**** BEGIN LOGGING AT Fri Jun 05 02:59:57 2009 Jun 05 05:38:24 morning Jun 05 05:38:38 'morning Jun 05 07:30:34 mrmoku|away o/ Jun 05 07:31:39 lindi-: i know of two other important bugs still present in kernel: one is ASOC related (and a workaround is implemented in frameworkd already), we should try to ask broonie to fix it, the other is battery charging nonsense discovered by DocScrutinizer and there's Werner's analisys on that (not sure who'll be fixing it, probably i'll try to do something when i'll be near my computer again). Jun 05 07:34:27 hey PaulFertser o/ Jun 05 07:38:01 Ainulindale: yo Jun 05 07:41:56 hello there Jun 05 07:43:34 'lo Jun 05 07:44:15 Blu3: i did not found your mail yesterday. would you mind to resend it? Jun 05 07:44:18 lindi-: also a "final touch" needed to avoid _any_ kernel messages (there's still some debugging stuff) on Qi default loglevel. Jun 05 07:44:29 k Jun 05 07:46:11 PaulFertser: what's the ASOC issue? Jun 05 07:48:48 sent Jun 05 07:49:41 thanks Jun 05 07:50:16 thank you Jun 05 07:54:50 PaulFertser: are you aware of the status of touchscreen wakeup kernel project? Jun 05 07:57:22 khiraly1: i can't understand what you're talking about Jun 05 07:57:53 there were some discussion some month ago, that touchscreen could wake up the phone Jun 05 07:58:01 lindi-: "Problems with ASoC and Bluetooth routing" on kernel ML. Jun 05 07:58:05 (it is only possible via the power button now) Jun 05 07:58:24 khiraly1: even if it was possible, i wouldn't want to have it. Button is ok for me. Jun 05 08:03:30 tracfeed: Ticket #505 (Split navit's ipk in 3 parts) updated Jun 05 08:03:49 * PaulFertser away Jun 05 08:25:13 morning....the saveenv in my u-boot aren't saved any more,...want to update then u-boot (dfu-util)...has someone a http:-link for a relative new uboot fro freerunner? Jun 05 08:25:59 I gooled a lot without finding one... Jun 05 08:26:10 googled... Jun 05 08:26:40 or can I buils it with my shr-unstable...? Jun 05 08:28:11 schasch, look at the releases of om2008.x get the latest and it should contain the u-boot bin Jun 05 08:42:09 m0nt0:tankx...ther I found the version 1.3.1...but I thought the newest is 1.3.2...this one I do not find... Jun 05 08:43:20 schasch, i know that u-boot is quite abandoned infact the next gen BL should be Qi Jun 05 08:44:02 so try to understand the differences between the 1.3.1 and the 1.3.2, if they are not relevant flash the 1,3,1 Jun 05 08:45:07 I did...but I change "§boot_menu_timeout to 240" and a saveenv...but it is not save at next booting... .-( Jun 05 08:47:57 schasch, i always used uboot from 1.3.1 i think, the one that comes with that releases, never got problems, the timeout can be adjusted setting an env var Jun 05 08:48:22 m0nt0:in my gta01 I changed it with cu -l /dev/ttyA... and a saveenv...ereything was OK... Jun 05 08:48:57 ah i got gta02, may be something is different Jun 05 08:49:20 m0nt0:I did in my GTA02 ::: setenv boot_menu_timeout 240 Jun 05 08:49:34 m0nt0: printenv to show...it was OK Jun 05 08:49:48 m0nt0:saveenv....he said saving.... Jun 05 08:50:06 m0nt0:but at reboot it was gone... Jun 05 08:50:42 schasch, are you using uboot from NAND, the one without logo right? Jun 05 08:51:15 last time i saw something like that was a friend of mine booting from NOR Jun 05 08:51:43 You should boot nand uboot, not nor Jun 05 08:52:15 mOntO:I did the aux#power buttons thant is NOR or? Jun 05 08:52:50 How can I change the env fron the NAND? Jun 05 08:53:41 schasch, aux then power = NOR, power then aux = NAND Jun 05 08:54:20 oh got....I'll try....I com from gta01 and have no a gta02 too..... .-) Jun 05 08:54:31 so you should do everything with power and after a little moment press aux Jun 05 08:54:50 if the u-boot is without logo then you are ok Jun 05 08:54:58 s/no/now Jun 05 08:55:20 mOntO:thankx...that the solution.... Jun 05 08:55:26 :D Jun 05 08:55:31 no problem Jun 05 09:18:31 raster, !!! Jun 05 09:18:34 please pm me! Jun 05 09:48:44 i've some strange issues with fso - when i enable dimming the display on idle, it does not get bright again, when i'm tapping the screen - however this works, when i _do not_ ignore the accelorometer-event-data Jun 05 09:49:14 heyho Jun 05 09:49:25 which means: by having this line ("ignoreinput = 2,3,4") tapping on screen (which is event1) is not recognized Jun 05 10:02:43 mirko: sounds like a bug Jun 05 10:04:23 mirko: where to put this line? Jun 05 10:38:40 SHR: 03seba.dos1 07shr-settings * r7169335d19e3 10/shr_settings_modules/shr_splash.py: [splash] add module to select shr-splash theme Jun 05 10:38:40 SHR: 03seba.dos1 07shr-settings * r385c681ebab5 10/shr-settings: [core] Others: remove Test, add Splash Jun 05 10:56:37 mrmoku: ping Jun 05 11:02:57 dos1|neo, pong Jun 05 11:03:31 klacl klack. klack klack. mrmoku 1:0 dos1|neo Jun 05 11:03:45 dos1|neo, pong Jun 05 11:03:46 dos1|neo, pong Jun 05 11:03:46 dos1|neo, pong Jun 05 11:04:03 4:1 Jun 05 11:04:26 game over... mrmoku won ;) Jun 05 11:04:32 dos1|neo, brb Jun 05 11:04:33 mrmoku: ping ping ping ping Jun 05 11:04:48 hehe Jun 05 11:04:53 * mrmoku won't rebuild anything :P Jun 05 11:04:59 mrmoku: how to easly send oe commits as mails? Jun 05 11:05:07 s/oe/oe git/ Jun 05 11:05:07 dos1|neo meant: mrmoku: how to easly send oe git commits as mails? Jun 05 11:05:26 i' playing with shr-splash recipes Jun 05 11:05:32 s/i'/i'm/ Jun 05 11:05:33 dos1|neo meant: i'm playing with shr-splash recipes Jun 05 11:05:35 dos1|neo, in 10 minutes? brb Jun 05 11:05:40 mrmoku: ok Jun 05 11:15:53 dos1|neo, ok Jun 05 11:16:06 dos1|neo, you should add a local branch for your changes Jun 05 11:16:17 dos1|neo, then commit to it Jun 05 11:16:21 dos1|neo, TEST it ;) Jun 05 11:16:49 dos1|neo, and then you can do 'git format-patch shr/import..your-local-branch Jun 05 11:16:53 ' Jun 05 11:17:04 which will give you a series of patches... one patch per commit Jun 05 11:17:33 dos1|neo, or you can let git send the patches directly Jun 05 11:20:02 dos1|neo, not sure how to do that... ptitjes did that Jun 05 11:21:29 SHR: 03seba.dos1 07shr-themes * r82ead6296338 10/e-wm/e-wm-illume-dict-pl/Polish.dic: e-wm-illume-dict-pl: initial commit of polish dictionary for Illume keyboard Jun 05 11:21:53 dos1, do you have 'git send-email' command? Jun 05 11:22:16 mrmoku: git: 'send-email' is not a git-command. See 'git --help'. Jun 05 11:22:56 dos1, yeah... don't have that one too... Jun 05 11:23:11 it is mentioned in the man page of git-format-patch though Jun 05 11:23:56 dos1, ahh, ok. extra package Jun 05 11:24:01 yum install git-email :-) Jun 05 11:24:25 aptitude install git-email :D Jun 05 11:25:52 dos1, ok... then do as described above and then 'git send-email your-patch Jun 05 11:25:55 ' Jun 05 11:26:11 man git-send-email tells you all ;) Jun 05 11:28:07 mrmoku: thanks :) Jun 05 11:28:14 dos1, np :) Jun 05 11:28:19 SHR: 03seba.dos1 07shr-themes * r79cd79c1ca29 10/e-wm/e-wm-illume-dict-pl/Polish.dic: e-wm-illume-dict-polish: convert from DOS to Unix format Jun 05 11:39:46 *brrr* moin Jun 05 11:56:32 shr script for deleting all sms's? Jun 05 11:56:41 mdbus -s something something.................... Jun 05 11:57:31 badloud: you can delete all smses in shr-settings with single click ;) Jun 05 11:57:58 duh!? Jun 05 11:58:18 Settings -> Phone -> Message book -> clean Jun 05 11:58:26 wow :-D Jun 05 11:58:40 haha Jun 05 11:59:46 hope you got a safety nag-requester! Not going to test that now ;-D Jun 05 12:02:44 wha?? Jun 05 12:02:48 sewiously? Jun 05 12:02:52 DocScrutinizer: that wasn't me who implemented that ;) Jun 05 12:03:25 badloud: take out from your mouth that carrot! Jun 05 12:03:37 ha Jun 05 12:03:38 done Jun 05 12:03:38 dos1: whom to spank then? ;-) Jun 05 12:03:49 well, look at that Jun 05 12:03:50 DocScrutinizer: yoyo Jun 05 12:03:52 ~seen yoyo Jun 05 12:03:53 dos1: i haven't seen 'yoyo' Jun 05 12:03:58 he wasn't pulling my leg Jun 05 12:03:58 ~seen yoyoek1 Jun 05 12:03:59 dos1: i haven't seen 'yoyoek1' Jun 05 12:04:02 ~seen yoyoek Jun 05 12:04:03 i haven't seen 'yoyoek', dos1 Jun 05 12:04:05 ~seen yoyo1 Jun 05 12:04:06 dos1: i haven't seen 'yoyo1' Jun 05 12:04:07 hmm Jun 05 12:04:18 ~seen God Jun 05 12:04:18 i haven't seen 'god', badloud Jun 05 12:04:21 k, he's hiding :-) Jun 05 12:04:23 I have Jun 05 12:04:29 (joking) Jun 05 12:04:38 dossie Jun 05 12:05:01 got me a kveshchun for ya Jun 05 12:05:46 badloud: carrot!!! Jun 05 12:07:41 ok, ok Jun 05 12:07:47 a 'question' Jun 05 12:07:56 /msg nickserv rename badcloud bugsbunnie Jun 05 12:08:24 very clever... Jun 05 12:08:31 dos1, you free? Jun 05 12:09:45 ~seen a_stable_release_of_stable_hybrid_release_somewhere Jun 05 12:09:45 badloud: i haven't seen 'a_stable_release_of_stable_hybrid_release_somewhere' Jun 05 12:10:14 badloud: it depends what you mean under "stable" Jun 05 12:10:23 badloud: unstable is stable enough for daily use ;) Jun 05 12:10:31 dos1, no it's not. Jun 05 12:10:35 then why dub it unstable? Jun 05 12:10:40 badloud: but we didn't release shr-stable image right now Jun 05 12:10:47 and I agree with TAsn Jun 05 12:10:53 I'm still using testing for that reason Jun 05 12:10:54 dos1, there's the dbus bug Jun 05 12:11:00 oh, yep Jun 05 12:11:04 forgot about it :X Jun 05 12:11:07 and the call waiting bug Jun 05 12:11:13 oemerge_20090529 is ok Jun 05 12:11:14 both are the real blockers Jun 05 12:11:26 and there's opkg... Jun 05 12:11:30 which segfaults Jun 05 12:11:30 bahhh Jun 05 12:11:31 :) Jun 05 12:11:39 I think I tried that, Doco Jun 05 12:11:39 dbus bug not yet fixed, for any distro Jun 05 12:11:50 I had some problem with outgoing sound qual Jun 05 12:11:51 TAsn: opkg segfaults in every distro :P Jun 05 12:11:55 opkg install pidgin-data Jun 05 12:11:55 Upgrading pidgin-data on root from 2.5.3-r0 to 2.5.3-r1... Jun 05 12:11:55 Downloading http://build.shr-project.org/shr-unstable/ipk//armv4t/pidgin-data_2.5.3-r1_armv4t.ipk Jun 05 12:11:55 opkg: fork failed Jun 05 12:11:55 Segmentation fault Jun 05 12:12:01 and there were other problems Jun 05 12:12:03 dos1, which only means no distro is stable Jun 05 12:12:07 not that shr is. Jun 05 12:12:15 dos1: nonsense Jun 05 12:12:27 omg Jun 05 12:12:30 /etc/init.d/xserver-nodm stop Jun 05 12:12:32 after a few retries Jun 05 12:12:36 it works Jun 05 12:12:36 swapon /swapfile Jun 05 12:12:39 opkg upgrade Jun 05 12:12:40 :P Jun 05 12:12:40 used opkg to install some pkgs Jun 05 12:12:45 dos1, too much work ;) Jun 05 12:12:46 no segfault Jun 05 12:12:59 DocScrutinizer: it segfaults every time it goes out of memory :P Jun 05 12:13:05 and it goes every bigger upgrade Jun 05 12:13:06 sure Jun 05 12:13:18 because for some reason Jun 05 12:13:23 who ever programmed it Jun 05 12:13:28 opkg upgrade prolly does Jun 05 12:13:28 added a fork bomb feature ;) Jun 05 12:13:42 muahahaaa Jun 05 12:14:15 dos1, what format does this create? 