**** BEGIN LOGGING AT Sat Nov 03 03:00:01 2012 Nov 03 03:01:38 well, closeish Nov 03 03:12:13 hah Nov 03 03:12:20 note the USB socket Nov 03 03:12:33 with through hole anchors Nov 03 03:13:21 Genius Nov 03 03:13:29 Too bad Nokia didn't think of it and patent it. Nov 03 04:51:59 the specs are better than the n900 Nov 03 04:52:02 wifi n, 8gb of nand Nov 03 05:44:13 O_O Nov 03 05:44:22 Wow, big channel Nov 03 05:45:22 Hello, I am a wanna-be N9 enthusiast. Nov 03 05:45:30 Can anyone help me? Nov 03 05:46:41 #harmattan Nov 03 05:47:02 Thanks Nov 03 06:47:35 hello, is there an app to let the interface of in-coming call shows caller's info (company, title, etc.) in contacts database? Nov 03 13:56:03 How do I enable the lockscreen? The thing which normally requires swiping from left to right to unlock? Nov 03 13:59:18 enable? Nov 03 13:59:51 It should pop-up if you press the powerbutton, if locked Nov 03 14:07:11 MrPingu: yes, but it doesn't. Nov 03 14:07:54 I also would like to replace the lockscreen with some other lockscreen, if possible. Nov 03 14:08:03 qtlockscreen? Nov 03 14:08:16 I tried the htc lockscreen, IIRC. Nov 03 14:08:21 But it didn't actually do anything. Nov 03 14:08:49 I can also install qtlockscreen, but I am not so much looking to configure anything, but more to use something premade. Nov 03 14:09:10 And for now, I am cracking the DES key on my phone :P Nov 03 14:09:32 IIRC, there are premade configs available Nov 03 14:10:10 My lockscreen doesn't show up on 1 press either, I always double or triple push the powerbutton Nov 03 14:13:14 Ok, that was easy to crack o.O Nov 03 14:20:33 MrPingu: I set QtLockscreen to 'sensation', but when I press the 'Lock screen and keys button', the screen goes black, but it's not locked. Nov 03 14:20:46 what do you expect of a 5-digit DES key? Nov 03 14:21:23 you could even brute force it manually, given you're a patient character Nov 03 14:22:01 fasta: That sounds weird Nov 03 14:22:38 Lock screen and keys does exactly that: blank screen and lock input Nov 03 14:22:46 it doesn't activate devicelock Nov 03 14:23:22 but yeah, you're not talking about device lock either Nov 03 14:23:46 so what do you mean by "but it's not locked"? Nov 03 14:23:58 DocScrutinizer05: so, how do I lock it in such a state that I see qtlockscreen? Nov 03 14:24:11 DocScrutinizer05: or any lock screen, like the default one? Nov 03 14:24:12 do you say your screen goes black but still takes touch events? Nov 03 14:24:26 I don't think it takes touch events. Nov 03 14:24:40 then it's exactly expected behaviour Nov 03 14:25:09 DocScrutinizer05: that's ok, but I still don't know how I get it to show this 'lock screen'. Nov 03 14:25:44 the 'lockscreen' with the slider is exclusively to avoid unlocking screen by accidentally preesing power button when screen blanked&locked Nov 03 14:26:35 so to see lockscreen you might lock device e.g. by locksliderswitch, then press powerbutton once Nov 03 14:26:49 DocScrutinizer05: what is locksliderswitch? Nov 03 14:27:00 the slider switch at side of device Nov 03 14:27:16 above headset receptacle Nov 03 14:27:30 DocScrutinizer05: excellent Nov 03 14:27:35 you slide it down to unlock/lock Nov 03 14:27:50 it wil NOT show the lockscreen on unlock Nov 03 14:28:11 lockscreen ONLY shows when pressing power button Nov 03 14:28:25 ...on locked device Nov 03 14:29:14 (at least that's how my devices are configured, I think you can tweak this behaviour in mce.ini) Nov 03 14:29:17 javispedro, did you see the 20% off all future MetaWatch products? Nov 03 14:29:32 yep