**** BEGIN LOGGING AT Sun Nov 25 02:59:56 2007
Nov 25 03:29:49 <ewon> time warner cable is a pile of shite
Nov 25 07:26:27 <impresive> äàðîôôôô
Nov 25 09:19:33 <Taco> Has anyone seen the following problem with their neo:   I can make outgoing phone calls successfully, using the dialer GUI, and incoming calls produce a ring.   But I can't answer the phone.   Regardless of which icon I click on, the call does not get answered, and shows up in the "missed calls" list.
Nov 25 09:20:05 <CM> Heh, for me it doesn't ring, only buzz, but I can answer incoming calls
Nov 25 09:20:20 <Taco> I'll trade ya...
Nov 25 09:31:59 <raster> moo
Nov 25 09:32:11 <CM> :)
Nov 25 09:32:50 <raster> (:
Nov 25 09:33:04 <CM> Hehe.. I know you're "down under" already :P
Nov 25 09:36:07 <raster> i am not under someone's desk
Nov 25 09:36:14 <raster> (:
Nov 25 09:36:21 <Q_Continuum> haha
Nov 25 09:45:48 <Taco> Does anyone know if the dropbear ssh client built into the standard rootfs distributions can forward X11 traffic?
Nov 25 09:47:40 <rwhitby> Taco: I don't think it can - that's why I always replace dropbear with openssh, for X11 and authentication agent forwarding.
Nov 25 09:49:22 <Taco> Thanks.   THat's sorta what I thought needed to be done.   Evidently, one can build dropbear in a way that allows X11 forwarding, but that does not appear to be what has been done in the standard OM distribution.
Nov 25 09:49:54 <Taco> I'd rather have "real" ssh.
Nov 25 09:50:25 <CM> Taco: I think rwhitby wrote in his blog how to change to openssh :)
Nov 25 09:50:43 <Taco> Cool - I'll look there.   Thanks for the tip.
Nov 25 09:51:28 <CM> Taco: http://www.rwhitby.net/blog/nslu2-linux/replacing-dropbear-with-openssh.html
Nov 25 09:52:05 <Taco> Gee - now that's service!   Thanks again.
Nov 25 09:52:25 <rwhitby> ~emulate rwhitby
Nov 25 09:52:26 <apt> Another Satisfied Customer!
Nov 25 09:53:11 <CM> ~praise rwhitby 
Nov 25 09:53:13 <apt> All hail rwhitby!
Nov 25 09:53:16 <CM> :D
Nov 25 09:53:32 <AzraiL> :)(
Nov 25 09:59:11 <Taco> Since I'm having good luck getting my questions answered, I'll risk asking another.   Sometimes when I boot my neo, all works well, and I get a popup box saying it has connected to the GSM network.   However, more often than not, after booting it shows many bars on the graphic signal strength indicator, but it never actually connects to the network.   Nothing from gsmd shows up in /var/log/messages.   Any idea what is wrong when it 
Nov 25 09:59:11 <Taco> has good signal strength, but doesn't connect to the network?
Nov 25 10:09:17 <Taco> It's absolutely amazing to see "ipkg install" work on a cell phone.
Nov 25 10:09:43 <ljp> dont need to be registered to the network to actuall receive signal strength indications
Nov 25 10:11:35 <Taco> /etc/init.d/gsmd restart doesn't fix the problem, but rebooting sometimes does.   Is there any way to interactively force a complete restart of the whole GSM software stack short of rebooting?
Nov 25 10:33:24 <Taco> Here's dumb question: What is the small, gold-plated connector roughly two centimeters above the USB port on the neo?
Nov 25 10:33:42 <Taco> Make that 3 centimeters...
Nov 25 10:38:12 <tuukkah> Taco, external gps antenna connector
Nov 25 10:38:32 <Taco> Oh.   Thanks!
Nov 25 10:46:37 <tuukkah> i'll add connectors to the getting started wiki page
Nov 25 10:47:08 <Taco> Tuukkah: You're a good man.
Nov 25 10:47:48 <tuukkah> no prob, thanks!
Nov 25 11:10:38 <aapo> hi, is there easy way to connect network with openmoko running under qemu?
Nov 25 11:12:10 <CM> aapo: http://wiki.openmoko.org/wiki/OpenMoko_under_QEMU#Setting_up_USB_connection
Nov 25 11:12:45 <CM> aapo: It's not very easy, but the steps are quite well described :)
Nov 25 11:13:15 <NineX> someone tried to compile irssi for openmoko?
Nov 25 11:13:48 <aapo> cm, thanks, i hoped there is something easier
Nov 25 11:14:22 <IfChinsCouldKill> Any experience with kqemu?
Nov 25 11:14:29 <CM> NineX: It's already in oe
Nov 25 11:15:00 <NineX> CM: oe? (please explain - my english still sucka ;) )
Nov 25 11:15:10 <CM> Heh, it's not your english
Nov 25 11:15:18 <CM> oe = OpenEmbedded
Nov 25 11:15:30 <NineX> CM: ok
Nov 25 11:15:39 <CM> The "distro buiding tool" that OpenMoko is using
Nov 25 11:15:52 <tuukkah> CM, for easier networking, what about say pppd over serial?
Nov 25 11:15:58 <borg_> NineX: if you have a neo you can install irssi with the application manager
Nov 25 11:16:05 <borg_> or with ipkg install irssi from the console
Nov 25 11:16:31 <CM> tuukkah: Hmm.. Might be, but I don't know much about that stuff 
Nov 25 11:17:28 <tuukkah> aapo, would you be interested in that?
Nov 25 11:17:34 <aapo> yes
Nov 25 11:18:26 <NineX> borg_: thanks. 
Nov 25 11:18:47 <CM> Heh, irssi was smaller than I thought. Only 600k to install
Nov 25 11:19:00 <aapo> tuukkah, I have qemu running, host system is ubuntu gutsy
Nov 25 11:19:34 <CM> NineX: If ipkg complains that it can't find the program, add ScaredyCat's repository: http://wiki.openmoko.org/wiki/Repositories
Nov 25 11:19:49 <NineX> what about screen orientation change?
Nov 25 11:20:05 <CM> Hold in the aux button and select it from the menu :)
Nov 25 11:20:38 <NineX> CM: thanks for help i will have lot of fun with neo phone :>
Nov 25 11:20:51 <CM> You're welcome :)
Nov 25 11:21:12 <Taco> You will have fun.   It's not a phone, it's a hobby!
Nov 25 11:22:16 <tuukkah> aapo, i don't know if this is easy either. but it could be worth a try to use the qemu "-serial pty" option and then one pppd within qemu and one on the host
Nov 25 11:22:38 <NineX> i'm thinkink about syslog change (writing logs into nand flash will decarase lifetime of this memory i think)
Nov 25 11:23:04 <NineX> s/tinkik/thinking/
Nov 25 11:23:13 <NineX> s/i'm/i/
Nov 25 11:23:13 <apt> NineX meant: i thinkink about syslog change (writing logs into nand flash will decarase lifetime of this memory i think)
Nov 25 11:23:46 <balrog-kun> tuukkah: i think two pppd's won't work on the same modem
Nov 25 11:24:36 <tuukkah> balrog-kun, what do you mean? of course on the different sides of the pty
Nov 25 11:25:19 <balrog-kun> tuukkah: oh, you mean without physiacl ppp connection?
Nov 25 11:26:00 * balrog-kun didn't know pppd can work as a ppp server to share internet
Nov 25 11:26:25 <tuukkah> pppd is point-to-point or peer-to-peer :-)
Nov 25 11:26:50 <balrog-kun> cool
Nov 25 11:27:17 <aapo> ok, I started qemu with -serial pty, now its waits (starting kernel...)
Nov 25 11:31:13 <NineX> apt: i will trye to reconfigure syslog to write log in microsd card, , cause if system make lot of logs better ides is put system log into microsd card (that can be changed) than use nand flash that habe 100000 erase cycles (i think)
Nov 25 11:31:14 <apt> You will trye to reconfigure syslog to write log in microsd card, , cause if system make lot of logs better ides is put system log into microsd card (that can be changed) than use nand flash that habe 100000 erase cycles (i think)?
Nov 25 11:31:35 <NineX> s/ides/idea/
Nov 25 11:31:35 <apt> NineX meant: apt: i will trye to reconfigure syslog to write log in microsd card, , cause if system make lot of logs better idea is put system log into microsd card (that can be changed) than use nand flash that habe 100000 erase cycles (i think)
Nov 25 11:33:41 <NineX> stupid bot :S
Nov 25 11:33:46 <zilux> NineX: log are in tmpfs (log -> volatile/log)
Nov 25 11:37:53 <Taco> Is tmpfs in RAM then?
Nov 25 11:41:14 <tuukkah> aapo, qemu prints a pty file name when it starts, can you see what it is?
Nov 25 11:42:18 <tuukkah> aapo, i see "char device redirected to /dev/pts/5" so the boot continues when i read from that with say "cat /dev/pts/5"
Nov 25 11:45:27 <aapo> yes, now it starts
Nov 25 11:46:00 <zilux> Taco: yes,  more info: http://lxr.linux.no/source/Documentation/filesystems/tmpfs.txt 
Nov 25 11:46:24 <Taco> zilux: Thanks for the pointer.
Nov 25 11:52:50 <tuukkah> aapo, i'm getting some kind of a pppd connection by killing gsmd inside qemu and running pppd nodetach debug /dev/ttySAC0
Nov 25 11:53:21 <tuukkah> then on the host pppd nodetach debug 192.168.100.1:192.168.100.2 noauth /dev/pts/5
Nov 25 11:53:41 <aapo> i will try
Nov 25 11:56:01 <cdbot2> * * OM Bug 1023 has been created by hns(AT)computer.org
Nov 25 11:56:02 <cdbot2> * * CDC/RNDIS Ethernet over USB driver provides random Ethernet address after each reboot which makes MacOS X config
Nov 25 11:56:03 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1023
Nov 25 11:58:38 <tuukkah> hmm, ping 192.168.100.2 worked, but ssh root@192.168.100.2 was slow and eventually crashed the qemu...
Nov 25 11:59:05 <aapo> openmoko says "tcflush failed: bad file descriptor" "tcsetattr: invalid argumetn (line 1003)"
Nov 25 12:00:59 <tuukkah> i think i got something like at some point and just tried again
Nov 25 12:01:12 <aapo> now ping works
Nov 25 12:01:12 <tuukkah> s/like/like that/
Nov 25 12:01:13 <apt> tuukkah meant: i think i got something like that at some point and just tried again
Nov 25 12:01:51 <aapo> maybe host pppd must start first
Nov 25 12:04:20 <aapo> i got ssh working moko -> host
Nov 25 12:04:36 <tuukkah> did you try in the other direction?
Nov 25 12:04:51 <aapo> but, now qemu crashes
Nov 25 12:05:15 <tuukkah> ok, so same as here :-(
Nov 25 12:05:26 <aapo> but this is good start
Nov 25 12:05:44 <tuukkah> yeah, it works pretty easily... in theory :-)
Nov 25 12:06:11 <aapo> yes, next step will be network redirecting
Nov 25 12:06:24 <aapo> i meen internet sharing
Nov 25 12:06:27 <tuukkah> i'm now trying with "-serial vc -serial pty" so the kernel output and gsmd won't interfere
Nov 25 12:06:49 <tuukkah> aapo, that should be easy. i'm more worried about the qemu crash
Nov 25 12:06:53 <edistar> is it possible to access the internet with OM through qemu?
Nov 25 12:07:17 <tuukkah> edistar, it is, but we're trying to find an easy way
Nov 25 12:07:51 <edistar> tuukkah: ok, cool
Nov 25 12:08:00 <tuukkah> edistar, seems the current suggested way is this a bit complicated http://wiki.openmoko.org/wiki/QEMU#Setting_up_USB_connection
Nov 25 12:08:34 <edistar> yes, it's a bit of work
Nov 25 12:26:08 <aapo> I tried ssh host -> moko. connected succesfully, but almost immediatly qemu crashed (segmentation fault)
Nov 25 12:27:29 <tuukkah> aapo, how recent versions do you have of qemu and openmoko?
Nov 25 12:28:03 <tuukkah> also, did you try the two "-serial" options?
Nov 25 12:29:47 <aapo> http://wiki.openmoko.org/wiki/Automatic_emulation_in_Ubuntu
Nov 25 12:29:59 <aapo> i made that ubuntu magic
Nov 25 12:30:38 <aapo> so the openmoko is newest, but i do not know my qemus version
Nov 25 12:32:08 <tuukkah> aapo, actually, the other way ;-) that will build the latest qemu, but use "the lastest release" of openmoko
Nov 25 12:34:00 <aapo> i did not try two -serial options
Nov 25 12:36:04 <tuukkah> i didn't have luck, perhaps because my qemu is old, but with "-serial vc -serial pty" the pty should be visible in the emulator as /dev/ttySAC1
Nov 25 12:38:00 <aapo> ok, if you find something, write it to wiki
Nov 25 12:43:25 <PBeck> hi
Nov 25 13:31:05 <tuukkah> aapo_away, seems to work fine that way
Nov 25 13:51:32 <balrog-kun> hmm the segfault shouldn't happen
Nov 25 13:52:56 <balrog-kun> it might not work with ttySAC0 because kernel dumps dmesg there and because the modem is there but SAC1 should be ok
Nov 25 13:53:26 <tuukkah> yeah, SAC1 seems to work fine
Nov 25 13:54:32 <tuukkah> do you think it's somehow possible to use SAC2 or something that won't conflict with gps in the future?
Nov 25 13:54:33 <rangen> is there a discussion about out of memory handling in openmoko somewhere?
Nov 25 13:54:51 <balrog-kun> tuukkah: yes, ttySAC2 should also work
Nov 25 13:55:56 <balrog-kun> -serial stdio -serial none -serial /dev/ptswhatever
Nov 25 13:56:49 <tuukkah> balrog-kun, i created /dev/ttySAC2 too, but can't use it: No such device or address
Nov 25 13:57:51 <tuukkah> actually i used "-serial null", "void device"
Nov 25 13:59:29 <balrog-kun> ah, the om kernel registers only two ports
Nov 25 13:59:55 <balrog-kun> it may be possible to override it from boot params line
Nov 25 14:00:22 <balrog-kun> but likely isn't :/
Nov 25 14:00:33 <tuukkah> so perhaps for now SAC1 is fine
Nov 25 14:01:49 <balrog-kun> SAC1 is unused as long as /sys/devices/platform/gta01.gps-pm/power reads 0
Nov 25 14:12:58 <tuukkah> http://wiki.openmoko.org/wiki/OpenMoko_under_QEMU#Simple_network_connection_with_pppd
Nov 25 14:16:31 <balrog-kun> cool
Nov 25 14:16:33 <balrog-kun> tuukkah: thanks
Nov 25 14:16:55 <edistar> tuukkah: did you write that just now?
Nov 25 14:17:00 <tuukkah> yeah
Nov 25 14:17:03 <edistar> cool
Nov 25 14:17:04 <edistar> thx
Nov 25 14:17:10 <tuukkah> np
Nov 25 14:18:48 <borg_> someone here who wants to test my display lock? :)
Nov 25 14:19:24 <tuukkah> i'd love to if i wasn't busy right now
Nov 25 14:19:32 <borg_> ;)
Nov 25 14:21:27 <tuukkah> hmm, do we ship three or four pairs of the earbud rubber parts?
Nov 25 14:25:57 <PBeck> borg_: ui have you screenshots? :)
Nov 25 14:26:14 <borg_> http://homepages.upb.de/olaf/dinge/openmoko/display_lock/openmoko-display-lock-readme.txt
Nov 25 14:26:20 <PBeck> i have no neo, sorry :(
Nov 25 14:26:30 <borg_> i need someone who makes me some images too :)
Nov 25 14:27:05 <borg_> something like a key and key hole or a cat and a basket
Nov 25 14:27:12 <borg_> something that one can put in another thing.. ;)
Nov 25 14:27:26 <tuukkah> PBeck, there's a screenshot linked: http://homepages.upb.de/olaf/dinge/openmoko/display_lock/some_dragging.png
Nov 25 14:27:50 <tuukkah> borg_, so you got drag and drop working ?-)
Nov 25 14:27:54 <borg_> yes
Nov 25 14:27:58 <borg_> it works totally nice :)
Nov 25 14:28:00 <borg_> imo ;)
Nov 25 14:28:06 <borg_> but i need some pictures
Nov 25 14:28:23 <edistar> borg_: you need a video!
