**** BEGIN LOGGING AT Sun Aug 12 02:59:57 2007 Aug 12 03:16:37 is there a way to force the Neo into sleep mode? Aug 12 03:16:58 apm -s Aug 12 03:17:44 thanks Aug 12 03:41:24 hey all Aug 12 03:41:48 when building autotools packages is it important that the autotools in the openembeded staging area be run and not the ones installed locally? Aug 12 03:43:08 orzo: i'm not much of an autotools kinda guy Aug 12 03:58:15 anyone tried the ubuntu openmoko VMware image? Aug 12 03:58:27 what do i do to start the emu? Aug 12 04:02:45 minthome, haven't seen that, where is it? Aug 12 04:04:27 on the mailing list Aug 12 04:04:59 http://easterniowashpe.org/openmoko/ Aug 12 04:05:19 or you could be nice and BT it http://tracker.unsads.com/torrents-details.php?id=3 Aug 12 04:05:30 u/p: openmoko Aug 12 04:06:08 using qemu, but i can't figure out how to start it Aug 12 04:10:50 it tells you how to start it when it finished compiling Aug 12 04:12:28 yep, on it... sorry for my ignorance Aug 12 04:12:29 http://wiki.openmoko.org/wiki/OpenMoko_under_QEMU Aug 12 04:16:19 you are doing it in linux, right? Aug 12 04:17:00 xamindar, yeah Aug 12 04:17:33 no actual gui emulation though?. all i get and qemu prompts Aug 12 04:18:15 after running flash.sh Aug 12 04:19:23 prompt GTA01Bv4 Aug 12 04:21:54 got it Aug 12 04:31:23 Does MokoMakefile not install everything that the official downloaded image does? Aug 12 04:31:34 i'm looking in my staging area and don't see libvorbis headers for some reason Aug 12 04:35:22 how does MokoMakefile decide which packages to build? Aug 12 04:35:40 are the packages listed somewhere? Aug 12 06:13:08 evening gentlemen :) Aug 12 06:13:21 i build an mpd pkg Aug 12 06:13:27 and its running on my phone heh Aug 12 06:13:31 mpd? Aug 12 06:13:37 music player daemon Aug 12 06:13:41 ah Aug 12 06:13:51 controlable by remote Aug 12 06:14:06 i dont have a client on the phone Aug 12 06:14:11 so i have to control it from the computer heh Aug 12 06:14:51 shouldn't it be the other way around? :) Aug 12 06:14:56 ya Aug 12 06:14:59 probably Aug 12 06:16:04 well i was thinking i'd build a client Aug 12 06:16:19 maybe cook up a new one Aug 12 06:29:39 ugh, was there a netsplit or did my host just choke? Aug 12 06:36:39 ugh Aug 12 07:21:09 anybody able to successfully sync up with the monotone repo? Aug 12 07:24:27 moin Aug 12 07:41:31 doc|home: you built OpenMoko from scratch? Aug 12 07:42:04 rtyler: nope, I'm merely a prospective owner Aug 12 07:42:20 haven't had time yet to play with an emulator Aug 12 07:42:26 damnit Aug 12 07:42:50 this fucking OpenEmbedded monotone distribution won't migrate properly on my ubuntu box Aug 12 08:15:15 re Aug 12 08:16:33 ahoy Aug 12 08:18:07 hi rtyler Aug 12 08:18:26 * hhf423 still browses the internets for a windows dfu util Aug 12 08:19:07 cygwin? Aug 12 08:20:28 you think so? Aug 12 08:20:54 hhf423: i saw something about a port in progress Aug 12 08:21:00 have you tried? :) Aug 12 08:21:15 rtyler: is the hardware access low level enuf? Aug 12 08:21:31 no, I did not even come accross it yet Aug 12 08:21:32 bahhh ;) Aug 12 08:22:52 is the flashing happening via serial over USB? Aug 12 08:23:29 I just want the frakking OE chain to build on my system, monotone versions seem to be incompatible Aug 12 08:23:48 yeah - the OE crap ... Aug 12 08:24:16 rxr: do you know what version of monotone they're using? I'd sure love to be able to build stuff for my phone now :/ Aug 12 08:24:21 ah, bah, I'll just do it via knoppix and be done with it. Aug 12 08:25:03 rtyler: no, I never got into the OE stuff Aug 12 08:25:16 in fact I just work on getting a Neo 1973 target into the T2 SDE Aug 12 08:25:32 because I find both, OE nor the Moko Gtk+ custom widget crap that appealing ... Aug 12 08:25:47 http://www.t2-project.org/targets/neo1973/ Aug 12 08:26:05 rtyler: http://www.openembedded.org/wiki/RequiredSoftware says recommended is at least 0.32. I'm using 0.35 with no problems since a while. Aug 12 08:26:41 i am building with 29 no problems so far.... still building though Aug 12 08:26:44 hrm, ubuntu gave me 0.31 from packages, damnit Aug 12 08:29:16 rxr: so is T2 actually got a "product" out yet? Aug 12 08:30:00 well more to the point can it boot on a neo Aug 12 08:30:25 xzcvczx: T2 is older than OE Aug 12 08:30:35 xzcvczx: there are more products and companies using T2 than OE Aug 12 08:30:53 including Puppy Linux since 2.10 and a swiss document archiving appliance Aug 12 08:31:28 xzcvczx: or do you mean for the Neo? Aug 12 08:31:57 rxr: for the neo Aug 12 08:32:00 xzcvczx: and T2 is not a company or so, it has no "products" it is an open source, GPL, community project Aug 12 08:32:16 rxr: yeah hence the "" Aug 12 08:32:20 xzcvczx: with even normal live and install CDs for x86, x86-64, PowerPC, SPARC etc. availabble :-) Aug 12 08:32:55 xzcvczx: we schedule to have a E17 based T2 target available that can be used to actaully make phonecalls by the end of this month Aug 12 08:33:24 hi all Aug 12 08:33:42 rxr: ah ok is there any screenshots of what you have currently? Aug 12 08:34:55 nope, as I said the phone just arrived this week, and we did not started with the Qemu emualator before because up to that point I considered the phone vapoware and did not wanted to started a target for a dead horse :-) Aug 12 08:35:19 xzcvczx: I'll post screenshots and make an announement when the T2/Neo1973 target is in end-user shape :-) Aug 12 08:36:57 * xkr47 reformatted the 512M flash to ext3 ^^ Aug 12 08:37:36 ext3 on microsd? that's gotta be slow.. Aug 12 08:37:42 is it ? Aug 12 08:37:46 what would be suggested then ? Aug 12 08:37:55 fat ?---) Aug 12 08:38:10 anything that's not journaled for starters, I just use ext2. Aug 12 08:38:19 thx Aug 12 08:38:37 heh no tune2fs binary.. reformat :) Aug 12 08:41:12 60kB/s goes onto the ext2 sd now over scp Aug 12 08:41:26 mmcqd eats 75% cpu, dropbear 15%, scp 4% Aug 12 08:46:23 Hello everyone! Can any1 help me? Where can i get Win32 pre-built binaries of openmoko? http://mdk.linux.org.tw/~jserv/openmoko/openmoko-emulator-win32-bin-20070625.zip not responding (( Aug 12 08:51:51 strange, now it started transferring files at 330kB/s Aug 12 08:52:04 mmcqd no longer eats any cpu, and dropbear is up to 60% Aug 12 08:52:17 the thingie now idles 30% Aug 12 08:53:24 the file isn't growing anymore :O Aug 12 08:54:22 ok errors in the dmesg Aug 12 08:54:31 pH5: using 0.36, looks like things are working better Aug 12 08:56:30 pasted here: http://pastebin.ca/654306 Aug 12 08:57:07 is this normal? any suggestions on how to proceed ? Aug 12 08:59:15 rxr: neo is arm920t not arm720t :) Aug 12 08:59:18 hmmm i really have to get a nicer build machine Aug 12 08:59:30 xzcvczx: for what specifically? Aug 12 08:59:45 rtyler: so it doesnt take so damn long to compile a devel-image Aug 12 09:00:05 ljp: it is ? than some openmoko.org page must be incorrect :-) as I got that information from there ... Aug 12 09:00:10 unfortunately I don't have my colo box up yet, or a port open to the corporate firewall for my dev workstation Aug 12 09:00:18 but dualcore opterons are so nice ;) Aug 12 09:00:23 root@fic-gta01:/media/card/bin$ cat /proc/cpuinfo Aug 12 09:00:23 Processor : ARM920T rev 0 (v4l) Aug 12 09:00:26 yes, indeed Aug 12 09:00:49 model name : AMD Athlon(tm) 64 X2 Dual Core Processor 3800+ Aug 12 09:00:51 * rtyler dances Aug 12 09:01:17 rxr: the GTA02 is going to be 400Mhz ARM right? Aug 12 09:01:19 * xzcvczx hacks into rtyler's box Aug 12 09:01:32 rtyler: no idea, I'm not a FIC person ... Aug 12 09:02:04 xzcvczx: I wish I could help you out really, we have plenty of hefty machines around here for servin' web pages ;) Aug 12 09:02:18 (see: quad dual core Opterons) Aug 12 09:02:33 xkr47: there are some reports somehwere on the wiki i thought of how you do get io errors doing too much io to the flash card Aug 12 09:03:25 rtyler: i am tempted to jsut go up to the compsci labs at my local uni boot all the ones that arent being used at the time into linux and just get some distcc or something happening Aug 12 09:03:27 :P Aug 12 09:03:35 thx xzcvczx Aug 12 09:03:35 xzcvczx: why not :D Aug 12 09:03:45 rtyler, quad xeons are even nicer Aug 12 09:03:54 Dmitry_Platonov: except for that whole "Xeon" bit :D Aug 12 09:04:08 with the exception of the Core 2 Duos, not a big intel fan Aug 12 09:04:14 although XScale is impressive Aug 12 09:04:33 rtyler: well the stupid thing is is that i am using a 1.3ghz for my build machine and i have a 1.7 duo which i cant get to work iwth distcc as its mac dammiy Aug 12 09:04:35 dammit Aug 12 09:04:45 xzcvczx: you can still use that IIRC Aug 12 09:04:52 just make sure your gcc versions match up Aug 12 09:04:57 rtyler: yeah its damn hard to work out how to do it thouhg] Aug 12 09:05:06 xzcvczx: true Aug 12 09:05:26 YAY! MokoMakefile processing... Aug 12 09:05:28 rtyler: the closest i ever saw was icecream and that wouldnt suit moko anyway as moko builds its own gcc Aug 12 09:05:59 xzcvczx: why not create a Knoppix image with the necessary tools and go use the labs? Aug 12 09:06:12 I would dig the image, I'd do the same thing aronud here ;) Aug 12 09:06:29 xzcvczx: that, or use gentoo, I hear it's OMG SUPAR FAST! Aug 12 09:07:00 rtyler: lol all my linux boxes that i have had at all really have been gentoo Aug 12 09:07:09 :) Aug 12 09:08:59 ljp: where did you spotted the arm710t? in the target sources ? Aug 12 09:10:09 FWIW, I posted some 2007.2 vs. 2007.1 photos after trying out some of the newer images: http://www.unethicalblogger.com/posts/r_tyler_ballance/openmoko_2007_2_preview Aug 12 09:10:29 rxr: running scripts/Config and selecting the neo project Aug 12 09:12:34 ljp: heh :-) the target was just started, do not expect it to yield a shiny E17 phone application .-) Aug 12 09:12:38 rtyler: thx :P Aug 12 09:12:56 its the system I am interested in Aug 12 09:13:03 2007.2 is *hot* Aug 12 09:13:10 still kind of broke though :P Aug 12 09:13:47 rtyler: yes, impressive what proken software is provided here after a year or so of work ... :-( Aug 12 09:14:35 rxr: I didn't mean it negatively, this is alpha software, I understand that :) Aug 12 09:15:13 * ljp has another gui for on top Aug 12 09:15:31 I do not beleive that "phone" part of neo would be consumer-ready before start of 2008 Aug 12 09:15:56 Basheri: I need to learn how to hold the phone steady in from of my Mac, the pictures are all a tad bit blurry :P Aug 12 09:15:57 hello Aug 12 09:18:04 SWEET! My first "you're a moron" blog response: http://itconservative.blogspot.com/2007/08/openmoko-experiences.html Aug 12 09:18:11 that guy really doesn't like Linux on your phone Aug 12 09:18:21 and doesn't seem to get that the software is pre-alpha Aug 12 09:20:45 all: why when I implement more code in panel-applet I start to gen G_MAXLONG ? Aug 12 09:20:47 rtyler: that guy looks like he's missing a couple of braincells ... actually make that big chunks of braincells. Aug 12 09:21:03 HERESY! YOU FREE SOFTWARE IDIOT OMGROFL Aug 12 09:21:05 * rtyler chuckles Aug 12 09:21:23 I think he's mistaking an open development process for "broken software" Aug 12 09:22:03 Anyway, I'd prefer GUI dialer & sound to work 8) Aug 12 09:22:15 well, look at the guy's profile: "free software is for poor people" Aug 12 09:22:19 * rtyler nods Aug 12 09:22:33 Dmitry_Platonov: the GUI dialer works for me Aug 12 09:22:35 ...and his taste in music. he's clearly lacking brain Aug 12 09:22:44 still not sure why my sound doth not work ;) Aug 12 09:22:44 rtyler, 2007.1 one? Aug 12 09:22:45 Anny idea? Aug 12 09:22:49 Dmitry_Platonov: both Aug 12 09:23:14 also, FWIW, I added all my photos to the OpenMoko flickr group (http://www.flickr.com/groups/openmoko/) Aug 12 09:23:20 * rtyler makes a note to keep adding them there Aug 12 09:27:09 who write's openmoko-panel? I need more allocated space in memory for plugin :...( Aug 12 09:28:46 yoyo: have you got like a fractal generator or somehting you want in the panel? Aug 12 09:28:50 yoyo: maybe you should cut your memory profile ;) Aug 12 09:28:51 hehehe Aug 12 09:29:01 xzcvczx: HEY! I like fractals! ;) Aug 12 09:29:08 * doc|home thinks they're silly Aug 12 09:31:21 rtyler: yeah but i also heard you like pong :P Aug 12 09:31:37 I'm writing a bluetooth plugin. yesterday when a had on/off option everything who's ok. now I add radio (public, not public, public for ~3min.) and I'w got G_MAXLONG Aug 12 09:32:21 creating filechooser makes the same problem Aug 12 09:37:02 fina-friggin-ly its built Aug 12 09:40:32 ljp: you said it's the system you are interested in, did you mean E17 or T2 or something else ? Aug 12 09:45:52 rxr: I am interested in T2 as build system, cause thats what I do Aug 12 09:46:52 ljp: for openmoko or other systems? did you know T2 before or is it new to you ? Aug 12 09:47:02 new to me Aug 12 09:47:09 (btw. the optimization setting was fixed in the target/neo1973/config.in, now :-) Aug 12 09:47:10 for other syste,s Aug 12 09:47:16 thanks for the pointer :-) Aug 12 09:47:22 no worries Aug 12 09:47:33 i like to see highly optimized things :) Aug 12 09:47:35 t2 has lots of docuemtnation, 100p++ handbook and such Aug 12 09:47:56 if you have trouble just ask on #t2 or the mailing-list :-) Aug 12 09:50:38 Any one who knows anything about the unavailable glibc-2.4 (the default replacement v2.6 failes to compile, and when forcing 2.5 using "PREFERRED_VERSION_glibc" and freinds the compile breaks due to a python exception inside glibc_2.5.bb) Aug 12 09:52:07 rxr: have you build an image for neo using uclibc? Aug 12 09:53:47 ljp: it's just on the way Aug 12 09:54:02 but so far I rsync over ssh the files on the /media/card of the Neo Aug 12 09:54:11 image creation comes next Aug 12 09:54:29 ljp: so far I still think about what init system to start with Aug 12 09:54:40 I do not exactly like the 2m boottime of the OpenMoko with sysvinit Aug 12 09:54:52 I used minit for minimal systems in the past Aug 12 09:55:11 but still hesitate to use it in combination with the UI parts ... Aug 12 09:55:15 busybox has init Aug 12 09:55:21 any votes? Aug 12 09:55:31 yeah - or just the busybox one ... Aug 12 09:56:55 well, if you are going to run bootchart, dont use the busybox init, as it wont work Aug 12 09:57:37 bootchart is? the OpenMoko progress screen ? Aug 12 09:57:47 or: Aug 12 09:57:52 http://www.bootchart.org/ Aug 12 09:57:57 yep Aug 12 10:00:00 Is anyone out there using either a hosted or their own local syncml server? I'm thinking of setting up funambol server on my own box but would take a less-effort route if one existed... Aug 12 10:09:27 small step :) Aug 12 10:16:20 jgm: Aug 12 10:16:22 ping Aug 12 10:16:43 jgm is alive (as much as you can expect for a Sunday morning, anyway) Aug 12 10:17:20 lol - just watched some crappy iphone 'speaker' thing on 'Something for the weekend' .... Aug 12 10:17:37 funambol has a simple 'all you need' install Aug 12 10:17:53 for either winders on linux Aug 12 10:17:54 http://www.funambol.com/opensource/downloads.html Aug 12 10:18:15 Yeah, except that I don't want to have to worry about maintaining the whole thing (I already run web services and email for people and keeping the infrastructure running eats enough of my time as it is) Aug 12 10:18:16 it's a 10-15 minute job. Aug 12 10:18:30 So I'll do it if I need to but if someone else has already done it and is selling accounts I'd buy one Aug 12 10:18:34 (And if not why not???) Aug 12 10:18:35 ahh I thought you were just going to experiment Aug 12 10:19:42 haa :) I'w got it :) Aug 12 10:19:49 I figure that if I sync all of my various contact databases to one source it'll be ready when I come to using my neo as my primary 'phone Aug 12 10:20:01 http://www.simchronise.com/ Aug 12 10:21:08 Looks interesting, will take a look. Thank Aug 12 10:21:11 s/Thank/Thanks/ Aug 12 10:21:11 jgm meant: Looks interesting, will take a look. Thanks Aug 12 10:21:53 http://www.fusionone.com/index.php Aug 12 10:21:58 also Aug 12 10:22:40 but it's much more fun running your own server :) Aug 12 10:46:47 is there an openmoko emulator or smth like that? Aug 12 10:47:01 i want to write some apps, but i don't have a neo phone Aug 12 10:47:36 Qemu :) Aug 12 10:47:57 http://wiki.openmoko.org/wiki/Running_OpenMoko_on_PC Aug 12 10:48:15 great, thanks :) Aug 12 10:48:20 hihi public not public :) Aug 12 10:48:26 SliMM: http://wiki.openmoko.org/wiki/OpenMoko_under_QEMU Aug 12 10:48:45 can it emulate the gps and other functions of neo? Aug 12 10:48:56 how on earth would it do that? :P Aug 12 10:48:57 just let the mokomakefile do the job Aug 12 10:49:40 :D Aug 12 10:50:55 rtyler: random satelite data :) Aug 12 10:51:13 oh, i'm running windows, is that ok? Aug 12 10:51:44 SliMM: connect to some random gpsd running somewhere? Aug 12 10:51:55 i have a debian virtual machine Aug 12 10:52:18 cjb_ie: no, just generate random or close-to-random gps data Aug 12 10:52:54 i'll use the virtual machine Aug 12 10:54:22 not to self, when running the MokoMakefile for the first time, get ready for a long wait, have other activities planned for about 40 minutes or so >_< Aug 12 10:54:26 i can probably dig up an nmea file a couple of hours long, if you were to pass that to gpsd one line at a time via a pipe maybe? Aug 12 10:56:10 i have arm-elf-gcc in my virtual machine Aug 12 10:56:37 is it good for compiling? Aug 12 10:59:33 rtyler: ha ha yeah for me i am more looking at ~ 24 hrs Aug 12 11:00:04 rwhitby: FYI: ETA on announcement + snapshot + (if you have a chance) MokoMakeFile changes = Aug. 19th Aug 12 11:00:10 and i'll be running it in VMware player :) Aug 12 11:00:15 omg it's mickeyl Aug 12 11:00:23 morning guys Aug 12 11:01:00 * mickeyl in a really good temper today Aug 12 11:01:08 hmm Aug 12 11:01:11 * SpeedEvil wonders if that's straw temper. Aug 12 11:01:13 good mood Aug 12 11:01:14 shall we ruin it? :) Aug 12 11:01:15 bad temper Aug 12 11:01:15 :) Aug 12 11:01:22 not good temper :) Aug 12 11:01:26 raster: ok, so much for dict.leo.org Aug 12 11:01:28 don't ask Aug 12 11:01:30 it's english Aug 12 11:01:40 i didnt' invent it Aug 12 11:01:46 mickeyl: if in doubt, say it in german and we'll translate appropriately ;) Aug 12 11:01:46 right Aug 12 11:01:47 my pc is a mess.. Aug 12 11:01:52 rtyler: :D Aug 12 11:01:54 ich kann es bloß Aug 12 11:01:59 hehe Aug 12 11:02:07 * cjb_ie has lived in an english-speaking country for over 25 years and is hopelessly confused by the above... Aug 12 11:02:08 ein bißchen sprecehn Aug 12 11:02:19 mickeyl: does that mean we will have openmoko-*2 in 1 week? Aug 12 11:02:36 in mokomakefile Aug 12 11:02:41 xzcvczx: Yes. Aug 12 11:02:45 sweet Aug 12 11:02:59 why does it have to be a sunday night :( Aug 12 11:03:01 :P Aug 12 11:03:16 might make it for Friday, lets see Aug 12 11:03:24 sweet Aug 12 11:03:27 mickeyl: is there a server running something like cruisecontrol or something to build regular snapshots? Aug 12 11:03:37 mickeyl: nah dont worry about me...... its just these damn timezones :S Aug 12 11:03:40 snapshots would be nice. Aug 12 11:04:00 rtyler: internally, yes. we're planning to make these snapshots available on a regular basis soon. Aug 12 11:04:18 thanks, I would like to fuck up my phone on a more regular basis ;) Aug 12 11:04:24 hehe Aug 12 11:04:28 excellent Aug 12 11:04:44 once the official P1 snapshot is out, we'll have a feed anyways Aug 12 11:04:46 or, danke, ich werde gerne mein Handy verbricken Aug 12 11:04:51 to make ipkg upgrade working Aug 12 11:04:57 hehe Aug 12 11:05:05 s/verbricken/in einen Backstein verwandeln/ Aug 12 11:05:24 Backstein verwandeln ? Aug 12 11:05:30 O_o Aug 12 11:05:40 ya, that's what we germans say for "bricking it" Aug 12 11:05:41 has anyone else had the issue with their phone that if the front cover is on properly it prevents the touchscreen working properly Aug 12 11:05:41 turn your phone into a brick wall O_O Aug 12 11:05:45 my german's a bit rusty... Aug 12 11:05:49 ah Aug 12 11:05:52 heh :-D Aug 12 11:06:04 they are encrypting their messages Aug 12 11:06:19 mickeyl: I think you should all start adopting the term "verbricken", mostly because that'll make me giggle everytime i see it ;) Aug 12 11:06:32 xzcvczx: darn. manufacturing problem. please remove the torx and the cover, and put it on again, that should help. might as well reseat the buttons while you're at it Aug 12 11:06:35 mickeyl: in english we just call it oh Aug 12 11:06:51 mickeyl: in english we just call it oh **** i just ****ed up my phone and dont have a debug board Aug 12 11:06:55 hehe Aug 12 11:07:01 debug boards are overrated Aug 12 11:07:02 * mickeyl runs Aug 12 11:07:12 yeah, I'm hoping I don't brick this, given the apparent shortage >_< Aug 12 11:07:40 mickeyl: i have done that a couple of times now and have now just resorted to leave the case not properly screwed in and sitting about a mm out from being seated properly Aug 12 11:07:48 common guys, it's not like a PSP, there are tools to unbrick your phones Aug 12 11:08:04 anybody using the 200070810 image from nslu2-linux.org? Aug 12 11:08:07 Markinoko: like what.... a sledgehammer? Aug 12 11:08:09 (http://ipkg.nslu2-linux.org/feeds/openmoko/images/OpenMoko-openmoko-devel-image-glibc-.dev-snapshot-20070810-fic-gta01.rootfs.jffs2) Aug 12 11:08:24 started building qemu for neo :) Aug 12 11:08:26 xzcvczx: well, that's one yes :-P Aug 12 11:08:34 or, a hefty bottle of vodka :) Aug 12 11:08:44 that'll solve all your Neo problems by presenting new ones ;) Aug 12 11:08:49 rtyler: planning on getting your phone drunk ? Aug 12 11:08:52 how long will it take? Aug 12 11:08:55 usr/bin/mplayer Aug 12 11:08:55 1015808 41% 293.66kB/s 0:00:04 Aug 12 11:08:58 rtyler: yeah then you can drink the vodka and smash the crap out of the phone with the bottle Aug 12 11:09:11 xzcvczx: ahhh, reminds me of moscow :D Aug 12 11:09:26 SliMM: on a good machine a couple of minutes Aug 12 11:09:26 * xzcvczx smacks up rtyler for putting that link in such an annoying way Aug 12 11:09:37 xzcvczx: whazzat? Aug 12 11:09:42 brackets Aug 12 11:09:46 oh, heh Aug 12 11:09:47 pointeless Aug 12 11:09:53 and break irc clients Aug 12 11:10:03 xzcvczx: no problems here with pidgin Aug 12 11:10:10 rtyler: in a virtual debian machine running on a dirty broken windows machine? :P Aug 12 11:10:16 Markinoko: well i am on mirc Aug 12 11:10:20 SliMM: twelve weeks Aug 12 11:10:22 hmmm thats a point does moko have irssi Aug 12 11:10:23 * rtyler chuckles Aug 12 11:10:30 dangit Aug 12 11:10:37 xzcvczx: iif you build it..they will come :) Aug 12 11:10:38 thanks :P Aug 12 11:10:48 xzcvczx: the answer is, openmoko has a cross compiler Aug 12 11:10:49 * rtyler is running irssi now, albeit on OpenBSD Aug 12 11:10:56 * ScaredyCat is the master of verbricken Aug 12 11:11:02 at least thaey have a real ps client in mickeyl's build Aug 12 11:11:08 s/client// Aug 12 11:11:10 xzcvczx meant: at least thaey have a real ps in mickeyl's build Aug 12 11:11:10 SliMM, you've bricked your neo? :) Aug 12 11:11:18 * mickeyl flashes a new uboot to his Bv Aug 12 11:11:19 3 Aug 12 11:11:26 # mplayer power.mp3 Aug 12 11:11:26 MPlayer dev-SVN-rUNKNOWN-4.2.1 (C) 2000-2007 MPlayer Team Aug 12 11:11:26 CPU: ARM Aug 12 11:11:34 i know it's a bit off-topic, but what does albeit mean? Aug 12 11:11:35 ^- T2/uClibC/Neo1973 ... Aug 12 11:11:44 lol i have upgraded my uboot once shitting myself the whole time Aug 12 11:11:51 SliMM: although? Aug 12 11:11:55 i think Aug 12 11:11:55 rxr: so now it's not a 15tb elf image :) Aug 12 11:12:03 mmazur: no, i don't have a neo, i'll wait until