**** BEGIN LOGGING AT Wed Jul 15 02:59:57 2009 Jul 15 03:23:23 shr-installer-->"what do you want to do TODAY"?!?! WTF!!!!!!! Remove that, it's highly dangerous for Neo, might break the device. I know a few guys (incl myself) who go nuts and stomp on the device whenever those redmondisms show up Jul 15 04:09:43 DocScrutinizer: what DO you want to do today? Jul 15 06:32:29 jc: Hi. Who is in charge of wiki.openmoko.org server now? Its ssl certificate expired. Jul 15 06:33:22 PaulFertser: gismo is the one maintaining the certificate. Jul 15 06:33:41 I will ping him today, and this should be fixed. Jul 15 06:34:05 jc: Thanks! :) Jul 15 06:35:26 PaulFertser: no problem. Jul 15 07:16:26 johnsu01: Hi :) Jul 15 07:26:21 hmm. mokonnect saying "could not find wifi in connmann devices".. Jul 15 07:26:38 iwconfig shows eth0 just fine Jul 15 07:27:18 FiXion_: editing wpa_supplicant.conf with vim works just fine ;) Jul 15 07:27:49 PaulFertser: is it suppose to be in /etc/wpa_supplicant ? Jul 15 07:28:50 FiXion_: it depends on how you start wpa_supplicant Jul 15 07:29:12 how do you start it? Jul 15 07:30:36 FiXion_: wpa_supplicant -Dwext -ieth0 -c/etc/wpa_supplicant/wpa_supplicant.conf usually Jul 15 07:31:46 do you send a dbus "enable wifi" first - or do you manually enable it? Jul 15 07:32:28 and don't you add a & in the end or a -B to the options? Jul 15 07:35:18 FiXion_: yes, i enable wifi first. Probably the best way is to do fsoraw -r WiFi wpa_supplicant ... Jul 15 07:35:59 FiXion_: and i usually keep it running in the terminal. Or send to background later... Most probably the way i do that is suboptimal. Jul 15 07:36:09 do you often get this: http://pastie.org/546511 Jul 15 07:36:17 often the wifi module dies - won't scan Jul 15 07:36:24 when using wpa_supplicant on shr (latest) Jul 15 07:36:48 then some magic combination of rmmod and modprobe ar600 and ifconfig eth0 down && up makes it scan again Jul 15 07:38:01 do you check for some state or do anything to make sure it works and doesn't "hit" the wifi at a bad time :) Jul 15 07:38:03 FiXion_: i don't get that often but yes, the module firmware sometimes goes into inoperable state, no known workaround except reloading the module :( Jul 15 07:38:45 PaulFertser: care to share your wifi "up" script? Jul 15 07:38:47 FiXion_: and i never do ifconfig eth0 up/down commands Jul 15 07:39:16 didn't need to this time. Jul 15 07:39:31 FiXion_: i'm afraid i type the command i just posted by hand or sometimes it's available from bash history. Jul 15 07:39:33 just saw an issue (wasn't shr) where I need to up/down it - before I could ifconfig it Jul 15 07:39:45 np then Jul 15 07:39:48 if it's just that line Jul 15 07:41:03 I guess you do an ifup afterwards - otherwise no one runs dhcp :) Jul 15 07:41:16 but it works like a charm with the fsoraw in front Jul 15 07:41:20 thanks a bundle Jul 15 07:41:22 FiXion_: i think fsoraw approach will work the best. If scan doesn't work, you kill wpa_supplicant, fso-raw releases resource, FSO unloads the module (well, unbinds in fact but it's the same), then you restart fsoraw ... wpa_supplicant and everthing works. Jul 15 07:41:41 FiXion_: haha, i run dhcpclient by hand too :D Jul 15 07:44:01 also fix routes in hand? Jul 15 07:44:23 mine stays at usb default gw Jul 15 07:44:27 kinda useless when on wifi :) Jul 15 07:44:46 and something else is spawning wpa_supplicant ..grr Jul 15 07:44:59 and I can't remove mofi apperantly :( Jul 15 07:45:05 task-shr-apps says it needs it Jul 15 07:45:22 hello there Jul 15 07:46:38 PaulFertser: any ideas what settings I need to disable/kick to have it NOT start wpa_supplicant by itself Jul 15 07:48:27 crap just crashed the neo Jul 15 07:48:37 doing fsoraw... with a ifup afterwards Jul 15 07:48:53 wpa was complaining about not being able to apscan etc. Jul 15 07:49:08 * FiXion_ adds an rmmod && modprobe ar6000 to the script :) Jul 15 07:53:56 arghh. how do I find out what is starting wpa_supplicant -u ? Jul 15 07:55:01 FiXion_: most probably ifup Jul 15 07:55:17 FiXion_: i don't need to fix routes, as eth1 route appears later, it takes priority. Jul 15 07:56:03 FiXion_: rmmod and modprobe shouldn't be necessary, FSO does an functional equivalent by itself (but nobody except fsoraw should have this resource requested). Jul 15 08:01:14 it was necessary here... but it might have been ifup stepping all over Jul 15 08:01:19 I'll try just to run dh_client Jul 15 08:05:19 tracfeed: Ticket #554 (usb0 configuration) updated Jul 15 08:10:06 I hate not having manpages Jul 15 08:10:13 udhcpd it seems is the dhcp client to use Jul 15 08:10:30 but it only reveals 2 options: -fS [configfile] Jul 15 08:10:36 not exactly helpful :( Jul 15 08:10:56 how do I tell it what interface to do dhcp on Jul 15 08:12:25 PaulFertser: fsoraw does not do the equivalent of rmmod&&modprobe : http://pastie.org/546537 Jul 15 08:12:45 after I ran it the first time - and there was a /var/run/wpa_supplicant/eth0 file Jul 15 08:13:57 if I try to rmmod ar6000 with a wpa_supplicant daemon running it totally dies. Jul 15 08:14:10 seems like some kind of race. rmmod never finished and I can't kill -9 wpa_supplicant process Jul 15 08:14:24 where would I open a bug report for this? Jul 15 08:14:38 I have a backtrace from dmesg Jul 15 08:16:01 kernel bug at sdio.c it says Jul 15 08:16:13 sdio_io.c 28 Jul 15 08:16:29 is the sdio_io.c driver part of the patches added to the stock kernel? Jul 15 08:23:03 morning all Jul 15 08:26:39 morning Jul 15 09:06:46 morning. Does someone know how to disable time update with gps on shr ? gllin makes it useless and broken :( Jul 15 09:08:13 SHR: 03mok 07shr-makefile * r19d46ec09b0e 10/Makefile: Makefile: remove deps on non existing shr overlay for shr-testing Jul 15 09:09:37 rtp: in frameworkd.conf Jul 15 09:09:54 [otimed] timesources = NTP Jul 15 09:10:04 (instead of GPS,NTP) Jul 15 09:11:30 SHR: 03seba.dos1 07shr-installer * r2c83f82855b4 10/shr-installer: Make errors looking better ;) Jul 15 09:11:30 SHR: 03seba.dos1 07shr-installer * r1e18bc34ae05 10/shr-installer: Some not important things, blah. Jul 15 09:11:47 rtp: you can also set timesources = NONE if you don't want frameworkd to touch your time Jul 15 09:11:48 mrmoku: I've no otimed entry in my frameworkd.conf. will add. Thanks :) Jul 15 09:13:25 mrmoku: I'm ok to use ntp but not gps. unlike on gta02, on gta01 it's a bad idea. Reset time to 00:00 until first sat and then get time wrong Jul 15 09:15:07 FiXion_: it's not fsoraw that does equivalent of module reloading, it's FSO itself that should do that. Jul 15 09:15:24 FiXion_: oops with ar6000 are known and stay unfixed for now... Jul 15 09:16:02 so nothing new I can add :) Jul 15 09:16:25 except perhaps a me too :) Jul 15 09:16:47 FiXion_: hehe, unfortunately, yes Jul 15 09:17:27 FiXion_: when WiFi resource in FSO is not used, it unbinds the driver. And binds on request. According to Werner it's equivalent to modprobe/rmmod. Jul 15 09:24:32 appearently not so Jul 15 09:34:40 FiXion_: i suspect it's just the resource is not really released. Jul 15 09:36:54 PaulFertser: sounds plausible Jul 15 09:37:52 TAsn: Sorry, had no time in the end yesterday :-( Jul 15 09:46:23 Ainulindale: hey. What is the proper way to alter default bluetooth/audio.conf bluez config in SHR? Jul 15 09:46:33 If I knew... :-) Jul 15 09:46:46 I always thought bluetooth headsets made people look weird Jul 15 09:48:17 They do. Jul 15 09:54:11 That's why I don't use them =) Jul 15 09:55:56 how do I add locales to shr ? Jul 15 09:56:20 no dpkg-reconfigure locales I'm guessing :) Jul 15 09:58:34 FiXion_: can you use locale-gen script? Jul 15 09:59:13 I'll try. Jul 15 09:59:20 there seems to be some glibc locale packages Jul 15 09:59:42 FiXion_: opkg install Jul 15 10:00:17 opkg install da_dk ? Jul 15 10:00:32 or en_dk (hopefully) Jul 15 10:01:22 FiXion_: replace locale.ipk by something like glibc-localedate-*.ipk Jul 15 10:01:45 those are the ones I'm installing Jul 15 10:01:50 locale -a still doesn't show them though Jul 15 10:02:49 rtp: how do you switch locale? Jul 15 10:02:55 just find the config file and edit it Jul 15 10:02:59 or is there some cli/gui tool Jul 15 10:03:48 PaulFertser: no locale-gen script Jul 15 10:04:31 FiXion_: you can borrow it from any desktop system, it's important to have localedef, then it'll work. Jul 15 10:04:40 FiXion_: fwiw I've these packages here : http://pastebin.ca/1495607 Jul 15 10:05:09 PaulFertser: it has localedef Jul 15 10:07:30 rtp: I have the same Jul 15 10:07:38 I just need to figure out where to change the default en_US Jul 15 10:08:06 FiXion_: to switch to an other locale, once locale -a is happy, edit /etc/profile Jul 15 10:08:53 me tries copying locale-gen over Jul 15 10:09:23 PaulFertser: running locale-gen gave no output Jul 15 10:10:03 I'll try running it with bash Jul 15 10:10:12 FiXion_: there's necessary an ipk for your locale. we only need to find out which :) Jul 15 10:10:32 there is glibc-localedata-en-dk Jul 15 10:10:36 and I installed it just fine Jul 15 10:10:39 but there must bemore Jul 15 10:10:49 perhaps binary-localedata.. Jul 15 10:11:44 yeah, this one looks "interesting". Jul 15 10:12:07 FiXion_: if you want to use locale-gen you're supposed to have /etc/locale.gen config file for it. Jul 15 10:12:17 ohh :) Jul 15 10:12:28 * FiXion_ obviously didn't read the source well enough :) Jul 15 10:12:28 mrmoku: ping Jul 15 10:13:41 rtp: the binary one was it Jul 15 10:13:51 can anyone check if libfso-glib builds with current fso-vala? Jul 15 10:14:35 FiXion_: great. if you manage to switch to your locale, imho, would be nice to put it on the shr wiki :) Jul 15 10:14:47 rtp: I'll do that :) Jul 15 10:15:50 thanks :) Jul 15 10:18:17 it worked Jul 15 10:20:47 put on wiki Jul 15 10:20:54 don't know how to format it nicely though Jul 15 10:21:36 FiXion_: look at http://shr-project.org/trac/wiki/Tweaks :) Jul 15 10:22:36 dos1: pong Jul 15 10:22:53 mrmoku: i think you can put shr-installer into feed Jul 15 10:23:11 dos1: with old packagekit? Jul 15 10:23:16 mrmoku: yep Jul 15 10:23:37 mrmoku: refreshing package list, installing and removing single packages works Jul 15 10:23:42 mrmoku: will you fix default /etc/bluetooth/audio.conf in SHR please? Jul 15 10:25:27 PaulFertser: if you tell me what fix it needs? Jul 15 10:26:18 mrmoku: yes, uncomment SCORoute=PCM line Jul 15 10:26:42 mrmoku: our BT chip's eeprom configuration requires that setting. Jul 15 10:29:14 mrmoku: could you please do something for me ?? Jul 15 10:29:16 :p Jul 15 10:29:28 good morning everybody btw Jul 15 10:29:44 ptitjes: no way ;) Jul 15 10:29:53 :D Jul 15 10:30:20 I will tell to daddy Ainulindale !!! Jul 15 10:30:23 :p Jul 15 10:30:55 mrmoku: it happens I forgot to bump the vala recipe on OE shr/import.. Jul 15 10:30:57 ptitjes: take care... he will spank you ;) Jul 15 10:31:08 mrmoku: could you please bump it ? Jul 15 10:31:27 the new version is : 0.7.5-fso1 Jul 15 10:31:32 ok Jul 15 10:31:39 thank you very much :D Jul 15 10:32:01 btw does the testing cooking goes as expected ? Jul 15 10:34:45 ptitjes: started this morning... and no, not yet:P Jul 15 10:34:57 :) Jul 15 10:35:05 I'm all with you Jul 15 10:35:52 PaulFertser: hmm... one thing though Jul 15 10:36:01 what about other hardware... like eten800? Jul 15 10:37:53 ptitjes, what is the vala base release? still 0.7.2? Jul 15 10:38:11 the one from upstream ? Jul 15 10:38:14 0.7.4 Jul 15 10:38:23 ok Jul 15 10:38:35 so it is 0.7.4+0.7.5-fso1 Jul 15 10:38:55 mrmoku: we always take the bumped version number that we postfix with -fsoX Jul 15 10:39:03 mrmoku: hum yeah... Jul 15 10:39:05 maybe Jul 15 10:39:11 I always forget :) Jul 15 10:39:18 just do the same as it was :) Jul 15 10:40:28 yep Jul 15 11:08:11 Deubeuliou: ping Jul 15 11:13:55 PaulFertser, if it is something machine specific I would rather enable that option for om-gta only Jul 15 11:14:04 PaulFertser, what do you think? Jul 15 11:17:40 Hey here Jul 15 11:17:45 hey Jul 15 11:18:20 Ainulindale, do you know if do_install_append_om-gta* is supposed to work? Jul 15 11:18:31 Yep it is Jul 15 11:18:35 good Jul 15 11:18:38 errr Jul 15 11:18:39 wait Jul 15 11:18:42 om-gta* as is ? Jul 15 11:18:42 bad ;) Jul 15 11:18:45 yep Jul 15 11:18:48 No it won't Jul 15 11:18:53 om-gta* isn't in OVERRIDES Jul 15 11:18:59 You have to use om-gta02, om-gta01 Jul 15 11:19:04 But it's easy to call one within the other Jul 15 11:19:12 I'd suggest declare a function Jul 15 11:19:15 and append it to do_install Jul 15 11:19:20 well... it's a oneliner sed to adjust audio.conf Jul 15 11:19:23 That's how it's supposed to be done Jul 15 11:19:24 so... Jul 15 11:19:28 then copy it :-) Jul 15 11:19:31 yep :) Jul 15 11:19:47 How's life apart from that? Jul 15 11:21:57 fine and busy Jul 15 11:22:01 yours? Jul 15 11:22:04 Well same here Jul 15 11:22:05 still on vacation? Jul 15 11:22:09 Been busy as hell last week Jul 15 11:22:22 Quite good reports on SHR from the community Jul 15 11:22:40 Funny to meet some people for real Jul 15 11:22:43 yep... astonishing, because we still could do much better ;) Jul 15 11:22:49 Yep Jul 15 11:23:00 and will :D Jul 15 11:24:04 I hope so :-) Jul 15 11:24:04 PaulFertser: ping Jul 15 11:25:49 Ainulindale: Do you have/know people testing for GTA01? Jul 15 11:27:28 Me Jul 15 11:27:38 I wasn't able to recently, or able to do anything for that matter Jul 15 11:27:43 But I was used to Jul 15 11:27:45 wurp2|lazy too Jul 15 11:27:45 Ainulindale: Okay Jul 15 11:27:46 and mwester Jul 15 11:28:07 talking about mwester... didn't see him for a long time... Jul 15 11:28:10 Ainulindale: Because the newer kernel doesn't (seem to) work with GTA01 Jul 15 11:28:16 alphaone: Yeah known issue Jul 15 11:28:18 mrmoku worked on that Jul 15 11:28:26 Ainulindale, nah... different issue Jul 15 11:28:26 It's ABI something Jul 15 11:28:30 ah? Jul 15 11:28:33 Then we should fix the rev Jul 15 11:28:36 IIRC alphaone's kernel does not even boot Jul 15 11:28:37 And the older one doesn't seem to build at the moment Jul 15 11:28:42 Yes Jul 15 11:28:57 It's not that it panics when going in userland Jul 15 11:29:04 It doesn't come up at all Jul 15 11:29:05 alphaone, I built a kernel with OABI some day ago and it seems to work Jul 15 11:29:23 mrmoku: Can you post the config somewhere? Jul 15 11:29:40 Or how did you build? yourself or with OE? Jul 15 11:29:57 alphaone, OE with shr/import Jul 15 11:30:00 no different config Jul 15 11:30:13 mrmoku: oh, interesting Jul 15 11:30:15 Thanks Jul 15 11:30:16 alphaone: pong. I don't know what could have changed so it doesn't work anymore. Jul 15 11:30:19 Will try that Jul 15 11:30:27 PaulFertser: Okay, thanks anyway Jul 15 11:30:28 I just fixed angstrom.inc to set ARM_KEEP_OABI for our kernel Jul 15 11:31:38 mrmoku: it's machine-specific, yes. Jul 15 11:31:50 mrmoku: (bluetooth setting) Jul 15 11:31:54 sweet Jul 15 11:31:58 mrmoku: I can confirm that the kernel you've uploaded with oabi si working here :) Jul 15 11:32:21 :) Jul 15 11:32:23 rtp: for gta01 or 02? Jul 15 11:32:24 rtp: mrmoku THat's 2.6.28? Jul 15 11:32:28 for gta01 Jul 15 11:32:31 ? Jul 15 11:32:32 shoragan: gta01 Jul 15 11:32:37 mrmoku: hows the testing image doing? Jul 15 11:32:50 alphaone, we build the same kernel rev for 01 and 02 I think Jul 15 11:32:56 linux-openmoko-shr-devel Jul 15 11:33:01 alphaone: "Linux version 2.6.29-rc3 (shr@opmbuild) (gcc version 4.1.2) #1 PREEMPT Sun Jul 12 20:21:03 CEST 2009" Jul 15 11:33:02 and it should be 2.6.29... Jul 15 11:33:22 okay, great Jul 15 11:34:21 alphaone: by the way Jul 15 11:34:23 fso-sounds is Jul 15 11:34:26 1) Horrid Jul 15 11:34:30 2) Gigantic Jul 15 11:34:36 3) Annoying as we need to remove it Jul 15 11:34:41 4) Darth Vader Jul 15 11:34:54 please strike the useless mention Jul 15 11:35:09 darth vader? Jul 15 11:35:15 TAsn-moko: Cheater Jul 15 11:35:34 I'll strike all prime numbers Jul 15 11:35:52 Ainulindale: did ya fix the makefile? Jul 15 11:36:12 TAsn-moko: No, as I told you I wasn't able to yesterday in the end Jul 15 11:36:17 I'm tired