**** BEGIN LOGGING AT Tue Nov 10 02:59:57 2009 Nov 10 03:11:32 hi,all.how to use the modules.kernel_version*.tar.gz file? scp to openmoko and extract it ? Nov 10 03:22:12 Breese_Kay: if you do not know it you do not need it i guess Nov 10 03:22:20 ;) Nov 10 03:23:17 Breese_Kay: but yes, the idea is that if you need particular kernel modules you copy it to your rootfs and etract from /. Nov 10 03:29:41 is there a howto on installing shr to a mini-sd from my laptop? Nov 10 03:33:38 PaulFertser: OK,I think I don't need it maybe . Nov 10 06:22:42 good morrow. qucik question: i am using latest mrmoku unstable with all upgrades till the feeds were moved. can i just edit my opkg feeds to mrmoku oemerge and upgrade? Nov 10 06:30:38 * JaMa would suggest reflash.. Nov 10 06:32:29 juri_: yes, it's quite easy. Nov 10 06:32:38 juri_: SHR releases both jffs2 and tarballs. Nov 10 06:33:03 juri_: you basically need to untar a tarball to ext2/ext3 partition on SD Nov 10 06:33:54 juri_: but if you want to use a kernel from SD you need to either make a small first fat partition or to update u-boot to the latest or to upgrade to Qi (upgrading BL on gta01 can be dangerous if you lack a debug board). Nov 10 06:33:58 JaMa, so quite a lot of changes... how well does the latest oemerge image work? Nov 10 06:35:38 JesusMcCloud: for me it works as shr/import did.. but ie you need to restart frameworkd sometimes.. Nov 10 06:37:50 hmmm. i think i will try tonight... Nov 10 07:57:02 morning Nov 10 07:57:56 JaMa|Wrk: frameworkd restarts shouldn't be needed anymore Nov 10 07:58:25 mickey fixed the race condition that prevented dbus autostart to work in ousaged Nov 10 07:58:29 fsousaged Nov 10 08:03:14 hmm right.. seen commit but haven't tested yet :) Nov 10 08:18:49 raster: awake? there is an elementary question on our shr-devel list. Nov 10 08:18:52 http://lists.shr-project.org/pipermail/shr-devel/2009-November/001166.html Nov 10 08:19:14 wants to create smaller windows similar to the power-off dialog. Nov 10 08:19:26 do you have time to give a pointer? Nov 10 08:30:49 morning Nov 10 09:18:56 does grep ERROR /var/log/frameworkd.log containt also "could not open channel UNSOL|CALL|MISC, retrying in 2 seconds" 4 times a minute for you? Nov 10 09:23:33 JaMa|Wrk: flashing a new image. can't test now Nov 10 09:32:42 misc-vapi fails, I do hate AUTOREV'ing did I say that already :) Nov 10 09:37:51 JaMa|Wrk: head ache today? ;) Nov 10 09:45:49 huh... reading one day backlog is quite long :P Nov 10 09:50:26 spaetz: not at all :P Nov 10 09:51:01 spaetz: which is really great after 4 hours of sleep :) Nov 10 09:51:44 hehe Nov 10 09:51:50 JaMa|Wrk: seen this patch? http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=e6ee9ab2e4370d6b006a1578210b38fbd72fb91a Nov 10 09:51:56 I guess the same is needed for eglibc Nov 10 09:52:44 as eglibc has the same .bbclass file Nov 10 09:52:57 spaetz: seen it but I'm not sure if its needed or just updated to new order Nov 10 09:53:15 * JaMa|Wrk havent built eglibc for about 5 days now! :) Nov 10 09:53:33 probably just update to new order, but if glibc needs it eglibc will need it too. (I guess) Nov 10 09:53:52 btw yesterday I found out that update-alternatives on neo are wrong .. Nov 10 09:53:52 I ping RP on #oe about it Nov 10 09:53:59 they are? Nov 10 09:54:05 mmh, all of them? Nov 10 09:54:11 opkg - 0.1.6+svnr240-r16.3 - /usr/bin/update-alternatives Nov 10 09:54:11 update-alternatives-cworth - 0.99.154-r2.4 - /usr/sbin/update-alternatives Nov 10 09:54:22 first one is first in PATH Nov 10 09:55:09 with my patch the second works better for me.. we should ship just one update-alternatives impl to confuse user a bit less Nov 10 09:55:49 agreed Nov 10 09:56:10 and as first one is probably binary (I already deleted it to check it now) it should be faster Nov 10 09:56:56 why does elementary RDEPENDS elementary-tests? it's suggestion not a depencie... Nov 10 09:57:38 spaetz: what is expected behavior if you install another alternative with the same priority as the one with highest already installed? Nov 10 09:57:59 no clue what update-alternatives wupld/should do in that case Nov 10 09:58:18 spaetz: for me: I expect to use the one which was added last (last update-alternatives --install call) Nov 10 09:58:20 Heinervdm: it probably shouldn't be RDEPENDS, agree Nov 10 09:58:36 I'd rather kill it off in at least the -lite builds anyway. Nov 10 09:58:44 spaetz: because then I can just use opkg install -force-reinstall gtk-theme-neo to switch alternatives link :) Nov 10 09:58:52 actually, at some point, I did kill it off in the recipe Nov 10 09:59:11 Heinervdm: what about puting it to RRECOMMENDS? Nov 10 09:59:27 JaMa|Wrk: makes sense to me. But no clue what debians update-alternatives would do in this case Nov 10 09:59:27 JaMa|Wrk: it's put into both Nov 10 09:59:37 delete from RDEPENDS +1 Nov 10 09:59:57 if we want it, we can always put it in task-shr.bb Nov 10 09:59:59 but bitbake will inlcude it into the image when it is build and it's in RRECOMENDS Nov 10 10:00:10 so we should remove both and add it to RSUGGESTS Nov 10 10:00:26 Heinervdm: right. Is there way to prevent some RRECOMMEDS to be included? Nov 10 10:00:38 Heinervdm: agreed Nov 10 10:00:42 alway wondered about that Nov 10 10:00:46 RSUGGESTS +1 Nov 10 10:01:24 spaetz: arrgh, that update-alternatives from opkg is also cworth's :) Nov 10 10:02:07 heh :) Nov 10 10:02:41 cookie sent Nov 10 10:03:04 * JaMa|Wrk eating Nov 10 10:03:31 * Heinervdm marking as eaten Nov 10 10:03:59 * mrmoku setting as accepted Nov 10 10:04:16 heh.. no Nov 10 10:04:20 :) Nov 10 10:04:28 :P Nov 10 10:04:50 talking of cookies.. (and cake).. Nov 10 10:04:58 TAsnn: ^^^ Nov 10 10:05:00 is anyone still having resume slowness with xf86-video-glamo? Nov 10 10:05:16 i'm having no suspend at all :D Nov 10 10:05:24 Weiss: another sleep found? :) Nov 10 10:05:34 Weiss: resumes fast for me Nov 10 10:06:14 nope, but I think someone (Heinervdm?) was saying they still had slowness after pulling the sleeps out Nov 10 10:06:33 that was JaMa i think Nov 10 10:06:50 mrmoku: mickey fixed fsousaged dbus autostart Nov 10 10:06:52 no I said that it was fast for me even with sleeps :) Nov 10 10:07:09 JaMa|Wrk: your FR seems to be unique and special :S Nov 10 10:07:13 it was a race condition in his code (dbus init put in some idle.add loop which screwed up things) Nov 10 10:07:23 its special... thats sure :P Nov 10 10:07:35 next on my "attack list" is EXA busywait-elimination Nov 10 10:07:38 spaetz: yep read that :) Nov 10 10:07:55 * spaetz cheers. Go Weiss! Nov 10 10:08:24 did dos build an image? Nov 10 10:08:35 * JaMa|Wrk mumbles about walnut-drm cake Nov 10 10:08:52 mrmoku: not after mickeys fix, I believe Nov 10 10:09:01 he build an image (onmy request) before that :) Nov 10 10:09:31 ok, will start another build then Nov 10 10:09:51 JaMa|Wrk: DRM-cake would be a bit chewy and metallic, I think. all those ioctls.. Nov 10 10:09:57 so, build a new image now and push it so shr-unstable. I'll write an announcement and say that upgrade path is most probably broken. Nov 10 10:10:22 :) Nov 10 10:10:36 better 'definitely broken'... people will try it anyway :P Nov 10 10:10:41 yes Nov 10 10:10:42 :) Nov 10 10:11:47 JaMa|Wrk: did you only eat or push too? Nov 10 10:12:44 I have a code