**** BEGIN LOGGING AT Sat Apr 25 02:59:57 2009 Apr 25 03:26:09 Is the latest Qi still at http://people.openmoko.org/andy/ ? Apr 25 03:26:18 its dated 3rd Marcxh ... Apr 25 03:26:32 s/Marcxh/March/ Apr 25 03:26:33 BillK meant: its dated 3rd March ... Apr 25 03:31:42 Andy is no longer with Openmoko, so I suspect that's as updated as it gets... Apr 25 03:33:01 hmm, so uboot is deprecated in favour of Qi, Qi looks abandoned ... Apr 25 03:33:34 I though I saw that some Qi work was being done Apr 25 03:34:03 on this channel a couple of days back Apr 25 03:34:56 There have been updates to uboot source code @ openmoko. I think some folks in the community have been playing with Qi. Andy said he was going to use Qi on another project, so somewhere on the internet is the re-birth of Qi (although who knows it if will have support for Om hardware)... Apr 25 03:38:53 k, tkx Apr 25 06:16:47 wakerd: error while loading shared libraries: libecore.so.0: cannot open shared object file: No such file or directory Apr 25 06:16:56 enlightenment: error while loading shared libraries: libecore_con.so.0: cannot open shared object file: No such file or directory Apr 25 06:16:57 run-parts: /etc/X11/Xsession.d/90xXWindowManager exited with code 127 Apr 25 06:17:07 Marvelous. Apr 25 06:17:30 Hopefully someone knows how to fix this. Apr 25 06:17:50 * mwester tosses the gta02 into the trash bin, and goes off to get some sleep. Apr 25 06:49:12 mwester: lib doesnt exist.. maybe? :) Apr 25 06:53:21 mwester: and to me.. that says packaging problems. a binary has been packaged without its attendant shared lib deps Apr 25 06:53:22 :) Apr 25 07:37:40 when changing to another kernel, I just need to Apr 25 07:37:46 flash the new kernel Apr 25 07:37:51 and uncompress Apr 25 07:38:02 the new modules, reboot Apr 25 07:38:15 and run modules-update right? Apr 25 08:13:00 SHR borks completely (somehow) when disabling "GSM antenna" Apr 25 08:17:10 dos1: do you think "settings-phone-gsmantenna:off" operates directly on sysfs without telling framework about modem going down?? ;D Apr 25 08:17:12 DocScrutinizer: PIN? Apr 25 08:17:34 yea it asks for pin again Apr 25 08:18:05 you cant kill the pin enzty window Apr 25 08:18:22 DocScrutinizer: no, it call SetAntennaPoweredf(or smthg like that) ingFSO Apr 25 08:18:27 in FSO* Apr 25 08:18:42 it looks like ophonekitd problem Apr 25 08:18:44 so it's a fso-bug then Apr 25 08:19:25 ophonekitd - for me no difference Apr 25 08:19:43 DocScrutizer: there is difference Apr 25 08:19:53 ophonekitd is SHR stuff Apr 25 08:19:58 not FSO Apr 25 08:20:18 it's a FSO client Apr 25 08:20:22 I know ;) Apr 25 08:20:42 DocScrutinizer: yes, it also hangs when SHR-settings is trying to access FSO stuff Apr 25 08:20:51 when you have a hardware dismounted GSM antenna Apr 25 08:20:54 :P Apr 25 08:21:25 md2k7: hangs? it should display error Apr 25 08:21:50 dos1: well, it didn't react for minutes, the python app just hanged Apr 25 08:21:55 pff, I have a lot of devices here with accitentally missed to reassemble GSM-ant Apr 25 08:22:49 md2k7: still looks like FSO problem Apr 25 08:23:11 yes, I guessed so too Apr 25 08:23:45 i must rewrite GSM module in SHR Settings Apr 25 08:24:04 it's code is really ugly now... Apr 25 08:24:44 dos1: power setting didn't show any sliders do disable autodim Apr 25 08:25:34 DocScrutinizer: ophonekitd crashed? ;p Apr 25 08:25:42 dos1: or is it me needing coffe Apr 25 08:25:42 restart shr settings Apr 25 08:25:44 ? Apr 25 08:26:09 it should allow you to restart ophonekitd via gui Apr 25 08:27:29 bah, just want to switch off modem to watch charging without modem drawing blind sidecurrent Apr 25 08:29:00 you should start ophonekitd Apr 25 08:29:09 I restart shitty ophonekitd, i disable modem again, it crashes once more? won't go that path ;D Apr 25 08:29:28 and then from mdbus call its releaseresource GSM Apr 25 08:29:50 disabling modem in shr-settings isn't done right Apr 25 08:30:10 at least not before breakfast, or even lunch Apr 25 08:31:02 DocScrutinizer: btw, pinging you as you requested ;) Apr 25 08:31:32 * DocScrutinizer away for coffee Apr 25 08:36:13 * roh mumbles something not understandable... anybody a clue which availabe 3d cad can open all our iges or step casefile from freerunner? Apr 25 08:36:46 roh: brl-cad is the only one i think Apr 25 08:37:01 this cad stuff is driving me mad... so much pain... and so many problems and bad workflow... how did people survive that mess at all for longer periods of time? Apr 25 08:38:02 moin roh :) Apr 25 08:39:32 dos1: we should dare to fix USB_curlim setting logic - in uboot/kernel whereever Apr 25 08:40:20 see, there's a lot of devices out there drawing 500mA (or even more) without clean enum on usb Apr 25 08:40:55 dos1: if we attach to a standard usb host, we may draw 100mA Apr 25 08:41:15 ah btw, are there some details on this "battery full while on charger"-discharge problem? Apr 25 08:41:21 I can't seem to find the bug for it Apr 25 08:41:43 dos1: uBoot and kernel manage to negotiate eventually to get 500 Apr 25 08:41:55 md2k7: doesn't hapen here Apr 25 08:42:21 DocScrutinizer: I have some 10-30 mA when battery is full Apr 25 08:42:28 md2k7 do you by any chance run an ancient uboot? Apr 25 08:42:35 I watched it for hours, for days, now. with SHR Apr 25 08:42:46 roh: good idea, I'll update Apr 25 08:42:50 thx Apr 25 08:42:51 i darkly remember a bug which resulted in some register not set correctly in some borderline cases Apr 25 08:42:57 md2k7: so what, I have this as well, and when down to 75% it restarts charging Apr 25 08:43:31 DocScrutinizer: I'm not sure I want a device to drain to 3/4 charge level while on charger Apr 25 08:43:38 hello, i'm trying to learn how to use mdbus. I want to call http://git.freesmartphone.org/?p=specs.git;a=blob_plain;f=html/org.freesmartphone.GSM.SIM.html;hb=HEAD#ListPhonebooks for example. so i run : mdbus -s org.freesmartphone.ogsmd /org/freesmartphone/GSM/SIM org.freesmartphone.GSM.SIM.ListPhonebooks . but i get an org.freedesktop.DBus.Error.UnknownMethod, why please ? Apr 25 08:43:54 roh: can i ask you a question about configuration of the community mailing list? Apr 25 08:44:06 md2k7: you probably want, as herwise your battery wil wear in 6months to death Apr 25 08:44:09 mmmh.. lets see what brl-cad can do on output/input formats Apr 25 08:44:15 PaulFertser ofcourse Apr 25 08:44:40 roh: is there any chance mangling of Reply-To header will be fixed? Apr 25 08:44:45 DocScrutinizer: isn't it worse for a battery to be cycled (if only 1/4) than to be left alone (full)? Apr 25 08:45:04 md2k7: nope, not for LiIon Apr 25 08:45:06 PaulFertser: 'sudo mdbus -s org.freesmartphone.ogsmd' does not list /org/freesmartphone/GSM/SIM Apr 25 08:45:10 argh Apr 25 08:45:15 PierreN: see above ^ Apr 25 08:45:22 DocScrutinizer: interesting Apr 25 08:45:23 PaulFertser i think we have it globally disabled beside the 'list for disble' *ducks* Apr 25 08:45:23 PierreN: mdbus -s org.freesmartphone.ogsmd /org/freesmartphone/GSM/Device org.freesmartphone.GSM.SIM.ListPhonebooks Apr 25 08:45:26 eh disabled Apr 25 08:45:43 md2k7: read the wikipedia on Lithium-ion. It describes the issue in detail Apr 25 08:45:59 community has it off i think because there are too many 'starters of using email' out there. but in general i agree. reply-to shouldnt be mangled Apr 25 08:46:04 roh: ah, so community is list for disabled and therefore "enabled" should ignore it completely or have an individual preference in MUAs? Apr 25 08:46:32 BillK: hey Apr 25 08:46:43 PaulFertser don't get me wrong.. but real development happens not on community. Apr 25 08:46:53 roh: ok then. IMHO mangling does more harm than good even in this case (as users should be educated, not workarounded) but it's just my humble opinion. Apr 25 08:47:02 BillK: you tried to post to shr-devel a question which would have been better in shr-user I think Apr 25 08:47:24 PaulFertser: ok so /org/freesmartphone/GSM/Device never change, assuming that i want to use gsm functions ? Apr 25 08:47:31 PierreN: yes Apr 25 08:47:39 md2k7: see (bat full & restart) see http://www.telit.com/module/infopool/download.php?id=623, p.29, 5.2.1.4, §"The last phase is not properly a charging phase..." & §"This last phase is not implemented in the" Apr 25 08:47:42 PierreN: unless API is changed by FSO devs, of course :) Apr 25 08:47:52 PaulFertser i fully agree, but trust me.. the reason to leave it on was because the current state 'minimized' the complaints in total. if i open that box of pandoras again it doesnt help you or me and just generates a lot of noise Apr 25 08:48:21 ok ^^ Apr 25 08:48:32 we can just silently change it if we ever move the lists somewhere else (who knows) Apr 25 08:48:36 ;) Apr 25 08:48:57 roh: that gives me some hope! :) Thanks :) Apr 25 08:49:13 DocScrutinizer: I get PHP errors in the beginning, forcing firefox to view the binary garbage :-(# Apr 25 08:49:30 * DocScrutinizer is glad to see PaulFertser satisfied wrt RFC conformance ;-) Apr 25 08:49:57 md2k7: known issue. Sometimes it works Apr 25 08:51:03 This last phase is not implemented in the UC864-E/G/WD/WDU internal charging Apr 25 08:51:05 algorithm so once-charged battery is left discharging down to a certain threshold. It is Apr 25 08:51:06 cycled from full charge to slight discharge even if the battery charger is inserted. This Apr 25 08:51:08 guarantees that the remaining charge in the battery is a good percentage and that Apr 25 08:51:09 the battery is not damaged by keeping it always fully charged (Li-Ion rechargeable Apr 25 08:51:11 batteries usually deteriorate when kept fully charged). Apr 25 08:52:01 md2k7: see last two lines Apr 25 08:52:41 DocScrutinizer: got the PDF thx Apr 25 08:53:24 md2k7: Nokia handles LiIon this way, keeping my battery good like a new one *0.8 after 7(!)years of coninued usage Apr 25 08:53:39 Telit handles bat this way Apr 25 08:53:48 we should do the same Apr 25 08:53:55 ok ok I'm convinced :-) Apr 25 08:54:15 especially since PCF50633 has an automatism for this as well Apr 25 08:55:12 only problem with pcf50633: some geeks switched it off in former days, in uboot or kernel. Today it seems fixed Apr 25 08:55:22 as Roh mentioned above Apr 25 08:55:25 somehow funny that aÃall freely available cad solutions are freakware and often not available as packet Apr 25 08:55:36 DocScrutinizer: is it possible my Acer laptop doesn't do that because my LiIon battery is about 1600 probably when it should be 4800 (after two years of mostly wall-plugged usage)? Apr 25 08:55:50 yup Apr 25 08:56:00 nown problem with laptops Apr 25 08:56:31 * PaulFertser wonders why so many things in life suck for no apparent reason... Apr 25 08:56:40 PaulFertser: to make things much much worse, bat is kept at >90% charge @ ~40°..60°C Apr 25 08:57:38 so you guys would suggest plugging in and out the laptop?^^ Apr 25 08:57:48 you can store a bat at fridge with 75% charge (not more!!) in a airtight bag, at ~2°..10° for years and it won't age Apr 25 08:58:23 wow, mine does. never noticed that. just unplugged and have 74 % Apr 25 08:58:32 my bats hang under the roof all day though ;-) Apr 25 08:58:54 some laptops (thinkpads) have firmware that allows you to set the target charge level, so you can keep it at 50% or whatever, even when plugged in Apr 25 08:58:57 the whole bat army Apr 25 08:59:33 btw: is high cpu usage of gstreamer a known issue? Apr 25 08:59:36 budfive: yup, I mentioned this as "prolonged bat live option" some months ago Apr 25 08:59:45 or is it my (self-coded) player app... Apr 25 08:59:59 md2k7: ulseaudio??? Apr 25 09:00:08 PulseAudio? Apr 25 09:00:22 how to determine where it is connected to?^^ Apr 25 09:00:33 I didn't give it much thought, just used playbin Apr 25 09:00:42 gstreamer isn't known for this, pulse is actually Apr 25 09:00:50 nope, no pulseaudio running here Apr 25 09:00:51 pulseaudio is deprecated Apr 25 09:01:07 Ainulindale: ok, I'll resend there. Apr 25 09:01:15 DocScrutinizer: was pulseaudio used on the freerunner at some point? That seems like a bad idea Apr 25 09:01:21 yup Apr 25 09:01:31 DocScrutinizer: for what? Apr 25 09:01:32 former days there was pulseaudio Apr 25 09:01:43 had to go for above mentioned reasons Apr 25 09:01:53 DocScrutinizer: i think sound states stuff should finally be RFCd. I think i have something to add to my text to allow for cooperative force "scenario" unloading scheme. Apr 25 09:02:18 budfive: for what? dunno. Poor ALSA knowledge maybe Apr 25 09:03:04 budfive: imho PA looks good in theory. But some arm (esp VFP-less) implementation bugs prevented it from being used efficiently. Apr 25 09:03:08 my app uses about 55 % CPU when playing :-( Apr 25 09:03:10 i'm planning to do double-blind tests between nokia and freerunner wrt. audio quality Apr 25 09:03:25 results in some serious lagging when you type on illume keyboard, switch apps, etc Apr 25 09:03:35 md2k7: what are you playing? Apr 25 09:03:42 PaulFertser: it has to be much more than just state-RFC. It starts with "use ALSA, don't use pulse" and doesn't end at "we can't use alsactl and statefiles for reason x, y, and z" Apr 25 09:03:43 128 kbps mp3, mostly Apr 25 09:03:50 md2k7: playing mp3? Some resampling maybe or you use a codec that wants floating point? Apr 25 09:04:16 PaulFertser: *sigh* I didn't want to give that so much thought :-( I use playbin of gstreamer Apr 25 09:04:24 DocScrutinizer: i'm not talking about PA at all, nobody cares about it for now :) Apr 25 09:04:48 md2k7: play wavs then :) Apr 25 09:05:08 which is supposedly a sorrow-less "play everything" gstreamer object Apr 25 09:05:12 PaulFertser: yup. still we need to explain how to use audio, not how to use statefiles Apr 25 09:05:13 md2k7: can you try with something like mplayer and see what the usage is then? I recall that the codec choice made a huge difference, although i still couldn't get it below 40% or so playing "normal" mp3s Apr 25 09:05:13 md2k7: at least with tremor ogg vorbis plays with < 30% cpu usage Apr 25 09:05:54 DocScrutinizer: in fact i think i know how to do everything in a way to allow e.g. mplayer to be wrapped in a bash script and to be cooperatively put on pause during incoming call. Apr 25 09:05:54 in fact state files are deprecated. Just they don't know yet ;D Apr 25 09:06:05 I can't understand even the 30 % cpu though, microcontrollers are supposed to be able to sw-decode mp3 with ~ 40 MHz Apr 25 09:06:14 --> shouldn't cause more than ~ 10% CPU load Apr 25 09:06:25 budfive: try mpg321 which uses libmad (integer implementation). Apr 25 09:06:54 PaulFertser: sounds interesting. Tell me more about it, maybe CC me on some PoC Apr 25 09:07:09 or toss me a pastebin link Apr 25 09:07:23 * md2k7 's mplayer uses 30 % CPU Apr 25 09:07:24 PaulFertser: yeah. "mplayer -ac mad ..." made things faster. Will play with mpg321 at some point later, but it shoul dbe the same as mplayer Apr 25 09:07:38 md2k7: it depends on the code Apr 25 09:07:40 and the cpu Apr 25 09:07:50 if its using floating point to decode then its using emulated fp Apr 25 09:07:54 DocScrutinizer: no PoC yet :( just various thoughts in my head... I guess i should finally write some code to be able to test all that ideas... Apr 25 09:08:02 and that is many times slower than an integer decoder Apr 25 09:08:05 yeah, I get the idea, have to dig into codecs and stuff Apr 25 09:08:08 DocScrutinizer: it's a pity nobody seem to care about it (hopefully yet). Apr 25 09:08:18 also remember microcontrollers will use an instructionset designed for signal decode Apr 25 09:08:23 understandable as the fr doesn't have an FPU Apr 25 09:08:30 BillK: plus you didn't subscribe to it anyway :-) Apr 25 09:08:46 PaulFertser: first goal is to stick close to alsa stndard means of doing things as far as possible Apr 25 09:08:49 eg a dsp like the dsp you get on omap2/3 emb4edded soc's is able to decode audo.video at muc lower clockrates and power consumption than the arm core Apr 25 09:09:04 as its specifically intended for processing streams of digital data and putting it through transforms Apr 25 09:09:15 so you can't compare arm and dsp Apr 25 09:09:20 chips designed for different things Apr 25 09:09:21 DocScrutinizer: and it's not interesting, unfortunately. I can't see how to make various sound apps co-exist nicely without dirty tricks. Apr 25 09:09:26 mind u newer arms get better and better Apr 25 09:09:27 :( I thought arm cores are for embedded devices for which playing audio is not an uncommon task Apr 25 09:09:35 but the biggest problem is really floating point Apr 25 09:09:50 md2k7: most audio is done in dsp's Apr 25 09:09:52 or dedicated chips Apr 25 09:10:00 and actually mp3 is not common in "embedded" Apr 25 09:10:07 PaulFertser: yup, there definitely has to be some "dirty hacks" eventually Apr 25 09:10:09 its common in mp3 players Apr 25 09:10:14 and newer mobile phones Apr 25 09:10:26 newer? has been there for years Apr 25 09:10:27 also.. even if you use 40% cpu to play mp3 Apr 25 09:10:28 thats fine Apr 25 09:10:32 as long as it CAN play Apr 25 09:10:34 nope, that's not Apr 25 09:10:45 if I open illume stuff, switch apps etc., it lags Apr 25 09:10:49 if you are stuck to using the arm core to do it - u wont get many lowe power states be it 10% 20% or 40% cpu Apr 25 09:11:01 thats not exactly the "smooth iphone feeling" coming up :D Apr 25 09:11:01 PaulFertser: anyway these dirty hacks need to be such that *not* doing them just has minimal impact on the rest of functionality Apr 25 09:11:08 no it isnt Apr 25 09:11:21 first - for embedded its designed fro 1 task at a time Apr 25 09:11:23 md2k7: just give it -20 priority Apr 25 09:11:27 if it can play mp3 with 40% cpu that is