**** BEGIN LOGGING AT Tue Sep 25 02:59:56 2007 Sep 25 03:13:46 are there supposed to be 2 Torx screws or just one in the GTA01bv4? Sep 25 03:15:21 * gcb77 wonders if the warranty covers the guitar pick ;) Sep 25 03:31:46 morning. Anyway to upgrade the OpenMoko phone to be usable as a phone yet? Sep 25 03:31:58 qtopia? Sep 25 03:32:08 (ducks) Sep 25 03:32:10 I see.. well thanks. Sep 25 03:32:39 still need power management and more stability in the gsm handling Sep 25 03:33:46 so as a novice is there anything I can do then? Sep 25 03:34:11 you can file bugs or correct things in the wiki Sep 25 03:34:16 qtopia is good Sep 25 03:34:35 i am going to issue a qtopia update soon Sep 25 03:34:38 hey you're not impartial :) Sep 25 03:34:42 cool Sep 25 03:34:45 nope. i am not Sep 25 03:34:57 well the current image I have is so... unusable that I do not dare to file anything Sep 25 03:35:11 does contact sync work? Sep 25 03:36:50 ok.. well I'll try qtopia to see how it looks like then.. thanks Sep 25 03:37:48 turn suspend off :) Sep 25 03:38:22 what does suspend do? Sep 25 03:38:56 ljp: updated to the 2.6.22 kernel? Sep 25 03:39:10 (or does not do) Sep 25 03:39:23 BLUG-fred: suspend works fine, it's the resume that is broken Sep 25 03:39:40 haha... ok Sep 25 03:41:39 yes, updated kernel Sep 25 03:42:18 ljp: I was able to ipkg upgrade and replace the kernel and qtopia kept working Sep 25 03:42:20 :) Dual-boot should be easier then Sep 25 03:42:27 but no qtopia components were upgraded :( Sep 25 03:43:36 fix it. :) Sep 25 03:47:44 would the agps on the neo work as an external gps device like to plug it into google earth through bluetooth? Sep 25 03:48:07 i mean could that functionality be added? Sep 25 03:48:28 xLOGiC yes but initially it would be a bit of a scripted up hackish thing Sep 25 03:49:22 I did something similar with gumstix and hacked it to allow two client connections at once Sep 25 03:51:17 cool.. i wish i knew how to code and / or owned a neo.. Sep 25 03:51:35 I know where you can buy one, xL0GiC Sep 25 03:51:54 gps repeater should be in the gui at some point Sep 25 03:52:12 ljp: any suggestions on howto test the qtopia 4.3 ui without a neo or greenphone? Sep 25 03:52:49 mwester-laptop: wasn't it sold out? Sep 25 03:53:21 Apparently another batch will be shipping out shortly Sep 25 03:54:47 hmmm Sep 25 03:55:37 i'll look into that. but still gta02 looks like it's worth the wait. specially for non-devs Sep 25 03:55:40 jannu: if you just want to test the ui, you can dl the source, and compile it (for qvfb) and run it with the Neo skin Sep 25 03:58:01 ljp: ok. i'm mostly interested in the chosen ui concepts&approaches to usual smartphone stuff Sep 25 04:02:25 ljp: i.e. general ui choices, voice-based communication, text-based communication, calendar, etc Sep 25 04:02:42 is there any getting started guide that showcase what's doable, what can be install and so on? Sep 25 04:03:00 I found something, but there are so many warnings.. that I felt that was not for me Sep 25 04:18:30 (script) wiki RecentChanges: [[Automatic_emulation_in_Ubuntu]] Sep 25 04:57:35 moin Sep 25 05:15:35 morning Sep 25 06:02:16 anyone here gone through the un-brick procedure? Sep 25 06:12:20 gcb77: yep Sep 25 06:13:05 rwhitby: may I trouble you with a few questions? Sep 25 06:13:49 I flashed a dev-version uBoot yesterday and the NEO won't start up any more Sep 25 06:13:50 go for it - reponses may be delayed since I'm at work. Sep 25 06:14:08 cool, thank you Sep 25 06:14:13 Here's where I'm at: Sep 25 06:14:19 first of all, are you sure it's not a battery charge issue? Sep 25 06:14:39 yes, tested both batteries, both should be at least half full Sep 25 06:15:25 ok, where are you at with devirginator and the debug board then Sep 25 06:15:34 I have the NEO dis-assembled, hooked up to the debug board, openocd set up and configured Sep 25 06:15:46 I'm using the MokoMakefile version Sep 25 06:15:54 by setting PATH and LD_LIBRARY_PATH Sep 25 06:16:01 ok Sep 25 06:16:07 I run openocd and get: Sep 25 06:16:25 Error: jtag.c:1253 jtag_examine_chain(): JTAG communication failure, check connection, JTAG interface, target power etc. Sep 25 06:16:29 Error: jtag.c:1440 jtag_init(): trying to validate configured JTAG chain anyway... Sep 25 06:16:34 Error: jtag.c:1346 jtag_validate_chain(): Error validating JTAG scan chain, IR mismatch, scan returned 0x00 Sep 25 06:16:42 then eventually Sep 25 06:16:43 Error: jtag.c:1448 jtag_init(): Could not validate JTAG chain, exit Sep 25 06:17:01 ok, pastebin your openocd config file Sep 25 06:17:55 http://pastebin.com/m6506411 Sep 25 06:18:11 Am I supposed to be pushing the power button on the NEO? Sep 25 06:18:29 (script) wiki RecentChanges: [[Mokopedia]] Sep 25 06:19:04 line 10 looks weird Sep 25 06:21:00 in that its only supposed to have and ? Sep 25 06:22:10 gcb77 http://pastebin.com/m46993da0 Sep 25 06:22:17 thats what i'm using Sep 25 06:24:01 I get the same error with that file... I get the feeling I'm missing a step here Sep 25 06:24:25 well.. try re-seating the fpc connectors Sep 25 06:24:30 Once I plug the board in to the USB is there any buttons to push? Sep 25 06:24:33 theyre a bit tricky Sep 25 06:24:49 gcb77 the button on the board should reset the moko Sep 25 06:25:16 also do sudo modprobe ftdi_sio vendor=0x1457 product=0x5118 Sep 25 06:25:27 that gives you 2 serials /dev/ttyUSB0 and 1 Sep 25 06:25:40 when you start openocd the first (0) vanishes Sep 25 06:26:00 the second (/dev/ttyUSB1) is where the real serial of the moko is connected. Sep 25 06:26:14 Is the moko supposed to be hooked up to the battery? Sep 25 06:26:21 use screen /dev/ttyUSB1 115200 or a similar tool to connect there Sep 25 06:26:23 jap Sep 25 06:26:40 hmm, power. Thought I might be missing something Sep 25 06:26:41 insert fpc, connect debugboard to power, insert battery, connect usb to moko Sep 25 06:26:47 i cant sync with http://buildhost.openmoko.org/snapshots/2007.08/ipk Sep 25 06:26:59 but i can ping openmoko.org ? what could this be ? Sep 25 06:27:27 best is when you use real differen ports for the debugport and the moko. i need to use one usb on the right and one on the left of my thinkpad Sep 25 06:31:47 yes, openocd is now connected! Sep 25 06:32:38 :) Sep 25 06:33:58 then do the reset_and_halt, upload lowlevel-foo, run it till breakpoint, upload uboot, run it, flash it with itself-game ;) Sep 25 06:39:18 Is there a way now to just flash a new uBoot, or do I have to reset NAND? Sep 25 06:43:09 any folks who could hook me up with a phone around? Sep 25 06:44:19 * clee wonders where fradeve is Sep 25 06:50:34 gcb77_ you should be able to just flash uboot if the rest of the flash is fine (oob areas, uboot env) Sep 25 06:56:46 roh: I think uboot is bad... would I flash it using the debug board? Sep 25 06:58:10 cu: creat (/var/lock/TMP00000073b2): Permission denied Sep 25 06:58:10 cu: /dev/ttyACM0: Line in use Sep 25 06:58:16 why is this ? Sep 25 07:03:59 gcb77_ jap. Sep 25 07:07:26 hm.. seems i should reencode this abcnews flash snippet to something people can watch Sep 25 07:09:23 roh: do I just run script.ocd, or does that reset all of NAND? Sep 25 07:10:33 oeh.. i dunno which script that is Sep 25 07:13:03 ah, I see. If I read this right, running Stage 1 should get me a new uboot, and leave the other NAND blocks alone Sep 25 07:13:58 gcb77_ http://wiki.openmoko.org/wiki/Bootloader#Using_JTAG_to_boot_from_RAM should have the correct stuff Sep 25 07:14:19 load lowlevel_foo, execute these few hundred bytes to get sdram up Sep 25 07:14:37 then uboot gets into ram and executed to be used to flash itself into nand again Sep 25 07:15:37 I'm using kermit to connecto to /dev/ttyUSB1... should I be getting a prompt there? Sep 25 07:15:58 when uboot is up and running, yes Sep 25 07:16:12 eh.. dunno about kermit..l never used that. i use screen for serials Sep 25 07:17:42 Switched to screen, and I get the same. I see "20480 bytes read: OK", then nothing Sep 25 07:19:27 My guess is I need to run through at at least Phase 1 of the devirginator, get a good uboot back Sep 25 07:21:33 sorry for beeing not much help right now.. i just got up and will need some time and caffeine to boot my brains through Sep 25 07:21:47 * gcb77_ knows that feeling Sep 25 07:23:50 heh.. works Sep 25 07:23:51 it seems that in my case roaming is not working Sep 25 07:23:56 i'm in US with an