**** BEGIN LOGGING AT Sun Oct 07 02:59:57 2007 Oct 07 03:02:23 when will the neo1973 gta01 release? Oct 07 03:06:52 ljp: huh ? what sign ? Oct 07 03:07:06 winu1: read the topic Oct 07 03:08:58 and the GTA02 ? Oct 07 03:09:05 december Oct 07 03:12:06 hopefully Oct 07 03:15:42 CVirus: a sign you should buy a Neo Oct 07 03:15:55 * doc|home want wifi :| Oct 07 03:16:03 in fact, it's the main reason I want a neo Oct 07 03:22:18 CVirus: think of it this way: the neo can't really be used as a *phone* at the moment :P Oct 07 04:00:42 openmoko: 03nbd * r3096 10/developers/nbd/ (47 files in 9 dirs): first attempt at ar6k code cleanup Oct 07 04:16:10 orospakr: yes it can. Oct 07 04:21:31 so, i put qtopia on it, and it turned off, and now it wont boot up or recongize thru usb (neo1973) any ideas guys? Oct 07 04:22:08 i do have the debug board Oct 07 04:22:20 but i want to know if theres anything else i can do Oct 07 04:22:31 i've been holding down he power button to no avail Oct 07 04:22:57 what is the mini firefox that was proposed to be put on the moko? Oct 07 04:23:30 bkruse_home: minimo? Oct 07 04:23:49 anyone have any idea? Oct 07 04:24:16 bluelightning: yes! thank you Oct 07 04:24:52 bluelightning: didnt they have some slick website? Oct 07 04:25:16 bkruse_home: not sure... I'm more interested in something based on webkit myself Oct 07 04:32:15 meshuga: take out the battery, put it back in, charge it Oct 07 04:33:35 ljp: i've done that, and when i plug it into a linux box it doesnt show up under dmesg at all..shouldnt it if its charging? Oct 07 04:34:11 yes Oct 07 04:34:22 do you have cdc-ether module loaded? Oct 07 04:35:43 ljp: well, the GSM modem is very buggy. it won't even register for me, with various SIM cards. Oct 07 04:36:38 meshuga: it won't show up in dmesg until it has enough charge to boot and respond to the usb enumeration Oct 07 04:37:01 you have to let it trickle charge for a while from a dead battery to give it enough charge to boot and request the fast charge Oct 07 04:38:14 i wonder if i use my blackberry charger if it'll be any faster then my linux laptop Oct 07 04:38:22 i wish they had a charging light Oct 07 05:03:50 orospakr: tried qtopia? :) Oct 07 05:04:00 ljp: yup Oct 07 05:04:21 again, it's not in OpenMoko, but rather the TI modem itself stops responding Oct 07 05:07:01 orospakr: does the device suspend? Oct 07 05:07:48 ljp: hm, sort of, I suppose. I can tap "Full PM" in OpenMoko (built from source, from two days ago and have it blank Oct 07 05:09:19 in my experience, the modem doesn;t survive a suspend/resume Oct 07 06:54:40 could anyone flash their local qemu's getting errors Oct 07 06:58:08 could anyone flash their local qemu. i am getting time out errors Oct 07 07:06:26 * gcb77 is going under, hoping to re-emerge running openSUSE 10.3 Oct 07 07:15:01 nbd, pong Oct 07 07:27:56 ossman: i've done some more work on ar6k and in the original version it can make requests without having to wait for them to finish Oct 07 07:28:05 ossman: do you think it would be worth adding that to the linux sdio stack? Oct 07 07:28:57 I would guess no, but show me some code and some numbers and I might be convinced Oct 07 07:29:20 i have neither Oct 07 07:29:23 :) Oct 07 07:30:00 just checking. for now i simply used the sdio_* functions and added FIXME comments Oct 07 07:30:10 ok Oct 07 07:30:18 it gets through most of the driver init now, but apparently sdio irqs are not working Oct 07 07:30:24 (on sdhci) Oct 07 07:30:36 with real interrupts? Oct 07 07:30:39 at least they aren't working with this card Oct 07 07:31:17 what do you mean? Oct 07 07:31:40 do you have the patch giving you real interrupts, or are you polling them? Oct 07 07:31:58 i have the latest code from the sdio branch of your git repo Oct 07 07:32:18 so including your patch for sdio irq on sdhci Oct 07 07:32:23 it's a ricoh controller, btw Oct 07 07:32:30 ok, then you should have real interrupts Oct 07 07:32:39 i'm not getting any Oct 07 07:33:05 I don't know if the ricoh controller has been tested. You could try removing that caps bit and see if things work. Oct 07 07:33:06 but i'm not really sure that it's the controller driver at all Oct 07 07:33:29 there have been quite a few hw issues with these cards ;) Oct 07 07:33:34 :) Oct 07 07:33:57 you have an ar6k card, right? Oct 07 07:33:59 any activity from the core openmoko folks, or are you still doing this on your own? Oct 07 07:34:01 yup Oct 07 07:34:08 i'm doing this on my own Oct 07 07:34:11 but I haven't had time for it Oct 07 07:34:20 my latest code is checked into developers/nbd in the openmoko svn Oct 07 07:34:37 i did quite a bit of cleanup on the driver source Oct 07 07:34:42 yes, you gave me an url some time ago Oct 07 07:35:16 did lots of sed on it and removed quite a bit of abstraction crap Oct 07 07:35:23 looks a bit more like a linux driver now Oct 07 07:35:26 though still quite messy Oct 07 07:35:33 heh Oct 07 07:35:51 the downside being that you'll have harder to incorporate changes from atherors Oct 07 07:35:52 -r Oct 07 07:36:08 i can deal with that Oct 07 07:36:41 i'd rather port all changes manually than deal with the code in its original crappiness Oct 07 07:37:13 that seems to be a common trait of vendor drivers :) Oct 07 07:37:45 nbd, care to talk me through how to build the driver? Oct 07 07:38:32 it builds like any regular kernel driver tree, so make -C $KERNELDIR SUBDIRS="$AR6KDIR" Oct 07 07:38:45 ah Oct 07 07:43:13 openmoko: 03werner * r3097 10/trunk/src/host/usbpath/ (5 files): Oct 07 07:43:13 openmoko: - switched to using autotools. Oct 07 07:43:13 openmoko: - figured out how to build dfu-util without libusbpath.pc Oct 07 07:45:42 nbd: Oct 07 07:45:43 [17179632.388000] BMI Get Target ID: Enter (device: 0x914ec800) Oct 07 07:45:43 [17179633.520000] Unable to decrement the command credit count register: -110 Oct 07 07:45:43 [17179633.520000] Unable to read from the device Oct 07 07:45:43 [17179633.520000] sdio_wlan: probe of mmc1:0000:1 failed with error 1 Oct 07 07:46:10 over SPI. my native controller is a bit flaky Oct 07 07:46:21 ossman: yep, i had that one quite frequently on my device as well Oct 07 07:46:47 ossman: but after retrying a few times, it got further Oct 07 07:47:22 hmmm Oct 07 07:47:26 the driver wedged something Oct 07 07:47:41 controller isn't responding to insertions anymore Oct 07 07:47:57 yep. same happened to me Oct 07 07:48:34 i wonder if it's simply the hardware being crazy Oct 07 07:48:52 no, hardware can never wedge the kernel Oct 07 07:48:56 at least not in this way Oct 07 07:49:06 so there is a driver bug somewhere Oct 07 07:49:13 (a stray host lock probably) Oct 07 07:49:17 i mean, obviously the driver is buggy as well Oct 07 07:50:12 hardware buginess usually increases with the size of the company that produced it :) Oct 07 07:50:17 ;) Oct 07 07:50:41 but now it's breakfast time. keep up the good work :) Oct 07 07:50:47 the thing is, the original driver had a delay of 1 second between hw powerup and driver init Oct 07 07:50:57 if i add that, then it dies much sooner during the init Oct 07 07:51:09 only gets to like 3 or 4 commands until everything timeouts Oct 07 09:32:39 * ScaredyCat peeks around Oct 07 09:33:06 * CM too Oct 07 09:33:23 * SpeedEvil peeks through the triangular window. Oct 07 09:33:30 no gsmd dev :( Oct 07 09:33:48 * XorA bombs the peekers Oct 07 09:33:57 * CM feels pretty bombed Oct 07 09:34:20 * ScaredyCat gets dressed Oct 07 09:35:09 * ScaredyCat promotes SpeedEvil to Senior GSMd developer Oct 07 09:35:20 openmoko: 03werner * r3098 10/trunk/src/host/usbpath/README: - README: added build instructions Oct 07 09:35:41 * ScaredyCat looks at that last comit with distain Oct 07 09:35:56 +m Oct 07 09:37:19 commies with commitment! Oct 07 09:37:41 openmoko: 03shoragan * r3099 10/branches/src/target/kernel/2.6.23.x/patches/ (gta01-vibrator.patch gta02-core.patch series): Oct 07 09:37:41 openmoko: Remove gta01-vibrator patch Oct 07 09:37:41 openmoko: Ben Dooks recommends to reimplement it using the s3c24xx leds driver. Oct 07 09:37:43 -viq Oct 07 09:37:46 :D Oct 07 09:38:31 mm, vibrators Oct 07 09:38:35 <-- random Oct 07 09:39:11 <_Shoragan> that driver didn't work for .23 anyway Oct 07 09:39:45 openmoko: 03werner * r3100 10/trunk/src/host/usbreset/ (. Makefile.am README autogen.sh configure.ac usbreset.c): Oct 07 09:39:45 openmoko: - usbreset/: moved from developers/werner to trunk/src/host Oct 07 09:39:45 openmoko: - usbreset.c: added copyright header Oct 07 09:39:45 openmoko: - Makefile: removed, using autotools instead Oct 07 09:39:45 openmoko: - usbreset.c (main): added support for selecting a specific device Oct 07 09:39:49 openmoko: - README: usage and build instructions Oct 07 10:42:13 does projects.openmoko.org GForge have a wiki plugin or should we use wiki.openmoko.org? Oct 07 10:43:14 I'd say use the wiki.opemoko.org Oct 07 