**** BEGIN LOGGING AT Thu Dec 24 02:59:57 2009 Dec 24 03:02:10 any apps use that data yet? Dec 24 03:06:53 not that i know of. they probably don't have enough data. someone should come up with something though Dec 24 03:07:17 it would probably help its popularity. Dec 24 08:05:56 lindi-, do you have more info on the suspend failed bug with gsm failed to suspend? Dec 24 08:07:12 i experience that often now, running shr-testing from nand Dec 24 08:28:39 hi Dec 24 08:28:52 any one has idea about apmd in ubuntu Dec 24 08:59:02 satish: it's easier to simply ask what you want Dec 24 09:03:54 sybren : i didn't get u Dec 24 09:30:24 satish: you're asking whether we know about apm in ubuntu. It's easier just to ask what you want to know about apm in ubuntu, and see who answers. Dec 24 10:04:58 spaetz: ping Dec 24 10:09:37 morning channel Dec 24 10:09:46 is it possible to preload python-elementary? (c_elementary.so) Dec 24 10:09:55 so that python elementary apps starts up faster? Dec 24 10:10:00 *start Dec 24 10:19:07 good morning daniele_athome Dec 24 10:19:18 (sorry, no idea about preloading) Dec 24 10:22:26 I seem to remember a python starter developed 1.5years ago. The idea aiui was to have *one python interpreter that loads all the foo once and starts the apps then with preloaded 'environment' Dec 24 10:24:51 hi sybren Dec 24 10:25:04 DocScrutinizer51, a sort of centralized python executable Dec 24 10:25:14 that loads all python applications? Dec 24 10:25:20 yep Dec 24 10:25:27 but if python segv... Dec 24 10:25:31 every app is killed :D Dec 24 10:25:43 yep Dec 24 10:27:18 still sounds like a bloody good thing to have, starting Python takes long Dec 24 10:27:29 sybren, actually importing library is the longest issue Dec 24 10:27:41 the tiny python interpreter starts in a very short time Dec 24 10:27:44 0.19 ms or so Dec 24 10:28:07 I mean, Cython with no extensions embedded Dec 24 10:28:21 about 250kb Dec 24 10:33:41 daniele_athome: "python < /dev/null" takes 770ms on my Neo Freerunner Dec 24 10:34:27 sybren, tried now on mine Dec 24 10:34:30 0.56 ms Dec 24 10:34:32 2.6.29 kernel Dec 24 10:34:35 shr-u Dec 24 10:34:42 same here Dec 24 10:35:21 after a few starts it averages at around 0.8 seconds wall clock time Dec 24 10:35:27 definitely not 0.5 ms Dec 24 10:35:28 but try with -S Dec 24 10:35:35 I know that is useless Dec 24 10:35:38 without site Dec 24 10:35:46 but it's some start... Dec 24 10:35:55 that's around 0.3 seconds, still not half a millisecond Dec 24 10:36:29 0.19 ms or so Dec 24 10:36:36 sorry I was wrong Dec 24 10:36:40 not ms but s Dec 24 10:36:49 I thought as much ;-) Dec 24 10:36:49 i mistyped Dec 24 10:37:06 and there I was hoping that Python could one day start in under a millisecond... Dec 24 10:37:12 *snif* Dec 24 10:37:15 T_T Dec 24 10:38:42 sybren, by the way I need to try python on 2.6.31 Dec 24 10:38:52 because shr boot time was halfed on that kernel Dec 24 10:38:57 nice Dec 24 10:38:59 maybe python also could starts faster Dec 24 10:51:53 guys in e-wm there a illume2 module, what is it? Dec 24 10:59:43 sybren, python with -S is 150ms Dec 24 11:20:43 sooner or later we won't need python for the core stuff, so things will fly eventually ;) Dec 24 11:21:33 mickey|XmasCold, why not? Dec 24 11:21:42 I mean why not focusing on making python faster Dec 24 11:21:47 instead of porting everything to C? Dec 24 11:22:03 I know that python is already fast the way it is Dec 24 11:22:14 and