**** BEGIN LOGGING AT Sun Aug 23 02:59:56 2009 Aug 23 04:00:07 hi all. I just upgraded my shr-u install and am having alot of problems Aug 23 04:00:27 it seems none of the dbus resources are working Aug 23 04:00:39 is anyone else having this problem? Aug 23 07:46:15 max_posedon: removing gsm0710muxd should be enough Aug 23 07:46:28 max_posedon: then ensure you have fso-abyss set as a muxer in frameworkd.conf Aug 23 07:46:32 that fixes the gsm not available thing? Aug 23 07:46:50 max_posedon: then try to start fso-abyss from the command line and see if there're any interesting errors. It should be obvious how to fix those. Aug 23 07:46:53 farhaven: yes Aug 23 07:46:59 great Aug 23 07:49:45 crap, that package is not installed Aug 23 07:52:46 farhaven: quite possibly something was broken due to recent fso-abyss changes so possibly one needs to dig deeper to solve the issue. Aug 23 07:53:00 ugh Aug 23 07:54:49 hmm, what service provides the dbus services to frameworkd? Aug 23 07:55:03 farhaven: which services? Aug 23 07:55:03 because it's complaining about services not being provided by .service files a lot Aug 23 07:55:14 CPI, GSM, Display, Bluetooth and WiFi Aug 23 07:55:18 *CPU Aug 23 07:55:26 which would boil down to the missing stuff Aug 23 07:55:26 farhaven: that's what frameworkd provides itself Aug 23 07:55:29 hmm Aug 23 07:56:00 farhaven: probably there's some more meaningful error in the beginning of the log. Aug 23 07:56:13 ah Aug 23 07:56:28 "the name org.freesmartphone.ogmsg was not provided by any .service file" Aug 23 07:56:43 that's the first error/warning message i have in my log Aug 23 07:57:07 the next are said complaints about dbus Aug 23 08:02:40 farhaven: hm, strange Aug 23 08:02:57 farhaven: are you sure that's "ogmsg" instead of "ogsmd"? Aug 23 08:03:09 heh, typo Aug 23 08:03:13 it's ogsmd Aug 23 08:03:34 i got a little bit further though Aug 23 08:03:41 How? Aug 23 08:04:03 in /etc/frameworkd.conf, ousaged was disabled Aug 23 08:04:19 farhaven: that's how it should be, fso-usage should be used instead Aug 23 08:04:19 after enabling it, ophonekitd no longer complains about no GSM being available Aug 23 08:04:23 ah Aug 23 08:04:33 do i need to enable that somewhere? Aug 23 08:04:50 farhaven: hm, probably you have problems with fso-usage then? Yes, it should be enabled explicitly in frameworkd conf. Aug 23 08:05:28 farhaven: http://www.mail-archive.com/smartphones-userland@linuxtogo.org/msg01440.html Aug 23 08:07:23 hmm, that's already in there Aug 23 08:08:43 farhaven: try starting fso-usaged by hand, probably it will complain about something. Aug 23 08:08:52 i hope so :) Aug 23 08:09:23 ah Aug 23 08:09:27 ERROR:subsystem.c:1286:fso_framework_dbus_subsystem_real_registerServiceObject: assertion failed: (conn != NULL) Aug 23 08:09:37 looks like dbus-fail Aug 23 08:10:04 farhaven: no, looks like it can't register itself as a service. Is frameworkd with old ousaged currently running? Aug 23 08:10:17 might be, i'll restart it Aug 23 08:10:48 farhaven: better kill Aug 23 08:10:50 it segfaults Aug 23 08:11:01 fsousaged that is Aug 23 08:11:12 farhaven: now that's some evidence. Aug 23 08:11:27 farhaven: can you try to reinstall fsousaged and all of its dependencies? Aug 23 08:11:32 yup Aug 23 08:11:41 how does that dependencies thing work? Aug 23 08:12:26 as in, how do i get the dependencies? Aug 23 08:12:28 farhaven: i'm not opkg user, just find some command to list dependencies. Aug 23 08:12:33 k Aug 23 08:17:47 farhaven: if that fails, then start fsousaged in gdb, look if the backtrace is anything meaningful. If not, install relevant -dbg packages and retry. Aug 23 08:18:03 will do Aug 23 08:25:53 hmm Aug 23 08:26:00 the backtrace isn't really useful Aug 23 08:26:18 http://farhaven.pastebin.com/d28b593ee Aug 23 08:31:06 farhaven: no, it's indeed meaningful. Aug 23 08:31:13 cool :D Aug 23 08:31:57 farhaven: there was some logger-related problem. Are you really sure you've reinstalled the libs that fsousaged depends upon? Aug 23 08:32:54 yes, all of them Aug 23 08:34:07 farhaven: btw, do you know "bt full" gdb command? Aug 23 08:34:21 nope Aug 23 08:34:36 hmm, doesn't yield any more information Aug 23 08:34:43 only mentions "no locals" a few times Aug 23 08:34:59 farhaven: it shows all available information when gdb knows names of variables on stack. Aug 23 08:35:12 neat Aug 23 08:38:29 farhaven: this one was about logging: http://git.freesmartphone.org/?p=cornucopia.git;a=commitdiff;h=d751bb0cf4ebf18ecadb0ad5bf9bff3df0d38641 Aug 23 08:39:54 hmm, but how would that commit make logger.info() segfault? Aug 23 08:45:04 farhaven: binary api was changed, that commit made everything work again, so you need to make sure your fsousaged and libfsobasics are newer than 10 days ago. Aug 23 08:45:28 hmm, i reinstalled both half an hour ago Aug 23 08:45:34 so they should be quite bleeding edge Aug 23 08:45:53 i'll do a feeds update, just to be sure Aug 23 08:51:22 could it be possible that the opkg feeds on build.shr-project.org are not up to date? Aug 23 09:01:36 farhaven: ask dos1 :) Aug 23 09:01:51 hehe Aug 23 09:01:58 farhaven: can't you determine the build date of the packages you have? Aug 23 09:02:09 i can determine the git revision, i think Aug 23 09:02:57 hmm, fosusaged has gitr394+4893e3cfa02722b0a082c1c9d03bba7e30193fea-r6.0 and libfsobasics has gitr394+4893e3cfa02722b0a082c1c9d03bba7e30193fea-r6.1 Aug 23 09:03:06 is the 6.0 versus 6.1 relevant? Aug 23 09:04:48 farhaven: hm, looks like both are the latest... Probably mickey|zzZZzz can have an idea what could be wrong, try to show him your backtrace... Aug 23 09:05:10 once he's awake :D Aug 23 09:17:01 hi Aug 23 09:17:17 is anybody aware of the location of python elementary test apps? Aug 23 09:18:28 khiraly1: in svn repo of python-elementary Aug 23 09:19:00 in enlightenment's svn? Aug 23 09:19:48 here?: http://trac.enlightenment.org/e/browser/trunk/BINDINGS/python/python-elementary Aug 23 09:22:18 khiraly1: yeap in tests Aug 23 09:23:56 hm gsm is dead with latest -unstable ? Aug 23 09:24:12 yes Aug 23 09:24:27 soltys: do you know how is it possible to change bit depth runtime? (X11 vs X11_16) Aug 23 09:24:29 it's fsousaged segfaulting Aug 23 09:24:41 khiraly1: shr-settings can do it Aug 23 09:26:24 hm gsm is down but wifi works... Aug 23 09:26:26 ;P Aug 23 09:26:37 hehe, voip for the win :D Aug 23 09:38:19 blah ;P Aug 23 09:39:06 crap, everything is going down Aug 23 09:39:14 first the phone stops working, now quake live is fubar Aug 23 09:40:29 khiraly1: do you probably know why nytowl doesn't want to unite efforts with SHR? Aug 23 09:41:17 PaulFertser: no idea at all Aug 23 09:41:56 (11.24.41) farhaven: khiraly1: shr-settings can do it <-- it messes with shell variables (through /etc/profile.d/elementary.sh) Aug 23 09:42:29 well, if you want to change the rendering engine at runtime, the wrench menu can do that Aug 23 09:42:40 though i don't know whether that changes all aspects Aug 23 09:42:45 farhaven: it only changes for illume itself Aug 23 09:42:59 not subsequent application Aug 23 09:43:23 khiraly1: i could never understand why you guys are pushing your own "distro" om2009 instead of maintaining the distro parts of SHR testing. Aug 23 09:44:26 so if you do an evas only application (no elementary), you choose the engine at startup, and the wrench does not change the color depth for you. Not sure about elementary apps Aug 23 09:45:00 hmm, i want portage on my phone Aug 23 09:45:11 that would make installing older versions of fsousage easier Aug 23 09:45:11 PaulFertser: current om2009 is on an older kernel (that means: no windows networking (forgotten patch), and hardly working wifi (I never succeed)). Aug 23 09:45:58 Thats why Angus is working on 2.6.30 kernel integration. (suspend does not work currently, seems like an fso bug, and MMC card is not recognised sometime) Aug 23 09:46:12 however I didnt see any effort from shr to migrate over newer kernel Aug 23 09:46:31 khiraly1: (new kernel) right, just because it doesn't make much sense. Aug 23 09:46:41 khiraly1: we're all waiting for 2.6.31 mainline release. Aug 23 09:47:04 khiraly1: that'll be base for all our patches, mainly those larsc prepared. Aug 23 09:49:05 in my dreams, there would be a core distro, really shr agnostic. And every distro would build up from this. (there are some other effort other than shr, there is some qt based one) Aug 23 09:49:51 khiraly1: nothing in shr is really shr-specific, i can't see why can't shr be used as a base for any OE distro for gta0x devices. Aug 23 09:50:14 there are some messing here and there. Mostly namings. But its only my point of view Aug 23 09:50:25 (sound scenarios for example) Aug 23 09:51:00 khiraly1: well, that's mostly non-issues, can't be easily fixed after some minor discussion with interested