**** BEGIN LOGGING AT Mon Aug 27 02:59:57 2007
Aug 27 03:19:04 <rwhitby> technil: it looks for source tarballs cached on the downloads server.  not finding it is a common case, in which case it just goes and checks it out itself from svn
Aug 27 03:19:12 <rwhitby> so what you are seeing is normal.
Aug 27 04:49:18 <blindcoder> moin
Aug 27 04:49:34 <CM> Morning
Aug 27 05:12:40 <linux_galore> ooh the local LUG (www.slug.org) is holding a talk on the OM
Aug 27 05:12:51 <linux_galore> at the end of this month
Aug 27 05:42:18 <cheriff> Hi, has anyone played with the QVGA patches and gotten them to work?
Aug 27 05:56:07 <gcb77> anyone having sound problems on the NEO?  My sound isn't working, and I'm wondering if its me or something else
Aug 27 06:02:10 * rwhitby disables openmoko-dialer2 from /etc/matchbox/session in his quest to find out which app is really at the bottom of the gsm registration problem.
Aug 27 06:02:32 <CM> gcb77: I did, with my own builds
Aug 27 06:02:44 <CM> gcb77: But with the offical images it works
Aug 27 06:03:00 <CM> Maybe I'm missing some kernel modules or something
Aug 27 06:03:22 <gcb77> I flashed it yesterday with new kernel image, but I played MP3 after
Aug 27 06:03:45 <CM> Ok
Aug 27 06:04:02 <gcb77> I played with alsamixer, that's probably what's messed up
Aug 27 06:09:25 <jeddy3> gcb77, start playing something, and fiddle around in alsamixer until you hear it again :)
Aug 27 06:32:16 <cheriff> someone here have wiki edit privilige?
Aug 27 06:32:32 <cheriff> http://wiki.openmoko.org/wiki/Kernel:  "The Kernel on the GTA01 is based on a vanilla 2.6.21.3 Linux kernel from kernel.org"
Aug 27 06:32:43 <cheriff> should now be 2.6.22.1
Aug 27 06:33:00 <cheriff> else patchset doesnot apply
Aug 27 06:34:03 <hads> You should be able to register an account and edit yourself.
Aug 27 06:34:18 <cheriff> okay
Aug 27 06:34:53 <hads> Hasn't it been switched back to 2.6..21 again though?
Aug 27 06:35:51 <cheriff> one of the patches in the series attempts to modify a file found only in .22, it seems
Aug 27 06:37:20 <cheriff> s3c_mci.patch, to be exact - the sd fixes
Aug 27 06:58:23 <gcb77> whew, sound just came back :)
Aug 27 06:58:46 <gcb77> I tried to make a call... and when I pushed the hangup button it beeped
Aug 27 07:01:00 <ljp> beep is good
Aug 27 07:01:24 <gcb77> yeah I was thinking flashing
Aug 27 07:08:08 <jeddy3> sweet 82 updates over the weekend :)
Aug 27 07:08:39 <CM> jeddy3: Nice. Any significant ones?
Aug 27 07:10:02 <jeddy3> lets see, gtk+, lots of kernel modules, all of the openmoko-utils
Aug 27 07:12:13 <CM> Just flashed the 070826 images this morning, but didn't have time to look through everything :)
Aug 27 07:18:55 <Dmitry_Platonov> I'm getting alot of "undefined reference"s in uicmoc4-native-4.3.1-r0. Any solution?
Aug 27 07:28:30 <cw666> has anyone seen a 'moko which fails to power up?
Aug 27 07:30:25 <cibomahto> Dmitry_Platonov: not that i know of, maybe someone here knows better though... there is a bug report: http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=747
Aug 27 07:33:07 <cibomahto> Dmitry_Platonov: I had some minor success modifying some of the qmake files in the project by hand, but thats probably not a good way to go about it.
Aug 27 07:37:47 <roh> cibomahto 2 things: compile with gcc4 
Aug 27 07:38:21 <cibomahto> roh: ok, i have been thinking about trying that
Aug 27 07:38:24 <roh> and second: gcc3 is only needed for qemu (as well the emu for the moko as the qemu*-native for the locales)
Aug 27 07:40:00 <roh> also there was some bug with uicmoc lately which afaik turned out to be a unneeded dependency on xx11-libs. now it builds qt-embedded which does not need this (in the end nothing should be linked to qt, only the webkit needs the build-tools
Aug 27 07:42:54 <cdbot2> * * OM Bug 756 has been created by cw(AT)f00f.org
Aug 27 07:42:55 <cdbot2> * * Device appears to not charge or have failed
Aug 27 07:42:56 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=756
Aug 27 07:43:37 <cibomahto> roh: maybe I am confusing things here; does the build system build a new gcc for the host as well as the target system?  I said i was using gcc 3.4.6 because that is what i am using on my host pc
Aug 27 07:43:53 <cdbot2> * * OM Bug 757 has been created by rod(AT)whitby.id.au
Aug 27 07:43:54 <cdbot2> * * Patch to openmoko-dialer to handle CREG: 2 responses
Aug 27 07:43:55 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=757
Aug 27 07:43:58 <ScaredyCat> svn server farked ?
Aug 27 07:44:32 <cibomahto> sorry if i am mixing things up, i am rather new to the OE environment
Aug 27 07:44:34 <roh> cibomahto it builds a new compiler for staging the crosscompiler and the crosscompiler. i dunno which one is used for the -native stuff.. some oe guy propably knows
Aug 27 07:44:42 <cibomahto> ok :-)
Aug 27 07:45:42 <rwhitby> Here is my proposed fix for the gsm registration problem: http://pastebin.ca/672048
Aug 27 07:47:32 <Dmitry_Platonov> roh, I'm compiling with gcc 4.2.0 and getting same error in uicmoc4-native-4.3.1-r0
Aug 27 07:52:03 <cibomahto> roh: it seems like it is a problem with the configuration, when i edit the qmake source for the first offending program in qt, i can get that one to compile
Aug 27 07:53:50 <cibomahto> roh: but the rest of them are still set up incorrectly so they don't compile.  the missing references are in another static library, it just doesnt get included for some reason
Aug 27 07:55:19 <roh> i'll rebuild that package and see what happens.
Aug 27 07:55:25 <cibomahto> ok
Aug 27 07:56:31 <cibomahto> thanks for looking at it
Aug 27 07:59:14 <rwhitby> ok, with my patch at http://pastebin.ca/672064 and restarting gsmd in the matchbox session, the phone now at least registers with the network, and I get some bars on the gsm icon.
Aug 27 08:01:15 <cibomahto> rwhitby: awesome!
Aug 27 08:01:39 <rwhitby> no, not awesome at all.  it still doesn't make a call successfully.
Aug 27 08:02:31 <cibomahto> oh, bummer
Aug 27 08:06:07 <roh> cibomahto nope... builds fine here
Aug 27 08:07:16 <cibomahto> roh: ok, i will switch to a newer gcc and rebuild the environment, and see if that clears the issue.
Aug 27 08:07:55 <roh> let me know if that helped
Aug 27 08:09:05 <cibomahto> will do
Aug 27 08:11:54 <cdbot2> * * OM Bug 758 has been created by rod(AT)whitby.id.au
Aug 27 08:11:55 <cdbot2> * * libmokogsmd2 doesn't handle CB and DS events
Aug 27 08:11:56 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=758
Aug 27 08:12:49 <rwhitby> patch for that is at http://pastebin.ca/672079
Aug 27 08:13:14 <rwhitby> roh: which core developer is looking after libgsmd, libmokogsmd2 and openmoko-dialer2 ?
Aug 27 08:14:35 <FuzzyCat> anyone else having difficulty svn updating from svn.openmoko.org?
Aug 27 08:16:08 <Dmitry_Platonov> for me, it sometimes closes connection. re-sync helps.
Aug 27 08:16:50 <FuzzyCat> mmm.. been trying for ages, just seems t o drop the connection ...
Aug 27 08:16:55 * FuzzyCat tries more
Aug 27 08:17:21 <FuzzyCat> I suppose they could be doing backups ....
Aug 27 08:17:29 <FuzzyCat> roflmao
Aug 27 08:53:43 <rwhitby> LaF0rge: could someone take a look at bug 575, with patch attached.  without it, registration never succeeds on my provider's network.
Aug 27 08:53:49 <rwhitby> s/575/757/
Aug 27 08:53:50 <apt> rwhitby meant: LaF0rge: could someone take a look at bug 757, with patch attached.  without it, registration never succeeds on my provider's network.
Aug 27 09:20:23 <grasu> is the svn server down?
Aug 27 09:32:58 <komar007> ls
Aug 27 09:34:55 <zecke> .
Aug 27 09:34:57 <zecke> ..
Aug 27 09:36:12 <FuzzyCat> ...?
Aug 27 09:36:24 <LittleIdiot> >) < pacman that eats your dots!
Aug 27 09:36:58 <FuzzyCat> yoink!
Aug 27 09:40:19 <florian> good morning
Aug 27 09:43:53 <cdbot2> * * OM Bug 759 has been created by rod(AT)whitby.id.au
Aug 27 09:43:54 <cdbot2> * * GSM_REGISTER_TIMEOUT is too short
Aug 27 09:43:55 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=759
Aug 27 09:47:57 <mdt> hi
Aug 27 09:48:35 <rwhitby> first gsm call immediately after boot with the patches in those bug reports I've entered in the last couple of hours
Aug 27 09:48:55 <rwhitby> (757, 758, 759)
Aug 27 09:49:25 * FuzzyCat stabs svn server
Aug 27 09:52:25 <rwhitby> still need to restart gsmd in matchbox session for it to work though
Aug 27 09:56:18 <Stephmw> mornin'
Aug 27 09:59:28 <CM> rwhitby: Nice work! :-)
Aug 27 11:10:47 <cesarb> rwhitby: do the patches you posted fix the problem when the symptom is a continuous stream of "AT+COPS=0" in the gsmd log?
Aug 27 11:11:36 <rwhitby> cesarb: yep
Aug 27 11:11:53 <rwhitby> (in conjunction with restarting gsmd in /etc/matchbox/session
Aug 27 11:11:54 <cesarb> rwhitby: cool, since that's what happens on mine
Aug 27 11:12:06 <rwhitby> Do you have a CREG:2 in there too?
Aug 27 11:12:11 <cesarb> rwhitby: I managed to make gsmd work by adding a "sleep 120" just before it on its init script...
Aug 27 11:12:20 <cesarb> rwhitby: dunno, the beginning of the log somehow seems to get truncated
Aug 27 11:12:57 <rwhitby> First the CREG:2 response was causing openmoko-dialer to continue trying to register over and over again.
Aug 27 11:13:10 <rwhitby> such that the real CREG:1 response never came
Aug 27 11:13:52 <rwhitby> then the very first COPS=0 was too soon after the CFUN=1 and that screwed up gsmd it seemed.
Aug 27 11:14:35 <rwhitby> so I doubled the timeout in openmoko-dialer and made it aware of the UNREG_BUSY state such that it didn't keep trying to register when the modem said "hang on, I'm trying to register already, just give me a chance will ya!"
Aug 27 11:15:18 <rwhitby> now I just need to wait for a battery recharge to keep testing the reliability of my current setup
Aug 27 11:16:19 <SpeedEvil> :)
Aug 27 11:17:55 <cesarb> rwhitby: that takes me closer to the time when I can start using it instead of the razr as my primary phone (what I need before that: working voice calls and working SMS, including theses "popup" ones which open immediately)
Aug 27 11:20:23 <cesarb> (oh, and a working input method would be fine too)
Aug 27 11:20:33 <mil_> hi there.. 2007.2 hangs at the very end of the loading screen. i did flash the kernel and rootfs. any tips? the loading screen sucks..i cant even see at which point it hangs..
Aug 27 11:21:55 <mil_> first tried a compile one.. didnt worked out.. so i flashed the snapshots of the buildhost and same error.. e.g. no error because the loader screen just stops at the very end
Aug 27 11:22:11 <mil_> compile/compiled
Aug 27 11:23:32 <ynezz> similar problems have been reported already
Aug 27 11:23:40 <ynezz> but dunno what's the problem
Aug 27 11:24:12 <ynezz> did you 'nand erase rootfs' before you flashed rootfs?
Aug 27 11:25:32 <mil_> ok thanks.. but i dont need a different uboot?
Aug 27 11:25:56 <mil_> ynezz: no i didnt. but the dfu-util finished the flashing without an error
Aug 27 11:26:08 <borg_> mil_: try nand erase rootfs, i had the same problem
Aug 27 11:26:22 <borg_> after updating to 2007.2
Aug 27 11:26:31 <mil_> borg_: thanks
Aug 27 11:27:57 <cesarb> mil_: unless you updated your uboot to the snapshot version, you need the "nand erase rootfs"
Aug 27 11:28:22 <cesarb> mil_: since the uboot from the latest 2007.2 snapshot already has the fix
Aug 27 11:29:16 <mil_> well hopes thats problem :> .. is it "nand erase rootfs" or "nand erase clean rootfs" like on the wiki page`?
Aug 27 11:30:16 <CM> mil_: Use cu or minicom to connect to /dev/ttyACM0 like the wiki says, then nand erase rootfs
Aug 27 11:31:26 <cesarb> mil_: I have only heard of it being "nand erase rootfs" here on IRC, never tried it (opted to upgrade uboot instead)
Aug 27 11:32:54 <cesarb> BTW, any news on the GSM850 issue? (has anyone reported it to bugzilla yet?)
Aug 27 11:32:59 <CM> I've done it a few times, but usually it's not needed.
Aug 27 11:33:30 <CM> Sometimes it worked just flashing the same rootfs again, but then I have a broken uboot :-(
Aug 27 11:33:48 <Tyrael> Hi all, can I do an ot question abuot GPL and license compatibility?
Aug 27 11:34:45 <CM> Ask away, and hopefully someone knows enough to answer :) (Probably not me)
Aug 27 11:38:49 <Tyrael> "A license p is compatible with a license q (or is q-compatible) if:
Aug 27 11:38:49 <Tyrael>     A work licensed under p can be distributed under the terms of q.
Aug 27 11:38:49 <Tyrael> For example, the X11 license is compatible with the GPL version 2, because works licensed under the X11 license, can be distributed under the terms of the GPL."
Aug 27 11:40:00 <Tyrael> means that I can take code from an X11 licensed project, modify and include it in my work that I release under gpl2?
Aug 27 11:40:13 <mjr> yes
Aug 27 11:40:26 <cesarb> Tyrael: http://www.fsf.org/licensing/licenses/ could be helpful for you
Aug 27 11:40:44 <Tyrael> I take it from that pages
Aug 27 11:40:57 <Tyrael> http://www.fsf.org/licensing/licenses/compat.html
Aug 27 11:41:01 <aloril> (script) wiki RecentChanges: [[USB_Networking]] [[User_talk:Hhf423]] [[MokoMakefile]] [[Hardware:Neo1973:Alternate_Cases:Camera]] [[Phase_1_Software_Testing]] [[Talk:Manually_using_GPRS]] [[Repositories]] [[Python_ncurses]] [[Buying_Interest_List]] [[Hardware:Neo1973:Alternate_Cases:Digital_gaming]] and other changes
Aug 27 11:41:04 <Tyrael> thanks
Aug 27 11:44:23 <mil_> cesarb: did the nand erase rootfs now. still waiting for the flashing to be finished. also wanted to update the uboot but i have no debug board and iam a bit afraid of bricking it with a work uboot image :\
Aug 27 11:44:36 <andylockran> morning guys
Aug 27 11:44:36 <mil_> my uboot is outdated its from May 2007 :D
Aug 27 11:47:35 <freepenguin> messaggio
Aug 27 11:50:29 <jgm> Hi is there a known problem with compiling libgsmd at the moment?  It seems to be giving me an empty package 
Aug 27 11:54:54 <cdbot2> * * OM Bug 760 has been created by cayco(AT)cayco.pl
Aug 27 11:54:54 <cdbot2> * * /etc/default/gsmd: 6: Syntax error
Aug 27 11:54:55 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=760
Aug 27 11:57:01 <freepenguin> www.freepenguin.it/index-en.html
Aug 27 12:04:53 <cdbot2> * * OM Bug 761 has been created by cayco(AT)cayco.pl
Aug 27 12:04:55 <cdbot2> * * gsdm does not start
Aug 27 12:04:56 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=761
Aug 27 12:12:50 <Writchie> cesarb: What is the 850 gsm issue you referred to earlier?
Aug 27 12:16:27 <cindy> Can someone point me to a wiki page to answer a build question? I can't figure out how to run bitbake on a package without it rebuilding (almost) everything. I use MokoMakefile to setup/build the system. That works fine (after 12 hours or so). 
Aug 27 12:17:10 <cindy> Then I go to the build directory and run "bitbake -c compile <package-name>" and it rebuilds 1000 packages.
Aug 27 12:17:36 <CM> cindy: What is it you are trying to rebuild?
Aug 27 12:17:52 <cindy> I just want to play with openmoko-feedreader2 right now. I make a one line change.
Aug 27 12:18:14 <FuzzyCat> cindy, : from your OMDIR : cd build ; . ../setup-env ; bitbake theappname
Aug 27 12:19:38 <CM> Hmm.. the libgsmd-too says "U       Unregister from netowrk" but it feels like a too silly bug to mention.. :P
Aug 27 12:20:01 <CM> s/oo/ool/
Aug 27 12:20:01 <apt> CM meant: Hmm.. the libgsmd-tool says "U       Unregister from netowrk" but it feels like a too silly bug to mention.. :P
Aug 27 12:20:19 <cindy> I think I did that (perhaps I didn't cd to build??) and I ran <bitbake openmoko-feedreader2> Thanks for verifying the steps. I guess I am doing the right thing
Aug 27 12:20:46 <FuzzyCat> you need to be in build and you *must* use the setup-env
Aug 27 12:20:49 <CIA-20> openmoko: 03laforge * r2834 10/trunk/src/target/u-boot/patches/ (7 files): 
Aug 27 12:20:49 <CIA-20> openmoko: * split s3c2442 support from s3c2440 (different PLL config)
Aug 27 12:20:49 <CIA-20> openmoko: * fix s3c2440/2442 pll speed calculation
Aug 27 12:20:49 <CIA-20> openmoko: * add pll speed settings to cmd_s3c24xx for s3c2440 and s3c2442
Aug 27 12:20:49 <CIA-20> openmoko: * fix hand-calculated UPLL initializers for 2442 (user manual is wrong)
Aug 27 12:21:21 <FuzzyCat> otherwise the directories it will use will be new ones.. 
