Re: Yaouh! out (update for tangogps maps)
On Fri, Jan 16, 2009 at 04:12:19PM +0100, Tilman Baumann wrote: Untested code: import httplib conn = httplib.HTTPConnection(tile.openstreetmap.org) conn.request(HEAD, /file...) r1 = conn.getresponse() print r1.status, r1.reason etag = getheader(ETag) print etag And if stuff was new, make GET instead of HEAD It sure would be more efficient if OSM supported the If-Modified-Since: header in a GET request. Then you wouldn't need the HEAD request. Also, consider using the same connection for multiple requests. It ought to help a lot with such small files. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Booting Debian from SD-card (Was: [fso] sending pin failed)
On Sat, Jan 17, 2009 at 09:07:08PM +0100, Tomas Nackaerts wrote: i was also unable to boot a distribution like debian or hackable from a 8gb sd-card, but i'm not sure it is related. You're being cryptic here. Please explain how you know it isn't working. Anyway, here's how I installed the Debian distribution to an 8 GB SD-card on a Freerunner (still with Om 2007.2 in the NAND flash ROM): --- begin Debian on Freerunner HOWTO --- Purpose: Install Debian on Neo Freerunner/1973 such that the default u-boot setup works and dpkg, which doesn't support vfat partitions, is still happy. Strategy: Create two partitions (vfat+ext2) but install as if we only use one partition. Manually copy the kernel image to the vfat partition after installation or upgrade of the kernel package. The details: Read the instructions from the Debian Wiki and download install.sh for your Neo: https://wiki.debian.org/DebianOnFreeRunner https://wiki.debian.org/DebianOnNeo1973 Remap the partitions for install.sh: # ln -s /dev/mmcblk0p2 /tmp/mmcblk0p1 # ln -s /dev/mmcblk0 /tmp/mmcblk0 Optionally: Verify your setup and set the system time: # SD_PART1_FS=vfat SD_DEVICE=/tmp/mmcblk0 ./install.sh testing time Partition and format your SDcard. Either do so yourself or use install.sh: # SD_PART1_FS=vfat ./install.sh partition format Run the rest of the installation. Feel free to add options as you see fit: # SINGLE_PART=true SD_DEVICE=/tmp/mmcblk0 ./install.sh mount debian apt fso configuration kernel Copy the kernel image to the vfat partition: # mount -t vfat /dev/mmcblk0p1 /media/card # cp -p /mnt/debian/boot/uImage.bin /media/card # umount /media/card Optionally: Copy your existing SSH setup: # cp -a ~root/.ssh /mnt/debian/root Unmount the Debian partition: # SINGLE_PART=true SD_DEVICE=/tmp/mmcblk0 ./install.sh unmount Clean up: # rm -f /tmp/mmcblk0 /tmp/mmcblk0p1 ./install.sh --- end Debian on Freerunner HOWTO --- -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [fso] sending pin failed
On Sun, Jan 18, 2009 at 08:33:47AM +0900, William Kenworthy wrote: Try 2008.12 - I find that FSO and SHR which use the gsm0710muxd just lock up the gsm chipset and dont work - ever :( There are still bugs being fixed in gsm0710muxd, e.g. this one: https://lists.linuxtogo.org/pipermail/smartphones-userland/2009-January/000874.html I.e. make sure you have at least version 0.9.2.2. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Yaouh! out (update for tangogps maps)
On Fri, Jan 16, 2009 at 04:12:19PM +0100, Tilman Baumann wrote: Untested code: import httplib conn = httplib.HTTPConnection(tile.openstreetmap.org) conn.request(HEAD, /file...) r1 = conn.getresponse() print r1.status, r1.reason etag = getheader(ETag) print etag And if stuff was new, make GET instead of HEAD The 'If-None-Match:' header field of the GET request appears to do the right thing. http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.26 Example using wget when we don't have the tile: $ wget --server-response http://tile.openstreetmap.org/5/6/7.png --2009-01-22 18:14:19-- http://tile.openstreetmap.org/5/6/7.png Løser tile.openstreetmap.org...128.40.58.204 Connecting to tile.openstreetmap.org|128.40.58.204|:80... forbundet. HTTP forespørgsel sendt, afventer svar... HTTP/1.1 200 OK Date: Thu, 22 Jan 2009 17:14:19 GMT Server: Apache/2.2.4 (Ubuntu) ETag: 865cd3d3cc1cf8ae3cf920234f539d91 Content-Length: 5822 Cache-Control: max-age=518361 Expires: Wed, 28 Jan 2009 17:13:40 GMT Keep-Alive: timeout=3, max=100 Connection: Keep-Alive Content-Type: image/png Længde: 5822 (5,7K) [image/png] Saving to: `7.png' 2009-01-22 18:14:19 (179 KB/s) - `7.png' saved [5822/5822] Example where we have an up-to-date tile already: $ wget --server-response \ --header 'If-None-Match: 865cd3d3cc1cf8ae3cf920234f539d91' \ http://tile.openstreetmap.org/5/6/7.png --2009-01-22 18:15:25-- http://tile.openstreetmap.org/5/6/7.png Løser tile.openstreetmap.org...128.40.58.204 Connecting to tile.openstreetmap.org|128.40.58.204|:80... forbundet. HTTP forespørgsel sendt, afventer svar... HTTP/1.1 304 Not Modified Date: Thu, 22 Jan 2009 17:15:25 GMT Server: Apache/2.2.4 (Ubuntu) Connection: Keep-Alive Keep-Alive: timeout=3, max=100 ETag: 865cd3d3cc1cf8ae3cf920234f539d91 Expires: Wed, 28 Jan 2009 18:17:50 GMT Cache-Control: max-age=522144 2009-01-22 18:15:25 FEJL 304: Not Modified. Example where we have an outdated tile: $ wget --server-response \ --header 'If-None-Match: 865cd3d3cc1cf8ae3cf920234f539d90' \ http://tile.openstreetmap.org/5/6/7.png --2009-01-22 18:23:10-- http://tile.openstreetmap.org/5/6/7.png Løser tile.openstreetmap.org...128.40.58.204 Connecting to tile.openstreetmap.org|128.40.58.204|:80... forbundet. HTTP forespørgsel sendt, afventer svar... HTTP/1.1 200 OK Date: Thu, 22 Jan 2009 17:23:10 GMT Server: Apache/2.2.4 (Ubuntu) ETag: 865cd3d3cc1cf8ae3cf920234f539d91 Content-Length: 5822 Cache-Control: max-age=522491 Expires: Wed, 28 Jan 2009 18:31:22 GMT Keep-Alive: timeout=3, max=100 Connection: Keep-Alive Content-Type: image/png Længde: 5822 (5,7K) [image/png] Saving to: `7.png' 2009-01-22 18:23:10 (129 KB/s) - `7.png' saved [5822/5822] -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Yaouh! out (update for tangogps maps)
On Tue, Jan 20, 2009 at 08:20:59AM +0900, William Kenworthy wrote: ext2/3 is not really suited to storing such a large number of small files. The main problem is the fixed number of inodes which I ran out of despite having 2Gb still free on the SD card partition :( To work around, create the file system with -b 1024 -i 1024 for the maximum number of inodes - unfortunately this cant be changed after the FS is created. This is another problem that would be solved by using ReiserFS. It is a pity that it is not compiled into the kernel as shipped by the distributions. (I don't understand why ext2/3 is used at all these days, as it has been obsolete for years. Except for one bug 8 years ago[1], ReiserFS has worked flawlessly for me on my desktop systems.) [1] http://lkml.indiana.edu/hypermail/linux/kernel/0003.1/0714.html -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [debian] zhone not working
On Sat, Jan 24, 2009 at 12:55:47PM -0800, vale wrote: Zhone does not longer work after upgrading debian, [snip] ImportError: libevas.so.0: cannot open shared object file: No such file or directory $ apt-file find libevas.so.0 libevas-dbg: /usr/lib/debug/usr/lib/libevas.so.0.9.9 libevas0: /usr/lib/libevas.so.0 libevas0: /usr/lib/libevas.so.0.9.9 So try to install the libevas0 package. It should have been installed automatically, though, through a dependency from zhone. What does $ dpkg --status zhone give? Anyway, last thing I heard was that the libevas0 package in Debian doesn't work. Someone was kind enough to provide a fixed one, but I have forgotten the URL. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Questions about the usability of GTA02
the estimate is too high. [2] mod.super-baz from http://www.exotica.org.uk/wiki/Super_Cars_II [3] See thread at https://lists.openmoko.org/pipermail/support/2009-January/004464.html And, will there be a (software) fix (soon) for the issue, that the battery gets uncharged, even if plugged in, as soon as the battery is fully charged? (i got no problem charging the phone every night, but when it's fully charged at 2am, it would have been running 7 hours on battery power by my arrival at university, so the battery would be low sooner) The hardware will resume charging when the battery voltage drops to 96%, but capacity might have dropped too much when that happens. A patch was posted to instead resume charging from software after 15 minutes: https://lists.openmoko.org/pipermail/openmoko-kernel/2008-December/007354.html https://lists.openmoko.org/pipermail/openmoko-kernel/2008-December/007355.html (I think that patch needs an on/off switch, as otherwise, the Neo will no longer resume charging when suspended or turned off.) IIRC, the parts that take power directly from the battery are GSM, the audio amplifier and the vibrator motor. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR] latest unstable does really work???