'mdbus -s org.freesmartphone.ogsmd /org/freesmartphone/GSM/Device org.freesmartphone.GSM.SIM.RetrievePhonebo ok 'contacts' > /media/mmcblk0p2/addressbook1.vcf' Jun 05 12:14:48 (disregard the .vcf, wishful thinking) Jun 05 12:14:54 aaah Jun 05 12:15:00 badloud: just tuples with contacts Jun 05 12:15:09 tuples? Jun 05 12:15:17 I'd like to have a vcard of my sim contacts Jun 05 12:15:31 badloud: ['name':'blabla', 'phone':'3535345'] Jun 05 12:15:34 something like tha Jun 05 12:15:35 t Jun 05 12:15:37 right Jun 05 12:15:43 how do I convert that to vcf? Jun 05 12:15:46 it isn't any format Jun 05 12:15:48 write script :P Jun 05 12:15:53 or find some Jun 05 12:15:54 badloud: use sed/awk? Jun 05 12:16:35 ummm, I'll have to google them Jun 05 12:16:39 thanks Jun 05 12:17:19 programming languages? Jun 05 12:17:26 DocScrutinizer: what do you mean? Jun 05 12:17:36 wayyyy out of my league Jun 05 12:17:48 wikipedia 'AWK is a programming language' Jun 05 12:17:58 dos1: what about additional fields for pim-contacts? like e.g. "address" ;-D Jun 05 12:18:00 sed isn't Jun 05 12:18:34 DocScrutinizer: talking about cleaning or backuping? Jun 05 12:18:42 so how would I use sed to convert mdbus's output to vcf?? Jun 05 12:18:53 sorry for the complete lack of ability on my part Jun 05 12:19:08 dos1: about having additional fields in e.g. SQL backend - and screen of course Jun 05 12:19:29 DocScrutinizer: then don't use ogsmd to request that ;) Jun 05 12:19:39 huh? Jun 05 12:19:52 I use "contacts" Jun 05 12:20:15 "new" Jun 05 12:20:30 DocScrutinizer: uh, where? Jun 05 12:20:37 still just "name:" & "number:" Jun 05 12:20:40 there is no opimd integration right now Jun 05 12:20:46 SHR? maybe? Jun 05 12:20:56 FSCK Jun 05 12:20:59 only in libframeworkd-phonegui-efl2, but that gui allows only for name and number too Jun 05 12:21:31 for testing i'm using my small python script, pim.py Jun 05 12:22:08 so, mdbus's output won't import into hackable1? Jun 05 12:22:11 * DocScrutinizer kicks his FR Jun 05 12:22:34 wtf is hackable?? Jun 05 12:22:42 never heard of ;-) Jun 05 12:22:54 ha Jun 05 12:22:55 does it do "address:"? Jun 05 12:22:57 DocScrutinizer: kind of cross-compiling debian packages ;) Jun 05 12:23:03 seriously though Jun 05 12:23:07 I'm itching to use it Jun 05 12:23:18 hmm, i have a silly question. a while back, stefan mentioned that SHR wasn hoping to be rid of the distribution maintenance, but with openmoko not doing it, that task was falling back on them. this made me wonder: with SHR and FSO focusing on the actual content of the distribution, couldn't that task of actually maintaining the distribution itself be "outsourced" to the debian or gentoo folks ? (debian may have a larger following Jun 05 12:23:18 , but i'm not sure) Jun 05 12:23:21 but not all of my sim contacts are shown Jun 05 12:23:46 DocScrutinizer: haven't you been at f9n ? :) Jun 05 12:23:55 so what? Jun 05 12:24:17 DocScrutinizer: i think hackable was mentioned quite a bit ;-) Jun 05 12:24:22 bah Jun 05 12:24:26 wpwrak: SHR will do the distro stuff. FSO will stop producing own images and focus on the code. Jun 05 12:24:29 developers crap Jun 05 12:25:48 I think SHR is just enough hackable for me ;-) Jun 05 12:26:24 stefan_schmidt: so the distro stuff isn't a burden in the end ? i.e., if the debian folks helped, that wouldn't be much of a help ? or are there deeper reasons why this wouldn't work ? Jun 05 12:27:25 wpwrak: debian has a complete different idea about a distro. Well suited for computing devices, but not for a phone imho. Jun 05 12:27:48 wpwrak: But I won't discuss this. It's my opinion. Jun 05 12:27:50 kernel-issue: Think we need ramconsole (or whatever it's called) by default. As kernel panics got down in frequency, and aren't exactly reproducable any more, we need each report to get to a goal of ZERO panics Jun 05 12:28:04 * mrmoku agrees with stefan_schmidt :) Jun 05 12:28:30 * DocScrutinizer too Jun 05 12:28:33 * stefan_schmidt loves debian on all his working machines but not on his phone. Jun 05 12:28:44 Anyway, /me is away for a bit. later. Jun 05 12:29:25 wpwrak, on FSOSHRUDCON we agreed upon a 'freesmartphone' branch in OE which should be used by all OE and FSO based distros Jun 05 12:29:27 stefan_schmidt: i'm just curious :) Jun 05 12:30:21 stefan_schmidt: plus, i may have to answer the "why not debian" question myself. so i'm looking for something better than "uh, historical reasons" or "personal preferences". Jun 05 12:31:23 DocScrutinizer: yeah, what happened with that little project ? it looked quite sensible but then somehow just vanished from the radar Jun 05 12:31:45 wpwrak: well, try it out. Perhaps it suits your needs. It's great for developing. But for a highly customized image debian does not work well. (IMHO) Jun 05 12:31:50 * stefan_schmidt is really off Jun 05 12:32:37 wpwrak: afaik it works, but isn't in [default] Jun 05 12:33:24 wpwrak: PaulFertser should know about it a little. For sure others (e.g. mrmoku) as well Jun 05 12:33:24 mrmoku: so ... do you know where the debian problems creep in ? are the dependencies too greedy ? using ubuntu, which is probably worse in terms of technical purity, i find the degree of customization quite acceptable. i do notice that it installs a fair amount of things that i don't think i need. Jun 05 12:34:20 wpwrak, I installed debian once a long time ago on my FR... so I'm certainly not an expert regarding debian on FR Jun 05 12:34:42 as far as I understand it... it is using official debian precompiled packages Jun 05 12:34:45 mrmr Jun 05 12:34:54 mrmoku: ramconsole? Jun 05 12:34:58 which by definition eliminates the possibility to tweak stuff to your needs Jun 05 12:35:07 DocScrutinizer, no idea about ramconsole... never heard of it :( Jun 05 12:35:16 aah, ok Jun 05 12:35:33 wpwrak, in contrast with OE you can add patches to packages depending on what type of HW you build for Jun 05 12:35:49 PaulFertser: ping? Jun 05 12:36:33 mrmoku: hmm, what sort of patches would that be ? separating man pages and such ? i.e., is it the package granularity ? Jun 05 12:36:33 hmm, weekend :-S Jun 05 12:36:51 DocScrutinizer: now that you mention it ... Jun 05 12:37:24 bah, you're not even awake - are you? Jun 05 12:38:24 BUE: 09:38, Moskow: 15:38 Jun 05 12:38:37 DocScrutinizer: just got up and having my first coke. so no, not really awake Jun 05 12:38:49 what I thought Jun 05 12:39:03 * DocScrutinizer away Jun 05 12:45:53 tracfeed: Ticket #504 (The revenge of the segfault - contacts + messages (but calls still ...) updated Jun 05 12:50:41 tracfeed: Ticket #504 (contacts and messages segfault (but calls are still possible)) updated  || Ticket #504 (contacts and messages segfault (but calls are still possible)) updated Jun 05 13:00:44 tracfeed: ophonekitd.2.log attached to Ticket #504  || frameworkd.2.log attached to Ticket #504 Jun 05 13:02:44 tracfeed: Ticket #504 (contacts and messages segfault (but calls are still possible)) updated Jun 05 13:29:58 mrmoku|away: Ainulindale: (ramconsole for default PLEASE) http://docs.openmoko.org/trac/ticket/2135 Jun 05 13:30:09 lindi-: what's current sttaus of this? Jun 05 13:30:36 wpwrak: ^^ to you as well Jun 05 13:33:37 DocScrutinizer: sorry for jumping without knowing all details but can't we use mtdoops instead ? Jun 05 13:33:42 DocScrutinizer: unknown. but i think it's a good idea. tried to do the same for x86 something like 16 years ago, only to be defeated by the BIOS clearing/messing up all memory on reboot Jun 05 13:37:00 rtp: no idea Jun 05 13:37:16 wpwrak: well according to lindi it seems to work Jun 05 13:38:31 and for sure we need it - e.g. I got a kernel panic yesterday. But no idea how to help fix it. No way to reproduce as it came "out of nothin" Jun 05 13:41:08 for random panics neither debubo nor any ticket will help, and I think eventually we want to get rid of these random panics as well. Will never happen if we don't get sth like ramconsole for default Jun 05 13:42:46 DocScrutinizer: ok. as it's storing the oops on some flash partition it survives poweroff which is nice but you need to have some place in a flash. I fear the flash is full on the FR :( Jun 05 13:43:04 * DocScrutinizer shudders thinking about all the hidden races we may still have on kernel, wrt suspend/resume, IRQ handling etc pp Jun 05 13:44:16 rtp: not exactly, there is some space left on mtd-partitons, e.g. factory Jun 05 13:44:32 rtp: i think the advantage of not using flash is mainly that it's simpler to just scribble into ram. the less infrastructure you need to store your crash information, the better. Jun 05 13:44:47 ~seen Jun 05 13:44:49 DocScrutinizer: it's not used to store some things like mac address ? Jun 05 13:45:04 ~seen mickeyl Jun 05 13:45:04 mickeyl was last seen on IRC in channel #openmoko-cdevel, 4d 23h 13m 42s ago, saying: 'bbl'. Jun 05 13:45:05 yup it is Jun 05 13:45:26 still I think there's plenty of free space Jun 05 13:45:42 wpwrak: ram content does not survive a power off so it may be annoying to use ram Jun 05 13:45:46 ~seen mickey|sun Jun 05 13:45:46 mickey|sun is currently on #openmoko-cdevel, last said: 'argh, yet another channel? :)'. Jun 05 13:45:55 mmmh Jun 05 13:45:58 thanks Jun 05 13:46:03 rtp: you can find free space at a few places, but it's generally a bit messy. things like end_of_partition_X - N_non_bad_blocks or such Jun 05 13:46:19 rtp: nand just sucks ;) Jun 05 13:47:12 wpwrak: keep a dummy file, that points directly to the blocks to use? Jun 05 13:47:35 keep it open all the time, meme-mapped or sth? Jun 05 13:48:11 wpwrak: just an idea of the moment Jun 05 13:48:38 DocScrutinizer: been there, done that ;-) (FIBMAP) too complex, particularly with jffs2 migrating blocks. Jun 05 13:48:51 ouch Jun 05 13:49:00 wpwrak: mtdoops already helped me fix some kernel bugs so that's why I was thinking about it. You're free to choose whatever you prefer :) Jun 05 13:50:24 ramconsole works and has been recommended several times to spot panics. Just saying we should get it on SHR for default Jun 05 13:50:26 rtp: i think just a ram-based approach is the easiest solution. if you need to save things to NAND, you can pick it up from RAM upon the reboot and then just save with regular user space tools. or you could use the AF447 approach and send it out over the air ;-) Jun 05 13:50:55 huh? :-D Jun 05 13:51:45 (there are also outright crazy approaches like kdump. and they even seem to work, which i find very cool. but we don't have to go there ;-) Jun 05 13:51:58 wpwrak: so you need a tool to copy the ram on reboot. with mtdoosp, you don't need such a tool Jun 05 13:52:06 DocScrutinizer: the list of advantages and shortcomings in the bug report should be accurate. I just don't have time/skills to make ramconsole more suitable for inclusion at least right now Jun 05 13:52:21 kdump/kexec may work. should be funny to use on the neo :) Jun 05 13:52:38 DocScrutinizer: the plane that crashed in the atlantic a few days ago. during its descent some automatic monitoring systems sent some problem reports via satellite. Jun 05 13:52:39 rtp: I am still waiting for kdump source code for arm Jun 05 13:52:55 rtp: a guy was porting it to arm in a company and said he could not release the source yet Jun 05 13:52:58 aaaah, this one yup Jun 05 13:53:42 rtp: I did investigate all available options I could find before writing ramconsole patch :P Jun 05 13:53:43 lindi-: oh ? I though it was working on arm :( Jun 05 13:53:49 rtp: it's not Jun 05 13:53:55 rtp: kexec works, kdump not Jun 05 13:53:57 wpwrak: didn't I mention autopilot? would expect some similar issue here Jun 05 13:54:35 hmm, can in-reply-to have a list of message-ids ? *scratch, scratch* Jun 05 13:54:49 lindi-: where's the problem to get it for default on SHR? Jun 05 13:55:17 lindi-: writing kdump on arm would be too hard ? Jun 05 13:55:39 DocScrutinizer: did you read the points at http://docs.openmoko.org/trac/ticket/2135#comment:21 ? Jun 05 13:55:44 rtp: apparently Jun 05 13:55:52 yup, quickread Jun 05 13:55:53 DocScrutinizer: err, the plane crash ? naw, still very unclear what happened. they haven't even found the right wreckage yet, just some unrelated flotsam. Jun 05 13:56:15 rtp: besides, that'd waste ram. you would need to have the dump kernel in ram all the tmie Jun 05 13:56:54 lindi-: sure but I though you're reserving some ram in the ramconsole too. right ? Jun 05 13:57:16 rtp: yep but log messages consume significantly less than a complete kernel? Jun 05 13:57:16 wpwrak: (airbus) mishap started with autopilot deactivation, last auto-dbg-msg was pressure loss of cabin Jun 05 13:57:31 DocScrutinizer: i think those shortcomings make it unsuitable for being on by default Jun 05 13:57:50 DocScrutinizer: it'll silently corrupt data if you forget mem=127M for example Jun 05 13:58:03 lindi-: could you elaborate? I don't exactly get it Jun 05 13:58:19 DocScrutinizer: it uses the last megabyte as a buffer right now Jun 05 13:58:32 DocScrutinizer: so kernel needs to be told not to use it for anything else... Jun 05 13:58:34 of RAM, right? Jun 05 13:58:38 DocScrutinizer: yep Jun 05 13:59:06 ok, So we need uBoot env setting then, right? Jun 05 13:59:12 lindi-: probably. kernels logs are 10k to 20k vs several Mo for kernel Jun 05 13:59:28 rtp: yep, and we are talking about something that'd be enabled by default for everyone Jun 05 13:59:46 DocScrutinizer: for example, also nor uboot could not used to boot anymore Jun 05 14:00:06 DocScrutinizer: some kernel guru could probably rewrite ramconsole in a much more sane way Jun 05 14:00:07 what's going to happen if we did? Jun 05 14:00:10 DocScrutinizer: yes, but a) nobody knows if those events were cause or effect and b) timing and even order aren't guarenteed. there are about 2500 postings at http://www.airliners.net/aviation-forums/general_aviation/ if you're interested in the details :) Jun 05 14:00:29 DocScrutinizer: did what? Jun 05 14:00:40 use nor Jun 05 14:00:55 or boot without ram=127 Jun 05 14:01:00 lindi-: i suppose you could just detect if the last MB is available ? Jun 05 14:02:02 lindi-: i briefly looked for a suitable allocator for the gta03-6410 camera driver, but didn't spot anything. so the mem=xxx hack may indeed be the best solution. Jun 05 14:02:14 or have a userspace tool to allocate that 1MB Jun 05 14:02:38 aaah, ok Jun 05 14:03:23 SHR: 03seba.dos1 07shr-settings * rb908e07c6fb1 10/shr_settings_modules/shr_clock.py: Revert "[clock] use fso api instead of calling directly date. Patch by Michael Scherer (misc)" Jun 05 14:03:33 SHR: 03seba.dos1 07shr-settings * r3c87e8e47946 10/shr_settings_modules/ (shr_clock.py shr_splash.py): [splash][clock] little changes in comments Jun 05 14:03:39 DocScrutinizer: user space tool won't help Jun 05 14:03:50 well, don't we have a rather determined amount of RAM on FR? ;-) So what about a dirty kernel patch? Jun 05 14:03:56 wpwrak: i think there's some way to allocate memory very early Jun 05 14:04:03 wpwrak: on x86 there is :) Jun 05 14:04:28 DocScrutinizer: yes the kernel patch could add mem=127M to be a default option too :) Jun 05 14:04:42 exactly Jun 05 14:05:26 so, I don't think we can't include it to default for that reason Jun 05 14:05:57 actually even patching uBoot env should be easy Jun 05 14:06:05 DocScrutinizer: ok, maybe you could. at least it's stable Jun 05 14:06:06 lindi-: I *thought* there was. so i was a bit surprised that i didn't see it. but maybe i didn't search well enough. Jun 05 14:06:30 wpwrak: i looked very hard too Jun 05 14:07:01 wpwrak: is there some way to read ram from userspace? I think /dev/mem failed somehow here Jun 05 14:07:37 since if there was a way to read the ramconsole buffer from userland then it would not need to dump the messages with printk on boot which would fix issuse #3 and #4 Jun 05 14:08:19 lindi-: could you take a little care to push this further? Jun 05 14:08:20 wpwrak: actually that'd fix #6 too since you could implement the logic nicely in userland Jun 05 14:09:07 lindi-: I think we really need this in default Jun 05 14:09:52 DocScrutinizer: i'll try to find time. i only have one phone and hooking it to debug board means I can't really use it for calls. also i don't have extra partitions set so if the primary partition does not boot with Qi I need to recover by booting from flash Jun 05 14:10:20 I should also start doing my master's thesis, one more excuse :) Jun 05 14:11:19 well, I see. maybe find other ppl to do actual tests ;-), and ask some of the SHR-guys to help on including it to SHR-default Jun 05 14:11:56 maybe i should start using kexec to test kernels Jun 05 14:12:04 lindi-: you can read ram from user space. see svn.openmoko.org/developers/werner/poke/. you just mmap /dev/mem and there's not limit to what you can do to the memory ;-) Jun 05 14:13:02 lindi-: doing this is the excuse for not starting the master's thesis, i suppose ? or have students changed all that much lately ? ;-) Jun 05 14:13:12 wpwrak: ko so plain read() won't work but I need to mmap it? Jun 05 14:13:41 wpwrak: well i only started to evaluate if the topic is right this week Jun 05 14:13:42 wpwrak: might we try to push this a little (i.e. could you (continue to ) mentor/help)? Jun 05 14:19:01 DocScrutinizer: well, you can catch my attention here between mails :) Jun 05 14:19:07 wpwrak: hmm, i would have wanted to test that program but I can't start wifi (hit bug #2277 Wireless does not work with the 2.6.29 kernel) Jun 05 14:19:55 lindi-: i think read should also work, but i generally find that mmap gives me more control. e.g., if i need to do a byte read, will read(1) really do that for me ? Jun 05 14:25:29 lindi-: wpwrak: AIUI we don't need anything more than just poke, to create the dump after boot? Jun 05 14:26:20 sure it's not optimized and probably dead slow Jun 05 14:26:27 so what...? Jun 05 14:27:20 yep those ingredients should be enough Jun 05 14:28:14 also PaulFertser told me You know of decent ways to mod ubbot env from userland Jun 05 14:28:24 DocScrutinizer: sure but not NOR uboot Jun 05 14:28:43 DocScrutinizer: i'm using Qi now for only one reason. it doesn't talk to glamo Jun 05 14:28:57 so I guess we got everything for inclusion of ramconsole to default Jun 05 14:29:18 "rmmod ar6000" causes kernel oops :) Jun 05 14:29:56 and maybe we even could talk some kernel crack into writing a dirty patch to have mem=127 for default Jun 05 14:30:24 PaulFertser: ^ ;-) Jun 05 14:31:10 ok "ifconfig" won't work anymore. time to reboot :( Jun 05 15:29:02 excuse me which is the difference between shrUNSTABLE and shrOEMERGE? Jun 05 15:29:07 can anyone tell me? Jun 05 15:33:00 oh yea, I'd like to know as well ;-) Jun 05 15:33:59 why does the oemerge directory still exists? Jun 05 15:37:04 Heinervdm: maybe to get a hold on oemerge-20090529 which is considered semi-stable? Honestly not sure about the policy of SHR wrt unstable|oemerge|testing. Suggested to move oemerge-20090529 to testing some 2 days ago. Ainulindale promised SHR-crew would provide a semi-stable release (like afore mentioned 0529) last night (or was it 2 nights ago?) Jun 05 15:38:19 DocScrutinizer: as i know shr-oemerge was renamed to shr-unstalbe 4 or 5 days ago... Jun 05 15:38:41 no idea, really. Jun 05 15:39:47 I just know testing-20090416 was quite stable, but bitrot in less than a week so nobody was able to dl a semi-decent image anymore Jun 05 15:41:00 but unstable now it's usable? Jun 05 15:41:07 I still got no clue about policy, as mentioned above Jun 05 15:41:23 and there is still a problem with wifi Jun 05 15:41:33 after suspend it doesn't work Jun 05 15:42:05 and the problem with HAL and the Battery applet Jun 05 15:42:13 pbaxter: unstable never is useable - by definition Jun 05 15:42:24 but it was... Jun 05 15:43:42 as soon as it is considered "useable" it should move to testing immediately - yet development should continue on unstable, never on testing Jun 05 15:45:23 so Docscrtinizer, you think now it's unusable? Jun 05 15:48:53 DocScrutinizer: something like http://bjorn.haxx.se/debian/testing.pl?package=gcc is also very useful Jun 05 15:50:07 maybe. dunno this URL Jun 05 15:51:38 DocScrutinizer: yep it's unofficial script that just has links to the relevant official pages Jun 05 15:51:49 pbaxter: no idea. see: I don't install unstable 3 times a day, doing power testing rest of my time Jun 05 15:52:26 you do power test? Jun 05 15:52:59 I don't!! Jun 05 15:53:10 got better things to do Jun 05 15:54:12 e.g. do in deep tests (well sorta) of oemerge20090529 Jun 05 15:54:54 rather than start tests of unstable over and over on basic things, 3 times a day Jun 05 15:55:29 after all that's job of the developer to do this kinda test Jun 05 15:55:39 ok excuse me DocScrutinizer Jun 05 15:55:46 i'm tto much questioner man+ Jun 05 15:55:55 excuse me( Jun 05 16:01:11 no reason for excuses :-) Jun 05 16:06:55 DocScrutinizer: kernel crack? ;) Jun 05 16:07:12 hehe - sure ;D Jun 05 16:08:03 do you think you could hack a "mem=127" into kernel, for kernel startup parameters? Jun 05 16:10:05 of course it has to be override-able by any actual "mem=" on kernel cmdline Jun 05 16:11:11 PaulFertser: I guess you're just reading backscroll? Jun 05 16:11:51 PaulFertser: sorry, have to run Jun 05 16:11:56 bbl Jun 05 16:23:22 DocScrutinizer: over gprs through ssh. not exactly easy Jun 05 16:31:10 