fine Apr 25 09:11:35 lindi-: then my keyboard is unusable :D Apr 25 09:11:36 as u dont do multiepl tasks generally Apr 25 09:11:49 also iphone is 2 generations of arm cpu later Apr 25 09:11:52 md2k7: really? here i can use keyboard and play audio at the same time without lagging Apr 25 09:11:57 and has a gfx unit several tiems faster than the gta02 Apr 25 09:12:07 *sigh* Apr 25 09:12:07 also the iphone has to draw 1/2 the pixels gta02 has to Apr 25 09:12:21 * md2k7 was calculating MHz against MHz and was wondering why it was so much faster Apr 25 09:12:24 i thight you will find that the arm cpu on an iphone is about 2-3x the processing power Apr 25 09:12:27 it ALSO has an FPU Apr 25 09:12:49 so any flaoting point math will be better Apr 25 09:12:49 it has to draw HALF the ppixels the gta02 does Apr 25 09:13:00 yeah, I get the point Apr 25 09:13:06 so for gfx at least u have now about 5-6x the apparent performance Apr 25 09:13:10 if you ONLY did it with the cpu Apr 25 09:13:24 not to mention anything floating-point would be about 10-20x the speed Apr 25 09:13:28 as you now have hw fp Apr 25 09:13:45 now add to that that it has a gfx chip that is ble to offloat most of the gfx work Apr 25 09:13:47 offload Apr 25 09:13:53 (imagination mbx) Apr 25 09:14:07 it is definitely more powerful than glamo and much more capable Apr 25 09:14:08 md2k7: (mhz/mhz) a usual miscnception promoted by industry Apr 25 09:14:21 the 3d unit on the glamo wouldnt be sane to do the advanced gfx you see on an iphone Apr 25 09:14:28 you cant compare mehz Apr 25 09:14:36 until desktop the embedded worls inst in a race to more mhz Apr 25 09:14:41 because if you double the mhzx Apr 25 09:14:48 s/industry/intel/ Apr 25 09:14:49 DocScrutinizer meant: md2k7: (mhz/mhz) a usual miscnception promoted by intel Apr 25 09:14:51 I wonder why the glamo is used at all, as it's only slowing down things (am I missing sth here?) Apr 25 09:14:52 you increase power consumption by much more than x2 Apr 25 09:14:54 like 4x Apr 25 09:14:58 its exxponential Apr 25 09:15:10 so in embedded evertyone is trying to squeeze more in the the same mhz Apr 25 09:15:14 wider mem busses Apr 25 09:15:18 lower cycle latencies Apr 25 09:15:25 or speard it out between multiple cores Apr 25 09:15:54 omfg http://brlcad.org/w/images/6/66/Converting_Geometry.pdf has details about cad ive never wanted to know Apr 25 09:15:57 md2k7: (glamo) just because we can't simple reverse our error to implement it at first instance Apr 25 09:16:04 i have heard from one chip maker than they can increase the numebr of cores for NO extra power usage Apr 25 09:16:16 but increasing clockrate exponentially eats through power Apr 25 09:16:21 so everyone is into the idea of adding cores Apr 25 09:16:24 2 4 or more Apr 25 09:16:29 and often they are specialised cores Apr 25 09:16:31 ~glamo Apr 25 09:16:31 it has been said that glamo is http://www.google.com/search?q=site%3Alists.openmoko.org+glamo+raster Apr 25 09:16:36 the 2442 in gta02 is single core Apr 25 09:16:39 nothign else Apr 25 09:17:08 omap3 for example have an arm core, neon simd core, sgx 3d core, dsp core, and i think at least another imaging processor core Apr 25 09:17:16 raster: yu still don't have a macro or script to spit out this stuff on hitting F11 ?? ;-D Apr 25 09:17:26 DocScrutinizer: nah Apr 25 09:17:30 i have gotten a bit bored of it Apr 25 09:17:36 eventually i'll lose my patience Apr 25 09:17:37 :) Apr 25 09:20:09 rastr: but why glamo is bad? we only need an accelerated drivers!!!1 Apr 25 09:20:18 (joking ofc ;)) Apr 25 09:20:21 you already do Apr 25 09:20:25 :-P Apr 25 09:20:27 dos1: SHUT UP! troll! Apr 25 09:20:35 ;D Apr 25 09:20:46 * md2k7 is hiding in a corner having started a war :P Apr 25 09:21:15 md2k7: also note the 2442 in gta02 runs at 400mhz with 100mhz memory bus Apr 25 09:21:20 it CAN do 500mhz Apr 25 09:21:32 but that was not enabled because it wasnt stable Apr 25 09:21:42 but punishment on membus clock going down Apr 25 09:22:02 i believe that werner and andy thnig it was actually the extra power drain/load on the bus that caused instability Apr 25 09:22:04 ie glamo Apr 25 09:22:08 i coudn't even boot after clocking it @ 500MHz in u-boot Apr 25 09:22:08 you get no 100MHz clock divider for 500MHz cpuclk Apr 25 09:22:11 if glamo was not attached it could do 500mhz Apr 25 09:22:19 mmmh.. somebody got a machine with mips and sgi irix and a pro-e seat for it? *ducks* Apr 25 09:22:34 DocScrutinizer: sure. Apr 25 09:23:26 raster could be.. but who knows what funny rf-adventures one could find inside the 0.8mm of pcb Apr 25 09:23:55 not worth the gain (also eats significantly more power if i remember correctly) Apr 25 09:24:39 so it's like going 400->500 for cpuclk means you either have 125MHz for membus which isn't stable, or you need to go for sth like 66MHz for membus and system thruput goes *down* overall Apr 25 09:24:55 roh: sure. but if the goal is speed... Apr 25 09:25:37 DocScrutinizer: my understanding is that werner and/or andy thought it was the glamo load on the membus that caused it not to be stable @ 125mhz (@500mhz) Apr 25 09:25:48 yu Apr 25 09:25:51 p Apr 25 09:25:51 if that load was taken off - it'd work Apr 25 09:26:12 that is my understanding - i am no hw expert, but it makes sense as the glamo is tagged onto the 244'2 mem bus Apr 25 09:26:17 sure, might be the case. though nobody can tell for sure ;) Apr 25 09:26:21 raster also the ram is intended for 100 isnt it? Apr 25 09:26:33 roh: that - i don't know Apr 25 09:26:36 ack, afaik Apr 25 09:26:45 we will never know unless we had a glamoless gta02 Apr 25 09:26:49 if we did - we'd see Apr 25 09:26:56 have fun Apr 25 09:26:58 i know i ran an early gta02 at 500mhz Apr 25 09:27:00 but it wasnt stable Apr 25 09:27:14 * dos11 is migrating from neo to pc Apr 25 09:27:24 but samsung did intend the 2442 to be able to do 500mhz + 125mhz mem bus Apr 25 09:27:29 overclockers' paradise ;D Apr 25 09:27:51 technically if we assume for now that they were right that glamo load could be the tipping factor in instability @500/125 Apr 25 09:27:55 raster samsung intended it for 64mb ram Apr 25 09:28:08 raster there is also external ram AND nor flash besides glamo Apr 25 09:28:19 the msp only brings the nand and half the flash Apr 25 09:28:19 sure Apr 25 09:28:21 eh ram Apr 25 09:28:30 tho flash isnt attached to the mem bus - right? Apr 25 09:28:31 and rated @10MHz afaik Apr 25 09:28:36 ie its not mmaped Apr 25 09:28:44 100 even Apr 25 09:29:01 raster dunno exactly.. would need to check.. afaik its memory mapped Apr 25 09:29:10 i dont think it is Apr 25 09:29:11 the ram is nevertheless Apr 25 09:29:23 ram is definitely Apr 25 09:29:24 :) Apr 25 09:29:28 eh Apr 25 09:29:37 and it's 100MHz ram Apr 25 09:29:42 why that discussion didn't happened before inventing gta02? :( Apr 25 09:30:04 i wasnt @ om Apr 25 09:30:12 when i came.. galmo was already firmly there Apr 25 09:30:12 dos1: don't ask questions you don't like the answer ;-) Apr 25 09:30:20 if i had been there before Apr 25 09:30:32 raster: i know Apr 25 09:30:46 i would have said "glamo may buy extar sdio 0- but from the specs i dont see it buying gfx performance, and it will create a mountain of work" Apr 25 09:31:01 and i will always avoid the "zero gain, but mountain of work" method Apr 25 09:31:14 i'd have pushed to get 500/125mhz tested (and working) Apr 25 09:31:21 sur e- while phone is on it east power Apr 25 09:31:23 raster: how do I disable illume keyboard animation? Apr 25 09:31:27 but a phone spends 99% of its time suspended Apr 25 09:31:27 raster the nor is memory mapped.. the nand is msp-internal but in the end also memory mapped jedec shit Apr 25 09:31:33 wrench -> animation settings -> keyboard: off doesn't seem to work Apr 25 09:31:34 md2k7: in wrench. "Animations" menu :P Apr 25 09:31:36 nah , I prefer he "zero work, but sme gain for free" method :D Apr 25 09:31:39 not really mapped.. but paged via controller Apr 25 09:31:42 md2k7: there is an animatyion config dialog in display in e's config panel Apr 25 09:31:46 md3k7: it should be disabled by default in SHR now Apr 25 09:32:46 it works for me here (xephyr, emulated on desktop) Apr 25 09:33:04 DocScrutinizer: i prefer zero wrk, but gbig gain too! Apr 25 09:33:05 for me it works too Apr 25 09:33:06 but big Apr 25 09:33:07 :) Apr 25 09:33:41 so raster: why didn't you come to Om earlier? :D Apr 25 09:33:50 but normally its a tradeoff of X work and X+N gain Apr 25 09:33:54 (is "animation off" meaning the keys don't "jump" any?) Apr 25 09:34:00 or just that they are not animated jumping Apr 25 09:34:02 md2k7: no Apr 25 09:34:16 (jump = get larger) Apr 25 09:34:18 dos1: dunno. Apr 25 09:34:29 md2k7: animation off means that popping up and down whole keyboard isn't animated Apr 25 09:34:52 I would want to disable popping of single keys Apr 25 09:34:56 does that work somehow? Apr 25 09:34:59 dos1: because OM mgmt never realized where to throw manpower at, and where to be reluctant Apr 25 09:35:02 dos1: but i do rememebr making an early appraisal of glamo and it was "we'll do a lot of work and really get no substantial gain wirth mentioning, some things will be faster, some slower, int he end you'll even out" Apr 25 09:35:18 and after my initial excitement at reading the short-list specs "3d acceleration" and al the rest Apr 25 09:35:24 my excitement dropped to 0 Apr 25 09:35:45 ie the chip was a shin-hole for work and pain, and will never produce any real benefit./ Apr 25 09:36:09 i because particularly negative about glamo when i heard no successor was being made Apr 25 09:36:23 with no successor any driver work would be throw-away ffor gta02 only - ever Apr 25 09:36:42 and it would never see the light of day on another om device Apr 25 09:36:55 and no one else i knew of was using glamo +x + linux Apr 25 09:37:00 thus even less useful Apr 25 09:37:01 on *any* device Apr 25 09:37:09 DocScrutinizer: well 1 other device used it Apr 25 09:37:17 some korean electronic dictionary Apr 25 09:37:17 duh? Apr 25 09:37:25 hehe Apr 25 09:37:28 ;D Apr 25 09:37:30 but it didnt use x or linux Apr 25 09:37:35 thus driver work was useless Apr 25 09:37:36 lol Apr 25 09:37:42 if glamo2 was being planned Apr 25 09:37:55 we might have hope of input on it - its featureset and a faster chip in future Apr 25 09:38:13 thus i'd consider work for it while useless now, woudl be a baseline for building on for glamo 2 Apr 25 09:38:15 and a benefit then Apr 25 09:38:24 but that wasnt happening either Apr 25 09:38:50 so i woudl rather just see glamo sink to the bottom of the pacific and never see the light of day again to save everyone the pain :) Apr 25 09:38:51 seems I've read this a few days ago Apr 25 09:38:55 ;) Apr 25 09:39:17 [gta03] maybe? Apr 25 09:39:20 we'd all better off with a gta02 that had a simple dumb fb like gta01 Apr 25 09:39:26 it'd be reliable and solid Apr 25 09:39:29 wel lmore solid Apr 25 09:39:31 raster: so is there a way to disable single keys on illume keyboard jumping when typed? Apr 25 09:39:44 it'd perform faster imho for most things Apr 25 09:39:51 raster: start creating a mocro finally! XD Apr 25 09:39:57 macro Apr 25 09:40:06 md2k7: fiddle with the theme (the anim is done in the theme - for all keys) Apr 25 09:40:30 ok.. brlcad is hard.. i mean.. come on.. gazillion of shell utils? thats quite 70th unix-ie Apr 25 09:41:32 roh: ooh i rememebr that Apr 25 09:41:35 shell utls for cad seems weird Apr 25 09:41:37 man did that put me off Apr 25 09:41:54 "so i am meant to design using cmd-line utils? hooray!" Apr 25 09:43:17 hehe, edit the project files with sed XD Apr 25 09:44:02 * DocScrutinizer thinks it's time for breakfast Apr 25 09:44:25 raster well.. actually its quite interresting Apr 25 09:44:42 read http://brlcad.org/w/images/6/66/Converting_Geometry.pdf Apr 25 09:44:57 errr, eeew Apr 25 09:45:06 raster: you mean the illume-shr.edc file inside the .edj theme? Apr 25 09:45:24 I don't really see what should do the animation here Apr 25 09:45:25 raster that stuff was written for the us-military and can be used for lots more than normal cas does.. and it explains what fundamental concepts exist in cad and where they come from Apr 25 09:46:05 great for cad-developers Apr 25 09:46:29 or for nuke-CAE Apr 25 09:46:47 ack. Apr 25 09:46:51 probably NASA also would like it Apr 25 09:47:01 md2k7: dont knwo what shr does.. but illume.edc is in e's src Apr 25 09:47:07 for people who need to calculate when it just blows up before it should Apr 25 09:47:23 DocScrutinizer they had their own stuff.. but also use it it seems Apr 25 09:47:32 raster, md2k7: illume-shr.edc is edited copy of illume.edc Apr 25 09:47:59 and modified are only "loading" layer, battery indicator and clock AFAIR Apr 25 09:49:35 bat-indicator is shitty. virtually can't click it Apr 25 09:49:53 (I'm under the impression that DocScrutinizer keeps whining :-p) Apr 25 09:49:57 active area intersects with right-arrow Apr 25 09:50:26 * md2k7 has clock, battery, gsm and right button, so he can click battery Apr 25 09:50:41 but battery overlaps clock, a bit Apr 25 09:50:51 bah, my caffeine level isn't ballanced to my adrenaline level. usual state when breakfasting Apr 25 09:51:06 :-) Apr 25 09:51:09 DocScrutinizer: how's life? Apr 25 09:51:22 >:-/ Apr 25 09:51:33 == as usual ;) Apr 25 09:51:37 From your smiley I'm guessing you're a punk? Apr 25 09:51:50 always been Apr 25 09:52:04 DocScrutinizer: it works for me Apr 25 09:52:35 Ok, installing hardware sensors Apr 25 09:52:39 I'm trying to make this freaking desktop silent Apr 25 09:52:47 And I have to say that Noctua CPU fans are great Apr 25 09:53:17 applying efl1 patches Apr 25 09:53:17 error: patch failed: recipes/efl1/evas_svn.bb:1 Apr 25 09:53:17 error: recipes/efl1/evas_svn.bb: patch does not apply Apr 25 09:53:19 pff, when it takes 14h to swap that fan... Apr 25 09:53:29 DocScrutinizer: it was another fan :-) Apr 25 09:53:40 (my server fan) Apr 25 09:54:14 * md2k7 is enjoying his 100 uf audio capacitors Apr 25 09:56:27 md2k7: are they tasty? Apr 25 09:56:53 :D Apr 25 09:56:58 * md2k7 doesn't eat tantalum caps Apr 25 09:57:00 * DocScrutinizer throws a handful of 2440 chip at md2k7 Apr 25 09:57:33 need some guacamole? Apr 25 09:57:45 Meh :-) Apr 25 09:58:33 SHR: 03seba.dos1 07shr-settings * r3d96733131c1 10/TODO: [TODO] update Apr 25 10:00:31 raster: ping Apr 25 10:02:11 ja Apr 25 10:02:48 raster: why is in elementary slider "delay,changed" callback for? Apr 25 10:03:51 raster: thanks for those snapshots :) Apr 25 10:04:25 snapshots? pr0n? ;-) Apr 25 10:04:35 efl/e17 Apr 25 10:05:07 raster: where are they? could they be built on the desktop and tested with Xephyr? Apr 25 10:08:43 * DocScrutinizer [rolldown, clrX, rolldown, rolldown] Apr 25 10:09:28 dos1: try and found out :) Apr 25 10:09:42 dos1: basically smeone (probably UBoot *and* kernel) should assume there's a 500mA-capable charger id USB-enum fails Apr 25 10:09:44 i tried, and now commiting :D Apr 25 10:10:07 DocScrutinizer: (delayed means.. its delayed from the actual slide - its so you dont react to every single move - if the user drags it around and keeps dragging u dont react - the signal doesnt get called until the users stops dragging it Apr 25 10:10:08 s/ id/ if/ Apr 25 10:10:10 DocScrutinizer meant: dos1: basically smeone (probably UBoot *and* kernel) should assume there's a 500mA-capable charger if USB-enum fails Apr 25 10:10:10 (lifts fingeR) Apr 25 10:10:21 or stops actually moving the thing for a bit (while still pressed) Apr 25 10:10:46 so you just react when they stop - elementary is handling "wait until the user has stopped changing his mind" code for you Apr 25 10:10:52 tilman: np Apr 25 10:11:00 raster: why you're teling me? Apr 25 10:11:01 tmzt: the snaps? Apr 25 10:11:09 DocScrutinizer: oops. tab oops Apr 25 10:11:13 dos1: for you above Apr 25 10:11:14 :) Apr 25 10:11:33 SHR: 03seba.dos1 07shr-settings * r4192063b1e17 10/shr_settings_modules/shr_display.py: [Display] change callback from changed to delay,changed Apr 25 10:11:40 raster: ;) Apr 25 10:11:42 yes Apr 25 10:12:19 dos1: next we could use something like the "slow charge daemon" or similar thing to enable 1A charging Apr 25 10:12:31 hoping for debian or gentoo pkgs but I can build from source Apr 25 10:12:40 dos1: 0.5->1A that is Apr 25 10:12:42 tmzt: of course u can. Apr 25 10:12:43 gentoo on arm I mean Apr 25 10:12:52 hello Apr 25 10:12:56 (built on desktop etc.) Apr 25 10:13:02 I use ubuntu on desktop Apr 25 10:13:07 url? Apr 25 10:13:18 www.enlightenment.org Apr 25 10:13:22 look @ news Apr 25 10:13:31 ok Apr 25 10:13:37 does anyone knows how to use adb to send keys signals to android (to use pc keyboard instead of the freerunner keyboard) ? Is there something like /dev/kdb ? thank you Apr 25 10:13:56 kaoul would like to know how to enable usb host Apr 25 10:14:23 maybe, what can you do with "usb host" ? Apr 25 10:14:39 I don't really see what the point Apr 25 10:14:42 use a keyboard with powered hub Apr 25 10:14:43 *what's Apr 25 10:14:47 kaoul: use FR like any other USB-host Apr 25 10:15:02 kaoul: yep, plug keyboard into neo Apr 25 10:15:10 kaoul: attach USB-devices, like kbd Apr 25 10:15:27 Yes, and because I don't have any usb keyboard (mobile computer), I'm not interested in usb host ;-) Apr 25 10:15:39 tmzt: why *powerd* hub? why any hub at all? Apr 25 10:15:39 oh sorry Apr 25 10:15:59 it doesn't have to be powered? Apr 25 10:16:18 tmzt: FR powers usb port when you are in host mode Apr 25 10:16:23 FR has 500mA power supply at usb Apr 25 10:16:34 most people don't have the injectors, but you still need miniA to B cable Apr 25 10:16:39 oh Apr 25 10:16:41 cool Apr 25 10:16:47 Do you guys