**** BEGIN LOGGING AT Mon Sep 24 02:59:59 2012 Sep 24 04:40:14 heh Sep 24 04:40:25 maybe Sep 24 05:50:05 no Sep 24 05:50:12 no more n950 Sep 24 05:51:18 DocScrutinizer05: why not, as if it is usefull :D Sep 24 05:52:04 DocScrutinizer05: I am thinking to try to upstream thumb patch, any idea where to look for the process? Sep 24 06:05:29 Anyone seen a thread about running Debian natively on N900? I can't find it Sep 24 06:14:31 int_ua: it's not worth it, anyway Sep 24 06:22:47 kerio: Why? :) Sep 24 06:23:12 int_ua: because half of the hardware doesn't work Sep 24 06:59:56 kerio: what hardware doesn't work? Sep 24 07:00:21 Sicelo: phone calls Sep 24 07:00:32 and last time i heard, the gps was still kinda meh Sep 24 07:01:10 * entitled has never made a single phone call with N900 Sep 24 07:01:37 purely for ssh. browsing and such Sep 24 07:02:33 int_ua: https://elektranox.org/N900 , http://elinux.org/N900 & look at NIN101's pages .. i forgot the link .. but same place RescueOS lives Sep 24 07:05:01 which reminds me.. i must find out how to switch between 2G-3G using ofono for when i'm on debian Sep 24 07:07:28 Sicelo: yeah, but no thread on Talks? Sep 24 07:07:52 Sicelo: so you have it running natively? Sep 24 07:08:00 there should be, but TMO Sep 24 07:08:06 yes, i do :) Sep 24 07:10:13 as kerio said, phne calls don't work (well), but you can use the modem for internet, which covers my needs. Sep 24 07:10:25 Sicelo: which DE are you using? Sep 24 07:10:56 i used E7 at first, but am using Xfce now Sep 24 07:11:06 why not hildon? :P Sep 24 07:11:58 i haven't seen it packaged for debian, yet. otherwise i would love that, or even plain old matchbox Sep 24 07:12:24 why not openbox? Sep 24 07:13:51 i'll try that Sep 24 07:14:10 what are you using to right click, btw? Sep 24 07:14:20 nothing, lol Sep 24 07:14:36 then you're not going to like openbox :3 Sep 24 07:15:10 yeah, openbox needs rightclick. Sep 24 07:15:18 but there is a workaround Sep 24 07:15:40 Sicelo: something with xorg.conf hacking Sep 24 07:17:01 thanks. 'll have to google that. fact is, i have nfc about some deep things Sep 24 08:35:35 deepy: btw, i just realized that without fixing the kbd slider, you can't use portrait mode Sep 24 08:35:36 ever Sep 24 08:35:52 ooo Sep 24 08:35:59 I didn't know there was a portrait mode Sep 24 08:36:12 Time to get a soldering iron, eh? Sep 24 08:36:37 note that it's mostly useless Sep 24 08:36:50 imo it's only useful with openmediaplayer Sep 24 08:36:55 to choose a song with one hand Sep 24 08:45:44 oooh, that reminds me Sep 24 08:45:55 I gotta find some SFW show so I can test the battery life and the stan Sep 24 08:45:57 at the same time Sep 24 09:01:54 deepy: fun fact: myfreecams.com works on microB Sep 24 09:02:02 and that's a website that makes netbooks choke Sep 24 09:03:00 microB? Sep 24 09:03:08 also it sounds incredibly fishy :P Sep 24 09:03:25 microB is "Web" Sep 24 09:03:27 the default browser Sep 24 09:03:30 Ah Sep 24 09:03:36 nah, it's legit Sep 24 09:03:37 The website still sounds fishy Sep 24 09:03:48 it's not exactly sfw, though Sep 24 09:04:47 Ever since the copypaste incident of '11 I'm not opening websites that involve the word 'cam' when I'm not at home Sep 24 09:05:01 oh, job Sep 24 09:05:02 pf Sep 24 09:05:04 pffffffffff Sep 24 09:05:13 pffff all you want Sep 24 09:05:16 I'll pffffff harder Sep 24 09:05:36 OSHI- Sep 24 09:05:39 Sent a PC to the wrong country Sep 24 09:05:45 * deepy whistles Sep 24 09:06:02 deepy: could you wrongfully send a PC to my house? Sep 24 09:06:23 kerio: you can already virtualize ten of these Sep 24 09:06:44 well, i do need a home server... Sep 24 09:07:16 what do you intend to do on that home server? Sep 24 09:07:22 I know of $19/yr VPS Sep 24 09:07:50 i already have one, but it's kinda bad Sep 24 09:07:56 file storage and torrents Sep 24 09:08:06 and DNS tunneling Sep 24 09:08:14 DNS tunneling is so 80s Sep 24 09:08:29 u jelly Sep 24 09:08:41 I'm at work chatting to you from a server in Sweden Sep 24 09:08:43 ;-) Sep 24 09:08:52 One of us is getting paid during this conversation Sep 24 09:16:22 I just sent two monitors to a girl Sep 24 09:16:32 two WIDESCREEN monitors Sep 24 09:16:40 I'm sitting here with a flickering 19" Sep 24 09:17:03 Not fair :-( Sep 24 09:17:26 sluta slöa! börja jobba! Sep 24 09:18:01 I am working! Sep 24 09:18:22 sure! ;) Sep 24 09:18:24 Just sent a battery to someone Sep 24 09:18:44 sweet, to me? i need one to my oqo01! Sep 24 09:18:51 Yes Sep 24 09:19:08 It's a pity it'll only fit x200s/x201 Sep 24 09:19:18 But hey, it's a 9cell Li-Ion! Sep 24 09:19:27 hehe Sep 24 09:20:15 im cleaning my bellybutton :D Sep 24 09:20:59 deepy: send us technology Sep 24 09:21:05 for free Sep 24 09:21:06 kerio: I can get you a job here Sep 24 09:21:10 for pay Sep 24 09:21:15 i'd have to work Sep 24 09:21:18 that sucks Sep 24 09:21:24 Does it look like I am working? Sep 24 09:21:38 I played nethack half of my time here Sep 24 09:30:17 play FTL! Sep 24 09:30:53 I don't know what that is Sep 24 09:31:00 a cute game Sep 24 09:31:00 I heard it's a roguelike in space