name for the new shr-unstable. "Was lange waehrt, wird endlich gut" :) Nov 10 10:12:59 Heinervdm: autosuspend not working for you? or suspend in general? Nov 10 10:13:04 this will keep google translate busy for a while :) Nov 10 10:13:10 * mrmoku knows why autosuspend is not working :P Nov 10 10:13:18 why? Nov 10 10:13:34 because on common request I removed the handling for it from phonefsod... Nov 10 10:13:38 mrmoku: tried suspend and it shuts down Nov 10 10:13:42 and I guess nobody added the correct rules for it Nov 10 10:14:04 Heinervdm: hmm... that is strange Nov 10 10:14:15 tried only one time Nov 10 10:14:29 mrmoku: http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?h=shr/merge&id=51d8c69272483d1749587ba589f038e81718afc3 pushed... Nov 10 10:14:34 JaMa|Wrk: nvm... got confused that your merge is last commit... and not Heinervdm's :) Nov 10 10:15:31 spaetz: maybe we should add a autosuspend rule before building the new image Nov 10 10:16:16 # Additional content II (can be masked with BAD_RECOMMENDS) Nov 10 10:16:52 so we can remove avahi with BAD_RECOMMENDS = "avahi-daemon avahi-autoipd" Nov 10 10:22:44 mrmoku: yes, that could help :) Nov 10 10:22:48 Is there none? Nov 10 10:23:12 nop Nov 10 10:23:15 Heinervdm: YES, let's remove Avahi! Another 2 seconds startup savings and less RAM Nov 10 10:23:22 mrmoku: surprising. Good catch :) Nov 10 10:23:28 * spaetz drafts a blog post Nov 10 10:23:39 you have been using avahi? ;) Nov 10 10:23:39 dos1 said he has some nice rule for that... but he's not around :| Nov 10 10:23:51 lindi-: yes, it's installed by default Nov 10 10:23:53 lindi-: using would be an overstatement ;) Nov 10 10:24:03 spaetz: "by default" in what? ;) Nov 10 10:24:14 by default in basically all OE images. Nov 10 10:24:28 its "RECOMMEND=" in some basic OE config file Nov 10 10:24:34 ah Nov 10 10:24:40 so you need to explicitely remove that and nobody bothered Nov 10 10:25:06 avahi isn't especially useful on the FR :) Nov 10 10:25:20 Hi there. Do you know of any documentation where one can get information of how "size_hint_weight_set" and "size_hint_align_set" should be used in python elementary? pls Nov 10 10:25:39 spaetz: well the first thing I do one a memory constrained system is to run "ps axuf" and remove anything extra Nov 10 10:25:44 s/one/on/ Nov 10 10:25:45 lindi- meant: spaetz: well the first thing I do on a memory constrained system is to run "ps axuf" and remove anything extra Nov 10 10:25:50 ok, what are the major changes: 1)xorg 2)eglibc 3)gry theme 4)phonefsod 5)phoneui Nov 10 10:26:03 lindi-: right. Nov 10 10:26:13 anything forgotten? Nov 10 10:26:21 lindi-: that's how it should be Nov 10 10:26:37 ahh, 6) fsousaged Nov 10 10:26:50 spaetz: not uservisible... but 7) based on org.oe.dev Nov 10 10:28:11 spaetz: neo theme :) Nov 10 10:28:55 spaetz: bumped efl.. so no more opkg.org packages :) Nov 10 10:29:20 spaetz: and I wish that extra could include dbus :-P Nov 10 10:29:50 ftn: the is no documentation, what do you want to know? Nov 10 10:30:00 spaetz: please write there that non-shr builds won't probably work and that stuff not available in shr repo should go to track with bbfiles :) Nov 10 10:30:34 s/builds/built packages/ Nov 10 10:30:34 JaMa|Wrk meant: spaetz: please write there that non-shr built packages won't probably work and that stuff not available in shr repo should go to track with bbfiles :) Nov 10 10:30:35 Heinervdm I'd like to place two buttons in the same row Nov 10 10:30:53 ftn: use a horizontal box Nov 10 10:32:16 Heinervdm alright but so far I've only seen elementary.Box. Is there any elementary.HBox or how do I do it horizontal? Nov 10 10:32:41 ftn: it has an option to make it horizontal Nov 10 10:33:39 why does edje-utils depends on embryo-tests? Nov 10 10:33:57 Heinervdm alright, I'll look for it. Thanks! Nov 10 10:34:13 mrmoku: good. org.oe.dev. JaMa|Wrk neo theme is not installed by default, right? Nov 10 10:34:43 spaetz: that's for the feed changelog ;) Nov 10 10:38:28 spaetz: yes its just in feeds Nov 10 10:39:12 ok, will add neo theme Nov 10 10:39:17 spaetz: but gpe icons/libframeworkd is messed up because of it.. so let users know that thay can install it with opkg install shr-theme-neo now :) Nov 10 10:39:33 JaMa|Wrk: will also comment that external packages will probably be broken and should be tracked in OE Nov 10 10:39:52 great Nov 10 10:41:12 what opimd backend is on by default? Nov 10 10:41:18 SIM card or sqlite? Nov 10 10:42:17 spaetz: SIM for Messaes, sqlite for everything else Nov 10 10:44:18 spaetz: on as in load data from it or on as default? Nov 10 10:46:08 as in save data Nov 10 10:46:12 thanks Heinervdm Nov 10 10:51:57 draft (pastebin.ca is SLOOOW) http://pastebin.ca/1664704 Nov 10 10:54:13 spaetz: aren't we using the latest available kernel? is that worth mentioning? Nov 10 10:54:26 sure. Will add that Nov 10 10:54:34 what are we on 2.6.29, right Nov 10 10:54:59 and don't try to squeeze in 2.6.31 before pushing this out :) Nov 10 10:55:08 huh? /me must be confused then, I thought we were already on .31 Nov 10 10:56:59 nahh, not in shr/merge. I think Jama experimented with .31 Nov 10 10:58:11 ahh, I'll mention that the time is off and ffalarms crashes. Nov 10 10:58:16 just to avoid those bug reports Nov 10 10:59:02 spaetz: when you mention neo theme, then you can mention niebie too Nov 10 10:59:02 ah, I was thinking that because I remembered seeing some packages being downloaded from the 31 branch, though I must have seen something different :D Nov 10 10:59:35 Heinervdm: sure. added Nov 10 11:00:26 Heinervdm: opkg install shr-theme-niebiee? Nov 10 11:00:52 spaetz: don't know if tere is a meta package Nov 10 11:09:32 just checked, there is. ok Nov 10 11:14:51 mrmoku: can we get a current -lite image, a full image, and a feed and push that to shr-unstable, please? Nov 10 11:16:36 spaetz: no suspend rule? Nov 10 11:16:55 ok, will build without then Nov 10 11:17:29 i think we shouldn't push to shr-unstable without suspend rule Nov 10 11:17:30 build started Nov 10 11:18:12 Heinervdm: feel that too Nov 10 11:18:21 anyway... will let it build Nov 10 11:18:37 to see if fsousaged works Nov 10 11:18:53 mrmoku: it's working Nov 10 11:18:58 meanwhile kingergarten and lunch Nov 10 11:19:00 Heinervdm: good :) Nov 10 11:19:12 bbl Nov 10 11:22:23 Heinervdm: ok agreed, suspend should should be included Nov 10 11:22:45 but that is the last change that we are going to make before pushing it out :). Any takers Nov 10 11:22:48 ? Nov 10 11:22:51 so, who can add this feature? Nov 10 11:23:27 perhaps i will remove avahi too Nov 10 11:23:42 I have no clue about rules Nov 10 11:23:53 what is required to remove AVAHI you say? Nov 10 11:23:59 sth with BAD_RRECOMMENDS Nov 10 11:24:20 if that can be added to shr.conf. I can do that. Nov 10 11:24:29 do you have any info? Nov 10 11:24:57 i try to find out how it works Nov 10 11:25:48 I'll add BAD_RECOMMENDS=avahi to my shr.conf and see if it goes away Nov 10 11:26:42 spaetz: yeah meta packages should be for all themes now.. Nov 10 11:26:43 spaetz: tried that already Nov 10 11:26:54 that does not work Nov 10 11:27:21 spaetz: you can also add info about KMS kernel and 2.6.31 if mrmoku adds few more dirs on buildhost where is just virtual/kernel built Nov 10 11:27:53 spaetz: there is kernel