Nov 25 14:28:52 <borg_> omg, i cant even make images for it.. ;)
Nov 25 14:28:53 <_buz> isnt there a qtopia video that shows that
Nov 25 14:29:15 <borg_> edistar: http://homepages.upb.de/olaf/dinge/openmoko/display_lock/neod_0.1.0+svnr3421-r6_neo1973.ipk
Nov 25 14:29:18 <borg_> you can test it for yourself ;)
Nov 25 14:29:21 <PBeck> borg_: hehe cool :)
Nov 25 14:29:39 <PBeck> tuukkah: thanks, i have found it in the readme :)
Nov 25 14:31:02 <tuukkah> what would be a fancier name than 'hole' for the place where the microphone is located on the case?
Nov 25 14:31:31 <borg_> plug?
Nov 25 14:32:00 <tuukkah> no i mean there's this hole that goes through the phone, where you can attach the lanyard too
Nov 25 14:32:07 <borg_> ah
Nov 25 14:32:09 <tuukkah> the loop?
Nov 25 14:32:09 <borg_> mh
Nov 25 14:32:18 <PBeck> http://wiki.openmoko.org/wiki/User:Quicksand#zedlock borg_ 
Nov 25 14:32:24 <PBeck> thats a good solution too :)
Nov 25 14:32:28 <abraxa_> tuukkah: "oval opening"?
Nov 25 14:33:20 <borg_> PBeck: but mine is ready :P
Nov 25 14:33:26 <tuukkah> abraxa_, thanks!
Nov 25 14:33:28 <borg_> but i can put some status information in it too
Nov 25 14:34:22 <PBeck> borg_: yes ;)
Nov 25 14:35:02 <PBeck> borg_: i like your idea,  you need only a few graphics ;)
Nov 25 14:35:12 <PBeck> borg_: see qtopia for example
Nov 25 14:35:17 <PBeck> key in keyhole
Nov 25 14:43:06 <tuukkah> entering pin would be too boring?
Nov 25 14:43:28 <PBeck> tuukkah: that should be possible too, i think
Nov 25 14:43:46 <PBeck> that should be a good security solution
Nov 25 14:44:04 <tuukkah> while we're at it, why not require solving a sudoku ;-)
Nov 25 14:44:24 <borg_> i dont want to enter my pin every time the display of my phone gets locked
Nov 25 14:44:27 <borg_> dont know
Nov 25 14:44:36 <borg_> but we can make it user configureable
Nov 25 14:44:58 <borg_> but then, there is no settings application at all at the moment... ;)
Nov 25 14:45:00 <tuukkah> yeah i think that's something that should be configurable
Nov 25 14:45:15 <tuukkah> for now, what you have is cool
Nov 25 14:45:57 <borg_> mickey|sports: where should i put patches for neod?
Nov 25 14:46:00 <PBeck> borg_: yes of course configurable
Nov 25 14:46:37 <tuukkah> borg_, have you checked if there's already a bugzilla entry for screen lock?
Nov 25 14:46:44 <abraxa_> borg_: Bugzilla :)
Nov 25 14:46:59 <borg_> ok
Nov 25 14:49:01 <ScaredyCat> borg_:  you have a .bb ?
Nov 25 14:49:34 <borg_> ScaredyCat: yes
Nov 25 14:49:48 <ScaredyCat> are you sharing it :)
Nov 25 14:50:52 <borg_> patches: http://homepages.upb.de/olaf/dinge/openmoko/display_lock/patches/
Nov 25 14:51:02 <borg_> bb: http://homepages.upb.de/olaf/dinge/openmoko/display_lock/neod_svn.bb
Nov 25 14:51:14 <borg_> the images are in here too: http://homepages.upb.de/olaf/dinge/openmoko/display_lock/
Nov 25 14:51:18 <borg_> but they are bad
Nov 25 14:51:26 <borg_> i hope someone can make some nice looking :)
Nov 25 14:53:57 <abraxa_> borg_: To unlock, one has to drag the white box from bottom to top?
Nov 25 14:54:02 <borg_> yes
Nov 25 14:54:51 <borg_> abraxa_: you could drag a key in a keyhole ar a cat in a basket or something
Nov 25 14:54:55 <borg_> but i cant make images :)
Nov 25 14:55:03 <abraxa_> Ah
Nov 25 14:55:36 <borg_> abraxa_: http://homepages.upb.de/olaf/dinge/openmoko/display_lock/openmoko-display-lock-readme.txt
Nov 25 14:56:18 <borg_> here we are: http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=572
Nov 25 14:57:17 <abraxa_> Hmm
Nov 25 14:59:05 <abraxa_> Personally, I'd prefer using zedlock and while hardcoding this feature in neod works for now, it can't be the long-term solution
Nov 25 14:59:38 <abraxa_> So if you have the time it would be nice if you could create d-bus bindings within neod and a lock-daemon that acts on these d-bus events
Nov 25 15:00:14 <abraxa_> That would make things flexible and future-proof :)
Nov 25 15:00:31 <borg_> abraxa_: yes, neod is bad :)
Nov 25 15:00:34 <tuukkah> is d-bus needed or is it enough to run a program which returns 0 to unlock or non-0 to keep locked
Nov 25 15:00:35 <roh> gnome-screensaver-api via dbus?
Nov 25 15:00:53 <abraxa_> tuukkah: d-bus will be better
Nov 25 15:00:56 <roh> has all the features
Nov 25 15:01:18 <abraxa_> roh: Good idea... so we just need d-bus bindings for that in neod?
Nov 25 15:01:23 <tuukkah> roh, if it has pluggable unlocking, that's great
Nov 25 15:03:20 <borg_> the question is how neod should look like in the future
Nov 25 15:03:43 <tuukkah> did some reorg and added http://wiki.openmoko.org/wiki/Getting_Started_with_your_Neo1973#Interconnections
Nov 25 15:03:48 <abraxa_> borg_: neod will go
Nov 25 15:05:11 <tuukkah> hmm, before "Getting shell access", there should perhaps be a section about how to use the gui %-)
Nov 25 15:05:39 <borg_> abraxa_: where will it go? ;)
Nov 25 15:05:56 <roh> http://svn.gnome.org/viewvc/gnome-screensaver/trunk/src/gs-listener-dbus.h?revision=994&view=markup
Nov 25 15:06:04 <roh> http://svn.gnome.org/viewvc/gnome-screensaver/trunk/src/gs-listener-dbus.c?revision=1294&view=markup
Nov 25 15:06:17 <roh> pretty simple interface..
Nov 25 15:06:38 <roh> "org.gnome.ScreenSaver"
Nov 25 15:11:04 <abraxa_> wb mickeyl - got any comments on the issue?
Nov 25 15:13:46 <mickeyl> issues?
Nov 25 15:14:06 * mickeyl needs more context
Nov 25 15:14:10 <mickeyl> what are we talking about?
Nov 25 15:15:11 <mickeyl> aha, something in between suspend and dim-only
Nov 25 15:15:22 <mickeyl> lock-but-don't-suspend, so to speak
Nov 25 15:15:26 <borg_> mickeyl: http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=572
Nov 25 15:16:32 <mickeyl> hmm
Nov 25 15:16:33 <mickeyl> aha
Nov 25 15:16:37 <mickeyl> wrong component
Nov 25 15:16:44 <mickeyl> that's why i didn't got that bug
Nov 25 15:16:50 * mickeyl changes component
Nov 25 15:20:22 <mickeyl> no comments offhand, need to think about it. doing some work on gsm applet atm.
Nov 25 15:21:10 <borg_> mickeyl: http://homepages.upb.de/olaf/dinge/openmoko/display_lock/neod_0.1.0+svnr3421-r6_neo1973.ipk
Nov 25 15:21:17 <borg_> if you want to test it shortly ;)
Nov 25 15:22:06 <tuukkah> mickeyl, could you check the component of this old bug too: http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=847
Nov 25 15:25:07 <borg_> abraxa_: i could deamonize neod and add dbus bindings and outsource the aux/power menu and the lock display, but i wont do it when neod will go anyway :)
Nov 25 15:25:55 <borg_> so it is like with every other application in openmoko, no information about the future of them at all.
Nov 25 15:26:30 <tuukkah> borg_, there's the neod readme: http://svnweb.openmoko.org/trunk/src/target/OM-2007.2/daemons/neod/README?view=auto
Nov 25 15:27:08 <borg_> ..and what does it say?
Nov 25 15:27:29 <borg_> there might be something somewhere
Nov 25 15:27:41 <borg_> it doesnt say what should be done
Nov 25 15:28:00 <tuukkah> right, no definite plan
Nov 25 15:28:08 <mickeyl> borg_: will test, thsnk
Nov 25 15:29:04 <mickeyl> tuukkah: hmm, 847 is interesting. i'm not sure which component should trigger the shutdown though. that might be a task for neod
Nov 25 15:29:35 <borg_> mickeyl: can you say something about the future of neod?
Nov 25 15:29:45 <borg_> i am willing to deamonize it and make some dbus bindings
Nov 25 15:30:06 <borg_> but only if it wont be replaced by something else in 1 month or so.. ;)
Nov 25 15:30:11 <mickeyl> borg_: sounds like a good plan
Nov 25 15:30:39 <mickeyl> my plan is to replace it with a generic peripheral and power management daemon, however this won't happen in 1 month
Nov 25 15:30:54 <mickeyl> so i think improving neod is a good thing, because that's what works now#
Nov 25 15:32:58 <goxboxlive> hi mickeyl : In the latest openmoko images i dont have any gtkrc file. I want to test the moko-engine on htcuniversal because of speed issue. Do you know how i do that?
Nov 25 15:34:49 <abraxa_> goxboxlive: The gtkrc file for the moko-GTK-engine should be installed to a specific theme upon installation of the engine
Nov 25 15:35:37 <mickeyl> goxboxlive: set the theme via gconf
Nov 25 15:35:50 <goxboxlive> thanks both, i'll test
Nov 25 15:36:27 <abraxa_> goxboxlive: http://svnweb.openmoko.org/trunk/src/target/OM-2007.2/libraries/moko-gtk-engine/Makefile.am?rev=3398&view=markup
Nov 25 15:36:38 <mickeyl> gconftool-2 --type string --set /desktop/poky/interface/theme Moko
Nov 25 15:36:41 <mickeyl> should be it
Nov 25 15:41:26 <mickeyl> is anyone else seeing gsmd oscillating between non-connected and connected ?
Nov 25 15:41:38 <borg_> mickeyl: yes
Nov 25 15:41:48 <mickeyl> this is new, isn't it?
Nov 25 15:41:49 <borg_> thats the normal behavior of gsmd since 2 month
Nov 25 15:41:52 <borg_> no
Nov 25 15:41:52 <borg_> :)
Nov 25 15:41:57 <mickeyl> hmm
Nov 25 15:42:03 <mickeyl> i wonder what's wrong there
Nov 25 15:42:19 <mickeyl> i mean this is sent by network
Nov 25 15:42:25 <mickeyl> so it can't really be gsmd at fault here
Nov 25 15:43:12 <mickeyl> hmm
Nov 25 15:43:14 <mickeyl> well
Nov 25 15:43:23 <mickeyl> why are we constantly sending SMS commands?
Nov 25 15:43:42 <mickeyl> maybe these are triggering the network reregistration
Nov 25 15:44:17 <hdr> are there any news about car-navigation with openmoko?
Nov 25 15:44:23 <mickeyl> hmm, let me see what happens if I patch this out
Nov 25 15:44:52 <balrog-kun> mickeyl: we are seding some sms setup commands on every network registration because they can't be sent before that
Nov 25 15:45:00 <borg_> hdr: http://scap.linuxtogo.org/
Nov 25 15:45:04 <borg_> look at pyroute
Nov 25 15:45:07 <borg_> looks nice imo :)
Nov 25 15:45:12 <balrog-kun> mickeyl: in current svn i think we also send them spuriously on evry cell change
Nov 25 15:46:24 <mickeyl> ooh
Nov 25 15:46:31 <mickeyl> that could be possible
Nov 25 15:46:44 <mickeyl> could they trigger network reregistration
Nov 25 15:46:45 <mickeyl> ?
Nov 25 15:46:59 <mickeyl> i'm trying to identify the cause, not the effect
Nov 25 15:47:00 <balrog-kun> mickeyl: they shouldn't
Nov 25 15:47:25 <mickeyl> i remember that last time i looked, it stayed connected for long when i didn't move
Nov 25 15:47:32 <mickeyl> s/connected/registrated/
Nov 25 15:47:32 <apt> mickeyl meant: i remember that last time i looked, it stayed registrated for long when i didn't move
Nov 25 15:48:21 <balrog-kun> they can cause unwanted call disconnection, which why there are patches for it on the gsmd list ;)
Nov 25 15:49:22 <hdr> borg_: nice ;) what's about maps of germany? i guess you're german?
Nov 25 15:50:08 <borg_> hdr: yes, i am german, but i dont know what maps they are using
Nov 25 15:51:04 <hdr> (names
Nov 25 15:51:25 <hdr> borg_: ok danke ;)
Nov 25 15:51:26 <mickeyl> balrog-kun: ok, then this will probably be applied soon
Nov 25 15:51:40 <mickeyl> for now i really need to find out how to fix this continuous reregistration
Nov 25 15:51:45 <ScaredyCat> borg_: check out scap
Nov 25 15:52:28 <borg_> ScaredyCat: nice !
Nov 25 15:52:29 <borg_> :D
Nov 25 15:52:33 <borg_> i like it
Nov 25 15:53:04 <ScaredyCat> not sure what 'south' should represent
Nov 25 15:53:58 <mickeyl> balrog-kun: got a link to the mail where you posted this patch?
Nov 25 15:56:43 <mickeyl> jserv--: i have a problem with gsmd
Nov 25 15:57:11 <balrog-kun> one sec
Nov 25 15:57:14 <mickeyl> jserv--: i don't know when it started, but it looks like it's constantly dropping registration and reregistration
Nov 25 15:57:24 <mickeyl> jserv--: i.e. even if i don't move
Nov 25 15:57:32 <jserv--> mickeyl, ouch
Nov 25 15:57:38 <mickeyl> any idea what could cause it?
Nov 25 15:57:41 <Stephmw> mickeyl: network trying to tell you to stay out?
Nov 25 15:58:00 <jserv--> mickeyl, it seems that we have to figure out the problem by monitoring TI debug tool.
Nov 25 15:58:01 <mickeyl> Stephmw: network sending CREG=0, then immediately sending reergistrating
Nov 25 15:58:17 <mickeyl> let me post you something
Nov 25 15:58:27 <balrog-kun> mickeyl: turns out i didn't post a patch, but posted a pointer to it in http://lists.openmoko.org/pipermail/gsmd-devel/2007-October/000299.html
Nov 25 15:59:13 <mickeyl> jserv--: this is with moko5 firmware
Nov 25 15:59:21 <mickeyl> http://pastebin.ca/796919
Nov 25 15:59:38 <mickeyl> creg0, creg1, creg0, creg1
Nov 25 15:59:46 <mickeyl> without me sending any commands nor moving the station
Nov 25 16:01:02 <jserv--> mickeyl, ok, I'll check it tomorrow.
Nov 25 16:01:17 * jserv-- is about to enter factory tomorrow. :-)
Nov 25 16:01:30 <ScaredyCat> ditched the south - or at least made it black borg_ ...
Nov 25 16:01:46 <mickeyl> jserv--: ok. for reference, it's
Nov 25 16:01:48 <mickeyl> !ombug 1024
Nov 25 16:01:50 <cdbot2> * * Bug 1024, Status: NEW, Created: Unknown
Nov 25 16:01:51 <cdbot2> * * mickey(AT)vanille-media.de: gsm modem oscillating between registrated / not-registrated
Nov 25 16:01:52 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1024
Nov 25 16:01:52 <ScaredyCat> not 'black borg_'
Nov 25 16:01:59 <balrog-kun> mickeyl: this is strange
Nov 25 16:02:01 <cdbot2> * * OM Bug 1024 has been created by mickey(AT)vanille-media.de
Nov 25 16:02:02 <cdbot2> * * gsm modem oscillating between registrated / not-registrated
Nov 25 16:02:03 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1024
Nov 25 16:02:13 <mickeyl> balrog-kun: indeed. worrying me
Nov 25 16:02:17 <borg_> ScaredyCat: looks good :)
Nov 25 16:02:18 <mickeyl> balrog-kun: could my hardware be faulty?
