**** BEGIN LOGGING AT Mon Sep 10 02:59:57 2007 Sep 10 03:00:58 this gsmd just doesn't like to stay running Sep 10 03:01:20 mine just died too Sep 10 03:02:24 anyone heard rumors about 800/850 GSM being disabled in P1 phones? Sep 10 03:02:35 (FCC related) Sep 10 03:02:49 I hadn't -- 800/850 mhz? Sep 10 03:03:06 I can find no suggestion of the such on the FCC site Sep 10 03:03:28 filings there detail all bands Sep 10 03:03:33 where'd you hear that? doesn't sound likely.. Sep 10 03:03:58 someone was in here a few days ago suggesting such a thing Sep 10 03:03:58 it's like the modem just isn't saying anything back Sep 10 03:04:48 the FCC approval docs don't include 850 Sep 10 03:05:00 the assumption is that the firmware has 850 disabled for that reason Sep 10 03:05:05 ''write(2, "interval expired, starting next alive inquiry\n", 46) = 46'' Sep 10 03:05:11 oh right.. my memory must be bad then, i thought .au used 850 for some reason Sep 10 03:05:49 there is no official word from FIC/OpenMoko on the issue (which is probably due to the bad press which would result from such an official announcement) Sep 10 03:06:06 I have no inside information - all of that is idle speculation Sep 10 03:06:20 (except the FCC bit - you can check the docs yourself) Sep 10 03:07:22 vallor: joe is joe in OE Sep 10 03:07:30 want me to add it to the feed.bb ? Sep 10 03:07:46 rwhitby: say the only gsm provider in my area operated on the 850 band only, then an error of CME 30 "no network available" makes perfect sense Sep 10 03:07:54 cesarb: nice patch...dial out with dialer works for me now (with manual gsmd restart and x-server restart)...though I did rebuild the whole thing to fix my webkit fail :-/ Sep 10 03:08:55 I think that has to be it. Sep 10 03:12:03 * technil will have to get a phone tomorrow :( Sep 10 03:13:56 okay, I've got a gsm.log that shows what's going on for me... Sep 10 03:14:10 http://ponzo.net/gsm.log Sep 10 03:18:19 rwhitby: if it were available to ipkg, would be nice -- but I've got an nfs mount to my workstation, so I can edit files here using the nfs mount as a staging area Sep 10 03:19:27 vallor: have you removed console=ttySAC0 from u-boot env? Sep 10 03:19:32 vallor: check out http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=574 Sep 10 03:20:39 rwhitby: yes indeed Sep 10 03:22:27 ooooo, thanks technil Sep 10 03:22:32 still reading.... Sep 10 03:22:54 * rwhitby sees if emacs will cross-compile for the neo ... Sep 10 03:26:26 hmm... Sep 10 03:26:29 root@fic-gta01:/eva$ cu -l /dev/ttySAC0 Sep 10 03:26:29 cu: open (/dev/ttySAC0): Permission denied Sep 10 03:26:41 cu: /dev/ttySAC0: Line in use Sep 10 03:26:49 but gsmd isn't running Sep 10 03:30:03 I get the same exact error vallor Sep 10 03:32:02 maybe it's just like that Sep 10 03:32:26 lsof | grep ttySAC0 returns nothing Sep 10 03:32:39 nor does fuser Sep 10 03:32:55 doesn't it seem like you should be able to use cu on the serial port? Sep 10 03:36:15 i recall someone saying something about how the modem was placed in the design.. Sep 10 03:36:42 arrgh, can't remember Sep 10 03:41:36 dmesg shows this, also: Sep 10 03:41:43 <4>Alignment trap: libgsmd-tool (1383) PC=0x0000982c Instr=0xe594e004 Address=0xbee16633 FSR 0x013 Sep 10 03:41:43 <7>modem wakeup interrupt Sep 10 03:42:13 vallor: cu will say permission denied unless you do the chmod dance on the wiki page Sep 10 03:43:36 vallor: alignment trap is bad... unless the kernel has alignment fixup enabled, the process will read in some random way and probably die due to the invalid value Sep 10 03:44:02 vallor: and alignment fixup is expensive, so it's almost always a bug Sep 10 03:44:17 vallor: (the bug would be in libgsmd-tool in this case) Sep 10 03:44:32 * cesarb goes to sleep Sep 10 03:44:33 hmm, k Sep 10 03:44:41 thank you cesarb! Sep 10 03:50:31 * moko-bunny wonders how to set the correct time on his Neo Sep 10 03:53:58 I think it's usually done with /etc/localtime, but I'm not familiar with embedded linux that much Sep 10 03:54:00 moko-bunny: http://wiki.openmoko.org/wiki/Getting_Started_with_your_Neo1973#Setting_date_and_time Sep 10 03:54:46 technij: sweet, thanks! (found /proc/driver/rtc BTW that shows if it took :) ) Sep 10 03:55:13 ntpdate would be a handy thing on the phone ;P Sep 10 03:55:35 have it interpolate based on gps data Sep 10 03:55:53 set the timezone correct wherever you are Sep 10 03:56:12 yeah gps gives time, i'm pretty sure you can get the time from the mobile network too Sep 10 03:57:04 * technil wanders around the fcc site .... Sep 10 04:01:55 for timezones, look at http://www.rwhitby.net/blog/nslu2-linux/setting-the-openmoko-timezone.html Sep 10 04:12:01 aha.. FCC ID# EUNGTA01BV4 filings for 1850.2 to 1909.8 and 2402.0 to 2480.0 Sep 10 04:15:14 yep, no 850 bands Sep 10 04:15:47 well I guess I'm screwed until new firmware / filings Sep 10 04:16:41 hey all Sep 10 04:17:04 running into problems building MokoMakefile according to the wiki instructions, using Gentoo, anyone have a sec to help me out? Sep 10 04:17:15 * technil needs to sleep Sep 10 04:17:27 make openmoko-devel-image gives me these notes: Sep 10 04:17:35 NOTE: Resolving missing task queue dependencies Sep 10 04:17:35 NOTE: preferred version 2.5 of glibc not available (for item virtual/arm-angstrom-linux-gnueabi-libc-for-gcc) Sep 10 04:19:02 vallor: it was fun, maybe i'll see you again on irc.. Sep 10 04:19:07 ultimately it fails giving me other errors, and won't complete Sep 10 04:19:26 take care Sep 10 04:19:50 meanwhile, something just failed catastophically, where "ls: Permission denied" is popping up... Sep 10 04:19:51 rwhitby thanks illuminating the 850 issue. Sep 10 04:20:01 dooh Sep 10 04:20:07 night Sep 10 04:20:18 er whatever time it is where you are Sep 10 04:20:25 iandouglas736: the NOTE is innocuous Sep 10 04:20:35 iandouglas736: pastebin the first error Sep 10 04:20:41 ~pastebin Sep 10 04:20:42 i heard pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.ca, or http://channels.debian.net/paste, or http://rafb.net/paste/, or http://pastebin.com is usually painfully too slow and unresponsive to use, use one of the other pastebin sites, or dpaste.com is a very nice pastebin as well Sep 10 04:21:09 * rwhitby sees if emacs will now install on the neo ... Sep 10 04:21:46 the emacs .ipk is 16MB Sep 10 04:26:08 rwhitby: the log file created is 1.1MB of text, pastebin will only allow 150kb, I'll post it on my own web site for you to view Sep 10 04:26:15 * mwester will settle for "joe" - considerably smaller... Sep 10 04:26:43 iandouglas736: just paste the errors, not the whole thing Sep 10 04:27:16 mwester: it's just something that *has* to be done. once. Sep 10 04:27:31 :-D Sep 10 04:27:58 * rwhitby wonders if 48MB of ramdisk will be enough for the expanded package ... Sep 10 04:28:09 nope Sep 10 04:28:43 (script) wiki RecentChanges: [[Community_nominees_for_GTA02_P0]] [[Phase_1_Software_Testing]] [[USB_Networking]] Sep 10 04:28:45 will have to wait until tonight and try it on the other neo after booting from SD card Sep 10 04:29:49 jserv2: will you integrate Telepathy Framework ? Sep 10 04:31:57 right, I saw a soc project for integrating telepathy framework Sep 10 04:32:35 PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND Sep 10 04:32:36 1328 root 15 0 8348 5016 2784 S 50.8 4.0 0:17.01 emacs-22.0.50 Sep 10 04:32:48 ok, once was enough. Sep 10 04:33:33 at least now we can truthfully answer the question "Will it run Emacs?" with a "Yes!" Sep 10 04:34:04 It runs emacs. Whatever problems it may have, if it can run emacs, it gets a thumbs-up from me! Sep 10 04:34:48 * rwhitby refrains from adding the emacs package to task-openmoko-feed ... Sep 10 04:35:17 Ah, but is joe in the feed yet? Sep 10 04:35:38 Hehe Sep 10 04:35:52 nope, but it builds fine Sep 10 04:36:17 rwhitby: Don't add that horrific program to the feed! :P (*starts a little morning flame*) ;) Sep 10 04:36:58 It's *NOT* a horrific program. It's an operating system with an integrated text editor. :p :p Sep 10 04:37:22 :D Sep 10 04:37:39 it's an operating system abstraction layer Sep 10 04:44:17 * mwester idly wonders if it would get his key revoked if he added xemacs to the openmoko image... ;) Sep 10 04:51:38 ScaredyCat: you there? Sep 10 04:56:37 rwhitby: http://pastebin.ca/689634 Sep 10 05:43:59 iandouglas736: That's a known issue, look at http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=747 Sep 10 05:44:59 basically, add network to the QT += at line 6 in src/tools/uic3/uic3.pro Sep 10 05:58:09 CM: thanks much Sep 10 05:58:39 :) Sep 10 06:02:18 hi all guys Sep 10 06:03:05 morging Sep 10 06:08:16 anyone knows how to use -dbg libraries Sep 10 06:08:21 i mean how to link to them Sep 10 06:14:36 how well is openmoko running on the a780? Sep 10 06:18:31 ColdFyre: You mean koens blog post? Sep 10 06:18:45 yup Sep 10 06:19:17 ive got openezx running, but i still see a lack of sound, phone is still a brick Sep 10 06:19:59 Maybe ask around in #oe or some ezx channel? Sep 10 06:20:22 I have no clue how it's going with getting 2.6 working Sep 10 06:21:00 nah, i was just saying it's not supported yet, just polling to see who's got what neat things doing stuff:) Sep 10 06:21:14 Hehe :) Sep 10 06:24:07 heya, does the latest openmoko image not kick right into matchbox? Sep 10 06:24:23 or rather, I'm just updating from ipkg Sep 10 06:25:52 I tried a clean build from scratch on saturday, and that one worked so-so Sep 10 06:26:16 But the gsmd issues might have been fixed by now Sep 10 06:26:19 hmmm...mines getting stuck in the neod (neo daemon) Sep 10 06:26:20 hmmm Sep 10 06:26:31 Wish there was a no-breakage policy Sep 10 06:26:35 ... Sep 10 06:26:37 I haven't tried updating via ipkg Sep 10 06:26:50 CM: so you just re-image it? Sep 10 06:26:52 I wish they tried to test it at least once before commiting Sep 10 06:26:57 rejon: Yes Sep 10 06:27:11 CM: Yes, I totally wish so too Sep 10 06:27:21 who is on here from the openmoko team Sep 10 06:27:23 ? Sep 10 06:27:30 Since that seems to give different results compared to ipkg, and more people use ipgk Sep 10 06:27:47 I can blog about it and go through the planet.openmoko, but maybe that would be negative press Sep 10 06:28:07 man, we should just be able to do: ipkg update && ipkg upgrade and this thing should just work Sep 10 06:28:11 and most definitely not crack out Sep 10 06:28:14 that should be policy IMO Sep 10 06:28:18 Well, stefan_schmidt, roh, mickey|away, alphaone and some more use to be here Sep 10 06:28:23 that would be too easy Sep 10 06:28:41 There's also a locked dev channel I think Sep 10 06:28:48 yeah, I saw that Sep 10 06:28:51 suck Sep 10 06:29:41 rejon: Or maybe have one stable and one unstable feed Sep 10 06:29:50 Yes, I totally agree Sep 10 06:30:01 So that people who want to try testing and bugfixing set it on the -dev feed Sep 10 06:30:07 agree Sep 10 06:30:08 The rest get a working phone Sep 10 06:30:31 stefan_schmidt mickey|away roh: what do you guys think? Sep 10 06:30:54 rejon: have you removed console=ttySAC0 from u-boot bootargs_base ? Sep 10 06:31:28 rejon: It's increased work to maintain two feeds though. It's like gentoos stable/unstable keywords. Sep 10 06:31:30 as for stability, the OM _svn.bb recipes should reference the most recent *tested* svn revision for each tool. Sep 10 06:31:37 Maybe it's too early Sep 10 06:31:46 rwhitby: agreed :) Sep 10 06:32:03 still thought, no breakage should be law Sep 10 06:32:10 just like *any* good software project Sep 10 06:32:15 You can't stop developers from checking in to svn, but you *can* use the new SVNREV facility in OE to isolate that and put the responsibility on the developer to only update the SVNREV in the .bb file when some rudimentary testing has been done Sep 10 06:32:17 hello guys Sep 10 06:32:34 u probably a Sep 10 06:32:44 rwhitby: hell yeah, can stop devs from committing broken code...that is just bad policy Sep 10 06:32:54 or rather not stop, but