**** BEGIN LOGGING AT Mon Sep 15 02:59:57 2008 Sep 15 05:56:40 hi all... i'm trying to build a custom kernel with the patches. . i've changed the linux-openmoko_2.6.24+git.bb to include the patches in SRC_URI. the problem is that make build-package-linux builds only 2.6.21 kernel and just wouldn't go to 2.6.24. I've changed the PREFERRED_PROVIDER_virtual/kernel to linux-openmoko and PREFERRED_VERSION_linux-openmoko to 2.6.24+git ... no luck Sep 15 05:56:48 can anyone please guide me on this? Sep 15 06:42:01 hi Sep 15 06:43:32 hello Sep 15 06:47:27 i'm using the qmeu-neo1973 simulator and I would like to read the disk image, is there any way to mount it? i'm looking for something in the web, but nothing works Sep 15 06:48:46 aleix: do you mean the rootfs of openmoko in qemu? Sep 15 06:49:05 yes, the jffs2 converted to qcow2 Sep 15 06:49:19 hello Sep 15 06:49:20 i know how to mount the jffs2 Sep 15 06:49:30 but not the qemu image Sep 15 06:49:55 i'm trying to start the matchbox-wm, but it fails, i'd like to read the log files Sep 15 06:50:18 http://wiki.openmoko.org/wiki/Booting_from_SD#Possibility_3_:_Convert_a_jff2_image_to_a_tarfile Sep 15 06:50:47 I haven't tried it myself but since it's "official", it should work :) Sep 15 06:51:01 hil__: I don't have a problem with the jffs2 Sep 15 06:51:27 hil__: I can read it. but the simulator uses its qcow2 format Sep 15 06:52:03 hil__: that is, jff2 is converted to qcow2, and the jffs2 image is not used again Sep 15 06:52:33 hil__: so, when something fails, it is written to the qcow2 Sep 15 06:52:45 oh. Sep 15 06:53:06 thanks, anyway Sep 15 06:53:18 i have no idea about qcow2 ... I work directly with Freerunner since it doesn't have qemu support yet Sep 15 06:53:27 yep Sep 15 06:54:00 i too have a question. see if you know something about it. Sep 15 06:54:14 i'm trying to build a custom kernel with the patches. . i've changed the linux-openmoko_2.6.24+git.bb to include the patches in SRC_URI. the problem is that make build-package-linux builds only 2.6.21 kernel and just wouldn't go to 2.6.24. I've changed the PREFERRED_PROVIDER_virtual/kernel to linux-openmoko and PREFERRED_VERSION_linux-openmoko to 2.6.24+git ... no luck Sep 15 06:54:24 i've been trying to get an answer for days now :( Sep 15 06:54:45 hey, does anybody know if the xglamo-xserver is required to run X or is it just for acceleration? Sep 15 07:03:48 u can use xfbdev Sep 15 07:03:52 but it wil be slower Sep 15 07:04:03 by quite a lot Sep 15 07:14:27 hil__: where have you changed the PREFERRED_VERSION? Sep 15 07:17:25 aleix: yes. I have. Sep 15 07:17:36 PREFERRED_VERSION_linux-openmoko = "2.6.24+git" Sep 15 07:18:26 (i also tried VERSION_linx = "..." cause I wasn't sure about linux/linux-openmoko... it still builds 2.6.21...) Sep 15 07:18:57 hil__: yes, but where have you chnaged that? Sep 15 07:19:10 openembedded/conf/machine/om-gta02.conf Sep 15 07:19:30 That's my machine. (bitbake says so when it starts.. so that's confirmed, right?) Sep 15 07:20:16 are you building fso? Sep 15 07:20:38 no. om2008.8 ...the default (usig mokomakefile) Sep 15 07:26:09 hil__: haven't used that, sorry Sep 15 07:26:44 alright :) Sep 15 07:26:58 i'll wait for someone else to wake up :) Sep 15 07:28:35 yes Sep 15 07:36:27 can I just do an "apt-get upgrade" on Debian FSO, to upgrade packages for the Neo? Sep 15 07:41:53 raster: normal xserver with vesa-driver or sth. like that would not be possible? Sep 15 07:43:05 cant do vesa Sep 15 07:43:17 its nota "vesa card" with a bios.... Sep 15 07:43:18 :) Sep 15 07:43:22 or vga Sep 15 07:43:44 u either have an accelerated driver that uses the glamo's features to do work for you Sep 15 07:43:51 or u use a dumb framebuffer xserver Sep 15 07:43:55 ie xfbdev Sep 15 07:43:59 (whihc is kdrive) Sep 15 07:44:20 aa fully modular xserver isnt probably worth it on the device as gfx chip will never change Sep 15 07:49:55 hmm - my intention is link xglamo against the uclibc Sep 15 07:50:08 it compiles fine, but while starting it crashes the whole device Sep 15 07:50:28 even jtag is not possible at this state anymore for debugging Sep 15 07:50:47 ideas for debugging are out - now i'm searching for alternatives Sep 15 07:51:33 brb Sep 15 07:52:10 raster: any ideas? did you eve rlook at the xglamo-source? Sep 15 07:52:42 its possible that uclibc doesnt implement some libc funciton the same way or as com pletely as glibc Sep 15 07:52:47 what one it is - i can't tell u Sep 15 07:52:56 but uclibc is not a 100% drop-in replacement Sep 15 07:53:08 it can have issues - espeically if apps make assumptions on the libc behavior Sep 15 07:53:13 i know - that's why i'm searching for the cause Sep 15 07:53:13 and i've looked at it Sep 15 07:53:19 but it could be anywhere... Sep 15 07:53:35 but the normal kdrive is working fine with uclibc Sep 15 07:53:42 process of elimination Sep 15 07:53:43 and x glamo is more or les sjust a fork of kdrive Sep 15 07:53:54 basically get it to abort setup at points and see what point u get to Sep 15 07:53:59 it'll take a while Sep 15 07:54:09 does xfbdev work? Sep 15 07:54:16 i tried lot's of printf-debugging Sep 15 07:54:18 several days Sep 15 07:54:22 no success :/ Sep 15 07:54:29 didn't try xfbdev Sep 15 07:54:50 but xfbdev can only run one toolkit at the same time i heared Sep 15 07:54:52 try Sep 15 07:54:53 :) Sep 15 07:55:00 totally incorrect Sep 15 07:55:09 xfbdev is simply an xserver that just uses /dev/fb Sep 15 07:55:16 instead of trying to accelerate Sep 15 07:55:18 oh, ok Sep 15 07:55:24 ie its slower - but should work fine Sep 15 07:55:34 so i just use kdrive with fbdev as driver? Sep 15 07:55:36 it tries nothnig special beyond what u see on the linux fb-base text console Sep 15 07:55:40 ok, will try it out - thanks :) Sep 15 07:55:42 yup Sep 15 07:55:52 then u can eithe rknow its a general x prblem Sep 15 07:56:00 (ie with the code kdrive builds from ghenerically) Sep 15 07:56:05 or something specific to the accel in xglamo Sep 15 07:56:13 yeah Sep 15 07:56:28 ok, will give it a try :) Sep 15 07:56:42 or any other changes made to xglamo vs core kdrive Sep 15 08:02:59 do yu know where i can get a working x-config for use with fb? Sep 15 08:06:40 u dont need any config for kdrive Sep 15 08:06:50 its a single binary u just run and works Sep 15 08:06:59 any config is on cmd-line if u really need it Sep 15 08:14:53 raster: and how do i define which driver will be used? Sep 15 08:25:15 Hey there Sep 15 08:25:37 AndreasD, please remember libframeworkd-glib repository is now on FSO git Sep 15 08:25:53 So please don't modify files on my SVN Sep 15 08:26:53 Ainulindale: I know, but couldn't get in contact with you Sep 15 08:27:40 Ainulindale: what is the progress for getting code into the git repo? Sep 15 08:27:55 Ainulindale: FSO trac seems to be closed Sep 15 08:31:36 Huh ? Sep 15 08:32:10 ah, sorry s/progress/process Sep 15 08:32:10 The code is in the git repo already Sep 15 08:32:16 Ah, that Sep 15 08:32:21 yes but patches Sep 15 08:32:22 You propose a git patch Sep 15 08:32:30 And someone (probably me) accept it Sep 15 08:33:10 but how? trac seems to be closed :s Sep 15 08:33:18 You mail the ml :-) Sep 15 08:33:32 either SHR or smartphone-handhelds Sep 15 08:33:39 smartphone would be better Sep 15 08:34:07 okay fine :-) Sep 15 08:49:55 hey MarcOChapeau Sep 15 08:53:44 Ainulindale: yo Sep 15 09:00:27 how went the thing with bearstech MarcOChapeau ? Sep 15 09:00:47 couldn't go. I left work at 10pm Sep 15 09:01:08 I'm currently having a wonrderful time at work Sep 15 09:01:08 Arf :-) Sep 15 09:01:15 I can see that Sep 15 09:01:21 I seriously consider quiting Sep 15 09:02:42 Hmmm if you're interested my company is looking for people proficient with network security Sep 15 09:03:09 (For MINEFI) Sep 15 09:03:17 +the Sep 15 09:06:52 MarcOChapeau: you have a git repository where you work on the gtk dialer? Sep 15 09:08:59 b0ef: svn repository Sep 15 09:09:01 The SHR svn Sep 15 09:09:59 Ainulindale: thanks Sep 15 09:10:13 no plans to move to git?;) Sep 15 09:14:01 b0ef: yep we'll move to git as soon as someone volunteer to do the migration Sep 15 09:14:07 it's easy, and I could do it, but I'm lazy Sep 15 09:58:22 Ainulindale: where would you host shr git repo? Sep 15 10:40:29 hi...I have an openmoko-devel-image. When I try build-package-linux, the makefile builds linux-2.6.21 and not 2.6.24+git but if I perform a make clean, then build-package-linux, it builds 2.6.24+git. Can anyone help me with this problem? Sep 15 12:13:51 mickey|dinner, here? Sep 15 12:16:46 quickdev: so as I was saying Sep 15 12:16:47 Ainulindale, let's continue our discussion ;) Sep 15 12:16:56 Implementing another daemon is your choice Sep 15 12:17:06 using ophonekitd makes sense to me Sep 15 12:17:07 I don't think it would have any added value to do that though Sep 15 12:17:24 B ut as I'm sure you