italian card Sep 25 07:24:02 it says it has some signal Sep 25 07:24:08 but nobody can reach me Sep 25 07:24:18 http://yamato.hyte.de/moko/ has the abc video in some more usable formats Sep 25 07:39:58 I'm not having much luck with the debug board Sep 25 07:42:37 soo .. finally my neo is going to be uptodate Sep 25 07:44:53 gcb77_: what's the problem? Sep 25 07:45:19 drath: I telnet to localhost on 4444, it connects, and get nothing Sep 25 07:45:56 you mean no output except for the telnet client telling you it connected? Sep 25 07:46:21 yes Sep 25 07:48:30 hum, it should greet you with Open On-Chip Debugger Sep 25 07:48:31 > Sep 25 07:48:48 did you change your .cfg since you pasted it an hour ago? Sep 25 07:49:04 I'm back to the original one I pasted Sep 25 07:49:13 I tried the one roh posted, with the same results Sep 25 07:50:21 can you run the OpenOCD with "-d -l ", and paste that? Sep 25 07:50:32 let it run for a few seconds, then kill it with SIGINT Sep 25 07:51:38 Error: server.c:159 add_service(): couldn't bind to socket: Address already in use Sep 25 07:51:48 kill all OpenOCD instances Sep 25 07:55:29 drath: ok, this time I got a prompt... somehow Sep 25 07:56:54 ok, you should always make sure that there's no leftover instance running Sep 25 07:57:04 make sure you're using the latest libftdi Sep 25 07:57:29 drath: yes, between the screen sessions, kermit, and telnet I must have locked something up Sep 25 07:57:43 I'm running with libs and binaries made by MokoMakefile Sep 25 07:59:03 ok, guess they should have the current version Sep 25 08:01:51 drath: is it safe to run "devirginate -1" or will that erase NAND? Sep 25 08:03:35 sorry, i don't know much about moko Sep 25 08:05:51 he he, I just wrote uboot at 0 Sep 25 08:09:32 * bedboi is away: sleeping Sep 25 08:26:29 does anyone have a solution to this? http://lists.openmoko.org/pipermail/openmoko-devel/2007-September/001558.html Sep 25 08:29:11 goodday #openmoko Sep 25 08:29:50 rwhitby, roh, drath thank you very much for the help, I got to the smiley face stage and will continue tomorow Sep 25 08:31:13 gcb77_: once you get the smiley face, the rest should be dfu-util stuff Sep 25 08:35:08 any suggestions on how to package GTK+ programs with .bb in the build environment (OM2007.2) ? Sep 25 08:49:25 roh: Nice report on abc, thanks for the link :) Sep 25 08:52:08 mbuf do you have a makefile which does linking and setting up cflags? Sep 25 08:52:35 roh, yes that part works Sep 25 08:52:50 roh, i get a binary ARM file, which I can run on the target; Sep 25 08:53:08 roh, i would like to create a .bb file that i would like end-users to just install and test Sep 25 08:53:44 roh, this works, http://free-opensource.qvantel.net/mediawiki/index.php/OpenMoko_Hello_World_Debian_20072#Standalone_Makefile Sep 25 08:55:43 roh, i am stuck here, http://lists.openmoko.org/pipermail/openmoko-devel/2007-September/001558.html Sep 25 08:56:24 mbuf hm.. i'm not sure.. but can you try DEPENDS = instead of DEPENDS += ? Sep 25 08:58:13 roh, nah, i get same error, gtk/gtk.h: No such file or directory Sep 25 08:59:01 hm... well.. i'll run into the same problems later when i need to build a package for a friend. ;) Sep 25 08:59:43 so i'll reply to your mail if not anyone with more bb experience did in the meantime.. need to move now.. Sep 25 08:59:46 bbl Sep 25 08:59:51 roh, i saw some of the .bb files in the oe sources, all just include "inherit openmoko" Sep 25 08:59:59 roh, thanks for your time Sep 25 09:00:20 mbuf yeah.. ive searched where the gtk includes are from, but it seems they come from libmokoui2 .. Sep 25 09:00:43 which makes sense.. i just don't get where c and ldflags are set via autoconf/automake Sep 25 09:01:09 thats where the includepath needs to come from so gcc can find gtk/gtk.h Sep 25 09:01:27 later Sep 25 09:01:35 roh, yup, later Sep 25 09:07:27 this is interesting for me too. i often wonder how that works. isn't there a wiki about that? how do developers at om work? do they always make bui..., package, install? Sep 25 09:07:53 emdete, most of their packages are integrated with the build environment, so i am not sure Sep 25 09:08:09 emdete, i believe ScaredyCat wrote the wiki HOWTO on local/overlay packages? Sep 25 09:08:50 for my package i set absolut path in the makefile - not the right solution Sep 25 09:09:32 emdete, i get an executable, that is fine; but I can't expect end-users to just test run it; it easier for them to install a package; Sep 25 09:10:19 same for me - i get an executable and rsync that to the neo Sep 25 09:11:03 or i get an ipk to install. but i source one of the scripts that bitbake generates.. :D also not the right solution Sep 25 09:12:49 i also wrote a configure script which just writes env & parameters to a logfile to understand where the pathes come from - did not see those :/ Sep 25 09:15:00 mbuf: have you build webkit? it uses qmake or am i wrong - how does /that/ work? i dont understand the dir structure of gdklauncher at all.... Sep 25 09:15:51 emdete, no idea about webkit; have to ask Mickey to use qmake; i like to stick with make :) Sep 25 09:17:04 is qmake from trolltech? Sep 25 09:19:39 emdete: yes Sep 25 09:30:25 strange... i don't understand gdklauncher at all :/ Sep 25 09:44:09 Happy Mid-Autumn Day, everyone! Sep 25 09:47:16 The dark half of the year is here.. :( Sep 25 09:48:21 or the hot half... Sep 25 09:52:30 99.5% of the earths population live south of my latiude Sep 25 09:54:05 been there.. done that Sep 25 09:54:40 Oh well, I have a bright monitor! :D Sep 25 09:55:06 i kinda miss the snow... Sep 25 09:55:32 Yes, it's worst before everything turns white Sep 25 09:56:01 nothin like a moonlight ski Sep 25 10:02:56 ljp: howdy :) don't want to brother you, but any news for me? Sep 25 10:18:49 (script) wiki RecentChanges: [[Press_Coverage]] Sep 25 10:27:22 Hi Sep 25 10:29:51 hello Sep 25 10:30:06 hi Sep 25 11:22:45 hrm... why does openembedded break every single time i try to build a new version? Sep 25 11:23:08 * nbd wonders how to fix the webkit-gtk compile breakage Sep 25 11:23:36 nbd: it's not OE, it's webkit-gtk which follows SVN and tends to break a lot... Sep 25 11:24:02 last time i had to fix 5 packages from OE before it compiled Sep 25 11:26:00 grüße aus Östereich :) Sep 25 11:26:08 how are you guys Sep 25 11:26:39 well me, I have managed to finaly brick my neo Sep 25 11:26:47 Congratulations! Sep 25 11:26:52 What did you do? Sep 25 11:27:00 i have been throug all the wiki s now Sep 25 11:27:16 And do you have a debug board. Sep 25 11:27:31 what the hack is OpenOcd and JTag Sep 25 11:27:40 yes i do hava a debug board Sep 25 11:27:45 What did you do? Sep 25 11:28:02 i connected the debug board to my neo Sep 25 11:28:24 i connected the debug board to my host Sep 25 11:28:41 i have my /dev/ttyUSB0 and 1 up now Sep 25 11:29:24 What initially did you do in order to brick it? Sep 25 11:29:58 the wrong thing Sep 25 11:30:12 i had to build libftdi-0.10 'cause ubuntu newses ist 0.7 Sep 25 11:30:18 well, i dont know Sep 25 11:30:32 I just hat it run for a couple of days Sep 25 11:30:35 doning nothing Sep 25 11:30:52 than i started my host again and gone it was Sep 25 11:30:55 Are you sure it's simply not a flat battery? Sep 25 11:31:03 hu? Sep 25 11:31:17 well i allway was connected Sep 25 11:31:27 but sure you are right Sep 25 11:31:38 Do you have a multimeter? Sep 25 11:32:11 Unless you were actively doing stuff - flashing uboot, or doing dd randomly to the flash device, it should not be brickable Sep 25 11:32:12 well, i think so Sep 25 11:32:26 how long do it need to load Sep 25 11:32:54 does it load while still not responding? Sep 25 11:33:53 http://wiki.openmoko.org/wiki/Neo1973_Battery_Charger Sep 25 11:34:03 thx Sep 25 11:34:09 See the info at the bottom Sep 25 11:41:23 the debug board dont work without a loaded battery in the neo, righ? Sep 25 11:43:29 possibly not. Sep 25 11:43:43 Certainly not I think actuallyh Sep 25 11:48:28 the thing is, i had neo starting the complete system before it stalled and never got back on Sep 25 11:50:00 what do you mean by complete system, and stalled? Sep 25 11:51:25 well it did boot and while starting up the UI it stopped playing the start sound, while making a nasty noise - than switch itself off Sep 25 11:51:49 That's exactly what would be expected if the battery was flat. Sep 25 11:51:56 oh Sep 25 