**** BEGIN LOGGING AT Tue Jan 22 02:59:56 2008 Jan 22 03:00:15 The front page of www.openmoko.org is the old GUI. Jan 22 03:02:56 ahh Jan 22 03:30:04 btw I just noticed an update to powermanagement, hopefully this is promising Jan 22 05:22:41 hellow fellow mokomonkeys! Jan 22 05:43:35 * raster throw poo @ holycow Jan 22 05:44:43 * SpeedEvil almost did that joke. Jan 22 05:45:33 * raster took the poo and ran with it Jan 22 05:47:33 * SpeedEvil puts down the cup labelled 2monkeys, and walks away. Jan 22 05:52:52 DYNAMIC ENTRY Jan 22 05:54:16 does the toolkit not have alsa headers? Jan 22 05:54:32 the one from here http://wiki.openmoko.org/wiki/Toolchain (toolchain rather) not have alsa headers? Jan 22 05:59:26 also, who would I go to, to help get a library included Jan 22 06:10:05 ~logs Jan 22 06:10:07 [logs] apt/ibot/infobot/jbot/purl all log daily to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz Jan 22 06:20:37 Hi, I update gsmd Patches When I try for testing AT commands I got error merrage "unable to handle packet type=2" Why ? My gsmd version is 0.0.1+3612 Jan 22 06:27:34 * bkruse_home tries to find gsm.h Jan 22 06:30:18 purple : I am not getting Jan 22 07:12:12 Good morning, good morning, good morning Jan 22 07:13:08 good moko to you too kind sir Jan 22 07:13:31 morning :) Jan 22 07:17:43 * bkruse_home is still awake from the night before Jan 22 07:17:44 :d Jan 22 07:18:19 holycow: i'm not a lord i'm a lady Jan 22 07:18:27 good morning none the less Jan 22 07:18:44 *roh* *roh* Jan 22 07:18:50 sorry! Jan 22 07:18:50 i just flashed the latest OpenMoko images in QEMU Jan 22 07:19:01 does anyone here use QEMU? Jan 22 07:19:08 hehe Jan 22 07:19:09 bkruse_home: heh Jan 22 07:19:11 zivjo miha. Jan 22 07:19:13 qemu is for whimps Jan 22 07:19:40 real men write down arm machine code on a notepad and execute it by hand with pencil and paper Jan 22 07:19:41 :) Jan 22 07:19:51 :) Jan 22 07:20:05 in machine code right? Jan 22 07:20:08 10101010101111 Jan 22 07:20:10 she said she was a lady, so Jan 22 07:20:21 yeah, good point Jan 22 07:20:24 idioterna: doesn't mean she can't try and be a "real man" :) Jan 22 07:20:26 hehe Jan 22 07:20:44 ok we are getting away from the point Jan 22 07:20:57 used qemu once for arm Jan 22 07:21:01 used it many times fo rx86 Jan 22 07:21:05 i find it women seldom try that for the fear of insulting all the men that can't do it Jan 22 07:21:08 but i tend to just rn on my actual device Jan 22 07:21:35 anyway, i don't really use qemu, but i flash my neo regularly :) Jan 22 07:21:39 idioterna: life's no fun without insulting people along the way :) Jan 22 07:21:43 ok, well i needed some help with QEMU Jan 22 07:21:54 but if none of you people have used it recently Jan 22 07:22:02 i should probably try later Jan 22 07:22:26 try the paper and pencil method in the meantime Jan 22 07:22:31 it'll keep you busy Jan 22 07:22:39 at any rate Jan 22 07:22:39 :) Jan 22 07:22:51 yes, great idea Jan 22 07:23:01 that will definately happen Jan 22 07:23:36 * raster pulls out his pencil and paper and starts scribbling wildly Jan 22 07:26:46 * bkruse_home is dancing in cross-compile land Jan 22 07:27:26 * cosmin fell asleep while using monotone Jan 22 07:30:36 haha Jan 22 07:30:41 I hate those days :/ Jan 22 07:37:41 * bkruse_home has school in 6 hours :[ Jan 22 07:38:11 ... and your stil awake? uuh :D Jan 22 07:47:30 ya I know :/ Jan 22 07:47:38 voip soft client has me thinking :] Jan 22 08:14:22 good morning Jan 22 08:15:30 good morning Hopscotch Jan 22 08:22:40 hey guys Jan 22 08:22:40 i used this doc (http://wiki.openmoko.org/wiki/Toolchain) to make my first build which required multiple libraries not included, and I got them all to cross compile, but the tutorial just says compile them on the native system, them scp your resulting ipkg over to the neo Jan 22 08:22:40 and of course, the neo does not have the libraries Jan 22 08:34:28 alsamixer: function snd_ctl_open failed for default: No such file or directory Jan 22 08:34:29 ideas? Jan 22 08:35:21 when might the consumer version of a moko device come to market? Jan 22 08:35:35 mine is dying and i'd hate to buy something for the interim Jan 22 08:36:33 can't you just get a 4.99 cheapo phone? Jan 22 08:37:17 well thats a damn fine point, it's a lot more fun anticpating the moko tho Jan 22 08:37:17 lol Jan 22 08:38:17 :) Jan 22 08:39:16 are there any reasonable guesses on release date tho? Jan 22 08:40:07 gta02 is expected in march... Jan 22 08:40:43 but judging by gta01 I wouldn't expect a consumer version of the software until december Jan 22 08:41:03 aha, danke Jan 22 08:41:03 i only want a gta02 so i can write software for it, not use it as a consumer device Jan 22 08:41:14 that's a guess, but progress has been slow to now Jan 22 08:41:49 well, you don't need the device to write software for it torpor :) Jan 22 08:42:15 yeah i do Jan 22 08:42:25 i hate writing software i can't test on an end user system Jan 22 08:43:30 maybe just maybe gta02 might come in a format with a foldout keyboard/thumbpad Jan 22 08:43:44 but then just seeing it on the market will be cool Jan 22 08:46:50 :( Jan 22 08:47:25 pandora might be what you are looking for if you want a keyboarded device. Jan 22 08:47:28 that would be nice, but... in reality :( Jan 22 08:47:42 lol, that's just butt ugly torpor ... Jan 22 08:48:32 * holycow googles pandora Jan 22 08:49:09 mostly what i'm looking for is an open phone with some kind of kb that our staff can use Jan 22 08:49:20 instead of going to a blackberry or one of the other devices Jan 22 08:49:42 wel, I know it's a lame work around, but you can use a BT keyboard Jan 22 08:49:53 which one youd recommend? Jan 22 08:50:05 torpor: what is this pandora thing your describing? Jan 22 08:50:27 yeah i thought abaout bt kb ... but it would haveto be in some sort of sleeve Jan 22 08:50:58 it doesn't exactly let them stand and typeaway as they are crossing a busy street without looking tho! Jan 22 08:51:18 yes our it dept has been folded into the hr dept Jan 22 08:51:24 itst how we thin the herd Jan 22 08:51:26 *cough* Jan 22 08:51:27 yeah, but it might stop them getting kiled Jan 22 08:51:29 no seriously Jan 22 08:51:29 kiled Jan 22 08:51:41 lll Jan 22 08:51:42 lllll Jan 22 08:51:46 l works Jan 22 08:51:48 i really really am working hard to avoid that bb crap and wince and all that nonsense Jan 22 08:51:49 * buz only finds the online music radio pandora Jan 22 08:52:13 torpor: yeah i can't find any red: pandora either Jan 22 08:52:17 http://openpandora.ca/ Jan 22 08:52:17 holycow: if you find a linux device with qwerty be sure to drop us a line ;P Jan 22 08:52:29 oh there's the guppy Jan 22 08:52:34 but it's BUTT ugly Jan 22 08:52:56 buz: lol i think moko is our first real opportunity here Jan 22 08:53:25 oh heh Jan 22 08:53:28 what the hell is that? Jan 22 08:53:40 i don't care how ugly it is, does it work? its a cell phone? Jan 22 08:53:47 it is a phone Jan 22 08:53:51 if only i could find it Jan 22 08:53:55 saw it on linuxdevices Jan 22 08:54:10 oh this is a phone eh? Jan 22 08:54:25 hard to tell from pics, looks like my umpc Jan 22 08:54:45 those are 3d renders is this in production at all? Jan 22 08:55:04 * FuzzyCat points at ROAD Jan 22 08:55:10 road is vaporware