**** BEGIN LOGGING AT Fri Oct 17 02:59:57 2008 Oct 17 05:10:48 freesmartphone.org: 03charlie 07framework * rde6c31e37c48 10/ (4 files in 2 dirs): oeventsd: DBusTrigger can be created from the rules file Oct 17 10:06:18 freesmartphone.org: 03charlie 07framework * ra72889abf344 10/framework/subsystems/opimd/ (9 files): opimd: settings merged with framework settings Oct 17 10:35:17 freesmartphone.org: 03mickey 07framework * r1c74c39b7662 10/framework/subsystems/ogsmd/modems/ti_calypso/unsolicited.py: ogsmd: [TI CALYPSO]: lower recamping detect factor to 0.3 :/ Oct 17 11:14:52 rwhitby ? Oct 17 11:15:09 . Oct 17 11:15:42 Heh :-) Oct 17 11:15:56 Do you know why all the packages are suffixed by .1 on the SHR buildhost ? Oct 17 11:16:10 Was there some change recently on OE ? Oct 17 11:17:06 (example : Upgrading bzip2 on root from 1.0.2-r1 to 1.0.2-r1.1... Oct 17 11:17:07 ) Oct 17 11:19:46 AinuFallback: I think this was probably zecke's --hash-style change for better shared library performance. Oct 17 11:20:34 Ok :-) Oct 17 11:20:39 Just wondering what it was Oct 17 11:21:13 basically, all the packages were recompiled with different linker options. Oct 17 11:21:40 Ok thanks :-) Oct 17 11:21:47 mickeyl: ping? Oct 17 13:16:02 Hey Oct 17 13:16:09 Ainulindale, back? Oct 17 13:17:32 rwhitby, here? Oct 17 13:18:06 . Oct 17 13:18:16 Makefile: needs update Oct 17 13:18:17 error: Entry 'Makefile' not uptodate. Cannot merge. Oct 17 13:18:21 could you correct that? Oct 17 13:18:26 it's on the buildhost Oct 17 13:18:40 quickdev: that's cause someone has manually edited the Makefile and so git cannot merge it. Oct 17 13:18:53 they need to git reset --hard HEAD in common Oct 17 13:19:04 ok, will fix it Oct 17 13:20:26 the .1 on all packages is due to OE adding a distro_pr and then incrementing it Oct 17 13:25:02 moin Oct 17 13:49:03 quickdev: is makefile the reason there's no autobuild for GTA02 today, only GTA01 ? Oct 17 13:49:44 quickdev: and tail-log is empty too :/ Oct 17 13:49:45 sicu, afaik there was issue with the kernel....images for gta02 should be there soon Oct 17 13:49:57 just building some things manually Oct 17 13:50:03 ah, ok then =] Oct 17 13:54:36 quickdev: fixed that prettyprinting thing, thanks Oct 17 13:54:44 fine :) Oct 17 13:57:58 has anyone encountered the following error? Oct 17 13:58:07 fatal: The remote end hung up unexpectedly Oct 17 13:58:07 make: *** [update-openembedded] Fehler 1 Oct 17 13:58:27 its caused by git pull in the openembedded directory Oct 17 13:58:39 but i have no idea how i could get rid of it Oct 17 13:59:40 Nobody96, you need the new version of the makefile Oct 17 13:59:55 cd shr/common ; git pull Oct 17 13:59:55 is there still a makefile-update? Oct 17 14:00:07 hmm no Oct 17 14:00:11 k Oct 17 14:00:26 oh, no common in my shr dir Oct 17 14:01:39 same error with new makefile Oct 17 14:03:16 ah, found the common dir Oct 17 14:07:02 quickdev: did you get my mail ? Oct 17 14:07:07 If not check please Oct 17 14:07:14 When exiting some screen, ophonekitd exits Oct 17 14:07:15 nope, even the newest makefile shows this error Oct 17 14:07:20 I have to go back to my conference now Oct 17 14:07:25 See you in an hour and a half or so Oct 17 14:07:47 ok, see you Oct 17 14:08:43 a direct "git pull" in the openembedded dir also results in "fatal: The remote end hung up unexpectedly" Oct 17 14:09:38 Nobody96, have you updated the common directory? Oct 17 14:09:45 yes Oct 17 14:11:13 i will check if the new OE location is used Oct 17 14:12:47 ok, the openembedded/.git/config is still pointing to the old location Oct 17 14:15:45 rm -rf openembedded and than make setup-openembedded will hopefully fix the problem Oct 17 14:21:29 ok, thanks quickdev its working again :) Oct 17 15:06:19 ah, finally something happening on the vala list wrt. async dbus calls Oct 17 15:29:16 quickdev: here again Oct 17 15:29:21 mickeyl: YOU THERE ! :-) Oct 17 15:29:44 kind of Oct 17 15:29:46 mickeyl: you probably didn't backlog, so is there any known problem about Messagebook capacities ? Oct 17 15:29:46 :) Oct 17 15:29:46 welcome back ;) Oct 17 15:30:01 AinuFallback: hmm, not that i know of. what prolbem are you having? Oct 17 15:30:02 AinuFallback, because of 16 messages? Oct 17 15:30:03 Because it tells me it has a 30 messages capacity Oct 17 15:30:12 And even with that Oct 17 15:30:25 quickdev's interface displays only 16 messages but it is a bug on our side Oct 17 15:30:30 (I guess) Oct 17 15:30:32 yes, it's our bug! Oct 17 15:30:36 don't spank mickeyl :) Oct 17 15:30:39 quickdev: our own proper bug Oct 17 15:30:41 I'm proud of us Oct 17 15:30:42 *phew* Oct 17 15:30:44 In any case Oct 17 15:30:44 :) Oct 17 15:30:49 I don't haave a 30 messages capacity Oct 17 15:30:51 I have way more Oct 17 15:30:57 Is there any way I can be certain of that ? Oct 17 15:31:00 sure Oct 17 15:31:11 open mickeyterm and do the following command for me Oct 17 15:31:18 Though now that I think about it Oct 17 15:31:25 I guess that's because it's stored in SIM and not