**** BEGIN LOGGING AT Wed Nov 19 02:59:57 2008 Nov 19 04:40:18 eureaka success! http://scap.linuxtogo.org/files/abb6b29eefcd835b791fe0f7498853f1.png Nov 19 05:50:29 Nov 19 05:58:32 Ainulindale: http://pastebin.ca/1261160 Nov 19 05:58:55 Patch to prevent the seg faults when there's an error with retreiving the message data. Nov 19 05:59:24 (There's still something wrong, but at least it doesn't seg fault anymore) Nov 19 08:28:36 ping, anyone oe dudes here? Nov 19 08:28:45 *any Nov 19 08:47:02 mwester: this isn't logical Nov 19 08:47:46 mwester: But thanks anyway :-) Nov 19 08:59:21 hi all... anyone has an idea how I could detect an attempt to access the internet, and tell frameworkd to start the GPRS connection? (I don't want to repeately connect, as my provider tends to disconnect me every 600 seconds if I'm idle) Nov 19 08:59:53 I thought maybe something with iptables or so.... (I don't want to use the pppd demand option, as then I can't control it via the framework) Nov 19 09:21:35 morning Nov 19 09:33:57 beniwtv: you can do a script that every 600 second ping google :) Nov 19 09:39:13 Hire: That's what I'm doing right now... my script is checking PDP.GetStatus() every x seconds, and reconnects if it has hungup. The problem is, if I'm idle, my provider forces a disconnect (C DEACT IP), so my script reconnects, then they disconnect again, etc... Nov 19 09:39:25 and endless loop Nov 19 09:39:58 So I tought to only connect when some program wants to use Internet Nov 19 09:40:11 So that I don't drive my provider mad :P Nov 19 09:40:56 beniwtv: I think Hires idea is to not get idle in first place Nov 19 09:41:13 so you don't get disconnected Nov 19 09:41:25 and you achieve this by pinging something Nov 19 09:41:54 Though that isn't very elegant ;) Nov 19 09:42:00 hmmm... yea, that could be a solution, the problem is I can't ping over my GPRS connection somehow. I think they block it :( Nov 19 09:43:00 but it might be enough so that they don't disconnect me Nov 19 09:43:12 * beniwtv is trying that... Nov 19 09:44:53 beniwtv: then get a webpage Nov 19 09:44:56 google.com for instance Nov 19 09:45:03 you can use cron/curl or lwp or whatever to do that Nov 19 09:45:09 even wget Nov 19 09:47:06 mickey|zzZZzz, alphaone, stefan_schmidt? Nov 19 09:47:07 Anyone alive? :-) Nov 19 09:48:42 Woooooh Nov 19 09:48:50 I just found out my rules decision system of choice in python Nov 19 09:53:15 i am alive Nov 19 10:07:49 Ainulindale: sorry, I'm beginning to think I'm dumb. But I tried to setup the sources, so I can start changing something. But it won't build. See http://pastebin.com/m53e38f5f. (And here is my .bb file: http://pastebin.com/m4bb3a7ac) Nov 19 10:13:24 dolf1074: I have time, wait a minute so I have a look Nov 19 10:13:31 This is git-native Nov 19 10:13:43 It fails to build for quickdev too Nov 19 10:13:56 Find the git-native bb recipe Nov 19 10:13:57 And remove the patch which is failing Nov 19 10:14:07 stefan_schmidt: ping Nov 19 10:14:20 (patch is git-gui-install-mode-arg-spaces.patch) Nov 19 10:16:15 dolf1074: it's in the SRC_URI, just remove it Nov 19 10:16:42 ok, will try Nov 19 10:23:38 Ainulindale: seems like it is working :) Nov 19 10:25:43 Ainulindale: alive and kicking Nov 19 10:27:43 dolf1074: cool then Nov 19 10:27:48 alphaone: sample question here Nov 19 10:27:48 I'm having fun with oeventsd Nov 19 10:27:56 Namely I created an action for rules.yaml Nov 19 10:28:14 okay Nov 19 10:28:14 I just have problems with the yaml parser, I'm not sure if you're confident enough with it to help me out Nov 19 10:28:22 to represent *args Nov 19 10:30:36 hmm Nov 19 10:31:02 I can take a look at the code if you show it to me Nov 19 10:31:39 alphaone: In fact I don't know how this should be properly done Nov 19 10:31:39 http://git.freesmartphone.org/?p=framework.git;a=blob;f=framework/subsystems/oeventsd/fso_actions.py;h=64e19e536e3059722d4a53129a5a2cf41fad90d0;hb=HEAD <= see the fso_actions Nov 19 10:31:40 I created an actions with *args Nov 19 10:31:57 It's parsed here Nov 19 10:31:57 http://git.freesmartphone.org/?p=framework.git;a=blob;f=framework/subsystems/oeventsd/parser.py;h=b135990a44d421dfb8a74075e3150971baaf200a;hb=HEAD Nov 19 10:32:31 My problem is : how can I specify a variable list of arguments? Nov 19 10:32:48 I have no output from the yaml parser regarding my rule Nov 19 10:33:32 And either using MyAction(arg, arg, arg, [vararg, vararg] or MyAction(arg, arg, arg, vararg, vararg) doesn't work Nov 19 10:33:32 Ainulindale, richy, Hire: So far GPRS has not disconnected. Though I may want to check how much they are chargin me :P Nov 19 10:33:32 (with the good changes indeed) Nov 19 10:33:34 beniwtv: yes it'd be a good idea Nov 19 10:38:16 Ainulindale: good morning and heeeeeeeeelp :) Nov 19 10:38:36 Ainulindale: I've just flashed the GSM chip and done an opkg upgrade... Nov 19 10:38:45 Ainulindale: and now, this is happening: Nov 19 10:39:28 (with ophonekitd): Nov 19 10:39:28 ** (process:1312): DEBUG: get_messagebook_info_callback() Nov 19 10:39:28 Segmentation fault Nov 19 10:40:39 I still can't understand why glib doesn't handle this hash properly Nov 19 10:42:13 which hash? Nov 19 10:43:37 freesmartphone.org: 03ainulindale 07libframeworkd-glib * rfbb1ae7e80b8 10/src/ogsmd/frameworkd-glib-ogsmd-sim.c: Commited fix for hashtable in messagebook_info. Thanks mwester for spotting. Nov 19 10:44:20 Ainulindale: doesnt MyAction(*varargs) work? Nov 19 10:44:20 polz: Wait a minute then update Nov 19 10:44:29 I'll tell you when Nov 19 10:44:29 cool, thanks Nov 19 10:44:33 Ainulindale: Next problem, the do_install of libframeworkd-phonegui-efl failed. see http://pastebin.com/m17f614ea (do_install_log: http://pastebin.com/m7b6555b8) Nov 19 10:44:42 dolf1074: show me your bb file again please? Nov 19 10:45:05 http://pastebin.com/m5a0195ef Nov 19 10:46:28 Ok this is wrong Nov 19 10:46:30 You shouldn't have put the SRC_URI here Nov 19 10:47:19 modify the shr.bbclass file Nov 19 10:47:19 hm... is ogpsd in the framework supposed to work yet with the UBX protocol? Nov 19 10:47:19 There's a commented out tag for local work Nov 19 10:47:29 And in that file there's also a S= variable Nov 19 10:47:38 S = "${WORKDIR}/git/${PN}" Nov 19 10:47:40 You should remove git there Nov 19 10:47:42 And replace the SRC_URI in the bbclass by the appropriate stuff (which you did in your bb file) Nov 19 10:48:13 Though it'll break other builds IIRC Nov 19 10:48:13 (because of bitbake) Nov 19 10:48:52 but I'll find out a way to do that more properly Nov 19 10:48:54 okay, will try that. (now I'm feeling my really dumb ;) ) Nov 19 10:49:02 You could try to put a SRCREV="local" Nov 19 10:49:02 in that file Nov 19 10:49:19 polz: package index is regenerating Nov 19 10:51:05 polz: you should update Nov 19 10:53:32 thanks. Nov 19 10:56:24 Ainulindale: opkg upgrade doesn't do anything; which package should I re-install? Nov 19 10:57:07 libframeworkd-glib Nov 19 10:57:07 This is getting on my nerves Nov 19 10:57:09 Wait a minute Nov 19 10:57:09 Bitbake borks again Nov 19 10:57:26 Ainulindale: SRCREV didn't revolve the breakage: http://pastebin.com/mf1a7dd7 (ps my shr.bbclass: http://pastebin.com/m136b9c18) Nov 19 10:58:01 Can't look now Nov 19 10:58:01 I have to go Nov 19 10:59:00 Ainulindale: sorry for taking so much of your time. Nov 19 10:59:01 me to Nov 19 11:01:01 I'm guessing the correct package name would be libframeworkd-glib0 Nov 19 11:03:02 nope, it still segfaults after a reinstall Nov 19 11:21:02 I was trying to build shr image last night, but i got this error, any ideas? http://rafb.net/p/AbUpcJ88.html Nov 19 11:35:44 it didn't found some receipes Nov 19 11:36:01 do you have a updated openembedded repo? Nov 19 11:36:16 ls Nov 19 11:39:14 methril: im sure i do, i followed this a couple of times http://shr.bearstech.com/trac/wiki/Building SHR?version=3 Nov 19 11:40:42 for the message looks like is missing some raster (illume) patches Nov 19 11:42:10 hm, maybe the patching went haywire Nov 19 11:42:12 i had to change a script there, it was using git-apply but i didn't have that binding, used git apply instead Nov 19 11:42:30 methril: btw, i was using the 2nd solution on that page, for patching. Nov 19 11:43:20 the alternative one? Nov 19 11:43:33 yeah Nov 19 11:43:54 it said i should use that one for now, right? Nov 19 11:43:54 do you have in your conf/local.conf the "require conf/distro/include/xxx.inc"? Nov 19 11:45:12 http://rafb.net/p/Avvvtj67.html Nov 19 11:45:12 i'm trying to compile now Nov 19 11:46:08 oki Nov 19 11:51:26 Here I am again Nov 19 11:51:28 With a full belly Nov 19 11:51:54 polz: seems it didn't update properly anyway Nov 19 11:52:58 Ainulindale: I've given up on it and am now trying to flash the nov. 17. image to the phone Nov 19 11:53:03 polz: It'll give you the same thing Nov 19 11:53:03 _and_ I have a lecture in 8 minutes Nov 19 11:53:17 huh? How come that never happened yesterday or the day before? Nov 19 11:53:32 * lindi- runs apt-get dist-upgrade during lectures Nov 19 11:54:53 polz: No idea at all but the package wasn't upgrade properly here Nov 19 11:54:54 So if you have time to wait for the recompilation Nov 19 11:54:54 +d Nov 19 11:54:54 Ainulindale: anyway, I still have to have some sort of working image on the phone in order to upgrade Nov 19 11:55:07 and no, unfortunately, I don't have the time. Nov 19 11:55:43 Ok then please bug me later for that Nov 19 11:55:44 I guess I'll just be phoneless for the next hour and a half or so :) Nov 19 11:56:30 btw, by working I mean "it boots", not "It rarely crashes and you can make phonecalls with it" Nov 19 11:56:46 I know Nov 19 11:56:46 Don't worry I'll fix that today Nov 19 11:56:46 anyway, see you all later. Nov 19 11:56:58 This is a pity Nov 19 11:57:00 Bitbake doesn't seem to properly get the SRCREV anyway Nov 19 11:57:17 Ainulindale: well, it's your free time we're wasting here, so please, don't feel pressured to fix it right away. Nov 19 11:57:48 I'll fix that, this is a blocking issue Nov 19 11:58:25 alphaone? Nov 19 11:59:42 Ainulindale: ? Nov 19 11:59:42 aah :-) Nov 19 11:59:43 alphaone: ok I managed to find out how to declare a [ ] thing with yaml Nov 19 12:00:28 But what's weird is that the constructor of the action seems to fail Nov 19 12:00:28 With no apparent messages Nov 19 12:00:28 Any hints on how I could test that? Nov 19 12:01:14 alphaone: my example is as follows Nov 19 12:01:15 If I do a super().__init__(things, *args) Nov 19 12:01:15 It won't work Nov 19 12:01:16 If I do a super().__init__(things, "arg1", "arg2") Nov 19 12:01:16 it'll work Nov 19 12:01:17 So either I'm passing the args using an invalid manner Nov 19 12:01:18 Ainulindale: I'll be back in a couple of minutes. Got to go shopping now Nov 19 12:01:18 a Nov 19 12:01:18 Or I'm doing something wrong here Nov 19 12:01:18 alphaone: no problem :-) Nov 19 12:01:20 Ainulindale: Try **args Nov 19 12:01:56 alphaone: this is a [ ] Nov 19 12:01:56 There's something with stars in python Nov 19 12:01:56 not a {} Nov 19 12:01:56 hmm Nov 19 12:01:56 I thought ** was for {} Nov 19 12:03:30 methril: so, it worked for you? Nov 19 12:03:42 no Nov 19 12:03:54 same error? Nov 19 12:04:13 shoragan: should know about * and ** in python Nov 19 12:04:13 no, another one Nov 19 12:04:15 hm Nov 19 12:04:15 alphaone: I read the tutorial anyway Nov 19 12:04:22 I didn't Nov 19 12:04:24 :-) Nov 19 12:04:29 And that's not that much helpful regarding this issue Nov 19 12:04:42 It finds the args Nov 19 12:05:05 2008.11.20 13:01:04 oeventsd INFO ['system', 'org.freesmartphone.ogsmd', '/org/freesmartphone/GSM/Device', 'org.freesmartphone.GSM.Call', 'Initiate', ['123', 'voice']] Nov 19 12:05:05 hmm Nov 19 12:05:05 it finds the func Nov 19 12:05:05 2008.11.20 13:01:04 oeventsd INFO Nov 19 12:05:05 but func(*args) gives Nov 19 12:05:05 2008.11.20 13:01:04 oeventsd INFO [] Nov 19 12:05:24 * is for variable (non-kw) args, ** is for variable kw args Nov 19 12:05:42 shoragan: ok so I wasn't mistaken here Nov 19 12:05:43 Here, I have the args I pasted above Nov 19 12:05:43 I want to create an object with the following signature : Obj(arg1, arg2, *args) Nov 19 12:05:49 It just seems to fail Nov 19 12:05:55 shoragan: And you can also have def foo(*args) Nov 19 12:07:50 >>> class C(object): Nov 19 12:07:50 ... def __init__(self, foo, bar, *args): Nov 19 12:07:50 ... print self, foo, bar, args Nov 19 12:07:50 ... Nov 19 12:08:23 >>> c = C(1,2) Nov 19 12:08:23 <__main__.C object at 0x7fb48025c390> 1 2 () Nov 19 12:08:23 >>> c = C(1,2,3) Nov 19 12:08:23 <__main__.C object at 0x7fb48025c490> 1 2 (3,) Nov 19 12:08:23 >>> c = C(1,2,3,4) Nov 19 12:08:24 <__main__.C object at 0x7fb48025c390> 1 2 (3, 4) Nov 19 12:08:42 >>> vars = [3,4,5] Nov 19 12:08:42 >>> c = C(1,2,*vars) Nov 19 12:08:42 <__main__.C object at 0x7fb48025c490> 1 2 (3, 4, 5) Nov 19 12:09:14 shoragan: yeah, this I understand Nov 19 12:09:27 >>> class D(object): Nov 19 12:09:27 ... def __init__(self, foo, bar, **args): Nov 19 12:09:27 ... print self, foo, bar, args Nov 19 12:09:27 ... Nov 19 12:09:37 >>> c = D(1,2,baz=100,gargel=20) Nov 19 12:09:37 <__main__.D object at 0x7fb48025c510> 1 2 {'gargel': 20, 'baz': 100} Nov 19 12:09:39 and you can combine both Nov 19 12:10:54 >>> class E(object): Nov 19 12:10:54 ... def __init__(self, foo, bar, *vargs, **kargs): Nov 19 12:10:54 ... print self, foo, bar, vargs, kargs Nov 19 12:10:54 ... Nov 19 12:11:12 >>> c = E(1,2,3,4,baz=100,gargel=20) Nov 19 12:11:12 <__main__.E object at 0x7fb48025c550> 1 2 (3, 4) {'gargel': 20, 'baz': 100} Nov 19 12:11:12 AH Nov 19 12:11:12 Yep, saw that Nov 19 12:11:20 But in the right order Nov 19 12:11:20 pastebin Nov 19 12:11:42 the kw-args have to be at the end Nov 19 12:11:50 yep Nov 19 12:11:52 That I know Nov 19 12:14:44 then i don't understand your problem ;) Nov 19 12:14:44 Neither do I :-p Nov 19 12:14:45 *sigh* Nov 19 12:14:46 * Hire wants opimd Nov 19 12:14:47 shoragan: Wait a minute I'll show you Nov 19 12:15:11 shoragan: [on InputTrigger if And(switch == AUX,event == released,duration == 0) then ExternalDBusAction(bus = , service = org.freesmartphone.ogsmd, obj = /org/freesmartphone/GSM/Device, itf = org.freesmartphone.GSM.Call, method = ['Initiate', '123'])] Nov 19 12:15:12 This is what I generate Nov 19 12:15:12 Ah Nov 19 12:15:13 Looking at that Nov 19 12:15:13 I obviously made a sad mistake Nov 19 12:15:23 I should read outputs Nov 19 12:17:15 Thought it'll give the same results anyway Nov 19 12:18:15 Ok Nov 19 12:18:15 [on InputTrigger if And(switch == AUX,event == released,duration == 0) then ExternalDBusAction(bus = , service = org.freesmartphone.ogsmd, obj = /org/freesmartphone/GSM/Device, itf = org.freesmartphone.GSM.Call, method = Initiate)] Nov 19 12:18:49 the args are ["123", "voice"] Nov 19 12:18:58 They are detected as I pasted before Nov 19 12:20:01 And now I think I know why, again Nov 19 12:21:09 (will try) Nov 19 12:21:48 shoragan: see that Nov 19 12:21:48 2008.11.20 13:17:37 oeventsd DEBUG trigger on InputTrigger if And(switch == AUX,event == released,duration == 0) then ExternalDBusAction(bus = , service = org.freesmartphone.ogsmd, obj = /org/freesmartphone/GSM/Device, itf = org.freesmartphone.GSM.Call, method = Initiate) Nov 19 12:21:49 2008.11.20 13:17:37 oeventsd INFO call dbus signal /org/freesmartphone/GSM/Device Initiate((['123', 'voice'],)) Nov 19 12:21:49 it tries to call a method with the args, and not the *args Nov 19 12:21:50 Thus giving me Nov 19 12:21:50 2008.11.20 13:17:38 dbus.connection ERROR Unable to set arguments (['123', 'voice'],) according to signature u'ss': : Expected a string or unicode object Nov 19 12:21:50 gmm Nov 19 12:21:51 whats in args? Nov 19 12:23:14 the [] thing Nov 19 12:23:23 2008.11.20 13:19:44 oeventsd INFO ['system', 'org.freesmartphone.ogsmd', '/org/freesmartphone/GSM/Device', 'org.freesmartphone.GSM.Call', 'Initiate', '123', 'voice'] Nov 19 12:23:27 empty list Nov 19 12:23:27 This is the full thing parsed by yaml Nov 19 12:23:29 *arg represents (supposedly) "123" and "voice" Nov 19 12:23:29 hey guys. Have you seen: http://n2.nabble.com/FSO-M4---nice-GSM-sound-tp1486414p1518726.html - I just testet it and it works perfectly. any chance that gets integrated into shr? Nov 19 12:23:29 that semms ok Nov 19 12:24:17 richy: if it is included in frameworkd it'll get included in SH Nov 19 12:24:17 +R Nov 19 12:24:25 shoragan: seems ok to me too Nov 19 