02:14] <koen> raster: I think u-boot was still ppcboot when I last used haret :)
  • [07:02:14] <eFfeM> morning everyone
  • [07:02:18] <koen> hey eFfeM
  • [07:02:23] <eFfeM> koen do you ever sleep ?
  • [07:02:32] <raster> koen: HAHAHAHAH
  • [07:02:36] <GregorR> ds2: I've never understood who can type on these keyboards that are between "thumb-type" and "real-type" size.
  • [07:02:44] <GregorR> ds2: I find the Eee's keyboard completely insufferable.
  • [07:02:46] <geckosenator> GregorR: small fingers?
  • [07:03:05] <geckosenator> I have a keyboard that is 95% normal width, it's not too bad
  • [07:03:08] <eFfeM> GregorR: they are made for taiwanese fingers
  • [07:03:12] <ds2> GergorR: there are some that are quite usable... the eeePC just isn't it... the Lifebook P series is much better
  • [07:03:20] <eFfeM> i can work on it but that's it
  • [07:03:33] <geckosenator> ds2: maybe you can make netbooks without lcds
  • [07:03:42] <geckosenator> ds2: have you ever seen a blind person surfing the web?
  • [07:03:53] <raster> GregorR: try type on a blackberry then
  • [07:03:53] <raster> :)
  • [07:04:02] <raster> you'll be drooling after an eee kbd :)
  • [07:04:07] <ds2> geckosenator: actually yes... I also know of a blind Linux developer doing a lot of ARM stuff too
  • [07:04:15] <geckosenator> ds2: well my point is you can do it
  • [07:04:27] <geckosenator> ds2: so all you need is headphones :-P
  • [07:04:31] <ds2> heh
  • [07:04:36] <koen> ds2: there are actually to arm kernel hackers called nico and both are blind :)
  • [07:04:44] <koen> s/to/two/
  • [07:05:02] <ds2> geckosenator: what do you think of a tablet with a eink on one side and LCD on the other?
  • [07:05:10] <ds2> koen: 2? I know of one
  • [07:05:18] * k3v|n is now known as K3v|n
  • [07:05:20] <geckosenator> ds2: that would be cool too
  • [07:05:31] <geckosenator> ds2: but maybe it would be best to be able to view both screens at the same time
  • [07:05:35] <koen> ds2: pitre and mcguire
  • [07:05:36] <ds2> geckosenator: you think it'd be useful to have it on hte back?
  • [07:05:47] * K3v|n is now known as K3v1n
  • [07:05:48] <ds2> koen: oh... didn't know about mcguire
  • [07:06:06] <geckosenator> I would like to learn how they do it
  • [07:06:27] <ds2> geckosenator: what would the eink display be for, IYO
  • [07:06:49] * K3v1n is now known as ik3vin
  • [07:06:52] <geckosenator> ds2: another framebuffer..
  • [07:06:56] * ik3vin is now known as ik3v1n
  • [07:06:56] <geckosenator> ds2: how is it driven? usb?
  • [07:07:09] <ds2> geckosenator: SPI
  • [07:07:17] <geckosenator> well if you make it a linux framebuffer
  • [07:07:22] <ds2> geckosenator: I mean what function? what would it show?
  • [07:07:25] <geckosenator> then you could use it for anything
  • [07:07:35] <geckosenator> irc?
  • [07:07:39] <geckosenator> they just do black and white?
  • [07:07:44] * ik3v1n is now known as ikev
  • [07:07:45] <ds2> update is too slow for irc
  • [07:07:49] <geckosenator> really?
  • [07:07:57] * ikev is now known as i
  • [07:07:58] <ds2> updates are on the order of 0.500sec or so
  • [07:08:02] <eFfeM> koen, apparently otg host is not working on .29, do you know anything about that? is there a patch ?
  • [07:08:03] <geckosenator> that would work
  • [07:08:11] <ds2> it takes time to muck with the stuff
  • [07:08:12] <geckosenator> ds2: I would use it as another console
  • [07:08:18] * i is now known as ikev
  • [07:08:37] <GregorR> 0.5sec would be fine for IRC if you made a client that didn't work in the standard scrollback mode.
  • [07:08:47] <Russ> I like the story about nico working on the LCD driver and asking his wife, how about now? "its upside down", etc
  • [07:09:05] <geckosenator> GregorR: it just wraps to the top?
  • [07:09:07] <ds2> Russ: hahah
  • [07:09:40] <GregorR> geckosenator: Idonno, something like that, the point is that there's nothing about IRC that requires a good refresh rate, it's just how the clients are written *shrugs*
  • [07:10:08] <ds2> scrolling wouldbe painful. think some guy at LPC talked about using eink and one issue was stock software SUCKS in that it triesto do too many redraws
  • [07:10:11] <ikev> hi all, is there a way to trade B6 for C?
  • [07:10:28] <eFfeM> ikev: yes: sell B6 on ebay, buy C2 from digikey
  • [07:10:31] <Russ> ds2: I think you have to replace scrolling with flipping (pgup/pgdown)
  • [07:10:41] <koen> eFfeM: try http://www.sakoman.net/cgi-bin/gitweb.cgi?p=openembedded.git;a=blob_plain;f=packages/linux/linux-omap3/musb-otg.patch;h=612183e51bab6bd8f8f1da601e3b9d8fa24c5552;hb=7a7dd16bb670a437dc1b428d3e3828eaf448decf
  • [07:10:43] <ds2> ikev: yep. the upgrade fee is a mere $149 at digikey
  • [07:10:52] <ds2> Russ: that's alot of work
  • [07:10:58] <eFfeM> ds2, lol
  • [07:11:07] <Russ> ds2: just hit pgdown in firefox :)
  • [07:11:16] <ikev> oh man
  • [07:11:24] <ds2> Russ: firefox was one of the worse offenders
  • [07:11:26] <eFfeM> koen, thanks, will give it a try
  • [07:11:29] <koen> ds2: the problem with eink is that most (if not all) optimized drawing methods are patented
  • [07:11:39] <GregorR> Bleh, lame :(
  • [07:11:44] <geckosenator> er, they patented drawing methods?
  • [07:11:48] <ikev> i bought B6 like 3 weeks before C is released LOL
  • [07:11:48] <ds2> koen: that is the least of the problems
  • [07:11:51] <ikev> so lame
  • [07:11:51] <GregorR> geckosenator: GOD BLESS AMERICA
  • [07:11:57] <eFfeM> i still have some issues with usb
  • [07:12:06] <ds2> koen: actually, I was thinking of a eink like display that is available in small quantities w/o NREs
  • [07:12:17] <ds2> basically a zero powerto retain display
  • [07:12:25] <koen> geckosenator: the method to track the stylus nicely when drawing is patented
  • [07:12:48] <geckosenator> a lot of things are patented
  • [07:12:50] <koen> ds2: the beagle-eink project looks promising, but $3k to get a display.....
  • [07:13:02] <geckosenator> in fact, they frequently issue patents for things that are already patented
  • [07:13:05] <ds2> koen: the one I am thinking of is $15 for a display
  • [07:13:11] <geckosenator> ds2: link?
  • [07:13:28] <ds2> geckosenator: it is somewhere in my pile of stuff I got at ESC...
  • [07:14:10] <raster> eink is great... until you want color... and want to update it...
  • [07:14:11] <raster> :)
  • [07:14:19] <raster> tho i believe color just came out
  • [07:14:24] <ds2> raster: that's why I want it as a secondary display
  • [07:14:28] * ikev (n=ikevin@adsl-69-108-110-199.dsl.irvnca.pacbell.net) Quit ()
  • [07:14:29] <raster> still dont know of epaper tat has "good refresh"
  • [07:14:35] <raster> ds2: thats a good idea
  • [07:14:44] * ikev (n=ikevin@adsl-69-108-110-199.dsl.irvnca.pacbell.net) has joined #beagle
  • [07:14:47] <koen> raster: AIUI colour eink suffers from burn-in
  • [07:14:47] <raster> ds2: do u know what makes it so pricey?
  • [07:14:51] <raster> is it failure rate?
  • [07:14:52] <geckosenator> I would like to use both displays at the same time though
  • [07:14:57] <koen> raster: at least the ones from 2 years ago
  • [07:14:58] <ds2> raster: NREs
  • [07:15:00] <raster> koen: poopie!
  • [07:15:06] <ds2> idiotic designs
  • [07:15:06] <raster> ds2: nres?
  • [07:15:13] * ikev (n=ikevin@adsl-69-108-110-199.dsl.irvnca.pacbell.net) Quit (Client Quit)
  • [07:15:14] * koen is a few years behind on the eink curve
  • [07:15:23] <ds2> NREs == Non Reoccuring Engineering costs
  • [07:15:27] <raster> aaah
  • [07:15:29] <raster> those
  • [07:15:52] <raster> until eink is at vast quantity levels for a screen...
  • [07:15:58] <raster> i was wondering if it was yeild
  • [07:16:00] <ds2> while the BB may not be a money looser on a per unit basis, I doubt it is that way when NREs are factored in
  • [07:16:04] <raster> ie "too many dead pixels"
  • [07:16:25] <ds2> it can't be that bad... the esquire cover for example was cheap
  • [07:16:39] <geckosenator> I don't think they are getting rich off the BB though
  • [07:17:01] <raster> ds2: i had always mulled - if it was wquality/yield problems
  • [07:17:08] <raster> just make a cheap-arse e-ink
  • [07:17:15] <raster> and put it ont he BACK of a cellphone
  • [07:17:21] <raster> if u have dead pixels - dont worry
  • [07:17:25] <ds2> not getting rich but it isn't as selfcontained as one would think
  • [07:17:26] <raster> its a standard defect in the "Case'
  • [07:17:35] <raster> but u'd have a software-modifiable case design
  • [07:17:42] * _gm__ (n=_gm_@190.173.65.250) Quit (Read error: 110 (Connection timed out))
  • [07:17:44] <ds2> raster: the $3K price is for the dev kit... that includes a bunch of stuff that isn't really needed
  • [07:17:50] <raster> you dont need a case design to update often so refresh is zero issue
  • [07:18:02] <ds2> yep
  • [07:18:15] <raster> and it coudl double to read books on your phone
  • [07:18:18] <raster> so when idle
  • [07:18:22] <raster> read a book on the back of the phone
  • [07:18:30] <raster> when done itgoes back to its case design
  • [07:18:34] <raster> problem is cost
  • [07:18:35] <eFfeM> this eink device looks pretty cool: http://www.plasticlogic.com/product.html
  • [07:18:38] <raster> if u could do it for $50
  • [07:18:41] <raster> it might be worth it
  • [07:18:48] <raster> as an optional "extra back cover"
  • [07:18:56] <ds2> raster: $50, sure... just buy probally 10K or 100K units at once
  • [07:19:30] <geckosenator> I would be happy if my only display were eink
  • [07:19:46] <Russ> I think japan has some cell phones now where the cover is B/W e-ink so that you can download designs for it
  • [07:19:56] * _gm__ (n=_gm_@190.173.64.90) has joined #beagle
  • [07:20:01] <geckosenator> Russ: like the display isn't flat?
  • [07:20:04] <eFfeM> i'll be damned, my debian beagle does recognize the mouse but not the kbd, although linux recognizes the kbd (boot log says it is found) anyone a clue how to get X to recognize it ?