in PIM Oct 17 15:31:28 AinuFallback, we somehow have to handle full message storage, because new SMS will not arrive if there's no space left. Oct 17 15:31:43 And I was reasoning considering what my previous phone did Oct 17 15:31:43 we would need the pimd to store the messages Oct 17 15:31:53 heh Oct 17 15:31:58 quickdev: opimd is going into frameworkd these days Oct 17 15:32:03 I'll try to help if I'm able to Oct 17 15:32:04 SIM.messagebook is of course limited by the SIM Oct 17 15:32:05 I saw the comits :) Oct 17 15:32:12 and 30 sounds perfectly valid for current SIMs Oct 17 15:32:17 but lets try the command: Oct 17 15:32:47 at+CPMS="SM","SM","SM" Oct 17 15:33:05 Hmmmmmm Oct 17 15:33:05 SM Oct 17 15:33:08 Sounds nasty Oct 17 15:33:20 +CPMS: 30,30,30,30,30,30 Oct 17 15:33:22 Ok then Oct 17 15:33:25 Sounds right to me :-) Oct 17 15:33:31 yep Oct 17 15:33:50 the "real" messagebook and phonebook will be covered by PIM then Oct 17 15:33:50 Now I'll spank you for the lack of opimd integration as of now Oct 17 15:34:01 heh, not my task Oct 17 15:34:04 I know :-) Oct 17 15:34:10 at last Oct 17 15:34:16 a multi-line entry widget... comes to life Oct 17 15:34:18 MUHAHAHAHAH Oct 17 15:34:19 mickeyl: we talked about that with charlie Oct 17 15:34:23 And raster as a matter of fact Oct 17 15:34:29 good Oct 17 15:34:35 quickdev: font size should be fixed for messages Oct 17 15:34:45 And you should have a "compose" option from contacts Oct 17 15:35:00 AinuFallback: all depends on the ui :) Oct 17 15:35:04 raster: my gift to you Oct 17 15:35:06 http://www.eyecandycan.com/ Oct 17 15:35:19 raster: And I would like your opinion on our UI if you have any kind of time Oct 17 15:35:23 HAHAHAHAHA Oct 17 15:35:25 Just to give some general advices Oct 17 15:35:29 hmm Oct 17 15:35:31 screenshots? Oct 17 15:35:33 raster, the text view in ETK is broken ;) Oct 17 15:35:34 or something? Oct 17 15:35:43 raster: Could you opkg install the package ? Oct 17 15:35:52 quickdev: not surprised. it didnt really ever work as best i know Oct 17 15:35:54 single-line yes Oct 17 15:35:57 multi-line not. Oct 17 15:35:59 I can do screenshots if necessary but well Oct 17 15:36:07 edje now can have textlbok parts editable Oct 17 15:36:09 they do work Oct 17 15:36:15 elementary exposes a nice entry widget Oct 17 15:36:31 AinuFallback: i have nothing flahsed with an fso upsteram feed Oct 17 15:36:41 sounds fine, maybe I should reimplement ETK's text widget based on that Oct 17 15:36:51 if u want a comment really soon/now Oct 17 15:36:54 or just using EWL, hehe Oct 17 15:36:54 raster: opkg install frameworkd ophonekitd libframeworkd-glib gsm0710muxd libframeworkd-phonegui-efl will do Oct 17 15:36:57 otherwise it'll need to go on my queue Oct 17 15:36:58 as it's Oct 17 15:37:03 2:37am Oct 17 15:37:03 :) Oct 17 15:37:11 And if you can't I'll try to do screenshots Oct 17 15:37:15 AinuFallback: no working feed Oct 17 15:37:19 the opkg config is broken Oct 17 15:37:24 shr.bearstech.com ? Oct 17 15:37:57 raster: http://shr.bearstech.com/shr-sources/openembedded/packages/images/shr-image/etc/opkg/ Oct 17 15:38:00 If you wish to try it Oct 17 15:38:28 whats the feed.conf line? Oct 17 15:38:28 I'd rather have your own feeling than your feeling based on my screenshot skills Oct 17 15:38:53 feed.conf line ? Oct 17 15:38:59 you have the whole file there Oct 17 15:39:08 quickdev: we should fix this theme issue Oct 17 15:39:17 I reinstalled some package and my theme was lost Oct 17 15:39:18 will need much mroe Oct 17 15:39:22 like gst packages etc. Oct 17 15:39:27 i only hjave my minimal illume images Oct 17 15:39:34 raster: everything is on SHR Oct 17 15:39:35 thats the problem Oct 17 15:39:49 It's mainly FSO plus our packages Oct 17 15:39:51 ooh wait Oct 17 15:39:52 conf Oct 17 15:39:53 there Oct 17 15:39:53 tnx Oct 17 15:40:01 now need wifi Oct 17 15:40:06 I knew you didn't look :-) Oct 17 15:40:11 Lazy ass of you ! Oct 17 15:40:13 AinuFallback, what's the cause of the theme issue? Oct 17 15:40:23 quickdev: some package is pushing its default.edje thing Oct 17 15:40:27 So white on white again Oct 17 15:40:34 We should either package a new thing for the image Oct 17 15:40:54 Or build a new package for that Oct 17 15:41:04 Your choice, I don't know the details Oct 17 15:42:06 quickdev: as I'm totally wasted I don't know if I'll be available tonight Oct 17 15:42:14 So please fire your questions now in case I'm not Oct 17 15:43:27 AinuFallback, about which theme do we talk? Ui apps? etk? illume? Oct 17 15:44:21 the default.edje file you gave me Oct 17 15:45:22 Downloading http://shr.bearstech.com/shr-testing/ipk/all/Packages.gz Oct 17 15:45:23 ... Oct 17 15:45:25 waiting... Oct 17 15:46:11 AinuFallback, I don't think that the file is the problem...try an autobuilder image Oct 17 15:46:29 Ok will try then quickdev Oct 17 15:46:31 raster: resolv.conf Oct 17 15:46:43 Or ipforward Oct 17 15:46:43 its fine Oct 17 15:46:45 (just to check) Oct 17 15:46:47 bah Oct 17 15:46:49 wifi is fucked Oct 17 