october Aug 12 11:12:14 all be it Aug 12 11:12:17 Blah. Aug 12 11:12:23 ScaredyCat, you've bricked your neo? :) Aug 12 11:12:23 mmazur: i just want to take a look at writting apps for openmoko Aug 12 11:12:36 mmazur: bricked and unbricked Aug 12 11:12:48 well, part-bricked Aug 12 11:12:49 mmazur: i also need to get some money for buying a neo Aug 12 11:13:13 get to work then ;) Aug 12 11:13:15 haha, built Aug 12 11:13:26 is there realy a way to completely brick a neo ? Aug 12 11:13:36 With a brick. Aug 12 11:13:36 Markinoko: encase it in mortar :D Aug 12 11:13:47 Markinoko: by stuffing your uboot and not having a devel board Aug 12 11:14:05 you could hook AC up to the battery terminals maybe Aug 12 11:14:50 I meant with a debug board... well I guess the vodka technic would work Aug 12 11:14:50 is there any reason why "nand erase rootfs" is not a function of dfu-util Aug 12 11:15:00 how do i install lynx? Aug 12 11:15:11 SliMM: make install Aug 12 11:15:17 awesome! Aug 12 11:15:19 Markinoko: vodka's often good for cleaning out the results of water damage to electronics Aug 12 11:15:25 not dropping frames Aug 12 11:15:34 tiny tiny window Aug 12 11:15:43 cjb_ie: so it's actually an unbricking technic Aug 12 11:15:47 markinoko: don't i have to download it first? Aug 12 11:16:00 SliMM: well, yes ;-) Aug 12 11:16:19 Markinoko: good eh ;-) Aug 12 11:16:30 SliMM: but it should be straightforward to compile Aug 12 11:17:07 and will it work by using the lynx command? Aug 12 11:17:41 SliMM: a bad surprise is possible, but unlikely Aug 12 11:17:43 13:42 < xkr47> hmh now plugging the usb jammed the device and I get "device descriptor read/64, error -110" in the host computer Aug 12 11:17:46 13:42 < xkr47> or jammed == does not respond to toucscreen Aug 12 11:17:49 13:47 < xkr47> same thing happened again Aug 12 11:17:51 13:47 < xkr47> I tail -f /var/log/messages this time on the moko Aug 12 11:17:54 13:47 < xkr47> I got a single ^@ i.e. nul byte Aug 12 11:17:56 13:48 < xkr47> I guess the kernel uImage-2.6.21.5-r2-fic-gta01.bin has some problem.. Aug 12 11:17:56 seee a (famyli dyner) Aug 12 11:18:07 xkr47: !!!!! Aug 12 11:18:28 http://wiki.openmoko.org/wiki/Bricked Aug 12 11:18:28 mickeyl: do you know if they're plans to market the Neo/OpenMoko in the US? Aug 12 11:18:36 woot.. Aug 12 11:18:39 I had that error on my host Aug 12 11:18:42 rtyler: You can buy it online. Aug 12 11:19:02 rtyler: selling through operators won't happen till the very earliest mid 2008. Aug 12 11:19:03 ScaredyCat, it's not bricked.. I shut it down normally and boot it back up .. works fine again Aug 12 11:19:15 ScaredyCat, it just locks up when I unplug, wait 30 sec, replug Aug 12 11:19:15 i wasn't suggesting that... Aug 12 11:19:22 could you please give me a link to lynx' source, i can't find it on google Aug 12 11:19:29 Probably more into 2009. Aug 12 11:19:37 lynx.org Aug 12 11:20:31 SpeedEvil: That's an ISP Aug 12 11:20:48 SliMM: hold on Aug 12 11:20:49 lynx.isc.org Aug 12 11:20:56 by 'lynx', do you mean the text-based browser? Aug 12 11:21:20 SliMM: http://lynx.isc.org/release/ Aug 12 11:21:23 ScaredyCat: do you find gsmd crashes regularly in .2? Aug 12 11:21:53 I find it's pretty useless Aug 12 11:21:53 well anyone with .2 Aug 12 11:22:40 not even sure if most of the .2 stuff is just show and doesn;t do antthing Aug 12 11:22:47 hav you made a call with .2? Aug 12 11:22:54 markinoko: is this the 'lynx' used for flashing the neo emu? Aug 12 11:23:05 SliMM: hu ? Aug 12 11:23:29 ScaredyCat: well i am not too worried about gui stuff at the moment i am more looking at gsmd backend Aug 12 11:23:31 SliMM: so you were not asking for the text based browser I guess... Aug 12 11:23:42 i don't know Aug 12 11:23:56 ScaredyCat, cg on getting your neo back :) Aug 12 11:24:07 * SliMM quotes http://wiki.openmoko.org/wiki/OpenMoko_under_QEMU : Two scripts are provided to generate a firmware for your phone, as realistic as possible. First make sure you have lynx installed, then run Aug 12 11:24:08 SliMM: to flash qemu, all the tools are given with the mokomakefile Aug 12 11:24:15 yeah.. was a bit hairy for a bit there xkr47 Aug 12 11:24:20 aaaaaaahhhhh !! Aug 12 11:24:23 this Aug 12 11:24:27 looks so, yeah.. Aug 12 11:24:28 thanks to cesarb, anrp and SpeedEvil Aug 12 11:24:36 SliMM: I suppose it's to download an official image Aug 12 11:24:45 SliMM: but I thought they used wget ? Aug 12 11:24:49 ScaredyCat, thanks for documenting the effort :) Aug 12 11:25:02 markinoko: so it's the browser? Aug 12 11:25:03 hehe... Aug 12 11:25:16 did raise some interesting points though.. Aug 12 11:25:26 at least we got something out of it.. Aug 12 11:25:30 ScaredyCat: what happened with your phone? Aug 12 11:25:32 SliMM: well, I guess it is. I'm still surprised. I don't have lynx on my laptop and I still managed to flash my qemu Aug 12 11:25:44 ScaredyCat, yeah Aug 12 11:25:46 it nearly went down the tilet Aug 12 11:25:49 toilet Aug 12 11:25:51 :O Aug 12 11:26:03 the phone or the experience ?-P Aug 12 11:26:04 markinoko: well, maybe you didn't use that scripts Aug 12 11:26:04 I made a typo xzcvczx... Aug 12 11:26:05 ScaredyCat: lol cant say i have taken it near my toilet Aug 12 11:26:28 http://wiki.openmoko.org/wiki/Bricked <-- describes some of it Aug 12 11:26:39 SpeedEvil: I was just curious, wondering if the carriers will even acknowledge the product ;) Aug 12 11:27:06 I was just lucky that the part of memory overwritten still allowed it to boot Aug 12 11:27:22 lol Aug 12 11:27:32 SliMM: you want to read the part that says "Using MokoMakefile" Aug 12 11:27:51 SliMM: it is much more simple with mokomakefile to get a qemu up and running Aug 12 11:27:53 * xzcvczx thinks we should all /ignore ScaredyCat for being well an idiot and remind him that the phone is not meant to be put in the toilet Aug 12 11:28:17 I've left a few bricks in there before.... Aug 12 11:28:21 >_> Aug 12 11:28:23 usually on Boxing day Aug 12 11:28:28 lol Aug 12 11:30:05 markinoko: how do i get mokomakefile :d Aug 12 11:30:12 ? Aug 12 11:30:32 http://wiki.openmoko.org/wiki/MokoMakefile Aug 12 11:30:35 SliMM: read the wiki article Aug 12 11:30:56 SliMM: you don't need to build a whole system, only do the part about qemu Aug 12 11:31:24 SliMM: instead of compiling the whole image yourself, you can use a precompiled official image Aug 12 11:31:35 ok Aug 12 11:31:39 thanks Aug 12 11:32:21 SliMM: but I read that you'll still need lynx... and I do have it on my computer, just checked... my bad Aug 12 11:32:24 xzcvczx: what we did think was that you could modify the splash by stuffing it into qemu, then dd'ing out the env and 3 blocks... Aug 12 11:32:42 SliMM: although you should easily find a package for your distro Aug 12 11:32:46 but qemu sd card emulation is broken, so there's no way to get it off the emulation Aug 12 11:33:04 markinoko: i have debian in a virtual machine Aug 12 11:33:19 SliMM: debian has a package, I'm sure about that Aug 12 11:33:58 ok, i'll take a look over it a bit later, i now need to earn some money for buying a neo :P Aug 12 11:35:46 bye now, and thanks Aug 12 11:35:53 i'll be back later :P Aug 12 11:36:36 ScaredyCat: so did it physically go in the toilet? Aug 12 11:40:18 no - that was a joke Aug 12 11:40:39 s/joke/failed joke/ Aug 12 11:40:39 ScaredyCat meant: no - that was a failed joke Aug 12 11:40:43 * hhf423 still browses the internets for a windows dfu util Aug 12 11:42:08 anrp: was doing one hhf423 Aug 12 11:42:13 * hhf423 installs ubunto while searching for a win32 dfu util Aug 12 11:43:04 anrp: you did a win32 dfu util? Aug 12 11:43:27 ScaredyCat: thanks Aug 12 11:47:13 hrm, copying the Aug 10th image over via ext2load in the console seems to have either hung, or is taking an insanely long time :/ Aug 12 11:48:05 does someone know the cause for the note 'NOTE: Upload support is currently broken' for dfu-util? Aug 12 11:48:48 should I flash a new u-boot? I have the one which is one the phone from factory now. Aug 12 11:49:09 hhf423: no.. Aug 12 11:49:14 why would you do that Aug 12 11:49:20 verbricken! Aug 12 11:49:40 ScaredyCat: because the one I have seems to power down after 30 or so sec? Aug 12 11:49:51 they all do Aug 12 11:49:59 ah, ok. no worries then Aug 12 11:50:11 press the aux button every so often Aug 12 11:50:29 ScaredyCat: so what did you actually do to b0rk it Aug 12 11:51:15 A typing error in the splashimage environment variable resulted in the splash image being read into the memory starting at 0x200000 rather than 0x32000000. The change was written, with the error being unnoticed, using saveenv. Aug 12 11:52:28 ScaredyCat: so what did you actually do to b0rk it Aug 12 11:52:43 * xzcvczx smacks up his keyboard Aug 12 11:52:56 * ScaredyCat has a feeling of deja vu Aug 12 11:53:06 lol Aug 12 11:53:16 it's the matrix! Aug 12 11:53:32 yup i just edited your nuts off Aug 12 11:53:37 :( Aug 12 11:53:41 they are about 2 feet in front of you now Aug 12 11:53:58 the built in accelerometers surely should have detected that Aug 12 11:54:08 sliiightly off-topic Aug 12 11:54:16 * ScaredyCat flys around in slow-mo Aug 12 11:55:24 are there any kiwis left here still waiting on their phone to be shipped Aug 12 11:55:30 now, who said a free phone would drive you nuts... Aug 12 11:55:38 kiwis ?-P Aug 12 11:55:59 yeah fat birds who cant fly Aug 12 11:56:13 nocturnal fat birds who cant fly Aug 12 11:56:17 which are the most sensible images to use at the moment? for rootfs etc? Aug 12 11:56:28 fat birds only come out at night? Aug 12 11:56:39 ScaredyCat: lol these ones do Aug 12 11:57:08 actually, I think kiwis have dissappeared Aug 12 11:57:25 http://news.nationalgeographic.com/news/2006/02/images/060217_kiwi.jpg Aug 12 11:57:28 how can you tell? they only come out at night... then it's dark Aug 12 11:57:30 :P Aug 12 11:57:59 dont you have "kiwi's" in the usa that are a "fruit"? Aug 12 11:58:14 hhf423: latest buildhost images worked fine for me. Aug 12 11:58:19 and a fruit is a faggot? Aug 12 11:58:40 and a faggot is a bundle of sticks that you got? Aug 12 11:58:41 * ScaredyCat can't keep up with 1000's of different meanings Aug 12 11:59:03 * Markinoko just can't keep up Aug 12 11:59:03 actually, faggots are like meatballs... Aug 12 11:59:14 kiwis are the cause for the big hole in the neo so they can carry it with their beak? Aug 12 11:59:14 Kensan: you did not use http://chooseopen.com/openmoko/build/ ? Aug 12 11:59:16 you just want to suck on them? Aug 12 11:59:32 errm Aug 12 11:59:35 http://www.bbc.co.uk/food/recipes/database/faggotswithoniongrav_3899.shtml Aug 12 12:00:05 stop the food / animal chat already Aug 12 12:00:08 pls Aug 12 12:00:09 hhf423: http://buildhost.openmoko.org/tmp/deploy/images/?C=M;O=D Aug 12 12:00:12 lol on-topic discussion is so much fun Aug 12 12:00:42 xkr47: they started it i was talking on-topic at the start by asking if there were any kiwi's still waiting on phones Aug 12 12:01:20 kiwis ?-P Aug 12 12:02:06 kiwi is a name for nzers based on one of our native endagered species the kiwi Aug 12 12:02:25 nzers ? Aug 12 12:02:34 New Zealanders Aug 12 12:02:39 thanks Kensan Aug 12 12:02:39 New? Aug 12 12:02:43 at least someone knoes Aug 12 12:02:44 knows Aug 12 12:03:09 ScaredyCat: it's the aussies vs. the kiwis Aug 12 12:03:13 ;) Aug 12 12:03:36 well, the aussies would win.... they're all crims ..... Aug 12 12:03:39 * ScaredyCat hides Aug 12 12:04:06 isn't nz just au without the nasty bugs Aug 12 12:04:28 ScaredyCat: pretty much Aug 12 12:04:41 and a lot less money Aug 12 12:04:48 hi Aug 12 12:04:53 and people... Aug 12 12:04:58 but more sheep Aug 12 12:04:59 Kensan: so you used openmoko-devel-image-fic-gta01-20070803091138.rootfs.jffs2 and uImage-2.6.21.6-moko11-r1_0_0_2388_0-fic-gta01.bin ? Aug 12 12:05:04 * xzcvczx is never compiling ghc again Aug 12 12:05:14 7hrs and counting Aug 12 12:05:15 hi Aug 12 12:05:52 someone added nfs into the build and broke it ? Aug 12 12:05:58 lo PBeck Aug 12 12:06:21 xzcvczx: yes one can really wast cpu cycles building moko Aug 12 12:06:52 Talpa: well that too but nah this is wasting precious cpu cycles for building openmoko building a haskell compiler Aug 12 12:07:03 hhf423: yes, but feel free to flash any image you like. As long as you don't brick uBoot you can always put a different image on it. Aug 12 12:07:05 * hhf423 just screwed up his ubuntu installation, start from scratch, arg Aug 12 12:07:10 hi ScaredyCat :) Aug 12 12:07:44 Kensan: ok Aug 12 12:07:49 I see you pop in (just as I'm leaving) and always say 'hi' Aug 12 12:07:53 hhf423: good for you... now you have the chance to install gentoo Aug 12 12:08:17 lol Aug 12 12:08:22 now gentoo is good at wasting cycles Aug 12 12:08:22 Talpa: Gentoo is not the right distribution to "get to know linux". Aug 12 12:08:25 * xzcvczx gives Talpa a beer Aug 12 12:08:33 Kensan: i disagree Aug 12 12:08:49 Kensan: its the only thing that actually got me to learn linux to a decent degree Aug 12 12:08:58 the rest dont teach you much at all Aug 12 12:09:02 xzcvczx: is't gentoo best at using cpu cycles Aug 12 12:09:05 Talpa: I totally do not care about distributions, I just want to flash my Neo Aug 12 12:09:06 no. Aug 12 12:09:15 gentoo is the worst for that... Aug 12 12:09:24 xzcvczx: everything you know is a lie Aug 12 12:09:25 xzcvczx: let me explain: getting to know linux in the sense of making your first baby steps into the linux world. Aug 12 12:09:49 * hhf423 totally does not care about operating systems since the OS wars in the early nineties Aug 12 12:10:01 Kensan: with the gui installtion now its easier Aug 12 12:10:09 yeah, when os/2 warp won Aug 12 12:10:16 exactly! Aug 12 12:10:48 * hhf423 can patiently wait for linux, windows and macosx to slowly become what OS/2 once was Aug 12 12:10:53 xzcvczx: yes, sure. but there's too much breakage *all* the time. It's not geared towards people wanting to give linux a try. Aug 12 12:11:10 but each to his/her own. Aug 12 12:11:24 well, that unmask stuff is crap.. Aug 12 12:11:37 If I want to install stuff , *I* want to install it Aug 12 12:12:01 ScaredyCat: hence why you add it to unmask Aug 12 12:12:28 no, wget src-tarball Aug 12 12:12:44 ScaredyCat: lol you can still do that anyway Aug 12 12:12:53 untar... ./configure && make & make cheese Aug 12 12:13:10 ScaredyCat: there is nothing to stop you doing that Aug 12 12:13:27 there's a whole generation of people who'll never understand dependancy hell Aug 12 12:13:46 ScaredyCat: what do you use? Aug 12 12:13:56 lfs ;) Aug 12 12:14:12 thats true Aug 12 12:14:16 the better we make our tools Aug 12 12:14:19 the less peolpe know Aug 12 12:14:26 ScaredyCat: go out one friday night and find yourself a lady friend Aug 12 12:14:28 as they just expect things to 'be done for them" Aug 12 12:14:40 hahaha, good reason for making crap tools Aug 12 12:14:45 there is a new generation of programmers/developers who dont know that the stack and the heap are Aug 12 12:14:51 xzcvczx: I'm married... she might object Aug 12 12:14:57 thewy dont know what a page is Aug 12 12:14:59 I'll use that one the next time someone complains about a tool :-) Aug 12 12:15:07 and actually have no clue what cache and buffers are Aug 12 12:15:12 * ScaredyCat complains about hhf423 Aug 12 12:15:16 all they know is how to catch an exception Aug 12 12:15:23 yeah.. Aug 12 12:15:24 or use some java class Aug 12 12:15:33 if its harder than chosing what java class to use Aug 12 12:15:36 its obviously too hard Aug 12 12:15:46 * ScaredyCat barfs at the sound of coffee related language Aug 12 12:15:49 it scares me to think - this is where things are going Aug 12 12:15:59 where is the clue all going to go? Aug 12 12:16:01 raster: it just allows for the ones that do know shine that much brighter ;) Aug 12 12:16:05 uhmm coffee Aug 12 12:16:11 raster: we'll be dead by then Aug 12 12:16:17 Jiten: true. job security :) Aug 12 12:16:36 YES its done building nano! Aug 12 12:16:44 wtf! Aug 12 12:16:51 what's wrong with vi! Aug 12 12:16:54 Talpa: howd you build nano Aug 12 12:16:58 ScaredyCat: too use to nano Aug 12 12:17:00 bitbake nano Aug 12 12:17:06 lol Aug 12 12:17:08 ahhh you did it that way Aug 12 12:17:10 vi -- beginners pronounce it "Why I?" Aug 12 12:17:29 no, that's people from Newcastle Aug 12 12:17:38 xzcvczx: he obviously did it the easy way... Aug 12 12:17:39 disk usage only 727MB .... Aug 12 12:18:01 xzcvczx: were you deliberately trying to make it hard for yourself then ? Aug 12 12:18:03 ScaredyCat: if vim would build on my host i might start using it but its not building at the moment Aug 12 12:18:23 ScaredyCat: no just so many ipks are created i cbf installing them all Aug 12 12:18:51 hehe Aug 12 12:20:13 what do people use for c programming in linux cmdline? Aug 12 12:20:29 emacs? Aug 12 12:20:37 * xzcvczx waits for half the people to say emacs and the other half to say vi Aug 12 12:21:04 vi Aug 12 12:21:12 xzcvczx: jed Aug 12 12:21:20 should't they say emacs or vim? Aug 12 12:21:20 copy con helloworld.c Aug 12 12:21:20 * xzcvczx thumps ScaredyCat for just saying that for the hell of it Aug 12 12:21:33 i bet there's a joe user out there too Aug 12 12:21:36 Talpa: i was meaning vi derivative/clone Aug 12 12:21:49 xzcvczx: i only use vi Aug 12 12:22:39 xzcvczx, vim. Aug 12 12:23:02 ppfpt amateur ;) Aug 12 12:23:10 :D Aug 12 12:23:28 ...but who switched of '.' in the vi on the neo?? Aug 12 12:23:59 * hhf423 prods his old cdrom to copy faster Aug 12 12:24:18 mdt: its busybox vi Aug 12 12:24:21 hhf423: try yelling at it Aug 12 12:24:31 robtaylor, i know - so what? Aug 12 12:24:37 mdt: not-quite-vi Aug 12 12:24:42 it's not real Aug 12 12:24:42 Talpa: :-) Aug 12 12:25:13 robtaylor, i saw one busybox-vi that was capable to do so... i assume it's switched off somehow... Aug 12 12:25:50 mdt: hmm, check the busybox config Aug 12 12:25:59 i already do so... :) Aug 12 12:26:13 * robtaylor would just install vim Aug 12 12:27:00 * Talpa tries to do a "bitbake vim" Aug 12 12:27:56 NOTE: package vim-7.0: completed Aug 12 12:28:10 well it compiles Aug 12 12:31:42 yes, they switched it off: # CONFIG_FEATURE_VI_DOT_CMD is not set Aug 12 12:32:41 counter Aug 12 12:32:41 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter Got it, get yours from openmoko.com (and read topic) (1800;252) Aug 12 12:36:27 mdt: i think its probably off by default in busybox Aug 12 12:36:43 xzcvczx, i think oe did so... Aug 12 12:40:35 btw. am I the only one that has problems using more than one cpu when using bitbake? Aug 12 12:41:11 when it runs make it used the specified number of cpus Aug 12 12:41:18 Talpa: send me the computer and i will have it instead :P Aug 12 12:41:52 anybody getting errors like this when loading files in the console from the SD card? Aug 12 12:41:55 http://pastebin.ca/654433 Aug 12 12:41:57 * rtyler grumbles Aug 12 12:42:53 sounds like a non properly compiled ext2load Aug 12 12:43:03 as in not for cpu in phone Aug 12 12:43:23 considering I've not changed u-boot ever, I don't see how that would work ;) Aug 12 12:43:34 I performed this same action with a different image yesterday >_< Aug 12 12:44:38 xzcvczx: but I need all the cores I can get my hands on for my self Aug 12 12:44:58 so why not ext2 support in the neo ,other than u-boot :/ Aug 12 12:45:05 s/not/no/ Aug 12 12:45:05 ScaredyCat meant: so why no ext2 support in the neo ,other than u-boot :/ Aug 12 12:45:37 * xzcvczx wishes someone would configure a shutup command for apt Aug 12 12:45:40 i wouldn;t have to buy a card reader then Aug 12 12:45:47 apt: stfu Aug 12 12:45:49 :( Aug 12 12:45:58 apt: piss of Aug 12 12:45:59 * apt of goes behind Beans & Brews and waters a tree Aug 12 12:46:00 off even Aug 12 12:46:02 ScaredyCat: that works? Aug 12 12:46:09 well, that's convenient Aug 12 12:46:10 dunno Aug 12 12:46:20 s/works// Aug 12 12:46:21 s/dunno/puppy/ Aug 12 12:46:21 xzcvczx meant: ScaredyCat: that ? Aug 12 12:46:21 ScaredyCat meant: puppy Aug 12 12:46:24 no Aug 12 12:46:28 grrrr Aug 12 12:46:39 apt: /ignore apt Aug 12 12:46:46 s///quit/ Aug 12 12:46:59 PCW: Is there a planned list of applications that will ship with the consumer version in October? In particular, which Web browser, e-mail client, and calendar will it ship with? Aug 12 12:46:59 OpenMoko: Yes, we do have a list. Aug 12 12:46:59 As for particulars, we write the interface ourselves, but a lot of the base code comes from existing FOSS [Free and Open Source Software] projects. E-mail will be based on the TinyMail framework. Our browser is based on WebKit (the same stuff that Safari is using). Aug 12 12:47:01 :O Aug 12 12:56:22 we really gotta do something about this boottime :P Aug 12 12:56:37 rtyler: its better than i thought it would be Aug 12 12:56:42 I could almost fit the entire Daft Punk song in 'Robot Rock' when I recorded a video of it booting Aug 12 12:57:01 xzcvczx: but it's booting from flash, it seems like it should be faster Aug 12 12:57:14 guess it's Linux + 266Mhz of arm love Aug 12 12:57:43 my old desktop 266 would never boot as fast as the moko Aug 12 12:57:57 your old desktop wasn't running off a flash disk was it? :) Aug 12 12:58:08 was running off something even faster Aug 12 12:58:15 a bigfoot Aug 12 12:58:18 Coaxial netboot? :P Aug 12 12:58:30 e Aug 12 12:58:38 * rtyler mentions http://www.unethicalblogger.com/posts/r_tyler_ballance/video_hello_moko Aug 12 12:58:44 * rtyler hums some daft punk Aug 12 12:58:58 I'm incredibly bored waiting for this build to finish Aug 12 13:00:27 rtyler: me too Aug 12 13:00:47 Have you considered finding Jesus? Valuable cash prizes if you find him! Challenging clues! Aug 12 13:01:16 SpeedEvil: I think I saw him selling cars down at that place on Market St., you seen it? "Jesus Chrysler" I think it's called :D Aug 12 13:01:30 SpeedEvil: why is he hiding Aug 12 13:01:46 xzcvczx: marketing ploy to get people to buy his book :D Aug 12 13:01:57 lol Aug 12 13:02:25 xzcvczx: the Roman police it looking for him Aug 12 13:02:30 haha Aug 12 13:02:33 I liked mine better :P Aug 12 13:03:17 rtyler: that lazy bastard, he did't even write his own book himself Aug 12 13:03:19 * ScaredyCat builds minimo Aug 12 13:03:30 holy ghost writer Aug 12 13:03:38 well played sir, well played Aug 12 13:06:05 how hard would it be to put gcc onto the phone? Aug 12 13:06:09 Is Jesus fighting for freedom of speech ? Aug 12 13:06:32 xzcvczx: I can't imagine it'd be too hard, but cross compiling will always be much faster Aug 12 13:06:44 much much much Aug 12 13:08:10 wonder if anyone has done a video of 2007.2 Aug 12 13:08:16 rtyler: yea i know but still would be convinient to have it on there Aug 12 13:08:31 Linux_Galore: I can probably make on after I reflash my Neo :P Aug 12 13:08:45 rtyler: thanks before hand Aug 12 13:08:59 although, stacking the books so my laptop's camera points straight down will kind of suck >_< Aug 12 13:09:07 * Linux_Galore twidles fingers while waiting for the TA02 Aug 12 13:09:13 GTA02* Aug 12 13:09:24 that's lots of finger twiddling Aug 12 13:09:42 yeah, getting a bit sad hacking my N95 Aug 12 13:14:29 Linux_Galore: by a GP2X and force it to run openmoko? Aug 12 13:14:34 buy Aug 12 13:16:54 http://www.flickr.com/photos/xbeepx/1088888503/ Aug 12 13:17:07 Talpa: I already have a GP2X Aug 12 13:17:56 and a Zaurus 5600 Aug 12 13:18:18 both running debian right now Aug 12 13:19:27 I just need a phone and then I will have a complete Linux working enviroment Aug 12 13:20:03 I was going to get a Moto Linux phone untill I heard about OM Aug 12 13:20:17 you seem to have