**** BEGIN LOGGING AT Thu Aug 30 02:59:57 2007
Aug 30 03:09:34 <xzcvczx> ah dangit.... worst thing possible..... linux doesnt recognise my network card
Aug 30 03:10:26 <legion_> is that even possible
Aug 30 03:10:44 <shackan> you're not even supposed to be here right now then
Aug 30 03:11:27 <cjb> what card is it?
Aug 30 03:12:45 <legion_> its an isa ethernet card. no 8p8c connecter, just a bnc
Aug 30 03:14:33 <Aria> Man, 10base2. Those were the days.
Aug 30 03:14:34 <Aria> (not)
Aug 30 03:15:53 <xzcvczx> lol its a Attrensic thing\
Aug 30 03:16:11 <xzcvczx> not supported til 22 without patching and my liveCD is only built on 19
Aug 30 03:17:32 <rwhitby> cesarb: won't that echo fail with command line too long in a worst case?
Aug 30 03:18:49 <rwhitby> (that's the reason I used find.  I should put a depth limit on it I guess)
Aug 30 03:30:37 <arw_> bye
Aug 30 03:31:29 * gcb77 finally got a sim card and used the NEO as a phone
Aug 30 03:35:39 <Aria> Woo!
Aug 30 03:36:22 <Aria> Am I the only one who thinks of "DFU" as "De-Fuck-Up"?
Aug 30 03:42:25 <raster> homies
Aug 30 03:42:53 <gcb77> doh, we can receive sms yes, can we :(
Aug 30 03:43:15 <gcb77> lol, that came out totally wrong... let me try again
Aug 30 03:43:20 <Aria> Nope.
Aug 30 03:43:22 <Aria> Can't.
Aug 30 03:43:38 <gcb77> I guess I won't be getting my account password then
Aug 30 03:44:36 <Aria> Slip it in another phone?
Aug 30 03:44:53 <raster> passwords are for the weak and feeble
Aug 30 03:44:56 <gcb77> My other phones are Verizon
Aug 30 03:45:31 <gcb77> that was horrible... When I was picking up the SIM card I saw my KRZR in the store, with Java Apps
Aug 30 03:45:45 <gcb77> I have the locked/brewed up Verizon version of it
Aug 30 03:45:50 <Aria> Oh, lovely.
Aug 30 03:46:05 <Aria> The carriers do such evil things to otherwise acceptable items
Aug 30 03:47:07 <gcb77> I was about to push the enter button and start flashing it with Alltel firmware, when somebody pointed me to the OpenMoko site :D
Aug 30 03:47:18 <raster> mostly a us/jp problem.
Aug 30 03:47:39 <raster> the rest of the world has it much better
Aug 30 03:47:48 <Aria> Yeah. For sure.
Aug 30 03:48:05 <roh> hm.. who did the gsm power off measurements recently?
Aug 30 03:49:45 <gcb77> roh: I just tried it and it didn't crash the panel
Aug 30 03:50:20 <roh> gcb77 i know.. i just have a idea what else to try to make it take less power in power-off
Aug 30 03:51:04 <roh> SpeedEvil 
Aug 30 03:51:13 <roh> at least my log says so ;)
Aug 30 04:19:16 <gcb77> Is there a way to tell what the phone is consuming?
Aug 30 04:19:31 <SpeedEvil> Not without external hardware.
Aug 30 04:19:51 <SpeedEvil> roh: I did - a bit.
Aug 30 04:20:44 <SpeedEvil> Briefly 30mA, when 'off' (power button/shutdown -h), 15mA with modem off too. 
Aug 30 04:21:08 <SpeedEvil> I'm assuming that the remaining drains are the audio chips.
Aug 30 04:21:45 <legion_> can you not switch them off?
Aug 30 04:22:12 <legion_> i messed with the creative labs audigy card and i know that dsp had an off command
Aug 30 04:22:40 <SpeedEvil> In principle, they can be switched off.
Aug 30 04:22:51 <SpeedEvil> Finding the right magic to turn them off isn't trivial.
Aug 30 04:23:15 <legion_> well whats the chip
Aug 30 04:24:18 <SpeedEvil> the LM* chip and the WM*
Aug 30 04:24:47 <SpeedEvil> briefly the 'soundcard' and the 'speaker amp'
Aug 30 04:27:14 <cw666> SpeedEvil: this chip is very dumb
Aug 30 04:27:15 <blindcoder> moin
Aug 30 04:27:24 <cw666> any power on/off would be external to the chip i assume
Aug 30 04:27:50 <bedboi> any dbus-er out there?
Aug 30 04:28:08 <SpeedEvil> cw666: which chip?
Aug 30 04:28:13 <legion_> http://www.wolfsonmicro.com/uploads/documents/en/WM8753.pdf page 72
Aug 30 04:28:21 <SpeedEvil> Both can be commanded off via I2C
Aug 30 04:28:24 <roh> SpeedEvil repeat the at@poff measurements. and then do one thing:
Aug 30 04:28:45 <roh> echo 0 > /sys/bus/platform/devices/gta01-pm-gsm.0/power_on
Aug 30 04:28:51 <roh> after the @poff
Aug 30 04:29:01 <roh> and retry if its still 30mA
Aug 30 04:29:12 <SpeedEvil> No, it went down to 15mA after @poff
Aug 30 04:29:50 <cw666> legion_: ah, good to know
Aug 30 04:29:54 <roh> 15 with or without the 0> ?
Aug 30 04:30:06 <SpeedEvil> I diddn't - as I should have - try calling the phone, I assumed it was off due to the lower current and lack of response to commands
Aug 30 04:30:23 <SpeedEvil> without - isn't that the normal state?
Aug 30 04:30:36 <SpeedEvil> wouldn't it just go back on if the state was '1'?
Aug 30 04:30:46 <roh> SpeedEvil how to you measure the current.. from the battery when connected via BT for ip?
Aug 30 04:31:03 <SpeedEvil> No, cut power wire, and use usb-net
Aug 30 04:31:30 <roh> ah.. so not really current to the device bug to the phone
Aug 30 04:31:48 <SpeedEvil> cut USB power wire
Aug 30 04:32:00 <cw666> roh: he's measuring it from between the battery and the device
Aug 30 04:32:06 <roh> AAAH
Aug 30 04:32:11 <roh> now it makes sense to me
Aug 30 04:32:14 <cw666> which is really what counts
Aug 30 04:32:25 <roh> cw666 of course
Aug 30 04:32:46 <cw666> roh: btw, i'm logging on my device power values for charging, etc
Aug 30 04:32:48 <SpeedEvil> Oddly I found that at 100Mhz, the CPU can go down to 1200mV, and still mostly work. 
Aug 30 04:32:56 <SpeedEvil> Audio seems to drop out at 1500
Aug 30 04:32:58 <cw666> im worried right now that it's continually charging actually
Aug 30 04:33:11 <SpeedEvil> chgcur has an offset I think.
Aug 30 04:33:18 <SpeedEvil> And could probably do with more filtering.
Aug 30 04:33:23 <cw666> chgcur is almost useless as-is
Aug 30 04:33:32 <cw666> it needs so many measurements to get anything sane from it
Aug 30 04:33:35 <SpeedEvil> If it was in fact continulaly charging, it would rise over 4200mV
Aug 30 04:33:50 <cw666> 1188443790 4230 480 12 fast_cccv 0 0 1 0 5050
Aug 30 04:33:50 <cw666> 1188443851 4218 746 12 fast_cccv 0 0 1 0 5050
Aug 30 04:33:50 <cw666> 1188443911 4236 746 12 fast_cccv 0 0 1 0 5050
Aug 30 04:33:50 <cw666> 1188443972 4230 746 12 fast_cccv 0 0 1 0 5050
Aug 30 04:33:50 <cw666> 1188444032 4207 853 12 fast_cccv 0 0 1 0 5050
Aug 30 04:33:52 <roh> SpeedEvil the clocks have influence on the plls whcih are needed for i2s data streaming to the codec... at least that is what i assume
Aug 30 04:33:53 <cw666> 1188444093 4218 480 12 fast_cccv 0 0 1 0 5050
Aug 30 04:33:55 <cw666> 1188444153 4218 746 12 fast_cccv 0 0 1 0 5050
Aug 30 04:33:58 <cw666> 1188444214 4212 800 10 fast_cccv 0 0 1 0 5050
Aug 30 04:34:00 <cw666> SpeedEvil: so it is over 4.2V
Aug 30 04:34:15 <SpeedEvil> cw666: I mean 4.5V or so
Aug 30 04:34:37 <cw666> SpeedEvil: well, given the battery was quite warm i powered it off
Aug 30 04:34:44 <SpeedEvil> 4212 is probably not measuring >4.2 with precision. 
Aug 30 04:34:48 <cw666> until i get a steel bucket to put it in
Aug 30 04:34:52 <cw666> of course
Aug 30 04:35:25 <cw666> anyhow, i didnt measure current draw but the batter is quite warm to the touch
Aug 30 04:35:29 <SpeedEvil> roh: Not playing with the clocks - just reducing the core voltage - but screwing up the PLLs is a fair assumption as to the reason.
Aug 30 04:35:54 <SpeedEvil> cw666: 40C, 60C, 80C?
Aug 30 04:35:59 <roh> youre playing woth a p1?
Aug 30 04:36:04 <SpeedEvil> roh: yes.
Aug 30 04:36:16 <cw666> SpeedEvil: touch :-)   touch can't measure temperature reliably
Aug 30 04:36:20 <roh> ok.. just checking which gsm firmware youre using ;=)
Aug 30 04:36:21 <SpeedEvil> I understand that there is some voltage granularity, I need to investigate that more.
Aug 30 04:36:38 <cw666> roh: how do you check the gsm firmware version?  i wondered about that here
Aug 30 04:36:53 <cw666> the firmware i have gets all upset and i have to power cycles the gsm module to get it back
Aug 30 04:37:08 <SpeedEvil> cw666: Sure it can. 40C is a little warmer than body heat, 60C is comfortable to touch for only short periods, and 80C is 'ow'.
Aug 30 04:37:30 <cw666> SpeedEvil: yes and no, fingers aren't 40C, they can in fact be a lot lower
Aug 30 04:37:37 <SpeedEvil> I know.
Aug 30 04:37:46 <roh> cw666 AT+CGMR
Aug 30 04:37:52 <cw666> and it's hot as hell in SF right now, so it's entirely subjective, except to say that it felt warm to my hot-me
Aug 30 04:37:54 <SpeedEvil> It's warmer than sticking your fingers into your armpit.
Aug 30 04:37:56 <cw666> very hot
Aug 30 04:37:58 <cw666> yes, im hot
Aug 30 04:38:20 <cw666> SpeedEvil: i dont know actually, i would guess yes but it's off now
Aug 30 04:38:22 <cw666> i might measure it tomorrow
Aug 30 04:38:26 <SpeedEvil> Or IIRC the core body temperature of a chicken.
Aug 30 04:38:34 <cw666> SpeedEvil: i dont have any chickens handy
Aug 30 04:38:48 <cw666> at least not alive ones that i feel like i need to probe deeply
Aug 30 04:38:55 <legion_> woah this audio amp does do i2c, everything is smart these days
Aug 30 04:39:20 <cw666> roh: can the firmware be upgrade in the field? via some magic AT sequence?
Aug 30 04:39:36 <SpeedEvil> Annoyingly - mp3 audio playback _almost_ works at 1.5Vcore at 100Mhz - but the sound is slightly intermittent.
Aug 30 04:40:15 <roh> cw666 not via AT, but in theory thats possible.
Aug 30 04:40:56 <cw666> roh: ok, i ask this because i wonder if it's not a tad quirky/buggy and if newer firmware might appear sometime
Aug 30 04:41:01 <roh> cw666 we cannot release a firmware flasher yet for 3rd party reasons yet. i'm sorry.. but thats all i can tell you about it
Aug 30 04:42:21 <cw666> SpeedEvil: actually, shutdown failed, so it's still on ... and yes, much warmer than my armpit
Aug 30 04:42:39 <cw666> of course the device has crashed now so i have to rm the battery
Aug 30 04:43:12 <cw666> roh: btw, it would be nice for the next phones if the BT could be NON-USB connected
Aug 30 04:43:28 <cw666> roh: something that uses a less power-unfriendly icky bus
Aug 30 04:44:05 <SpeedEvil> It's not that bad.
Aug 30 04:44:16 <legion_> blue tooth is a usb device?
Aug 30 04:44:17 <rwhitby> looking at the PCB, the BT is on the same interface SDIO) as the wifif
Aug 30 04:44:18 <SpeedEvil> As long as it can raise wake signals.
Aug 30 04:44:22 <roh> cw666 we know. but since the alternative currently is spi, atleast with the chipsets i know thats no solution for high-bandwith bluetooth apps
Aug 30 04:44:23 <SpeedEvil> which it can.
Aug 30 04:44:25 <cw666> SpeedEvil: it can't wake the CPU
Aug 30 04:44:40 <SpeedEvil> however, not in the present implementation.
Aug 30 04:44:52 <SpeedEvil> if there were a couple of extra wires on the FPC, it could.
Aug 30 04:44:57 <cw666> SpeedEvil: you're saying side-band a wake signal?
Aug 30 04:45:15 <hozer> so.. um, how do I upgrade u-boot nicely?
Aug 30 04:45:16 <SpeedEvil> the bluetooth module supports 'turn off USB' input, and 'I have a event' output
Aug 30 04:45:28 <SpeedEvil> (but does not have those wires exported
Aug 30 04:45:32 <roh> SpeedEvil the extra gpio are used for wifi co-op already in newer hardware
Aug 30 04:45:49 <SpeedEvil> so in principle, GTA03 could have that feature, but not 02
Aug 30 04:46:15 <cw666> SpeedEvil: btw, if you get a chance i'd would be a nice to see a pic of your copper-tape battery thang for reference
Aug 30 04:46:27 * hozer reads flashing openmoko page
Aug 30 04:46:50 <cw666> hozer: dont unless you need to
Aug 30 04:47:31 <hozer> cw666: I am getting annoyed with the uboot turning off after 60 seconds or so
Aug 30 04:47:37 <roh> hozer you flash it via dfu-util. if you don't have a debug board it would be a good idea to stick to tested builds
Aug 30 04:47:48 * hozer has a debug board
Aug 30 04:47:55 <cw666> oh, go for it then
Aug 30 04:48:02 <hozer> I'm just wondering what are the tested builds
Aug 30 04:48:15 <cw666> i have one being dropped off tomorrow so i'll probably do it myself then, because yeah, the 60s turnoff is annoying
Aug 30 04:48:19 <roh> then have fun. ive flashed one on my bv4 built yesterday
Aug 30 04:48:27 <hozer> I suppose I should just go test the one I just built
Aug 30 04:48:41 <cw666> qemu might not be a bad idea to test uboot
Aug 30 04:48:54 <roh> u-boot-gta01bv4-1.2.0+git3e66c078003607a7d1d214c15a5f262bc1b4032f+svn20070829-r2.bin works for me
Aug 30 04:49:15 <SpeedEvil> cw666: Quite simple. Take 8cm*2mm of copper tape. 4cm*4mm of copper tape backing (card would work better) stick copper tape along the long side, then back up, and then cut off the end. So you have a copper strip on either side of the card. Attach wires to the copper tape. Tape to the battery over the "+" electrode, along the length of the battery, with enough over that it can be folded down over the battery contact.
Aug 30 04:49:26 <roh> if you want i can upload it quickly
Aug 30 04:50:09 <cw666> SpeedEvil: ok, so did you figure out what the last 15mA or so was caused by?
Aug 30 04:50:41 <SpeedEvil> It's a supposition.
Aug 30 04:50:47 <roh> http://people.openmoko.org/roh/u-boot-gta01bv4-1.2.0+git3e66c078003607a7d1d214c15a5f262bc1b4032f+svn20070829-r2.bin
Aug 30 04:51:26 <SpeedEvil> Actually, it can probably only be the LM* chip - as the WM* chip isn't connected to teh battery directly.
Aug 30 04:52:55 <legion_> yeah that draw seems to match what the doc said for any mode other than 0
Aug 30 04:53:43 <roh> I(SD) should be 0.2-3 uA
Aug 30 04:54:12 <SpeedEvil> suspend-RAM shouldn't be more than a milliamp tops.
Aug 30 04:55:00 <cw666> roh: HW: GTA01Bv4, GSM: gsm_ac_gp_fd_pu_em_cph_ds_vc_cal35_ri_36_amd8_ts0-Moko1
Aug 30 04:55:07 <cw666> which i assume is the most recent
Aug 30 04:55:20 <roh> cw666 it is what we delivered with P1
Aug 30 04:55:44 <SpeedEvil> It is a bit annoying that the calypso chipset seemingly draws rather a lot in standby. (compared to forex my nk3300)
Aug 30 04:56:23 <SpeedEvil> Though to be fair that may only be dualband - I'm unsure.
Aug 30 04:57:16 <cw666> SpeedEvil: what does it draw when idle?
Aug 30 04:59:03 <SpeedEvil> I don't know - I know it gets 5 days on a 800mAh battery.
Aug 30 04:59:16 <SpeedEvil> Or about 8mA
Aug 30 04:59:24 <cw666> idle or off?
Aug 30 04:59:28 <SpeedEvil> standby
Aug 30 04:59:33 <SpeedEvil> (the nokia)
Aug 30 04:59:42 <cw666> right, my nokia gets something like that too
Aug 30 04:59:53 <cw666> i mean the calypso, what does it do?
Aug 30 04:59:55 <hozer> hrrm... my latest uboot build seems to have some issues with /dev/ttyACM0
Aug 30 05:00:03 <SpeedEvil> Seems to be 15mA
Aug 30 05:00:13 <cw666> SpeedEvil: that sounds right, other modules claim 17mA or so
Aug 30 05:02:56 <cw666> roh: some commands take a tad over 5s sometimes, i suspect newer firmware might address that
Aug 30 05:03:08 <cw666> (given similar devices respond fairly quickly)
Aug 30 05:03:50 <hozer> roh: have you tried cu with /dev/ttyACM0 with that uboot build?
Aug 30 05:04:04 <roh> cw666 i dunno
Aug 30 05:04:21 <roh> hozer nope... but i can 
Aug 30 05:05:33 <hozer> try it..
Aug 30 05:05:37 <hozer> it seems .. weird
Aug 30 05:05:47 <hozer> do you use minicom?
Aug 30 05:05:51 <cw666> roh: basically, the delays are long enough that gsmd falls over pretty often
Aug 30 05:06:43 <hozer> troy@lyta:/extra/moko/images/openmoko$ cu -l /dev/ttyACM0
Aug 30 05:06:43 <hozer> Connected.
Aug 30 05:06:43 <hozer> In:    usbtty
Aug 30 05:06:43 <hozer> OIn:    usbtty
Aug 30 05:06:43 <hozer> O_?Y?U?U?W?E?U*????U??q?L?U??In:ut:   usbtty
Aug 30 05:06:43 <hozer> Err:   usbtty
Aug 30 05:06:45 <hozer> DEVICE_CONFIGURED: ?U?_1
Aug 30 05:06:47 <hozer> Enabling auto  usbtty
Aug 30 05:07:16 <roh> hozer uuuuh.. thats indeed a bug
Aug 30 05:07:30 <roh> same here
Aug 30 05:07:55 <hozer> so, you want to file that, or should I?
Aug 30 05:08:10 <hozer> that's my third bug today :0
Aug 30 05:08:15 <roh> hozer feel free
Aug 30 05:08:29 <hozer> I was hoping you'd do it ;0
Aug 30 05:08:33 <hozer> err ;)
Aug 30 05:08:50 <cw666> roh: so is there are list of moko specific AT+ magic?
Aug 30 05:08:53 <cw666> i didnt see one
Aug 30 05:09:17 <roh> https://wiki.openmoko.org/wiki/GTA01_gsm_modem has some
Aug 30 05:09:46 <cw666> thanks
Aug 30 05:10:14 <roh> https://wiki.openmoko.org/wiki/Hardware:AT_Commands has some more
Aug 30 05:11:40 <hozer> okay, now that works *much* better than the emulator
Aug 30 05:11:47 * hozer just flashed stuff
Aug 30 05:12:26 <cw666> roh: sidetone levels mean what?
Aug 30 05:12:57 <mickey|ICE676> *yawn*
Aug 30 05:12:57 <mickey|ICE676> morning folks
Aug 30 05:12:57 <mickey|ICE676> abraxa_: ping
Aug 30 05:13:17 <abraxa_> mickey|ICE676: I'm here, what's up? :)
Aug 30 05:13:18 <roh> cw666 the mix from mic to reciever. people need that to not believe its broken.
Aug 30 05:13:34 <xzcvczx> mickey|ICE676: how was the witch
Aug 30 05:13:36 <xzcvczx> oops
Aug 30 05:13:39 <mickey|ICE676> abraxa_: ah cool. can you mail me a sample playlist? i want to check whether playing works :)
Aug 30 05:13:42 <xzcvczx> s/witch/mother in law
Aug 30 05:13:46 <mickey|ICE676> heh
Aug 30 05:14:06 <roh> morning mickey
Aug 30 05:14:08 <abraxa_> mickey|ICE676: Sure thing
Aug 30 05:14:16 <mickey|ICE676> xzcvczx: actually she's very nice.... NOW, that is. she wasn't in the early years, but now she gave in and realized there won't be another man :D
Aug 30 05:14:23 <xzcvczx> lol
Aug 30 05:15:47 * cw666 fights with GPRS and T-Mobile some more
Aug 30 05:16:29 <mickey|ICE676> i guess now that i'm a dr. and my bank account is >= 0 i'm good enough for her daughter
Aug 30 05:16:29 <mickey|ICE676> heh
Aug 30 05:17:50 <mickey|ICE676> roh: morning!