15:46:53 Heh Oct 17 15:47:11 Why use wifi when you can have a nice and cancer-free connection ? Oct 17 15:47:16 (:-p) Oct 17 15:47:42 hmm Oct 17 15:47:44 pings work Oct 17 15:48:01 ping shr.bearstech.com then ? Oct 17 15:48:47 yeah Oct 17 15:48:53 but opkg update is stsalled Oct 17 15:48:57 oh bolucks Oct 17 15:49:05 this is to much for 3:45am Oct 17 15:49:10 err Oct 17 15:49:12 2:45am Oct 17 15:49:43 That's what you have when you use too much UIs Oct 17 15:49:49 It's tiring for the eye Oct 17 15:49:59 :-> Oct 17 15:50:31 ok Oct 17 15:50:33 slightly happier Oct 17 15:51:39 AinuFallback, I'm searching a way for deploying the apps I'm developing more quick Oct 17 15:52:03 atm I'm bitbaking a package on my local host, transfer them to my neo...install and test them...but that's veery slow Oct 17 15:52:06 any suggestions? Oct 17 15:53:26 quickdev: develop natively with x86 and use xnest Oct 17 15:53:53 quickdev: scp binaries Oct 17 15:53:55 that's what I do Oct 17 15:54:03 AinuFallback, and that's slow Oct 17 15:54:08 Not when you do sh scripts Oct 17 15:54:20 AinuFallback: ok missing pakcages in repo Oct 17 15:54:22 packages Oct 17 15:54:32 can't install Oct 17 15:54:35 If you don't like that try to mount your toolchain directory in NFS Oct 17 15:54:35 * ERROR: Cannot satisfy the following dependencies for frameworkd: Oct 17 15:54:35 * frameworkd-config * Oct 17 15:54:35 * Cannot find package libframeworkd-glib. Oct 17 15:54:35 * Cannot find package libframeworkd-phonegui-efl. Oct 17 15:54:35 * Cannot find package frameworkd-config. Oct 17 15:54:49 Heh frameworkd-config is not in all but in gta02 Oct 17 15:55:08 http://shr.bearstech.com/shr-testing/ipk/om-gta02/ Oct 17 15:55:22 and I'm sorry it's libframeworkd-glib0 Oct 17 15:55:27 and libframeworkd-phonegui-efl0 Oct 17 15:55:32 (this thing bugs me) Oct 17 15:55:32 u had them commented out Oct 17 15:55:33 :) Oct 17 15:57:01 ok Oct 17 15:57:08 messing up my beautiful image Oct 17 15:57:08 :) Oct 17 15:57:13 raster: not at all :-p Oct 17 15:57:20 raster: steps to launch the screens are : Oct 17 15:57:28 /etc/init.d/frameworkd start Oct 17 15:57:29 and Oct 17 15:57:33 DISPLAY=":0" ophonekitd& Oct 17 15:57:42 And you're good to go it should launch itself automatically asking for PIN Oct 17 15:58:04 And waiting for calls and such if PIN has already been asked for Oct 17 15:58:55 oooh crap Oct 17 15:58:56 more Oct 17 15:59:02 Collected errors: Oct 17 15:59:02 * Cannot find package libframeworkd-glib. Oct 17 15:59:02 * Cannot find package libframeworkd-phonegui-efl. Oct 17 15:59:02 * Cannot find package libframeworkd-glib. Oct 17 15:59:11 0 missing? Oct 17 15:59:28 Yep Oct 17 15:59:33 17:55 < AinuFallback> and I'm sorry it's libframeworkd-glib0 Oct 17 15:59:33 17:55 < AinuFallback> and libframeworkd-phonegui-efl0 Oct 17 15:59:44 raster: Congrats on the front-page /. article! Oct 17 15:59:59 wurp2|working, which article? Oct 17 16:01:00 * raster adds 0's Oct 17 16:03:07 ooh bugger Oct 17 16:05:18 AinuFallback, is it somehow possible to connect the libframeworkd-glib on my x86 host to my neo's dbus? Should be possible, shouldn't it? Oct 17 16:05:22 AinuFallback: no dbus available Oct 17 16:05:24 not working Oct 17 16:05:33 ugh' Oct 17 16:05:36 Heh, you have to start dbus then Oct 17 16:05:37 Listen Oct 17 16:05:38 It's late Oct 17 16:05:42 You're obviously tired Oct 17 16:05:46 I'll do some screenshots Oct 17 16:05:48 Don't bother with that Oct 17 16:06:02 quickdev: I don't know :-) Oct 17 16:06:06 hmm Oct 17 16:06:08 frameworkd dies Oct 17 16:06:10 somewhere Oct 17 16:06:18 dbus is up Oct 17 16:06:30 44 1295 1 0 Oct14 ? 00:08:28 /usr/bin/dbus-daemon --system Oct 17 16:06:32 etc. Oct 17 16:07:19 well that's weird Oct 17 16:07:33 quickdev: http://shr.bearstech.com/trac/ticket/28 Oct 17 16:07:38 Do you know about this echo fix thing ? Oct 17 16:08:00 maybe the call echo? frameworkd issue? Oct 17 16:08:29 we should fix some trac tickets, hehe Oct 17 16:08:35 Yep Oct 17 16:08:38 Just doing that in fact Oct 17 16:09:49 ok Oct 17 16:09:52 reboot and still not happy Oct 17 16:09:54 not sure whats up Oct 17 16:10:00 tail frameworkd.log ? Oct 17 16:10:05 i'd need to dig deeper Oct 17 16:10:20 where is it? Oct 17 16:10:31 I think default is to syslog Oct 17 16:10:33 so logread -f Oct 17 16:11:14 nothing in log Oct 17 16:11:16 naada Oct 17 16:11:38 well Oct 17 16:11:46 aha! Oct 17 16:11:49 log_to = file Oct 17 16:11:49 log_destination = /tmp/frameworkd.log Oct 17 16:11:52 in frameworkd.conf Oct 17 16:11:54 ImportError: No module named optparse Oct 17 16:11:59 heh Oct 17 16:11:59 poackaging problem Oct 17 16:12:01 deps Oct 17 16:12:02 Yep Oct 17 16:12:04 mickeyl ! Oct 17 16:12:08 Free hilight Oct 17 16:14:16 python-textutils Oct 17 16:14:20 is the missing offender Oct 17 16:15:28 mickeyl: regarding the echo "fix", did you modify frameworkd ? Oct 17 16:17:03 ugh Oct 17 16:17:08 difflib Oct 17 16:17:08 quickdev: bug there Oct 17 16:17:13 answered to call quickly Oct 17 16:17:28 and I had the incoming call