**** BEGIN LOGGING AT Tue Jan 29 02:59:57 2008 Jan 29 03:06:46 i call the police on you Jan 29 03:38:57 there, incoming text messaging event, and I can even read it. libgsmd-tool level. Jan 29 04:24:08 openmoko: 03werner * r3969 10/trunk/src/target/u-boot/patches/gta02-nor.patch: (log message trimmed) Jan 29 04:24:08 openmoko: This patch adds the minimum infrastructure necessary for having a correct Jan 29 04:24:08 openmoko: "mtdparts" variable that contains the NOR partition, and still making sure that Jan 29 04:24:08 openmoko: u-boot understands the partition layout. Jan 29 04:24:08 openmoko: CFG_MTDPARTS_PREFIX is admittedly a bit of a hack. Should probably extend the Jan 29 04:24:12 openmoko: dynparts logic to handle multiple devices. Jan 29 04:24:14 openmoko: Note that this patch isn't in "series" yet. There's more to come. Jan 29 05:07:07 yay, second step succeeded, called a method in libgsmd and saw what gsmd tries to do in my testing framework Jan 29 05:08:20 next step: produce a response and see it arrive back via libgsmd Jan 29 05:20:30 which immediately brings up the question, wtf are libgsmd methods returning in this int? and where does the string from gsmd go... Jan 29 06:00:44 raster: only on the modified phone I have Jan 29 06:12:14 <_Krieger_> i have a fail on 'make all' with MokoMakefile on building cacao. What does my sytem lack? Java environment? http://www.pastebin.ca/876962 Jan 29 06:13:37 What version of the java sdk do you have? Jan 29 07:15:08 hi ever since I updated to the image from 23/01 my OM segfault/freezes when I try to do any operations that are IO intensive, e.g. ipkg update, scp, etc Jan 29 07:15:12 any ideas on how to fix it? Jan 29 07:59:51 good morning Jan 29 08:22:42 morning Jan 29 08:23:41 morning Jan 29 08:41:55 openmoko: 03werner * r3970 10/branches/src/target/kernel/2.6.24.x/patches/gta02-acc.patch: Jan 29 08:41:55 openmoko: sensors-inactive-unless-event-open.patch Jan 29 08:41:55 openmoko: From: warmcat Jan 29 08:41:55 openmoko: Stop the global slowdown action of the current motion sensor patch: Jan 29 08:41:55 openmoko: only accept interrupts if the event interface is opened by something Jan 29 08:41:56 openmoko: Signed-off-by: Andy Green Jan 29 08:44:57 morning Jan 29 09:09:34 openmoko: 03werner * r3971 10/branches/src/target/kernel/2.6.24.x/patches/gta02-acc.patch: (log message trimmed) Jan 29 09:09:34 openmoko: motion-sensors-irq-spi.patch Jan 29 09:09:34 openmoko: From: Andy Green Jan 29 09:09:34 openmoko: This patch removes the workqueue method originally implemented for Jan 29 09:09:34 openmoko: motion sensor service and replaces with with a platform-defined Jan 29 09:09:36 openmoko: callback which is used in the ISR. Jan 29 09:09:40 openmoko: An implementation for the callback is also provided which performs Jan 29 09:38:19 how lonk does it take to make openmoko-devel-image Jan 29 09:40:25 gambler 5-7h on a fast recent machine with fast internet. Jan 29 09:48:56 roh what does that leave me with? will i be able to bring up an emulator when its finished Jan 29 09:52:02 <`pwgen`> gambler: you can build your emulator in maybe 20 min http://wiki.openmoko.org/wiki/OpenMoko_under_QEMU Jan 29 09:55:16 gambler the emulator is seperate. that just takes minutes to build. Jan 29 09:55:51 openmoko-devel-image builds a complete toolchain (crossgcc, binutils, etc) and a 'world' means linux kernel, filesystem full of system and applications Jan 29 09:57:43 roh: including things like strace Jan 29 09:58:41 cesarb dunno. need to find out Jan 29 09:58:54 but you can surely build that if you need it Jan 29 09:59:04 balrog-kun youre there? Jan 29 10:00:54 roh: I just used strace on the emulated device yesterday Jan 29 10:01:08 roh: I found it so unexpected it was already there I had to mention it ;-) Jan 29 10:02:36 cesarb: thats why its called devel-image Jan 29 10:03:09 ;) Jan 29 10:03:22 <- just adding balrog to planet Jan 29 10:04:17 did we miss somebody else? Jan 29 10:16:22 roh *couhg cough* Jan 29 10:25:31 * ScaredyCat posts pr0n images on xcasex's blog Jan 29 10:27:15 * XorA chuckles Jan 29 10:27:22 still no mokopr0n out there Jan 29 10:27:31 there is gp2x pr0n, why no moko pr0n Jan 29 10:28:35 o rly Jan 29 10:28:39 * cesarb adds code to have a sysfs file showing all enabled clocks Jan 29 10:29:09 On qemu, it looks like all blocks except spi (wtf?) are enabled... I wonder if it's the same on the real device Jan 29 10:29:59 * cesarb wonders for the Nth time if *anyone* was worried about power usage when writing the uboot/kernel code Jan 29 10:30:23 I was.... :P Jan 29 10:30:31 k Jan 29 10:30:47 but i didn;t write it... Jan 29 10:30:56 * ScaredyCat pokes cesarb with a stick Jan 29 10:31:39 ScaredyCat: tss only pics on my blahwg are of my two bonus sons Jan 29 10:32:01 did you win them ? Jan 29 10:32:10 bonus ? Jan 29 10:32:13 eh? Jan 29 10:32:15 nah Jan 29 10:32:22 how to resolve this error Jan 29 10:32:22 cesarb: I toyed with powering up/down audio on demand, but need support from gsmd for that Jan 29 10:32:23 * ScaredyCat falls over, confused Jan 29 10:32:24 http://pastebin.ca/877451 Jan 29 10:32:26 its my gfs sons from a previous relationship, and well Jan 29 10:32:31 ahh Jan 29 10:32:33 you get the idea Jan 29 10:32:38 roh any ideas? Jan 29 10:32:39 cesarb: its actually only a few lines of code in audio driver Jan 29 10:37:12 XorA: actually, audio is almost completely a separate pair of chips Jan 29 10:37:23 XorA: I'm talking about blocks within the SoC chip Jan 29 10:37:42 cesarb: take a step back, think and look at audio driver author :-) Jan 29 10:37:52 XorA: things like rtc (which we don't use - we use the PMU RTC - but still seems to have been left enabled) Jan 29 10:37:56 cesarb: and the I2S unit is inside s3c24xx Jan 29 10:38:43 Also, UART2. There's nothing using UART2 (in fact, it shares lines with RTS/CTS of UART1, so it's useless). Why is it (at least on qemu) enabled? Jan 29 10:39:41 cesarb: please file a bug with this Jan 29 10:39:55 cesarb: that way the kernel guys can answer or go DOH!!! Jan 29 10:41:29 gambler sorry.. no clue. do you have the rest of the error? Jan 29 10:41:39 XorA: I have to check with the real device later tonight Jan 29 10:41:49 XorA: since qemu can sometimes be a bit wonky Jan 29 10:42:10 cesarb: for some strange reason s3c24xx boots with everything enabled, you have to remever to turn off devices in driver init Jan 29 10:42:19 cesarb: that one drove me mad for a bit Jan 29 10:42:34 * cesarb wonders how many uW would disabling UART2 and RTC save Jan 29 10:43:03 XorA: for kernel stuff, one should use om-kernel imho. Otherwise, it has high chances to be missed Jan 29 10:44:47 roh blog.asystemapart.com :p Jan 29 10:46:49 rtp: doesnt filing a bug against kernel assign it to openmoko-kernel anyway? Jan 29 10:48:34 roh, the full output: http://pastebin.ca/877524 Jan 29 10:49:31 XorA i think its best to file a bug, but without a mail to the system-ml (openmoko-kernel) it could go unnoticed Jan 29 10:49:37 XorA: I doubt. In a recent thread on om-kernel, it was said that the kernel guys were not really looking at the bugzilla Jan 29 10:49:54 rtp they are looking if somebody hands them a link ;) Jan 29 10:49:54 XorA: http://lists.openmoko.org/pipermail/openmoko-kernel/2008-January/000639.html Jan 29 10:50:48 gambler some problem with locale generation via local qemu invocation Jan 29 10:51:04 gambler but why it crashed i dunno. Jan 29 10:51:56 roh: which doesn't seem to happen. Also, other