10:43:46 One wiki to rule them all.. ;) Oct 07 10:44:51 openmoko: 03shoragan * r3101 10/branches/src/target/kernel/2.6.23.x/patches/ (gta-vibrator.patch series): Add support for the vibrator using the generic s3c24xx led support. Oct 07 10:48:34 CM, thanks Oct 07 11:28:47 wookie wookie Oct 07 11:39:32 openmoko: 03werner * r3102 10/developers/werner/usbreset/: - usbreset/ has moved to trunk/src/host/ Oct 07 11:44:45 XorA: There's a patch for gst-plugins-ugly I need to have applied for OM. It should be of general use for OE as well, so I need to get it into the OE repo somehow. How do I go about this aside from bugging Mickey? ;) Oct 07 11:47:25 post it into OE bugtracker abraxa_ Oct 07 11:47:43 hrw: Okay, will do that then Oct 07 11:51:24 has there been any significant changes in libgsmd since OM2007.1 ? Oct 07 11:53:45 has anyone tried libgsmdtool on recent builds? Oct 07 11:55:02 Folks, the mediaplayer is too slow playing mp3 & ogg files. Any workaround? Oct 07 11:55:46 emre_: Replace your libgstmad.so in /usr/lib/gstreamer-0.10 with this one: http://abraxa.dyndns.org:81/random/libgstmad.so Oct 07 11:56:29 abraxa_: Ok, I'm trying.. Oct 07 11:58:50 abraxa_: works great! Oct 07 11:59:02 abraxa_: bugs.openembedded.org and tell us the bug number Oct 07 11:59:46 zecke: okay, will do Oct 07 12:00:26 abraxa_: No luck, still the same. Oct 07 12:01:58 emre_: Make sure your version of the media player is >= r3034, I assume it's not the case for your build currently Oct 07 12:02:18 abraxa_: I built it a couple days ago. Oct 07 12:03:43 libgsmd -m atcmd, AT STR='AT' doesn't produce any RSTR response, what could be the problem? Oct 07 12:22:40 emre_: What symptoms do you see? e.g. what CPU usage do you see for the media player and pulseaudio on top? Oct 07 12:23:01 * * OM Bug 915 has been created by shakthimaan(AT)gmail.com Oct 07 12:23:02 * * No RSTR responses from libgsmdtool -m atcmd Oct 07 12:23:03 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=915 Oct 07 12:25:01 abraxa_: Having 1 ssh connection, mediaplayer & today app. are open. mediaplayer takes 68%cpu, pulseaudio takes 19% cpu. Oct 07 12:25:44 emre_: Did you restart the player after replacing libgstmad.so? Oct 07 12:25:54 abraxa_: sure. Oct 07 12:25:58 Hmm. Oct 07 12:26:21 emdete: nothing stupid like you're running at 100Mhz? Oct 07 12:26:29 emr: Oct 07 12:26:46 emre_: Can you please try putting http://abraxa.dyndns.org:81/random/openmoko-mediaplayer in /usr/bin and try again Oct 07 12:26:48 speedevil: Not sure, I didn't do anything to change the clock freq. Oct 07 12:26:58 shouldn't have then Oct 07 12:27:41 cat /proc/cpuinfo says bogomips: 132.71 Oct 07 12:30:39 abraxa_: I tried your mediaplayer bin, still the same.. Oct 07 12:32:22 emre_: Would you mind rebooting just to make sure all libraries are reloaded? Oct 07 12:32:39 abraxa_: Ok, rebooting now. Oct 07 12:32:45 emre_: Thanks :) Oct 07 12:35:48 abraxa_: Maybe mediaplayer gui should be less responsive, because the text players works fine. Oct 07 12:37:42 emre_: CPU usages of 50+% and 20+% respectively are a sign that both gstreamer and libgstmad aren't running in modes optimized for embedded systems - updating libgstmad and using a recent version of the media player should change that Oct 07 12:38:13 emre_: It's interesting that it doesn't do it for you because it worked for everyone else so far Oct 07 12:39:19 abraxa_: Nop, still the same. It seems like a buffer problem. when I pause and resume for the first 3-4 secs it works fine, then problems begin. Oct 07 12:51:52 openmoko: 03werner * r3103 10/trunk/src/host/usbpath/ (README devnum2path.c path2devnum.c usbpath.c usbpath.h): Oct 07 12:51:52 openmoko: Converted numbering scheme from bus+port+... with zero-based ports to Oct 07 12:51:52 openmoko: bus-port. ... .port with one-based ports, as used by the kernel. Oct 07 12:53:01 emre_: Can you try running the media player from console with GST_DEBUG=4 and pastebin the output? Oct 07 12:53:18 abraxa_: sure. Oct 07 12:53:42 emre_: Thanks :) Oct 07 12:53:54 usb numering is insane anyway Oct 07 12:57:07 abraxa_: It's around 19k lines ;) Oct 07 12:57:20 emre_: Yeah, I know, it's insane also ;) Oct 07 12:58:01 It's the only way to see what the plugins do, though... Oct 07 13:06:14 abraxa_: Here it is: http://emreturkay.googlepages.com/log.txt.bz2 Oct 07 13:06:29 emre_: Thank you Oct 07 13:08:27 can someone point me to the instructions where i can flash a rootimage only to ram? Oct 07 13:15:33 folken23: What do you mean by "only to ram"? How would you run anything if the RAM is filled with the rootfs? Maybe such a thing does exist but I haven't heard of something like this before :) Oct 07 13:16:05 probably he means sdcard? Oct 07 13:16:20 folken23: You can however put the rootfs on the SD Oct 07 13:16:29 emre_: Possibly Oct 07 13:16:47 abraxa_: i thought there was a quick way of flashing for testing purposes only. Or does this only apply to the kernel? Oct 07 13:17:10 Actually, that only applies to u-boot from what I know Oct 07 13:17:33 ah. Oct 07 13:17:44 folken23: http://wiki.openmoko.org/wiki/U-boot#Using_JTAG_to_boot_from_RAM Oct 07 13:18:00 Is that what you had in mind? Oct 07 13:19:07 abraxa_: yes, but it seems only to apply to u-boot as you stated. Oct 07 13:19:36 Unfortunately, yeah Oct 07 13:21:01 abraxa_: since the network support is inside the kernel, i could boot of nfs.. Oct 07 13:21:20 That, too Oct 07 13:28:36 heey Oct 07 13:29:55 openmoko: 03werner * r3104 10/developers/werner/dfu-util-r3103.patch: (log message trimmed) Oct 07 13:29:55 openmoko: configure.ac: added libusbpath.a and usbpath.h dependency Oct 07 13:29:55 openmoko: src/main.c: added search filter DFU_IFF_DEVNUM (bus, devnum) Oct 07 13:29:55 openmoko: src/main.c: added pseudo search filter DFU_IFF_PATH (path) Oct 07 13:29:55 openmoko: src/main.c (get_first_dfu_device, count_dfu_devices): abstracted common code Oct 07 13:29:57 openmoko: into iterate_dfu_devices Oct 07 13:29:59 openmoko: src/main.c (iterate_dfu_devices): added filtering by USB device number Oct 07 13:30:22 Hey Ellen Oct 07 13:33:03 * folken23 gets now to test the debug board, because he was stupid enough to enter 'nand erase'. :) Oct 07 13:34:51 ;D Oct 07 13:34:57 folken23: time to devirginate? Oct 07 13:37:14 folken23: hmm, when i accidentally entered nand erase clean it was kind enough to ask me whether I really meant what I just typed Oct 07 13:40:23 hi Oct 07 13:40:27 Hey nbd, How's the ar6k driver going? Oct 07 13:52:24 Henryk: it did ask me too.. but i was too interested in using the jtag once. :) Oct 07 13:53:34 hrw: yep. Dissasembly was no big deal however. Oct 07 14:35:57 emre_: Did you only try .ogg or .mp3 as well? Oct 07 14:47:28 abraxa_: did you say you have a gstreamer fix/patch that needs to go into OE? Oct 07 14:48:26 mwester: Yeah, for libgstmad in gst-plugins-ugly Oct 07 14:49:32 (love those suffixes!) Oct 07 14:50:11 If you have a OE bug number, I should be able to get synced up and commit something today. Oct 07 14:50:34 Thanks, I just want to make sure it's the Right Thing (TM) to do Oct 07 14:51:16 If it's a general fix, I think it's right -- if it's good for small ARM devices (such as the NSLU2) I'm very interested :) Oct 07 14:52:54 It is, it however only affects mp3 playback - ogg playback is still not as smooth as it could be so I'm wondering whether it's better to patch the decoders (mad/vorbisdec) or whether patching pulsesink is proving just as efficient - if so, it seems to be a better way since it would work with other decoders out-of-the-box Oct 07 15:06:10 abraxa_: what decoder are you using for ogg? tremor? Oct 07 15:07:10 LaF0rge: You're flying back? Have a safe trip if so :) Oct 07 15:07:15 Henryk: Yup Oct 07 15:07:45 abraxa_: hmm, ok Oct 07 15:07:50 Henryk: Why? Oct 07 15:07:57 I am in taipei Oct 07 15:08:11 abraxa_: because, as you said, it's basically impossible to playback ogg Oct 07 15:08:25 LaF0rge: Ohhh... okay. Well I hope you have a safe trip whenever you return then ;) Oct 07 15:08:39 thanks Oct 07 15:08:41 LaF0rge: hi! Oct 07 15:08:45 hi zecke Oct 07 15:09:11 LaF0rge: I'm glad you are not stucked in singapore Oct 07 15:09:32 zecke: yes, apparently one day earlier taipei airport was closed, but today only ~1hour delay Oct 07 15:10:11 something seems to have seriously affected taiwans international bandwidth. I'm not getting more than 60kbit from germany to here (hinet) Oct 07 15:14:55 LaF0rge: that might have something to do with the weather, perhaps? :) Oct 07 15:15:30 not that international bandwidth from Taipei has ever been particularly _good_, in my experience Oct 07 15:16:03 LaF0rge: ever tried downloading through multiple connections at once? maybe latency is killing the throughput when using only one connection ;) Oct 07 15:16:07 hi, btw. :) Oct 07 15:16:23 dwmw2: depends on the ISP. APTG/APOL is very good to europe Oct 07 