**** BEGIN LOGGING AT Wed May 13 02:59:59 2015 May 13 05:32:04 Can't locate PerlIO.pm in @INC May 13 05:32:04 T_T May 13 05:32:32 oh May 13 06:08:50 KotCzarny: lol mediabox sucks May 13 06:09:10 what's mediabox? May 13 06:09:25 https://bpaste.net/show/cd54b9d3719c May 13 06:09:30 the stock player May 13 06:09:30 L29Ah: it does. May 13 06:10:08 https://bpaste.net/show/2f85e21fac9c measuring program May 13 06:10:09 I installed it on my and I was disappointed. May 13 06:10:21 wow May 13 06:10:38 decoding on cpu is more power effective? May 13 06:10:39 :) May 13 06:10:49 oh well. it also sucks battery. May 13 06:11:07 also, mind you, pygtk is very unoptimized May 13 06:11:44 i wonder if one can sigstop it when screen is off May 13 06:12:24 i might add check for screen off and simply stop updating gui May 13 06:25:36 * L29Ah wonders how to register to the wiki May 13 06:34:03 drathir: it doesnt matter which core is more loaded (cpu vs dsp) because it has to do similar work, and cpu decoding is quite effective nowadays May 13 06:34:33 *efficient May 13 06:34:35 :] May 13 06:34:40 pf May 13 06:34:45 but right May 13 06:35:08 L29Ah: mind pasting those results in oscp tmo thread? May 13 06:35:27 link? May 13 06:35:36 http://talk.maemo.org/showthread.php?t=94590 May 13 06:36:00 also, you might also test omp (open media player) May 13 06:36:12 which is usually recommended over stock media player May 13 06:36:39 also, is that mp3 constant bitrate? May 13 06:37:38 dunno May 13 06:37:43 KotCzarny: (CPU decoding) - no way cortex A8 is as efficient as C64x DSP in de/coding of audio-video, even when using NEON May 13 06:37:51 i played it from the beginning every time tho May 13 06:38:07 DSP has special instructions to deal with the most common tasks in HW May 13 06:38:25 freemangordon: then it must be gui drawing that does more work May 13 06:39:36 KotCzarny: I couldn't find the context in the backscroll, but anyway, offloading multimedia stuff to DSK is 1 - battery efficient, 2 - frees CPU to do the job it is meant for May 13 06:39:37 also, i wonder how much power would be saved bypassing pusleaudio May 13 06:39:42 https://bpaste.net/raw/cd54b9d3719c May 13 06:39:46 KotCzarny: depends May 13 06:39:51 that's the link i'm commenting May 13 06:40:50 PA uses more power when the output is on speakers because of the x-prot May 13 06:41:05 but results are similar with headphones on too May 13 06:41:12 it is not PA to balme, but the flawed HW design May 13 06:41:21 similar? May 13 06:41:28 100mW diff May 13 06:41:35 ie. stock player using 5-10% power more May 13 06:42:04 which one is the stock player? May 13 06:42:07 mediabox May 13 06:42:18 never heard about it :D May 13 06:42:23 is that on n900? May 13 06:42:28 yes May 13 06:42:37 stock player is called mediaplayer May 13 06:42:44 but it is gui only May 13 06:43:01 the real one is ... err... dammit May 13 06:43:02 does it have an option to play in background? May 13 06:43:06 gstreamer May 13 06:43:08 no May 13 06:43:17 just a second to remember May 13 06:43:21 it doesnt use gst as a backend? May 13 06:43:27 hmm, i had mediabox preinstalled on my thing May 13 06:43:28 it uses gst May 13 06:43:45 L29Ah: it is not in the FW May 13 06:44:09 * freemangordon feels stupid May 13 06:44:16 :) May 13 06:44:18 oooh May 13 06:44:27 new day, new knowledge May 13 06:44:29 it is the so-called mafw framework May 13 06:44:43 isnt mafw spawn of evil? May 13 06:44:52 http://wiki.maemo.org/Documentation/Maemo_5_Developer_Guide/Using_Multimedia_Components/Media_Application_Framework_%28MAFW%29 May 13 06:45:08 doesn't matter, this is what the stock player uses May 13 06:45:15 also, why evil? May 13 06:45:16 okay i'll go measure mediaplayer May 13 06:45:41 can't find how to make it play a file May 13 06:45:41 i remember some people having issues with mafw being hard to use May 13 06:45:52 L29Ah: make sure tracker has settled down before you start measuring May 13 06:46:16 tracker starts everytime you open stockplayer? May 13 06:46:40 depends May 13 06:46:44 damn can i do this at all? May 13 06:46:52 if there is new stuff to be indexed May 13 06:47:10 freemangordon: but it has to start to check if there is new stuff, right? May 13 06:47:14 L29Ah: just open mediaplayer and select a song to be played :) May 13 06:47:23 my question is, does it do it every time you open stockplayer? May 13 06:47:29 > (no songs) May 13 06:47:33 :> May 13 06:47:37 i killed the tracker thing and forgot how May 13 06:47:45 this is the first time you start it? May 13 06:47:49 yep May 13 06:47:49 aaah May 13 06:47:55 killed... May 13 06:48:12 not as in "kill(1)" May 13 06:48:12 mafw uses tracker db to feed mediaplayer May 13 06:48:53 L29Ah: you should find a way to revive it :) May 13 06:49:09 i remember there was some gui to configure it May 13 06:49:17 tracker cfg :) May 13 06:49:35 yeah May 13 06:49:43 i disabled tracker on mine too May 13 06:49:49 why? May 13 06:49:55 because i hate stock player May 13 06:50:08 that is not a reason :P May 13 06:50:15 i killed it because it was thrashing on my gentoo chroot May 13 06:50:17 and because i have a gui to play audio/video without it May 13 06:50:30 tracker dose more stuff than audio/video May 13 06:50:34 *does May 13 06:50:41 way more May 13 06:50:46 freemangordon: i dont need it, no app i use needs it May 13 06:51:03 I assume you don't really use your n900 for anything but a toy May 13 06:51:05 i hate all those trackers/indexers May 13 06:51:28 freemangordon: i use it for sporadical internet browsing, a/v player, phone May 13 06:51:44 your daily device? May 13 06:51:48 and sometimes shell work May 13 06:51:54 yes, its my main phone May 13 06:52:00 since 2009 May 13 06:52:15 so, you don;t use e-mail, do not take shots? May 13 06:52:20 nope May 13 06:52:30 ah, right, camera too May 13 06:52:35 but no email/social stuff May 13 06:52:48 for social i use ssh/irc May 13 06:52:51 so, aiui you use ~10 percent of what it is capable May 13 06:53:12 nope, some gaming too May 13 06:53:12 but for 100% usage tracker is a must May 13 06:53:30 I mean producing, not consuming :) May 13 06:53:32 you still didnt say what needs it May 13 06:53:39 email search May 13 06:53:43 image gallery May 13 06:53:46 i use web based email May 13 06:54:00 for image gallery i use my own app May 13 06:54:12 :) May 13 06:54:14 yeah May 13 06:54:15 more? May 13 06:54:45 yes, there are more iirc, but need to dig deeper to find what else May 13 06:54:47 also, constant tracking/indexing is evil on mobile app, it should be done when actual needs arise May 13 06:54:59 who said it does constant indexing? May 13 06:55:00 if you need to dig more, then its not that essential May 13 06:55:11 constant as in 'you start an app, tracker starts' May 13 06:55:20 it is not how it works May 13 06:55:49 I warned L29Ah because I was almost sure it will be the first time he'll use MP May 13 06:55:55 415mW on mediaplayer on speakers May 13 06:55:59 he? she? May 13 06:56:07 he May 13 06:56:15 who cares? May 13 06:56:19 :) May 13 06:56:22 :) May 13 06:56:31 so, after all stock MP is not that bad :) May 13 06:56:37 L29Ah: now with headphones May 13 06:56:44 ye May 13 06:56:55 freemangordon: yeah, but its usability is flawed May 13 06:57:09 i hate media libraries, i organize things by directory/file name May 13 06:57:13 depend on your use cases May 13 06:57:17 *depends May 13 06:57:40 ie. most general genre/detailed genre/artists/albums/tracks May 13 06:57:44 262mW May 13 06:57:46 o_O May 13 06:57:49 KotCzarny: so, what stops you from having file manager opened? May 13 06:57:51 :) May 13 06:57:59 freemangordon: i use mc May 13 06:58:10 as file manager May 13 06:58:12 so? May 13 06:58:21 so i dont need trackerd May 13 06:58:22 bumped the wiki May 13 06:58:36 tell mc to open MP when you press enter on a media file May 13 06:58:45 i use mplayer for enter commands May 13 06:58:51 anyway May 13 06:58:57 it is your device :) May 13 06:59:14 L29Ah: what FW is that? May 13 06:59:23 stock? CSSU? CSSU-thumb? May 13 06:59:26 yes, i know, but i like to have choice, and not have things shuffled by force because 'its needed by ..' May 13 06:59:36 mediaplayer seems to stop the gui when the screen is off May 13 06:59:41 freemangordon: dunno May 13 06:59:46 oh, come on May 13 06:59:48 non-thumb i guess May 13 06:59:55 check in control panel May 13 07:00:27 21.2011.38-1 May 13 07:00:28 L29Ah: also, which kernel May 13 07:00:30 w/ power kernel May 13 07:00:36 profile? May 13 07:00:48 wat May 13 07:00:52 :) May 13 07:00:53 kernel profile that is May 13 07:01:01 dunno what's kernel profile May 13 07:01:12 do you have kernel-power-settings installed? May 13 07:01:18 ~kernel-power May 13 07:01:29 ~lady May 13 07:01:33 ~ping May 13 07:01:33 ~pong May 13 07:01:45 ~kernelpower May 13 07:01:49 ~power-kernel May 13 07:01:49 rumour has it, power-kernel is http://wiki.maemo.org/Kernel_Power May 13 07:01:56 10x May 13 07:01:57 ye May 13 07:01:58 hm May 13 07:02:12 i remember hitting `kernel-load` on some customized file May 13 07:02:37 hmm, kernel-load is ages old May 13 07:02:44 which KP version is that? May 13 07:02:52 freemangordon: https://bpaste.net/show/28515c0e339f May 13 07:02:57 2.6.28.10-power53 May 13 07:03:11 