**** BEGIN LOGGING AT Tue Apr 05 02:59:58 2011 Apr 05 04:39:06 yay, that damn build actually finished successfully :-) Apr 05 04:39:11 and moin Apr 05 04:40:12 * mrmoku off to a client and can't wait to try it when he's back Apr 05 07:36:05 freesmartphone.org: 03morphis 07cornucopia * ra62fbca67dd7 10/fsogsmd/ (3 files in 2 dirs): fsogsmd: connman: use correct version for plugin definition Apr 05 07:36:08 freesmartphone.org: 03morphis 07cornucopia * r0485dc75d30b 10/fsogsmd/conf/palm_pre/fsogsmd.conf: fsogsmd: switch to internal ppp stack for palmpre machine Apr 05 07:36:09 freesmartphone.org: 03morphis 07cornucopia * r5e52276b4dc8 10/fsogsmd/ (src/connman/plugin.vala vapi/connman.vapi): fsogsmd: connman plugin: register network device driver on initialization Apr 05 07:36:11 freesmartphone.org: 03morphis 07cornucopia * rbc835f6d910c 10/fsogsmd/src/connman/ (Makefile.am modemhandler.vala plugin.vala): fsogsmd: connman plugin: add modemhandler which requests GSM resource at startup Apr 05 07:43:13 heyho Apr 05 07:43:25 hey Apr 05 07:44:29 <[Rui]> hey! bye! Apr 05 08:45:31 mickeyl: my image does not work :/ Apr 05 08:46:34 mickeyl: interestingly I get 2011-04-05T00:02:08.356843Z [DEBUG] fsogsmd : GLib : gisicomm.vala:102: NETLINK STATE = PN_LINK_DOWN Apr 05 08:46:40 directly after SendAuth Apr 05 08:46:46 even before trying to register Apr 05 08:51:27 mickeyl: the only difference (apart from fso being autoreved) is that libnl1 is missing in my image... but not the working one Apr 05 09:10:02 mickeyl: ping :) Apr 05 09:10:06 10:45 < mrmoku> mickeyl: my image does not work :/ Apr 05 09:10:06 10:46 < mrmoku> mickeyl: interestingly I get 2011-04-05T00:02:08.356843Z [DEBUG] fsogsmd : GLib : gisicomm.vala:102: NETLINK STATE = PN_LINK_DOWN Apr 05 09:10:09 10:46 < mrmoku> directly after SendAuth Apr 05 09:10:11 10:46 < mrmoku> even before trying to register Apr 05 09:10:14 10:51 < mrmoku> mickeyl: the only difference (apart from fso being autoreved) is that libnl1 is missing in my image... but not the working one Apr 05 09:11:09 hmm Apr 05 09:11:10 that's bad Apr 05 09:11:58 mickey|office: good morning Apr 05 09:12:02 morning morphis Apr 05 09:12:08 mickey|office: PPP in the Palm Pre now works fine Apr 05 09:12:16 even establishing the default route Apr 05 09:12:38 good news Apr 05 09:13:15 jepp Apr 05 09:14:34 what are the major non-telephony issues we have on the pre? Apr 05 09:15:20 mrmoku: ok, i guess next try has to be me creating another image. perhaps we have build host infection somewhere Apr 05 09:15:27 mplayer is freezing the device (or general playback of encoded audio files) Apr 05 09:15:33 * bluetooth is not working Apr 05 09:15:50 * wifi is dieing after some time (we need to limit the bandwidth) Apr 05 09:16:02 or try newer compat-wireless if it works with the palm pre Apr 05 09:16:14 or make it possible to load the webOS wifi driver + firmware Apr 05 09:16:15 mickey|office: yeah, I will take a closer look at libnl1/2 differences meanwhile Apr 05 09:16:55 for bluetooth we need a working bluetooth support in the kernel as it compiles not with the patch from Palm (they don't use kernel for driving bluetooth) Apr 05 09:17:28 other non-telephony issues don't come in my mind actually Apr 05 09:17:52 everything else is beauty correction Apr 05 09:18:00 ok, not bad, not bad Apr 05 09:18:05 :) Apr 05 09:18:36 btw. connman plugin now can request the usage resources Apr 05 09:19:01 heh, good Apr 05 09:21:50 so boys go and buy a Palm pre! Apr 05 09:22:08 mickeyl: ah mickey I forgot GPS Apr 05 09:22:27 but that all are things which should not take months to get them running Apr 05 09:24:13 famous last words ;) Apr 05 09:25:07 :D Apr 05 09:26:28 mrmoku: which OE rev did you build? Apr 05 09:27:40 mickey|office: 8f4d4e7b5a4b9f63958ac7d1037ccc21a99bdf06 Apr 05 09:27:47 a merge of yours :) Apr 05 09:28:10 hmm, looks good Apr 05 09:28:14 i went for skiing on 5th Apr 05 09:28:20 and i recall that things worked before Apr 05 09:28:23 (for me) Apr 05 09:28:33 yeah, you told us to test audio before leaving :) Apr 05 09:29:20 heh, right Apr 05 09:29:30 and the rev you built must have been before cc14abd3baa79dcff056679e2479523f96bd8969 (tzdata pr bump) Apr 05 09:29:38 ok Apr 05 09:29:52 so i will launch an image with that very rev now Apr 05 09:30:04 and after 64503bd9bc15c8dc699f6bd3cb7147c128a86c4b (busybox PR bump) Apr 05 09:30:10 according to installed-packages.txt Apr 05 09:33:22 hm ok, debugging connman -> fsogsmd -> msmcommd -> serial_forward -> modem is not so easy as I thought Apr 05 09:34:00 so I am leaving to catch my train Apr 05 09:34:02 bye Apr 05 09:46:16 ok, build started Apr 05 10:41:10 SHR: 03Martin.Jansa 07shr-chroot * rfa9e330f29e6 10/ (923 files in 103 dirs): system upgrade Apr 05 11:48:38 mickey|office: still around ? Apr 05 11:52:34 yep Apr 05 11:57:29 mickey|office: did you come across this ? Apr 05 11:57:36 NOTE: package fsogsmd-1_0.5.0+gitr2+bc835f6d910c4e8a9774471d73d340b76e7cb447-r5.0: task do_configure: Failed Apr 05 11:57:42 | Requested 'connman >= 0.72' but version of connman is 0.68 Apr 05 11:59:53 either include fso-autorev.inc Apr 05 