and sick of having no time Jul 15 11:37:02 sorry, didnt see :) Jul 15 11:37:40 Seriously I'm eagerly waiting to change jobs now Jul 15 11:38:09 :D Jul 15 11:38:20 hello Ainu Jul 15 11:38:25 ptitjes: hey you're going to laugh Jul 15 11:38:33 relly ? Jul 15 11:38:34 My new job laptop is... a macbook Jul 15 11:38:35 +a Jul 15 11:38:40 héhéhé Jul 15 11:38:42 And first thing I did is Jul 15 11:38:47 I wiped out the harddrive Jul 15 11:38:48 And put gentoo on it Jul 15 11:38:53 nice Jul 15 11:39:02 Quite difficult to make alsa work, but I did Jul 15 11:39:06 Everything is working properly now Jul 15 11:39:15 mazeltof Jul 15 11:39:18 Although I remapped the qwerty keyboard to azerty in software Jul 15 11:39:25 So it's a bit unsettling for people Jul 15 11:39:28 Ainulindale: sucks! (mac) though gj (gentoo) Jul 15 11:39:28 But I don't care about people Jul 15 11:39:30 oh it is a uk mac ? Jul 15 11:39:39 Yep Jul 15 11:39:41 Macbook 5.1 Jul 15 11:39:52 4Gb RAM, 250Gb, dual core my ass Jul 15 11:39:58 I don't know what I'll do with all that stuff Jul 15 11:40:02 I mainly do web Jul 15 11:40:16 lol Jul 15 11:40:34 And I have to say that backlit keyboard is useless Jul 15 11:40:40 Ainulindale: you'll appreciate it as a knew warmer for long winter nights Jul 15 11:40:41 Utterly useless Jul 15 11:40:48 cya Jul 15 11:40:49 -w+e Jul 15 11:40:52 ptitjes: Not with cpufreqd & acpid Jul 15 11:41:01 Underclocked properly, I have no issue Jul 15 11:41:04 haa Jul 15 11:41:11 Even underclocked video card Jul 15 11:41:20 wow Jul 15 11:41:20 As the most difficult thing I display is my wallpaper Jul 15 11:41:31 :p Jul 15 11:41:47 ptitjes: but if you want some help with yours, I can haz gentoo now Jul 15 11:42:02 why not :) Jul 15 11:42:04 ptitjes: oh and by the way Jul 15 11:42:08 http://openmoko-fr.org/forum/viewtopic.php?pid=7061 <= Jul 15 11:42:10 but I already maintain mine :) Jul 15 11:42:23 With Gentoo/Prefix IIRC Jul 15 11:42:25 Not Gentoo Jul 15 11:42:46 ail t'as raté ton annonce officielle ? Jul 15 11:43:00 Ainulindale: Gentoo/Parallels Jul 15 11:43:29 though I may move to VBox, or... someday Jul 15 11:45:47 mrmoku: what about shr-installer? Jul 15 11:47:05 dos1, one step after the other ;) Jul 15 11:47:27 Ainulindale, how to make a package arch dependent? Jul 15 11:47:41 the om-gta02 override has no effect for bluez4 Jul 15 11:47:53 alphaone, or maybe you have a better solution? Jul 15 11:48:19 mrmoku: the override should be there for both gta01 and gta02 please :) Jul 15 11:48:20 PACKAGE_ARCH Jul 15 11:48:25 Check frameworkd-config Jul 15 11:48:35 PaulFertser, yep have done it for both Jul 15 11:48:37 Ainulindale, ok, thanks Jul 15 11:51:58 mrmoku: BTW, i've sent patch for opimd-utils recipe ~week ago to shr-devel Jul 15 11:52:36 dos1, I had seen that one... and forgot about it, sorry Jul 15 11:53:37 mrmoku: no problem ;) Jul 15 11:56:04 Ah I have time to check that Jul 15 11:56:05 Ainulindale, would it make sense to do PACKAGE_ARCH_shr = "${MACHINE_ARCH}" for bluez4? Jul 15 11:56:06 I'll recheck patchwork Jul 15 11:56:15 Why would you want to do that mrmoku? Jul 15 11:56:31 To ship an extraconfig? Jul 15 11:56:33 because we need a custom audio.conf for gta01/02 Jul 15 11:56:37 Ok then don't do that Jul 15 11:56:40 Add a package Jul 15 11:56:40 Ainulindale: to match the hardware Jul 15 11:56:42 not extraconfig... changed config Jul 15 11:56:43 Ship the file in this package Jul 15 11:56:47 And make it arch dependent Jul 15 11:56:55 I'd do that myself Jul 15 11:57:38 Ainulindale, would you have time to do that? I pushed the necessary change to bluez4 Jul 15 11:57:45 it just has no effect ;) Jul 15 11:58:18 To do what? The extra package? Jul 15 11:58:24 yep Jul 15 11:58:45 Hmpf I could but I won't be able to test a thing right now Jul 15 11:59:38 hmm... not sure an extra package will work out Jul 15 12:00:05 the config files are part of the bluez4 base package Jul 15 12:00:22 ahem Jul 15 12:00:33 mrmoku: let me show you an example :-) Jul 15 12:01:03 http://cgit.openembedded.net/cgit.cgi/openembedded/tree/recipes/freesmartphone/frameworkd_git.bb?h=shr/import Jul 15 12:01:12 line 69 Jul 15 12:01:22 * dos1 eats Jul 15 12:01:27 All you need to know is there Jul 15 12:01:42 You may even be able to copy paste it to a certain extent Jul 15 12:02:56 playya_: http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?h=shr/import&id=f2fea82349714231ad2253b0c2d29fc7dd1cfa52 Jul 15 12:02:59 :) Jul 15 12:03:09 Ainulindale, well I did not say it would not be possible ;) Jul 15 12:03:23 mrmoku: I hope so else I'd have spanked you Jul 15 12:03:36 but I don't know if it is worth all that just to remove one char in one conffile :P Jul 15 12:03:53 Well it might be useful for the future Jul 15 12:04:02 I don't know how bluez packaging is done Jul 15 12:04:04 So I'm just suggesting Jul 15 12:05:16 Ainulindale, are etc files auto-packaged? Jul 15 12:06:06 Ainulindale: why are SHR devs' efforts are concentrated on writing elementary apps instead of writing an emacs mode? ;) Jul 15 12:06:50 PaulFertser: ncurses Jul 15 12:07:10 ptitjes: You suck :-) Jul 15 12:07:19 mrmoku: with autotools yes if they're declared that way Jul 15 12:07:38 Ainulindale, yep, that's how bluez packages its config files then Jul 15 12:07:51 Ainulindale: how's libframeworkd-phonegui-ncurses going? ;) Jul 15 12:07:52 Ainulindale: guess what. I seriously looked at the ncurses APIs :) Jul 15 12:08:18 I bet I can program something quickly in Vala =p Jul 15 12:08:57 Well I have all the bindings for mouse ptitjes Jul 15 12:09:11 ohoh Jul 15 12:10:14 ptitjes: not quite. Emacs provides you with a ready-to-use scripting infrastructure along with a familiar keybindings. Just imagine a contact list in emacs. You can use inceremental search there without any additional effort. You can split windows, hide/show, whatever you need. Why do you consider X+EFL (or ncurses) to be a better framework to build phone apps? Jul 15 12:10:52 mrmoku: Patchwork is working now Jul 15 12:11:00 * mrmoku is in no way familiar with emacs keybindings ;) Jul 15 12:11:10 I ended to have problems with the 20090709 Jul 15 12:11:14 unstable Jul 15 12:11:30 mrmoku: not a big deal. Learn once, use everywhere (gnu readline default is emacs-like). Jul 15 12:11:47 PaulFertser: Nothing more than what toolkits can provide Jul 15 12:11:57 mrmoku: Check patchwork Jul 15 12:12:00 You should see my latest patch Jul 15 12:12:00 PaulFertser: just the burden of a scripting language Jul 15 12:12:07 ptitjes: exactly, emacs is not toolkit that's why it so much better ;) Jul 15 12:12:28 PaulFertser: so please write libframeworkd-phonegui-emacs :D Jul 15 12:12:45 :) Jul 15 12:13:06 hi! Jul 15 12:13:16 dos1: there's already an emacs mode by johnsu01. And guess what, it doesn't use any of SHR api, pure FSO. Jul 15 12:13:20 PaulFertser: mrmoku: [11:34] FiXion_: i suspect it's just the resource is not really released. Jul 15 12:13:20 hi methril|work Jul 15 12:13:22 Ainulindale, but I don't see the one from dos1 ;) Jul 15 12:13:44 I suggested to have lsresources some time ago Jul 15 12:13:45 ptitjes: you're the vala guru? :) Jul 15 12:13:52 mrmoku: Yes I just enabled it Jul 15 12:13:52 humm Jul 15 12:14:20 dos1, could you try to resend your patch please? Jul 15 12:14:21 methril|work: ask and we'll see ;) Jul 15 12:14:52 mrmoku: ok Jul 15 12:14:54 mrmoku: I just assigned the patch to you Jul 15 12:14:56 Do you have a mail? Jul 15 12:15:06 Ainulindale: what's patchwork url? Jul 15 12:16:08 http://patchwork.dev.bearstech.com/project/SHR/list/ Jul 15 12:16:13 DocScrutinizer: exactly to the point Jul 15 12:17:01 ptitjes: http://pastebin.com/m1b8bd8f0 Jul 15 12:17:06 dos1: change the filters to see the patches Jul 15 12:17:18 methril|work: update vala Jul 15 12:17:19 http://patchwork.dev.bearstech.com/project/SHR/list/?state=* Jul 15 12:17:20 PaulFertser: seems I need to open a ticket similar to fsoraw ;-) Jul 15 12:17:21 for instance Jul 15 12:17:34 methril|work: http://www.freesmartphone.org/sources Jul 15 12:17:38 DocScrutinizer: :) Jul 15 12:17:40 mrmoku: done Jul 15 12:17:50 Yes! Jul 15 12:17:53 Working nicely :-) Jul 15 12:18:40 methril|work: or is it ie OE ? Jul 15 12:18:47 uhm..... it's OE Jul 15 12:18:49 Ah, that's done, cool Jul 15 12:18:50 with other machine Jul 15 12:18:54 Ok so now checking the makefile Jul 15 12:18:58 methril|work: then update OE :) Jul 15 12:19:03 i'm doing some test to add other system Jul 15 12:19:06 i updated :( Jul 15 12:19:17 as when I started this topic, evrybody said "good idea" and we even discussed about what to actually show (instead of "python") for rexource-users Jul 15 12:19:23 it compiles well with this machine but it doesn't with the other one Jul 15 12:19:25 sonner than 1 hour ago ? Jul 15 12:19:36 nop Jul 15 12:19:39 i'll try Jul 15 12:19:42 PaulFertser: alas nobody kept track of it Jul 15 12:19:53 mrmoku: uh, i fscked something with that patch, PR change is in comments Jul 15 12:20:00 methril|work: see http://cgit.openembedded.org/cgit.cgi/openembedded/log/?h=shr/import the commit before the last one Jul 15 12:20:16 Ah mrmoku was quicker than me Jul 15 12:20:20 TAsn: mrmoku updated the makefile Jul 15 12:20:26 You can start working on fixing revs Jul 15 12:20:31 mrmoku: thanks a lot Jul 15 12:20:43 PaulFertser: so almost same disappointing result as with ramconsole Jul 15 12:20:50 ptitjes: nice!! thank you Jul 15 12:21:00 mrmoku, Ainulindale: remove that patch :P Jul 15 12:21:07 i still didn't understand how it compiles with gta02 machine :) Jul 15 12:21:12 i'll send it again, from git send-email Jul 15 12:21:35 Ok I set it to superseded Jul 15 12:21:55 methril|work: hum. I did not understand ? Jul 15 12:21:55 DocScrutinizer: yes, that's one of the downsides of working on whatever one's want at any particular moment. Someone who can do that looses interest or tries to do something else and good ideas are forgotten. Jul 15 12:22:23 ptitjes: i'm checking the build with other embedded system Jul 15 12:22:53 ptitjes: but i try the same oe repo with the Freerunner system as an target and it works Jul 15 12:23:31 methril|work: which systems ? Jul 15 12:24:11 sorry I do not understand :( Jul 15 12:24:13 mrmoku, Ainulindale: ok, now it's ok :) Jul 15 12:24:13 dos1: http://patchwork.dev.bearstech.com/patch/5/ Jul 15 12:24:24 Ainulindale: i've seen ;) Jul 15 12:24:25 ptitjes: machine=om-gta01 Jul 15 12:24:31 dos1: useful, isn't it? Jul 15 12:24:38 yep :) Jul 15 12:24:44 Ainulindale: nice Jul 15 12:25:05 I checked with a normal mail to advert that aswell Jul 15 12:25:42 methril|work: and it fails ? Jul 15 12:26:56 ptitjes Jul 15 12:27:06 ptitjes: it fails with machine=em-x270 Jul 15 12:27:25 methril|work: DISTRO=SHR branch shr/import ? Jul 15 12:27:42 yes Jul 15 12:28:00 dos1, patch under review now ;) Jul 15 12:28:09 mrmoku: haha Jul 15 12:28:16 You're an ass Jul 15 12:28:21 ;D Jul 15 12:28:27 methril|work: libfso build fails with the reference out parameters in yielding functions ? Jul 15 12:28:29 right ? Jul 15 12:31:20 dos1, and applied :) Jul 15 12:31:25 DocScrutinizer: do you know anything about the buzz fix activities at debconf9? I read http://wiki.openmoko.org/wiki/BuzzFixParty#Caceres_.28Debconf09.29 but the ticket can not be found from the shop anymore. I also sent mail to david Jul 15 12:34:32 lindi-: sorry, no info arrived here yet Jul 15 12:36:40 mrmoku: hum maybe the recipe should have been named vala-native_0.7.4+fso1.bb... Jul 15 12:37:33 mrmoku: IMHO this will clash if ever we deliver an official 0.7.5 (but I don't know if that will happen) Jul 15 12:38:24 ptitjes, hmm... I don't think so Jul 15 12:38:32 because it does not matter how the recipe is called ;) Jul 15 12:38:50 the PV is what matters and that is 0.7.4+0.7.5-fso1 Jul 15 12:39:48 mrmoku: ok but now we are forced to have one minor version more than the official Jul 15 12:40:03 because 0.7.5 does not exist yet :) Jul 15 12:46:45 ptitjes: right Jul 15 12:47:47 lindi-: i'm the iron man at the buzz fix at debconf09 Jul 15 12:48:00 methril|work: so there is something that does not select the correct valac Jul 15 12:48:37 ptitjes: it's working with the last update, but it's broken in another way :) Jul 15 12:48:39 ptitjes, I don't get you ;) Jul 15 12:49:29 mrmoku: when official 0.7.5 will be out we will have to name our recipe _0.7.6+... Jul 15 12:50:07 methril|work: is it still possible to get a buzz fix ticket? ;) Jul 15 12:50:49 ptitjes, well, it will *be* 0.7.6 then, no? Jul 15 12:51:49 mrmoku: I thought git revision was added above last release number in recipe names Jul 15 12:52:08 lindi-: i think so, i see that the tuxbrain shop doesn't have buzz fix tickets anymore but it has to be an error Jul 15 12:53:04 mrmoku: said another way, it is more a 0.7.4 plus some added features than a 0.7.5 without some features Jul 15 12:53:05 :D Jul 15 12:54:19 mrmoku: hence the 0.7.4+fsoX Jul 15 12:54:59 anyway Jul 15 12:57:14 methril|work: what other way ? Jul 15 13:03:40 hi guys Jul 15 13:03:59 ptitjes: kernel receipe Jul 15 13:04:07 ha Jul 15 13:04:13 :) Jul 15 13:05:42 how can i change the current profile over config files? i want to write a script which switches the profile e. g. from vibrate to ring Jul 15 13:07:00 I've read the program managing this is exposure but i didn't find any config files for this Jul 15 13:07:30 do we plan to merge shr/import with openembedded.dev? Jul 15 13:12:10 has nobody an idea? Am I the first who want to switch the profile over shell? Jul 15 13:13:18 mrmoku: is this known that the contact list does not load completely when shr-messages is launched too soon ? Jul 15 13:13:42 Deadman44: have you tried to look at FSO API description? Jul 15 13:13:57 Deadman44: there's mdbus that allows easy scripting of FSO API calls. Jul 15 13:14:37 PaulFertser, I will have a look at this. Thanks Jul 15 13:14:39 and strangely my calls are not working ?? Jul 15 13:15:40 mrmoku, did something important changed in unstable just before the 20090709 release ? Jul 15 13:17:54 ptitjes: yeah, your mamma Jul 15 13:18:09 ptitjes: for #1 question this is a known issue Jul 15 13:19:00 and what do you mean by calls are not working? Jul 15 13:19:05 That doesn't mean a thing and you know that =) Jul 15 13:22:13 :) Jul 15 13:22:29 I can't understand what happened Jul 15 13:22:47 you tried to place a call and what did it do? Jul 15 13:22:55 Did you upgraded to latest? Jul 15 13:23:10 first call: active call screen appaers and then disappears Jul 15 13:23:31 second call and after: no active call screen at all Jul 15 13:23:37 ophonekitd borked Jul 15 13:23:39 done from contacts app Jul 15 13:24:16 Ainulindale: I noticed some strange long unresponsive delays with 0709 unstable Jul 15 13:24:32 I've been told that's because of ologicd Jul 15 13:24:42 oh yeah that explains Jul 15 13:24:45 ;) Jul 15 13:24:54 Delays while doing what? Jul 15 13:25:03 (you are the worst bug reporter ever!) Jul 15 13:25:05 sometime nothing Jul 15 13:25:19 How can you notice a delay when you do nothing? Jul 15 13:25:24 héhéhé Jul 15 13:25:26 What were you doing exactly? Jul 15 13:25:37 because I try to do something and it is blocked Jul 15 13:25:45 So you tried to do something Jul 15 13:25:46 What? Jul 15 13:26:10 sorry I can't remember but this happenned while making different things Jul 15 13:26:16 first boot? Jul 15 13:26:21 maybe Jul 15 13:26:23 flashed? upgraded? Jul 15 13:26:26 flashed Jul 15 13:26:35 but upgraded Jul 15 13:26:38 oups Jul 15 13:26:42 sorry I forgot Jul 15 13:27:10 I reflashed 0710 in the afternoon Jul 15 13:27:13 -re Jul 15 13:30:16 re Jul 15 13:30:20 hey mwester Jul 15 13:30:23 howdy? Jul 15 13:30:25 salutations Jul 15 13:30:46 :-) Jul 15 13:31:14 I plan to wipe out my shr repo, and do a clean build and install today... what are the chances that it will all work? Jul 15 13:31:26 Very good chances Jul 15 13:31:30 excellent! Jul 15 13:31:34 I tried that myself a couple of days ago, with no issue Jul 15 13:33:36 mwester: long time no see, welcome back :D Jul 15 13:34:24 thanks! I see you've been