uh May 13 07:03:14 900 MHz May 13 07:03:45 some wiki page said smartreflex thing doesn't work on bigger freqs May 13 07:03:50 so i didn't try them May 13 07:03:55 why "uh"? May 13 07:04:07 * freemangordon says it doesn't work for anything > 805 May 13 07:04:19 900 is too much May 13 07:04:46 lol then smartreflex sucks and doesn't matter as consumption at 900 = consumption at 250 / 250 * 900 May 13 07:05:09 doesn't work == not reliable :) May 13 07:05:22 i'm yet to see it crashing May 13 07:05:36 yes, because your DSP is downclocked May 13 07:05:49 or rather not OC enough May 13 07:05:51 Since Kernel Power 50 (KP50), despite some disagreements amongst some of the more senior Maemo developers, SmartReflex can now be considered stable up 900MHz with stock DSP frequencies and up to 600Mhz for overclocked DSP May 13 07:05:59 wiki need to be updated then May 13 07:06:08 I know I just said SR is stable up to 900MHz however some users have reported that on their devices SR is only stable up to lower frequencies (720-805MHz for example). For more examples on tuning SR see the KP50 article. May 13 07:06:59 * L29Ah hugs his brick May 13 07:07:10 you could check who made the latest SR related patches, no need for a wiki ;) May 13 07:07:26 hehe May 13 07:08:33 900 is not even considered in kernel May 13 07:08:45 it uses the voltage for 805, iirc May 13 07:09:05 so, L29Ah has a very stable device May 13 07:09:52 but that stability could be because of the usage pattern as well May 13 07:10:19 i was compiling gentoo for hours at 900MHz May 13 07:11:14 oh, and battery usage on 900 max as about 1.5 times more than on 805 and about 2 times more than on 500. iirc, there were some graphics on TMO made by vi May 13 07:11:48 sounds bullshitty May 13 07:11:52 lemme check 805 May 13 07:11:53 hmm? May 13 07:11:58 :) May 13 07:12:10 L29Ah: load "dsp" profile May 13 07:12:15 kernel-config load dsp May 13 07:12:52 why? May 13 07:13:20 because this is the profile KP53 is supposed to work woth for 250-805 range May 13 07:13:26 *with May 13 07:13:37 it should "just work" May 13 07:13:44 i go mess w/ scaling_max_freq and run some infinite loops May 13 07:13:56 most of the times, unless you;re unlucky and your DSP doesn;t tolerate SR May 13 07:13:58 infinite loop will only check cpu May 13 07:14:00 and not dsp May 13 07:14:06 who cares about dsp? May 13 07:14:11 me May 13 07:14:13 :) May 13 07:14:30 and all the others who play videos on n900 May 13 07:14:47 yeah, videos are the only reason to care May 13 07:14:57 it failed to play my 720p porn and i'm too lazy to install hacks from tmo May 13 07:15:09 hacks? May 13 07:15:15 come on May 13 07:15:18 1064mW at 900MHz May 13 07:15:34 you locked it @900 May 13 07:15:37 |? May 13 07:15:39 ? May 13 07:15:50 infinite loop should lock it at max May 13 07:15:54 oh May 13 07:17:17 903mW at 805MHz May 13 07:17:21 freemangordon ← he sucks May 13 07:18:36 L29Ah: this is not how you measure n900 power consumption May 13 07:19:03 a real usage pattern never puts CPU @max constantly May 13 07:19:36 L29Ah: check the 600mhz too May 13 07:19:47 no, 500 May 13 07:19:52 isnt 600 the stock one? May 13 07:19:57 it is not linear May 13 07:20:01 it is May 13 07:20:05 freemangordon: yeah i need to hack in some intermittent load simulator May 13 07:20:18 almost linear anyway May 13 07:20:25 no, it is not May 13 07:20:25 278@250 May 13 07:20:41 the "sweet spot" is 500 May 13 07:20:44 λ> 1064 / 900 May 13 07:20:44 1.1822222222222223 May 13 07:20:44 λ> 903/805 May 13 07:20:44 1.1217391304347826 May 13 07:20:44 λ> 278/250 May 13 07:20:45 1.112 May 13 07:20:54 performance/vs mA May 13 07:20:56 805 looks quite sweet May 13 07:21:25 might be some additional things kicking in May 13 07:21:37 L29Ah: your math misses he time needed to finish the job May 13 07:21:43 *the May 13 07:22:25 anyway /me has to run to the office May 13 07:22:31 bye May 13 07:23:10 how so? T_T May 13 07:24:20 because if your compile needs 600s at 900mhz or 900s at 600mhz May 13 07:24:45 so? May 13 07:25:10 so it uses more power at 900, so you would ave to count mW used May 13 07:25:14 mWh May 13 07:25:24 that's why i divided by frequency May 13 07:26:46 leaving the time away and counting milliwatts per megahertz or nanojoules per clock cycle May 13 07:27:38 524@500 May 13 07:27:45 that's assuming mhz relates directly to time May 13 07:27:45 this is too synthetic, what you really need is the energy needed to finish a job J May 13 07:27:48 yeah sweet May 13 07:28:03 L29Ah: oh, so it is > 2x May 13 07:28:07 500 is more efficient than 250 May 13 07:28:12 yes, it is May 13 07:28:14 freemangordon: so you suck May 13 07:28:17 hmm? May 13 07:28:22 why is that? May 13 07:28:35 1.048 * 2 ≠ 1.1822222222222223 May 13 07:28:59 no, it is your math that suck May 13 07:29:02 if we are to measure power consumption per equal jobs May 13 07:29:04 how so? May 13 07:29:18 I still don;t see the time in your equations May 13 07:29:26 you suck May 13 07:29:33 yeah. right May 13 07:29:37 * freemangordon is out May 13 07:29:44 KotCzarny: explain him plz :3 May 13 07:29:57 few lines above May 13 07:30:00 that's assuming mhz relates directly to time May 13 07:30:04 wat May 13 07:30:06 ah May 13 07:30:20 and he says it doesnt May 13 07:30:24 and you says it does May 13 07:30:26 yeah i suck May 13 07:30:39 if we're adding i/o May 13 07:32:10 but if cpu is waiting idle, it should downclock, thats why real test for doing clean+compile would tell real number May 13 07:32:28 better yet 2 or 3 times May 13 07:33:18 compiling is a bad performance criteria for n900 uses May 13 07:33:32 maybe an hour of oscp playback will do May 13 07:33:32 but its the usecase most interesting to you May 13 07:33:38 not really May 13 07:33:55 i'm not going to compile much off grid May 13 07:34:35 you might use oscp core and gstreamer as a usecase May 13 07:38:35 (both callable via ssh) May 13 07:39:04 oh, and could you test ncurses mode for oscp? May 13 07:39:25 should be close to gui-off, but im curious May 13 07:40:17 nah i ran out of fuel, go measure yourself ;p May 13 07:42:11 :) May 13 07:42:24 i wrote it :P May 13 08:32:57 what's the best way to launch ssh-agent? May 13 08:33:13 maybe .profile May 13 08:33:19 ~/.profile May 13 08:33:38 though i dont know if n900's shell reads that file May 13 08:33:56 it does, but every new terminal window is a login shell May 13 08:34:57 ahm, no, looking at my files it worked different May 13 08:35:14 i have ~/.sa containing: May 13 08:35:14 SSH_AUTH_SOCK=/tmp/ssh-ilYYun2285/agent.2285; export SSH_AUTH_SOCK; May 13 08:35:15 SSH_AGENT_PID=2286; export SSH_AGENT_PID; May 13 08:35:15 echo Agent pid 2286; May 13 08:35:25 and it's sourced by .profile? May 13 08:35:25 and in .profile: . /home/ariel/.sa May 13 08:35:32 yeah but what launches ssh-agent? May 13 08:35:48 cant find a trace May 13 08:35:55 might be hand launched May 13 08:35:59 because if you're launching it manually, might as well have a static SSH_AUTH_SOCK=/home/user/.ssh/agent May 13 08:36:12 and launch `ssh-agent -a .ssh/agent` manually May 13 08:36:20 im using hibernation so it was running all the time May 13 08:36:53 dat feel when os x's ssh setup is the best May 13 08:36:53 but as above, you can add some if/else magic May 13 08:37:37 dat feel, when you start to be baffled by f*ckton of weirs osx directories May 13 08:37:41 *weird May 13 08:40:21 moin :) May 13 08:40:28 grroink May 13 08:40:38 KotCzarny: 'bry :) May 13 08:40:48 elo ;) May 13 08:49:34 i don't understand ppl who use ssh-agent May 13 08:50:00 its useful if you use passphrase but dont want it enter more than once per session May 13 08:50:09 use session sharing ffs May 13 08:50:19 it's different May 13 08:50:24 also pubkeys ftw May 13 08:50:27 bencoh: yeah i know May 13 08:50:31 from nokia smlv3 'this product is of superior design and craftmanship and should be treated with care' May 13 08:50:43 L29Ah: what May 13 08:50:44 :3 May 13 08:50:50 wtf is session sharing May 13 08:50:51 KotCzarny: how cute :) May 13 08:50:58 controlmaster? May 13 08:50:59 kerio: man ssh, /master May 13 08:51:02 that's a different thing May 13 08:51:02 yeah May 13 08:51:14 and it's pretty damn awkward if your connection isn't really reliable May 13 08:51:38 * L29Ah drops his pubkey the first time he ssh'es to a machine May 13 08:51:47 there were some tcp/ip tweaks for ssh over flaky connections May 13 08:51:59 l29ah: passphrase May 13 08:52:17 besides, the ssh version on fremantle doesn't really have the fancy stuff that make connection muxing bearable May 13 08:52:27 with sshagent you enter passphrase only once May 13 08:52:39 like controlpersist May 13 08:53:26 also, agent forwarding only works if you're using an actual agent May 13 08:53:32 wow ControlPersist May 13 08:53:36 :* May 13 08:53:39 :) May 13 08:53:45 new day, new knowledge May 13 08:53:55 and to think it was all the time in the man page! May 13 08:54:04 not all the time I guess May 13 08:54:04 L29Ah: if you're doing things right, your private key takes *seconds* to be decrypted May 13 08:54:09 on a real computer May 13 08:54:18 http://www.openssh.com/txt/release-5.6 May 13 08:54:20 i've read the man some years ago but no May 13 08:54:34 ssh on n900 is 5.1 May 13 08:54:38 kerio: huh? May 13 08:54:44 he got n900 only recently May 13 08:55:11 new-format keys allow for multiple rounds of bcrypt May 13 08:55:18 i have unencrypted private keys lying in an encfs, reads instantly May 13 08:55:29 dear lord .