Nov 25 16:02:22 <borg_> can you send me the pictures?
Nov 25 16:02:28 <ScaredyCat> sure
Nov 25 16:02:31 <borg_> i will include them in the ipk
Nov 25 16:02:34 <borg_> olaf@uni-paderborn.de
Nov 25 16:02:48 <balrog-kun> mickeyl: i don't know, is there a way to check if it happens on a different phone with the same sim?
Nov 25 16:03:35 <mickeyl> balrog-kun: hmm, yes. i have one other neo.
Nov 25 16:03:42 <borg_> ScaredyCat: and it works ok for you?
Nov 25 16:03:42 * mickeyl flashes same software to it
Nov 25 16:04:04 <balrog-kun> mickeyl: no, i mean some mass market phone :)
Nov 25 16:04:15 <mickeyl> oh
Nov 25 16:04:21 <balrog-kun> but i guess it may be happening silently and other phones may not let the user know at all
Nov 25 16:04:35 <mickeyl> yeah
Nov 25 16:04:37 <mickeyl> _however_
Nov 25 16:04:47 <mickeyl> i think i didn't see this previously
Nov 25 16:05:12 <ScaredyCat> borg_: yeah, works well...
Nov 25 16:05:16 <ScaredyCat> http://buildhost.automated.it/lock-images.tar.gz
Nov 25 16:05:22 <borg_> ok :)
Nov 25 16:07:50 <ScaredyCat> brb
Nov 25 16:08:12 <Risto> Hi there, currently trying to compile OM, but it fails to build xserver-kdrive-1.3.0.0 -> http://pastebin.com/d60d412ae Anybody know how to solve this?
Nov 25 16:08:15 <balrog-kun> mickeyl: i notice in your log AT+COPS hasn't been issued yet?
Nov 25 16:08:33 <mickeyl> balrog-kun: this is just an excerpt
Nov 25 16:08:41 <mickeyl> at+cops has been sent 
Nov 25 16:09:10 <balrog-kun> mickeyl: but AT%CPI is there and AT+COPS i think should be later but i may be wrong
Nov 25 16:10:42 <mickeyl> balrog-kun: please see the corresponding libgsmd-tool log
Nov 25 16:10:46 <mickeyl> http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1024
Nov 25 16:10:54 <mickeyl> i did AT+COPS manually to be sure it has been issued
Nov 25 16:11:04 <mickeyl> and it shows the oscillation as well
Nov 25 16:11:20 <ScaredyCat> Risto: clean xserver-kdrive
Nov 25 16:11:26 <mickeyl> this is a completely new behavour, i never have seen this problem
Nov 25 16:11:50 <balrog-kun> mickeyl: hmm i see
Nov 25 16:12:03 <balrog-kun> i haven't seen this behaviour before either
Nov 25 16:12:12 <mickeyl> ok, thanks
Nov 25 16:12:17 <mickeyl> may be my hardware
Nov 25 16:12:23 <mickeyl> because i also have seen some other strange things
Nov 25 16:12:30 <mickeyl> CME Error: 23
Nov 25 16:12:32 <mickeyl> uh oh
Nov 25 16:12:38 <mickeyl> "Memory Failure"
Nov 25 16:12:41 <mickeyl> did you see that one?
Nov 25 16:12:59 <balrog-kun> ouch, nope never
Nov 25 16:13:02 <mickeyl> ok
Nov 25 16:13:06 <mickeyl> looks like hardware failure then
Nov 25 16:13:10 <mickeyl> *sigh*
Nov 25 16:13:23 <balrog-kun> also CSQ: 12, 99 is pretty much at the border of useable signal quality i think
Nov 25 16:14:37 <ScaredyCat> Risto: if you are using the mokomakefile then : make clean-package-renderproto; make clean-package-xserver-kdrive
Nov 25 16:14:42 <mickeyl> hmm, so perhaps the antenna?
Nov 25 16:15:05 <ScaredyCat> not just a bad signal?
Nov 25 16:15:13 <mickeyl> no
Nov 25 16:15:21 <mickeyl> like i said i'm working here since long
Nov 25 16:15:25 <Stephmw> balrog-kun: get the impression it's swapping between base stations?
Nov 25 16:15:31 <mickeyl> and never have seen this
Nov 25 16:15:38 <PBeck> hdr: the mapdata is from openstreetmap
Nov 25 16:15:40 <mickeyl> hmm
Nov 25 16:15:46 <mickeyl> Stephmw: you mean multiple base stations, same cell id?
Nov 25 16:15:48 <PBeck> hdr: see http://www.openstreetmap.org
Nov 25 16:15:50 <ScaredyCat> no, but if power has failed to a tower, then you;d have issues
Nov 25 16:15:51 <hdr> PBeck: thx
Nov 25 16:15:57 <Stephmw> mickeyl: *nod*
Nov 25 16:16:00 <mickeyl> oh
Nov 25 16:16:20 <mickeyl> i didn't know that existed
Nov 25 16:16:24 <PBeck> hdr: help and create your city ;)
Nov 25 16:16:31 <mickeyl> i was under the impression, every base station corresponds to a cell
Nov 25 16:16:37 <mickeyl> there goes my gsm knowledge
Nov 25 16:16:38 <mickeyl> heh
Nov 25 16:16:38 <balrog-kun> i didn't know either
Nov 25 16:16:45 <balrog-kun> but it's reporting %CSQ: 99, 99 in between, which means no signal at all
Nov 25 16:16:50 <ScaredyCat> Stephmw: from the same providr?
Nov 25 16:16:58 <Stephmw> mickeyl: I can confirm this for sure tomorrow
Nov 25 16:17:09 <hdr> PBeck: as soon as i have the openmoko ;)
Nov 25 16:17:14 <Stephmw> ScaredyCat: it's a bit bizarre, agreed
Nov 25 16:17:26 <Stephmw> ScaredyCat: it could also be a software glitch on a single station
Nov 25 16:17:27 <balrog-kun> mickeyl: are you geographically in the same place as alway also?
Nov 25 16:17:32 <balrog-kun> :)
Nov 25 16:18:06 <Stephmw> ScaredyCat: god knows, the agent systems for managing slot allocation and connectivity don't just depend on signal strength
Nov 25 16:18:15 <PBeck> hdr: hehe i track my city since the last weeks ;)
Nov 25 16:18:16 <mickeyl> balrog-kun: yes. not moving.
Nov 25 16:18:28 <PBeck> hdr: one german city more in openstreetmap ;)
Nov 25 16:18:35 <ScaredyCat> strange shit ! 
Nov 25 16:18:35 <hdr> PBeck: how does it work?
Nov 25 16:18:55 <SpeedEvil> hdr: you basically take a GPS track you've made
Nov 25 16:19:00 <balrog-kun> PBeck: using the Neo?
Nov 25 16:19:09 <ScaredyCat> right
Nov 25 16:19:11 <borg_> there are just a dozen of streets from the city where i live in openstreetmap :(
Nov 25 16:19:13 <SpeedEvil> hdr: then upload it to the server, and trace over it using an editor
Nov 25 16:19:18 * ScaredyCat is off for tea and crumpets
Nov 25 16:19:31 <hdr> SpeedEvil: hm no gps-mouse :/
Nov 25 16:19:31 <SpeedEvil> hdr: or you trace from ariel imagery - yahoo allows this
Nov 25 16:19:43 * SpeedEvil uses his neo.
Nov 25 16:20:20 <PBeck> hdr: you collect with a gps reciever tracks and trace it with josm or potlach (online editor) and describe the streets 
Nov 25 16:20:23 <SpeedEvil> (for GPS/bluetooth GPS logging onbly)
Nov 25 16:20:41 <SpeedEvil> PBeck: you could do it with a theodolite if you're feeling retero.
Nov 25 16:20:57 <balrog-kun> has someone successfully contributed to openstreetmap using Neo?
Nov 25 16:21:01 <SpeedEvil> yes
Nov 25 16:21:05 <hdr> only gps-receiver, my dad has is in his car ;)
Nov 25 16:21:22 <balrog-kun> SpeedEvil: cool, i will need to try it too
Nov 25 16:21:29 <PBeck> balrog-kun: i havent a neo ;)
Nov 25 16:21:46 <Stephmw> mickeyl: do you have an areaid for your cellid?
Nov 25 16:21:46 <PBeck> SpeedEvil: theodolite?
Nov 25 16:21:54 <balrog-kun> i have no other GPS than the one in my neo
Nov 25 16:22:01 <Stephmw> mickeyl: turns out cellids are only unique within an area
Nov 25 16:22:09 <rxr> SpeedEvil: is there now something to parse the numbers from the GPS chip ?
Nov 25 16:22:25 <Stephmw> mickeyl: sooo... if your primary station's coverage shrinks, you could be entering another areaid quite easily
Nov 25 16:22:30 <balrog-kun> Stephmw: "000F","0343" this is mickeyl's
Nov 25 16:23:16 <Stephmw> balrog-kun: hmmm, yeah
Nov 25 16:23:22 <Stephmw> balrog-kun: so same op, same area
Nov 25 16:23:23 <SpeedEvil> rmoravcik: no - just the binary driver
Nov 25 16:24:06 <rxr> SpeedEvil: I thought the binary driver is not re-distributable?
Nov 25 16:24:13 <SpeedEvil> PBeck: you basically measure the angles between known reference points, and use lots of telescopic observations to determine a position.
Nov 25 16:24:24 <SpeedEvil> rxr: it's not
Nov 25 16:24:32 <PBeck> SpeedEvil: lol
Nov 25 16:24:39 * SpeedEvil wonders if there has been some changes.
Nov 25 16:24:46 <SpeedEvil> PBeck: it's what they used to use.
Nov 25 16:24:55 <PBeck> SpeedEvil: thats a brilliant idea :>
Nov 25 16:25:02 <PBeck> i only need the hardware :>
Nov 25 16:25:12 <PBeck> perhaps its easier with gps ;)
Nov 25 16:25:22 <rxr> SpeedEvil: can this binary blob be "found" somewhere ?
Nov 25 16:25:54 <borg_> rxr: if you post your email adress here, perhaps someone will accidantially send it to you
Nov 25 16:25:57 <borg_> :X
Nov 25 16:26:23 <balrog-kun> SpeedEvil: is there a guide to using Neo with openstreetmap?
Nov 25 16:26:48 <SpeedEvil> As to tracks: http://openstreetmap.org/traces/tag/fife
Nov 25 16:26:57 <balrog-kun> rxr: i saw it on someone's website downloadable
Nov 25 16:27:27 <borg_> SpeedEvil: look scap
Nov 25 16:27:46 * balrog-kun notices SpeedEvil is going in circles a lot
Nov 25 16:27:50 <SpeedEvil> http://wiki.openstreetmap.org/index.php/Beginners%27_Guide is the basic guide from OSM
Nov 25 16:28:02 <borg_> the mouse is to hard to see i guess
Nov 25 16:28:13 <SpeedEvil> Then http://www.mauve.plus.com/dogll is my trivial wrapper script
Nov 25 16:28:16 <borg_> and the alpha channel is flickering when moving :(
Nov 25 16:29:05 <balrog-kun> SpeedEvil: thanks
Nov 25 16:29:21 <SpeedEvil> borg_: scap?
Nov 25 16:29:37 <borg_> SpeedEvil: sorry, e meant ScaredyCat 
Nov 25 16:29:46 <borg_> s/e/i/
Nov 25 16:29:48 <apt> borg_ meant: SpiedEvil: sorry, e meant ScaredyCat 
Nov 25 16:29:52 <borg_> lol
Nov 25 16:29:52 <borg_> :)
Nov 25 16:30:11 <hdr> helping osm without gps-receiver is only possible via yahoo?
Nov 25 16:30:13 <borg_> ScaredyCat: look scap
Nov 25 16:30:13 <Stephmw> balrog-kun: what about frequency hopping? same cellid, but kaput on another band?
Nov 25 16:30:24 <SpeedEvil> (then you take that nmea trace and transform it with gpsbabel -i nmea -o gpx -f $1 -F $1.gpx
Nov 25 16:30:32 <SpeedEvil> and then upload it or do josm stuff
Nov 25 16:30:54 <SpeedEvil> hdr: pretty much - or naming streets existing in OSM that are not named
Nov 25 16:30:57 <rxr> borg_: rene <at> t2-project <dot> org <- for anyone who as a gps blob to spare .-)
Nov 25 16:31:28 <balrog-kun> Stephmw: i guess it's possible too, but this is black magic for me
Nov 25 16:32:36 <Stephmw> mickeyl: is there any way to tell what freq the gsm is using?
Nov 25 16:33:38 * SpeedEvil passes Stephmw a spectrum analyser
Nov 25 16:34:08 <Stephmw> SpeedEvil: give it to mickeyl, I still haven't got myself a neo :)
Nov 25 16:34:29 <Stephmw> SpeedEvil: just wondering if that info can be obtained from the phone
Nov 25 16:35:13 <mickeyl> 1800 here
Nov 25 16:35:13 <Stephmw> finally... does .de have a site like http://www.sitefinder.ofcom.org.uk/ that allows you to identify base stations?
Nov 25 16:35:22 <Risto> ScaredyCat: Thanks Very Much! It's working again :-) 
Nov 25 16:35:33 <Stephmw> mickeyl: when it flipflops, does that change?
Nov 25 16:35:49 <mickeyl> oh
Nov 25 16:35:53 <mickeyl> no idea, frankly
Nov 25 16:36:02 <mickeyl> i have no equipment to test that
Nov 25 16:36:22 <mickeyl> the good thing though is
Nov 25 16:36:28 <mickeyl> it doesn't happen on another device with a different SIM
Nov 25 16:36:50 <mickeyl> will cross-test, i'm afraid either my preproduction gta01bv4 is broken or the SIM
Nov 25 16:36:59 <Stephmw> that's my current theory: your near station has 1+ bands it can use... one of which is dead, the other which is weak
Nov 25 16:37:17 <SpeedEvil> Alternative test equipment.
Nov 25 16:37:20 <SpeedEvil> http://www.dealextreme.com/details.dx/sku.7059
Nov 25 16:37:23 <SpeedEvil> Cheep!
Nov 25 16:38:43 <balrog-kun> i think the FEM project's aim is to make a database like this, just worldwide
Nov 25 16:39:12 <Stephmw> interesting... that db is missing some picocells I know of
Nov 25 16:39:17 <mickeyl> http://pastebin.ca/796947
Nov 25 16:39:32 <mickeyl> other Neo (production model this time), same place, different SIM
Nov 25 16:39:41 <mickeyl> still oscillates between cells
Nov 25 16:39:41 <Stephmw> mickeyl: quite a few cells
Nov 25 16:39:45 <mickeyl> but doesn't unreg
Nov 25 16:40:19 <balrog-kun> i think it's normal forphones to oscillate between cells for no obvious reason
Nov 25 16:40:27 <mickeyl> maybe the antenna in the preproduction model has problems
Nov 25 16:40:27 <SpeedEvil> yeah
Nov 25 16:40:34 <SpeedEvil> especially if there is a plurality of cells
Nov 25 16:40:39 <mickeyl> should try an external one
Nov 25 16:46:48 <ScaredyCat> lol borg_
Nov 25 16:50:01 <mickeyl> If someone of you can clarify 1025, I'd be grateful
Nov 25 16:50:46 <ScaredyCat> !ombug 1025
Nov 25 16:50:47 <cdbot2> * * Bug 1025, Status: NEW, Created: Unknown
Nov 25 16:50:48 <cdbot2> * * mickey(AT)vanille-media.de: gsmd initialization questions
Nov 25 16:50:49 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1025
Nov 25 16:51:00 <cdbot2> * * OM Bug 1025 has been created by mickey(AT)vanille-media.de
Nov 25 16:51:01 <cdbot2> * * gsmd initialization questions
Nov 25 16:51:02 <cdbot2> * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1025
Nov 25 16:51:10 <borg_> ScaredyCat: i updated the ipk now to this http://scap.linuxtogo.org/files/1d892edca9bc93ba417fd8bc274aeff5.png
Nov 25 16:51:20 <borg_> it is too much flickering with alpha channel :\
Nov 25 16:53:23 <PBeck> borg_: hehe great ;)
Nov 25 16:53:35 <borg_> :)
Nov 25 16:57:15 <mickeyl> gsm is completely strange
Nov 25 16:57:26 <mickeyl> why is the signal event so completely undeterministic
Nov 25 16:57:38 <mickeyl> sometimes it fires every couple of seconds
Nov 25 16:57:44 <mickeyl> sometimes it's absent for minutes
Nov 25 16:59:17 <Writchie> mickeyl: IIRC you get events for multiple operators
Nov 25 16:59:52 <emdete> mickeyl: are you the mickeyl from oe? if so: i made a patch to get ctypes compiles for arm. are you interested?
