**** BEGIN LOGGING AT Sun Nov 21 02:59:58 2010 Nov 21 04:09:11 is there a good reason to use ubi rather than jffs2?... Nov 21 07:28:48 antrik: http://www.linux-mtd.infradead.org/doc/ubifs.html#L_overview Nov 21 07:39:31 gena2x: I already read that some time ago, and it sounded rather "meh" to me... Nov 21 07:40:11 the only features compared to jffs2 seem to be journalling and faster mounts of large devices. is either of these relevant for openmoko?... Nov 21 07:43:51 antrik: ubifs is significantly faster Nov 21 07:45:17 especially on write Nov 21 07:45:48 i am using qtmoko with ubifs and it if for sure faster Nov 21 07:46:06 even it uses slowest possible algorithm Nov 21 07:46:18 *compression algorithm Nov 21 07:46:38 *it is Nov 21 07:47:23 is this noticable with typical usage patterns of the phone?... Nov 21 07:48:59 nobody did direct mathmatically correct comparison, but its better, also allow mapping files, also consumes less ram. Nov 21 07:49:15 nitice that http://www.linux-mtd.infradead.org/misc.html Nov 21 07:49:41 in qtmoko, slowest compression used (zlib + ubifs) Nov 21 07:49:59 if you planning to do any comparison Nov 21 07:50:22 but really, boot time for example decreased Nov 21 07:50:42 *notice Nov 21 07:51:41 antrik: you can boot from sd, format your flash and try to compare fses with some bonnie++ for example Nov 21 07:52:00 would be interesting to read results Nov 21 07:54:24 nah, I'm not interested in artificial benchmarks... I'd like to know whether it's actually noticably faster in everyday usage :-) Nov 21 07:54:46 yes :) Nov 21 07:55:27 i added +20% by hw ecc and disabling crc checks Nov 21 07:55:30 alsa Nov 21 07:55:36 also Nov 21 08:04:29 OK, I'll give it a try... Nov 21 09:28:15 gena2x: http://gitorious.org/~jama/htc-msm-2-6-32/openmoko-kernel/commits/shr-2.6.34 doesn't look like useless all-in-one patch Nov 21 09:28:58 and all patches from om-gta02-2.6.32 branch there should be included, but I'll check Nov 21 09:44:12 hrm... I have 31 MB less space available than with jffs2... Nov 21 09:45:06 is that expected? Nov 21 09:58:35 <[Rui]> JaMa|Off: moin, I'm getting a quite familiar white screen bug! Nov 21 10:01:34 jffs2 doesn't know how much space left by design Nov 21 10:01:53 df just show "expected" but nobody knows how much space free Nov 21 10:05:49 nah, df seems to show the amount of space actually used (compressed) Nov 21 10:06:59 the difference is in the "total": while jffs2 shows something around 250, with ubifs it's only 221... Nov 21 10:07:36 I hope it doesn't use up 30 MB for the journal?... Nov 21 10:08:51 (boot is really *way* faster BTW... though unfortunately launching X and e still needs a lot of time, so boot is only somewhat less annoying) Nov 21 10:43:34 hi PaulFertser Nov 21 10:53:37 hi DocScrutinizer Nov 21 11:03:06 GNUtoo|laptop: sorry, was already sleeping in peace last night Nov 21 11:03:19 did you get gena2x's message regarding depmod? Nov 21 11:03:19 ok np I solved the problem Nov 21 11:03:22 with some help Nov 21 11:03:22 ok Nov 21 11:03:32 basically the kenrel I flashed was 0 in size Nov 21 11:03:45 so it kept the previous kernel Nov 21 11:03:51 which didn't match the module dir Nov 21 11:04:08 ahh, ok Nov 21 11:04:39 mrmoku, I tried usb host, same problem than before Nov 21 11:04:43 still not working for me Nov 21 11:05:15 hmm... can't help you with hostmode yet Nov 21 11:05:21 did not really try it yet Nov 21 11:05:48 I must wait for PaulFertser or DocScrutinizer51 Nov 21 11:08:41 morning Nov 21 11:08:47 moin daniele_athome Nov 21 11:09:05 hi mrmoku Nov 21 11:11:02 GNUtoo|laptop: hmm... usb does not work for me... though I guess it is more a problem with ArchLinux on my laptop... as it is the same with gta02 and n900 and worked before Nov 21 11:11:24 usb host or client? Nov 21 11:11:35 ok Nov 21 11:11:35 ssh over usb Nov 21 11:11:45 ok Nov 21 11:11:51 ip is fine on both sides Nov 21 11:11:55 ping not working though Nov 21 11:11:57 ok Nov 21 11:12:02 if you see usb0 Nov 21 11:12:06 and ip is fine Nov 21 11:12:11 it's not a problem on n900 Nov 21 11:12:15 anyway do that : Nov 21 11:12:30 *put a script for extracting the firmwares Nov 21 11:12:35 connect to wifi Nov 21 11:12:40 right now I'm on gta02 Nov 21 11:12:47 ok Nov 21 11:12:55 the script is available on the shr wiki Nov 21 11:13:00 ok Nov 21 11:13:09 good... as the firmware is still missing on my n900 :) Nov 21 11:14:23 also be sure to extract the firmwares before connecting Nov 21 11:14:42 btw we might want to modify the 70-persistant-net-rules Nov 21 11:14:44 for n900 Nov 21 11:14:57 because the wlan0 still goes to wlan1,wlan2 etc... Nov 21 11:15:20 mrmoku, could you trigger mokojeweled rebuild please? Nov 21 11:15:29 it hasn't passed through latest EFL bump Nov 21 11:15:34 .so name changed Nov 21 11:16:05 daniele_athome: guess we have to bump efl again for the keyboard sizing fix Nov 21 11:16:10 or do we already have that? Nov 21 11:16:23 mrmoku, i don't know Nov 21 11:16:43 mrmoku, if the top-bar would be fixed it would be great Nov 21 11:16:51 could a bump fix that? Nov 21 11:16:53 