**** BEGIN LOGGING AT Wed Sep 12 02:59:56 2007 Sep 12 04:39:31 moin Sep 12 05:01:56 hey Sep 12 05:02:14 I get an error while compiling my openmoko-devel-image: | configure: error: cannot find sources (gdbmdefs.h) in /media/nfs/moko/build/tmp/work/i686-linux/gdbm-native-1.8.3-r2/gdbm-1.8.3 or .. Sep 12 05:02:22 can anyone help me? Sep 12 05:03:07 uuhm.. youre building via nfs? Sep 12 05:06:53 roh: well, the files are remote, but that shouldn't be the problem Sep 12 05:08:06 edistar that sometimes provokes strange things... did you try rebuilding gdbm-native? Sep 12 05:08:21 make rebuild-package-gdbm-native Sep 12 05:08:30 if youre on mokomakefile Sep 12 05:09:35 roh: yes, I am Sep 12 05:09:38 I'll try Sep 12 05:11:24 roh: same error: configure: error: cannot find sources (gdbmdefs.h) in /media/nfs/moko/build/tmp/work/i686-linux/gdbm-native-1.8.3-r2/gdbm-1.8.3 Sep 12 05:19:39 roh: a bit strange, it tells me at the beginning that it's building for i486 altough local.conf has i686 configured, and then it looks for */i686-linux/gdbm*/ Sep 12 05:23:03 edistar: I know people have had problems with that before Sep 12 05:23:14 But not sure if it was gdbm. Sep 12 05:23:23 Try building for i486 maybe? Sep 12 05:24:32 CM: just change i686 to i486? Sep 12 05:25:07 If I remember correctly, someone here said they got around their problem that way Sep 12 05:25:29 I'm a bitbake newbie myself and haven't really set myself into it yet Sep 12 05:27:28 HI, the kernel driver of USB serial converter chip in debug board v2 is what? It seems that my kernel can't recognise it Sep 12 05:28:29 I should remember what I had to change, but I did it 2 month ago.. Sep 12 05:28:34 Is it in the wiki? Sep 12 05:29:37 CM: if you remember, please tell me ;) Sep 12 05:30:48 Device Drivers -> USB support -> USB Modem (CDC ACM) support Sep 12 05:31:17 The wiki says the usb to serial chip is part of FT2232, but where is the driver of it? Sep 12 05:31:38 agui: The one I just mentioned :) Sep 12 05:32:29 It's an usb modem, and you connect to /dev/ttyACM0 Sep 12 05:35:25 Do you mean use drivers/usb/class/cdc-acm.c? Sep 12 05:38:07 Oh, sorry.. Guess I'm not really awake :( Sep 12 05:38:19 agui: Didn't see that you said the debug board Sep 12 05:39:27 agui: http://wiki.openmoko.org/wiki/Neo1973_Debug_Board_v2 Sep 12 05:41:41 I have read that WiKi, everything is ok except that there is no /dev/ttyUSB0 or /dev/ttyUSB1 device file or something else appear Sep 12 05:42:04 agui: Ok, then it's the usb modem drivers I guess Sep 12 05:42:47 Try adding them, and if that's not enough, wait around a bit until the other people who have debug boards and know more wake up ;) Sep 12 05:44:08 Hi, I got the kernel message "drivers/usb/class/cdc-acm.c: This device cannot do calls on its own. It is no modem" Sep 12 05:44:52 I just have a Neo, and it's a modem, but I don't know about the debug board Sep 12 05:45:11 agui thats normal. it just means that is just a serial and nothing with ri and cd lines Sep 12 05:45:53 CM: thank you a lot Sep 12 05:46:12 agui for the debugboard use this: Sep 12 05:46:15 modprobe ftdi_sio vendor=0x1457 product=0x5118 Sep 12 05:46:27 agui: np, but roh is someone who really knows :) Sep 12 05:46:59 then it detects the debugboard as 2 serials and as soon as openocd starts the first one vanishes. the second one is where the real serial console of the moko comes out Sep 12 05:47:06 Ok, I will try, thanks! :) Sep 12 05:47:30 agui: That part was in the wiki too ;) Sep 12 06:17:41 hi Sep 12 06:52:55 ...still building! Task 1993... hope this will last till the end of that build Sep 12 07:24:37 boo Sep 12 07:42:50 hi raster... Sep 12 07:43:48 emdete, yoyo Sep 12 07:46:00 * bedboi is away: sleep mode Sep 12 07:48:57 raster: we are playing around with edje, performance is indeed better than expected... Sep 12 07:49:04 raster: but development is a pain ;) Sep 12 07:49:13 aarg Sep 12 07:49:24 have to run. Sep 12 07:49:34 would love to chat. Sep 12 07:49:45 meeting foo Sep 12 07:49:54 :D Sep 12 08:06:45 I tried to build a *ext2 image of OpenMoko, but I get this error: genext2fs: couldn't allocate a block (no free space) Sep 12 08:07:11 where I can change the size? Sep 12 08:12:56 heh... harald used SOCK_SEQPACKET in gsmd/libgsmd.. which means that only works with > 2.6.4 Sep 12 08:25:20 re Sep 12 08:40:14 oh, now there are free tools for unlocking the iphone. Sep 12 08:40:53 is someone working on a openmokoport already? :) Sep 12 08:47:01 :-) Sep 12 08:47:52 1,5 months. not bad, although I actually thought it would be faster Sep 12 08:49:07 nslu2-linux is porting Optware to the iphone Sep 12 08:49:29 http://svn.nslu2-linux.org/svnroot/optware/trunk/platforms/toolchain-iphone.mk Sep 12 08:50:40 free iphones for 399, not a bad deal at all Sep 12 08:50:43 hhf423: 1,5 months until phase2 sells? Sep 12 08:51:20 dcordes_: you wish. 1,5 months from iphone sales to unlock Sep 12 08:51:38 399? where did you get that price from? Sep 12 08:51:42 hmmmmmmm Sep 12 08:51:55 I need one badly Sep 12 08:52:02 you have to add 2 years contract... Sep 12 08:52:11 emdete: 399 is the price of an iphone and unlocking is now free Sep 12 08:52:33 emdete: you do not have to add a 2 years contract. you never had to do that. Sep 12 08:52:51 emdete: you were supposed to, but thats not the same Sep 12 08:52:52 hhf423: so the contract starts with the activation oonly? Sep 12 08:53:29 emdete: yes. you buy the phone without contract, then usually choose a contract upon activation. Sep 12 08:53:48 hey guys Sep 12 08:53:51 hhf423: wow, that's different Sep 12 08:54:03 now the neo seems to be very, very expensive Sep 12 08:54:07 i have a question Sep 12 08:54:25 emdete: where did you get the impression that it was different? Sep 12 08:54:41 hhf423: i thought the contract starts with the purchase Sep 12 08:54:46 i was building that new devel-enviorment Sep 12 08:54:58 and got this finally: Sep 12 08:55:02 | checking for MOKOUI... configure: error: Package requirements (libmokoui2 vte) were not met: Sep 12 08:55:02 | Sep 12 08:55:02 | No package 'vte' found Sep 12 08:55:13 emdete: well, after the massive (33%) price cut from apple, the Neo is actually expensive, feature wise. If you do not count the value of GPL drivers etc. Sep 12 08:55:27 hhf423: who does.... Sep 12 08:55:55 RalphEichelberge: I'm updating now, I'll see if I get that error too Sep 12 08:55:59 emdete: no, that was never the case. Thats the big different from iphone to other phones brought in the states, in the shop, you just buy the phone. Sep 12 08:56:19 RalphEichelberge, try to build vte first Sep 12 08:56:40 well, how? Sep 12 08:57:00 bitbake vte Sep 12 08:57:12 hhf423: actually, I do. I think there is value in open drivers and such. But of course the question is how many people care and how much (as in $$$ much) they care :-) Sep 12 08:57:24 openmoko@openmoko-celsius:~/moko$ bitbake vte Sep 12 08:57:24 ERROR: Please set the 'CACHE' variable. Sep 12 08:57:55 RalphEichelberge, i don't know how to do that with mokomakefile Sep 12 08:57:59 hhf423: shure, me too. but fic looks for mass market Sep 12 08:58:46 RalphEichelberge: you need to source ./setup-env first Sep 12 08:59:30 well i dont know enough about the building process Sep 12 08:59:38 the MokoMakefile wiki page shows how to run bitbake manually - have you read it? Sep 12 09:00:10 emdete: right. that will only work with a cool device (hardware) and cool software. And because they do not have the cachet/prestige of Apple, even with the same featureset and slickness of GUI they will have to charge less Sep 12 09:00:16 well i read the wiki - but u know it has many sites :) Sep 12 09:00:28 i will search for that Sep 12 09:01:20 ok rwhitby ist baking Sep 12 09:01:31 http://wiki.openmoko.org/wiki/MokoMakefile under "Developing with MokoMakefile" - I guess part of that could be moved to an earlier section. Sep 12 09:01:41 what is vte anyway? Sep 12 09:01:52 Please add it where you would have expected to see it on that page. Sep 12 09:02:28 well that page i found serveral times in wiki Sep 12 09:02:48 and i have read it too Sep 12 09:03:20 but you know - there are new aspects to the same manual every time to have a new issue :) Sep 12 09:05:11 hhf423: shure, cool device is important. but when i show the neo around most people don't like it very much. okay, the included gps is great for some but the display and the high edge around is really ugly Sep 12 09:11:06 RalphEichelberge: vte = virtual terminal emulation Sep 12 09:13:23 libmokogsmd2 seems broken now.. Sep 12 09:13:58 http://rafb.net/p/OoeKHb43.html Sep 12 09:17:31 i guys Sep 12 09:18:15 ello :) Sep 12 09:18:26 i Marcin :D Sep 12 09:19:41 someone remember which uboot rev is safe to use? Sep 12 09:20:20 2007.8 snapshot Sep 12 09:22:18 ok Sep 12 09:24:04 hi all Sep 12 09:25:32 I have a question, I have upgraded packages from buildhost.openmoko.org and since then stopping gsmdaemon hangs whole phone, does anybody know this bug? Sep 12 09:25:50 zdanek, yup, same here Sep 12 09:26:14 see http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=788 maybe? Sep 12 09:26:35 zdanek, can't even do ipkg upgrade since it restarts deamon when upgrading =P Sep 12 09:26:46 Sep 12 11:25:45 home kernel: [55434.961200] usb 2-2: usbfs: USBDEVFS_CONTROL failed cmd dfu-util rqt 161 rq 3 len 6 ret -62 Sep 12 09:26:49 argh Sep 12 09:29:41 yes, that's the bug Sep 12 09:32:23 by the way I'll ask my old question did someone manage to make gprs connection on 