Aug 27 12:21:27 <FuzzyCat> hence a full rebuild..
Aug 27 12:21:44 <cindy> Thanks Fuzzycat, I did use setup-env but possibly didn't cd to build. Thanks.
Aug 27 12:21:53 <FuzzyCat> k..
Aug 27 12:22:02 <FuzzyCat> thank me twhen it works ;)
Aug 27 12:22:03 <CM> Ah, first call with libgsmd-tool to my other mobile
Aug 27 12:22:33 <CM> FuzzyCat: Sometimes you have to clean the tmp/work files and touch some stamps, but I never get when.
Aug 27 12:22:39 <FuzzyCat> *sigh*  - I've done so little work today :(
Aug 27 12:23:42 <FuzzyCat> I've never had that, I've only had the bit where I wasn't paying attention and started a build without the setup-env and ended up with a new full build int he root of my moko build user
Aug 27 12:23:53 <zecke> stefan_schmidt: mickey|offline : how was the froscon?
Aug 27 12:24:07 <FuzzyCat> then I wondered where all my disk space went
Aug 27 12:24:38 <ynezz> cindy: next time try, bitbake -b /path/to/file.bb -f -c compile
Aug 27 12:25:07 <FuzzyCat> you can actually use the mokomakefgile to build the package too
Aug 27 12:25:18 <stefan_schmidt> zecke: HI. Only mickey|offline was there. I need at least _some_ days at home. Heading to .dk on thursday.
Aug 27 12:25:24 <write_erase> Hi, Are there known hardware bugs on the current GTA01 ?
Aug 27 12:25:26 <FuzzyCat> make build-package-blah
Aug 27 12:25:55 <cindy> FuzzyCat, that made a huge difference! I guess I missed the "cd build" Arggh!  
Aug 27 12:26:40 <FuzzyCat> :)
Aug 27 12:26:48 <FuzzyCat> btw ... read up 1 line
Aug 27 12:27:10 <stefan_schmidt> write_erase: No. Why do you ask?
Aug 27 12:27:35 <CIA-20> openmoko: 03laforge * r2835 10/trunk/oe/packages/uboot/uboot-gta01_svn.bb: 
Aug 27 12:27:35 <CIA-20> openmoko: * build from latest u-boot mainline head and our current patch repository trunk
Aug 27 12:27:35 <CIA-20> openmoko: * build u-boot for gta02v2
Aug 27 12:28:18 <write_erase> stefan_schmidt, 'Cause I'm wondering if I wait for the next release or if I order one now. I have not much money, and this hardware will be an investment :-)
Aug 27 12:28:44 <Writchie> write_erase: stefan_schmidt: i consider only charging at 40ma a possible hardware bug as apparent inability to operate on gsm 850 - and at&t SIM problems
Aug 27 12:28:49 * CM wants a new uboot but is too afraid to brick it
Aug 27 12:28:57 <stefan_schmidt> write_erase: If you only have money for one. I would wait for GTA02.
Aug 27 12:29:22 <write_erase> stefan_schmidt, Yeah, but no idea where it"s released. 
Aug 27 12:30:00 <write_erase> stefan_schmidt, I wish I can start hacking this device as soon as possible 
Aug 27 12:30:02 <stefan_schmidt> Writchie: gsm 850: First time I hear of it. AT&T SIM needs a GSM firmware fix and 40ma charging is also the first time I hear of it.
Aug 27 12:30:37 <stefan_schmidt> moment, /me fetches food
Aug 27 12:30:44 <SpeedEvil> 40ma is the default charge current when plugged in 
Aug 27 12:30:53 <write_erase> can you tell me how long the battery last while GSM is on ( phone call) 
Aug 27 12:31:13 <SpeedEvil> Umm.
Aug 27 12:31:19 <SpeedEvil> Probably around 2-3 hours
Aug 27 12:31:32 <SpeedEvil> Based on some very quick hack measurements I made.
Aug 27 12:31:41 <Writchie> stefan_schmidt: AFAIK everyone attempting to connect in GSM 850 only environment fails. FCC test report shows no testing below 1900 Mhz. Actually operation on 850 GSM remains unconfirmed
Aug 27 12:31:43 <SpeedEvil> This is in a poor signal area.
Aug 27 12:32:18 <write_erase> ok that's perfect ...
Aug 27 12:32:40 <stefan_schmidt> Writchie: hmm, can people please fill a bug about this? With details.
Aug 27 12:32:41 <Writchie> stefan_schmidt: I get same results as SpeedEvil: Default charge is only 40ma (GSM modem drains 30ma so net charge can be as low as 10ma unless battery removed.
Aug 27 12:33:00 <write_erase> Someone know the "10 septembre shipping date" is still valid ? or there are extra delays for delivery ? 
Aug 27 12:33:21 <zash_se> october?
Aug 27 12:33:31 <stefan_schmidt> Writchie: hmm, please also a bugreport.
Aug 27 12:34:08 <Writchie> stefan_schmidt: OK, is the nature of the AT&T SIM problem understood?
Aug 27 12:34:20 <stefan_schmidt> I'm not a hardware guy, but without real bugreports we can't handle this. People just talking about it on IRC are not good to track. :)
Aug 27 12:34:46 <zash_se> write_erase: "The Consumer Version will be available in October." -- http://www.openmoko.com/products-neo-base-00-stdkit.html
Aug 27 12:35:00 <stefan_schmidt> Writchie: My last status on this was that this has to be checked with TI engineers...
Aug 27 12:36:18 <Writchie> stefan_schmidt: then the statement "needs a GSM firmware fix" is a just a guess?
Aug 27 12:36:41 <Writchie> stefan_schmidt: in other words this could require a hardware fix
Aug 27 12:37:46 <stefan_schmidt> Writchie: Of course it is a guess. You never really know it until it is fixed. It's hardware after all.
Aug 27 12:37:49 <rxr> yes, the GSM is closed
Aug 27 12:37:54 <rxr> we all hope for the best
Aug 27 12:38:13 <rxr> stefan_schmidt: well - hardware with a firmware software part
Aug 27 12:38:34 * XorA points people at bug #666
Aug 27 12:38:49 <stefan_schmidt> rxr: Yeah, but getting software right is easier then hardware. :)
Aug 27 12:39:35 <stefan_schmidt> XorA: thanks
Aug 27 12:40:00 <XorA> stefan_schmidt: If only I had access to my previous employers kit, I could probably nail that bug in about 10 mins
Aug 27 12:41:06 <stefan_schmidt> XorA: I know. :) But I'm more pissed of the long waiting times with TI...
Aug 27 12:41:31 <Writchie> I thought FIC had the full dev tools for the calypso
Aug 27 12:41:46 <SpeedEvil> stefan_schmidt: Do you know if this is a 'ship affected phones back to FIC for reflashing' issue, or can the user do it?
Aug 27 12:42:05 <stefan_schmidt> Writchie: XorA refers to some special GSM testing equipment.
Aug 27 12:42:08 <XorA> Writchie: part of the problem is SIMs are not as standard as people are lead to beleive, neither are handsets
Aug 27 12:42:39 <stefan_schmidt> SpeedEvil: No idea. We only can decide this once we have a fix.
Aug 27 12:42:44 <XorA> Writchie: in my previous life I wrote and maintained a SIM, so I got used to the quirks
Aug 27 12:42:59 <Writchie> every handset mfg is used to the quirks
Aug 27 12:43:34 <Writchie> of course this is why phase 1 is necessary
Aug 27 12:43:40 <XorA> Writchie: not true, nokia fight ETSI to get any form of standardised testing, as it shows their bugs in bad light and they would actually have to test before release :-)
Aug 27 12:44:37 <Writchie> by used to the quirks I meant is used to having to test a boatload of sims prior to release
Aug 27 12:44:51 <Writchie> s/release/release of the handset/
Aug 27 12:44:51 <apt> Writchie meant: by used to the quirks I meant is used to having to test a boatload of sims prior to release of the handset
Aug 27 12:48:52 <Writchie> I can't personally test GSM850 yet - has anybody had a working call on an 850GSM network.
Aug 27 12:52:10 <write_erase> Are there GPS navigation opensource software that can be included in openmoko ? 
Aug 27 12:52:25 <write_erase> s/included in/ported to/
Aug 27 12:52:25 <apt> write_erase meant: Are there GPS navigation opensource software that can be ported to openmoko ? 
Aug 27 12:52:56 <CM> write_erase: I guess MaemoMapper could be a good candidate
Aug 27 12:53:04 <CM> Or something OpenStreetmap related
Aug 27 12:53:45 <cb22> There's also a project up on the GForge
Aug 27 12:55:33 <maistk0> www.freepenguin.it/index-en.html
Aug 27 13:02:53 <SpeedEvil> write: basically, the problem is the maps, not the mapper.
Aug 27 13:03:10 <SpeedEvil> wr: over most of the world, there are no freely available maps. And openstreetmap sucks.
Aug 27 13:04:04 <cb22> a nice idea would be to have various backends
Aug 27 13:04:53 <Writchie> Bug 762 filed on 40ma actual default charge rate for GTA01V4
Aug 27 13:04:54 <cdbot2> * * OM Bug 762 has been created by wally.ritchie(AT)gmail.com
Aug 27 13:04:54 <cdbot2> * * GTA01v4 Only Charges at 40ma
Aug 27 13:04:55 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=762
Aug 27 13:07:33 <FuzzyCat>  bitbake gpsdrive
Aug 27 13:07:56 <FuzzyCat> you might want a monster disk though
Aug 27 13:10:26 <SpeedEvil> gpsdrive is basically abusing google though.
Aug 27 13:10:30 <SpeedEvil> AIUI.
Aug 27 13:11:48 <stefan_schmidt> Writchie: thnaks
Aug 27 13:11:48 <El_Salvador> <SpeedEvil> wr: over most of the world, there are no freely available maps. And openstreetmap sucks.
Aug 27 13:11:52 <stefan_schmidt> thanks even
Aug 27 13:11:53 <El_Salvador> why it sucks?
Aug 27 13:13:45 <FuzzyCat> El_Salvador, lack of data
Aug 27 13:14:40 <El_Salvador> ah
Aug 27 13:14:43 <El_Salvador> thansk
Aug 27 13:14:47 <El_Salvador> thanks
Aug 27 13:19:17 <rwhitby> cindy: the reason why the "build-package-foo" target was added to the MokoMakefile was specifically so people didn't need to remember which directory you have to be in when you type bitbake manually.
Aug 27 13:41:22 <aloril> (script) wiki RecentChanges: [[OpenMoko_Local_Groups:_Saint-Petersburg]] [[OpenMoko_Local_Groups:]] [[OpenMoko_Local_Groups]] [[Hardware:Neo1973:Alternate_Cases:Camera]] [[USB_Networking]]
Aug 27 13:50:42 <smallfoot-> does openmoko is only on Neo1973?
Aug 27 13:50:43 <aloril> smallfoot-: See http://wiki.openmoko.org/wiki/FAQ (and also SH1, SH1_FAQ, Neo1973, Developer_preview and Neo1973_Hardware#GTA01Bv4 pages) (Hopefully these links answers your question. This autotrigger for new nick and question during first 15 minutes done only once.)
Aug 27 13:51:13 <woglinde> smallfoot hm there are same gys working on it for htc and ezx
Aug 27 13:51:26 <smallfoot-> whats htc and ecx?
Aug 27 13:51:42 <woglinde> ezx are the linux motorolas
Aug 27 13:51:43 <CM> I guess the main problem is that OpenMoko has a 480x640 screen
Aug 27 13:51:48 <woglinde> and htc is htc
Aug 27 13:51:53 <smallfoot-> will Nokia, Samsung, LG, Sony-Ericsson, Motorola, or anyone else support/use OpenMoko?
Aug 27 13:52:01 <smallfoot-> oh
Aug 27 13:52:25 <CM> smallfoot-: Probably never, but that doesn't mean that they can not run OM some day
Aug 27 13:52:32 <woglinde> smallfoot which of them runs linux und you has access too?
Aug 27 13:52:48 <cjb_ie> motorola had a linux phone didn't they?
Aug 27 13:53:06 <CM> They have several
Aug 27 13:53:24 <CM> But they're not exactly "open"
Aug 27 13:53:42 <cjb_ie> pfeh, with a good enough logic analyser... :-)
Aug 27 13:53:52 <CM> Hehe
Aug 27 13:53:53 <smallfoot-> woglinde, i dont have access to any, i dont even own a cellphone :(
Aug 27 13:54:25 <woglinde> smallfoot oh hm
Aug 27 13:56:13 <aloril> autofaq did answer question ;-)
Aug 27 13:56:28 <zash_se> smallfoot-: there are rumors that google is interested in openmoko for their gphone
Aug 27 13:56:34 <smallfoot-> sweet
Aug 27 13:56:36 <smallfoot-> gphone? wow
Aug 27 13:58:29 <rxr> aren't rumours that they release something in the next weeks?
Aug 27 13:58:45 <rxr> someone at openmoko should know in that case already if Google ships the Neo or not ...
Aug 27 13:58:48 * CM would like to see some facts surrounding those rumours
Aug 27 13:59:02 <Writchie> rxr: that's what I hear
Aug 27 13:59:20 <woglinde> its only rumors
Aug 27 13:59:23 <Writchie> and openmoko is no where near two weeks
Aug 27 13:59:25 <woglinde> I cant belevie it
Aug 27 13:59:38 <rxr> yeah - but as OpenMoko is onst moe the most secret Inc.'s around we probably will learn here on the day Google announces it :-)
Aug 27 13:59:49 <CM> And I wonder why google would want to build a phone? They're not a hw-company
Aug 27 13:59:49 <aloril> well, openmoko will announce something, hmm ... actually they already did: 2007.2 (2007.8 more recently I think)
Aug 27 13:59:50 <Writchie> other rumors are that google has spent $100 million on this
Aug 27 14:00:07 <CM> The only hw they sell is that yellow search box, and it's actually a dell machine
Aug 27 14:00:30 <rxr> Writchie: yes, ack. I have the "feeling" their thing is not about OpenMoko ...
Aug 27 14:00:36 <rxr> but that is just a feeling
Aug 27 14:00:39 <woglinde> CM lol
Aug 27 14:00:54 <woglinde> CM they pay 500.000 bucks at htc and will have theire hw
Aug 27 14:00:59 <cjb_ie> omg it's teh google iphone-killer!
Aug 27 14:01:05 <rxr> the same feeling that tells me that I have to live with this damn high pitch noisy crappy MacBook for quite a while until Apple releases some ultra portable that is rumoured since last december ...
Aug 27 14:01:07 <CM> But I do think they'll be releasing more mobile apps, like the gmail client
Aug 27 14:01:09 <rxr> sigh
Aug 27 14:01:14 <CM> But some map thing or so
Aug 27 14:01:46 <CM> rxr: I'd like that one too. I guess they're waiting for the price of SSD's to drop a bit more
Aug 27 14:02:00 <juri_> wow. looks like my openmoko arrives today.
Aug 27 14:02:11 <CM> :)
Aug 27 14:02:28 * CM points juri_ to http://wiki.openmoko.org/wiki/P1_Owners
Aug 27 14:02:43 <rxr> CM: the MacBook rumour ?
Aug 27 14:02:50 <CM> rxr: Yes
Aug 27 14:02:52 <rxr> CM: this damn MacBook I type on is one of the worst hardware ever ...
Aug 27 14:02:54 <rxr> damn crap ...
Aug 27 14:02:59 <rxr> and I had iBook for ages ...
Aug 27 14:03:02 <juri_> its not here yet. i'm waiting for ups to decide itn eeds to go to south africa, now. ;P
Aug 27 14:03:17 <rxr> (if I would not had to code on one OS X App I would not even had this stinking crap of a plastic ...)
Aug 27 14:03:27 <woglinde> hms where is the opera vs iphone flash gone
Aug 27 14:03:33 <CM> rxr: My other option is the XPS1330 dell one
Aug 27 14:03:45 <rxr> well - that is what you get when you are vendor locked-in ..
Aug 27 14:05:04 <woglinde> ah http://www.operamini.com/beta/video/
Aug 27 14:30:53 <CM> Just noticed that cups is built when using the makefile.. Bluetooth or maybe even usb printing of pdf's and such would be nice
Aug 27 14:32:09 <SpeedEvil> Indeed.
Aug 27 14:32:12 <SpeedEvil> Or fax
Aug 27 14:32:18 <SpeedEvil> If supported
Aug 27 14:32:29 <zecke> CM: cups is compiled for the Gtk+ cups printing backend
Aug 27 14:32:38 <zecke> CM: bluez gets compiled with cups support as well
Aug 27 14:32:47 <CM> zecke: Nice :D
Aug 27 14:32:57 <zecke> (not installed by default though)
Aug 27 14:33:06 <CM> Just have to make it easy to configure
Aug 27 14:33:49 <CM> Would definitely look funnt to plug in my big usb laser printer into the neo :)
Aug 27 14:33:57 <CM> s/nnt/nny/
Aug 27 14:33:58 <apt> CM meant: Would definitely look funny to plug in my big usb laser printer into the neo :)
Aug 27 14:49:01 <hhf423> re
Aug 27 14:51:09 <riot> zecke: install some python-stuff, you'll likely have cups in some of the dependencies (which is sort of a bug..)
Aug 27 14:52:15 <zecke> riot: if you identify the package, please report that in the oe bugtracker
Aug 27 14:57:10 <riot> zecke: next time, i install the stuff, i'm gonna watch out :)
Aug 27 14:57:44 <mdt> riot, are you shure it's a dependency? it may be a recommended package also....
Aug 27 14:59:10 * FuzzyCat really would like to hang whoever thought that boot sound was a good idea
Aug 27 14:59:29 * mdt agrees
Aug 27 14:59:32 <cjb_ie> there's a fix for that if you scroll up :)
Aug 27 14:59:35 * cb22 seconds
Aug 27 14:59:44 <mdt> is there a easy way to shut up the neo?