Jan 22 08:55:14 (mostly) Jan 22 08:55:17 oh heh Jan 22 08:55:19 agreed Jan 22 08:55:26 pandora.bluwiki.org/ Jan 22 08:55:36 i honestly don't care how ugly it is Jan 22 08:55:48 something like that would be absolutely perfectc Jan 22 08:55:52 http://www.linuxdevices.com/news/NS6946488091.html Jan 22 08:56:50 pandora is in prototype stages right now .. Jan 22 08:57:10 pandora is not a phone from what i see? Jan 22 08:57:11 really? Jan 22 08:57:25 torpor: what is REALLY cool is the game joystick at the top of kb Jan 22 08:57:27 buz: I'd buy that phone Jan 22 08:57:39 the gupp thing? Jan 22 08:57:40 i have a fujitsy umpc ... i'ave wanted those buttons there from day one Jan 22 08:57:48 torpor: do you work for them or ... ? Jan 22 08:57:50 its not a phone but it will run VOIP Jan 22 08:58:00 may as well buy a n810 then Jan 22 08:58:04 that's shipping Jan 22 08:58:05 buz: yep, to replace my Treo650 Jan 22 08:58:09 no i don't... i've been a big fan of the GP2X scene for a long time and i'm following a lot of linux gadgets these days Jan 22 08:58:10 and actually looks quite nice Jan 22 08:58:23 gp2x info: gbax.com Jan 22 08:58:45 pardon my ignorance, what is gp2x ? Jan 22 08:58:53 opensource gameboy Jan 22 08:58:54 yay pandora Jan 22 08:58:58 http://gp2x.co.uk/ Jan 22 08:59:16 oh and then there's the ultimate vaporware Jan 22 08:59:20 e28 2881 Jan 22 08:59:39 pity pandora doesnt have CF, I have a CF GSM card right here Jan 22 08:59:40 http://www.e28.com/e28mobile/mobile_e2881.htm Jan 22 09:00:08 torpor: if this thing can make phone calls and is cheap enough Jan 22 09:00:12 it really could be what i'm looking for Jan 22 09:00:27 holycow: it won't make calls out of the box, but you can add that feature i'm sure .. ;) Jan 22 09:00:51 hmm Jan 22 09:01:04 one day there is sure to be SDIO gpm modems Jan 22 09:01:06 pandora with a decent keyboard and a vga out might give the eee a run for its money Jan 22 09:01:30 thats running linux too? Jan 22 09:01:31 holy moly Jan 22 09:02:07 holy crap that e28 thing is cool, i haven't seen that before Jan 22 09:03:03 lol all those features and it doesn't do imap! Jan 22 09:03:58 woooow indeed Jan 22 09:04:28 i would buy THAT Jan 22 09:04:44 the e28 doesnt exist Jan 22 09:04:52 it's been floating around for at least year Jan 22 09:04:54 i would buy that right now infact Jan 22 09:04:58 and never ever been seen by anyone Jan 22 09:05:07 just look at the specs and its size Jan 22 09:05:11 way too good to be true Jan 22 09:05:19 pretty much i think Jan 22 09:05:50 it may in theory be possible to build such a phone Jan 22 09:06:06 but probably only for a tier 1 company Jan 22 09:06:16 nokia could pull it off Jan 22 09:06:50 well the technology isn't really all that far away Jan 22 09:06:57 the hardware specs of the moko can run all that Jan 22 09:07:04 the problem isn't hte hardware its the software Jan 22 09:07:11 sounds like they are doing that google thing Jan 22 09:08:03 e28? no their have their own linux stack Jan 22 09:08:07 they do ship some phones Jan 22 09:08:10 and have so for a while Jan 22 09:08:15 just not the 2881 Jan 22 09:09:32 someone should beat nokia to senses Jan 22 09:09:36 that thing will be expensive Jan 22 09:09:44 why are we still waiting for the gsm enabled n810 Jan 22 09:09:56 can't be because they CAN'T do it Jan 22 09:10:28 heh Jan 22 09:10:32 buz: internal market politics Jan 22 09:10:42 yeah but still Jan 22 09:10:42 well they are toying aroung with the n880 or whatever it is Jan 22 09:10:43 it's stupid Jan 22 09:10:46 finally it has a kb, thats fine Jan 22 09:10:49 and carrier relations Jan 22 09:10:51 yeah but still no gsm Jan 22 09:10:57 but its not like people aren't telling them what they want Jan 22 09:11:15 the day they release the n810 with gsm, i'll be first in line Jan 22 09:11:33 that would beat the crap out of any htc device Jan 22 09:12:54 or you'll just use dun :) Jan 22 09:12:57 or pan Jan 22 09:13:03 means i need to carry 2 devices Jan 22 09:13:07 which is exactly what i dont want Jan 22 09:13:11 buy a packback Jan 22 09:13:21 well there is a serial port hidden in the battery compartment, connect a gsm modem to that and sell it as a third party upgrade :-) Jan 22 09:13:22 n810 is big too Jan 22 09:13:25 i replaced my laptop with a umpc, its great for pretty much 855 of my workload Jan 22 09:13:32 i could live with an n810 as phone Jan 22 09:13:37 now just to frickin avoid crackberries, trios and wince devices Jan 22 09:13:46 it's a bit big but considering what it does... Jan 22 09:13:47 and i have a clear path to linux everywhere all the time Jan 22 09:13:58 buz: I guess they didnt want any ngage fiasco (taco talking) Jan 22 09:14:12 no it's pure politics Jan 22 09:14:18 speaking of which Jan 22 09:14:23 are there pluggable cel cards? Jan 22 09:14:25 anyone with a half brain would figure out the n810 would be a hit with gsm Jan 22 09:14:45 holycow: I have a Compact Flash GSM Jan 22 09:14:45 i can't imagine it would be unheard of to have some sort of external cel network card that accepts a sim pluggable via usb Jan 22 09:14:55 those are widely available Jan 22 09:15:00 really? Jan 22 09:15:00 holycow: there are usb ones as well Jan 22 09:15:05 oh what? Jan 22 09:15:05 generally go by the name of GSM/UMTS modems Jan 22 09:15:15 oh! no shit? Jan 22 09:15:17 google for mobile broadband Jan 22 09:15:22 thanks! Jan 22 09:15:24 buz: but I tell ya before it was launched I tried it and talked with a dev from nokia. I loved it already then Jan 22 09:15:38 holycow: leave the 80's we are in the 00's now :-D Jan 22 09:15:40 also exist in pcmcia, expresscard and cf form factors Jan 22 09:15:45 buz: also I tried it later and it was still great :) Jan 22 09:15:52 XorA: hehe :) Jan 22 09:15:54 yeah but it wont do as a phone replacement Jan 22 09:15:59 * holycow puts away his brick cell phone Jan 22 09:16:00 now if nokia would add a cf slot... Jan 22 09:16:00 lol Jan 22 09:16:07 are any of these linux compatible? Jan 22 09:16:25 the RTM-8800 is, and the one t-mobile UK sell is AFAIK Jan 22 09:16:39 http://www.intomobile.com/wp-content/uploads/2007/09/e51.jpg which one is the best color? Jan 22 09:16:45 aha Jan 22 09:17:03 i like the black one Jan 22 09:17:10 white one is quite cool looking too Jan 22 09:17:13 and i'm seriously considering getting an e51 Jan 22 09:17:19 black would be 100% safe choice Jan 22 09:17:30 Im getting it to be my bt dun modem and sms machine :p Jan 22 09:17:39 black one sort of looks like a thinkpad turned into phone Jan 22 09:17:44 oh! damn! Jan 22 09:17:52 I love thinkpads :D Jan 22 09:17:52 which is good :P Jan 22 09:18:06 already drooling over that thinkpad x300 Jan 22 09:18:10 yeah! Jan 22 09:18:27 I have a wannabe thinkpad :P Jan 22 09:18:28 but i only bought a laptop in august Jan 22 09:18:34 so no chance buying another one Jan 22 09:18:44 http://www.notebookcheck.it/typo3temp/pics/994558fd26.jpg Jan 22 09:18:54 hey guys Jan 22 09:18:58 I have my gta01 and i want to copy all the modules over but I only get about 3 of them then I run out of space :/ Jan 22 09:18:59 * FuzzyCat cried when T20 died :( Jan 22 09:19:00 ideas? Jan 22 09:19:03 :( Jan 22 09:19:15 oh anyway, black it is Jan 22 09:20:28 is there something I have to do to get my sound card to work with