11:51:57 ok Sep 25 11:52:07 The battery charging stuff is broken at the moment. Sep 25 11:52:37 well, how to charge it then? Sep 25 11:52:44 openmoko: 03thomas * r3036 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (3 files in 2 dirs): Sep 25 11:52:44 openmoko: * libmokoui/moko-finger-scroll.c: Sep 25 11:52:44 openmoko: * libmokoui/moko-search-bar.c: Sep 25 11:52:44 openmoko: Fix some typos in the documentation Sep 25 11:54:28 isnt it enought to let it load 10 hours now? Sep 25 12:07:12 I GOT IT BACK ON! Sep 25 12:07:28 how to get to the fast loading mode? Sep 25 12:07:44 i am in the u-boot menu now Sep 25 12:08:40 RalphEichelberge: Isn't it an option for that? Sep 25 12:08:53 ? Sep 25 12:08:58 I can't remember, since my uboot menu is broken :( Sep 25 12:09:13 i am sorry for you :( Sep 25 12:09:31 but no there isnt a optinon in the u boot Sep 25 12:09:34 RalphEichelberge: I guess http://wiki.openmoko.org/wiki/Neo1973_Battery_Charger says how to do it? Sep 25 12:10:05 RalphEichelberge: Ah, it's not default Sep 25 12:10:30 RalphEichelberge: Check http://wiki.openmoko.org/wiki/Bootloader#Battery_charger_related Sep 25 12:10:36 no there is only that there is a driver Sep 25 12:11:01 what mean set console to usb btw? Sep 25 12:13:24 however, guys, it is up again Sep 25 12:13:36 thx for beeing with me :) Sep 25 12:16:59 RalphEichelberge: Set console to usb is what you have to to to be able to use cu or minicom to /dev/ttyACM0 to access the neo in u-boot mode Sep 25 12:17:32 i have booted linux and prevented it from startig up X - now it ist in fast chage mode Sep 25 12:17:33 root@fic-gta01:/$ cat /sys/devices/platform/s3c2410-i2c/i2c-adapter/i2c-0/0-0008/chgmode Sep 25 12:17:33 fast_cccv Sep 25 12:17:51 i see - if i go to serial what happens? Sep 25 12:18:10 there is no serial, rightr? Sep 25 12:18:36 It's serial over usb :) Sep 25 12:18:45 really? Sep 25 12:18:47 ok Sep 25 12:18:59 now my mission is to read from SIM Sep 25 12:19:07 The thing is that it's the same serial that's used for communicating with the gsm modem Sep 25 12:19:10 how would you start ? Sep 25 12:19:19 What do you want to read from the sim? Sep 25 12:19:22 Doesn't that work? Sep 25 12:19:30 the SIM-Operator ID Sep 25 12:19:42 I'd use the gsmd console Sep 25 12:19:47 hey, how about somone to do somthing like this for the NEO ;P Sep 25 12:20:20 RalphEichelberge: http://wiki.openmoko.org/wiki/Gsmd#Usage_of_shell_mode Sep 25 12:21:52 well nice the Sep 25 12:21:56 sonar Sep 25 12:22:10 moin Sep 25 12:22:12 i did the shell mode allready Sep 25 12:22:32 santaclaus666: Heh, cool thing Sep 25 12:22:36 but there is no option to read the SIM's Data Sep 25 12:22:56 i guess there are AT-command Sep 25 12:22:59 s Sep 25 12:23:01 to do so Sep 25 12:23:02 santaclaus666: I wonder if you can use the stereo speakers on the neo for even better accuracy Sep 25 12:23:14 There's a list of all AT-commands in the wiki too Sep 25 12:23:30 yes - i got the pdf too Sep 25 12:23:36 yah ;) probably it can be combined in some way with the 3d acceleromerter and the gps ;) Sep 25 12:23:46 now i only have to find a working image again Sep 25 12:23:50 RalphEichelberge: Hey, you're a step ahead every time now ;) Sep 25 12:24:12 CM :) Sep 25 12:24:53 there was a URL with uptodate snapshots of the image - do you kown it? Sep 25 12:25:29 hi Sep 25 12:28:29 RalphEichelberge: You mean ScaredyCat's images? Sep 25 12:28:35 Or the official build host? Sep 25 12:29:55 hey placid Sep 25 12:30:00 I don't know if there's any images availible with mwester's patches that fix gsmd Sep 25 12:30:39 I think that it may be skipped by build host because of the sane rev mechanism Sep 25 12:31:23 Heh, I haven't seen a sane rev for almost a month.. ;) Sep 25 12:32:03 RalphEichelberge: Here's some fairly recent images at least: http://buildhost.automated.it/OM2007.2/ Sep 25 12:49:40 CM : mwester's fix, have you seen it? Would it be something like: stty -F /dev/ttyACM0 -echo ? Sep 25 12:50:55 sannes: Look at http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=788#c20 Sep 25 12:52:04 ok again me Sep 25 12:52:06 Haven't tried that latest patch either Sep 25 12:52:14 Oh no! :P Sep 25 12:53:40 i am running a neo v4 Sep 25 12:53:55 and have that problem Bug 256 Sep 25 12:54:06 is there a way to get it to work? Sep 25 12:54:24 i know - with qtopia it works fine Sep 25 12:55:47 is there a difference in the kernel? Sep 25 12:57:02 No, it's the same kernel Sep 25 12:57:18 well where is the difference? Sep 25 12:57:40 gsmd vs qtopias way to handle the gsm modem Sep 25 12:58:17 yes i know Sep 25 12:58:22 they dont have gsmd Sep 25 12:58:41 do you know how they do communicate with the modem? Sep 25 12:58:51 MAybe ask ljp about it? Or in #qtopia Sep 25 12:59:00 i see Sep 25 12:59:08 thx ill try Sep 25 12:59:24 :) Sep 25 13:02:03 RalphEichelberge: what do you mean? Sep 25 13:02:25 emdete Sep 25 13:02:43 i just wanted to talk to my modem in my neo Sep 25 13:02:50 but it hangs Sep 25 13:03:05 RalphEichelberge: qtopia supports gsm0710 in userland. the use there event mechanism between the apps that use the modem Sep 25 13:03:21 RalphEichelberge: no hope to use that model in om Sep 25 13:03:30 i know Sep 25 13:03:36 RalphEichelberge: did you follow the wiki? Sep 25 13:03:44 but if i want to use cu it hangs Sep 25 13:03:51 like in that buch 256 Sep 25 13:03:55 bug Sep 25 13:03:56 emdete: well the usage of QCOP is just an implementation detail, with probably a one line change it is using dbus Sep 25 13:03:59 did you reset via /sys...? Sep 25 13:04:33 i did what i found in http://wiki.openmoko.org/wiki/Manually_using_GSM Sep 25 13:04:37 zecke: you won't want to use d-bus for every 32byte package especially for gprs/ppp Sep 25 13:04:59 does anbody have an idea how i can get a known working webkit-gtk version in my oe/om build tree? Sep 25 13:05:07 emdete: what makes you think grps is going to use QCOP/dbus? Sep 25 13:05:29 nbd: the last time I checked webkit-gtk was pinned to a certain revision Sep 25 13:05:36 zecke: ppp is on top of one channel of gsm0710 or am i wrong? Sep 25 13:05:49 zecke: well, mine apparently builds a broken one Sep 25 13:06:10 complains about missing libc symbols Sep 25 13:06:10 emdete: right, but at least they used to exec a helper to use such a channel exclusively Sep 25 13:06:28 emdete: so for PPPD it used to dial the number and then passed control over to pppd Sep 25 13:06:40 zecke: i would like a more unix-like aproach better Sep 25 13:07:10 zecke: using qcop all the time per package as i understand that Sep 25 13:08:25 also if i use libgsmd-tool -m atcmd Sep 25 13:08:56 i get this: Sep 25 13:08:58 root@fic-gta01:~$ libgsmd-tool -m atcmd Sep 25 13:08:58 libgsm-tool - (C) 2006 by Harald Welte Sep 25 13:08:58 This program is Free Software and has ABSOLUTELY NO WARRANTY Sep 25 13:08:58 at Sep 25 13:08:58 STR=`at' Sep 25 13:08:59 root@fic-gta01:~$ libgsmd-tool -m atcmd Sep 25 13:09:01 libgsm-tool - (C) 2006 by Harald Welte Sep 25 13:09:03 This program is Free Software and has ABSOLUTELY NO WARRANTY Sep 25 13:09:05 Can't connect to gsmd Sep 25 13:09:07 root@fic-gta01:~$ Sep 25 13:09:17 RalphEichelberge: then gsmd is not running at all Sep 25 13:09:20 ? Sep 25 13:09:33 well it ran before Sep 25 13:09:35 RalphEichelberge: i would start wth trying to use at in /dev/ttyACM0 drectly Sep 25 13:09:43 RalphEichelberge: it dies often Sep 25 13:09:47 see /tmp/gsm.log Sep 25 13:09:57 you will notice the 'modem dead' message Sep 25 13:10:13 after that gsmd exits - don't know why Sep 25 13:10:51 yes Sep 25 13:11:11 did you try http://wiki.openmoko.org/wiki/Manually_using_GSM? Sep 25 13:11:13 and before that : gsmd_usock_user_cb() EOF, this client has just vanished Sep 25 13:11:22 yes Sep 25 13:11:24 RalphEichelberge: yes, that ist libtool... Sep 25 13:11:55 you know what - libgsmd-tool worked fine in an other build Sep 25 13:12:02 hmm, last build on buildhost from the 17th Sep 25 13:12:06 still Sep 25 13:12:19 can u give me that url Sep 25 13:12:26 please Sep 25 13:12:27 two more days and we're without a new build for ten days Sep 25 13:13:53 and now my neo goes suspend and kills my ssh session every time Sep 25 13:14:12 suspend? wow - cool Sep 25 13:14:17 :-/ Sep 25 13:14:50 i just pressed the power button than it flashed completey white Sep 25 13:14:59 and went back to UI Sep 25 13:15:34 strange Sep 25 13:15:41 where is that