busy with jtag and such... Jul 15 13:35:09 mwester: not me, it was Psi who accidentally borked his gta01 u-boot. I just suggested an easy method of recovering without a dboard. Jul 15 13:35:16 mwester: i cleanly build this morning and it works Jul 15 13:36:06 PaulFertser, that looks a lot like the wiggler circuit -- real easy to do. And if you add a cell phone 3,3v-serial-to-usb cable, you can replace the debug board for very little money.l Jul 15 13:36:26 anybody seen before the bootscreen being spilled with TONS of " has a timestamp in the future" ? Jul 15 13:36:27 mwester: sure, provided you have a spare parallel port. Jul 15 13:37:12 I have lots of old obsolete computers, so I have loads of parallel ports! (but they are rather rare for those who have new stuff, that is true) Jul 15 13:37:34 mwester: hey still alive :-) Jul 15 13:37:49 mwester: i'm about to implement dumb battery handling for gta02 based on gta01 driver. So far i see no good solution except for manually unbinding bq27000 driver and binding gta01-battery. Do you have any ideas about how to do it in a more fancy way? Jul 15 13:38:00 yep. back from the fringes of GSM coverage to the real world for a while... Jul 15 13:38:25 mwester: found a new job that keeps you busy all the day and night? Jul 15 13:38:26 PaulFertser, that's exactly the issue I ran into with the bq27000 driver when I looked at that. Jul 15 13:39:10 mwester: for me binding/unbinding manually is ok at works. But DocScrutinizer thinks it's too inelegant. He'd prefer to have an automatic way of switching whenever HDQ communication constantly fails. Jul 15 13:39:23 s/ok at works/ok and works/ Jul 15 13:39:23 PaulFertser meant: mwester: for me binding/unbinding manually is ok and works. But DocScrutinizer thinks it's too inelegant. He'd prefer to have an automatic way of switching whenever HDQ communication constantly fails. Jul 15 13:39:38 TAsn: how's work on shr-testing going? Jul 15 13:39:39 PaulFertser: compromise, not job :( There are no jobs available, so I am instead "reducing costs" -- I am painting an old building right now that normally I would have paid someone else to do properly. :( Jul 15 13:40:04 PaulFertser: absolutely Jul 15 13:40:15 PaulFertser, after following the discussions with doc and others, I am no longer convinced that the bq drivers adds value. Jul 15 13:40:20 mwester: ah. OTOH, DIY spirit ftw :D Jul 15 13:41:14 mwester: it gives end-user more accurate estimate of remaining capacity comparing to approximation based solely on voltage vs. charge graph. Jul 15 13:41:23 PaulFertser: detecting HDQ-fail is<'t any problem Jul 15 13:41:37 PaulFertser, DIY is right. Painting is easy; it's the "other stuff" that is such a pain in the butt. For example, I have now discovered that much of the stairway needs to be rebuilt, as it seems that it's only the old paint that is keeping it from falling into the basement! :D Jul 15 13:41:49 mwester: also we don't have a way to measure the current on gta02. (not that it works anywhere near good on gta01 ;) ) Jul 15 13:41:56 Ah ok. Jul 15 13:42:03 Then we need a single unified driver. Jul 15 13:42:11 DocScrutinizer: yes, one just needs to write a userspace daemon watching the syslog and switch the drivers. Jul 15 13:42:14 DocScrutinizer: so no kernel work. Jul 15 13:42:16 It needs to handle the "state changes" itself. Jul 15 13:42:37 mwester: we're sol with dumb batteries anyway. No ability to measure current on both gta01 and gta02. Jul 15 13:42:51 Clarify? Jul 15 13:42:57 We can measure current. Jul 15 13:43:27 Are you saying we are not really measuring the charge current (because of the way its wired)? Jul 15 13:43:35 mwester: dumb batteries do not have an intergrated shunt to measure the current. gta01 has one but it's so noisy that it makes it useless. Jul 15 13:43:42 ah. Jul 15 13:43:51 we can filter the noise to some degree Jul 15 13:43:54 mwester: we can only measure the current with bq27000 on gta02. Jul 15 13:44:03 mwester: +-100mA is hardly filterable :) Jul 15 13:44:12 is it really that bad? Jul 15 13:44:20 I've not noticed, although I've not looked closely either. Jul 15 13:44:22 mwester: from what i saw, yes. Jul 15 13:44:28 wow Jul 15 13:44:39 mwester: people showed me the logs (i unfortunately don't have a gta01 myself (yet)) Jul 15 13:44:57 :( Jul 15 13:46:01 mwester: i'm afraid that making bq27000 + dumb battery driver is out of scope due to layering violation. I can't see a clean way here. Either we load bq27000 or we load gta01-battery driver. Userspace can simplify that by switching automatically. Jul 15 13:46:26 Well, that's easier too. Jul 15 13:47:01 Sorry, gtg. See you later :D Jul 15 13:47:11 later Jul 15 13:47:12 PaulFertser: it smells like feature request for FSO Jul 15 14:01:49 PaulFertser: dos1: IIRC we got a "present" attrib for bat. So caould we have 2 bat - one GTA02 and one GTA01 style, and set present accordingly? Jul 15 14:02:34 (in /sys all this) Jul 15 14:05:27 mrmoku: ping Jul 15 14:05:56 DocScrutinizer: makes sense for me Jul 15 14:07:32 nytowl, pong Jul 15 14:07:45 dos1: "present" must fail anyway, when HDQ (02) or thermistor readout (01) borks Jul 15 14:08:09 mrmoku, have you been trying to integrate 2.6.30 ? Jul 15 14:08:31 nytowl, hmm, no Jul 15 14:08:38 maybe lock gta01 by asking state of gta02-present Jul 15 14:08:39 we take whatever andy-tracking gives us ;) Jul 15 14:09:02 ah k, so no one in SHR is looking at 2.6.30 ? Jul 15 14:11:23 mrmoku: looks like andy-tracking is going to be obsolete Jul 15 14:12:52 as working on andy-tracking makes preparing patches for upstream inclusion harder Jul 15 14:12:58 (that's what people are saying) Jul 15 14:13:57 looks like moving to 2.6.30 + patchset for gta01/2 drivers is better solution now Jul 15 14:14:07 with one patch for one driver Jul 15 14:14:32 2.6.30 is booting nicely and can talk to the modem and GPS ( with modified sys paths Jul 15 14:14:33 and someone did that already (look at openwrt patches), but only for gta02 ATM Jul 15 14:15:15 I'm working on alsa and wifi now and was wondering if anyone else had looked at it Jul 15 14:15:19 nytowl, plain 2.6.30? Jul 15 14:15:41 nope the on in ms5.5 the recipe applies all of the openwrt patches Jul 15 14:23:22 dos1, (testing image) no idea, ask mrmoku, I asked him earlier, though I can't remember what his answer was :) Jul 15 14:23:36 TAsn: he did his job Jul 15 14:23:41 that's what Ainulindale said ;) Jul 15 14:24:05 dos1, not quite Jul 15 14:24:10 he did something else I asked from him Jul 15 14:24:10 [14:19] TAsn: mrmoku updated the makefile Jul 15 14:24:12 [14:19] You can start working on fixing revs Jul 15 14:24:26 saw that Jul 15 14:24:50 though I asked Ainulindale for something else Jul 15 14:25:08 unrelated Jul 15 14:25:20 I mean, I asked him to fix the makefile to work with the OE shr branch Jul 15 14:26:06 and I think that's what mrmoku did Jul 15 14:26:06 as for the testing image Jul 15 14:26:06 I'm no package maintainer Jul 15 14:26:06 I have no idea how to even start working with OE correctly Jul 15 14:26:06 mrmoku is the expert :) I'm just a programmer/troll ;) Jul 15 14:28:46 TAsn, ??? Jul 15 14:29:12 he asked me about the testing image Jul 15 14:29:54 and I said that I have no idea since you are the expert Jul 15 14:30:05 PaulFertser, (audio.conf) I asked on oe devel how to do that Jul 15 14:30:18 preferred version would be that blueprobe autodetects that ;) Jul 15 14:30:23 (if possible) Jul 15 14:30:44 bbl see you Jul 15 14:31:03 mrmoku: i think i'll start to implement volume slider in libframeworkd-phonegui-efl ;) Jul 15 14:31:05 TAsn, yeah, but what are the Makefile changes you want from Ainulindale Jul 15 14:31:11 mrmoku, while we are at it, how is the testing image going? Jul 15 14:31:14 dos1, you can't do that Jul 15 14:31:19 mrmoku: why? Jul 15 14:31:20 mrmoku, I wanted the makefile to use the OE shr repo Jul 15 14:31:23 TAsn, working on it Jul 15 14:31:46 i.e, I wanted the makefile to work out of the box for creating an OE build env Jul 15 14:31:46 :) Jul 15 14:31:48 dos1, because frameworkd does not abstract that yet, or does it? Jul 15 14:32:03 TAsn, it does not? Jul 15 14:32:09 oh, that way Jul 15 14:32:38 dos1, the direct alsa way would break horribly on other hardware ;) Jul 15 14:32:52 mrmoku: yep, thought about that few seconds ago Jul 15 14:32:54 ;) Jul 15 14:33:02 hmm, ok. when we'll use SetSpeakerVolume we'll be spanked by DocScrutinizer ;D Jul 15 14:33:11 yep Jul 15 14:33:13 mrmoku, no idea, I just talked about it Jul 15 14:33:13 and then Ainulindale said it doesn't Jul 15 14:33:13 so I asked him to fix it :) Jul 15 14:33:13 and that was about it ;) Jul 15 14:33:14 we're doomed :P Jul 15 14:33:35 What? Jul 15 14:33:36 Ainulindale, ^^ ??? Jul 15 14:33:38 :) Jul 15 14:33:48 TAsn: I checked latest commits Jul 15 14:33:56 Apparently mrmoku stripped testing of its overlay Jul 15 14:33:59 so to me it's good Jul 15 14:34:24 * mrmoku likes stripped stuff... especially female stripped stuff :P Jul 15 14:34:55 mrmoku: ;> Jul 15 14:35:28 mrmoku, bottom line Jul 15 14:35:50 on the unstable branch Jul 15 14:35:50 can I just Jul 15 14:35:50 make setup Jul 15 14:35:50 and get a working OE env? Jul 15 14:35:51 :) Jul 15 14:35:53 mrmoku: shouldn't you s/female/my wife's/, or you're sure she won't read it? ;) Jul 15 14:36:09 TAsn: yes Jul 15 14:36:12 and yes for testing too Jul 15 14:36:14 and dos1 Jul 15 14:36:18 get on your volume slider now Jul 15 14:36:20 or feel my wrath Jul 15 14:37:22 Ainulindale: so what method i should use to setting volume? Jul 15 14:37:52 dos1, first of all Jul 15 14:37:58 wrap with your own function Jul 15 14:38:17 so you'll be able to easily change to whatever we'll decide in the future :) Jul 15 14:38:18 dos1, she does not understand english... so I'm fine off ;) Jul 15 14:38:25 dos1: well I'd suggest put it into a conf file Jul 15 14:38:33 for the time being Jul 15 14:38:41 (the whole system command) Jul 15 14:38:49 that way we'll be able to use alsactl or mdbus or whatever Jul 15 14:38:57 Although it would have no immediate impact Jul 15 14:39:01 system command sucks Jul 15 14:39:07 :) Jul 15 14:39:15 So does your mamma Jul 15 14:39:20 just wrap it around! Jul 15 14:39:38 bleh :-p Jul 15 14:39:48 dos1: ask mickey|lunch what is the new way to do that Jul 15 14:39:54 dos1: because there is one Jul 15 14:40:42 Ainulindale, lol. Jul 15 14:41:36 TAsn: "lol" yourself! Jul 15 14:41:48 I am. Jul 15 14:41:55 lol Jul 15 14:41:57 ;) Jul 15 14:42:06 mickey|lunch: i'm asking you Jul 15 14:42:08 ;) Jul 15 14:42:35 btw... did you notice that mickey always eats lunch AGES? ;x Jul 15 14:43:02 dos1: ok found it Jul 15 14:43:06 German sausage digestion shouldn't be hastened Jul 15 14:43:47 ptitjes: so? ;) Jul 15 14:43:56 2 sec Jul 15 14:44:07 1 Jul 15 14:44:08 2 Jul 15 14:44:10 you're late :P Jul 15 14:45:43 dos1: ok you have to ask mickey to open the mixer controls in the APIs Jul 15 14:46:01 dos1: he wanted to do it, but it seems he did not yet Jul 15 14:46:36 dos1: wrap a alsactl call in a function for now, you'll replace it by a dbus call later Jul 15 14:47:54 dos1: but opening the two configurable mixer controls per scenario as he wanted to do might not take so much time in cornucopia Jul 15 14:52:54 amixer -c 0 set Speaker 100% Jul 15 14:52:56 ? Jul 15 14:53:24 dos1: and there should be a configurable min and max for each control Jul 15 14:53:54 ptitjes: well, amixer does it with percentage value Jul 15 14:54:36 dos1: yeah and ? can't you have a configurable min and max percentage per mixer control ? Jul 15 14:54:51 but why for? Jul 15 14:55:06 dos1: some values are not to used Jul 15 14:55:41 dos1: gui must give a possibility to change between decent values Jul 15 14:55:57 ok, but after implemeting basics Jul 15 14:56:21 dos1: and remember that although amixer is manipulated with percentages Jul 15 14:56:43 dos1: some values are less than sixty-valued Jul 15 14:57:08 sixty-valued? Jul 15 14:57:24 dos1: that would be a pain for user if the slider can go in between positions and not snap to positions, as the toochscreen is crappy Jul 15 14:58:07 some mixer controls have a tiny number of values Jul 15 14:58:13 ptitjes: oh, i don't think elementary slider can snap to positions Jul 15 14:58:18 dos1: also use alsactl directly Jul 15 14:58:26 my 2 cents Jul 15 14:58:38 dos1: this is not an excuse Jul 15 14:58:47 (for the elemntary sliders) Jul 15 14:58:48 :D Jul 15 14:59:25 ptitjes: so give me example, how to use alsactl for that Jul 15 14:59:26 :P Jul 15 14:59:38 man alsactl Jul 15 14:59:52 :p Jul 15 15:00:08 ptitjes: read, but still don't know Jul 15 15:02:13 ptitjes: ? Jul 15 15:02:30 our midori doesn't do ssl Jul 15 15:02:30 right? Jul 15 15:02:32 can't remember would have to google Jul 15 15:03:09 dos1, just google, I mean Jul 15 15:03:43 there are plenty of cool alsa examples Jul 15 15:03:43 there's pyalsa Jul 15 15:03:43 or something like that Jul 15 15:03:48 alsactl? Jul 15 15:04:03 googled, but still don't know how to just change one control with it ;P Jul 15 15:04:12 from command line, without specifing scenario file Jul 15 15:08:51 TAsn: no, i'm asking how to do it with alsactl, cause ptitjes said i should use alsactl directly :P Jul 15 15:09:35 dos1: or if it is from settings and there is pyalsa Jul 15 15:09:36 dos1, that's because ptitjes is a hacker :) Jul 15 15:09:39 then use pyalsa Jul 15 15:09:47 ptitjes, :) Jul 15 15:09:50 ptitjes: no, not settings Jul 15 15:10:01 dos1, quick tip: Jul 15 15:10:03 ptitjes: libframeworkd-phonegui-efl, active call screen :P Jul 15 15:10:04 TAsn: beware, I'm going to hack you Jul 15 15:10:06 :p Jul 15 15:10:20 libcanberra Jul 15 15:10:27 I remember finding a pygtk+pyalsa audio mixer Jul 15 15:10:32 or the other one from Doc Jul 15 15:10:57 dos1: libalsa Jul 15 15:11:20 please don't do command line when you can do C calls Jul 15 15:11:46 ptitjes: but that's only temporiary solution until it'll be implemented in fso! ;x Jul 15 15:11:51 * dos1 checks how it was done in 2007.2 Jul 15 15:12:04 try googling for that Jul 15 15:12:04 the "hard" part is actually making it persistent (i.e saving it to the correct state after you are done) Jul 15 15:12:20 ptitjes, ++ :) Jul 15 15:12:28 dos1, DON'T! :) Jul 15 15:12:57 although 2007.2 did things ok Jul 15 15:12:57 I bet under the hod Jul 15 15:12:57 hood* Jul 15 15:12:57 things were a bit hackish because of tight deadlines :) Jul 15 15:13:24 dos1: then prepare common infrastructure from both solution Jul 15 15:13:39 (fso-based or direct libalsa calls) Jul 15 15:13:48 "" please don't do command line when you can do C calls"" please elaborate Jul 15 15:13:53 dos1: and libalsa calls are really simple Jul 15 15:13:56 ptitjes, ++. Jul 15 15:14:09 dos1, or just implement it in frameworkd ;) Jul 15 15:14:11 I made a JNI wrapper for that some years before Jul 15 15:14:19 DocScrutinizer, he meant Jul 15 15:14:24 don't use system() Jul 15 15:14:29 when you can just call a function :) Jul 15 15:14:34 mrmoku: cornucopia Jul 15 15:14:42 not in frameworkd Jul 15 15:14:45 please Jul 15 15:14:55 hmm, ok. At least as long it's exactly 100:100 Jul 15 15:15:02 ptitjes: i reminds you that we're still using frameworkd ;p Jul 15 15:15:15 s/reminds/remind/ Jul 15 15:15:15 dos1 meant: ptitjes: i remind you that we're still using frameworkd ;p Jul 15 15:15:37 in some days, fsodeviced will be able to run alongside frameworkd without device subsytem Jul 15 15:15:52 dos1: mickey said these days Jul 15 15:15:58 and fsodeviced still isn't ready for inclusion to SHR Jul 15 15:15:59 maybe next week it will be ready Jul 15 15:16:06 maybe Jul 15 15:16:22 dos1: I said maybe meaning if not sooner Jul 15 15:16:36 we're already using fsousaged... but fsodeviced is terrible now Jul 15 15:16:36 dos1: and you won't have the infrastructure ready at that time Jul 15 15:16:47 dos1: so please do the important stuff Jul 15 15:16:56 dos1: put settings in preferences Jul 15 15:17:13 settings = min, max, current Jul 15 15:17:17 for each control Jul 15 15:17:50 (controls= mic