**** BEGIN LOGGING AT Thu Apr 15 02:59:56 2010 Apr 15 04:25:44 ~ping Apr 15 04:26:21 ~chaninfo Apr 15 04:26:22 I'm on 3 channels: #openmoko-cdevel/91, ##guleague/11, #gsoc/1 Apr 15 04:26:22 i've cached 103 users, 101 unique users, distributed over 3 channels. Apr 15 04:27:03 ~join #openmoko Apr 15 05:48:53 Is it just me or Cornucopia documentation really lacking "distro integrator" info? I can't quite figure out what parts are supposed to be started by dbus activation and what aren't. mrmoku|away any hint? Apr 15 06:13:29 PaulFertser: iirc only fsodeviced is supposed to be started asap (through init.d) and the rest is ok with dbus activation (but for SHR we disabled fsogsmd dbus activation, because we have it installed but still using ogsmd from frameworkd by default) Apr 15 06:13:45 JaMa: ok, thanks for the info Apr 15 06:23:40 JaMa: do you by any chance want to fix this little opimd-cli utf-8 handling glitch or should i rather wait for dos? http://paste.debian.net/68960/ (dirty hack from a python noob) Apr 15 06:24:50 I would keep it for dos.. as I don't even have R/W to opimd repo :) Apr 15 06:48:46 spaetz: ping Apr 15 07:28:40 pong Apr 15 07:28:47 graziano: I'm here now and had coffee Apr 15 07:28:56 so I am prepared for horror messages :) Apr 15 07:28:57 that's good! :) Apr 15 07:29:33 spaetz: yesterday I flashed shr-t from April 14, that bug about accidentally launching icons when scrolling apps in home screen is gone Apr 15 07:29:34 but Apr 15 07:29:51 I tried to install pingus, and gnuchess+xboard Apr 15 07:29:57 installation went fine but Apr 15 07:30:51 a) to see the pingus icon I had to reboot, and pingus keeps exiting when I start it, I think this should be fixed in shr-u but I don't know Apr 15 07:31:33 ok, first of all I am glad that it works in general, I was afraid everything is broken :-) Apr 15 07:31:44 http://git.openembedded.org/cgit.cgi/openembedded/commit/?h=shr/unstable&id=907c4dbbccba448c930f885c3b20c990c1be2351 Apr 15 07:31:46 So if "just" pingus is broken, I am happy :) Apr 15 07:31:52 I will look at pingus specifically Apr 15 07:32:02 ahh, thanks JaMa Apr 15 07:32:06 will pick that. Apr 15 07:32:10 yes that is what I was talking about Apr 15 07:32:14 then Apr 15 07:32:20 http://git.openembedded.org/cgit.cgi/openembedded/commit/?h=shr/unstable&id=780b100bd0c6ae69fa4f9f1fff0768a6b9f494aa Apr 15 07:32:35 yes Apr 15 07:32:36 :) Apr 15 07:32:44 JaMa: many thanks :) Apr 15 07:34:30 JaMa: indeed, thanks. Apr 15 07:34:31 Hm, i hoped zhone2 uses opimd :) Apr 15 07:34:39 graziano: fixes are in the tree, and updating Apr 15 07:34:48 Should be available soon Apr 15 07:35:04 once the buildhost has build them. Apr 15 07:35:20 thanks, let me see if I had something else to whine about :D Apr 15 07:35:46 * graziano checks its stupid page: http://wiki.openmoko.org/wiki/User:Graziano Apr 15 07:36:13 PaulFertser: is zhone2 possible o build now? Apr 15 07:37:57 graziano: it's not stupid. Apr 15 07:38:07 YOu are our QA department! Apr 15 07:38:11 spaetz: any possibility for shr-t to ship with debugging options disabled in the kernel? Apr 15 07:38:13 * graziano hides Apr 15 07:38:19 0_o Apr 15 07:38:44 graziano: I had them disabled and we got all kind of feedback with weird kernel errors that ONLY occur when debugging is turned off Apr 15 07:38:55 graziano: we tried that once and then nothing worked Apr 15 07:38:58 YOu are our QA department! <--- too much pressure!!! :D Apr 15 07:39:08 spaetz: Heinervdm k,k :) Apr 15 07:39:09 graziano: we will switch to non-debug with the .32 kernel Apr 15 07:39:11 and that is something I cannot have. Apr 15 07:39:55 graziano: seems