Aug 27 14:59:47 <FuzzyCat> i know the fix
Aug 27 14:59:56 <FuzzyCat> you just comment it out of the server.conf
Aug 27 14:59:57 <cb22> its on the wiki IIRC
Aug 27 15:00:02 <FuzzyCat> I just wish it wasn;t there 
Aug 27 15:00:09 <FuzzyCat> I've just flashed it
Aug 27 15:00:16 <FuzzyCat> hence  - annoying noise
Aug 27 15:01:02 <aloril_> (script) wiki RecentChanges: [[Buying_Interest_List]] [[Hardware:Neo1973:Alternate_Cases:Camera]]
Aug 27 15:03:57 <FuzzyCat> ok, it's called session now, rather than server.conf
Aug 27 15:05:42 <mdt> FuzzyCat, is there a package to deinstall? i deinstalled psplash to get rid of the splash screen
Aug 27 15:08:27 <hhf423> or just rename the sound file in /usr/share/openmoko/sounds or so
Aug 27 15:08:31 <FuzzyCat> mdt, just edit /etc/pulse/session
Aug 27 15:08:35 <FuzzyCat> at the bottom
Aug 27 15:08:43 <FuzzyCat> comment out the sound lines
Aug 27 15:09:28 <ScaredyCat> http://wiki.openmoko.org/wiki/Boot_sound
Aug 27 15:09:34 <FuzzyCat> as per that link
Aug 27 15:12:08 <hhf423> if I want ipkg upgrade to log into a file, how do I do that?
Aug 27 15:12:37 <FuzzyCat> ipkg upgrade >mylogfile.log ?
Aug 27 15:13:00 <hhf423> > or >>?
Aug 27 15:13:21 <daxxar> > if you want to overwrite mylogfile.log
Aug 27 15:13:24 <daxxar> >> if you want to append
Aug 27 15:13:41 <FuzzyCat> yes
Aug 27 15:13:53 <hhf423> ah, right, thanks
Aug 27 15:14:22 <hhf423> oh
Aug 27 15:14:28 <hhf423> "Nothing to be done"
Aug 27 15:14:45 <FuzzyCat> did you update first?
Aug 27 15:15:02 <hhf423> there I go. No changes in the last 12 hours, the community is slowing down?
Aug 27 15:15:11 <FuzzyCat> ipkg update && ipkg upgrade
Aug 27 15:15:23 <hhf423> yes, I did the update first
Aug 27 15:15:37 <FuzzyCat> if you're bored I added asterisk to my repository... and vnc :)
Aug 27 15:15:55 <hhf423> asterisk?!1!
Aug 27 15:16:04 <hhf423> you _are_ crazy
Aug 27 15:16:12 <hhf423> but I like that in a cat :-)
Aug 27 15:16:31 * cjb_ie has three cats of whom at least two are certifiable
Aug 27 15:17:10 <FuzzyCat> i can has cheezburger?
Aug 27 15:17:53 <hhf423> there is strictly nothing to be done. And I soo hoped BT might work better today
Aug 27 15:19:34 <hhf423> next saturday I am off on vacation and I want proper BT to work!
Aug 27 15:20:03 <rxr> haha
Aug 27 15:20:39 <hhf423> that would make me very happy
Aug 27 15:28:41 <riot> re: startsound : I'd just like some less annoying sounds. I think i'm gonna build some new audio-theme this week
Aug 27 15:29:03 <riot> i'm thinking of some industrial/minimalized stuff, like old mobile phone-beepers
Aug 27 15:29:33 <riot> gotta play around with my tb303 emulation :)
Aug 27 15:29:42 <don-o> how is the making/taking calls functionality coming? i figure a month worth of having P1s in the field must be helping things.
Aug 27 15:30:03 <hhf423> don-o: see P1 software testing
Aug 27 15:30:12 <hhf423> don-o: getting better but slowly
Aug 27 15:31:18 <don-o> hhf423: any news on P2?
Aug 27 15:32:14 <don-o> hhf423: found the P1 software testing wiki page, thanks for the tip.
Aug 27 15:33:19 <hhf423> don-o: nothing I know on P2
Aug 27 15:36:31 <PBeck> hi
Aug 27 15:39:52 <hhf423> hi
Aug 27 15:41:16 * don-o catches up on the mailing list
Aug 27 15:49:07 <cayco> hi guys! does anyone managed to get gsmd to work in latest 2007.2 code?
Aug 27 15:51:17 <don-o> Harold W. says "GTA02 phase-1 will probably be october."
Aug 27 15:51:28 <ewon> impressive
Aug 27 15:51:37 <don-o> from an email 8 days ago
Aug 27 15:51:38 <ewon> I had the feeling it was going to be delayed
Aug 27 15:51:53 <don-o> ewon: yeah thats encouraging, although i dont know what 'phase-1' means
Aug 27 15:52:00 <don-o> (in that context)
Aug 27 15:52:02 <ewon> hw will be ready, but will the software be solid enough for consumer release?
Aug 27 15:52:02 <Tyrael> cayco: it "works"
Aug 27 15:52:11 <hhf423> cayco: sort of, see Phase 1 software testing in the wiki
Aug 27 15:53:00 <zblach> hey all
Aug 27 15:53:05 <cayco> Tyrael: hmm, it does not to me and my friend (who also has neo): http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=761
Aug 27 15:53:26 <don-o> ewon: http://lists.openmoko.org/pipermail/community/2007-August/009491.html
Aug 27 15:53:50 <Tyrael> cayco: tried sto stop it, x and restart?
Aug 27 15:54:03 <cayco> Tyrael: of course :)
Aug 27 15:54:25 <Tyrael> what output lsmod?
Aug 27 15:54:51 <ewon> don-o: ah :)
Aug 27 15:55:00 <cayco> Tyrael: now, when i read this i think it could be related to my fix in another ticket: http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=760
Aug 27 15:55:14 <cayco> Tyrael: i just assumed that it does not work for everyone
Aug 27 15:55:28 <cayco> Tyrael: let me check this and I get back :)
Aug 27 15:55:36 <Tyrael> lsmod?
Aug 27 15:55:36 <ewon> hmm, I'll buy a GTA02 when it comes out, hopefully I can contribute, although I'm NDA'd to my eyes on a lot of network stuff
Aug 27 15:55:55 * cayco left neo at work ;)
Aug 27 15:56:07 <Tyrael> ah
Aug 27 15:56:30 <cayco> ok, thx for confirmation that this realy works ;)
Aug 27 15:56:38 <don-o> reading the software testing page, its pretty funny to see people excited that outbound calls work some of the time.
Aug 27 15:57:22 <hhf423> don-o: start learning to be happy about the small things in live. :-)
Aug 27 16:04:26 <FuzzyCat> mmm. I still have that sd card i/o bug...
Aug 27 16:10:55 <torpor> hi all
Aug 27 16:11:26 <Tyrael> I have a laptop with bluetooth. There is a mode to emulate a bluetooth keyboard to test it with the Neo?
Aug 27 16:11:58 <hhf423> Tyrael: not heard of that
Aug 27 16:13:19 <don-o> Tyrael: that would be cool.
Aug 27 16:13:27 <Tyrael> yep
Aug 27 16:15:46 <hhf423> why? If I can link from my Neo to my PC/Laptop, I can also ssh in and use the keyboard
Aug 27 16:16:53 <Tyrael> but you can't use it for input on the gui application
Aug 27 16:20:41 <pjz> ...unless you run something like a vncserver on the Neo and a client on your PC/Laptop
Aug 27 16:32:20 <Elrond> Hmm, the openmoko bugzilla doesn't allow voting, right?
Aug 27 16:33:50 <SpeedEvil> Not formally, I think.
Aug 27 16:33:51 <SpeedEvil> Why?
Aug 27 16:35:16 <Elrond> Well, I have a few bugs I'd like to say "Yes, I care for it". But I don't want to start out stupid "me three" one comments.
Aug 27 16:35:31 <Elrond> s/one/one-line
Aug 27 16:37:23 <cb22> hmm, was there a bot that changed sed style expression in here?
Aug 27 16:37:27 <cb22> s/was/wasnt/
Aug 27 16:37:27 <apt> cb22 meant: hmm, wasnt there a bot that changed sed style expression in here?
Aug 27 16:39:20 <Tyrael> I found something for the keyboard: http://www.mulliner.org/bluetooth/xkbdbthid.php
Aug 27 16:42:21 <borg_> Tyrael: i think the keys are to small
Aug 27 16:44:12 * hhf423 yawns
Aug 27 16:44:14 <mbuf> which versions of u-boot, openmoko-devel-image and uImage do I need for OM2007.2? http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/images/fic-gta01/
Aug 27 16:44:55 <hhf423> mbuf: see files here: http://wiki.openmoko.org/wiki/Phase_1_Software_Testing#Snapshot_20070825_with_2.6.21.6-moko11-r2
Aug 27 16:45:55 <Elrond> chrustinho - Yes, there was. But I choose to not use it. :)
Aug 27 16:45:57 <mbuf> hhf423, no need of u-boot ?
Aug 27 16:46:34 <hhf423> mbuf: no, u-boot does not necessarily be changed for 2007.2
Aug 27 16:47:16 <mbuf> hhf423, but, there was mention of some bug that was fixed?
Aug 27 16:47:49 <hhf423> mbuf: yes, the bug that created issues when you were flashing a smaller image than was on the phone before
Aug 27 16:48:08 <hhf423> mbuf: but that has nothing to do with mbuf
Aug 27 16:48:17 <hhf423> err, with 2007.2
Aug 27 16:48:36 <jroman> hi!
Aug 27 16:48:40 <hhf423> I flashed the u-boot here to get rid of the bug: http://buildhost.openmoko.org/snapshots/2007.08/images/
Aug 27 16:49:14 <mbuf> hhf423, thanks
Aug 27 16:49:33 <hhf423> mbuf: as you can brick your neo with a wrong uboot, I'd not do that if you dont feel like adventures :-)
Aug 27 16:49:52 <Elrond> and unless you have a debug board.
Aug 27 16:49:55 <torpor> is that URL the source of the latest builds, btw?
Aug 27 16:50:08 <mbuf> Elrond, i have a debug board, haven't had time to play with it,  yet
Aug 27 16:50:21 <Elrond> mbuf - Make sure you know to use it. :)
Aug 27 16:50:23 <Tyrael> borg_: not for pen on rotated screen, but I search this for my laptop to test neo bluetooth
Aug 27 16:50:34 <hhf423> torpor: no, that is not the url with the latest builds
Aug 27 16:50:44 <torpor> oh. anyone know aht it is?
Aug 27 16:50:44 <mbuf> Elrond, as long as you guys are here, i don't have to worry :)
Aug 27 16:51:02 <hhf423> torpor here: http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/images/fic-gta01/?C=M;O=D
Aug 27 16:51:36 <hhf423> torpor: and a bit more current but a bit more unofficial is here: http://ipkg.nslu2-linux.org/feeds/openmoko/images/?C=M;O=D
Aug 27 16:51:41 <Elrond> mbuf - hehe. I never used the debugboard.
Aug 27 16:52:49 <jroman> I've deleted moko/openembedded/packages/qemu/ directory, how can I restore it?
Aug 27 16:53:29 <Elrond> SpeedEvil - I decided for "CC"-Voting. Added myself to the CC and thus said "I care for this bug". ;-)
Aug 27 16:54:29 <torpor> hhf423: cool thanks mate.
Aug 27 16:54:54 <torpor> of course i'm building my own localimages too .. but i want to try the official ones and compare
Aug 27 16:55:58 * k-s[WORK] is away: lunch
Aug 27 16:59:36 <jroman> ok, mtn revert --missing
Aug 27 17:00:18 <mbuf> what could be the reason for determining device status: state = dfuERROR, satus = 14, dfuERROR, clearing state
Aug 27 17:00:38 <mbuf> using dfu-util for uImage-2.6.21.6-moko11-r2-fic-gta01.bin on debian etch
Aug 27 17:00:47 <mbuf> dmesg shows detection of CDC_ACM device
Aug 27 17:01:41 <aloril> (script) wiki RecentChanges: [[Phase_1_Software_Testing]] [[Supported_microSD_cards]] [[Talk:Manually_using_Bluetooth]] [[Buying_Interest_List]] [[Boot_sound]]
Aug 27 17:03:40 <hhf423> mbuf did you sudo ./dfu-util ?
Aug 27 17:03:40 <jseghers> Hi all.  I put a T-mobile SIM into a 2007.2 build.  It worked fine initially.  Both outgoing and incoming calls worked.  Then it stopped.  I've ssh'd to the phone and restarted X so that I have stdout and stderr coming to the ssh connection.  I'm getting "Requesting registration" about every 4 seconds.  I tried stopping gsmd, but the "Requesting registration" kept going.  Anyone have an idea what it the problem here?
Aug 27 17:04:25 <Aria> ditto jseghers question here ;-)
Aug 27 17:04:42 <mbuf> hhf423: no, disconnected, plugged it back, it worked
Aug 27 17:04:52 <borg_> mtn: warning: attach node 2147500159 blocked by unversioned path 'packages/openssl/openssl-0.9.7g/armeb.patch.lock'
Aug 27 17:04:55 <borg_> uhm
Aug 27 17:05:07 <borg_> does someone have this too?
Aug 27 17:05:15 <torpor> nope
Aug 27 17:05:19 <borg_> mh
Aug 27 17:05:26 <Aria> bet the lock file needs removing
Aug 27 17:05:32 <mdt> jserv2, look into /tmp/gsmd.log - probably u will se a modem dead message there....
Aug 27 17:05:33 <borg_> perhaps i should delete openssl
Aug 27 17:06:35 <mdt> s/jserv/jseghers/
Aug 27 17:06:36 <apt> mdt meant: jseghers2, look into /tmp/gsmd.log - probably u will se a modem dead message there....
Aug 27 17:07:12 <hp__> the responsiveness in the new build on qemu is allot worse now :(
Aug 27 17:07:34 <jseghers> mdt: I don't see anything there indicating a dead modem
Aug 27 17:08:43 <mdt> jseghers, but what do u seee there?
Aug 27 17:08:57 <jseghers> mdt: http://pastebin.com/d62560f0e is the complete contents of that log
Aug 27 17:10:03 <hp__> i must have missed something
Aug 27 17:10:15 <hp__> qemu won't even respond to my mouse clicks
Aug 27 17:10:21 * hhf423 flashes
Aug 27 17:10:39 <mbuf> hp__, same here
Aug 27 17:12:02 <mdt> jseghers, hm, you stopped it - did you start it again also?
Aug 27 17:12:17 <jseghers> yes, it had no effect.
Aug 27 17:12:38 <jseghers> I just stopped X again, and it stopped the polling
Aug 27 17:13:32 <mbuf> root.jffs2 takes long time to flash for 2007.2 :)
Aug 27 17:13:56 <jseghers> I noticed I had two sessions of the /usr/bin/dbus-deamon --fork.... processes running--these get launched by X's startup--so I'm killing those.
Aug 27 17:14:33 <jseghers> There are also currently 2 processes of dropbear running, but I don't know what dropbear is...
Aug 27 17:15:08 <hp__> jseghers.. google dropbear :)
Aug 27 17:15:31 <jseghers> hp: yep, jsut did... That would explain it :)
Aug 27 17:15:59 <hhf423> mbuf: 40meg over usb1.1
Aug 27 17:16:46 <edistar> I like dropbear
Aug 27 17:16:58 <edistar> better than the openssh-server on embedded
Aug 27 17:16:58 <mbuf> hhf423, the console boot messages used to show what was happening, the progress bar isn't very informative :)
Aug 27 17:18:22 <hhf423> mbuf: you can have it pretty or you can have it informative :-)
Aug 27 17:26:09 * hhf423 reboots
Aug 27 17:27:41 * hhf423 waits while teh UI starts up
Aug 27 17:27:56 * hhf423 enters PIN
Aug 27 17:29:36 * hhf423 makes a call to the neo and from the neo ok
Aug 27 17:41:25 <torpor> guys - how can i know what the 'real lib names are in openmoko, for inclusion in a configure.ac file?
Aug 27 17:41:47 <torpor> is there some empirical way of determining these words?
Aug 27 17:44:01 <borg_> root@fic-gta01:/$ alsactl -f /etc/stereoout.state restore
Aug 27 17:44:01 <borg_> alsactl: load_state:1327: No soundcards found...
Aug 27 17:44:07 <borg_> aaah :D
Aug 27 17:44:09 <borg_> whats this?
Aug 27 17:44:15 <torpor> kernel does not match modules onboard
Aug 27 17:44:29 <borg_> so, no sound at the moment?
Aug 27 17:44:41 <torpor> you upgraded something that puts new modules onboard and you do'nat have a compatible kernel
Aug 27 17:44:46 <torpor> upgrade to the kernel image needed.
Aug 27 17:44:57 <borg_> mh
Aug 27 17:47:19 <Writchie> SpeedEvil: Are you convinced the charger is working correctly. I had a Neo go dead over the weekend (wouldn't boot). Then I charged it up for about 6 hours at 40ma and it went to 0ma. Now I boot and it draws 490ma. This suggest (at least to me) that the 40ma charge doesn't actually charge up the battery. 
Aug 27 17:55:12 <borg_> torpor: thank you! :)
Aug 27 17:55:24 <borg_> 2.6.21.6 works nice
Aug 27 17:55:28 <borg_> 2.6.21.5 didnt
Aug 27 17:57:27 <torpor> glad you got it sorted
Aug 27 17:59:13 * k-s[AWAY_WORK] is back.
Aug 27 18:01:10 <Writchie> Anybody gotton the neo RNDIS to work with Windows XP or Vista ?
Aug 27 18:05:44 <hp__> RNDIS?
Aug 27 18:06:56 <Writchie> hp__ Remote NDIS - aka USB Networking
Aug 27 18:11:22 <hp__> aahh no idea.. did you search the wiki?