means that I could make my mobile computer act like a usb keyboard linked to the freerunner ? Apr 25 10:17:02 kbd is only 100 or so mA Apr 25 10:17:10 no Apr 25 10:17:16 kaoul: seems nobody understood exactly what's your question and usecase Apr 25 10:17:23 please elaboprate Apr 25 10:17:41 adb ?? Aple Desktop Bus? Apr 25 10:17:49 DocScrutinizer i think andoid hacked up some proprietary usb debuggingfoo and even messed up the naming Apr 25 10:17:53 Well, I use android debug bridge Apr 25 10:18:06 a d b Apr 25 10:18:09 ah.. thats what i means.. Apr 25 10:18:17 no android here Apr 25 10:18:26 he wants to use ui apps in android with keys from another computer, when he said pc keyboard I missread as usb Apr 25 10:18:36 never heard, never going to use. Sorry Apr 25 10:18:41 kaoul try #freerunner-android Apr 25 10:18:42 it's just a shell connection Apr 25 10:19:10 lol I was on #android and tmzt advized me to ask here ;) Apr 25 10:19:13 if google would have been smart they would have used cdc-acm like the rest if the world Apr 25 10:19:24 if that exists that's better, sorry I had the wrong channel Apr 25 10:19:32 n.p. Apr 25 10:19:43 no problem, I try #freerunner-android ! Apr 25 10:19:51 thank you guys Apr 25 10:20:18 ok.. freecad can open the moko cad in step format Apr 25 10:20:29 but it really kills my machine with load Apr 25 10:25:31 ;-D Apr 25 10:25:48 still loading? ;-> Apr 25 10:27:30 nop.. but rendering takes ages and has bugs Apr 25 10:28:24 same as it ever was ;) Apr 25 10:28:52 SHR: 03seba.dos1 07shr-settings * r26be75564803 10/data/po/it/ (shr-settings.mo shr-settings.po): [i18n] update italian translation Apr 25 10:31:34 SHR: 03seba.dos1 07sh * rec0997443978 10/libframeworkd-phonegui-efl/ (configure.ac po/it.po): libframeworkd-phonegui-efl: add italian translation by Marco Franceschetti (von_fritz) Apr 25 10:34:28 dos1: wurp2|gone : basically the risks of setting a higher USB_CURLIM (>100, >500) are largely overrated Apr 25 10:35:45 for sure OM can't recommend it as they would take all sorts of unevaluable risks and most likely get sued evetually Apr 25 10:35:56 raster: why does it say eina is missing, it's the ecore pkg Apr 25 10:36:27 not a upstream problem Apr 25 10:36:31 packaging problem if anything Apr 25 10:37:25 dos1: wurp2|gone: but for all practical cases forcing USB-CURLIM to 500 after USB enum failed would be a safe thing afaikt Apr 25 10:37:41 s/ikt/ict/ Apr 25 10:37:41 DocScrutinizer meant: dos1: wurp2|gone: but for all practical cases forcing USB-CURLIM to 500 after USB enum failed would be a safe thing afaict Apr 25 10:38:10 I see, well I'm trying to install to opt but I thought there was on tgz to download now I see there are many Apr 25 10:38:58 likely u havent fixed your build env vars Apr 25 10:39:02 PKG_CONFIG_PATH Apr 25 10:39:09 later u'll need PATH too so it can edec edje_cc Apr 25 10:39:12 etc. Apr 25 10:42:08 dos1: or let me put it this way: 100 is ok during USB enum, and if result of enum is it's a host and it can't go beyond 100. all other cases I'd push up curlim to 500 and check if it still delivers decent voltage a Vbus-USB. *IF* NOT, switch OFF USB (curlim=0) for 10 sec, then resume with 100 Apr 25 11:15:56 I want to turn off the screen (via sysfs spi disable), but X is turning it on again on tap Apr 25 11:16:03 how can I disable this? Apr 25 11:16:30 or maybe disable touchscreen (?) Apr 25 11:25:14 what builds illume_start or is that not needed anymore? Apr 25 11:31:56 md2k7: chvt 1? Apr 25 11:32:38 lindi-: thx Apr 25 11:35:10 lindi-: nope, X still comes back :-( Apr 25 11:35:28 doesn't switch, whyever Apr 25 11:35:42 md2k7: why does it not switch? Apr 25 11:52:44 md2k7: if you're running frameworkd, it's oeventsd that is turning on screen on touch. Apr 25 11:53:34 I've run mdbus -s org.freesmartphone.ousaged /org/freesmartphone/Usage org.freesmartphone.Usage.SetResourcePolicy Display disabled Apr 25 11:53:54 I've kinda noticed it is turning the backlight on again Apr 25 11:54:04 SHR: 03seba.dos1 07shr-overlay * rce29344cfac1 10/openembedded/recipes/images/shr-image.inc: shr-image.inc: fix generating rootfs. Apr 25 11:54:30 it's not my fault this time :D Apr 25 11:55:12 md2k7: not disabled - use enabled Apr 25 11:55:52 dos1: ??? Apr 25 11:56:01 mdbus -s org.freesmartphone.ousaged /org/freesmartphone/Usage org.freesmartphone.Usage.SetResourcePolicy Display enabled Apr 25 11:56:09 yeah, but why exactly? Apr 25 11:56:19 with that frameworkd shouldn't touch backlight Apr 25 11:56:34 from fso docs: "disabled" - using this resource is currently not allowed, Apr 25 11:56:42 "enabled" - the resource is enabled even without a client. Apr 25 11:56:50 I thought enabled means display = on Apr 25 11:56:57 md2k7: yep, and you want to have it enabled Apr 25 11:57:02 display and CPU are somewhat special resources Apr 25 11:57:09 they do influence the way the idlenotifier works Apr 25 11:57:27 if you _enable_ the display resource, it means the framework will no longer dim it Apr 25 11:57:30 heh Apr 25 11:57:34 ophonekitd crashed Apr 25 11:57:37 "Out of memory" Apr 25 11:57:44 talk about stability :) Apr 25 11:57:46 i think it's time to swap :P Apr 25 11:57:46 if you _enable_ CPU, it means the framework will no longer send out SUSPEND events Apr 25 11:58:15 mickeyl: thanks for the info :-) CPU is somewhat clear, display is inverted to my logic Apr 25 11:58:30 md2k7: any details about your bass caps rework? Apr 25 11:58:31 yeah Apr 25 11:58:37 Display has the same logic as CPU for me :X Apr 25 11:58:45 the problem is "enable" and "disable" does not make sense for these resources Apr 25 11:58:52 rather only refer to "request" and "release" Apr 25 11:58:56 it makes more sense then Apr 25 11:59:10 PaulFertser: I've pretty much done the same thing as you Apr 25 12:00:09 md2k7: nice you enjoy it because i use sandisk sansa e250 as a portable player, so for me the rework was somewhat useless ;) Apr 25 12:00:58 PaulFertser: I had an old mp3 player which could only read from SD cards <= 1 GB Apr 25 12:09:03 PaulFertser: had you made a sound rework for the neo, I mean bass sound? Apr 25 12:09:14 Q-Master: http://wiki.openmoko.org/wiki/GTA02_bass_fix Apr 25 12:09:32 thnx Apr 25 12:09:59 Q-Master: i guess i was first to publish the photos of performed rework :) Apr 25 12:12:09 PaulFertser: Andy's rework is smaller. 8) Apr 25 12:12:30 Q-Master: mine is even larger :P Apr 25 12:12:42 I used 16V tantalum caps because thats what I got (ordered 10V) Apr 25 12:12:54 had to glue them sideways Apr 25 12:13:08 otherwise the width of the 2 caps wouldn't fit Apr 25 12:13:17 Q-Master: yes, but he did it as demo only, i guess he didn't test it with headphones at all :) And it has enough limitations. Apr 25 12:13:31 people using ASU look like they stopped at previous age ;) Apr 25 12:13:42 asking for things, which are there for a while Apr 25 12:13:44 dos1++ :) Apr 25 12:14:36 PaulFertser: I've tried headphones, but the state which should be loaded to phone through them stayed unknown to me Apr 25 12:14:37 8) Apr 25 12:14:49 Q-Master: headset.state Apr 25 12:14:58 Q-Master: the state to do what exactly? Listen to the music? Apr 25 12:14:59 in /usr/share/openmoko/scenarios/ Apr 25 12:15:16 and to using gsm via headphones: gsmheadset.state Apr 25 12:15:48 Q-Master: on FSO-based distro it works OOB, regular stereout.state with one knob tweaked automatically by oeventsd whenever you insert the headphones. Apr 25 12:15:57 dos1: while we're at it: there is no state for talking on GSM via headset, is there? Apr 25 12:16:17 md2k7: there is, but it doesn't matter Apr 25 12:16:24 PaulFertser: why not? Apr 25 12:16:30 md2k7: there is. gsmheadset.state. and it works for me. Apr 25 12:16:32 md2k7: headset mic will buzz the hell out of the other party :) Apr 25 12:16:38 PaulFertser: I have buzz fix :) Apr 25 12:16:46 md2k7: won't help, see schem Apr 25 12:17:26 PaulFertser: :( Apr 25 12:17:54 md2k7: fixing gsm buzz for wired mic is not practically possible on all FR out there (and A8 should come buzz-free from the factory, though i doubt it's out already). Apr 25 12:18:18 PaulFertser: what did they do to it? black magic? PCB redesign? Apr 25 12:18:24 hmm Apr 25 12:18:44 md2k7: (mass-fixing, that is. Probably some dedicated guy can do that) They just place ferrite beads on all paths to prevent RF from entering the can. Apr 25 12:19:03 wow, what an invention Apr 25 12:19:09 i have warranty for my Neo to july 2009 Apr 25 12:19:28 md2k7: heh. I guess it'd be practical to use internal buzz-fixed mic with external headphones though. Or a BT headset. Apr 25 12:20:14 i wonder if that buzz would be a reason to replace my gta02v5 to v8 Apr 25 12:20:52 but... i really love this device :x Apr 25 12:21:01 hmm Apr 25 12:21:34 does someone know, how openmoko "mass-fixing-project" could be done? Apr 25 12:22:10 dos1: well, I read on the ML that openmoko itself isn't very interested to do that Apr 25 12:22:47 i read that they only had to do some logistics until they can start it Apr 25 12:22:54 mhm Apr 25 12:32:02 PaulFertser: SHR also? Apr 25 12:32:13 Q-Master: also what? Apr 25 12:32:32 also switches to headset if i insert it? Apr 25 12:32:33 Q-Master: switch amplifier on/off? Yes, sure. :) Apr 25 12:32:53 the only thing I need now is a PIM Apr 25 12:34:35 BTW, is it possible to change virtual keyboard in Navit to support russian keys? Apr 25 12:35:01 Q-Master: does navit use some other non-illume keyboard? Apr 25 12:37:11 PaulFertser: yep. it has it's own internal one Apr 25 12:37:54 Q-Master: can't it be disabled then? Apr 25 12:38:30 quickdev was working on efl navit backend Apr 25 12:38:34 ~seen quickdev Apr 25 12:38:34 quickdev was last seen on IRC in channel #openmoko-cdevel, 58d 22h 3m 14s ago, saying: 'and for you?'. Apr 25 12:38:40 :/ Apr 25 12:39:35 PaulFertser: don't know. Apr 25 12:42:19 mwester: is andy-tracking branch now in a state where it could be tested? ;) Apr 25 12:43:06 lindi-: we are using andy-tracking in shr Apr 25 12:43:12 so it looks like it is :P Apr 25 12:43:17 ok Apr 25 12:44:53 dos1: Heinervdm's sqlite backend looks like working one. 8) Apr 25 12:45:17 Q-Master: opimd backend? Apr 25 12:45:21 Q-Master: did you tested it? Apr 25 12:45:29 dos1: yep. Apr 25 12:45:35 Heinervdm: nop. 8) Apr 25 12:45:39 :D Apr 25 12:45:48 Heinervdm: hmm... where can i get it? Apr 25 12:45:51 I have android for now. Apr 25 12:45:57 i had no programm to test it Apr 25 12:46:09 i would like to test it Apr 25 12:46:10 :) Apr 25 12:46:18 dos1: i will post it on pastbin Apr 25 12:46:43 Heinervdm: if it works and it's written well, i will commit it to FSO Apr 25 12:46:47 ;) Apr 25 12:46:48 http://pastebin.com/m41c64947 Apr 25 12:46:54 Heinervdm: the only thing it needs is deleting and editing Apr 25 12:47:03 dos1: that was my first python script ;) Apr 25 12:47:07 dos1: it is written rather good Apr 25 12:47:20 opimd lacks whole deleting and editing support at the moment Apr 25 12:47:21 ;) Apr 25 12:47:37 Q-Master: first editing and deleting has to be implemented in opimd Apr 25 12:48:16 anyway one need to implement an import from vcard script. it'll be easy i think Apr 25 12:48:17 Heinervdm: code looks good :) Apr 25 12:48:45 Heinervdm: so if it works, i will commit it Apr 25 12:48:55 and i'll test it now Apr 25 12:48:56 dos1: i used 2 tables one for uniqe elements and one for entrys where multiple values are possible Apr 25 12:48:56 dos1: check it plz Apr 25 12:49:15 Heinervdm: the design is good also Apr 25 12:49:41 it's mostly from the cvs backend Apr 25 12:50:12 Heinervdm: i wanted to do the same, but i'm lazy, so i'm happy that someone did it for me :DDD Apr 25 12:50:38 ok, i will test it after dinner ;) Apr 25 12:50:44 dos1: how do you test it? is there a application using opimd? Apr 25 12:50:58 i have my small test python script Apr 25 12:51:05 ok Apr 25 12:51:09 and there is libframeworkd-phonegui-efl2 Apr 25 12:51:12 using opimd Apr 25 12:51:33 * Q-Master needs to write an vcard import script Apr 25 12:51:59 Q-Master: it shouldn't be hard Apr 25 12:52:06 it's nice to see that there is some opimd work :) Apr 25 12:52:20 yep Apr 25 12:53:02 at the time i ve written this, there was no one to ask about opimd Apr 25 12:53:23 so i've stoped, but now i can do further work Apr 25 12:54:51 Ainulindale: http://pastebin.com/m41c64947, Heinervdm's sqlite contacts backend for opimd :) Apr 25 12:55:10 mrmoku: ^^^ Apr 25 12:55:14 :-) Apr 25 12:55:59 Heinervdm: yeah... further work please :-)( Apr 25 12:56:07 what is contact uri looks like? Apr 25 12:56:21 dbus:// ??? Apr 25 12:56:28 this are questions i had too ;) Apr 25 12:57:03 no... dbus:// has to be dropped Apr 25 12:57:22 How should it look like? Apr 25 12:57:26 it would be nice if in the documentation were examples for every entry Apr 25 12:57:26 never seen it Apr 25 12:58:03 /org/freesmartphone/opimd/Contacts/123 ??? Apr 25 12:58:06 Heinervdm: you should use /home/root/ or ~/ in _SQLITE_FILE_NAME Apr 25 12:58:25 123 is what?// Apr 25 12:58:31 the contact id Apr 25 12:58:37 Q-Master: i think i tried it, but id didn't work with ~/ Apr 25 12:58:40 (which is not permanent as far as I understood) Apr 25 12:59:15 mrmoku: ok Apr 25 13:00:10 if opimd is getting some momentum again now... I will concentrate on libframeworkd-phonegui-efl2 Apr 25 13:00:12 and i don't like a hard coded path to a user dir ;) Apr 25 13:01:32 hmm Apr 25 13:01:56 Heinervdm: looks like your backend is the first, which will have editing and deleting support Apr 25 13:02:08 as it's the easiest to do with sql backend :) Apr 25 13:02:21 8) Apr 25 13:02:46 Matthias Felsche around here? someone knows where to IRC him? Apr 25 13:03:00 Heinervdm: and now it will create a sqlite file in the current pwd Apr 25 13:03:05 Heinervdm: not good Apr 25 13:05:20 Heinervdm: it'll better be hardcoded then random Apr 25 13:12:49 Has anyone fixed the broken e library problem in unstable yet? Apr 25 13:12:58 * mwester didn't see pertinent commits :( Apr 25 13:13:08 mwester: it works for me Apr 25 13:13:16 mwester: commits are in openembedded Apr 25 13:13:35 Q-Master: no, it's not in currect pwd Apr 25 13:13:37 I built it twice late last night Apr 25 13:13:41 from scratch Apr 25 13:13:43 correction: commit _have_ been in OE, but have been reverted Apr 25 13:13:49 commits are still in mickey/org.oe.dev Apr 25 13:13:50 oh Apr 25 13:14:02 dos1: why? Apr 25 13:14:11 Q-Master: it's in /etc/freesmartphone/opim/ Apr 25 13:14:14 mickey_away: I shall not ask why. :( Apr 25 13:14:20 mwester: read my latest blog entry Apr 25 13:14:29 dos1: not good either Apr 25 13:14:33 dos1: that's right it's easy with sqlite Apr 25 13:14:41 mwester: i will take some time to think whether we can afford a fork Apr 25 13:14:53 there are many people fed up with the situation Apr 25 13:14:57 Q-Master: that'S right too, but we can change it :) Apr 25 13:15:23 mickey_away: Just collectively kick the other ones out Apr 25 13:15:31 Heinervdm: let's do it? Apr 25 13:16:08 alphaone: if it was so easy, we had done it years ago Apr 25 13:16:08 mickey_away: I get rather worked up over these issues; I had hoped things like the ease of merge/branch with git would help, and that the stable branch would be a clear solution. But... I guess not. Apr 25 13:16:25 alphaone: hehe! Apr 25 13:16:27 Q-Master: dos1 can change it to /home/root/.opimd or something like that before he commits it Apr 25 13:16:29 mwester: yes, i had hoped that as well Apr 25 13:16:48 Heinervdm, Q-Master: now it should be in the same directory as CVS file Apr 25 13:16:56 mickey_away: i hate seeing projects go down the drain that way :( Apr 25 13:16:57 good Apr 25 13:17:07 mickey_away: So what is the problem there? Apr 25 13:17:08 Wonka: yes, it's very sad Apr 25 13:17:25 there is obviously no majority for draconian consequences Apr 25 13:17:25 * Q-Master is studying python-vcard to write an import/export script Apr 25 13:17:37 and as such i no longer want to be a member of this "core team" Apr 25 13:17:46 *sigh* Apr 25 13:17:56 lindi-: (kernel) yes, shr-unstable _and_ shr-testing are both on andy-tracking; the huge problem right now is the re-appearance of the WSoD (or variation thereof) with u-boot. :( Apr 25 13:18:46 mwester: ah yes i saw some discussion about it Apr 25 13:20:17 Unfortunately, the news of that came out after the switch to andy-tracking, or I think we would have left testing on 2.6.28 (without the WSoD). Apr 25 13:24:45 AT+CSMS=1=>['+CMS ERROR: 314'] Apr 25 13:25:14 pcf50633-rtc pcf50633-rtc: PCF_TIME: 25.04.09 13:23:48 Apr 25 13:25:18 pcf50633-rtc pcf50633-rtc: RTC_TIME: 25.3.109 13:23:48 Apr 25 13:25:23 what's that pcf time? Apr 25 13:30:12 lindi-: I assme PCF is rtc of PCF50633 actual cmos-clock. Don't ask me what's RTC-time though Apr 25 13:32:50 DocScrutinizer: anyways, when i upgraded my kernel the clock went off-by-month Apr 25 13:33:45 lindi-: it's correct Apr 25 13:34:00 PCF vs RTC in those messages are the before and after conversion values. RTC is the format the hardware wants, PCF is the linux format Apr 25 13:34:05 month in RTC is saved from 0 to 11 Apr 25 13:34:15 Those are correct figures. Apr 25 13:34:16 and year is from 1900 Apr 25 13:35:01 WTF?? omg Apr 25 13:36:01 I know the kernel sematics of time-structure type are quite nuts. BUT WHO WANT's TO SEE THIS????? Apr 25 13:36:37 especially in this fucked-up format? Apr 25 13:37:58 SHR: 03mwester 07shr-overlay * rf1a9d0de0e21 10/patches/ (2 files in 2 dirs): shr/patches: update patches for upstream-updates to PR numbers. Apr 25 13:39:22 nobody wants to see that, now that it's been debugged. lower the loglevel. Apr 25 13:39:39 * DocScrutinizer curses Kontact. >:-( SEGFAULT Apr 25 13:40:40 Well, switch to