Sep 24 09:31:10 it's a roguelike-like Sep 24 09:31:49 permadeath, randomly generated map and encounters, fairly constant challenge throughout the game Sep 24 09:33:27 It seems like I may enjoy dwarf fortress more Sep 24 09:49:25 static.maemo.org is down Sep 24 09:49:40 so wiki.maemo.org not working... Sep 24 10:15:55 worksforme Sep 24 10:16:20 Pali: how's the bme replacement project going? Sep 24 10:16:46 kerio, hald-addon-bme is implemented and tested by me - forking fine Sep 24 10:17:05 neat Sep 24 10:17:06 kernel modules bq2415x_charger, bq27x00_battery and rx51_battery working fine too Sep 24 10:17:07 what does it use? Sep 24 10:17:26 libbmeeipc is implemented, but not tested yet Sep 24 10:17:50 bme-rx51 (meta)(pseudo)package is written too, but not tested Sep 24 10:18:02 hald-addon-bme is used by more programs (also 3rd party) Sep 24 10:18:16 it export battery & charger properties to HAL Sep 24 10:18:39 for example default nokia battery status menu plugin Sep 24 10:18:51 all 3rd battery widgets, ... Sep 24 10:19:16 how is it packaged? Sep 24 10:19:26 in deb package Sep 24 10:20:06 each package has same name, same library/binary as original nokia Sep 24 10:20:38 did you make a fake /usr/sbin/bme_RX-51? :O Sep 24 10:20:51 libraries export same function names & binaries export needed interface Sep 24 10:21:09 /usr/sbin/bme_RX-51 was removed, not needed Sep 24 10:21:27 no application/library using it Sep 24 10:21:29 i'm still not convinced that your "bme-rx-51" should be called like that Sep 24 10:22:03 original bme-rx-51 has 2 system files: 1) event.d script 2) bme binary Sep 24 10:22:17 does the rest of the stuff fail gracefully if the modules aren't loaded? Sep 24 10:22:22 kinda like if bme wasn't running Sep 24 10:22:50 original event.d script start charging (spawn bme and wait for soemthing...) Sep 24 10:23:05 new event.d script start charging too (it load kernel modules) Sep 24 10:23:36 Pali: does it touch /etc/bme-disabled or whatever usbmode.sh checks? Sep 24 10:24:03 no, usbmode.sh was changed to check if new version of bme-rx-51 is installed Sep 24 10:24:25 * kerio groans Sep 24 10:24:39 every old nokia bme package has version << 1.0 Sep 24 10:24:56 and every bme replacement version is >= 1.0 Sep 24 10:25:55 (if somebody for some unknown stupid reason want to increase version of bme packages - he can - but not higer than 1.0) Sep 24 10:36:25 kerio, you can look at project page: https://gitorious.org/rx51-bme-replacement Sep 24 10:54:34 kerio: guess what? sbc decoding and encoding binaries actually exists for C64x dsp. Sep 24 10:55:24 what's sbc? Sep 24 10:59:31 Bluetooth low complexity, sub band codec Sep 24 11:01:43 the a2dp talk from yesterday, remember? Sep 24 11:07:28 oh right Sep 24 11:07:39 well, if sbc is done by the bt chip... Sep 24 11:10:03 Morning Sep 24 11:10:31 morning! Sep 24 11:10:42 moin Sep 24 11:11:41 Jaffa: sup Sep 24 11:12:01 Jaffa: i like to read your sad morning tweets :P Sep 24 11:12:23 Skry, do you know about aptx bluetooth codec? Sep 24 11:12:40 Pali: almost nothing supports it :( Sep 24 11:13:02 is there any specification? Sep 24 11:13:38 I see that my bluetooth headset has support for aptx Sep 24 11:13:57 but I did not find anything about aptx codec... Sep 24 11:16:34 Pali: not much. Sep 24 11:17:15 i bet it's something not open Sep 24 11:17:38 there are products that use c64x dsp for the encoding but it's very proprietary stuff Sep 24 11:17:55 yeah, not open at all Sep 24 11:18:41 so there is no way to encode aptx on n900? Sep 24 11:18:56 not that I know about it Sep 24 11:18:59 or are there any ti dsp binaries for that? Sep 24 11:19:07 not from ti at least Sep 24 11:20:13 Pali: you could try to ask them for some info if there is binary for c64x http://www.dr-embedded.com/servlet/publicSite.displayIP?ipId=SW_MW_1200 Sep 24 11:20:34 and you have the skills for userspace parts Sep 24 11:21:07 what c64x mean? Sep 24 11:21:10 ~c64x Sep 24 11:21:46 n900 uses TMS320C64x dsp/isp Sep 24 11:24:37 c64x+ to be more exact Sep 24 11:28:27 Skry: do re really need DSP codec node for SBC? Sep 24 11:28:46 bluetoothd has nice NEON optimized codec Sep 24 11:29:28 freemangordon, how is dsp binary codec connected to system? Sep 24 11:29:37 gst-dsp Sep 24 11:30:02 gstreamer that is. though TI-OMX could be used too Sep 24 11:30:05 gst-dsp has special code for each codec? Sep 24 11:30:09 yes Sep 24 11:30:25 because each codec has different set of parameters :) Sep 24 11:30:42 year ago I saw project for theora support Sep 24 11:30:51 they created dsp codec binary Sep 24 11:30:59 hmm, IIRC it was DSP side Sep 24 11:31:09 freemangordon: no i dont need one, we just had this talk about it yesterday Sep 24 11:31:41 I do not know state of that project, but will be possible to use it? Sep 24 11:31:45 Pali: I tried to search for that one, disappeared :\ Sep 24 11:32:30 Pali: it could be usefull, but is there any problem with MPU side theora codec? Sep 24 11:33:19 all this talking reminds me that it is time to return to gst-dsp :) Sep 24 11:33:31 development that is Sep 24 11:34:36 Skry, I found link