12:24:25 I obviously did something wrong somewhere Nov 19 12:24:25 directly before calling the dbus function, please print all paramenters Nov 19 12:25:06 shoragan: http://pastebin.com/m74f95252 Nov 19 12:25:15 this is the code Nov 19 12:25:15 (with ugly debug) Nov 19 12:25:35 this class is instantiated wit hthe arguments I pasted before Nov 19 12:26:11 http://pastebin.com/m763491f6 Nov 19 12:26:13 shoragan: 2008.11.20 13:24:13 oeventsd INFO call dbus signal /org/freesmartphone/GSM/Device Initiate((['123', 'voice'],)) Nov 19 12:26:31 This is just before the dbus call Nov 19 12:26:35 uhu? Nov 19 12:26:52 you don't have the session bus in the framework Nov 19 12:26:52 logger.info("call dbus signal %s %s(%s)", self.obj, self.method, self.args) Nov 19 12:27:01 that wont work anyway Nov 19 12:27:01 I'm using the system bus Nov 19 12:27:05 And it did work using it Nov 19 12:27:05 Just not using *args Nov 19 12:27:07 but then why have an ExternalDBusAction? Nov 19 12:27:17 This is purely a language issue, I think, here Nov 19 12:27:18 shoragan: because I have other needs for that :-) Nov 19 12:27:40 And there's no current way to call even frameworkd methods with the rules Nov 19 12:27:40 no dbus actions Nov 19 12:27:49 the dbus action itself takes the bus as a python object Nov 19 12:28:14 It shouldn't be done that way, should it? Nov 19 12:28:37 Ainulindale: Yes there Nov 19 12:28:37 is Nov 19 12:28:37 we're currently using the rules to set the leds for example Nov 19 12:28:38 There is a specific action for the leds IIRC Nov 19 12:28:56 (and anyway it's fun to try) Nov 19 12:29:09 Ainulindale: Ah, no Nov 19 12:29:32 there's a DBus Trigger, though Nov 19 12:29:32 Yep Nov 19 12:29:32 But not a dbus action Nov 19 12:29:41 Don't you think it's sad? :-) Nov 19 12:29:41 right Nov 19 12:29:41 Ainulindale: Very :-) Nov 19 12:29:49 alphaone: goal being to trigger illume screensaving feature Nov 19 12:29:54 to forget about setbrightness once and for all Nov 19 12:29:54 (for SHR) Nov 19 12:30:02 why not just use screensaver of X? Nov 19 12:30:03 Ainulindale, it wont work easily Nov 19 12:30:03 Even if it doesn't work Nov 19 12:30:12 I just want to fiddle with it to understand how it works Nov 19 12:30:12 because you don't have access to the session bus Nov 19 12:30:12 THis is a good example Nov 19 12:30:23 that's by design of dbus Nov 19 12:30:23 shoragan: Let's say I just want to use the system one :-) Nov 19 12:30:29 It still gives me an issue here Nov 19 12:31:58 alphaone: I would also like to be able to detect several inputs at a time Nov 19 12:31:58 i've added a line to your paste Nov 19 12:32:08 I thought what I pasted before should be enough Nov 19 12:32:38 shoragan: adding it Nov 19 12:32:49 somehow the wrapping and unwrapping doesn't match Nov 19 12:33:11 2008.11.20 13:31:07 oeventsd INFO ('123', 'voice') Nov 19 12:33:51 that's good Nov 19 12:33:52 Yep that's what I expected Nov 19 12:34:40 Though it'll still give me Nov 19 12:34:40 2008.11.20 13:32:57 oeventsd INFO call dbus signal /org/freesmartphone/GSM/Device Initiate((['123', 'voice'],)) Nov 19 12:35:27 => Nov 19 12:35:27 logger.info("call dbus signal %s %s(%s)", self.obj, self.method, self.args) Nov 19 12:35:35 (action.py) Nov 19 12:36:02 could you also log args in DBusAction.__init__? Nov 19 12:37:19 I could Nov 19 12:37:21 Have you been nice? Nov 19 12:37:23 If you didn't, no present this year Nov 19 12:37:24 :) Nov 19 12:37:45 somewhere near there it gets mangled Nov 19 12:38:33 2008.11.20 13:36:51 oeventsd INFO ('123', 'voice') Nov 19 12:38:33 2008.11.20 13:36:51 oeventsd INFO ('123', 'voice') Nov 19 12:38:43 also ok Nov 19 12:39:00 ah Nov 19 12:39:01 logger.info("call dbus signal %s %s(%s)", self.obj, self.method, self.args) Nov 19 12:39:04 change that to: Nov 19 12:39:48 ah, no Nov 19 12:40:24 str(self.args)? Nov 19 12:40:45 yes Nov 19 12:40:47 logger.info("call dbus signal %s %s%s", self.obj, self.method, str(self.args)) Nov 19 12:41:11 ah, the logger function may mangle it Nov 19 12:42:39 That is to say? Nov 19 12:42:41 still 2008.11.20 13:40:40 oeventsd INFO call dbus signal /org/freesmartphone/GSM/Device Initiate((['123', 'voice'],)) Nov 19 12:42:57 it does something to format the parameters for logging Nov 19 12:42:58 I'll add some debug Nov 19 12:43:01 maybe thats broken Nov 19 12:43:34 I bet that's because of charlie *nudge nudge* Nov 19 12:44:05 C is easier.. sheesh -:) Nov 19 12:44:36 2008.11.20 13:42:56 oeventsd INFO (['123', 'voice'],) Nov 19 12:44:36 2008.11.20 13:42:56 oeventsd INFO call dbus signal /org/freesmartphone/GSM/Device Initiate((['123', 'voice'],)) Nov 19 12:45:12 so where does the additional [] come from? Nov 19 12:45:55 (I can't understand why it doesn't regenerate the pyo systematically) Nov 19 12:46:45 shoragan: Ok now it works :-) Nov 19 12:46:46 That's weird Nov 19 12:47:29 Now, let's remove frameworkd and add the actions from scratch Nov 19 12:47:44 what was the problem? Nov 19 12:47:48 No idea Nov 19 12:48:01 I regenerated the pyo by removing them Nov 19 12:48:14 hm Nov 19 12:48:15 wrong permissions? Nov 19 12:48:26 I'm root Nov 19 12:48:26 (I'm using it on the device) Nov 19 12:51:55 Ok it's working Nov 19 12:52:21 alphaone: interested in the commit? Nov 19 12:52:21 Or should I wait for mickeyl? Nov 19 12:54:09 commit? Nov 19 12:54:10 * alphaone wasn't listening... Nov 19 12:54:44 alphaone: Meeeh :-) Nov 19 12:54:44 I just built a DBusAction for oeventsd Nov 19 12:54:48 It's working nicely here, should I commit it? Nov 19 12:55:27 alphaone: (I thought this would be useful anyway for me to understand oeventsd) Nov 19 12:55:28 (and indeed it was) Nov 19 12:57:36 Ainulindale: Sure, commit to master Nov 19 12:57:37 alphaone: by the way Nov 19 12:57:37 mickey|zzZZzz: ping :) Nov 19 12:57:37 I think I broke master anyway... Nov 19 12:57:37 Be VERY wary on reports about no ringtone Nov 19 12:57:38 Yesterday I had this big problem Nov 19 12:58:00 Ainulindale: ah yeah, sorry Nov 19 12:58:00 Turned out it was just speech-dispatcher (navit) Nov 19 12:58:01 I've noticed that as well Nov 19 12:58:01 Which was hogging the audio resource Nov 19 12:58:01 After a suspend Nov 19 12:58:36 Unfortunately Nov 19 12:58:36 But busy with this f***ing MS4.1 release Nov 19 12:58:36 It'll trigger a deadlock in oeventsd Nov 19 12:58:54 So if there's any pulseaudio thing active before the suspend Nov 19 12:58:54 it'll bork oeventsd Nov 19 12:58:54 It's really bothering me that I don't get any new work done Nov 19 12:59:03 Just battling with image generation Nov 19 12:59:03 git-native too? Nov 19 12:59:23 no Nov 19 12:59:32 alphaone: please consider the fact that I spent the whole evening yesterday on this speech-dispatcher thing Nov 19 12:59:33 Instead of developing Nov 19 12:59:38 So don't whine :-) Nov 19 12:59:39 cherry-picking the ENTERPRISE_DISTRO changes to the ms4 oe checkout Nov 19 12:59:40 And various other fixes Nov 19 13:00:32 build, test, see that all my batteries are flat, recharge, wait, go shopping, whine a little and then hopefully flash and test Nov 19 13:00:36 find bugs, goto start Nov 19 13:01:02 Ainulindale: Yeah, this sort of stuff is annoying... Nov 19 13:01:15 freesmartphone.org: 03ainulindale 07framework * r1a5a5db680c7 10/framework/subsystems/oeventsd/fso_actions.py: Nov 19 13:01:15 freesmartphone.org: Added an oeventsd action to call DBus methods. Syntax is : ExternalDBusAction("system"|"session", service, device, Nov 19 13:01:15 freesmartphone.org: interface, method, *arguments). Nov 19 13:01:15 freesmartphone.org: Session bus is untested and shouldn't work as is (FIXME). Nov 19 13:01:19 freesmartphone.org: Thanks shoragan for the step-by-step thing ;-). Nov 19 13:02:02 session bus *WONT WORK* Nov 19 13:02:02 alphaone: yeah I almost have no time to develop anything inbetween Nov 19 13:02:20 shoragan: there's a way to get the session bus Nov 19 13:02:20 So it "could" work :-) Nov 19 13:02:37 it a complete hack Nov 19 13:02:46 I agree Nov 19 13:03:02 shoragan: I'll make it throw an error on session bus if necessary Nov 19 13:03:19 the "right" way is to have an agent in the session, which then offers services also on the system bus Nov 19 13:03:19 Though I think it's weird not being able to find this session bus Nov 19 13:03:19 see networkmanager, bluetooth... Nov 19 13:03:33 can anoyone point me to a recent zhone ipkg to install on shr? I like to check out if the backend or frontend causes my problem by using a different phone app Nov 19 13:03:56 DPThought: compiling it now Nov 19 13:03:56 Ainulindale, there may be no session bus or even many of them Nov 19 13:04:14 DPThought: which version of frameworkd? Nov 19 13:04:15 shoragan: yep, I know Nov 19 13:04:15 Ainulindale: ah, great Nov 19 13:04:15 DPThought: i have simple frontend app Nov 19 13:04:23 lindi-: the latest in the shr feed Nov 19 13:04:23 so the frameworkd has nothing to do with a session bus Nov 19 13:04:32 DPThought: i don't know what that has. do you have a version number? Nov 19 13:04:32 shoragan: Yep I get that :-) Nov 19 13:04:32 we talked about that at length when we were in taipei Nov 19 13:04:32 lindi-: latest stabilization Nov 19 13:04:36 Anyone from the framework team can clarify this for me? When I tried GPS with both zhone and fso-gpsd, I never got a fix. It turns out that I needed to /sys/devices/platform/s3c2440-i2c/i2c-adapter/i2c-0/0-0073/neo1973-pm-gps.0/pwron * before* requesting the resource. So my guess is that zhone and fso-gpsd do not power on the gps via d-bus calls, when they should. Any thought? Or is frameworkd supposed to power on automatically the GPS when reque Nov 19 13:04:47 shoragan: Ok then I'll let mickey|zzZZzz beat me a little and remove that :-) Nov 19 13:05:24 :P Nov 19 13:05:31 lindi-: it says 1.0.0 in /usr/bin/frameworkd, but this is probably not hat you want :) Nov 19 13:05:31 shoragan: what's great though is that in the ended, I eventually had a look at that Nov 19 13:05:49 DPThought: opkg update/opkg install zhone Nov 19 13:05:49 freesmartphone.org: 03mickey 07framework * r00c000c9ba11 10/framework/ (3 files in 2 dirs): ogsmd: split server and device object implementation into two files Nov 19 13:05:59 mickey|sofa: heya Nov 19 13:06:08 morning Nov 19 13:06:15 mickey|sofa: I just commited the dbus action we talked about Nov 19 13:06:15 Ainulindale: ah, did you add it to the feed? nice Nov 19 13:06:41 Ainulindale: cool Nov 19 13:06:41 Ainulindale, your implementation just does the same as the existing DBusAction Nov 19 13:06:54 shoragan: except that I added session/system Nov 19 13:06:54 even if it does naught Nov 19 13:07:04 system should have worked before Nov 19 13:07:05 no Nov 19 13:07:12 the bus object is a python object Nov 19 13:07:17 DPThought: no Nov 19 13:07:27 DPThought: what's the version number of the package? Nov 19 13:07:34 shoragan: that's why there was a need for something like that, even so basic Nov 19 13:08:03 105 trigger: DbusTrigger(system, 'org.freesmartphone.odeviced', 0, 'org.freesmartphone.Device.IdleNotifier', 'State') Nov 19 13:08:03 106 actions: Debug("dbus trigger test") Nov 19 13:08:12 shoragan: trigger Nov 19 13:08:13 you need to supply system as an identifyer, not as a string Nov 19 13:08:13 lindi-: 0.8.4.3+gitr752+e594989233f2db669821c900964bdee561a9a2db-r0.1 Nov 19 13:08:21 DPThought: ok Nov 19 13:08:49 DPThought: http://iki.fi/lindi/openmoko/kapula might be compatible then Nov 19 13:08:49 shoragan: I bet you just want to bug me with that :-p Nov 19 13:09:00 ah Nov 19 13:09:12 see the code for DBusTrigger here: Nov 19 13:09:12 http://git.freesmartphone.org/?p=framework.git;a=blob;f=framework/subsystems/oeventsd/trigger.py;h=840fbc81a2e97d91ab41566c0c2fa804f7b764e2;hb=1a5a5db680c7695c4116df0e92b9df4dfe53c943 Nov 19 13:09:13 lindi-: oh, sorry. Ainulindale just added it in the shr opkg feed, no need anymore. Nov 19 13:09:13 (Yeah I was reading it) Nov 19 13:10:02 so that could simply go into DBusAction Nov 19 13:10:03 shoragan: indeed Nov 19 13:10:04 I'll do that after my cigarette =) Nov 19 13:10:05 (SHR tends to break oneself) Nov 19 13:10:05 now that, they are people here, anyone know why i keep getting this? http://dpaste.com/91746/ its making me weep Nov 19 13:10:05 (I started smoking again) Nov 19 13:10:08 Ainulindale: no more gums? :-) Nov 19 13:10:21 DocScrutinizer: nope Nov 19 13:10:46 anywone here who can read and translate a few chinese words? Nov 19 13:11:41 lindi-: oh, thats not zhone i see. But i have no gtk bindings at the moment an very scarse flash space Nov 19 13:11:42 lindi-: what does this app? Nov 19 13:12:10 DPThought: calls Nov 19 13:12:19 lindi-: just what i need :) Nov 19 13:12:49 DocScrutinizer: try Google Translate? Nov 19 13:12:50 mickey|sofa: I think I've found a bug in zhone and fso-gpsd... They do not turn on the GPS chip before requesting the GPS resource. But I'm wondering if this should not be handled by frameworkd. Any thoughts? Nov 19 13:13:30 beniwtv: surely is should be handled by it Nov 19 13:13:30 ajuonline: if only I knew how to enter the "characters" Nov 19 13:13:55 beniwtv, it is handled by ogpsd Nov 19 13:14:04 which gpsdevice did you configure in frameworkd.conf Nov 19 13:14:04 ? Nov 19 13:14:31 shoragan: hmm... then it's a bug there :) I have to manually /sys/devices/platform/s3c2440-i2c/i2c-adapter/i2c-0/0-0073/neo1973-pm-gps.0/pwron before requesting the GPS resource Nov 19 13:14:39 otherwise I do not get any fix Nov 19 13:14:48 (nor any data are received by ogpsd) Nov 19 13:14:48 beniwtv, it worked for alphaone and my Nov 19 13:14:52 which gpsdevice did you configure in frameworkd.conf? Nov 19 13:15:03 shoragan: UBXDevice, SerialChannel Nov 19 13:15:16 ah Nov 19 13:15:17 please use GTA02Device Nov 19 13:15:42 shoragan: Thanks, Trying it now.... Nov 19 13:15:45 http://git.openembedded.net/?p=openembedded.git;a=blob;f=packages/freesmartphone/frameworkd/om-gta02/frameworkd.conf;h=a2210a890ebc3289f8a229a381a18d0b98587141;hb=HEAD Nov 19 13:17:05 ha... stupid me... I've updated the local tree this morning, but didn't sync the .conf :P Nov 19 13:17:09 strange, when i launch zhone all native phone apps work too. Looks like something just needed to request the resources. Whats job would this be normally on shr? Nov 19 13:17:53 DPThought, any application which uses the DBus APIs should request the resource Nov 19 13:19:09 DPThought: ophonekitd Nov 19 13:19:09 mickey|sofa? Nov 19 13:19:11 yup, just had the idea. It was in fact not running Nov 19 13:19:46 mickey|sofa: why were the python C bindings for libyaml not use for oeventsd? Nov 19 13:19:46 +d Nov 19 13:20:27 Ainulindale: charlie said there were some problems wrt. parsing Nov 19 13:20:33 Hmmm Nov 19 13:20:49 we'd love to use them Nov 19 13:20:49 Too bad, it's taking a very long time Nov 19 13:20:49 but apparantly it's not straightforwad Nov 19 13:23:49 mickey|sofa: ah and Nov 19 13:23:49 regarding rules Nov 19 13:23:58 Do you plan to include contextual rules/variable declarations? Nov 19 13:24:00 (in the rework) Nov 19 13:24:02 e.g. Nov 19 13:24:10 InputEvent(): e, CallEvent(): f Nov 19 13:24:27 To be able to target the events you're testing the attrs of Nov 19 13:25:30 mmh, i don't know Nov 19 13:25:31 i'm scared of making the rules too powerful Nov 19 13:26:11 mickey|sofa: I had a nice discussion about that with ptitjes Nov 19 13:26:15 Why not writing simple bindings for the rules Nov 19 13:26:16 Ainulindale, you only have one active event (the one you triggerd on) Nov 19 13:26:17 the problem is we all are looking from a programmer's POV Nov 19 13:26:39 which we should not do Nov 19 13:26:40 the rules are for people who don't want to program Nov 19 13:26:40 mickey|sofa: sorry to disturb you but have you noticed http://trac.freesmartphone.org/ticket/240 ? Nov 19 13:26:40 hence we really really need to keep 'em simple (yet powerful enough) Nov 19 13:26:44 i.e. being able to tell to the rule system which events it can listen to Nov 19 13:26:45 (ReleaseResource("GSM") causes frameworkd to consume 100% cpu time) Nov 19 13:26:59 and then just give him the necessary information it needs Nov 19 13:27:03 and handle the result of the rule processing? Nov 19 13:27:25 In order to allow the user to just "plug" rules decision systems Nov 19 13:27:34 lindi-: no i didn't -- interesting, thanks. i'll test that asap Nov 19 13:27:46 That way, you could ship with a default rule processing system Nov 19 13:27:47 And plug this or