2007.2? Sep 12 09:32:37 zdanek: there is doc in wiki Sep 12 09:32:43 especially hrw because you have polish sim Sep 12 09:33:03 I have some problems as I wrote on community mail list Sep 12 09:33:07 (PL) Sep 12 09:33:17 emdete: yes, thats right Sep 12 09:33:23 * hrw do not read community too often Sep 12 09:33:28 :) Sep 12 09:33:33 zdanek: see "Phase 1 software testing" in the wiki Sep 12 09:34:03 I have Sep 12 09:34:03 [LCP ConfReq id=0x1 ] Sep 12 09:34:03 and then Sep 12 09:34:03 tcsetattr: Invalid argument (line 964) Sep 12 09:34:05 zdanek: did you do the change to boot variables? Sep 12 09:34:22 sorry, what do you mean? Sep 12 09:34:34 zdanek: I also get that gprs error sometimes, I usually have to try two to five times Sep 12 09:35:16 I tried many times, never successful Sep 12 09:35:25 zdanek: did you remove "console=ttySAC0,115200" from bootargs_base while in uboot? Sep 12 09:35:33 weird. I found a totally broken island but the rendering was fine. I assume someone has broken is recently Sep 12 09:36:45 hhf423: nope Sep 12 09:38:29 * hrw wants Neo1973 holder from carkit Sep 12 09:41:18 hhf423: I briefly red wiki's Phase 1 soft testing and i'll try to fix some probles at my neo Sep 12 09:41:27 but suddenly I found this http://scap.linuxtogo.org/files/0c933096b179fc81f91595ccf4d3e681.png Sep 12 09:41:33 what's this app? Sep 12 09:41:39 any heard about gconf support in python? in debian it comes from python-gnome2 and gnome-python Sep 12 09:42:15 zdanek: thats "SettingsGui", look it up in the wiki Sep 12 09:43:34 hhf423: thank you Sep 12 09:45:25 zdanek: from the same author there is also "smstool" which will allow you to send sms via gui Sep 12 09:46:18 thx :D Sep 12 09:54:40 mickey|sports: ping Sep 12 09:54:41 hi Sep 12 09:54:52 hi Sep 12 10:11:45 ok, just updated neo...now X hangs instead of libgsmd =P Sep 12 10:15:20 jeddy3: did you remove "console=ttySAC0,115200" from bootargs_base while in uboot? Sep 12 10:15:56 hhf423, aah no :| Sep 12 10:20:06 jeddy3: you also need to find a black chicken and kill it under full moon Sep 12 10:20:16 hhf423, xD Sep 12 10:22:31 hhf423, you have any reference to that uboot fix? Sep 12 10:23:32 jeddy3: err, its in one of the bugs in bugzilla, let me check Sep 12 10:25:06 788 over and over Sep 12 10:26:56 !ombug 788 Sep 12 10:27:13 wow, Gravenreuth muss in den knast? Sep 12 10:27:17 oops, sorry Sep 12 10:28:25 oh! it's same bug as before...didn't behave the same though :| Sep 12 10:28:42 thanks for help, hhf423, tuukkah Sep 12 10:29:52 jeddy3: welcome Sep 12 10:37:27 mickeyl: should mokomakefile build the frozen version uboot-openmoko, or from uboot-openmoko_svn.bb, or should we create a new frozen version (cause the existing frozen version doesn't patch for me) Sep 12 10:40:01 mickeyl: also, when are you guys going to institute a fixed SRCREV in each _svn file, which gets bumped by the author when a new svn revision has been tested? Sep 12 10:40:16 (or alternatively, in a preferred-openmoko-versions file) Sep 12 10:42:30 rwhitby: the frozen version should be preferred. it builds fine here, btw. Sep 12 10:42:43 Hmm. Sep 12 10:42:58 fixed SRCREVs would be nice. hopefully next month, when the OE dude starts Sep 12 10:43:28 do you agree with the bumping of a srcrev equating to some rudimentary testing of the new value? Sep 12 10:44:21 ya Sep 12 10:45:29 mickeyl: what's the latest thoughts in the core team about gsmd and removing kernel console= vs stty vs something else? (ombug 788) Sep 12 10:45:57 which should we be recommending to each new person who arrives here with the hung neo problem? Sep 12 10:46:15 rwhitby: my personal take on that is removing kernel console in environment Sep 12 10:46:19 i proposed this ages ago Sep 12 10:46:40 since i always had the gut feeling that the automatic switching was not perfectly working Sep 12 10:47:06 the relatively new complete hang finally proves my gut feeling :/ Sep 12 10:47:22 can we enable the fw_env tool so that people can modify the u-boot environment from userland? Sep 12 10:47:42 fw_env? Sep 12 10:48:01 yeah, userland tool that does printenv and setenv of the uboot environment variables Sep 12 10:48:31 interesting. why not just having someone download the fixed environment via dfu ? Sep 12 10:48:59 someone with a debug board should test whether dfu now correctly uploads Sep 12 10:49:18 then we could just take this uploaded env and put it online Sep 12 10:49:22 (after testing) Sep 12 10:49:36 What about the uboot_env? (mtdblock1) Sep 12 10:50:04 I've uploaded and then downloaded the uboot-env partition using dfu-util when I unbricked the neo from Sydney Sep 12 10:50:49 Guess I'll have to try that next Sep 12 10:51:19 Before, scp:ing a working env to the neo and dd it to /dev/mtdblock1 worked, but no longer Sep 12 10:51:23 rwhitby: could you include a check for the uboot image in qemu-flash-local? without that file flash.sh will fail horribly. Alternatively, just include build-package-uboot-openmoko as a prerequisite for qemu-flash-local. kthx. Sep 12 10:52:15 ixs: uboot is now built by image, so that problem should resolve itself. Sep 12 10:52:31 rwhitby: ahhh okay. in that case, nvm. .D Sep 12 10:54:34 rwhitby: pushing the frozen version to a new date Sep 12 10:54:55 thx Sep 12 10:57:39 hello! is "amd64 double core + fc6 + binutils-cross-2.17 + mokomakefile won't compile" a faq already? Sep 12 10:58:03 hmmmm Sep 12 10:59:35 rwhitby: can you put this known-working environment on the wiki please? Sep 12 10:59:35 Batalyx: http://bugs.openembedded.org/show_bug.cgi?id=2928 Sep 12 10:59:50 who ever did spread the rumour that recent uboot builds will not auto-poweroff after 30secs? they even do that in qemu? Sep 12 11:00:25 mickeyl: you mean the one that I flashed on the sydney brick? sure. which page? Sep 12 11:00:36 Batalyx, there can be many problems with builds Sep 12 11:00:56 rwhitby: does it still contains the serial console? Sep 12 11:01:05 CM: thanks. I'll try that... Sep 12 11:01:09 what's the easiest way for me to create and build my own version of gmsd or neod? Sep 12 11:01:19 mickeyl: it won't after I upload it again from my neo, then download it again to test it Sep 12 11:02:11 do i create a local overlay then what? Sep 12 11:02:16 rwhitby: well, perhaps it would be even better on the web. i guess i'll put it to the 2007.08 snapshot directory Sep 12 11:02:23 or hey Sep 12 11:02:26 right into OE :) Sep 12 11:02:29 hey zecke Sep 12 11:02:42 ixs: obviously someone with no ida Sep 12 11:02:58 zecke: right on your way to Holland? Sep 12 11:03:08 hhf423: yeah, someone even put that info into the wiki on SH1 Sep 12 11:03:15 mickeyl: you could put it in the openmoko svn, then I can add a mokomakefile target to dfu-util flash it Sep 12 11:03:47 ixs: setenv boot_menu_timeout 65000' to keep it from powering down Sep 12 11:03:50 mickeyl: yes! the train just departed and I found a power plug... Sep 12 11:04:23 * hhf423 congratulates zecke to have gotten a tain with poweroutlets Sep 12 11:04:30 train even Sep 12 11:04:45 hhf423: I know that command, but as I said, rumour is, that "broken uboot installs on some neos" power down Sep 12 11:04:57 the places for the disabled most of the time have poweroutlets... Sep 12 11:05:07 ixs: funny Sep 12 11:06:59 rwhitby: sounds good Sep 12 11:07:00 hhf423: "* Sep 12 11:07:01 Note: Some devices have been reported to turn off after a short period of inactivity while on the uboot menu. To avoid this, simply press the AUX button every 30 seconds or so. " Sep 12 11:07:10 from http://wiki.openmoko.org/wiki/Flashing_openmoko Sep 12 11:07:16 well Sep 12 11:07:24 it is like that Sep 12 11:07:35 turning off while inactive is standard Sep 12 11:07:50 turning off while a dfu-load is in progress would be a major bug Sep 12 11:07:58 i didn't see the latter one since months though Sep 12 11:07:58 yes, but "some devices"?! all devices! Sep 12 11:08:11 ya, some devices is bogus Sep 12 11:08:27 changing it Sep 12 11:08:34 thanks Sep 12 11:08:41 ixs: will you change that to all and put in something like unless you setenv boot_menu_timeout 65000' to keep it from powering down Sep 12 11:08:58 (saveenv afterwards) Sep 12 11:12:32 done http://wiki.openmoko.org/wiki/Flashing_openmoko#Preparing_the_device_for_DFU_mode Sep 12 11:17:26 someone know how to disable GSM broadcast messages? Sep 12 11:18:19 You mean cell broadcasts2D? Sep 12 11:24:14 yes Sep 12 11:24:39 Heyah sends lot of them - mostly sex offers Sep 12 11:24:42 there is a way of saying 'I only want cell broadcasts that are x,y,z Sep 12 11:24:46 ID Sep 12 11:28:07 Batalyx, zipola, if you come to the flug-jkl meeting today we can hack on the neo Sep 12 11:29:44 I'm afraid I'll have to recover from flu yet another day. Sep 12 11:30:11 zipola: Äh, make it a flu-jkl meeting :P Sep 12 11:30:44 * CM hides in shame in the corner after delivering that awful joke Sep 12 11:32:49 tuukkah: sorry, can't come there today Sep 12 11:32:53 hehe Sep 12 12:15:28 anyone know where i can get gst-meta-audio which openmoko-mediaplayer2 needs? Sep 12 12:19:53 ok the source is gst-meta-base Sep 12 12:53:29 morning Sep 12 13:07:40 hi kristian-m Sep 12 13:08:07 kristian-m: looks like gsmd and libgsmd became even more erratic after the last chenges Sep 12 13:08:24 good morning Sep 12 13:08:28 and yikes Sep 12 13:13:56 does AT+CSQ works in neo? Sep 12 13:15:49 hhf423 hmm, it'll develop - I'm implementing a passive access for settingsgui using the atcmd mode. - so stop interfering with gsmd in the next