level and speaker level for each of the scenarios (BT headset, normal, handset ...)) Jul 15 15:18:13 then make prototype wrapper Jul 15 15:18:19 do gui around that Jul 15 15:18:31 and this will have been days fsodeviced runs correctly Jul 15 15:18:32 :D Jul 15 15:19:06 dos1: don't you agree ? Jul 15 15:20:23 ptitjes: i was thinking only on simple elementary slider with one-line callback, which will be changed into another one-line dbus callback later ;x Jul 15 15:21:07 dos1: no conf, nothing ? Jul 15 15:21:27 dos1: you would force us to remember the exact settings and redo them every time we call ? Jul 15 15:21:39 not now (shouldn't conf be handled by framework?) Jul 15 15:21:50 ptitjes: it would be better than now ;p Jul 15 15:22:00 dos1: then I would say, implementing nothing is better than implementing something than won't be useful Jul 15 15:22:16 dos1: then design the other way round Jul 15 15:22:23 dos1: features to implementation Jul 15 15:22:28 not the other way round Jul 15 15:22:57 ptitjes: but adding slider to gui has to be done whatever infrastructure you do Jul 15 15:23:04 :P Jul 15 15:23:18 dos1: you told it: just adding a slider is nothing Jul 15 15:23:24 dos1, actually, best thing Jul 15 15:23:24 would be Jul 15 15:23:24 writing an app Jul 15 15:23:24 (in the meanwhile) Jul 15 15:23:27 dos1: then implement the rest Jul 15 15:23:30 but it has to be done Jul 15 15:23:32 that allows you to change major parameters Jul 15 15:23:35 it is not so much additional work Jul 15 15:23:42 and then dump them to a custom state file Jul 15 15:23:42 but is what will make the feature usefull Jul 15 15:24:07 I don't want to have to tell every of my parties: hey please let adjust the volumes... do you here me correctly ? Jul 15 15:24:08 this will let us "play" with the values Jul 15 15:24:08 easily Jul 15 15:24:08 in a real time environment Jul 15 15:24:08 and not in "the lab<" Jul 15 15:24:23 ptitjes: hmm... but what are you doing now? :P Jul 15 15:24:30 anyhow, I'm off. Jul 15 15:24:39 "hey, please let me go into terminal, open alsa mixer and adjust volume"? ;P Jul 15 15:24:45 dos1: I adjust the volumes with amixer, save them once for all Jul 15 15:25:02 dos1: what you have to implement is something for which I don't have to do that Jul 15 15:25:13 dos1: and additionnaly fine-tune occasionaly Jul 15 15:25:20 i have? :x Jul 15 15:25:23 dos1: fine-tune means fine-tune Jul 15 15:25:29 not re-tune from scratch Jul 15 15:25:35 thus -> preferences!!!! Jul 15 15:25:48 then write it ;) Jul 15 15:25:51 dos1: I mean if you accept your mission ;) Jul 15 15:25:55 hehe Jul 15 15:26:13 but that's long term mission for me, as i'm not C programmer Jul 15 15:26:17 that IRC channel will auto-destroy itself in 5 Jul 15 15:26:22 ... Jul 15 15:26:47 well.. maybe not long term, but not short term too :P Jul 15 15:26:49 dos1: oh really this is a nice choice for a first project Jul 15 15:26:53 :p Jul 15 15:27:22 (and obviously one of the most wanted missing feature ;)) Jul 15 15:27:24 for the time being Jul 15 15:27:30 a volume slider while on call is enouygh Jul 15 15:27:33 -y Jul 15 15:27:37 first step before Jul 15 15:27:46 then we'll talk about saving =) Jul 15 15:27:54 pleaaaase Jul 15 15:27:54 Ainulindale: ++! Jul 15 15:27:55 :) Jul 15 15:28:01 ptitjes: nah Jul 15 15:28:09 storing|retrieving values is two dbus calls more Jul 15 15:28:09 ptitjes: You're always keen on doing everything at once Jul 15 15:28:13 That's not how it works that's all Jul 15 15:28:15 and all i want to do at the moment is being able to simple turn volume down when i'm in very quiet room with other person Jul 15 15:28:21 please don't do interim stuff that stays as is for months Jul 15 15:28:26 i don't want him to hear everything my caller says Jul 15 15:28:27 That's not interim stuff Jul 15 15:28:32 We'll always need a volume slider Jul 15 15:28:49 It's available on all the phones I know about Jul 15 15:29:16 Ainulindale: exactly and on these phones Jul 15 15:29:24 modified levels are stored Jul 15 15:29:32 Then we'll do that on a second step as I just told Jul 15 15:29:50 and there are decent min and max that are at the 0 and 100% position of the slider Jul 15 15:29:56 As Rome, FSO wasn't build in one day, it's incremental, why should we do any other way? Jul 15 15:30:42 ptitjes: If you want that, do it yourself, as dos1 doesn't feel comfortable with C, it's sensible to do it in small steps Jul 15 15:30:42 min and max are only a math "base de 3" (don't know the name in english) Jul 15 15:30:49 Furthermore, everybody I know does it that way Jul 15 15:30:52 But you Jul 15 15:30:53 ptitjes: so. i'll add slider, and you'll add that two dbus calls more. ok? :P Jul 15 15:31:02 So.. :-> Jul 15 15:31:04 Ainulindale: as with Rome as with anything to do with OM occasionally got knocked down and re-implemented :) Jul 15 15:31:14 tig|: meh :-) Jul 15 15:31:25 ptitjes: règle de 3 you mean? Jul 15 15:31:35 yep Jul 15 15:31:44 tig|: exactly Jul 15 15:31:45 trojan rule? :-> Jul 15 15:32:03 tig|: everytime the same train Jul 15 15:32:19 ptitjes: You always try to tell that there is no way we can do that incrementally Jul 15 15:32:22 Whereas ther eis Jul 15 15:32:23 It's obvious Jul 15 15:32:28 So dos1 will do that his way Jul 15 15:32:35 If you're not satisfied, go ahead and add two dbus calls =) Jul 15 15:32:51 won't do that in C Jul 15 15:32:55 and then I can use my phone on the same train :) Jul 15 15:33:01 Anyways I'm sure he'll do it when he'll be satisfied with his stuff Jul 15 15:33:07 Plus, ptitjes Jul 15 15:33:11 When you develop, when you test Jul 15 15:33:17 You don't build a wall as is Jul 15 15:33:20 You use bricks Jul 15 15:33:26 yep Jul 15 15:33:28 Step by step, incremental, bricks, it's the same Jul 15 15:33:34 So let him do it that way, stop whining, get to work =) Jul 15 15:33:35 Ainulindale: but those bricks are not rect-shaped Jul 15 15:33:46 but they have shapes that must be thought before Jul 15 15:33:48 ptitjes: stop fiddling around my metaphors Jul 15 15:33:53 in order they imbricate correctly Jul 15 15:34:04 A volume slider is something we need, we talked about that six months ago, we even have mockups screens for that Jul 15 15:34:13 Ainulindale: indeed this was a good metaphore :) Jul 15 15:34:15 * mwester has missed the inane chatter whilst he was travelling... :p :D Jul 15 15:34:16 There is no need to digress endlessly Jul 15 15:34:26 http://shr.bearstech.com/Makefile <--- latest makefile, right? Jul 15 15:34:31 Ainulindale: what mockups? Jul 15 15:34:33 TAsn: no Jul 15 15:34:34 but surely with enough mortar to glue them together we can fix the two different brick shapes together to build a strong wall :) Jul 15 15:34:37 Ainulindale: yeah but what is important is what is under the slider Jul 15 15:34:38 TAsn: but will update to latest Jul 15 15:34:41 * mwester still thinks Ainulindale needs a large whip. ;) Jul 15 15:34:44 ptitjes: no Jul 15 15:34:45 mockups don't say everything Jul 15 15:34:46 Ainulindale, tell me when you are done ;) Jul 15 15:34:53 I want to get and start building :) Jul 15 15:34:55 else we can add a slider and make it inoperant :) Jul 15 15:35:09 I won't elaborate on that, dos1 feels comfortable with his way, I utterly agree with him, and it's a sensible way to do things Jul 15 15:35:10 just beep me when it becomes the latest :) Jul 15 15:35:17 So to me it's end of topic, I have work to do here Jul 15 15:35:23 mwester: why is that? :-) Jul 15 15:35:26 :) Jul 15 15:35:27 TAsn: it's done Jul 15 15:35:30 thanks. Jul 15 15:35:35 TAsn: I just told you Jul 15 15:35:37 Ainulindale, to keep your workers in line! Jul 15 15:35:39 it will update to latest Jul 15 15:35:43 automagically Jul 15 15:35:46 (and to use on your boss, too) Jul 15 15:35:54 Ainulindale, it seems that I'm laggy Jul 15 15:35:54 meh Jul 15 15:36:06 TAsn: nah I was a bit difficult to read :-) Jul 15 15:36:20 dos1: on trac Jul 15 15:36:23 for the mockups Jul 15 15:36:26 :) Jul 15 15:36:36 I'm a bit busy here reading specs Jul 15 15:36:37 Boring specs Jul 15 15:37:55 dos1: http://www.alasal.be/openmoko/shr2/ Jul 15 15:38:31 the slider is missing, it was on an older version IIRC Jul 15 15:38:47 but alasal (dolf) is missing in action Jul 15 15:39:28 Ainulindale, very nice pics :) Jul 15 15:39:34 Yep Jul 15 15:39:45 * TAsn is thinking about stealing the design :) Jul 15 15:39:45 And you didn't see my girlfriend yet Jul 15 15:39:55 who's the designer and can we use it? Jul 15 15:40:16 dolf Jul 15 15:40:19 and yes you can Jul 15 15:40:20 as we need a more phonish design, and less a regular GUI design for basic phone apps Jul 15 15:40:21 it's open Jul 15 15:40:23 it was on a ticket Jul 15 15:40:27 "better visual hierarchy" Jul 15 15:40:29 Ainulindale, super sweet :) Jul 15 15:40:32 consider it stolen Jul 15 15:40:33 :) Jul 15 15:40:37 Good =) Jul 15 15:40:48 And meanwhile Jul 15 15:40:54 Could you add an easter egg for me? Jul 15 15:40:57 I have a test tomorrow and after than I'm free :) Jul 15 15:41:03 Ainulindale, what do you need? :) Jul 15 15:41:06 Two dozens of useless sliders when phone detects it's being used by ptitjes Jul 15 15:41:13 Just to piss him off Jul 15 15:41:15 :DD Jul 15 15:41:20 Or better Jul 15 15:41:26 Two dozens of randomly assigned sliders Jul 15 15:41:27 lol Jul 15 15:41:59 :p Jul 15 15:42:28 Ainulindale, will de Jul 15 15:42:30 do Jul 15 15:42:33 I need one thing though Jul 15 15:42:38 you gotta get his imei Jul 15 15:42:44 Peh, easy Jul 15 15:43:00 I logged everything from his phone Jul 15 15:43:04 Although he can changed imei Jul 15 15:43:06 then it's a done deal :) will hopefully be included in the first testing image :) Jul 15 15:43:06 -d Jul 15 15:43:10 alphaone|gone, how? Jul 15 15:43:13 Ainulindale, * Jul 15 15:43:22 well it's doable on the moko IIRC Jul 15 15:43:59 super awesome! :) Jul 15 15:44:05 Ainulindale, then please get me his IMSI Jul 15 15:44:18 and let's just hope he's too lazy to get himself another sim Jul 15 15:44:18 :) Jul 15 15:44:36 well with the trouble he had making his one work... =) Jul 15 15:44:52 great! Jul 15 15:44:56 but more seriously Jul 15 15:44:59 ? Jul 15 15:45:00 I'm all for new design yes Jul 15 15:45:06 I'd also like an alternative sixteen Jul 15 15:45:08 blue/gray Jul 15 15:45:11 something lighter than black Jul 15 15:45:14 Ainulindale, my plans for this weekend: Jul 15 15:45:28 fix opkg's config handling Jul 15 15:45:36 what's to fix? Jul 15 15:45:37 start a major phonegui rewrite Jul 15 15:45:42 Ainulindale, I have a ticket open Jul 15 15:45:44 which language? Jul 15 15:45:46 #434 or something Jul 15 15:45:54 Ainulindale, C Jul 15 15:45:56 will remain in C Jul 15 15:45:57 :) Jul 15 15:46:02 Ok Jul 15 15:46:05 not really a complete rewrite Jul 15 15:46:07 just partial rewrites Jul 15 15:46:10 and fixes Jul 15 15:46:13 and enhancements Jul 15 15:46:22 hopefully I'll get a lot done Jul 15 15:46:24 :) Jul 15 15:46:25 to handle opimd? Jul 15 15:46:32 yeah Jul 15 15:46:35 that's great Jul 15 15:46:38 I'm all for that Jul 15 15:46:39 but not only. Jul 15 15:46:43 that is? Jul 15 15:46:49 ophonekitd needs also changes to handle opimd Jul 15 15:46:50 Ainulindale, I need a fucking working phone already :) Jul 15 15:46:53 (signal handling :P) Jul 15 15:46:58 dos1, will do that as well Jul 15 15:47:02 Well I could do that if necessary Jul 15 15:47:05 and remove phonelog from ophonekitd Jul 15 15:47:07 This week-end i'll hopefully have time Jul 15 15:47:12 Ainulindale, same here Jul 15 15:47:13 nice! Jul 15 15:47:14 I have big plans Jul 15 15:47:14 :) Jul 15 15:47:19 though not sure about the time I have :) Jul 15 15:47:20 TAsn: can't find the ticket for conf Jul 15 15:47:25 Ainulindale, sec. Jul 15 15:47:34 big plans are good, even if only small part is really done after all ;) Jul 15 15:47:45 http://shr.bearstech.com/trac/ticket/484 Jul 15 15:47:51 the fix would be Jul 15 15:47:57 fixing opkg's configuration handling Jul 15 15:48:01 Well ask me if I can help you now Jul 15 15:48:03 instead of individual packages Jul 15 15:48:04 Because starting august Jul 15 15:48:08 I'll have a huge learning curve for H:1 Jul 15 15:48:15 Ainulindale, H:1, why? Jul 15 15:48:28 Ainulindale, are you ditching shr? :) Jul 15 15:48:39 (traitor) Jul 15 15:48:41 No Jul 15 15:48:46 I've been hired by bearstech Jul 15 15:48:55 really? Jul 15 15:48:57 supersweet :) Jul 15 15:49:00 Part of my time will be being H:1 project leader Jul 15 15:49:16 I can't see any problem in this ticket TAsn Jul 15 15:49:16 Ainulindale, cool. Jul 15 15:49:24 Configuration files SHOULD be in packages Jul 15 15:49:25 Ainulindale, why not? Jul 15 15:49:32 look at this scenario: Jul 15 15:49:33 What's the problem with that? Jul 15 15:49:35 Ainulindale: i'll explain you Jul 15 15:49:36 you got phonegui.conf Jul 15 15:49:40 you change the home prefixes Jul 15 15:49:47 next update to phonegui Jul 15 15:49:51 Yeah? Jul 15 15:49:52 will tell you you have a clash Jul 15 15:49:57 in the configurations Jul 15 15:49:57 Ainulindale: we want to make opkg asking ONLY when conf in package has been really changed Jul 15 15:50:00 So what when he asks me what to do Jul 15 15:50:02 I diff Jul 15 15:50:03 And that's all Jul 15 15:50:07 Ainulindale, NO! Jul 15 15:50:09 dos1: that's already the case Jul 15 15:50:09 this sucks Jul 15 15:50:10 badly Jul 15 15:50:14 it's annoying Jul 15 15:50:15 Ainulindale, no Jul 15 15:50:18 Yes iti s Jul 15 15:50:20 dos1, explain as I gtg :) Jul 15 15:50:24 s/ti s/t is/ Jul 15 15:50:25 Ainulindale, dos1 will explain Jul 15 15:50:26 :) Jul 15 15:50:26 Ainulindale meant: Yes it is Jul 15 15:50:36 trust me Ainulindale I usually don't bitch without a reason ;) Jul 15 15:50:39 Ainulindale: if user changed conf, and package has still the same conf, then opkg shouldn't ask Jul 15 15:50:45 Ainulindale: it asks now :P Jul 15 15:50:59 dos1: two things herre Jul 15 15:51:00 -r Jul 15 15:51:04 dos1, good and to the point. ;) Jul 15 15:51:05 opkg install a Jul 15 15:51:12 opkg install -force-reinstall a Jul 15 15:51:14 will not ask Jul 15 15:51:17 opkg install a Jul 15 15:51:25 echo "boo" > /etc/a Jul 15 15:51:28 opkg install -force-reinstall a Jul 15 15:51:29 will ask Jul 15 15:51:35 exactly Jul 15 15:51:36 And that's normal Jul 15 15:51:38 which is BAD Jul 15 15:51:39 And that's intended Jul 15 15:51:40 no it's not Jul 15 15:51:40 no Jul 15 15:51:45 Yes it is Jul 15 15:51:49 take the phonegui conf for example Jul 15 15:51:52 Because you can install alternative package versions Jul 15 15:51:58 Whether the conf has changed or not Jul 15 15:52:01 dos1, please explain :) Jul 15 15:52:03 I'm really going Jul 15 15:52:04 ciao. Jul 15 15:52:11 And you can't manage conf versions through the different package revisions Jul 15 15:52:20 You only can do one thing which might be of interest to you but which is difficult Jul 15 15:52:23 Ainulindale: dpkg is doing it correctly Jul 15 15:52:28 It's to handle upgrade revisions at postinst Jul 15 15:52:32 That's what I was willing to do Jul 15 15:52:33 That is Jul 15 15:52:38 if I upgrade from X to Y Jul 15 15:52:39 Then do that Jul 15 15:52:53 That's doable, that's interesting, but that won't change the problem you'll have for that specific case Jul 15 15:53:02 Although you could sed or whatnot Jul 15 15:53:02 Ainulindale: "IF user_has_changed_conf AND new_package_brings_new_conf THEN ask" Jul 15 15:53:20 dos1: you can't know user has changed conf from opkg point of view Jul 15 15:53:29 "IF user_has_changed_conf AND new_package_has_still_the_same_conf_as_before THEN dont_ask" Jul 15 15:53:30 Not directly Jul 15 15:53:38 rpm does that ;) Jul 15 15:53:45 and dpkg does that too Jul 15 15:53:48 Well I guess it has samples Jul 15 15:53:51 opkg doesn't Jul 15 15:54:03 I think you're focusing on something gadget Jul 15 15:54:04 then we should implement samples in opkg ;) Jul 15 15:54:07 First step to me should be Jul 15 15:54:13 Handle upgrade postinst Jul 15 15:54:22 That is being able to do postinst operations when we upgrade from X to Y Jul 15 15:54:28 (c.f. e .so) Jul 15 15:54:32 