build host ? Sep 25 13:15:52 hhf423! Sep 25 13:16:03 second Sep 25 13:16:22 i am building one my own at the moment - will last till tomorrow i guess Sep 25 13:16:32 http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/images/fic-gta01/?C=M;O=D Sep 25 13:16:51 the list of all repositories is here: http://wiki.openmoko.org/wiki/Repositories Sep 25 13:17:40 thx hhf423 Sep 25 13:18:50 how about ScaredyCat's builds? Sep 25 13:18:52 does someone have ipk from gsmd with all the recommended patches? Sep 25 13:21:29 SaredyCat can I take the uImage from 11 Sep while using the jffs2 from 25 th? Sep 25 13:26:28 What value does PCLK usually have? Sep 25 13:26:34 in GTA01 Sep 25 13:49:46 any ETA on the consumer version of the openmoko? Sep 25 13:50:43 i heard of a "working dialer" - where is it? Sep 25 13:50:57 Tjikkun_work http://wiki.openmoko.org/wiki/Neo1973#.22Phase_2.22_.28GTA02.2C_.22Mass_Market.22.29 Sep 25 13:51:57 but that date does not necesarily include all the software to make it realy a mass-market device... Sep 25 13:52:12 ok - some now how to kill the starting sound on neo? Sep 25 13:52:48 http://www.rwhitby.net/blog/openmoko/stopping-the-openmoko-startup-sound.html Sep 25 13:53:14 goxboxlive: thanks Sep 25 13:53:49 welcome Sep 25 13:53:51 Mek: I don't care :) I will also write/port some of my own software Sep 25 13:53:54 openmoko: 03thomas * r3037 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (4 files in 2 dirs): Sep 25 13:53:54 openmoko: * libmokoui/moko-finger-scroll.h: Sep 25 13:53:54 openmoko: * libmokoui/moko-hint-entry.h: Sep 25 13:53:54 openmoko: * libmokoui/moko-search-bar.h: Sep 25 13:53:54 openmoko: Document enums and structs Sep 25 13:54:02 thx - wow it didnt anyway Sep 25 13:54:18 scardycat probably allready did that patch :) Sep 25 14:01:38 openmoko: 03chris * r3038 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (ChangeLog libmokoui/moko-search-bar.c): Sep 25 14:01:38 openmoko: * libmokoui/moko-search-bar.c: (moko_search_bar_class_init): Sep 25 14:01:38 openmoko: Document signals Sep 25 14:02:35 Can anyone build webkit? Sep 25 14:02:49 shure Sep 25 14:03:32 what's your problem with it? Sep 25 14:03:57 .../libgdk-x11-2.0.so: undefined reference to `cairo_xlib_surface_create_for_bitmap Sep 25 14:04:02 and may more like it. Sep 25 14:04:07 s/may/many/ Sep 25 14:04:07 mwester meant: and many more like it. Sep 25 14:04:19 strange Sep 25 14:04:41 I synced up about 8 hours ago and started it, I guess I'll try again. Sep 25 14:05:07 (frustrating, 'cause I have a patch for the dialer "can't set antenna power" dialog box problem, and I can't test it) Sep 25 14:06:25 openmoko: 03thomas * r3039 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (3 files in 2 dirs): Sep 25 14:06:25 openmoko: * libmokoui/moko-finger-scroll.h: Remove redunent comment Sep 25 14:06:25 openmoko: * libmokoui/moko-hint-entry.c: (moko_hint_entry_class_init): Add missing Sep 25 14:06:25 openmoko: documentation Sep 25 14:07:41 thomasg: do any of the commits on libmokoui2 fix the problem where one can't start the dialer? Sep 25 14:08:10 mwester: I'm sure you do not mean me :) Sep 25 14:09:12 Oops, sorry! Sep 25 14:10:08 I build webkit just a while ago, and it works, but midori is broken Sep 25 14:13:26 hello CM Sep 25 14:13:45 if I go: Sep 25 14:13:45 cat /dev/ttySAC0 & Sep 25 14:13:50 and do Sep 25 14:14:05 echo "AT" > /dev/ttySAC0 Sep 25 14:14:09 Anyone able to access a sd-image from u-boot in qemu? Sep 25 14:14:28 i get a "AT" back - but no "OK" - why? Sep 25 14:15:51 i did everything like in wiki manually using GSM Sep 25 14:15:59 but cu is not installed so.. Sep 25 14:16:15 cu? Sep 25 14:16:32 You can't use the serial modem at the same time as the gsm, iirc Sep 25 14:16:48 access the GSM Modem UART root@fic-gta01:~$ cu -l /dev/ttySAC0 Connected. AT-Command Interpreter Ready OK Sep 25 14:17:05 wiki says Sep 25 14:17:10 Ah, sorry. I thought you meant from the pc Sep 25 14:17:21 Yes, from the phone it should work Sep 25 14:17:44 yes - and my variation should work too Sep 25 14:17:56 but i just get the echo - no responce Sep 25 14:18:00 Did you try diabling the hw flow control? Sep 25 14:18:28 check http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=788#c15 Sep 25 14:18:29 (script) wiki RecentChanges: [[User:ChrisLord]] [[User:Zash]] Sep 25 14:18:31 i did stty -F /dev/ttySAC0 crtscts Sep 25 14:18:40 Ok, and that didn't help? Sep 25 14:18:42 RalphEichelberge: You need to do much more than "echo" to get the GSM modem to work. Disable echo as well. Sep 25 14:18:44 Weird.. Sep 25 14:19:05 RalphEichelberge: There's some who knows! :) Sep 25 14:19:08 stty -F /dev/ttySAC0 crtscts -echo Sep 25 14:19:09 how to get a OK afer a AT Sep 25 14:19:24 Also, make sure that you enable the gsm modem Sep 25 14:19:41 The best way is to use gsmd; cu will work as well. Sep 25 14:21:45 What image are you running on your phone that doesn't have cu? Sep 25 14:23:43 http Sep 25 14:23:51 sorry Sep 25 14:25:26 does anybody else has a problem flashing using flash.sh with qemu-neo1973? Sep 25 14:25:38 dpavlin: yes Sep 25 14:27:41 mwester: OpenMoko-scaredycat-openmoko-devel-image-glibc-ipk-P1-September-Snapshot-20070925-fic-gta01.rootfs Sep 25 14:27:52 viq: Turning off gadgetfs and rebuilding it might help. Sep 25 14:28:25 gadgetfs? Where is that? Sep 25 14:29:13 if you want to use networking between linux host and qemu... Sep 25 14:29:36 it works for me when gadgetfs is disabled (by default it is), but when I compile it with gadgetfs it doesn't... Sep 25 14:29:58 gadgetfs is explained at http://wiki.openmoko.org/wiki/Qemu#Setting_up_USB_connection Sep 25 14:31:19 oh, thanks Sep 25 14:31:20 hi Sep 25 14:36:20 dpavlin: it doesn't seem gadgetfs is enabled, at least I didn't do anything to enable it. Sep 25 14:37:32 RalphEichelberge: check to see if /usr/bin/cu exists Sep 25 14:38:21 dpavlin: i use gadget drivers. currently wogking on g_audio.o for the moko (and other arm devices..) Sep 25 14:39:05 g_audio? would that make the phone appear as a usb audio device i take it? Sep 25 14:39:13 youp Sep 25 14:39:15 yes Sep 25 14:39:22 nice Sep 25 14:40:27 24 audio soundcard. Sep 25 14:40:30 plug it in, whatever you want for routing, etc. Sep 25 14:40:35 when it works. Sep 25 14:40:54 midi too, so 'total integration' .. Sep 25 14:41:04 torpor you start from g_midi ? Sep 25 14:41:21 sort of want to push idi into the neo to stave off the OSC crowd from doing naughty deeds with bluetooth .. ;) Sep 25 14:42:12 BabelO0: yes, i have g_midi up too .. MidiShare integration is not done on neo, but thats what i'm working on at the moment, anyway. Sep 25 14:42:43 How good is the neo's speaker? Sep 25 14:42:48 torpor : so good i ve no sound card at work, so may be usefull ;) Sep 25 14:43:02 heh heh Sep 25 14:43:12 plug in the headphones to listen to your PC's soundcard .. Sep 25 14:43:42 that can definitely be done through gadget, and the nice audio capabilities are highly exploitable, imho. Sep 25 14:43:51 musicaly imean. Sep 25 14:44:30 torpor yes agree with you last time i plug my HTC Blueangel on office computer load module gmidi and play so mid file :) Sep 25 14:44:47 great! Sep 25 14:45:26 where should I submit patch for --enable-gadgetfs for qemu-neo1973? bugs? Sep 25 14:45:49 straight into the openmoko tree, i reckon, or even through rafb.net if you want testing. Sep 25 14:46:11 there is a bugtrac too but ican never remember what it is to be frank. Sep 25 14:46:51 torpor: care to enaborate on that since I don't have svn commit access? Sep 25 14:49:24 i don't either thats the thing. Sep 25 14:49:41 if you have a patchset and want me to test compiling/running it, i can grab it from you on rafb.net, tho. Sep 25 14:50:36 it's trivial change to configure, I don't think that testing is what is needed :-) Sep 25 14:51:13 oh i thought you meant a module for gadget or something .. Sep 25 14:52:26 torpor: http://rafb.net/p/t4YwsN14.html is the patch... Sep 25 14:52:52 I will try to submit it upstream too... Sep 25 14:53:11 see i have no idea about the public svn, i guess i keep missing that stuff on my rush to ignore the wiki. Sep 25 14:53:23 i'm sure there's a bot in here tho, who knows the keyword.. Sep 25 14:56:27 openmoko: 03thomas * r3040 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (7 files in 3 dirs): Sep 25 14:56:27 openmoko: * doc/reference/tmpl/moko-stock.sgml: Add stock