like lots of apps listed in http://wiki.openmoko.org/wiki/User:Graziano as from opkg.org are also in shr-u feeds.. whould be nice to sort them right Apr 15 07:40:14 graziano: well, I never wanted to maintain shr-t, you never wanted to be our QA department. :) Apr 15 07:40:29 vanous123: of course, i built it on debian. Apr 15 07:40:35 vanous123: needs vala and efl-vala bindings Apr 15 07:40:51 Unfortunately, i can't answer a single call with it :( Apr 15 07:41:12 i see Apr 15 07:41:17 just curious Apr 15 07:41:23 And shr apps in debian do not allow me that either (at least with fsogsmd) Apr 15 07:41:41 i see Apr 15 07:41:44 spaetz: oh no I am honoured to be, I like to be of help when I can, so I really like to test your amazing work Apr 15 07:41:58 PaulFertser: i think you need another version for fsogsmd, but i'm not sure Apr 15 07:41:59 PaulFertser: one would need those tweaked phoneui apps... Apr 15 07:42:10 PaulFertser: have you tred them? Apr 15 07:42:20 s/tred/tried/ Apr 15 07:42:20 vanous123 meant: PaulFertser: have you tried them? Apr 15 07:42:46 and I would like to thank all the developers here that are so kind to me, I know it is hard when people come and bother you about non-working things without providing solutions :) Apr 15 07:43:27 ok some other thoughts Apr 15 07:43:46 PaulFertser: have you used fso-glib branch for building phonefsod? Apr 15 07:44:05 PaulFertser: mrmoku's porting to fsogsmd is mostly there Apr 15 07:44:16 trying to connect to a wifi network is a really PITA atm using mokonnect and I know you already know this, possible solutions? I remember someone suggesting that NWA thing Apr 15 07:44:32 http://wiki.openmoko.org/wiki/NWA Apr 15 07:45:55 graziano: i think wifi has lower priority for all developers, because of that no one of us will fix it now Apr 15 07:46:51 :(((((((( Apr 15 07:47:30 phone is top priority Apr 15 07:47:37 sure I imagined that Apr 15 07:47:49 that has to be stable, after that one can look for the other things Apr 15 07:48:04 or someone in the community will solve it before Apr 15 07:53:33 graziano: what's the problem with wifi for you? it does not work at all or it's not user-friendly enough? Apr 15 07:54:46 JaMa: both, mokonnect is too complex, too many fields etc. AND it often fails to connect and I have to retry again and again (I saw other people complaining with this on the list) Apr 15 07:55:03 JaMa: have you ever seen how android manage wifi? Apr 15 07:56:22 no.. and I usually use vi to edit wpa_supplicant.conf and then CLI commands.. so I'm not the right one for user-friendliness requests :) Apr 15 07:57:03 graziano: I agree, mokonnect is too complex. I would love to have a simpler add that did ust wifi, but that good Apr 15 07:57:07 JaMa: about this bug: http://trac.shr-project.org/trac/ticket/789 Apr 15 07:57:08 just wifi... Apr 15 07:57:35 JaMa: in testing e-wm depends on virtual-locale, but that should be build after that eglibc bump Apr 15 07:58:22 Heinervdm: no, wifi is not lower priority for me, but a) it works for my hotspots and b) I tried to code a new wifi manager in vala and got stuck due to vala issues (incompencies) Apr 15 07:58:42 and eglibc in testing depends on it too Apr 15 07:58:46 I might break down and just go for a python replacement if i have some time (which I have not) Apr 15 07:59:16 Heinervdm: virtual-locale is separate recipe? or just PACKAGE in eglibc? Apr 15 07:59:49 JaMa: eglibc has it in Depends: in the packge (not recipe) Apr 15 08:00:13 and only in shr-t? Apr 15 08:00:17 JaMa: e.g. virtual-locale-zh-cn Apr 15 08:00:21 graziano: pingus and gnuchess fixes are not in the feed Apr 15 