**** BEGIN LOGGING AT Tue May 03 02:59:58 2011
May 03 04:21:07 <lindi-> antrik: firefox can ptrace my terminal emulator
May 03 06:37:31 <morphis> heyho
May 03 06:39:42 <JaMa|Zzz> moin
May 03 07:47:11 <morphis> mickeyl: ping
May 03 10:19:10 <morphis> mickey|office: heyho
May 03 10:19:14 <mickey|office> morning
May 03 10:19:23 <morphis> GNUtoo: heyho
May 03 10:19:29 <morphis> mickey|office: we have a little problem
May 03 10:19:50 <GNUtoo> hi
May 03 10:20:12 <morphis> we have so many problem with python-pyside as it configuration search for QtCore, qt4 and so on but we have QtCoreE, qtopia, ... with the qt4-embedded package
May 03 10:20:12 <GNUtoo> I hope the problem is really small
May 03 10:20:48 <morphis> do you know (or anyone else) why qt4-embedded name everything different than qt4-x11-free?
May 03 10:21:00 <GNUtoo> yes it does
May 03 10:21:12 <morphis> darkh: heyho
May 03 10:21:12 <mickey|office> i guess they chose that way to ease parallel installation
May 03 10:21:16 <darkh> hi
May 03 10:21:19 <morphis> darkh: I found the problem 
May 03 10:21:21 <mickey|office> i remember that they have been using the same names in earlier times
May 03 10:21:29 <morphis> they do
May 03 10:21:39 <morphis> dark: <morphis> we have so many problem with python-pyside as it configuration search for QtCore, qt4 and so on but we have QtCoreE, qtopia, ... with the qt4-embedded packag
May 03 10:21:44 <GNUtoo> I think it's to make both libs installable on the same machine
May 03 10:21:56 <morphis> but is that useful?
May 03 10:21:59 <GNUtoo> qte is for framebuffer
May 03 10:22:07 <GNUtoo> and qt-x11 is for the x11 version
May 03 10:22:18 <morphis> yes but who wants both on the same machine?
May 03 10:22:38 <GNUtoo> hmmm
May 03 10:22:38 <darkh> thats the problem
May 03 10:22:39 <mickey|office> i don't know, but that's the way it is atm.
May 03 10:22:53 <morphis> mickey|office: maybe we should change it
May 03 10:23:18 <GNUtoo> morphis, maybe look at the reason inside the commits
May 03 10:23:22 <morphis> we can make the variables QT_DIR_NAME and QT_LIBINFIX in qt4-embedded.inc weak and assign them in the distro config
May 03 10:23:38 <morphis> GNUtoo: if the change is not too old ...
May 03 10:24:55 <morphis> "Changed the name of the qtdemo exectable installed
May 03 10:24:56 <morphis>       in the /usr/bin directory to qtdemoE for Qt
May 03 10:24:56 <morphis>       embedded builds.  This allows the Qt embedded
May 03 10:24:56 <morphis>       version of the demos package to co-exist with
May 03 10:24:56 <morphis>       the Qt X11 version."
May 03 10:25:27 <mickey|office> morphis: i'd rather fix the detection than change their names in order not to deviate too much from the packages for other distros
May 03 10:25:38 <morphis> mickey|office: ok
May 03 10:26:28 <morphis> then we have to make to recipes for pyside
May 03 10:26:34 <morphis> pyside-embedded and pyside-x11
May 03 10:27:34 <mickey|office> yes+
May 03 10:27:40 <darkh> and both set some hint to point findqt4.cmake to the right location?
May 03 10:28:04 <darkh> i'm not sure if i still can follow :/
May 03 10:28:07 <GNUtoo> darkh, you're the one making the meta-kde4?
May 03 10:28:11 <darkh> yes
May 03 10:28:32 <GNUtoo> ok,maybe I should try it
May 03 10:28:37 <GNUtoo> on the n900
May 03 10:28:44 <morphis> darkh: with qt4-embedded all libs has the infix E
May 03 10:28:56 <morphis> and even /usr/include/qt4 is named /usr/include/qtopia
May 03 10:29:04 <darkh> i know/guessed
May 03 10:29:32 <darkh> but what is the problem? findqt4.cmake looking for the wroing libs to link against?
May 03 10:29:43 <morphis> it looking for the wrong libs
May 03 10:29:53 <darkh> that can and should be fixed
May 03 10:29:53 <morphis> and it only finds them in host sysroot
May 03 10:30:10 <darkh> that is what i still don't understand
May 03 10:30:11 <morphis> as they are named correctly there
May 03 10:30:21 <darkh> ah .. ^^
May 03 10:30:24 <darkh> search order
May 03 10:30:27 <morphis> there is a qt4-libs-native
May 03 10:30:36 <darkh> and naming
May 03 10:30:39 <morphis> it compiles qt4-anywhere for some native tools which needs qt
May 03 10:30:51 <morphis> naming is the main problem
May 03 10:31:15 <morphis> if the FindQt4 would search for the correct libs (like QCoreE) it would only find them in target sysroot and everything is fine
May 03 10:31:21 <darkh> morphis:  the problem is CMake does not a "staging target directory" but not an "staging native dir"
May 03 10:31:41 <morphis> but you have the FIND_ROOT_PATH right?
May 03 10:31:47 <morphis> and both are already in there
May 03 10:31:48 <darkh> no, wait
May 03 10:32:07 <morphis> so if we make FindQt4 to look for the correct libs everything should be fine
May 03 10:32:19 <darkh> the FIND_ROOT_PATH needs to point to both STAGING_DIR_TARGET and STAGIN_DIR_NATIVE
May 03 10:32:26 <morphis> it does already
May 03 10:33:12 <morphis> my plan is like the following:
May 03 10:33:17 <darkh> morphis: yes pointing it to the right libs would stop searching for them in the wrong place
May 03 10:33:34 <morphis> * add -DQT_LIBINFIX=E to EXTRA_OECMAKE in the python-pyside recipe
May 03 10:33:49 <darkh> that would do the trick
May 03 10:33:58 <morphis> * change the FindQt4.cmake to use the LIBINFIX wherever it needs it
May 03 10:34:01 <darkh> and double-check findqt4.cmake respects it
May 03 10:34:14 <darkh> :)
May 03 10:36:18 <morphis> jepp
May 03 10:38:10 <mickey|office> that change may as well go upstream. it looks like noone ever tried to build pyside against Qt/E, so they might be interested
May 03 10:38:49 <darkh> what do you mean with upstream? the findqt4.cmake, the recipe?
May 03 10:39:38 <mickey|office> yes
May 03 10:39:57 <mickey|office> the cmake portion
May 03 10:41:23 <darkh> findqt4.cmake should rather be bundled with cmake-native recipe... it is painfull if every project ships its own half fixed and hacked version
May 03 10:41:46 <darkh> kde does it multiple times, so i have to fix every subproject
May 03 10:42:14 <morphis> the findqt4 should be fixed for other packages too
May 03 10:42:30 <morphis> but for the first it will stay in recipes/pyside/files
May 03 10:42:48 <morphis> then it should be merged into cmake itself later
May 03 10:42:53 <darkh> i need a copy too... want to have it in sync
May 03 10:43:17 <darkh> my hacked findqt4.cmake is the kde version...
May 03 10:43:35 <darkh> no idea why it is not in cmake yet
May 03 10:45:03 <morphis> darkh: I will commit everything to the aurora/development branch in the git repository hosted here: http://git.freesmartphone.org/?p=openembedded.git;a=summary
May 03 10:45:24 <darkh> ok
May 03 10:46:03 <morphis> so changes to FindQt4.cmake are done, now I have to wait for my qt4-embedded build to complete
May 03 10:46:26 * darkh hopes it works this time
May 03 10:46:52 <morphis> I hope too
May 03 10:48:27 <darkh> i have to fix some packaging issues for kde :( the *.policy file don't want to get packaged
May 03 10:49:18 <darkh> but at least the most vital seems to build now... until my next hack breaks something again :/
May 03 10:49:26 <darkh> +parts
May 03 10:59:11 <darkh> have to leave now, see you later
May 03 11:50:49 <dcordes> hi
May 03 11:51:17 <GNUoo|n900> hi
May 03 11:51:31 <dcordes> GNUoo|n900: how are you ??
May 03 11:51:38 <GNUoo|n900> fine
May 03 11:51:45 <GNUoo|n900> but still tired
May 03 11:52:06 <dcordes> I'm also a bit ill
May 03 11:52:23 <dcordes> I have a flu like bad feeling
May 03 11:52:44 <dcordes> stress+10°C temperatur drop ...
May 03 11:52:46 <GNUoo|n900> ok,too bad....
May 03 11:53:17 <GNUoo|n900> I hope you will go well soon
May 03 11:53:20 <dcordes> no good for dcordes' organism
May 03 11:53:23 <dcordes> thanks you too
May 03 11:54:54 <dcordes> need to read logs
May 03 11:56:25 <dcordes> GNUoo|n900: did you try to compile nexusone kernel with the bcm4329_204 from http://developer.htc.com ?
May 03 11:57:15 <GNUoo|n900> is it the same in git?
May 03 11:57:24 <dcordes> captainigloo: did you try fso with htcleo settings on bravo without sim ?
May 03 11:57:37 <dcordes> GNUoo|n900: htc has no git
May 03 11:57:55 <GNUoo|n900> i mean from wice git
May 03 11:58:55 <dcordes> are you asking if developer.htc.com kernels have same bcm4329 that gives working iwlist scan on htcleo ?
May 03 12:00:53 <GNUoo|n900> no
May 03 12:01:13 <GNUoo|n900> you gave me a link to wice kernel for leo
May 03 12:01:22 <GNUoo|n900> it had 204
May 03 12:01:27 <dcordes> wice ?