parties. Aug 23 09:51:29 s/can't/can/ Aug 23 09:51:29 PaulFertser meant: khiraly1: well, that's mostly non-issues, can be easily fixed after some minor discussion with interested parties. Aug 23 09:51:40 but I really cant speak of Angus, I only work on paroli not on om2009 itself. As far as Im concerned, if paroli would be installable on shr, I would love to test it on shr too. Aug 23 09:52:39 khiraly1: do you percieve this channel (#openmoko-cdevel) as somewhat unfriendly towards non-shr projects? Aug 23 09:53:19 and I never talked with Angus about this. So dunno what is his ideas. I only fell in love with paroli, so it was fairly obvious for me, that I use om2009 (no other distro provide paroli (by default)) Aug 23 09:55:11 but suspend-resume is superior experience on om2009, and I dont see any advantage of shr's kernel. But if 2.6.31 will be used on both distro and paroli could be installable on shr, for me it does not make much difference;) Aug 23 09:55:42 khiraly1: what's better about suspend/resume on om2009? Aug 23 09:56:02 a bit faster, and no output on console screen. Aug 23 09:56:26 khiraly1: shr is no different Aug 23 09:56:45 khiraly1: output on console screen depends on bootloader passing kernel proper command line. Aug 23 09:57:08 khiraly1: or one can change that runtime with dmesg -n X or by other means. Aug 23 10:00:04 I use almost two months old om2009 as my daily phone, and I develop paroli on top of it. It uses 2.6.28 (define itself as 2.6.29), so Im affraid of upgrading of newer testing image (which use this very kernel). And 2.6.30 does not work (suspend, sometime MMC card initialization). Shr is no different it use a very old kernel too (2.6.29)... Aug 23 10:00:14 morning Aug 23 10:02:36 khiraly1: shr uses latest andy-tracking which is based on .29, that's true. And doesn't include latest-and-greatest glamo patches. But why hurry, we'll just wait for .31 which will include much of gta02 support o-o-b and we'll apply a relatively small set of patches on top of it targeting upstream submission. Aug 23 10:03:20 But why hurry <-- because fso will break on it, and would be nice to know this things in advance Aug 23 10:03:43 khiraly1: fso breakage is minimal and easily fixable. Aug 23 10:04:29 anyway, as far as I know, I dont expect Angus to come to shr folks and ask for a merge/naming. The other way around will probably work Aug 23 10:05:23 but Im not competent of speak about it. I can only speak of paroli (if it interests anybody;), if shr will provide paroli, Im happy to test on shr too Aug 23 10:06:57 khiraly1: Angus is one man, SHR is joint community effort, i can't see why you say that :S Aug 23 10:07:51 khiraly1: you know there was time when paroli worked out-of-the-box on SHR? Yes, that's true. Aug 23 10:08:12 khiraly1: then something's changed and it broke. Ainulindale asked paroli guys to fix that but got no reply. Aug 23 10:10:33 dunno about it. I only mess with paroli since I got my freerunner (june? may?). If it happened before I have no idea. But if it happens now, Im ready to collaborate and fix those issues Aug 23 10:12:21 khiraly1: are you ready to collaborate right now? To check if paroli recipe is buildable, to see if it work etc? Obviously most SHR guys are not _that_ interested to do it themselves but i'm sure they're not going to build any obstacles and they're pretty responsive so you can get something working pretty fast. Aug 23 10:14:43 touche Aug 23 10:14:50 an url of bb recipe? Aug 23 10:15:01 khiraly1: sure Aug 23 10:17:27 khiraly1: (E lib names change) btw, at the time of the change SHR guys were in close communication with raster so in my understanding they did exactly what upstream suggested. Aug 23 10:19:00 khiraly1: i guess SHR uses this one directly: http://cgit.openembedded.net/cgit.cgi/openembedded/tree/recipes/openmoko-projects/paroli_git.bb Aug 23 10:19:59 khiraly1: i'm afraid to actually help with getting paroli in some usable state into SHR you'll need to install the full building environment to be able to test changes and resulting images/packages. It's supposed to be easy enough, just follow http://shr-project.org/trac/wiki/Building%20SHR Aug 23 10:20:45 paroli only uses python-elementary if Im right (and evas, but its nowhere really used), so dunno what could have been E lib names change Aug 23 10:21:37 khiraly1: the E lib names change hit the guys that package apps themselves (such as intone). Aug 23 10:36:44 SHR: 03seba.dos1 07opimd-utils * r4a9ddff8c0cc 10/opimd-notes: opimd-notes: implement searching Aug 23 10:36:45 SHR: 03seba.dos1 07opimd-utils * rd88605604b20 10/ (TODO opimd-notes): opimd-notes: temporialy disable anchors in tags Aug 23 10:36:46 SHR: 03seba.dos1 07opimd-utils * r4597722b9e25 10/opimd-notes: opimd-notes: popup keyboard when searching Aug 23 10:36:46 SHR: 03seba.dos1 07opimd-utils * r793ad088907c 10/opimd-notes: opimd-notes: fix traceback when editing note Aug 23 10:36:48 SHR: 03seba.dos1 07opimd-utils * r2321a603d79e 10/opimd-notes: opimd-notes: few interface changes Aug 23 10:36:56 SHR: 03seba.dos1 07opimd-utils * r1fccea7852e1 10/opimd-notes: opimd-notes: cosmetic changes to topbar Aug 23 10:38:10 dos1: hi Aug 23 10:38:23 dos1: i've got one nice suggestion from an "end-user" Aug 23 10:39:32 dos1: can you probably add a note to shr settings stating that "power settings are not persistent" (on the power settings page of course). And the same for backlight and blanking interval. Probably mention editing rules.yaml or something... What do you think about it? Aug 23 10:40:10 well, adding such notes will be close to making mess ;x Aug 23 10:40:17 there is user manual Aug 23 10:41:13 dos1: just a small note like "changes are not persistent" Aug 23 10:42:58 dos1: or just save that settings and restore them after reboot ;) Aug 23 10:43:05 soltys: no Aug 23 10:43:20 khiraly1: FYI Aug 23 10:43:23 why not ? Aug 23 10:43:24 damnit, i want a working phone :/ Aug 23 10:43:29 Angus & I collaborated a lot for our OE branches Aug 23 10:43:34 He stole some of our stuff, we stole his Aug 23 10:43:42 PaulFertser: this building looks really complicated. Some days of work... Aug 23 10:43:48 The problem wasn't on packaging but on software with compatible versions IIRC Aug 23 10:43:57 khiraly1: it's only work for you machine, not for your brain. Aug 23 10:44:57 Ainulindale: and migrating over 2.6.30 (or 2.6.31) does help on compatible versions? Aug 23 10:45:22 really om2009 should turn into shr testing or shr stable. Aug 23 10:45:28 khiraly1: (kernel) irrelevant Aug 23 10:45:46 khiraly1: 2.6.30 of what? kernel? no link Aug 23 10:45:57 yepp kernel Aug 23 10:46:07 khiraly1: IIRC they developed their eleme,ntary bindings in parallel to the mainline Aug 23 10:46:24 That was problematic too as we're using them and they weren't stable enough Aug 23 10:47:28 only paroli uses elementary if Im right. And that is fixable if there is some incompatibilities Aug 23 10:49:29 khiraly1: well IIRC it was fixed later they rejoined the mainline Aug 23 10:49:35 I can't recall what was latest problem Aug 23 10:49:39 I'm triggering a build to see Aug 23 10:50:05 khiraly1: One thing I want you to be sure about Aug 23 10:50:16 There's no ill feeling between paroli, SHR, or whatever for that matter Aug 23 10:50:25 vegyraupe even lent me some cigarettes on our conference that's a sign :-) Aug 23 10:50:58 hm is there a way to downgrade packages ? because -force-downgrade don't work Aug 23 10:51:04 As of today we focus on our stack and that's all :-) Aug 23 10:51:11 As far as paroli is concerned I can give you the paroli dir->system dir relationship where paroli is installed (or expected to be installed) system wise. But what PaulFertser suggested seems a lot of work (for me) at once Aug 23 10:51:21 If there is a strong demand for Paroli on SHR, even paroli replacing our stack, I'm not bounded to anythin Aug 23 10:51:24 +g Aug 23 10:51:32 khiraly1: that is? Aug 23 10:51:56 as installing? paroli does not have any (big) requirements. there are some .edj building Aug 23 10:52:15 that wasn't my question Aug 23 10:52:19 what was his suggestion? Aug 23 10:52:24 apart from the dumb choice of debian? :-) Aug 23 10:52:47 PaulFertser: by the way in bearstech we trimmed down the h:1 image to ~70M IIRC Aug 23 10:52:53 Ainulindale: i suggested him to build SHR and check what's needed to get paroli running. Aug 23 10:53:04 Ah Aug 23 10:53:05 Easy to do Aug 23 10:53:08 I should build and test paroli on shr, and contribute the newest bb recipe Aug 23 10:53:17 khiraly1: well the "newest" bb recipe should be on OE Aug 23 10:53:24 It shouldn't differ that much from the old one IMHO Aug 23 10:53:31 Ainulindale: yes, i gave him url to OE cgit. Aug 23 10:53:34 THe problem is related to the dependencies Aug 23 10:53:44 Well khiraly1 Aug 23 10:53:58 I