as now it's REALLY BAD Jul 15 15:54:39 That is the first, simple step, of this kind of thing Jul 15 15:54:40 as you can have you're own framework configuration Jul 15 15:54:44 It's pr etty straightforward Jul 15 15:54:46 Yes we can Jul 15 15:54:48 proof is, I do Jul 15 15:54:51 It's just annoying Jul 15 15:54:54 Ainulindale: no Jul 15 15:55:01 Well prove me I can't do it Jul 15 15:55:06 I've been doing it for 6 months :-) Jul 15 15:55:14 Ainulindale: it's not only annoing, it's bad Jul 15 15:55:17 Ainulindale: cause Jul 15 15:55:20 Don't do your extremist :-) Jul 15 15:55:35 dos1: Is it annoying? Yes Jul 15 15:55:40 Ainulindale: imagine situation, when new subsystem comes to frameworkd Jul 15 15:55:44 Does it forbid you to manage your system normally? No Jul 15 15:55:58 Ainulindale: and it needs some default config Jul 15 15:55:59 Does not being able to follow upgrade path is more important? Yes :-) Jul 15 15:56:06 dos1: diff Jul 15 15:56:11 that's what I did today :-) Jul 15 15:56:12 (for opimd call) Jul 15 15:56:27 dos1: I got your point don't worry Jul 15 15:56:29 Just telling you that Jul 15 15:56:31 In my opinion Jul 15 15:56:33 Ainulindale: are you looking at diffs every upgrade? Jul 15 15:56:37 Yes Jul 15 15:56:42 In my opinion, was I saying Jul 15 15:56:44 Ainulindale: in frameworkd.conf it's really simple to miss something Jul 15 15:56:57 it's more important to have something to handle properly upgrade path Jul 15 15:56:59 as it gets rearranged after changing some settings cia dbus interface Jul 15 15:57:01 Which is the first step of your idea Jul 15 15:57:06 Then do what you're describing Jul 15 15:57:12 You won't be able to if you can't handle upgrade path properly Jul 15 15:57:14 heh ;) Jul 15 15:57:23 And that is something lacking today Jul 15 15:57:25 i'm for implementing both :P Jul 15 15:57:37 dos1: yeah but conf management depends on upgrade path sadly Jul 15 15:57:53 dos1: for your issue with frameworkd.conf Jul 15 15:57:59 if it gets rearranged not in the same order Jul 15 15:58:00 Then it's bad Jul 15 15:58:06 And shouldn't happen Jul 15 15:58:12 It should always be written in the same order Jul 15 15:58:14 BTW. when config has changed, and user didn't change conf, it shouldn't ask too Jul 15 15:58:25 opkg asks, rpm and dpkg don't ask Jul 15 15:58:39 dos1: On that specific one I agree, still, diff does it job Jul 15 15:58:44 What I want to say is, to sum up Jul 15 15:58:46 You can do that with a diff Jul 15 15:58:49 Sure it's bad and annoying Jul 15 15:58:50 But you can Jul 15 15:58:56 You can't manage upgrade path though Jul 15 15:58:59 And that may break your system Jul 15 15:59:03 That's more important in my opinion Jul 15 15:59:11 Ainulindale: you can answer calls with vala-terminal and mdbus call Jul 15 15:59:11 We have more users complaining about a broken system Jul 15 15:59:17 it's annoying, but you can ;) Jul 15 15:59:23 dos1: yes of course Jul 15 15:59:24 and guess what Jul 15 15:59:30 how did I do at the beginning? Jul 15 15:59:34 :) Jul 15 15:59:38 While I was developing libframeworkd-phonegui Jul 15 15:59:48 (I'm a bit lying there I was using zhone :-p) Jul 15 15:59:52 But still Jul 15 15:59:59 Sure it's annoying Jul 15 16:00:03 But it's not THAT annoying Jul 15 16:00:09 Whereas a broken system is VERY annoying :-) Jul 15 16:00:16 Especially for a non technical user Jul 15 16:00:38 i'm specific - broken system means "more fun" for me :D Jul 15 16:00:53 and anyways dos1 what I just meant is, you won't be able to do what you want if you're not able to have some package information in postinst Jul 15 16:01:02 And that will allow you to do upgrade path management Jul 15 16:01:08 that is from_version to_version Jul 15 16:01:21 which is the first step to handle your thing, IMHO Jul 15 16:01:33 so be a nice guy with me and give me some candy :-p Jul 15 16:10:44 DocScrutinizer: (have two bats drivers at the same time) i thought about that. Possible solution, in fact i'd say desirable. I wonder what raster will say about it though. Jul 15 16:13:56 hi!, where pyphonelog store it db, and why sometimes after opkg update I loose info there? Jul 15 16:13:56 mrmoku: (audio.conf) In fact this setting doesn't have much sense at all. When set to HCI it disallows Play() method. When set to PCM it allows it. IIRC it doesn't do anything more. And patching bluez to autodetect that is hard and would work only for CSR chips anyway. Jul 15 16:15:38 max_posedon: cause recently it changed from ophonekitd database to opimd interface Jul 15 16:15:56 PaulFertser: I don't see what can be the problem. The driver are reporting if the battery is present or not. imho it should already be handled. it's useless to try to report the charge for an absent battery :) Jul 15 16:16:34 dos1, nice, but, wasn't it possible to moving without "clear"? Jul 15 16:16:51 max_posedon: I love it: "sometimes" :-) Jul 15 16:16:58 max_posedon: you're just a whiny user after all :-p Jul 15 16:17:28 max_posedon: well, old database is still on your neo :P Jul 15 16:17:29 rtp: but if bq27000 is there then the battery considered present. And at the same time this battery voltage can be measured externally, so "gta01-battery" is also present. Jul 15 16:18:45 PaulFertser: you mean that the present sysfs file for both drivers is always reporting '1' ? Jul 15 16:18:48 ptitjes: by the way, still waiting for the videos Jul 15 16:18:48 dos1, may be you can suggest export/import command?) Jul 15 16:18:55 rtp: yes Jul 15 16:19:09 Ainulindale, I like that I can update without reflashing) Jul 15 16:19:11 max_posedon: write python script for that, it shouldn't be hard :P Jul 15 16:19:13 PaulFertser: that's surely a bug :((( Jul 15 16:19:16 Ainulindale: then ask deubeuliou Jul 15 16:19:20 max_posedon: so do I Jul 15 16:19:31 ptitjes: he supposedly uploaded them somewhere and give you the link Jul 15 16:19:35 s/give/gave/ Jul 15 16:19:35 Ainulindale meant: ptitjes: he supposedly uploaded them somewhere and gave you the link Jul 15 16:19:42 That's what he told me anyways Jul 15 16:19:50 nah Jul 15 16:19:51 He's on holiday so... Jul 15 16:20:08 Damn it's been several months, can't you guy see each other? =) Jul 15 16:20:16 rtp: if it is a bug, then how do you suppose it should work properly? Jul 15 16:20:34 rtp: i mean how technically to do that? Which driver modify and how? Jul 15 16:20:58 i'm quite sure present is not set when the battery is not inserted Jul 15 16:21:21 larsc: for gta02 battery yes Jul 15 16:21:25 yes Jul 15 16:21:28 ah Jul 15 16:21:40 larsc: we're talking about having an alternative driver for "dumb" batteries. Jul 15 16:21:51 ptitjes: http://alt-tab.org/?post/1258 <= you'd like that I'm sure =) Jul 15 16:21:58 larsc: i want to use gta01-battery driver when bq27000 is not available. Jul 15 16:22:44 larsc: but the question is how to do it in a nice way. unbinding bq27000 and binding gta01-battery works well but some folks would like to have that automated whenever communication with bq27000 permanently fails. Jul 15 16:24:10 larsc: i'm going to add a simple linear lookup approximation of charge level based on voltage to the gta01-battery driver (as it's needed by the userspace anyway). So it'll be a good candidate for use on gta02 when you use a dumb battery. Jul 15 16:24:39 PaulFertser: the sysf present file is meant to report if the battery is present or not, so if it's always '1', the driver is buggy Jul 15 16:25:44 rtp: if you take out the battery from a running gta01 you're screwed anyway. Jul 15 16:26:18 dos1, where I can find opimd db? Jul 15 16:26:27 rtp: and when a smart battery inserted in gta02 it's present in the sense that both drivers (gta01-battery and bq27000) can work with it. Jul 15 16:26:43 rtp: that way we'll have 2 present batteries and it's wrong. Jul 15 16:26:49 max_posedon: mdbus ;P Jul 15 16:26:59 max_posedon: it's dbus interface Jul 15 16:28:28 its easy to say, but hard to use, because of deprecate worning( Jul 15 16:28:41 max_posedon: wtf? Jul 15 16:29:05 "Using **pending_return in dbus_connection_send_with_reply_setup() without pending_setup is deprecated and strongly discouraged" annoing me Jul 15 16:29:15 max_posedon: just ignore it Jul 15 16:29:15 ok, I do 2>/dev/null now) Jul 15 16:29:56 well, we could pass the simple driver platform data which contains a pointer to the bq27000(or any other advanced battery driver) and the simple driver must query the advanced driver if the advanced driver recongnises the current battery. And only if not the simple driver takes control of the battery. Jul 15 16:30:10 PaulFertser: that's the problem.. in a perfect world the gta01 driver on gta02 should"refuse" to handle the battery. Or maybe the bq27000 driver should "claim" the battery, preventing the gta01 to handle it Jul 15 16:30:29 dos1, as I said its easy to say) hard to ignore it, when I see it. Jul 15 16:30:40 rtp: i don't think that power_supply interface was developed with "claiming" in mind. Jul 15 16:30:57 also, I can't find opimd dbus API to store that data Jul 15 16:30:59 larsc: that has no chances to be merged imho Jul 15 16:31:09 (don't see how enter timestamp) Jul 15 16:31:20 larsc: too complicated. I prefer userspace binding/unbinding corresponding drivers based on kernel messages about HDQ communication. Jul 15 16:31:33 dos1, can you point me to correct API? Jul 15 16:32:06 PaulFertser: I know. I was thinking about making a platform device claimed by both drivers or something like that Jul 15 16:32:31 max_posedon: i'm ignoring it easly all the time ;P Jul 15 16:32:48 max_posedon: and enter timestamp in seconds after unix epoch Jul 15 16:33:18 ok, but, where? I don't see opimd api to do this( Jul 15 16:34:12 org.freesmartphone.opimd /org/freesmartphone/PIM/Messages org.freesmartphone.PIM.Messages({HERE-INSERT-DATA}) Jul 15 16:34:23 and that {HERE... is just dict Jul 15 16:34:38 {'Direction': 'in', 'Timestamp': 123456, ...} Jul 15 16:34:42 Messages? or Calls? Jul 15 16:34:52 I always thought messages is sms Jul 15 16:36:17 Initialized empty Git repository in /home/tom/projects/openmoko/oe/downloads/git/git.kernel.org.pub.scm.network.connman.connman.git/.git/ Jul 15 16:36:50 max_posedon: sorry, of couse calls Jul 15 16:37:01 and the last one should be org.freesmartphone.PIM.Calls.Add(... Jul 15 16:37:46 bah, stuck at this for about 40 mins now ;\ Jul 15 16:43:36 mrmoku: gimmie shr-installer in feeds! Jul 15 16:43:37 :D Jul 15 16:44:07 i'm compiling now phonegui-efl with useless slider ;) Jul 15 16:46:23 oh, i have to get etk compiled firstly :( Jul 15 16:49:54 has anyone stumbled across this issue? (the one I'm describing, this stuck on the line I pasted for hours) Jul 15 16:51:02 * mrmoku had to eat :) Jul 15 16:51:26 TAsn, unusual for kernel git Jul 15 16:51:41 mrmoku: http://tinderbox.openembedded.net/public/logs/task/1843033.txt Jul 15 16:51:55 mrmoku: could you help me with that? :x Jul 15 16:52:07 mrmoku, thanks (I don't know if I should be happy that it's usual, or not) Jul 15 16:52:10 food. Jul 15 16:52:52 TAsn, guess it would be easier for you if it would be usual :P Jul 15 16:53:00 TAsn: he said that it's *un*usual ;) Jul 15 16:53:15 dos1, recipe? Jul 15 16:53:30 mrmoku: etk Jul 15 16:54:00 from shr/import, without any changed Jul 15 16:54:04 -d+s Jul 15 16:54:18 hmm Jul 15 16:55:21 dos1, trying to reproduce Jul 15 16:58:13 mrmoku: i have that since ages, but now it's blocker :( Jul 15 17:01:58 dos1, comparing with my configure log shows me that Jul 15 17:01:59 configure.ac:23: installing `./config.rpath' Jul 15 17:02:09 config.rpath is there for me but not for you Jul 15 17:02:24 dos1, old build tree? Jul 15 17:02:38 or do you rebuild from scratch once in a while? Jul 15 17:02:47 mrmoku: i'll clean and try again Jul 15 17:03:03 mrmoku: old tree, but after oe merge Jul 15 17:03:04 rm -rf tmp; bitbake etk ;) Jul 15 17:03:18 the nice thing is... that will give you time to do nice other stuff :P Jul 15 17:03:18 noooo.... it will take ages :x Jul 15 17:03:43 you could try to clean the native autotools packages... Jul 15 17:03:43 mrmoku: i need libframeworkd-phonegui-efl compilling to do nice stuff :P Jul 15 17:03:56 mrmoku: but it depends on etk Jul 15 17:04:20 * dos1 prefers python way of developing ;D Jul 15 17:05:02 mrmoku, dos1, oh, oops, I saw "usual" :) Jul 15 17:05:06 * mrmoku prefers the C way of running apps ;) Jul 15 17:05:10 anyhow, for some reason it's working now :) Jul 15 17:05:11 mrmoku: how far have you got with libframeworkd-phonegui-efl2? Jul 15 17:05:29 dos1, duh, that's easy :) Jul 15 17:05:51 (python dev) Jul 15 17:05:56 as you can easily do trial and error Jul 15 17:05:57 :) Jul 15 17:05:58 bumbl, I'm still stuck with strange segfaults inside lfg... put it on ice some days ago Jul 15 17:06:01 and add debug info Jul 15 17:06:23 mrmoku: even after my change regarding paths? Jul 15 17:06:23 bumbl, but now ptitjes has provided me with an async capable libso-glib Jul 15 17:06:41 * dos1 agrees with mrmoku about C way of running apps, but at this stage of development /me prefers easier dev than faster apps :P Jul 15 17:06:44 will try that one after finishing the testing-image-blues ;) Jul 15 17:08:02 TAsn, I think after fixing the bluez4 issue and correctly depending on and using fso-abyss I will build an unstable image Jul 15 17:08:18 if that one works I will stuff all the revs from that image into shr-testing and build that Jul 15 17:08:35 hmm well - have i mentioned i dislike the C(++) way of coding? atm i have to read a book about c++ - gc i miss you Jul 15 17:09:12 although gc for sure kills performance Jul 15 17:09:16 mrmoku, great! :) Jul 15 17:09:32 you rock ;) Jul 15 17:09:37 bumbl, gc? Jul 15 17:10:03 garbage collection? Jul 15 17:11:11 bumbl, well, programmers are lazy people, but don't bee too lazy ;) Jul 15 17:11:14 I don't like gc ;) Jul 15 17:11:33 pieterc, oh ;) Jul 15 17:11:40 * bumbl_ hates forced reconnects Jul 15 17:11:41 pieterc, I also hate gcs. Jul 15 17:12:01 update for bumbl : Jul 15 17:12:02 bumbl, gc? Jul 15 17:12:02 garbage collection? Jul 15 17:12:02 bumbl, well, programmers are lazy people, but don't bee too lazy ;) Jul 15 17:12:02 I don't like gc ;) Jul 15 17:12:10 +1 for hating gc Jul 15 17:12:18 I prefer RAII way Jul 15 17:15:20 pieterc: yep garbage collector Jul 15 17:15:32 ~lart gc Jul 15 17:15:32 * bzzbot_ smacks gc up side the head with a clue-by-4 Jul 15 17:17:00 Hi, can somebody point me where i can find DBoard specification ? Jul 15 17:17:13 hey, I've been reading Rebel Code, (a book thing) about the history of FOSS and linux's kernel in particular Jul 15 17:17:16 anyone read that one? Jul 15 17:18:02 it says how good free software / open source is, but it is not written that well Jul 15 17:18:24 so i thought, lets help that guy out... seems that he doesn't release his book under an open license ... Jul 15 17:18:27 bummer ;) Jul 15 17:21:30 is this right forum to ask about DBoard ? Jul 15 17:22:09 fabian81: yep Jul 15 17:22:24 have you searched the wiki? Jul 15 17:23:02 bumbl_: i did but might have overlooked it. Jul 15 17:23:49 i don't own a dboard but http://wiki.openmoko.org/wiki/Debug_Board seems to be a good starting point Jul 15 17:24:46 then you usually might check out v3 Jul 15 17:24:46 freesmartphone.org: 03mickey 07framework * rb3f5e4b380c9 10/framework/subsystems/ogsmd/ (3 files in 2 dirs): Merge commit 'origin/mickey/full-csd-handling' Jul 15 17:24:58 :) Jul 15 17:25:26 bumbl_: thanks.. Jul 15 17:25:31 dos1: talk to me about slider underlaying function Jul 15 17:28:56 dos1, yeah talk with DocScrutinizer about sliding :P Jul 15 17:29:00 fabian81: look people.openmoko.org/joerg/schematics Jul 15 17:30:17 mrmoku: ¿ :P ? Jul 15 17:32:16 DocScrutunizer: Thanks a ton Jul 15 17:32:27 freesmartphone.org: 03mickey 07framework * r7c09174fb4e8 10/ (ChangeLog setup.py): (log message trimmed) Jul 15 17:32:27 freesmartphone.org: * Merge mickey/full-csd-handling branch into master. This adds a new Jul 15 17:32:27 freesmartphone.org: configuration entry 'data-call-handler' in section [ogsmd] where