that system Nov 19 13:29:35 freesmartphone.org: 03ainulindale 07framework * r5e6f7a5a6337 10/framework/subsystems/oeventsd/fso_actions.py: oeventsd/ExternalDBusAction : proper handling of the bus argument. Nov 19 13:29:39 freesmartphone.org: 03ainulindale 07framework * rc414dad70adf 10/framework/ (3 files in 2 dirs): Merge branch 'master' of git@git.freesmartphone.org:framework Nov 19 13:29:40 Damn merge Nov 19 13:29:52 shoragan: here it is Nov 19 13:29:57 Forgot to thank you this time, sorry :-) Nov 19 13:30:44 Forgot to check method anyway, will commit that later :-) Nov 19 13:30:59 Ainulindale, always use gitk before pushing Nov 19 13:30:59 gitk? Nov 19 13:31:04 gitk rocks Nov 19 13:31:37 What's that? Nov 19 13:31:51 also we don't need an additional class for this Nov 19 13:31:59 just put it into DBusActinon Nov 19 13:32:36 so it matches with DBusTrigger Nov 19 13:32:53 shoragan: are you waiting for my commits on purpose? :-) Nov 19 13:33:16 :-) Nov 19 13:33:23 Well anyway, what is gitk? Nov 19 13:33:35 i wasn't clear enough before, it seems :) Nov 19 13:33:36 apt-get install gitk Nov 19 13:33:40 apt-get? what's that? :-) Nov 19 13:33:54 (no ebuild) Nov 19 13:33:54 a visualisation tool for git Nov 19 13:33:54 alphaone? Nov 19 13:34:30 ok will try to find an ebuild then Nov 19 13:36:45 Ainulindale: compile git with useflag tk enabled Nov 19 13:36:58 oh, tk then Nov 19 13:37:03 So I guess it's graphical Nov 19 13:38:55 freesmartphone.org: 03tobias-git 07framework * r13f6f1ab7a7e 10/ (2 files in 2 dirs): Nov 19 13:38:55 freesmartphone.org: Add support for audio file options. Nov 19 13:38:55 freesmartphone.org: Options can be appended to audio file names, seperated via semicolons. Nov 19 13:38:55 freesmartphone.org: As an example, the ringtone now plays a different tune from the defaul Nov 19 13:38:55 freesmartphone.org: SID file. Nov 19 13:38:55 freesmartphone.org: 03mickey 07framework * r8b1232cb39b5 10/ (2 files in 2 dirs): update ChangeLog, change default incoming SMS notification tone Nov 19 13:38:55 Ainulindale: yeah, a nice tool to look at the branches Nov 19 13:39:02 Yeah but just so you know Nov 19 13:39:14 I'm currently on a windows computer Nov 19 13:39:16 connecting to my own computer through a vpn Nov 19 13:39:17 and putty Nov 19 13:39:30 and using vim Nov 19 13:39:36 So I'm not sure using any kind of visualization thingy is a good idea :-) Nov 19 13:39:52 yay :-) Nov 19 13:40:50 Ok now I'm bored, I did my thing Nov 19 13:41:00 I think I'll check opimd then Nov 19 13:41:09 yay Nov 19 13:41:13 :) Nov 19 13:41:15 Yay, more borking in the code Nov 19 13:41:39 * mickey|sofa needs to reping abraxa Nov 19 13:41:44 this guy is hard to reach Nov 19 13:43:26 the shutdown problems are starting to anoy me. Does anyone know why they happen and if there is some quickfix? Nov 19 13:43:35 DPThought: shutdown? Nov 19 13:43:59 if i just reboot or shutdown X stays running and then some time later the screen just freezes, until i hard reset it. I can however reboot if i stop x before Nov 19 13:45:03 DPThought: it's taking some time before the actual shutdown Nov 19 13:45:05 here it does the same, but shuts down in the end Nov 19 13:50:12 Ainulindale: i thouhght i waited long enough, but that might be just the case. I will check that out next time Nov 19 13:50:24 and regarding non running ophonekitd. http://nopaste.com/p/aquUbVxHeb Nov 19 13:50:54 OK, the lecture is over Nov 19 13:50:54 frameworkd is runnin though, and zhone comes up nicely too. Looks like maybe some race condition Nov 19 13:51:08 DPThought: are you running in debug? Nov 19 13:51:30 (I know we should display a warning of that situation, and ask to restart the connection) Nov 19 13:52:51 you mean frameworkd in debug? i was just about to switch this on Nov 19 13:52:53 DPThought: first please update Nov 19 13:53:01 Then killall -9 ophonekitd Nov 19 13:53:04 then killall -9 gsm0710muxd Nov 19 13:53:10 then /etc/init.d/frameworkd start Nov 19 13:53:14 then DISPLAY=":0" ophonekitd& Nov 19 13:53:15 And retry Nov 19 13:54:17 no new packages Nov 19 13:55:09 DPThought: that's weird, libframeworkd-glib Nov 19 13:55:44 Ii don't know what bitbake is doing Nov 19 13:55:44 But this is bugging me a lot Nov 19 13:55:52 mickey|sofa: I'm about to do something horrible Nov 19 13:56:01 mickey|sofa: launching frameworkd on windows Nov 19 13:56:34 Ainulindale: w00t! Nov 19 13:56:38 polz: is it working? Nov 19 13:56:40 Package: libframeworkd-glib0 Nov 19 13:56:40 Version: 0.0.1+r3-gitr64+fbb1ae7e80b880f9817cc5240a440bd1ca58dd02-r3.1 Nov 19 13:56:47 Ainulindale: just now, the ophoned _didn't_ crash :) Nov 19 13:56:48 ophonekitd Nov 19 13:56:56 let's see if I can place a call Nov 19 13:56:56 But that's great Nov 19 13:56:56 sorry Nov 19 13:56:56 Though it doesn't explain why it was failing in the first place Nov 19 13:57:17 DPThought: that seems right Nov 19 13:57:17 I'm too lazy to type the whole name correctly ;) Nov 19 13:57:42 it still crashes if it tries to go into suspend Nov 19 13:57:44 That's known Nov 19 13:57:45 Ainulindale: (windows) you're crazy .) Nov 19 13:57:46 Though as mwester said it shouldn't happen with latest kernel Nov 19 13:57:46 mickey|sofa: Did you doubt I was? Nov 19 13:57:55 hehe Nov 19 13:58:03 Ainulindale: i will try a sleep 30 before ophonekitd in the xsession script, maybe it is just racy Nov 19 13:58:04 Ainulindale: where should I download that :) Nov 19 13:58:12 ? Nov 19 13:58:19 DPThought: no Nov 19 13:58:47 this has nothing to do Nov 19 13:58:47 here, the dbus call is failing Nov 19 13:59:58 Ainulindale: could that be because frameworkd is not up and running right now? the dbus system itself seems to work fine, zhone which is starting at the same time has no problems Nov 19 14:00:23 Ainulindale: btw. you are right, it did reboot eventually Nov 19 14:01:10 Ainulindale: hmm, this is weird: I can see the WiFi setting in the Settings app again. Is this also expected ? Nov 19 14:04:47 Ainulindale: you are right, did not help Nov 19 14:05:22 anyway, worktime is over, going home time is now. Nov 19 14:06:28 polz: no, I removed it Nov 19 14:06:30 But it seems bitbake is mocking me Nov 19 14:06:55 at least you've managed to build an image already Nov 19 14:12:30 Hi! Nov 19 14:12:52 hey ptitjes Nov 19 14:13:17 how r u ? Nov 19 14:13:33 fine, thanks Nov 19 14:14:29 Do u think it is easier that I open a enhancement ticket for the logrotate stuff ? Nov 19 14:14:52 No, if this works I'll work on itt Nov 19 14:14:52 -t Nov 19 14:15:09 and to keep track of my dumb patch ? Nov 19 14:15:38 btw how should I proceed to directly put the logrotate in place in the rootfs ? Nov 19 14:16:25 I mean I would like to add to the SHR scm the config directives for frameworkd so that it gets automagicaly in the autobuilded rootfs Nov 19 14:16:26 ptitjes: check the image definition on git Nov 19 14:17:10 mickey|sofa, you once said dbus tcp connections are quite unstable, didn't you? Nov 19 14:18:27 mickey|sofa: any idea on the 100% cpu consumption issue? Nov 19 14:19:04 lindi-: not before i had time to look into it Nov 19 14:19:18 ok :) Nov 19 14:19:34 quickdev: not unstable per se, but very "involved", i.e. absolutely not straightforward Nov 19 14:19:52 btw., did you ever get gsm forwarding reliably? Nov 19 14:19:53 mickey|sofa: hey the patch for the prevent suspend while in call is nice but is it truly necessary? Nov 19 14:20:02 Ainulindale: yes and no Nov 19 14:20:04 I mean, is it the responsability of frameworkd to handle that? Nov 19 14:20:16 As we may want to suspend while on call Nov 19 14:20:17 suspending during call would be useful Nov 19 14:20:17 saves power Nov 19 14:20:19 Or we may not want to suspend while using gps Nov 19 14:20:42 lindi-: etc. Nov 19 14:20:47 Ainulindale: why not? i want gps to keep its fix Nov 19 14:20:56 i don't like the patch per se, it voilates layering Nov 19 14:20:57 mickey|sofa, have you heard of fusd? I think that kernel module should work for proper forwarding.....but dbus tcp would be even better Nov 19 14:20:58 mickey|sofa: indeed Nov 19 14:20:58 but i do like the idea of a subsystem being able to veto against a suspend Nov 19 14:21:10 lindi-: why not what? Nov 19 14:21:11 mickey|sofa: yep, that's what we did in ophonekitd Nov 19 14:21:11 quickdev: never heard of that Nov 19 14:21:15 quickdev: sounds interesting Nov 19 14:21:25 quickdev: got a link handy? Nov 19 14:21:31 mickey|sofa: and in fact, the patch is exactly what we do Nov 19 14:22:00 Though the rule is redundant Nov 19 14:22:41 mickey|sofa: there should be a way to "rulify" the suspendability Nov 19 14:22:51 at runtime Nov 19 14:23:33 Ainunlidale: Ok found and understood the shr-image.inc! What is the best approach from the following : Nov 19 14:23:33 yeah Nov 19 14:23:37 multiple approaches here: Nov 19 14:23:37 mickey|sofa, http://www.circlemud.org/~jelson/software/fusd/ Nov 19 14:23:38 mickey|sofa: I'm quite happy with my recent changes about timeouts by the way it's working great Nov 19 14:23:39 a) remove the rule and just handle it on your own Nov 19 14:23:40 or Nov 19 14:23:47 mickey|sofa: and shr-settings allows the user to change suspend this way Nov 19 14:24:16 b) add resource-vetoing-against suspend and allow configuring the subsystems Nov 19 14:24:16 still thinking about b) Nov 19 14:24:25 1) Patch the frameworkd package and directly add logrotate dep. and config directives (hmmm...) Nov 19 14:24:25 Ainulindale: (timeouts) cool Nov 19 14:24:29 mickey|sofa: I think this should be specified when requesting the resource Nov 19 14:24:29 quickdev: thanks Nov 19 14:24:46 This is less invasive, and easy to do at first Nov 19 14:24:47 2) Add files to the shr-image/etc/ to overide standard ones Nov 19 14:25:11 i.e. first request an unsuspendable resource or not Nov 19 14:25:12 3) Apply patches on post-processed to modify standard ones ? Nov 19 14:25:24 then, be able to handle rules for the suspendability of suspendable resources Nov 19 14:25:25 the first case covers the GPS use case Nov 19 14:25:35 ptitjes: I won't accept patches :-p Nov 19 14:25:49 ptitjes: logrotate should be desolidarized from frameworkd, shouldn't it? Nov 19 14:26:03 so 2) or 3) Nov 19 14:26:31 mickey|sofa: each resource, IMHO Nov 19 14:26:49 mickey|sofa: quickdev: that doesn't make any sense - dbus over TCP is functionally identical to dbus over UNIX sockets, once the socket is created Nov 19 14:26:49 mickey|sofa: should give the full extent of its information to a rule handler Nov 19 14:27:44 Ainulindale, btw...fetching python-wifi failed here Nov 19 14:27:54 quickdev: hehe, yeah Nov 19 14:28:03 quickdev: hmmm it worked here Nov 19 14:28:04 quickdev: this is correct Nov 19 14:28:04 fetch? Nov 19 14:28:19 quickdev: but it's meaningless at the same time Nov 19 14:28:32 since you will run into lots of troubles Nov 19 14:28:33 Ainulindale, http://rafb.net/p/qqj1ju43.html Nov 19 14:28:33 for once, you can either listen on UDP _or_ TCP Nov 19 14:28:53 which means, you need to run two system daemons on your device, if you want to have local operation still working Nov 19 14:29:13 and this is where it gets nasty Nov 19 14:29:30 I don't need local operations if it works on my host box. Is there any other issue you know of? Nov 19 14:29:48 wait wait Nov 19 14:29:48 you do need local operation Nov 19 14:30:07 on the device Nov 19 14:30:20 what for? Nov 19 14:30:34 intersubsystem communication Nov 19 14:31:00 what does that mean? :) Nov 19 14:31:04 two processes talking via dbus Nov 19 14:31:05 on the device Nov 19 14:31:10 if you only listen on tcp, they will never hear each other Nov 19 14:31:28 at least i had major problems getting that to work Nov 19 14:31:28 keep me updated Nov 19 14:32:07 i still think serial port forwarding would be more appealing though Nov 19 14:32:07 *shrug* Nov 19 14:32:07 but it just doesn't work *sigh* Nov 19 14:32:08 reliably, that is Nov 19 14:32:10 perhaps i should try another MUXer Nov 19 14:32:36 mickey|sofa, I'll try tcp things soon..would be the best solution in my eyes Nov 19 14:32:53 yeah, for everything except working on the muxer itself :) Nov 19 14:33:15 * mickey|sofa would love to run the frameworkd on the host Nov 19 14:33:31 quickdev: add the insecure download things to your local.conf please Nov 19 14:36:05 mickey|sofa, haven't thought about gsmmuxd yet...it's using a dbus interface and needs to run on the neo, right? argh.. :) Nov 19 14:36:13 yes Nov 19 14:36:31 unless... Nov 19 14:36:35 we could get serial port forwarding done :) Nov 19 14:36:47 * mickey|sofa hesitating to port the qt muxer Nov 19 14:36:55 don't want to touch C... Nov 19 14:37:03 meh:-) Nov 19 14:37:49 freesmartphone.org: 03mickey 07framework * r250d21919d93 10/framework/subsystems/ogsmd/gsm/const.py: ogsmd: add documentation for call type 128 Nov 19 14:39:09 Ainulindale, phyton-wifi compiled (btw...I started frmo scratch this morning) Nov 19 14:39:25 Again? Nov 19 14:39:25 You're mad Nov 19 14:39:31 mickey|sofa: almost done for the windows thingy Nov 19 14:39:31 :-p Nov 19 14:41:12 python-wifi ... Nov 19 14:43:54 freesmartphone.org: 03mickey 07specs * r9fad458d18ae 10/ (2 files in 2 dirs): odeviced: mention format-specific options in audio resource specifications for org.freesmartphone.Device.Audio.* Nov 19 14:45:25 Now, dbus-glib Nov 19 14:46:26 Ainulindale: Do you have time? You have said to adapt my shr.bbclass. Now I did that (http://pastebin.com/m516317f6) but my bitbake is still failing (http://pastebin.com/m243587d8). Nov 19 14:46:48 yeah I know Nov 19 14:46:48 that's because there's a SRCREV in the bb file Nov 19 14:47:00 I don't know how to fix that apart from totally reemoving the SRCREV of the file Nov 19 14:47:09 every file? Nov 19 14:47:09 alphaone, mickey|sofa : any wonderful idea? Nov 19 14:47:44 mickey|sofa: just so you know Nov 19 14:47:44 D:\git\framework\framework>python frameworkd Nov 19 14:47:50 2008.11.19 15:47:26 root INFO Installprefix is / Nov 19 14:49:07 okay, I will change than change every .bb file :( Nov 19 14:50:15 freesmartphone.org: 03mickey 07framework * ra56d84727315 10/framework/subsystems/testing/ (__init__.py testing.py): testing: add new subsystem solely for testing purposes Nov 19 14:51:03 Ainulindale: SRCREVs should never appear in recipes Nov 19 14:51:04 they are solely to be configured in .conf files Nov 19 14:51:04 (or things like sane-srcrevs.inc) Nov 19 14:51:33 mickey|sofa, how to include the git or svn revision in the package name otherwise? Nov 19 14:51:53 i meant an assignment Nov 19 14:51:53 SRCREV = foo Nov 19 14:51:57 using it on the right side is ok, of course Nov 19 14:52:49 back to your error... are you sure you defined SRC_URI as a git or svn repo? Nov 19 14:53:37 How do you mean? Nov 19 14:54:20 mickey|sofa, he used local sources. That's why the error appeared. Nov 19 14:54:23 oh Nov 19 14:55:02 Ainulindale wanted to ask wheter there's a way to make the recipes compatible...with changing from svn to local and back Nov 19 14:55:02 ok Nov 19 14:55:12 in that case i think we could perhaps enhance bitbake to ignore SRCREV on file:// sources Nov 19 14:55:19 need to take that up with the bb hackers Nov 19 14:55:28 as a quick workaround reomve the sanity check Nov 19 14:55:55 (grep for the error message, i doN't know the location offhand) Nov 19 14:57:19 that's the error message http://pastebin.com/m243587d8, but I don't find the sanity check Nov 19 14:58:07 did you grep? Nov 19 14:58:24 grep to what? Nov 19 14:58:34 for the error message Nov 19 14:58:43 in Oe and bb Nov 19 14:58:52 it might be somewhere in the fetchers Nov 19 14:59:14 (in bitbake) Nov 19 14:59:19 sorry, I don't have much experience with bitbake and I don't understand 'grep for error message in Oe and bb' Nov 19 14:59:46 dolf1074, just remove SRCREV from the bb recipes you're trying to compile for now Nov 19 14:59:49 quickdev: ok Nov 19 15:01:02 grep for error message means using the 'grep' and 'find' linux tools Nov 19 15:01:05 that way you can identify the file that where the error message is created Nov 19 15:01:16 and then patch that to remove the check Nov 19 15:03:19 any hint/links for a conf to get my opkg get packages in my shr/oe build tree ? Nov 19 15:04:35 mickey|sofa: frameworkd starts =) Nov 19 15:04:36 though it fails a lot :-p Nov 19 15:08:57 But I'll be able to work from my hotel in toulouse next time Nov 19 15:10:00 nice Nov 19 15:11:01 It still can't find fcntl Nov 19 