strongly suggest you to try & build the image using our makefile Aug 23 10:54:07 Then I'm all for trying to build a -paroli version of the image Aug 23 10:54:20 As long as you document your stuff :-p Aug 23 10:54:41 Hmm, maybe I simply download and install shr, and run paroli (without installing), thats the way how I develop it anyway Aug 23 10:54:59 if it does not start, there are some more issues other than packaging Aug 23 10:55:02 Well yeah try that Aug 23 10:55:05 And give me the logs Aug 23 10:55:06 Ainulindale: any idea why fsousaged segfaults ? Aug 23 10:55:12 If it doesn't work properly Aug 23 10:55:15 soltys: none Aug 23 10:55:28 Although if dos1 synced the feed Aug 23 10:55:31 And if the feed was borked Aug 23 10:55:38 due to the former bug we had Aug 23 10:55:45 I'd bet it's the same Aug 23 10:56:04 (bug in libfsosomething) Aug 23 10:56:28 libfsobasics0 afaik Aug 23 10:56:37 could be yes Aug 23 10:56:39 Ainulindale: it was fixed Aug 23 10:56:43 Ok :-) Aug 23 10:56:46 Did you sync the feed? Aug 23 10:57:02 yup Aug 23 10:57:10 Ainulindale: btw, is H:1 still using the old libgsmd etc? What are the projects goals atm? Aug 23 10:57:10 Did the produced image worked properly? Aug 23 10:57:12 so it should work after upgrading now? Aug 23 10:57:17 PaulFertser: I started a blog Aug 23 10:57:26 PaulFertser: we'll be switching to FSO after rev5 (mid september) Aug 23 10:57:28 It's already in devel Aug 23 10:57:36 PaulFertser: blog.hackable1.org Aug 23 10:57:49 Ainulindale: thanks for the link! Aug 23 10:57:53 PaulFertser: you're welcome Aug 23 10:57:57 (it's on planet though) Aug 23 10:58:05 * PaulFertser doesn't read planet Aug 23 10:58:17 http://tinderbox.openembedded.net/public/logs/task/2564177.txt <= khiraly1 Aug 23 10:58:31 edj problem Aug 23 10:58:38 shouldn't be that hard to fix Aug 23 10:59:19 so dos1 Aug 23 10:59:21 did the image work? Aug 23 10:59:57 Ainulindale: data/paroli.edj Aug 23 11:00:06 that is the location in paroli git Aug 23 11:00:26 /data/e-config/paroli.edj -> /data/paroli.edj Aug 23 11:00:40 Ainulindale: which image? Aug 23 11:00:46 dos1: the image you generated Aug 23 11:00:52 Ainulindale: i didn't generate image Aug 23 11:00:55 Ainulindale: gconf? Oh. How is it any better than windows registry? :S Aug 23 11:00:58 dos1: ah! Aug 23 11:00:58 Ainulindale: not yet Aug 23 11:01:07 dos1: anyway folks complain that fsousaged segfaults for them Aug 23 11:01:10 PaulFertser: well valid point, could you comment on it? :-) Aug 23 11:01:13 dos1: http://farhaven.pastebin.com/d28b593ee Aug 23 11:01:31 Ainulindale: about why windows-registry style settings management is bad? Aug 23 11:01:45 khiraly1: I don't even have the edj file anywhere Aug 23 11:01:54 PaulFertser: about why gconf is Aug 23 11:01:57 PaulFertser: when that segfault happens? Aug 23 11:02:07 jbl chose to do that, he's sensible, he'll listen Aug 23 11:02:10 Ainulindale: http://git.paroli-project.org/?p=paroli.git;a=tree;f=data;h=e3d872859094f42dcdf2ae4a7e1513659ce70880;hb=HEAD Aug 23 11:02:13 dos1: if one tries to start fsousaged from command line. Aug 23 11:02:17 there is a paroli.edj Aug 23 11:02:25 khiraly1: well that must mean that we set an old rev Aug 23 11:02:56 mickey|zzZZzz ;x Aug 23 11:03:03 why can't he be here when he's needed? ;P Aug 23 11:03:12 hmmm weird no Aug 23 11:03:31 ah yeah Aug 23 11:03:37 sane-srcrevs of paroli is outdated Aug 23 11:04:31 Just switched it to autorev as a test Aug 23 11:04:40 Although it'll probably fail due to the same reasons Aug 23 11:05:01 Oh! Aug 23 11:05:03 About the logger problem Aug 23 11:05:08 I know where it comes from Aug 23 11:05:13 Deubeuliou hit the problem too Aug 23 11:05:13 ! Aug 23 11:05:18 do tell :) Aug 23 11:05:20 It's on blog.hackable1.org Aug 23 11:05:32 Ainulindale: well, i don't know much about gconf vs windows registry differences but i can't see why a way like that could be beneficial. You loose quite many essential things comparing to plain text files configs: you're not able to use a sane text-editor or sed to do amendments, you're not able to store comments or alternative settings in a sane way, you're not able to conveniently diff between updates etc... I can't see any benefits in ... Aug 23 11:05:38 ... registry-like settings except that it makes changing them from GUI easier. But since GUI way of changing settings is inherently limited i can't see that as a major issue. Aug 23 11:06:40 Ainulindale: also i see little need for an app controlling "power" state of various subsystems since that way of doing it is deprecated anyway. One should either use FSO directly to request resources on demand or use an fsoraw wrapper, at least that's what seems to be sane to me. Aug 23 11:06:59 PaulFertser: well that was done considering non FSO way Aug 23 11:06:59 farhaven: http://blog.hackable1.org/2009/08/run-ogsmd-on-hackable1.html#more Aug 23 11:07:06 thx Aug 23 11:07:15 farhaven: /etc/abyss.conf Aug 23 11:07:15 So it's temporary Aug 23 11:07:26 great, thanks a lot Aug 23 11:09:59 Ainulindale: and btw i don't know about many hackers who love gnome on desktop so i don't see a reason why they would be happy to use it on a smartphone. Esp. since it will be fso-based anyway. What will make H:1 different? PIM with Evolution instead of opimd, is it a really good idea? Aug 23 11:13:32 PaulFertser: nope Aug 23 11:13:40 Really different : gnome mobile & debian Aug 23 11:13:41 among others Aug 23 11:13:49 That and a kick ass leader Aug 23 11:14:54 Ainulindale: unfortunately i couldn't see what makes it really different and kicking ass from the blog. Aug 23 11:15:14 agreed Aug 23 11:15:22 we're in a setup time these days Aug 23 11:15:28 I'm setting up my position Aug 23 11:15:32 And we're fixing bugs Aug 23 11:15:41 After that we'll think about truly different features Aug 23 11:16:55 Ainulindale: probably small write-up explaining how and why are you going to kick ass would be nice to have. Aug 23 11:17:31 PaulFertser: ah by the way Aug 23 11:17:51 One can only look forward to a Manowar concert if and only if he's into dwarfs in leather pants Aug 23 11:18:38 rofl Aug 23 11:18:57 That's my opinion though :-) Aug 23 11:20:16 PaulFertser: fsousaged still segfaults, even with /etc/abyss.conf in place Aug 23 11:20:56 farhaven: you should ask to Deubeuliou Aug 23 11:22:04 Ainulindale: i think that's that mickey|zzZZzz's constructors stuff Aug 23 11:22:09 hm on ml someone is saying that with frameworkd.conf from 18.08 gsm works but suspend is only via 'apm -s' Aug 23 11:22:11 ;P Aug 23 11:22:30 soltys: seems reasonable Aug 23 11:22:41 Deubeuliou: help! :) Aug 23 11:23:01 dos1: why so ? Aug 23 11:23:51 mickey|zzZZzz, do you want to be called "Turbo Breaking Man"? :P Aug 23 11:24:12 "Dr." Turbo breaking man Aug 23 11:24:57 ;) Aug 23 11:25:31 soltys: it segfaults on initializing lowlevel plugins Aug 23 11:25:42 soltys: with old config lowlevel plugins aren't initialized Aug 23 11:25:47 hi Aug 23 11:25:57 soltys: but on suspend fsousaged wants to use lowlevel plugin Aug 23 11:26:07 soltys: and as it's uninitialized, it segfaults :P Aug 23 11:26:11 so to conclude: Aug 23 11:26:17 WE'RE DOOMED!!! Aug 23 11:26:19 ;) Aug 23 11:26:21 hehe Aug 23 11:26:49 dos1: and that is your fault ;/ Aug 23 11:26:50 very good timing :) Aug 23 11:27:28 soltys: partly mickey|zzZZzz's, partly my ;x Aug 23 11:27:36 ;) Aug 23 11:27:40 *sigh* Aug 23 11:28:03 * DocScrutinizer searching for the whip Aug 23 11:28:48 oh shit Aug 23 11:28:52 i need working phone today ;x Aug 23 11:29:11 uh, i'll back to ousaged temporialy :P Aug 23 11:29:27 how would one do that? Aug 23 11:29:28 farhaven: what kind of help do you need N Aug 23 11:29:38 i need help getting fsousaged to not crash Aug 23 11:29:45 the stuff on the h:1 blog didn't really help Aug 23 11:29:58 * Wonka hands dos1 his old Motorola M3888 Aug 23 11:30:10 farhaven: [fsousaged] disable = 1, [ousaged] disable = 0 Aug 23 11:30:11 here. works. all the time. with 4x AA cells if necessary. Aug 23 11:30:18 dos1: thx Aug 23 11:30:22 Wonka: does it openBmap logging? ;) Aug 23 11:30:37 no. it phones. but that reliably. Aug 23 11:30:48 :p Aug 23 11:30:53 i have nokia 6230i somewhere Aug 23 11:31:05 «dos1> i need working phone today ;x» Aug 23 11:31:23 Wonka: yup, i need Aug 23 11:31:26 no word about GPS ;) Aug 23 11:31:33 ;] Aug 23 11:31:36 Wonka: but it would be nice to have also gps ;) Aug 23 11:31:46 i'll switch to ousaged for today Aug 23 11:31:59 when mickey|zzZZzz will return, please bite him, ok? :P Aug 23 11:32:04 * Wonka wants a N900 Aug 23 11:32:38 btw Aug 23 11:32:52 Wonka: wtf FOUR times AA??? 3 should be enough, definitely Aug 23 11:33:21 Wonka: aah sorry. nevermind Aug 23 11:33:33 farhaven: I guess so. The segfaults is in libgsm0710d, which, I guess, is not usesd by ousaged Aug 23 11:33:37 DocScrutinizer: that's what it takes. two