people might want to look at the patches. They likely won't get this link Jan 29 10:53:31 rtp yes. thats why i meant, post it to bugzilla with as much info as possible and then send a mail to openmoko-kernel with a link to the bugzilla entry. Jan 29 10:54:28 gambler you seem not to have psycho installed (python jit compiler) which would speed things up massively.. youre on x86? Jan 29 10:54:54 gambler also it cries about not having qemu.. the question for me is: why didnt it get built by OE? Jan 29 10:55:10 XorA any idea? ERROR: Nothing PROVIDES 'uboot-openmoko' (but '[]' DEPENDS on or otherwise requires it) Jan 29 10:55:24 roh: update MokoMakefile Jan 29 10:55:44 roh: ok. You make a point. But this won't work for reviewing patches. Jan 29 10:55:46 gambler there you have it ;) Jan 29 10:56:22 rtp as soon as somebody is really working on a ticket it should be assigned to the respecting devel, means he also gets mail on adding new patches. Jan 29 10:56:43 rtp: internally we are only allowed to work on unscheduled work if it is assigned a bug Jan 29 10:58:25 gambler: building on fedore core? Jan 29 10:58:45 you know, I always thought gkt programming would be hard... the programming isn't the hard bit... it that stupid layout crap.... Jan 29 10:58:50 gtk Jan 29 10:59:07 XorA: oh... I didn't know that Jan 29 10:59:28 anyone know of a simple ui design tool that works properly with gtk ? Jan 29 11:00:22 ScaredyCat: glade? Jan 29 11:00:29 glade sucks Jan 29 11:00:46 I don;t want another lib for a start... plus it lies.. Jan 29 11:00:50 ScaredyCat: you should have said a non sucky simple :-) Jan 29 11:03:05 well, glade seems to support what is effectively 'colspan' but it seems to be totally ignored when I use the C code Jan 29 11:04:28 ScaredyCat: to be honest I havent tried glade since gtk 1 Jan 29 11:04:41 what do you use? Jan 29 11:04:43 paper? Jan 29 11:05:49 * ScaredyCat just thinks it would be wrong to use a fixed container Jan 29 11:06:20 ScaredyCat: I dont do UI work, its easier that way Jan 29 11:06:36 lol Jan 29 11:25:51 roh, XorA, hey yep building on fedora Core...yep Jan 29 11:25:54 updaing MokoMakefile Jan 29 11:28:23 gambler: qemu arm doesnt play well with fedora core, you need to replace your kernel Jan 29 11:29:18 argh...ok Jan 29 11:38:38 roh: awesome Jan 29 11:43:16 balrog-kun :) Jan 29 11:43:30 lol, looks funny in that order... Jan 29 11:43:47 CM saying you aren't there... then you being there, below him Jan 29 11:43:57 balrog-kun if somebody should be in there just mail gismo or me in the future ;) Jan 29 11:45:06 roh: okays, thanks :) Jan 29 11:57:07 openmoko: 03thomas * r3972 10/trunk/src/target/opkg/opkg_download.c: opkg: protect against total being zero when calculation percentage Jan 29 12:25:15 * tonyg catches up on logs Jan 29 12:25:32 ScaredyCat, glade-3 is what I used for the (extremely ugly, mind) UI for my erlang userland Jan 29 12:25:38 It seemed entirely adequate. Jan 29 12:26:12 For traditional, desktop-oriented gui toolkits, that is :-) Jan 29 12:27:46 btw, re: yesterday's discussion of at@poff: I can confirm that the battery survived nicely overnight after I made the system at@poff before shelling out to "poweroff". Jan 29 12:28:13 some few percentage points of drop, but nothing to worry about (esp considering the power draw when it's switched on :-) ) Jan 29 12:28:16 the battery use is about 100uA when off - amp and GSM Jan 29 12:28:31 This is a battery life of aroung a year Jan 29 12:36:24 openmoko: 03werner * r3973 10/trunk/src/target/u-boot/patches/nor-default-env.patch: (log message trimmed) Jan 29 12:36:24 openmoko: This patch makes u-boot use the default environment is booting from NOR. That Jan 29 12:36:24 openmoko: way, we can be sure no contamination occurs from a possibly severely upset Jan 29 12:36:24 openmoko: NAND. Jan 29 12:36:24 openmoko: nor-default-env.patch: Jan 29 12:36:26 openmoko: - cpu/arm920t/start.S, common/env_common.c (env_relocate): new configuration Jan 29 12:36:30 openmoko: option CFG_DEFAULT_ENV_IF_NOR to use the default environment (and not try to Jan 29 12:37:53 openmoko: 03chris * r3974 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (6 files in 2 dirs): Jan 29 12:37:53 openmoko: * bindings/python/Makefile.am: Jan 29 12:37:53 openmoko: * bindings/python/moko.override: Jan 29 12:37:53 openmoko: * bindings/python/mokomodule.c: Jan 29 12:37:53 openmoko: Rename 'moko' to 'mokoui' in python bindings Jan 29 12:39:07 r Jan 29 12:39:09 e Jan 29 12:43:05 openmoko: 03chris * r3975 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/bindings/python/Makefile.am: Unbreak python bindings makefile Jan 29 12:51:17 Is it posible to post to gsmd mailing list through gmane without subscription? Jan 29 12:52:51 does it have to go through gmane? Jan 29 12:53:36 openmoko: 03chris * r3976 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/bindings/python/ (mokoui.override mokouimodule.c): *Really* unbreak python bindings Jan 29 12:55:56 I prefer to read it through gmane... but when I try to post to gmane.comp.handhelds.openmoko.gsmd it doesn't return any "error" e-mail neither my post appears in the mailing list... Jan 29 12:58:30 pauliusz thats ok. i mean. the list is gsmd-devel@lists.openmoko.org its not hosted at gmane. Jan 29 13:00:56 are the MLs broken again? Jan 29 13:01:31 Vegar nope. not that i know of. and the admin sits across to me and i can login onto the admin if just now Jan 29 13:01:53 I haven't received any mails today Jan 29 13:02:40 and youre sure its not your mta? Jan 29 13:03:14 might be, but I'm assuming that gmail is pretty stable Jan 29 13:03:31 whats your address and which lists? Jan 29 13:04:01 but the gmane is just gateway... It means that if I post to gmane.comp.handhelds.openmoko.gsmd it should just resend my post to the gsmd-devel@lists.openmoko.org Jan 29 13:04:27 roh: would you mind if I sent the address in msg? Jan 29 13:04:53 maybe there is some blacklist for gmane in this mailing list? Jan 29 13:04:56 lists -svn, -distro-devel, -kernel, -community, -device-owners Jan 29 13:07:36 Vegar n.p. just /msg me. Jan 29 14:00:50 openmoko: 03werner * r3977 10/trunk/src/target/u-boot/patches/usb-vendor.patch: (log message trimmed) Jan 29 14:00:50 openmoko: OpenMoko, Inc. now has its own USB vendor ID. Changing the ID creates a Jan 29 14:00:50 openmoko: compatibility barrier for DFU, so we only do this for GTA02. Jan 29 14:00:50 openmoko: To make it easier to locally back out this change, it gets its own patch Jan 29 14:00:50 openmoko: for a while. To be merged into uboot-gta02.patch later. Jan 29 14:00:54 openmoko: usb-vendor.patch: Jan 29 14:00:56 openmoko: - include/configs/neo1973_gta02.h (CONFIG_USBD_VENDORID), Jan 29 14:31:42 freesmartphone.org: 03emdete * r64 10/trunk/software/py-proto/pygsmd.py: fault tolerant restart with muxer implementation started Jan 29 15:02:09 looks like mailing list is down? Jan 29 15:04:34 or I must subscibe to be able to post to it? Jan 29 15:06:21 look on archives Jan 29 15:06:29 is it updating? Jan 29 15:06:35 if not, it may be down Jan 29 15:06:55 or it's moderated for those who aren't subsribed Jan 29 15:07:10 s/subsribed/subscribed/ Jan 29 15:07:11 ynezz meant: or it's moderated for those who aren't subscribed Jan 29 15:08:36 Don't actually know Jan 29 15:09:11 I vaguely recall it bounciing mail back to me saying it's waiting for admin if I got the sending name wrong Jan 29 15:09:20 kyes it is Jan 29 15:09:57 not subbed == admin approval (afaik doesn't happen) Jan 29 15:10:07 ie message never delivered Jan 29 15:12:09 ScaredyCat: thanks for info. Jan 29 15:12:35 np Jan 29 15:12:38 abc Jan 29 15:12:44 s/a\(.