Nov 25 17:02:35 <mickeyl> emdete: yes, that's me. yeah, pretty cool, please attach it to the oe bugzilla, i'll take care of it
Nov 25 17:12:19 <emdete> mickeyl: is there a ticket on ctypes already?
Nov 25 17:12:38 <mickeyl> i don't know offhand
Nov 25 17:13:03 <zecke|flu> emdete: ctypes the header file?
Nov 25 17:13:40 <emdete> zecke|flu: didn't understand that...
Nov 25 17:14:22 <balrog-kun> emdete: i.e. what do you refer to with ctypes
Nov 25 17:15:02 <emdete> ctypes is a module of standard 2.5 python... that's what i refering to
Nov 25 17:15:46 <balrog-kun> ah ok, in C it is a library of.. well, C types :)
Nov 25 17:16:33 <emdete> :D oups... yes, shure. sry, forget to mention python.
Nov 25 17:17:05 * mwester has set out to learn python, as painful as it seems to be...
Nov 25 17:18:34 <abraxa_> borg_: Might wanna contact http://wiki.openmoko.org/wiki/User:Nenadgrujicic to get some nice GUI images
Nov 25 17:18:40 <emdete> python painful? oh...
Nov 25 17:18:57 <tuukkah> perhaps it's not the python part but the learning part :-)
Nov 25 17:19:28 <mwester> :P  No, it's the throwback to the worst misfeature of FORTRAN that hurts about it.
Nov 25 17:19:47 <emdete> apropos fortran - does OM compile again?
Nov 25 17:20:04 <borg_> abraxa_: ok, i am sending a mail to the openmoko-apps list right now
Nov 25 17:20:21 <emdete> mwester: love it or hate it, there is nothing in between. but fortran is far from python...
Nov 25 17:20:53 <mwester> Sigh.  I don't know.  I think the change that broke the build was backed out of the gcc recipe, but not the gcc-cross recipe.  So it will only build, I think, if you have a working mpfr package installed on your distro.
Nov 25 17:21:27 <mwester> Python is far, far better than FORTRAN, but I have nightmares about column 6.
Nov 25 17:22:03 <emdete> mwester: oh, in python you may do anything you like with col 6 ;)
Nov 25 17:22:06 <tuukkah> so the pain of unlearning =)
Nov 25 17:22:38 <mwester> emdete:  No, not so -- you must carefully COUNT the spaces, so that you place your statement on the correct column, otherwise it will be part of the wrong block!
Nov 25 17:23:14 <emdete> mwester: this does my editor for me... but i have a similar problem with lua: it counts arrays from 1
Nov 25 17:23:18 <mwester> To be specific, column-sensitivity to determine nesting seems so, so wrong anymore!
Nov 25 17:23:37 <emdete> mwester: for me it seems so, so right ;)
Nov 25 17:23:38 <tuukkah> mwester, don't you indent all your code exactly in any language?
Nov 25 17:23:39 <mwester> I didn't know that about lua.  That's another problem.
Nov 25 17:24:42 <mwester> tuukkah:  I do -- usually.  When I'm debugging, however, I may wrap a block in an if statement, or remove an else clause, and I do not wish to adjust the indentation of everything just because I've added a debug conditional.
Nov 25 17:25:45 <tuukkah> mwester, in that sense, you need more of editor features: comment/uncomment block, indent/outdent block
Nov 25 17:26:10 <emdete> anyway - what is about build om and this mpfr package? i can't find that for debian...
Nov 25 17:26:13 <mwester> I'll get used to it, I suppose.  But habits die hard, and I'm accustomed to writing code and debugging it, and then AFTERWARDS going in and pretty-printing it to fix up all the tabs/spaces and indentations.
Nov 25 17:26:28 <emdete> is it libmpfr-dev?
Nov 25 17:26:58 <tuukkah> emdete, looks like it
Nov 25 17:27:33 <emdete> mickeyl: i don't find a ticket - shall i create one? 
Nov 25 17:28:13 <mwester> emdete:  Yes, probably libmpfr-dev.  It doesn't exist at all on Fedora.  Package "mpfr-native" is build by the build process, but something is wrong in the bitbake recipe, and gcc-cross can't find the one that OE builds.  Apparently it will "work" if it instead finds the one installed in your /usr/lib/ directory.  Although that too should fail some sanity test, I would think.
Nov 25 17:30:19 <mwester> Ah, let me correct myself -- I think that linking to your host library instead of the correct one will not be found by the sanity checker because gcc-cross is a native package, and is not packaged as an ipkg or part of the final image.
Nov 25 17:31:56 <emdete> hm, no that didn't work. there is another error now...
Nov 25 17:32:15 <emdete> in libgfortran...
Nov 25 17:32:36 <emdete> can i somehow easily kick this fortran-... out of the buile
Nov 25 17:33:06 <emdete> s/buile/build/
Nov 25 17:33:07 <apt> emdete meant: can i somehow easily kick this fortran-... out of the build
Nov 25 17:36:55 <mwester> emdete:  What package, specifically, is failing?
Nov 25 17:39:23 <CVirus> no community updates yet .. damn
Nov 25 17:40:48 <emdete> still gcc-cross-4.1.2-r11 in build/tmp/work/armv4t-angstrom-linux-gnueabi/gcc-cross-4.1.2-r11/gcc-4.1.2/libgfortran/intrinsics/selected_int_kind.f90:35
Nov 25 17:41:07 <mwester> Ok.  Hang on a few minutes.
Nov 25 17:56:35 <mickeyl> emdete: yes, please go ahead with the ticket.
Nov 25 17:58:31 <CVirus> youtube ... Http/1.1 Service Unavailable
Nov 25 17:58:32 <CVirus> hehe
Nov 25 17:59:04 <ScaredyCat> mmm
Nov 25 17:59:10 <ScaredyCat> you tube broke?
Nov 25 18:00:33 <IfChinsCouldKill> Borg?
Nov 25 18:00:45 <borg_> IfChinsCouldKill: ?
Nov 25 18:00:50 <IfChinsCouldKill> http://home.arcor.de/finishlast/zeugs/openmokochoice.png
Nov 25 18:00:58 <IfChinsCouldKill> How about this?
Nov 25 18:01:00 <borg_> NICE :D
Nov 25 18:01:11 <borg_> wow
Nov 25 18:01:33 <ScaredyCat> lol
Nov 25 18:01:45 <borg_> so nice
Nov 25 18:02:16 <IfChinsCouldKill> it
Nov 25 18:02:26 <IfChinsCouldKill> it's based on this here http://www.blogfrei.de/metalust/Freedom2.jpg
Nov 25 18:02:37 <borg_> i see
Nov 25 18:05:23 <IfChinsCouldKill> I made it into this north south drag things http://home.arcor.de/finishlast/zeugs/mokochoice.zip
Nov 25 18:05:37 <borg_> IfChinsCouldKill: i will make it in the ipk
Nov 25 18:05:38 <borg_> mom
Nov 25 18:06:22 <emdete> mickeyl: done (http://bugs.openembedded.org/show_bug.cgi?id=3368). how fast does this go into om now?
Nov 25 18:06:49 <mickeyl> can't predict
Nov 25 18:06:52 <mickeyl> depends on when i have time
Nov 25 18:07:21 <mickeyl> hopefully sometime next week
Nov 25 18:09:05 <emdete> mickeyl: if i can build again i probably can fix this issue i mention in the comment until then...
Nov 25 18:14:40 <borg_> IfChinsCouldKill: http://scap.linuxtogo.org/
Nov 25 18:14:46 <borg_> http://homepages.upb.de/olaf/dinge/openmoko/display_lock/neod_0.1.0+svnr3421-r6_neo1973.ipk
Nov 25 18:16:41 <IfChinsCouldKill> hehe nice
Nov 25 18:16:58 <IfChinsCouldKill> Borg: You should make a video an put it on youtube
Nov 25 18:17:07 <borg_> IfChinsCouldKill: i will try
Nov 25 18:17:25 <borg_> IfChinsCouldKill: the alpha channel if flickering a bit, i will try to fix this first
Nov 25 18:17:42 <IfChinsCouldKill> scaredycat can make an android logo and a trashcan.
Nov 25 18:17:48 <borg_> but i think thats a problem with the gtk drag and drap thing, so i cant do anything against it
Nov 25 18:17:51 <borg_> :)
Nov 25 18:18:42 <borg_> or am iphone and a blender
Nov 25 18:18:44 <borg_> 8)
Nov 25 18:19:30 <torpor> yo
Nov 25 18:20:40 <orzel> hello
Nov 25 18:28:50 <borg_> mickeyl: have you already tried the display lock?
Nov 25 18:29:05 <borg_> you have to use the new one with the images of IfChinsCouldKill when not
Nov 25 18:29:07 <borg_> it is so nice :)
Nov 25 18:35:59 <mickeyl> i have no time to build anything atm. working on gsm applet
Nov 25 18:36:06 <mickeyl> do you have a prebuilt image i can try on one of my neos?
Nov 25 18:36:24 <borg_> http://homepages.upb.de/olaf/dinge/openmoko/display_lock/neod_0.1.0+svnr3421-r6_neo1973.ipk
Nov 25 18:36:45 <mickeyl> images...
Nov 25 18:36:45 <borg_> just du ipkg -force-overwrite -install neod..
Nov 25 18:36:52 <borg_> no, sry ;)
Nov 25 18:37:03 <mickeyl> ah, k. will try when i'm finished with today's hacking sessino
Nov 25 18:37:40 <mickeyl> hmm
Nov 25 18:37:49 <mickeyl> struct gsmd_msg_hdr { is missing the union for the data packages
Nov 25 18:37:50 <mickeyl> *sigh*
Nov 25 18:38:03 <mickeyl> let's go add 'em then...
Nov 25 18:40:30 <demaledetti> borg_: it rocks, I love it, much needed!
Nov 25 18:40:48 <torpor> what is it?
Nov 25 18:40:55 <borg_> demaledetti: thanks :)
Nov 25 18:41:05 <mwester> emdete:  Sync up once again, and try the build -- it should rebuild gcc-cross anew.
Nov 25 18:41:07 <demaledetti> torpor: see the ipk link above
Nov 25 18:41:11 <torpor> what is it though?
Nov 25 18:41:16 <borg_> demaledetti: the neo is very much useable as a mp3 player
Nov 25 18:41:29 <demaledetti> torpor: and find the new screen locker under the power menu
Nov 25 18:41:32 <torpor> yup, i'm using mine as an mp3 player right now .. 
Nov 25 18:41:33 <borg_> but the lack of a display lock made it bad.. ;)
Nov 25 18:41:47 <borg_> torpor: then you should install the new neod :)
Nov 25 18:41:55 <demaledetti> torpor: with very cool graphics too
Nov 25 18:42:19 <torpor> when will this be in the man tree?  and are there screenshots?
Nov 25 18:42:20 <demaledetti> borg_: yeah it's useful even just for carrying the phone around in a pocket
Nov 25 18:42:47 <borg_> torpor: http://scap.linuxtogo.org/files/79bc6f5f1990dc39de783132ab633847.png
Nov 25 18:43:22 <torpor>  ah, i see .. 
Nov 25 18:43:24 <torpor> cool!
Nov 25 18:43:35 <demaledetti> totally
Nov 25 18:43:43 <borg_> dont know if it comes in mainline, i made an attachemend to the bugzilla and i am harassing mickeyl here in irc ;)
Nov 25 18:43:46 <borg_> i cant do more :D
Nov 25 18:44:40 <borg_> torpor: http://scap.linuxtogo.org/files/1d892edca9bc93ba417fd8bc274aeff5.png
Nov 25 18:44:42 <torpor> well i'll wait until it goes into the main tree though .. i'm trying to make a kickass bootable SD to put online for people to check out .. 
Nov 25 18:44:44 <borg_> you can use this one too :)
Nov 25 18:44:55 <torpor> nice!
Nov 25 18:45:07 <borg_> :P
Nov 25 18:49:18 <mickeyl> hehe
Nov 25 18:49:26 <mickeyl> ok
Nov 25 18:49:33 <mickeyl> let me check
Nov 25 18:49:34 <zecke|flu> LaF0rge: got a second? Should I get cash for india here or there? Should I get a credit card or will maestro (EC-Karte) do?
Nov 25 18:51:41 <mickeyl> hahahaha
Nov 25 18:51:45 <mickeyl> ok, guys, i love that :D
Nov 25 18:51:48 <mickeyl> it's so cool
Nov 25 18:52:11 <mickeyl> we just need a different background picture
Nov 25 18:52:19 <mickeyl> this one is a bit... uhm... too che guevara
Nov 25 18:52:20 <torpor>  /clear
Nov 25 18:52:31 <torpor> how doe the images work?
Nov 25 18:52:46 <torpor> are they separate layers, or is it just one image that is utomatically skinned?
Nov 25 18:52:50 <torpor> +a
Nov 25 18:53:01 <balrog-kun> it's great :)
Nov 25 18:54:45 <borg_> is there something in oe with which i can make some sort of screencast of the openmoko display?
Nov 25 18:54:58 <SpeedEvil> X?
Nov 25 18:55:04 <SpeedEvil> DISPALY=?
Nov 25 18:56:54 <matsimoto> are there any events that could wake up an application when a SMS or a phone call comes in ? ... what I mean is wake up from standby.
Nov 25 18:57:11 <borg_> torpor: http://homepages.upb.de/olaf/dinge/openmoko/display_lock/openmoko-display-lock-readme.txt
Nov 25 18:57:27 <borg_> there you can read what pictures are needed
Nov 25 18:57:29 <borg_> and in which size
Nov 25 18:57:40 <borg_> you can just put them in /usr/share/neod
Nov 25 18:58:38 <IfChinsCouldKill> lol too much che guevara, oh boy:D
Nov 25 18:58:54 <borg_> IfChinsCouldKill: hehe ;)
Nov 25 18:58:56 <mickeyl> could someone come up with a less offensive one?
Nov 25 18:59:05 <mickeyl> this patch would make it into the snapshot then :)
Nov 25 18:59:50 <IfChinsCouldKill> Google forced me to make it this way
Nov 25 19:00:17 <borg_> i have no skills in making graphics ;)
Nov 25 19:00:35 <borg_> red and green rectangles was the best i could do ;)
Nov 25 19:02:40 <SpeedEvil> I would suggest: http://en.wikipedia.org/wiki/Image:Kathmandu-05.JPG
Nov 25 19:02:42 <SpeedEvil> as an image
Nov 25 19:03:20 <borg_> mh.. there is nothing where somthing could be dragged into ;)
Nov 25 19:04:37 <Vegar> regarding that lock..
Nov 25 19:04:53 <Vegar> would it be possible to limit the distance the "thing" can be dragged?
Nov 25 19:05:18 <Vegar> to prevent accidental unlocking
Nov 25 19:05:40 <borg_> how do you mean that?
Nov 25 19:05:48 <borg_> if you stop dragging the thing falls back down
Nov 25 19:05:51 <Vegar> yeah
Nov 25 19:05:55 <Vegar> but what if you don't?
Nov 25 19:06:07 <SpeedEvil> Make it a little maze!
Nov 25 19:06:20 <borg_> i dont think that is needed
Nov 25 19:06:36 <borg_> at least there was no accidential unlocking in my jogging-test ;)
Nov 25 19:07:48 <Vegar> borg_: you should do a dance test too
Nov 25 19:07:58 <SpeedEvil> And put it on youtube.
