**** BEGIN LOGGING AT Sun Jan 13 02:59:57 2008 Jan 13 08:59:10 good morning Jan 13 10:52:06 Hi Jan 13 14:14:48 Nice, new u-boot :) Jan 13 14:16:32 can it boot from sdhc? :) Jan 13 14:17:06 CM, any important feature adds? Jan 13 14:18:21 Obri: ping Jan 13 14:20:44 accumulated bug fixes Jan 13 14:21:20 in ubootß Jan 13 14:21:21 ? Jan 13 14:21:46 ya Jan 13 14:21:55 wow, never even found one ^^ Jan 13 14:22:26 well, there was the usb-hang regression, some sd problems, some gfx regressions Jan 13 14:22:54 nice to know them fixed but me has no debugboard anymore :-D Jan 13 14:23:32 which version do you run currently? Jan 13 14:23:46 a very recent one... wait... Jan 13 14:24:46 svnr3692 - 27. dec 07 Jan 13 14:25:17 and even my old one from july or so never did sth unexpected ^^ Jan 13 14:26:10 * josch observed that the "select" button on the neo is much easier to press with the case opened Jan 13 14:27:45 yeah, your newer one has fixed some regressions and your older one didn't expose the bugs Jan 13 14:27:48 lucky one Jan 13 14:28:28 * CM tries to get a shelf up so he can finally unpack his computer Jan 13 14:29:14 CM, you have moved your stuff to STHLM? Jan 13 14:29:25 jeddy3: Yes, all of it :) Jan 13 14:29:34 CM, nic Jan 13 14:29:36 e Jan 13 14:29:53 you live at your GF's place? Jan 13 14:29:54 Bor vid Sabbatsbergs sjukhus Jan 13 14:30:07 Yes, 38 roomy sqm... Jan 13 14:31:21 CM relativt centralt iaf :) Jan 13 14:31:32 eller väldigt centralt Jan 13 14:31:42 Heh, jo, en kvart att gå till jobbet Jan 13 14:32:04 en kvart? what's that? Jan 13 14:32:09 15 minutes or 250 meters? Jan 13 14:32:11 :D Jan 13 14:32:15 15 min Jan 13 14:32:17 Vegar, 15 minuters Jan 13 14:32:18 -r Jan 13 14:32:27 I see Jan 13 14:32:34 jeddy3: Hehe.. Minuters låter bra ;) Jan 13 14:32:53 CM mmm...lite svengelska där :P Jan 13 14:37:24 O_o Jan 13 15:19:21 hello i have an answer, i install the latest kernel and rootfs for my neo1973. Jan 13 15:20:02 Date: 12 january, but when i reestart the phone, sounds don't work, the touchscreen sound and call sound don't work Jan 13 15:20:14 sorry but i'm spanish and my english is very poor Jan 13 15:21:46 What software version have you installed? Jan 13 15:22:53 mm uImage-2.6.22.5-moko11+svnr3801-r13-neo1973.bin and OpenMoko-openmoko-devel-image-glibc-ipk-P1-Snapshot-20080112-fic-gta01.rootfs.jffs2 Jan 13 15:24:12 Crashbit: others have reported that it doesn't work in recent builds - I haven't investigated Jan 13 15:28:46 m thanks SpeedEvil in other recent builds the sound not work. Jan 13 15:30:54 Not if ahy some solution to be able to speak with the telephone, since the sounds do not work Jan 13 15:31:13 I'm sorry - I haven't tried to fix it - so I can't comment. Jan 13 15:32:45 ok Jan 13 15:35:02 slm Jan 13 15:35:31 . Jan 13 15:36:28 halo all Jan 13 15:37:47 only one asks more, the problem roots in the kernel or in the rootfs? Jan 13 15:38:01 *have Jan 13 15:38:30 Don't know. Have you played with alsactl - to change the volumes? Jan 13 15:42:26 when i use alsamixer it says : function snd_ctl_open failed for default:no such device Jan 13 15:48:00 sounds like you have a mismatch of kernel and modules Jan 13 15:48:05 check /lib/modules Jan 13 15:48:06 and uname -a Jan 13 15:56:58 hiz Jan 13 16:50:06 Do we