08:00:36 now in the feed Apr 15 08:00:44 mrmoku|away: JaMa http://trac.freesmartphone.org/ticket/542 shouldn't be ophoned left turned on in SHR? It seems it works for me here and it's essential for those who wants bluetooth. Apr 15 08:00:47 spaetz: thank you Apr 15 08:01:07 spaetz: may I ask a few other things, maybe you know the answer or someone else here do Apr 15 08:01:27 JaMa: hmm, no in unstable too Apr 15 08:02:21 PaulFertser: it would be usefull for sip too Apr 15 08:02:37 sure, asking you can always do. :) Apr 15 08:03:08 Heinervdm: mrmoku|away told me he disabled ophoned because it wasn't working but now it apparently works, so probably it should be reenabled. Apr 15 08:03:32 JaMa: but why does installing locale works in unstable but not in testing? Apr 15 08:03:48 git grep shows that eglibc also provides it.. but strange that maybe it's renamed to runtime dependency locale-base-cs-cz.iso-8859-2+eglibc-binary-localedata-cs-cz.iso88592 only in shr-u Apr 15 08:03:54 PaulFertser: ah, ok then we can try Apr 15 08:03:56 if I understand it right Apr 15 08:05:02 Heinervdm: at least i initiated a call and it was allright (i haven't tried bluetooth, just saw correct messages in frameworkd.log) (after i changed it to match against fsogsmd statuses) Apr 15 08:05:05 JaMa: it seems so, there is no virtual-locale-* package in armv4t dir Apr 15 08:06:57 Hm, shit, no wonder i can't receive calls with zhone2 :( Apr 15 08:07:09 spaetz: keep me updated if you code something wifi related, I want to test :), now the other questions Apr 15 08:07:12 rotation Apr 15 08:07:31 Heinervdm: "GLIBC_GENERATE_LOCALES = "en_US.UTF-8"" is only in shr-t local.conf Apr 15 08:07:39 will it ever be integrated in the system without need to install third party sw? Apr 15 08:07:59 JaMa: oh, that's bad Apr 15 08:08:11 Heinervdm: that's probably why it won't generate other locales at all Apr 15 08:08:27 No clue, I have tried omnewrotate and found that I dislike the frequent automatic rotations. Apr 15 08:09:05 spaetz: do you want me to rebuild eglibc for shr-t? to fix locales? Apr 15 08:09:16 spaetz: I heard something related to rotation and illume2 but I may be wrong.. Apr 15 08:09:25 JaMa: don't we need to rebuild everything? Apr 15 08:09:35 JaMa: sure. Just let me rebuild gnuchess before that. Apr 15 08:09:50 Heinervdm: imho no.. this varible is hopefully used only for glibc locales Apr 15 08:09:54 Although I have heard no complaints about broken locales in shr-t yet Apr 15 08:10:00 Heinervdm: and ie those mc-locale-cs were generated Apr 15 08:10:08 spaetz: http://trac.shr-project.org/trac/ticket/789 Apr 15 08:10:17 ahh :) Apr 15 08:15:04 spaetz: started gnuchess Apr 15 08:15:23 thanks Apr 15 08:16:10 strange how slow buildhost is nowadays.. no cpu/mem usage but all IO is slow as hell Apr 15 08:17:52 GarthPS: seen your thread.. Apr 15 08:18:16 JaMa: yes :-( Apr 15 08:18:30 GarthPS: but to add something as complex as phonon is not that easy/quick as bump SRCREV for ventura Apr 15 08:18:31 JaMa: Hi,about PHonon backend or about the beers? Apr 15 08:18:45 both Apr 15 08:18:46 gnuchess finished. Apr 15 08:19:06 phonon? That is KDE tech, right? Apr 15 08:19:09 JaMa: oh yeah! I don't doubt it! :p I was juste wondering if this was interesting some one Apr 15 08:19:29 And it is an abstraction layer if I remember correctly Apr 15 08:19:36 spaetz: qt-phonon is here too Apr 15 08:19:55 spaetz: it is a part of Qt Apr 15 08:20:00 spaetz: kde-phonon is only some extension to it or something like that Apr 15 08:20:51 So yes it is an abstraction layer and I use it in my Qt app Apr 15 08:21:21 I see Apr 15 08:21:58 