for that theora binary: http://code.entropywave.com/download/leonora/demo/theoradec.dll64P Sep 24 11:34:52 nice! Sep 24 11:35:09 but where is source code I do not know... Sep 24 11:35:22 but I remember that source code was in some git repo Sep 24 11:35:33 but on my hdd I cannot find it Sep 24 11:35:45 I found only that dll64P and link Sep 24 11:37:47 freemangordon: we don't even know if it's the CPU that does the sbc conversion Sep 24 11:37:55 it could be the bt chip itself Sep 24 11:37:58 here is original post: http://blog.mjg.im/2010/04/16/theora-on-n900.html Sep 24 11:41:32 Pali: http://code.entropywave.com/git?p=leonora.git;a=summary Sep 24 11:42:05 and http://code.entropywave.com/git?p=theora-c64x.git;a=summary Sep 24 11:42:31 nice, thanks! Sep 24 11:46:59 freemangordon, so for testing (and comparing) we only need gst-dsp binary which will use that theoradec.dll64P? Sep 24 11:48:35 i think you need to make theora specific code for gst-dsp first? Sep 24 11:51:11 hm... there is theora code for gst-openmax: http://code.entropywave.com/git?p=gstreamer/old/gst-openmax.git;a=summary Sep 24 11:57:34 google's android has modified bluez which support aptx codec via library libbt-aptx-4.0.3.so Sep 24 11:58:06 search for libbt-aptx-2.3.so or libbt-aptx-4.0.3.so Sep 24 12:14:43 Pali: in theory yes, though it depends for which DSP BIOS is codec node compiled for Sep 24 12:15:15 kerio: SBC on n900 is done through MPU, look in bluetoothd source code Sep 24 12:15:55 we have source code of theora dsp codec. can we recompile it for our dsp bios? Sep 24 12:16:27 aah, never tried to recompile DSP codec. AIUI is is a tough job :) Sep 24 12:19:30 ZogG_laptop: Glad I can provide :-) Sep 24 14:29:32 How do I get audible bell working with terminal? Sep 24 14:29:42 in ssh to irssi shell.. Sep 24 14:30:05 bells and all it's settings should be ok Sep 24 15:12:08 hello Sep 24 15:12:25 entitled: not sure it can even do a bell Sep 24 15:12:48 xterm on the n900 is pretty, say, underwhelming Sep 24 15:13:35 I decide to install maemo on moto me863 ,where can I download the source code of maemo 5? Sep 24 15:18:53 yang_: Nokia servers? Sep 24 15:27:26 lava_croft: yeah, I figured. I looked it up before and found it impossible but my friend claimed he had done it before.. but he confused it with something else Sep 24 15:28:17 Jaffa: it's just always something bad happens to you and when i read i just imagine you cold to what saying and just informing sarcastically Sep 24 15:37:24 Lava_Croft: yeh, no bell in osso-xterm. But just maybe you could train bash to do some tricks on seeing ^G Sep 24 15:38:39 in the end it all will call paly-sound Sep 24 15:38:52 play-sound* Sep 24 15:38:58 sorry Pali Sep 24 16:46:00 hello. I'm looking for a handbrake settings with great quality that will work on both my xbox 360 and n900. anyone have one of these presets? Sep 24 16:46:37 depsite looking at page on maemo.org and trying some of those presets, they didn't work for me Sep 24 16:46:50 and this assumes the generic presets aren't worth using Sep 24 16:47:26 i want to encode some dvds that i want to get rid of because they take up far more than compressed video Sep 24 16:47:35 for more space* Sep 24 16:49:59 ~seen zeq Sep 24 16:50:11 zeq <~s_j_newbu@2a01:348:1e3:1:e6ec:10ff:fe9a:d418> was last seen on IRC in channel #maemo-ssu, 11d 2h 47m 2s ago, saying: 'hi freemangordon'. Sep 24 16:50:19 :( Sep 24 17:03:17 toldya, he ran with his sponsored N900, probably to Mexico Sep 24 17:03:34 yeah :( Sep 24 17:10:22 jk4: h264 baseline 3.0, aac audio, mp4 container, target size 700-1400 Sep 24 17:10:50 i tried the baseline and it didn't work Sep 24 17:12:15 always seem to get either something like invalid media format or bad audio codec Sep 24 17:12:18 in so many words.. Sep 24 17:16:56 jk4: I think there was some problems with (weighted?) b-frames but dont remember anymore exactly, also try dropping baseline level, and ofc autocrop and 2-pass encoding. Sep 24 17:20:18 jk4: http://talk.maemo.org/showpost.php?p=1136402&postcount=4 Sep 24 17:21:43 actually saw that post already Sep 24 17:21:46 and I ment _use_ autocrop and 2-pass encoding Sep 24 17:21:48 maybe i did something incorrectly Sep 24 17:22:54 weighted p-frames needs to be disabled too Sep 24 17:23:32 these mostly affect file size, yes? Sep 24 17:24:51 usually everything affects quality, if you dont care about file size you can just crank bitrate to whatever insane value you like :) Sep 24 17:25:33 anyways, i need to go but if you dont get h264 working, use mpeg4 simple profile Sep 24 17:33:41 what audio mixes does n900 support? Sep 24 17:33:48 if you don't care about file size you can just flac dat shit Sep 24 17:37:43 i prefer flac,but i doubt xbox supports flac Sep 24 17:37:54 have to go lowest common denominator for audio sadly Sep 24 17:38:11 could still add a flac version of audio.. Sep 24 17:38:34 flac option doesn't seem to be enabled.. huh.. Sep 24 17:43:53 jk4: and don't OC or undervolt! Sep 24 17:44:33 since a barfing up DSP/GFX-accel will cause exactly same msg Sep 24 17:45:53 OC? Undervolt? :) Sep 