Nov 25 19:08:16 <Aria> Hee
Nov 25 19:08:39 <Vegar> of course
Nov 25 19:09:23 <abraxa_> SpeedEvil: http://lists.openmoko.org/pipermail/openmoko-devel/2007-October/001607.html
Nov 25 19:09:27 <abraxa_> That's more like a maze ;3
Nov 25 19:11:44 <SpeedEvil> Ideally I'd like an x-stroke like recogniser to do it and a passphrase option - but...
Nov 25 19:11:46 <mickeyl> hmm
Nov 25 19:11:59 <SpeedEvil> that sounds logical
Nov 25 19:22:27 <matsimoto> Th wiki says: "The GSM daemon configures the GSM Modem so as to wake the CPU when events happen." is this true ?
Nov 25 19:22:41 <SpeedEvil> matsimoto: probably not
Nov 25 19:22:51 <SpeedEvil> matsimoto: much of the wiki is merely hopes
Nov 25 19:23:07 <SpeedEvil> power managment at the moment is of the 'point and laugh' class.
Nov 25 19:23:24 <Writchie> press and press PM
Nov 25 19:23:29 <Writchie> press and pray PM
Nov 25 19:23:52 <matsimoto> Ok that is an explanation why I cannot find anything in gsmd ... is there another source where I can find something on CPU wakeup ?
Nov 25 19:24:10 <SpeedEvil> not really AIUI
Nov 25 19:24:18 <SpeedEvil> the datasheet of the SOC and PMU chip
Nov 25 19:24:25 <Writchie> seriously, eventually the gsmd or its successor will stop all all unsolicited AT messages except those of interest for wakeup. 
Nov 25 19:24:50 <Writchie> PM will need some coordination among user space and kernel drivers. 
Nov 25 19:27:36 <matsimoto> Ok, I am more a software engineer than an electronics engineer :(, does anyone has some hints were to begin as a sw-engineer when trying to wakeup on GSM events ?
Nov 25 19:28:25 * SpeedEvil pases matsimoto Brave Sir Robins minstrels.
Nov 25 19:28:30 <SpeedEvil> Umm.
Nov 25 19:28:48 <Writchie> unfortunately, I think you will have to wait until the PM is more full debugged. 
Nov 25 19:29:31 <Writchie> you can probably count on their eventually being a system dbus message that will inform applications of a wakeup. 
Nov 25 19:30:30 <Writchie> however, there may not be any reason for an application to care. It should just respond to normal events. 
Nov 25 19:30:52 <Writchie> If every application tries to get involved with PM in a free for all it will never work. 
Nov 25 19:31:33 <Writchie> applications should generally let the daemons and the kernel take care of it. 
Nov 25 19:31:50 <matsimoto> allright, but is somebody working on this or is this something everybody is waiting for ?
Nov 25 19:32:15 <SpeedEvil> matsimoto: We haven't been told. To expand.
Nov 25 19:32:49 <borg_> matsimoto: in my opinion thats the biggest problem here, nobody knows what is going on :)
Nov 25 19:32:52 <Writchie> kernel work seems to be in a bit of a limbo. Unfortunately, no-one outside of OM can help much because it involves GTA02 hardware changes not yet disclosed. 
Nov 25 19:33:37 <SpeedEvil> Dirstly. for GTA01 at least - it should in principle be possible to wake it from suspend-RAM.
Nov 25 19:34:06 <matsimoto> ok, but I think they need applications that cooperate in a PM environment before they can go to market with GTA02.
Nov 25 19:34:24 <SpeedEvil> you basically look for which pin the GSM modem is wired to, then work out how to configure it as a 'wake from' interrupt before you go to sleep/.
Nov 25 19:34:33 <Writchie> matsimoto: yes of course
Nov 25 19:34:58 <Writchie> a couple of hours of battery life makes a useless MOBILE phone
Nov 25 19:36:11 <Writchie> unfortunately, uboot - kernel - and hardware are all related and there are several issues due to differing hardware on GTA01 and GTA02
Nov 25 19:37:00 <matsimoto> In my point of view a lot of other mobiles enter the market: iPhone, HTC Touch and you have to love open phones to pay a price for it where you can get a HTC Touch. Though, I think time is running o..
Nov 25 19:37:16 <Writchie> the PM chips for example are different and there is no published info on the PM chip used in GTA02 not to mention the wiring. 
Nov 25 19:37:30 <edistar> matsimoto: yeah, OM needs to speed up a lot
Nov 25 19:37:41 <SpeedEvil> That's _really_ sucky - I diddn't realise that.
Nov 25 19:37:43 <Writchie> matsimoto: time never runs out and the NEO is not really in the market. 
Nov 25 19:37:49 <SpeedEvil> I hope that some docs will be provided.
Nov 25 19:38:20 <Writchie> the NEO is and will remain a developer prototype, at least for several more months. 
Nov 25 19:38:34 <matsimoto> Writchie: yep, you are right what I mean is that there coming other phones with SDKs that will occupy market the NEO would fit perfectly.
Nov 25 19:38:52 <Writchie> but if it takes until june of next year - the world will not end. 
Nov 25 19:39:07 <SpeedEvil> athlon XP2500
Nov 25 19:39:10 <SpeedEvil> oops
Nov 25 19:39:19 <_buz> if google were to in fact opensource all of android, june is way too late
Nov 25 19:39:22 <borg_> matsimoto: you were the guy that wanted an interface to neod, right?
Nov 25 19:39:34 <matsimoto> borg_: yes
Nov 25 19:39:42 <borg_> ok
Nov 25 19:39:45 <_buz> (i have serious doubts that google would do that but who knows)
Nov 25 19:40:15 <borg_> i am thinking of making neod in a real daemon and outsourcing the aux menu, the power menu and the display locker
Nov 25 19:40:36 <borg_> and perhaps rename it in "phoned" or something
Nov 25 19:40:45 <MetaMorfoziS> Is there any news about the (central?)european store of neo?
Nov 25 19:40:49 <borg_> because it allready runs on more then the neo
Nov 25 19:41:13 <borg_> matsimoto: whould it be a good idea to send all events over dbus?
Nov 25 19:41:47 <matsimoto> borg_: yes, I think so. I am not very expereinced with dbus, but it sounds like right tool for the job
Nov 25 19:42:41 <Writchie> borg_: all of the system level thingy's should probably be on dbus. 
Nov 25 19:43:05 <matsimoto> Writchie: I hope you are righ. The other phones will have an advantage over the NEO. But of course I will buy one ;) even if it is Octobre.
Nov 25 19:43:45 <Writchie> the main problem is that they should all probably be defined and reveiwed before being implemented (i.e. they should be designed) IMNSO
Nov 25 19:43:49 <borg_> so, i just get all the events (aux, power button pressend, earphones inserted, 10 seconds no input from user, ...) and put them into dbus
Nov 25 19:43:55 <anrp> what?
Nov 25 19:43:57 <anrp> designed?!
Nov 25 19:44:03 <anrp> heretic! get 'im!
Nov 25 19:44:05 <borg_> and make the aux menu and so an normal apps using this
Nov 25 19:44:07 <borg_> right?
Nov 25 19:44:15 <Writchie> i know - its not the FOSS way ;)
Nov 25 19:44:45 <Writchie> which is why FOSS is better but takes longer (and sometimes infinitely longer)
Nov 25 19:45:09 <anrp> is it still better if it takes an infinite amount of time, though?
Nov 25 19:45:16 * anrp eyes the calculus with suspicion
Nov 25 19:45:23 <Writchie> yes - if you live long enuf
Nov 25 19:45:24 * flexd divides time by zero
Nov 25 19:45:50 <borg_> i think that sounds doable for me ;)
Nov 25 19:46:01 <borg_> i will start programming a neod replacement then
Nov 25 19:46:27 <anrp> good luck D:
Nov 25 19:46:44 <edistar> what does neod do?
Nov 25 19:46:44 <Writchie> borg_: have you looked at how BlueZ does it?
Nov 25 19:47:38 <borg_> edistar: watching for the button events of the phone and starting the aux/power menu and so on
Nov 25 19:47:41 <borg_> Writchie: no
Nov 25 19:47:53 <edistar> borg_: ah
Nov 25 19:48:17 <borg_> but there is no way for other programms to get these events atm
Nov 25 19:48:29 <Writchie> although not terribly well documented, they use d-bus in a very powerful and flexible way. 
Nov 25 19:48:37 <edistar> borg_: why not?
Nov 25 19:49:01 <mwester> emdete:  Is it building correctly now?
Nov 25 19:49:09 <borg_> Writchie: do you have a little bit more specific place where i should look in bluez? ;)
Nov 25 19:49:10 <Writchie> borg_: you generally want to permit applications to subscribe to whatever system events they are interested in. 
Nov 25 19:49:24 <Writchie> borg_: wait one
Nov 25 19:49:40 <borg_> edistar: because it is not implemented ;)
Nov 25 19:50:18 <borg_> edistar: i mean, everyone can read the events from /dev/input/event%d, but....
Nov 25 19:50:18 <Writchie> borg_: http://wiki.bluez.org/
Nov 25 19:50:45 <borg_> Writchie: i will try to make some sort of document in the wiki first
Nov 25 19:51:38 <borg_> but it sounds pretty much doable, there are not that many events from neod
Nov 25 19:51:47 <Writchie> borg_: the problem with what I call the application free for all is that in a mobile platform you need some daemon in charge of the platform as a whole due to PM and mobility issues. 
Nov 25 19:52:05 <SpeedEvil> and usability
Nov 25 19:52:17 <matsimoto> borg_: dont forget your screenlocked event
Nov 25 19:52:24 <borg_> matsimoto: yes ;)
Nov 25 19:52:29 <Writchie> so you really need to define the "platform" and not just treat it as a desktop
Nov 25 19:52:29 <SpeedEvil> the star-map program doesn't get to pop up if a call is ringing as venus is about to transit the sun.
Nov 25 19:52:53 <anrp> i'd say that depends on the caller
Nov 25 19:52:56 <Writchie> SpeedEvil: exactly - or do GPRS if bank account < $100,000
Nov 25 19:52:56 <anrp> ;)
Nov 25 19:53:08 <matsimoto> Writchie: does this mean that the GSM Events should be handled from the phoned too
Nov 25 19:53:20 <SpeedEvil> andred_: true of course.
Nov 25 19:53:40 * anrp pokes SpeedEvil with tab completion
Nov 25 19:53:51 <Writchie> matsimoto: IMHO yes. 
Nov 25 19:54:02 <Writchie> with some exceptions
Nov 25 19:54:29 <borg_> i will start with the button and the timeout events from neod
Nov 25 19:54:31 <borg_> then we can see
Nov 25 19:55:07 <Writchie> it may even be necessary at some point to have some interactions with the linux scheduler. 
Nov 25 19:56:35 <Writchie> ATM when userspace is resumed all process become runnable. 
Nov 25 19:56:37 <matsimoto> Writchie: Do you think everything should be put in one daemon or should there something like a daemon infra structure connected with each other ?
Nov 25 19:57:00 <SpeedEvil> Writchie: initrd
Nov 25 19:57:18 <Writchie> maybe - 
Nov 25 19:57:26 <SpeedEvil> or something simila
Nov 25 19:57:27 <SpeedEvil> r
Nov 25 19:57:31 <Writchie> SpeedEvil: it could become a runlevel thingy
Nov 25 19:57:55 <Writchie> runlevel 5 is burn baby burn 
Nov 25 19:58:11 <Writchie> runlevel 2 is conserve power as best possible.
Nov 25 19:58:24 <Aria> :w
Nov 25 19:58:27 <Writchie> runlevel 1 stay alive until battery is close to dead
Nov 25 19:58:27 <borg_> sounds interesting
Nov 25 19:58:44 <borg_> but you have to do a lot of changing in a lot of things for that ;)
Nov 25 19:59:22 <edistar> but it sounds simple..
Nov 25 19:59:43 <edistar> the concept..
Nov 25 20:00:18 <Writchie> you could have runlevel 4 where the gui runs only with power conserving apps. 
Nov 25 20:00:59 <CIA-14> 03mickey * r3504 10/trunk/src/target/gsm/include/gsmd/usock.h: 
Nov 25 20:00:59 <CIA-14> gsmd: start with gsmd_msg_auxdata union
Nov 25 20:00:59 <CIA-14> Using a union here (similar to gsmd_evt_auxdata) allows for compiler checks and
Nov 25 20:00:59 <CIA-14> is much less prone to fail than using pointer arithmetics.
Nov 25 20:03:44 <CIA-14> 03mickey * r3505 10/trunk/src/target/OM-2007.2/libraries/libmokogsmd2/ (3 files in 2 dirs): 
Nov 25 20:03:44 <CIA-14> libmokogsmd2:
Nov 25 20:03:44 <CIA-14>  Register network message handler with libgsmd to get
Nov 25 20:03:44 <CIA-14>  answers for trigger messages.
Nov 25 20:03:44 <CIA-14>  Keep trying to connect to gsmd, if not available on startup.
Nov 25 20:03:45 <CIA-14>  Add new g_signal for incoming operator name
Nov 25 20:05:11 <CIA-14> 03mickey * r3506 10/trunk/src/target/OM-2007.2/panel-plugins/openmoko-panel-gsm/ (ChangeLog src/openmoko-panel-gsm.c): 
Nov 25 20:05:11 <CIA-14> openmoko-panel-gsm:
Nov 25 20:05:11 <CIA-14>  * Query operator name on network registration event.
Nov 25 20:05:11 <CIA-14>  * Show operator name in notification.
Nov 25 20:25:32 <borg_> should it be possible for other applications to tell phoned to lower the brigthness of the backlight e.g.?
Nov 25 20:25:48 <edistar> borg_: yes, of course..
Nov 25 20:26:11 <edistar> borg_: if you have an app that doesn't need full backlight all the time.. how would OM know that
Nov 25 20:26:13 <hdr> echo 8 > /sys/backlight/brightness ;)
Nov 25 20:26:37 <borg_> hdr: yes, that is how it is done at the moment, thats bad ;)
Nov 25 20:26:45 <edistar> borg_: why is that bad?
Nov 25 20:27:02 <hdr> it's the same way on my laptop ;)
Nov 25 20:27:09 <borg_> edistar: because you have to change every application if this path changes
Nov 25 20:27:10 <balrog-kun> edistar: it's linux specific, non-portable
Nov 25 20:27:13 <borg_> or the distribution changes
Nov 25 20:27:14 <borg_> or....
Nov 25 20:27:34 <edistar> balrog-kun: ah.. I didn't think that far..
Nov 25 20:27:44 <balrog-kun> but i don't know if backlight belongs in phoned :)
Nov 25 20:27:52 <balrog-kun> i guess not all phones need to have an LCD
Nov 25 20:28:16 <borg_> balrog-kun: then nothing happens in this situation.. ;)
Nov 25 20:28:58 <balrog-kun> yeah, or an error is returned
Nov 25 20:29:20 <borg_> what i meant was, should the other applications really use dbus or should there be some sort of libphone wher some simple functions can be used?
Nov 25 20:31:55 <pavelm> Hello, everyone.
Nov 25 20:32:20 <aevin> hi pavelm 
Nov 25 20:32:54 <pavelm> ...planet.openmoko says power management is usable now ("lasts a day")
Nov 25 20:33:07 <borg_> oh really?
Nov 25 20:33:07 <borg_> cool
Nov 25 20:33:35 <pavelm> ...phase0 has broken hw, can't be repaired.
Nov 25 20:33:52 <pavelm> ...phase 1 has broken hw, but I was told difference is very small -- one transistor?
Nov 25 20:34:16 <pavelm> ...is there "how to fix phase1" document available somewhere?
Nov 25 20:34:52 <CM> pavelm: Wasn't the case that you can remove one transistor to fix that? I think roh talked hachi through it
Nov 25 20:35:29 <CM> pavelm: http://people.openmoko.org/roh/Gta01b_v4_front_Q1001.jpg
Nov 25 20:35:53 <CM> If you have one of the red boards that is.. I think
Nov 25 20:36:30 <balrog-kun> there's only a couple of billions of transistors on this picture, pick the right one and remove it
Nov 25 20:37:27 <IfChinsCouldKill> Borg: a less offensive one... http://home.arcor.de/finishlast/zeugs/mokopuzzle.png
Nov 25 20:37:27 <pavelm> aha, the picture even has green circle on it, good.