JaMa: so do you think it could be in feed a day? by the way , if it is so complex perhaps jump a step and package VLC backend directly :) Apr 15 08:22:38 JaMa: does this will be in openembedded or in shr feed? (if it will be a day.) Apr 15 08:23:31 oe.dev branch == SHR Apr 15 08:23:42 so it will be available for whoever builds it Apr 15 08:23:56 the recipe is "share" among all distros Apr 15 08:24:01 +d Apr 15 08:25:29 JaMa: coool thanks! even if i send you a a bitbake recipe for my app it will be availiable for oe? Apr 15 08:25:45 JaMa: "+d" ?? Apr 15 08:25:46 sure ALL recipes are shared :) Apr 15 08:25:57 "share" +d :) Apr 15 08:28:16 JaMa: SOoo good!! "+d" hum sorry but, en bon francais, I don't understand the +d what does it stand for? Apr 15 08:30:15 GarthPS: I just forget to type 'd' in "share" so maybe I should add s/"share"/"shared"/g instead shorter "+d" :) Apr 15 08:30:47 JaMa: LOL! ok my bad sorry :p Apr 15 08:31:17 no problem :) Apr 15 08:33:17 JaMa: hum i don't wnat to bother you but I thank you didn't answer on of my question.. :p "do you think it could be in feed a day?" Apr 15 08:33:18 this time it was shorter surely :)) Apr 15 08:34:10 vanous123: hehe :) Apr 15 08:34:21 GarthPS: yes it could be in feed Apr 15 08:34:57 JaMa: ;) and about vlc backend directly? perhaps it is more easy .. Apr 15 08:35:04 GarthPS: but the recipe needs few updates (ie no DEPENDS is bad) and I don't expect it to build so easily Apr 15 08:35:46 GarthPS: and there can be some hidden problem in source (like arm optimizations for newer arm arch or whatever) preventing us to build it without modifications... Apr 15 08:36:27 JaMa: ho the recipe I shown? I am realy not sure it is a good one! perhaps a start but not a usefull for now Apr 15 08:36:35 JaMa: yeah.. Apr 15 08:36:53 yeah.. but it's a template at least :) Apr 15 08:41:48 GarthPS: are you aware of this? http://paste.pocoo.org/show/202071/ Apr 15 08:42:50 JaMa: you mean th gstreamer backend ? Apr 15 08:42:56 GarthPS: looks like you only need that vlc backend, but I'm not sure how advanced is vlc support in OE (1.0.2 recipe is here) Apr 15 08:44:28 GarthPS: what you sent looks like recipe for whole phonon Apr 15 08:45:10 GarthPS: http://gitorious.org/phonon/phonon-vlc and this looks like just the vlc backend as you said before.. so I'm just pointing that first part is already in OE Apr 15 08:45:18 JaMa: yeah.. I touhg that.. Apr 15 08:45:53 JaMa: the first part , you mean libphonon Apr 15 08:46:22 yes, built with qt4 Apr 15 08:46:40 but I belive that them vlc backend need libvlc 1.1 wich is still in a beta stage Apr 15 08:46:51 but perhaps.. Apr 15 08:47:15 so I will update my recite to match the vlc-backend Apr 15 08:48:37 no need to resend that... Apr 15 08:50:29 JaMa: anyway, thank you for pointing me this out. I will look at vlc backend instead of xine . but for xine it was the same thing i.e only missing the backend part Apr 15 08:51:31 JaMa: you mean a right start for a recipe of vlc backend? :) Apr 15 08:52:30 yes.. I've already changed that... Apr 15 09:00:03 JaMa: you definitively deserve a bunch of beers! (or some thing else if you want :p) Apr 15 09:00:52 PaulFertser: ping Apr 15 09:13:11 mickey|office: pong Apr 15 09:13:13 :)) Apr 15 09:14:56 PaulFertser: good morning. could you add more details to your last bug report, please? :) Apr 15 09:15:07 Heh, that STATUS thing is the single reason that brought most breakage to my everyday phone :) Now after binary-patching shr libs i can use phoneuid again. Apr 15 09:15:17 mickey|office: not able to answer with zhone2? Apr 15 09:15:26 yeah Apr 15 09:15:39 mickey|office: someone calls me, i want to answer :) Apr 15 09:15:46 this functionality is implemented, so i wonder why it doesn't work Apr 15 09:16:05 ooh Apr 15 09:16:09 the scenario is not changed Apr 15 09:16:11 could it be that? Apr 15 09:16:19 mickey|office: hm, how so? I've looked at the code and it was definetely not there, i attached a patch even to the ticket, have you seen it? Apr 15 09:16:29 let me check Apr 15 09:17:08 oops Apr 15 09:17:11 of course Apr 15 09:17:19 thanks, will commit asap Apr 15 09:17:20 :) Apr 15 09:17:25 :D Apr 15 09:18:33 mickey|office: what's your immediate plan about ophoned btw? mrmoku|away told me he disabled it in SHR, but it seems it should be enabled now as it looks like it's working reasonably and is needed for BT, right? Apr 15 09:27:33 Felix Huber did some work on it and he's using it actively, so I would not mind it being reenabled. For FSO2, I need to take a thorough look into how the communication with bluetooth actually works before deciding whether it fits naturally into fsogsmd or whether we need an equivalent of ophoned again. Apr 15 09:34:03 PaulFertser: I'm starting three things via init scripts: fsodeviced, fsogsmd and fsotdld Apr 15 09:34:42 fsogsmd might be started via dbus-activation too... would have to add some dummy call to phonefsod to do that I think Apr 15 09:36:04 with my shadow resources this dummy call will be obsolete Apr 15 09:36:12 but they're not ready for consumption Apr 15 09:36:40 mrmoku: here everything but fsodeviced starts by activation Apr 15 09:37:34 mickey|office: ok, will just keep the initscript until then Apr 15 09:37:54 PaulFertser: what activates fsogsmd? Apr 15 09:38:03 hmm Apr 15 09:38:17 mickey|office: would trying to get a proxy to register a signal handler activate it? Apr 15 09:38:38 mickey|office: when you'll be taking a look at bluetooth, do not forget to use some "new" method with unix sockets or something like that instead of Play() function to activate SCO over PCM. Apr 15 09:38:43 mrmoku: no idea... Apr 15 09:39:12 mrmoku: only method calls activate. you can always use org.freedesktop.Peer.Ping() as activation method Apr 15 09:39:19 every dbus object implements this method Apr 15 09:39:35 PaulFertser: what do you use as phonestack? phonefsod/phoneuid? Apr 15 09:40:34 mrmoku: yes, some old version in debian 20100213, it activates gsm but doesn't actually register to network, zhone2 does. Apr 15 09:41:19 ok, adding that Ping is easy enought that I might just want to do that instead of the initscript Apr 15 09:41:23 -t Apr 15 09:42:05 And yes, fsotdld is not starting here automatically, should be explicitly run. Apr 15 09:43:06 PaulFertser: the old version does not use SetFunctionality... which is the reason why it activates it, but does not register Apr 15 09:43:14 Of course Apr 15 09:43:32 mrmoku: it doesn't properly handle call statuses anyway, so np :) Apr 15 09:44:10 mrmoku: new phoneuid interface (comparing to ophonekitd) is quite nice btw, but still sluggish ;) Apr 15 09:44:29 PaulFertser: one thing about the call statuses that changed is that the status is upercase now... if you use oeventsd you have to fix rules.yaml accordingly Apr 15 09:44:46 it has to look for INCOMING instead of incoming Apr 15 09:44:54 same for outgoing and active Apr 15 09:45:03 mrmoku: LOL, i've changed everything already, including ophoned and binary patched shr libs ;) Apr 15 09:45:18 :P Apr 15 09:45:38 And filed meaningful tickets ;) Apr 15 09:45:45 good thing :) Apr 15 09:46:26 mmm what about rotation and illume2? Apr 15 09:46:42 I remember about something related but I'm not sure Apr 15 09:48:20 mickey|office: sections for subsystems not existing in frameworkd.conf implicitely disable it, right? Apr 15 09:52:41 mrmoku: must confess i no longer know... let me look at the code Apr 15 09:54:03 mrmoku: no, frameworkd scans through the directories, need the 'disable' Apr 15 09:54:08 the config-based approach only works in fso2 Apr 15 09:54:11 mrmoku: can you look at this an tell me why the genlist is empty? http://shr.pastebin.com/Gb53wqsq Apr 15 09:54:12 ok, thanks Apr 15 09:58:37 Heinervdm: do you have the full thing? I can't see the feeding part in the patch Apr 15 10:08:05 mrmoku: http://heinervdm.dyndns.info/shr-git/libphone-ui-shr/src/view/sim-manager-view.c Apr 15 10:15:36 Heinervdm: can't see it on a fast look... have to go to the kindergarten now and eat... will take a close look afterwards Apr 15 10:15:55 mrmoku: thx Apr 15 10:51:07 *facepalm* Apr 15 10:51:13 android mailclient Apr 15 10:51:41 start writing an email, save it as template, reopen it later, have all & replaced by & Apr 15 10:52:10 you know, sometimes I wonder how google ever managed to create a decent search engine... Apr 15 10:53:02 by allocating the dummy coders to side project such as android :) ? Apr 15 10:53:31 probably Apr 15 10:53:40 come on, even _I_ could have made that better Apr 15 10:53:53 doesn't anyone ever TEST the applications? Apr 15 11:04:08 maybe I should apply a QA tester at google for android applications, though. I've found so much crap in 2 weeks that it'll keep them busy for a wwhile Apr 15 11:06:07 blindcoder: just make apps Apr 15 11:08:01 blindcoder: they just don't use & in e-mails. 8) Apr 15 11:08:08 lol Apr 15 11:08:18 nobody is testing exactly that for sure :) Apr 15 11:08:46 and if there is no automated test for that.. nobody will notice it while testing it by-hand Apr 15 11:13:00 will shr buildhost finish eglibc rebuild in less then 3hours? I guess no.. now it's >2:50 :/ Apr 15 11:21:28 Heinervdm: I think I know Apr 15 11:21:57 yes? Apr 15 11:22:01 Heinervdm: in contact_select_view you have struct _contact_select_pack as local variable Apr 15 11:22:11 which will vanish when that function finished Apr 15 11:22:26 use a pointer with malloc instead Apr 15 11:22:38 mrmoku: so i have to make that static? Apr 15 11:22:45 well no Apr 15 11:22:45 ah Apr 15 11:22:47 ok Apr 15 11:22:49 just make it a pointe3r Apr 15 11:23:10 Heinervdm: and don't forget to free it when not needed anymore :P Apr 15 11:23:28 btw. why do you do a new window/view ? Apr 15 11:23:32 mrmoku: i will forget it :P Apr 15 11:23:35 instead of using a pager Apr 15 11:24:06 i wanted to use an inwin, but i can't get inwin working with a layout Apr 15 11:24:21 Heinervdm: for that I think I have a fix :) Apr 15 11:24:30 there is one thing missing in ui_utils_view_init Apr 15 11:24:36 which I fixed in my libfso-glib branch Apr 15 11:24:45 the layout itself does not get resized properly Apr 15 11:24:58 thus leading to various kind of strange sizing errors Apr 15 11:25:48 i haven't got a inwin at all Apr 15 11:26:02 nothing happes after starting that funciton... Apr 15 11:26:05 forgot elm_win_inwin_activate? Apr 15 11:26:13 no that was there Apr 15 11:26:43 now i have to use: VIEW_PTR(*pack) Apr 15 11:26:48 yep Apr 15 11:33:35 hey do you guys know if the 2.6.32.11 in http://build.shr-project.org/tests/mrmoku/2.6.32/images/om-gta02/ is supposed to be working properly? The xserver in shr-t doesn't work and i've applied the script in the readme. dmesg and xorg.0.log don't show anything interesting Apr 15 11:33:51 eglibc rebuild finished! :) 3:30 Apr 15 11:34:09 zeroedout: you have to update event device numbers in xorg.conf