24 17:46:07 well, nm then, better you don't know ;-D Sep 24 17:46:43 ~omap-oc Sep 24 17:46:44 [omap-oc] http://mg.pov.lt/maemo-irclog/%23maemo.2010-08-01.log.html#t2010-08-01T22:16:05 read that!, or and this http://mg.pov.lt/maemo-irclog/%23maemo.2011-03-11.log.html#t2011-03-11T03:04:11 Sep 24 17:46:51 ~oc Sep 24 17:46:52 i heard oc is an Optical Carrier, An OC1 has 672 channels or 44.736 Mbit/s Sep 24 17:46:55 meh Sep 24 17:47:00 ~overclocking Sep 24 17:47:00 "OK, listen up. This is your CPU." apt drops the CPU into a hot frying pan. "This is your CPU on overclocking. Any questions?" Sep 24 18:09:08 jk4: you can pump up the dsp if you overclock Sep 24 18:09:16 but it's kinda dangerous Sep 24 18:09:30 that's not what i want Sep 24 18:10:56 that's usually not what people want, unless they want 720p recording/playback Sep 24 18:11:21 but, as DocScrutinizer05 said, overclocking can also give out error messages Sep 24 18:11:55 Foxcon workers win Guinness award for largerst mass battle since british hools Sep 24 18:12:07 heh Sep 24 18:12:10 what happened? Sep 24 18:12:32 2000 engaged in a wresting battle Sep 24 18:12:49 they must have real masses of strong beer there Sep 24 18:13:13 oh, i was hoping it was a mass riot Sep 24 18:13:37 yeah, it was Sep 24 18:13:53 neat Sep 24 18:14:05 they left the place wrecked Sep 24 18:14:08 is there any footage? Sep 24 18:14:11 mass punchup Sep 24 18:14:50 hello guys, does anyone have touch.facebook.com running in microb? its just loads infinitly here Sep 24 18:15:08 never worked properly for me, i use m. Sep 24 18:15:13 or the main site Sep 24 18:15:18 (it loads perfectly fine) Sep 24 18:16:01 ok thanks Sep 24 18:16:19 you might be luckier with opera mobile or fennec Sep 24 18:18:25 dont open in pera mobile (says something about xml error) and i dont like fenec as its slow and unmaintained anymore i think Sep 24 18:33:32 asem: well, do i have a treat for you! Sep 24 18:33:54 thumb-compiled fennec 17 for FREE* Sep 24 18:34:05 (*: must install cssu-thumb) Sep 24 18:36:44 kerio: i will check it out, thanks man :) Sep 24 18:37:11 complain to freemangordon for any problems Sep 24 19:03:46 * DocScrutinizer05 would feel happy if kerio was a tad more reasonable and user oriented with his recommendations. Sep 24 19:06:20 * vi__ would appreciate $50 and a blowjob. Sep 24 19:06:21 DocScrutinizer05: to be fair, fennec thumb-compiled is almost usable Sep 24 19:06:52 vi__: you can find a blowjob and $-50 together, often Sep 24 19:07:26 apt-cache policy blowjob Sep 24 19:08:02 sudo blow_me Sep 24 19:08:48 unzip && mount && fsck && sleep Sep 24 19:09:47 to be fair (to users you don't have any F*ing clue about ther level of experience), your suggestions was like "here's some funny substance, it is warm all the time and emits blue light in the dark. See what you can do with it" Sep 24 19:10:46 "but uh, here's that extremely heavy shirt, you'll like it" Sep 24 19:11:22 At least the lead clothing was provided. :p Sep 24 19:11:32 kerio: lol, what did you do know? Sep 24 19:11:47 DocScrutinizer05: but maybe we'll get a thorium fluoride reactor out of it! Sep 24 19:12:02 vi__: a dude was complaining about fennec being slow, i suggested fennec from cssu-thumb Sep 24 19:13:34 kerio: reasonable suggestion. Sep 24 19:13:42 i know, right? Sep 24 19:14:04 I assume you framed it by noting the caveats of using cssu-thumb? Sep 24 19:14:06 Requires a metric fuckton of caveat emptor, but whatevs. Sep 24 19:14:28 vi__: ...maybe Sep 24 19:18:49 Anyone try that port of the Webkit browser on the n900 -- Doobie? Sep 24 19:18:51 * Dooble Sep 24 19:22:48 great another half finished soon to be abandoned port. Sep 24 19:23:10 shall I file it with midori, tear or helium? Sep 24 19:26:47 Is there something I should know before trying pali's u-boot with cssu-thumb? I think installing kernel-cssu-bootimg should be enough? Sep 24 19:32:46 vi__: I guess Tear served the 8xx well, though. Sep 24 19:33:16 I never had an 8x0. Sep 24 19:33:30 Although I liked the size better Sep 24 19:34:10 wtf is !!Linux kernel for power user (boot Image) breaks the 3rd party package policy? Sep 24 19:34:21 qwazix__: some HAM bullshit Sep 24 19:35:02 hmm... now infinite loops come to my mind. Sep 24 19:35:47 If I force install it, and then I want to update cssu-thumb, I'll find it in front of me, and I'll have to remove bootimg, but if I have u-boot, bootimg is essential Sep 24 19:35:56 aargh! Sep 24 19:36:57 apt-get indeed doesn't complain. I'll risk it. Sep 24 19:37:02 qwazix__: what? Sep 24 19:37:09 which version is that? Sep 24 19:37:24 the one in devel, let me see Sep 24 19:37:37 no,no, cssu-thumb you have installed? Sep 24 19:37:42 1:2.6.28-10cssu3 Sep 24 19:37:54 ah, latest Sep 24 19:37:57 not the kernel :) Sep 24 19:38:21 no wrong. 