Aug 27 18:13:18 <roh> Writchie i think the standart linux usb cdc has that feature and its activated in out builds... i didn't test i though
Aug 27 18:14:26 <Writchie> roh: the inf file is nowhere near matching what the USB gadget (as modified) reports over the USB
Aug 27 18:14:59 <Writchie> roh: even after hacking this to match - windows fails to start the device with error 10
Aug 27 18:14:59 <roh> i see..  where did you get it from?
Aug 27 18:15:27 <Writchie> the linux.inf is part of the sources
Aug 27 18:15:56 <Writchie> but it has the defaults used by the USB gadget folks
Aug 27 18:16:42 <Writchie> just looking to see if anyone has ANY RNDIS gadget working on XP or Vista before I start spending time on this/
Aug 27 18:17:02 <roh> i don't know of any modifications on that part of the kernel. i think till now it worked fine for anybody who did cdc.
Aug 27 18:17:38 <roh> Writchie when there is a problem, its propably also upstream. which does not mean we can or should not fix it ;)
Aug 27 18:18:22 <Writchie> well it surely wasn't upstream that changed the vendor code to FIC and device codes
Aug 27 18:18:25 <zecke> *yawn*
Aug 27 18:18:40 <roh> Writchie ack. but thats a trivial one and shurely not the bug
Aug 27 18:22:22 <serenity> hi
Aug 27 18:22:49 <serenity> is there a list of command which a special in openmoko (ts_calibrate and so on)
Aug 27 18:23:47 <jseghers> Writchie: I got RNDIS installed on XP...but it only suffices to engage the 500ma charge :)
Aug 27 18:24:18 <jseghers> Writchie: could not get actual communications to occur.
Aug 27 18:24:36 <Writchie> jseghers: did windows attempt to install drivers?
Aug 27 18:25:02 <jseghers> Writchie: yes, installed, but failed... 
Aug 27 18:25:25 <Writchie> jseghers: the neo is probably satisfied enough that it is enumerating to let the charge go
Aug 27 18:25:29 <jseghers> Like I said, it was enough that it is able to negotiate high speed charge, but nothing mor
Aug 27 18:25:34 <jseghers> yep
Aug 27 18:25:35 <Writchie> this is actually a bug
Aug 27 18:25:53 <Writchie> maybe
Aug 27 18:26:20 <jseghers> Writchie: yes, it is.  It also seems to be a pissing contest between the RNDIS people and Microsoft.
Aug 27 18:26:45 <Writchie> how so 
Aug 27 18:26:46 <serenity> which broswer will be used in openmoko?
Aug 27 18:27:38 <zecke> openmoko browser
Aug 27 18:27:42 <jseghers> Writchie: from what I've read online, it seems the RNDIS people say they've implemented the RFC correctly, so it's up to Microsoft to fix their driver template.
Aug 27 18:28:04 <jseghers> Writchie: and apparently Microsoft hasn't seen fit to do so.
Aug 27 18:28:21 <Writchie> well MS is just following the golden rule - he who has the gold makes the rules
Aug 27 18:28:45 <Writchie> i'm not sure its just a template problem
Aug 27 18:28:45 <serenity> zecke: and openmoko browser is based on what?
Aug 27 18:29:10 <jseghers> I think it's going to take someone with USB driver experience to figure out what's really going on on both ends.
Aug 27 18:30:01 <Writchie> this is why I wanted to know what I'm getting into. Better to know it doesn't work first
Aug 27 18:30:08 <zecke> serenity: gtk+, a hand full of magic and an existing browser engine
Aug 27 18:30:14 <serenity> zecke: ;)
Aug 27 18:30:18 <hhf423> Writchie: my experience with windows xp and rndis is in the wiki under discussion
Aug 27 18:30:19 <Writchie> it seemed the gumstix guys may have had this working
Aug 27 18:30:30 * serenity is looking for the openmoko fairy
Aug 27 18:30:53 <hhf423> Writchie: http://wiki.openmoko.org/wiki/Talk:Neo1973_and_Windows
Aug 27 18:31:06 <zecke> serenity: If I'm the one to decide I would use WebKit. But one can think of the EAL approach (it is nokia's gobject/gtk browser abstraction layer)
Aug 27 18:31:31 <hhf423> Writchie: I did not ry all on thsi page, tho: http://handhelds.org/moin/moin.cgi/WindowsXpUsbNetworkHowTo
Aug 27 18:31:45 <hhf423> I did not try all on this page: http://handhelds.org/moin/moin.cgi/WindowsXpUsbNetworkHowTo
Aug 27 18:31:56 <serenity> zecke: me to, but maybe there will be a webkit Browser ported one day
Aug 27 18:32:12 <Writchie> hhf423: I remember seeing your comment on the wiki. Which inf file did you use.
Aug 27 18:32:26 <hhf423> Writchie: the one referenced in the wiki
Aug 27 18:32:33 <zecke> serenity: or we will just go the webkit way...
Aug 27 18:32:44 <serenity> zecke: ?
Aug 27 18:32:57 <Writchie> linux.inf from where?
Aug 27 18:33:13 <hhf423> Writchie: http://privat.wmo.de/~c_schweers/linux.inf
Aug 27 18:33:32 <zecke> serenity: well, Firefox on Linux/Gtk+ is a second class citizen, the mozilla API is unstable so it feels natural to use something where you are the first class citizen
Aug 27 18:33:52 <serenity> zecke: :-D
Aug 27 18:34:29 <hhf423> Writchie: I was actually more counting on bluetooth connectivity. but this also does not work at the moment, not with windows at least
Aug 27 18:34:35 <zecke> serenity: e.g. my feedreader is already using webkit, so it might be wise to use it for a browser as well, instead of adding mozilla to the stack
Aug 27 18:34:45 <Writchie> hhf423: thanks. I'll compare to mine. 
Aug 27 18:34:55 <Writchie> I'm just begining to test BT
Aug 27 18:35:10 <Writchie> what hardware did you have trouble with 
Aug 27 18:35:17 <Writchie> on the XP/Vista side
Aug 27 18:35:40 <hhf423> Writchie: see my comments on BT here: http://wiki.openmoko.org/wiki/Phase_1_Software_Testing#Bluetooth
Aug 27 18:36:34 <Writchie> OK. I was able to transfer to P800 and nokia phones and between neo and ubuntu no problem.
Aug 27 18:36:37 <hhf423> Writchie: BT on my windows box works fine with a number of devices. The Neo will not hcitool cc with anything for longer than a few seconds, though. Not with my Windows, not with my SE phone etc. pp
Aug 27 18:37:03 <hhf423> Writchie: oh, really? Neo BT and 2007.2?
Aug 27 18:37:20 <Writchie> i did the testing on 2007.1
Aug 27 18:37:35 <hhf423> Writchie: ah, see. that could be it. 
Aug 27 18:37:52 <Writchie> right now i have one 2007.2 and 2007.1 and have been reluctant to give up the 2007.1 until I get to at least the same point on gsm and bt
Aug 27 18:38:06 <hhf423> Writchie: from the wiki I got the idea that BT actually worked in 2007.1, but does no longer in 2007.2
Aug 27 18:38:38 <Writchie> it definitely worked for obex and there were reports of PAN working as well
Aug 27 18:39:01 <Writchie> so i guess i'll keep one device on 2007.1 for a while longer
Aug 27 18:39:05 <hhf423> Writchie: yes, thats what it looks like if you read the wiki.
Aug 27 18:39:13 <mjr> PAN worked for me in 2007.1, in the last .2 image I tried it (for a BT keyboard) it worked, but I was required to load modules manually
Aug 27 18:39:27 <hhf423> mjr: which modules did you have to load?
Aug 27 18:39:54 <mjr> don't recall exactly, ls -R /lib/modules, see about the usb and BT support
Aug 27 18:39:55 <hhf423> Writchie, mjr what seems to be necessary is the .6 kernel
Aug 27 18:40:30 <hhf423> with the .6 kernel you can see the BT interface e.g. using hciconfig. with the .5 kernel no luck
Aug 27 18:42:28 <Writchie> i did notice 2007.2 seemed to go backward to 0.5 - what was that all about?
Aug 27 18:42:56 <hhf423> no idea. all seems to work fine with .6 
Aug 27 18:43:18 <hhf423> the only difference I can see between 0.6 and .5 is that with .6, bt interface is there
Aug 27 18:45:33 <hhf423> the default auth PIN for BT is 0000, right?
Aug 27 18:45:43 <Writchie> usually
Aug 27 18:45:52 <Writchie> once in while 1234
Aug 27 18:46:22 <hhf423> well, my SE phone always says that the bt pin I entered was not the same as on the Neo
Aug 27 18:46:30 <hhf423> and I tried 0000 and 1234
Aug 27 18:47:00 <ScaredyCat> hhf423: p910i ?
Aug 27 18:47:15 <hhf423> k610i
Aug 27 18:48:16 <ScaredyCat> k, I get the same problem wiht my p910i and nokia 770 - apparently it's because SE can't actually make proper phones.
Aug 27 18:48:35 * ScaredyCat points at 990i
Aug 27 18:48:59 <hhf423> ScaredyCat: awww, SE was the first one who properly implemented BT. Nokia never got it right
Aug 27 18:49:29 <ScaredyCat> well, it doesn't work on a lot of their phones now...
Aug 27 18:50:08 <ScaredyCat> and of course the problem is, even if nokia got it wrong, they need to be compatible with nokia... 
Aug 27 18:50:21 <ScaredyCat> you just have to look at windows to see that
Aug 27 18:50:26 <hhf423> all I know is that so far I could pair my k610i and n770 with anything else I have and I can not with the Neo, so what do you think I think is broken? :-)
Aug 27 18:50:35 <ScaredyCat> brb making a cuppa
Aug 27 18:51:33 <hhf423> you all used the passkey-agent to set a bt passkey? Writchie, ScaredyCat ?
Aug 27 18:52:18 <Writchie> hhf423: what is in your /etc/bluetooth/pin file
Aug 27 18:53:10 <hhf423> root@fic-gta01:~$ cat /etc/bluetooth/pin
Aug 27 18:53:12 <hhf423> cat: /etc/bluetooth/pin: No such file or directory
Aug 27 18:53:46 <Writchie> you may need either this or a working pin helper
Aug 27 18:54:10 <hhf423> what should be in the pin file? just like 0000?
Aug 27 18:54:54 <ScaredyCat> hhf423: i just set it to 1234
Aug 27 18:55:30 <hhf423> awwwwwwwwwwwwwwwwwwwww
Aug 27 18:55:40 <hhf423> there is a hciconf
Aug 27 18:55:46 <hhf423>  # Default PIN code for incoming connections
Aug 27 18:55:47 <hhf423>         passkey "BlueZ";
Aug 27 18:56:02 <ScaredyCat> yeah, that's just dumb
Aug 27 18:56:24 <ScaredyCat> since I don;t know a phone that'll allow text ass a p/w
Aug 27 18:56:51 <borg_> root@fic-gta01:/media$ mount card
Aug 27 18:56:51 <borg_> /dev/mmcblk0p1: Input/output error
Aug 27 18:56:53 <borg_> ooops
Aug 27 18:57:17 <borg_> i think i broke my sd card
Aug 27 18:57:20 <borg_> :(
Aug 27 18:57:29 <ScaredyCat> rebewted?
Aug 27 18:57:40 <ScaredyCat> of rmmod/modprobe
Aug 27 18:57:43 <ScaredyCat> or
Aug 27 18:57:45 <hhf423> ok, next try to link
Aug 27 18:57:53 <ozarka> borg_, just buy a new Neo.  You get a free SD card with it.
Aug 27 18:57:57 <borg_> :D
Aug 27 18:58:26 <hhf423> hahahaha
Aug 27 18:58:31 <hhf423> that worked!1!
Aug 27 18:58:41 <borg_> ScaredyCat: it got input/output error while i tranfered a video to it
Aug 27 18:58:44 <borg_> then i unmounted it
Aug 27 18:58:50 <borg_> and now i cant mount it :)
Aug 27 18:58:55 <ScaredyCat> borg_: from another pc?
Aug 27 18:58:59 <ScaredyCat> ie scp'd it?
Aug 27 18:59:01 <borg_> yes
Aug 27 18:59:06 <ScaredyCat> that's normal
Aug 27 18:59:14 <borg_> mh
Aug 27 18:59:15 <borg_> why?
Aug 27 18:59:25 <borg_> whats the right way to do this?
Aug 27 18:59:33 <Writchie> hhf423: from pin in config file or /etc/bluetooth/pin ?
Aug 27 18:59:47 <ScaredyCat> he driver sucks badger cock
Aug 27 18:59:48 <ScaredyCat> the
Aug 27 18:59:48 <ScaredyCat> there isn't - it's broken
Aug 27 18:59:58 <hhf423> Writchie: one or the other
Aug 27 19:00:00 <ScaredyCat> a reboot or rmmod/modprobe should fix it
Aug 27 19:00:08 <hhf423> Writchie: let me kill the pin file and see
Aug 27 19:00:09 <Writchie> ah - you did both ;)
Aug 27 19:00:26 <borg_> ScaredyCat: i dont need to mount it when i cant transfer something to it :)
Aug 27 19:01:00 <ScaredyCat> use an externl usb dev on your pc
Aug 27 19:01:06 <eggsy> is anyone else having trouble building uicmoc4-native-4.3.1: undefined reference to QAbstractSocket::flush()
Aug 27 19:01:23 <CM> eggsy: Yes, me too
Aug 27 19:01:26 <Writchie> hhf423: it looks like you could also use a one line script as a pin helper with    echo "PIN:1234"  
Aug 27 19:01:29 <CM> !ombug 747
Aug 27 19:01:30 <cdbot2> * * Bug 747, Status: NEW, Created: Unknown
Aug 27 19:01:31 <cdbot2> * * yoyo(AT)ghost.pl: uicmoc4-native_4.3.1.bb
Aug 27 19:01:32 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=747
Aug 27 19:01:42 <aloril> (script) wiki RecentChanges: [[OpenMoko_Local_Groups:_Bucharest]] [[MacOS_X]] [[OpenMoko_under_QEMU_on_MacOSX]] [[Today/2007.2]] [[Today/2007]] [[Dialer/2007.2]] [[Contacts/2007.2]] [[Dialer/2007]] [[ChangeLog]] [[Category:OpenMoko2007]] [[OpenMoko2007]] [[Category:OpenMoko2007.1]] and other changes
Aug 27 19:01:48 <Writchie> eggsy: are you on ubuntu 7.04
Aug 27 19:02:59 <eggsy> Writchie: no gentoo 32 and 64 bit versions
Aug 27 19:03:00 <hhf423> Writchie: let me kill the pin file and see
Aug 27 19:03:29 <hhf423> Writchie: I killed the pin file and it still works
Aug 27 19:03:34 <Writchie> do you have /usr/lib/libXext.so  ?
Aug 27 19:03:36 <hhf423> Writchie: I'll update the wiki
Aug 27 19:03:37 <CM> eggsy: I'm on ~amd64 too, gcc 4.2.0
Aug 27 19:03:58 <eggsy> CM: I'd looked on OE bugzilla let me read openmoko bug
Aug 27 19:04:30 <CM> Writchie: I have /usr/lib/libXext.so -> libXext.so.6.4.0
Aug 27 19:13:02 <eggsy> CM: yeah, this is the bug I have. I'll add to the list. What uses this tool anyway?
Aug 27 19:13:44 <CM> I have no idea, but people have had troubles building it before too
Aug 27 19:16:16 <yoyo_> hi
Aug 27 19:16:43 <ScaredyCat> pesky 64 bit gear
Aug 27 19:16:57 <CM> yoyo_: Hello. Did you manage to build uicmoc4?
Aug 27 19:17:34 <yoyo_> !
Aug 27 19:17:42 <yoyo_> :/
Aug 27 19:17:48 <CM> :(
Aug 27 19:17:50 <ScaredyCat> :-
Aug 27 19:17:59 <CM> ScaredyCat: :P
Aug 27 19:18:19 <ScaredyCat> you broke it ;)]
Aug 27 19:18:55 <yoyo_> at present time Im trying to build i686 compiler to build it on not x68_64 
Aug 27 19:19:11 <jkline> Using the Debug Board, how does one reprogam uboot? (one of our guys accidently ran nand erase -- erassing all of the nand flashing including uboot)
Aug 27 19:19:11 <aloril> jkline: See http://wiki.openmoko.org/wiki/FAQ (and also SH1, SH1_FAQ, Neo1973, Developer_preview and Neo1973_Hardware#GTA01Bv4 pages) (Hopefully these links answers your question. This autotrigger for new nick and question during first 15 minutes done only once.)
Aug 27 19:19:58 <komar007> eggsy: I've built uicmoc following the 747 bug: QT += network in uic3/uic.pro
Aug 27 19:20:03 <hp__> jkline accidently ran nand erase? auch!
Aug 27 19:20:54 <yoyo_> about uicmoc4 someone have the same problem?
Aug 27 19:21:10 <CM> Yes, both me and eggsy 
Aug 27 19:21:16 <eggsy> komar007: I'll try that but -lQtNetwork is in the lib path
Aug 27 19:21:17 <CM> And more people too 
Aug 27 19:21:43 <komar007> eggsy: I know... but if you add network to QT, it'll work;) Magick.
Aug 27 19:22:28 <hhf423> Writchie: ok, I can now pair with the Neo. Any idea how pand could start working
Aug 27 19:22:48 <yoyo_> stinky think :P that uicmoc
Aug 27 19:22:55 <hp__> Qt on the neo.. sweet..
Aug 27 19:23:04 <jkline> hp__, ya.. well... things happen... any way to actually force uboot to be reinstalled w/ the dev board
Aug 27 19:23:14 <hhf423> Writchie: ip a add 10.0.0.2/24 dev bnep0 gives no such interface
Aug 27 19:23:17 <aloril> jkline: see devirginator
Aug 27 19:23:24 <aloril> (or something like that)
Aug 27 19:23:41 <ScaredyCat> hhf423: 
Aug 27 19:23:53 <ScaredyCat> the wiki is misleading
Aug 27 19:24:20 <ScaredyCat> don't do that bit until you start the server + client parts
Aug 27 19:24:34 <hhf423> ScaredyCat: how do I do that?