rows of two AA cells. original NiMH battery is that format too. Aug 23 11:33:51 hmm Aug 23 11:33:59 dos1, my neo broken! it doesn't work after yesterday update Aug 23 11:34:00 * DocScrutinizer should stop habit to read backscroll bottom to top :-P Aug 23 11:34:05 have anybody using his terminal (on desktop), what makes the same completion as ipython doeS? (you type the beginning of the command, and it look up in the history all the command which begins the typed characters) Aug 23 11:34:05 is there an alternative to that lib? Aug 23 11:34:21 khiraly1: you mean reverse-i-search? Aug 23 11:34:31 dos1, what is current status, do you know reason? Aug 23 11:34:38 max_posedon: ... Aug 23 11:34:44 max_posedon: perfect timing ;P Aug 23 11:35:00 max_posedon: we're discussing it at the moment, for about 15 minutes Aug 23 11:35:01 farhaven: hmm, yeah, thats about it, just it needs an extra kombo. But thanks. Aug 23 11:35:12 ^r in bash Aug 23 11:35:59 so, today neo unusable?, can't you fix it Aug 23 11:36:15 ousaged looks good Aug 23 11:36:16 farhaven: hmm, it is not that usable. It does not cycle through the history, and pron to mess up Aug 23 11:36:26 cycle? Aug 23 11:36:40 you can press ^r repeatedly to cycle through history Aug 23 11:37:04 ehh. how counterintuitive is it Aug 23 11:37:11 i like it :) Aug 23 11:37:13 farhaven: the segfaults in libgsm0710muxd has been fixed in the latest git rev Aug 23 11:37:17 farhaven: try ipython once, and see the difference Aug 23 11:37:19 k Aug 23 11:37:23 max_posedon: just use ousaged Aug 23 11:37:31 farhaven: but thanks. Aug 23 11:37:31 Deubeuliou: any prediction on when it will hit the opkg feeds? Aug 23 11:37:57 farhaven, Deubeuliou: you're wrong! Aug 23 11:37:58 I only need to have an ability to save the opened tabs with history in each. Like firefox do. (Im using gnome-terminal) Aug 23 11:38:11 dos1: hmm? Aug 23 11:38:11 farhaven, Deubeuliou: fso-abyss segfault != fsousaged segfault Aug 23 11:38:22 fso-abyss segfault is already fixed Aug 23 11:38:29 k Aug 23 11:38:47 fsousaged segfault was introduces about 15 hours ago in feeds, and you can workaround it by using ousaged temporialy Aug 23 11:38:59 dos1: I know Aug 23 11:39:03 dos1, ok, fine, how i should switch? Aug 23 11:39:05 13:33 < Deubeuliou> farhaven: I guess so. The segfaults is in libgsm0710d, which, I guess, is not usesd by ousaged Aug 23 11:39:06 dos1: yup, doing that atm Aug 23 11:39:21 max_posedon: it was also said here, few minutes ago! Aug 23 11:39:29 whee Aug 23 11:39:32 phone works :D Aug 23 11:39:42 ok, confirmed, ousaged solution works Aug 23 11:39:59 but keep in mind ousaged is buggy regarding RequestResource Aug 23 11:40:41 oh, sorry, I just woke up... Aug 23 11:41:03 s/Resource/Resource done by oeventsd/ Aug 23 11:41:03 dos1 meant: but keep in mind ousaged is buggy regarding RequestResource done by oeventsd Aug 23 11:41:28 hehe, being a bit buggy > crashing Aug 23 11:43:49 freesmartphone.org: 03mickey 07cornucopia * ref66ce26b4f5 10/fsodeviced/src/plugins/kernel26_rtc/plugin.vala: fsodeviced: kernel26_rtc: add logging output for set and get methods Aug 23 11:43:49 freesmartphone.org: 03mickey 07cornucopia * re85a17c5e8b3 10/libfsoframework/fsoframework/object.vala: Aug 23 11:43:49 freesmartphone.org: libfsoframework: Vala objects created w/ Object.new() do NOT (I repeat _not_) Aug 23 11:43:50 freesmartphone.org: call any default constructor (like Foo()), but just the construct {} constructor. Aug 23 11:44:35 morning folks Aug 23 11:44:39 hi Aug 23 11:44:40 morning Aug 23 11:44:52 dos1, if latest fsousaged fixed, may be you can just update packages in feed may be? Aug 23 11:44:58 is that commit to fix fso another time? ;-) Aug 23 11:45:23 max_posedon: we already have latest fsousaged (except those two commits above :P) Aug 23 11:45:27 * soltys wondering if that commits will stop fsousaged from segfaulting... Aug 23 11:45:48 mickeyl: so? ;) Aug 23 11:45:52 yep Aug 23 11:45:57 soltys: mickeyl: [13:44] is that commit to fix fso another time? ;-) Aug 23 11:46:07 aah Aug 23 11:46:10 ;) Aug 23 11:46:10 :-)) Aug 23 11:47:30 Ainulindale: what do you do on screen? can i build fsousaged and friends, and sync feeds? Aug 23 11:47:47 Ainulindale: paroli looks like built ;) Aug 23 11:48:42 i'll cut the first set of releases for the cornucopia stack next month. from then on i advise using release tarballs on the stable/testing distro and only using git versions on the unstable one Aug 23 11:48:52 NIICE. elementary test app crashed enlightenment;) Aug 23 11:51:49 mickeyl: valaterm has flaws on redrawing screen correctly on get_focus with busybox (adds a few annoying newlines on bottom effectively blanking screen). With bash lineedit aka cursormovement on multiline input is borked (home off a number of chars, linewraps don't get skipped correctly). Also there's a report valaterm stops bash from sourcing .profile Aug 23 11:52:37 mickeyl: any chances we see a fix for all that? Aug 23 11:53:07 dos1: I didn't Aug 23 11:53:10 hmpf Aug 23 11:53:13 unclear. could check whether there's any more we can tweak in libvte, but apart from the last thing everything sounds like VTE issues Aug 23 11:53:53 perhaps we're just configuring it wrong Aug 23 11:54:10 do we have all that as bugs on bugzilla? Aug 23 11:54:17 let me check whether there's a component for vala terminal Aug 23 11:54:26 s/bugzilla/trac/ Aug 23 11:54:39 not from me (trac) Aug 23 11:54:46 mickeyl: you too wished it was bugzilla? ;) Aug 23 11:54:55 heh, not really Aug 23 11:55:01 i'm fine with either Aug 23 11:55:05 but trac's so much easier to setup Aug 23 11:55:14 ok, now there is a component Aug 23 11:55:19 feel free to add bugs to it Aug 23 11:55:47 heh;) for bugzilla an entire life is not enough to learn it;) Aug 23 11:56:33 khiraly1: at least it'd be nice to use e-mail for commenting/altering tickets. Apparently trac fails to provide that :-/ Aug 23 11:56:44 khiraly1: apparently paroli git built Aug 23 11:56:52 Ainulindale: nice! Aug 23 11:56:56 so check with dos1 when the package will be available on the feed Aug 23 11:57:07 I'm off to play cards this afternoon Aug 23 11:57:22 khiraly1: it'll be really soon Aug 23 11:57:27 as fsousaged build is finishing Aug 23 11:58:34 dos1: and it will work ? Aug 23 11:58:37 ;P Aug 23 11:58:40 super. It was time to try out my newest 4G sd card (and abandone the default 512MB one) Aug 23 11:59:21 maybe I move from 512 to 2G uSD card... Aug 23 12:00:10 khiraly1: paroli synced Aug 23 12:00:15 others: fsousaged too Aug 23 12:00:19 upgrade :) Aug 23 12:01:38 thanks dos1! I will try it out a bit later (after I mess a bit more with elementary) Aug 23 12:01:54 you actually have to update libfsoframework to fix the problem in fsousaged Aug 23 12:02:14 and fsousage might have to be rebuilt after that Aug 23 12:03:04 dos1: testing;P Aug 23 12:03:59 aha! using x11_16 engine the Icon Transparent demo crashes enlightenment (as a whole) Aug 23 12:04:13 (using python-elementary demo suite) Aug 23 12:04:55 mickeyl: both upgraded and rebuilt Aug 23 12:04:58 ok, going out now Aug 23 12:05:02 byt Aug 23 12:05:05 bye* Aug 23 12:05:33 dos1, Ainulindale: thank for you support! Aug 23 12:06:56 khiraly1: x11_16 engine crashes even with some E settings menus. But otoh it's fast. Aug 23 12:07:34 ;) Aug 23 12:07:43 hm now I have gsm! :D Aug 23 12:10:17 E settings menus? You mean the wrench? Aug 23 12:10:26 khiraly1: yes Aug 23 12:12:39 hi, is this a kinf of föauncher or? http://scap.linuxtogo.org/files/f916ec6fca9bdd0655daf3645a8eb590.png Aug 23 12:19:25 Upgrading libfsoframework0 on root from 0.1.0.0+gitr394+4893e3cfa02722b0a082c1c9d03bba7e30193fea-r6.2 to 0.1.0.0+gitr396+e85a17c5e8b3e4bdc88aa5f56cb0e35c9d0c6387-r6.2... Aug 23 12:19:32 Not installing libfsoframework0 (0.1.0.0+gitr396+e85a17c5e8b3e4bdc88aa5f56cb0e35c9d0c6387-r6.2) on root -- already installed Aug 23 12:19:39 I think its strange Aug 23 12:33:02 max_posedon: --force-reinstall? Aug 23 12:33:22 max_posedon: though it's strange for sure Aug 23 12:49:10 hmm Aug 23 12:54:00 yo, hmm Aug 23 12:54:16 * DocScrutinizer ->brunch. bbl Aug 23 13:19:44 please can one of the kernel-savvy check if we switch SPI_MOSI1 and relatives to low or highZ when shutting down GSENSOR_3V3 Aug 23 13:22:25 and while you're on it, please check which supply voltage LDO1 actually is set to when we enable GSENSOR_3V3. Aug 23 13:25:18 Problems with Categories? I just two icons of Sudoko and Midori in SHR-u. If I remove GTK; from the midori.desktop Categories=GTK;Network; and change in Soduku from Categories=Applications;Game;ENLIGHTENMENT; to Categories=Game; It will only show one icon instead.. Aug 23 13:26:44 DocScrutinizer: is it not 3.3? Aug 23 13:28:05 larsc: please refer to recent mail on [HW] Aug 23 13:28:16 DocScrutinizer: hmmm mach-gta02 (andy-tracking has) .min_uV = 