**** BEGIN LOGGING AT Sun May 10 02:59:57 2009 May 10 05:57:07 DocScrutinizer: There's simply no way to create packed structures, period. That's one of the reasons developers must sort all the structure members to be aligned (every struct member should be on the address dividable by its size) manually. And it still doesn't guarantee anything really, just a good approximation that mostly works. May 10 05:59:57 And afaik ANSI doesn't say anything about alignment at all, so for strict binary compatibility a programmer should access struct members via explicit offsets (and _most_ architectures don't allow unaligned access or it results in a significant slowdown). May 10 08:40:17 SHR: 03seba.dos1 07shr-themes * r9b8ae86e2f98 10/e-wm/e-wm-config-illume-shr/e.src: e-wm-config-illume-shr: use syscon again May 10 08:55:00 "I'm working on CPU frequency scaling support. Slowing the CPU to 100 MHz and 1.1 V core power supply gives:" May 10 09:08:32 noone interested in cpu scaling support on gta02? :p May 10 09:10:06 I am May 10 09:10:16 but I also am in a hurry ;-) May 10 09:24:55 omg May 10 09:24:57 PaulFertser: #pragma pack May 10 09:25:13 have you seen new finger-scrolling in elementary (shr-unstable)? May 10 09:25:21 funny ;D May 10 09:25:49 DocScrutinizer: #pgragma pack works with gcc? May 10 09:25:58 dunno May 10 09:26:41 anyway it seems to be an ansi-c defined extension May 10 09:27:47 dos1: let me guess .... May 10 09:27:50 iirc there was an initiative to make gcc ansi-compliant last century May 10 09:27:51 39750 May 10 09:28:46 I remember something like __attribute__ May 10 09:29:59 Flyser_: no, 40567 May 10 09:30:02 found it http://gcc.gnu.org/onlinedocs/gcc-4.4.0/gcc/Variable-Attributes.html#Variable-Attributes May 10 09:30:23 Flyser_: it wasn't here in 40478 May 10 09:31:24 dos1: I found this: http://lists.openmoko.org/pipermail/community/2009-April/045179.html May 10 09:31:33 Why did it increase? May 10 09:32:25 Flyser_: i suppose some cpu usage... but why are you asking me? :x May 10 09:33:43 who else to ask? :p May 10 09:34:32 Flyser_: maybe aevin... he is first on list ;) May 10 09:35:23 Flyser_: i'm just FR user, which is programming something for SHR or FSO for time to time ;) May 10 09:35:28 s/which/who/ May 10 09:35:28 dos1 meant: Flyser_: i'm just FR user, who is programming something for SHR or FSO for time to time ;) May 10 09:35:31 ;x May 10 09:35:39 okay ^^ May 10 09:36:16 s/for/from/ May 10 09:36:24 or something like that May 10 09:36:27 ;) May 10 09:40:30 dos1: a little early ? May 10 09:40:31 :) May 10 09:40:56 ptitjes: what early? May 10 09:41:05 hour? May 10 09:41:08 yeha May 10 09:41:13 hehe May 10 09:41:33 hello all May 10 09:41:36 hello May 10 09:41:37 mickey|tv, hi, around May 10 09:41:40 ptitjes, yo May 10 09:41:48 i don't know why sometimes i wake up late, and sometimes i can't sleep anymore early ;) May 10 09:42:28 you're becomming old ;) May 10 09:43:08 we should get some public release policy straight May 10 09:43:27 so that people know what's up and what they can expect May 10 09:44:14 hello mrmoku May 10 09:45:07 ptitjes: hey May 10 09:45:31 so much frustration around :( May 10 09:46:06 ptitjes: (old) grrrrr May 10 09:46:06 mrmoku: what happened ? May 10 09:46:19 ptitjes: read MLs.... May 10 09:46:23 DocScrutinizer: no as old as you for sure May 10 09:46:37 mrmoku: the e problem ? May 10 09:46:50 yeah... and the discussions around it May 10 09:47:41 ptitjes: wait another 364 days! ;-) May 10 09:47:45 mrmoku: we are missing a stable release that's all May 10 09:48:01 ptitjes: yes... but we have to make that even more clear May 10 09:48:04 ooops 365 May 10 09:49:01 explain the difference between testing and unstable and that all can break ;) May 10 09:49:10 mrmoku: or releasing a stabel May 10 09:49:18 s/stabel/stable/ May 10 09:49:18 ptitjes meant: mrmoku: or releasing a stable May 10 09:50:04 mrmoku: but I know this will make yet much more work May 10 09:50:17 anyway... have to go now... bbl May 10 09:50:25 bye May 10 09:50:28 gtg too May 10 09:50:30 ciao May 10 09:50:49 @shr: I wonder if we should add gsm-flashing as a package to SHR May 10 09:50:49 DocScrutinizer: Error: "shr:" is not a valid command. May 10 09:51:11 muahaha May 10 09:53:22 pfff, all are "away". RUN, duck and cover! cowards! ;) May 10 09:53:50 @help May 10 09:53:50 dos1: (help [] []) -- This command gives a useful description of what does. is only necessary if the command is in more than one plugin. May 10 09:56:46 I strongly suggest this stupid bot takes some classes with bzzbot May 10 09:56:58 ~help May 10 09:57:45 where can i find shr? May 10 09:58:03 shr is http://shr-project.org/ May 10 09:58:08 where can i find shr? May 10 09:58:09 :P May 10 09:58:12 otherwise odds are wugbranler gets /ignore'd or even kicked/banned May 10 10:00:32 mrmoku|away: ping pang pong May 10 10:00:43 dos1: huh??? May 10 10:02:17 ping :P May 10 10:02:54 dos1: you should issue a '~help' yourself to see what I was talking about. compare to @help then May 10 10:03:44 DocScrutinizer: i know. i issued it directly to bzzbot. i only read " I learn mainly by observing declarative statements such as "x is at http://www.xxx.com", and then reply when people ask things like "where can i find x?"" May 10 10:03:46 Shouldn't bots always reply with a notice? May 10 10:03:46 oh, at May 10 10:03:58 shr is at http://shr-project.org/ May 10 10:04:03 where can i find shr? May 10 10:04:08 :( May 10 10:04:24 dos1: I mean you should issue it *here*, otherwise you won't see the diff May 10 10:04:32 ~help May 10 10:04:42 DocScrutinizer: it's the same... May 10 10:04:48 i don't understand you :P May 10 10:04:53 nope it isn't May 10 10:04:53 @help May 10 10:04:54 dos1: (help [] []) -- This command gives a useful description of what does. is only necessary if the command is in more than one plugin. May 10 10:05:03 @list May 10 10:05:04 dos1: Admin, Channel, Config, Misc, Owner, RSS, TracBot, and User May 10 10:05:45 don't you notice the difference??? May 10 10:06:19 bzzbot answers in private chat May 10 10:06:45 wugbranler spams the channel :-( May 10 10:07:06 DocScrutinizer: oh, i knew that already... :) May 10 10:07:21 so though you mean something else ;) May 10 10:09:11 CIA-75: help May 10 10:09:33 coma May 10 10:10:50 but that's OK :-D May 10 10:10:55 CIA-75: help May 10 10:10:58 ::P May 10 10:11:41 wugbranler should learn netiquette ;-> May 10 10:14:05 DocScrutinizer: non-standard, that's first. Second, it doesn't guarantee proper data aligment that is necessary anyway for most archs. Don't forget, implicit struct padding is there for a reason. May 10 10:16:20 BTW, has anyone noticed a strange thing: sometimes i resume and _immediately_ after that i hear incoming sms sound and then the message appears. It happened to me several times already but i was foolish enough (for unknown reasons) to not check logs... May 10 10:17:32 sounds like resume on sms is borked, no? :-/ May 10 10:20:47 DocScrutinizer: yeah, that's what i'm afraid of. But it doesn't happen every time, just sometimes (rarely). And i assume there might be a race where the incoming sms indication is already received but suspend is activated and therefore the message actually comes before the suspend, just processing is delayed. May 10 10:22:34 hmm, a very critical point. I know about quite some considerations and thoughts to avoid this case. Probably nobody has tested though May 10 10:24:50 maybe another flaw introduced by edge-trigger irq? May 10 10:27:34 level IRQ always blasts as soon as it's enabled even when trigger level was set prior to this. edge doesn't - edges occuring prior to enabling irq go unnoticed May 10 10:28:48 PaulFertser: that's why I argued several times we *must* use level IRQ instead of the current braindead both-edge-trigger-then-poll scheme May 10 10:30:08 it's incredibly stupid and wrong May 10 10:31:08 DocScrutinizer: first we disable flow control, then irq comes, then we suspend and never resume? Well, iirc it checks this condition very late on suspend, so timeslot for the race is very small. And i don't remember it ever happening with incoming call (after the moko11 fix). May 10 10:34:57 without some decent gdb-breakpoints set to stop shutdown in the "right" state, I don't see a way to test/debug this issue anyway May 10 10:38:32 PaulFertser: you noticed my comments regarding ALSA last night? May 10 10:39:13 DocScrutinizer: yeah, some dmix vs hw:0 strange difference. To be honest, i didn't understand it :) May 10 10:40:08 PaulFertser: I wonder if it's better to *resume* audio playback, or rather abort the playback app with a IO-error May 10 10:40:44 PaulFertser: now playback app freezes and doesn't recover at resume. that's not good for sure May 10 10:42:08 PaulFertser: the dmix bug is a completely different issue though May 10 10:42:35 DocScrutinizer: Just tried: started mpg321, heard music, suspended, resumed, music continued with "ALSA: underrun, at least 0ms." May 10 10:42:44 PaulFertser: just try the two cmds I suggested, and hear the diff May 10 10:43:07 PaulFertser: that's ok May 10 10:44:31 I tried with "while true; do aplay foo.wav; done" and aplay froze in half of the suspends May 10 10:45:52 PaulFertser: ^Z^C broke alsa completely as aplay didn't free the audioresource May 10 10:46:18 DocScrutinizer: hm, it'd be nice if broonie looked into this, after all he has hardware now. May 10 10:46:31 ack May 10 10:46:56 I need to create a set of tickets for all these issues May 10 10:51:19 BUG: stream = 0, pos = 0x2488000, buffer size = 0x4000, period size = 0x400 May 10 10:51:34 Don't have any idea what that means though. May 10 10:52:17 basically nothing useful May 10 10:53:42 you can tell the app is using 10 buffer segments in the ringbuffer, each segment (period) with a size of 0x400 May 10 10:56:49 PaulFertser: which app, how to reproduce? May 10 10:58:20 DocScrutinizer: i did while true; do aplay -D plug:dmix ./paroli/data/sounds/alarm.wav; done and then randomly suspended pressing power button (oeventsd handled it as usual). May 10 10:58:58 -D plug:dmix... also an interesting setup May 10 11:23:13 hmm... "1622 ? Z 0:00 [xset] " May 10 11:23:54 DocScrutinizer: frameworkd up to date? May 10 11:24:34 dos1: dunno what's "up to date". prolly not - this is shr-t-04-16 May 10 11:24:50 DocScrutinizer: few days ago there was fix for that May 10 11:25:14 k May 10 11:26:15 PaulFertser: it turns out "-D plughw:0,0" introduces a kind of echo to playback May 10 11:26:33 while true; do aplay -D plughw:0,0 /usr/share/sounds/purple/alert.wav ;done May 10 11:26:36 vs May 10 11:27:10 @!