**** BEGIN LOGGING AT Sun Mar 11 02:59:56 2007 Mar 11 04:01:18 any way to stop the device auto-powering off within u-boot? Mar 11 04:07:20 can anyone confirm I haven't just borked my device? http://pastebin.ca/390104 Mar 11 04:13:40 it's fine. even if you had, you have jtag. Mar 11 04:18:23 looks like I have done something wrong :( Mar 11 04:18:57 splash changed, boot menu has only two items in it... and boot gives errors and then powers down Mar 11 04:19:23 well, I would have JTAG if I had the appropriate cable Mar 11 04:22:15 hmm, aux button is sticking as well Mar 11 04:23:22 dissasemble, ... Mar 11 04:23:48 yep, have done... it's the button itself, not the plastic bit in front of it Mar 11 04:24:22 easiest thing probably is just to leave the plastic bit out so I can pull it out without opening it again Mar 11 04:33:45 or, just leave the front cover off completely Mar 11 04:37:34 Can you receive Bluetooth OBEX messages when the neo1973 is idle? Mar 11 04:39:22 so, erm... assuming this latest bootloader I have flashed on (u-boot-gta01v4-r3_0_0_1321.bin) is working, what happened to the USB console menu item? Mar 11 04:39:39 for those that have a neo, is there a keyguard function for it yet? Mar 11 04:40:24 i'm curious how it is implented with a touchpad Mar 11 04:42:19 ~lart silence Mar 11 04:42:20 * apt plops silence into a giant vat of herring Mar 11 04:48:30 press a pin to enable touchpad, or touch specific areas in sequence. Mar 11 04:48:41 Or press 'aux' button maybe Mar 11 04:55:30 what's the GUI called again? Mar 11 04:58:49 what gui, matchbox? Mar 11 04:59:59 GUI could apply to lots of things Mar 11 05:07:48 matchbox, thanks Mar 11 07:15:13 counter Mar 11 07:15:13 http://wiki.openmoko.org/wiki/Wishlist:Neo1973_P0_Review P1: 3 weeks 04:44:46 (21.198 +-7.0 days) (see faq) (517) Mar 11 07:26:38 mornin Mar 11 07:26:40 dang Mar 11 07:38:21 * rwhitby finally worked out how to get into u-boot without the debug board Mar 11 07:39:58 rwhitby - mokomakefile rocks Mar 11 07:40:07 I had it working manually before Mar 11 07:40:21 but I switched to mokomakefile so my install would be more common with other devs. Mar 11 07:40:26 standardization is goodness. Mar 11 07:42:48 noidd: thx Mar 11 07:43:20 only one missing dep on gento Mar 11 07:43:23 which was quilt Mar 11 07:43:32 but emerge -v quilt and I was done Mar 11 07:43:38 let me know if there is other stuff that you think it should build. I'm working on an openmoko-devel-tools target right now to build the host-based tools required to debug the neo Mar 11 07:43:56 noidd: was that listed on the OE and your distro setup wiki page? Mar 11 07:44:15 Not that I saw Mar 11 07:44:38 it just says "emerge bitbake" Mar 11 07:44:42 I'll edit Mar 11 07:44:55 http://www.openembedded.org/wiki/OEandYourDistro lists quilt for gentoo. Mar 11 07:45:20 err, not on the page I'm viewing Mar 11 07:45:31 oh, bitbake will not bring in any of the other tools. Mar 11 07:45:36 Or put it this way, it says "Old INstructions are...: Mar 11 07:45:46 yeah, that's bogus. Mar 11 07:45:59 When I read it I thought the only thing I needed to emerge was bit bake Mar 11 07:46:01 just add bitbake to the top of that list, and it's *not* old instructions. Mar 11 07:46:15 * noidd nods, will do. Mar 11 07:46:24 thx Mar 11 07:47:16 Done. Mar 11 08:01:28 * rwhitby sets bootdelay to -1 in u-boot so the phone always comes up in the bootloader, and can be operated on (and booted into Linux) remotely Mar 11 08:11:57 anyone know how to get from uboot Runtime Mode to uboot DFU Mode? Mar 11 08:17:21 ah, dfu-util does it itself. Mar 11 08:46:30 hoi :D Mar 11 08:46:37 any fancy screenshots of openmoko so far? Mar 11 08:47:30 screenshots? Mar 11 09:02:15 rwhitby: The makefile still fails at db3-native because of this old bug: http://bugs.openembedded.org/show_bug.cgi?id=1765 Mar 11 09:02:47 Can anyone give me a hint how I use that patch attatched in comment 1? Mar 11 09:20:27 good morning Mar 11 09:24:54 hi Mar 11 09:28:31 CM: apply that patch in the bug report to the file in your openembedded directory Mar 11 09:29:02 the MokoMakefile will retain your local change to that file. Mar 11 09:29:14 rwhitby: Ok, thanks. I'll look around in the OE wiki :) Mar 11 09:29:16 then lobby koen to include it in OE proper. Mar 11 09:29:26 I think it is in OE.. Mar 11 09:29:43 edit /home/moko/openembedded/site/x86_64-linux Mar 11 09:29:49 I've asked about it a month ago or so, and then someone (might have been koen) sait that "it's in dev already" Mar 11 09:29:53 Ok, thanks a lot Mar 11 09:30:16 nope, it's not in .dev Mar 11 09:30:23 ok Mar 11 09:33:19 I can't put it in myself, cause it's different to what's already there, and is different to what's in the normal ix86-common file, so it may be an indicator of a deeper problem in 64bit toolchain support. Mar 11 09:33:55 Well, as the patch comment says it would never work anyway since there's two } on that line Mar 11 09:34:01 (i.e. you have to ask the question "why should it be changed in the 64-bit case only, when the change has nothing to do with 32bit vs 64bit) Mar 11 09:34:34 maybe try it with a single } first Mar 11 09:34:50 (if *that* works, I can definitely check that in) Mar 11 09:35:17 It failed with the change from the patch.. Mar 11 09:36:33 Hmm.. forgot to remove the tmp dir I guess.. Mar 11 09:39:56 * CM cleans the whole build/tmp and restarts Mar 11 09:44:32 CM: that's a drastic restart action Mar 11 09:44:51 you could have just done "bitbake -c clean db3-native" Mar 11 09:45:10 Heh, guess that was the knowledge I was missing ;-) Mar 11 09:45:34 Hmm.. Maybe that could be added to the makefile? Some kind of clean-package thing? Mar 11 09:46:08 Or at least to set the BBPATH env in the shell or so.. Hmm Mar 11 09:48:02 no, MokoMakefile will not change the parent shell, cause that would cause grief with people developing other OE distros alongside openmoko, who might want BBPATH set a different way. Mar 11 09:48:58 I could do a clean-package-* pattern rule which runs bitbake -c clean in the build dir with the environment set up properly. Mar 11 09:57:55 counter Mar 11 09:57:55 http://wiki.openmoko.org/wiki/Wishlist:Neo1973_P0_Review P1: 3 weeks 02:02:04 (21.085 +-7.0 days) (see faq) (518) Mar 11 10:00:35 count faster, damnit. Mar 11 10:00:35 :) Mar 11 10:55:26 warning: dfu-util is broken on big-endian host architectures. Mar 11 10:55:45 (endian issues, already marked in the code, so the developers are fully aware of it) Mar 11 10:55:58 * rwhitby switches from SlugOS/BE to SlugOS/LE on the NSLU2 ... Mar 11 11:30:56 openmoko: 03rwhitby * r51 10mokomakefile/trunk/ (4 files in 3 dirs): Added bitbake patch that retrieves remote versions and only builds if they have changed Mar 11 11:33:50 run "make update-makefile ; make update ; make setup" if you want to use that new mokomakefile functionality Mar 11 12:21:36 counter Mar 11 12:21:36 http://wiki.openmoko.org/wiki/Wishlist:Neo1973_P0_Review P1: 2 weeks, 6 days 23:38:23 (20.985 +-7.0 days) (see faq) (519) Mar 11 12:46:33 in 2 weeks counter starts running in 1/2 speed (or P1 is available) Mar 11 12:47:29 counter should run faster now. :) Mar 11 12:49:03 aloril2: why? Any particular reason? Mar 11 13:12:38 morning Mar 11 13:13:38 Morning. Mar 11 13:13:46 morning :D Mar 11 13:16:11 hi Mac Mar 11 13:16:21 TRIsoft: will you sell phase1? Mar 11 13:16:50 hi hrw, let's say, we're heavily working on this issue ;-) Mar 11 13:16:59 ok Mar 11 13:17:42 hrw: And you're having fun with your Phase ß Neo ? Is it still alive ? :D Mar 11 13:17:53 *Phase 0 Mar 11 13:18:16 anyone