5.1+thumb0 Sep 24 19:38:29 hmm, installing KP should not break it Sep 24 19:38:44 freemangordon: he wants to install the bootimg Sep 24 19:38:45 not the flasher Sep 24 19:38:48 I'm not installing KP, just kcssu bootimg Sep 24 19:38:51 does not matter Sep 24 19:39:08 which 3rd party policy? Sep 24 19:39:23 it should be listed under "problems" tab Sep 24 19:40:24 oh, I always forget problems tab, I already installed it, let me remove it and check again Sep 24 19:40:40 qwazix__: BTW you should have either KP or KCSSU flasher installed Sep 24 19:40:52 otherwise it won't work Sep 24 19:41:01 HAM has a weird thing where third-party packages can't depend on exact versions of packages listed in the metapackage Sep 24 19:41:20 hmm Sep 24 19:41:37 it can be ignored via redpill Sep 24 19:41:45 It's installed Sep 24 19:42:02 the flasher, that is Sep 24 19:42:47 but you have problems installing bootimg? WTF? Sep 24 19:43:09 * qwazix__ is waiting patiently for HAM to do whatever is doing Sep 24 19:43:18 hehe Sep 24 19:43:40 kerio: wtf, r u high? Sep 24 19:43:48 qwazix__: though it should be a bit faster with -thumb you have installed Sep 24 19:43:56 vi__: ? Sep 24 19:43:58 why? Sep 24 19:44:45 qwazix__: I predict you will be waiting some time. Sep 24 19:44:53 yeah, it is much better, but with devel enabled it's still slow. I wonder if somebody has hidden code to compute fractals inside HAM Sep 24 19:45:11 apt 4 lyfe, yo Sep 24 19:45:56 the problems tab has the following extremely useful information Sep 24 19:46:29 Unable to install 'Linux kernel for Community SSU (boot image for u-boot and multiboot)' Sep 24 19:46:45 it's because the 3rd party policy is half-assed Sep 24 19:47:27 I'll upgrade to latest cssu-thumb and try again Sep 24 19:48:36 freemangordon, ^^ (problems tab) Sep 24 19:48:39 ham can suck my nuts. Sep 24 19:48:39 I must make a confession. Sep 24 19:48:39 I installed thumb by usihg apt. Sep 24 19:48:39 I know. Sep 24 19:48:39 I wish xterm worked properly. Sep 24 19:49:02 qwazix__: yeah, saw it Sep 24 19:49:13 anything in the HAM log? Sep 24 19:49:33 where's the HAM log? Sep 24 19:49:41 in the menu Sep 24 19:50:21 "Checking for updates, please wait"... Sep 24 19:50:35 :) Sep 24 19:54:22 Except the same error message about breaking 3rd party policy, there is a load of errors "Ignoring version from wrong domain: libsdl-mixer1.2 1:1.2.6-5+0m5+ogg+mp3 Sep 24 19:54:54 qwazix__: hmm, something is wrong there Sep 24 19:55:30 what does apt-get upgrade say? Sep 24 19:56:25 BTW i'll try to install -bootimg on my seconde device, lets see hot it will go here Sep 24 19:56:34 s/hot/how/ Sep 24 19:56:35 freemangordon meant: BTW i'll try to install -bootimg on my seconde device, lets see how it will go here Sep 24 19:58:14 libqtm-12-whatever, libsdl-mixer1.2, mypaint, mad-developer kept back, mp-fremantle-community-pr, busybox power, dosfstools and some apps to be upgraded Sep 24 20:00:00 I'll upgrade cssu-thumb and try again after the upgrade Sep 24 20:00:08 ok Sep 24 20:02:32 qwazix__: no problem here :( Sep 24 20:03:31 hmm, I had the same problem when trying to upgrade cssu-thumb, and removed bootimg by hand. DrFrostDk on tmo had the same problem IIRC Sep 24 20:03:56 could it be libqtm? Sep 24 20:04:06 anyway, I can live with installing bootimg with apt-get, but I am curious if the same thing will happen next time I upgrade cssu-thumb Sep 24 20:04:15 possibly Sep 24 20:04:40 http://wiki.maemo.org/Red_Pill_mode Sep 24 20:04:51 qwazix__: 5.1-thumb0 to 5.1-thumb1 should have no problems (and it does not have here) Sep 24 20:04:53 I really missed it :-D Sep 24 20:05:19 It doesn't seem to have any problems here either. (without the bootimg) Sep 24 20:05:33 I mean, typing "matrix" into some URL, if that's not geeky then I dunno Sep 24 20:06:28 DocScrutinizer05, I missed it too especially after PR1.whatever where it was the only way for me to install debs outside of the repos (had no idea of dpkg then) Sep 24 20:06:50 qwazix__: hmm, that is really strange. it could be that HAM plays it smart and tries to autoremove cssu-kernel-flasher. though i can see a reason why it want to do that, -bootomg does not provide kernel-feature-errata-... Sep 24 20:07:11 *i can not see Sep 24 20:07:42 DocScrutinizer05: the hackerbox is a lot geekier Sep 24 20:10:35 freemangordon, I am in thumb1 successfully, flasher is still installed, let's see what HAM says about bootimg Sep 24 20:11:33 "WARNING: Red Pill is not intended for use by users, power-users, nor the vast majority of developers. Red Pill mode is very likely to break your tablet and should not be used unless you know exactly what you are doing." Sep 24 20:11:42 but rootsh is totally fine, isn't it Sep 24 20:12:31 kerio, I love the message that sudo gainroot spits out without rootsh installed... Sep 24 20:22:32 * kerio just optified /var/lib/apt and /var/lib/dpkg Sep 24 20:22:44 66MB free on rootfs Sep 24 20:22:59 with a size of 228MB, so it's more like 80 Sep 24 20:30:12 freemangordon, still not installable neither kcssu-bootimg, nor kp-bootimg Sep 24 20:30:47 qwazix__: NFC :( Sep 24 20:30:59 it's probably some apt-spaghetti I've got on my device. Nevermind. Sep 24 20:31:12 yeah Sep 24 21:04:46 hm, why is IRC network acyclic? Sep 24 21:05:36 freemangordon, you need kernel-power from git Sep 24 21:05:51 Pali: from git? Sep 24 21:05:52 which provides kernel-feature-rx51-battery Sep 