worked far more on your daemon that I worked on mine Sep 15 12:17:28 As I had no UI to test it Sep 15 12:17:35 Feel free to commit your changes on ophonekitd Sep 15 12:17:43 which should, in the end, migrate to FSO git I think Sep 15 12:17:47 commit changes = send patches to you? Sep 15 12:17:50 Nope Sep 15 12:17:59 ophonekitd is on SHR SVN Sep 15 12:18:06 Which will become SHR GIT soon enough Sep 15 12:18:16 (as soon as someone will migrate it) Sep 15 12:18:16 and shr svn is accessible by everyone? Sep 15 12:18:20 nope Sep 15 12:18:23 just send me your public key Sep 15 12:18:27 and you'll have an account Sep 15 12:18:37 right now it's hosted on a server of mine Sep 15 12:18:47 and it will migrate on the bearstech shr buildhost Sep 15 12:19:00 as soon as someone has enough time to do the migration process and create the accounts Sep 15 12:20:18 Ainulindale, do I have to create a ordinary rsa key? I'm not used to access svn with a public key Sep 15 12:20:25 yep Sep 15 12:20:41 if you use fingerprint for you day to day ssh usage it will fit Sep 15 12:21:20 Ainulindale: why migrate from svn to git? Sep 15 12:22:07 Because git allows far more flexible things Sep 15 12:42:02 quickdev: And what about the toolkit ? Sep 15 12:42:26 Ainulindale, which toolkit? Sep 15 12:42:47 graphical toolkit Sep 15 12:43:35 I'm using the efl Sep 15 12:43:44 Oh ok :-) Sep 15 12:43:55 If I can just tell you how the library thing is supposed to work Sep 15 12:44:13 http://www.calaquendi.org/om/ophonekitd.png Sep 15 12:44:17 it's not that exact but hey... Sep 15 12:44:29 lfg (libframeworkd-glib) is a bunch of primitives Sep 15 12:44:37 it's supposed to free the memory and handle the connection to dbus Sep 15 12:45:06 lfp (libframeworkd-phonegui) is a set of abstract primitives such as "display_pin", "display_talking" Sep 15 12:45:06 I know, I'm using it ;) Sep 15 12:45:20 So if you want to implement a library which will work immediately with ophonekitd Sep 15 12:45:23 you just have to fill the gaps Sep 15 12:45:28 and link ophonekitd against it Sep 15 12:45:29 but why is ophonekitd on top of lf-phonegui? shouldn't that be switched? Sep 15 12:45:49 this drawing is a bit difficult to explain just like that Sep 15 12:45:53 I'll build a good one today Sep 15 12:45:59 it has changed a lot Sep 15 12:47:00 ok ;) Sep 15 12:47:11 btw. are you talking german? Sep 15 12:47:25 Not at all Sep 15 12:47:29 But I know how to say I like potatoes Sep 15 12:47:39 (I'm french) Sep 15 12:48:03 quickdev: ophonekitd is there just to be the link between "model" (lfg) and "view" (lfp) Sep 15 12:48:23 Because I don't want to see the UI talking that much with frameworkd Sep 15 12:48:24 yeah, I understood the function ;) Sep 15 12:49:00 I'll see how well it integrates Sep 15 12:51:28 I think the idea is pretty simple Sep 15 12:51:31 Ainulindale, could you remove libframeworkd-glib from the svn repository when AndreasD_ saved his changes and committed them to the git repo? Sep 15 12:54:30 quickdev: i'll do that, I planned to do that a while ago but I'm too lazy Sep 15 12:54:47 won't be much work, hehe ;) Sep 15 12:59:48 Ainulindale, could you check a little thing according libframeworkd-glib for me? Sep 15 13:00:17 Of course Sep 15 13:00:48 frameworkd-glib-sim.c, sim_get_auth_status_callback(), free(status), is the free correct in that function? It produces a segfault, because of an invalid pointer Sep 15 13:01:56 that does only happen when my daemon (ophonekitd functionality) is started right after dbus is available. Sep 15 13:03:23 (wait a minute, boss alert) Sep 15 13:11:35 quickdev: it's right Sep 15 13:11:46 it has to be there Sep 15 13:11:57 Ainulindale, have you any explanation why it could end up with a segfault there? Sep 15 13:12:15 It must be null Sep 15 13:25:45 @quickdev: actually it could be any invalid pointer not just null. Actually if free gets a null pointer it should not segfault. Sep 15 13:26:12 man 2 free: free() frees the memory space pointed to by ptr, which must have been returned by a previous call to malloc(), calloc() or realloc(). Otherwise, or if free(ptr) has already been called before, undefined behavior occurs. If ptr is Sep 15 13:26:12 NULL, no operation is performed. Sep 15 13:26:29 yeah, it indicated that it was a arbitrary address Sep 15 13:27:23 If the pointer is valid, then make sure you didn't accidently write just before where the pointer points. Some malloc implementations put information about the allocated memory there. Sep 15 13:27:43 Oh Sep 15 13:27:45 I know the problem Sep 15 13:27:49 free(status) shouldn't be there Sep 15 13:28:01 it should be in the else Sep 15 13:28:13 this value has a meaning only if there's no error Sep 15 13:28:36 bug caught :) Sep 15 13:28:59 I think this behaviour is to be seen a lot of times Sep 15 13:28:59 I've got another function with a segfault, one moment Sep 15 13:29:17 (I'm drawing this diagram) Sep 15 13:32:06 Ainulindale, device_get_info_callback(), frameworkd-glib-device.c, g_hash_table_destory() fails if there's an error, too Sep 15 13:36:01 quickdev: http://www.calaquendi.org/om/lf.png Sep 15 13:36:33 more precise Sep 15 13:36:40 wurp2|working: hilight Sep 15 13:39:35 Is it clearer ? Sep 15 13:42:44 Is anybody able to get gestures recognised in FDOM ? Sep 15 13:42:52 Ainulindale, for me the upper part is a bit confusing Sep 15 13:43:24 Ainulindale, lfp-gtk - libframework-phonegui-gtk? Sep 15 13:43:51 Ainulindale, openmoko-dialer3 doesn't use it's own ui? Sep 15 13:44:01 so Sep 15 13:44:08 lfp-gtk is libframeworkd-phonegui-gtk yes Sep 15 13:44:19 you can implement any toolkit by implementing lfp primitives Sep 15 13:44:22 (it's a bunch of headers) Sep 15 13:44:32 openmokod-dialer3 is supposed to be a placeholder Sep 15 13:44:52 a binary linked with lfp and just calling the right method to display the dialer Sep 15 13:45:07 That way all the UI is in the same place Sep 15 13:45:38 @Ainulindale: What program did you draw that diagram with? Sep 15 13:45:56 I'd rather not say that Sep 15 13:46:02 :-) Sep 15 13:46:07 I don't have OO here Sep 15 13:46:10 windows paint :) Sep 15 13:46:23 Paint can't do line antialiasing. Sep 15 13:46:25 And I'm supposed to be able to use visio Sep 15 13:46:31 Liar ;-) Sep 15 13:46:31 wdraxinger, you can do it by hand Sep 15 13:46:54 LOL, subsampling the single pixels by hand with the pencil tool, or what? Sep 15 13:47:03 wdraxinger: visio then Sep 15 13:47:15 which is awful Sep 15 13:47:25 but I'm supposed to be able to use it (because other people here use it) Sep 15 13:47:25 Ainulindale, openmoko-dialer3 has the "controller" code without any ui things? The ui is borrowed from libframeworkd-phonegui-gtk, right? Sep 15 13:47:38 wdraxinger, exactly ;) Sep 15 13:47:44 quickdev: dialer3 is main() { display_dialer(); } Sep 15 13:49:38 Don't be ashamed. Did you ever use LabVIEW. Now that's some punishment. I mean: At first it looks cool and one begins to really like it. But then all those nasty inconveniences pop up. Like that LabVIEW can do only 1D FFT. Anything in higher dimensions must be build by stacking loops each doing FFTs on the rows in a dimension. And then one has to transpose in all dimensional permutations and do that again. Sep 15 13:49:41 I'll put inserts to explain that Sep 15 13:49:50 Why not simply implement a muldidimensional FFT. Sep 15 13:50:03 FFT are evil things by definition Sep 15 13:50:16 Though I was happy once I used an FFT to reduce some noise on a sample I did Sep 15 13:50:44 First time, and only time in fact, knowing what an FFT was helped me at all Sep 15 13:50:47 FFTs are cool, if you know, how to use them. And they can be handy for all sorts of things. As long you don't have to implement them yourself. I leave that to the experts. Sep 15 13:51:04 Please don't make me cry :-) Sep 15 13:52:25 * Weiss has an ipkgs of FFTW3 and GSL is they're any help to anyone :) Sep 15 13:52:50 i ported a piece of my research software to OM for the comedy value Sep 15 13:53:25 An IPGK of that http://cr.yp.to/djbfft.html would be handy. Sep 15 13:56:00 Ainulindale, is it expected that anyone codes a libframeworkd-phonegui-efl and openmoko-dialer3 could use that? Sep 15 13:56:14 I like the DJB way (though I admit, that DJBs social skills are underdeveloped ;-)), but the way how his software works, shows how good software design is meant. Thats why my current OpenMoko related project is getting minit http://www.fefe.de/minit/ up and running to speed up that boot. mini has been developed the DJB way, and it's way cool. Sep 15 13:59:01 wdraxinger, tried some things in that direction a few weeks ago Sep 15 13:59:30 how was your milage? Sep 15 13:59:35 would be very appreciated if you find a proper setup :) Sep 15 14:00:36 intersecting with OpenMoko is my start of a cross compiling aware package manager. Highly inspired by Paludis, but meant to be able