know how to fix the rebooting every seven seconds due to a low battery charge problem? Mar 11 13:18:42 rwhitby: iirc use the fast charge uboot command Mar 11 13:18:52 'neo1973 charge fast' or something like that Mar 11 13:19:00 how do you do that when it reboots faster than you can type that command? Mar 11 13:19:02 TRIsoft: still works as I lack time to brick it Mar 11 13:19:14 hrw: lol Mar 11 13:19:24 (i have it set up so that it stops at the bootloader and doesn't automatically boot Linux) Mar 11 13:20:03 rwhitby: Werner described a method in the bugreport. Mar 11 13:20:05 rwhitby: paste the command via (usb-)serial? Mar 11 13:20:13 rwhitby: IIRC press power button once. Mar 11 13:20:45 rwhitby: bug 212 Mar 11 13:25:14 thx, that seems to work. Mar 11 13:28:13 np Mar 11 13:29:27 rwhitby: I left it powered off on USB for a couple of hours to recharge and it fixed itself. Mar 11 13:29:50 I thought I'd bricked it when it gave me the splash screen, then black, then spash, then blank etc Mar 11 13:30:29 if you get a splash screen it isn't bricked Mar 11 13:30:44 splash screen == working uboot != brick Mar 11 13:31:14 and current neo owners are equipped with debug boards iirc Mar 11 13:31:15 I know now, originally I just had that "Oh, I broke it within 2 hours of recieving it" sinking feeling Mar 11 13:31:20 I wonder if I can use the splashcmd variable to turn on autofast charging automatically on every u-boot boot ... Mar 11 13:31:36 hrw: not all got debug boards. dwery (kernel rtc maintainer) for instance didn't. Mar 11 13:31:51 rwhitby: be carefull with connecting the neo to non 500mA ports Mar 11 13:31:58 (e.g. 3 year old dell hardware) Mar 11 13:32:07 you *could* brick it and still have the uboot splash Mar 11 13:32:10 koen|recife: nslu2 can give 2A out the port :-) Mar 11 13:32:37 rwhitby: :) Mar 11 13:32:42 (and you can power the nslu2 the other way through the port :-) Mar 11 13:33:04 we blew up numerous lab computers by drawing 500mA and unplugging the usb cable Mar 11 13:33:15 Wow Mar 11 13:34:07 iirc it was some bug in an intel chipset dell used Mar 11 13:35:34 * rwhitby starts testing dfu-util again, this time on a little-endian NSLU2 build (cause it don't work big-endian) Mar 11 13:37:14 hi Mar 11 13:38:51 openmoko: 03rwhitby * r52 10mokomakefile/trunk/patches/openmoko-HEAD/ (remove-obsolete-image-recipe.patch series): Mar 11 13:38:51 openmoko: Added a patch which fixes the 'ERROR: Could not include required file Mar 11 13:38:51 openmoko: openmoko-image.bb while parsing Mar 11 13:38:51 openmoko: /home/moko/oe/packages/images/openmoko-devel-image.bb' problem by deprecating Mar 11 13:38:51 openmoko: the openmoko-devel-image.bb in openmoko svn in favour of the one in Mar 11 13:38:55 openmoko: openembedded. Mar 11 13:39:49 an interesting aspect of bbcollections - you can require a file from the directory which you have overlaid. Mar 11 13:39:55 s/can/can't/ Mar 11 13:39:56 rwhitby meant: an interesting aspect of bbcollections - you can't require a file from the directory which you have overlaid. Mar 11 13:57:46 I was just wondering if openmoko UI apps are based on some prior work or do they exist at all? Mar 11 13:58:00 * noidd tickles hardware bug 212 Mar 11 13:58:06 damn Mar 11 13:58:07 there are screenshots but are they mock ups? Mar 11 13:58:36 Moo: asome are real Mar 11 13:58:40 some are mockups Mar 11 13:58:58 and yes, some of the applications are based on prior work. Mar 11 13:59:15 we'd be idiots to write everything from scratch. This is an opensource project after all :-) Mar 11 13:59:24 * Moo^^^ nods Mar 11 13:59:33 I mean touch screen apps like dialer Mar 11 13:59:48 did you write it yourself or was there some other project? Mar 11 13:59:58 Well, I didn't write it :-) Mar 11 14:00:24 it is my understanding that the dialer was/is being written by us (meaning the openmoko project) Mar 11 14:00:45 ok. that was my guess. I just wanted to make sure. Mar 11 14:00:59 who is your UI cannon or do you have one? Mar 11 14:01:07 I'm prepared to be (and often am) wrong. Mar 11 14:01:39 your hints shed light to my darkness of unawary :) Mar 11 14:05:20 Moo^^^: http://stuff.pimp-your-penguin.org/screenshots/Neo1973/20070310/ <- real dialer screenshots Mar 11 14:09:49 cu Mar 11 14:10:07 cu hrw Mar 11 14:10:14 hrw: l8r Mar 11 14:12:08 sounds like there will be at least one german distributor nor the neo =) Mar 11 14:12:24 nice Mar 11 14:12:24 I would bet $5 on it :D Mar 11 14:12:25 rly? Mar 11 14:12:29 anything about a price yet? Mar 11 14:13:06 Sorry, no hard facts actually. Too many things to work on. Mar 11 14:13:42 somehow, the fonts on the buttons are much too small? Mar 11 14:14:05 whereas the one for the phonenumber seems kinda big Mar 11 14:14:11 (doesnt really fit long numbers) Mar 11 14:14:55 buz: the dialer is _far_ from finished and actually unusable w/ the moko theme :\ Mar 11 14:15:03 doooh Mar 11 14:16:16 is there a current qemu image available somewhere? Mar 11 14:17:45 oh i should RTFF Mar 11 14:18:13 Is this talk of P2 distributors? Mar 11 14:18:57 ?? Mar 11 14:21:14 ok, the AUX button on p0 is the one on the top left of tyhe phone right? Mar 11 14:21:22 at about 10/11 o'clock Mar 11 14:22:24 noidd: right Mar 11 14:22:54 the one which looks like an IRDA cover on the black phone Mar 11 14:23:01 I've tried 3/4 times to get uboot to kick in Mar 11 14:23:08 I'm doing something wrongh obviously Mar 11 14:24:25 i'm missing something fundimental I think Mar 11 14:25:34 noidd: well, the phone must be _off_. Then press and hold AUX, then press and hold PWR until the phone turns on Mar 11 14:26:05 after a shutdown -h now Mar 11 14:26:09 right Mar 11 14:26:12 and pulling the USB and battery Mar 11 14:26:19 I can't be any more powered off than that I guess Mar 11 14:26:23 hehe Mar 11 14:26:24 gonna try again Mar 11 14:27:01 sadly due to the short boot-delay, it is impossibly to enter u-boot before the kernel is loaded Mar 11 14:28:26 sorry, are you saying that uboot appears after the kernel has booted? Mar 11 14:28:53 no, the OM splashscreen == u-boot Mar 11 14:29:07 thats what I thought Mar 11 14:29:16 u-boot then waits $BOOTDELAY and then it starts the kernel Mar 11 14:29:31 CoreDump|home - Hold the AUX button while powerin gon --> boot menu. Mar 11 14:29:37 if $BOOTDELAY is long enough you can enter the u-boot console and interupt the boot =) Mar 11 14:29:40 did everyone get his P0 phone? Mar 11 14:29:54 loufoque - Mostly everyone, I think. Mar 11 14:29:59 Elrond: I know, noidd is having problems w/ it Mar 11 14:30:04 See, I'm doing that and its imgoring me Mar 11 14:30:32 So, let me just double check the everything i'm doing. Mar 11 14:30:36 1) shutdown -h now Mar 11 14:30:40 noidd - And $BOOTDELAY is too short? Mar 11 14:30:51 The splashscreen is there for 7-10 seconds Mar 11 14:31:04 CoreDump|home: it's ugly Mar 11 14:31:17 should I be holding the AUX button *WHILE* I'm pressing the power on? Mar 11 14:31:28 noidd: of course Mar 11 14:31:30 cos thats what i'm doing Mar 11 14:31:43 Ok, so I pull ethernet and battery Mar 11 14:31:51 sorry, usb and battery :-}) Mar 11 14:31:54 BTW: Anyone with another distro than debian at hand and wireshark 0.99.