for KMS (from drm-tracking) Nov 10 11:28:10 and 2.6.31 should be shr-built too for users to switch easily.. Nov 10 11:28:18 Heinervdm: trying BAD_RECOMMENDS+="avahi-daemon avahi-autoipd" now Nov 10 11:28:29 JaMa|Wrk: we can add that later anytime. Nov 10 11:29:09 only problem why not build them in task-shr-feed is that ie kernel-modules-snd module is named the same and if you install 2.6.31 it will opkg install just latest module.. Nov 10 11:29:09 JaMa|Wrk: will add information on KMS, although i hardly know anything about that. Feel free to post me a line in the changelog :) Nov 10 11:29:29 spaetz: nooo please wait for mrmoku virtual/kernel images... Nov 10 11:30:02 spaetz: he did that already in shr-kms or something.. but it's missing in tests/mrmoku now :( Nov 10 11:30:55 and you can have all 3 installed ok if you download uImage + modules.tar.bz2 Nov 10 11:31:01 there is always stuff to wait for. If we do that we'll never get a snapshot out Nov 10 11:31:29 we'll have to make a cut at some point. branch off a -testing candidate and then go on brekaing shr-unstable Nov 10 11:31:30 hmm no big cake from Weiss for users? Nov 10 11:31:51 spaetz: this is not about breaking -unstable.. Nov 10 11:32:39 no, it's about getting a new -unstable out, which we haven't done since september :) Nov 10 11:32:52 alright.. as you wish.. Nov 10 11:34:51 JaMa|Wrk: hehe, if everyone else disagrees, fine. But I really want to get an image OUT this week. Nov 10 11:35:35 and branching of a -testing candidate Nov 10 11:37:06 spaetz: that's just about buildhost speed for building virtual/kernel and toolchan is the same.. so mrmoku can just copy tmpdir, purge deploy dir build virtual/kernel with small change in conf/distro/include/shr-om-gta02.conf Nov 10 11:37:20 in which packages are the rules that should contain the suspend? Nov 10 11:37:50 JaMa|Wrk: sure, if you get it in quickly, do it :) Or have mrmoku do it. Nov 10 11:38:01 spaetz: RP pushed simillar patch for eglibc.. Nov 10 11:38:12 JaMa|Wrk: he notified me. I pinged him. Nov 10 11:38:13 cool Nov 10 11:38:22 spaetz: I've no account on buildhost.. Nov 10 11:38:53 has the ffalarm segfault issue been solved? as that might be a bit of an issue for users Nov 10 11:39:05 sicu: no, but it's release noted :) Nov 10 11:39:28 I suspect it's another -O2 vs -Os issue and it might not crash if we compiled with O2 Nov 10 11:39:55 JaMa|Wrk: neither have I :) Nov 10 11:40:25 * spaetz can't find the package which contains suspend rules shr git contains way too many repositories :) Nov 10 11:40:35 spaetz: ok then you can put there comment that download dir for shiny 2.6.31 and KMS kernel will be provided later.. Nov 10 11:41:00 JaMa|Wrk: noted Nov 10 11:42:30 We will be providing a possibilitiy to upgrade the kernel to 2.6.31 (including KMS goodness, for increased visual pleasure) for adventorous users some time after this release. We just had to make a cut before this made it in. Nov 10 11:48:30 spaetz: not sure about visual pleasure :) Nov 10 11:53:33 so, what is the advantage of KMS for the FR? Nov 10 11:53:53 speed? les flickering? Nov 10 11:53:54 less Nov 10 11:55:17 spaetz: found the error :D Nov 10 11:55:29 it's spelled wrong in minimal.conf Nov 10 11:55:35 BAD_RECOMMENDATIONS = "avahi-daemon avahi-autoipd" is the right way Nov 10 11:55:36 ahh Nov 10 11:55:54 heh, so we need to patch org.oe.dev? Nov 10 11:56:07 what's the misspelling? Nov 10 11:57:39 BAD_RECOMMENDS -> BAD_RECOMMENDATIONS Nov 10 11:58:53 OK, I'll prepare a patch for OE Nov 10 11:59:02 spaetz: advantage is that kernel can show panic when something went wrong and you're still in graphical mode and that apps can share graphical ship acceleration (not used now, but -vo glamo should be replaced with vdpau module using drm) Nov 10 11:59:16 spaetz: best would be to point to Weiss info page http://www.bitwiz.org.uk/s/how-dri-and-drm-work.html Nov 10 12:17:21 JaMa|Wrk: cool. included Nov 10 12:24:23 no clue how an autosuspend rule should be crafted Nov 10 12:24:39 I just found our rules.yaml again. It's in shr-themes.git. Nov 10 12:24:39 * Heinervdm either Nov 10 12:25:02 but then I don't find suspend rules for older frameworkd versions either. Nov 10 12:25:11 dos1: do you know how to do autosuspend with fso? Nov 10 12:26:26 the trigger must be IdleStatus Nov 10 12:26:43 filter on "suspend" Nov 10 12:26:52 then call the SuspendAction Nov 10 12:27:23 trigger: IdleState() Nov 10 12:27:34 fileters: HasAttr(status, "suspend") Nov 10 12:27:45 actions: SuspendAction() Nov 10 12:27:47 ? Nov 10 12:27:49 trying Nov 10 12:28:03 s/fileters/filters) Nov 10 12:28:51 hmmm yummy Nov 10 12:29:22 hehe Nov 10 12:33:04 yo DocScrutinizer-8, what's the latest saying on your new toy? Nov 10 12:33:12 as in ETA... Nov 10 12:33:22 humpf Nov 10 12:33:41 webshop still 11-25 Nov 10 12:34:25 #maemo there's rumour shiping starts today - no idea where in the world though Nov 10 12:34:43 spaetz, mickey|office: that is missing the check for active calls then though? Nov 10 12:35:04 yeah Nov 10 12:35:17 well, no Nov 10 12:35:23 this is orthogonal Nov 10 12:35:36 you just need to claim the CpuResource on active Nov 10 12:35:37 no suspend idle state anymore while having active calls? Nov 10 12:35:40 ahh Nov 10 12:35:49 I think we're already doing that Nov 10 12:36:50 allocate display from dialer? ;-) Nov 10 12:37:07 bbl Nov 10 12:37:35 dos1: could you commit your auto-suspend rule please? Nov 10 12:37:43 spaetz is now known as spaetz->fileters Nov 10 12:40:03 Heinervdm: what is the state regarding incoming SMS? Nov 10 12:40:36 mrmoku: don't know Nov 10 12:40:46 haven't recived any in the last 2 days Nov 10 12:41:12 hmm... me neither Nov 10 12:41:13 mrmoku: what about another virtual/kernel on buildhost? Nov 10 12:41:27 Subject: Re: kernel 2.6.3x? Where can I download the kernel 2.6.3x? Nov 10 12:42:08 JaMa|Wrk: what to do? Nov 10 12:42:27 guys! Send "INFO" to 5667 (or sth similar) Nov 10 12:42:36 mrmoku: i'm getting PIN dialog twice Nov 10 12:42:54 Heinervdm: double-security ;) Nov 10 12:43:07 ;) Nov 10 12:43:10 mrmoku: do you want to share toolchain or is it ok to rebuild it? Nov 10 12:43:46 sharing would be faster :P Nov 10 12:44:12 but a bit dangerous.. but should be ok.. Nov 10 12:44:25 I can rebuild too... if that is better Nov 10 12:44:40 I guess I have to rebuild anyway Nov 10 12:44:42 just as you did shr-kms before Nov 10 12:45:16 12:36:48 < JaMa|Wrk> spaetz: that's just about buildhost speed for building virtual/kernel and toolchan is the same.. so mrmoku can just copy tmpdir, purge deploy dir build virtual/kernel with small change in conf/distro/include/shr-om-gta02.conf Nov 10 12:45:38 and in second tmpdir change conf/distro/include/shr-om-gta02.conf preferred kernel to linux-openmoko-2.6.31 Nov 10 12:45:58 (5667) =loop, for O2. Free of charge Nov 10 12:45:58 mickey|office: btw... what do PalmPre owners do when they travel abroad and start to do roaming? Nov 10 12:45:59 would be nice to have clean rebuild Nov 10 12:46:28 mrmoku: because otherwise there would be unneeded stuff in stagging area Nov 10 12:46:31 DocScrutinizer-8: me is eplus with contract Nov 10 12:46:45 JaMa|Wrk: yeah, doing that now Nov 10 12:46:58 mrmoku: good question, right now i would not recommend that. rwhity probably knows more Nov 10 12:47:02 