redisplayed over the active call window Oct 17 16:18:05 and gst Oct 17 16:19:16 AinuFallback, please ticket Oct 17 16:20:56 ok Oct 17 16:20:59 fwd try again Oct 17 16:23:20 ugh Oct 17 16:23:23 needs sim Oct 17 16:24:02 AinuFallback: ok. it's doing absolutely nothing Oct 17 16:24:34 give me screenies Oct 17 16:26:46 frameworkd didn't launch ? Oct 17 16:26:51 Did you launch ophonekitd ? Oct 17 16:27:42 fwd is running without complaint Oct 17 16:27:49 ophonekitd just stis spweing complaints Oct 17 16:27:58 like dbus not available etc. Oct 17 16:28:39 Huh ? Oct 17 16:28:45 DISPLAY=":0" ophonekitd isn't working ? Oct 17 16:28:47 That's weird Oct 17 16:28:47 raster: are you putting the phone apps into your illume-image? Oct 17 16:28:52 "dbus not available try in 5 sec" Oct 17 16:28:57 dbus si definitely there Oct 17 16:29:06 Ah ! That's normal Oct 17 16:29:12 Wait for frameworkd to settle down Oct 17 16:29:23 And check that ogsmd has been launched Oct 17 16:29:28 fwd has been running for about 5+mins Oct 17 16:29:31 its settled Oct 17 16:29:37 system is essentially idle Oct 17 16:29:48 is ogsmd launched ? Oct 17 16:29:51 nup Oct 17 16:29:52 no ogsmd Oct 17 16:29:58 Hence ophonekitd won't launch Oct 17 16:30:07 err, attach to ogsmd Oct 17 16:30:15 inf act Oct 17 16:30:18 ogsmd doesnt exist Oct 17 16:30:22 do you have a SIM ? :-) Oct 17 16:30:33 If you do, please check that gsm0710muxd is launched Oct 17 16:30:35 frameworkd logs.. Oct 17 16:30:48 If it isn't, kill frameworkd, then launch gsm0710muxd then launch frameworkd Oct 17 16:30:51 yup Oct 17 16:30:53 gsmmux is ther Oct 17 16:30:56 But it should launch gsm0710muxd by itself Oct 17 16:30:56 eno ogsm on the system Oct 17 16:31:03 raster: that's a frameworkd issue then Oct 17 16:31:21 try to kill frameworkd, relaunch gsm0710muxd then relaunch frameworkd Oct 17 16:31:24 And check if ogsmd is there Oct 17 16:31:25 just get screenies :) Oct 17 16:31:32 I thought that much :-) Oct 17 16:32:12 thee is no ogsmd executable anywehre Oct 17 16:32:23 python sit packages for subsystems for fwd Oct 17 16:32:41 Hmmm quickdev Oct 17 16:32:43 Ticket isn't your fault Oct 17 16:32:51 raster: ogsmd is a subsystem of frameworkd Oct 17 16:32:58 mdbus -s will tell you if it's there Oct 17 16:33:17 quickdev: I have two signals for incoming Oct 17 16:33:40 Oct 16 01:42:47 om-gta02 user.warn 2008.10.16 01:42:47 ogsmd WARNING (can't read from modem. giving up) Oct 17 16:34:15 so there is no way for me to test or see anything without gsm/sim all being 100% perfect it seems Oct 17 16:34:22 screenies Oct 17 16:34:30 in fact Oct 17 16:34:31 sleep Oct 17 16:34:46 spent almost an hour just to see ui images Oct 17 16:34:49 raster: i'll do screens don't worry Oct 17 16:34:50 :) Oct 17 16:34:59 Thanks anyway for your time Oct 17 16:35:22 tomorrow Oct 17 16:35:29 i have a 10am conf call Oct 17 16:35:30 3:30am Oct 17 16:35:36 me <- not happy Oct 17 16:36:22 mickeyl: two incoming signals are sent for an incoming call Oct 17 16:36:42 (i have logs !) Oct 17 16:41:51 quickdev: http://mobile.slashdot.org/article.pl?sid=08/10/16/2055235 Oct 17 16:42:05 sorry for the slow response; actually working for once Oct 17 16:45:51 raster: and you wanna do that conf call on the freerunner? *heh* Oct 17 16:49:44 quickdev: did you fix the quitting issue ? Oct 17 16:50:14 AinuFallback, incoming sms quitting? no, just setting up a wiki page for fast development for us. Oct 17 16:53:16 quickdev: I think I experienced this elsewhere but I'm not convinced about that Oct 17 16:53:45 quickdev: we should focus on the settings application though (and pim after that) Oct 17 16:53:54 yes Oct 17 16:54:04 Especially silent mode Oct 17 16:54:05 btw..there's a settings application in ASU...not? Oct 17 16:54:10 I don't know Oct 17 16:54:16 In any case it doesn't use frameworkd Oct 17 16:54:20 (yet) Oct 17 16:54:40 yeah..but we could use that code..if it's available somewhere Oct 17 16:54:54 let's not reinvent the wheel Oct 17 16:54:57 I guess it is Oct 17 16:54:58 indeed Oct 17 16:55:00 But problem is Oct 17 16:55:12 Most of the settings thing is frameworkd specific Oct 17 16:55:21 AinuFallback, could we change the trac theme to standard? The current one is not very good.. Oct 17 16:55:21 So in my eyes it will have to be developed Oct 17 16:55:57 quickdev: do as you wish Oct 17 16:55:58 I don't care Oct 17 16:56:31 quickdev: nice thing by the way on the wiki Oct 17 16:56:36 Didn't know we could do that Oct 17 16:56:40 It's interesting, indeed Oct 17 16:56:45 what thing? Oct 17 16:56:50 DBus thing Oct 17 16:56:59 yeah...I will complete it in some minutes Oct 17 16:59:42 damn mickeyl, still doing stuff but not answering :-) Oct 17 17:01:12 AinuFallback, could you write bearstech to change the theme of trac to standard? I'd need ssh access Oct 17 17:02:01 We won't have SSH access Oct 17 17:02:08 yeah, so they need to do it Oct 17 17:02:27 Write me what you need and what you want to do and I'll do the messager :-) Oct 17 17:03:09 AinuFallback, in order