gnome. Apr 25 13:40:51 dos1: but the output of "date" is off-by-month too Apr 25 13:41:15 Any software system that konstanly puts a "k" where a "c" should be is suspekt. :p Apr 25 13:41:21 task-shr-minimal-x 2.0-r4 is broken... depends on glibc-binary-localedata-en-gb, which is not available Apr 25 13:41:24 lindi-: your rtc was off previously. Apr 25 13:41:38 lindi-: probably correct Apr 25 13:41:47 no glibc-binary-localedata-* is available. many glibc-localedata-* are. Apr 25 13:41:49 lindi-: you need to reset your hwclock (date && hwclock --systohc) Apr 25 13:43:39 mwester: yep i am doing that Apr 25 13:43:47 mwester: i was just wondering if that was expected or not Apr 25 13:43:48 actually that should be done automatically whenever there's a trusted "timeserver" (like GPS) offering a timestamp Apr 25 13:44:31 and it MUST NOT be done in init.d stopscripts Apr 25 13:45:27 ultimately, yes. Apr 25 13:45:36 lindi-: btw, same time you should `rm /etc/adjtime` Apr 25 13:46:01 for right now, I'm far less concerned with the accuracy of the clock as I am with the WSoD problem returning, and the apparent total lack of progress on fixing that. Apr 25 13:47:11 hmm, I had my share on trying to fix it for good. No approved kernel-dev here ;-) Apr 25 13:47:33 mwester: i thought Nicolas is working on it. Probably we i can send him a letter asking for some kind of status report :) Apr 25 13:48:18 basically the solution was to power down LCM by LDO-switchoff afair Apr 25 13:48:43 if someone messed it up again, I don't know how to help Apr 25 13:48:59 I just sent an email to the kernel list. Apr 25 13:49:07 (power down) instead of going into LCM-suspend Apr 25 13:49:45 it's all in the relating wsod-tracticket I think. What else can we do? Apr 25 13:50:01 I rather had hoped that someone who's been working with that code could sort this; it would take a lot of time (and be more risky) if someone new were to start hacking on that driver as well. It's just too fragile. Apr 25 13:50:56 alas it seems nobody left over of the ones "who's been working with that code" Apr 25 13:51:36 so exactly that has happened already. It's fragile amnd someone unused to it broke it Apr 25 13:52:16 * DocScrutinizer shrugs Apr 25 13:55:37 mwester: (fedora) as far as i know, fedora project has a clear goal of being a testbed for fancy modern features (including desktop-related), some of which will be later incorporated into RHEL. And all ubuntu does is fucking up Debian (like "fail-safe x-server" ewww) for no apparent gain... Apr 25 13:56:08 And spamming google with useless "solutions" from ubuntuforums :-/ Apr 25 13:56:17 mwester: if I stated "no kernel hw-driver patches commit without me reviewing them" - do you think anybody cared (beyond starting a flamewar on me being arogant)? Apr 25 13:57:13 DocScrutinizer: I think if you offered to review same, Nelson would probably solicit your input. Apr 25 13:58:17 PaulFertser: I thought ubuntu's goal was to make Debian "user-friendly" Apr 25 13:59:05 bunt-kuh, bah Apr 25 13:59:24 BTW, I was practicing my soldering skills and managed to repair an old motherboard -- I just installed Debian on it. I can't say that I see what all the fuss is about. Apr 25 13:59:25 mwester: to me it looks like that's just another wording of what i wrote. Because they seem to understand user-friendliness in some very perverted way... Apr 25 13:59:30 ok Apr 25 13:59:41 i fighted with opimd and managed it to use sqlite backend Apr 25 13:59:43 NameError: global name 'add_contact_to_db' is not defined Apr 25 13:59:55 so it has problems, but i will try to fix Apr 25 14:00:07 Well, i think all the distro wars are silly. We should all switch to Windows Vista. Apr 25 14:00:09 :p Apr 25 14:00:35 * mrmoku would rather buy some sheeps and herd them ;) Apr 25 14:00:42 mwester: They became less silly when you're choosing a distro to install on your new machine or to recommend to a friend. Apr 25 14:00:57 bah, console from linus branch Apr 25 14:00:58 Well, except when there are so many. Apr 25 14:01:14 Fedora vs Debian -- ok, I can see that - very different goals. Apr 25 14:01:33 But ubuntu, kubuntu, and how many other buntus? Apr 25 14:01:46 But here we go creating our own for the phone. :D Apr 25 14:01:50 or gentoo XD Apr 25 14:02:06 buntkuehe Apr 25 14:02:11 BTW, fellow girls i know (all 3 of them :) ) all installed Gentoo recently. Apr 25 14:02:17 Hehe! Isn't that just a word for "build it yourself" (gentoo) Apr 25 14:02:18 ? Apr 25 14:02:34 yup Apr 25 14:02:38 No, Gentoo is nice. Easier than OE on an order of magnitude. And as flexible. Apr 25 14:02:42 Heinervdm: ok, it has problems Apr 25 14:02:42 Or does it actually mean "no too alike, ever"? Apr 25 14:02:54 Heinervdm: KeyErrors Apr 25 14:02:54 mwester: no, gentoo has ebuilds. DIY would be linux-from-scratch Apr 25 14:02:57 s/too/two/ Apr 25 14:02:58 mwester meant: Or does it actually mean "no two alike, ever"? Apr 25 14:03:12 surname = contact_data['Surame']; del contact_data['Surname'] Apr 25 14:03:13 as well Apr 25 14:03:22 what if there is no "Surname" in contact_data? Apr 25 14:03:57 a third meaning is "no need for runtime config, it's all compiletime" Apr 25 14:04:21 Gentoo is great. Apr 25 14:04:25 (Argument of authority) Apr 25 14:04:35 (As every knows I'm the master around here) Apr 25 14:04:39 that's what "flexibility" is all about, isn't it Apr 25 14:04:40 +body Apr 25 14:04:40 Aha! Apr 25 14:04:55 Fix SHR! Apr 25 14:05:30 make my APN sticky Apr 25 14:06:14 or set up an option in this gentoo compiletime config file and build a distro for me ;-) Apr 25 14:06:38 By the way Apr 25 14:06:46 I realized something Apr 25 14:06:58 The most noisy device in my desktop is now an old hard drive Apr 25 14:07:07 hehe Apr 25 14:07:11 I was wondering how I could copy my system over another drive properly Apr 25 14:07:47 that's not that trivial in fact. depends on your fstab-option Apr 25 14:07:49 dos1: it will rise an exception imo Apr 25 14:08:02 Install Windows XP, then use something like Seagate's installation wizard on the CDROM you get with your new drive. Apr 25 14:08:04 Q-Master: yes. it's raising now :P Apr 25 14:08:13 but i think it shouldn't Apr 25 14:08:21 yep Apr 25 14:08:40 it should, but it must be prevented from doing so Apr 25 14:08:41 8) Apr 25 14:08:43 dos1: well nothing special Apr 25 14:08:45 hmpf Apr 25 14:08:47 DocScrutinizer: nothing special Apr 25 14:09:26 check if hd's are mounted by name, or by dunno what else Apr 25 14:09:39 well mounted as ever, by path Apr 25 14:09:44 (/dev/sdthingy) Apr 25 14:09:59 if you keep partition size, you just could dd then Apr 25 14:10:08 Well certainly not Apr 25 14:10:13 dd is an evil tool Apr 25 14:10:23 Plus I won't keep partition size. Apr 25 14:10:26 duh, why? Apr 25 14:10:31 ok Apr 25 14:10:34 DocScrutinizer: Well because I say so :-p Apr 25 14:10:35 then copy Apr 25 14:10:46 yeah but what about character devices? Apr 25 14:10:52 Can't recall how to do that properly Apr 25 14:11:01 Ainulindale: i transfered system using tar with proper parameters (to preserve device nodes, users, etc). Then reinstalled bootloader from livecd. Apr 25 14:11:09 it's nasty Apr 25 14:11:22 best do it from an auxiliary system Apr 25 14:11:32 PaulFertser: well that's what I was intending to do Apr 25 14:11:37 But I lost my tarcopy macro somewhere Apr 25 14:12:27 Ainulindale: or you could probably attach the second drive, partition it and then use rsync. Apr 25 14:13:06 you always get props with /dev & /sys Apr 25 14:13:31 and open files Apr 25 14:13:48 probs even Apr 25 14:14:00 DocScrutinizer: modern livecd's are damn nice. Apr 25 14:14:20 aux system was a synonym Apr 25 14:14:31 PaulFertser: hmmm didn't think about rsync Apr 25 14:14:56 Meanwhile, my white tea is gently infusing Apr 25 14:15:01 Ainulindale: never tried it to do a copy of device nodes, but i'm sure it has some cool option for that. And it's fast. Apr 25 14:15:13 I'd dd it, then modify partiton size with some smart tool Apr 25 14:15:25 DocScrutinizer: fyi it's gparted :) Apr 25 14:15:39 there's a g in that name Apr 25 14:15:43 one of them, yes Apr 25 14:15:46 it's almost as awful as software with k in their names Apr 25 14:16:25 Ainulindale: can you avoid using glib in your life? ;) Apr 25 14:16:34 try yast then, no k no g Apr 25 14:17:49 it seems rsync will take care of everything Apr 25 14:17:49 partiton magic is another one without g or k Apr 25 14:18:06 there's a g in that DocScrutinizer Apr 25 14:18:30 *cough* you're right Apr 25 14:18:48 go fix my APN! ;-D Apr 25 14:18:57 Your what? Apr 25 14:19:09 Acces Point Name Apr 25 14:19:17 What's that for? Apr 25 14:19:26 GPRS? Apr 25 14:19:27 This new thing called wifi people keeps talking about? Apr 25 14:19:35 Ah, the *other* new thing Apr 25 14:19:42 yup ;) Apr 25 14:19:45 Can't you use IRDA like everyone? Apr 25 14:20:05 nah, the hw-mod is WIP Apr 25 14:20:58 for GTA04 I suggested a laserpointer coupled to IRDA ;) Apr 25 14:21:06 meh :-) Apr 25 14:21:13 Can't wait for the GTA42 then Apr 25 14:21:24 I hope it'll be black, big, and will have "Don't Panic" on it somewhere. Apr 25 14:21:34 By the way this is a good case design IMHO :-) Apr 25 14:22:24 I'm still waiting for someone really doing the "DON'T PANIC" bootscreen I suggested Apr 25 14:22:38 Well if you find someone I'll put it as a default Apr 25 14:22:43 I'd rather use that than the present one Apr 25 14:22:54 I'll even do my fascist dictator Apr 25 14:23:04 Yeaaaaaaaaaah SOUND CHAAAASSEEER Apr 25 14:23:06 Heinervdm, Q-Master: lot of bugs, but it seems it works now to me :) Apr 25 14:23:38 dos1: have you fixed it? 8) Apr 25 14:24:03 CHA CHA CHA CHA ! CHA CHA ! Apr 25 14:24:18 Q-Master: no, there is still problem... with loading data at start Apr 25 14:24:43 but adding works Apr 25 14:24:46 ;) Apr 25 14:24:50 dos1: what's the prob? Apr 25 14:25:14 keys = line.keys() Apr 25 14:25:14 AttributeError: 'tuple' object has no attribute 'keys' Apr 25 14:26:42 Nope. Still no xserver starting. Unstable is truely busted. Apr 25 14:28:24 btw, whomever it may bother: uboot has a bug not noticing powerbutton releases. It's sufficient to start the device to press power shortly, then press a second time when the 8sec delay is expiring. Not the way it's supposed to work Apr 25 14:28:51 mwester: what's wrong? Apr 25 14:29:04 wakerd: error while loading shared libraries: libecore.so.0: cannot open shared object file: No such file or directory Apr 25 14:29:11 mwester: what did you guys did? :-) Apr 25 14:29:20 And indeed, there is no libecore.so Apr 25 14:29:29 used unstable ;-) Apr 25 14:29:30 Just lots of other libecore_*.so Apr 25 14:29:41 Since i know nothing of guis, I cannot fix. Apr 25 14:29:51 hahahaha Apr 25 14:29:53 What an ass Apr 25 14:30:04 mwester: Don't try to be me! Apr 25 14:30:10 That's my line :-) Apr 25 14:30:26 :P Apr 25 14:30:39 fucki*ng load_entries_from_db... Apr 25 14:30:45 hmm Apr 25 14:31:12 /usr/bin/enlightenment: error while loading shared libraries: libecore_con.so.0: cannot open shared object file: No such file or directory Apr 25 14:31:16 what's the "*" for? ;D Apr 25 14:31:35 And, yes indeed -- no libecore_con.so -- lots of other libecore_*, though. Apr 25 14:31:54 DocScrutinizer: hmm, know i don't know too ;D Apr 25 14:32:36 Ok, I updated the patches -- now who's going to figure out what dependency is missing so that I get my GUI libraries back??? Apr 25 14:33:06 (Don't go making my learn GUI stuff, that will make your lives muchly complicated in the long run!) Apr 25 14:38:05 dos1: yep. that's a problem. Apr 25 14:38:32 dos1: another thing to fix in opimd is being able to call InitAllEntries() twice Apr 25 14:39:17 mrmoku: now you can't? Apr 25 14:39:30 Grrrrrrrrr Apr 25 14:40:20 dos1: you can... but you will be left with doubled contacts ;) Apr 25 14:40:39 it does not clear the contacts first... just adds them all again Apr 25 14:40:48 mrmoku: ok Apr 25 14:41:31 dos1: might be interesting to have some function to get the state of a backend Apr 25 14:41:43 (think that was one more planned but not yet implemented thing) Apr 25 14:41:57 the dbus method is there... but did not work for me Apr 25 14:43:25 mrmoku: i said sth like "you need to know when to invalidate a cache" some hours ago Apr 25 14:43:39 why doesn't anybody listen? Apr 25 14:45:42 well because you talk like an old man :! Apr 25 14:46:36 DocScrutinizer: who was talking about cache? Apr 25 14:47:24 [2009-04-24 23:00:34] There's *no* sane reason for any kind of cache for contacts Apr 25 14:47:26 [2009-04-24 23:00:45] hmpf Apr 25 14:47:27 [2009-04-24 23:36:33] cache for contacts is mere nonsense if you don't consider when and how to invalidate that cache. *IF* you do, you'll find caching is a poor concept for most of the benefit you hope to gain Apr 25 14:47:29 [2009-04-24 23:38:43] either youwant to query a (remote) database, or you sync local and remote eventually Apr 25 14:47:45 DocScrutinizer: yeah... last night... but now? Apr 25 14:48:45 if any "init" gives you duplcate contact records, that's actually a sync doen the wrong way Apr 25 14:49:23 (my tought too) Apr 25 14:49:25 +h Apr 25 14:49:32 a correct sync needs to handle doublicates Apr 25 14:49:33 DocScrutinizer: well... just missing to clear the list of contacts first Apr 25 14:49:42 as well as conflicts Apr 25 14:50:35 that's an action I don't see for a local database synced to any "remote" sim or whatever. you never "clear the list" Apr 25 14:51:26 (agreed too) Apr 25 14:51:32 that's why my thought was Apr 25 14:51:38 first thing to do is to build a local representation (i.e. cache) Apr 25 14:51:42 with a unicity handling Apr 25 14:51:48 source, id, hash, whatever Apr 25 14:51:55 mrmoku: obviously you moved from "cache" to sqlite, but still you do a paradigma of "caching" Apr 25 14:52:22 I'd like to set some ldap, truly Apr 25 14:52:24 DocScrutinizer: no... did not move anywhere ;) Heinervdm did another opimd backend (sqlite) Apr 25 14:52:50 but what I was talking about is opimd startup... you have to initialize backends once it is started Apr 25 14:52:52 so this needs an "init"??? Apr 25 14:53:07 opimd needs to initialize the backends Apr 25 14:53:26 which loads the contacts (in memory me thinks) Apr 25 14:53:26 how TF could this "init" duplicate sqlite records then? Apr 25 14:53:47 don't know for sqlite records... but it duplicates the contacts from SIM Apr 25 14:54:18 that's WRONG. any init MUST NOT read data from SIM! Apr 25 14:54:33 tell that abraxxa ;) Apr 25 14:54:45 importing data from sim is best done on user interaction Apr 25 14:55:42 OR it has to handle duplicates, resolve conflicts (which is involving user interaction once again), etc pp Apr 25 14:56:08 s/OR/AND/ Apr 25 14:56:10 DocScrutinizer meant: AND it has to handle duplicates, resolve conflicts (which is involving user interaction once again), etc pp Apr 25 14:56:12 DocScrutinizer: (usb current limit) can you please take a look at this: http://git.openmoko.org/?p=kernel.git;a=commitdiff;h=9029dff1f370018665a6e2999632a34fd0518f4d Apr 25 14:57:12 so... does that mean we cannot start to use opimd before we have that nice unicity handling representation? Apr 25 14:57:18 (which we all agree is what we want)( Apr 25 14:57:38 PaulFertser: sounds as weird as all comments I've read regarding this. Apr 25 14:57:39 mrmoku: why not just avoid to read contacts from sim until explicitly asked? Apr 25 14:58:04 (asked) YUP, exactly Apr 25 14:58:05 DocScrutinizer: You turn off your FR and it suddenly starts to draw 500mA. Do you think it's impossible? Apr 25 14:59:03 nah, it's possible depending on what shutdown does to PMU registers. But I don't know of any dumb charger able to offer less than 500mA Apr 25 14:59:28 and there's no sense in the whole chargecurrent naggedynagg Apr 25 14:59:35 PaulFertser: again... ask that abraxxa. Maybe I misunderstood the opimd code wrong (my python foo is very limited) Apr 25 14:59:56 Just know for sure that when you call InitAllEntries twice... contacts are doubled when you query them Apr 25 15:00:25 mrmoku: i know next to nothing about opimd. But i'm sure what Doc (and Ainu) says about syncing that kind of data is true. Apr 25 15:02:04 DocScrutinizer: do you mean there's no sense at all in limiting usb current to 100mA except for the time while FR attempts to be enumerated? Then yes, probably caring about drawing less than 500mA in OFF state is pointless. Anyway, do you see any harm from that? Apr 25 15:02:32 not if it's done correctly Apr 25 15:02:37 PaulFertser: agree on that. We want that. But does that mean we cannot start using it? Apr 25 15:02:45 DocScrutinizer: How it is not? Apr 25 15:02:57 If we don't start using opimd it will never get fixed Apr 25 15:03:02 except of senseless complication of code Apr 25 15:03:11 mrmoku, ++ Apr 25 15:03:18 I'm in favor of using opimd. Apr 25 15:03:22 mrmoku: i guess dos1 will promptly reply you :) He'll just disable reading contact from SIM for now, i guess and it'll be all right :) Apr 25 15:03:25 PaulFertser: (not) dunno Apr 25 15:03:40 PaulFertser: maybe it's done correctly Apr 25 15:03:57 PaulFertser: but it's completely pointless Apr 25 15:04:01 DocScrutinizer: the code really looks ok. Just after any change of usb current limit, charging limit is altered accordingly. I see no reason to be concerned here. Apr 25 15:05:00 PaulFertser: I'm more concerned about the idea to shutdown the device during charging at all Apr 25 15:05:46 Q-Master: ok, it works :D Apr 25 15:05:54 DocScrutinizer: i think i can imagine a dumb charger not suitable for providing 500mA. Now consider following scenario: you hook up your FR to it, look at the current, see 100mA, think ok, that will charge it slowly, i'm ok with that. Turn your FR off, and go to sleep. Then it starts to draw 500mA and dumb charger bursts in flames. Apr 25 15:06:42 just add a smoke detector! Apr 25 15:06:49 lindi-: do you have one at home? Apr 25 15:06:53 there's no such thing like a dumb charger busting nto flames @ 500mA Apr 25 15:07:26 DocScrutinizer: but it's not what you would expect from a device anyway (suddenly change charging current on turning off) Apr 25 15:08:50 and -nota bene- there's no decent way to tell about a dumb charger that is able to ofer nly 100mA - it's DUMB. So best thing we can do is what I metioned in my recent post to ML: assume any dumb charger being able to ofer USB-STANDARD of 500mA Apr 25 15:09:18 PaulFertser: as that's exactly what a million of other stupid USB-gadgets do as well Apr 25 15:09:57 like USB-teacup-warmers, usb-fans, USB-vibrators... ;-) Apr 25 15:10:04 DocScrutinizer: imho we should ensure consistency between charging current in ON and OFF modes anyway. That's exactly what the patch does. Apr 25 15:10:18 huh? Apr 25 15:10:55 DocScrutinizer: i agree if we never (except while trying to enumerate) use 100mA mode, then the patch makes no sense. Apr 25 15:11:16 why care about charging current? We need to care about USB-current - *IF* *anything* Apr 25 15:11:55 dos1: hacker. 