12:00:15 or bump the fsogsmd version, which now has lowered this requirement Apr 05 12:00:57 grep autorev conf/local.conf Apr 05 12:00:57 require conf/distro/include/fso-autorev.inc Apr 05 12:00:57 require conf/distro/include/shr-autorev.inc Apr 05 12:01:01 mrmoku: if you have fso autorev'ed I reckon you had to bump vala manually to get it to compile, right? Apr 05 12:01:12 then I must bitbake -c rebuild freesmartphone ? Apr 05 12:02:24 freesmartphone.org: 03mickey 07cornucopia * r1bc0a2d5332d 10/fsogsmd/configure.ac: fsogsmd: lower connman requirement to 0.68 Apr 05 12:02:28 dcordes: try again now ;) Apr 05 12:03:09 freesmartphone.org: 03mickey 07cornucopia * r458306121e45 10/.gitignore: gitignore++ Apr 05 12:03:57 mickey|office: what's the correct command to rebuild it fetching the new source ? Apr 05 12:04:17 i always run clean and then build again Apr 05 12:04:23 (bitbake -c clean ...) Apr 05 12:05:25 mickey|office: I didn't set the requirement of connman to 0.68? Apr 05 12:10:34 mickey|office: is it sufficient to do it only for fsogsmd Apr 05 12:10:36 ? Apr 05 12:18:54 mickey|office: hmm... no. It did compile without manually bumping Apr 05 12:19:01 strange thing indeed Apr 05 12:19:38 actually I wanted to disable fso-autorev and thought I failed, because it did not build the wanted rev. Apr 05 12:19:46 Did not check if it actually was autorev Apr 05 12:19:49 * mrmoku checks now Apr 05 12:20:29 mickey|office: duh... nvm I'm stupid :/ Apr 05 12:20:56 freesmartphone.org: 03morphis 07cornucopia * ra53427b12aaa 10/fsogsmd/ (src/connman/modemhandler.vala vapi/connman.vapi): fsogsmd: connman plugin: implement network registration on correct modem status Apr 05 12:22:38 mickey|office: I built some older rev instead... building the correct one now to see if it works then Apr 05 12:29:00 mickey|office: hmm... the correct rev does not build Apr 05 12:29:00 | Requested 'fso-glib-1.0 >= 2011.03.14.1' but version of libfso-glib is 2011.01.13.1 Apr 05 12:29:19 have to set that one too... Apr 05 12:30:17 * JaMa down to 11 errors in shr-core ;) Apr 05 12:30:26 * mrmoku hopes that won't fail then due to fso-specs... Apr 05 12:30:43 JaMa: yay Apr 05 12:30:43 dcordes: yes, just rebuilding fsogsmd should do Apr 05 12:31:05 morphis: no, you just forgot i guess :) Apr 05 12:32:37 as expected... | Requested 'fso-specs >= 2011.03.14.1' but version of fso-specs is 2011.02.01.1 Apr 05 12:34:24 c9faafeca2ebfc060b9affd92ba8a159d90429a1 then Apr 05 12:34:31 mickey|office: no problem :) Apr 05 12:36:44 well, i'm still building 8f4d4e7b5a4b9f63958ac7d1037ccc21a99bdf06, i have added a new vala and fsosystem Apr 05 12:56:44 and --disable-connman-support Apr 05 12:56:48 in fsogsmd Apr 05 13:11:26 mickey|office: ok that worked. thanks Apr 05 13:12:05 does anybody know how to bitbake transmission successfully? configure will always fail with the error that either gtk or cli option must be specified Apr 05 13:14:27 sounds like you need an EXTRA_OECONF = "--with-whatever-option" Apr 05 13:23:41 mrmoku: ok, here's my build http://amethyst.openembedded.net/~mickey/oe/n900/tmp/deploy/images/nokia900/ Apr 05 13:25:32 mickey|office: hm maybe it would be better to write a default value for that config in the recipe, and enable an override via extra conf or whatever Apr 05 13:25:38 mickey|office: so that it's buildable always Apr 05 13:25:53 dcordes: sure Apr 05 13:27:56 mickey|office: ok, will try if my updated image works and then your image Apr 05 13:33:36 mickey|office: 2011-04-05T00:07:18.341365Z [INFO] fsogsmd : GLib : gisicomm.vala:860: OPER = M-net Apr 05 13:33:40 2011-04-05T00:07:18.349972Z [INFO] NokiaIsiModem : Modem Status changed to FSO_GSM_MODEM_STATUS_ALIVE_REGISTERED Apr 05 13:33:47 ok, with the correct fso revs the image works :-) Apr 05 13:34:21 root@nokia900:~# opkg list_installed | grep libnl Apr 05 13:34:21 libnl-genl2 - 1:2.0-r2 Apr 05 13:34:21 libnl-nf2 - 1:2.0-r2 Apr 05 13:34:21 libnl-route2 - 1:2.0-r2 Apr 05 13:34:21 libnl2 - 1:2.0-r2 Apr 05 13:34:33 mickey|office, GNUtoo: which means our problem is _not_ libnl related Apr 05 13:34:51 ok Apr 05 13:40:57 * mrmoku installing mickeys new image Apr 05 13:41:18 yoh Apr 05 13:47:19 if we were to add the new illume keyboard layout in SHR I guess it must first find its way into e17 right ? Apr 05 13:49:02 dcordes: why not here? http://git.shr-project.org/git/?p=illume-keyboards.git;a=summary Apr 05 13:49:26 mrmoku: good job /me happy it's not libnl Apr 05 13:49:33 mrmoku: ah, interesting. ok, that confirms my suspicion Apr 05 13:50:15 JaMa: is it used in org.oe.dev for shr-images ? Apr 05 13:50:25 mrmoku: i hope it's not gcc or something :/ Apr 05 13:50:58 dcordes: yes Apr 05 13:51:56 dcordes: recipes/e17/illume-keyboards-shr_git.bb Apr 05 13:54:12 mickey|office: hmm... yours is giving me the following error: Apr 05 13:54:12 2011-04-05T14:28:21.949920Z [DEBUG] fsogsmd : GLib : mediators.vala:343: error = 224 Apr 05 13:54:49 \o/ 224 Apr 05 13:54:58 mrmoku: bummer, that i had as well with a recent image Apr 05 13:54:59 ;-D Apr 05 13:55:57 well, probably 343/224 is kinda semi-decent clue Apr 05 13:56:10 not really Apr 05 13:56:19 