\)/a\1\1/ Jan 29 15:12:44 123 Jan 29 15:12:49 bah Jan 29 15:13:42 someone should set proper settings for gmane that posting without subsciptio is not allowed... Jan 29 15:25:46 I finished that little program to detect low charge rate and ask if you want to crank it up Jan 29 15:26:07 I'll test it out for a couple of days and then post it to the wiki and/or make an ipkg Jan 29 15:35:09 lol bug? Jan 29 15:35:10 root@fic-gta01:~$ ls /var/volatile/tmp/x.log -lha Jan 29 15:35:10 -rw-r--r-- 1 root root 61.5M Jan 29 09:10 /var/volatile/tmp/x.log Jan 29 15:35:27 why is x.log 61.5MB ???? Jan 29 15:35:55 it eats up my whole /var/volatile Jan 29 15:37:20 okay my x.log is filled up with only this message: Jan 29 15:37:21 ** (phone-kit:1079): WARNING **: Unhandled IO condition Jan 29 15:38:37 josch, re Jan 29 15:39:03 hi Sup3rkiddo Jan 29 15:45:53 Is there a better way to programatically unlock the phone than 'killall neod'? Jan 29 15:52:12 openmoko: 03werner * r3978 10/trunk/src/target/u-boot/patches/ (6 files): (log message trimmed) Jan 29 15:52:12 openmoko: Implement new partition layout and enable patches that Jan 29 15:52:12 openmoko: - change the vendor ID, Jan 29 15:52:12 openmoko: - make "mtdparts" work smoothly, and Jan 29 15:52:12 openmoko: - that make NOR boot use the default environment Jan 29 15:52:12 openmoko: Details: Jan 29 15:52:14 openmoko: uboot-gta02.patch: Jan 29 15:54:08 logging is soooo much fun. should say "message repeated 103,258 times in the last minute" Jan 29 15:55:12 wurp2|in_and_out: not that I know of. But I do not recell that behaviour for the last dozen times I tried (seen it gobble up 100% CPU before, tho...) Jan 29 15:56:52 Kero: Do not recall what behavior? Jan 29 15:57:09 Oh, it's not "locking up" Jan 29 15:57:17 wurp2|in_and_out: I you can't do it, maybe someone should work on a/the d-bus interface to neod. Jan 29 15:57:37 I mean get rid of the cool matrix-y image that blocks the screen Jan 29 15:57:57 Ebbe: I just don't know how to do it, and the fastest way to find out seemed to be to ask here Jan 29 15:58:26 I wrote a little program to monitor if you're doing slow charge, and throw up a GUI to ask if you want to fast charge Jan 29 15:58:27 wurp2|in_and_out: I also meant the main developers should look at it :) Jan 29 15:58:46 but I want it to unlock the phone and turn on the lcd when it shows the gui Jan 29 15:59:02 ah Jan 29 15:59:32 turning on the lcd I can figure out; just find the right proc file and send the right string to it Jan 29 15:59:39 wurp2|in_and_out, it should probably just turn on the lcd and give up on the user after a few seconds and turn it off again... they might not be looking Jan 29 15:59:56 openmoko: 03werner * r3979 10/trunk/src/target/u-boot/scripts/build: build: upgraded to SVN_REV 3978 Jan 29 16:00:05 tonyg: I didn't plan to keep it on; just let it time out the same way it normally does Jan 29 16:00:27 tonyg: Good point - it would suck to have the program for helping you charge your battery kill your battery faster Jan 29 16:09:09 Is there someone that is familiar with usb drivers for the openmoko. I'm running my usb port in host mode and it works with a keyboard. Then i build a driver for a usb radio receiver, the module load correctly on openmoko but when I connect the radio there are some errors Jan 29 16:09:39 The same driver works fine on my laptop Jan 29 16:10:17 what sort of radio? Jan 29 16:11:08 It is a DAB receiver Jan 29 16:11:22 what are the errors? Jan 29 16:11:38 I assume it doesn't require USB2? Jan 29 16:12:13 When I first connect the device, it should load a firmware then reset the device Jan 29 16:12:51 but it doesn't do it... the vender:device ids stays to the value of the "no firmware" device Jan 29 16:13:17 This device existed before USB2.0 so it is not the problem :) Jan 29 16:13:53 do you get errors from the firmware loader step? Jan 29 16:14:01 jmichel: is the firmware file in the moko? Jan 29 16:14:44 XorA: the firmware is built inside the .ko file Jan 29 16:14:46 openmoko: 03werner * r3980 10/trunk/src/target/u-boot/patches/gta02-splash.patch: Jan 29 16:14:46 openmoko: Reference to CONFIG_GTA01_REVISION broke GTA01 builds of u-boot. Jan 29 16:14:46 openmoko: gta02-splash.patch: Jan 29 16:14:46 openmoko: - board/neo1973/common/jbt6k74.c (top-level, jbt6k74_init): unfortunately, Jan 29 16:14:46 openmoko: there is no CONFIG_GTA01_REVISION. Use CONFIG_GTA02_REVISION instead. Jan 29 16:15:36 jmichel: any reports of that driver working on arm? Jan 29 16:15:47 SpeedEvil: It seems the firmware is loaded... there is no error at this step. Then the code sends a signal to the device to reset then nothing happens Jan 29 16:16:26 Ah Jan 29 16:16:29 Hmm Jan 29 16:16:33 XorA: No but I'm rebuilding the driver for the arm so I thought it would work Jan 29 16:16:56 jmichel: someone might have fucked up the packing/alignment for the firmware and the driver is sending garbage Jan 29 16:16:59 Actually no - my first thought would be tha thte bus reset would be usually done by turning off hte power, which won't work. Jan 29 16:17:06 But of course that's not hte case Jan 29 16:17:08 jmichel: that sort of error is still far too common in linux Jan 29 16:17:34 XorA: well the same driver works on my laptop with OpenSuse Jan 29 16:17:47 if the device is self powered unplugging it should do a reset :-) Jan 29 16:17:53 from usb that is Jan 29 16:18:14 jmichel: most driver do work on x86, but there are quiet a few dont work right on arm due to shoddy coding Jan 29 16:19:06 XorA: I tried to unplug/replug only the USB but it still see the device with the vendor:device ids as if the firmware was not loaded Jan 29 16:19:34 jmichel: that kind of indicates firmware isnt being squirted right Jan 29 16:20:29 XorA: What do you mean exactly? The firware is not read correctly on the arm ? Jan 29 16:20:59 jmichel: the driver for any number of reasons does not correctly send firmware to device Jan 29 16:21:12 Could it be an issue of integer size on the arm vs x86 ? Jan 29 16:21:24 jmichel: you'll probably have to stick in some debugging and see what happens Jan 29 16:25:11 XorA: I'll try to read the code of the driver again and add some debugging info to find the exact location of the problem Jan 29 16:25:55 jmichel: when you get it working you can claim credit for a better than FM receiver integrated with Neo :-) Jan 29 16:26:08 DAB is only sometimes better than FM Jan 29 16:26:30 96K MP2 is _not_ better than an even semi-decent FM signal Jan 29 16:28:10 SpeedEvil: This receiver would also be good for DMB video... which could be nice... if OpenMoko can decode it. Jan 29 16:28:21 Well... Jan 29 16:28:35 It can do VCD class video - just about Jan 29 16:31:44 (The UK regulator recently removed the 128K constraint on broadcasters on DAB - they are now free to reduce it as much as they like Jan 29 16:32:00 And under 128K MP2 isn't pretty Jan 29 16:32:55 under 128k no audio codec is pretty Jan 29 16:33:11 There are less bad ones. Jan 29 16:33:13 I guess that is a good thing no other countries adopted DAB before last year. Now they can go with DAB+ which should be better... it uses the AAC+ codec Jan 29 16:33:17 thats true Jan 29 16:33:26 but even 128k is bad with most Jan 29 16:33:44 yeah, because they think that quantity is better than quality Jan 29 