that's reason I suggested " (or similar) " Nov 10 12:48:59 mickey|office: another very strong reason to use a different software on it :P Nov 10 12:49:02 btw I guess everybody is absolutely aware you can send SMS to yourself Nov 10 12:49:19 DocScrutinizer-8: and pay 30c for each... Nov 10 12:49:31 that's the downside Nov 10 12:49:32 mrmoku: true Nov 10 12:50:19 e.g. config msg which you can trigger on some service side of your provider is free of charge Nov 10 12:50:45 btw, you pay 30c / sms ?? o.O Nov 10 12:51:13 DocScrutinizer-8: let me check Nov 10 12:54:37 DocScrutinizer-8: worse even Nov 10 12:54:38 39c Nov 10 12:54:47 WTF?? Nov 10 12:55:09 find a better provider! ;-) Nov 10 12:55:15 you send sms abroaf?! Nov 10 12:55:25 already did... just waiting for time to pass Nov 10 12:55:27 *abroad Nov 10 12:55:32 max_posedon: that is NOT abroad Nov 10 12:55:54 DocScrutinizer-8: contract still running till 07/10 Nov 10 12:56:49 you bot phone with contract? Nov 10 12:57:01 its extrimelly expensive Nov 10 12:57:22 max_posedon: I don't care about SMS... as I don't write any Nov 10 12:57:30 just for testing it would be nice :) Nov 10 12:58:56 mrmoku: considered sipgate.de? Nov 10 13:01:50 DocScrutinizer-8: that is 7,9c ... much better yes Nov 10 13:01:51 rather convenient to send SMS from a real machine with decent kbd Nov 10 13:07:14 (related decent kbd) on nokia 6210 they did not support , so I had to do a hack and create a template out of an inbound sms, that contained just that , to insert into new SMS Nov 10 13:08:27 DocScrutinizer-8: you have O2 with InternetPack M? Nov 10 13:08:35 L Nov 10 13:09:20 what is the difference? more traffic allowed before tuning down to GPRS speed? Nov 10 13:09:41 though the L pack expired yesterday. Time to buy my monthly 30 bucks fillup Nov 10 13:09:59 ahh.. prepaid stuff Nov 10 13:10:10 * mrmoku never liked prepaid stuff Nov 10 13:10:24 which gives me 25bucks for INTERNET L, plus 35 bucks for standard calls and SMS Nov 10 13:10:50 (never liked) so beat the bill above Nov 10 13:11:44 5GB data plus 35 bucks worth of calls and SMS, for 30 bucks a month Nov 10 13:11:51 I did not say that I would pay less Nov 10 13:12:06 DocScrutinizer-8: how much traffic do you produce a month? Nov 10 13:12:41 a lot less for now (u know, GPRS :-/ ) Nov 10 13:14:24 and the "bonuskonto" where the 30bucks extra for calls and SMS sum up, now is at >150bucks Nov 10 13:18:35 mrmoku: I'd guess 5GB is plenty for any mobile usage Nov 10 13:19:52 I've got 6GB of 7.2Mbps with 10 euros over here.. as long as you have a good signal :) whatever you have over 6GB works with 256Kbps Nov 10 13:20:42 but it's on a dedicated sim card, without any telephony activated on it Nov 10 13:23:18 DocScrutinizer-8: is the UMTS speed difference between PalmPre and the N900 relevant? Nov 10 13:23:21 well, the datarate isn't limited here either. can use whatever BTS offers (which usually is HSPA on O2). Not sure what's going to happen when I exceed the 5GB limit Nov 10 13:23:55 mrmoku: how should I tell? I got neither yet ;-) Nov 10 13:24:29 nah, in theory :P Nov 10 13:25:48 in theory mine is: the RTT (ping time) is the bottleneck in UMTS, so HSPA isn't a big improvement for user experience over UMTS Nov 10 13:27:20 DocScrutinizer-8: ok Nov 10 13:28:00 with ping times >>500ms it's totally irrelevant if it take 5 or 1tms to transfer the requested data Nov 10 13:28:25 s/1t/15 Nov 10 13:32:03 my knowledge of IP is limited, but shouldn't RTT not matter (much)? I mean, if it takes 20ms or 500ms to request data and receive the first reply packet doesn't matter because the packets 2..n should come directly afterwards, no? Nov 10 13:32:15 btw that's the same problem as with IP over sat I guess Nov 10 13:33:53 blindcoder: not if you simply got no follow up packets, or worse: the embedded png can be requested only (creating another 500ms lag) *after* the html data arrived Nov 10 13:34:39 DocScrutinizer-8: ah, okay. Nov 10 13:34:41 if it uses tcp, it will send a bunch of packets (how many fit in the tcp window) and then wait for arrival confirmation, afaik Nov 10 13:34:55 DocScrutinizer-8: yeah, that would add up, right. Nov 10 13:38:11 rebelThor: I'm also no TCP guru, but I guess that's async, and will retransmit packets that got no ACK, after a certain timeout, out of sequence Nov 10 13:40:02 DocScrutinizer-8: yep, if it doesn't receive the ACK, it will lower the window size, and retransmit packets, and then wait some more.. i guess if it's a really slow connection, it will reach a really small window size, so it'll have to wait a lot for acknowledgements, and delaying the whole thing Nov 10 13:40:11 ok, another son pickup service to fulfill... bbl Nov 10 13:42:06 I hate it when I implement feature requests and they show bugs in totally unrelated pieces of code Nov 10 13:43:42 in this case, hackersdiet new feature: multiple profiles shows that the preferences get initialised in the wrong place Nov 10 13:44:00 though that's the common scenario Nov 10 13:51:27 hmmm, 15 days til N900 shipping date. Time for Nokia to postpone the date once more - they never postponed the date less than 15 days from shipping yet Nov 10 13:57:28 what is the expected price(without a plan)? Nov 10 14:04:35 600EUR at Nokia, 500 at resellers Nov 10 14:04:54 www.bada.com. Muhahaha Nov 10 14:05:59 it's open, but it seems impossible to even find out which OS paradigm it follows Nov 10 14:06:33 * DocScrutinizer wonders what's Samsungs definition of "open" Nov 10 14:07:38 mrmoku|away: http://build.shr-project.org/sources/ should contain sources used while oemerge building asap.. Nov 10 14:08:06 mrmoku|away: people with forbiden git:// svn:// protocols cannot fetch sources :) Nov 10 14:10:06 maybe it's open like ooxml Nov 10 14:11:47 In OM200x I was setup /sys/devices/platform/s3c2410-ohci/host_mode to 1 to provide power to a USB device. Is there any equivalent in SHR because I can't find any? Nov 10 14:12:50 use "find /sys -name host_mode" Nov 10 14:13:25 or look at wiki sysfs page, the addendum at bottom Nov 10 14:14:26 btw SHR has a slider to do all hostmode foo, in settings connectivity Nov 10 14:15:39 DocScrutinizer-8: There is a setting to set USB mode in "host" but not to set the host_mode in 1 Nov 10 14:15:53 huh?? Nov 10 14:16:04 the find didn't return any "host_mode" in /sys Nov 10 14:16:27 2here is which setting? (or not) Nov 10 14:16:58 0lease use the slider in SHR settings - connectivity Nov 10 14:17:05 or refer to wiki Nov 10 14:17:06 DocScrutinizer-8: By default in host mode the FR doesn't power the device. "usb_mode" can be set is settings but not "host_mode" Nov 10 14:17:30 jm3474: that's not quite precise Nov 10 14:18:33 in FR host mode is defined as "set protocol to host, enable power". This are 2 separate sysfs nodes to switch Nov 10 14:19:26 the names of those 2 nodes have changes quite some time ago, but they are still there Nov 10 14:24:52 DocScrutinizer-8: Where could I find the name of these 2 sysfs nodes? I searched the wiki but can't find so far Nov 10 14:25:06 and for me the SHRsettings-Connectivity "USB mode:" slider does the switching to e.g. connect a kbd quite conveniently (on SHR-U0808) Nov 10 14:25:18 *sigh* Nov 10 14:26:27 DocScrutinizer-8: You are right it switches both (usb host and power) I just tried it