guys x11 engine on freerunner is damn slow Nov 21 11:16:57 TAsn: any hint on usb with FR under ArchLinux? Nov 21 11:17:00 it's a shame no one wants to maintain it anymore Nov 21 11:17:03 GNUtoo|laptop: hmm... no idea Nov 21 11:17:05 raster even wants to kill it :( Nov 21 11:17:19 i mean killing x11-16 Nov 21 11:33:07 <[Rui]> daniele_athome: hi Nov 21 11:33:12 hi [Rui] Nov 21 12:03:53 GNUtoo|laptop: hmm... the battery script is missing in the wiki Nov 21 12:04:26 use fuel cells! Nov 21 12:04:49 :P Nov 21 12:05:57 ok Nov 21 12:05:58 then add it Nov 21 12:06:28 * DocScrutinizer ponders idly to tweak N900 BSI into a HDQ for smart battery :-D Nov 21 12:06:44 hmm.... yes Nov 21 12:06:55 and integrate charging in fso Nov 21 12:07:09 a pity there's no smart BL-5J out there Nov 21 12:08:32 yeah, I know Mickey thinks fso should manage battery Nov 21 12:08:46 a really weird concept Nov 21 12:09:19 not more weird than bme managing it ;) Nov 21 12:09:29 you think it should be handled in kernel only? Nov 21 12:09:52 sure more weird, as fso framework is even further away from low level Nov 21 12:10:24 well... fsodeviced is not further away IMHO Nov 21 12:10:36 I think charging should work in init S Nov 21 12:10:41 it is directly one level up from kernel Nov 21 12:10:46 hahaha Nov 21 12:10:50 yeah, then it should be handled in kernel Nov 21 12:11:01 anyway... brb Nov 21 12:11:47 when fso is one level above kernel, then I assume kernel is hurd and based on d-bus? Nov 21 12:14:46 DocScrutinizer, hi Nov 21 12:14:58 yes, meanwhile I'm supporting the idea to have bq27200 basic management in kernel (as in "a wd kicker comes with those sysnodes used to set bq27200 values") Nov 21 12:14:59 I can't get usb host working, what's the way to make it work? Nov 21 12:15:24 DocScrutinizer, good idea, I started to do kenrel work with bq27 readings Nov 21 12:15:32 but I didn't test it yet Nov 21 12:15:48 install powerkernel, install h-e-n or follow the "manual instructions" using echo and vboost script Nov 21 12:15:51 I need a better setup for that Nov 21 12:15:56 that's what I did Nov 21 12:16:05 what is vboost? Nov 21 12:16:18 sh booston ? Nov 21 12:16:21 enable VBUS Nov 21 12:16:26 yes but how? Nov 21 12:16:42 because I seem to have the sames issues than last time Nov 21 12:16:48 lots of errors Nov 21 12:16:50 etc... Nov 21 12:17:35 GNUtoo|laptop: check http://talk.maemo.org/showthread.php?t=65232 Nov 21 12:18:14 90% success, nobody had unsolvable issues making hostmode work for *some* devices Nov 21 12:19:48 GNUtoo|laptop: please don't ask like "doesn't work for me - please give generic instructions!" Nov 21 12:20:06 I've done one time too often Nov 21 12:20:10 ok Nov 21 12:20:24 I followed the instructions at: Nov 21 12:20:38 https://garage.maemo.org/plugins/ggit/browse.php/?p=h-e-n;a=commit;h=d7f76e505b16f7e9305c59c51d02fb1473ab5b2c Nov 21 12:20:50 either you come with a particular problem with detailed error description, or - better - you make it work like described n this thread Nov 21 12:21:26 I'll pastebin what I did + dmesg Nov 21 12:21:56 and you think I'll debug your process? Nov 21 12:22:44 and I do a 'diff' between your process and the way it worked for everybody? Nov 21 12:24:32 I thinks if you change the process then it's up to you to do that "diff" and come up with a detailed analysis and a proper good question why it fails in one particular point (and describe in which way it fails at that point) though you tried to make sure it should work same way like in http://talk.maemo.org/showthread.php?t=65232 Nov 21 12:25:59 ok Nov 21 12:26:06 I'm reading the code of the GUI right now Nov 21 12:26:35 what I wanted to know is how to do it without the GUI Nov 21 12:26:42 but I'll finish reading the GUI Nov 21 12:27:13 DocScrutinizer: above kernel because it is interacting with the kernel interfaces directly... sysfsnodes... rfkill... Nov 21 12:27:14 the gui is just doing a sophisticated set of echo foo >bar and system() Nov 21 12:27:33 yes but I don't find how to activate vbus Nov 21 12:28:01 mrmoku: no doubt it's above kernel. Question is how far above. Can it run in init S ? Nov 21 12:28:09 jrbme will Nov 21 12:28:11 ah ok Nov 21 12:28:14 dummy = os.popen('/root/vboost.sh &').read() Nov 21 12:28:18 vhoost.sh Nov 21 12:29:02 though jrbme is a silly scriptie Nov 21 12:29:20 DocScrutinizer: so you think the correct solution would be a kernel module exporting state via sysfs + a small daemon handling the charging? Nov 21 12:29:29 yes Nov 21 12:29:50 where can I find vboost.sh? Nov 21 12:29:52 what state is the bq module in the power kernel in Nov 21 12:29:59 would that be a start to continue upon? Nov 21 12:30:25 inside the h-e-n pkg, on h-e-n git/docs in paul's files... Nov 21 12:30:32 ok Nov 21 12:30:58 mrmoku: there's no bq24150 kernel module yet Nov 21 12:31:04 (I was looking there: https://garage.maemo.org/frs/?group_id=1660 ) Nov 21 12:31:45 harmattan allegedly comes with such a module. I'm convinced it will be crap once more, like all things Nokia does to "fix" shit on kernel level Nov 21 12:33:55 mrmoku: I created a API spec for bq24150 kernel module (sysfs nodes sematics) - finished 70%. jacekowski had a look into it, but stopped development when he realized it breaks bme (same occupies-I2C issue like bq27200 module) Nov 21 12:34:33 ok Nov 21 12:34:58 DocScrutinizer: power-kernel has a patch for that i2c-occupied problem btw. Nov 21 12:35:27 the "daemon" is jrbme, and I'm trying to encapsulate bq24150 access there in, so it's easily adapted to new sysfs situation Nov 21 12:35:46 mrmoku: I know it has... for bq27200 Nov 21 12:35:50 playya__: i thought about it, but we can't just change the interface to include the 'sender'. It doesn't make sense to have the client give that information explicitly. If we do that, clients could register for other clients - e.g. clients which are not even running - and this makes everything fall apart. We need to convince Jürg to add that functionality. Nov 21 12:35:56 * mrmoku got to go with family to a museum now Nov 21 12:36:02 enjoy Nov 21 12:36:03 we have to have access to the caller Nov 21 12:36:10 DocScrutinizer: well... will try :P Nov 21 12:36:16 (to enjoy) Nov 21 12:36:19 lo mickeyl Nov 21 12:36:24 hi doc Nov 21 12:36:24 bbl Nov 21 12:49:27 DocScrutinizer, thanks a lot for the vboost info Nov 21 12:49:31 it seem to work Nov 21 12:49:48 http://pastebin.com/YXwB6Uxu Nov 21 12:51:28 * GNUtoo|laptop will read to find out if hubs are suported Nov 21 12:51:31 specially powered ones Nov 21 12:52:05 GNUtoo|laptop: hey Nov 21 12:52:07 GNUtoo|laptop: i'm back :) Nov 21 12:52:11 hi Nov 21 12:52:14 I'm trying usb host Nov 21 12:52:18 GNUtoo|laptop: nice Nov 21 12:52:21 I got it to recognize my hub Nov 21 12:52:31 but it doesn't find the stuff that is connected to the hub Nov 21 12:53:07 GNUtoo|laptop: hm, that's strange. After the hub is enumerated everything else should get recognised automatically. Nov 21 12:53:14 GNUtoo|laptop: is it HS or FS hub you use? Nov 21 12:53:29 2.0 Nov 21 12:53:33 so... Nov 21 12:53:41 GNUtoo|laptop: so High Speed. Nov 21 12:53:42 echo hosth > /sys/devices/platform/musb_hdrc/mode Nov 21 12:53:55 GNUtoo|laptop: yes, then enable vboost, the enumerate. Nov 21 12:54:05 ok Nov 21 12:54:11 I did that Nov 21 12:54:15 let me try something Nov 21 12:56:03 PaulFertser: any clue why mohammad, in his gui, doesn't enable vboost automatically, after maybe a sleep 2 ? Nov 21 12:56:37 DocScrutinizer: probably to not spend extra power when the hub/device is self-powered. Nov 21 12:56:52 DocScrutinizer: i mean it should work if it feeds some power to the N900 side too. Nov 21 12:56:56 PaulFertser: btw isn't F = force session, not force enum? Nov 21 12:57:22 DocScrutinizer: well, i think i abused it :) Nov 21 12:58:10 PaulFertser: yes, that's what I thought as well, but then this gui is done in an odd way, as it should either apply vbus automatically, or instruct user in clear letters to do that with external power Nov 21 12:58:23 PaulFertser, still the same issue, here what I did: Nov 21 12:58:28 reboot Nov 21 12:58:31 PaulFertser: ohnoes? Why? Nov 21 12:58:48 echo hosth > /sys/devices/platform/musb_hdrc/mode Nov 21 12:58:50 you redefined F to mimic E ? Nov 21 12:58:56 sh vboost.sh Nov 21 12:59:10 echo F > /proc/driver/musb_hdrc Nov 21 12:59:47 GNUtoo|laptop: you tried same setup with original h-e-n? Nov 21 12:59:53 http://pastebin.com/qYkUZY37 Nov 21 12:59:55 DocScrutinizer: there was no special "E" thing anyway. I felt that what "F" did was useless for our purposes so i made it do not only force session but also to mimic a CONNECT request. Nov 21 13:00:03 I'll retry with the original h-e-n script Nov 21 13:00:11 DocScrutinizer: there was no ready-to-use "E". Nov 21 13:00:13 PaulFertser: there was Nov 21 13:00:23 DocScrutinizer: it didn't do what was needed then. Nov 21 13:00:57 GNUtoo|laptop: everything's right, only your defconfig is wrong :) Nov 21 13:00:57 but I didn't figure out if the vboost.sh should be run before or after doHostmodeC( Nov 21 13:01:05 PaulFertser, ah? Nov 21 13:01:13 what is missing? Nov 21 13:01:17 GNUtoo|laptop: you did everything right in the userspace, but your config blacklists hubs. Nov 21 13:01:18 ahhh Nov 21 13:01:23 I got it Nov 21 13:01:29 627 * e enumerate Nov 21 13:01:30 usb_bus_start_enum(bus,bus->otg_port); Nov 21 13:01:31 I took mrmoku's defconfig Nov 21 13:01:48 and I forgott to enable the right CONFIG_ Nov 21 13:03:51 GNUtoo|laptop: see? that's why every devel refuses to debug derived work unless the user with the problem made sure it's consistent with original pkg behaviour Nov 21 13:04:38 I.E. use the original pkg and test if it works there. Then do a proper 'diff' Nov 21 13:06:25 PaulFertser: I'm not happy with breakage of sarahn's 'API' - I can't run *my* hostmode scripts anymore Nov 21 13:11:36 also see Moh's HeFHIH sequence Nov 21 13:11:43 'e'!! Nov 21 13:21:26 PaulFertser: see https://garage.maemo.org/pipermail/h-e-n-devel/2010/000022.html for a complete "commented" list of /proc/driver/musb_hdrc API Nov 21 13:22:44 I think 'a lot of' people got used to that, and it proved to be incredibly useful