16:33:46 vorbis is pretty okay for my ears with default -q 3, which is a bit less Jan 29 16:33:49 DAB _sucks_ in terms of spectral efficiency and bits/J/Hz Jan 29 16:33:51 though Jan 29 16:34:20 AAC+ at 128k is pretty good... it's the defautl Itune encoding quality Jan 29 16:34:34 which is a rather sorry proposition if you ask me Jan 29 16:34:42 (which means that a similar sized transmitter goes much less far) Jan 29 16:35:11 mmmm... Jan 29 16:35:16 * ScaredyCat wonders where notify_notification_new is Jan 29 16:35:29 ScaredyCat: you weren't told? Jan 29 16:35:55 noone notified me :P Jan 29 16:36:01 then again considering what radio generally airs, they could use 56k wma for all i care Jan 29 16:36:22 or god forbid realaudio Jan 29 16:36:31 ew Jan 29 16:36:53 or flac and i still wouldt listen to it Jan 29 16:37:30 * SpeedEvil likes some of Radio 1's output. Jan 29 16:37:34 (UK) Jan 29 16:38:00 Though much of the time I'm listening to Radio 4 - speech based network. Jan 29 16:38:14 speech is precisely the reason i never listen to radio Jan 29 16:38:27 It's not that sort of speech radio network. Jan 29 16:39:22 For example - http://www.bbc.co.uk/radio4/science/thematerialworld.shtml Jan 29 16:39:43 why does radio and tv still exist when there is the intetubes! Jan 29 16:40:08 s/inte/inter Jan 29 16:40:17 because supposedly downloading stuff aint quite legal Jan 29 16:40:32 and IP is horribly flawed way of broadcasting stuff Jan 29 16:40:51 IMO _good_ channels have a future. Jan 29 16:41:01 But primarily as a content aggregation service Jan 29 16:41:16 With the programs being available elsewhere. Jan 29 16:41:43 And Internet streaming for mobile devices like phones is still not a solution Jan 29 16:42:21 so radio still has this market Jan 29 16:43:22 Internet streaming over wireless is certainly not a solution for thousands of listeners in a relatively small space Jan 29 16:48:48 Some numbers I did back when I found what channel 5 (UK) paid for 'Joey' indicated 10p/episode - which I suspect many'd be willing to pay. Jan 29 16:53:35 tach Jan 29 17:03:45 i was probably the only person that liked 'joey' Jan 29 17:08:04 ScaredyCat: nope, there is 2 of us Jan 29 17:11:56 :) Jan 29 17:12:37 who or what is joey Jan 29 17:12:51 Sup3rkiddo: the talented one from friends in his own sitcom Jan 29 17:13:18 ok...:/, Jan 29 17:13:47 crazy american sitcoms :P Jan 29 17:17:08 That show got cancelled halfway into the first season, I thought. Jan 29 17:17:20 for a reason Jan 29 17:17:33 I never knew it existed Jan 29 17:17:35 awful show Jan 29 17:17:50 I feel dimished somehow for knowing now Jan 29 17:18:39 kdean06: it made it to halfway through second season :-) Jan 29 17:19:50 Of course, I still play D&D, so good taste and I are two ships passing in the night Jan 29 17:20:01 I'm not sure which concerns me more, that you know that or that the show lasted that long. :P Jan 29 17:22:07 I was basically meaning as an example Jan 29 17:22:26 It was a 'prime' show - maybe 5m viewers for the first few episodes in the UK Jan 29 17:22:37 wurp2: nothing wrong with D&D Jan 29 17:22:58 Oh, I'm only teasing, I tend not to care too much either way about TV. :) Jan 29 17:22:58 * SpeedEvil places all D&D detractors in a portable hole. Jan 29 17:23:15 * XorA is eagerly await 4th Ed Jan 29 17:23:53 I haven't seen anything about 4E that turns me on yet Jan 29 17:24:02 But, really, we haven't seen anything substantive about 4E yet Jan 29 17:24:18 it's too expensive Jan 29 17:24:21 In a lot of ways I like 1st ed better than 3.5E Jan 29 17:24:32 if they dont bottle out, the changes are IMO good, but then maybe thats because they make my style of character not cost insane amounts of XP Jan 29 17:24:40 'cept melee. They really did melee right in 3.5. And saving throws. Jan 29 17:26:16 wurp2: yeah melee ran well in 3.5 I hope they dont break it in 4E Jan 29 17:26:54 must be time for someone to write MokoDungeonMaster Jan 29 17:29:21 is there something like a roadmap for the stages Jan 29 17:38:43 Media player works in the latest ScaredyCat images but I'm still not getting an audio ring on incoming calls. :( Jan 29 17:39:06 kdean06: Have you an updated kernel? Jan 29 17:40:11 Yes, the kernel image in the automated buildhost from the same date... The 27th's snapshot... I HAVE sound, but not on that specific event. Jan 29 17:40:19 I can make calls and use the media player. Jan 29 17:41:57 ScaredyCat, Good job on this latest image. :) Jan 29 17:42:18 kdean06: Why are you going w/ ScaredyCat vs main build? Jan 29 17:42:22 Any hint where I might set an audio ringtone though? Jan 29 17:42:34 wurp2, Because as of the 24th, the main build didn't have working SMS. Jan 29 17:42:34 I tried two ScaredyCat images recently and both had big sound problems Jan 29 17:42:39 Of course, so did the main builds Jan 29 17:42:44 orly Jan 29 17:42:57 I thought ScaredyCat just had some extra packages Jan 29 17:43:21 like ScummVM, which is nice Jan 29 17:43:43 I've never had problems with the main builds, but there's less working features. ScaredyCat's build has working SMS but had audio issues on the 24th. Today I've got audio except for that specific thing, which is kinda big but isn't horrifical. Jan 29 17:44:04 Yeah, I'm still working off a build from about the 24th Jan 29 17:44:05 main build Jan 29 17:44:10 I had to hack on it to get sound in Jan 29 17:44:34 wurp2, I'd LOVE to know what ScaredyCat does differently. He said he used "the bleeding edge stuff" but I assumed that MokoMakefile did too. :( Jan 29 17:44:40 Oh, I wrote that little app to monitor for slow charge and ask me if I want to kick it to fast Jan 29 17:45:12 I'm going to test it for a day or 2 and add a couple of very minor features, then I'll ipkg it up or put it on the wiki Jan 29 17:45:18 depending on how lazy I feel at the time Jan 29 17:45:54 I've been flashing builds every day, but I've somehow not taken the time to sit down and figure out the build system. I've made openmoko-devel-images but that's basic and offers no control. Jan 29 17:46:24 I thought the mainline build on the 26th was supposed to be quite good Jan 29 17:46:34 including SMS, my sometimes faulty memory suggests Jan 29 17:46:39 Hrm... Jan 29 17:47:28 Any info on today's? Jan 29 17:47:36 Not that I know of Jan 29 17:47:56 We need to create a wiki page that lists builds, w/ a link to a little page for each where people can make notes Jan 29 17:47:57 I'll try. Jan 29 17:48:35 I thought there was one, but it was horribly out of date. Jan 29 17:48:49 I hate wiki. :( Jan 29 17:48:54 really? Jan 29 17:48:58 I love it Jan 29 17:49:02 At least OpenMoko's is decently organized. Jan 29 17:49:09 Way better than the common alternative, which is nothing Jan 29 17:49:35 From said wiki: Jan 29 17:49:38 # Several (non openmoko) people report that the daily image from 27 Jan 08 is quite stable for use on GTA01 and that with it you can both SMS and make and receive voice calls with good stability although (understandably) tweaks to the UI are far from complete. However, SeanC reported that under pressure testing of the GSM daemon (ie cycling it through hundreds of commands) there are still failures so some way to go yet until the GSM side is sorted. Jan 29 17:49:57 It used to say the 26th build was the good one, but he (JW) updated it today to say the 27th Jan 29 17:50:16 wurp2, Allow me to rephrase... I think Wiki is great for static info based on collective input. However, I think it's horrible for