[45] on it? I need to know, that it's a debian packaging issue. :) Mar 11 14:31:59 so, I plug back in usb, then battery Mar 11 14:32:05 hold AUX down Mar 11 14:32:05 Elrond: yes, it defaults to one second....I set it to 5 which allows me to enter u-boot and abort the boot Mar 11 14:32:11 and hold power down for 4-5 seconds Mar 11 14:32:21 why is the 6 bolder than the other numbers? Mar 11 14:32:27 is it because it is the number of the beast? Mar 11 14:32:38 Elrond: Er... I got Wireshark 0.99.4 on Gentoo here Mar 11 14:32:43 shouldn't that do it? Mar 11 14:32:50 loufoque: heh, because the theme sucks / is not finished yet Mar 11 14:33:04 noidd: yes it should Mar 11 14:33:31 noidd: if it doesn't turn on, remove battery + charger and wait a bit Mar 11 14:33:41 Lumpio- - Cool. Can you do Menu -> Edit -> Preferences Look for Protocols/KRB5. Does it have an entry for some keytab file? Mar 11 14:33:46 it turns on, it jsut boots normally Mar 11 14:34:05 noidd: then your AUX button is not really pressed me thinks Mar 11 14:34:16 noidd - Does the aux button work inside linux? Mar 11 14:34:19 Elrond: Nope, that page only has one checkbox on it Mar 11 14:34:29 What is its bahaviour meant to be in linux? Mar 11 14:34:31 About reassembling messages that span multiple TCP packets Mar 11 14:34:57 Lumpio- - Okay, thanks. So this is some wireshark upstream issue. ethereal used to be able to decrypt kerberos... Mar 11 14:35:08 got it Mar 11 14:35:25 looks like I needed to leave the battery / USB out a lot longer than I did before Mar 11 14:35:27 thanks guys Mar 11 14:35:31 for the sanity check Mar 11 14:35:46 noidd - Can you document this in the wiki, please? Mar 11 14:36:05 Yup Mar 11 14:37:39 somehow editing X config to rotate the X server doesnt help? Mar 11 14:37:45 done Mar 11 14:38:37 noidd: type "setenv bootdelay 5" Mar 11 14:38:41 then "saveenv" Mar 11 14:38:57 so you can about a normal boot via ttyACM0 Mar 11 14:39:04 s/about/abort Mar 11 14:39:39 Will do. I'm currently hunting down the kernel module cdc_acm Mar 11 14:39:43 its not in my kernel sources Mar 11 14:46:53 Ok, i'm an idiot Mar 11 14:47:15 cdc_acm != cdc-acm Mar 11 14:48:27 buz: rotation is done via environment variables Mar 11 14:50:33 hello, who have the file fstests_svn.o-hand.com_.repos.misc.trunk__20070311.tar.gz? i can't find and checkout it. Mar 11 14:53:05 * CoreDump|home only has ___20061122.tar.gz Mar 11 14:54:58 why it is needed by MokoMakefile? Mar 11 14:56:47 Readon: do you have any other version of that file? Mar 11 14:57:23 i find it in downloads.openmoko.org Mar 11 14:57:37 it is 20060814 Mar 11 15:32:24 CoreDump|home: like what env variable? Mar 11 15:32:40 buz: check /etc/X11/Xserver Mar 11 15:32:46 ok Mar 11 15:33:26 since kdrive does not have a config file, all settings are either from the env or --switches to kdrive Mar 11 15:38:44 counter Mar 11 15:38:45 http://wiki.openmoko.org/wiki/Wishlist:Neo1973_P0_Review P1: 2 weeks, 6 days 20:21:14 (20.848 +-7.0 days) (see faq) (520) Mar 11 15:50:53 SpeedEvil: in 2 weeks counter runs at 1/2 speed because it starts to reduce uncertainty =(upper_limit-lower_limit)/2 Mar 11 15:51:22 Hah, fuzzy counter. Mar 11 15:52:42 Does it stop at the upper limit? Mar 11 16:27:26 SpeedEvil: yes, counter will claim P1 has been released at upper limit Mar 11 16:29:10 of course any of this doesn't take into account manual adjustments as new info becomes available Mar 11 16:30:47 Hmm, how does the P1 plan look? Does "Manually using GSM" work for someone? Mar 11 16:38:41 hm. it doesn't work for me. after echoing 1 into gta01-pm-gsm.0/power_on, ttySAC0 is still dead. Mar 11 16:39:33 pH5 - Some friend of stefan_schmidt got at least that to work. Mar 11 16:41:03 question: where can I see the status of the basic apps that are being developed at the moment? Mar 11 16:42:59 pH5: Disabled in /etc/inittab? dmeg -n1? Following the GSM manually page in the wiki works here. Mar 11 16:44:13 stefan_schmidt - "works here" includes "Doing a real phone call as in talking to a human like being on the other end"? Mar 11 16:44:41 Elrond: Yes, with the right alse config files even this works. :) Mar 11 16:45:37 stefan_schmidt - So there is some hope of P1 phones actually being useable to phone someone? Mar 11 16:45:38 pH5: You get a connected from ttySAC0 when doing root@fic-gta01:~$ cu -l /dev/ttySAC0 ? Mar 11 16:47:30 stefan_schmidt: yes, but then it's dead: http://en.pastebin.ca/390679 Mar 11 16:50:44 Elrond: I hope so. :) It's more a question how the whole chain of GSM -> gsmd -> libgsmd -> dialer is working. LaF0rge just started again work on gsmd and when mickey|austria is back we hopefully will have some tests with the dialer. Mar 11 16:51:25 pH5: OK, that could be a real problem. Let me explain. Mar 11 16:52:38 pH5: My phone has exactly the same problem. (I use another one atm) No connection to GSM seems to work. Mar 11 16:53:20 pH5: I'll send the broken one to LaF0rge next week and he takes a look what can be broken here. Mar 11 16:53:42 pH5: You're the only other one I heard about with the same problem. Mar 11 16:54:08 pH5: It would be very helpfull if you could take the time and verify this stuff once again. Mar 11 16:54:35 pH5: I like to be sure we don't show up with a user-bug here. :) Mar 11 16:55:03 The 64MB thing shouldn't impact on this, right? Mar 11 16:55:35 Elrond: No, it should not. 2 of our 3 phones here work fine. Mar 11 16:56:09 I still hope this is some kind of software bug and not a QA hardware problem. Future will show. Mar 11 16:56:09 noone wants to answer my question? Or am I just to stupid? Mar 11 16:56:20 edistar: moment Mar 11 16:56:36 ok, thanks Mar 11 16:56:55 mickeyl! Mar 11 16:57:00 cheers guys Mar 11 16:57:03 wb mickeyl Mar 11 16:57:04 stefan_schmidt - Talking about the devil. ;o)) Mar 11 16:57:15 ;) Mar 11 16:57:15 Hey, welcome back Mickey Mar 11 16:57:19 cheers TRIsoft Mar 11 16:57:21 mickeyl: Hope you had some relaxed holiday. Mar 11 16:57:21 * CM cheers Mar 11 16:57:27 hey CM Mar 11 16:57:31 mickeyl - Was you holiday good? Mar 11 16:57:33 hey mickeyl Mar 11 16:57:34 stefan_schmidt: absolutely. Mar 11 16:57:45 although i worked 8h a day it was great Mar 11 16:57:53 stefan_schmidt: same as before: http://en.pastebin.ca/390694. I'll reboot and verify, but I think it is dead all right. Mar 11 16:57:55 <-- needs to go to austria sometime again. :) Mar 11 16:57:56 edistar: Problem is that all apps are work-in-progress the whole time Mar 11 16:58:10 even my cold is gone Mar 11 16:58:16 * CM liked Bad Gastein and Salzburg in the spring Mar 11 16:58:19 looking forward to start working on OM next week again Mar 11 16:58:22 mickeyl: wow, that sounds nice. Mar 11 16:58:41 my wife and the guys had a hard time with the snow though Mar 11 16:58:44 CM - I was in Hall and Vienna. :-) Mar 11 16:58:44 it was hardly enough Mar 11 16:58:51 heh Mar 11 16:59:07 are they working a bit then yet? Mar 11 16:59:07 * stefan_schmidt is sad that he was not snowboarding the whole winter. Mar 11 16:59:10 and finally on the day we left (yesterday) it started snowing like hell Mar 11 16:59:16 s/hell/winter/ Mar 11 16:59:17 mickeyl meant: and finally on the day we left (yesterday) it started snowing like winter Mar 11 16:59:19 edistar: Some yes, some no. Mar 11 16:59:19 :) Mar 11 16:59:36 Elrond: Was in Wien too, nice town and I really liked the Natural history museum :) Mar 11 16:59:37 where can I see some status? on a blog or so? Mar 11 16:59:37 * mickeyl was in Salzburg on Thursday Mar 11 17:00:22 CM - I was in Wien three times now. Didn't get to the museums yet. ;o) Mar 11 17:01:02 mickeyl: Hi Mar 11 17:01:21 cheers alphaone. got the package oK? Mar 11 17:01:33 mickeyl: Yeah, thanks! Mar 11 17:01:36 edistar: What kind of status you mean? This functions is working and this one not? Mar 11 17:01:49 yes Mar 11 17:01:53 mickeyl: There was one Free&Easy SIM card in it Mar 11 17:02:09 argh Mar 11 17:02:16 heh Mar 11 17:02:25 in which one? Mar 