for experimenting with USB. I don't think it's a good idea to break it for no good reason Nov 21 13:23:24 mickeyl, ping Nov 21 13:23:28 it's especially deprecated to change semantics without further notice Nov 21 13:23:42 mickeyl, you know, i'm starting to think that i should use a separate config file for those xml tags i proposed Nov 21 13:23:51 :) Nov 21 13:24:17 but some attributes should pass through anyway Nov 21 13:25:19 e.g. Nov 21 13:25:31 it's just a simple attribute to add to the DTD Nov 21 13:26:48 afternoon Nov 21 13:27:29 does anybody have xrandr running on anything else than a gta02? Nov 21 13:27:41 * zub would like to do a little test Nov 21 13:28:55 so I'm looking for a volunteer :) Nov 21 13:42:57 wow I got an usb keyboard working Nov 21 13:43:07 GNUtoo|laptop: what about the hub? Nov 21 13:43:26 n900<->hub<->keyboard Nov 21 13:44:39 DocScrutinizer: in fact all those commands were never there. To the contrary, my latest patch adds some new actually useful commands. Only thing that i agree is wrong is changing the 'F' behaviour. But most probably it's not that useful for experimentation anyway. Nov 21 13:45:15 wat means "never there"? o.O I used them! Nov 21 13:45:31 DocScrutinizer: they were in sarahn's patches only, not in mainlie. Nov 21 13:46:09 we're not exactly talking about mainline, when we do h-e-n, right? Nov 21 13:46:11 DocScrutinizer: if you want, i can prepare some patch to change the proc node behaviour. Nov 21 13:46:32 DocScrutinizer: well, my goal was to provide minimal patchset on top of stock maemo kernel. Nov 21 13:47:49 apreciated. But introducing a changed semantics using an already existing API - though not in mainline - isn't exactly a good idea anyway Nov 21 13:49:17 DocScrutinizer: i assumed that using 'F' for what it was used in sarahn's patches is totally irrelevant because it was just testing kernel for experimentation, there were very few users of the API etc etc. Nov 21 13:50:24 until today I thought h-e-n hostmode enabled kernel would keep the sarahn's patches as they don't hurt and you even seemed to use them. Or at very least they would be easy to apply to whatever kernel is out there, if you are interested in using them. Won't work for new PF h-e-n kernel as it has a conflicting semantics of F, without any good reason Nov 21 13:50:29 DocScrutinizer: also i have never seen the sarahn "api" helping with actually understanding the mechanics of the hostmode, it wasn't useful for me at all, so i didn't feel any need to reimplement it. Nov 21 13:51:33 sarahn's patches to hurt because they are not really comprehensible and also trying to do useless and probably harmful/conflicting things. Nov 21 13:52:00 well, *I* used them and they helped to achieve proper hostmode: https://garage.maemo.org/pipermail/h-e-n-devel/2010/000025.html Nov 21 13:52:39 DocScrutinizer: if you tell me to reimplement the sarahn's procfs nodes, i'll do that. Nov 21 13:53:14 DocScrutinizer: i'm not sure that was "proper" hostmode. It looked more like the trick i've seen when i re-enabled interrupts. Nov 21 13:53:24 I'm not to tell you anything - I just can utter my thoughts and concerns Nov 21 13:53:29 DocScrutinizer: that is, during operation the FORCED_HOST was actually off. Nov 21 13:55:09 DocScrutinizer: if you want to do more experimentation, let's discuss what exact facilities you need for that, i'd be happy to add those. Nov 21 13:55:46 Currently it looked like everybody's happy with the semi-stable solution i arrived at. Nov 21 13:57:15 well, I suggested quite some BS to people having problems with it, due to being unaware the things I suggested were'nt there or even changed their semantics Nov 21 13:57:16 DocScrutinizer: i mean when one is really involved in low-level stuff, trying to think about hw state machine transitions etc etc, it doesn't really matter what letter exactly is used for the debugging/experimentation purposes, it's more like his own private API he changes all the time, trying different approaches etc. Nov 21 13:57:58 made me look like an idiot Nov 21 13:58:00 That's how i see it and that's the exact reason i felt free to "break" it. I might be wrong. Nov 21 13:58:13 I'm sorry. Nov 21 14:00:56 mrmoku, ping Nov 21 14:01:11 nota bene I'm not interested in anything beyond what's sufficiently specified with my summary in https://garage.maemo.org/pipermail/h-e-n-devel/2010/000022.html, for procfs/sysfs Nov 21 14:01:34 no patches to the statemachine itself or anything Nov 21 14:02:01 just the API to call the different procedures / set values of vars Nov 21 14:05:03 mrmoku, could you remove peripherical whitelist and usb hub blacklist in usb host stuff for your linux-nokia900 kernel Nov 21 14:05:04 ? Nov 21 14:05:14 because i've tested and it works Nov 21 14:05:20 (usb host) Nov 21 14:05:44 but in order to work the defconfig must be changed as I jsut said upper Nov 21 15:24:39 GNUtoo|laptop: ok Nov 21 15:36:00 PaulFertser: CONFIG_USB_OTG_BLACKLIST_HUB=y that one must be n, right ? Nov 21 15:36:11 (or not set) Nov 21 15:36:18 mrmoku: right Nov 21 15:36:32 