version... Sep 12 13:16:15 kristian-m: ah, thats probably a good idea Sep 12 13:17:54 hhf423 currently it initiates a power-on and register to the modem on connection Sep 12 13:17:57 hrw: yes - AT+CSQ will give you present signal quality - propritary command AT%CSQ IIRC enables unsolicited SQ reports Sep 12 13:20:39 thx Sep 12 13:34:59 * Sup3rkiddo has his neo delivered...dances up and down Sep 12 13:35:18 when order? Sep 12 13:36:13 SpeedEvil, that was long long long ago...3 days from the day it was announced Sep 12 13:37:25 Sup3rkiddo: Guess you got one from the new batch or? Sep 12 13:46:01 is the new batch different? Sep 12 13:47:50 CM, dunno mate, the insititute authorities are having it, i just got to touch it and had to give it to them...another 3 days to flash it Sep 12 13:50:08 ok Sep 12 13:58:08 hey all anyone getting screwed whilst building curl? Sep 12 14:00:12 http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=813 Sep 12 14:01:11 Sup3rkiddo: checking.. Sep 12 14:01:18 NOTE: package curl-7.16.4-r0: task do_build: completed Sep 12 14:02:11 hrw, dude, could you please check out the bug report and guide me on whatsup?, cleaning stuff doesnt help, Sep 12 14:08:06 no idea Sep 12 14:08:09 works for me Sep 12 14:11:10 Sup3rkiddo: have you tried `make clean-package-curl`? Sep 12 14:11:30 gordonsyme, yup Sep 12 14:11:41 and is it still curl that breaks? Sep 12 14:11:47 gordonsyme, yeah Sep 12 14:12:23 odd, I had a situation a while back where loads of packages exhibited that behaviour. Cleaning them and rebuilding worked for them all Sep 12 14:12:53 I've also just rebuild the curl package... Sorry, can't be further help, it works for me Sep 12 14:14:00 hi...has anyone tried to build an OpenMoko 2007.2 *.ext2 image ? Sep 12 14:14:06 gordonsyme, ok man, thanks..make clobber has worked for me in the past in similar scenarios, but grrrr....the amount of time..:( Sep 12 14:14:10 I tried I but there is a error Sep 12 14:14:41 Sup3rkiddo: I feel your pain, done that a good few times Sep 12 14:15:02 zefanja, and has it to do with curl? Sep 12 14:15:07 gordonsyme, =( Sep 12 14:18:00 sorry...I don't understand Sep 12 14:18:20 zefanja, er..sorry, paste the errors somewhere... Sep 12 14:18:47 wait Sep 12 14:19:41 genext2fs: couldn't allocate a block (no free space) Sep 12 14:20:02 zefanja: IMAGE_ROOTFS_SIZE is too small Sep 12 14:20:05 i think i have to change to size of the ext2 image file Sep 12 14:20:46 so i have to added this option in local.conf? Sep 12 14:21:01 yes Sep 12 14:21:05 thx Sep 12 14:21:54 the size in blocks? Sep 12 14:22:18 so 102400 for 100MB? Sep 12 14:22:52 well, the bill is in for repairs to the car :/ Sep 12 14:23:21 ScaredyCat: how much? Sep 12 14:23:32 2500 euro :( Sep 12 14:24:14 ScaredyCat: what did you do to it? Sep 12 14:24:17 openmoko: 03mickey * r2956 10/trunk/src/target/OM-2007.2/daemons/neod/src/buttonactions.c: Sep 12 14:24:17 openmoko: neod: get the name of the (first) backlight device node from sysfs Sep 12 14:24:17 openmoko: This makes neod controlling brightness on all devices with LCD backlight class Sep 12 14:24:25 ScaredyCat: Yikes, what car is it? Sep 12 14:24:26 ScaredyCat: That's a lot of work! Sep 12 14:24:35 * CM bought his car for that Sep 12 14:24:36 it wasn't me wibbit - someone broke into it. Sep 12 14:24:46 it's a Volvo S80 CM Sep 12 14:24:59 mwester: they made a real mess of it Sep 12 14:25:00 ScaredyCat: I had some one kick in 3 of my doors, repair bill was 2500 gbp Sep 12 14:25:14 ScaredyCat: Ah.. I'm sitting att the factory where they build them right now Sep 12 14:25:56 I love it.. it's the first car I've ever brought that wasn't a pile of crap. Sep 12 14:26:18 bedore this they were all sub 1000 gbp. Sep 12 14:26:25 before Sep 12 14:26:28 ScaredyCat: not to be a pedant, but it's bought... Sep 12 14:26:46 well, I bought it, then brought it home Sep 12 14:26:50 LOL Sep 12 14:27:27 kristian-m: there was also, I think Writchie thinking loud about connecting gsm to d-bus so you could then listen to that Sep 12 14:27:30 it's the first time I think that the labour costs have been lower than the parts cost Sep 12 14:27:46 CM: you work for Volvo? Sep 12 14:28:22 ScaredyCat: No, consultant company. But I've been assigned to work on Volvo's factory planning Sep 12 14:28:40 I'm responsible for the systems where they plan all future cars Sep 12 14:28:59 All cars at all that is Sep 12 14:29:15 cool Sep 12 14:29:25 * CM knows Volvo's cars in detail to 2016, and the vague volumes up to 2022 Sep 12 14:29:43 CM: I once listened to a gal from volvo that did the only-women-designed concept car Sep 12 14:30:09 hhf423: Ah, they got a lot of free pr for that one Sep 12 14:30:11 that was a cool car, it was dead practical Sep 12 14:30:24 you think you could slip a little bug in the design software that tricks the manufacturing equipment inot makeing me one and shipping it to me.. ? :P Sep 12 14:30:59 ScaredyCat: Hehe.. I can actually update the MPS (master production schedule) as I want.. Sep 12 14:31:19 * amarsh04 works where Harley-Davidson motorcycle wheels are made Sep 12 14:31:25 cool...that's even better... you can put it at the top of the list too then :P Sep 12 14:31:51 So the car would get planned an materials ordered and so on, but I'd need to assign an order number to it before they start building Sep 12 14:31:52 they make Chilli sauce next door... Sep 12 14:32:04 Takes about 2 days to build an XC90 Sep 12 14:32:14 I can wait 2 days :) Sep 12 14:32:18 Hehe Sep 12 14:32:36 surely you have to do 'test' runs ;) Sep 12 14:32:48 depending on the wind direction at work, we either get the smell of a wood veneer factory or a biscuit factory Sep 12 14:33:00 amarsh04: *lol* Sep 12 14:33:37 * CM sneaks out from work early Sep 12 14:33:42 :o Sep 12 14:33:51 with a volvo in his pocket.. Sep 12 14:34:05 Of course.. :P Sep 12 14:34:25 'is that a volvo in your pocket or are you just pleased to see me' Sep 12 14:35:39 my load car is a Peugeot 206... Sep 12 14:36:22 they guy on the phone kept saying "it's really small, small, I mean it's tiny, small, really - sorry" Sep 12 14:37:11 wibbit: they kicked in 3 of your doors? Sep 12 14:37:25 Yup Sep 12 14:37:28 road rage Sep 12 14:37:32 :O Sep 12 14:37:43 big boot prints in the side of three dorrs Sep 12 14:37:45 doors Sep 12 14:37:47 why did they miss the last one then? Sep 12 14:38:09 or had you run em over by then... Sep 12 14:38:27 ScaredyCat: no idea to be honest with you Sep 12 14:38:38 a bit ott though... Sep 12 14:38:55 maybe by the time they got to number 4 they'd calmed down Sep 12 14:39:01 The most anoying this was I phone the police there and then, and the stupid woman on the other end of the phone told me it was not a 999 issue, and that I should report it to the local police station Sep 12 14:39:02 ScaredyCat: me and gf have a 207. Her choice :) Sep 12 14:39:17 The fact that we were stuck in traffic, and they were in front of me for another phew minutes, and could have been cough Sep 12 14:39:25 amarsh04: Would you prefer having a stonemason next to your office who likes to work outside? :) For us the wind direction is irelevant. Even the wheather does not matter. :)) Sep 12 14:39:28 wibbit: atleast in Sweden,youre allowed to call 112 for "crime in progress" or accidents Sep 12 14:39:42 sandos: interesting Sep 12 14:39:45 wibbit: lazy fuzz Sep 12 14:39:48 I would assume kicking on somebodies car is a crime Sep 12 14:39:51 sandos: I think that is what 999 is for in the UK Sep 12 14:41:10 DPThought, I'm (mostly) happy with the environment at work (-: Sep 12 14:42:11 sandos: I just feel very very vunerable in something that small... Sep 12 14:42:18 we have 000 for "life threatening" emergencies and 131444 for "police attendance" Sep 12 14:42:37 ScaredyCat: heh, ok. The 207 has 5 star NCAP rating so.. Sep 12 14:42:57 there's at least 10 acres between me and the front in the S80... in the 207 it seems like it's right there Sep 12 14:42:59 I think its prety ok. Although I guess I wouldnt want to be in a head-on collision with for example a XC90 :) Sep 12 14:43:25 I don;t think anyone wants to get hit by a 4x4 ... can't be good Sep 12 14:43:54 It isn't Sep 12 14:44:03 * amarsh04 used in-vehicle gps navigation for the first time yesterday, and wonders what cpu power and storage is needed for such an application ( > current FIC Neo hardware?) Sep 12 14:44:37 amarsh04: look at the spaces of a tomtom go .. or clonish thing... Sep 12 14:44:46 gta02 spec is pretty close Sep 12 14:45:16 the unit I used was called a "navman" Sep 12 14:45:19 mickey|bbiab: Any possibilities we can discuss modifying the gsmd script to eliminate the gsm/device lockups (bug 788), in addition to changing the uboot env? That would resolve the problem for those who would rather not give up the console. (details in the bug report)... Sep 12 14:45:44 I did see a tv program on some car crashes... there was one where a truck driver smashed into the back of a car... Sep 12 14:46:20 amarsh04: The Garmin routing/mapping GPSen typically use a 100-130MHz ARM cpu.. current use is actually rather minimal-- my ex's 60CS does about 9 hours on 2x AAs. Sep 12 14:46:23 it wasn't untill the fire crew got there that they discovered that there had been another car *in between* the two Sep 12 14:46:29 ScaredyCat: ouch Sep 12 14:46:52 yes, that person was very very - not alive. Sep 12 14:47:25 * baird should be getting a new eTrex Legend tomorrow in the mail... Sep 12 14:48:07 I generally take the