Nov 10 14:26:39 But I'd like to add this control to my app Nov 10 14:27:17 http://wiki.openmoko.org/wiki/GTA02_sysfs Nov 10 14:29:08 DocScrutinizer: ok thanks for the link Nov 10 14:29:42 http://wiki.openmoko.org/wiki/GTA02_sysfs#old:_.2Fsys.2Fdevices.2Fplatform.2Fneo1973-pm-host.0.2Fhostmode Nov 10 14:36:50 jm3474: echo host > find $(/sys/devices -name usb_mode) Nov 10 14:37:11 echo 1 > `find /sys/devices -name hostmode` Nov 10 14:37:25 JaMa|Wrk: hmm... suspend is actually broken on last image Nov 10 14:37:26 oops Nov 10 14:37:27 DocScrutinizer: Thanks it is working now Nov 10 14:37:38 jm3474: echo host > $(find /sys/devices -name usb_mode) Nov 10 14:37:42 JaMa|Wrk: Can't read RTC alarm. Returning 0 Nov 10 14:37:50 (from frameworkd) Nov 10 14:37:59 looks like result = pyrtc.rtcReadAlarm() Nov 10 14:38:01 does not work Nov 10 14:38:53 >>> print pyrtc.rtcReadAlarm() Nov 10 14:38:53 /dev/rtc0: Device or resource busy Nov 10 14:39:26 ffatd? Nov 10 14:39:52 Did you break something again, mrmoku? :p Nov 10 14:39:57 I think sms is commented out in phonefsod, so I don't know if it works Nov 10 14:40:03 back, btw :) Nov 10 14:41:09 mwester: yup, still have to understand what though... to add it to my Breaklog ;) Nov 10 14:41:24 mrmoku: would that patch solve suspend? http://dpaste.com/118671/ Nov 10 14:41:34 in all rules.yaml variations of course Nov 10 14:41:42 DocScrutinizer-8: good guess Nov 10 14:41:43 1427 ? S 0:00 /usr/sbin/atd /var/spool/at Nov 10 14:42:19 after killing it... Nov 10 14:42:20 >>> print pyrtc.rtcReadAlarm() Nov 10 14:42:20 ioctl(RTC_WKALM_RD): Invalid argument Nov 10 14:42:53 mwester: what is needed to make the kernel understand that ^^^ ? Nov 10 14:43:32 spaetz: patch looks good Nov 10 14:43:40 though suspend does not work at all :P Nov 10 14:45:42 mrmoku, no idea... do we have all the right defconfig settings? I would have thought we had all the RTC stuff implemented, but perhaps that's an old interface? (just speculating) Nov 10 14:46:31 JaMa|Wrk: ^^^ any idea? Nov 10 14:47:26 just flashed a new image, will apply my patch now to rules.yaml and check if it works Nov 10 14:47:35 and I thought we use a fso based implementation of atd Nov 10 14:48:07 DocScrutinizer-8: yep... we did... looks like it got lost :| Nov 10 14:48:21 which might explain the ffalarms segfaults too... Nov 10 14:48:37 so seems your breaklog entry gets more clear ;-) Nov 10 14:48:40 nah, ffalarms strace segfaults when reading a nonexisting file Nov 10 14:48:47 I doubt it has somthing to do with atd Nov 10 14:48:55 although it might be... Nov 10 14:49:48 wouldn't fopen on an unexisting file simply return an easily catchable NULL? Nov 10 14:50:39 catchABLE Nov 10 14:51:56 that's what separates the good coders from the sloppy ones - which don't check *each* returncode Nov 10 14:52:40 there's a small handful of return codes I don't check Nov 10 14:52:50 e.g. ALSA you should even check close() of a pcm-dev Nov 10 14:52:56 malloc, for example, because the best I can do in the case it fails is crash anyway Nov 10 14:53:39 or wait for memory to become available Nov 10 14:53:39 malloc is the only case where checking is not always mandatory Nov 10 14:54:11 while ((buf = (char *)malloc(256)) == NULL){ sleep(1);} // :-) Nov 10 14:54:15 but if you check you at least know where it crashed Nov 10 14:54:33 larsc: that can be gathered from the core file Nov 10 14:54:44 though even for malloc you can do a "shortcut" to exit() at best Nov 10 14:55:25 though a explicit exit(ERR_NOMEM) for sure is better choice Nov 10 14:57:31 it's better for sure than e.g. contine to use a linked list where last elem is marked by elem.successor = NULL Nov 10 14:58:57 I usually end my lists with a NULL pointer, so the check for data->next == NULL is done anyway, the problem is when the malloc fails and the initialising code begins assuming malloc succeedes. data->next->next = NULL; // segfault galore Nov 10 14:59:33 hrm Nov 10 14:59:35 I need to wrap malloc Nov 10 14:59:54 see.... Nov 10 14:59:58 :-D Nov 10 15:00:00 preferrably by a macro, but how do I macro buf = (char *)malloc(128); in a reasonable way Nov 10 15:00:13 preferrable with a printf including __FILE__ and __LINE__ Nov 10 15:00:29 hrm Nov 10 15:01:08 with gcc extensions its easy Nov 10 15:01:37 I'm compiling with -pedantic -Wall Nov 10 15:01:37 mrmoku: nice patch but it does not work. no suspend Nov 10 15:01:45 I'd prefer to keep it that way Nov 10 15:02:02 I do get this in the log: 2009.11.10 12:21:21.58 oeventsd.fso_triggers INFO Receive IdleState, status = suspend Nov 10 15:02:05 spaetz: well... as I said suspend does not work in no way Nov 10 15:02:21 but it does nt seem to trigger the rule Nov 10 15:02:43 hah Nov 10 15:02:43 I'll add a different action now, to see if the rule is triggered at all Nov 10 15:02:53 #define malloc(x) mymalloc(x, __FILE__, __LINE__) Nov 10 15:03:04 that should work Nov 10 15:04:19 blindcoder: probably any printf will fail when malloc failed before. Won't hurt to try though ;-). Use your private my-malloc() ? Nov 10 15:04:32 aaah, too slow Nov 10 15:06:15 mrmoku: I just inserted a xset -display localhost:0 s reset as Suspend action Nov 10 15:06:27 and the screen saver went off. So the rule is triggered Nov 10 15:06:33 but the SuspendAction is broken... Nov 10 15:06:57 yup, due to the rtc problem Nov 10 15:07:08 we have normal atd in the image... instead of atd-over-fso Nov 10 15:07:18 and I just understood why Nov 10 15:07:49 ahh, ok now I get it Nov 10 15:09:08 but the suspend rule is probably still needed. Nov 10 15:09:18 it just worked in shr/import, because there we had a PREFERRED_VERSION for atd Nov 10 15:09:24 matching the version of atd-over-fso Nov 10 15:09:25 mrmoku: I don't have commit access to shr-themes, I just send a patch to shr-devel Nov 10 15:09:33 that PREFERRED-VERSION got lost Nov 10 15:09:34 I see Nov 10 15:09:38 spaetz: ok, will commit Nov 10 15:09:47 so a preferred provider should be used? Nov 10 15:09:53 what we actually want instead is a PREFERRED_PROVIDER_atd="atd-over-fso" Nov 10 15:09:57 yup Nov 10 15:09:59 right Nov 10 15:10:11 heyho Nov 10 15:10:13 committed that one Nov 10 15:10:16 morphis_: hey Nov 10 15:10:26 suspend cookie: http://patchwork.dev.bearstech.com/patch/429/ Nov 10 15:10:41 * morphis_ is now connected from his Palm Pre :-) Nov 10 15:11:26 interersting to see IF this will fix ffalarms Nov 10 15:11:37 morphis_: with the wrong software running? Nov 10 15:11:43 btw I prefer "assert( (bla=malloc(foo)), x, y) Nov 10 15:12:21 mrmoku: with a webOS App ... Nov 10 15:12:21 spaetz: I'm manually building atd-over-fso now... to test Nov 10 15:12:35 morphis_: as I said... *wrong* software :P Nov 10 15:12:54 for everything else you will have to wait a little bit Nov 10 15:13:10 morphis_: I had an interesting thought today... what do you do when you go abroad... and have to roam? Nov 10 15:13:14 that will get expensive ;) Nov 10 15:13:37 but I am currently working on implementing the modem protocol Nov 10 15:13:43 :) Nov 10 15:14:51 puuh, don't know but I am not planing to go abroad in the next time Nov 10 15:15:12 morphis_: better finish the protocol implementation first :) Nov 10 15:15:54 hehe Nov 10 15:15:56 hehe Nov 10 15:16:25 modem protocol implementation? Nov 10 15:16:26 I hope I get the link layer