  • [07:20:10] <raster> ds2: i was thinking.. you make a prohne to start with a few extra connectors wired to the back
  • [07:20:14] <raster> but not plugged in
  • [07:20:20] <raster> just the normal cheap plastic back cover
  • [07:20:28] <Russ> geckosenator: right
  • [07:20:29] <raster> but with unused spi or i2c lines
  • [07:20:36] <raster> not sure what u'd wire ti to
  • [07:20:36] <geckosenator> Russ: that's pretty crazy
  • [07:20:44] <raster> then you can buy the optional "ebook back"
  • [07:20:53] <raster> and it finally connects those lines up
  • [07:20:56] <raster> might be doable
  • [07:21:04] <ds2> raster: right... it'll be a real skin but you still need to pay for NREs
  • [07:21:11] <raster> yeah
  • [07:21:14] <raster> thats the problem
  • [07:21:17] <Russ> http://www.cameraphonesplaza.com/wp-content/uploads/2008/07/e-ink-gadgets-2.jpg
  • [07:21:21] <raster> but if you have sold 50k of your phone
  • [07:21:27] <raster> you know there are 50k units in the wild
  • [07:21:34] <raster> and for $50 they become real ebook readers
  • [07:21:45] <raster> that shoudl net you 10-20k sales...
  • [07:21:46] <Russ> http://sparkingtech.com/gadgets/hitachi-w61h/
  • [07:21:47] <ds2> I am guessing NREs run around $10K-$50K
  • [07:22:09] <raster> Russ: damn the japanese! they stole my idea!
  • [07:22:13] <raster> kddi did it again
  • [07:22:17] <raster> damn them and their cool designers
  • [07:23:18] <raster> in asia blinging your phone
  • [07:23:29] <raster> (adding sparkley dimaonds, patterns, shapes) is "in"
  • [07:23:34] <raster> half the people do it
  • [07:23:40] <raster> theres entire stores devoted to doing it
  • [07:23:52] <raster> customising a phone is a good business there
  • [07:24:09] <raster> i see kddi cottoned on
  • [07:24:12] <ds2> indeed
  • [07:24:26] <ds2> raster: I was thinking of something more drastic
  • [07:24:38] <raster> not just a flat panel section
  • [07:24:43] <raster> but curved around the whole body?
  • [07:24:43] <ds2> entire case made of eink
  • [07:24:46] <raster> yeah
  • [07:24:47] <ds2> yep
  • [07:24:50] <raster> i wanted that too
  • [07:24:55] <raster> but i had no idea if it was feasible
  • [07:24:58] <ds2> no labeling, just a eink "block"
  • [07:25:01] <raster> yup
  • [07:25:05] <raster> i'd go for that
  • [07:25:07] <ds2> it is feasible
  • [07:25:14] <raster> well a good lcd for the 'Fast updates"
  • [07:25:22] <geckosenator> can you print eink on a printer?
  • [07:25:22] <raster> but everythgin that isnt lcd - is eink
  • [07:25:27] <ds2> normally "black" but goes transparent to show the display when the phone is active
  • [07:25:36] <ds2> the LCD would be below the eink
  • [07:25:36] <raster> ooooh
  • [07:25:43] <raster> hmmm
  • [07:25:48] <raster> i am not sure thats doable
  • [07:25:50] <koen> can eink become transparent?
  • [07:25:53] <ds2> phone rings, or the user wakes it up... it goes transparent
  • [07:25:57] <ds2> koen: yep
  • [07:25:58] <geckosenator> can you make eink tatoos for your body?
  • [07:26:03] <raster> as eink relies on solid min-balls with one side black, one white
  • [07:26:12] <raster> u'd need one side blak, on a translucent ball
  • [07:26:13] <ds2> geckosenator: it isn't that flexible
  • [07:26:20] <geckosenator> ds2: maybe someday :-P
  • [07:26:23] <raster> but that'd not work as no matter how its rotated it'd block the underneath
  • [07:26:30] <ds2> well, the competing technology I saw is possible
  • [07:26:37] <raster> aaaah ok
  • [07:26:38] <raster> cool
  • [07:26:43] <ds2> and what's more the entire case should be capactive sensitive
  • [07:26:46] <raster> i'd agre with ui there
  • [07:26:48] <raster> that'd be perfect
  • [07:27:01] <ds2> and maybe hermetically sealed so it is waterproof
  • [07:27:04] <raster> if it is simply a mater of black vs "show whats underneath" reliably
  • [07:27:16] <raster> (much like lcd - but permanent withotu needign power)
  • [07:27:26] <raster> agreed
  • [07:27:32] <raster> tho waterproof might be hard
  • [07:27:44] <ds2> why?
  • [07:27:45] <raster> but the rest i'd definitely think would make for an awesome device
  • [07:27:47] <geckosenator> I'm still going with the clear drybags for that..
  • [07:27:55] <raster> well still have battery replacable
  • [07:28:04] <raster> have any way to charge it or connect it to stuff
  • [07:28:06] <ds2> have it screw together with an O-ring gasket
  • [07:28:07] * koen loves inductive charging
  • [07:28:15] <raster> u could use conductiing cols for chargin
  • [07:28:16] <ds2> inductive charging; Bluetooth syncing
  • [07:28:25] <raster> err inductive
  • [07:28:26] <raster> yeah
  • [07:28:30] <ds2> it'll be a "live" brick
  • [07:28:37] <geckosenator> I would just get a waterproof power plug
  • [07:28:42] <geckosenator> since I care about efficiency
  • [07:28:45] <ds2> i should just prototype such a unit
  • [07:28:50] <raster> what about speakers and mic?
  • [07:28:56] * _AV500_ "charges" his pots and pans inductively all the time
  • [07:29:08] <ds2> no speakers or mic.. bt handsfree kit :D
  • [07:29:16] <raster> hmmmm
  • [07:29:25] <ds2> _AV500_: you like eddy currents? :D
  • [07:29:25] <geckosenator> _AV500_: you have an inductive stove?
  • [07:29:38] <_AV500_> yes and yes
  • [07:29:47] <geckosenator> what frequency does it use?
  • [07:30:12] <raster> ds2: seems doable. i think waterproof mic/speaker would be possible tho
  • [07:30:13] * _gm__ (n=_gm_@190.173.64.90) Quit (Read error: 104 (Connection reset by peer))
  • [07:30:13] <_AV500_> no idea :-)
  • [07:30:30] <geckosenator> my stove is resistive
  • [07:30:34] <raster> with good deals around the bay
  • [07:30:38] <_AV500_> pah, old school
  • [07:30:54] <_AV500_> capacitive is all therage
  • [07:34:32] <ds2> raster: find the parts...
  • [07:35:09] <raster> err seal
  • [07:35:12] <raster> i meant
  • [07:35:26] <raster> ds2: dude. find the e-ink that can be an entire case first! :)
  • [07:35:54] <ds2> raster: eink is flexible
  • [07:36:14] <ds2> I think I can build up something cheap
  • [07:36:36] <ds2> just can't do it with the corners using my cheap method
  • [07:37:16] <raster> ds2: maybe something that is more cylindrical?
  • [07:37:23] <raster> ie the fases and sides and back curve
  • [07:37:33] <raster> but the top and bottom are "regular uninked" ?
  • [07:40:37] <ds2> it is a matter of finding out the min. bend radius
  • [07:41:25] <raster> gah!
  • [07:41:31] * raster kicks haret
  • [07:42:14] <raster> u know... i've never checked
  • [07:42:23] <raster> wtf is the exact diff between uImage and zImage?
  • [07:42:49] <ds2> uImage is decompressed by Uboot; zImage is self contained, IIRC
  • [07:43:01] <raster> hmm
  • [07:43:20] <ds2> uImage also has a header for identifying things by U-boot
  • [07:43:58] * raster grumbles
  • [07:44:05] <raster> this is.. not working as planned
  • [07:44:25] <ds2> almost anything besides u-boot will want either a zImage or bzImage
  • [07:44:40] <raster> thats my current dilemma
  • [07:44:48] <raster> i'd happily provide a zimage.. if i could
  • [07:44:56] <raster> but i only habe a binary uimage
  • [07:45:00] <geckosenator> raster: just convert it
  • [07:45:02] <raster> i am not going to get source :(
  • [07:45:02] <geckosenator> that's what I do
  • [07:45:06] <raster> geckosenator: how?
  • [07:45:09] <geckosenator> raster: mkimage
  • [07:45:12] <ds2> Oh... that is nasty
  • [07:45:14] <raster> been googling
  • [07:45:25] <geckosenator> apt-get?
  • [07:45:28] <ds2> raster: any chance you can get the vmlinux file associated with it?
  • [07:45:29] <geckosenator> ds2: it works
  • [07:45:30] <raster> ds2: i know. it's a slow diplomatic process to get them to come around
  • [07:45:49] <raster> decreally.. thanks! much obliged. u googled for a bit but didnt find anything
  • [07:45:59] <raster> if i do this
  • [07:46:09] <raster> i will have a fully working linux kernel for my i780
  • [07:46:19] <raster> from there up i have src
  • [07:46:30] <raster> foir telephony stack (userspace part of 3g driver) etc. etc.
  • [07:46:34] <geckosenator> raster: uboot-mkimage
  • [07:46:43] <geckosenator> raster: mkimage -A arm -O linux -T kernel -C none -a 0x80008000 -e 0x80008000 -n "Linux" -d arch/arm/boot/Image uImage
  • [07:47:21] <ds2> geckosenator: I think he has a uImage but needs a zImage
  • [07:47:30] <geckosenator> oh
  • [07:47:35] <ds2> IIRC, vmlinux is the common format between them
  • [07:48:13] <ds2> raster: tried this - load it with u-boot in another board, then use u-boot to do a memory save
  • [07:48:45] * _gm__ (n=_gm_@190.173.66.55) has joined #beagle
  • [07:49:08] <raster> ds2: unfortunately i dont have any other pxa310 hw... :(
  • [07:49:16] <raster> ooh wait
  • [07:49:18] <raster> just the load bit
  • [07:49:20] <raster> not the boot bit
  • [07:49:35] <raster> geckosenator: yes! i need to go in reverse
  • [07:49:37] <raster> :(
  • [07:49:40] <ds2> try it on a PXA270 board
  • [07:49:50] <raster> ds2: i'll try
  • [07:49:58] <raster> but that seems as goo an idea as any
  • [07:50:19] <raster> oh crap
  • [07:50:24] <ds2> if that fails, hack U-boot to bypass the image type checks
  • [07:50:26] <raster> my pxa270 also doesn thave uboot
  • [07:50:29] <raster> just cocoboot
  • [07:50:30] <ds2> the beagle is probally good for that
  • [07:50:30] <raster> bugger
  • [07:50:31] <raster> hmmm
  • [07:50:42] <ds2> ARGGG... got a Gumstix Verdex?