15:11:59 Oh what the heck Nov 19 15:11:59 I'll just disable ogsmd Nov 19 15:12:43 Though I still think some dummy gsm modem should be developed Nov 19 15:12:47 Isn't there a gsm emulator available somewhere? Nov 19 15:14:24 (dummy gsm modem) sure thing, on our list of things to do Nov 19 15:14:31 looong list Nov 19 15:14:41 i want to connect it with the qtopia modem tools Nov 19 15:14:48 they're quite nice Nov 19 15:15:07 what can they do? Nov 19 15:16:36 a) they have a modem emulator talking AT Nov 19 15:16:51 b) they have a UI tool to "inject" several things such as sending CBs, SMSs, ringing et. al. Nov 19 15:17:20 Nice indeed Nov 19 15:18:23 mickey|sofa: modem emulator that acts over? Nov 19 15:18:37 when you need to emulate the voice..... Nov 19 15:18:37 ? Nov 19 15:19:01 mickey|sofa: would you happen to know how I can set a policy for dbus for all users? Nov 19 15:19:43 freesmartphone.org: 03mickey 07framework * r307aa69bf4bf 10/framework/ (resource.py subsystems/testing/testing.py): Nov 19 15:19:43 freesmartphone.org: testing: add SetResourceBehaviour method Nov 19 15:19:43 freesmartphone.org: add testing interface to cli-framework Nov 19 15:19:44 Ainulindale: no idea Nov 19 15:19:45 Same here =) Nov 19 15:19:51 methril: no voice of course Nov 19 15:19:59 but you can call the emulator Nov 19 15:19:59 mickey|sofa: is ResourceBehaviour a stub for the suspend thing we were talking about? Nov 19 15:19:59 and it will call back Nov 19 15:20:00 etc. Nov 19 15:20:12 very handy for automated testing Nov 19 15:20:39 well, it could be adapted to use it as a VoIP device Nov 19 15:20:39 Ainulindale: this is just a testing resource where I can exeriment with the suspendveto stuff. i need to make up my mind Nov 19 15:20:39 methril: heh, yeah Nov 19 15:20:52 then you could change from one source to other without changing the API :D Nov 19 15:20:53 mickey|sofa: My point was, earlier Nov 19 15:21:11 There are cases when an active resource may or may not be suspended depending on the app which requested it Nov 19 15:21:41 Thus a simple addition is a request resource with a prevent suspend flag Nov 19 15:21:51 It won't fix the "no suspend if()" case Nov 19 15:21:57 But it's simple for the GPS/navit etc Nov 19 15:22:23 i see what you mean, i wonder though whether this is too broad of a sword Nov 19 15:22:27 so if navit uses it Nov 19 15:22:28 we can never suspend the device Nov 19 15:22:45 if we stop suspending on a suspendveto of a subsystem Nov 19 15:22:45 Well mickey|sofa Nov 19 15:22:47 If you explicitely tell not to suspend a device Nov 19 15:22:56 I think this is a fair thing not to suspend Nov 19 15:22:56 :-) Nov 19 15:23:14 not to suspend at all? Nov 19 15:23:14 Yep Nov 19 15:23:31 well in that case i'd rahter not call the suspend call Nov 19 15:23:37 in the first place Nov 19 15:23:50 well, makes sense, if you are driving with your neo-GPS sw, you don't want the suspend Nov 19 15:23:50 well, i need to play with some of that behaviour Nov 19 15:23:59 methril: i think that's a too simple conclusion Nov 19 15:24:08 mickey|sofa: then you would need a rule to see if there's unsuspendable resources IMHO Nov 19 15:24:11 but that sums up to the same thing Nov 19 15:24:26 except that in your case, you check using the rules Nov 19 15:24:26 all this leads into customization hell Nov 19 15:24:30 too many parameters, too many cases :/ Nov 19 15:24:48 mickey|sofa: That's why I think it would be a better thing to leave this to the resource requester Nov 19 15:24:53 when i'm with the neo connected to the PC i don't like to get suspend Nov 19 15:24:56 If the requester wants to prevent suspend, it's user triggered Nov 19 15:25:02 So it's logical not to suspend Nov 19 15:25:24 Now if we think about specific cases such as "do not suspend while on call" Nov 19 15:25:25 that's another thing, indeed Nov 19 15:25:41 We (SHR) did this our way Nov 19 15:25:46 But a way to "rule" this out (sic') would be great Nov 19 15:26:02 Because after all, it's common sense Nov 19 15:26:41 but if youu choose that kind of approach you are going to get out of battery really fast Nov 19 15:26:53 mickey|sofa: and for that I think that if you want to prevent too many customizations, you would have to register some sort of "active" queue Nov 19 15:27:05 i.e. if an object is in this queue/set/whatever Nov 19 15:27:06 needs a more complex power management approach to save Nov 19 15:27:06 then resource is active Nov 19 15:27:22 methril: whatever you do, if you want to prevent suspend while a resource is active, it will be power consuming Nov 19 15:28:04 but while you are speaking, usually you didn't need the GPS, neither some other resources Nov 19 15:28:20 What if you receive a call while using GPS as a tracking device? Nov 19 15:28:47 And what if you start the GPS while on call, then release the call? Nov 19 15:28:47 It'll sum up to the same issue Nov 19 15:29:18 It'd be a lot easier if there was some way to manage counters in the rules Nov 19 15:29:27 or arrays Nov 19 15:29:47 (for that kind of stuff) Nov 19 15:29:47 then when you start to use one resource, you have to register that kind of Timer Nov 19 15:29:48 or an unsuspendable way to request a resource Nov 19 15:29:55 timer? Nov 19 15:29:55 why timer? Nov 19 15:30:17 s/timer/your array Nov 19 15:30:31 There are many answers to that kind of situation Nov 19 15:30:44 To me, the simplest is not to bother at all Nov 19 15:31:18 freesmartphone.org: 03mickey 07framework * rcaeedce5f51e 10/ (3 files in 3 dirs): enable testing subsystem Nov 19 15:31:20 what daemon manages the suspend? Nov 19 15:31:29 frameworkd Nov 19 15:32:42 mickey|sofa: I can't understand why this bugs you by the ay Nov 19 15:32:55 +w Nov 19 15:32:55 what? Nov 19 15:32:55 (having an unsuspendable resource) Nov 19 15:33:15 this leads to no transversal layer action I know about Nov 19 15:33:29 And it's pretty simple to solve quickly the basic stuff before any kind of real action Nov 19 15:33:43 I feel as if you're thinking about some problem I can't se :-) Nov 19 15:33:44 +e Nov 19 15:33:56 (damn keyboard & connection) Nov 19 15:34:14 i did not made up my mind about the suspend veto at all and you start tossing unsuspendable resources in, that's too quick for my mind :) Nov 19 15:34:25 Heh :-) Nov 19 15:34:30 then the question becomes a semantic one Nov 19 15:34:30 mickey|sofa: well anyway Nov 19 15:34:35 The question isn't to veto, to me Nov 19 15:34:46 does an unsuspendable resource prohibits suspending at lal Nov 19 15:34:47 or not Nov 19 15:34:47 The question is that the suspend rule is wrong Nov 19 15:34:47 that needs to be though tout Nov 19 15:34:47 logrotate works well!! Nov 19 15:34:48 And that's one thing Nov 19 15:35:11 So either you have a way to write true rules Nov 19 15:35:18 And this will become frameworkd business logic Nov 19 15:35:18 i.e. not customizable Nov 19 15:35:31 or you have a way to make the applications ask for that kind of logic Nov 19 15:35:49 yep Nov 19 15:35:49 that's the question Nov 19 15:35:49 Giving to the users the opportunity to script its suspend behaviour is dumb Nov 19 15:35:51 i think it has to be customizables Nov 19 15:36:05 Because everyone will use the same behaviour Nov 19 15:36:05 i have two hats here Nov 19 15:36:12 And we're falling once again on the suspend issue Nov 19 15:36:24 one hat is to make frameworkd self-contained in a minimal environment Nov 19 15:36:24 that's why i have the rules doing what they do Nov 19 15:36:29 What if I want to suspend/unsuspend on GPS or whatever? Nov 19 15:36:29 Though, to me Nov 19 15:36:29 the other one is providing services for apps Nov 19 15:36:35 which is the case for less rules Nov 19 15:36:38 This has to be defined on the requester side Nov 19 15:36:38 and more application vehaviour Nov 19 15:36:47 i.e. checkbox "authorize suspend while using navit" Nov 19 15:36:48 something like that Nov 19 15:37:12 (checkbox) ack. Nov 19 15:37:17 :) hey that's my usecase!! Nov 19 15:37:18 if apps were to use this feature, they need to do that Nov 19 15:37:18 I don't think scripting the behavior of the "veto" is wise Nov 19 15:37:45 Because in the end, it'll give you loads of work Nov 19 15:37:49 For something which will be, sure, customizable Nov 19 15:37:57 i'm not adding the veto to make it accessablefrom the rules Nov 19 15:37:57 But the same for everyone Nov 19 15:37:59 i'm adding it because we need it sooner or later Nov 19 15:38:18 Ok then, you want to add a veto Nov 19 15:38:19 The question is how :-) Nov 19 15:38:21 excactly Nov 19 15:38:22 And I'm bugging you with my considerations Nov 19 15:38:28 Is that it? :-p Nov 19 15:38:28 yeah Nov 19 15:38:41 heh :-) Nov 19 15:38:41 you too fast for me Nov 19 15:38:45 That's because I'm bored Nov 19 15:38:45 i just started doing some testing code Nov 19 15:38:50 because for one, we do not care about failing suspend atm. Nov 19 15:39:10 which is important before the veto thing Nov 19 15:39:12 so one step after another :) Nov 19 15:39:36 Right Nov 19 15:39:37 we have the problem w/ gsm for one Nov 19 15:39:38 I.e.? Nov 19 15:39:46 mickey|sofa: oh and Nov 19 15:39:47 for suspend, by the way Nov 19 15:40:18 lets say we try to configure the modem for suspend, but it responds with errors Nov 19 15:40:18 what now Nov 19 15:40:19 It would be a nice thing to have something like /etc/freesmartphone/suspend.d Nov 19 15:40:19 suspend anyways? Nov 19 15:40:19 retry? Nov 19 15:40:36 not usspend? Nov 19 15:40:36 (I'm thinking about pulseaudio here) Nov 19 15:40:41 Ainulindale: see our trac. iirc there's a task called "play apmd" there Nov 19 15:40:49 Yeah I see what you mean Nov 19 15:41:04 Heh :-) Nov 19 15:41:37 mickey|sofa: see that's what happens if I'm bored :-) Nov 19 15:41:37 -s Nov 19 15:41:38 hehe Nov 19 15:41:49 * mickey|sofa points to opimd ;) Nov 19 15:42:18 suspend/resume is a nasty crosslayer thing Nov 19 15:42:18 i hate it Nov 19 15:42:28 why can't we have better batteries? Nov 19 15:42:43 Why crosslayer? Nov 19 15:42:43 Ainunlindale: can I disturb once more ? Nov 19 15:42:43 You have to handle the suspend issues of specific resources, yes Nov 19 15:42:56 But you already have something to segregate that properly Nov 19 15:42:56 ptitjes: yes Nov 19 15:43:46 crosslayer because you have to work on every layer of the stack. unfortunately it's not just kernel (which would be cool), but also userland, and even there not containable in the service layer, since it needs to be application-aware Nov 19 15:43:56 anyways ,need to run some errands Nov 19 15:43:56 groceries and such Nov 19 15:44:05 bbl Nov 19 15:44:05 Can I add an enhancement ticket for the addition of cron and logrotate and a /etc/ file addition ? Nov 19 15:44:06 to the SHR trac ? Nov 19 15:44:06 mickey|sofa: don't forget to buy some coffee pills Nov 19 15:44:10 haha Nov 19 15:44:11 righto Nov 19 15:44:18 ptitjes: you can, but it'll be faster to ask me to give you git access Nov 19 15:44:19 did you see ticket #243? Nov 19 15:44:41 heu why not... Nov 19 15:44:46 ptitjes: IF you tried the generation Nov 19 15:44:56 And IF frameworkd.log rotates properlyu :-p Nov 19 15:45:18 but I should even open a ticket and close it myself so that you can have an history of the things done to the distro Nov 19 15:45:43 Ok, do that if you want to Nov 19 15:45:43 And mail me your pubkey Nov 19 15:45:50 logrotate does rotate frameworkd.log and ophonekitd.log properly!! :) Nov 19 15:45:56 That's a pretty good thing indeed :-) Nov 19 15:46:20 maybe those should be moved to /var/log again, no ? Nov 19 15:46:32 We'll check that later on Nov 19 15:47:08 also doesn't frameworkd deletes log at startup ? Nov 19 15:47:09 if yes that should be removed Nov 19 15:47:09 Ainulindale, ptitjes, do we really want persistent logs? Nov 19 15:47:18 atm they are cleaned and reboot Nov 19 15:47:27 *at reboot Nov 19 15:47:56 quickdev: I don't Nov 19 15:48:26 but now they are not cleaned during their use Nov 19 15:48:26 thus triggering a full memory if frameworkd is in debug Nov 19 15:48:44 and they can be compressed Nov 19 15:48:44 ptitjes: no it doesn't Nov 19 15:48:49 what it doesn't ? Nov 19 15:48:49 Ainulindale, full memory even if logs are cleaned at reboot? Nov 19 15:48:52 quickdev: my neo has been running for several days once Nov 19 15:48:58 frameworkd.log was ~10M Nov 19 15:49:11 (GPS outputs a lot of things) Nov 19 15:49:11 logrotate can rotate and compress and clean log after a certain period Nov 19 15:49:18 10mb = memory full? ;) Nov 19 15:49:27 quickdev: no, but this is something to do anyway Nov 19 15:49:27 this is cool if you want to have logs when an error occured Nov 19 15:50:15 yeah, ok :) Nov 19 15:50:16 so much time I had a bug of frameworkd (or ophonekitd) and restart made me loose that log!! Nov 19 15:50:49 (16:46:34) Ainulindale: ptitjes: no it doesn't ??????? Nov 19 15:51:05 ptitjes: the logs are on volatile Nov 19 15:51:11 the system cleans them at reboot Nov 19 15:51:13 frameworkd doesn't remove its log Nov 19 15:51:20 and this is normal Nov 19 15:51:29 ok /tmp is wiped out Nov 19 15:52:22 * polz is back Nov 19 15:57:07 well, time to go home Nov 19 15:57:07 see you Nov 19 16:01:23 Ainulindale: would it be possible to remove all the locales except en-us from the shr-lite images? Nov 19 16:01:23 Ainulindale: my key is in your mailbox Nov 19 16:05:56 oops Nov 19 16:05:57 i have do a little error Nov 19 16:06:14 i don't have compiled my edc Nov 19 16:06:40 http://scap.linuxtogo.org/files/651b5addf7e339ac9fce861d1f87b06e.png Nov 19 16:07:51 where is the frameworkd-config package in shr tree ? Nov 19 16:08:04 in oe tree ? Nov 19 16:08:52 yep Nov 19 16:09:13 polz: it would be, and it is a good idea, though some packages install their locales I think Nov 19 16:09:23 I should check how to do that Nov 19 16:09:31 Hire: pretty nice Nov 19 16:10:12 mickey|zzZZzz: when you'll be there again I'm interested in some info on opimd Nov 19 16:10:12 do you like that black? :o Nov 19 16:10:39 It doesn't bother me :-) Nov 19 16:10:48 Ainulindale: I've successfully removed the locales manually and everything seems to be working fine... Nov 19 16:10:49 polz: Yeah but the image generation is different :-) Nov 19 16:10:52 Ainulindale: and I've gained about 6M of disk space Nov 19 16:11:16 hmm.. I wish I knew what a home-built image looks like ;) Nov 19 16:11:34 Stills having errors then? Nov 19 16:11:56 yep. Nov 19 16:11:58 at least it failed yesterday Nov 19 16:12:10 if I want to give it another go, what should I do? Nov 19 16:12:18 Retry to make it and tell me the error Nov 19 16:12:34 shouldn't I update something? Nov 19 16:12:59 No Nov 19 16:13:54 OK, I've just started . setup-env; make image in the shr-testing dir. Nov 19 16:14:29 bitbake shr-lite-image instead of make image Nov 19 16:14:29 it'll be faster Nov 19 16:15:48 Hire: looking good, but one of them looks dangerously like a horizontally-flipped microsoft logo, exact same colours ;p Nov 19 16:16:07 Hire: didn't know you were a MS fanboi =p Nov 19 16:16:09 polz: your key is here Nov 19 16:16:10 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * rfafe2c680957 10/fsod/ (12 files in 5 dirs): Nov 19 16:16:10 freesmartphone.org: 1.) add config.vapi to wrap config.h Nov 19 16:16:10 freesmartphone.org: No hardcoding of paths in the code Nov 19 16:16:10 freesmartphone.org: 2.) Numerous cleanups in Makefile.ams Nov 19 16:16:10 freesmartphone.org: 3.) Add some command line options Nov 19 16:16:11 freesmartphone.org: Thanks to cbrake for the patch. Nov 19 16:17:30 Ainulindale: what key? Nov 19 16:17:52 err Nov 19 16:17:54 s/polz/ptitjes/ Nov 19 16:17:54 sicu: naaahh... it isn't a ms-scum :o Nov 19 16:17:57 anyway, it fails with ...shr-testing/openembedded/packages/git/git-native_1.5.2.3.bb' failed Nov 19 16:18:17 polz: ok, you have to edit the bb recipe Nov 19 16:18:17 it's borked Nov 19 16:18:22 there's a patch which can't be applied Nov 19 16:18:27 the output tells which Nov 19 16:18:32 remove it from the SRC_URI Nov 19 16:18:41 OK.. Nov 19 16:19:13 now... how exactly is this build system supposed to work and why are there constantly packages failing like this? Nov 19 16:20:00 you should ask that to stefan_schmidt :-) Nov 19 16:20:50 stefan_schmidt: help? :) Nov 19 16:21:29 I mean seriously, I've had this phone for a little over a year now and I've only been able to build an image for it _once_ Nov 19 16:21:31 Hire: have you got the build system going locally, or