May 03 12:01:41 <GNUoo|n900> yes linux on wince
May 03 12:02:03 <dcordes> ooh oh ok
May 03 12:02:37 <dcordes> well it is for a device that has a radio firmware that was originally designed for wince on system processor
May 03 12:02:46 <dcordes> so wince kernel is misleading
May 03 12:02:59 <dcordes> git://gitorious.org/linux-on-wince-htc/linux_on_wince_htc.git linux_on_wince_htc
May 03 12:03:07 <dcordes> this is the kernel I use on my htcleo
May 03 12:03:43 <dcordes> and this kernel, compiled with htcleo_defconfig gives bcm4329_204 which was originally extracted from htc evo kernel. this gives working wilist scan on htcleo
May 03 12:04:17 <GNUoo|n900> ok
May 03 12:04:22 <GNUoo|n900> i tried that
May 03 12:04:25 <dcordes> htc evo = htc evo 4g
May 03 12:05:42 <dcordes> above git is 2.6.32 . but the kernel you wish to use on nexus is 2.6.37 (?) , the bcm4329_204 does not compile
May 03 12:06:05 <dcordes> I recommend to browse http://developer.htc.com/
May 03 12:06:20 <dcordes> pick up the two most recent kernel version tarballs they have
May 03 12:06:28 <dcordes> see if they havem bcm4329 and try them instead
May 03 12:06:58 <GNUoo|n900> ok
May 03 12:06:59 <dcordes> they might have the same characteristics as evo 4g's bcm4329 (because both are from same website)
May 03 12:07:07 <GNUoo|n900> ok
May 03 12:07:10 <GNUoo|n900> bbl
May 03 12:07:28 <GNUoo|n900> (eating)
May 03 12:07:47 <dcordes> desire s (saga) http://dl4.htc.com/RomCode/Source_and_Binaries/saga-2.6.35-crc.tar.gz
May 03 12:08:15 <dcordes> ncredible s (vivo) http://dl4.htc.com/RomCode/Source_and_Binaries/vivo-2.6.35-g89aa373.tar.gz
May 03 12:08:19 <dcordes> I'm also off
May 03 12:08:26 <dcordes> see you later
May 03 12:08:27 <mickey|office> dcordes: get well sooin
May 03 12:08:34 <dcordes> mickey|office: thanks a lot
May 03 12:08:39 <GNUoo|n900> ok see you later
May 03 12:56:28 <CIA-10> freesmartphone.org: 03k3dar7 07utilities * r1606e0e42844 10/palmpre/fso-installer/Makefile: 
May 03 12:56:28 <CIA-10> freesmartphone.org: fso-installer: some small fixes
May 03 12:56:28 <CIA-10> freesmartphone.org: * check for utilities to extract webos doctor before using them
May 03 12:56:28 <CIA-10> freesmartphone.org: * use unzip if jar utility is not available
May 03 12:56:28 <CIA-10> freesmartphone.org: * add a sleep of 20 seconds to wait for the erasure of the flashed image
May 03 12:56:30 <CIA-10> freesmartphone.org: 03morphis 07utilities * rf26281a1b4bc 10/palmpre/fso-installer/Makefile: 
May 03 12:56:30 <CIA-10> freesmartphone.org: fso-installer: some small enhancements
May 03 12:56:30 <CIA-10> freesmartphone.org: * use which utility to determine if a utility is available or not
May 03 12:56:31 <CIA-10> freesmartphone.org: * memboot: extract doctor first before booting the device into recovery mode
May 03 13:34:24 * morphis is crossing his fingers that the next build of python-pyside succeeds ...
May 03 13:34:46 * mickey|office helps with more fingers
May 03 13:35:19 * PaulFertser adds toes
May 03 13:35:46 <morphis> :)
May 03 13:38:39 <morphis> hm some tiny errors due to my unconcentration 
May 03 13:38:47 <morphis> next build is running ...
May 03 13:39:47 <morphis> yeah! "error while loading shared libraries: libgenrunner.so.0.6: cannot open shared object file: No such file or directory"
May 03 13:39:52 <morphis> no Qt error!
May 03 13:40:26 <morphis> and libpyside was builded successfully!
May 03 13:44:32 <mickey|office> excellent. that's progress
May 03 13:44:54 <morphis> jepp
May 03 13:45:04 <morphis> hm, so generatorrunner-native does not work correctly
May 03 13:45:21 <morphis> seems like LD_LIBRARY_PATH is not set correctly
May 03 13:49:08 <morphis> mickey|office: any idea?
May 03 13:52:46 <morphis> with doing a manual "export LD_LIBRARY_PATH=/home/morphis/oe/buildroot/tmp/sysroots/x86_64-linux/usr/lib" it compiles
May 03 13:55:01 <morphis> mickey|office, JaMa|Off, GNUoo|n900: is that even the right way to set the LD_LIBRARY_PATH in the recipe or there another way to do so?
May 03 13:55:57 <mickey|office> morphis: not really
May 03 13:56:04 <morphis> ok
May 03 13:56:13 <mickey|office> it's wrong, but i personally wouldn't mind if it works
May 03 13:56:19 <mickey|office> the right way is to fix qmake here
May 03 13:56:29 <mickey|office> obviously it doesn't submit the correct -rpath argument
May 03 13:56:32 <mickey|office> @ linking time
May 03 13:56:36 <morphis> ok
May 03 13:56:55 <GNUtoo> morphis, rpath?
May 03 13:57:18 <morphis> GNUtoo: jep
May 03 14:01:55 <GarthPS> morphis: hi! what are you doing ?
May 03 14:02:04 <morphis> mickey|office: qmake? but generatorrunner is not compiled with qmake
May 03 14:02:28 <mickey|office> sure?
May 03 14:02:35 <mickey|office> then it's cmake
May 03 14:02:47 <morphis> GarthPS: trying to get pyside (python qt binding) to compile with OE
May 03 14:03:14 <GarthPS> morphis: hm ok still for aurora right ?
May 03 14:04:09 <morphis> GarthPS: yes
May 03 14:04:27 <morphis> GarthPS: what are you expecting me to do? working on more Palm Pre features? :)
May 03 14:04:56 <morphis> mickey|office: will include the LD_LIBRARY_PATH hack for now and will hope darkh can help us later to remove it 
May 03 14:06:32 <GarthPS> morphis: lol :) hm I hope I will finally do some decent video of it this week as my moving (of appartement) is almost done
May 03 14:09:26 <morphis> GarthPS: ah, your moving :)
May 03 14:09:58 <GarthPS> ?
May 03 14:13:48 <GarthPS> morphis: is there a way to wget latest makefile ?
May 03 14:14:10 <GarthPS> wget http://git.freesmartphone.org/?p=utilities.git;a=blob_plain;f=palmpre/fso-installer/Makefile;hb=HEAD does not obviously work
May 03 14:14:16 <mrmoku> time for some FOSS :-)
May 03 14:14:42 <mickey|office> mrmoku: yay
May 03 14:14:50 * mickey|office still standing by with his drum...
May 03 14:16:42 <morphis> GarthPS: some like this http://git.freesmartphone.org/?p=utilities.git;a=blob_plain;f=palmpre/fso-installer/Makefile;h=7fc861a7cdfadc48f00b7ee1b545247c64df5f36;hb=f26281a1b4bcce7834c62bcf380adbf040c90a20 shoudl work
May 03 14:18:25 <morphis> darkh: welcome back!
May 03 14:18:40 <darkh> not really back
May 03 14:18:51 <darkh> just for some minutes
May 03 14:19:06 <darkh> any success so far with pytside
May 03 14:19:47 <morphis> yes
May 03 14:19:48 <morphis> it compiles
May 03 14:19:59 <mrmoku> mickey|office: have to find some segfaults first :)
May 03 14:20:00 <morphis> the configuration fixes I did works
May 03 14:20:17 <darkh> good to hear
May 03 14:21:01 <morphis> but I have another problem for you
May 03 14:21:09 <morphis> I have a recipe generatorrunner-native
May 03 14:21:14 <morphis> it's even using camke
May 03 14:21:21 <morphis> and it get the rpath wrong
May 03 14:21:34 <morphis>  "error while loading shared libraries: libgenrunner.so.0.6: cannot open shared object file: No such file or directory"
May 03 14:21:48 <morphis> with a manual  "export LD_LIBRARY_PATH=/home/morphis/oe/buildroot/tmp/sysroots/x86_64-linux/usr/lib" it works fine
May 03 14:22:12 <darkh> did not see such a problem so far
May 03 14:22:47 <darkh> seems as the cmake buildsystem does not tell cmake the right thing
May 03 14:23:18 <darkh> actually cmake.bbclass should take care
May 03 14:23:53 <darkh> does the generatorrunner's cmake files set something strange?
May 03 14:24:14 <morphis> will take a look later
May 03 14:26:33 <GarthPS> morphis: yeap but I mean a generic one, not the current last one
May 03 14:26:53 <GarthPS> is it possible to git clone a file ?
May 03 14:28:56 <morphis> GarthPS: no
May 03 14:29:10 <morphis> you can only clone the whole repostiroy
May 03 14:29:40 <morphis> but we can add a hook to the repository so it makes the Makefile available out of the tree everytime when some one pushes a commit
May 03 14:32:38 <mrmoku> mickey|office: btw. the AtPdpSetCredentials mediator does not fit... it's trying to send some AT command
May 03 14:32:54 <mrmoku> so I implemented that one as IsiPdpSetCredentials
May 03 14:33:38 <morphis> mrmoku: how does pdp work with isi?
May 03 14:33:45 <morphis> is there some multiplexing?
May 03 14:37:31 <antrik> lindi-: hm, right, the default policy allows processes to debug all other processes belonging to the same user. the answer is to run dangerous processes under different user IDs. this is possible even on Linux, with some limitations. on systems like the Hurd it would be no problem at all...
May 03 14:38:21 <GarthPS> morphis: ok . I think It could ne cool. like that It will be easier to checkout the last makefile for the install process (perhaps we could even fill the update target of the makefile)7
May 03 14:39:12 <morphis> ok, will do that
May 03 14:39:47 <mrmoku> morphis: isi internal
May 03 14:40:06 <mrmoku> you create a pipe via isi and get a gprs0 iface then
May 03 14:40:27 <morphis> mrmoku: ah ok
May 03 14:41:11 <lindi-> antrik: it is not really easy on any current gnu/linux distro
May 03 14:41:18 <lindi-> antrik: since they use X
May 03 14:42:47 <mrmoku> mickey|office: #0  0x407457e4 in fso_gsm_ipdp_handler_activate (self=0x0, _callback_=0x4071894c <fso_gsm_at_pdp_activate_context_run_ready>, _user_data_=0x1fcec8) at pdp.c:728
May 03 14:42:58 <mrmoku> self being null does not seem to be a good thing :P
May 03 14:46:11 <GarthPS> morphis: thx. tell me when you have done it , I will update the howto then
May 03 14:46:16 <morphis> mrmoku: you created the handler in the modem class?