  • [07:50:43] <raster> yeah
  • [07:50:48] <raster> beagle maybe
  • [07:50:52] <raster> nup
  • [07:50:55] <raster> overao
  • [07:50:57] <raster> no verdex
  • [07:51:03] <ds2> the beagle might need a hacked up u-boot
  • [07:51:08] <ds2> but that is easy to do and recover from
  • [07:51:17] <raster> let me see what i get out of this
  • [07:54:22] <raster> ok
  • [07:55:14] <raster> hmm
  • [07:56:04] <raster> dont seem to have a uboot command to load and decode a uimage
  • [07:56:09] <raster> but not boot it
  • [07:57:18] <ds2> hmmm
  • [07:57:29] <raster> i can lopad a whole binary blob
  • [07:57:31] <raster> that i knew
  • [07:57:45] <raster> ie dump a partition or file from fat to a memory loc
  • [07:57:49] <ds2> if it was on a 270 board, I'd say jtag break point at kernel startt
  • [07:57:57] <raster> (nasty hack for doing splashscreens)
  • [07:58:03] <ds2> but as I say... you can hack U-boot
  • [07:58:04] <raster> yeah
  • [07:58:26] <raster> a uimage tho will have a header.
  • [07:58:37] <raster> after that header should be a "zimage"
  • [07:58:40] <raster> just not.. z'd'
  • [07:58:57] <raster> hmm
  • [07:59:02] <raster> mkimage...
  • [07:59:09] <raster> if u can make a uimage from a zimage...
  • [07:59:23] <ds2> no
  • [07:59:26] <raster> i must be able to reverse it
  • [08:00:08] <raster> should be no different to memroy after uboot has done the load...
  • [08:00:14] <raster> uboots info comes from the same file?
  • [08:00:19] <Russ> I always thought a uImage was just a zImage wrapped in some u-boot header
  • [08:00:24] <raster> so its the same thing - just without using uboot
  • [08:00:46] <ds2> zImage includes a self decompressor
  • [08:00:53] <ds2> uboot does the decompression
  • [08:01:05] <ds2> so you should get something that kind of looks like vmlinux
  • [08:01:41] <raster> thats fine by me
  • [08:01:44] <raster> already uncompressed
  • [08:01:50] <raster> that would be the result of hacking uboot anyway
  • [08:05:36] <Russ> then why does it say, "Uncompressing Linux............................................................................................ done, booting the kernel."
  • [08:05:51] <Russ> and "Data Size: 1419840 Bytes = 1.4 MB"
  • [08:06:05] <Russ> my zImage is 1.4M
  • [08:06:11] <ds2> let me look at the code and I'll retract my statement otherwise :D
  • [08:06:27] <Russ> so the claim of "Image Type: ARM Linux Kernel Image (uncompressed)" seems most definitely fase
  • [08:06:31] <Russ> er, false
  • [08:08:43] * _gm__ (n=_gm_@190.173.66.55) Quit (Read error: 110 (Connection timed out))
  • [08:09:36] <koen> Russ: uboot has its own compression scheme
  • [08:09:54] <koen> Russ: to it is indeed an uncompresses uImage, containing a zImage
  • [08:09:58] <Russ> koen: then it seems rather a waste
  • [08:10:01] <koen> s/d/d/ somewhere
  • [08:10:19] * _gm__ (n=_gm_@190.173.67.79) has joined #beagle
  • [08:10:27] <koen> Russ: if u-boot doesn't set up the cache properly, it's a few seconds slower
  • [08:10:31] <Russ> maybe thats part of why it takes solong
  • [08:10:40] <Russ> a double decompression
  • [08:10:48] <ds2> Russ: I stand corrected, you are riight
  • [08:10:49] <koen> Russ: it took me a while to get that point across to the openmoko engineers 2 years ago
  • [08:11:46] <Russ> nope, its not double compressed
  • [08:11:54] <Russ> mkimage is invoked with -C none
  • [08:13:05] <raster> koen: it took them years to realise that just skipping uboot decomp and letting the kernel do its own was.. much better
  • [08:13:48] <raster> ooh
  • [08:13:53] <raster> uiumage is a bit more complex than i thought
  • [08:14:01] <koen> raster: approx 4 seconds gained on gta01 if you used zImage + -C none
  • [08:14:03] <raster> supports multiple files even
  • [08:15:05] <raster> bugger
  • [08:15:07] <koen> but still no support for automatically executing scripts tacked onto the uImage
  • [08:15:15] <raster> Image Type: ARM Linux Kernel Image (uncompressed)
  • [08:15:15] <raster> Data Size: 1874648 Bytes = 1830.71 kB = 1.79 MB
  • [08:15:56] * koen reads about openmoko on slashdot
  • [08:16:24] <raster> koen: news to the public
  • [08:16:33] <raster> been lurking amongst thopse who knew for a few weeks
  • [08:16:49] <raster> not totally surprised
  • [08:16:57] * koen wasn't surprised at all
  • [08:16:58] <raster> sean found a new shiny bauble to play with
  • [08:17:24] <raster> oooh wait
  • [08:18:51] <raster> ooooh
  • [08:18:52] <raster> ok
  • [08:19:05] <raster> bugger
  • [08:20:49] * pcgeil (n=steffen@p549E4A47.dip.t-dialin.net) has joined #beagle
  • [08:21:20] <koen> heh
  • [08:21:46] <Russ> wow, I'm down to 18mA@5V (90mW)
  • [08:21:54] <koen> some dutch politician suggests that busdrivers should be allowed to use pepperspray on agressive travellers
  • [08:22:12] <koen> I don't want to be inside a bus where that is going to be used
  • [08:22:36] <koen> Russ: removed the leds?
  • [08:22:55] <Russ> no, I think it has to do with the MUSB regulators
  • [08:23:06] <Russ> I start without a cable plugged in, plug the cable in, then suspend
  • [08:23:22] <Russ> if I do it the other way around, it uses more power than if I never plug a cable in
  • [08:23:36] * dirk (n=dirk@p5B040840.dip0.t-ipconnect.de) Quit ("Konversation terminated!")
  • [08:23:49] <Russ> so I definitely need to get regulators working right
  • [08:24:05] <Russ> but yes, most of that 18mA would be the power LED
  • [08:24:19] <Russ> I haven't gotten my temp controlled iron in the mail yet
  • [08:24:45] <geckosenator> I wish my butane one were temp controlled
  • [08:24:57] <geckosenator> but maybe it's not possible
  • [08:25:09] <Russ> geckosenator: not unless it is very expensive
  • [08:25:18] <geckosenator> well I have an electric one that is
  • [08:25:37] <geckosenator> it's also very powerful, it heats up in 8 seconds
  • [08:26:03] <geckosenator> so it senses loads and can overdrive a lot
  • [08:26:20] <geckosenator> the butane one just has a thermal mass so it can't react very fast
  • [08:26:37] <koen> Russ: you might get David Brownell interested in MUSB regulators
  • [08:26:51] <Russ> koen: do you know if he has a git tree?
  • [08:27:01] <Russ> I want to try some of his stuff, but he has tons of scattered patches
  • [08:27:07] <koen> no idea on that
  • [08:27:25] <koen> I just know he's everywhere I turn when hacking new hardware :)
  • [08:27:45] <Russ> do you have a quick suggestion for a nntp reader that can save mbox files? (eg, for patches)?
  • [08:28:17] <raster> Russ: 90mw where, dowing what what is 90mw?
  • [08:28:27] <Russ> raster: sleep
  • [08:28:28] <koen> I only use thunderbird for nntp, and it sucks
  • [08:28:37] <raster> Russ: where? the bb?
  • [08:28:45] <Russ> yes
  • [08:28:45] <raster> rsalveti: the whole bb? or just soc core?
  • [08:28:53] <raster> and sleep in what way?
  • [08:28:55] <raster> suspend?
  • [08:29:04] <raster> or zero-clokc, tickless idle?
  • [08:29:05] <koen> Russ: have you see http://patchwork.kernel.org/ ?
  • [08:29:06] <Russ> the whole bb
  • [08:29:17] <Russ> poweroff for the chip
  • [08:29:24] <Russ> but retain for the SDRAM
  • [08:29:28] <koen> s/see/seen/
  • [08:29:36] <Russ> koen: yes, I love patchwork, but they don't have the linux-usb list
  • [08:30:39] <koen> drat
  • [08:31:46] <raster> Russ: ok. so suspend.
  • [08:31:46] <Russ> ok, I'm trying out the gmane nntp archives of the list
  • [08:31:46] <Russ> in evolution
  • [08:31:59] <raster> Russ: do u know whats eating most of the 90mw?
  • [08:32:29] <koen> powerled on the 5V rail
  • [08:33:13] <Russ> raster: the TWL4030, the 26MHz oscillator (~15mW), the power LED (~50mW), the 232 converters (3.3mW), the TFP410 (1.6mW)
  • [08:33:22] <Russ> and of course the SDRAM
  • [08:33:49] <Russ> thats at least what I get from looking at the datasheets
  • [08:34:21] <geckosenator> can the oscillator stop?
  • [08:34:21] <raster> hmmm
  • [08:34:22] <raster> ok
  • [08:34:30] <raster> the uimage i have is uncimpressed for uboot
  • [08:34:43] <raster> that means its really a zimage inside a uimage wrapper
  • [08:34:48] <Russ> there are two models, one drops the power by half, the other drops it into the microamps range
  • [08:34:49] <raster> so all i need is to snarf away the uboot header
  • [08:34:52] <raster> erruimage
  • [08:34:54] <koen> Russ: http://www.gmane.org/export.php
  • [08:34:55] <Russ> the first model is what is on the beagle
  • [08:34:58] <raster> now.. how big is this header
  • [08:35:02] <koen> Russ: gmane allows mbox exports :)
  • [08:35:08] <Russ> and the stop pin is tied to a voltage rail
  • [08:35:29] <raster> Russ: hmm so actually.. if you reall yhad a suspended system we'd be @ 15mw
  • [08:35:30] <geckosenator> Russ: is there a resistor?
  • [08:35:36] <raster> assuming we didnt find any lower power modes for the 4030
  • [08:36:01] <geckosenator> how long does it take to wake from this mode?
  • [08:36:50] <Russ> geckosenator: nope
  • [08:36:54] <raster> oooh
  • [08:36:55] <Russ> geckosenator: dunno how long it takes
  • [08:36:56] <raster> this is too easy
  • [08:37:35] <Russ> raster: I want to tie the power led to the 3.3V rail and then tie the 3.3V LDO enable to a GPIO of the TWL4030 or the OMAP
  • [08:38:08] <geckosenator> that will allow it to kill itself?
  • [08:38:19] <Russ> ?
  • [08:38:35] <geckosenator> you lose the 3.3v rail and the omap turns off right?
  • [08:38:39] <raster> 128 bytes of uimage header
  • [08:38:43] <raster> walk in the park
  • [08:38:54] <raster> it even was a beatiful multiple of lines in od
  • [08:38:54] <Russ> geckosenator: the 3.3V rail only powers the TFP410 and the RS232 level converters
  • [08:39:07] <raster> didnt even need to finish reading the src for mkimage
  • [08:39:07] <raster> :)
  • [08:40:01] <raster> Russ: have u tried zero-clock and tickless kernels
  • [08:40:16] <raster> ie ti's uber-cool power management that also puts ram into self-refresh when idle?
  • [08:40:21] <raster> so u never need to suspend
  • [08:40:25] <raster> simply keep the system idle
  • [08:40:32] <Russ> I have been using those, but I really don't care about power usage while running
  • [08:40:54] * koen spots approx 25 patches to make musb behave for 2.6.29
  • [08:40:55] <Russ> my project is to reduce power consumption by cutting SDRAM sleep consumption in half
  • [08:41:08] <Russ> right now, that would get lost in the noise and it would look quite unimpressive
  • [08:41:48] <raster> Russ: sure..