Nov 03 14:29:36 javispedro: DUH!!! Alive :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) Nov 03 14:29:39 DocScrutinizer05: so, by default, it requires two user actions to activate. Nov 03 14:29:47 javispedro: HI pal! Nov 03 14:30:16 GeneralAntilles: but that does not include bastards like me who didn't buy the strata ;) Nov 03 14:30:27 fasta: lockscreen only shows on locked device Nov 03 14:30:39 DocScrutinizer05: yiiha :) Nov 03 14:30:46 it doesn't make sense to show it on an actively used device Nov 03 14:31:01 I'm interested to see what it is they come up with for 2011 buyers Nov 03 14:31:13 fasta: usually devices auto-lock after a few minutes Nov 03 14:31:20 I queried them on Twitter about it. Nov 03 14:31:58 DocScrutinizer05: how do I really lock it (i.e. require some kind of code) for it to unlock? Nov 03 14:31:59 javispedro, if I don't use a discount voucher you're welcome to it, though. Nov 03 14:32:12 Since sowatch changed my life. :D Nov 03 14:32:18 I will probably need to buy one in the future Nov 03 14:32:23 fasta: I use power button only to look up time Nov 03 14:32:24 fasta: the lockslider on the side, skips the lockscreen Nov 03 14:32:34 2 out of 4 pins are now rusted on mine :/ Nov 03 14:32:37 then immediately put back device to holster Nov 03 14:32:41 DocScrutinizer05: smae here :P Nov 03 14:32:46 *same Nov 03 14:33:24 javispedro: creepy voltage? Nov 03 14:33:29 javispedro, email Dave. He'll hook you up with a replacement. Nov 03 14:33:31 What I don't get is that both qtlockscreen and the default are active at the same time. Nov 03 14:33:34 javispedro: electrolytic stain? Nov 03 14:33:41 GeneralAntilles: we'll see, they'lll probably follow you on that 2011-buyers-benefit-program, we can't be that many. Nov 03 14:34:01 DocScrutinizer05: along with human sweat :) Nov 03 14:34:07 sure Nov 03 14:34:29 insane to keep voltage on sth that has contact to skin Nov 03 14:34:34 supposedly they were not able to tristate or similar the pins on this design Nov 03 14:34:55 will etch off your skin too Nov 03 14:35:36 one with Na+ the other with Cl- ;-P Nov 03 14:35:59 fortunately I don't notice anything because the pins are actually slightly protruded inwards Nov 03 14:36:08 but there are some scary pictures on the forums Nov 03 14:36:11 you're lucky Nov 03 14:36:25 I's ue them outa the water Nov 03 14:36:34 javispedro, seriously, though, if it stops charging contact Dave. Nov 03 14:36:39 never work again in my life (only for Americans) Nov 03 14:36:50 I'd sue* Nov 03 14:36:55 :) Nov 03 14:37:53 By the way, how much extra effort is it to support STRATA? Nov 03 14:38:30 javispedro: how's life? Long time no see :-/ Nov 03 14:38:55 fed up with maemo? Nov 03 14:39:07 did not had much time to play :( Nov 03 14:39:16 oh Nov 03 14:39:23 so, nothing new from my side Nov 03 14:39:51 GeneralAntilles: supposedly it will be the same Nov 03 14:39:51 already though we lost you to whatever other team Nov 03 14:39:56 thought* Nov 03 14:40:18 GeneralAntilles: tbh I am not happy with what metawatch is doing atm, with the surprise factor and all that Nov 03 14:40:35 GeneralAntilles: no firmware commits for months despite the fact they are surely rushing right now Nov 03 14:40:52 they'll probably throw it over the wall when the first watches arrive Nov 03 14:41:30 so, no idea what exactly they will use the new hardware (aka more RAM) for Nov 03 14:42:21 DocScrutinizer05: not yet Nov 03 14:42:24 :) Nov 03 14:43:10 good to hear, though it's a pity when your work is as demanding as mine regarding time left over for #maemo* Nov 03 14:43:22 GeneralAntilles: battery life with the latest firmwares has increased spectacularly on my watch; the strata is quite probably going to reach 2 weeks or so. Nov 03 14:43:36 Crazy Nov 03 14:46:04 fasta: That's why I don't bother with qtlockscreen Nov 03 14:46:20 DocScrutinizer05: more like #* :( Nov 03 14:46:21 Only eating extra memory Nov 03 14:46:35 javispedro: indeed Nov 03 14:47:00 javispedro: heard the latest bad news? Nov 03 14:47:24 javispedro: seems I have to do 6 months @ jail err council Nov 03 14:47:50 Haha Nov 03 14:48:59 DocScrutinizer05: you'll do better than some by not showing to work .... :/ Nov 03 14:49:00 * javispedro hides Nov 03 14:49:39 *by _just_ not showing to work, I mean. Nov 03 14:49:51 that's the plan ;-P Nov 03 14:49:54 Evil. Nov 03 14:50:13 I think that there's more pressing matters on the foundation side Nov 03 14:50:31 if nothing changes within 6 months there would be not community to "council" over. Nov 03 14:50:41 in fact my time available to assign to council topics is extremily limited Nov 03 14:50:53 Yeah, foundation's got its work cut out for. Nov 03 14:50:57 it' Nov 03 14:51:24 DocScrutinizer05, get somebody to update the freaking council wiki docs. Nov 03 14:51:57 first I'm awaiting my sentence Nov 03 14:52:25 right now it's still the leaving council in duty Nov 03 14:52:55 GeneralAntilles: what exactly needs update? Nov 03 14:53:07 iirc, didn't one of the "leaving council" members get a "get out of jail free" card? Nov 03 14:53:28 The list of hot topics hasn't been changed since I was on the council. :P Nov 03 14:53:43 lol, yeah, noticed that several times Nov 03 14:53:53 maybe there are no hot topics anymore Nov 03 14:54:19 just put a script there that dumps the hot topics from TMO Nov 03 14:54:27 eeew Nov 03 14:54:42 so that you get things like "My whatscrapp login is broken!!!!!!!!!!!!!!!!!!!one" Nov 03 14:54:55 indeed Nov 03 14:55:22 better tear down the whole wiki page Nov 03 14:55:36 s/better/rather/. Nov 03 14:55:56 nah, just remove the topics as soon as you raise to power Nov 03 14:56:32 btw, getting into business... Nov 03 14:56:48 DocScrutinizer05: do you have a link to the musb/mentor graphics datasheet? Nov 03 14:56:58 haha Nov 03 14:57:12 nope, I doubt there's *anything* available Nov 03 14:57:25 thought that :( Nov 03 14:57:27 meh Nov 03 14:58:05 all info about fsckng mg musb been from OMAP TRM Nov 03 14:58:16 back when I did h-e-n Nov 03 14:58:25 http://forum.meego.com/showpost.php?p=41826&postcount=72 Nov 03 14:58:28 ^^ saw that? Nov 03 14:58:41 plus some crappy press release level papers from MB Nov 03 14:58:44 MG Nov 03 14:59:12 DAFAQ fmg still active?? Nov 03 14:59:17 I know that you are not happy if it's OTG instead of real hostmode... but I suspect this might be easier to implement on the kernel. Nov 03 14:59:38 (and since I have at most a few hours to hack on it...) Nov 03 15:00:57 it actually might work OOTB Nov 03 15:01:18 at least with OTG enabled kernel, and OTG is sufficently hostmody Nov 03 15:01:19 it doesn't, but I'm trying to figure out the changes Nov 03 15:01:32 the ID pin comparator was gated, but I fixed that Nov 03 15:01:42 hey :-D Nov 03 15:02:00 trying to lure me into HARM tasks again? Nov 03 15:02:06 heehee :) Nov 03 15:02:36 I got it to advance to a_idle, just by plugging the modded cable Nov 03 15:02:52 but then neither vbus pump starts Nov 03 15:03:23 btw (re your fmc post) that's not exactly a "simple cable mod" since I haven't found a 5wire cable with micro-usb plug so far Nov 03 15:03:36 I have