May 03 14:46:24 <morphis> GarthPS: for sure
May 03 14:48:53 <antrik> lindi-: and how does qubes solve that problem? by running separate X instances I suppose? that should work on any system...
May 03 14:51:10 <lindi-> antrik: yes separate X servers and shared composition buffers over xen shared memory
May 03 14:51:22 <lindi-> antrik: and yes, qubes is software, it can run on most existing hardware :)
May 03 14:52:51 <morphis> hm, pyside is still compiling since 10 minutes without any error message
May 03 14:58:11 <antrik> lindi-: what does hardware got to do with it? I was saying that you don't need a hypervisor to run separate X instances
May 03 15:04:32 <lindi-> antrik: chroots or containers then?
May 03 15:05:00 <mrmoku> morphis: no I have an extra plugin
May 03 15:05:38 <lindi-> antrik: but you are right, you can just take the GUI isolation if you want. but then you need to worry about bugs in linux/dhcp client/etc.. again
May 03 15:06:26 <JaMa|Off> mrmoku: hi, any luck with gdbus-codegen yesterday?
May 03 15:06:31 <mrmoku> no :/
May 03 15:06:40 <mrmoku> no luck whatsoever yesterday
May 03 15:06:57 <JaMa|Off> ok
May 03 15:06:57 <mrmoku> though for me it might be python3 related too
May 03 15:07:00 <antrik> lindi-: I don't think you need chroot to run separate X instances
May 03 15:07:22 <JaMa|Off> mrmoku: but shr-chroot should use python2.7 by default
May 03 15:07:26 <JaMa|Off> mrmoku: no python3
May 03 15:07:28 <JaMa|Off> not
May 03 15:07:31 <lindi-> antrik: but without chroot you have the same dbus-daemon
May 03 15:07:40 <mrmoku> yeah, but building does not generate
May 03 15:07:43 <mrmoku> I did that manually
May 03 15:07:53 <lindi-> antrik: a lot of attack surface :)
May 03 15:07:53 <antrik> I don't know about dbus security
May 03 15:08:02 <mrmoku> gdbus-codegen is only needed for the proxy glue class stuff :/
May 03 15:08:10 <mrmoku> (once it is generated)
May 03 15:08:14 <lindi-> antrik: indeed, desktop stuff is not really designed with this in mind
May 03 15:08:50 <mrmoku> JaMa|Off: if you call 'make gen-bindings' it generates the glue stuff
May 03 15:08:57 <lindi-> antrik: so using hypervisor with very simple interfaces between VMs is a good strategy for making it easy to audit but still usable with common software
May 03 15:09:04 <mrmoku> which then needs to be linked against the lib from gdbus-codegen
May 03 15:09:39 <antrik> actually my impression of dbus so far is that they reinvented the wheel, ignoring several decades of research on simple, powerful, efficient, and secure IPC mechanisms conducted by the microkernel community; instead creating something that's inferior in every one of these regards...
May 03 15:10:01 <lindi-> antrik: well they at least went ahead and implemented it :)
May 03 15:10:34 <mrmoku> morphis: this is my pdp handler class: http://git.freesmartphone.org/?p=cornucopia.git;a=blob;f=fsogsmd/src/plugins/pdp_nokia_isi/plugin.vala;h=0edb6cda2957ed7adaa468b76b50a77fb5d48593;hb=2d7ffb516c9182ef169a76c91e5bcb09f8007b17
May 03 15:11:00 <lindi-> antrik: the hypervisor approach also means that you can use linux's hardware support but still use something else for other appVMs
May 03 15:12:12 <antrik> lindi-: again, my point is that you don't need a hypervisor for any of this. could just as well do it with a traditional microkernel. as I said, the *only* fundamental difference between a microkernel and a hypervisor is that the latter tries to create an environment looking more or less like a "real" machine (though only partially with paravirtualisation), so existing monolithic kernels can be run with relatively little modification
May 03 15:12:51 <antrik> lindi-: note that all that has been done with L4
May 03 15:13:16 <antrik> (running multiple Linux instances, including drivers)
May 03 15:14:14 <lindi-> antrik: qubes does discuss L4 if you read the docs
May 03 15:14:37 <JaMa|Off> mrmoku: so should I try with gdbus-codegen from glib repo or you've tried this too and it didn't work?
May 03 15:15:06 <mrmoku> JaMa|Off: I tried that and failed... if you have old python it would be worth a try though
May 03 15:15:38 <JaMa|Off> I have 2.7 in shr-chroot :)
May 03 15:15:57 <lindi-> antrik: maybe xen had better hardware support than L4? I don't know
May 03 15:16:33 <morphis> mrmoku: you need to instantiate the handler in the initPdpHandler function of the base modem class
May 03 15:16:40 <morphis> http://git.freesmartphone.org/?p=cornucopia.git;a=blob;f=fsogsmd/src/lib/modem.vala;h=0a8c15c00b5a975453b125b1d201790137e3c683;hb=2d7ffb516c9182ef169a76c91e5bcb09f8007b17#l383
May 03 15:16:45 <antrik> Xen has more commercial support/hype; that's the only advantage I can think of :-)
May 03 15:16:59 <lindi-> antrik: yes, it works :)
May 03 15:17:03 <lindi-> antrik: and it's in distributions
May 03 15:17:17 <antrik> L4 and L4Linux work too
May 03 15:17:38 <mrmoku> morphis: duhh... thanks :-)
May 03 15:17:55 <morphis> no problem
May 03 15:18:44 <lindi-> antrik: maybe, I've just never heard of anybody running it
May 03 15:18:53 <antrik> as I said, the idea of qubes isn't really unique. and it doesn't solve any of the fundamental problems microkernel systems are facing as far as I can tell...
May 03 15:19:05 <lindi-> antrik: it solves practical issues :)
May 03 15:19:34 <antrik> and introduces practical issues, just like traditional microkernel systems
May 03 15:20:02 <lindi-> antrik: surely it's tradeoff always
May 03 15:20:41 <lindi-> antrik: but I think in qubes case it is pretty nice tradeoff. I don't need to worry about bugs in linux but I can still run most of the applications that I use normally :)
May 03 15:20:44 <antrik> of course. and I believe the tradeoff is often worth it. but apparently most people don't, or we would all be running microkernel systems
May 03 15:21:43 <lindi-> antrik: well surely microkernel systems have had larger issues? like 32-bit gnu/hurd not supporting disks larger than 4GB around year 2000?
May 03 15:22:34 <misc> I am not sure this is linked to the micro kernel architecture 
May 03 15:23:08 <lindi-> misc: not directly of course but if you talk about "most people" then the implementations of course play a large role
May 03 15:23:22 <morphis> mickey|office: "NOTE: package python-pyside-embedded-1.0.2-r0.0: task do_build: Succeeded"
May 03 15:24:13 <antrik> lindi-: well, not exactly, but you are basically right. the Hurd is tricky, as it has independent implementations of most stuff, which is obviously a lot of work. but as I pointed out, that is not necessary; L4 can run a somewhat modified (you could say paravirtualized) Linux just fine
May 03 15:24:34 <lindi-> antrik: yep, qubes docs point out that you could use L4 instead of xen
May 03 15:25:51 <antrik> lindi-: but as misc pointed out, that is just an implementation issue. it's not that microkernels haven't entered mainstream because of such implementation issues. it's rather that the implementation issues have not been fixed because microkernel systems are not mainstream, due to some fundamental difficulties -- which qubes doesn't solve either
May 03 15:26:35 <misc> ( personnaly, I think microkernel are not upstream because the architecture is more complex and so fewer people can work on it for now, hence lack of coder, etc )
May 03 15:26:47 <misc> ( but maybe I am wrong on this )
May 03 15:27:07 <antrik> (for general-purpose systems that is... in the simpler embedded space microkernel systems have been mainstream for ages)
May 03 15:27:47 <misc> yup, qnx for example is a micro kernel, no ?
May 03 15:27:58 <lindi-> qnx is non-free though
May 03 15:27:59 <morphis> mickey|office: now lets try if it works on the device ...
May 03 15:28:42 <lindi-> antrik: qubes is just collecting all the pieces and glueing them together with qubes specific xorg drivers for example
May 03 15:29:30 <lindi-> antrik: I had been looking for such a GUI isolation system for almost two years. Using composition buffers over xen shared memory never occured to me
May 03 15:30:00 <lindi-> antrik: I was looking at trusted solaris and selinux X extensions which did not seem to go anywhere
May 03 15:30:21 <mickey|office> morphis: congrats... holding my breath for the device test...
May 03 15:30:29 <morphis> :)
May 03 15:30:50 <morphis> takes some minutes as I have to build the image with python support
May 03 15:31:37 <antrik> lindi-: the L4 people implemented a proof-of-concept system with separate domains and integrated windowing system approach many years ago.
May 03 15:32:00 <lindi-> antrik: interesting, I never heard about it
May 03 15:32:05 <lindi-> antrik: are the developers still using it?
May 03 15:32:13 <antrik> right. because it's not Xen, so it doesn't have the hype ;-)
May 03 15:32:19 <antrik> no idea
May 03 15:32:27 <lindi-> antrik: yeah, but some amount of hype is important
May 03 15:32:55 <lindi-> antrik: does it have a git repo?
May 03 15:32:55 <antrik> yes of course. I already said that the hype is a good thing
May 03 15:33:13 <antrik> I'm just pointing out that the qubes concept is not all that new really
May 03 15:33:54 <lindi-> antrik: the idea is probably not new
May 03 15:33:58 <antrik> (and that some of it could be implemented even with stock Linux, if people cared enough)
May 03 15:34:05 <lindi-> antrik: but actually writing the Xorg modules that are needed is?