gta01? Jan 22 09:20:46 maito: nice ... i have one of these --> http://geekstuff4u.com/product_info.php?manufacturers_id=&products_id=582 Jan 22 09:21:22 maito: dont you think the e51 is a bit expensive just as an bt modem? Jan 22 09:21:38 if you dont need the wlan, consider nokia 6120 Jan 22 09:21:40 bkruse_home: boot off SD card .. Jan 22 09:21:46 http://www.sierrawireless.com/faq/ShowFAQ.aspx?ID=601 <-- interesting Jan 22 09:22:04 torpor, ? Jan 22 09:22:05 wtf Jan 22 09:22:06 buz: other choices were: e90, samsung g800, and htc tytn II Jan 22 09:22:17 e51 then Jan 22 09:22:26 buz: I want something quality, it will have to last for 2 years and it needs to do HSDPA Jan 22 09:22:44 thats cool, opensource drivers Jan 22 09:23:04 vodafone are good at that too, they have an sf project Jan 22 09:23:05 * maito dropped his nokia 7110, yes, the first WAP device, for like 1000 times Jan 22 09:23:21 heh Jan 22 09:23:23 FuzzyCat: what? Jan 22 09:23:26 after more than 4 years my p900 still lives Jan 22 09:23:33 FuzzyCat: where are we wrt audio atm? Jan 22 09:23:34 bkruse_home: boot off SD card .. Jan 22 09:23:35 and this way Id get 3g speeds on openmoko when the freerun is released :) Jan 22 09:23:41 to solve the no audio? Jan 22 09:23:46 s/freerun/freerunner Jan 22 09:23:52 xcasex, still broken.. Jan 22 09:23:58 FuzzyCat: no, to solve the no disk space dilemna.. Jan 22 09:24:15 ahh, delayed answer from you then Jan 22 09:24:34 holycow: nice umpc Jan 22 09:24:36 :/ Jan 22 09:24:45 is there something I have to do to get my sound card to work with gta01? Jan 22 09:24:47 bkruse_home: boot off SD card .. Jan 22 09:24:49 lol Jan 22 09:24:49 okay, away time... gotta do some work. Jan 22 09:25:22 holycow: but Id like to have something that I can carry with me and that has more power - I bought a 12" core2duo Jan 22 09:25:48 if you want one with audio you might try the 1207 image xcasex .. Jan 22 09:25:56 iirc that was the last one that worked ok Jan 22 09:26:34 maito: yeah *nod*. i just couldn't get around having to carry around a full laptop Jan 22 09:26:37 or even a subnotebook Jan 22 09:26:42 heh Jan 22 09:26:50 * buz carries around a 15" machine when he really needs it Jan 22 09:27:02 there just arent any decent 12" ones below 3000USD Jan 22 09:27:05 my boss just got a 12" panasonic w7 Jan 22 09:27:09 Im tall and strong ;) I can carry a subnotebook + a tablet/pda + a phone with me Jan 22 09:27:11 very light but still way too big Jan 22 09:27:13 for me Jan 22 09:27:19 maito: lol! Jan 22 09:27:26 really I am Jan 22 09:27:34 buz: x-serie thinkpad Jan 22 09:27:39 * holycow sends over a palatte of spinach Jan 22 09:27:40 :) Jan 22 09:27:44 x300 would be nice ;) Jan 22 09:28:52 x series thinkpad have lousy 1024*769 screen Jan 22 09:28:52 also with subnotebooks, you can spend countless hours trying to find one with the right price + linux comptibility + durability Jan 22 09:29:14 http://www.linuxquestions.org/hcl/showproduct.php/product/3947 Jan 22 09:29:17 looking at the specs, the x300 should be very linux compatible Jan 22 09:29:17 veeeeery interesting Jan 22 09:29:22 honestly I dislike the 1280x800, 1366x768 from sony (11.1") is also bad but better Jan 22 09:29:28 thank you guys for the heads up ... this could be a huge help Jan 22 09:29:47 yeah 1280x800 is lousy too Jan 22 09:29:55 which is why i love the 1440x900 in the x300 Jan 22 09:29:56 FuzzyCat: ya, I was kind of confused about that answer too :P Jan 22 09:29:59 i just decided all resolutions are lousy Jan 22 09:30:02 looking at the x300, it wont be under 3000usd when its worth to order :D Jan 22 09:30:04 (of course, my 15" has 1920x1200) Jan 22 09:30:09 i have dual monitors at work and its still not enough realasted Jan 22 09:30:13 buz: ah! thats what we need on 12" Jan 22 09:30:21 i'm not contemplating a quad 24 inc moni setup and i'm sure i'm still going to complain Jan 22 09:30:26 holycow: and you can do with 1024x600? O_o Jan 22 09:30:32 on your umpc Jan 22 09:30:40 yes actually Jan 22 09:31:00 linux desktops really helps of course Jan 22 09:31:01 whats the dpi of openmoko screen? Jan 22 09:31:31 285 Jan 22 09:31:45 whereas laptop screens are usually better of computers, 120dpi or less Jan 22 09:31:50 and desktops can be 96 Jan 22 09:32:20 okay now, bye Jan 22 09:33:18 bye Jan 22 09:33:30 ugh this audio thing is killing me Jan 22 09:33:44 anyone here uses OpenMoko w/ QEMU? Jan 22 09:33:44 i need an affordable 24" Jan 22 09:34:57 is there any reason / should already be full? Jan 22 09:35:11 of course the drivers wont fit for the sd card either Jan 22 09:35:12 ugh! Jan 22 09:38:16 is it normal to have the module snd_pcm_oss loaded? Jan 22 09:44:07 openmoko: 03werner * r3911 10/branches/src/target/kernel/2.6.24.x/patches/series: Jan 22 09:44:07 openmoko: Temporarily reinstating the disable-hwecc patch, since we get streams of Jan 22 09:44:07 openmoko: ECC errors in some cases that still need investigating. Jan 22 09:44:07 openmoko: series: s3c2440-nand-disable-hwecc.patch Jan 22 09:44:28 openmoko: 03werner * r3912 10/branches/src/target/kernel/2.6.24.x/scripts/build: scripts/build: updated to KERNEL 2.6.24-rc8 and SVN_REV 3911 Jan 22 09:50:58 how do I erase the old rootfs? Jan 22 09:51:12 cu -l /dev/ttyACM0 wont work for me Jan 22 09:58:54 Hi, I update gsmd patches both on host and on target but while executing AT command application get hangs Jan 22 10:50:00 bkruse_home: why do you want to erase it? why not just put a new one over the old one using dfu-util? Jan 22 10:50:46 Field: new question, how can you debug applications? Jan 22 10:50:58 My app seems to be dying and crashing the whole phone with it :/ Jan 22 10:55:35 i want to know how to debug applications Jan 22 10:55:47 gdb isn't up to anything but the simplest Jan 22 10:55:57 gdb was what I thought Jan 22 10:56:02 is it in the repos? Jan 22 10:56:10 yes Jan 22 10:56:14 rocking, perfect Jan 22 10:56:21 hopefully I can keep it from crashing my phone Jan 22 10:56:57 yes, well i tried gdb on the phone and it worked for simple apps Jan 22 10:57:04 interesting Jan 22 10:57:05 ty :] Jan 22 10:57:06 are u able to test the app on your desktop? what kind of crash is it? Jan 22 10:57:25 it locks the phone completely Jan 22 10:57:41 adding some debug statements, does not happen on my desktop, my guess is because its a library isue Jan 22 10:58:23 is a C app? Jan 22 10:58:24 miha: why do you say gdb is only for simple apps? Jan 22 10:58:48 while i understand how difficult debugging python can be using gdb, it's still a formidable tool Jan 22 10:58:58 zedstar: yes Jan 22 10:59:26 idioterna: because it just didn't work, it did not set the breakopints where i wanted Jan 22 10:59:57 miha: was your application written in C? Jan 22 11:00:14 i mean the one you were trying to debug Jan 22 11:00:15 and it's also difficult to use gdb on the phone because you have to type using the software keyboard Jan 22 11:00:21 yes it is in C Jan 22 11:00:25 bkruse_home: why can u add some printf or some kind of debug though to the desktop version? Jan 22 11:00:29 cant Jan 22 11:00:30 well then just break file.c:linenumber Jan 22 11:00:43 zedstar: I can Jan 22 11:00:44 and run gdb over USB network link Jan 22 11:00:47 the desktop version does not crash Jan 22 11:00:47 so i tried