know if Mickey ever got the gsmd working on the a780? Jan 13 17:06:32 mickeyl: Hi! Does the gsmd work on the a780? Jan 13 17:06:59 hi. unfortunately not yet. there's still some problems with the in-kernel MUXer on the EZX platform Jan 13 17:07:46 Ah, bummer. I got an a780 to have 850MHz band access. I was hopeful. Do you know how QTopia runs on it? I'll probably just stick with OpenEZX, but I thought I could share development efforts. Jan 13 17:13:47 never tried qtopia on it Jan 13 17:14:02 i don't think there's much missing from openezx to become usable Jan 13 17:14:08 we're close Jan 13 17:14:16 hopefully just a couple of months Jan 13 17:30:38 /server -m fc.irc.su:+7000 Jan 13 18:13:02 * Sup3rkiddo is tempted to flash uboot Jan 13 18:13:56 * CoreDump is missing a bv3 version :\ Jan 13 18:17:05 openmoko: 03werner * r3821 10/trunk/src/target/u-boot/patches/ (5 files): (log message trimmed) Jan 13 18:17:05 openmoko: Apply global addition of (NOLOAD) to BSS section in linker scripts. Jan 13 18:17:05 openmoko: (In recent upstream. Catch-up still in progress.) Jan 13 18:17:05 openmoko: uboot-20061030-neo1973.patch: Jan 13 18:17:05 openmoko: - board/neo1973/gta01/u-boot.lds: set .bss type to NOLOAD Jan 13 18:17:06 openmoko: uboot-hxd8.patch: Jan 13 18:17:10 openmoko: - board/hxd8/u-boot.lds: set .bss type to NOLOAD Jan 13 18:20:08 FWIW, I just flashed said u-boot and it works Jan 13 18:21:16 openmoko: 03werner * r3822 10/trunk/src/target/u-boot/scripts/build: Updated to GIT_HEAD 6d714f82fb4b8bb7e267e9c71b8009bc670bfe88 and SVN_REV 3821. Jan 13 18:37:38 Great, there are TWO different HCLK on the LDC controller chapter, and no effort is made to distinguish between them Jan 13 18:39:26 It also has utterly unexplained magic values at some points Jan 13 18:42:38 hi Jan 13 19:01:05 If I wanted to develop an application for my phone, where should I start? Jan 13 19:01:35 Is OpenGL available or going to be available? Jan 13 19:02:04 Who invented liquid soap, and why? Jan 13 19:02:28 opengl. on a phone. with only few mhz. how slow do you want graphics to appear? :) Jan 13 19:03:01 RayDar: you should look at the toolchain http://wiki.openmoko.org/wiki/Toolchain Jan 13 19:03:31 RayDar: there will be a gpu in FreeRunner, so opengl will be possible there someday, probably Jan 13 19:03:38 The drivers shouldn't be expected any time soon, but the gta02 gpu will be opengl es capable, though only with a 512x512 rendering window without driver workarounds Jan 13 19:04:40 RayDar: liquid soap was introduced in 1980 by Minnetonka Corporation Jan 13 19:07:45 Ganneff: The idea is to speed up the graphics. Especially eye-candy stuff that consumers are becoming more expectant of. Jan 13 19:08:07 i wonder why all those "consumers" are like that. its a phone, damnit :) Jan 13 19:08:45 mjr: Thanks for the nfo. I am still not sure where to start in developing an app. There must be an SDK or something. Jan 13 19:09:39 borg_: I was suspecting the Rand corporation might have had something to do with it. Jan 13 19:16:05 Does anyone know if there's a problem in driving the LCM with a slower refresh frequency? CLKVAL is already near the minimum (HCLK/4; the only other step available would be HCLK/2), and lowering the FCLK/HCLK will lower the clock