Aug 30 05:17:52 * mickey|ICE676 has no idea what to tell the audience @ MDD07
Aug 30 05:17:52 <mickey|ICE676> i even don't know which kind of audience this is
Aug 30 05:17:52 <mickey|ICE676> looks like symbian and WM weenies
Aug 30 05:17:53 <mickey|ICE676> ;)
Aug 30 05:18:05 <roh> hrrrr
Aug 30 05:18:20 <abraxa_> mickey|ICE676: Mail sent
Aug 30 05:18:32 <SpeedEvil> plagarism. "I have a dream"...
Aug 30 05:19:06 <roh> SpeedEvil we don't dream.. we just do it... through concrete walls with a bulldozer if needed
Aug 30 05:19:37 <cw666> does anyone know if anyone on the list has T-Mobile GPRS working?  something is slightly off here
Aug 30 05:19:51 <mickey|ICE676> well, TT and Motorola are cosponsoring the event, so i guess i don't need to tell about why to put OSS on a phone
Aug 30 05:19:51 <mickey|ICE676> but rather _how_ it should be
Aug 30 05:19:51 * mickey|ICE676 thinks about the subtitle "OSS on mobile done right", but that may be too offensive
Aug 30 05:19:53 <cw666> t-mobile tried to help but it took like 15 minutes before they could tell me the APN :-/
Aug 30 05:20:13 <stefan_schmidt> *yaaaawwwwnnnn*
Aug 30 05:20:16 <mickey|ICE676> roh: heh, i like that metaphor :)
Aug 30 05:20:22 <mickey|ICE676> heyyyy stefan_schmidt
Aug 30 05:20:23 <stefan_schmidt> morning mickey|ICE676 
Aug 30 05:20:24 <mickey|ICE676> good morning!
Aug 30 05:20:25 <cw666> mickey|ICE676: you can't claim it *is* done right
Aug 30 05:20:30 <mickey|ICE676> cw666: right
Aug 30 05:20:50 <stefan_schmidt> mickey|ICE676: "LInux is now mobile"? *caught*
Aug 30 05:20:55 <mickey|ICE676> stefan_schmidt: carrier 12, seat 43.
Aug 30 05:21:01 <mickey|ICE676> stefan_schmidt: hehe
Aug 30 05:21:27 <mickey|ICE676> stefan_schmidt: seats 31
Aug 30 05:21:27 <roh> mickey why not be offensive? now where apple followers are starting to insult us and spread FUD it seams they already see us as rival
Aug 30 05:21:37 <mickey|ICE676> + 32 were not next to each other
Aug 30 05:21:39 <mickey|ICE676> so i moved
Aug 30 05:21:40 <cw666> roh: what is the point of AT@POFF? given you can just externally power it down?
Aug 30 05:21:51 <stefan_schmidt> mickey|ICE676: ok, table?
Aug 30 05:21:53 <mickey|ICE676> roh: yeah, that report is pretty polemic :/
Aug 30 05:21:59 <roh> cw666 nope. in gta01 there is no main power switch
Aug 30 05:22:03 <mickey|ICE676> stefan_schmidt: unfortunately not. all tables are reserved
Aug 30 05:22:09 <mickey|ICE676> (that's probably why we couldn't get one)
Aug 30 05:22:12 <stefan_schmidt> mickey|ICE676: next train
Aug 30 05:22:16 <hozer> what are the apple weenies saying?
Aug 30 05:22:34 <stefan_schmidt> anyway, shutdown to catch my first train. Later guys
Aug 30 05:22:40 <roh> mickey i see that report as 'credits'. it means they are in enough panic already
Aug 30 05:23:22 <cw666> roh: what is /sys/bus/platform/devices/gta01-pm-gsm.0/power_on for then?
Aug 30 05:23:31 <SpeedEvil> It's a virtual 'power button'
Aug 30 05:23:36 <mickey|ICE676> yep
Aug 30 05:23:36 <mickey|ICE676> where was this FUD blog entry?
Aug 30 05:23:36 <mickey|ICE676> don't have it offhand
Aug 30 05:23:42 <cw666> huh? it seems to work
Aug 30 05:23:44 <roh> cw666 pressing the power-on button on the pmu inside the gsm blackbox
Aug 30 05:23:44 <SpeedEvil> the chipset was designed to have a physical button to turn it on
Aug 30 05:23:53 <cw666> in that if i power it off/on it looks like the module restarts
Aug 30 05:24:13 <roh> cw666 also the sysctl switches the latch between gsm and serial to the debugboard
Aug 30 05:24:32 <cw666> roh: i use that to 'reset' the module when it gets upset ... so for that it works
Aug 30 05:24:47 <cw666> but i take it that when it's 'off' it doesnt really prevent power draw then?
Aug 30 05:24:53 <SpeedEvil> No.
Aug 30 05:24:57 <roh> cw666 its a soft reset on rising flank
Aug 30 05:25:00 <SpeedEvil> If you power down (off button)
Aug 30 05:25:05 <SpeedEvil> then call the phone
Aug 30 05:25:20 <cw666> oh, i never tried that
Aug 30 05:25:24 <SpeedEvil> it will be engaged, not 'off'
Aug 30 05:25:35 <cw666> i need to find my number to test that :)
Aug 30 05:28:39 <hozer> is there any equivalent to the 'linux-headers-2.6.x.foo' debian package for OM?
Aug 30 05:33:53 <rwhitby> mickey|ICE676: can we consider changing dropbear to openssh in openmoko, since you need openssh-sftp for sshfs ?
Aug 30 05:35:04 <abraxa_> I take it this is the FUD comparison mentioned: http://www.roughlydrafted.com/RD/RDM.Tech.Q3.07/B10AE668-EAD3-46DC-A042-5EF3461D63EF.html
Aug 30 05:36:39 <CM> abraxa_: I just looked through the logs to find it, but I guess I'm too slow before my morning coffee. ;)
Aug 30 05:37:49 <mickey|ICE676> abraxa_: that's it
Aug 30 05:37:53 <abraxa_> CM: Don't worry, this is a link that doesn't need being memorized anyway (not like the URL would make it easy anyway)
Aug 30 05:38:02 <mickey|ICE676> rwhitby: yes, I'll think about it. we should have enough space on the neo
Aug 30 05:38:32 <mickey|ICE676> sshfs is incredibly useful
Aug 30 05:39:32 <hozer> well.. instead of changeing it... can you make openssh be an optional ipkg
Aug 30 05:39:53 <hozer> which does an /etc/alternatives replacement for dropbear
Aug 30 05:40:24 <mickey|ICE676> abraxa_: btw, we need to make it a bit more solid, e.g. it crashes if it can't build the gst pipeline. we should rather open a dialog box and tell the user to install gst-meta-audio or so
Aug 30 05:40:28 <rwhitby> hozer: the trouble is the if you do that changeover in a naive manner from a usbnet connection, your connection dies and you can't get back in
Aug 30 05:40:47 <rwhitby> you have to use nohup or && on a single line
Aug 30 05:41:19 <hozer> so.. what happens on a debian box if you install dropbear
Aug 30 05:41:34 <rwhitby> on the nslu2, at least, openssh is also three times faster than dropbear.
Aug 30 05:41:51 <hozer> I upgrade SSH all the time remotely on debian boxes
Aug 30 05:41:53 <CM> What's the size difference?
Aug 30 05:42:19 <rwhitby> hozer: openembedded is not the same as debian.  at the moment, if you do it on openssh the connection dies
Aug 30 05:42:47 <abraxa_> mickey|ICE676: Yeah, agreed - I'm not too sure on how verbose the player should be, though... do messages on stderr suffice for such things or do we want to nag the user about such errors all the time?
Aug 30 05:42:57 <hozer> so, why can't openembedded do the 'right' thing like debian does
Aug 30 05:43:26 <hozer> but I suppose if it's not much bigger (memory footprint wise) , then switch
Aug 30 05:43:54 <rwhitby> hozer: nslu2 has 32 mb ram and 8mb flash, and we decided to switch :-)
Aug 30 05:45:09 <abraxa_> mickey|ICE676: Another issue I want resolved is the problem that when started and resuming playback, the player starts gstreamer playback while the UI is still being built - I couldn't find a proper way to ask GTK when it has finished doing that
Aug 30 05:46:24 <mickey|ICE676> really?
Aug 30 05:46:28 <mickey|ICE676> that's surprising me
Aug 30 05:46:37 <mickey|ICE676> dropbear was supposed to be light _and_ fast
Aug 30 05:47:19 <mickey|ICE676> abraxa_: ok, once i can get it to play something i'll think about that
Aug 30 05:47:46 <rwhitby> mickey|ICE676: the speed comparison was 1.5 years ago, and we put it down to openssh having arm optimisations for the crypto algorithms.
Aug 30 05:48:07 <rwhitby> dropbear may have got better.  Dunno, cause I never use it anymore :-)
Aug 30 05:48:19 <rwhitby> (sshfs is too convenient)
Aug 30 05:51:22 <hozer> and.. um it's not shutting down cleanly :P
Aug 30 05:51:43 <hozer> I did halt, and now the phone is kinda sitting there doing who the heck knows what
Aug 30 05:59:58 <mickey|ICE676> strange. i don't have problems with shutting done since months
Aug 30 05:59:58 <mickey|ICE676> are you using the power button menu?
Aug 30 06:01:59 <cw666> so gprs on this fails on the macbook too
Aug 30 06:02:01 <cw666> i think
Aug 30 06:02:12 <cw666> if i could figure out how to su there i would tcpdump and look harder
Aug 30 06:02:17 <gcb77> mickey|ICE676: yes, I use the power button menu all the time
Aug 30 06:02:32 <Aria> sudo -s
Aug 30 06:02:34 <gcb77> mickey|ICE676: I was wondering if there was a way to swap orientation from the terminal
Aug 30 06:02:36 <CM> hozer: That happens to me all the time if I touch the /dev/mtd* blocks
Aug 30 06:02:44 <Aria> gcb77: xrandr 
Aug 30 06:03:03 <gcb77> Aria: awesome, thank you
Aug 30 06:03:27 <cw666> gcb77: xrandr -o <n>
Aug 30 06:03:35 <mickey|ICE676> strange. are you sure you have the new uboot that erases the complete rootfs?
Aug 30 06:04:13 <mickey|ICE676> good morning LaF0rge
Aug 30 06:05:08 <roh> morning LaF0rge
Aug 30 06:05:15 <mickey|ICE676> gcb77: I'll shortly code an own openmoko-terminal based on vte and that will get a menu for most-important-commands
Aug 30 06:06:25 <gcb77> mickey|ICE676: that's great, thank you.  This thing's really coming together.
Aug 30 06:07:32 <gcb77> woo hoo, this thing really does have a GPS! (I'm running the satscan tool posted on the lists)
Aug 30 06:10:07 <mickey|ICE676> is tabbed terminal important for you guys?
Aug 30 06:10:25 <CM> I don't think so
Aug 30 06:10:37 <CM> Since you can have several terminals open anyway
Aug 30 06:10:49 <gcb77> I've used it a few times, but with the new UI its easy to get another one going
Aug 30 06:12:41 <aloril> (script) planet: Thomas Wood: OpenMoko 2007.2 http://blogs.gnome.org/thos/2007/08/21/openmoko-20072/
Aug 30 06:14:59 <CIA-20> openmoko: 03abraxa * r2854 10/trunk/src/target/OM-2007.2/applications/openmoko-mediaplayer2/src/ (12 files): 
Aug 30 06:14:59 <CIA-20> openmoko: Cleared dialog titles
Aug 30 06:14:59 <CIA-20> openmoko: Added fade-in on session resume
Aug 30 06:14:59 <CIA-20> openmoko: Improved gstreamer error handling
Aug 30 06:14:59 <CIA-20> openmoko: Various minor improvements/bugfixes
Aug 30 06:16:05 <CM> abraxa_: Is the media player available in the current builds? Any screenshots? :D
Aug 30 06:16:47 <abraxa_> CM: It is available now, however currently only handles pre-made playlists - fixing that is next on my agenda
Aug 30 06:16:57 <CM> Ok, cool :)
Aug 30 06:17:22 <mickeyl_> anyone knowing lennart p.? i need to talk w/ him. the gstreamer pulseaudio sink is taking a whole lot of resources. we can only play gstreamer directly over alsa
Aug 30 06:17:35 <cw666> roh: one other thing i wondered about, right now on calls there is bad echo ... might AT@AUL be used to address this at a later time?
Aug 30 06:17:45 <rwhitby> nslu2-linux autobuilder is currently building task-openmoko-native-sdk and task-openmoko-feed (and all the packages those two rdepend on)
Aug 30 06:18:10 <abraxa_> CM: As about the UI - I'm pretty much using the same layout as http://wiki.openmoko.org/wiki/Image:Screen_music_player.png so there's not too much new to see honestly
Aug 30 06:19:24 <CM> abraxa_: Ok, fair enough. Just curious what it looked like. Thanks :)
Aug 30 06:19:35 <abraxa_> mickeyl_: Is that the reason why playback uses >80% CPU?
Aug 30 06:20:07 <mickeyl_> abraxa_: depends on how you build the pipeline
Aug 30 06:20:13 <mickeyl_> the best way to build it right now is
Aug 30 06:20:20 <mickeyl_> filesrc ! mad ! alsa
Aug 30 06:20:29 <mickeyl_> takes about 25%
Aug 30 06:20:43 <abraxa_> Should I hardcode that instead of using the playbin?
Aug 30 06:20:49 <CM> abraxa_: Btw, I'm a big fan of Muine's UI: http://www.muine-player.org/wiki/Image:CMMainPageScreenshot.png  ;)
Aug 30 06:20:51 <cw666> wtf; my RAZR can surf whilst it has a GPRS connection over BT going
Aug 30 06:20:53 <abraxa_> Using the decodebin, that is
Aug 30 06:21:00 <cw666> i wonder how it does that, some alternative for web cruft?
Aug 30 06:21:16 <mickeyl_> abraxa_: yes, once you detect a mp3, you should use this pipeline instead of the generic playbin
Aug 30 06:21:39 <mickeyl_> same goes for ogg (but w/ vorbis or whatever instead of mad)
Aug 30 06:21:53 <abraxa_> mickeyl_: Okay, I'll do that then :)
Aug 30 06:22:00 <mickeyl_> cool, thanks
Aug 30 06:22:17 <mickeyl_> for gta01 we really need to use specialized pipelines
Aug 30 06:22:22 <mickeyl_> gta02 should be able to just use the playbin
Aug 30 06:22:38 <mickeyl_> althoug specialized ones are beneficial anyways because they save energy
Aug 30 06:22:51 <cw666> mickeyl_: what do you mean here by specialize pipelines?
Aug 30 06:22:59 <CM> mickeyl_: I don't know lennart, but the email stated on his homepage says mzvzcerffhz@0pointer.de
Aug 30 06:23:01 <cw666> +d
Aug 30 06:23:12 <mickeyl_> specialized = you completely specify the complete pipeline
Aug 30 06:23:26 <cw666> how does that save energy?
Aug 30 06:23:29 <mickeyl_> standard = you use smart elements like playbin, decodebin, etc.
Aug 30 06:23:38 <mickeyl_> less CPU usage, less energy consimption
Aug 30 06:23:54 <cw666> ah, k
Aug 30 06:24:39 <rwhitby> dsilva: http://ipkg.nslu2-linux.org/feeds/openmoko/unstable/armv4t/openssh-sftp_4.6p1-r2_armv4t.ipk (for sshfs)
Aug 30 06:25:07 <mickeyl_> anyone ever played with aufs?
Aug 30 06:25:15 <mickeyl_> they claim to be better than unionfs
Aug 30 06:25:27 <abraxa_> CM: The UI is interesting indeed, though to me it lacks important features (=buttons) ;x
Aug 30 06:25:47 <CM> abraxa_: What buttons?
Aug 30 06:26:44 <abraxa_> CM: Volume up/down (or a slider), ffwd, rew, shuffle, repeat and such
Aug 30 06:27:37 <CM> abraxa_: I agree on the shuffle/repeat thing, but the forward and rev is there, and the volume button is a slider when you press it
Aug 30 06:27:57 <abraxa_> Interesting concept
Aug 30 06:28:15 <abraxa_> Might be a tad awkward to use with just a finger, though?
Aug 30 06:28:52 <CM> abraxa_: Hmm.. Might be, I haven't tried. That volume button is the same as YouTube and all those ones have
Aug 30 06:30:19 <abraxa_> CM: My g/f wants to have a shot at improving the UI so I'll show her the screenshot of Muine and see what she makes of it :)
Aug 30 06:30:40 <CM> abraxa_: Why not install it and let her play around with it? ;)
Aug 30 06:30:54 <CM> abraxa_: It's a mono app though
Aug 30 06:31:20 <abraxa_> CM: If I can get mono to build then I'll do that - it's been a while since I tried
Aug 30 06:31:50 <CM> abraxa_: Thanks for being open about UI suggestions
Aug 30 06:33:15 <abraxa_> CM: Oh heh, the 2007.1 UI isn't quite en par with the 2007.2 applications so of course I want to remedy that - just wanted to get things working for the SoC first ;)
Aug 30 06:33:21 <cw666> ok, i have GPRS sorta working
Aug 30 06:33:35 <cw666> sweet
Aug 30 06:35:27 <CM> abraxa_: Is it some kind of back-end/front-end design so that you could easily make your own gui using the "official" openmoko mediaplayer backend?
Aug 30 06:36:07 <dsilva> rwhitby: thanks
Aug 30 06:36:42 * mickeyl_ thinks OM 2008 will be where almost every application will have this backend/frontend split and you will be able to completely change the UI for every app
Aug 30 06:37:15 <abraxa_> CM: Not really, that is however on the to-do list as Mickey just said
Aug 30 06:37:36 <CM> Ok, sounds reasonable. :)
Aug 30 06:38:12 * CM thinks of the first Nokia 770 release compared to the curent software
Aug 30 06:40:00 <roh> cw666 AT@ST is for the sidetone
Aug 30 06:40:28 <mickeyl_> cw666: congrats.
Aug 30 06:42:15 <roh> cw666 see https://wiki.openmoko.org/wiki/GTA01_gsm_modem
Aug 30 07:04:37 <cw666> roh: you think it's a bad sidetone setting that is causing the echo effect(s)?
Aug 30 07:08:41 <CIA-20> openmoko: 03laforge * r2855 10/trunk/src/target/u-boot/patches/ (series uboot-s3c2410-warnings-fix.patch): uboot has merged the warnings-fix patch into mainline
Aug 30 07:13:53 <cdbot2> * * OM Bug 779 has been created by roh(AT)openmoko.org
Aug 30 07:13:54 <cdbot2> * * acm console broken - spits out garbage
Aug 30 07:13:55 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=779
Aug 30 07:16:29 <roh> cw666 could be.. there were reports about that.
Aug 30 07:16:53 <roh> cw666 we usually disable the sidetone on the modem and generate it ourselves via a mixer setting. at least that is/was the plan
Aug 30 07:21:05 <cw666> roh: ok, the AT@AUL details are only in effect when sidetone is on or those are orthogonal?
Aug 30 07:21:19 * cw666 wonders how long he can keep GPRS sessions up
Aug 30 07:22:17 <roh> cw666 i don't have a clue. i do not have more informations about how it works exactly than in the wiki there
Aug 30 07:23:03 <roh> i'm shure we will take care of correct settings there once gsmd is stable and calling works everytime in all directions
Aug 30 07:23:40 <cw666> yeah, im in the process of replacing gsmd potentially, so ill have to keep an eye on those changes
Aug 30 07:27:34 <trains> hi, i'm building openmoko for the first time. does it take too long? (in and average 2 years laptop)
Aug 30 07:28:07 <CM> trains: Yes
Aug 30 07:28:36 <trains> yes? :) how long does it took you?
Aug 30 07:28:36 <CM> trains: On my 3000+ amd 64-bit with 1gb ram it takes about 8h
Aug 30 07:28:47 <trains> :O
Aug 30 07:28:51 <trains> 8h?
Aug 30 07:29:01 <CM> On a new core 2 duo you can get it down to 4h or so
Aug 30 07:29:05 <CM> Yes
Aug 30 07:29:17 <CM> That is from scratch, with the makefile
Aug 30 07:29:38 <trains> i forgot to mention i'm compiling on feisty running on vmware :)
Aug 30 07:29:45 <cw666> trains: yeah, sadly it's a tad insane in disk and time requirements
Aug 30 07:30:02 <CM> Hehe.. trains Probably even longer in that case
Aug 30 07:30:09 <trains> yeah
Aug 30 07:30:38 <CM> Just leave it on over night or so :)
Aug 30 07:30:58 <trains> i can stop it and resume, right?
Aug 30 07:31:03 <CM> Yes
Aug 30 07:31:06 <trains> ok
Aug 30 07:31:22 <Xx13> is there and .ebuild yet?  ;)
Aug 30 07:31:34 <ynezz> what's that
Aug 30 07:31:37 <borg_> trains: you are building a whole distribution
Aug 30 07:31:42 <borg_> that takes a while :)
Aug 30 07:31:45 <CM> gentoo package
Aug 30 07:31:52 <ynezz> meh
Aug 30 07:32:04 <trains> Xx13: .ebuild ? where?