it's a timeout in the ISI stack Apr 05 13:56:25 which means we're lost Apr 05 13:56:25 :-S Apr 05 13:56:32 yep :-/ Apr 05 13:57:11 mickey|office: that is a different problem then? Apr 05 13:57:28 or the same with a different symptom? Apr 05 13:57:34 because the link does not go down Apr 05 13:57:36 mrmoku: no, i'm afraid it's somewhat related, but just another artefact Apr 05 13:57:56 but i'm not sure, tbh Apr 05 13:58:12 debugging blackboxes. Always *real* fun Apr 05 13:58:36 exactly Apr 05 13:58:40 it's driving me crazy Apr 05 13:58:59 mickey|office: ok, could you please add shr ssh://git@git.shr-project.org/n900-oe.git (push) Apr 05 13:58:59 that's about the most unpleasant work we have to do here Apr 05 13:59:17 as remote and push with what you built as new branch there? Apr 05 13:59:26 JaMa: ok nice Apr 05 13:59:56 mrmoku: you mean fsosystem, fsogsmd changes, and the new vala? Apr 05 14:00:03 JaMa: but I think we have to wait until [Rui] adds the new alternative key features in upstream Apr 05 14:00:17 JaMa: else the large button keyboard layout will be not very usable Apr 05 14:00:24 mickey|office: yeah, just do a dirty commit of the state you built with Apr 05 14:00:36 mrmoku: okay Apr 05 14:01:02 despite the layout, are there some ways to configure the illume-keyboard behaviour? like sensitivity etc Apr 05 14:01:08 in the end it will be the new vala :-P Apr 05 14:01:20 sensitivity?? Apr 05 14:01:21 brb Apr 05 14:02:55 DocScrutinizer: like for how long a key has to be pressed (same thing is configurable for the app icons in illume-home) Apr 05 14:03:12 aah Apr 05 14:03:32 kinda like debounce_time Apr 05 14:04:03 thought you asked for pressure etc Apr 05 14:05:32 mrmoku: To ssh://git@git.shr-project.org/n900-oe.git Apr 05 14:05:32 * [new branch] testing -> testing Apr 05 14:06:05 mrmoku: the new vala... i would not rule that out, yeah :/ Apr 05 14:07:02 mickey|office: vala is every time good for new bugs ... Apr 05 14:07:16 mickey|office: btw. do you integrated fsoaudio API already in zhone2? Apr 05 14:07:24 morphis: not yet, no Apr 05 14:07:32 ok Apr 05 14:07:54 i had to set up my EFL tree again Apr 05 14:07:58 DocScrutinizer: don't know about debounce_time , no I don't mean ts driver specific things Apr 05 14:07:59 and compiling this by hand always takes a long Apr 05 14:08:17 do you use ubuntu or debian? Apr 05 14:08:24 ubuntu Apr 05 14:08:37 as debian already has the 1.0.0 packages in it's repository Apr 05 14:09:46 next time you're in contact with palm, ask them to open mojo Apr 05 14:10:59 I have a sysfs device that will make the phone vibrate for n ms if you write n > /sys/class/devices/vibrate (or something like that) . would it be possible to include a function in illume-keyboard that writes such a value on every key press ? maybe it could execute a script so it is configurable for different devices. I know some vibrator drivers have an on or off state Apr 05 14:11:08 AKA haptic feedback Apr 05 14:11:43 mickeyl: mojo? mojo is only the javascript framework ... but I know what you mean :) Apr 05 14:11:57 and mojo will die Apr 05 14:12:12 they already introduced their new javascript framework for the comming years Apr 05 14:12:15 Enjo Apr 05 14:12:25 i mean the UI toolkit specifically Apr 05 14:13:24 don't know if it is really a UI toolkit, all applications you see are html+css+javascript, the top bar + notifacation seems to be hard coded and based upon their own vector graphics library Piranha Apr 05 14:25:36 mickey|office: ok, will update vala in my working image now Apr 05 14:25:48 mrmoku: yes, good. and only that, please Apr 05 14:25:52 and rebuild ... hmm... rebuild what? Apr 05 14:25:54 only fsogsmd? Apr 05 14:25:59 or libfso-glib? Apr 05 14:26:05 start with fsogsmd and libgisicomm Apr 05 14:26:10 ok Apr 05 14:26:21 those two are tightly coupled for the n900 Apr 05 14:28:51 mickey|office: you have checksums disabled? Apr 05 14:28:53 | ERROR: Function 'Fetch failed: http://downloads.freesmartphone.org/sources/vala-0.11.7.4.tar.bz2;name=archive checksum mismatch.' failed Apr 05 14:36:22 baah I hate this "libtool: link: `/lib/x86_64-linux-gnu/libgcrypt.la' is not a valid libtool archive" Apr 05 14:39:45 mrmoku: possible, i didn't bother to change them Apr 05 14:40:02 mrmoku: i just cp'ed the file from 0.11.something Apr 05 14:40:38 playya_: ping Apr 05 14:43:50 s/ link: `/lib/x86_64-linux-gnu/libgcrypt.la' is not a valid libtool archive"/"!