item documentation Sep 25 14:56:27 openmoko: * libmokoui/moko-finger-scroll.c: Sep 25 14:56:27 openmoko: * libmokoui/moko-hint-entry.c: Sep 25 14:56:27 openmoko: * libmokoui/moko-search-bar.c: Sep 25 14:56:29 openmoko: * libmokoui/moko-stock.c: Sep 25 14:56:31 openmoko: Add section documentation Sep 25 15:05:52 so fingerscroll is a widget .. ? Sep 25 15:12:45 torpor it is a gtk-scroll-window i suspect Sep 25 15:16:44 kristian-m: looks like it .. might try to use it for something .. Sep 25 15:18:48 torpor hmm, it's a gtk container... Sep 25 15:28:43 quick test (sorry) :) Sep 25 15:30:03 failed Sep 25 15:31:30 depends on whether he was testing connectivity or his netiquette skills. :) Sep 25 15:34:01 hey all Sep 25 15:34:43 can anyone tell me what comes in the advanced kit that you dont get in the base? Sep 25 15:35:22 debug board Sep 25 15:35:27 fiberspy: dude, read the website Sep 25 15:35:38 fiberspy: http://openmoko.com/products-index.html Sep 25 15:36:06 thx, was looking at the wiki and shop page...lol Sep 25 15:48:40 mwester: ping Sep 25 15:59:44 abraxa_: pong Sep 25 16:00:03 * bedboi is back (gone 07:50:31) Sep 25 16:00:29 mwester: Today's commits didn't fix the libmokoui2 bug and I couldn't find a bugzilla entry either. Do you happen to know of one? Sep 25 16:03:31 (script) planet: Thomas Wood: libmokoui2 documentation http://blogs.gnome.org/thos/2007/09/25/libmokoui2-documentation/ Sep 25 16:04:02 abraxa_: I haven't looked for a bugzilla entry. I think we need to create one, even if its a duplicate. Sep 25 16:04:31 mwester: Okay, I'll do that then in a bit - the ohand guys are probably home already anyway Sep 25 16:05:47 (funny how Thomas just blogged and raves about libmokoui2, too, heh :) Sep 25 16:18:12 (script) wiki RecentChanges: [[Mokopedia]] [[Hardware:Neo1973:Alternate_Cases:Camera]] [[Hardware:AT_Commands]] Sep 25 16:20:08 Is there an overview of the address space of the s3c24xx? Sep 25 16:23:18 The datasheet? Sep 25 16:24:16 yep, I have that. But I thought maybe someone has done a nice graphic in the wiki or something like that :) Sep 25 17:11:58 * hhf423 sighs Sep 25 17:12:51 any news on GPS? Sep 25 17:14:59 I think mickey|portugal was working with Broadcom on that, so I don't expect anything new for a couple of weeks. (and the cynic in me says it's doubtful that Broadcom will have done anything at all other than stall this effort more by then) Sep 25 17:18:37 hi guys, any news? Sep 25 17:19:29 ~seen zdanek Sep 25 17:19:37 zdanek was last seen on IRC in channel #openmoko, 3d 19h 18m ago, saying: 'cu'. Sep 25 17:23:41 counter Sep 25 17:23:42 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter Got it, get yours from openmoko.com (and read topic) (1812;252) Sep 25 17:49:53 mickey|portugal: If you're adding packages to the Sept image, add fsck.vfat and the kernel vfat module (for microSD cards) Sep 25 17:50:52 gnu: I'm pretty sure the vfat module is already in the image, or am I mistaken? Sep 25 17:51:07 mwester: it's not in the August image. Sep 25 17:51:25 (fsck.vfat would be good, especially given how difficult it is to remove the card) Sep 25 17:52:40 -rw-rw-r-- 1 root root 12596 Jan 1 1970 vfat.ko Sep 25 17:53:01 Looks like it's in the current versions already. Sep 25 18:15:40 so whats the current status of the Neo1973? Sep 25 18:15:40 narcoclepsy: 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.) Sep 25 18:18:12 (script) wiki RecentChanges: [[Translation]] [[Main_Page/sk]] [[Buying_Interest_List]] Sep 25 18:18:14 ok to be more verbose having read what it can currently do... can it reliably make and receive phone calls at this juncture? Sep 25 18:18:58 should the 2.6.22.5-moko11 kernel support sdhc? Sep 25 18:19:08 it seems the neo1973 certainly can do that at least with qtopia software, I believe the openmoko software is still not that reliable... Sep 25 18:19:43 I tried with a sandisk 6gb, and while it is correctly recognized (the /sys stuff seems fine), all reads from mmcblk0 only give "error 1 transferring data" Sep 25 18:23:18 pH5, stefan tested it with a 4GB card, which worked ok Sep 25 18:24:42 pH5: thomasg had a 6 gb microsd sdhc card and had no problems Sep 25 18:25:02 I still have :) Sep 25 18:25:25 pH5: I have no problem at all (except booting, but this seems to be uboots fault) Sep 25 18:25:31 *has Sep 25 18:25:44 thomasg: with the 22.5-moko11 kernel? Sep 25 18:25:49 yes Sep 25 18:25:56 moko11-r2 Sep 25 18:26:20 ok, thanks. time to recompile with mmc debug enabled, then. Sep 25 18:26:25 :) Sep 25 18:26:43 you can take a look in the wiki (supported microSD cards) to see the exact type of card I have Sep 25 18:37:54 ye frigging goods i just had to respond to an sms using keypad, pure horror. and people complain about touchscreens *rolleyes* Sep 25 18:40:41 buz: depends on what you are used to :) Sep 25 18:40:47 touchscreen of course Sep 25 18:41:24 and just by what standard it is considered sane behavior to keep on writing capitals after writing one is beyond me, too Sep 25 18:43:57 What's a good page for like "current status on important features"? Sep 25 18:44:03 (on the wiki) Sep 25 18:45:20 jorgenpt: there is no such page Sep 25 18:45:38 only bits and pieces on the lists, bugzilla, and here Sep 25 18:46:46 What's the page that has tests of various builds? Sep 25 18:50:10 hi @ all.. I have a question to someone who owns a Neo already... How easily do you touch a touchbutton with your ear while doing a phone call? or does this not really happen? Sep 25 18:51:57 unique: with other touchscreen phones, it does happen. but i guess you could easily solve it by better interface design Sep 25 18:52:20 suppose you steal the idea from qtopia, and use a phone and handset that need to meet to hang up Sep 25 18:52:37 thining about it, that's genius ;) Sep 25 18:53:26 buz: well I guess I'll wait for the mass market version and ask that question then again ;) Sep 25 18:54:40 unique: Haven't had problems with it. Sep 25 18:55:33 LarstiQ: well, that's good Sep 25 18:58:17 I'm to 80% sure that I'll buy a Neo when it is finished, but first I'll wait for tests, and reports. Sep 25 19:03:01 http://www.linuxdevices.com/news/NS9554238577.html the qwerty one looks interesting save for the 2.4 stuff Sep 25 19:12:12 Nice kb: http://www.reghardware.co.uk/2007/09/25/slim_bluetooth_keypad/ Sep 25 19:13:34 cm yeah nice Sep 25 19:15:15 now add real buttons instead of that razr'esqe stuff and i'll buy it Sep 25 19:16:01 there actually is a thing this small with real buttons but it uses usb, on the plus side, it's <20usd Sep 25 19:16:56 http://www.cartft.com/catalog/il/478 Sep 25 19:17:06 (i've seen it chaper than that for sure) Sep 25 19:18:05 in fact, shrink it some more and glue it to neos front ;) Sep 25 19:20:25 buz: hehe Sep 25 19:44:32 Any idea why the panel crashes with the battery-plugin in it? Sep 25 19:46:42 hello Sep 25 19:46:52 I'm almost more surprised when the panel doesn't crash ;) Sep 25 19:48:19 CM - Remove the battery thing from its startup. Sep 25 19:48:29 * CM nods Sep 25 19:49:40 Which is a mess, because it's the most useful part in the panel. Sep 25 19:49:56 Okay, next to the keyboard. Sep 25 19:52:46 thats odd, the battery plugin worked allready years ago Sep 25 19:53:10 borg_ - Right. Sep 25 19:53:22 one step forward two steps back Sep 25 19:53:41 matchbox-panel-2 --start-applets systray,startup \ --end-applets openmoko-panel-battery,keyboard --titlebar & Sep 25 19:54:03 That's my current start-up line. Which shows a battery symbol for some little time, than crashes. Sep 25 19:56:34 i have problems with packaging openmoko-devel-image Sep 25 19:56:40 It's a pity that you could not extend the contract. ....... Sep 25 19:56:44 ups Sep 25 19:56:45 wrong paste ;) Sep 25 19:56:50 | ERROR: Cannot satisfy the following dependencies for task-openmoko-pim: Sep 25 19:56:50 | libspiff0 (>= 0.7.2) liburiparser0 (>= 0.5.1) Sep 25 19:56:54 any ideas? Sep 25 20:04:25 Elrond: good catch - I was trying to debug the gsm plugin on the panel, because I just jumped to the conclusion that was why it was crashing! Sep 25 20:04:30 * ScaredyCat blinks Sep 25 20:05:06 * mwester would have thought ScaredyCat would jump when startled, not blink... :p Sep 25 20:05:34 * ScaredyCat sleeps Sep 25 20:06:08 * ScaredyCat jumps startled Sep 25 20:18:52 (script) wiki RecentChanges: [[Current_events]] [[Package_management]] Sep 25 20:31:55 