Aug 30 07:32:29 <Xx13> i dunno, just askin
Aug 30 07:32:32 <CM> There's no ebuild for openmoko
Aug 30 07:32:41 <Xx13> would be nice to have
Aug 30 07:32:45 <trains> what's ebuild?
Aug 30 07:32:54 <CM> But there is for monotone, sqlite3, pysqlite2 and all the other requirements
Aug 30 07:33:08 <CM> trains: It's the package files for Gentoo
Aug 30 07:33:35 <CM> Xx13: What would be nice to have?
Aug 30 07:33:35 <trains> like .deb for debian?
Aug 30 07:33:56 <Dmitry_Platonov> trains, more like .bb for openembedded
Aug 30 07:34:07 <CM> trains: Not really, since you compile it. As Dmitry_Platonov said
Aug 30 07:34:51 <trains> i'm not very familiar with openembedded. don't know what .bb is
Aug 30 07:35:20 <CM> trains: "BitBake", a recipie for how to build a program and all its requirements
Aug 30 07:35:28 <jeddy3> hmm, ipkg seems to download new kernels, but how do i update to it
Aug 30 07:35:50 <CM> jeddy3: Where do you run the ipkg update? On the Neo?
Aug 30 07:36:03 <Xx13> trains: ebuild is a script that fetches the files and patches, compiles and installs
Aug 30 07:36:05 * CM still haven't gotten the routing to the neo working
Aug 30 07:36:25 <trains> ok. thanks
Aug 30 07:37:20 <trains> i want to start help coding. Any idea where i should start helping?
Aug 30 07:37:39 <Xx13> trains: that's why compiling overnite is nothing new to gentoo ppl
Aug 30 07:38:06 <CM> Xx13: Or that it's no problem using the computer at the same time as you compile. :)
Aug 30 07:38:30 <Xx13> depends on what you do
Aug 30 07:38:56 <Dmitry_Platonov> trains, would be nice to have working gsmd/libgsmd. Also, in-kernel power management (freq. scaling, etc.) would be nice 8)
Aug 30 07:39:21 <CM> Dmitry_Platonov: Hehe.. I agree, and I wish I had your optimism ;)
Aug 30 07:40:40 <Xx13> is it possible to powerdown gsm module completely, so the phone can be used on a plane?
Aug 30 07:40:55 <cw666> Xx13: yes
Aug 30 07:41:17 <Xx13> cool
Aug 30 07:41:21 <CM> Xx13: I have enough trouble getting it to power up! :P
Aug 30 07:41:30 <trains> Dmitry_Platonov: i don't have a neo yet. so i was thinking about tasks that don't require a device. and i know nothing about of kernel devel
Aug 30 07:41:52 <cw666> CM: it's quirky as anything at times (gsmd that is)
Aug 30 07:42:06 <cw666> im tempted to start work on an alterantive
Aug 30 07:42:09 <jeddy3> CM, yes
Aug 30 07:42:21 <CM> trains: There's still a lot you can do using qemu
Aug 30 07:42:55 <jgm> cw666: writchie is looking to provide something similar, maybe talk to him
Aug 30 07:42:59 <trains> i could help with gsmd/libgsmd
Aug 30 07:43:27 <CM> jeddy3: Ok. Is it some good wiki-page how to do the network forwarding?
Aug 30 07:43:29 <cw666> jgm: similar? a replacement to gsmd?
Aug 30 07:43:30 * CM is lazy
Aug 30 07:43:39 <cw666> CM: network forwarding? you mean IP networking?
Aug 30 07:43:41 <trains> is there a list of open bugs for gsmd?
Aug 30 07:44:06 <CM> cw666: Yes, to easily use ipkg, or surf or whatever on the neo using teh usbnet
Aug 30 07:44:24 <cw666> CM: masquerade/SNAT on the host machine
Aug 30 07:44:25 <roh> CM you mean a /etc/network/interfaces sample which sets up iptables for you?
Aug 30 07:44:43 <CM> roh: Yes
Aug 30 07:44:50 <jgm> cw666: Yep.  gsmd seems to be okay as-is but is missing a bunch of things like arbitration, which is what I believe writchie is looking to provide as a wrapper library with a D-Bus interface
Aug 30 07:44:54 <roh> i have that... lets see where it would fit into the wiki
Aug 30 07:45:04 <cw666> jgm: ok, then i dont want to talk to him
Aug 30 07:45:09 <cw666> i loath dbus to the extreme
Aug 30 07:45:10 <CM> roh: Thanks, would be great. :)
Aug 30 07:45:18 <jeddy3> CM, i havent found one...i tried manually routing, but gave up and installed firestarter
Aug 30 07:45:33 <hads> I think there already is one
Aug 30 07:45:33 <cw666> i want some small and sane, not crippled with dbus/gnome/whatever
Aug 30 07:45:45 <jgm> cw666: that's a shame because an awful lot of the stuff being built is using D-Bus
Aug 30 07:45:50 <CM> jeddy3: I haven't managed either. I've installed shorewall, as I've used that before, but haven't gotten around to try the routing yet
Aug 30 07:45:57 <jgm> What would you recommend instead for a generic IPC mechanism?
Aug 30 07:46:16 <roh> CM its already there... but its buggy.. mompls
Aug 30 07:46:17 <cw666> jgm: im trying to get enough people interested it doing an entire alternate software stack right now
Aug 30 07:46:24 <jeddy3> works ok in firestarter, although it does not want to disable the firewall part of it, so i cannot mount desktop via nfs on neo when i route internet :p
Aug 30 07:46:33 <hads> cw666: Why?
Aug 30 07:46:55 <cw666> hads: because the current arch/design calls for a lot of large slow processes and whilst that's fine for a lot of people, it's not what im after
Aug 30 07:47:04 <jgm> cw666: well good luck with that, most people are focused on producing the current one first
Aug 30 07:47:21 <cw666> jgm: well, i have the advantage i can copy/steal what they have that workds
Aug 30 07:47:22 <cw666> works
Aug 30 07:47:31 * CM wishes that MediaWiki had a working search function
Aug 30 07:48:05 <jgm> cw666: and none of the other current 'phone OS activities fit your bill either?
Aug 30 07:48:22 <jeddy3> anyone know how to update kernel with ipkg?
Aug 30 07:48:45 <cw666> jgm: which ones are those?  i want something linux based and hackable, but i dont care for a cut-down gnome desktop
Aug 30 07:48:51 <cw666> there is a LOT in between
Aug 30 07:49:08 <cw666> one major problem i have doing this is i18n ... that's so hard to do and probably will bloat things no matter what
Aug 30 07:49:08 <CM> It's crazy that searching for USB or Network or both doesn't find USB_Networking
Aug 30 07:49:16 <jgm> You could always not start the desktop...
Aug 30 07:49:27 <jgm> ubuntu-mobile?
Aug 30 07:49:59 <cw666> im not sure, right now i just want to get some core daemons/libs to drive the hw and do PM for now
Aug 30 07:50:24 <jgm> Well if you do that for openmoko I'm sure that they will happily take your code and put it in the main tree
Aug 30 07:50:24 <cw666> the gui i honestly wont ever be able to do (i suck at guis) so unless i get help ill steal another one
Aug 30 07:50:53 <cw666> jgm: we'll see...  given im not that interested in openembedded, bitbake, etc. someone will have to integrate anything i do
Aug 30 07:51:26 <jgm> If you write a fully-functional power management daemon then I suspect someone else will go to that effort to get it integrated into their image
Aug 30 07:51:31 <hads> *yawn*
Aug 30 07:51:57 <cw666> jgm: it's a long way from that, kernel changes are needed first
Aug 30 07:52:15 <cw666> and also, gsmd might come a long way before i get much going, so it would be silly not to use that or at least steal from that
Aug 30 07:52:34 <hads> Or perhaps... even... contribute to that.
Aug 30 07:53:14 <cw666> maybe, the code isn't too bad at all
Aug 30 07:53:15 <FuzzyCat> *blink*
Aug 30 07:54:50 <roh> CM https://wiki.openmoko.org/wiki/USB_Networking#Automatic_method
Aug 30 07:55:03 <roh> CM i just fixed the ubuntu part...
Aug 30 07:55:11 <CM> roh: Thanks :D
Aug 30 07:57:30 <roh> cw666 i'd recommend you read about dbus first and spare youre ranting till afterwards. actually it really fixes a lot of the ipc problems out there quite well and _uniform_ while still keeping dependencys down at a absolute minimum. i quite like it.
Aug 30 07:57:32 * ScaredyCat stabs roh
Aug 30 07:57:52 * CM likes dbus too
Aug 30 07:58:10 * hads like collaboration rather than duplicated effort
Aug 30 07:58:12 <jgm> After using a lot of other IPC methods out there D-Bus does seem to balance functionality and simplicity quite well
Aug 30 07:58:14 <jeddy3> roh, sweet!
Aug 30 07:58:17 * ScaredyCat prefers dtrain
Aug 30 07:58:29 <jeddy3> x)
Aug 30 07:58:44 <roh> compared to 20 people trying to break their fingers trying to get unix sockets right its sane... especially if you keep multiple clients and restarting daemons in mind. (reconnect problems, etc)
Aug 30 07:58:55 <jgm> Although at the end of the day I'll use whatever is 'standard' for a given system, so for openmoko D-Bus it is
Aug 30 07:59:21 <cw666> roh: you're missing the point, but that's ok ... the point of open access to the device is that i get to choose what i like nad you get to use what you like
Aug 30 07:59:47 <roh> jgm we'll see. actually we need to use it since even bluez depends on it, but harald does not quite like it somehow.
Aug 30 08:00:16 <jgm> cw666: I think the point is more that if you're developing as a community you need to pick something as a community and stick to it
Aug 30 08:00:33 <jgm> roh: moving away from D-Bus at this stage would be a very bad idea
Aug 30 08:00:36 <roh> cw666 nobody hinders you. but keep in mind that throwning rocks in the way of people they will stumble
Aug 30 08:00:39 <cw666> jgm: im not developing as a community though, ill do what i can to solve my needs and others are welcome to help
Aug 30 08:00:51 <jgm> cw666: yeah I noticed that...
Aug 30 08:00:55 <Placid> morning
Aug 30 08:01:51 <roh> don't get me wrong here. i don't wanna push dbus or anything. there are far too many people who think it will solve all their problems. i just see some opportunitys here and there and we will see how it performs and works.
Aug 30 08:02:18 <roh> in the end thats what opensource is about. evolutionary development with a lot of trial and error ;)
Aug 30 08:02:38 <cw666> roh: dbus might be fine, it's the least of my worries right now
Aug 30 08:03:00 <cw666> the total impact of dbus is pretty small compared to the footprint and requirements of some other components
Aug 30 08:03:12 <cw666> but for initial developement im not that excited about it
Aug 30 08:04:59 <jeddy3> roh, reisty autoconfiguration works perfectly (after patching bug in ifupdown udev rule)
Aug 30 08:05:16 <jeddy3> roh, thanks nice work
Aug 30 08:05:41 <cw666> roh: when you get a chance, try "LD_TRACE_LOADED_OBJECTS=1 /usr/bin/openmoko-dialer"
Aug 30 08:05:49 <roh> jeddy3 i still need to figure out some way when using more than pone phone ;)
Aug 30 08:06:04 <roh> cw666 sounds scary
Aug 30 08:06:12 <jeddy3> roh, ah :)
Aug 30 08:06:34 <roh> cw666 mompls.. need to boot the phone
Aug 30 08:07:36 <roh> jeddy3 i also have a patch in my mtn repo that adds a resolv.conf to images i build which points to the 192.168.0.200 where i run a dnscache
Aug 30 08:07:55 <jeddy3> roh, nice!
Aug 30 08:08:40 * ScaredyCat used dnrd
Aug 30 08:10:36 <rwhitby> roh: we should make that standard - something that works for smart developers is better than a lack of resolv.conf which works for no-one.
Aug 30 08:13:14 <rwhitby> roh: especially because at home I plug the neo into my router's usb port to charge and get on the network :-)
Aug 30 08:15:09 * ScaredyCat points rwhitby at the wiki 
Aug 30 08:15:11 <ScaredyCat> ;)
Aug 30 08:16:13 <jeddy3> ok, if an update installed all this 2.6.22-moko5 modules, that must mean i can flash moko with a 2.6.22-moko5-uimage, and it will work?
Aug 30 08:16:36 <rwhitby> ScaredyCat: why the wiki?  I'm strongly considering just changing it in the OE tree.
Aug 30 08:16:47 <rwhitby> jeddy3: should do
Aug 30 08:17:40 <ScaredyCat> but then you'll break other stuff that uses oe no?
Aug 30 08:18:05 <ScaredyCat> isn't this what we get bitched at for doing?
Aug 30 08:21:43 <jeddy3> hmm, whats the difference between a uImage and a zImage on buildhost?
Aug 30 08:23:55 <XorA> uImage is for uboot, zImage is propber kernel image
Aug 30 08:24:12 <jeddy3> XorA, come again?
Aug 30 08:24:22 <jeddy3> XorA, it's uimage i wan't right?
Aug 30 08:24:34 <XorA> jeddy3: use uImage to flash, use zImage if you understand the difference :-)
Aug 30 08:24:53 <jeddy3> uimage is some kind of "patched" zimage for uboot?
Aug 30 08:24:55 <rwhitby> ScaredyCat: no, you can do stuff in OE on an image or distro basis. 
Aug 30 08:24:58 <jeddy3> XorA, nod
Aug 30 08:25:05 <rwhitby> in this case it would be for the openmoko distro or openmoko image (adding an /etc/resolv.conf which matches the default usb networking setup)
Aug 30 08:25:09 <XorA> jeddy3: uImage has some extra header for uboot added
Aug 30 08:25:15 <cw666> SpeedEvil: about still?
Aug 30 08:25:17 <jeddy3> XorA, got it, thanks
Aug 30 08:25:24 <cw666> i
Aug 30 08:25:28 <cw666> oops
Aug 30 08:32:45 <roh> rwhitby i'll put that on my list for later
Aug 30 08:33:10 <squalyl> hi all
Aug 30 08:33:48 <roh> rwhitby but since i also not yet have a nice idea how to make it detachable and usable via pan without having 2 complete set of configs... 
Aug 30 08:34:09 <roh> eventually just bridge the opan and the usbnet together and put the l3-ip on br0
Aug 30 08:34:39 <rwhitby> with udev rules to add or remove interfaces from the bridge
Aug 30 08:34:52 <cw666> roh: oan?
Aug 30 08:34:57 <cw666> err, opan?
Aug 30 08:35:21 <rwhitby> he probably means bnep
Aug 30 08:36:06 <cw666> ive not tested that yet;  people have it working ok?
Aug 30 08:38:15 <squalyl> OM2007.2 doesn't build here due to do_compile with git-native. Is this "normal" ?
Aug 30 08:38:32 <cw666> squalyl: it built this morning for me
Aug 30 08:38:49 <cw666> is your tree older than that or more recent?
Aug 30 08:39:16 <cw666> squalyl: you have git-core, monotone, etc installed?
Aug 30 08:39:38 <squalyl> cw666: I have all installed
Aug 30 08:40:22 <squalyl> http://pastebin.org/1409
Aug 30 08:40:51 <squalyl> and I just pulled the last OE.mtn and ran mtn update
Aug 30 08:41:16 <roh> cw666 pand/bnep yep bnep0 and usb0 on a bridge should work *hmm*
Aug 30 08:43:22 <squalyl> cw666: maybe I lack a specific perl module?
Aug 30 08:44:22 <cw666> squalyl: im not sure, im doing an update and rebuild right now
Aug 30 08:45:07 <squalyl> I just tried to install CPAN which downloaded a dependency "ExtUtils::MakeMaker" this may help me ;)
Aug 30 08:47:13 <cw666> squalyl: svn.berlios.de is down right now so i can't update to retest
Aug 30 08:47:33 <squalyl> Don't worry i fixed the issue
Aug 30 08:47:51 <squalyl> perl module ExtUtils::MakeMaker is needed to build git-native
Aug 30 08:48:05 <jas4711> cw666: 'make -k update' works here, it skips updating bitbake but probably it hasn't changed
Aug 30 08:48:37 <jas4711> although my build stops when building mtd... at least a few hours ago
Aug 30 08:48:44 <squalyl> I'm putting a small note in [[Talk:OpenMoko2007.2]]
Aug 30 09:02:03 <aloril> (script) wiki RecentChanges: [[Web_Browser]] [[Talk:OpenMoko2007.2]] [[User:ThomasWood]] [[User:Javiroman]] [[Press_Coverage]] [[User_talk:Emdete]] [[USB_Networking]] [[User_talk:HaraldWelte]]
Aug 30 09:02:58 <florian> good morning
Aug 30 09:31:27 <roh> wtf?
Aug 30 09:31:32 <roh> http://www.engadget.com/2007/08/29/nokias-iphone-no-seriously/
Aug 30 09:31:38 <roh> muhhahha
Aug 30 09:32:02 <jeddy3> yeah, thats pretty sick
Aug 30 09:32:18 <jeddy3> worse than any chinese ripoff e
Aug 30 09:32:25 <jeddy3> -e
Aug 30 09:37:57 <ynezz> it's just a joke, isn't it? :p
Aug 30 09:38:21 <Tyrael> hi,  my neo don't boot on power up or reboot. It is blocked on the splash screen.  I need to power up with aux button and then select boot.
Aug 30 09:38:32 <Tyrael> doesn't (sorry)
Aug 30 09:39:07 <torpor> morning folks
Aug 30 09:39:11 <torpor> morning folks
Aug 30 09:39:41 <ynezz> Tyrael: have you erased your rootfs before flashing the new one?
Aug 30 09:41:16 <Tyrael> ynezz: this time no. but before flashing the previuos one yes (that was38,x MB). And this was 39,xMB)
Aug 30 09:42:07 <Tyrael> ah, the UBOOT splash screen, don't the "kernel spash screen" 
Aug 30 09:42:23 <jeddy3> Tyrael, i reasoned so too at last update, but it still borked
Aug 30 09:42:50 <jeddy3> Tyrael, new image when i flashed was slightly larger, still it borked
Aug 30 09:44:05 <Tyrael> jeddy3: ah ok. Now I erase kernel and rootfs
Aug 30 09:46:25 <torpor> whatsup?
Aug 30 09:46:25 <torpor> svn: Can't connect to host 'svn.berlios.de': Connection refused
Aug 30 09:47:14 <Tyrael> wow....  nand erase clean kernel made a lot of ouput....
Aug 30 09:49:02 <trains> I need a little help understanding something about OM bugzilla. See this bug http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=107 . It has 2 comments with patches attached that supposedly solve the problem or part of it, but the bug is still marked as NEW.  
Aug 30 09:51:58 <CM> Tyrael: I have that problem all the time
Aug 30 09:52:21 <ynezz> trains: that just means that either responsible person don't care, or have no time
Aug 30 09:52:28 <ynezz> no more, no less
Aug 30 09:52:58 <trains> so, can i apply the patch?
Aug 30 09:53:17 <ynezz> right, you can do whatever you want
Aug 30 09:54:08 <trains> ? they don't have the time to mark the bug as solved or they don't have the time to submit the patch?
Aug 30 09:56:36 <wibbit> trains: I would assume, they don't have time to test the patch, to make sure it works appropriately, and then apply
Aug 30 09:56:45 <trains> ok
Aug 30 09:57:17 <wibbit> though, that is an assumption.
Aug 30 09:57:34 <trains> but, can i do that? what do i need to submit patches?
Aug 30 09:57:55 <wibbit> But I can believe there is a fair amount of work to take suggested patches, apply, ensure it does what is expected, and then update the bugzilla + regression testing etc
Aug 30 09:58:09 <wibbit> To submit or apply
Aug 30 09:58:35 <wibbit> The people who have suggested patches, I would assume have already "submitted" them, to apply them, you'd need SVN/CVS access
Aug 30 09:59:04 <wibbit> correction, svn/cvs commit access
Aug 30 09:59:23 <wibbit> but, I do not know what openmoko's policies are on testing patches, regression etc
Aug 30 10:00:21 <cb22> So, should I buy a P1 device now on eBay, or wait and buy the P2?
Aug 30 10:00:37 <wibbit> cb22: depends what you want it for :)
Aug 30 10:00:46 <aloril> (script) wiki RecentChanges: [[MacOS_X]] [[User:ThomasWood]] [[Talk:MacOS_X]]
Aug 30 10:00:54 <cb22> wibbit, i dont know myself :)
Aug 30 10:01:05 <wibbit> cb22: however, I got the impression that gps usage, may be problematic with the P1's due to a hardware driver issue.
Aug 30 10:01:10 <cb22> i'm decent with dev / linux stuff, so an incomplete OS is no biggy
Aug 30 10:01:29 <zash_se> P[12] s/P/GTA0 ?
Aug 30 10:01:43 <cb22> hmm, you see i also really wanted wifi
Aug 30 10:02:12 <cb22> but, if the GTA02 (correct?) comes out later then october, i'd rather just get a GTA01
Aug 30 10:02:25 <wibbit> I've seen comment, expecting it in jan
Aug 30 10:02:31 <cjb_ie> yesterday somebody said GTA02 probably jan
Aug 30 10:02:33 <cb22> at the moment i'm without a phone you see
Aug 30 10:02:38 <cjb_ie> wibbit: damn your nimble fingers
Aug 30 10:02:44 <wibbit> hee hee
Aug 30 10:02:50 <cb22> hmm, thanks guys
Aug 30 10:03:17 <cb22> i supose i could always just use a usb wifi stick!