encourage ;) Sep 10 06:33:00 have other topics Sep 10 06:33:15 rejon: right - in theory, and in policy. if you isolate using .bb SVNREV, then you can do it in practise too ... Sep 10 06:33:27 but I am new to openmoko and have to compile that newes image Sep 10 06:33:38 I've amazingly stayed out of hacking on the openmoko for some months ;) been super preoccupied with creative commons stuff Sep 10 06:34:01 well, except team at CC hacking on liblicense which will go on openmoko too Sep 10 06:34:25 rejon: so I agree the policy should be no breakage, but I say the implementation mechanism should not rely only on SVN stability on trunk. Sep 10 06:34:40 rwhitby: so, I guess I can't just use ipkg to keep up-to-date...suck Sep 10 06:34:52 well yes Sep 10 06:35:01 rejon: I'm using snapshots and ipkg upgrade right now Sep 10 06:35:12 u obiously know about that problem :) Sep 10 06:35:24 see my latest post to device-owners for the recipe I use for a working phone Sep 10 06:35:41 RalphEichelberge: How far along have you gotten, and what is it you want to do? Run OpenMoko in qemu? Sep 10 06:35:46 where can i find that? Sep 10 06:35:47 RalphEichelberge: 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 10 06:36:22 well that wiki didnt help me Sep 10 06:36:40 i just made that make update .... thing Sep 10 06:36:57 rwhitby: do you have a url to that post? Sep 10 06:37:00 Is it possible to increase the sensitivity of the microphone in the Neo1973 to use it as a voice recorder, say for recording uni lectures? Sep 10 06:37:37 Man, so you are saying that there is a customized image and then ipkg you use to get this thing to work Sep 10 06:37:44 OSError: [Errno 2] No such file or directory: '/home/openmoko/moko/build/tmp/work/armv4t-angstrom-linux-gnueabi/cups-1.2.10-r2/install/cups-locale' Sep 10 06:37:44 ERROR: Task 1430 (/home/openmoko/moko/openembedded/packages/cups/cups_1.2.10.bb, do_package_write_ipk) failed Sep 10 06:37:47 why don't they/we just use your custom thing as law Sep 10 06:37:59 ugh...just want to get the base image, and ipkg that ish Sep 10 06:38:08 seems like too many points of failure otherwise Sep 10 06:38:16 rejon: no, the 2007.08 OM snapshot release, then ipkg upgrade from buildhost from there Sep 10 06:38:41 man, I need to read up Sep 10 06:38:43 (plus some tweaks to work-around bugs) Sep 10 06:38:43 i just did, what i found at http://wiki.openmoko.org/wiki/MokoMakefile Sep 10 06:39:12 RalphEichelberge: when did you last "make update image" ? Sep 10 06:39:48 rejon: http://lists.openmoko.org/pipermail/device-owners/2007-September/000449.html Sep 10 06:40:01 Friday evening Sep 10 06:40:27 you need to do it three times in a row before reporting a bug here Sep 10 06:40:44 well i did Sep 10 06:40:45 (since OE moves so fast - that bug was fixed over the weekend) Sep 10 06:41:01 really? Sep 10 06:41:05 no you didn't - you just said Friday evening - that's three days ago Sep 10 06:41:50 but that means that i have to start from scratch with the compiling - 8 hours? Sep 10 06:41:54 three times in a row means three times in a row, one immediately after the other, where the last two will take less than 5 minutes each usually Sep 10 06:42:34 RalphEichelberge: Yes, sometimes there's old cruft in your build/tmp/ folder, and cleaning that out and starting from scratch is the only way Sep 10 06:42:40 At least for me Sep 10 06:43:11 but usually it's not. the "three times in a row" requirement on the MokoMakefile page does not require rebuilding from scratch three times Sep 10 06:43:21 hi there. Sep 10 06:43:26 ok - so you think, if I am going to do that make clean and update it will compile through this time? Sep 10 06:43:55 anyone here is using google perftools? Sep 10 06:43:56 who said anything about make clean? Sep 10 06:44:03 No-one able to answer that? Sep 10 06:44:16 make update image Sep 10 06:44:18 ? Sep 10 06:44:27 rwhitby: thx...trying these steps Sep 10 06:44:37 make update image is very much *not* the same as make clean Sep 10 06:44:59 ok i try :) thanks Sep 10 06:45:10 Demonax: If you have hardware you can play around with the mixer settings - that's as much of an answer you'll probably get at the moment Sep 10 06:45:37 btw: paulproteus myself lessig and some others had dinner with RMS last week and he talked at length about openmoko and wanting to use it :) Sep 10 06:45:43 RalphEichelberge: Also, if you get stuck on task 1360, uicmoc4-native, look at bug 747 in the openmoko bugzilla Sep 10 06:46:12 aha - ok i will Sep 10 06:46:16 rejon: those steps work for me on Bv3 and Bv4- Sep 10 06:46:26 rejon: RMS? Sep 10 06:46:31 stallman Sep 10 06:46:37 Oh Sep 10 06:46:56 is that the boing bug - huh? :) Sep 10 06:47:22 RalphEichelberge: Yes. Bug 666 is another favourite that's still open. ;) Sep 10 06:48:25 uuuh! Sep 10 06:48:31 love this Sep 10 06:48:52 666 the bug of the beast Sep 10 06:48:59 just intrest - where u from, CM? Sep 10 06:49:07 RalphEichelberge: Sweden Sep 10 06:49:14 i c Sep 10 06:49:21 Sweden is the shit! Sep 10 06:49:33 well linux ... Sep 10 06:49:41 nokia Sep 10 06:49:53 skype Sep 10 06:50:01 all from sweden Sep 10 06:50:07 Well, skype is swedish, the others are Finnish :P Sep 10 06:50:07 no Sep 10 06:50:19 oh sorry Sep 10 06:50:32 But Sweden did own Finland for a couple of years... :D Sep 10 06:50:34 its just north for me - i know+ Sep 10 06:50:35 hmm, skype swedish? Sep 10 06:50:36 yep :) Sep 10 06:50:47 god damn ssglagg! Sep 10 06:50:50 CM: so did russia :P Sep 10 06:50:52 whats his name again? Sep 10 06:51:02 s/ssg/ssh Sep 10 06:51:06 skype swedish/danish Sep 10 06:51:11 and I don't see anyone mixing up those two :) Sep 10 06:51:11 fatal-: Yes, Nicklas Zennström, the Kazaa and Joost guy was behind it Sep 10 06:51:21 zenström? Sep 10 06:51:26 ok Sep 10 06:51:35 mhm Sep 10 06:51:36 and Janus Friis Sep 10 06:51:38 i read about Sep 10 06:52:08 El_Salvador: Yes Sep 10 06:52:09 is dialer supposed to keep working if you shut down pulseaudio ? Sep 10 06:52:10 abraxa_: I suppose it'd somehow be possible to attach a USB mic that would make it possible if the internal one can't do it, right? Sep 10 06:52:25 isnt there a big mainframe that could make the builds for moko? Sep 10 06:52:36 El_Salvador: But he's from Denmark ;) Sep 10 06:52:38 Demonax: I didn't know USB host mode worked now Sep 10 06:53:00 CM: I know Sep 10 06:53:04 skype swedish/danish Sep 10 06:53:21 Ah, sorry, missed that line Sep 10 06:53:31 polz: no, I just mean in the future, I don't have a Neo1973 yet, saving for the final release in December... Hopefully I can get one around January. Sep 10 06:53:36 no problem :) Sep 10 06:53:36 did anyone allready used the neo as a phone? Sep 10 06:53:52 * CM encourages people to update http://wiki.openmoko.org/wiki/Community_nominees_for_GTA02_P0 Sep 10 06:55:10 * rwhitby wonders whether people assume he is one of "the core developers on OpenMoko and OE" Sep 10 06:57:53 rwhitby: Well, I think you're an obvious choice, even if you're not officially hired. Besides, it shouldn't only be me nominating people and I assumed someone would add you. :) Sep 10 06:58:24 hey guyes - what is the difference between make image and make openmoko-devel-image ? Sep 10 06:58:30 rwhitby: you're behind mokomake correct? Sep 10 06:58:40 mhm Sep 10 06:58:54 Demonax: Among many things, yes Sep 10 06:59:22 what is the savest way to get a working image? Sep 10 06:59:41 RalphEichelberge: Download it from the build host? Sep 10 06:59:49 RalphEichelberge: For qemu or? Sep 10 06:59:57 no for the neo Sep 10 07:00:30 Demonax: Yeah, gta02 will have powered host-mode USB so if the mic is too weak you can just hook up an external one, given it draws <100mA Sep 10 07:00:32 Same thing I guess, get a tested image from a build host Sep 10 07:00:43 RalphEichelberge: those two targets do the same thing. image is a convenient shortcut Sep 10 07:00:44 good morning Sep 10 07:00:55 i see Sep 10 07:01:08 DerCorny: yes, I wrote and maintain MokoMakefile Sep 10 07:01:14 thx rwhitby Sep 10 07:01:46 i launced the make again Sep 10 07:01:49 RalphEichelberge: "safest" way at the moment is to use the 2007.08 snapshot, and *not* run ipk upgrade Sep 10 07:02:00 where can i find the newes build? Sep 10 07:02:19 most practical way is http://lists.openmoko.org/pipermail/device-owners/2007-September/000449.html Sep 10 07:02:39 you can find the newest build in build/tmp when your "make image" completes ;-) Sep 10 07:03:22 Note that the only things that are tested on buildhost are the images in http://buildhost.openmoko.org/snapshots/2007.08/ Sep 10 07:03:28 yes that my own :) Sep 10 07:03:49 the stuff in http://buildhost.openmoko.org/OM2007.2/ is no more tested than the images you build yourself. Sep 10 07:04:30 yeah, but they are old, right? Sep 10 07:04:47 yep Sep 10 07:05:17 8 hours later, maybe I will have a new build Sep 10 07:05:22 you can have old and "working" or new and "broken" (for various definitions of working and broken) at the moment Sep 10 07:05:35 hey, don't you have a build for me? Sep 10 07:05:47 that i could flash Sep 10 07:06:36 I personally start with http://buildhost.openmoko.org/snapshots/2007.08/ and then ipkg upgrade and do other stuff as documented in http://lists.openmoko.org/pipermail/device-owners/2007-September/000449.html Sep 10 07:07:05 now i understand that ipgk Sep 10 07:07:21 I have a builder that runs continuously at http://ipkg.nslu2-linux.org/feeds/openmoko/images/, but those images are even more untested than http://buildhost.openmoko.org/OM2007.2/ Sep 10 07:07:47 select the level of "bleeding edge" which achieves your goals ... Sep 10 07:15:00 you know, i am not a end-customer ... i just want the newest stuff :) Sep 10 07:15:12 so that builder is awesome Sep 10 07:16:35 RalphEichelberge: Nominate rwhitby for a next gen neo then ;) http://wiki.openmoko.org/wiki/Community_nominees_for_GTA02_P0 Sep 10 07:22:24 well i guess his afforts in doing some support will give him that advantage :) - i wil Sep 10 07:44:12 when i flash the image, i don't have to flash the kernel, because its allready on that image, right? Sep 10 07:45:51 RalphEichelberge: No, you have to flash the kernel too Sep 10 07:46:31 RalphEichelberge: I think that's a bit annoying though, but that's how it's made Sep 10 07:46:43 You have to flash the kernel matching the rootfs Sep 10 07:46:57 ok Sep 10 07:53:57 CM: I would be surprised if rwhitby wasn't already on the list ;-) Sep 10 07:54:13 aloril: Me too Sep 10 07:54:58 But still, if there's 50 GTA02's, there's some left to send out even after the core ppl get theirs :) Sep 10 08:05:15 * XorA watches the pirahna tank Sep 10 08:28:25 (script) wiki RecentChanges: [[Community_nominees_for_GTA02_P0]] [[User:BBlack]] Sep 10 08:41:40 can anyone tell me what's going on vis-a-vis opengl es/acceleration plann(ing/ed) with GTA02? Sep 10 08:41:55 I'm interested on working on this in some capacity Sep 10 08:52:33 notserpe: there has been no public information about the drivers for the video accel on GTA02 (apart from the fact that they will be written by OpenMoko staff under NDA, but will be open source with full source code availabiilty of course, and no binary parts to the driver) Sep 10 08:53:01 if you have significant experience, I suggest you contact Mickey or Harald directly. Sep 10 08:53:07 And that they've hired some known hacker to fix the drivers Sep 10 08:53:13 To be announced soon Sep 10 08:53:38 Hmm - have they announced the OE dev yet? Sep 10 08:54:03 No, I think mickeyl said he'd say it the 19th on some conference or so Sep 10 08:54:28 * rwhitby is guessing raster for the video dev and XorA for the OE dev. No inside information, just idle speculation. Sep 10 08:54:50 Sounds likely to me too :) Sep 10 08:54:55 * XorA sees his name in vain Sep 10 08:55:02 Damn :P Sep 10 08:55:12 ~hail XorA Sep 10 08:55:13 * apt bows down to XorA and chants, "I'M NOT WORTHY!!" Sep 10 08:56:16 ~lart evolution Sep 10 08:56:16 * apt --purges evolution Sep 10 08:56:28 hmm, good advice from apt there Sep 10 08:56:38 indeed Sep 10 08:58:01 rwhitby: hehe, I tought raster too when I read that mail about video dev ;-) Sep 10 09:02:36 OT, but coolest rc thing of the day: http://www.festo.com/INetDomino/coorp_sites/de/22ff224c0cba8e40c12572d60033d076.htm Sep 10 09:05:00 man, thats looks really odd Sep 10 09:06:55 CM: not cool Sep 10 09:07:01 a bit of wind and the thing is a goner. Sep 10 09:07:14 guess why they were flying it indors in the demo. Sep 10 09:07:31 ixs: True, and it's big. But I'd like a smaller one like that and fly it around at the office ;) Sep 10 09:07:37 * CM is bored at work Sep 10 09:07:58 CM: let's fit it with a little bomb-chute and sell it to helpdesks worldwide. Sep 10 09:08:05 Hehe Sep 10 09:10:50 rwhitby: thanks for the heads up Sep 10 09:15:27 (script) wiki RecentChanges: [[OpenMoko_Local_Groups:_Montreal]] Sep 10 09:15:52 seems like you need giant room for that Sep 10 09:27:26 (script) openmoko-neo1973: Harald Welte Re: SIM Card/Bugzilla 666 status? Sep 10 09:31:46 (script) openmoko-community: Harald Welte Re: ATI to provide specs (was: Re: SMedia 3362) Sep 10 09:36:22 LaF0rge: didn't know you were able to release documentation too for SMedia 3362: seems like best possible result Sep 10 09:36:30 LaF0rge: great work Sep 10 09:36:38 Indeed. Sep 10 09:36:46 * buz nods Sep 10 09:36:53 * rwhitby concurs Sep 10 09:36:53 Well - almost. Sep 10 09:37:22 * XorA pokes SpeedEvil in eyeball :-D Sep 10 09:38:03 XorA: you got one of those finger tip pointers then ;.