ok Nov 21 15:41:48 mrmoku, yes Nov 21 15:42:00 mrmoku, but there is also a CONFIG WHITELIST that must be N Nov 21 15:45:22 ahh, ok Nov 21 15:54:15 mrmoku, yes, just look in the om wiki Nov 21 15:54:23 mrmoku, it's the same as in any other distro though. Nov 21 16:53:42 hi, is headphones supposed to work(that is to say mplayer doesn't block) on om-gta02? Nov 21 16:54:11 elisa42 is not there so I can't ask Nov 21 16:55:41 GNUtoo|laptop: is that some shr specific question? Nov 21 16:55:52 I think so Nov 21 16:55:55 but I'm not sure Nov 21 16:56:53 the log is full of: Nov 21 16:56:57 nanosleep({0, 32000000}, NULL) = 0 Nov 21 16:56:58 ioctl(5, 0x806c4120, 0xbea373d8) = 0 Nov 21 16:57:07 so maybe that's a kernel problem Nov 21 16:57:23 yuck Nov 21 16:59:00 JaMa|Off, here's the patchset: Nov 21 16:59:04 Source: http://kernel.org/pub/linux/kernel/v2.6/linux-2.6.34.tar.bz2;name=kernel http://kernel.org/pub/linux/kernel/v2.6/patch-2.6.34.7.bz2;apply=yes;name=stablepatch file://ARM-Add-support-for-LZMA-compressed-kernel-images.patch;status=pending file://defconfig file://openmoko.patch file://shr.patch Nov 21 16:59:20 it's the standard just opkg upgraded image Nov 21 16:59:26 from shr feeds Nov 21 16:59:46 (It is running stock and official shr-unstable, not built by me) Nov 21 17:46:04 Is anyone on? I have questions Nov 21 17:47:04 mrmoku: just ask Nov 21 17:47:08 MarioMaster100: ^ Nov 21 17:48:15 Ok, this might seem stupid, but: what are the benefits if the shr and the idroid? That is still unclear to me. Nov 21 17:49:54 Also being on my itouch on webchat I'll have to periodically logout and back to see the chat log as it eventually stops showing it :/ Nov 21 17:50:44 MarioMaster100: can you rephrase the question in other words please? Nov 21 17:52:08 Ok simpler -> what benefits does the shr and the idroid give to the (since I only have this device) itouch? Nov 21 17:52:37 MarioMaster100: comparing to what? Nov 21 17:53:55 :( it stopped showing chatlog how about you pm me until the pm log stops showing on webchat on my device? I can't see what you said :( Nov 21 17:54:45 MarioMaster100: to iPhoneOS? Nov 21 18:06:41 * MarioMaster100 thinks it's showing the log again Nov 21 18:52:22 mrmoku, hi Nov 21 18:55:55 mrmoku, can I prepare and push your kernel Nov 21 19:14:18 GNUtoo|laptop, headsets plug&play isn't working anymore for me since a pair of weeks ago Nov 21 19:14:49 I think [Rui] said same yesterday or 2 days ago Nov 21 19:15:10 ok Nov 21 19:15:11 thanks Nov 21 19:15:12 no headset detection is bad, as I want to try enjoy! Nov 21 19:15:16 T_T Nov 21 19:16:25 I want to enjoy in general, what should i do? I'm afraid fixing headset detection wouldn't be enough. Nov 21 19:17:35 GNUtoo|laptop: though radare is cool, i do not enjoy it much, i'm still rather far from the objective i had in mind. Nov 21 19:17:54 ok Nov 21 19:18:02 PaulFertser, well, it would be enough for me :D Nov 21 19:18:45 no headset detection mean no music in the underground in the morning :( Nov 21 19:19:18 pespin: why can't you switch alsa profile manually? Nov 21 19:19:27 pespin: and why can't you enjoy reading instead of music. Nov 21 19:19:39 pespin: btw, what do you plan to listen to tomorrow? Nov 21 19:19:48 PaulFertser, Depending on the day I read or I listen to music :) Nov 21 19:21:27 but no book atm, had to return "operating System Concepts" book which I was reading to the univ bibliotheque as I couldn't have it more days. Have to go to borrow it another time, but haven't had time this week :P Nov 21 19:22:17 pespin: you can download something from the net and read it on your FR with FBReader ;) Nov 21 19:22:38 PaulFertser, I tried, but screen is too small for me. Don't like reading too much there Nov 21 19:24:40 btw, with "enjoy", I meant a new EFL music player which we have managed to use in SHR since some weeks ago, and still has to be pushed into OE. I wanted to try how well dos it play in daily basis and take best gstreamer backend :) Nov 21 19:24:57 and set that in OE recipe Nov 21 19:25:23 JaMa|Off: i didn't mean 'useless' exactly. i mean it is not as easy to browse all-in-one patch file with 20 patches. with your link it's much easier :) But anyway, patch is missing, that's why people getting white screens. Nov 21 19:25:31 GNUtoo|laptop: sure Nov 21 19:25:58 ok Nov 21 19:27:12 mrmoku, well, how's arch treating ya? Nov 21 19:27:35 TAsn: heh, good so far... apart from my 32bit and my usb problems Nov 21 19:27:46 for 32bit I had to install a chroot Nov 21 19:27:47 lol, so not good at all Nov 21 19:27:52 mrmoku, really? why is that? Nov 21 19:27:59 because multilib does not have all needed libs Nov 21 19:28:03 I got a multilib sys Nov 21 19:28:05 mrmoku, oh Nov 21 19:28:08 the missing one for me was libusb Nov 21 19:28:25 mrmoku, I change the kernel for nokia900 machine? Nov 21 19:28:27 anyway... chroot works fine too for what I need Nov 21 19:28:33 s/kernel/default kernel Nov 21 19:28:57 mrmoku, cool. Nov 21 19:29:00 GNUtoo|laptop: hmm... who's using it? Nov 21 19:29:10 for SHR changing it is what we want Nov 21 19:29:24 