/ Apr 05 14:45:21 mrmoku: funny wording anyway Apr 05 14:52:36 mickey|office: ok, rebuilt fsogsmd + libgisi with newer vala... and what do I get? Apr 05 14:52:43 2011-04-04T23:47:03.320465Z [DEBUG] fsogsmd : GLib : gisicomm.vala:102: NETLINK STATE = PN_LINK_DOWN Apr 05 14:52:46 2011-04-04T23:47:03.321441Z [DEBUG] fsogsmd : GLib : gisicomm.vala:102: NETLINK STATE = PN_LINK_UP Apr 05 14:53:27 * mickey|office turns pale Apr 05 14:53:48 can you switch back to the old libgisi Apr 05 14:53:50 ? Apr 05 14:54:03 fsogsmd is _HUGE_ Apr 05 14:54:09 it better be libgisi Apr 05 14:54:12 :) Apr 05 14:54:15 (or comm, for that matter) Apr 05 14:54:16 ok Apr 05 14:54:19 thanx Apr 05 14:55:46 hmm... there is no morgue no more? Apr 05 14:56:45 that makes it a bit more work... Apr 05 14:56:49 * mrmoku reinstalls fresh image Apr 05 14:57:46 when you switch forward and back to verify everything, can you do me another favor please? Apr 05 14:57:54 sure? Apr 05 14:58:08 it'd be very helpful if you could log the packets going over phonet0 for a successful and for an unsuccessful run Apr 05 14:58:19 with tcpdump? Apr 05 14:58:20 e.g. in some init script calling tcpdump ... Apr 05 14:58:21 yeah Apr 05 14:58:45 ok, can put that into the cmt initscript Apr 05 14:59:03 mickey|office: any parameters I have to feed to tcpdump? Apr 05 14:59:10 because i can only imagine that for some reason the messages get sent in a different order or get destroyd Apr 05 14:59:11 let me check Apr 05 15:00:20 tcpdump -i phonet0 -w >somefile Apr 05 15:00:22 err Apr 05 15:00:23 without > Apr 05 15:02:24 ok, mplayer does not freeze the palm pre device anymore Apr 05 15:02:47 mickey|office: ok Apr 05 15:02:55 morphis: yay, good work Apr 05 15:04:20 mickey|office: I just compiled the 1.0rc2 version Apr 05 15:04:30 as the rc4 does not work Apr 05 15:06:19 hum mplayer git uses "git://repo.or.cz/mplayer/glamo.git" Apr 05 15:06:22 why glamo? Apr 05 15:06:38 wasn't that the name of the om-gta02 gpu? Apr 05 15:07:18 JaMa: ping Apr 05 15:07:24 morphis: I think so Apr 05 15:07:25 morphis: because I have RW there Apr 05 15:07:43 * dcordes remembers xglamo Apr 05 15:07:52 morphis: dont worry about name, it's just upstream mplayer2.git + patch for -vo glamo Apr 05 15:08:03 morphis: so won't break other devices Apr 05 15:08:07 JaMa: but it is behind mplayer upstream Apr 05 15:08:25 I know, I didn't have time to merge upstream lately Apr 05 15:08:41 but I plan it now after switching to libav Apr 05 15:08:55 JaMa: ok, as the current version causing the palmpre to freeze Apr 05 15:09:18 ah sorry, so I'll try harder to find time Apr 05 15:09:23 I will switch then only locally to upstream Apr 05 15:09:29 ok Apr 05 15:09:32 ok Apr 05 15:09:41 no problem, you have like too much things to do :) Apr 05 15:10:22 shr-core looks much better today :) Apr 05 15:10:30 JaMa: btw. the shr-chroot is great :) Apr 05 15:10:33 really? Apr 05 15:10:46 I tried to set it up yesterday but that failed Apr 05 15:10:55 morphis: yup fixed about 30 recipes today Apr 05 15:11:01 ok Apr 05 15:11:58 mickey|office: interesting.... Apr 05 15:12:04 yes? Apr 05 15:12:09 with old gisi it still shows me the providers Apr 05 15:12:15 but does not register Apr 05 15:12:18 link goes down Apr 05 15:12:41 duh.. my phonet.dump is empty Apr 05 15:12:46 hmm Apr 05 15:15:16 morphis: how it failed? Apr 05 15:15:35 mickey|office: ok, have to interrupt for a moment... family duty... bbiab Apr 05 15:15:52 JaMa: while 'make setup-shr-core' Apr 05 15:15:58 will search for the correct error later Apr 05 15:16:03 or wait a moment .. Apr 05 15:22:42 ok, was faster than expected :) Apr 05 15:24:27 JaMa: http://pastie.org/1759151 Apr 05 15:28:04 morphis: do you have current common? and from branch master? Apr 05 15:28:35 I hope so, will look Apr 05 15:28:40 morphis: make update-common should do it Apr 05 15:30:47 last 3 :) Apr 05 15:30:48 ERROR: 'virtual:native:/OE/shr-core/meta-shr/recipes-devtools/vala-dbus-binding-tool/vala-dbus-binding-tool_git.bb' failed Apr 05 15:30:51 ERROR: '/OE/shr-core/meta-shr/recipes-shr/3rdparty/ffphonelog_git.bb' failed Apr 05 15:30:54 ERROR: '/OE/shr-core/meta-shr/recipes-shr/3rdparty/iliwi_git.bb' failed Apr 05 15:31:23 mickey|office: do I need some special magic for valac to look into sysroot for .vapi? Apr 05 15:32:37 mickey|office: sysroot paths are changed a bit http://paste.pocoo.org/show/365940/ Apr 05 15:33:48 JaMa: take a look at vala.bbclass Apr 05 15:33:49 # Upstream Vala >= 0.11 looks in XDG_DATA_DIRS for .vapi files Apr 05 15:33:49 export XDG_DATA_DIRS = "${STAGING_DATADIR}" Apr 05 15:33:59 ok, thanks Apr 05 15:36:17 ffphonelog wasn't using vala.bbclass and still it worked before, and it works (partially again with this export) Apr 05 15:37:48 fails in c level the same as iliwi, maybe libeflvala needs update http://paste.pocoo.org/show/365943/ Apr 05 15:37:53 hy guys! someone know if there is a FSO library for accelerometer data retrieval? Apr 05 15:38:05 n3v3rmind: for a game? Apr 05 15:38:16 not...university application :) Apr 05 15:38:31 n3v3rmind: what does it do? ;) Apr 05 15:38:33 there's no FSO abstraction for raw data, we just provide orientation data. Apr 05 15:38:45 should think about something like that... Apr 05 15:38:49 mickey|office: is that over dbus or is there something else? Apr 05 15:39:00 lindi-: orientation data over dbus, yes. what else :) Apr 05 15:39:09 mickey|office: think games :P Apr 05 15:39:16 lindi-:i've to monitor a flow of every kind of object such a plane...aircraft :) Apr 05 15:39:27 lindi-:i've to monitor a flight of every kind of object such a plane...aircraft :) Apr 05 15:39:40 lindi-: yes, i know. if i were to come up with an abstraction for raw data, i'd provide a library Apr 05 15:39:50 the current way is just orientation via dbus Apr 05 15:41:02 btw....i've just found a tutorial on