24 21:06:04 yes, I did not rebuilded pre release yet Sep 24 21:06:06 isn't the one in TMO thread ok? Sep 24 21:06:31 missing provdes line in debian/control Sep 24 21:06:33 well, is it only the dependensy that is missing? Sep 24 21:06:41 yes Sep 24 21:06:53 i'll dpkg --force-depends it Sep 24 21:06:58 ok Sep 24 21:07:01 (or whatever the switch is ) Sep 24 21:07:12 btw which package needs that? Sep 24 21:07:13 did anyone who has updated to flash10 have problems with no sound? Sep 24 21:07:21 hald-addon-bme needs dsme library for compile Sep 24 21:07:22 vi__: lots of people Sep 24 21:07:35 vi__: just increase the volume Sep 24 21:07:45 Pali: ok Sep 24 21:08:37 there is also fake bme-rx-51 package which replace bme daemon Sep 24 21:08:43 but it is not tested yet Sep 24 21:08:46 Pali: do i need it? Sep 24 21:08:54 ah, ok Sep 24 21:08:57 for testing hald-addon-bme and libbmeipc not Sep 24 21:09:13 Pali: how to restart hald-addon-bme? Sep 24 21:09:27 I didn't find a way when I was playing with it :) Sep 24 21:09:31 freemangordon: I have done. On the flashplayer, globally, alsa, everywhere! Sep 24 21:09:32 for testing you need to stop bme and manually load kernel modules Sep 24 21:09:42 to restart hald-addon-bme use: Sep 24 21:09:51 0. stop hal Sep 24 21:09:52 yes, but how to restart addon without reboot? Sep 24 21:09:57 ok Sep 24 21:10:02 1. /etc/init.d/hal start Sep 24 21:10:20 wait Sep 24 21:10:28 too late Sep 24 21:10:34 :D Sep 24 21:10:39 stop hal must be called first Sep 24 21:10:47 but it only mark hal as stopped Sep 24 21:10:53 in upstart Sep 24 21:11:04 but it still running Sep 24 21:11:13 so? killall hald? Sep 24 21:11:16 so then you need to kill hal (all daemons) manually Sep 24 21:11:23 ps | grep hal Sep 24 21:11:31 and kill pids... Sep 24 21:11:41 and then use only /etc/init.d/hal Sep 24 21:11:49 ok Sep 24 21:12:07 do not use start|stop hal --> it is buggy and working only at startup Sep 24 21:12:51 note that log file is still enabled and hald-addon-bme print it to /tmp/ Sep 24 21:12:59 ok Sep 24 21:13:15 so I need bq_battery, bq_charger and rx51_battery? Sep 24 21:13:30 yes Sep 24 21:14:10 insmod: error inserting 'rx51_battery.ko': -1 File exists Sep 24 21:14:12 ?!? Sep 24 21:14:41 rx51_battery is loaded at startup Sep 24 21:14:46 ok Sep 24 21:14:52 automatically, not blacklisted Sep 24 21:15:06 use modprobe instead insmod Sep 24 21:15:16 insmod did not loaded dependency modules Sep 24 21:15:18 all is in place Sep 24 21:15:34 power-supply is already loaded because of rx51_battery Sep 24 21:15:36 modprobe also ignore loaded modules Sep 24 21:15:38 ok Sep 24 21:15:49 anyway, everything is loaded Sep 24 21:15:54 what now :D? Sep 24 21:16:10 replace hald-addon-bme Sep 24 21:16:13 and restart hal Sep 24 21:16:16 already did it Sep 24 21:16:22 so it should work Sep 24 21:16:24 and libbmeipc too Sep 24 21:16:35 see lshal Sep 24 21:16:43 check battery status menu applet Sep 24 21:16:49 battery.reporting.design = 1303 Sep 24 21:17:08 battery.charge_level.design = 5172 Sep 24 21:17:11 check value against /sys/class/power_supply/... Sep 24 21:17:12 ?!? Sep 24 21:17:49 give me output of uevent of power_supply drivers first Sep 24 21:18:33 POWER_SUPPLY_CHARGE_FULL=1554556 Sep 24 21:18:44 POWER_SUPPLY_CHARGE_NOW=1338750 Sep 24 21:18:49 that is bq_battery Sep 24 21:19:04 what else? Sep 24 21:19:06 turns out I went full retard. Sep 24 21:19:18 silent mode overides everything. Sep 24 21:19:25 >> battery.reporting.design = 1303 << this is OK I belive Sep 24 21:19:29 but media player Sep 24 21:19:34 -_-' Sep 24 21:19:42 *mafw Sep 24 21:19:48 should be, this is standart Nokia battery Sep 24 21:19:58 Pali: a typo "POWER_SUPPLY_MODEL_NAME=chip bq24150a, revision 1.3, vender code 010" Sep 24 21:20:05 in charger, "vender" Sep 24 21:20:18 it is not typo :D Sep 24 21:20:24 Doc already reported it Sep 24 21:20:31 what?!? Sep 24 21:20:32 vender is correct too Sep 24 21:20:37 see dict Sep 24 21:20:42 the fuck :D:D:D Sep 24 21:20:46 and in datasheet is used vender Sep 24 21:20:50 ok,ok Sep 24 21:21:04 so I wrote vender to driver Sep 24 21:21:44 hmm, POWER_SUPPLY_CURRENT_NOW=39270 Sep 24 21:21:52 is that 39mA? Sep 24 21:22:13 which driver? Sep 24 21:22:18 bq_battery Sep 24 21:22:27 it could be Sep 24 21:22:46 hmm, strange, I have only wifi connected Sep 24 21:23:10 ah I see, battery.charge_level.design is number 1..8 for standard battery Sep 24 21:23:30 ~5172/150 Sep 24 21:23:30 34.48 Sep 24 21:23:44 the fuck "HWRecoveryResetSGX: SGX Hardware Recovery triggered" :( Sep 24 21:23:57 ~1338750/150000 Sep 24 21:23:58 8.925 Sep 24 21:24:04 ok Sep 24 21:25:03 freemangordon, I think I know where is problem Sep 24 21:25:07 Pali: I think it should be hardcoded to 8 Sep 24 21:25:18 I think no Sep 24 21:25:32 otherwise battery applet will be confused Sep 24 21:26:48 https://gitorious.org/rx51-bme-replacement/hald-addon-bme/blobs/master/README.properties#line85 Sep 24 21:27:32 Pali: ok, will see Sep 24 21:29:52 Pali: battery.charge_level.current = 0 (0x0) (int) Sep 24 21:30:13 Pali: how often is it polled? Sep 24 21:30:23 60sec? Sep 24 