a Jul 15 17:32:27 freesmartphone.org: handler binary can be specified that gets called whenever a data call Jul 15 17:32:28 freesmartphone.org: has been setup. This handler will be called with params , Jul 15 17:32:30 freesmartphone.org: where direction is "incoming" or "outgoing" (depending on whether you are Jul 15 17:32:32 freesmartphone.org: the data call initiator or peer), and port is a device node that can be Jul 15 17:32:47 fabian81: If you need further detail, then ping me here Jul 15 17:33:31 CSD - yyyaahhoOOOOO!! :-DD Jul 15 17:34:26 DocScrutinizer: Correct me if I am wrong, a DBoard is meant to be used to connect the NeoFreeRunner to the PC..DBoard acts as a bridge for low-level debugging . Right ? Jul 15 17:34:43 yes Jul 15 17:35:05 :) Jul 15 17:35:25 okay..I misunderstood that it is development board..good i came here..thanks Jul 15 17:35:41 basically it offers tty for system-console. And JTAG to unbrick Jul 15 17:36:43 fabian81: but also you get convenient way to access FR I2C-bus, and (iirc) IIS audiobus Jul 15 17:36:51 DocScrutinizer: I want to develop an application and port it on NeoFreerunner, what is the best place to begin ? Jul 15 17:37:12 your desktop X-system Jul 15 17:37:37 fabian81: for applications you don't need debug board Jul 15 17:38:10 all the e* stuff is X-based and runs on PC same way as it does on FR Jul 15 17:38:50 dos1: yes, i figured it out jst now. Jul 15 17:38:56 fabian81, unless your developing a core app and want to use the framework Jul 15 17:39:05 :) Jul 15 17:39:05 you're* Jul 15 17:39:29 DocScrutinizer: I have the app ready on my ubuntu, now I want to put it on real phone, is there a simulator i can first get it working on ? Jul 15 17:40:26 fabian81: it's *much* easier to go for real phone, than to do a interim qemu step Jul 15 17:40:34 My app is basically a simple audio in , some processing and audio out , with some fancy UI.. Jul 15 17:41:16 audio can't be emulated nor tested on PC Jul 15 17:42:14 dos1: I have a few function-"prototypes" for you ;-) Jul 15 17:42:30 DocScrutinizer: hmm? ;) Jul 15 17:43:08 oh well, seems nobody is interested Jul 15 17:43:28 DocScrutinizer: in which? Jul 15 17:45:51 DocScrutinizer: thanks for help, I will check the wiki for more info and land my questions here if any.. Jul 15 17:47:49 TAsn, mrmoku, just finished adding powering on and off of devices... testing it now... :) Jul 15 17:48:00 F4t, super sweet :) Jul 15 17:48:03 how did you do it? Jul 15 17:48:10 fso dbus Jul 15 17:48:18 because enabling and disabling the resource is BAD Jul 15 17:48:26 you should request and release the resource Jul 15 17:48:43 F4t, great Jul 15 17:48:48 look at the last screenshot on the page: http://www.assembla.com/wiki/show/shrdev/Screenshots Jul 15 17:48:50 F4t, bad. ;\ Jul 15 17:49:01 you should when you want to get the resource Jul 15 17:49:02 somebody should teach connman now to recognize it :P Jul 15 17:49:15 TAsn, request the resource? O_o Jul 15 17:49:26 to start a daemon that'll request the resource Jul 15 17:49:35 and when you want to release it, tell the daemon to release it Jul 15 17:49:50 how exactly do i request wifi resource? Jul 15 17:49:53 btw Jul 15 17:49:53 if you want to request it and it's off Jul 15 17:49:53 you should turn it on Jul 15 17:49:56 F4t: fsoraw Jul 15 17:49:59 but before turnning it off Jul 15 17:50:00 e.g. Jul 15 17:50:04 you should check the refcount :) Jul 15 17:50:04 fabian81: best would be if you could test stuff on a real neo - no need for a debug board if you just want to do app development though Jul 15 17:50:13 F4t, in other words, it's nasty business :) Jul 15 17:50:21 TAsn, why da hell should i implement resource handling logic? Jul 15 17:50:32 that's not resource handling logic Jul 15 17:50:36 well kinda Jul 15 17:50:43 I mean, the framework should handle it Jul 15 17:50:52 my job is either to turn it on when i need it, and turn it off when i dont, or reques wifi when i need it, and release it when i dont... NOT BOTH Jul 15 17:51:05 but because the way wifi is handled (being turned on and off with another app iirc) Jul 15 17:51:12 you gotta hack it to be correct Jul 15 17:51:28 i couldnt care less who else turns on and off wifi in the case of mokonnect... Jul 15 17:51:31 F4t, since you are a connection manager Jul 15 17:51:36 TAsn: F4t: FSOResourceAllocationWrapper ! Jul 15 17:51:42 i ama connection maker... not maintainer... Jul 15 17:51:43 you should turn the wifi on when you need it Jul 15 17:52:12 after ur done with making a wifi connection with mokonnect you can go fire a different application and turn off wifi for all i care... Jul 15 17:52:29 but after you asked for it Jul 15 17:52:29 other apps may ask it as well Jul 15 17:52:29 so you can't just turn it off Jul 15 17:52:29 though iirc there's an issue that causes a big power leak if you don't turn it off (and leave it automatic) Jul 15 17:52:39 wait for someone else's confirmation cause I'm talking about a 3 months old out of my head info Jul 15 17:52:39 :) Jul 15 17:53:15 DocScrutinizer, ^ Jul 15 17:53:17 i leave the option to the user, if he doesnt want to turn it of, mokonnect wont... Jul 15 17:53:31 brb, test session 1 Jul 15 17:53:58 TAsn: F4t, paul elaborated on this this very morning, epically Jul 15 17:54:03 I mean, that's why wifi is off and not set to automatic in the first place, isn't it? Jul 15 17:55:58 TAsn: see modprobe vs err unregister/unlink/dunno. As we have no other way to powerdown AR6000 Jul 15 17:56:02 ask PaulFertser Jul 15 17:56:13 or simply read backscroll Jul 15 17:56:27 haha Jul 15 17:56:33 I knew my memory didn't fail me ;) Jul 15 17:56:38 F4t, but I agree Jul 15 17:56:46 you shouldn't care about the edge cases Jul 15 17:56:54 since you are the connection manager Jul 15 17:56:56 no one else should touch it imho. Jul 15 17:57:50 TAsn: bah, conmngr should be called implicitly by allocating resource Jul 15 17:58:07 that's the correct chain of dependencies Jul 15 17:58:21 DocScrutinizer, he shouldn't start covering up a bug he's not responsible for Jul 15 17:58:30 ack Jul 15 17:59:48 TAsn: but what I say is: not conmngr should switch on(/off) wifi, rather switching on wifi (by resource alloc) should start conmngr Jul 15 18:00:01 mrmoku: please just give me ./config.rpath from etk... i just need quick and dirty solution :P Jul 15 18:00:52 TAsn: i think connection manager should call SetResourcePolicy, not RequestResource Jul 15 18:01:03 nope Jul 15 18:01:29 DocScrutinizer, he's a gui Jul 15 18:01:30 TAsn: you are supposed to be still connected to wifi after closing mokonnect :P Jul 15 18:01:40 dos1, and you will be Jul 15 18:02:13 bah, does *anybody* read what I write? Jul 15 18:02:25 DocScrutinizer: i read, but i don't get your point Jul 15 18:02:28 DocScrutinizer, too long for meh Jul 15 18:02:43 dos1, you just turn it on. Jul 15 18:02:46 DocScrutinizer, I also replied. Jul 15 18:02:47 :) Jul 15 18:02:55 F4t: please move to taiwan Jul 15 18:03:25 DocScrutinizer, i agree that in a perfect world i should request wifi device, and then release it.... but we live in a FAR FAR OH SO FAR from perfect world here... Jul 15 18:04:08 dos1, hehe Jul 15 18:04:12 if I start xchat, browser, ominstaller, whatever, then this app needs to *allocate* connectivity resource. Jul 15 18:04:31 DocScrutinizer, OH Jul 15 18:04:34 now I got you ;) Jul 15 18:04:35 F4t: maybe you Jul 15 18:04:41 DocScrutinizer: now i got you too ;D Jul 15 18:04:44 well yeah, it should be called Jul 15 18:04:48 if a connection is not available Jul 15 18:04:54 :) Jul 15 18:04:55 mrmoku: i just want to build libframeworkd-phonegui-efl :P Jul 15 18:04:58 i.e we should make a script called Jul 15 18:05:08 I_need_da_intranet Jul 15 18:05:26 dos1, doing -c configure now Jul 15 18:05:26 which will be called everytime an internet app is launched Jul 15 18:05:34 if there's no connection Jul 15 18:05:41 it'll launch mokonnect or whatever ;) Jul 15 18:05:42 DocScrutinizer, won't it require a specialised code on the APP side? Jul 15 18:05:49 DocScrutinizer, is this what you meant? Jul 15 18:05:55 F4t, no, it's linux Jul 15 18:06:01 F4t: needs fsoraw Jul 15 18:06:01 you just wrap with scripts :) Jul 15 18:06:11 TAsn, either that or a driver.... decide... Jul 15 18:06:20 TAsn: fsoraw Jul 15 18:06:28 something is gonna have to trigger connection making once midori tries to create a socket.... Jul 15 18:07:03 PLEASE see fsoraw, trac.fso Jul 15 18:07:15 * TAsn puts DocScrutinizer on ignore Jul 15 18:07:17 (j/k) Jul 15 18:07:19 ;) Jul 15 18:07:23 DocScrutinizer, I don't know fsoraw Jul 15 18:07:29 btw, pure linux-wise.... if i do "killall connmand && connmand".... or "killall connmand; connamnd" whats the difference? Jul 15 18:07:31 DocScrutinizer, the needed resource does not exist though Jul 15 18:07:53 F4t: first will start connmand only, if killall connmand will be successfull Jul 15 18:08:00 I don't want midori to request wifi when I have a working usb network for example Jul 15 18:08:05 ohhhhh..... that explains some problems :P Jul 15 18:08:06 mrmoku: possible. That's a flaw then in fso-spec Jul 15 18:08:06 so i'll just go study Jul 15 18:08:06 gl everyone Jul 15 18:08:06 and listen to the Doc :) Jul 15 18:08:09 F4t: seconds will start connmand, no matter what happend with killall :P Jul 15 18:08:12 so the requested resource should be Network Jul 15 18:09:37 mrmoku: exactly Jul 15 18:09:43 right....off to testing session 2 Jul 15 18:10:02 mrmoku, what if midori requests a local address? Jul 15 18:11:06 and this network resource probably will start conmngr, which in turn is supposed to have entries like "enable wifi" "dhcp to usb-network-host" etc Jul 15 18:12:48 F4t, good question :) Jul 15 18:13:00 though that's version 2.3, v0.4 could be to have a fixed AP configured and allocation goes to WiFi witch starts up with ifup script Jul 15 18:13:17 that's the most rudimentary version Jul 15 18:13:23 DocScrutinizer, yeah, depending on what is set in the current profile it might even fire up a gprs connection :_) Jul 15 18:13:37 exactly Jul 15 18:13:58 aaah, a pleasure to talk to system architects Jul 15 18:14:34 we just need people to make it real :P Jul 15 18:17:20 F4t: TAsn: http://trac.shr-project.org/trac/ticket/172, http://trac.freesmartphone.org/ticket/393 Jul 15 18:18:05 heyho Jul 15 18:18:32 mrmoku: if our destination is 5km to the south, first thing to do is stop people from running direction NE Jul 15 18:18:35 ;-) Jul 15 18:18:43 mickeyl: ping Jul 15 18:19:20 is there a frickin way i can save my preference of dont suspend the phone?! Jul 15 18:19:46 mrmoku: in the end if everybody walked south we might have been there meanwhile ;D Jul 15 18:20:04 morphis: pong Jul 15 18:21:28 F4t: for *this* you might use set_resource_policy on system start - as a botch until dos1|away will implement a decent persistent settings sorage Jul 15 18:21:41 sTorage Jul 15 18:21:49 DocScrutinizer, oh, I know that :) Jul 15 18:21:59 now that's fsoraw Jul 15 18:22:00 ffs :) Jul 15 18:22:11 okie :) Jul 15 18:22:17 is set_resource_plicy an application? cause if its not... screw this... Jul 15 18:22:52 mdus -s blafasl-app-policy-set Jul 15 18:23:36 err resource-policy-set, dunno. I bet blindly on mdbus Jul 15 18:25:57 dos1|away, got my mail? Jul 15 18:27:55 DocScrutinizer, if 5km south is a huge lake of water, but 100 meters to the NE is a tiny pond, no man who is thirsty gonna walk south.... Jul 15 18:28:11 morphis: pong Jul 15 18:28:58 mickeyl: great Jul 15 18:29:26 mickeyl: I saw you created some project repository on git.freesmartphone.org related to umaf Jul 15 18:29:54 F4t: please apply same amazing insight to our resources management problem we see next to everywhere on a embedded system like Neo Jul 15 18:30:03 somewhat, yeah. I plan to do some UI stuff (hopefully) in the near future Jul 15 18:30:12 reviving the original ideas for UMAF Jul 15 18:30:45 DocScrutinizer, all im saying is, if people need basic wifi, they first gonna make basic wifi, then think about ways to make it uber cool and uber correct... Jul 15 18:30:50 mickeyl: I played with the same thoughts but don't have some time until now Jul 15 18:31:30 mickeyl, which toolkit/language do you want to use? Jul 15 18:31:48 mickeyl: You already have some concret thoughts what to implement? Jul 15 18:32:12 whats the name of the service that registers to the GSM network at the booting of the phone? Jul 15 18:32:27 is it odeviced? Jul 15 18:32:31 F4t: I agree on that being a widespread custom. I have to state it's a known bug that often is introduced in design phase Jul 15 18:33:29 morphis: not really concrete. i want to start with a dialer w/ contacts and messages. While doing the work necessarry, create the infrastructure bits on the fly Jul 15 18:33:30 mickeyl: I still playing with the idea to implement some daemon which handles application lifecycle like android does Jul 15 18:33:56 playya: FSO 2.0 for the backend. For the frontend, i'm undecided. I almost settled on EFL and then I saw QML Jul 15 18:33:56 even offers some functionalities to create background tasks etc. Jul 15 18:34:10 more a general solution for users to create easily applications Jul 15 18:34:11 F4t, not quite Jul 15 18:34:16 morphis: something like a real init system? Jul 15 18:34:17 ophonekitd does that in shr Jul 15 18:34:22 morphis: upstart++ ? :) Jul 15 18:34:29 though it requests ogsmd (and odeviced?) Jul 15 18:34:31 mickeyl: not an init system Jul 15 18:34:34 iirc. Jul 15 18:34:43 morphis: then a launcher backend? Jul 15 18:35:29 morphis: sounds like our work is complementing that rather than overlapping. that's good Jul 15 18:35:30 something like: the applications starts, app runs, another app starts, the former app gets suspended Jul 15 18:35:54 *nod* Jul 15 18:36:08 very close to the lifecycle android defines for its applications Jul 15 18:36:20 mickeyl, I'll try to send my thought about ophoned to smartphone-userland the next days Jul 15 18:36:32 playya: good! Jul 15 18:36:57 mickeyl: I think I will do the same as playya about my thoughts cause I am running currently out of time Jul 15 18:37:31 morphis: cool. thanks Jul 15 18:38:04 ok I have to leave now Jul 15 18:38:07 byebye Jul 15 18:38:21 imho it would be nice to regsiter some regex to the message widget. like: +4923459793845 -> {send sms, add to contact, call} Jul 15 18:40:12 TAsn, anyway, i was just wondering, why cant i power on the wifi device, or get its state, or anything, while the phone tries to register to the GSM network... only when i get the message that there is no sim, then the dbus starts working again.... Jul 15 18:40:38 what exactly there is single threaded? Jul 15 18:40:44 ot blocking? or i dont know... Jul 15 18:40:47 or* Jul 15 18:41:11 F4t, I'm not sure about the single thread, though there might Jul 15 18:41:27 F4t, mickeyl is the frameworks main dev ask him ;) Jul 15 18:41:55 playya, aye. I think there's even a bug about it in the shr trac. Jul 15 18:42:12 don't you worry, it'll happen asap hopefully :) Jul 15 18:42:12 hmm? Jul 15 18:42:23 (your regex suggestion) Jul 15 18:42:40 F4t: from mdbus? Jul 15 18:42:42 s/bug/ticket/ Jul 15 18:42:42 :) Jul 15 18:42:43 F4t: or from your app? Jul 15 18:42:55 i had a patch for a number regex in the message dialog. Jul 15 18:42:58 nothing is blocking in FSO Jul 15 18:43:05 maybe its somewhere on my hdd Jul 15 18:43:19 although if everything is fully busy, it may appear as being blocked until the CPU has some cycles free again Jul 15 18:43:47 this will improve with FSO 2.0 Jul 15 18:44:05 as we can schedule more fine granular w/ task priorities Jul 15 18:45:31 mickeyl, from shr-settings Jul 15 18:45:59 perhaps shr-settings is using sync. dbus calls Jul 15 18:46:05 in that case everything is serialized Jul 15 18:46:23 mickeyl, yes but shr-settings isnt the one thats registering gsm now is it? Jul 15 18:46:40 i'm not familiar with shr-settings, sorry Jul 15 18:46:58 the gsm registration starts at boot, and while it is still trying... i cant access any fso.org devices... :-/ Jul 15 