15:16:31 nbd: we had great luck with that a few months ago. Split large file into 64KiB chunks and download in parallel Oct 07 15:16:39 we were trying to download full OS images for OLPC Oct 07 15:16:50 dwmw2: but openmoko taipei is completely offline at the moment, so I can't test their peering status Oct 07 15:17:06 the fact that half the undersea cables were broken in an earthquake last year didn't help much :) Oct 07 15:17:38 hinet goes taiwan->japan->US->Europe most of the time, APTG/APOL goes straight taipei->hongkong->frankfurt Oct 07 15:17:47 ah, cool. Oct 07 15:18:57 dwmw2: I'm currently trying to fully understand the entire mess with regard to large-page-nand, hardware ecc, linux mtd and yaffs2 ;) Oct 07 15:19:11 oh, and u-boot yaffs2 writing, not to forget. Oct 07 15:19:37 did you check out ubi? might be a good choice as well Oct 07 15:19:43 it really seems mtd nand support is very 512byte page centric. Oct 07 15:19:44 nbd: too new. Oct 07 15:20:09 yeah, maybe. but the code is very clean, imho Oct 07 15:20:13 the main issue is that the oob layour and even ecc algorithms don't really have a common standard all over the place Oct 07 15:20:24 logfs? Oct 07 15:20:30 LaF0rge: large-page support shouldn't be an issue Oct 07 15:20:47 we _have_ to have a flexible oob layout in order to support hardware ECC Oct 07 15:20:57 because there are numerous ways that the hardware can want its stuff laid out Oct 07 15:21:39 but we're running JFFS2 on large-page flash (1GiB of it) on OLPC, with hardware ECC. Oct 07 15:21:58 dwmw2: I didn't get that properly running here :( Oct 07 15:22:11 what did you have problems with? Oct 07 15:23:00 mh, don't remember the exact error message that's being spit out for every block at boot time. Oct 07 15:23:34 send me a unit with large-page flash (or just a socket where the flash should be; I can find chips) and I'll make it work :) Oct 07 15:24:13 dwmw2: shouldn't be that much of a problem. are you in the UK currently? Oct 07 15:24:25 I will be by the time it gets there :) Oct 07 15:24:32 Boston at the moment, visiting OLPC Oct 07 15:24:48 ok, will try to make them assemble one unit for you. Oct 07 15:24:52 ok Oct 07 15:25:01 The plan for OLPC in the long term is probably logfs Oct 07 15:25:07 although there's life in JFFS2 yet. Oct 07 15:25:11 LogFS too new? Oct 07 15:25:31 the log-structured approach doesn't scale wonderfully, but it's doing OK for now on 1GiB and there are a bunch of optimisations we can make Oct 07 15:25:36 yeah, logfs is definitely too new :) Oct 07 15:25:51 i think i will try ubi soon, because i rather like the concept Oct 07 15:26:06 Jörn has been doing good work though, and it's mostly arisen out of the design discussions we had about 'jffs3'. That's where I want to go next Oct 07 15:26:36 yeah, ubi is interesting. We will probably use ubi+jffs2 for a while, to give us partitioning &c, before we're brave enough to ship logfs Oct 07 15:29:17 dwmw2_gone: btw. just for fun - we have openwrt running on olpc, with X Oct 07 15:29:24 cool Oct 07 15:29:54 now if we could just get the libertas device working in master mode... Oct 07 15:41:08 abraxa_: I tried an mp3 now and it works far better. Has pretty much acceptable quality. But not crystal clear. Oct 07 15:42:57 abraxa_: The text players work great, but this works average. If I try to do anything even a simple tapping breaks the sound. Oct 07 15:47:52 emre_ well, i think the soundchip is somehow strange - when i play with the potentiometer of my headfones - the volume doesn't change much but the bass-level Oct 07 15:58:30 kristian-m: Note that for some reasons recent images enable "Amp Spk 3D" which makes music sound strange Oct 07 16:02:11 Henryk hmm, i tryed about every amp setting in alsamixer, no way Oct 07 16:03:43 i also found that using connectors with 3 contacts have to be manipulated to work - should not be fully pulled in Oct 07 16:04:06 maybe there still is a problem with my jack then Oct 07 16:45:51 hi Oct 07 16:47:25 Heya AciD Oct 07 16:47:53 I just heard about noe1973, it looks _really_ cool Oct 07 16:47:57 It is :) Oct 07 16:48:14 Do you have any questions? Oct 07 16:48:15 perhaps I ll wait 3 months before buying a pda :) Oct 07 16:48:39 * cjb_ie hopes his current phone lasts that long too :) Oct 07 16:49:12 not really Oct 07 16:49:28 I m just looking forward to see any news from the production :0 Oct 07 16:49:35 We all are, hehe Oct 07 16:49:44 openmoko: 03laforge * r3105 10/trunk/src/target/kernel/patches/ (lis302dl.patch series): add preliminary lis302 accelerometer input device driver for GTA02 Oct 07 16:49:56 Woot! Oct 07 16:50:17 * mjr flashes qtopia to try it out, possibly use while waiting on gsmd Oct 07 16:54:12 oooh big channel... Oct 07 16:54:24 someone convince me to buy an openmoko phone and not an iphone please :) Oct 07 16:54:43 buy both Oct 07 16:55:08 nbd, i can't even afford one of them yet :) Oct 07 16:56:36 Moniker42: Is there anything you want to do that the 1.1.1 firmware for the iPhone doesn't permit? (e.g. running 3rd-party programs) Oct 07 16:57:16 yes, i'm sure within twenty minutes of having an iphone in my pocket i'd be able to find hundreds of things i can't do on it i'd be able to do with some linux tinkering Oct 07 16:57:43 plus i live in the UK and the only option is a £40pcm (that's about $80USD) contract with O2 Oct 07 16:57:48 which i can't afford :P Oct 07 16:57:51 What more arguments do you need then? ;) Oct 07 16:58:09 iphone has a camera... Oct 07 16:58:20 and multitouch, and a bunch of accelerometer gizmos etc Oct 07 16:58:57 GTAO2 has accelerometers, too Oct 07 16:59:04 No camera and no multitouch, though Oct 07 16:59:16 multitouch might be possible Oct 07 16:59:26 Do we get a new touchscreen? Oct 07 16:59:33 no Oct 07 16:59:38 is the openmoko actually available for purchase in the UK? Oct 07 16:59:45 How would multitouch be possible then? Oct 07 16:59:56 Moniker42: GTA01 is but 02 not Oct 07 17:00:04 hmm Oct 07 17:00:08 in time for christmas? :) Oct 07 17:00:14 Yes, you should wait for 02 ;) Oct 07 17:00:40 abraxa_: http://lists.openmoko.org/pipermail/community/2006-December/000576.html Oct 07 17:00:54 "The vendor of the touch panel we are using said it might be possible. But it would be a considerable driver effort." Oct 07 17:00:58 it being multitouch Oct 07 17:01:37 Vegar: Interesting Oct 07 17:01:55 There are four wires from the touch panel. Oct 07 17:02:03 this cannot give you generic multitouch Oct 07 17:02:13 you can get more than signle-touch. Oct 07 17:02:29 Vegar: I hope it's not just some "our marketing department says that it *might* be possible without actually consulting our engineers" Oct 07 17:02:31 something like total pressure in the bounding box of all touches Oct 07 17:02:56 'proper' Multitouch certainly is not possible with the GTA01 hardware. Oct 07 17:03:23 SpeedEvil: Is it because of resistive vs. capacitive measurement? Oct 07 17:03:27 SpeedEvil: would it be possible to get the size of the bounding box? Oct 07 17:03:33 Vegar: yes Oct 07 17:03:42 where bounding box is a loose term Oct 07 17:03:47 as it's not quite that simple Oct 07 17:03:52 it's a rough idea though Oct 07 17:03:53 Moniker42: not that much openmoko related but worth the read: http://diveintomark.org/archives/2007/10/04/if-wishes-were-iphones Oct 07 17:03:56 SpeedEvil: so stylus/finger detection is possible? Oct 07 17:04:17 Vortor: _maybe_ - it depends on the rest of the harfware Oct 07 17:04:40 SpeedEvil how have you found taht out? Oct 07 17:04:51 SpeedEvil datasheet of the touch screen? Oct 07 17:07:38 OpenMoko 2007.1 is uclibc, right? Oct 07 17:08:38 kristian-m: no - knowledge of how resistive touchscreens work and the fact it's got four electrodes. I don't know what the rest of the hardware supports. Oct 07 17:08:51 Henryk, no Oct 07 17:09:25 Basically. There are two sets of panels, each with electrodes on parallel edges, and at right angles to each other. Oct 07 17:10:05 Both panels have resistance between each electrode, and when a voltage is applied between the parallel electrodes, you get a voltage gradient. Oct 07 17:10:24 So, you drive the X side, and then attempt to measure a voltage on the Y side. Oct 07 17:10:39 * roh does not get what all people want with multitouch. ive seen enough devices here in berlin which have it and none makes use of it Oct 07 17:10:47 whee, managed to lock my prepaid sim and they don't have a puk code with the package Oct 07 17:10:49 If you succeed - this means there is a touch, and you have a value for the X value. Oct 07 17:11:01 Repeat for the other axes, and you have an XY Oct 07 17:11:08 s/makes use of ot/makes more productive use than single touch of it/ Oct 07 17:11:22 measure the resistance total between both pairs of electrodes in aggregate and you get the total pressure Oct 07 17:11:36 (where pressure is ill-defined) Oct 07 17:12:00 Do