On Fri, Jan 30, 2009 at 03:18:10PM +0100, Alexandre Girard wrote: I can ssh but can't ping google.com from the OM, any idea? I've tried to set the ip of my gateway machine (192.168.0.200) in /etc/ Resolv.conf, with no success. route add default gw 192.168.0.200 -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Freesmartphone.org Milestone 5 'Calling All Stations' Release
On Tue, Feb 03, 2009 at 03:07:00PM +0100, Petr Vanek wrote: It's a good guess... does mount say that it's still ro? i have flashed u-boot and it starts OK. reflashed qi and have the problem again. interestingly enough i tried to boot fso from nor menu once and then qi was able to boot it ok afterwards. so i am not sure what exactly happens there for the first time... Perhaps this is the same initscripts bug that was introduced with OM2008.9, causing it to fail to boot if you upgraded from OM2008.8 with opkg instead of installing a new image? The bug is that /etc/init.d/udev tries to write to the rootfs before it has been remounted rw. Example patch which I use to get OM2008 going after the upgrade: --- /mnt/flash/etc/init.d/udev~ 2008-11-24 04:20:49.0 +0100 +++ /mnt/flash/etc/init.d/udev 2008-11-25 11:28:14.0 +0100 @@ -53,12 +53,16 @@ fi } +# rootfs might be read only at this point make_extra_nodes() { if [ $(echo /lib/udev/devices/*) != /lib/udev/devices/* ]; then cp -a /lib/udev/devices/* /$1/ elif [ -f /lib/udev/devices.tar.gz ]; then -tar -C /lib/udev -zxf /lib/udev/devices.tar.gz -cp -a /lib/udev/devices/* /$1/ +#tar -C /lib/udev -zxf /lib/udev/devices.tar.gz +#cp -a /lib/udev/devices/* /$1/ +tar -C /$1/ -zxf /lib/udev/devices.tar.gz +mv /$1/devices/* /$1/ +rmdir /$1/devices fi [ -e /etc/udev/links.conf ] || return 0 (You'd get a message something like 'tar: Unable to open devices'.) the trouble is that i don't know how to change the ro flag for qi (i The rootfs should always be mounted ro at first. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Questions about the usability of GTA02
On Sun, Feb 01, 2009 at 11:46:31PM +0100, Rask Ingemann Lambertsen wrote: [on battery power] The following is with the Debian distribution (with ancient 2.6.24 kernel), xscreensaver not installed and GSM off: Suspended: 20 % battery charge remaining after 3 days. If sitting idle (not suspended) with the screen blanked: 20 hours[1]. But 7 hours drain 38 % of capacity, which gives about 18.5 hours instaed. I've checked out the 2.6.28 kernel from the andy-tracking branch, and it does better. After 12 hours, there's still 37 % charge left in the battery with an estimated 7.5 hours left. Where the 2.6.24 kernel would draw 56 mA from a fully charged battery, the 2.6.28 kernel is down to 49 mA and with 37 % carge left, it is only up to 55 mA at 3.78 V. IIRC, the parts that take power directly from the battery are GSM, the audio amplifier and the vibrator motor. USB 5 V power supply too, btw. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: opkg-deb
On Wed, Feb 04, 2009 at 09:48:42AM +0100, kimaidou wrote: Hi all I think this idea is great. I am using SHR from the start, and now I installed Hackable 1 on my SDcard (I need a PIM solution). So I miss the goodies in www.opkg.org. I've installed both usbmode and linball[1] on Debian from the .ipk package. You need to pass --force-architecture and perhaps --force-depends to dpkg. It will probably work fine for many packages. [1] Hangs on exit without -n to turn off sound (which doesn't work anyway). -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [Om2008.x] UMTS based on USB adapter
On Tue, Feb 03, 2009 at 03:41:47PM +, Al Johnson wrote: On Tuesday 03 February 2009, Matthias Apitz wrote: is there any way to use something like an USB-splitter to connect the USB of the FR to the modem and at the same time to a laptop? I don't think that'll be possible. You need the FR to be a USB host to connect to the modem, but a USB client to connect to the laptop. It is switchable between the two modes, but can't be both at the same time. A workaround would be to connect the Neo and the laptop over 802.11 (WLAN) or Bluetooth. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Low earpiece volume (Was: Questions about the usability of GTA02)
On Tue, Feb 10, 2009 at 09:35:07AM +0100, Erland Lewin wrote: My main problem with the FR at the moment is that the volume in the built-in earpiece, and wired headset as well are way too low. It works in a quiet indoor environment, but on a busy street I can hardly hear that someone is talking on the phone at all, and in a room with other people talking in the background, it is difficult to hear the person at the other end. Just wondering: What happens if you run this as root during a call (on a distrubution which uses FSO)? # (sleep 5 echo -e 'AT+CLVL?;+CLVL=255\r' sleep 1) | mickeyterm You should see output something like this: - read history from /root/.mickeyterm_history --- Mickey's Term V2.9.1 @ /dev/pts/2 --- AT-Command Interpreter ready AT+CMEE=2;+CRC=1 OK AT+CLVL?;+CLVL=255 +CLVL: 175 OK CTRL-D Good Bye. - It should report the current modem output volume (175 here) and then set it to 255, which is the loudest possible on the TI Calypso. In your /usr/share/openmoko/scenarios/gsmhandset.state, what are the values of these controls? 1) Control 4 Speaker Playback Volume (LOUT2VOL and ROUT2VOL in [2]). 2) Control 6 Bypass playback volume. 3) Control 52 ROUT2 Phase (should be Inverted). 4) Control 92 DAPM Handset Spk (should be true). (Bypass refers to the tree lines at the top of the drawing in [2] while side tone refers to the line that goes right from MIC MUX in the lower left of the same drawing.) There's also the possibility of using a larger speaker: # alsactl restore -f /usr/share/openmoko/scenarios/gsmspeakerout.state Of particular interest are: 1) Control 3 Headphone Playback Volume (LOUT1VOL and ROUT1VOL in [2]). 2) Control 6 Bypass playback volume. 3) Control 87 DAPM Stereo Out (should be true [3]). 4) Control 92 DAPM Handset Spk Switch (should be false [3][4]). 5) Control 93 Amp State Switch (should be true). 6) Control 94 Amp Spk Switch (should be true). [1] https://wiki.openmoko.org/wiki/Neo_Freerunner_audio_subsystem [2] https://wiki.openmoko.org/wiki/Neo_1973_audio_subsystem#ALSA_Channels [3] Mine, as installed from the Debian distribution, has true/false swapped. Both of control 87 and control 93 must be set to true to enable sound from headphones or the large speaker. [4] Or set it to true to use both speakers and check control 4. My impression is that this is a hardware limitation, and that even with max volume from the GSM chip to the audio codec (CLVL=255), and maximum Alsa volumes, the volume won't be high enough from the earpiece.. I think the only way around this is to use a Bluetooth headset, but I don't think the Bluetooth headset profile works in the version of the bluetooth stack (bluez-3) that's on the Freerunner now. Sound from the modem can be captured, so it ought to be possible to increase volume in software with e.g. sox and the 'gain', 'loudness' or 'vol' effects. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR unstable] Openmoko headset
On Fri, Feb 13, 2009 at 12:55:07PM +0100, Tony Berth wrote: Dear List, how can I avoid to have the Openmoko headset AND the speaker working at the same time? Where I plug-in the headset, people around me can listen to the conversation from the speaker! $ amixer sset 'DAPM Handset Spk' mute or perhaps better $ /usr/sbin/alsactl restore -f /usr/share/openmoko/scenarios/gsmheadset.state But assuming you want this to be handled automatically, the file you're after is /etc/freesmartphone/oevents/rules.yaml. Is there any documentation for this file? Also, how can I increase the volume in the headset cause the default one is very low! You can hardly understand the person you are calling. Hopefully it is enough to adjust control 3 Headphone Playback Volume and control 6 Bypass Playback Volume in the gsmheadset.state file. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: AGPS question
On Sun, Feb 15, 2009 at 12:14:45PM +0100, Leonti Bielski wrote: Or are also time limitations? I've read somewhere that aGPS data is enough for 4 hours or so. The satellites move, which obsoletes the aGPS data. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Speech dispatcher output to bluetooth (navit) ?
On Thu, Jan 22, 2009 at 09:38:41PM +0100, Lothar Behrens wrote: Has no one got the output of speech-dispatcher into a bouetooth headset ? Using mplayer -ao alsa:device=bluetooth works, but not spd-say. It always comes out of the speaker. What does the speech-dispatcher documentation say about selecting a sound card for output? Also try some of the stuff from the wiki: http://wiki.openmoko.org/wiki/Neo_1973_audio_subsystem#Driver_Status_3 Specifically these: http://wiki.openmoko.org/images/3/3d/Btcpu_gta02.state.txt http://handheldshell.com/bluetooth_pcm http://opensource.wolfsonmicro.com/~gg/bluetooth-pcm/bluetooth_pcm.c Usage would be something like this: $ alsactl restore -f Btcpu_gta02.state.txt $ ./bluetooth_pcm [play some sound here] $ killall bluetooth_pcm The interesting controls of the state file are these: 20 'Sidetone Capture Volume' - 0 (increase as needed). 64 'Capture Right Mux' - 'Sidetone' to route playback mixer sound into right ADC. 66 'Capture Right Mixer' - 'Analogue Mix Right' to record right channel from right ADC. 67 'Capture Left Mixer' - 'Analogue Mix Right' to record left channel also from right ADC. 68 'Playback Mixer Voice Capture Sw' - false to not mix GSM output into the sidetone. 69 'Playback Mixer Left Capture Swi' - true to mix left output into the sidetone. 70 'Playback Mixer Right Capture Sw' - true to mix right output into the sidetone. (Maybe look at control 22 'Capture Volume' too?) -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [Query] Known Good SD cards
On Thu, Feb 19, 2009 at 11:57:27AM +0500, Shaz wrote: Hi everyone, I need some guidance on uSD cards because i tried a few locally available to my town but none are working. My understanding is that its the quality or I am making a mistake in compatibility. I need some help in understanding this issue. If there are known to be working SD cards then I would order from some other place if not available locally. I bought an 8 GB speed class 4 SanDisk card marked 0732502849DZG (in very small letters - bring a magnifying glass to the shop). It came bundled with a combined MicroSDHC and M2 card reader with USB connector. I've had the card for about two months with a Debian installation and it works fine with u-boot (as shipped in NOR and NAND flash) and kernel versions 2.6.24 and andy-tracking 2.6.28. No boot parameters required. There were two different SanDisk packages of an 8 GB card and a USB card reader I could choose from. According to the salesman the differences were only in the packaging. But anyway, the one I bought has 80-13-02494 printed at the bottom of each page of the user guide. The barcode on the back of the package reads 6 1965904025 3 with SDSDQR-8192-E12M printed above it and 80-56-04874-8192 printed below it. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [Debian]Voice recorder - ecasound -- Help
On Sat, Feb 21, 2009 at 11:40:18AM +0100, kimaidou wrote: Hi ! I totally agree with you, i will implement this for the 0.2 version For the mono thing, I would like it but I don't know yet how to do it. This is why when you open a wav recorded file with audacity, you can see the 2 left and right side, but only the right has data... This is not good. Have a look at controls 66 and 67. Depending on how your state file routes the mic input, set them both to right or left. Example: $ amixer sset 'Capture Right Mixer' 'Analogue Mix Right' $ amixer sset 'Capture Left Mixer' 'Analogue Mix Right' (Why do we have both? They are forced to the same value.) Then select the same ADC as capture source for both channels (control 51): $ amixer sset 'ADC Data Select' 'Right ADC' I have attached two state files, one to capture from the headset mic and one to capture from the handset mic. Both use automatic level control and the one to capture from the headset mic also uses the noise gate (squelsh) control. The signal level at the handset mic is unfortunately too low for that. Please also refer to the wiki as there are a _lot_ of ways to route sound to and from various places with the WM8753L: https://wiki.openmoko.org/wiki/Neo_1973_audio_subsystem It would be really great to have a mixer app where you could point out the controls on a diagram such as the one on the wiki page. So if someone knwos the parameter to put in arecord to record only one mono sound, this would help ! -c 1, but the WM8753L doesn't support mono. You can work around it by using '-D plughw:0,0' instead of '-D hw:0,0'. Alternatively: $ arecord --help | grep -F -e 'separate-channels' -I, --separate-channels one file for each channel -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [Debian]Voice recorder - ecasound -- Help
On Sun, Feb 22, 2009 at 02:33:09PM +0100, Rask Ingemann Lambertsen wrote: I have attached two state files, this time around. Btw, the state files have the speaker/headphone output and amplifier turned on (controls 87 and 93) for ease of testing. Just turn them off if you don't need them to save power. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year capturehandset.state.gz Description: GNU Zip compressed data captureheadset.state.gz Description: GNU Zip compressed data ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [Debian]Voice recorder - ecasound -- Help
On Sun, Feb 22, 2009 at 01:30:52PM +0100, kimaidou wrote: Hi guys I have a question : why not just pass the good parameters to arecord so that it recors in 8bit, Mono, 8kHz as : arecord -D hw -f S16_LE -t wav -c 1 -r 8000 myfile.wav $ arecord -D hw -f S16_LE -t wav -c 1 -r 8000 myfile.wav Recording WAVE 'myfile.wav' : Signed 16 bit Little Endian, Rate 8000 Hz, Mono arecord: set_params:923: Channels count non available $ arecord -D hw -f S8 -t wav -c 2 -r 8000 myfile.wav Recording WAVE 'myfile.wav' : Signed 8 bit, Rate 8000 Hz, Stereo arecord: set_params:918: Sample format non available $ arecord -D hw -f U8 -t wav -c 2 -r 8000 myfile.wav Recording WAVE 'myfile.wav' : Unsigned 8 bit, Rate 8000 Hz, Stereo arecord: set_params:918: Sample format non available -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: (SHR) again
On Sun, Feb 22, 2009 at 05:19:29PM +0100, Francesco de Virgilio wrote: Klaus Kurzmann ha scritto: * Francesco de Virgilio fradev...@gmail.com [090222 15:42]: Same here! SHR unstable it's fantastic, stable and really usable as a daily phone. Just one problem: I've a cheap BL-5C battery which works like a charm with all my nokia phones, but with the Neo FreeRunner is not recognized. That battery with OM 2008.12 doesn't show any info with apm, but worked ok. On SHR unstable, even with the battery fully charged and connected to USB, after 5 minutes, SHR autonomously power off the phone. I set all the values in Settings - Power Management to -1, but nothing. The only battery state indication you have with a non-BQ27000 battery is the voltage as measured by the charger. I don't know where that has gone in the 2.6.28 kernels, but in 2.6.24, it was something like /sys/devices/platform/neo1973-pcf50633/battvolt. You can try to file a feature request with the SHR and FSO people to use it when you're not using the standard GTA02 battery. Pardon, I answer myself[1]: the GPS is powered on autoagically on boot, so... is there a way to default power it off on boot? [1] http://shr-project.org/trac/wiki/Draft%3ASHRSettingsApp#lookfeel That's not what it says. It says now gps is starting automatically, which on FSO-based distributions usually means off unless in use. You should have a sysfs file to check: $ cat /sys/bus/platform/devices/neo1973-pm-gps.0/power_on 0 (If it's not there in the kernel you're running, try to find it with the command find /sys -name 'neo1973-pm-gps*' -print.) -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Voicenote - new software
On Sun, Feb 22, 2009 at 04:21:07PM +0100, Francesco de Virgilio wrote: r...@om-gta02 ~ $ sh /usr/bin/voicenote.sh zenity: error while loading shared libraries: libgnomecanvas-2.so.0: cannot open shared object file: No such file or directory Missing dependency in the zenity package. apt-file is your friend: $ apt-file find libgnomecanvas-2.so.0 libgnomecanvas2-0: /usr/lib/libgnomecanvas-2.so.0 libgnomecanvas2-0: /usr/lib/libgnomecanvas-2.so.0.2001.0 -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: GUI for SHR fast charge mode: possible?