11 17:02:46 I think the E680 Mar 11 17:02:59 ~lart mickeyl Mar 11 17:03:00 * apt installs PocketPC on mickeyl's PDA Mar 11 17:03:15 i guess that one has been disabled anyways Mar 11 17:03:25 ~lart eplus Mar 11 17:03:25 * apt DoSes eplus Mar 11 17:03:34 mickeyl: You want me to send it back to you? Mar 11 17:03:48 alphaone: let us compare the numbers please Mar 11 17:03:54 Okay Mar 11 17:04:09 can you give me the SIM serial? Mar 11 17:05:03 edistar: That would mean spending much time on documentation instead of coding. This is fine for a later phase of the project, but not yet. So I think we don't have the kind of info you're searching. Sorry. Mar 11 17:05:21 no problem, just asking Mar 11 17:05:33 some kind of CVS maybe? Mar 11 17:05:47 edistar: of course. It's all in svn.openmoko.org Mar 11 17:05:53 root@fic-gta01:~$ cu -l /dev/ttySAC0 Mar 11 17:05:53 Connected. Mar 11 17:05:53 AT-Command Interpreter ready Mar 11 17:06:02 just starting to get involved with linux in this way now, so it's all quite new for me, sry if I ask stupid things Mar 11 17:06:04 CoreDump|home: So yours is working Mar 11 17:06:24 I managed to make real phone calls with the Neo Mar 11 17:06:52 stefan_schmidt: same as before, http://en.pastebin.ca/390710 and then silence. Mar 11 17:06:56 edistar: No question is stupid. But people like to see you try to inform yourself with FAQ and some search about it before asking. Mar 11 17:07:26 pH5: damn Mar 11 17:07:34 mickeyl - You're in FFM sometime I heard? Mar 11 17:07:39 wasn't there a wiki page explainng Mar 11 17:07:39 pH5: Please open a bug about it. I'll comment Mar 11 17:07:48 explaining the modem commands? Mar 11 17:07:54 stefan_schmidt: will do. btw, is gsmd running on ezx already? Mar 11 17:08:01 CoreDump|home - Search for "AT" Mar 11 17:08:16 CoreDump|home: Yeah, but the problem is GSM don't even start to say it'S ready. Mar 11 17:08:30 ok, thanks Mar 11 17:08:35 pH5: No, not yet. Just not enough time. :( Mar 11 17:08:55 Elrond: I'm trying Mar 11 17:09:17 pH5: If I get a GSoC fond for the port I work on it at least in the summer. Hopefully before. Mar 11 17:09:18 ok got it Mar 11 17:09:59 why do you need Google to work on open source projects? Mar 11 17:10:19 CoreDump|home - I assume you already know http://wiki.openmoko.org/wiki/Manually_using_GSM Mar 11 17:10:24 CoreDump|home: Standard AT command set Mar 11 17:10:49 Elrond: thanks! Mar 11 17:10:51 CoreDump|home - http://wiki.openmoko.org/wiki/Hardware:AT_Commands Mar 11 17:11:18 loufoque: I work on it anyway, since more then a year now, but getting money for work is nice as I don't have to care earning money other ways for food and living Mar 11 17:11:22 (don't ask me, why Search for "AT" does not show this page. Bug in the wiki probably.) Mar 11 17:13:50 wb mickeyl Mar 11 17:18:22 probably because "AT" is too short of a word and would return too many results. ("at" is quite a common word in english) Mar 11 17:18:30 i don't think it is case sensitive is it? Mar 11 17:18:45 A list of AT commands: Go to http://pda.etsi.org/pda/queryform.asp and search for GSM 07.07 Mar 11 17:20:20 or were you being facetious Elrond? Mar 11 17:22:47 stefan_schmidt: yeah, that'd be great! I filed bug #256. Mar 11 17:24:36 SpeedEvil - You there? Mar 11 17:25:38 pH5: thanks Mar 11 17:34:24 Did anyone get the gps chip to talk any single byte? Mar 11 17:40:34 not me yet Mar 11 17:41:46 Hmmm, so it needs some init sequence. :-| Mar 11 17:42:21 Which is good in some ways. It potentially means more accurate initial position on switch-on. Mar 11 17:46:40 what are the chances that it's even powered up? Mar 11 17:47:01 There is a power switch Mar 11 17:47:06 re Mar 11 17:47:22 It's probably in low-power mode till you wake it somehow though. Mar 11 17:47:37 it needs to be enabled first through the pcf50606 and then wake up Mar 11 17:47:39 afaik Mar 11 17:47:54 the power management sequences are documented Mar 11 17:47:56 then maybe inited Mar 11 17:50:01 The power-init-sequence is in the kernel patches. Mar 11 17:50:27 Except for the reset-pin issue, where I'm still waiting for comments. Mar 11 17:50:27 IIRC, a few days ago koen did that, and it did nothing. Mar 11 17:52:09 Right. Mar 11 17:53:02 SpeedEvil - BTW: Do I read http://wiki.openmoko.org/wiki/Neo1973_Audio_Subsystem#using_Bluetooth_headset right? The BT module has a connection to the WM*? Mar 11 17:53:17 AIUI, yes, it's been stated that it does. Mar 11 17:53:21 it datasheet says "When the UART is in SLEEP mode then it will ignore any bytes transmitted by the host except for INTACK (interrupt acknowledge - 0xFA) and WAKE (0xFB)." Mar 11 17:53:31 http://wiki.openmoko.org/wiki/Neo1973_Hardware#Bluetooth does not mention that connection. Mar 11 17:53:35 BT audio doesn't have to go through the linux but. Mar 11 17:53:42 bit Mar 11 17:54:02 I mean that the CPU doesn't need to deal with it if it's setup, and the mixers twiddled right. Mar 11 17:54:36 Yeah, that's my understanding too. Just think, that the later URL misses that info. Should I add it? Mar 11 17:55:14 Sounds reasonable. Mar 11 17:55:29 I don't know exactly how the hardware/software stack in BT breaks down though. Mar 11 17:55:53 It'd be nice perhaps to be able to run two or more headsets at once in some circumstances. Mar 11 17:56:39 http://wiki.openmoko.org/wiki/Neo1973_Hardware#Bluetooth also suggests that connection. Mar 11 17:57:08 I think it was discussed here, I forget. Mar 11 17:57:35 I'd grep the logs for grep -i -10 bluetooth * |grep -2 direct Mar 11 17:58:14 c Mar 11 17:58:20 urrently don't have logs online Mar 11 18:00:06 I'm already doing that grep. Mar 11 18:00:36 so, according to the a-gps docs, if you send a 0xfa, it should always answer with a 0x7a or 0x7b, if it's operational Mar 11 18:03:22 SpeedEvil - Yeah. You asked about using the neo as intercom base station. No result. ;o) Mar 11 18:07:47 what a-gps docs? Mar 11 18:08:22 It's got other apps than intercom - handing off calls to other people closeby. Mar 11 18:09:53 "Product Specification" Mar 11 18:11:05 hummm.. so it cannot be distributed, looking at the first page Mar 11 18:11:09 i didn't notice that Mar 11 18:11:10 Sigh. Mar 11 18:11:14 I suspected so. Mar 11 18:11:34 it still doesn't have any useful information really Mar 11 18:12:05 Does it go into 'these are the per-satellite data blocks, and here is the format' stuff? Mar 11 18:13:38 no, it goes into: clock signal has to be connected here and uart here and has to be soldered in not more than X centigrades Mar 11 18:13:59 Ah - it's just the electrical, rather than the interface specs then. Mar 11 18:14:20 yeah Mar 11 18:14:26 it gives timint for the uart, the i2c, the other pins, too Mar 11 18:14:29 timings* Mar 11 18:15:20 What's the uart speed - perhaps it was being listened to at the wrong speed by Koen? Is it RS232? Or some other sort of serial. Mar 11 18:15:22 Oh well. Mar 11 18:15:34 I suppose all this will come out eventually. Mar 11 18:16:40 also lots of charts about signal conditions for the satellite signals Mar 11 18:16:43 lemme see Mar 11 18:18:08 "NCO-based baud rate generator that aneables generation of any baud rate from 1/4095th up to 1/16.5nd of the supplied reference clock frequency", then it says "Automatic baud rate matching after reset" Mar 11 18:18:23 fixed 8N1 operation Mar 11 18:18:37 Right - that sounds like it wants you to output some sort of sync sequence to it. Mar 11 18:18:55 which would make sense with the '0xFA' perhaps. Mar 11 18:19:09 After the Hammerhead is brought out of reset, the baud rate generator requires a series