more complex stuff, and you can get the approximate bounding box of all touches. Oct 07 17:13:10 Then there is driver stuff you can do to infer multitouch. Oct 07 17:13:29 if the touch point moves halfway across the screen instantly, then there has been another touch. Oct 07 17:13:38 But this gets rapidly messy. Oct 07 17:14:18 roh: multitouch is a nice buzzword, people want to have it because they've been told that it's cool ;) Oct 07 17:14:49 neo doesnt need multitouch because we can tap in morse while moving pointer Oct 07 17:15:21 EBDIC! Oct 07 17:15:25 * Henryk detex: XorA can speak morse or at least wishes he could Oct 07 17:15:26 nbd well.. i'm still waiting for any ui which makes use of that and is not just an experiment Oct 07 17:15:39 Henryk: I can spell SMS in morse :-) Oct 07 17:15:41 "multitouch" is a great way of moving one's pointer really really quickly on a laptop with a trackpad Oct 07 17:15:50 XorA: and SOS I hope Oct 07 17:16:01 Henryk: thats not as much used Oct 07 17:16:48 i don't know if that's multitouch, but being able to scroll on a trackpad simply by using two fingers instead of one, is certainly useful Oct 07 17:17:44 roh: will you be joining oedem tomorrow? Oct 07 17:18:05 Henryk i want, but i'm feeling sick somehow Oct 07 17:18:14 roh: uh Oct 07 17:26:01 Henryk so youre in novalisstr now? Oct 07 17:27:46 roh: yes Oct 07 17:28:14 whats the agenda for today? Oct 07 17:28:26 roh: sitting around and waiting for richard Oct 07 17:28:31 * roh needs to decide if he goes outside or better stays in bed Oct 07 17:28:35 SpeedEvil funny, thats exatly what I've been thinking about recently. - but i thought the driver only gives you x,y coords. - well to use the "jump fo the second finger" method we'd need a realtime extension Oct 07 17:29:26 SpeedEvil and we couuld only use "half" multitouch where one finger has to stay at the same spot Oct 07 17:30:13 SpeedEvil and we better would not use fingers at all as they touch multible pixels at once - so two - or at least - one finger would have to be a pen or a finger nail. Oct 07 17:30:28 I don't think multitouch is very useful with a 2.8" screen Oct 07 17:30:36 roh: the agenda for the next days has been changed shortly: http://www.openembedded.org/wiki/OEDEM2007 Oct 07 17:30:57 wall-sized LCD panels might have a use for multitouch Oct 07 17:30:59 hello Oct 07 17:31:00 Vegar it makes perfect sense - as the screen has no space for other input elements Oct 07 17:31:01 i have an idea Oct 07 17:31:16 how to use dvb-t Oct 07 17:31:32 * berkus jumps in anticipation of his neo adv. box.... arrr Oct 07 17:31:34 http://www.winner-netshop.de/product_info.php?info=p23144_TERRATEC-CINERGY-T-USB-XE-WHITE-BOX.html is a usb 1.1 compatible usb dvb-t stick Oct 07 17:32:41 maybe with a bit hacking of the phone (currently i have none) and adding another thing that steps the voltage up to 5V woudl be something Oct 07 17:32:57 thomasgruebler eats way too much power and besides that we have no possibile solution to decode D1 sized mpeg2 on the device Oct 07 17:33:24 mhm Oct 07 17:33:36 but with the gta02 i think mpeg should work Oct 07 17:33:48 not that i see... also what for Oct 07 17:34:06 what for? watching videos? Oct 07 17:34:06 what about mpeg4? Oct 07 17:34:07 moviez Oct 07 17:34:22 400mhz do not decode mpeg2 in D1, not even with fpu and yuf-conversion in hardware Oct 07 17:34:53 Vegar the same. not in D1 (720x576 !square pixels) Oct 07 17:35:03 my Pentium II 366MHz could do DVD!!1!1! (well, barely) Oct 07 17:35:19 Henryk: your pentium had floating point Oct 07 17:35:20 Henryk not really, it was skipping every 5th frame or so ;) Oct 07 17:35:21 PII has MMX and two cores Oct 07 17:35:32 And a _much_ bigger cache prolly Oct 07 17:35:43 is it so complicate? Oct 07 17:36:08 However - 320*240 pixels - most stuff can _almost_ be played with stock mplayer. Oct 07 17:36:15 i read in the wiki that mplayer should work to play videos. but i cannot remember what format Oct 07 17:36:16 Everything except h.264 Oct 07 17:36:36 It's around the 20-25fps range in most cases Oct 07 17:36:41 what about with hard framedrops? Oct 07 17:36:41 ok Oct 07 17:36:44 its just fscking much to do for a cpu and needs floats as well as lots of fast memory. e.g. it is not possible to decode hdtv h264 without using ddr ram. with sdram it just is not possible Oct 07 17:37:00 SpeedEvil could you switch the screen to 320x240? Oct 07 17:37:10 kristian-m: yes - sort of. Oct 07 17:37:18 You can't rotate in hardware though. Oct 07 17:37:21 Which is a problem. Oct 07 17:37:28 http://www.smediatech.com/product3370.htm doesn't the graphics chip have mpeg4 decoding support? Oct 07 17:37:30 There would need to be an optimal rotator written Oct 07 17:37:35 SpeedEvil i see - using mplayer or fbset? Oct 07 17:37:38 Maledictus: that's for 02 Oct 07 17:37:42 yes Oct 07 17:37:45 kristian-m: mplayer only Oct 07 17:37:47 mplayer already has an otimal rotator Oct 07 17:37:54 Maledictus no. smedia has no chip which does vga AND mpeg4 Oct 07 17:38:02 XorA: I have doubts. Oct 07 17:38:09 SpeedEvil: I dont Oct 07 17:38:13 * SpeedEvil hasn't looked at it. Oct 07 17:38:32 roh, don't you mean vga and h.264? Oct 07 17:38:33 roh: according to the table on that page the 3362 should do it Oct 07 17:38:37 besides that, even if you have qpal-mpeg2 or mpeg4 decoding. it does not help. dvb-t is only D1 mpeg2 Oct 07 17:38:45 at least mostly Oct 07 17:39:04 careful you'll wake zecke the master of DVB-H Oct 07 17:39:29 I suspect that _very_ low bitrate dvb-t streams might work. Oct 07 17:39:44 but not your average. Oct 07 17:39:51 Maledictus i dunno that table, ive read the datasheet and cannot remember any mpeg4 with vga operations Oct 07 17:39:59 SpeedEvil there are none. Oct 07 17:40:09 SpeedEvil and even them are fullres and not only qpal. Oct 07 17:40:12 roh: there are actually in some markets :/ Oct 07 17:40:17 XorA: psst I'm tired Oct 07 17:40:27 Even qpal I'm pretty sure. Oct 07 17:40:44 XorA: it is too loud :) Oct 07 17:40:54 zecke: heh heh, damn yanks :-D Oct 07 17:41:02 (though admittedly the only use I'm aware of is for on-screen deaf signer' Oct 07 17:41:06 SpeedEvil well.. have fun finding one. ive got that dvb reciever stuff covered. and its not happening atm. Oct 07 17:41:27 Not practical on near-term, yes. Oct 07 17:41:53 and after youve read the mpeg la licensing terms you never wanna touch it again ;) Oct 07 17:42:00 Sigh. Oct 07 17:42:10 what's this talk about gta02 in november? Oct 07 17:44:40 cb22: ha, just searched for you. I got a mail today that someone has bootet netbsd with output over serial in qemu-neo1973 :) Oct 07 17:45:01 haha, nice Oct 07 17:45:16 hows the port going along? Oct 07 17:46:01 I have not time right now because of university. But with help from the netbsd guys it should get better next week. Oct 07 17:46:46 yeah, it seems as if USB is already supported Oct 07 17:46:53 and the bluetooth device is picked up Oct 07 17:47:01 i currently read that the dvb-t receiver needs 1,4ghz cpu :D Oct 07 17:47:08 ok thats a lot Oct 07 17:47:09 where do you read that? Oct 07 17:47:16 in the manual Oct 07 17:47:18 of the stick Oct 07 17:47:25 sorry, meant cb22 Oct 07 17:47:27 Maledictus, the netbsd wiki page Oct 07 17:47:35 Maledictus: ok Oct 07 17:47:41 ah, ok Oct 07 17:47:43 http://wiki.openmoko.org/wiki/NetBSD Oct 07 17:47:45 theres a dmesg Oct 07 17:47:51 its standard ohci Oct 07 17:48:03 thomasgruebler 1.4 is bullshit for endusers. its just a question of enough cpu for mpeg2 decoding and data transfers Oct 07 17:48:19 roh: i know Oct 07 17:48:23 and usb sucks when it comes to low-cpu usage ;) Oct 07 17:48:42 there's a lot of cpu power that is needed by windows and the other parts of the surely big application Oct 07 17:48:51 cb22: yep, should always work. like bluetooth. I'm more worried about jffs2 and wifi Oct 07 17:49:13 doesn't freebsd have a flash fs? Oct 07 17:49:21 and wifi... hmm Oct 07 17:49:28 Ar6001 AFAIK Oct 07 17:49:34 nope Oct 07 17:49:49 600Mhz or ssomething PII should do fine. Oct 07 17:49:54 ramdrive + microsd card :) Oct 07 17:50:52 i had to look what's the problem at two sattelite receivers. Oct 07 17:50:57 i work for a man Oct 07 17:51:20 and there was an MPEG2decoder ic. with the ARM logo on it Oct 07 17:51:28 cb22: yep. at least at first I will only use sd. Oct 07 18:14:38 hi, is there already a picture of that HXD8 unit available? Oct 07 18:15:10 afaik no Oct 07 18:17:37 k Oct 07 18:21:37 are there any examples how to wait for I/O in a GTK+ application and do not waste CPU cycles (Battery)? Is gtk_idle_add an appropriate way ? Oct 07 18:23:15 matsimoto you could use a thread to handle the data it's gobject.timeout_add in pygobject Oct 07 18:24:50 kritian-m: I use C, but what exactly do you mean ? Iam fairly new to GTK+. Oct 07 18:25:12 kristian-m: I use C, but what exactly do you mean ? Iam fairly