  • [08:41:57] <raster> i have alwasy been fascinated by omap's zeor-clock stuff
  • [08:42:01] <Russ> 480uA@1.8V vs 200uA@1.8V
  • [08:42:02] <raster> as thats just super-waesome
  • [08:42:07] <raster> being able to avoid suspending.. ever
  • [08:42:21] <Russ> if I do enable all that stuff, I have to mash the keyboard to get it to wake up
  • [08:42:30] * vsr (i=7aa60de8@gateway/web/ajax/mibbit.com/x-cc0d63a74e9395cd) has joined #beagle
  • [08:42:34] <geckosenator> 480uA si low enough
  • [08:42:43] <geckosenator> it's the same as my battery leakage
  • [08:43:33] <Russ> geckosenator: thats how much power the SDRAM uses while in self refresh mode
  • [08:43:46] * _gm___ (n=_gm_@190.173.66.226) has joined #beagle
  • [08:43:47] <Russ> geckosenator: it uses 200uA if you disable 7/8th of the array
  • [08:43:54] <geckosenator> do you lose the ram?
  • [08:44:08] <Russ> with 480uA, you keep it all
  • [08:44:14] <geckosenator> well that's what I need
  • [08:44:23] * _gm__ (n=_gm_@190.173.67.79) Quit (Connection timed out)
  • [08:44:23] <geckosenator> maybe you could wake up the os and tell it to swap more out
  • [08:44:29] <geckosenator> so you can disable it
  • [08:44:50] <Russ> thats kinda the idea, except to do all that before sleeping
  • [08:44:54] <Russ> and without the swap
  • [08:45:04] <geckosenator> well.. you might need swap
  • [08:45:11] <geckosenator> unless you just don't use much memory
  • [08:45:33] <geckosenator> 7/8th of 256 gives you like 32megs
  • [08:45:49] <Russ> the usage model would be for a system that needs lots of RAM for caches and file backed pages
  • [08:46:08] <Russ> and then compcache (now ramzswap) can be used to compress down the remaining RAM
  • [08:46:10] <geckosenator> you could put a small solar cell on the board to completely make up for this so it wouldn't drain the batteries
  • [08:46:29] <geckosenator> well if you have swap space it's much easier
  • [08:46:32] * lcuk (i=lcuk@cpc3-oldh7-0-0-cust590.manc.cable.ntl.com) has joined #beagle
  • [08:46:38] <geckosenator> can't you suspend to flash too?
  • [08:46:50] <geckosenator> copy all the ram, then you can completely shut down
  • [08:47:03] <Russ> you can do that too, but that'd be slow
  • [08:47:15] <koen> sloooooooooooooooooooow even
  • [08:47:16] <Russ> and would require a u-boot start and a kernel boot
  • [08:47:22] <geckosenator> yes
  • [08:47:45] <raster> yes.. yes.. YES!!!!!
  • [08:47:48] <raster> muhahah
  • [08:47:54] <geckosenator> well it might take a few secons to ccompress your ram and copy it
  • [08:50:00] <Russ> and it would take so much power to do so, that it might be a wash
  • [08:55:01] <raster> bugger
  • [08:55:07] <raster> now i have a zimage...
  • [08:55:08] <raster> nada
  • [08:55:09] <raster> poo
  • [08:57:34] <raster> hmm
  • [08:58:05] <raster> i think my ramaddr was wrong
  • [08:58:36] <Moshy> Hey all :)
  • [08:58:49] * montamer|away (n=vijay@203.199.213.3) Quit ()
  • [08:58:49] * montamer (n=vijay@203.199.213.3) Quit (Remote closed the connection)
  • [09:01:34] * mib_0w7cjd (i=5601b2b8@gateway/web/ajax/mibbit.com/x-246829cef7c56cd3) has joined #beagle
  • [09:03:00] <mib_0w7cjd> 1
  • [09:03:09] * vsr (i=7aa60de8@gateway/web/ajax/mibbit.com/x-cc0d63a74e9395cd) Quit ("http://www.mibbit.com ajax IRC Client")
  • [09:03:11] <raster> poo
  • [09:03:15] <raster> i thiink i have a zimage
  • [09:03:18] <raster> but its not working
  • [09:03:23] <raster> most poopie
  • [09:03:30] * mib_0w7cjd (i=5601b2b8@gateway/web/ajax/mibbit.com/x-246829cef7c56cd3) Quit (Client Quit)
  • [09:03:46] <Russ> it needs to be copied to the exact right address
  • [09:03:57] <Russ> and caches need to be off before running it
  • [09:03:58] * vsr (i=7aa60de8@gateway/web/ajax/mibbit.com/x-61080407f51c7b52) has joined #beagle
  • [09:05:07] <Moshy> Guys I'm an ULTRA newbie in here and I have some questions :)
  • [09:05:36] <Moshy> 1. I've got running Angstrom on the Beagle board
  • [09:05:57] <Moshy> but my Mouse and the Keyboard aren't working (goosh what a newb) :D
  • [09:06:12] <Moshy> I spent hours reading about this issue an all i found
  • [09:06:27] <Moshy> is that the boards OTG usb must be in host mode
  • [09:06:53] <Moshy> (I've got a self powered USB hub (
  • [09:07:11] <Moshy> (Edimax EU-HB4S)
  • [09:07:18] <Russ> which rev board?
  • [09:07:23] <Moshy> B7 :)
  • [09:07:54] <Moshy> It was too late when i saw that the rev C is coming in like 3 days :D
  • [09:07:59] <Moshy> :(
  • [09:08:24] <Russ> ok
  • [09:08:33] <Russ> do you have a OTG cable with a mini-A plug?
  • [09:08:45] <Moshy> well thats the problem :)
  • [09:08:56] <Russ> it has a B on the end that plugs into the beagle?
  • [09:08:58] <Moshy> i was not able to find A plug so i'm using B
  • [09:09:03] <Moshy> with shorted 5 and 4
  • [09:09:04] <Moshy> on the beagle
  • [09:09:06] <Russ> ah
  • [09:09:13] <Russ> I don't know much about the shorting
  • [09:09:21] <Moshy> well
  • [09:09:28] <Russ> do you have everything plugged in when you power on?
  • [09:09:34] <Moshy> yup :)
  • [09:09:52] <Moshy> well i'll go around the city and i'll look for mini a
  • [09:09:59] <Moshy> to mini B
  • [09:10:10] <Moshy> (the hub've got a mini B)
  • [09:10:22] <Russ> I think you should be ok with the shorted pins
  • [09:10:34] <Russ> do kernel messages or lsusb tell you anything interesting?
  • [09:10:44] <raster> hmm
  • [09:10:49] <raster> console=tty - right?
  • [09:10:51] <Russ> and if the hub has a self/bus switch somewhere, is it on self?
  • [09:10:54] <Moshy> yup :)
  • [09:11:06] <Moshy> i've got console :)
  • [09:11:28] <Russ> does the light on the mouse turn on?
  • [09:11:36] <Moshy> yes :)
  • [09:11:46] <Moshy> but nothing from the keyboard :)
  • [09:11:52] <Moshy> (sorry for the :) ... )
  • [09:14:34] * martin_filip (i=d9d084a6@gateway/web/ajax/mibbit.com/x-8cfa16b53ad33007) has joined #beagle
  • [09:14:49] * bar_foo (n=falseadd@ufc.paivola.fi) has left #beagle
  • [09:14:56] <Moshy> well'll i'll be back later an i'll check some more things on the angstrom
  • [09:15:06] <Moshy> Thanks :)
  • [09:19:26] <vsr> unable to build beagle angstrom image. e-places seems to be broken
  • [09:19:58] <raster> hmm
  • [09:20:17] <raster> any reason a kernel wouldnt give me a contsole on the fb... even if i have console=tty0
  • [09:20:26] <raster> the kernel seems to know about the fb as it slaps up a logo
  • [09:20:41] <Russ> its fun watching the back and forth with dbrownell and sshtylyov
  • [09:20:57] <Russ> I still can't tell if they are joking, or if they really hate each other
  • [09:21:30] <Russ> raster: you don't have the kernel option for virtual consoles?
  • [09:21:47] <raster> just standard stuff
  • [09:21:58] <raster> root=/dev/mmcblk0p2 rootfstype=ext2 rw mem=128M console=tty0
  • [09:22:09] <Russ> I mean in your .config
  • [09:22:26] <raster> this is my problem
  • [09:22:31] <raster> i haveth no kernel src
  • [09:22:39] <Russ> good times
  • [09:22:40] <raster> i've just binary munged a uimage into a zimage
  • [09:22:49] <Russ> is there a /proc/config.gz after you boot?
  • [09:22:55] <Russ> or maybe a config.ko kernel module?
  • [09:22:59] <raster> dont get that far
  • [09:23:04] <raster> kernel seems to do somehting
  • [09:23:09] <raster> lcd does some resertting
  • [09:23:12] <raster> a logo pops up
  • [09:23:14] <raster> then nada
  • [09:23:29] <raster> the usb port doesnt seem to bring up any usb gadgets
  • [09:23:39] <raster> so i'm pretty stumped from there
  • [09:23:41] <raster> zero output
  • [09:23:45] <Russ> maybe add an init=/bin/sh?
  • [09:23:49] <raster> dont even know if it moutns the fs
  • [09:24:16] <Russ> it could be compiled with quiet or something
  • [09:24:33] <raster> any way to force noisiness?
  • [09:24:37] <Russ> hmm
  • [09:24:37] <raster> debug=something
  • [09:24:47] * uberfry (n=spinl0ck@vodsl-10890.vo.lu) has joined #beagle
  • [09:24:56] <raster> noo\
  • [09:24:59] <raster> loglevel
  • [09:25:33] <Russ> ignore_loglevel
  • [09:27:05] <Russ> is it one of the kernels from ESC?
  • [09:28:02] * abitos (n=nixgibts@dslb-084-057-154-013.pools.arcor-ip.net) has joined #beagle
  • [09:28:03] <raster> esc?
  • [09:28:15] <Russ> the conference last week
  • [09:28:19] <raster> oooh no no
  • [09:28:23] <raster> from.. elsewhere
  • [09:28:30] <Russ> so its hot?
  • [09:28:38] <raster> hot as in?
  • [09:28:51] <Russ> stolen or illicit goods
  • [09:29:16] <raster> oh no
  • [09:29:30] <raster> licitly obtained through all the right channels
  • [09:29:48] <Russ> then demand the source dammit
  • [09:30:34] <raster> i asked
  • [09:30:41] <raster> but as i am under nda with said customer
  • [09:30:47] <raster> and they pay me...
  • [09:30:56] <raster> i am not going to demand anything :)
  • [09:31:17] <raster> irrespective of gpl and rights
  • [09:31:19] <Russ> you could try poking the kernel with sysrq over serial
  • [09:31:28] <raster> and that is arguable as they may have furnished me with a binary
  • [09:31:39] <raster> but i am under contract...