So commits get mentioned in this channel? Sep 25 20:34:07 Yep. Sometimes more than once. Sep 25 20:35:19 mwester: interesting, just from devs, or from all the projects? Sep 25 20:36:03 * psymin waits for GTA02. Sep 25 20:36:12 mwester - That was my first conclusion. Especialy, as my neo doesn't have gsmd. ;o) So I removed the gsm-plugin. But the panel still crashed a lot. Sep 25 20:36:17 * bkruse has been waiting for awhile Sep 25 20:36:27 bkruse: On this channel the bot alerts on the commits by the developers. Sep 25 20:36:36 mwester: ahh, ok Sep 25 20:37:01 I haven't seen any commit messages for the projects, just the core software. Sep 25 20:37:42 mwester: that would be cool for the projects too, then you could /ignore if you wanted... Sep 25 20:38:19 I think so. And if any project gets so much traffic that it's disturbing, well, they probably will have an IRC channel for that project anyway. Sep 25 20:40:46 I'd guess, that each project has to setup the cia stuff on their own? Sep 25 20:41:24 I would presume so. There is at least one website that can do that for you, but I don't remember the URL. Sep 25 20:41:29 mwester: exactly, im interested in some without having to go check on them, i believe the project manager has that option Sep 25 21:36:33 Is buying a Neo Advanced the only way for the public to get a debug board (for unbricking)? Sep 25 21:37:01 you can't brick the mass-market neo without one Sep 25 21:37:30 good deal :) Sep 25 21:37:44 at least, you know, by flashing Sep 25 21:37:54 you can still brick it with a brick Sep 25 21:38:36 * psymin chuckles. Sep 25 21:40:39 or $liquid_of_choice ;) Sep 25 21:41:27 (oh yeah, flashing it 100k times will probably also do it) Sep 25 21:41:50 mjr: what, you're a flasher? ;) Sep 25 21:42:37 I'm a little bit wii, and a little bit woo. I'm a geezer. Sep 25 21:42:41 mostly I'm interested in the bricking methods that actually allow UNbricking with the debug board .. if it can unbrick from $liquid_of_choice then I'm sold on it :) Sep 25 21:43:18 sure, just use $paper_towel_of_choice Sep 25 21:43:26 thats currently the best solutions Sep 25 21:43:42 yeah anyway, the mass-market neo has a known working backup bootloader in nor flash, and you can't reflash (ie. mess up) that without the debug board Sep 25 21:43:58 Thats great! Sep 25 21:44:22 now, lets just wait tiill gta02 comes out :) Sep 25 21:44:43 sounds like Christmas for Mass Market GTA02? Sep 25 21:45:08 sure hope so Sep 25 21:45:19 would make a nice presant Sep 25 21:45:24 at least to get Sep 25 21:58:31 mwester: I also get mokomakefile commits to come here. Sep 25 21:58:59 Elrond: I know how to set it up if any other projects want to notify commit messages here. mickey|portugal said he was happy for that to happen. Sep 25 22:19:11 (script) wiki RecentChanges: [[PyFeedRead]] Sep 25 22:49:56 is there any portable shared memory IPC library? Sep 25 22:52:18 how portable? There's the oll SysV SHM, and POSIX SHM Sep 25 22:55:36 windows, linux, bsds Sep 25 22:55:43 well linux, bsds, windows :) Sep 25 22:56:45 windows pretty much breaks the compatibility here... Sep 25 22:56:54 yep i know. Sep 25 22:57:01 dbus is too slow for me Sep 25 22:57:12 it adds 1.6-3.0 ms of latency Sep 25 23:01:25 What are you doing, bedboi? Sep 25 23:03:16 I'd probably just do my own shm wrapper functions for windows and standard OSes separately Sep 25 23:05:32 bedboi: Well there is the old mm library that was used in apache: http://www.ossp.org/pkg/lib/mm/ Sep 25 23:06:07 bedboi: But there is no port to win32 to the best of my knowledge Sep 25 23:06:15 ah Sep 25 23:06:18 so it's useless Sep 25 23:06:28 the only think that i could think of is dbus Sep 25 23:06:52 'course you could check cygwin for sysv/posix shm implementations Sep 25 23:07:46 the problem is that i'm measuring 3-4 ms of delays Sep 25 23:07:49 in D-BUS Sep 25 23:08:21 bedboi well.. it also depends on what you want to do. dbus is not for bulk data, only for ipc/rpc Sep 25 23:08:41 yep Sep 25 23:08:58 i need to send at very high rate data from one module to another Sep 25 23:09:11 actually i have 1 - N communication Sep 25 23:09:20 N modules communicate with this guy Sep 25 23:09:21 bedboi then dbus is propably not your solution Sep 25 23:09:32 so then what? Sep 25 23:09:37 socket? Sep 25 23:09:49 well.. what is it what you want to do? Sep 25 23:09:54 might I ask how any of this is related to openmoko? :P Sep 25 23:10:06 * thseiler points our that windows does not have unix domain sockets Sep 25 23:10:25 Mek: Interesting question ;-) Sep 25 23:10:46 Mek: the application that i'm going to run is going to work with openmoko Sep 25 23:10:59 but it has to be portable Sep 25 23:11:06 (i always write portable stuff) Sep 25 23:11:38 * mwester goes for pragmatic Sep 25 23:11:48 on the neo i'm going to expect huge performance degradation Sep 25 23:12:05 Mek: so i guess it's quite related to openmoko Sep 25 23:12:24 bedboi, what's the app about? Sep 25 23:12:26 Writing portable stuff is good, but some operating systems just are hostile to portability, so *shrug* :] Sep 25 23:12:33 cb22: it's a networking application Sep 25 23:12:34 okay, I wasn't really critizizing, just curious :) Sep 25 23:12:59 anyway, still'd go with the custom wrappers, shouldn't be too much of a problem Sep 25 23:13:03 cb22: it's something very new. I'm working in a research project founded by BBN and NSF Sep 25 23:13:26 BBN and NSF? Sep 25 23:13:31 yep Sep 25 23:13:43 BBN is the one that deployed the internet as we know Sep 25 23:13:50 NSF is National Science Foundation Sep 25 23:13:52 bedboi: You might be able to build a wrapper around CreateFileMapping (win32) and mmap (linux) but it will be #ifdef hell, depending on how much locking you will have Sep 25 23:14:01 yep Sep 25 23:14:12 sounds interesting Sep 25 23:14:38 cb22: yes it is. it is actually VERY interesting. It's redesigning the whole concept of internet. Sep 25 23:14:54 the project calls Future INternet Design (FIND) Sep 25 23:15:20 wow, got any url / link with more infos ? Sep 25 23:15:26 yeah Sep 25 23:15:26 wait Sep 25 23:15:32 ind.isi.edu Sep 25 23:15:35 find.isi.edu Sep 25 23:15:46 http://www.tna.cnri.net/Welcome.html Sep 25 23:15:51 this is better Sep 26 00:13:16 * cesarb is starting to wonder if he being the only one not hearing the startup sound might be because of some bashism somewhere, or something being broken on x86-64 Sep 26 00:13:44 Could someone with a standard build get me the output of "find build/tmp/rootfs/"? Sep 26 00:13:57 I'll do that Sep 26 00:17:02 cesarb: http://cl1p.net/find.build.tmp.rootfs/ Sep 26 00:17:26 cesarb: I may have a few other things in there, but it should be a clean build from ~12 hrs ago Sep 26 00:20:25 sagacis: looks like your uploaded file was cut off in the middle :( Sep 26 00:20:37 look at the last line: build/tmp/rootfs/usr/lib/ipkg/info/gsmd-plugin-v Sep 26 00:20:54 (script) wiki RecentChanges: [[Press_Coverage]] [[Neo1973_Hardware]] Sep 26 00:22:02 hm, I do have a correct build/tmp/rootfs/etc/pulse/session... Sep 26 00:22:43 * cesarb wonders if pulseaudio is running on the neo... Sep 26 00:24:24 my neo has been left plugged into the computer for days and still it wont turn on Sep 26 00:25:17 Have you pulled the battery first? Sep 26 00:25:19 orzo: some people had these problems with neos left plugged for too long Sep 26 00:25:57 well i pulled the battery now and reinserted teh batter and replugged Sep 26 00:26:01 is that going to do anything? Sep 26 00:26:56 cesarb: I'll try with a file. Sep 26 00:27:28 hm, pulseaudio is running, so it's not that... Sep 26 00:28:44 cesarb: again. I attached a file to http://cl1p.net/find.build.tmp.rootfs/ Sep 26 00:28:45 orzo: do you own a multimeter? Sep 26 00:28:50 no Sep 26 00:29:23 cesarb: You should be happy to not get blasted away every time you start the phone. I removed it from the session on mine. Sep 26 00:29:26 it was plugged in a long tmie, but then the computer it was plugged into lost power Sep 26 00:29:36 and when i finalyl came back to it, it would not power up Sep 26 00:29:43 and its been like that ever since Sep 26 00:32:43 Ok - that's pretty much expected - it will discharge so far it won't boot enough to charge properly. Sep 26 00:33:14 Have you - after it wouldn't power up - done the remove battery for 10 mins, put back in, put in USB charge lead for several hours thing? Sep 