new to GTK+. Oct 07 18:26:25 matsimoto: no, gtk_idle is not ok Oct 07 18:26:25 matsimoto: Is it blocking I/O? If so, running it in a separate thread (via glib) might be a good idea Oct 07 18:26:26 matsimoto i'd keep the data shared and use a thread for the I/O then you can us a function registered by timeout_add to update the UI from the collected data Oct 07 18:26:54 matsimoto: depends on what you want to wait for. but generally: if it can be used in select() that would be the way to go Oct 07 18:28:47 I can use select to wait for the data in a thread but how can I display it quickly ? Is a timeout a good choice? sounds like polling for me. Oct 07 18:29:13 matsimoto: just a moment, have to look the exact function name up Oct 07 18:29:49 there should be some way to wake up the gui thread when it is waiting in the event loop Oct 07 18:30:13 it shouldn't be necessary to use timers Oct 07 18:30:31 in python it's gobject.io_add_watch Oct 07 18:30:33 matsimoto hmm, well it's not that much overhead - 0.5 seconds is a good choice and even in python i don't find any slowdown Oct 07 18:30:46 Henryk oh, thx I'll look into that Oct 07 18:31:12 matsimoto: generally the gtk main event loop already does select() for you, so there's no need for polling, waiting, threading, or anything like that. you only need to tell it the fd to select on and a callback Oct 07 18:31:12 but doesn't this waste the highest value we have ... battery power ? Oct 07 18:32:07 Henryk: ahh ok, the gtk main loop can check via select a fd is this correct ? Oct 07 18:32:37 As I said fairly new to gtk. Oct 07 18:33:33 Henryk: coudl you give me pointer to a function or something ? Oct 07 18:34:04 matsimoto: right. I'm currently looking for the docs Oct 07 18:34:52 0x70892919 ? Oct 07 18:36:06 Henryk: Do you mean: g_io_channel_unix_new () ? Oct 07 18:36:39 the python docs would be http://library.gnome.org/devel/pygobject/stable/gobject-functions.html#function-gobject--io-add-watch but I can't find that in the c docs, apparently it got shuffled around. again. Oct 07 18:37:33 matsimoto: yep, looks right Oct 07 18:37:38 Henryk: Ok, I think I understand how this works. Thanks a lot. I will try this. Oct 07 18:39:23 hi all Oct 07 18:40:07 hiho Oct 07 19:04:44 Can anyone play music with qtopia rootfs? Oct 07 19:06:43 emre_, did you see the mailing list post by ljp? Oct 07 19:07:13 yes, but he's only talking about mp3. As far as I remember I tried ogg and it wasn't working either.. Oct 07 19:15:01 if someone was going to build a soundapp on openmoko, should they build it against gstreamer or alsa or asound? Oct 07 19:15:43 meshuga: gstreamer Oct 07 19:16:15 ok thx Oct 07 19:16:22 meshuga and gstreamer should then connect to pulseaudio Oct 07 19:16:38 or alsa to pulseaudio? Oct 07 19:16:39 i got a build prior to september 3rd (as per scaredycats instructions) and now i am going to load mediaplayer on it Oct 07 19:17:04 kristian-m: gstreamer->PulseAudio->ALSA, yeap Oct 07 19:17:09 and if they wanted said app to work in, say qtopia too, would gstream>pulsaudio still apply? if you dont know its understandable Oct 07 19:18:01 hmm, did laforge do a chown -R laforge:webmaster /space/www/people on sita? Oct 07 19:18:08 meshuga: Yeah, the media player has pulsesink hardcoded Oct 07 19:19:38 meshuga if you only want to output sound and not use any codecs from gstreamer you also could only use pulse, which eases up the programming of 'have pcm, wanna put it somewhere' massively Oct 07 19:20:38 meshuga gst is for codecs and filters, pulse is for mixing and transparency, alsa is just the kernel-api for pcm and mixer/hw routing Oct 07 19:20:58 ok thx Oct 07 19:21:40 my cofounder of my company bought one, and wants to port our voip app to it, so I'm trying to collect info on it for him Oct 07 19:22:08 meshuga: What API does the app use? Oct 07 19:22:33 right now there's not a linux port of it Oct 07 19:22:38 Oh Oct 07 19:22:41 so i'm unsure of exactly why my boss bought it... Oct 07 19:22:56 but he knows i love embedded linux devices, and he enjoys new toys :) Oct 07 19:22:58 In that case, using PulseAudio directly makes the most sense Oct 07 19:23:14 jap Oct 07 19:23:28 abraxa_: sweet, i've emailed him about it. he's working on it in ubuntu and i'm supposed give him the handheld end of the week to have him start messing with it directly Oct 07 19:23:41 thx for your help guys, you answered all my sound related questions Oct 07 19:24:06 Let us know when you have more question marks ;) Oct 07 19:24:53 oh i'm sure I will :) Oct 07 19:25:08 is scaredycat's message about 'use a build prior to sept 3rd' still revelant? Oct 07 19:25:14 or should i try out one of the october builds Oct 07 19:25:31 Of what? Qtopia? Oct 07 19:26:09 oh, of openmoko Oct 07 19:26:21 qtopia worked nice, but that suspend bug plauged me so i switched back Oct 07 19:26:58 Hm, I wasn't aware ScaredyCat also provides OM images - the images on buildhost (the "official" ones) are older than Oct Oct 07 19:26:59 and the mediaplayer was giving me 'unsupported filetype' for mp3s Oct 07 19:27:09 suspend bug? Oct 07 19:27:25 abraxa_: *nods* i tried a 17th build of that and it was pretty flaky...scaredycat has 'unofficial' builds of his own Oct 07 19:27:28 alot of them Oct 07 19:27:38 ljp: ya, qtopia does like unsuspending, or going into suspend Oct 07 19:27:41 which drains the battery Oct 07 19:27:45 doesnt like^ Oct 07 19:28:16 http://buildhost.automated.it/OM2007.2/ Oct 07 19:28:19 here is scaredycat's repo Oct 07 19:28:31 and it says: Current images after 3rd September are not as good as the one from the 3rd which seems to be most stable and you can make calls from. Oct 07 19:28:46 meshuga: I myself am not having any problems with recent builds but your mileage may vary of course Oct 07 19:28:55 when i use openocd version 188 with the supplied openocd.cfg from http://wiki.openmoko.org/wiki/OpenOCD#openocd.cfg Oct 07 19:29:06 i get Error: jtag.c:1346 jtag_validate_chain(): Error validating JTAG scan chain, IR mismatch, scan returned 0x20 Oct 07 19:29:10 the qtopia image doesn't suspend because the modem doesn't survive, which also effects openmoko, I believe Oct 07 19:29:40 abraxa_: recent build being from sept 17th (the official buildhost) or recent as in latest? Oct 07 19:29:49 inmy own testing OM doesn't resume succesfully either Oct 07 19:30:30 ljp: Suspend support is very flaky. The neo often shuts off from a suspend state. is what http://wiki.openmoko.org/wiki/Qtopia_on_Neo_1973 says Oct 07 19:30:42 and i've better luck with openmoko then qtopia personally Oct 07 19:31:17 meshuga: Recent as in SVN Oct 07 19:31:22 SVN HEAD* Oct 07 19:32:50 nevermind found the error Oct 07 19:33:23 meshuga: all qtopia does is echo mem > /sys/power/state Oct 07 19:35:12 abraxa_: ok i'll try that, thanks again Oct 07 19:41:49 can somebody give me instructions to the JTAG? Does the seperate usb cable need to be connected to the neo? Should the battery be set in or not? Oct 07 19:49:23 folken: all is in wiki Oct 07 19:50:46 Hi Oct 07 19:51:00 hrw: http://wiki.openmoko.org/wiki/Debug_Board does not seem to be it. Oct 07 19:51:30 I've got some strange problem with my Neo1973 Hardware Oct 07 19:51:44 Hey lorenz Oct 07 19:51:50 I installed Qtopia *shame on me* Oct 07 19:52:02 hrw: http://wiki.openmoko.org/wiki/Connecting_Neo1973_with_Debug_Board_v2 only states the jtag cable.. Oct 07 19:52:12 and the powermanagement suspended Oct 07 19:52:16 folken: look at phase0 pages Oct 07 19:52:48 now waking up I only get a whitescreen (of death ;-)) and a vibration Oct 07 19:53:24 folken: search for "devirginate" Oct 07 19:53:28 Any ideas, what went wrong (and how to fix that) Oct 07 19:53:56 folken: and first link has proper description which works Oct 07 19:54:10 looking at http://lists.openmoko.org/pipermail/openmoko-kernel/2007-July/000183.html I guess it's some powermanagement bug Oct 07 19:54:58 but I do not have any idea on how to get any kind of console (i do have a debug board, but minicom only tells me: not connected) Oct 07 19:55:20 hrw: i found it thank you. Oct 07 19:55:48 so at the moment I have a really expensive paperweight Oct 07 19:56:32 lorenz: See http://wiki.openmoko.org/wiki/Qtopia_on_Neo_1973 under "Tips", third paragraph Oct 07 19:57:15 thank you, I read that, that won't help me any more Oct 07 19:57:27 as i do not get any connection to the phone Oct 07 19:57:37 lorenz: Maybe ljp knows more Oct 07 19:59:03 hrw: thank you, that got it back from the dead. :) Oct 07 20:05:02 lorenz: maybe try taking out the battery for a few minutes Oct 07 20:05:21 hrw: Error: target.c:1243 handle_wait_halt_command(): timed out while waiting for target halt Oct 07 20:05:25 hrw: any ideas? Oct 07 20:05:42 nope Oct 07 20:05:58 folken: its over half of year since I needed to devirginate Oct 07 20:06:02 phone Oct 07 20:06:49 XorA: I haven't had the chance to