@help May 10 11:27:10 Ainulindale: Error: "!@help" is not a valid command. May 10 11:27:24 SHR-WugBranler: dumbass. May 10 11:27:24 Ainulindale: Error: "dumbass." is not a valid command. May 10 11:27:24 while true; do aplay /usr/share/sounds/purple/alert.wav ;done May 10 11:27:44 Ainulindale: ACK!!! May 10 11:28:26 SHR-WugBranler: dumbass. May 10 11:28:28 /kick shr-wugbranler May 10 11:28:33 Better. May 10 11:28:51 SHR-WugBranler: help May 10 11:28:52 dos1: (help [] []) -- This command gives a useful description of what does. is only necessary if the command is in more than one plugin. May 10 11:28:54 @ainulindale: what did U do? May 10 11:28:57 @shr May 10 11:29:07 ahaaaa May 10 11:29:09 @help May 10 11:29:09 dos1: (help [] []) -- This command gives a useful description of what does. is only necessary if the command is in more than one plugin. May 10 11:29:23 WTF?? May 10 11:29:42 Could you please avoid creating a mess when I'm testing? May 10 11:29:53 lol May 10 11:31:09 @?est May 10 11:31:14 Good. May 10 11:31:18 @!@test May 10 11:31:21 Better. May 10 11:31:30 @!@help May 10 11:31:30 Ainulindale: (help [] []) -- This command gives a useful description of what does. is only necessary if the command is in more than one plugin. May 10 11:31:37 Perfect. May 10 11:33:00 DocScrutinizer: hm, too many problems in too many areas... May 10 11:34:49 PaulFertser: hmm. well. May 10 11:35:35 PaulFertser: at least suspend tests are a PITA via ssh. And a not much minor pita via FR-terminal May 10 11:37:51 $ update-alternatives --install /usr/share/shr-splash/theme shr-splash-theme /usr/share/shr-splash/themes/shr-splash-theme-simple 10 May 10 11:37:51 update-alternatives: Linking //usr/share/shr-splash/theme to /usr/share/shr-splash/themes/{} May 10 11:37:52 why? May 10 11:37:54 :x May 10 11:39:08 oh... May 10 11:39:13 ok, there was no question May 10 11:39:16 ;D May 10 11:41:03 SHR: 03seba.dos1 07shr-overlay * r20ff2bb3f270 10/openembedded/recipes/shr/ (3 files): shr-splash-*: fix update-alternatives May 10 11:43:06 * PaulFertser resorts to sleeping -- a universal method of workarounding any problems May 10 11:44:36 DocScrutinizer: why? Here ssh connection over usb is resumed properly. May 10 11:44:54 not here May 10 11:45:01 host loses route May 10 11:45:10 SHR: 03seba.dos1 07shr-overlay * rd05d19cc35bc 10/openembedded/recipes/shr/ (3 files): Revert "shr-splash-*: fix update-alternatives" May 10 11:45:38 DocScrutinizer: with "allow-hotplug" in /etc/network/interfaces it works here on Debian May 10 11:45:52 (on host side ofc) May 10 11:45:53 DocScrutinizer: because i'm using an obsolete distro i have while true; do sleep 1; /sbin/ifconfig usb0 192.168.0.200 up; done running and it works here for me. May 10 11:46:20 hmm, might be part of the issue, yes May 10 11:47:04 And yes, on Debian host allow-hotplug is enough. May 10 12:04:04 no /etc/network here. KNetworkManager :-/ May 10 12:07:59 DocScrutinizer: /etc/network/ is debian specific May 10 12:09:30 probably on osuse yast nework devices would be the place to configure usb0 "card" May 10 12:49:09 dos1, who upped the call rec recipe? (I don't remember who I talked with about it) May 10 12:49:23 anyhow, there's an issue with it, for some reason, the patches are not moved to the app data dir May 10 12:49:24 ;\ May 10 12:49:36 TAsn: callrec? let's check May 10 12:49:51 it's supposed to be done with autotools May 10 12:49:56 furthermore, they get copied here May 10 12:50:07 I mean, when I build the pack from the same recipe. May 10 12:50:18 dos@sowa:~/openmoko/git/shr-overlay/openembedded/recipes$ find . | grep call May 10 12:50:20 nothing :x May 10 12:50:50 or at least I think May 10 12:51:01 dos1, opkg list | grep callrec May 10 12:51:29 TAsn: maybe it's in openembedded May 10 12:51:29 supposed to work anyway, no idea why it doesn't. ;\ May 10 12:51:34 dos1, may be. May 10 12:51:47 though anyhow, got any idea why would it happen? May 10 12:52:31 ;\ May 10 12:54:25 what's even worst is that some files are copied May 10 12:54:28 and some aren't. May 10 12:54:37 TAsn: i don't have any idea... May 10 12:54:42 ;\ May 10 12:58:31 I have a problem charging my GTA02s battery. I don't think it's the problem with the battery being fully discharged, since I tried to recharge it at 14% (according to the UI in SHR). I've tried both connecting a usb-cable to my computer and using the adapter, nothing. May 10 12:59:43 Oh, wrong channel. May 10 13:01:05 m_abs: no, why wrong chan? May 10 13:02:10 DocScrutinizer: lost the pointer to your pdf about buzz fix May 10 13:02:18 may do it during the afternoon May 10 13:03:02 ptitjes: search people.om/joerg May 10 13:03:03 arf first link on google May 10 13:03:09 :) May 10 13:03:10 hehe May 10 13:18:29 DocScrutinizer: also I can't find the procedure to open the case May 10 13:18:49 ptitjes: OMG o.O May 10 13:18:52 :-D May 10 13:19:18 search wiki for disassebling Neo 1973 May 10 13:19:42 I succeeded in fact :) May 10 13:20:36 opening the case is heard! May 10 13:20:58 needed to buy special screwdrivers and when opening with the pick it sounds like you are breaking it. May 10 13:23:59 SHR: 03seba.dos1 07shr-themes * re2a43fff9ae8 10/shr-splash/shr-splash-theme-dontpanic/ (6 files in 2 dirs): shr-splash-theme-dontpanic: artwork update May 10 13:24:45 DocScrutinizer: do you want dontpanic version without halo and text on bottom? or something else? May 10 13:27:46 dos1: I like a version with much more halo so converting to 5/6/5 won't erase the nice halo May 10 13:28:19 DocScrutinizer: and text on bottom (it's booting)... with or without? May 10 13:29:08 I'm ok with that text, tho it's mangled to ilegibility by conversion May 10 13:30:51 DocScrutinizer: so without text will be better option? May 10 13:35:18 I wouldn't mind this text going away either May 10 13:35:49 Ainulindale: ping May 10 13:36:22 or replace it by address of the manufacturer of the original H2G2 May 10 13:36:35 somewhere on Altair or sth ;-D May 10 13:37:36 bah, fucking idiots. May 10 13:37:39 dos1, I found the problem May 10 13:37:44 though I don't know how to pass it. May 10 13:37:55 maybe even write a small "hitchhiker's guide to the galaxy" obove of "don't panic" May 10 13:37:57 for some reason it doesn't try to update sources from svn, but uses the cache May 10 13:39:09 DocScrutinizer: right now i have: http://openmoko.opendevice.org/~dos/dontpanic-uboot.png May 10 13:40:27 someone with SHR buildhost access here? May 10 13:40:32 i want to send mail about shr-splash May 10 13:40:36 dos1: the halo isn't stronger it seems, it's just much more blurry (which doesn't look as good as in 1. version) May 10 13:40:49 but shr-splash isn't generated on buildhost May 10 13:40:54 DocScrutinizer: hmm... May 10 13:42:08 dos1: the problem with halo is the cutoff of lower 3 bits of each color (2 for green) is converting slight color of halo to black May 10 13:42:30 DocScrutinizer: refresh May 10 13:42:32 it needs more saturation to survive conversion May 10 13:43:06 dos1: looks good May 10 13:43:55 dos1: check if the most delicate colors of halo are still 0b00001xxx at least for red channel May 10 13:43:57 What's a good Elementary widget to implement a command prompt-like thingie? Entry? May 10 13:44:14 thrashold: yep, entry May 10 13:44:29 Thanks :) May 10 13:44:34 everything less suturated will convert to black May 10 13:44:42 saturated May 10 13:45:03 dos1, can you rebuild callrec and make it download from svn May 10 13:45:10 and not use a stupid cache? May 10 13:45:15 TAsn: i don't have access to shr buildhost May 10 13:45:19 shit. May 10 13:45:24 mrmoku|away, ping me when you are here. May 10 13:45:26 Ainulindale, mrmoku|away: ping May 10 13:45:46 TAsn: i also need someone with buildhost access ATM :D May 10 13:47:38 dos1, btw, thanks for the tip May 10 13:47:41 using nm-applet ;] May 10 13:50:39 DocScrutinizer: I just read 2k2 May 10 13:50:51 huh? May 10 13:50:53 DocScrutinizer: It seems the value of the resistor is correct May 10 13:51:00 wow May 10 13:51:15 so you're lucky :) May 10 13:51:25 DocScrutinizer: between the - and + arrows that you have drawn on the first picture May 10 13:51:29 just need to add big C May 10 13:51:30 good May 10 13:51:33 :D May 10 13:51:59 ptitjes: between case of mic, and GND? May 10 13:52:49 yep May 10 13:54:48 ptitjes: you could check by setting #53 to "mic3", #63 to "mic1"(iirc) May 10 13:55:13 if the R is ok, you should get audio from mic May 10 13:55:40 if the R is 0 ohm, you won't get audio May 10 13:55:41 DocScrutinizer: sorry but I don't get what you speak about... May 10 13:55:51 #53 ? May 10 13:56:09 control.53 in alsa May 10 13:56:19 DocScrutinizer: phone is off May 10 13:56:47 yeah, just a silly suggestion. nevermind May 10 13:56:51 :D May 10 14:01:58 DocScrutinizer: what do you think about solder the wire on the cap first May 10 14:02:00 ? May 10 14:02:24 ptitjes: very good idea May 10 14:26:22 done May 10 14:26:23 :D May 10 14:27:37 ptitjes: DONE?? you fixed buzz? May 10 14:27:45 DocScrutinizer: yeah` May 10 14:27:48 :) May 10 