a lot. Jan 13 19:47:06 RayDar, borg_: that's a quite US-centric view. Liquid soap is much older, e.g. over 30 years ago I heard about liquid soaps in germany (e.g. "Vorwerk Schmierseife"). Jan 13 20:07:01 * * OM Bug 1179 has been created by audriusa(AT)bluewin.ch Jan 13 20:07:02 * * Mobile phone hands after 'shutdown now' from the command line Jan 13 20:07:03 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1179 Jan 13 20:07:25 schurig_home: right, that was just the first google hit ;) Jan 13 20:07:51 schurig_home: i am from germany.. :P and right, i know schmierseife of course :) Jan 13 20:11:44 borg_: do you also still know the TV game show "Spiel ohne Grenzen" (Games without limits/borders) ??? They used liberal amounts of Schmierseife (smear soap) :-) Jan 13 20:21:31 hello all Jan 13 20:28:12 schurig_home: sure :) Jan 13 20:48:58 sleep Jan 13 20:55:08 hello Jan 13 20:55:14 i want to know more about openmoko Jan 13 20:55:18 it's an amazing project Jan 13 20:55:54 i have been reading some articles Jan 13 20:56:06 i will try to boot neo with qemu Jan 13 21:39:57 does the 1973 currently support SDHC? Jan 13 21:43:22 yes Jan 13 21:43:28 i use a 4gb sdhc from sandisk Jan 13 21:43:32 works like a charm :) Jan 13 21:43:50 there is a list of working sd cards in the wiki Jan 13 21:44:52 http://wiki.openmoko.org/wiki/Supported_microSD_cards Jan 13 21:46:40 i have a 8gb sandisk Jan 13 21:47:13 yes it does SDHC Jan 13 21:47:17 does yours work? Jan 13 21:47:45 don't have the phone Jan 13 21:50:03 matthew: mail me it, and I'll check it, and return it in a couple of months :) Jan 13 21:50:24 heh, i'm sure you would Jan 13 21:50:44 SpeedEvil: hi Jan 13 21:50:49 hi Jan 13 21:51:06 * mjr still has his on order, the local store isn't getting them very fast... (8G µsdhc) Jan 13 21:51:07 SpeedEvil: coded so far: core, time, serial, lcd. Tested so far: nothing. Jan 13 21:51:16 cesarb: :) Jan 13 21:51:23 cesarb: you've a debug board I assume? Jan 13 21:51:31 * SpeedEvil is awaiting mail. Jan 13 21:51:32 SpeedEvil: yeah. never used it, however Jan 13 21:51:52 SpeedEvil: didn't even break the small piece of paper with a red X ;-) Jan 13 22:01:34 hm, flash-qemu-local seems to be broken Jan 13 22:15:08 cesarb: how so? Jan 13 22:15:25 * raster yawns Jan 13 22:15:48 * ljp hands raster a cuppa tea Jan 13 22:17:07 rwhitby: the new uboot is probably misworking Jan 13 22:17:24 rwhitby: it timeouts after doing what seems no forward progress at all Jan 13 22:17:43 hrrrm Jan 13 22:17:53 why has my desktop suddenly decided its 1hr earlier Jan 13 22:18:23 rwhitby: I'm using my slow saturated connection to do a flash-qemu-official, which should work after a couple of hours of downloading Jan 13 22:18:28 raster: outdated tzdata? Jan 13 22:18:53 cesarb: apt-get dist-upgraded yesterday Jan 13 22:18:56 shouldnt be Jan 13 22:18:58 ... Jan 13 22:19:00 :/ Jan 13 22:19:18 raster: if you are in debian, that only applies if you have the volatile repository Jan 13 22:19:23 raster: which is NOT the default Jan 13 22:19:34 raster: on ubuntu, you need -updates; -security isn't enough Jan 13 22:20:14 hmm Jan 13 22:20:18 well ran ntpdate by hand Jan 13 22:20:20 happy now Jan 13 22:20:26 * cesarb always has to pay attention to tzdata; Brazil changes