congratulate you on the new job yet - congrats! Oct 07 20:07:13 * rwhitby is currently in Urbana, IL - just near uiuc.edu Oct 07 20:07:37 rwhitby: cheers Oct 07 20:12:14 something weird with charging: battvolt:3697 chgmode:idle chgstate: fast_enabled Oct 07 20:12:51 that is, why isn't it charging as it's not full? Oct 07 20:13:03 It's plugged into a USB host? Oct 07 20:13:16 SpeedEvil, yes Oct 07 20:13:24 For how long Oct 07 20:13:26 ? Oct 07 20:14:08 half an hour or so, the battery was completely empty Oct 07 20:15:04 ljp: already tried that Oct 07 20:15:07 is batvolt going up? Oct 07 20:16:13 tt Oct 07 20:16:17 i think it was charging at first, but then the phone rebooted, and now it seems to be going down Oct 07 20:16:24 Odd. Oct 07 20:16:28 I'd try replugging USB Oct 07 20:18:00 that helped, now it's battvolt:3791 chgmode:fast_cccv chgstate: fast_enabled present Oct 07 20:18:19 charging then rebooting when first charging after plugging in battery was a common sight on gta01bv2. i think that got fixed in the meantime Oct 07 20:20:38 Henryk, hmm, this is v4 though. do you remember what it was about, or if there was a bug number for that? Oct 07 20:21:09 anyone any other idea on how to make a hard-reset for the neo? Oct 07 20:21:37 i'm getting some pulsaudio/gstreamer error when i run Media Player. It works though since I hear screentaps and the opening sound Oct 07 20:21:38 unplug battery, wait 10 min, replug Oct 07 20:21:41 any ideas? Oct 07 20:22:02 the dialog box goes too quick so i cant see the exact error Oct 07 20:22:36 ifor some reason even though it has an 'ok' box it just disappears immediately Oct 07 20:24:08 meshuga: Try "ipkg install gst-plugin-pulse" Oct 07 20:25:41 XorA|gone: Mind adding the depency of gst-plugin-pulse to openmoko-mediaplayer2_svn.bb? Oct 07 20:25:46 lorenz, can you try devirgination? Oct 07 20:27:00 tuukkah, what do you mean with devirgination? Oct 07 20:27:12 abraxa_: will do, thx Oct 07 20:27:20 lorenz, http://wiki.openmoko.org/wiki/Devirginator Oct 07 20:29:23 well actually, I already tried connecting with the debug board with serial, and currently I am desperately trying to make OpenOCD compile and/or (for the precompiled) find the debug board Oct 07 20:30:21 what about, did you try serial over usb? Oct 07 20:32:31 serial over debug board Oct 07 20:32:40 (and thus over usb) Oct 07 20:32:56 which sshd does openmoko use? i can ping it but it doesnt appear to have sshd running Oct 07 20:33:14 dropbear Oct 07 20:33:34 ok thx Oct 07 20:34:38 weird, running 'dropbear' doesnt error but nothing happens Oct 07 20:35:28 damn, i can ssh locally, but not via usb, odd. Oct 07 20:35:57 have uyou done the ifconfig usb0 192.168.0.200 thing? Oct 07 20:36:05 ya Oct 07 20:36:07 i can ping fine Oct 07 20:36:33 wait some moment.. reverse lookups? Oct 07 20:38:24 i can ping from the pc to the handheld but not back Oct 07 20:38:40 and i know there isnt any firewalls or anything Oct 07 20:38:53 i did have to add a explicit .202 route since i use 192.168.0.x Oct 07 20:40:24 before I try to devirginate (hopefully I will get that OpenOCD compiled and configured to work with the debugboard): does anyone know what the small battery next to the AUX button is for Oct 07 20:41:24 (might it be enough to remove that in order to reset the neo?) Oct 07 20:42:09 Any GTK+ specialists around? Oct 07 20:42:26 woot, got it working on 192.168.5.x Oct 07 20:47:31 aplay may work or not work Oct 07 20:47:42 sometimes there is this: Oct 07 20:47:48 aplay: main:545: audio open error: Device or resource busy Oct 07 20:47:59 because pulseaudio is using /dev/snd/pcmC0D0p Oct 07 20:48:29 using the alsa-plugins for pulseaudio, it is ok Oct 07 20:48:44 aplay -Dpulse file.wav Oct 07 20:49:15 but these packages are not installed by default :/ Oct 07 20:51:08 hi, I have a question. How can I remove the rootfs partition? Oct 07 20:52:57 bl4d3: http://wiki.openmoko.org/wiki/Nand_erase Oct 07 20:53:20 bl4d3: be careful with this command :-) Oct 07 20:54:11 ok, thasnk you :) Oct 07 20:54:45 thanks Oct 07 20:55:33 roh: I take it the use of the GTK+ pixmap engine is deprecated in favor of Thomas' GTK+ engine? Oct 07 21:12:58 * abraxa_ seriously wonders what good gobject's flexibility is if GTK+ introduces the concept of "style properties" to create an artificial barrier Oct 07 21:14:08 ok any other idea except devirginating? I am not able to connect via JTAG Oct 07 21:14:19 seems it is really a brokenmoko now Oct 07 21:15:04 sounds pretty bad then Oct 07 21:16:08 indeed, that summing of the vibrator really sounds bad (at least I have not lost my humor) Oct 07 21:18:30 zilux, I erased it, thanks Oct 07 21:19:56 ok :-) Oct 07 21:20:45 lorenz: Am I getting this right? When you turn it on it keeps vibrating but nothing else happens? Oct 07 21:21:12 it keeps vibrating and has the screen backlight on, fully white Oct 07 21:21:18 nothing else happens Oct 07 21:21:31 no JTAG, serial or usb connectivity Oct 07 21:21:34 Sounds to me like something *really* low-level is broken... Oct 07 21:22:09 ok, that will make me test the FIC support Oct 07 21:22:09 * abraxa_ wonders what the layer below u-boot is Oct 07 21:22:29 now i'm getting 'gstreamer error: Resource not found' Oct 07 21:22:39 meshuga: When doing what? Oct 07 21:22:49 loading up media player Oct 07 21:23:05 i did a ipkg update && ipkg upgrade and then installed the pkg you told me too Oct 07 21:23:08 meshuga: Where was the file located that you played when you closed it? Oct 07 21:23:17 oh shit Oct 07 21:23:20 thats the problem, thx :) Oct 07 21:23:24 ;) Oct 07 21:23:30 for some reason my microSD isnt reading Oct 07 21:23:36 that's a really good question, i'd assume: either there is a really minimal first stage bootloader located at the start address of the processor Oct 07 21:23:57 or there is no stage below u-boot, and u-boot is located there Oct 07 21:24:30 but I'd rather think there is a first-stage bootloader Oct 07 21:24:30 lorenz: I'd assume that there's something below u-boot and it's borked Oct 07 21:24:45 I highly doubt the CPU itself is broken Oct 07 21:27:48 ok i copied the mp3 over now i get a 'gstreamer: internal data flow error" Oct 07 21:28:01 and this one doesnt go away Oct 07 21:28:53 well, does any one have his "service card" here, I have mine at work, and I don't know the @brokenmoko.com mail-address Oct 07 21:29:07 meshuga: Can you pastebin your contents of /usr/lib/gstreamer-0.10/ please? Oct 07 21:29:17 and the earlier I ask for replacement the earlier I get one ;-) Oct 07 21:29:20 help@brokenmoko.com Oct 07 21:29:26 thanks Oct 07 21:29:29 np Oct 07 21:30:43 http://www.pastebin.ca/729331 Oct 07 21:31:00 * * OM Bug 916 has been created by gcasse(AT)oralux.org Oct 07 21:31:01 * * Alsa plufin for pulseaudio Oct 07 21:31:02 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=916 Oct 07 21:38:35 So let's see how good the support from FIC is (and I hope they have a sense of humor, I could not resist to use "Just another brick on the desk" as topic) Oct 07 21:39:03 hehe :) Oct 07 21:39:23 lorenz: there is a 'kickstart' or something ROM in the SoC Oct 07 21:39:28 jumpstart? Oct 07 21:39:32 Flashdance? Oct 07 21:39:37 anyway - something like that Oct 07 21:39:51 48K or so of mask-programmed ROM - which it runs first Oct 07 21:40:30 SpeedEvil: What are the chances of that being borked? Oct 07 21:40:35 abraxa_: zero Oct 07 21:40:41 abraxa_: any luck with that pastebin? Oct 07 21:41:00 abraxa_: it's mask programmed - it cannot be altered anymore than you can add a new floating point unit. Oct 07 21:41:23 SpeedEvil: Well yeah. I'm just wondering cause his unit did run before, from what I understand... Oct 07 21:41:52 abraxa_: we have verified that the voltage on the battery is OK? Oct 07 21:42:43 meshuga: Hm. You have all plugins that gstreamer needs so if that file really is an mp3 then it should play Oct 07 21:43:03 meshuga: Try another file? Oct 07 21:43:06 abraxa_: it is. i took the microsd out of my blackberry Oct 07 21:43:08 i've tried all of them Oct 07 21:43:12 like 30 or 40 Oct 07 21:43:21 and i listen to these tunes everyday Oct 07 21:43:23 ok and it is really help@brokenmoko.com ? because rt@openmoko.org automatically replied that it could not route my mail Oct 07 21:43:25 Odd. Oct 07 21:43:41 lorenz: thats what the card says. Oct 07 21:43:45 email support Oct 07 21:43:49 Sigh. Oct 07 21:43:50 support@openmoko.com Oct 07 21:44:12 ok, one more "-" for FIC on my list Oct 07 21:44:59 abraxa_: seems like a common gstreamer issue :\ https://bugs.launchpad.net/ubuntu/+source/gst-plugins-ugly0.10/+bug/136264 Oct 07 21:48:13 lorenz: brokenmoko.com isn't officially in use, so no "-" for FIC when trying to send mail there ;) Oct 07 21:49:33 * SpeedEvil notes brokemywookie.com is not registered. Oct 07 21:50:32 well, but it's officially printed on that paper-card that lies in the box Oct 07 