view that every other driver on the road is actually a small 5 year old child - prone to totally irrational behaviour at any point in time Sep 12 14:48:43 amarsh04: Navicore for the Nokia 770 comes with a 1GB mmc card. The only additional hardware needed is a bluetooth gps receiver. Cpu is slow (220Mhz arm, 64mb ram). It works ok. TomTom has a additional acelorameter. Sep 12 14:48:44 ... it seems to work (he says touching a very lrge peice of wood) Sep 12 14:48:59 hhf423 hmm, an other idea would be to implement a python interface to the socket... - hmm - okay first I'll fight with the nonblocking stdout of the atcmd-mode of libgsmd-tool Sep 12 14:49:06 ScaredyCat: haha Sep 12 14:49:40 i hope navicore ports it's stuff to openmoko. Should be no big deal after they did the maemo thing Sep 12 14:49:59 ScaredyCat: that's the first rule I gave to my daughters when I taught them to drive: treat every other driver on the road as if they were asleep, stupid, or drunk; you may be mistaken most times, but you'll be prepared when you're right. Sep 12 14:50:26 too right Sep 12 14:50:40 some are even all 3 Sep 12 14:50:42 thanks baird, DPThought Sep 12 14:50:43 is anyone having trouble checking out the openmoko svn: http://pastebin.ca/693947 Sep 12 14:51:31 I'm puzzled why the checkout fails, but I can "svn ls http://svn.o-hand.com/repos/tasks/trunk/libkoto" just fine Sep 12 14:51:37 mwester: just about to leave for the rest of the evening. sure if there would be a way for the gsmd script to not produce lockups, I would be all ears. however i tried almost every combination and failed. it is my belief that 90% of our problems will vanish once gsmd will be refactored to using sync processing though. Sep 12 14:51:57 the rest of the 10% may be kernel / serial related Sep 12 14:52:07 mwester: lets discuss this via bugzilla Sep 12 14:52:16 Did you try the stty stuff in the last notes on the bugzilla report? Sep 12 14:52:31 no, but gsmd does exactly that Sep 12 14:52:36 setting hardflow and 115200 Sep 12 14:52:42 so i'd be surprised if it would change anything Sep 12 14:52:48 No, the solution is to *disable* flowcontrol. Sep 12 14:52:56 oh Sep 12 14:53:00 well, need to run now Sep 12 14:53:01 lets talk later Sep 12 14:53:05 ok Sep 12 14:55:48 i don't need no steenkin flow control, nuh sir! Sep 12 14:56:13 ScaredyCat: Re: "every other driver on the road" ... As a commuting cyclist, my view is that 50% of all car drivers have an I.Q below 100... :) Sep 12 14:56:31 i ride a motorcycle D: Sep 12 14:56:45 so i get the worst of both! Sep 12 14:57:41 I get all the crap motorcyclists get, plus an easy target for thrown beer bottles. :/ Sep 12 14:57:57 anrp: are you one of those that weaves in between the stopped traffic to thread ones way up to the front of the line at a traffic signal? (outside of that behavior, most cyclists seem to be fairly reasonable on the road) Sep 12 14:58:11 i've only done that once Sep 12 14:58:16 when there was a 5 mile long backup on a highway Sep 12 14:59:17 someone spit their gum at me but other than that it was uneventful ~_~ Sep 12 15:01:32 anrp: people tend to be jealous when they see other people's choices work better than their own :P Sep 12 15:01:41 heh :/ Sep 12 15:01:56 the reason for the backup was that a state trooper had gotten absolutely nailed at high speed Sep 12 15:02:06 so crap all over the road Sep 12 15:02:29 lanesplitting's illegal for motorbike riders in NSW Australia. Cyclists can do it, and hook-turns, too. Sep 12 15:02:47 lanesplitting is illegal in all of the US except for california Sep 12 15:02:53 for motorcycles Sep 12 15:05:14 what is lanesplitting? Sep 12 15:08:10 car + bike in the same lane Sep 12 15:08:21 iiiek. sounds scary Sep 12 15:08:26 DPThought: usually a bike driving around other traffic in the same lane Sep 12 15:08:32 usually at very low speeds Sep 12 15:09:00 I always thought it was supposed to be a perk ... Sep 12 15:09:03 ah, ok. this is common here. But i don't think it's legal :) Sep 12 15:09:14 ..and car owners know they can get away with 'accidently' opening their car doors without warning :) Sep 12 15:09:15 in lieu of staying dry in thunder storms Sep 12 15:09:29 baird: it depends... Sep 12 15:10:07 if there's no gap on the inside, don't hop onto the pavement and then get upset when a door opens Sep 12 15:11:01 I don't mind cyclists. Sep 12 15:11:08 When traffic comes to a stop on the autobahn all cars have to bould a alley for rescue vehicles. Bikers use this alley when it is free. Sep 12 15:11:34 so not so much sharing the lane but misusing the rescue lane Sep 12 15:11:46 I do mind that it's always the car drivers fault Sep 12 15:12:00 DPThought: no hard shoulder? Sep 12 15:16:26 SpeedEvil: driving the shoulder is very unpopular and widely considered offensive. No biker would do this. :) Sep 12 15:16:51 germany is strange with its rules and regulations Sep 12 15:18:04 In AU, cyclists who ride the shoulder are called 'gutter bunnies'. :) Sep 12 15:18:37 driving the alley is probably as much illegal as driving the shoulder lane. But if you drive at the shoulder you are a asshole and if you drive the alley you are just opportunistic *g* Sep 12 15:19:27 ok...I tried it again to build an ext2 image again but now I get this: ERROR: '/home/OM/openembedded/packages/psplash/psplash_svn.bb' failed Sep 12 15:20:35 http://pastebin.ca/694013 Sep 12 15:34:15 hey: I have the following error: | configure: error: cannot find sources (gdbmdefs.h) in /media/nfs/moko/build/tmp/work/i486-linux/gdbm-native-1.8.3-r2/gdbm-1.8.3 or .. Sep 12 15:34:41 I built for i686 before and had the same error... but then for i686, can anyone help me? Sep 12 15:34:56 nfs D: Sep 12 15:45:48 anrp: so? everything else works.. Sep 12 15:52:25 anyone else having trouble compiling psplash? psplash.c errors: HAND_IMG_WIDTH, HAND_IMG_HEIGHT, HAND_IMG_BYTES_PER_PIXEL and HAND_IMG_REL_PIXEL_DATA all undeclared Sep 12 15:57:00 jseghers_: yes, me too Sep 12 15:57:38 http://pastebin.ca/694013 Sep 12 15:57:42 zefanja: Ok... someone must have forgotten to check something in...I hope they fix it soon Sep 12 15:57:58 yep, exactly what I have Sep 12 15:57:58 I hope so Sep 12 16:02:39 say, does anyone know where / if i could get the openmoko gtk2 theme? Sep 12 16:03:51 anyone getting build errors related to patches in all packages, only to disappear when a clean-package- is done? Sep 12 16:09:31 zefanja: take a look at http://pastebin.ca/694071 Sep 12 16:09:50 zefanja: I'm currently building (it's wading through bitbake files at the moment) Sep 12 16:09:56 but I think that's the problem. Sep 12 16:10:53 zefanja: I spoke too soon.... Sep 12 16:11:28 bye Sep 12 16:12:10 zefanja: actually, that did fix it, I just needed to do a rebuild-package-psplash for it to pick up the change Sep 12 16:12:31 jpp...thx. I will try it Sep 12 16:13:03 A change was made some hours ago to nail down the specific version of the psplash package; looks like it isn't quite right. Sep 12 16:13:38 mwester: yes, someone didn't convert the "hand" image correctly Sep 12 16:13:55 mwester see http://pastebin.ca/694071 Sep 12 16:15:08 now I'm getting errors on "libidn" Sep 12 16:24:17 hi Sep 12 16:24:28 hi Sep 12 16:24:38 hey jseghers_ Sep 12 16:25:46 After fixing psplash, I ran into errors on libidn and popt... both of them, rebuilding worked... Sep 12 16:32:46 hi all Sep 12 16:33:06 zdanek, hi mate Sep 12 16:33:12 I applied fix to bug 788 Sep 12 16:33:19 no hang ups Sep 12 16:33:42 but no clock, antenna, keybord etc. panel in upper right Sep 12 16:33:51 do you have the same? Sep 12 16:34:18 like when you power up gsm antenna and whole panel is gone then Sep 12 16:34:20 same bug Sep 12 16:34:25 is it connected? Sep 12 16:36:30 anybody has any idea? Sep 12 16:36:41 zdanek, no it's not Sep 12 16:36:54 zdanek, it sounds like you have an old version of the gsm applet Sep 12 16:37:04 zdanek, either that, or it still segfaults when you try to use it Sep 12 16:40:05 i have latest packages from buildhost.openmoko.org Sep 12 16:40:21 i had even keyboard icon Sep 12 16:41:07 no i stopped gsmd to try gprs call (i never succeeded) and my neo failed again to connect and died after some seconds Sep 12 16:41:10 strange Sep 12 16:41:12 blank Sep 12 16:44:12 thos: where should I seek for segfault? in /var/log/messages? Sep 12 16:44:23 I watching it and no info Sep 12 16:45:11 :| Sep 12 16:45:35 I found reason Sep 12 16:45:53 I removed from /etc/matchbox/session gsmd restarting lines Sep 12 16:46:11 after putting them again the panel is back Sep 12 16:46:35 it's even asked me for PIN Sep 12 16:47:00 got gsm network! Sep 12 16:47:01 :D Sep 12 16:47:37 dialer works again xD Sep 12 16:48:18 * J-23 :| Sep 12 16:48:48 how are phoen calls going? Sep 12 16:48:52 any reliable yet? Sep 12 16:49:21 * mwester has had reliable phone calls for a couple of days now... (reasonably so, anyway) Sep 12 16:51:33 dcordes_: it's okay Sep 12 16:51:48 sound quality - like normal phone Sep 12 16:52:07 when doing outgoing call from neo - everything's ok Sep 12 16:52:32 but when incoming call hits neo sometimes dialer doesn't popup Sep 12 16:52:53 in calling phone you hear "calling tone" Sep 12 16:53:12 but on neo no dialer response Sep 12 16:53:29 it acts like no incoming call was at the momenyt Sep 12 16:53:50 I didn't try smses yet Sep 12 16:54:00 but now I will install app to do so Sep 12 16:57:33 oooh, a new MS BT Mouse, how cool Sep 12 16:57:54 zdanek: good luck :) Sep 12 16:57:55 hhf423: did you