the summer time dates almost every single year, AND does so with only a few weeks to spare usually Jan 13 22:20:48 aaah the joy of daylight savingsa Jan 13 22:20:55 i wish it would just be killed off Jan 13 22:21:10 +1 from me. Jan 13 22:23:48 raster: ah, and check if your distro is set to use the RTC as GMT instead of local time; having the RTC with local time (as some other operating system does) tends to cause trouble with summer time/DST transitions Jan 13 22:24:20 * cesarb has read somewhere that the other operating system asks on boot what the correct time is if you boot it in the middle of a transition, which is ridiculous Jan 13 22:31:34 cesarb: well its ubuntu Jan 13 22:31:40 so dunno Jan 13 22:31:45 i havent paid much attention to that Jan 13 22:31:47 either way Jan 13 22:31:50 clock is fixed now Jan 13 22:31:51 happy Jan 13 22:31:52 :) Jan 13 22:41:07 * ljp is happy he lives where daylight savings is not used Jan 13 22:42:44 ljp: lucky you Jan 13 22:55:31 Is the GTA02 out? Can I buy it and get a plan with some telecom etc, and it will work? Jan 13 22:58:23 no,no,no Jan 13 22:58:31 well, no,no,maybe Jan 13 22:58:50 the last one only if your telco allows u to get just a sim card Jan 13 22:58:56 and an account with that Jan 13 22:59:22 tho technically u can get a plan - you may be forced to buy a phone with the plan as the teclo has no "byo" schemes Jan 13 23:12:36 unstable: it depends highly from where you are Jan 13 23:12:52 in germany it is very common to just buy a sim card without a phone Jan 13 23:13:01 in the usa, as i heard, it is very uncommon Jan 13 23:13:18 * cesarb notes everyone forgot to point unstable the topic Jan 13 23:13:20 in japan it's impossible Jan 13 23:13:41 thats bad for the japanese people Jan 13 23:13:42 borg_: I'm in New York Jan 13 23:13:53 borg_: they dont seem to mind Jan 13 23:13:54 ATT, Verizon, T-Mobile, Sprint Jan 13 23:14:37 they have some of the most advanced phones around Jan 13 23:14:50 just get the cheapest, nastiest phone possible with your plan Jan 13 23:15:00 AT&T gave me a RAZR free when I signed up Jan 13 23:15:08 I yanked the sim out and put it in my 6310i Jan 13 23:15:16 they have 2+mpixel cameras when most "western" phones were still considering "vga" res cameras to be "high quality" Jan 13 23:18:49 ewon: that can be unlocked, and sell it on ebay Jan 13 23:27:42 looks like the NAND controller will be a problem for cpufreq. The debug line printed while running on qemu says Twrph0=8, which is already the maximum (and would have to be increased for lower clocks) Jan 13 23:29:21 cesarb: as I understood it - the CPU main bus will simply slow down, as will the DRAM. Jan 13 23:29:54 cesarb: at 50Mhz - the bus will be running at 50Mhz - which will make the RAM consume mildly more power per bit output. Jan 13 23:30:40 cesarb: I note that 's3c410 speed 50' in the u-boot serial - actually does work. Jan 13 23:30:50 (exact command wrong) Jan 13 23:31:50 SpeedEvil: I'm looking at NAND, not DRAM Jan 13 23:31:56 cesarb: ah Jan 13 23:32:03 cesarb: I don't think I looked at that. Jan 13 23:32:03 SpeedEvil: the NAND controller has three timing parameters Jan 13 23:32:28 SpeedEvil: one of them is already at the maximum for gta01-core.patch, if the kernel I just ran on qemu