26 00:33:20 sagacis: no difference which could explain why I don't get gui sounds Sep 26 00:33:24 sagacis: thx Sep 26 00:33:31 no Sep 26 00:33:38 i removed the batetry for a few seconds Sep 26 00:33:59 should i unplug it and remove the battery? Sep 26 00:34:13 is there a wiki page on this whole issue? Sep 26 00:35:46 yes Sep 26 00:35:49 hang on Sep 26 00:37:26 hmm. i just plugged in my openmoko to charge... and its screen is flashing off and on rapidly. is this normal? Sep 26 00:38:19 http://wiki.openmoko.org/wiki/Neo1973_Power_Management - USB. Sep 26 00:38:33 -> juri_, orzo Sep 26 00:38:44 it may do that in some charge states juri_. Sep 26 00:39:42 You do have a battery in it, right? (juri_) Sep 26 00:40:06 (the power management is, um, "bad" if it is plugged in sans battery) Sep 26 00:40:35 SpeedEvil: perhaps juri's PMU is noticing that a USB charger has been plugged and trying to run uboot? Sep 26 00:40:38 right. ;) Sep 26 00:40:57 SpeedEvil: which would happily think it has lots of juice and proceed to enable all the blinkenlichten... Sep 26 00:41:09 its not a "steady flash". sometimes it misses a turn-on or turn-off cycle. pretty randomly. Sep 26 00:41:12 Wheee! Finally got an image built! So now I have an image that boots, starts gsmd and the dialer without error. :) Sep 26 00:41:31 The panel occaisionally dies, but that's another problem. Sep 26 00:41:32 lemme try it with the other battery... Sep 26 00:41:47 SpeedEvil: if that's the case, uboot's lowlevel init should be made a bit smarter... checking first with the charger if it's ok to proceed or if it should halt the cpu instead Sep 26 00:42:54 Looks like someone committed what is claimed to be a less intrusive startup sound -- but they commited it to the wrong place (2007.1 I think, not the 2007.2) Sep 26 00:43:11 is it possible to get a beefier battery for the phase 1 neo? Sep 26 00:43:58 hmm. just a dead battery. Sep 26 00:44:14 i'll charge my "good" battery, then try and charge this mostly dead one. Sep 26 00:44:31 Not quite "consumer-ready", huh :) Sep 26 00:44:51 orzo: a beefier battery wouldn't help with insane failure modes like these Sep 26 00:45:00 BTW, anyone know where the "hangup" button is? Sep 26 00:45:18 cesarb: well i want it to last longer without having to implement power saving tricks Sep 26 00:45:48 how's the suspend feature development progressing? Sep 26 00:45:48 wilymage: 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.) Sep 26 00:45:55 does anybody know why i get a "network registration denied" via gsmd shell? Sep 26 00:45:58 mwester I have your patch. That goes in the kernel directory and build, right? Sep 26 00:46:18 if i use one sim, its fine, with the other it does answer with that error... Sep 26 00:46:41 LittleIdiot: first things first: is the failing SIM for a GSM850 carrier? Sep 26 00:48:41 uhm... dunno, i'll have a look if o2 germany uses the 850 band... Sep 26 00:49:10 cdes: yes - basically that all needs to be done. It's broken at the moment. And it needs to turn off before using all the juice. Sep 26 00:49:20 but these two sims are from the same carrier, only difference is that the one not working has it's "homezone" where i am Sep 26 00:52:16 LittleIdiot: check /tmp/gsmd.log Sep 26 00:53:02 its in the 1800 band Sep 26 00:53:04 LittleIdiot: even if it doesn't register, you should be getting the signal strengths (the CSQ stuff), and you could find the error message there too Sep 26 00:55:48 ok, rebooting and inserting the other sim... might take some time ;) Sep 26 00:56:23 LittleIdiot: if you are booting with the failing SIM, be sure to pastebin the gsm log, since it can be useful for people to help debugging Sep 26 01:02:03 if i write some assembler code to put the cpu in slow mode, will it stay that way and iwll the phone be operational? Sep 26 01:04:12 juri_: I just had an idea, if you have a debug board you could try using it to halt uboot before it decides to turn on the backlight inverter, for long enought to do a basic charge Sep 26 01:05:35 sagacis: There are multiple ways to do it. Do you want to build it yourself, or just download a kernel with the patch already in it? Sep 26 01:06:24 sagacis: http://www.mwester.net/moko for a kernel that's pre-patched. (and a rootfs that's also patched, if anyone wishes). Sep 26 01:07:34 But do double-check the md5sums before you flash; my ISP doesn't understand that TCP is supposed to be a reliable end-to-end connection, and twiddles a couple bits every few GBytes of data. Sep 26 01:09:12 shouldnt it be able to turn on even with a dead battery as long as i have usb power supplied? Sep 26 01:09:45 it can't power up far enough to be able to charge faster than it discharges if the battery is dead Sep 26 01:10:23 mwester: the tcp checksum is what's supposed to make it reliable... Sep 26 01:10:45 mwester: (of course, with the current data rates, it might not be enough) Sep 26 01:12:05 Yeah, but it's only a checksum. My ISP has some bad equipment that twiddles bits in pairs, in a small set of patterns that go undetected. They insist it's noise - I point out several hundred examples I've collected with *exactly* the same bits, in the same positions relative to one-another -- and they harumph, and mutter, and say "We'll get back to you" and do nothing. Growl. Sep 26 01:13:18 mwester: ouch, such evil networking equipment can be a pain Sep 26 01:13:28 Mek: I just wrote up a bug report on it: 891 Sep 26 01:13:31 !ombug 891 Sep 26 01:13:33 Especially when it's tower-mounted. Sep 26 01:13:41 hm, where's the bot? Sep 26 01:13:59 ~botsnack Sep 26 01:14:00 aw, gee, mwester Sep 26 01:14:10 ~ombug 891 Sep 26 01:14:23 apt: help Sep 26 01:16:06 mwester: the ombug bot is the other one Sep 26 01:16:10 mwester: not apt Sep 26 01:16:28 We have multiple bots, then? Sep 26 01:16:57 * mwester wonders if it's possible to get the two bots to "talk" to one-another Sep 26 01:17:18 cesarb: CME ERROR: 15 Sep 26 01:17:26 according to etsi that is "SIM wrong" Sep 26 01:17:48 pin is definitively correct Sep 26 01:17:51 mwester: the bot which answers to ombug is too behaved for that... Sep 26 01:19:46 LittleIdiot: the PIM being sent which you can see on /tmp/gsmd.log is the correct one? the UI is still a bit dodgy, I wouldn't trust it too much... Sep 26 01:20:29 hm, wouldn't the wrong PIM be error _16_ instead of _15_? Sep 26 01:20:39 s/PIM/PIN/ Sep 26 01:20:39 cesarb meant: hm, wouldn't the wrong PIN be error _16_ instead of _15_? Sep 26 01:20:58 error 15 says _SIM_ wrong... as to what that means, I've no idea... Sep 26 01:21:02 cakeface_: here's the log http://pastebin.com/f67ce2c25 Sep 26 01:21:35 but it's right next to errors 13 (SIM failure) and 14 (SIM busy) Sep 26 01:23:08 SIM failure? Sounds like it might be a mechanical problem; the connector is known to be very particular. Sep 26 01:23:15 LittleIdiot: hm, right after that error, it sends again the AT+CFUN=1, which was what requested the PIN before, and it succeeds this time, so the SIM seems to have accepted the PIN Sep 26 01:24:06 LittleIdiot: the real error seems to be 32 (network not allowed - emergency calls only) Sep 26 01:25:26 LittleIdiot: and +CREG gets 2 (searching) and then 3 (registration denied) Sep 26 01:25:43 what was that command to show the current operator... AT+COPS? wasn't it? Sep 26 01:25:50 RSTR=`+CREG: 2,3' Sep 26 01:26:00 ...response to CREG Sep 26 01:26:14 cops returns 0 Sep 26 01:26:29 LittleIdiot: you could try libgsmd-tool -m atcmd and run AT+COPS? on it Sep 26 01:26:46 i did - returns: RSTR=`+COPS: 0' Sep 26 01:27:45 LittleIdiot: sorry, just checked the manual, that isn't the correct command, it'll always return 0 Sep 26 01:28:18 LittleIdiot: you could try "AT+COPS=?" (lists operators, can take several minutes, any other command of yours will cancel the search and confuse gsmd badly) Sep 26 01:29:05 mwester, sorry. Got called away to put a new sim card in wife's phone. I would prefer to build my own kernel (with ext2 support for booting from SD Card) Sep 26 01:29:07 isn't it "AT+COPS?"