  • [09:31:50] <raster> so in gray land
  • [09:31:57] <Russ> so they didn't furnish you, they furnished your company
  • [09:31:57] <raster> problem is serial
  • [09:31:59] <raster> dont have serial
  • [09:32:01] <raster> wish i did
  • [09:32:10] <raster> well actually furnished me
  • [09:32:11] <raster> directly
  • [09:32:18] <raster> *I* am under contract
  • [09:32:28] <raster> independant contractor
  • [09:32:36] <raster> i dont have a company set up for this (yet)
  • [09:32:49] <Russ> you don't even have your own LLC?
  • [09:32:55] <raster> nup
  • [09:32:57] <raster> havent bothered
  • [09:33:08] <raster> i was going to set up an offshore in a tax haven
  • [09:33:15] <Russ> heh
  • [09:33:21] <raster> but wasn't necessary - so i wasnt going to pay the costs
  • [09:33:29] <raster> i'll get to it eventually
  • [09:33:46] <Russ> LLC is so easy and cheap, there isn't a point to not have one
  • [09:33:56] <raster> in .au it's easy too
  • [09:34:09] <raster> i just havent bothered
  • [09:34:39] <raster> in fact in .au if you have > 50% of your business from 1 client you ARE considered an employee
  • [09:34:41] <raster> company or not
  • [09:34:43] <raster> doesnt matter
  • [09:34:49] <Russ> lame
  • [09:34:55] <raster> all the legal entities in the world wont change anything
  • [09:35:09] <Russ> do they count hours or total billed?
  • [09:35:14] <raster> its to stop people setting up these companies for 1 man bands that then just contract to 1 company to avoid tax etc.
  • [09:35:19] <raster> total billed
  • [09:35:28] <Russ> ah, caus I could pay you to sleep
  • [09:35:35] <raster> you could set up some bizarre system where the 1 client has 2 legal entities
  • [09:35:36] <raster> or 3
  • [09:35:42] <raster> and you bill the varieous entities they have
  • [09:35:44] <raster> it gets complex
  • [09:35:57] <raster> the solution is an offshore
  • [09:36:05] <raster> so your company is outside .au jurisdiction
  • [09:36:09] <raster> thus .au law doesnt appluy
  • [09:36:18] <Russ> I like an LLC because it allows you to set up your own benefits structure easily and with out tax disadvantage
  • [09:36:21] <raster> legally nothing stops you from workign for an offshore and residing in.au
  • [09:36:32] <raster> so your company lets say bills $200k/yer for stuff
  • [09:36:35] <Russ> and then you always keep the same benefits no matter where you move/work
  • [09:36:37] <raster> but you pay yoursel $50k
  • [09:36:43] <raster> the other $150k gets parked in the company as profit
  • [09:36:52] <raster> if its in a tax haven - the company pays no company tax
  • [09:37:02] <raster> legally your income is 50k - so pay income tax on that
  • [09:37:15] <raster> the rest is just asset accumulation (liquid asset - cash) in the company.
  • [09:37:25] <Russ> its not really that necessary in the US because with an LLC, any pension plan you have set up for your LLC reduces your tax liability
  • [09:37:34] <raster> the problem then is if you cease to be a resident you are deemed to have disposed of all assets
  • [09:37:37] <raster> your company included
  • [09:37:44] <raster> thus u have to pay tax on the value of those assets
  • [09:37:57] <raster> if u remaint
  • [09:38:08] <raster> sooner or later to make use of the money the company has - u have to transfer it to yourself
  • [09:38:18] <Russ> consider a move to ireland
  • [09:38:39] <raster> also fringe benefits tax nukes the ability for the company to provide cars, houses etc. to you.. without you paying tax on those benefits
  • [09:38:45] <raster> as if it were money
  • [09:39:00] <Russ> thats a given
  • [09:39:03] <raster> actually been considering .ch
  • [09:39:08] <raster> uk is good too
  • [09:39:13] <Russ> but working from home still allows you to deduct a percentage
  • [09:39:14] <raster> all offshore income is tax free
  • [09:39:33] <raster> for the moment the money is being parked anyway
  • [09:39:45] <raster> i need to figure somethng to do with it before end of june
  • [09:40:05] <raster> likely i';ll set up the offshore before then
  • [09:40:39] <raster> take the $ - invest into the company, claim as a business expense - thus nixing the income, and then pay myself some small portion of it that i need
  • [09:46:51] <raster> gaaah
  • [09:46:54] <raster> still no console!
  • [09:47:22] <Russ> what if you make a script on the flash card that does 'dmesg > /dmesg'
  • [09:47:30] <Russ> and then do init=<script>
  • [09:47:54] <raster> i dont even know if it finds root
  • [09:48:06] <Russ> if it did work, it would tell you alot
  • [09:48:12] <raster> true
  • [09:50:47] * flo_lap (n=fuchs@f049172180.adsl.alicedsl.de) has joined #beagle
  • [09:51:26] <raster> ok
  • [09:51:28] <raster> lets see
  • [09:51:30] <raster> wait a bit...
  • [09:51:35] <raster> let it mount and do its dirty foo
  • [09:51:38] <raster> if it does...
  • [09:52:38] <raster> nup
  • [09:52:42] <raster> didnt get as far as init
  • [09:54:33] <Russ> is JTAG an option?
  • [09:54:39] <raster> nup
  • [09:54:43] <raster> no jtag equipment
  • [09:54:48] <Moshy> Hey Its me again :)
  • [09:54:49] <raster> and consumer level device
  • [09:54:56] <raster> havent opened it up even yet
  • [09:55:03] <Moshy> I've found an usb cable
  • [09:55:18] <Moshy> but it says USB A 5pin MINI-B
  • [09:55:36] <Moshy> isn't the mini-B 4 pin?
  • [09:55:59] <Russ> its also 5 pin
  • [09:56:44] <Moshy> will it work? :)
  • [09:56:50] <Russ> ya
  • [09:56:54] <Moshy> i need 5 pin instead of the original 4 pin?
  • [09:57:31] <Moshy> good
  • [09:57:46] <Russ> all mini connectors have "5 pins"
  • [09:57:55] <Russ> on mini b, I don't think the ID pin is populated at all
  • [09:58:42] * mib_6db6sx (i=55510a11@gateway/web/ajax/mibbit.com/x-4404e46e3d54c181) has joined #beagle
  • [09:58:49] * mib_6db6sx (i=55510a11@gateway/web/ajax/mibbit.com/x-4404e46e3d54c181) Quit (Client Quit)
  • [09:59:05] <Moshy> well i've "opened" one mini-b and i dont think it
  • [09:59:11] <Moshy> it's*
  • [09:59:15] * xpg (n=pf@0x55510a11.adsl.cybercity.dk) has joined #Beagle
  • [09:59:26] <Moshy> because it has only 4 wires coming out of it
  • [10:00:12] <Russ> there are 5 spots where pins could be in the connector
  • [10:00:44] <Moshy> yes
  • [10:03:36] <Moshy> http://blogs.forum.nokia.com/blog/kate-alholas-forum-nokia-blog/2008/01/21/usb-on-the-go
  • [10:05:14] <Moshy> I'm going to "patch" my adapter :] brb
  • [10:06:50] * valhalla (n=valhalla@81-174-23-93.dynamic.ngi.it) has joined #beagle
  • [10:13:31] * FuL|OUT is now known as fulgas
  • [10:22:42] * pcgeil (n=steffen@p549E4A47.dip.t-dialin.net) Quit ()
  • [10:25:49] <raster> ok
  • [10:25:51] <raster> enough
  • [10:25:55] <raster> 'real life (tm)
  • [10:36:08] <eFfeM> raster: real life is an illusion caused by a lack of alcohol :-)
  • [10:37:39] <vsr> raster: the latest ssl native changes are causing e to beak
  • [10:37:41] * koen wonders whether the leopardboard is going to use dm357 or dm365 in the near future
  • [10:37:58] * kozak (n=subbu@117.192.15.35) Quit ("Leaving.")
  • [10:46:15] <Russ> koen: something ate my patches
  • [10:47:55] <Russ> the gmame download mbox thing doesn't work so well
  • [10:47:58] <Russ> it eats patches
  • [10:48:48] * shankar (n=shankar@122.172.34.138) has joined #beagle
  • [10:50:08] <koen> Russ: it would have been too easy
  • [10:50:42] <koen> I still don't get the hypocrisy about mailclients from the kernel folks
  • [10:50:42] * shankar (n=shankar@122.172.34.138) Quit (Read error: 104 (Connection reset by peer))
  • [10:51:00] * shankar (n=shankar@122.172.34.138) has joined #beagle
  • [10:51:01] <koen> if your mailer linewraps patches they say "get a decent mailclient"
  • [10:51:12] <Russ> no, it actually eats them
  • [10:51:17] <Russ> whole chunks missing
  • [10:51:29] <Russ> maybe the bottom portion of 1 in 5
  • [10:51:47] <koen> but if you send the patch as attachment, they say "my mailclient can't display them" and you're not allowed to say "get a decent mailclient"
  • [10:54:37] <mru> koen: decent mail clients don't linewrap patches *and* are able to display attachments
  • [10:54:38] * shankar (n=shankar@122.172.34.138) Quit (Read error: 104 (Connection reset by peer))
  • [10:54:59] * shankar (n=shankar@122.172.34.138) has joined #beagle
  • [10:56:36] <Russ> mru: and they have a single horn in the center of their head
  • [10:57:58] <mru> mine seems to work
  • [10:58:12] <eFfeM> koen the musb patch you mentioned before does not do the trick, there is definitely more
  • [10:58:13] <mru> brb, got to feed the unicorns
  • [10:59:01] <ali_as> Do you put a cork on the horn too? Looks kinda stupid but it's so much safer to be around them.
  • [10:59:44] <eFfeM> nah, that is only on american ones
  • [10:59:56] <eFfeM> otherwise they'll sue your ass off if something happens
  • [11:00:23] <eFfeM> mru is in the good old UIK, there it is just folklore
  • [11:00:37] <eFfeM> s/UIK/UK/
  • [11:00:37] * shankar (n=shankar@122.172.34.138) Quit (Read error: 54 (Connection reset by peer))
  • [11:00:59] * shankar (n=shankar@122.172.34.138) has joined #beagle
  • [11:03:23] * shankar (n=shankar@122.172.34.138) Quit (Read error: 104 (Connection reset by peer))
  • [11:17:48] * suihkulokki deals with a funny corporate mailserver
  • [11:18:36] <suihkulokki> it linewraps and converts tabs to spaces. if you attact them as text/plain, it "only" converts linefeeds to msdos style
  • [11:19:22] <Russ> suihkulokki: use zip? or do they block those...
  • [11:50:17] * booxter (n=booxter@80.249.95.245) Quit (Remote closed the connection)
  • [11:59:37] * garren (n=garren@dsl-246-33-211.telkomadsl.co.za) Quit (Read error: 110 (Connection timed out))
  • [12:05:54] * garren (n=garren@dsl-246-33-211.telkomadsl.co.za) has joined #beagle
  • [12:09:41] * dirk2 (n=dirk@p5B0422F8.dip0.t-ipconnect.de) has joined #beagle
  • [12:17:24] * montamer (n=vijay@203.199.213.3) has joined #beagle
  • [12:32:14] * pcgeil (n=steffen@p549E4A47.dip.t-dialin.net) has joined #beagle
  • [12:38:35] <djlewis> its good morning from CDT
  • [12:39:18] * dirk2 (n=dirk@p5B0422F8.dip0.t-ipconnect.de) Quit ("Konversation terminated!")