a good paying job Aug 12 13:20:37 SrRaven: Im in marketing Aug 12 13:20:56 I used to be in IT services, not much money in that Aug 12 13:20:57 thats a yes Aug 12 13:21:30 Linux_Galore: are you around here often? Aug 12 13:22:33 not really, I pop in and out but I leave the client logged in Aug 12 13:23:42 Only reason Im on right now is Im downloading some podcasts for my iPod Aug 12 13:24:18 Linux_Galore: if you add the openmoko category feed from my stupid blog to your feed reader, I'm sure you'll be notified when the 2007.2 video is up ;) Aug 12 13:24:24 * rtyler already posted some photos of it Aug 12 13:24:47 is anyone else here keen to have the ability to network the phone up to a windoze machine? Aug 12 13:26:10 * ScaredyCat stares blankly at xzcvczx Aug 12 13:26:18 Ive got a perl script that scans flickr for openmoko images and tells me if there is a new one Aug 12 13:26:33 ScaredyCat: eat me Aug 12 13:26:44 *chomp* *chomp* Aug 12 13:26:45 errm, no. Aug 12 13:27:10 xzcvczx: once I'm able to get things built for my Neo, I'll start working on a .NET/Mono based sync app Aug 12 13:27:26 xzcvczx: not exactly a big rush to work on making a completely open platform work with a closed one Aug 12 13:27:49 it needs to be done... Aug 12 13:28:01 unfortunately yes Aug 12 13:28:13 does USB networking even exist on Windows? Aug 12 13:28:18 need a Linux client too Aug 12 13:28:19 rtyler: I'd wait until that desicion has been made (what wouldbe used) Aug 12 13:28:23 Linux_Galore: mostly for while i am at work where i only have windows machines and at home as my only linux box is in the other room and i only have a windows laptop so is more useful if i have full controll over the phone physical and networked Aug 12 13:28:34 syncml is suggested, but nothing appears set in stone Aug 12 13:28:37 rtyler: yes Aug 12 13:28:42 xzcvczx: I have Linux at work and home Aug 12 13:28:49 Linux_Galore: well i dont Aug 12 13:29:00 xzcvczx: cry me a river Aug 12 13:29:07 ScaredyCat: I was going to write a little daemon that sits on top of the OpenMokoFramework Aug 12 13:29:14 Linux_Galore: don't sit around and flame plz kthx Aug 12 13:29:35 surely that's 'kthxbai' Aug 12 13:29:41 lol Aug 12 13:29:56 I dont use linux either but interested in openmoko since its start Aug 12 13:30:16 Sounds like syncing is going to be something based on Funambol (think I read it from somewhere also http://www.funambol.com/partners/ gives a hint) Aug 12 13:30:18 SrRaven: well you will still probably need linux for flashing the phone at this stage Aug 12 13:30:33 like id buy a neo1973 at the current state Aug 12 13:30:41 id prolly never buy one,looks to cheap for me Aug 12 13:31:02 SrRaven: what you planning on using openmoko on? Aug 12 13:31:22 probably the next mobile that will be able to use it Aug 12 13:31:25 http://i0005.photobucket.com/albums/0005/icanhascheezburger/2007/7/15/12/128290020895277500phoolishcatu.jpg Aug 12 13:31:52 SrRaven: people are more interested in paying for an open phone than how pretty it is Aug 12 13:32:03 well I am superficial like that Aug 12 13:32:07 SrRaven: if you want pretty get a locked iPhone Aug 12 13:32:16 iphone is shit Aug 12 13:32:36 SrRaven: why that ? Aug 12 13:32:40 SrRaven: or get any phone, they are all locked Aug 12 13:32:42 im more of the mda vario 2 type for looks Aug 12 13:32:54 i suppose i am very optimistic asking whether busybox vi has syntax highlighting? Aug 12 13:32:55 rxr cant stand apple Aug 12 13:33:12 overpriced shit Aug 12 13:33:15 xzcvczx: haha Aug 12 13:33:46 Ive figured out apples design method, remove all keys and edges and cover it with a shiny finnish Aug 12 13:33:46 Markinoko: dang it Aug 12 13:33:57 xzcvczx: busybox vi may be as good as a solaris vi Aug 12 13:34:02 the new iMac is a great example Aug 12 13:34:34 * rtyler gets out his -1, Offtopic mod Aug 12 13:34:42 heh Aug 12 13:35:04 * xzcvczx stabs vim Aug 12 13:35:12 I actually like vim Aug 12 13:35:44 Linux_Galore: i am not stabbing it for usage/feature issues i am stabbing it for compile issues Aug 12 13:36:09 Linux_Galore: peolpe like shiny - bling bling Aug 12 13:36:10 :) Aug 12 13:36:11 vim+bootcd+chroot=fixed system with corrupted linux kernel image Aug 12 13:36:43 * Linux_Galore spray paints an OM gold Aug 12 13:37:12 * Linux_Galore glues 25 carats of diamons to the front Aug 12 13:37:16 i have got too use to nano unfortunately Aug 12 13:37:18 diamonds* Aug 12 13:37:29 there OM with bling Aug 12 13:38:22 bling bling Aug 12 13:38:44 that will be 50k please Aug 12 13:38:57 * SpeedEvil glues 25 carrots to his neo. Aug 12 13:39:12 SpeedEvil: call it Bugs Aug 12 13:39:51 Not sure about the brown theme for the OM though Aug 12 13:40:44 I like blue/metalic grey Aug 12 13:41:27 heh, do a LCARS theme for the OM, few thousand Trekie geeks will want one Aug 12 13:41:46 if I recall correctly, you make themes for OpenMoko; snap to it then Aug 12 13:42:02 I need one first Aug 12 13:42:19 no you don't, run QEMU Aug 12 13:42:31 http://wiki.openmoko.org/wiki/OpenMoko_under_QEMU Aug 12 13:42:33 have fun Aug 12 13:42:43 rtyler: QEMU doesnt show real perspective for artwork Aug 12 13:43:04 also you need to see how the art and colour look "on the phone" Aug 12 13:43:33 I think the more important thing here is that Linux_Galore knows that the star trek interface is called LCARS Aug 12 13:43:47 * ScaredyCat points at the trekkie Aug 12 13:44:12 ScaredyCat: no I know the LCARS theme mainly because the N800 has an optional UI called LCARS with sounds Aug 12 13:44:41 yes, we believe you... Aug 12 13:44:43 convenient excuse ^_^ Aug 12 13:44:43 :P Aug 12 13:45:12 http://www.youtube.com/watch?v=YwXBPjLdJnU Aug 12 13:45:22 you just made that Aug 12 13:45:26 ;) Aug 12 13:45:26 LCARS theme on N770 ^^ Aug 12 13:45:44 * Linux_Galore slam dunks Aug 12 13:46:07 omg nerdz Aug 12 13:46:44 heh Aug 12 13:47:22 blew me away when I saw that, also learnt LCARS is the name of the Star Trek layout when I did a google Aug 12 13:48:07 ok, I have *not* just installed it to *not* try it Aug 12 13:50:36 lol lcars iPhone http://www.youtube.com/watch?v=mafumN10Tm0 Aug 12 13:53:29 putting LCARS on any phone with sound is like sticking a bib neon sign above your head with and arrow and the text NERD Aug 12 13:53:35 big* Aug 12 13:53:46 so is mentioning it on IRC Aug 12 13:54:01 I just find it funny Aug 12 13:54:42 http://www.google.com/trends?q=neo1973%2C+openmoko&ctab=0&geo=all&date=2007&sort=0 Aug 12 13:55:33 Dmitry_Platonov: more importantly: http://www.google.com/trends?q=neo1973%2C+openmoko%2C+iphone&ctab=0&geo=all&date=2007&sort=0 ;) Aug 12 13:56:13 wow, russians are really into the OM Aug 12 13:57:19 rtyler, seen that :-/ Aug 12 13:57:32 Im out Aug 12 13:57:39 of to bed, see ya all Aug 12 13:58:09 Linux_Galore, russians are into gadgets Aug 12 14:00:50 anybody had their host machine completely fux0r when they connect the device? Aug 12 14:01:04 my Linux box just refuses to bring up /dev/ttyACM0 anymore Aug 12 14:01:12 makes accessing the console quite troublesome :P Aug 12 14:01:22 modprobe cdc_acm Aug 12 14:01:41 ccnet% sudo modprobe cdc_acm Aug 12 14:01:41 ccnet% Aug 12 14:01:42 hrm Aug 12 14:01:57 ccnet% lsmod | grep cdccdc_acm 15904 0 usbcore 134280 7 cdc_acm,usb_storage,libusual,usbhid,ohci_hcd,ehci_hcdccnet% Aug 12 14:02:02 gack, screen fucked that up Aug 12 14:02:55 ccnet% lsmod | grep cdc Aug 12 14:02:55 cdc_acm 15904 0 Aug 12 14:02:55 usbcore 134280 7 cdc_acm,usb_storage,libusual,usbhid,ohci_hcd,ehci_hcd Aug 12 14:02:58 there we go Aug 12 14:03:01 ccnet% Aug 12 14:03:21 tYou do know you need to boot with the button held down? Aug 12 14:03:31 yes yes, it goes into u-boot mod Aug 12 14:03:32 mode* Aug 12 14:03:44 what does lsusb then show? Aug 12 14:04:06 it doesn't show anything other than the other USB devices, but four sets of this get logged into /var/log/messages: Aug 12 14:04:09 Aug 12 06:59:47 ccnet kernel: [116060.564000] usb 2-7: new full speed USB device using ohci_hcd and address 17 Aug 12 14:04:43 it doesn't show the neo? Aug 12 14:04:51 nope Aug 12 14:04:55 Odd. Aug 12 14:05:20 the first couple times I flashed the device, I was able to from this computer, I'm not sure what's gone wrong with it Aug 12 14:05:31 Stupid stuff first. Aug 12 14:05:48 You are in fact sure you're plugging it into a working USB port, and the otehr end is plugged into the neo? Aug 12 14:06:11 yes yes, plugged into the same set of front USB port as my workng SD reader is :) Aug 12 14:06:17 K. Aug 12 14:06:24 is your battery low by any chance - I've noticed USB gadget not working when battery is low Aug 12 14:06:37 Writchie: nah, battery shows full, and it's a 500ma port Aug 12 14:06:53 Battery shows full in what? Aug 12 14:07:00 when I boot the neo Aug 12 14:07:22 huzzah, moko's built, that only took 3 hours >_> Aug 12 14:07:27 :) Aug 12 14:07:43 so, woohoo, let's get this working again so I can flash :D Aug 12 14:08:04 Hmm. Aug 12 14:08:09 SpeedEvil: so Linux is recognizing *a* device, it's just not registering it as a cdc_acm device Aug 12 14:08:19 what does lsusb show? Aug 12 14:09:06 SpeedEvil: http://pastebin.ca/654488 Aug 12 14:09:11 Good morning. Aug 12 14:09:18 morning your highness Aug 12 14:09:22 heh Aug 12 14:09:44 anybody seen htc touch interface? Aug 12 14:09:50 that's not showing it then. Aug 12 14:10:04 how is the video player working? Aug 12 14:10:12 SpeedEvil: nope :/ Aug 12 14:10:13 not in real life Aug 12 14:10:14 but videos Aug 12 14:10:23 SliMM: basically - primitively. Aug 12 14:10:27 "If you die in Canada, you die in real life!" Aug 12 14:10:27 Will the Neo draw 500ma in charging mode, or will it draw more? Do we need to use a current limiting charger device? Aug 12 14:10:46 DukeOfURL: it will draw up to 500mA Aug 12 14:10:51 at 320*240 with sound, does it run at at least 24 fps? Aug 12 14:10:57 yes. Aug 12 14:11:01 However. Aug 12 14:11:05 Not with rotation. Aug 12 14:11:16 It would need an optimised rotator to be written. Aug 12 14:11:25 And for P2, this is not an issue at all. Aug 12 14:11:31 I'm looking at my Motorola mini USB cigarette lighter adapter. I'm pretty sure the output is 5V. Aug 12 14:11:34 With the faster CPU and harfware accel. Aug 12 14:11:53 you should take a look over the mpeg player at http://www.rockbox.org Aug 12 14:12:00 the gigabeat has similar processor Aug 12 14:12:13 and it works great rotated Aug 12 14:12:25 Does the gigabeat have any video hardware? Aug 12 14:12:31 Would the battery get too hot will an unlimited supply? Aug 12 14:12:31 none Aug 12 14:12:34 writing a rotated output handler isnt hard Aug 12 14:12:35 And what MHz is the CPU? Aug 12 14:12:38 making it fast Aug 12 14:12:39 ScaredyCat: just read your new wiki page... The reason editing the environment by hand does not work well is not the missing variable, but that the CRC doesn't match anymore, which makes it ignore the whole environment Aug 12 14:12:45 about 300 Aug 12 14:12:50 DukeOfURL: no - it draws what it needs. Aug 12 14:12:59 i don't know if it's the same cpu, but it's arm Aug 12 14:13:09 raster: no - but it requires per-codec tweaking. Aug 12 14:13:13 it's over 250 mhz Aug 12 14:13:15 And for someone to do it. Aug 12 14:13:28 SpeedEvil: hmmm not that i can think of Aug 12 14:13:34 per yuv format - yes Aug 12 14:13:40 And as this work will be essentially useless for P2, this is not going to be done by FIC. Aug 12 14:13:55 invariably a doced will produce a yuv image of some sort Aug 12 14:14:14 err Aug 12 14:14:20 codec will produce Aug 12 14:14:32 so u really need to do a yuv->rgb + rotator Aug 12 14:14:57 +scaler Aug 12 14:15:04 SpeedEvil: I can connect using minicom on my Mac, but my Linux box is the build machine, so I'm wondering if maybe the cdc_acm module just got fux0rd on the host somehow? Aug 12 14:15:10 since it certainly doesn't seem to be the device Aug 12 14:15:17 rtyler: lsusb does not show the device Aug 12 14:15:24 yar Aug 12 14:15:32 Dmitry_Platonov: yeah - that too Aug 12 14:15:35 rtyler: it will - even if the device driver doesn't pick it up. Aug 12 14:15:40 u will roll those into 2 convertor Aug 12 14:15:45 yikes Aug 12 14:15:55 maybe usbcore is fucked? Aug 12 14:16:31 in this case u'd have a specific yv12->rgb565+90rot+scale convertor Aug 12 14:16:46 rockbox's mpegplayer is very optimized Aug 12 14:16:54 it works great with low resources Aug 12 14:17:38 but only mpeg Aug 12 14:17:46 how many bpp is neo's screen? Aug 12 14:17:47 things liek xine or mplayer handle tonnes fo codecs Aug 12 14:17:56 and in the logn run are more useful Aug 12 14:17:57 16bpp Aug 12 14:18:13 raster: mpeg is a good place to start Aug 12 14:18:16 :P Aug 12 14:18:57 unfortunately, the mpeg player isn't codec-based yet Aug 12 14:19:07 so maybe that's why it works so well Aug 12 14:19:54 cesarb: k, updating... Aug 12 14:19:59 what's you rn ? Aug 12 14:20:24 s/you/your/ Aug 12 14:20:24 ScaredyCat meant: what's your rn ? Aug 12 14:20:43 is the next neo going to have a better processor and some sort of graphics acceleration? Aug 12 14:20:47 yes. Aug 12 14:20:51 faq? Aug 12 14:20:51 See http://wiki.openmoko.org/wiki/FAQ (and also SH1, SH1_FAQ, Neo1973, Developer_preview and Neo1973_Hardware#GTA01Bv4 pages) Aug 12 14:21:05 so, here I go. Some linux booted, downloaded dfu util and the images Aug 12 14:21:16 and also accelerometers, right? Aug 12 14:21:36 and when I want to flash, I get @can not claim interface Aug 12 14:21:48 SliMM: yes Aug 12 14:22:25 i will turn it into a wii-like remote control for my lego mindsorms *smug* Aug 12 14:22:53 I get "can not claim interface 2" Aug 12 14:25:21 dfu --list gives me Found Runtime: [0x1457:0x5119] devnum=0, cfg=0, intf=2, alt=0, name="UNDEFINED" Aug 12 14:26:19 but there should be like a list of five Aug 12 14:27:23 I think it listed the same for me Aug 12 14:27:25 and flashing worked Aug 12 14:27:36 I think the wiki documentation does not match the latest sources Aug 12 14:27:41 SpeedEvil: so no ideas? :/ Aug 12 14:28:35 rxr: so you started flashing by pressing the two buttons to get into uboot and the start dfu util? Aug 12 14:29:44 yellow Aug 12 14:29:51 after typing openmoko/flash.sh, i get thi message: Aug 12 14:29:53 Splash needs 'pngtopnm' and 'ppmtorgb3' - ,ake sure that they're in $PATH. Aug 12 14:30:02 what does this mean? Aug 12 14:30:13 this* Aug 12 14:30:20 it means you need to install those apps Aug 12 14:30:32 oh, great Aug 12 14:30:40 ScaredyCat: rn? Aug 12 14:30:55 real name Aug 12 14:31:21 but it's ok Aug 12 14:32:03 ScaredyCat: Cesar Barros (yeah, I use my real name as my nickname :D ) Aug 12 14:32:17 weird ;) Aug 12 14:32:26 hhf423_ub: yes Aug 12 14:32:41 rxr: what does dfu show when you hook up your neo now? Aug 12 14:32:52 hhf423_ub: it's booted into Linux anyway Aug 12 14:33:12 ah Aug 12 14:33:39 but if there is no list of alt shown, dfu can not flash to alt 3 no? Aug 12 14:33:46 hhf423_ub: if it is vitally important for you I could boot it into u-boot ... Aug 12 14:34:07 ScaredyCat: in the wiki, it would be more useful to put my wiki username (CesarB) instead of the real name Aug 12 14:34:09 rxr: well right now I have no idea whats wrong here Aug 12 14:34:15 iihaa fully control public bt and unpublic bt action from gui :) Aug 12 14:34:41 rxr: so it would help, i d appreciate that Aug 12 14:35:00 * rxr wonders where the official openmoko devels are hiding ... Aug 12 14:35:09 ScaredyCat: the "discoveries" part, item 2, should also say something about the CRC ;-) Aug 12 14:35:13 it's the weekend, they might be resting ;) Aug 12 14:35:36 * rtyler furiously pounds on his keyboard and screams at his kernel Aug 12 14:35:44 some were on the CCC camp and are probably traveling now Aug 12 14:35:52 hhf423_ub: /me rebooting Aug 12 14:36:03 rxr: thanks Aug 12 14:36:05 rtyler: aren't they in some event? if so, it might not be because of being a weekend, but because the event finished and they are on a plane/resting Aug 12 14:36:16 Chaos Camp Aug 12 14:36:17 better cesarb? Aug 12 14:36:30 * hhf423_ub wonders why the neo can not just stay in uboot mode but powers off after a while Aug 12 14:36:56 http://t2-project.org/ # dfu-util --list Aug 12 14:36:58 dfu-util - (C) 2007 by OpenMoko Inc. Aug 12 14:36:58 This program is Free Software and has ABSOLUTELY NO WARRANTY Aug 12 14:36:58 Found Runtime: [0x1457:0x5119] devnum=13, cfg=0, intf=2, alt=0, name="UNDEFINED" Aug 12 14:37:54 ok, my devnum is 0 but thats all the difference Aug 12 14:38:00 oh just so you know it looks like a patch will come in soon to dfu-util which will replace -a 3 with kernel and -a 5 with rootfs Aug 12 14:38:04 or seomthing like that Aug 12 14:38:20 so how did you execute dfu util to flash a kernel? Aug 12 14:38:33 about time too xzcvczx you lazy sod Aug 12 14:38:33 hhf423_ub: like it was listed in the wiki Aug 12 14:38:57 rxr: like ./dfu-util -a 3 -R -D ./uImage-2.6...blabla ? Aug 12 14:39:08 well - I would have to open a browser to confirm Aug 12 14:39:16 I do not want to confirm that magic numbers from the top of my head :-) Aug 12 14:39:26 ScaredyCat: lol not me i just googled the thing and i found a patch in someones people.openmoko dir that says its going to be a future patch Aug 12 14:39:39 hhehe Aug 12 14:39:51 rxr: thats verbatim from the web: http://wiki.openmoko.org/wiki/Flashing_openmoko Aug 12 14:39:54 well proposed patch Aug 12 14:40:01 hhf423_ub: yes, like on that side Aug 12 14:40:08 the dfu example printout differs there, right ?= Aug 12 14:40:16 rxr: yes Aug 12 14:40:28 yeah - I noticed too, but the flashing worked notheless Aug 12 14:40:37 funny Aug 12 14:41:10 not so funny is that apparently most decent SIM cards to not cooperate with the TI GSM chip ... Aug 12 14:41:28 bug #666 Aug 12 14:42:39 well, at this point I time I can not even worry about that... Aug 12 14:46:20 ah Aug 12 14:46:27 GOD DAMNIT Aug 12 14:46:36 the problem must be the "can not claim interface 2" Aug 12 14:46:58 I can't copy shit off the SD card without this error: http://pastebin.ca/654520 Aug 12 14:47:01 ARGH! Aug 12 14:47:38 rtyler: rip the ext2load exec off the neo onto a computer and run file ext2load Aug 12 14:48:07 xzcvczx: that's great in there, but I can't get the damn thing to connect properly to my Linux box as is :P Aug 12 14:48:21 isn't the ext2load in the u-boot silliness anyways? Aug 12 14:48:22 hahaha, me not root... sudo does the trick Aug 12 14:48:43 SpeedEvil: following up from the LCARS link earlier... abuse the accelerometers to trigger "Whoooooooooooosh!" sounds for Star Wars fans... Aug 12 14:48:51 * hhf423_ub hits himself with a root Aug 12 14:48:57 rtyler: not being run while your booted i wouldnt think Aug 12 14:49:10 ok ignore that Aug 12 14:49:16 unless your using a custom image Aug 12 14:49:45 I'm using one the one linked from the wiki from when you first have to put rootfs on the device Aug 12 14:50:22 nowwww we re talkin Aug 12 14:50:30 Stephmw: There is much work on core applications to be done before that. Aug 12 14:50:38 Light-sabre sounds, for example. Aug 12 14:50:59 rtyler: are you using it when booted? Aug 12 14:51:26 I'm using it in u-boot Aug 12 14:51:33 oh ok Aug 12 14:51:34 lol Aug 12 14:51:43 why are you trying to load it there? Aug 12 14:52:07 because I'm loading the rootfs, as well as uImage since dfu-util won't work for me Aug 12 14:52:14 SpeedEvil: oh, I wasn't implying it should go to the top of the priority list :D Aug 12 14:52:30 SpeedEvil: I'm sure some fan will have that done without needing encouragement Aug 12 14:54:29 I'll try something different after I get in from a smoke, waiting for my images to rsync to a public site Aug 12 14:54:34 * rtyler stomps off Aug 12 15:03:08 so, all flashed. great Aug 12 15:03:55 just out of curiousity, I haven't been following much lately, what's the status of the software? Aug 12 15:04:58 mega awesome Aug 12 15:05:00 far from usable ... Aug 12 15:06:19 * hhf423_ub is booting up Aug 12 15:06:24 :) Aug 12 15:06:44 summatusmentis: if you want a portable xterm it's great - mega awesome mobile terminal emualtor :-) Aug 12 15:06:45 rtyler: of course it's mega awesome, this is openmoko we're talking about Aug 12 15:06:54 * hhf423_ub calibrates screen Aug 12 15:07:01 rxr: so even the GPE backend isn't usable really? Aug 12 15:07:07 GPE ? Aug 12 15:07:08 * summatusmentis is used to GPE on the zaurus Aug 12 15:07:15 I saw no GPE in OpenMoko Aug 12 15:07:20 I assume that's what openmoko is using for X? Aug 12 15:07:32 Basically. Aug 12 15:07:34 yes, with crappy, incomplete custom widgets Aug 12 15:07:36 the pictures on flickr suggest it's gpe Aug 12 15:07:40 where barely anthing works Aug 12 15:07:46 It functions just fine as a small desktop linux box. Aug 12 15:07:53 I see Aug 12 15:07:56 if you expect to make phonecalls wait some month, flash GPE or help coding ... Aug 12 15:07:57 with all the peripherals working as you would expect. Aug 12 15:08:09 However. Aug 12 15:08:20 The phone software is at the moment, minimal and buggy. Aug 12 15:08:25 yeah - adjusting ALSA volume over ssh and initiating calls after manually starting the gsmd ... Aug 12 15:08:39 I can make calls using the dialer. Aug 12 15:08:48 I could in the .1 version Aug 12 15:08:49 rxr: that's how _real_ users make calls :-) Aug 12 15:08:55 in the .2 version I do not even get the PIN dialog ... Aug 12 15:09:20 but to be honest I do not care, I'm right now working on my own EFL (enlightenment 17) based apps ... Aug 12 15:09:44 has there been a solution for finger based text entry yet? or is that planned for once the base apps are working? Aug 12 15:09:49 summatusmentis: the hardware is mostly good, only that the TI chip does not accept most recent SIM cards Aug 12 15:09:52 see bug #666 Aug 12 15:09:57 now, how do I make it register on the GSM network? it does not ask for my PIN? Aug 12 15:10:00 I hope FIC and TI get this fixed with some firmware update Aug 12 15:10:18 hhf423_ub: in the .2 software it doesn't ask me for a pin Aug 12 15:10:23 with the .1 software it did ... Aug 12 15:10:32 hhf423_ub: also most images do not set the audio levels Aug 12 15:10:39 * rtyler is tempted to get a debug board now Aug 12 15:10:41 ah so thats normal? Aug 12 15:10:50 you have to either use the barely useful rxvt and a lens or ssh the load some useful audio levels Aug 12 15:11:01 so how can I tell the SIM my PIN so that it can register on the network? Aug 12 15:11:03 and yes, I knwo this is just rediculous for this old project Aug 12 15:11:15 hhf423_ub: look the wiki for manual using GSM Aug 12 15:11:20 ok Aug 12 15:12:06 hhf423_ub: note I'm not a FIC / OpenMoko person, I'm just a 2 day phone owner who is just starting to write an alternative sofware stack Aug 12 15:12:25 rxr: why an alternative stack? Aug 12 15:12:27 as apparently the OpenMoko project is not really progressing on the software site ... Aug 12 15:12:36 summatusmentis: because OpenMoko sucks from the bottom to the top ? Aug 12 15:12:39 glibc on a phone ? Aug 12 15:12:46 yet other custom Gtk+ widgets ? Aug 12 15:13:16 well, to be fair, there were a lot of issues at the early stages Aug 12 15:13:17 and after over a year of drafting it's not even starting with sensible ALSA setups or even asking for the PIN in all images ? Aug 12 15:13:30 uploaded my recently built images FWIW: http://monkeypox.org/openmoko/images/ Aug 12 15:13:38 summatusmentis: if they where sane they would at least have used GPE or so ... Aug 12 15:13:47 but not reinvent the Whole Gtk+ mobile application game yet again ... Aug 12 15:14:04 Fork off if you don't like it. Aug 12 15:14:08 rxr: when