there is no way using the actual implementation to make it faster Dec 24 11:22:29 but one could implement python in another, lighter way Dec 24 11:22:51 given the outlook, rewriting it in compiled language (not necessarily C) is less work Dec 24 11:23:03 trust me on that, i know the domain and have all eyes on things like pypy and jit etc. Dec 24 11:23:10 but i know how long these guys have been working on that Dec 24 11:23:15 and know the prospectives Dec 24 11:23:42 so all in all we will benefit more from the port, since it's also not just about speed Dec 24 11:23:52 but also about memory footprint, security, etc. Dec 24 11:24:35 last but not least, i enjoy coding in Vala :) Dec 24 11:25:03 mickey|XmasCold, eh... Dec 24 11:28:35 wrt. illume2, i think that's the supersecret stuff raster is working on for some company Dec 24 11:28:57 lol Dec 24 11:29:11 mickey|XmasCold, it's supersecret and he puts it on the public repository? :D lol Dec 24 11:29:34 hehe Dec 24 11:29:37 i think it's a good deal Dec 24 11:29:41 he's not allowed to talk about it Dec 24 11:29:46 but he's allowed to code on it in public Dec 24 11:29:52 better than otherwise IMO Dec 24 11:30:55 twice lol Dec 24 11:30:56 :D Dec 24 11:33:48 daniele_athome: that looks quite a bit faster - assuming that's really milliseconds ;-) Dec 24 11:34:00 sybren, yes they are :P Dec 24 11:34:46 illume2 is restructurised illume Dec 24 11:34:54 divided into modules Dec 24 11:35:11 and with cleaner code :P Dec 24 11:35:25 AFAIR what raster was explaining :D Dec 24 11:36:29 dos1, talking about illume2 UI, doesn't change much Dec 24 11:36:32 does it? Dec 24 11:36:48 daniele_athome: no, i don't think UI will really change Dec 24 11:36:52 (dos1, any news about the window key_up callback?) Dec 24 11:37:00 but it will be easier to change illume2 UI than illume1 :P Dec 24 11:37:06 dos1, that's sad :( Dec 24 11:37:13 so, for instance, you'll be able to use your own launcher instead of bundled one Dec 24 11:37:20 i will have to write my own UI from scratch, or modify illume2 Dec 24 11:37:38 I only hope they adapt the config dialogs for small screens Dec 24 11:39:12 dos1, do you remember? Dec 24 11:39:20 yup Dec 24 11:39:31 but well, i failed to do that with standard way Dec 24 11:39:45 oh... Dec 24 11:39:45 but i think it could be possible using hackish way ;d Dec 24 11:39:52 dos1, that is? Dec 24 11:39:56 using focus on a widget? Dec 24 11:40:02 use one big elementary.Layout Dec 24 11:40:13 and use SWALLOW to display your UI in it Dec 24 11:40:22 and then listen to signals of this Layout Dec 24 11:40:26 that will work for sure :D Dec 24 11:40:48 dos1, ok.. better than nothing :) Dec 24 11:40:53 thanks Dec 24 11:41:20 hmm Dec 24 11:41:38 ~lart TI Calypso not being able to switch from MUX mode back to AT mode without powercycling Dec 24 11:41:39 * apt shoots TI Calypso not being able to switch from MUX mode back to AT mode without powercycling in his sleep Dec 24 11:42:19 that's really obnoxious when developing Dec 24 11:42:21 *sigh* Dec 24 11:43:04 mickey|XmasCold, is it a hardware issue? Dec 24 11:44:45 hardware or firmware, yes Dec 24 11:44:48 well Dec 24 11:44:50 rather firmware Dec 24 11:44:59 but something we can't change Dec 24 11:45:04 easily Dec 24 11:45:10 mickey|XmasCold: What would that feature be used for? Dec 24 11:45:18 well Dec 24 11:46:04 the problem is that on shutdown of the MUX session Dec 24 11:46:07 whereas all modems fall back to AT Dec 24 