14:28:02 DocScrutinizer: I still hear some "wind" on the other party side May 10 14:28:05 cooool, and - does it work? May 10 14:28:15 it does work May 10 14:28:29 yet I have to adjust all the alsa values now I suppose May 10 14:28:35 so you should double-check your gsmhandset.state now May 10 14:28:42 :-) May 10 14:28:44 DocScrutinizer: because I hear some breath on the other side party May 10 14:28:52 and also I have echo May 10 14:29:15 hmm, so you haven't the right AT%Nxxxx handling then May 10 14:29:28 I assume May 10 14:30:27 hum I'll to dig all the afternoon for all the information I guess May 10 14:30:39 or there is a central place fo this May 10 14:30:43 s/fo/for/ May 10 14:32:59 ptitjes: you should visit http://wiki.openmoko.org/wiki/Neo_Freerunner_audio_subsystem#Alsamixer_channel_controls May 10 14:34:54 ptitjes: ping May 10 14:35:01 hello mickeyl May 10 14:35:14 ptitjes meant: or there is a central place for this May 10 14:35:58 ptitjes: if you have some time over the next days, could you update our branches? Jürg commited a huge bunch of posix patches which I had previously in a posixextra.vapi. May 10 14:36:46 yep no problem May 10 14:37:12 I'll do that today, after I suppress the echo on my buzz-fixed fr May 10 14:37:14 :) May 10 14:37:35 mickeyl: btw, I fear the way you declared the common errors won't bind correctly to Vala May 10 14:37:50 ptitjes: that's sad -- why not? May 10 14:38:02 mickeyl: I don't if one can write public void foo() throws MyErrorDomain.MY_ERROR May 10 14:38:22 mickeyl: this is what I tried to explain you some days ago May 10 14:38:45 throws clause in Vala defines an error domain that can be throw, not an individual error code May 10 14:39:04 sure, but you can throw multiple error domains May 10 14:39:20 mickeyl: this is why I adviced you to group your individual errors by application-kind of error ("domain") May 10 14:39:37 you don't get me May 10 14:39:42 correct May 10 14:39:43 look at what you commited there: May 10 14:40:16 http://git.freesmartphone.org/?p=specs.git;a=blobdiff;f=org.freesmartphone.Device/org.freesmartphone.Device.LED.xml.in;h=1724a6738eb175f01ea45678546c089a684bbb0d;hp=d88d65e3e98c37f09c691fd85028dc38095e2f78;hb=e52a73c7e446fef2419f35d7796abb12a3d5046f;hpb=47ac37a1c3d7167eb2776eed2222e37d6052bbef May 10 14:40:20 oups long url May 10 14:40:32 May 10 14:40:37 this is equivalent to May 10 14:41:02 .... throws MyError.Unsupported May 10 14:41:09 ah, that's wrong May 10 14:41:10 ok, then we need to change this to May 10 14:41:29 and by the way I also fear that the name "org.freesmartphone" won't bind too May 10 14:41:37 as it is the namespace name May 10 14:41:39 only May 10 14:42:00 hmm, that's sad again May 10 14:42:16 mickeyl: I understand that you can not change the already existing error names May 10 14:42:27 mickeyl: so I might introduce a vala xmlns May 10 14:42:39 well May 10 14:42:42 mickeyl: vala:name="CommonError" May 10 14:42:45 it's not so much about existing APIs May 10 14:42:53 it's more about the style i like for the error May 10 14:43:09 if the API throws May 10 14:43:12 org.freesmartphone.Unsupported May 10 14:43:20 we need to point people to this in the xml spec May 10 14:43:21 mickeyl: but IMHO that style (no error grouping and only individual errors) is loose May 10 14:43:29 it does not transport so much information May 10 14:43:54 i prefer this much over adding one error per method for invalid parameters May 10 14:44:00 or per interface May 10 14:44:08 feels too redundant to me May 10 14:44:25 mickeyl: I did not say you have to do that to May 10 14:44:31 :) May 10 14:44:55 mickeyl: but if there is some fso-specific technical errors May 10 14:45:04 then there might be an error domain for that May 10 14:45:14 sure May 10 14:45:15 mickeyl: but then there might also be other errors May 10 14:45:20 and i thought the domain might have the proper name May 10 14:45:27 but if that doesn't work due to namespace problems May 10 14:45:32 then we have to change the name of the error domain May 10 14:45:44 mickeyl: then calling the technical errors errordomain "org.freesmartphone" does block for other common error domain May 10 14:45:56 and does not convey the information it is a technical error May 10 14:46:25 what exactly do you mean with technical error? May 10 14:47:00 I would say InvalidParameters, InternalError, SystemError and Unsupported are technical errors May 10 14:47:09 they are not application errors May 10 14:47:17 like NumberNotFound May 10 14:47:25 correct May 10 14:47:26 (I invent there but I think you get me) May 10 14:47:44 thus I would call the errordomain May 10 14:48:02 org.freesmartphone.TechnicalError May 10 14:48:06 or something better maybe May 10 14:48:15 mickeyl: yet I am completely open May 10 14:48:24 I just want you to see what bugs me May 10 14:48:54 do you have a problem with the dbus name or the errordomain name or both? May 10 14:50:20 mickeyl: I would say the dbus error name is too weak but yet this is not my problem May 10 14:50:31 if you like it like this (no pun) this is ok for me May 10 14:51:27 well May 10 14:51:30 i like hierarchical errors May 10 14:51:35 as well as interfaces May 10 14:51:42 that's why we have org.freesmartphone.GSM.SIM.Foo May 10 14:52:02 which should raise specific errors like org.freesmartphone.GSM.SIM.Problem May 10 14:52:05 but also generic ones like May 10 14:52:08 which should raise specific errors like org.freesmartphone.GSM.Problem May 10 14:52:15 so it came natural to me to call generic errors May 10 14:52:18 which should raise specific errors like org.freesmartphone.Problem May 10 14:52:28 err May 10 14:52:36 s/which should raise specific errors like/ May 10 14:52:55 yeah I understand you pov May 10 14:53:09 I've no fixed opinion May 10 14:53:17 mickeyl: I need to pull my specs.git May 10 14:53:26 regen libfso to what it will produce May 10 14:53:32 as for the errordomain lets go for Common then May 10 14:53:45 then we'll see May 10 14:53:46 although it will surprise users May 10 14:53:56 when the errordomain is different than the actual dbus error prefixes May 10 14:53:58 yeah May 10 14:54:04 lets fix the type first though May 10 14:54:14 where did i do it wrong? May 10 14:54:17 everywhere? :) May 10 14:55:05 just network and devic May 10 14:55:12 mickeyl: I will try to regenerate it later today, and I'll tell you what happens May 10 14:55:17 ok May 10 14:55:27 i'll fix the things which are definitely wrong though in any case May 10 14:55:53 mickeyl: btw I love that you renamed the folders :) May 10 14:55:57 :) May 10 14:55:59 was about time May 10 14:56:28 I hope you don't mind I watch your commits ;) May 10 14:56:36 of course not May 10 14:56:42 we're here to improve things May 10 14:56:48 (btw I hope you do the same with mine :) ) May 10 14:57:35 freesmartphone.org: 03mickey 07specs * raba44ae3bebc 10/ (2 files in 2 dirs): May 10 14:57:35 freesmartphone.org: org.freesmartphone.Device.LED / org.freesmartphone.Network: fix fso:throws type to raise an error category, not a specific error May 10 14:57:35 freesmartphone.org: NOTE: This is tentative and not final and may change at any time May 10 14:57:36 sure May 10 14:57:46 mickeyl: what is the best way: using alsamixer or modifying the state files May 10 14:57:48 ? May 10 14:57:56 to fix my echo May 10 14:58:18 well, the state files overwrite the mixer settings May 10 14:58:25 oki May 10 14:58:29 so at the end of the day you want to have things fixed in the statae May 10 14:58:33 use alsamixer to fix it May 10 14:58:37 then store from card to file May 10 14:58:40 (with alsactl) May 10 14:58:47 mickeyl: also do you know if the correct AT% thing is in frameworkd ? May 10 14:58:53 ptitjes: there IS NO WAY to fix echo via mixer settings! May 10 14:58:56 there is no such thing as the "correct" AT% May 10 14:59:15 but we export it via setting May 10 14:59:20 ti_calypso_dsp_mode (or so) May 10 14:59:27 the valid values are in our example conf May 10 14:59:33 in the framework git tree May 10 14:59:34 DocScrutinizer: don't shout at me!! I did nothing :( May 10 14:59:43 sorry May 10 14:59:52 :) May 10 14:59:56 ptitjes: and by default it's maximum suppression. May 10 15:00:08 arf but I still have it May 10 15:00:21 just I get mad hearing "fix echo by alsa setting" over and over and over and over again May 10 15:01:08 DocScrutinizer: heh don't forget I'm a stupid noob May 10 15:01:20 (that does not read the MLs above all :) ) May 10 15:03:13 ptitjes: btw, wrt to hardware reasons that lead to echo (you might want to experiment a bit): when there's only PCB, there's no feedback; when the case is half-assembled, there is more; when it's fully assembled, even more; when the screwes are in place and tightened, it's maximum. And in my tests all feedback was eliminated by %N0187 (you can even try to use mickeyterm during the call to experiment with various values). May 10 15:03:14 ptitjes: AT%Nxxxx is used to suppress echo. It needs to be issued after modem power-up and after every modem-reset and after evry call, it's not persistent regrettably May 10 15:03:34 And frameworkd already issues the command every reasonable time. May 10 15:04:00 oups I forget to put the screws back!!! May 10 15:04:20 ptitjes: i don't use them for months, they're not strictly needed. May 10 15:04:33 yup May 10 15:04:48 cool May 10 15:05:00 yeah it seems the box stay assembled without them May 10 15:06:15 So i guess some creative guy can almost completely mitigate feedback (and therefore use less aggressive echo suppression) after applying some foam to some places... May 10 15:06:25 PaulFertser: did you experiment with leaving out he rubber ring of mic? and placing a foam ring on backside of earpiece transducer? May 10 15:06:52 DocScrutinizer: sadly, no May 10 15:11:25 hmm May 10 15:11:26 Post by non-member to a