-) Sep 10 09:38:22 buz: you dont diss dah man :-) Sep 10 09:38:44 i meant for poking Sep 10 09:38:52 The most ideal case would be s3 saying 'we were only kidding, here's a GPL driver :)' Sep 10 09:39:00 all: thanks. Sep 10 09:39:08 But this is awesome Sep 10 09:39:25 And needless to say very promising for future chips. Sep 10 09:39:25 * CM waits for the mail to reach the archives Sep 10 09:39:28 buz: anyway, Id probably use that lethal neo stylus :-D Sep 10 09:39:40 maybe turn it into a class 3 laser eh Sep 10 09:40:21 Is there a link to a datasheet now, or sometime soon. Sep 10 09:40:32 To read the mail, I'd have to get out of bed... Sep 10 09:41:05 now if gta02 had a cam Sep 10 09:41:06 SCNR Sep 10 09:41:13 and a v8, of course Sep 10 09:41:43 No. It should have a fuel cell, that can run from v8. Sep 10 09:42:14 i dont think you can run anything from v8 Sep 10 09:42:16 gruesome stuff Sep 10 09:42:44 But it's healthy! Sep 10 09:43:26 maybe Sep 10 09:43:47 they used to say that about OJ as well whereas in reality, it has way too much sugar in it Sep 10 09:50:12 what's the latest on the next shipment of gta01? Sep 10 09:50:50 3rd week of september Sep 10 09:51:11 tuukkah: Latest I've read is this mail: http://lists.openmoko.org/pipermail/community/2007-September/010153.html Sep 10 09:51:27 is now 2nd or 3rd week? Sep 10 09:51:39 2:nd Sep 10 09:51:40 2nd i'd say Sep 10 09:55:39 ok, so next week Sep 10 09:57:09 my university ordered one advanced kit to evaluate the possibility of using it in projects or courses Sep 10 10:00:03 :) Sep 10 10:00:54 It's quite flexible really - everything from using JTAG, to asm programming. Sep 10 10:06:16 .win 3 Sep 10 10:06:18 Does the mail mean that there is/will no public doc (be) available for the wifi? Sep 10 10:08:36 Maledikt: the doc is written in C :-) Sep 10 10:08:50 grmpf ;) Sep 10 10:09:39 Maledikt: perhaps you can talk to Atheros directly and convince them to release the doc ;-) Sep 10 10:09:58 lol, atheros sucks Sep 10 10:11:15 * rwhitby can never bring himself to take anyone who says " sucks" seriously ... Sep 10 10:12:04 s/sucks/are not open source friendly people/ better? ;) Sep 10 10:13:28 interesting. So you're on #openmoko, saying that the single most open-source friendly wifi solution that the OM core team could find is not open source friendly, as if some other company is more open source friendly ... Sep 10 10:13:59 Guess you're going to have an internal conflict to resolve if you buy a GTA02 ... Sep 10 10:14:08 no, didn't mean that. And don't want to say that. I think it was the best choice they could make Sep 10 10:14:26 but there is always room for improvment Sep 10 10:14:44 ah, so you're saying that all wifi manufacturers suck, but Atheros sucks least? Sep 10 10:14:59 when i hear atheros i always have to think of their blob for there laptop and desktop hardware Sep 10 10:15:54 no, afaik there are companys who provide open doc and even hw to os developers. but thats desktop and laptop Sep 10 10:17:11 * Maledikt is searching for a name as example Sep 10 10:19:34 can't find it Sep 10 10:20:36 ubuntu? Sep 10 10:21:00 but i guess you probably meant hardware companies. Sep 10 10:21:06 yes Sep 10 10:21:32 Intel has gone fairly OSS friendly and has wireless chips, no? Sep 10 10:21:32 Currently sold Neo1973 GTA01B_v04 (P1) doesn't have WiFi, GTA02 (P2) which is scheduled for October will have WiFi with free (libre) drivers (Atheros AR6K): http://wiki.openmoko.org/wiki/Neo1973 http://wiki.openmoko.org/wiki/FAQ#Q:_What_is_the_rationale_behind_the_exclusion_of_WiFi.3F Sep 10 10:22:20 ralink! Sep 10 10:22:33 afaik thats a company who provides docs and hw Sep 10 10:22:41 Lynet: the wireless isnt so open afaik Sep 10 10:22:43 binary blobs Sep 10 10:23:01 Lynet: their firmware isn't open, their drivers are Sep 10 11:13:26 hi all Sep 10 11:13:54 who tried OpenMoko on hx4700? Sep 10 11:14:09 I wanted to do the same, but the kernel doesn't boot :( Sep 10 11:15:23 :/ Sep 10 11:15:33 well. Have you a kernel that boots? Sep 10 11:15:40 er... Sep 10 11:15:47 (script) wiki RecentChanges: [[Repositories]] [[Phase_1_Software_Testing]] [[Talk:Community_nominees_for_GTA02_P0]] Sep 10 11:15:57 I tried GPE's kernel and it does the same excepts GPE displays a "tux" icon Sep 10 11:16:07 I get no messages on the serial console Sep 10 11:16:31 (I'm using haRET, a wince bootloader) Sep 10 11:17:04 As far as I know - nobodies tried. I may be wrong. Sep 10 11:17:31 First you need a booting linux kernel on your device. Sep 10 11:17:46 "koen" tried and posted something on his blog, but there's no technical info Sep 10 11:17:48 Then you install the root filesystem of OM to where the kernel has been told to find it. Sep 10 11:18:14 I built Openmoko changing "fic-gta01" to "hx4700" in local.conf Sep 10 11:18:19 all went well. Sep 10 11:18:36 I have a rootfs on an ext2 compactflash, all I need is a working uImage ! Sep 10 11:19:35 haRET boot avoids flashing anything! Sep 10 11:19:51 but... sigh Sep 10 11:21:45 the only "openmoko hx4700" resource is this : http://dominion.kabel.utwente.nl/koen/cms/openmoko-running-on-an-ipaq Sep 10 11:53:08 now thaqt was very un-fun Sep 10 11:53:22 om's build managed to fill up 450gb of disk space... Sep 10 11:53:26 with 1 log file. Sep 10 11:53:27 woot Sep 10 11:54:21 Hehe.. yikes Sep 10 11:55:38 some build got into an endless loop while i was away waiting for it Sep 10 11:55:39 woot Sep 10 11:55:45 deletign that crap took a while Sep 10 11:56:05 interesting Sep 10 11:56:44 actually Sep 10 11:56:49 520gb of log Sep 10 11:56:57 poor disk Sep 10 11:58:31 at least it tested gthe codepath in e when trying to write its config Sep 10 11:58:33 and failing Sep 10 11:58:37 because your disk is full Sep 10 11:58:43 and it managed to not lose the config Sep 10 11:58:51 and nicely pop up a dialog to let you know Sep 10 11:59:49 yo Sep 10 12:06:52 anyone got any clue as to what the state of the current moko tree is like, build-wise? i finished another full build on thursday and am only just getting back to catchup .. Sep 10 12:07:11 i guess i should look at the bugbase, but TBH, i don't wanna use my web browser right now. ;) Sep 10 12:07:13 cat soup! Sep 10 12:10:45 torpor: I finished a build this morning Sep 10 12:28:38 CM cool i've just started an update and a new make .. Sep 10 12:28:52 < -- want to boot my own OM image today, and put some modules in. Sep 10 12:30:13 :-) Sep 10 12:51:08 any word about the next batch of gta01's coming? Sep 10 12:55:19 gamin: Yes, hopefully next week Sep 10 12:56:05 gamin: http://lists.openmoko.org/pipermail/community/2007-September/010153.html Sep 10 13:03:03 CM: thanks, I hadn't seen that mail Sep 10 13:15:28 (script) wiki RecentChanges: [[Talk:Neo1973_and_Windows]] [[Neo1973_and_Windows]] [[Talk:Community_nominees_for_GTA02_P0]] [[Talk:Hear_Me_FIC]] Sep 10 13:53:58 shhhhwing http://www.engadget.com/2007/09/10/3ms-dms800-short-throw-projector-cum-alien-probe/#comments Sep 10 14:12:38 mickeyl: are you around? Sep 10 14:29:32 re Sep 10 14:30:29 * hhf423_w can finally hook up his Neo to his windows system via usb networking Sep 10 14:30:33 yes it is unstable that newest build :) Sep 10 14:30:57 now a dfu-util for windows is what we still need Sep 10 14:31:13 y? Sep 10 14:31:29 there is no need for windows anyway Sep 10 14:32:20 RalphEichelberge: unless you want to sell to 95% of the people in the universe Sep 10 14:33:20 hhf423_w: there was someone who was working on a windows build of dfu-util Sep 10 14:33:38 and I think harald fixed a problem in the kernel which was preventing that from working Sep 10 14:34:11 rwhitby: the neo kernel is now fixed so that usb networking works Sep 10 14:34:41 we probably only need a windows developer who can build a dfu-util for windows Sep 10 14:36:17 ooh - new face on the OM wiki Sep 10 14:36:27 hmm? Sep 10 14:42:07 sorry - i didn't want to be cruel because of that M$ thing Sep 10 14:45:14 woah, "Turned Off iPhone Gets $4800 Bill from AT&T" http://hardware.slashdot.org/hardware/07/09/10/1216224.shtml Sep 10 14:45:38 Off ain't. Sep 10 14:45:56 yeah and taking battery out is not simple thing either Sep 10 14:46:03 hahaha Sep 10 14:46:08 so .. only way to avoid that is to leave it at home Sep 10 14:46:11 good lessons from this Sep 10 14:46:24 or put it inside tinfoil ;-) Sep 10 14:46:48 or microwave oven Sep 10 14:46:49 new product, portable iPhone emi cage Sep 10 14:46:58 Foil baggie. Sep 10 14:47:22 emi cage many more $$$$ than foil baggie ;) Sep 10 14:47:40 or turn off email Sep 10 14:47:41 do we have javame ported? i could get a nice situation to promote the neo if we do =) Sep 10 14:50:13 SpeedEvil: on more serious note we do absolutely need a concept of connection cost - merely being connected is not sufficient - current cost of connectivity will need to be available to apps. Sep 10 14:50:40 Indeed - there are notes of that on the wiki Sep 10 14:52:25 meeeeeeeeeeeeeeh, now the date for the new shipment is 24th sept X_x Sep 10 14:53:58 Fatal: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=260157395019 Sep 10 14:54:37 heh, if it was only I then maybe, sadly there are two coworker in the same order Sep 10 15:00:28 hhf423_w: I just got internet connection sharing to work on windows with the neo Sep 10 15:00:47 hhf423_: u-boot cdc serial black screens my XP-SP2 machines - have you had similar experience. I think this is pre-requisite to Dfu-Util Sep 10 15:03:53 rwhitby: via the registry hack or otherwise? Sep 10 15:04:36 Writchie: sorry, how exactly did you get the Neo to bluescreen your windows? Sep 10 15:05:23 fatal for P1 shipping I presume? Sep 10 15:05:32 hhf423_w: during install of drivers for the cdc serial - not bluescreen - black screen Sep 10 15:06:20 Writchie: so you did boot into uboot and then connected to windows and then got black screen? Sep 10 15:06:21 hhf423_w: I created a new bridge ("USB Network") with just the usb connection to the phone in it Sep 10 15:06:33 then I enable ICS on the wan interface, and told it to use that bridge for the lan Sep 10 15:06:59 it changes the lan address to 192.168.0.1 when you do that, so I then just went back in to the bridge IP settings and changed it back to 192.168.0.200 Sep 10 15:07:04 now I get routing and DNS proxying through the wan interface Sep 10 15:07:25 hhf423_w: yes - I used a modified inf file that called up the standard windows drivers for cdc serial Sep 10 