Aug 30 10:03:26 <cb22> if i could get a power source
Aug 30 10:03:35 <cb22> and when the switching gets implimented
Aug 30 10:04:10 <cb22> but, wibbit, what are these hardware GPS errors you speak of?
Aug 30 10:04:49 <ynezz> it's not usable
Aug 30 10:05:45 <cb22> due to the binary gps not being provided, or other issues?
Aug 30 10:07:16 <wibbit> cb22: I don't know the whole story, so some one, please correct me.
Aug 30 10:07:46 <wibbit> cb22: The people that were originally providing the GPS hardware, have been bought by another company, the other company does not wish to honour the original agreement to provide binary drivers.
Aug 30 10:08:04 <wibbit> cb22: but as I said, I may not (probably not) know the whole story
Aug 30 10:08:10 <cb22> hmph
Aug 30 10:08:33 <cb22> oh well, 50 people still have the binary driver, hopefully they will be able to RE something
Aug 30 10:09:07 <wibbit> as many have stated, that may take a while
Aug 30 10:09:40 <cb22> yeah, true, but atleast it is a ray of hope :)
Aug 30 10:10:10 <wibbit> Yup, like hoping, whilst one hurtles towards the sun, that it a big wind may come by and blow it out :)
Aug 30 10:10:18 <cb22> lol
Aug 30 10:10:24 <cjb_ie> how big is the binary?
Aug 30 10:10:41 <cb22> although im suprized it hasnt been leaked futher
Aug 30 10:10:56 <cb22> those phase 0 people must be good boys
Aug 30 10:11:11 <mjr> wibbit, hm, I didn't have the impression that it was a lost cause to get the binary drivers released, merely a matter of (*sigh*) more negotiation
Aug 30 10:11:13 <Dmitry_Platonov> GTA02 will come with different GPS chip
Aug 30 10:11:14 * aloril mumbles something about searching for irc logs
Aug 30 10:11:22 <mjr> but of course I don't have the full story either
Aug 30 10:11:33 <XorA> cb22: most of us P0 guys didnt know it was there and destroyed it
Aug 30 10:12:04 <cjb_ie> XorA: as in, it was written in by default and got overwritten?
Aug 30 10:12:06 * mjr happens to have an illegal copy of gllin (ssh)
Aug 30 10:12:07 <wibbit> mjr: The fact that, as Dmitry_Platonov states, GTA02, will come with a different GPS chip (and I assume from another manufacturer), the whole negotiation thing will probably fail.
Aug 30 10:12:23 <XorA> cjb_ie: yeah, first thing I did was flash phone, didnt even look at what was there
Aug 30 10:12:37 <cjb_ie> any idea why the new company doesn't want to honour the agreement?
Aug 30 10:12:38 <mjr> wibbit, that's jumping to conclusions
Aug 30 10:12:45 <wibbit> mjr: yup
Aug 30 10:13:03 <cjb_ie> XorA: ah, the gps chipset doesn't have its own firmware area?
Aug 30 10:13:40 <Dmitry_Platonov> as I understand, gllin will not work on 2007.2
Aug 30 10:13:57 <XorA> Dmitry_Platonov: only in a OABI chroot
Aug 30 10:14:20 <cjb_ie> s/firmware area/flash/
Aug 30 10:14:20 <mjr> there are plenty other reasons for wanting a new chip; for example, simply the fact that it's been so difficult so far with the Hammerhead, and maybe they don't want to deliver more chips at the current terms, or maybe (just maybe ;) they just want rid of the non-free userspace driver
Aug 30 10:14:20 <apt> cjb_ie meant: XorA: ah, the gps chipset doesn't have its own flash?
Aug 30 10:15:06 <mjr> so I really wouldn't yet assume that there won't be a gllin for GTA01, though it's a bit discouraging certainly that it's not already here
Aug 30 10:15:11 <cjb_ie> the sirf iii is well regarded in userland, too. lots of openstreetmap folks swear by it, and none of them's heard of hammerhead.
Aug 30 10:15:49 <ewon> sirf_iii++;
Aug 30 10:15:58 <ewon> nice little chip. gets locks indoors.
Aug 30 10:16:07 <cb22> downside?
Aug 30 10:16:24 <ewon> maybe it eats more power than the hammerhead? no real idea.
Aug 30 10:16:28 <wibbit> ewon: that, sounds VERY interesting
Aug 30 10:16:37 <roh> ewon nope. no sirf
Aug 30 10:16:44 <mjr> I think roh said it's not sir... right ;)
Aug 30 10:16:49 <ewon> wibbit: by indoors I mean a standard semi detached, no idea about a corporate office
Aug 30 10:16:59 <wibbit> That's still impressive
Aug 30 10:17:12 <wibbit> increase the likelyhood of getting a lock along the streets of london
Aug 30 10:17:16 <ewon> I've a couple of sirf-iii devices
Aug 30 10:17:19 <Dmitry_Platonov> did you read hammerhead vs sirf III comparison? hammerhead did slightly beeter
Aug 30 10:17:44 <ewon> Dmitry_Platonov: impressive
Aug 30 10:18:04 <Dmitry_Platonov> I bet new chip would be on same level
Aug 30 10:19:55 <cjb_ie> ewon: i got a streetpilot c550 last night, its sirf3 is a _lot_ better at picking up signal than my holux sirf3 usb dongle
Aug 30 10:20:09 <ewon> cjb_ie: odd. better antenna?
Aug 30 10:20:34 <cjb_ie> ewon: i suspect it's got more amplification, less constrained power budget
Aug 30 10:20:55 <ewon> cjb_ie: wait a sec, ZOMG, you but a closed-source mapping solution? doesn't this mean the other openstreetmap people have to cut your balls off?
Aug 30 10:21:02 <ewon> s/but/bought/
Aug 30 10:21:02 <apt> ewon meant: cjb_ie: wait a sec, ZOMG, you bought a closed-source mapping solution? doesn't this mean the other openstreetmap people have to cut your balls off?
Aug 30 10:21:38 <cjb_ie> my place is low down, on the north side of a tall building, with only north-facing windows; and the streetpilot got a fix, and only got unhappy when i had it sitting in front of the laptop and i was crouching over it.
Aug 30 10:22:16 <cjb_ie> ewon: :p
Aug 30 10:22:52 <cjb_ie> the guy's letting me have it for half what he paid for it - and that was already a reduced price. can't argue with that.
Aug 30 10:22:57 <ewon> heh
Aug 30 10:23:05 <roh> if youve not seen it already... i can advice haralds talk at the camp
Aug 30 10:23:24 <roh> the sound is totally crappy, but its worth it http://people.openmoko.org/ninjutsu/Haraldccc07.mpg
Aug 30 10:40:01 <ynezz> roh: it's in english?
Aug 30 10:40:51 <CIA-20> openmoko: 03zecke * r2856 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (ChangeLog libmokoui/moko-search-bar.h): 
Aug 30 10:40:51 <CIA-20> openmoko: 2007-08-30 Holger Hans Peter Freyther <zecke@selfish.org>
Aug 30 10:40:51 <CIA-20> openmoko:  Change the header file to be parsable by h2def.py, which will
Aug 30 10:40:51 <CIA-20> openmoko:  reduce the manual work needed to create bindings.
Aug 30 10:40:51 <CIA-20> openmoko:  * libmokoui/moko-search-bar.h:
Aug 30 10:41:24 <mmazur> roh, if someone took the sound and filtered it out or sth.
Aug 30 10:41:30 <mmazur> roh, I can hardly understand anything.
Aug 30 10:45:38 <roh> mmazur does not help. the decompression-recompression alone would loose more than the filter gains. so you need to filter it at playback
Aug 30 10:50:56 <zecke> ~seen Cwiiis
Aug 30 10:50:58 <apt> cwiiis is currently on #maemo (13h 12m 33s) #oe (13h 12m 33s), last said: 'zecke_: You're writing an HTML rendering engine?'.
Aug 30 11:01:01 * FuzzyCat stabs berlios.de
Aug 30 11:01:49 <wibbit> ~seen wibbit
Aug 30 11:01:50 <apt> wibbit is currently on #openmoko (1h 34m 28s). Has said a total of 22 messages. Is idling for 1s, last said: '~seen wibbit'.
Aug 30 11:02:34 * FuzzyCat hands wibbit  a mirror
Aug 30 11:03:12 * wibbit admires the intricate lines appearing in the mirror as it starts to crack
Aug 30 11:03:16 <raster> meep
Aug 30 11:41:14 <CM> roh: Heh, the audio sure is noisy.. People sawing in there and everything :D
Aug 30 11:45:23 <CIA-20> openmoko: 03chris * r2857 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (3 files in 2 dirs): MokoSearchBar: Add new API to count and clear categories
Aug 30 11:50:46 <jonathan_br_ba> guyz i have that problem, i think is calibration, how i fix it?
Aug 30 11:51:12 <jonathan_br_ba> i cant load terminal
Aug 30 11:51:24 <jgm> jonathan_br_ba: can you log in over ssh?
Aug 30 11:51:26 <SpeedEvil> can you ssh in?
Aug 30 11:51:35 <SpeedEvil> oops, gtg
Aug 30 11:51:43 <jonathan_br_ba> i dont know
Aug 30 11:51:51 <jonathan_br_ba> how i do this?
Aug 30 11:52:29 <jgm> Start with https://wiki.openmoko.org/wiki/USB_Networking
Aug 30 11:53:33 <_diego_> i'm using OM 2007.08, and when i try to lounch the input manager, it starts calculator 
Aug 30 11:54:09 <jgm> _diego_: yeah it seems like the app switcher doesn't work out that you've changed category.  Try changing your category to 'All' and pick it from there
Aug 30 11:57:00 <_diego_> i open 2 input manager, how can i close one?
Aug 30 11:57:04 <nnpiggy> Could anybody tell me off-hand how to get control characters on the on-screen keyboard?
Aug 30 11:57:16 <jgm> No idea (for either of the above)
Aug 30 11:57:57 <DPThought> nnpiggy: what do you mean by control characters? command+XX combos?
Aug 30 11:58:16 <nnpiggy> ctl-p, ctl-e, etc...
Aug 30 11:58:28 <nnpiggy> Not the ALT characters...
Aug 30 11:58:53 <_diego_> use the full keyboard on the display
Aug 30 11:59:00 <CM> _diego_: That's a pain, but basically, run ps -u root to find the pid of matchbox-keyboard and then kill it
Aug 30 11:59:19 <CM> nnpiggy: The second kb from the drop down menu, as _diego_ said
Aug 30 11:59:54 <DPThought> nnpiggy: if you have launched input manager. the right side of the white blob is a dropdown. there you can selecte a more sophisticated keyboard with command, alt and all the keys you need
Aug 30 12:00:11 <nnpiggy> oh ho! Thanks!
Aug 30 12:00:48 <nnpiggy> Yikes! I have both keyboards now!
Aug 30 12:00:48 <jonathan_br_ba> jgm: i tryid but usb0 is not exists
Aug 30 12:01:35 <_diego_> nnpiggy: welcome to the club ^^
Aug 30 12:02:41 <jonathan_br_ba> guys help me ssh my qemu?
Aug 30 12:02:50 <jonathan_br_ba> i tried with tutorial but
Aug 30 12:03:00 <jonathan_br_ba> my terminal says that usb0 dont exists
Aug 30 12:03:45 <_diego_> do you have usb net support? check the kernel 
Aug 30 12:04:03 <nnpiggy> Thanks, _diego_. Brekkie time.
Aug 30 12:04:45 <jonathan_br_ba> _diego_: man xcuse but, how i check?
Aug 30 12:05:01 <_diego_> nnpiggy: btw if u have ssh, look for a process mbinputmgr and kill it ^^
Aug 30 12:07:17 <_diego_> jonathan_br_ba: lsmod. You should have cdc_ether and usbnet loaded
Aug 30 12:07:27 <jonathan_br_ba> ok
Aug 30 12:09:08 <jonathan_br_ba> _diego_: i didnt have but i load it with modprobe
Aug 30 12:09:40 <_diego_> jonathan_br_ba: retry now
Aug 30 12:11:14 <jonathan_br_ba> _diego_: i tried: ifconfig usb0 192.168.0.200 netmask 255.255.255.0 | returns -> usb0: ERROR while getting interface flags:
Aug 30 12:11:58 <torpor> anyone know why /media/card is no longer working?
Aug 30 12:12:11 <torpor> 2.6.21.6-moko11
Aug 30 12:12:51 <aloril> torpor: tried taking battery out and rebooting?
Aug 30 12:14:11 <CIA-20> openmoko: 03laforge * r2858 10/trunk/src/target/kernel/patches/ (gta02-core.patch s3c2442b-cpuid.patch series): implement core kernel support for GTA02v2 (S3C2442B)
Aug 30 12:14:19 <roh> torpor because you use a oild build which still has bugs with sdcards. fixed some days now
Aug 30 12:15:15 <torpor> yeah
Aug 30 12:15:25 <torpor> i'll try the new build then.
Aug 30 12:15:26 <torpor> thx
Aug 30 12:15:26 <_diego_> jonathan_br_ba: try to see if usb is recognized. Look at lsusb and dmesg, if u see something interesting
Aug 30 12:20:08 <CIA-20> openmoko: 03laforge * r2859 10/trunk/src/target/kernel/patches/smedia-glamo.patch: 
Aug 30 12:20:08 <CIA-20> openmoko: glamo3362 driver:
Aug 30 12:20:08 <CIA-20> openmoko: * temporarily deactivate irq_chip demultiplexer code
Aug 30 12:20:08 <CIA-20> openmoko: * add gta02v2 support (pull LCM out of reset before using it)
Aug 30 12:20:08 <CIA-20> openmoko: * add incomplete hardware accelerated cursor support (currently inactive)
Aug 30 12:22:15 <nnpiggy> Thanks, _diego_--I killed it.
Aug 30 12:45:57 <FuzzyCat> roh, sd card not nixed
Aug 30 12:46:17 <FuzzyCat> fixed
Aug 30 12:47:58 <roh> FuzzyCat for me it is.. ive not seen a single ioerror since
Aug 30 12:48:16 <FuzzyCat> with lots of files
Aug 30 12:48:18 <FuzzyCat> ?
Aug 30 12:48:53 <FuzzyCat> try with >100mb
Aug 30 12:48:55 <roh> FuzzyCat ive copied a mp3 collection up on the card multiple times.. failed consistently all the time after 2 ord 3 files
Aug 30 12:49:16 <roh> FuzzyCat ive used atleast 400mb per test
Aug 30 12:49:30 <FuzzyCat> mmmm... I 
Aug 30 12:49:51 * FuzzyCat reflashes just in case
Aug 30 12:50:46 <roh> FuzzyCat svn rev 2831/2832 is the patch in question
Aug 30 12:50:50 <SpeedEvil> roh: Hack. scp -l 1000 * phone:/media/card/mp3/
Aug 30 12:51:02 <SpeedEvil> though updating is the right way
Aug 30 12:51:05 <roh> and no.. nobody bumped the bitbake-revision
Aug 30 12:51:08 <SpeedEvil> to limit bandwidth
Aug 30 12:51:36 <roh> Speeei do not want to linit the bandwith. i want to provoke the error
Aug 30 12:52:20 <SpeedEvil> dd if=/dev/zero works well for that
Aug 30 13:01:41 <aloril> (script) wiki RecentChanges: [[Main_Page/ar]] [[Template:Languages/ar]] [[Media_Content/ar]] [[Media_Content]] [[Template:Improve/ar]] [[Template:Unimplemented/ar]] [[Translation_HOWTO/ar]] [[Translation_HOWTO]]
Aug 30 13:02:23 <FuzzyCat> gah scrollback tooo short
Aug 30 13:02:34 <FuzzyCat> mtd-utils.... ?
Aug 30 13:07:01 <FuzzyCat> isn't that a path error?
Aug 30 13:17:57 * FuzzyCat elbows CM
Aug 30 13:18:33 <mdt> strange: in /etc/fstab /tmp is a tmpfs but /etc/init.d/bootmisc.sh wants to make a symlink for it.... hm
Aug 30 13:20:35 <milo_> since i had that errpr with scp 'scp: /media/card/movies: No space left on device' the card doesnt mount anymore. still not at reboot :(
Aug 30 13:21:05 <mdt> milo_, do a mkdosfs?
Aug 30 13:21:13 <FuzzyCat> Error: Too much pr0n on device.
Aug 30 13:21:19 <milo_> when i check for the space left with 'df -h' it seams not mounted
Aug 30 13:23:07 <mdt> milo_, did you try to mount it?
Aug 30 13:23:34 <milo_> should i rm /media/card and reboot? where is the mounting logged?
Aug 30 13:23:49 <milo_> mdt: no i didnt. how?
Aug 30 13:24:43 <mdt> mount /dev/mmcblk0p1 /media/mmcblk0p1
Aug 30 13:25:01 <mdt> you will see the message while the mount failed... or it mounts it
Aug 30 13:25:49 <milo_> "mount: mount point /media/mmcblk0p1 does not exist"
Aug 30 13:26:14 <mdt> milo_, so just do mount /dev/mmcblk0p1 /media
Aug 30 13:26:52 <milo_> "unknown filesystem type 'vfat'"
Aug 30 13:27:03 <mdt> so the kernel module is missing
Aug 30 13:27:18 <mdt> do a ipkg list_installed|grep fat
Aug 30 13:27:28 <milo_> ah ok. but it worked since 10 min ago
Aug 30 13:27:40 <mdt> :D
Aug 30 13:27:48 <milo_> "kernel-module-fat - 2.6.22.5-moko11-r1 - fat kernel module
Aug 30 13:27:50 <milo_> kernel-module-vfat - 2.6.22.5-moko11-r1 - vfat kernel module; VFAT filesystem support"
Aug 30 13:27:52 <mdt> 10 minutes are a lot, you have 266MHz....
Aug 30 13:27:54 <mdt> fine
Aug 30 13:28:13 <mdt> what does `lsmod|grep fat` say?
Aug 30 13:28:23 <milo_> i tried to mount it on a different host and it worked
Aug 30 13:28:42 <mdt> 'different host' means what?
Aug 30 13:28:50 <milo_> lsmod | grep fat return nothing. how can i load it
Aug 30 13:29:01 <mdt> modprobe vfat
Aug 30 13:29:02 <milo_> another machine.. gentoo
Aug 30 13:29:07 <mdt> milo_, :D
Aug 30 13:29:22 <mdt> milo_, different machine has different behavior... shure
Aug 30 13:29:30 <milo_> "WARNING: Could not open '/lib/modules/2.6.21.6-moko11/kernel/fs/nls/nls_base.ko': No such file or directory
Aug 30 13:29:33 <milo_> WARNING: Could not open '/lib/modules/2.6.21.6-moko11/kernel/fs/fat/fat.ko': No such file or directory
Aug 30 13:29:37 <milo_> FATAL: Could not open '/lib/modules/2.6.21.6-moko11/kernel/fs/vfat/vfat.ko': No such file or directory"
Aug 30 13:29:56 <milo_> so maybe reinstall the modules?
Aug 30 13:30:06 <mdt> oups
Aug 30 13:30:10 <mdt> do a uname -a
Aug 30 13:30:10 <torpor> you upgraded the filesystem but not the kernel.
Aug 30 13:30:17 <mdt> yes looks like :)
Aug 30 13:30:18 <torpor> <-- has the same problem.
Aug 30 13:30:34 <milo_> "Linux fic-gta01 2.6.21.6-moko11 #1 PREEMPT Sat Aug 25 20:36:00 CEST 2007 armv4tl unknown"
Aug 30 13:30:59 <mdt> you have to loook for a kernel image with version 2.6.22.5-moko11-r1....
Aug 30 13:31:08 <torpor> actually there are no such modules.
Aug 30 13:31:24 <mdt> your modules are 2.6.22, your kernel 2.6.21
Aug 30 13:31:47 <mdt> another hint that the openmoko developer did a very good job: you still can login...! :D
Aug 30 13:31:57 <milo_> oh ok. i cant update the kernel with ipkg? ok so i ll flash the ne kernel thanks
Aug 30 13:32:25 <mdt> yes... interesting, a new kernel exists. good to know :D
Aug 30 13:34:07 <torpor> what the URL for the 'latest ' stuff from openmoko?
Aug 30 13:34:14 <torpor> i never know where to go to get the latest builds .. 
Aug 30 13:35:09 <mdt> same to me.... i try http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/ipk/armv4t
Aug 30 13:35:33 <mdt> but i build most packages myself...
Aug 30 13:35:40 <milo_> i think i did the ipkg upgrade and it installed all latest stuff which need the ne kernel. right? mh :( good to know.. ipkg is lovely but this sucked
Aug 30 13:36:44 <torpor> mdt: thats just for package s.. i'm talking about images for burning ..
Aug 30 13:37:10 <mdt> here are the new modules... http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/ipk/fic-gta01/, http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/images/fic-gta01/ are the images
Aug 30 13:37:46 * FuzzyCat slaps mdt
Aug 30 13:37:51 <FuzzyCat> -utils
Aug 30 13:38:32 <mdt> FuzzyCat, ?
Aug 30 13:39:33 <torpor> how do you use th emodules packages listed?