-. May 13 08:58:11 oO May 13 08:59:59 n900 allows for intermittent operation from -40C to +85C May 13 09:00:01 :) May 13 09:00:19 without the damage, but without guarantee to work May 13 09:00:33 uh May 13 09:07:05 someone asked for schematics: http://www.cpkb.org/wiki/Nokia_N900_service_manual_and_schematic_download May 13 09:09:33 i think many things about n900 could be added to maemo.org wiki May 13 09:09:47 http://www.cpkb.org/wiki/Nokia_N900_pinout for example May 13 09:10:04 true May 13 09:10:06 yeah May 13 09:10:20 and added as in copied, as the page could go missing May 13 09:11:17 http://www.cpkb.org/wiki/Nokia_N900_RX-51_firmware_data_package_download May 13 09:11:18 hehe May 13 09:12:15 also, i wonder what is the legal status of the service manuals May 13 09:12:27 since n900 is abandoned May 13 09:13:05 microsoft will sue and see you in hell May 13 09:13:13 nothing is "abandoned" from the legal view May 13 09:15:18 http://plan9.stanleylieber.com/hardware/ nice collection May 13 09:15:46 * L29Ah found schematics for his thinkpads May 13 09:16:15 the plan9 port for n900 neither really started :) May 13 09:16:58 it'd take some dev-years to do :/ May 13 09:17:39 probably not that much, if plan9 runs on some arm board May 13 09:17:52 I think it does May 13 09:18:22 http://plan9.bell-labs.com/sources/plan9/sys/src/9/omap/ May 13 09:21:19 i wonder how hard it'd be to make highpass in dsp May 13 09:22:40 i guess you could find code examples May 13 09:23:08 and implement it to maemo May 13 09:25:01 http://talk.maemo.org/showpost.php?p=1326668&postcount=182 May 13 09:25:02 hmm May 13 09:25:24 i wonder. May 13 09:25:54 ah i guess nevermind as one won't play music thru speakers anyway May 13 09:26:21 :D May 13 09:26:37 some do May 13 09:27:45 L29Ah: what is wrong with using ssh-agent? May 13 09:28:39 l29ah, i guess you could try using gstreamer as an highpass filter May 13 09:28:55 zGrr: nothing May 13 09:29:18 i wonder if i should use gst+alsa instead of pa May 13 09:29:25 and what would be the outcome May 13 09:30:11 okay there are speakers on ebay May 13 09:30:24 need to make the phone work thru alsa tho May 13 09:31:10 l29ah: try setting ~/.aoconfig May 13 09:31:28 i don't have such file May 13 09:32:16 let me check May 13 09:32:33 https://xiph.org/ao/doc/config.html May 13 09:32:47 oh it uses libao? nice May 13 09:32:47 echo default_driver=alsa >~/.libao May 13 09:33:12 but i dont remember if n900's libao has alsa backend compiled in May 13 09:34:49 yeah, no alsa backend May 13 09:34:52 ;[ May 13 09:34:54 only pulse May 13 09:35:21 maybe i should recompile pa and just share the .so May 13 09:35:53 also, remember it will block the audio device May 13 09:36:03 dmix! May 13 09:36:18 and some hook to get rid of dmix on calls May 13 09:36:23 hmm May 13 09:36:55 needs exeprimenting May 13 09:37:07 or redirect calls to alsa May 13 09:38:33 okay, put it on my list after playing w/ modern kernel May 13 09:38:46 l29ah: check that libao config, maybe alsa is built in May 13 09:39:01 what config? May 13 09:39:09 echo default_driver=alsa >~/.libao May 13 09:39:16 and if oscp plays anything May 13 09:39:27 if it doesnt just echo default_driver=pulse >~/.libao May 13 09:40:14 where is the source of http://repository.maemo.org/extras/pool/fremantle/free/liba/libao0/ ? May 13 09:41:34 it does but still uses pa May 13 09:41:41 Hi guys. I've got an issue with my N900, which suddenly refuses to boot. When I turn it on, I have the brief white Nokia splash, blue let stays on, and screen turns black (but on) for like 10 sec, and then everything is off. It does not reboot. May 13 09:41:43 hrm May 13 09:41:54 Last time I used it was yesterday, it ran out of battery May 13 09:41:58 kabouik: battery dead May 13 09:42:11 I charged it using an external charger this morning, sounds weird May 13 09:42:11 do not turn it on, let it charge till green May 13 09:42:13 aha! /etc/asound.conf specifies pa output May 13 09:42:18 O.o May 13 09:42:23 i'll try killing it May 13 09:42:43 i wonder if that would make phoneapp output to alsa directly May 13 09:42:45 :) May 13 09:43:11 I'll try another battery tonight KotCzarny, but my external charger said the battery was full this morning May 13 09:43:43 kabouik: battery could be dead anyway, how old is it? May 13 09:43:53 And when I plug the N900 to my laptop here, it won't charge (no breathing led). The screen is turned on (very dim), and then it goes into reboot loop May 13 09:44:17 still PA :/ May 13 09:44:25 Well it's old KotCzarny! But it did a few cycles the past week with no issues (I've not been using my N900 since a long time) May 13 09:44:59 L29Ah: you can try my old oscp version that was compiled with alsa direclty May 13 09:45:28 what do i press? May 13 09:45:46 i should look where i put it away May 13 09:46:03 or just recompile it apparently May 13 09:47:00 Anyway I can force charge with just a usb cable? It just goes into reboot loop and I have no charging led, and no other batteries or external charger at work May 13 09:47:07 Any way* May 13 09:47:59 Kabouik_: remove battery, wait 5s, insert it, connect charger and wait (without turning on) May 13 09:48:00 I have Backupmenu installed, bit it complains about watchdog not found and then turns off (this N900 unit was bought used and I have not done a clean install, it seemed to be in a good shape software-side) May 13 09:48:11 if it gets to pulsing yellow, good May 13 09:48:46 Nope May 13 09:48:47 and its good to have always reflash to factory image when getting used electronics May 13 09:48:52 wait ~30-60s May 13 09:49:03 I guess the battery needs a minimal charge level to initialize charge mode May 13 09:49:20 someone said failsafe charging kicks in after 30s May 13 09:49:30 Yes but it's already in reboot loop May 13 09:49:51 try to reflash then, assuming you dont have anything important on it May 13 09:50:25 I'll try another battery tonight first, because I don't see what would have cause that. I've not installed anything yesterday, it just ran ouf of battery during a call May 13 09:51:22 I thought there might be a 'force charge' mode in Backupmenu, but not sure, and anyway I can't start Backupmenu because of this watchdog not found May 13 09:57:48 n900 makes funky sounds when it's low on battery May 13 10:01:21 does it? May 13 10:01:46 l29ah: check if that one works https://sourceforge.net/projects/oscp/files/tmp/ May 13 10:02:04 recompiled to use alsa May 13 10:07:29 ./oscp-n900-alsa --help May 13 10:07:29 * Unrecognized option: --help May 13 10:07:29 [daemonize enabled, disabling ncurses ui] May 13 10:07:31 this sucks May 13 10:07:50 -h May 13 10:08:00 -d 0 to enable ncursesz May 13 10:08:12 killall oscp first tho May 13 10:08:53 ncurses crashes May 13 10:08:55 i did May 13 10:09:18 crashes as in segfault? May 13 10:09:25 oscp-n900-alsa: simple.c:398: snd_mixer_selem_get_playback_volume: Assertion `elem' failed. May 13 10:09:33 ;[ May 13 10:09:34 h,, May 13 10:09:36 hmm May 13 10:11:26 oh well, /me goes to find his n900 May 13 10:50:56 hrm May 13 10:51:17 i remember having working alsa version, now it hangs during card selection May 13 11:04:28 https://wiki.maemo.org/LED_patterns#Lysti_Format_Engine_Patterns_and_Commands sick stuff May 13 11:05:37 http://talk.maemo.org/showthread.php?t=77695 re 720p vidyas May 13 11:06:08 oh it's freemangordon (: May 13 11:10:05 stock player doesn't stop on unplug ;[ May 13 11:12:03 l29ah: check open media player May 13 11:12:21 it's stock player replacement May 13 11:15:19 Resolving downloads.maemo.nokia.com... failed: Name or service not known. May 13 11:15:20 mmkay May 13 11:15:32 nokia is no more, replace it with some mirror May 13 11:15:58 i guess in need to repack those .deb's or smth May 13 11:16:02 s/in/I/ May 13 11:16:02 L29Ah meant: i guess I need to repack those .deb's or smth May 13 11:16:11 no, you dont need to uninstall i think May 13 11:16:50 huh May 13 11:17:01 http://talk.maemo.org/showthread.php?t=77695 i'm talking about these 1kb debs May 13 11:17:09 ahm May 13 11:18:03 files are attached May 13 11:18:18 so why do you need nokia repos? May 13 11:18:21 and they try to download some stuff when dpkg -i May 13 11:18:47 O.o May 13 11:18:49 wut? May 13 11:19:12 ahm, maybe they have some install scripts May 13 11:19:55 ye May 13 11:21:58 nvm i've read the forum thread May 13 11:29:03 * KotCzarny ran alsamixer May 13 11:29:09 so much stuff May 13 11:29:10 O.o May 13 11:29:23 alsamixer on n900 ? May 13 11:29:28 sure, why not? May 13 11:29:51 alsa is a backend for pulseaudio anyway May 13 11:30:22 all I see is pulseaudio :) May 13 11:30:31 (alsa uses the PA backend by default here) May 13 11:31:00 hmm May 13 11:31:54 i disabled /etc/asound.conf beforehand May 13 11:32:07 and kernel knows nothing about pulseaudio May 13 11:32:12 but alsa is working May 13 11:32:34 so you see alsa going to pulseaudio which goes to barebones alsa May 13 11:33:33 :)) May 13 11:33:40 whoa adc hpf May 13 