to: Sep 15 14:01:02 * Install all required toolchain into the developer's system. Sep 15 14:01:19 * Install cross compiled packages into a cross - chroot Sep 15 14:01:22 wdraxinger, I tested some things with runit. Most of the startup time was spend with the kernel initialization, udev and frameworkd Sep 15 14:01:40 quickdev: yes it is Sep 15 14:01:56 Yes, the kernel startup still sucks. But 5 seconds are just wasted by the rootdelay=5 setting. Sep 15 14:02:02 it is my main purpose with this architecture in fact Sep 15 14:02:14 allowing the developer to easily plug libraries to manage telephony events Sep 15 14:02:17 or even other events Sep 15 14:02:27 http://www.calaquendi.org/om/lf.png <= updated version Sep 15 14:02:34 have to go now, see you this evening Sep 15 14:02:54 wdraxinger, keep getting that working :) Sep 15 14:02:58 check the updated version just to see if it's easier to understand :-) Sep 15 14:03:16 I think a lot could be done to speed things up, by having some initramfs (not to confuse with initrd, which requires filesystems compiled into the kernel) Sep 15 14:03:53 in the initramfs even some part of the userland init could take place, before mounting the actual root device and re-rooting. Sep 15 14:04:09 Ainulindale, I think it's better - will have a more precise look this evening Sep 15 14:04:24 ok :-) Sep 15 14:04:34 wdraxinger, try it :) Sep 15 14:04:46 see ya Sep 15 14:05:00 That would allow to use the otherwise wasted rootdelay doing something usefull. Also I think that as much drivers as possible should be modularized, but still some parts of the GTA02 stuff refuses to compile as module. Sep 15 14:15:42 http://www.cyriak.co.uk/lhc/lhc-webcams.html <= marf Sep 15 14:20:07 Another thing: A good friend of me just got his degree in "visual communication". Thats an assortment of topics covering photography (mostly design/still lifes), product design (the appearance of a product communicates it's purpose), and so on. Naturally in that field he's dealing with a lot of Apple computer stuff. But interestingly he likes the FreeRunner more than the iPhone. And he even considered, that he mig Sep 15 14:20:07 ht design a new case for it. Sep 15 14:20:31 mickey|dinner: we should talk about lfg announcement Sep 15 14:20:59 And I have a bunch of questions related to MS3 Sep 15 14:21:05 First one being I can't even access SMS :-) Sep 15 14:21:14 i'm trying to load matchbox-window-manager without luck, does anyone knows if it works? Sep 15 14:21:38 (and the silent mode doesn't work) Sep 15 14:22:17 IMHO matchbox somehow sucks. Whenever I tried it, either embedded or on my desktop in a Xnest/Xephyr it has some weird bugs and crashes randomly... Sep 15 14:22:37 wdraxinger: great :-) Sep 15 14:23:49 Ainulindale: thanks for the drawing. Sep 15 14:24:19 Ainulindale: do you know what was the idea for shr? was it to use matchbox, enlightenment....? Sep 15 14:24:28 Best bug so far: starting matchbox-desktop standalone: works. starting matchbox-desktop with mwm, kwin, metacity, whatever as a WM: works. starting matchbox-wm standalone works. starting matchbox-desktop when matchbox-wm is running: matchbox-wm crashes. starting matchbox-wm when matchbox-desktop runs: both crash. Sep 15 14:24:45 aleix: first we wished to use matchbox Sep 15 14:24:48 but I think we will use illume Sep 15 14:24:52 actually they use enlightenment. Sep 15 14:25:02 we should work on matchbox though for people who prefer it Sep 15 14:25:33 Ainulindale: i was trying to setup the startup X script to load matchbox for SHR, but no luck Sep 15 14:25:52 Ainulindale: illume worked fine in the simulator Sep 15 14:26:02 I didn't work with matchbox Sep 15 14:26:04 I got a long long list of projects here to do for the OpenMoko and embedded in general. One topic on the list is creating an embedded WM that doesn't suck. Enlightenment ist fine, but even as resource friendly it is, it's IMHO still too bloated. Sep 15 14:26:28 (yay, troll time :-) ) Sep 15 14:27:02 Don't get me wrong: I love enlightenment and when I first thought about how to build a embedded desktop that WM came to my mind. Sep 15 14:28:05 But a mobile phone is something special. People expect the UI to be ultra responsive. On a desktop somehow everybody accepts little delays imposed by loading resources or so. But on a phone things must be "snappy". Sep 15 14:28:46 wdraxinger: I agree, but here the problem is related to the glamo if I recall correctly Sep 15 14:29:50 One drawback of enlightenment I see here, as I do in matchbox and similair is: They are totally configurable in all aspects. Parsing and processing configurable stuff takes time. IMHO there should be a "use the source luke" policy employed, i.e. configure the stuff at build time. Sep 15 14:31:00 And for an open source phone that's not really a problem: You can modify stuff in the source and recompile. To make things exchangable the whole configuration could be implemented in terms of shared libraries. Sep 15 14:31:55 ls -rtlh Sep 15 14:32:01 sorry Sep 15 14:32:01 (that's the direction we choose for lfg) Sep 15 14:32:27 Apropos glamo. Is there any chance to get some documentation on that thing. I somehow don't get it, why one would use a heavy NDA component in a open source phone. Sep 15 14:35:51 I actually don't even get it, why _hardware_ manufactureres are so closed about documenting their shit. I mean, who'd buy a cat in the sack. And to me a component without documentation I can read IS a cat in a sack. Sep 15 14:36:24 wdraxinger: that was discussed to death on OM-Community. OM can't give the specs (that they've been given) to us, but they can write their own specs which we can then see, but there isn't enough manpower to do that for now Sep 15 14:36:30 at least, that's how i understand it Sep 15 14:36:47 agreed about HW manufacturers being shady about things :/ Sep 15 14:37:53 I knew that. So what do I have to do, to get the OM guys writing their specs (and I don't see any reason why the original spec is then still closed. Any spec is as good, as long it covers every detail). Ah lawyers and logic, that doesn't mix. Sep 15 14:46:20 Oh, there's logic with lawyers. It's all quite understandable, and makes complete sense, when one notes that everything a lawyer does is designed to ensure the need for more lawyers -- thus ensuring that their pocketbooks are constantly being filled. Sep 15 14:59:29 bye Sep 15 15:33:20 ls -rtlhtail -f ibid Sep 15 15:35:30 Ainulindale: hilight? Sep 15 15:47:22 wurp2|working: My drawing Sep 15 15:48:17 Oh, yep, I was gonna put a link to it from the SHR devel page Sep 15 15:48:24 Very nice :-) Sep 15 16:04:32 hi, the example toolchain has a sample with a gtk application. Is there any sample for nice black 2008.08 applications? Sep 15 16:14:14 jrei: You mean an efl example? Sep 15 16:14:50 i guess so, it that is the lib the "2008.08 style" applications are build Sep 15 16:15:41 i wasn't shure if that was some sort of qt or eft Sep 15 16:17:36 i want to build some apps and i want them to fit into the 2008.08 style Sep 15 16:21:34 wurp2|working: hey. i just got the simulator running and trying to setup matchbox for the shr image Sep 15 16:26:31 aleix: Cool! Sep 15 16:26:58 jrei: that's efl aka e aka e17 (iirc) Sep 15 16:27:47 k Sep 15 16:27:59 wurp3: thanks Sep 15 16:28:05 3-- Sep 15 16:44:32 the gui style guidlines are outdated with the new eft style applications, arn't they? Or is this still the style we should use? Sep 15 17:05:37 jrei: Sorry, I have no idea Sep 15 17:18:28 hi Sep 15 17:18:40 is there any chance to acces dbus as non root user ? Sep 15 17:22:21 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r22877bb1abab 10/fsod/ (15 files in 3 dirs): start work on fsod, frameworkd Sep 15 17:22:21 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * rcfaea0eb4119 10/fsod/ (6 files in 3 dirs): add initial subsystem manager and vapi's Sep 15 17:22:24 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r84df58688ea9 10/fsod/src/subsystems/Device/ (42 files in 10 dirs): replicate Device subsystem structure from odeviced Sep 15 17:22:24 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * re84e65cae7d2 10/fsod/ (5 files in 4 dirs): Fix build errors Sep 15 17:22:26 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r63103c55c366 10/fsod/ (4 files in 2 dirs): add helpers.vala Sep 15 17:22:29 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r53e8d27a0297 10/fsod/src/Makefile.am: add Makefile.am Sep 15 17:22:32 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r547ef045012c 10/fsod/ (12 files in 5 dirs): Initial Device subsystem with input plugin. Others to follow Sep 15 17:26:06 no ideas ? Sep 15 17:26:32 peter-b: You tried it? Sep 15 17:26:42 yupp Sep 15 17:27:04 should it work ? Sep 15 17:27:26 I know that I've read about security config for dbus, but I really don't know anything