Nov 25 20:38:17 <pavelm> borg: can you repost the offensive one? :-)
Nov 25 20:38:43 <borg_> pavelm: http://scap.linuxtogo.org/files/79bc6f5f1990dc39de783132ab633847.png
Nov 25 20:38:45 <borg_> this one? ;)
Nov 25 20:38:59 <IfChinsCouldKill> http://home.arcor.de/finishlast/zeugs/mokopuzzle.zip north south kid thingy
Nov 25 20:39:05 <IfChinsCouldKill> mid even
Nov 25 20:39:17 <borg_> pavelm: http://homepages.upb.de/olaf/dinge/openmoko/display_lock/neod_0.1.0+svnr3421-r6_neo1973.ipk
Nov 25 20:39:43 <borg_> IfChinsCouldKill: cool too :)
Nov 25 20:40:27 <borg_> IfChinsCouldKill: can you put a drag.png in it too?
Nov 25 20:40:47 <pavelm> borg: that one is actually pretty cool.
Nov 25 20:40:55 <borg_> pavelm: yes, it is :)
Nov 25 20:41:18 <balrog-kun> what do the monkeys do there? :)
Nov 25 20:42:09 <IfChinsCouldKill> updated the zip file with the drag thingie
Nov 25 20:42:14 <borg_> ok
Nov 25 20:42:27 <IfChinsCouldKill> balrog-kun: was running out of freedom words ;)
Nov 25 20:43:42 <emdete> mwester: still building - i did a make clobber
Nov 25 20:44:13 <mwester> :(
Nov 25 20:44:28 <borg_> mickey|sofa: are you still available? ;)
Nov 25 20:44:31 <balrog-kun> heh
Nov 25 20:44:38 <emdete> mwester: sry, this is a slow machine here. is generating locales before or after that point?
Nov 25 20:44:39 <borg_> whats about this design? http://homepages.upb.de/olaf/dinge/openmoko/display_lock/moko_puzzle/mokopuzzle.png
Nov 25 20:45:15 <mwester> emdete:  Dunno, but I think it's after.  Ok, I guess it is fixed.  For now. 
Nov 25 20:45:52 <_buz> how about a simple svg or maybe even png theming ability for it?
Nov 25 20:46:37 <emdete> mwester: thnx alot...
Nov 25 20:49:11 <Writchie> pavelm: where did you see that GTA01 phase 1 hardware requires hardware fix for PM?
Nov 25 20:50:19 <CM> Writchie: I'm not sure, but I think it's pre-series phones
Nov 25 20:51:56 <balrog-kun> what does the fix fix?
Nov 25 20:52:53 <pavelm> writchie: heard it on irc. It is not in wiki :-(
Nov 25 20:53:02 <Writchie> CM: pre as in prior to GTA01Bv4
Nov 25 20:53:17 <weez> whats the easiest way to see what rev your hardware is?
Nov 25 20:53:45 <Writchie> all purchased hardware is GTA01Bv4 AFAIK
Nov 25 20:53:55 <weez> ahh
Nov 25 20:54:20 <weez> nothing in my chat logs...
Nov 25 20:54:23 <IfChinsCouldKill> Borg: I edited the pictures, took away some line on some puzzle tiles, just download the zip file again...
Nov 25 20:54:39 <weez> i had a couple of months logged
Nov 25 20:54:46 <weez> but i lost them in a tragic lvm resize accident
Nov 25 20:54:51 <Writchie> and if this requires hardware changes to PM to work - that would be very bad. 
Nov 25 20:54:56 <weez> now i only have since Nov 20
Nov 25 20:55:53 <borg_> ok
Nov 25 20:56:07 <pavelm> try searching for 'transistor'
Nov 25 20:56:37 <weez> nothing in the last 5 days
Nov 25 20:56:49 <weez> just what came up a few minutes ago
Nov 25 20:57:19 <weez> FreeNode-#openmoko.log:Nov 25 14:58:41 <pavelm> try searching for 'transistor'
Nov 25 20:57:36 <weez> that worked ;-)
Nov 25 20:58:01 <pavelm> weez: that was more like 3 months ago.
Nov 25 20:58:06 <weez> damn
Nov 25 20:58:16 <weez> damn LVM! DAMN YOU
Nov 25 20:58:21 <roh> hm?
Nov 25 20:58:31 <Aria> Why LVM?
Nov 25 20:58:37 <Writchie> well the transistor in the green circle appears on the gta01bv4 photo.
Nov 25 20:58:50 <borg_> IfChinsCouldKill: on the phone this one looks even more nice
Nov 25 20:58:54 <borg_> mhmh mickey|sofa ?
Nov 25 20:58:57 <Writchie> was this maybe a discussion on the private #openmoko-dev group?
Nov 25 20:58:57 <weez> i renamed a snapshot source
Nov 25 20:59:10 <weez> and everything (the source) and the snapshot went bye bye
Nov 25 20:59:13 <borg_> and it makes fun :)
Nov 25 20:59:25 <weez> with all my xchat logs
Nov 25 20:59:32 <weez> it was the only thing in there i cared about..
Nov 25 20:59:35 <weez> sniff..
Nov 25 21:00:39 <IfChinsCouldKill> I like the che guevera one ;)
Nov 25 21:00:47 <borg_> yes i like it too
Nov 25 21:00:49 * edistar hands weez a hanky
Nov 25 21:00:52 <borg_> i will use the che guevera one ;)
Nov 25 21:00:59 <CM> borg_: Don't poke him too much, it's sunday evening
Nov 25 21:00:59 <demaledetti> +1 for the che guevara one
Nov 25 21:01:01 <weez> edistar: thanks
Nov 25 21:01:02 <borg_> but i think mickey|sofa can put the puzzle one in mainline
Nov 25 21:01:02 <edistar> borg_: me too ;)
Nov 25 21:01:02 <pavelm> that one transistor should be difference between phase 1 sent to developers and phase 1 sold
Nov 25 21:01:18 <CM> pavelm: That's what I figured too
Nov 25 21:01:33 <borg_> CM: ok ;)
Nov 25 21:01:41 <CM> pavelm: Just heard roh and hachi discuss it, and hachi removed his
Nov 25 21:01:57 <Writchie> CM: pavelm: ok - then it means the GTA01B_v4 photo on the wiki is wrong. 
Nov 25 21:02:44 <pavelm> CM: Ok, I guess I should create wiki page about it
Nov 25 21:03:13 <Writchie> if it only affects give away phones - it might confuse more than help. 
Nov 25 21:03:30 <roh> Writchie which one
Nov 25 21:03:44 <CM> roh: Ah, good. Can you help pavelm along with that transistor?
Nov 25 21:03:49 <roh> Writchie there are red pcb and green pcb GTA01B_v4
Nov 25 21:03:51 <CM> The one hachi removed?
Nov 25 21:03:58 <Writchie> http://wiki.openmoko.org/wiki/Image:Gta01b_v4_front.jpg
Nov 25 21:04:11 <roh> the ones sold for money should all be green and never had q1001 placed, thus no bug.
Nov 25 21:04:13 * hozer has not been around for awhile... has anyone here looked at google's android stuff?
Nov 25 21:04:17 <Writchie> the red one shows the transiter in question. 
Nov 25 21:04:30 <CM> hozer: Look in the #android channel.. ;)
Nov 25 21:04:33 <Writchie> roh: good - 
Nov 25 21:05:30 <roh> the red pcb ones had this transistor placed afaik and can be fixed by removing it. (but its a nontrivial task, requiring experience and proper tools)
Nov 25 21:05:55 <Writchie> roh: sure - and removal of the can lids
Nov 25 21:06:24 <pavelm> I'm adding this to wiki:
Nov 25 21:06:25 <pavelm> ==== Errata =====
Nov 25 21:06:25 <pavelm> There were two versions of GTA01Bv4: some were given away (red mainboard), and some were sold (green mainboard). The given-away have [[one transistor too many]].
Nov 25 21:06:26 <roh> but since the bug is no showstopper to develop software with it it was only fixed for MP.
Nov 25 21:07:10 <Writchie> roh: am I correct in assuming the only issues on GTA01b4 at the moment are gsm firmware related?
Nov 25 21:07:18 <balrog-kun> +1 for the mice (and more importantly cheese)
Nov 25 21:07:19 <roh> meaning not swapping and reworking all giveaway-phones.
Nov 25 21:07:41 <roh> Writchie speaking from a hardware-errata point of view, yes.
Nov 25 21:07:55 <Writchie> roh: yes - hardware/firmware
Nov 25 21:08:14 <roh> Writchie i must say i am quite impressed that we did not find any new errors coming from the hardware
Nov 25 21:08:34 <balrog-kun> roh: what was the bug exactly?
Nov 25 21:08:37 <Writchie> roh: so far (fingers crossed)
Nov 25 21:08:58 * balrog-kun just looked and his pre-MP phone's pcb is red :(
Nov 25 21:09:08 <Writchie> roh: until we see full PM suspend/resume working its still an unknown. 
Nov 25 21:09:31 <Writchie> roh: OLPC kinds needed hardware changes at last minute over PM IIRC
Nov 25 21:09:47 <Writchie> s/kinds/guys/
Nov 25 21:09:47 <apt> Writchie meant: roh: OLPC guys needed hardware changes at last minute over PM IIRC
Nov 25 21:09:49 <pavelm> roh: http://wiki.openmoko.org/wiki/One_transistor_too_many ...can you check that it is basically correct?
Nov 25 21:09:49 <roh> balrog-kun the transistor pulls a 'reset' line on the gsm pmu. this pin on the pmu is not debounced, the other one for power (same wiring, other transistor) is debounced.
Nov 25 21:10:29 <balrog-kun> roh: meaning that gsm would be constantly in reset?
Nov 25 21:10:51 <Writchie> roh: the gsm modem is not responding to the mux command to go into low power state. Does it does this automatically on no activity?
Nov 25 21:11:05 <roh> when the pmu is shut down for suspend-to-ram the samsung somehow generates a hard, short spike on all gpio, which would, due to no debouncing reset the gsm, which wants to send data on the serial and wakes up the samsmung -> loop.
Nov 25 21:11:35 <balrog-kun> roh: ah okay
Nov 25 21:11:52 <roh> balrog-kun not constantly, but its like a oscillation through 2 subsystems (gsm and pmu) and the soc.
Nov 25 21:12:00 <pavelm> roh: Can I quote you in the wiki?
Nov 25 21:12:03 <balrog-kun> hmm but i'm able to suspent/resume the phone corrently, with modem on or off
Nov 25 21:13:13 <roh> since ti says the reset is only for testing, not for regular use, we removed q1001 from the placement map. just the name was not bumped to gta01bv5 since that would require changes to the gerber (labeling) and just not placing the transistor does not. its basically a layout-revision
Nov 25 21:13:21 <Writchie> balrog-kun: what current is being drawn in suspend?
Nov 25 21:13:44 <roh> pavelm you can.
Nov 25 21:13:50 <pavelm> roh: thanks.
Nov 25 21:14:09 <roh> pavelm i think that page needs a big, fat warning, only to do this if you know what youre doing etc. yadda,yadda
Nov 25 21:14:09 <balrog-kun> Writchie: judging from the battery life time improvement.. well there's not improvement
Nov 25 21:14:34 <roh> i would not have the tools for that at home. atleast not without destroying the shielding on that corner
Nov 25 21:14:35 <Writchie> gsm idle should draw just a few miliamps, 3 or 4
Nov 25 21:14:46 <roh> (and i solder some pretty sick stuff by hand)
Nov 25 21:14:52 <balrog-kun> i have no precise measurements, but i decided it wasn't worth suspending this device between copmiling and uploading new binaries for test, like it is done on other arm devices
Nov 25 21:15:34 <Writchie> balrog-kun: which kernel are you using for suspend/resume testing
Nov 25 21:15:50 <Writchie> 2.6.23?
Nov 25 21:16:05 <borg_> http://wiki.openmoko.org/wiki/Display_Locker
Nov 25 21:16:25 <roh> the 2.6.22.5 or so which was what we called stable till recently does fine with suspend to ram... are there issues i do not know about?
Nov 25 21:16:48 * hozer is so far not impressed by android
Nov 25 21:17:16 <balrog-kun> Writchie: i think we were at 2.6.19 when i decided that
Nov 25 21:17:20 <Writchie> roh: i've seen some issues on 2.6.23 PM serial on i386. 
Nov 25 21:17:41 <ScaredyCat> awwwno little mousey borg_ :(
Nov 25 21:17:51 <hozer> it reminds me of the effort for the first PPC40x kernel port... a huge pile of code was developed in a vacuum from the community and then dumped out there to the public assuming that the kernel people would just take it
Nov 25 21:17:58 <roh> hozer well.. google has another point of view about the word 'open' ;)
Nov 25 21:18:33 <Writchie> roh: is there a known stable kernel to use at this point - i've been assuming 2.6.23 will be coming shortly for GTA02
Nov 25 21:18:41 <torpor> the puzzle pic is awesome.
Nov 25 21:18:48 <IfChinsCouldKill> ScaredyCat: your mouse will be included as an easteregg ;)
Nov 25 21:18:58 <edistar> torpor: do you have a link?
Nov 25 21:18:58 <ScaredyCat> hhehehe
Nov 25 21:19:08 <hozer> roh: I suppose it will lead to a lot of phones on the market that could be hacked to run openmoko ;)
Nov 25 21:19:15 <Writchie> roh: the advantage of 2.6.23 is its the stock fedora 8 kernel
Nov 25 21:19:20 <edistar> hozer: good luck finding the drivers
Nov 25 21:19:21 <roh> Writchie ack. the patches are all ported as far as i can see, but not really tested and debugged.
Nov 25 21:20:05 <roh> Writchie well.. i do not care anything about what somebody runs as his host-kernel. the target is my interrest there
Nov 25 21:20:06 <Writchie> only you guys can test both GTA01 and GTA02
Nov 25 21:20:12 <Writchie> roh: of course
Nov 25 21:20:31 <hozer> edistar: based on how people have reverse-egineered drivers for mac-linux, it will be a pain, but possible
Nov 25 21:20:32 <torpor> edistar: see the wiki entry for devicelock, posted above
Nov 25 21:21:03 <edistar> torpor: ok
Nov 25 21:21:04 <edistar> :)
Nov 25 21:21:09 <Writchie> roh: except that some release points are better than others from a bug point of view. 
Nov 25 21:21:37 <roh> hozer well.. what ive seen so far from google means they will not care for gpl anymore than they must due to the license. there will be huge binary blobs which do hw-abstraction and no source for it. the kernel just proxys some stuff or allows mmaping it from userspace
Nov 25 21:21:43 <pavelm> roh: http://wiki.openmoko.org/wiki/One_transistor_too_many was updated with your words ... and with some warnings.
Nov 25 21:21:51 <edistar> +1 for the puzzle!
Nov 25 21:21:59 <SpeedEvil> Writchie: by 'gsm current should be' do you mean 'that'd be nice', or do you mean 'I've observed this? 
Nov 25 21:22:20 <SpeedEvil> Writchie: because for me, it's about an order of magnitude over 3-4mA
Nov 25 21:22:25 <Writchie> SpeedEvil: some calypso based modules are 3 to 4 amps in gsm idle
Nov 25 21:22:30 <SpeedEvil> ah
Nov 25 21:22:33 <hozer> roh: good, then FIC will have a very stable long-term market niche ;)
Nov 25 21:22:34 <SpeedEvil> right.
Nov 25 21:22:39 <roh> pavelm you can add that people who are crazy enough to try it can/should contact me
Nov 25 21:22:48 <Writchie> 3 to 4 ma
Nov 25 21:22:51 <SpeedEvil> That's more sane - compared to my nokia.
Nov 25 21:22:55 <IfChinsCouldKill> Borg: renewed the zip for the last time, I removed some non puzzle lines in north and mid.
Nov 25 21:23:05 <Writchie> 3 to 4 amps is more like my present neo ;)
Nov 25 21:23:10 <roh> hozer exact. i do not fear google yet. they wilder in microsofts, symbians and apples market, not in ours
Nov 25 21:23:14 <pavelm> ...I guess "I'll just take a sharp knife and kill that transistor" was a bit too naive?