wpwrak: thanks a lot for the mmap hint. I can now read the buffer from userland Jun 05 16:31:22 wpwrak: i'll write a tool to dump the contents of ramconsole Jun 05 16:33:01 lindi-: kewl, glad to have been of use :) Jun 05 16:41:42 tracfeed: Ticket #493 (Set Date/Time in shr-settings doesn't work) closed Jun 05 16:54:46 dos1: hmm, we got *date* now as well ? ;-) Jun 05 16:55:36 DocScrutinizer: no, there is another ticket for that Jun 05 16:55:38 ;) Jun 05 16:56:37 "date/time isn't actually DATE and time" ? :) Jun 05 16:56:45 ok, no matter which ticket, I suggested to have a TZ selector on Date/time as well :-) Jun 05 16:58:07 and goddamn kick the "automatic gsm-based TZ setting" off for default. It constantly borks Jun 05 16:58:21 opkg: fork failed Jun 05 16:58:21 Segmentation fault Jun 05 16:58:31 ahh, we have such a special shortbus too Jun 05 16:58:32 l Jun 05 16:58:36 => Jun 05 16:59:11 i like my auto gsm TZ. it's either +2400 or +0000, neither of which are anything close to correct Jun 05 16:59:39 heh, cool ;-D Jun 05 17:00:48 so obviously it takes much less experienced user to enable this auto-TZ, than it needs for fixing borked auto-TZ setting Jun 05 17:01:12 so for default we shouldn't enable this freakin option Jun 05 17:01:46 auto gsm TZ works ok for me ;> Jun 05 17:01:55 and "date/time" is name of category Jun 05 17:01:59 module is "clock" Jun 05 17:02:02 so you're wrong ;) Jun 05 17:02:12 bah, for me it borked when swapping SIM Jun 05 17:03:19 caption is "Date/time", so I'm right Jun 05 17:04:02 nifty, my fone done locked up Jun 05 17:04:28 DocScrutinizer: /me is checking Jun 05 17:04:31 and though caption mentions "Date" there is no date anywhere Jun 05 17:04:46 caption says "Time settings" Jun 05 17:05:01 maybe yours, maybe now Jun 05 17:05:06 on main screen there are JUST categories Jun 05 17:05:17 so we need date module Jun 05 17:05:31 not fixing/enhancing time module Jun 05 17:05:32 ;) Jun 05 17:05:41 the two are not separate Jun 05 17:05:56 arrrrrrgh Jun 05 17:06:01 * DocScrutinizer away Jun 05 17:06:08 Blu3: why? Jun 05 17:06:15 date is a less granular portion of time Jun 05 17:06:20 er Jun 05 17:06:33 do anybody know how can I disable vertical sliding in .edc? Jun 05 17:06:38 wpwrak: DocScrutinizer: http://iki.fi/lindi/openmoko/ramconsole-dump.c seems to work. next i'll update the kernel patch so that it does not try to print the buffer on boot anymore Jun 05 17:06:41 Blu3: eh, from programmer point of view not ;) Jun 05 17:06:50 you need to have different widget for date and for time Jun 05 17:07:08 time() returns # of seconds since 1979-1-1 Jun 05 17:07:22 Blu3: in GUI? Jun 05 17:07:26 that naturally exceeds 12/24 hours represented by a clock Jun 05 17:07:35 from a programmer point of view Jun 05 17:07:48 Blu3: do you want users typing time in unix timestamp? Jun 05 17:07:58 Blu3: so err, from GUI programmer point of view ;> Jun 05 17:08:14 and nearly all date/time interfaces include both Jun 05 17:08:23 Blu3, DocScrutinizer: and elementary is still missing calendar widget Jun 05 17:08:31 maybe if you're a gnome person who feels less is better Jun 05 17:08:34 so bug raster for that 8) Jun 05 17:08:43 Blu3: no, not less Jun 05 17:09:03 Blu3: but i want to use modularity of shr-settings Jun 05 17:09:22 Blu3: so on one "date/time" category screen you have two modules Jun 05 17:09:25 Date and Time Jun 05 17:09:26 too many modules is silly. i already have to use tiny desktop icons because there are too many Jun 05 17:09:43 "date" and "time" represent the same thing, they should be one module Jun 05 17:09:59 that's like saying we should have a separate module for the hour, minute, and second Jun 05 17:10:09 Blu3: i don't see difference in used area when using one module for date and time, and two separate modules for that Jun 05 17:10:28 Blu3: no, as you have one widget for hour, minute and second in elementary Jun 05 17:10:29 :P Jun 05 17:10:48 there is widget Clock, and there will be widget Calendar Jun 05 17:10:50 and maybe that's a design flaw of elementary Jun 05 17:11:01 so i just want to use that Jun 05 17:11:05 hi everyone :-) Jun 05 17:11:10 hey onen Jun 05 17:11:35 onen|openBmap: my GPS isn't working :/ i had a nice log trip, but i couldn't log gsm cells :'( Jun 05 17:11:41 s/log/long/ Jun 05 17:11:43 dos1 meant: onen|openBmap: my GPS isn't working :/ i had a nice long trip, but i couldn't log gsm cells :'( Jun 05 17:11:50 * onen|openBmap is writing all the answers about obm/ch thread in openmoko community ML Jun 05 17:12:15 dos1: hi! Jun 05 17:12:41 dos1: what do you mean? A. your gps does not work at all? B. the gps does not work in openBmap logger? Jun 05 17:14:26 Onen_away: don't worry, my GPS doesn't work at all Jun 05 17:14:27 ;) Jun 05 17:19:53 the 'don't panic' bootsplash is fairly pointless if it only shows for about one second then gets blotched out by kernel text as it continues booting Jun 05 17:20:59 Blu3: boot with "loglevel=1 splash" Jun 05 17:21:41 what if i actually want to capture those messages in dmesg after bootup is complete? Jun 05 17:23:42 Blu3: they are in dmesg Jun 05 17:25:09 meh, we need a real bootsplash Jun 05 17:26:33 Blu3: no, we don't want to have real bootsplash Jun 05 17:26:46 yes we do Jun 05 17:26:55 Blu3: no we don't :P Jun 05 17:26:59 yes we do Jun 05 17:27:03 real bootsplash - slower boot Jun 05 17:27:11 and don't showing boot messages when it's needed Jun 05 17:27:15 s/don't/not/ Jun 05 17:27:15 dos1 meant: and not showing boot messages when it's needed Jun 05 17:27:29 one image displayed on start is optimal Jun 05 17:27:30 * Blu3 isn't sure how that would really affect the lengthy boot as it is, fb text output is what takes forever Jun 05 17:27:46 this topic was discussed Jun 05 17:27:58 Blu3: that's why we don't produce fb text output Jun 05 17:28:06 default kernel options have "loglevel=1 splash" Jun 05 17:28:28 not on the current kernel Jun 05 17:28:39 someone must have forgotten that when they pushed it Jun 05 17:28:52 Blu3: no Jun 05 17:28:58 are you using u-boot? Jun 05 17:29:01 uhm, yeah. Jun 05 17:29:04 yep Jun 05 17:29:09 for qi there is correct append-GTA02 file Jun 05 17:29:18 Blu3: so we can't change your environment :P Jun 05 17:29:21 cause it's your Jun 05 17:29:37 then you want to add "loglevel=1 splash" to your kernel cmdline Jun 05 17:29:58 speaking of qi, and not to change the subject, but what're the advantages/disadvantages between qi and uboot? Jun 05 17:30:13 rtp: DocScrutinizer: new patch and userland tool is at https://docs.openmoko.org/trac/ticket/2135 -- feedback and test reports are appreciated :) Jun 05 17:30:15 i think booting from NOR uboot should work too, as it have low loglevel too Jun 05 17:30:23 Blu3: Qi is simplier Jun 05 17:30:32 it doesn't touch glamo and many other things Jun 05 17:30:36 it leaves everything to linux Jun 05 17:30:41 which is good Jun 05 17:30:52 less work doing by bootloader - more speed, less bugs Jun 05 17:31:36 oh joy, my flash partition needs fscked Jun 05 17:32:06 somedays OM annoyeth me Jun 05 17:33:01 dos1: the only problem is that you can't easily test new kernel parameters Jun 05 17:34:46 hmm Jun 05 17:34:53 we have logrotate configured as weekly Jun 05 17:35:12 * Blu3 has yet to reach that milestone *grin* Jun 05 17:41:59 tracfeed: Ticket #508 (Store frameworkd and ophonekitd logs from past) updated Jun 05 17:45:44 tracfeed: Ticket #509 (Change resolution to 240x320 fails) created Jun 05 17:47:05 lindi-: you're awesome :-) Jun 05 17:50:31 SHR: 03seba.dos1 07shr-themes * r60bf60a9c8ca 10/e-wm/e-wm-config-illume-shr/module.battery.src: e-wm-config-illume-shr: use Internal method for checking battery state Jun 05 17:52:02 tracfeed: Ticket #509 (Change resolution to 240x320 fails) closed Jun 05 18:02:56 -testing should be updated really soon Jun 05 18:03:11 most of patches from shr-testing overlay are in fso/milestone5.5 Jun 05 18:04:24 dos1: it never worked? I had once the GPS not working after flashing my FR. After a new flashing of it, it worked. maybe you should re-flash? Jun 05 18:05:00 onen|openBmap: it worked, but now it doesn't work. i flashed few days ago shr-oemerge image, in which GPS worked for some other people, but for me not :( Jun 05 18:05:26 onen|openBmap: but it didn't looked at all as every geek should yet (lack of time) Jun 05 18:05:35 onen|openBmap: so i don't complain (yet :D) ;) Jun 05 18:06:14 cause maybe it's my fault... or hw fault... or something else. firstly i have to figure out, what it is ;) Jun 05 18:06:28 i was only sad because i couldn't log gsm cells for openbmap Jun 05 18:09:40 * dos1 tries fix for popping illume keyboard automatically in gtk+ apps Jun 05 18:09:54 mickeyl: ping Jun 05 18:14:24 dos1: pong Jun 05 18:15:00 mickeyl: could i have access to openembedded to work on shr branch? :) Jun 05 18:16:33 you need to request this via oe-list Jun 05 18:16:37 point to your previous work for reference Jun 05 18:17:31 mickeyl: ok, i'll do that. right now i'm sending patches to shr-devel maillist :X Jun 05 18:27:00 hi everyone, i'm using latest shr unstable and have a buzzfixed gta02v5; are the standard scenarii the ones i want? Jun 05 18:34:17 dos1, let me know your results Jun 05 18:41:39 Zorkman: hi. thanks for the nice support :-) Jun 05 18:41:55 onen|openBmap: np, just stating my opinion Jun 05 18:42:28 Zorkman: hope you like all the precise answers from ch ;-) Jun 05 18:42:52 hehe Jun 05 18:43:12 Zorkman: why do you have your freerunner back again and mine is still on its way although i live much nearer to munich than you do Jun 05 18:43:15 grr Jun 05 18:43:46 slooow hermes delivery Jun 05 18:44:02 lol, i had the same bumbl :) Jun 05 18:44:17 and the order status doesn't get updated properly Jun 05 18:45:25 mickeyl: could you please toss a pointer to actual suspend command invocation (the point and next neighbourhood where device stops/resumes) Jun 05 18:45:43 DocScrutinizer: in frameworkd? Jun 05 18:45:55 i hope they deliver tomorrow Jun 05 18:46:05 I think so - would assume frameworkd is calling suspend Jun 05 18:46:40 because if not i will get it on tuesday as they normally only deliver for big companies on mondays Jun 05 18:47:00 or maybe frameworkd is calling some other command that does a lot of tricky scripting prior/after suspend/resume Jun 05 18:47:46 I'm interested in executing some tasks immediately prior to /after suspen-resume Jun 05 18:48:16 DocScrutinizer: /etc/apm.d/resume.d/ ? Jun 05 18:48:23 btw onen|openBmap, did I really only contribute 486???? thought on one trip i contributed almost 1000 Jun 05 18:48:24 would help Jun 05 18:48:35 thought apm is obsolete though Jun 05 18:48:44 onen|openBmap: you even posted the number in this chan once, and