Aug 27 19:24:48 <ScaredyCat> pand -s and pand -c iirc
Aug 27 19:25:19 <Writchie> hhf423: look at services available from the xp box
Aug 27 19:26:08 <ScaredyCat> the pand -c needs the address as a param
Aug 27 19:26:09 <Writchie> it needs to be showing a PAN -NAP or PAN-USER
Aug 27 19:27:29 <hhf423> whooo, pand -s makes the xp box see a pan
Aug 27 19:27:30 <eggsy> komar007: black magic more like. That worked - let me compare the command line to see why
Aug 27 19:28:11 <komar007> eggsy: I didn't even compare... I'm new to moko. Just wanted to compile and start helping;)
Aug 27 19:28:55 <hhf423> uh, oh, explorer just sort of locked up
Aug 27 19:29:14 <Writchie> windows standard behaviour ;)
Aug 27 19:29:44 <hhf423> :-)
Aug 27 19:38:13 <juri_> yay! my openmoko is at home!
Aug 27 19:38:17 * juri_ heads home early.
Aug 27 19:39:13 <polz> has anyone here managed to make and/or receive calls using a neo1973 ?
Aug 27 19:39:22 <polz> preferably without using a terminal
Aug 27 19:40:30 <hhf423> polz: of course
Aug 27 19:41:03 <hhf423> polz: see: http://wiki.openmoko.org/wiki/Phase_1_Software_Testing#Snapshot_20070825_with_2.6.21.6-moko11-r2
Aug 27 19:41:11 <polz> so... where do I enter the PIN ?
Aug 27 19:41:12 <polz> I can't find that on the wiki
Aug 27 19:41:12 <polz> 10x
Aug 27 19:41:58 <hhf423> polz: which software version do you have?
Aug 27 19:43:19 <polz> ugh... is there any way to check ?
Aug 27 19:43:36 <polz> I'm using the openmoko-devel-image-fic-gta01.jffs2
Aug 27 19:43:39 <polz> as the root
Aug 27 19:44:19 <polz> I got it off of buildhost.openmoko.org/OM2007/tmp/deploy/glibc/images/fic-gta-01/
Aug 27 19:45:04 <polz> btw, is there a known good u-boot image I should upgrade to ? Right now, the phone shuts down after about 30s of inactivity during the boot prompt
Aug 27 19:45:40 <polz> the timestamp on the root image I'm using says it was put there on 26. 8.
Aug 27 19:46:21 <hhf423> polz: 26.08 should be ok
Aug 27 19:46:39 <hhf423> did you do ipkg update && ipkg upgrade from the Neo?
Aug 27 19:46:53 <polz> nope
Aug 27 19:47:11 <hhf423> do that, reboot afterwards for good measure
Aug 27 19:47:20 <hhf423> also, did you put '/etc/init.d/gsmd stop; /etc/init.d/gsmd start' in the /etc/matchbox/session file before 'openmoko-dialer &'  ?
Aug 27 19:47:22 <polz> I'd have to set up the networking properly first, I presume...
Aug 27 19:47:41 <polz> umm... no ?
Aug 27 19:47:54 <hhf423> do that
Aug 27 19:47:57 <polz> should I have ?
Aug 27 19:47:59 <polz> heh...
Aug 27 19:48:03 <polz> OK, justasec
Aug 27 19:48:05 <hhf423> usb networking is properly explained here: http://wiki.openmoko.org/wiki/USB_Networking
Aug 27 19:48:30 <hhf423> polz: well, its explained in phase 1 software testing link I gave you :-)
Aug 27 19:48:35 <polz> for now, I've only set it up so I can ssh to it
Aug 27 19:49:18 <polz> the link says: The ScaredyCat images above from August 10 are the first images which actually asked me for a PIN and registered on the network.
Aug 27 19:49:32 <hhf423> awwww
Aug 27 19:49:39 <ScaredyCat> :D
Aug 27 19:49:39 <polz> and then something about some workaround
Aug 27 19:49:41 <hhf423> no, thats no whats the link says
Aug 27 19:50:03 <hhf423> this link: http://wiki.openmoko.org/wiki/Phase_1_Software_Testing#Snapshot_20070825_with_2.6.21.6-moko11-r2
Aug 27 19:50:31 <hhf423> which points to  Snapshot 20070825 with 2.6.21.6-moko11-r2
Aug 27 19:51:45 * ScaredyCat mutters at not having a .2 image tested
Aug 27 19:51:52 <Writchie> hhf423: looking on the windows side at the device descriptor info it looks all messed up - may be a line ending problem
Aug 27 19:52:01 <polz> ooooh, now I see that
Aug 27 19:52:28 <Writchie> or tabs, etc. bcdDevice in wrong place - protocol is wrong, etc.
Aug 27 19:52:42 * polz plays the idiot.swf for himself ... "you are an idiot, hahahahahahahhaaa haha haha haaaa"
Aug 27 19:53:08 <Writchie> for my last install i modified a nokia 770 *.inf that supposed works on XP with the linux RNDIS gadget
Aug 27 19:54:10 <hhf423> Writchie: you talking about the RNDIS?
Aug 27 19:54:29 <Writchie> yeah - sure doesn't look like the problem is on the windows side
Aug 27 19:54:35 <polz> wow, cool, 10x
Aug 27 19:54:51 <hhf423> Writchie: so there is hope. we can fix everything but windows :-)
Aug 27 19:55:00 <Writchie> yup
Aug 27 19:55:10 <polz> hhf423: why is this workaround needed ? I mean... what should I read tu understand this stuff ?
Aug 27 19:55:35 <Writchie> my guess is the edits to the gadget were borked
Aug 27 19:55:49 * polz only received his phone today and wasn't really compelled to use the slooooooowwww emulator
Aug 27 19:57:49 <polz> is picking up a call supposed to work, too ?
Aug 27 19:58:01 <cjb> I've picked up a call successfully
Aug 27 19:58:05 <cjb> just not made an outgoing :)
Aug 27 19:58:25 <eggsy> for uicmoc4-native bug - Adding +network to uic3.pro changes order of libraries in the linking and seems to work 
Aug 27 19:58:33 <roh> Writchie do you have a pointer for kernelpatches against rndis which make it work?
Aug 27 19:58:34 <polz> for me, the calling itself seems to work
Aug 27 19:58:40 <eggsy> don't know why it worked on some platforms thoug
Aug 27 19:58:43 <polz> I mean.. the called phone rings :)
Aug 27 19:58:49 <eggsy> s/thoug/though/
Aug 27 19:58:49 <apt> eggsy meant: don't know why it worked on some platforms though
Aug 27 19:58:56 <Writchie> roh??
Aug 27 19:59:07 <hhf423> polz: just read every sentence in the wiki quite carefully :-)
Aug 27 19:59:23 <roh> eggsy i think it could be that it seems everybody with that problem uses gcc 3.4 instead of 4 for some reason
Aug 27 19:59:47 <polz> hhf423: just the page you gave me, or the whole wiki ? 'cause the whole wiki seems a biiiiiit too huge :)
Aug 27 19:59:47 <Writchie> roh: I'm suspecting that the OM pathes are what make it not work. 
Aug 27 19:59:57 <roh> Writchie when it works with the N800 and linux and not on the moko and linux then there must be a difference in the code
Aug 27 20:00:04 <roh> pathes?
Aug 27 20:00:41 <roh> Writchie can you give me a sudo lsusb -vvv from the moko which does not work and the N800 that works?
Aug 27 20:00:41 <Writchie> roh: OM patches at least the product ID, everybody else seems to use the default linux vendor ID and product ID
Aug 27 20:01:17 <Writchie> i'll look at the patches later and see what else they do. 
Aug 27 20:01:25 <roh> Writchie vendor ids are in the .inf file. when you patch them it works thats nothing which is or should be fixed in the driver
Aug 27 20:01:53 <Writchie> roh: the produce ID's and vendor ID's are in the code for the gadget
Aug 27 20:02:00 <Writchie> inf file is windows only
Aug 27 20:02:27 <polz> OK, next question: is there any way to compile and test the applications I'd eventually want to run on openmoko....
Aug 27 20:02:33 <Writchie> the device reports that its a fic vendor
Aug 27 20:02:40 <polz> without using qemu ?
Aug 27 20:02:42 <Writchie> but the product id seems screwed up
Aug 27 20:02:50 <Writchie> among other things
Aug 27 20:02:51 <polz> I mean... I'd like to compile the apps natively
Aug 27 20:02:59 <polz> and then recompile them for the neo
Aug 27 20:03:03 <roh> Writchie yes. and the linux driver on the host side does not care because the class and subclass are important
Aug 27 20:03:17 <Writchie> roh: that's what i understand
Aug 27 20:03:21 <torpor> can someone with the patience for such things please read this draft and tell me what you think?
Aug 27 20:03:24 <torpor> http://w1xer.de/OpenMoko_DevNotesDRAFT.txt
Aug 27 20:03:29 <Writchie> windows uses these to locate the "best" matching driver
Aug 27 20:03:42 <roh> thats how it works also with mice and keyboards, usb connected modems, serials, bluetooth recievers and so on
Aug 27 20:04:22 <Writchie> vendor ID and product ID must be properly implemented to meet the USB specs
Aug 27 20:04:37 <torpor> i'm going to HTML'ize it and make it proper for the web a bit later once i've finished it  .. but i'd like some comments now .. 
Aug 27 20:04:40 <Writchie> even if linux USB host is more tolerant
Aug 27 20:04:51 <roh> and they are. these are actually registerd at the usb forum
Aug 27 20:05:39 <roh> Writchie that has nothing to do with tolerance. its how usb works and is supposed to work (and does.. or do you think windows has a table for every usbkeyboard on the planet)
Aug 27 20:06:14 <Writchie> windows will use a class driver unless a better driver is found 
Aug 27 20:06:23 <SpeedEvil> Is there a camera class?
Aug 27 20:06:48 <Writchie> SpeedEvil: i think there is a "image device" class or something like that
Aug 27 20:06:53 <torpor> yes
Aug 27 20:06:55 <torpor> CDC Image
Aug 27 20:07:04 <torpor> CDC Image Capture, rather.
Aug 27 20:07:06 <SpeedEvil> Ah.
Aug 27 20:07:11 <torpor> doesn't have to be a camera, could also be a CCD ..
Aug 27 20:07:18 <SpeedEvil> I wonder if that might be what an advertised 'driver free' webcam is.
Aug 27 20:07:22 <Writchie> properly implementing the idVendor and bcdDevice is part of the usb spec
Aug 27 20:07:31 <roh> when implemented right usb is pretty cool to some degree...
Aug 27 20:07:35 <torpor> you can't get a USB logo without doing it.
Aug 27 20:07:41 <meus> torpor: thankyou!
Aug 27 20:07:44 <torpor> <-- has written firmward for USB devices...
Aug 27 20:07:52 <Writchie> i don't see any USB logo on the neo ;)
Aug 27 20:07:55 <torpor> meus: you're welcoem.  but for what?
Aug 27 20:08:13 <SpeedEvil> http://dealextreme.com/details.dx/sku.5883 for example.
Aug 27 20:08:38 <meus> torpor: the notes :) the ones you posted 5 min ago :)
Aug 27 20:08:46 <roh> torpor i did not mean the logo, with right i meant proper designed interfaces and usage of standards like cdc ethernet or acm
Aug 27 20:08:58 <torpor> oh.  let me know what you think!  i'm not done, this is just a first draft, i wanna get some feedback before i work on completing it tomorrow.
Aug 27 20:09:17 <roh> instead of developing a selfhacked serial protocol like some vendors do and supply bad coded drivers for every os
Aug 27 20:09:38 <Writchie> SpeedEvil: this should use a standard class driver
Aug 27 20:10:05 * SpeedEvil wonders if finally cheap webcams may start working under linux reliably.
Aug 27 20:10:18 <torpor> they have done so for years, SpeedEvil ..
Aug 27 20:10:50 <SpeedEvil> Indeed - for name brand ones.
Aug 27 20:11:02 <meus> torpor: willdo!
Aug 27 20:11:11 <yoyo_> echo mem > /sys/power/state 
Aug 27 20:11:18 <SpeedEvil> Not so much for random 'fuki camera' ones, or they sort-of-work.
Aug 27 20:11:24 <Writchie> roh: i think the problem is that windows does not supply a standard usb serial port class driver
Aug 27 20:11:45 <ewon> !etl
Aug 27 20:11:49 <ewon> oops
Aug 27 20:11:50 <Writchie> RNDIS is the only communication class driver AFAIK that MS has supplied
Aug 27 20:12:32 <torpor> windows is crap for CDC
Aug 27 20:12:32 <polz> OK, next question: is the GSM signal supposed to be clearly audible when making a call ?
Aug 27 20:12:39 <torpor> actually windows is utterly crap for USB in general, anyway.
Aug 27 20:12:59 <SpeedEvil> polz: Ideally, no. In practice I found it moderately so.
Aug 27 20:13:17 <SpeedEvil> If this is the fault of the hardware or the mixer settings is a good question.
Aug 27 20:13:21 <polz> hmm.. here, it's a bit loud, in my oppinion...
Aug 27 20:13:28 <SpeedEvil> If it's hardware, there is nothing that can be done.
Aug 27 20:13:45 <Writchie> SpeedEvil: it probably fails to meet GSM specs if you can hear it?
Aug 27 20:13:48 <SpeedEvil> It's louder than on any other phone I've had.
Aug 27 20:13:53 <polz> maybe add some shielding to the wire that leads to the speaker
Aug 27 20:14:13 <SpeedEvil> Speaker leads are least likely to be the problem.
Aug 27 20:15:16 <polz> OMG, I'm glad the calling worked right out of the box
Aug 27 20:15:25 * polz has just run alsamixer for the first time
Aug 27 20:15:26 <ScaredyCat> SpeedEvil: http://www.theinquirer.net/default.aspx?article=39291
Aug 27 20:15:40 <ScaredyCat> http://mxhaard.free.fr/spca5xx.html
Aug 27 20:15:46 <Writchie> SpeedEvil: i need to find the name of FIC's test lab so I can use them ;)
Aug 27 20:15:52 <polz> how was anyone able to guess a working setting with _this_ many channels ? :)
Aug 27 20:16:56 <SpeedEvil> polz: you don't guess
Aug 27 20:17:07 <Writchie> maybe u-boot menu is considered "normal configuration of the device"
Aug 27 20:17:23 <SpeedEvil> polz: if you have a full diagram, and all of the settings mentioned, then it's fairly simple to generate mixer settings.
Aug 27 20:17:59 * polz has a loooot of reading ahead of himself
Aug 27 20:18:07 <polz> SpeedEvil: 10x for all the help so far
Aug 27 20:18:22 <SpeedEvil> roh: Is there an internal 'this is a picture of the soundcard mixers and switches with everything labelled with the alsamixer names, and what's connected to them' that is around internally?
Aug 27 20:19:18 <roh> does https://wiki.openmoko.org/wiki/Neo1973_Audio_Subsystem not help?
Aug 27 20:19:29 <SpeedEvil> roh: It helps. Somewhat.
Aug 27 20:19:52 <SpeedEvil> But which mixer on alsamixer is the four-way switch to the top-left (from memory), for example.
Aug 27 20:20:02 <eggsy> roh: on X86_64 I was using gcc 4.1.2 with same error. The uic3.pro hack fixed here as well as 32 bit gcc 3.4.6
Aug 27 20:20:39 <roh> eggsy and on 32bit gcc4 it just works without patch?
Aug 27 20:20:48 <Waldmokomeister> hi!
Aug 27 20:21:39 <dcordes> Waldmokomeister: ho
Aug 27 20:22:26 <Writchie> SpeedEvil: default charge gives up long before battery is fully charged. 
Aug 27 20:22:42 <SpeedEvil> Writchie: I don't think so.
Aug 27 20:22:53 <SpeedEvil> Writchie: unless batvolt lies.
Aug 27 20:23:03 <Writchie> i think it does
Aug 27 20:23:07 <SpeedEvil> It stays at 4200mV for me. (which is actually a concern)
Aug 27 20:23:38 <Writchie> i had a neo run out over the weekend. - charged at 40ma for 7 hours - then was able to reboot and its been drawing 490ma for the last two hours
Aug 27 20:23:44 <roh> SpeedEvil which four way switch?
Aug 27 20:24:16 <Waldmokomeister> dcordes: hi. i want to dial out, but gsmd -p /dev/ttySAC0 -s 115200 -F says can't open unix socket"
Aug 27 20:24:20 <Writchie> i think you can float the battery up to 4.2V with well below a full charge in it.
Aug 27 20:24:24 <SpeedEvil> roh: I mean in general - the mixer names and the labels on the diagram don't match.
Aug 27 20:24:50 <roh> SpeedEvil eh.. good question. i think we need to lokok that up in the source
Aug 27 20:24:51 <Waldmokomeister> does the dialer work at moment?
Aug 27 20:25:05 <SpeedEvil> Writchie: You can't. you have about 80% of charge once it hits 4.2V
Aug 27 20:25:18 <SpeedEvil> This is inherent in the electrochemistry.
Aug 27 20:25:24 <torpor> okay i'm out for the night - if someone has the time to review it and can send me some notes, please check out: http://w1xer.de/OpenMoko_DevNotesDRAFT.txt and let me know .. i'll be working on it again tomorrow, and hopefully will be finished with it tomorrow evening, for inclusion in the wiki ..
Aug 27 20:26:30 <Writchie> SpeedEvil: are you sure that you actually triclkle charge li-ion from dead to 4.2 without giving up more than 20%
Aug 27 20:26:50 <hhf423> ok, I can set up a PAN with my k610i and then the neo connects to the internets via the k610i
Aug 27 20:26:54 <SpeedEvil> Writchie: yes - if you keep the voltage at 4.2, it's fully charged in an hour.
Aug 27 20:27:10 <SpeedEvil> This is actually bad for the battery - it shouldn't be floated at 4.2
Aug 27 20:27:14 * SpeedEvil checks
Aug 27 20:27:22 <cw666> SpeedEvil: can you leave it connected 24/7 safely?
Aug 27 20:27:45 <cw666> i mean to the usb cable
Aug 27 20:28:10 <SpeedEvil> cw66: there is no actual risk - at worst it will degrade the battery faster.