your project is walking before the end of the month you can start shitting on openmoko but i think you should just not be so quick to abuse it when you dont have a working alternative yet Aug 12 15:14:09 rxr: do I have to go through the manual GSM stuff to make the phone register on the network? Aug 12 15:14:22 hhf423_ub: depends on what images you flashed Aug 12 15:14:26 right now I have a "?" next to the antenna symbol Aug 12 15:14:27 on some the UI asked for a PIN Aug 12 15:14:34 in the .2 images I was not asked and had to do it manually Aug 12 15:14:43 of course in a pefect world the UI should ask you Aug 12 15:15:13 xzcvczx: you speek with a 10 year Linux coder here Aug 12 15:15:43 xzcvczx: I did and contributed to a lot of projects Aug 12 15:16:21 hmm Aug 12 15:17:47 SpeedEvil: btw. noticed how many people these days have a phone and are in this channel, and yet the software is far from a useable state ? Aug 12 15:19:38 I find this strange at best Aug 12 15:19:39 rxr: i dont care how many years you have been coding linux for, and development will move faster if you contribute to openmoko not start your own project against openmoko Aug 12 15:19:43 rxr: i thought it was clear that P1 would by no means be in a useable state? Aug 12 15:20:11 drath: I do not care that _I_ have no software, I'm not intested in it anyway Aug 12 15:20:33 but there is something very wrong when a project that large, with some many people "involved" does not have much to show off Aug 12 15:20:33 rxr: heard the saying, dont reinvent the wheel? Aug 12 15:21:12 rxr: well most of them are due to experienced coders like yourself jumping ship to a competing project so they both get completed half as fast Aug 12 15:21:16 well, all sorts of custom moko Gtk+ widgets clearly is reinventing the wheel Aug 12 15:21:52 GTK on a 2.8" display is not pleasant. Aug 12 15:22:00 SpeedEvil: ? Aug 12 15:22:04 SpeedEvil: i am tempted to try kde :P Aug 12 15:23:19 but maybe the problem comes from the fact that there are so many thing done behind the communities back in "upcomming" secret announcements ... Aug 12 15:23:47 that secret stuff is mostly about upcoming hardware platforms, rather than openmoko itself, isn't it? Aug 12 15:23:56 no idea Aug 12 15:23:59 No. Aug 12 15:24:14 The secret stuff to some extent has already been a problem. Aug 12 15:24:24 the last thing I read is "new famour graphic artist" to be comming and what not Aug 12 15:24:30 For example, with GTA02, it's lead to not being able to wake on bluetooth. Aug 12 15:24:49 Or at least that could potentially have been avoided. Aug 12 15:25:04 btw. are the whole hardware schematics somewhere ? Aug 12 15:25:08 sorry, but open source doesn't necessarily mean open company Aug 12 15:25:08 No,. Aug 12 15:25:14 I know. Aug 12 15:25:28 some things are strategic and marketing blurb Aug 12 15:25:43 which is... unfortunately, but i have to admit it... necessary to keep funding going Aug 12 15:26:02 Indeed, we all understand this, it's just disagreeing about the balance. Which is largely irrelevant, as it's managment driven anyway. Aug 12 15:26:08 mickeyl: are there hopes bug #666 can be resolved by a future gsm firmware update ? Aug 12 15:26:14 !ombug 666 Aug 12 15:26:15 * * Bug 666, Status: NEW, Created: Unknown Aug 12 15:26:16 * * steven.demetrius(AT)fiwwi.com: 3G SIM capability Aug 12 15:26:17 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=666 Aug 12 15:26:17 that is if you know anything about it Aug 12 15:26:41 * drath finds it rather amusing that bug got id 666 assigned :) Aug 12 15:26:48 drath: /me too Aug 12 15:26:52 sorry, i don't know anything about that. Best I can do is promise to pester the hardware team to comment on it Aug 12 15:27:01 Good enough. Aug 12 15:27:15 * SpeedEvil hopes it's fixable without RMAs. Aug 12 15:27:21 * mickeyl too Aug 12 15:27:24 SpeedEvil: isnt that issue with the 02 you mentioned hardware thought anyway? Aug 12 15:27:50 xzcvczx: yes - however, if the requirement was put in early enough, the hardware would have been relatively easy to change. Aug 12 15:28:00 (As I understand it). Aug 12 15:28:07 At this stage - it's basically cast in stone. Aug 12 15:29:37 * xzcvczx tries bitbake gcc Aug 12 15:30:04 Rene, you obviously overestimate the amount of people that have been working on the project. Aug 12 15:30:20 mickeyl: do you happen to know if TI is still unwilling to allow users to reflash? Aug 12 15:30:23 mickeyl: you haven't had any problems getting Linux to bloody recognize the device? Aug 12 15:30:32 and obviously underestimate the amount of internal restructuring and problems the coreteam has faced. Aug 12 15:30:36 rtyler: Try another system. Aug 12 15:30:37 rtyler: does it show up in lsusb? Aug 12 15:30:39 but anyways, feel free to come up with something own Aug 12 15:30:46 xzcvczx: of course not >_< Aug 12 15:30:54 rtyler: occasionally I've found some USB devices don't like some computers. Aug 12 15:30:55 i guess that means we no longer need to talk about doing some work together on it? Aug 12 15:30:59 rtyler: tried a different usb port? Aug 12 15:31:08 SpeedEvil: I wish I had more Linux boxen around, but my Mac and Linux box are it :/ Aug 12 15:31:08 mickeyl: from this channel list and the mailing list there are many people to count Aug 12 15:31:12 xzcvczx: yep, no dice Aug 12 15:31:21 rxr: we're trying to fix the bug with the help from TI though. Aug 12 15:31:30 yes, and those many people had their phones for like... 10 days Aug 12 15:31:30 rtyler: windows will find it. Aug 12 15:31:30 SpeedEvil: it worked once upon a time, I did the first couple of flashes from this machine Aug 12 15:31:32 big deal Aug 12 15:31:41 rtyler: it will ask for a driver CD Aug 12 15:31:48 (or whatever it does nowadays) Aug 12 15:31:53 * jserv2 just reviewed the firmware last week. Aug 12 15:31:54 what will ask for a driver CD? Aug 12 15:31:56 rtyler: does other usb stuff work on that machine Aug 12 15:32:01 rtyler: windows Aug 12 15:32:19 * mickeyl shrugs and continues cleaning up the office Aug 12 15:32:25 xzcvczx: yep, that's how I've gotten stuff onto the SD card ;) Aug 12 15:32:36 * xzcvczx now knows what they are wasting time doing.... cleaning up their offices Aug 12 15:32:42 just joking mickeyl Aug 12 15:32:44 :P Aug 12 15:33:05 * rtyler waits for mickeyl to autoop and kickban xzcvczx :P Aug 12 15:33:14 lol Aug 12 15:33:33 * xzcvczx gets on his knees and prays to mickeyl for forgivness Aug 12 15:33:40 oh shit Aug 12 15:33:46 * xzcvczx hides Aug 12 15:33:55 heh. takes more than that Aug 12 15:33:57 mickeyl: you now, I even was around at the computex, but harald did not wanted to meet with us although FIC desperatedly wanted to ... Aug 12 15:34:02 Well - they are wasting their time on random stuff. The blog posts have made that _Very_ clear. Aug 12 15:34:08 did someone get scummvm compiled yet? Aug 12 15:34:29 I mean - ordering computers and sorting out cabling - madness. Aug 12 15:34:33 ScaredyCat: yes. Aug 12 15:34:43 can I have a copy :) Aug 12 15:34:55 rxr: that's not correct information. We just arrived after 13h with lots of jetlag and weren't fancy to go to a business meeting like zombies Aug 12 15:34:55 heh Aug 12 15:34:59 What is the 'web' thing based on? Aug 12 15:35:00 ScaredyCat: That person should send me bb-file and patches today. Aug 12 15:35:14 ahhh goodie good Aug 12 15:35:45 ScaredyCat: unfortunately he has been partying till late last night, so it might take a while... ;) Aug 12 15:35:49 mickeyl: and ? Aug 12 15:35:52 gah! Aug 12 15:35:58 slacker! Aug 12 15:36:11 * xzcvczx returns from hiding Aug 12 15:36:22 rxr: nothing and. that's the story. Aug 12 15:36:35 i didn't know what you were into at that time anyways Aug 12 15:36:43 we had over 4 days or so to meet in .tw ... Aug 12 15:37:10 mickeyl: nothing I had other business, but the FIC people appeared to really wanted help from more people it was their idea to meet ... Aug 12 15:37:12 you can't imagine my schedule. i was the first time in .tw then Aug 12 15:37:23 mickeyl: I likewise ... Aug 12 15:39:18 Hey was DFU utils ever fixed to accomidate other endains? Aug 12 15:39:32 so a meeting that didn't happen for whatsoever reasons is the reason you are bashing the project? sounds a bit strange to me Aug 12 15:39:40 mickeyl: just watched some videos, any idea on a if/when directfb or whatever you were playing with would come in? Aug 12 15:39:56 calamous_: we don't support other endian tribes last I heard ;) Aug 12 15:39:59 The proper QVGA patch would be nice. Aug 12 15:40:14 rtyler: it's there. you can install it, but there isn't any specific OM work going on there based on it. Aug 12 15:40:21 gotcha Aug 12 15:40:29 rtyler: so it only works on little endian? Aug 12 15:40:41 we are going to introduce non-GTK+ stuff into the platform somewhere early next year Aug 12 15:40:44 Then 'all' that remains to be done is to write an optimal rotator. Aug 12 15:40:45 from my experiences, I've read it doesn't, haven't tried it on say, a PPC box Aug 12 15:40:49 rtyler: can't we go through the source and fix it? Aug 12 15:40:54 calamous_: have fun :D Aug 12 15:43:13 mickeyl: I do not bash the project Aug 12 15:43:22 rtyler: I think I saw it being changed by harald to fix some endianess issues some time ago Aug 12 15:43:28 rtyler: on the commitlog list Aug 12 15:43:29 mickeyl: in fact I help quite some new users as good as I could the last days Aug 12 15:44:27 rxr: ok, sorry then, i just seen lines like "and after over a year of drafting it's not even starting with sensible ALSA setups or even asking for the PIN in all images ?" and " btw. noticed how many people these days have a phone and are in this channel, and yet the software is far from a useable state ?" Aug 12 15:44:35 some people would think that counts as bashing Aug 12 15:44:45 *shrug* Aug 12 15:44:58 * rtyler hugs mickeyl Aug 12 15:45:02 I still likes you :) Aug 12 15:45:10 * mickeyl grins Aug 12 15:45:27 mickeyl: it is not that shipping images that ask for the PIN and loads some mixer values should be too much of a problem Aug 12 15:45:36 HOLY SHIT Aug 12 15:45:42 Linux box recognizes it now Aug 12 15:45:59 rtyler: what fixed it? Aug 12 15:46:10 rxr: expecting this to be release quality software is not realistic. Aug 12 15:46:12 rtyler: oh btw did you check it was plugged in :P Aug 12 15:46:18 rxr: It's not, it's pre-alpha. Aug 12 15:46:19 xzcvczx: yes i did -_- Aug 12 15:46:29 mickeyl: I could also say "that's how a linux phone should be operated like", but probably defect stuff should be pointed out Aug 12 15:46:32 rtyler: lol what got it working? Aug 12 15:46:33 xzcvczx: plugging it into a different set of ports on a different USB controller -_- Aug 12 15:46:40 rtyler: lol Aug 12 15:46:45 rtyler: nice Aug 12 15:47:01 [- - -] <-- ports I *was* trying ports that worked -> [- -] Aug 12 15:47:21 * rtyler writes Linus a strongly worded email Aug 12 15:47:41 *sigh* so you don't think we know about these problems? Aug 12 15:47:43 lol i doubt its linus's fault..... the ports working at all? Aug 12 15:47:49 mickeyl: has anybody started working on getting the OpenMokoFramework to speak SyncML? Aug 12 15:47:58 xzcvczx: yeah, those ports were working fine Aug 12 15:48:09 I think the usb subsystem just got all gunked up on that bus Aug 12 15:48:10 rtyler: funambol hopefully, but i'm not up to date as to their status Aug 12 15:48:31 mickeyl: this are "one-liner" problems we talk about that make the alpha-quality software even harder to use ... Aug 12 15:48:32 mickeyl: Global Locate - any progress? Aug 12 15:48:33 are they going to be a little more open source? I saw registration forms all over their site for access content? Aug 12 15:49:16 as long as the device speaks proper SyncML i'll be happy, and I'll start hacking on a sync desktop app ^_^ Aug 12 15:49:42 mickeyl: I got some stuff todo, when you need one more hand you probably know where you can find me ... Aug 12 15:50:07 xzcvczx: I've never been so excited to see USB jibberish in syslog before :P Aug 12 15:50:08 we didn't exactly have a good start, but yes, lets talk at some other point of time Aug 12 15:50:14 two days of absolute frustration Aug 12 15:50:19 oh and mickeyl: why was glibc choose as C library ? Aug 12 15:50:27 to spite you! muharharhar Aug 12 15:50:40 mickeyl: we as we both or openmoko :-) ? Aug 12 15:50:48 rxr: all :D Aug 12 15:50:51 lol Aug 12 15:51:09 rxr: as for glibc, because we wanted to keep the amount of variables at a low minimum Aug 12 15:51:22 it's not as if anything is set in stone at this point of time Aug 12 15:51:28 this is an iterative process Aug 12 15:51:46 variables as in "when porting desktop software to the phone platform" I guess ? Aug 12 15:51:51 FWIW I think glibc was the right choice - at this point Aug 12 15:51:55 yes. Aug 12 15:52:01 there are different opinions when it comes to a phone Aug 12 15:52:07 see the GMAE Aug 12 15:52:13 certainly makes sense with Gtk+ imgo Aug 12 15:52:15 imho too Aug 12 15:52:25 glibc has a large memory and other footprint for many phones. However, this is not an especially small phone. The laptop I'm typing on has the same RAM. Aug 12 15:52:51 And I do not (often) expect to have 34 firefox tabs open on the phone. Aug 12 15:53:02 yeah - but with 20%++ saved RAM you can do more interesting things :-) Aug 12 15:53:04 whatever Aug 12 15:53:32 are you kidding me? I'm going to use my phone as a mobile statistics analysis platform ;) Aug 12 15:53:32 you may never get to such things without glibc Aug 12 15:53:48 * rtyler swaps out Aug 12 15:53:50 Writchie: even KDE compiles with uClibC .-) Aug 12 15:54:39 no way doubts the uClibC or whatever could be made to work, the point is we have enough memory and glibc lets many more apps work with minimum mods Aug 12 15:54:48 thats the whole point behind the VGA Aug 12 15:55:04 IMO - VGA is 1/10th of the battle. Aug 12 15:55:17 Yes, you can run an app with unusably small widgets. Aug 12 15:55:20 (for a phone). Aug 12 15:56:13 I think the point is that you get to sort of working very fast. Yes, you have to rework for size, but the speed at which you get something WORKING encourages continued work on the opimal solution. Aug 12 15:56:25 Indeed. Aug 12 15:56:25 Anyway, that's what I've noticed. Aug 12 15:56:46 The footprint does make be puke Aug 12 15:56:48 * mickeyl nods to Writchie Aug 12 15:57:06 but memory is cheap and Moore's law still reigns Aug 12 15:57:08 some even think debian compatibility is a must Aug 12 15:57:17 which kind of rules uClibC out as well Aug 12 15:57:41 "debian compatibility" ? Aug 12 15:58:01 I would like to see a better framebuffer level and switching between X and straight framebuffer Aug 12 15:58:13 Ctrl+Alt+F2? :P Aug 12 15:58:14 s/level/layer Aug 12 15:59:06 the handset/pda mode switching needs a better think Aug 12 15:59:41 the architecture should support a phone with built-in PDA or a PDA with built-in phone Aug 12 15:59:49 two somewhat different philosophies Aug 12 15:59:56 with pervasive implications Aug 12 16:00:25 great, quilt 0.35 and 0.36 are different enough to break MokoMakefile Aug 12 16:02:47 Kensan: who was it btw, perhaps I can help encourage Aug 12 16:06:55 ScaredyCat: it's JadeFalcon. I wrote him a text-message, but I believe he is still recovering ;) Aug 12 16:07:20 can't hack the pace eh? :P Aug 12 16:08:08 * cesarb is on the PDA-with-phone camp Aug 12 16:08:33 i'm definitely in the phone with PDA camp Aug 12 16:08:57 coexistance is possible Aug 12 16:09:31 Writchie: especially it being very easy to reflash the device with a completely different distribution ;-) Aug 12 16:09:55 reflashing, unfortunately, can't overcome hardware screwups Aug 12 16:10:18 Writchie: so the PDA-with-phone people like me can have their crazy memory-gobbling glibc-plush-bash-and-coreutils, and the phone-with-PDA can chose ulibc-and-busybox Aug 12 16:10:24 Writchie: since in the openmoko architecture you control the application processor, the thought of pda with phone capabilities is probably predominating. Aug 12 16:10:33 whooo, at+cpin OK Aug 12 16:11:19 the phone with PDA is definitely more prevalent, but I think this is mainly because of lack of knowledge is the GSM space. Aug 12 16:11:25 which BTW is not an insult Aug 12 16:12:19 s/is/in/ Aug 12 16:12:19 Writchie meant: which BTW in not an insult Aug 12 16:12:56 Writchie: I agree. I was more speaking of how I see the mindset of most (community) devs. Aug 12 16:12:58 dumb bot Aug 12 16:13:13 yes - the mindset often gets in the way Aug 12 16:13:25 hmm, should "autoregister with network" work? Aug 12 16:13:59 good engineering process should not be forsaken just because its open source Aug 12 16:14:09 Writchie: I think more than lack of GSM knowledge, it's the lack of a full-featured gsmd which is problematic Aug 12 16:14:25 gsmd at this point is just a concept demo Aug 12 16:14:28 Writchie: since with a full-featured gsmd people would simply use it Aug 12 16:15:00 yes, the handset (or MT) layer really needs to be done well Aug 12 16:15:10 and the MMI can be left to whatever Aug 12 16:16:09 i think much of the disappoint at this point comes from not being able to use the device as a basic phone. Aug 12 16:16:20 which hurt's creditbility Aug 12 16:16:26 when you show the device Aug 12 16:16:27 * Dmitry_Platonov agrees with Writchie Aug 12 16:16:42 gsm+audio Aug 12 16:16:47 unfortunately, most people are completely unable to visualize Aug 12 16:17:03 gest, gsm+audio + basic mobility Aug 12 16:17:26 so you think we should have postponed selling the hardware even more? Aug 12 16:17:29 phew Aug 12 16:17:37 no- Aug 12 16:17:48 aha. so? Aug 12 16:17:51 i fully agree with letting the hardware go Aug 12 16:17:51 i fail to see options now Aug 12 16:18:10 there are always options Aug 12 16:18:21 mickeyl, just put more effort into gsmd/dialer/sms Aug 12 16:18:46 heh, yeah mickeyl, we in IRC have a better handle on how you should be spending your time ;) Aug 12 16:18:52 the project needs a focus on gsmd Aug 12 16:19:01 there are capable people willing to help Aug 12 16:19:01 yes, the basis things, dialer (pim) and audio settings shoudl be done right Aug 12 16:19:17 and the images only release whit this basic functions tested Aug 12 16:19:26 well Aug 12 16:19:30 we didn't release any images Aug 12 16:19:32 rxr: no, the images should be automatic daily builds Aug 12 16:19:37 so i fail to see the point of this complaint Aug 12 16:19:43 i think we need a quick and dirty dbus layer for the phone Aug 12 16:19:45 rxr: since it's the best way to find bugs Aug 12 16:19:51 ok - whatever Aug 12 16:19:51 to buy time to get gsmd right Aug 12 16:20:05 * xzcvczx is surprised mickeyl is still here Aug 12 16:20:05 then at least the basic one-liner bugs should be addressed Aug 12 16:20:13 you and your one-liner bugs Aug 12 16:20:21 right now most people get an image that does not set audio, does not ask the pim or doesn#t start gsmd at all Aug 12 16:20:22 where are your bugtracker attachements? Aug 12 16:20:33 anybod had dfu-util time out? http://pastebin.ca/654600 Aug 12 16:20:35 the p1 snapshot release image IS NOT READY Aug 12 16:20:39 mickeyl, complaint is that gsmd does not see much improvements, and unlikely to be in good shape in october. Aug 12 16:20:59 mickeyl: note that I didn't start this time :-) Aug 12 16:21:03 right. Aug 12 16:21:16 rtyler: the phone is still on? Aug 12 16:21:18 l8er. i don't want my good mood today to be destroyed even more Aug 12 16:21:22 yes Aug 12 16:21:31 mickeyl: I just want to say "making sure that just pin entering and dialing works and audio elvels" in most images would already be a big step Aug 12 16:21:40 learn when to shut up you lot Aug 12 16:21:57 hm Aug 12 16:22:48 I have a "?" next to the GSM antenna symbol, anyone knows what that means? Aug 12 16:22:57 rtyler: your usb stack seems fucked Aug 12 16:23:00 hhf423: you're screwed Aug 12 16:23:06 xzcvczx: no kidding >_< Aug 12 16:23:09 hhf423: means that the gui isnt connected to the network Aug 12 16:23:11 and ATE1 does not get ERROR, but gets OK Aug 12 16:23:15 at least, he listen. kudos to him. Aug 12 16:23:15 those damned default ubuntu kernel images Aug 12 16:23:17 rtyler: you have rebooted i assume Aug 12 16:23:37 xzcvczx: moving around USB ports again and replugging seems to have fixed it, ROFL Aug 12 16:23:57 Dmitry_Platonov: yeah but he doesnt come here just to get why doesnt this work/that work/this is the most important thing i am sure they have an internal roadmap that they are working as hard as they can to stick to Aug 12 16:24:06 and my PIN does not get acceppted I think, AT+CPIN="mypin" does get ERROR Aug 12 16:24:26 does your sim have a pin? Aug 12 16:24:29 hhf423: if your PIN is not accepted you might face bug #666 Aug 12 16:24:34 well your network Aug 12 16:25:00 well whatever part the pin is used for as we dont have pins here Aug 12 16:25:23 xzcvczx, if they made "internal" roadmap not so internal, there would be not so many complaints. Aug 12 16:25:26 * mwester thinks that some of those who complain about the Neo should give their phones to some of those who would be delighted to have one, regardless of usability. Aug 12 16:25:41 yes Aug 12 16:25:48 Dmitry_Platonov: also if he only gets shit when he comes in here hes not going to listen any more Aug 12 16:25:48 * mwester puts his name on the head of the "delighted to have one" list... Aug 12 16:25:55 heh Aug 12 16:25:55 mwester: me! me! me! I promise not to complain (much) :) Aug 12 16:26:07 :) I'm first in line! Aug 12 16:26:27 the noise level is very high because of expectations that got of hand Aug 12 16:26:29 xzcvczx, what he gets here is not "shit", it's community feedback Aug 12 16:26:33 although I admit to not even having ordered a P1 due to announced lack of functionalityf Aug 12 16:26:52 80% of input is crying about AT&T SIMs Aug 12 16:27:09 aaaaah Aug 12 16:27:11 /data/openmoko/openmoko/tmp/work/fic-gta01-angstrom-linux-gnueabi/libgsmd-0.1+svnrNOT IMPLEMENTED-r18/ Aug 12 16:27:12 this is a developer preview Aug 12 16:27:14 yes, this SIM compatibility issue is a bummer ... Aug 12 16:27:16 cant someone remove the blank here? Aug 12 16:27:19 That was an unexpected problem, and a very serious issue. Aug 12 16:27:23 so go get t-mobile Aug 12 16:27:27 ah, cool, I can dial Aug 12 16:27:27 it isnt compiling because of this for 5 days now... Aug 12 16:27:28 and do some development Aug 12 16:27:31 mickeyl, just put more effort into gsmd/dialer/sms <----- i am sure its not his choice what he is spending time on. Aug 12 16:27:46 I can dial even though AT+CPIN returned ERROR Aug 12 16:27:57 that's also not even the level of the project he's really responsible for Aug 12 16:27:59 http://wiki.openmoko.org/wiki/User:Mickey Aug 12 16:28:08 don't be such bullies next time guys Aug 12 16:28:13 i fail to see options now Aug 12 16:28:43 it's an option Aug 12 16:31:18 omg omg openmoko might be finally building on my Mac Aug 12 16:31:20 anybody know where the source for the steppingstone boot code is located? Aug 12 16:31:47 samsung. Aug 12 16:31:52 It's in ROM. Aug 12 16:32:08 in ROM loads the first page of flash - the source for the first page of flash Aug 12 16:32:43 Ah. Aug 12 16:32:43 GAK, Python errors! Aug 12 16:32:47 It's u-boot isn't it? Aug 12 16:32:56 I don't think so. Aug 12 16:33:10 there is a 4K loader that loads u-boot Aug 12 16:33:17 ah Aug 12 16:33:18 the PM wakeup has to be done there Aug 12 16:33:23 Yeah. Aug 12 16:33:33 i think I see a way to make the BT wakeup work Aug 12 16:33:46 Impossible AIUI. Aug 12 16:33:50 and divide the PM and boot functions Aug 12 16:33:55 There is no interrupt line connected at the moment. Aug 12 16:34:01 don't need one Aug 12 16:34:04 just need a PIO Aug 12 16:34:06 Writchie: I'm working on it Aug 12 16:34:09 Or PIO. Aug 12 16:34:21 wake up every second and look around Aug 12 16:34:23 The 'wakeup' line is not brought out to the chip. Aug 12 16:34:37 yoyo_: working on the boot code? Aug 12 16:34:57 i think I see a way to make the BT wakeup work Aug 12 16:35:21 Rather. Aug 12 16:35:37 on GT01 there are two lines brought out Aug 12 16:35:40 The 'wakeup' line is not brought out through the connector of the bluetooth device onto the main PCB. Aug 12 16:35:40 to PIO Aug 12 16:35:41 * cesarb hopes there will be no problems with brazilian SIM chips... as it doesn't look like anyone in Brazil got a phone yet, or at least nobody in Brazil wrote about it on the wiki Aug 12 16:36:08 Writchie: however, it's been stated that these are bluetooth/wifi coexistance of some sort. Not wakeup. Aug 12 16:36:12 cesarb: this does not appear to be region based, I think it hasa to do with the generation of SIM card Aug 12 16:36:23 not in GT01 Aug 12 16:36:30 cesarb: if your SIM is older than 6m than it will probably work Aug 12 16:36:31 Writchie: As waking up from bluetooth wasn't thought to be interesting for GTA02. Aug 12 16:36:49 on GT02 it depends on whether one of these line makes it to a GPIO Aug 12 16:36:51 SpeedEvil: can you stop ranting :-) Aug 12 16:37:08 Writchie: it's been stated that it's not. (for GTA02) Aug 12 16:37:26 rxr: well, IIRC I have my Razr for more than a year, so the SIM is older than 6 months... Aug 12 16:37:42 I'm not ranting. I'm responding to Writchies points, with paraphrases from mails from Harald to the hardware list. Aug 12 16:37:45 cesarb: probably that SIM shoudl work then Aug 12 16:38:00 cesarb: but I do not garuanteee, I'm an extern newcommer here ... Aug 12 16:38:05 (no TI inssurance company :-) Aug 12 16:38:14 SpeedEvil: that was a joke .