21:30:43 30s Sep 24 21:30:49 check log Sep 24 21:30:55 cat /tmp/bme... Sep 24 21:31:08 /tmp/hald-addon-bme.log Sep 24 21:32:42 log looks ok Sep 24 21:33:17 but battery.charge_level.current = 0 (0x0) (int) Sep 24 21:33:57 capacity % ? Sep 24 21:34:15 battery.charge_level.percentage = 0 (0x0) (int) Sep 24 21:34:28 POWER_SUPPLY_CAPACITY=84 Sep 24 21:34:49 give me full output of hal bme properties (lshal -u /something/bme/...) and all 3 uevent files Sep 24 21:34:52 to pastebin... Sep 24 21:34:58 ok Sep 24 21:35:16 and also /tmp/hald-addon-bme.log Sep 24 21:36:25 Pali: hmm, which hal device? Sep 24 21:36:33 bme? Sep 24 21:36:40 aah, ny bad then :D Sep 24 21:36:44 *my bad Sep 24 21:36:46 some device which has bme substring Sep 24 21:37:54 this one looks almos ok Sep 24 21:37:57 *almost Sep 24 21:38:02 besides: Sep 24 21:38:18 battery.reporting.current = 1304 (0x518) (int) Sep 24 21:38:25 battery.reporting.design = 1297 (0x511) (int) Sep 24 21:38:35 but I think we can live with that Sep 24 21:39:20 battery.charge_level.last_full = 10 (0xa) (int) Sep 24 21:39:30 looks suspicios too Sep 24 21:40:16 can you give me full output? Sep 24 21:41:26 http://pastebin.com/z5qY4gnL Sep 24 21:43:11 Pali: what is the upper temp limit in rx51_battery? Sep 24 21:43:46 48°C Sep 24 21:43:57 hmm, too low Sep 24 21:44:19 we should cover at least 60-65 deg Sep 24 21:44:39 if value is low, then 49°C is reported Sep 24 21:45:02 you mean if it is above 48? Sep 24 21:45:38 if RAW value is too low, then 49°C is reported Sep 24 21:45:44 but AIUI there is no way to trigger thermal shutdiwn if we report incorrect values Sep 24 21:46:06 yes, we need some approximation formula Sep 24 21:46:19 but I did not implemented anything... Sep 24 21:46:30 Pali: I sent it to you, do you keep it? Sep 24 21:46:43 4th order polinomial Sep 24 21:46:44 I do not have it Sep 24 21:46:47 ok Sep 24 21:47:19 hmm, a better idea (doc will like it) Sep 24 21:47:35 https://meego.gitorious.org/~pali/meego-device-adaptation/pali-n900_libbme/blobs/master/include/em_isi.h#line1566 Sep 24 21:47:46 seems that bars are really 1-8 Sep 24 21:47:51 I will change code Sep 24 21:48:07 i can use some 10th order polinomial to approximate range from -30-90 deg Sep 24 21:48:18 gimme some toime, will pastebin the result Sep 24 21:48:44 -30 to 90 should be ok, ain't DocScrutinizer05? Sep 24 21:52:09 friggin netsplits! Sep 24 21:53:01 but why it reported this??? battery.charge_level.design = 5172 Sep 24 22:00:21 freemangordon, can you upload also /tmp/ log file? Sep 24 22:00:41 I'd like to know why it reported battery.charge_level.design = 5172 Sep 24 22:01:18 will reboot, to remove the noise Sep 24 22:05:16 Pali: have in mind that was for /org/freedesktop/Hal/devices/computer_power_supply_battery_rx51_battery Sep 24 22:05:50 ok Sep 24 22:06:15 so it is something in rx51_battery driver Sep 24 22:07:15 though seems like uninitialized variable, there is no way you can know design capacity in rx51_... Sep 24 22:15:16 we can ignore /org/freedesktop/Hal/devices/computer_power_supply_battery_rx51_battery Sep 24 22:15:20 Pali: is there any chance to get some measurement at lets say 55 deg? Sep 24 22:15:58 freemangordon, we can patch twlmadc kernel driver to reported fixed RAW value (set from user space) Sep 24 22:15:58 or even higher? Sep 24 22:16:09 and then ask BME for converted temperature Sep 24 22:16:16 hmm, yes, you are right Sep 24 22:16:25 no need to do approximation then Sep 24 22:16:31 yes Sep 24 22:16:45 but it needs to turn off some mce/dsme/... protection Sep 24 22:17:11 the problem is that depending on ppolinomial order, I get too different values above 64 deg Sep 24 22:18:23 freemangordon, I can ask bme devs (again) how to use this API: https://meego.gitorious.org/~pali/meego-device-adaptation/pali-n900_libbme/blobs/master/include/bmeisa.h#line115 Sep 24 22:18:28 well, no need, once we have more points, the approximation should be correct to lets say 80 deg. that should be pretty enough for all practical purposes Sep 24 22:19:11 BME sent me some data for ADCIN0 struct bmeisa_adc_cal_data Sep 24 22:19:26 but I do not know which union is used Sep 24 22:19:50 and what that bmeisa_adc_*_conv_cal_par_t means... Sep 24 22:21:10 wow, I just met a RL troll Sep 24 22:22:07 (entering troll mode myself) WTF is fixed RAW value Sep 24 22:24:19 hope you still have the pastebin URL of my freezer enterprise Sep 24 22:24:59 I really think there's no use in any nifty polinomail formula since we got a LUT anyway Sep 24 22:25:37 a LUT with *for sure* less than 150 tuples Sep 24 22:25:54 2 * int16 Sep 24 22:26:28 you basically can't implement any decent polinomial in that number of bytes Sep 24 22:28:10 if you need values for 50, 60 , even 90°C, just take a sheet of paper and extrapolate the curve strictily human-manually Sep 24 22:29:04 a) we're not interested in precise absolute values Sep 24 22:29:27 b) there's a noise of like +/-4°C anyway Sep 24 22:30:58 c) if anybody comes claiming we needed a globaly applicable formula, I just answer it's nonsense, since the coefficients for that formula need to