11:33:41 I know this is madness May 13 11:34:05 why we need pa when hpf is done by the audio codec? :3 May 13 11:34:11 maybe its broken May 13 11:34:30 see xprotect May 13 11:34:40 and the few related PA modules May 13 11:34:48 what's a xprotect? May 13 11:35:04 google is unhelpful May 13 11:35:14 hmm, maybe it's not "xprotect" May 13 11:35:23 xprot* May 13 11:36:13 http://wiki.maemo.org/Porting/Audio/Q_and_A_RE-PA May 13 11:36:17 http://talk.maemo.org/showthread.php?t=91325 May 13 11:36:20 i found this one May 13 11:36:58 yeah May 13 11:38:09 but for headphones only using direct alsa is something to consider May 13 11:39:10 i'd love to see any reports of actual damage May 13 11:39:22 as i'm going to run plain alsa May 13 11:39:26 be the first to check it May 13 11:39:50 i think i'm not first May 13 11:39:58 considering ppl running their debians on n900 May 13 11:40:12 what I'd like to know is whether you get a noticeable quality difference May 13 11:40:17 but they go through emulated alsa May 13 11:40:23 which goes to pa May 13 11:40:37 because they dont run full debian but chrooted one May 13 11:40:48 there's no emulated alsa if they're chrooted or boot debian instead of maemo May 13 11:40:54 bencoh: you get back 30% of cpu May 13 11:40:57 for starters May 13 11:40:59 as the emulation goes thru alsalib May 13 11:41:09 ie 100-150mA power back May 13 11:41:24 err, if you have some high-quality audio sample (and can actually hear the difference between highquality and garbage music) May 13 11:41:27 mW i'd say May 13 11:41:37 L29Ah: most install prepared debian chroot, which might /etc/asound.conf predefined to use pa May 13 11:41:50 L29Ah: im talking about your results May 13 11:42:01 ah, right May 13 11:42:03 mW May 13 11:42:06 my bad May 13 11:43:11 getting used switching fingertips to fingernails going from android phone to n900 May 13 11:46:58 whoa 230mW on openmediaplayer on headphones May 13 11:47:56 fingertip works well enough May 13 11:47:56 remember you played with power profiles May 13 11:48:04 set it back to 900 to have consistent results May 13 11:48:10 dear lord May 13 11:48:15 why do you need to overclock May 13 11:48:15 or redo all tests May 13 11:48:29 ye i have it set at 900 max May 13 11:49:04 and that also answers your earlier question if dsp is needed May 13 11:49:19 still, 230 is close to 273 May 13 11:49:27 L29Ah: fingernails <3 May 13 11:49:35 (do use a screen protector though) May 13 11:49:49 +1 for screenprotector May 13 11:50:06 which reminds me ... I cant find any in EU on amazon May 13 11:50:14 out of stock afaict :( May 13 11:50:25 buy cheapest one and use advanced tool called scissors May 13 11:50:32 or sharp knife May 13 11:50:44 i'd use a transparent household tape May 13 11:50:53 hmm May 13 11:50:58 tape would need glue residues May 13 11:51:03 s/need/leave/ May 13 11:51:03 KotCzarny meant: tape would leave glue residues May 13 11:51:20 i have isopropanol to get rid of them May 13 11:51:22 well, modern phone screens are allegedly wider, but ... May 13 11:51:38 tho it might not cover all the display :[ May 13 11:52:08 switched to 500MHz max, got 229 May 13 11:52:31 l29ah: check oscp too May 13 11:52:34 with lower freq May 13 11:52:42 k May 13 11:52:48 though i think it doesnt bump the freq over 250 on mp3s May 13 11:52:56 so max doesnt matter May 13 11:53:21 i remember on n800 using dsp ALWAYS bumped cpu to the max May 13 11:53:36 dont know about n900 May 13 11:54:20 as for not covering whole display, you mean 'faceplate' because display is standard 4" May 13 11:54:45 i don't remember the tape size and i'm in a forest atm May 13 11:55:27 282mW@500MHz May 13 11:55:39 hehe, more than earlier 273? May 13 11:55:44 which was at 900 May 13 11:55:55 L29Ah: get away from irc and try and speak to the forest animals, then ... you're being rude to them ;p May 13 11:55:56 doesn't seem to bump freq at all May 13 11:56:09 and i have 250 at the lower end because ppl say phone stops working @125 May 13 11:56:20 bencoh: they troll me May 13 11:56:24 125mhz isn't a good idea May 13 11:56:25 haha May 13 11:56:31 I'm sure they dont like systemd May 13 11:56:41 pulseaudio May 13 11:56:56 phone doesnt stop at 125mhz, but it's freaking slow until it jumps at max (with ondemand) May 13 11:57:21 so should i put 125 as a min freq? May 13 11:57:26 definitely no May 13 11:57:27 for ondemand May 13 11:57:34 why? May 13 11:57:46 becase you'd stay at max most of the time May 13 11:58:01 no intermediate freqs? May 13 11:58:15 huh May 13 11:58:18 i don't get it May 13 11:58:21 and it would still feel slow when changing app if you were just reading (for instance) May 13 11:58:31 KotCzarny: not with ondemand, no May 13 11:58:50 not that I know of May 13 11:59:33 okay i just set it and see myself :3 May 13 11:59:36 l29ah: it would jump between 125 and 800 May 13 11:59:57 and if 125 be too slow it would stay at 800 most of the time May 13 12:00:08 somehow it doesn't go lower ;[ May 13 12:00:12 KotCzarny: no May 13 12:00:26 fire powertop and check out freq stats May 13 12:00:39 just cat /sys/something/curspeed May 13 12:00:52 Nokia-N900:/sys/devices/system/cpu/cpu0/cpufreq# cat scaling_cur_freq May 13 12:00:52 250000 May 13 12:00:52 Nokia-N900:/sys/devices/system/cpu/cpu0/cpufreq# cat scaling_min_freq May 13 12:00:52 125000 May 13 12:00:59 wtf it won't lower May 13 12:00:59 uname -a ? May 13 12:01:02 hmm iirc there's something preventing it from running at 125 in some case but I forgot which May 13 12:01:04 and available freqs? May 13 12:01:06 2.6.28.10-power53 May 13 12:01:17 1150000 1100000 1000000 950000 900000 850000 805000 720000 600000 550000 500000 250000 125000 May 13 12:03:47 L29Ah: have you changed the ondemand/ attributes ? May 13 12:04:11 ondemand/avoid_frequencies aha May 13 12:06:52 powertop says shit ;[ May 13 12:06:56 kernel's too old i guess May 13 12:07:55 music skips in oscp sometimes May 13 12:08:14 at 125? May 13 12:08:18 ye May 13 12:08:27 285mW May 13 12:08:31 could be too much with pa May 13 12:08:41 hacking at pure alsa May 13 12:08:55 or it keeps jumping back and forth May 13 12:09:00 from min to max May 13 12:09:01 ye top says cpu is 70% loaded May 13 12:09:15 could be interrupt starving too May 13 12:10:28 ao_alsa WARNING: Unable to open surround playback. Trying default device... May 13 12:10:29 hrm May 13 12:10:33 the heck? May 13 12:11:16 no 5.1 sound for u May 13 12:12:01 i remember my 200mhz vortex86 machine skipped on some loads with 100hz kernel May 13 12:12:05 bumping to 300 fixed it May 13 12:12:38 openmediaplayer is 232mW and doesn't skip May 13 12:13:34 decoding happens in dsp May 13 12:14:55 https://zfh.so/hell/0a41c3ced24213f10198552ccc13ce3ce0ab7e01.jpg :3 May 13 12:15:19 whoa 45mW idling @125MHz May 13 12:15:25 i'm buying it May 13 12:15:31 :) May 13 12:15:57 idling ... screen off and locked ? May 13 12:16:17 ye May 13 12:16:25 it shouldnt have any influence on it, unless you have some app periodically waking up the cpu May 13 12:16:37 (or widgets) May 13 12:16:44 i have no widgets May 13 12:16:50 but i have ssh over wifi May 13 12:16:56 oh, okay :) May 13 12:17:22 try without it May 13 12:21:51 34mW May 13 12:23:52 battery eye shows lowest usage at 8.6mA May 13 12:24:04 with 3.7V its ~31.8mW May 13 12:30:56 im too stupid to create working asound.conf May 13 12:32:59 empty one should work May 13 12:33:58 yeah, works, but no sound May 13 12:35:57 did you kill pa? May 13 12:36:00 hmm May 13 12:36:03 good catch May 13 12:36:10 though it would say device busy May 13 12:36:29 gentoo page says alsa works out of the box May 13 12:36:40 might be some mixer setting then May 13 12:36:46 KotCzarny: isnt alsa supposed to be able to mix between multiple sources ? May 13 12:36:54 why should you get "device busy" ? May 13 12:36:57 bencoh: if you enable software mixing May 13 12:37:03 oh, true May 13 12:37:07 otherwise n900 offers only 1 playback stream May 13 12:37:28 normal cards that do hw mixing offer multiple playback streams May 13 12:42:08 aplay doesnt work neither May 13 12:42:22 -n May 13 12:42:35 Try `aplay --help' for more information. May 13 12:43:02 https://bpaste.net/show/249470588b01 my amixer state May 13 12:43:11 hm, wait May 13 12:43:49 -bash: diff: command not found May 13 12:43:49 ffs May 13 12:44:13 aha May 13 12:44:25 ‰ diff a b May 13 12:44:25 84,85c84,85 May 13 12:44:25 < Front Left: Playback 0 [0%] [-99999.99dB] [on] May 13 12:44:25 < Front Right: Playback 0 [0%] [-99999.99dB] [on] May 13 12:44:25 --- May 13 12:44:25 > Front Left: Playback 117 [99%] [0.50dB] [on] May 13 12:44:25 > Front Right: Playback 117 [99%] [0.50dB] [on] May 13 12:44:34 on speakers May 13 12:44:59 gets enabled when you run sound playing thing May 13 12:45:28 right, wonder i dont see it in alsamixer May 13 12:45:30 https://bpaste.net/show/1f2c664cef4d on headphones May 13 12:46:06 but May 13 12:46:12 which one it is? May 13 12:46:19 please diff -u3 May 13 12:46:34 or even -u6 May 13 12:46:38 second paste didn't help" May 13 12:46:39 ? May 13 12:46:52 it's against silence May 13 12:47:23 context cut a bit May 13 12:47:24 https://bpaste.net/show/ade73c0ca8c8 u100 May 13 12:47:37 much better May 13 12:49:51 ok, i have sound now, even if i