of synchronization (0x80) characters to be sent from the host to determine the baud rate" Mar 11 18:19:25 0xFA is to query for interrupt status Mar 11 18:19:28 That's interesting. Mar 11 18:19:48 Does it imply you get a stream of chars out, or that it's a protocol? Mar 11 18:20:21 are bluetooth enabled light switches? (for 230V lights) Mar 11 18:20:28 *there Mar 11 18:20:54 i don't see anything that would indicate either stream of chars or anything else Mar 11 18:21:54 K, thanks. It's useful that someone with hardware can potentially now try sending lots of 0x80's at it, and see if it wakes up in a useful state. Mar 11 18:22:09 buz: ebay search for bluetooth and x10 together shows nothing. Mar 11 18:22:26 i dont want no x10. just plain switch. Mar 11 18:22:30 but i cant find any Mar 11 18:22:51 SpeedEvil: it mentions "packets" Mar 11 18:24:01 http://209.85.135.104/search?q=cache:Ri8QM1wcK1gJ:www.ezurio.com/dl/%3Fid%3D30+bluetooth+light+switch&hl=en&ct=clnk&cd=12&ie=UTF-8 make your own i guess Mar 11 18:24:49 There would need to be some sort of packet format - but is it two dir or one dir... Need to get someone attacking it with minicom to see if it responds I guess. Mar 11 18:27:10 i think there are basically two forms of uart communication, one is the unknown protocol that has packets and perhaps commands and responses, and the couple of (like five) special values that are for setting up the uart and sleep states Mar 11 18:27:53 the only concrete values in the whole pdf are 0x80, 0xfa, 0xfb, 0x7a, 0x7b Mar 11 18:28:15 Sigh. Sounds like 'we' really want the other one. Mar 11 18:28:31 If they even distribute that. Mar 11 18:29:03 yeah, but the 0xfa -> 0x7a/0x7b sequence should be enough to get an idea if the chip is alive Mar 11 18:29:11 Probably, yes. Mar 11 18:29:23 And with luck, it'll start spewing out a stream of data. Mar 11 18:29:46 Which varies when it's taken out of a microwave oven. (shielded box) Mar 11 18:30:42 buz: the standard being pushed for that sort of stuff is 'ZigBee'. Mar 11 18:30:49 i know Mar 11 18:30:55 but no zigbee on neo Mar 11 18:31:20 using a gumstix to do the job is "slight" overkil Mar 11 18:32:13 As an example of its current popularity, as I see it, there is one zigbee device on ebay at the moment. Mar 11 18:33:14 exactly, yes Mar 11 18:33:21 but 100 million bluetooth phones Mar 11 18:34:12 There are bluetooth -> motor devices - but they are not packaged as light switches. Mar 11 18:35:56 a bluetooth serial thingy that can drive serial without any further feedback would be good enough i guess Mar 11 18:36:26 (thetoy.co.uk) (nsfw) Mar 11 18:36:48 i laughed about that one numerous times Mar 11 18:36:54 i wonder if it actually sells Mar 11 19:09:14 Ok, anyone know how to switch uboot into dfu mode? Mar 11 19:11:50 noidd: dfu-util does that for your Mar 11 19:15:24 I did a dfu-util --list and I only see one entry Mar 11 19:15:56 noidd: --list only works after you have flashed something and didn't use -R(eset) Mar 11 19:16:08 what do you want to flash? Mar 11 19:16:18 my new rootfs image Mar 11 19:16:23 jffs2 Mar 11 19:16:35 ./dfu-util_static -a5 -R -D openmoko-devel-image-fic-gta01.jffs2 Mar 11 19:18:27 awesome, she's flashing Mar 11 19:18:29 thanks Mar 11 19:19:02 noidd: Do you have a current uboot for that? Mar 11 19:19:09 heya, everyone. Mar 11 19:19:27 hi pavelm Mar 11 19:19:42 noidd: np ;) Mar 11 19:20:35 alphaone: you mean uImage? Mar 11 19:20:55 CoreDump|home: No. Mar 11 19:21:00 Does anyone have usable dialing/call answering/get-me-gprs connection/ read email code? Mar 11 19:21:15 I'd like to help with testing ;-). Mar 11 19:21:21 reading email is probably a lot easier than making phone calls right now Mar 11 19:21:40 pavelm: ahoj :-) Mar 11 19:21:45 calling will be like an extra feature in the future ;) Mar 11 19:21:47 * buz begins to think that maybe the easiest way to get a bluetooth module is to simply take a cheap phone Mar 11 19:22:09 Last night I tried to flash a rootfs with the old version of uboot and bug 245 messed it up Mar 11 19:22:22 Actually. Mar 11 19:22:30 Bluetooth modules are pretty cheap. Mar 11 19:22:37 now you've made me nervous Mar 11 19:22:40 ahoj, mikesh. Does it mean we can speak czech now? Mar 11 19:22:49 alphaone: really? I flashed one rootfs w/ original u-boot w/o problems Mar 11 19:22:59 alphaone: I'm flashing my moko with the default uboot - should I have upgraded that first? Mar 11 19:23:00 SpeedEvil: yes, usb ones Mar 11 19:23:08 CoreDump|home: strange Mar 11 19:23:14 pavelm: pro jistotu #openmoko-cz Mar 11 19:23:15 http://www.sparkfun.com/commerce/advanced_search_result.php?keywords=bluetooth&x=0&y=0 Mar 11 19:23:16 maybe I was lucky =) Mar 11 19:23:17 then again, old phones are basically free Mar 11 19:23:18 noid: making calls should be easy; set up mixer, ATD and there you go Mar 11 19:23:27 noidd: You can always upgrade it later if you run into problems Mar 11 19:23:34 old phone ones have no docs. Mar 11 19:23:39 And are usually hard to attach. Mar 11 19:23:52 be advised that I could _not_ update u-boot with the original u-boot installed Mar 11 19:24:07 Do not touch u-boot unless you have JTAG access :\ Mar 11 19:24:13 SpeedEvil: after dfu-util just reset ? Mar 11 19:24:20 CoreDump|home: I was able to do that :-) Mar 11 19:24:21 Does anyone have scripts for getting gprs connection to work? Mar 11 19:24:39 alphaone: hehe Mar 11 19:24:40 ...I can probably add pin entering into scripts from my computer, but... Mar 11 19:25:01 noidd: yes Mar 11 19:26:05 pages of jffs2 errors about directories being hard links. Looks like I need to re-check my build env Mar 11 19:26:29 LaF0rge: The 200$ Hackers Lunchbox should contain a second USB cable for the debug board ;) Mar 11 19:26:44 noidd: you didn't read the errata =) Mar 11 19:27:00 you have to erase the rootfs in u-boot first Mar 11 19:27:49 haha Mar 11 19:27:50 awesome Mar 11 19:27:50 pavelm: I _think_ the dialer is first priority, after comes GPRS. Mar 11 19:27:56 Just from what I've read. Mar 11 19:28:32 CoreDump: if I was errata where would I live? Mar 11 19:28:57 speed: Yes, I guess so. GPRS should be simple enough to get running with pppd & chat.. without special software. I guess I should just do that. Mar 11 19:29:07 speed: ...then, mail application will be needed. Mar 11 19:29:17 Yeah. Mar 11 19:29:28 oh yeah, in the errata page on the wiki Mar 11 19:29:42 Unfortunately - everything is _horribly_ uncoordinated at the moment. Mar 11 19:30:15 The main push is - as I understand it - to get the thing sort of working as a phone, by the time phase 1 hits. (hardware to all with $350 in some 3 weeksish) Mar 11 19:30:24 noidd: http://wiki.openmoko.org/wiki/Neo1973_Phase_0#Errata Mar 11 19:31:22 BTW what is that GPE phone edition project? They seem to have dialer. Anyone tried it? Mar 11 19:31:54 maybe i could get neo sbcs as controllers ;) Mar 11 19:32:02 can even drop the gsm Mar 11 19:32:11 mhh thinking about it, that might actually be a good idea for fic Mar 11 19:32:29 sell the thing as sbc, a lot of infrastructure will be around for it soon Mar 11 19:32:41 ... I was playing with gomunicator at some point, and it had working dialer, too. Mar 11 19:34:14 The idea is to get a working dialer that is actually native, and will intgrate with the rest of the software nicely in a coherent manner once it's written, rather than jamming a random dialer in AIUI. Mar 11 19:34:37 I'm yet to get a system capable of running a build tree though at the moment. Mar 11 19:34:48 having assorted anoying problems. Mar 11 19:35:13 speed: Yep, I