dunno if there are still others in OE land caring about n900 Nov 21 19:29:31 mrmoku, I don't know, I bet only people who want compatibility with maemo Nov 21 19:29:53 GNUtoo|laptop: who's maintainer for the nokia900.conf? Nov 21 19:30:01 no idea Nov 21 19:30:04 <[Rui]> pespin: yes, doesn't work for me with two or 3 weeks old image Nov 21 19:30:07 we can see the commiters Nov 21 19:30:11 but maintainers.... Nov 21 19:30:21 <[Rui]> I updated and haven't tested headset again... Nov 21 19:30:21 TAsn: the usb problems are strange... works with maemo, but does not with SHR on n900 and gta02 Nov 21 19:30:29 <[Rui]> need to reboot for preupgrade into Fedor 14 Nov 21 19:30:45 <[Rui]> will take a while, but hopefully allright Nov 21 19:30:51 basically: Nov 21 19:30:52 [Rui], ok :) uhm I remember telling you 2 things yesterday night. One was about headset, but don't remember the other one :S Nov 21 19:30:53 [Rui]: I just dumped my Fedora 14 in favour of ARchLinux :P Nov 21 19:31:03 *hrw: Nov 21 19:31:03 mrmoku, did you check usb0 vs eth1? Nov 21 19:31:07 (something I fell for) Nov 21 19:31:14 I'm not sure he still do oe work or still uses oe Nov 21 19:31:18 as he work for canonical Nov 21 19:31:21 TAsn: I'm using networkmanager... I don't care ;) Nov 21 19:31:29 i.c. Nov 21 19:31:31 btw Nov 21 19:31:32 the ip is set correctly on both sides Nov 21 19:31:39 rumor has it connman is getting matured Nov 21 19:31:43 *mature Nov 21 19:31:50 would be about time :P Nov 21 19:31:57 and rschuster who also uses it for maemo compat Nov 21 19:32:04 is there something ready to use for connman in efl land? Nov 21 19:32:10 woglinde said I could go ahread when changing the defconfig Nov 21 19:32:13 some time ago Nov 21 19:32:22 GNUtoo|laptop: ok, just change it then :) Nov 21 19:32:25 ok Nov 21 19:32:35 that count as an ack? Nov 21 19:32:48 yes Nov 21 19:33:19 http://pastebin.com/JfVdyG8P Nov 21 19:33:21 for reference Nov 21 19:33:22 ok Nov 21 19:33:31 mrmoku, yes. Nov 21 19:33:43 mrmoku, the default networking manager which it's name I can't remember Nov 21 19:33:46 uses connman Nov 21 19:33:51 just ask @ #e Nov 21 19:34:03 GNUtoo|laptop: looks good Nov 21 19:34:03 exalt? Nov 21 19:34:29 * pespin uses wicd on his netbook with archlinux Nov 21 19:34:53 pespin: that sounds like wireless only? Nov 21 19:35:02 mrmoku, it accepts wired too :) Nov 21 19:35:08 openvpn? Nov 21 19:35:11 it's faar less intrusive than network-manager Nov 21 19:35:33 mrmoku, uhm I configure openvpn with config file + /etc/rc.d/openvpn start Nov 21 19:35:39 wicd is crap. Nov 21 19:35:40 :P Nov 21 19:35:42 honestly. Nov 21 19:35:44 I hate it. Nov 21 19:35:55 mrmoku, we used it (or at least I used it) on the fr Nov 21 19:35:58 for a short while Nov 21 19:36:19 pespin, iirc it's the one that mangles with wpa_suppliant.conf? Nov 21 19:36:24 and for eduroam APs I use a script which calls wpa_supplicant Nov 21 19:36:30 in a shitty matter Nov 21 19:36:33 pespin: I have lots of servers I access via openvpn... would get mad by changing configs around Nov 21 19:36:40 with nm it's just two clicks Nov 21 19:36:48 hmm no idea then hehe Nov 21 19:37:00 let me check Nov 21 19:37:52 mrmoku, I think pespin got the name right (exalt) Nov 21 19:37:56 I honestly can't remember though Nov 21 19:38:07 because I've hated connman for too long Nov 21 19:38:08 :) Nov 21 19:38:21 TAsn: so we should bump connman and try that exalt thing Nov 21 19:38:26 TAsn, I think wicd can use wpa_supplicant, but I'm not sure Nov 21 19:38:59 pespin, iirc they had a hack Nov 21 19:39:03 but that was a long time ago Nov 21 19:39:07 like 2-3 years Nov 21 19:39:08 :P Nov 21 19:39:22 hm.. maybe 2 Nov 21 19:39:32 yeah, it uses it: wpa_supplicant -B -i wlan0 -c /var/lib/wicd/configurations/00a0c5814c2b -D wext Nov 21 19:40:01 I remember they were more hackish :P Nov 21 19:40:05 but anyhow, cool. Nov 21 19:40:20 I'd still refrain from using a gtk+ ui Nov 21 19:42:47 btw, the other day I tried wifi-radar in the FR, and it made the Xserver crash afair Nov 21 19:42:59 but this should be fixed now, right? Nov 21 19:43:32 once the "no more WSOD" patch is in, I'll opkg update and try it :) Nov 21 19:45:49 wifi-radar was also hackish an crap Nov 21 19:45:50 iirc Nov 21 19:46:16 and honestly: nm-applet is super awesome, no reason to use anything else (except if that something else uses efl) Nov 21 19:47:13 TAsn: we could try NM again on the FR :P Nov 21 19:47:47 did we ever? Nov 21 19:47:55 because devilhorns Nov 21 19:47:58 well not officially Nov 21 19:48:03 fixed a bug with systray icons Nov 21 19:48:03 I played with it long ago Nov 21 19:48:06 only recently Nov 21 19:48:10 (2 wks ago) Nov 21 19:51:38 I'd like to try empathy too, but current versions in OE doesn't compile right and newer versions require glib-2.26 u.u Nov 21 19:52:14 ohh... 2.26 Nov 21 19:52:18 everybody needs that :P Nov 21 19:52:28 yeah Nov 21 19:52:36 I tried, but there are some patches that don't apply anymore Nov 21 19:52:54 pespin, hi were you interested in enjoy, what's the status? Nov 21 19:52:57 and I'm not the appropriate