is that not needed for working on the theme ? Nov 19 16:21:46 and that took a whole day to finish. Nov 19 16:21:47 polz: that's normal Nov 19 16:22:00 yeah compiling images is a $@#%^# Nov 19 16:22:20 Ainulindale: it's normal with this distribution. Openwrt builds in about 20 minutes, even if you're using a svn snapshot. Nov 19 16:22:21 Hire: because looking at all the problems ppl have with setting it up, i can't really be arsed to do it here until it becomes a little straight forward :/ Nov 19 16:22:27 sicu: i have a right env to build an image but i don't use it for the theme :) Nov 19 16:22:35 polz: in debian build failure is serious policy violation and will get the package removed from the distro if not fixed before release :) Nov 19 16:22:48 we need to ask nasa if we can borrow their super cluster as a build host Nov 19 16:23:17 why nasa? Nov 19 16:23:28 lindi-: are you telling me to try and set up an emdebian buildroot? :) Nov 19 16:23:32 nasa columbia Nov 19 16:23:33 we have the roadrunner :D Nov 19 16:24:10 Hire: so do you have any links/tutorials/how-to's for working with the theme ? or did you did you just read the enlightenment docs ? Nov 19 16:24:12 http://www.nas.nasa.gov/Resources/Systems/columbia.html Nov 19 16:24:23 polz: http://iki.fi/lindi/openmoko/qemu-distcc.txt Nov 19 16:24:43 nice lindi- Nov 19 16:24:43 how can I git-commit a new directory ? Nov 19 16:24:50 i have just read the E docs :) Nov 19 16:25:02 and build E from svn Nov 19 16:25:02 Hire: cool ... thx Nov 19 16:25:19 polz: git add Nov 19 16:25:19 grmbl Nov 19 16:25:19 ptitjes: git add Nov 19 16:25:19 gracias Nov 19 16:25:23 de nada Nov 19 16:25:26 i don't know your knowledge but it is pretty easy to build E :D Nov 19 16:26:03 Hire: i have E built already =] Nov 19 16:26:03 instead is very hard to hack an edc file Nov 19 16:26:03 ahh.. ok Nov 19 16:26:25 you can play with edje_editor if you want but it is pretty limited Nov 19 16:26:38 and you can decompress an edje file with edje_decc Nov 19 16:27:00 next is you want to compile an edc files you have to run ./build.sh Nov 19 16:27:17 i prefer if you work on theme created by quickdev for shr apps Nov 19 16:27:56 EDC files are quite simple Nov 19 16:28:29 hehe, probably for a programmer, yes =p Nov 19 16:28:36 lindi-: I really, _really_ don't like the idea of having to use qemu in order to compile every package Nov 19 16:28:49 lindi-: just the thought of that makes me feel dirty Nov 19 16:29:42 polz: then struggle with your build failures :) Nov 19 16:29:57 polz: currently rebuilding without the locales Nov 19 16:30:05 polz: THough I expect you to thoroughly test it =) Nov 19 16:30:24 polz: the compiler is not running under qemu so it is not that slow Nov 19 16:30:25 quickdev: what about your build by the way? Nov 19 16:30:43 NOTE: Running task 3624 of 5671 (ID: 3136, /home/deluxe/shr/shr-testing/openembedded/packages/xorg-lib/libxpm_3.5.7.bb, do_compile) Nov 19 16:30:44 quickdev: damn :-) Nov 19 16:30:46 It's weird, my build was faster Nov 19 16:31:46 argh Nov 19 16:32:33 ok Nov 19 16:33:04 Ainulindale: grrr, that's emotional blackmail, getting me to test your images one after the other ;) Nov 19 16:33:42 polz: heh that's your fault Nov 19 16:33:42 You asked for locales removal Nov 19 16:34:57 does someone here have publish rights on openembedded ? Nov 19 16:34:57 * polz bows his head in shame. OK, OK, I'll do it Nov 19 16:34:58 ;) Nov 19 16:35:21 I have a patch for logrotate.bb Nov 19 16:35:40 btw, has anyone here built the latest enlightenment and SHR packages for Debian yet ? Nov 19 16:36:08 Ainulindale: I'm ready to push to shr git Nov 19 16:36:48 ptitjes: we will someday Nov 19 16:36:49 meanwhile stefan_schmidt has Nov 19 16:36:49 and mickey|zzZZzz Nov 19 16:36:58 stefan_schmidt: ? Nov 19 16:37:26 moin Nov 19 16:38:20 stefan_schmidt: I have a patch for logrotate bb recipe - it just removes an option to an install command Nov 19 16:39:59 someone else had the problem before but nothing commited yet : http://www.mail-archive.com/openembedded-users@lists.openembedded.org/msg00468.html Nov 19 16:40:00 ptitjes: Cosmetic or fix an issue? Nov 19 16:40:48 in fact this may be the host 'install' command that is used there so the -s option to install can't strip arm elf files Nov 19 16:40:56 so it fails to install Nov 19 16:41:09 ptitjes: hmm, strange that I don't see it failing here. Nov 19 16:41:35 see http://pastebin.com/m2e8fb8a5 for the patch Nov 19 16:41:35 ptitjes: Anyway, please send me the patch: stefan@datenfreihafen.org Nov 19 16:41:38 haaa Nov 19 16:41:41 :) Nov 19 16:42:05 do u yet want it by mail ? Nov 19 16:42:14 ptitjes: Will ask some other devs before commiting it. Will come back to you one way or another Nov 19 16:42:49 ptitjes: Well, as it is a one-liner pastebin is also fine Nov 19 16:43:25 okay i'd like to commit to shr the use of logrotate so I would need it asap if its possible Nov 19 16:43:39 :-] Nov 19 16:43:59 thanks anyway Nov 19 16:44:16 ptitjes: You have it fixed locally so no time pressure, right? Nov 19 16:45:04 no time pressure Nov 19 16:45:04 ptitjes: I'll ask some other people before commiting it. Not my part of expertise Nov 19 16:45:15 ptitjes: Also SHR can pick it up before if they have trouble. Nov 19 16:45:26 ptitjes: I'm just wondering why it does not fail for us here Nov 19 16:46:19 stefan_schmidt: same for git-native by the way Nov 19 16:47:01 I don't know either but the guy the mail archive (see above) never had any answer... Nov 19 16:47:05 yeah, git-native patches do not apply Nov 19 16:48:58 Ainulindale: Same strip problem? Nov 19 16:49:00 ptitjes: ok, give me the pastebin, I'll ask around Nov 19 16:49:00 why isn't there any post on the shr-devel mailing list ? Nov 19 16:49:16 stefan_schmidt: http://pastebin.com/m2e8fb8a5 Nov 19 16:50:17 stefan_schmidt, sorry to poke, but do you have any idea why get this? http://dpaste.com/91746/ Nov 19 16:51:17 Sup3rkiddo: huh Nov 19 16:51:41 Sup3rkiddo: Let me try to build it. Nov 19 16:51:52 stefan_schmidt, cool..thanks... Nov 19 16:56:50 Sup3rkiddo: ok, will eat now. Come back to this later Nov 19 16:57:16 stefan_schmidt, sire..bon appetit Nov 19 17:20:28 for me edc isn't so clear Nov 19 17:21:06 Hire, what's your problem with it? ;) Nov 19 17:21:33 nothing, I am studying it :) Nov 19 17:21:46 quickdev: currently cleaning the image of the useless stuff Nov 19 17:21:46 and i do some test Nov 19 17:22:12 Ainulindale, fine... :) Nov 19 17:23:24 quickdev: Found out our way to build it was less than satisfying Nov 19 17:24:52 Ainulindale, be more precise :) Nov 19 17:29:34 alphaone, I'd like to see the output of bitbake's do_configure. Do you know how to output that? Nov 19 17:29:39 the do_configure stamp is empty Nov 19 17:31:33 quickdev, theres a debug flag to bitbake you can use Nov 19 17:31:37 very verbose Nov 19 17:34:07 I alreayd tried the verbose flag. Didn't try debug flag...I'm doing it now :) Nov 19 17:34:23 quickdev: it's on log.configure Nov 19 17:34:48 ah, great Nov 19 17:36:59 quickdev: for your question about the way to build Nov 19 17:37:05 I wrote bitbake tasks Nov 19 17:37:12 To handle the things properly Nov 19 17:45:23 rwhitby: ping Nov 19 17:46:15 mickeyl: isn't there a problem on line 135 of parser.py of oeventd ? Nov 19 17:46:30 i'm not familiar with the code, but let me check Nov 19 17:46:57 I hope I broke the thing Nov 19 17:47:15 I like to break things Nov 19 17:47:20 ptitjes: no problem here Nov 19 17:47:34 (to me) Nov 19 17:47:43 Got a parse error with my IDE : Nov 19 17:47:44 alphaone, another question: Nov 19 17:47:44 NOTE: the following files were installed but not shipped in any package: Nov 19 17:47:44 NOTE: /usr/lib/navit/graphics/libgraphics_null.so.0.0.0 Nov 19 17:47:48 trigger = r['trigger'] if not while_rule else r['while'] Nov 19 17:47:48 how do I ship all things in /usr/lib/navit ? Nov 19 17:47:56 looks good Nov 19 17:47:56 quickdev: FIELS_${PN}= Nov 19 17:48:04 with FILES_${PN} +=? Nov 19 17:48:04 FILES Nov 19 17:48:04 ah, yeah Nov 19 17:48:28 ptitjes: looks ok to me, why do you think there's a probleM? Nov 19 17:49:10 (you know the new ternary operator in Python, right?) Nov 19 17:49:48 mickeyl: no! this is because I'm a python newbye Nov 19 17:49:50 ah Nov 19 17:49:51 s/newbye/newbie Nov 19 17:49:58 and also because my IDE don't parse this!!! Nov 19 17:50:09 [expr] if [bool] else [bool] Nov 19 17:50:09 :( Nov 19 17:50:10 err Nov 19 17:50:10 [expr] if [bool] else [expr] Nov 19 17:50:23 came with python 2.5 iirc Nov 19 17:50:23 perhaps 2.4 Nov 19 17:50:38 python 2.5 syntax is said to be supported Nov 19 17:50:38 pfff Nov 19 17:50:51 freesmartphone.org: 03mickey 07framework * r1cd7057890de 10/framework/resource.py: fix resource not being able to report errors correctly Nov 19 17:51:08 rwhitby: can i have access to downloads.freesmartphone.org ? Nov 19 17:51:23 rwhitby: i want to create a handpopulated stable dir Nov 19 17:53:16 polz? Nov 19 17:53:24 ku? Nov 19 17:53:42 Ainulindale: you've summoned me, oh master? Nov 19 17:54:37 polz: I'll commit my changes in some minutes Nov 19 17:54:43 And compile the lite image Nov 19 17:54:44 Will you be able to try it? Nov 19 17:58:13 Sup3rkiddo: hmm, fsod built here Nov 19 17:58:19 mickeyl: Any idea bout this one: http://dpaste.com/91746/ ? Nov 19 17:58:51 eeks Nov 19 17:58:51 stefan_schmidt, it did? the recipe works if i fetch it from the git repository...but borks if i use file:/// shall i pastebin the recipe i am using? Nov 19 17:58:53 quickdev: new proper way to build an image online Nov 19 17:59:02 looks like something in Makefile.am Nov 19 17:59:07 or configure.ac Nov 19 17:59:07 Ainulindale, thanks Nov 19 17:59:07 Sup3rkiddo: ah, only did git here of course Nov 19 17:59:08 quickdev: what are you thanking me for? Nov 19 17:59:24 Sup3rkiddo: did you build in the same dir for your host before? Nov 19 17:59:45 http://shr.bearstech.com/git/?p=shr-overlay.git;a=blob;f=openembedded/packages/images/shr-image.inc;h=23df8009e5564ec85b906651968c1f8e05ee4191;hb=96eeaf94fc5b6ec674ea539e70b581a5250c8e78 <= looks a lot nicer :-) Nov 19 17:59:58 Sup3rkiddo: Can you try with a fresh local checkout and file:// ? Nov 19 18:00:22 stefan_schmidt, yeah... Nov 19 18:00:52 stefan_schmidt: by the way Nov 19 18:00:52 Ainulindale: I'll do my best. Nov 19 18:01:05 Ainulindale: btw, make image just failed in ophonekit.bb Nov 19 18:01:49 How can we build a set of bbclass/bb recipes which could easily be switched from/to local/git Nov 19 18:01:58 polz: what now? Nov 19 18:03:13 cc1: error: ophonekitd-main.d: No such file or directory Nov 19 18:03:22 quickdev? :-) Nov 19 18:03:50 Do you remember what we did to fix that? Nov 19 18:04:55 polz: regenerating the gta01 image as we speak Nov 19 18:04:55 Ainulindale: hmm Nov 19 18:05:24 * stefan_schmidt points to mickeyl and zecke for bbclass stuff. Never touched them. :) Nov 19 18:05:25 Heh :-) Nov 19 18:05:26 Ainulindale: I get the idea. Nov 19 18:05:40 mickeyl? Nov 19 18:05:44 stefan_schmidt, fresh checkout works.........:/..man, this is embarrassing. I have been with this for a day now Nov 19 18:05:48 Ainulindale: is it planned to move shr settings into the illume wrench? Nov 19 18:05:50 Ainulindale, that's exactly my problem. I think I haven'T got a fix so far. Nov 19 18:06:16 polz: please pastebin cat conf/* Nov 19 18:06:20 bumbl: that's a question you should ask to raster Nov 19 18:06:25 but I don't think so, for now Nov 19 18:06:29 ah ok Nov 19 18:06:34 Though it is doable Nov 19 18:06:41 It's not on his topmost priority list :-) Nov 19 18:07:16 Sup3rkiddo: It's sometimes easier from an outer perspective. Have fun with it. :) Nov 19 18:07:16 because it would be better if we would have settings in a single, central place (I know this is low priority and a lot of work) Nov 19 18:07:22 bumbl: I don't agree with you :-) Nov 19 18:07:32 To me, things are better when they're redundant (in this case) Nov 19 18:07:32 Ainulindale: you can use base_conditional Nov 19 18:07:38 stefan_schmidt, thanks :D Nov 19 18:07:38 mickeyl: example somewhere? Nov 19 18:07:40 or write a full fledged python function Nov 19 18:07:54 Ainulindale: all over the classes Nov 19 18:08:01 mickeyl: errr Nov 19 18:08:06 mickeyl: problem here is Nov 19 18:08:06 Aj Nov 19 18:08:12 if SRC_URI contains git:// something Nov 19 18:08:12 and we replace it by file Nov 19 18:08:18 the bb recipe won't parse because of SRCREV in PR Nov 19 18:08:22 (PV sorry) Nov 19 18:08:36 ya, got it a couple of hours ago Nov 19 18:08:36 two solutions Nov 19 18:08:45 a) make bitbake ignore that case Nov 19 18:08:50 As you said before Nov 19 18:09:07 b) write a bitbake class that yanks SRCREV in the local case Nov 19 18:09:13 Ainulindale: as you are the coders what you say is law :) Nov 19 18:09:14 Aouch Nov 19 18:09:24 bumbl: my point is anyway Nov 19 18:09:24 bbiab, wifey comes home Nov 19 18:09:42 I agree this should be settable through illume Nov 19 18:09:46 bumbl: I'm just wondering "what about not using illume?" Nov 19 18:09:49 mickeyl: see you then :-) Nov 19 18:10:15 Ainulindale: http://pastebin.ca/1261543ž Nov 19 18:10:15 Ainulindale: http://pastebin.ca/1261543ž Nov 19 18:10:15 Ainulindale: http://pastebin.ca/1261543 Nov 19 18:10:24 yep Nov 19 18:10:24 sorry Nov 19 18:11:10 polz: seems right to me Nov 19 18:11:15 I had this problem, quickdev had it Nov 19 18:11:17 We settled it Nov 19 18:11:28 And I still can't see why it works at time and not at other Nov 19 18:11:31 Could you try removing the glibc locales? Nov 19 18:11:37 Or just try to do : Nov 19 18:11:40 (before anything else) Nov 19 18:11:41 where do I do that? Nov 19 18:11:44 bitbake ophonekitd -c clean Nov 19 18:11:49 bitbake ophonekitd -c mroper Nov 19 18:11:50 grmbl Nov 19 18:11:53 bitbake ophonekitd -c mrproper Nov 19 18:11:56 (clean and proper) Nov 19 18:12:10 then rm tmp/stamps/arm*/ophonekitd* Nov 19 18:12:24 then verify that there's nothing in downloads Nov 19 18:14:04 And if it still doesn't work Nov 19 18:14:05 remove the parallel make features and retry Nov 19 18:14:29 (and the bb number threads) Nov 19 18:15:02 polz: the jffs2 for shr-lite-image is there Nov 19 18:15:03 Though it's bigger than the last one Nov 19 18:15:16 I may have added things out of unluckiness :-) Nov 19 18:15:34 damn, I need to learn how to type faster Nov 19 18:15:54 polz: it may be because of the locales Nov 19 18:15:58 There's only the en-US locales Nov 19 18:16:07 You'll tell me anyway Nov 19 18:16:25 I'll reenable the testlab feature tonight just to see Nov 19 18:17:58 alphaone, FILES_${PN} += "${libdir}/${PN}/*" <- should that install /usr/lib/navit/graphics/test.so ? Nov 19 18:17:58 where are the downloads supposed to be? Nov 19 18:18:35 forget that, I've found it Nov 19 18:18:41 quickdev: yes Nov 19 18:18:46 in fact Nov 19 18:18:59 It should just declare the fact that the file is packaged in an ipk Nov 19 18:19:08 though installing it Nov 19 18:19:10 (though isn't the right word here) Nov 19 18:19:55 anyway...things that are installed with "install" in the makefile.am should be in the package as default, right? Nov 19 18:20:43 if you used autotools, yes Nov 19 18:21:12 Ainulindale: I'm downloading the image, this is going to take a few minutes. Nov 19 18:21:29 No problem here Nov 19 18:21:41 I may be gone by then but you'll leave me a message e:-) Nov 19 18:22:14 ERROR: QA Issue: non -dev package contains symlink .so: navit-elm path '/work/armv4t-angstrom-linux-gnueabi/navit-elm-0.1.0-r1/install/navit-elm/usr/lib/navit/graphics/libgraphics_evas.so' Nov 19 18:22:59 That just proves you'd better work on lfp-efl :-p Nov 19 18:23:06 obviously he does not like symlinks there Nov 19 18:23:20 quickdev: Will that be an elm frontend for navit? Nov 19 18:23:25 stefan_schmidt, yes Nov 19 18:23:29 ~hail quickdev Nov 19 18:23:29 ~praise quickdev Nov 19 18:23:33 Ainulindale, there you see Nov 19 18:23:40 heh :-) Nov 19 18:23:40 that's stefan_schmidt Nov 19 18:23:40 he's weird Nov 19 18:23:41 quickdev: I _always_ was wishing something like this Nov 19 18:23:47 Never came to it. :) Nov 19 18:24:36 quickdev: Let me now when you like to get some testing. I'm here. :) Nov 19 18:24:37 Ainulindale: And that from you. ;) Nov 19 18:24:37 stefan_schmidt: then try SHR Nov 19 18:24:39 If you have time Nov 19 18:24:44 Instead of being lazy Nov 19 18:24:47 stefan_schmidt, I need to solve the symlink problem Nov 19 18:25:02 Ainulindale: yeah, still