implementation done this weekend Nov 10 15:16:29 mrmoku: should we keep obsolete git repositories on shr around? Nov 10 15:16:38 for Palm Pre Nov 10 15:16:39 spaetz: which ones? Nov 10 15:16:40 or kill them off (archive them somewhere) Nov 10 15:17:02 there's so many git repositories now, that it's hard to see which ones are the active ones. hold on Nov 10 15:17:04 spaetz: TAsn2 and me marked some of them as obsolete in git web Nov 10 15:17:20 (or even removed them from there... don't remember) Nov 10 15:17:34 well, starting with libframeworkd-phonegui-efl2.git Nov 10 15:17:41 libframeworkd-phonegui-gtk2.git Nov 10 15:17:54 libmodulo still active? Nov 10 15:18:02 last checking 5 months ago Nov 10 15:18:15 libxsb-frontend.git 9 months Nov 10 15:18:16 spaetz: well... we don't use it... but it is not obsolete Nov 10 15:18:25 ok then, just asking Nov 10 15:18:29 :) Nov 10 15:27:13 spaetz: atd-over-fso built Nov 10 15:27:32 I have to do another taxi service now... If you want to check if it fixes suspend and/or ffalarms? Nov 10 15:27:43 otherwise will do it myself when I'm back in half an hour Nov 10 15:27:52 bbl Nov 10 15:27:53 I only have 30 minutes left Nov 10 15:27:56 might try Nov 10 15:27:59 cu Nov 10 15:32:26 I'd like to repeat my suggestion to integrate shr-dates and ffalarms (and shr-notes) Nov 10 15:33:04 http://www.vanille-media.de/site/index.php/2009/11/10/towards-the-end-of-2009/ Nov 10 15:33:09 brief FSO2 status summary Nov 10 15:33:21 feel free to add questions, if there's something missing. I'll enhance the main article then Nov 10 15:35:15 mickey|office: thanks, will have a look Nov 10 15:35:34 k Nov 10 15:36:22 I'll include the status update in our blog/mail announcing a new shr-unstable snapshot too Nov 10 15:39:02 *gnaaah* Nov 10 15:39:12 uboot crapped itself again Nov 10 15:39:22 though I start to think it's the memory, not uBoot Nov 10 15:40:24 and no pc capable of flashing in reach Nov 10 15:40:26 spaetz: is a new official image coming out? :) Nov 10 15:41:05 rebelThor: as soon as things like suspend work. I hope Nov 10 15:41:19 mickey|office: how did you like Cambridge? Nov 10 15:43:32 something is weird with that suspendaction in oeventd Nov 10 15:43:38 this works: mdbus -s org.freesmartphone.ousaged /org/freesmartphone/Usage Nov 10 15:43:56 org.freesmartphone.Usage.Suspend Nov 10 15:44:06 but triggering a SuspendAction doesn't Nov 10 15:44:21 although it's supposed to do the same Nov 10 15:46:38 Weiss: beautiful landscapes and nice houses in the city and vincinity, a bit hard to get to from Frankfurt, but other than that i liked it a lot Nov 10 15:46:55 doh Nov 10 15:46:58 it works Nov 10 15:47:08 It's Suspend() not SuspendAction() Nov 10 15:47:11 oops :) Nov 10 15:47:19 hehe Nov 10 15:47:29 could somebody please toss me the mdbus cmdline to initiate a call Nov 10 15:47:35 mrmoku, I bake an updated cookie now Nov 10 15:48:20 mmmh, cookies Nov 10 15:49:04 cookjes and fileters Nov 10 15:49:35 for vampi Nov 10 15:52:33 mrmoku|away: please apply http://patchwork.dev.bearstech.com/patch/430/ to shr-themes Nov 10 15:52:40 confirmed to work with suspend Nov 10 15:53:00 I used it together with atd-over-fso, no clue whether just atd would have worked too Nov 10 15:54:46 ffalarms still crashing though Nov 10 15:54:55 DocScrutinizer-8: hehe Nov 10 15:56:42 dial via mdbus, anybody? Nov 10 15:56:52 sorry, no clue. Nov 10 15:57:56 I could do via debug call injecting a ATD to modem, but that's a little :-/ Nov 10 15:58:26 that should be something like Nov 10 15:58:59 mdbus -s org.freesmartphone.odeviced /org/freesmartphone/GSM/Device org.freesmartphone.Call.Initiate "+4912345678" "voice" Nov 10 15:59:10 mickey|away: thanks :-) Nov 10 15:59:22 hmm rather Nov 10 15:59:26 mdbus -s org.freesmartphone.odeviced /org/freesmartphone/GSM/Device org.freesmartphone.GSM-Call.Initiate "+4912345678" "voice" Nov 10 15:59:28 err Nov 10 15:59:29 no hypen Nov 10 15:59:31 a dot Nov 10 16:00:02 I had something like that from shr-contacts source, but didn't get right the exact spelling Nov 10 16:00:33 mickey|office: fsodeviced we'll have to look into CPU consumption at some poit Nov 10 16:00:41 10-30% on the FR now Nov 10 16:00:50 eeeeek Nov 10 16:00:50 gone for the next hour or so Nov 10 16:00:50 when? Nov 10 16:01:02 in idle mode? Nov 10 16:01:05 now, trying on shr-unstable :) and always Nov 10 16:01:10 even in idle mode, I guess Nov 10 16:01:14 do you have an image where i can observe this behaviour? Nov 10 16:01:21 no time, no image. Nov 10 16:01:25 talk to you later Nov 10 16:01:36 * spaetzMeeting is gone"now Nov 10 16:01:48 so I get my true speeddial icons in a few minutes. Finally speeddial :-D Nov 10 16:01:48 heh **** BEGIN LOGGING AT Tue Nov 10 19:53:47 2009 Nov 10 19:54:02 mickeyl, if its ok, can you pastebin your local.conf? here is mine http://dpaste.com/118792/ Nov 10 19:54:26 did some emacs foo to stip comments... so kindly ignore the ';'s Nov 10 19:55:31 http://vala.pastebin.com/d69d8bf40 Nov 10 19:55:39 i'm afraid that won't help you much ;) Nov 10 19:55:43 i have some stuff in my site.conf Nov 10 19:55:46 let me dig that out as well Nov 10 19:56:06 mickeyl, that is it? Nov 10 19:56:12 http://vala.pastebin.com/d5d7254f4 Nov 10 19:57:06 that is it?... Nov 10 19:58:11 yeah, local.conf typically only needs 10 lines or so Nov 10 19:58:27 i leave most values at their default as found in bitbake.conf Nov 10 19:58:53 hmm, my new image does not work, SIGSEGV all over the place, i guess i'm missing some eglibc patches still Nov 10 19:58:58 * mickeyl checks out shr/merge Nov 10 19:59:13 ah ok Nov 10 20:00:23 local.conf.sample may look scary, but in reality most defaults are ok Nov 10 20:00:33 bbiab, wife calls Nov 10 20:00:48 thanks for the headsup... If its ok with you, can I take this up in the mailing list if need arises?... someone else might find it useful Nov 10 20:03:21 for sure Nov 10 20:05:14 mickeyl, oh before you leave... what does aurora.cof under distro/ contain.. I see that you have set DISTRO to aurora Nov 10 20:06:42 wpwrak: well... she's cheating only while playing cards... so not that bad :P Nov 10 20:14:31 mickeyl: still segfaults? drat. Nov 10 20:15:12 mrmoku: so better lock the cards away while you're not around :) Nov 10 20:17:40 :) Nov 10 20:17:55 wpwrak: if she cheats others... the better for me ;) Nov 10 20:20:41 mrmoku: have you heard of the "broken window syndrome" ? :) Nov 10 20:23:41 * mrmoku fires wikipedia Nov 10 20:24:36 wpwrak: I see :) Nov 10 20:24:51 actually it got a looot better since when I got to know her Nov 10 20:25:03 she cheats less nowadays ;) Nov 10 20:25:35 mrmoku: or maybe she got a lot better ;-) Nov 10 20:26:11 heh, yes possible :P Nov 10 20:37:00 TAsn: hmm... why do we have apmd in the image? :P Nov 10 20:37:19 no idea ;] Nov 10 20:37:27 not my fault ;] Nov 10 20:37:50 JesusMcCloud, I wasn't able to add the keyboard icon correctly, (was ugly) please add it yourself :) Nov 10 20:38:27 TAsn, i saw it... *uärg* Nov 10 20:38:39 :) Nov 10 20:38:41 saw the scap? ;) Nov 10 20:38:45 yup Nov 10 20:40:38 have you even trie dthe 3-key-icon from kde? afair it looked really nice in mokel32's theme... i am still nto satisfied with the icons i looked at. by did u commit your chenges? i need to know so i can synch (i made soem changes which are not yet in