try to connect it to neo? Sep 12 16:58:01 so incoming calls are unstable? Sep 12 16:58:05 thx Sep 12 16:58:17 zdanek: you can not buy it yet. but I should try with my other BT mouse Sep 12 16:58:18 wait I'll try to call Sep 12 16:58:41 hhf423: maybe sbd will try keyboard Sep 12 16:58:44 that should be cool Sep 12 16:59:48 zdanek, what sms application? Sep 12 16:59:52 zdanek, I don't think there is one yet Sep 12 17:00:03 * bedboi is back (gone 00:27:39) Sep 12 17:01:17 does anyone know if there are problems with the parallel build mechanisms? I'm tending to get errors where if I then go and rebuild individual packages things tend to work... Sep 12 17:01:35 thos: their is a small python app with which you can send sms Sep 12 17:01:44 borg_, fair enough Sep 12 17:19:16 I do confirm - if you change from dialer to "today" applet - no incoming calls are seen by dialer Sep 12 17:19:47 I can hear dialing tone in my phone but no reaction on neo Sep 12 17:20:22 hmm, ipkg upgrade gives: ERROR: Cannot satisfy the following dependencies for task-openmoko-pim: libspiff0 (>= 0.7.2) liburiparser0 (>= 0.5.1) Sep 12 17:21:51 and dialing out of neo doesn't work anymore Sep 12 17:21:52 :( Sep 12 17:22:01 brb Sep 12 17:25:27 I can still dial out Sep 12 17:27:04 but, right, if Neo is in Today, it rings but dialer pops not up Sep 12 17:27:13 zdanek: will you file a bug? Sep 12 17:28:18 hmmm Sep 12 17:28:22 | psplash.c: In function 'main': Sep 12 17:28:22 | psplash.c:260: error: 'HAND_IMG_WIDTH' undeclared (first use in this function) Sep 12 17:28:28 openmoko: 03jserv * r2957 10/trunk/src/target/gsm/src/util/atcmd.c: Sep 12 17:28:28 openmoko: Fixlet to libgsmd-tool does not flush stdout in atcmd mode. Sep 12 17:28:28 openmoko: PR833 - http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=833 Sep 12 17:28:28 known problem? Sep 12 17:29:32 openmoko: 03mickey * r2958 10/trunk/src/target/OM-2007.2/libraries/libmokopanelui2/libmokopanelui/ (Makefile.in moko-panel-applet.c): libmokopanelui: remove generated Makefile.in, try to fix rescaling bug wrt. MBPanelScaledImage Sep 12 17:29:43 ixs: a number of folks have encountered that today. Sep 12 17:29:54 oops Sep 12 17:29:56 ixs yes... look at http://pastebin.ca/694071 Sep 12 17:29:59 sorry, will fix in a minute Sep 12 17:30:52 thx Sep 12 17:39:07 hhf423: you mean to create a ticket on bugzilla? Sep 12 17:42:52 yes Sep 12 17:48:05 mtpull Sep 12 17:48:11 psplash bug fixed Sep 12 17:48:34 mickeyl: mtn finds two heads? WTF? Sep 12 17:48:43 What am I doing wrong? I ran make update-makefile; make setup update openmoko-devel-image... However, I'm getting frequent stops where a package foo refuses to build, If I run make rebuild-package-foo, it succeeds and I can make a it more progress, but then another package won't build... Sep 12 17:48:53 repeat ad nausium Sep 12 17:48:53 ixs: welcome to monotone Sep 12 17:48:56 this can happen Sep 12 17:49:01 just chose one Sep 12 17:49:08 or wait until someone merges Sep 12 17:49:11 and then pull again Sep 12 17:49:43 mickeyl: YAY! :D Sep 12 17:49:54 I'm feeling right at home. Sep 12 17:49:59 same chaos as in my room. :) Sep 12 17:50:12 well Sep 12 17:50:17 s/chaos/distributed development/ Sep 12 17:50:25 it's inevitable Sep 12 17:50:34 but so much better than svn Sep 12 17:50:36 :D Sep 12 17:50:45 distributed development ~= chaos Sep 12 17:51:12 but it should be chaos moving in the same general direction... Sep 12 17:51:40 mickeyl: but someone merged it already Sep 12 17:51:55 ya Sep 12 17:54:06 it merges itself every 10 mins Sep 12 17:54:19 mickeyl: did you get that? with the newest everything, the dialer does not pop up when a call comes in when the user is e.g. in Today Sep 12 17:54:40 that's unfortunate Sep 12 17:54:51 I would say so :-) Sep 12 17:55:18 there's a lot in flux atm. wrt to phone functionality Sep 12 17:55:22 are there any news concerning fic's phase2 progress? Sep 12 17:59:01 if you mean gta02, no news except v3 has gone gerber this week Sep 12 17:59:58 went to Moto developer's conference yesterday Sep 12 18:00:20 a Linux playform is planned, but not really available yet Sep 12 18:00:42 basically, you'll develop your apps in JavaME for the time being Sep 12 18:00:58 well, i lost my hope in MOTO, to be honest Sep 12 18:01:11 i hope OM is there to kick enough ass before they release something Sep 12 18:01:12 and then try to get Motorola to certify your app, and then get the carriers to certify Sep 12 18:01:17 jump through hoops Sep 12 18:01:18 heh, fun Sep 12 18:01:55 given how good OpenMoko runs on EZX, i expect it to be a full replacement for the Motorola system in less than 6 months Sep 12 18:01:56 DukeOfURL: that's been Moto's stance forever... they're using J2ME as their app sandbox so they don't have to figure out the security issues for linux. Sep 12 18:01:57 there's also UIQ on Simbian, but there's not much interest Sep 12 18:02:36 if you beg and plead, they will give you a developers cert for you handset--it unloks all the APIs Sep 12 18:04:00 DukeOfURL; you have the coolest /nick ever Sep 12 18:04:42 most aren't old enough to remeber 60s music Sep 12 18:07:21 mickeyl: i'm still having problems with psplash: http://pastebin.org/2307 Sep 12 18:07:44 my cmd is: make update-makefile;make update setup;make openmoko-devel-image Sep 12 18:12:07 DukeOfURL: same problem here, I will fix shortly if mickeyl is not around ... Sep 12 18:12:18 thx Sep 12 18:13:06 forgot to mention... certifying your JavaME app with Motorola and/or your carrier costs big bucks $$$ Sep 12 18:13:36 have to recertify each app within each locale Sep 12 18:15:43 that's purely an operator/moto issue, it's not inherent in the technolody Sep 12 18:15:52 it's a hassle though Sep 12 18:19:20 hi all Sep 12 18:21:35 DukeOfURL: its already been fixed -- update your build and try again Sep 12 18:32:31 they are rather large phones no?(EZX i mean) Sep 12 18:33:22 and no GPS Sep 12 18:33:52 is that a must have these days? Sep 12 18:34:06 i like the htc touch.. its nice and small + no gps.. Sep 12 18:34:13 cbrake: still no joy: http://pastebin.org/2311 Sep 12 18:34:42 how do I re-download the psplash package? Sep 12 18:34:58 anyone else having problems building openmoko-dates2 ? Sep 12 18:35:58 Motorola downplays GPS on their phones. Sep 12 18:35:58 oh, and can someone here tell me how to see a list of .bb files bitbake can find ? Sep 12 18:36:08 DukeOfURL: try the following: source setup-env; cd build; bitbake psplash -c rebuild Sep 12 18:36:21 for some reason, http://wiki.openmoko.org/wiki/Application_Development_Crash_Course isn't working for me. Sep 12 18:36:49 polz, find openembedded |grep bb$ Sep 12 18:37:10 DukeOfURL: the version number was not updated with the fix, so OE does not know to rebuild Sep 12 18:37:28 There is no way to supply ephemeris data to the Motorola GPS Sep 12 18:37:46 Fake GPS satellite. Sep 12 18:37:48 They spent a lot of time explaining GPS using base station signal strength Sep 12 18:37:51 DukeOfURL: after the above completes, then continue with your build as before Sep 12 18:37:51 polz: I'm doing a "make update" with the mokomakefile, then I'll (re-)try making it... Sep 12 18:37:53 (no, not sensible) Sep 12 18:38:22 DukeOfURL, remove the source packge from the sources dir, then remove the build/tmp/stamps/arm4vt/psplash* files (or something like that) Sep 12 18:38:26 Ghiottone: now.. what if I've altered the BBPATH ? Sep 12 18:38:54 polz, altered how? Sep 12 18:38:55 Ghiottone: I'd like to somehow check whether bitbake is actually looking in my local dir or not Sep 12 18:39:24 polz, oh. then create an empty .bb file in one of the dir you put in BBPATH and try to build it Sep 12 18:39:24 Ghiottone: according to the Application_Development_Crash_Course in the wiki Sep 12 18:39:56 cbrake: that worked! thanks. Sep 12 18:40:01 Ghiottone: I've tried that, it can't find it if I'm in the build dir (as mokomakefile seems to do it) Sep 12 18:40:01 DukeOfURL: great! Sep 12 18:41:40 polz, did you change setup-env ' Sep 12 18:42:26 Ghiottone: yep. Sep 12 18:42:52 Ghiottone: I'm pretty sure I've messed something up, that's why I'd like bitbake to tell me where it's looking for .bb files Sep 12 18:44:29 polz, do this: . setup-env; cd build; bitbake -e | grep BBPATH Sep 12 18:44:40 10x Sep 12 18:44:43 polz, I have my local package dir at the end of the path Sep 12 18:50:55 okay, here's the error I'm (still) getting when I try to build: ''ERROR: '/Desktop/Openmoko/moko/openembedded/packages/binutils/binutils-cross_2.18.bb' failed'' Sep 12 18:51:29 ...and the makefile stops. Sep 12 18:51:43 Try make rebuild-package-binutils-cross Sep 12 18:51:47 See if that works. Sep 12 18:51:57 okay, I shall! Sep 12 18:52:32 I'm having terrible problems with the makefile dying and needing to rebuild the offending package and then restart the openmoko-devel-image build. Sep 12 18:53:52 hmm, it's complaining about things missing in tmp/cross/lib Sep 12 18:54:00 libiberty.a is one... Sep 12 18:54:06 Ghiottone: I have it at the start of mine... Sep 12 18:54:15 I'll c/p the error for your perusal... Sep 12 18:54:46 just where the Allmighty Wiki told me to put it ;) Sep 12 18:56:01 http://ponzo.net/newshawk/for_your_perusal.txt Sep 12 18:56:36 ''ERROR: Exception: Message:('Function failed in task: /home/scott/Desktop/Openmoko/moko/build/tmp/work/armv4t-angstrom-linux-gnueabi/binutils-cross-2.18-r0/temp/log.do_stage.8041', )'' Sep 12 18:56:38 Did someone change something that is forcing all of