calculated everything right. however, Jan 13 23:33:02 it looks like it's a conservative value - the NAND chip datasheet (assuming it's the correct datasheet) is giving me values of 40 to 25 (the value on gta01-core.patch is 60) Jan 13 23:33:22 (the exact value depends on the voltage and on another of the values) Jan 13 23:33:57 SpeedEvil: so, it's quite possible that we cannot run the NAND controller on slow mode :( (but I should make some more calculations before being sure) Jan 13 23:34:04 Is that a minimum? Jan 13 23:34:19 SpeedEvil: yes, and there's no maximum Jan 13 23:35:39 SpeedEvil: page 14 of the flash datasheet (ds_k9f1208x0b_rev03.pdf) and page 213 of the soc datasheet (993579um_s3c2410a_rev11.pdf) Jan 13 23:37:06 the SoC values are TACLS, TWRPH0, TWRPH1, which correspond to tCS, tWP, tCH on the flash datasheet Jan 13 23:37:33 I can't look at them at the moment. Jan 13 23:37:48 (lousy small display) Jan 13 23:37:52 (no, not 2.8") Jan 13 23:37:58 SpeedEvil: lol Jan 13 23:39:42 SpeedEvil: the values for tCS and tCH are 10, while on gta01-core they are 20 (tCS can be 0 but then you need a longer tWP) Jan 13 23:40:17 SpeedEvil: the values for tWP are 40, 25, 25 for 1.8V, 2.7V, 3.3V Jan 13 23:41:58 SpeedEvil: and on gta01-core it's 60 Jan 13 23:42:24 SpeedEvil: and here's what qemu outputs: s3c2410-nand s3c2410-nand: Tacls=3, 22ns Twrph0=8 60ns, Twrph1=3 22ns Jan 13 23:43:50 SpeedEvil: the maximum TACLS is 4, the maximum for the other two are 8. They are multiplied by HCLK (they are counting HCLK cycles) Jan 13 23:43:53 cesarb: I don't know - I can only say that it does in fact work at 50Mhz Jan 13 23:44:30 SpeedEvil: the kernel doesn't give an error while setting these registers? Jan 13 23:44:39 The kernel isn't awake then Jan 13 23:44:44 u-boot sets them Jan 13 23:44:54 it's documented on the u-boot page of the wiki Jan 13 23:45:11 I don't think the kernel touches them at all Jan 13 23:47:10 hm, booting at 50MHz on qemu gives me "s3c2410-nand s3c2410-nand: Tacls=2, 40ns Twrph0=4 80ns, Twrph1=2 40ns"... the multipliers DECREASED? Jan 13 23:47:33 they are not multipliers I thought - but divisors Jan 13 23:47:39 SpeedEvil: the kernel does change the NAND registers... Jan 13 23:48:01 cesarb: well - it booted and worked 'normally' (screen was flickery) Jan 13 23:48:19 SpeedEvil: makes sense. I might have gotten confused with 1/s versus s Jan 13 23:49:03 SpeedEvil: so, in fact, running on SLOW the times would only get even longer, with no adverse effects Jan 13 23:49:35 If hte datasheet for the nand does not give maximum times Jan 13 23:49:44 * cesarb has done so far: cpufreq core, timer, serial, framebuffer, nand. Tested so far: nothing ;-) Jan 13 23:49:53 Some chips are happy with the inputs changing statically - down to 1Hz or below. Jan 13 23:50:06 As long as all the timing constraints on the edges are met. Jan 13 23:50:25 SpeedEvil: there are no maximum times, so that's no problem. It might only get confusing if the edge is too slow, but I don't think that would be an issue even down to 1MHz Jan 13 23:50:41 SpeedEvil: typed faster than me ;-) Jan 13 23:51:26 * cesarb wonders how many drivers left Jan 13 23:57:00 "PCLK should be more than 20MHz to use USB Device Controller stably." <- great, so no