**** BEGIN LOGGING AT Wed Apr 25 02:59:58 2012 Apr 25 03:49:13 how can you see when a contact was added? Apr 25 09:05:01 hahaha review of nokia lumia 800 Apr 25 09:05:02 pros: windows phone 7.5 Apr 25 09:05:04 cons: windows phone 7.5 Apr 25 09:07:22 very detailed report too Apr 25 09:07:29 rather review Apr 25 09:08:40 heh Apr 25 09:08:45 just the end result Apr 25 09:15:19 well, apparently sales are picking up now Apr 25 09:18:53 the last I want to hear are windows users come screaming in for support on maemo/harmattan just like those who wanted to install maemo on Apr 25 09:28:44 jacekowski: "sales are picking up" is a dubious statement. it just means that one person bought a device. Apr 25 09:39:40 hi all Apr 25 09:40:03 o/ Apr 25 09:40:23 why if i do apt-get remove libmodest-dbus-client, about 44 packages are also to be removed? Apr 25 09:40:35 packages depending on it Apr 25 09:40:42 including hilden-applications-manager Apr 25 09:40:48 yes Apr 25 09:40:52 i understand, that it's because of deps Apr 25 09:41:05 but why that package have so many deps? Apr 25 09:41:25 all the nokia packages are depending on each other Apr 25 09:41:35 isn't libmodest-dbus-client related only to modest email app? Apr 25 09:42:44 modest is the primary email app, if you remove it addressbook brakes if you remove addressbook hildon-desktop brakes if you remove hildon-desktop hildon-home brakes and so on Apr 25 09:42:46 it's strange for me, that removing one modest related package cleans-up all system Apr 25 09:43:33 chem|st, ok, thanks Apr 25 09:43:39 vdv: its like removing totem on a gnome system... you will end up with a server like config Apr 25 09:44:07 lol Apr 25 09:44:31 vdv: for maemo even more strickt Apr 25 09:44:41 stricken... Apr 25 09:44:48 * vdv is happy that he uses slackware on desktop Apr 25 09:45:27 * chem|st is happy to have her sister leave slackware alone for now... Apr 25 09:54:09 vdv: try removing any i18n package... Apr 25 09:54:55 StyXman, no, thanks :) Apr 25 09:55:20 * ruskie has a simple solution for that... dummy packages that install nothing but replace the existing ones Apr 25 09:56:26 heh Apr 25 09:56:36 have a script that generates the relevant stuff Apr 25 09:57:40 http://sprunge.us/FUSO?bash <-- in case someone cares about it Apr 25 09:58:23 aha, there is a maemo channel Apr 25 09:59:33 you pass it what you want(gen_backup or gen_break) and a package name and it will generate either a backup of the currently installed package or an empty package Apr 25 10:01:20 hmm gen_backup function, don't tell me you've reinvented the wheel which dpkg-repack does ;) Apr 25 10:02:08 psycho_oreos, I believe the limitation was it needs to work on vanilla maemo Apr 25 10:02:49 psycho_oreos, which doesn't have dpkg-repack Apr 25 10:03:04 it reinvents a lot of the stuff Apr 25 10:03:07 ruskie, not that I've tried out your script but I wonder what would be the difference, for countless amounts of times I've relied on dpkg-repack when it comes to repacking those installed debs Apr 25 10:03:36 I doubt there would be much Apr 25 10:03:41 maybe a bit more metadata Apr 25 10:04:08 sounds interesting Apr 25 10:04:10 yeah I do know dpkg-repack complains alot about all the extra metadata, usually those x-maemo-icon, etc Apr 25 10:05:16 and frankly I don't really care about reinventing the wheel or not Apr 25 10:05:21 it was a nice learning experience Apr 25 10:05:26 on how to hand craft .debs Apr 25 10:05:50 which I might need for some other project I had on the backburner as well Apr 25 10:08:04 looking at the script, its not just sounds interesting to me, its actually very interesting. It doesn't heavily rely on dpkg-deb which dpkg-repack heavily relies on.. let alone the build functionality out of that dpkg-deb. Reminds