On Sun, Feb 22, 2009 at 01:20:05PM -0500, Joel Newkirk wrote: On Sun, 22 Feb 2009 17:37:36 +0100 Francesco de Virgilio fradev...@gmail.com wrote: IOError: [Errno 2] No such file or directory: '/sys/class/i2c-adapter/i2c-0/0-0073/charger_type' /sys/class/power_supply/adapter/type (I think...) There's no direct equivalent to charger_type. $ head /sys/class/power_supply/*/type == /sys/class/power_supply/ac/type == Mains == /sys/class/power_supply/adapter/type == Mains == /sys/class/power_supply/battery/type == Battery == /sys/class/power_supply/usb/type == USB $ head /sys/class/power_supply/*/online == /sys/class/power_supply/ac/online == 1 == /sys/class/power_supply/adapter/online == 0 == /sys/class/power_supply/battery/online == 1 == /sys/class/power_supply/usb/online == 0 Note that unless you solder in such a thing yourself, the GTA01 and GTA02 don't have an adapter input. Also, in latest andy-tracking the USB vs. ac detection is broken (as is the case in the example above). For me the fast-charging mode is really important to use the Neo in my car, which adaptor for default provides only 100 mhA. Is there a way to set quickly fast-charge mode on SHR unstable? /sys/class/power_supply/ac/device/force_usb_limit_dangerous That will be /sys/class/power_supply/ac/device/usb_curlim in 2.6.29. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [fso] no sound card
On Wed, Mar 04, 2009 at 05:23:34PM +0100, GNUtoo wrote: On Wed, 2009-03-04 at 16:17 +, Hans Zimmerman wrote: Any pointers where to look for the problem/solution? on SHR i had a problem with sound too: no /dev/dsp was present That might not be a sign of trouble because /dev/dsp is the old Open Sound System (OSS) interface. These days, ALSA is the norm and lives in the /dev/snd directory. Three useful commands to try out (and example output with just the built in sound card available): $ lsmod | grep -E -e '^sound|^snd[_ ]' snd_soc_neo1973_gta02_wm8753 8808 0 snd_soc_s3c24xx_i2s 4568 1 snd_soc_neo1973_gta02_wm8753 snd_soc_s3c24xx 5192 1 snd_soc_neo1973_gta02_wm8753 snd_soc_wm8753 34228 2 snd_soc_neo1973_gta02_wm8753 snd_soc_core 46548 4 snd_soc_neo1973_gta02_wm8753,snd_soc_s3c24xx_i2s,snd_soc_s3c24xx,snd_soc_wm8753 snd_pcm_oss45736 0 snd_mixer_oss 15304 1 snd_pcm_oss snd_pcm75564 4 snd_soc_s3c24xx,snd_soc_core,snd_pcm_oss snd_timer 21212 1 snd_pcm snd_page_alloc 6824 1 snd_pcm snd50852 8 snd_soc_neo1973_gta02_wm8753,snd_soc_wm8753,snd_soc_core,snd_pcm_oss,snd_mixer_oss,snd_pcm,snd_timer soundcore 6796 1 snd (The modules snd_pcm_oss and snd_mixer_oss are for the kernel level OSS emulation.) $ ls -l /dev/snd total 0 crw-rw 1 root audio 116, 0 Jan 30 00:04 controlC0 crw-rw 1 root audio 116, 24 Jan 30 00:04 pcmC0D0c crw-rw 1 root audio 116, 16 Jan 30 00:04 pcmC0D0p crw-rw 1 root audio 116, 25 Jan 30 00:04 pcmC0D1c crw-rw 1 root audio 116, 17 Jan 30 00:04 pcmC0D1p crw-rw 1 root audio 116, 33 Jan 30 00:04 timer $ cat /proc/asound/cards 0 [neo1973gta02 ]: WM8753 - neo1973-gta02 neo1973-gta02 (WM8753) -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Your package is broken in opkg.org
On Sun, Mar 08, 2009 at 04:01:45PM +0200, Risto H. Kurppa wrote: (cc:d the authors of some broken packages I was able to extract the e-mail address from..) Hi! If you have packaged some of the apps listed below, I just want to let you know that the package in opkg.org is broken and not recognized as an installable file (to try this, download the packages from opkg.org and run file package.ipk - it should return Debian Binary but for the ones listed below, it returns data/empty or gzip) Another type of breakage I've seen is this one (data and control archives in wrong order): $ ar tv fluid_0.0+svn20070817-r2_armv4t_eabi.ipk rw-r--r-- 1000/1000 4 Aug 17 02:28 2007 debian-binary rw-r--r-- 1000/1000 45830 Aug 17 02:28 2007 data.tar.gz rw-r--r-- 1000/1000411 Aug 17 02:28 2007 control.tar.gz While this seems to work fine with the Debian tools, it would be preferable to fix the tool which builds the packages that way. Packages broken this way can be fixed like this: $ cd empty_directory $ ar xov broken_package.ipk $ rm broken_package.ipk $ ar qcv broken_package.ipk debian-binary control.tar.gz data.tar.gz $ rm debian-binary control.tar.gz data.tar.gz GZIP yphonekitd_0.4.3_any.ipk zomg_0.0.6-r2_armv4t.ipk [snip] They might be gzipped tar archives. An example case I came across: $ cd empty_directory $ tar -zxf pingus_0.7.2-r0_armv4t_DolfFix.ipk $ rm pingus_0.7.2-r0_armv4t_DolfFix.ipk $ ar qcv pingus_0.7.2-r0_armv4t_DolfFix.ipk debian-binary control.tar.gz data.tar.gz $ rm debian-binary control.tar.gz data.tar.gz -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Troubles with opkg.org repo
On Sat, Mar 07, 2009 at 09:09:15AM +0100, Tobias Kündig wrote: This is a little summary of what I do: http://www.opkg.org/repo.html Debian has this tool called dpkg-scanpackages in the dpkg-dev package. It might simplify your create_repo script. Example usage: $ cat updateindex #!/bin/sh ( dpkg-scanpackages . /dev/null \ ; dpkg-scanpackages -t opk . /dev/null \ ; dpkg-scanpackages -t ipk . /dev/null ) \ | sed -e 's_^Filename: ./_Filename: _' \ | gzip --best Packages.gz -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Open IPKG packages from command line
On Tue, Feb 17, 2009 at 10:10:14AM +0100, Kasper Johansen wrote: Hi list. Is it possible to extract a IPK-package (some_package.ipk)? I do not want to install anything - just inspect the package on my Ubuntu-machine. man dpkg-deb I have been able to extract some packages by using tar -zxvf after renaming the extension to tar.gz - though I cant with all packages (and I dont know why). Because they are not supposed to be gzipped tar archives. The ones you can extract with tar -zxvf are broken. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Bass fix (Was: Pulster fixe(s) and rework)
On Thu, Mar 05, 2009 at 12:56:06PM +0100, Helge Hafting wrote: Another question - is a single big capacitor enough, if it is put into the ground line instead of having one cap for each of the stereo channels? Or will that wreck stereo sound? One could then use a even bigger cap. I think it won't solve the problem of DC through the speakers. Think of the case where the left output is at 0 V and the right output is at 3 V. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Flashing GSM firmware
On Thu, Mar 05, 2009 at 10:25:40AM +0300, Paul Fertser wrote: All this steps except starting fluid.exe are not strictly necessary and i'd recommend to start fluid like that: FLUID_PORT=/dev/ttySAC0 fluid.exe -oo -od13,13 -f path_to_moko11_bin Notice i use ROM bootloader and moko11 firmware. I flashed the moko11 GSM firmware today and found the same thing: -oo without FLUID_FLOWCONTROL works. I couldn't not get any of the other combinations to work. I first backed up the old firmware, then flashed the new one: # time FLUID_PORT=/dev/ttySAC0 fluid.exe -oo -r 0x..0x0040 -f ~rask/download/firmware/calypso-firmware-backup.m0 (real 6m28.286s user0m11.200s sys 0m4.315s) # time FLUID_PORT=/dev/ttySAC0 fluid.exe -oo -f ~rask/download/firmware/calypso-moko11.m0 (real 2m56.764s user0m16.070s sys 0m0.990s) I did need two power off/on cycles between the two fluid invocations, though. I did not use s3c24xx-gpio b7=0 as according to the schematics, pin GPIOB7 is just the MODEM_ON signal. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
[debian] Installing *.ipk/*.opk part 1: Dependencies
When installing opk/ipk packages from e.g. http://www.opkg.org/, there will be problems with unmet dependencies because Debian uses other package names for some of the packages providing the dependencies. I've created a few small packages which create a mapping of the package names so tricks like --force-depends/--ignore-depends should not be necessary. See: http://nospamnospam.homepage.dk/software/download/oedepmappings/ It works as a repository too if you run this as root: # echo 'deb http://nospamnospam.homepage.dk/software/download/oedepmappings/ /' /etc/apt/sources.list.d/ril-oedepmappings.list -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [debian] Installing *.ipk/*.opk part 1: Dependencies
On Wed, Mar 11, 2009 at 08:00:51PM +0100, Davide Scaini wrote: Nice it would be grat in future if we (debian users) could just add opkg repos as deb ones... maybe with different priority... it may be difficult and risky, but indeed useful! d Once the Packages.gz on opkg.org is fixed, this ought to do the trick: # echo 'deb http://www.opkg.org/packages/ /' /etc/apt/sources.list.d/opkg.org.list Currently, the procedure is a bit more involved: 1. Update the package list. # apt-get update 2. Delete empty dependency lists. 3. Delete junk between package sections. # sed -i -n -e '/^Depends: *$/ n' -e '/^Package: /,/^$/ p' /var/lib/apt/lists/www.opkg.org_packages_Packages 4. Fix missing commas in dependency lists. # sed -i -e '/^Depends: / s/\([[:alnum:]]\) \([[:alnum:]]\)/\1, \2/g' /var/lib/apt/lists/www.opkg.org_packages_Packages 5. Change the architecture to armel. # sed -i -e 's/^Architecture: arm.*/Architecture: armel/' /var/lib/apt/lists/www.opkg.org_packages_Packages However, apt-get seems to dislike the www.opkg.org server: # apt-get -o 'DPkg::Options={--force-architecture;}' install gwaterpas Reading package lists... Done Building dependency tree Reading state information... Done The following extra packages will be installed: libfltk The following NEW packages will be installed: gwaterpas libfltk 0 upgraded, 2 newly installed, 0 to remove and 210 not upgraded. Need to get 570B of archives. After this operation, 0B of additional disk space will be used. Do you want to continue [Y/n]? WARNING: The following packages cannot be authenticated! libfltk gwaterpas Install these packages without verification [y/N]? y Get:1 http://nospamnospam.homepage.dk libfltk 1.1.9-6 [570B] Get:2 http://www.opkg.org gwaterpas 0.2 [8620B] Fetched 9190B in 0s (15.4kB/s) Failed to fetch http://www.opkg.org/packages/gwaterpas_0.2_armv4t.ipk Size mismatch E: Unable to fetch some archives, maybe run apt-get update or try with --fix-missing? -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Yet Another Screen Of Death
On Thu, Mar 12, 2009 at 11:03:57AM -0300, Tomas Riveros Schober wrote: try to use u-boot instead, and remember to install the kernel modules after flashing I'd instal the modules _before_ flashing the kernel to eliminate the window where you don't have any modules for the newly flashed kernel. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [Debian] read GPS logger i-blue 747 on freerunner over USB
On Wed, Mar 11, 2009 at 05:46:19PM +0100, Carsten Gerlach wrote: The kernel has the right module for this chipset, cp2101, wich I can load with modprobe cp2101. It should happen automatically. I know, that i have to set the Freerunner to USB host mode. I used the openmoko-panel to switch to USB host mode. That hasn't worked for a few months now. To switch to USB host mode: # echo /sys/devices/platform/s3c2410-ohci/usb_mode host # echo /sys/class/i2c-adapter/i2c-0/0-0073/neo1973-pm-host.0/hostmode 1 To switch back to USB device mode: # echo /sys/class/i2c-adapter/i2c-0/0-0073/neo1973-pm-host.0/hostmode 0 # echo /sys/devices/platform/s3c2410-ohci/usb_mode device -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: date and GPS related questions