on my list. I suck sometimes Nov 19 18:25:07 Ainulindale: lazy? Don't know this word. Nov 19 18:25:21 :-) Nov 19 18:25:28 stefan_schmidt: it's being generated Nov 19 18:25:46 hmm Nov 19 18:25:46 93M, damn Nov 19 18:25:57 Though the lite image will be enough :-) Nov 19 18:26:02 * stefan_schmidt gets off his ass and starts a download Nov 19 18:26:18 stefan_schmidt: wait a minute then Nov 19 18:26:22 Ainulindale: sure Nov 19 18:26:28 it's writing it as we speak Nov 19 18:26:42 ah, just ended Nov 19 18:26:50 quickdev: The error above? Nov 19 18:26:51 Ainulindale: after changing the config file, is there anything else I have to do to get bitbake to build one package at a time? Nov 19 18:26:51 Ah I actually reduced the image by 5mo Nov 19 18:26:57 (locales) Nov 19 18:26:58 polz: bitbake mypackage Nov 19 18:27:05 stefan_schmidt, yes Nov 19 18:27:05 (with . setup-env) Nov 19 18:27:13 Ainulindale: I don't care about size :) Nov 19 18:27:25 stefan_schmidt: I heard that somewhere Nov 19 18:27:46 Ainulindale: It's all technique Nov 19 18:28:11 Yeah, I'm sure about that :-) Nov 19 18:28:12 By the way Nov 19 18:28:14 quickdev: hmm, can pastebin a bit more context? Nov 19 18:28:14 I had a look at opimd Nov 19 18:28:14 OK, I've just run bitbake ophonekitd Nov 19 18:28:14 Ainulindale: And? Nov 19 18:28:14 Wasn't there any plan to include a journal? Nov 19 18:28:16 Calls, and such Nov 19 18:28:19 Because right now, there's no such thing Nov 19 18:28:27 And it's not explicit at all Nov 19 18:28:28 :( Nov 19 18:28:41 stefan_schmidt, I added FILES_${PN} += "/usr/lib/navit/* /usr/share/navit/*" to the package....and there are some symlinks in /usr/lib/navit/ Nov 19 18:28:41 "out of date cache found"... sure hope this isn't a bad omen Nov 19 18:28:42 We need one, yes. Nov 19 18:28:46 polz: not at all Nov 19 18:28:53 It's about time to get abraxa back from studies Nov 19 18:28:54 stefan_schmidt: so, isn't opimd the way to go for that? Nov 19 18:29:07 Good evening Nov 19 18:29:16 rotating slashes and dashes in the shell usually mean "you'll grow a beard before this is done" ;) Nov 19 18:29:26 Ainulindale: We still hope Nov 19 18:29:42 Ainulindale: PIM gives us a bit headache though. We need someone dedicated for this Nov 19 18:29:42 polz: I don't care about that, I already have a beard Nov 19 18:29:55 Nothing we can do next to the usual stuff Nov 19 18:30:04 stefan_schmidt: Well, my problem with PIM is that what abraxa did isn't self explicit Nov 19 18:30:07 A main would be of interest Nov 19 18:30:15 hmm Nov 19 18:30:28 (maybe I just didn't find it) Nov 19 18:30:34 If I have that I'll write test cases Nov 19 18:30:54 Ainulindale: Give me a moment with quickdevs problem Nov 19 18:31:08 No problem Nov 19 18:31:58 quickdev: Why do you want to add local files to the package? Nov 19 18:32:13 quickdev: Aren't they for x86? Nov 19 18:32:34 quickdev: Other way around. What do you want to do? Nov 19 18:32:43 stefan_schmidt, damn, you're right. But I already tried it with the variables ${libdir} Nov 19 18:33:41 stefan_schmidt, I'd like to get some modules into the package. They are normally installed to /usr/lib/navit/ Nov 19 18:33:41 quickdev: You what to add some more files into the package? That is the goal? Nov 19 18:33:41 ok Nov 19 18:33:42 but wait Nov 19 18:33:42 Then you should use ${D} IIRC Nov 19 18:33:48 quickdev: there's an example of that in one of our packages Nov 19 18:34:06 I tried it on my x86 and things were installed (modules) Nov 19 18:34:09 but when using bitbake that modules did not go into the package Nov 19 18:34:14 Ainulindale: it's still failing to build ophonekitd_git Nov 19 18:34:14 quickdev: http://shr.bearstech.com/git/?p=shr-overlay.git;a=blob;f=openembedded/packages/mofi/mofi_0.02.bb;h=70584df202dbf3372ba78f9a5afedeaf25da05ad;hb=HEAD Nov 19 18:34:49 polz: will check that in ten minutes Nov 19 18:34:56 polz: just tell me if it's still the same error Nov 19 18:35:31 stefan_schmidt, things that are installed with the install utility in an autotools makefile should be integrated into the package, not? Nov 19 18:36:29 quickdev: yes Nov 19 18:36:50 quickdev: Are you sure you listed the files correctly? Nov 19 18:37:03 In the autotools stuff that is Nov 19 18:37:53 it's the navit source package...at least their modules should be listed correctly <- anyway..how I said: I installed it on my local box and modules appeared in the folder they should have been installed to Nov 19 18:37:54 Ainulindale: yep, cc1: error: ophonekitd-main.d: No such file or directory Nov 19 18:39:30 quickdev: Only your new modules do not get packaged or also the navit ones already there? Nov 19 18:39:31 the ones already there are not installed, too Nov 19 18:39:33 quickdev: hu Nov 19 18:39:49 quickdev: That is the recpie from OE? Nov 19 18:39:58 I modified the recipe Nov 19 18:40:11 quickdev: Did it work before? Nov 19 18:40:56 polz: hmmmpf Nov 19 18:41:06 yeah...I should have a look what I changed Nov 19 18:41:09 quickdev: :D Nov 19 18:41:24 polz: that's something which the autotools are supposed to generate Nov 19 18:41:28 OK, I'm flashing the image now. Hope this works :). Nov 19 18:41:32 I'm away for an hour or so. Will be back later. We can talk then if the problem is still there. Nov 19 18:42:22 Ainulindale: which autotools, the ones on my sistem or is the buildsystem trying to use it's own version? Nov 19 18:42:28 polz: will you please pastebin ls tmp/work/armv4t-angstrom-linux-gnueabi/ophonekitd-0.0.1+r9-gitr* ? Nov 19 18:42:42 And then the git directory which should be there? Nov 19 18:43:03 polz: the one in your staging directory supposedly Nov 19 18:43:13 stefan_schmidt: happy wii'ing ;) Nov 19 18:43:32 Kensan: nah, not this time :) Nov 19 18:44:03 polz: here it compiles without any kind of problem Nov 19 18:44:26 polz: will you please try also bitbake --version? Nov 19 18:44:26 And tell me which version you're using Nov 19 18:45:10 stefan_schmidt: allright, but try to relax a little ;) Nov 19 18:46:15 quickdev, Ainulindale: I uploaded a new rasterpatches.tar.bz2 to the wiki Nov 19 18:46:15 with not applying patches removed Nov 19 18:46:30 mrmoku, great, thanks! Nov 19 18:46:43 just a sec Nov 19 18:47:06 quickdev: could you clean the two files? I don't have permissions to overwrite files :) Nov 19 18:47:19 mrmoku: good, thank you very much :-) Nov 19 18:47:26 stefan_schmidt: hilight Nov 19 18:47:32 stefan_schmidt: 19:46:10 < mrmoku> quickdev, Ainulindale: I uploaded a new rasterpatches.tar.bz2 to the wiki Nov 19 18:47:38 Should be of interest to you too Nov 19 18:47:45 hey, thank _you_ :) Nov 19 18:48:00 polz: I have to leave so be quick please :-) Nov 19 18:48:07 mrmoku, files cleaned Nov 19 18:48:36 ok, great Nov 19 18:49:02 quickdev: latest images are there by the way Nov 19 18:49:07 :) Nov 19 18:49:42 Ainulindale: 80ophonekitd git temp Nov 19 18:49:56 what do people think about starling? I find it quite finger-unfriendly... Nov 19 18:50:02 1.8.11 Nov 19 18:50:12 that's the bitbake version Nov 19 18:50:15 mrmoku, it is...and it has some segfualts ;) Nov 19 18:50:24 and no volume adjuster ;) Nov 19 18:50:56 so it would be worth to take a look at openmoko-mediaplayer2 ? Nov 19 18:55:40 heh Nov 19 18:56:07 /etc/X11/Xserver: cannot create /tmp/x.log: nonexistent directory Nov 19 18:57:13 did somebody see the openmoko-mediaplayer2 in action? Is it more finger-friendly? If it is worth the effort, I will take a look... Nov 19 19:04:32 quickdev: heh, you just removed the rasterpatches.tar.bz2? (from the wiki) Nov 19 19:05:26 mrmoku, you said I should clean the files...so that you can upload it with the correct name, because you haven't got permissions to overwrite a file? :) Nov 19 19:05:58 sorry, I ment clean up - remove the old one - rename the new one ;) I Will readd :) Nov 19 19:06:02 hello, is there the person who compiled fennec for the openmoko here...because I'm trying to compile it with org.openembedded.dev and it seems that the thing to fix are never ending...because we don't have automake 2.13 but 2.61 Nov 19 19:14:39 Ainulindale: with openmoko-shr-lite-image-glibc-ipk--20081119-om-gta01.rootfs.jffs2, the phone runs out of space during first boot. Nov 19 19:15:11 Ainulindale: I think you might have enabled some additional packages. Nov 19 19:17:10 freesmartphone.org: 03mickey 07stabilization/milestone4 * r34e935a5b762 10framework/framework/resource.py: Nov 19 19:17:10 freesmartphone.org: fix resource not being able to report errors correctly Nov 19 19:17:10 freesmartphone.org: Conflicts: Nov 19 19:17:10 freesmartphone.org: framework/resource.py Nov 19 19:22:52 is there any image that provides call forwarding to hands-free set, say, Sony Ericssons' HPM-64 ? Nov 19 19:30:32 how can I make audio output to come only out of earphones, rather than external speakers? Nov 19 19:33:24 mbuf: are you talking about the headphones that came with the phone? Nov 19 19:34:57 hi, someone's using MokoMakeFile with qemu on ubuntu? Nov 19 19:38:36 mbuf: it should automatically switch from speaker to earphone, if you plug the earphones in Nov 19 19:43:32 mrmoku: only if you have software to do it Nov 19 19:44:20 how could I change python path for frameworkd ? Nov 19 19:44:57 is it possible ? Nov 19 19:55:15 lindi-: yep, should have said that I'm talking about shr images Nov 19 19:59:28 polz: damn =) Nov 19 19:59:43 polz: my bet is that the locales are doing bad things Nov 19 19:59:44 polz: I'll enable the testlab Nov 19 19:59:55 Ainulindale: eating people's children? :P Nov 19 20:00:01 Could be :-) Nov 19 20:00:26 mrmoku: yes, the headphones that came with the phone Nov 19 20:00:47 mrmoku: for which image is that? I am using FDOM Nov 19 20:01:00 it's really weird. The older image works fine. Are you sure you didn't accidentaly enable some additional packages? Nov 19 20:01:50 polz: I just explicitely told it to use the en-us locales Nov 19 20:01:58 whereas before, no locales were set Nov 19 20:02:11 (apart from the glib one) Nov 19 20:03:03 mbuf: SHR or FSO should do it Nov 19 20:04:45 mrmoku: i see Nov 19 20:13:58 mrmoku: where can I find instructions on installing FSO image on Neo? Nov 19 20:14:07 i see the images here, http://downloads.freesmartphone.org/fso-testing/images/ Nov 19 20:15:17 never mind, found it here, http://wiki.openmoko.org/wiki/OpenmokoFramework/Status_Update_5#Installation Nov 19 20:15:36 re Nov 19 20:17:44 anyone, isn't the full python library provided in the python-2.5 package ? I miss urllib module... Nov 19 20:18:15 Ainulindale: I have a missing elementary-alarm (while bitbaking an image). Is this current state, or am I doing something wrong? Nov 19 20:18:15 what is the difference between the fso-image-om-gta02.jffs2 and openmoko-fso-image-glibc-ipk...gta02.rootfs.jffs2? Nov 19 20:19:05 mwester: the echo/buzz issues are amazing, did a test with my girlfriend today: we went through the city, she was 10-15 m away from me, sometimes she stopped with her phone and everything was perfect, then she stopped 5 meters away and the echo/buzz was there Nov 19 20:19:27 yes, that is exactly what would be expected. Nov 19 20:19:48 mwester: she went back at the "working" position and it was fine Nov 19 20:19:57 yep. Nov 19 20:20:59 Ainulindale: is there a real need for the ice icons? the new screenshots of hire are amazing! Nov 19 20:21:45 new scren? Nov 19 20:21:45 screen Nov 19 20:22:08 +s Nov 19 20:22:22 hire: http://scap.linuxtogo.org/files/651b5addf7e339ac9fce861d1f87b06e.png Nov 19 20:23:31 Hire: are they contained in the new image? Nov 19 20:24:38 mwester: you said that a bt headset should solve both problems right? Nov 19 20:24:47 yes Nov 19 20:24:52 mwester: did you tried it? Nov 19 20:25:38 ah... yes... i did it Nov 19 20:25:42 mwester: and which bt headset should i get? dont know much about bt ... has bt to be activated all the time, security issues there? Nov 19 20:25:56 Hire: nice, flashing Nov 19 20:28:14 mwester: i dont think that this issues will be solved on the freerunner Nov 19 20:31:22 Ainulindale, Hire, is the icon for numptyphysics fixed? It was missing sometime.. Nov 19 20:31:52 now, it isn't fixed Nov 19 20:31:57 i named it "star.png" Nov 19 20:32:10 you have to change the numpty*.desktop Nov 19 20:32:50 BluesLee: i have to study how works edc Nov 19 20:33:02 Hire, we can't change the .desktop files. Imagine your theme is installed on a different distribution. Nov 19 20:33:15 You should name the icon as it is defined in the .desktop file. Nov 19 20:33:21 ok? :) Nov 19 20:34:32 can a python fellow tell me where the hell I can find the collections module ? Nov 19 20:34:49 erhm... Nov 19 20:34:58 into numpty*.desktop Nov 19 20:35:06 Icon=star.png Nov 19 20:35:16 this is weird Nov 19 20:36:05 i named it star.png... it is right because into the .desktop the path is right Nov 19 20:36:56 Ainulindale: mickeyl will review some more of rasters patches on which the other depend. Hopefully we get this sorted out over the next days. Nov 19 20:38:28 Here am I again Nov 19 20:38:52 Ok, polz Nov 19 20:39:07 There's a testlab for GTA01 lite Nov 19 20:39:09 My guess is that it fills up memory writing the font cache Nov 19 20:39:34 quickdev: this should be fixed in the .bb file (numpty) Nov 19 20:39:56 THere was a path before Nov 19 20:39:58 But it is sed'ed on rootfs Nov 19 20:41:18 Ainulindale, mh? Nov 19 20:42:27 stefan_schmidt, btw...oe package is now working....but now I'm struggling with a segfault that happens on my neo, but not on the host :) Nov 19 20:42:44 quickdev: Ah, what was it? Nov 19 20:43:12 FILES_${PN} += "${libdir}/navit/* ${datadir}/navit/*" Nov 19 20:43:38 quickdev: gdb and -dbg packages are your friend. Helped me a lot with the illume segfault. And of course alphaone gdb knowledge Nov 19 20:43:39 yes, I know Nov 19 20:43:40 quickdev: yeah, thta looks better :) Nov 19 20:46:05 stefan_schmidt: by the way Nov 19 20:46:14 what are the size specified in the testlab features? Nov 19 20:46:15 -s Nov 19 20:46:26 grmbl, bad english Nov 19 20:46:26 http://shr.bearstech.com/shr-testing/images/neo1973/openmoko-shr-lite-image-glibc-ipk--20081119-om-gta01-testlab/installed-package-sizes.txt <= here Nov 19 20:46:32 There are sizes for the packages Nov 19 20:46:36 Ainulindale: No idea. Never used testlab Nov 19 20:46:36 Is it truly kb? Nov 19 20:46:55 A font is doing 10mo ! Nov 19 20:46:55 arphic-uming Nov 19 20:47:06 (mb) Nov 19 20:47:07 fonts are big Nov 19 20:47:41 Well I surely don't want that in my image :-) Nov 19 20:47:52 stefan_schmidt: and by the way Nov 19 20:47:52 I did a proper image Nov 19 20:48:00 It may be of use for you Nov 19 20:48:00 If you're interested Nov 19 20:48:01 You guys have tons of OE stuff for me everyday. When should I come to write code? :) Nov 19 20:48:01 (proper bb recipes) Nov 19 20:49:08 Ah, I see my problem, and for that I need your help stefan_schmidt Nov 19 20:49:08 See Nov 19 20:49:08 RDEPENDS_${PN}_append_om-gta02 = "\ ttf-arphic-uming \ Nov 19 20:49:08 " Nov 19 20:49:13 Though, I have the font on my gta01 package Nov 19 20:49:23 mickeyl: I don't understand how I must select what part of the core python library I need in my python build Nov 19 20:50:39 Ainulindale: hmm Nov 19 20:51:09 polz: fixing your issue / ... \ ... / Nov 19 20:51:10 Ainulindale: The only idea I would have is that it somehow sneaks in as 01 and 02 are in the same class of devices Nov 19 20:51:11 (This is a hasty picture of stefan_schmidt's train of thought using slashes and dots) Nov 19 20:51:15 Hire: okay, the topshelf icons are missing in the image, rest is there Nov 19 20:51:18 Ainulindale: But the append explicit for 02 should make this right Nov 19 20:51:34 stefan_schmidt: seems logical to me too Nov 19 20:51:36 hmm Nov 19 20:51:53 http://shr.bearstech.com/git/?p=shr-overlay.git;a=tree;f=openembedded/packages/tasks;h=efda4d6ae78392b7d01b7ca5c11afd7937306ca2;hb=HEAD Nov 19 20:52:01 It's in minimal-x Nov 19 20:52:01 i am working on topshelf icons... Nov 19 20:52:01 If you want to double check Nov 19 20:52:11 when I finished I will post some screen Nov 19 20:52:44 Hire: i like those of the last screenshot, what is the problem with them? Nov 19 20:52:50 and if quickdev or Ainulindale likes and wants them can put it into SHR Nov 19 20:52:58 Indeed Hire Nov 19 20:53:15 Ainulindale: thanks :) Nov 19 20:53:16 Ainulindale: I would like to postpone this. Poke me tomorrow again if it is still there. Nov 19 20:53:20 which