Nov 25 21:23:17 <SpeedEvil> Writchie: any thoughts on where the current is to?
Nov 25 21:23:19 <borg_> IfChinsCouldKill: ah ok ;)
Nov 25 21:23:29 <pavelm> roh: Care to give me contact info? Or should I just write "roh on rc"?
Nov 25 21:23:34 <pavelm> "roh on irc"?
Nov 25 21:23:41 <anrp> huh
Nov 25 21:23:42 <_buz> roh: it sort of depends how much of a premium the linux crowd will really put on a purely gpl device
Nov 25 21:23:45 <borg_> IfChinsCouldKill: http://wiki.openmoko.org/wiki/Display_Locker
Nov 25 21:23:48 <roh> either that or write mail via the wiki
Nov 25 21:23:51 <anrp> low heat/low flow hot air would probably do it
Nov 25 21:23:51 <Writchie> gsm idle is can be low duty cycle 
Nov 25 21:23:56 <Writchie> relatively
Nov 25 21:23:59 <anrp> i could try it tomorrow ..
Nov 25 21:24:09 <Writchie> the chipset can sleep 
Nov 25 21:24:31 <SpeedEvil> Writchie: it's got to be idle all of the time - the reciever has to be active in large portions of the time.
Nov 25 21:24:37 <pavelm> roh: ok..
Nov 25 21:24:37 <roh> pavelm my mailaddr is in the wiki
Nov 25 21:25:21 <roh> _buz well.. there seem to be lots of interrested people. my focus currently of getting it done. ;)
Nov 25 21:25:23 <Writchie> idle means only monitoring for an incoming connection 
Nov 25 21:25:46 <Writchie> on the current signalling channel. 
Nov 25 21:25:53 <_buz> well interest doesnt necessarily mean "i'll pay 450USD"
Nov 25 21:25:58 <SpeedEvil> Writchie: yes 
Nov 25 21:26:08 <pavelm> roh: Thanks!
Nov 25 21:26:18 <SpeedEvil> _buz: for a phone that's getting less functional - compared to what $450 used to buy you
Nov 25 21:26:23 <Writchie> this can be done with 10mw
Nov 25 21:26:50 <_buz> SpeedEvil: yeah. i wouldnt mind it being a triband very much, but it having been announced at quadband gives a bad taste to it
Nov 25 21:26:56 <Writchie> i'm not sure if the receiver is always on or wakes ups up periodically
Nov 25 21:27:03 <SpeedEvil> _buz: not what I meant
Nov 25 21:27:12 <hozer> speaking of which... who do I contact to find out what a lot of between 100 and 1000 Neo's would cost
Nov 25 21:27:26 <SpeedEvil> _buz: I mean that $450 - when it was announced bought you a fair amount less phone than it will in March?
Nov 25 21:27:38 <hozer> I have a proposal idea that needs some budget numbers
Nov 25 21:27:40 <_buz> SpeedEvil: oh yes, of course
Nov 25 21:27:58 <_buz> imho it missed the window
Nov 25 21:28:08 <Writchie> SpeedEvil: the 40ma question is if the calypso goes automatically into low power idle - especially since it does not obey the low power state commands from the mux? 
Nov 25 21:28:11 <SpeedEvil> the triband thing should be ameliorated for most people by having two versions for different markets, which is _almost_ free
Nov 25 21:28:29 <SpeedEvil> Writchie: ah
Nov 25 21:28:43 <_buz> but it doesnt inspire confidence at all. "So what else is not really working??
Nov 25 21:28:44 <SpeedEvil> (almost free for FIC to change)
Nov 25 21:28:49 <mwester-laptop> speaking of which, I'd have expected an announcement as to FIC/OM's intentions on that by now.
Nov 25 21:29:16 <_buz> mwester-laptop: i would have expected announcements on a great many issues by now
Nov 25 21:29:37 <borg_> IfChinsCouldKill: i think the puzzle theme is really userfriendly, one knows instantly what has to be done
Nov 25 21:30:13 <Writchie> mwester-laptop: this is really a no-brainer - you have to have European and American tri-band variants. 
Nov 25 21:31:10 <pavelm> ...actually, qtopia is gpl, so we can use their lock/key variant.
Nov 25 21:31:10 <_buz> i'm a lot more worried about the continuing issues with PM and, on another scale, PR
Nov 25 21:31:17 <Writchie> although personally, if I had the choice of 3 bands I'd go for 800/900/1900
Nov 25 21:31:27 <_buz> Writchie: NFW
Nov 25 21:31:31 <IfChinsCouldKill> Borg: unfortunatelly I'm too stupid to integrate in in my qemu thing, so I wait for the you tube vid ;)
Nov 25 21:31:34 <mwester-laptop> :) I know...  and it shouldn't take more than a week for engineers to figure out the costs involved, and even less for the business managers to figure out their market, at which point the head honcho makes his decision.
Nov 25 21:31:39 <_buz> there's 1800 only networks in europe
Nov 25 21:31:44 <balrog-kun> pavelm: the theme yes, the code not really
Nov 25 21:31:48 <mwester-laptop> So why are we still waiting to know?
Nov 25 21:32:00 <balrog-kun> but we can do better than them too ;)
Nov 25 21:32:05 <_buz> i'm beginning to think there are no head honchos
Nov 25 21:32:13 <pavelm> balrog? qtopia should be gpl by now.
Nov 25 21:32:15 <borg_> IfChinsCouldKill: i will try to make one ;)
Nov 25 21:33:00 <_buz> in fact, any kind of leadership seems conspicously absent
Nov 25 21:33:04 <Writchie> _buz: but if given choice between 850 and 1800 anyone in US would choose the 850
Nov 25 21:33:06 <pavelm> does suspend/resume actually work for someone here?
Nov 25 21:33:07 <balrog-kun> pavelm: i mean it is all gpl and very cool, but we can't use all of the code because the qtopia design is quite different from openmoko
Nov 25 21:33:11 <Writchie> which is why you really need quadband.
Nov 25 21:33:30 <pavelm> balrog: aha, ok, true; sorry for confusion.
Nov 25 21:33:32 <balrog-kun> pavelm: different in almost every bit, starting from the programming language
Nov 25 21:33:59 <hozer> ah crud.. can someone explain exactly what the quad vs tri band thing is?
Nov 25 21:34:03 <hozer> on a wiki page?
Nov 25 21:34:10 <balrog-kun> suspend/resuming the cpu works for me, but not waking up on modem events (but i haven't try doing anything about it yeT)
Nov 25 21:34:17 <hozer> is this a board layout issue, firmware issue, or combination of both?
Nov 25 21:34:32 <hozer> ... this explains why I get crappy T-mobile reception... 
Nov 25 21:34:33 <pavelm> hozer: quadband vs. triband? Different hw.
Nov 25 21:34:40 <Writchie> balrog-kun: if I understand the licensing right none of qtopia can be used in lgpl code. 
Nov 25 21:34:41 <_buz> broken design
Nov 25 21:34:47 <_buz> (hardware)
Nov 25 21:35:03 <anrp> t mobile is mostly 1900 in the us
Nov 25 21:35:12 <anrp> which works more or less fine
Nov 25 21:35:16 <balrog-kun> Writchie: hmmm, i don't know, that's possible but that would be also true for all gpl code in openmoko and any other distro
Nov 25 21:35:19 <Writchie> balrog-kun: i don't think the proper code is in the driver. 
Nov 25 21:35:32 <hozer> I'd like to see some description of exactly how the hardware is broken for GTA01
Nov 25 21:35:46 <balrog-kun> Writchie: which code?
Nov 25 21:35:46 <Writchie> balrog-kun: yes - true for the gpl code but not the lgpl code 
Nov 25 21:35:58 <balrog-kun> or, which driver?
Nov 25 21:36:07 <Writchie> the samsung serial port driver
Nov 25 21:36:25 <pavelm> writchie: true. qtopia is deliberately GPL. You can't include it in LGPL code.
Nov 25 21:36:30 <Writchie> there needs to be some fiddling - or I arrived at that opinion when I last looked at it.
Nov 25 21:37:13 <balrog-kun> Writchie: right, there seems to be no code for serial wake-up but i suspect it would be possible to do with current hardware and modifying just the sefotware, through smart use of the gpio pins
Nov 25 21:37:20 <balrog-kun> i'm not sure
Nov 25 21:38:08 <Writchie> balrog-kun: I'll be doing this in the mux code - it may not need changes to the samsung driver if its handled by the line discipline.
Nov 25 21:38:11 <zecke|flu> pavelm: you can include it in LGPL code, as LGPL can be relicensed to GPL ;)
Nov 25 21:38:44 <pavelm> zecke: Ok, but result will be GPL... which is probalby not what you wanted.
Nov 25 21:38:47 <pavelm> ;-)
Nov 25 21:38:52 <balrog-kun> Writchie: you mean with the modified modem firmware?
Nov 25 21:39:10 <Writchie> balrog-kun: modifed modem firmware? 
Nov 25 21:39:38 <pavelm> I'd say that it is easier to remove the transistor than make gsm chip ignore the reset line.
Nov 25 21:39:51 <balrog-kun> Writchie: accroding to the wiki in current gta01v4b modem wake-up was not taken into accoutn at all and a firmware update was required
Nov 25 21:40:11 <balrog-kun> i was suspecting that it may be possible anyway but others were rather sceptical about it
Nov 25 21:40:16 <Writchie> balrog-kun: i didn't see that 
Nov 25 21:40:58 <Writchie> IIRC changes were put in the firmware prior to shipment that would permit RTS wakeup
Nov 25 21:41:07 <balrog-kun> Writchie: it was discussed on the list, but there's also a summary of the discussion at http://wiki.openmoko.org/wiki/GTA01_gsm_modem#Wakeup_of_CPU_from_GSM_Modem
Nov 25 21:41:15 <balrog-kun> put up by mwester-laptop i believe
Nov 25 21:42:23 <balrog-kun> Writchie: ah, that's possible too
Nov 25 21:42:29 <roh> hozer if you need to contact sales for bigger volumes just mail them.
Nov 25 21:43:07 <balrog-kun> my phone, which is not from the mass production batch is already confirmed to have the old, un-updated firmware :/
Nov 25 21:43:18 <roh> hozer sales@openmoko.com
Nov 25 21:44:03 <pavelm> balrog: so there's more difference between green and red mainboards?
Nov 25 21:44:37 <balrog-kun> pavelm: the mass production boards are green? then yes
Nov 25 21:45:20 <roh> pavelm the green and the red should not have any difference in the hardware besides that transistor as far as i know.
Nov 25 21:45:43 <Writchie> balrog-kun: IIRC this change is in there - I sure hope so. 
Nov 25 21:45:44 <roh> pavelm thats also why we refrained of bumping the number
Nov 25 21:46:06 <pavelm> roh: I mean.. is the gsm firmware different?
Nov 25 21:46:09 <balrog-kun> Writchie: yes, http://wiki.openmoko.org/wiki/GTA01_gsm_modem#AT.25VER would seem to indicate so
Nov 25 21:46:19 <roh> hmm.. gsm firmware version could be higher on the mp ones, yesÃ
Nov 25 21:46:23 <Writchie> roh: do you know if the modem firmware for GPIO output IO1 is in the shipped firmware
Nov 25 21:46:24 <roh> but thats no hardware thing
Nov 25 21:46:44 <balrog-kun> http://wiki.openmoko.org/wiki/GTA01_gsm_modem#AT.25VER lists the gsm firmware timestamps for different phases
Nov 25 21:48:39 <Writchie> roh: it would be nice to have a change log for the gsm modem firmware versions. 
Nov 25 21:49:04 <Writchie> roh: nicer still the ability to reflash ;)
Nov 25 21:50:10 <pavelm> yep... reflash is important... I can find someone to remove transistor, but flashing gsm chips is probably harder
Nov 25 21:51:27 <roh> Writchie heh.. i am sure as soon as we solved the flashing problem we can also somehow put a changelog into the public
Nov 25 21:52:19 <pavelm> hmmm... http://wiki.openmoko.org/wiki/GTA01_gsm_modem is indeed interesting. It allows "netmonitor" implementation....
Nov 25 21:54:21 <Writchie> balrog-kun: it appears that we should not enter suspend unless the gsmd is in a proper state, otherwise we will just come back of
Nov 25 21:54:28 <Writchie> out of suspend with data loss. 
Nov 25 21:56:26 <roh> pavelm yes.... btw.. i somehow feel i need to point out that this information about the ti proprieatary commands does not come from fic or om ;) that was put into the wiki by someone else.
Nov 25 21:57:04 <anrp> the mysterious anonymous
Nov 25 21:58:30 <Writchie> roh: yes it comes from others manfacturers of calypso based products that may interpret the NDA differently or have rights to release things that FIC does not. 
Nov 25 21:58:33 <balrog-kun> it was put into the wiki by me and i have signed no NDA :)
Nov 25 21:58:56 <roh> anrp well.. they are not anonymous, but from documentation which was not under nda, but for different products, using the same stuff
Nov 25 21:59:19 <roh> balrog-kun well.. you also did not use the stuff from ti which we signed the nda for, right? ;)
Nov 25 22:00:22 <balrog-kun> roh: right, besides none of the documents i've seen aout the modem had any information abotu any at commands :(
Nov 25 22:02:07 <roh> ah
Nov 25 22:02:44 <pavelm> roh: I understand... NDAs are ugly stuff.
Nov 25 22:02:48 <balrog-kun> actually there weren't any docs about the modem i think
Nov 25 22:03:05 <balrog-kun> except perhaps the source code that harald and sean have access to
Nov 25 22:04:26 <pavelm> ...aha, and GTA01_gsm_modem explains that I need both removed transistor and updated firmware :-(.
Nov 25 22:05:08 <roh> balrog-kun well.. there is stuff. but its kept contained to only the few burned people who have access ;)
Nov 25 22:06:47 <balrog-kun> roh: teh wiki page proves that TI docs for the modem are not really necessary for anything
Nov 25 22:13:50 <Writchie> roh: the RTSCTS flow control on the wiki doesn't look right. Isn't it bidirectional hardware flow control. 
Nov 25 22:17:32 <zecke|flu> ljp: ping?
Nov 25 22:17:39 <ljp> zecke|flu: pong
Nov 25 22:21:28 <roh> Writchie ?
Nov 25 22:22:54 <Writchie> true RTSCTS flow control is compatible with S3C2410A's auto flow control. 
Nov 25 22:23:09 <Writchie> RTS is the neos "Ready to Receive" 
Nov 25 22:23:47 <Writchie> CTS is the modems Ready to Receive
Nov 25 22:23:54 <Writchie> each independant of the other
Nov 25 22:24:43 <Writchie> in other words - the modem won't send to the NEO when RTS is low
Nov 25 22:25:05 <Writchie> and CTS may be high, even when RTS is low
Nov 25 22:26:08 <roh> Writchie the modem wakes up the cpu via an irq when the flowcontrol signals the cpu cannot recieve (e.g. in suspend to ram)
Nov 25 22:27:29 <roh> Writchie exactly that is triggered when the modem resets due to the q1001 -bug ;)
Nov 25 22:28:23 <Writchie> roh: but this is RTS from the neo - correct?
Nov 25 22:28:43 <Writchie> and CTS on the neo is the ready to receive from the modem. 
Nov 25 22:29:26 <Writchie> if the modem drops CTS and the neo keeps sending it will overrun. 
Nov 25 22:30:02 <roh> Writchie when the modem drops cts the neo should stop sending data to it.. 
Nov 25 22:30:28 <roh> hm.. to be exact i do not know if thats high or low-active logic there.. would need to fiddle that out of the docs.
Nov 25 22:30:59 <Writchie> i think the levels are correct. 
Nov 25 22:31:32 <Writchie> it just that the wiki implies classe RTS from terminal CTS from modem operation and that not what it does - its a classic RTS/CTS flow control. 
Nov 25 22:31:50 <roh> well.. ive seen it working (suspend to ram) .. ive not tried wakeup by gsm recently, but that worked as well afaik
Nov 25 22:32:16 <Writchie> have you run any high data rate streams like fax or HSCSD
Nov 25 22:32:27 <Writchie> that is where the flow control will be an issue. 