to make libframeworkd-glib use a custom dbus address, we need to pass the address somehow.....we could do that through an environement variable or with a program parameter (which would be passed to frameworkd-glib..argc/argv)....what do you think? Oct 17 17:04:23 I think an environment variable would be more suitable...because it's a setting that describes the environment, not? Oct 17 17:04:31 quickdev: I don't think this is a must do Oct 17 17:04:41 what is not a must do? Oct 17 17:04:42 I think dbus is configurable on the client device Oct 17 17:04:54 (on the .service thing) Oct 17 17:04:57 But I'm not sure about that Oct 17 17:05:06 I don't think so.. Oct 17 17:05:15 could you find that out? Oct 17 17:05:18 have to leave in 3 minutes.. Oct 17 17:05:21 yep Oct 17 17:08:07 AinuFallback, what when different clients want to connection to different dbus daemons? Wouldn't work with a config file in /etc/dbus-1, right? Oct 17 17:10:38 I don't know at all :-) Oct 17 17:11:38 so implement the environment variable thing ;) Oct 17 17:14:31 quickdev: in fact I think we just have to modify the DBUS_SYSTEM_BUS env var Oct 17 17:14:58 ah, yeah..try it :) Oct 17 17:15:04 have to leave now Oct 17 17:15:05 Or something like that Oct 17 17:15:07 Just see that Oct 17 17:15:11 DBUS_SESSION_BUS_ADDRESS=unix:abstract=/tmp/dbus-vrWWxpQCa7,guid=125fb844b1df495f044fd92448f8917a Oct 17 17:15:12 girlfriend ;) Oct 17 17:15:17 Ok good night :-) Oct 17 17:15:18 yeah..that's it Oct 17 17:15:26 Except we use the session bus Oct 17 17:15:29 system Oct 17 17:15:30 see you later(12) or tomorrow Oct 17 17:15:46 ok see you Oct 17 19:36:50 hmmm, i just flashed the latest SHR image, and it's not playing nice Oct 17 19:36:58 *restart* Oct 17 19:38:23 * mwester has flashbacks to horrors from the help-desk days -- "It's not playing nice!" -- response from helpdesk "Is it plugged into the wall?"... :D Oct 17 19:38:59 sicu: what does "not playing nice" mean? Oct 17 19:42:17 mwester: heh ot cure those flashbacks you should probably turn back to your java project ;) Oct 17 19:49:40 mwester: heh ... sorry to bring up horrors from the past :p Oct 17 19:51:36 mwester: hey there Oct 17 19:51:46 I read somewhere about a echo fix and it was talking about you Oct 17 19:51:55 Do you have any information on that ? :-) Oct 17 19:53:38 mwester: cannot recieve sms, cannot send sms (stays stuck on 'loading'), cannot make call, cannot pick up call (pressing answer does nothing, it just keeps ringing until caller hangs up) Oct 17 19:54:01 sicu: latest ? Oct 17 19:54:05 That's weird it's supposed to work Oct 17 19:54:09 I'm using it here in fact Oct 17 19:54:21 What does "cannot make call" means ? Oct 17 19:54:24 Do you have logs ? Oct 17 19:54:28 yeh ... flashed the GTA02 image of today Oct 17 19:54:37 Or is it because you're trying to call someone with your phone under your shoe ? Oct 17 19:55:02 Gimme logs, lazy bastard ! :-) Oct 17 19:55:13 in the middle of a reflash atm ... so will get back to ya ;) Oct 17 19:55:22 Does it ask for pin ? Oct 17 19:55:34 sicu: it may be that you have an unstable version of frameworkd Oct 17 19:55:34 it did after a few mins .... Oct 17 19:55:56 If it asks for PIN though, everything is supposed to work Oct 17 19:56:07 As long as illume will display the network Oct 17 19:56:10 It will work Oct 17 19:56:27 well. perhaps it was playing me ;p Oct 17 19:56:44 it was working fine with an old SHR image (few days ago) Oct 17 19:56:46 sicu: I seriously use it on a day to day basis for a week Oct 17 19:56:52 So I guess that's some ponctual problem Oct 17 19:57:17 yeh ... i will check things out when the flash is done Oct 17 19:57:24 ok then Oct 17 19:57:31 sicu: If you have any problem feel free to ask Oct 17 19:57:35 I'm there to help Oct 17 19:57:59 Ainulindale: cheers =] Oct 17 19:58:27 related to SHR of course Oct 17 19:58:33 I can't help you with anything else :-p Oct 17 19:59:06 hehe ... i've been using only SHR for over a week now ... so i'll stick to it =] Oct 17 19:59:19 I guess I'll just have to put up with my erectile dysfunction, then Oct 17 19:59:26 sicu: oh by the way Oct 17 19:59:32 sicu: Installed on SD, or flash? Oct 17 19:59:33 If you have suggestions for the content of the image Oct 17 19:59:36 regarding packages Oct 17 19:59:39 feel free to tell Oct 17 20:00:01 I was thinking about doing something similar to alphaone Oct 17 20:00:20 and that would be...? Oct 17 20:00:27 wurp2|working: flash Oct 17 20:00:32 Ah, you didn't try his image then Oct 17 20:00:45 A lot of useful packages Oct 17 20:01:05 Ainulindale: nah, must've mist that one :/ uri? Oct 17 20:01:05 sicu: When I tried flash, and I did opkg install -force-reinstall blah blah, it ran out of space Oct 17 20:01:40 Ainulindale: In that case I recommend a flash image + an SD image Oct 17 20:01:45 For my convenience ;-) Oct 17 20:02:00 Or, far better, do flash only, but do it the way mwester was suggesting Oct 17 20:02:40 Where you put the rootfs on the flash drive as a file, and the kernel