screenshots? :) Nov 19 20:53:37 * stefan_schmidt needs to have a llok at the OpenEZX patches again which should get submitted next friday Nov 19 20:53:38 Hire: i think they must be put into the image otherwise there is no harmony Nov 19 20:54:07 but for now i have to finished it and next I will hack some edc files Nov 19 20:54:15 Ainulindale: I get again the problem with building python-etk. I tried the commands you suggested last time, but it won't work. What should I do? (http://pastebin.com/m71ba025e, compile log: http://pastebin.com/m561edc3f) Nov 19 20:54:24 the screenshots with the colored illume thing? Nov 19 20:54:28 That's weird Nov 19 20:54:37 mrmoku: FSO is nice; audio is good Nov 19 20:54:39 BluesLee: for that reason I am working on it Nov 19 20:54:39 quickdev: yes Nov 19 20:54:39 Looks like a missing dependency dolf1074 Nov 19 20:54:42 bitbake eina please Nov 19 20:54:57 stefan_schmidt: before you leave Nov 19 20:54:57 Hire, nothing against you, but the icons alone are much better :) Nov 19 20:55:01 see the previous log from dolf1074 Nov 19 20:55:01 I like the illume bar to be neutral Nov 19 20:55:14 quickdev: http://scap.linuxtogo.org/files/651b5addf7e339ac9fce861d1f87b06e.png Nov 19 20:55:14 * Hire is going to watch 1x08 of californication Nov 19 20:55:16 python_etk seems to lack some eina dependency Nov 19 20:55:23 quickdev: i like them colored;-) Nov 19 20:55:27 Ainulindale, eina was added recently so it might be the case ;) Nov 19 20:55:37 quickdev: sure Nov 19 20:56:16 Ainulindale: .bb : http://pastebin.com/m7944062d Nov 19 20:56:31 Huh? Nov 19 20:56:31 Why did you modify the .bb ? Nov 19 20:56:42 Ah it's shr-settings Nov 19 20:57:14 dolf1074: well anyway, this is a eina issue Nov 19 20:57:14 so bitbake eina please Nov 19 20:57:25 polz: image is regenerating Nov 19 20:57:40 And it's there polz Nov 19 20:57:42 You can download it now Nov 19 20:57:42 Ainulindale: hmm, why should python-etk depend on eina? Nov 19 20:57:52 polz: -10Mo Nov 19 20:58:09 stefan_schmidt: I don't know, but dolf1074 has an error Nov 19 20:58:10 it lacks eina.h Nov 19 20:58:27 http://pastebin.com/m561edc3f Nov 19 20:58:28 Ainulindale: It's just a python wrapper for etk, which instead should depend on it Nov 19 20:58:36 stefan_schmidt: line Nov 19 20:58:44 136 Nov 19 20:58:56 Hmmm though it may be python_evas Nov 19 20:59:08 Ainulindale, there's a user who did the python-elementary work Nov 19 20:59:08 (juste saw the no such file or directory) Nov 19 20:59:08 and he made a recipe Nov 19 20:59:17 quickdev: which work? Nov 19 20:59:34 Do you guys use AUTOREV for python-* Nov 19 20:59:34 ? Nov 19 20:59:34 could we allow him to use our repository? Nov 19 20:59:38 quickdev: what did he do? Nov 19 20:59:38 Ainulindale, elementary bindings for python Nov 19 20:59:38 IIRC there was some breakage Nov 19 20:59:38 Oh Nov 19 20:59:39 Of course he can Nov 19 20:59:47 stefan_schmidt: I don't think so Nov 19 20:59:47 hmm Nov 19 20:59:56 stefan_schmidt: I had this problem a week ago Nov 19 20:59:56 Forgot about it Nov 19 21:00:12 And after bitbake illume Nov 19 21:00:13 Let me check the other python-e* bb's Nov 19 21:00:13 (or e-wm) Nov 19 21:00:13 it went well Nov 19 21:00:14 Ainulindale, we need his public key, right? Nov 19 21:00:18 dolf1074: please try to bitbake eina and confirm if you can bitbake python-etk after that Nov 19 21:00:18 yes Nov 19 21:00:23 Ainulindale: I'm doing that Nov 19 21:00:23 quickdev: create an extra repo for him Nov 19 21:00:28 Ainulindale, for the recipe only Nov 19 21:00:28 but there are 400 tasks to do Nov 19 21:00:30 Oh Nov 19 21:00:30 Just the recipe Nov 19 21:00:30 That's good then Nov 19 21:00:30 so no access? Nov 19 21:00:40 No I thought the whole bindings Nov 19 21:00:42 If he wants the whole sources on our repo Nov 19 21:00:48 Let's create a repo for him Nov 19 21:00:55 heyho Nov 19 21:00:55 Ainulindale, morphis is it Nov 19 21:00:56 If he just wants to put the bb file somewhere Nov 19 21:01:01 Then it's good Nov 19 21:01:01 morphis: hi :-) Nov 19 21:01:11 dolf1074: then I guess you'll wait :-) Nov 19 21:01:17 morphis: What do you need exactly? Nov 19 21:01:17 morphis, I'm managing the repository with Ainulindale Nov 19 21:01:22 hey Ainulinda Nov 19 21:01:30 ok Nov 19 21:01:44 I have written the python-bindings for rasters toolkit elementary Nov 19 21:02:24 and it would be fine, if you could include it in the shr Nov 19 21:02:25 polz: please tell me if this image works Nov 19 21:02:43 morphis: that's no problem, but my question wasn't that Nov 19 21:02:44 What do you need? Nov 19 21:02:45 A repo for your sources or your bb recipe? Nov 19 21:02:54 Ainulindale: yup waiting on my comp. (Why isn't there a redistribution program for that. I have here 3 comps that are doing nothing :) So If I could divide the task, it would go faster :) ) Nov 19 21:03:28 ok I missunderstood you Nov 19 21:03:50 I only want you to include the bb-receipe in the shr-overlay for the daily builds of the shr Nov 19 21:03:58 a repository I already have Nov 19 21:04:03 It's not daily builds anymore but that's not an issue of course Nov 19 21:04:16 morphis: indeed, no problem with me Nov 19 21:04:25 perfect Nov 19 21:04:25 Though you don't need access for that Nov 19 21:04:30 Just mail us the bb recipe we'll commit it Nov 19 21:04:30 no I don't think so Nov 19 21:04:36 And quickdev and I are here most of the time so if you need changes... :-) Nov 19 21:05:06 morphis: Though right now, as I just told, we don't daily-build anything Nov 19 21:05:06 ok, then I'll add it if things are clarified :) Nov 19 21:05:06 I'm manually generating stuff Nov 19 21:05:10 I will send the bb-receipe to quickdev and tell you the svn rev Nov 19 21:05:14 No problem with that morphis Nov 19 21:05:14 hmm, python-evas has evas and eina as dep Nov 19 21:05:19 aye Nov 19 21:05:21 mickeyl: ? Nov 19 21:05:21 And if it goes well we'll be able to commit it to OE :-) Nov 19 21:05:40 this would be great :) Nov 19 21:05:40 stefan_schmidt: Then something went wrong :-) Nov 19 21:06:40 quickdev: you have post :) Nov 19 21:06:45 yeah, indeed :) Nov 19 21:07:21 mickeyl: You take care about python-efl in OE? Nov 19 21:07:41 stefan_schmidt: see though, dolf1074 tried to compile python-etk and it didn't compile eina as it's regenerating it as we speak Nov 19 21:07:52 mickeyl: Should eina be in DEPENDS in python-efl.inc? It seems all would need it Nov 19 21:08:30 Ainulindale: Got me. I'm now thinking were to put the dep. In the inc or separate for the bb's Nov 19 21:11:49 Ainulindale: I put it on my list. Please stay with building eina by hand as workaround for today. I'll take care of it tomorrow. Nov 19 21:11:54 * stefan_schmidt is now really off to other work. Nov 19 21:12:15 someone can help me to understand how python packages are produced in oe ? Nov 19 21:12:40 Ainulindale: bitbake -c python-etk is still failing Nov 19 21:13:23 I don't understand from where the python-codec package comes from - I would need python-netclient so I might like to do a similar package definition Nov 19 21:14:58 Ainulindale: bitbake command: http://pastebin.com/m571fcc59. Compile log: http://pastebin.com/m70efcfa1 Nov 19 21:17:51 Ainulindale: Do you know how core python package are builded ? Nov 19 21:17:56 s/package/packages Nov 19 21:18:45 dolf1074: bitbake -c clean python-etk; bitbake python-etk Nov 19 21:19:10 That may be due to autoconf not finding the headers, and you didn't relaunch the configure Nov 19 21:21:36 dolf1074: so? Nov 19 21:21:45 (sorry by the way for this really difficult thing) Nov 19 21:21:52 polz: interested in comments about latest image Nov 19 21:22:01 Ainulindale: I think he's doing fine Nov 19 21:22:01 Ainulindale, I added python elementary. Nov 19 21:22:12 quickdev: nice :-) Nov 19 21:22:12 but he has to generate 1800 tasks Nov 19 21:22:31 stefan_schmidt: confirmed, it was eina it seems Nov 19 21:22:31 Ainulindale, could you regenerate it? :) Nov 19 21:22:52 Ainulindale: wait a little bit longer. In about 5 minutes I know for sure Nov 19 21:23:19 quickdev: it's goin on Nov 19 21:23:20 great Nov 19 21:23:33 +gf Nov 19 21:23:34 Grmbl Nov 19 21:23:53 quickdev: did you fix your ophonekitd issue? Nov 19 21:24:27 quickdev: thinking about it I'm afraid it may be due to the fact we're not specifying the version of the autotools we need Nov 19 21:24:34 And in that case I'm interested in the output of configure Nov 19 21:24:48 yeah..could very well be the case Nov 19 21:24:49 does someone know where does the python-xml, python-codes, python-readline, ... come from ? Nov 19 21:24:57 ptitjes: from OE Nov 19 21:25:09 but there is not bb recipe Nov 19 21:25:22 ptitjes: some packages provide several packages Nov 19 21:25:25 I found a manifest-....inc Nov 19 21:25:57 freesmartphone.org: 03morphis * r326 10/trunk/python-elementary/ (3 files in 2 dirs): Update test application Nov 19 21:25:57 ptitjes: that's the one Nov 19 21:25:57 so how do you know where to search for a particular package say python-netclient ? Nov 19 21:26:04 you don't Nov 19 21:26:13 At least I don't Nov 19 21:26:13 I use grep Nov 19 21:26:25 ptitjes: see in the manifest.inc file Nov 19 21:26:30 there's a PACKAGES= directive Nov 19 21:26:42 that's the packages built by the recipe Nov 19 21:26:48 Ainulindale, how to specify autotools version? Nov 19 21:26:48 yes but opkg install python-client return -200xxx Nov 19 21:26:55 quickdev: no idea Nov 19 21:27:09 quickdev: but the output of configure will be of interest Nov 19 21:27:09 s/python-net/python-netclient Nov 19 21:27:10 ptitjes: using SHR? Nov 19 21:27:32 Ainulindale, where are the logs? Nov 19 21:27:47 quickdev: tmp/work/thingythingy/ophonekitd*/temp Nov 19 21:27:48 root@om-gta02 /usr/lib/python2.5 $ opkg install python-netclient Nov 19 21:27:48 An error ocurred, return value: -49900. Nov 19 21:28:18 ptitjes: please verify you're not using a proxy by mistake or anything like that Nov 19 21:28:18 and update Nov 19 21:28:37 a proxy by mistake ? i'm at home Nov 19 21:28:37 ptitjes: http://shr.bearstech.com/shr-testing/ipk/armv4t/python-netclient_2.5.2-ml0.1_armv4t.ipk Nov 19 21:28:46 It's there Nov 19 21:29:03 So update and try again Nov 19 21:29:08 can i bitbake the kernel uImage? Nov 19 21:29:44 cause i was hoping to add some special kmods Nov 19 21:29:53 Ainulindale: ok update works but what is this opkg result number!! Nov 19 21:30:03 Ainulindale: thx Nov 19 21:30:04 No idea Nov 19 21:30:44 I took one hour to search fo that package to build it... :( Nov 19 21:31:23 * mwester wanders in, kicks the trash cans over just to make a mess, and shouts "Good Morning!" Nov 19 21:31:31 mwester: hey there Nov 19 21:31:39 'allo Nov 19 21:31:40 your patch is in lfg, and generated on latest image Nov 19 21:31:48 Scary patch. Nov 19 21:31:48 polz says it's working nicely Nov 19 21:31:49 Though I can't understand why it segfaults in hash_destroy Nov 19 21:31:50 either has is null and it shouldn't do a thing Nov 19 21:31:50 With the patch, I don't see the number of messages printed in the log. Nov 19 21:31:51 or it isn't null Nov 19 21:32:05 So I think we get a bad dbus message. Nov 19 21:32:23 mwester: I didn't apply your "full" patch Nov 19 21:32:36 Right - debug noise. Nov 19 21:32:41 I just couldn't find the pastebin :-p Nov 19 21:32:45 So I put a test there Nov 19 21:32:45 mwester: morning :) Nov 19 21:32:54 The operative part was if we have a dbus_error non-null, we can't hash_destroy anything. Which implies that the hash argument was corrupted or bad in some way. Nov 19 21:33:05 mwester: indeed Nov 19 21:33:20 well anyway I'll do a bit of a review tomorrow for that kind of issue Nov 19 21:33:38 In between Nov 19 21:33:38 is it crashing? Nov 19 21:33:39 greeting O Mr. Kensan. Nov 19 21:34:02 No, it's not crashing anymore. Nov 19 21:34:39 BTW, on the kernel list, Andy says they changed "bat" to "battery" in the stable-tracking branch ------- to satisfy Android's hard-coded requirements. Nov 19 21:35:08 Meh :-) Nov 19 21:35:25 Just for gta01? Nov 19 21:36:23 No, they did it for gta02, and Andy want's to know if it should be "battery" for gta01 as well. Nov 19 21:36:35 Ok then Nov 19 21:36:35 Good to know Nov 19 21:36:56 Well I seriously have to lay my neck down Nov 19 21:36:57 It hurts like hell Nov 19 21:36:58 I'm feeling old Nov 19 21:37:11 See you in half an hour or so Nov 19 21:39:57 Ainulindale, do_configure: http://rafb.net/p/gxjYC464.html Nov 19 21:41:28 hey raster :) Nov 19 21:44:09 quickdev: yo Nov 19 21:44:56 hey raster Nov 19 21:46:19 quickdev: what's this mess about audiofile and libmikmod? Nov 19 21:46:30 mh? don't know Nov 19 21:47:27 raster, does evas_object_image_data_set(this->image, data); modify data in any way? Nov 19 21:47:45 raster: I am currently writing a python binding for your elementary widget set, the first version will be included in the next shr build Nov 19 21:48:08 raster: i have too many icons on one screen, i need more screens;-) Nov 19 21:48:09 morphis: already is Nov 19 21:48:58 Ainulinda: as a package? Nov 19 21:50:48 raster, already solved it. You don't have to answer :) Nov 19 21:50:54 quickdev: it tells evas that the image object's pixel data now lives at the pointer you gave it Nov 19 21:50:55 thats all Nov 19 21:51:48 \morplease... keep the binding 1:1 identical to the c api... unlike what everyone else likes to do (invent new calls and conventions, only implement 1/2 the api etc. etc.) Nov 19 21:52:08 BluesLee: scroll :) Nov 19 21:52:56 raster: I only encapsulate the c function calls in classes. the names are the same as in c Nov 19 21:53:57 raster: yeah, everytime i scroll i hit an icon and an app starts:-) Nov 19 21:55:29 raster: I keep as close to the rest of e python bindings as possible Nov 19 21:56:14 raster: introducing widgets should solve this or not? home1, home2 ... widgets Nov 19 22:00:02 so good night erverybody Nov 19 22:00:45 blu ? Nov 19 22:05:01 yet another thing I don't understand... collections.py imports a module _collections but I don't have that Nov 19 22:06:40 ImportError: No module named _collections Nov 19 22:15:13 Ainulindale: sorry, I left my computer there for an hour or so. I'll download it right away. Nov 19 22:16:27 Ok :-) Nov 19 22:20:02 Ainulindale: Yahoou! I succeded to remote debuging frameworkd's python code from Eclipse Nov 19 22:20:38 heh Nov 19 22:21:09 I have now to determine where to put breakpoints for the missing outgoing call signal Nov 19 22:22:38 You have this issue? Nov 19 22:23:15 btw, I've reinstalled yesterday's image, updated, upgraded, added the magic echo cancelation fix someone posted on the mailing lists... Nov 19 22:23:24 and now the messages app sometimes crashes Nov 19 22:23:31 Interested in the logs Nov 19 22:24:07 I'll try the latest image now... Nov 19 22:24:07 Ainulindale: I don't know who you are talking to Nov 19 22:24:07 I think mwester mentioned this issue with GTA01 Nov 19 22:24:08 On message received or something like that Nov 19 22:24:16 ptitjes: regarding what? Nov 19 22:24:42 (23:20:39) Ainulindale: You have this issue? Nov 19 22:24:48 That was for you Nov 19 22:24:51 The "no outgoing" signal Nov 19 22:24:57 Do you have this issue? Nov 19 22:25:13 I have this issue till the begining (I mean when I got a good SIM card) Nov 19 22:25:23 Ainulindale: I think it just crashed randomly, after a few restarts I received an SMS and then it crashed again :) Nov 19 22:25:23 polz: Systematic and reproducible? Nov 19 22:25:25 nope :/ Nov 19 22:25:38 ptitjes: same question for you Nov 19 22:25:39 polz: hmpf Nov 19 22:25:46 polz: if this happens again Nov 19 22:25:54 yes it is - I have no gui that show up Nov 19 22:25:55 please copy the frameworkd logs immediately somewhere Nov 19 22:26:01 and nothing weird by ophonekitd, at least as far as I could see Nov 19 22:26:07 will do. Nov 19 22:26:14 polz: And try to copy ophonekit logs too Nov 19 22:26:28 And I'd be interested in any kind of reproduction of this phenomena with logs Nov 19 22:26:38 But my bet is that it's slow to respond Nov 19 22:26:52 I will try the new image - BTW I can commit the logrotate stuff when you want Nov 19 22:27:18 Your key is already there ptitjes Nov 19 22:28:52 ha ok thx - should I wait for the oe logrotate patch to be applied or can you carry it in the shr tree ? Nov 19 22:29:05 If you commit the bb recipe with it I can carry it locally Nov 19 22:30:41 you mean adding a logrotate recipe to the shr git ? Nov 