me that this could very well be an alternative for me to get gnu-tar installed on my harmattan as that has now become a real pita on pr1.1 Apr 25 10:09:09 well as said... the limitation was vanilla maemo 5 Apr 25 10:11:53 I know I know, just looking at it, that could be something that I could base my bit of work around to make it working on harmattan. Thanks for the link! Apr 25 10:12:02 yw Apr 25 10:12:44 I'll probably/eventually do something similar for rpms as well... will have to see about slackware but might do it as well eventually Apr 25 10:18:32 hi Apr 25 10:58:26 if modest doesn't autoupdate inside of xephyr, does that mean, that on real device that will happen too? Apr 25 10:59:54 * ShadowJK wouldn't count on it Apr 25 11:02:49 ShadowJK, ok Apr 25 11:25:19 does modest permanently delete messages? there's no trash folder, and it doesn't use gmail's trash folder too... Apr 25 11:56:04 http://www.theregister.co.uk/2012/04/25/cook_settle_patents/ <-- heh Apr 25 12:21:28 Solutions stirred with SCP-1067-c continue to stir themselves for up to four (4) hours after the initial stirring, even when obstructed with solid objects. Any liquids that come into contact with SCP-1067-c immediately become oolong tea. Apr 25 12:23:26 RST38h: SCP is sooooo good Apr 25 12:25:59 scp? Apr 25 12:34:53 http://scp-wiki.wikidot.com/ Apr 25 15:11:51 Weird problem: my scratchbox doesn't have dh_make? Apr 25 15:12:10 (debhelper is installed) Apr 25 15:12:49 dpkg -L debhelper | grep -c -E 'dh_make$' ---> 0 Apr 25 15:13:06 there is no dh_make Apr 25 15:13:49 or maybe there is Apr 25 15:14:58 if there is no dh_make then how are we supposed to roll deb packages? Apr 25 15:15:10 I can run it from the host system ofc, but seems odd that it's not in scratchbox Apr 25 15:15:34 My host system has dh_make in a package named dh-make btw, but that's not in the scratchbox Apr 25 15:15:40 it's in different package Apr 25 15:19:04 but i've got no idea where Apr 25 15:20:12 dpkg-dev Apr 25 15:20:12 hmm, no Apr 25 15:20:13 http://wiki.maemo.org/Packaging#Packaging_a_.deb Apr 25 15:20:18 Note: dh_make doesn't seem to be provided by the Squeeze devkit, so you'll have to run it outside of Scratchbox to get a debian/ folder based on debhelper 7. Apr 25 15:20:29 so you have to use older devki Apr 25 15:20:31 devkit Apr 25 15:22:35 Confusing. So which would be better, run dh_make from outside scratchbox, or install debhelper7? Apr 25 15:44:21 bah, I forgot how long Gentoo takes to build glibc :P Apr 25 15:46:29 jonwil, How long would that be? Apr 25 15:46:47 a couple hours Apr 25 15:46:51 on my box anyway Apr 25 15:47:33 Wow. I entertained the idea of switching to gentoo for a bit, but settled for debian sid instead eventually. Seems I made the right choice :) Apr 25 15:48:01 I'm on an Atom powered netbook. What takes hours for normal people to compile takes days here ;) Apr 25 15:48:18 My box is a P4 1.4 Apr 25 15:48:28 with I think 1GB of ram Apr 25 15:48:31 Ah, so not exactly "normal" either Apr 25 15:49:01 http://www.anandtech.com/show/5770/lava-xolo-x900-review-the-first-intel-medfield-phone Apr 25 15:49:04 * SpeedEvil wonders. Apr 25 15:49:41 * ErwinJunge wonders what SpeedEvil wonders Apr 25 15:51:27 If it's open enough Apr 25 15:53:24 I wouldn't hold my breath Apr 25 15:53:53 opensh: operation not permitted. wtf? i don't remember any dev mode updates to be installed recently Apr 25 16:05:38 bah, no-wonder my gentoo compiles take so long, this box only has 512mb of RAM, not the 1GB I thought it had Apr 25 16:05:55 Not that I would be able to get more RAM for the thing though, its too old for that :P Apr 25 16:06:31 bah, its up to package 8 of 258 Apr 25 16:06:57 jonwil, what is your estimate of the total compile duration? Apr 25 16:07:00 no