the kind of information generated by this project, since it changes SO rapidly. Jan 29 17:50:25 Link? Jan 29 17:50:41 http://wiki.openmoko.org/wiki/Community_Updates Jan 29 17:50:51 I check http://wiki.openmoko.org/wiki/Special:Recentchanges a couple of times a day Jan 29 17:51:20 kdean06: Agreed, except that I still think the likely alternatives are even more out of date html pages and (ugh) forums Jan 29 17:51:45 You are 100% right that wiki is WAY less useful for info that changes fast as it does on this project Jan 29 17:52:04 Maybe we should write a wiki that colors the background of text based on how old it is Jan 29 17:52:05 I've had the wiki vs forums debate before. But for me, Wiki is "documentation" and mailing lists are collaboration. Jan 29 17:52:34 wurp2: or even an ordered table. Jan 29 17:52:42 wurp2, THough, truth be told, the problem is less the format and more "not enough people writing their experiences down". Jan 29 17:52:45 the problem is that the poor organization common in wikis is replaced with 0 organization for a mailing list/forum Jan 29 17:53:12 SpeedEvil: except that often the info only makes sense in the context it was initially created Jan 29 17:53:32 hi Jan 29 17:53:49 You would prolly want a 'refresh' button to tell people the info was still fresh w/o updates Jan 29 17:54:03 or an everfresh option for info you know is stable Jan 29 17:54:28 wurp2, Would it make sense to establish a certain list of tests that could be used to compare images? Jan 29 17:54:52 kdean06: That's a great idea Jan 29 17:55:18 A grid on each "release notes" page with the results of basic activities Jan 29 17:55:31 * SpeedEvil passes kdean06 "#!/bin/bash" to begin his test suite. Jan 29 17:55:42 sms, make phone calls, recv same Jan 29 17:55:50 etc Jan 29 17:57:49 SpeedEvil, Would scripting the tests really make sense? I mean, for me, I look at this as moving towards a consumer device, so actually using the interface myself gives me a better feel. I'm sure we could automate things, to standardize how to "burn" out the battery, for instance to test bettery life. :P Jan 29 17:57:55 kdean06: I was assuming you meant manual tests ppl would... yeah, that Jan 29 17:58:09 I am a big fan of automated tests Jan 29 17:58:26 But for UI, they are much less useful Jan 29 17:58:37 kdean06: yes - but hard Jan 29 17:58:40 and it's very hard to automate tests for things like battery life, phone calls, etc. Jan 29 17:58:47 No, I like automation, don't get me wrong. :) Consistancy is a great thing to have for benchmarks. :) Jan 29 17:59:04 kdean06: or are you voulanteering to do it every night manually? Jan 29 17:59:24 Sort of, yes. Jan 29 17:59:27 :) Jan 29 17:59:28 Wow Jan 29 17:59:39 I thought you were just suggesting a list of tests we would put on the wiki Jan 29 17:59:49 that ppl could update with their experinces for that build Jan 29 18:00:03 SpeedEvil, I've been writing crappy articles that I hope might seed documentation at some point. Jan 29 18:00:18 crappy beats none Jan 29 18:00:20 http://www.monochromementality.com/index.php/blog/show/Anatomy-of-an-OpenMoko-Application.html Jan 29 18:01:02 I plan to break down each application, take screenshots, provide "feedback" but at this point I have no real basis to judge since I only have my own use scenario. Jan 29 18:01:43 SpeedEvil: There are some, although very out of date Jan 29 18:01:46 at togaware Jan 29 18:02:07 They were very helpful for me when I 1st got my phone Jan 29 18:02:20 to answer questions like 'how do I exit an app', etc. Jan 29 18:02:47 ssh in, kill $pid Jan 29 18:02:53 lol Jan 29 18:03:38 I love your finesse, SpeedEvil Jan 29 18:03:47 moo! Jan 29 18:03:57 You playing with apt again? Jan 29 18:04:14 no, dbus Jan 29 18:05:39 * ScaredyCat muwhahaah's Jan 29 18:06:02 dbus moos too? Jan 29 18:08:03 anyone working on gps dbus integration? Jan 29 18:08:56 dbus can moo.... if you make it Jan 29 18:15:07 Hrm... Multi-tap input. :( Jan 29 18:15:18 This is the one Chris hacked together, I think... Jan 29 18:16:18 OK, stupid question Jan 29 18:16:25 What linux distro is openmoko based upon? Jan 29 18:16:29 Hrm. I wonder if that wasn't more to close bug 1121 Jan 29 18:16:34 lfs? Jan 29 18:16:40 wurp2: angstrom Jan 29 18:16:56 XorA: Thanks, I didn't know angstrom was a distro :-) Jan 29 18:17:16 wurp2: and angstrom/OE originate from open zaurus Jan 29 18:17:56 what's the relationship between ångström and OE? Jan 29 18:18:07 just that they both originate from open zaurus? Jan 29 18:18:18 Angstrom distro is built from OE build system Jan 29 18:18:31 same as OpenMoko distro is built from OE build system Jan 29 18:18:36 Why wasn't OpenMoko started as a fork of QTopia? Jan 29 18:18:39 OE can also build distros other than Angstrom. Jan 29 18:18:48 <_Krieger_> do developers know about such error: "ERROR: Nothing PROVIDES 'uboot-openmoko' (but '[]' DEPENDS on or otherwise requires it)"? Jan 29 18:18:54 wurp2: The two are vastly different in basics Jan 29 18:19:04 wurp2: qtopia isnt a distro Jan 29 18:19:04 _Krieger_: Don't worry about it Jan 29 18:19:06 qtopia wasn't free at the time, plus it's more limited Jan 29 18:19:18 for a while there was both uboot-openmoko and u-boot-openmoko Jan 29 18:19:20 so it builds both Jan 29 18:19:31 XorA: I actually knew that ;-) Jan 29 18:19:47 mjr: I understand the 'not free' part, that shoots it down altogether Jan 29 18:19:56 mjr: But in what way is it more limited? Jan 29 18:19:58 NOTE: package qtopia-phone-4.3.0-r1: task do_compile: started Jan 29 18:20:09 Not to start a flame war betwee peeps w/ three letter names Jan 29 18:20:57 mwester: They are vastly different in basics, but why? What made the openmoko team decide to start something vastly different, rather than update an existing project? Jan 29 18:21:08 (besides the non-free part, which we've already covered) Jan 29 18:21:44 wurp2, That was the key reason Gnome was started against KDE... I think saying "other than that" is a big deal. :P Jan 29 18:21:47 Clearly openmoko would be much closer to being consumer ready on neo1973 if it had started from QTopia Jan 29 18:21:54 qtopia4 wasnt available, gtk2+ was at the time Jan 29 18:22:13 That said, there are arguements for QT and against, and for GTK and against, and for EFL and against. Jan 29 18:22:51 Though I know those aren't 1-to-1 comparisons. :) Jan 29 18:23:05 wurp2: qt is based on a single large entity that handles all gui interaction, and requires that apps be written to the qt API in C++. OM chose to use the standard X11 framework instead. There are rational arguments for both, and drawbacks for both, but in general they do not play nice together. :( Jan 29 18:23:05 * mmp notes that no-one said anything against console-based interface :-P Jan 29 18:23:30 mwester: OK, I'll buy that Jan 29 18:23:46 qtopia is officially coming to openmoko real soon now Jan 29 18:24:08 but will it co-exist with the existing X apps? Jan 29 18:24:26 prolly not Jan 29 18:24:46 no, it will be a seperate choice Jan 29 18:24:53 XorA: You mean, qtopia is officially coming to the neo? Jan 29 18:25:14 'cause I thought this whole conversation was about how openmoko and QTopia are incompatible Jan 29 18:25:19 wurp2: its already on the neo, but OpenMoko distro should include it as an option Jan 29 18:25:51 I knew it was already on the neo... I thought 'officially' meant that TrollTech would be distributing it Jan 29 18:26:18 Nokia. ;P Jan 29 18:26:25 everything is GPL or other FOSS license, either