15:07:43 rwhitby: put it into the wiki as an alternative way to mine Sep 10 15:07:54 hhf423_w: which page? Sep 10 15:08:18 Now I can plug the neo into either the linux box or the windows box, and not have to change any settings on the neo. Sep 10 15:08:19 rwhitby: http://wiki.openmoko.org/wiki/Neo1973_and_Windows Sep 10 15:08:56 rwhitby: right, same here, although I needed to add a route in my DSL router Sep 10 15:09:16 hhf423_w: with this method, windows ICS does nat I expect Sep 10 15:09:47 rwhitby: yes, I also would expect that Sep 10 15:10:15 rwhitby: its better than my option because no routes need to be added anywhere Sep 10 15:12:56 thanks hhf423_w :-) Sep 10 15:13:11 hhf423_w: I've added it to the page. It can probably do with some screen shots and better description. Sep 10 15:13:47 kristian-m: :-) Sep 10 15:15:27 * rwhitby wonders whether windows will handle two Neo1973 USB Ethernet/RNDIS connections ... Sep 10 15:15:29 (script) wiki RecentChanges: [[Neo1973_and_Windows]] [[Repositories]] [[User:Hhf423]] [[No_Linux]] [[GSM]] [[Neod]] [[Neo1973_Hardware]] [[User:Kriss]] [[Talk:Community_nominees_for_GTA02_P0]] Sep 10 15:16:22 rwhitby: i see no reason why not - I will test when I get working kernel on both of my units - BTW works fine on Linux Sep 10 15:16:57 night all Sep 10 15:17:02 cya Sep 10 15:17:19 rwhitby: night - any progress on u-boot?? Sep 10 15:19:16 rwhitby: good night Sep 10 15:19:33 Writchie: uboot-openmoko still fails to build for me Sep 10 15:19:42 tx Sep 10 15:38:14 hmmm Sep 10 15:41:12 is this pretty much still accurate or not? Sep 10 15:41:12 http://wiki.openmoko.org/wiki/Building_OpenMoko_from_scratch Sep 10 15:41:57 torpor: think so Sep 10 15:43:13 hmm. Sep 10 15:43:52 i had my mokomakefile setup working perfectly last week, then i rm -rf'ed build/tmp, did what i thought was a successful rebuild "make update && make all" .. and now nothing is working. Sep 10 15:44:18 That happens Sep 10 15:44:24 annoying. Sep 10 15:44:52 c compiler cannot create executables .. wtf was it doing for 12 hours then, matey!! ;) Sep 10 15:44:56 There is hope; it's certainly gotten much more stable recently. Sep 10 15:45:28 Althought the compiler problems were sorted out a week and half ago, I thought Sep 10 15:45:42 so i suppose i should update my makefile and try again .. Sep 10 15:46:25 I tell you, what is annoying: http://scap.linuxtogo.org/files/d39e0c54b8a43650751d0ae19b78f676.png and then http://scap.linuxtogo.org/files/0690c9bd00f72b082edee75db1f5ff51.png Sep 10 15:47:00 at least those are apps that are in development. Sep 10 15:47:20 but for sure as hell, this needs to be bulletproof. and its a shock and a shame that it isn't by now. Sep 10 15:47:41 i think there are other priorities. the guys that should be fixing it are playing with hardware instead. ;) Sep 10 15:47:49 hi - im building openmoko using mokomakefile and its failing on packageicu-native-3.6 in task do_compile - anyone else having this problem? Sep 10 15:48:28 nope Sep 10 15:48:53 whats the log output say? Sep 10 15:49:10 i compiled it successfully last night Sep 10 15:50:25 okay now its working again. wtf?! Sep 10 15:50:31 * torpor goes for a beer. Sep 10 15:51:59 log output is in http://pastebin.com/m8ad6605 ... Sep 10 15:52:23 hi all Sep 10 15:53:32 wonder what this: /usr/bin/../lib/libstdc++.so.6: undefined reference to `_Unwind_GetIPInfo@GCC_4.2.0' means? Sep 10 15:53:33 mickeyl: I think cesarb said that there now is a patch for the Dialer crazieness as described here: http://wiki.openmoko.org/wiki/Phase_1_Software_Testing#GSM_10 when will there be a new build on builthost? Sep 10 15:54:18 jpcass: bitbake icu-native-3.6 -c mrproper Sep 10 15:54:20 jpcass, it has problems while linking to your host's libstdc++ Sep 10 15:56:10 torpor: mrproper worked, am now trying again but have not changed anything re: my libstdc++ ... Sep 10 15:57:39 so the question for today is: Is there any news for the shipment dates (today is 10!)? Sep 10 15:58:21 task do_compile started ... wait... Sep 10 15:58:40 bedboi: couple of weeks maybe Sep 10 15:58:50 And yes, the banner still says the 10th Sep 10 15:59:16 same error :-( Sep 10 16:00:47 ive got two libstdc++: 5.0.7 and 6.0.8, i guess its linking to 6 - do others have something different? Sep 10 16:01:14 SpeedEvil: so i definitely have to cancel the order Sep 10 16:01:22 i'm quite pissed off Sep 10 16:01:26 :/ Sep 10 16:01:30 bedboi: what's wrong? :( Sep 10 16:01:38 a couple of weeks man Sep 10 16:01:43 it's not a couple of days Sep 10 16:01:47 it's not serious Sep 10 16:01:50 That's as far as I'm aware, I haven't got sources for that. Sep 10 16:02:01 short time in politics, long time in opensource... Sep 10 16:02:01 i'll get a motorola Sep 10 16:02:12 It's on teh mailing list. Sep 10 16:02:29 jpcass: it's not about politics. Sep 10 16:02:43 hhf423_w: try http://obri.sygroup.ch/deploy/ipk/armv4t/openmoko-dialer2_0.1.0+svnr2946-r5_armv4t.ipk Sep 10 16:02:44 i'm going to go back in italy at the end of the month Sep 10 16:02:59 jpcass: i'm quite sure that the package won't arrive till the end of the month Sep 10 16:03:10 plus, i don't have a mobile phone right now. Sep 10 16:03:36 mine arrived 2 days after i left london for a us trip. i could have picked the damn thing up personally! Sep 10 16:03:53 bedboi: ps: not having a mobile phone is very much like having a neo ;-) Sep 10 16:04:13 jpcass: btw at least you have something Sep 10 16:04:31 you can make it work Sep 10 16:04:41 btw that's not the main point Sep 10 16:04:56 Mine hangs when trying to start X with an image built today (known error at least) Sep 10 16:05:30 Error: "Not connected to network. Current Status: Home network registered" ? Sep 10 16:05:32 Obri: how can I install that easily? Sep 10 16:06:16 ah, ok Sep 10 16:06:17 hhf423_w: download the file to the neo and install it with ipkg install /path/to/openmoko-dialer2_0.1.0+svnr2946-r5_armv4t.ipk Sep 10 16:07:41 ipkg update && upgrade from today, w/ gsmd restart hack Sep 10 16:08:11 enodev: see the msg fro Obri , try the new dialer Sep 10 16:08:39 could you report the url? i joined too late Sep 10 16:08:53 http://obri.sygroup.ch/deploy/ipk/armv4t/openmoko-dialer2_0.1.0+svnr2946-r5_armv4t.ipk Sep 10 16:09:56 Obri: thanks, do I have to use any --force statements? It says: Multiple packages (openmoko-dialer2 and openmoko-dialer2) providing same name marked HOLD or PREFER. Using latest. Sep 10 16:11:09 hrm even after ipkg -force_depends remove openmoko-dialer2, i get the same error as hhf423_w Sep 10 16:11:25 hhf423_w: hm, good question Sep 10 16:13:17 ah, well, lets see what it does after a reboot Sep 10 16:14:01 cd /etc/ipkg && wget http://obri.sygroup.ch/obri.conf && ipkg update && ipkg install openmoko-dialer2 Sep 10 16:14:04 anyone got ideas why a link to my host libstdc++ would fail in package icu-native??? Sep 10 16:14:19 with force_overwrite it works Sep 10 16:15:01 i've sent a mail to the orders guys Sep 10 16:15:08 let's see what happens Sep 10 16:15:56 oooh, I am allowed to do calls again! Sep 10 16:16:15 Obri: it worked after a reboot even without force statements Sep 10 16:16:41 I was just about to open a bug, no need to do that now :-) Sep 10 16:18:07 what the hell is icu-native anyway? and why is it getting in the way of me making an image for my neo! Sep 10 16:20:39 jpcass: "Who is General Failure and why is he reading my harddisk?" :-) Sep 10 16:21:35 "Who is Colonel Panic?" Sep 10 16:21:45 i wish General Failure would kick the ass of this obnoxious package! Sep 10 16:21:54 hhf423_w: the colonel reports to the general Sep 10 16:22:14 why are there no instant messengers in the repos? Sep 10 16:22:26 Whose side are the zombie processes on, though? Sep 10 16:23:10 hahaha Sep 10 16:23:27 dcordes: good question. a pidgin would be nice Sep 10 16:23:36 :) Sep 10 16:24:34 will someone who CAN compile icu-native please let me know what version libstdc++ they have? Sep 10 16:29:40 jpcass: I struggeled with that a while ago, never got around it until I restarted from scratch.. :-/ Sep 10 16:30:46 CM: what do you mean 'scratch' ? i've done a bitbake mrproper which is about as scratchy as you can get! Sep 10 16:31:06 CM: did you reinstall your host OS? Sep 10 16:31:11 Hehe, no Sep 10 16:31:14 jpcass: wget ...Makefile is what i do for 'from scratch'... Sep 10 16:31:38 I built it with the make file like emdete said Sep 10 16:31:46 New dir, wget makefile, ... Sep 10 16:32:15 hmm.. thats what Im doing...its all 'fresh'... Sep 10 16:32:18 I suck at using bitbake, so I'm sure there's some better way Sep 10 16:32:35 Is it nothing in bugzilla about your problem? Sep 10 16:32:59 CM: havent checked - will look now, but no one here seems to have it... Sep 10 16:33:56 "Zarro Boogs found." - dont think its a bug, some kind of configuration mismatch I think ... Sep 10 16:34:23 CM: what version libstdc++ do you have in /usr/lib ? Sep 10 16:35:23 libstdc++-v3-3.3.6 Sep 10 16:35:53 Ehh.. libstdc++.so.5.0.7 Sep 10 16:36:10 gentoo 64-bit Sep 10 16:36:27 CM: thanks - i have two, that and 6.0.8 - i have removed 6.0.8 temporarily see if it works... Sep 10 16:39:12 do_compile started ... wait ... Sep 10 16:44:09 hi Sep 10 16:46:26 nope still failing - its looking for libstdc++.so.6 ! Sep 10 16:52:03 * hhf423_w hums a song Sep 10 17:15:43 hhf423_w: Thanks :D (Updating that page every now and then keeps it on the Recent Changes page.. :) Sep 10 17:15:50 (script) wiki RecentChanges: [[Community_nominees_for_GTA02_P0]] [[User:ChrisLord]] [[PulseAudio]] [[Power_management_menu]] [[Phase_1_Software_Testing]] [[Neod]] [[Functions_menu]] [[SMSTool]] [[SettingsGUI]] [[Neo1973_power_button]] [[Neo1973_and_Windows]] Sep 10 17:19:38 okay, now the damned thing keeps shutting off, even when fully booted up Sep 10 17:19:49 and with the usb plugged in Sep 10 17:20:09 vallor: Mine doesn't even start any more, hangs on loading the GUI Sep 10 17:20:21 * CM waits a day or two for things to settle Sep 10 17:21:07 hhf423_w: (sorry for late reply) yes Sep 10 17:21:17 CM: there's a change you can make in your boot environment that might fix that -- or at least, fixed it for me Sep 10 17:21:32 hello mokoers Sep 10 17:21:45 CM: do you know where the shutoff-timer config would be located? Sep 10 17:22:24 Nope, but what if you disable the powermanager? Sep 10 17:22:48 Iirc there's some issue with that, try turning it off Sep 10 17:24:37 That is by holding in the power button while the neo is on and disable it Sep 10 17:25:22 ah, will try Sep 10 17:25:44 woot! Sep 10 17:25:53 it was set to "full PM" Sep 10 17:26:04 changed to "no PM" Sep 10 17:26:10 Ah, that should help Sep 10 17:27:22 oh, btw -- whoever was porting emacs to the thing -- there's an editor that can act like a "basic emacs", it's jmacs -- which is "joe" with special keybindings Sep 10 17:27:47 joe is small and wiry, so it makes sense for an embedded app Sep 10 17:29:34 You really need a bt keyboard to use emacs on it Sep 10 17:32:21 vallor: joe is now in the feeds. Sep 10 17:32:50 is it possible to configure matchbox input manager so the xkbd does not pop in everytime i change the term? Sep 10 17:33:07 mwester: !!!! thank you :) Sep 10 17:33:25 :) I didn't do it; I just noticed that rwhitby did it last night. Sep 10 17:33:30 I've been a joe fan since I first started using Linux -- way back in '92 (boy I feel old) Sep 10 17:33:39 oh, hmm Sep 10 17:33:54 do you know the package name? Sep 10 17:34:42 If the autobuilder got to it yet, it should just be "ipkg install joe" Sep 10 17:34:57 oh, tried that a few minutes ago, no joy Sep 10 17:35:10 no joy for joe, but I'll be patient :) Sep 10 17:35:42 CM: np Sep 10 17:36:08 Fatal outch Sep 10 17:36:18 hi Ghiottone Sep 10 17:37:54 hey hhf423_w Sep 10 17:38:31 no, joe is not there Sep 10 17:38:47 but vi is enough! Sep 10 17:39:13 "vi -- beginners pronounce it 'Why I?'" Sep 10 17:39:21 let's not get into religious editor debates -- we use what we like to use, and