May 03 15:34:20 <lindi-> antrik: and integrating this to trusted grub and things like that
May 03 15:36:53 <antrik> the special Xorg modules are only necessary because of Xen. on a traditional system, you could just use Xephyr
May 03 15:37:34 <antrik> as for "trusted", leave me alone with that. any claims that this is actually useful for protecting your privacy are bullshit
May 03 15:38:32 <mrmoku> mickey|office: /usr/sbin/fsogsmd: symbol lookup error: /usr/lib/cornucopia/modules/fsogsm/pdp_nokia_isi.so: undefined symbol: gisi_comm_modem_access_get_type
May 03 15:38:45 <mrmoku> hmm
May 03 15:38:51 <mickey|office> not linked against libgisi?
May 03 15:38:55 <mickey|office> (comm)
May 03 15:39:21 * mrmoku forgot to add isi to ld
May 03 15:40:01 <mrmoku> mickey|office: do I need that one too? $(top_srcdir)/src/lib/libfsogsm.la 
May 03 15:41:03 <lindi-> antrik: how exactly would you use xephyr?
May 03 15:41:15 <mickey|office> mrmoku: if you use something from libfsogsm, then yes
May 03 15:41:30 <mrmoku> hmm... not that I would be aware off :P
May 03 15:41:42 <lindi-> antrik: it uses trusted grub to prevent the storageVM from tampering with files in /boot
May 03 15:42:04 <lindi-> antrik: encryption keys in TPM and all that trusted computing stuff is finally put to a good use :)
May 03 15:42:58 * mrmoku picking up son... keep the drums warm mickey|office :)
May 03 15:43:34 <antrik> lindi-: I don't know the details about Xephyr, nor do I really care... we already carried this off-topic discussion here way too long :-)
May 03 15:44:35 <lindi-> antrik: http://www.invisiblethingslab.com/resources/misc-2010/xorg-large-memory-attacks.pdf is from qubes authors and attacks xephyr based system
May 03 15:44:49 <lindi-> antrik: care to move it to #freesoftware? ;)
May 03 15:45:31 <lindi-> antrik: many of your comments have been valuable, don't get me wrong even if I try to defend qubes here :)
May 03 15:53:11 <lindi-> antrik: I would have never heard of L4Linux without you
May 03 15:53:32 <antrik> so would most people... which is really a pity
May 03 15:55:35 <lindi-> antrik: is it in some distro? ;)
May 03 15:56:24 <antrik> no idea. most of the L4 stuff is reasearch only, unfortunately. nobody outside the small microkernel community even knows about it
May 03 15:57:09 <lindi-> chicken and egg problem then
May 03 15:57:24 <antrik> there is a lot of interesting stuff in the L4 world. for example one of the L4 groups has something they call a "microhypervisor" -- an L4-based system with a couple of userspace processes providing the functionality a "traditional" hypervisor like Xen offers
May 03 15:58:28 <lindi-> antrik: sounds potentially cool
May 03 15:59:45 <antrik> another interesting project is a microkernel operating system emphasising isolation, running on top of L4 (or even Linux if you prefer)
May 03 16:01:34 <lindi-> antrik: those need to be polished to the point where they can be added to distros so that you can get more users
May 03 16:01:41 <lindi-> antrik: that's the only way to get more developers
May 03 16:05:03 <morphis> mickeyl: import PySide works
May 03 16:06:11 <GNUtoo> wow nice
May 03 16:06:16 <GNUtoo> congratulations
May 03 16:07:22 <antrik> I think few people realize the close connection between microkernel systems an virtualisation. they are going from opposite directions (microkernel: how can we change system design to get isolated components with tightly controlled communication channels? virtualisation: we have completely isolated VMs; how can we add controlled communication channels, to make it look more like an integrated operating system?); but in the end it's al
May 03 16:07:57 <morphis> GNUtoo: jepp example to show a simple window works too
May 03 16:08:15 <lindi-> antrik: your line was too long
May 03 16:08:16 <GNUtoo> git add; git commit -s;git push 
May 03 16:08:25 <morphis> needs some cleanup
May 03 16:08:41 <morphis> before I can push everything
May 03 16:09:02 <GNUtoo> don't loose the working version tough
May 03 16:09:51 <lindi-> antrik: it ended with "but in the end it's al" ...
May 03 16:11:04 <antrik> hrm. I'll never understand why this irssi thing praised by everyone doesn't even manage the line length, while the crusty simplistic client I used before just worked...
May 03 16:11:05 <morphis> GNUtoo: no, I will push it, but not to upstream OE
May 03 16:11:25 <GNUtoo> ah?
May 03 16:11:49 <antrik> anyways, ... but in the end it's all working towards the same goal
May 03 16:12:50 <lindi-> antrik: I think qubes is an excellent opportunity to get rid of linux :)
May 03 16:13:20 <lindi-> antrik: I can run debian gnu/kfreebsd in an appVM and still use linux's device drivers in storageVM and netVM
May 03 16:13:54 <morphis> GNUtoo: I want mickeyl to look at the recipes first
May 03 16:13:59 <GNUtoo> ok
May 03 16:14:40 <morphis> GNUtoo: look here: http://git.freesmartphone.org/?p=openembedded.git;a=shortlog;h=refs/heads/aurora/development
May 03 16:16:01 <GNUtoo> ok
May 03 16:17:18 <morphis> GNUtoo: but most changes from the aurora/development branch should go upstream as fast as possible
May 03 16:17:27 <GNUtoo> ok
May 03 16:17:36 <GNUtoo> else it could be lost in the oe-core transition
May 03 16:18:39 <morphis> it will not
May 03 16:18:49 <morphis> as I want to move to oe-core after the first release of aurora
May 03 16:19:13 <mrmoku> next try
May 03 16:20:38 <GNUtoo> ok nice
May 03 16:22:23 <mrmoku> still the same undefined symbol :/
May 03 16:22:24 <morphis> but for now we want to base our work on a stable version of classic OE
May 03 16:22:36 <antrik> lindi-: BTW, running a complete Linux in a VM to use it's drivers is another thing L4 people did as a proof of concept long ago... I think it was even before Xen came up
May 03 16:23:03 <GNUtoo> ok,  I should start migrating the games to oe-core
May 03 16:25:12 <mrmoku> duh
May 03 16:25:12 <mrmoku> make[4]: Warning: File `Makefile' has modification time 14 s in the future
May 03 16:39:59 <lindi-> antrik: great
May 03 16:40:15 <lindi-> antrik: so I guess all it lacked was the necessary hype to get users :)
May 03 16:40:51 <lindi-> antrik: and maybe it was too proof of concept only. Joanna actually uses qubes on her normal machine for work
May 03 17:07:59 <playya_> argh. I lost my µSD adapter
May 03 17:08:15 <mrmoku> 1970-01-01T00:55:41.577178Z [WARN]  fsogsmd : GLib <gisicomm>: gisicomm.vala:1672: failed to create PEP
May 03 17:09:12 <mrmoku> picking up the other son now
May 03 17:09:18 * mrmoku feels like a taxi driver :/
May 03 17:09:25 <angelox_123> playya,i have 3.... two are failed :(
May 03 17:10:16 <playya_> where're my cameras
May 03 17:24:37 <GNUtoo>  mickeyl hi
May 03 17:25:46 <mickeyl> morphis: excellent, congrats
May 03 17:25:48 <mickeyl> hi GNUtoo 
May 03 17:25:55 <mickeyl> mrmoku: hehe
May 03 17:26:09 <mickeyl> mrmoku: feel free to commit the present state of things and let me take a look
May 03 17:26:31 <GNUtoo> mickeyl, did you start making the alsa backend of cmtspeechtest?
May 03 17:26:45 <mickeyl> GNUtoo: unfortunately not
May 03 17:26:48 <mickeyl> no time
May 03 17:26:51 <GNUtoo> ok
May 03 17:27:05 <GNUtoo> if I do it badly, is it ok?
May 03 17:27:40 <morphis> mickeyl: jepp but can you take a look at the recipes before I commit them upstream
May 03 17:27:55 <mickeyl> morphis: sure, let me see
May 03 17:28:25 <mickeyl> GNUtoo: of course. if you want, freel free to start.
May 03 17:28:45 <morphis> mickeyl: you already found them?
May 03 17:28:56 <GNUtoo> ok
May 03 17:29:02 <mickeyl> morphis: fso git?
May 03 17:29:05 <morphis> http://git.freesmartphone.org/?p=openembedded.git;a=shortlog;h=refs/heads/aurora/development
May 03 17:29:09 <mickeyl> ah, there
May 03 17:30:01 <mickeyl> ah, you didn't use mine as template
May 03 17:31:01 <mickeyl> i packaged apiextrator and generatorrunner in universal recipes, but for me it's ok to have them only -native
May 03 17:32:21 <mickeyl> hehe, EXTRA_OECMAKE = " \ ...
May 03 17:32:22 <morphis> mickeyl: I used your recipes
May 03 17:32:23 <mickeyl> that's scary
May 03 17:32:42 <morphis> but removed the BBCLASSEXTEND cause it's not needed in a non-native version
May 03 17:32:46 <morphis> jepp
May 03 17:32:48 <mickeyl>        -DQT_DBUSXML2CPP_EXECUTABLE=/usr/bin/qdbusxml2cpp \
May 03 17:32:49 <mickeyl>   32         -DQT_DBUSCPP2XML_EXECUTABLE=/usr/bin/qdbuscpp2xml \
May 03 17:32:52 <mickeyl> absolute paths?
May 03 17:32:59 <mickeyl> that will be frowned upon
May 03 17:33:17 <morphis> oh yes
May 03 17:33:22 <morphis> will fix both
May 03 17:34:56 <mickeyl> for unconditional append we rather prefere += these days
May 03 17:35:04 <mickeyl> but it's your call
May 03 17:35:18 <morphis> where?
May 03 17:35:29 <mickeyl> libshiboken SRC_URI
May 03 17:35:40 <ChristW> mrmoku: I've generated an SSH key, and I'm ready to commit my changes (I guess...)
May 03 17:36:04 <mickeyl> and add the NOTE for why we are doing export LD_LIBRARY_PATH = "${STAGING_LIBDIR_NATIVE}" also in shiboken.inc
May 03 17:36:11 <mickeyl> err, bs,
May 03 17:36:19 <mickeyl> in shiboken-native i mean
May 03 17:36:31 <ChristW> (is there a 'revert' command in Git? One of the files that are changed doesn't need any changes...)