-) Aug 12 16:38:57 harold was going to look into it for GTA02 - but its doubtful Aug 12 16:39:04 but GTA01 is doable Aug 12 16:39:09 and probably GTA02 Aug 12 16:39:23 s/GTA02/GTA03 Aug 12 16:40:08 IMHO GTA03 will need both BT Class 1 and WiFi and new TI chip to support UMA Aug 12 16:40:43 UMA == what ? Aug 12 16:41:06 unlicensed mobile access, running GSM through the Internet Aug 12 16:41:12 UMA works today Aug 12 16:41:13 rxr: http://en.wikipedia.org/wiki/Unlicensed_Mobile_Access Aug 12 16:41:18 ah Aug 12 16:41:29 rxr: The Wiki Has All The Answers :D Aug 12 16:41:53 yeah, tried wiki.openmoko.org and id did not had UMA .-) Aug 12 16:41:53 unfortunately, many of the answers are wrong ;) Aug 12 16:41:58 lol Aug 12 16:42:00 oh, UMA soudns very interesting Aug 12 16:42:07 anyone knows how the chaos camp is/was going ? Aug 12 16:42:18 re Aug 12 16:42:22 bi Aug 12 16:42:25 UMA benefits both the operators and the users Aug 12 16:44:28 yoyo_: are you working on the PM? Aug 12 16:44:42 what is PM? Aug 12 16:44:46 P Aug 12 16:44:50 power management - wakeup Aug 12 16:44:50 woer managment. Aug 12 16:45:27 Writchie: I'm working on a UI manager (high level) not low level Aug 12 16:45:40 OK tnx Aug 12 16:46:16 wake up on BT sux Aug 12 16:46:27 how? Aug 12 16:46:34 generally - or on NEO Aug 12 16:46:53 heh.. paranoid moko phone went nutz waking up on every bt request :) Aug 12 16:47:20 Writchie: on the planet in many stupid people which don't even now what is a bluetooth Aug 12 16:47:50 true - but how does that make BT wakeup suck Aug 12 16:47:58 Writchie: and the send everything to everything Aug 12 16:48:13 you don't have wakeup from them Aug 12 16:48:22 you only wake up from your devices Aug 12 16:48:28 if you implement it right Aug 12 16:48:35 Writchie: you can't know which device it is before you wake up... Aug 12 16:48:41 of course you can Aug 12 16:48:49 Writchie: there is a possibility to get some junk Aug 12 16:49:25 Writchie: can the BT chip filter so you will only wake up on a page scan from a list of approved devices? Aug 12 16:49:35 yes Aug 12 16:49:42 Writchie: cool Aug 12 16:50:32 bluecore4 is about the 7th generation BT chip Aug 12 16:50:49 * cesarb wonders if it would be possible to wakeup not on BT, but on a 1Hz timer, and in the 12MHz mode, to poll the BT chip Aug 12 16:51:16 12Mhz mode is too much power. Aug 12 16:51:19 Writchie: ah, but is it possible to do that filtering while using HCI-USB (which is what the neo uses IIRC)? Aug 12 16:51:39 Writchie: or does the filtering need bluecore's proprietary mode? Aug 12 16:51:43 Hi! Has anyone successfully paid for GTA01 with an Amex card? Aug 12 16:52:19 you need one of the proprietary commands, IIRC Aug 12 16:52:33 not really proprietary Aug 12 16:52:40 just not vanilla HCI Aug 12 16:52:47 vendor specific is more correct Aug 12 16:53:31 if you don't have a least one wakeup line available, then you have to bring-up the USB host Aug 12 16:53:39 which will consume power Aug 12 16:53:48 I read somewhere that the phones are running at 200Mhz and there was a u-boot console command to change this to 266? Aug 12 16:53:57 quizz: C language. how to init a array of a structs in a struct? Aug 12 16:54:25 init to what? Aug 12 16:54:48 Writchie: a vendor-specific command is less annoying than having to switch to a completely different communication mode ;-) Aug 12 16:55:11 yoyo_: just use the correct nesting of {} Aug 12 16:55:56 yoyo_: with C99, you can even use named structure fields in the initializer, which is way more readable (the linux kernel does it all the time) Aug 12 16:56:35 (the linux kernel did it even before C99, using a slightly different GCC extension) Aug 12 16:57:54 rtyler: wasn't the slower speed in P0? Aug 12 16:58:44 200 was mandated as the early debug board did not work at 266 Aug 12 16:58:47 whoooaaaaa Aug 12 16:58:52 monster storm Aug 12 17:00:07 cesarb: not sure Aug 12 17:02:38 I have struct with have a array of objects I need to init it again biger Aug 12 17:04:41 linked lists FTW Aug 12 17:11:55 yoyo_: if you need to resize the array, either use the trick of having it at the end of the struct (with 0 elements, or with 1 elements if you want to follow the old stardards), allocate the struct as its sizeof plus the size of the array, and use realloc later to resize it, or use no trick at all and allocate the array with malloc (so you can realloc it later) Aug 12 17:12:05 yoyo_: but that's sort of OT for this channel Aug 12 17:21:35 SpeedEvil: I'm no longer convinced that wakeup from BT via USB is not reasible. Aug 12 17:22:00 s/reasible/feasible Aug 12 17:26:02 see a cinema time Aug 12 17:45:29 is libgtkstylus documented anywhere? Aug 12 17:46:15 Writchie: you mean basically by micro-wakeups, and rapid polling? Aug 12 17:46:24 That'd be useful for many other things. Aug 12 17:47:01 SpeedEvil: basically yes. I'm looking at splitting the bootstrap into wakeup and coldboot sections Aug 12 17:47:17 and it looks like we may be able to detect a resume from the USB hardware Aug 12 17:47:21 at the asm level Aug 12 17:47:28 before really waking up the cpu Aug 12 17:47:41 s/cpu/linux Aug 12 17:47:54 we have 4K and 2 32bit regs to work with Aug 12 17:48:11 polling resolution will be only 1 second Aug 12 17:48:19 because of the funky RTC Aug 12 17:48:55 power off is really suspend Aug 12 17:49:14 a warm start is when to two 32 bit regs are zero Aug 12 17:49:33 we need part 1 and part 2 wakeup code. Aug 12 17:49:49 part 1 in the stepping stone Aug 12 17:49:59 part 2 in RAM Aug 12 17:50:28 we can almost doulble the memory by using another NAND block for cold boot code Aug 12 17:50:53 and overlay the image loaded by the CPU ROM Aug 12 17:51:24 D:? Aug 12 17:51:30 what are you trying to do? Aug 12 17:52:42 anrp: are you asking me? Aug 12 17:53:22 yeah Aug 12 17:53:54 trying to get rational sleep, wake, wakeup, to support BT and GSM Aug 12 17:54:21 want battery to last maybe 2 days in a BT wakeable mode Aug 12 17:54:41 at least need to wake up from BT in order to do what any BT capable phone is able to do. Aug 12 17:54:42 rational? Aug 12 17:54:57 It can wakeup on GSM AIUI already. Aug 12 17:55:00 (in principle) Aug 12 17:55:01 consume under 25ma and we wakeupable Aug 12 17:55:10 yes i hink GSM is probably OK Aug 12 17:55:46 How fast is 'normal' resume from suspend-RAM? Aug 12 17:56:05 Writchie: did you see http://www.pkblogs.com/orzlab/2007/05/openmoko-steppingstone-code.html Aug 12 17:56:06 don't know but I think its gots lots of work Aug 12 17:56:44 Kensan: thanks Aug 12 17:56:58 do you know what encoding? Aug 12 17:57:13 Writchie: maybe it's worth joining #orzlab Aug 12 17:57:33 utf-8 Aug 12 17:57:38 I don't know how #orzlab? Aug 12 17:57:44 #orzlab? Aug 12 17:57:59 is that chinese? doesn't look like japanese :/ Aug 12 17:58:13 it's chinese afaik. Aug 12 17:58:22 its showing ? for me, let me load chinese support Aug 12 17:58:41 chinese apparently Aug 12 17:59:34 Writchie: the page mentions #orzlab here on freenode. Maybe you can ask around there Aug 12 17:59:59 we are on the verge of mainland knockoffs being better than the real thing ;) Aug 12 18:00:13 :) Aug 12 18:01:01 Writchie: it seems that jserv is part of that team. Aug 12 18:01:07 At some point in the next decades, if transport doesn't get harder, wages hasve to equalise. Aug 12 18:01:16 http://translate.google.com/translate?u=http%3A%2F%2Fwww.pkblogs.com%2Forzlab%2F2007%2F05%2Fopenmoko-steppingstone-code.html&langpair=zh%7Cen&hl=en&ie=UTF8 Aug 12 18:01:40 they are equalizing, factor workers are up to $0.60/hour Aug 12 18:01:56 I think it is not difficult to transplant, So pegged guarantee simulator can be used, so she decided OpenMoko/S3C2410 play! Aug 12 18:02:33 LG built some lcd factories or something in Poland hoping for cheap labor. Well, they misscalculated. Aug 12 18:02:59 I love translated chinese Aug 12 18:03:01 EU qeualizes wages inside EU. And EU's expanding. Aug 12 18:03:25 There is a big delta across the EU. Aug 12 18:03:25 Writchie: it's funny with other languages too *heh* Aug 12 18:03:27 machine translation is actually better than most translators ;) Aug 12 18:03:28 Still. Aug 12 18:03:44 You can sort-of understand what he's talking about. Aug 12 18:04:04 After several passes through the text. Aug 12 18:04:25 its actually easy to understand if you have worked with many chinese engineers before Aug 12 18:05:25 Writchie: talk to jserv2. I also read on the ML that he is being hired to do gsm bits. Aug 12 18:06:28 yes - i think its another week or two before he starts Aug 12 18:06:37 but there is a lot of stuff on the plate Aug 12 18:07:10 * SpeedEvil wonders what 'suggest a better translation' does to translation ouput. Aug 12 18:07:43 Writchie: well try their channel, etc. jserv2 was around here a couple hours ago I think. Aug 12 18:08:02 which channel? Aug 12 18:08:08 orzlab? Aug 12 18:08:20 Writchie: the one mentioned on the site, yes. Aug 12 18:26:36 so, i've still not been able to build om2007.2 Aug 12 18:26:44 http://dev.damaestro.us/log.do_compile.4497 Aug 12 18:26:56 any ideas? Aug 12 18:31:58 daMaestro: try this patch http://pastebin.ca/654717 Aug 12 18:32:24 testing now Aug 12 18:32:40 daMaestro: i had the same problem Aug 12 18:32:55 evening Aug 12 18:32:59 from the error i could assume this is the issue Aug 12 18:33:09 what about the "declared and not used" issues? Aug 12 18:34:17 daMaestro: this problem was reported to oe bugzilla http://bugs.openembedded.org/show_bug.cgi?id=2708 Aug 12 18:34:48 thanks. Aug 12 18:34:49 daMaestro: i'm using fedora 7 too Aug 12 18:34:54 i need to get on all of the trackers Aug 12 18:35:08 rmoravcik, i have packages for dfu-util and qemu-1973 Aug 12 18:35:17 qemu-neo1973* Aug 12 18:35:27 i'm still working out what to do about a "buildsys" package Aug 12 18:35:31 want the srpms? Aug 12 18:35:39 i've not yet submitted them Aug 12 18:36:09 and if you have ideas about what to do about packages for the build sys... that would be sweet Aug 12 18:36:30 i was thinking about just porting over the makefile commands... well exposing them as global commands Aug 12 18:36:35 like openmoko-prime Aug 12 18:36:38 like openmoko-build Aug 12 18:36:47 but, yeah Aug 12 18:37:47 grrr, rmoravcik how/where can i apply this patch Aug 12 18:37:53 i don't see the source exploded anywhere Aug 12 18:38:46 sources are extracted under tmp/work/fic-gta01-angstrom-linux-gnueabi/gtk+-2.10.14-r3/gtk+-2.10.14/... Aug 12 18:41:09 ah, i figured that would be cleaned Aug 12 18:41:10 thanks. Aug 12 18:45:29 SpeedEvil: USB host is showing wakeup enabled Aug 12 18:49:20 rmoravcik, ok... it is actually building Aug 12 18:49:53 rmoravcik, http://repo.damaestro.us/fedora/personal/7/x86_64/dfu-util-2573.svn-1.fc7.src.rpm http://repo.damaestro.us/fedora/personal/7/x86_64/qemu-neo1973-2588.svn-1.fc7.src.rpm Aug 12 18:50:11 rmoravcik, instructions on how to update are in the spec, if you know rpm specs at all Aug 12 18:50:25 daMaestro: thanks Aug 12 18:50:57 rmoravcik, sure. once i figure out what to do about a buildsys package those will be yum'able Aug 12 18:52:55 wooot, I did something and the phone asked me for the PIN and now is attached to the network. no idea how I did that Aug 12 18:54:00 Writchie: ? Aug 12 18:54:26 Writchie: Keeping the USB alive is quite expensive. Aug 12 18:54:36 don't need to keep it alive Aug 12 18:54:47 How not? Aug 12 18:55:14 If it's able to wakeup the CPU, the USB controller has to be on. Aug 12 18:55:29 should be able to disconnect the USB when going into suspend Aug 12 18:55:44 Yes, in which case it won't wake. Aug 12 18:55:47 may be able to just check the OHCI registers during the peek Aug 12 18:56:01 the wakeup state on the PINS should come from the BT module Aug 12 18:56:26 But it was stated that teh pins from the module are not the wakeup pins. Aug 12 18:56:38 no D+ and D- Aug 12 18:56:39 They are BT/wifi coexistance pins. Aug 12 18:57:19 Keeping the host controller on is quite expensive - don't you have to basically have the entire chip powered on - admittedly with the CPU in idle mode. Aug 12 18:57:31 with the USB PLL going at 48MHz, ... Aug 12 18:57:36 the OHCI is not really on, in the suspend state Aug 12 18:57:54 rmoravcik: your patch for gtk+ works here for me too :) Aug 12 18:58:01 You mean there is some sort of interrupt over the D+ or D- pins, that can wake? Aug 12 18:58:31 there is a standard wakeup from suspend over the normal USB Aug 12 18:58:41 Right - is it in fact implemented. Aug 12 18:59:03 I don't recall anything about that from the datasheet. Aug 12 18:59:04 i'm not sure if implemented or if the whole state machine of the controller has to running to detect the wakeup request Aug 12 18:59:22 what datasheet? it just says one OHCI controller? Aug 12 18:59:33 unlesss you have something different Aug 12 18:59:44 yeah - I mean, it doesn't mention it as a wake source anywhere Aug 12 19:00:06 It would be under 'what can wake the CPU from various modes' table. Aug 12 19:00:14 it's part of the OHCI spec. Aug 12 19:00:27 For sure, we cannot directly wake up the CPU Aug 12 19:01:03 but if we can detect the inband wakeup request after a 1 second poll wakeup, then we're good to go Aug 12 19:01:27 we would wakeup on RTC alarm every second. Aug 12 19:01:36 take a peek and go back to sleep Aug 12 19:01:39 Does the poll request remain active till you answer it - on the USB side? Aug 12 19:01:41 unless usb wakeup Aug 12 19:01:56 And does the bluetooth module actually support this. Aug 12 19:02:06 i'm looking into it now, IIRC its a static handshake Aug 12 19:02:14 Would be nice. Aug 12 19:02:16 the BT module can support thi Aug 12 19:02:25 s/thi/this Aug 12 19:02:29 :) Aug 12 19:03:11 this is why I'm looking for the steppingstone code, because I have to test this from that level Aug 12 19:03:16 Yeah. Aug 12 19:03:36 Being able to twiddle the ROM would be nice. Aug 12 19:03:38 48K. Aug 12 19:03:47 i think there is no hope of that Aug 12 19:03:49 But committing to a mask run is _slightly_ expensive. Aug 12 19:03:56 but the 4K will work Aug 12 19:04:31 Hmm. Aug 12 19:04:33 on power up or wakeup the CPU loads the 4K from first block on the NAND and runs it Aug 12 19:04:41 Yeah, I know. Aug 12 19:04:45 * mjr settles for "hey, great if you can pull it off with reasonable power drain" and goes back to having no fucking clue if it'll work Aug 12 19:04:47 that enough Aug 12 19:05:06 It'd be nice to have some RAM. Aug 12 19:05:19 it runs from RAM so you have it Aug 12 19:05:20 But, probably not absolutely required. Aug 12 19:05:32 there are also two words for state Aug 12 19:05:35 Oh - right - I thought this was running from locked-down cache. Aug 12 19:05:50 the SDRAM is still in auto-refresh at this point Aug 12 19:06:02 we want to keep it that way if we're not really waking up Aug 12 19:06:06 Yeah, Aug 12 19:06:11 Less fragile. Aug 12 19:08:09 i'm hoping that if the linux drive is reporting wakeup enabled, that the low level bits in the OHCI controller are what they need to be. Aug 12 19:09:00 It'd be on the powerdown status bits page. Aug 12 19:09:08 Now, where did I put that datasheet. Aug 12 19:11:52 whoow.. this channel is crowded.. man!! Aug 12 19:12:01 with awesome people Aug 12 19:12:10 ...//.../....... Aug 12 19:12:27 good one :) Aug 12 19:12:50 The first page of the clock and power managment page says that everything is turned off but the RTC. Aug 12 19:13:41 but what does turned off mean. Aug 12 19:13:50 there is some state Aug 12 19:14:00 the 'on' bit is set to 0? :P Aug 12 19:14:25 off but holding state transitions to on with some state different than reset Aug 12 19:14:51 its possible if not likely the USB controller is reset on wakeup Aug 12 19:14:55 even that is not fatal Aug 12 19:15:19 we just need to be able to test if resume is being signalled before we start up the real resume Aug 12 19:15:35 yeah. Aug 12 19:16:22 s/resume/wakeup Aug 12 19:19:03 SpeedEvil: bad news, looks like resume is a pulse Aug 12 19:19:11 How long? Aug 12 19:19:15 15ms Aug 12 19:19:34 I dunno if the GPS RTC can interrupt :) Aug 12 19:19:50 yes, but only on second boundaries Aug 12 19:20:27 GPS, not P* RTC Anyway - not really practical. Aug 12 19:20:30 On a related topic. Aug 12 19:20:43 How accurately can the P* RTC be read? Is it just second? Aug 12 19:21:07 i think its seconds Aug 12 19:21:19 it counts the 32768 clock Aug 12 19:21:32 i don't think you can read the counter Aug 12 19:21:32 It's not possible to read the clock phase? Aug 12 19:21:35 :/ Aug 12 19:21:37 Oh well. Aug 12 19:22:12 Would have been nice for the GPS. Aug 12 19:22:30 i don't like the RTC Aug 12 19:22:32 grrr git-native-1.5.2.3-r4 still fails to build Aug 12 19:22:46 Being able to read the clock to within 1/6th of a ms or so would simplify intermittent reads. Aug 12 19:22:50 who needs that crap in the hardware while taking away the basics Aug 12 19:23:20 i mean the RTC on the 2410 Aug 12 19:24:08 I was really meaning the RTC on the PMU. I need to have anohter look at that. Aug 12 19:24:10 solve y2K in the hardware but take away the RTC counter - reason to kill marketing Aug 12 19:24:17 ah Aug 12 19:24:26 that should still be alive Aug 12 19:25:38 i wonder if we could just run the code in the steppingstone as another LP state? Aug 12 19:26:39 ultra slow mode Aug 12 19:27:51 That may not be a win. Aug 12 19:28:05 LP mode is much more expensive clock for clock. Aug 12 19:28:13 However, it may come up faster. Aug 12 19:28:15 why? no SDRAM, no peripheral clocks Aug 12 19:28:28 LP mode more expensive that what Aug 12 19:28:30 The CPU core. Aug 12 19:28:44 the CPU core with clocks off should be modest Aug 12 19:28:51 and a function of freq Aug 12 19:28:52 Clock for clock - the 266Mhz mode is much cheaper than 12Mhz. Aug 12 19:29:17 doesn't sound right? Aug 12 19:29:24 is this based on measurement Aug 12 19:29:32 No, based on the datasheet. Aug 12 19:30:41 err Aug 12 19:30:59 it shows 33mW in 12Mhz Aug 12 19:31:04 with busses running Aug 12 19:31:18 yeah Aug 12 19:31:27 266/12 = 22 Aug 12 19:31:31 22*33=726mW Aug 12 19:31:41 which is quite a bit more. Aug 12 19:31:47 more than reality. Aug 12 19:32:01 * juri_ waits patiently for her neo1973 to arrive. Aug 12 19:32:07 I could live with 33mW Aug 12 19:32:11 How things change without busses is another question. Aug 12 19:32:12 plus GSM and BT Aug 12 19:33:24 hopefully GTA03 will get things right - OFF-MODE is the ideal Aug 12 19:34:31 the 12mhz running in steppingstone mode could be something like a bios ACPI Aug 12 19:34:46 concerned with wakeup Aug 12 19:35:27 we also need to monitor battery once in a while and start up the system in order to shut it down completely when battery goes south of critical Aug 12 19:35:45 this is tricky with li-ion Aug 12 19:36:26 Not really. Aug 12 19:36:36 the knee can be pretty sharp Aug 12 19:37:06 you need to be sure you have enough to bring it up completely and then shut it down completely Aug 12 19:38:26 The knee isn't really that sharp. Aug 12 19:38:29 rmoravcik, is git-native failing for you also? Aug 12 19:38:31 might be better to kill the power to the devices in case the startup doesn't have enough power to succeed. Aug 12 19:38:41 You've got several minutes after the knee. Aug 12 19:38:50 the knee varies with device and life Aug 12 19:38:54 True. Aug 12 19:39:05 minutes @ 310ma? Aug 12 19:39:08 But, if you're polling every second, it's not really an issue. Aug 12 19:39:17 hangon, I've got a graph Aug 12 19:39:37 not to hand. Aug 12 19:39:58 It is several minutes though. But if you are polling every second, you have to be really stupid for this to sneak up on you :) Aug 12 19:40:16 agreed - if we can last 2 minutes it will work Aug 12 19:40:39 I'm planning on a fairly advanced battery monitoring thingy - doing impedance spectroscopy to work out the resistance of the battery, and determine its condition. Aug 12 19:40:42 but i think you trade total power if you set the point too high Aug 12 19:41:07 would be nice, but true monitoring need current in and out plus temp Aug 12 19:41:13 Yeah. Aug 12 19:41:27 There are some ways round that, just using voltage. Aug 12 19:41:57 i don't think the R(int) is constant Aug 12 19:42:12 No, it's not quite. Aug 12 19:42:20 But it varys in relatively predictable ways. Aug 12 19:42:32 things are probably better when the USB power is not supplied Aug 12 19:42:35 if the "task bar" or what ever the top bar on the Neo is called, dies on me, can I resurrect it? Aug 12 19:42:40 s/better/more predicable/ Aug 12 19:42:40 Writchie meant: things are probably more predicable when the USB power is not supplied Aug 12 19:43:19 hhf423: not without ssh AIUI Aug 12 19:43:22 the problem with not monitoring current is not knowing the true state of the battery during typical plug in/ unplug/ etc. Aug 12 19:43:30 Writchie: You don't care. Aug 12 19:43:37 daMaestro: no Aug 12 19:44:06 Writchie: voltage is a _really_ good proxy for charge-state - once you add in temperature and estimated discharge, and last known impedance at full charge. Aug 12 19:44:25 rmoravcik, do you know how i can kill the git-native task and make it run again? Aug 12 19:45:10 SpeedEvil: I think voltage alone will get you maybe within 20% Aug 12 19:45:19 If you have a decent estimate of current at any point. Aug 12 19:45:26 that's the problem Aug 12 19:46:18 For the suspend case, you know - pretty much - the current though. Aug 12 19:46:25 yeah Aug 12 19:46:39 but what's left to start with Aug 12 19:46:49 estimates might