8) Apr 25 15:11:58 charging current is strictly to cope with maximum charge current battery cell can handle. Apr 25 15:11:58 DocScrutinizer: and i agree it makes sense to use the scheme you proposed here today (wait for some time, if we can't enumerate, switch to 500mA anyway). We can't control usb current in "pcf standby mode", so the patch controls charging current instead. Apr 25 15:12:39 freesmartphone.org: 03seba.dos1 07framework * r6bfd89af907e 10/framework/subsystems/opimd/ (__init__.py pimb_sqlite_contacts.py): opimd: add sqlite contacts backend by Heinervdm and modify it to work Apr 25 15:12:50 I would prefer set usb_curlim to 0 in PMU-standby mode. Apr 25 15:13:27 well it's not the best idea as we might have troubles to boot on low bat hen ;-) Apr 25 15:13:34 o shit Apr 25 15:13:54 anyway the device shouldn't go to shutdown (=pmu-standby) during charging at all Apr 25 15:14:10 DocScrutinizer: even if you explicitly ask it? Apr 25 15:14:16 it's a nonsense usecase Apr 25 15:14:32 However it was a real life observation :) Apr 25 15:15:12 if you force it to shutdown (which by he way is the only way to shutdown ;), then charging should be stopped. I'd never expect a switched-off device to chage Apr 25 15:15:16 The guy who reported it noticed some led of his charger changing state indicating that suddenly the current became more. Apr 25 15:15:36 WTF, I don't care!!! Apr 25 15:15:52 he's NOT SUPPOSED to shutdown Apr 25 15:16:07 DocScrutinizer: can you please propose that (completely switch off charging on shutdown) on hardware mailing list? Apr 25 15:16:20 freesmartphone.org: 03seba.dos1 07framework * r12de19b81e0f 10/framework/subsystems/opimd/pimb_sqlite_contacts.py: opimd: convert sqlite backend from windows to unix format Apr 25 15:16:31 and if he does, I recommend to set bat_curlim to 0mA in that very moment Apr 25 15:16:41 ok, /me needs to go out Apr 25 15:17:22 That sounds reasonable to me. But will it be consistent enough with turned off device after replugging a battery? (i haven't read pcf datasheet in details). Apr 25 15:17:32 PaulFertser: the device for very good reason boots automatically when inserting USB. It's for enabling controlling charge Apr 25 15:18:31 PaulFertser: (replugging bat) there's very little we can do for hat case *anyway* Apr 25 15:18:49 If you disable charging while shutdown, then how does one recover from the totally flat battery problem? Apr 25 15:19:01 and, as mentioned above, device ots/resumes on USB insertion Apr 25 15:19:06 DocScrutinizer: (auto-boot) yeah, i agree that's good. But it should try to set curlim to 1A if booted with wall-charger, that's a major issue for a flat battery problem. Apr 25 15:19:15 And BTW, I *do* expect devices turned off to be able to charge. Apr 25 15:19:34 I don't expect SMART charging, but I expect it to trickle to maintain the battery. Apr 25 15:19:59 mwester: btw, i checked that myself. Most of the time i'm able to start NOR u-boot on my v6 without a battery. That proves i can charge any flat battery with it. Apr 25 15:20:02 is shr-devel dead? I didn't get any mail since the move. Apr 25 15:20:18 mwester: I recover from low bat exactly the same way it's done right now: uBoot setting up PMU registers to start charging Apr 25 15:20:46 Ok. Then it's just the user expectation of what it does shutdown but plugged in. Apr 25 15:21:31 I think users expect that it will at least trickle-charge. But I think that they do not expect it to suddenly pull MORE than it was when it was running, that is a more serious problem. Apr 25 15:22:55 and I really don't care about whether the device is charging or not during shutdown state. I simply say it's nonsense to assume we need to handle usbcurrents of less than 500mA during shutdown state. The way it's done now you also don't know what's going to happen when you change charger to a weaker one while shutdown Apr 25 15:25:54 that's so completely non-issue I don't know how anybody can create tons of patches with flase assumtions to correct a hw-behavior thats good enough for all other devices. And "correct" here means replace a simple working scheme by a complex senseless workarround Apr 25 15:26:44 mwester: my se P1i is not trying to charge while off Apr 25 15:26:45 Well, here you and I are in agreement. Apr 25 15:27:07 again: THERE IS NO SUB-500mA DUMB CHAGER Apr 25 15:27:27 Q-Master: I don't doubt that there exist some devices that do that; I simply point out that my expectation and experience in general is that devices do charge while turned off. Apr 25 15:28:12 not all Apr 25 15:28:23 i've often wondered about the maniacal focus on the "must not pull >100ma unless we have been given permission" thing -- few devices follow that rule. Apr 25 15:28:25 but this should be a good feature Apr 25 15:28:43 om so dn't touch the whole shit, let do PMU it's thing, and hapy we are (except the one guy with his virtual nonexistent 150mA dumb charger) Apr 25 15:29:09 s/om so dn't/OK so don't/ Apr 25 15:29:10 DocScrutinizer meant: OK so don't touch the whole shit, let do PMU it's thing, and hapy we are (except the one guy with his virtual nonexistent 150mA dumb charger) Apr 25 15:29:32 * mwester has a few pcmcia usb cards that cannot do 500ma, but they do not light up in flames when devices attempt to pull that much current either. Apr 25 15:30:16 * mwester thinks he as found the trouble with unstable and E Apr 25 15:30:51 * mwester suspects the experts will find it shortly, when they attempt to flash unstable. Apr 25 15:31:17 my suggestion was: try enable 500mA, see if Vusb is dropping by more than a certain limit (=overload), *IF* overload detected switch off USB current for 10sec and start over with lower USB_CURLIM Apr 25 15:32:07 That's a far better heuristic than we have at present -- it is even useful for dumb chargers (solar panels, etc) Apr 25 15:32:17 yup Apr 25 15:32:45 and (btw) it's nspired by USB2.0 chager spec Apr 25 15:32:51 I rather like it, the more I think about it. Apr 25 15:34:12 so if there's any dumb charger outside with a writing on it "Do *NOT* use with USB2.0 devices" I'd recommend not to use it for FR either. For all the other dumb chagers the whle thing is an *annoying* non-issue Apr 25 15:35:36 no country on this wide earth of God is allowing a charger that goes up in smoke on overload Apr 25 15:35:39 * PaulFertser wished Doc came with all that insightful comments on kernel mailing list when the issue was raised first (no sarcasm intended) Apr 25 15:36:04 * Blu3 gives doc a saturday happy pill Apr 25 15:37:02 DocScrutinizer: also, this complex scheme involving measuring Vusb and timeouts is probably too inelegant to be integrated in kernel. It'd be perfectly ok to see it implemented in userspace though. Apr 25 15:37:18 yup Apr 25 15:37:49 especially as it's good for >500mAdumb chargers as well Apr 25 15:38:14 my usb battery can only do 450 mA :) Apr 25 15:38:25 but maybe that counts as 500 mA Apr 25 15:38:32 It should be prototyped in userspace, at least. But I can't see that it would really add much more to the already complicated PCF driver... and it would make things lots better for the user of any distro. Apr 25 15:38:56 kernel should kept simple, just setting PMU regs to bare minimum needed to make PMU do it's thing which is good enough for most of the time anyway Apr 25 15:39:30 wel, I don't have a strong opinion here either way :) Apr 25 15:39:43 mwester just likes to care about distros that are too conservative/lacking manpower/dumb to do the right thing in userspace ;) Apr 25 15:39:58 Yeah. Like debian. :p Apr 25 15:40:07 i'm generally of the opinion that the kernel driver should be thin and implement/enforce policy that comes from userland Apr 25 15:40:30 yu Apr 25 15:40:31 I generally agree. Apr 25 15:40:32 p Apr 25 15:40:39 i don't like seeing everything done in the kernel, makes it bloated Apr 25 15:40:44 But for certain critical devices, i'm willing to make an exception. Apr 25 15:40:47 which, makes it prone to error Apr 25 15:41:10 yes, some things necessarily should be done in kernel land Apr 25 15:41:15 see the recent "patch" for nonexistent problem Apr 25 15:41:50 * Blu3 has had one hour of quasi sleep, where is this patch? Apr 25 15:41:58 Blu3: you can say that about opening up the kernel interfaces to user-space; the paranoid me imagines some user setting insane values and breaking things. Apr 25 15:42:12 Blu3: I think Doc talks about http://git.openmoko.org/?p=kernel.git;a=commitdiff;h=9029dff1f370018665a6e2999632a34fd0518f4d Apr 25 15:42:16 which, I bet, would spoil my day if I decided I want to charge my cell with only 59mA Apr 25 15:42:32 yes, i think all apis of the kernel should only accept sane values Apr 25 15:42:57 mwester: pilot error Apr 25 15:43:01 only 59ma? heheh, so uhm, you'd never -use- your fone? :> Apr 25 15:43:41 maybe I want to test a completely different battery? Apr 25 15:44:22 or heal a deep-discharged one the safe way? Apr 25 15:44:36 aye Apr 25 15:44:45 i'll bbiab, we're @ target Apr 25 15:45:27 I'm away now, exploit the last daylight to get my first *decent* coffe Apr 25 15:50:02 * Q-Master starting to hate googlecoders.... Apr 25 15:50:27 mirko-paroli: hi Apr 25 15:50:43 hi Q-Master Apr 25 16:03:20 * Q-Master had built adb for powerpc linux... that was fun... Apr 25 16:08:57 not connecting to neo Apr 25 16:29:30 hi, I've a question about kdrive, I've a device with kdrive 1.3...what should I do to the scripts for making it use my hardware keyboard? it works if I use Xfbdev :0.0 -fb /dev/fb1 tty Apr 25 16:29:51 but I tried modifying scripts in /etc/X11 but strangely it doesn't work Apr 25 16:47:05 howdy! Apr 25 16:47:44 Hehe :) Apr 25 16:47:44 DocMobilizer: hehe... had your good coffee? Apr 25 16:49:04 just about. ircing on FR Apr 25 16:52:20 pidgin is a little crammed with all these menu/tool-bars Apr 25 16:53:11 just 5 lines left for actual channel log Apr 25 16:57:03 éåññññññññññññññññññññ Apr 25 16:57:06 sorry Apr 25 16:57:10 Ainulindale, I've added Intone package request Apr 25 16:57:15 http://www.shr-project.org/trac/ticket/415 Apr 25 16:57:25 * Q-Master shouts YESSSSSSSSSSSSSSSSSSSSS Apr 25 16:57:32 adb works Apr 25 16:59:56 docmobile, pidgin is handy but yes, it's ui isn't really suited to a small device Apr 25 17:01:46 DocMobilizer, Blu3 is there any alternative to pidgin ? Apr 25 17:02:18 i haven't found anything that i like Apr 25 17:02:26 hmm shame Apr 25 17:02:33 which is unfortunate since pidgin is turning into dinosaur-ware Apr 25 17:02:54 by the time pidgin gets A/V, the rest of the world will have 3d holographic IM Apr 25 17:03:13 LOL Apr 25 17:03:46 just testing IRC on FR for very first time Apr 25 17:03:52 :> Apr 25 17:04:11 i did some jabbering, but that wasn't too handy either Apr 25 17:04:33 i mean, how many years has it been that we've had A/V, themes - wallpapers etc, in other IM clients? Apr 25 17:06:05 prefs on pidgin is inoperable since leaving screan Apr 25 17:06:38 s/ean/een/ Apr 25 17:06:39 DocMobilizer meant: prefs on pidgin is inoperable since leaving screen Apr 25 17:07:05 DocMobilizer: with onscreen keyb or external? Apr 25 17:07:30 onscreen terminal Apr 25 17:08:00 * mrmoku thinks transparent keyboards would be very handy for that Apr 25 17:08:05 but no kbd when pref Apr 25 17:08:37 there are a number of dialogs where i rely on "" to exit them since the "ok" button is off screen, and you can't scroll to reach it Apr 25 17:11:15 'cmd mozilla %s not found' - lol Apr 25 17:11:25 DocMobilizer, there's also xchat for the openmoko, isn't it? Apr 25 17:11:38 dunno Apr 25 17:11:51 I've seen some photo of it running, but never got it on my fr Apr 25 17:12:03 would prefer that i guess Apr 25 17:12:56 it's on OE Apr 25 17:12:59 anyway kbd-shortcuts where handy Apr 25 17:13:27 mrmoku, what about building xchat package from OE repos? :D Apr 25 17:13:31 ok, latest andy-tracking indeed has WSOD Apr 25 17:13:32 like 'pinternet.interkom.dee' Apr 25 17:14:36 lindi__: someone messed it up :-) Apr 25 17:14:56 mirko-paroli: what's new with sqlite in paroli? Apr 25 17:15:05 hehe Apr 25 17:15:06 * mrmoku tries exactly that in this moment :-) Apr 25 17:15:16 you really want sql in paroli eh :) Apr 25 17:15:46 mirko-paroli: I want contacts-not-from-sim-card Apr 25 17:15:51 dunno what for Apr 25 17:15:54 that paroli has Apr 25 17:16:01 just not with a sql backend Apr 25 17:16:09 and with too few fields for you iirc Apr 25 17:16:18 mirko-paroli: text file? Apr 25 17:16:29 use opimd! Apr 25 17:16:36 pickly, but easily changable to text file, yes Apr 25 17:16:46 pickle Apr 25 17:17:08 mirko-paroli: use opimd. it is now able to store contacts in sqlite. 8) Apr 25 17:17:13 last time I enquired it still was not ready Apr 25 17:17:16 SHR: 03mok 07libframeworkd-phonegui-efl2 * rde6857adedea 10/ (15 files in 4 dirs): remove toolchain craft, fix build and copy po from efl Apr 25 17:17:26 SHR: 03mok 07libframeworkd-phonegui-efl2 * r202fe0b492a5 10/src/view/message-show-view.c: message-show: fix segfault when message has no timestamp Apr 25 17:17:58 Sharwin_F: built fine (xchat) Apr 25 17:18:09 can I save, retrieve, edit etc ? Apr 25 17:18:17 talking contacts for now Apr 25 17:18:27 mirko-paroli: retrieve and add for now 8) Apr 25 17:18:41 installing Apr 25 17:18:41 right forgot delete Apr 25 17:18:41 no edit, no delete yet Apr 25 17:18:49 Sharwin_F: built it locally ;) Apr 25 17:18:58 meh Apr 25 17:19:22 :P Apr 25 17:19:31 installing and testing now... if it works... then Apr 25 17:19:37 edit = create + delete Apr 25 17:19:50 DocMobilizer: delete does not work yet :P Apr 25 17:20:35 "Running IRC as root is stupid!" is what it tells me ;) Apr 25 17:20:35 should be easy Apr 25 17:21:03 and it's damn right Apr 25 17:21:48 hope it quits after that Apr 25 17:23:29 hmm Apr 25 17:23:44 hmmmmmmmmmm Apr 25 17:23:51 font is too big Apr 25 17:26:09 but it works Apr 25 17:28:06 actually for pidgin i kinda second this Apr 25 17:28:48 though all the toolbars are annoying as well Apr 25 17:29:57 might be tweakable... will add it to the SHR feed... after dinner ;) Apr 25 17:31:14 that is something that bothers me, everything in shr runs as root Apr 25 17:31:28 terminal kbd is wasting 10% of screen for a useless bold header Apr 25 17:32:19 Blu3: pester Ainulindale to add decent users :) Apr 25 17:32:37 i'll join in Apr 25 17:33:32 ;> Apr 25 17:34:48 prolly everything borks as FR/OE never had users and app-devels never cared Apr 25 17:35:34 probably Apr 25 17:37:52 wolfgang always said 'community will add that eventually'. pah. like building a house and care about cellar last Apr 25 17:38:57 yeah, really Apr 25 17:38:59 :/ Apr 25 17:39:41 it's really better to build a strong foundation/structure, and let the community add the pretty wall decorations Apr 25 17:40:11 well, suspending now. let's see if pidgin will recover on resume in 10min Apr 25 17:40:22 bbl Apr 25 17:40:50 maybe, pidgin crashed on me last night when i did that Apr 25 17:41:45 i'd expext the same ;-> Apr 25 17:46:23 WAKE UP SHR-devs!! Apr 25 17:46:51 * mwester tosses an angry wet cat into Ainulindale's living room to get his attention Apr 25 17:46:53 mwester: is there anything i could do wrt. the WSOD bug? Apr 25 17:48:11 lindi-: Dunno - I see you have a FR that can reproduce the problem, at least. Apr 25 17:48:17 Do you use u-boot or qi? Apr 25 17:48:32 mwester: lol Apr 25 17:48:37 Ah u-boot. Apr 25 17:48:46 * mwester read the whole bug report now. Apr 25 17:49:42 mwester: s/cat/tasmanian devil/ Apr 25 17:50:01 ;-) Is there much of a difference, really? Apr 25 17:50:12 * mwester prefers dogs over cats anytime. Apr 25 17:51:13 mwester: depends on preparation Apr 25 17:51:25 hehe!! Apr 25 17:51:41 You've spent entirely too much time in certain parts of Asia. Apr 25 17:51:49 hehe Apr 25 17:52:24 actually my best cangaroo i ate here in Germany Apr 25 17:53:15 crocodile wasn't that tasty tho Apr 25 17:53:35 (was a mixed plate ;) Apr 25 17:54:13 mwester: uboot Apr 25 17:54:13 or is "dish" the right word Apr 25 17:54:18 ? Apr 25 17:59:06 (wsod) my uneducated guess: as WSOD recovers now on a second suspend-resume cycle, I assume it's actually powered down by LDO6 now. As it never did back in Nov. Had to remove LCM-FPC and reinsert to make it recover without powerdown of whole device Apr 25 17:59:55 so next guess is: race between powerup LDO and reset/init of LCM at resume time Apr 25 18:00:55 as when