gdbserver Jan 22 11:00:59 bkruse: so it runs fine on desktop but just crashes phone? Jan 22 11:01:08 zedstar: yep Jan 22 11:01:14 * miha goes to lunch Jan 22 11:02:04 bkruse: the same debug output not showing u how far it is getting? Jan 22 11:04:26 I think to where it is doing some audio initilization Jan 22 11:24:05 Hi, I update gsmd patches both on host and on target but while executing AT command application get hangs Jan 22 11:49:39 are there no prebuilt qemu images? Jan 22 11:59:08 openmoko: 03chris * r3913 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (9 files in 3 dirs): (log message trimmed) Jan 22 11:59:08 openmoko: Patch by: Frank Li Jan 22 11:59:08 openmoko: * Makefile.am: Jan 22 11:59:08 openmoko: * bindings/python/Makefile.am: Jan 22 11:59:08 openmoko: * bindings/python/moko.override: Jan 22 11:59:09 openmoko: * bindings/python/mokomodule.c: (initmoko): Jan 22 11:59:11 openmoko: * configure.ac: Jan 22 12:24:11 xcasex: there were - I think it may be very old by now. Search on Qemu in the wiki. Jan 22 12:24:31 Atul_: what 'AT' command are you trying to send? Jan 22 12:24:32 ungh Jan 22 12:26:10 openmoko: 03werner * r3914 10/branches/src/target/kernel/2.6.24.x/patches/gta02-core.patch: (log message trimmed) Jan 22 12:26:10 openmoko: 2.6.24 did not match the vibrator LED device to the right driver Jan 22 12:26:10 openmoko: This fixes that and allows control of the vibrator by the following Jan 22 12:26:10 openmoko: path: Jan 22 12:26:10 openmoko: echo 127 >/sys/bus/platform/devices/neo1973-vibrator.0/leds/neo1973\:vibrator/brightness Jan 22 12:26:14 openmoko: Signed-off-by: warmcat Jan 22 12:26:16 openmoko: gta02-core.patch Jan 22 12:59:37 hi Jan 22 13:00:31 anyone here? Jan 22 13:01:29 No. Jan 22 13:01:42 not me Jan 22 13:01:50 Be afraid, Be very afraid! In space, noone can hear you scream! Jan 22 13:01:52 oops Jan 22 13:02:22 you could be marooned, just like that!! Jan 22 13:02:27 I'd like some information :) Jan 22 13:03:45 "don't ask to ask. ask!" Jan 22 13:03:56 Or pay to ask. Jan 22 13:04:28 I'm a college student, I have to do a thesi on operating system Jan 22 13:04:49 Operating systems are bad! Jan 22 13:04:58 I'd like the hw ref about the samsung processor on neo1973 Jan 22 13:05:08 look on the wiki Jan 22 13:05:21 under neo1973 -> neo1973 hardware Jan 22 13:05:25 yeah... but there is the full docs Jan 22 13:05:32 yes tehre is Jan 22 13:06:32 only the user manual http://www.samsung.com/global/business/semiconductor/productRightMenuDown.do?doc_file=26537 Jan 22 13:07:29 what do you mean? Jan 22 13:07:34 by full manual? Jan 22 13:07:50 if that's the one I'm thinking of, it's got details of the peripherals, register maps, ... Jan 22 13:16:04 on samsung support site there are a lot of docs but for security reason there aren't available for download :( Jan 22 13:16:25 I think that the register maps, etc are in these docs... Jan 22 13:18:52 google um_s3c2410a_rev11.pdf Jan 22 13:23:58 only 3 links but not so useful http://www.google.it/search?hl=it&q=um_s3c2410a_rev11.pdf&btnG=Cerca+con+Google&meta= Jan 22 13:24:18 hi all, i guess the there are a lot of guys from germany, i heard that courses in technical degrees are in german and not in english..is that true? Jan 22 13:26:04 Iko974629: google s3c2410a datasheet Jan 22 13:26:38 Sup3rkiddo: in germany? Jan 22 13:26:53 SpeedEvil, yeah Jan 22 13:29:56 ok thx Jan 22 13:40:40 openmoko: 03erin_yueh * r3915 10/trunk/src/target/gsm/ (4 files in 3 dirs): gsmd: retrieve imsi value before sending at command (Sean Chiang) Jan 22 13:41:09 hi there. Jan 22 13:46:09 ahem. Jan 22 13:46:33 what else is bothering the dpi setting except the /etc/X11/Xserver file Jan 22 13:48:43 is anyone else having trouble with 'make flash-qemu-local' from MokoMakefile? Jan 22 13:54:41 hi bedboi Jan 22 13:54:49 hey florian Jan 22 13:59:52 SpeedEvil, AT Command like AT, AT+CGMI, AT+CGMR Jan 22 14:15:31 freesmartphone.org: 03emdete * r52 10/trunk/software/py-proto/ (13 files): (log message trimmed) Jan 22 14:15:31 freesmartphone.org: added gprs enabling daemon Jan 22 14:15:31 freesmartphone.org: added cellid-to-location daemon Jan 22 14:15:31 freesmartphone.org: simplified dbus api Jan 22 14:15:31 freesmartphone.org: send signals only for status-changes and allow request for status Jan 22 14:15:32 freesmartphone.org: added code for activation from muxer Jan 22 14:15:36 freesmartphone.org: seperated config into an ini file Jan 22 14:24:17 whats going on this the FreeRunner? Jan 22 14:28:33 Not much. Jan 22 14:28:42 Driver development work in-house at OpenMoko Jan 22 14:32:11 openmoko: 03chris * r3916 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (ChangeLog libmokoui/moko-finger-scroll.c): Jan 22 14:32:11 openmoko: * libmokoui/moko-finger-scroll.c: (parent_set_cb), Jan 22 14:32:11 openmoko: (moko_finger_scroll_add), (moko_finger_scroll_dispose), Jan 22 14:32:11 openmoko: (moko_finger_scroll_init), (moko_finger_scroll_mode_get_type): Jan 22 14:32:11 openmoko: Fix removal of widgets from a MokoFingerScroll causing crashes/weird Jan 22 14:32:15 openmoko: behaviour Jan 22 14:49:10 openmoko: 03thomas * r3917 10/trunk/src/target/opkg/opkg_download.c: opkg: Fix #ifdef Jan 22 15:06:40 openmoko: 03chris * r3918 10/trunk/src/target/OM-2007.2/applications/openmoko-messages2/ (6 files in 2 dirs): (log message trimmed) Jan 22 15:06:40 openmoko: * src/sms.h: Jan 22 15:06:40 openmoko: * src/sms-contacts.c: (sms_contacts_page_new): Jan 22 15:06:40 openmoko: Store the MokoFingerScroll the treeview is placed in Jan 22 15:06:40 openmoko: * src/sms-notes.c: (forward_clicked_cb): Jan 22 15:06:42 openmoko: Show a contacts chooser for picking which contact to forward to Jan 22 15:06:44 openmoko: * src/sms-utils.c: (sms_contact_picker_dialog): Jan 22 15:13:10 freesmartphone.org: 03emdete * r53 10/trunk/software/py-proto/ (7 files): fixed minor bugs in init.d script and the sample Jan 22 15:31:00 openmoko: 03chris * r3919 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (ChangeLog configure.ac): Jan 22 15:31:00 openmoko: * configure.ac: Jan 22 15:31:00 openmoko: python bindings disabled by default Jan 22 15:33:01 openmoko: 03chris * r3920 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/libmokoui/moko-type.h: Add missing file Jan 22 16:27:04 openmoko: 03sameo * r3921 10/developers/sameo/patches/ar6k-atheros-2.0/2.6.24/ (atheros_2_0_function.patch atheros_2_0_sdio_stack.patch): Jan 22 16:27:04 openmoko: ar6k-atheros-2.0: 2.6.24 patches update. Jan 22 16:27:04 openmoko: This fixes the iperf bug on 2.6.24. Jan 22 16:29:46 openmoko: 03chris * r3922 10/trunk/src/target/OM-2007.2/libraries/libmokoui2/ (4 files in 2 dirs): (log message trimmed) Jan 22 16:29:46 openmoko: Patch by: Frank Li Jan 22 16:29:46 openmoko: * libmokoui/moko-finger-scroll.c: (moko_get_time_delta), Jan 22 16:29:46 openmoko: (moko_finger_scroll_button_press_cb), (moko_finger_scroll_refresh), Jan 22 16:29:46 openmoko: (moko_finger_scroll_timeout), Jan 22 16:29:48 openmoko: (moko_finger_scroll_motion_notify_cb), Jan 22 16:29:52 openmoko: (moko_finger_scroll_button_release_cb), (parent_set_cb), Jan 22 16:30:41 openmoko: 03sameo * r3923 10/developers/sameo/patches/ar6k-atheros-2.0/2.6.24/atheros_2_0_hcd.patch: ar6k-atheros-2.0: Forgot the HCD patch. Jan 22 