May 03 17:37:14 <morphis> mickeyl: why was it needed for shiboken-native?
May 03 17:37:25 <morphis> even because of generatorrunner?
May 03 17:37:29 <mickeyl> i think so
May 03 17:37:33 <morphis> ok
May 03 17:37:34 <mickeyl> shiboken-native calls generatorrunner
May 03 17:37:41 <antrik> ChristW: there is a "revert" command, but I'm not sure whether that's what you want
May 03 17:37:44 <mickeyl> generatorrunner only calls it for the tests/examples iirc
May 03 17:37:51 <mickeyl> ChristW: git reset / git checkout, if you don't have committed yet
May 03 17:38:02 <morphis> ok
May 03 17:38:10 <ChristW> Yes, some googling suggests 'git checkout' on the file.
May 03 17:38:50 <mickeyl> i also succeeded
May 03 17:38:52 <mickeyl> >>> from PyQt4 import QtCore
May 03 17:38:52 <mickeyl> >>> from PyQt4 import QtGui 
May 03 17:38:53 <mickeyl> >>>
May 03 17:38:58 <morphis> yeah!
May 03 17:39:03 <antrik> ChristW: so you want to drop the changes you made to the file completely? if so, checkout is probably right
May 03 17:39:07 <mickeyl> argh:
May 03 17:39:14 <mickeyl> >>> from PyQt4 import QtDeclarative
May 03 17:39:14 <mickeyl> Traceback (most recent call last):
May 03 17:39:14 <mickeyl>   File "<stdin>", line 1, in <module>
May 03 17:39:14 <mickeyl> ImportError: PyQt4/QtDeclarative.so: undefined symbol: _ZTI33QPyDeclarativePropertyValueSource
May 03 17:39:18 <ChristW> antrik: Yes, I put in debugging, but don't need it afterwards.
May 03 17:39:21 <morphis> :)
May 03 17:39:31 <ChristW> mickeyl: _Almost_ there....
May 03 17:39:38 <mickeyl> yeah
May 03 17:40:29 <angelox_123> mickeyl,are you trying  what now? :)
May 03 17:43:17 <mickeyl> angelox_123: i'm updating PyQt in OE. it doesn't build since months
May 03 17:43:27 <mickeyl> but since they changed a lot, i need to catch up 
May 03 17:43:32 <mickeyl> since I'm not using their buildsystem
May 03 17:43:40 <mickeyl> but rather creating the build files myself
May 03 17:43:49 <mickeyl> it's a bit of effort...
May 03 17:44:11 <angelox_123> hmm,thanks :)
May 03 17:44:25 <mickeyl> not that we need it, but someone might 
May 03 17:44:35 <morphis> mickeyl: what are they using for the buildsystem?
May 03 17:44:36 <mrmoku> ChristW: mail it to mok (at) fluxnetz.de please
May 03 17:45:14 <mickeyl> morphis: self-invented build files (.sbf) which create .pro files which is then compiled with qmake :/
May 03 17:45:32 <morphis> sounds ugly
May 03 17:45:35 <mickeyl> indeed
May 03 17:45:42 <morphis> the only way?
May 03 17:45:57 <mrmoku> mickeyl: now the problem is that creating the PEP fails 
May 03 17:46:13 <mickeyl> morphis: for now yes, for next version they will probably get rid of that crazy thing
May 03 17:46:21 <morphis> ah ok
May 03 17:46:23 <mickeyl> mrmoku: hmm, why?
May 03 17:46:39 <ChristW> mrmoku: On its way.
May 03 17:47:05 <morphis> mickeyl: I even thought about the way aurora should look&feel a little bit more
May 03 17:47:10 <mickeyl> mrmoku: are we using it wrong? did you compare the C files?
May 03 17:47:14 <mickeyl> morphis: good
May 03 17:47:19 <mickeyl> morphis: any sketches? :)
May 03 17:47:22 <morphis> mickeyl: do you ever looked at the windows mobile phone 7 system?
May 03 17:47:28 <CIA-10> freesmartphone.org: 03mok 07cornucopia * r01c463649521 10/fsogsmd/src/lib/modem.vala: 
May 03 17:47:28 <CIA-10> freesmartphone.org: fsogsmd: lib/modem: add nokia_isi pdp typename
May 03 17:47:28 <CIA-10> freesmartphone.org: Signed-off-by: Klaus Kurzmann <mok@fluxnetz.de>
May 03 17:47:29 <CIA-10> freesmartphone.org: 03mok 07cornucopia * r10fdfefc212a 10/fsogsmd/src/plugins/pdp_nokia_isi/Makefile.am: 
May 03 17:47:29 <CIA-10> freesmartphone.org: fsogsmd: pdp_nokia_isi give it the ISI_LIBS
May 03 17:47:29 <CIA-10> freesmartphone.org: Signed-off-by: Klaus Kurzmann <mok@fluxnetz.de>
May 03 17:47:46 <morphis> mickeyl: with the information hubs they use?
May 03 17:48:13 <CIA-10> freesmartphone.org: 03mok 07libgisi * r00b2d0b16359 10/gisicomm/gisicomm.vala: 
May 03 17:48:13 <CIA-10> freesmartphone.org: gisicomm: gpdsd: check if creating the pipe worked out
May 03 17:48:13 <CIA-10> freesmartphone.org: Signed-off-by: Klaus Kurzmann <mok@fluxnetz.de>
May 03 17:48:38 <mrmoku> mickeyl: going to do that now... take a close look (again) at what ofono does...
May 03 17:48:42 <mrmoku> ChristW: ok
May 03 17:50:27 <mrmoku> cd->pep = g_isi_pep_create(cd->idx, NULL, NULL); 
May 03 17:50:54 <mrmoku> mickeyl: pep = GIsi.PEP.create( ll.modem, ( pep ) => { yield; }, null ); 
May 03 17:50:57 <mrmoku> is this wrong maybe?
May 03 17:51:38 <mickeyl> i can't parse this without indenting :)
May 03 17:52:06 <mickeyl> what's the callback for?
May 03 17:52:19 <mickeyl> it doesn't yield anything meaningful or does it?
May 03 17:52:24 <mrmoku> it returns the pep (too)
May 03 17:52:28 <mickeyl> do you mean to call yield <somefunc>.callback instead?
May 03 17:52:50 <mickeyl> let me take a look at the code in context
May 03 17:52:51 <mrmoku> don't think so... just want it to be done before doing the next step
May 03 17:52:58 <mickeyl> is it already checkedin?
May 03 17:54:32 <mrmoku> yup
May 03 17:54:36 <mrmoku> all pushed
May 03 17:55:05 <mrmoku> ChristW: ok, you need to change your .git/config
May 03 17:55:19 <mrmoku> url = git+ssh://git@git.shr-project.org/libphone-ui-shr.git
May 03 17:56:45 <mrmoku> mickeyl: does the inline callback make it wait until the callback is called? or does it continue?
May 03 17:57:09 <ChristW> mrmoku: Did that.
May 03 17:57:29 <mickeyl> mrmoku: ok, first thing. you are not allowed to call yield in a non-async function
May 03 17:57:37 <mickeyl> strange that vala did let you do that
May 03 17:58:09 <mickeyl> make activate and deactivate async functions
May 03 17:58:11 <mrmoku> hehe, ok
May 03 17:58:22 <mickeyl> 2nd, you call an anonymous yield
May 03 17:58:36 <mickeyl> before scheduling any callback
May 03 17:58:48 <mickeyl> so instead of 
May 03 17:58:54 <mickeyl>             pep = GIsi.PEP.create( ll.modem, ( pep ) => { yield; }, null );
May 03 17:58:56 <mickeyl> what you actually want is
May 03 17:59:08 <mickeyl>  pep = GIsi.PEP.create( ll.modem, ( pep ) => { activate.callback(); }, null );
May 03 17:59:09 <mickeyl> yield;
May 03 17:59:39 <mrmoku> oh, yes
May 03 17:59:41 <mrmoku> thanks :)
May 03 18:00:12 <morphis> mickeyl: http://amethyst.openembedded.net/~morphis/aurora-01.jpg
May 03 18:00:13 <mickeyl> np, that may already be it
May 03 18:00:20 <morphis> mickeyl: some fast sketch
May 03 18:00:30 <mickeyl> excellent, this is the 'today/main' screen?
May 03 18:00:32 <mrmoku> mickeyl: see... I need your book :-)
May 03 18:00:33 <morphis> I am not a very good drawer
May 03 18:00:38 <mickeyl> mrmoku: hehe...
May 03 18:00:48 <morphis> mickeyl: no, it's the main screen, today and launcher in one thing
May 03 18:00:51 <mickeyl> mrmoku: i'm soon a dad, so i will have more time then to write it *cough*
May 03 18:01:00 <morphis> mickeyl: on the main screen is a list of boxes
May 03 18:01:14 <morphis> each box is a "hub" which provides information about a topic
May 03 18:01:16 <ChristW> (Wrong window to hit ^W to...)
May 03 18:01:18 <mrmoku> mickeyl: hehe, right... until you will be a taxi driver too ;)
May 03 18:01:28 <morphis> if you click the hub the corresponding application will start
May 03 18:01:34 <mickeyl> *nod*
May 03 18:01:43 <morphis> if you click at the first box in the sketch the dialer will start
May 03 18:01:48 <morphis> and so on
May 03 18:01:51 <ChristW> morphis: Looks nice. Have you see Notion Ink Adam's interface?
May 03 18:01:57 <morphis> ChanServ: no
May 03 18:02:04 <morphis> ChristW: no
May 03 18:02:25 <morphis> (autocompletion don't know who I want to ping ...)
May 03 18:02:37 <CIA-10> freesmartphone.org: 03mok 07libgisi * r0e8f5d0eff4c 10/gisicomm/gisicomm.vala: 
May 03 18:02:37 <CIA-10> freesmartphone.org: gisicomm: make gpdsd.activate async and use yield correctly
May 03 18:02:37 <CIA-10> freesmartphone.org: Signed-off-by: Klaus Kurzmann <mok@fluxnetz.de>
May 03 18:02:53 <ChristW> It runs apps in Panels. Home Panel shows 'updates' that can be pressed to run an app in a panel.