21:50:50 It...is?! Oct 07 21:50:50 The card lies! Oct 07 21:51:18 Well okay, that's just not so smart on FIC's side then. Oct 07 21:51:33 I got a card too. Oct 07 21:51:56 well that's a small little minus, ok?! ;-) Oct 07 21:52:02 'for any other customer support issues, please contact: support@openmoko.com" Oct 07 21:53:39 hmm, I'll check that as soon as, I can see my card (it's at work), but I am really sure that for a broken neo the address was [something]@brokenmoko.com Oct 07 21:53:57 no Oct 07 21:54:01 SpeedEvil: you should start a pr0n site with brokemywookie.com Oct 07 21:54:05 'HAveing trouble?' Oct 07 21:54:35 maybe there are different revisions of that card Oct 07 21:54:53 if you suspect the device shipped to you is defective, within 14 days of your shipment date, please contact help@brokenmoko.com Oct 07 21:54:54 i got this neo last week Oct 07 21:55:04 yup, thats what my card says too Oct 07 21:55:10 ok, mine's from August Oct 07 21:55:17 the daily quick question: Ekiga is the only VoIP/SIP program actually running on OpenMoko? Oct 07 21:55:22 Mine too (aug) Oct 07 21:55:40 meshuga: Does "gst-launch playbin uri=file:///foo.mp3" play? Oct 07 22:00:52 I guess you don't have a clue either, SpeedEvil, howto unbrick/unbreak/what ever my phone? or at least what happend to it? or do you know, what the battery next to the AUX button saves? maybe unsoldering might reset some memmory and make u-boot boot? Oct 07 22:05:00 ok, anyway, thanks to all for their help, I'll have some sleep Oct 07 22:05:45 lorenz: the battery next to the aux button saves power managment chip stuff, and the RTC Oct 07 22:05:52 It should not cause it to not boot Oct 07 22:06:11 abraxa_: http://www.pastebin.ca/729364 Oct 07 22:06:14 remove the battery overnight - if you ahve't done that Oct 07 22:06:16 abraxa_: no, same error, more verbosity though Oct 07 22:06:41 lorenz: ya try removing the battery overnight. i thought i bricked mine. wouldnt respond to anything. finally just plugged it into my blackberry charger overnight Oct 07 22:06:44 now it works like a champ Oct 07 22:07:20 ok, I'll do that Oct 07 22:07:28 meshuga: If you can, pastebin me the output of "GST_DEBUG=4 gst-launch ...", please Oct 07 22:08:20 SpeedEvil, are you sure powermanagement cannot cause that trouble? as it started right after some bricky suspend qtopia does Oct 07 22:08:33 bricky? Oct 07 22:08:40 No. Oct 07 22:08:58 I suspect you can configure the power chip in such a way it will not charge or boot. Oct 07 22:09:12 so cat 'mem' > /sys/power/state is bricky? Oct 07 22:09:47 don't know what qtopia does, only know that wiki mentions it to be, let's call it unstable Oct 07 22:10:09 well, then suspend/resume on the neo is unstable Oct 07 22:10:10 lorenz: do you have a multimeter? Oct 07 22:10:24 /DMM Oct 07 22:10:30 yes of cause Oct 07 22:10:56 (even an oscilloscope, if that will help) Oct 07 22:11:08 Have you measured the battery voltage? Oct 07 22:11:34 of cause, it's within tolerance Oct 07 22:11:42 Which is? Oct 07 22:12:06 (about) Oct 07 22:12:43 both batteries are ok, they are around the 3m7V printed on them Oct 07 22:12:53 s/3m7/3.// Oct 07 22:12:58 3.7V is ratehr low. Oct 07 22:13:05 But it should boot. Oct 07 22:13:11 Odd. Oct 07 22:13:37 i can try some nokia battery tomorrow Oct 07 22:13:58 that's almost identical, if i am right Oct 07 22:14:01 It should boot down to 3.2V or so. Oct 07 22:14:11 maybe a bit lower Oct 07 22:15:44 well, obviously it has quite a lot of power to fuel the screen and vibrator Oct 07 22:15:56 so I don't assume the battery is too low Oct 07 22:17:13 It often will do something even if the battery is low. Oct 07 22:17:28 It can get into wierd states. Oct 07 22:17:46 anyway, i have to work hard tomorrow (though the part with the neo will be postponed, i guess) Oct 07 22:17:51 When it did it to me, I hooked the battery up to a lab PSU set at 4.20V and current limited at 0.5A Oct 07 22:17:58 And charged it that way. Oct 07 22:18:02 And then it worked fine Oct 07 22:18:26 ok I'll do that tomorrow Oct 07 22:19:31 SpeedEvil: for your in-phone current measurements, how did you connect the probes to the battery/Neo circuit? Oct 07 22:19:39 anyway: if that's really the problem: that should not happen for the poor end-consumer Oct 07 22:20:13 I'm expecting to use foil tape on opposite sides of a thin plastic, and place between battery and Neo contacts, but is there a better way? Oct 07 22:20:32 mwester: that's what I basically did. Oct 07 22:20:44 Take thin bit of card. Oct 07 22:20:54 4mm*40mm or so Oct 07 22:21:09 and now some copper tape - 2mm or so. Oct 07 22:21:16 bye Oct 07 22:21:19 put down the middle of one side, and back up the other side. Oct 07 22:21:35 firm down heavily to get it stuck best. Oct 07 22:21:41 now, cut off the end. Oct 07 22:21:53 (leaving you with two bits of copper) Oct 07 22:22:18 solder connections to the ends Oct 07 22:22:29 (away from the battery side Oct 07 22:22:39 now, attach to meter. Oct 07 22:22:49 Any reason to measure anything from the "T" terminal? (that's just a sensor of some sort for the charge circuit, right?) Oct 07 22:22:55 yes Oct 07 22:22:57 I diddn't Oct 07 22:23:05 temperature/thermistor Oct 07 22:23:11 ah Oct 07 22:24:15 Thanks. I'm going to test some simple hacks to the gsmd script to power off the GSM modem when we shut down gsmd. Oct 07 22:24:40 It seems the only sure way to know if the modem is off is to measure the current draw. Oct 07 22:25:15 just at@poff Oct 07 22:25:19 and it goes off Oct 07 22:25:21 also. Oct 07 22:25:29 if you don't do that and just shut down Oct 07 22:25:39 the modem doesn't know you've shut down Oct 07 22:25:45 and will ring Oct 07 22:26:07 So it's not only still on, it's on the network... Oct 07 22:26:36 yes Oct 07 22:26:40 which is a serious problem on an airplane. Oct 07 22:27:09 Or near certain construction sites in mountainous regions (blasting zones) Oct 07 22:27:39 there's an app we need, remote detonator ;) Oct 07 22:28:03 AntonTakk, that'll get me sent to Gitmo in a hurry! :) Oct 07 22:28:11 heh Oct 07 22:28:28 abraxa_: hrm probably 30 screens of info and > didnt worked to a txt file, any ideas the best way to capture it? Oct 07 22:28:33 maybe i should turn debugging down Oct 07 22:28:39 who doesn't want a free trip to the caribbean :) Oct 07 22:29:54 * mwester-laptop wonders if the American dollar is doing better against the various caribbean currencies... Oct 07 22:29:59 meshuga: Level 4 is needed cause we otherwise don't see why the plugins don't mingle - gstreamer seems to fork when being launched, so stdio redirections don't seem to work as usual... unfortunately I don't know how to catch it either Oct 07 22:30:19 abraxa_: ok i'll figure something out Oct 07 22:33:17 meshuga: have you tried redirecting stderr too? Oct 07 22:34:13 maybe it's still strong against the lire Oct 07 22:54:12 Vegar: i did not Oct 07 22:54:23 abraxa_: www.hadm.net/~kenneth/putty.log Oct 07 23:02:45 meshuga: (gst-launch-0.10:1245): GStreamer-WARNING **: pad mad0:src returned caps which are not a real subset of its template caps Oct 07 23:03:01 meshuga: So it's libgstmad's fault, apparently... that is weird. Oct 07 23:03:44 meshuga: Does it still occur if you use my library from http://abraxa.dyndns.org:81/random/libgstmad.so ? (put it in /usr/lib/gstreamer-0.10/) Oct 07 23:10:03 works! Oct 07 23:10:11 now i'll try it from the gui Oct 07 23:12:32 abraxa_: sucess! thank you soo much Oct 07 23:13:23 meshuga: Good, though that error was a little strange... but okay Oct 07 23:15:58 im impressed with the speaker quality, actually Oct 07 23:16:10 stereo too Oct 07 23:16:48 You're the first one to say that Oct 07 23:17:06 abraxa_: when my deck got stolen i listened to music on a samsung a920 for several months in my car Oct 07 23:17:08 Personally the display still to this day keeps me in awe :) Oct 07 23:17:12 these speakers are *way* better Oct 07 23:17:15 ya definitely Oct 07 23:17:23 this screen is on par with my zaurus one :) Oct 07 23:17:39 sadly GTA02 loses stereo speakers to make room for wifi circuitry, or so i heard Oct 07 23:17:45 Yup Oct 07 23:17:51 hrm this is skipping out though Oct 07 23:18:12 like glitching once or two every 30 seconds Oct 07 23:18:13 or more Oct 07 23:18:31 meshuga: You might not be using a recent enough version of the media player then Oct 07 23:18:55 http://abraxa.dyndns.org:81/random/openmoko-mediaplayer in /usr/bin should fix that Oct 07 23:18:56 abraxa_: i just ran 'ipkg upgrade' after updating about 30 mins ago Oct 07 23:19:01 did the package name change? Oct 07 23:19:06 Nope Oct 07 23:19:13 i'll try it anyways Oct 07 23:19:24 But those packages come from buildhost unless you added alternative feeds Oct 07 23:19:33 Buildhost hasn't updated for weeks Oct 07 23:19:35 ahh i have not yet Oct 07 23:20:10 do you know of any good feeds to add or is setting up a