16:38:09 openmoko: 03werner * r3924 10/branches/src/target/kernel/2.6.24.x/patches/ (atheros_2_0_hcd-patch atheros_2_0_hcd.patch series): Jan 22 16:38:09 openmoko: series, atheros_2_0_hcd-patch: renamed atheros_2_0_hcd-patch to Jan 22 16:38:09 openmoko: atheros_2_0_hcd.patch Jan 22 17:00:51 hi Jan 22 17:03:21 am I right, that the openmoko-kernel reallizes the console over the framebuffer by simply aktivating "framebuffer console" in die kernel-config-file? Jan 22 17:04:28 i should read what I am writing before hitting return ... but I'm sure everyone understands, what i mean =) Jan 22 17:06:28 little background: I use my own micro-kernel with a linux on top of this - the parameter "vga=0x301" is reaching my kernel (cat /proc/cmdline) but I actually could only see the console via serial interface Jan 22 17:10:10 openmoko: 03thomas * r3925 10/trunk/src/target/opkg/ (libopkg.h opkg_download.c): opkg: include the current url being downloaded in the progress callback Jan 22 17:33:32 openmoko: 03thomas * r3926 10/trunk/src/target/opkg/opkg_download.c: opkg: don't report the same download percentage multiple times Jan 22 17:42:34 openmoko: 03sameo * r3927 10/developers/sameo/patches/ar6k-atheros-2.0/2.6.24/atheros_2_0_function.patch: ar6k-atheros-2.0: We don't unregister our netdev if it doesn't exist. Jan 22 18:18:12 any gtk programming types about ? Jan 22 18:23:25 :d Jan 22 18:33:13 does g_timeout_add_seconds just do 1 call and then terminate, or is it supposed to recurr ? Jan 22 18:39:24 FuzzyCat: "The function is called repeatedly until it returns FALSE, at which point the timeout is automatically destroyed and the function will not be called again." Jan 22 18:39:27 http://library.gnome.org/devel/glib/stable/glib-The-Main-Event-Loop.html#g-timeout-add-seconds Jan 22 18:40:57 hmm... I wonder why libpcre-4.4 is the only/newest version in openembedded/openmoko... that's a version from over 4 years old... Jan 22 18:41:55 (and I have a program that has a dependency on libpcre >= 4.5) :) Jan 22 18:42:41 Mek: well, update OE :) Jan 22 18:42:57 Mek: to answer your question, pbly because nobody needed a newer version Jan 22 18:44:23 if I could figure out how to update it to at least 4.5 in my local copy I'd do that... but there seems to be a lot of magic stuff going on automaticaly in bitbake that I don't understand/can't find documentation for... Jan 22 18:52:43 hmm... pcre4.5 fails to compile because it doesn't seem to be defining EXPORT... Jan 22 18:53:11 Mek: the best (only?) documentation for bitbake/oe consists of other people's bitbake recipes. Go read your packages and includes directories. ;) Jan 22 18:57:56 ah, libpcre7.5 (the latest release) does compile cleanly... Jan 22 19:00:52 yo Jan 22 19:11:15 <_Krieger_> This page introduces you to the most efficient way to create new software for the OpenMoko platform. Note that there is a VMware image where this environment has been prebuilt for you. --- where is that image??? Jan 22 19:12:12 this is not finished yet Jan 22 19:12:18 i was a bit premature with editign this page Jan 22 19:13:02 woops :) Jan 22 19:20:01 i'd be happy to share my parallels image if thats of any help . Jan 22 19:20:24 cool, why not Jan 22 19:20:28 i'm getting a mac soon Jan 22 19:20:32 this week hopefully Jan 22 19:20:35 can you upload somewhere? Jan 22 19:33:40 mickeyl: yeah i'll do that .. Jan 22 19:39:17 cool, thanks Jan 22 20:04:23 /away Jan 22 20:04:34 rgr that Jan 22 20:04:44 ajbrun: actually back :-) Jan 22 20:04:55 you tricked me :( Jan 22 20:07:00 haakeyar__ : yeah, thanks - I just had a hot bath and thought about it in there... came out sorted the problem Jan 22 20:20:56 anyone know about a 256MB memory limit for qemu ARM? Jan 22 20:22:00 Stephmw: how do you mean? Jan 22 20:23:00 philippe: someone I know tried setting the ram value to 512 and it stops booting (this isn't OM-specific btw) Jan 22 20:23:38 Stephmw: never noticed that, but never did it either so it might be a problem Jan 22 20:23:58 philippe: don't suppose you have an image you could try that on quickly? Jan 22 20:25:25 the question is whether the real versatile/pb handles that much, maybe it's impossible anyway Jan 22 20:26:11 right, it's impossible Jan 22 20:26:28 the sysctl registers are at 0x10000000 PA in the ARM designs Jan 22 20:26:34 ahh Jan 22 20:27:32 thanks for checking Jan 22 20:27:50 you would have to map a second chip through some kind of flash controller or something, outside the RAM space and it would be slower (on real hw) Jan 22 20:37:17 Has anyone here tried to hook up a Neo to a stereo system by a 2.5mm-to-3.5mm (4-conductor-to-3-conductor) stereo cable? I get bad results. Jan 22 20:38:05 Loud scratching sounds in the stereo, and neod doesn't select the headphone-only output. Jan 22 20:38:33 Is the hardware expecting to see a microphone (or similar impedance) attached? Jan 22 20:38:46 i've only tried with headphones and a 2.5mm-to-3.5mm adapter, it sounded ok Jan 22 20:39:02 it was able to go much louder with the headphones than any other hardware i have at home Jan 22 20:39:14 freesmartphone.org: 03emdete * r54 10/trunk/software/py-proto/pygsmd.py: workaround for strange response on AT+CIMI added Jan 22 20:39:35 Headphones work OK for me too. Hooking up to something active, not so much. Jan 22 20:39:45 but i have not tried routing gsm calls to the headphones, only music from Linux Jan 22 20:40:27 (I have only tried the included headphones, though, not via adapter -- my 3.5mm end has the wrong gender for that) Jan 22 20:40:31 And yes, this is all via the media player, not calls. One thing at a time! Jan 22 21:02:41 openmoko: 03andrew * r3928 10/trunk/src/host/qemu-neo1973/ (10 files in 3 dirs): Sync from CVS. Jan 22 21:32:00 hi Jan 22 21:32:21 hi Jan 22 21:33:21 how can i rotate the screen orientation on a low level? i want to rotate the screen and set is to the resolution 320x240 Jan 22 21:33:39 i saw this in the wiki: chvt 4 && echo qvga-normal > /sys/devices/platform/s3c24xx-spi-gpio.1/spi0.0/state && fbset qvga Jan 22 21:34:02 but it doesnt rotate it. and i dont know how this could be achieved on a low level Jan 22 21:37:28 not possible with gta01 at least Jan 22 21:38:13 the framebuffer is always portrait, you need to rotate in software Jan 22 21:39:35 Or hardware Jan 22 21:39:52 (though that solution does reduce the available width) Jan 22 21:40:11 But seriously. Jan 22 21:40:16 eek, can you? Jan 22 21:40:41 yes, you have to do it in software, unless turning the phone 90 degrees from what you were hoping works. Jan 22 21:40:57 A nice optimised rotator would be so nice. Jan 22 21:41:04 *facepalm* Jan 22 21:41:09 annyway... Jan 22 21:41:11 optimised for cache and stuff Jan 22 21:41:55 is there no way to redirect the frame buffer to landscape mode? Jan 22 21:41:59 no Jan 22 21:42:18 sorry, I was being silly. You can rotate the phone Jan 22 21:42:24 (physically) Jan 22 21:42:40 But you can't make it look like a 320x240 display Jan 22 21:42:46 rather than a 240x230 Jan 22 21:42:54 (or the equivalent VGA)O Jan 22 21:43:03 xrandr! Jan 22 21:43:20 I'm unsure xrandr is very optimised Jan 22 21:43:45 some trivial tests I did and basic sums seemed to indicate to me that it should get better performance than it was. Jan 22 21:44:11 This could of course be my limited understanding rather than lack of optimisation. Jan 22 21:44:59 when i use xrandr i get the error: http://pastebin.org/16556 any hints? Jan 22 22:10:20 * mjr recodes prerotated test material for qvga video playing Jan 22 22:11:01 :) Jan 22 22:43:30 fookin' ay Jan 22 22:44:11 got decent full-screen (as much as is possible aspect-wise) video playing juuust fine Jan 22 22:44:34 yep it works just fine with gta01 + mpeg4 Jan 22 22:44:55 mjr, did you use my mencoder settings I posted in the wiki? Jan 22 22:45:41 actually I cooked some up myself, probably not too optimal (and mpeg1 only), might see about that Jan 22 22:46:43 Bleh, and I can't even get mp3 to play without skipping. Jan 22 22:47:02 with the gui mediaplayer? Jan 22 22:47:11 !