idea Apr 25 16:07:08 hours/days/weeks? Apr 25 16:07:23 probably a day or so depending on how large these packages end up being Apr 25 16:08:04 e.g. if those 258 packages include things like GCC or KDE or QT, its going to take longer Apr 25 16:08:06 ArkanoiD_: don't run binaries from the fat partition. Apr 25 16:09:11 Thankfully I dont have any of the REALLY big packages installed like OO.o or Firefox Apr 25 16:09:40 oh snap, now the thing is telling me I am running out of disk space :( Apr 25 16:15:23 oops, wrong channel :-) it was for #harmattan Apr 25 16:58:35 hello! Can someone help with "Network Based Location Awareness" ? Apr 25 16:59:15 swiftkick: ? Apr 25 16:59:39 r00t|home: ! Apr 25 16:59:56 r00t|home: unable to contact user-defined location server. Using h-slp.mnc410.mcc310.pub.3gppnetwork.org instead. Apr 25 17:00:29 Whenever I use mapping applications, or often times upon booting the phone, I get the phone asking if I want to enable "Network Based Location Awareness". Apr 25 17:00:31 If I enable it, Apr 25 17:00:39 the above error message pops up over... and over... and over Apr 25 17:01:15 I know where the setting is to turn off this feature or put in a user-defined server adress. Apr 25 17:01:37 But when I call my provider, AT&T, they dont seem to know what the hell I'm talking about. }=^> Apr 25 17:02:01 r00t|home: any ideas? Apr 25 17:02:10 Ive been googling on it and not coming up with much. Apr 25 17:03:02 3gppnetwork.org doesnt seem to exist, or at least, it doesnt have a web server at any address I can find. Apr 25 17:04:04 the server is unrelated to your provider in any case Apr 25 17:04:31 the default one is run by nokia, and there was some alternative from another company that's supposedly more reliable Apr 25 17:05:32 as far as i know (aka, wild guess), the system works by gps-enabled phone submitting mappings between gps-location and cell-id, which the server uses to locate cells Apr 25 17:05:46 in any case, you don't need a server provided by your provider Apr 25 17:49:33 jonwil, Had to go for a while, but have one more question about your gentoo on P4 install: Why? Apr 25 17:50:17 I inherited this computer from someone to set up as a Linux box and dabble in Linux Apr 25 17:50:25 they had bought something new Apr 25 17:50:37 so in this case its "free hardware" Apr 25 17:51:01 I meant "why gentoo" :) Apr 25 17:51:13 as for why Gentoo instead of say Ubuntu, I happen to like Gentoo Apr 25 17:51:36 and I like the idea that I can recompile any package I like with different options easily Apr 25 17:51:54 "easily" being very relative in this case ;) Apr 25 17:52:04 it may be slow but it does compile :) Apr 25 17:52:34 and besides, watching large packages compile can be a good way to take your mind off stuff if you are frustrated by something else Apr 25 17:52:58 like not being able to figure out the inner workings of connui-home-cellular :P Apr 25 17:53:35 Haha :) Is the system usable while it's compiling? Apr 25 17:54:55 yeah it sort of is Apr 25 17:54:59 command line at least Apr 25 17:57:36 Hm... so how often do you do these upgrades? Since even with a weekly upgrade it would mean a nearly unusable system for 1/7th of the time :S Apr 25 17:59:17 as a Gentoo dev I don't find my system unusable and I'm building almost continuously! Apr 25 17:59:22 wow Apr 25 17:59:25 nokia drive is great Apr 25 17:59:38 it's not like ovi maps suckiness Apr 25 17:59:52 zeq, what are your specs? Apr 25 17:59:54 it's better than gnav.. you get to download the maps for free :) Apr 25 18:00:23 Macer, is that on n900? Apr 25 18:00:30 ErwinJunge: no. lumia 900 Apr 25 18:00:42 would be nice if they had it for n900 Apr 25 18:00:45 instead of lame ass ovimaps Apr 25 18:01:02 i got lost a bunch of times trying to use