1300000, .max_uV = 1300000 for PCF50633_REGULATOR_LDO1 Aug 23 13:28:27 i think this is fishy isn't it? Aug 23 13:29:02 sorry, no time to check right now. but yes sounds strange Aug 23 13:29:48 bbl Aug 23 13:31:09 sleipnir: that's fixedup before initalising the regulator driver Aug 23 13:31:19 so, ich mach mich auf nach russland Aug 23 13:31:26 err, sorry, wrong chan Aug 23 13:31:47 larsc: yeah i see, but .max_uV is still 1300000, isn't is? Aug 23 13:32:07 larsc: we set min_uV two times... i think this is a typo Aug 23 13:33:28 yes it is Aug 23 13:33:45 should i send a patch? Aug 23 13:34:23 if you send a patch make 3.3 the default and only change it to 1.3 for rev1 Aug 23 13:35:07 but probably you could drop support for rev1 completly Aug 23 13:35:37 larsc: i thought about killing the mangle function completely as rev1 should be killed anyway Aug 23 13:35:45 exactly Aug 23 13:35:55 yup Aug 23 13:38:45 i'm on it Aug 23 13:42:42 hmmm the "om-2.6.31-rc5" branch already has the correct code Aug 23 13:43:12 so what should i do? fix it for andy-tracking, leave everything as is? Aug 23 13:43:35 it would be really nice if someone could cleanup the git mess Aug 23 13:53:52 using shr-u all phone apps segfault, when run from terminal i get "Failed to handle dbus error: Resource Device is not enabled" any ideas? Aug 23 13:56:57 arg0_, you should update Aug 23 13:57:03 bug was fixed 1 hour ago Aug 23 13:57:52 larsc: all new branches have the correct setting. what should i do? Aug 23 13:58:32 larsc: new = *-2.6.31-rc5 Aug 23 13:59:12 just did the upgrade, any ideas? Aug 23 14:00:11 hi Aug 23 14:00:56 with opimd-notes I always get an " ValueError: Unable to guess signature from an empty dict" error.. how can I resolve this issue? Aug 23 14:18:46 hmm Aug 23 14:24:44 sleipnir: andy-tracking will be killed as soon as .31 is out upstream. Aug 23 14:27:04 http://pastebin.com/m44219445 Aug 23 14:27:19 could somebody open a python shell, and type the above 5 commands? Aug 23 14:27:43 Im interested if on shr evas_get() returns with None too. Aug 23 14:29:50 PaulFertser, does it mean, that .31 have all most important neo patches? Aug 23 14:30:05 not just `base platform support` Aug 23 14:30:48 max_posedon: no, it means that it'll be the base for the further work. Aug 23 14:40:44 mickey|bbl: I know it's a kernel issue first instance, but fso might request it "officially": we have no means yet to configure power management & IO for peripherals during suspend. I.e: is GPS during suspend: a)on, b)off, or c)in low_power (aka sleepmode)? how can I keep an LED shining during suspend? Are gmeters during suspend a)on, b)off, c)on and IRQ resume-reason enabled?. Same for all peripherals... Aug 23 14:41:52 mickey|bbl: I think fso resource management has to learn about that, and kernel has to support a more detailed means to configure every single powerline / IO for suspend Aug 23 14:42:33 or did I simply miserably miss anything basic here? Aug 23 14:46:03 DocScrutinizer: i'm not sure led can be configured to shine during suspend. Aug 23 14:46:23 dos1|away: ping Aug 23 14:46:47 PaulFertser: atm it can't as kernel does a hard switchdown of IO. hw is well capable of handling that though Aug 23 14:47:10 that's all my point Aug 23 14:48:10 aiui there's hardcoded assumptions in kernel about what suspend state means wrt peripherals. that's utterly wrong Aug 23 14:50:35 I don't say I wrapped my head around the whole sw side of this issue. It's just what I gathered for sparse info about that topic up til now Aug 23 14:52:52 DocScrutinizer: i thought that in suspend SoC GPIO is either high or low but not blinking. Aug 23 14:54:29 PaulFertser: see [HW] and LDO1 for instance. If we want gmeter to trigger resume we want to keep it powered. If we don't want that, we need to highyZ the MOSI and CLK otherwise we might see the "switching off causes more power consumption thru pulled down IO than powered up" issue Aug 23 14:55:32 DocScrutinizer: i don't say your point as a whole is wrong, i'm just saying that i don't know if blinking leds in suspended state is possible. Aug 23 14:55:33 PaulFertser: (LED blink) true, that's what I've been told as well. Nevertheless a 100% on should always be possible Aug 23 14:56:31 PaulFertser: how could it be possible? Aug 23 14:56:47 PaulFertser: I always assumed the hw-pwm could do that dimming / blinking during suspend as well. I stand corrected we allegedly have no clock for it Aug 23 14:58:11 PaulFertser: btw that's one of the reasons why I asked you yesterday if you know how PWM is implemented (for vibrator) - via hw-countor or via software and IRQ Aug 23 15:04:48 wpwrak: could you shed a little bit of light on that. It's been you who told me we can't use hw-PWM during suspend, and maybe you know if we use it at all? Aug 23 15:08:12 I lightly admit we might miss cpu-masterclock for pwm-counter to operate during suspend (and I wonder if that's a design flaw and we could have used any 32k RTC clock instead) Aug 23 15:14:20 DocScrutinizer: FIQ handler does it afaik Aug 23 15:14:47 lindi-: that's exactly what I expect. It's abysmal wrong Aug 23 15:15:57 another "hack" of our gifted kernel guys when they were without a clue how to put hw to decent operation ;-) Aug 23 15:16:30 at least that's what I suspect Aug 23 15:18:31 probably hw-pwm failed on "missing base-R" issue which lead to readback of IO always was 0 despite the output was set to 1. So they implemented that hack with "double-buffering" IO state to toggle it, in sw, with FIQ. Aug 23 15:19:01 I guess now we got those base-R we might put hw-pwm back to purpose Aug 23 15:21:14 or it turns out it's actually a flaw in chip design and never worked under no conditions. Then we need to clearly document it, so this topic is done once and for ever Aug 23 15:21:45 (with 2442b ;-) Aug 23 15:23:28 anybody checked yet if we actualy disable SPI_MOSI1 / SPI_CLK1 when shutting down LDO1 ? (gmeter issue) Aug 23 15:32:07 Not yet, sorry Aug 23 15:46:37 PaulFertser: please join [hw] thread Aug 23 15:46:55 I sent mail 1min ago Aug 23 15:47:25 bbl, cya Aug 23 16:04:55 DocScrutinizer: yes, you should be able to do hw-pwm, but only when not suspended. in suspend, the 2442 has no clock Aug 23 16:05:16 that's what I assumed above Aug 23 16:05:59 DocScrutinizer: also, one of the timers is used for fiqs (for hdq). that's why it's done through software Aug 23 16:06:21 I can't imagine Samsung didn't think about that. So I suspect EE did implement a flaw there (probably we had to route PMU CLK_32k to SoC for that) Aug 23 16:07:17 wpwrak: ooh, thas a valid excuse for *one* of the hw-counters then :-D Aug 23 16:07:29 (fiq) Aug 23 16:07:57 i don't know if this helps. also an external clock may be gated. the 2442 manual doesn't mention a scenario where you could have the timers operate in suspend Aug 23 16:08:22 hmmm :-/ Aug 23 16:09:34 anyway, thanks for explaining that far. Is it only the one pwm-output whose timer is needed for FIQ which doesn't run hw-pwn then? or does none of them hw-pwm? Aug 23 16:11:29 TCLK1 is available gta02 (nGSM_OC), so you could try to inject a signals and see what happens :) Aug 23 16:11:58 wpwrak: wat's your opinion to he general suspend&peripherals topic I raised with "" [2009-08-23 16:40:43] mickey|bbl: I know it's a kernel issue..." ? Aug 23 16:12:34 err yes, the suspend is generally also a kernel issue ;-) Aug 23 16:13:44 looool GSM_OC. My *special* friend >:-( Aug 23 16:14:08 now we see we have had a better purpose for that GPIO ;-) Aug 23 16:14:08 why now SetScenario(gsmhandset) is commented out from rules.yaml? Aug 23 16:14:59 Q-Master: that's done by dialer afaik Aug 23 16:15:41 anyway it's totaly misplaced in fso o(events/device/foo)d Aug 23 16:16:53 DocScrutinizer: hm. seems that it is done by the efl0 lib, efl2 doesn't know anything about it. so I don't here any sound after dialout Aug 23 16:17:06 Q-Master: as a general rule of thumb the ap that wants to do audio needs to care about correct scenario to be restored. No-one will do it on app's behalf Aug 23 16:17:53 ok Aug 23 16:18:12 Q-Master: so here the app is diealer Aug 23 16:18:32 humm a nice freudian Aug 23 16:18:35 anyway. seems that was not a dialer, but a lib Aug 23 16:19:18 I'm not maintainer neither of dialer nor of fso. Sorry you need ask someone else for more detail Aug 23 16:20:32 I just remember a dialog here in cdevel that I tried to replay Aug 23 16:20:47 thnx Aug 23 16:20:59 (except the rule of thumb which is a genuine jOERG statement ;-D ) Aug 23 16:30:00 Q-Master: to provide you with a second general rule: no app should ever use a hardcoded "default" for alsa-audio-device. The mandatory way is every app SHOULD have a confic option for name of audiodev. If app miises on that, then it MUST use a *unique* hardcoded audiodev name and ship with a matching .asoundrc amendment to map that unique devname to arbitrary audidevices Aug 23 16:37:17 [shr-u] basic question about building shr-u. Following instructions from http://trac.shr-project.org/trac/wiki/Getting%20started%20developing%20SHR including the line "require local-builds.inc" into shr-unstable/conf/local.conf produces an error as multiple .bb files are providing xserver. I've followed the other instructions on that page? Aug 23 16:39:03 DocScrutinizer: (SPI_MOSI1 / SPI_CLK1) the pins are set to 0 but the internal pullup is not enabled during suspend (if i read the kernel source correctly) Aug 23 16:41:08 any kernel dev here? Aug 23 16:49:09 * PaulFertser feels to be incapable of anything this weekend :( Aug 23 16:49:46 PaulFertser: maybe you can help me :-P Aug 23 16:50:36 if a pin of the s3c2442 should be highZ I need to call s3c2410_gpio_pullup!? Aug 23 16:50:58 shit Aug 23 16:51:08 hmmm? Aug 23 16:51:15 no channel from fso-abyss after update Aug 23 16:51:34 there was a solution for that on the h:1 blog Aug 23 16:53:17 wpwrak: ping Aug 23 16:53:58 is fsousaged working for you? :) Aug 23 16:54:35 dos1: seems that it works Aug 23 16:54:45 sleipnir: i think to make it high-z you change it to input. Aug 23 16:55:00 sleipnir: like this: s3c2410_gpio_cfgpin(S3C2410_GPH1, S3C2410_GPIO_INPUT); Aug 23 16:55:06 sleipnir: at least that's what Andy did. Aug 23 16:55:10 sleipnir: url? Aug 23 16:55:44 Q-Master: http://blog.hackable1.org/2009/08/run-ogsmd-on-hackable1.html Aug 23 16:58:42 sleipnir: that seems it's just like s3c24xx works: if the pin is output it might be pulled or pulled down. If the pin is input it can be either high-z or pulled-down (thats 2442b). Or it can be managed internally by being assigned to primary or secondary function. Aug 23 16:59:35 2.6.31-rc7 is out Aug 23 16:59:47 one or two rc, and the final 2.6.31 will be out Aug 23 17:00:09 PaulFertser: so to set it highz i need to s3c2410_gpio_setpin($pin, 0); s3c2410_gpio_cfgpin($pin, S3C2410_GPIO_INPUT); ? Aug 23 17:00:26 PaulFertser: sorry i'm a noob there :-S Aug 23 17:00:36 dos1: wtf is with abyss? Aug 23 17:01:29 Q-Master: works here Aug 23 17:01:37 sleipnir: btw, do you have gta02 yourself? Aug 23 17:01:42 Q-Master: check if you have running gsm0710muxd too, it shouldn't be running Aug 23 17:01:46 PaulFertser: yes Aug 23 17:01:53 fso-abyss: error while loading shared libraries: libfsotransport.so.0: cannot open shared object file: No such file or directory Aug 23 17:01:58 wtf? Aug 23 17:02:09 sleipnir: in case you don't know there's one handy utility by Werner, s3c24xx-gpio, comes preinstalled in SHR. Aug 23 17:02:23 sleipnir: allows to fiddle with all gpios from the userspace. Aug 23 17:02:43 PaulFertser: do you have a URL for the script? i'm using Gentoo on the phone Aug 23 17:03:00 Q-Master: then it's something bad with your instalation Aug 23 17:03:07 Q-Master: opkg install libfsotransport0? :P Aug 23 17:03:18 sleipnir: sure http://svn.openmoko.org/trunk/src/target/gpio/gpio.c Aug 23 17:03:20 dos1: it worked before opkg update/upgrade Aug 23 17:03:50 sleipnir: you're also interested in another handy utility: http://svn.openmoko.org/developers/werner/pmu/ Aug 23 17:03:53 max_posedon: about that "already installed" - no, it wasn't strange :P Aug 23 17:04:50 Q-Master: so? Aug 23 17:05:20 dos1: you've ruined everything! 8)) Aug 23 17:06:21 dos1, look at versions, still strange Aug 23 17:06:51 max_posedon: i still can't see anything strange Aug 23 17:07:02 max_posedon: one package wanted upgrade of libfsoframework0 Aug 23 17:07:06 max_posedon: it was upgraded Aug 23 17:07:18 max_posedon: and then other package wants to upgrade the same package Aug 23 17:07:24 max_posedon: so it displays "Not installing ..." Aug 23 17:07:30 dos1: does shr-utils already have a .desktop files? Aug 23 17:07:45 max_posedon: probably that "Not installing" msg doesn't have sense at all, other package managers just don't display anything like that Aug 23 17:07:58 Q-Master: ??? Aug 23 17:08:07 dos1: sorry, opimd-utils Aug 23 17:08:54 PaulFertser: thx for the links Aug 23 17:08:58 Q-Master: well, but still ??? :P Aug 23 17:09:09 Q-Master: there were desktop files for ages Aug 23 17:09:24 since beginning ;P Aug 23 17:09:30 dos1: too dumb today. found everything myself. thnx Aug 23 17:10:11 sleipnir: so i think to ensure the pin is high-z you need to "s3c2410_gpio_pullup($pin, 1); s3c2410_gpio_cfgpin($pin, S3C2410_GPIO_INPUT);" Aug 23 17:10:19 sleipnir: if you are going to write a patch keep in mind that s3c2410_gpio_{set,cfg}pin has been deprecated in favour of the generic gpio functions. Aug 23 17:10:23 max_posedon: but still, in opkg behaviour there is nothing strange for me Aug 23 17:10:54 dos1: btw, i've looked over frameworkd and understood that it will be very hard to make opimd normally understand >1 phone per person. 8( Aug 23 17:11:00 larsc: i would first try to fix it with the old functions and then rewrite the whole beast Aug 23 17:11:33 Q-Master: hmm? multiple phone numbers are possible even now Aug 23 17:11:39 PaulFertser: ok. i just wonder if there is a documentation somewhere... Aug 23 17:11:52 sleipnir: i guess that's SoC's datasheet Aug 23 17:11:52 dos1: have you tried to edit them? Aug 23 17:12:00 larsc: when will .31 be out? Aug 23 17:12:12 Q-Master: yes, it was unimplemented, but i fixed it yesterday Aug 23 17:12:26 it was just sqlite backend, not opimd Aug 23 17:13:05 dos1: I've looked through opimd src and understood that it will be rather hard to edit 1 number from a several ones. The best idea will be delete the contact from db and then recreate it, but that suxx. 8( Aug 23 17:13:26 so how it works in opimd-notes? ::P Aug 23 17:13:38 tags, the same architecture as for phones Aug 23 17:14:22 dos1: can you plz think something out to make opimd contacts be more vcard conforming? Aug 23 17:14:38 notes are notes, but contacts are contacts Aug 23 17:14:50 PaulFertser: whenever linus feels like it ;) or what was your question? Aug 23 17:14:56 editing 1 contact will be the pain Aug 23 17:14:58 * Q-Master off Aug 23 17:15:01 Q-Master: the same architecture is the same architecture Aug 23 17:15:16 PaulFertser: but i guess in 2-3 weeks Aug 23 17:15:24 he's still giving me shit something won't work, and then i'm implementing frontend and guess what? Aug 23 17:15:25 it works ;P Aug 23 17:15:45 larsc: ok, just what i wanted to know. I guess i will try to phycologically prepare for that to be able to contribute in some sane way. Aug 23 17:18:14 PaulFertser: p.17 of the 2442 manual says that GPG7 (which is ACCEL_CLK) is high-Z during sleep Aug 23 17:18:36 PaulFertser: so do we need to set it "by hand"? Aug 23 17:19:02 sleipnir: i need to read the manual really careful to say for sure. Aug 23 17:19:56 PaulFertser: hmmm Aug 23 17:20:55 freesmartphone.org: 03seba.dos1 07framework * r6d41a5427845 10/framework/subsystems/opimd/ (4 files): opimd: SQLite-*: fix traceback when editing Aug 23 17:25:55 nice boys ! :-))) Aug 23 17:26:15 I see you did weel while daddy was out Aug 23 17:26:20 well even Aug 23 17:26:32 :-P Aug 23 17:27:54 DocScrutinizer: hey, can you probably give me a hint how to prove that gta01 will be severely damaged if powered via USB without a battery? Aug 23 17:27:54 DocScrutinizer: ???? :-P Aug 23 17:28:34 PaulFertser: what do you need? a video showing it explode??? :-P Aug 23 17:28:52 DocScrutinizer: no, just some keywords so i can google a mail confirming the experience Aug 23 17:29:12 scream of death is the buzzword Aug 23 17:29:44 DocScrutinizer: cool Aug 23 17:31:13 yeah I'm the living encyclopedia of useless knowledhe ;-D Aug 23 17:33:06 is there _any_ single point in the kernel that shows how to make a pin high Z? even google finds nothing :-( Aug 23 17:33:41 PaulFertser: http://www.google.de/search?q=openmoko+%22scream+of+death%22&ie=UTF-8&oe=UTF-8 Aug 23 17:34:29 DocScrutinizer: i'm dumb but not _that_ dumb ;) Aug 23 17:34:44 sleipnir: set it to input and disable all pull-ups and -downs, i guess Aug 23 17:34:45 sleipnir: switch to input, switch off any pull-up/down resistors Aug 23 17:35:08 larsc: hey Aug 23 17:35:17 ok three time the same opinion is an empirical proof :-P Aug 23 17:35:29 PaulFertser: just for your convenience Aug 23 17:36:42 sleipnir: for the particular purpose to manage GMETER_3V3 powerdown it's sufficient to set the lines to 0 (output low level) Aug 23 17:36:45 DocScrutinizer: otoh i'm unable to find anything except "search archives for scream of death" and "be aware of scream of death, that's probably pcf50606 was destroyed but it was before my time". Aug 23 17:37:06 PaulFertser: hmmm Aug 23 17:37:09 mompl Aug 23 17:39:53 PaulFertser: Aug 23 17:39:56 For all I can imagine here's nothing else than the PMU PCF50606 that's going Aug 23 17:39:58 to fry. Battery isn't connected