11:46:15 the Calypso falls into a coma Dec 24 11:46:19 and does not answer anymore Dec 24 11:46:24 neither to MUX nor AT commands Dec 24 11:46:48 so i will have to patch out the "disconnect from MUX" for the calypso while developing Dec 24 11:46:59 which might work, but is somewhat annoying Dec 24 11:47:15 mickey|XmasCold, does this issue affect performance? Dec 24 11:47:21 no, not at all Dec 24 11:47:23 I mean power cycling takes some time Dec 24 11:47:26 ok Dec 24 11:47:29 it's acutally completely irrelevant for the users Dec 24 11:47:40 since typically you don't shutdown the mux session at all Dec 24 11:47:45 once it's established Dec 24 11:47:55 so it's just cumbersome for the developer :) Dec 24 11:48:06 which would be... me Dec 24 11:48:48 since I'm working on the TI Calypso modem abstraction next in fsogsmd Dec 24 11:52:48 mickey|XmasCold, but when I have to send modem commands Dec 24 11:52:53 should I have to switch to AT mode? Dec 24 12:01:30 daniele_athome: you did ogpsd stuff? Dec 24 12:03:22 lindi-, no gsm Dec 24 12:03:32 i'm talking with mickey|XmasCold about calypso modem Dec 24 12:03:41 daniele_athome: yes yes but in 2008? Dec 24 12:03:55 lindi-, i don't get you Dec 24 12:04:00 or was that alphaone|gone? Dec 24 12:04:05 daniele_athome: no, you can send AT fine in MUX mode as the libgsm0710mux is transparently converting this into the binary format Dec 24 12:04:23 daniele_athome: ah, daniele != Daniel Willmann? Dec 24 12:04:44 lindi-, yes :) Dec 24 12:04:46 i'm not willmann Dec 24 12:04:48 mickey|XmasCold, ok Dec 24 12:04:50 daniele_athome: ok Dec 24 12:05:12 mickey|XmasCold: calypso supports some binary protocol? Dec 24 12:05:22 * mickey|XmasCold still hopes for a libublox from Daniel... Dec 24 12:05:31 i was asking exactly about that Dec 24 12:05:34 lindi-: 07.10 Dec 24 12:05:40 mickey|XmasCold: ah that one Dec 24 12:06:04 like i complained above, you just can't shut it down properly Dec 24 12:06:07 foobared in firmware Dec 24 12:06:12 *shrug* Dec 24 12:07:05 mickey|XmasCold: btw, if resume reason is GSM it would be nice if muxer could tell me the channel that caused the resume Dec 24 12:07:27 mickey|XmasCold: this way if I get GPRS data I can just go back to suspend immediately and save energy Dec 24 12:07:54 yes, makes sense Dec 24 12:09:23 mickey|XmasCold: also, what's the status of that fso watchdog thing? i just realized I had to do some changes to my own quick'n'dirty watchdog script and thought they might be useful ideas for fso too Dec 24 12:09:54 playya did lots of work on fso-monitord, but I'm afraid it's too huge for my kind Dec 24 12:10:01 i planned something much more lightweight Dec 24 12:10:16 mickey|XmasCold: (basically just polling for AT+COPS? is not enough. I also need to poll AT+CPIN? or else the SIM might sometimes decide that the PIN needs to be resent and all calls fail until you do that) Dec 24 12:10:17 so i'm afraid we will rip the watchdog functionality out Dec 24 12:10:25 and put it into a small project Dec 24 12:10:43 ah Dec 24 12:10:44 well Dec 24 12:10:58 that's more fsogsmd watchdog than fso watchdog Dec 24 12:11:20 mickey|XmasCold: ok, what's the state of that? Dec 24 12:11:57 no work done, but it's going to be present in fsogsmd Dec 24 12:12:05 an optional watchdog thread Dec 24 12:12:41 i'm not feeling like introducing additional stuff in ogsmd Dec 24 12:12:54 but fsogsmd is another story Dec 24 12:13:03 mickey|XmasCold: ok, if you have some todo