On Thu, Mar 12, 2009 at 10:19:05PM +0100, Fernando Martins wrote: Hi, 1) I got the date reset to 1970 a few times. I was not paying attention but I guess this happens when battery is removed, i.e., there's no battery specific for the clock? There was a kernel bug that would cause failure to read the hardware clock on the following dates: March 1st, 2nd and 3rd. May, July, October, December 1st. Additionally, if the device was off or suspended across the beginning of a month, you might have seen it lose or gain a day. Behaviour during January was undefined. All of this because the hardware clock was one month behind: https://lists.openmoko.org/pipermail/openmoko-kernel/2009-March/009147.html I don't which distributions ship a kernel with this bug fixed. I'm also still looking for someone with a GTA01 to confirm that the bug also exists with the pcf50606-rtc driver as well. 2) Several posts mentioned that GPS could only get the fix if date was correct, the requirements being within 1 sec precision. That's only when feeding the GPS with data to speed up getting a fix. Unassisted GPS works fine with incorrect date and time. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: GPS emergency call standards
On Tue, Mar 17, 2009 at 02:58:10PM +0100, Tilman Baumann wrote: Harald Welte wrote: On Tue, Feb 24, 2009 at 12:06:20PM +0100, Tilman Baumann wrote: PS: According to Wikipedia, 112 works on all GSM networks no matter if the number is a emergency number in tie state. that depends on what the network operator does. Yep, but there seems to be some international agreement on the significance of 112. I don't have any quote yet, but as far as I understood it is even required to by the GSM standards. But that might be wrong. A D112 AT command is mentioned in the 3GPP TS 07.07 specification which you can get from here: http://www.3gpp.org/ftp/Specs/html-info/0707.htm Quoting section 8.3Enter PIN +CPIN: NOTE: Commands which interact with ME that are accepted when ME is pending SIM PIN, SIM PUK, or PH‑SIM are: +CGMI, +CGMM, +CGMR, +CGSN, D112; (emergency call), +CPAS, +CFUN, +CPIN, +CDIS (read and test command only), and +CIND (read and test command only). I don't know where the D112 command is documented. I also haven't looked through the commands to see if there is a loophole such that you can dial 112 without making an emergency call. Btw, a few days ago danish media had a news story about unintended emergency calls. It appears that the answering machine offered by many telcos is partly to blame. What happens is that you press and hold 1 to speed dial your answering machine, press 1 to listen to a message and press 2 to delete a message. Doing so leaves the number 112 on the display... -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: OT: iPhone howto
On Tue, Feb 24, 2009 at 07:15:22PM +0100, Leonti Bielski wrote: Call me boring but I don't find funny mindless destruction. Maybe this girl is saying something that makes it funny? I don't speak German :( She says: This is how you turn it on. This is your music. This is your email. This is the Internet. And this is a twelve inch wrench. After which destruction follows. And I wouldn't call it mindless destruction. The purpose is to show off a fit and healthy member of the female population. :-) If you don't like that, there are alternatives: http://www.youtube.com/v/DLxq90xmYUs (Still destroying an iPhone, of course.) -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Voicenote - 3rd version - Can play wav files
On Fri, Mar 06, 2009 at 04:16:28PM +0100, kimaidou wrote: Depends: zenity alsa-utils-aplay A comma is missing. It should be like this: Depends: zenity, alsa-utils-aplay -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR-testing] Problems with battery
On Wed, Mar 25, 2009 at 05:26:40PM +0100, Francesco de Virgilio wrote: Hi list, a small problem: I use my Nokia 6630 compatible battery (BL-5C) in the Neo with SHR-testing, but Neo doesn't recognize the battery chip, so it says that it is empty even it is full. Reporting that the battery is empty is an FSO bug. The 2.6.24 kernels do report battery voltage for all batteries, including alien ones, but that has disappeared in the 2.6.28 series. It's on my TODO list to get it back in. I've edited /etc/freesmartphone/opreferences/rules.yaml and commented the section that powers off the phone with low battery level. Definitely a silly bug - for instance, how are you supposed to be able to charge a depleted battery in the Neo, then? But, it still remains a problem: the battery doesn't charge, why? How do you know it doesn't charge the battery? Which kernel are you using? -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [All] Boot with power button freezes the Neo
On Thu, Mar 26, 2009 at 04:55:56PM +0100, Francesco de Virgilio wrote: Thank for the quick hint but... is there a smart way to do this? I've read the wiki entries for uboot but it seems quite complicated and I don't know from which point start :| Boot to the NOR boot menu, connect to the USB console from another computer with a terminal probram and use the printenv command. Save the output somewhere. Repeat for the NAND boot menu. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: test result of battery current against display brightness and GPS power mode
On Fri, Mar 27, 2009 at 05:21:59AM +0800, Qingyou Meng wrote: To save power, we set display brightness to 0% by locking screen, but OS still consumes 95 mA, leaving at most ~10 hours battery life! You must have done something wrong. I frequently test power comsumption and I get 54 mA on a fully charged battery, dropping slowly as the battery discharges[1] and nearly 20 hours battery life. FWIW, I test with this command on a Debian installation: sleep 120 cat /sys/class/power_supply/battery/{status,current_now,voltage_now,capacity} [1] It's supposed to be the other way around - current increasing as the battery discharges - but there's a current leak somewhere. It was down to 46 mA not too long ago with a kernel from the andy-tracking branch. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: question about kernel image build
On Wed, Mar 25, 2009 at 09:57:29AM +0800, mx li wrote: i see the message: can't open kernel image You chose the 'moredrivers' config. That kernel image is huge. U-Boot doesn't yet automatically detect the kernel image size when loading from flash ROM, so you'll need to increase the load size from the default 0x20. Besides, you really should also install the modules corresponding to the kernel you're using. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: test result of battery current against display brightness and GPS power mode
On Fri, Mar 27, 2009 at 05:21:59AM +0800, Qingyou Meng wrote: To set brightness: write (brightness_percent / 100 * 255) to file /sys/class/backlight/gta02-bl/brightness I think that's why you get so high currents. This # echo /sys/class/backlight/gta02-bl/brightness 0 doesn't do what you hope it does. You should try # echo /sys/class/i2c-adapter/i2c-0/0-0073/pcf50633-regltr.9/glamo3362.0/glamo-spi-gpio.0/spi2.0/state sleep also. A shorthand for that file is /sys/bus/spi/devices/spi2.0/state. When the display isn't blanked, it reads 'normal'. But, IMHO, consider using a higher-level interface (such as freesmartphone.org) to turn off the display instead of trying to find all the places to mess with under /sys yourself. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: test result of battery current against display brightness and GPS power mode
On Sat, Mar 28, 2009 at 07:32:24PM +0100, Richard Kralovic wrote: It may be the case that the fixes for the current leak were introduced in devel branch (linux-openmoko-devel). On kernel 2.6.29-rc3, my tests show a drop from cca 80mA to cca 47mA. Do you know which git revision that kernel is? Alternatively, where did you get that kernel from? -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Solar panel on Freerunner back cover
On Thu, Apr 02, 2009 at 10:17:23AM +0200, Radek Polak wrote: I'd like to have 7 pieces. Those could do 7*0.57=3.99V and 180uA current. Hopefully that's 180 mA. Would it be possible to attach output from this solar panel straight to battery connector? Do you think that this can work? You'd want a diode in there to prevent the charger from charging your solar panels. Alternatively, if you're good with a soldering iron, you could try to use the unused adapter input of the PCF50633. Please see figure 4 and/or figure 38 in the PCF50633 user manual. The adapter input is not used on the Freerunner. You might want to ask questions like this on the hardware list. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Internal USB devices (Was: Using 3G USB dongle)
On Fri, Apr 03, 2009 at 12:02:06PM +0300, Timo Juhani Lindfors wrote: I'd love to be able to fit USB devices inside the case but I have not looked at how to reroute USB easily inside the device. There has been some discussion about this already: https://lists.openmoko.org/pipermail/hardware/2009-January/000951.html https://lists.openmoko.org/pipermail/hardware/2009-January/000953.html https://lists.openmoko.org/pipermail/hardware/2009-January/000955.html http://www.instructables.com/id/Four-port-USB-hub/ Also, the LDO3 voltage regulator on the PCF50633 is unused and capable of supplying 50 mA. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Graphics Performance
On Thu, Apr 02, 2009 at 09:19:45PM -0400, Iain B. FIndleton wrote: Well, that clears things up a bit. So, there is no way to get rid of the draping one sees when the display is refreshed? My stuff uses double buffering, but your comments appear to indicate that that is a waste of time. The Glamo doesn't have a vblank interrupt. Try to search the bug tracker, though, because there was a mention of an alternative means of getting double buffering to work. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Graphics Performance
On Fri, Apr 03, 2009 at 09:14:16PM +0100, Ian Stirling wrote: Iain B. FIndleton wrote: I wonder what the issue is in moving the frame buffer data? At the speeds available on the FR, moving 640 x 480 x 2 = 614,400 bytes from memory to a video buffer at 30 Hz needs about 18 MB/sec. What is the bandwidth for memory moves? About 6-8 or so - with 100% CPU utilisation There are four DMA channels in the s3c422b, so you shouldn't use the CPU for straight memcpy() of such large blocks. The kernel doesn't yet support mem-mem DMA transfers, but I'm working on that to use it in the glamo SD card driver. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Sean's speech at ESC about making a 3G device
On Fri, Apr 03, 2009 at 11:27:02PM -0700, Steve Mosher wrote: The GTA03 as a design had a recamping bug. it kept changing. and changes of course, are perceived as easy. after all, hardware is just like software.. I won't go through all the iterations of GTA03. But it's processor changed, the RFs changed, in the end the design had no Wifi and no GPS. Raise your hand if think that a GTA03 with no wifi and no GPS is good thing to sell to the community? Raise your hand if you think it was a mistake to kill this? Yes, I think GPS and Wifi/WLAN can be considered a requirement these days for anything else than a basic cell phone. I kind of like the GTA02.5 idea (see Tim Dobson's message) as a follow up to the GTA02: Upgrade S3C2442B to S3C6410. Take out the Glamo, including - using the S3C LCM interface for the display. - using the S3C SDIO interface for the SD card connector. This means finding somewhere else to hook up the Wifi chip (and in turn finding another Wifi chip I guess). Upgrade Calypso GSM to Cinterion MC75i. Time, space and cost permitting: Camera on S3C6410 camera interface. Stylus with middle and right mouse buttons. Increased maximum display brightness. Increased maximum speaker volume. Generally keep everything else as in GTA02v8. Most of this was already being worked on for the GTA03 - certainly the kernel patches suggest so. So I'm curious what changed so much in the intended usages of the GTA03 (compared to GTA02) that it had to be cance^W^H postponed. The Neo1973 was released in July 2007 and the Freerunner in July 2008, so with about the same number of changes for the new device as between the two previous ones, perhaps a GTA2.5 could have been out in about July 2009. It must have been more than just the volume up/down buttons that did it. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Bass fix (Was: buzz fix)