as a file, and uboot is smart enough to mount them Oct 17 20:02:51 So you don't have to format your flash drive every time you upgrade :-( Oct 17 20:03:30 wurp2|working: sounds like a good idea Oct 17 20:04:20 well, gotta go to a meeting now, but will be back to test this later .... cya Oct 17 20:04:20 I agree, wish it had been mine Oct 17 20:04:27 Actually, he stole it from Palm Oct 17 20:04:31 have fun Oct 17 20:52:44 freesmartphone.org: 03mickey 07specs * r8910efcb0c91 10/ (10 files in 5 dirs): ousage: add error specs Oct 17 20:56:33 mickeyl: hey Oct 17 20:57:22 http://shr.bearstech.com/trac/ticket/34 <= mickeyl Oct 17 20:57:28 SHould be of interest for you Oct 17 21:00:30 hmm Oct 17 21:00:35 i can't see any problem in the dump Oct 17 21:00:40 it's incoming Oct 17 21:00:42 then active Oct 17 21:00:44 then release Oct 17 21:04:36 Check agin Oct 17 21:04:38 +a Oct 17 21:04:39 Two incoming Oct 17 21:04:59 Line 37 and line 50 Oct 17 21:05:27 With different parameters orders Oct 17 21:05:46 (which may be irrelevant but heh) Oct 17 21:06:15 that's just twice reporting that the same call is incoming Oct 17 21:06:27 you'll get one signal per modem RING Oct 17 21:06:32 mickeyl: signal is received twice Oct 17 21:06:37 (on the ophonekitd part) Oct 17 21:07:22 I checked lfg and ophonekitd and wasn't able to see any mistake there Oct 17 21:07:32 sure, the signal comes frequently until you actually make it active or release Oct 17 21:07:39 the ID is still '1' Oct 17 21:07:45 Ah, that is an interesting information Oct 17 21:07:53 [the signal being fired multiple times is stated in the docs :)] Oct 17 21:08:07 Wasn't when I first developed lfg maybe then :-) Oct 17 21:08:11 Hey Oct 17 21:08:15 possibly :) Oct 17 21:08:24 mickeyl: That's good to know in any case Oct 17 21:08:29 quickdev: Oct 17 21:08:30 23:07:33 < mickeyl> sure, the signal comes frequently until you actually make it active or release Oct 17 21:08:30 yep Oct 17 21:08:33 23:07:40 < mickeyl> the ID is still '1' Oct 17 21:08:36 still working? fine! :) Oct 17 21:09:14 the active signals comes frequently? Oct 17 21:09:22 no, "incoming" will come frequently Oct 17 21:09:31 (on every modem RING) Oct 17 21:09:37 quickdev: you're lucky Oct 17 21:09:47 You'll have to do a ghashtable implementation :-) Oct 17 21:09:54 Get to work :-p Oct 17 21:10:03 why that? Oct 17 21:10:10 to keep track of incoming calls Oct 17 21:10:20 for a list as small as that, walking an int array may be faster Oct 17 21:10:29 if ghashtable is a pain in the ass Oct 17 21:10:32 wurp2|working: yeah and managing memory manually Oct 17 21:10:33 * mickeyl notes it a hash may be overkill for a maximum of 3 connections Oct 17 21:10:44 2 even Oct 17 21:10:49 I'd rather skip memory management Oct 17 21:10:49 at least on GSM Oct 17 21:10:53 VoIP may be another story Oct 17 21:10:57 Even if I understand what you're telling Oct 17 21:11:16 Ainulindale, couldn't you do it? :) Oct 17 21:11:35 quickdev: My eyes are trying to rip themselves off their orbits Oct 17 21:11:46 That's why I don't think it is a good idea :-p Oct 17 21:11:54 not today.. Oct 17 21:12:03 Ok will do then Oct 17 21:12:11 quickdev: in any case I found our solution for dbus things Oct 17 21:12:28 tell me :) Oct 17 21:12:28 mickeyl: do you know if it 's possible to replace easily the dbus-glib DBUS_SYSTEM_BUS constant ? Oct 17 21:12:37 (i.e. env var ?) Oct 17 21:14:30 I thought you found the solution? :) Oct 17 21:14:38 I found the solution Oct 17 21:14:41 Not the means to use it Oct 17 21:14:44 Ainulindale: uh no idea Oct 17 21:14:51 why would you want to do that? Oct 17 21:15:05 mickeyl: to be able to use frameworkd dbus on a remote system Oct 17 21:15:15 hence to develop faster (no deployment) Oct 17 21:15:37 ah Oct 17 21:15:45 yah, been toying with this idea as well Oct 17 21:15:53 tried some things Oct 17 21:15:57 but ended up not getting anywhere Oct 17 21:16:09 mickeyl: well Oct 17 21:16:12 tcp-support in dbus is just too broken / underdeveloped Oct 17 21:16:16 Oh Oct 17 21:16:19 I thought it was working Oct 17 21:16:21 Well Oct 17 21:16:21 argh.. Oct 17 21:16:22 not really Oct 17 21:16:26 That puts an end to quickdev's idea then Oct 17 21:16:27 you will run into major problems Oct 17 21:16:39 there's another low hanging fruit though Oct 17 21:16:43 i'd rather have someone fixing serial port forwarding Oct 17 21:16:49 then you could just use the neo's modem on your host Oct 17 21:16:56 running gsm0710muxd and the whole frameworkd on your host Oct 17 21:17:09 hm.. Oct 17 21:17:16 it's nearly working Oct 17 21:17:23 there's just some signalling problems w/ the serial forwarding Oct 17 21:17:26 so it only works once Oct 17 21:17:35 and then you have to restart the whole installation Oct 17 21:17:48 (reset modem, relaunch forwarder, listener, clients) Oct 17 21:18:03 I'd say either doing this or Oct 17 21:18:09 doing a full generic dbus forwarder Oct 17 21:18:15 which may be tempting as well Oct 17 21:18:22 should not be too hard in python Oct 17 21:18:29 