list you might want to add AT+CPIN? polling there Dec 24 12:13:13 mickey|XmasCold: and let me know if you find a way around that :) Dec 24 12:13:17 hmm Dec 24 12:13:29 i wonder about the conditions when this can happen at all Dec 24 12:13:39 I have here a log file that shows how PIN is entered and after 7 hours all calls fail Dec 24 12:13:41 the only thing that can make the modem reset the SIM is a +CFUN command Dec 24 12:14:01 and it took me a long time to figure out that setting PIN again makes it possible to answer calls again Dec 24 12:14:03 that's interesting. i'd like to see that log Dec 24 12:14:10 awesome Dec 24 12:14:12 scary though Dec 24 12:14:23 uploading Dec 24 12:14:51 mickey|XmasCold: also note that you need to have periodic RTC wakeups to poll for that Dec 24 12:15:44 *nod* Dec 24 12:15:47 which frequency are you using? Dec 24 12:17:23 mickey|XmasCold: i have not implemented it with RTC yet. when the device is running its every 2 minutes Dec 24 12:24:43 mickey|XmasCold: https://docs.openmoko.org/trac/ticket/2323 Dec 24 12:25:11 thanks Dec 24 12:27:02 lindi-: hmm Dec 24 12:27:13 lindi-: were you on battery power? Dec 24 12:30:29 anyways a very interesting log Dec 24 12:30:53 we need to seamlessly recover the modem going down and up anyways Dec 24 12:31:15 at least it didn't went down completely otherwise it would have relaunched in AT mode Dec 24 12:31:54 mickey|XmasCold: yes battery power Dec 24 12:32:29 fun Dec 24 12:32:31 mickey|XmasCold: see $ grep resuming syslog |nc paste.dyndns.org 1234 Dec 24 12:32:31 http://paste.debian.net/54820/ Dec 24 12:32:57 mickey|XmasCold: that 13-20 mA consumption indicates that it was on battery power Dec 24 12:33:03 ok, but definitely not low battery Dec 24 12:33:42 mickey|XmasCold: nope, this will automatically shutdown itself gracefully on low battery Dec 24 12:34:02 starts to warn at 20% and shutdowns at 3% Dec 24 12:34:37 with nice synthesised "battery almost empty" voice :) Dec 24 12:35:06 hehe Dec 24 12:36:48 * mickey|XmasCold would appreciate doing some UI level coding again Dec 24 12:37:05 don't see that coming for at least another 6 months though Dec 24 12:42:53 heh Dec 24 12:43:15 mickey|XmasCold: how easy it would be possible to use the new vala-rewritten tools to create a really simple gsmd that does not need dbus? ;) Dec 24 12:50:18 basically you would need to write a new plugin that does not forward the dbus calls to mediator invocations, but contains another means to receive commands that then invoce the correct mediators. Dec 24 12:50:46 should be relativey simple Dec 24 12:50:51 take a look at the dbus service plugin Dec 24 12:51:06 you see it's 1:1 mapping dbus calls to mediator invocations Dec 24 12:51:23 now you just need to call the mediators based on your needs Dec 24 12:51:40 http://git.freesmartphone.org/?p=cornucopia.git;a=blob;f=fsogsmd/src/plugins/dbus_service/plugin.vala;h=60e8b71df5eea8e0fc249202f66fff2652764c0b;hb=8dd650901b76ded09b5dc0cd893384b9252e64c0 Dec 24 12:51:50 line 118ff Dec 24 12:52:51 i'd wait until the TI calypso plugin is done though before starting with that :) Dec 24 12:53:11 although you can already use the singleline abstraction with the Calypso Dec 24 12:54:42 i figure this could also be very interesting for modems that e.g. do not need all the configuration, but just call one number, sending some DTMF, and be done Dec 24 12:54:48 M2M Dec 24 12:56:04 mickey|XmasCold, hey, I have a question/semiproblem