about it :-( Sep 15 17:27:31 You tried both as root & non-root Sep 15 17:28:02 peter-b, there is Sep 15 17:28:12 peter-b, in your policy file change allow_user Sep 15 17:28:17 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r3bdd1fa3bdcf 10/fsod/Makefile.am: Add DBus policy file Sep 15 17:28:33 freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r4481a36d99d8 10/fsod/etc/ (dbus-1/system.d/fsod.conf fsod.conf): Add DBus policy file Sep 15 17:28:56 peter-b, http://git.freesmartphone.org/?p=specs.git;a=blob_plain;f=html/org.freesmartphone.GSM.Call.html;hb=HEAD#CallStatus, it says "released" Sep 15 19:23:49 hmm, docs might be right, but implementation might be wrong Sep 15 19:24:02 could we figure that out somehow? Sep 15 19:24:19 use mdbus -s -l and run through a call Sep 15 19:24:40 mickeyl, frameworkd definetely implements "release", but is that right? Sep 15 19:25:39 I ask because I just wanted to send a patch to smartphone userland mailing list for libframeword-glib Sep 15 19:25:53 "right" is relative. it's what zhone expects atm. and the framework sends Sep 15 19:26:09 i'm going to patch the docs rather than the frameworkd and the clients Sep 15 19:26:37 ok, let's use "release" then Sep 15 19:27:41 yeah Sep 15 19:28:03 quickdev: will you also fix ophonekitd? Sep 15 19:28:37 AndreasD, sooner or later, yes, Ainu gave me commit rights some hours ago Sep 15 19:28:56 great :-) Sep 15 19:44:24 quickdev: I'm there Sep 15 19:44:27 But not for long Sep 15 19:44:28 Be quick Sep 15 19:44:56 the free(status) we were talking about Sep 15 19:45:09 Yes ? Sep 15 19:45:18 it belongs into the if(data->callback != NULL) and if(dbus_error==NULL) statement, right? Sep 15 19:45:27 In the else statement Sep 15 19:45:29 There's only one else Sep 15 19:45:32 You can't make a mistake Sep 15 19:45:43 generally, are arguments initialized if no callback is set? Sep 15 19:46:02 All the arguments are initialized if dbus_error is NULL Sep 15 19:46:08 if it isn't NULL, I have no idea Sep 15 19:46:18 It belongs to the dbus-glib way of doing things Sep 15 19:46:21 Which is to say a crappy way Sep 15 19:46:35 ok, so we can only call free if dbus_error is NULL Sep 15 19:46:45 That's what I said :-) Sep 15 19:46:53 I'll send a patch for that to the smartphone userland mailing list, ok? Sep 15 19:47:18 or to you...you have commit rights, ok :) Sep 15 19:47:25 mail it to me Sep 15 19:47:40 Because I'd have to subscribe to the aforementioned mailing list first Sep 15 19:47:44 And I'm lazy Sep 15 19:47:49 And I'll eat in a few minutes Sep 15 19:48:12 ok, have a nice evening, thanks Sep 15 19:49:20 hi Sep 15 19:49:30 hi aleix Sep 15 19:50:36 wurp2|working: ping **** ENDING LOGGING AT Mon Sep 15 19:57:33 2008 **** BEGIN LOGGING AT Mon Sep 15 20:22:16 2008 Sep 15 20:32:12 wurp2|working: hey. what was the problem with dialer3? it's showing up in the simulator Sep 15 22:13:25 Does anyone know of any execution profiling software that can be used to determine execution times of functions within a program running on a Freerunner? Sep 15 22:14:27 you could try oprofile Sep 15 22:14:39 but you'll need to compile a new kernel to add support for it Sep 15 22:16:37 thos, Thanks. I was hoping to avoid building a custom kernel, though. Sep 15 22:17:26 other option is to use printf() or similar Sep 15 22:18:59 I used this technique on the freerunner: http://blogs.gnome.org/thos/2008/06/01/profiling-gnome-appearance-properties-start-up-time/ Sep 15 22:19:55 thos, Ooo - that's interesting. Thanks for the pointer! Sep 15 22:21:04 Ken_Young, yeah you can create funky graphs :-) Sep 15 22:21:38 I used it to profile startup times on the freerunner, but you have to be careful because the strace overhead is a lot bigger on the freerunner than on a desktop Sep 15 22:21:54 I think I restricted strace using -e access Sep 15 22:22:05 or whatever this syscall is that it uses Sep 15 22:53:07 thos, hi. any news from iain? Sep 15 22:53:34 greetings Sep 15 22:53:53 greetings earthling Sep 15 22:54:19 I asked on the mailing list with no response -- anyone know what fso-frameworkd does that keeps touchscreen input away from apps until the stylus is lifted? Sep 15 22:55:01 example being something like xournal or gimp -- draw a line, but gimp doesn't see it until you lift. might be gtk level? Sep 15 22:58:48 awwaiid, When the stylus is lifted, does the application get all the motion notify events it should have gotten if the stylus was moved while touching the screen? Sep 15 23:03:42 yes Sep 15 23:04:42 so the whole path is drawn Sep 16 00:28:37 awwaiid: I don't know hot it fso does, but in priciple _any_ X client can grab any X event, so that other applications don't recieve it. So there could be an application that greedy grabs all touchscreen input, and if it has not been processed specially replays it by synthetic events. Sep 16 00:31:23 X11 does not protect applications against each other (on the X11 level), so you can do quite crazy things with it. It's also how window managers work. Those programs attach themself to the root window and set the event mask to anything related about windows created by other clients. And then, to add titlebars, border, etc., a window manager will create another window, and reparent the original clients window into t Sep 16 00:31:23 hat new one, which is used to draw titlebars and such. Sep 16 00:31:38 (this is called reparenting window management). Sep 16 01:08:00 Hey I'm in hour 18 of my moko build... My first task is to patch glamo.c to correct for the bad touch screen coordinates when rotating between all modes except "normal". Sep 16 01:08:15 Has anyone else already built an update for that package? Sep 16 01:09:05 From what I can tell xserver-kdrive-glamo-r24 is the latest compiled binary (arm4) on the repo. which does not include the patch. Sep 16 01:59:54 NOTE: Applying patch 'enable-xcalibrate.patch' (/home/scott/code/mokobuild/openembedded/packages/xorg-xserver/xserver-kdrive-1.3.0.0/enable-xcalibrate.patch) Sep 16 02:00:31 Awesome.. I'm 90% done with the mokobuild... and it looks like my wish came true. i didn't even have to do anything. Sep 16 02:00:38 Thanks to whomever did that. **** ENDING LOGGING AT Tue Sep 16 02:05:38 2008 **** BEGIN LOGGING AT Tue Sep 16 07:02:31 2008 Sep 16 07:02:41 maddy: if you apt-get install openssh-server and -redir tcp:5022::22 then you can "ssh localhost -p 5022" to access the arm system which is kind of nice Sep 16 07:04:34 lindi-, ok Sep 16 09:44:10 did someone had problems with package perl-native on om-asu-image ? Sep 16 09:44:34 do_compile fails with : Sep 16 09:44:54 NOTE : make -e MAKEFLAGS= Sep 16 09:44:56 makefile:952: *** missing separator. Stop. Sep 16 09:47:08 line 952 is the last line of the makefile Sep 16 09:47:29 I can't remember if it needs a return at end or not ??? Sep 16 09:49:18 can someone try make build-package-perl-native please ??? Sep 16 09:49:54 ptitjes: i'm building package-linux.. only a few tasks left Sep 16 09:50:10 i'll try it out and tell you in a while (unless someone else does it first :) Sep 16 09:50:34 hil_: thanks :) Sep 16 09:52:30 What I find strange is that the last lines of the makefile are : Sep 16 09:52:38 # AUTOMATICALLY GENERATED MAKE DEPENDENCIES--PUT NOTHING BELOW THIS LINE Sep 16 09:52:38 # If this runs make out of memory, delete /usr/include lines. Sep 16 09:52:49 < line 952 Sep 16 09:53:01 Maybe dependencies are not generated Sep 16 09:53:20 hnn... that seems to be a problem of the do_configure task, no? Sep 16 09:53:37 I though so... :) Sep 16 09:56:46 I have so frequently errors while building asu image (because of floating versions choice as says rwhitby - which I find is a non-sense) that I really have whished bitbake could postpone some packages and continue building the rest of the packages until I find the errors Sep 16 10:06:29 hil__: I'll quit as I must go Sep 16 10:07:43 maybe if you find what the problem is with building perl-native could you make a mail to the community list ? Sep 16 10:07:50 thanks anyway Sep 16 10:07:53 bye Sep 16 11:17:15 playya, hey, what did you want from iain? Sep 16 11:34:46 hmm - looks like there's a sane-srcrevs.inc entry missing for illume in OE ... Sep 16 11:34:57 mickey|tv: are you getting a failure on that? Sep 16 11:41:20 raster: ping Sep 16 11:41:42 raster: you forgot to add an entry to sane-srcrevs.inc when you added illume_svn.bb Sep 16 11:42:35 ummmm Sep 16 11:42:39 i know i did Sep 16 11:43:01 I'm referring to OE, not OM Sep 16 11:43:06 (fso build) Sep 16 11:43:53 I can copy it across if you like ..,. Sep 16 11:44:16 raster: so are you building from OM git repo or OE git repo? Sep 16 11:44:54 9:44PM ~/oe/org.openembedded.dev > grep illume conf/distro/include/sane-srcrevs.inc | grep EFL Sep 16 11:44:54 SRCREV_pn-illume ?= "${EFL_SRCREV}" Sep 16 11:45:01 9:44PM ~/oe/org.openembedded.dev > grep "EFL_SRCREV ?