the localization packages to be built? It seems that many (most?) of these packages that I am having to rebuild are ones with tons of localization packages. Sep 12 19:00:34 polz, so is it working now? Sep 12 19:01:25 vallor, I think the gcc version has changes. libiberty is part of its package. try rstarting from scratch if you can Sep 12 19:01:35 s/changes/changed/ Sep 12 19:01:35 Ghiottone meant: vallor, I think the gcc version has changed. libiberty is part of its package. try rstarting from scratch if you can Sep 12 19:03:09 polz: did you get past openmoko-dates2? My build stopped there but a rebuild-package-openmoko-dates2 fixed it. Sep 12 19:03:17 Ghiottone: okay, I shall Sep 12 19:03:34 apt: I shall, thank you Sep 12 19:03:35 You shall, thank you? Sep 12 19:03:54 thank you very much! :) :) Sep 12 19:04:01 | make[1]: Entering directory `/home/polz/openmoko/build/tmp/work/armv4t-angstrom-linux-gnueabi/openmoko-dates2-0.1.0+svnr565-r0/jana' Sep 12 19:04:02 | make[1]: *** No rule to make target `install'. Stop. Sep 12 19:04:09 whaf... ? Sep 12 19:04:23 who's jana? :) Sep 12 19:04:28 Ghiottone: I'm trying to get past the build error Sep 12 19:04:34 apt: did I trigger a macro, sir? Sep 12 19:04:43 I have 0 idea Sep 12 19:05:23 well, I am grateful, sirs Sep 12 19:06:24 how can this sort of crap get into the official svn repository ? Sep 12 19:08:13 oh cripes -- here I was, I thought I offended apt -- but it's a bloody bot, garrr! Sep 12 19:09:53 polz, try with make clean-package-openmoko-dates2 Sep 12 19:10:05 polz, then make rebuild-package-openmoko-dates2 Sep 12 19:10:28 lol @apt :) Sep 12 19:10:32 Ghiottone: tried it. I've now removed the .bb files in the hopes that the thing is fixed in svn, but it's not getting fetched Sep 12 19:11:05 polz, you don't have to do it. put the .bb file back in place Sep 12 19:11:29 polz, instead, remove any reference to openmoko-dates2 in the sources directory Sep 12 19:11:59 polz, and from the build/tmp/stamps/armv4t-angstrom-linux-gnueabi/ directory Sep 12 19:12:13 polz, that done, the rebuild will fetch the package again Sep 12 19:12:49 how can I force mokomakefile to fetch a single file from the repository ? Sep 12 19:13:16 I'll try that Sep 12 19:13:55 polz, it will do it as part of the rebuild-package if you remove the package_name_do_fetch stamp Sep 12 19:13:59 hi there Sep 12 19:14:05 hey BadRobot Sep 12 19:14:15 so how is the OpenMoko project going? Sep 12 19:14:16 BadRobot: 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 12 19:18:31 CWow very nice Sep 12 19:18:48 I loved the look of the boot Sep 12 19:19:08 yeah, it looks great Sep 12 19:19:19 and you get to look at it for a long time, and often :P Sep 12 19:20:40 ;) Sep 12 19:20:52 I like it,but how the technology goes I think it would be great to have Wifi ,HSPDA or 3G Sep 12 19:21:38 so then it could compate with symbian and brew phones Sep 12 19:22:01 but anyway it is a very great start Sep 12 19:22:33 dude, the mass market version will have wifi Sep 12 19:22:47 ok that is nice Sep 12 19:23:03 meanwhile, I'm getting a message repetitively in /var/log/messages: http://ponzo.net/newshawk/problem1.txt Sep 12 19:23:04 can I install any linux program or application on it? Sep 12 19:23:10 is anyone aware of a library that gives you all resolved and unresolved symbols of a shared object or an object file ? Sep 12 19:23:31 something similar to what objdump or nm do Sep 12 19:23:37 pretty much anything (for ARM/EABI) Sep 12 19:24:27 ldd ? Sep 12 19:24:46 oh wait, that's libraries Sep 12 19:26:11 or would "ldd -d -r" do the trick? :\ Sep 12 19:27:20 SpeedEvil: ping Sep 12 19:27:30 check this out http://www.mobile-review.com/fullnews/main/index_eng.shtml#15671 Sep 12 19:27:31 ong Sep 12 19:27:52 they could change a bit the look of OpenMoko Sep 12 19:28:12 i came across a spec on an nxp 802.11b/g chip. Power Consumption 1.6mw standby, 385mW receive, 250mW transmit Sep 12 19:28:27 Atheros should be in the same ballpark Sep 12 19:29:02 more interesting is the power saving modes Sep 12 19:29:21 sleep 200 uW Sep 12 19:29:33 I mean functional Sep 12 19:29:52 ? Sep 12 19:30:21 powerdown RX when not needed Sep 12 19:30:34 perhaps in cooperation with an AP Sep 12 19:30:44 like a BT park? Sep 12 19:31:37 Writchie: it's interesting RX is so high Sep 12 19:31:59 this is true also for BT - signal processing at high speed cost mucho power Sep 12 19:33:09 sort of Sep 12 19:33:16 well, RX also might have some chatter back to the sender Sep 12 19:33:45 Also fake sleep mode, with a server somewhere (ideally local) Sep 12 19:34:19 that sort of power use is actually pretty high, i wonder if that's typical? Sep 12 19:34:26 it limits what you can do with it greatly Sep 12 19:34:30 turn on radio for 1s, login to AP, poll your server to see if you have waiting packets, if not, go to sleep. Sep 12 19:34:34 yes, it's typical Sep 12 19:34:41 or rather low Sep 12 19:35:06 I also came across some comments where the OLPC is seeing 10ms sleep time and 400ms resume. Sep 12 19:35:07 that's wifi for you Sep 12 19:35:19 cw666: the NXP is actually very low Sep 12 19:35:43 previous generation was 415mW receive 731mw@15dBm for transmit 802.11b only Sep 12 19:39:32 utlimate solution will need BT Class 1 and WiFi sharing the same antenna - Class BT + WiFi is necessary to meet UMA Spec Sep 12 19:41:00 Writchie: got a URL for that? Sep 12 19:41:20 nm, found it Sep 12 19:42:59 ^^^ windows user Sep 12 19:43:05 :( Sep 12 19:43:12 xchat doesnt like me Sep 12 19:43:27 still doing the checkout -- and just realized, I'm doing it over my slow connection, woops! Sep 12 19:46:30 hm. rwhitby there? Sep 12 19:55:11 is there a way to restart the panel? Sep 12 19:56:34 restart x Sep 12 20:14:23 DukeOfURL: /etc/init.d/xserver-nodm restart Sep 12 20:15:11 yes, got that. I'm in libgsmd-tool trying to get GSM to register with T-Mob Sep 12 20:16:01 when I do AT+COPS=0 it hangs and won't accept another AT command Sep 12 20:16:49 DukeOfURL: I saw that before -- thing to do is in libgsmd-tool, say "O" (capital oh) to turn the radio on Sep 12 20:17:05 if you tail /tmp/gsm.log you'll get an idea of when the radio is on Sep 12 20:17:14 then you can "r" Sep 12 20:17:32 I'm giving it AT+CFUN=0 to turn off and =1 to turn on Sep 12 20:18:27 DukeOfURL: if you examine /etc/init.d/gsmd , you'll see that there is a magic /proc file that controls the hardware itself -- echo "1" into that file Sep 12 20:18:48 sorry /etc/default/gsmd is where that variable is kept Sep 12 20:19:07 here it is: /sys/bus/platform/devices/gta01-pm-gsm.0/power_on Sep 12 20:20:24 * vallor got all kinds of help getting gsmd to work, hopefully he can help out with others.. Sep 12 20:20:26 ... Sep 12 20:22:02 I can get gsmd running, but giving it AT+COPS=0 command hangs the modem Sep 12 20:22:10 DukeOfURL: did you do the boot variable change? Sep 12 20:22:58 DukeOfURL: did you remove "console=ttySAC0,115200" from bootargs_base Sep 12 20:23:02 bootargs_base? yes. Sep 12 20:23:36 mwester: are you running oe on a treo? Sep 12 20:23:46 oh, new ipkg update / upgrades available Sep 12 20:24:45 how do I run mediaplayer? Sep 12 20:25:12 hhf423: oe = open embedded ;P Sep 12 20:26:04 oh, its in the apps list Sep 12 20:26:29 vallor: no, just on irc from a treo as I travel. Sep 12 20:26:49 snazzy :) Sep 12 20:27:33 how do I create a playlist for that thing? Sep 12 20:28:09 hhf423: you're using mplayer? Sep 12 20:29:44 any other option? Sep 12 20:30:10 madplay Sep 12 20:30:14 but it is text-mode, too Sep 12 20:32:11 in fact -- is there any matchbox-capable media player out there somewhere? Sep 12 20:32:14 gmplayer should im principle work Sep 12 20:33:14 SpeedEvil: alas, it isn't in ipkg yet... Sep 12 20:36:43 how does a xspf playlist look like? Sep 12 20:36:54 is anyone getting audio on a completed call? Sep 12 20:38:30 ah, found it Sep 12 20:38:54 now I get "gstreamer error: ressource busy or not available" Sep 12 20:41:42 mickeyl: http://scap.linuxtogo.org/files/c4b7ac0fecbca7b4d066dfb262979ffb.png Sep 12 20:41:56 DukeOfURL: what do you mean by that? Sep 12 20:42:13 hey...today there are some errors...i want to build a *.ext2 image and got this: http://pastebin.ca/694507 Sep 12 20:42:22 hhf423: I can complete a call to another phone, but there is no audio. Sep 12 20:42:50 with the current images? Sep 12 20:43:11 yes, compiled about two hours ago. Sep 12 20:43:27 DukeOfURL: does the neo make sounds when you tap on the dialer etc? Sep 12 20:43:33 yes Sep 12 20:43:53 well, then probably dialer forgets to change alsa profiles Sep 12 20:43:58 it works for me, here Sep 12 20:44:06 hmmm Sep 12 20:51:57 hey all Sep 12 20:52:13 hhf423, killall pulseaudio Sep 12 20:52:59 how do I choose the headset mike instead of the handset mike ? Sep 12 20:52:59 DukeOfURL, there are alsa profiles in the /etc/ directory Sep 12 20:53:33 alsactl restore -f /etc/.state Sep 12 20:53:47 I think there is a problem with the new date application Sep 12 20:58:39 zefanja, try with make rebuild-package-openmoko-dates2 Sep 12 20:58:54 Ghiottone: then I do not get the error message anymore, but I also do not get any sound Sep 12 20:59:51 Anyone see a problem with lobflac7 and loboggflac dependencies when building their openmoko-devel-image? Sep 12 21:00:07 s/lob/lib/ Sep 12 21:00:08 jpcass meant: Anyone see a problem with libflac7 and loboggflac dependencies when building their openmoko-devel-image? Sep 12 21:00:42 hhf423, strange. it works for me after restarting the media player