@#% PulseAudio Jan 22 22:47:22 yes, pulseaudio hogs cpu like a hog Jan 22 22:47:27 Yeah. Jan 22 22:47:46 it's no bug - it's a feature! Jan 22 22:47:51 I like pulse on the desktop, but on this machine . . . not so much. Jan 22 22:49:48 mjr, try this: mencoder input.avi -ovc lavc -lavcopts vcodec=mpeg4:vhq:vbitrate=300 -vf crop=xx:xx:xx:xx,rotate=2,scale=-2:320 -oac mp3lame -lameopts br=64:cbr -o output2.avi Jan 22 22:50:00 yeah, I got that from the wiki and running now Jan 22 22:50:02 I watched 300 this way on my neo just fine :-D Jan 22 22:50:59 now one would like an mplayer ui with as much kludging as needed... Jan 22 22:51:37 isn't the new mediaplayer by abraxa_ mplayer based? Jan 22 22:51:40 Nice. Snarfed. Jan 22 22:55:09 how is playback performance on the neo? Jan 22 22:55:20 not _too_ bad Jan 22 22:55:34 Basically on a very early unhacked mplayer. Jan 22 22:55:37 coptix, with qvga - prerotated it's just fine Jan 22 22:55:54 It gets 20+FPS on your average 150K/s MPg Jan 22 22:56:41 Yeah - prerotated, with the appropriate codec, it's quite possible. Jan 22 22:57:03 * josch was able to watch 300kBit/sec mpeg4 prerotated qvga with full fps Jan 22 22:57:17 I mean kbytes/s Jan 22 22:57:21 ah kk Jan 22 22:57:33 teh average 'VCD' quality stuff played OK without any hacking at all Jan 22 22:57:39 dropping the occasional frame. Jan 22 22:57:52 sounds good - so freerunner should have no problem? Jan 22 22:58:11 coptix, at least less problems ^^ Jan 22 22:58:17 That's a seperate issue. Jan 22 22:58:33 The GPU on freerunner has a very slow bus. Jan 22 22:58:40 (30 megabytes/second) Jan 22 22:58:55 So if you're pushing raw pixels to it, that may be a limiting factor. Jan 22 22:59:21 However, at QVGA that isn't so much of a problem Jan 22 22:59:29 not with qvga and nobody needs more for video Jan 22 22:59:31 The GPU can do hardware rotate Jan 22 22:59:55 SpeedEvil, does mplayer support this feature?? Jan 22 23:00:01 probably not Jan 22 23:00:03 And the GPU can do IDCT and stuff that higher level video codecs may be able to use to output partially cooked output to it. Jan 22 23:00:11 support what feature? Jan 22 23:00:14 rotation Jan 22 23:00:17 nice Jan 22 23:00:25 the om media player is based on gstreamer, isn't it? Jan 22 23:00:41 quicksand: yes Jan 22 23:00:57 josch: the mplayer I was referring to was a _very_ early one simply running on GTA01s raw framebuffer unrotated Jan 22 23:01:28 SpeedEvil, never used mplayer on my neo in an other way than unrotated framebuffer Jan 22 23:01:30 I should try installing mplayer and playing mp3s that way. It's a little less convenient to operate from the command line while driving, though. Jan 22 23:02:25 mplayer - with the right codec (I forget) plays mp3 at about 25% load Jan 22 23:02:36 128K Jan 22 23:02:47 one should just try them all out then Jan 22 23:02:56 It easily manages with CPU clock at 100Mhz Jan 22 23:03:10 problem is it may bve a special codec, I forget. Jan 22 23:04:07 since mp3 in videos seems to be no problem for mplayer without any codec hacking mp3 without video kindof /has/ to work without lagging Jan 22 23:04:43 In the earlier example I disabled audio - it wouldn't _quite_ manage smooth audio with Jan 22 23:05:09 had no problems with mplayer + mpeg4 + mp3 Jan 22 23:05:46 This was simply a 'will the assorted media files I have play' test - not specifically created ones with sensible bitrates. Jan 22 23:06:10 Some tests on that indicated you can go really quite low in terms of bitrate. Jan 22 23:06:15 On a 2.8" screen. Jan 22 23:06:37 I found that 160*120 was in most cases quite acceptable. Jan 22 23:06:45 agreed Jan 22 23:07:27 the screen is too tiny as if anything more would make much sense as you are not able to see the difference Jan 22 23:07:49 Oh, I'm sure that's true. I'll tolerate quite a bit of picture grunge if I'm just trying to watch a TV show on an airplane. Jan 22 23:08:17 and the encode finishes quicker with lower resolution ;) Jan 22 23:08:34 I have done on-neo transcoding. Jan 22 23:08:37 More will fit on my micro-SD that way too! Jan 22 23:08:39 It's not actually _that_ bad. Jan 22 23:08:40 quicksand, just be suprised that there is no noticeable difference between qvga and four times the resolution (vga) Jan 22 23:09:09 But is of course not realtime Jan 22 23:09:17 regarding microSD - whats the largest card that's been reported to work with the neo? Jan 22 23:09:18 josch: you upped the bitrate too? Jan 22 23:09:26 CoreDump|afk: 6G I think Jan 22 23:09:29 cop: Jan 22 23:09:52 Someone mentioned 8G - I don't know that that was actually attached though. Jan 22 23:09:56 maito, it's quite hard to maintain the same bitrate if you quadrupel the resolution Jan 22 23:10:04 ah ok Jan 22 23:10:20 (In principle if 6 works, 8 should) Jan 22 23:10:22 SpeedEvil, I own an 8G microsd Jan 22 23:10:23 josch: Depends on the source. I would posit that most VGA video came originally from 480i (or 525i) sources and hence is missing true VGA detail anyway. Jan 22 23:10:31 josch: of course you need to up the bitrate too when you up the resolution. maybe big time Jan 22 23:10:33 I have an 8G on the way, and I'll report back with results. Jan 22 23:10:37 SpeedEvil, sandisk and works fine as my rootfs Jan 22 23:10:42 :) Jan 22 23:10:45 josch: speed? Jan 22 23:10:46 quicksand, sandisk? Jan 22 23:10:57 Yes. Jan 22 23:11:01 thats good to hear Jan 22 23:11:04 josch: meg/second read and write in the neo that is. Jan 22 23:11:09 quicksand, so it will work :-) Jan 22 23:11:24 SpeedEvil, just tell my how to benchmark it for you Jan 22 23:11:35 Awesome. I had assumed so but it's nice to get confirmation. Jan 22 23:12:02 time dd if=/dev/zero of=/media/card/testfile bs=1024 count=102400 Jan 22 23:12:15 will give write speed for 100M of zeros Jan 22 23:12:17 how do i set the neo to qvga to get it working with mplayer? Jan 22 23:12:55 shred, http://wiki.openmoko.org/wiki/Video_Player Jan 22 23:15:23 josch: then i get a black screen with "state_write: Switch to state:3 jbt6k74_enter_state: entering(old_state=normal_to_sleep: entering ..............) Jan 22 23:16:32 SpeedEvil, 1m 17.76s Jan 22 23:16:48 actually i can start mplayer from ssh. but is there a way to preserve the gui at the lower resolution Jan 22 23:16:53 josch: ? Jan 22 23:17:06 SpeedEvil, the time to write to the sdcard Jan 22 23:17:08 shred: not without starting another X I think Jan 22 23:17:13 josch: ah Jan 22 23:17:28 josch: so that looks like on the order of 800k/second write Jan 22 