ovimaps lol Apr 25 18:01:06 Yup, that's why I asked :) Apr 25 18:01:19 Never got lost, it's just a terrible ui mostly. Apr 25 18:01:25 ovi maps has to be the worst i have ever used Apr 25 18:01:27 yeah heh Apr 25 18:01:31 ErwinJunge: various :) Apr 25 18:01:33 i don't know what they were thinking with that ui Apr 25 18:01:56 but nokia has nokia drive on the ms market Apr 25 18:02:01 lowest I attempted was a PXA270 ;) Apr 25 18:02:06 and it's awesome heh.. i'm surprised they're giving it away for free Apr 25 18:02:20 i guess to compete with google and gnav.. which is awesome as well minus the need for a net connection Apr 25 18:02:22 zeq, I'm on an Atom N450 powered netbook, which is the reason I never tried Gentoo. It seems prohibitively much compiletime. Apr 25 18:03:01 if it's your only system, then yes, that's prohibitive Apr 25 18:03:30 if you have a secondary system with some spare cycles distcc helps a lot Apr 25 18:03:46 lol... i thought the point of gentoo was to make bins that were without a doubt made for your system Apr 25 18:03:56 In practice it is. I also have a C2D desktop standing around, but need to be able to pack up and leave with the pc quickly, which is kind of hard with a desktop ;) Apr 25 18:03:58 in order to make more efficient bins Apr 25 18:04:03 distcc doesn't change that Apr 25 18:04:05 seems like a moot point if you have to build everything heh Apr 25 18:04:39 you only *need* to re-emerge for security fixes Apr 25 18:04:50 all that efficiency is lost in the initial installing of apps Apr 25 18:04:56 Oh, so I could farm out the compiling to my c2d? Apr 25 18:05:00 it's like the ethanol of linux :) Apr 25 18:05:03 invested :) Apr 25 18:06:58 I guess I should dive into it again sometime. Took the time to read through the guide a while ago, but googling on "gentoo compile time on atom" turned me off to the idea ;) Apr 25 18:07:17 yep, on modern laptop/desktop hw it's pretty quick to emerge most things, okay not OpenOffice... Apr 25 18:07:38 Don't like or use openoffice anyway, so that's not an issue for me Apr 25 18:09:17 Btw, current project (which might be of interest to some people here): porting aqualung to maemo. Apr 25 18:10:53 Which is turning out to be relatively easy. I've got an initial version working already and it's also gapless on device (hooray!). Which leaves me with two remaining issues: (1) packaging (2) porting a desktop gtk ui to hildon Apr 25 18:11:08 Any good tips on either heartily appreciated btw ;) Apr 25 18:13:10 has anybody ported nss_mdns to maemo? Apr 25 18:16:08 apt-cache search nss-mdns Apr 25 18:16:09 libnss-mdns - NSS module for Multicast DNS name resolution Apr 25 18:19:35 which is apparently in http://repository.maemo.org fremantle/free Apr 25 18:19:50 version 0.10-3.1maemo1 Apr 25 18:19:56 Anything else you want to know? :) Apr 25 18:23:51 fyi /usr/bin/mdns-scan just built for harmattan Apr 25 18:23:54 oh, wasn't there last time I checked... lets see Apr 25 18:30:33 that's funny, I must be missing a repo Apr 25 18:32:53 Pali, ping Apr 25 18:33:01 freemangordon, pong Apr 25 18:33:29 Pali, what exactly is doing this: Apr 25 18:33:32 twl4030_i2c_write_u8(TWL4030_CHIP_PM_MASTER, 0x02, Apr 25 18:33:32 TWL4030_PM_MASTER_PB_CFG); Apr 25 18:33:42 that is from u-boot rx51.c Apr 25 18:36:25 freemangordon, it enable i2c access to powerbus Apr 25 18:36:42 so then we can set twl4030 regulator mode Apr 25 18:45:47 Isn't http://repository.maemo.org fremantle/free the official maemo repo that should be on all devices at all times? Apr 25 18:47:45 Pali, I tried backporting g-audio from 2.6.31 (first appearance) but got about 300 lines of problems when trying to compile :) Apparently, lot's of stuff changes in 3 kernel releases. Would it be a good idea to simply copy-paste