understand. But if anyone has binary version of working dialer (I don't care about integration ;-), I'll be happy to play with it. Mar 11 19:35:26 Yeah - :) Mar 11 19:35:46 haha, hitting return again repeats a command... Mar 11 19:35:50 so, nand erase... Mar 11 19:35:53 dfu-util Mar 11 19:35:54 I've considered writing a trivial ncursed GUI for the phone. Mar 11 19:35:59 cu -l Mar 11 19:36:03 noidd: yeah :\ Mar 11 19:36:04 and it erased what I flashed Mar 11 19:36:08 that's hillarious Mar 11 19:36:22 speed: I _did_ write trivial (almost ncursed) GUI for phone. tui.sf.net. Mar 11 19:36:23 happened to me as well.....twice.... =) Mar 11 19:36:27 I'm guessing i'm not the first person to do that Mar 11 19:36:33 noidd: You need a new uboot Mar 11 19:36:42 That were the same errors I had Mar 11 19:36:44 speed: It assumes keyboard at the moment... Mar 11 19:36:47 alphaone: is that wise without a debug board? Mar 11 19:36:58 ncurses + xterm 'mouse support'. Mar 11 19:37:10 Insane though. Mar 11 19:37:12 noidd: Hmm, maybe not :-/ Mar 11 19:37:13 noidd: I flashed u-boot it two times with dfu-util, no problems. Mar 11 19:37:17 s/it// Mar 11 19:37:17 pH5 meant: noidd: I flashed u-boot two times with dfu-util, no problems. Mar 11 19:37:20 speed: yes. I did not get to using that mouse support _yet_. Mar 11 19:37:32 noidd: Make sure it doesn't power off Mar 11 19:37:34 speed: Insane -- maybe, but if it makes the hw usable for now.... :-) Mar 11 19:38:00 speed: and it is not lost work because it will likely be useful on sx1 Mar 11 19:38:06 I'd be more motivated if I had hardware to do silly things like that - at the moment I'm putting the walls back on my house. Mar 11 19:38:14 (http://www.mauve.plus.com/walls.html ) Mar 11 19:39:51 I'm kinda scared to do it with a binary version I compiled tho Mar 11 19:45:44 hey koen Mar 11 19:46:09 will battery charge when I unplug it, plug it, then plug in charger? Mar 11 19:46:15 ...or should I boot linux? Mar 11 19:46:22 ...seems that linux will not boot :(. Mar 11 19:46:47 good morning all Mar 11 19:47:04 fic is the ODM themselves right? Mar 11 19:47:06 neo1973 == instant rockstar status over here Mar 11 19:47:15 the neo screen is too small: http://www.flickr.com/photos/7337648@N08/417853278/ Mar 11 19:48:26 buz: afaik, yes Mar 11 19:49:46 woah neat, civ on a phone Mar 11 19:49:47 pavelm: i don't know but you can enable/disable charging from u-boot too Mar 11 19:51:20 balrog: it seems it gathered enough energy to boot linux, good. Mar 11 19:51:55 pH5: Damn Mar 11 19:53:05 koen|away: did you see the earlier GPS conversation? Mar 11 19:53:58 SpeedEvil: nope, too busy sweating and eating Mar 11 19:54:02 I.E. waking up the GPS, then sending 0x80's (quite a lot of) syncs the UART on the GPS to whatever speed the UART on the SOC is set to. Mar 11 20:02:02 gtg cya Mar 11 20:17:06 counter Mar 11 20:17:07 http://wiki.openmoko.org/wiki/Wishlist:Neo1973_P0_Review P1: 2 weeks, 6 days 15:42:53 (20.655 +-7.0 days) (see faq) (521) Mar 11 20:30:34 nice pH5 Mar 11 20:30:44 is that an exported display or running natively there Mar 11 20:31:06 faq Mar 11 20:31:06 See http://wiki.openmoko.org/wiki/FAQ Mar 11 20:40:18 noidd: natively, of course :) I needed something to test my cross toolchain. Mar 11 20:40:28 hehe Mar 11 20:41:35 Any idea how long phase0 openmoko survives in standby? Mar 11 20:41:36 pavelm-away: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Mar 11 20:42:13 Is there way to get bigger font in rxvt? Mar 11 20:42:26 pavelm: no idea. is there any way to wake it up at all without the debug board? Mar 11 20:43:21 (rxvt --font 10x20 works on PC but not on moko; I'd like to get usable size of test) Mar 11 20:43:35 ph5: 911 button wakes it just fine Mar 11 20:44:02 ph5: ..but I meant "how long does it last, running but not doing anything, backlight off" Mar 11 20:47:21 pavelm: ouch, I didn't even try pressing the aux button. Mar 11 20:47:24 but I don't know the answer to that either. Mar 11 20:48:02 ph5: just trying :-). Mar 11 20:48:14 hm... I get s3c2410-i2c errors during resume Mar 11 20:48:18 ph5 : I guess I'll just test it, but it should take few hours :-). Mar 11 20:48:20 "cannot get bus (error -110) Mar 11 20:48:38 ph5: it seemed to more or less work for me. reboot & retry ;-). Mar 11 20:50:43 pavelm: the current version has _major_ hardware bugs with regards to power consumption. Mar 11 20:51:04 However, most of these are not really present when it's running at normal clock speed. Mar 11 20:51:28 With the RAM and LCD idle, and the backlight off, and nothing else on, you're looking at around 150mW Mar 11 20:51:56 Some 24 hours Mar 11 20:53:15 This is with everything other than the CPU off, and the CPU idle, without using much bus traffic. Mar 11 20:54:33 speed: 24hours standby... is pretty much okay, for my use anyway. Mar 11 20:54:35 speed: Good. Mar 11 20:54:47 This is for the numbers for final hardware though. Mar 11 20:55:08 http://wiki.openmoko.org/wiki/Neo1973_Power_Management has some guesstimates at the bottom. Mar 11 21:01:13 ..yep; I was interested for numbers for phase0 hw. Mar 11 21:01:51 As I understand it. For idle/active mode, these should be about right. Mar 11 21:02:09 For 'slow' mode - they are around double the nominal power. Mar 11 21:02:19 Aha, good. So I'm looking at 24hours standby if I get software right. Mar 11 21:02:37 That's not _good_, but it is what I'm getting out of my nokia6230, anyway. Mar 11 21:02:44 And in suspend, it may be that some 'sensible' sources of interrupts can't wake it from suspend to RAM or disk. Mar 11 21:02:46 err Mar 11 21:02:49 SD Mar 11 21:03:06 Suspend to SD is a rare case though. Mar 11 21:03:30 speed: No, I do not think we want to suspend to SD on openmoko. Mar 11 21:03:45 speed: We'd prefer to still take the calls, right? Mar 11 21:03:53 Though it offers potentially interesting cases. For example, smash the screen, suspend to SD, swap SD, resume Mar 11 21:04:10 speed: and resume from disk takes too long ... Mar 11 21:04:20 speed: Interesting -- yes. Mar 11 21:04:33 RAM will be the vastly commoner and more important case. Mar 11 21:06:08 speed: Does suspend-to-RAM mean "CPU in slow mode" in openmoko case? Mar 11 21:06:18 speed: ...i.e. not really suspended but very very very slow? Mar 11 21:06:21 As I understand it, I don't know. Mar 11 21:06:36 It is possioble to put the RAM in 'self-refresh' rate. Mar 11 21:06:40 s/rate/mode Mar 11 21:06:51 this leads to a battery life of around a year. Mar 11 21:07:05 I don't know if the power system is capable of doing this though. Mar 11 21:07:41 And CPU very slow unfortunately doesn't take that much power. 