with frameworkd Dec 24 12:56:37 I'm wrting a bluetooth manager and I want fso to power up bluetooth, so I request Bluetooth resource Dec 24 12:57:20 when frameworkd returns the message and I want to start using bluez, though, it isn't yet fully loaded and running Dec 24 12:57:58 so I may ask if this is a wrong behaivour and frameworkd should return after bluez is fully loaded Dec 24 12:58:28 or if it is posible to have some kind of signal for it, or just repeat the call untill bluez responds Dec 24 13:01:10 what exactly is the state of it after fsousaged returns the call? Dec 24 13:01:29 "isn't yet fully loaded and running" means exactly? Dec 24 13:01:48 the semantics is it should be ready for action after the call returns Dec 24 13:02:04 mickey|XmasCold, after requesting it, I call org.bluez / Adapter.GetDefaultDevice or something similar Dec 24 13:02:14 and I get null pointer to dbus rteplymessage Dec 24 13:02:23 ah, darn Dec 24 13:02:27 ok, i think i know what happens Dec 24 13:02:58 FSO turns it on Dec 24 13:03:03 then returns Dec 24 13:03:06 however Dec 24 13:03:12 bluetooth is a USB device on the FR Dec 24 13:03:15 meaning it appears on the bus Dec 24 13:03:16 gets enumerated Dec 24 13:03:20 and registered by bluez Dec 24 13:03:26 only some time after it's powered on Dec 24 13:03:40 yep Dec 24 13:04:17 ok, i'd like to fix that Dec 24 13:04:23 just need to think about the best way Dec 24 13:04:29 so the question is, should frameworkd be aware of this or I should do it in my app? Dec 24 13:05:01 for now I'm afraid you need to do it in your app, like, accepting that there might be a gap until BLuez sees the device Dec 24 13:05:10 I was thinking to just call the method untill it doesn't return a null pointer hehe Dec 24 13:05:18 yeah, not the worst idea Dec 24 13:05:22 just add delay Dec 24 13:05:24 so that you don't hammer it Dec 24 13:05:35 i could imagine that FSO cooperates more tightly with Bluez eventually to fix that Dec 24 13:05:43 so we would not return until bluez sees the device Dec 24 13:05:51 and if it doesn't after a timeout, we correctly report an error Dec 24 13:05:59 hmm how can I make a delay with C? Dec 24 13:05:59 would make more sense from an integration standpoint imo Dec 24 13:06:07 g_usleep() Dec 24 13:06:15 without glib? :P Dec 24 13:06:22 usleep :D Dec 24 13:06:26 xD ok Dec 24 13:06:55 by all means though, please add a bug report with your findings to the FSO bugtracker Dec 24 13:07:01 so that we don't forget this issue Dec 24 13:07:15 ok :) Dec 24 13:07:17 thanks Dec 24 13:12:27 mickey|XmasCold, which component I attach the bug to? fsousaged? Dec 24 13:12:55 i think it's fsodeviced job to correctly handle that Dec 24 13:13:04 as Bluetooth resource is registered in fsousaged by fsodeviced ;) Dec 24 13:13:34 pespin: fsodeviced for now, please Dec 24 13:13:38 dos1 is correct Dec 24 13:13:54 ;) Dec 24 13:17:57 tremendously off topic question: can anybody of you still use thepiratebay.org? I always get a login screen, even when i select "register" Dec 24 13:27:31 mickey|XmasCold, http://trac.freesmartphone.org/ticket/499 Dec 24 13:28:27 dos1, is the theme with shr logo style in the feeds? Dec 24 13:28:36 pespin: right Dec 24 13:28:39 pespin: that one from scap? Dec 24 13:28:46 pespin: yup, e-wm-theme-illume-shrlogo Dec 24 13:29:31 dos1, I just have to use shr-settings to use it, right? :) Dec 24 13:29:44 pespin: no, there is no shrlogo elementary theme yet Dec 24 