Sep 12 21:00:49 polz: choose the appropriate alsa profile Sep 12 21:01:19 Ghiottone: so I kill the player, I kill pulseaudio and I start the player and it should work? Sep 12 21:02:44 I see problems in the current build with webkit-gtk not building. Known problem? Sep 12 21:03:38 hhf423, I kill pulseaudio and then the player Sep 12 21:03:47 hhf423, but it should be the same, I think. Sep 12 21:04:20 wow. cutecom is actually very cute :) Sep 12 21:04:22 let me reboot Sep 12 21:06:36 Ghiottone: now I do not have no bootup sound Sep 12 21:07:26 hhf423, wow. I did not mean to kill it *this* hard Sep 12 21:08:00 hahaha Sep 12 21:08:00 try resetting volumes. alsa save its state between reboots Sep 12 21:08:02 crap Sep 12 21:10:22 ha, alsactl restore -f /etc/stereoout.state did the trick Sep 12 21:11:31 now let me call the neo Sep 12 21:11:39 while it playing music Sep 12 21:12:24 cool, dialer pops up, steals audio from mediaplayer Sep 12 21:12:34 but does not return audio to mediaplayer Sep 12 21:14:36 * hhf423 sings to the music Sep 12 21:14:53 hhf423, nice :) Sep 12 21:15:15 this is fun. I need a bigger MMC :-) Sep 12 21:15:42 i wonder if Tomtom wil ever get ported Sep 12 21:15:52 to neo.. that would be awsome Sep 12 21:16:12 hp___: its there for the N770, no? So its not impossible Sep 12 21:17:16 ok, off to bed. Sep 12 21:17:29 * hhf423 waves good night Sep 12 21:19:54 hhf423, nite :) Sep 12 21:20:21 g'nite :) Sep 12 21:20:58 thx Sep 12 21:24:29 hhf423 did you get mediaplayer to work ? I had to add an /etc/asound.conf to get it to work together with touch-click etc. Sep 12 21:24:49 ... for mixing ! Sep 12 21:33:40 4gb Micro SD cards.. unbelievable.. Sep 12 21:33:56 thats just insane Sep 12 21:34:08 is that working? Sep 12 21:34:09 52 Euro.. hmm maybee i should get me one Sep 12 21:34:25 1TB soon :p Sep 12 21:35:00 hehe its going insane so who knows :) Sep 12 21:35:11 last i heard they could get 32Gb on one of them cards.. so Sep 12 21:35:21 not for sale yet i think Sep 12 21:37:13 http://www.engadget.com/2007/05/16/samsung-develops-8gb-microsd-card/ <<-- so who konws Sep 12 21:38:53 http://www.amazon.com/SanDisk-MicroSD-Transflash-Adapter-Packaging/dp/B000SMVQK8 Sep 12 21:39:06 okie now i'm just going to order it and thats it.. Sep 12 21:44:16 i'd just like cheaper larger faster (UDMA) CF cards right now Sep 12 21:45:43 morning Sep 12 21:46:35 hp: i just ordered one online for 53USD with adapter Sep 12 21:47:09 er, the 4gb that is. Sep 12 21:47:57 do those cards do wear leveling? Sep 12 21:49:16 <_buz> they should Sep 12 21:49:46 really? i was under the impression a lot of cheaper cards didnt Sep 12 21:49:56 <_buz> according to WP they should have Sep 12 21:49:59 <_buz> also sd cards Sep 12 21:50:07 WP? Sep 12 21:52:07 notserpe: same here.. with addapter and usb card reader: $52 Sep 12 21:52:39 what's the state of development for my NeoPod? :-) Sep 12 21:52:51 Pod?? :) Sep 12 21:52:52 er maybe that should be neonano? Sep 12 21:52:58 hehe Sep 12 21:53:05 time for some ZzzzZ over here Sep 12 21:53:08 g'nite :) Sep 12 21:53:12 I'm assuming somebody must be working on a media/music player instead of something more important Sep 12 21:53:27 hehe like who Sep 12 21:53:28 'nite Sep 12 21:54:17 okay, when my 4gb arrives, it'll be me ;-) Sep 12 21:54:28 notserpe: The mediaplayer somewhat works, still early (=buggy and limited), but goot progress the last few days. Sep 12 21:56:30 anyone have any thoughts on security? selinux polcies for everything? Sep 12 22:01:25 anyone here for answering about issues with projects.openmoko.org? Sep 12 22:05:57 notserpe: wine and selinux do not get along Sep 12 22:06:26 oh wait, I'm in #openmoko Sep 12 22:07:18 dammit, and here I thought he was going to port wine to ARM Sep 12 22:09:28 wine and arm :D ? Sep 12 22:11:08 Qemu's ARM host CPU status seems to be Testing, incidentally :] Sep 12 22:12:16 there's always libwine, of course... Sep 12 22:20:48 Does anyone know what order number they are now filling? Are they shipping any? Sep 12 22:20:49 jeremym1: 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 12 22:48:45 what package is this "media player" you guys were using? Sep 12 22:50:23 openmoko-mediaplayer2 Sep 12 22:50:48 but its more a musicplayer then a mediaplayer Sep 12 22:50:53 at the momment Sep 12 22:51:24 ah okay Sep 12 22:51:26 thank you :) Sep 12 22:51:34 sweet Sep 12 22:51:38 usb host works Sep 12 22:51:45 <4>changing usb to host Sep 12 22:51:45 <6>usb 1-2: new full speed USB device using s3c2410-ohci and address 9 Sep 12 22:51:45 <6>usb 1-2: configuration #1 chosen from 1 choice Sep 12 22:51:46 <6>hub 1-2:1.0: USB hub found Sep 12 22:52:01 a question though: Sep 12 22:52:11 does it actually function though. Sep 12 22:52:21 what you see if a hum plugged into it Sep 12 22:52:24 *hub Sep 12 22:52:31 that's the root hub Sep 12 22:52:49 there are two host controller Sep 12 22:52:50 ah Sep 12 22:52:51 there are two host controllers Sep 12 22:52:52 <6>usbcore: registered new interface driver usb-storage Sep 12 22:52:52 <6>USB Mass Storage support registered. Sep 12 22:52:56 yes Sep 12 22:52:58 yes it does Sep 12 22:53:02 is that plugging something in? Sep 12 22:53:07 yep Sep 12 22:53:10 dang, the media player doesn't seem to have made it to ipkg yet Sep 12 22:53:12 What did you have to do? Sep 12 22:53:17 i even mounted my ausb stick Sep 12 22:53:20 *usb Sep 12 22:53:25 *man, i can't type Sep 12 22:53:32 the patch in SVN, and my proposed one? Sep 12 22:53:35 i tweaked the patch in svn a bit Sep 12 22:53:45 was that yours? Sep 12 22:53:46 to set the pullups Sep 12 22:53:46 yeah Sep 12 22:53:59 it also, well, kinda needed one other thing Sep 12 22:54:01 I posted a tiny patch on the ML to do the pullups Sep 12 22:54:05 yeah Sep 12 22:54:12 That I hadn't actually compiled. Sep 12 22:54:22 well, it does Sep 12 22:54:29 vallor: It's in packages/openmoko2 ! Sep 12 22:54:46 oh -- thank you! Sep 12 22:54:50 i was under the impression it was some kind of convention that /sys knobs could handle trailing /n's and the like Sep 12 22:55:06 that was the thing that i couldn't get for a bit Sep 12 22:55:31 i needed to "echo -n" where i was "echo"-ing Sep 12 22:55:54 so i made the strcmp into a strncmp, and that did it Sep 12 22:56:15 All we need now is to find the secret USB2 mode :/ Sep 12 22:56:25 Though 1.1 is undeniably handy. Sep 12 22:56:32 waaait... can the hardware really do that? Sep 12 22:56:42 no friggin way... Sep 12 22:56:43 No. Sep 12 22:56:54 Sorry. Sep 12 22:56:56 darn Sep 12 22:57:04 Annoyingly the next version can't either. Sep 12 22:57:15 USB2 would be so nice for some things. Sep 12 22:57:27 my next task is to get v4l going so i can have a camera on this thing Sep 12 22:57:44 * SpeedEvil passes orpheus a roll of duct tape. Sep 12 22:58:03 What camera are you thinking of? Sep 12 22:58:07 actually, i've been directed at a 3d printed and told to re-make the back cover Sep 12 22:58:50 whichever camera is closest, at the moment, a logitech something Sep 12 22:58:50 Much 3d printed stuff unfortunately has mechanically sucky properties. Sep 12 22:59:22 this one lays it out layer by layer, so the layers are stong, but the bonding between them is not Sep 12 22:59:30 but still, good enough Sep 12 22:59:39 cameraphone, here we come Sep 12 22:59:43 :) Sep 12 22:59:48 but yeah Sep 12 22:59:56 what do I do with this patch now? Sep 12 23:00:05 dunno. Sep 12 23:00:07 :/ Sep 12 23:00:13 bugzilla? Sep 12 23:00:19 ... could work Sep 12 23:00:59 does anyone here have svn access and want to add usb host happiness? Sep 12 23:01:33 also, after I discovered that the trailing /n was a problem, i tried something on a hunch Sep 12 23:01:53 i hadn't been able to force the phone to fast charge before Sep 12 23:02:32 but it's the same there, if you "echo -n " to /sys/devices/platform/s3c2410-i2c/i2c-adapter/i2c-0/0-0008/chgmod, it works Sep 12 23:02:43 Ah. Sep 12 23:02:47 That's a bug IMO. Sep 12 23:03:02 so I'm wondering, yeah, is it supposed to be that way? Sep 12 23:03:03 echo -n fast_cccv you mean? Sep 12 23:03:09 yeah Sep 12 23:03:43 this is needed, since i'm now on a charger, and my usb lines are going crazy places (to the hub) Sep 12 23:03:45 so when I enter a playlist in the media player, it tanks Sep 12 23:03:59 mpayer -playlist file? Sep 12 23:04:04 works for me Sep 12 23:04:13 SpeedEvil: no, the gui media player Sep 12 23:04:23 ah Sep 12 23:04:24 SpeedEvil: it just came out Sep 12 23:04:25 mpayer, it's a tool for paying media... Sep 12 23:05:14 vallor: You need to create ~/playlist/ Sep 12 23:05:18 vallor: You need to create ~/playlists/ Sep 12 23:05:21 speedevil: thanks for making that change to the usb patch, i never would have sorted that out Sep 12 23:05:26 frma71: !! okay Sep 12 23:05:32 * orpheus goes to try out a few more things Sep 12 23:06:15 also: is /proc/bus/usb/devices deprecated or something? Sep 12 23:06:23 the neo doesn't have it... Sep 12 23:06:41 it's a fake filesystem Sep 12 23:06:49 aah Sep 12 23:06:49 you've got to mount /proc/bus/usb/devices -t none Sep 12 23:06:51 or something Sep 12 23:07:07 you've got to mount /proc/bus/usb/ -t none Sep 12 23:07:17 procbususb on /proc/bus/usb type usbfs (rw) Sep 12 23:07:20 thanks Sep 12 23:07:38 /sbin/mount -v usbfs /proc/bus/usb -t usbfs Sep 12 23:08:16 I was needing to work out how to turn the charger completely off - which that hint should do. Sep 12 23:08:28 (to put internal power to the 5V pin) Sep 12 23:08:53 speedevil: hunh? Sep 12 