On Thu, Jan 29, 2009 at 01:51:04PM +0100, Helge Hafting wrote: My hope is that the plug is so big, that re-soldering it won't require special skills. And the small caps can apperently be shorted without soldering, as suggested on the hw list. Do you have a pointer to that? -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Graphics Performance
On Fri, Apr 03, 2009 at 11:36:39AM +0100, Thomas White wrote: We do have some acceleration already - both XGlamo (the Kdrive X server) and xf86-video-glamo (the Glamo driver for Xorg) make use of Glamo's 2D engine to accelerate tasks such as flood-filling large areas and moving blocks of data around the screen or onto the screen from offscreen. Btw, what is the state of xf86-video-glamo? Does XVIDEO work? Does screen rotation work (without getting huge fonts, messed up DPI values)? There are many limitations of the chip, but I don't see them as a reason to give up on this kind of thing. For example, it's often mentioned that the 3D engine won't render to a buffer larger than 511x511 pixels. That would seem to rule out such graphical fanciness at the native resolution of 480x640, but how about we just cover a 480x511 region of the screen with accelerated graphics and make the remaining area into some kind of tool or status bar? No, we absolutely need to have a 3D accelerated keyboard. :-) -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Using 3G USB dongle [Re: Sean's speech at ESC about making a 3G device]
On Fri, Apr 03, 2009 at 09:59:49AM +0200, Sander wrote: Timo Juhani Lindfors wrote (ao): I have been using E169 dongle quite a lot with freerunner. What do you use to connect the dongle and the FR? I'm looking for a converter between the small FR port and the large dongle port. https://wiki.openmoko.org/wiki/Specialized_USB_cables -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: test result of battery current against display brightness and GPS power mode
On Mon, Mar 30, 2009 at 02:41:54PM +0200, Richard Kralovic wrote: Rask Ingemann Lambertsen wrote: On Sat, Mar 28, 2009 at 07:32:24PM +0100, Richard Kralovic wrote: It may be the case that the fixes for the current leak were introduced in devel branch (linux-openmoko-devel). On kernel 2.6.29-rc3, my tests show a drop from cca 80mA to cca 47mA. Do you know which git revision that kernel is? Alternatively, where did I am using gitr1e257a0e99817a338e3706708ebb5036518e46d8, I compiled it myself. OK, I think I've found something. If I boot up Debian with LXDE (with xscreensaver and pcmanfm uninstalled) and just run the test there, I once in a while get 53 mA, but usually it's 55 mA. Now I stopped some userspace stuff: # /etc/init.d/apmd stop # /etc/init.d/nodm stop $ sleep 120 cat /sys/class/power_supply/battery/{status,current_now,voltage_now,capacity} Discharging 47250 4151000 100 I.e. 8 mA lost because of something in userspace. :-( I'm working on CPU frequency scaling support. Slowing the CPU to 100 MHz and 1.1 V core power supply gives: $ sleep 120 cat /sys/class/power_supply/battery/{status,current_now,voltage_now,capacity} Discharging 39750 4148000 100 :-) -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR-Unstable] opkg upgrade fail
On Wed, Apr 08, 2009 at 05:04:17PM -0500, The Digital Pioneer wrote: opkg: fork failed Segmentation fault How do I get it to upgrade? :\ Perhaps it runs out of memory. Do you have a swap partition or swap file? -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Slashdotted
On Tue, Apr 07, 2009 at 11:58:33AM +0200, arne anka wrote: Die Entwicklung der Mobiltelefone wird eingestellt Development of mobile phones stopped big topic (many people don't read more) Openmoko gibt freie Handy-Entwicklung auf Openmoko stopps development of free mobile phones In my opinion it suggerates the wrong message. so? phone development means gta03 -- and that indeed stopped. I think you've just given a very good example of why the GTA0x naming had to be abandonned. If even people who read Openmoko mailing lists daily get it wrong, surely a lot of other people will get it wrong too. No, phone development doesn't mean GTA03. Openmoko can release a device which is not a successor to the GTA02. gta02 does not see any development, just support and fixes. I'm sure not everybody agrees with you that bug fixes or improvements to existing features is not development. Perhaps you could clarify what you think counts as development to an already released product (GTA02). -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR-Unstable] opkg upgrade fail
On Thu, Apr 09, 2009 at 11:41:36AM -0500, The Digital Pioneer wrote: Ahh, swapfile did the trick it seems. How can I make it automount my swapfile on boot? # grep -F -e swap /etc/fstab /dev/mmcblk0p4 swapswapdefaults0 0 ^^ name of swap file/partition -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Camera module as back case (was Re: Slashdotted)
On Mon, Apr 06, 2009 at 11:41:57AM +0200, Lothar Behrens wrote: About the camera issue: I have read something about a small computer base case that has a jack or bus system ontop of it's case. That way it is extendable with cameras or what ever. I don't know where it was but it was cool. Why not think about a back case that includes a cam and using contacts in the inner housing to connect the cam to, say an I2C bus or even an USB hub (USB2)? But how will you get the image pixels across to the Neo? There has been some discussion about this on the hardware list, starting with the message Weird hardware mod ??? back in January. To sum it up: 1) USB. There are many USB webcams available, but the difficulty is finding one that works well on a USB 1.1 bus, because USB 2.0 is fast enough that the cameras can simply dump the image across the cable with little or no buffering in the camera. So most cameras won't have enough buffer space to hold a good still image, and for the same reason also don't have the buffer space needed for any of the good video compression codecs. 2) A generic image sensor interface with I2C for control and 8-bit parallel port for data transfer. Example camera chip: https://www.sparkfun.com/commerce/product_info.php?products_id=8668 The options: 2a) The Glamo has an unused camera interface, even with all the important pins connected to a resistor network, but the documentation is under NDA. 2b) Roll your own somehow. The SC32442B has lots of unused GPIO - e.g. most of the LCD interface - but the ones that appear to be connected to somewhere accessible are scattered all over the place in register space. rant index=1 It would have been a lot easier if the SC32442B camera interface pins hadn't been frittered away for GPIO assignments that could just as well have been served by the pins of the unused LCD interface instead. Even if you don't want to use them for a camera, the 13 of them are nicely lumped together in GPIO port J. They would have been fine for many projects. /rant rant index=2 Of the only two external DMA request pins, one is a dead end and the other one is connected to the GPS as a simple output. Again lots of other less valuable GPIO pins are available for such purposes. /rant -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Slashdotted
On Sat, Apr 11, 2009 at 07:34:04PM +0200, arne anka wrote: phone development means gta03 -- and that indeed stopped. it wrong, surely a lot of other people will get it wrong too. No, phone development doesn't mean GTA03. Openmoko can release a device which is not a successor to the GTA02. i never said, it has to be a successor to gta02. the only other phone in the pipeline had the name gta03 and its development has stopped. period. The misrepresentation - which you seem to think is not a misrepresentation - in the news coverage is that they're saying development of open phones at OM has been cancelled, while all that's been decided is to cancel the GTA03, not development of more open phones. gta02 does not see any development, just support and fixes. Perhaps you could clarify what you think counts as development to an already released product (GTA02). one or more of the following - newer/better hardware - improved functionality (f ex wrt standby time, power consumption) - different device Dieter Spaar is working on the deep sleep mode of the Calypso. If he suceeds, it will improve standby time and power consumption. https://lists.openmoko.org/pipermail/hardware/2009-March/001086.html https://lists.openmoko.org/pipermail/hardware/2009-April/001106.html i am sure, we can spend days on end with attempts to find a definition of development that everybody agrees to, but what purpose would it serve? No, I just want to make sure I understand you correctly before I claim you are wrong. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: latest and greatest, progress mail 6
Latest and greatest of what? -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [debian] problems with debian and u-boot
On Sun, Apr 12, 2009 at 12:33:52PM +1200, Robin Paulson wrote: now, when i try to start it, either by pressing the power button alone, or going through the nand menu, i get an error about 'wrong format' of a partition. i'm assuming this is to do with the fat/ext issue for the kernel, but as i followed the instructions to use an ext partition, i'm not sure what's wrong. it feels like i'm missing something silly - i got this working last year Make sure the kernel is on a partition completely below 4 GB or use a U-Boot built with this patch: https://lists.openmoko.org/pipermail/openmoko-kernel/2009-March/009573.html FWIW, even with the above patch, I'm having no success getting U-Boot to read a kernel from a 1 GB partition starting at 6 GB. I'm trying to debug it, though. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Slashdotted
On Sun, Apr 05, 2009 at 05:14:03PM +0100, Joerg Reisenweber wrote: for your Q about project files instead of pdf: OM is making money by selling hw, so there's not much sense in publishing data that doesn't help EE guys in community to understand the hw but instead is only needed for production purposes. In the end you can't do anything on a single-device basis with layout or schem proj data you couldn't do without it. Or are you the guy who's etching 8-layer at home and soldering uBGA by hand? ;-) You can't patch a ready-done 8layer PCB, no matter what your document files are (sw POV on hw!). Actually, it would be useful to know which signals are available from a trace on the surface. With only the schematics and component placement available, you're pretty much out of luck unless the signal happens to be available at a test point, resistor, capacitor or some other component with exposed terminals. The Big-C buzz fix is an example of a mod which makes use of a surface trace. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [shr] Headset detect?