one :) Nov 03 15:03:41 duh! Nov 03 15:03:48 they're sold on ebay Nov 03 15:03:58 wow, where? Nov 03 15:04:05 they call those "samsung otg cables" Nov 03 15:04:12 mhmmmm Nov 03 15:04:16 seems that samsung is also not a big lover of the microa/b port duality Nov 03 15:04:30 so they just put microb ports and call it otg Nov 03 15:04:40 wonder why they usb guys didn't jump on them already Nov 03 15:04:42 haha Nov 03 15:04:44 *the usb guys Nov 03 15:05:20 maybe they don't print a OTG badge on their box? Nov 03 15:05:49 heh, probably Nov 03 15:06:10 AIUI USB-cert rules don't forbid checking ID pin on usb_b and acting accordingly Nov 03 15:07:03 hm Nov 03 15:07:48 they just say that official OTG devices need a usb_a(/b) receptacle Nov 03 15:08:02 well Nov 03 15:08:23 I know that the musb can be configured to NOT trigger the gaia/twl4030 vbus pump Nov 03 15:08:41 errr, yep Nov 03 15:08:45 probably Nov 03 15:08:51 we'll need to figure out how to do the entire vbus pump setup, with either your bq2xxxx scripts or using gaia Nov 03 15:09:08 needs kernel patches Nov 03 15:09:22 bq2xxxx certainly, but using gata might not need any change at all Nov 03 15:09:29 just check how bme reacts to that Nov 03 15:09:33 or a userland auxilary driver Nov 03 15:09:35 hopefully not by frying the device Nov 03 15:10:02 it WILL fry the device Nov 03 15:10:11 bme still seems to use twl4030 to do charger detect Nov 03 15:10:14 no kidding Nov 03 15:10:21 and IF twl4030 sense ID pin, then it NEVER reports charger detect Nov 03 15:10:31 not that I would test that theory on my device :/ Nov 03 15:10:33 oooh Nov 03 15:11:57 that'd be indeed a rather sane solution, almost as sane as when bme would switch bq24xxx from charging to boost mode on ID detect. Errr nope scratch that, will not play nice with OTG Nov 03 15:12:36 javispedro: pali did a kernel driver for bq24xxx Nov 03 15:12:46 based on my original draft Nov 03 15:12:52 aiui Nov 03 15:13:08 just integrating that with harmkernel is probably already out of my freetime slot :( Nov 03 15:13:13 DocScrutinizer05, yesterday I sent new version for upstreaming to LKML Nov 03 15:13:25 Pali: noticed that Nov 03 15:13:34 btw Nov 03 15:13:59 so, how long would it take to fry my device if shit happens (aka bme charges with bq2xxxx or gaia vbus pump active)? Nov 03 15:14:14 ok, I'm staying here for one more coffee, before I have to leave for RL Nov 03 15:14:25 (should've been gone 2h ago) Nov 03 15:14:29 haha Nov 03 15:14:59 DocScrutinizer05: feel free to leave, I know how vengeful RL can be... Nov 03 15:15:00 javispedro: (how long) usual overtemp issues, so between several seconds and never Nov 03 15:15:32 nah, much to excited to talk to you like first time this year Nov 03 15:15:42 because I'm thinking on testing and see what bme does Nov 03 15:16:31 javispedro: basically the problem is when twl4030 chargepump power gets eaten completely by bw24xxx trying to charge Nov 03 15:16:55 because I presume bme will disable bq2xxx pump, so the problem is only with the twl4030 one Nov 03 15:17:27 that's why I was trying to find out how I can disable the twl4030, which, according to docs, will turn on by itself on id sense Nov 03 15:17:28 possible outcome: twl4030 detects overcurrent and shuts down, bme detects bad powersource and stops charging, or twl4030 overheats due to overloading Nov 03 15:19:03 javispedro: for a start you should be fine with simply stopping bme and waiting til bw24xxx 32s timeout Nov 03 15:19:40 !! Nov 03 15:19:40 MUSB_ULPI_USE_EXTVBUS Nov 03 15:19:51 unless the chip is actually entering