be fairly good Aug 12 19:46:51 The current only gets important when you discharge fast. Otherwise the voltage is a good measure. Aug 12 19:46:55 until we factor in the radios Aug 12 19:47:05 After several minutes 'resting' from being on. Aug 12 19:48:18 when operating, we at 0.3C Aug 12 19:48:59 but you're right, we are probably there for a couple of minutes. Aug 12 19:49:37 and the voltage when we go into standby from there can be compared to a profile Aug 12 19:49:52 to estamate mAhr's remaining Aug 12 19:50:16 actually, this crap belongs in the battery anyway Aug 12 19:50:45 it's probably there but we can't get at it Aug 12 19:51:09 No, we can't. Aug 12 19:51:18 It's - AIUI - a fairly dumb battery. Aug 12 19:51:30 It's got a thermistor, and an internal low-voltage protection chip. Aug 12 19:51:33 that's it. Aug 12 19:51:49 bare minimum Aug 12 19:52:00 (found out about the low-voltage protection today) Aug 12 19:52:10 but the protection chips these days have integrated fuel guage Aug 12 19:52:21 Not the bottom end. Aug 12 19:52:28 playing with the third wire I see Aug 12 19:52:38 No. Aug 12 19:52:54 Ran the phone down all the way to see what the voltage hit. Aug 12 19:53:01 disconnected the battery, and it read 0. Aug 12 19:53:09 that's good Aug 12 19:53:22 if it goes below 2.5V its pretty much toast Aug 12 19:53:31 Indeed. Aug 12 19:53:52 it must also have current protections Aug 12 19:54:06 it looks like a nokia standard battery Aug 12 19:54:27 but it doesn't respond to nokia hardware Aug 12 19:55:25 have you measured how long to re-charge to 4.1V Aug 12 19:55:26 * ljp thinks laser pointers and cats are just plain fun Aug 12 19:55:42 Some 3h IIRC, fully, with it off. Aug 12 19:55:50 with the phone on - rather Aug 12 19:55:54 ljp: mine got bored of it fairly quickly Aug 12 19:56:00 should be 20hr with phone off Aug 12 19:56:03 ljp: she realised there wasn't anything to catch Aug 12 19:56:12 am I the only person on the planet worried about battery life? Aug 12 19:56:12 matthew22: See http://wiki.openmoko.org/wiki/FAQ (and also SH1, SH1_FAQ, Neo1973, Developer_preview and Neo1973_Hardware#GTA01Bv4 pages) (Hopefully these links answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Aug 12 19:56:26 * SpeedEvil stabs matthew22. Aug 12 19:56:27 all 3 of mine just love it. even the 11 year old one Aug 12 19:56:35 We've just been talking about it for the last few mins. Aug 12 19:56:47 hah, good Aug 12 19:56:55 there is nothing about it on the wiki and that's disturbing Aug 12 19:57:01 Yes there is. Aug 12 19:57:04 lots Aug 12 19:57:08 search on 'power managment' Aug 12 19:57:10 for a device that's trying to do absolutely everything under the sun.. Aug 12 19:57:18 An extensive page. Aug 12 19:58:11 * ljp likes when they bite at it Aug 12 19:59:07 see the article on upgrading your flashlight with a high power laser diode from a DVD writer - then your cat will notice Aug 12 19:59:42 will the neo charge while the computer the usb is hooked into is turned off? Aug 12 19:59:45 * SpeedEvil wonders if you can de-hair a cat with that laser remotely. Aug 12 19:59:45 and you will too - with the one good eye you have left Aug 12 19:59:51 matt: slowly. Aug 12 20:00:08 * ewon wants an infra-red laser pointer Aug 12 20:00:11 so it charges faster when the computer is on? Aug 12 20:00:16 mat: unplug USB, open case, remove battery, wait 10s, replace battery. Aug 12 20:00:29 ewon: easy enough Aug 12 20:00:29 hello Aug 12 20:00:32 concerning laserpointers, look at this: http://www.eblogx.de/flashvideo_6977_laser-pointer.html Aug 12 20:00:33 plug into USB. Aug 12 20:00:33 ewon: remove the IR filter Aug 12 20:00:42 Then it will charge slowly. Aug 12 20:00:44 ewon: be ready to fry your eyeballs with 150mW+ Aug 12 20:00:56 Stephmw: only needs to project a normal spot at 18m Aug 12 20:01:14 SpeedEvil: can the usb charging be disabled? Aug 12 20:01:16 leave 1h or so, then boot it, and have it connected to a computer port which will regognise it. Aug 12 20:01:20 noon: why? Aug 12 20:01:27 the thing is it has be be invisible to the naked eye but still visible on a webcam Aug 12 20:01:36 ewon: that's still dangerous Aug 12 20:01:41 ewon: they are readily available Aug 12 20:01:46 google IR laser pointer Aug 12 20:02:06 because i don't want to geht it charged when i connect it to my notbook Aug 12 20:02:09 * SpeedEvil passes noon a 20cm USB cable with the red wire cut. Aug 12 20:02:11 SpeedEvil: will do Aug 12 20:02:21 just out of interest, why does the stylus even come with all that stuff? :) Aug 12 20:02:56 to play with cats Aug 12 20:02:57 SpeedEvil: so it can't be done with software? Aug 12 20:03:03 doc|home: because it's nearly free. Aug 12 20:03:10 noon: Not with current software. Aug 12 20:03:14 AIUI. Aug 12 20:03:17 $1.50 Aug 12 20:03:19 can I put it into the low current charge mode Aug 12 20:03:30 Unsure. Aug 12 20:03:30 SpeedEvil: 2$ on ebay Aug 12 20:03:39 Indeed. Aug 12 20:03:53 I would actually like a decent quality version of the stylus. Aug 12 20:04:00 mine comes apart a bit. Aug 12 20:04:08 the pen sux Aug 12 20:04:24 but I think i'll order a case Aug 12 20:04:37 i like the flashlight Aug 12 20:04:57 is there any way to sent texts with spaces in them using libgsmd-tool -m shell? Aug 12 20:04:58 but we already have one Aug 12 20:05:01 in the backlight Aug 12 20:05:03 * doc|home isn't a fan of disposable stuff when it's something that can be better made to last longer for the same price Aug 12 20:05:10 except its always a dead battery Aug 12 20:06:42 doc|home: planned obscolesence is alas the currently dominent business model Aug 12 20:07:14 Stephmw: fine for the phone, because the tech will change anyway, but I'd like the stylus to last longer than a month :) Aug 12 20:08:00 use whatever stylus you want, none fits the device Aug 12 20:11:42 Actually. Aug 12 20:11:56 I have noticed that you could drill a hole for a small stylus I think. Aug 12 20:12:12 heh Aug 12 20:12:19 SpeedEvil: if we can to 100mW, we will have 48hrs standby, 20hrs standby + 2 hrs talk, or 3.5 hr talk. Aug 12 20:12:47 True. Aug 12 20:13:16 1/2 the minimum desired, but at least workable Aug 12 20:13:51 anyone working on a web browser? i think dillo would make a good start Aug 12 20:14:35 pick 'web' from the menu, with an active GPRS connection. Aug 12 20:14:37 It's... Aug 12 20:14:47 Well, it displays some pictures. Aug 12 20:16:18 i think something webkit based is the way to go Aug 12 20:16:31 browser wars! Aug 12 20:16:37 almost as exciting as editor wars Aug 12 20:16:55 just because webkit is getting its own gtkwidget and it'll be easy getting a browser for free Aug 12 20:16:56 not familiar with webkit Aug 12 20:17:06 webkit = safari Aug 12 20:17:08 without the burden of geecko Aug 12 20:17:09 derived from khtml Aug 12 20:17:20 HOLY SHIT ON A STICK Aug 12 20:17:30 root@fic-gta01:~$ mono --version Aug 12 20:17:30 Mono JIT compiler version 1.2.2.1, (C) 2002-2006 Novell, Inc and Contributors. www.mono-project.com TLS: normal GC: Included Boehm (with typed GC) SIGSEGV: normal Disabled: none Aug 12 20:17:34 root@fic-gta01:~$ Aug 12 20:17:37 * rtyler dances Aug 12 20:17:37 ... Aug 12 20:17:39 boo Aug 12 20:17:49 this makes me happy ;) Aug 12 20:17:51 well, not boo Aug 12 20:17:56 Does it have stereo? Aug 12 20:17:57 still mmh c# ..~_~ Aug 12 20:18:05 not yet anyways ;) http://www.mono-project.com/Boo Aug 12 20:18:16 whatever the case, i would think a browser would be a high priority since phase2 will have wifi. Aug 12 20:18:26 err Aug 12 20:18:39 anrp: .NET CF apps are relatively prominent, having support would be nice ;) Aug 12 20:18:45 How large rtyler ? Aug 12 20:18:51 is the runtime? Aug 12 20:19:40 yes Aug 12 20:19:44 the binary is only 1.6MB, I need to start testing some of my base level applications to see what memory consumption will be like Aug 12 20:19:57 matthew22: using existing stuff, webkit gecko etc and work done on them it should be easy to write a simple browser Aug 12 20:19:58 Not bad. Aug 12 20:20:19 i agree defer. Aug 12 20:20:23 SpeedEvil: you can embed the Mono runtime in C-based applications, it's only ~4MB D Aug 12 20:20:26 :D* Aug 12 20:20:28 that's why i think dillo would be a good simple start. Aug 12 20:20:38 isn't dillo dead >_> Aug 12 20:20:47 and somebody's working on a WebKit-based browser Aug 12 20:20:55 (Acid2 compliance FTW) Aug 12 20:21:09 webkit is the smart choice IMO Aug 12 20:21:23 * SpeedEvil uses FF on a machine with the same RAM as the neo. Aug 12 20:21:25 as someone who runs WebKit nightlies on his Mac, I agree ;) Aug 12 20:22:08 i don't have a problem with webkit, just think dillo would be easier since it's only like 400k Aug 12 20:22:24 What is 'web' based on? Aug 12 20:22:29 gtkhtml? Aug 12 20:23:09 gtkhtml isn't that complete, dillo isn't a fully-featured browser either Aug 12 20:23:41 right... but can a fully feature browser be written before phase2 is released? i don't think so. Aug 12 20:24:05 i'm pretty sure it can Aug 12 20:24:21 well, that's the beauty of open-source, we can do both Aug 12 20:24:22 you're just really getting webkit or gecko to compile and wrap an interface around it Aug 12 20:24:36 the hardest work is done Aug 12 20:25:04 of course Aug 12 20:25:33 but phase2 is 2 months away, and as far as I know no one has even started on a browser Aug 12 20:25:39 maybe someone has thogh Aug 12 20:25:46 phase 2 is more than two months away Aug 12 20:25:53 if phase one is anything to go by Aug 12 20:25:54 hi Aug 12 20:26:05 is there a good java vm for openmoko? Aug 12 20:26:15 quick answer: no Aug 12 20:26:17 long answer: not yet Aug 12 20:26:19 wbx: working on it ;) Aug 12 20:26:37 rtyler: which vm? Aug 12 20:26:48 I'm going to play with IKVM Aug 12 20:26:52 rtyler: i like to get cachewolf working. Aug 12 20:26:53 (on top of Mono) Aug 12 20:26:57 i've compiled gcj for arm before, but i have never used gcj Aug 12 20:27:05 rtyler: i try jamvm with gnu classpath. Aug 12 20:27:18 rtyler: seems maintained. Aug 12 20:27:33 rtyler: mono seems to slow.. ;) Aug 12 20:27:41 -1, Flamebait :D Aug 12 20:27:43 ooh careful Aug 12 20:27:47 nerd fight! Aug 12 20:27:50 * anrp popcorn Aug 12 20:27:52 heh Aug 12 20:27:58 MY VM IS BETTER THAN UR VM Aug 12 20:28:12 anyways, wbx have you tried Kaffee? Aug 12 20:28:18 bye Aug 12 20:28:54 and what do you want to do with cachewolf, with no GPS subsystem yet? :P Aug 12 20:31:38 rtyler: yes, lavazza, just some minutes ago, wonderful. my sleep is now away ;) Aug 12 20:31:48 rtyler: is gps broken? Aug 12 20:32:01 if broken means not finished, then yes ;) Aug 12 20:32:46 rtyler: that is the reason I pay money for a great c hacker ;) Aug 12 20:33:15 rtyler: he will fix it for me. he already likes to get my zaurus for his next geo cache search *g* Aug 12 20:33:43 but first step is to get a good boot logo. is there a description in which format a boot logo should be? Aug 12 20:37:33 hmm. 480x640 would be good ;) Aug 12 20:40:56 my first guess would be something cattable to the fb Aug 12 20:41:16 * defer brb Aug 12 20:44:23 wbx: the boot logo is some sort of raw gzipped format Aug 12 20:44:59 wbx: look at build/qemu/openmoko/Makefile on a mokomakefile tree, it's the splash.gz target Aug 12 20:45:40 cesarb: i created a 480x640 png of http://www.snowcrest.net/fox/tar5.JPG, and used ./splashimg.pl | gzip -9 >splash.gz Aug 12 20:45:41 * cesarb looks at the orzlab page mentioned several hours ago Aug 12 20:45:50 ello all Aug 12 20:45:52 but only 1/4 of the image is now displayed on boot Aug 12 20:46:17 all the rest is black. hmm. Aug 12 20:46:22 Why are these people reverse engineering lowlevel_foo? Isn't its source parte of openmoko's uboot+patches? Aug 12 20:46:49 -rw-r--r-- 1 wbx wbx 58527 2007-08-12 22:40 splash.gz Aug 12 20:46:49 -rw-r--r-- 1 wbx wbx 17216 2007-08-12 12:57 splash.gz.orig Aug 12 20:46:56 is my logo to big? Aug 12 20:47:03 possibly Aug 12 20:47:19 wbx: no, the partition, as discussed on this channel yesterday, is ridiculously large Aug 12 20:47:20 hmm. i will check maximum flash size. Aug 12 20:47:35 cesarb: hmm. yesterday i was not online ;) Aug 12 20:47:46 wbx: even the smaller partition on the qemu setup is more than 200k if I calculated it correctly Aug 12 20:48:00 cesarb: sorry asking RTFM stuff. is there an entry in the wiki for the splash stuff? Aug 12 20:48:14 cesarb: i have no qemu. i have a real orange device :} Aug 12 20:48:55 wbx: Special:Allpages finds nothing on the splash Aug 12 20:49:48 wbx: found the cause for your problem Aug 12 20:49:50 cesarb: there is a readme in openmoko/src/host/splash. Aug 12 20:49:54 cesarb: yes? Aug 12 20:50:03 orange D: Aug 12 20:50:15 anrp: at least it's not OLPC green Aug 12 20:50:19 grr Aug 12 20:50:22 i love olpc green Aug 12 20:50:36 i so wish that the phones came in the green/white that they originally showed Aug 12 20:50:43 would make me soo happy Aug 12 20:50:46 wbx: at least on qemu's setup, there is a length hardcoded on the uboot environment for how much will it read from the splash partition Aug 12 20:51:01 wbx: on the default qemu setup, it's 20480 bytes Aug 12 20:51:05 ah. do you know the variable name? Aug 12 20:51:14 will check. Aug 12 20:51:29 cesarb: There doesn't seem to be any source for the 4K image, loaded by the steppingstone, which then loads u-boot Aug 12 20:51:32 wbx: it's in the middle of the splashimage command, but BE CAREFUL when changing it Aug 12 20:51:47 cesarb: i have the advanced moko ;) Aug 12 20:51:53 wbx: heh Aug 12 20:52:06 cesarb: and i already know how to use uboot and jtag on ppc systems ;) Aug 12 20:52:20 wbx: someone yesterday tyoped one number on that line, and lost DFU (and USB serial; uboot crashed with USB plugged in) Aug 12 20:52:36 cesarb: yeah, life is dangerous ;) Aug 12 20:52:53 jeez apparently i've missed alot Aug 12 20:53:04 wbx: since you have the debug board, recovery is easier; you can just use the real serial port and change the variable back :D Aug 12 20:53:37 wbx: back to your question, look at build/qemu/openmoko/flash.sh, where it creates the splashimg line, and compare to yours: Aug 12 20:53:38 cesarb: i hope i do not need it yet. may be later ;) Aug 12 20:53:39 SpeedEvil: there may be hope. if the OHCI controller meets the 1.0 spec as advertised, it will recognize transition from usb suspend to usb resume Aug 12 20:53:49 setenv splashimage 'nand read.e $splash_addr splash $splash_size; unzip $splash_ Aug 12 20:53:50 addr 0x33d00000 0x96000' Aug 12 20:54:01 sorry for the accidental line break, it's a single line Aug 12 20:54:16 BryceLeo: everything is in the wiki: http://wiki.openmoko.org/wiki/Bricked Aug 12 20:54:20 ;) Aug 12 20:54:25 cesarb: thx. this makes sense to me. I'll report any success or failure in a few minutes. Aug 12 20:54:27 SpeedEvil: if this doesn't work, it suggestion to put the USB host into USB Suspend for power down wouldn't make sense. Aug 12 20:54:53 of course, failing to include a wakeup interrupt for this doesn't make sense either Aug 12 20:55:01 Kensan: sweet! Aug 12 20:55:27 my mom had a grapefruit sized cyst removed from her liver, so i've been a bit off line Aug 12 20:55:46 nice Aug 12 20:55:57 cesarb: seems that my battery is empty. :( Aug 12 20:56:20 wbx: don't you have _two_ of them? ;-) Aug 12 20:56:23 cesarb: is it possible to see on the pc if the battery gets loaded? Aug 12 20:56:34 cesarb: the first was empty in the morning ;) Aug 12 20:56:47 cesarb: i already have my moko 3 days *g* Aug 12 20:57:03 cesarb: need to find a usb hub and fast charge :} Aug 12 20:57:34 wbx: I don't think you can see in the PC if the battery is being charged or not, but did you try the trick often repeated in this channel? Aug 12 20:57:46 cesarb: which trick? Aug 12 20:57:58 power off, remove battery, wait 10s, put battery back in, plug usb, wait some amount of minutes I can't remember right now Aug 12 20:58:11 cesarb: i am relatively new to openmoko, because i love to play with hardware i own ;) Aug 12 20:58:40 wbx: mine should be in the next batch, if nothing goes wrong and I'm counting the ticket numbers correctly Aug 12 20:59:14 (advanced too, as knowing myself I will sooner or later end up playing with either uboot or the kernel) Aug 12 20:59:29 (and, as we learned yesterday, even playing with uboot's environment can be dangerous!) Aug 12 20:59:39 cesarb: hehe. Aug 12 21:00:39 anyone know status on 2007.2 Aug 12 21:00:41 lol @http://hardware.slashdot.org/hardware/07/08/12/193221.shtml Aug 12 21:01:39 cesarb: my openmoko has a white apple sticker on his back ;) Aug 12 21:01:41 yeah, somewhat funny Aug 12 21:01:55 * mjr was thinking an Ubuntu sticker, but that would just be untrue Aug 12 21:03:02 I reckon you can get OE stickers somewhere... Aug 12 21:03:51 i've never been able to find any nice small stickers Aug 12 21:03:57 i didn't realize that "some assembly required" referred to arm assembler ;) Aug 12 21:04:02 only like the large bumper sticker sized ones on cafe press Aug 12 21:04:58 no reason you can't cut them down Aug 12 21:05:01 Writchie: no, it refers only to having to flash the working image Aug 12 21:05:17 Writchie: which, of course, does not exist yet (mickey said so some pages ago) ;-) Aug 12 21:05:20 anrp: when you're chopping off the logo and the words??? Aug 12 21:06:08 IMHO before we're done it will be necessary to reflash u-boot, the steppingstone boot, and maybe even the GSM Aug 12 21:06:09 er just put a bunch of stickers on the same run Aug 12 21:06:50 Writchie: is the steppingstone ROM flashable? Aug 12 21:07:11 Writchie: (or are you refering to the first page loaded by the steppingstone, which is part of uboot?) Aug 12 21:07:14 its the first physical page of the flash Aug 12 21:07:31 (or am I mixing the nomenclature up?) Aug 12 21:07:36 are you saying this is part of u-boot, Aug 12 21:07:44 i can't even find the image Aug 12 21:07:51 much less the source Aug 12 21:07:58 please enlighten me Aug 12 21:08:14 Writchie: I thought it was part of uboot... I might be wrong. Aug 12 21:08:25 it's what loads u-boot Aug 12 21:08:35 there is a ROM bootloader on the 2410 Aug 12 21:08:42 that loads 4K into RAM Aug 12 21:08:51 that then loads u-boot Aug 12 21:08:57 which then does it thing Aug 12 21:09:14 yeah, arm cores that use primarily nand are like that Aug 12 21:09:25 3 stage bootloader Aug 12 21:09:38 same deal on an at91rm9200 board i have here Aug 12 21:10:17 boot from nand is relatively new, but almost all have code to boot from something, UART, i2C, etc Aug 12 21:11:53 in any case, we shouldn't have to disassemble this code in order to get a workable PM. Aug 12 21:13:16 /dev/mtdblock0 - first 4 blocks should be the image? Aug 12 21:14:28 Writchie: did you see this patch to uboot? http://svn.openmoko.org/trunk/src/target/u-boot/patches/boot-from-ram-and-nand.patch Aug 12 21:14:40 SpeedEvil: we should be able to load 4K of executable in the steppingstone buffer, do the shut-down except for actually entering shutdown state, then jmp to code in steppingstone Aug 12 21:14:46 Writchie: take a look at cpu/arm920t/start.S, it looks interesting Aug 12 21:15:13 Writchie: it seems to at least have code to take SDRAM out of self-refresh Aug 12 21:15:40 would booting from ram be faster than booting from NAND? Aug 12 21:16:20 in low power modes, we want to shut down SDRAM Aug 12 21:16:32 and run from the onchip steppingstone RAM Aug 12 21:16:49 Kensan: thanks Aug 12 21:16:50 Is it possible to lock down the cache? Aug 12 21:16:53 I dunno. Aug 12 21:17:04 ah Aug 12 21:19:48 SpeedEvil: actually entering power off state needs code in the steppingstone FLASH to handle resume Aug 12 21:20:19 SpeedEvil: but ultra-slow mode with SDRAM off can be done now by loading code to the RAM buffer Aug 12 21:20:34 some other SoC's use the Ethernet buffer for a similar purpose Aug 12 21:21:11 hmm. networkmanager in feisty is annoying, it switches to usb network always when my neo boots. Aug 12 21:21:14 Oh - right. Aug 12 21:22:50 cesarb: paste and copy in uboot is evil ;) Aug 12 21:22:56 of course, we need the basic ACPI functions to work, and this is tricky stuff and I have no idea who is working on it Aug 12 21:24:30 * mjr installed an ad-hoc screen protector. Now let's see if it interferes with the ts... Aug 12 21:25:17 Writchie: based on a cursory look at the uboot code, I would guess the first 16k of the NAND is a copy of the first 16k of some old uboot Aug 12 21:25:30 Writchie: but the code is kinda messy Aug 12 21:26:03 hey! Aug 12 21:26:32 Writchie: the way to be sure would be to dd out these first 16k and compare Aug 12 21:28:47 cesarb: the first few physical pages of NAND may be hidden Aug 12 21:29:08 stefan_schmidt: hey! Aug 12 21:30:15 SpeedEvil: well, just tried suspending a stock Ubuntu and it doesn't work, and neo keeps charging while USB is suspended, and ubuntu doesn't come of suspend. Aug 12 21:30:21 so lots of work is ahead Aug 12 21:31:28 Writchie: It seems that Werner Almesberger is doing uboot related work. Aug 12 21:31:35 does anyone have a solid working resume and suspend (with USB devices) Aug 12 21:31:39 Kensan: yes Aug 12 21:32:26 Writchie: I think they are simply outside of all partitions, much like a PC's MBR Aug 12 21:32:50 Writchie: but I have no neo to test, only a lousy qemu Aug 12 21:32:52 may be outside of the disk device as well Aug 12 21:33:06 there are interactions with bad block table etc. Aug 12 21:34:13 cesarb: hardware can always be physically read, though, if necessary Aug 12 21:34:56 there is just so much stuff here and so little architectural documentation. Aug 12 21:35:11 i sick of this just read the source b.s. Aug 12 21:35:19 this is not engineering Aug 12 21:35:33 Writchie: IIRC, the BBT is at the end of the flash, so it shouldn't be a real problem Aug 12 21:36:09 Writchie: going by the wiki (http://wiki.openmoko.org/wiki/S3C2410_Steppingstone) the source is not available. Aug 12 21:36:14 yes, the BBT is at end IIRC and the stepping stone at begining, and I think both are hidden in the exposed MTD device Aug 12 21:36:36 u-boot should be able to read it out of the nand if the linux kernel hides it Aug 12 21:36:49 the stepping stone code on the SoC is not published Aug 12 21:37:04 although other ARM SoC's do publish theirs Aug 12 21:37:04 oh, there's flash inside the cpu? Aug 12 21:37:16 Writchie: I fail to see why would the first few k of the NAND hidden, other than by being outside of all partitions Aug 12 21:37:16 thought it was rom Aug 12 21:37:31 but we are talking about the first page from the nand which is what the steppingstone loads. Aug 12 21:37:47 cesarb: I would certainly hide it from DD Aug 12 21:38:07 if you trash it you'll be starting over from scratch. Aug 12 21:38:22 Writchie: I wouldn't, the same way the MBR isn't hidden from dd if=/dev/hda Aug 12 21:38:35 Writchie: trashing it is just as evil as trashing the uboot Aug 12 21:39:24 it's similar to the code in the MBR Aug 12 21:39:30 but it has to be aware of bad blocks Aug 12 21:39:52 this is done on an SD card internally Aug 12 21:42:19 Kensan: the wiki is a bit incomplete Aug 12 21:42:40 the steppingstone is also involved in the resume processing Aug 12 21:45:54 Writchie: yes, judging by the edit history it's never been updated since last year (!). I was pointing at the wording "some hidden microcode" which hints that there's no src available. Aug 12 21:46:43 got ya. the actuall bootloader is samsung. The steppingstone is either FIC/openmoko.com or openmoko.org Aug 12 22:07:13 hey Aug 12 22:07:18 what controls screen blanking? Aug 12 22:07:24 