=" conf/distro/include/sane-srcrevs.inc Sep 16 11:45:01 EFL_SRCREV ?= "35961" Sep 16 11:45:10 ah, is that from mtn or git? Sep 16 11:45:17 from git Sep 16 11:45:19 i sent patches Sep 16 11:45:24 i suspect they didnt get merged right Sep 16 11:45:33 i made illume track efl srcrev Sep 16 11:45:42 (as it just makes much mroe sense considering...) Sep 16 11:46:01 http://git.openembedded.net/?p=org.openembedded.dev.git;a=blob;f=conf/distro/include/sane-srcrevs.inc;h=e21ed7fef522d68e8571ed5bbf23a2bb3b1ea529;hb=HEAD Sep 16 11:46:06 it's missing from OE git head Sep 16 11:46:27 EFL_SRCREV ?= "35984" Sep 16 11:46:52 raster: so are you building from OM git repo or OE git repo? Sep 16 11:47:01 oe Sep 16 11:47:08 (thats what i use to do my fso builds) Sep 16 11:47:33 git clone git://git.openembedded.net/org.openembedded.dev.git Sep 16 11:47:35 that git repo Sep 16 11:48:51 so that link i just pasted is supposed to be the head of that repo, and it doesn't match what you pasted Sep 16 11:48:58 so what's the delta? Sep 16 11:49:57 just a sec Sep 16 11:51:00 35984 seems to build fine ... Sep 16 11:51:49 hmm Sep 16 11:51:53 sosmoen removed illume Sep 16 11:51:58 i just git-pull'd Sep 16 11:53:10 they removed the last lines Sep 16 11:53:21 mickey|tv: http://git.openembedded.net/?p=org.openembedded.dev.git;a=commitdiff;h=625f68c086ef6084c98b9a247903c724d26e9bad Sep 16 11:53:28 that's the commit which removed it Sep 16 11:54:45 yup Sep 16 11:54:49 looks like mickey broke it Sep 16 11:54:51 it WAS there Sep 16 11:54:52 :) Sep 16 11:54:56 :-) Sep 16 11:56:06 ok Sep 16 11:56:12 i have a fecking conflict Sep 16 11:56:14 i resolve it Sep 16 11:56:22 how do i get git to shut up and move the fuck on? Sep 16 11:56:42 there's some --resolve flag I think Sep 16 11:56:44 9:54PM ~/oe/org.openembedded.dev > git-pull Sep 16 11:56:45 fatal: Entry 'conf/distro/include/sane-srcrevs.inc' would be overwritten by merge. Cannot merge. Sep 16 11:56:45 Merge with strategy recursive failed. Sep 16 11:56:55 9:55PM ~/oe/org.openembedded.dev > git-mergetool Sep 16 11:56:56 merge tool candidates: kdiff3 tkdiff xxdiff meld gvimdiff opendiff emerge vimdiff Sep 16 11:56:56 No files need merging Sep 16 11:57:07 * rwhitby dunno Sep 16 11:57:46 fucking git Sep 16 11:57:55 i have come to hate git Sep 16 11:57:59 with a passion Sep 16 11:59:41 cant commit it Sep 16 11:59:43 raster: I can push it to OE mtn if you like Sep 16 11:59:45 nothnig to merge Sep 16 11:59:46 cant pull Sep 16 11:59:58 fucking git Sep 16 12:01:11 well reversing those 4 lines removed will fix the build Sep 16 12:01:18 also it really shoudl use EFL_SRCREV Sep 16 12:01:53 + Sep 16 12:01:53 +SRCREV_pn-illume ?= "${EFL_SRCREV}" Sep 16 12:02:06 that's what I will commit to OE mtn Sep 16 12:02:34 (as soon as monotone takes half an hour to pull down 49 revs ...) Sep 16 12:02:38 that will fix it Sep 16 12:02:49 i'd generat u a diff Sep 16 12:02:53 but git is now just being a bitch Sep 16 12:03:03 it does this to me so often Sep 16 12:03:05 I've already made that change and tested it Sep 16 12:03:10 the image built fine Sep 16 12:03:16 it should Sep 16 12:03:20 illume and efl will keep in sync Sep 16 12:03:24 i make sure of that Sep 16 12:03:24 :) Sep 16 12:03:36 once i fix some pending nasties Sep 16 12:03:40 and do this illume theme stuff for e Sep 16 12:03:44 and e's new default theme Sep 16 12:03:49 i'll bump a lot of it up Sep 16 12:06:47 pushing now Sep 16 12:07:40 tnx Sep 16 12:07:49 i'll try see if my git will work again some other time Sep 16 12:07:53 for now i give up Sep 16 12:08:09 (and will just hope one day eithe git stops benig a bitch or someoen starts an svn tree...) Sep 16 12:15:19 * rwhitby waits for CIA to report the push ... Sep 16 12:28:31 hi rwhitby, could you read my post concerning Python & FSO on devel list ? I'm stuck . any idea ? Sep 16 12:28:52 write_erase: just catching up on mailing list email at the moment actually Sep 16 12:30:46 rwhitby, ok... your help welcome :-) Sep 16 12:31:19 write_erase: what host OS? Sep 16 12:33:07 Linux Debian Lenny, up-to-date with dpkg-list checked Sep 16 12:34:55 write_erase: is that all of the log.do_install, or did you cut some of it? Sep 16 12:36:17 I cut the begining of the log.. not this end... I'm pushing the whole log on my webserver Sep 16 12:36:56 write_erase: did you try a bitbake -c rebuild python ? Sep 16 12:37:37 I'll try that right now Sep 16 12:40:07 rwhitby, same again ... it stops here: Sep 16 12:40:11 Compiling /home/olivier/moko/fso-testing/tmp/work/armv4t-angstrom-linux-gnueabi/python-2.5.2-ml10/image/usr/lib/python2.5/zipfile.py ... Sep 16 12:40:11 make: *** [libinstall] Error 1 Sep 16 12:40:11 FATAL: oe_runmake failed Sep 16 12:40:44 can you rm -rf that work directory, and the corresponding stamps files and try again? Sep 16 12:40:57 (just the perl dir under work) Sep 16 12:41:01 (not all of work) Sep 16 12:41:11 s/perl/python/ Sep 16 12:41:23 what's stamp file ? Sep 16 12:41:42 e.g. rm -rf /home/olivier/moko/fso-testing/tmp/work/armv4t-angstrom-linux-gnueabi/python-2.5.2-ml10 /home/olivier/moko/fso-testing/tmp/stamps/armv4t-angstrom-linux-gnueabi/python-2.5.2-ml10* Sep 16 12:42:03 rebuild is supposed to do that, but just in case ... Sep 16 12:43:45 if that doesn't work, then I dunno. Sep 16 12:46:22 rwhitby, same again ... I'll ask python guys Sep 16 13:29:40 anyone happen to know how to get the SDIO-WLAN debug messages to show up? in the source, they look like DBG_PRINT(SDDBG_TRACE, ..) (e.g. drivers/sdio/stack/busdriver/sdio_bus_misc.c) Sep 16 13:30:42 mickeyl: ping - what do you want to do about fso3 feeds? Sep 16 13:34:29 will issue out ms 3.1 Sep 16 13:41:20 mickeyl: current images on shr.bearstech.com are up to date after the push I made to fix the missing illume sane-srcrev Sep 16 13:41:33 rwhitby: ok, cool Sep 16 13:41:38 thanks Sep 16 13:42:31 rwhitby: so the shr.bearstech.com images are e.g. fso ms3.0.1 Sep 16 13:42:46 bumbl: that's mickeyl's call Sep 16 13:43:06 mickeyl: will you fix the my-distribution.org thing first? Sep 16 13:43:44 mickeyl: oh, actually I just realised I haven't checked shr.bearstech.com, I've been working on the nslu2-linux autobuilder Sep 16 13:44:02 (which we can cname to a freesmartphone.org host if you like) Sep 16 13:44:03 rwhitby: nothing to fix there Sep 16 13:44:14 my-distribution.org is coming when the builder forgets to fix his local.conf Sep 16 13:44:34 my builds did have shr.bearstech.com as feed Sep 16 13:44:42 ah, ok. Sep 16 13:45:15 so I need to add some lines to fso-testing and fso-unstable builds then. Sep 16 13:45:23 do you have some boilerplate I can copy? Sep 16 13:46:13 DISTRO_FEED_PREFIX = "fso-testing" Sep 16 13:46:40 DISTRO_FEED_URI = "http://shr.bearstech.com/fso-unstable/ipk/" Sep 16 13:46:45 (from my local.conf) Sep 16 13:46:48 err Sep 16 13:46:52 fso-testing -> fso-unstable Sep 16 13:47:11 nod Sep 16 13:47:21 Hey Sep 16 13:47:56 mickeyl: so what do you think about a cname for .freesmartphone.org to point to the nslu2-linux feed? Sep 16 13:48:20 sure, why not Sep 16 13:48:24 how about Sep 16 13:48:27 software.freesmartphone.org Sep 16 13:48:37 i can point that to any IP if you want Sep 16 13:48:45 not downloads. or ipkg. ? Sep 16 13:49:10 *shrug* Sep 16 13:49:12 downloads then Sep 16 13:49:24 any preferred directory structure? Sep 16 13:49:41 not really Sep 16 13:50:43 hey mickeyl Sep 16 13:51:05 Ainulindale, do you have some time? Sep 16 13:51:10 A bit yes Sep 16 13:52:10 Ainulindale, atm ophonekitd uses the libframeworkd-phonegui-gtk includes for compiling. Wouldn't it be better to define some extern function(XX); to keep it flexible? Sep 16 13:52:51 The includes should not be in that packages as discussed before Sep 16 13:52:56 They should be in a placeholder package Sep 16 13:53:30 And we should be able to link ophonekitd against a specific version of the library Sep 16 13:53:41 Or use dlload and so on Sep 16 13:54:37 Ainulindale, ok, another thing: I want my "daemon" to retry connection if the framworkd (dbus) isn't launched already (at startup) to get the earliest connection. That's only possible if dbus_error_handler() doesn't exit() if there's a dbus error. Shouldn't the dbus error be passed to the user callback? Sep 16 13:55:45 ophonekitd does that Sep 16 13:56:16 and the exit has been there because I couldn't take the time to figure what to do in its stead Sep 16 13:56:52 And no, the dbus error should not be passed to the user callback, there should be an error realm specific to unknown errors Sep 16 13:57:13 (or specific errors such as "no method by that name" and so on) Sep 16 13:57:35 atm ophonekitd is not retrying Sep 16 13:58:06 Oh ? I thought so Sep 16 13:58:13 I can't really have a look at the code right now Sep 16 13:58:15 I have to go for a bit Sep 16 13:58:19 I'll be back in ten minutes Sep 16 13:58:28 Ainulindale, zhone handles