but will cost you a couple of days Oct 17 21:19:19 in any case don't let me put you down on using tcp in dbus. it's pretty broken but perhaps it can be beaten into shape. there have been mutiple threads on that this year on the dbus mailing list Oct 17 21:20:37 isn't there a way to export the unix domain socket via tcp? and converting it on the other side? Oct 17 21:21:08 yes Oct 17 21:21:23 you will run into problems with dbus auth though Oct 17 21:21:33 and you will only be able to do _either_ udp _or_ tcp Oct 17 21:21:34 quickdev: ok implemented the thing Oct 17 21:21:45 so it will mess up everything on the target Oct 17 21:21:55 what about NFS? Oct 17 21:22:20 domain sockets are no real files Oct 17 21:22:21 Ainulindale, fine, thanks :) Oct 17 21:23:05 the generic dbus forwarder might be most practical Oct 17 21:23:16 acting as proxy on your host Oct 17 21:23:22 appearing as client to the target Oct 17 21:26:21 mickeyl, are you certain that nfs won't work? Oct 17 21:26:40 well Oct 17 21:26:41 i never tried Oct 17 21:26:49 but my intuition says a unix domain socket is a special file Oct 17 21:26:52 that can't be exported via NFS Oct 17 21:27:01 as much as device files Oct 17 21:27:02 I think he meant a nfs mount with a binary on it Oct 17 21:27:04 etc. Oct 17 21:27:13 Which will work if you manage permissions well enough Oct 17 21:27:26 what binary? Oct 17 21:27:27 but NFS is a pain Oct 17 21:27:35 well, I'm using NFS all the time here Oct 17 21:27:47 mounted my fso git tree over NFS to the Neo Oct 17 21:27:47 goal being to launch your binary on the phone Oct 17 21:28:00 mickeyl: the point was to have a better way to develop Oct 17 21:28:04 ooh Oct 17 21:28:10 i though you wanted to export the dbus domain socket Oct 17 21:28:15 to get on the bus Oct 17 21:28:17 from your host Oct 17 21:28:33 sure, running binaries via NFS works fine Oct 17 21:28:36 doing it all the time here Oct 17 21:28:50 root@om-gta02:/local/pkg/fso/framework/tools# mount Oct 17 21:28:55 192.168.0.200:/local/pkg on /local/pkg type nfs (rw,vers=3,rsize=32768,wsize=32768,soft,nointr,nolock,proto=udp,timeo=7,retrans=3,sec=sys,addr=192.168.0.200) Oct 17 21:28:57 :) Oct 17 21:29:16 Ainulindale, that should be faster? hehe.. Oct 17 21:29:33 Indeed Oct 17 21:29:36 so you can even do it w/ OE Oct 17 21:29:37 But I'm too lazy to setup NFS Oct 17 21:29:42 bitbake -c devshell yourProgram Oct 17 21:29:44 then edit Oct 17 21:29:48 then make Oct 17 21:29:50 then run on the device Oct 17 21:29:52 edit again Oct 17 21:29:53 etc. Oct 17 21:29:59 devshell? Oct 17 21:30:03 if your finished, make a patch, and commit it Oct 17 21:30:22 devshell opens a terminal window with the environment set up as "inside" an OE build Oct 17 21:30:36 a lean-and-mean way to crosscompile and edit stuff "in place" Oct 17 21:30:48 a, very fine! we should've asked you before Oct 17 21:31:00 a really good mean to forget to copy your sources before any -c clean Oct 17 21:31:08 hehe, ya there's some risk Oct 17 21:31:10 (this is true experience) Oct 17 21:31:23 it's actually just abusing OE to be a poor man's development environment ;) Oct 17 21:31:35 works fine as long as you're careful though Oct 17 21:32:14 alternatively you can just change the SRC_URI Oct 17 21:32:19 to point to a directory in your filesystem Oct 17 21:32:23 and then OE will grab the sources from there Oct 17 21:32:31 mickeyl, SRC_URI is already local in SHR Oct 17 21:32:35 ah, good Oct 17 21:32:41 that's a pain too Oct 17 21:32:46 because it won't regenerate each time Oct 17 21:32:51 We sought a way to do that Oct 17 21:32:54 But couldn't figure out Oct 17 21:33:15 yeah, SRCREV does not exists on the local directory Oct 17 21:34:27 well Oct 17 21:34:30 you can remove the stamps Oct 17 21:34:41 or remove the tarball in the source archive Oct 17 21:34:55 there must be a way to prevent caching iirc Oct 17 21:35:05 so it would grab the sources from the URI everytime Oct 17 21:35:20 mickeyl: is outgoing doing the same thing incoming does ? Oct 17 21:35:20 yes, but the URI is local...and thus there's no SRCREV? Oct 17 21:35:32 Ainulindale: no, outgoing only triggers once Oct 17 21:35:35 Ok :-) Oct 17 21:36:27 quickdev: yes, it would always have the same SRCREV, but who cares if you are just running the binaries -- you don't want to opkg install the package everytime, right? Oct 17 21:37:05 again, if someone fixes the serial port forwarding all that is moot :) Oct 17 21:37:11 and we can easily develop the whole stack on the PC Oct 17 21:37:26 Oct 17 21:37:31 mickeyl: how can I do if I have a mac ? Oct 17 21:37:32 Oct 17 21:37:53 hmm, point taken Oct 17 21:38:09 Point yourself ! Oct 17 21:38:27 mickeyl: your remark about incoming closes two tickets Oct 17 21:38:32 (no suspend after a while) Oct 17 21:38:42 (because of the active_calls tracking on incoming calls) Oct 17 21:38:56 serial port forwarding? a forwarder for the dbus things via tcp? Oct 17 21:39:17 no, serial port forwarding was the other thing i talked about Oct 17 21:39:25 