git) Nov 10 20:41:51 TAsn, but no hurry at all since i am opkg upgradign atm, and i am quite sure my fr will end up fudged up Nov 10 20:42:14 JesusMcCloud, didn't commit mine Nov 10 20:42:16 and I'm not gonna Nov 10 20:42:19 as it's ugly ;] Nov 10 20:42:27 kk Nov 10 20:42:32 and JesusMcCloud please don't change the keyboard icon in the same commit Nov 10 20:42:36 as the other changes Nov 10 20:42:47 ? Nov 10 20:42:49 as I really like the idea of a keyboard icon, and really hate the other changes Nov 10 20:43:10 and i'd like to make myself a custom build without them and with the keyboard ;] Nov 10 20:43:29 furthermore, it's the correct way to commit (commit different stuff in different commits) Nov 10 20:43:53 sounds sound Nov 10 20:44:37 sounds sound? :) Nov 10 20:44:50 TAsn, what's the deal with the themecontest now? it'll still continue, right?! Nov 10 20:45:11 I can't remember what's the deadline, but I don't think we reached it yet. Nov 10 20:46:05 TAsn, i'm just wondering since gry* is default now... sort of Nov 10 20:47:22 JesusMcCloud, You won until we'll vote for a winner ;] Nov 10 20:47:38 The theme contest will choose the default theme *after* the end of the theme contest. Nov 10 20:47:46 until then it's like the wild wild west. Nov 10 20:48:03 all clear now Nov 10 20:51:17 sounds sound ... /me likes =] Nov 10 20:55:38 hi Nov 10 20:55:45 what can i do when he dont ask for pin? Nov 10 20:55:59 u know what i mean? ^^ Nov 10 20:56:36 Hardy: punch him Nov 10 20:57:28 he wouldnt forgive me anymore Nov 10 21:01:32 Hardy: He's allforgiving, didn't They told you? Nov 10 21:04:05 no they dont Nov 10 21:04:20 last time he slaped me when i was looking arround Nov 10 21:04:52 Hardy: that doesn't mean he doesn't love you, trust me Nov 10 21:04:59 lol Nov 10 21:05:09 hehe Nov 10 21:05:12 okay, i believe ya Nov 10 21:06:05 Hardy: you might want to provide a little more info if you want help Nov 10 21:07:37 which distro? Nov 10 21:08:20 okay thanks Nov 10 21:08:21 SHR Nov 10 21:08:35 SIM-Card is inside Nov 10 21:08:43 but he dont ask for PIN Nov 10 21:09:20 wait an moment Nov 10 21:09:26 this sim card got no pin xD Nov 10 21:09:50 he actually logged in to the provider Nov 10 21:10:13 so you don't actually have a problem anymore then? =] Nov 10 21:10:15 Hardy: great :) Nov 10 21:10:15 but, i dont recognized because the messenger and Contactlist breaks when started Nov 10 21:10:19 xD Nov 10 21:10:22 Hardy: you should wait Nov 10 21:10:37 Hardy: until the sim is fully initialized, that's known bug in old shr. Nov 10 21:10:48 my shr is updated Nov 10 21:11:42 i know on older versions, when no Sim card is inside, the contacts and the messenger will break because he cant read from it Nov 10 21:12:00 thats why i was so comfused ^^ Nov 10 21:12:56 so you get an error upon opening contactlist/messages? or just an empty list? Nov 10 21:13:07 error Nov 10 21:13:12 it breaks without reason Nov 10 21:13:27 segmentation fault Nov 10 21:13:43 the dialer is running Nov 10 21:23:33 * DocScrutinizer-8 wonders how much update SHR can take ATM, and to which state of functionality Nov 10 21:25:29 DocScrutinizer, how old your system is? Nov 10 21:25:44 if you still use glibc, update to eglibc isn't possible Nov 10 21:26:00 and right now, shr works for me, except.. Nov 10 21:26:08 it don't sustepd automatically Nov 10 21:26:14 and I'm not sure about sms's Nov 10 21:26:18 DocScrutinizer, here it seems liek EVERY SINGLE file is being replaced (opkg upgrading atm) Nov 10 21:26:37 well, I stopped upgrading 0808 somewhere around 2009-09 Nov 10 21:26:47 JesusMcCloud, and at the end your system won't be bootable Nov 10 21:27:08 I taked me 24h for update and fix a lot of issues, and nothing at the end Nov 10 21:27:28 JesusMcCloud: yep, normal Nov 10 21:27:45 max_posedon, i am expecting about the same, but i consider it an experimetn Nov 10 21:28:11 btw, does SHR finally update shr-unstable? Nov 10 21:28:41 nope, still only mrmoku/oemerge alive Nov 10 21:28:51 mrmoku, what's our unstable sync status? Nov 10 21:29:00 :) Nov 10 21:29:31 TAsnn: it's not funny if *you* play that running gag Nov 10 21:29:36 ;-D Nov 10 21:30:44 s/if/when Nov 10 21:30:44 :) Nov 10 21:31:18 that's the role of occasional visitors Nov 10 21:31:50 and we had it just 5h ago today ;-) Nov 10 21:32:40 TAsnn: suspend does not work for me Nov 10 21:32:48 TAsnn: for spaetz it seems to work... Nov 10 21:32:57 DocScrutinizer, next time I'll connect using another client with different name ;] Nov 10 21:33:16 as I want to be part of the game as well. ;\ Nov 10 21:33:46 and syncing an image that does not suspend does not make me happy ;) Nov 10 21:33:49 mrmoku, any leads? any differences? (hw revisions/fixes or maybe nand vs sd?) Nov 10 21:34:02 no idea... and spaetz is not around Nov 10 21:34:26 for me the screen blanks and it does not suspend Nov 10 21:34:37 and then the screen is just black until reboot Nov 10 21:34:56 (guess I could revive it via some magic sysfs nodes...) Nov 10 21:35:22 spaetz: ahh :) Nov 10 21:35:28 why in shr-settings I see `auto-sustend OFF`? Nov 10 21:35:29 * mrmoku just wanted to go to bed :P Nov 10 21:35:30 back ;) Nov 10 21:35:35 (usb cable detached!) Nov 10 21:35:59 heh, i am about to go to bed too Nov 10 21:36:00 max_posedon: fso borked Nov 10 21:36:03 spaetz: what is the difference between your and my moko? Nov 10 21:36:14 had a business dinner. Nov 10 21:36:18 DocScrutinizer-8, nope, it isn't Nov 10 21:36:32 diff? mine resumes and yours not... Nov 10 21:36:43 hehe, mine does not even suspend Nov 10 21:36:47 max_posedon: try switch it back to on Nov 10 21:36:53 I discovered that just the screen turns black :P Nov 10 21:37:10 seriously,a stock shr/merge image with a patched rules.yaml Nov 10 21:37:14 I just tested all phone functionality works Nov 10 21:37:15 max_posedon: use lsfsor Nov 10 21:37:22 outch Nov 10 21:37:24 spaetz: did you try a manual suspend too? Nov 10 21:37:27 spaetz, did you install it on nand or sd? Nov 10 21:37:51 mrmoku, did you use the same image or just built from the same recipes? Nov 10 21:37:53 yes, manual suspend via mdbus ousaged Nov 10 21:37:58 so, rules.yaml (or smth else) just didn't allowed suspend on cable detached! Nov 10 21:38:01 spaetz, special hw fixes? hw revision? Nov 10 21:38:01 TAsnn: buildhost image Nov 10 21:38:03 on sd card Nov 10 21:38:14 spaetz: but not via syscon menu Nov 10 21:38:14 both buildhost image on sd? Nov 10 21:38:17 * mrmoku on sd too Nov 10 21:38:21 a6, no hw fixes Nov 10 21:38:21 mrmoku, who responsible on such action? Nov 10 21:38:25 (mine is nand) Nov 10 21:38:31 a6 with buzz fix Nov 10 21:38:55 hm... this seems to be the only difference, though I really doubt the buzz fix has anything to do with it. Nov 10 21:38:55 max_posedon: for manual suspend via syscon menu it is /etc/enlightenment/suspend.sh Nov 10 21:39:03 Qi or uboot? Nov 10 21:39:06 Qi Nov 10 21:39:09 haven't tried suspend via syscon menu Nov 10 21:39:12 qi Nov 10 21:39:24 hm.. Nov 10 21:39:33 spaetz: maybe try that... and I try the mdbus way Nov 10 21:39:33 you practically got the same device ;] Nov 10 21:39:40 local image her, nnot buildhost Nov 10 21:39:57 but stock shr/merge full image Nov 10 21:40:02 max_posedon: which image do you have installed? Nov 10 21:40:17 spaetz, oh... Nov 10 21:40:25 