come fom board parameter file just same way the 150 tuples for LUT have to Sep 24 22:32:18 and no, there's absolutely no need to calculate any nonsense like sensor current or whatever Sep 24 22:37:08 tbh we could as well get away with just THREE tuples in LUT: -> "too cold" -> 0°K;a -> "it's fucking alright" -> 290°K; -> "too hot" -> 400°K; Sep 24 22:38:04 where "too cold" is like -25°C, and "too hot" is like 70°C Sep 24 22:38:12 http://www.ipazzport.com/18BR.html Sep 24 22:38:13 nice Sep 24 22:38:15 nothing else matters Sep 24 22:38:24 looks like the n900 keypad heh Sep 24 22:38:51 so that's what nokia did with all those spare parts! Sep 24 22:38:53 :) Sep 24 22:41:25 actually make that TWO tuples, since it's just 3 ranges Sep 24 22:41:36 2 of them open Sep 24 22:41:53 makes 2 borders aka tuples in LUT Sep 24 22:43:21 GGon: I pretty much wish my N900 had THAT keymat Sep 24 22:43:33 heh Sep 24 22:43:48 well yeah. it has the F keys Sep 24 22:43:52 and esc ;) Sep 24 22:44:03 caps and tab too heh Sep 24 22:44:16 nokia should make a wp phone with that :-P Sep 24 22:44:32 here is BME output for temperature adcin calibration data: https://meego.gitorious.org/~pali/meego-device-adaptation/pali-n900_libbme/blobs/master/include/bmemsg.h#line335 Sep 24 22:44:32 you're aware keymat is one coherent component, are you? Sep 24 22:44:33 0x05 0x03 0x00 0x00 0x64 0x00 0x34 0x0d 0x2a 0x01 0x00 0x00 0x03 0x00 0x00 0x00 Sep 24 22:45:03 one coheren component? Sep 24 22:45:20 Pali: sorry, friend. I'm too grumpy right now to even comment on that Sep 24 22:46:00 GGon: one fsckng sheet of plastic Sep 24 22:46:21 Pali: hmm, seems like some coefficient, not LUT :( Sep 24 22:46:32 so what? Sep 24 22:46:53 or it looks like cell's last cheers and moans Sep 24 22:47:21 here is union: https://meego.gitorious.org/~pali/meego-device-adaptation/pali-n900_libbme/blobs/master/include/bmeisa.h#line115 Sep 24 22:47:31 DocScrutinizer05: meh.. i want it for xmbc ;) Sep 24 22:47:51 honestly I offer a bet about 12 bottles Klipshift that there's no calibration data for R1110 Sep 24 22:48:37 it's plain silly to calibrate such a component, with THAT amount of noise Sep 24 22:49:10 DocScrutinizer05: AIUI we are trying to understand what BME does Sep 24 22:49:11 and I've always seem *multiple* BME records in CAL Sep 24 22:49:50 freemangordon: maybe you do, I always stated we better refuse to even look at bme since it can't get worse than that Sep 24 22:50:53 search chanlogs foe " NOOO, pretty please DO NOT disclose BME sourcecode! We don't want to see it!" Sep 24 22:51:22 Pali: hacking adc raw in twl seems the best option, we'll directly get LUT Sep 24 22:51:40 hacking WUT? Sep 24 22:52:05 faking raw register readouts? Sep 24 22:52:08 ok Sep 24 22:52:09 congrats! Sep 24 22:52:22 I wrote this more times! Sep 24 22:52:26 freemangordon, we can patch twlmadc kernel driver to reported fixed RAW value (set from user space) Sep 24 22:52:28 yes Sep 24 22:52:48 WTF!!! is a "fixed RAW value"??? Sep 24 22:52:59 you tell driver some fixed value Sep 24 22:53:00 it's nonsense Sep 24 22:53:06 and driver will report it Sep 24 22:53:27 aaah, ok Sep 24 22:53:29 nm Sep 24 22:53:51 thought you gonna "fix" the raw values reported Sep 24 22:53:54 Pali: seems charging works too, though I would not leave it overnight :D Sep 24 22:54:12 then it doesn't work Sep 24 22:54:33 did I ask something? Sep 24 22:54:42 Pali: POWER_SUPPLY_CURRENT_NOW=-426793 Sep 24 22:54:57 its ok Sep 24 22:55:08 when charging it reports negatove value Sep 24 22:55:18 I know, i wonder about the value Sep 24 22:55:28 isn't that like 500mA? Sep 24 22:55:44 -426.793mA Sep 24 22:55:56 connected to wall charger? Sep 24 22:56:04 looks bad to me Sep 24 22:56:05 sounds good Sep 24 22:56:20 isn't that supposed to be something like >800mA Sep 24 22:56:41 it is from bq27200 chip Sep 24 22:56:53 *burp* Sep 24 22:57:14 aah, ok Sep 24 22:57:29 seems I have to take some sleep :D Sep 24 22:57:37 http://wiki.maemo.org/User:Joerg_rw/jrbme/bme_discharge_learn_log Sep 24 22:57:56 we should not overcharge(overcurrent) the battery Sep 24 22:59:28 http://wiki.maemo.org/User:Joerg_rw/jrbme/bme_microcycle_log Sep 24 22:59:38 ORLY?! Sep 24 23:00:48 pshshst! maybe you missed it: **we can't really overcurrent the battery!** Sep 24 23:01:29 since bq24150 doesn't provide more than 0.9C Sep 24 23:01:43 0.8? Sep 24 23:02:24 and thanks bq24150 we also can't really severely overvolt the battery Sep 24 23:03:22 and I posted same statement some 30 months ago, and nobody except ShadowJK and SpeedEvil understood Sep 24 23:03:36 freemangordon, I pushed hald-addon-bme change which fix max number of bars to 8 Sep 24 23:04:11 Pali: ok, will test it tomorrow, going to bed now :) Sep 24 23:04:11 probably also a board parameter X-P Sep 24 23:04:18 ok Sep 24 23:05:18 night Sep 25 02:28:39 * ShadowJK sees 850mA or more charge current with his charge script and power kernel Sep 25 02:29:06 can be tweaked between ~100 and 1200ish Sep 25 02:29:32 higher end only achievable with a good battery that's empty **** ENDING LOGGING AT Tue Sep 25 02:59:58 2012