emergency chage mode then, when seeing VBUS (from twl), and draws so much power that twl chargepump goes down Nov 03 15:20:11 DocScrutinizer05: can it enter emergency charge without charge detect? Nov 03 15:20:45 The only sure way is to tell it not to charge. Nov 03 15:20:46 USE_EXTVBUS looks exactly what I wanted Nov 03 15:20:47 that's the idea, but for hw charger detect (d+- short) on N900 we got the 1707 Nov 03 15:20:57 since twl was broken re that Nov 03 15:21:13 (ext_vbus) :nod: Nov 03 15:21:13 without software control it reverts to eating 100mA on input side Nov 03 15:21:42 ShadowJK: yep, but only if Vbat < 3.7V Nov 03 15:22:49 ShadowJK: otherwise yep, to stop bq24xxx completely, you need to tell it to stay idle every 15s Nov 03 15:23:35 (coffee, bbl) Nov 03 15:31:16 dang, this could've been the first on-topic convo in #maemo-hostmode-discussion since _ages_ Nov 03 15:33:45 well, I guess some IRC channels simply die from obsolescence eventually Nov 03 15:34:36 hey, #meego still has some people around :) Nov 03 15:34:55 indeed, 170 Nov 03 15:36:05 and a real 5 lines of convo yesterday Nov 03 15:36:38 err, thursday Nov 03 15:36:44 yesterday nuttin Nov 03 15:38:05 javispedro: if you're interested - I still have a N9 here that's ready to test potentially dangerous stuff like that Nov 03 15:38:22 that's what this device been meant for Nov 03 15:38:31 meh, I'm testing on N950, that's how much I appreciate it =) Nov 03 15:38:41 hehehehe Nov 03 15:39:03 Pali: does the bq driver you posted on mailing list differ from the one you have on gitorious? Nov 03 15:39:11 you fool, better sell it for 1800GBP Nov 03 15:39:27 Skry, no, it is last version from gitorious Nov 03 15:39:49 DocScrutinizer05: for testing what? Nov 03 15:39:58 hostmode Nov 03 15:40:15 on n9/50? Nov 03 15:40:20 VBUS chargepunp/boost actually Nov 03 15:40:24 Pali: ok, thanks. Nov 03 15:40:24 yep Nov 03 15:40:34 well damn. Nov 03 15:41:25 Pali: is there any chance your policy file for jack bias could stop tv-out working? Nov 03 15:41:49 not completely impossible Nov 03 15:42:09 tv-out is same jack pin as mic(-bias) Nov 03 15:42:19 just to remember, rm-680 is n950 and rm-696 is n9? Nov 03 15:42:29 err Nov 03 15:42:30 google says so Nov 03 15:42:37 then yes ;-) Nov 03 15:42:48 * javispedro patches both Nov 03 15:43:09 vi__, my patch to policy file no, but you know that I disassembled binary file, so there can be problem Nov 03 15:44:00 curiously enough, the n9/50 kernel defines a 100mA power limit for boost Nov 03 15:44:02 but I tested new compiled file in swi prolog interpretor and it had all predicats Nov 03 15:44:05 what you say is "not my changes but the disassembly process might have introduced bug" Nov 03 15:44:10 right? Nov 03 15:44:11 hi. is someone using yappari? Nov 03 15:44:24 vi__, so I think that problem is not here... Nov 03 15:44:35 Pali: ok, thank you. Nov 03 15:44:51 vi__: KP? Nov 03 15:45:01 vi__: err CSSU? Nov 03 15:45:13 my problem is tv-out would not work when I hooked my n900 up to my projector. Nov 03 15:45:28 DocScrutinizer05, disassembling can intruduce bug in way that some predicate is missing Nov 03 15:45:28 vi__: you're using CSSU-T ? Nov 03 15:45:39 Pali: yep, got that Nov 03 15:45:40 but not that code of predicate was changed Nov 03 15:46:03 although my projector has strated becomming a bit flakey in its old age and I have not tried it on any other composite input. Nov 03 15:46:05 but I tested if new compiled version has same predicates as old nokia Nov 03 15:46:17 I have cssu-t and kernel-cssu3 Nov 03 15:46:18 