Aug 30 13:39:40 <torpor> manually copy over and untar?  or what?
Aug 30 13:39:55 <torpor> .e.g this one:
Aug 30 13:39:55 <torpor> http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/images/fic-gta01/modules-2.6.22.5-moko11-moko11-r2-fic-gta01.tgz
Aug 30 13:40:06 <torpor> no auto-install method?  requires manual intervention or what?
Aug 30 13:40:12 <FuzzyCat> mdt-utils not you mdt :)
Aug 30 13:40:24 <FuzzyCat> seems that it's busted
Aug 30 13:40:57 <FuzzyCat> fatal: not a git repository - basically 404
Aug 30 13:41:36 <Fatal> FuzzyCat: If you try to push something into me I will be unpleased
Aug 30 13:42:11 <FuzzyCat> :D
Aug 30 13:47:47 <mdt> anyone an idea about my /tmp problem?
Aug 30 13:47:58 <torpor> what /tmp problem?
Aug 30 13:49:07 <mdt>  /tmp is mounted tmpfs by /etc/fstab. /etc/init.d/bootmisc.sh tries to remove/symlink that to /var/tmp. /var/tmp is symlinked to /var/volatile/tmp...
Aug 30 13:51:26 <mwester> mdt: what's the problem?
Aug 30 13:54:28 <mdt> mwester, as i described. the removal & symlinking of /tmp failes because it's mounted.
Aug 30 13:54:43 <mdt> mwester, look into /tmp, you will find a /tmp/tmp symlinked to /var/tmp
Aug 30 13:55:20 <mwester> Is this in the standard openmoko image then?
Aug 30 13:55:27 <mdt> yes
Aug 30 13:55:33 <mwester> Ah, ok. 
Aug 30 13:56:22 <mwester> I guess someone should fix that then.  I think the idea is that /var/volatile is tmpfs, and the mounting of /tmp on tmpfs is unnecessary.
Aug 30 13:56:36 <mdt> yes, i thinks so
Aug 30 13:56:49 <mdt> i fixed it this way in my image.
Aug 30 13:57:06 <mdt> tmpfs is also mounted to /media/ram... what is that for?
Aug 30 13:57:58 <mwester> That's been a point of discussion on other OE platforms too.  I personally find it unnecessary, but I think it's for historical reasons.
Aug 30 13:58:41 <mwester> It doesn't hurt anything, apparently -- the resources required to keep that mounted are very small (unless one actually uses that tmpfs space)
Aug 30 13:59:21 * mwester notices the time and ducks out to a meeting...
Aug 30 13:59:28 <milo_> mdt: flashed uImage-2.6.22.5-moko11-r2-fic-gta01.bin and it works when i mount mmcblk0p1 by hand. but its not loaded at boot
Aug 30 14:02:00 <mdt> milo_, ...hm, don't know how automount is handled in openmoko. fstab uses /media/card but has noauto flag. 
Aug 30 14:04:36 <milo_> thanks works again.. mplayer is fun :) even with gta01's slow cpu
Aug 30 14:04:55 * milo_ dreams of a giant battery pack :D
Aug 30 14:05:08 <mdt> milo_, yes, libmad is a briliant project...
Aug 30 14:09:22 <ScaredyCat> has anyone tried a build recently ?
Aug 30 14:10:08 <thomasg> I did a build before about 16 hours
Aug 30 14:11:07 <ScaredyCat> seems to be looking for a mdt-utils git revision that doesn;t exist
Aug 30 14:11:14 <thomasg> yes
Aug 30 14:11:20 <thomasg> here's the fix:
Aug 30 14:11:38 <thomasg> hier ist der patch: http://bugs.openembedded.org/attachment.cgi?id=1776
Aug 30 14:12:08 <thomasg> that patch doesn't work for me, but using it manually works
Aug 30 14:12:47 <ScaredyCat> thomasg++
Aug 30 14:15:28 <ScaredyCat> mmm....
Aug 30 14:15:34 <ScaredyCat> that appears to have applied
Aug 30 14:17:48 <ScaredyCat> thomasg: where was it getting it from ?
Aug 30 14:17:52 <ScaredyCat> is
Aug 30 14:19:53 <cdbot2> * * OM Bug 780 has been created by hozer(AT)hozed.org
Aug 30 14:19:54 <cdbot2> * * uboot usb serial broken
Aug 30 14:19:55 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=780
Aug 30 14:20:58 <hozer> ScaredyCat: I had an mtd-utils problem with include paths
Aug 30 14:21:12 <hozer> but just running 'make' in the directory fixed it
Aug 30 14:22:01 <aloril> (script) openmoko-apps: Sean Moss-Pultz <sean at openmoko.com> Re: About QVGA theme for OM.2
Aug 30 14:22:02 <ScaredyCat> http://pastebin.ca/675798 is my issue
Aug 30 14:24:05 <hozer> blah, it looks like the tar.gz does not include git metadata
Aug 30 14:25:31 <hozer> OM needs something like tinderbox
Aug 30 14:26:19 <hozer> and a repository that contains *all* the sources that requires that a build succeed before allowing the commit
Aug 30 14:30:02 <SpaMMER> ;e;e;lelel;g
Aug 30 14:30:02 <SpaMMER> gr
Aug 30 14:30:02 <SpaMMER> df
Aug 30 14:30:02 <SpaMMER> df
Aug 30 14:30:03 <SpaMMER> dsf
Aug 30 14:30:03 <SpaMMER> sdd
Aug 30 14:30:05 <SpaMMER> sdsf
Aug 30 14:30:07 <SpaMMER> dfs
Aug 30 14:30:09 <SpaMMER> ddfs
Aug 30 14:30:11 <SpaMMER> ds
Aug 30 14:30:13 <SpaMMER> sd
Aug 30 14:30:15 <SpaMMER> sd
Aug 30 14:30:17 <SpaMMER> ds
Aug 30 14:30:19 <SpaMMER> sd
Aug 30 14:30:21 <SpaMMER> dssd
Aug 30 14:30:23 <SpaMMER> ds
Aug 30 14:30:25 <SpaMMER> ds
Aug 30 14:30:27 <SpaMMER> ds
Aug 30 14:30:28 <SpaMMER> fd
Aug 30 14:30:31 <SpaMMER> ds
Aug 30 14:30:33 <SpaMMER> sds
Aug 30 14:31:24 <ewon> right
Aug 30 14:31:48 <ixs> mhm. svn.berlios.de still down?
Aug 30 14:32:00 <juco> svn: Can't connect to host 'svn.berlios.de': Connection refused
Aug 30 14:32:06 <juco> that's what I get
Aug 30 14:32:50 <ScaredyCat> yeah...
Aug 30 14:32:55 <ScaredyCat> it's broken
Aug 30 14:32:59 <ScaredyCat> has been all day
Aug 30 14:33:31 * ScaredyCat blames.....
Aug 30 14:33:33 <ScaredyCat> ......
Aug 30 14:33:33 <ixs> ScaredyCat: I concur...
Aug 30 14:33:38 <loufoque> what does openmoko does that GPE Phone does not?
Aug 30 14:33:40 <ScaredyCat> ...... ewon
Aug 30 14:33:54 <ScaredyCat> crash?
Aug 30 14:34:13 <ixs> loufoque: It does NOT do calls. And if you order now, you will NOT get a gps. Our lines are open!
Aug 30 14:34:25 <SpeedEvil> You get the GPS hardware.
Aug 30 14:34:29 <ScaredyCat> or wifi
Aug 30 14:34:44 <ixs> SpeedEvil: yeah, you do get hardware, but no soft.
Aug 30 14:34:51 <ScaredyCat> but you can play scummvm games :D
Aug 30 14:35:06 <ixs> RIGHT!
Aug 30 14:35:24 <ixs> anyway, has anyone ever fiddled with the greenphone?
Aug 30 14:35:33 * XorA points at zecke 
Aug 30 14:35:35 <loufoque> why not working on improving GPE Phone for the Neo 1973 instead of creating the new code that is openmoko?
Aug 30 14:35:59 <ScaredyCat> because the neo is just one target for open moko...
Aug 30 14:36:05 <ScaredyCat> s/ /s
Aug 30 14:36:06 <ScaredyCat> vmvjfc#
Aug 30 14:36:44 <SpeedEvil> The idea is that eventually OM will run on many phones.
Aug 30 14:36:57 <zecke> ixs: OE own's one greenphone and I have one
Aug 30 14:36:58 <SpeedEvil> Ideally, Nokia et al pick it up.
Aug 30 14:37:43 <ScaredyCat> i wish someone would :/
Aug 30 14:37:47 <loufoque> I still don't understand the need to create openmoko, since there is GPE Phone which is the same thing and that already supports quite some hardware.
Aug 30 14:37:59 <ScaredyCat> then they could throw a few coders at it
Aug 30 14:38:45 <SpeedEvil> GPE phone came later.
Aug 30 14:38:49 <ScaredyCat> loufoque: because fic want something different ?
Aug 30 14:39:20 <SpeedEvil> GPE phone started in march
Aug 30 14:39:32 <loufoque> GPE itself existed before that
Aug 30 14:40:03 <edistar> if I buy a device in oct. or november, will it be software-upgradeable to the finished consumer product sometime?
Aug 30 14:40:11 <SpeedEvil> yes.
Aug 30 14:40:13 <SpeedEvil> Or now.
Aug 30 14:40:28 <SpeedEvil> The idea is that the whole ball of wax is basically open-source software.
Aug 30 14:40:37 <SpeedEvil> you can install it on any supported phone.
Aug 30 14:40:45 <ozarka> Can you buy a GPE Phone today?
Aug 30 14:41:04 <zash_se> MMN-o!
Aug 30 14:41:14 <SpeedEvil> And changes to the OM applications in the future will just be installable on your existing phone.
Aug 30 14:41:18 <edistar> I mean will the hardware basically stay the same?
Aug 30 14:41:30 <SpeedEvil> edistar: probably. 
Aug 30 14:41:42 <SpeedEvil> edistar: that's to a large extent not the point though.
Aug 30 14:41:56 <MMN-o> zash_se!
Aug 30 14:42:23 <SpeedEvil> The eventual ideal is that you buy a phone you like (supported one of course), and install whatever stuff you want. 
Aug 30 14:42:31 <SpeedEvil> Or take the preinstalled image and just use that.
Aug 30 14:42:46 <edistar> SpeedEvil: I know.. but I have the problem at the moment that my n770 doesn't run all the software for the n800
Aug 30 14:43:16 <SpeedEvil> That's presumably as some of the apps are closed source. 
Aug 30 14:43:23 <SpeedEvil> Or the drivers are not there, or the hardware.
Aug 30 14:43:39 <edistar> SpeedEvil: well... maybe.. but it was also e.g. wesnoth
Aug 30 14:43:42 <SpeedEvil> If you have opensource software, and hardware with drivers, it all should work.
Aug 30 14:43:48 <SpeedEvil> wesnoth?
Aug 30 14:44:08 <edistar> battle for wesnoth, an OSS game ;)
Aug 30 14:44:15 <edistar> just an unimportant example
Aug 30 14:44:27 <SpeedEvil> Well - given OS operating system, and application, you can fix it yourself.
Aug 30 14:44:45 <SpeedEvil> In principle, the idea is for all the core apps at least to be very portable.
Aug 30 14:44:50 <edistar> SpeedEvil: ok
Aug 30 14:44:56 <SpeedEvil> Games that use specific hardware won't be.
Aug 30 14:45:01 <Writchie> edistar: Does you n770 connect successfully with Windows XP using the RNDIS 
Aug 30 14:45:06 <SpeedEvil> If they require an accel, for example.
Aug 30 14:46:52 <Tyrae1> Someone tried SettingsGUI?
Aug 30 14:48:36 <edistar> Writchie: don't have xp
Aug 30 14:49:05 <edistar> Writchie: running debian here
Aug 30 14:49:22 <Writchie> ok thanks, that uses a different interface than RNDIS
Aug 30 14:55:19 <dcordes> hi
Aug 30 14:59:42 <aloril> (script) wiki RecentChanges: [[MokoMakefile]] [[User:ThomasWood]]
Aug 30 15:00:30 <dcordes> is it possible to build an openmoko image with spare oe yet? (without using the script)
Aug 30 15:35:47 <thomasg> ScaredyCat: did the patch help?
Aug 30 15:36:28 <folken> do i correctly assume that since there has been a kernel update (ipkg) i need to reflash the kernel with dfu-util? 
Aug 30 15:43:49 <ewon> I'm being blamed for whatnow?
Aug 30 15:44:02 <XorA> ~blame ewon for something
Aug 30 15:44:03 * apt blames paulproteus and perhaps ewon for something and all the evil in the world
Aug 30 15:48:19 <ScaredyCat> thomasg: no....
Aug 30 15:51:11 <ScaredyCat> the patch was applied ok...
Aug 30 15:51:20 <ScaredyCat> it's bad data...
Aug 30 15:51:30 <ccube_> hi all
Aug 30 16:05:10 <ScaredyCat> s;v
Aug 30 16:06:37 <ScaredyCat> holy crap - berlios.de ... it's alive!!!
Aug 30 16:07:39 <ewon> zomg
Aug 30 16:08:29 <FuzzyCat> ~lart svn.berlios.de
Aug 30 16:08:30 * apt takes a rusty axe and swings it violently, taking svn.berlios.de's head off
Aug 30 16:34:16 <MarcelPentium> salut
Aug 30 16:34:45 <torpor> guys .. is there a really easy way to set up internet on bluetooth?  i'm using ajzaurusUSB on Mac OSX and getting kernel panics (on my macbook) whenever i try to do big transfers .. so i want to swap to bluetooth and see if that gives me less hassle... anyone know if this is working on openmoko yet, and if so, how so?
Aug 30 16:42:01 <xkr47> is there any plans to get the kinetic scrolling into webkit ?
Aug 30 16:42:03 <xkr47> are*
Aug 30 16:46:09 <b1303s__> torpor: you can try http://wiki.openmoko.org/wiki/Manually_using_Bluetooth
Aug 30 16:46:41 <b1303s__> dunno with a mac ..
Aug 30 17:38:58 <geohot> hey
Aug 30 17:39:28 <geohot> are the schematics for the openmoko available anywhere?
Aug 30 17:39:28 <aloril> geohot: See http://wiki.openmoko.org/wiki/FAQ (and also SH1, SH1_FAQ, Neo1973, Developer_preview and Neo1973_Hardware#GTA01Bv4 pages) (Hopefully these links answers your question. This autotrigger for new nick and question during first 15 minutes done only once.)
Aug 30 17:40:32 <geohot> are the schematics for the boards of the openmoko available anywhere?
Aug 30 17:41:19 <SpeedEvil> no.
Aug 30 17:41:25 <SpeedEvil> And openmoko is the software
Aug 30 17:41:32 <SpeedEvil> Neo1973 is the hardware
Aug 30 17:41:37 <geohot> well the neo1973 the
Aug 30 17:41:39 <geohot> *then
Aug 30 17:42:01 <geohot> will they ever be available?
Aug 30 17:42:11 <SpeedEvil> probably not.
Aug 30 17:42:33 <hozer> FIC *is* a hardware company after all ;)
Aug 30 17:49:21 <aloril> (script) wiki RecentChanges: [[FAQ/fr]] [[SettingsGUI]]
Aug 30 18:08:15 <folken> cool.. statisfying to make the first working calls from the gui. 
Aug 30 18:11:35 <bedboi> hi there.
Aug 30 18:11:41 <bedboi> any dbus expert out there?
Aug 30 18:11:55 <torpor> folken: what gui?  calls to what?
Aug 30 18:13:06 <folken> torpor: regular openmoko gui with dialer. Calls to/from landlines and gsm. 
Aug 30 18:16:48 <hozer> folken: awesome.. this is current svn?
Aug 30 18:17:28 <folken> hozer: current image updated with ikpg and new kernel 2.6.22
Aug 30 18:17:55 <folken> hozer: however i need to still kill gsmd and xserver.. then restart gsmd and xserver. Then it asks for pin and just works. 
Aug 30 18:18:06 <hozer> did you build the kernel?
Aug 30 18:18:28 <hozer> I built a kernel and made a uImage, and it seemed to boot, but never gave me the splash screen
Aug 30 18:19:24 <folken> hozer: yes its a selfbuilt kernel build from mokomakefile. 
Aug 30 18:19:54 <folken> hozer: currently the splashscreen only appears partialy. Something is broken there.. 
Aug 30 18:30:08 <hozer>  hrrm.. did you flash your kernel?
Aug 30 18:30:32 <hozer> I booted it from ram, and it sat for 20 minutes at least and never really booted
Aug 30 18:31:14 <folken> hozer: i flashed it. 
Aug 30 18:52:46 <folken> hmm dialer is completley broken now (today) when i dial 0 or asterisk it segfaults instantly. 
Aug 30 18:54:58 <folken> but only if its the first digit. 
Aug 30 18:57:29 <folken> yacc: hoi. :)
Aug 30 18:58:19 <yacc> folken?
Aug 30 18:58:39 <folken> yacc: nur wil .ch 
Aug 30 19:00:14 <yacc> folken :-P
Aug 30 19:02:33 <cw666> someone mentioned a satscan URL was posted to the lists but i dont see it, anyone recall the URL/reference?
Aug 30 19:05:45 <gcb77> cw666: http://members.shaw.ca/mmontour/satscan/satscan_1.0-r0_armv4t.ipk
Aug 30 19:06:37 <gcb77> I was just playing with it... nothing inside, but as soon as I walked outside all these stars came up on the screen :)
Aug 30 19:07:59 <cb22> gcb77, satscan?
Aug 30 19:08:46 <gcb77> It was posted on the device-owners list, one of the sphyrna project hackers wrote a really simple app to test out the GPS
Aug 30 19:09:02 <cb22> nice
Aug 30 19:09:07 * cb22 checks the lists
Aug 30 19:10:54 <baze> anyone knows how robust the case and the touch screen of the neo are?
Aug 30 19:11:34 <Aria> Pretty damn solid. # how's that for empirical
Aug 30 19:11:38 <XorA|gone> case is pretty damn solid
Aug 30 19:11:44 <XorA|gone> I havent hammered the touchscreen though
Aug 30 19:12:07 <XorA|gone> my bv2 doesnt even have any screws in it and its never fallen apart
Aug 30 19:12:43 <baze> sounds good :)
Aug 30 19:12:47 <SpeedEvil> baze: there are drop reports on the wiki. See 
Aug 30 19:12:51 <SpeedEvil> Robustness
Aug 30 19:12:54 <SpeedEvil> or search for it
Aug 30 19:12:56 <baze> thanks
Aug 30 19:12:57 <XorA|gone> someone needs to blend one
Aug 30 19:13:11 <cb22> neo smoothie
Aug 30 19:13:53 <cb22> ouch, how can someone purposfully damage one :(
Aug 30 19:13:56 <cb22> thats sick
Aug 30 19:14:09 <XorA|gone> not seend the video of someone blending an iPhone?
Aug 30 19:14:29 <cb22> an iPhone is ok to blend
Aug 30 19:14:33 <cb22> but a neo - never!
Aug 30 19:14:40 <baze> on willitblend?
Aug 30 19:15:10 <guaqua> has someone bleded the iphone already? :D
Aug 30 19:15:25 <XorA|gone> guaqua: months ago
Aug 30 19:15:30 <guaqua> ooh
Aug 30 19:16:14 <folken> this satscan thing is cool. :) 
Aug 30 19:16:25 <folken> now if i only knew how to interpret the output. 
Aug 30 19:18:30 <gcb77> folken: there's a readme, but basically you have sat numbers on the left, freq on top
Aug 30 19:18:51 <gcb77> and power graph: _-=^*
Aug 30 19:18:56 <gcb77> _ is 0
Aug 30 19:19:05 <gcb77> * is over 80% signal strength
Aug 30 19:20:13 <folken> gcb77: thank you. 
Aug 30 19:22:05 <ScaredyCat> *blink*
Aug 30 19:24:18 <cw666> gcb77: hmm... how long did it take before you saw it find something?
Aug 30 19:24:45 <gcb77> cw666: right away outside, nothing inside.
Aug 30 19:25:28 <gcb77> It takes almost a minute to scan every sat on every freq and start over
Aug 30 19:26:12 <cw666> gcb77: yeah, nothing here, might be the building & balcony above
Aug 30 19:27:03 <gcb77> cw666: Maybe they forgot to put the GPS chip in yours :0 (just kidding, ofcourse)
Aug 30 19:29:08 <thomasg> if I have equally named .bb-files in different folders in $BBPATH, is there a prefered path (the order of the paths in $BBPATH?) - or will it result in conflicts?
Aug 30 19:48:40 <aloril> (script) wiki RecentChanges: [[MokoMakefile]] [[OpenMoko_Local_Groups:_Linkoping]] [[IPhone/es]] [[Booting_from_SD]] [[FAQ/fr]]
Aug 30 19:58:08 <folken> heh, i just got the mediaplayer to work. Wrote a playlist in the editor and uploaded it: docs here: http://www.xspf.org/quickstart/
Aug 30 19:58:27 <folken> soundquality of the headset rocks. :)
Aug 30 19:58:35 <meus> folken: sweet
Aug 30 19:58:41 <ScaredyCat> heheh
Aug 30 19:58:49 <ScaredyCat> that's folken sweet
Aug 30 19:58:51 <gcb77> folken: does it still play through both the speaker and headset?