On Tue, Apr 14, 2009 at 02:20:26PM +0200, Helge Hafting wrote: No phone shop seems to have such a connector. Of course, these people doesn't actually seem to know that 2.5mm and 3.5mm plugs are standardized sizes, they keep asking what brand of phone it is. :-( Yes, the sizes are standardized, but the pinout is not, thus the question which brand of phone connects to the 2.5 mm end of the converter. See also the Wiki: https://wiki.openmoko.org/wiki/Analog_wired_Headset -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Bying a Freerunner with the buzz-fix on it
On Sat, Apr 04, 2009 at 02:33:14PM -0700, Steve Mosher wrote: Cédric Berger wrote: And if it is the buzz fix (by adding the cap) that causes potential problems... what about the SOP currently applied to freerunners ? They may be impacted by incorrect ALSA states ? Yes, that was a concern of mine as well, so this too had to be checked and verified. The complication being that a reworked A6 can only be distinguished AFAIK by visual inspection and not by checking a bit in the hardware. Perhaps you could turn on MICBIAS and measure for how long the current is higher than normal as the capacitor charges. See MICBIAS CURRENT DETECT in the WM8753L manual. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Om2009 testing images
On Mon, Apr 13, 2009 at 01:34:51PM -0600, Angus Ainslie wrote: As part of the Om2009 release plan testing images were going to start at the beginning of April. For those brave souls that would like to help us test Paroli, FSO milestone 5.5 and the Om2009.1 kernel there is a Om2009 information page that contains links to the images and flashing instructions [1]. Are there any chances of building a more recent U-Boot? The one you are shipping is 8 months old (2008-08-10). -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR-Unstable] Forcing fast-charge
On Sat, Apr 18, 2009 at 05:11:36PM -0700, Mike Montour wrote: chg_curlim controls the battery charger, while usb_curlim controls the total current that can be drawn from the USB port (charging + the current used by the Freerunner). Note that whenever you set usb_curlim (directly or indirectly by plugging in power), chg_curlim is set to the new value of usb_curlim. usb_curlim can be 0, 100, 500 or 1000 while chg_curlim is between 0 and 996 (in steps of about 4). -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: backing up rootfs fails - after 40m FreeRunner continues booting
On Mon, Apr 20, 2009 at 12:07:55AM +0200, Bram Mertens wrote: Hi, I'm about ready to flash a new distro onto the FreRuuner main memory but since it'll be the first time I wanted to try backing up the existing distro first. So I followed the steps on http://wiki.openmoko.org/wiki/Backup . Which version of U-Boot are you using? DFU Upload was broken at least until 2008-10-17. 2008-10-07 Harald Welteu-boot: Fix DFU upload in u-boot http://git.openmoko.org/?p=u-boot.git;a=commitdiff;h=414367bf5b65942947dd5d569c27d2a8e8e5e562 -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR-Unstable] Forcing fast-charge
On Tue, Apr 21, 2009 at 02:19:04AM +0100, Joerg Reisenweber wrote: Am So 19. April 2009 schrieb Rask Ingemann Lambertsen: usb_curlim can be 0, 100, 500 or 1000 while chg_curlim is between 0 and 996 (in steps of about 4). checked 8.12.6.9, Tbl 95, Tbl94, and it's mentioning : 255/255 × Ich(ref) for max current. Please explain to me. Probably I'm missing some details The range of the sysfs entry is 0 to 996. I too think the scaling in the pcf50633 driver is slightly off (by about 0.4 %). -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [debian] problems with debian and u-boot
On Tue, Apr 14, 2009 at 11:05:55AM +1200, Robin Paulson wrote: 2009/4/14 Rask Ingemann Lambertsen r...@sygehus.dk: Make sure the kernel is on a partition completely below 4 GB or use a U-Boot built with this patch: https://lists.openmoko.org/pipermail/openmoko-kernel/2009-March/009573.html agh, that's why. it was on an 8gb partition. i must have missed that in the debian documentation/om wiki I'm not sure it was a known limitation. A newer U-Boot with the 4 GB patch is now available in the unstable directory: http://downloads.openmoko.org/distro/unstable/NeoFreerunner/u-boot-gta02v5-1.3.1+gitrb20cc520a22715fe7ff069b36b7dfdbb925a8e5a-r1.bin FWIW, even with the above patch, I'm having no success getting U-Boot to read a kernel from a 1 GB partition starting at 6 GB. I'm trying to debug it, though. using qi now, but not tried to boot debian yet. cheers rask The second problem was that U-Boot can't read ext2 filesystems revision 1 where inodes might not be 128 bytes. In my case they are 256 bytes, so the inode position is miscalculated. Patch posted: https://lists.openmoko.org/pipermail/openmoko-kernel/2009-April/010051.html -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: USB Kernel problem with latest SHR unstable
On Sat, Apr 25, 2009 at 08:37:50PM +1200, Robin Paulson wrote: i'm not entirely sure what it was, but i think it was related to some fiddling i did with depmod and modprobe on the phone. i'm guessing some usb modules weren't loaded by default The SHR packager forgot to run depmod before tarring up the rootfs. Alternatively, the initscripts should run depmod on startup. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR-Unstable] Forcing fast-charge
On Sat, Apr 25, 2009 at 01:16:17PM +0100, Joerg Reisenweber wrote: Am Di 21. April 2009 schrieb Paul Fertser: drivers/power/pcf50633-charger.c: /* * We limit the charging current to be the USB current limit. * The reason is that on pcf50633, when it enters PMU Standby mode, * which it does when the device goes off, the USB current limit * reverts to the variant default. In at least one common case, that * default is 500mA. By setting the charging current to be the same * as the USB limit we set here before PMU standby, we enforce it only * using the correct amount of current even when the USB current limit * gets reset to the wrong thing */ Whoever wrote this amazingly puzzling comment, I think he got something severely wrong with operating principles of PMU PCF50633. Datasheet of PMU clearly states there's no situation whatever that could result in batcharge current overloading the USB_CURLIM, The comment doesn't claim there is. as *allways* there will be priority on serving system by providing up to 100% of usb current to power it. Serving the system takes 0 mA in this particular case, because the device is off. Bat charge will get whatever might remain after that, *up_to* the charging limit programmed into PMU. Exactly. As we may charge our battery with 1C (=1200mA) it's perfectly safe to set bat chg curlim to that value, and rely on PMU managing distribution of actual USB supply current to system and charging according to the momentary needs. The problem is not that of staying under the 1200 mA permitted for the battery. The problem is staying under the maximum USB current, which may be as little as 100 mA. We just can't do anything about the USB current limit being reset to 500 mA, but we _can_ keep the charging current limited to 100 mA, which is good enough when the only consumer is the the battery charger. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Asterisk on Freerunner was: voip on Debian
On Sat, Apr 18, 2009 at 05:49:05PM +0200, Nicola Mfb wrote: I will be happy to write an AMI gui but now I'm hold having problems with the alsa channel. Using the pcm default is not compatible with the default shipped /etc/asound.conf, so I just tried to use plughw:dnsoop and plughw:dmix, the result is that there freerunner does not ring on incoming call (and you cannot hear the other peer), while audio transmitting is perfect. Using plughw:0,0 for input/output works but I have stuttered audio (from freerunner to peer). Why are you not using hw:0,0? -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Will my phone buzz?
On Sun, May 03, 2009 at 02:18:57PM +0300, Shachar Shemesh wrote: I'm sorry if I'm asking a silly question. When I do cat /proc/cpuinfo I get Revision: 24420360, so presumably I'm okay. Then again, while trying to figure out what field I was supposed to be looking at, I stumbled upon http://wiki.openmoko.org/wiki/Neo_FreeRunner_Hardware. From there, it seems that the user space gets that data from the kernel, which in turn gets it from the boot loader, which doesn't get it from anywhere at all hardware related. It _is_ read from the hardware by the kernel. Please see function gta02_get_pcb_revision() in arch/arm/mach-s3c2442/mach-gta02.c. I have the debug board, which implicitly means I have the screw driver needed to open the device (already did once). I was more looking for something that says this is the change you need to do in order to fix the buzz, and then just open the device and have a look whether that change is already in place. Easy enough. Unscrew the two screws, remove the front cover and check the area left of the mic. If there isn't a huge (relatively - 3.2 mm by 1.6 mm or more) capacitor left of or above the small resistors and capacitors, then you don't have the buzz fix. Examples: http://anyotherkey.googlepages.com/c http://vanous.penguin.cz/files/om/SOP/ Note: This does not apply to the GTA02v8 which is buzz-fixed differently. Btw, do we really need two wiki pages for the same thing? https://wiki.openmoko.org/wiki/GSM_buzz https://wiki.openmoko.org/wiki/Buzz_Fix -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR-Unstable] Forcing fast-charge
On Sat, May 02, 2009 at 03:30:10AM +0200, Joerg Reisenweber wrote: Am Mi 29. April 2009 schrieb Cédric Berger: On Tue, Apr 28, 2009 at 15:36, Joerg Reisenweber jo...@openmoko.org wrote: That's nonsense as there is no dumb charger out there not being capable of supplying 500mA. OTOH any management of maximum ingress current to USB by whatever means will inevitably fail when unplugging the charger device connected to FR during powerdown and replace it with a (nonexistent) weaker one. No. For better or for worse, the PCF50633 variant we use will power up the Frerunner and thus run the boot loader when you swap chargers. If the boot loader doesn't set the current limit correctly for the newly inserted charger, then that's where the bug is. TODO: Check U-Boot and Qi. (Long term TODO: Instead of fixing each and every bug three times (kernel, U-Boot and Qi), try to fit a bare-bones kernel and rootfs with boot menu into the 512 kB we currently use for boot loader.) Or a solar panel ? How could setting BAT_CURLIM to USB_CURLIM help? Because USB_CURLIM is reset to 500 mA when the Freerunner is turned off, while BAT_CURLIM remains at what it was last set to. Or you have 2 phone charging from the same usb port ? or I'm using the usb extension in my usb-coffemug-warmer? Or create a shortcircuit? It's not a valid usecase and I don't see how to accomplish that. Don't tell me you haven't seen one of those 4-port unpowered USB hubs which can only supply 100 mA per port. Note than doing this (I used a little GUI with 1 set button for each value), I must set usb current limit, then only I set charge current limit. And in the meanwhile, since charge limit=usb limit=1A, it starts pulling about 1A and I cannot avoid that. Sorry I don't understand this statement. The problem is when you use e.g. a 750 mAh Nokia battery and set chg_curlim to 750 mA like you're supposed to. If you then plug in the charger, usb_curlim will be set to 1000 mA and so will chg_curlim. Automatically setting chg_curlim to 1000 mA is a bug since the user said max 750 mA for the battery. The kernel needs to remember the user specified chg_curlim. TODO: Kernel patch. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR-unstable]Blinking AUX button LED
On Thu, Apr 30, 2009 at 04:26:30PM +0300, Timo Juhani Lindfors wrote: Vasco Névoa vasco.ne...@sapo.pt writes: Possible errors that lead to that situation: - bad rootfs kernel parameters in u-boot environment; - check if you need to change them; This can't surely cause missing filesystem support? It can if you no longer have the correct mtdparts option. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Compatibility between deb and opkg (Was: Ain't it funny..)
On Tue, May 05, 2009 at 01:31:47PM +0200, Pander wrote: For better compatability between deb and opkg, I am still in favour of using the debian/ubuntu categories in opkg. See this thread from December: http://lists.openmoko.org/pipermail/community/2008-December/038472.html and http://lists.openmoko.org/pipermail/community/2008-December/038546.html A significant incompatibility is that deb uses architectures 'armel' and 'any' while opkg uses 'armv4t' and 'all'. AFAIK there is no way of specifying to dpkg/apt that they are equivalent. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [debian] Installing *.ipk/*.opk part 1: Dependencies
On Wed, Mar 11, 2009 at 10:16:06PM -0400, Stefan Monnier wrote: When installing opk/ipk packages from e.g. http://www.opkg.org/, there will be problems with unmet dependencies because Debian uses other package names for some of the packages providing the dependencies. I've created a few small packages which create a mapping of the package names so tricks like --force-depends/--ignore-depends should not be necessary. See: http://nospamnospam.homepage.dk/software/download/oedepmappings/ [snip] Also, it would probably be preferable for those packages to have a name that makes it clear that they're not real. E.g. you might call them something like adapter-opkg-foo, and then have them provide the required feature. I would if APT supported it. Unfortunately, the APT support for Provides: is buggy enough that it only works for packages which are already installed, and I'm hoping to eventually be able to do something like # apt-install-package-file http://www.opkg.org/package.opk and have APT install the dependencies automatically (including having them marked as such). -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Speed of NAND flash vs. SD card (Was: Ain't it funny..)