entity can make images for Neo Jan 29 18:26:26 <_Krieger_> i have a fail on 'make all' with MokoMakefile on building cacao. What does my sytem lack? Java environment? http://www.pastebin.ca/876962 I have installed dev-java/sun-jdk-1.6.0.03 and dev-java/sun-jre-bin-1.6.0.03 Jan 29 18:26:35 But now I am more confused. I thought openmoko meant the non-QTopia software & OS stack that we install on the neo Jan 29 18:26:44 kdean06: heh Jan 29 18:27:30 hey mickeyl Jan 29 18:27:36 hi XorA Jan 29 18:27:38 how are things? Jan 29 18:27:52 mickeyl: slow, C++ compiles take ages Jan 29 18:27:56 true Jan 29 18:28:13 broke angstrom 2008.1 Jan 29 18:28:22 locales dont compile in glibc 2.6.1 Jan 29 18:28:27 wurp2, SMS works on today's snapshot. Audio works for making and recieving calls, Media player works (thought I'm in my office so I don't know if it skips). They buggy matchbox keyboard has been replaced with a multi-tap keyboard. Jan 29 18:28:41 Audio alert on incoming call, startup sound. Jan 29 18:28:44 kdean06: Great!! Jan 29 18:28:55 I never had the matchbox kbd bug out on me Jan 29 18:29:06 And I think I prefer it to multi-tap, usually Jan 29 18:29:23 I know what I'm doing when I get home tonight Jan 29 18:29:51 I'm using this rootfs -> http://buildhost.openmoko.org/daily/neo1973/deploy/glibc/images/neo1973/OpenMoko-openmoko-devel-image-glibc-ipk-P1-Snapshot-20080129-fic-gta01.rootfs.jffs2 and this kernel -> http://buildhost.automated.it/OM2007.2/uImage-2.6.22.5-moko11+svnr3874-r13-neo1973.bin Jan 29 18:30:05 mickeyl: that dbus call did work, much better after food. ... Jan 29 18:30:15 ScaredyCat: glad to hear. Jan 29 18:31:06 I was looking at the .xml files from your svn .... Jan 29 18:31:41 I'm hoping that you'll add list cell-id's to the network one Jan 29 18:32:02 so that you can get a lit of the cell id's and their signal strengths Jan 29 18:32:17 s/lit/list/ Jan 29 18:32:18 ScaredyCat meant: so that you can get a list of the cell id's and their signal strengths Jan 29 18:35:42 http://www.mauve.plus.com/gsmlog is a trivial awk script to log the fields emitted by the modem about the local cell-sites repeatedly Jan 29 18:37:02 <_Krieger_> do developers know about such error: "ERROR: Nothing PROVIDES 'uboot-openmoko' (but '[]' DEPENDS on or otherwise requires it)"? Jan 29 18:37:12 <_Krieger_> sorry, wrong msg Jan 29 18:37:14 <_Krieger_> i have a fail on 'make all' with MokoMakefile on building cacao. What does my sytem lack? Java environment? http://www.pastebin.ca/876962 I have installed dev-java/sun-jdk-1.6.0.03 and dev-java/sun-jre-bin-1.6.0.03 Jan 29 18:39:58 Does anyone know a simple way to programatically unlock the phone? Jan 29 18:40:26 i.e. get rid of the screen that shows after a while when you choose 'dim, then lock' Jan 29 18:41:24 a loock screen? Jan 29 18:41:40 Yeah, the one that kind of looks like the screens from the Matrix Jan 29 18:41:56 and shows up in the app drop-down as Neod Jan 29 18:42:08 I recall talk of sliding something fom one side to the other Jan 29 18:42:24 bottom to bot Jan 29 18:42:37 SpeedEvil: he wants a friendlier script version of killall neod-locker-program Jan 29 18:42:44 I think he wants to automate that so, for instance, it unlocks on incoming call. Jan 29 18:42:45 aah :) Jan 29 18:42:54 * XorA guess someone might have/will add dbus interface :-) Jan 29 18:42:58 SpeedEvil: I said programatically Jan 29 18:43:32 One presumes from the dripping sarcasm that you know all about it, and hence can answer my question ;-) Jan 29 18:43:55 XorA: Yeah, but I want to do it before then :-) Jan 29 18:44:55 XorA: surely neod should have a dbus interface Jan 29 18:45:36 where might such a thing be documented Jan 29 18:45:51 ? Jan 29 18:46:29 did you just say documented? Jan 29 18:46:41 I did indeed Jan 29 18:47:00 * SpeedEvil plays a certain star-wars clip. Jan 29 18:47:09 well, the only programming docs seem to be embedded in the source code Jan 29 18:47:17 Documentation is one of my favorite fictitious things Jan 29 18:47:21 * ScaredyCat pats SpeedEvil Jan 29 18:47:30 way cooler than Santa Clause or the Tooth Fairy Jan 29 18:48:03 or even DukeNukem Forever Jan 29 18:49:27 *crickets chirp* Jan 29 18:49:54 huh EU sw patent thingie is still around ? Jan 29 18:50:10 well, it's unharmonized so far Jan 29 18:50:39 so basically up to the country (even though, of course, a _honest_ reading of the european patent convention disallows them) Jan 29 18:53:13 wurp2: you could use xev Jan 29 19:04:16 ScaredyCat: Thanks, let me look up info on that :-) Jan 29 19:05:45 would be a hack... I'm just reading up on the dbus stuff, since it seems to make much more sense to do it that way... then other apps can register callbacks when things like that happen Jan 29 19:09:27 ScaredyCat: Yeah, I see now. I hadn't used xev before. I will probably just 'ps' around and find the name of the gui part of the screen locker for now, and whack that Jan 29 19:09:45 I don't think I really want to emulate dragging the icon for this little app Jan 29 19:10:01 And I can update it to use dbus when the interface is available. Thanks much for your input! Jan 29 19:10:19 np.. good luck :) Jan 29 19:10:25 dbus=awesome Jan 29 19:13:13 dbus=opaque for now :-( Jan 29 19:14:39 :[ Jan 29 19:41:59 XorA: Regarding my usb driver, it seems the formware is not loading correctly. It is defined as a serie of "unsigned short" values. Where could I find the info about the arm architecture nad the size of the basic types? Jan 29 19:52:18 jmichel: if it is in C, type defining should not matter, as the compiler is what takes it from c -> arm asm Jan 29 20:10:13 jmichel: is it a struct? Jan 29 20:11:47 yes Jan 29 20:12:33 do you get any warnings if you compile with -Wpadded? Jan 29 20:14:44 Vegar: I don't have any new warnings with this option Jan 29 20:15:54 ok Jan 29 20:43:52 Anyone know offhand what phone-kit does, or a safe way to restart it? Jan 29 20:44:07 It seems to be gobbling lots of cpu right now and I wanna listen to my music :-) Jan 29 20:50:41 <_Krieger_> guys, please, pay attention to my issue. did anybody have a fail on cacao compilation? Jan 29 20:51:12 _Krieger_: sorry, no Jan 29 20:51:21 I think mine fails well before that :-/ Jan 29 20:51:47 <_Krieger_> can is somehow skip that package on make all with mokomakefile? Jan 29 20:52:04 <_Krieger_> i suggest that is not critical package? Jan 29 20:52:25 _Krieger_: If you want to bypass MokeMakefile, this may work for you: http://wiki.openmoko.org/wiki/Toolchain#Downloading_and_installing Jan 29 21:10:36 Hello Jan 29 21:10:48 I have a few questions about Neo1973 hardware Jan 29 21:10:56 I subscribed to the community mailing-list Jan 29 21:11:10 but no answer. My mail is not sent Jan 29 21:11:21 And I think the subscription does not work Jan 29 21:11:58 subscription seems to work for me Jan 29 21:12:05 How long ago did you send the email? Jan 29 21:17:16 I don't receive any confirmation or anything Jan 29 21:17:39 There have been reports about broken email lists. Jan 29 21:17:48 have you looked at the wiki hardware pges? Jan 29 21:17:53 Yes Jan 29 21:18:23 But I wanted to know is there's a way to plug a FM Tuner with RDS Jan 29 21:18:30 (USB or anything else) Jan 29 21:18:37 and if somebody did it Jan 29 21:18:41 In theory Jan 29 21:18:52 any USB device should work - with external power Jan 29 21:19:00 If it supports usb1.1 Jan 29 21:19:01 I also wanted to know if there is a DSP for sound processing Jan 29 21:19:06 no Jan 29 