Aug 30 19:59:00 <meus> i wonder how complete the software will be when gta02 arrives
Aug 30 19:59:05 <folken> gcb77: unfortunatly yes. 
Aug 30 19:59:40 <folken> gcb77: question is wheter the switching is implemented in the driver or in alsa
Aug 30 20:01:34 <stefan|aalborg> folken: mickey|zzZZzz plans to implement speaker muting when headset is inserted while we are in denmark. Gives you hopefully an even better experience. :)
Aug 30 20:02:35 <ScaredyCat> is there a fix for the mtd-utils build problem?
Aug 30 20:02:51 <gcb77> doh, trying to refill my prepaid TMobile card... there's just no way to do it
Aug 30 20:02:54 <thomasg> ScaredyCat: what about the OE patch I told you?
Aug 30 20:03:09 <ScaredyCat> thomasg: it was applied, still broken
Aug 30 20:03:17 <folken> stefan|aalborg: thx for the info. :) 
Aug 30 20:03:19 <thomasg> ScaredyCat: still works perfectly for me
Aug 30 20:03:47 <mwester> gcb77:  you serious?  You can buy refill cards at the local drugstore, no?
Aug 30 20:03:52 <thomasg> ScaredyCat: just add the line: TAG = "9845d92440bd87739c89edd000fd6e0c47fab185"
Aug 30 20:04:12 <ScaredyCat> I get a 404 for the download.. then a 'fatal: Not a git repository'
Aug 30 20:04:19 <thomasg> and modify the SRC_URI line to look like this: SRC_URI = "git://git.infradead.org/mtd-utils.git;protocol=git;tag=${TAG}
Aug 30 20:04:27 <thomasg> that's all and it should work
Aug 30 20:04:39 <gcb77> mwester: I can't get the password to log in online, since that's a text message.
Aug 30 20:04:58 <gcb77> mwester: And I cant use *233 to call, since I cant send tones for selections
Aug 30 20:05:01 <Ghiottone> hello mokoers
Aug 30 20:07:05 <mwester> gcb77:  Ah, ok.  Now that I think about it, I had to borrow my daughter's unlocked RAZR to get the password, 'cause my cell phone is locked to Cingular
Aug 30 20:07:34 <ScaredyCat> thomasg: I have TAG = "9845d92440bd87739c89edd000fd6e0c47fab185"
Aug 30 20:08:26 <ScaredyCat> the src_uri has SRC_URI = "git://git.infradead.org/mtd-utils.git;protocol=git;tag=${TAG} \ plus the other fix lines
Aug 30 20:08:58 <thomasg> I did my last update before 2 hours and it worked for me
Aug 30 20:09:46 <ScaredyCat> :(
Aug 30 20:11:14 <thomasg> can you nopaste me the error log?
Aug 30 20:11:22 <ScaredyCat> ya
Aug 30 20:12:41 <Ghiottone> is there a way to configure gnome (or gtk I don't know) font size without using gnome-control-center?
Aug 30 20:13:01 <Aria> tweak the gconf registry or edit the theme.
Aug 30 20:13:22 <kristian-m> folken try http://mput.de/~kristian/.openmoko/stereospeakers.state and http://mput.de/~kristian/.openmoko/stereoheadphones.state - to activate only speakers or headphones
Aug 30 20:13:22 <ScaredyCat> http://pastebin.ca/676136
Aug 30 20:15:56 <folken> kristian-m: works perfectly. :)
Aug 30 20:16:19 <Ghiottone> kristian-m, hello. does the SettingsGUI package work?
Aug 30 20:17:10 <kristian-m> Ghiottone oh, i haven't tryed the package yet
Aug 30 20:17:29 <folken> i just wonder what happens when i recieve a call now.. 
Aug 30 20:17:48 <kristian-m> Ghiottone version 0.5 is finished, I'll try out the package right now
Aug 30 20:17:58 <folken> kristian-m: have you tried to make calls with your headset profile ? 
Aug 30 20:18:11 <thomasg> ScaredyCat: can you also paste the mtd-utils_1.0.0+git.bb file?
Aug 30 20:18:19 <ScaredyCat> sure
Aug 30 20:18:19 <thomasg> the patched one
Aug 30 20:18:39 <quicksand> Woohoo!  USB RNDIS finally works with Windows XP.  (My employer's PC.)   I just updated the wiki.
Aug 30 20:19:33 <quicksand> Kernel 2.6.22.5-moko11 seems to have fixed it.
Aug 30 20:19:39 <Aria> woot!
Aug 30 20:19:43 <ScaredyCat> http://pastebin.ca/676148
Aug 30 20:20:12 <kristian-m> folken this one is not for the gsm-modul - only for pcm out
Aug 30 20:20:28 <folken> kristian-m: ok 
Aug 30 20:20:28 <kristian-m> Ghiottone it's working, thanks :-)
Aug 30 20:21:06 <folken> the mixer kind of loses me.. :)
Aug 30 20:21:21 <folken> just way too many channels. 
Aug 30 20:21:58 <kristian-m> folken use the "amp" settings
Aug 30 20:23:15 <folken> kristian-m: ah, found it. thx. 
Aug 30 20:24:45 <polz> does the neo1973 come with any sort of warranty ?
Aug 30 20:25:12 <ScaredyCat> 'You break it you own the pieces' warranty
Aug 30 20:25:50 <mickeyl> evening.
Aug 30 20:25:55 <mickeyl> "warranty void if not open"
Aug 30 20:26:05 <happycube> lol!
Aug 30 20:26:16 <CM> Hehe
Aug 30 20:26:16 <ScaredyCat> does that inglude the gps .. 
Aug 30 20:26:18 <ScaredyCat> ahem
Aug 30 20:26:24 <polz> how about "sorry we sent you broken earphones, let us replace them" warranty ?
Aug 30 20:26:25 <Stephmw> mickeyl: "it compiled, ship it!"
Aug 30 20:26:25 <ScaredyCat> include
Aug 30 20:26:30 <mickeyl> Stephmw: :D
Aug 30 20:26:53 <Stephmw> mickeyl: ok, I've never been anywhere quite that bad
Aug 30 20:27:05 <Stephmw> mickeyl: last place we purified, linted and code inspected
Aug 30 20:28:21 <mickeyl> right
Aug 30 20:28:36 <mickeyl> ok, now lets see which key the headphone sends on insert
Aug 30 20:29:01 <Stephmw> <hangup>
Aug 30 20:29:03 * Stephmw hides
Aug 30 20:29:57 <ScaredyCat> prolly the power key
Aug 30 20:31:12 <ScaredyCat> thomasg: the same right  ?
Aug 30 20:31:45 <thomasg> ScaredyCat: no, it's not the same problem as for me :(
Aug 30 20:31:59 <ScaredyCat> I mena the files are the same ?
Aug 30 20:32:00 <thomasg> it seems it just fails at downloading one single file
Aug 30 20:32:06 <thomasg> yeah, it's all correct
Aug 30 20:32:35 <thomasg> on the om download server I couldn't find the file he is looking for
Aug 30 20:36:20 <ScaredyCat> no, it's not there...
Aug 30 20:36:48 <bedboi> anyone good at dbus (glib bindings)?
Aug 30 20:39:53 <Ghiottone> Aria I think I have missed the answer, if there was
Aug 30 20:42:44 <mickeyl> hmm
Aug 30 20:43:38 <CIA-20> openmoko: 03mickey * r2860 10/trunk/src/target/OM-2007.2/daemons/neod/src/buttonactions.c: neod: try to capture headphone insertion / removal
Aug 30 20:45:29 <gcb77> \
Aug 30 20:47:21 <mdt> bedboi, what do u want to know?
Aug 30 20:53:09 <CIA-20> openmoko: 03mickey * r2861 10/trunk/src/target/OM-2007.2/daemons/neod/src/buttonactions.c: neod: fix reversed logic in headphone detection *cough*
Aug 30 20:57:23 <Aria> Ghiottone: tweak the gconf registry or edit the theme.
Aug 30 20:57:26 <mickeyl> ok, detection works. now lets see how to tweak one alsa mixer control
Aug 30 20:58:46 <Ghiottone> Aria, where are the files?
Aug 30 20:59:45 <Aria> Ghiottone: Not sure on the neo
Aug 30 21:00:26 <Ghiottone> in root's home there aren't any
Aug 30 21:01:04 * XorA|gone hears the scary term alsa
Aug 30 21:01:43 <daxxar> Is there any page that has like 'status' on various ``core'' features?
Aug 30 21:01:56 <mickeyl> XorA|gone: don't worry, i won't ask you to do it :)
Aug 30 21:03:08 <XorA|gone> mickeyl: scripting or ioctl
Aug 30 21:03:30 <mickeyl> XorA|gone: i need to control just one mixer param. can i do that w/ alsactl ?
Aug 30 21:03:44 <XorA|gone> mickeyl: amixer
Aug 30 21:04:12 <mickeyl> aah good
Aug 30 21:04:15 <daxxar> Hum, aren't there any wiki pages on a proposal for an SMS application?
Aug 30 21:04:16 <mickeyl> going to use that for a while
Aug 30 21:04:18 <mickeyl> thanks
Aug 30 21:04:41 <kristian-m> hmm, maybe the best way to do this in the long run would be to include a boolean mixer-setting - "disable speakers on headset" directly into the sound driver...
Aug 30 21:05:12 <XorA|gone> kristian-m: your way behind
Aug 30 21:05:30 <kristian-m> oh, kay
Aug 30 21:05:40 <abraxa_> daxxar: Not that I'm aware of
Aug 30 21:06:17 <daxxar> And what's the status on the jitter issue?
Aug 30 21:06:33 * daxxar asks bugzilla
Aug 30 21:06:50 <Ghiottone> how do I turn the gps chip on?
Aug 30 21:06:52 <daxxar> Hm, no hits on 'jitter'. What words should I look for?
Aug 30 21:07:28 <XorA|gone> amixer sset "Amp Mode" "Headphones"
Aug 30 21:07:48 <XorA|gone> amixer sset "Amp Mode" "Stereo Speakers"
Aug 30 21:07:52 <XorA|gone> mickeyl: there :-)
Aug 30 21:08:19 <mickeyl> XorA|gone: cool, merci
Aug 30 21:08:29 * mickeyl adds to neod
Aug 30 21:09:03 <XorA|gone> although its gonna be funny if someone is in a gsm call and removes headphones :-)
Aug 30 21:13:36 <ScaredyCat> thomasg: do you have the tarball ?
Aug 30 21:15:24 <bedboi> mdt: i'm writing a componentized application
Aug 30 21:15:28 <CIA-20> openmoko: 03mickey * r2862 10/trunk/src/target/OM-2007.2/daemons/neod/src/buttonactions.c: neod: change alsa amp mode on headphone insertion / removal. thanks XorA
Aug 30 21:15:37 <bedboi> mdt: the problem is that one component must be run as root
Aug 30 21:15:59 <bedboi> so i should use the SYSTEM bus 
Aug 30 21:16:00 <mdt> bedboi, you want to use system or session dbus?
Aug 30 21:17:09 <cw666> what daemon is responsible for the PM stuff?
Aug 30 21:17:10 <cw666> neod?
Aug 30 21:17:22 <mdt> bedboi,  no, you don't need to. you may start a session bus for root also
Aug 30 21:17:24 <bedboi> mdt: the problem is that the root component won't see the other components if all the other components are sitting on the session bus
Aug 30 21:17:39 <bedboi> mdt: but the two buses won't see each other
Aug 30 21:18:15 <mdt> bedboi, ah, ohne one component runs as root... understand. the other run as what?
Aug 30 21:18:24 <bedboi> as normal user.
Aug 30 21:18:41 <bedboi> there is no problem for me to use the system bus.
Aug 30 21:18:57 <CIA-20> openmoko: 03mickey * r2863 10/trunk/src/target/OM-2007.2/daemons/neod/src/buttonactions.c: neod: make it compile
Aug 30 21:19:10 <mdt> bedboi, what does 'normal' mean, you have only root on the neo today ;)
Aug 30 21:19:12 <bedboi> the problem is this: i have objects whose names are foo.bar.something_%d
Aug 30 21:19:40 <bedboi> where %d is an integer
Aug 30 21:19:41 <dsilva> rwhitby: the openssh-sshd package is missing the setup: set /etc/ssh/sshd_config to allow root login, permit password login, permit empty passwords, and then run /etc/init.d/sshd start
Aug 30 21:20:33 <mdt> bedboi, okay... i used python/glib/dbus. this should not be a problem... or does it?
Aug 30 21:21:22 <polz> can anyone here suggest how I can test the microphone on my earphones ?
Aug 30 21:21:27 <bedboi> mdt:the problem is that when you use the system bus there are security restrictions
Aug 30 21:21:42 <polz> I've just repaired the speakers (by opening the connector and soldering back a loose wire
Aug 30 21:21:46 <polz> ) 
Aug 30 21:21:59 <polz> and now I'm not sure whether the mike still works
Aug 30 21:22:33 <pjz> polz: run some sound-recording app?
Aug 30 21:22:53 <bedboi> mdt: are you able to send a message using glib bindings just knowing the object path?
Aug 30 21:23:04 <mdt> bedboi, hm, i did not use the system bus... don't know. perhaps the %d is not the right idea?
Aug 30 21:23:06 <bedboi> mdt: so with objects that haven't registered a name on the bus
Aug 30 21:23:18 <polz> pjz: how can I select the headphones mike as the source ?
Aug 30 21:23:21 <bedboi> mdt: something like foo_1234
Aug 30 21:23:24 <mdt> bedboi, in python i could. yes.
Aug 30 21:23:33 <bedboi> mdt: what do you use?
Aug 30 21:24:29 <dsilva> rwhitby: other than that though (lots of tapping on the touchscreen terminal), sshfs works great
Aug 30 21:24:31 <mdt> bedboi, in c i avoid glib...
Aug 30 21:25:18 <mickeyl> yay
Aug 30 21:25:21 <mickeyl> it workz
Aug 30 21:25:42 <mdt> bedboi, in c i avoid glib... but i only use events in c and yes, they are not registered.
Aug 30 21:25:52 <XorA|gone> ~praise OH and their claws-mail hack
Aug 30 21:25:53 <apt> All hail OH and their claws-mail hack!
Aug 30 21:27:26 <bedboi> mdt: do you have any example in C?
Aug 30 21:28:08 <mdt> bedboi, for events only...
Aug 30 21:29:49 <Writchie> anybody having trouble with build - internal consistency failure - while compiling nf_nat-H323.c  in 2.6.22.5-moko11 ??
Aug 30 21:31:58 <XorA|gone> hmm, midori runs as root and not as user :-(
Aug 30 21:33:04 <_diego_> how can i download file from a openmoko shell? 
Aug 30 21:33:26 <mickeyl> wget 
Aug 30 21:33:28 <Ghiottone> wget?
Aug 30 21:34:00 <_diego_> ok... it seems i can't route to internet...
Aug 30 21:34:12 <hads> Well yes, you'll need that first...
Aug 30 21:36:15 <Ghiottone> the usb networking page on the wiki has some good tips, iirc
Aug 30 21:38:12 <mdt> bedboi, here is my c-code: http://rafb.net/p/Qvxk8328.html - i reduced it a little but this is more or less what's needed. dbus needs a own thread because the message read blocks. i didn't find a handle to use select()... :/
Aug 30 21:40:16 <Ghiottone> kristian-m, what's new with SettingsGUI 0.5?
Aug 30 21:41:39 * XorA|gone tests SanDisk Ultra II in neo
Aug 30 21:41:43 <rwhitby> morning
Aug 30 21:41:50 <XorA|gone> yo rwhitby 
Aug 30 21:41:54 <Ghiottone> hello rwhitby 
Aug 30 21:43:38 * rwhitby finally works out how to wrap /sbin/hotplug on the Freecom FSG-3 router to automatically bring up usb networking when the neo is plugged in
Aug 30 21:44:11 <kristian-m> Ghiottone gsm cell and state information
Aug 30 21:46:09 <quicksand> XorA: what's the claws-mail hack?  I have that app on my moko but it doesn't fit the screen well at all, nor can I access the context menu.
Aug 30 21:46:29 <rwhitby> has anyone tried the native-sdk ipkg install on the neo yet?
Aug 30 21:47:33 <quicksand> rwhitby: the deafening silence suggests not.
Aug 30 21:48:16 <Ghiottone> kristian-m, oh. I did not notice cause I don't have a sim in my neo.
Aug 30 21:48:33 <Ghiottone> kristian-m, I have uploaded the package for 0.5 and updated the link in my user page
Aug 30 21:48:40 <aloril> (script) wiki RecentChanges: [[User:Alessandro]] [[Talk:FAQ]] [[SettingsGUI]] [[Neo1973_and_Windows]]
Aug 30 21:48:54 <rwhitby> 128MB ram should be enough to self-host MokoMakefile on the neo ;-)
Aug 30 21:49:03 <XorA|gone> quicksand: to make the folders and messages different screens
Aug 30 21:49:23 <kristian-m> Ghiottone thanks
Aug 30 21:49:40 <mjr> rwhitby, that's what swap over nfs is for
Aug 30 21:49:54 <cw666> mjr: except it would deadlock :-)
Aug 30 21:50:10 <XorA|gone> why not swap on the USB harddisk?
Aug 30 21:50:17 <Ghiottone> rwhitby, it would need at least a 32GB microSD card....
Aug 30 21:50:35 <kristian-m> Ghiottone i'll be a while until 0.6 i plan to include ppp configuration for gprs
Aug 30 21:50:35 <SpeedEvil> swapping onto SD is not a problem.
Aug 30 21:51:19 <Ghiottone> kristian-m, great
Aug 30 21:51:33 <Ghiottone> kristian-m, some buttons to connect and disconnect would be very nice too
Aug 30 21:52:32 <Writchie> SpeedEvil: seems to me a 128MB swap space on the SD would rather nice
Aug 30 21:53:12 <Writchie> as soon as bugfix for the SD writes confirmed
Aug 30 21:53:13 <kristian-m> Ghiottone for gsm or gprs? - for gprs, it's on the todo list
Aug 30 21:54:18 <Ghiottone> kristian-m, gprs. but it will need gsm to be active to work
Aug 30 21:55:13 <stefan|aalborg> Writchie: It is confirmed. With the patch I have not seen any problem under high load tests for the SDHC card
Aug 30 21:55:15 <XorA|gone> Writchie: 133M written to card, no errors yet
Aug 30 21:55:36 <XorA|gone> Writchie: with a Sandisk Ultra II
Aug 30 21:55:37 <Writchie> great
Aug 30 21:55:52 <Writchie> kritian-m: are you aware of the AT%EM commands
Aug 30 21:56:25 <Writchie> s/kritian-m/kristian-m/
Aug 30 21:56:26 <apt> Writchie meant: kristian-m: are you aware of the AT%EM commands
Aug 30 21:56:59 <quicksand> XorA: Ah.  Well that does seem to work.  Actually the whole application seems to work pretty well, except for that not-able-to-use-it thing.
Aug 30 21:57:24 <quicksand> (actually, I should give it another try now that I have adjusted my dpi setting to something more sane)
Aug 30 21:57:48 <kristian-m> Writchie what do you mean?
Aug 30 21:58:41 <Writchie> kristian-m: it's an engineering mode command which gives all kinds of additional info e.g. neighboring cells, number of pdp contexts allowed, etc.
Aug 30 21:59:10 <kristian-m> Writchie oh, that sounds fun :-) - is there a list?
Aug 30 22:00:08 <kristian-m> Writchie I am in 2 cells here with the same location ID - would be interesting to find out what else there is and how the signal stenght to other cells is...
Aug 30 22:00:30 <rwhitby> woohoo!  I can now simply plug my neo into my home router, and it's instantly on the lan :-)
Aug 30 22:00:34 <Writchie> kristian-m: I found docs from a gsm module that appears to use a somewhat later chipset than the neo (with EDGE) but most of the commands seem to work and match the documentation.
Aug 30 22:00:50 <Writchie> hold on i'll get a link for you. 
Aug 30 22:01:07 * rwhitby waits for the link with much interest
Aug 30 22:01:38 <thseiler> regarding location of  gsm cells, i found this site: www.cellspotting.com
Aug 30 22:01:51 <CIA-20> openmoko: 03mickey * r2864 10/trunk/src/target/OM-2007.2/applications/openmoko-firststart2/ (24 files in 4 dirs): initial import of openmoko-firststart2
Aug 30 22:02:44 <Writchie> kristian-m: rwhitby: http://www.enfora.com/media/EDG0108TN001_-_Enabler_II-E_Engineering_Mode_Manual.pdf
Aug 30 22:02:54 <CIA-20> openmoko: 03mickey * r2865 10/trunk/src/target/OM-2007.2/artwork/themes/openmoko-standard-2/gtk-2.0/ (gtkassistant gtkrc): openmoko-standard-2: add theming for gtkassistant
Aug 30 22:04:24 <kristian-m> Writchie thanks, i'll have a look at this
Aug 30 22:04:42 <rwhitby> Writchie: awesome, thx.  Looks like you get direct access to the mibs and sibs.
Aug 30 22:05:00 <Writchie> kristian-m: AT%EM=2,1 through 2,13 should work more or less on the neo
Aug 30 22:05:34 * polz finally understands why he was unable to place a call using the headset earlyer
Aug 30 22:06:14 <polz> in case anyone wonders, the mixer "channel" you have to tweak is Neo Mode
Aug 30 22:08:32 <XorA|gone> polz: use the pre-defined states in /etc/alsa
Aug 30 22:08:49 <Writchie> XorA: are these working now?