the wiki...someone need it? Apr 05 15:44:43 JaMa: I have latest common but with a little modification to the Makefile everything works fine http://pastie.org/1759223 Apr 05 15:48:56 morphis: you're right, thanks, pushed Apr 05 15:49:02 SHR: 03morphis 07shr-makefile * r446f7d1b5287 10/Makefile: Makefile: fix path to common/conf Apr 05 15:50:14 JaMa: no problem :) Apr 05 15:51:40 JaMa: "Pseudo is not present but is required, building this first before the main build" - this means it will build pseudo on it's own? Apr 05 15:52:28 morphis: are you using setup-env from shr-core? Apr 05 15:52:38 jepp Apr 05 15:52:51 cd shr-core && . ./setup-env Apr 05 15:52:56 which bitbake shows openembedded-core/scripts/bitbake? Apr 05 15:53:25 `which bitbake` :) Apr 05 15:53:40 "/OE/shr-core/openembedded-core/scripts/bitbake" Apr 05 15:53:47 then you should be fine Apr 05 15:53:56 ok Apr 05 15:54:38 wuuhu, 133 tasks for building msmcommd? Apr 05 15:54:46 thats ok? Apr 05 15:54:56 libfsobasics fails to find libnl2 lib, but that has to wait http://paste.pocoo.org/show/365954/ Apr 05 15:55:26 strange, I would expect MUCH mor Apr 05 15:55:27 e Apr 05 15:55:45 133 is just pseudo-native Apr 05 15:55:56 then it will restart and show task count for msmcommd Apr 05 15:56:14 gtg Apr 05 15:56:21 ah ok Apr 05 15:58:28 btw some local modifications for meta-oe/oe-core are available at https://gitorious.org/shr Apr 05 15:58:45 but iirc nothing from it is needed for msmcommd Apr 05 16:04:54 JaMa|Off: http://pastie.org/1759305 Apr 05 16:05:06 perl-native fails cause file to fetch is not available Apr 05 16:14:42 JaMa|Off: a D_P_palmpre = "5" in a recipe is ok? Apr 05 16:44:24 mickeyl: possible that tcpdump does not work with local netlink? Apr 05 16:48:16 mrmoku: is a D_P_palmpre = "..." in a recipe ok? Apr 05 16:49:12 mickeyl: nvm Apr 05 16:49:23 morphis: hmm... I have no idea Apr 05 16:49:45 actually I think we do something like that somewhere... Apr 05 16:49:48 not sure though Apr 05 16:51:42 mrmoku: ok Apr 05 17:06:44 bye Apr 05 17:27:28 mickeyl: ping Apr 05 17:30:30 freesmartphone.org: 03morphis 07framework * rdfeb4381aea8 10/etc/freesmartphone/oevents/palmpre/rules.yaml: oevents: update configuration for palmpre machine Apr 05 17:34:58 freesmartphone.org: 03morphis 07cornucopia * ra19483afb552 10/fsodeviced/conf/palm_pre/fsodeviced.conf: fsodeviced: palmpre config: let the power key be the aux key for now Apr 05 17:43:23 mickeyl: ping :-) Apr 05 17:44:43 mickeyl: http://build.shr-project.org/tests/mrmoku/phonet-dump/ Apr 05 17:45:23 I re-verified by upgrading libgisi to be built with new vala and it failed again Apr 05 17:49:21 hi morphis Apr 05 17:50:13 you forgott to tell why you needed that mplayer version Apr 05 17:50:51 GNUtoo|nexusone, because older versions freeze the palm pre afaik Apr 05 17:50:56 in the oe mplayer preffered version Apr 05 17:51:00 GNUtoo|nexusone: uups Apr 05 17:51:07 ah ok Apr 05 17:51:14 it's the latet one currently working for palmpre Apr 05 17:51:21 ok Apr 05 17:51:23 the 1.0rc4 let the palmpre freeze Apr 05 17:51:33 and latest git is not compilable Apr 05 17:51:40 neon? Apr 05 17:51:47 and JaMa want to take care about it already Apr 05 17:52:01 ok Apr 05 17:52:04 removed configure flags and some other compilation errors Apr 05 17:52:18 so I decided to switch to 1.0rc2 for now Apr 05 17:52:27 ok Apr 05 17:52:38 but you are right I should write this next time in the commit itself ... Apr 05 17:52:41 hi GNUtoo|nexusone Apr 05 17:52:45 hi Apr 05 17:53:00 do you have a phonet enabled tcpdump at hand? Apr 05 17:53:10 news on the n900? Apr 05 17:53:15 well, yeah Apr 05 17:53:18 getting closer Apr 05 17:53:23 ok nice Apr 05 17:53:31 looks like libgisi compiled with the new vala behaves badly Apr 05 17:53:49 you mean tcpdump and phonet Apr 05 17:54:02 yeah, I have two tcpdump raw dumps Apr 05 17:54:09 ahh... Apr 05 17:54:12 that was wireshart Apr 05 17:54:13 k Apr 05 17:54:21 with the phonet plugin Apr 05 17:54:23 hmm Apr 05 17:54:26 ok Apr 05 17:54:30 nvm then Apr 05 17:54:35 i don'y Apr 05 17:54:38 ok Apr 05 17:54:47 *don't Apr 05 18:53:04 mickeyl: nice thing is I have rm_work off :-) Apr 05 18:53:17 have the c source of both... working and not working libgisi Apr 05 18:53:52 * mrmoku wonders if trying 0.12.0 would be worth a shot Apr 05 18:59:53 freesmartphone.org: 03morphis 07cornucopia * r5eebf6d1aa75 10/fsogsmd/src/plugins/modem_qualcomm_palm/ (4 files): fsogsmd: modem_qualcomm_palm: report network status update only once and include mcc/mnc Apr 05 19:02:20 * mrmoku tries without asking the recipe maintainer as advised in the recipe ;) Apr 05 19:08:31 morphis, hmmm maybe you should really solve this neon issue Apr 05 19:08:39 that would require getting serial console etc... Apr 05 19:08:48 and looking trough kernels patches Apr 05 19:09:14 btw didn't palm use some NEON in pulseaudio? Apr 05 19:12:05 GNUtoo: which neon issue? the one with evas? Apr 05 19:12:14 not only evas Apr 05 19:12:16 but mplayer Apr 05 19:12:19 evas Apr 05 19:12:22 etc... Apr 05 19:12:31 GNUtoo: with mplayer must not be a neon issue Apr 05 19:12:37 ah ? Apr 05 19:12:49 it occurs even