members-only list May 10 15:11:40 Ainulindale: can you let this through and whitelist my real address,please? May 10 15:11:46 i'm not using the openmoko address anymore May 10 15:12:21 mrmoku|away, Ainulindale: any hint on the release I should upgrade too to test my buzz-fixed phone ? May 10 15:12:27 BTW, JC agreed with my reasoning that in general mailing lists should be open to posting by non-subscribed users. May 10 15:13:42 sounds good as long as there is a full time person administrating the spam filter May 10 15:15:19 mickeyl: have you ever seen a spam with "PASS" SPF result? May 10 15:16:03 no idea, i don't check the headers of my spam May 10 15:17:19 I did a random check on my spam folder. No spam letter i saw was SPF pass. And for those real users for whom spf fails, i'd say fix your mail domain configuration. But spammers seem to never do that. May 10 15:18:35 I got 3 spam on my om account today :-( May 10 15:19:19 if ML will develop similar habit, I have to unregister May 10 15:19:33 mickeyl: can you somehow explain that sometimes (very rarely) _immediately_ after resume i hear the incoming message sound and see the message? May 10 15:20:30 are you using SIM buffering or direct? May 10 15:20:33 mickeyl: sorry for not providing logs, i don't know why i didn't check / save them... May 10 15:20:45 mickeyl: Hm, zhone May 10 15:21:27 SIM buffering then. In theory we leave SMS delivery messages enabled during suspend May 10 15:21:55 Can there be a race like the incoming message unsol received during suspend but it's already activated? May 10 15:22:07 And therefore processed only after resume. May 10 15:22:47 And yes, most of the time incoming message initiates resume, just it doesn't seem to be always. May 10 15:23:11 possible. there's a short period of time where GSM is in 'suspending' state, where signals are queued May 10 15:23:20 race on IRQ handling due to edge trigger madness in kernel May 10 15:23:25 the queue is only processed after resuming May 10 15:23:39 ya, might as well be a kernel race May 10 15:24:55 the whole concept of linux suspend/resume makes things very difficult for a phone May 10 15:25:06 I don't have a hard proof but after moko11 was out i've never experienced symptoms of "incoming call doesn't resume", of course it doesn't prove anything. May 10 15:25:08 if you check the android kernel sources, you will see some pointers in that direction May 10 15:25:36 usually you should: 1) disable gsm-tty, 2) process queue to empty, 3) enable IRQ. On 3) a resume reason should trigger if modem received a sms meanwhile May 10 15:25:40 I don't think android is a good example, i don't even want to think about it. May 10 15:25:54 heh May 10 15:26:04 i think it's an example for the complexity of the suspend/resume concept May 10 15:26:14 whether good or bad i can't judge May 10 15:26:53 Probably unnecessary complexity for non-technical reasons. They never aimed to build a good unix-like system after all, all they wanted is to use a good kernel for their non-standard stuff. May 10 15:27:19 What we could do is change the flow control order May 10 15:27:23 alas with our braindead concept of edge-trigger any pending IRQ might go unnoticed on enabling the IRQ May 10 15:27:40 right now we touch the flow control sysfs node very late during suspending May 10 15:27:57 mickeyl: see above May 10 15:28:07 I'm afraid if we do it earlier we'll just miss that irq. May 10 15:28:28 yeah May 10 15:28:54 what we could do is do a round of queue checking before we actually suspend May 10 15:28:57 but it wouldn't help much May 10 15:29:14 since by that time, the GSM is suspended and the modem will not accept any commands, e.g. to read the actual incoming message May 10 15:29:21 so that won't help May 10 15:29:47 s/GSM/GSM resource/ May 10 15:30:16 if IRQ handling would be done correctly, we could queue a modem-wake-IRQ manually May 10 15:31:10 "Last chance: abort if GSM already interrupted" May 10 15:31:44 This kernel function is called very late in a suspend process, it should take care to terminate it if there's incoming irq. May 10 15:32:31 oh how i'd love to just halt the CPU May 10 15:32:34 If flowcontrol asserted, abort if GSM already interrupted May 10 15:32:35 ~shoot suspend/resume May 10 15:32:41 it also should take care to terminate if there's a msg in the queue May 10 15:32:47 That's a comment in the sources. May 10 15:33:28 mickeyl: ACK :-) May 10 15:33:30 So, i don't see problems kernel-wise, only a tiny timeslot (between late suspend hook and actual suspend) is left for races. May 10 15:34:11 this mustn't be there as well. We need level IRQ May 10 15:34:33 not this nasty bastard madness May 10 15:34:45 of edge and poll May 10 15:34:48 Tell that Harald :-/ May 10 15:35:05 duh, why not do it better than he did? May 10 15:35:29 Let me check it first... May 10 15:36:41 DocScrutinizer: why do you think that modem wakeup irq is ever polled? May 10 15:37:04 I can't see anything like that in the code. May 10 15:38:12 So, buttons irq handling is definetely nasty, but i can't see how rising edge for modem irq can lead to any more troubles than level. May 10 15:42:18 Well, i just simulated a case where nothing fails :) May 10 15:42:41 PaulFertser: edge is *wrong*. As mentioned above, a level-IRQ will fire immediately on prev level change the moment you enable the cpu-IRQ May 10 15:43:25 Did echo 1 > flowcontrolled; called the phone, tried to suspend by echo mem > /sys/power/state, got "PM: Device neo1973-pm-gsm.0 failed to suspend: error -16" in dmesg, so no suspend. echo 0 > flowcontrolled, zhone showed the call... May 10 15:43:37 whereas a edge ever goes unnoticed if it happens prior to enabling the IRQ-handler May 10 15:45:01 DocScrutinizer: look, if we calypso is not flowcontrolled, it won't do irq, if it is, then the irq will be catched by the kernel (irq handler is always running) and if suspend is initiated it will refuse to suspend. May 10 15:46:31 there's a gap. there's always a gap. you can mathematically proove it's impossible to write a decent irq-handler if you use poll anywhere May 10 15:46:42 There's no poll anywhere. May 10 15:47:15 huh, so this IRQ is level-triggered? May 10 15:47:25 No, it's edge triggered but never polled. May 10 15:47:51 The irq is never disabled therefore always handled. May 10 15:48:03 if it's edge triggered it can't work May 10 15:48:44 every IRQ is disabled in a small timewindow May 10 15:48:46 I agree there's a small gap. May 10 15:49:05 as small gap is a gap May 10 15:49:36 there's no way to say "this gap is small enough we don't need to care" May 10 15:49:55 Well, i agree. And i think i see how it can be done properly with level, you're right. May 10 15:50:04 race will always happen eventually May 10 15:51:38 Though i still believe frameworkd in its current state is much more problematic wrt this than kernel. May 10 15:52:33 frameworkd simply needs to abort suspend if there's a msg in queue May 10 15:52:55 process message May 10 15:53:13 then eventually a new suspend may happen later on May 10 15:53:33 msg in queue is "wake-cause" May 10 15:55:55 [17:25] usually you should: 1) disable gsm-tty, 2) process queue to empty, 3) enable IRQ. On 3) a resume reason should trigger if modem received a sms meanwhile May 10 15:56:28 DocScrutinizer: but the way i can see level implemented would break broken userspace. I see it like this: initially irq set to level and enabled. On irq we disable irq and set a global var to 1. With this global var set to 1 suspend will always abort. To set it back to 0 we must find a place to re-enable irq and then set the var to 0, i propose to do it when we echo 0 > flowcontrolled. That will break some userpace for sure. May 10 15:57:35 basically you're right May 10 15:59:05 IRQ needs to be reset (which means read from modem). after device causing IRQ has deasserted level due to being serviced, you reenable the irq-handler on cpu May 10 15:59:35 But if we passed the last late_suspend check for the global var we should disable disabling irq in irq handler... this gets messy again. May 10 16:00:23 :) May 10 16:00:28 nah, I don't see this May 10 16:00:32 deja vu May 10 16:01:47 As you guys discuss, if you find a yellow-handled screwdriver, it's probably mine -- I lost it a year ago when following this same path through the kernel suspend/resume and GSM IRQ stuff. :) May 10 16:02:34 BTW, the current handling is sub-optimal, but it's the best I could get Andy to accept. May 10 16:03:02 Another (probably better) possibility: we don't process irq at all ever (empty handler), just configure it as a wake-up source. And don't care about it anymore wherever, just high level on the pin will always resume and it's up to the userspace to clear it. It's more user-space friendly but... irq storm is bad too... If only we could configure it as a wakeup source but didn't have to service it ever. May 10 16:06:39 PaulFertser: btw there is no "global var" it's just IRQ is disabled automatically (by cpu) on entering irq-handler, and usually you enable IRQ again on exit of handler. not though if you trigger a worker to do the actual possibly timeconsuming servicing of modem. then this worker had to signal "irq device serviced and reset, now reenable IRQ handling". anyway this could also be done by a second irq-handler for low-level that gets trigger May 10 16:06:39 ed on modem deasserting IRQlevel - this second handler then simply would set IRQ callback reg to IRQhandler#1 again May 10 16:07:41 DocScrutinizer: if you got level_high irq and don't disable it, you'll get it again and again. If you disable it and suspend, you won't wakeup. May 10 16:07:55 Has anyone looked at the GSM IRQ handler? May 10 16:08:00 mwester: i did May 10 16:08:10 It consists of maybe... 