didnt change anything May 13 12:49:52 O.o May 13 12:49:58 :D May 13 12:50:16 argh, wrong binary May 13 12:50:24 did it run pa? May 13 12:50:33 yeah, through pa May 13 12:50:47 kill the damn thing May 13 12:50:48 reupped alsa enabled oscp May 13 12:50:58 it must be messing w/ mixer May 13 12:51:01 https://sourceforge.net/projects/oscp/files/tmp/ May 13 12:51:02 here May 13 12:51:41 how do i stop it? May 13 12:52:24 btw you can save and load mixer states w/ alsactl May 13 12:52:26 killall? May 13 12:52:33 it gets restarted May 13 12:53:01 hmm, it's owned by init May 13 12:53:15 mv it :3 May 13 12:55:41 and risk infinite loop? May 13 12:56:08 ye why not? May 13 12:56:13 might be hardcoded into some binary May 13 12:56:48 i just hope it wont enter reboot loop after i disabled it May 13 12:56:49 hehe May 13 12:57:48 KotCzarny: with "stop" May 13 12:57:55 (upstart stuff) May 13 12:58:02 bencoh magic word/path? May 13 12:58:08 "stop" May 13 12:58:14 sto ;) May 13 12:58:16 you mean stop pulseaudio ? May 13 12:58:21 probably May 13 12:58:47 find /etc/event* -name "*pulse*" to make sure ;) May 13 12:58:48 anyway, stopped with moving binary May 13 12:58:54 meh May 13 12:58:57 that's crude May 13 12:59:07 no cpu hogs in top? May 13 12:59:29 apparently not May 13 13:00:02 but most likely if someone calls me it would fail May 13 13:00:37 will reboot after testing May 13 13:01:43 just re`start` it May 13 13:05:03 interestingly headphone control maxes out at 100 (-5.40dB) May 13 13:06:01 lol May 13 13:06:07 switching Left DAC Mux produced sound May 13 13:08:15 hmm, i've an idea May 13 13:08:27 * L29Ah refreshes his asound.conf syntax knowledge May 13 13:09:14 pcm.!default { May 13 13:09:14 type plug May 13 13:09:14 slave.pcm "hw:0,1" May 13 13:09:14 } May 13 13:09:16 try this May 13 13:09:24 what does plug mean? May 13 13:09:37 means it just redirects stuff May 13 13:09:44 to a slave May 13 13:09:52 that is an existing alsa device May 13 13:10:06 anyway, grab oscp-n900-alsa and test the power pls May 13 13:10:29 im curious of the outcome May 13 13:10:32 eh May 13 13:10:38 did you get playback? May 13 13:10:42 yes May 13 13:10:44 k May 13 13:10:49 after switching Left DAC Mux May 13 13:11:55 plug resulted in same behavior May 13 13:12:14 beautiful, 8.7% cpu usage May 13 13:12:16 and no PA May 13 13:12:26 and cpu is at 250mhz May 13 13:12:27 :> May 13 13:13:20 Left DAC Mux needs to be set to DAC_L2 to produce a sound (same for right) May 13 13:13:41 also, remember not to play any loud clips May 13 13:13:57 as d-class amp might produce freqs that break your poor speakers May 13 13:14:26 same assertion thing May 13 13:14:46 whats your ~/.libao and /etc/asound.conf ? May 13 13:15:09 default_driver=alsa May 13 13:15:15 all commented out May 13 13:15:19 hmm May 13 13:15:48 maybe ~/.asoundrc ? May 13 13:16:00 no such thing May 13 13:23:34 hrm May 13 13:26:46 also, you dont need to kill pulseaudio, as long system is not playing anything audio device is free May 13 13:27:22 are you sure your /etc/asound.conf is all commented out? May 13 13:27:36 ye, and tried mving it, didn't change anything May 13 13:28:19 killall oscp ? May 13 13:29:06 nope May 13 13:29:12 and no oscp-related stuff in `ps aux May 13 13:29:13 ` May 13 13:29:28 i remember that playing with alsa on n900 failed, then suddenly it started working May 13 13:29:37 i run it as `TERM=xterm ./oscp-n900-alsa -d 0` May 13 13:30:10 are you sure you've uploaded a fresh thing? May 13 13:30:16 md5sum oscp-n900-alsa May 13 13:30:17 d582f677a5daab728b74acc02a689566 oscp-n900-alsa May 13 13:30:20 -rwxrwxr-x 1 1003 102 5290591 May 13 12:30 /opt/oscp/oscp May 13 13:30:27 d582f677a5daab728b74acc02a689566 /opt/oscp/oscp May 13 13:30:56 ah, right May 13 13:30:58 enable softvol May 13 13:31:03 -v 1 -V 90 May 13 13:33:05 okay it works May 13 13:34:54 headphones, max vol, 399mW May 13 13:34:59 O.o May 13 13:35:59 w8, it seems pa's max volume ≠ alsa max volume May 13 13:36:03 i'll go cmp May 13 13:37:29 screen off? May 13 13:39:13 ye May 13 13:39:20 wtf pa doesn't want to play anymore May 13 13:39:30 stop alsamixer and oscp May 13 13:39:46 also, change mixer values back May 13 13:41:06 anyway, packaged oscp doesnt use libao on n900, uses libpulseaudio May 13 13:41:27 it doesn't work at all, the progressbar doesn't move May 13 13:41:55 fixed by restarting pa May 13 13:41:57 ;[ May 13 13:42:06 :) May 13 13:43:19 444mW on pa May 13 13:43:27 it seems like ncurses ui sucks May 13 13:43:30 :) May 13 13:43:32 hehe May 13 13:43:45 btw. you can run remote on pc and control core on n900 May 13 13:44:23 added to todo to check why ncurses is so bad May 13 13:45:40 retested w/ pulseaudio's mixer state: 341mW May 13 13:45:46 O.o May 13 13:46:39 maybe somehow related to 44.1→48 conversion May 13 13:47:36 on the other hand, it doesn't skip @125MHz May 13 13:47:40 hehe May 13 13:47:43 it shouldnt May 13 13:47:51 it uses only ~10% of cpu at 250 May 13 13:47:56 CPU: 28.0% usr 8.3% sys 0.0% nice 63.0% idle 0.0% io 0.0% irq 0.5% softirq May 13 13:49:25 eh, i suck May 13 13:49:29 :) May 13 13:49:33 341mW is great May 13 13:49:40 retesting w/ gui May 13 13:51:04 or, better, w/o gui ran thru gui May 13 13:51:23 you can start core with -d 1 switch and control/reconnect from pc May 13 13:51:44 or run two cores on diff. ports and start/stop them when needed May 13 13:53:27 195mW May 13 13:53:30 :3 May 13 13:53:30 o.o May 13 13:55:07 also i'm not sure dmix is not running May 13 13:55:53 and oscp doesn't seem to like being run two times May 13 13:56:05 aplay: main:608: audio open error: Device or resource busy May 13 13:56:06 aha, no dmix May 13 13:56:52 in stop state device should be freed May 13 13:56:58 though im not 100% sure May 13 14:00:25 i must think out some multiplexing thing May 13 14:00:34 to be able to wake myself with a phone call May 13 14:00:55 install dmix? May 13 14:01:15 then make pa use dmixed alsa? May 13 14:01:41 wtf May 13 14:02:28 too lazy to compile it myself May 13 14:02:36 will do this in gentoo ;] May 13 14:02:51 glibc incompability May 13 14:03:06 you can try grabbing debian-armel one May 13 14:03:17 i'll try grabbing ofono thing May 13 14:03:43 anyway dunno if i get to do it at all, first i'll play w/ fresh kernel and find a sd card to experiment on May 13 14:05:05 I think you just want to toss maemo away ;p May 13 14:05:13 sure i want May 13 14:05:20 but it's too good May 13 14:05:21 aww, poor us May 13 14:05:40 so i do the mutually useful stuff first May 13 14:05:48 I think mostly everyone here thought of it at some point, but .... May 13 14:17:04 bencoh, maemo is nice, but has its annoyances May 13 14:17:16 the blobs May 13 14:17:36 that too, but i talk about media library based player and trackerd May 13 14:17:45 that's what i hate personally May 13 14:18:30 i shall make native xlib gui one day May 13 14:23:59 :)) May 13 14:26:23 make it use the dsp first :] May 13 14:26:46 that needs recent gcc :/ May 13 14:26:58 4.2 is too old and available 4.6 is unstable May 13 14:27:03 i don't see no problem May 13 14:27:06 take 4.9 May 13 14:27:09 it's stable May 13 14:27:10 there is 4.7 May 13 14:27:14 doesnt compile with eglibc2.5 May 13 14:27:19 ;[ May 13 14:27:21 go gentoo May 13 14:27:27 bencoh, how do i install it in sb? May 13 14:27:34 there is a page for that in wiki May 13 14:27:47 it's mostly about thumb, but you can use it without thumb iirc May 13 14:27:55 and why broken 4.6 is in repos but not 4.7? May 13 14:28:36 because from what I've heard, properly adding a compiler/toolchain to scratchbox is a pain May 13 14:28:43 hehe May 13 14:29:02 i wouldnt mind compiling it myself, but i've tried and it failed May 13 14:29:11 and nobody cares about on-device compiling, at least for the people contributing to repos :) May 13 14:29:31 so it wont go to extras/armel either May 13 14:29:41 no, sb uses repos May 13 14:29:53 and that's what i care about May 13 14:30:31 http://wiki.maemo.org/CSSU-thumb_toolchain_setup_%28gcc4.7.2-linaro%29 May 13 14:30:33 this one? May 13 14:30:36 yeah May 13 14:31:47 though it would only allow me to use neon asm for codecs May 13 14:31:58 still, would be quite a big improvement May 13 14:32:09 hm I think you need an up-to-date libgcc on device May 13 14:32:11 KotCzarny: no May 13 14:32:12 why don't you make use of gstreamer? May 13 14:32:25 L29Ah: because not all codecs are supported May 13 14:32:31 also, not all file formats May 13 14:32:43 it gives better asm/neon intrinsics/optims for C code May 13 14:32:56 if you check what file formats supported by oscp May 13 14:33:28 but intrisics are hardly efficient for audio/video codecs May 13 14:33:49 bencoh, i had to disable all neon asm in libav because 4.2 was too old May 13 14:34:01 and arm asm too May 13 14:34:13 what ? May 13 14:34:13 so it's quite pure c routines May 13 14:34:41 i have to pass --disable-asm to libav or it doesnt compile May 13 14:35:01 so you know, if 4.7 works out, oscp will be quite a lot faster May 13 14:38:39 KotCzarny: which version of ffmpeg ? that's strange May 13 14:38:40 why don't you make use of gstreamer? May 13 14:39:32 ffmpeg-0.5 (from extras .... yeah, it's freaking old) is explicitely build with --enable-neon May 13 14:39:35 built* May 