Aug 27 20:28:19 <Writchie> SpeedEvil: then I wonder where my 2 Watt Hours over the last two hours has gone?
Aug 27 20:28:41 <cw666> also, *what* stays drawing power when the device is off?  the battery drains very quickly
Aug 27 20:28:57 <Writchie> it's not heat
Aug 27 20:29:18 <SpeedEvil> Writchie: what do you mean?
Aug 27 20:29:50 <Writchie> I've been charging at 490ma for more than two hours - the neo runs on < 300ma
Aug 27 20:30:13 <Writchie> this is after the 40ma charge went to zero.
Aug 27 20:30:42 <roh> cw666 i dunno if the gsm gets switched off properly. since that happens via at-command it could be that it doesn't i case gsmd is broken in the image
Aug 27 20:31:03 <__Ace__> anyone here that can give me ideas on how to calculate with precent when I cant use either doable or float?
Aug 27 20:31:15 <cw666> roh: oh, good point, it almost certainly doesnt sometimes
Aug 27 20:31:25 <cw666> gsmd craps out a lot, i need to fix that then
Aug 27 20:31:59 <cw666> __Ace__: what do you mean?  and fwiw, doesnt -msoft-float work?
Aug 27 20:32:22 <Waldmokomeister> okay. found it. how do i configure my gsm-connection? should the dialer work from scretch?
Aug 27 20:32:41 <polz> __Ace__: multiply everything by 100, calculate, then divide by 100
Aug 27 20:33:17 <__Ace__> nah, -msoft-float doesnt seem to work in the crosscompiler I use since that seems to have standard libs created with HP
Aug 27 20:33:25 <polz> Waldmokomeister: http://wiki.openmoko.org/wiki/Phase_1_Software_Testing#Snapshot_20070825_with_2.6.21.6-moko11-r2
Aug 27 20:33:46 <__Ace__> well, hardware float..
Aug 27 20:42:05 <SpeedEvil> Writchie: see http://www.ibt-power.com/Battery_packs/Li_Ion/Lithium_ion_tech.html
Aug 27 20:44:18 <cw666> __Ace__: HP?
Aug 27 20:45:34 <roh> cw666 when you like to take measurements and have time.. see https://wiki.openmoko.org/wiki/GTA01_gsm_modem#AT.40POFF
Aug 27 20:45:50 <roh> bluetooth and lcm also eats a lot of power
Aug 27 20:46:09 <cw666> roh: the kicker is i dont really have a way to measure power draw
Aug 27 20:46:31 <cw666> well, not without making something...  also, i notice checking the charging current the value is all over the place
Aug 27 20:46:39 <cw666> so im not even sure that driver is correct
Aug 27 20:47:13 <SpeedEvil> chgcur is horrible.
Aug 27 20:47:26 <SpeedEvil> You need to average some dozens of measurements.
Aug 27 20:47:38 <SpeedEvil> then it means something - but not for power draw.
Aug 27 20:48:02 <SpeedEvil> roh: @poff is different from the echo 0 >/sys/wherever?
Aug 27 20:48:08 * SpeedEvil checks.
Aug 27 20:48:52 * SpeedEvil has a multimeter.
Aug 27 20:48:58 <cw666> SpeedEvil: for draw i was just thinking about using and external meter actually
Aug 27 20:49:37 <SpeedEvil> I've got a breakin to measure battery current.
Aug 27 20:50:14 <roh> SpeedEvil yes.
Aug 27 20:51:03 <roh> SpeedEvil the entry in the sysfs only is a line which presses the power-on button on the gsm-pmu. but there is no way to switch it off via a line (design problem of the gsm-pmu)
Aug 27 20:51:11 <cw666> SpeedEvil: breakin?
Aug 27 20:51:18 <__Ace__> polz that doesnt work
Aug 27 20:51:33 <__Ace__> I wanna translate 0-870 to 0-320
Aug 27 20:51:37 <SpeedEvil> cw666: a way of connecting the meter in series with the battery.
Aug 27 20:51:54 <roh> it was intended for featurephones with a button and people pressing it.
Aug 27 20:51:57 <SpeedEvil> roh: ah - right. That explains a lot.
Aug 27 20:53:05 <roh> SpeedEvil and since its on when gsm is on it is also used for switching the latch between debug-board serial and gsm modem
Aug 27 20:53:59 * SpeedEvil hooks meter up and waits for boot.
Aug 27 20:55:34 * ScaredyCat gives SpeedEvil the boot
Aug 27 20:58:52 <SpeedEvil> Hokay - it's drawing under 10mA with @poff activated. 
Aug 27 20:58:55 <SpeedEvil> :)
Aug 27 20:59:08 <SpeedEvil> Compared to 30 before
Aug 27 21:00:11 <roh> SpeedEvil in which state?
Aug 27 21:00:34 <roh> meaning.. what parts did you leave on
Aug 27 21:00:50 <SpeedEvil> shutdown -h 
Aug 27 21:01:01 <aloril> (script) wiki RecentChanges: [[User:Hhf423]] [[Phase_1_Software_Testing]] [[Buying_Interest_List]] [[Manually_using_Bluetooth]] [[P1_Owners]] [[USB_Networking]] [[Disassembling_Neo1973]]
Aug 27 21:01:07 <SpeedEvil> getting more accurate measurement means.
Aug 27 21:01:54 <roh> SpeedEvil can you try switching off all subsystems via sysfs before?
Aug 27 21:02:15 <SpeedEvil> First I'm checking poweroff doing nothing accurately, poweroff doing @poff accurately
Aug 27 21:02:51 <SpeedEvil> then more - if it's not ~zero
Aug 27 21:07:29 <SpeedEvil> Ok - first confirmed at some 25mA at 4.2V. (just off via off button)
Aug 27 21:10:14 <SpeedEvil> After @poff - 12mA
Aug 27 21:11:03 <SpeedEvil> Which implies that if it's gettable to 'zero' other than the modem, somewhere in the region of 100 hours should be plausible for standby time.
Aug 27 21:22:32 <Writchie> SpeedEvil: the CC portion of the cycle is nominally 1C - i've seen curves for 0.3C on some batteries but 40ma is 0.03C 
Aug 27 21:23:55 <SpeedEvil> roh: I can't find anything other than bluetooth, GPS, and the @poff technique - are there more power off sysctls?
Aug 27 21:25:38 <Writchie> SpeedEvil: anyway, it looks to me like this very low current charge eventually raises the battery voltage to 4.2 where the charger then cutsoff. But this won't hold for very long. As soon as you boot, the voltage drops considerable and a normal CC/CV cycle kicks in. Only after 3 hours now am I down to 410 ma total current from the USB
Aug 27 21:25:38 <SpeedEvil> Other than echo 0 > /sys/bus/i2c/devices/0-0008/voltage* - which seems extreme.
Aug 27 21:26:45 <SpeedEvil> You can't raise the voltage on any rechargable battery to its maximum terminal voltage, for hours, and not have it charged.
Aug 27 21:27:10 <SpeedEvil> The electrochemistry doesn't permit it.
Aug 27 21:27:39 <Writchie> Writchie: it's charged for sure, but at what % of its capacity under proper charge conditions?
Aug 27 21:28:14 <SpeedEvil> Some 80-90% - if it's indeed at 4.2V
Aug 27 21:28:29 <SpeedEvil> If it remains at 4.2V for 2 hours or so, it will be fully charged.
Aug 27 21:28:54 <SpeedEvil> Basically, it's an almost perfect integrator.
Aug 27 21:28:57 <Writchie> I can't be sure what voltage the battery attained, only that the PMU killed charging
Aug 27 21:29:17 <SpeedEvil> If you charge it at 10mA for 120 hours, or 100mA for 12 hours, you get the same thing.
Aug 27 21:29:41 <SpeedEvil> This starts to break down at high rates, where energy is lost as heat
Aug 27 21:30:05 <Writchie> then what's the reason for not trickle charging li-ion?
Aug 27 21:30:50 <SpeedEvil> Writchie: the reason is you don't (ideally) keep the float charge at 4.2V going on forever, as this damages the battery over time, and rapidly (well, more than normal) reduces its capacity.
Aug 27 21:31:42 <SpeedEvil> The right thing to do is to turn off the charger, allow the battery voltage to be whatever it is, without drawing current from it, until it reaches 4.1V (?) at which point you turn the charger back on.
Aug 27 21:31:59 <SpeedEvil> this may be after quite a long time.
Aug 27 21:32:13 <Writchie> Writchie: yes, that's what I see in other charger spec sheets except 4.05 is usually the voltage to start a top off
Aug 27 21:32:44 <Writchie> in any case, the charger for sure cuts off after 7 hour at 40ma which is not enough for a full charge but enough to boot. 
Aug 27 21:33:45 <mdt> cat someone enlighten me about the content of /etc/ipkg in the image Snapshot-20070819? it contains 6 config files where packages are doubled. while i installed a kernel 2.6.21.5-moko10 that came with the rootfs the ipkg update installed modules for 2.6.21.6-moko11-r2... what did i wrong?
Aug 27 21:35:46 <SpeedEvil> Writchie: there is a timer in the PC* chip, which limits the time.
Aug 27 21:36:00 <SpeedEvil> As I understand it anyway.
Aug 27 21:38:36 <dcordes> how is the openmoko session initialized on bootup?
Aug 27 21:39:07 <dcordes> is this any documentated?
Aug 27 21:39:24 <hads> What session are you referring to?
Aug 27 21:39:43 <hads> You may want to look at /etc/matchbox/session
Aug 27 21:39:47 <dcordes> "that X11 thing" that pops up after you boot
Aug 27 21:39:51 <dcordes> ok thanks
Aug 27 21:44:31 <dcordes> hads: ok that looks interesting but I don't know how to solve my problem wiht it. Thing is: I switched from gpe to openmoko2
Aug 27 21:45:06 <dcordes> and when I remove gpe-login, starting X11 results in a flashing screen
Aug 27 21:45:22 * hads knows nothing about gpe
Aug 27 21:45:23 <dcordes> the X cursor appears then the screen is erased and started again in a loop
Aug 27 21:45:42 <dcordes> I installed openmoko2-session after it but that doesn't work
Aug 27 21:53:10 <msfrank> hiya!  i'm interested in hacking on the bluetooth subsystem, specificially the hands-free and phonebook-access profiles, and i was wondering if anyone knows how much development has progressed on that front (and where i could find the code)
Aug 27 21:53:58 <juco2> http://wiki.openmoko.org/wiki/Manually_using_Bluetooth
Aug 27 21:54:11 <bmidgley2> msfrank: the latest bluez utils should have a workable headset impl
Aug 27 21:54:24 <msfrank> ah thnx
Aug 27 21:54:40 <juco2> I don't see anything about phonebook access
Aug 27 21:54:44 <msfrank> headset and handsfree are different iirc, i'm interested in interfacing with my car
Aug 27 21:55:14 <juco2> I think many of those profiles are missing from bluez
Aug 27 21:55:57 <juco2> http://www.bluez.org/profiles.html
Aug 27 21:56:20 <bmidgley2> the plan is for the voice support in bluez to support handsfree
Aug 27 21:56:30 <bmidgley2> it will use hf if available otherwise hs
Aug 27 21:56:50 * hads wishes for A2DP
Aug 27 21:56:53 <msfrank> interesting, i'll have to look more closely at bluezs roadmap
Aug 27 21:57:35 <msfrank> i'd like the phonebook support too, so my car stereo will display caller id =)
Aug 27 21:57:37 <bmidgley2> hads: a2dp is very close. if the codec scales down it would work with just a little hacking
Aug 27 21:57:43 <eggsy> roh: with gcc 4.1.1 on 32 bit I still need the uic3.pro patch to build uicmoc4-native. I'll try with gcc 4.2 overnight
Aug 27 21:58:00 <hads> bmidgley2: Nice, I thought it was still a way off.
Aug 27 21:58:05 <juco2> phonebook sits on top of rfcomm, right?
Aug 27 21:58:23 <msfrank> yeah
Aug 27 21:58:26 <bmidgley2> hads it depends on our codec. could use some optimization but i'm out of my depth
Aug 27 21:58:34 <hhf423> wayyy to go!
Aug 27 21:58:34 <msfrank> well, everything is on top of rfcomm, pretty much
Aug 27 21:58:47 <hads> bmidgley2: That's way out of my depth :)
Aug 27 21:58:48 * hhf423 can ssh from his windows box to the Neo via BT
Aug 27 21:58:49 <msfrank> unless you are using the bt protocol over a wired protocol for some reason
Aug 27 21:59:01 * hhf423 dances the bluetooth dance
Aug 27 21:59:17 <msfrank> hhf423: yes, it is a dance alright ;)
Aug 27 21:59:23 <bmidgley2> hhf423: using PAN?
Aug 27 21:59:27 <hhf423> yes
Aug 27 21:59:42 <bmidgley2> write it up! :)
Aug 27 21:59:53 * hhf423 can PAN from Neo to Windows (or, to be exact, from Windows to Neo)
Aug 27 22:00:14 <hhf423> its a bit bizarre, but it works. I'll now update the wiki
Aug 27 22:01:20 <juco2> http://wiki.bluez.org/wiki/Serial
Aug 27 22:01:21 <SpeedEvil> Interesting and odd. Reducing the set CPU voltage by 10mV at a time reveals three distinct power states. 
Aug 27 22:01:30 <riot> mhmmm.. even the gtk.calendar-object looks good. but the fonts are all too BIG
Aug 27 22:01:36 <juco2> looks like pbap for the neo would be something fairly straightforward to implement
Aug 27 22:01:41 <riot> how to get them smaller??
Aug 27 22:01:51 <juco2> if you call reading the bluetooth spec straightforward
Aug 27 22:01:56 <SpeedEvil> 2100-1850mV = 190mA
Aug 27 22:02:08 <msfrank> juco2: seriously.  the core spec is a 12mb pdf.
Aug 27 22:02:14 <ewon> nic
Aug 27 22:02:16 <ewon> e
Aug 27 22:02:16 <SpeedEvil> 1850-1560 or so - 192
Aug 27 22:02:24 <SpeedEvil> 1560 and below - 170
Aug 27 22:02:57 <SpeedEvil> this is md5suming a file after each reduction
Aug 27 22:03:24 <SpeedEvil> md5sum segfaulted at 1480mV
Aug 27 22:03:53 <anrp> that's probably bad D:
Aug 27 22:03:58 <riot> seriously, how did the core-developers realize such tiny fonts? all i get in pygtk is HUGE.
Aug 27 22:05:43 <SpeedEvil> And the system wedged - or at least fell off cdc_ether at 1460mV
Aug 27 22:06:34 <anrp> SpeedEvil: the PCF chip has very limited resolution at high v
Aug 27 22:06:45 <anrp> you get 0.025V steps up to 1.5
Aug 27 22:06:53 <anrp> then you get 1.8, 2.1, 2.4
Aug 27 22:06:57 <anrp> which is ...
Aug 27 22:07:02 <SpeedEvil> That makes sense.
Aug 27 22:07:39 <anrp> the chip is run at 2.1v normally but the arm core isn't really rated for it
Aug 27 22:07:40 <roh> SpeedEvil did you adjust the pll for different clocking?
Aug 27 22:07:47 <anrp> which is why it disappeared from the later datasheets
Aug 27 22:08:03 <anrp> yeah, the SOA graph tells you allowable speed vs voltage
Aug 27 22:08:07 <roh> eggsy i use 4.1.2 and 3.4.6 here
Aug 27 22:08:31 <SpeedEvil> roh: no.
Aug 27 22:08:39 <SpeedEvil> This is at 266?
Aug 27 22:08:47 <SpeedEvil> whatever the stock shipped version was.
Aug 27 22:08:49 <anrp> 266 is only allowed at 2.1
Aug 27 22:08:49 <SpeedEvil> Interesting.
Aug 27 22:08:54 <anrp> by the old datasheet
Aug 27 22:08:56 <SpeedEvil> It seems stable at 1500.
Aug 27 22:09:00 <roh> on gta01bv4 / P1 266mhz is default, yep
Aug 27 22:09:01 <anrp> well it might work
Aug 27 22:09:06 <anrp> but that there's overclockin land
Aug 27 22:09:10 <anrp> or underclocking
Aug 27 22:09:20 <SpeedEvil> serious undervolting.
Aug 27 22:09:27 <anrp> D:
Aug 27 22:09:36 <anrp> also, even if the current is the same
Aug 27 22:09:50 <roh> anrp the 2410 can be run on 200mhz also.. i would need to look up at which voltage we did that
Aug 27 22:09:51 <SpeedEvil> that was the battery current I was measuring.
Aug 27 22:09:54 <anrp> the voltage output goes down. as long as it's one of the switching regulators, that means the power consumption goes down
Aug 27 22:09:54 * SpeedEvil rechecks.
Aug 27 22:10:04 <anrp> o >_> yeah, probably battery
Aug 27 22:10:06 <anrp> which is overall
Aug 27 22:10:34 <anrp> the 2410 can be run at anything from 20 to 200, by design
Aug 27 22:10:46 <anrp> samsung used to rate chips to run at 266 with a higher v
Aug 27 22:10:49 <anrp> but not anymore
Aug 27 22:10:49 <riot> no app-developers awake?
Aug 27 22:10:59 <SpeedEvil> anrp: no - there are constraints on using a 12Mhz external clock
Aug 27 22:11:10 <anrp> only if you need usb
Aug 27 22:11:16 <anrp> and that's a seperate PLL anyway
Aug 27 22:11:23 <SpeedEvil> Not, read the constraints on the clock PLL
Aug 27 22:11:32 <SpeedEvil> it says 'minimum of 4* input frequency'
Aug 27 22:11:43 <anrp> oh, so 48-200 then ~_~
Aug 27 22:11:45 <anrp> if the crystal is 12
Aug 27 22:11:52 <SpeedEvil> though it violates that on the recommended clocks table, and goes down to nearly 3* IIRC
Aug 27 22:11:55 <anrp> or slow @ 12mhz
Aug 27 22:12:04 <dcordes> when I press fn on my zaurus, openmoko-terminal changes input mode to greek-iso :(
Aug 27 22:12:13 <anrp> greek, you say D:
Aug 27 22:12:34 <guaqua> it's still greek, not chinese :)
Aug 27 22:12:38 <dcordes> s/openmoko-terminal/mrxvt
Aug 27 22:12:39 <anrp> you can go pretty far outside ratings before things break
Aug 27 22:12:46 <anrp> but... it's not dependable ~_~
Aug 27 22:12:58 <dcordes> I swear
Aug 27 22:13:12 <MacNorth> me too! fvck !