23:18:39 SpeedEvil, is there a comparison of read/write speeds with several card types on the neo? Jan 22 23:18:53 Not that I'm aware of. Jan 22 23:19:42 I'm pretty sure that tehre was a _significant_ regression in speed as the fix for the write-hang went in Jan 22 23:20:00 SpeedEvil: isn't it more? Jan 22 23:20:32 Vegar: no - less Jan 22 23:20:36 quicksand, i managed to get mp3s working without any gaps through the openmoko media player as soon as I simply reniced pulseaudio to -10. I wonder why that's not the default. Jan 22 23:20:45 Vegar: I'm sure I recall speeds of 2M+/second Jan 22 23:21:01 * * OM Bug 1193 has been created by mjr(AT)iki.fi Jan 22 23:21:01 * * Severe performance problems with Pulseaudio Jan 22 23:21:02 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1193 Jan 22 23:21:10 Wow - that was quick. Jan 22 23:21:21 :-) Jan 22 23:21:26 My kingston 2G card gives 1m 9.02s -- same order of magnitude. Jan 22 23:21:28 well Jan 22 23:21:36 that should rather be in the pulseaudio bugtracker Jan 22 23:21:41 My kingston 512M card gives... Jan 22 23:21:45 there's something wrong Jan 22 23:21:45 * SpeedEvil taps foot Jan 22 23:21:48 PA should not behave so badly Jan 22 23:22:03 unfortunately 0.9.8 doesn't work w/ gstreamer Jan 22 23:22:23 as i said, it works ok for me, if it has enough priority :) Jan 22 23:22:27 real 1m 59.35s Jan 22 23:22:33 toxx_: I'll try that, and maybe put the nice-ness in the launcher if it works Jan 22 23:22:43 I think it's kingston anyway Jan 22 23:22:48 SpeedEvil: empty card? Jan 22 23:23:05 no - ex2fs - it shouldn't matter that much Jan 22 23:23:10 ext2 or fat? Jan 22 23:23:22 * josch also ext2 Jan 22 23:23:24 toxx_: well it's burning a lot of cycles somewhere Jan 22 23:23:38 + gstreamer not making it better Jan 22 23:24:14 mickeyl, ok, I agree this is rather a workaround than a fix. i just thought it would be nice to share my findings as i was also disappointed by my neos music performace. Jan 22 23:25:01 My 2G card has two partitions -- I get 1:09.02 for the fat and 0:56.53 for the ext2 Jan 22 23:25:14 what about read? Jan 22 23:26:08 This is odd - successive reads, 22, 35,47s Jan 22 23:26:43 * SpeedEvil winds it up to 200M to ensure it's not cached. Jan 22 23:26:54 reading 100M: time dd if=/dev/mmcblk0p1 of=/dev/null bs=1024 count=102400 Jan 22 23:27:04 time cat testfile > /dev/null gives me 40.37 (ext2, first try) Jan 22 23:27:08 0m 37.93s Jan 22 23:27:47 we definately should document this, shouldn't we? Jan 22 23:27:51 42.96 (ext2, second try) Jan 22 23:28:05 IRC log! Jan 22 23:28:11 haha Jan 22 23:32:00 On the fat partition I get 40.32 and 39.16 -- pretty consistent Jan 22 23:34:35 GTG. Somebody wanna update the wiki? ;) Jan 22 23:34:45 I'm doing a trivial test script Jan 22 23:34:54 SpeedEvil, great Jan 22 23:35:26 but shouldn't benchmarks be done without a file system in between? like writing plainly onto the device? Jan 22 23:35:35 In theory yes. Jan 22 23:35:41 If you post it somewhere I'll run it when I'm back. Jan 22 23:35:51 In practice, a whole lot more people will run it if they don't need to nuke all their files. Jan 22 23:35:56 haha Jan 22 23:37:18 Benchmarking writes with 'dd' may not be accurate. You want something that will do an fsync() at the end to flush any data that's cached in RAM. Read testing should also flush caches before each measurement. Jan 22 23:37:40 yes - script addresses than in a most complete and accurate way. Jan 22 23:37:46 It uses 128M test file Jan 22 23:43:14 * cesarb prepares for one more crazy test - what will happen if I use the "conservative" governor? and "ondemand"? Jan 22 23:43:33 you got it running, cesarb? Jan 22 23:44:19 Vegar: I did test it already a bit on the real hardware, but with the "userspace" governor Jan 22 23:44:31 Vegar: see http://wiki.openmoko.org/wiki/User:CesarB/cpufreq Jan 22 23:45:06 * cesarb hopes the battery doesn't die during the tests Jan 22 23:45:15 this indicator is ridiculously innacurate Jan 22 23:45:48 55%=0 Jan 22 23:45:59 Oops - I need to get that patch found, compiled and off Jan 22 23:46:01 Ok, i just checked out the svn is there a way to compile everything at once? Jan 22 23:46:44 On my 2G Kingston card I get a write rate of 2.4 MiB/s and a read rate of 3.2 MiB/s. The card is formatted with ext3 and mounted async,noatime. Jan 22 23:47:00 kernel? Jan 22 23:47:06 * cesarb should be at 66% which would be 22% then Jan 22 23:48:14 my kernel? 2.6.22.5-moko11+svnr3874-r13 Jan 22 23:48:36 uname -a Jan 22 23:48:36 Linux fic-gta01 2.6.22.5-moko11 #1 PREEMPT Wed Dec 19 20:33:49 GMT 2007 armv4tl unknown Jan 22 23:50:10 my 'uname -a' is "Linux fic-gta01 2.6.22.5-moko11 #1 PREEMPT Thu Jan 17 21:30:24 PST 2008 armv4tl unknown", but it's not very useful as it doesn't show the svn version Jan 22 23:52:01 ok - /proc/version not uname -a Jan 22 23:52:18 cool, "conservative" works fine (since I didn't make the anti-flicker patch for 2.6.22.5, I could see by the flicker) Jan 22 23:52:39 cat /proc/version: Linux version 2.6.22.5-moko11 (mmontour@mythtv) (gcc version 4.1.2) #1 PREEMPT Thu Jan 17 21:30:24 PST 2008 Jan 22 23:52:47 OTOH, ondemand worked for a bit and then got stuck at what I can by the flicker pattern identify as the minimum, 12000 (I comented out anything lower than that for this test) Jan 22 23:52:50 cesarb: is this a work in terms of 'saves lots of power', or 'not explode'? Jan 22 23:53:45 SpeedEvil: a works in terms of "selects a somewhat correct speed, which I can see by the flicker pattern and by cating scaling_cur_speed" Jan 22 23:53:54 :) Jan 22 23:54:12 does setfb help with reducing flicker? Jan 22 23:54:14 is it possible that 12MHz is too slow for ondemand? Jan 22 23:54:33 SpeedEvil: where's the CPU? the battery seems cold to the touch, but I want to check the part of the case where the CPU is Jan 22 23:54:43 Umm Jan 22 23:54:51 under the battery - ... Jan 22 23:54:56 SpeedEvil: cool, even it being stuck at 12000, I can still do an orderly shutdown with ondemand Jan 22 23:54:58 bah... the version of libxine in openembedded is also quite outdated it seems... Jan 22 23:55:04 SpeedEvil: hah, no way I can check then :D Jan 22 23:56:38 why are all packages in openembedded > 2 years old :( Jan 22 23:56:42 SpeedEvil: below the battery is cold to the touch too, perhaps I didn't test for long enough... Jan 22 23:57:24 I think 'under the battery' is about as good as you can get. Jan 22 23:57:35 without opening the can Jan 22 23:59:01 * SpeedEvil finds the reason for his odd benchmark results. Jan 22 23:59:12 * SpeedEvil wonders why gsmd is using 99% CPU Jan 22 23:59:23 SpeedEvil: it's calling home Jan 22 23:59:30 sigreturn() = ? (mask now [ILL SEGV ALRM TERM STKFLT CONT STOP TSTP TTIN TTOU URG XFSZ VTALRM PROF SYS]) Jan 22 23:59:30 --- SIGSEGV (Segmentation fault) @ 0 (0) --- Jan 22 23:59:32 ... Jan 23 00:01:25 ok, added to http://wiki.openmoko.org/wiki/User:CesarB/cpufreq#Testing_report (this seems to slowly be becoming a blog... :D ) Jan 23 00:02:02 I notice that the speed of the blogging speeds up and slows down in an irregular manner :) Jan 23 00:03:02 Mek: that's an exaggaration Jan 23 00:03:07 but if so, then... stability? :D Jan 23 00:03:09 patches welcome Jan 23 00:03:55 mickeyl: the version of libxine in openembedded is from