= ;) Sep 26 01:29:26 LittleIdiot: ? shows the current value, =? shows what it can be set to Sep 26 01:29:39 LittleIdiot: in the case of AT+COPS, =? shows all the available carriers Sep 26 01:29:56 LittleIdiot: and IIRC it also says which one you are currently using... Sep 26 01:30:05 ok, COPS=? returns CME error 15 again... Sep 26 01:30:12 LittleIdiot: but I don't know if it'll work when the net reg has been denied Sep 26 01:32:41 hm, a completely unrelated manual found on google says as a troubleshooting step for CME ERROR 15 "Check SIM, it must be a GSM SIM." Sep 26 01:32:56 the error itself is described as "SIM is wrong type" Sep 26 01:33:26 ok, well the sim also works with umts, maybe that's why Sep 26 01:34:14 i'll check the logs for the correct operator id and will try to issue an operator manually... Sep 26 01:34:25 LittleIdiot: tell FIC, they might want to know more about that SIM of yours... it's not the first problematic one (they have been having trouble with some US SIMs too, I heard rumors a modem firmware revision is in the works) Sep 26 01:35:00 LittleIdiot: they have been having trouble with some SIMs _and_ GSM850 (but they have been completely and utterly silent on the GSM850 issue) Sep 26 01:37:08 LittleIdiot: what's most interesting is that it's answering with error 15 when you send it the PIN, and then later goes ahead as if either the PIN succeeded or you simply didn't have a SIM installed Sep 26 01:37:24 i've seen somebody reporting a similar problem with an uk network, which was also not 850... he got no answer on the ml so far Sep 26 01:38:19 sagacis: http://pastebin.com/m1fac6ee3 Sep 26 01:38:42 LittleIdiot: perhaps the modem cannot notice the problem before you send the PIN (since the SIM won't talk without it), and then with the SIM unlocked it notices something wrong, and pretends the SIM is not even there so you can still do emergency calls Sep 26 01:39:34 sagacis: make those changes to your openembedded directory, then rebuild -- it'll create a new kernel.. Sep 26 01:42:01 mwester: Cool! I was almost there. I didn't know the monotone stuff. This is good info. Thanks. Sep 26 01:43:41 bedboi: I updated the images on the web site - sorry for the delay; there were some OE problems building today :( Sep 26 01:44:18 cesarb: ok, thank you so far ;) Sep 26 01:44:41 i did the no battery for 10 minutes thing Sep 26 01:44:49 and have had it plugged in for a while and i just turned it on Sep 26 01:44:58 it booted to the logo and is stuck there Sep 26 02:00:39 do the no battery thing, then leave it for longer Sep 26 02:02:16 mwester: You got the dialer without error? Did you use an older libmokoui2 or am I just too dense to see the fix on SVN? :) Sep 26 02:04:44 I can't commit the fix to SVN, I don't have SVN access :( Sep 26 02:05:44 There are two fixes, one is to revert to the older libmokoui2, and there's another that patches libgsmd, they fix the inability to start the dialer, and the dialer's "can't power up the antenna" problem. Sep 26 02:05:49 respectively. Sep 26 02:06:55 And Elrond found the panel problem (the one where the matchbox panel sometimes (usually) won't start) -- it's the battery applet, not the GSM applet. Sep 26 02:07:38 So I have an image now that seems to start very reliably. :) So much so that I've suddenly realized that I can't find the hangup button to end a call! Sep 26 02:08:11 mwester: Did you make a patch for libmokoui2 or are you referring to the gsmd patch? Sep 26 02:09:08 I have another patch for libgsmd, and you have to also edit sane-srcrevs to go back to an older libmokoui2 (and clean up the build remnants from the newer, broken one, before you rebuild). Sep 26 02:10:04 So now there's a kernel patch to fix the gsmd flow-control lockups, a patch to fix the gsmd startup problems, and another gsmd patch to fix the dialer problem. It's becoming quite "patchwork" :-D Sep 26 02:10:05 mwester: Ok, then I'll make a patch for libmokoui2 and open a bug Sep 26 02:10:22 Yeah, heh :) Sep 26 02:11:01 I think there were some updates to libmokoui2 committed to SVN, I'll try one of those tomorrow and see if they work. Forward is better than backwards... Sep 26 02:11:24 They're just documentation Sep 26 02:12:34 I had a manager once who thought that sufficient documentation of a problem would result in a resolution of same... Sep 26 02:13:00 Ohh, fun Sep 26 02:17:23 We had a software engineer who fixed a bug by updating comments in the code this year. Sep 26 02:17:39 When the bug came up for analysis, we couldn't get it to reproduce. Sep 26 02:18:19 We finally found one of the engineers had inadvertently flipped a bool comparison while writing comments on the line above, and he fixed the bug! Sep 26 02:22:36 mwester: what? Sep 26 02:23:35 sagacis: wow Sep 26 02:24:10 mwester: that's because a documented issue is not a bug, it's at most a misfeature ;-) Sep 26 02:24:17 How... do you flip a bool comparison by writing a comment above? Sep 26 02:24:33 abraxa_: fat fingers? Sep 26 02:25:02 *shudders* Sep 26 02:26:43 In his defense, it was mostly intentional. He was writing the comment for the bool comparison and realized it wasn't correct. He made it match the comment and forgot later what he had done. Sep 26 02:27:09 When he committed, his comment said "added comments to lots of files", but that also contained the fix. Sep 26 02:27:20 he didn't know there was a filed bug for the problem. Sep 26 02:41:44 ohhh okay - glad to know he did it on purpose Sep 26 02:43:21 On that note... why do people use "dnl" in autotools instead of "#"? Sep 26 02:44:57 From the gentoo dev manual: (dnl is actually a macro which discards the remainder of the input — it stands for "discard new line"). Sep 26 02:45:33 Hm... thanks, though I fail to see how that would be preferred to # Sep 26 02:45:58 I think it depends on what file you are in, but I'm no autotools guru Sep 26 02:46:44 how slow is the mmc writing? Sep 26 02:49:27 ~3mb/sec Sep 26 02:50:08 I'm trying to cp the /opt/Qtopia to the sd card. It's been at it for 30 mins or so. Sep 26 02:50:37 What FS is on the card? Sep 26 02:50:48 ext2 or ext3, I can't remember offhand Sep 26 02:50:55 Ah, hm Sep 26 02:50:56 I think it's ext2 now. Sep 26 02:51:08 Did you mount it sync or not? Sep 26 02:51:59 sync Sep 26 02:52:43 Mounting it without sync gives you a significant speed improvement, so you might wanna edit you fstab accordingly and try again Sep 26 02:52:48 your* Sep 26 02:53:00 abraxa_: there's a fundamental difference between dnl and # Sep 26 02:53:13 cesarb: Oh? Sep 26 02:53:29 abraxa_: dnl is a true comment: not only nothing is output, but the newline is also discarded (dnl = discart till new like or something like that) Sep 26 02:53:37 darn. sorry I wasted 30 minutes Sep 26 02:53:51 abraxa_: for #, it (IIRC) disables macro expansion, BUT does not discard the line... it gets sent to the output Sep 26 02:54:44 sagacis: fear not, I think I've seen on the commits list for OE (somewhere at lists.linuxtogo) someone changing the default fstab to async ;-) Sep 26 02:54:47 sagacis: today ;-) Sep 26 02:54:58 cesarb: Ahh... okay, that makes... complete sense, considering autotools' partial retardedness :) Sep 26 02:55:02 heh.... Just in time. Sep 26 02:55:10 I'll have to update my phone. Sep 26 02:55:31 abraxa_: think of autoconf as a macro preprocessor for shell scripts on steroids, and you'll have the right idea ;-) Sep 26 02:55:50 Yeeeeahhhh... m4, anyone? Sep 26 02:56:05 abraxa_: the "retardness" is due to excessive preocupation with portability (wtf is a "Kubota Titan"?) Sep 26 02:56:42 abraxa_: for instance, it duplicates code all over (leading to a multi-megabyte shell script). Why? Because some VERY OLD shells don't have functions ;-) Sep 26 02:57:41 ...yeah, and it bugs me that it copies the "missing" scripts to each and every project instead of just making symlinks or magically using system-wide installed ones... Sep 26 02:58:13 in today's "all the world uses Linux" environment, such preocupations are excessive... making it work with the last 5-6 years of Linux distributions, assorted BSDs, MacOS X releases and Unix-like windows environments would cover over 99% of the user base Sep 26 02:58:27 Anyway, I just didn't understand why people were using "dnl" instead of "#" when "dnl" significantly reduces readability in my opionion. Sep 26 02:58:58 abraxa_: autoconf copies the scripts because it wants to generate a self-contained tarball, which doesn't depend on the user having autotools Sep 26 02:59:35 abraxa_: if you use the correct macro, it can put all but 2-3 of them in a separate directory (the documentation even warns you that you should NOT name that directory "aux") **** ENDING LOGGING AT Wed Sep 26 02:59:56 2007