retrying, ophonekitd only tries to power up antenna after 5 seconds Sep 16 13:59:07 mickeyl: can you point downloads.freesmartphone.org to 140.211.169.169 ? Sep 16 13:59:56 * mickeyl notes Sep 16 14:00:59 you should find some empty dirs there, which will be populated after an rsync Sep 16 14:02:19 done Sep 16 14:06:57 mickeyl: which package do I need to bitbake -c rebuild to pick up those feed settings? Sep 16 14:07:09 distro-feed-config Sep 16 14:07:15 +s ? Sep 16 14:07:27 ya, +s Sep 16 14:08:05 mickeyl: can you review the directory structure please? Sep 16 14:08:49 that looks well! Sep 16 14:11:43 I'll add a README to the top-level, like on shr.bearstech.com Sep 16 14:12:38 quickdev: that's because the dbus error exits the program Sep 16 14:12:56 Ainulindale, I sent you a mail Sep 16 14:12:58 Modifying that should correct the problem Sep 16 14:13:34 mickeyl: did you see my hilight about how zhone didn't allow me to view messages and sound profiles didn't work ? Sep 16 14:13:50 no, missed that Sep 16 14:13:59 fresh MS3 Sep 16 14:14:05 sound profiles i never tested, tbh. Sep 16 14:14:10 I have recurrent sound alerts Sep 16 14:14:13 view messages should work though Sep 16 14:14:17 gimme logs Ö:) Sep 16 14:14:18 Which I matched with incoming SMSes Sep 16 14:14:28 No logs whatsoever about that (didn't run debug though) Sep 16 14:14:30 But what is weird is Sep 16 14:14:39 ZHone launches, get the dbus connection Sep 16 14:14:48 Displays the dialer function Sep 16 14:14:53 Then wait for some time Sep 16 14:14:57 Displays the addressbook function Sep 16 14:15:06 But whatever I do the messages function is greyed Sep 16 14:15:14 interesting Sep 16 14:15:22 that looks like it never can retreive the sms from your SIM Sep 16 14:15:24 I tried to unload/reload it Sep 16 14:15:29 please run w/ debug Sep 16 14:15:43 mickeyl: what's weird is that I have the sound for OM2007.2 Sep 16 14:15:50 Telling me, I suppose, I received SMSes Sep 16 14:15:59 heh, no Sep 16 14:16:01 (you know the startup sound) Sep 16 14:16:04 yeah Sep 16 14:16:09 fun thing is: Sep 16 14:16:18 the notification sound is getting done by the framework Sep 16 14:16:21 through the rules :) Sep 16 14:16:28 so the framework works Sep 16 14:16:34 and zhone has problems ;) Sep 16 14:16:38 [as expected] Sep 16 14:16:59 we all rely on you guys for a nice UI Sep 16 14:17:07 zhone has no [software] design Sep 16 14:17:30 quickdev: I don't like that patch :-p Sep 16 14:17:37 quickdev: I don't want to propagate dbus errors as is Sep 16 14:17:51 Ainulindale, would you make a little patch which passes an appropriate error? Sep 16 14:17:54 I want to translate them into some simple stuff with debug logging Sep 16 14:18:05 I could but I truly don't have any time for that right now Sep 16 14:18:16 quickdev: see what I did on the sim errors and such ? Sep 16 14:18:29 I think what is necessary is to define an error domain for typical dbus errors and do the same Sep 16 14:18:45 i.e. define an enumeration of errors we want to propagate Sep 16 14:18:50 and default to unknown Sep 16 14:19:29 Ainulindale, that is what I tried first, but I ran into a design problem as far as I remember Sep 16 14:19:40 Please detail that :-) Sep 16 14:20:02 I would suggest to build a specific dbus-glib_handle_errors Sep 16 14:20:21 which would translate the errors appropriately Sep 16 14:20:30 (I truly hate not being able to do that) Sep 16 14:20:42 mickeyl: by debug you mean verbose for ogsmd ? Sep 16 14:20:51 ya Sep 16 14:20:54 [ogsmd] Sep 16 14:20:58 log_level = DEBUG Sep 16 14:20:59 (yeah, yeah :-p) Sep 16 14:21:06 and run zhone on the console Sep 16 14:21:10 so we see its output as well Sep 16 14:21:16 doesn't it log to /tmp ? Sep 16 14:21:24 right Sep 16 14:21:27 it should iir Sep 16 14:21:28 c Sep 16 14:21:36 Pfffft ! You don't even know what you did :-) Sep 16 14:21:50 a symptom of me doing too much Sep 16 14:21:53 too many context switches :) Sep 16 14:22:01 Ainulindale, sim_handle_errors() for dbus? And: #define SIM_ERROR g_quark_from_static_string(SIM_INTERFACE) - what do I use instead of SIM_INTERFACE here? Any arbitrary value indicating that it relates to dbus? Sep 16 14:22:05 Well Sep 16 14:22:10 I have a little problem it seems =) Sep 16 14:22:18 Screen isn't res^ponding :-) Sep 16 14:22:42 screen must be off because of frameworkd Sep 16 14:22:52 quickdev: exactly Sep 16 14:23:05 Just put DBUS_REALM Sep 16 14:23:10 and some random string Sep 16 14:23:12 ok, I'm going to do that Sep 16 14:23:23 Usually I use things like 'Oh yeah, I do really like panties' Sep 16 14:23:27 mickeyl: http://downloads.freesmartphone.org/, first line: "official" or "unofficial" ? Sep 16 14:23:27 But you can put something more appropriate Sep 16 14:24:16 mickeyl: how do I reenable the screen manually ? Sep 16 14:24:20 (display) Sep 16 14:24:30 rwhitby: as much as FSO is independent, this may as well be the 'official' feeds. Sep 16 14:24:45 Ainulindale: no idea i'm afraid. me not using screen too much Sep 16 14:25:03 You're not using the screen ? :-) Sep 16 14:25:05 mickeyl: ok, I'll organise access for yourself to the autobuilder and feeds machine Sep 16 14:25:12 rwhitby: that'd be great Sep 16 14:25:16 odeviced dimmed/disabled it and I can't find how to reenable it in fact Sep 16 14:25:20 Ainulindale: aaah Sep 16 14:25:22 mickeyl: you know ka6sox already, right? Sep 16 14:25:25 i thought you meant the 'screen' util Sep 16 14:26:04 Ainulindale: send org.freesmartphone.Device.IdleNotifier.SetState("BUSY") to /org/freesmartphone/Device/IdleNotifier/0 Sep 16 14:26:11 rwhitby: i should, but i'm not sure Sep 16 14:26:27 mickeyl: Tom King, who runs all the nslu2-linux.org and nas-central.org infrastructure Sep 16 14:26:38 ah, ya, name rings a bell Sep 16 14:26:56 if ever you can't reach me, he can do anything you need on downloads.freesmartphone.org Sep 16 14:27:04 great. thsnk Sep 16 14:27:39 hmmm the illume plugin for network status keeps being inquisitive Sep 16 14:28:22 Or is it Sep 16 14:28:29 My framework seems to be in some kind of loop Sep 16 14:28:39 ya, it hammers the framework while it startsup Sep 16 14:28:44 should stop in a few seconds Sep 16 14:28:49 it didn't Sep 16 14:28:55 yep Sep 16 14:28:55 but then I killed frameworkd like a pig Sep 16 14:29:04 and now zhone won't connect Sep 16 14:29:57 hmmm that's weird Sep 16 14:30:06 I started gsm0710muxd then frameworkd Sep 16 14:30:12 without anything specific launched Sep 16 14:30:16 a quick question: Where do I have to put my application in order to be displayed by the illume launcher? Sep 16 14:30:18 and it keeps being hammered on the getstatus method Sep 16 14:30:25 and getsignal strength Sep 16 14:30:31 whereas I don't have the shelf plugin Sep 16 14:30:42 Ainulindale: here it only hammers it for a couple of seconds Sep 16 14:30:55 and here it won't stop without zhone or the shelf plugin Sep 16 14:30:59 quickdev: categoriy=Office Sep 16 14:31:09 And now zhone won't even connect to dbus Sep 16 14:31:11 quickdev: or somewhere in the fd.o stuff Sep 16 14:31:18 Ainulindale: sounds your setup is a bit b0rked Sep 16 14:31:25 Oh, no, I see why Sep 16 14:31:29 2008-09-16 14:28:57,229 WARNING could not create proxy for org.freesmartphone.opreferencesd:/org/freesmartphone/Preferences Sep 16 14:31:32 mickeyl, category=Office, which directory is that file? Sep 16 14:31:37 quickdev: the .desktop file Sep 16 14:31:41 quickdev: for said app Sep 16 14:31:45 mickeyl, ah, ok, thanks Sep 16 14:33:16 mickeyl: seems unloading the plugin from illume doesn't unload it from memory Sep 16 14:35:20 ok registered to network there Sep 16 14:36:06 mickeyl: http://rafb.net/p/3u8Dya26.html Sep 16 14:36:14 and now it didn't even load contacts Sep 16 14:37:17 error is weird Sep 16 14:40:06 mickeyl: BTW, openembedded/packages/openmoko-projects/diversity-nav_svn.bb needs fixing in OE unstable. It's pointing to a non-existent svn dir, which causes bitbake autorev to fail parsing Sep 16 14:44:09 Turning on antenna is awfully long Sep 16 14:45:27 Well, that's awfully weird Sep 16 14:45:34 Can someone tell me how to redirect frameworkd's output into a file? fsom3 crashes more or less regularly while the screen is turned of and I cannot seem to produce useful logs... :) Sep 16 14:47:44 Hmmm Sep 16 14:47:50 Looks as if there's some inconsistency there Sep 16 14:48:15 mickeyl: got a preference for the builder when reporting fso builds to tinderbox? I was thinking 'freesmartphone.org' ... Sep 16 14:48:28 Tanuva: read openmokoframework on the wiki Sep 16 14:48:34 rwhitby: sounds good Sep 16 14:49:09 mickeyl: fresh start, step by step : gsm0710muxd start; frameworkd -s ogsmd,odeviced,opreferencesd,oeventsd,ousaged,ogpsd; zhone Sep 16 14:49:15 asks for pin Sep 16 14:49:18 then doesn't do a thing Sep 16 14:49:28 except SimRetrieveMessageBook Sep 16 14:49:40 Won't activate any button but settings and