18:46:59 F4t, no, it's ophonekitd Jul 15 18:47:16 that certainly doesn't look normal Jul 15 18:47:23 i can do all kinds of things in parallel here Jul 15 18:47:29 whoever wants can try this, boot the shr without a sim card, and try to go to the wifi settings.... Jul 15 18:47:59 you'll keep getting "dbus is borked" message... until a notification will appear about gsm not able to register.. Jul 15 18:48:03 and then it all works... Jul 15 18:48:18 I'm afraid that sounds more like a problem with the shr infrastructure than with fso Jul 15 18:48:25 i wrote a bb for elmdentica: http://pastebin.com/f6e3ac1d8 Jul 15 18:48:33 mickeyl, could be... Jul 15 18:48:33 can someone test it? Jul 15 18:48:45 F4t, mickeyl, to me it seems Jul 15 18:48:51 that the framework just didn't finish loading Jul 15 18:49:17 * TAsn is defending the shr infrastructure! :) Jul 15 18:49:56 TAsn, if being registered to gsm, or failing registration to gsm is a condition to its loading, that would explain the blocking... Jul 15 18:50:03 it should be async... Jul 15 18:51:27 F4t, are you sure the gsm failure/success is the cause Jul 15 18:51:28 ? Jul 15 18:51:41 I think it's just the framework taking ages to load. Jul 15 18:53:46 playya, trying it now Jul 15 18:53:56 ok. thx Jul 15 18:54:12 did you already update the fso-vala? Jul 15 18:54:41 $ bitbake ophonekitd libphonegui-efl Jul 15 18:54:41 NOTE: Handling BitBake files: \ (0922/6568) [14 %] Jul 15 18:54:45 stuck here forever now Jul 15 18:54:59 :( Jul 15 18:54:59 man I hate this shit :) Jul 15 18:55:07 press enter Jul 15 18:55:10 F4t, hmm... booted without SIM and it seems to work all nicely (apart from telephony ;) Jul 15 18:56:12 F4t, what should I do exactly to test it? Jul 15 18:56:47 mrmoku, as soon as it boots, goto settings, and click on connectivity... Jul 15 18:57:04 instead of the regular wifi off->on thingy u'll see dbus is borked Jul 15 18:57:31 oh. mrmoku can you change the PV from 0.0.1 to 0.6.0? Jul 15 18:57:54 playya, before comitting, yeah Jul 15 18:58:00 ok Jul 15 18:58:09 F4t, especially you are talking about as soon as it boots Jul 15 18:58:18 I suspect it's the framework not finishing to load. Jul 15 18:58:22 F4t: thanks, you reminded me to change that "dbus is borked" messages Jul 15 18:58:37 people think that dbus is not running, or something like that when they see it :P Jul 15 18:58:50 TAsn, i dont believe in coincedence when it starts working the instance i get the gsm sim notification Jul 15 18:58:52 instead it is ophonekitd not up yet? Jul 15 18:59:26 F4t, you get the sim notification Jul 15 18:59:28 when everything is up Jul 15 18:59:54 as I said Jul 15 18:59:56 so you want to tell me, i can run shr-settings, goto wifi.. twice... before its booted? Jul 15 19:00:24 or more accurately, if i do have SIM card inserted, it still takes the same time? Jul 15 19:03:59 the framework just doesn't finish loading Jul 15 19:03:59 (or so it seems) Jul 15 19:03:59 even if it's a framework bug, I'm not sure it's important enough to dwell on it, since the framework is being rewritten Jul 15 19:04:01 and this is only a minor issue Jul 15 19:04:03 (few seconds hang at boot) Jul 15 19:04:22 few secodns = big delay :( Jul 15 19:09:49 F4t, yes Jul 15 19:09:52 even with sim Jul 15 19:09:53 SHR: 03seba.dos1 07shr-settings * r2f7184cc93c9 10/ (14 files in 2 dirs): Change "dbus is borked" or "cannot connect to dbus" messages to "Couldn't connect to FSO". Jul 15 19:10:22 it still takes the same time. Jul 15 19:10:22 I bet that when mrmoku said everything is normal, he assumed you knew about this lag :) Jul 15 19:10:41 :-/ could be... Jul 15 19:10:49 it sucks having to read/reply to sms messages with at commands Jul 15 19:11:06 F4t: X started != FSO started Jul 15 19:11:44 mrmoku: argh! add shr-installer to feed! Jul 15 19:11:46 :P Jul 15 19:11:57 dos1, STOP! Jul 15 19:12:00 dos1, got my mail? Jul 15 19:12:02 don't interrupt mrmoku Jul 15 19:12:06 mrmoku: got ;) Jul 15 19:12:09 I think he's playing with the testing image :) Jul 15 19:12:13 TAsn: he just added elmdentica to feed Jul 15 19:12:18 TAsn, nah... I want shr-installer in testing too ;) Jul 15 19:12:19 TAsn: that's why i'm angry :P Jul 15 19:13:06 mrmoku: about mail: thanks, i'm trying now Jul 15 19:14:07 mrmoku, you leave me no choice. Jul 15 19:14:18 dos1, you sholud stop developing altogether until mrmoku is done :) Jul 15 19:14:18 dos1, got it Jul 15 19:14:21 should* Jul 15 19:14:42 mrmoku, bah, I'm so tired + not ready for this test :) Jul 15 19:14:50 I'm hoping to ace it, though I'm so tireeed... ;) Jul 15 19:14:53 mrmoku: thanks ;) Jul 15 19:15:22 mrmoku: is it autorev? Jul 15 19:15:30 TAsn, stop bothering dos1 then and get some sleep ;) Jul 15 19:15:32 dos1, yes Jul 15 19:15:37 ok ;) Jul 15 19:15:48 then i'm trying to get libframeworkd-phonegui-efl compilling Jul 15 19:15:59 mrmoku, NO! I'm determined to shipping a new testing image this friday :) Jul 15 19:16:04 *on Jul 15 19:16:32 i want to be online today for about 5 hours, so i have time for developing ;) Jul 15 19:17:44 slider in phonegui-efl -> regexps in opimd -> looking at fixing packagekit bugs Jul 15 19:17:55 that's my todays todo Jul 15 19:18:42 unfortunately my OE env doesn't seem to build Jul 15 19:18:50 therefore I'm a bit stuck :) Jul 15 19:19:38 hm opkg upgrade stoped on frameworkd-config-shr ;/ Jul 15 19:22:10 dos1: it's your fault! ;> Jul 15 19:24:18 Blu3: hey, you're a dev after all. And you understand that amount of effort to get the ability to disable BT headset is neglegible comparing to get it working in the first place! Jul 15 19:24:45 heheh Jul 15 19:24:54 ok... who wants to test new mokonnect? :P Jul 15 19:25:13 now comes with the power of power Jul 15 19:25:16 lol Jul 15 19:25:52 :) Jul 15 19:26:09 F4t, one has still to restart connman manually though, right? Jul 15 19:27:00 Blu3: my last message was meant as a rant against your trolling e-mail on community ML, not as a joke for you to laugh ;) Jul 15 19:27:48 my last email was humor. "it should work" was followed by "you have to reboot to switch back to headset" Jul 15 19:28:01 which in my interpretation ... isn't working :> Jul 15 19:28:02 mrmoku, nop Jul 15 19:28:50 :D Jul 15 19:28:55 PaulFertser: wasn't that you who elaborated on connman & fsoraw and kill -9 etc this morning? Jul 15 19:28:58 F4t, do you check if wifi is already up Jul 15 19:29:03 before you decide to restart connmand? Jul 15 19:29:08 TAsn, yep Jul 15 19:29:15 was just verifying :) Jul 15 19:29:37 btw, cool, nice work :) Jul 15 19:29:42 mrmoku, is mokonnect default? Jul 15 19:29:47 (in non-lite) Jul 15 19:29:52 i even check if connman is no longer buggy and can pick up on the wifi device without restarting... Jul 15 19:30:11 F4t, lol Jul 15 19:30:20 that's a bit optimistic, don't you think? Jul 15 19:30:20 :) Jul 15 19:30:34 it actually can after the first time.... Jul 15 19:30:54 my thoery is that connmand is started too early in the boot process Jul 15 19:31:18 have you tried making it start later? Jul 15 19:31:28 nop Jul 15 19:31:32 :) Jul 15 19:31:53 but i have tried restarting it before powering on wifi.. Jul 15 19:32:33 DocScrutinizer: yes, it was me. When WiFi resource is not claimed, frameword unbinds s3c-sdio driver that in turn cuts the power to wlan chip and disables its driver. So it's equivalent to turning it off (or suspending the whole device). Jul 15 19:33:22 Blu3: nevertheless i hope to see some nice gui from you soon ;) Jul 15 19:34:18 PaulFertser: could you do F4t a favour and re-post your explanation about fsraw and how it fits in? ;-) Jul 15 19:34:23 nice gui comes after i get something that works which comes when i have time which comes after i get my phone back ;) Jul 15 19:34:52 PaulFertser, or do me a favour and not? :P Jul 15 19:35:36 F4t, interesting Jul 15 19:36:02 mrmoku: also interesting is fact, that etk passed configure and now it's compiling ;) Jul 15 19:36:05 mrmoku, ? Jul 15 19:36:12 and it passed compilation not :D Jul 15 19:36:14 dos1, :) Jul 15 19:36:18 :( Jul 15 19:36:20 mrmoku: big thanks :) Jul 15 19:36:24 s/not/now/ Jul 15 19:36:28 :) Jul 15 19:36:28 ahh :D Jul 15 19:36:39 just typo ;D Jul 15 19:37:09 populate_staging :) Jul 15 19:37:20 * dos1 is happy :D Jul 15 19:37:21 mrmoku, could u gen a new mokonnect package? im done with testing.... Jul 15 19:37:48 F4t, yeah, will soon start building stuff Jul 15 19:38:52 mrmoku, do you think we'll be able to ship on friday? Jul 15 19:38:55 ok, should i mark off v0.3? or wait until i finish setup.py and moving config to ~/mokonnect? Jul 15 19:39:02 ~/.mokonnect* Jul 15 19:39:09 completed! :D Jul 15 19:40:13 F4t, wait :) Jul 15 19:40:24 k :) Jul 15 19:40:24 actually don't Jul 15 19:40:27 ?? Jul 15 19:40:27 you made great progress Jul 15 19:40:46 but then i'd have to bug mrmoku to change the bitbake to 0.3 :P Jul 15 19:40:46 and feature changes Jul 15 19:41:06 TAsn: hey Jul 15 19:41:12 Blu3, hey ;) Jul 15 19:41:15 BluesLee, * Jul 15 19:41:18 'hallo Jul 15 19:41:23 F4t, he's used to people bugging him Jul 15 19:41:34 ;D Jul 15 19:41:36 PaulFertser, heh Koen Kooi responded to my mail regarding the bluetooth config Jul 15 19:41:41 Blueprobe can already do autodetection and more, so I don't see what the Jul 15 19:41:44 problem is. Jul 15 19:41:45 add mrmoku to your beeping names in your irc client and you'll want to die ;) Jul 15 19:41:57 beep Jul 15 19:42:02 TAsn, btw, whats the other way to take the version number from the svn? not everyone is writing in python...... Jul 15 19:42:20 F4t, huh? what do you mean? Jul 15 19:42:30 setup.py won't affect PV for some reason Jul 15 19:42:37 although I bet it should and we are just doing stuff wrong :) Jul 15 19:42:42 if i wanted version number to change without bugging mr moku, and im not using python, how would i do that? Jul 15 19:42:48 F4t, hm.. autoconfig scripts Jul 15 19:42:54 ? :) Jul 15 19:43:02 no idea Jul 15 19:43:08 I'm no bb expert Jul 15 19:43:19 any bb experts here? Jul 15 19:43:35 TAsn: its not stable;-) Jul 15 19:43:45 mrmoku: you better not mess with Koen. He knows everything! And if he doesn't he's right nevertheless! Ask mickey|dinner if unsure ;) Jul 15 19:43:50 F4t: np Jul 15 19:43:50 BluesLee, what's not stable? Jul 15 19:44:00 PaulFertser, how? Jul 15 19:44:06 TAsn: unstable is not stable Jul 15 19:44:17 therefore its calles unstable i guess Jul 15 19:44:22 hehe Jul 15 19:44:29 PaulFertser, hehe, so what do I do now? Jul 15 19:44:36 * mwester laughs as he reads PaulFertser's comment re: Koen Jul 15 19:44:39 mickey|dinner also eats dinner ages, did you notice that? :D Jul 15 19:44:53 dos1, yeah, he can eat for days ;) Jul 15 19:45:01 good food needs its time Jul 15 19:45:03 mwester: hey, give me a link please Jul 15 19:45:07 oh, PaulFertser, u changed it manually.. :-/ Jul 15 19:45:16 BluesLee, not in shr :) Jul 15 19:45:17 in shr Jul 15 19:45:22 TAsn: i had the following situation yesterday Jul 15 19:45:27 testing = experimental Jul 15 19:45:35 unstable = testing :) Jul 15 19:45:35 F4t: so, FSO has a list of resources. Each resource has a policy (auto|enabled|disabled). Default is auto. When resource is not "requested" (or claimed) by anything, it's disabled. After booting frameworkd turns off all the resources and policy is auto and nobody claims them. Jul 15 19:46:01 F4t, bumped the version Jul 15 19:46:28 TAsn: canola was active, incoming call, release, canola doesnt play anymore, suspend and gsm is broken Jul 15 19:46:58 TAsn: but your workaround with "sleep 3" works after a reboot Jul 15 19:47:09 BluesLee, :) Jul 15 19:47:10 F4t: fsoraw is the program that claims the resource you want on start and then starts the program you want. After fsoraw exits, the resource is released. When frameworkd sees that nobody uses the resource anymore and its policy is auto, it turns the resource off. Jul 15 19:47:30 PaulFertser, ok, I know what to do Jul 15 19:47:34 mrmoku, for future reference, how would one go about doing the version bumping useing a file from my svn? Jul 15 19:47:54 BluesLee, sounds like a bug to me :) Jul 15 19:48:04 F4t, as long as we build from svn we have to do that manually Jul 15 19:48:09 TAsn: gsm daemons stop working from time to time Jul 15 19:48:23 so you either can tag it (and hope we notice ;) or just tell us Jul 15 19:48:28 F4t: are you satisfied with this explanation? Jul 15 19:48:59 PaulFertser, yes, is it currently working? i mean can i requesta wifi resource now, hold it while i need it, release it and it would all work? Jul 15 19:49:17 BluesLee, are you using fsousage + fso-abyss? Jul 15 19:49:17 no idea if it's related Jul 15 19:49:17 mrmoku, I'm really not sure about it. Jul 15 19:49:17 I mean, I would assume the part that parses setup.py Jul 15 19:49:24 also parses the PV from the setup.py Jul 15 19:49:29 mrmoku, can the bitbake get another bitbake script from the svn and then use it? Jul 15 19:49:31 btw. i have script, which changes my ringtone every call ;D Jul 15 19:49:34 F4t: it should work provided you don't change resource's policy to "enabled" by other means (SHR settings probably can do that). Jul 15 19:49:36 we should just find how to use it :) Jul 15 19:50:22 PaulFertser, will it turn on the wifi for me if its off? Jul 15 19:50:33 F4t: who? Jul 15 19:50:53 PaulFertser, whoever gets my "gimme wifi resource" request... Jul 15 19:51:03 PaulFertser: shr settings calls odeviced IIRC. i'll change it to ophonekitd.Usage interface Jul 15 19:51:08 F4t: sure, feel free to reread my explanation earlier. Jul 15 19:51:17 F4t: ;) Jul 15 19:51:41 well, once dos1 provides a good example of how to do it, ill change too :P Jul 15 19:51:50 TAsn: whats that? Jul 15 19:51:52 F4t, no Jul 15 19:52:05 dos1: the idea is that if you just wrap wpa_supplicant in fsoraw, you shouldn't enable the resource at all, including any switches in shr settings. Jul 15 19:52:27 mrmoku, but.... but....... Jul 15 19:52:59 TAsn, don't think so... the PV is set in the bb and I don't think it gets overwritten by something else Jul 15 19:53:24 mrmoku, I can only hope ;) Jul 15 19:53:38 I don't recall seeing any documentation for the setup.py bitbake tools Jul 15 19:53:44 so I guess reading the source is must ;) Jul 15 19:53:49 will do this weekend Jul 15 19:54:01 F4t: you should keep wifi disabled in SHR Settings. And then just use fsoraw. Jul 15 19:54:37 * PaulFertser wonders why he's so poorly understood by F4t today. "Is it just the weather, or am i going mad?" Jul 15 19:57:18 hmm... OE git re-started to hang up on first connect :( Jul 15 19:59:47 mrmoku, is it normal that bitbake is still stuck here: NOTE: Handling BitBake files: \ (0922/6568) [14 %] ? Jul 15 20:00:15 usually means it's stuck trying to evaluate an AUTOREV entry Jul 15 20:01:56 mwester, how do I get past that? :) Jul 15 20:03:01 you can run bitbake with the correct debug options so that it will tell you which repo it is trying to reach to find out the latest rev, and then you can manually address the problem... Jul 15 20:05:48 PaulFertser, i understand you, its just that u came here too late.... maybe future versions will use that method.... Jul 15 20:05:48 PaulFertser, although the usage scenario of the mokonnect doesnt exactly match the resource taking idea.... Jul 15 20:05:49 Does somebody report problems with framework after yesterday/today updates? Jul 15 20:06:22 neo become very unstable in framework part, usually I can't make more then 2-3 calls, after this framework stops working. Jul 15 20:06:45 mrmoku, TAsn, and the rest, v0.3 is taged and bagged.... with wiki updated on the changelog :P http://www.assembla.com/wiki/show/shrdev/ChangeLog Jul 15 20:07:12 cool. Jul 15 20:07:13 :) Jul 15 20:07:19 max_posedon: i don't have such problems Jul 15 20:07:34 mwester, thanks. Jul 15 20:07:34 max_posedon: could you try to comment OccypyResource parts in rules.yaml and then test? Jul 15 20:07:35 http://pastebin.ca/1496161 Jul 15 20:07:45 last lines in frameworkd.log Jul 15 20:08:16 poke me when its in repo and someone tried the new features... Jul 15 20:08:51 dos1, I can