21:19:22 Ok Jan 29 21:19:36 That's annoying because I intended to try some speech recognition stuffs Jan 29 21:20:11 Well - basically - think of it as similar to a pentium 100 or P150 class device Jan 29 21:20:29 maybe a bit faster in some cases, a bit less in others Jan 29 21:21:34 mmm Jan 29 21:22:05 so speech recognition must be possible with enough optimization Jan 29 21:22:11 it will be hard Jan 29 21:22:26 * OJW thinks of all the games he played on 25MHz PCs... Jan 29 21:22:53 Mine was 33MHz ;-) Jan 29 21:24:28 What do you mean by speaker recognition? Jan 29 21:24:42 A) simple word matching with maybe 10 words Jan 29 21:24:53 B) speaker dependant limited dictionary Jan 29 21:25:04 C) speaker independant large dictionary? Jan 29 21:26:07 I'm still not sure Jan 29 21:26:18 it will probably be B Jan 29 21:26:29 is there any software to do A? Jan 29 21:26:33 How much work have you done in the past? Jan 29 21:27:38 I mean on speech Jan 29 21:27:51 SpeedEvil: I'm not the specialist Jan 29 21:27:54 we are a team Jan 29 21:28:07 we must choose an hardware platform Jan 29 21:28:13 And I proposed the OpenMoko Jan 29 21:28:15 Ah. Borg of speech recognition. Jan 29 21:28:40 My work is more about user interaction ;-) Jan 29 21:28:52 GTA02 - the next version - will go from 266-400MHz Jan 29 21:29:03 but that may be months out. Jan 29 21:29:21 But nothing directly related to speech. Jan 29 21:29:47 But I can buy a GTA01 now and port it to 02 later, isn't it ? Jan 29 21:29:51 What would be _lovely_ would be speaker independant voice recognition that could transcribe incoming calls. But that's not really likely I think. Jan 29 21:29:59 Basically - the same exact code should run. Jan 29 21:30:59 so if we buy the 01 now, we could get it next week (in Europe ? or more ?) and work on it before the 02 Jan 29 21:31:20 at least, I'm trying to convince my boss that it's a good idea ;-) Jan 29 21:31:27 Recent orders have been happening in that sort of timescale as I understand it. Jan 29 21:31:49 * SpeedEvil pokes anyone who's had a recent neo for an update. Jan 29 21:32:15 So, I have my answer for the DSP/voice recognition, thanks :-) Jan 29 21:32:32 It'd be nice. :) Jan 29 21:32:59 Now, if you ever heard about someone who played with an FM tuner on OpenMoko, I'm highly interested Jan 29 21:33:16 There was someone on earlier trying to get a DAB radio working. Jan 29 21:33:55 having endian issues with the firmware upload it looked like. Jan 29 21:36:32 I bought my neo about 4 or 5 weeks ago Jan 29 21:36:42 It took right at a week to get it Jan 29 21:36:51 But that was in the US... I think it's longer in EU Jan 29 21:37:32 The carrier company was never the bottleneck. Jan 29 21:37:40 Ah yes, DAB radio interests me a lot Jan 29 21:38:16 All those discussion are on hardware list, community or mainly IRC ? Jan 29 21:38:48 Because if I get one, you will get me more than a few minutes at the end of the day ;-) Jan 29 21:38:52 IRC - I'm unsure if the mailing list is up. Jan 29 21:40:28 I have gotten emails within the last couple of days Jan 29 21:40:45 from community & hardware at least Jan 29 21:40:52 Although I seem to get a lot of dupes Jan 29 21:47:35 yep but no way to subscribe Jan 29 21:48:34 You can't subscribe from the mailing list web site? Jan 29 22:07:45 good night Jan 29 22:51:13 Hiz ! Jan 29 22:53:06 I have a problem. I've a neo1973 with kernel 2.6.22-5. Jan 29 22:54:16 The bluetooth is on and my other phone detects neo193 but the phone asks for a key Jan 29 22:54:47 i Not that key I have to put, I have tried with 1234 0000 Jan 29 22:55:31 Have you read "Manually using Bluetooth" on the wiki? Jan 29 22:55:48 yes I read this manual Jan 29 22:56:33 unfortunately I haven't looked into getting bluetooth going on my neo, so I don't have any useful advice :-( Jan 29 22:57:51 need a bluetooth pin helper for om in the long run, you probably could hack it to a constant somehow for now Jan 29 23:02:38 On mobile phones. Jan 29 23:02:57 BT has revealed that 200 phone-boxes in scotland were not used in the past year Jan 29 23:04:50 LiF might get you a 6V cell Jan 29 23:04:53 oops Jan 29 23:05:17 zNaima: there is a wiki page for this: http://wiki.openmoko.org/wiki/Manually_using_Bluetooth Jan 29 23:06:11 (sorry wurp2 for the forked comment) Jan 29 23:06:50 zilux: yes I read this page but the phone asks for a key Jan 29 23:06:57 I don't have what a key Jan 29 23:07:24 I write 1234 0000 and I put in the neo passkey-agent 0000 Jan 29 23:07:34 but i don't conect the two phones Jan 29 23:09:38 what do you want to connect the phones as? Jan 29 23:10:37 I want to synchronize the contacts Jan 29 23:10:57 I think that's optimistic. Jan 29 23:11:03 Unless you plan to write that first. Jan 29 23:12:42 I don't understand Jan 29 23:12:57 :-) Jan 29 23:13:18 sofwares in current development Jan 29 23:13:26 ah ... ok ok Jan 29 23:13:30 or expected as is Jan 29 23:14:40 With qtopia could have synchronized all the contacts but with openmoko not Jan 29 23:14:56 here I simply type four digits using a bt keyboard Jan 29 23:20:05 otherwise you may perhaps want to check if the connexion is encrypted, and change the conf file accordingly Jan 29 23:22:02 gn Jan 30 02:17:09 may i ask how many people have neo1973 in here? Jan 30 02:22:56 The channel has 300 people. What do you expect, hundreds of "yeah I do"s? Jan 30 02:23:25 s/expect/want/, perhaps Jan 30 02:24:14 I don't :P Jan 30 02:24:21 Waiting for GTA03, I guess. Jan 30 02:25:27 waiting long then, since it seems likely there'll be no such thing Jan 30 02:31:31 Then what is the next step after GTA02? Jan 30 02:31:47 i have 4 Jan 30 02:31:51 something else, one presumes Jan 30 02:32:45 afaik they've always referred to Neo1973 revisions Jan 30 02:32:50 Anyway, I am waiting for HSPA and proper software (voice recognition and all those other things that make the phone more usable) before getting one. Jan 30 02:33:18 The first I get, the latter I don't. But have fun waiting. Jan 30 02:33:53 What do you mean? You don't see the use for voice recognition? Jan 30 02:34:35 I don't see why one would spesifically wait for a frill like that Jan 30 02:35:51 I am quite happy with my current Nokia phone (N72), and also the one that I had before this one (6230, that now serves as a backup), so I am in no hurry to get another phone. Jan 30 02:36:15 Voice recognition is probably the feature that I use the most, maybe after the answer call button. Jan 30 02:36:35 It's much easier just to say the name of the person you want to dial, instead of browsing the address book. Jan 30 02:36:56 if you don't really want a free phone, I'm sure Nokia will think of something else "must-have" to keep you tied in that's difficult for a free system to implement Jan 30 02:37:18 I never got into that free software religion. Jan 30 02:37:46 I never got into that condescending manner of putting things down as "religions" when they aren't Jan 30 02:37:47 then the neo is defninitely not for you Jan 30 02:37:51 its barely usable as a phone Jan 30 02:37:59 I want the best one, no matter if it is free or not. The free ones tend to get better quite quicly, so that's why I often end up using them. Jan 30 02:38:22 raster: I know, but I believe that it will be good. Jan 30 02:38:55 Righto, good to know you're uninteresting then. Jan 30 02:38:55 mjr: RMS worship then? ;) Jan 30 02:39:15 Tronic: u'll be waiting a long time. software-wise it is still in its infancy Jan 30 02:39:36 * ljp thinks neo is very