that's okay, because it is the Unix way! Sep 10 17:39:44 vallor: if it does not work, reboot is the unix way... Sep 10 17:40:18 actually, reboot is the windows way Sep 10 17:40:33 vallor: then you have not been around long enough Sep 10 17:40:34 Linux systems can generally be "fixed" without rebooting Sep 10 17:40:57 or I'm better at good uptimes ;P Sep 10 17:41:00 * hhf423_w can still remember the says when SUNs were booting fast. Sep 10 17:41:42 or, as some people said, the only thing a SUN does good, is, it boots fast Sep 10 17:41:47 btw, what's funny is I've got another embedded Linux system that I just started working on -- a meraki Sep 10 17:41:57 if you want to see a SUN at its best, reboot it! Sep 10 17:42:15 but that was ages ago. now SUNs do not even boot fast anymore Sep 10 17:42:16 well, that's sun Sep 10 17:42:38 * hhf423_w points out that the Neo also does not boot fast ;-) Sep 10 17:43:02 alas, you are correct Sep 10 17:43:54 mmm I booted the neo. the panel wasn't showing. probably crashed Sep 10 17:44:08 then I ipkg updated && upgraded and rebooted Sep 10 17:44:19 but it is still not showing !?!? Sep 10 17:44:24 any known issue? Sep 10 17:44:44 Ghiottone: the top menu panel or what? Sep 10 17:45:13 yesterday, some kind folks helped me with removing the "console=" line from the boot parameters Sep 10 17:45:15 top right panel. that which contains the applets Sep 10 17:45:29 after an X server restart it works Sep 10 17:46:31 Ghiottone: I have my status documented in phase 1 software testing. a change to uboot parameters and the usual in /etc/matchbox/session Sep 10 17:46:51 btw, how come the netmask on the Neo is set to 255.255.255.0 by default ? Wouldn't 255.255.255.192 make more sense ? Sep 10 17:47:19 hhf423_w, the gsmd restart thing? I thought that that issue was closed as from recent commits Sep 10 17:47:22 polz: because 192.168.0.x is originally class c? Sep 10 17:47:37 Ghiottone: not really if you ask me :-) Sep 10 17:47:38 recent = today's... maybe buildhost has not compiled it yet Sep 10 17:48:01 Ghiottone: right, the most resent image is from yesterday or so Sep 10 17:48:17 autsch: http://www.phoronix.com/scan.php?page=news_item&px=NjA0NQ Sep 10 17:48:28 UPS does not deliver december 25 in Germany Sep 10 17:49:04 in any case, december 25 I am already in berlin Sep 10 17:50:00 Ghiottone: oh, new images just showed up on buildhost Sep 10 17:50:21 Just a little question. I found nothing about DECT in your Wiki? Isn't it planned? Sep 10 17:50:34 hhf423_w, that means that it has just finished rebuilding all the packages Sep 10 17:51:02 chris^^: I think DECT is not planned, no Sep 10 17:51:31 chris^^: at home, you will probably be able to use WLAN for dialing out etc Sep 10 17:51:40 chris^^: using VoIP Sep 10 17:51:48 well, but I'm often at events, friends, parents and so on Sep 10 17:51:55 somewhere I have DECT but no wireless Sep 10 17:53:45 http://wiki.openmoko.org/wiki/Wish_List_-_Hardware#DECT.2FGAP Sep 10 17:54:29 anyone got a link for the user manual for the S3C2442 ? Sep 10 18:02:33 hhf423_w: yes, just a wish, not a plan to implement ;) Sep 10 18:04:58 chris^^: right. I deem it unlikely Sep 10 18:05:30 chris^^: also, UMTS mico cells may come fast and will replace DECT Sep 10 18:05:43 hopefully Sep 10 18:05:46 :) Sep 10 18:06:33 chris^^: and probably one should talk to e.g. people like eventphone and see if VoIP phoce can not be integrated, so that people could use DECT or WLAN Sep 10 18:06:55 that can be done Sep 10 18:07:02 but it isn't only eventphone Sep 10 18:07:15 DECT can't be done. Sep 10 18:07:25 As you can't talk DECT without a DECT radio. Sep 10 18:07:31 which the thing does not have. Sep 10 18:11:42 is "neod" the X server for the beasty -- and if so, can I send it a HUP to restart X? Sep 10 18:11:54 no. Sep 10 18:11:58 Xfvdev Sep 10 18:12:03 Xfbdev Sep 10 18:12:11 ah, there I see that Sep 10 18:12:58 dang, sending it a hup didn't re-exec it, it just shut down Sep 10 18:14:28 Xserver I think to restart Sep 10 18:15:47 just sussed that out by looking in /etc/X11 :) :) Sep 10 18:16:28 how can I write the changed environment in uboot Sep 10 18:16:37 saveenv Sep 10 18:17:03 10x, I'm obviously an analphabet Sep 10 18:17:31 vallor: to restart x do: Sep 10 18:18:12 do /etc/init.d/xserver-nodm restart Sep 10 18:18:25 oh okay Sep 10 18:19:44 I'm good :) Sep 10 18:19:59 I even have my applet panel back :) Sep 10 18:20:54 looks like something is ptracing all the X apps... Sep 10 18:23:50 I'm wondering if gsmd needs a re-write :\ Sep 10 18:24:14 vallor: don't wonder Sep 10 18:24:44 vallor: gsmd needs a rewrite, everyone knows that, even the OpenMoko team. Sep 10 18:25:12 but knowing something and acting on the knowledge are two things Sep 10 18:26:32 I was asking in terms of me getting off my lazy a** and doing something, like giving gsmd a once-over Sep 10 18:27:27 vallor: lots of beer await the one who fixes gsmd :-) Sep 10 18:28:09 hehe, now dialer says "not connected to network. Current status = Home network registerd Sep 10 18:28:14 " Sep 10 18:31:46 http://buildhost.openmoko.org/snapshots/2007.08/ <--- place to get latest snapshot? Sep 10 18:32:20 there are also daily snapshots Sep 10 18:32:57 ah "MokoMakefile" is what I need :) :) Sep 10 18:43:52 the new frontpage of the wiki is stunning Sep 10 18:47:06 vallor: see http://wiki.openmoko.org/wiki/Phase_1_Software_Testing#GSM_10 Sep 10 18:47:25 okay Sep 10 18:48:43 vallor: the latest semiofficial stuff is always here: http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/images/fic-gta01/?C=M;O=D Sep 10 18:49:00 for other repositories see: http://wiki.openmoko.org/wiki/Repositories Sep 10 18:50:42 thank you! I got the new dialer and installed... Sep 10 18:52:22 the dialer dialed! Sep 10 18:53:07 meanwhile "make setup" is getting http://www.openembedded.org/snapshots/OE-this-is-for-mtn-0.35.mtn.bz2 Sep 10 18:55:56 vallor: Did anyone answer your call? Sep 10 18:58:30 * SpeedEvil discovers a TV producer that needs to die. Sep 10 18:58:34 UK - channel 4 - a bit on the first mobile phones. Sep 10 18:58:41 (which they placed 25 years ago) Sep 10 18:58:45 ozarka: oh, I just called my treo then hung up w/out answering Sep 10 18:58:51 with text speak as subtitles. Sep 10 18:59:43 Sad. First call with the Neo and it didn't even get to talk to anyone. Sep 10 19:00:13 SpeedEvil: where have you the information for the gta02 release from? http://wiki.openmoko.org/wiki/Neo1973 Sep 10 19:01:19 ozarka: aw! I made calls last night, using libgsmd-tool Sep 10 19:01:44 will test dialer... Sep 10 19:01:54 vallor: Don't try. It's not the same. The moment is gone. :-( Sep 10 19:02:11 PBeck: mickey l here in channel Sep 10 19:02:27 and last time timelines for GTA01 Sep 10 19:02:37 hmm, it isn't dialing now... Sep 10 19:02:42 it was slightly fudged. Sep 10 19:02:58 gsmd died :\ Sep 10 19:03:06 but I'd like to see reasons why any stage should be much faster. Sep 10 19:03:51 SpeedEvil: hm really? where can i find the timelines? Sep 10 19:04:20 Sigh. Sep 10 19:04:57 Unfortunately, due to the lack of annoucements at the time, and the drips of information that happened last time, I can't cite a nice source. Sep 10 19:05:26 IRC and mailing list is about as good as it gets. Sep 10 19:06:20 well, okay -- dialer worked the first time, but now isn't working -- seems to do everything ui-wise, but doesn't actually go off-hook and dial Sep 10 19:06:40 SpeedEvil: hum a german magazine has gripped the informations about the timeline Sep 10 19:07:23 Sigh. Sep 10 19:07:43 PBeck: which magazine Sep 10 19:07:45 I am in the process of logging in to comment on that Sep 10 19:07:54 hhf423_w: pro-linux.de Sep 10 19:08:10 vallor: did you do the uboot variables change? Sep 10 19:08:18 I'd hoped that it was clear enough that it was an estimated timeline by the 'this is an estimated timeline' before it. Sep 10 19:08:25 SpeedEvil: no person reads the inofficial part ;) Sep 10 19:08:34 hhf423_w: if you mean removing the "console=blahblah" line -- yes indeed, change made last night :) Sep 10 19:08:35 PBeck: ah, yes, I saw that Sep 10 19:08:38 And the package service. Sep 10 19:08:55 vallor: hmm, dialer does not die after a call for me Sep 10 19:09:19 oh, dialer isn't dying, it just doesn't seem to do anything but display the ui Sep 10 19:09:24 I'll try calling in Sep 10 19:09:27 SpeedEvil: thats sadly Sep 10 19:09:47 vallor: well, what I meant. I can make multiple calls in a row Sep 10 19:10:22 hhf423_w: I wonder if the fact that I hung up on the first call w/out letting it answer might have cornfused it Sep 10 19:11:03 vallor: well, it is of course easy to confuse dialer and gsmd, but hanging up before someone picks up also works for me Sep 10 19:11:11 oh, heck Sep 10 19:11:17 SpeedEvil: i hope we get a few official informations in the near feature :) Sep 10 19:16:32 (script) wiki RecentChanges: [[FreeBSD]] [[Talk:MacOS_X]] [[GTA02_Openness]] [[GPS]] [[Current_events]] [[Press_Coverage]] Sep 10 19:18:39 SpeedEvil: it's also on lwn.net Sep 10 19:31:36 _[/Desktop/Openmoko/moko]_(scott@eva)_ Sep 10 19:31:36 $ make openmoko-devel-image Sep 10 19:31:39 here it goes! Sep 10 19:33:02 "NOTE: Psyco JIT Compiler (http://psyco.sf.net) not available. Install it to increase performance." <---- is it worth bothering with to install this PJC? Sep 10 19:34:11 vallor yep. psycho really speeds up the buildsys Sep 10 19:34:16 vallor, it's a lot faster with it Sep 10 19:34:23 okay, will get it Sep 10 19:34:36 Unfortunately it's not available for x86_64 Sep 10 19:35:26 alphaone hehe... yep.. too sad it works well without, is just slow.. esle there woul be enough 'leidesdruck' to port it Sep 10 19:35:29 well, I could build Psyco Sep 10 19:35:32 +n Sep 10 19:35:50 roh: Right :-) Sep 10 19:38:22 c/codegen.h:16:3: error: #error "Sorry, non-32-bit platforms are not supported at all." Sep 10 19:38:22 c/codegen.h:17:3: error: #error "You may try with a Python compiled in 32-bit " blah blah .... ugh Sep 10 19:38:54 psyco? Sep 10 19:39:01 just ditch it, not required Sep 10 19:39:06 vallor, it's a jit compiler for python Sep 10 19:39:14 but doesn't it speed things up? Sep 10 19:39:53 (meanwhile, I just restarted the make openmoko-devel-image ) Sep 10 19:40:06 it make a difference mostly for parsing the .bb files Sep 10 19:40:24 the rest of bitbake it not cpu-bound Sep 10 19:40:30 Shoragan: oh -- it's up to 55% on that already Sep 10 19:40:54 anyone tried SettingsGUI recently? Sep 10 19:41:33 this is a core 2 quad 2.4GHz, clocked at 3.05GHz, with DDR-3 1333 memory, and the FSB set to 1333 Sep 10 19:41:45 so I'm curious how fast it does the make Sep 10 19:42:25 grml. sh != bash Sep 10 19:43:39 vallor, did you remember to configure parallel compiling? Sep 10 19:43:50 tuukkah: I did indeed :) Sep 10 19:48:43 oh this is weird -- SELinux is complaining about some of the things this make process is doing... Sep 10 19:48:54 (I have it set to permissive, so no problem -- just weird it would pop up) Sep 10 19:49:10 PBeck: seeming unlikely from what mikeyl and ro h were saying Sep 10 19:49:35 The issue seems to be that they think that releasing no date is better than releasing dates which may slip. Sep 10 19:49:44 Which I fundamentally disagree with. Sep 10 19:49:44 vallor: selinux is evil. Turn it off... Sep 10 19:50:17 mwester: I used to think that, but it's handy in permissive mode with my distro Sep 10 19:50:29 let's me know when "unusual" things happen :) Sep 10 19:50:41 SpeedEvil: releasing dates which slip is not good PR, but no informationen for a open phone is not a good idea i think Sep 10 19:50:56 ,too Sep 10 19:51:00 SpeedEvil: I agree with you; I've already had a conversation with certain core team folks that have convinced me that while they have great technical ability, they need a marketing genius on staff. Sep 10 19:51:18 IMO, dates with additional caveats are fine, as long as they are very public. Sep 10 19:51:27 And the caveats are public. Sep 