  • [12:44:03] <djlewis> Russ: are you the russ that helped me yesterday with connman?
  • [12:44:42] <djlewis> just looked at your whois and yep.
  • [12:46:01] <djlewis> I left gpe-info running in angstrom and it took 165MBytes down to 8MBytes before I shut it down.
  • [12:46:39] <djlewis> turns out it has logging going on and it does grab memory by its very nature.
  • [12:47:27] <djlewis> it needs a check box to disable logging if someone wants to leave it on.
  • [12:47:40] <djlewis> Just an observation.
  • [12:49:54] <djlewis> I suppoe that would be an angstrom issue?
  • [12:50:43] <djlewis> Or is that a thing that a volunteer would correct in a kernal fix?
  • [12:57:32] * montamer (n=vijay@203.199.213.3) Quit (Read error: 54 (Connection reset by peer))
  • [12:59:24] * jkridner (n=jason@c-76-31-18-64.hsd1.tx.comcast.net) has joined #beagle
  • [12:59:24] <BeagleTest> jkridner is http://beagleboard.org/user/blog.hangerhead.com
  • [13:00:01] * Wiedi (n=wiedi@newton-air.w.fruky.net) has joined #beagle
  • [13:00:45] <jkridner> good morning all.
  • [13:16:52] * vsr (i=7aa60de8@gateway/web/ajax/mibbit.com/x-61080407f51c7b52) Quit ("http://www.mibbit.com ajax IRC Client")
  • [13:18:48] * tzhau (n=tzhau@dsl-245-80-220.telkomadsl.co.za) Quit (Read error: 110 (Connection timed out))
  • [13:18:55] * tzhau_ (n=tzhau@dsl-245-80-220.telkomadsl.co.za) has joined #beagle
  • [13:19:37] * vsr (i=7aa60de8@gateway/web/ajax/mibbit.com/x-7d412d026c3aa440) has joined #beagle
  • [13:48:14] * raster_ (n=raster@124-170-175-169.dyn.iinet.net.au) has joined #beagle
  • [13:48:52] * djlewis (i=4b0f4002@gateway/web/ajax/mibbit.com/x-02c2276d30c79ab5) Quit ("http://www.mibbit.com ajax IRC Client")
  • [13:55:40] * djlewis (i=4b0f4002@gateway/web/ajax/mibbit.com/x-f14be2e528101041) has joined #beagle
  • [13:56:51] * nabax (n=nabax@89.129.96.71) has joined #beagle
  • [13:58:31] * nabax (n=nabax@89.129.96.71) Quit (Client Quit)
  • [13:58:34] * nabax_ (n=nabax@89.129.96.71) has joined #beagle
  • [13:58:39] * nabax_ (n=nabax@89.129.96.71) Quit (Read error: 104 (Connection reset by peer))
  • [13:58:47] * nabax (n=nabax@89.129.96.71) has joined #beagle
  • [14:04:03] <_AV500_> jkridner: good "morning"
  • [14:05:13] * raster (n=raster@enlightenment/developer/raster) Quit (Read error: 101 (Network is unreachable))
  • [14:06:27] * eFfeM has to wait about 14 hrs before he can say good morning again :-)
  • [14:06:35] <eFfeM> so good afternoon everyone :-)
  • [14:06:48] <mru> it is always morning on irc
  • [14:06:57] * martin_filip (i=d9d084a6@gateway/web/ajax/mibbit.com/x-8cfa16b53ad33007) Quit ("http://www.mibbit.com ajax IRC Client")
  • [14:07:01] <_AV500_> xactly
  • [14:07:20] <_AV500_> and we always have the NEON lights on!
  • [14:07:22] <eFfeM> i know, mips morning, but this is arm :-)
  • [14:07:29] <djlewis> that explains why I see some 24/7 users ;)
  • [14:07:44] <eFfeM> you're talking about koen again :-)
  • [14:08:19] <djlewis> and russ
  • [14:08:28] <_AV500_> koen bot passes the turing test
  • [14:08:51] <eFfeM> guess koen is an avatar powered by beaglebot :-)
  • [14:09:11] <eFfeM> living in UGT: http://www.total-knowledge.com/~ilya/mips/ugt.html
  • [14:09:14] <djlewis> I read all about beaglebot
  • [14:09:29] <djlewis> that is my interest
  • [14:14:46] <koen> jkridner: good morning!
  • [14:15:20] <mru> morning koen
  • [14:16:00] <koen> hey mru
  • [14:16:52] <koen> mru: I see other people are doing ffmpeg NEON stuff now :)
  • [14:17:11] <mru> yeah, david conrad
  • [14:18:11] <mru> he'll be doing some neon for x264
  • [14:19:28] <koen> sweet
  • [14:19:53] <koen> let's see if neon can beat TI c64x stuff again
  • [14:24:28] * zedstar (n=john@fsf/member/zedstar) has joined #beagle
  • [14:24:52] <vsr> which svn version of enlightment are beagle-demo images using?
  • [14:24:56] * kulve (i=acb0af5e@emperor.pingu.fi) Quit (Remote closed the connection)
  • [14:25:22] * kulve (i=8ae16525@194.187.213.45) has joined #beagle
  • [14:26:01] <koen> vsr: load up an image and do 'opkg info e-wm'
  • [14:26:17] <koen> vsr: but it depends on whether you did 'opkg upgrade' or not
  • [14:26:38] <koen> you know, being package based means it can have updates from the intarweb and all
  • [14:27:11] <vsr> i saw on youtube recently e-places working, i built a beagle demo image, i dont see the e-places
  • [14:27:14] <_AV500_> koen: there is a cpu that beats both, the neon and the dsp
  • [14:27:46] <koen> _AV500_: core2quad?
  • [14:28:08] <_AV500_> it is the 9xx inside the dsp, you can have it do stuff like color conv in the backgnd, while both cores do other stuiff
  • [14:28:17] <vsr> some where in between e-places got broken as well, i keep updating oe repo and everytime something or else which was working earlier gets broken
  • [14:28:29] <koen> _AV500_: inside the IVA, not inside the dsp, right?
  • [14:28:34] <djlewis> q
  • [14:28:46] <koen> vsr: why not install stuff from the angstrom feeds?
  • [14:28:53] <koen> vsr: much easier than trying to build your own stuff
  • [14:29:19] <_AV500_> koen: yes, the so called sequencer
  • [14:29:49] <vsr> koen, you mean narcissus? i would like to try out various combinations and more over narcissus does not give all packages
  • [14:30:13] <koen> vsr: no, I mean with opkg, but narcissus is an option as well
  • [14:31:01] <mru> _AV500_: what clock frequency does the arm968 run at?
  • [14:31:30] <_AV500_> no idea, 200 or so
  • [14:31:33] <vsr> i just need a stable oe tree, where i can compile most of the packages
  • [14:31:59] <_AV500_> the idea is to use it to load stuff like dma or the hw accels without disturbing the dsp
  • [14:32:10] <koen> vsr: a stable tree just like http://www.angstrom-distribution.org/building-??ngstr??m tells you to use?
  • [14:32:15] <_AV500_> seq is there since dm320 dsp
  • [14:32:32] <koen> vsr: and narcissus has a 'beagleboard demo' checkbox that gets you everything that's in the demo
  • [14:32:53] <mru> _AV500_: yes, I know roughly how it's used
  • [14:33:43] <mru> do I remember correctly that it doesn't have access to the main ram other than through the dma engines?
  • [14:34:06] <_AV500_> could be, i would have to check
  • [14:34:15] <_AV500_> but that is ok for stuff like color conv
  • [14:36:30] <_AV500_> it also shows up in the jtag scan chain, that had us confused when we went from dv to omap3, we tried to load the sw to the poor arm9 all the time...
  • [14:36:43] <mru> hehe
  • [14:37:42] <mru> looks like it has 32k L2 memory
  • [14:38:08] <mru> 8k ITCM, 4k DTCM
  • [14:38:36] <_AV500_> 1/2 c64
  • [14:39:32] <mru> clock is half of dsp clock
  • [14:39:55] <_AV500_> so 200 was about right :-)
  • [14:43:23] <vsr> koen, does narcassius use the stable 2009 branch?
  • [14:43:24] * flo_lap (n=fuchs@f049172180.adsl.alicedsl.de) Quit (No route to host)
  • [14:43:58] * flo_lap (n=fuchs@f049172180.adsl.alicedsl.de) has joined #beagle
  • [14:45:27] <djlewis> Hcan i copy the sd image partitions to a PC for backup?
  • [14:45:54] * raster_ is now known as raster
  • [14:46:06] <djlewis> I expect I would have to redo the uboot partition for MLO to work.
  • [14:46:45] <djlewis> upon restore
  • [14:47:25] <_AV500_> cp?
  • [14:48:11] <djlewis> well the install instruc's specify un tarring on the card. will a copy of the card work if cp'd back?
  • [14:48:39] <mru> of course you can make a raw copy of the card
  • [14:49:36] <djlewis> I am card poor at thei time so this is why I ask as I want to try different builds and have a backup.
  • [14:58:08] * valhalla (n=valhalla@81-174-23-93.dynamic.ngi.it) Quit ("Leaving")
  • [15:01:13] <_AV500_> tar with the correct flags should archive the partition quite ok
  • [15:04:34] * emeb (n=ericb@ip72-223-90-212.ph.ph.cox.net) has joined #beagle
  • [15:10:46] <djlewis> tnx
  • [15:17:49] * TAK2004 (n=thomas@dslb-088-074-054-145.pools.arcor-ip.net) has joined #beagle
  • [15:17:58] <koen> vsr: narcissus uses binary packages, it's seperate from OE
  • [15:20:37] <vsr> okay, thanks
  • [15:20:53] <vsr> thought it builds from sources on the fly
  • [15:30:29] * Crofton (n=balister@wsip-70-164-137-164.lv.lv.cox.net) Quit ("Leaving")
  • [15:41:20] * valhalla (n=valhalla@81-174-23-93.dynamic.ngi.it) has joined #beagle
  • [15:43:01] * kozak (n=subbu@117.192.19.20) has joined #beagle
  • [15:44:39] * sakoman (n=sakoman@static-74-41-60-154.dsl1.pco.ca.frontiernet.net) Quit (Read error: 104 (Connection reset by peer))
  • [15:44:59] * sakoman (n=sakoman@static-74-41-60-154.dsl1.pco.ca.frontiernet.net) has joined #beagle
  • [16:00:43] * dirk2 (n=dirk@p5B04078F.dip0.t-ipconnect.de) has joined #beagle
  • [16:02:17] * ScriptRipper (n=martin@opensuse/member/MartinMohring) Quit (Remote closed the connection)
  • [16:02:25] * ScriptRipper (n=martin@host-82-135-36-197.customer.m-online.net) has joined #beagle
  • [16:11:14] <vsr> how can i use the sgx modules
  • [16:11:30] <vsr> do i need to compile powervr
  • [16:14:52] * ogra (n=ogra@ubuntu/member/ogra) Quit (Read error: 104 (Connection reset by peer))
  • [16:15:15] * ogra (n=ogra@ubuntu/member/ogra) has joined #beagle
  • [16:19:35] * flo_lap (n=fuchs@f049172180.adsl.alicedsl.de) Quit (Read error: 113 (No route to host))
  • [16:22:23] * BThompsonD (n=bernie@cpe-72-190-75-99.tx.res.rr.com) has joined #beagle
  • [16:29:14] * Openfree` (n=Openfree@222.65.175.96) Quit (Remote closed the connection)
  • [16:30:01] * eFfeM (n=frans@195-241-226-180.ip.telfort.nl) Quit ("Leaving.")