and I also started ohm debugger to dump data Nov 03 15:46:21 vi__: you got me on your ignore list? Nov 03 15:46:23 *on device Nov 03 15:46:25 aah Nov 03 15:46:44 vi__: in cssu-t you need to twiddle a bit with advanced tv-out settings Nov 03 15:46:47 so I think disassembling was correct Nov 03 15:46:49 no, I am on the ol' n900 and typing stuff takes time! Nov 03 15:46:52 it doesn't work on 100% zoom Nov 03 15:47:05 default zoom is 90% Nov 03 15:47:15 also try to change to pal Nov 03 15:47:29 and click on enable button :-) Nov 03 15:47:29 well hal reported that video cable was inserted. Nov 03 15:47:55 dmesg said cable was inserted but nothing was coming out. Nov 03 15:48:01 vi__: also it seems you need to switch back and foce all settings of tv-out once, since the new applet doesn't sync automatically with old settings Nov 03 15:48:21 cssu applet listening to HAL for changes Nov 03 15:48:29 DocScrutinizer05: interesting. Nov 03 15:48:42 vi__: so: change every parameter in settings-TV and store. Then revert to original settings Nov 03 15:49:03 it check HAL property jack (there should be list where is video string) Nov 03 15:49:06 I must try it on other composite inputs before I can say for certain it is my n900. Nov 03 15:49:14 also try to reboot n900 Nov 03 15:49:18 Pali: yes, it said video. Nov 03 15:49:33 vi__, and is cssu applet visible in menu? Nov 03 15:49:36 status menu? Nov 03 15:49:52 it is visible only when cable is connected Nov 03 15:50:20 mine been visible but didn't work until I switched to disabled and then back to enabled Nov 03 15:50:29 I cannot recall, I do not think so. Nov 03 15:50:46 let me just go test right now. Nov 03 15:50:48 brb. Nov 03 15:51:24 vi__: also keep in mind you need to frist plug cable to TV, only then to N900 Nov 03 15:51:33 otherwise detection fails Nov 03 15:51:41 first* Nov 03 15:53:04 the hw is testing resistance from 3rd pin to GND. For TV plugged this should probably be >0 but I'd actually guess it's 50R nominal Nov 03 15:53:28 for TV Nov 03 15:53:40 and 2kR for headset mic Nov 03 15:54:05 with cable not plugged in to TV it's infinte (open circuit) Nov 03 15:54:19 with headphones it's zero Nov 03 15:54:27 aka shortcircuit Nov 03 15:55:08 well fsck me. Nov 03 15:55:12 it worked. Nov 03 15:55:18 hehe Nov 03 15:55:41 after redoing all the settings from the applet perhaps, I cannot be certain. Nov 03 15:55:57 i didnt test before. :/ Nov 03 15:56:00 toldya you need to redo settings Nov 03 15:56:18 DocScrutinizer05: what would happen if both bq2xxxx and gaia pump were to turn on? Nov 03 15:56:27 nothing bad Nov 03 15:56:53 increased battery drain Nov 03 15:56:57 negligible Nov 03 15:57:23 and more power available on VBUS Nov 03 15:58:48 inclear though what chips' logic (esp any possible softstart) would say when there's already full VBUS when boost about to start Nov 03 15:59:12 probably no impact either Nov 03 15:59:54 worst thing to happen was an error condition of either of the both chips Nov 03 16:00:04 like "overvoltage" or whatever Nov 03 16:01:00 OTG HNP though will fail miseably Nov 03 16:01:12 yeah... Nov 03 16:01:17 as will OTG suspend mode Nov 03 16:01:31 still stuck at a_idle, even with bq2xxx Nov 03 16:01:33 pump Nov 03 16:01:34 luckily we're not interested in either Nov 03 16:02:05 :-/ Nov 03 16:02:12 no auto ENUM then? Nov 03 16:02:35 you have set it to OTG in /proc or /sys ? Nov 03 16:02:40 (forgot where) Nov 03 16:02:55 since default state probably is "device" Nov 03 16:03:46 well, a_idle seems to indicate you did Nov 03 16:03:50 id pin sense did Nov 