May 03 18:03:08 <angelox_123> bbl
May 03 18:03:09 <ChristW> morphis: My autocomplete complains that there are more users...
May 03 18:03:52 <mrmoku> mickeyl: another question... I would like to delete pep and pipe in deactivate... but vala complains
May 03 18:04:02 <morphis> ChristW: I don't like my auto completion as it don't know automatically who I want to select without pressing TAB more than one time :)
May 03 18:04:14 <morphis> mickeyl: do you like the idea?
May 03 18:05:28 <mrmoku> mickeyl: committed to early :P
May 03 18:05:29 <mrmoku> | gisicomm.vala:1669.59-1669.75: error: Access to instance member `GIsiComm.GPDS.activate' denied
May 03 18:05:33 <mrmoku> |             pep = GIsi.PEP.create( ll.modem, ( pep ) => { activate.callback(); }, null );
May 03 18:06:39 <ChristW> mrmoku: Can I now just 'push'?
May 03 18:07:19 <morphis> mickeyl: it's even very keyboard friendly
May 03 18:07:46 <ChristW> I think I mean 'commit', by the way...
May 03 18:07:49 <mickeyl> mrmoku: i don't see the async qualifier
May 03 18:08:20 <mickeyl> ah, now
May 03 18:08:26 <mickeyl> hmm
May 03 18:08:27 <mickeyl> you messed up indenting :)
May 03 18:08:36 <mickeyl> ~lart tabs
May 03 18:08:36 * apt whips tabs with a wet and grimy noodle just because
May 03 18:08:42 <mickeyl> let me fix that
May 03 18:10:06 <mickeyl> hmm, i don't understand the error message
May 03 18:11:05 <mrmoku> mickeyl: am I allowed to add vim tags?
May 03 18:11:17 <mickeyl> sure
May 03 18:11:50 <morphis> we should even think about adding vim tags in all vala sources
May 03 18:11:57 <mrmoku> public async void activate
May 03 18:11:59 <mrmoku> isn't correct?
May 03 18:12:03 <mickeyl> it is
May 03 18:12:04 <mrmoku> morphis: +1
May 03 18:12:07 <morphis> should ease life
May 03 18:12:19 <morphis> and should be able to do it automatically
May 03 18:12:27 <morphis> so any volunteer?
May 03 18:12:28 <morphis> :)
May 03 18:12:49 <mrmoku> hmm... guess that would be me then :P
May 03 18:13:11 <ChristW> So, did I commit anything? :-)
May 03 18:13:26 <mickeyl> mrmoku: hmm, i don't understand that code
May 03 18:13:31 <mrmoku> no?
May 03 18:13:43 <mrmoku> ChristW: at least you did not push
May 03 18:13:43 <mickeyl> PEP.create is returning the same as it gives as parameter in the callback?
May 03 18:14:14 <mrmoku> public static unowned GIsi.PEP create (GIsi.Modem modem, GIsi.PEPCallback cb, void* data);
May 03 18:14:19 <mrmoku> is in the vapi
May 03 18:14:28 <mrmoku> public delegate void PEPCallback (GIsi.PEP pep, void* opaque); 
May 03 18:14:36 <mickeyl> ah
May 03 18:14:49 <mickeyl> vala may be confused about that shadowed parameter
May 03 18:15:12 <mickeyl> no, that's not it
May 03 18:15:19 <mickeyl> ah
May 03 18:15:30 <mrmoku> hmm
May 03 18:15:32 <mickeyl> ok, the vapi is wrong
May 03 18:15:44 <mickeyl> we don't have a target param for the callback
May 03 18:15:47 <mickeyl> so the callback has no object
May 03 18:15:50 <mrmoku> pep_create does not even call the callback
May 03 18:15:50 <mickeyl> so it can't access anything
May 03 18:15:53 <mrmoku> it just stores it
May 03 18:16:20 <mickeyl> let me fix the create definition
May 03 18:16:34 <mrmoku> ok
May 03 18:16:48 <mickeyl> vala uses the data param to carry its instance
May 03 18:17:01 <mickeyl> so you are not allowed to use it for yourself
May 03 18:17:13 <mickeyl> i.e. it needs to be removed from the function signature
May 03 18:17:48 <mrmoku> ok
May 03 18:18:57 <mrmoku> mickeyl: the callback is used for g_io_add_watch
May 03 18:19:18 <mrmoku> so the yield there is wrong anyway
May 03 18:19:50 <morphis> mickeyl: whats your opinion about my idea?
May 03 18:20:21 <mickeyl> morphis: busy with mrmoku
May 03 18:20:23 <mickeyl> give me some time
May 03 18:20:32 <morphis> sorry :)
May 03 18:20:57 <darkh> morphis: what's the current status of the cmake problems?
May 03 18:21:02 <mrmoku> mickeyl: same for pipe
May 03 18:21:10 <morphis> darkh: it's solved
May 03 18:21:17 <morphis> python-pyside builds fine
May 03 18:21:30 <darkh> and the generator thing?
May 03 18:21:30 <morphis> darkh: http://git.freesmartphone.org/?p=openembedded.git;a=shortlog;h=refs/heads/aurora/development
May 03 18:21:33 <mickeyl> mrmoku: hmm, not quite
May 03 18:21:38 <mickeyl> i can't see the data param in Pipe.create
May 03 18:21:44 <morphis> darkh: I fixed it with export LD_...
May 03 18:21:51 <mrmoku> yeah, I mean the callback is stored and used later on
May 03 18:21:59 <mrmoku> so I shouldn't yield for it
May 03 18:22:12 <mickeyl> ah ok
May 03 18:22:14 <mrmoku> have to add a pep and a pipe handler function
May 03 18:22:20 <mickeyl> ah ok
May 03 18:22:25 <mickeyl> but they're not necessary to continue?
May 03 18:22:31 <mickeyl> (i.e. to be called)
May 03 18:22:34 <mrmoku> no, don't think so
May 03 18:22:38 <mrmoku> trying without 
May 03 18:22:40 <mickeyl> ok, in that case we don't have to suspend
May 03 18:22:58 <mickeyl> it's ok that pipe.create is using a static callback without data* argument then
May 03 18:23:34 * mrmoku passes mickeyl to morphis and tries :-)
May 03 18:23:48 <mickeyl> ok, it builds now
May 03 18:23:52 <mickeyl> committing
May 03 18:23:58 <morphis> org.freesmartphone.Usage.RequestResource("mickeyl")
May 03 18:24:33 <CIA-10> freesmartphone.org: 03mickey 07libgisi * rbdbdbac74771 10/data/libgisi.vapi: libgisi: fix definition of Pipe.create
May 03 18:24:34 <CIA-10> freesmartphone.org: 03mickey 07libgisi * r1cccef402226 10/gisicomm/gisicomm.vala: gisicomm: PDP.activate: no need to wait for the callback to be called
May 03 18:24:43 <mickeyl> that might be it
May 03 18:24:49 <mickeyl> what was your problem for deactivating?
May 03 18:24:54 <mrmoku> ahh, yeah
May 03 18:25:03 <mrmoku> delete does not work on GIsi.PEP and Pipe
May 03 18:25:19 <mrmoku> ofono calls g_isi_pep_destroy
May 03 18:25:37 <mrmoku> which is the free_function
May 03 18:25:47 <mickeyl> ok
May 03 18:26:03 <mrmoku> but when I tried that vala disliked me
May 03 18:26:27 <mickeyl> ok, let me fix that
May 03 18:26:35 <mickeyl> your basic problem is that you did create them as unowned
May 03 18:26:38 <mickeyl> they shouldn't be
May 03 18:26:41 <mickeyl> we want them to be owned
May 03 18:26:48 <mickeyl> then, to invoke the destroy
May 03 18:26:51 <mickeyl> we just set them to null
May 03 18:26:56 <mickeyl> on it..
May 03 18:27:11 <mrmoku> oh, ok
May 03 18:27:31 <mrmoku> a difficult thing to do for a C coder :P
May 03 18:28:03 <mickeyl> do you want deactivate to be without a callback?
May 03 18:28:14 <mickeyl> (to mediator)
May 03 18:28:18 <mrmoku> did not yet think about it
May 03 18:28:26 <mickeyl> ok, i'll leave the cb outcommented then
May 03 18:28:27 <mrmoku> wanted to take care when activation works
May 03 18:29:34 <CIA-10> freesmartphone.org: 03mickey 07libgisi * rfd41cd5eb912 10/ (data/libgisi.vapi gisicomm/gisicomm.vala): gisicomm: fix PDP.deactivate
May 03 18:29:35 <mickeyl> something like this?
May 03 18:29:40 * mrmoku pulls
May 03 18:30:31 <mickeyl> morphis: sounds nice
May 03 18:30:34 <mrmoku> mickeyl: yeah, looks good
May 03 18:31:21 <morphis> mickeyl: ok, so you like the idea; think about the same only using it with a keyboard and arrows keys
May 03 18:32:47 <mickeyl> yes, the joystick can be used to move a cursor on the non-ts devices
May 03 18:32:58 <mickeyl> and the box will be highlighted to indicate the current cursor
May 03 18:33:04 <morphis> jepp
May 03 18:33:13 <mickeyl> cool
May 03 18:33:23 <morphis> so we can use mostly the same UI for both types of devices
May 03 18:33:44 <morphis> ... and need some "back button" some where
May 03 18:34:05 <morphis> but I will start implementing something for aurora the rest of this week
May 03 18:34:11 <mickeyl> yeah, of course if we add _many_ blocks to the front page this gets a bit annoying for the keyboard devices, but might as well use pages then where we "swipe" right and left
May 03 18:34:28 <mickeyl> so n pages a 4 or 5 blocks
May 03 18:34:31 <morphis> jepp
May 03 18:34:46 <mickeyl> pretty cool, can't wait to join in
May 03 18:34:50 <morphis> but I don't think we will get so many blocks in the near future
May 03 18:34:59 <mickeyl> right, keep it simple and clean
May 03 18:35:05 <morphis> :)
May 03 18:35:10 <mickeyl> i want a weather report though ;)
May 03 18:35:29 <mickeyl> and a stock ticker
May 03 18:35:31 <mickeyl> and and and
May 03 18:35:31 <morphis> me too
May 03 18:35:32 <mickeyl> :D
May 03 18:35:49 <morphis> ideas we have enough I think
May 03 18:35:52 <mickeyl> for sure
May 03 18:35:55 <morphis> just missing the time to implement everything
May 03 18:36:17 <morphis> so I have to leave, sport is waiting
May 03 18:36:18 <morphis> bye
May 03 18:36:25 <mickeyl> ok, have fun, good work!