12Mhz - the nominal slow mode - uses a third of that at 200Mhz. Mar 11 21:08:08 Hmm, third of 200MHz is still only 3days of standby. Mar 11 21:08:23 If it was possible to do suspend to RAM in this way, a battery life of over a week is easily possible, with GSM on and listening for calls. Mar 11 21:08:44 speed: li-ion is not going to last a year. Mar 11 21:09:09 Do you mean from self-discharge point of view? Mar 11 21:09:25 speed: yes. Mar 11 21:09:34 It mostly does. Mar 11 21:09:46 It loses 20-40% of energy depending on heat. Mar 11 21:10:01 And 10-15% of capacity. AIUI. Mar 11 21:11:10 Anyway - months, in deep suspend. Mar 11 21:11:24 speed: Ok, but that's w/o gsm running. Mar 11 21:11:46 Yes. it's around a week or so with GSM and nothing else on. Mar 11 21:11:47 speed: -> not too useful, except for alarm clock and calendar functions, where it is nice. Mar 11 21:12:01 speed: zauruses and so on "only" last about 10 days. Mar 11 21:12:04 Unless you can poll for missed calls every few mins. Mar 11 21:13:18 And still, few will want to do that. The important bit ios how much power the GSM modem uses. Mar 11 21:13:55 speed: agreed. Mar 11 21:14:09 I would guess that booting the GSM chip and connecting to the GSM network every few minutes draws much more energy than just leaving it running. But I can't test that as long as it doesn't work at all. Mar 11 21:14:41 I don't think so. Mar 11 21:14:47 ph5: plus you can't really poll for lost calls that wway. Mar 11 21:14:50 It uses _maximum_ 2W when transmitting. Mar 11 21:15:04 On some providers they text you saying 'missed call from foo' Mar 11 21:15:25 ph5 is right; registering to the network is normally done every 5 minutes Mar 11 21:15:44 ph5 so one power up of GSM is normally as costly as ~5 minutes of standby Mar 11 21:16:44 If it transmits for a second when logging in, then you're looking at 2Ws, if it uses 30mW, then you're winning if you're asleep for more than a minute. Mar 11 21:17:18 ph5 assuming strong signal and not moving receiver Mar 11 21:17:33 full power power is used for gsm network registration. Mar 11 21:17:53 Yes, but it does not take long. Mar 11 21:18:39 If you're transmitting for a second, that's 2 watt-seconds. If you use .03W/second in standby, then if you sleep for more than a minute, you get lower power. Though lose the ability to take calls. Mar 11 21:24:36 * SpeedEvil does some tests with his nk3330 Mar 11 21:26:17 Place it near speakers to hear some gsm ;-). Mar 11 21:30:30 I was actually going to hook it up to a PSU, in place of the battery, and integrate the energy. Alas, I've misplaced the power supply. Mar 11 21:31:52 ..so you have smoking piece of plastic instead of nk3330? Mar 11 21:32:39 No, I cannot find my power supply that has a nice meter on it, that I can use to get a good idea of the total startup energy of the phone. Mar 11 21:42:28 counter Mar 11 21:42:28 http://wiki.openmoko.org/wiki/Wishlist:Neo1973_P0_Review P1: 2 weeks, 6 days 14:17:32 (20.596 +-7.0 days) (see faq) (522) Mar 11 21:42:56 wtf is counter? Mar 11 21:43:39 Release date Mar 11 21:43:59 2 weeks 6 days to nominal release date for the semi-washed masses. Mar 11 21:47:31 speed: I'd not trust that at this point. sw is not going to be ready within 3 weeks. Mar 11 21:47:42 No. Mar 11 21:47:50 speed: not sure if even fixed hardware is going to be ready at that point. Mar 11 21:48:03 There may be a minimal dialer app working though. Mar 11 21:48:25 * LarstiQ can deal with the software. Mar 11 21:48:26 There has been no announcement on delays, and they've been quite good on announcing delays befdore. Mar 11 21:48:54 speed: ok, maybe. Mar 11 21:49:08 speed: I'm actually trying to get _minimal_ dialer working ;-))))). Mar 11 21:55:33 As long as the hw is correct I can't imagine that they'd delay it unless they have huge problems with uboot Mar 11 21:57:56 CM: I certainly hope so Mar 11 21:59:08 u-boot is the critical part. Mar 11 22:00:10 Personally. Mar 11 22:00:36 I'd be happy with working hardware, and software that does not have major bugs. Mar 11 22:00:36 they would need to cause regression to have huge problems with it now, as it's pertty functional at this point already Mar 11 22:00:48 (brick bugs that'd need JTAG) Mar 11 22:00:49 u-boot i mean Mar 11 22:01:17 Hello, magon. Seems there's big interest in openmoko in czech rep. Mar 11 22:01:19 I have interesting stuff to hack on the phone. Mar 11 22:01:30 have charging issues been resolved yet? Mar 11 22:01:33 For example, I'd like to play with alternate methods of text entry. Mar 11 22:02:47 As it is - it needs minimal software to do a front end for mplayer, ... Mar 11 22:04:36 Hmm, the phone is comparable to a Pentoum at 133 MHz? Mar 11 22:04:50 That's the sort-of comparison I've been using internally. Mar 11 22:04:53 I could even consider compiling stuff on it, without cross. ;-) Mar 11 22:05:13 It's fast enough that it's usable as a general purpose computer. Mar 11 22:05:22 Especially with the small screen. Mar 11 22:05:33 * balrog-kun built a working linux kernel on a 60MHz ARM :p Mar 11 22:05:53 "16Mhz 386" Mar 11 22:06:01 sx Mar 11 22:06:18 the screen was too small to run menuconfig (it requires 80x25 chars) but make config worked :p Mar 11 22:06:50 how long did it take? :D Mar 11 22:07:11 5+ hours.. Mar 11 22:07:50 ;o) Mar 11 22:09:22 SpeedEvil - The screen is actually my major problem. ;) From a technical point of view for normal humans, it's awesome. But developers love to use the highres.... and I can't read anything then. ;o) Mar 11 22:10:17 All you need is a monocular mount. Mar 11 22:10:28 that's what i was thinking Mar 11 22:10:29 A powerful lens, and the neo held 15cm from your eye. Mar 11 22:11:12 a VR helmet Mar 11 22:11:15 SpeedEvil - Guess what I wear round my neck. Mar 11 22:11:31 I have also wondered about a mode where the neo in most apps goes to 240*320, for power saving. Mar 11 22:11:34 SpeedEvil - a lanyard with a 8x20 monocular. Mar 11 22:11:51 RAM bandwidth costs power. Mar 11 22:13:05 Yeah, that mode would be nicer. :) Mar 11 22:14:44 elrond: actually the fonts are _way too small_ for a cellphone. Mar 11 22:14:57 i Mar 11 22:15:02 want the option. Mar 11 22:15:09 elrond; cellphones are used in environments with vibrations etc... you just can't use the small font there. Mar 11 22:15:22 speed: Yep, being able to select small font is nice when you are home. Mar 11 22:15:36 However, there needs to be a _really_ easy way to make it big. Mar 11 22:16:03 Ouhc.... openmoko refuses to launch binaries I compile :-((((((((((((((((. Mar 11 22:16:19 EABI vs. some other ABI?! Mar 11 22:16:51 pavelm - Yeah. I hope some developers recognize that. Mar 11 22:17:27 elrond: I believe they will. Mar 11 22:18:36 elrond: Or I'll take them to the horseback riding, go trot, and ask them to do anything with their shiny new toy. Mar 11 22:18:57 ...is there easy way to get compiler that produces working openmoko binaries? Mar 11 22:24:13 pavelm - MokoMakefile is one. openembeded is assumed to be the other. Mar 11 22:24:45 scratchbox, crossdev, crosstoolchain are more ways to produce the toolchain Mar 11 22:27:35 Will /scratchbox/compilers/arm-gcc-3.3.4-glibc-2.3.2/arm-linux/bin/gcc do the trick? Mar 11 22:28:04 ...no, because that's what I'm currently using. Too bad. Mar 11 22:28:14 Ok, enough for today, have to get some sleep. Mar 11 22:28:15 good night. Mar 11 22:28:27 has anyone tried to build openmoko-devel-image for running with Xoo ? Mar 11 22:29:29 i.e. using MACHINE="x86" Mar 11 22:30:11 Elrond: MokoMakefile is just a wrapper around OpenEmbedded - they are *not* two alternative build systems Mar 11 22:30:29 Night Paavo Mar 11 22:30:41 rwhitby - :-)) Mar 11 22:38:49 rwhitby: what does "openmoko-devel-tools" (cf. MokoMakefile) contain? Mar 11 22:39:03 devirginator.. and ? Mar 11 22:42:57 splash, perhaps Mar 11 22:42:58 nm. i see now it's the dfu-util-native, ftdi-eeprom-native and openocd-native. handy :-) Mar 11 22:44:02 and anything else we need on the host will be added Mar 11 22:44:34 great. Mar 11 22:51:40 Have any white/orange P0 phones been seen? Mar 11 22:53:58 i've seen one Mar 11 22:54:01 the one in OE booth at FOSDEM07 was white/orange, i don't know if it was P0 Mar 11 22:54:18 oh, yeah, it might've been pre-P0 Mar 11 22:55:41 Probably pre-P0 Mar 11 22:56:54 it was pre-p0 Mar 11 22:57:08 I would bet it was mickeys neo =) Mar 11 22:57:15 hehe. ;) Mar 11 23:30:59 wansti, this one is for you: http://sicherheitsschwankung.de/gallery/v/openmoko/IMG_8413.JPG.html