**** BEGIN LOGGING AT Thu Dec 02 02:59:58 2010 Dec 02 03:22:09 demarchi: wouldn't "for key in props" work like "for key in list(props.keys())"? Dec 02 03:22:49 you might also need to replace stuff like "% (sys.argv[0])" with "% (sys.argv[0],)" iirc Dec 02 04:00:42 yang_office: i think setlocale() is more reliable than querying LANG because it will query some other standard environment variables as well as LANG if the other ones are not set Dec 02 07:04:01 balrog-k1n: Do you mean I can use setlocale(LC_ALL, NULL) to retrieve the locale info? I tried this and got the returned string "C", which means a POSIX locale. Dec 02 07:04:42 I wonder how to get the format of "[language[_territory][.codeset][@modifier]]", so that I can retrieve then language that follows the ISO-639-1 (two characters) definition. Dec 02 07:05:50 balrog-k1n: By the way, I referred the webpage: http://www.opengroup.org/onlinepubs/009695399/basedefs/xbd_chap08.html Dec 02 10:31:36 balrog-k1n: yes, but it's the suggested way for conversions like this Dec 02 10:31:55 balrog-k1n: because props.keys() does not return a list anymore Dec 02 10:32:46 about the sys.argv conversion, i don't see the point Dec 02 12:12:07 denkenz: hi Dec 02 12:12:53 denkenz: i sent the patches for frequency band selection as a RFC so we can agree about the proper interface Dec 02 12:13:48 denkenz: in order to implement it for huawei (the only modem i have with this feature) i need the radio-settings patch to enter first Dec 02 12:16:08 denkenz: so, i think i'm taking another item from TODO while you review the others. How are you regarding your flu? Dec 02 13:43:24 denkenz: hello, I am taking (via patch mail) the "Add support for GPS power control atom" item from TODO. If somebody else has some work in progress I would like to know. I guess no because that item in TODO does not have owner yet. Dec 02 13:43:53 rafa_: Nobody has been working on that one yet. Dec 02 13:44:01 At least not that I know of. Dec 02 13:44:10 holtmann: ah, great. I will take then. Thanks Dec 02 13:44:24 Send a patch for the TODO file ;) Dec 02 13:44:34 holtmann: yes, I will :) Dec 02 14:20:38 hello all Dec 02 14:24:06 uh oh, was about to ask how to put a conf call on hold Dec 02 14:24:32 but just realized that I guess that's PrivateChat() Dec 02 14:26:52 .. and SwapCalls()... sorry for the noise Dec 02 17:23:38 pessi: So I'm seeing two race conditions with Andras' code Dec 02 17:24:08 if we're going offline and emergency number is dialed, we proceed but then the rx is cut off and the call fails -> oops Dec 02 17:25:23 less fatal one is if we're going online and the emergency number is dialed, that way we set online twice and the call proceeds before the Emergency property is set Dec 02 17:26:04 actually nevermind on the emergency part Dec 02 17:26:30 This also interacts with sim removal Dec 02 17:26:51 have you guys thought about any of these? Dec 02 17:27:56 denkenz/holtmann: i have a comment regarding DisconnectReason() in cdma-voicecall-manager-api.txt Dec 02 17:28:57 yes? Dec 02 17:29:03 it has been defined as a property rather than a signal Dec 02 17:29:37 i'm not sure if we want to keep it as a persistent property, but a one shot signal Dec 02 17:30:13 no it was just an oversight, fixed Dec 02 17:30:31 ahh ok! thank you Dec 02 17:59:53 denkenz: probably not, I'll have a peek on Andras' latest patches Dec 02 19:11:37 denkenz: ping Dec 02 19:11:47 padovan: pong Dec 02 19:11:59 any word on my patches? Dec 02 19:12:54 I'm still not 100%, so doing selective reviews only Dec 02 19:13:15 Lets see if I can look at yours today Dec 02 19:14:05 Ok, great. Dec 02 19:17:16 padovan: So AFAIK gatchat can handle unrefing from within the notify callback Dec 02 19:17:33 Are you g_idle_adding because this isn't so? Dec 02 19:20:37 Yeah, I'm having problem unrefing the gatchat inside the notify callback. Dec 02 19:20:54 It screws up the notify_list Dec 02 19:21:09 I thought I fixed this already Dec 02 19:22:44 padovan: Do you grok enough of gatchat to debug this with me? Dec 02 19:23:07 err s/with/for Dec 02 19:25:56 I don't think so, but I will take and see if I can fix that. Dec 02 19:26:41 let me apply the patch and see what happens Dec 02 19:26:53 In theory I fixed this for other cases, so wonder why yours is screwing up Dec 02 19:27:32 you have to remove the g_idle_add from there if want to see the bug Dec 02 19:27:48 yep Dec 02 19:34:07 (ofonod:20216): GLib-CRITICAL **: g_hash_table_iter_next: assertion `ri->version == ri->hash_table->version' failed Dec 02 19:34:52 hm, this is working for me Dec 02 19:35:15 glib version maybe? Dec 02 19:35:24 but yes, I do get the assert Dec 02 19:35:27 for whatever reason Dec 02 19:38:23 I really hate glib, that PoS needs to be rewritten ;) Dec 02 19:38:36 what do you mean by whatever reason? Dec 02 19:38:39 PoS? Dec 02 19:40:39 Piece of.. Dec 02 19:41:04 software? Dec 02 19:41:16 hehe Dec 02 19:41:23 :-) Dec 02 19:41:28 not quite what I had in mind ;) Dec 02 19:49:27 Hah Dec 02 19:49:40 the real reason is that we trigger powered(true) Dec 02 19:50:03 Which inserts some more things into the hash table Dec 02 19:51:02 causing glib to assert, which is just plain stupid Dec 02 20:02:47 padovan: For the first patch, have a think about race conditions Dec 02 20:03:29 Right now if the modem is in the process of going online / offline, calling ofono_modem_set_powered would treat it like we're doing a powered on / off Dec 02 20:03:57 So this might just work accidentally, or not.. Dec 02 20:26:51 Jeevaka: around? Dec 03 00:51:49 denkenz: I was able to use my freerunner with ofonod running on my PC. My debian in freerunner was using 2.6.32 kernel, but then I read a bit about /sys used.. and I noted few differences. So I switched 2.6.29 openmoko andy kernel and now ofonod is happy with freerunner. I am doing more tests, but I wanted to tell you that. Dec 03 00:52:08 good Dec 03 00:52:36 That code is a little unmaintained, glad it still works Dec 03 00:54:15 denkenz: yes, I see that as well. But well, glad as well it works, I was a bit frustrated yesterday :) Dec 03 00:55:09 nod, it takes a little time to get used to it Dec 03 00:55:24 the freerunner over tcp is one of the most fidgety cases Dec 03 00:55:46 most others, plug in and it works ;) Dec 03 00:56:57 (over tcp) yes, it was useful I found some mails in ofono mailing list archives **** ENDING LOGGING AT Fri Dec 03 03:00:00 2010