23:09:06 * SpeedEvil has a soldering iron. Sep 12 23:09:43 you mean, how to turn charging off? Sep 12 23:09:48 why would you do that? Sep 12 23:10:19 just make the funny three-headed cable... Sep 12 23:10:35 Because if you're supplying 5V power from the battery in the neo, then leaving the charger enabled would be bad. Sep 12 23:11:09 that's the thing, the port on the side of the neo can't put out power on those lines Sep 12 23:11:18 or so I thought Sep 12 23:11:22 * SpeedEvil has a soldering iron. Sep 12 23:11:37 you don't even get a 5V source, do you? Sep 12 23:11:42 No. Sep 12 23:12:08 i mean, you could cross your fingers and hope that a usb device will run off of 3.7, or get a voltage multiplier Sep 12 23:12:28 but aren't voltage multipliers pretty inefficient? Sep 12 23:12:33 I'm planning a for-sale accelerometer, USB power, and battery monitor module. Sep 12 23:12:38 90%+ in many cases. Sep 12 23:13:07 hurm Sep 12 23:13:09 cool Sep 12 23:13:17 compromise on mass, volume, price, and efficiency Sep 12 23:13:22 as usual. Sep 12 23:13:25 heh Sep 12 23:17:30 if only it wasn't such a pain in the neck to get at the removable media Sep 12 23:17:33 *sigh* Sep 12 23:20:40 speedevil: timeframe? Sep 12 23:21:01 notserpe: at best a month Sep 12 23:25:11 sounds good, how much? Sep 12 23:26:18 Umm. Sep 12 23:26:42 I think I signed up on your interest page on the wiki IIRC. Sep 12 23:26:46 Probably around $50 Sep 13 00:28:39 frma71: Can I see the asound.conf you created? I'm kinda wondering how to get the audio apps to share resources properly myself Sep 13 00:30:27 it's more than asound.conf Sep 13 00:30:50 it will need a special daemon to route the inputs and outputs with the available A/D and D/As Sep 13 00:31:02 eventually Sep 13 00:31:37 esd? Sep 13 00:35:35 I thought either esd or dmix would be used but it seems to not be the case Sep 13 00:35:58 As for PulseAudio... I'm not sure that it's the right choice for an embedded device Sep 13 00:36:34 Which driver model is the current sound drivers using? If ALSA, I think that already includes a mixer; dunno if it is versatile enough though. Sep 13 00:39:17 Lynet: alsa... and the mixer is dmix, which was mentioned above Sep 13 00:45:12 The mixer is a small part of it. Sep 13 00:45:31 it's not a conventional soundcard, where you have one or two basic channels, and can mix stuff into them Sep 13 00:45:55 you have several inputs and outputs, and A/D and D/As, and can route audio flexibly between devices. Sep 13 00:46:29 For example, in principle, you can have GSM audio going direct from microphone-GSM module, and GSM-module to earpiece, with the CPU off. Sep 13 00:46:42 If ALSA is already using dmix then why do we get "resource busy" errors? Sep 13 00:47:22 Or you can have it going from GSM module-A/D - software voice processing - D/A - earpiece, and the same for the microphone Sep 13 00:47:29 Horribly complex. Sep 13 00:47:51 and at the same time you might be streaming music to a sound system. Sep 13 00:52:01 clear Sep 13 01:09:49 abraxa_: because we are not yet using dmix Sep 13 01:10:00 which i'm fixing as we speak Sep 13 01:10:02 :D Sep 13 01:10:06 Oh, wonderful =) Sep 13 01:10:10 Thank you Sep 13 01:11:01 i just hope we can use it at all Sep 13 01:11:12 since every mixing stage introduces more cpuload and latency Sep 13 01:11:22 and we are already at max. w/ gta01 Sep 13 01:12:03 Yeah - which reminds me... didn't you say that PulseAudio was buggy and responsible for causing such a high load when using gstreamer? Sep 13 01:14:43 no Sep 13 01:14:51 at least i didn't put it that way Sep 13 01:14:59 the high gstreamer load depends on Sep 13 01:15:03 a) the proper pipeline Sep 13 01:15:09 (or rather improper) Sep 13 01:15:10 and Sep 13 01:15:26 b) remixing Sep 13 01:15:44 gstreamer->pulseaudio->alsa doesn't cut it on gta01 Sep 13 01:15:50 so the alternatives are Sep 13 01:16:03 a) not mixing, i.e. autoreleasing audio when idle Sep 13 01:16:03 or Sep 13 01:16:27 b) finding a route through the mixer that doesn't convert too many times Sep 13 01:18:08 Ah, okay. I was wondering because I'm now forcing the playbin to use alsasink but it's still showing the same poor performance. Looks like I'll have to manually construct a more optimized pipeline than playbin does after all Sep 13 01:18:15 yah Sep 13 01:18:23 just use: Sep 13 01:18:28 filesrc ! mad ! alsasink Sep 13 01:18:31 for mp3 Sep 13 01:18:49 this has accetable load when going directly to alsa Sep 13 01:18:54 ~25% cpu Sep 13 01:19:11 when going through pulseaudio to alsa, it stutters horribly Sep 13 01:19:20 so i'm checking whether dmix is better Sep 13 01:19:25 (which i doubt, honestly) Sep 13 01:20:22 Should I prepare several pipelines for different stream types (mp3/ogg/wav/etc.) or should I construct/destroy them with every file load? Not sure which is better so maybe you have some experience with that Sep 13 01:20:43 constructing a pipeline is reasonably lightweight Sep 13 01:20:49 so better do it on-demand Sep 13 01:20:55 okay, will do that then :) Sep 13 01:20:56 s/lightweight/fast/ Sep 13 01:37:27 hmm Sep 13 01:37:35 root@fic-gta01:/etc$ gst-launch filesrc location=/tmp/foo.mp3 ! mad ! alsasink device=dmix Sep 13 01:37:42 ERROR: from element /pipeline0/filesrc0: Internal data flow error. Sep 13 01:37:45 *sigh* Sep 13 01:45:19 Ouch Sep 13 01:46:19 i can't get gstreamer to use alsa in dmix mode Sep 13 01:46:22 odd Sep 13 01:50:48 mickeyl: Can other apps use dmix then? Sep 13 01:51:46 aplay seems to can Sep 13 01:51:58 i can run aplay -D dmix multiple times Sep 13 01:52:09 so it doesn't look like a problem with dmix per se Sep 13 01:54:32 hey :D Sep 13 01:56:28 i really like the openMoko idea but i just wonder if neo is the right hardware i would liek for my next mobile phone Sep 13 01:58:51 DrHalan: Hellos! What do you want to know? Sep 13 01:59:45 if i easily can use openMoko on a diffrent mobile device :P Sep 13 01:59:47 mickeyl: frma71 seems to have a working asound.conf so it seems possible - maybe compare with his? Sep 13 02:00:13 abraxa_: sure. i'd be surprised if it was a configuration problem, but who knows. where can i get that? Sep 13 02:00:28 DrHalan: See http://dominion.kabel.utwente.nl/koen/cms/angstrom-running-on-an-motorola-a780 for example Sep 13 02:00:50 mickeyl: I've been waiting for a response from him myself, unfortunately Sep 13 02:03:24 ah so its also usable on non-touchscreen-devices? Sep 13 02:03:25 DrHalan: 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 13 02:05:47 DrHalan: I can't give a definite "yes" or "no" on that one but as it's open source I'm sure that someone will make it so - if not the core team itself, once the base distribution is ready for mass market Sep 13 02:14:42 ok, we can forget dmix Sep 13 02:15:04 we need to pipe the mad data through audioresample ! audioreconvert to make that work Sep 13 02:15:09 since mad outputs 32bit int Sep 13 02:15:16 and dmix wants float Sep 13 02:16:03 Just curious... how did you figure that one out? Sep 13 02:16:23 friendly guys on #gstreamer Sep 13 02:16:49 so gta01 is doomed to exclusive alsa access Sep 13 02:17:19 all i can do is decrease the pulseaudio unloading timeout Sep 13 02:17:50 well Sep 13 02:18:01 we could ask pulseaudio to explicitly unload its driver Sep 13 02:18:05 from the media player Sep 13 02:18:12 but that's nasty :/ Sep 13 02:19:34 Hmm... my main concern is: how to make the dialer wait with ringtone playback? The media player would have to pause first - sounds to me like we'd need several d-bus events for that scenario alone Sep 13 02:20:15 nasty nasty Sep 13 02:20:20 Indeed... Sep 13 02:20:43 now if we could hack the mad plugin to output 16bit Sep 13 02:20:50 then we could feed that into the pulseaudio sink Sep 13 02:20:58 which accepts 16bit int Sep 13 02:21:07 just not 32bit. Sep 13 02:21:36 Why PulseAudio? Sep 13 02:22:21 because pulseaudio does a lot what we want Sep 13 02:22:23 it mixes Sep 13 02:22:26 it contains a sample storage Sep 13 02:22:35 and some interesting plugins Sep 13 02:22:45 e.g. to feed the x11 bell into audio Sep 13 02:22:49 and lot more Sep 13 02:22:55 plus it's lightweight Sep 13 02:23:03 and the mixing code is faster than dmix'es Sep 13 02:23:26 Oh okay, I didn't realize it did mixing as well :) Sep 13 02:23:32 gstreamer ! mad ! audioconvert ! pulseaudio _nearly_ works Sep 13 02:24:01 it's just stuttering, so if we could get mad to output 16bit, everything should work even on gta01 Sep 13 02:24:08 Okay Sep 13 02:24:25 I'll see if I can get it done myself - if not I'll let you know; is that okay with you? Sep 13 02:24:41 sounds pretty good. Sep 13 02:25:00 that will involve some nice gstreamer hacking which i absolutely don't have time for :) Sep 13 02:25:52 ah, well, one thing Sep 13 02:26:10 it might be easier to hack gst-plugin-pulse to input 32 bit than to hack gst-plugin-mad to output 16 bit Sep 13 02:26:14 *shrug* Sep 13 02:27:02 Okay, I'll decide on that after looking at the code Sep 13 02:27:54 cool Sep 13 02:39:43 hmm. 04:40 Sep 13 02:39:45 time to sleep Sep 13 02:39:46 g'night folks Sep 13 02:43:15 Night mickey Sep 13 02:55:24 so how are phone calls doing with recent builds? Sep 13 02:55:53 * hozer just got an iwireless phone so I need to yank the SIM card and try it Sep 13 02:56:37 As long as you don't do anything weird (like try to use the "today" application) it works nicely. :D Sep 13 02:57:06 The first trick is to get a recent build, though. **** ENDING LOGGING AT Thu Sep 13 02:59:57 2007