Apr 15 11:34:50 zeroedout: http://paste.pocoo.org/show/202109/ Apr 15 11:35:40 mrmoku: do you have it the same now ^^ ? I would update README there if so Apr 15 11:35:52 sweet, thanks Apr 15 11:36:41 JaMa: yep Apr 15 11:38:56 added there to README Apr 15 11:45:01 mickey|office: SIM.ReadyStatus signal is gone forever? Apr 15 11:57:55 * graziano reads -devel email where a great guy develop a gui tool to connect to wifi Apr 15 11:57:56 \o/ Apr 15 11:58:37 mrmoku: and all, if it's fine could it be integrated in shr? Apr 15 11:59:21 * mrmoku searches that mail Apr 15 11:59:52 graziano: which -devel email? shr-devel? Apr 15 11:59:54 graziano: when it's working we can put it in the feeds Apr 15 12:00:13 mrmoku: yes Apr 15 12:00:23 http://github.com/Ebbe/Iliwi Apr 15 12:00:24 Esben Damgaard here? Apr 15 12:00:38 * JaMa already building that ;) Apr 15 12:01:07 :) Apr 15 12:01:09 JaMa: \o/ Apr 15 12:01:27 * graziano updates his page with today findings Apr 15 12:02:12 how does it work? it doesn't depend on wpa_supplicant nor on connman, nor on networkmanager Apr 15 12:02:54 graziano: can you try this? http://jama.homelinux.org/org.openembedded.shr/armv4t/iliwi_0.0.1+gitr0+adae747061e8d885bbe9084814dad169fb00fae2-r0.4_armv4t.ipk Apr 15 12:05:56 mrmoku: view/idle-view.c:165: error: too few arguments to function 'edje_edit_part_selected_state_set' Apr 15 12:06:24 Heinervdm: hehe Apr 15 12:06:26 API changed there Apr 15 12:06:37 the "default 1.0" is "default", 1.0 now Apr 15 12:06:42 (or whatever there is) Apr 15 12:06:43 seems so Apr 15 12:06:59 JaMa: installed... trying now :) Apr 15 12:07:17 will you change it, or should i? Apr 15 12:07:45 Heinervdm: let me see if I can just cherry-pick from my branch Apr 15 12:07:54 ok Apr 15 12:09:31 Heinervdm: do you have fbf52ad40ca3078d392884aa90de5ad2c5539d39 ? Apr 15 12:09:57 mrmoku: no Apr 15 12:10:10 ok, will stash my changes, switch branch and cherr-pick it then Apr 15 12:10:55 SHR: 03mok 07libphone-ui-shr * r25a03a38e578 10/src/view/idle-view.c: idle-view.c: adjust to changed semantics of edje_edit_part_selected_state_set Apr 15 12:10:55 done Apr 15 12:11:26 ok Apr 15 12:16:12 JaMa: hmm... somehow it makes illume segfault on exit :/ Apr 15 12:17:35 hmm :/ Apr 15 12:18:42 other than that (and the missing desktop icon) it looks like a nice alternative :) Apr 15 12:19:15 JaMa: I am @work now but I will try it when I come home in a few hours Apr 15 12:21:35 ah mrmoku already did it Apr 15 12:21:55 maybe someone could report this to the mailing list so its developer can fix it Apr 15 12:23:49 *** glibc detected *** /usr/bin/enlightenment: munmap_chunk(): invalid pointer: 0x00211bcc *** Apr 15 12:23:58 * mrmoku tries to add an issue to github Apr 15 12:24:02 * JaMa also at work, and only built that.. Apr 15 12:24:13 eh Apr 15 12:24:17 needs login :/ Apr 15 12:36:51 iliwi uses wpa_supplicant and udhcpc directly as I see it Apr 15 12:36:54 NICE Apr 15 12:36:58 if it works Apr 15 12:37:09 let's get rid of connman Apr 15 12:37:12 if it works :) Apr 15 12:37:21 at least it finds my network even up here Apr 15 12:37:30 can't connect though... might just be too far away Apr 15 12:37:34 wpa_supplicant over dbus, or over commandline? Apr 15 12:37:58 I _think_ over command line Apr 15 12:38:36 commandline, definitely Apr 15 12:38:47 \o/ Apr 15 12:38:47 \o/ Apr 15 12:38:48 \o/ Apr 15 12:38:52 * graziano jumps around Apr 15 12:39:07 * graziano sees a new wifi management app Apr 15 12:39:17 * graziano sees developers are almost happy with it Apr 15 12:39:20 * mrmoku is going downstairs to see if it connects Apr 15 12:39:33 lots of hardcoded assumptions. Apr 