13:29:51 pespin: you have to select it in illume wrench Dec 24 13:29:54 ok Dec 24 14:50:42 freesmartphone.org: 03mickey 07cornucopia * r113490a4d765 10/docs/ (6 files): docs: add configuration file samples for the new world order... Dec 24 15:12:46 *yawn* Dec 24 15:12:50 merry chrismas Dec 24 15:19:02 ooh yeah. I wonder if I might get some food at turkish or asian restaurant at least Dec 24 15:45:01 freesmartphone.org: 03mickey 07cornucopia * rf18b1ca67349 10/fsogsmd/src/ (4 files in 3 dirs): fsogsmd: introduce notion of 'main' channel Dec 24 15:59:53 mickey|XmasCold, at what point is fsogsmd? Dec 24 15:59:55 :) Dec 24 16:00:30 and what do you think about ofono as a replacement for opimd and fsogsmd? Dec 24 16:00:39 and eventually fsodeviced, since ofono is pluggable... Dec 24 16:04:30 blindcoder: you tried openttd some days ago? Dec 24 16:07:22 daniele_athome: ofono replaces opimd?? hardly Dec 24 16:08:16 DocScrutinizer51, it could Dec 24 16:08:26 is has backend interfaces for messages and calls Dec 24 16:08:34 the fact is there are no backend actually implemented :D Dec 24 16:08:43 *backends Dec 24 16:09:18 So why should a non-working implementation replace a working one? Dec 24 16:09:53 Heinervdm, I just want to know your opinion Dec 24 16:09:59 about a potentially competitor Dec 24 16:10:03 *potential Dec 24 16:10:21 and I underline _potential_ don't get it wrong :) Dec 24 16:10:53 an opinion from a programmer to other programmers Dec 24 16:13:09 * daniele_athome buon natale a tutti!!!! :) Dec 24 16:13:18 sorry :D Dec 24 16:13:27 (damn xchat commands :S ) Dec 24 16:20:33 AME? ;-P Dec 24 16:21:24 DocScrutinizer51, yes :S Dec 24 16:21:39 I thought it sends the message to only one network Dec 24 16:21:41 not to all Dec 24 16:21:41 :S Dec 24 16:26:42 well if u dont like it ->settings, usercmds Dec 24 16:28:36 daniele_athome: (opimd) fso framework is not only already production state. It is actually so much more than ofono even is designed to be eventually Dec 24 16:29:15 opimd also has contacts and callendar, no? Dec 24 16:30:21 DocScrutinizer51, yes it has Dec 24 16:30:36 does ofono? Dec 24 16:30:40 opimd Dec 24 16:30:47 oh Dec 24 16:30:49 A: no Dec 24 16:31:04 ofono has backends interface for messages and calls Dec 24 16:31:11 but I don't remember for contacts and calendar Dec 24 16:31:14 but I guess not Dec 24 16:31:28 DocScrutinizer51, the difference is fso is a complete smartphone framework Dec 24 16:31:29 you said that Dec 24 16:31:48 ofono is just an interface to gsm related things Dec 24 16:31:59 exactly Dec 24 16:32:06 although it has also support for other network methods Dec 24 16:32:11 like wireless management, etc. Dec 24 16:32:17 or more 'will eventually be' Dec 24 16:32:22 yes... Dec 24 16:33:13 k I'm going to walk a while now Dec 24 16:33:17 bbl Dec 24 16:33:33 bye DocScrutinizer51 Dec 24 20:33:47 ? Dec 24 21:57:04 hi, i am having a strange problem, each time i try to register to the network I see "sim failure" ( error code 13 ), and since the card used to work on my freerunner some time ago, and still work fine on other phones, I wonder where the problem is :/ Dec 24 21:57:40 ( it is on debian on sdcard, but I am about to see if the problem is the same on shr and qtmoko ) Dec 24 21:58:10 ( and the modem firmware is moko11, the latest iirc ) Dec 24 22:08:38 ok, on shr, it work Dec 24 23:22:24 Merry Christmas! Dec 25 02:59:10 BillK, pong **** ENDING LOGGING AT Fri Dec 25 02:59:57 2009