buildenviro the only way to stay current? Oct 07 23:20:25 Is anybody else having trouble building today? I'm failing at webkit-gtk Oct 07 23:22:03 meshuga: #1: Download http://buildhost.automated.it/OM2007.2/packages/scaredycat.conf and put it in /etc/ipkg/ Oct 07 23:23:14 sweet thats the one i planned on using :) Oct 07 23:23:58 it does still pop a little bit, but it sounds better then the old one Oct 07 23:25:04 meshuga: #2: Create a similar file, just with http://ipkg.nslu2-linux.org/feeds/openmoko/unstable/ as base path Oct 07 23:25:10 at around 40% cpu Oct 07 23:25:24 ScaredyCat's is enough already but always better to have more than one repo Oct 07 23:30:02 should i remove the buildhost ones since they're outdated? Oct 07 23:30:32 They'll be up-to-date again soon so no need Oct 07 23:30:39 oh ok Oct 07 23:30:45 I'm sure ipkg is smart enough to not use it till then Oct 07 23:32:18 wow i added both those repos, did update and upgrade said 'nothing to be done' Oct 07 23:32:36 and i can tell its downloading from both those repos Oct 07 23:32:58 http://buildhost.automated.it/OM2007.2/OpenMoko-openmoko-devel-image-glibc-ipk-P1-September-Snapshot-20071007-fic-gta01.rootfs.jffs2 Oct 07 23:33:01 this is the rootfs i used Oct 07 23:33:28 abraxa_: thanks again man for all your help. i don't know how I would of got sound working without irc :) Oct 07 23:39:30 meshuga: Yeah, no prob - the patched libgstmad isn't in the OpenEmbedded repo (yet), so consider this a temporary situation Oct 07 23:41:11 if I remove my "packages"-dir in my local openembedded tree: will it be regenerated by mtn pull && mtn update? Oct 07 23:44:05 thomasg: Rename it and try? Oct 07 23:45:54 abraxa_: good idea :) Oct 07 23:47:31 thomasg: you could also use mtn revert Oct 07 23:48:41 rwhitby: By the way - it would be very convenient if you could provide a ready-to-use config file we can put in /etc/ipkg, just like http://buildhost.automated.it/OM2007.2/packages/scaredycat.conf Oct 07 23:49:09 rwhitby: OE already told me, but thanks Oct 07 23:49:28 mtn revert --missing for restoring the whole tree Oct 07 23:53:29 Anyone know why the status bar app at the top fails to start on boot sometimes? Oct 07 23:54:41 tRubl: It supposedly happens because of the battery panel plugin, though the bug is still unfixed Oct 08 00:19:36 hrm now i'm trying to get gsm working. I decided to just use cu and see if i can get the 'at-command interpreter ready' prompt before trying to dial from the handset Oct 08 00:19:43 and it just sits at 'Connected' Oct 08 00:19:54 i tried powering the gsm modem down and then back up, but same results Oct 08 00:20:48 abraxa_: thanks, I'll check it out Oct 08 00:24:07 saw this in the wiki Oct 08 00:24:08 http://wiki.openmoko.org/wiki/Manually_using_GSM#Connecting_to_GSM_Modem_.28using_cu.29 Oct 08 00:25:14 meshuga: which might not mean good things for you Oct 08 00:25:28 but I had to power off and power on my phone a billion times until it worked right Oct 08 00:30:00 paroneayea: thats what i was trying Oct 08 00:30:05 i will try a billion times :P Oct 08 00:34:26 meshuga: GTA01 or GTA02? Oct 08 00:35:49 meshuga: On GTA01, examine the most recent /etc/init.d/gmsd script for the proper power-up sequence (skip the sleep statements). Oct 08 00:43:56 mwester: ok will try that (i have GTA01) Oct 08 00:51:22 http://www.pastebin.ca/729466 Oct 08 00:51:30 gsmd says 'modem dead!' Oct 08 00:51:36 that isn't a good sign is it :P Oct 08 01:06:11 meshuga: check your /etc/init.d/gsmd script - if it does not contain a line similar to "stty -F $GSM_DEV -echo" then you need to upgrade to the latest libgsmd package. Oct 08 01:09:01 k i'll check that Oct 08 01:09:01 root@fic-gta01:~$ ipkg install openmoko-browser2 Oct 08 01:09:02 Bus error Oct 08 01:09:05 any idea on that? Oct 08 01:09:31 tty -F "$GSM_DEV" -crtscts Oct 08 01:09:31 [ -n "$GSM_POW" ] && ( echo "0" >$GSM_POW; sleep 1 ) Oct 08 01:09:31 stty -F "$GSM_DEV" -echo Oct 08 01:09:33 public transportation employees on strike? Oct 08 01:09:35 mh Oct 08 01:09:38 so this looks ok? Oct 08 01:09:39 i think my neo is broken Oct 08 01:09:58 In what way? Oct 08 01:10:01 Did you blend it? Oct 08 01:10:10 it doesnt do anything Oct 08 01:10:14 meshuga: yes, looks ok. Oct 08 01:10:18 you can use it to test the blender now, i guess Oct 08 01:10:23 borg_: how does this make you feel? Oct 08 01:10:34 pretty bad Oct 08 01:10:44 borg_: More seriously - what symptoms. Oct 08 01:10:52 i made an ipkg upgrade Oct 08 01:11:02 and after 2 minutes or so it stucked Oct 08 01:11:09 * mwester-laptop sees that Speedevil was emulating a news reporter Oct 08 01:11:10 then i looked at the neo and it was off Oct 08 01:11:26 Actually 'Eliza'/ Oct 08 01:11:29 i thougt it was gone into standby or something Oct 08 01:11:40 i pulled the battery in and out Oct 08 01:11:46 but it doesnt start Oct 08 01:12:07 you need to pull the battery, for several mins, then leave charge for an hour Oct 08 01:12:17 i will try it Oct 08 01:12:24 It can get confused and not charge from USB Oct 08 01:12:26 but the battery should be full Oct 08 01:12:47 dmesg on my pc says Oct 08 01:12:49 [36823.691019] usb 3-2: device descriptor read/64, error -71 Oct 08 01:12:53 when i put it in Oct 08 01:13:16 no idea what this means Oct 08 01:14:28 Possibly means that u-boot on the neo is not able to initialize the USB on the neo end. Oct 08 01:14:57 i do not have the debug board Oct 08 01:15:07 but i cant have done this with the ipkg upgrade, can i? Oct 08 01:15:17 Nothing so far indicates that you would need the debug board. Oct 08 01:15:46 but why should it do something good to charge it? Oct 08 01:15:52 i hang on my pc for weeks Oct 08 01:16:01 the battery should be full ;) Oct 08 01:16:07 s/i/it/ Oct 08 01:16:11 Follow what Speedevil suggested, and see if it doesn't work. It may have stopped charging some hours ago, and have a dead battery. Oct 08 01:16:28 i will try it of course ;) Oct 08 01:19:25 damn every package i install gives me 'bus error' Oct 08 01:19:40 is arm4l a bad architecture? Oct 08 01:20:09 meshuga: you probably upgraded a library upon which ipkg depends. There's a limit to what you can upgrade with ipkg it seems. Oct 08 01:21:39 openmoko: 03werner * r3106 10/trunk/src/host/dfu-util/ (configure.ac src/main.c): (log message trimmed) Oct 08 01:21:39 openmoko: configure.ac: added libusbpath.a and usbpath.h dependency Oct 08 01:21:39 openmoko: src/main.c: added search filter DFU_IFF_DEVNUM (bus, devnum) Oct 08 01:21:39 openmoko: src/main.c: added pseudo search filter DFU_IFF_PATH (path) Oct 08 01:21:39 openmoko: src/main.c (get_first_dfu_device, count_dfu_devices): abstracted common code Oct 08 01:21:43 openmoko: into iterate_dfu_devices Oct 08 01:21:45 openmoko: src/main.c (iterate_dfu_devices): added filtering by USB device number Oct 08 01:21:54 root@fic-gta01:~$ ipkg info ipkg Oct 08 01:21:55 Bus error Oct 08 01:21:57 damnit :P Oct 08 01:22:12 maybe i should reflash Oct 08 01:35:29 oh rebooting fixed it Oct 08 02:22:25 online with gprs again, finally Oct 08 02:25:19 sagacis|neo: i hope you have unlimited data rates! Oct 08 02:25:19 :) Oct 08 02:26:10 I order it every month. AT&T has yet to get my bill right. Oct 08 02:27:19 bummer. browser2 crashed Oct 08 02:27:45 pidgin works pretty well Oct 08 02:28:09 any screenshots of browser2? :) Oct 08 02:28:31 sure. what do you want to see? Oct 08 02:29:16 anything really, the only screenshots that i saw of a browser on the new were the galeon ones i think. Oct 08 02:29:24 s/new/neo/ Oct 08 02:29:25 xaid meant: anything really, the only screenshots that i saw of a browser on the neo were the galeon ones i think. Oct 08 02:30:54 http://scap.linuxtogo.org Oct 08 02:31:32 * SpeedEvil should have taken some screenshots of FF Oct 08 02:31:49 sagacis|rebuild: thats perfect, thank you :) Oct 08 02:34:48 ok. 2 more, showing the search/addr pane and the tab list pane Oct 08 02:35:12 @ http.scap.linuxtogo.com Oct 08 02:35:22 errrm Oct 08 02:36:36 that looks pretty straightforward to use. i wasn't sure how the multiple tabs would be implemented on a mobile device with limited screen space. Oct 08 02:37:30 by the way, the on screen keyboard, does it support anything other than english? if you set the LC_ALL to something other than english, would it show the correct characters? Oct 08 02:38:18 Don't know. Oct 08 02:38:48 I'll get qemu running and check it out. thanks again for posting the screenshots. Oct 08 02:39:05 Just posted a recursive screenshot. :D Oct 08 02:39:39 :D Oct 08 02:40:59 hi Oct 08 02:51:42 well,I accomplished one of my goals. I'm going to hit the sack. Oct 08 02:56:00 * * OM Bug 917 has been created by gnomecn(AT)gmail.com Oct 08 02:56:01 * * libmokojournal2 build failure: cann't find gtk-doc.make Oct 08 02:56:02 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=917 **** ENDING LOGGING AT Mon Oct 08 02:59:57 2007