person to do that :P Nov 21 19:53:01 shall we ask the guys in #oe to do it? Nov 21 19:53:31 pespin: that would be a good thing :-) Nov 21 19:53:53 GNUtoo|laptop, haven't played much more with it since you sent the recipe to fidencio. My headsets doesn't work automatically and haven't had time this weeks to look into it Nov 21 19:54:11 I pinged fidencio an hour ago to know when he's going to push the recipe Nov 21 19:55:12 ok Nov 21 19:55:20 did you have a response? Nov 21 19:55:23 nop Nov 21 19:55:57 maybe we think he's the one to push it and he thinks the opposite xD Nov 21 19:58:13 ok Nov 21 19:58:16 btw, some days ago a person in university showed me the samsung wave, which is using the Vada SO (which afaik uses linux+EFL) Nov 21 19:58:18 maybe I should finish it? Nov 21 19:58:28 wow!!!!!! Nov 21 19:58:31 GNUtoo|laptop, I'd wait for his response first Nov 21 19:58:33 is there some specs? Nov 21 19:58:40 for the samsung wave Nov 21 19:58:45 some details? Nov 21 19:58:46 etc... Nov 21 19:58:52 hmm let's search :P Nov 21 19:58:54 is there a GUI for opkg? Nov 21 19:59:06 they are not using opkg afaik hehe Nov 21 19:59:15 what do they use? Nov 21 19:59:24 no openembedded? Nov 21 19:59:28 install packages from web afaik Nov 21 19:59:29 or openembedded with dpkg? Nov 21 19:59:32 ? Nov 21 19:59:36 from web? Nov 21 19:59:42 but that needs a package manager Nov 21 19:59:48 yeah, you press a link and it install from web browser Nov 21 19:59:55 yeah, no idea sorry, person didn't know Nov 21 19:59:59 interesting thing Nov 21 20:00:00 yes that needs a package manager Nov 21 20:00:20 was that he got it for 50€ + some "points" from the telephony company Nov 21 20:00:29 lol Nov 21 20:00:30 ok Nov 21 20:00:32 I think it has a 1Ghz cpu Nov 21 20:00:37 ok Nov 21 20:00:42 so... really intersting Nov 21 20:00:45 * GNUtoo|laptop must find some infos Nov 21 20:00:48 let me look for some information Nov 21 20:00:55 ok Nov 21 20:01:50 seem to have no keyboard Nov 21 20:02:06 this person told me the great thing of the mobile is the camera Nov 21 20:02:13 ok Nov 21 20:02:19 the camera of n900 is also good Nov 21 20:02:28 5 mpixels, and has "augmented reality" applications (don't know if it's called like this) Nov 21 20:02:37 the one from the nexusone too but it needs proprietary stuff Nov 21 20:05:00 something like showing a simbol to the camera and then you sawy on the screen a tower instead of that simbol, with tanks/panzers and this sort of things shooting each other Nov 21 20:07:21 android has the same kind of stuff Nov 21 20:07:22 I bet Nov 21 20:07:26 I never tries it tough Nov 21 20:08:42 hehe we could try asking people like the one which did free dirvers for kinect :P Nov 21 20:15:32 pespin, http://opensource.samsung.com/ Nov 21 20:17:11 pespin, are you sure it runs the linux kernel Nov 21 20:17:11 ? Nov 21 20:17:20 what did uname -a return? Nov 21 20:17:32 afair, and couldn't run that Nov 21 20:17:39 ah? Nov 21 20:17:40 why? Nov 21 20:17:47 beause he didn't have a terminal installed Nov 21 20:17:49 :P Nov 21 20:17:52 ah ok Nov 21 20:18:11 so no real proof it runs the linux kernel Nov 21 20:18:15 nop Nov 21 20:18:20 ok Nov 21 20:18:25 I'll digg ino it in a few minuts, I'm helping my mother now hehe Nov 21 20:18:36 pespin, the Wave Nov 21 20:18:44 uses BaDa (with a B) Nov 21 20:18:44 ok Nov 21 20:18:55 and it's not EFL (or linux) Nov 21 20:18:56 based Nov 21 20:19:07 indeed that's what I thought Nov 21 20:19:09 regarding EFL I'm 100% sure, regarding linux I'm not. Nov 21 20:19:18 about linux Nov 21 20:19:33 I'm 99.9% sure about linux Nov 21 20:19:35 as well Nov 21 20:19:50 But except for the shitty OS Nov 21 20:19:51 ok Nov 21 20:19:52 ok maybe I just did remember wrong Nov 21 20:19:59 the Wave is an awesome device Nov 21 20:20:08 Used it for a while Nov 21 20:20:46 googling shows bada is linux based Nov 21 20:20:50 which as I said, I find odd Nov 21 20:21:16 heh +1 point for my memory :P Nov 21 20:21:34 http://www.linuxfordevices.com/c/a/News/Samsung-Wave/ Nov 21 20:21:39 but you can't really tell if it's true or not Nov 21 20:21:52 I bet what wikipedia said is right Nov 21 20:22:19 basically that the os supports a Linux or another kernel(probably BSD based) Nov 21 20:22:20 Samsung has said that the kernel-configurable Bada platform can be based on a Linux kernel or another real-time OS (RTOS). The company did not list the kernel configuration for the Wave, but a number of technology sites,as well as Wikipedia, flatly refer to Bada as being Linux-based, and an advanced smartphone such as the Wave would very likely require Linux. Nov 21 20:22:54 So at least the wave is proprietary based Nov 21 20:22:55 (I think) Nov 21 20:23:06 or if it's linux Nov 21 20:23:13 it's a crippled-linux (even worst than android) Nov 21 20:23:15 uhm can it run linux without having a visible linux repo? Nov 21 20:23:30 I mean easy to find Nov 21 20:23:55 GNUtoo|laptop, http://www.gsmarena.com/samsung_s8500_wave-3146.php Nov 21 20:24:15 pespin, that's the official specs Nov 21 20:24:27 I was looking for a more in-depth analysis