powerup whole device, PMU sequencing is determining when LDOs get enabled. Not so on resume Apr 25 18:01:25 (thet's for difference of boot never showing WSOD, but resume does) Apr 25 18:01:42 thAt Apr 25 18:02:22 ergo: delay LCM init and let LDO voltage stabilize, on resume Apr 25 18:02:58 mwester: agree? for the logic at least Apr 25 18:08:49 DocScrutinizer: does that fit with the observation made by many, that just flashing Qi as a bootloader resolves the issue? Apr 25 18:09:20 yup, as Qi afail doesn't nit LCM at all. Does it? Apr 25 18:09:31 afaik Apr 25 18:09:40 init Apr 25 18:09:43 waaah Apr 25 18:10:20 so maybe kernel does it right, but uBoot doesn't know how to handle it correctly Apr 25 18:13:05 mirko-paroli: if opimd is not ready... why aren't you working on opimd then? Apr 25 18:13:24 dos1|away: coz I need to fix other things Apr 25 18:13:53 so another case where kernel doing the right thing causes fail in bootloader. Seems I've heard same story before ;-) Apr 25 18:13:56 mirko-paroli: no. then you are developing you own custom thing, which will be deprecated later Apr 25 18:14:19 s/you own/your own/ Apr 25 18:14:21 dos1 meant: mirko-paroli: no. then you are developing your own custom thing, which will be deprecated later Apr 25 18:14:33 hi mirko-paroli Apr 25 18:14:34 dos1: sry I don't feel like your bickering today, I am not doing anything related to PIM right now as there are pther issues in paroli Apr 25 18:14:41 hi DocScrutinizer Apr 25 18:16:11 mirko-paroli: right now. but writing editing and deleting support to opimd looks like easier than writing own PIM system. That's only my opinion, don't get me wrong :P Apr 25 18:16:13 dos1: seems that 1st steps in import script are done. 8) it successfully prints the contents of vcf file. 8) Apr 25 18:16:48 dos1: sure, but as I said, right now I am not doing any PIM stuff Apr 25 18:16:54 Q-Master: so now it should be easy :) python? Apr 25 18:17:01 yep Apr 25 18:17:32 Q-Master: i can send you code for adding contact Apr 25 18:18:15 dos1: it'll be great. I now just use std sqlite api 8) Apr 25 18:18:41 mirko-paroli: ok. i don't want to fight with you ;) i'm only telling my opinion ;] Apr 25 18:19:07 dos1: and I am saying that I never had any other opinion and don't like people to say otherwise Apr 25 18:19:28 ok :) Apr 25 18:19:34 that's all i wanted to know Apr 25 18:19:48 dos1: next time you want to know something, just ask Apr 25 18:19:52 please Apr 25 18:26:26 freesmartphone.org: 03mickey 07framework * r9f17c047cb47 10/ (ChangeLog framework/resource.py): ousaged: Serialize resources with an asynchronous worker queue. This should ease fixing FSO #404. Apr 25 18:26:27 mirko-paroli: low mood today? Apr 25 18:27:11 s/low/foul/ Apr 25 18:27:12 DocScrutinizer meant: mirko-paroli: foul mood today? Apr 25 18:27:22 DocScrutinizer: hehe, no, there is just some folks that tend to accuse me with some stupid stuff and I am just not willing to listen and smile any longer Apr 25 18:27:33 I am actually rather happy Apr 25 18:27:37 had a great day :) Apr 25 18:27:39 how about you? Apr 25 18:27:47 fine! Apr 25 18:28:26 nice Apr 25 18:28:28 :) Apr 25 18:29:49 made a point about using improved USB-powermanagement scheme that convinced even mwester - so how could this day get any better :-) Apr 25 18:29:56 hehe Apr 25 18:30:33 actually I love it when others apreciate my ideas Apr 25 18:30:47 who wouldn't ? ;-) Apr 25 18:30:58 my thought exactly :) Apr 25 18:49:59 Q-Master: contacts = getDbusObject (bus, "org.freesmartphone.opimd", "/org/freesmartphone/PIM/Contacts", "org.freesmartphone.PIM.Contacts") Apr 25 18:49:59 contacts.Add({'Name':'dos', 'Number':'+48663571389'}) Apr 25 18:50:10 getDbusObject is my helper Apr 25 18:50:36 which i've stolen from some app which i found on maillist ;) Apr 25 18:50:55 looks nice :-) Apr 25 18:51:11 dos1: what modules should I import? Apr 25 18:51:20 dbus Apr 25 18:51:23 and helper: Apr 25 18:51:37 def getDbusObject (bus, busname , objectpath , interface): Apr 25 18:51:37 dbusObject = bus.get_object(busname, objectpath) Apr 25 18:51:37 return dbus.Interface(dbusObject, dbus_interface=interface) Apr 25 18:51:39 and that's all Apr 25 18:51:45 import dbus Apr 25 18:51:48 then define helper Apr 25 18:51:53 and then add contact Apr 25 18:51:53 ;) Apr 25 18:52:48 * dos1 is going to post his opimd test app to some wiki Apr 25 18:53:45 dos1: is there some Cell number, home number, etc? Apr 25 18:54:19 * Q-Master is crowling through python-vobject examples and docs.... horrible thing Apr 25 18:54:27 python-vobject? Apr 25 18:54:42 i don't know fields Apr 25 18:54:59 but they should be in opimd sources Apr 25 18:55:17 some txt file with fields Apr 25 18:55:54 fields are documented in docs dir in opimd Apr 25 18:56:38 dos1: one thing... Apr 25 18:56:49 [OT] reign over me is a nice movie Apr 25 18:56:50 fields of an opimd contact can come from different backends Apr 25 18:57:45 dos1: yep. python-vobject.... horrible thing Apr 25 18:57:56 mrmoku: the answer is: i don't know. i don't know opimd well. i'm studying code when i'm changing, when i'm changing it :) Apr 25 18:58:17 s/code when/code which/ Apr 25 18:58:19 dos1 meant: mrmoku: the answer is: i don't know. i don't know opimd well. i'm studying code which i'm changing, when i'm changing it :) Apr 25 18:58:29 dos1: was no question ;) Apr 25 18:58:46 mrmoku: but it's universal answer. Apr 25 18:58:46 :D Apr 25 18:59:02 even for no questions ;) Apr 25 18:59:06 more importantly... you can have multiple fields of the same type Apr 25 18:59:18 like an arbitrary number of CellPhone numbers Apr 25 19:01:55 mrmoku: now i want to have basics of opimd working on actual design. then we can think about new design, as now i don't feel enough power to change it's design :D Apr 25 19:02:25 fine for me :) Apr 25 19:03:57 seems that it is mainly focused on vcalendar Apr 25 19:08:18 hehe, combining different sources with complementary info into a single contact is rather tricky Apr 25 19:10:00 DocScrutinizer: guess that's why it is just designed that way... and not yet implemented :P Apr 25 19:10:43 mrmoku: imagine editing such a contact ;) Apr 25 19:13:46 i have python script which searches for contacts by name given in argument Apr 25 19:13:52 using opimd Apr 25 19:14:26 when you request "do" you will get "dos", "dom", "dot", "dotty" etc. Apr 25 19:15:32 so that type of requests can be usable for filtering in contact app ;) Apr 25 19:16:58 dos1: yup, but it should also search instring. to be honest it should do a regex match over all fields Apr 25 19:18:12 that is one thing I did not find out how to do it... search over multiple fields Apr 25 19:18:23 (with one expression) Apr 25 19:18:26 or, alternatively, restrict the regex to a set of fields only. finding name for a number (substring) is a common demand Apr 25 19:19:31 freesmartphone.org: 03mok 07libframeworkd-glib * r6f2e7ce113a0 10/ (13 files in 4 dirs): Apr 25 19:19:31 freesmartphone.org: opimd: start implementation of wrappers for opimd calls Apr 25 19:19:31 freesmartphone.org: This is completely untested... Should compile though. Apr 25 19:19:31 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 25 19:19:33 freesmartphone.org: 03mok 07libframeworkd-glib * re945a1ff12ca 10/src/ (7 files in 3 dirs): Apr 25 19:19:35 freesmartphone.org: opimd: add wrappers fro sources interface Apr 25 19:19:37 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 25 19:19:39 mrmoku: which interface? Apr 25 19:19:39 freesmartphone.org: 03mok 07libframeworkd-glib * raae989c07be4 10/src/ (frameworkd-glib-dbus.c frameworkd-glib-dbus.h): Apr 25 19:19:42 freesmartphone.org: dbus: const correctnes for connect_to_interface Apr 25 19:19:44 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 25 19:19:46 freesmartphone.org: 03mok 07libframeworkd-glib * r8176de3c87f0 10/src/opimd/ (2 files): Apr 25 19:19:48 freesmartphone.org: opimd/contacts: keep the DBusGProxy for queries Apr 25 19:19:50 freesmartphone.org: This is needed, because opimd remembers the query cursor Apr 25 19:19:52 freesmartphone.org: for each client connection... Apr 25 19:19:54 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 25 19:19:58 freesmartphone.org: 03mok 07libframeworkd-glib * r9c43c1aa4116 10/src/opimd/ (frameworkd-glib-opimd-dbus.c frameworkd-glib-opimd-dbus.h): Apr 25 19:20:01 DocScrutinizer: opimd Query Apr 25 19:20:01 freesmartphone.org: opimd/sources: add missing dbus connect functions Apr 25 19:20:03 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 25 19:20:05 freesmartphone.org: 03mok 07libframeworkd-glib * r32aca2c5a5c4 10/src/opimd/dbus/ (contacts.h contacts.xml): Apr 25 19:20:10 freesmartphone.org: opimd/contacts: implement GetMultipleResults Apr 25 19:20:12 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 25 19:20:14 freesmartphone.org: 03mok 07libframeworkd-glib * r8abecf8adf5e 10/src/opimd/frameworkd-glib-opimd-contacts.c: Apr 25 19:20:17 freesmartphone.org: opimd/contacts: fix messages/contacts typo and add missing cast Apr 25 19:20:22 mrmoku: maybe spec flaw? Apr 25 19:20:23 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 25 19:20:25 freesmartphone.org: 03mok 07libframeworkd-glib * r9c2bae8f2a12 10/ (20 files in 4 dirs): Merge branch 'opimd' Apr 25 19:20:28 freesmartphone.org: 03mok 07libframeworkd-glib * r7301d4e1ec7f 10/src/opimd/ (4 files): Apr 25 19:20:30 freesmartphone.org: opimd: add G_BEGIN_DECLS / G_END_DECLS Apr 25 19:20:32 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 25 19:20:48 DocScrutinizer: or just not documented / implemented... don't know Apr 25 19:21:03 dos1: see? I did it ;) Apr 25 19:21:26 regarding complexity of the whole thing I wouldn't be surprised Apr 25 19:22:50 DocScrutinizer: substring works too. when i requests "os" it returns "dos" Apr 25 19:23:32 hmm, that's half the beef Apr 25 19:24:34 DocScrutinizer: and after seeing code (not tested) - it sorts results, so the best matching is on top Apr 25 19:24:47 cool Apr 25 19:25:25 DocScrutinizer: could lcd code be completely skipped by uboot on resume? Apr 25 19:26:01 so all we need is selecting fieldnames Apr 25 19:26:30 http://openmoko.opendevice.org/~dos/pim.py Apr 25 19:26:46 mrmoku: so now - bb recipe for libframeworkd-phonegui-efl2! :D Apr 25 19:26:48 lindi-: I guess it could, but then we get problems with bootmenu Apr 25 19:27:02 DocScrutinizer: but it is never shown on resume anyway? Apr 25 19:27:11 dos1: yep... have it already locally Apr 25 19:27:13 DocScrutinizer: ping Apr 25 19:27:20 pong Apr 25 19:27:32 Q-Master: http://openmoko.opendevice.org/~dos/pim.py Apr 25 19:27:42 hi Apr 25 19:27:53 lindi-: yup, but uboot has no idea I guess whether it's boot or resume Apr 25 19:28:03 I just installed SHR-testing of April 22 Apr 25 19:28:15 DocScrutinizer: What could be the reason if the call volume of the GTA02_A5 is real low after rework? Apr 25 19:28:16 DocScrutinizer: but it has to have? otherwise it wouldn't know when to show the menu and when not? Apr 25 19:28:25 looks good, am able to make calls, open SSH connection so it's looking good Apr 25 19:28:32 SHR: 03seba.dos1 07shr-settings * rf94ee9084ea2 10/playground/wifiman.py: [playground] wifiman.py: fix crashes on networks with hidden ssid Apr 25 19:28:43 hmm, menu is shown by holding Aux Apr 25 19:28:48 however I can't figure out how to install other apps Apr 25 19:29:02 maybe not on resume. no idea Apr 25 19:29:02 (is the first distro I get working on the FreeRunner) Apr 25 19:29:32 opkg list and opkg list_installed return basically the same list of packages Apr 25 19:29:45 m8ram: opkg update? :P Apr 25 19:30:00 alphaone: pong Apr 25 19:30:05 DocScrutinizer: What could be the reason if the call volume of the GTA02_A5 is real low after rework? Apr 25 19:30:14 really Apr 25 19:30:18 dos1: opkg update returns: Apr 25 19:30:19 mic? Apr 25 19:30:23 yeah Apr 25 19:30:29 http://openmoko.opendevice.org/~dos/pim.py Apr 25 19:30:31 It's really noticable... Apr 25 19:30:34 wrong statefile? Apr 25 19:30:46 m8ram: i don't think it returns that link ;) Apr 25 19:31:07 (and similar for armv4, armv4t and om-gta02 Apr 25 19:31:12 Resistance between mic and GND is 2.1k Apr 25 19:31:17 then http://openmoko.opendevice.org/~dos/pim.py Apr 25 19:31:30 control63 not mic2? Apr 25 19:31:32 DocScrutinizer: Well, it always worked good with the test image I flashed... Apr 25 19:31:35 nope, wrong copy buffer: * Failed to download http://build.shr-project.org/shr-testing/ipk//all/Packages.gz, error 0 Apr 25 19:31:36 which reminds me... Apr 25 19:31:37 m8ram: looks like you really like my script :D Apr 25 19:31:43 Maybe he flashed the wrong one.. Apr 25 19:31:53 m8ram: can you ping some internet server from FR? Apr 25 19:31:59 :) Apr 25 19:32:15 that was my next try and it looks like that's the issue: Apr 25 19:32:24 dos1: here you go Apr 25 19:32:28 m8ram: cat /etc/resolv.conf Apr 25 19:32:32 SHR: 03mok 07shr-overlay * r640f6388655c 10/openembedded/ (2 files in 2 dirs): add recipe for libframeworkd-phonegui-efl2 Apr 25 19:32:36 :> Apr 25 19:32:42 mrmoku: testing testing testing testing Apr 25 19:32:42 :D Apr 25 19:32:50 cat /etc/resolv.conf Apr 25 19:32:51 nameserver 195.130.129.164 Apr 25 19:32:51 nameserver 195.130.130.164 Apr 25 19:32:51 nameserver 194.7.38.242 Apr 25 19:32:54 works nicely to install in parallel to old efl Apr 25 19:33:03 but nslookup looks like it's timing out Apr 25 19:33:06 and then switch via /etc/frameworkd-phonegui.conf Apr 25 19:33:06 m8ram: can you test with opendns servers? Apr 25 19:33:18 nslookup www.google.com Apr 25 19:33:19 Server: 195.130.129.164 Apr 25 19:33:19 Address 1: 195.130.129.164 Apr 25 19:33:29 you mean the ones that were in the default file? Apr 25 19:33:31 DocScrutinizer: I'll refresh Apr 25 19:33:50 check control63 Apr 25 19:33:51 these servers are the ones from my ISP and are working on my laptop (I copied them from this machine) Apr 25 19:33:56 m8ram: yep Apr 25 19:34:33 m8ram: have you configured your laptop to provide internet connection to neo? Apr 25 19:34:41 cat /etc/resolv.conf Apr 25 19:34:42 nameserver 208.67.222.222 Apr 25 19:34:42 nameserver 208.67.220.220 Apr 25 19:34:42 nameserver 208.67.222.222 Apr 25 19:34:42 nameserver 208.67.220.220 Apr 25 19:34:48 nslookup www.google.com Apr 25 19:34:48 Server: 208.67.222.222 Apr 25 19:34:48 Address 1: 208.67.222.222 Apr 25 19:34:57 hehe, see shr-trac Apr 25 19:35:09 DocScrutinizer: will do Apr 25 19:35:27 mrmoku: are you building libframeworkd-phonegui-efl2 now? Apr 25 19:35:32 on buildhost? Apr 25 19:35:48 can do so... yeah. have to build xchat anyway Apr 25 19:36:16 dos1: ah the IP forwarding? Apr 25 19:36:31 alphaone: a hw cause might be leak current in bigC reducing micbias voltage Apr 25 19:37:00 DocScrutinizer: leak current where? Apr 25 19:37:01 maybe cooked till it's weldone? Apr 25 19:37:04 m8ram: yep Apr 25 19:37:17 alphaone: in big C Apr 25 19:37:22 hmm Apr 25 19:37:25 SHR: 03mok 07shr-overlay * r239a75006a41 10/openembedded/recipes/tasks/task-shr-feed.bb: feed: add xchat Apr 25 19:37:50 dos1: don't think I have, thanks will see if I can figure that bit out Apr 25 19:39:30 mrmoku: please tell me when it's done :) Apr 25 19:40:09 yeah... started it now Apr 25 19:42:51 alphaone: check voltage of 100uF when gsmhandset restored Apr 25 19:43:12 alphaone: compare to known good device Apr 25 19:43:54 dos1: just checking before I try this, if I understand correctly all I need is "iptables -A POSTROUTING -t nat -j MASQUERADE -s 192.168.0.0/24" and "echo 1 > /proc/sys/net/ipv4/ip_forward", right? Apr 25 19:44:13 is that the same no matter how this laptop is connected to the internet? Apr 25 19:44:37 in my case the laptop connects to my router, not directly to the net Apr 25 19:44:59 m8ram: no problem Apr 25 19:45:00 DocScrutinizer: Good idea, will do Apr 25 19:45:26 DocScrutinizer: thx Apr 25 19:47:24 dos1: what should address:// url look like in opimd? Apr 25 19:47:31 doesn't seem to help Apr 25 19:47:51 still unable to resolve www.google.com from the FreeRunner Apr 25 19:48:11 first try to ping some IP Apr 25 19:48:44 Q-Master: i don't know... i think that urls aren't so good Apr 25 19:49:26 but if you want to use it, simply do "addres://dos1's street, London" :P Apr 25 19:49:27 dos1: I just thinking now about how to map vcard to opim Apr 25 19:49:46 ping 208.67.222.222 works Apr 25 19:50:33 dos1: does not build :( due to eina-native Apr 25 19:50:36 so it's a resolv.conf issue it seems Apr 25 19:50:39 dos1: also I don't really know what to do with photos Apr 25 19:51:20 m8ram: on FR :-) Apr 25 19:52:08 could it be that DNS port is blocked? that's 53 TCP/UDP right? Apr 25 19:52:17 any special rule I need for that as well? Apr 25 19:52:32 Q-Master: don't do it at the moment... Apr 25 19:52:45 Q-Master: we need to write some specification of opimd Apr 25 19:52:47 or is the forward rule sufficient? (My iptables knowledge is a bit rusty...) Apr 25 19:53:26 dos1: yep. just figuring out what should i do and how Apr 25 19:55:19 mrmoku: i have libframework-phonegui-efl builded on my buildhost today. i will try with efl2. if it won't work, then can you send a binary to me? Apr 25 19:55:53 mrmoku: ERROR: SRCREV was used yet no valid SCM was found in SRC_URI Apr 25 19:56:04 dos1: yeah... wait one sec :P Apr 25 19:56:11 in libframeworkd-phonegui-efl2_git.bb Apr 25 19:56:12 ok Apr 25 19:56:21 m8ram: seems you got the right ideas to fix the issue. I can't help, sorry. Apr 25 19:56:37 SHR: 03mok 07shr-overlay * r70d0c49a09a3 10/openembedded/recipes/shr/libframeworkd-phonegui-efl2_git.bb: libframeworkd-phonegui-efl2: fix SRCREV Apr 25 19:56:59 ok, thanks for putting me in the right direction Apr 25 19:57:02 * mrmoku has problems with using