when starting mplayer without any media file attached Apr 05 19:12:52 mplayer has some neon assembly inside, at least recent versions Apr 05 19:12:59 ok Apr 05 19:13:06 then I should really look into it Apr 05 19:13:23 jepp Palm uses neon somewhere Apr 05 19:13:28 because else it would be problematic to lock old software for the pre Apr 05 19:13:40 in their libsdl patch and pulseaudio too Apr 05 19:13:44 for sure Apr 05 19:13:46 because at the end everything will be locked to old versions Apr 05 19:13:50 *would Apr 05 19:13:55 ok Apr 05 19:15:16 then I should take my old pre and try to find serial + jtag again Apr 05 19:15:55 YAY YAY YAY Apr 05 19:16:07 mickeyl, GNUtoo: with vala-0.12.0 it works :-D Apr 05 19:19:09 wpw Apr 05 19:19:11 *wow Apr 05 19:19:14 !!!!! Apr 05 19:19:39 JaMa|Off: something to watch out for when switching distro? Apr 05 19:22:19 2011-04-05T19:22:00.519072Z [DEBUG] FsoGsmModemHandler <>: Establishing GSM PDP connection ... Apr 05 19:22:19 2011-04-05T19:22:00.519112Z [DEBUG] FsoGsmModemHandler <>: mccmnc = 26202 Apr 05 19:22:19 2011-04-05T19:22:00.533469Z [DEBUG] FsoGsmModemHandler <>: Using apn = "mail.partner.de", username = "", password = "" Apr 05 19:22:35 yeah! activating FSO PDP conneciton with connman works! Apr 05 19:22:53 with auto-configuration, no more setting the apn/username/password Apr 05 19:24:35 morphis, 3g works on the pre? Apr 05 19:27:43 freesmartphone.org: 03morphis 07cornucopia * r284ba58dfd7d 10/fsogsmd/src/plugins/modem_qualcomm_palm/mediatorhelpers.vala: fsogsmd: modem_qualcomm_palm: format mcc/mnc correct when printing them together Apr 05 19:27:46 freesmartphone.org: 03morphis 07cornucopia * ra5a0ef6e5ddd 10/fsogsmd/src/connman/ (modemhandler.vala plugin.vala): fsogsmd: connman plugin: activate pdp connection when network is connecting Apr 05 19:30:00 GNUtoo: jepp Apr 05 19:30:27 wow nice Apr 05 19:31:11 thats why I am working on the connman plugin Apr 05 19:31:24 ok Apr 05 19:31:42 but fso had data of apn and providers right? Apr 05 19:32:07 anyway, Apr 05 19:32:22 under android there is the option to enable or not data Apr 05 19:32:26 and it's important Apr 05 19:32:33 use case: Apr 05 19:32:37 I've per day data rate Apr 05 19:32:44 if I use it I pay 1E Apr 05 19:32:48 if I don't I pay nothing Apr 05 19:32:55 use beeing make a data conenction Apr 05 19:33:04 *making Apr 05 19:33:48 congratulation you got your device to be the second one to get 3g Apr 05 19:33:52 along with htc dream Apr 05 19:34:48 :) Apr 05 19:35:06 SHR: 03lukasmaerdian 07shr-themes * rc2bc72a01653 10/e-wm/e-wm-config-illume2-shr/ (Makefile.am palm_pre/module.battery.src): e-wm-config-illume2-shr: palmpre: use dbus backend for battery Apr 05 19:35:06 the use case is correct, but thats an option that should solve the user of the FSO API Apr 05 19:35:17 ok Apr 05 19:35:47 such as shr-settings Apr 05 19:35:48 wow, Slyon did a dbus backend for the battery! Apr 05 19:35:49 and similar Apr 05 19:35:51 GNUtoo: jepp Apr 05 19:36:05 wow Apr 05 19:36:07 for the battery Apr 05 19:36:56 Slyon: heyho Apr 05 19:37:00 ~hail Slyon Apr 05 19:37:02 morphis, hey Apr 05 19:37:04 Slyon: I saw your commit message Apr 05 19:37:09 mrmoku, hmn? Apr 05 19:37:19 ~ping Apr 05 19:37:21 hmm Apr 05 19:37:28 morphis, yeah... our fso battery is almost working Apr 05 19:37:36 Slyon: wanted apt to thank you for that battery thing :) Apr 05 19:37:42 ah heh :) Apr 05 19:37:46 Slyon: great work! Apr 05 19:37:58 Slyon: SHR: lukasmaerdian shr-themes * rc2bc72a01653 /e-wm/e-wm-config-illume2-shr/ (Makefile.am palm_pre/module.battery.src): e-wm-config-illume2-shr: palmpre: use dbus backend for battery Apr 05 19:38:35 yes. that is needed as well for the fso battery. Apr 05 19:38:46 i extended the e battery gadget Apr 05 19:38:59 do we want it as a patch in OE or should i try to get it upstream? Apr 05 19:39:36 try to get it upstream Apr 05 19:39:40 if it is possible Apr 05 19:39:47 but for now we can patch it in OE Apr 05 19:39:56 how big is the patch? Apr 05 19:40:20 11,2 KB Apr 05 19:40:24 wow Apr 05 19:40:31 can you paste it somewhere? Apr 05 19:40:38 sure Apr 05 19:41:03 morphis, http://pastie.org/1760129 Apr 05 19:41:10 btw. we have now 3G connections for the Palm Pre ready Apr 05 19:41:21 morphis, yay! great work! Apr 05 19:41:39 to use the patch, you need some OE_EXTRACONF Apr 05 19:42:25 --enable-fso --enable-device-udev=no Apr 05 19:42:46 which makes it device specific Apr 05 19:42:59 makes e-wm device-specific* Apr 05 19:44:26 you already did the OE work? Apr 05 19:45:49 almost only thing missing is the MACHINE_ARCH thing. but I can do this Apr 05 19:46:17 it should be enabled for all fso devices Apr 05 19:46:22 not just for palm-pre Apr 05 19:46:33 we could do it on SHR basis as well Apr 05 19:46:38 all shr devices Apr 05 19:46:44 that's what I meant Apr 05 19:46:46 shr distro Apr 05 19:46:52 not device specidic Apr 05 19:47:01 like _shr = "--enable-fso" Apr 05 19:47:13 yup Apr 05 19:48:19 morphis, there are some problems... 