add, that I can call to neo, hear usual bip, but no any messages in logs Jul 15 20:08:53 max_posedon: NO CARRIER? Jul 15 20:09:03 as you see Jul 15 20:10:46 I don't think that OccypyResource will help Jul 15 20:10:52 I have an easier idea. Jul 15 20:11:03 nvm I don't. Jul 15 20:11:11 SHR: 03seba.dos1 07shr-settings * r6f302dc8325c 10/data/po/ (pl/shr-settings.mo pl/shr-settings.po shr-settings.pot): [data] update pot file and polish translation Jul 15 20:11:33 mwester, http://pastebin.com/m3f6b955d Jul 15 20:11:50 so, looks like I'm alone with this? Jul 15 20:12:02 it's stuck gitting the framework ;) Jul 15 20:12:07 Ouch Jul 15 20:12:24 Well, if it was something you didn't really need, you could BBMASK it, but I think you need that one... Jul 15 20:14:03 max_posedon: yep... Jul 15 20:14:30 mrmoku: thanks, my LOCAL libframeworkd-phonegui-efl built nicely ;) Jul 15 20:14:45 dos1, how can I help find a root of this problem? some logs, running smth with some params? Jul 15 20:15:08 max_posedon: ask mickey|dinner when he ends eat his dinner ;) Jul 15 20:15:12 (e.g. I'm going to try make this bug reploducable) Jul 15 20:17:54 Hey there again Jul 15 20:21:01 SHR: 03mok 07shr-themes * r4993ea2e653d 10/frameworkd/frameworkd-config-shr/ (om-gta01/frameworkd.conf om-gta02/frameworkd.conf): frameworkd-config-shr: turn on the abyss Jul 15 20:21:11 SHR: 03mok 07shr-themes * r44bc3e259e18 10/ (28 files in 10 dirs): Merge branch 'master' of git+ssh://git@shr.bearstech.com/shr-themes Jul 15 20:22:18 mrmoku, poke, build packages, poke.... Jul 15 20:23:02 mwester, ;\ Jul 15 20:23:19 doesn't do anything Jul 15 20:23:19 just stays there ;\ Jul 15 20:24:02 F4t, yeah yeah... already doing that ;) Jul 15 20:32:04 johnsu01: ping. Could you suggest me the best place to get emacs23? Jul 15 20:32:27 slider is displayed :) Jul 15 20:32:32 crap connection... Jul 15 20:32:37 it doesn't do anything, but it is! : Jul 15 20:32:38 :D Jul 15 20:32:45 mrmoku, poke, build packages, poke.... Jul 15 20:32:46 PaulFertser: for the FR? I just build it Jul 15 20:33:07 22:24 < mrmoku> F4t, yeah yeah... already doing that ;) Jul 15 20:33:15 ahh cool, that i didnt see :P Jul 15 20:33:19 :) Jul 15 20:33:53 johnsu01: looking forward to use your fso.el :) I think SHR guys are doing all wrong, they should have used emacs as a frameworkd from the beginning ;) Jul 15 20:35:04 PaulFertser: Heh, I have to warn you there is some bitrot and things not working quite right :) Jul 15 20:35:15 PaulFertser: I should be able to get back to working on it more shortly Jul 15 20:36:38 johnsu01: is the UI really fast? Jul 15 20:36:56 johnsu01: or limited by elisp interpreter etc? Jul 15 20:37:09 PaulFertser: well, for some things, but for other things the dbus signals were taking quite a while to get through sometimes and I hadn't fully debugged it Jul 15 20:37:41 johnsu01: dbus speed can be improved, that's not really an issue. But if i look at Zhone e.g. i'm sure slowness comes from the UI side of it. Jul 15 20:38:00 yep, zhone is slow Jul 15 20:38:03 PaulFertser: Yeah, just like selecting menu options and moving around and things is pretty quick Jul 15 20:38:16 due to gui and x11 engine Jul 15 20:38:26 it redraws everything on every page change Jul 15 20:38:41 PaulFertser: dialing right now requires typing a number, which is quick but a little inconvenient Jul 15 20:38:47 PaulFertser: a "dialpad" is in the works Jul 15 20:39:05 also one nice thing about the emacs ui is that you get all of the history functions already, so you can easily scroll to previous numbers etc Jul 15 20:39:10 dos1: i've heard searching inside contacts is dead slow and inefficient in SHR too now. Jul 15 20:39:43 dos1: and in emacs i'd just press C-s and start typing for incremental search. You get it for free just being inside emacs. Jul 15 20:39:59 johnsu01: what about dialling numbers for contacts? Jul 15 20:40:10 PaulFertser: there is no search in shr now Jul 15 20:40:32 dos1: there is. You painfully scroll, look, compare using your eyes, scroll, look etc. Jul 15 20:40:52 PaulFertser: nothing for contacts yet, is on the todo list Jul 15 20:41:12 PaulFertser: urgency is low for me because I already store all my contacts in a plain text file Jul 15 20:41:28 johnsu01: that's ok, and how do you dial them? Jul 15 20:41:28 PaulFertser: so I just search and copy and paste the number -- but I do intend to make it easier Jul 15 20:41:50 probably for starters just a dial-number-at-point function Jul 15 20:42:24 PaulFertser, that is intended.... and called pattern-recognition-trainer ;) Jul 15 20:43:16 mrmoku: i still can't understand why not just implement an emacs mode as johnsu01 does... Jul 15 20:43:48 well... there are some downsides to emacs -- startup time for example Jul 15 20:44:39 johnsu01: and lack of threading Jul 15 20:44:50 * johnsu01 nods Jul 15 20:44:55 johnsu01: and ancient dialect of LISP Jul 15 20:45:02 emacs sucks, i know Jul 15 20:45:04 ;) Jul 15 20:45:13 PaulFertser: well, I meant using it on the FR had some downsides in particular :) Jul 15 20:45:23 I don't actually have a problem with emacs start time on my other systems Jul 15 20:45:27 emacs --daemon :) Jul 15 20:46:59 johnsu01: if one thinks of emacs as of main application to run on FR then startup time is neglegible too. You just start it on boot and it always runs. You access shell from it, you split windows, you switch buffers... One doesn't even need a WM. You write mail in Gnus, chat in irc.el and jabber.el... If only FR had a keyboard! :( Jul 15 20:47:31 mwester, got any clue how to solve my issue? Jul 15 20:47:51 PaulFertser: yeah, I was using erc+bitlbee on it for IM already Jul 15 20:47:56 PaulFertser: and gnus Jul 15 20:48:11 and w3m-el Jul 15 20:50:13 johnsu01: are there downsides? Jul 15 20:50:40 PaulFertser: it can be a downside sometimes to be trying to do everything in one window like that Jul 15 20:50:46 on a small device Jul 15 20:51:01 that could be partly solved by having multiple frames Jul 15 20:51:17 lack of a keyboard is more of an issue in some modes than others Jul 15 20:56:04 johnsu01: i wish there was some modern emacs reimplementation... I know of one effort to do that using Scheme but i'm not sure it's anywhere near comparable. Jul 15 20:56:49 PaulFertser: True, Emacs is getting a lot of modern things lately though Jul 15 20:56:57 so I have hopes for the future Jul 15 20:57:04 new maintainership and all Jul 15 20:57:24 Emacs is good operating system, but it lacks good text editor :D Jul 15 20:57:49 johnsu01: haha, i remember how RMS said he saw Eclipse and that emacs should borrow some stuff from it. Shortly after he transfered maintainership :) Jul 15 20:58:19 yep :) Jul 15 20:59:39 pure that others ide's don't takes main emacs features Jul 15 21:00:15 dos1: you just never tried it. Jul 15 21:00:17 like separating file/data and buffers Jul 15 21:00:48 I still don't know any editor except emacs, where I can easy change file in to different parts in one time. Jul 15 21:00:59 PaulFertser: i was just repeating popular joke ;) Jul 15 21:01:08 dos1: i know :D Jul 15 21:01:09 hashmap.vala:160.50-160.68: error: use of possibly unassigned local variable `next' Jul 15 21:01:09 for (node = (owned) _nodes[i]; node != null; node = (owned) next) { Jul 15 21:01:09 ^^^^^^^^^^^^^^^^^^^ Jul 15 21:01:21 libgee fails to build... who broke it??!!! Jul 15 21:03:22 damn OE :( Jul 15 21:03:35 mwester, do you think if i'll give it time Jul 15 21:03:38 it'll just work? Jul 15 21:04:38 mwester, are you building unstable or testing? Jul 15 21:05:08 hmm... probably ptitjes broke that :( Jul 15 21:05:12 ptitjes, ping ^^^ Jul 15 21:06:32 mrmoku, got any ideas on how to solve my OE issue? Jul 15 21:06:54 not really... did you already interrupt and restart it? Jul 15 21:07:01 a couple of times Jul 15 21:07:16 and it always hangs while cloning frameworkd? Jul 15 21:07:43 yes. Jul 15 21:08:05 :| Jul 15 21:09:00 TAsn, can you manually git clone it? Jul 15 21:11:24 anyway... bedtime Jul 15 21:11:52 the lite image is almost finished... the rest shall build without me watching it ;) Jul 15 21:11:59 gnight all Jul 15 21:12:03 TAsn, and good luck :P Jul 15 21:13:02 night ;) Jul 15 21:15:11 TAsn, look in the git subdir in your downloads dir for anything to do with framework, and delete it. This will force it to redownload all the framework stuff... that's the only thing I know to do. Jul 15 21:15:16 mrmoku|away, unstable. Jul 15 21:15:22 and it is, in fact, unstable! :D Jul 15 21:17:20 * dos1 is compiling libframeworkd-phonegui-efl with (hopefully) working volume slider Jul 15 21:17:34 it's ugly now, but i'll try to make it better ;) Jul 15 21:17:54 (code) Jul 15 21:22:14 will the phonegui have volume for both spkr and mic? Jul 15 21:26:19 Blu3: now only for speaker Jul 15 21:26:27 but it can be easly added Jul 15 21:26:46 aye Jul 15 21:29:01 mrmoku|away, !!! I hope you added the sleep3 workaronud :) Jul 15 21:32:19 <__Jan__> hi there... Jul 15 21:32:27 <__Jan__> i have a little problem setting up the SHR build environment. Jul 15 21:32:34 <__Jan__> downloaded the makefile. make setup works fine, but then: http://pastebin.org/2329 Jul 15 21:32:43 <__Jan__> http://pastebin.org/2327 is the mentioned logfile - anyone here which saw this before? Jul 15 21:48:22 <__Jan__> anyone? :) Jul 15 22:12:24 PaulFertser: "" PaulFertser wonders why he's so poorly understood by F4t "" you're not alone. That's why I handed over ;-) Jul 15 22:13:15 TAsn: sleep 3 workaround? Jul 15 22:13:54 aye. Jul 15 22:14:08 I are speak bad lanugage yes? Jul 15 22:14:10 bug 551 iirc Jul 15 22:14:24 PaulFertser, you should attempt speaking in russian :) Jul 15 22:14:29 no understanding you correctnessness Jul 15 22:14:33 :P Jul 15 22:15:27 i remember when i was young, and some man in supermarked asked me something in english... when he saw that i don't understand, he asked "do you speak english?" Jul 15 22:15:32 i answered "little" ;D Jul 15 22:16:53 i still speak english "little" regarding tenses and vocabulary ;) Jul 15 22:17:10 dont we all... Jul 15 22:18:59 * dos1 fixed all compilation errors :D (ah, moving from python to C is painful :P) Jul 15 22:19:17 dos1, why would anyone do that? Jul 15 22:19:30 F4t: what? Jul 15 22:19:52 moving from python to c Jul 15 22:20:27 F4t: i was working mostly in python for year, but now i'm trying to implement volume sliders in phonegui-efl, which is written in C Jul 15 22:20:32 C should be smaller and faster Jul 15 22:21:01 Blu3, faster executed, not faster to write in... Jul 15 22:21:09 :P Jul 15 22:21:28 * dos1 looks forward into learning vala Jul 15 22:21:45 but that after i'll finish working on opimd :) Jul 15 22:22:18 is vala a python-like c? Jul 15 22:22:38 F4t, no Jul 15 22:22:43 it's a c# like c Jul 15 22:22:51 oh the horror.... Jul 15 22:22:58 :) Jul 15 22:23:08 it just gives you compiled c# Jul 15 22:23:12 give or take Jul 15 22:23:20 which I guess is nice for some people Jul 15 22:23:23 I bet I'll hate it ;) Jul 15 22:23:49 <__Jan__> you don't _want_ c# ;) Jul 15 22:23:50 not from what i hear, i heard vala gives you the power of scripted languages with c.... any 3rd opinion? Jul 15 22:24:24 F4t, vala docs? :) Jul 15 22:24:33 F4t: vala just brings you high-level language with low-level speed and memory footprint Jul 15 22:25:03 dos1, high level.. as in python? or c#? thats a very crucial point... Jul 15 22:25:16 F4t: dunno, ask mickey|dinner or ptitjes Jul 15 22:25:24 or just rtfm ;) Jul 15 22:25:31 wiki/vala or something :) Jul 15 22:25:45 http://en.wikipedia.org/wiki/Vala_%28programming_language%29 Jul 15 22:25:55 vala is compiled to c???? :P Jul 15 22:26:07 F4t: yep Jul 15 22:26:52 thats.... interesting... Jul 15 22:27:43 yeah. ;) Jul 15 22:28:02 vala just converts vala code to C and then calls gcc iirc ;) Jul 15 22:29:57 anyhow everyone Jul 15 22:30:02 night. Jul 15 22:30:27 let's just hope I'll ace the test tomorrow :) anyhow starting from tomorrow night I'm free to dev awesome stuff ;) Jul 15 22:40:06 is there a vala book? or a bunch of documentation? Jul 15 23:04:32 TAsn: even more awesome yet? I'm not sure we will survive that ;-D Jul 15 23:05:01 Anyway, good luck for your test! Jul 15 23:05:01 DocScrutinizer, is the build over? Jul 15 23:05:20 F4t: asking ME ? Jul 15 23:05:37 i dunno, u seem like someone who would know how and where to check... Jul 15 23:05:46 ANNNND, ur online now... :) Jul 15 23:07:21 F4t: THREE misconceptions ;-) Jul 15 23:08:12 so many... :-/ Jul 15 23:09:49 F4t: all I can say is: that's among the youngest files on http://build.shr-project.org/shr-unstable/images/om-gta02 -> shr-shr-image-glibc-ipk--20090710-om-gta02.rootfs.jffs2 Jul 15 23:10:08 :( Jul 15 23:10:20 even i can say more, since mokonnect 0.3 didnt appear in the repo.... Jul 15 23:11:15 see, you discovered by yourself 2 of the three miscons. Jul 15 23:11:37 * DocScrutinizer makes clear #3 also isn't correct :-) Jul 15 23:11:55 that you are online? Jul 15 23:16:54 volume slider works :) Jul 15 23:19:12 dos1, maybe you know whats the build progress? Jul 15 23:19:50 F4t: dunno, maybe check on tinderbox Jul 15 23:20:06 wheres that? Jul 15 23:21:10 tinderbox.openembedded.org? Jul 15 23:22:30 + 2009-07-15 21:17:17 shr.bearstech.com shr om-gta02 Failed Jul 15 23:25:17 F4t: so you know the answer ;) Jul 15 23:25:50 i do :( Jul 15 23:26:38 freesmartphone.org: 03mickey 07cornucopia * r84b31c35d6ae 10/libfsoframework/ (3 files in 2 dirs): fsoframework: enable GNU extensions Jul 15 23:27:49 freesmartphone.org: 03mickey 07fso-abyss * rff68be158106 10/configure.ac: enable GNU extensions Jul 15 23:31:40 freesmartphone.org: 03mickey 07cornucopia * r7eff20eef31e 10/fsodeviced/ (configure.ac src/bin/Makefile.am src/lib/Makefile.am): fsodeviced: enable GNU extensions Jul 15 23:31:41 freesmartphone.org: 03mickey 07cornucopia * rd28bdb55f336 10/fsonetworkd/configure.ac: fsonetwork: enable GNU extensions Jul 15 23:31:42 freesmartphone.org: 03mickey 07cornucopia * r88771df1c70a 10/fsotimed/ (configure.ac src/bin/main.vala): fsotimed: enable GNU extensions Jul 15 23:31:44 freesmartphone.org: 03mickey 07cornucopia * r6546df25d7b7 10/fsogsmd/configure.ac: fsogsmd: enable GNU extensions Jul 15 23:31:45 freesmartphone.org: 03mickey 07cornucopia * r971647ae24ce 10/fsousaged/ (configure.ac src/bin/Makefile.am src/lib/Makefile.am): fsousaged: enable GNU extensions; fix EXTRA_DIST declaration Jul 16 00:10:22 DieMumie1: i merged drm-kms into drm-tracking, in case you're interested :) Jul 16 00:44:24 * dos1 was using "self" in C code Jul 16 00:44:29 isn't it sign that it's late? :D Jul 16 01:04:34 <__Jan__> it's a sign that c laks python syntax ;) Jul 16 01:16:13 * DocScrutinizer wonders how many words there are for "self"/"this"/"me"/... Jul 16 01:16:52 DocScrutinizer: now, it was data Jul 16 01:17:10 my brain just changed it to self :P Jul 16 01:17:56 duh, so "data" isn't exactly a generic self-reference? Jul 16 01:19:56 * DocScrutinizer knows "data" as a keyword only as complementary to "code" Jul 16 01:20:45 don't ask me the name of that weird language Jul 16 01:21:06 probably some assembler Jul 16 01:21:13 freesmartphone.org: 03Frederik.Sdun 07dbus-hlid * reffb134c3f9d 10fso-monitord/src/ (framework.vala monitor.vala): Get pid of the daemons on startup Jul 16 01:21:14 freesmartphone.org: 03Frederik.Sdun 07dbus-hlid * r8198bcc78572 10fso-monitord/src/framework.vala: Spawn daemons async Jul 16 01:21:26 DocScrutinizer: it wasn't for self reference, just struct containing all data, which will be "transfered" over functions Jul 16 01:21:54 hmm, a simple name then Jul 16 01:22:28 * dos1 installs his package with slider based on alsa-lib Jul 16 01:25:08 ophonekitd: symbol lookup error: /usr/lib/libframeworkd-phonegui-efl.so.0: undefined symbol: snd_hctl_open Jul 16 01:29:33 compiling once again... Jul 16 01:36:21 fuck Jul 16 01:36:35 code seems to be done, just linking :/ Jul 16 01:36:45 ok, i'll ask about that tomorow Jul 16 01:36:47 now good night Jul 16 01:37:01 good night **** ENDING LOGGING AT Thu Jul 16 02:59:57 2009