the missing stuff from 2.6.31 into g-audio, or will that likely lead me nowhere? Apr 25 18:50:17 I have it now. Working well :) much easier than remembering local IPv6 addresses :P Apr 25 18:51:35 Haha, it definitely is :) Apr 25 18:51:51 ErwinJunge, backport what is needed Apr 25 18:52:49 Wouldn't that require rebuilding the kernel? I was hoping to just do this as a module. Apr 25 18:52:52 anyone knows, where gtk icons stored physically? where does gtk_icon_theme_load_icon function look up? Apr 25 18:54:31 vdv /usr/share/icons (I think) Apr 25 18:55:42 i've checked that dir Apr 25 18:56:12 icons for that concrete app are not there Apr 25 18:56:13 And your problem is... Apr 25 18:56:15 Ah Apr 25 18:56:38 look under /opt also Apr 25 18:56:42 dpkg -L packagename --> lists files installed by package Apr 25 18:57:05 icons don't come with package Apr 25 18:57:15 i have sources Apr 25 18:57:30 they're not in source tarball too Apr 25 18:57:33 ok, be more specific. what package are we talking about and what icons are you missing? Apr 25 18:57:47 modest Apr 25 18:58:07 i just want to know where icons are located Apr 25 18:58:26 modest uses gtk_icon_theme_load_icon function Apr 25 18:58:33 to fetch icons by name Apr 25 18:58:41 and the name of the icon is... Apr 25 18:59:12 for example, mail_reply Apr 25 19:01:24 [sbox-FREMANTLE_ARMEL: /usr] > find -name '*mail*reply*' Apr 25 19:01:24 ./share/icons/hicolor/48x48/hildon/email_message_reply_all.png Apr 25 19:01:24 ./share/icons/hicolor/48x48/hildon/email_message_reply.png Apr 25 19:01:51 Either of those fit what you mean? Apr 25 19:03:29 ah Apr 25 19:03:51 file name don't correspond icon name in sources Apr 25 19:05:02 ErwinJunge, thanks a lot Apr 25 19:08:14 vdv, no problem :) For future reference, the names used in the sources of modest are defined in src/hildon2/modest-icon-names.h Apr 25 19:08:17 As found by: Apr 25 19:08:28 [sbox-FREMANTLE_ARMEL: ~/modest/modest-3.4.8+0m5] > grep -R 'email_message_reply' * Apr 25 19:08:28 src/hildon2/modest-icon-names.h:#define MODEST_TOOLBAR_ICON_REPLY "email_message_reply" Apr 25 19:08:28 src/hildon2/modest-icon-names.h:#define MODEST_TOOLBAR_ICON_REPLY_ALL "email_message_reply_all" Apr 25 19:08:59 ErwinJunge, there's also email_reply macro Apr 25 19:09:43 grep and find are your friends in this case :) Apr 25 19:09:47 ah, it's src/gtk/* Apr 25 19:09:51 That's in gnome, not hildon :) Apr 25 19:09:56 yep Apr 25 19:10:11 my mistake Apr 25 19:10:22 no problem, happy to help Apr 25 19:10:48 Pali, Wouldn't that require rebuilding the kernel? I was hoping to just do this as a module. Apr 25 19:10:59 maybe yes Apr 25 19:11:20 if that module using something new, then you need to patch kernel Apr 25 19:11:43 if not, then you maybe can compile without modifing kernel Apr 25 19:13:28 would you mind looking at my make output to see if that's the case? I don't feel comfortably patching the kernel. Apr 25 19:13:42 s/comfortably/comfortable/ Apr 25 19:13:46 ErwinJunge meant: would you mind looking at my make output to see if that's the case? I don't feel comfortable patching the kernel. Apr 25 19:15:09 http://pastebin.com/cFviCabj Apr 25 19:17:02 seems, that you need to patch kernel Apr 25 19:18:20 Well, that Apr 25 19:18:49 's off the table then. Seems a bit too dangerous with my zero kernel-mucking experience. Apr 25 19:19:07 Are you accepting feature requests in powerkernel? ;) Apr 25 19:22:28 no time for backporting it Apr 25 19:23:02 if sombody has time and write me patch I can include it in kernel-power Apr 25 19:24:30 Alright then, was a nice idea, maybe I'll get around to it someday. How dangerous would it be? (i.e. can I permanently brick my device?) Apr 25 19:24:45 what? Apr 25 19:29:03 patching the kernel in an attempt to get g-audio