git:// as url when actually protocol is http:// :P Apr 25 19:57:26 m8ram: welcome. tho I didn't actually do anything ;-) Apr 25 19:58:56 DocScrutinizer: something else, untill I fix this can I download the packages somehow on my laptop and then install them locally? Apr 25 19:59:12 just so I can at least test the FreeRunner a bit Apr 25 19:59:52 m8ram: sorry that's totally different domain ;-) hw department here Apr 25 20:00:13 I'm sure someone else here can help Apr 25 20:00:21 m8ram, you can install downloaded packages with opkg, if you mean that: opkg install file.ipk Apr 25 20:00:55 DocScrutinizer: indeed, someone does! :) Apr 25 20:00:59 Sharwin_F thanks Apr 25 20:01:14 m8ram, similar as dpkg -i file.deb Apr 25 20:01:20 so ou'll have to provide all the dependencies manually Apr 25 20:01:20 so I can download ipk packages from opkg.org and then install them locally Apr 25 20:01:22 great Apr 25 20:01:35 yes Apr 25 20:04:39 Sharwin_F: and that's where the fun starts off course... :) Apr 25 20:05:07 m8ram, yes, the problem mught be dependencies hehe Apr 25 20:05:12 might* Apr 25 20:05:46 indeed, that's why we have package managers isn't it? Apr 25 20:06:00 ok, guess I'll have to freshen up my iptables knowledge... Apr 25 20:07:30 arghh, xchat fails to compile with OE on my computer Apr 25 20:08:15 is .opk also a valid extension? Apr 25 20:08:50 pyefl-sudoku comes as a "sudoku_r49.opk" file... Apr 25 20:08:54 m8ram: I'd clean up resolv.conf ( echo -n >resolv.conf) then reboot FR. maybe something with resolver lib or network borked Apr 25 20:09:06 dos1: the types of fields, posted in opimd docs is rather strange IMO Apr 25 20:09:14 though according to http://projects.openmoko.org/frs/?group_id=157 it's an ipk file Apr 25 20:09:23 m8ram, yep, opkg works with .ipk and .opk Apr 25 20:09:28 DocScrutinizer: I'll give that a try thanks Apr 25 20:09:41 Sharwin_F: In Linux, extensions don't really matter. Only with badly written programs ;) Apr 25 20:10:52 beniwtv, well, they matter if you put the wrong one on a file, because then you try opening a file with a program which doesn't know what to to with it xD Apr 25 20:11:02 m8ram: I don't think iptables is bloccking port53 specifically Apr 25 20:12:19 xchat fail log, if someone is bored: http://pastebin.com/m551d2b1 Apr 25 20:12:21 Sharwin_F: so linux doesn't care, but user does XD Apr 25 20:12:25 beniwtv: I know, but extensions do often "indicate" a type of file, and I wasn't certain .opk was the same as .ipk Apr 25 20:12:27 that is Apr 25 20:12:33 hehe Apr 25 20:12:42 that's the matter Apr 25 20:13:12 the package installed, but it immediately fails when I try to run it: An exit code of 1 was returned from sudoku_main.py. Apr 25 20:13:16 Sharwin_F: Any good program would refuse that file then. And your file manages shouldn't open a file by extension, but by type. I never got that nonsense with extensions in Windows. It even hides them, and from that come all those nasty viruses, which are really .exe's but have .zip.exe. Gah :-/ Apr 25 20:13:24 but that's a different matter... Apr 25 20:14:19 beniwtv, yes, but I don't think you get any succes passing a image.opk as paramter to opkg, if the opk file is really an image... :P Apr 25 20:14:32 Sharwin_F: :P Apr 25 20:14:35 so extensions matter for he user Apr 25 20:16:12 dos1: after cleaning eina-native it looks like it is building now... will take some time though Apr 25 20:16:15 building complete e Apr 25 20:17:09 mrmoku, why did you success with xchat and I get an error? did you have to solve them? Apr 25 20:17:33 Sharwin_F: you get an error while building? Apr 25 20:17:39 mrmoku, Apr 25 20:17:40 yes Apr 25 20:17:50 mrmoku, log -> http://pastebin.com/m551d2b1 Apr 25 20:18:09 I've never been able to build anything, so I suppouse I'm doing something wrong... xD Apr 25 20:18:40 hmm... missing ssl Apr 25 20:18:43 Sharwin_F: Seems you miss openssl Apr 25 20:19:00 beniwtv, but it should be installed by bitbake, shouldn't it? Apr 25 20:19:17 not if xchat is missing it as dependency Apr 25 20:19:31 then it would build on the buildhost (because ssl is build for other stuff already) Apr 25 20:19:38 and would fail for you me thinks Apr 25 20:19:56 hmm ok Apr 25 20:20:05 bitbake ouput -> http://pastebin.com/m62fb0750 Apr 25 20:20:09 I'll try to build it then Apr 25 20:20:37 so it's a problem of the .bb file? Apr 25 20:21:00 would be then... Apr 25 20:21:02 Sharwin_F: Could be Apr 25 20:23:44 damn: after shutting down the device it doesn't boot anymore Apr 25 20:24:06 it stops at "INIT: No inittab file found" Apr 25 20:24:12 Enter runlevel: Apr 25 20:24:47 anybody know how to enter anything during the boot stage? Apr 25 20:24:57 xchat is a lot nicer than pidgin :p Apr 25 20:26:24 and any idea how the inittab got lost? Apr 25 20:26:37 nah, strange Apr 25 20:26:38 could it be because of the package I installed manually? Apr 25 20:26:48 m8ram, which package? Apr 25 20:26:59 mrmobil: for irc, not for jabber or icq 8) Apr 25 20:27:01 the sudoku_r49.opk Apr 25 20:27:24 m8ram, don't think so Apr 25 20:27:30 mrmobil: The only issue I have with it (at least @ work on OSX) that it doesn't tab-complete user names very well. So for me it's pidgin. Oh, and +1 to what Q-Master said Apr 25 20:27:34 only interested in irc Apr 25 20:27:39 it's the only one I could install, moko-sudoku_0.21_armv4t.ipk complained about missing libmokoui2-0 Apr 25 20:28:01 xchat for irc is good Apr 25 20:28:07 anyway, is there a way to proceed? Apr 25 20:28:16 beniwtv, no works here Apr 25 20:28:26 or do I have to power down, boot another distro and somehow create/restore inittab? Apr 25 20:28:40 mrmobil: on OSX? Apr 25 20:28:59 m8ram: If you don't have a usb keyboard I would guess not Apr 25 20:29:20 beniwtv, on shr ;) Apr 25 20:29:35 (since I assume ssh hasn't started without init) Apr 25 20:30:03 not likely indeed Apr 25 20:30:12 mrmobil: yeah, SHR is better for the FR than OSX ;) Apr 25 20:30:24 and no usb keyboard available, at least not with a small usb connector Apr 25 20:31:02 and I don't suppose a usb connecter with a large and a small usb plug on a regular keayboard would work would it? Apr 25 20:32:24 beniwtv, indeed :) Apr 25 20:33:08 m8ram: I can't see why not. The USB plugs are the same, only the mini usb is smaller. So If you have a USB<->mini-usb plug it should work. Apr 25 20:34:46 I doubt that will pan out. USB is in device mode per default I'd guess Apr 25 20:35:20 m8ram: ah, yeah, forgot about the usb mode Apr 25 20:35:30 Sharwin_F, building xchat on buildhost now Apr 25 20:35:36 scrap that plan: the keyboards I have available are all still PS2... Apr 25 20:35:47 only my new wireless keyboard/mouse combo is usb Apr 25 20:35:49 errr... s/m8ram/DocScrutinizer/ Apr 25 20:36:04 never mind, I'll reboot to FDOM then Apr 25 20:36:26 m8ram: probably there are exactly 2 things you can possibly do Apr 25 20:36:33 reflash Apr 25 20:36:40 or boot from uSD Apr 25 20:36:48 uSD? Apr 25 20:37:19 see wiki Apr 25 20:37:34 e.g gsm/flashing Apr 25 20:37:35 do you mean an SD card? Apr 25 20:37:49 yup Apr 25 20:37:53 because this installation of SHR is on an SD card Apr 25 20:38:06 the device's NAND still runs FDOM Apr 25 20:38:18 mrmobil, I'm building openssl right now :P Apr 25 20:38:31 oh, so it's easy, isn't it? Apr 25 20:38:41 because that came with the device I haven't touched it yet because it has proven the only distro so far I could reliably SSH into... Apr 25 20:38:49 insert card to PC and fix the inittab issue Apr 25 20:38:55 dos1: building package-index now :-) Apr 25 20:39:08 listening to music and some other apps open while building, an my computer is not very new.. so it builds slow :( Apr 25 20:39:32 hopefully that's the only issue... Apr 25 20:40:04 dos1: done Apr 25 20:40:08 freesmartphone.org: 03seba.dos1 07framework * ra100a6b73023 10/framework/subsystems/opimd/backend_manager.py: opimd: don't fail when default backend isn't specified. Use some, which supports specified domain. Apr 25 20:40:44 m8ram: still I wonder how inittab can get messed/deleted Apr 25 20:41:31 it hasn't, just ssh'd into FDOM, mounted opkg install and there's a seemingly valid inittab in etc Apr 25 20:41:36 strange... Apr 25 20:41:56 after all the first time I booted this distro after installing it everything went fine... Apr 25 20:42:25 and all I did so far was to set date, link /etc/localtime and install that sudoku package Apr 25 20:43:29 oh, and I messed around with the resolv.conf Apr 25 20:43:32 dos1: is there a possibility in python to search in lists? Apr 25 20:43:37 easy way Apr 25 20:44:08 Q-Master: "string" in [list] ? Apr 25 20:44:15 yep Apr 25 20:44:35 mrmoku, argh, I get an error bitbaking openssl too Apr 25 20:44:48 Q-Master: the above returns True when "string" is an element of the [list] :) Apr 25 20:45:07 Toaster`: I've understood already. thanks Apr 25 20:45:11 np Apr 25 20:45:13 forgot python a bit. 8) Apr 25 20:45:28 Sharwin_F: hmm... guess something is broken in your setup Apr 25 20:46:10 anybody an idea as to why I'd get this error while inttab is available? Apr 25 20:46:14 mrmobil, bitbake oputput -> http://pastebin.com/m21587fd8 Apr 25 20:47:18 Sharwin_F: missing ncurses? Apr 25 20:47:44 m8ram: wrong bootparams? Apr 25 20:47:58 where are those stored? Apr 25 20:48:01 in /boot? Apr 25 20:48:13 or in the NAND? Apr 25 20:48:30 nah in uBoot environment (some nand partition iirc) Apr 25 20:48:43 OK, and how do I check those? Apr 25 20:48:45 mrmoku, I get exactly the same output when doing bitbake ncurses Apr 25 20:48:52 do I need dfu-util for that? Apr 25 20:49:11 or do I need to update U-Boot as well? Apr 25 20:49:23 U-Boot is not really related a distro is it? Apr 25 20:49:38 m8ram: uboot cli, see wiki uboot Apr 25 20:49:46 m8ram: you can tweak u-boot environment both from user-space and from usb console (while u-boot is working on device). Apr 25 20:49:52 because I haven't updated that today when I installed this distro Apr 25 20:50:09 is everyone having the "full battery reports 33% in battery widget" problem (with recent kernels), or is it just me? i wonder if it's to do with my screwing around with the kernel. if everyone has the same problem, i'll stop worrying... Apr 25 20:51:06 never seen. which distro Apr 25 20:51:15 DocScrutinizer, PaulFertser: thanks, I'll check that later, have to go now Apr 25 20:51:48 DocScrutinizer: SHR-testing, but i had it with "old SHR-testing" as well.. Apr 25 20:52:13 it didn't happen with the freshly flashed image, but when i installed my own kernel it went wrong, and i can't put it back Apr 25 20:52:19 nah, maybe your bat needs some calibrating? Apr 25 20:52:24 Wiss: some people have this problem, some not Apr 25 20:52:26 dos1: I think that the list of the fields in opimd doc should be rewritten. It is too far away from vcard3 or even vcard2 spec Apr 25 20:52:38 Q-Master: i think that too ;) Apr 25 20:52:43 hmm, ok then. dunno Apr 25 20:52:48 DocScrutinizer: it was definitely fine before i flashed my new kernel.. Apr 25 20:52:59 Q-Master: i'm opimd developer... since today, cause I decided to be :D Apr 25 20:53:00 dos1: the main question is "When?" 8) Apr 25 20:53:06 bye and thanks again for the help Apr 25 20:53:20 Q-Master: and i don't know it so well Apr 25 20:53:28 * Q-Master is playing with vcards 8) Apr 25 20:53:29 dos1, installed efl2? Apr 25 20:54:01 dos1: seems that there is no hardcoded fields there. Apr 25 20:54:20 Weiss: any chance you can check contents of /sys/class/power_supply/battery ? Apr 25 20:54:23 Weiss: ;) Apr 25 20:54:45 dos1: except Name and Phone in sim backend Apr 25 20:55:14 long time no see BluesLee Apr 25 20:55:36 PaulFertser: anything in particular? Apr 25 20:56:11 Weiss: actually my widget and settings report identical values, but icon doesn't change accordingly. dunno if it's supposed to at all Apr 25 20:56:32 Weiss: capacity, i guess. Apr 25 20:56:48 PaulFertser: it says '100' Apr 25 20:56:56 DocScrutinizer: interestingly, "apm" always reports the right value Apr 25 20:57:04 odd... my FR isn't accepting my ssh keys anymore... only passwords Apr 25 20:57:51 raster: ping Apr 25 20:58:10 mrmobil: yes, testing it now Apr 25 20:58:25 dos1: ok :-) Apr 25 20:58:49 quatrox: bad time for austalia Apr 25 20:58:56 ahh Apr 25 20:59:04 dos1: http://en.wikipedia.org/wiki/VCard Apr 25 20:59:46 Weiss: probably that widget is using hal, so lshal can help to nail down the problem... Apr 25 21:00:12 mrmoku: PIN wrong Apr 25 21:00:31 PIN wrong? Apr 25 21:01:05 mrmoku: yep. it works with efl Apr 25 21:01:21 dos1: you mean the SIM PIN? Apr 25 21:01:22 but with efl2 it always says "PIN wrong" after entering pin Apr 25 21:01:24 yep Apr 25 21:01:38 hmm... strange Apr 25 21:01:46 I have no PIN... so it might as well be Apr 25 21:01:52 but I did not change anything there Apr 25 21:02:46 dos1: utf8 ? Apr 25 21:04:01 mrmobil? Apr 25 21:04:13 DocScrutinizer: what utf8? opimd? Apr 25 21:04:32 PIN wrong Apr 25 21:04:45 Weiss: there's a setting in E concerning battery. Wrench->Advanced->Battery meter->Advanced->Hardware. I have "Auto Detect" there. Apr 25 21:04:52 a wild guess Apr 25 21:05:11 PaulFertser: heh Apr 25 21:05:14 DocScrutinizer: PIN is numerical... Apr 25 21:05:19 mrmoku: about utf8 in opimd Apr 25 21:05:22 it's only sim backend Apr 25 21:05:26 name = name.encode('ascii', 'ignore') Apr 25 21:05:45 mrmoku: line 83, pimb_sim_contacts_fso.py Apr 25 21:05:57 can you test it without that line? Apr 25 21:06:10 mrmoku: are you mrmobil? Apr 25 21:06:20 mrmoku: openmoko-messages crashes with efl2 :( Apr 25 21:06:21 it worked Apr 25 21:06:29 but then i added new contact Apr 25 21:06:31 and it crashed Apr 25 21:06:46 mrmoku: s/messages/contacts/ Apr 25 21:08:10 btw I killed quite some SIMs by entering the wrong pin and then hitting enter repeatedly because of delay in feedback Apr 25 21:08:23 not on SHR tho Apr 25 21:13:12 PaulFertser: me too. switching it to HAL keeps the problem, but switching to "Internal" fixes it Apr 25 21:13:44 duh Apr 25 21:13:53 auto here as well Apr 25 21:14:00 Weiss: well, luckily i don't run hal at all :) Apr 25 21:14:16 Weiss: and you do as you're using xorg obviously :) Apr 25 21:14:24 hmm, maybe that's the difference Apr 25 21:14:27 mrmoku: oh, i know Apr 25 21:14:29 Weiss: i guess most other users don't, that's why they don't see the problem. Apr 25 21:14:39 i was setting "Number" instead of "Phone" :) Apr 25 21:14:52 evenin Apr 25 21:15:08 SHR: 03mwester 07shr-overlay * r2c37bbbd8d9f 10/patches/efl1/0002-temporary-fixup-of-e-renaming.patch: SHR patches: add a patch to ensure that the ecore* libs are correctly packaged; Apr 25 21:15:16 PaulFertser: correct, i'm running Xorg.. but i have it configured without HAL. Apr 25 21:15:30 and.. grr... it switches back to Auto after a few seconds :/ Apr 25 21:15:32 mwester: \o/ Apr 25 21:15:39 Hacky Apr 25 21:16:08 Weiss: wtf! Apr 25 21:16:23 Maybe nytowl|weekend can look at it and see if it applies to fso/milestone5.5 -- if not, maybe he can tell us what we are doing wrong that the packages are not done correctly here. :) Apr 25 21:16:32 Weiss: do you have hal installed? Do you have lshal command? Apr 25 21:17:16 etc/initd/hal staus? Apr 25 21:17:27 PaulFertser: yes.. what i mean is that i configured Xorg not to use HAL for input devices (probably unrelated) Apr 25 21:17:28 status even Apr 25 21:18:27 hmm, confusing - i have lshal but no /etc/init.d/hal Apr 25 21:18:32 cool, paul solved another mystery Apr 25 21:18:46 Weiss: there was some hal-related problem reported, it was related to sysfs not having some nodes commonly found on laptops iirc. But nobody was pushing it, so it's left unnoticed. Can you provide lshal output? I'll try to find the ticket now. Apr 25 21:18:57 maybe its called differently Apr 25 21:18:58 DocScrutinizer: if all mysteries were that easy... Apr 25 21:19:02 PaulFertser: coming up.. Apr 25 21:19:29 DocScrutinizer: and i knew the answer from the beginning (due to that bugreport)... Apr 25 21:19:47 cheater ;) Apr 25 21:22:18 well the shitty widget icon at least is showing a 100graphic at 68%, and a 2/3graphic at 40% :-( Apr 25 21:22:38 http://pastebin.com/d427e7b09 (lshal -t), and http://www.bitwiz.org.uk/hal.txt (the full whack) Apr 25 21:23:16 should gsmd still work with newer omoko kernel pathes? i do everything as instructed in openmoko gsmd wiki-entry but keep getting "Modem power-up failed: 262145" when i start to power-up the antenna with libgsmd-tool Apr 25 21:23:24 and when I ever manage to hit it all to the left border, percentage comes up white-outline on white. eye cancer Apr 25 21:24:04 doxy2: gsmd? If not updated it will fail to power up the modem itself, i guess. Why do you ask, are you a H:1 dev? Apr 25 21:24:51 PaulFertser: i have own kernel and "distro" and just tryin to get things to work Apr 25 21:25:11 checked out the source of gsmd from svb Apr 25 21:25:12 svn Apr 25 21:25:40 doxy2: ok, feel free to ask me any modern kernel-related questions. Apr 25 21:25:52 * DocScrutinizer thought gsmd being deprecated like old fish Apr 25 21:26:12 maybe I'm totally wrong Apr 25 21:26:18 getting tired Apr 25 21:26:28 yeah, so is gsmd so deprecated it doesn't even work with newer kernel updates? Apr 25 21:27:36 well rumor is /sys seen changes Apr 25 21:27:37 doxy2: kernel updates made some slight changes to the sysfs paths wrt gsm modem. Apr 25 21:28:04 doxy2: you should ask H:1 folks if they made it work with any new kernel. Or fix paths yourself. Apr 25 21:28:24 PaulFertser: ummm...H:1 :) ? Apr 25 21:28:57 hackable:1, the necrophiles ;-) Apr 25 21:29:12 heh Apr 25 21:29:46 well, i don't necessarily have to use gsmd, whats the modern one? Apr 25 21:29:59 fso Apr 25 21:30:35 mrmoku: ping Apr 