usable as a phone if you use qtopia Jan 30 02:39:40 Tronic: an open pone is not about GPL and religoion and rms worship Jan 30 02:39:41 raster: I have a few years at least, before getting another phone. Jan 30 02:39:53 its aboutr technical peopl ebeing ABLE to do with a device they bought - what they want Jan 30 02:40:07 without the manufacturer saynig "ooooh - you can't do that because we just won't let you!" Jan 30 02:40:26 I would expect the development to speed up rapidly as the phone goes mass market (like has happened with Symbian, even though it is a PITA to develop for it). Jan 30 02:40:34 it hasw happened too many times that you are shoe-horned into either having to use a language and api set they vendor defines Jan 30 02:40:45 even if you dont like it - you are forced to use it Jan 30 02:40:48 OR Jan 30 02:41:13 ljp: I wonder what will happen to Qtopia now that Nokia bought Trolltech. Jan 30 02:41:14 they simply dont let you do a chunk of things - the device can do. it does it in the proprietary parts of the phone, but does not make it available to you (eg GPS unit) Jan 30 02:41:25 * mjr thinks ljp is annoying with his constant qtopia plugs, we know you love it already Jan 30 02:41:25 openmoko is about removing all those barriers Jan 30 02:41:32 Tronic: nothing. it will be the same as it always was Jan 30 02:41:45 that's actually what I suspect will happen to qtopia Jan 30 02:41:46 ljp: They could put more effort into it as well. Jan 30 02:41:51 allowing developers ultimate freedoms to make their device do just what they wanted - the only limits literally being "hardware literally can't do that - sorry" Jan 30 02:41:54 it will remain as it is :] Jan 30 02:41:57 mjr: just pointing out a better alternative Jan 30 02:42:31 raster: Symbian gets quite close to that. Jan 30 02:42:43 heh symbian is as closed as closed can be Jan 30 02:42:44 except it doesn't really Jan 30 02:43:00 You can make the phone do pretty much everything you want with Symbian software. Jan 30 02:43:08 wow, ljp has a sane view of something Jan 30 02:43:22 :) because i been assimlated by borg Jan 30 02:43:25 Sure it still is closed as hell, but raster was speaking of making the phone do things, not about it being open. Jan 30 02:44:17 i cant fix hardware drivers on symbian Jan 30 02:44:39 ljp: Are there problems with the hardware drivers on Nokia phones? I haven't noticed any. Jan 30 02:44:45 yes :] Jan 30 02:44:46 openness allows that kind of lower level tinkering Jan 30 02:45:16 ljp: Can you still access the GSM or GPS module firmwares on Neo? Jan 30 02:45:37 Tronic: but symbian itself is a nightmare to code for - have u seen their posix/libc implementation? Jan 30 02:45:52 ljp: There will always be a layer that you can't access. A hardware driver should be fairly easy to get right and once it works, there is absolutely no reason to go tinkering with it. Jan 30 02:46:03 i doubt you'll ever be able to access the GSM firmware on any phone due to regulations Jan 30 02:46:10 Tronic: as for do what i want Jan 30 02:46:10 no, but that's a disingenious argument - "Neo isn't fully open so nyah, doesn't matter that Symbian has much more closed stuff" Jan 30 02:46:18 what if the phonebook is missing some feature i want Jan 30 02:46:21 how ill i add it? Jan 30 02:46:23 i can't Jan 30 02:46:25 no soruce Jan 30 02:46:30 i need to entirely write a new one Jan 30 02:46:39 then depending on the phone i can't just replace the existing one Jan 30 02:46:47 i cant put qtopia on symbian........ yet :) Jan 30 02:46:49 raster: I know that Symbian is hell for developers, and I find it amazing that there still is that much third party software available. Jan 30 02:46:50 i need to have an auxiliary app for it Jan 30 02:47:25 raster: I would greatly prefer GTK or possibly even Qt, not to mention C++, over the Symbian/C++ language. Jan 30 02:47:39 same Jan 30 02:47:43 and openmoko is just that Jan 30 02:47:53 if i dont like the address book- i can modify it Jan 30 02:47:55 or replace it Jan 30 02:48:03 hell i can replace the windowmanager and everything Jan 30 02:48:03 raster: You can do custom phonebook on Symbian. Jan 30 02:48:10 Fring is one program that does. Jan 30 02:48:14 i can modify just about anything there Jan 30 02:48:32 i can modify the EXISTING one Jan 30 02:48:34 if i want Jan 30 02:49:23 bah! who needs a window manager Jan 30 02:49:40 or a standard display system Jan 30 02:50:06 framebuffer displays just fine :) Jan 30 02:53:33 Btw, how fast is OpenMoko on GTA01? Jan 30 02:53:42 faster than on qemu, to be sure Jan 30 02:54:12 Are there significant delays (over few hundred milliseconds) anywhere in the UI, browsing menus, etc? Jan 30 02:54:16 sluggish, but tolerable (to me anyway) Jan 30 02:54:16 ljp: :-P~ Jan 30 02:54:26 Hmmh. Jan 30 02:54:41 faster than Speedy Gonzales Jan 30 02:54:42 Tronic: its absymally slow/bad Jan 30 02:54:54 (by my standards anyway) Jan 30 02:55:10 raster: you are too used to desktops Jan 30 02:55:11 Is this because of GTK? Jan 30 02:55:20 :) probably Jan 30 02:55:23 ljp: Desktops are way too slow nowadays. Jan 30 02:55:24 or X Jan 30 02:55:39 I'd like phones to be like they used to, five years ago (in terms of speed). Jan 30 02:56:00 with a one color LCD? Jan 30 02:56:03 At that time everything just happened instantly. Jan 30 02:56:10 no it didn't Jan 30 02:56:17 (but generally more so than these days yes) Jan 30 02:56:22 ljp: no - i just know how to make x do my bidding... efficiently Jan 30 02:56:23 :) Jan 30 02:56:25 The only thing that took some time was loading contacts from the SIM. Jan 30 02:56:33 i happen to know exactly what x is doing under he covers for me Jan 30 02:56:47 And starting the phone took a little while (much, much shorter time than starting Symbian phones). Jan 30 02:56:56 * mjr is glad raster's on board, incidentally Jan 30 02:56:57 Tronic: gtk itself is not really fast Jan 30 02:57:04 but the theme used is pretty heavy too Jan 30 02:57:10 combined it leads to non-fastness Jan 30 02:57:29 openmoko being as "standard" as possible gui-wise is good Jan 30 02:57:32 it eases development Jan 30 02:57:43 your desktop app will "just work" on your phone when crosscompiled Jan 30 02:57:47 as for symbian, I'm used to an n-gage, so the neo isn't that bad for me ;) Jan 30 02:57:47 It would be nice if a lightweight UI library could be developed, with the GTK API. Jan 30 02:57:50 (if it can crap into the screen res) Jan 30 02:57:58 so special languages/toolkits etc. needed Jan 30 02:58:17 Tronic: unlikely Jan 30 02:58:18 its gtk Jan 30 02:58:25 (one that does not use XML, SVG, PNG or other heavy stuff, but rather renders the widgets internally by source code) Jan 30 02:58:35 pngs is not a problem Jan 30 02:58:37 svg is heavy Jan 30 02:58:40 dont worry Jan 30 02:58:43 we are working on stuff Jan 30 02:58:50 there are toolkits that dont do that Jan 30 02:58:54 PNG decoding is very slow on PC with libpng at least. Maybe there are better libs, though. Jan 30 02:59:06 no - png decode is not bad Jan 30 02:59:11 encode is horrid though Jan 30 02:59:13 (or maybe libpng has gotten faster in the past few years) Jan 30 02:59:22 it's not like the theme is going to need to decode stuff very many times anyway... Jan 30 02:59:35 mjr: depends how a theme works Jan 30 02:59:36 :) Jan 30 02:59:41 in my world it does Jan 30 02:59:44 well, I said "_need_" :] Jan 30 02:59:45 but thats what u have caching for Jan 30 02:59:45 :) **** ENDING LOGGING AT Wed Jan 30 02:59:56 2008