On Tue, May 05, 2009 at 07:12:25AM -0600, Angus Ainslie wrote: You could always use the .tar.gz on an sd card ( It boots way faster that way anyway :) NAND flash ought to be a lot faster than Glamo's MMC interface. Where is the bug? FWIW, I tried a simple hdparm run on NOR flash, NAND flash and SD card: r...@debian-gta02:~# hdparm --direct -t /dev/mtdblock0 # NOR flash /dev/mtdblock0: Timing O_DIRECT disk reads:2 MB in 1.24 seconds = 1.61 MB/sec r...@debian-gta02:~# hdparm --direct -t /dev/mtdblock3 # NAND flash /dev/mtdblock3: Timing O_DIRECT disk reads:8 MB in 2.89 seconds = 2.77 MB/sec r...@debian-gta02:~# hdparm --direct -t /dev/mmcblk0# Glamo MMC /dev/mmcblk0: Timing O_DIRECT disk reads:8 MB in 3.46 seconds = 2.31 MB/sec -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: How to use Google Map on tangoGPS
On Fri, May 08, 2009 at 05:58:52AM +0800, Daniel.Li wrote: On Thu, 2009-05-07 at 08:53 -0700, jeremy jozwik wrote: i use it all the time for editing the los angeles area. works great! it can get a bit noisy when walking though Yes, if you use GoogleMap. I just walking into a river :( But I didn't analysis the data yet. But how can I walk into the river??? :) I believe it's what they call an urban canyon. When you have something tall - usually a building - which reflects the GPS signals on one side of you, you will appear to be in a different position than where you really are. Whatch your HDOP values - it's the third of the slash separated values shown by TangoGPS on the main screen. If it's as much as 4.5, the position shown can easily be 30 - 40 m off. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Mokomaze as non-root
On Wed, May 06, 2009 at 10:59:20PM -0500, The Digital Pioneer wrote: Not sure what all issues you're experiencing, but the accelerometers are probably root:root access only (they are on mine, SHR-unstable), if that's your problem. You can probably chmod them manually, but I'm sure you can mod some conf somewhere to change it. Not sure what... Comparing SHR unstable (from about 3 weeks ago) /etc/udev/rules.d/permissions.rules # input devices KERNEL==event[0-9]*, ATTRS{name}==*dvb*|*DVB*|* IR * \ MODE=0664,GROUP=video KERNEL==js[0-9]*, MODE=0664 KERNEL==lirc[0-9]*, GROUP=video with Debian /lib/udev/rules.d/91-permissions.rules # input devices KERNEL==event[0-9]*, ATTRS{name}==*dvb*|*DVB*|* IR * \ MODE=0664,GROUP=video KERNEL==js[0-9]*, MODE=0664 KERNEL==event[0-9]*, ENV{ID_CLASS}==joystick, \ MODE=0664,GROUP=audio KERNEL==lirc[0-9]*, GROUP=video there's a difference in the input device section and indeed I have some input devices with group audio and read permission for everybody: # ls -l /dev/input/ total 0 crw-rw 1 root root 13, 64 May 5 11:39 event0 crw-rw 1 root root 13, 65 May 5 11:39 event1 crw-rw-r-- 1 root audio 13, 66 May 5 11:39 event2 crw-rw-r-- 1 root audio 13, 67 May 5 11:39 event3 crw-rw 1 root root 13, 68 May 5 11:39 event4 crw-rw 1 root root 13, 63 May 5 11:39 mice -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: nandwrite problem in writing uboot env (/dev/mtd2)
On Fri, May 08, 2009 at 05:34:59PM +0200, giacomo giotti mariani wrote: [snip] Where am I wrong? You're not using uboot-envedit from the fso-utils package. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR-Unstable] Forcing fast-charge
On Thu, May 07, 2009 at 12:42:42AM +0200, Joerg Reisenweber wrote: So taking in account your comments above this means the problem case is intentionally shutting down the device during charging from solarpanel. 1) I don't see why anybody should do this To save energy for later. 2) this should be handled during shutdown, rather than all the time by syncing BAT_CURLIM with USB_CURLIM. I'll give it a try. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: nandwrite problem in writing uboot env (/dev/mtd2)
On Sat, May 09, 2009 at 02:48:33AM +0300, Timo Juhani Lindfors wrote: Rask Ingemann Lambertsen r...@sygehus.dk writes: You're not using uboot-envedit from the fso-utils package. uboot-envedit is horribly slow. fw_setenv is in debian main and works faster. But fw_setenv doesn't do what he wants it to, does it? With uboot-envedit you can get your current environment into a plain text file where you can edit it with your text editor of choice. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR] I18N
On Sat, May 09, 2009 at 02:43:00AM +0400, ivvmm wrote: Like in every normal GNU/Linux distribution create an executable sh script in /etc/profile.d/ (better to name it lang.sh) where put lines #!/bin/sh No: $ head -n 3 /etc/profile.d/lang.sh # /etc/profile.d/lang.sh - set i18n stuff sourced=0 export LANG=de_DE.UTF-8 do not forget to make it executable. No: $ ls -l /etc/profile.d/lang.sh -rw-r--r-- 1 root root 2693 11 nov 17:59 /etc/profile.d/lang.sh /etc/profile is responsible for launching that script. No, it includes the script. Otherwise, something like lang.sh wouldn't work. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: nandwrite problem in writing uboot env (/dev/mtd2)
On Fri, May 08, 2009 at 05:34:59PM +0200, giacomo giotti mariani wrote: Hello list, some day ago I decided to change some uboot environment variable (ext2 fs in ext3 fs) from my OM2008.12 booted from the uSD. Sorry, I didn't see this at first. Why not just fix the U-Boot environment? The U-Boot environtment shipped by OM is a mess. Instead of bootcmd=setenv bootargs ${bootargs_base} ... menu_1=...: setenv bootargs ${bootargs_base} rootfstype=ext2 root=... bootargs_base=rootfstype=jffs2 root=/dev/mtdblock6 console=... change it to some thing like this: bootcmd=setenv bootargs rootfstype=jffs2 root=/dev/mtdblock6 ${bootargs_base} ${mtdparts} ... menu_1=...: setenv bootargs ${bootargs_base} root=... bootargs_base=console=... Then you can use ext2/3/4, reiserfs or whatever else your kernel supports on your SD card without changing the U-Boot environment. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Finger friendly keyboards
On Sat, May 09, 2009 at 06:59:30PM +1000, Denis Johnson wrote: On Sat, May 9, 2009 at 7:21 AM, Jon Levell openm...@coralbark.net wrote: So far on my FR, I've mostly been typing non-dictionary words so the predictive keyboard has been getting in my way. I knew a few people had created finger friendly keyboards so I had a bit of a play. Have you tried Qwo ? http://www.opkg.org/package_84.html for me despite it's unconventional approach and learning curve, this is now my favourite keyboard for sms. I'm using it on Debian, including for terminal use. The only features I miss are - autorepeat. - a smaller version for stylus use so you have more of the screen available. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: test result of battery current against display brightness and GPS power mode
On Fri, Mar 27, 2009 at 05:21:59AM +0800, Qingyou Meng wrote: Because display and GPS chip may be powered on for a long while, I choose to test them here. [snip] Here is the results: 1. when GPS chip is powered off, test brightness vs. battery current: * brightness = 100%: battery current ~= 203 mA * brightness = 75%: battery current ~= 153 mA * brightness = 50%: battery current ~= 116 mA * brightness = 25%: battery current ~= 101 mA * brightness = 0%: battery current ~= 95 mA See also http://lists.openmoko.org/pipermail/openmoko-kernel/2008-October/005542.html I.e. you're not turning the display off by just setting the backlight brightness to 0. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
[Debian] Save 8 mA. Apmd considered harmful? (Was: test result of battery current against display brightness and GPS power mode)
On Fri, Mar 27, 2009 at 01:56:26AM +0100, Rask Ingemann Lambertsen wrote: You must have done something wrong. I frequently test power comsumption and I get 54 mA on a fully charged battery, dropping slowly as the battery discharges[1] and nearly 20 hours battery life. FWIW, I test with this command on a Debian installation: sleep 120 cat /sys/class/power_supply/battery/{status,current_now,voltage_now,capacity} [1] It's supposed to be the other way around - current increasing as the battery discharges - but there's a current leak somewhere. It was down to 46 mA not too long ago with a kernel from the andy-tracking branch. This mysterious current leak seems to be fixed by not running apmd! # /etc/init.d/apmd stop # update-rc.d -f apmd remove # update-rc.d apmd stop 20 0 1 2 3 4 5 6 . -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
[debian] Installing *.ipk/*.opk part 2: apt-get-file
For part 1, please see the mailing list archive: http://lists.openmoko.org/pipermail/community/2009-March/043753.html Below you will find a shell script to install a package from a .deb, .opk or .ipk file on a Debian system. It will - download the package if necessary. - repack gzipped tar archives if necessary (some *.ipk). - edit the control file for Debian compatibility (*.opk and *.ipk). - jump through hoops so apt-get can be used for installation, so dependencies can be resolved necessary. - delete all temporary files created. Dependencies: dpkg-dev wget file binutils tar gzip Usage: apt-get-file [--verbose] install file [apt-get options] I'm sure there are still small bugs around the edges, so this is only alpha quality. I'm posting the script now because it mostly does what it's supposed to do and I want to get some feedback. Example run: # ~rask/bin/apt-get-file install http://projects.openmoko.org/frs/download.php/575/ledclock_0.6_all.ipk --2009-05-10 06:49:26-- http://projects.openmoko.org/frs/download.php/575/ledclock_0.6_all.ipk Resolving projects.openmoko.org... 88.198.93.218 Connecting to projects.openmoko.org|88.198.93.218|:80... connected. HTTP request sent, awaiting response... 200 OK Length: 13214 (13K) [application/binary] Saving to: `ledclock_0.6_all.ipk' 100%[] 13,214 --.-K/s in 0.1s 2009-05-10 06:49:28 (99.5 KB/s) - `ledclock_0.6_all.ipk' saved [13214/13214] ** Packages in archive but missing from override file: ** ledclock Wrote 1 entries to output Packages file. Reading package lists... Done Building dependency tree Reading state information... Done The following extra packages will be installed: python-pygtk python-subprocess python2.6-minimal Suggested packages: binfmt-support Recommended packages: python2.6 The following NEW packages will be installed: ledclock python-pygtk python-subprocess python2.6-minimal 0 upgraded, 4 newly installed, 0 to remove and 112 not upgraded. Need to get 1331kB/1344kB of archives. After this operation, 4780kB of additional disk space will be used. Do you want to continue [Y/n]? WARNING: The following packages cannot be authenticated! python-pygtk python-subprocess ledclock Install these packages without verification [y/N]? y Get:1 http://ftp.dk.debian.org experimental/main python2.6-minimal 2.6.1-3 [1330kB] Get:2 http://nospamnospam.homepage.dk python-pygtk 2.14.1-2 [576B] Get:3 http://nospamnospam.homepage.dk python-subprocess 2.6.1-3 [582B] Fetched 1331kB in 2s (451kB/s) debconf: delaying package configuration, since apt-utils is not installed Selecting previously deselected package python-pygtk. (Reading database ... 45694 files and directories currently installed.) Unpacking python-pygtk (from .../python-pygtk_2.14.1-2_all.deb) ... Selecting previously deselected package python2.6-minimal. Unpacking python2.6-minimal (from .../python2.6-minimal_2.6.1-3_armel.deb) ... Selecting previously deselected package python-subprocess. Unpacking python-subprocess (from .../python-subprocess_2.6.1-3_all.deb) ... Selecting previously deselected package ledclock. Unpacking ledclock (from .../ledclock_0.6_all.deb) ... Processing triggers for man-db ... Setting up python-pygtk (2.14.1-2) ... Setting up python2.6-minimal (2.6.1-3) ... Could not find platform dependent libraries exec_prefix Consider setting $PYTHONHOME to prefix[:exec_prefix] Setting up python-subprocess (2.6.1-3) ... Setting up ledclock (0.6) ... # And the script itself: #!/bin/sh # apt-get-file - Install package from .deb, .opk or .ipk file. # Copyright (c) 2009 Rask Ingemann Lambertsen ccc94...@vip.cybercity.dk # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU General Public License version 3 as # published by the Free Software Foundation. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with this program. If not, see http://www.gnu.org/licenses/. exit_clean () { cd / || true [ -z ${APTSRC} ] || rm -rf -- ${APTSRC} [ -z ${APTLIST} ] || rm -rf -- ${APTLIST} [ -z ${PKG_TMPDIR} ] || rm -rf -- ${PKG_TMPDIR} exit 1 } trap exit_clean ERR set -o errexit set -o errtrace if [ --verbose = $1 ]; then shift AR_VERBOSE=v TAR_VERBOSE=v set -o xtrace else AR_VERBOSE= TAR_VERBOSE= fi if [ install != $1 ]; then echo Usage: $0 [--verbose] install file apt-get options exit 1 fi shift PKG_URL=$1; shift PKG_FILE=$(basename ${PKG_URL
Re: Booting GTA02v6 without battery (was: Re: Need battery help in DC area)
On Sun, May 10, 2009 at 04:40:38PM +0800, Daniel.Li wrote: Well, I think maybe u hw has fixed problem. Make sure your battery never discharges completely. This is an issue because the internal charging circuitry can not be turned on until the FreeRunner has booted, and booting through USB power alone does not work on early A5/6 devices. From http://wiki.openmoko.org/wiki/Neo_FreeRunner_Battery#know_issues FWIW, I have an early GTA02v6 (date code: 20080722) and it does usually boot from USB power and always boots from the charger. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Hardware mod for starting GTA02v5 without battery? (Was: Need battery help in DC area)
On Sat, May 09, 2009 at 10:36:19PM -0400, Paul Buede wrote: Also, does anyone have instruction on how to fix my Freerunner so it can boot off USB or wall charger? Is it a hardware mod or firmwae upgrade or something? It's a hardware mod. It's capacitor C1767, page four of the schematics[1] at the bottom. There's a (rather poor) picture of a reworked A5 here: http://lists.openmoko.org/pipermail/openmoko-kernel/2008-July/003963.html Better picture (of the whole A5 PCB without the rework): http://wiki.openmoko.org/wiki/Image:Gta02a5_pcba_cs.JPG Electrically, the rework is simple enough: Just connect a larger capacitor in parallel to the existing one. But it sits under the main can, so making it fit isn't easy. The picture of the reworked device isn't very good, but it looks like a capacitor has been connected to the VB_SYS side of R1768 and to the GND side of C1722. See URL:http://people.openmoko.org/werner/gta02-chg/ for an explanation of the problems GTA02v5 has starting without battery power. Btw, where is C1767 itself? I can't find it on either the assembly silkscreen[2] or the component placement drawing[3]. At 47 uF on the A6, it ought to be easy to spot. [1] http://people.openmoko.org/joerg/schematics/GTA02/Schematics_Freerunner-GTA02_A5-A7cumulative_public_RC0.pdf [2] http://people.openmoko.org/joerg/schematics/GTA02/GTA02-MB-A6(50-71481-00)_assembly%20silkscreen.pdf [3] http://people.openmoko.org/joerg/schematics/GTA02/Component-placement_Freerunner-GTA02-MB-A6.pdf -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [debian] Installing *.ipk/*.opk part 2: apt-get-file
On Sun, May 10, 2009 at 06:48:20PM +0200, kimaidou wrote: Waou ! This is great ! Thanks for this Hum, I just tried and I got this error: debian-gta02:~# apt-get-file install http://projects.openmoko.org/frs/download.php/575/ledclock_0.6_all.ipk trap: 26: ERR: bad trap Any clue ? Maybe. What happens if you replace this #!/bin/sh with this #!/bin/bash (assuming you have bash installed)? -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: what hardware version do I REALLY have?