it was >1000 (not to nagg, just curious) Jun 05 18:48:53 DocScrutinizer: # ---------------> Good Night! Jun 05 18:48:53 open( "/sys/power/state", "w" ).write( "mem\n" ) Jun 05 18:48:53 # ---------------< Good Morning! Jun 05 18:49:13 k, thanks. will try to find my way prior to continue pestering other guys Jun 05 18:49:29 Zorkman: 486 is nice number :D Jun 05 18:50:01 dos1::D Jun 05 18:50:14 didn't even notice it :) Jun 05 18:50:17 lindi-: thanks, that's what I expected how thinks might work. Where to find this? Jun 05 18:50:39 DocScrutinizer: ./framework/subsystems/ousaged/generic.py: open( "/sys/power/state", "w" ).write( "mem\n" ) Jun 05 18:50:49 thanks a lot Jun 05 18:51:02 DocScrutinizer: the only problem is that it requires you to run the python code as root :( Jun 05 18:51:14 DocScrutinizer: with os.system("apm -s") it was enough for apm to be suidroot Jun 05 18:51:18 obviously ;-) Jun 05 18:51:47 hm.. that's odd Jun 05 18:52:00 was the caching time of gstreamer reduced? Jun 05 18:52:08 or is my phone just responding faster? Jun 05 18:52:14 or maybe a nice value change? Jun 05 18:52:26 TAsn: i think your phone has good mood today :) Jun 05 18:52:33 hehe ;) Jun 05 18:53:07 Zorkman: did I? probably Nick stole some of yours to keep in front of me ;-) Jun 05 18:53:25 TAsn: maybe the buzz fix fixed more ;) Jun 05 18:53:42 my phone is not buzz fixed :) Jun 05 18:53:50 because it never suffered from buzzing :) Jun 05 18:53:52 only echo. Jun 05 18:54:07 Zorkman: I'll clarify with Nick. I see two explanations, stats were buggy (were on test until one or two days ago); stats are buggy; you are mistaken Jun 05 18:54:29 * onen|openBmap does not count correctly Jun 05 18:54:35 onen|openBmap: i really don't find it important, no need to bother him Jun 05 18:54:57 i might be going on a 3000km roadtrip in the near future, i'll map some cells on that trip :P Jun 05 18:55:26 Zorkman: no, I want to clarify, if this is a bug, it is interesting. And some people are interested in stats (see ch success) Jun 05 18:55:34 onen|openBmap, got a big dbase from shrfsoudcon? Jun 05 18:56:12 Zorkman: hope you have one sim for your way there and another one on the way back ;-) (I bought a new sim to be able to cover more cells :-O ) Jun 05 18:56:41 onen|openBmap: well as long as no data dissolves into nowhereit is no problem, isn't it? Jun 05 18:56:43 onen|openBmap: I HAVE :D Jun 05 18:56:56 but it won't matter, because i won't be returning the same way Jun 05 18:57:03 TAsn: to be honnest, kind of disappointed by the numbers. I think people who did log a lot, were people logging already before. Jun 05 18:57:25 TAsn: other loggers were using ch, and heard about obm during fsoshrudcon Jun 05 18:57:40 TAsn: I hope they will convert to our project from now Jun 05 18:57:51 :( Jun 05 18:57:52 TAsn: already some new accounts from people met there Jun 05 18:57:58 cool. Jun 05 18:58:42 the problem i see with cell logging is privacy - i'd have to upload my whole track (when i have been where) Jun 05 18:58:57 TAsn: ch always is so proud of his numbers. we have neighbour cells since a few weeks only. we went from 16K to 24K since then Jun 05 18:59:22 bumbl: what do you mean about data dissolving? Jun 05 18:59:29 onen|openBmap, why don't you and ch share the db? Jun 05 18:59:33 or at least copy to each other? Jun 05 18:59:57 onen|openBmap: lost data Jun 05 19:00:04 data corruption,... Jun 05 19:00:59 bumbl: I talk about the privacy concern in the summary of the manual. I personnaly do start my logger far from my home, and far before my home Jun 05 19:01:46 bumbl: that is the main reason I put "start/stop" buttons. So that I can keep control of what I log or not easily. otherwise I would simply have logged when app is running Jun 05 19:01:59 onen|openBmap: are you aware of any efforts to ask information about cell locations from providers/governments? Jun 05 19:02:11 bumbl: by the way, this is in no way different from uploading a GPX file to openstreetmap. Jun 05 19:02:20 I tried the institute that regulates it for belgium, already mailed them 5 times over the past half year, never got a response Jun 05 19:03:20 TAsn: please see ongoing thread on om community ML + archive. ch does not want to share his db. he wants to keep the control of it, because it is easier for him to make modifications if he wants. otherwise he would have to ask others before doing so Jun 05 19:03:20 onen|openBmap: that's why i haven't uploaded any gpx files to openstreetmap (i just named some streets) Jun 05 19:03:56 bumbl: scared that lazy lulu is going to come for you too? : Jun 05 19:03:58 *:p Jun 05 19:03:59 onen|openBmap, that's a lame excuse Jun 05 19:04:03 ask him to mirror it once Jun 05 19:04:03 Zorkman: yep Jun 05 19:04:05 TAsn: in a way we will copy. ch declares he will upload his data to oci. we do import oci data from time to time (but not often, lot of work) Jun 05 19:04:05 (only now) Jun 05 19:04:35 this sounds like a classic case of using the community for personal gain (ch) Jun 05 19:04:46 (not willing to share the data) Jun 05 19:04:48 Zorkman: + i have watched Steve Rambam - Privacy Is Dead - Get Over It Jun 05 19:04:55 bumbl: sorry I don't understand what you meant with this: onen|openBmap: well as long as no data dissolves into nowhereit is no problem, isn't it? Jun 05 19:05:01 I'm not saying this is the case, I'm just saying it reminds me of darker times :) Jun 05 19:05:01 onen|openBmap: hay Jun 05 19:05:04 anyhow, i'm off. Jun 05 19:05:50 btw: those that use the Fr to sms, what keyboards do you use? Jun 05 19:05:50 Zorkman: (ask gov/providers about cell location). Some countries have tower locations (switzerland, and I think France too). But not cell coverage Jun 05 19:05:56 onen|openBmap: i meant that as long as just the stats are wrong there is no problem - but if the problem is not with the stats algo but with the data (because of corruption,...) then you have a problem Jun 05 19:07:05 Zorkman: having talked with people working in big provider here in France, they don't want to share this info, and they are pissed off by google having logged it Jun 05 19:07:11 onen|openBmap: think i found a bug.... I'll try to reproduce lataer today, but it goes like this. collect data, select upload, type wrong password, get error message, select no to the question about deleting processed logs, when i try to upload once more with correct password, all the collected data is gone Jun 05 19:07:11 Zorkman: you should write email about your tries, might give idea to others. to bad you got no answers :-( Jun 05 19:07:43 onen|openBmap: did he say why they didn't like it? Jun 05 19:08:06 onen|openBmap: can't immediatly think of a reason how this could harm a provider Jun 05 19:08:14 Zorkman: i use the old large qwerty keyboard Jun 05 19:08:19 that was there first Jun 05 19:08:50 TAsn: I do think so. I pushed hard (maybe to hard, and he thought I wanted to swallow his project into our...) for having only one db with all the info both projects want, and one logger or at least one logging engine Jun 05 19:09:04 TAsn: but he answered he had no time for a collaborating project Jun 05 19:09:35 TAsn: he does not want to share his database, but the data are freely available for download Jun 05 19:09:45 seen it before Jun 05 19:09:52 we had an OSM like project in israel Jun 05 19:09:56 the dude shared the info Jun 05 19:10:02 and you could only get it from his db Jun 05 19:10:06 *software Jun 05 19:10:48 he won't share it no matter what. Jun 05 19:10:54 anyhow, I'm off, ciao. Jun 05 19:11:04 bye TAsn Jun 05 19:11:23 bumbl: (privacy) I usually only do gpx of strolling, or OSM mapping trip. so no privacy revealed there. Jun 05 19:11:50 onen|openBmap: well you reveal that your useraccount was there at a certain point of time Jun 05 19:13:08 bumbl: the data are only bound to you accountname. so this is a first protection (in obm). another we settled for is: we will propose raw data for download. But with a file with the list of contributors. gps/gsm logs won't be linked to logins. thus only us (obm) could know who did what Jun 05 19:13:45 bumbl: well not completely true, if you are the only one to upload sth between two updates from the raw data package, it could be linked to your login... Jun 05 19:13:58 fredrin: hi, how are things going? Jun 05 19:14:17 relative good i guess Jun 05 19:14:50 bumbl: (stats/ data corruption), no data is reliable, as it is produced by my logger 8-) Jun 05 19:15:01 did you read the message about the bug? Jun 05 19:15:03 s/no/no, / Jun 05 19:15:03 onen|openBmap meant: bumbl: (stats/ data corruption), no, data is reliable, as it is produced by my logger 8-) Jun 05 19:15:54 onen|openBmap: well as i said i am paranoid after the privacy is dead talk Jun 05 19:16:08 * onen|openBmap is bad at following many subjects at the same time, sorry Jun 05 19:17:04 fredrin: collect data, (stop logging?), select upload... ? Jun 05 19:17:22 type wrong password, but right username Jun 05 19:17:32 get error Jun 05 19:17:40 and answer now to delete processed logs Jun 05 19:17:44 fredrin: I mean, do you stop the logger before upload? Jun 05 19:17:54 i just hit the upload Jun 05 19:18:27 Zorkman: this is sth of great value. many usage depends on the data. if not available, you have to pay to get access to it. Jun 05 19:18:54 Zorkman: maybe they don't want to let competition where is good/bad coverage of their network too.... ? Jun 05 19:19:32 onen|openBmap: fuck that; if i buy a car, I also want to know were there are no gass stations... Jun 05 19:19:46 but i understand what you're trying to say, i just find it very shortsighted of the providers Jun 05 19:20:54 onen|openBmap: i'll do some more digging into the problem. I just sucks to loose data Jun 05 19:21:04 TAsn: was it the app written in some java compiled to c or sth like this, based on roadmap software or sth? Jun 05 19:21:04 TAsn: bye (wow, I am so quick :-) ) Jun 05 19:21:10 bumbl: imagine you shuffle data, and remove everything related to time. that would protect you. Jun 05 19:21:22 yep Jun 05 19:21:27 bumbl: but we need data to know how old are the data (cells appear/disappear) Jun 05 19:21:45 bumbl: that makes me think, maybe month+year would be enough Jun 05 19:22:08 i might write a script for that Jun 05 19:22:26 changing the date + time within half a year Jun 05 19:22:33 randomly for each cell Jun 05 19:23:08 bumbl: I am very concerned about privacy (very!). But privacy will never be as it was. too many sources. even a friend who puts on a website a photo with you and write what he did throws your (old way) privacy away :-( Jun 05 19:23:30 well yes Jun 05 19:24:25 Zorkman: hey, I can understand why they do, but I try to make it change ;-) the day the value won't be enough (because obm goog enough), maybe they will give it away? Jun 05 19:24:29 * onen|openBmap is optimistic Jun 05 19:25:00 although i have refused to join facebook for months there are some photos showing me up there Jun 05 19:25:04 fredrin: do you have your FR at hand right now? Jun 05 19:25:13 always Jun 05 19:25:16 :) Jun 05 19:25:25 fredrin: look at HOME/.openBmap/Logs|Processed_logs directories Jun 05 19:26:44 bumbl: please talk with us first about this (obm sourceforge forum?). I just want to be sure this would not bring problems (I don't think so, but I prefer to check first). I will ask Nick what he thinks Jun 05 19:27:10 onen|openBmap: i checked, both was empty Jun 05 19:27:10 bumbl: but I would say year+month would be very much enough. maybe even semester+year Jun 05 19:27:21 fredrin: s*cks :-( Jun 05 19:27:40 i remember i checked them after the error happend aswell Jun 05 19:28:15 bumbl: I am working on social network things for a living, but not even subscribed to it ;-) Jun 05 19:28:30 maybe it's something with the checking of user/pass and then it continues to do stuff even if it's wrong username or password Jun 05 19:29:29 fredrin: how it (is supposed to) works: try sending, A. gets ok, move file to Processed_logs/. Jun 05 19:30:35 onen|openBmap: but i got the question if i wanted to do stuff with the prosessend logs even if i had the wrong password. Jun 05 19:30:39 fredrin: B. gets "already exists", suppose it was already uploaded and move it as if succesfully uploaded Jun 05 19:30:40 fredrin: C. sth else, don't move it, it is an error Jun 05 19:30:44 onen|openBmap: well social networks have many advantages Jun 05 19:30:58 fredrin: yes that is normal behaviour Jun 05 19:31:02 but they all come at their price Jun 05 19:31:08 bumbl: indeed Jun 05 19:31:37 bumbl: much too long subject, and out of topic in this channel anyway ;-) don't want to start on this :-P Jun 05 19:31:51 bumbl: plus, I have a bug to hunt down to death Jun 05 19:31:55 onen|openBmap: shoudn't it be like, if wrong username/password, do nothing with the logs, just give error. Jun 05 19:32:18 hehe yes Jun 05 19:32:29 fredrin: if wrong u/p, then error then file stays in [..]/Logs Jun 05 19:33:00 fredrin: hiting "yes" to the question will only do sth with files located in [..]/Processed_logs (PL) Jun 05 19:33:20 fredrin: did you by luck not relaunch the app since the pb occured? Jun 05 19:33:38 i relaunched Jun 05 19:33:56 fredrin: home/.openBmap/openBmap.log is the application log file. you could see here what happens in the app. it is overwritten at every launch :-( Jun 05 19:34:56 fredrin: you got error message you said? what exaclty? Jun 05 19:35:24 fredrin: weird, weird, weird Jun 05 19:35:35 fredrin: I note it, and try to figure it out Jun 05 19:35:45 fredrin: a quick fix: use correct password :-D Jun 05 19:35:51 onen|openBmap: is that true? is it always overwritten? can't be that hard to code it in a way that if there already is a opebBmap.log, it creates a openBmap1. log, no? Jun 05 19:36:21 onen|openBmap: i'll try to reproduce it today, but the message i got, was, 1) wrong password 2) If i wanted to move the logs 2a) i said no, then i try to upload again, with correct password, and there was no data Jun 05 19:37:16 * onen|openBmap is annoyed by users, they find bugs. Without users, no bugs Jun 05 19:39:00 Zorkman: I know it is not satifactory. is on my todo list. rotating log file is provided by python. it just needs more attention than the "standard" log I use Jun 05 19:39:02 lol Jun 05 19:39:17 Zorkman: just did not take time to take care of this. but patches are welcome 8-) Jun 05 19:39:58 onen|openBmap: if i could code it... Jun 05 19:40:00 Zorkman: another pb with current state, log file keeps growing until restart, no limit :-( Jun 05 19:40:03 onen|openBmap: looking at the code now myself Jun 05 19:40:29 Zorkman: what prevents you from doing it? Jun 05 19:40:34 btw: is there a handy linux command that will let me sort every word in a text file in alphabetical order (to make a dictionary file) Jun 05 19:40:56 onen|openBmap: time, full- time looking for a job atm :) Jun 05 19:41:07 lame reason i know, but it fills my days quite good :) Jun 05 19:43:05 Zorkman: sort Jun 05 19:43:30 very powerful Jun 05 19:44:27 * onen|openBmap got disconnected, last seen message: onen|openBmap: time, full- time looking for a job atm :) Jun 05 19:45:06 onen|openBmap: you didn't miss anything Jun 05 19:45:08 bumbl: thanks Jun 05 19:45:14 just found it while googling Jun 05 19:45:15 np Jun 05 19:48:37 onen|openBmap: hmmm, maybe i b0rked something when i logged the data, because now it's working..... I think i was just me that did not hit the 'generate' button when i was out logging. I forgot i had to do that when i was out logging.... Jun 05 19:49:32 "Start Logging" would be a more intuitive name than 'Generate' thought IMHO Jun 05 19:49:48 Zorkman: thanks for letting me know Jun 05 19:50:17 fredrin: what is the category for such bug "user" ? Jun 05 19:50:22 fredrin: ;-) Jun 05 19:50:43 onen|openBmap: from now on, blame the uses if there is a bug Jun 05 19:50:52 user Jun 05 19:51:42 fredrin: have to change this too (work on the GUI this weekend), remove generate, and put "start". generate is misleading Jun 05 19:51:42 fredrin: plus, buttons have text embedded in image, which prevent internationalisation Jun 05 19:52:00 fredrin: well I don't need to, there is *no* bug Jun 05 19:52:05 yeah, nice Jun 05 19:52:13 fredrin: just kidding Jun 05 19:52:22 fredrin: I appreciate your help! Jun 05 19:53:04 sure man Jun 05 19:53:22 just found it while googling Jun 05 19:53:32 (sorry, wrong up and enter :)) Jun 05 19:53:45 I have a question: Jun 05 19:53:53 I made my own dic file (a dutch one) Jun 05 19:54:06 initially there are no frequencies, just a plain list of words Jun 05 19:54:23 but even after ten times typing a particullar word, it still doesn't get a frequency Jun 05 19:54:30 thought illume did that automaticall? Jun 05 19:56:43 Zorkman: yes in the user dict in your home dir Jun 05 19:56:51 look in .e/ Jun 05 19:57:29 bumbl: so i don't need to place my dictioanry in /usr/lib/enlightenment/modules/illume/dicts ? Jun 05 19:57:58 not nesicarily afaik Jun 05 19:58:27 Zorkman: how did you make the dict? Jun 05 19:58:36 the dicts dir is even empty, but everything is being logged in the dicts-dynamic dir Jun 05 19:59:17 i once made an empty dict for /usr/lib/enlightenment/modules/illume/dicts Jun 05 19:59:21 fredrin: i went to a site that has two large files with dutch words in it, i merged them, and sorted them alphabetically, named it dutch.dic and that's it Jun 05 19:59:30 and then trained it Jun 05 19:59:46 which is imo the perfect solution for all not english languages Jun 05 19:59:48 i want to make a dict myself, but based on a mix of languages, english, spanish and norwegian. but based on the sms inbox and outbox Jun 05 20:00:43 that should be no problem Jun 05 20:01:48 * onen|openBmap_ got disconnected (again), last seen message: yeah, nice Jun 05 20:02:16 onen|openBmap_: after that message i have here two answermessages from you Jun 05 20:02:24 fredrin: don't know if you got my last message: just kidding || I appreciate your help! Jun 05 20:02:28 bumbl: can you elaborate a bit more?, copied it to the e/ditcs directory, but that dic file still doesn't get updated Jun 05 20:02:38 only te dynamic-dicts Jun 05 20:02:41 onen|openBmap_: got it :) Jun 05 20:02:54 bumbl: or is that the way it's supposed to be+? Jun 05 20:03:04 bumbl: yes, I meant the last message received seen, sorry Jun 05 20:03:18 Zorkman: that's the way it's supposed to be Jun 05 20:03:31 dynamic-dicts is for updating Jun 05 20:03:49 so when reflashing, i should always back up both files, the dictionary and the dynamic-dicts Jun 05 20:03:52 check Jun 05 20:05:05 dos1: for a buzzfixed A5, is the standard statefile from the latest unstable good? Jun 05 20:05:11 Zorkman: dunno Jun 05 20:05:39 :) ok, because i' kinda lost with all the different ones posted in ML and trac Jun 05 20:06:27 Zorkman: ithink that's how it is intended Jun 05 20:07:00 Zorkman: use the a7 one Jun 05 20:08:29 bumbl: http://docs.openmoko.org/trac/attachment/ticket/2121/gsmhandset-a7.state? Jun 05 20:08:58 yep Jun 05 20:09:15 that's what DocScrutinizer told me Jun 05 20:12:08 * onen|openBmap_ finally went through all the responses to the ML thread :-S Jun 05 20:12:46 bumbl; i don't have a /usr/share/openmoko/scenarios/ , has the dir changed? Jun 05 20:13:36 Zorkman: /usr/share/shr/scenarii/ Jun 05 20:18:32 thanks again :) Jun 05 20:18:44 g2g now, good luck all! Jun 05 20:21:05 TAsn: was it the app written in some java compiled to c or sth like this, based on roadmap software or sth? Jun 05 20:21:12 I think it's roadmap based Jun 05 20:26:27 onen|openBmap_: Hi there! Do you have any guesses what makes openBmap fail in OM2009? Jun 05 20:26:35 TAsn: ok I think I see what you are talking about, I have always been wondering from where it got all its data for israel Jun 05 20:26:58 onen|openBmap_, the dude here doesn't share his data Jun 05 20:26:59 rhkfin: OM2009 is FSO based, right? Jun 05 20:27:04 so I doubt it's there Jun 05 20:27:05 onen|openBmap_: yes Jun 05 20:27:12 TAsn: my preciouuuuuuuuus, data Jun 05 20:27:19 :) Jun 05 20:27:53 onen|openBmap_: can the stopping the logger dialog not be modal please? Jun 05 20:28:07 it's not possile to switch to a different applicatino while it is shown Jun 05 20:28:13 possible Jun 05 20:29:15 dos1, applied your stuff Jun 05 20:29:48 I'm back from my conference. one week offline,... Jun 05 20:29:50 onen|openBmap_: 'reported' this at http://www.opkg.org/package_186.html - it gives no messages (exit code 1), it just doesn't start Jun 05 20:30:28 mrmoku: thanks Jun 05 20:30:31 TAsn: mmh? I meant http://www.freemap.co.il/map.html Jun 05 20:30:39 mrmoku: http://pastebin.com/m59ae6469 - why it fails? Jun 05 20:30:51 SRCREV is 1 Jun 05 20:30:52 onen|openBmap_, yeah Jun 05 20:30:54 that's the dude. Jun 05 20:31:00 onen|openBmap_: btw congrats on being now the biggest & beautiful :) Jun 05 20:31:01 and if i want to use SRCPV, it doesn't parse Jun 05 20:31:02 he's an ass. Jun 05 20:31:03 rhkfin: what happens when you run it? Jun 05 20:31:04 I hate him. Jun 05 20:31:15 spaetz: why? Jun 05 20:31:27 and he is calling it Jun 05 20:31:33 spaetz: sorry, you already said Jun 05 20:31:42 the free (as in freedom, not beer) israeli mapping project Jun 05 20:31:43 spaetz: I think you can... Jun 05 20:31:47 he should really get sued. Jun 05 20:32:01 onen|openBmap_: not with the illume app switching arrows Jun 05 20:32:11 TAsn: yeah, use free and open are buzz words nowadays Jun 05 