july 2005, the version of libpcre was even older... okay, probably some packages are maintained/up-to-date... :) Jan 23 00:04:05 well Jan 23 00:04:08 libxine on arm is a nightmare Jan 23 00:04:39 look at all the patches we need Jan 23 00:04:39 i guess we just gave up on that Jan 23 00:04:39 that's why no one updated it Jan 23 00:04:45 mickeyl: not to mention volatile packages like tzdata, which can rarely ever cause breakage (but it not being up-to-date is a sort of breakage) Jan 23 00:04:59 well Jan 23 00:05:02 count the number of packages Jan 23 00:05:06 count the number of people Jan 23 00:05:10 okay, libxine might indeed be an exception than :) Jan 23 00:05:12 then you come to a conclusion that's pretty insane Jan 23 00:05:48 OE is not Ubuntu, OE has ~20 people working on 5000 recipes Jan 23 00:05:51 divide that Jan 23 00:05:52 and have fun Jan 23 00:06:12 I just don't want to learn yet another version-control-system, fighting bitbake with no docs, and more stuff like that... Jan 23 00:09:59 well Jan 23 00:10:01 bitbake has docs Jan 23 00:10:07 and you don't need to learn monotone to submit patches Jan 23 00:10:19 if you don't want to contribute, ok Jan 23 00:10:23 but please don't try to find excuses Jan 23 00:10:44 I'm just saying it could have been made easier for someone to contribute, not trying to find excuses Jan 23 00:11:02 making this easier is a lot of work in itself Jan 23 00:11:06 s/this/things/ Jan 23 00:11:07 mickeyl meant: making things easier is a lot of work in itself Jan 23 00:11:21 a full fledged distribution builder that flexible is just not an easy thing Jan 23 00:11:24 it's a complex thing Jan 23 00:11:34 inherently Jan 23 00:17:27 hm, is gsmd supposed to work in a compile from late last week? Jan 23 00:17:37 s/in/on/ Jan 23 00:17:37 cesarb meant: hm, is gsmd supposed to work on a compile from late last week? Jan 23 00:18:16 * cesarb doesn't know if gsmd died due to his cpufreq stuff or because it just dies a lot Jan 23 00:20:41 Quick poll - does awk -F\" '/\+CNUM:/{print $(NF-1);exit}' /var/volatile/tmp/gsm.log Jan 23 00:20:45 print your phone number? Jan 23 00:21:01 Or am I misusing something that won't generally work Jan 23 00:21:17 SpeedEvil: I am not using a SIM (or even a SD card) on these tests for now ;-) Jan 23 00:22:58 (mostly because I only have one SIM... sure, I could borrow one of my relative's, only that it's for a 850 network :D ) Jan 23 00:24:04 anyone have scaling issues with xoo; it is huge Jan 23 00:29:06 There is no way to read - simply - the manufacturer info of a SD card? Jan 23 00:30:17 dunno Jan 23 00:35:54 openmoko: 03werner * r3929 10/branches/src/target/kernel/2.6.24.x/patches/ (3 files): Jan 23 00:35:54 openmoko: We have to be more careful about pending SDIO interrupts from the host side. Jan 23 00:35:54 openmoko: This patch fixes the iperf bug Nod has been reporting. Jan 23 00:35:54 openmoko: Signed-off-by: Samuel Ortiz Jan 23 00:35:54 openmoko: [ And does some general housekeeping as well. - Werner ] Jan 23 00:40:12 great, after testing on several speeds, the modem fails to answer precisely at 12000 Jan 23 00:42:52 openmoko: 03werner * r3930 10/branches/src/target/kernel/2.6.24.x/patches/atheros_2_0_function.patch: Jan 23 00:42:52 openmoko: Whenever our ar6k chip can not be identified, we are trying to unregister Jan 23 00:42:52 openmoko: our netdevice although it was never registered. Jan 23 00:42:52 openmoko: This patch fixes this bug. Jan 23 00:42:52 openmoko: Signed-off-by: Samuel Ortiz Jan 23 00:53:15 Sigh. Jan 23 00:53:34 I just rebooted my neo three times, trying to work out why the screen wasn't working. Jan 23 00:53:45 I have it on my desk. Jan 23 00:54:04 Screen is much more visible when it's facing upwards. Jan 23 00:54:13 lol Jan 23 00:54:32 I was tapping the back, and wondering why the sound wasn't working. Jan 23 00:54:36 openmoko: 03werner * r3931 10/branches/src/target/kernel/2.6.24.x/patches/gta01-vibrator.patch: Jan 23 00:54:36 openmoko: We have to be more careful about pending SDIO interrupts from the host Jan 23 00:54:36 openmoko: side. This patch fixes the iperf bug Nod has been reporting. Jan 23 00:54:36 openmoko: Signed-off-by: Samuel Ortiz Jan 23 00:54:36 openmoko: [ And does some general housekeeping as well. - Werner ] Jan 23 00:54:39 openmoko: gta01-vibrator.patch: global renaming of "gta01" to "neo1973", except for Jan 23 00:54:41 openmoko: asm/arch/gta01.h and machine_is_neo1973_gta01 Jan 23 00:56:05 * SpeedEvil wonders why <6>s3c2410-sdi s3c2410-sdi: running at 16625kHz (requested: 25000kHz). Jan 23 00:56:12 not PCLK/4 Jan 23 00:56:24 (22.16Mhz) Jan 23 01:01:45 openmoko: 03werner * r3932 10/branches/src/target/kernel/2.6.24.x/patches/gta01-vibrator.patch: (log message trimmed) Jan 23 01:01:45 openmoko: vib-trashes-resume-fix.patch: Jan 23 01:01:45 openmoko: The resume stuff for vibrator was broken --> OOPS... I guess it was never Jan 23 01:01:45 openmoko: tested on GTA02 yet considering it did not have an active device associated Jan 23 01:01:45 openmoko: with the driver until today. Fixed. Jan 23 01:01:49 openmoko: Signed-off-by: Andy Green Jan 23 01:01:51 openmoko: gta01-vibrator.patch: Jan 23 01:40:32 WTF... 'After 10 minutes at 450C, one of the SIM cards remained readable'. Jan 23 01:40:56 (of recovering data from SIM cards) Jan 23 01:41:23 450*C ? Jan 23 01:41:37 If that's the phone+SIM - I can sort-of-see it, as the phone will melt and shield it to a degree Jan 23 01:41:49 If it's the bare SIM I don't understand. Jan 23 01:41:49 450 celsius degrees? Jan 23 01:41:53 yes Jan 23 01:42:33 useful Jan 23 01:42:40 http://ukpress.google.com/article/ALeqM5iJFbCLMvdDiXnVe0FXS9sKVU75cA Jan 23 01:42:56 slightly OT - but how can I revoke a gpg key with the revoke certificate? Jan 23 01:43:26 Using acid, the researchers gently stripped the protective epoxy moulding surrounding each chip in order to expose its circuitry. Jan 23 01:43:26 The cards were then placed in heated air for 10 minutes, were allowed to cool and the researchers then sought to recover data by attaching tiny probes to the circuit and reading its contents via an interface pad. Jan 23 01:43:27 Six cards were heated to around 180 C (356 F) and could be read after rewiring with no loss of data. Jan 23 01:43:32 but SpeedEvil Jan 23 01:43:57 it has metal too? just a bit.. Jan 23 01:44:01 Five were cooked to 450 C (842 F). Four of them could not be read by the researchers, but the fifth one could, briefly. Jan 23 01:44:23 So this is 'what can the chip take' - not 'what's practical with a whole card' Jan 23 01:44:38 The limiting case in otherwords Jan 23 02:49:57 Crap. The IDE drive in my IBM X40 just died. Jan 23 02:50:17 IBM-- # weird, non-stadard drive form factor Jan 23 02:50:21 well, it Jan 23 02:50:38 1.8"? Jan 23 02:50:41 not a standard 2.5"? Jan 23 02:50:43 shit Jan 23 02:50:44 Yep. Jan 23 02:51:03 I guess it is a standard, but he x40 is the only laptop I know of that uses it. Jan 23 02:51:41 almost all the 1.8" drives use a ZIF connector, but this one uses a standard IDE pinout Jan 23 02:53:57 nah Jan 23 02:54:02 oooh Jan 23 02:54:08 1.8" but different connector Jan 23 02:54:08 aaah Jan 23 02:55:00 Only the (expensive) Hitachi drives will work. **** ENDING LOGGING AT Wed Jan 23 02:59:57 2008