On Sun, May 10, 2009 at 12:25:15PM -0700, Doug Jones wrote: [snip] ...which means hardware version 5, I think. Did they really ship me an older hardware version? The date code printed inside the phone is 20080605. As far as I can tell, the relationship between date code and PCB revision is this: Date code = 20080718 is GTA02v5. Date code = 20080722 is GTA02v6. Someone else (on community, support or hardware list) disassembled a GTA02 with date code 20080718 and found a PCB revision A5. I disassembled my GTA02 with date code 20080722 two weeks ago and found a PCB revision A6. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [debian] Installing *.ipk/*.opk part 2: apt-get-file
On Mon, May 11, 2009 at 10:55:39AM +0200, kimaidou wrote: My last pbm is my hackable apt-get wich does not want to install some packages.. but it is not related to your script. Did you add the repository from part 1? # echo /etc/apt/sources.list.d/ril-oedepmappings.list \ # 'deb http://nospamnospam.homepage.dk/software/download/oedepmappings/ /' -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [QTEi] battery duration
On Mon, May 11, 2009 at 09:05:39AM +0200, Franky Van Liedekerke wrote: Maybe recharging when connected to the usb should be optional ... This works on any recent kernel: # echo /sys/class/power_supply/usb/device/chg_curlim 0 (Although currently it is reset when you plug in power supply to the USB port. That will be fixed RSN.) -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [debian] Installing *.ipk/*.opk part 2: apt-get-file
On Mon, May 11, 2009 at 12:13:39PM +0200, kimaidou wrote: No I did not. But now it it done, and after a apt-get update, I still get: You might want to run `apt-get -f install' to correct these: The following packages have unmet dependencies: ledclock: Depends: python-pygtk but it is not going to be installed Depends: python-subprocess but it is not going to be installed E: Unmet dependencies. Try 'apt-get -f install' with no packages (or specify a solution). Is it a specific hackable:1 problem ? I don't know. I've seen that but it is not going to be installed message many times when trying to install packages (from the official Debian repository) on Debian. Apt-get is most unhelpful when it comes to error messages, basicly just saying: I know exactly what the problem is, but I won't tell you. :-( But, to try to debug the problem, what happens if you say # apt-get --dry-run install python-pygtk python-subprocess ? -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [shr-testing] sim not recognised
On Fri, May 15, 2009 at 11:37:17AM +1200, Robin Paulson wrote: what's happening, is there anything i can do to further debug things? GSM firmware revision? Have you flashed the moko11 version yet? http://wiki.openmoko.org/wiki/GSM/Flashing -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [shr-testing] sim not recognised
On Fri, May 15, 2009 at 09:47:18PM +1200, Robin Paulson wrote: right, everyone's convinced me. now, problem is i don't have a card reader, so i'll have to flash my micro sd via the freerunner. as the file is 270MB, i can't do that from the fr flash, so i'll be doing it from my desktop, over the usb connection. i'm guessing this is going to involve sshfs, but beyond that i'm in over my head... can i get some advice on how to do this? If you have the image unpacked[1] somewhere you can get to it from the FR, one of # scp host:/path/to/image /dev/mmcblk0 or # wget URL-to-image -O /dev/mmcblk0 on the FR ought to do the trick. Alternatively, from outside the FR: $ scp /path/to/image r...@gta02:/dev/mmcblk0 [1] It's a .tar.gz archive. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: BL-5C Nokia battery information NA, why?
On Fri, May 15, 2009 at 05:03:32PM +0200, Laszlo KREKACS wrote: Hi! Maybe the wiki is a bit disambiguous, but none of the replacement battery works in freerunner. Please define works. Note that you can simulate a non-GTA02 battery simply by putting a piece of paper or plastic between the middle pad of the battery and the middle pin of the battery connector. That said, the battery in freerunner has a Coulomb counter, which is required by the freerunner. Please define required. Note that you can quite easily try it out as explained above. All the other battery does not contain a coulumb counter (nokia bl-5c, 6c, etc), therefore freerunner CAN NOT charge it. Please define CAN NOT charge. Note that you can quite easily try it out as explained above. Simply put in a battery with middle pad covered and see if it charges. When you want to check the capacity, just take out the piece of paper or plastic. (If you can read schematics, also note how the coulomb counter is not connected to the charger circuit.) And CANT display the remaining capacity. True (sort of - with the 2.6.24 kernel you could at least read the battery voltage, which gives an indication of the charge level). The only way to use this batteries atm, is you charge the battery EXTERNALLY (via an another nokia phone or external charger), then you put into the freerunner and use it. Stop! Just before, you said: none of the replacement battery works in freerunner. However you will not have any indication about the remaining capacity. True (sort of - with the 2.6.24 kernel you could at least read the battery voltage, which gives an indication of the charge level). That said, if somebody write a correct driver it can be charged this type of batteries too. The problem with it, is the exact specification of the battery is not available, and if you charge incorrectly it can even explode. Already true. External power permitting, the Freerunner will try to charge whatever battery you insert to 4.15 V with a 1000 mA limit. So writing a correct driver requires many trial-and-error, and the possibility to destroy the battery and the freerunner too. Because you need to estimate how much you already charged the battery (the coulomb counter says exactly how many charges it took), and the remaining battery indicator would be just an approximation. (like the 3-4 bar on the mobile phone) Please explain why you need to know how much you already charged the battery. Hint: Search for charging procedure at URL:http://en.wikipedia.org/wiki/Lithium-ion_battery. See also the specifications for the Sanyo UF653450S as used in the Freerunner: http://battery.sanyo.com/en/spec/ion/UF653450S.pdf Btw, the coulomb counter doesn't know how much charge is lost to internal leakage in the battery. It is just a speculation, but imho writing this driver is not as difficult as it seems at the first look. True. In particular, there is already a GTA01 dumb battery driver. If somebody spy how the nokia phone charges its battery, we could replicate the process. We should simply monitor the voltage and the current of the battery. Needs some simple circuit(couple of wires and a soldering iron, and proper multimeter). Sure, but you'll find no surprices there. Correct me if Im wrong. In this particular case, it's easier to point out the things you're right about. :-( -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [all] usb headset connector?
On Fri, May 15, 2009 at 01:05:46PM +0200, arne anka wrote: i remember having asked that a long time ago, but here goes again: is there any way/ work going on to make one of those usb-to-headset-connectors working, most likely with a kernel module? something like that: http://www.amazon.com/Cingular-Stereo-Converter-Straight-Connector/dp/B000UCKPGW/ref=pd_rhf_f_i_cs_3/184-0114134-5978960 Tell you what: If you add the 11-pin mini USB headset port to the Neo so the adapters will fit, I'll do the kernel module. ;-) (Do read the reviews before buying, btw.) -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [all] usb headset connector?
On Sun, May 17, 2009 at 06:00:12PM +0200, arne anka wrote: ok, if i understand correctly, all of _these_ adapters[1] work by using additional pins not specified in the usb standard, ie not by some driver magic that sends sound via the standard pins, right? [1] not just the cingular one, but also http://www.amazon.com/3-5mm-Stereo-Adapter-T-Mobile-Diamond/dp/B0011WDLES/ref=pd_sbs_a_2_img/184-2544850-3162110 or http://www.amazon.com/OEM-Converter-Cingular-T-Mobile-Cavalier/dp/B001RXTDP2/ref=pd_cp_e_1_img/184-2544850-3162110?pf_rd_m=ATVPDKIKX0DERpf_rd_s=center-41pf_rd_r=03ZX1CK3AZE3RG0W2AVJpf_rd_t=201pf_rd_p=413863501pf_rd_i=B000UHY6XU Yes. On the first page you can clearly see the non-standard connector on the left of the adapter. On the second page it is even spelled out: HTC ExtUSB. http://en.wikipedia.org/wiki/Universal_Serial_Bus#Proprietary_connectors_and_formats Pinout: https://www.sparkfun.com/commerce/product_info.php?products_id=9078 -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: New Life in Openmoko Phones
On Tue, May 19, 2009 at 10:52:22AM -0400, Stefan Monnier wrote: That still doesn't explain why removing one of the two accelerometers is a good idea. What is the benefit? Why not remove them both? Is it that all the programs that use the accelerometers (as of now) only use one of the two? Is it that having two accelerometers introduces layout difficulties? Is it that there aren't enough interrupt lines on the SoC to properly support the two accelerometers? ... I don't think so. We already have both EINT8/GPG0 and EINT16/GPG8 reserved for the second accelerometer, but not connected it to EINT16/GPG8 (R1547 = NC). Last time I counted[1], there wasn't a shortage of interrupt or GPIO pins. [1] http://lists.openmoko.org/pipermail/gta03/2009-April/74.html If we're really going to mess with the accelerometers, why not move them off the SPI1 bus and onto GPIO pins? We're currently using the bitbanging GPIO_SPI driver anyway. That way, we could keep a GTA01/GTA02 compatible debug connector (because WLAN could use SPI1 instead of SPI0). I actually have the same question for the audio-amp: why remove it? But that one is a bit more complicated, because I'm not sure what is this audio-amp anyway (is it the thing that drives the headphone plug?) It drives either the headphone speakers or the bottom handset speaker, depending on the presence of the headphone plug. It's the LM4853 (U4101 on page 7 of the schematics). IIUC, the GTA03 was going to drive both from the WM8753L sound chip directly as suggested in the WM8753L datasheet. I wanted to compare the output power of the two, but I can't find the exact LM4853 variant we're using. -- Rask Ingemann Lambertsen Danish law requires addresses in e-mail to be logged and stored for a year ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community