working Apr 25 19:30:04 I think that if you do not touch something crytical the worst think will be that you will need to reflash device with flasher-3.5 Apr 25 19:30:52 I booted more times a lot my patches :D and my n900 is still working Apr 25 19:31:21 Ok, but did you have kernel hacking experience before? Apr 25 19:31:22 full reflash and external battery charger fixed all my problems Apr 25 19:31:40 some yes Apr 25 19:32:11 you can always test your kernel in qemu Apr 25 19:32:37 so if you creating something bad it will not boot in qemu Apr 25 19:32:42 Do you have a clear howto on that? Apr 25 19:33:05 on what? backporting g_audio? Apr 25 19:33:16 see errors from compiler Apr 25 19:33:27 and see what structures was changes and how Apr 25 19:33:35 no, trying the kernel on qemu :) Apr 25 19:33:42 install madde Apr 25 19:33:50 madde has n900 qemu image Apr 25 19:33:59 create kernel deb packages Apr 25 19:34:08 I can probably fix the backporting stuff, just scared to permanently destroy my device if I screw up. Apr 25 19:34:10 and install it in madde qemu Apr 25 19:34:49 then reboot qemu (like n900) and see if it boot again :D Apr 25 19:35:12 Ok, thanks for the info. If I have some time to spare I'll probably give it a shot. Apr 25 19:35:14 if not remove madde qemu image, install default and try another build :-) Apr 25 19:36:40 also, in madde you need special kernel module for 3D acceleration (needed by Xserver) Apr 25 19:36:44 source code are on: http://meego.gitorious.org/qemu-maemo/gles-libs/trees/master/kfgles2 Apr 25 19:37:05 default madde qemu image has it for default 2.6.28 kernel Apr 25 19:37:34 but if you build kernel-power, you need to recompile that module and do not remember to install it before reboot Apr 25 19:37:54 without that module madde qemu image will not boot into X and hildon-desktop Apr 25 19:41:54 I guess you mean "do not forget" instead of "do not remember"? Apr 25 19:47:41 Anyway, thanks for the instructions. I copy-pasted them somewhere safe as a howto so that I can give it a shot sometime when I have more time. I'll finish porting aqualung first, since that solves 95% of my problems anyway and is a lot simpler. Apr 25 20:15:43 DocScrutinizer, Would it be possible to automatically switch PSM on/off when connecting/disconnecting a power supply? Apr 25 20:16:29 The reason I'm asking is because this is the way my laptop works and it provides the best of both worlds, great batterylife when off power and great performance when on power Apr 25 20:40:24 ErwinJunge: hmm, quite possibly possible Apr 25 20:43:00 I'm musing along the line: 2 connection settings to same AP (one with, 1 w/o PSM=max), a dbus-script that triggers on USB/charger plug-in/out and switches connection(-priority) Apr 25 20:53:24 Something along those lines, yes. I can hack the dbus thing together in python without problems, but do the psm settings have dbus? Apr 25 21:06:27 dbus there's dbus scripts pkg Apr 25 21:08:56 I didn't suggest to change PSM via dbus (nor any other way) - I suggested to switch from "MyAP-PSM" WLAN connection to "MyAP-USBpowered" WLAN connecton. There are methods to tell ICD2 to switch to a particular connection Apr 25 21:09:10 probably even dbus Apr 25 21:21:34 Ah, so make 2 connections for same network, one with and one without psm and switch between those. Smart. Apr 25 21:52:17 ErwinJunge: http://maemo.org/api_refs/5.0/beta/icd2/group__dbus__api.html#g2d518bd08002c0a8a92c8ba00546360f Apr 25 22:03:13 DocScrutinizer, Thanks for looking that up :) I'll see if I can hack something together based on that tomorrow. Apr 25 22:05:06 installing dbus-script 'app', the whole thing probably consists of two dbus script files with one dbus-send cmdline each **** ENDING LOGGING AT Thu Apr 26 03:00:00 2012