2 instructions? May 10 16:08:15 nah, no way May 10 16:08:17 No worker threads. May 10 16:08:22 DocScrutinizer: Yep. May 10 16:08:25 mwester: yes May 10 16:08:32 It sets a variable, and that's all it does. May 10 16:08:49 Oh - and it used to printk, but that's probably removed by now. May 10 16:09:08 if IRQ is triggered, it's also disabled until exit of irq-handler May 10 16:09:25 mwester: it's not May 10 16:09:26 DocScrutinizer: It's not a serial IRQ -- it's just a GSM-interrupted IRQ, so there is no servicing of anything at all. May 10 16:09:36 irq-handler takes care about it won't trigger recursively May 10 16:09:51 PaulFertser: ok, I guess it doesn't matter much, but I think that printk is probably not useful anymore :) May 10 16:10:24 DocScrutinizer: in the level irq handler we should either disable irq or remove the reason that caused it. We can't do the second and if we do the first we shouldn't suspend. May 10 16:10:27 mwester: nonsense. the modems disasserts the irq line as soon as serial line gets enabled and serviced May 10 16:10:43 Yes, the irq-handler does some work -- but the actual IRQ service routine for the the GSM interrupt does nothing of any interest, it's about as close to a wake-only as you can get. May 10 16:11:03 DocScrutinizer: What does that have to do with the GSM IRQ? May 10 16:11:07 I think you are confused. May 10 16:11:26 PaulFertser: if we got IRQ we shouldn't suspend anyway ;-D May 10 16:11:28 But I must go - other things to do. As I said, I've been in this code a lot. May 10 16:11:34 Happy day! May 10 16:11:41 DocScrutinizer: we can't really prevent it. May 10 16:11:46 mwester: Luck! :) May 10 16:12:33 PaulFertser: what is it we cant prevent?? May 10 16:13:12 DocScrutinizer: how? i proposed to set a global var and check it and refuse to suspend. But it's as racy as the current behaviour. May 10 16:13:27 do you say we can't prevent the device going to suspend even if we got a wake-irq??? May 10 16:14:20 DocScrutinizer: i'm not sure we can do that in a race-free way, yes. May 10 16:14:51 I'm rather sure. there's always a racefree way May 10 16:15:42 at least as long as the same hw (cpu) has to do the suspend May 10 16:16:07 it can't do that as long as it is in irq-handler May 10 16:18:03 If there was a way to cancel suspend directly from an irq-handler... Probably there is but i'm unaware of it. May 10 16:18:29 DocScrutinizer: it seems people I call are pretty happy of voice quality May 10 16:18:41 DocScrutinizer: Real thanks to your creativity!!!!! May 10 16:18:41 great May 10 16:18:48 :-)) May 10 16:19:02 ptitjes: welcome May 10 16:19:03 DocScrutinizer: my neo is also fixed May 10 16:19:07 DocScrutinizer: So I might give some help at F9N May 10 16:19:25 DocScrutinizer: I think we can do a chain May 10 16:19:35 sure May 10 16:19:50 ptitjes: but you didn't have to do the hard part -- changing 0402 R. May 10 16:19:54 I'm the "desolder-R man" May 10 16:20:01 1) triage based on whether the existing resistor is 2ké May 10 16:20:03 2k2 May 10 16:20:12 2) you do the resistor stuff May 10 16:20:21 (that is really too tiny for me :) ) May 10 16:20:37 3) I can help to do come of the caps May 10 16:20:44 s/come/some/ May 10 16:20:45 ptitjes meant: 3) I can help to do some of the caps May 10 16:20:46 soldering is too tiny for me too, to be honest May 10 16:20:54 yeah for sure May 10 16:21:05 even my best loop/lamp wa hard to use May 10 16:21:39 you see, I'm close to being an old man now ;-) May 10 16:21:48 PaulFertser: yeah hopefully :) May 10 16:22:00 anyway. time for breakfast. May 10 16:22:03 cya all May 10 16:22:03 DocScrutinizer: naaah May 10 16:22:06 ptitjes: it seems you're the first guy that lucky May 10 16:22:10 ciao May 10 16:22:12 DocScrutinizer: enjoy :) May 10 16:22:16 DocScrutinizer: and thank you again :) May 10 16:22:39 I now have a real phone :) May 10 16:27:10 mickey|mday, just started working on onetworkd, since it doesnt need any plugins per se, is it ok if I keep it as simple as possible without using the typemodule infrastructure (if its possible) May 10 17:01:54 freesmartphone.org: 03jluebbe 07framework * redbe2cecece9 10/framework/subsystems/otimed/otimed.py: otimed: handle the case when we have no GSM time zone and are in a multi zone country May 10 17:23:02 Arhuaco: hey :) Not much feedback at LAKML, it can mean that either everything is ok and will be accepted soon or something is seriously wrong ;) I hope for the first ;) May 10 17:24:05 Hello PaulFertser. I hope for the first also. I'll resubmit now adding the right "From"s. May 10 17:24:25 Arhuaco: and S-o-b's :) May 10 17:25:01 Arhuaco: they seem to be already there though :) May 10 17:25:30 Yes PaulFertser. That's it. But I need to add my own S-o-b to every email. May 10 17:25:36 * Arhuaco did it. May 10 17:26:13 Arhuaco: yeah, that's what i'm talking about, as you pass those patches and can certify their origin. May 10 17:26:21 Ainulindale, mrmoku|away: ping May 10 17:26:28 Arhuaco: i was just a bit unsure so checked SubmittingPatches again :) May 10 17:29:02 Arhuaco: btw, i saw sometimes people omit full defconfigs from patch series, not sure if it should be done this time or not... In fact, great work, keep going! Upstream ftw! :) May 10 17:33:08 PaulFertser. At least I put the defconfig in a separate patch... I noticed Russell complained people who do not do that recently. May 10 17:34:14 PaulFertser. At least I put the defconfig in a separate patch... I noticed Russell complained people who do not do that recently. May 10 17:34:16 Arhuaco: Yes, saw that too but failed to understand the reasons. May 10 17:36:39 PaulFertser. Too much noise. Nobody really reads that :-) May 10 17:40:29 mrmobil: where you are? May 10 17:40:29 :P May 10 17:40:49 here ;) May 10 17:41:10 mrmobil: i wrote mail about shr-splash... and someone have to generate packages now ;x May 10 17:41:22 had 2l of good bavarian beer May 10 17:41:33 enjoying rl May 10 17:41:39 PaulFertser: funny I was thinking about IRQ the whole breakfast. You're right gap on suspend is impossible to solve with standard means. There has to be a dedicated means (similar to RETI return from interrupt) to enable wake-from-interrupt same time the CPU suspends. May 10 17:42:00 DocScrutinizer: :) May 10 17:42:09 mrmobil: Wow i wished i could do that too :) May 10 17:42:18 * PaulFertser is green with envy May 10 17:42:48 PaulFertser: I'll bring some real good franconian beer to F9N ;-) May 10 17:43:23 hmm keller bier May 10 17:43:25 DocScrutinizer: this reminds me to go fix modem power supply, thanks :) May 10 17:47:38 PaulFertser: (wake-from-interrupt) OTOH WAKE-from-IRQ isn't necesserily same as enable-IRQ. So if CPU is simply running beyond the SUSPEND instruction if there's a wake-level even on a IRQ-*dis*abled line, this would work perfectly May 10 17:48:10 DocScrutinizer: yeah, that's exactly what we need, as i said earlier. May 10 17:48:22 my neo moko seems to crash often lately May 10 17:48:26 DocScrutinizer: i just don't know if we can do that or not given SoC limitations. May 10 17:48:33 no it hangs often May 10 17:48:44 i have the feeling this has something to do with that memory card in the phone May 10 17:48:57 since if i want to access the card with ls /mnt/card my console hands May 10 17:48:59 hamgs May 10 17:49:00 hangs May 10 17:49:13 PaulFertser: if it's not a braindead CPU instruction set worse than 8088 then there has to be a way to handle this May 10 17:49:19 that is true for all processes needing to access that card. anyone an idea? May 10 17:50:04 qknight: no idea... does it panic? May 10 17:50:09 i even can't "kill 1231" since maybe the kernel needs to access the card or a reference on that card on 'kill' as well May 10 17:50:14 PaulFertser: I'm quite sure IRQ and wake-source are two distinct things May 10 17:50:19 DocScrutinizer: I'm afraid to look at SoC specs ;) May 10 17:50:26 PaulFertser: no it does not panic (sometimes it does but i think that is caused by a different issue) May 10 17:50:38 lol, I did, it's as few as ~900pp May 10 17:51:00 poorly structured May 10 17:51:03 [ 5239.885000] pcf50633-rtc pcf50633-rtc: PCF_TIME: 10.05.09 17:43:39 May 10 17:51:03 [ 5239.885000] pcf50633-rtc pcf50633-rtc: RTC_TIME: 10.4.109 17:43:39 May 10 17:51:03 [ 5239.905000] pcf50633 0-0073: Masking IRQ 7 May 10 17:51:03 [ 5239.925000] pcf50633 0-0073: Masking IRQ 7 May 10 17:51:04 Really, that's less than i expected. May 10 17:51:10 i get those often lately May 10 17:51:12 and fed up with bugs May 10 17:51:43 qknight: usb access hangs too? May 10 17:51:54 PaulFertser: no May 10 17:52:12 qknight: anything interesting in dmesg during "console hang"? May 10 17:52:32 PaulFertser: i'm not sure but what i have postet above seems to be there quite often May 10 17:52:40 qknight: irrelevant May 10 17:52:49 PaulFertser: ok May 10 17:52:50 sigh May 10 17:53:00 qknight: the first two (RTC) lines are completely meaningless May 10 17:53:00 qknight: tried fscking? May 10 17:53:03 i try to reinstall the card (made it work last time) May 10 17:53:17 qknight: do you have SIM installed? May 10 17:53:26 qknight: "masking IRQ7" doesn't look bad in itself May 10 17:53:28 PaulFertser: SIM installed and GSM connected May 10 17:53:45 qknight: ok, it's just that some devices have bad usd connection without sim. May 10 17:53:47 qknight: I suggest remount microsd, with async, never understand why shr mount it synced May 10 17:53:48 my neo won't shutdown May 10 17:53:54 it seems just like a hang of that sd card May 10 17:54:39 max_posedon1: sync? weird... May 10 17:54:43 max_posedon1: ok, i try the async thing first May 10 17:54:49 what causes that? May 10 17:55:54 i dont know, but it can help you to have this issue not often May 10 18:08:29 /dev/mmcblk0p1 /media/card auto defaults,async,noauto 0 0 May 10 18:08:33 ~that was default May 10 18:08:37 seems it is already async ;P damn May 10 18:18:00 