19 22:30:42 shr-overlay Nov 19 22:30:46 yes Nov 19 22:31:09 okay - go out with the dog and do that Nov 19 22:32:47 Ainulindale, have you seen the configure.log? Nov 19 22:33:14 Yep Nov 19 22:33:19 Would you please clean, then bitbake -c configure? Nov 19 22:33:20 I'll do the same here and we'll compare Nov 19 22:33:20 dolf1074: how is the build going? Nov 19 22:33:25 Ainulindale: It compiled ;) Nov 19 22:33:31 Nice :-) Nov 19 22:33:58 definitely :) Nov 19 22:35:43 booting image now... Nov 19 22:35:46 polz: STOP ! It'll fry your device ! Nov 19 22:35:46 No, I'm joking. Nov 19 22:36:15 (Just didn't wish you to unplug savagely) Nov 19 22:36:50 or maybe you aren't joking and laughing with his fried device :) Nov 19 22:37:23 Ainulindale: btw, on gta01, the image should never take up more than about 57M - there needs to be enough space for the user to install gllin which takes up a further 5M Nov 19 22:37:40 polz: then we'll pinpoint the problems by checking the testlab Nov 19 22:37:44 see what takes what Nov 19 22:37:58 polz: http://shr.bearstech.com/shr-testing/images/neo1973/openmoko-shr-lite-image-glibc-ipk--20081119-om-gta01-testlab/installed-package-sizes.txt Nov 19 22:39:20 the OM icon theme is quite big, for instance Nov 19 22:39:20 out of space again Nov 19 22:39:28 Damn Nov 19 22:39:47 which packages were you including before? Nov 19 22:39:48 and that font looks huge Nov 19 22:39:54 I removed it Nov 19 22:39:54 (already) Nov 19 22:40:17 polz: everything is on git Nov 19 22:40:27 http://shr.bearstech.com/git/?p=shr-overlay.git;a=blob;f=openembedded/packages/images/shr-lite-image.bb;h=415ed430766f67b726956175c3fb92e4d795546c;hb=HEAD <= Nov 19 22:40:32 then shr-image.inc Nov 19 22:40:36 and in the openembedded/packages/task Nov 19 22:40:46 you'll find the tasks and the packages installed Nov 19 22:41:05 Why is there a cpp package, I wonder Nov 19 22:41:25 X11, probably Nov 19 22:42:09 That's a compiler... Nov 19 22:42:17 used by edje-utils and cpp-symlinks Nov 19 22:42:29 I know, but AFAIK X uses it for some stuff Nov 19 22:42:29 dolf1074, how did you get it compile? Nov 19 22:42:33 quickdev: his problem was related to etk Nov 19 22:42:34 Ainulindale, I did clean before configure Nov 19 22:42:45 ah, ok Nov 19 22:43:03 I seel to remember X using it to compile keyboards or something Nov 19 22:43:03 quickdev: ok so now you have an empty dir with configure and not compile, right? Nov 19 22:43:12 Ainulindale: thanks for all the help so far. I have finally see my goal accomplished: edit a package, compile, install on my freerunner and run it ;) Nov 19 22:43:20 dolf1074: Sheesh :-) Nov 19 22:43:21 polz: no dependency here Nov 19 22:43:21 Ainulindale: but I don't see what could have changed in the last 2 days here Nov 19 22:43:23 * Package cpp is depended upon by packages: Nov 19 22:43:23 * cpp-symlinks Nov 19 22:43:23 * edje-utils Nov 19 22:43:28 Ainulindale, yeah Nov 19 22:43:37 polz: some stuff got bigger, some stuff changed, so well Nov 19 22:43:51 quickdev: ok then please pastebin me the list of your work directory (in the good S) Nov 19 22:44:12 Ainulindale: would a dump of list_installed on a working, upgraded image from the day before yesterday help ? Nov 19 22:44:34 quickdev: should be ls tmp/work/armv4t-angstrom-linux-gnueabi/ophonekitd-0.0.1+r9-gitr477+715f0f5e137b53dcf0c3e74409727d8cd3df1e1f-r9/git/ophonekitd/ Nov 19 22:44:44 polz: You can do that but it won't help me that much Nov 19 22:44:59 I'd rather compare what is useless here Nov 19 22:45:05 polz: but you can do it if you manage to pinpoint the changes Nov 19 22:45:11 I won't do it, not enough time Nov 19 22:45:23 Ainulindale, could you send me your ophonekitd folder? I've got a folder diff utility Nov 19 22:45:50 dir content: http://rafb.net/p/BFkdfs95.html Nov 19 22:46:02 quickdev: calaquendi.org/ophonekitd.tar Nov 19 22:46:14 quickdev: calaquendi.org/om/ophonekitd.tar Nov 19 22:46:39 Ainulindale: btw, talking about time - it's about time you went to bed ;) Nov 19 22:46:44 Nah Nov 19 22:46:53 Can't, won't Nov 19 22:47:07 quickdev: seems normal to me Nov 19 22:47:24 I'm trying -c compile Nov 19 22:47:33 Didn't talk about any .d Nov 19 22:47:53 And I now have the .o and the binary Nov 19 22:49:08 mickeyl: re: access to downloads.fso.org - please send me your ssh key Nov 19 22:49:34 quickdev: well, going outside for a while, keep me in touch Nov 19 22:49:34 polz: and you, young man Nov 19 22:49:43 please pinpoint the differences Nov 19 22:49:44 if that's not too much for you Nov 19 22:49:52 I'll be glad to remove the packages Nov 19 22:49:53 And for now, going out for a bit Nov 19 22:49:53 Ainulindale, wait Nov 19 22:50:02 Ainulindale: I don't think it's worth the effort to trim down an image for the gta01 flash right now -- it's quite reasonable to require an SD card on such a flash-limited device, and your time is too valuable to waste it on such trivalities today. Nov 19 22:50:02 JMO Nov 19 22:50:10 Ainulindale: I might be infantile, but unfortunately, I'm no longer young :) Nov 19 22:50:15 quickdev: yes? Nov 19 22:50:23 mwester: well I already have the correct setup for that Nov 19 22:50:24 So it's just a matter of doing dd Nov 19 22:50:37 ah, ok Nov 19 22:50:38 I'll leave soon Nov 19 22:50:39 quickdev: you had a question? Nov 19 22:50:40 Ainulindale: I didn't think you had a gta01 Nov 19 22:50:40 if we don't talk anymore..have a good night;) Nov 19 22:50:46 no ;) Nov 19 22:50:46 mwester: I don't Nov 19 22:50:55 mwester: that's why I can't test Nov 19 22:51:17 quickdev: ok, same to you then Nov 19 22:51:34 quickdev: please double check you're using bitbake 1.8.11 Nov 19 22:51:35 rwhitby: will do, thanks Nov 19 22:51:35 Well, right. I'm just pointing out that anybody can trim down the image to fit the GTA01, but only YOU can fix some of the other problems and bugs. :) Nov 19 22:51:44 mwester: then assign some tickets to me Nov 19 22:51:44 I'm bored Nov 19 22:51:44 And for now Nov 19 22:51:44 I'm off for ten minutes or so Nov 19 22:51:45 See yoy Nov 19 22:51:45 s/yoy/you/ Nov 19 22:52:21 * mwester assigns the "dialer has no idea a call is in progress" bug to Ainulindale Nov 19 22:55:55 Ainulindale, what's your libtool version? Nov 19 22:57:11 ok..that might not be the source. oe libtool should be used Nov 19 22:57:14 leave Ainulindale aloneeee Nov 19 22:57:20 leave him aloneeeeeeee Nov 19 22:57:28 ( and slap mwester for fun ) Nov 19 22:57:47 Hire, how long did the navit gtk take to redraw the screen when clicking on zoom in/out? Nov 19 22:58:11 * mwester adds "if (strcmp(username, "Hire") panic();" to the kernel. Nov 19 22:58:37 quickdev: depends on how much you want zoom Nov 19 22:58:48 mwester: !strcmp Nov 19 22:58:51 if you zoom a small region it takes 2-3 seconds Nov 19 22:58:54 hrr :) Nov 19 22:59:16 Hire, small = much zoom? Nov 19 22:59:17 mwester: Ok then I'll look at that tomorrow but my feeling is that it's more a job for mickeyl or alphaone :-) Nov 19 22:59:17 but if you want to zoom a city like Rome it will takes some minutes :) Nov 19 22:59:18 and less zoom? Nov 19 22:59:39 Ainulindale, dolf1074: mickeyl just pushed a change that should fix your eina problem Nov 19 22:59:48 Hire, ok, same here Nov 19 22:59:48 Ok good :-) Nov 19 22:59:49 yeah, much zoom :) Nov 19 22:59:50 stefan_schmidt: in between if you have an idea what a .d file could be missing for some and not for others Nov 19 22:59:53 quickdev: gtk is FUCKING slow Nov 19 22:59:54 Without any apparent reason Nov 19 23:00:09 To me this sounds like a dependency problem as building other stuff resolves the problem locally Nov 19 23:00:27 Hire, it's not gtk. It's cairo / the neo :) Nov 19 23:00:37 Ainulindale: hmm, you really hit every problem, right? ;) Nov 19 23:01:10 Ainulindale, I compared things. Almost no difference (x86, 64) Nov 19 23:02:34 oops Nov 19 23:02:40 my little mistake Nov 19 23:02:53 Hire, which maps are you using? Nov 19 23:02:53 reiseplanner Nov 19 23:02:55 mg Nov 19 23:02:57 is there a difference to OSM? Nov 19 23:03:23 yes... it is a tele atlas' map Nov 19 23:03:28 and it is complete... and closed Nov 19 23:03:28 I meant...speed-wise Nov 19 23:04:39 raster: evening Nov 19 23:04:40 it is big. the map is 1,3 gb and it contains Italy-Swiss Nov 19 23:04:46 i don't have tasted an osm map Nov 19 23:05:03 tested Nov 19 23:05:08 ok...I think it's too much data to render Nov 19 23:05:28 yeah :) Nov 19 23:05:28 this could be made faster to use less data Nov 19 23:05:41 you don't need smal streets when zooming out for example Nov 19 23:05:44 only the important things Nov 19 23:05:45 you are right :) Nov 19 23:06:14 that could mean some navit hacking..hehe Nov 19 23:07:11 stefan_schmidt: just curious: is the glofiish-ML "invite-only"? Nov 19 23:08:22 quickdev Nov 19 23:08:28 do you have tested the internal gui of navit? Nov 19 23:08:38 on my host, yes Nov 19 23:08:38 why? Nov 19 23:08:51 ptitjes: most python packages are built through the main package python_2.5.2.bb Nov 19 23:08:51 Kensan: At the moment yes. Should be changed soon. No idea when Harald has time for this. Nov 19 23:08:53 see python-2.5-manifest.inc Nov 19 23:08:55 g'night Nov 19 23:09:15 mickey|zzZZzz: night Nov 19 23:09:15 night mickey|zzZZzz Nov 19 23:09:24 do you want to do a thing like that with efl? Nov 19 23:09:42 stefan_schmidt: just stumbled on it and quickly wanted to check out the archives then saw that one had to register. Nov 19 23:10:03 an ui touch-friendly ( and fr-friendly ) ? Nov 19 23:10:04 Kensan: yes, will change Nov 19 23:10:13 Kensan: Nothing interesting on there anyway. Nov 19 23:10:30 stefan_schmidt: since I didn't win the X800 auction I don't have one anyway so... Nov 19 23:10:35 Kensan: All hacking is done in public. See Haralds blog posts and patches to om-kernel Nov 19 23:10:43 stefan_schmidt: yeah, I was just wondering. Nov 19 23:10:51 Hire, yes Nov 19 23:10:51 stefan_schmidt: yeah sure! No doubt about that Nov 19 23:11:00 that's why I'm dealing with navit Nov 19 23:11:08 nice :) Nov 19 23:11:13 * stefan_schmidt mumbles something about gnufiish.org Nov 19 23:11:13 :D Nov 19 23:11:17 Kensan: moo Nov 19 23:12:07 Kensan: Git repo will come once .tw internet is good enough for it. Harald already tried 4 times... Nov 19 23:14:55 stefan_schmidt: I may hit every problems Nov 19 23:14:56 But you're creating as much that's why :-p Nov 19 23:14:57 Ainulindale: heh Nov 19 23:15:11 Ainulindale: It's a pleasure for me doing so Nov 19 23:15:19 Well anyway the behavior is almost the same than python-etk Nov 19 23:15:30 IT fails until something gets compiled Nov 19 23:15:30 Then it works Nov 19 23:15:39 raster, is there a way to trigger full screen mode in elementary? Nov 19 23:15:48 So my guess is something is wrong in the dependencies Nov 19 23:15:50 It just won't generate a .d then complain about it Nov 19 23:16:07 (recipe is ophonekitd) Nov 19 23:16:17 stefan_schmidt: oh wow, sounds shaky... Nov 19 23:16:17 raster: what's up? Nov 19 23:16:55 stefan_schmidt: heh Harald starting a closed source project is right up there with Stalman installing Vista on his Macbook Pro *haha* Nov 19 23:17:39 apple sux Nov 19 23:17:39 Ainulindale: no idea Nov 19 23:17:41 microsoft sux Nov 19 23:18:37 stefan_schmidt: the thing is simple, it just uses autotools Nov 19 23:18:38 That's why I can't get that Nov 19 23:18:51 Snooze time here. Night guys. Nov 19 23:18:56 night :-) Nov 19 23:19:04 Ainulindale: Let us look tomorrow. To tired. Can't think Nov 19 23:19:04 Hire: no need to tell us, since we know :) Nov 19 23:19:21 stefan_schmidt: Go to sleep :-) Nov 19 23:19:22 stefan_schmidt: good night Nov 19 23:19:44 stefan_schmidt: I'll always find something to bug you with anyway Nov 19 23:20:12 raster: are all of your patches in oe now? Nov 19 23:20:13 quickdev: could you retry without -g3 -o0 by the way? Nov 19 23:21:52 And what's this -ldl in the makefile? Nov 19 23:23:22 Well anyway, going to watch House, I'll be there intermittently Nov 19 23:23:53 I'm doing "opkg install shr-settings_0.0.1-r0.1_armv4t.ipk -force-do Nov 19 23:23:54 wngrade -force-reinstall" and opkg still install the package online. Huh? Nov 19 23:25:07 quickdev: actually no.. there isn't. i can add one Nov 19 23:25:49 Kensan: gah... doing all sorts of cruft for the morning Nov 19 23:25:53 today.. is.. paperwork day Nov 19 23:25:53 raster, use case? navigation ssystem ;) Nov 19 23:26:29 raster: eww one of these days eh... Nov 19 23:26:44 quickdev: yes - i can see uses. i just havent done it Nov 19 23:26:53 ok, fine :) Nov 19 23:27:10 "games" Nov 19 23:27:10 although there are more important things...list widget for example :) Nov 19 23:27:10 "media players" Nov 19 23:27:10 ... Nov 19 23:27:22 fullscrene is trivial tho Nov 19 23:27:26 just hook up some calls Nov 19 23:28:18 great :) Nov 19 23:28:27 a lot of things are Nov 19 23:28:34 and a lot of thigns are mroe work Nov 19 23:28:36 i was doing entry stuff yesterday Nov 19 23:28:40 x copy and paste now all works Nov 19 23:28:53 i still havent solved splititng hilighting text vs scrolling tho Nov 19 23:29:55 raster: by the way Nov 19 23:30:09 Did I tell you the timeout settings were persistent now? Nov 19 23:30:31 (Checking, I don't remember if I did) Nov 19 23:31:16 (that is if you deal with them, frameworkd will not have to set brightness by itself) Nov 19 23:32:34 ok Nov 19 23:32:34 :) Nov 19 23:35:21 ahhh Nov 19 23:37:54 * Hire is pretty sad, as usual Nov 19 23:40:58 good night guys Nov 19 23:40:59 * sicu passes Hire a prozac Nov 19 23:41:05 nite quickdev ! Nov 19 23:47:31 i need a new life Nov 19 23:47:35 i hate this life Nov 19 23:47:50 but for now I have to go Nov 19 23:47:52 night guys Nov 19 23:59:16 heh ... that's killing the mood ... Nov 20 00:08:50 raster: just telling you that to stop you from whining about frameworkd hogging X :-) Nov 20 00:12:08 freesmartphone.org: 03tobias-git 07stabilization/milestone4 * re4f8bb524475 10framework/ (2 files in 2 dirs): Nov 20 00:12:08 freesmartphone.org: Add support for audio file options. Nov 20 00:12:08 freesmartphone.org: Options can be appended to audio file names, seperated via semicolons. Nov 20 00:12:08 freesmartphone.org: As an example, the ringtone now plays a different tune from the defaul Nov 20 00:12:15 freesmartphone.org: SID file. Nov 20 00:12:15 freesmartphone.org: 03mickey 07stabilization/milestone4 * rbfddd4c28c6d 10framework/ (2 files in 2 dirs): update ChangeLog, change default incoming SMS notification tone Nov 20 00:28:57 okay, incoming SMS rehave sound Nov 20 00:29:08 incoming calls ring Nov 20 00:29:41 I can call out Nov 20 00:30:50 GPS is still working Nov 20 00:30:53 Suspend works Nov 20 00:31:07 It's time to get this stupid update out Nov 20 00:52:07 It still won't tell the call direction... Nov 20 00:52:07 :( Nov 20 00:52:34 mwester: ? Nov 20 00:53:41 the problem where the dialer places a call, but the call progress messages don't have a direction associated with them. Works for most people, but a number of folks reported this. Nov 20 00:54:23 oh, okay Nov 20 00:54:37 mwester: Have you already reported that in fso trac? Nov 20 00:54:42 call direction... hanging a bit to the left Nov 20 00:55:38 I've reported to SHR; Ainulindale hasn't yet looked at it in detail but seems to think it's a framework issue. So, no I haven't put it in the FSO trac. Nov 20 00:56:22 Do you want it reported as I discovered it, or do you need me to place a call using mdbus commands ;) Nov 20 01:01:07 alphaone: The behavior isn't that Nov 20 01:01:11 There's no outgoing signal Nov 20 01:01:35 I bugged you times and times for that Nov 20 01:01:35 Some people have reported that Nov 20 01:01:39 namely mwester and ptitjes Nov 20 01:01:40 Watching the logs : no outgoing signal Nov 20 01:01:40 (full debug) Nov 20 01:01:48 I don't have any log at hand but mwester should be able to produce one Nov 20 01:03:09 Ainulindale: Sorry, I'm really preoccupied with that damn release Nov 20 01:03:15 It bugs me to no ends Nov 20 01:03:45 in fact I'm going to quit xchat and not come back until this stuff is off my mind Nov 20 01:04:00 hahaha Nov 20 01:07:13 * mwester starts to cry... Nov 20 01:07:17 waaaaah! Nov 20 01:13:57 alphaone: Don't be sorry I understand :-) Nov 20 01:14:01 alphaone: I saw you were preoccupied, don't worry :-) Nov 20 01:14:02 Well anyway Nov 20 01:14:03 I have to get some sleep Nov 20 01:14:21 See you tomorrow **** ENDING LOGGING AT Thu Nov 20 02:59:56 2008