slow mode when plugged in Jan 13 23:58:11 cesarb: PCLK is a divisor off FCLK isn't it? Jan 13 23:58:33 SpeedEvil: PCLK is divided from HCLK which is divided from FCLK Jan 13 23:58:44 SpeedEvil: even if we use the 1:1:1 mode, SLOW is too... slow. Jan 13 23:59:04 It'd be nice if the clocking was a hair more flexible. Jan 13 23:59:22 SpeedEvil: agreed :D Jan 13 23:59:30 From memory, the slowest 'recommended' clock speed is 36Mhz Jan 13 23:59:37 SpeedEvil: or at least if we didn't have to "stop the world" before changing the clocks Jan 13 23:59:38 the output of the PLL that is. Jan 13 23:59:54 SpeedEvil: yeah, to get lower than that you have to bypass the PLL Jan 14 00:00:30 cesarb: or use non-recommended speeds Jan 14 00:00:30 SpeedEvil: and there isn't a 36MHz clock in the table IIRC, which means the lowest one is more like 40something Jan 14 00:02:24 With GTA02 the voltage reduction shoulf really help at that speed. Jan 14 00:03:30 * SpeedEvil wants a glamo datasheet. Jan 14 00:09:42 SpeedEvil: hm... you said the screen was flickery with slow speeds... isn't it possible that the flicker wasn't the screen, but the PWM timer which controls the backlight? Jan 14 00:09:59 too much depends on these clocks :| Jan 14 00:10:03 Umm. Jan 14 00:10:13 No - it was at maximum brightness Jan 14 00:12:05 SpeedEvil: it's not the brightness, it's that the timer which controls the PWM has to be fast enough to avoid flicker. If it got too slow, it's possible its frequency went into the visible range. Jan 14 00:12:24 cesarb: don't think so. Jan 14 00:12:39 cesarb: if it's set to 100% duty cycle, it won't flicker at all Jan 14 00:12:42 that only applys when the brightness isnt at 100% Jan 14 00:12:48 heh snap Jan 14 00:19:26 wel, so far the only missing ones are: backlight, vibrator, touchscreen, SD/MMC, I2C, IIS, SPI, actually compiling everything, and testing. Jan 14 00:35:40 openmoko: 03werner * r3823 10/branches/src/target/kernel/2.6.24.x/patches/ (fail-unless-uimage.patch series): Jan 14 00:35:40 openmoko: Fail the build noisily if "mkimage" can't be found, e.g., if we forgot to add Jan 14 00:35:40 openmoko: the u-boot directory to PATH. Jan 14 00:35:40 openmoko: fail-unless-uimage.patch: Jan 14 00:35:40 openmoko: - scripts/mkuboot.sh: on failure, exit 1, not 0 Jan 14 00:43:29 hi, is it just me or has planet.openmoko.org died? Jan 14 00:44:44 you mean being offline or no new blogs? Jan 14 00:45:21 openmoko: 03werner * r3824 10/branches/src/target/kernel/2.6.24.x/ (config/ config/defconfig-2.6.24-rc7 scripts/ scripts/build): Jan 14 00:45:21 openmoko: scripts/build: build script adapted for 2.6.24-rc7 kernel (for use during Jan 14 00:45:21 openmoko: transition) Jan 14 00:45:21 openmoko: config/defconfig-2.6.24-rc7: default configuration for 2.6.24-rc7 Jan 14 00:45:21 openmoko: - removed CONFIG_SND_USB_AUDIO Jan 14 00:45:22 openmoko: - removed CONFIG_BRIDGE Jan 14 00:46:14 Mononoke: no new blogs Jan 14 00:46:53 Mononoke: btw you from Japan? Jan 14 00:47:05 just your name sounds very much like a famous Japanese movie :) Jan 14 00:47:43 duffyd: yes there are no new blog, maybe mark of people being bussy and I'm not from japan Jan 14 00:49:06 Mononoke: true seems bit strange, though **** ENDING LOGGING AT Mon Jan 14 02:59:56 2008