13 14:57:12 bencoh: not ffmpeg, libav May 13 14:57:32 ffmpeg is a parasite living on libav May 13 14:57:45 L29Ah: oscp -H May 13 14:57:49 that's why May 13 14:58:01 I strongly believe it is the other way around, and all this is FUD May 13 14:58:05 nope May 13 14:58:08 but I dont really want to feed this shitty troll May 13 14:58:18 the core devs that actually worked on things split May 13 14:58:25 I know the story May 13 14:58:42 kk May 13 14:58:53 i'm convinced by libav side of the story tho May 13 14:59:31 I also know niedermayer is still working on ffmpeg, btw May 13 14:59:41 (among quite a few others) May 13 14:59:51 anyway, enough said about that May 13 15:00:13 yeah, anyway, gcc 4.2 is too old to compile recent ffmpeg/libav asm May 13 15:00:25 ~nokia May 13 15:00:25 nokia is, like, a cellphone company from Finland. May 13 15:00:27 hmm May 13 15:00:30 I'm quite surprised, but okay May 13 15:00:31 ~nokiarepos May 13 15:00:44 what was the magic word? May 13 15:00:47 ~repos May 13 15:00:52 what for ? May 13 15:00:57 nokia repos mirror May 13 15:01:06 I dont think there is one May 13 15:01:16 there is at least 2 May 13 15:01:31 I dont think there is a magic word* May 13 15:02:57 and which nokia mirrors do you use? May 13 15:05:18 l29ah: answer to fremantle1.3 not working: http://talk.maemo.org/showpost.php?p=1333227&postcount=44 May 13 15:05:49 KotCzarny: muarf.org May 13 15:06:39 mind pasting the lines? May 13 15:06:52 see TMO :) May 13 15:06:56 googling May 13 15:09:07 should i use mr0 and ssu too? May 13 15:09:09 or just apps? May 13 15:11:20 hey hi.. Suddenly i am not being able to visit tmo, it says my ip has been put in anti-spam list May 13 15:14:03 i hate debian, bencoh, mind pasting the lines? May 13 15:24:39 E: Couldn't find package nokia-binaries May 13 15:24:40 hrm May 13 15:26:30 http://talk.maemo.org/showpost.php?p=1467566&postcount=112 May 13 15:26:54 scratchbox May 13 15:26:57 oh May 13 15:26:59 ie. i cant use manager May 13 15:27:00 you dont need it May 13 15:27:09 Accept Maemo5 SDK EULA and copy the URL. Add nokia-binaries repository URL to /etc/apt/sources.list May 13 15:27:12 http://wiki.maemo.org/CSSU-thumb_toolchain_setup_%28gcc4.7.2-linaro%29 May 13 15:27:15 apparently i do May 13 15:27:21 unless i don't May 13 15:27:25 nokia-binaries is hosted on repository.maemo.org May 13 15:27:43 you dont need a nokia mirror :) May 13 15:27:52 maybe, but it says it cant find it May 13 15:28:14 deb http://repository.maemo.org/ fremantle/sdk free non-free May 13 15:28:14 deb-src http://repository.maemo.org/ fremantle/sdk free May 13 15:28:14 deb http://repository.maemo.org/ fremantle/tools free non-free May 13 15:28:14 deb-src http://repository.maemo.org/ fremantle/tools free May 13 15:28:21 that's what sb has May 13 15:29:37 http://repository.maemo.org/dists/fremantle/ token code is written there May 13 15:29:47 what's token code? May 13 15:29:48 add deb http://repository.maemo.org/ fremantle/ nokia-binaries May 13 15:30:07 can i copy it from other sb target? May 13 15:32:27 k, worked May 13 15:32:56 is that token identical for everyone? May 13 15:33:59 yes May 13 15:36:38 i wonder if i need cssu in sb May 13 15:37:00 no May 13 15:37:17 but you will have troubles with gcc-4.7 May 13 15:37:25 will i? May 13 15:37:26 because of libgcc iirc May 13 15:37:40 or something along that line May 13 15:37:44 first i want to check if it compiles May 13 15:37:55 then i would be hacking at libgcc May 13 15:38:42 hrm, cssu uses same libc or something incompatible with stock libc? May 13 15:38:53 ask android_808 on TMO how he's managed to build ffmpeg 2.6 May 13 15:39:10 dunno May 13 15:39:16 most likely he used newer gcc May 13 15:39:30 unless he is arm asm wizard May 13 15:39:40 libc should be compatible between stock and cssu May 13 15:48:09 eh May 13 15:51:32 libc too old May 13 15:51:34 fun May 13 15:51:58 egg and chicken May 13 15:55:39 The problem is connected to glibc-2.5, after upgrading to glibc-2.6 it went away. The bug is now fixed since gcc-4.3 was updated to depend on glibc-2.6. May 13 15:55:40 fun May 13 15:58:47 -fgnu89-inline to the rescue! May 13 16:00:24 w00t, it compiled May 13 16:01:01 i wonder if i can compile newer gcc that way May 13 16:01:10 i think the error was similar May 13 16:08:26 hmm, not much difference May 13 16:11:26 mh, what's the most recent debian that can be used with 2.6.28? May 13 16:11:37 at some point there's got to be glibc incompatibilities May 13 16:12:18 oh, lol, apparently my kernel is using 1150mhz May 13 16:16:50 or not May 13 16:17:13 even if its set as max, time_in_state shows that max used is 600 May 13 16:18:16 right, pmconfig sets max as 600 May 13 16:22:30 do i need to reboot after changing pmconfig or there is some command ? May 13 16:34:35 hmm May 13 16:34:43 smartreflex seriously lowered voltage May 13 16:34:48 cool May 13 16:34:57 smartreflex is cool May 13 16:35:12 what does sr_vdd1_dsp_boost do? May 13 16:35:40 http://talk.maemo.org/showpost.php?p=1157313&postcount=80 May 13 16:35:43 nvm, found it May 13 16:36:02 whoops May 13 16:40:37 1 is at 23 and 2 at 30 May 13 16:53:10 hmm, fremantle1.2, 401 auth required? May 13 16:53:14 (on muarf.org) May 13 16:54:32 for ovi, yes May 13 16:57:52 what's muarf.org May 13 16:57:56 and why does it use ssh-dss May 13 17:14:49 wtf May 13 17:15:05 app manager forces me to create backup on install of ssu updates May 13 17:15:14 yet its not proceeding with updates May 13 17:15:43 screw app manager May 13 17:15:58 use fapman, the manly package mangler May 13 17:15:59 ! May 13 17:16:10 fapman is a timebomb May 13 17:16:19 how so? May 13 17:16:24 ~fapman May 13 17:16:25 somebody said fapman was Faster Application Manager, a frontend for apt which uses own repositories catalog, and shouldn't be used to do system upgrades (like CSSU), or actually for anything since ~speedyHAM. It also does "apt-get autoremove" after every operation, by default. In short, it's been identified as source of system corruption and thus deprecated, or see ~hamvsfam May 13 17:16:47 sounds fun May 13 17:16:49 yeah May 13 17:16:51 ~hamvsfam May 13 17:16:52 hamvsfam is, like, https://mg.pov.lt/maemo-irclog/%23maemo.2013-10-28.log.html#t2013-10-28T10:44:33, or http://talk.maemo.org/showthread.php?t=93227 May 13 17:16:53 debian autoremove is evil May 13 17:17:00 i guess i just go apt-get May 13 17:17:23 ~speedyHAM May 13 17:17:24 hmm... speedyham is 30 times faster than HAM http://maemo.merlin1991.at/cssu/community-devel/pool/free/h/hildon-application-manager/hildon-application-manager_2.2.73-2_armel.deb May 13 17:18:28 yeah, apt for me May 13 17:20:27 no speedyham in repos May 13 17:20:35 you need to go to cssu May 13 17:20:48 ie. upgrade to cssu version May 13 17:21:00 wat May 13 17:21:11 ~cssu May 13 17:21:12 i heard cssu is http://wiki.maemo.org/Community_SSU, or (Community Seamless Software Update) May 13 17:22:48 in -testing May 13 17:23:01 will it work on stock? May 13 17:23:20 dunno May 13 17:24:58 i have the community repos on but no speedyham May 13 17:25:08 extras-testing May 13 17:25:13 Use the link above? May 13 17:25:26 i did May 13 17:25:35 extras-testing enabled as well May 13 17:25:39 not extras-testing! CSSU-Testing May 13 17:25:44 ahm May 13 17:25:57 deb http://repository.maemo.org/community-testing/ fremantle free non-free May 13 17:25:57 in my config May 13 17:26:14 anyway, apt-get updated kernel-power, reboot, here i come! May 13 17:26:31 huh? May 13 17:26:46 always nervous when firmware upgrading May 13 17:28:09 hrm May 13 17:28:17 booting longer than usual May 13 17:29:15 5 dots blinking, which boot stage is that? May 13 17:31:51 wiki for kernel-power doesn't specify neither patches that are on nor the source code ;[ May 13 17:33:53 drat, its not booting May 13 17:40:18 wth May 13 17:42:52 ok, where is flasher-3.5 for linux located? May 13 17:42:55 ~flasher May 13 17:42:55 [flasher] at http://www.jedge.com/n810/flasher/maemo_flasher-3.5_2.5.2.2.tar.gz (also .exe!), or http://www.chakra-project.org/ccr/packages.php?ID=5027 or generally http://www.google.com/search?q=maemo_flasher-3.5_2.5.2.2.tar.gz. HARMattan(N9): https://aur.archlinux.org/packages/fl/flasher-harmattan/flasher-harmattan.tar.gz, or -- list of filenames/md5sums: http://pastebin.com/sYKdNJSH, or http://galif.eu/nokia/ May 13 17:46:43 why flasher, there's 0xFFFF May 13 17:46:48 to boot it May 13 17:46:56 right now it's stuck on 5dots May 13 17:47:07 KotCzarny ← likes to run trojans on his box May 13 17:51:11 rescueos booted May 13 17:51:13 what now, hrm May 13 17:54:48 ControlPersist is evil May 13 17:55:00 it takes a lot of time for ssh to learn the connection is dead May 13 17:55:13 and i can't easily kill just the one i want killed May 13 17:55:54 you know ~. ssh kill sequence? May 13 17:56:24 yep May 13 17:56:25 and there is a ssh option to ping server and disconnect on x failures May 13 17:56:30 and it doesn't kill the master May 13 17:56:44 yep and it sucks when i just suspend for a short time May 13 18:00:28 suspending shouldnt kill the tcp session May 13 18:02:20 yep but we're talking killing on keep-alive ping failure May 13 18:24:53 l29ah: but you want two contradicting features May 13 18:25:10 one to