Mar 11 23:33:41 alphaone: how's the game controlled on the neo? Mar 11 23:34:01 aevin: unfortunately not at all atm :-/ Mar 11 23:34:14 But it's great that it works ootb Mar 11 23:34:21 just bitbake supertux Mar 11 23:36:04 And dound does not work either Mar 11 23:36:26 okay. Mar 11 23:37:20 alphaone: so is there anything special you're working on with the OpenMoko @ Neo? Mar 11 23:37:40 a particular area of focus.. Mar 11 23:39:58 aevin: Just random bugsquashing at the moment. Mar 11 23:40:32 okay. i guess there are plenty of those around :) Mar 11 23:41:01 I'll try to tackle some of the application segfaults. Mar 11 23:41:19 the openmoko-* apps? Mar 11 23:41:22 yeah Mar 11 23:42:05 bug 245 was annoying me but luckily that got fixed yesterday where I wasn't able to do anything openmoko related Mar 11 23:43:37 Hi wansti :-) Mar 11 23:43:52 hey Elrond Mar 11 23:44:52 alphaone - Have you tried to power-on the gps yet? Mar 11 23:47:56 Elrond: Well, I echoed power_up into /sys/bus/platform/devices/gta01-pm-gps.0/power_sequence Mar 11 23:48:37 now connect to the GPS serial port, and send (at 8N1) a lot of 0x80's Mar 11 23:48:50 This trains the uart on the GPS to the right speed Mar 11 23:49:06 minicom should work Mar 11 23:50:18 How does one get 0x80 into minicom? Mar 11 23:50:47 copy paste? Mar 11 23:51:11 Hmm, might work. Mar 11 23:52:10 "Congratulations. Openmoko has better version-control then Microsoft" =) Mar 11 23:52:22 ctrl-a s then 'ascii' may work Mar 11 23:52:24 * LuitvD laughs microsoft in the face Mar 11 23:52:24 with a file Mar 11 23:52:56 how's development on the project with some P0 developers? Mar 11 23:53:24 LuitvD: where's that quote from? Mar 11 23:53:45 rwhitby: just thought of it reading THIS: http://opinion.zdnet.co.uk/leader/0,1000002208,39286244,00.htm Mar 11 23:54:31 rwhitby: practically every subversion-based project has better version/revision control then Microsoft's OneCare Mar 11 23:54:47 fixing a bug in 0.5, and letting it reappear in 1.5 XD Mar 11 23:54:58 *G* Mar 11 23:54:59 bloody stupid, that's what I think it is Mar 11 23:56:52 * LuitvD silently hopes for microsoft to go on like this, and completely fall apart Mar 11 23:57:38 LuitvD - windows customers want it that way. ;) Mar 11 23:57:40 (or for microsoft's customers to stop being ignorant and blind) Mar 11 23:58:00 LuitvD: honestly, I wouldn't think this happened because of an SCM Mar 11 23:58:06 Elrond: yeah, that's the part I find the hardest to understand Mar 11 23:58:20 SCM? Mar 11 23:58:41 scm/vcs/rcs Mar 11 23:59:27 LarstiQ: how do you explain this then? Mar 11 23:59:39 shear bad luck? Mar 11 23:59:57 bad bug fixing protocols. Mar 12 00:00:30 I think many things come down to bad communication... Mar 12 00:00:39 LuitvD: a bug can be reintroduced all kinds of ways, a testsuite is one way to catch regressions Mar 12 00:00:42 (good thing we have this channel here :) ) Mar 12 00:01:21 SpeedEvil: Like that? echo -ne "\x80" > /dev/ttySAC1 Mar 12 00:01:27 sure, having svn will prevent lots of reintroductions over just copying directories around :) Mar 12 00:01:39 It's supposed to need lots of 0x80's. Mar 12 00:01:48 100 should be safe. Mar 12 00:01:58 LuitvD: but I've seen the same thinko being made more than once in version controlled projects Mar 12 00:02:54 i think this OneCare-deletion-bug-reappearance is either because of using the wrong code (bad codebase management), or because of someone reintroducing a 'feature' who didn't know the dangers of (bad communication) Mar 12 00:02:58 or both :) Mar 12 00:03:47 guess both apply to microsoft because they try to manage every single aspect of an operating system, using a gigantic team of developers that don't really code for the things they want to code for Mar 12 00:04:24 SpeedEvil: Doesn't seem to affect it at all Mar 12 00:04:24 *G* Mar 12 00:04:35 alphaone - It wont answer. ;-) Mar 12 00:04:42 * LuitvD starts to run warm for his talk about FOSS in education Mar 12 00:04:49 alphaone - Now send a 0xfa und see, if it answers. :) Mar 12 00:05:29 Sigh Mar 12 00:06:20 or send as 'ascii' /dev/urandom Mar 12 00:06:24 alphaone - It should answer with a single byte. Mar 12 00:06:32 SpeedEvil - *G* Mar 12 00:06:41 Nope Mar 12 00:06:54 hmpf. Mar 12 00:08:06 :-/ Mar 12 00:08:33 alphaone - How have you done "send 0xfa, wait for one char"? Mar 12 00:12:08 Elrond: Yes Mar 12 00:12:26 echo > on one shell and cat on the other Mar 12 00:12:49 Ahh. :-) Mar 12 00:14:03 I never tried this echo/cat combination... Mar 12 00:17:58 Hmmm. Mar 12 00:17:58 Is cat buffering? Mar 12 00:19:24 dd if=/dev/ttysomething bs=1 Mar 12 00:19:25 ? Mar 12 00:20:03 Not to mention I haven't used stty to set 8n1 for ages. Mar 12 00:29:54 Elrond: That always worked for me.. Mar 12 00:30:08 What should I set the tty to? Mar 12 00:31:46 /dev/ttySAC1 ? Mar 12 00:33:08 SpeedEvil: What speed Mar 12 00:34:42 It's relatively unimportant. Mar 12 00:34:45 it autosyncs Mar 12 00:34:56 tot eh stream of 0x80's it's sent. Mar 12 00:35:27 well, I tried that without any success Mar 12 00:35:28 38400 should be safe. Mar 12 00:35:30 Well, the speed should be probably not less than 4800. Mar 12 00:35:56 Right, 38.4 sounds good. Mar 12 00:37:18 alphaone - And 8n1. Mar 12 00:38:06 yeah Mar 12 00:38:14 anyone here got an untouched /e/n/interfaces on his neo and could pastebin it? Mar 12 00:38:20 I'd try the /dev/urandom thing next if it doesn't. Mar 12 00:38:37 SpeedEvil: Just did that, too Mar 12 00:39:30 hmmm. Mar 12 00:40:29 I'd now try putting a multimeyter on the battery to read current iut - to see if gps was really on. Mar 12 00:40:36 but... Mar 12 00:42:15 Or put a dso probe at the gps chip. ;o) Mar 12 00:42:41 power is easier Mar 12 00:44:07 apropos: Did you find your PSU? Mar 12 00:47:20 all of this adds to my suspicion, that the reset pin on the gps is not set correctly in the power_up-cycle. Mar 12 01:01:59 Night people. Mar 12 01:02:19 | -- Mar 12 01:02:20 | Successfully terminated. Mar 12 01:02:20 | Collected errors: Mar 12 01:02:20 | ERROR: Cannot satisfy the following dependencies for task-openmoko-demo: Mar 12 01:02:20 | mplayer Mar 12 01:02:20 | -- Mar 12 01:02:38 Night Elrond Mar 12 01:02:46 Night Elrond Mar 12 01:03:14 I have never built successfully.... Mar 12 01:10:33 Readon: are you using MokoMakefile? Mar 12 01:12:02 aevin, yes Mar 12 01:12:57 is it a long time since yourlast source update, i.e. "make update"? Mar 12 01:13:41 i read something a couple of days ago that mplayer was taken out of the openmoko-devel-image build receipe. Mar 12 01:14:39 i updated it yesterday Mar 12 01:14:43 Readon: btw, what target machine are you building for? (in your build/conf/local.conf ) Mar 12 01:15:26 fic-gta01 i think Mar 12 01:15:33 okay. Mar 12 01:16:11 i think you should try an update and then rebuild. Mar 12 01:16:37 i just rebuild it totally yesterday. Mar 12 01:19:23 i have no idea, then. isn't there a log file with further error info? Mar 12 01:20:11 the mplayer one might be a real problem Mar 12 01:21:14 hmm - no, my rebuild overnight worked. Mar 12 01:21:35 Readon: please pastebin the output of "make update-makefile ; make update ; make setup ; make all" Mar 12 01:22:02 ok, i will do it in a minutes Mar 12 01:23:47 * rwhitby creates http://wiki.openmoko.org/wiki/MokoMakefile#Reporting_Problems Mar 12 02:01:27 i see someone said the word mplayer Mar 12 02:02:47 yes, we are Mar 12 02:03:02 | ERROR: Cannot satisfy the following dependencies for task-openmoko-demo: Mar 12 02:03:02 | mplayer **** ENDING LOGGING AT Mon Mar 12 02:59:57 2007