1.: fsodeviced status reporting (e.g. charging/discharging) is very slow (signal comes very late) Apr 05 19:48:20 we should do it for all SHR devices Apr 05 19:48:39 you mean when plugging the charger in? Apr 05 19:49:12 morphis, jep. it takes a while until fsodeviced sends the charging singal after pluging in the charger Apr 05 19:49:40 2.: I don't know if my patch is fully stable yet. and it's pretty hacky Apr 05 19:49:54 [fsodevice.palmpre_quirks/battery] poll_timeout = 10 Apr 05 19:51:02 thats the default value Apr 05 19:51:06 you can try to increase it Apr 05 19:51:41 Slyon: ok, then we should test it a little bit more before submitting it upstream Apr 05 19:52:09 yes, and maybe TAsn or somebody else with some E skills should look at it Apr 05 19:52:20 I got my eyes poked out Apr 05 19:52:38 can't take a look at anything Apr 05 19:52:43 heh Apr 05 19:52:46 luckily for me, I get irc wireless Apr 05 19:53:14 TAsn, maybe you can get this wirless as well ;) http://pastie.org/1760129 Apr 05 19:53:30 * TAsn is reading, sec. Apr 05 19:53:42 *wirelessly probing that link Apr 05 19:54:45 I wonder if this can go upstream or not. Apr 05 19:54:59 best thing would probably be sending it to e-devel Apr 05 19:55:40 ok. but probably we should first test it for a while and squash out bugs Apr 05 19:55:45 yep. Apr 05 19:56:23 as I didn't have any experience with neither e-modules nor e_dbus nor a lot of C Apr 05 19:56:30 before this work Apr 05 19:56:35 :) Apr 05 19:58:38 morphis, was [fsodevice.palmpre_quirks/powersupply/battery] what you meant? Apr 05 19:58:52 as i don't have any poll_timeout in my fsodeviced.conf Apr 05 19:59:13 Slyon, btw, I have a couple of comments Apr 05 19:59:26 +#else Apr 05 19:59:26 #ifdef HAVE_EEZE Apr 05 19:59:32 #elif defined(HAVE_EEZE) Apr 05 19:59:37 should be used instead Apr 05 20:00:03 but making it an #else Apr 05 20:00:05 is bad anyway Apr 05 20:00:16 cause you don't want to disable eeze if you build for fso Apr 05 20:00:26 the reason eeze disables hal is because it deprecates it... Apr 05 20:01:04 but you shouldn't disable anything Apr 05 20:01:19 Slyon, ^ Apr 05 20:01:38 haven't reviewed anything more Apr 05 20:02:10 TAsn, yes i had quiet a few problems with that in the beginning and that was the way i got it to build. but i'll look at it again Apr 05 20:02:20 cool. Apr 05 20:02:27 you should check what's enabled in the configuration Apr 05 20:03:00 and enable according to that Apr 05 20:03:08 but everything should be enabled in configuration Apr 05 20:03:22 grr Apr 05 20:03:29 but everything should be enabled in compilation Apr 05 20:03:50 ok. you mean everything should be enabled at build time and selected at run time via config Apr 05 20:03:53 right? Apr 05 20:04:07 yep. Apr 05 20:04:14 look at the configuration options you got now Apr 05 20:04:19 and check how taht's done. Apr 05 20:04:25 yes, will do Apr 05 20:05:09 but +1 for the initiative. :) Apr 05 20:05:22 thanks :) Apr 05 20:08:56 Slyon: yes Apr 05 20:11:32 morphis: strange it should be on shr buildhost available Apr 05 20:11:40 morphis: (perl-native) Apr 05 20:11:52 mrmoku: with same tmpdir? Apr 05 20:12:08 JaMa|Off: yup Apr 05 20:12:09 mrmoku: bad idea, you should rebuild from scratch when switching distro Apr 05 20:12:13 ok :) Apr 05 20:12:24 I knew I better ask :P Apr 05 20:12:53 morphis, are warnings like this one OK on fsodeviced startup? 2010-06-14T16:43:29.506317Z [WARN] subsystem : Could not load plugin: Could not load /usr/lib/cornucopia/modules/fsodevice/palmpre_quirks/powersupply/battery.so: /usr/lib/cornucopia/modules/fsodevice/palmpre_quirks/powersupply/battery.so: cannot open shared object file: No such file or directory Apr 05 20:13:04 well mostly it's about pinned versions, but it could be also thumb/no-thumb policy etc.. Apr 05 20:13:43 yeah, I will take the safe (slow) route and rebuild from scratch then Apr 05 20:13:50 just to be sure Apr 05 20:14:53 Slyon: no, they are not ok Apr 05 20:15:09 morphis, so something is wrong with my config? Apr 05 20:15:18 but the date of the logging entry seems to be very old 2010-06-14 Apr 05 20:15:39 it's from now. but i haven't synced my time on the pre Apr 05 20:16:23 ok Apr 05 20:16:29 they appear when i killall fsodeviced; fsodeviced & Apr 05 20:24:59 hm Apr 05 20:26:12 Slyon: config should look like this: http://pastie.org/1760353 Apr 05 20:27:02 morphis, okay thanks. i'll try with this one Apr 05 20:30:09 SHR: 03lukasmaerdian 07shr-themes * rb303244ac9a3 10/e-wm/ (6 files in 6 dirs): e-wm-theme-illume-*: remove default text 'time' in the battery gadget Apr 05 20:33:52 Slyon: when upstream does not accept the patch we can do our own battery gadget in shr-e-modules Apr 05 20:35:03 morphis, well I'm planning to do it the right way (TM). and it should then be accepted upstream. but this could take a while, as i'll have to dig in a lot of stuff Apr 05 20:37:39 ok Apr 05 20:38:02 but good to hear someone is working on it Apr 05 20:38:58 yeah. I've got some more weeks of holidays left :) Apr 05 20:50:31 hi folks, looks like you all got amazing progress today Apr 05 20:50:47 mrmoku: unbelievable. can you get me the diff of the C file? Apr 05 20:50:50 morphis: congrats Apr 05 20:56:29 mickeyl: :) Apr 05 20:56:55 mickeyl: all in all it's not finished, it