detect broken session, second to keep alive even for long sleep durations May 13 18:25:58 also, ping is client feature, not server one, so if you suspend, it should work still May 13 18:27:37 anyway, any idea how to figure out why it's not booting? May 13 18:27:53 as there is no syslog etc May 13 18:28:19 i've updated kernel-power via apt-get then rebooted May 13 18:32:47 KotCzarny: try boot the vanilla kernel via flasher May 13 18:32:58 how do i do that? May 13 18:33:54 um.. haven't played with flasher for over a year now.. but if you `flasher-3.5 -h` it should show you the argument .. -l i think May 13 18:34:17 "load" May 13 18:34:38 yeah, but it still needs file on pc May 13 18:34:52 thought it can load something already on device May 13 18:35:19 download the vanilla kernel to pc? the deb is on the repos May 13 18:36:02 maybe i don't understand what you're saying May 13 18:41:33 and where is the initrd for stock kernel? May 13 18:43:22 hmm, i was suggesting for you to attempt to boot Maemo with the stock kernel May 13 18:43:40 yes, but stock kernel need initrd (afair) May 13 18:43:47 no May 13 18:43:54 ok, gonna try May 13 18:44:51 stuck at 'nokia' May 13 18:45:30 and rebooted apparently May 13 18:46:18 hmm .. maybe you no longer have 'stock' modules in your system. by the way, you got in this by just installing power kernel? May 13 18:46:21 ah, right wrong command line May 13 18:46:26 ah, ok May 13 18:46:29 can someone paste /proc/cmdline ? May 13 18:46:53 flasher takes care of that .. May 13 18:47:07 or are you booting some other OS? May 13 18:48:47 init=/sbin/preinit ubi.mtd=rootfs root=ubi0:rootfs rootfstype=ubifs rootflags=bulk_read,no_chk_data_crc rw console=ttyMTD,log console=tty0 snd-soc-rx51.hp_lim=42 snd-soc-tlv320aic3x.hp_dac_lim=6 May 13 19:05:41 anyone have a rough idea when n900 neo will be released May 13 19:05:52 no one :) May 13 19:06:16 I bought two brand new n900's today May 13 19:06:25 ready to install the boards :( May 13 19:06:47 nice "excess flood" flood May 13 19:07:43 are we talking months years? May 13 19:07:46 decades??? :( May 13 19:08:14 I need a !@#% You NSA phone May 13 19:14:28 l29ah: thx, sicelo: it hung at nokia screen without cmdline and i had to take out the battery May 13 19:15:03 brand new N900 can't exist, can they? May 13 19:15:10 can May 13 19:15:13 shop stock May 13 19:15:14 etc May 13 19:15:34 * Sicelo would be very skeptical May 13 19:15:51 so it booted fine with cmdline? May 13 19:15:55 dont know yet May 13 19:16:09 as a sidenote, it doesnt even turn on without battery May 13 19:16:11 sicelo its one from where i work we never sold it May 13 19:16:18 and it was in storage all this time May 13 19:16:21 want to see a picture? May 13 19:16:21 :) May 13 19:16:26 lol, of course it shouldn't turn on without battery May 13 19:17:05 someone claimed it booted without battery, just turning off after bme start May 13 19:17:41 heh May 13 19:18:01 i can only confirm it is possible to hotswap batteries May 13 19:18:22 hrm, still stuck at 5dots May 13 19:19:15 must be some rc breakage May 13 19:19:21 hrm May 13 19:19:31 rc means? May 13 19:19:40 system startup scripts May 13 19:19:46 ah May 13 19:20:15 enable fbcon May 13 19:20:19 how? May 13 19:20:51 use rescueos, mount maemo rootfs May 13 19:22:05 that's assuming you have framebuffer console compiled into kernel May 13 19:22:41 there's a 2.6.28 with fb compiled ... may also be in repos .. May 13 19:23:08 and all KP have that.. so can also just try with the KP you were installing May 13 19:23:16 k May 13 19:23:21 now, how do i enable it? May 13 19:25:57 let me check on my N900 where i had it May 13 19:30:07 lol, can't really find it now ;) May 13 19:30:15 in the meantime, http://talk.maemo.org/showthread.php?t=37420 May 13 19:30:20 useful stuff May 13 19:30:25 what i love about linux is it's verbosity, what i hate about all that cute-fluff is hiding info from users May 13 19:31:06 dmesg is obscene! May 13 19:31:18 sure, as long you can get to shell May 13 19:32:02 Open /sbin/preinit May 13 19:32:02 Go to the init_system() function. May 13 19:32:03 above of the "}" insert: modprobe fbcon. May 13 19:32:23 that's from Open /sbin/preinit May 13 19:32:24 Go to the init_system() function. May 13 19:32:41 that's from http://n900.quitesimple.org/crypto/encrypted_home/README ... rescueos guy May 13 19:33:35 enabled r&d, printed some info on nokia screen then went to 5dots May 13 19:34:24 try fbcon. if that don't help, then you're fuckt.. you'll have to guess where the problem could be May 13 19:35:01 i think i have to disable that 5dots animation too May 13 19:35:38 i don't remember that it can be disabled easily.. might be wrong May 13 19:36:11 gonna fix rescueos initrd a bit May 13 19:36:18 but that tomorrow May 13 19:36:51 one thing that's critically missing is tab key May 13 19:37:28 poke me when you find out how to screw the dots thing :] May 13 19:37:40 :) May 13 19:37:42 will do May 13 19:38:42 tmo mentions it's easy, via /etc/init.d/fb-progress.sh May 13 19:38:48 i've never tried that myself May 13 19:39:24 http://i57.tinypic.com/zuqq1l.jpg my n900 May 13 19:39:25 :/ May 13 19:39:31 mint May 13 19:40:08 for fucks sake, how to exit vi?! May 13 19:40:13 no escape key May 13 19:40:15 i wonder what's the `act dead` feature for May 13 19:40:15 :!q May 13 19:40:21 hehe May 13 19:40:26 without esc key May 13 19:40:31 :!q May 13 19:40:35 KotCzarny, :!q May 13 19:40:45 im in edit mode May 13 19:40:50 first press escape May 13 19:40:53 then try May 13 19:40:56 #maemo> without esc key May 13 19:41:00 ^[ May 13 19:41:03 ;h :( May 13 19:41:16 l29ah: no [ key on kb May 13 19:41:24 paste in toe ascii control for escape May 13 19:41:24 so cant press ctrl-[ May 13 19:41:30 in the* May 13 19:41:40 you're screwed May 13 19:41:43 a bit May 13 19:42:02 hmm May 13 19:42:06 act_dead is something to do with the state when the N900 is half-on.. when charging with the device switched off. as you know, the software must be 'up' to an extent May 13 19:42:06 ^C works in vim, try it May 13 19:42:42 i could be confusing act_dead with some other state however May 13 19:42:58 nah sounds adequate May 13 19:43:01 ok, shift-ctrl-, somehow exited edit mode May 13 19:43:10 now how to save file and quit? May 13 19:43:18 ZZ May 13 19:43:33 quickest way, if you know you typed everything correction May 13 19:43:34 'ZZ' is not implemented May 13 19:43:38 um May 13 19:43:38 :wq May 13 19:43:45 yeah May 13 19:43:47 thanks May 13 19:43:50 phew May 13 19:44:00 it's like escaping from dungeon monster May 13 19:44:09 now, to disable boot anim May 13 19:45:18 sno__: how much did it cost? May 13 19:45:48 its from work I just too it since its unsellable May 13 19:45:52 took* May 13 19:46:04 they're old phones not May 13 19:46:04 unsellable? huh? May 13 19:46:08 nobody wants to buy them here May 13 19:46:21 its all about these nsa spy phones not May 13 19:46:24 er iphone May 13 19:46:53 not = now May 13 19:47:07 we have a few of them lying around May 13 19:47:57 I haven't even seen a working prototype for neo 900 so who knows how this will pan out May 13 19:48:52 I wish it to be true and that one comes out soon but I won't hold my breath May 13 19:49:21 fricking 5dots May 13 19:49:32 disabled them but still showing May 13 19:49:33 do you have extra k€? May 13 19:49:51 hmm? May 13 19:50:34 I have access to about 8 of them and I can take them all for nothing May 13 19:50:44 since we don't sell phones much anymore only repair them May 13 19:50:50 we repair iphones May 13 19:50:52 i mean neo900 thing May 13 19:51:27 neo900 does it exist? May 13 19:51:34 no May 13 19:51:36 i want one May 13 19:51:38 take my money May 13 19:51:47 haha May 13 19:51:51 sno__: /j #neo900 :) May 13 19:51:57 project's still alive May 13 19:52:04 but they havent released/delivered yet May 13 19:53:57 sicelo, got fb runnig, but how do i safely disable that boot anim? May 13 19:55:10 which one ? May 13 19:55:17 the video "handshake" ? May 13 19:55:30 5dots May 13 19:55:38 dunno May 13 19:55:40 i have disable handshake long time ago May 13 19:55:44 *disabled May 13 19:57:25 hmm May 13 19:57:37 r&d says 'no initrd \o/' May 13 19:59:27 backing up 28GB will be pita May 13 20:00:02 why would you back that up? you can flash rootfs alone iirc May 13 20:00:29 I don't believe you can remove 5 dots, I have never heard of anyone being able to do so of the people that ask. May 13 20:00:43 O.o May 13 20:01:02 those 5dots are overriding fbconsole May 13 20:01:24 yeah .. that's why i eventually disabled fbcon anyway :) May 13 20:01:52 sicelo, but rootfs mounts ok in rescueos May 13 20:02:32 if you're careful, play with that /etc/init.d/fb-progress.sh for the 5 dots business, although it might not be much help itself May 13 20:04:36 hmm May 13 20:06:30 http://talk.maemo.org/showthread.php?t=89687 May 13 20:06:37 pretty interesting May 13 20:07:04 ~boot-process May 13 20:07:05 see boot-process May 13 20:07:18 er May 13 20:07:33 infobot: thx, that was ... yeah, really useful May 13 20:07:50 ~listkeys #maemo boot May 13 20:07:51 Factoid search of '#maemo boot' by key (2): #maemo boot-process ;; #maemo bootloop. May 13 20:08:01 ~test May 13 20:08:01 somebody said test was