May 03 18:36:29 <morphis|sport> thx :)
May 03 18:36:46 * mickeyl similarly excited as when he saw 'zed' the first time
May 03 18:37:01 <mickeyl> (pyneo thingy)
May 03 18:40:15 <ChristW> mrmoku: I 'only' did a commit for the changes I made. I presume those are made locally? And that I now need to 'push' to get them in the main repo?
May 03 18:41:19 <mickeyl> correct. in contrast to centralized SCMs, git is a decentralized, which means you are in posession of a while repository, not just a working copy. so commit does commit only to your local repository. if you want to push to a remote one, you need git push
May 03 18:41:28 <mickeyl> s/while/whole/
May 03 18:41:28 <apt> mickeyl meant: correct. in contrast to centralized SCMs, git is a decentralized, which means you are in posession of a whole repository, not just a working copy. so commit does commit only to your local repository. if you want to push to a remote one, you need git push
May 03 18:41:38 <mickeyl> bbl
May 03 18:44:31 <mickey|tv> mrmoku: btw., i'm really appreciating your work on that. it is always much more motivating if you're not a single-player. so, thanks!
May 03 18:45:14 <ChristW> mrmoku: I get an error when I push: http://pastebin.com/B1nEw5ZN
May 03 18:48:11 <mrmoku> mickey|tv: yeah, I can figure that... and very much welcome :)
May 03 18:48:55 <mrmoku> ChristW: hmm
May 03 18:49:47 <ChristW> I guess you added my public key to the server?
May 03 18:49:53 <mrmoku> yup
May 03 18:50:00 <mrmoku> you're pushing to libphone-ui-shr, right?
May 03 18:50:24 <mrmoku> can you show me the uri of the remote in .git/config ?
May 03 18:51:25 <ChristW> http://pastebin.com/Rkpdu2cX
May 03 18:52:03 <mrmoku> looks good
May 03 18:52:17 <ChristW> ssh -v tells me lots of stuff as well...
May 03 18:53:03 <ChristW> http://pastebin.com/3ubiSEP7
May 03 18:54:26 <ChristW> Wait... where did that key come from? It's not in that dir! (/OE/.ssh)
May 03 18:54:42 <mrmoku> :P
May 03 18:54:54 <ChristW> I generated an SSH key for my host system, not for the chrooted env.!
May 03 18:55:07 <ChristW> Let's do that again ;-)
May 03 18:55:59 <ChristW> shame on me...
May 03 18:56:56 <ChristW> Anyway... on its way!
May 03 18:57:41 <mrmoku> ChristW: I mount the sources into the chroot env
May 03 18:57:53 <pespin_> hi there 
May 03 18:57:53 <mrmoku> and push from outside
May 03 18:57:56 <mrmoku> hi pespin_ 
May 03 18:58:37 <ChristW> Since I only use this virtual host for development, there's not much of a difference between pushing frominside the chroot and outside of it...
May 03 18:58:42 <pespin_> I'd like to say that I don't think I will have time to write more blog entries during next weeks
May 03 18:58:51 <ChristW> Except for the key, naturally :-D
May 03 18:58:58 <mrmoku> ChristW: for me there is one big difference... as I love git gui
May 03 18:59:01 <mrmoku> and that needs X :)
May 03 18:59:05 <pespin_> so if someone wants to write some, feel free to do it :)
May 03 18:59:19 <mrmoku> pespin_: hopefully I have something to write soon :)
May 03 18:59:28 <pespin_> mrmoku, nice, thanks ;)
May 03 18:59:51 <ChristW> I'll retry that push in a few minutes, then... bbl!
May 03 19:00:52 <pespin_> I'm haveing lots of exams and several projects at univ lately u.u
May 03 19:03:36 <ChristW> pespin_ Good luck!
May 03 19:04:58 <GNUtoo|laptop> hi morphis_ 
May 03 19:05:08 <GNUtoo|laptop> you made a library in fsoaudiod/src/lib
May 03 19:05:22 <GNUtoo|laptop> I'm specially interested in the alsa.vala
May 03 19:05:30 <GNUtoo|laptop> specifically the function named:
May 03 19:05:44 <GNUtoo|laptop>     public static SoundDevice create( string cardname = "default" ) throws SoundError
May 03 19:05:54 <GNUtoo|laptop> what should I do?
May 03 19:06:08 <GNUtoo|laptop> I bet copy-pasting the function in my code is not the solution
May 03 19:06:10 <mrmoku> pespin_: I thought you're all busy with el classico :P
May 03 19:06:25 <GNUtoo|laptop> also I bet I can't use only that lib since I didn't find the writei function
May 03 19:06:36 <pespin_> mrmoku, yeah, that too hahah I'm watching it with the netbook in the sofa
May 03 19:06:40 <pespin_> :P
May 03 19:06:47 <mrmoku> :)
May 03 19:06:52 <mrmoku> hope you have enough crisps
May 03 19:07:02 <mrmoku> guess it will be quite amusing :)
May 03 19:08:02 <pespin_> yeah it is :) We'll shut up mourinho forever hehe
May 03 19:08:02 <mrmoku> ChristW: so I exchange the key?
May 03 19:08:12 <mrmoku> hehe... he deserves it ;)
May 03 19:09:42 <ChristW> mrmoku: Probably, yes...
May 03 19:09:59 <ChristW> The key I sent you the second time is the key from the chrooted host env.
May 03 19:10:07 <mrmoku> ok, moment
May 03 19:10:21 <GNUtoo|laptop> does someone knows in how long morphis will come in front of his screen?
May 03 19:10:30 <GNUtoo|laptop> did he said for how long he was away?
May 03 19:11:09 <mrmoku> ChristW: hmm... it does not differ
May 03 19:11:39 <ChristW> Weird...
May 03 19:11:58 <GNUtoo|laptop> is he doing sport?
May 03 19:12:23 <GNUtoo|laptop> in other words should I do something else like some recipes or some buglabs stuff
May 03 19:12:23 <GNUtoo|laptop> ?
May 03 19:12:29 <mrmoku> 20:35 < morphis> so I have to leave, sport is waiting
May 03 19:12:36 <GNUtoo|laptop> ah ok
May 03 19:12:45 <GNUtoo|laptop> so I'll do the alsa thing another day
May 03 19:12:52 <mrmoku> so no idea... maybe he's not doing sports himself but watching el classico too :P
May 03 19:13:03 <GNUtoo|laptop> what's el classico?
May 03 19:13:08 <antrik> ChristW: you can use the same key for several systems BTW
May 03 19:13:13 <mrmoku> FC Barcelona - Real Madrid
May 03 19:13:33 <mrmoku> GNUtoo|laptop: the hottest soccer possible in spain
May 03 19:13:34 <GNUtoo|laptop> ah ok
May 03 19:14:22 <GNUtoo|laptop> strange
May 03 19:14:26 <GNUtoo|laptop> mickey|tv, wrote this file
May 03 19:14:30 <GNUtoo|laptop> not morphis_ 
May 03 19:15:07 <mrmoku> ChristW: yeah, antrik is right... just copy the key to your chroot
May 03 19:20:45 <mrmoku> GNUtoo|laptop: we should document that damn u-boot thing
May 03 19:21:03 <mrmoku> GNUtoo|laptop: if it is creating problems even for mickey|tv ...
May 03 19:21:08 <mrmoku> that's bad
May 03 19:21:13 <GNUtoo|laptop> mrmoku, what uboot thing?
May 03 19:21:21 <GNUtoo|laptop> the boot image?
May 03 19:21:26 <GNUtoo|laptop> boot.scr
May 03 19:21:27 <GNUtoo|laptop> rather
May 03 19:21:37 <mrmoku> document to install u-boot to have a working dual boot
May 03 19:21:44 <mrmoku> or at least a working shr boot :P
May 03 19:22:07 <mrmoku> from flashing to booting
May 03 19:22:37 <mrmoku> mickey|tv: still... 1970-01-01T01:35:46.054596Z [WARN]  fsogsmd : GLib <gisicomm>: gisicomm.vala:1672: failed to create PEP
May 03 19:23:41 * mrmoku digs into ofono
May 03 19:24:02 <GNUtoo|laptop> mrmoku, why not patching the kernel instead?
May 03 19:24:37 <mrmoku> GNUtoo|laptop: to make it small enough?
May 03 19:24:46 <mrmoku> or what do you mean?
May 03 19:24:46 <GNUtoo|laptop> no, to make it boot with uboot 
May 03 19:24:53 <GNUtoo|laptop> /dev/mmcblkNp1
May 03 19:25:00 <GNUtoo|laptop> the root=
May 03 19:25:02 <GNUtoo|laptop> patch to make that work
May 03 19:25:06 <mrmoku> hmm... the thing is... it works fine for me
May 03 19:25:29 <GNUtoo|laptop> 1) what uboot did you flash
May 03 19:25:31 <mrmoku> maybe it's just a different env
May 03 19:25:35 <GNUtoo|laptop> 2)does it work fine for mickeyl?
May 03 19:25:35 <mrmoku> own built
May 03 19:25:39 <GNUtoo|laptop> ok
May 03 19:25:50 <GNUtoo|laptop> what's the setup?
May 03 19:25:54 <mrmoku> the problem is I don't remember which one of my builds I did flash
May 03 19:25:58 <GNUtoo|laptop> could you share it with mickeyl
May 03 19:25:59 <mrmoku> let me see ...
May 03 19:26:00 <GNUtoo|laptop> ah ok
May 03 19:26:11 <GNUtoo|laptop> then download it, md5sum it
May 03 19:27:39 <mrmoku> GNUtoo|laptop: heh, git diff in my u-boot tree tells it all
May 03 19:27:49 <mrmoku> http://paste.pocoo.org/show/382395/
May 03 19:28:33 <GNUtoo|laptop> ok
May 03 19:28:36 <GNUtoo|laptop> what do we do then?