needs some more polishing to get everything with connman right Apr 05 20:56:59 i'll leave for today. gn8 guys! Apr 05 20:57:03 sadly there is not documentation at all .. Apr 05 20:57:09 Slyon: gn8! Apr 05 20:58:13 mickeyl: yeah, but we don't need it anymore :) Apr 05 20:58:31 mickeyl: I can provide with the two tarballs of the workdir Apr 05 20:58:43 +you Apr 05 20:58:55 mrmoku: yes, please. i'm still curious, as i don't understand it yet Apr 05 20:59:02 ok, moment Apr 05 21:01:01 mickeyl: http://build.shr-project.org/tests/mrmoku/n900/libgisi-not-working.tar.bz2 Apr 05 21:01:08 http://build.shr-project.org/tests/mrmoku/n900/libgisi-working.tar.bz2 Apr 05 21:01:22 thanx Apr 05 21:03:30 mickeyl: maybe I have somebody who does a little bit of publicity/web restructuring work for us Apr 05 21:03:43 that'd be helpful Apr 05 21:05:14 hmm Apr 05 21:05:22 ok, thanks, but that diff is unreadable Apr 05 21:05:45 anyways, now we can finally move on with audio, can we? Apr 05 21:06:39 brb Apr 05 21:12:14 mickeyl: unreadable yeah... like vala produced c code in general :-P Apr 05 21:13:17 mickeyl: (audio) yeah, could you bump vala please? Apr 05 21:14:14 vala? Apr 05 21:14:23 oh you mean the required version? Apr 05 21:14:32 so I am off for today Apr 05 21:14:33 gn8 Apr 05 21:14:37 n8 morphis Apr 05 21:14:37 is vala-0.12.0 in oe already? Apr 05 21:14:40 yes Apr 05 21:14:43 ahh, ok Apr 05 21:14:53 i will add a requirement on .12 though Apr 05 21:14:58 good Apr 05 21:15:19 and I will merge current oe master into our n900 exploration branch Apr 05 21:15:28 night morphis Apr 05 21:15:30 excellent Apr 05 21:15:48 things move again finally :-) Apr 05 21:16:44 freesmartphone.org: 03mickey 07libgisi * r713ab918a710 10/configure.ac: bump vala requirement to 0.12, certain 0.11.x versions generate broken code Apr 05 21:16:50 yeah. so what's our plan of attack wrt. audio? Apr 05 21:16:54 call audio, that is Apr 05 21:17:17 explore libcmtspeech probably Apr 05 21:17:28 yep Apr 05 21:17:34 did not yet look at it and actually have no idea Apr 05 21:17:51 btw. would it be difficult to implement SetFunctionality? Apr 05 21:18:33 depends, it's not exactly straightforward, but i can tackle it next Apr 05 21:19:04 ok, because then I could use our ui for the calls... maybe Apr 05 21:19:09 at least pin dialog pops up Apr 05 21:19:25 yeah, i can imagine. i'll keep it in mind Apr 05 21:19:30 good Apr 05 21:19:36 fwiw, libcmtspeechdata is in oe and i started writing a vapi for it Apr 05 21:20:02 that's where i left when going skiing Apr 05 21:20:04 yeah, saw it Apr 05 21:20:15 while searching for the correct commit to build :) Apr 05 21:20:20 hehe Apr 05 21:20:32 * mrmoku checks if he already has a checkout of cmtspeech Apr 05 21:22:17 no, cloning Apr 05 21:28:08 mickeyl: looks like being quite straight forward Apr 05 21:28:42 question is if libcmtspeechdata is enough though Apr 05 21:29:10 yeah. thought so as well. we'll learn it, i guess Apr 05 21:30:56 probably the hard way like always :) Apr 05 21:31:18 *nod* Apr 05 21:33:03 given the current state w/ palm pre and n900, we should really start planning a get-together-weekend soon Apr 05 21:38:21 oh yes Apr 05 21:46:02 mickeyl: ok, tomorrow I will build a fresh SHR image and start to play with calls Apr 05 21:46:49 awesome. i don't have much time the coming days, but i'll try to take care of SetFunctionality frst Apr 05 21:46:51 g'night Apr 05 21:46:55 IIUC libcmtspeech gets audio data from the modem and we must handle that Apr 05 21:47:07 yep. it looks like a tight copying loop Apr 05 21:47:23 that we actually get the pcm data is pretty exciting Apr 05 21:47:34 yup Apr 05 21:47:47 yeah Apr 05 21:47:49 * mrmoku bed too Apr 05 21:47:54 gnight:) Apr 05 21:48:02 n8 Apr 05 22:49:07 finally shr-core/tmp/deploy/images/nokia900/shr-lite-eglibc-ipk--20110405-nokia900.rootfs.tar.gz Apr 05 22:49:37 gnight Apr 05 22:55:41 hi mrmoku Apr 05 22:56:01 I want to rebase the changes of the n900 branch over oe-dev, how di I do that? Apr 05 22:58:21 GNUtoo: git rebase -i origin/master Apr 05 22:59:19 ok Apr 05 22:59:34 without the -i I bet Apr 05 22:59:42 as -i is for interactive Apr 05 23:00:01 how do you want to rebase it? Apr 05 23:00:18 automatically Apr 05 23:00:31 like if I cherry-picked the n900 commits on top of origin/master Apr 05 23:01:32 but there are multiple commits which are actually just one change merged multiple times.. so normally I have to squash such changes (during interactive rebase) Apr 05 23:01:42 ok Apr 05 23:01:46 then I'll do that Apr 05 23:02:14 and expect that you will have to add -f to git push Apr 05 23:02:47 I know Apr 06 01:09:40 anybody knows what the new mechanism is to add apps in latest (in oe) elfe revision ? Apr 06 01:09:56 I accidently showed up the menu but can't figure how Apr 06 01:16:22 ah ok Apr 06 01:16:38 the puzzle icon in the bottom right has to be tapped _once_ Apr 06 01:16:55 if you go hectic and tap it twice it won't work again until restart of the program Apr 06 02:46:50 JaMa|Zzz: if the same commit is merged multiple times, Git should just ignore it... **** ENDING LOGGING AT Wed Apr 06 02:59:58 2011