gps Sep 16 14:49:40 mickeyl: actually, I'll do testing@freesmartphone.org and unstable@freesmartphone.org, so you can check those builds independently Sep 16 14:49:46 ah, even better Sep 16 14:49:53 Ainulindale: warmstart will not work Sep 16 14:50:00 please kill gsm0710muxd before doing warmstart Sep 16 14:50:07 gsm is sending out a "SIM ready" signal Sep 16 14:50:10 which you won't receive on warmstart Sep 16 14:50:13 only on coldstart Sep 16 14:50:20 That's what I did Sep 16 14:50:22 zhone enables the buttons once it gets the SIM ready signal Sep 16 14:50:23 I killed it Sep 16 14:50:26 oh Sep 16 14:50:30 logs? Sep 16 14:51:07 http://rafb.net/p/J6sn1F45.html Sep 16 14:51:11 you'll have my pin Sep 16 14:51:18 Too bad :-) Sep 16 14:51:34 * mwester catches a key line in this conversation, which explains a *lot* Sep 16 14:51:34 it timeouts visibly Sep 16 14:51:45 mwester: ? Sep 16 14:51:54 warmstart/coldstart Sep 16 14:51:57 ah Sep 16 14:52:21 calypso sends some info once it copied SIM information into its internal memory Sep 16 14:52:38 if you try to access SIM memory before that, you get errors Sep 16 14:52:39 I only killed and restarted zhone and frameworkd, and was pulling out my hair the other evening. :D (and I'm starting to lose my hair, so I wish to avoid that!) Sep 16 14:52:43 mickeyl: on my steps to reproduce bugs I always do as if I was starting from scratch Sep 16 14:52:52 so I kill everything, and do the aforementioned things Sep 16 14:52:57 mickeyl: I did, but there's no exact advice and even frameworkd &> fwd.log doesn't redirect anything Sep 16 14:53:42 Ainulindale: that looks like you don't have any data on SIM: 321 is 'invalid memory index' Sep 16 14:53:44 What bothers me is the timeout Sep 16 14:53:54 mickeyl: It was working approximatively ten minutes ago :-) Sep 16 14:54:09 well, there are some commands where the default dbus timeout doesn't cut it Sep 16 14:54:14 you need to increase it for some commands Sep 16 14:54:51 mickeyl: this morning I had and received calls Sep 16 14:54:59 I also had inconsistent behaviours for calls Sep 16 14:55:08 Er, are you debugging the issue where CMGL gets a CMS 321 error? Sep 16 14:55:25 Sometimes zhone wouldn't call ("dial" => talking => hang screen cinematic) Sep 16 14:55:32 (cms 321) not much to debug there Sep 16 14:55:45 if you query a category of messages that is empty, you get 321 Sep 16 14:55:51 that's default behaviour Sep 16 14:55:55 Ainulindale: nasty Sep 16 14:55:57 mickeyl: right now it didn't even register to the network it seems Sep 16 14:55:58 Ainulindale: didn't see that one yet Sep 16 14:56:00 mickeyl: http://downloads.freesmartphone.org/fso-testing/conf/, so you always know what's being built Sep 16 14:56:08 Whereas I started from scratch Sep 16 14:56:12 (that's rsynced on each build) Sep 16 14:56:19 ah, good Sep 16 14:56:34 Right. I hacked mediator.py to accept either "OK" or CMS: ERROR 321 as responses; and got rid of that endless exchange of CMGL/CMS ERROR messages. Sep 16 14:56:40 Well, I killed everything. Retrying. Sep 16 14:56:56 That's mightily weird. Sep 16 14:57:08 sometimes the modem is really fscked up Sep 16 14:57:16 then it only helps removing power for a while Sep 16 14:57:21 but that's _very_ rare Sep 16 14:57:23 ogsmd DEBUG : TIMEOUT 'ATZ' => ??? Sep 16 14:57:24 ogsmd DEBUG : TIMEOUT 'AT+COLP=1' => ??? Sep 16 14:57:27 That leads me to this Sep 16 14:57:37 ok, that's not good Sep 16 14:57:43 You think ? :-) Sep 16 14:57:48 check that you don't have an accidantal frameworkd process in background Sep 16 14:57:54 sometimes this happens for me Sep 16 14:57:58 * mwester debugs by watching the data send/rcvd by the kernel serial driver. Sep 16 14:58:08 i try to kill the process, but it still doesn't release the VCs Sep 16 14:58:13 and calypso only has 4VCs Sep 16 14:58:20 mickeyl: nope, no process Sep 16 14:58:33 tried to launch zhone, and no antenna power Sep 16 14:58:34 unfortunately it will happily allocate lots of VCs Sep 16 14:58:35 I think I'll reboot Sep 16 14:58:38 but only the first 4 will work Sep 16 14:58:47 ~lart 07.07 Sep 16 14:58:50 err Sep 16 14:58:51 07.10 even Sep 16 14:59:13 No bot, you'll have to lart it yourself ;-) :D Sep 16 14:59:26 right Sep 16 14:59:38 * mickeyl curses the gsm 07.10 standard Sep 16 14:59:44 hehe Sep 16 15:00:26 Ok, f or a reason I won't be able to explain Sep 16 15:00:35 frameworkd launched a second gsm0710muxd Sep 16 15:00:46 Which is explaining a lot indeed Sep 16 15:01:50 ogsmd DEBUG |...|...|...|...|...|...|...|...|...|...|...|...|...|...> SimSendAuthCode._handleCmeCmsExtError: ENTER ('+CME ERROR: 16',),{} Sep 16 15:01:50 oooh Sep 16 15:01:52 that's no good :) Sep 16 15:01:55 And that's weird too Sep 16 15:02:10 16... incorrect passwrd Sep 16 15:02:12 ogsmd DEBUG : COMPLETED 'AT+CFUN=1' => ['+CME ERROR: 11'] Sep 16 15:02:19 That's the good one Sep 16 15:02:40 ya Sep 16 15:02:43 Well, retried, and it works Sep 16 15:02:49 Can't explain that Sep 16 15:03:10 * mickeyl dances a Calypso Sep 16 15:03:10 Oh Sep 16 15:03:11 >Got a stack Sep 16 15:03:24 And my header won't go back to let me see this stack. Sep 16 15:03:27 mlgkrpokpsdf ! Sep 16 15:03:42 And still 321 on SMS Sep 16 15:04:20 321 in response to which CGML? Sep 16 15:04:26 if it's 4 then it's normal Sep 16 15:04:34 4 is unsent iirc Sep 16 15:04:41 zhone now periodically checks for unsent SMS Sep 16 15:04:49 I'll output the logs to a file Sep 16 15:04:55 good thinking Sep 16 15:04:56 If I can remember how to do that (I'm tired here :-) ) Sep 16 15:05:00 heh Sep 16 15:05:05 check the mdbus wiki page Sep 16 15:05:07 its shown there Sep 16 15:05:19 i need to add a debug_destination to frameworkd Sep 16 15:05:26 atm. you can only change destination via dbus at runtime Sep 16 15:05:42 should do that so that we can log to file right from the start Sep 16 15:06:08 mickeyl did you remember that i ask for more devices support in the FSO framework? Sep 16 15:06:25 ya Sep 16 15:06:29 what device do you want? Sep 16 15:06:37 and what kind of support? Sep 16 15:06:47 adding basic suport is easy Sep 16 15:06:52 modem support may be complicated Sep 16 15:06:52 well, i'm helping the Ida system guys to get openmoko sw working into the em-x270 Sep 16 15:06:56 and i have some advances Sep 16 15:07:14 i made a fso image with the oe repositories Sep 16 15:07:17 Ainulindale, I sent you an email with the changes andreasd made to the svn repo Sep 16 15:07:22 Got the stack Sep 16 15:07:29 quickdev: I'll check that tonight Sep 16 15:07:30 methril: ok, so the tough part will be a modem class Sep 16 15:07:33 mickeyl: any idea how to set "Metadata branch" and "Metadata revision" on http://tinderbox.openembedded.net/builders/testing@freesmartphone.org/ ? Sep 16 15:07:52 rwhitby: no idea, never did anything serious w/ tinderbox. Koen, hrw, or zecke should now Sep 16 15:07:52 Ainulindale, ok, they should be commitable without problems Sep 16 15:07:53 know Sep 16 15:07:58 mickeyl yes Sep 16 15:08:22 methril: luckily we have a couple of examples Sep 16 15:08:27 nice Sep 16 15:08:34 if your device is implicit MUX, then you can base on freescale neptune Sep 16 15:08:43 hi Sep 16 15:08:44 if your device is explicit MUX, then you can base on muxed4line or calypso Sep 16 15:08:47 i have to look deeper Sep 16 15:08:55 if your device is not using MUX at all (baaaaaaad), then you might base on singleline Sep 16 15:08:57 Bad newsd mickeyl it's a real problem :-p Sep 16 15:08:58 but singleline is buggy Sep 16 15:09:06 mickeyl: http://rafb.net/p/NBFKCP32.html Sep 16 15:09:22 Ainulindale: aah Sep 16 15:09:23 that's good Sep 16 15:09:26 unicode I guess ? Sep 16 15:09:27 "good" Sep 16 15:09:30 mickeyl an then i send some patches to the FSO-standard ml? Sep 16 15:09:41 or another ml? Sep 16 15:09:50 methril: yes. although smartphones-userland may be more appropriate Sep 16 15:09:57 smartphones-standards is more API discussion Sep 16 15:10:02 ok Sep 16 15:10:29 Ainulindale: it means you have an SMS that has characters in it that are outside the boundaries of the gsm alphabet Sep 16 15:10:33 i start this week and see how i go through the FSO ;) Sep 16 15:10:42 mickeyl: How's that possible ? Sep 16 15:10:52 Ainulindale: gsm alphabet is just one possible alphabets Sep 16 15:10:52 MMS perhaps ? Sep 16 15:10:58 Oh Sep 16 15:11:01 Ainulindale: we need to the PDU data for your SMS Sep 16 15:11:01 mickeyl: if you watch http://tinderbox.openembedded.net/builders/testing@freesmartphone.org/, then once om-gta01 and om-gta02 have built images and package feeds, then it should rsync to downloads. I'm off to bed now. Sep 16 15:11:13 Ainulindale: please mail to the list about that. you need to take that up with our PDU expert alphaone Sep 16 15:11:14 mickeyl: how can I get that for you ? Sep 16 15:11:40 mickeyl: it shouldn't take too long since everything should be up to date, so it's just a