03 16:04:04 otherwise should be b_idle iirc, for "device" Nov 03 16:04:10 yep Nov 03 16:04:27 javispedro: nonono, it still needs right mode Nov 03 16:04:47 it won't change from 'device' to 'OTG' mode automagically Nov 03 16:05:09 echo OTG >/*/mode Nov 03 16:05:27 mode is what returns "a_idle" Nov 03 16:05:44 ummm Nov 03 16:05:45 the other one (that was on gaia's sysfs iirc on n900) is gone Nov 03 16:05:52 dang Nov 03 16:06:05 well, if it's not trivial Nov 03 16:06:12 let me enable debug output .. Nov 03 16:12:30 javispedro: echo OTG >/sys/devices/platform/musb_hdrc/mode Nov 03 16:12:48 iirc Nov 03 16:13:23 oh, it works Nov 03 16:13:28 I mean, the write Nov 03 16:13:35 the rest is still stuck at a_idle Nov 03 16:14:06 t900:~# echo hostf >/sys/devices/platform/musb_hdrc/mode Nov 03 16:14:08 t900:~# cat /sys/devices/platform/musb_hdrc/mode Nov 03 16:14:09 b_idle Nov 03 16:14:45 t900:~# echo OTG >/sys/devices/platform/musb_hdrc/mode Nov 03 16:14:47 t900:~# cat /sys/devices/platform/musb_hdrc/mode Nov 03 16:14:48 b_idle Nov 03 16:15:36 original valid parameters to echo been 'host', 'OTG', 'device' Nov 03 16:15:39 afaik Nov 03 16:15:43 lowercase Nov 03 16:16:04 :nod: Nov 03 16:16:25 coffee over Nov 03 16:16:53 no more excuse to procrastinate Nov 03 16:17:16 bbl o/ Nov 03 16:17:31 cya DocScrutinizer05 Nov 03 16:17:36 maybe next year ;) Nov 03 16:19:26 I seriously hope to not wait that long Nov 03 16:20:00 you might find a 15 min every other weekend Nov 03 16:24:21 host[h|f|l] and peripheral is on n900 Nov 03 16:27:21 how hard would it be to change the policy that governs playing audio while on a phone call? Nov 03 16:28:07 meh, removing the gadget driver tears down usb _host_ support, how stupid Nov 03 16:28:19 I used ionice to give rtcom-call-ui realtime priority. Is there any reason I should do that? Nov 03 16:33:39 vi__, first RE/disassemble libplayback library and alsaped daemon and then understand nokia pulseaudio, ohm, alsa, ... patches :-) Nov 03 16:34:23 javispedro, some usb gadget (e.g. g_file_storage) must be loaded for usb host mode Nov 03 17:29:15 Pali: let me put it another way. Nov 03 17:29:23 Pali: Il buy you a pizza. Nov 03 17:37:34 actually I think the missing docs and closed nature at large of audio handling on fremantle is one of Nokia's mortal sins Nov 03 17:38:22 since, dang, it's a *phone* and a *mediaplayer*, and also maybe a recorder Nov 03 17:38:24 well if anything superior voice quality is the "secret sauce" on any phone Nov 03 17:38:47 voice quality not related to policy handling Nov 03 17:39:21 merlin1991: what is this? 2001? no one uses 'phones' for 'phone-calls' anymore. Nov 03 17:39:23 not like if it's a secret thing at least Nov 03 17:40:37 there needs to be clear specs how any arbitrary audio app is supposed to either step back or take priority over e.g. inbound phonecall Nov 03 17:41:54 "we're doing that for you. You got game-audio-class, and system-audio-class - so smile and be happy" is kinda like telling devels "F U, don't dare to mess with 'our' audio" Nov 03 17:43:41 I want finegrained config options whether ringtone is supposed to get played back via BT, via wired headphone, via wired headset, via TV-out Nov 03 17:44:03 I want to decide whether inbound call stops camera, or camera stops inbound call Nov 03 17:44:29 or maybe inbound call just doesn't ring when camera active Nov 03 17:44:48 but still is signalling the call on display Nov 03 17:45:46 all that is not at all related to *quality* of call audio, but merely about policy management and handling Nov 03 17:47:06