mrmoku/oemerge and just updated Nov 10 21:40:40 max_posedon: does manual suspend work for you? Nov 10 21:40:47 i can't try syscon now without wwaking up wife. fr is in bed room Nov 10 21:40:47 yes Nov 10 21:41:02 haha, seems I never seen two buildhosts creating identical images Nov 10 21:41:04 but I'm not sure does it wake ups or not, because screen was blank Nov 10 21:41:11 max_posedon, from the illume menu? Nov 10 21:41:16 TAsnn, yes Nov 10 21:41:31 root@om-gta02 ~ $ mdbus -s org.freesmartphone.ousaged /org/freesmartphone/Usage org.freesmartphone.Usage.Suspend Nov 10 21:41:36 just does nothing :| Nov 10 21:41:53 DocScrutinizer-8, because of different build times which results in different last change time on files which results in a different checksum ;] Nov 10 21:42:06 bah Nov 10 21:42:09 but you probably were talking about working vs non-working. Nov 10 21:42:09 ;] Nov 10 21:42:17 and I'm just being a smartass. Nov 10 21:42:21 mrmoku, sucks. Nov 10 21:42:21 worked here... atd-over fso? Nov 10 21:42:21 fascinating.. I poked my Glamo chip the wrong way, and now I have noisy horizontal lines (rapidly changing) over everything Nov 10 21:42:29 I just say libasound Nov 10 21:42:47 DocScrutinizer-8, what about the script you promised me? Nov 10 21:42:50 spaetz: yup, kicked atd in favour of atd-over-fso... and even removed apmd Nov 10 21:42:53 :) Nov 10 21:43:17 so you're going to use it? Nov 10 21:43:28 ok, so ousaged suspend is even broken for you. sucks Nov 10 21:43:31 DocScrutinizer-8, only the save part Nov 10 21:43:32 hmm... kicking apmd had the effect that the screen does not blank anymore on suspend :) Nov 10 21:43:37 we already discussed about that ;) Nov 10 21:43:44 that won't work Nov 10 21:43:53 save STATENAME CONTROLNAME Nov 10 21:43:57 i might not havve apmd running Nov 10 21:43:59 i.e Nov 10 21:44:12 but need to check tomorrow Nov 10 21:44:21 yeah, lets check tomorrow Nov 10 21:44:24 snoozie time ;) Nov 10 21:44:33 same here Nov 10 21:44:40 ./save gsmhandset "Mono Playback Volume" Nov 10 21:44:42 have to (finally) bring my father in law to the train station tomorrow morning early :P Nov 10 21:44:51 we'll ship it tomorrow Nov 10 21:44:55 hehe Nov 10 21:45:03 DocScrutinizer, why not? Nov 10 21:45:05 but only if it supsends ;) Nov 10 21:45:12 anway... gnight all :) Nov 10 21:45:16 +y Nov 10 21:45:19 yep, sure Nov 10 21:45:23 nigt Nov 10 21:45:23 TAsn: THERE ARE NO states Nov 10 21:45:29 mrmoku, night. Nov 10 21:45:55 hm, I lied Nov 10 21:46:05 my phone doesn't suspends Nov 10 21:46:10 just turns off screen Nov 10 21:46:13 DocScrutinizer-8, I don't think I quite get you. Nov 10 21:46:24 max_posedon, sucks. :| Nov 10 21:46:36 or wakeuped immidiatly Nov 10 21:46:59 http://pastebin.ca/1665488 Nov 10 21:47:04 last lines of dmesh Nov 10 21:47:39 DocScrutinizer, ? Nov 10 21:47:43 and screen is blank forewer I think Nov 10 21:48:29 mrmoku|away, max_posedon has the same issue as you... Nov 10 21:49:17 but I really also dislike Nov 10 21:49:18 2009.11.10 23:43:28.203 odeviced.kernel26 ERROR Can't read RTC alarm. Returning 0 Nov 10 21:49:30 may be corrent kernel brokes rtc? Nov 10 21:51:29 no idea Nov 10 21:51:30 I'm off. Nov 10 21:51:52 shr-setting's not resasable windows makes ssh -X useless Nov 10 21:52:19 why shr-settings do WM work?! Nov 10 21:53:34 TAsn: what you think about as "scenarios" might change anytime without further notice (e.g. when plugin of headset). This in turn may change meaning and setting of a physical control. I can't and won't implement store of physical control to "scenario". You ask the audiodaemon what's physical control number of an abstract control like "playback master" for *current "scenario"*, and you can't store that later on the way yo Nov 10 21:53:34 u suggested Nov 10 21:54:24 max_posedon, non-resizable? Nov 10 21:54:36 yes Nov 10 21:54:37 that is illume Nov 10 21:55:02 which doesn't offer resizable wwinddows Nov 10 21:55:12 illume as widget toolkit? Nov 10 21:55:22 because my WM is KDE Nov 10 21:55:29 (kwin4) Nov 10 21:55:42 I thought illume is wm Nov 10 21:55:42 ahh Nov 10 21:55:45 DocScrutinizer, I already keep track (actually manage) scenarios Nov 10 21:55:52 in phoneuid Nov 10 21:55:54 while in a call. Nov 10 21:56:03 yes, illume is wm. Nov 10 21:56:15 TAsn: and I won't support that dead end concept Nov 10 21:56:15 I ask for scenarios and change back, so I should always be able to determine what's the current scenario Nov 10 21:58:10 DocScrutinizer, I'm off, we'll talk about it tomorrow, I'm too tired, sorry. ;\ Nov 10 21:59:06 even ffalarms resizable Nov 10 21:59:15 dos1, !!! Nov 10 22:01:04 max_posedon: hmm? Nov 10 22:01:33 max_posedon: illume just doesn't allow resizable windows Nov 10 22:01:35 shr-settings not resable, why? Nov 10 22:01:43 once more, I use kwin4 Nov 10 22:01:43 max_posedon: it has very simple policy Nov 10 22:01:51 max_posedon: oh, that's what you mean Nov 10 22:01:53 and runs a lot of apps through ssh -X Nov 10 22:01:57 max_posedon: hmm... dunno... Nov 10 22:02:05 all resisable, except this Nov 10 22:02:09 max_posedon: i'll have to check it Nov 10 22:33:13 interesting.. Nov 10 22:33:25 I'm getting between 1 and 3 interrupts each time I expected one Nov 10 22:34:02 better than the other thing, I suppose Nov 10 22:36:06 That's getting more for your money! Nov 10 22:36:11 A good deal Nov 10 22:36:19 hehe Nov 10 23:45:31 am I the only one ofr who ffmpeg does not compile? "/ffmpeg-1_0.5.0+r0+svnr20231-r0/trunk/libswscale/swscale.c:462: error: 'PIX_FMT_VDPAU_MPEG4' undeclared " Nov 10 23:45:40 ? Nov 11 00:09:13 the merge branch of shr-u no longer keeps the rootfs directory populated after a Nov 11 00:09:26 successful build (did it ever?) its Nov 11 00:09:47 quite handy - is there a way to make it not cleanup after a good build? Nov 11 00:46:13 Weiss: I always suspected IRQ handling on OM kernel is FUBAR Nov 11 00:47:04 Weiss: I even have proven this suspicion for button IRQs at least, which really are totally insane Nov 11 00:50:59 I couldn't be surprised less to find the weird "trigger on both edges and then *poll*(!!) IRQ line in handler" scheme would create surplus IRQs Nov 11 01:01:15 a long time ago when I discussed that with Andy and asked him why we are using edge instead of level triggered IRQ on FR at all, he implicitly admitted level was the right thing to do, as he stated he tried to change from edge to level trigger and everything slowed down to a grinding halt. So instead of investigating what's wrong with level (and there is a lot that could be wrong if we'd only change hw trigger setup without Nov 11 01:01:16 complete sanitizing of IRQ handler code) we (they / he) simply turned back to edge trigger madness. Nov 11 01:04:16 our kernel will never stabilize due to all the races, losses, and duplications of IRQ waiting there to rise their ugly head Nov 11 01:38:46 DocScrutinizer, isn't moving to .31(and upstreaming) kind of fixing such kind of issues? Nov 11 01:48:42 freesmartphone.org: 03mickey 07cornucopia * r39c725db31c6 10/fsonetworkd/src/plugins/sharing/Makefile.am: fsonetworkd: build fix Nov 11 01:49:56 g'night Nov 11 01:54:45 max_posedon: I heard nothing about IRQ sanitizing (except here a few days ago, where somebody mentioned IRQ scheduled workers or sth) **** ENDING LOGGING AT Wed Nov 11 02:59:56 2009