Aug 27 22:13:12 <dcordes> I press fn => [Greek: iso]
Aug 27 22:52:20 * lSpeedEvi sighs.
Aug 27 22:54:18 <dcordes> how can I launch the openmoko xsession from command line?
Aug 27 23:00:08 <riot> dcordes: so, xserver-nodm is kdrive, actually?
Aug 27 23:00:15 <juri_> yay! i have an openmoko! :)
Aug 27 23:00:20 * juri_ cheers.
Aug 27 23:00:27 <riot> juri_: congrats
Aug 27 23:00:31 * juri_ does somersaults
Aug 27 23:00:39 <dcordes> riot: riot er I think xserver-nodm is just that startmethod
Aug 27 23:00:44 <dcordes> but no idea to be honest
Aug 27 23:00:57 <juri_> ... after 14 days of shipping, including going to germany...
Aug 27 23:01:01 <riot> dcordes: yeah, sure. but the underlying xserver in the default image.. is that kdrive?
Aug 27 23:01:41 <riot> oh, yes it is
Aug 27 23:01:50 * dcordes has no idea
Aug 27 23:02:01 <aloril> (script) wiki RecentChanges: [[Manually_using_Bluetooth]] [[Neo1973_and_Windows]] [[Phase_1_Software_Testing]] [[Buying_Interest_List]] [[MokoMakefile]]
Aug 27 23:03:26 <riot> xserver-kdrive - 1:1.3.0.0-r16 - X server from freedesktop.org
Aug 27 23:03:35 <riot> quite obviously, it is.
Aug 27 23:04:13 <dcordes> what's obvious?
Aug 27 23:04:26 <riot> that ipkg-info about your xserver ;)
Aug 27 23:04:38 <riot> not the fact itself, actually :>
Aug 27 23:05:28 <dcordes> riot: how do you change themes in openmoko?
Aug 27 23:06:02 <amirikin> i dont think there are other themes yet
Aug 27 23:06:19 <SpeedEvil> vi :)
Aug 27 23:06:35 <dcordes> amirikin: there are.
Aug 27 23:06:38 <dcordes> the gpe themes...
Aug 27 23:06:54 <dcordes> ipkg list | grep gpe-theme
Aug 27 23:07:05 <dcordes> possibly not in openmoko :D
Aug 27 23:07:29 <dcordes> SpeedEvil: so where's the setting file for that?
Aug 27 23:08:17 <dcordes> setting the theme
Aug 27 23:09:02 <SpeedEvil> dcordes: I don't really knwo - I've been looking at low level stuff mainly. 
Aug 27 23:09:05 <SpeedEvil> Nott he apps.
Aug 27 23:09:34 <dcordes> ok I did not have the them installed somehow
Aug 27 23:09:48 <dcordes> SpeedEvil: how would you start the openmoko xsession from command line?
Aug 27 23:09:59 <SpeedEvil> shutdown -r -t 0 now
Aug 27 23:10:10 <dcordes> no other way?`
Aug 27 23:10:13 * SpeedEvil tries to remember where he found the script
Aug 27 23:10:35 <SpeedEvil> Sorry - at the moment my server, to which my phone is attached is rebooting and fscking
Aug 27 23:10:46 <dcordes> k
Aug 27 23:10:49 <SpeedEvil> I think somewhere in /etc/rc*
Aug 27 23:10:57 <SpeedEvil> Xsession?
Aug 27 23:11:01 <SpeedEvil> xsession?
Aug 27 23:12:01 <dcordes> SpeedEvil: perfect!
Aug 27 23:12:02 <dcordes> thanks
Aug 27 23:12:20 <polz> dcordes: /etc/init.d/x-something
Aug 27 23:12:32 <dcordes> there is some breakage about my theme path
Aug 27 23:12:46 <SpeedEvil> :)
Aug 27 23:13:18 <dcordes> there is an error messge right after session launch: "Failed to open file '/usr/share/pixmaps//lock.png': Not found
Aug 27 23:13:23 <polz> btw, is there supposed to be any way to use mrxvt, or is it just loaded by default as a demo ?
Aug 27 23:13:24 <dcordes> notice the 2x/
Aug 27 23:13:49 <dcordes> I like mrxvt but the new tab and close buttons could be bigger
Aug 27 23:13:49 <polz> I mean... the thing doesn't seem to have a keyboard anywhere
Aug 27 23:14:00 <polz> or am I missing something ?
Aug 27 23:14:16 <dcordes> polz: there is not much IM in openmoko2007.2 at all
Aug 27 23:14:25 <dcordes> afaik
Aug 27 23:14:38 <polz> instant messaging ??
Aug 27 23:14:48 * riot is working on an im, which shall have plugin-ui's
Aug 27 23:14:53 <dcordes> polz: input method.
Aug 27 23:14:58 <polz> oh.. OK
Aug 27 23:15:08 <dcordes> like display keyboard
Aug 27 23:15:16 <polz> but the search widgets seem to have at least some sort of keyboard
Aug 27 23:15:25 <polz> it's a bit silly that the terminal doesn't
Aug 27 23:15:39 <riot> dcordes: there is the oe-input-manager. Works quite fine, but has a strange bug: you must start it via group "ALL" - not utilities
Aug 27 23:15:45 <polz> ah, well, that should come along pretty soon, right ?
Aug 27 23:16:02 <linux_galore> yeah, the UI on the OM seriously needs a bit of attention its still too primitive, if you compare it to a N800 you will see what I mean
Aug 27 23:16:06 <dcordes> riot: I don't have groups
Aug 27 23:16:24 <dcordes> but my device also have qwerty keyboard
Aug 27 23:16:27 <riot> dcordes: on the today-launcher? in 2007.2? you should have them
Aug 27 23:16:28 <SpeedEvil> linux_galore: N800 has had >>>100* the funding.
Aug 27 23:16:35 <dcordes> riot: nope.
Aug 27 23:16:42 <dcordes> riot: actually I'm on angstrom
Aug 27 23:16:45 <dcordes> :P
Aug 27 23:16:52 <riot> dcordes: now.. go away :P
Aug 27 23:17:09 <riot> dcordes: still you should be able to use the IM
Aug 27 23:17:20 <polz> btw, has anyone managed to get the USB working as a host yet ?
Aug 27 23:17:44 <dcordes> polz: modprobe ohci_hcd ?
Aug 27 23:17:50 <dcordes> that should do the magic
Aug 27 23:17:54 <linux_galore> SpeedEvil: you do realise there are other FOSS UI's with no funding and they get it right
Aug 27 23:18:02 <riot> that is possible? the openmoko-wiki doesn't mention THAT one..
Aug 27 23:18:10 <polz> dcordes: I'll do that in the morning
Aug 27 23:18:16 <riot> linux_galore: name one.
Aug 27 23:18:27 <polz> and see if I can attach a keyboard to the phone
Aug 27 23:18:29 <dcordes> polz: you will need a powered usb hub
Aug 27 23:18:32 <linux_galore> matchbox
Aug 27 23:18:33 <cesarb> I just saw on the backlog that someone complained that nobody filed a bug on GSM850 not working...
Aug 27 23:18:46 <polz> dcordes: can't I just wire a 5V line into a modified cable ?
Aug 27 23:19:13 <cesarb> It would be better if someone who actually tried it filed the bug, as I so far only failed to see the GSM850 carrier on a network scan
Aug 27 23:19:43 <linux_galore> whats the Linux UI they have on the Palm when you want to run Linux, that  just looks awesome 
Aug 27 23:20:02 <riot> dcordes: are you sure, the neo supports host-usb?
Aug 27 23:20:15 <dcordes> polz: I do it that way: wire a 5v line to the host cable of my mini usb hub
Aug 27 23:20:20 <dcordes> riot: no
Aug 27 23:20:30 <riot> dcordes: what platform are you using? zaurus?
Aug 27 23:20:49 * riot 'd love host usb on the neo...
Aug 27 23:21:04 * polz is going to sleep now
Aug 27 23:21:04 <dcordes> yes
Aug 27 23:21:09 <cesarb> riot: the kernel driver for host usb on the neo AFAIK is not ready yet
Aug 27 23:21:11 <polz> so.. good night
Aug 27 23:21:18 <dcordes> riot: neo has usb host
Aug 27 23:21:22 <dcordes> polz
Aug 27 23:21:27 <dcordes> wait a second
Aug 27 23:21:32 <polz> yes ?
Aug 27 23:21:36 <riot> oh. Many dreams coming true in the near future, as it sounds.
Aug 27 23:21:50 <polz> riot: don't count on it being near
Aug 27 23:21:55 <cesarb> riot and host usb on the neo needs an Evil cable
Aug 27 23:21:57 <linux_galore> riot: then you have Poky http://www.pokylinux.org/
Aug 27 23:22:03 <riot> like my EOS350D as a Mobile-Camera :D
Aug 27 23:22:16 * polz just hopes openmoko doesn't fold before the software is ready to be used
Aug 27 23:22:48 <riot> polz: uh, just some thousands of new hackers just received their neo phase1-device..
Aug 27 23:23:09 <cesarb> polz: *knocks on wood* shhhh
Aug 27 23:23:30 <riot> np. The community is growing fast paced, imho.
Aug 27 23:23:46 <linux_galore> Im surprised they dont try and integrate poky into the OM UI
Aug 27 23:23:54 <riot> just today i convinced three friends just by showing up with the device :)
Aug 27 23:23:57 <dcordes> polz: http://img390.imageshack.us/my.php?image=z1xy6.jpg
Aug 27 23:24:18 <cesarb> I have two friends who are just waiting for the software to be usable enough
Aug 27 23:24:25 <riot> dcordes: eew.. that looks.. hardcore..
Aug 27 23:24:56 <riot> dcordes: what is that? wlan, bluetooth and what?
Aug 27 23:25:09 * Kero wondes whether the Neo has to run out of battery 1) when it's off, and esp 2) when it's on and connected to USB
Aug 27 23:25:20 <amirikin> im waiting until the hardware update
Aug 27 23:25:31 <dcordes> riot: device is sharp zaurus sl-c1000 with usb hub connected to its usb host
Aug 27 23:25:45 <dcordes> silver cable with little red and black wire
Aug 27 23:25:55 <linux_galore> yeah, GTA02 will open allot of doors for better graphics
Aug 27 23:25:55 <dcordes> these go to 2000mah li ion battery
Aug 27 23:25:56 <polz> dcordes: that's waaay prettier than what I would have imagined
Aug 27 23:25:58 <riot> dcordes: yeah, i can see that, but what usb-devices are connected? weird bunch :)
Aug 27 23:26:18 <dcordes> rt73 based 802.11g device
Aug 27 23:26:26 <riot> portable madness *bg*
Aug 27 23:26:53 * polz is off
Aug 27 23:26:56 <dcordes> I can't wait getting my phase02 to try
Aug 27 23:26:57 <riot> gdn8 polz 
Aug 27 23:27:02 <dcordes> cya polz 
Aug 27 23:27:46 <riot> yeah, we're sure all getting some shitloads of fun with the next generation neos.. this one's already coming with a great funfactor
Aug 27 23:28:06 <amirikin> i just don't want to get a G2.5 device when they are going to turn around and release a g3 a few months later
Aug 27 23:28:19 <amirikin> with a faster processor
Aug 27 23:28:28 <riot> amirikin: the current device is G2 only.
Aug 27 23:28:35 <linux_galore> amirikin: yeah but it will cost more
Aug 27 23:28:47 <riot> if i'm not mistaken, there'll be no g2.5.
Aug 27 23:28:58 <amirikin> cost more than buying this both this device and the next device
Aug 27 23:29:00 <amirikin> ?
Aug 27 23:29:06 <Kero> gprs is 2.5, no?
Aug 27 23:29:09 <dcordes> riot: do you have any idea why my system is looking for wrong theme path?
Aug 27 23:29:24 <riot> no.. edge is 2.5, afaik. GSM +GPRS are g2.
Aug 27 23:29:34 <linux_galore> Im sort of put of that the GTA02 isnt G3
Aug 27 23:29:37 <riot> dcordes: uh. no
Aug 27 23:29:42 <dcordes> nvm
Aug 27 23:29:44 <riot> linux_galore: yep.
Aug 27 23:29:50 <amirikin> gta02 isn't g3?
Aug 27 23:29:58 <riot> linux_galore: but all the 2008-devices shall be g3, according to laf0rge
Aug 27 23:30:04 <Kero> mm, as long as it has gprs, I'm happy enough
Aug 27 23:30:22 <riot> Kero: ack, same for me.
Aug 27 23:30:32 <Kero> now all I need is more time to hack.
Aug 27 23:30:45 * riot needs a cheap .de-gprs-flatrate
Aug 27 23:31:58 <linux_galore> yeah, there are flat rate pre paid data rates for G3 in many countries now
Aug 27 23:32:07 <dcordes> riot: wanna see mokoui on my zaurus?
Aug 27 23:32:13 <riot> Kero: if you're into developing apps, just switch to python. Takes half the time ;)
Aug 27 23:32:42 <riot> dcordes: if you have some shots ready... ;) i was pretty pissed off about them not selling the device here.
Aug 27 23:32:54 <linux_galore> dcordes: look on youtube  under openmoko zaurus, there is a Zaurus with OM UI running on it
Aug 27 23:33:04 <dcordes> riot: I got mine on ebay
Aug 27 23:33:17 <dcordes> linux_galore: that should be my one
Aug 27 23:33:37 <dcordes> note that it is running on angstrom :)
Aug 27 23:33:41 <riot> dcordes: i'm gonna have a look on youtube then
Aug 27 23:33:41 <linux_galore> dcordes: aah
Aug 27 23:33:59 <dcordes> also in the videos I did not unclutter the system. Looks much more like the real openmoko now :) I removed all the gpe stuff
Aug 27 23:34:23 <dcordes> riot: http://www.youtube.com/watch?v=5nSJo07MsZM
Aug 27 23:34:58 <gcb77> How can I prevent the NEO from connecting to the gsm network?
Aug 27 23:35:01 <Kero> riot: I prefer Ruby :)  But I guess this time I'll be changing apps, not making them
Aug 27 23:35:06 <linux_galore> there is also a few pics on flickr, saw a few with a HTC phone running OM
Aug 27 23:35:07 <Kero> If I can find the source
Aug 27 23:35:13 <gcb77> Right now it just keeps pinging the netowrk
Aug 27 23:35:21 <Kero> which last time I looked seemed to be *** nowhere
Aug 27 23:35:28 <dcordes> linux_galore: I bet that's goxboxlive's htc
Aug 27 23:35:34 <dcordes> unfortunately not an omap one
Aug 27 23:35:45 <dcordes> otherwise I would love totry one of his images ona friend's htc
Aug 27 23:35:49 <riot> Kero: sorry, no ruby on angstrom
Aug 27 23:36:09 <linux_galore> dcordes: looks ugly though
Aug 27 23:36:15 <Kero> riot: someone compiled it. and if I get the toolchain up, some day, I can too
Aug 27 23:36:27 <linux_galore> dcordes: wrong screen res
Aug 27 23:36:29 <Kero> for Neo
Aug 27 23:36:53 <Kero> but now, bedtime
Aug 27 23:36:55 <dcordes> linux_galore: htc pwhone for mokoui?
Aug 27 23:37:15 <linux_galore> dcordes: yeah, everything look chunky
Aug 27 23:37:16 <dcordes> most htcs are 320x240
Aug 27 23:37:19 <dcordes> I bet
Aug 27 23:37:47 <riot> dcordes: uh, on that video.. that is not dialer2
Aug 27 23:38:35 <dcordes> not?
Aug 27 23:38:37 <riot> dcordes: und hybsches kurz&gut im hintergrund ;)
Aug 27 23:38:47 <riot> dcordes: no, mine looks different ;)
Aug 27 23:39:26 <dcordes> riot: well everything looks bit unhybsch on both videos because as I already mentioned I didn't optimize thinks for mokoui when I made them
Aug 27 23:40:03 <dcordes> riot: you told me about oe-input-manager
Aug 27 23:40:22 <dcordes> unfortunately that is not present in the angstrom feeds and I can't build right now for arm
Aug 27 23:40:34 <dcordes> is there some repo I can fetch an ipk?
Aug 27 23:40:50 <riot> dcordes: hmm.. that was part of the image, i took from the buildhost
Aug 27 23:40:58 <riot> dcordes: there are many
Aug 27 23:41:11 <dcordes> what's the most recent binary repo?
Aug 27 23:42:13 <riot> http://buildhost.openmoko.org/snapshots/2007.08/ipk/fic-gta01 << if i don't err.. that is one of the config files of my ipkg
Aug 27 23:42:35 <dcordes> great thanks
Aug 27 23:42:37 <riot> oh, a week old
Aug 27 23:43:21 <riot> but thats the most recent there is, yes
Aug 27 23:43:24 <hads> http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/ipk/
Aug 27 23:43:31 <dcordes> riot: could you possibly mean matchbox-keyboard-inputmethod?
Aug 27 23:43:44 <riot> dcordes: possible, yes. I'm gonna have a look
Aug 27 23:43:58 <riot> yep
Aug 27 23:44:06 <riot> that one is installed, according to ipkg
Aug 27 23:44:12 <riot> and matchbox-applet-inputmanager
Aug 27 23:45:00 <dcordes> both up to date
Aug 27 23:45:01 <hads> riot: That is the snaphot (i.e release snapshot) URL you posted, so it will be dated from the snapshot (19th)
Aug 27 23:45:41 <dcordes> hads: which version of both packages do I need in order to get a working keyoard-im?
Aug 27 23:45:57 <hads> No idea
Aug 27 23:46:14 <dcordes> I can't run matchbox-keyboard
Aug 27 23:47:19 <dcordes> how do you guys start your keyboard inputmethod?