10 19:51:41 No, we can't cope with what third party publications will do. Sep 10 19:52:28 mwester i think the big point is that its not yet donw. thus we do no marketing because even if there would be real endusers, we would have no ressources at all supporting them Sep 10 19:52:41 SpeedEvil: i think so too. But the press has a different view Sep 10 19:52:59 Developers are users of a form too. Sep 10 19:53:19 Especially ones that are considering purchasing hardware. Sep 10 19:53:23 mwester currently we only can support developers, and even that only barely. the support is basically... eh.. yeah.. youre standing it it Sep 10 19:53:24 Developers *ARE* the users. Sep 10 19:53:44 mwester no. developers do not buy because of the color. they buy because of features. Sep 10 19:54:07 roh, that's simplifying a bit too much i think Sep 10 19:54:13 Developers may also get annoyed and wander away from the project if not fed any information at all. Sep 10 19:54:48 And yes, there is of course information on some things coming out. Sep 10 19:55:04 I fear we must simply agree to disagree, roh -- nobody wants this to succeed more than I, and I grow increasingly worried about the perception that the P1 owners are NOT customers... Sep 10 19:55:34 both developers and non-technical people do a lot of decisions based on the image they get of the company and the product Sep 10 19:56:12 I don't make my comments to discourage people, or to upset the core team -- rather I comment because I feel so strongly that these issues need to be discussed and addressed in order for this to be a success. Sep 10 19:56:30 are there any known issues with compiling on Gentoo with the openmoko-terminal2 package not finding 'vte' as a requirement? Sep 10 19:56:32 tuukkah i think not. its as ugly as that. we get loads of requests from people just searching for their new phone which sadly not have any clue about what were doing here. even kids who find the case nice and funny. but we don't have anybody who could possibly take care of customer support at that point. were just not there yet. besides that, customer support for a unfinished product ist messy and not helping as long as the promised features are not coded yet. Sep 10 19:57:13 i have vte v0.16.6-r1 installed via 'emerge' but compiling the openmoko-devel-image halts on openmoko-terminal2 saying 'vte' wasn't found Sep 10 19:57:36 mwester the p1 owners are customers. but not 'users' in a common sense. they are developers and thus have a total different recognignition of technology. Sep 10 19:57:45 P1 is like a development board Sep 10 19:57:48 iandouglas736, try make rebuild-package-vte Sep 10 19:57:57 no user would ever buy (or should) a develboard. Sep 10 19:57:58 roh, i'm not saying we should already advertise for non-technical users. i'm saying the openmoko brand can be hurt already before advertisement Sep 10 19:57:59 There is a difference between not providing customer support, and not spending 5% of the time you answer questions making a public FAQ. Sep 10 19:58:09 Ghiottone: thanks, trying now Sep 10 19:58:12 tuukkah yes. Sep 10 19:58:20 Even a brief explanation of why you can't release certain information. Sep 10 19:58:31 personally, I'm seeing a lot of "make" systems cohabitating, some of which I've never heard of -- that's okay, but I want to start looking at the gsmd code, not play "science fair" with this bitbake... <---- contructive criticism, please don't take personally! 8D Sep 10 19:58:51 SpeedEvil easiest one: because we do not know yet ourselves Sep 10 19:58:59 (script) wiki RecentChanges: [[Hardware:Neo1973:Alternate_Cases:Camera]] [[Community_nominees_for_GTA02_P0]] Sep 10 19:59:26 SpeedEvil faq do not help at all. believe me. people do not read them. Sep 10 20:00:03 There is some information that you certainly know. Will GTA02 go on sale next week - for example. There is presumably a minimum possible timeline that it might ship by if everything falls right. This is very unlikely to slip in the 'quicker' direction. Sep 10 20:00:09 SpeedEvil and i say that because i see the questions people have and answer them if i find the time, not seldomly ba giving them pointers to the right corner of the wiki Sep 10 20:00:17 No, of course you can't give an upper bound. Sep 10 20:01:18 SpeedEvil the best guess i can give you is 'there will be some hardware for the core devs this month' and 'sales will not start next week' ;) Sep 10 20:01:20 But, if for example someone might find GTA02 hardware useful for a project that starts on Oct 24, for example, simply knowing that it certainly won't be available is a good thing. Sep 10 20:01:28 roh: consider a strategy of identifying a set of developer/customers who can function as ambassadors/developer-support for OpenMoko -- then the core team can focus on answering *their* questions, and allow that "inner circle" of clued-in developers to support the rest of the community. Sep 10 20:01:52 roh: Many of that set of developers are here, on this channel... Sep 10 20:02:10 Thanks. I gathered probably not this week :) Sep 10 20:02:16 [ Sidenote: I'm very grateful that folks have been giving me wiki pointers in the channel here -- thanks guys! :) ] Sep 10 20:02:54 mwester i'm currently working with michael on a concept how to make the flow of informations steadier and getting more informations out regularly. please be patient, that does not work in 1-2 days. Sep 10 20:03:15 * SpeedEvil stops bugging roh. Sep 10 20:03:18 Understand - we'll wait. Sep 10 20:03:29 * SpeedEvil sets alarm clock for 2.00001 days. Sep 10 20:03:49 More seriously, thanks for all the work roh. Sep 10 20:03:55 And everyone. Sep 10 20:03:56 mwester and yes. we are totally buzzed and greatful to all the people helping us by maintaining in the wiki, answering questions and so on. we could not have come so far without. Sep 10 20:03:59 roh, https://direct.openmoko.com/ still starts with information about August. i'd suggest updating it to say a new batch is scheduled and put a new lower bound plus something like "expected to ship this month" Sep 10 20:04:23 Core devs != public availablity of GTA02 Sep 10 20:04:51 at least since we opened up Sep 10 20:05:15 tuukkah i have that on my todo already. Sep 10 20:05:28 cool Sep 10 20:05:52 A simple way would be 24*7 public webcams on all the core team. We can then document from there. :) Sep 10 20:06:59 hehe! Sep 10 20:07:44 hehe Sep 10 20:08:06 SpeedEvil i think there would be some womans arguing against that Sep 10 20:08:09 ;) Sep 10 20:08:17 and a german BundesTrojaner on every PC/Laptop so that we can see what they are working on. Sep 10 20:08:51 be honestt.. youre just missing our quality pr0n ;) Sep 10 20:09:32 roh, i hope it isnt taiwanize pr0n :p Sep 10 20:09:39 Actually, to be serious, one thing I'd certainly like would be a note on some of the bugs filed if the core team knows something that we don't -- e.g. if the core team has already started to rewrite/fix gsmd, then perhaps we can avoid wasting time working on it anymore... Sep 10 20:09:56 i hear up there the men get a little, eh, confused Sep 10 20:10:10 btw.. have you ever seen something like this with Neo1973: "kernel: usb 3-2: device descriptor read/64, error -62\nkernel: usb 3-2: new full speed USB device using ohci_hcd and address 9\nkernel: usb 3-2: device not accepting address 9, error -62" Sep 10 20:10:25 aloril: on the host or on the new? Sep 10 20:10:29 on host Sep 10 20:10:31 vallor, gsmd source is in package libgsmd Sep 10 20:10:34 mwester thats what bugzilla is for Sep 10 20:10:57 ghiottone: thank you :) Sep 10 20:11:01 mwester i think everybody in the coreteam gets bugzilla as mail also Sep 10 20:11:11 it was connected OK, then got disconnected (while did nothing on Neo1973 or cable) and then above when trying to reconnect Sep 10 20:11:13 meaning every change as a mail in a folder Sep 10 20:11:14 aloril: I've seen it when my usb hub acts up, I have to reboot the hub to clear the problem Sep 10 20:11:14 vallor, did you read what I wrote before? Sep 10 20:11:24 vallor, maybe I got a net problem :/ Sep 10 20:11:45 ghiottone: let me scroll up -- I've been fighting with getting the mokomakefile running Sep 10 20:11:51 mwester: if you want to know what people are working on, its easy just to look into the commitlog ml. there you can see that e.g. harald is still working on uboot, mickey on terminal2, sometimes dialer2 etc. and exactly noone is working on gsmd Sep 10 20:12:00 vallor: hmm.. same thing on other machine too, anyway rebooted, took battery out of Neo1973 (I wonder how long to be sure it has been fully reseted?) Sep 10 20:12:10 Roh: I would hope so - I'm just asking that a note in bugzilla such as "I think I know what this is, I will write a patch as soon as I have time, please don't anybody waste time trying to fix this without board schematics or my knowlege." That would be good information :) Sep 10 20:12:24 hhf423_w devels not only work on software which is tracked by bugzilla Sep 10 20:12:45 hhf423_w harald is mostly busy because of gta02. and sorry. i cannot provide details there Sep 10 20:13:32 ghiottone: I did post "thank you" :) -- I'm about to just drill down and look at the individual package, but I was hoping to have a successful "make world" before poking at things Sep 10 20:13:37 mwester: so far, if I have filed a bug in bugzilla, I got good response. e.g. on bluetooth issues Sep 10 20:13:44 mwester see bug 788, harald made a comment and put it onto his todo.. if someone fixes it before he does, be shure he will use that informations. Sep 10 20:14:29 vallor, ok my messages went lost then. when you have downloaded the MokoMakeFile just do a make setup update Sep 10 20:14:36 roh: that one is an excellent example of where a few moments spent by harald have been very very helpful! Sep 10 20:14:37 roh: but bugzilla is not the same as commitlog ?! Sep 10 20:14:39 vallor, this will set things up the correct way Sep 10 20:14:44 the problem I ran into (and fixed here) is that gnu-configize wasn't in the path Sep 10 20:14:57 bugzilla is great for such communications since it enforces some tabelaric view and a progress. makes it much easier to work with than loads of unsorted mails. especially since working mail and or stating the question or problem concrete seems to be a problem Sep 10 20:15:18 ghiottone: yes sir, that what I'm doing, and it's crapping out on "bunutils-cross" :\ Sep 10 20:15:22 er Sep 10 20:15:28 take two: ghiottone: yes sir, that what I'm doing, and it's crapping out on "binutils-cross" :\ Sep 10 20:15:45 vallor, lol. you don't have to build all to get the sources Sep 10 20:16:11 say, is there any news on whats beween us and a working gps? Sep 10 20:16:21 vallor, just do: . setup.env; cd build; bitbake -c fetch libgsmd ; bitbake -c unpack libgsmd Sep 10 20:16:41 ghiottone: okay, I'll do that -- thank you! :) Sep 10 20:16:42 hhf423_w commitlog is just a log what was done. nothing which helps communicating bi-directional Sep 10 20:17:08 vallor, np :) sources will be here tmp/work/armv4t-angstrom-linux-gnueabi/libgsmd-0.1+svnr2806-r24/gsm/src/gsmd Sep 10 20:19:13 I'm looking at the sources now... Sep 10 20:19:19 roh: sure, but you can see what part of the openmoko stuff people are actually working on. its informational, at least I think it is Sep 10 20:20:26 does anyone know if the graphical application manager is actually installable yet? ipkg install? Sep 10 20:29:00 ghiottone: pardon me if this is a dumb question, but are you the libgsmd/gsmd developer? Sep 10 20:29:31 * vallor is reading the pseudocode in the readme... Sep 10 20:29:49 vallor, no. just another moko owner Sep 10 20:29:57 :) ^5 Sep 10 20:30:10 vallor: there currently seems to be no gsmd developer Sep 10 20:30:18 huh Sep 10 20:31:23 seems like a pretty core item -- I'm looking to see if I can wrap my head around it Sep 10 20:31:38 huh? I thought somebody from OM was working on it? Sep 10 20:34:15 ghiottone: or maybe someone from this project just needs to liason with the OM developer Sep 10 20:34:24 * vallor looks Sep 10 20:35:01 ghiottone: vallor there was the rumor that OpenMoko is planning to hire someone or that someone from the staff would look at it Sep 10 20:35:22 oh Sep 10 20:35:24 I see Sep 10 20:35:30 but you can clearly see from the commitlog that there is basically nothing going on wrt gsmd Sep 10 20:35:36 ghiottone harald did the initial design, and mostly maintains it. but recently he is very thin on time. Sep 10 20:36:05 ghiottone there also were a lot of patches which he merged some time ago Sep 10 20:37:03 I thought I saw some commits on gsdm today regarding command sequences Sep 10 20:37:17 a week ago I would have had the hope that after putting some nice features that would help with gsmd development into uboot (mainly gsm passthrough) he would start cleaning up gsmd, but is looks like he is now working on other things Sep 10 20:37:36 something like fixing that only one command is sent at once... so I thought it was being developed Sep 10 20:38:02 hhf423_w, pity. I think that a working gsmd is the most wanted feature Sep 10 20:38:31 so that I can show the neo to friends without having to do long talks and tell excuses Sep 10 20:38:34 ghiottone: I agree Sep 10 20:38:51 ghiottone: exactly :-) Sep 10 20:38:53 hehehe Sep 10 20:39:04 or have gsmd crap out when you send it an sms... Sep 10 20:39:06 so far, only the friend of mine who work on free sw appreciated it Sep 10 20:39:10 (which just happened to me) Sep 10 20:39:56 vallor: did you use SMStool? Sep 10 20:40:18 vallor: from the commitlog, if any, thomas at sita.openmoko.org is working on gsmd related stuff Sep 10 20:40:33 hhf423_w: ah, thank you! Sep 10 20:40:53 hey El_Salvador Sep 10 20:41:17 where can I get SMStool? Sep 10 20:41:21 hmm, if I could extract commitlog somehow I could do trends on what feature receives most changes per week Sep 10 20:41:36 I'm not sure about the smstool -- I just used my treo to send an sms to the neo -- it didn't do anything, and upon checking, gsmd had stopped Sep 10 20:41:38 ghiottone: look up SMStool in wiki Sep 10 20:42:05 its from Kriss Sep 10 20:42:22 right. I can covnert my addressbook to evoltion stuff. Too bad I have to drop info (or is there a way to store >3 addresses, home, work, other? who thought up those limitations?) Sep 10 20:44:14 hhf423_w, ok. wow. Kriss is a working hard developer. maybe I can package smstool too, if he had not already done it Sep 10 20:44:39 at the beginning of august, gsmd received a lot of attention from laforge/harald, but not much since then Sep 10 20:44:54 hhf423_w, I bet there are a lot of statistic showing tools for subversion Sep 10 20:44:55 ghiottone: I think he has not done it Sep 10 20:44:58 hhf423_w, at least you can check out the svn repo and run svn log Sep 10 20:45:06 next, I'll add the phone numbers, and then on to the real stuff. Sep 10 20:45:12 ghiottone: would be cool if you could make it get a button in applications Sep 10 20:47:13 hhf423_w, that is fairly easy :) Sep 10 20:47:21 hhf423_w, I'll have a look right now Sep 10 20:47:38 but I can already see it s*cks. evolution contacts has no business-mobile. it has mobile, it has business, but not the combo. and yes, I know people who have all three. Sep 10 20:48:22 anynone happen to know whether evolution contacts are based on a standard? vcard or something? Sep 10 20:49:04 vallor: looks like most of the stuff that harald/laforge did put in gsmd and related code was coming from balrog at zabor.org Sep 10 20:51:33 hhf423_w: k , I'm thinking it's better to just coordinate with the maintainer, unless I have questions re: the balrog gentleman's changes Sep 10 20:52:00 is there an sms icon in the standard set? Sep 10 20:52:25 no idea Sep 10 20:54:02 vallor: looks like that gsmd stuff is horrid. Like "changed from ATE as first send command to AT because of some phones firmware crapping out", then: "removed AT as first send command because some other phones firmware craps out" etc. Sep 10 20:55:23 hmm, gsmd should be usable with any phone that supports AT commands, even those that don't run linux Sep 10 20:55:58 Indeed. what's probably needed is a config file Sep 10 20:56:06 absolutely Sep 10 20:56:09 'do x y z' to init modem Sep 10 20:56:48 now that i know how to use cu i connected to my series 40 phone :-) Sep 10 20:57:18 Hmm, sounds like the good old BBS days. Sep 10 20:57:34 gsmd command line options talk about "GSM modem vendor plugin" and "GSM modem machine plugin" Sep 10 20:58:12 Lynet: actually, we would do well to consider lessens learned from the old-timey dialup modem environment, regarding robustly talking to modems w/software... Sep 10 20:58:24 (cu is stupid in that it doesn't let root use the device unless it's owned by the user uucp...) Sep 10 20:58:39 vallor: Always reset to known good config, in other words. Sep 10 20:58:58 * SpeedEvil ponders if mail via uucp via gprs is completely insane. Sep 10 20:59:39 SpeedEvil: Tuck an ATM and a PPPoE layer in there too, and you're golden. ;-) Sep 10 21:00:35 well, cu doesn't like talking to my modem :\ -- here's the device: crwxrwxrwx 1 uucp dialout 204, 64 Sep 10 13:58 ttySAC0 Sep 10 21:01:02 cu connects, but I don't get any responses, until I kill cu, then re-run it Sep 10 21:01:03 I want POPPPoE Sep 10 21:01:13 may be non-trivial though. Sep 10 21:02:50 vallor, do you mean on the neo? cu has trouble with flow control, you need to run on another console stty -F /dev/ttySAC0 crtscts Sep 10 21:03:02 tuukkah: ah! Sep 10 21:03:18 vallor, wait a minute Sep 10 21:03:34 did you say it worked after re-running? Sep 10 21:03:55 no, it just dumped all the responses that had been saved up from the previous session Sep 10 21:04:09 (and were, I guess, waiting to be read from the device) Sep 10 21:04:16 Evenin' Sep 10 21:04:55 * hhf423_w thinks that uucp rocks Sep 10 21:05:38 * mwester is old enough to have run a uucp node Sep 10 21:06:07 wow Sep 10 21:06:15 * hhf423_w can still remember the days when nntp:all.all would only require 64kbit bandwidth Sep 10 21:06:41 hhf423_w: when we first started our service, we provided uucp email to several clients -- also had a "door" running under an account that would let folks grab newsgroups into FIDO bbs forums... Sep 10 21:07:18 doh, fido :-) Sep 10 21:07:20 now the stty trick isn't doing it for me... Sep 10 21:07:27 I remember when a steady 9600kbps stream from a satellite was enough... :) Sep 10 21:07:36 running stty second time it hangs Sep 10 21:07:41 how again did they call dorks/morons/trolls on fido? can not remember ... Sep 10 21:08:02 but they had some special term Sep 10 21:08:19 hehe, no idea, it's been years and years Sep 10 21:08:37 tuukkah: I haven't seen that yet Sep 10 21:09:45 but gsmd can still work so it's just another issue with cu Sep 10 21:10:31 hhf423_w: DAU, Elch? Sep 10 21:10:41 * hhf423_w is reading fake steve and falling from his chair, lughing Sep 10 21:11:06 Lynet: hmm, no, was something else Sep 10 21:11:27 Hmm, no wonder Jobs wanted 10% of the cell contract. Sep 10 21:14:22 great! package done at first try :) Sep 10 21:16:19 you can install smstool from my repo (look at my user page Alessandro) Sep 10 21:16:44 +CMS ERROR: 512 <--- how do you get out of this funky mode? Sep 10 21:16:55 gsmd doesn't know what to do with it :\ Sep 10 21:16:56 only thing that is good also for SettingsGUI is that it has to be installed in the same destionation (card or root) as python Sep 10 21:17:44 s/good /valid / Sep 10 21:17:44 ghiottone meant: only thing that is valid also for SettingsGUI is that it has to be installed in the same destionation (card or root) as python Sep 10 21:19:14 ah, toggling /sys/bus/platform/devices/gta01-pm-gsm.0/power_on did the trick Sep 10 21:22:43 maybe the reset part (which is in the script in init.d) should be put in gsmd as part of the initialization process Sep 10 21:24:42 ghiottone: put a comment into the smstool wiki page Sep 10 21:26:06 hhf423_w, I prefer to wait to ask kriss for this. he would undoubtely agree but.... Sep 10 21:26:46 kristian-m, ping ? Sep 10 21:27:09 ghiottone: wiki is not about asking, its about doing. If he does not like it, he can/will remove it :-) Sep 10 21:28:08 always yield to the hands-on imperative! Sep 10 21:28:51 I usually prefer asking people. it leads to a much more serene and respectful atmosphere Sep 10 21:29:05 but if you say so, then I'll do it Sep 10 21:29:36 if you want to be extra nice, put a note about the change onto his talk page Sep 10 21:31:17 hmm, I ipkg installed smstool but no icon? Sep 10 21:33:03 hhf423_w, xserver needs restarting Sep 10 21:33:13 give me good feedback before I commit the changes Sep 10 21:33:15 ghiottone: that I did Sep 10 21:33:35 did you install it in the card? Sep 10 21:33:40 yes Sep 10 21:33:47 ipkg-utils add smstool Sep 10 21:33:50 like python Sep 10 21:33:53 then restart again Sep 10 21:34:27 ops sry Sep 10 21:34:31 ipkg-link add smstool Sep 10 21:35:41 ok, let me try that Sep 10 21:36:11 ah, great! Sep 10 21:36:42 I assume it is working Sep 10 21:37:11 hmm, the icon is there Sep 10 21:37:18 but i can not start it Sep 10 21:37:44 I get the error: /usr/bin/python: error while loading shared libraries: libpython2.5.so.1.0: cannot open shared object file: No such file or directory Sep 10 21:37:54 mmmmm Sep 10 21:38:17 checking dependencies.... Sep 10 21:38:49 in the mean while try doing ipkg-link mount /media/card/ Sep 10 21:39:09 this should create the missing link of that library Sep 10 21:41:37 hhf423_w: ldconfig? Sep 10 21:41:54 doing ipkg-link mount /media/card/ now... Sep 10 21:42:03 I think ldconfig will be launched by ipkg-link Sep 10 21:42:09 ah, ok. Sep 10 21:42:52 ipkg-link sets up symbolic links to packages in the card. And after that it probably runs ldconfig to update library list Sep 10 21:43:09 at least this is what I think it does Sep 10 21:43:27 right, that worked Sep 10 21:43:40 great! Sep 10 21:43:59 adding the note Sep 10 21:44:43 summatusmentis: you are at @unm :) Sep 10 21:45:23 summatusmentis: i'm at unm too :) Sep 10 21:46:19 ok, will go to bed now. talk to you all tomorrow Sep 10 21:46:29 * hhf423_w waves goodnight Sep 10 21:47:53 bedboi: yeah, Morris campus Sep 10 21:48:17 summatusmentis: i'm at ece Sep 10 21:48:29 what city are you in? The Twin cities? Sep 10 21:48:48 i'm at abq Sep 10 21:49:08 AHAHA Sep 10 21:49:12 you are @ UMN Sep 10 21:49:13 lol Sep 10 21:49:15 im @ UNM Sep 10 21:49:37 I'm confused now... I'm at the University of Minnesota - Morris :-P Sep 10 21:49:44 UNM? where's that? Sep 10 21:50:03 University of New Mexico (Albuquerque) Sep 10 21:50:14 the land of the chile Sep 10 21:50:26 lol, oh. Yeah, I'm a bit of a ways away from there Sep 10 21:50:43 yep Sep 10 21:57:02 anyone good at CPU profiling? Sep 10 21:57:55 i'm trying goole proftools but it seems that they don't want to use glib debug symbols (installed in /usr/lib/debug/usr/lib) Sep 10 21:57:58 here's something better than cu Sep 10 21:58:02 socat - /dev/ttySAC0,crnl Sep 10 21:58:42 I've been reading the documentation at http://www.developershome.com/sms/cpmsCommand.asp <--- goldmine! :) Sep 10 21:58:48 AT sms handling Sep 10 21:59:19 (script) wiki RecentChanges: [[User:Alessandro]] [[User:Hhf423]] [[SMSTool]] [[WiFi_support_in_OpenMoko]] [[Power_management_menu]] [[WLAN]] [[WiFi]] Sep 10 21:59:43 i've found apatch Sep 10 21:59:45 cool Sep 10 22:03:16 is there a easy way to print images on full screen? a cli viewer app? Sep 10 22:05:44 probably something using the framebuffer Sep 10 22:06:25 Anyone heard anything about gllin yet ? Sep 10 22:08:50 hrm, booted machine a while ago, battery out of Neo1973 quite long time, still "kernel: usb 3-2: device descriptor read/64, error -62" in PC :-( Sep 10 22:09:47 aloril: I've seen that -- is it connected to a usb hub? Sep 10 22:09:58 no, directly to PC Sep 10 22:10:05 worked earlier just fine Sep 