Aug 30 22:09:08 <XorA|gone> Writchie: alphanone made some changes to make them work AFAIK
Aug 30 22:09:25 * XorA|gone loads a 212M pdf on Neo and all is fine
Aug 30 22:09:52 <Writchie> what is that - microsoft OOXML spec?
Aug 30 22:09:56 <polz> XorA|gone: the real problem was that someone forgot to solder 2 wires to a connector
Aug 30 22:10:02 <polz> in some factory in chine
Aug 30 22:10:06 <polz> chinA
Aug 30 22:10:29 <XorA|gone> polz: not a problem on my bv2
Aug 30 22:10:36 <polz> I've opened the bloody thing, soldered the wires on and glued the thing together
Aug 30 22:11:14 <polz> so now everything works, but the connector looks like crap now :)
Aug 30 22:11:24 <SpeedEvil> Anyone have a clue on if it's possible to get cell broadcasts in plain text?
Aug 30 22:11:52 <XorA|gone> anyway time for sleep
Aug 30 22:11:58 <Writchie> SpeedEvil: you should be able to get them in text mode?
Aug 30 22:12:03 <rwhitby> SpeedEvil: libgsmtool show the name of my suburb in plain text from the cell broadcast
Aug 30 22:12:06 <SpeedEvil> I have a partially written awk thingy to parse them.
Aug 30 22:12:23 <SpeedEvil> Writchie: As I understand it that's only normally supported for SMSs
Aug 30 22:12:30 <SpeedEvil> not cell broadcasts.
Aug 30 22:12:38 <SpeedEvil> rwhitby: with what args?
Aug 30 22:12:46 <rwhitby> when I register
Aug 30 22:14:25 <ScaredyCat> how long does it usually take for someone yo 
Aug 30 22:14:30 <ScaredyCat> oops
Aug 30 22:15:00 <ScaredyCat> how long does it usually take for someone to fix a missing file download from the openmoko servers?
Aug 30 22:15:31 <Writchie> SpeedEvil: AT+CPMS=? should return available memories.
Aug 30 22:15:43 <Writchie> it may depend on whether SIM supports it. 
Aug 30 22:15:57 <Writchie> I haven't seen T-Mobile US support CBM's 
Aug 30 22:16:24 <polz> now, the next question: how do I get rid of the horrible echo when using the neo as a phone ?
Aug 30 22:16:48 <SpeedEvil> Writchie: as I understand it, you can select whatever memory you like, and recieve messages in binary format for each, but SMS only works in ascii
Aug 30 22:16:59 <SpeedEvil> Writchie: as I understand it, you can select whatever memory you like, and recieve messages in binary format for each, but only SMS works in ascii
Aug 30 22:17:42 <SpeedEvil> polz: maybe turn down 'sidetone' in the mixer?
Aug 30 22:19:12 <SpeedEvil> rwhitby: how? I don't see any obvious commands.
Aug 30 22:20:25 <cw666> Writchie: RSTR=`+CPMS: ("ME","SM"),("ME","SM"),("ME","SM")'
Aug 30 22:20:35 <cw666> CBMs are what?
Aug 30 22:20:44 <cw666> cell broadcast messages?
Aug 30 22:20:50 <SpeedEvil> yes.
Aug 30 22:21:02 <cw666> ok, so the above means the sim i have can't/won't support them i take it?
Aug 30 22:21:06 <rwhitby> SpeedEvil: mine came unsolicited
Aug 30 22:21:15 <polz> SpeedEvil: I'll try that.
Aug 30 22:21:36 <polz> SpeedEvil: although I thought the sidetone was just some sort of "white noise"
Aug 30 22:21:42 <SpeedEvil> Hmm. my libgsmtool is crashing
Aug 30 22:21:59 <SpeedEvil> polz: it's feedback from mic to speaker to tell you that you are speaking
Aug 30 22:22:00 <polz> SpeedEvil: btw, I don't hear the echo, the people I talk to do.
Aug 30 22:22:07 <SpeedEvil> ah
Aug 30 22:22:28 <polz> and that's what I'd like to get rid of
Aug 30 22:22:43 <SpeedEvil> It should not actually feed into teh line.
Aug 30 22:22:50 <rwhitby> hmm - now I just need to work out how to distinguish between my two neo's via hotplug so I can set them up on separate subnets via usb0 and usb1 ...
Aug 30 22:23:00 <polz> I think the microphone picks up the speaker
Aug 30 22:23:15 <SpeedEvil> polz: try headset?
Aug 30 22:23:16 <polz> and sends the voice of whoever you're speaking to back
Aug 30 22:23:35 <polz> SpeedEvil: that fixes the problem, but it also makes the phone much less handy
Aug 30 22:23:39 <SpeedEvil> rwhitby: I think they are random MACs.
Aug 30 22:23:46 <rwhitby> yes, they are.
Aug 30 22:23:54 <rwhitby> that needs to be fixed.
Aug 30 22:23:56 <SpeedEvil> rwhitby: maybe set one phone up to use an IP other than 192.168.0.202?
Aug 30 22:24:12 <rwhitby> I've set the second up on 192.168.1.202
Aug 30 22:24:36 <rwhitby> but I need to know which one is plugged in, so I can do the correct ifconfig for the usbX network device.
Aug 30 22:24:58 <rwhitby> (and the appropriate iptables commands and such on the router)
Aug 30 22:25:00 <thomasg> mickeyl: is imlib2 in OE broken?
Aug 30 22:25:18 <mickeyl> not that i know of
Aug 30 22:25:21 <mickeyl> what's wrong?
Aug 30 22:25:35 <thomasg> it complains about the loaders modules
Aug 30 22:25:49 <Writchie> SpeedEvil: should get ("BM","ME","SM") memories when CBM is supported
Aug 30 22:26:13 <SpeedEvil> Ah.
Aug 30 22:26:28 <Writchie> Are you getting these message in pdu mode or not at all?
Aug 30 22:28:23 <mickeyl> thomasg: ah, it's plugin based, it may need OE_LT_RPATH_ALLOW
Aug 30 22:28:31 <mickeyl> thomasg: where does it searches its plugins?
Aug 30 22:28:51 <thomasg> mickeyl: wait a second - I removed the work dir for imlib2 and now it fails at another point :)
Aug 30 22:29:12 <thomasg> make DESTDIR=/OE/tmp/work/armv4t-angstrom-linux-gnueabi/imlib2-1.4.0.002-r0/image install
Aug 30 22:29:23 <thomasg> make: *** No rule to make target `install'.  Stop.
Aug 30 22:29:47 <mickeyl> srcdate changed
Aug 30 22:29:51 <mickeyl> it's past 12
Aug 30 22:29:54 <mickeyl> bitbake -c rebuild imlib2
Aug 30 22:30:02 <mickeyl> try adding the following:
Aug 30 22:30:18 <mickeyl> OE_LT_RPATH_ALLOW = "any"
Aug 30 22:30:20 <mickeyl> OE_LT_RPATH_ALLOW[export]="1"
Aug 30 22:30:22 <mickeyl> then rebuild
Aug 30 22:30:29 <mickeyl> and upgrade imlib + filters + modules
Aug 30 22:30:33 <mickeyl> they should load then
Aug 30 22:32:02 <mickeyl> bed time
Aug 30 22:32:04 <mickeyl> g'night
Aug 30 22:33:40 <ScaredyCat> thomasg: do you have that tarball that's missing from the openmoko servers?
Aug 30 22:34:41 <thomasg> ScaredyCat: no, sry, I have rm_work active and think it deleted it :(
Aug 30 22:35:07 <thomasg> I'll try a `find`
Aug 30 22:35:18 <thomasg> can you give me the exact name pls?
Aug 30 22:35:25 <ScaredyCat> so no one is actually updating and building at all ?
Aug 30 22:35:26 <Writchie> SpeedEvil: it looks to me like both ME and SM memory is supported for CBM
Aug 30 22:35:46 <ScaredyCat> git_git.infradead.org.mtd-utils.git_9845d92440bd87739c89edd000fd6e0c47fab185.tar.gz
Aug 30 22:36:41 <thomasg> ScaredyCat: found it
Aug 30 22:36:54 <thomasg> http://gstaedtner.net/git_git.infradead.org.mtd-utils.git_9845d92440bd87739c89edd000fd6e0c47fab185.tar.gz
Aug 30 22:37:53 <ScaredyCat> you're a star thomasg - thanks
Aug 30 22:38:01 <ScaredyCat> thomasg++
Aug 30 22:38:06 <thomasg> np :)
Aug 30 22:38:50 <Writchie> SpeedEvil: Looks like the default is no CBM messages received. 
Aug 30 22:39:06 <SpeedEvil> yeah - you've got to setup the set of CBMs you want to get.
Aug 30 22:39:15 <ScaredyCat> ooo that;s broken bad
Aug 30 22:39:18 <SpeedEvil> though you can also say which ones you don't want to get.
Aug 30 22:39:19 <Writchie> you may be able to change this to all messages received with AT+CSCB=1,"",""
Aug 30 22:40:44 <thomasg> I hate openembedded -.-
Aug 30 22:41:04 <Writchie> SpeedEvil: to you ever get any "Area Info" on you regular GSM phone?
Aug 30 22:42:06 <thomasg> Writchie: for nokia series60 phones, there's a freeware called celltrack. there's also a commercial software from nokia (very expensive), called net monitor
Aug 30 22:42:56 <Writchie> thomasg: what is it supposed to do?
Aug 30 22:44:03 <thomasg> it shows you the area information
Aug 30 22:44:21 <ixs> Writchie: it gives you cell-id, lac, network name etc..
Aug 30 22:44:25 <ixs> Writchie: signal monitor
Aug 30 22:44:50 <thomasg> net monitor can do much more, but for basic use celltrack is ok
Aug 30 22:45:25 <SpeedEvil> Writchie: I've never plugged this SIM into a regular phone.
Aug 30 22:45:34 <Writchie> thomasg: it looks like using AT%EM=x,y: we can get a lot info for later offline analysis 
Aug 30 22:46:01 <ScaredyCat> is there anyone here with access to the openmoko servers??
Aug 30 22:46:02 <Writchie> thomsag: could probably make a kick-ass open source GSM network monitor ;)
Aug 30 22:46:28 <thomasg> Writchie: yes, I'm really looking forward to this :)
Aug 30 22:46:28 * ScaredyCat too thomasg
Aug 30 22:46:42 <ScaredyCat> openembedded--
Aug 30 22:47:08 <Writchie> i found band switch AT commands but unfortunately they seem to be disabled
Aug 30 22:47:11 <ScaredyCat> thomasg: you ned to get that file pushed to the openmoko servers :)
Aug 30 22:47:24 <thomasg> ScaredyCat: did it help?
Aug 30 22:47:25 <SpeedEvil> Writchie: where are you getting docs?
Aug 30 22:47:27 <Writchie> anyone know if 850 has yet been confirmed to work or not work?
Aug 30 22:47:36 <Writchie> which docs?
Aug 30 22:47:46 <SpeedEvil> Writchie: Command reference docs
Aug 30 22:47:59 <ScaredyCat> yes, except I forgot that I cleared out my work dir, but not the stamps.. rebuilding now
Aug 30 22:48:03 <thomasg> Writchie: afaik it's disabled at moment, because there is no approval in the states
Aug 30 22:48:37 <Writchie> thomasg: that sucks big time - no one ever mentioned this - and I assume its a firmware reload to fix it. 
Aug 30 22:48:41 <thomasg> ScaredyCat: maybe you want to do a bugzilla entry. you can link to the file on my server if you want
Aug 30 22:48:46 <ScaredyCat> SpeedEvil: i have som on my site
Aug 30 22:48:48 <jroman> hi all, every day I've getting problems compiling openmoko, I have to rebuild over and over until, for example I am getting problems with webkit-gtk package, anyboy has the same problem?
Aug 30 22:49:16 <thseiler> The cellspotting.com project has a client for nokia series 60 which can also extract cellid, lac etc... They use it to build a global database of gsm cells. Interesting...
Aug 30 22:49:20 <thomasg> jroman: every time I solved one problem, I get another :)
Aug 30 22:49:24 <jroman> (ups, problems with my english)
Aug 30 22:49:55 <thomasg> at least my basic openmoko-devel-image is working at moment - but now I'm in trouble with imlib2...
Aug 30 22:49:58 * ScaredyCat too ..
Aug 30 22:50:05 <abraxa_> jroman: Do you use mokomakefile?
Aug 30 22:50:08 <Writchie> SpeedEvil: i've been scouring google and turned up some modules that seem to use TI chipsets from enfora.
Aug 30 22:50:12 <ScaredyCat> now I sleep while my precious builds
Aug 30 22:50:27 <Writchie> http://www.enfora.com/media/EDG0108TN001_-_Enabler_II-E_Engineering_Mode_Manual.pdf is a good one with AT%EM commands
Aug 30 22:50:33 <jroman> thomasg: the same symptoms like me :-)
Aug 30 22:50:41 <thomasg> Writchie: I wonder how much information the gsm-module is giving out to the open system
Aug 30 22:50:45 <jroman> abraxa: yes i do
Aug 30 22:50:51 <SpeedEvil> ah.
Aug 30 22:51:16 <Writchie> you can download their other at command references and they document what appear to be TI vendor specific commands.
Aug 30 22:51:40 <Writchie> thomasg: i don't consider these to be any great secrets
Aug 30 22:51:53 <ixs> mh, what is the trigger with mokomakefile to force a rebuild of the images?
Aug 30 22:52:34 <Writchie> however, the operators don't like ordinary people having the info of how bad their network sucks - especially in places like US where it sucks bigtime
Aug 30 22:52:40 <jroman> I'm getting problems with webkit-gtk right now (i always try  make update-makefile && make setup update openmoko-devel-image) and also I try make rebuild-package-<packagename>
Aug 30 22:53:05 <thomasg> Writchie: I work for a operator in europe, and our focus is on quality :)
Aug 30 22:53:34 <Writchie> thomasg: it probably is when you have competition and understand what business you are in 
Aug 30 22:54:12 <thomasg> Writchie: there's competition and quality is one of the "unique" selling point :)
Aug 30 22:54:17 <Writchie> US operators don't understand that they are in the network business - they think they are in the brainwashing business
Aug 30 22:54:37 <Writchie> the GSM model works very well, even with two carriers. 
Aug 30 22:54:58 <jroman> is it a problem with my system, is it a problem with openmoko repository?
Aug 30 22:54:59 <ixs> thomasg: and I always thought cheap is a selling point. :D
Aug 30 22:55:00 <Writchie> We only hope that at least AT&T and T-Mobile will end up competing on network 
Aug 30 22:55:04 <thomasg> well, with some effort it works with even more carriesrs :)
Aug 30 22:55:24 <ixs> jroman: delete your webkit build dir and your stamp files and your sources, then try make again.
Aug 30 22:55:35 <ixs> jroman: I did a rebuild earlier today which succeeded without problems.
Aug 30 22:56:16 <thomasg> ixs: my company doesn't like cheap, they like getting much money for few effort :)
Aug 30 22:56:27 <Writchie> there are good economic reasons for duopolies - but only where the handsets are open are people can change networks at will
Aug 30 22:57:10 <jroman> ixs: thx
Aug 30 22:58:04 <ixs> thomasg: mhm. expensive and no service... Which provider was that again? ;->
Aug 30 22:58:12 * jroman anyway the build system is not solid IMHO
Aug 30 22:58:12 <thomasg> Writchie: imho the best thing would be letting the state handle the infrastructure
Aug 30 22:58:25 <thomasg> ixs: I didn't say no service :)
Aug 30 22:58:33 <SpeedEvil> IMO, the state mandating the standards is probably the least bad.
Aug 30 22:58:41 <Writchie> thomasg: State screw up everything they touch. 
Aug 30 22:58:44 <ScaredyCat> errrm
Aug 30 22:58:44 <SpeedEvil> Interoperable GSM only.
Aug 30 22:58:49 <ixs> thomasg: expensive and bad service? It's either d2 or t-bim. :D
Aug 30 22:58:49 <thomasg> the service of the competitors is much worse
Aug 30 22:58:52 <SpeedEvil> Though that can really screw up.
Aug 30 22:59:05 <SpeedEvil> For example, UK now has digital audio broadcasting.
Aug 30 22:59:05 <thomasg> Writchie: well, that's the problem :(
Aug 30 22:59:15 <SpeedEvil> In MP2, under 128kbits/second often.
Aug 30 22:59:17 <ixs> jroman: depends. build system is quite nice right now. It#s been much worse the last week
Aug 30 22:59:40 <ScaredyCat> ixs: famous last worda
Aug 30 22:59:45 <SpeedEvil> The quality is usually worse than a decent FM signal.
Aug 30 22:59:56 <ScaredyCat> s/worda/words/
Aug 30 22:59:56 <apt> ScaredyCat meant: ixs: famous last words
Aug 30 23:00:10 <Writchie> Just mandate meeting standards liike GSM and prohibiting monopolistic activities like tie-in's of handsets and false advertising and the market would take care of the rest IMHO
Aug 30 23:00:29 <Writchie> it is essentially that all operators in a country be on the same standard
Aug 30 23:00:30 <SpeedEvil> Spectrum allocation is another issue.
Aug 30 23:00:45 <Writchie> otherwise they are not competitions - only the allusion of being competitors
Aug 30 23:01:09 <Writchie> s/sllusion/illusion/
Aug 30 23:01:19 <Writchie> s/illusion/illusion/
Aug 30 23:01:27 <SpeedEvil> :)
Aug 30 23:01:28 <Writchie> dumb bot should know what i mean not what i say
Aug 30 23:01:46 <ixs> ScaredyCat: of course. :-D
Aug 30 23:09:13 <cw666> did anyone make an mtr[-tiny] package yet?
Aug 30 23:11:57 <Writchie> Anybody else getting the fatal error: internal consistency failure - compiling nf_nat_h323.c in linux-2.6.22.5 ?????
Aug 30 23:12:54 <cw666> Writchie: gcc crashing?
Aug 30 23:13:03 <Writchie> looks like it
Aug 30 23:13:17 <Writchie> looks like it was rebuilt on this build
Aug 30 23:13:28 <Writchie> or it seemed that I saw that flashing by
Aug 30 23:13:33 <cw666> gah
Aug 30 23:13:48 <cw666> i so dislike things that touch/rebuild toolcahins all the time for that reason
Aug 30 23:15:16 <Writchie> i don't understand why we can't have stability at least in the compiler?
Aug 30 23:15:25 <Writchie> i.e. the compiler we use.
Aug 30 23:15:44 <Aria> Note to self: build crossgcc-arm-eabi for my distro tonight.
Aug 30 23:16:48 <cw666> Writchie: because then you wouldn't have to use 12GB and 5hrs+ for a rebuild?
Aug 30 23:17:10 <Writchie> oh - you mean i would actually get some useful work done?
Aug 30 23:17:25 <cw666> dont get me started
Aug 30 23:17:58 <cw666> Writchie: next you'll be wanting sane makefiles and sane build times for smaller apps
Aug 30 23:20:05 <Writchie> i'm starting to miss Visual Studio
Aug 30 23:20:21 <dsilva> is there an easy way to get profiling data (or just instruction counts) on binaries running on the phone?
Aug 30 23:24:46 <SpeedEvil> Hmm.
Aug 30 23:25:00 <SpeedEvil> at%em=2,3
Aug 30 23:25:10 <Writchie> cool stuff
Aug 30 23:25:30 <SpeedEvil> I wonder if this in fact gives range to the cells that it's considering logging on to.
Aug 30 23:25:42 <Writchie> crude
Aug 30 23:25:52 <SpeedEvil> Crude, of course.
Aug 30 23:26:17 <SpeedEvil> But significantly better than simple cell-ID/signal-strenght
Aug 30 23:26:58 * SpeedEvil needs to do some logging.
Aug 30 23:27:00 <Writchie> its only a single operator
Aug 30 23:27:04 <SpeedEvil> (next time I'm out.)
Aug 30 23:27:14 <Writchie> we're often lucky to get one cell
Aug 30 23:27:32 <Writchie> we definitely need a neo engineering mode
Aug 30 23:27:45 <Writchie> and track all this stuff - integrated with the call handling
Aug 30 23:27:50 <Writchie> and the gps
Aug 30 23:27:55 <Writchie> uber cool
Aug 30 23:28:03 <SpeedEvil> I'm getting three at the moment.
Aug 30 23:28:30 <juco2> http://www.google.com/support/jobs/bin/static.py?page=why-ny-ny.html&id=research
Aug 30 23:28:38 <juco2> has a link to: http://www.stat.purdue.edu/%7Echuanhai/docs/callModelBW.pdf
Aug 30 23:28:44 <juco2> I was looking at that the other day
Aug 30 23:29:05 <SpeedEvil> What is it?
Aug 30 23:29:14 <juco2> a paper about using the data you are looking at
Aug 30 23:29:19 <SpeedEvil> ah.
Aug 30 23:29:24 <juco2> A Statistical View of the Transient Signals that Support a
Aug 30 23:29:24 <juco2>                        Wireless Call
Aug 30 23:29:37 <Writchie> i think this was a cdma network
Aug 30 23:30:16 <Writchie> but principles probably similar
Aug 30 23:30:30 <juco2> oh, hmm, I only started reading it, I did see that they were talking about CDMA but I thought perhaps they were only speaking of CDMA vs TDMA
Aug 30 23:32:54 <LittleIdiot> http://www.okean.com/antispam/iptables/rc.firewall.sinokorea
Aug 30 23:33:20 <LittleIdiot> yeeks, wrong chan
Aug 30 23:33:47 <Writchie> you meant the great firewall of china channel
Aug 30 23:35:18 <SpeedEvil> Writchie: As I see it 2,3 has the majority of the interesting onfo.