Nov 21 20:24:38 for instance kernel sources, modem interfaces etc... Nov 21 20:24:46 Anyhow, no idea Nov 21 20:25:08 but I honestly think no linux version was released Nov 21 20:25:12 ;) Nov 21 20:25:15 indeed Nov 21 20:25:26 so there is no such thing Nov 21 20:25:36 so it's not worth to loose time on it Nov 21 20:25:40 but back to the main thing we talked about Nov 21 20:25:45 it's NOT efl based Nov 21 20:25:45 unless someone ports GNU/Linux on it Nov 21 20:25:45 :) Nov 21 20:27:31 anyhow, ciao. Nov 21 20:31:17 hi folks, i have some problems with my freerunner/shr and i don't find answers (easily) Nov 21 20:31:28 i can't seem to receive messages Nov 21 20:31:49 has the format of pim.db changed during the last couple of months, perhaps? Nov 21 20:35:10 TAsn, GNUtoo|laptop: (bada+EFL) ok, my fault read wrong this entry some days ago: http://www.osnews.com/story/22503/Samsung_Sponsors_Enlightenment_EFL_Development Nov 21 20:35:57 "It is very difficult to tell just how far along Samsung's involvement with Enlightenment is, and as such, if we will really see Enlightenment and EFL on Bada" Nov 21 20:49:05 could someone give me a hint or point me in the right direction? Nov 21 20:50:20 stevendh, I think it has changed, but when upgrading a script updating it should have been run automatically Nov 21 20:50:37 stevendh: the pim format changed twice in the past Nov 21 20:51:01 added paragraph 'disadvantages of qi' to wiki :) :http://wiki.openmoko.org/wiki/Qi#Advantages_of_Qi . Please don't kill me. Nov 21 20:51:19 pespin|netbook: IIRC for the first change it was not run automatically... not sure though Nov 21 20:51:19 just can't read it anymore Nov 21 20:51:57 pespin|netbook, mrmoku: i just used my old pim.db from my old testing and dropped it on a fresh unstable flash a few days ago Nov 21 20:52:06 could that be the cause? Nov 21 20:52:57 contraindication: my old messages/contacts/phonelog show just fine, and new messages i sent myself too Nov 21 20:55:02 stevendh: so, you receive messages, but not all? Nov 21 20:55:27 mrmoku: i haven't received a single message since reflash Nov 21 20:55:32 mrmoku, do you know when missing patched stuff against "WSOD" whill be on feeds? I'll wait till it's in to update, but I'd like to update as afaik there are great improvements in eve+https Nov 21 20:55:50 mrmoku: and i know some have been sent to me... Nov 21 20:56:32 pespin|netbook: you can compile kernel on your own Nov 21 20:57:15 pespin|netbook: or temporary use my kernel for example Nov 21 20:57:43 ah, no, mine will not fit, it is glamo-fb not kms Nov 21 20:59:00 pespin|netbook: I would like to let JaMa|Off handle that Nov 21 20:59:15 gena2x, "bitbake virtual/linux" in my shr buildtree should generate the needed kernel? or do I need to do something more? Nov 21 20:59:22 mrmoku, ok np :) Nov 21 20:59:30 stevendh: what do you mean with: "and new messages i sent myself too" ? Nov 21 20:59:40 ahh, you see them as sent messages Nov 21 20:59:41 yeah Nov 21 20:59:56 * pespin|netbook dinner time Nov 21 21:00:16 mrmoku: if i would send a message now, it would show up in the list, but i wouldn't receive the reply Nov 21 21:00:29 pespin|netbook: i don't know about shr as i am not shr guy, but you can just compile it manually -> get sources, cross-compiler and do few steps to compile Nov 21 21:00:48 stevendh: yup, it shows as sent message.. if you receive it you will have it twice Nov 21 21:01:11 TAsn: wrong pim.db format can cause messages not being received? Nov 21 21:01:36 stevendh: did you check if there is something suspicious in the logfiles? Nov 21 21:04:06 mrmoku: i checked, but i can't quite distinguish between suspicious and normal Nov 21 21:05:04 i'll have someone send a sms now, then i'll know what time to look for Nov 21 21:05:12 ok Nov 21 21:09:45 (logs seems to use UTC, right?) Nov 21 21:11:15 mrmoku: i have '[phonefsod] WARNING: GSM request still running...' Nov 21 21:13:35 and fsogsmd.log asks to report the lack of some URC handlers to Mickey Nov 21 21:18:11 stevendh: hmm Nov 21 21:18:35 stevendh: possible too that there is a message fsogsmd has problems to digest in some way... Nov 21 21:19:04 if you have another phone you could try to put the sim in there, receive the pending messages and retry with the FR Nov 21 21:19:39 or turn on DEBUG for fsogsmd and pastebin the log... or even better add a ticket to trac.freesmartphone.org Nov 21 21:20:21 mrmoku: the other direction works fine, i sent a message, it shows in the logs and it actually arrived Nov 21 21:22:44 (brb) Nov 21 21:26:05 ok, I'll push soon Nov 21 21:26:08 oops wrong channel Nov 21 21:42:24 * JaMa|Off looking at it now :) Nov 21 21:51:54 mrmoku: i've tried some calls, and now connectivity seems lost Nov 21 21:52:21 so, bug report on trac.fso.org (not shr.org?) Nov 21 21:53:25 oh, and is there some kind of script to anonymize the logs? Nov 21 22:39:28 SHR: 03mail 07libphone-ui-shr * ra9f2f3d3c51f 10/src/view/message-list-view.c: message-view: avoid to load data on two different scroll events **** ENDING LOGGING AT Mon Nov 22 02:59:58 2010