last night's shr update for text messaging - composition is wrapping at about 8 characters May 10 18:18:39 DocScrutinizer: does it ever happen in DE that devices suddenly get 380VAC from mains? May 10 18:19:03 rarely :-o May 10 18:19:45 DocScrutinizer: not that unusual in Russia unfortunately, every now and then somebody tells stories about fried devices :-/ May 10 18:20:15 a wellknown problem though, of old mains installation, as the neutral line might see much higher current than the R, S, T liifelines May 10 18:21:57 so if your N fries and blows, a coffecooker or heater may short a much weaker consumer - e,g, a monitor - to R instead of N. So the monitor sees R - S == 400V May 10 18:21:58 DocScrutinizer: probably because of the higher current the neutral line sometimes just loses connectivity (they call it burnt off) and voila, you have Phase A, your neighbour has Phase B, both have devices plugged in and the same neutral line. May 10 18:22:21 DocScrutinizer: exactly :) May 10 18:25:17 In Germany it's prohibited to use common N for nonsymetric consumers for long time now. I.E RST with a common N also have to use a 3-phase fuse and a 3-pase power outlet / consumer. No more separate consumers on same N but distinct L May 10 18:27:14 btw this issue is introduced by non-ohmish consumers like switched power supply etc only. Wasn't a problem back when there wasn't such non-linear loads May 10 18:28:16 as number of switched power supplies (without PFC) rises, you see more and more of such incidents May 10 18:28:58 so it's even deprecated or maybe already forbidden to sell SPS without PFC here May 10 18:30:10 PaulFertser: I'd suggest a decent UPS ;-) May 10 18:30:31 helps cure a few of those nasties May 10 18:31:01 and protects your expensive appliences May 10 18:35:10 PaulFertser: the real nasty about all this is N never is fuse-protected (for obvious reasons) and thus an overload might start a fire May 10 18:35:57 dos1: pong ;) (pang pling ping) May 10 18:36:00 basically fire is the only failure-mode ;-D May 10 18:36:26 mrmoku: please build shr-splash, shr-splash-theme-simple and shr-splash-theme-dontpanic May 10 18:36:31 ;) May 10 18:37:42 dos11: will do... but just because I'm drunk ;) May 10 18:38:00 DocScrutinizer: luckily i don't have anything expensive :) May 10 18:38:13 Is compiling drunk allowed by law? May 10 18:38:17 :D May 10 18:38:41 thrashold: guess... in france yes ;P May 10 18:38:58 it's mandatory by hacker laws ;-) May 10 18:39:16 and compiling completly trashed is entitled by law in the netherlands May 10 18:39:30 the thing is... have to do a VAT declaration too... which is much harder than building :( May 10 18:40:19 oph NOOOEES, this word again... VAAAAT May 10 18:40:39 declarificifation May 10 18:43:19 DocScrutinizer: bad thing... yeah. Have to do that every month... until 10th :( May 10 18:46:19 dos1: did you just change the nopanic theme? or the simple also? May 10 18:46:34 it automatically build nopanic while building the base thingie May 10 18:46:38 but not the simple one May 10 18:46:47 mrmoku: i changed dontpanic theme, and bb files for all shr-splash packages May 10 18:46:56 mrmoku: i know, but opkg install simple when installing base :x May 10 18:47:19 s/install/installs/ May 10 18:47:19 dos1 meant: mrmoku: i know, but opkg installs simple when installing base :x May 10 18:51:07 mrmoku: please ping me when it's done May 10 18:51:43 dos1: ok... splashes are there May 10 18:51:49 mrmoku: thanks :) May 10 18:51:59 so... sending mail to shr-user May 10 18:53:56 any OE guru around ? what is "S_pn-something = ${WORKDIR}/${PN}" (i.e, what is workdir, S and PN ?) May 10 18:54:22 I don't know what to put in local-builds.inc, it keeps failing May 10 18:54:42 complaining about not finding any valid SCM May 10 18:54:52 Deubeuliou: no guru... though... S is the location of the source May 10 18:55:03 PN is the name of the package (eg. libframeworkd-phonegui-gtk2) May 10 18:55:20 and WORKDIR is tmp/work/[arch]/PN May 10 18:55:30 with version number... May 10 18:55:53 the src/ subdir ine the package dir or the package's dir ? May 10 18:56:42 so S, for shr-apps *must* be ${WORKDIR}/shr-apps/${PN} ? May 10 18:57:24 ERROR: SRCREV was used yet no valid SCM was found in SRC_URI May 10 18:58:45 Deubeuliou: (regarding S) yeah, correct May 10 18:59:11 (regarding the error) that you get, when you have a SRCREV set, but the SRC_URI does not point to some scm May 10 18:59:23 ok. and for other ? (custom vala-lang , i.e) May 10 18:59:48 vala-lang? May 10 18:59:49 I did not set SRCREV anywhere May 10 19:00:06 and SRC_URI points to a local directory May 10 19:00:08 it used to work May 10 19:00:10 for vala SRCREV is set in the shr overlay May 10 19:01:16 but I override it May 10 19:01:17 wow... my mail from two days ago made it to e devel list ;) May 10 19:01:19 quit fast :( May 10 19:01:21 +e May 10 19:01:36 SRC_URI_pn-vala-lang = "file:///home/deubeuliou/dev/vala/vala-lang" May 10 19:01:36 SRCREV_pn-vala-lang = "LOCAL" May 10 19:01:37 S_pn-vala-lang = "${WORKDIR}/${PN} May 10 19:02:02 oh, wait. I should remove the inheritance from "shr" in the BB recipe ? May 10 19:02:07 Deubeuliou: vala-lang? May 10 19:02:13 it is called vala or vala-native May 10 19:02:13 yes May 10 19:02:17 mmmh May 10 19:02:28 the repo on fso.org is called vala-lang May 10 19:02:32 yes May 10 19:02:40 but the package is named vala-native May 10 19:02:46 mmh ok ! May 10 19:03:27 if it is built as local, it will override the "standard" package, right ? May 10 19:08:15 Deubeuliou: local means getting the source from a local place... May 10 19:08:25 in that sense... yeah it overrides the standard package May 10 19:08:38 ok. the error message is gone now May 10 19:09:36 ANNOUNCEMENT: do you have nice splash screen concept? Do you have nice gimp skills? Do you just want to try? Send 480x640 PNG file to shr-user maillist! We are waiting for your splash! :) May 10 19:09:37 :D May 10 19:09:50 but it looks like it ignores the local files and fetch them from a repo :/ May 10 19:10:13 dos1: are animated splashes supportes ? May 10 19:10:22 s/tes/ted/ May 10 19:10:23 Deubeuliou meant: dos1: are animated splashes supported ? May 10 19:10:25 Deubeuliou: no. we don't want to slow down booting May 10 19:10:29 :) May 10 19:10:34 dos1: Should the splash display "SHR" or something? Should it have text or not? May 10 19:10:59 thrashold: as you want, but i think Ainulindale would prefer SHR somewhere :D May 10 19:11:16 I'm a bit tired of (not) coding and waiting for stuff to compile, I might spent some time playing in GIMP :) May 10 19:13:10 what package has desktop files for shr? May 10 19:13:48 it would be great to have a gui ipkg-tools May 10 19:14:55 ooi, is there anyone (anywhere in the OpenEmbedded world) working on a better package manager than opkg? May 10 19:15:42 usr: there's a gtk one in gpe May 10 19:18:38 i don't know about gpe... but i thought one should write an ide for py-gtk & openmoko, wich includes gui for ipkg-tools May 10 19:18:57 should be based on python and py-gtk May 10 19:19:21 it could be a module for shr-settings, I thought asu had something like that? May 10 19:19:53 tmzt: ipkg-tools != opkg May 10 19:19:57 you mean assasin May 10 19:20:01 yeah May 10 19:20:10 was it a wrapper for the cli tools? May 10 19:20:30 but's not for creating ipkg's? May 10 19:20:42 tmzt: assasin is package manager which uses packagekit May 10 19:20:45 with opkg backend May 10 19:20:58 it's for installing, removing and updating, not for creating packages ;p May 10 19:21:01 oh, so usr is asking about a programming tool, I see May 10 19:21:43 yes for a complete ide May 10 19:21:55 dos1: what causes frameworkd to start the phonegui libary? how does it get DISPLAY? May 10 19:22:20 tmzt: frameworkd does *not* start phonegui library May 10 19:22:20 tmzt: it is not frameworkd. it is ophonekitd, which is started via Xsession.d May 10 19:22:28 oh, ok May 10 19:22:31 it is started by ophonekitd (which gets started by a x session script) May 10 19:23:03 where is the rasterman?! ;] May 10 19:23:12 damn, he's not around lately! ;[ May 10 19:23:22 he used to be here 24/7 ;[ May 10 19:24:07 ~seen raster May 10 19:24:07 raster was last seen on IRC in channel #openmoko-cdevel, 1d 6h 45m 3s ago, saying: 'brb'. May 10 19:24:23 TAsn: can wait... he will be right back ;) May 10 19:24:36 is there anyone working on an ide for py-gtk? May 10 19:25:02 usr: we don't need ipkg-tools May 10 19:25:19 usr: we are using bitbake to generate packages May 10 19:25:47 gui-based? May 10 19:26:06 nah... nothing gui ;) May 10 19:26:26 so, that's what i tought about... May 10 19:26:29 mrmoku, ;] May 10 19:27:03 usr: what do you need a gui for to build packages? May 10 19:27:16 to develop the source... maybe... but to build? May 10 19:27:43 it would be much easier for newcommers if there is an gui-based ide to creat new apps and generate packages May 10 19:31:14 * PaulFertser just changed a thermo-fuse in supply and now his USR Sportster almost works, just for some reason stuck in the bootloader :-/ May 10 19:32:09 i want all the functionality described in the pygtk-tutorial + automated integration for fso-dbus + generation of packages... in a gui-based ide May 10 19:34:05 would one like to write such an ide? May 10 19:34:55 mrmoku: peng May 10 19:37:40 dos1: is ophonekit all in vala? that's the current version in git? May 10 19:37:48 http://git.shr-project.org/git/?p=ophonekitd.git;a=tree May 10 19:37:58 tmzt: no, it's in C, but there is procedding rewrite in vala May 10 19:38:14 do you have a gitweb link? May 10 19:38:19 for the c version May 10 19:38:34 tmzt: it's in shr.git repo May 10 19:41:22 BluesLee: yo May 10 19:41:33 "Modem Code Corrupt" arghhh May 10 19:41:37 mrmoku: hi May 10 19:42:06 BluesLee: shoot ;) May 10 19:42:07 mrmoku: i am running