to anything else. PMU is hard to replace Aug 23 17:39:59 (vitualy impossible). Aug 23 17:40:01 All this only wild guess, not confirmed. Aug 23 17:40:02 /jOERG Aug 23 17:40:31 DocScrutinizer: that's what is done in the kernel: GPD12/13 (CS), GPG7 (CLK), GPG6 (MOSI) are set to 0 and GPG5 (MISO) is pulled down? Aug 23 17:40:52 that sounds ok then Aug 23 17:41:17 no way for any reverse-feed thru the signal lines Aug 23 17:41:26 DocScrutinizer: yes, i could only find that mail, nothing more relevant. Aug 23 17:41:47 so that 2V4 on GMETER_3V3 during suspend is even more mystery Aug 23 17:45:42 it's also quite interesting that the sensors work even if ldo0 is disabled Aug 23 17:46:45 PaulFertser: sorry I'm lost here. No idea how to find original quotes Aug 23 17:46:56 DocScrutinizer: np, i don't think it matters that much. Aug 23 17:47:15 one last idea... I'll scan the myriad of internal ML postings Aug 23 17:47:18 DocScrutinizer: esp. if the case wasn't investigated at all (since it'd be nice to employ some small hardware tweak to prevent damage). Aug 23 17:47:52 the tweak is pretty clear: add a capacitor in parallel to bat Aug 23 17:48:47 some 50.100uF or such Aug 23 17:53:04 DocScrutinizer: gta02 has only 10uF Aug 23 17:53:06 PaulFertser: hmm, not really. nevermind Aug 23 17:53:13 yeah Aug 23 17:53:17 mompl Aug 23 17:55:14 might be something around U1703 Aug 23 17:55:27 really strange Aug 23 17:56:53 C1701 loks fishy Aug 23 17:57:23 might create a phase-shift and feedback >1 Aug 23 17:57:37 on a certain level on gate Aug 23 17:58:36 hard to tell without lengthy test series in lab - probably kiling a few devices during that Aug 23 17:59:16 DocScrutinizer: i wondered about U1703 the first time i saw that but never actually got to try to understand. Aug 23 18:00:15 well it's PMU's exoscelleton robot booster Aug 23 18:01:14 DocScrutinizer: is it possible that if the GPIOs are not set to 0, the guy measures 2V4? Aug 23 18:01:30 anyway a C from source to gate looks strange here Aug 23 18:01:44 sleipnir: yes, absolutely Aug 23 18:02:28 DocScrutinizer: so i will ask him to post his kernel driver (he probably forgets/fails to set the gpios to 0 then) Aug 23 18:02:29 someone already tested opimd-notes? Aug 23 18:02:38 (keep in mind you have to have newest frameworkd) Aug 23 18:05:10 DocScrutinizer: if it's just a powerful MOSFET i can't understand why it is used that way. :( Aug 23 18:05:25 same here Aug 23 18:06:34 I kinda understand what it is supposed to do, but I miss the details on why it's done that way particularly Aug 23 18:07:12 probably it's the appnote of 50606, and they messed capacitance once again? ;-P Aug 23 18:08:50 oh, hardware discussions Aug 23 18:08:56 better to not interrupt ;) Aug 23 18:09:53 dos1: stay away! hot iron! Aug 23 18:10:08 :) Aug 23 18:10:17 DocScrutinizer: there's an application example on the 130 page of pcf50606 datasheet. I don't think it looks exactly like what is done in gta01. Aug 23 18:10:53 ok, so should I bother to look into that? Aug 23 18:11:05 DocScrutinizer: most probably no Aug 23 18:11:06 I've noted that whenever I upgrade frameworkd, /etc/frameworkd.conf has changed its ordering. Is there any way of keeping the same ordering in that file? It's quite a hassle to find out what's been changed. Aug 23 18:11:13 DocScrutinizer: just if you're really intrigued ;) Aug 23 18:11:25 not that much ;) Aug 23 18:11:56 I accept it's a genuine tpe invention and thus an immanent enigma Aug 23 18:12:52 DocScrutinizer: fyi gta01 has a 10uF parallel to the battery too. Aug 23 18:13:01 i know Aug 23 18:13:13 that's why I said "nevermind" Aug 23 18:14:09 lol Aug 23 18:14:38 if it was worth anything I'd recommend testing without that source-gate C Aug 23 18:15:13 or, even better, connect that to GND instead of source Aug 23 18:16:11 and probably it's not PMU that fries, but rather that FET Aug 23 18:16:42 I was obviously wrong on my former notion about that issue Aug 23 18:16:44 hmmmmm fries Aug 23 18:17:19 sybren: kill it, grill it Aug 23 18:19:26 DocScrutinizer: do you know anything about the frameworkd.conf file I asked about? Aug 23 18:19:42 nope, sorry Aug 23 18:20:00 DocScrutinizer: frying FET? How? Aug 23 18:20:34 mrmoku|away, ping Aug 23 18:20:48 F4t: you can ping, and ping, and ping... Aug 23 18:20:53 F4t: he's on long away Aug 23 18:20:58 overvolage on wild oscillation? overheating due to constant switching and dissipating power? Aug 23 18:20:59 F4t: on vacations ;) Aug 23 18:21:28 then who's gonna rebuild mokonnect? :0 Aug 23 18:21:37 or a full image for that matter Aug 23 18:24:02 F4t: i can do that Aug 23 18:24:07 F4t: i rebuilt mokonnect yesterdays evening Aug 23 18:24:18 F4t: do you want it again? :P Aug 23 18:24:20 and merge into unstable repo? Aug 23 18:24:21 DocScrutinizer: my friend says N900 would be great: you start with typing an extremely long (to make a good file) mencoder command line to capture some small video, then do actual shooting, then an even longer mplayer line to just see you screwed one of the parameters for encoding. ;) Aug 23 18:24:24 F4t: yup Aug 23 18:24:31 whats current svnr in repo? Aug 23 18:25:31 F4t: 110 Aug 23 18:25:49 PaulFertser: hehe! so we feel almost at home, no? Aug 23 18:25:53 ah 110 is good... Aug 23 18:26:23 DocScrutinizer: yes, except lots of proprietary drivers again. Aug 23 18:26:23 cool Aug 23 18:27:35 PaulFertser: sure, I'm aware of that. Still I guess hacking those proprietary drivers is less effort than building a device with obsolete chips that have FOSS drivers Aug 23 18:29:07 and tbh I don't care much about closed drivers as long as they don't interfere with core system and API is comprehensive and functional Aug 23 18:29:36 DocScrutinizer: and what do you think about the device raster promises? Aug 23 18:30:01 hmm, didn't hear much of that lately Aug 23 18:31:14 if it's exactly built to raters order it has inferior screen (like 320*200), otherwise will be identical :-P Aug 23 18:31:29 nah, crappy C-TS probably Aug 23 18:32:17 DocScrutinizer: he says that one high-profile company is going to release iPhone-like specd device in a semi-open way. Aug 23 18:32:55 and you're sure he didn't talk about Nokia? :-D Aug 23 18:33:04 DocScrutinizer: yes Aug 23 18:33:30 well, me too. I even think I know device and company Aug 23 18:46:20 \o/ just updated, GPRS works for me again Aug 23 18:55:02 sybren: congrats Aug 23 19:10:50 * dos1 wonders, why there is quite big amount of people who thinks SHR had close development and hard to contribute... Aug 23 19:11:10 we need some marketing! ;) Aug 23 19:12:51 dos1: hmmm Aug 23 19:14:39 s/had/has/ Aug 23 19:15:11 i just proved someone on maillist that he's wrong :P Aug 23 19:15:28 "You are right, there is no problem! SHR is an open distro, sorry by the mistake." Aug 23 19:17:16 dos1: Maybe update wiki.openmoko.org to be more SHR firendly.. Aug 23 19:21:37 or keep SHR out of wiki.openmoko.org and just use http://shr-project.org/trac/wiki Aug 23 19:23:23 DocScrutinizer: mmmh... N900 maybe? Aug 23 19:23:50 Wonka: maybe what? Aug 23 19:23:51 DocScrutinizer: one _might_ suspect that one after http://linux.slashdot.org/story/09/08/21/1155217/Nokia-Leaks-Phone-With-Full-GNULinux-Distribution?from=rss flew around for some days :> Aug 23 19:24:02 Problems with Categories in shr-u? I got two icons of Sudoko and Midori in SHR-u. If I remove GTK; from the midori.desktop Categories=GTK;Network; and change in Soduku from Categories=Applications;Game;ENLIGHTENMENT; to Categories=Game; It will only show one icon instead.. Aug 23 19:24:16 aah, raster? Nooo Aug 23 19:25:01 TeLLuS: known old "problem" Aug 23 19:25:39 TeLLuS: still nobody cared to fix it yet it seems Aug 23 19:31:23 though for e.g. midori I don't see the double icon on desktop with shr-u88 Aug 23 19:31:34 DocScrutinizer: Is it the right way to fix it? Should I open a ticket or bug someone about it.. Aug 23 19:31:55 hmm, good question Aug 23 19:32:39 I the double Midori after the last update today. and the Sudoko about a week ago. Aug 23 19:32:45 probably best way is to ask dos1 or mrmoku|away as soon as he's back from holiday Aug 23 19:33:12 dos1: ping Aug 23 19:35:52 TeLLuS: hmm Aug 23 19:36:16 i added GTK to displayed categories Aug 23 19:36:30 so navit icon is installed Aug 23 19:36:57 aaah-ha! Aug 23 19:37:19 ah Aug 23 19:37:23 s/installed/displayed/ Aug 23 19:37:49 So why cant applications belog to multiple categories and only be displayed once.. Aug 23 19:38:16 seems the root problem is that it's possible to display an icon twice Aug 23 19:39:04 Or just make applications use only one categorie Aug 23 19:39:25 TeLLuS: ok, i think i have fix for that Aug 23 19:39:35 that's not the correct fix Aug 23 19:39:50 (use only one) Aug 23 19:39:55 I agree.. Aug 23 19:40:37 SHR: 03seba.dos1 07shr-themes * rd8e72425ab94 