the phone is in a bad state Aug 12 22:07:29 it keeps blanking out the screen Aug 12 22:07:53 if i press the power button and wait a while, the screen comes back for about a second and blanks again Aug 12 22:08:20 Writchie: just took a look at the "derviginator" (which is what you would expect to write the first 16k of the NAND) Aug 12 22:08:32 Writchie: bizarrely, it doesn't seem to touch the first 16k of the NAND Aug 12 22:08:50 * cesarb is going "huh?" Aug 12 22:09:42 it would make sense that the first couple physical pages be reserved. Aug 12 22:10:16 Writchie: something has to write the magic bootloader on these pages. If the devirginator isn't writing it, then WHAT is writing it??? Aug 12 22:10:25 i wonder how many devices have been produces in the first batch.. Aug 12 22:10:33 hp_: 1000 Aug 12 22:10:50 whow.. nice :) Aug 12 22:11:28 hp_: all sold already Aug 12 22:11:51 cesarb: do you know the correct baudrate for ttyACM0 Aug 12 22:11:57 awsome to hear so many devs wanted to try it out :) Aug 12 22:11:59 cesarb: typing in u-boot is strange Aug 12 22:12:26 cesarb: Stage 0, perhaps Aug 12 22:12:59 cesarb: will look at source Aug 12 22:13:55 stuff/lowlevel_foo-gta01bv3-blah blah .bin may be it Aug 12 22:13:59 Writchie: no, stage0 only runs "nand scrub 0x3ff0000 0x10000" at the beginning of stage1 Aug 12 22:14:16 cesarb: that's what is says Aug 12 22:14:23 Writchie: I think lowlevel_foo is for memory booting only (loaded via jtag) Aug 12 22:15:04 * cesarb goes "huh?" again Aug 12 22:15:10 # switch on the backlight, so one can see that we're alive Aug 12 22:15:14 neo1973 backlight off Aug 12 22:15:16 wtf? Aug 12 22:18:31 cesarb: still doesn't make sense unless low-level foo is the binary for the steppingstone Aug 12 22:18:40 or u-boot includes this Aug 12 22:19:43 the bad block remap needs to be working on the u-boot load Aug 12 22:20:38 how do i get my phone out of this state? Aug 12 22:21:13 it keeps blanking the lcd and then a while later disconnecting from usb0 network Aug 12 22:21:16 what is the command to rotate the display 90 degrees? Aug 12 22:21:31 use xrandr -o Aug 12 22:21:37 and give it a number 3 Aug 12 22:21:40 thx Aug 12 22:21:41 xrandr -o 3 Aug 12 22:21:48 to go horizontal Aug 12 22:21:52 xrandr -o 0 Aug 12 22:21:53 to return Aug 12 22:22:02 ok Aug 12 22:22:09 orzo: Master Reset, i.e. open case, remove battery, wait 10 seconds Aug 12 22:22:35 orzo: you may have to charge for a while before you can get to powering up for fast charge Aug 12 22:22:50 Writchie: boot-from-ram-and-nand.patch seems to be the key, it mostly changes to check if (loaded from NAND & address == 0). Isn't address=0 the internal SRAM? Aug 12 22:23:34 when booting from nand, there is a different memory map. Aug 12 22:24:24 i'll have to look at the full code Aug 12 22:24:30 is this a recent patch? Aug 12 22:25:12 or just a patch in the openmoko differences sense Aug 12 22:25:21 I think that's different. Aug 12 22:25:26 Writchie: openmoko/trunk/src/target/u-boot/patches Aug 12 22:25:30 Writchie: it's a quilt patch Aug 12 22:25:31 Isnt't hat boot from an image loaded by ... Aug 12 22:25:33 err - no Aug 12 22:25:36 ignore me. Aug 12 22:27:01 Writchie, should I let it charge for a while before removing the battery and waiting? Aug 12 22:27:16 no. Aug 12 22:27:30 The reason is that the GSM modem may not be off. Aug 12 22:27:37 And other stuff. Aug 12 22:27:42 orzo: it's the opposite, you should remove the batery and wait before charging for a while Aug 12 22:29:39 cesarb: thanks a lot Aug 12 22:30:05 wbx: did it work? Aug 12 22:30:19 cesarb: setting it to setenv splashimage nand read.e 0x32000000 splash 0xff00 ... helped Aug 12 22:30:27 cesarb: yes Aug 12 22:30:30 i have no sim in it Aug 12 22:30:44 it'd fire up gsm modem even though there's no sim? Aug 12 22:30:50 orzo: yes Aug 12 22:30:52 cesarb: my time machine (dr. who) splash is ready. Aug 12 22:31:00 orzo: GSM is supposed to work even without a SIM Aug 12 22:31:05 orzo: for emergency calls for instance Aug 12 22:36:09 battery is at 79% acording to apm Aug 12 22:36:16 why did it go into that mode? Aug 12 22:39:10 cesarb: lowlevel_foo is looking to me like the steppingstone code Aug 12 22:39:23 does /etc/init.d/gsmd stop power off the modem? Aug 12 22:39:29 Writchie: it obviously would look similar even if it weren't Aug 12 22:39:34 Writchie: since the purpose is the same Aug 12 22:39:38 true Aug 12 22:39:47 it runs at effective 0 Aug 12 22:39:58 with the interrupt vectors Aug 12 22:40:01 orzo: maybe Aug 12 22:40:09 Writchie: but if it isn't the steppingstone code, it will not have code to read from NAND, while the real code will have Aug 12 22:40:15 orzo : the problem is tha some stuff is not powered off. Aug 12 22:40:26 orzo: and we don't actually know what. Aug 12 22:40:31 * cesarb still thinks the code you are looking for is probably nothing more than the first 16k of uboot itself Aug 12 22:40:35 Switched off - due to the menus - uses some 30mA. Aug 12 22:41:03 very possible Aug 12 22:41:04 Switched off - due to pulling the battery - uses <1mA. Aug 12 22:41:10 (once you plug the battery back in) Aug 12 22:41:13 Writchie: there is one particular thing which makes me think that Aug 12 22:41:27 Writchie: take a look at board/neo1973/gta01/u-boot.lds on uboot Aug 12 22:41:29 there's not enough stuff to load u-boot Aug 12 22:41:45 Writchie: notice that it suspiciously have a particular layout for the first section Aug 12 22:42:06 Writchie: if you know how to read gnu binutils linker scripts, you will see what I mean: Aug 12 22:42:13 cpu/arm920t/start.o (.text) Aug 12 22:42:17 cpu/arm920t/s3c24x0/nand_read.o (.text) Aug 12 22:42:22 *(.text) Aug 12 22:42:46 Don't you think it's very suspicious that it explicitly forces the code from nand_read.o to be at the very beginning of the image? Aug 12 22:42:51 can i run framebuffer aps without quitting x? Aug 12 22:43:21 yes, it means its going to be loaded there. Aug 12 22:43:27 (just after the start.S code, which *has* to be at the very beginning of the image or else it will not work) Aug 12 22:44:00 the stepping stone code executes at a different address after the switch Aug 12 22:44:04 I think it's done that way so the first 16k of the uboot binary will have both the startup code and the nand reading code Aug 12 22:44:30 And it looks at the pc to know if it's been loaded by the first stage loader or not Aug 12 22:44:40 do you see the load of u-boot itself in the u-boot code Aug 12 22:44:52 i just haven't looked yet Aug 12 22:44:57 orzo, yes, though obviously the draws will conflict with X's, if they both draw to the fb around the same time Aug 12 22:45:18 * SpeedEvil ponders. Aug 12 22:45:33 Writchie: I tried to, I saw code to read from the nand (it's at cpu/arm920t/start.S), but the code is WAY too loaded with #ifdef noise for me to follow it quickly Aug 12 22:45:36 chvt n works Aug 12 22:46:12 Xfbdev :2& Aug 12 22:46:13 Writchie: so I couldn't find where it actually jumps to the loaded code, to see if it jumps to the start or to some magic offset or something even crazier Aug 12 22:46:14 then chvt Aug 12 22:46:39 the console is on VT 1, the normal X on vt2, and the new X on vt3 Aug 12 22:47:11 SpeedEvil: you got chvt to work? Aug 12 22:47:17 yes Aug 12 22:47:28 how so with one fb device? Aug 12 22:47:34 Don't know. Aug 12 22:47:37 Just worked. Aug 12 22:47:42 Even then, I'm still trying to understand why the derviginator code doesn't seem to write anything to NAND address 0. My current theory is that the devirginator code is old, and in the past 'dynpart' put the uboot partition at address 0, instead of several kb later Aug 12 22:47:50 * SpeedEvil does not investigate things that just work closely. Aug 12 22:47:53 They might stop. Aug 12 22:48:08 SpeedEvil: when you write to the fb dev does it go to the screen Aug 12 22:48:51 For chvt to work, you need either 'openvt' or something else opening a new VT Aug 12 22:48:51 yes Aug 12 22:48:57 not really Aug 12 22:49:04 chvt n seems to open new vt Aug 12 22:49:19 actually - maybe there are 4 vts only Aug 12 22:49:28 anyway. Aug 12 22:49:43 chvt4, cat >/dev/fb0, chvt 2 gets you back to x Aug 12 22:49:47 SpeedEvil: hm, true, the manpage says chvt creates new VTs Aug 12 22:49:50 and chvt 4 gets you back to the noise Aug 12 22:49:51 SpeedEvil: sorry for the noise Aug 12 22:50:27 actually no - it wipes Aug 12 22:50:59 it looks like there is only one physical framebuffer fb device Aug 12 22:51:09 i hope i'm wrong Aug 12 22:51:13 why? Aug 12 22:51:27 i want to co-exist X and direct FB Aug 12 22:51:56 without resorting to a virtual FB in X Aug 12 22:52:27 yeah - it seems to overwrite if I start mplayer Aug 12 22:52:30 the X plus everything known to man stack is going to be unstable for a long time Aug 12 22:52:54 the easiest fix is by at least two fb devices Aug 12 22:53:01 dDo you need to be able to play movies? Aug 12 22:53:17 Two X works quite well. Aug 12 22:53:20 not so much for that as to have a stable phone under any conditions Aug 12 22:53:41 two X work with one fb dev Aug 12 22:53:47 yes. Aug 12 22:53:52 they seem to stop writing Aug 12 22:53:53 hmmm Aug 12 22:54:06 simple second X may be stable enuf Aug 12 22:54:08 and know when they are suspended Aug 12 22:54:35 I don't think I've seen an actual X crash yet. Aug 12 22:54:59 probably gtk Aug 12 22:55:19 actually - mainly bits of the panels falling out mainly. Aug 12 22:55:51 where do they go is the question? Aug 12 22:56:02 what do they fall on Aug 12 22:56:27 both your x have the touch screen? Aug 12 22:56:41 No, it works just like 'normal'. Aug 12 22:56:48 when you switch X on a PC Aug 12 22:57:18 you mean one at a time Aug 12 22:57:52 yes - the active one gets the events Aug 12 22:57:58 cool Aug 12 22:58:14 can you force which one is active from the outside? Aug 12 22:59:18 chvt Aug 12 22:59:51 strange, when i played with chvt it did nothing to the current X , i'll try it again Aug 12 22:59:55 a script to poll the top button, for example. Aug 12 23:00:17 we need a bailout to known working simple thingy Aug 12 23:00:19 If you just start a new X, it goes to a seperate VT. Aug 12 23:00:44 five clicks of the top button in 10s, forex. Aug 12 23:01:09 shake the mother real hard for GT02 Aug 12 23:01:44 Writchie: next time you are on uboot, could you do a 'dynenv get' and compare it with the mtdparts line on the environment? Aug 12 23:02:03 Writchie: on my qemu, it's pointing to the _uboot_ partition, not the _environment_ partition, which contradicts the wiki Aug 12 23:02:07 cesarb: OK, but it will probably be awhile Aug 12 23:03:11 Writchie, watch out, we're gonna violate etch'a'sketch patents Aug 12 23:03:15 * cesarb wonders if it's a bug, or some other piece of the steppingstone puzzle Aug 12 23:03:16 its 7PM sunday here, i'll get to it in the morning. Aug 12 23:03:46 cesarb: are you the one making qemu work Aug 12 23:04:05 Writchie: no, I only am playing with it because I have no real device yet ;-) Aug 12 23:04:40 nothing like the real device - for the low level stuff Aug 12 23:05:33 ah, got it Aug 12 23:05:37 the mtdparts was confusing me Aug 12 23:05:50 just run the "mtdparts" command on uboot and you will see Aug 12 23:06:00 the mtdparts thingy on the environment is not the offsets, it's the lengths Aug 12 23:06:18 Writchie: this means... the first page is the start of the uboot partition! Aug 12 23:06:39 then the steppingstone foo: would have to be in uboot? Aug 12 23:06:48 a true bootstrap Aug 12 23:07:06 Writchie: not only being on uboot, it's the first part of uboot! start.S and the nand stuff Aug 12 23:07:10 * juri_ waits patiently for her neo1973 to arrive. Aug 12 23:07:13 gtg, bbl Aug 12 23:07:50 cheers Aug 12 23:12:11 What's the fully-charged battery voltage? Aug 12 23:13:04 And the "knee" voltage Aug 12 23:14:21 I'm seeing 4025 atm Aug 12 23:14:29 DukeOfURL: That's about fully charged Aug 12 23:14:53 DukeOfURL: I think it's listed as a 3.7v battery or some such Aug 12 23:14:54 do you know what the "knee" voltage is? Aug 12 23:15:07 DukeOfURL: I"m not sure that has been learned yet. Aug 12 23:15:17 ok, thanks Aug 12 23:16:23 I was able to run portable gprs for about 4 hrs. Not sure when it stopped, tho. Aug 12 23:16:29 juri_: where are you? Aug 12 23:16:33 I plugged it into a new charger, but it didn't charge. Aug 12 23:17:14 I plugged mine into a Motorola 5V 550ma charger and it didn't charge Aug 12 23:17:50 ? Aug 12 23:17:57 This is a scosche 550ma charger from walley-world Aug 12 23:18:10 huh? Aug 12 23:18:29 wamart? Aug 12 23:18:32 Walmart Aug 12 23:18:33 walmart? Aug 12 23:18:55 I"m flashing right now. After booting, I'll try it again when the cpu is acting rationally Aug 12 23:19:28 is there a negotiation that goes on between the Neo and the USB port? Aug 12 23:20:05 I hope not. Aug 12 23:20:27 The wiki isn't very revealing Aug 12 23:21:18 There are some fast charge parameters in the bootload process Aug 12 23:21:24 try "cat /sys/bus/i2c/devices/0-0008/chgcur" Aug 12 23:21:55 1733 Aug 12 23:22:38 That's your charging current in ma, I think Aug 12 23:22:49 seems high Aug 12 23:22:59 chgcur is very unstable. Aug 12 23:23:06 Take ten readings and average Aug 12 23:23:08 Oh..... Aug 12 23:23:16 and I don't think it's actually mA. Aug 12 23:23:49 And I think it's not actually charging current, but charging current - operating current Aug 12 23:24:01 And I think it's not actually charging current, but charging current + operating current Aug 12 23:28:59 Heh... Aug 12 23:29:17 Fixed the -dpi in .1 build. Got a BIG mess on the today panel Aug 12 23:30:19 SpeedEvil: what do you call the flexible circuit film found in cameras? Aug 12 23:30:31 Damn stuff. Aug 12 23:30:51 Flexible printed connectors - fpc is one name Aug 12 23:31:01 or flexible pcb Aug 12 23:31:12 can ICs be attached to it? Aug 12 23:31:26 In theory. It's _expensive_ to get made though. Aug 12 23:33:19 or fpc flexible plastic connector Aug 12 23:33:55 been around for years, also very common to fail with phones that have a clamshell design Aug 12 23:33:57 i thought it was flexible printed circuit - oh well Aug 12 23:35:26 I used to sell panasonic Cordless phones year back the plastic ribbon connector between the two halves of the phone always failed after a year or two, pretty much the reason why there are so few clamshell cordless phones today Aug 12 23:35:41 s/year/years Aug 12 23:36:48 * linux_galore notes he just added another name to confuse people Aug 12 23:37:16 SpeedEvil: it does look like re-start from USB is possible and will fit in the u-boot code. Aug 12 23:37:24 linux_galore: not really. Aug 12 23:37:30 Ive also seen it being called a thin film connector Aug 12 23:37:47 linux_galore: there are many in europe - it's possible to do it right. Aug 12 23:37:55 The key is adequate coiling in the hinge Aug 12 23:38:47 bend radius is pretty important Aug 12 23:39:42 Ive talked to some mobile tech's and they say the clamshells fail before the brick phones Aug 12 23:40:10 clam shells fail for lots of reasons including hinges Aug 12 23:40:16 Indeed. Aug 12 23:40:22 they are not very robust when dropped Aug 12 23:40:48 Done right, they can be robust. Aug 12 23:40:51 I actually like clamshell phones heh Aug 12 23:40:54 Many are not done right. Aug 12 23:41:09 Moto ones seem to work Aug 12 23:41:13 they go in and out of vogue Aug 12 23:41:27 startac was pretty good gsm phone in its time Aug 12 23:41:53 Seen the new Nokia phones with their gem look Aug 12 23:42:26 they remind me of designs way back in the 80's were they were trying to make plastic look less bland Aug 12 23:42:49 back in late 90's in china, nokia sold a slim phone for the lucky number R.M.B 8888 Aug 12 23:42:54 hmmm what's the secret hand shake to type in the neo1973 termial or gain access via usb networking? I have googled without success so far. Aug 12 23:42:59 which was about $1100 US Aug 12 23:43:02 nothing. Aug 12 23:43:05 on host. Aug 12 23:43:05 and was very popular Aug 12 23:43:12 modprobe cdc_ether Aug 12 23:43:20 ifconfig usb0 192.168.0.200 Aug 12 23:43:25 ssh 192.168.0.202 Aug 12 23:43:35 like the wiki says Aug 12 23:43:46 google is a bit behind Aug 12 23:43:47 SpeedEvil so is there a password? Aug 12 23:43:47 anyway back to work Aug 12 23:43:51 blank Aug 12 23:44:02 Ahhh I was doing root@.... Aug 12 23:44:03 not really blank, none Aug 12 23:44:07 yeah Aug 12 23:44:14 space used to be called blank Aug 12 23:44:18 but that's a long time ago Aug 12 23:44:24 blankety blank. Aug 12 23:45:48 have to look at security on the OM soon Aug 12 23:45:51 SpeedEvil: I can't decide whether to work on PM or gsmd Aug 12 23:46:32 gsmd Aug 12 23:46:44 needs to be functional before it's nice IMO. Aug 12 23:46:49 i think that too Aug 12 23:47:03 there are a lot of people that can get the PM working Aug 12 23:47:37 My concern is what are the guys working on at ccc? Aug 12 23:47:50 I don't want to spend a week working on gsmd or PM or anything, to find another has already finished it. Aug 12 23:47:51 what they always do at ccc Aug 12 23:47:55 Drink beer? Aug 12 23:48:33 that's the delimma, and I don't think they understand that. The community is the largest resource Aug 12 23:48:54 Right. We need a way to multiply our efforts, not duplicate (I said that a few days ago) Aug 12 23:49:05 architects write paper not code Aug 12 23:51:32 i think that all of the core team is very hard working and very overworked. Aug 12 23:52:25 however, many potential contributions to the project are thwarted. Aug 12 23:52:49 DukeOfURL: On my device... Aug 12 23:52:53 Volatage (ma): 4224 Aug 12 23:52:53 Charging (ma): 560 Aug 12 23:52:55 perhaps they all of this internally Aug 12 23:53:05 Hmm... need to fix a script typo Aug 12 23:53:11 but they do not have enough internal people to do the project Aug 12 23:53:33 SpeedEvil thanks for your help Aug 12 23:53:38 :) Aug 12 23:53:51 unlikely you would 4.2 volts and be drawing 560ma. Aug 12 23:54:06 It goes up and down.... Aug 12 23:54:27 I'm using the /sys/.../chgcur and /battvolt Aug 12 23:54:51 SpeedEvil: do you know the highest voltage you.ve seen on vbat? Aug 12 23:54:56 sagacis: I'm seeing 4125 on mine Aug 12 23:54:58 4230ish Aug 12 23:55:19 4242 right now Aug 12 23:55:44 4253 is the max. Aug 12 23:55:50 You need to average voltage too. Aug 12 23:55:56 It's not as inaccurate as current. Aug 12 23:56:08 But, it's not noise-free either. Aug 12 23:56:27 Sure. I"m just dumping it to the console to see how my charger is doing.... Aug 12 23:56:42 The chargecurrent is almost meaningless without averaging Aug 12 23:57:09 I'm looking for more than 0. Aug 12 23:57:10 I.E. varying by a factor of 2. Aug 12 23:57:31 It rarely goes to 0. Aug 12 23:57:44 heh.. After plugging it in, I see a swing from 0 to 17714.... Aug 12 23:58:47 I don't think my 500ma wall adapter is working for the neo Aug 12 23:58:56 morning Aug 12 23:59:15 On my phone. Aug 12 23:59:18 sagacis: mine don't seem to work Aug 12 23:59:20 I'll let it sit for a couple hours and watch a movie. I'll see how it does Aug 12 23:59:21 The voltage noise is some 80mV Aug 12 23:59:46 Writchie: I've seen this one not seem to work. Trying to figure out if it really does or maybe why. Aug 12 23:59:54 I threw the packaging away, so I can't return it. Aug 13 00:00:04 I just bought another online. It would suck if it doesn't work also. Aug 13 00:00:45 the current noise is some 600 units. Aug 13 00:02:01 The charge current is some 1300 units dropping to 700 units. Aug 13 00:02:35 any idea what units is Aug 13 00:02:39 foos. Aug 13 00:03:01 foos or fubab's Aug 13 00:03:10 oh hang on, I can guess from voltage curve Aug 13 00:11:10 Eyeballing the graph. Aug 13 00:11:44 Charging while on. Aug 13 00:12:10 It seems to take 5 hours to finally complete charging. Aug 13 00:12:26 The 'zero' looks like 850 or so. Aug 13 00:12:43 The bulk is 1350 or so. Aug 13 00:13:05 So call it 500 over the zero. Aug 13 00:13:29 For 4 hours or so. Aug 13 00:13:46 So that'd be 2000?h Aug 13 00:14:28 So maybe 2*(reading-850) might be a reasonable guess at current. Aug 13 00:16:48 talking battery? Aug 13 00:16:48 Planet: See http://wiki.openmoko.org/wiki/FAQ (and also SH1, SH1_FAQ, Neo1973, Developer_preview and Neo1973_Hardware#GTA01Bv4 pages) (Hopefully these links answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) Aug 13 00:18:39 The fluctuations are so wild, 2(reading-850) will give me many negative numbers. Aug 13 00:18:48 So, can I even tell if it's not charging? Aug 13 00:20:22 sagacis: average a dozen readings Aug 13 00:20:45 I'm assuming anything over 17K is math error Aug 13 00:22:16 maybe ~100, with a few 0's. Aug 13 00:22:22 I'll have to log to a file to do this right. Aug 13 00:22:23 anyone able to vouge for new images? Aug 13 00:22:38 om2007.2 working images that are newer then 8-8? Aug 13 00:23:00 i a watching some build right now Aug 13 00:23:38 Look at http://www.mauve.plus.com/battery.gif Aug 13 00:23:46 for a graph of smoothed results Aug 13 00:25:00 green is voltage Aug 13 00:25:18 you can see that it means nothing when discharrging. Aug 13 00:25:37 240 is my current charging Aug 13 00:27:06 I"m definitely losing voltage off the battery charge over time. Aug 13 00:27:24 I guess my 500ma wall charger doesn't work with the phone. Wonder why? Aug 13 00:27:28 You cannot measure the charging current once. Aug 13 00:27:41 you've got to measure it a few dozen times, and aveage. Aug 13 00:27:53 sagacis : look at 'bootloader' Aug 13 00:28:23 SpeedEvil: I did. I"m also measuring the battery voltage. Since I plugged it in, my average has gone from 4200 to 4100, and that's about 30 minutes. Aug 13 00:28:24 Follow the recommendations to add an option to your bootloader menu to turn it on. Aug 13 00:28:42 fast charge that is Aug 13 00:28:49 Autofast is on.... I had it plugged into the usb, and it was fast charging. Aug 13 00:29:01 Unplugged from computer and plugged into wall adapter, no charging. Aug 13 00:29:16 I tried it because the battery icon doesn't show charging when on the wall adapter. Aug 13 00:29:18 Autodast is a once thing. Aug 13 00:29:22 Autofast Aug 13 00:29:26 unfortuanayely. Aug 13 00:29:40 it will, if plugged into a dumb charger, make it charge fast. Aug 13 00:29:52 there was supposed to be logic to detect impedance of a usb charger Aug 13 00:29:52 If you then turn off, unplug and replug, it won't charge fast Aug 13 00:29:58 There isn't. Aug 13 00:30:13 I didn't turn it off. I just moved the usb cable Aug 13 00:30:14 if there is it doesn't work Aug 13 00:30:23 Didn't even unplug from the device. Aug 13 00:30:38 if the source of 5V gors away, it stops charging fast Aug 13 00:30:50 goes Aug 13 00:31:00 Does it start again if the 5V returns? Aug 13 00:31:06 No. Aug 13 00:31:09 I'm pushing 5.5v from the wall adapter. Aug 13 00:31:20 So, you're saying if I plug into the PC's usb port now, it still won't charge? Aug 13 00:31:20 Doesn't matter. Aug 13 00:31:36 yes - it will - the PC will negotiate that it can take the current. Aug 13 00:31:41 And it will feel happy, and draw it. Aug 13 00:31:56 What does the PC have to say about the phone's charging? Aug 13 00:31:57 and then you can turn device off to get full current Aug 13 00:31:58 * sagacis is confused Aug 13 00:32:21 USB port only supplies 100ma until you ask very nicely for 500ma Aug 13 00:32:26 sagacis: USB has protocols to negotiate current. Aug 13 00:32:32 The neo follows them. Aug 13 00:32:32 Ohhh.... Aug 13 00:32:33 USB enumeration Aug 13 00:32:37 My dumb charger doesn't then? Aug 