20:32:23 ;\ Jun 05 20:32:30 spaetz: I will check, but I would say I did it Jun 05 20:32:43 anyhow, I'm off again Jun 05 20:32:48 going to get something to eat. Jun 05 20:32:48 spaetz: I will try, and come back to you (hope another lane to todo list :-( ) Jun 05 20:32:57 dos1, with what error does it fail? Jun 05 20:33:00 no hurry, no big thing Jun 05 20:33:17 onen|openBmap_: absolutely nothing happens. On command line I get no messages.. Jun 05 20:33:19 I was just somewhat annoyed as stopping the logger takes a few seconds :) Jun 05 20:33:44 mrmoku: http://pastebin.com/m1de97ee1 Jun 05 20:34:27 dos1, add AUTOREV for it Jun 05 20:34:36 mrmoku: oh Jun 05 20:34:41 :) Jun 05 20:34:45 mrmoku: i'm stupid :D Jun 05 20:34:58 dos1, nah... unconcentrated maybe :P Jun 05 20:35:16 rhkfin: thanks, but I do not take glory for this. It has been a great amount of work for Nick to import oci, and to me for repeating our message everywhere. it's a pity, as this should not had been necessary :'-( Jun 05 20:36:35 rhkfin: so if this is FSO based... do you have a (home)/.openBmap which has been created? Jun 05 20:36:39 onen|openBmap_: ok, well, thank Nick for me :) Jun 05 20:36:43 onen|openBmap_: I'll check. Jun 05 20:37:09 onen|openBmap_: yes, it's there Jun 05 20:37:32 rhkfin: good. by luck, an openBmap.log file in there? Jun 05 20:37:56 onen|openBmap_: yes, I'm looking at it.. I'll pastebin it.. Jun 05 20:38:46 rhkfin: are you the person who wrote the message on opkg.org? Jun 05 20:39:05 onen|openBmap_: http://pastebin.com/d58c39c2e (yes, I wrote it there) Jun 05 20:39:27 onen|openBmap_: looks like that it's doing stuff but no GUI is started.. Jun 05 20:40:11 rhkfin: I got no RSS email whatever about your message :-( sorry for not responding it Jun 05 20:40:58 rhkfin: hope the "selfish people" of your email was not for me ;-) Jun 05 20:40:59 onen|openBmap_: hmm. actually that's an old log file from 2 weeks ago... I have an oldish home dir. Jun 05 20:41:09 [2009-06-05 22:08:29] bumbl: http://docs.openmoko.org/trac/attachment/ticket/2121/gsmhandset-a7.state? Jun 05 20:41:10 [2009-06-05 22:08:58] yep Jun 05 20:41:12 [2009-06-05 22:09:15] that's what DocScrutinizer told me Jun 05 20:41:38 noooope! http://docs.openmoko.org/trac/attachment/ticket/2121/gsmhandset.state.new Jun 05 20:42:16 rhkfin: indeed, an old file Jun 05 20:42:30 onen|openBmap_: no, it's not you, I interpret that you've met a 'difficult project' and forked it. See http://risto.kurppa.fi/blog/the-open-source-project-that-made-me-annoyed-angry/ - I believe I know some of your feelings.. Jun 05 20:42:44 DocScrutinizer: can't you remove that -a7.. It's misleading people all the time.. Jun 05 20:43:23 it even tricked *me* :-/. But I can *not* remove it from trac :-((( Jun 05 20:43:35 onen|openBmap_: I renamed the .openBmap dir and recreated it and tried to run with no success. No log files created. I'll try to reinstall (if the installation installs the config files..) Jun 05 20:43:47 rhkfin: ok good, then I keep going helping you ;-) Jun 05 20:44:10 rhkfin: no Jun 05 20:44:16 DocScrutinizer: you or anyone? Make someone remove the file from the server (-> a broken link but it'll not mislead :) or symlink it to the .new :) Jun 05 20:44:28 rhkfin: at startup check if conf file exists. if not create one with default values Jun 05 20:44:33 roh: ping Jun 05 20:44:41 onen|openBmap_: ok... Jun 05 20:45:21 rhkfin: that's tricky, no message and no om2009 to test... Jun 05 20:45:29 onen|openBmap_: right... Jun 05 20:45:42 ah, now it created the files (reinstall & re-run.. Jun 05 20:45:44 roh: [2009-06-05 22:44:16] DocScrutinizer: you or anyone? Make someone remove the file from the server (-> a broken link but it'll not mislead :) or symlink it to the .new :) Jun 05 20:45:48 if i said "shr is oe-based", would that be a statement anyone would object to ? Jun 05 20:46:00 roh: http://docs.openmoko.org/trac/attachment/ticket/2121/gsmhandset-a7.state? Jun 05 20:46:06 onen|openBmap_: no, sorry.. Jun 05 20:46:51 onen|openBmap_: looked the wrong dir. So absolutely no messages (except Illume let's me save the error report that tells me that exit status was 1..) Jun 05 20:47:25 onen|openBmap_: this is what illume saves: http://pastebin.com/d1596873c Jun 05 20:48:30 onen|openBmap_: does it work on SHR? Jun 05 20:48:39 rhkfin: yes Jun 05 20:48:48 ok... Jun 05 20:48:58 hmm, nobody ? :-( Jun 05 20:49:09 rhkfin: I try to find an easy way to make you test it. but I don't know how far you are willing to spend time on this ;-) Jun 05 20:49:31 wpwrak: no objection Jun 05 20:49:33 wpwrak, no, don't think anybody could object Jun 05 20:49:42 rhkfin: openBmapGTK creates GUI and uses ....../site-packages/openbmap/logger.py file Jun 05 20:49:45 stefan_schmidt: excellent, thanks ! Jun 05 20:50:13 * stefan_schmidt just registered a new machine ID and has seen that wpwrak registered the 9's Jun 05 20:50:23 wpwrak: old guy, he? ;) Jun 05 20:50:27 stefan_schmidt: hallo! I am thrilled by developments of the thread! we go much further than last time ;-) Jun 05 20:50:45 stefan_schmidt: the Psion S5 ? ah yes, been there ;-) Jun 05 20:50:55 onen|openBmap_: heh, hi Jun 05 20:51:00 wpwrak: yup Jun 05 20:51:30 * stefan_schmidt just registered the Samsung SGH-i900 aka Omnia Jun 05 20:51:47 stefan_schmidt: it actually had pretty good power managenment. all kernel-based and with scary little /proc files, but it did after a few days of work pretty much what the neos are starting to get close to now as well :) Jun 05 20:51:50 onen|openBmap_: any idea where that site-packages might be? Jun 05 20:52:01 If anybody has infos about the protocol used for the dual port ram communication with the msm6xxx modems let me know! Jun 05 20:52:10 rhkfin: thanks for hint. Anyway I took another intention for doing a decent audiotool tonight, to help user fix statefile issues and esp mic gain (#48) and mic-level, according to procedure described here: http://wiki.openmoko.org/wiki/Neo_Freerunner_audio_subsystem#Alsamixer_channel_controls Jun 05 20:52:34 rhkfin: so an approach would be to launch a python console, import the file logger.py, and then run init() method from obm class (openBmapGTK will guide you at what to call for init). Is it ok, or too complicated for what you are willing to test? Jun 05 20:52:36 wpwrak: yup, it's sad that hardware ages to fast Jun 05 20:52:48 DocScrutinizer: Awesome! A tool is needed! Jun 05 20:52:57 rhkfin: /usr/lib/python-2.6/site-packages/openbmap/ (AFAIR) Jun 05 20:53:05 onen|openBmap_: btw, ARFCN would be something I like openbmap log as well Jun 05 20:53:15 onen|openBmap_: I could try.. Will need help but I'll see .) Jun 05 20:53:40 anybody knows about (1)sox for FR? Jun 05 20:53:54 stefan_schmidt: are you working on the omnia (IIRC the dual ram thing is about it, right?) ? Jun 05 20:54:01 sorry for asking Jun 05 20:54:25 onen|openBmap_: well, I gathered some infos about it today. (Should arrive on monday) Jun 05 20:54:37 Modem is the biggest unknown, as usual Jun 05 20:54:44 stefan_schmidt: pfffff, DocScrutinizer first with timing advance, take a ticket ;-) Jun 05 20:55:16 onen|openBmap_: heh, poke you later then Jun 05 20:55:21 Ainulindale, ping Jun 05 20:55:26 stefan_schmidt: ok, on todo list. Is there anything I should know about it (is it always valid, when read etc...?) Jun 05 20:55:28 onen|openBmap_: I think I'll need a step-by-step help here.. Want to go private, do you have time for this? But anyway, I found the site-packages/openbmap with 3 files (__init__.py size is 0) Jun 05 20:55:52 onen|openBmap_: should be valid Jun 05 20:56:31 rhkfin: yes, no pb, go ahead Jun 05 20:57:00 stefan_schmidt: ok, then it should be quite easy, now that I have monitoring serving/neighbour methods Jun 05 20:57:26 stefan_schmidt: have to see with nick, because we agreed on a V2 version of xml files, and arfcn is not there Jun 05 20:57:29 Si I now have the python console running and ran 'import logger' - then what..? Jun 05 20:57:40 onen|openBmap_: it's not urgent. Would just be nice to have it logged. Jun 05 20:58:28 BTw, the idea why we would like to have it is quite funny. Jun 05 20:58:56 stefan_schmidt: no pb, it is on todo list, and not a big deal for nick. he does not use it in the db, so he does not care if the raw data contains anything he does not use Jun 05 20:58:59 Harald, rms and Sean thought about a way to detect IMSI catcher Jun 05 20:59:28 rhkfin: I understood you wanted to go private... ? Jun 05 20:59:55 So if you detect a new ARFCN around you in a known area it is suspect. :) Jun 05 21:00:33 Still nothing you can base a lot on, but an interesting fact. Collecting them is cheap when you collect other things anyway. Jun 05 21:03:04 DocScrutinizer: what audio tool are you making? Jun 05 21:03:20 * Toaster` is back! Jun 05 21:03:20 or what's the idea? Gui editor for alsa.state or smarter? Jun 05 21:03:30 toast Jun 05 21:03:48 some silly script to setup #48 correctly, to check if alsa.statefile is ok, etc Jun 05 21:04:21 onen|openBmap_: Can you give me a short version what you are logging atm? Lost a bit track on this. Jun 05 21:04:28 actually gui is the part I don't like to tackle at all right now Jun 05 21:04:54 is it possible to have some sort of echo check or like make a test-tuner-call Jun 05 21:06:00 dos1: any idea what the root cause of the clock issue is? Jun 05 21:06:11 Toaster`: what clock issue? Jun 05 21:06:48 there was shr-settings bug regarding the time/date module Jun 05 21:07:27 Toaster`: ah some userland stuff Jun 05 21:07:37 lindi-: ish Jun 05 21:07:59 depends on the root cause... :) Jun 05 21:09:52 Toaster`: just that FSO call isn't for setting time :) Jun 05 21:09:57 it's for alarm i think Jun 05 21:10:10 mrmoku: sent 2 new patches Jun 05 21:10:16 ahh ok Jun 05 21:11:56 stefan_schmidt: I'll sent you an xml file sample per email, ok? Jun 05 21:12:12 onen|openBmap_: thanks Jun 05 21:12:34 mrmoku: uh, i think i forgot to bump PR in tasks. can you do that? :) Jun 05 21:12:42 stefan_schmidt: rms? Jun 05 21:12:55 onen|openBmap_: Richard Stallman Jun 05 21:14:08 stefan_schmidt: oh my god, I am getting closer and closer to light :-D Jun 05 21:14:15 mrmoku: oh! and add autorev! Jun 05 21:14:16 :/ Jun 05 21:14:22 onen|openBmap_: lol Jun 05 21:14:45 dos1, first of all I have to review your patches ;) Jun 05 21:15:19 stefan_schmidt: maybe openbmap will become the best partner to laforge's work on GSM BTS and so forth ;-) Jun 05 21:15:41 mrmoku: well. first is for recipe with polish dictionary for illume Jun 05 21:16:08 mrmoku: second is for adding matchbox-keyboard-im to image Jun 05 21:16:10 onen|openBmap_: The best partner would be a GSM modem for which we have the full source code and access to modify it :) Jun 