shr testing since 2 weeks now and overall behaviour is good May 10 19:42:52 mrmoku: there are some minor things May 10 19:42:54 BluesLee: nice to hear :) May 10 19:43:36 mrmoku: if i dont shutdown the device for a longer time gsm seems to deregister, i get a "sos only" May 10 19:44:07 BluesLee: longer time like in some hours... or some days? May 10 19:44:15 mrmoku: some days May 10 19:44:27 mrmoku: 1-2 days May 10 19:44:38 BluesLee: hmm... guess FSO guys would be interested in a debug log for that :P May 10 19:45:00 mrmoku: normally there is no problem as i turn off the device in the night May 10 19:45:01 yup May 10 19:45:29 mrmoku: but i have to recharge the device and its only possible when a system is up, unfortunately May 10 19:45:53 i used the gsm for >4days without shutdown or breakage May 10 19:46:02 * mrmoku too May 10 19:46:04 sounds good May 10 19:46:21 it happened twice to me in the last week May 10 19:46:29 the other issue May 10 19:46:47 is for an incoming call May 10 19:47:03 i accept it but for 2-3 seconds nothing happens May 10 19:47:27 this we all have :( May 10 19:47:28 i only hear the chosen phoen ring sound May 10 19:48:08 i had a 20sec delay few days ago. That's really been scary May 10 19:48:22 one more is the "gsm timezone", its local -2 for me May 10 19:48:50 BluesLee: well... that you can turn off in frameworkd-conf May 10 19:49:00 where? May 10 19:49:07 exactly May 10 19:49:10 [otimed] May 10 19:49:14 dos1: Nice work on the new splashs. very slick. May 10 19:49:41 zonesources = NONE May 10 19:50:25 mrmoku: should i add [otimed]? its not there May 10 19:50:30 yeah May 10 19:50:59 testing May 10 19:51:20 mrmoku: are there any aims to speed up shr?;-) May 10 19:51:47 hehe, speed is always good :P May 10 19:52:11 that would be great as it is my distribution of choice for now, it has all the apps i need May 10 19:52:28 one hope is to get speed by the frameworkd vala rewrite May 10 19:52:34 BluesLee: you can use x11-16, it provides big speedup of gui May 10 19:52:35 though that is still some way to go May 10 19:52:41 (and introduces uglyness :P) May 10 19:52:56 dos1: in the illume settings? May 10 19:53:28 BluesLee: yep, and in /etc/profile (ELM_ENGINE = x11-16) May 10 19:54:16 [otimed] May 10 19:54:16 zonesources = NONE May 10 19:54:21 dos1: does it work reliably nowadays? May 10 19:54:27 doesnt work May 10 19:54:40 nah... you have to setup your localtime yourself now ;) May 10 19:54:49 (and reboot) May 10 19:55:15 mrmoku: yep, it only crashes when i'm playing too much with non-fitting configs in illume wrench May 10 19:55:39 dos1: which setting it was in illume settings? May 10 19:55:48 dos1: we might think about adjusting our theme then... to look good on 16 May 10 19:56:45 dos1: engine to software_16? May 10 19:59:33 mrmoku: what is 16? May 10 19:59:51 BluesLee: yep, software_16 May 10 19:59:56 yep May 10 20:00:01 mrmoku: what does it mean? reduced colors? May 10 20:00:30 yep 16bit colors May 10 20:01:25 hey, it looks cool here May 10 20:02:19 dos1: i get a ELM-ENGINE not found when i do a ssh May 10 20:02:31 BluesLee: it's env variable... May 10 20:02:39 export ELM_ENGINE=x11-16 May 10 20:03:12 ELM_ENGINE = x11-16 May 10 20:03:12 export PATH PS1 OPIEDIR QPEDIR QTDIR EDITOR TERM ELM_ENGINE May 10 20:03:24 okay May 10 20:08:17 does anyone use pygtk in here? May 10 20:08:52 in python we are using python-elementary May 10 20:08:53 :P May 10 20:09:02 usr: I know it a little May 10 20:09:13 what is the problem ? May 10 20:09:51 i need an ide ^^ May 10 20:11:13 export ELM_ENGINE = x11-16 May 10 20:11:37 gives me -sh: export: line 46: : bad variable name May 10 20:11:57 the neo neglects to boot May 10 20:12:38 usr: uh ? why an IDE for pygtk ? May 10 20:12:53 you mean, for designing the app ? May 10 20:12:55 Glade May 10 20:13:39 because it would be easier to write things just one-time May 10 20:14:10 yes, to write apps May 10 20:14:59 i thought about to write one myself May 10 20:15:22 in python May 10 20:15:29 and pygtk May 10 20:16:36 to get "all" the commands from pygtk and fso-dbus under one hood May 10 20:20:04 rightt now i'm writing an app... and i thought about, what if i'm finish. i keept a few things, but in the end i was just costomizing an example from the pygtk tutorial May 10 20:20:53 mrmoku and dos1: ugly and it doesnt change much for speed issues May 10 20:21:24 the problem is with starting apps, it takes too much to pop them up May 10 20:24:44 so the thing which i thought about is, if i would spend a bit more time to integrate all the things i learn from the pygtk-tutorial in an ide + fso-dbus it would be much easier to click things together the next time i have to use the same May 10 20:25:48 i've choosen pygtk because it's easy to get installed, even on windows May 10 20:25:56 BluesLee: will try it for some day now... see what happens May 10 20:26:01 anyway... off to bed now May 10 20:26:04 gnight all May 10 20:28:40 does anyone read what i write? as an example if i would have to create a button for my app. i could write a script which does creat a button sth. like that May 10 20:31:14 mrmoku|away: gn8 May 10 20:33:59 when i finish the app i am at right now and i'll have to create an ipkg, i'll write a gui, where you easily can fill in all the parameters and press the button generate. in pygtk for shure... May 10 20:40:25 usr: have you looked at glade, anjunta, GtkBuilder? May 10 20:41:03 at least glade.... May 10 20:42:06 but it does use c bindings or sth. like that right? May 10 20:43:37 is there an ide written in python and pygtk for pygtk? May 10 20:44:31 which rotate app should i use with shr? May 10 20:45:12 Anjuta is a versatile Integrated Development Environment (IDE) for C and C++ May 10 20:45:25 wie alt wird er denn? May 10 20:45:44 wouldnt it be a good idea to use the aux button for things like "lock screen" (as before) or as a menue also for rotation May 10 20:46:46 who kicked off the "lock screen" aux button? May 10 20:47:55 usr: it has a gnome-mobile mode I think, maybe one for pygtk2.0 May 10 20:48:43 I just compiled libfso-glib but now, for a package that is depending on it, I get: checking for FSO_GLIB... configure: error: Package requirements (libfso-glib) were not met May 10 20:48:58 in my configure.ac, there is: PKG_CHECK_MODULES(FSO_GLIB, libfso-glib) May 10 20:49:26 it isn't the right name ? or is the problem elsewhere ? May 10 20:50:55 but if u create an app in pygtk it's much better if your ide is written in pygtk. so you can integrate things whilst you are using it May 10 20:54:59 so anyway it would be great to have an ide fpr openmoko, shr, fso and others written in python where you can easily add extensions for related stuff like dbus, etk, gtk etc. May 10 20:56:01 usr: I very much like Vim :) May 10 20:56:22 but I understand May 10 20:56:39 I think we are just not a point where it is a concern May 10 20:56:51 s/a point/at a point/ May 10 20:56:51 Deubeuliou meant: I think we are just not at a point where it is a concern May 10 20:58:54 Ainulindale: ping ? May 10 20:59:26 :Deubeuliou vim? i thought about to use the ide on om to May 10 21:01:31 on the device ? developping on an embedded device sounds challenging May 10 21:01:49 so... it has to be easy May 10 21:02:08 usr: I use SPE here. It's not a full IDE a-la Eclipse, but it works quite well. Other then not being able to connect directly to the phone. That is annoying. May 10 21:02:11 I think it musn't be done May 10 21:02:25 SPE? May 10 21:03:09 tmzt: Stani's Python Editor May 10 21:03:17 oh, ok May 10 21:03:22 is there a hompage for SPE? May 10 21:04:08 schade May 10 21:05:24 * PaulFertser asked my Ukranian friend, he connected via VNC to a windows host, installed modem flashing software, some shareware virtual com port utility, i used socat to share my serial port to TCP and voila, modem reflashed and works now! :) May 10 21:06:50 ptitjes: ping May 10 21:11:30 Deubeuliou: pong May 10 21:12:11 I built libfso-glib but it doesn't seem to have any do_install instruction May 10 21:12:17 spe uses wxwidgets i didn't found all the packages for o, thats why i got rid off Boa Constructor... May 10 21:12:43 so when I try to compile libframeworkd-phonegui-gtk2, it doesn't find the libfso-glib package. May 10 21:12:55 Deubeuliou: it uses autotools so it doesn't need any May 10 21:13:07 I tried to add inheritance to the rules "autotools autotools_stage" but it fails May 10 21:13:26 inherit pkgconfig May 10 21:13:30 already done May 10 21:13:41 I mean, it already had May 10 21:13:53 PaulFertser: you're talking weird, no? :-D May 10 21:13:58 Check for the directive in your staging dir May 10 21:14:10 I don't remember if it's supposed to be libfso-glib or fso-glib May 10 21:14:16 DocScrutinizer: well, i wanted my modem to work, i needed to reflash it. And software is windows only. :) May 10 21:14:27 the BB file is named libfso-glib May 10 21:14:32 but I'll check May 10 21:14:36 No link with what I'm saying. May 10 21:14:50 I was talking about packaged libraries and configure names. May 10 21:14:59 PaulFertser: aaaaah :-) May 10 21:15:00 DocScrutinizer: without the modem i can't receive faxes at home :) May 10 21:15:05 it seems to be libfso-glib-0.0.1 May 10 21:15:44 PaulFertser: geeky setup anyway May 10 21:16:08 right story for a blog ;) May 10 21:16:25 Ainulindale: which directory am I to check ? May 10 21:16:28 DocScrutinizer: (geeky) that's why i shared it here despite being highly off-topic May 10 21:16:30 Anyone know how to remove/replace the contents of an elementary.Frame widget? May 10 21:16:52 PaulFertser: maybe you can buffle your students with that? May 10 21:16:57 I tried simply adding another content object, but they just get layered May 10 21:17:23 Ainulindale: this?: ./tmp/work/armv4t-angstrom-linux-gnueabi/libfso-glib-0.0.1-gitr22+28f8a574abd532fd8d6b0cd0800f44583813256b-r0/git/fso-glib-1.0.pc.in