13 00:32:44 The dumb charger doesn't - and cannot. Aug 13 00:32:50 that's why it's "dumb" Aug 13 00:32:50 As it is not a USB host. Aug 13 00:32:51 except that the neo still draws full charge current when the PC goes into standby Aug 13 00:32:52 I see. Aug 13 00:33:08 There are no smart chargers - in that sense. Aug 13 00:33:10 wonder how much a smart wall charger would add cost Aug 13 00:33:23 $3 cost - $5 price. Aug 13 00:33:30 there are chargers smart enough to present the right impedance Aug 13 00:33:39 to trigger usb charging Aug 13 00:33:41 Which is not what the neo is looking for. Aug 13 00:33:56 neo is particular in that regard Aug 13 00:33:57 At least as far as I could determine. Aug 13 00:33:59 where did i see on the wiki about the 47K resisters and looking for 47K to ground Aug 13 00:34:05 I couldn't make it charge with a resistor. Aug 13 00:34:18 Writchie: assuming hte wiki is accurate is perhaps optimistic. Aug 13 00:34:30 The neo can't read the wiki. Aug 13 00:34:43 So, if I have the neo on, I can't charge from a wall-wart or car? Aug 13 00:34:48 they were going to source a charger - probably a moto knockoff Aug 13 00:35:03 i have a moto charger Aug 13 00:35:06 it's four resistors to make it work Aug 13 00:35:08 sagacis: I've not found a way to make it fast charge. Aug 13 00:35:15 You can slow-charge. Which will probably be less than the Neo eats, at least at this point. Aug 13 00:35:16 sagacis: you can go through u-boot. Aug 13 00:35:20 Yeah, the razor charger worked for me (borrowed a friend's) Aug 13 00:35:24 It's _much_ less. Aug 13 00:35:26 although i remember they were looking for 1% accuracy on the resistors Aug 13 00:35:31 It did? :) Aug 13 00:35:32 and using the wrong values by 2% Aug 13 00:35:58 i have a rzr charger and it doesn't seem to work Aug 13 00:36:01 I guess. I'll try again when I go to work. Aug 13 00:36:17 The fone was dead. I plugged it into the razr charger for a couple hours and it worked again. Aug 13 00:36:26 Maybe it just trikle charged enough. Aug 13 00:36:39 probably Aug 13 00:36:40 it will draw 40ma from anything Aug 13 00:36:43 including a straw Aug 13 00:36:54 when it's off, the trickle will of course slowly charge it Aug 13 00:37:03 slowly as in 20 hr Aug 13 00:37:07 from dead Aug 13 00:37:16 you may be able to get going in 3 or 4 Aug 13 00:38:00 anyway, as the u-boot can do autofast, seems "just" a matter of code to get linux do the same on request... Aug 13 00:38:38 indeed/ Aug 13 00:38:48 Surely Aug 13 00:38:52 but you need to get enough charge to sustain the boot Aug 13 00:39:08 I'm holding at 4.1v with the wall-wart. Aug 13 00:40:44 once you are running there is no problem Aug 13 00:41:05 the problem is getting a dead battery up to the point that the device can boot. Aug 13 00:41:11 without waiting like 4 hours Aug 13 00:41:40 that battery can take 1C which is 1200ma in current limit mode Aug 13 00:42:00 Nope. Aug 13 00:42:13 The charger won't in any circumstances draw >500mA Aug 13 00:42:21 that's true Aug 13 00:42:26 Well. Aug 13 00:42:42 All right.... I'll check it after watching "300" Aug 13 00:42:43 but in a well design charger you can charge this battery in 1 hour Aug 13 00:42:56 yeah. Aug 13 00:43:16 The charger may support 10V input. Aug 13 00:43:26 (it may also make your neo explode) Aug 13 00:43:27 i think it will Aug 13 00:43:37 but I don't think you need that much headroom Aug 13 00:43:55 and you have to worry about frying the USB host Aug 13 00:43:59 10V input would give 5W of input power - you'd need most of that to do 1h charging Aug 13 00:44:32 the '5V' line of the neoUSB is not connected to the SoC. Aug 13 00:45:06 What else it's connected to is another question. Aug 13 00:46:08 i would think 5V would give you the full 500ma into the battery Aug 13 00:46:22 at least when the bat is very low Aug 13 00:46:59 Ideally. Aug 13 00:47:10 >500 into the battery. Aug 13 00:47:43 the dumb charger text was removed from the wiki Aug 13 00:48:08 reliability of the info is questionable Aug 13 00:49:02 SpeedEvil: 48K ohm from pin 5 to ground? Aug 13 00:49:57 anybody got a 4GB microSD in their neo yet? Aug 13 00:50:30 SpeedEvil: sounds usb OTG like Aug 13 00:51:12 Writchie: the true spec for dumb chargers is D+ shorted to D- Aug 13 00:51:28 which true spec? Aug 13 00:51:39 the new USB draft? Aug 13 00:51:55 recent draft != true spec. Aug 13 00:51:56 or the one from CEA Aug 13 00:53:03 Writchie: not a draft, it's an "approved class specification" Aug 13 00:53:12 jeabird: I have a 2GB micro SD--$34 at Costco Aug 13 00:53:17 USB class charger Aug 13 00:53:24 Writchie: http://www.usb.org/developers/devclass_docs/batt_charging_1_0.zip Aug 13 00:53:39 Writchie: http://www.usb.org/developers/devclass_docs#approved Aug 13 00:54:10 also bug 685 Aug 13 00:55:35 the big question remains - what if anything will the neo respond to at the dead battery hardware level to go into fast charge? Aug 13 00:56:00 cesarb: btw, thanks - i found the mystery code Aug 13 00:56:09 at least i think i have Aug 13 00:56:11 Writchie: nothing Aug 13 00:56:33 Or maybe the P* chip can be setup to do autofast all the time. Aug 13 00:56:44 need to look at that Aug 13 00:56:54 the datasheet is insane. Aug 13 00:56:57 now where did i put that datasheet Aug 13 00:56:58 132 pages. Aug 13 00:57:08 with two different versions Aug 13 00:57:10 Writchie: if the USB battery charging spec is implemented, the detection of the short from D+ to D- can be done in hardware... but the current neo doesn't have the needed hardware Aug 13 00:57:56 there is no intelligence in the loop now other than the PM chip Aug 13 00:58:08 Writchie: I looked at detecting it at least when the application processor is running (uboot or linux), but I don't think there's a way to send a high level in one of the USB pins and read the result from the other Aug 13 00:58:27 DukeOfUrl: 68.50 for 4GB @ amazon Aug 13 00:58:31 there is not - but it might be done from pio Aug 13 00:58:42 but the big problem is when the ap is not running Aug 13 00:58:46 I don't think the USB pins are also PIO pins. Aug 13 00:58:56 Writchie: which is sad, because even if it needed uboot or linux running, it would be cool to be able to simply plug a motorola charger (which does short D+/D-) and have it charge at 500mA Aug 13 00:58:57 As I understand it, they are dedicated. Aug 13 00:58:58 they are not, but you could bridge them over Aug 13 00:59:08 IMO. Aug 13 00:59:10 SpeedEvil: yes, they are dedicated, I looked Aug 13 00:59:29 The breakage from charging if the host has not signalled us in 10s is small. Aug 13 00:59:32 shorting them to PIO pins might muck up the impdeance Aug 13 00:59:44 SpeedEvil: and I also didn't find a way in the spec to make the USB block do a gpio-like dance with them Aug 13 00:59:49 The only real case that it casuses problems with is shorting s suspended bus. Aug 13 01:00:16 the suspend isn't working now anyway, the neo continues to draw Aug 13 01:00:22 and that'd possibly be a suspended bus that won't wake up, if we do the wake thing. Aug 13 01:00:24 True. Aug 13 01:00:30 jaebird: plus shipping... whatever, I remember a 200MB disk drive was 24" wide, 36" deep, and 36" high. Used a voice coil to move the heads. Twenty surfaces. Today's technology just blows me away. Aug 13 01:00:35 500mA won't cause it to fail anymore than 50mA. Aug 13 01:00:39 will have to go over the PM chip again Aug 13 01:00:58 at this point I say charge away all the time if its doable Aug 13 01:01:09 IMO - no. Aug 13 01:01:20 Only if the host doesn't negotiate Aug 13 01:01:31 even better Aug 13 01:01:35 However, we need to have a mass storage stub device in there. Aug 13 01:01:40 but you can't negotiate from a dead battery Aug 13 01:01:45 yeah. Aug 13 01:02:32 A PIC would make this so much nicer. Aug 13 01:02:34 how about, as the battery dies we set up the PM to fast charge? Aug 13 01:02:43 s/pic/(pic|avr|...) Aug 13 01:02:50 that's not always optimal either. Aug 13 01:03:02 If hte only power supply you have won't supply 500mA... Aug 13 01:03:02 a PIC or msps430 yes is the way to handle PM Aug 13 01:03:16 the task is difficult enough to dedicate a programmable uC to the task Aug 13 01:03:29 for example, solar, or a USB unpowered hub. Aug 13 01:03:33 Writchie: it would be even better to have a uC core in the PM chip itself Aug 13 01:03:40 IMO - not. Aug 13 01:03:49 A non-dedicated one can do fun stuff too. Aug 13 01:03:58 personally, i like my PM dumb to be managed by a uC Aug 13 01:04:04 but you have to have that uC Aug 13 01:04:37 PIC + USB host/device controller over SPI :) Aug 13 01:04:57 maybe Aug 13 01:05:27 That's overkill though probably - it's another $4 or so. Aug 13 01:05:38 my battvolt seems to have peeked a 4177 Aug 13 01:05:52 DukeOfURL: the batvolt varys by 80mV or so. Aug 13 01:06:08 maybe not if you factor in A/D, restart reliability, etc. Aug 13 01:06:30 although the 2410 seems more reliable in startup than most arms Aug 13 01:07:28 with a uC you could detect >6 V input as a charger Aug 13 01:07:38 and just jump to fast charge Aug 13 01:07:51 in fact, a little hardware could probably do this now Aug 13 01:09:08 i guess nobody on the team ever got their ars kicked in a design review Aug 13 01:09:38 * cesarb still likes the idea of detecting a shorted D+ to D- as a charger better Aug 13 01:10:03 i'm still not sure this is not problems for OTG devices on the other end of cable Aug 13 01:10:39 at least at the margins for devices not conservatively designed Aug 13 01:11:16 Writchie: no, because if I read the spec correctly, the detection is made by a pullup similar to the one normally used for a USB device, but on the _other_ line (not the one which normally has the pullup) Aug 13 01:11:46 Writchie: a normal host will not notice (OTG or not), but with a charger shorting both lines together, you will get a high on the other line Aug 13 01:11:54 how are we going to set the time? CMOS (CMOS battery)? GSM? GPS? Aug 13 01:12:03 Writchie: you just need a weak pulldown on the other line and a comparator Aug 13 01:12:03 GPRS? Aug 13 01:12:26 DukeOfURL: the PM chip has a RTC Aug 13 01:12:39 DukeOfURL: (in fact, I think every single freaking chip on this thing has a RTC) Aug 13 01:12:49 The absolute dead battery case should be rareish. Aug 13 01:12:59 both batteries dead. Aug 13 01:13:47 I think there are only three RTCs. Aug 13 01:13:50 :) Aug 13 01:14:09 what do you mean by both? Aug 13 01:14:16 The backup too. Aug 13 01:14:22 what does this supply? Aug 13 01:14:26 RTC Aug 13 01:14:36 and some P* chip RAM Aug 13 01:15:05 if the main battery is dead, the ap won't run at all - correct Aug 13 01:15:13 yes. Aug 13 01:15:23 so it looks like a dead battery Aug 13 01:15:24 well - the RTC maybe. Aug 13 01:15:34 I'm not sure which RTC it uses. Aug 13 01:15:35 yes, i hope the RTC will continue Aug 13 01:16:05 The current scripts make it look like the PMU chips RTC is preferred. Aug 13 01:16:46 cesarb: i think an existing OTG A-device will have its limits exceeded Aug 13 01:17:01 but this is probably not a problem in practise Aug 13 01:17:49 with a uC you could detect faults e.g. voltage drop below 4.5V after you start fast charge Aug 13 01:18:07 you then know you're connected somewhere you're not supposed to be Aug 13 01:18:33 i guess for now its nokia knockoff chargers and spare batteries Aug 13 01:19:06 Writchie: a OTG-A device would not have its D+ shorted to its D-, so it would not be detected as a charger Aug 13 01:19:40 gah - latest oe is trying to build glibc-intermediate-*2.3.2* Aug 13 01:19:46 how long am i should i wait after the reset and what happens if i don't ? Aug 13 01:19:56 cesarb: i think the problem is under dead battery conditions Aug 13 01:20:16 it's covered in the specs Aug 13 01:21:53 Writchie: which is why the detection should be done by the PM chip, so it would work even with a dead battery Aug 13 01:22:10 yes Aug 13 01:22:14 Writchie: btw, the same specification says you can drain 100mA for as long as you need if you have a dead battery Aug 13 01:22:30 Writchie: (take a look, it's a cool specification... one of these which also explains how to implement it) Aug 13 01:22:37 it would help a lot if the neo would just draw 100ma Aug 13 01:22:41 instead of 40ma Aug 13 01:22:57 with some fraction of that going to anything that's still on Aug 13 01:23:40 is this a hardware only channel? Aug 13 01:23:46 No. Aug 13 01:23:48 the battery may be getting little more than 20ma Aug 13 01:23:50 ok good :) Aug 13 01:24:08 Software, firmware, wetware, hardware. Aug 13 01:24:17 even warez Aug 13 01:24:48 * Planet pretends he doesn't know what that is Aug 13 01:25:09 middleware and vaporware too Aug 13 01:25:15 lots of vaporware Aug 13 01:25:30 oh nice, diperware? Aug 13 01:25:57 i am here, looking for any fellow canadians, who have gotten gsm working Aug 13 01:26:04 let's have a show of hands. 2007.1 Aug 13 01:26:11 2007.2? Aug 13 01:26:43 i am trying to build .2 now, dont want to waste time poking around .1 if its on the way out Aug 13 01:27:01 others? Aug 13 01:27:12 i would guess that FIDO would work but i haven't heard for sure Aug 13 01:27:20 don't know about rogers Aug 13 01:27:35 yah, i can recieve, but can't dial with rogers Aug 13 01:27:47 get, no carrier every time Aug 13 01:27:54 can you talk to the SIM Aug 13 01:28:00 yah, i can recieve Aug 13 01:28:03 lookup phonebook Aug 13 01:28:13 does registration work Aug 13 01:28:16 yup Aug 13 01:28:25 then it should work Aug 13 01:28:25 i can call the phone, answer it, here myself Aug 13 01:28:37 but when i issue a dial, i get no carrier Aug 13 01:28:39 sounds like audio problem Aug 13 01:28:47 and your dial string Aug 13 01:28:52 i have tried them all Aug 13 01:29:11 have you tried +1npanxxxxxx Aug 13 01:29:53 first, are you talking to the gsm modem, the libgsm-tool, or the GUI Aug 13 01:30:07 ok i figured out how to fix it - angstrom wants glibc 2.5 and it needs to be set to 2.6 Aug 13 01:30:12 libgsm(shell, atcmd) and cu Aug 13 01:30:14 all do the same Aug 13 01:30:40 did you put the semicolon at the end? Aug 13 01:30:52 .....no... Aug 13 01:30:56 lol, you serious! Aug 13 01:31:13 you may be going into a data mode Aug 13 01:31:25 i didn't notice that is any gsm docs i found Aug 13 01:31:33 it's there ;-) Aug 13 01:32:07 hmm i need to track down a refill card; grr Aug 13 01:33:12 i cant believe i was beat'n by a semi-colon Aug 13 01:33:21 http://wiki.openmoko.org/wiki/Manually_using_GSM#Dialling_an_outgoing_call Aug 13 01:33:27 better than by a colon Aug 13 01:33:30 yah...i read those, plus a bunch Aug 13 01:33:38 well.. ok, the ; is there Aug 13 01:33:48 http://wiki.openmoko.org/wiki/Gsmd#Usage_of_shell_mode is a good read too Aug 13 01:33:53 yah...i dont know how i could of not put it Aug 13 01:34:32 Planet: yeah, according to the spec, a semicolon at the end means voice call Aug 13 01:34:42 Planet: without it, you probably were doing CSD ;-) Aug 13 01:34:47 arg! Aug 13 01:34:59 i only wasted a few hours.... Aug 13 01:35:02 if you called another cell phone, check your voice mail - you might have a calling tone Aug 13 01:35:02 Planet: did you read 07.07? It's all in there... Aug 13 01:35:06 yah.... Aug 13 01:35:29 i figured there was some rogers specific thing i had to do Aug 13 01:35:34 knowing rogers and all Aug 13 01:36:18 when you confirm working please add it to wiki if its not there Aug 13 01:36:31 trying it now Aug 13 01:37:02 Planet, http://wiki.openmoko.org/wiki/Carriers/Rogers Aug 13 01:37:33 hmm Aug 13 01:37:36 no carrier Aug 13 01:37:54 did you play with the call class Aug 13 01:37:56 ATD+1xxxxxxxxxx Aug 13 01:38:05 ATD+1xxxxxxxxxx; Aug 13 01:38:15 x, being 10 digit phone Aug 13 01:38:41 try ATDTnpaxxx; Aug 13 01:38:51 npaxxxxxxx; Aug 13 01:39:38 i'm sorry, ATDT1npaxxxxxx; Aug 13 01:40:04 do i put, npa, or does that stand for something Aug 13 01:40:10 anybody used the vmware/ubuntu image? Aug 13 01:40:12 area code Aug 13 01:40:18 north american number plan area Aug 13 01:40:23 ok, thats what i figured :) Aug 13 01:40:44 the no carrier means it thinks its a data call. Aug 13 01:40:53 otherwise you would be getting some call progress indications Aug 13 01:41:28 that string kills atcmd mode Aug 13 01:41:43 i get unknown/reject/release Aug 13 01:41:49 then it stops responding Aug 13 01:42:08 using the tool? Aug 13 01:42:14 yah Aug 13 01:42:17 i can try cu Aug 13 01:42:55 ok...i went for it :) the 4GB microSD Aug 13 01:43:14 i'll share my success or failure Aug 13 01:44:01 ok, in CU, with that dial string, ATDTnpaxxxxxxx; i get no carrier Aug 13 01:44:15 need 1 before npa Aug 13 01:44:25 sorry i used it Aug 13 01:44:32 hmmm Aug 13 01:44:43 it finds the number in my phonebook on the sim Aug 13 01:44:49 %CPI: Aug 13 01:45:14 you can dial from that Aug 13 01:45:26 ATD>? Aug 13 01:46:24 IIRC Aug 13 01:46:24 ATD>; Aug 13 01:46:32 where is entry number in phone book Aug 13 01:50:24 no carrier :( Aug 13 01:50:40 does it come back right away? Aug 13 01:50:45 yup Aug 13 01:50:56 but i can call it... Aug 13 01:51:13 Planet: atd isnt your first command is it? Aug 13 01:51:16 outgoing calls barred Aug 13 01:51:31 lol ignore that Aug 13 01:51:35 no its registered Aug 13 01:52:21 try a +CMOD=1 and try again Aug 13 01:52:23 does %CPI: 1,8,0,0,,,... NO CARRIER %CPI: 1,7,0,,,... Aug 13 01:52:26 ok Aug 13 01:55:00 Planet: are you getting signal strength reports Aug 13 01:55:31 csq's?, yes Aug 13 01:55:41 my phone in same location gets full Aug 13 01:55:55 what number are you getting? Aug 13 01:56:01 18,99 Aug 13 01:56:27 18 is about 2 bars but 99 is an undefined value Aug 13 01:56:38 99 could mean no signal Aug 13 01:57:12 do you have both 850 and 1900 at your location? Aug 13 01:57:17 or just 1900? Aug 13 01:57:25 i.e. is your other phone quadband Aug 13 01:57:38 samsung p207 Aug 13 01:57:43 it is i believe Aug 13 01:58:15 i'm not yet convinced that gsm850 works on the neo Aug 13 01:58:43 p207 is 850/1900/1800 Aug 13 01:58:49 or tri-band USA Aug 13 01:58:52 tri Aug 13 01:59:08 this means it will suck on a train in france or germany Aug 13 01:59:28 lol, i am in canada... Aug 13 01:59:57 if only the 1900 is working, your signal may be marginal Aug 13 02:00:07 very low compared to the signal on 850 Aug 13 02:00:12 but, if I can call it... Aug 13 02:00:17 and it rings, and i can hear myself ok Aug 13 02:00:21 Planet: I hope you get it working, I'm with Rogers now but thinking of moving over to Fido since its the same network and I want to know if the neo would work before actually getting one :) Aug 13 02:00:52 the threshold for an inbound call may be a bit lower Aug 13 02:01:09 my touchpad on my laptop suddenly stopped supporting scroll or touch-click Aug 13 02:01:54 if you have time tommorrow we can play with it some more, step by step direct to the gsm modem Aug 13 02:02:18 Planet: Aug 13 02:02:22 with cu? Aug 13 02:02:30 yes Aug 13 02:02:34 thats what I have been using Aug 13 02:02:57 you need to at least run through the same commands that gsmd does at startup Aug 13 02:03:11 fun/creg/cops? Aug 13 02:03:18 yeah Aug 13 02:03:19 etc? Aug 13 02:03:31 yah..i read many gsm docs...most had the same sequence Aug 13 02:03:48 at least your sim is working Aug 13 02:04:04 and inbound working is half of it. Aug 13 02:04:29 yah, i was worried about that, it doesn't have any of the serial numbers etc, on it Aug 13 02:05:18 i'd like to see at least one working gsm850 situation Aug 13 02:05:58 csq 15,99 Aug 13 02:06:01 and inbound works Aug 13 02:06:39 ATD>n; still returns no carrier Aug 13 02:06:50 n being phone book entry Aug 13 02:07:45 15 should work Aug 13 02:08:12 it's possible you can't originate calls on 1900 through their network? Aug 13 02:08:22 but i'm grasping at staws Aug 13 02:08:37 this I dont know....i was hoping someone here might have insight into rogers Aug 13 02:08:58 you are the guinea pig Aug 13 02:09:00 17,99 now, near window Aug 13 02:09:21 17 does not correspond with full bars on the samsung Aug 13 02:10:05 thats quite low i guess, -109 to -53 Aug 13 02:10:53 it should work on 1900 with a 17 Aug 13 02:11:07 is the neo only using 1900? Aug 13 02:11:16 I have doubts. Aug 13 02:11:22 its supposed to be both Aug 13 02:11:23 I have 50cm thick stone walls. Aug 13 02:11:36 castle? Aug 13 02:11:36 1900 does _not_ propagate through these. Aug 13 02:11:41 No. Aug 13 02:11:55 A 1500s cottage that was rebuilt several times. Aug 13 02:11:55 no 1900 in scotland Aug 13 02:11:59 1800 there Aug 13 02:12:05 s/1900/1800/ Aug 13 02:12:10 i got the neo in the window Aug 13 02:12:20 pretty much..a screen in the way, keeping bugs out Aug 13 02:12:30 metal screen? Aug 13 02:12:34 its too cold for bugs in canada Aug 13 02:12:37 hehe, i am not that stupid Aug 13 02:12:44 cold..i am cooking Aug 13 02:13:01 3 fans Aug 13 02:13:20 Oh, I fogot, August 10 - 20th is summer Aug 13 02:13:42 i blame 2007.2 building Aug 13 02:13:54 keeping cpu at 100% for hours Aug 13 02:13:56 EMI from the build Aug 13 02:14:14 that's interfering with you GSM ;) Aug 13 02:14:22 hehehe Aug 13 02:14:34 its got issues, wont build libgsmd Aug 13 02:15:00 i'm not going to touch it until success is reported Aug 13 02:15:11 i'll just keep working with the last build Aug 13 02:15:23 too many changes at the same time Aug 13 02:15:36 i didn't want to waste time..if it was all going to change Aug 13 02:15:43 can someone recommend a good and affordable hotel from sf? Aug 13 02:16:01 motel 8 Aug 13 02:16:37 damn you rogers! Aug 13 02:16:42 screwed me again Aug 13 02:17:18 planet: any specific one in mind? Aug 13 02:17:29 lol, i dont even know sf Aug 13 02:17:36 okay Aug 13 02:18:13 i dont use many hotels, just camp Aug 13 02:22:49 xaid|work: Have you tried yours? Aug 13 02:44:01 hello all Aug 13 02:44:06 ello Aug 13 02:45:32 I am hoping that my iax client will work...Do you think a properly coded client (SIP/IAX) would work well over GPRS/GSM/CDMA/other technology here? Aug 13 02:45:53 Should I start with an application and then integrate it into the dialer as a "Gateway" or the other way around... Aug 13 02:46:13 i'd run it over the wifi, the gprs is high latency Aug 13 02:46:43 BryceLeo: Thats exactly what I was thinking :[ Aug 13 02:46:49 but the wifi capability would be amazing. Aug 13 02:47:15 yeah it will be pretty sweet Aug 13 02:49:51 BryceLeo: Should I do an integration as a "gateway" in the dialer (eg dial through: iax/ tmobile (example)) or a whole nother application? Aug 13 02:51:20 hmm, you could do it as an app, however have it integrated into the panel of the dialer so that when you swap to it activates the application Aug 13 02:51:30 then you could activate as a normal app or through the panel on the dialer Aug 13 02:52:53 Right, I would probably seperate the two. Aug 13 02:53:15 Getting used to the whole development envirnment has an interesting learning curve Aug 13 02:53:43 haha yes i sympathise totally Aug 13 02:53:48 :[ Aug 13 02:53:55 All well, ill get used to it. Aug 13 02:54:06 Just makes the whole development process Slower :/ but its ok Aug 13 02:54:36 hahaha it's a learning process Aug 13 02:54:53 don't be afraid to add tips and tricks to the wiki Aug 13 02:54:59 tel lme about it and i'll include it Aug 13 02:55:54 BryceLeo: Awesome, sounds good. I am currently just following the guide right now, got so caught up in starting the gtk programming interface, I did not even read about how to integrate your app and build an image with it on there, hehe Aug 13 02:59:08 bkruse_home: which guide? **** ENDING LOGGING AT Mon Aug 13 02:59:56 2007