05 21:16:41 mrmoku: with that package keyboard is popping up automatically with gtk+ apps Jun 05 21:16:54 mrmoku: like in good old 2007.2 :) Jun 05 21:17:10 stefan_schmidt: I feel your frustration about m800 ;-) Jun 05 21:17:37 onen|openBmap_: That was not about this. Jun 05 21:17:51 onen|openBmap_: There is just _none_ modem we have enough control over. Jun 05 21:17:56 Not even the calypso Jun 05 21:17:57 dos1|neo: mmh keyboard in gtk means valaterm? Jun 05 21:18:00 stefan_schmidt: I see Jun 05 21:18:30 onen|openBmap_: The ideal counterpart to OpenBSC would be a 100% free GSM stack on the modem Jun 05 21:18:31 mrmoku: but i forgot to add AUTOREV for e-wm-illume-dict-pl, and to bump PV of tasks Jun 05 21:18:40 No idea if this will ever happen Jun 05 21:18:42 dos1|neo, ok, applied :) (and added AUTOREV + bumped PR for feed) Jun 05 21:18:43 spaetz: yep, in valaterm too Jun 05 21:18:51 spaetz: tested, it works :) Jun 05 21:19:01 dos1|neo, second? Jun 05 21:19:05 yay Jun 05 21:19:25 mrmoku: second patch Jun 05 21:19:38 mrmoku: thanks :) Jun 05 21:19:42 hit my mailbox now... after pushing the other one :P Jun 05 21:20:14 pushed the second one too Jun 05 21:21:13 stefan_schmidt: but can obm be the best parner until we got the open modem? (please?) Jun 05 21:21:47 mrmoku: so bump task-shr-minimal too :) Jun 05 21:22:14 * stefan_schmidt hands onen|openBmap_ the best OpenBSC partner medal Jun 05 21:22:24 hehe Jun 05 21:22:37 * onen|openBmap_ hopes it is chocolate medal :-) Jun 05 21:22:55 stefan_schmidt: jokes aside, sure, no big deal loggint it. will do it Jun 05 21:23:00 dos1|neo, next time I will refuse to apply patches without bumped PR ;) Jun 05 21:23:18 onen|openBmap_: thanks Jun 05 21:24:00 mrmoku: sorry :x i had to go from my pc :x Jun 05 21:24:18 * onen|openBmap_ runs with the medal and dig a hole in his garden Jun 05 21:24:21 onen|openBmap_: (rms) lol - the guy that never finds a harddisk meeting his high-spirited needs, as none has FOSS firmware ;-) Jun 05 21:24:26 mrmoku: and testing it took ages Jun 05 21:25:16 DocScrutinizer: ;-) Jun 05 21:25:41 but ok, that's all i wanted to have commited ATM :) Jun 05 21:25:53 :) Jun 05 21:26:04 dos1|neo, btw.... somebody spammed our wiki Jun 05 21:26:08 when new image will be built? :) Jun 05 21:26:15 (furester discovered that) Jun 05 21:26:29 added stupid amazon links to some pages Jun 05 21:26:34 mrmoku: which wiki? SHR? Jun 05 21:26:41 DocScrutinizer, yep Jun 05 21:26:43 mrmoku: well. somebody was spamming forum on my server Jun 05 21:27:05 It was done with two accounts Jun 05 21:27:07 mrmoku: i had to turn off mysql, cause build was slooooooow ;p Jun 05 21:27:14 (which are deleted now ;) Jun 05 21:27:39 dos1|neo, yeah... sometimes OE is a pain in the ass :P Jun 05 21:28:04 incredibly what effort those fsckng spammers do Jun 05 21:28:18 http://shr.bearstech.com/trac/wiki/Customize?action=diff&version=18 Jun 05 21:28:30 i was supriced, cause it never was as slow as today Jun 05 21:28:55 i noticed mysql was eating resorces Jun 05 21:29:10 yeah, mysql is quite hungry Jun 05 21:29:30 * Blu3 prefers postgresql Jun 05 21:29:32 and then i noticed full of spambots adding oferts to enlarge stuff every second ;x Jun 05 21:29:53 ouch :( Jun 05 21:30:19 i was using mysql only for that board Jun 05 21:30:38 which anyway was unused for some time ;) Jun 05 21:30:55 so i turned off mysql as workaround ;D Jun 05 21:32:01 dos1|neo: why not run a nmap & DoS against each bot IP? ;-) Jun 05 21:34:00 I even got a script for this (if it's not bit rotten meanwhile ;-) Jun 05 21:36:06 mrmoku: would you be kind enough to pastebin me the output from: opkg info openbmap-logger ? Jun 05 21:37:45 onen|openBmap_, from the one I uploaded some cells before? Jun 05 21:39:17 DocScrutinizer: later, today i didn't have time for that :( Jun 05 21:39:41 DocScrutinizer: i just wanted to test my package and send patches :P Jun 05 21:39:43 DocScrutinizer, want the IP? :P Jun 05 21:40:01 nah, not really Jun 05 21:40:30 btw it's most likely void meantime Jun 05 21:40:45 onen|openBmap_, uhh... did not yet upload :P Jun 05 21:41:03 mrmoku: well, from the package in your feed. just the opkg command: #opkg info openbmap-logger Jun 05 21:41:13 mrmoku: :-) Jun 05 21:41:37 mrmoku: I want to check dependencies of the paquet, as rhkfin met problems with om2009 about this Jun 05 21:41:52 mrmoku: add libframeworkd-phonegui-efl2 to testing autorev Jun 05 21:42:12 mrmoku: it's breaking build (as i see on maillist) Jun 05 21:42:16 onen|openBmap_, http://pastebin.com/d63b8f3ec Jun 05 21:42:43 mrmoku, dos1|neo: Would it not make sense to put the illume dict into upstream svn, together with other dict files? Jun 05 21:43:20 stefan_schmidt, that should be the final destination, yeah Jun 05 21:43:21 mrmoku: I have some recipes and updates for some swisscom package here. Should I just push them to shr/import? (No interaction with other stuff) Jun 05 21:43:37 stefan_schmidt, yes, just push it :) Jun 05 21:43:48 mrmoku: ok, then just poke raster. He should be glad to put it there. Jun 05 21:43:55 mrmoku: ack Jun 05 21:44:08 stefan_schmidt: i agree :) Jun 05 21:44:19 ~seen raster Jun 05 21:44:19 raster was last seen on IRC in channel #openmoko-cdevel, 1d 3h 42m 22s ago, saying: 'stefan_schmidt: awesome. big pixel. that's all me!'. Jun 05 21:44:19 mrmoku: stefan_schmidt: rhkfin: weird, dependencies are missing from the package in SHR feed Jun 05 21:44:36 mrmoku: thanks for the pastebin Jun 05 21:44:38 onen|openBmap_, what dependencies? Jun 05 21:44:54 mrmoku: stefan_schmidt: give me a second, I look at the recipe Jun 05 21:45:37 mrmoku: OM2009 feed: http://downloads.openmoko.org/repository/unstable/armv4t/Packages has these dependencies: Depends: python-subprocess, python-netclient, python-math, python-core, python-io, frameworkd Jun 05 21:46:23 mrmoku: python-pygtk and python-pygobject are missing Jun 05 21:46:47 mrmoku: compare to http://tinderbox.openembedded.net/packages/600903/ Jun 05 21:47:29 rhkfin, looks like opkg info shows just the RDEPENDS... and not the DEPENDS Jun 05 21:47:42 DEPENDS = "python python-dbus python-pygobject python-pygtk libglade" Jun 05 21:47:43 RDEPENDS = "python-subprocess python-netclient python-math python-core python-io frameworkd" Jun 05 21:47:46 mrmoku: stefan_schmidt: rhkfin: Stefan's recipe in fsoM5.5 feed contains Jun 05 21:47:53 mrmoku: yep Jun 05 21:47:57 :) Jun 05 21:48:12 and opkg info shows only the RDEPENDS Jun 05 21:49:03 mrmoku: stefan_schmidt: rhkfin: but your "opkg info openbmap-logger" output miss the dependencies located in rdepends, doesn't it? Jun 05 21:49:53 mrmoku: ok might be it does not show it simply Jun 05 21:50:28 yep Jun 05 21:50:42 mrmoku: rhkfin: fact is, rhkfin had dependencies which were not pulled in. weird. (nobody under FSO or SHR raised this problem to me... ) Jun 05 21:51:06 mrmoku: pushed Jun 05 21:51:12 hmm Jun 05 21:51:14 root@om-gta02 ~ $ opkg whatdepends python-pygtk Jun 05 21:51:23 does not show openbmap-logger though... Jun 05 21:51:27 it just shows phonelog Jun 05 21:51:31 mrmoku: ah, you see :-) Jun 05 21:51:44 stefan_schmidt, ok :) Jun 05 21:52:27 stefan_schmidt, could you sched some light on that dependency issue? I always assumed that DEPENDS automatically are RDEPENDS too... is that so? Jun 05 21:52:56 mrmoku: hmm, say should Jun 05 21:53:18 s/say/they/ Jun 05 21:53:19 stefan_schmidt meant: mrmoku: hmm, they should Jun 05 21:53:28 mickeyl: around? Jun 05 21:53:32 mrmoku: here dependencies seem ok: http://tinderbox.openembedded.net/packages/570514/ Jun 05 21:53:41 * mrmoku reading OE manual Jun 05 21:53:52 no... should not :P Jun 05 21:53:58 hu Jun 05 21:54:21 Then it was pure luck it worked until now :) Jun 05 21:54:36 onen|openBmap_, that is the build dependencies Jun 05 21:54:53 the recipe is wrong... because python-pygtk is a runtime dependency Jun 05 21:55:05 rhkfin: thanks for pointing this! Jun 05 21:55:06 and it works in SHR because we have pyphoneloge which pulls it in :P Jun 05 21:55:51 onen|openBmap_: nice to hear I did something useful :) I guess you're the people to take this high enough so all OM/SHR/etc users get the fix? Jun 05 21:56:02 hmm... not sure though... don't know about that python stuff ;) Jun 05 21:56:10 mrmoku: stefan_schmidt: lucky us :-) you know my OE knowledge is close to 0, so I have to ask for help on this :'-) Jun 05 21:56:44 * mrmoku tries to move python-pygtk to RDEPENDS and to rebuild then Jun 05 21:58:23 what's the difference of DEPENDS and RDEPENDS? Jun 05 22:00:08 DEPENDS is stuff needed to *build* a package - RDEPENDS is stuff needed to *run* a package Jun 05 22:00:20 ah, ok, understand, thanks. Jun 05 22:00:25 have to go now Jun 05 22:00:25 thanks all for you help Jun 05 22:00:25 will come back to you tomorrow Jun 05 22:00:25 good night everyone Jun 05 22:00:29 onen|openBmap_: CU Jun 05 22:00:33 onen|openBmap_, night :) Jun 05 22:00:35 mrmoku: then I will check if this is all correct tomorrow Jun 05 22:00:39 thanks agana Jun 05 22:00:40 gain Jun 05 22:00:42 night Jun 05 22:01:27 stefan_schmidt, how easy are the bubbles to hit? :P Jun 05 22:02:01 mrmoku: Haven't tried it yet. At least they have been designed together with some usability guys. :) Jun 05 22:02:18 Should be big enough Jun 05 22:02:19 uhh... that might be good... or bad ;) Jun 05 22:02:28 * stefan_schmidt thinks it is good. Jun 05 22:02:33 :) Jun 05 22:02:57 To much of our software is designed only by engineers who love the technical details and not the user experience :) Jun 05 22:02:58 mrmoku: did you make the changes to OE -> it'll be available for all distributions? Jun 05 22:03:05 stefan_schmidt: i agree! Jun 05 22:03:22 rhkfin, testing if it builds now Jun 05 22:03:27 mrmoku: great Jun 05 22:03:36 if that works out I will commit it to the shr/import branch in OE Jun 05 22:03:58 ok Jun 05 22:04:43 Om2009 is being built from the fso/ms5.5 branch though Jun 05 22:04:52 it will not pick it up automatically Jun 05 22:04:59 :( :( Jun 05 22:05:18 don't know who is in charge of Om2009 nowadays... Jun 05 22:05:23 mrmoku: rhkfin: I will take care of it being pushed everywhere Jun 05 22:05:33 mrmoku: nytowl for some days still.. Jun 05 22:05:40 real good night ;-) Jun 05 22:05:42 onen|openBmap_: great! Jun 05 22:05:53 rhkfin, ok, then onen|openBmap_ will just tell him to cherry pick it ;) Jun 05 22:09:18 hmm... it built nicely... though I'm not sure about that Jun 05 22:16:37 dos1|neo, triggered rebuild of images and feed Jun 05 22:16:46 and off to bed now :) Jun 05 22:16:49 good night all Jun 05 23:48:20 Hello, I wanna buy a bluetooth headset, Can anyone help to let me know which type of device FR supported? Thanks in advance **** ENDING LOGGING AT Sat Jun 06 02:59:58 2009