not funny May 13 20:08:11 :O May 13 20:08:22 ~boot-process May 13 20:08:23 see boot-process May 13 20:08:27 ~bootloop May 13 20:08:28 extra, extra, read all about it, bootloop is when your device has broken rootfilesystem, so during reboot it fails on some service startup or kernel module load and thus reboots. This *drains* battery! And you can't reflash to stop bootloop when battery is drained. Recharge your battery by other means before reflashing. E.g. using ~rescueOS. Or external charger or BL-5J compatible other device. May 13 20:08:29 hmm May 13 20:09:03 looks like we've got a pointer dereference loop ;p May 13 20:09:11 no, boot-process is http://talk.maemo.org/showthread.php?t=87959 May 13 20:09:28 infobot: no, boot-process is http://talk.maemo.org/showthread.php?t=87959 May 13 20:09:28 okay, sixwheeledbeast May 13 20:09:32 ~boot-process May 13 20:09:32 see boot-process May 13 20:09:33 ~slap May 13 20:09:34 * infobot thinks sixwheeledbeast should be slapped May 13 20:09:45 infobot: meh May 13 20:09:45 meh int. /meh/ [The Simpsons] An interjection expressing apathy, or mgeary's favorite phrase May 13 20:09:46 ~boot-process May 13 20:09:47 see boot-process May 13 20:09:53 grrrrrrr May 13 20:09:55 srsly :D May 13 20:10:27 infobot: no, #maemo boot-process is http://talk.maemo.org/showthread.php?t=87959 May 13 20:10:27 sixwheeledbeast: okay May 13 20:10:37 ~boot-process May 13 20:10:37 from memory, boot-process is http://talk.maemo.org/showthread.php?t=87959 May 13 20:10:39 Doc: knock some sense into your child bot May 13 20:10:52 \o/ May 13 20:10:57 channel issue May 13 20:11:12 swb: tyvm May 13 20:11:28 but sleepy time for me, gotta fight tomorrow May 13 20:11:56 KotCzarny: np May 13 20:16:05 Matchbox is started to display the 5 dots May 13 20:16:27 which means you're already in X11 May 13 20:16:40 yes, that's why i need to debug why it's not booting May 13 20:16:51 anyway, sleepy time May 13 20:16:54 I wonder why fmrx is started in ACTDEAD May 13 20:17:36 n900 cant act as a radio alarm clock ... right ? :] May 13 20:17:43 I mean, not if it's off (?) May 13 20:39:12 KotCzarny: (exit vi) :q! May 13 20:39:59 freemangordon: the issue was (howto) esc in fbcon (rescueOS I guess) May 13 20:40:32 (with no esc key) May 13 20:45:06 oh May 13 20:46:13 bencoh: BTW... May 13 20:46:31 look at the last 6-7 commits here https://github.com/community-ssu/clock-ui/commits/master May 13 20:46:45 yay May 13 20:46:55 then tell me again how bad Qt is :) May 13 20:47:13 if you still feel like :P May 13 20:47:37 no, it is not the car, but the device behind the wheel :) May 13 20:49:30 I still dont like the wheel ;) May 13 20:50:33 but ... yeah it's much better :) May 13 20:55:32 I meant the steering wheel, if it was not clear May 13 20:55:54 it was (clear) May 13 20:56:44 anyway, I still prefer Qt over GTK. mind you, I prefer C over C++ :). But I guess it is a matter of preference. May 13 20:58:07 the part I dont like about Qt mostly related to C++ to be honest May 13 20:58:12 is* May 13 20:59:32 bencoh: well, C++ is not *that* bad May 13 20:59:36 imo May 13 20:59:53 not to say it is waaay better then the alternatives May 13 21:00:13 shit that has GC and such May 13 21:01:15 there are times C++ allows you to make better SW than C May 13 21:01:33 not most of the times but still May 13 21:03:19 however, if I had looked at adeclock sources by the time he (and not only) to be included in CSSU, I would have done everything I can to prevent it entering CSSU in this form May 13 21:03:38 s/to/wanted to May 13 21:03:58 wont' make that mistake again May 13 21:05:42 :-) May 13 21:06:29 the old/current version looks like it defeats every C++/qt purpose May 13 21:06:37 feels like reading bad gtk code, but in C++ May 13 21:06:52 * L29Ah ♥ shit w/ GC May 13 21:07:27 proper memory management of a complex app is quite funky May 13 21:07:55 L29Ah: yeah, your math implies it :P May 13 21:08:02 ;[ May 13 21:08:32 bencoh: no, really, I've never seen such a bad code. honestly. May 13 21:09:15 even the php code produced by our office assistant looks better most of the times May 13 21:09:20 eww May 13 21:09:26 that's almost insulting :D May 13 21:09:29 but she is not programmer, after all May 13 21:10:02 sorry if someone feels insulted, but... that is the truth as I see it May 13 21:10:31 I dont :) May 13 21:10:56 in every damn file we have a copy of those ~40 lines https://github.com/community-ssu/clock-ui/commit/37cf2aaeefd05e41c8c6211697f43cccf07433d8#diff-1b0d45f41668aa21495babdc75b80923L16 May 13 21:10:57 i do! May 13 21:11:00 translation stuff May 13 21:11:07 meet you in court May 13 21:11:08 heh, at least the coder seems to have mentioned that the code was horibble May 13 21:11:17 Sicelo: whare? May 13 21:11:21 *where May 13 21:11:34 there's a link sixwheeledbeast gave a few days ago.. May 13 21:11:36 oO May 13 21:11:42 could you share it? May 13 21:12:12 which coder? ade? as he is not the one to blame actually May 13 21:12:14 er, I forget now, but IIRC I highlighted you. May 13 21:12:17 when was the first day that you the bad code was 'exposed'? then we can look in those logs May 13 21:12:25 ah, sixwheeledbeast is here :) May 13 21:12:29 what's clock-ui? May 13 21:12:33 :nod: May 13 21:12:57 * L29Ah hides May 13 21:12:57 ade explained he was not happy with the code and wished it to be looked over. May 13 21:13:00 hmm, can't remember, some 2 weeks ago May 13 21:13:01 sixwheeledbeast: here ? May 13 21:13:10 less than 2 weeks :) May 13 21:13:11 oh, 2w ago, dont have that in backlog May 13 21:13:16 L29Ah: no need May 13 21:13:20 ~logs May 13 21:13:21 All conversations are logged to http://infobot.rikers.org/%23maemo/ Lines starting with spaces are not logged. Logs are updated daily. http://mg.pov.lt/maemo-irclog/ May 13 21:13:27 ~clock-ui May 13 21:13:55 right, we have logs May 13 21:13:58 L29Ah: I doubt that will work... May 13 21:14:25 anyway, noone takes you seriously :). Though you'll have problems with kerio, as he is our troll on duty for the last...how much? 3 years? May 13 21:14:30 ~test May 13 21:14:31 [test] not funny May 13 21:14:41 u fucking wot m8 May 13 21:14:49 see :) May 13 21:14:57 :)) May 13 21:15:24 > Lines starting with spaces are not logged. May 13 21:15:24 ;[ May 13 21:15:40 penis May 13 21:15:46 i hope there was a space May 13 21:15:49 L29Ah: it's like with bash history May 13 21:15:51 i definetely inserted a space May 13 21:15:59 my " ~test" was logged May 13 21:16:26 did it contain a space? May 13 21:16:33 it did May 13 21:16:58 not in logs though ... funny May 13 21:17:51 kerio: well, what I saw inserted in IRC is not space, but penis. Never thought you can do that with IRC. May 13 21:18:11 it's new in IRCv3 May 13 21:18:57 kerio: you'd better take care of him, a man never knows May 13 21:20:24 freemangordon: the link on IRC was http://talk.maemo.org/showpost.php?p=1385196&postcount=452 BTW May 13 21:20:49 I found it again, not via the logs; but internal memory ;) May 13 21:21:27 10x May 13 21:21:37 as I said, it is not ade to blame May 13 21:21:51 but we lazy CSSU guys May 13 21:22:18 sounds like he had a rough time May 13 21:24:04 BTW, after spending ~2weeks (and expecting to spend ~3 more) on that code, I'll never touch any SW made by cecipepez (or whatever was this guy's nick) with a 10m long pole. sorry if someone feels offended, but... May 13 21:24:41 :D May 13 21:24:49 bencoh: yeah, the reasong being adeclock inclusion in CSSU took too much time May 13 21:24:57 *reason May 13 21:25:08 whatsapp users are grateful to cepiperez, generally May 13 21:25:37 good for them, and good for me not using that shit (whatsapp) May 13 21:25:48 *for not May 13 21:26:27 * freemangordon feels his typing skills degrading as he is typing :) May 13 21:31:39 why whatsapp ? is he the original author of yappari ? May 13 21:32:17 oh right May 13 21:34:47 no, but he helped with certain parts of the code, especially during protocol changes May 13 22:25:50 hence my switch from filebox to hamsterfiler years ago. May 13 22:27:16 I am no coder but looking at the source and you can see what's going on. props to gidzzz for nice looking code I can understand. May 13 22:28:41 wait, what? doc's karma is higher than mine?!? May 13 22:29:03 hmm... I'll look into that :D May 13 22:31:11 sixwheeledbeast: you don;t need to be a coder to get that copying the same code all over the place doesn't make much sense May 13 22:33:08 but anyway, I took over it, so you may expect some improvement :). so far the binary is rediced to ~480k from 620k. and there is still lots of work to be done May 13 22:33:59 I expect the final app to be ~250k and startup time to be reduced <.5 seconds, when using smoke and mirrors May 13 22:34:16 and ~1.5s if not May 13 22:35:26 (smoke and mirrors) the stock app gets preloaded like modest, phone app, etc May 13 22:35:55 I am still undecided if we should have it like that May 13 23:25:03 hamsterfiler is pure love anyway May 13 23:25:13 nice, lean, fast May 13 23:25:52 not that I really use a graphical file manager, but ... it's really excellent at what it does May 13 23:39:54 KotCzarny: http://talk.maemo.org/showthread.php?t=95479 :] **** ENDING LOGGING AT Thu May 14 02:59:58 2015