May 10 21:18:22 PaulFertser: please report on percentage of students even able to understand the geek brilliance of this hack :-) May 10 21:18:49 DocScrutinizer: i'm sure it'll be about 90% as i'm good at explaining :) May 10 21:18:59 hehe May 10 21:19:53 probably 89% will ask "why didn't just use windows?" right away May 10 21:19:59 Deubeuliou: Put fso-glib instead of libfso-glib May 10 21:20:46 Judging by mad noise outside (it's 01:20am here) Russians won the great hockey game May 10 21:20:47 that's what I did, indeed. retrying May 10 21:21:14 Ainulindale: the same (No package 'fso-glib' found) May 10 21:21:37 PaulFertser: The Russians won? Damnit! May 10 21:22:04 Toaster`: :P May 10 21:22:46 PaulFertser: well we can't win the all the hockey titles all the time... :( May 10 21:24:05 Toaster`: in fact i don't care much about professional sport. Just was surprised to hear so many honking cars so late. May 10 21:24:31 Professional sport is business. Quite unhealthy and unethical one. Why should we care... May 10 21:25:07 Ainulindale: on my desktop, the package's name found by pkg-config is "libfso-glib" May 10 21:27:01 Ainulindale: retrying with libfso-glib-0.0.1 May 10 21:27:47 Ainulindale: you have any experience with elementary.Frames? May 10 21:32:56 so here is an idea how an ide could work... as i'm not that into hacking i would write it in simple modules. one module could just create the code for a button and echo all this into a file named after the button. afterwards you read the file echo it and among different rules (which do the framework) into the final-script... is this a good idea? May 10 21:43:21 hmm May 10 21:43:25 what about my idea? would anyone like to write modules if i have created the framework for this ide? May 10 21:43:27 dos1: elm question for you May 10 21:43:48 what about May 10 21:43:51 wrr May 10 21:44:16 what about IM app using elementary and telepathy? May 10 21:44:53 * dos1 checks, how telepathy framework looks like May 10 21:45:03 Toaster`: ask :) May 10 21:45:04 how can I replace the contents of a frame/scroller? If I call .content_set a second time it layers the two. reading the source, suggests that it's supposed to delete the existing first, but I think it's not doing that. May 10 21:45:16 it would be nice if the tp and fso concept of channels could be combined May 10 21:46:10 Toaster`: dunno... i think we should ask raster May 10 21:46:27 ~seen raster May 10 21:46:27 raster was last seen on IRC in channel #openmoko-cdevel, 1d 9h 7m 23s ago, saying: 'brb'. May 10 21:46:33 but he is not here ATM... May 10 21:46:56 Toaster`: have you tried just "del old_content"? May 10 21:47:15 hrm.. nope. lets see... May 10 21:47:57 nope, no joy May 10 21:51:44 Ainulindale: according to http://git.shr-project.org/git/?p=ophonekitd.git;a=blob_plain;f=configure.ac;hb=HEAD , the package name is "fso-glib-1.0" May 10 21:51:48 but May 10 21:52:08 using "find" on that name only returns one line: ./tmp/work/armv4t-angstrom-linux-gnueabi/libfso-glib-0.0.1-gitr22+28f8a574abd532fd8d6b0cd0800f44583813256b-r0/git/fso-glib-1.0.pc.in May 10 21:52:23 since finding "libxml-2.0" returns more May 10 21:52:54 this one, for instance: ./tmp/staging/armv4t-angstrom-linux-gnueabi/usr/lib/pkgconfig/libxml-2.0.pc May 10 21:54:51 or ./tmp/work/armv4t-angstrom-linux-gnueabi/libxml2-2.7.2-r0/libxml2-2.7.2/libxml-2.0.pc May 10 21:59:21 I got something good cooking here ;] May 10 21:59:31 more rtl support for e17 ;] May 10 21:59:55 it's more than usable for rtl users ! ;] May 10 22:17:08 dos1: Any idea how to limit a scroller to only the vertical axis? May 10 22:18:46 Toaster`: it was limited, the newest EFL update broked that May 10 22:18:49 ;x May 10 22:18:54 freesmartphone.org: 03mickey 07cornucopia * rc6f066eaf035 10/docs/frameworkd.conf.sample: docs: document current state of frameworkd.conf May 10 22:18:55 freesmartphone.org: 03mickey 07cornucopia * r08db9ea98922 10/fsodeviced/src/plugins/kernel26_powersupply/plugin.vala: fsodevice: kernel26_powersupply: compute status if we're AC-powered May 10 22:18:55 freesmartphone.org: 03mickey 07cornucopia * ra2a8359b0e50 10/libfsoframework/fsoframework/ (fsoframework-2.0.vapi interfaces.vala): fsoframework: interfaces: add org.freesmartphone.Device.PowerSupply.GetType() -> s May 10 22:23:31 dos1: arg... ok. May 10 22:23:43 that would explain why everything is bouncing around May 10 22:26:16 and probably why the weights/expansion of widgets is all funny May 10 22:58:06 yay, will apply the patch soon! ;] May 10 22:59:01 TAsn: will it also fix the line break issues? :) May 10 22:59:16 don't think so May 10 22:59:21 darn May 10 22:59:22 I mean, it could be a cool side effect May 10 22:59:31 though it's not a wanted feature ;] May 10 22:59:42 *intended May 10 23:00:01 agreed May 10 23:10:51 arg... how does one tell git to revert a modified file back to the current version? May 10 23:12:35 co ? May 10 23:12:41 not sure though May 10 23:12:46 * TAsn <-- git noob. May 10 23:13:43 heh. Well I can figure out how to revert a commit, but I'd rather not have to make the commit if I don't have to. May 10 23:28:15 SHR: 03tom 07shr-overlay * r6017fcd2c91a 10/patches/efl1/0001-rtl-cursor-patch-evas.patch: added the rtl cursor patch for evas May 10 23:31:35 SHR: 03frazier.cameron 07shr-settings * r5a00607a51c7 10/playground/settings_toolbar.py: [Playground] Attempt to create a toolbar version of shr-settings, please test and comment May 10 23:31:35 SHR: 03frazier.cameron 07shr-settings * r56bedc17a1b5 10/shr_settings_modules/shr_wifi.py: [Wifi] General cleanup, added ability to call wifiman.py (currently commented out) May 10 23:31:36 SHR: 03frazier.cameron 07shr-settings * rb782ded3eb48 10/shr_settings_modules/shr_currentprofile.py: [CurrentProfile] Added filter to sound files listing to remove directories May 10 23:31:37 SHR: 03frazier.cameron 07shr-settings * r3c61d8396fe4 10/shr_settings_modules/shr_currentprofile.py: [CurrentProfile] Corrected a filtering error May 10 23:31:45 really?! ;] May 10 23:31:48 coooool!!! May 10 23:31:54 SHR: 03frazier.cameron 07shr-settings * r59f2319540de 10/shr_settings_modules/shr_bt.py: [Bluetooth] Removed need for timers May 10 23:31:55 we got shr-wifi ;] May 10 23:32:01 bulid the damn package! ;] May 10 23:32:26 TAsn: It's the standard shr-settings wifi file May 10 23:32:41 oh, sucks. May 10 23:32:46 :) May 10 23:33:04 I've just added in some convience stuff for wifiman.py, which is a very alpha connman manager May 10 23:33:17 Toaster`, yeah, tested it before May 10 23:33:18 it's nice. May 10 23:33:32 bah, anyhow, I'm off, gn. ;] May 10 23:33:35 already 2:23 ;\ May 10 23:33:37 Night May 10 23:33:38 time flies. May 10 23:35:39 omg! I just notice raster was here and I missed him! ;[ May 10 23:38:10 he was in/out in less than a minute IIRC May 10 23:41:14 Toaster`, this only means one thing, I'm not as sharp as I used to be May 10 23:41:20 I used to be a good raster hunter ;\ May 10 23:53:15 SHR: 03frazier.cameron 07shr-settings * redad304d6e1d 10/playground/ (settings_toolbar.py shr-settings-toolbar): [Playground] Renamed settings_toolbar.py to shr_settings_toolbar, and fixed it to work May 10 23:53:17 Toaster`: "git checkout somefile" should do what you want May 10 23:54:17 Weiss: Thanks. I did the brute force method of saving it from the web git interface, overwriting the modified version. Thank you though, I'll remember that for next time May 10 23:55:01 ick, definitely the hard way :) May 10 23:55:35 the meaning of "checkout" can be difficult if you've come from a different VCS. in git, it means "generally transfer a file from the repository to the working copy" May 10 23:56:05 hence it's used for checking out a branch, switching branch and reverting changes May 10 23:56:47 ok. Well my background in VCS' is none, so I appreciate the lesson :) May 10 23:58:35 no problem :) May 11 00:06:29 why does ophonekitd bind to org.shr.ophonekitd.Usage? May 11 00:08:26 tmzt: it has to do with the way the org.freesmartphone.Usage doesn't handle scripts; rather it expects daemons, of which things like shr-settings is not. May 11 00:08:36 At least that's how it was explained to me :) May 11 00:09:29 I would like to disable ousaged completely for now, I guess that's not going to work May 11 00:09:46 What are you trying to do? May 11 00:09:55 port shr to touch pro May 11 00:10:12 which has very different sysfs layout from fr kernel May 11 00:10:24 ahh ok May 11 00:11:10 IIRC only the auto-suspend/auto-dim controls in shr-settings would be affected May 11 00:11:49 ok, I'm going to try frameworkd -s ogsmd,ousaged and see if that works May 11 00:28:24 does somebody know the clone url for shr.git? May 11 00:28:45 tmzt: May 11 00:28:51 tmzt: which part? May 11 00:28:57 the whole thing May 11 00:28:59 shr.git May 11 00:29:15 http://shr.bearstech.com/repo/shr.git <-- That? May 11 00:29:30 no, http://git.shr-project.org/git/?p=shr.git;a=summary May 11 00:29:54 eh, I was close May 11 00:31:12 oh, it's related to sim card May 11 00:31:23 like everything else that doesn't work so far May 11 00:32:21 so the question is should I try to make the c version work on cdma or the vala one? May 11 00:34:19 20 phonegui_sim_auth_show = phonegui_get_function("phonegui_sim_auth_show"); May 11 00:34:22 21 phonegui_sim_auth_hide = phonegui_get_function("phonegui_sim_auth_hide"); May 11 00:38:07 is quickdev the only one familiar with ophonekitd? May 11 00:38:36 no, why did I ask that, this file has everybody's name on it May 11 00:38:44 If you want help with shr-settings I can help, anything else.. :( May 11 00:39:07 what I'm wondering is how does the main loop actually start all these dialogs, but I'll just read through it for a while **** ENDING LOGGING AT Mon May 11 02:59:57 2009