10 22:10:11 aloril: in my case, I had to reboot my usb hub (Dell monitor) to get it to come back Sep 10 22:10:30 did reboot PC Sep 10 22:10:56 aloril: I get the same thing when trying to sync my treo -- then the port is "spoiled" until the hardware is reset Sep 10 22:11:30 aloril: I used to move to a different port when that happened, but now it seems (at least with the neo) that rebooting the usb hub is a fix Sep 10 22:11:53 aloril: oh, didn't read that Sep 10 22:11:59 hmm... only rebooted, didn't power off PC Sep 10 22:12:11 ya, I think you have to reset the hardware... Sep 10 22:13:40 aloril, trying unloading and then reloading the usb bus module Sep 10 22:14:15 ghiottone: didn't help (and rebooting machine should do that anyway) Sep 10 22:14:25 ghiottone: he actually soft-booted the PC, which didn't clear up the problem Sep 10 22:14:51 vallor: I think you are right, it needs poweroff Sep 10 22:15:00 oh yeah right Sep 10 22:16:18 hmm, I wonder if this is a case where an external hub might be better? (then you can just reboot the hub, instead of the whole thing) -- would have to be a fairly decent hub that handled dfu correctly, of course... Sep 10 22:16:46 vallor: power off Neo1973, power on Neo1973, connect to unspoiled machine -> OK Sep 10 22:17:02 vallor: yeah, except flashing I think needs direct connection Sep 10 22:18:40 connect Neo1973 when spoiled state -> spoil other side Sep 10 22:19:27 aloril: yes, it might actually be a linux driver issue, where it needs to completely reset the port after unplugging Sep 10 22:19:56 i got the same problem today Sep 10 22:32:45 anyway, quite rare problem (first time ever for me and it has been connected to PC most of time of about month) Sep 10 22:33:13 i just happened to update uboot so i was afraid i broke it Sep 10 22:34:22 date -s in busybox doesn't accept any format i try to give it Sep 10 22:36:27 hh:mm:ss is a valid format, but the date is off by one too Sep 10 22:36:31 man, I hope we aren't trying to use the "PDU" mode in gsmd -- I've been reading about that for half an hour, and it's still making my head hurt Sep 10 22:36:49 for SMS I mean Sep 10 22:37:10 we probably should do the simples thing that works to get the messages out of the chip Sep 10 22:38:27 eek Sep 10 22:38:46 the _simplests thing would be to use the text mode Sep 10 22:38:54 i.e.: Sep 10 22:39:06 AT+CMGL="ALL" Sep 10 22:39:07 +CMGL: 2,"REC READ","453",,"07/09/09,15:18:09-28" Sep 10 22:39:07 Did you know? Message rates for myFaves apply per your rate plan. Go to my.t-mobile.com for your plan details. Sep 10 22:39:07 +CMGL: 3,"REC READ","453",,"07/09/09,15:18:10-28" Sep 10 22:39:29 that won't work for most people =) Sep 10 22:39:46 nono -- that's what gsmd would do Sep 10 22:40:18 so why don't we let gsmd do that? Sep 10 22:40:21 communicate with the gsm/gprs modem using the "text mode" -- much easier to debug and suss out Sep 10 22:40:52 tuukkah: that's what will be happening, but I'm looking at gsmd to see if it's "doing the right thing" Sep 10 22:41:10 (which it obviously isn't, it exits frequently for no reason) Sep 10 22:45:47 tuukkah: the busybox date command may be problematic; if it's serious enough of an issue, open a bug -- the build options for busybox can be altered, or (as has been done on other OE distros) the real "date" command can be added to the image. Sep 10 22:45:55 vallor: sorry, i just looked over this ... are you futzing about with t-mobile myFavs SMS formatting? Sep 10 22:46:14 * mwester has encountered date problems with busybox before. Sep 10 22:46:23 cw666: no, I'm just trying to get the modem to send an sms message Sep 10 22:46:35 and read, too Sep 10 22:46:47 that message was one I received from t-mobile Sep 10 22:48:48 vallor: ah, i've had that working w/ gsmd and also by hand Sep 10 22:49:28 did you have to do the hexidecimal encoding stuff (pdu mode), or were you successful in text mode? Sep 10 22:49:37 vallor: pdu mode Sep 10 22:49:46 hmm, k Sep 10 22:49:51 vallor: i actually didnt know there was a text mode, ive just played with this Sep 10 22:49:54 briefly Sep 10 22:50:50 vallor: i wondered if anyone knew the magic sms format and destination to update myFavs Sep 10 22:51:39 mwester, i'm reading the source :-) Sep 10 22:52:03 I used date to set my date/time Sep 10 22:52:19 there's no switch -- just do "date --help" to see the format it takes Sep 10 22:52:30 rdate 192.168.0.x Sep 10 22:52:30 (for the busybox date) Sep 10 22:52:37 is probably the easiest thing Sep 10 22:52:43 wow, it has rdate? Sep 10 22:52:47 (ie. have datetime enabled on the connecting machine) Sep 10 22:53:06 #!/bin/ash Sep 10 22:53:06 date `ssh 192.168.0.200 date "+%m%d%H%M%Y.%S"` Sep 10 22:53:15 oh - right Sep 10 22:54:23 gack -- brb, eye break Sep 10 22:56:04 date: invalid date `0911015407.30' Sep 10 22:56:14 why not use rdate? Sep 10 22:56:24 cw666: I diddn't realise it was there. Sep 10 22:56:26 looking at the wiki, i see '* Sep 10 22:56:27 i dont see why people have to get clever with ssh, it's also very heavy weight Sep 10 22:56:27 12MBit NOR flash for 'unbrickable emergency boot' feature. ' Sep 10 22:56:32 that true? Sep 10 22:56:41 for the gta02? Sep 10 22:57:03 cw666, i don't have datetime port open in firewall Sep 10 22:57:14 tuukkah: huh? you dont need it Sep 10 22:57:23 tuukkah: whatever maching the usb is connected to, enable datetime on that Sep 10 22:57:24 But teh above works for me Sep 10 22:57:27 cb22 ah.. where does that come from? Sep 10 22:57:37 from http://wiki.openmoko.org/wiki/Neo1973_GTA02_Hardware roh Sep 10 22:57:56 cw666, the machine the usb is connected to has the firewall :-) Sep 10 22:58:06 I'm unsure if the unbrick flash is inside the SoC Sep 10 22:58:14 The new one is a multi-chip-module Sep 10 22:58:26 but there is unbrick flash....? Sep 10 22:58:27 tuukkah: umm, so allow datetime then Sep 10 22:58:45 cb22 yep.. its true there will ne a new nor flash.. i dunno about the final size, but small Sep 10 22:59:04 only uboot needs to be in there. to be exact. one uboot Sep 10 22:59:05 now why does this phone just keep on sounding better and better :) Sep 10 22:59:43 Can we reflash the stored uboot? Sep 10 22:59:51 fsck. help says YY but it needs YYYY Sep 10 22:59:54 that was it Sep 10 23:00:00 And is it the normal boot mechanism. Sep 10 23:00:00 cb22 the nor flash will have its write-protect line not connected to the app cpu, but only to the debug port. so if youre able to erase it you will have the tools to unbrick it. Sep 10 23:00:22 cb22 someone without debug board and soldering iron should be unable to really brick his device then Sep 10 23:00:36 roh, thats nice, but there is still the normal uboot existing on the nand right? Sep 10 23:00:40 is there a plan on what to do about time zones? Sep 10 23:00:45 cb22 ack. Sep 10 23:01:05 in fact, that kicks ass, no risk of bricking unless you can recover it Sep 10 23:02:05 The new SoC has DRAM and NAND flash inside? as a multi-chip? Sep 10 23:02:14 jap. Sep 10 23:02:19 That should help lots with boardspace. Sep 10 23:02:22 well - a bit Sep 10 23:02:30 as you can lose the RAM/bus Sep 10 23:02:53 IIRC it only has 64mb built in though Sep 10 23:02:59 the other 64mb is external Sep 10 23:03:00 mbit? Sep 10 23:03:04 8MB Sep 10 23:03:05 of ram Sep 10 23:03:07 there will be a external ram, but it stell saves pcbspace for things like e.g. the 5V generating parts for the usb host mode Sep 10 23:03:08 that's enough for kernel and small rootfs Sep 10 23:03:10 Oh - right. Sep 10 23:03:25 5V generating parts aren't really bad. Sep 10 23:03:40 * SpeedEvil is currently waiting on bits for GTA01 USB power thingy Sep 10 23:04:33 * meus is booting openmoko 2007.2 under qemu for the first time. exciting :) Sep 10 23:04:34 And am trying to avoid adding too many features. Sep 10 23:04:48 'Oooh - a compass would be easy...' Sep 10 23:05:21 SpeedEvil, have you tried if it's a good idea to put "echo "AT@POFF" | socat - /dev/ttySAC0,crnl" in /etc/init.d/gsmd and "amixer sset "Amp Mode" "Off"" in /etc/init.d/alsa-state? you got the meter :-) Sep 10 23:05:38 I have not tried that. Sep 10 23:05:41 howerver. Sep 10 23:06:05 when you do that manually, and power off, the current drawn goes to under 100uA Sep 10 23:06:46 i was just thinking those two lines might be worth it Sep 10 23:07:07 Sure. Sep 10 23:10:01 hmmm.... handling all the sms-modem stuff might already be available in this: http://www.kannel.org/ Sep 10 23:10:06 actually the socat sin't able to proceed if the gsm modem is already off Sep 10 23:10:12 ''Kannel is a compact and very powerful open source WAP and SMS gateway, used widely across the globe both for serving millions of short messages (SMS) and pushing WAP service indications.'' Sep 10 23:12:12 I just use echo Sep 10 23:12:21 i use echo too Sep 10 23:12:41 hehe, you use echo to send sms's? Sep 10 23:12:48 yes, i have done Sep 10 23:12:53 why on earth do we first power off the modem and then stop gsmd %-) Sep 10 23:13:19 tuukkah, because thats what the lords of FIC demanded, and unless you sacrifice a cow, it wont change :p Sep 10 23:13:46 cb22: it makes sense to have gsmd turn off the modem on it's way down IMO Sep 10 23:14:27 sure :) i was just adding in a useless comment Sep 10 23:14:40 i tend to do those at 1:12am, when i have to wake up at 6am Sep 10 23:16:09 echo -e "AT@POFF\r" >>/dev/ttySAC0 Sep 10 23:29:21 zowie, this "kannel" has all the charset translation stuff for handling GSM PDU messages :) Sep 10 23:29:31 * vallor looks too see how its done in gsmd -- if at all... Sep 10 23:30:10 cd Sep 10 23:30:14 ups :P Sep 10 23:50:17 does the gui sound keep the /dev/dsp resource busy prevents madplay or mplayer from using it? how can i to stop it? Sep 10 23:51:16 milo_: do not stop it, join it... use aoss ;-) Sep 10 23:53:42 oh :) thanks Sep 10 23:55:07 ghiottone pong, thanks for the package - I'll try to continue work on both projects, new releases are planed at the end of this week. Sep 10 23:58:59 (script) wiki RecentChanges: [[Main_Page]] [[Current_events]] Sep 11 00:01:33 cesarb: AOSS as in Alsa's OSS interface Sep 11 00:01:34 ? Sep 11 00:02:48 abraxa_: aoss as in the aoss command Sep 11 00:03:01 Ah Sep 11 00:03:22 abraxa_: "man aoss" on most distros should give you more detail Sep 11 00:03:33 * cesarb tends to use artsdsp instead of aoss for some random reason Sep 11 00:04:24 er... would that be "aplay"? Sep 11 00:04:25 brb Sep 11 00:07:15 vallor: no, aplay is more like sox Sep 11 00:07:48 vallor: while artsdsp is arts's equivalent of aoss Sep 11 00:13:35 oh, hmm -- seems like there should be a standard agreed upon, so everybody's sound thingies don't collide later Sep 11 00:13:53 (assuming the driver doesn't just mix the audio)_ Sep 11 00:19:38 vallor: with alsa you have dmix (on recent alsa versions it's automatic), so there's no collision. It's only old OSS applications which don't play along (try running mplayer with the -ao set to alsa to check) Sep 11 00:21:24 * mjr seems to recall something about the pulseaudio sound server being used Sep 11 00:27:29 yeah, pulseaudio is in init.d Sep 11 00:28:01 i had to stop it to get madplay/mplayer to work with sound Sep 11 00:28:08 on the newer build Sep 11 00:28:42 heh Sep 11 00:29:37 also found alsamixer is on the new, didnt know about it Sep 11 00:29:41 new=neo Sep 11 00:35:45 and alsactl -- with various .state files in /etc Sep 11 01:50:22 hi there. Sep 11 02:12:00 okay, what the heck is this all about? Sep 11 02:12:02 File "/extra/moko/bitbake/lib/bb/runqueue.py", line 281, in prepare_runqueue Sep 11 02:12:02 listid = taskData.tasks_lookup[fnid][target[1]] Sep 11 02:12:02 KeyError: 'do_rm_work' Sep 11 02:13:40 openmoko: 03nbd * r2947 10/developers/nbd/ (7 files in 3 dirs): clean up patches, update sdio merge, some small ar6k bits Sep 11 02:48:59 hozer: you're missing the INHERIT += rm_work Sep 11 02:58:52 build is borked - again - package libelf_0.8.6.bb **** ENDING LOGGING AT Tue Sep 11 02:59:56 2007