Aug 27 23:47:34 <riot> geez. pygtk-dev is so simple... i wonder, why anyone would use something else..
Aug 27 23:47:51 <hads> dcordes: With the little white square thingy in the panel.
Aug 27 23:48:05 <dcordes> riot: don't have that
Aug 27 23:49:03 <dcordes> maybe the angstrom feeds need some upgrading
Aug 27 23:49:32 <dcordes> I bet there is something lacking in the /et/matchbox/session in order to get the keyboard "thingy" :)
Aug 27 23:51:20 <dcordes> riot: can you trace what happens when you press the squary thingy?
Aug 27 23:52:21 <dcordes> I'm off later
Aug 27 23:56:19 <smappy> 1~2~
Aug 27 23:56:43 <smappy> uit
Aug 27 23:57:21 <zash_se> whops, thunderstorm
Aug 27 23:57:35 <zash_se> Warning power loss detected on UPS overkill                                    
Aug 27 23:58:40 <riot> dcordes: uh. i could try
Aug 27 23:59:07 <riot> perhaps the debug-output of xdrive is sufficient
Aug 27 23:59:09 <riot> lets see
Aug 28 00:01:17 <riot> is a call to matchbox-keyboard-remote.c
Aug 28 00:02:05 <riot> (and the OM-keyboard is seriously broken in landscape mode..)
Aug 28 00:08:26 <riot>  http://www.youtube.com/watch?v=S5xj3FLb2Ew << LOL
Aug 28 00:16:48 <rwhitby> morning
Aug 28 00:19:27 <riot> geez.
Aug 28 00:19:37 <riot> anyone been playing video on the neo?
Aug 28 00:21:44 <don-o> riot: its very dated by now but mokoninka has a youtube video of Mickey trying to play an episode of Battlestar on the P1
Aug 28 00:21:54 <cdbot2> * * OM Bug 763 has been created by <autobuild-report>
Aug 28 00:21:55 <cdbot2> * * uboot-gta01-1.2.0+svnnow-r14_3e66c078003607a7d1d214c15a5f262bc1b4032f_0_2834-do_compile
Aug 28 00:21:56 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=763
Aug 28 00:23:17 <riot> don-o: yeah, i just saw that... didn't look god.
Aug 28 00:23:18 <riot> +o
Aug 28 00:24:36 <riot> don-o: and its not that dated.. two months old.
Aug 28 00:25:13 <riot> now well, i didn't expect full screen movie from my vcr, which is a 500 mhz via c3
Aug 28 00:25:22 <riot> (but it does that quite well)
Aug 28 00:32:46 <cheriff> does anyone know how I might find the proper timings to run the LCD at 240x320? I have the LCM in QVGA mode but there are artifacts due to bad timings from the s3c2410 controller ...
Aug 28 00:43:09 <SpeedEvil> how did you put it in that mode?
Aug 28 00:45:28 <cheriff> SpeedEvil: Stefan Schmidt has some patches to drivers/spi/jbt6k74.c which add support. However at the moment they are a compile-time option
Aug 28 00:45:45 <cheriff> (http://people.openmoko.org/stefan/patches/LCM-QVGA/ )
Aug 28 00:45:57 <SpeedEvil> ah
Aug 28 00:46:26 <cheriff> they are a little old, and dont all apply cleanly.. but those that fail are trivial enough to merge by hand
Aug 28 00:47:58 <cheriff> we have a little demo, of which about half the screen is mpeg playback and it struggles slightly at full VGA. Not sure if bus or cpu bound yet but QVGA plays much better
Aug 28 01:01:01 <aloril> (script) wiki RecentChanges: [[OpenMoko_Local_Groups:_Oslo]] [[Neo1973_and_Windows]]
Aug 28 01:31:55 <CIA-20> openmoko: 03laforge * r2836 10/trunk/src/target/u-boot/patches/uboot-gta02.patch: fix cosmetic issue leading to compilation problems
Aug 28 01:43:46 <abraxa_> Does anyone happen to know where I can submit .bb recipes for inclusion in the tree? Do I file a bug with the .bb files attached on the oe bugzilla? Or the openmoko one? Or how else do I go about it?
Aug 28 01:50:37 <roh> abraxa_ depends if its more openmoko or more openembedded. but attaching patches to bugzilla entries is always a good idea
Aug 28 01:50:58 <roh> there should also be the possibility to tag a bug as feature improvement
Aug 28 01:54:44 <abraxa_> roh: Well, it's two recipes: one for UriParser, one for libSpiff (XSPF playlist reference implementation) - I would assume that they might be of greater interest than just OpenMoko
Aug 28 01:58:40 <rwhitby> roh: who is meant to be looking at openmoko-dialer bug reports?
Aug 28 01:59:06 <cesarb> abraxa_: no matter which bugzilla you used, they would end up in openembedded, so that's not an issue
Aug 28 01:59:24 <rwhitby> (I have three reports, with two patches enclosed, which should make GSM calls work for many more people than they currently do, but no-one seems to be interested)
Aug 28 01:59:35 <abraxa_> cesarb: Ah okay, thanks 
Aug 28 01:59:46 <hads> rwhitby: It was only yesterday they were posted wasn't it?
Aug 28 01:59:48 <cesarb> abraxa_: but since they did not originate with openmoko, if the change is not openmoko-specific (for instance, to fix a bug which appeared only on neo so far), you should send them directly upstream (i.e. openembedded)
Aug 28 02:00:22 <rwhitby> hads: yes, but for a critical issue like actually getting the phone to work, I would have expected at least an ack by now.
Aug 28 02:00:27 <abraxa_> cesarb: Yeah, it's not directly OpenMoko related, so I'll do that. Thanks =)
Aug 28 02:00:31 <cesarb> rwhitby: commitlog says it's njp
Aug 28 02:01:15 <cesarb> rwhitby: and the last commit by him on commitlog is 2007-08-23
Aug 28 02:01:26 <hads> rwhitby: Fair enough, I'm used to things being slow moving :)
Aug 28 02:02:47 <rwhitby> hads: I'm used to projects that want community input and patches from community developers to at least respond to fixes for critical issues with a "got it, will look at it" within 24 hours :-)
Aug 28 02:03:13 <roh> rwhitby i think alphaone was working on the dialer, but i am not 100% certain who else.
Aug 28 02:03:28 * rwhitby wonders how many of the 1000 P1 owners still can't make a phone call directly out of boot yet
Aug 28 02:03:49 <roh> rwhitby propably all of them?
Aug 28 02:03:58 <jadams> rwhitby: I can't do it from any image I know of, but I can get it going out of boot in five minutes from a fresh install
Aug 28 02:04:19 <rwhitby> jadams: do you have a pin on your sim or not?
Aug 28 02:04:20 <roh> nothing which makes me wonder or shiver.
Aug 28 02:04:51 <cesarb> rwhitby: said projects don't have developers doing all the logistics...
Aug 28 02:05:41 <rwhitby> cesarb: yeah, I realise the current situation, that's why I'm looking for the name of the person to ping directly, rather than wasting everyone's time by complaining on a mailing list :-)
Aug 28 02:06:55 <roh> *sigh* alphaone is volunteer at the german thw (dunno how its called in english) and they had a lot of work to do after these stormy days.. pumping out cellars and stuff i think
Aug 28 02:06:57 <rwhitby> jadams: I have a theory (after looking at openmoko-dialer code for the last two nights after work) that it probably works for people who need to type in a sim pin (assuming their gsmd comes up cleanly), but probably doesn't work for people who don't have a pin on the sim, due to a timing issue.
Aug 28 02:07:24 <cesarb> rwhitby: makes sense
Aug 28 02:07:43 <cjb> ooh.
Aug 28 02:07:46 <cesarb> rwhitby: not only the PIM isn't enabled on mine, it's also the well-known default PIM for Claro
Aug 28 02:07:59 <jadams> rwhitby: I have no SIM
Aug 28 02:08:20 <jadams> but I know of two things to get it to work...one being stop/start in my matchbox session
Aug 28 02:08:28 <cesarb> rwhitby: my Razr fails at changing it (the long sequence of **something*something*something*** needed to change it is prematurely terminated), and I'm no longer as paranoid, so I don't care anymore
Aug 28 02:08:32 <jadams> and the other being in the logs from yesterday somewhere, but sounded more appropriate  :)
Aug 28 02:08:54 <rwhitby> roh: thanks for the info - that's what I was looking for - some indication of whether I should ping directly or not.  if the person responsible is busy doing other things then I won't bother them until they get back on line.
Aug 28 02:09:08 <cesarb> jadams: you know, the empty SIM holder also needs no password ;-) (just try to register to the network and you'll see)
Aug 28 02:09:40 <jadams> cesarb: huh?
Aug 28 02:09:42 <rwhitby> jadams: how are you making a call five minutes after boot with no SIM ?
Aug 28 02:09:44 * cesarb doesn't think anyone he knows use a SIM PIN
Aug 28 02:09:52 <jadams> rwhitby: with a SIM, just no PIN
Aug 28 02:09:54 <jadams> I can't type
Aug 28 02:09:55 <jadams> :)
Aug 28 02:10:36 <rwhitby> jadams: and your phone reliably registers with the network on every boot?
Aug 28 02:11:50 <rwhitby> roh: don't get me wrong - i'm not complaining about the response time - I just want to make sure that the person looking after the openmoko-dialer code doesn't need to wade through hundreds of outstanding bug reports to find the one I entered which has a patch to fix it
Aug 28 02:12:09 <cesarb> rwhitby: I'll take a bath right now, and later will add your patches to my overlay tree and see what happens with them
Aug 28 02:12:49 <roh> rwhitby that should not be a problem. if you find multiple bugs which should be one just say so and we tag the dups
Aug 28 02:12:54 <jadams> rwhitby: yup
Aug 28 02:13:13 <jadams> rwhitby: of coure, my SIM's never in it these days because I keep it in my real phone in case...someone...wants to call
Aug 28 02:13:21 <jadams> since incoming calls don't register for me most of the time
Aug 28 02:15:01 <hads> If I restart gsmd in the matchbox session I can mak a call (I have no SIM pin).
Aug 28 02:17:13 <Writchie> roh: looking at the RNDIS code, it looks like there are module options that are not provided which may be why garbage is sent to windows
Aug 28 02:17:19 <jadams> umm
Aug 28 02:17:23 <jadams> the sd card got wiped
Aug 28 02:17:27 <jadams> I don't know how that happened
Aug 28 02:17:37 <jadams> but I had a pygtk app I was writing on the phone that I hadn't backed up on the laptop yet
Aug 28 02:17:41 <jadams> sigh
Aug 28 02:17:48 <jadams> I didn't delete shit...my sd just got wiped
Aug 28 02:18:10 <jadams> hrm
Aug 28 02:18:17 <jadams> nm, seems to just not read the card right now
Aug 28 02:18:40 <ccube> morning
Aug 28 02:20:48 <jadams> morning ccube, somewhere  :)
Aug 28 02:21:11 <ccube> :)
Aug 28 02:21:25 <jadams> ahh, amsterdam?
Aug 28 02:22:00 <ccube> what time do u have?
Aug 28 02:22:45 <jadams> 9:22 PM
Aug 28 02:22:47 <jadams> CST
Aug 28 02:22:57 <jadams> GMT-6/DST
Aug 28 02:23:59 <ccube> uii, nice
Aug 28 02:23:59 <ccube> <-- 04:23 AM
Aug 28 02:23:59 <ccube> :(
Aug 28 02:23:59 <ccube> *tired*
Aug 28 02:24:05 <cesarb> jadams: you can always use PhotoRec if the worst happens with the data on your SD card
Aug 28 02:24:32 <cesarb> wow, just got a CSQ of 23... beating my previous record of 24
Aug 28 02:28:59 <jadams> hrm...doing an ipkg update, and:
Aug 28 02:28:59 <jadams> Configuring task-openmoko-pim
Aug 28 02:28:59 <jadams> Configuring task-openmoko-ui
Aug 28 02:28:59 <jadams> ldconfig: /usr/lib/libcurl.so.4 is not an ELF file - it has the wrong magic bytes at the start.
Aug 28 02:30:01 <SpeedEvil> that can be FS corruption
Aug 28 02:30:38 <SpeedEvil> if you write too much to the card too fast, it wedges 
Aug 28 02:30:42 <SpeedEvil> and needs a reboot
Aug 28 02:30:52 <SpeedEvil> the data isn't actually always lost.
Aug 28 02:35:43 <jadams> http://pastebin.com/m3d23bf1
Aug 28 02:36:46 <jadams> system.conf is a directory
Aug 28 02:36:58 <cesarb> rwhitby: damn, the patch had rejects
Aug 28 02:37:13 <cesarb> rwhitby: due to some evil whitespace probably
Aug 28 02:37:23 * cesarb learns not to use copy-paste to get patches from pastebin
Aug 28 02:38:03 <Aria> browser cut and paste is that exact kind of evil
Aug 28 02:38:21 <cesarb> Aria: no, it's the WHITESPACE that's evil
Aug 28 02:38:30 <rwhitby> cesarb: there is a raw view which you can wget
Aug 28 02:38:30 <cesarb> Aria: four whitespaces in an otherwise blank line
Aug 28 02:38:35 <cesarb> Aria: aka trailing whitespace
Aug 28 02:38:53 <cesarb> Aria: trailing whitespace in C source is ALWAYS evil. No exceptions.
Aug 28 02:39:09 <rwhitby> cesarb: I didn't do a patch for the doubling the timeout bug
Aug 28 02:39:22 <Aria> Well, that too. But browser cut and paste doesn't help.
Aug 28 02:39:30 <cesarb> rwhitby: I'm trying the first one only
Aug 28 02:39:40 <cesarb> rwhitby: since I have a hunch it'll be enough
Aug 28 02:39:49 <cesarb> rwhitby: if not, well, it's a useful data point ;-)
Aug 28 02:40:00 <rwhitby> cesarb: I needed to double the timeout between antenna power on and first +COPS
Aug 28 02:40:23 <rwhitby> if COPS got sent before gsmd reported power on, then nothing following would work properly
Aug 28 02:42:36 <cesarb> rwhitby: the patch is still failing with rejects for me in the places where there is evil trailing whitespace
Aug 28 02:42:49 <cesarb> rwhitby: I think it wasn't my copy-paste which ate them, but pastebin itself
Aug 28 02:42:59 <cesarb> rwhitby: can you attach it to the bugzilla report instead?
Aug 28 02:43:12 <jadams> mmm...running a 'resize image' on a phone feels good  :)
Aug 28 02:44:17 * cesarb thinks someone will have to run one of the kernel developer's anti-trailing-whitespace scripts on the whole openmoko tree...
Aug 28 02:48:11 <CIA-20> openmoko: 03rwhitby * r154 10mokomakefile/trunk/patches/openmoko-HEAD/ (libmokogsmd.patch openmoko-dialer.patch): Checked in the patches that I have submitted to bugzilla, so others can test them directly
Aug 28 02:48:25 <rwhitby> cesarb: grab them directly
Aug 28 02:51:32 <pjz> augh, why is mtn sooooo slooooow
Aug 28 02:53:18 <mwester> mtn encrypts the data, loses the key, and then resorts to a brute-force attack to decrypt the data when you need it -- nothing else can explain why it takes so much raw CPU with so relativly-little data.  At least that's my theory ;-)
Aug 28 02:53:24 <defer_> because it sucks, should have used git
Aug 28 02:54:16 * pjz LAUGHS at mwester
Aug 28 02:54:20 <pjz> defer_: I agree
Aug 28 02:54:36 <cesarb> rwhitby: the patch applied fine, compiling right now
Aug 28 02:55:09 <pjz> I'm waiting for the git guys to write git-mtn
Aug 28 02:55:18 <pjz> at which point mtn will be _over_
Aug 28 02:55:48 <cesarb> pjz: the same way SVN clobbered CVS?
Aug 28 02:56:05 <defer_> more like the same way git clobbered svn but some people didn't notice :)
Aug 28 02:56:16 * Clint chuckles.
Aug 28 02:56:22 <cesarb> defer_: nah, SVN has one special feature which git doesn't
Aug 28 02:56:26 <defer_> git-svn <3
Aug 28 02:56:29 <cesarb> defer_: if works better on Windows.
Aug 28 02:56:30 * cesarb runs
Aug 28 02:56:42 <pjz> defer_: exactly
Aug 28 02:56:53 <pjz> cesarb: not really
Aug 28 02:56:56 <mwester> everything works better on Windows!
Aug 28 02:56:59 <pjz> cesarb: I think they fixed that
Aug 28 02:57:01 <arw_> see you people
Aug 28 02:57:10 <defer_> git doesn't run on windows
Aug 28 02:57:15 <defer_> the official one at least
Aug 28 02:57:17 <mwester> See?
Aug 28 02:57:23 <cesarb> pjz: wake me when we see (urgh)explorer(urgh) integration like tortoise coes
Aug 28 02:57:30 <defer_> only on cygwin and there is a port afaik but it's not that good
Aug 28 02:57:31 <mwester> Visual SourceSafe!
Aug 28 02:58:06 <pjz> I think there's a pure java git implementation
Aug 28 02:58:11 <pjz> wonder how it works on winblows
Aug 28 02:58:26 <cesarb> pjz: java... and git?
Aug 28 02:58:34 <pjz> cesarb: oh ick, gui's get your fingers all sticky
Aug 28 02:58:36 * cesarb thinks you just mixed oil and water there
Aug 28 02:58:41 <pjz> cesarb: nah
Aug 28 02:58:50 <rwhitby> would java git be as slow as mtn ?
Aug 28 02:58:58 <defer_> no
Aug 28 02:59:03 <defer_> nothing is a slow as mtn
Aug 28 02:59:11 <cesarb> pjz: when I had to use windows, I used svn command line, but for some random reason everyone else used tortoisesvn
Aug 28 02:59:26 <defer_> and java is not _that_ slow it's actually quite fast once the vm is setup
Aug 28 02:59:32 <cesarb> pjz: fortunately, that has ended, and now I'm on my prefered kubuntu desktop
Aug 28 02:59:51 <pjz> hrm, looks like java git may have died
**** ENDING LOGGING AT Tue Aug 28 02:59:56 2007