10/e-wm/e-wm-theme-illume-niebiee/flaunch.edc: Merge branch 'master' of git+ssh://git@shr.bearstech.com/shr-themes Aug 23 19:40:41 DocScrutinizer: no, that's nor my fix Aug 23 19:40:42 illume needs to get smarter... don't disply duplicates, learn to sort icons Aug 23 19:40:43 ;) Aug 23 19:40:44 SHR: 03seba.dos1 07shr-themes * ra16b05529a51 10/e-wm/e-wm-menu-shr/applications.menu: e-wm-menu-shr: try to avoid duplicating some icons on desktop Aug 23 19:41:04 DocScrutinizer: that was wrong menu file Aug 23 19:41:10 TeLLuS: building package now Aug 23 19:42:12 Is it bitbake e-wm-menu-shr? I can build it myself.. Aug 23 19:43:18 TeLLuS: yes, but when you'll build it it won't hit repo ;) Aug 23 19:43:36 TeLLuS: building package-index now :P Aug 23 19:44:17 yes, but you know it is OK to test first.. :) Aug 23 19:45:05 TeLLuS: it won't break anything, so even if it won't help i think we can avoid testing ;) Aug 23 19:45:38 TeLLuS: opkg update; opkg install e-wm-menu-shr; killall -HUP enlightenment Aug 23 19:48:15 TeLLuS: nothing disappeared here ;) Aug 23 19:48:41 TeLLuS: but i didn't have duplicates (i don't have navit exposed on desktop, i have links instead), so you also have to check it Aug 23 19:55:07 dos1: you're about to fix a bug that crept in because of assumptions. Now you say it doesn't need testing because you assume it won't break anything ;-) Aug 23 19:56:26 DocScrutinizer: i'm just lazy and you know it, just "it won't break anything" sounds better than "i'm too lazy to test it before" :P Aug 23 19:56:27 Anyway verified now. Works for me.. Aug 23 19:56:29 dos1: If a learned anything about releasing, then it's the need to test *everything* Aug 23 19:56:42 DocScrutinizer: on -unstable? ;) Aug 23 19:57:01 yes Aug 23 19:57:17 and without release, just in repo Aug 23 19:57:27 even debian has sometimes problems with packages in unstable Aug 23 19:57:33 even unstable isn't the unpayed testers for you. Aug 23 19:58:03 (but we need -testing and -stable to make my previous sentence matching to SHR ;/) Aug 23 19:58:16 dos1: Do you know about Debian experimental :) Aug 23 20:00:09 TeLLuS: yes, i know ;) Aug 23 20:00:32 but -experimental will be too much for SHR at actual stage :D Aug 23 20:16:49 ok, going to bed Aug 23 20:16:51 night Aug 23 20:34:07 Will http://andreasdalsgaard.blogspot.com/2008/07/openmoko-development-in-5-minutes.html give me the toolchain required to do compile applications that run on SHR Unstable? Aug 23 20:43:24 yes but I think you wont be able to compile apps with certain librarys like elementary Aug 23 20:44:27 this should work for all apps http://shr-project.org/trac/wiki/Building%20SHR Aug 23 20:45:15 yeah, but that takes a cr*pload of time, and isn't the most stable Aug 23 20:45:47 i.e. I've tried it and couldn't get things to build properly. Aug 23 20:48:00 I was hoping I could just install some binaries and get compiling, like maemo Aug 23 20:49:46 then try the 5 minutes approach. it might work for you Aug 23 20:50:14 that's good to know :) Aug 23 20:50:46 sybren: without OE you won't get proper dependency management and most other necessary things anyway. Aug 23 20:50:55 sybren: i guess it's really better to use SHRMakefile... Aug 23 20:51:53 PaulFertser: last time I tried that, my machine let me know I had to clean my CPU fan by overheating and crashing Aug 23 20:51:55 is it save to upgrade to the newest fso upgrade from today? does GSM and suspend work with newest packages? Aug 23 20:52:35 Slyon: I just upgraded and things look fine Aug 23 20:52:52 sybren, thanks, so i'm gonna try it Aug 23 20:53:00 sybren: you're supposed to be using a working machine to compile OE, that's true ;) Aug 23 20:53:23 PaulFertser: :P Aug 23 22:28:43 I've got a failure compiling qemu-native. The error is: .../qemu-native-0.9.1+svnr4242-r14/trunk/target-i386/ops_sse.h:597: error: unable to find a register to spill in class `GENERAL_REGS' Aug 23 22:28:56 I've googled, and found it may have something to do with gcc version Aug 23 22:29:05 which gcc is suggested for OE? Aug 23 22:29:09 SHR, that is Aug 23 22:30:16 ah never mind, I should have read the 'troubleshooting' section first ;-) Aug 23 22:38:35 Has anyone tried using freerunner as a bootable usb flash drive? Sounds like fun. Aug 23 22:39:27 that would be quite useful, as my image tank is broken and I can't find my USB stick anywhere :P Aug 23 22:40:07 sybren: i've just tried to do that, no idea if it will boot on a real pc but "qemu -hda /dev/sda1" works ok. Slax rules :D Aug 23 22:40:26 sybren: i assume you disabled binary locale generation and avoided qemu build failure? Aug 23 22:40:56 hello all Aug 23 22:41:52 PaulFertser: I limited the locale to en_GB.utf-8, and changed qemu's configure file, then did "./configure; make", as described in the troubleshooting section of "Building SHR". Aug 23 22:43:08 sybren: ah, ok. Qemu is also used to build some tricky packages like emacs but generally you might avoid it, OE supports building locales on first boot on target. Aug 23 22:46:29 PaulFertser: any wiki page I can find that on? I'm just following the "Building SHR" to the letter Aug 23 22:47:14 btw, qemu just survived a "make", so I'm doing "bitbake shr-lite-image" again to see if it works now. Aug 23 22:50:30 NOTE: package qemu-native-0.9.1+svnr4242-r14: task do_compile: started Aug 23 22:50:43 why is it compiling when I just a "make" in its work directory? Aug 23 22:50:57 yep, another build failure on qemu-native Aug 23 22:52:02 sybren: OE uses "stamps" to look what to do next for each package. Aug 23 22:52:26 sybren: i think i saw somewhere a clear hint how to disable binary locale generation. Probably in the conf file itself. Aug 23 22:52:52 sybren: yup, obviously ENABLE_BINARY_LOCALE_GENERATION = "0" Aug 23 22:53:00 excuse me Aug 23 22:53:00 PaulFertser: what does that have to do with qemu? Aug 23 22:53:20 pbaxter: for what? Aug 23 22:53:23 can i ask help tp put other applications in teh bar on the down of the screen Aug 23 22:53:24 sybren: OE generates binary locales by using qemu. If you don't need it it won'ttry to build quemu. Aug 23 22:53:50 pbaxter: don't ask for what you can ask - just ask your question Aug 23 22:54:02 http://scap.linuxtogo.org/files/f916ec6fca9bdd0655daf3645a8eb590.png Aug 23 22:54:05 like this Aug 23 22:54:15 now i have only a Dialer in the middle Aug 23 22:54:15 PaulFertser: so the wiki is wrong? Changing configure, then a manual "./configure; make" is not enough? Aug 23 22:54:30 how do i put other apps into the other two fields? Aug 23 22:54:47 sybren: probably that's another way to avoid build failures. I prefer to not use qemu at all for that. Aug 23 22:54:54 pbaxter: no idea Aug 23 22:55:07 PaulFertser: well, it's not a way to avoid failures, since I did that and it still fails Aug 23 22:55:21 Paul, you know how to? Aug 23 22:56:22 PaulFertser: skipping locale generation, it seems to be the easiest way to get somewhere right now Aug 23 22:57:55 pbaxter: no, sorry Aug 23 22:58:13 pbaxter: it's already late night in Europe, you need for those guys to wake up :) Aug 23 22:58:22 ol Aug 23 22:58:25 i'0m in europe Aug 23 22:58:39 pbaxter: i'm low-level guy, used to bad UIs ;) Aug 23 22:58:48 ah ok PaulFertser Aug 23 22:58:53 thanks though Aug 23 22:59:02 pbaxter: here it's almost 3am already ;) Aug 23 22:59:22 here it's 1 in the morning Aug 23 22:59:35 pbaxter: there is no1 in the morning - it's night ;-) Aug 23 22:59:46 it's only morning after 6 Aug 23 22:59:57 ok sybren Aug 23 22:59:59 i'm in the night Aug 23 23:00:02 (personal grudge against people skipping the pleasant phase of the day called 'night') Aug 23 23:00:45 PaulFertser : you know "High Hopes" of Pink Floyd? Aug 23 23:01:41 PaulFertser : this is something strange for you http://www.youtube.com/watch?v=bpV5InLw52U Aug 23 23:03:04 pbaxter-AFK: i'll try to not forget to listen to that tomorrow, it's time for me to sleep... gotta work tomorrow. Aug 23 23:03:24 PaulFertser: sweet dreems :) Aug 23 23:04:15 sybren: thanks, good luck compiling OE. It's not easy for the first time and even sometimes for the 10th. If you're lucky you'll manage it. If not, we'll help you ;) Aug 23 23:05:52 PaulFertser: thanks :) It seems to continue beyond the point of failure, so that's a good thing Aug 23 23:06:02 PaulFertser: we'll see where it ended up tomorrow ;-) Aug 23 23:11:11 good night PaulFertser (to know it's something like Jethro tal) Aug 23 23:12:26 pbaxter: I really like Jethro Tull :D "Let me play you songs from the wood. To make you feel much better" :) Aug 23 23:12:51 but this is something similar and different Aug 23 23:12:57 so good night Paul and good work Aug 23 23:36:28 freesmartphone.org: 03Frederik.Sdun 07dbus-hlid * r5df7f49fe888 10/src/obj.vala: Add ListChildrenByObject **** ENDING LOGGING AT Mon Aug 24 02:59:57 2009