15 12:39:34 http://github.com/Ebbe/Iliwi/blob/master/src/wifi.vala#L224 Apr 15 12:40:05 wpa_supplicant provides a DBUS api, right? Apr 15 12:40:31 vala is good at making dbus calls easily, so that might be converted. Rather than hardcoding interface names and binary paths. Apr 15 12:41:07 wpa_supplicant dbus api can't make much Apr 15 12:42:02 but the way it's done there it's the same as mofi Apr 15 12:42:11 one should at least use wpa-cli Apr 15 12:48:02 so should I add wpa_supplicant&udhcpc to RDEPENDS? Apr 15 12:48:34 JaMa: yes Apr 15 12:49:08 /usr/bin/killall Apr 15 12:49:15 /usr/sbin/wpa_supplicant Apr 15 12:49:28 /sbin/udhcpc Apr 15 12:49:45 /sbin/iwlist (is that wireless-tools?) Apr 15 12:50:09 that seems all Apr 15 12:52:11 udhcpc is provided by busybox the rest is clear Apr 15 12:52:23 thanks spaetz Apr 15 12:56:37 * mrmoku had no luck Apr 15 12:56:47 and then... I never used wifi with the new kernel Apr 15 12:57:04 hmm not sure if everyone wants full udhcpc and killall... but for now I'll put it there Apr 15 12:57:27 wpa_supplicant can at least scan for networks over dbus: http://hostap.epitest.fi/gitweb/gitweb.cgi?p=hostap.git;a=blob;f=wpa_supplicant/examples/wpas-test.py Apr 15 12:58:20 it would be better if busybox is set to RPROVIDE them :/ Apr 15 13:02:02 mrmoku: i finally have contacts in the select screen ;) Apr 15 13:03:31 :) Apr 15 13:03:49 hmm... manually using wpa_supplicant works :/ Apr 15 13:04:25 mrmoku: can you check this with wpa_supplicant? http://hostap.epitest.fi/gitweb/gitweb.cgi?p=hostap.git;a=blob;f=wpa_supplicant/examples/wpas-test.py Apr 15 13:05:14 and this too: http://hostap.epitest.fi/gitweb/gitweb.cgi?p=hostap.git;a=blob;f=wpa_supplicant/examples/wpas-dbus-new.py Apr 15 13:05:36 then we can ask him if he would mind switching to dbus instead of cli calls Apr 15 13:05:59 ohh... now it worked with iliwi too - at least the association Apr 15 13:08:25 you guys built it yourself? Apr 15 13:08:58 vanous123: http://build.shr-project.org/tests/jama/ Apr 15 13:11:21 will test thanx Apr 15 13:11:54 Heinervdm: what to pass it as interface? Apr 15 13:11:57 eth0 does not work Apr 15 13:12:32 mrmoku: it should list the interfaces, if i understood the source right Apr 15 13:12:46 the first one not Apr 15 13:12:53 or does it need wpa_supplicant running?= Apr 15 13:12:56 with dbus interface? Apr 15 13:13:06 yes it needs it running Apr 15 13:13:11 ahh, ok Apr 15 13:13:12 with -u option Apr 15 13:14:18 dbus.exceptions.DBusException: org.freedesktop.DBus.Error.UnknownMethod: Method "scan" with signature "" on interface "fi.epitest.hostap.WPASupplicant.Interface" doesn't exist Apr 15 13:14:36 hmm Apr 15 13:14:36 before it found my network Apr 15 13:14:38 00:08:54:97:6b:40 :: ssid='MRMOKU2' wpa=no wpa2=yes quality=3% rate=0 freq=2432 Apr 15 13:14:51 do we have an old version of wpa_supplicant? Apr 15 13:15:37 so this can at least replace the iwlist call Apr 15 13:15:48 the second one gives me Apr 15 13:15:48 dbus.exceptions.DBusException: org.freedesktop.DBus.Error.ServiceUnknown: The name fi.w1.wpa_supplicant1 was not provided by any .service files Apr 15 13:16:41 oh we have wpa_supplicant 0.6.9 Apr 15 13:16:49 old? Apr 15 13:16:52 newest is 0.7.1 Apr 15 13:17:12 heh Apr 15 13:17:19 current fedora has 0.6.8 Apr 15 13:17:26 so it is not *that* old :P Apr 15 13:18:56 * added new DBus interface (fi.w1.wpa_supplicant1) Apr 15 13:19:03 ^ last item of 0.7.0 changelog Apr 15 13:19:24 and latest stable is supposed to be 0.6.10 Apr 15 13:19:52 so 0.7.1 should be interesting for us? Apr 15 13:20:45 never tried newer than 0.6.9.. but I guess yes.. Apr 15 13:21:00