Nov 25 22:32:32 <roh> the wiki is.. well the wiki. always press history is somethings bogus and take with a bucket of salt
Nov 25 22:32:41 <Writchie> roh: ack
Nov 25 22:33:05 <Writchie> i'm not sure the stock driver for the serial is setup for auto flow control but i'll check it out. 
Nov 25 22:33:53 <Writchie> roh: it will be necessary to take control away from autoflow control to handle entry to suspend sanely
Nov 25 22:34:19 <pavelm> I read http://wiki.openmoko.org/wiki/November_13%2C_2007_Community_Update ... if you have permission to release the gllin, just do that.
Nov 25 22:35:01 <pavelm> If your legal department is still not sure, tell me which version of gllin is okay to release and I'll place it on the web somewhere....
Nov 25 22:35:24 <Writchie> roh: what driver "owns" the interrupt input for GSM wakeup?
Nov 25 22:37:04 <roh> Writchie uh.. dunno
Nov 25 22:37:28 <roh> Writchie should be the serial one ... atleast thats where i would search
Nov 25 22:40:31 <CVirus> Where is this page that had the other devices that can run OpenMoko ?
Nov 25 22:41:30 <CVirus> found it http://wiki.openmoko.org/wiki/Supported_Hardware
Nov 25 22:43:19 <mwester-laptop> Writchie: There is nothing unusual about the hw flow control with the modem; if the wiki implies otherwise, then I wrote the text incorrectly. :(
Nov 25 22:43:49 <mwester-laptop> Writchie: but there might be some issues if one needs to "mess with" the flow control for suspend, as it is done by hardware, not the driver.
Nov 25 22:44:35 <anrp> flow control? pah :-)
Nov 25 22:44:39 <anrp> infinite buffers ftw
Nov 25 22:45:18 <Writchie> the part "when hardware flow control is enabled, a given endpoint is expected to assert its RTS line when it has data to transmit" is the part that is not class hardware flow control 
Nov 25 22:45:36 <mwester-laptop> How so?
Nov 25 22:45:39 <Writchie> for bidirectional
Nov 25 22:45:53 <Ghiottone> hi all
Nov 25 22:45:55 <Writchie> RTS/CTS is class for Terminal to Modem. 
Nov 25 22:46:12 <mwester-laptop> Yes.
Nov 25 22:46:14 <Writchie> s/class/classic/
Nov 25 22:46:14 <apt> Writchie meant: RTS/CTS is classic for Terminal to Modem. 
Nov 25 22:46:42 <Writchie> but this is really a two wire RTR control. 
Nov 25 22:46:54 <Writchie> modem asserts its line when its ready to receive. 
Nov 25 22:47:05 <Writchie> terminal asserts its line when its ready to receive. 
Nov 25 22:47:15 <Writchie> nobody says they are ready to send
Nov 25 22:47:20 <mwester-laptop> I can't speak modemese.  The rts on A is connnected to cts on B, and vice versa.
Nov 25 22:47:36 <Writchie> yes - 
Nov 25 22:47:50 <mwester-laptop> Feel free to correct the terminology in the wiki.
Nov 25 22:48:00 <Writchie> ok
Nov 25 22:48:34 <Writchie> the cross connected RTS/CTS type of hardware flow control involves no processor intervention. 
Nov 25 22:48:48 <Writchie> its' handled when the fifo's are near full on both sides. 
Nov 25 22:49:04 <Writchie> its the only one that reliable with small 16 byte fifos
Nov 25 22:49:08 <mwester-laptop> Yes.
Nov 25 22:49:09 <Writchie> running at higher speeds
Nov 25 22:57:42 * anrp introduces more acronyms, dte and dce
Nov 25 22:58:40 <mwester-laptop>  hmmmm.... I wonder what happens to the data in the HW fifo if the cpu suspends while flow control has the data flow stopped...
Nov 25 22:58:46 <mwester-laptop> I'm betting it's lost.
Nov 25 22:59:07 <Writchie> suspend is really CPU OFF so yes - it's gone
Nov 25 22:59:27 <SpeedEvil> there are assorted STR modes
Nov 25 22:59:35 <SpeedEvil> you can turn the CPU off but no the periphs
Nov 25 22:59:39 <Writchie> somehow I smell a race but I have to think it all the way through
Nov 25 22:59:40 <mwester-laptop> So we really need to assert that the fifo is empty before suspend occurs.
Nov 25 22:59:49 <SpeedEvil> but that isn't very low power
Nov 25 23:00:22 <SpeedEvil> It'd be _so_ much easier if there was a PIC controlling all of this stuff.
Nov 25 23:00:25 <Writchie> the deep suspend is CPU off except for one internal register everything is lost - and SDRAM is saved
Nov 25 23:00:27 <SpeedEvil> (IMO)
Nov 25 23:00:58 <anrp> yay, more processors
Nov 25 23:01:05 <anrp> not that i disagree, i think you're right
Nov 25 23:01:24 <SpeedEvil> For a phone selling millions, it's a bad call.
Nov 25 23:01:40 <SpeedEvil> you don't do that, you get it right the first time, and if need be make custom silicon.
Nov 25 23:01:55 <anrp> millions, eh :-)
Nov 25 23:02:22 <Writchie> i agree - given the usual unknowns reset an PM are best handled by a small reprogrammble low power microcontroller. 
Nov 25 23:02:26 <SpeedEvil> For a phone selling 10K-100K absolute tops, you may well want to take the tiny area and power hit in order to make it so much more flexible
Nov 25 23:02:53 <SpeedEvil> Given that the class of PIC involved is well under $1
Nov 25 23:03:08 <tuukkah> more ideas for gta03 :-)
Nov 25 23:03:29 <anrp> i wouldn't use a pic, simply because they have a terrible arch
Nov 25 23:03:32 <anrp> i like avr8 more
Nov 25 23:03:38 <anrp> but that's completely peripheral to the design issue
Nov 25 23:03:40 <Writchie> personally, i'd use an ti 430
Nov 25 23:04:00 <Writchie> and a xilinx cpld to route the hardware signals around more flexibly
Nov 25 23:04:12 <anrp> if you're feeling really evil you'd use an arm720t ;)
Nov 25 23:04:39 <Writchie> its a bit higher in power than the ti's
Nov 25 23:04:45 <Writchie> and less forgiving 
Nov 25 23:08:17 <Writchie> and one led from the microcontroller to tell you WTF is going on
Nov 25 23:08:36 <anrp> a signal, a light, a sound, anything!
Nov 25 23:08:50 <anrp> sadly, i think that there'd be nothing, simply keeping with the spirit of om up to now :(
Nov 25 23:09:03 <Writchie> my battery charge indicator is now three times the size of the phone
Nov 25 23:09:28 <balrog-kun> gta02 has not one, but two LEDs :)
Nov 25 23:09:37 <anrp> you could use an improperly designed switching regulator as a buzzer ..
Nov 25 23:09:53 <Writchie> except the whole f&$%8king world has to be running to drive the LED's
Nov 25 23:10:19 <Writchie> well - just u-boot i guess
Nov 25 23:10:26 <balrog-kun> not necesarily
Nov 25 23:10:41 <balrog-kun> i mean the whole world has to be running to change the state of the LEDs, yes
Nov 25 23:10:52 <balrog-kun> but you can leave them on when you go to sleep
Nov 25 23:11:15 <balrog-kun> it would be nice if the PMU could be told to driver the LEDs
Nov 25 23:11:20 <balrog-kun> *drive
Nov 25 23:11:20 <SpeedEvil> It can
Nov 25 23:11:25 <Writchie> the PMU has modes for this
Nov 25 23:11:29 <tuukkah> a "there's enough charge to light a LED" LED
Nov 25 23:11:32 <SpeedEvil> It can drive LEDs to flash in various patterns
Nov 25 23:11:35 <Writchie> but I don't think that's howe the LED's are wired. 
Nov 25 23:11:38 <balrog-kun> ah cool
Nov 25 23:11:44 <Writchie> hopefully I'm wrong. 
Nov 25 23:11:49 <balrog-kun> Writchie: :(
Nov 25 23:12:19 <SpeedEvil> Writchie: we have - AIUI - no docs on the GTA02 PMU
Nov 25 23:12:25 <Writchie> it's a different PMU on the GTA02 with no docs
Nov 25 23:12:48 <balrog-kun> the pcf50633?
Nov 25 23:13:08 <Writchie> i can't find them on NXP site  typically need to buy millions chip
Nov 25 23:13:46 <Writchie> could get them from a local rep but then I'd just be disappointed not knowing how its wired ;)
Nov 25 23:14:18 <balrog-kun> hmmm, indeed it says "Philips confidential" inthe background of every page
Nov 25 23:14:26 <balrog-kun> ;)
Nov 25 23:14:57 <Writchie> typical of parts for mobile industry
Nov 25 23:15:21 <Writchie> you don't want a knock off before you lock in your volume buyers.
Nov 25 23:15:44 <balrog-kun> in this version it includes an original I^2C logo on the firts page, unlike the 50606 datasheet iirc
Nov 25 23:16:12 <balrog-kun> i wonder if they paid the license fee for the logo or are only putting it on non-public docs
Nov 25 23:16:13 * Writchie wonders if the I2C IP went to NXP or kept by philips
Nov 25 23:18:09 * Writchie ponders a world without bullship patents and copyrights
Nov 25 23:18:24 <edistar> ja
Nov 25 23:18:34 * Writchie and without lawyers
Nov 25 23:20:22 <anrp> wot
Nov 25 23:20:58 <balrog-kun> or software bugs
Nov 25 23:21:08 <balrog-kun> or spinach!
Nov 25 23:21:40 <mwester-laptop> You know that there was never such a thing as the "first" lawyer, right?   There had to be a pair, because the only reason you need a lawyer is because someone else has one.
Nov 25 23:21:59 <Writchie> yeah - like nuclear weapons
Nov 25 23:22:27 <mwester-laptop> Which just goes to prove that lawyers are "in cahoots" with each other all the way back through time.
Nov 25 23:23:16 <Writchie> 2nd oldest profession
Nov 25 23:23:52 <mwester-laptop> One wonders which of the two professions is the more disreputable.
Nov 25 23:24:02 <Writchie> that's easy
Nov 25 23:24:40 <Writchie> one makes an honest living
Nov 25 23:25:23 <balrog-kun> hmm the pcf50633 has an RTC encryption mode
Nov 25 23:25:41 <balrog-kun> why would you want to occult the current time and date so much?
Nov 25 23:25:41 <SpeedEvil> A what?
Nov 25 23:25:47 <SpeedEvil> well
Nov 25 23:25:59 <SpeedEvil> if it's _good_ encryption - that could be really handy
Nov 25 23:26:03 <SpeedEvil> provable timestamps
Nov 25 23:26:14 <SpeedEvil> but it's probably not
Nov 25 23:26:24 <Writchie> trusted computing here we come
Nov 25 23:26:45 <Writchie> maybe its for android mode ;)
Nov 25 23:26:47 <SpeedEvil> for example, ship the chip a packet of data, it then signs it with the current time and a secret, and passes it back.
Nov 25 23:27:05 <SpeedEvil> you know the secret - you set it in the first place on your PC or wahtever
Nov 25 23:27:20 <SpeedEvil> so that you can verify that it was in fact timestamped at that time
Nov 25 23:27:54 <balrog-kun> hmm, doesn't sound like an extremely useful thing
Nov 25 23:28:12 <Writchie> since the PM battery will last the lifetime of the product this could be neat (and evil)
Nov 25 23:28:41 <balrog-kun> there's an Ambien Light Monitor input, too
Nov 25 23:28:43 <SpeedEvil> balrog-kun: consider if you want authenticated signatures from customers that their packet has arrived.
Nov 25 23:29:08 <SpeedEvil> balrog-kun: it's now _much_ harder for the delivery person to do stuff to the timestamp
Nov 25 23:29:27 <SpeedEvil> (as a most trivial application)
Nov 25 23:30:04 <balrog-kun> SpeedEvil: about as useful as encrypting any chosen part in the message
Nov 25 23:30:29 <SpeedEvil> no - you sign the signature of a given file. - which is the image of the signature.
Nov 25 23:30:30 <balrog-kun> it may be the timestamp or a part of some header or something else
Nov 25 23:30:39 <SpeedEvil> and its time
Nov 25 23:30:55 <SpeedEvil> and then you have a timestamped signature that it really happened at that time.
Nov 25 23:33:14 <balrog-kun> well, that might potentially have some use
Nov 25 23:33:43 <balrog-kun> it seems the pcf50633 also works as a 5V USB power supply
Nov 25 23:34:09 <Writchie> for powering the USB host or OTG
Nov 25 23:35:05 <balrog-kun> probably general-purpose
Nov 25 23:36:01 <Writchie> i mean that's probably why the chose it, 5V not very useful these days for much else
Nov 25 23:44:45 * pavelm disagrees about "just put a small PIC in there".
Nov 25 23:44:52 <pavelm> ...that's how the PC world works.
Nov 25 23:44:56 <pavelm> ...it is a nightmare.
Nov 25 23:46:37 <pavelm> Speed: encrypted timestamps...
Nov 25 23:46:51 <pavelm> ...you'd need quite good battery, very close to the RTC chip...
Nov 25 23:47:03 <pavelm> ...and RTC chip that does not need adjusting during its lifetime.
Nov 25 23:47:08 <pavelm> ...not too practical I'd say.
Nov 25 23:48:56 <pavelm> anyway, I think I need some sleep.
Nov 25 23:52:32 <balrog-kun> pavelm: good night
Nov 25 23:52:37 <pavelm> good night all.
Nov 26 00:32:56 <edistar> good night..
Nov 26 01:18:50 <CIA-14> 03mickey * r3507 10/trunk/src/target/OM-2007.2/panel-plugins/openmoko-panel-gsm/ (ChangeLog src/openmoko-panel-gsm.c): 
Nov 26 01:18:50 <CIA-14> openmoko-panel-gsm:
Nov 26 01:18:50 <CIA-14>  Add menu button for redisplaying status notification
Nov 26 01:18:50 <CIA-14>  Show some details in the notification
Nov 26 01:20:22 <CIA-14> 03mickey * r3508 10/trunk/src/target/OM-2007.2/artwork/themes/openmoko-standard-2/gtk-2.0/openmoko-panel-plugins: openmoko-standard-theme: make title menu bold
Nov 26 01:33:13 <borg_> mickeyl: http://homepages.upb.de/olaf/dinge/openmoko/display_lock/moko_puzzle/mokopuzzle.png
Nov 26 01:33:21 <mickeyl> :D
Nov 26 01:33:25 <borg_> :)
Nov 26 01:33:27 <mickeyl> adding it atm.
Nov 26 01:33:31 <borg_> nice
Nov 26 01:34:47 <mickeyl> who did the artwork?
Nov 26 01:34:52 <mickeyl> i want to add him to AUTHORS
Nov 26 01:34:53 <goodwill> I hate US phone companies
Nov 26 01:35:27 <borg_> mickeyl: "IfChinsCouldKill"
Nov 26 01:35:34 <borg_> dont know more about him ;)
Nov 26 01:42:00 <Crofton|home> goodwill, why?
Nov 26 01:43:38 <CIA-14> 03mickey * r3509 10/trunk/src/target/OM-2007.2/daemons/neod/ (9 files in 3 dirs): neod: add display locking, code by Olaf Lüke, artwork by 'IfChinsCouldKill'
Nov 26 01:44:06 <Crofton|home> mickeyl, ping
Nov 26 01:45:31 <mickeyl> pöng
Nov 26 01:46:01 <Crofton|home> my wife is interested in skiing conditions in EU :)
Nov 26 01:46:27 <Crofton|home> can you ask your wife for a "ski report" somtetime :)
Nov 26 01:46:37 <Crofton|home> must be awful later over there ..
Nov 26 01:48:31 <mickeyl> Crofton|home: sure thing, will give you a report soon
Nov 26 01:48:43 <Crofton|home> thanks, no hurry
Nov 26 01:48:57 <Crofton|home> she just spent an hour checking out skiing :)
Nov 26 01:55:15 <ljp> heres a ski report: cold and snowy
Nov 26 02:30:02 * HeLLoMoTo 4 TODO GRATIS PARA TU MOVIL http://www.canal-moviles.es
**** ENDING LOGGING AT Mon Nov 26 02:59:57 2007