rootfs rebuild for two machines. Sep 16 15:11:41 rwhitby: excellent. thanks and good night Sep 16 15:12:05 mickeyl: now I can't complain that FSO feeds are done wrong ;-) Sep 16 15:12:10 nice to see oestats for fso :) Sep 16 15:12:43 rwhitby: :D Sep 16 15:12:55 mickeyl: well so I have to insult alphaone, ok :-) Sep 16 15:13:03 Ainulindale: yep. but this is very important for us Sep 16 15:13:07 i bet we have more of these problems Sep 16 15:13:11 so we can iron out pdu support! Sep 16 15:13:20 pdu has been brandnew in ms3 Sep 16 15:14:40 I'll send you the whole log Sep 16 15:14:47 because unicode doesn't seem to work on phonebook either Sep 16 15:15:02 Oh, didn't say anything Sep 16 15:15:06 Just double checked it Sep 16 15:15:20 Logs are confusing :-) Sep 16 15:15:59 Ok got the content Sep 16 15:16:11 thanks a lot Sep 16 15:16:17 * mickeyl loves log Sep 16 15:16:21 the more the merrier ;) Sep 16 15:16:28 Hmmmpf Sep 16 15:16:34 How could I more disagree with that ? :-) Sep 16 15:16:37 I hate logs Sep 16 15:16:43 hehe Sep 16 15:16:45 If logs are useful it means the software doesn't work ! Sep 16 15:16:53 true that Sep 16 15:17:01 but without logs you have a hard time fixing software that doesn't work Sep 16 15:17:12 and foss is all about incremental development ;) Sep 16 15:17:15 that's why I don't fix it myself Sep 16 15:17:21 I'd rather build anew :-p Sep 16 15:17:32 * mwester has spent his life with logs, and by that definition, with defective software. ... sounds about right... Sep 16 15:17:50 mwester: that's because you're a masochist Sep 16 15:18:21 Yeah, well, I could have been a developer, but then I would have had to move to California, or Boston. So I ended up begin a field tech guy instead. Sep 16 15:18:41 s/begin/being/ Sep 16 15:19:18 But I don't think software developers have any easier life, in the end. Sep 16 15:20:28 alphaone: wake up, lazy boy Sep 16 15:21:14 Ainulindale: Not lazy, just busy with other stuff... Sep 16 15:21:44 But yeah, my mail and other OM related backlog needs to go doen Sep 16 15:21:46 down Sep 16 15:22:03 And it will for sure once I'm in .tw next week :-) Sep 16 15:22:20 Which mail should I send this log ? Sep 16 15:22:35 Knowing I'll let the numbers in so I won't mail it to any kind of mailing list Sep 16 15:27:38 Ainulindale: smartphones-userland Sep 16 15:27:55 mickeyl: you didn't read my second line did you ? :-) Sep 16 15:28:15 Ainulindale: mlauer@vanille-media.de Sep 16 15:28:26 and for alphaone ? Sep 16 15:28:30 (if he's still around) Sep 16 15:28:31 daniel@openmoko.org Sep 16 15:28:43 hmm Sep 16 15:28:46 or alphaone@openmoko.org Sep 16 15:28:48 can't remember that Sep 16 15:28:51 let me check Sep 16 15:29:12 Sep 16 15:29:14 Already sent, too bad :-p Sep 16 15:29:16 use that one for alphaone Sep 16 15:29:18 heh, nm Sep 16 15:29:42 I ought to work a bit on my things but I can't find the time Sep 16 15:29:52 sounds familiar to me Sep 16 15:30:39 Yeah but you're paid to do what you're doing Sep 16 15:30:47 I'm not paid for my work on openmoko stuff Sep 16 15:30:52 that's right Sep 16 15:31:09 although i find myself working twice the amount of what I'm getting paid for Sep 16 15:31:10 :D Sep 16 15:31:16 And right now there's a big contract to win for my company Sep 16 15:31:23 So I'm doing... visio diagrams Sep 16 15:31:28 What a nice way to spend time Sep 16 15:31:35 uh oh Sep 16 15:31:35 mickeyl: by the way Sep 16 15:31:44 http://www.calaquendi.org/om/lf.png Sep 16 15:31:46 did you see that ? Sep 16 15:31:58 It could be helpful for the announcement you should have post a while ago :-p Sep 16 15:32:03 +ed Sep 16 15:32:07 yeah, that's pretty cool Sep 16 15:32:10 going to include that Sep 16 15:32:16 ooh right Sep 16 15:32:19 i completely forgot about that Sep 16 15:32:20 ~lart me Sep 16 15:32:27 I can do it but then as I told before Sep 16 15:32:33 I'm not that well known on the "community" Sep 16 15:32:44 sure, no problem. i should do that Sep 16 15:32:44 And I'd rather have it done by someone who has some legitimity Sep 16 15:32:45 and i will Sep 16 15:32:57 devel@ and cc: shr? Sep 16 15:33:07 and probably smartphones-userland as well Sep 16 15:33:11 Yep :-) Sep 16 15:33:17 I should subscribe to the last one Sep 16 15:33:22 yea Sep 16 15:33:24 Can't remember the URL, if you have it near Sep 16 15:35:24 http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/smartphones-userland Sep 16 15:36:02 Done thanks :-) Sep 16 15:36:11 Did you think about the meeting we talked about a while ago ? Sep 16 15:36:24 wurp2|gone's going to be in switzerland soon enough it would be great to plan that Sep 16 15:36:42 I owe him a good spank Sep 16 15:39:18 hehe Sep 16 15:39:36 the rest of 08 is seriously booked Sep 16 15:39:42 flying to taipei on sunday Sep 16 15:39:46 3 weeks Sep 16 15:39:51 then munich for systems fair Sep 16 15:40:02 then Leipzig for a presentation Sep 16 15:40:05 and then the year is almost over Sep 16 15:40:26 how bout fosdem in february? Sep 16 15:40:35 That's what I was writing Sep 16 15:41:11 mickeyl: so no time for vorarlberg :p Sep 16 15:41:18 bumbl: unfortunately not, sorry Sep 16 15:41:32 i seriously need to cut down my travelling Sep 16 15:42:01 yeah Sep 16 15:42:59 from when to when is the fosdem? Sep 16 15:43:07 always the last weekend in februrary Sep 16 15:46:27 :( thats the week after my holidays Sep 16 15:55:13 bummer Sep 16 16:04:59 rwhitby, will you dare rm -fr $MOKO/fso-unstable/tmp on the tinderbox, so the next build is a full one ? Sep 16 16:10:22 Ainulindale: It'll probably be at least 3-4 months before I'm in Switz Sep 16 16:14:12 You'll be there in february then Sep 16 16:14:30 I haven't (and won't) start looking for work there until my stepdad's bad health situation is resolved - I might start looking in Dec or Jan regardless, but I suspect not. Google finally came back with an answer - if I want to go do an in-person interview in the US, they want me. But the Zurich office managers said my phone interview results weren't impressive enough for them to bring me in, since they're hiring well from the EU :-( Sep 16 16:14:57 Too bad :-/ Sep 16 16:15:11 Yeah, I can probably make February. I don't think my stepdad will make it through the year :-( Sep 16 16:15:21 hmpf :-/ Sep 16 16:15:46 I really don't feel I can leave w/o helping my mom through the time after he dies Sep 16 16:16:36 I have a friend who works in Geneve, maybe I should start asking him what he knows about getting qualified to be employed in Suisse Sep 16 16:17:05 He lives in Annecy Sep 16 16:31:55 I know some people there in the identity management field wurp2|working Sep 16 16:41:17 wurp2|working: hey Sep 16 16:47:38 aleix: Hey! Sep 16 16:47:56 Ainulindale: identity management... Sep 16 16:48:02 what is that? :-) Sep 16 16:48:13 Don't say LDAP Sep 16 20:10:38 hey Sep 16 20:13:34 ho Sep 16 20:34:31 Ainulindale, here? :) Sep 16 20:36:41 not for long Sep 16 20:37:35 :-) Sep 16 20:38:32 Ainulindale, I send you the call status "release" fix to your email, can be applied after the patch of andreasd Sep 16 20:44:24 I'll do that first thing tomorrow Sep 16 20:44:33 You have the legitimate right to whip me if I don't Sep 16 20:48:36 Ainulindale: what's with all the bondage today? Sep 16 20:48:54 Looking forward to an interesting night with the gf sometimes soon? Sep 16 20:50:20 -s Sep 16 20:57:45 If only Sep 16 20:58:04 Unfortunately these times she's experiencing some quality audits at work Sep 16 20:58:12 Which are tiring Sep 16 20:58:19 And I have lots of things to do Sep 16 21:03:14 ugh Sep 16 21:14:42 Ainulindale: Hey, did you ever get your invisible shield? Sep 16 21:44:15 wurp2|working: in the end yes Sep 16 21:46:34 Cool, how do you like it? Sep 16 22:03:13 did someone try to generate vala bindungs for libframworkd-glib? Sep 16 22:13:24 wdraxinger, any news according boot things? :) Sep 16 22:15:25 still working on getting a sane toolchain up and running, that has not been generated with MokoMakefile Sep 16 22:17:26 Well, actually the toolchain works. It's now the basic system and libraries. uClibc is built. so is libowfat and embutils Sep 16 22:24:14 libowfat? do you want to run opentracker on your FR? Sep 16 22:25:40 playya: I don't know how you associate libowfat with opentracker. Sep 16 22:26:13 It's basically a set of very slim written functions for standard tasks, especially for embedded systems. Sep 16 22:26:54 A reimplementation of DJB's excelent libraries under the GPL http://www.fefe.de/libowfat/ Sep 16 22:29:46 wurp2|working: it's nice but weird Sep 17 00:05:59 morning Sep 17 00:06:58 w_erase: are you talking about rebuilding the feed on shr.bearstech.com or the new one on downloads.freesmartphone.org ? Sep 17 01:48:43 mickey|dinner: fso-testing images and feeds should now all be up to date on downloads.freesmartphone.org **** ENDING LOGGING AT Wed Sep 17 02:59:57 2008