forwarding the gsm modem to a pty Oct 17 21:39:35 then running gsm0710muxd and the whole frameworkd on your PC Oct 17 21:39:49 ah, ok Oct 17 21:39:56 although that gives you complete different timing... :) Oct 17 21:40:15 Oh, nioce a bug Oct 17 21:40:22 [1] + Segmentation fault /usr/bin/ophonekitd Oct 17 21:40:27 Nice modification indeed Oct 17 21:40:51 Problems is, oeventsd won't stop the ringtone Oct 17 21:41:44 yep, i have a big todo item there Oct 17 21:41:49 if gsm resource gets released Oct 17 21:41:51 mickeyl, we are talking about having /dev/ttySAC1 for example on the x86 system....which sends/reads data from neo's /dev/ttySAC1? Oct 17 21:41:52 shut down gracefully Oct 17 21:41:59 quickdev: yes Oct 17 21:42:09 like i said, it works somehow Oct 17 21:42:10 mickeyl, that should be done very quickly? Oct 17 21:42:12 mickeyl: good thing though Oct 17 21:42:15 I changed this ringtone Oct 17 21:42:20 So at least I'm listening to some zappa Oct 17 21:42:28 quickdev: it works, but has some problems with serial signalling Oct 17 21:42:33 quickdev: check out andrews post on it Oct 17 21:43:48 mickeyl, I have no clue how to find it? Which mailing list? Any keywords to search for? Oct 17 21:43:51 http://unadventure.wordpress.com/2007/10/28/qpe-430-plus-qemu/ Oct 17 21:43:54 you need pty Oct 17 21:43:55 and nc Oct 17 21:44:07 and then patch gsm0710muxd not to try powering up/down the modem Oct 17 21:44:16 (which won't work on your host) Oct 17 21:44:30 the setup is a bit fragile, like i said Oct 17 21:44:33 but i got it working once Oct 17 21:44:41 very convenient Oct 17 21:45:05 Ok bug fixed Oct 17 21:45:43 quickdev: commited Oct 17 21:48:18 There are too many quickdev assigned tickets out there :-) Oct 17 21:48:21 Get to work ! :-) Oct 17 21:48:29 mickeyl: my other question was Oct 17 21:48:34 What about this echo issue fix ? Oct 17 21:48:39 hehe..let's build a fast development environment first Oct 17 21:50:24 Ainulindale: http://git.freesmartphone.org/?p=framework.git;a=blob;f=framework/subsystems/ogsmd/modems/ti_calypso/channel.py;h=15e44cfa61f80d9ee0d405a8d9917254cb157d32;hb=1c74c39b7662032f07e43bdfddc3d86c245cc44c Oct 17 21:50:39 still toying around with the right compromise Oct 17 21:50:43 6db or 12db Oct 17 21:50:49 AEC or only Noise Oct 17 21:50:51 or both Oct 17 21:50:52 etc. Oct 17 21:51:41 mickeyl, what I need to do: Run nc on my neo and pty on my x86 system and then patching gsm0710muxd? That's all? Oct 17 21:52:54 in theory yes Oct 17 21:53:07 in practice you need to manage reset and on/off on your neo manually now Oct 17 21:53:10 [modem] Oct 17 21:53:22 then you need to change dbus configuration so that gsm0710muxd can register its object path Oct 17 21:53:23 etc. Oct 17 21:53:38 and the timing and order is a bit fragile Oct 17 21:54:05 but this may be fixable once we found out what's making it hang the 2nd time Oct 17 21:54:58 mickeyl, why should the order differ? Oct 17 21:55:08 the modem is sensitive Oct 17 21:55:08 mickeyl: So I guess I can fix the ticket as fixed upstream :-) Oct 17 21:55:14 Ainulindale: :) Oct 17 21:55:23 quickdev: but just have a go Oct 17 21:55:28 and we talk about concrete problems :) Oct 17 21:55:48 ok :) Oct 17 21:55:48 i just tell you that it was not completely straightforward Oct 17 21:58:22 quickdev: please give a try to the latest ophonekitd :-) Oct 17 22:02:12 quickdev: incoming/outgoing should be grouped together for your edc files and some C files too I thinkl Oct 17 22:02:16 -l Oct 17 22:02:23 (And that would be easier to handle speaker out) Oct 17 22:02:53 bbtomorrow Oct 17 22:02:56 g'night Oct 17 22:03:01 Ainulindale, grouping what together? The screen when there's an active call or the whole thing? Oct 17 22:03:05 <[Rui]_> wanna save 2s from boot time? delete a few lines in /etc/init.d/xserver-nodm Oct 17 22:03:32 the whole thing quickdev Oct 17 22:03:39 doing a "call" screen Oct 17 22:03:45 and just setting the text to outgoing or incoming Oct 17 22:04:04 ok, will do it Oct 17 22:04:49 You should also define constants for the path of the main directory for the edc files Oct 17 22:04:56 s/constants/constant/ Oct 17 22:05:38 ie #define UI_PATH "/your/path/" Oct 17 22:05:52 Then #define UI_FILE to your path plus the file Oct 17 22:06:43 ok Oct 17 22:10:19 Just giving some basic stuff here quickdev Oct 17 22:10:26 if you don't like it that way don't do it Oct 17 22:10:35 of course you're right ;) Oct 17 22:10:35 That's what I would do to make my life easier :-) Oct 17 22:14:52 ok, i just booted into SHR, it asked for PIN and registered to network just fine =] Oct 17 22:15:18 but there's no wrench in the dropdown menu ... so how do i access settings ? Oct 17 22:16:18 infact, top-menu is empty apart from qwerty and network .... there's nothing else there Oct 17 22:26:42 have to leave Oct 17 22:26:47 see you tomorrow Oct 17 23:40:29 :/ i get errors during opkg update Oct 17 23:41:15 heh, and now it works ... wird Oct 17 23:41:24 s/wird/weird **** ENDING LOGGING AT Sat Oct 18 02:59:56 2008