  • [16:32:54] * eFfeM (n=Frans@195-241-226-180.ip.telfort.nl) has joined #beagle
  • [16:44:42] * bazbell (n=a0192809@nat/ti/x-10dbb070f50cd611) Quit (Remote closed the connection)
  • [16:47:32] * bazbell (n=a0192809@nat/ti/x-2acbeb30a470e77a) has joined #beagle
  • [16:50:48] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [16:51:45] <djlewis> Ooooh! I like this synergy tool
  • [16:52:05] <djlewis> much better than vnc
  • [16:52:13] <koen> jkridner: ping
  • [16:52:25] <mru> djlewis: it's not really solving the same problem though
  • [16:52:34] <djlewis> well it is if you have the monitors for it.
  • [16:53:04] <djlewis> I tried both and the x11vnc is sluggish
  • [16:53:20] <mru> vnc is meant for viewing graphical stuff on a remote machine
  • [16:53:26] <djlewis> I know
  • [16:53:28] <mru> the input forwarding thing is a hack
  • [16:54:11] <djlewis> I like using the same kybd and mouse
  • [16:54:30] <djlewis> synergy seems low on resources
  • [16:54:43] <mru> vnc is good when the other machine is 1000 miles away
  • [16:54:58] <mru> remote control of the local console won't do you much good then
  • [16:55:07] <djlewis> Oh sure, i use it all the time here going from machine to machine
  • [16:55:15] * ddompe (n=ddompe@ip221-27-10-190.ct.co.cr) has joined #beagle
  • [16:55:34] <djlewis> I go xp to linux and back a lot
  • [16:55:51] <djlewis> so yep gotta have a desktop
  • [16:56:25] <djlewis> but working beside the BB with a linux box is fine with synergy
  • [16:56:49] <mru> of course it is
  • [16:56:53] <mru> that's what synergy was meant for
  • [16:57:04] <djlewis> Well why the negativity?
  • [16:57:07] <mru> sharing input devices between machines with their own displays
  • [16:57:11] <mru> I'm not negative
  • [16:57:20] <mru> I'm saying you should use the right tool for the job
  • [16:57:29] <djlewis> thats me...
  • [16:57:47] <mru> and comparing two very different tools isn't really fair to either
  • [16:58:09] <djlewis> Well I left out too much content in my first msg for you to et my intent.
  • [16:58:11] <mru> they both solve the problem they were intended to solve very well
  • [16:59:32] * pcgeil (n=steffen@p549E4A47.dip.t-dialin.net) Quit ()
  • [17:00:11] <djlewis> I am still trying to figure out how to get one then two webcams to display on BB so I can move on to openCV
  • [17:08:20] <eFfeM> I had a webcam working but there are some artefacts probalby due to the usb driver
  • [17:08:41] <eFfeM> others have reported properly working webcams
  • [17:12:41] <koen> eFfeM: I'll have some time next week to look at MUSB patches (there are >20 q'd on the linux-usb list)
  • [17:13:32] <eFfeM> ah nice. i have hardly any knowledge there
  • [17:14:01] * koen neither, but applying patch, try, apply more patches, try again is easy to do :)
  • [17:14:35] <koen> and can be done as a background task :)
  • [17:14:36] <eFfeM> actually i have also issues with synchronous transfers with my webcam and with my usb tv tuner working on ehci but not on a hub connected to ehci
  • [17:14:51] * eFfeM likes to understand the problem
  • [17:14:58] <eFfeM> i know: bad habit
  • [17:15:26] <mru> usb hubs are generally evil
  • [17:15:28] <eFfeM> so put the issue aside for a moment and decided to work a little on planning the summer holiday
  • [17:15:30] <koen> bad silicon, incomplete docs and Sergei is a jackass
  • [17:15:37] <eFfeM> i tried 3 different ones
  • [17:15:54] <eFfeM> but ofc they could be the same chipset internally
  • [17:16:06] <eFfeM> same hub + devices works on opensuse if i recall correctly
  • [17:17:09] <koen> I need musb to work for well for the beaglecube project
  • [17:17:28] <koen> eFfeM: Bill should send me the cube case soon, together with your bits
  • [17:17:37] <eFfeM> ah great
  • [17:17:42] <eFfeM> love to see the cube case too
  • [17:21:53] * kozak (n=subbu@117.192.19.20) Quit ("Leaving.")
  • [17:26:25] * mib_u0txw2 (i=c10b883b@gateway/web/ajax/mibbit.com/x-451e0805d121a712) has joined #beagle
  • [17:29:19] * mib_u0txw2 (i=c10b883b@gateway/web/ajax/mibbit.com/x-451e0805d121a712) Quit (Client Quit)
  • [17:33:45] <mib_jn81mk> Are these usb hubs from digikey OK that were used in the beagle 101 workshop? Or are they sort of flakey... ?
  • [17:39:59] * eFfeM1 (n=Frans@195-241-226-180.ip.telfort.nl) has joined #beagle
  • [17:41:00] * DaQatz (n=db@c-66-30-48-150.hsd1.nh.comcast.net) Quit (Read error: 104 (Connection reset by peer))
  • [17:43:33] * mib_jn81mk is now known as MrB09
  • [17:46:02] * dirk2 (n=dirk@p5B04078F.dip0.t-ipconnect.de) Quit ("Konversation terminated!")
  • [17:51:05] * nabax (n=nabax@89.129.96.71) Quit (Remote closed the connection)
  • [17:56:43] * uberfry (n=spinl0ck@vodsl-10890.vo.lu) Quit (Read error: 110 (Connection timed out))
  • [17:56:45] * eFfeM (n=Frans@195-241-226-180.ip.telfort.nl) Quit (Read error: 110 (Connection timed out))
  • [18:05:28] * bmxr (n=bmxr@S01060018f3b11a22.vf.shawcable.net) has joined #beagle
  • [18:07:39] * florian (n=fuchs@f049172180.adsl.alicedsl.de) has joined #beagle
  • [18:31:57] * eFfeM1 is now known as eFfeM
  • [18:32:38] * djlewis (i=4b0f4002@gateway/web/ajax/mibbit.com/x-f14be2e528101041) Quit ("http://www.mibbit.com ajax IRC Client")
  • [18:41:12] <GregorR> OK, I have an S-Video cable, time to get my HMD working :)
  • [18:41:57] <GregorR> ... as soon as I figure out how to make it output via S-Video :)
  • [18:44:40] <GregorR> Sweet, I'm getting a (somewhat lackluster) signal via the TV :)
  • [18:45:05] <mru> s-video *is* lackluster
  • [18:45:11] <mru> that's a requirement
  • [18:49:11] <GregorR> Yeah, but it shouldn't be quite this bad, it's off-center et cetera.
  • [18:49:37] <GregorR> Whoot, froze the BB.
  • [18:53:54] <GregorR> Why are none of the settings in .../overlays usefully modifiable :P
  • [18:56:54] <GregorR> Oh, y'know what, this size looks like PAL, not NTSC, which explains why the display seems effed up vertically (mapping 480 lines to 574 lines ...)
  • [18:57:26] <GregorR> I can't change any of the values in .../displays, is there some trick I'm missing here? X_X
  • [18:57:34] * mib_q8rpp9 (i=4b0f4002@gateway/web/ajax/mibbit.com/x-6ebde6b874b21646) has joined #beagle
  • [18:59:42] <geckosenator> GregorR: echo to it?
  • [19:00:13] * kozak (n=subbu@117.192.19.20) has joined #beagle
  • [19:00:29] <GregorR> geckosenator: Tried, naturally.
  • [19:00:47] <GregorR> Now that I'm looking through some mailing list logs, it looks like NTSC isn't supported by DSS2 >_<
  • [19:00:55] <GregorR> So I'll have to switch to a kernel that uses the old display subsystem.
  • [19:01:41] <koen> the old display subsystem doesn't support s-video at all
  • [19:02:10] <GregorR> Oh, rock on :P
  • [19:02:46] * mib_q8rpp9 (i=4b0f4002@gateway/web/ajax/mibbit.com/x-6ebde6b874b21646) Quit ("http://www.mibbit.com ajax IRC Client")
  • [19:02:51] <GregorR> OK, so correction, I'll have to go hunt for a patch that adds NTSC support.
  • [19:03:12] * mib_q8rpp9 (i=4b0f4002@gateway/web/ajax/mibbit.com/x-45c539755adca0c7) has joined #beagle
  • [19:07:48] * mib_q8rpp9 (i=4b0f4002@gateway/web/ajax/mibbit.com/x-45c539755adca0c7) Quit (Client Quit)
  • [19:07:57] * fulgas is now known as FuL|OUT
  • [19:08:47] <tomba> GregorR: latest DSS2 supports NTSC
  • [19:09:11] <GregorR> tomba: Is that in the latest kernel snapshot that Angstrom uses?
  • [19:09:21] <tomba> that I don't know
  • [19:09:51] <tomba> you can check the kernel sources, Documentation/arm/OMAP/DSS. if NTSC is supported on that kernel, it contains a not about it
  • [19:10:01] <tomba> "a note"
  • [19:10:12] * mib_q8rpp9 (i=4b0f4002@gateway/web/ajax/mibbit.com/x-c2b0ec1f8baf3135) has joined #beagle
  • [19:10:40] <GregorR> OK, thanks *pokes around*
  • [19:11:21] <tomba> well, all you need to do is something like echo ntsc > /sys/something/displayx/mode, if I remember right. but it's good the check the documentation anyway, it's not very long =)
  • [19:12:25] <GregorR> Well, the only caveat to checking the documentation is that I just grabbed a precompiled kernel, so now I have to trudge around the intarwebs for the source :P
  • [19:12:55] <tomba> also, if you don't mind compiling kernel, it's easy to change ntsc as the default mode in older kernels also. only bit missing was changing the mode
  • [19:14:04] <GregorR> Ahh, OK.
  • [19:14:50] * mib_q8rpp9 (i=4b0f4002@gateway/web/ajax/mibbit.com/x-c2b0ec1f8baf3135) Quit (Client Quit)
  • [19:23:46] * TAK2004 (n=thomas@dslb-088-074-054-145.pools.arcor-ip.net) Quit (Remote closed the connection)
  • [19:30:42] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [19:37:18] * djlewis (i=4b0f4002@gateway/web/ajax/mibbit.com/x-921456830223ac54) has joined #beagle
  • [19:40:30] * FuL|OUT is now known as fulgas
  • [19:42:47] * ArteK (n=Artur@81.15.241.96) has left #beagle
  • [19:46:56] * recalcati (i=5d90c85d@gateway/web/ajax/mibbit.com/x-20af4218eb5a347c) has joined #beagle
  • [19:47:30] <recalcati> morning
  • [19:47:38] <mru> morning recalcati
  • [19:48:04] * kulve_ (n=kulve@emperor.pingu.fi) has joined #beagle
  • [19:49:07] <recalcati> mru: ok, BB run correctly a good midori with a locked window, now I have to add plugins , more exactly a download and an handler to correclt player
  • [19:52:28] <koen> tomba: the .29 kernel in OE has latest dss2, but usb otg isn't working yet on that, so it'll take a while to get distributed to the masses
  • [19:52:58] * kulve_ (n=kulve@emperor.pingu.fi) Quit (Remote closed the connection)
  • [19:53:21] * kulve_ (i=1ebd2d6d@emperor.pingu.fi) has joined #beagle
  • [19:58:05] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) has joined #beagle
  • [19:59:59] <eFfeM> koen does .29 support 720p without additional patch ?