Aug 30 23:35:47 <Writchie> IIRC there are some others very interesting when a call is in progress
Aug 30 23:35:53 <Writchie> or gprs session
Aug 30 23:36:39 <Writchie> unfortunately, there's a lot of work necessary just to handle calls correctly ;)
Aug 30 23:36:46 <SpeedEvil> Yeah - GPRS is interesting.
Aug 30 23:36:54 <SpeedEvil> but GPRS of course would imply a working mux.
Aug 30 23:37:10 <cesarb> SpeedEvil: has the "not really off" been resolved already?
Aug 30 23:37:11 <Writchie> enough data for a court case to get your money back ;)
Aug 30 23:37:28 * cesarb is thinking of looking at the kernel/uboot code responsible for the audio amp
Aug 30 23:37:35 <SpeedEvil> cesarb: no.
Aug 30 23:37:55 <SpeedEvil> cesarb: the modem turning off has been.
Aug 30 23:38:10 <SpeedEvil> you send it at@poff
Aug 30 23:38:12 <SpeedEvil> and it goes off
Aug 30 23:38:52 <cesarb> SpeedEvil: I'm suspecting that something on either the kernel or uboot is turning on the audio amp (which, according to something I saw you saying yesterday on my logs, is connected directly to the battery)
Aug 30 23:39:00 <cesarb> SpeedEvil: and NOTHING is turning it off
Aug 30 23:39:14 <cw666> SpeedEvil: there is a patch for a partial implementation of the mux code
Aug 30 23:39:26 <cesarb> SpeedEvil: which means, if it's true, that it's not only wasting power with the phone off, but also wasting power with it _on_
Aug 30 23:39:42 <cesarb> SpeedEvil: with obvious battery life consequences
Aug 30 23:40:08 <Writchie> prolly not much with nothing playing but every bit helps
Aug 30 23:40:18 <SpeedEvil> Indeed. 
Aug 30 23:40:41 <SpeedEvil> Determining when the amp should be off is not completely trivial though.
Aug 30 23:40:51 <cesarb> Writchie: IIRC, according to speedevil yesterday, it's the same the GSM modem uses... not really "not much" :P
Aug 30 23:41:20 <SpeedEvil> You'd have to talk to pulseaudio
Aug 30 23:41:22 <SpeedEvil> Well.
Aug 30 23:41:24 <Writchie> the combined total seems to be 40ma 
Aug 30 23:41:25 <SpeedEvil> It's 15mA
Aug 30 23:41:27 <SpeedEvil> ish
Aug 30 23:41:37 <SpeedEvil> so it's 1% of battery per hour
Aug 30 23:42:16 <Writchie> is any of the apm working
Aug 30 23:42:25 <SpeedEvil> VEry little.
Aug 30 23:42:36 <SpeedEvil> [[Neo1973 Power Managment]]
Aug 30 23:42:49 <SpeedEvil> 1/7 power managment thingies are implemented
Aug 30 23:43:16 <Writchie> IMO the BIG thing is shutting down and restarting the kernel
Aug 30 23:43:52 <SpeedEvil> The huge thing is rapidly and completely shutting down stuff when it's not needed.
Aug 30 23:44:38 <rwhitby> cw666: I'm adding mtr to task-openmoko-feed
Aug 30 23:44:39 <SpeedEvil> And of course varying clockspeeds.
Aug 30 23:45:00 <Writchie> doesn't every driver have to respond to some kind of apm event - and do so sanely
Aug 30 23:45:10 <SpeedEvil> In theory.
Aug 30 23:45:22 <Writchie> i didn't see much improvement from the clock speed
Aug 30 23:45:30 <SpeedEvil> There is some.
Aug 30 23:45:31 <cesarb> SpeedEvil: determining when the amp should be off should be easy in theory: when the device node is closed
Aug 30 23:45:42 <SpeedEvil> cesarb: not really.
Aug 30 23:45:51 <cesarb> SpeedEvil: if pulseaudio is like esd or arts, the device is closed after some time of inactivity
Aug 30 23:45:55 <SpeedEvil> you can route from the GSM modem or the BT port to the speakers
Aug 30 23:46:07 <SpeedEvil> and not have the CPU touching it.
Aug 30 23:46:08 <Aria> SpeedEvil's right on this one: when a call
Aug 30 23:46:13 <Writchie> big question - who is doing the apm stuff in the kernel?
Aug 30 23:46:18 <Aria> When a call's active, the GSM directly outputs sound, not through the CPU
Aug 30 23:46:19 <cesarb> SpeedEvil: hm, so it would have to be /sys ;-)
Aug 30 23:47:23 <SpeedEvil> The simple option would be a fake mixer.
Aug 30 23:47:29 <SpeedEvil> amp on or amp off
Aug 30 23:47:51 <cesarb> SpeedEvil: cool idea
Aug 30 23:48:03 <SpeedEvil> We don't have enough mixers.
Aug 30 23:48:06 <SpeedEvil> :)
Aug 30 23:48:17 <Writchie> can we go to 3 digits
Aug 30 23:48:40 <aloril> (script) wiki RecentChanges: [[Installation_guide]] [[Installation_emulation_automatic_ubuntu]] [[Application_Development_Crash_Course]] [[Talk:FAQ]]
Aug 30 23:53:19 <rwhitby> http://www.enfora.com/media/Enabler-IIEATCommandSetEDG0108PB004MAN.pdf <- AT command set Writchie was referring to
Aug 30 23:54:56 <rwhitby> http://www.enfora.com/media/Enfora_Enabler_II-E_Integration_Guide_EDG0108PB003MAN.pdf may be interesting too
Aug 31 00:00:02 * SpeedEvil just wants to know how to turn on 3D mode :)
Aug 31 00:00:09 <SpeedEvil> s/D/G/
Aug 31 00:00:48 <Writchie> i'd settle for EDGE 
Aug 31 00:01:45 <xaid> I'm trying to get OpenMoko working using the Qemu guide and I'm running into a small issue where I get a kernel panic when I try to boot the newly built image, something regarding a missing libm.so.6. I have libm.so in my /usr/lib and I was wondering if there is a way to get this working without creating a symlink :)
Aug 31 00:05:16 <xaid> ah, nevermind. its been a long day :)
Aug 31 00:09:07 <cesarb> xaid: if the error is within the emulated image, it has no relationship with libm.so outside the emulated image...
Aug 31 00:15:53 <xaid> cesarb: I realised that after I wrote it down. I think I know what the problem is, I'm running a hardened Gentoo toolchain and I see PaX errors being logged during 'make qemu'.
Aug 31 00:16:16 <xaid> openmoko/flash.sh: line 88: 24296 Terminated              ( sleep $2; kill $pid; sleep 1; kill -KILL $pid )
Aug 31 00:16:21 <cesarb> xaid: gentoo. always gentoo.
Aug 31 00:16:39 <cesarb> xaid: it seems everyone who has problems compiling qemu is using gentoo
Aug 31 00:16:50 <xaid> cesarb: lol, I'm pretty sure it'd be smooth if I was running the non-hardened toolchain.
Aug 31 00:17:36 <xaid> cesarb: I'll compile a non-PaX'ed kernel and give it a go. it should take 5 minutes :)
Aug 31 00:18:19 <xaid> there are a few modifications that I had to do to get it working on Gentoo though since the paths for certain binaries are different
Aug 31 00:19:13 <cesarb> xaid: the most common problem with gentoo is that qemu's autodetection of gcc-3.x fails (i.e. it cannot use gcc 3.x while the rest was compiled with 4.x)
Aug 31 00:19:53 <cesarb> xaid: and then people say "ok, let's just make the whole system use gcc 3.x" (gcc-config IIRC, it's been a while since I last used gentoo), and then they have problems because some things aren't quite as tested with gcc 3.x...
Aug 31 00:21:09 <xaid> cesarb: I'm actually using gcc 3.4 here on a system-wide scale since gcc-4.x is not hardened-friendly yet. I guess it'll be interesting on reboot to see if I can get a successful qemu image up and running.
Aug 31 00:24:04 <cesarb> SpeedEvil: found what I was looking for. Could you try setting the "Amp Mode" mixer to -5 (to which the kernel will add 5 to get to mode 0, which is the shutdown mode), and doing your current-measuring dance?
Aug 31 00:24:20 <cesarb> SpeedEvil: try also playing with the "Amp Fast Wakeup Playback Switch" one, should be a boolean
Aug 31 00:25:16 <SpeedEvil> I could.
Aug 31 00:25:22 * SpeedEvil shuts down neo
Aug 31 00:26:11 <cesarb> SpeedEvil: I'm not seeing any reboot notifier or something like that which would set it to 0 (shutdown, visible to userspace as -5) when rebooting or powering off
Aug 31 00:26:18 <cesarb> SpeedEvil: which would nicely explain the symptoms
Aug 31 00:26:22 <SpeedEvil> indeed.
Aug 31 00:27:48 * cesarb spots a wtf
Aug 31 00:28:04 <cesarb> SpeedEvil: _if_ you manage to set that control to minus five...
Aug 31 00:28:17 <cesarb> SpeedEvil: the alsa state file I'm looking at says it's ENUMERATED
Aug 31 00:28:22 <cesarb> SpeedEvil: from 0 to 4
Aug 31 00:29:47 <cesarb> SpeedEvil: the enumeration values match precisely with the mode table, modes 5 to 9...
Aug 31 00:30:27 <cesarb> SpeedEvil: mode 5 is "mono line out"... the alsa state file comment says "off"...
Aug 31 00:31:24 <rtyler> aw damnit, the latest kernel fails
Aug 31 00:31:32 <rtyler> on the netfilter module AFAIK
Aug 31 00:31:38 <rtyler> NOTE: Tasks Summary: Attempted 2992 tasks of which 2923 didn't need to be rerun and 1 failed.ERROR: '/home/tyler/source/openmoko/openembedded/packages/linux/linux-gta01_2.6.22.5.bb' failed
Aug 31 00:33:50 <SpeedEvil> ok - amp mode current if it truly turns off should be visible with the system on
Aug 31 00:34:23 <cesarb> SpeedEvil: just noticed I misread the code, you should set it to 0 (text should say "off"), it only adds 5 for the other 4 modes
Aug 31 00:35:11 <cesarb> SpeedEvil: (so, what we should try is setting "Amp Mode" to 0/Off, and playing with "Amp Fast Wakeup Playback Switch" to see if it makes any difference)
Aug 31 00:36:09 <cesarb> SpeedEvil: if it works, a quick-and-dirty script on the shutdown path to turn it off would workaround the issue, and it could be used by the UI too when nothing is playing
Aug 31 00:37:04 <cesarb> SpeedEvil: the right solution would be, however, to add a notifier to the kernel so it would turn if off when shutting down (or rebooting - IIRC the notifier path is the same)
Aug 31 00:39:02 <SpeedEvil> running.
Aug 31 00:39:13 <SpeedEvil> Amp mode 'off' reduces power use by some 10-15mA
Aug 31 00:39:25 <SpeedEvil> some of the amp modes use more power than others
Aug 31 00:39:41 * cesarb is not having luck with the build system... after fighting with the build of mtd-utils-native for hours yesterday, and because of it hitting the dreaded date rollover bug, I flash the neo to a crashing/hanging graphics mode (which somehow breaks cdc ether), and now after a update it decides to remake the toolchain...
Aug 31 00:40:01 <SpeedEvil> 'call speaker' is the cheapest at 8mA or so
Aug 31 00:40:17 <cesarb> SpeedEvil: cool, isn't 10-15mA near what you saw with the AP and GSM off?
Aug 31 00:41:03 <cesarb> SpeedEvil: isn't very surprising, since "call speaker" is the only one which is mono
Aug 31 00:41:25 <SpeedEvil> Whee!
Aug 31 00:41:40 <cesarb> SpeedEvil: try playing with "Amp Earpiece 6dB Playback Switch"
Aug 31 00:41:41 <SpeedEvil> I can't measure current after shutdown -h
Aug 31 00:41:51 <cesarb> SpeedEvil: it should make a difference when in the "call speaker" mode
Aug 31 00:42:01 <SpeedEvil> Which means that it's below .1mV across a 1.2 ohm resistor
Aug 31 00:42:04 <cesarb> SpeedEvil: (and only in that mode)
Aug 31 00:42:23 <SpeedEvil> Actually - it shouldn't make a difference unless you're playing audio on it
Aug 31 00:42:47 <SpeedEvil> So under a hundred microamps
Aug 31 00:42:50 <cesarb> SpeedEvil: the "Amp Fast Wakeup Playback Switch", on the other hand, I couldn't find on the datasheet which difference it makes... (perhaps for off->on transitions?)
Aug 31 00:43:15 <SpeedEvil> Dunno. Maybe.
Aug 31 00:43:17 <SpeedEvil> Anyway. 
Aug 31 00:43:19 <cesarb> SpeedEvil: "Amp Earpiece 6dB Playback Switch" seems to change the gain
Aug 31 00:43:29 <SpeedEvil> gain does not equate to power use.
Aug 31 00:43:36 <SpeedEvil> it means it's louder when it's playing
Aug 31 00:43:52 <cesarb> SpeedEvil: well, now if with "Amp Mode" off, and GSM off, and a shutdown, I wonder how little it should drain...
Aug 31 00:43:54 <SpeedEvil> it does not mean that it uses more current when silent.
Aug 31 00:44:01 <cesarb> SpeedEvil: makes sense
Aug 31 00:44:01 * SpeedEvil points to above.
Aug 31 00:44:09 <SpeedEvil> Under 100uA
Aug 31 00:44:22 <cesarb> SpeedEvil: below what you can measure?
Aug 31 00:44:27 <SpeedEvil> yes.
Aug 31 00:44:32 <cesarb> SpeedEvil: w00t!
Aug 31 00:44:42 <SpeedEvil> Indeed. Now going to try leaving the modem on
Aug 31 00:47:57 * BryceLeo wonders if i lost connection
Aug 31 00:48:01 * BryceLeo guesses not
Aug 31 00:49:46 <SpeedEvil> Annoyingly, crashed - resistor value is too high and it brownouts on transmit.
Aug 31 00:49:50 <SpeedEvil> retrying
Aug 31 00:50:39 <cesarb> SpeedEvil: 1.2ohm is too high?
Aug 31 00:52:14 <cesarb> SpeedEvil: I'm reporting the bug on bugzilla, now that we know it does make a difference...
Aug 31 00:56:04 <SpeedEvil> Well - it's trivial to fix at least basically
Aug 31 00:56:18 <SpeedEvil> stuff a alsactl line in the poweroff script
Aug 31 00:57:01 <zeuszz> entren aki aver q pueden hacer los hacker -----> /server irc.tucumanos.net
Aug 31 00:57:02 <zeuszz> entren aki aver q pueden hacer los hacker -----> /server irc.tucumanos.net
Aug 31 00:57:03 <zeuszz> entren aki aver q pueden hacer los hacker -----> /server irc.tucumanos.net
Aug 31 00:57:03 <zeuszz> entren aki aver q pueden hacer los hacker -----> /server irc.tucumanos.net
Aug 31 00:57:04 <SpeedEvil> yes - the absolute power peaks are something like 900mA maybe.
Aug 31 00:57:20 <SpeedEvil> OR maybe my wires are loose.
Aug 31 00:57:27 <SpeedEvil> Anyway, the interesting numbers.
Aug 31 00:57:58 <SpeedEvil> In echo mem >/sys/power/state - with amp off and modem on, the numbers are 50mA or so.
Aug 31 00:58:23 <SpeedEvil> In shutdown, with modem on, they are 15mA
Aug 31 00:58:49 <SpeedEvil> So for the near-term at least, that means a standby of 24h or so.
Aug 31 00:59:02 <SpeedEvil> which isn't terribly great, but not completely hideous.
Aug 31 00:59:33 <SpeedEvil> roh: under 100uA drain with @poff and with 'amp mode' in alsamixer to 0.
Aug 31 00:59:40 <SpeedEvil> roh: under 100uA drain with @poff and with 'amp mode' in alsamixer to off.
Aug 31 01:00:45 <SpeedEvil> And somewhere over 3 days, with the modem on, and some hypothetical fast-wake mechanism.
Aug 31 01:01:45 <cesarb> Bug 781 Submitted
Aug 31 01:01:53 <cdbot2> * * OM Bug 781 has been created by cesarb(AT)cesarb.net
Aug 31 01:01:54 <cdbot2> * * LM4857 should be turned off on a reboot notifier
Aug 31 01:01:55 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=781
Aug 31 01:02:27 <SpeedEvil> The suspend-RAM number is _way_ too high for some reason.
Aug 31 01:02:42 <SpeedEvil> As I understand it, it should be on the order of a milliamp.
Aug 31 01:03:05 <cesarb> SpeedEvil: well, I'd suspect the OTHER sound chip, the one which would be turned off by the PMU on a normal shutdown...
Aug 31 01:03:22 <SpeedEvil> Hmm.
Aug 31 01:03:24 <cesarb> SpeedEvil: as well as the other dozens of random milimiter-sized chips littering the board
Aug 31 01:03:33 * cesarb goes to look the magic setting for the other chip too
Aug 31 01:04:45 <SpeedEvil> And then there is the SC* config itself.
Aug 31 01:05:56 <cesarb> SpeedEvil: yeah, but that one is too scary
Aug 31 01:11:32 <SpeedEvil> What would be really nice would be to be able to setup the ROM in the SC* chip right.
Aug 31 01:11:37 <SpeedEvil> But that's expensive.
Aug 31 01:11:55 <cesarb> SpeedEvil: urgh, this codec has FOUR power control registers, plus a ridiculously high number of special "low power" mode settings
Aug 31 01:11:58 <SpeedEvil> I'd imagine you'd need to buy >100K
Aug 31 01:14:49 <cesarb> SpeedEvil: found the "magic setting" for the wm* driver... you have to put it on D3cold
Aug 31 01:15:12 <cesarb> SpeedEvil: (perhaps you also have to disable a few dozen mixer switches... I hope not)
Aug 31 01:15:45 <cesarb> SpeedEvil: but it seems it is already done on suspend...
Aug 31 01:15:51 <SpeedEvil> I wonder...
Aug 31 01:16:04 <SpeedEvil> The number for suspend-RAM is just about bang on 'idle' mode of the CPU.
Aug 31 01:16:44 <SpeedEvil> It says 124mW typ - which is 30mAish. 
Aug 31 01:17:13 <SpeedEvil> which is about what I measure (well, would be with the modem off)
Aug 31 01:18:06 <cesarb> SpeedEvil: well, if you can find the right /sys/... directory for the wm* chip, and find the power control file within it, and echo the correct magic value to it to make it suspend/resume, you might be able to measure the codec's contribution to the power use
Aug 31 01:19:04 <cesarb> gtg bbl
Aug 31 01:23:42 <SpeedEvil> there is no power control file for most stuff
Aug 31 01:35:48 <wotcha> account list
Aug 31 01:36:01 <wotcha> oops sorry
Aug 31 01:49:40 <aloril> (script) wiki RecentChanges: [[Neo1973_GTA01_Power_Management]]
Aug 31 01:56:06 <rwhitby> cw666: mtr fails to build for me
Aug 31 02:03:25 <SpeedEvil> roh: Should  STBY_1V8 - generated by d3reg -really be set at 2.1V? this seems outside the 1.8V+-0.15 that I see in the datasheet.
Aug 31 02:05:35 <technil> rwhitby, cw666 : mtd-utils-native?
Aug 31 02:05:44 <technil> oh sorry mtr
Aug 31 02:05:47 <technil> nm
Aug 31 02:06:18 <cesarb> technil: I saw a change today on OE which probably means mtd-utils-native has been fixed
Aug 31 02:09:39 <BryceLeo> anyone in here happy to know python and use the gtkmozembed?
Aug 31 02:10:47 <BryceLeo> s/happy/happen/g
Aug 31 02:10:48 <apt> BryceLeo meant: anyone in here happen to know python and use the gtkmozembed?
Aug 31 02:11:04 <BryceLeo> ]
Aug 31 02:54:41 <xaid> cesarb: it is working. I had to delete the content of my 'images' directory and after it redownloaded the images and built everything again, things are loading correctly now (albiet slow)..
Aug 31 02:55:56 <xaid> cesarb: I should be able to use the mouse to click and launch applications, right? nothing seems to happen when I click on the contacts icon or so.
Aug 31 02:57:21 <cesarb> xaid: it's broken on 2007.2
Aug 31 02:57:41 <cesarb> xaid: it's said there's a specific sequence of clicks on specific locations which allow you to get to the terminal
Aug 31 02:57:51 <cesarb> xaid: where you can simply run the screen calibration tool to fix it
Aug 31 02:58:11 <cesarb> xaid: I heard the trick is to realize the screen is in fact rotated
Aug 31 02:58:44 <xaid> cesarb: thanks for the tip :)
Aug 31 02:59:33 <cesarb> xaid: if you try hard enough you can find places where people discussed the tricks... perhaps on the mailing lists, perhaps on the wiki, perhaps on the irc logs... I don't recall where I heard of such tricks
**** ENDING LOGGING AT Fri Aug 31 02:59:56 2007