25 21:30:44 ** (process:3051): DEBUG: launched async query for a single contact (callback) ... Apr 25 21:30:44 ** (process:3051): DEBUG: name callbacking... Apr 25 21:30:44 *** glibc detected *** openmoko-messages: free(): invalid pointer: 0x402fa46c *** Apr 25 21:30:59 * DocScrutinizer is spanking himself to shutup Apr 25 21:31:00 Hey SHR doods --- is there really only one theme anymore? Apr 25 21:31:00 doxy2: and if you want to be on the front line, cornucopia (FSO rewrite in Vala), isn't really used anywhere yet. Apr 25 21:31:39 mwester: seems Apr 25 21:31:44 :-/ Apr 25 21:31:54 Ok. Then we're missing some icons. Apr 25 21:32:17 Contacts, Mofi, Vagalume all are blank. Apr 25 21:32:31 not here Apr 25 21:33:18 shr-unstable, past few days. Apr 25 21:33:37 Ever since the shr-illume theme went away, basically. Apr 25 21:33:39 dos1: pong Apr 25 21:33:44 yeah, you still use unstable ;-D Apr 25 21:33:52 But the SHR guys are ignoring me. Apr 25 21:33:57 yep... don't want you to view messages Apr 25 21:34:16 mwester: I have all those icons Apr 25 21:34:21 mrmoku: contacts works well. it's only segfaulting on adding new contact, but it is added and displayed Apr 25 21:34:28 mrmoku: editing and closing doesn't work, as expected Apr 25 21:34:31 (well vagalume don't know) Apr 25 21:35:01 mrmoku: btw, i've connected today to open network by wifiman.py Apr 25 21:35:10 vagalume ok Apr 25 21:35:39 I think one of the smart SHR guys (dos1, mrmoku, Ainulindale) needs to do a clean build of unstable, and flash that, so that we can get unstable working correctly again. It's rather a pain this way. Apr 25 21:35:52 dos1: Are there any non-SHR standard dependancies for wifiman.py? Apr 25 21:36:20 PaulFertser,DocScrutinizer: ok, thnx for the info! Apr 25 21:36:21 Toaster`: python-elementary, python-edbus, connman... nothing else i think Apr 25 21:36:28 hrm, ok Apr 25 21:36:30 Ok, for anyone interested here's the source of E battery monitor (including hal calls): https://svn.enlightenment.org/svn/e/trunk/e/src/modules/battery/e_mod_main.c Combined with the full hal output Weiss provided it can show why he had 38% all the time. I'll try to read it some more. Apr 25 21:36:34 maybe I had an old version Apr 25 21:36:44 doxy2: you're welcome. I suggest you to stick to FSO for now. Apr 25 21:37:00 mwester: agree, though my son has birthday tomorrow... want get around to do it before tomorrow evening Apr 25 21:37:04 hope Ainulindale is faster :P Apr 25 21:37:42 mrmoku: "rm tmp; cd ..; make update-shr-unstable; make shr-unstable" Then go off and enjoy your party. Apr 25 21:37:44 mwester: by clean build you mean build from scratch? Apr 25 21:37:47 ahh, ok Apr 25 21:37:53 mwester: i thought it was fixed, as i read somewhere on maillist... Apr 25 21:38:13 mwester: ok, will launch a local build of unstable on my laptop overnight Apr 25 21:38:23 I don't even know which theme we are supposed to use in the first place. Apr 25 21:38:31 mwester: Illume SHR Apr 25 21:38:34 Or why it offered me two before, and one now.. Apr 25 21:38:35 e-wm-theme-illume-shr Apr 25 21:38:44 dos1: what is a correct git url for opimd? Apr 25 21:38:58 Q-Master: the same as for frameworkd, as it's frameworkd part Apr 25 21:39:08 and? Apr 25 21:39:13 * mwester is not a GUI person. :( Apr 25 21:39:37 Q-Master: http://git.freesmartphone.org/?p=framework.git;a=tree;f=framework/subsystems/opimd;h=3833f88723670af2940bfa06ba56e375827ceb1a;hb=a100a6b73023abb4ed385cf0e3615c1f303f6cc4 Apr 25 21:39:49 * mwester thinks maybe a good kernel, and emacs on the phone is the way to go. Apr 25 21:39:51 thnx Apr 25 21:40:05 Hal thinks that USB power supply is a battery :-/ Apr 25 21:40:11 mwester: Ainulindale has moved e-wm-theme-illume-shr from shr.git to shr-themes.git Apr 25 21:40:23 mwester: emacsphone Apr 25 21:40:24 maybe that went wrong :x Apr 25 21:40:28 mwester: write libframework-phonegui-emacs then :P Apr 25 21:40:40 mwester: special edition of gnuphone Apr 25 21:41:08 Ainulindale was working on libframeworkd-phonegui-ncurses ;) Apr 25 21:41:09 PaulFertser: fsck hal Apr 25 21:41:13 Only problem with an emacsphone would be that it would need a full-size keyboard (meta-x-ctrl-pctrl-x-ctrl-d to dial) Apr 25 21:41:33 PaulFertser: hmm, i think raster mentioned something along those lines a few weeks ago.. Apr 25 21:42:14 Weiss: so E sees 2 batteries and 38 might be some result of computation. I need to find out why hal thinks usb is battery. Apr 25 21:42:40 mrmoku: which backend i should set as default in SHR? CVS or SQLite? Apr 25 21:42:54 PaulFertser: 33% is my value when full.. Apr 25 21:43:03 Weiss: ok, 33% :) Apr 25 21:43:20 sounds like 1/3 Apr 25 21:43:25 yeah Apr 25 21:43:31 PaulFertser: vague recollection.... i think the usb being a bat may have something to do with a request by Android folks. Check kernel mailing list archives. Apr 25 21:43:48 one full and two empty bats? Apr 25 21:43:59 dos1: sqlite Apr 25 21:44:05 There are three possible power sources. Apr 25 21:44:15 bat, ac(charger), usb Apr 25 21:44:24 waaaaaaaaaaaaah. FSCK android!! Apr 25 21:44:47 mwester: that's it Apr 25 21:45:05 Q-Master: sqlite is great for messages Apr 25 21:45:13 but for contacts... hmm Apr 25 21:45:18 i'll do it tomorrow Apr 25 21:45:19 Weiss: btw, it sees 3 batteries. One apm battery, one /sys/ battery, one usb :) Apr 25 21:45:40 hal uses apm bats? Apr 25 21:45:46 one full and two empty bats? Apr 25 21:45:47 dos1: also for contacts. I hate new phones storing contacts only in SIM Apr 25 21:45:50 that suxx Apr 25 21:46:11 Q-Master: i didn't mean SIM. I mean CVS :P Apr 25 21:46:11 mwester: yeah, i see apm in hal listing. And if apm emulation was disabled, it wouldn't be a problem. Apr 25 21:46:21 That's what I was thining. Apr 25 21:46:23 cvs also sux. 8) Apr 25 21:46:26 er, thinking. Apr 25 21:46:45 Q-Master: i need to test custom fields with sqlite backend. Apr 25 21:46:52 so i will do it tomorrow Apr 25 21:46:57 ok Apr 25 21:46:58 now i have to go Apr 25 21:47:02 good night :) Apr 25 21:47:08 hmm Apr 25 21:47:12 ~seen Dave Apr 25 21:47:13 dave was last seen on IRC in channel #openmoko-cdevel, 9d 8h 26m 18s ago, saying: 'dos1, I thought opimd was just a legend!? :o'. Apr 25 21:47:13 try to think of a vcard field names Apr 25 21:47:23 no, Dave, opimd is not legend ;) Apr 25 21:47:28 xD Apr 25 21:47:29 or compatible Apr 25 21:47:34 dos, how long does it usually take for commits to the SHR-Settings git to be included? Apr 25 21:47:48 Toaster`: until someone regenerate package ;) Apr 25 21:48:05 Toaster`: bug mrmoku or Ainulindale for doing that Apr 25 21:48:11 but now i really have to go Apr 25 21:48:11 Unstable is broken until somebody (hint hint!!!!!) does a full rebuild. Apr 25 21:48:18 cu :) Apr 25 21:48:31 * DocScrutinizer worries about where Dave left Apr 25 21:48:35 heh, ok, so it's not an automatic build? Apr 25 21:48:40 Nope Apr 25 21:48:48 cool. Apr 25 21:49:08 An autobuilder would be nice. Apr 25 21:49:26 * mwester wonders what the nslu2-linux autobuilder has been up to lately. Apr 25 21:51:04 Ok, http://cgit.freedesktop.org/hal/tree/hald/linux/device.c?id=HEAD clearly shows that HAL considers USB to be a battery. Apr 25 21:56:37 * mrmoku starts a build from scratch Apr 25 21:57:29 shr-testing update problem [e-wm-menu-shr/task-shr-minimal-x] Apr 25 21:57:32 http://pastebin.ca/1403122 Apr 25 21:57:38 is it known bug? Apr 25 21:59:10 max_posedon: try opkg -force-depends remove e-wm-menu; opkg install e-wm-menu-shr; opkg upgrade Apr 25 22:00:13 I asked a question in the #hal channel. Let's see if it gives us anything... Apr 25 22:00:29 PaulFertser: looking through the kernel archives, and can't find anything obvious. i'm curious as to why it seemed to appear only after i updated my kernel... Apr 25 22:01:09 mrmoku, thanks, first part is fixed Apr 25 22:01:16 :) Apr 25 22:01:17 what about missing glibc-binary-localedata-en-gb? Apr 25 22:01:59 Weiss: probably you had a kernel without APM emulation or hadn't HAL installed or smth like that? Apr 25 22:02:57 max_posedon: hmm... there was something broken with locales... don't know Apr 25 22:03:22 PaulFertser: ah, hang on, you're right. i installed Xorg at the same time as upgrading the kernel, and it pulled in some HAL bits.. (even though it shouldn't need them) Apr 25 22:03:36 max_posedon: you could try to -force-depends remove that one also Apr 25 22:04:54 mwester: error: patch failed: recipes/efl1/evas_svn.bb:1 Apr 25 22:04:59 you have that one too? Apr 25 22:05:19 Weiss: btw, it's only apm battery that contributes to your 33% level. power_supply battery is considered to be empty because it has battery.charge_level.last_full = 1072 but battery.charge_level.current = 0 and 0 diveded by anything is 0. Apr 25 22:05:43 mrmoku: I fixed that. Apr 25 22:06:11 hmmm Apr 25 22:06:18 how many space is needed on TMPDIR to build shr-unstable aprox? Apr 25 22:06:19 I restarted completely from scratch Apr 25 22:06:20 mrmoku: probably your git pull from OE died with that fatal error Apr 25 22:07:08 mwester: ok... make update fixed it :-) Apr 25 22:07:34 * mrmoku has to find out some way to break visible when OE fails Apr 25 22:07:38 Sheesh. I guess OE can't fix that problem with their server, so we probably need ot fix the makefile to die if that error happens. Apr 25 22:08:13 mrmoku: I'm having the same issue Apr 25 22:08:19 rwhitby would know how to do that in an instant. Apr 25 22:08:49 Toaster`: do the make update-whatever thing again; watch the output carefully to make sure that you do not get any text with the word "fatal" in it. Apr 25 22:08:58 * mrmoku checks the fso makefile Apr 25 22:10:09 I guess we should fix the kernel driver to behave more consistently with regular laptops, e.g. we lack design capacity and some other properties. Apr 25 22:10:14 mwester: hmm... fso makefile does update the same way... Apr 25 22:10:29 Yeah. So does the SlugOS makefile. Apr 25 22:10:36 And it fails constantly too. Apr 25 22:10:42 (I haven't fixed that either). Apr 25 22:12:17 oh my, lost 2 days of shr-user posts *_* Apr 25 22:12:49 mrmoku, with -force-depends issues was fixed. thanks Apr 25 22:12:59 great :-) Apr 25 22:13:11 btw, why shr haven't many intersted packages in feeds? Apr 25 22:13:31 for me it looks like OE have more, that shr Apr 25 22:13:58 because it takes a lot of effort to keep them all building. Apr 25 22:14:26 so, problem is that they doesn't work even in OE? Apr 25 22:14:48 Very often that is the case. Apr 25 22:15:09 PaulFertser: spot on! opkg remove (-force-depends) hal -> problem fixed (right channel this time :) Apr 25 22:15:11 ok, make shr-unstable-image running now Apr 25 22:15:24 OE builds many different distros; a recipes presence just means it builds on one of those, not necessarily all. Apr 25 22:16:24 Weiss: sure, but we need to co-exist nicely with HAL anyway. As now i know everything needed about why and what works and what doesn't i'll wait for some feedback from HAL folks (just posted a question at their chan, if it will go unanswered, i'll write to the ML tomorrow). Apr 25 22:17:12 PaulFertser: agreed - thanks for looking into it! Apr 25 22:17:44 Sharwin_F-om: how is xchat? :_) Apr 25 22:18:02 mrmoku, unusuable due to font size Apr 25 22:18:23 I changed it to Monospace 7 ... quite nice Apr 25 22:18:29 and the text log part doesn't show well sometimes Apr 25 22:19:23 mrmoku: my only problem is that i do not notice if somebody mentions my name Apr 25 22:19:23 go to have dinner, I'll try it later Apr 25 22:20:31 mrmoku: too manyy channels for them to fit to screen and xchat does not rearrange them based on activity or nythhhing Apr 25 22:20:32 lindi-: why that? Apr 25 22:20:55 (notice) Apr 25 22:21:31 Weiss: btw, can you tell us anything interesting about X-related stuff? Apr 25 22:21:40 lindi-: didn't you config a few trigger patterns? Apr 25 22:22:04 DocScrutinizer: how would the trigger? Apr 25 22:23:44 PaulFertser: what like? Apr 25 22:23:58 this weekend i'm working on EXA-via-DRI Apr 25 22:24:13 lindi-: preferences-alerts Apr 25 22:24:48 extra wors to highlight Apr 25 22:26:06 then sounds-keyword Apr 25 22:28:31 Weiss: like that interesting stuff (3D-acceleration afaict) you're working on :) Apr 25 22:30:53 i'm planning a bit of a hackathon once i've submitted my thesis, until then it's quite slow progress (especially since it seems to be just me) Apr 25 22:31:22 but i have all the basic initialisation and some of the memory management stuff working Apr 25 22:31:43 lindi-: :-) Apr 25 22:31:49 did not try more than one channel yet Apr 25 22:32:17 Weiss: do you have any particular goal in mind or do you just enjoy this kind of hacking? Apr 25 22:32:45 mwester: ok... build is running... will report back tomorrow Apr 25 22:32:53 bedtime now Apr 25 22:33:00 gnight Apr 25 22:33:16 Weiss: btw, thank you for being a sane accelerometer-using app dev :) Apr 25 22:33:52 PaulFertser: my main motivations are to get a unified access pathway to Glamo (seriously, accelerated X server + accelerated mplayer should make your Freerunner explode), and to play with 3D Apr 25 22:34:18 i see potential in UI stuff using the 3D engine, but i have raster's concerns in mind. it's still a stimulating project whatever the outcome Apr 25 22:34:32 hehe Apr 25 22:36:13 Weiss: is it possible you're interested in fixing one small bug in xorg driver? ;) Apr 25 22:36:34 the EXA debug message thing? Apr 25 22:36:50 Weiss: no, the QVGA misplaced image. Apr 25 22:37:15 hmm.. is there a ticket? Apr 25 22:37:40 Weiss: sure: http://docs.openmoko.org/trac/ticket//2263 Apr 25 22:38:55 Ainulindale: btw, what's the progress on including xorg? What are the difficulties? Can't you just include it in the build? Apr 25 22:40:10 hmm, i'll give that a try in a few minutes.. Apr 25 22:41:59 roh: Hi again :) Could you please give somebody (e.g. me) a permission to add Cc to the OM trac tickets? E.g. http://docs.openmoko.org/trac/ticket//2263 lacks Lars, the main driver developer :-/ Apr 25 22:42:44 interesting Xorg consideration: it appears to be quite difficult to properly hide the cursor Apr 25 22:44:34 roh: and btw OM trac is too closed imho. No way for reporter to close a ticket is somewhat too uncommon to be called sane imho :-/ Apr 25 22:45:36 the OM Trac doesn't even seem to want to let me add comments to tickets at the moment :( Apr 25 22:46:47 Weiss: you're supposed to be logged in first. Apr 25 22:47:17 i am Apr 25 22:47:40 it says "Warning: " at the top Apr 25 22:48:53 Weiss: :( Apr 25 22:49:55 Anyone know if there are any python FFT modules on the FR? Apr 25 22:49:58 PaulFertser: confirmed? Apr 25 22:50:45 DocScrutinizer: what? I haven't tried to add a comment because i have nothing to add anywhere. Apr 25 22:50:57 PaulFertser: ok, i see that problem happening in my setup (EXA disabled + drm-tracking kernel [different VRAM layout]) Apr 25 22:51:02 DocScrutinizer: but the fact OM trac is too closed is confirmed, yes. Apr 25 22:51:09 so I won't run to my PC now Apr 25 22:51:18 i don't even get a box to type a comment into... Apr 25 22:51:51 Weiss: and i do. Do you see your name as logged in? Apr 25 22:52:12 yep, top right Apr 25 22:53:08 Weiss: i just tried to log out and logged in again. Worked flawlessly and i see comment box in the bottom. Apr 25 22:54:10 hmm, it's been misbehaving for me for a long time Apr 25 22:56:57 Probably i'm guilty of using a JS-enabled firefox and that's the reason it works for me... Apr 25 23:17:56 PaulFertser: playing with the offset after doing xrandr. for me, the offset was large (about 1/3 of the height of the screen), but after turning the xglamo kernel hack off, it goes down to about 1px Apr 25 23:18:44 PaulFertser: also, do your illume icons bunch up a bit differently with xorg, as if a single pixel sliver of screen has been taken off the width of the screen? Apr 25 23:19:29 hmm, it's doing 32px now.. Apr 25 23:42:02 Hello, in latest andy-tracking I get segfault due to preemption during some ar6k function Apr 25 23:42:34 I don't have any technical means to provide more info than that, just letting you know preemption doesn't work yet :) Apr 26 00:11:54 When building the SHR-Unstable image, is it the SHR-lite version? Or completely bare? Apr 26 02:22:23 How long is a from-scratch SHR-Unstable image build supposed to take Apr 26 02:22:43 overnight is not unreasonable Apr 26 02:22:55 did you specifiy any parallism? Apr 26 02:23:03 so if it took ~3 hours I should see if something broke? Apr 26 02:23:03 parallelism? Apr 26 02:23:19 I can do about three hours on my quad-core Apr 26 02:23:35 ok, I set it to yous all 4 of my cores. Apr 26 02:23:44 s/yous/use/ Apr 26 02:23:44 Toaster` meant: ok, I set it to use all 4 of my cores. Apr 26 02:23:56 that makes sense Apr 26 02:23:58 Thanks Apr 26 02:24:03 There's a summary at the end; should say "0 errors" among other stuff. Apr 26 02:24:53 "0 failed" Looks good then Apr 26 02:25:09 Now, will it run? Apr 26 02:25:39 more like... Can Toaster` find the image? Apr 26 02:25:55 This is the first time I tried to build it Apr 26 02:26:11 tmp/deploy/glibc/images/om-gta02/* Apr 26 02:27:24 within the build directory? Apr 26 02:28:34 I'm blind Apr 26 02:28:36 nm Apr 26 02:31:23 Does the SHR makefile pull all the latest git commits? or is it a snapshot? Apr 26 02:39:06 unstable pulls in many many latest git and svn commits. testing less so. the release, when it happens, will be nailed down. Apr 26 02:39:16 prefect Apr 26 02:39:23 err.. perfect Apr 26 02:39:24 And, yes, that's a path inside the shr-unstable buld dir. Apr 26 02:39:54 Do a "find" for '*.jffs2' Apr 26 02:40:09 I've found it Apr 26 02:42:21 Now to see if it flashes and boots **** ENDING LOGGING AT Sun Apr 26 02:59:57 2009