  • [20:00:17] <eFfeM> hi recalcati
  • [20:16:36] * RogerMonk (n=a0740758@nat/ti/x-729ab0293753dd58) has joined #beagle
  • [20:17:07] * garren (n=garren@dsl-246-33-211.telkomadsl.co.za) Quit ("Ex-Chat")
  • [20:17:31] * ddompe (n=ddompe@ip221-27-10-190.ct.co.cr) Quit ()
  • [20:19:36] * kulve (i=8ae16525@194.187.213.45) Quit (Read error: 110 (Connection timed out))
  • [20:21:29] <recalcati> eFfeM: thx.
  • [20:21:43] <recalcati> I'll integrate when otg will work.
  • [20:22:33] <recalcati> Now, my version of 2.6.29 has otg working. so I suppose a later patch has create some pb on it
  • [20:34:40] * eFfeM (n=Frans@195-241-226-180.ip.telfort.nl) Quit (Read error: 104 (Connection reset by peer))
  • [20:40:15] * djlewis (i=4b0f4002@gateway/web/ajax/mibbit.com/x-921456830223ac54) Quit ("http://www.mibbit.com ajax IRC Client")
  • [20:43:48] * djlewis (i=4b0f4002@gateway/web/ajax/mibbit.com/x-1166dd75d6840a1a) has joined #beagle
  • [21:00:26] <GregorR> This is fuzzier than even S-Video ought to be ... is S-Video particularly bad on the BB? I mean, I know it's 640x480 at best, but I've used plenty of devices that used S-Video and weren't fuzzy?
  • [21:01:31] <GregorR> Is the BB's output progressive instead of interlaced?
  • [21:03:36] * mib_j4lx94 (i=7cab6904@gateway/web/ajax/mibbit.com/x-efba0c6c770e45fa) has joined #beagle
  • [21:07:23] * djlewis (i=4b0f4002@gateway/web/ajax/mibbit.com/x-1166dd75d6840a1a) Quit ("http://www.mibbit.com ajax IRC Client")
  • [21:07:28] * RogerMonk (n=a0740758@nat/ti/x-729ab0293753dd58) Quit (Remote closed the connection)
  • [21:11:14] <recalcati> don't know.
  • [21:11:48] <recalcati> I also have not nice S-video , but I try long time ago
  • [21:15:17] <ds> broadcast-safe NTSC signals are pretty fuzzy.
  • [21:15:56] <ds> if there's a broadcast filter somewhere in the system, it should be disabled
  • [21:16:31] <ds> and on another note, I finally got Orc to create working neon code. Yay.
  • [21:17:06] <mru> orc as in that research compiler?
  • [21:18:47] <GregorR> <ds> if there's a broadcast filter somewhere in the system, it should be disabled # Is this a hardware thing, or a software setting?
  • [21:21:34] <ds> mru: no
  • [21:22:30] <mru> ds: which orc then?
  • [21:22:34] <ds> mru: http://www.schleef.org/blog/2008/05/23/introducing-orc/
  • [21:23:30] <mru> ah
  • [21:23:44] <mru> confusion...
  • [21:24:05] <mru> I happened to be reading about open64 and the other orc just moments ago
  • [21:24:13] <ds> heh
  • [21:24:51] <mru> on second thought, I don't think open64/orc targets arm at all
  • [21:25:10] <mru> it's supposedly not too hard to add new targets though
  • [21:25:47] <ds> i thought it was itanium-only
  • [21:26:00] <mru> it was originally mips
  • [21:28:20] * valhalla (n=valhalla@81-174-23-93.dynamic.ngi.it) Quit ("Leaving")
  • [21:33:06] * djlewis (i=4b0f4002@gateway/web/ajax/mibbit.com/x-59922124c337a197) has joined #beagle
  • [21:35:38] * bmxr (n=bmxr@S01060018f3b11a22.vf.shawcable.net) Quit (Remote closed the connection)
  • [21:37:16] * mib_j4lx94 (i=7cab6904@gateway/web/ajax/mibbit.com/x-efba0c6c770e45fa) Quit ("http://www.mibbit.com ajax IRC Client")
  • [21:39:09] * arne-unicap (n=arne@p5481B53B.dip0.t-ipconnect.de) has joined #beagle
  • [21:41:33] <GregorR> I'm actually thinking this is progressive ... I wonder if my TV can tell me that ...
  • [21:42:29] <arne-unicap> Are there any ways to get usable performance out of the Beagle USB interface and/or the SDIO interface???? :-(
  • [21:44:49] * ddompe (n=ddompe@ip221-27-10-190.ct.co.cr) has joined #beagle
  • [21:45:22] * gregoiregentil (n=zonbu@adsl-71-135-118-139.dsl.pltn13.pacbell.net) has joined #beagle
  • [21:48:40] <GregorR> arne-unicap: I haven't seen any performance issues with the USB interface.
  • [21:49:04] <GregorR> (Are you referring to OTG or the host port, I've only used OTG so far)
  • [21:50:02] <arne-unicap> OTG
  • [21:50:07] <arne-unicap> I have only the RevB
  • [21:50:27] <GregorR> What performance problems are you having?
  • [21:50:45] <GregorR> (If you're using the OTG to host a device, that could be a power problem)
  • [21:50:50] <arne-unicap> when I dd to a USB hard disk, I only get ~6-8 MB/sec for writes
  • [21:50:58] * brolin (n=brolin@190.157.13.75) has joined #beagle
  • [21:50:59] <geckosenator> wow
  • [21:51:03] <geckosenator> I got like 20 mb/sec
  • [21:51:13] <arne-unicap> for writes?
  • [21:51:26] <geckosenator> don't remember
  • [21:51:26] <arne-unicap> which kernel did you use?
  • [21:51:33] <geckosenator> 2.6.28
  • [21:51:46] <geckosenator> I think just for reads
  • [21:51:51] <geckosenator> I used hdparm
  • [21:51:54] <arne-unicap> yes, reads are faster
  • [21:52:37] <arne-unicap> also I was not able to use any USB camera at a decent resolution and/or frame rate
  • [21:52:50] * Yuvi (n=yuvi@resnet-241-44.resnet.umbc.edu) Quit ()
  • [21:53:04] * Yuvi (n=yuvi@resnet-241-44.resnet.umbc.edu) has joined #beagle
  • [21:57:05] <geckosenator> well I had a lot of problems
  • [21:57:12] <geckosenator> since I couldn't get wifi or ethernet to work
  • [21:57:19] <geckosenator> but I did get bluetooth and harddrives
  • [21:58:14] * jrmuizel (n=jrmuizel@CPE001f5be79d0f-CM0017ee62f8b0.cpe.net.cable.rogers.com) Quit ()
  • [21:58:28] * brolin (n=brolin@190.157.13.75) Quit (Client Quit)
  • [22:01:16] * uberfry (n=spinl0ck@vodsl-10890.vo.lu) has joined #beagle
  • [22:05:59] <GregorR> Bleh, this is bugging me ... there's clearly something wonky about the output, my HMD refuses to accept it, and it's fuzzy on the TV (and not just S-Video fuzzy, "something-is-wrong" fuzzy)
  • [22:07:50] <geckosenator> heh
  • [22:10:41] <mru> GregorR: do you have a capture card?
  • [22:11:57] <GregorR> 'fraid not, I'm recording my TV with my camera to give an example :P
  • [22:12:10] <mru> that will add more fuzz
  • [22:12:29] <GregorR> I'm comparing it to something else using S-Video.
  • [22:12:48] <GregorR> And there's so damn much fuzz in the first place, I think it should be obvious what fuzz I'm referring to and what fuzz is the video.
  • [22:13:21] <mru> are you making a video of the TV?
  • [22:13:26] <mru> use a still camera instead
  • [22:14:32] <mru> if it has optical zoom, set it about halfway and stand so the tv almost fills the picture
  • [22:14:58] <mru> is the tv lcd or crt?
  • [22:18:40] <GregorR> Whoah hey
  • [22:18:43] <GregorR> Suddenly it's beautiful
  • [22:18:45] <GregorR> With no explanation
  • [22:18:55] <mru> it's magic
  • [22:19:01] <GregorR> Clearly :P
  • [22:19:04] <GregorR> MAGIC IS A-OK THOUGH
  • [22:19:08] <mru> I'm still curious
  • [22:19:15] <mru> did you get any images of the fuzz?
  • [22:19:42] <GregorR> Video, no still shot (I was rebooting to get a still shot when it started refusing to fail :P )
  • [22:20:47] <mru> well, post what you got somewhere
  • [22:20:49] <mru> I'll take a look
  • [22:21:11] <GregorR> Sure, just a minute.
  • [22:21:19] * abitos (n=nixgibts@dslb-084-057-154-013.pools.arcor-ip.net) Quit (Read error: 113 (No route to host))
  • [22:25:17] * mrc3 (n=ddiaz@189.157.115.100) Quit (Remote closed the connection)
  • [22:41:09] * arne-unicap (n=arne@p5481B53B.dip0.t-ipconnect.de) Quit (Read error: 60 (Operation timed out))
  • [22:42:18] * kulve_ (i=1ebd2d6d@emperor.pingu.fi) Quit (Read error: 113 (No route to host))
  • [22:52:04] * zuh_ (i=e1197ed6@emperor.pingu.fi) has joined #beagle
  • [22:56:31] * kulve (i=ee2d7adc@emperor.pingu.fi) has joined #beagle
  • [22:56:46] * zuh (i=78086211@emperor.pingu.fi) Quit (Read error: 113 (No route to host))
  • [22:57:41] * nabax (n=nabax@89.129.96.71) has joined #beagle
  • [22:57:53] * sakoman (n=sakoman@static-74-41-60-154.dsl1.pco.ca.frontiernet.net) Quit (Read error: 60 (Operation timed out))
  • [22:59:12] * zuh_ (i=e1197ed6@emperor.pingu.fi) Quit (Remote closed the connection)
  • [22:59:30] * zuh (i=3de0e5a1@emperor.pingu.fi) has joined #beagle
  • [23:04:08] * recalcati (i=5d90c85d@gateway/web/ajax/mibbit.com/x-20af4218eb5a347c) Quit ("http://www.mibbit.com ajax IRC Client")
  • [23:20:00] * uberfry (n=spinl0ck@vodsl-10890.vo.lu) Quit (Read error: 110 (Connection timed out))
  • [23:21:59] * kulve (i=ee2d7adc@emperor.pingu.fi) Quit (Success)
  • [23:26:57] * zuh (i=3de0e5a1@emperor.pingu.fi) Quit (Connection timed out)
  • [23:28:05] * fulgas is now known as FuL|OUT
  • [23:37:14] * florian (n=fuchs@f049172180.adsl.alicedsl.de) Quit (Read error: 113 (No route to host))