May 03 19:28:39 <GNUtoo|laptop> between:
May 03 19:28:42 <GNUtoo|laptop> *patching uboot
May 03 19:28:44 <mrmoku> dunno if the diff is against what we currently build
May 03 19:28:46 <GNUtoo|laptop> *patching the kernel
May 03 19:28:53 <mrmoku> but probably the env is set to mmcblk1p1
May 03 19:29:29 <mrmoku> JaMa: why do you have mmcblk1p1 in the u-boot env?
May 03 19:32:10 <GNUtoo|laptop> personally in the .ok bin yes
May 03 19:32:25 <GNUtoo|laptop> I think the best is to change the kenrel
May 03 19:32:34 <GNUtoo|laptop> because 
May 03 19:32:43 <GNUtoo|laptop> the mmc is removable
May 03 19:32:46 <GNUtoo|laptop> the OneNAND is not
May 03 19:32:52 <GNUtoo|laptop> so if OneNAND is 0
May 03 19:32:55 <GNUtoo|laptop> and mmc is 1
May 03 19:33:03 <GNUtoo|laptop> you can refer to something that doesn't move
May 03 19:33:07 <GNUtoo|laptop> which is better
May 03 19:33:32 <mrmoku> could you try to do that?
May 03 19:34:42 <GNUtoo|laptop> mrmoku, look in #oe
May 03 19:39:28 <ChristW> mrmoku: I had an insight... see your mail.
May 03 19:52:48 <morphis|sport> GNUtoo|laptop: ping
May 03 19:53:25 <GNUtoo|laptop> morphis, hi it was for alsa
May 03 19:53:34 <morphis> I read your lines
May 03 19:53:38 <GNUtoo|laptop> ok
May 03 19:53:38 <morphis> you need the function too?
May 03 19:53:56 <GNUtoo|laptop> yes I need the following function:
May 03 19:54:12 <GNUtoo|laptop> <GNUtoo|laptop>     public static SoundDevice create( string cardname = "default" ) throws SoundError
May 03 19:55:19 <morphis> for what?
May 03 19:55:29 <GNUtoo|laptop> for opening the card 
May 03 19:55:46 <morphis> GNUtoo|laptop: then you need the whole SoundDevice class right?
May 03 19:55:46 <GNUtoo|laptop> like
May 03 19:55:52 <GNUtoo|laptop> I think so
May 03 19:55:54 <morphis> ok
May 03 19:55:56 <GNUtoo|laptop> snd_pcm_open
May 03 19:56:00 <morphis> in which component?
May 03 19:56:02 <GNUtoo|laptop> snd_pcm_hw_params_malloc
May 03 19:56:10 <GNUtoo|laptop> etc...
May 03 19:56:21 <GNUtoo|laptop> and snd_pcm_writei
May 03 19:56:24 <GNUtoo|laptop> at the end
May 03 19:56:28 <morphis> the SoundDevice does not deal with the pcm device
May 03 19:56:29 <GNUtoo|laptop> and close of course
May 03 19:56:32 <morphis> it uses the card
May 03 19:56:33 <GNUtoo|laptop> ah?
May 03 19:56:38 <GNUtoo|laptop> ah ok
May 03 19:56:42 <GNUtoo|laptop> so I do it on my own
May 03 19:56:45 <GNUtoo|laptop> ?
May 03 19:56:53 <GNUtoo|laptop> or should I re-use some of your stuff?
May 03 19:57:08 <morphis> where do you need the pcm stuff?
May 03 19:57:28 <GNUtoo|laptop> what do you mean by where?
May 03 19:58:00 <GNUtoo|laptop> If I use the only vapis, is it ok?
May 03 19:59:50 <morphis> GNUtoo|laptop: btw. I got a response from palm: "We are currently in the process of fulfilling free Pre 2 requests this week.
May 03 19:59:50 <morphis> "
May 03 20:00:06 <GNUtoo|laptop> ok nice
May 03 20:00:20 <GNUtoo|laptop> do you have more details?
May 03 20:00:37 <GNUtoo|laptop> because I go in france starting from thursday
May 03 20:00:45 <GNUtoo|laptop> then come back on sunday
May 03 20:00:57 <GNUtoo|laptop> I hope I won't miss the pre 
May 03 20:02:01 <morphis> no, no details
May 03 20:02:33 <GNUtoo|laptop> ok
May 03 20:02:36 <GNUtoo|laptop> thanks a lot for the infos
May 03 20:02:39 <morphis> but there will be a later with duty information before you receive the device
May 03 20:02:47 <GNUtoo|laptop> ok
May 03 20:02:55 <morphis> GNUtoo|laptop: I mean, for what purpose do you need the alsa stuff?
May 03 20:03:00 <GNUtoo|laptop> ah ok
May 03 20:03:05 <GNUtoo|laptop> I'll explain
May 03 20:03:20 <GNUtoo|laptop> under the n900, as you know the modem speak a network protocol
May 03 20:03:31 <GNUtoo|laptop> the audio is also a protocol, not something wired
May 03 20:03:41 <GNUtoo|laptop> so we use a library called cmtspeech to extract the audio data
May 03 20:04:07 <GNUtoo|laptop> so far I wrote it on the disk and played with aplay as me and Doc tought/wanted to do
May 03 20:04:13 <GNUtoo|laptop> and it works
May 03 20:04:20 <GNUtoo|laptop> so in order to have a working telephony
May 03 20:04:28 <GNUtoo|laptop> I want alsa instead of aplay
May 03 20:07:24 <morphis> ah ok
May 03 20:07:58 <morphis> then bind the necessary function in a alsa-ext.vapi 
May 03 20:08:12 <GNUtoo|laptop> ok I started to do that
May 03 20:08:15 <morphis> ok
May 03 20:08:27 <morphis> when you are finished, please send a patch to vala upstream
May 03 20:08:28 <GNUtoo|laptop> (using alsa-ext.vapi copied from your fsoaudiod)
May 03 20:08:43 <GNUtoo|laptop> ah bind
May 03 20:08:46 <GNUtoo|laptop> I read badly
May 03 20:08:56 <GNUtoo|laptop> I think the necessary function were already there
May 03 20:08:59 <GNUtoo|laptop> if not I'll bind
May 03 20:09:31 <mrmoku> ChristW: ok, you can retry it
May 03 20:10:39 <morphis> GNUtoo|laptop: ok
May 03 20:10:55 <morphis> GNUtoo|laptop: if I have some more time the next days I can help you
May 03 20:11:56 <GNUtoo|laptop> ok
May 03 20:14:14 <GNUtoo|laptop> mrmoku, who does kexecboot,me or you?
May 03 20:14:18 <GNUtoo|laptop> or JaMa ?
May 03 20:15:21 <mrmoku> GNUtoo|laptop: I'm busy with libgisi
May 03 20:15:25 <GNUtoo|laptop> ok
May 03 20:15:29 <GNUtoo|laptop> then it's me or JaMa 
May 03 20:15:32 <GNUtoo|laptop> I'll do it then
May 03 20:15:35 <mrmoku> and not quite a sector where i'M expertised
May 03 20:15:39 <GNUtoo|laptop> and try to continue alsa another day
May 03 20:15:47 <GNUtoo|laptop> it's just a kernel config I bet
May 03 20:16:01 <mrmoku> probably alsa is even more important
May 03 20:16:10 <mrmoku> so better continue with that
May 03 20:16:12 <GNUtoo|laptop> but we need mickeyl with us
May 03 20:16:14 <GNUtoo|laptop> ok
May 03 20:16:24 <GNUtoo|laptop> then alsa
May 03 20:16:52 <mrmoku> yeah, I will send mickey|tv my combined.bin the next time he wants to play with the n900
May 03 20:17:06 <morphis> so I am off
May 03 20:17:07 <morphis> gn8
May 03 20:17:11 <mrmoku> night morphis 
May 03 20:17:16 <darkh> n8
May 03 20:41:46 <ChristW> mrmoku: You still there?
May 03 20:42:46 <mrmoku> yup
May 03 20:42:50 <mrmoku> not for long though
May 03 20:43:00 <mrmoku> ChristW: did you retry?
May 03 20:43:27 <ChristW> This time I sent you the correct SSH key, I think. I just 'push'ed, but I keep getting Permission denied errors.
May 03 20:43:47 <mrmoku> hmm
May 03 20:44:06 <ChristW> ssh -v _does_ connect, though...
May 03 20:44:06 * mrmoku checks if the correct key made it to the server
May 03 20:44:52 <ChristW> It asked for my passphrase, and did lots more than before.
May 03 20:45:09 <ChristW> 'git push' asks for no passphrase, and 'just' fails.
May 03 20:45:46 <mrmoku> does your key have a passphrase?
May 03 20:45:55 <ChristW> Yes. Shouldn't it?
May 03 20:45:59 <mrmoku> sure it should
May 03 20:46:02 <mrmoku> but then it should aks
May 03 20:46:04 <mrmoku> ask
May 03 20:46:19 <mrmoku> you're pushing as bitbake user?
May 03 20:46:25 <ChristW> I should just do 'git push', and git does the rest, right?
May 03 20:46:32 <ChristW> Yes, I'm bitbake...
May 03 20:46:35 <mrmoku> hmm... depends on the config
May 03 20:46:40 <mrmoku> git push origin master:master
May 03 20:47:27 <ChristW> Permission denied.
May 03 20:48:05 <mrmoku> if it doesn't ask you for the passphrase it looks like ssh is not configured correctly
May 03 20:48:35 <ChristW> Hum. Will look into that tomorrow then, time for ZZZzzz-mode... tomorrow I'm up at 06:00..
May 03 20:48:37 <mrmoku> your key is in /OE/ssh/id_rsa[.pub]
May 03 20:48:44 <ChristW> Yes, it is.
May 03 20:48:45 <mrmoku> yeah... zzz mode for me too :)
May 03 20:48:56 <mrmoku> we'll get it working tomorrow then
May 03 20:49:03 <mrmoku> gnight all
May 03 20:49:04 <ChristW> Ok, see you then!
May 03 20:49:07 <ChristW> gnight all!
May 03 21:01:21 <GNUtoo|laptop> now I'm stuck
May 03 21:01:28 <GNUtoo|laptop> there are unbinded stuff
May 03 21:01:34 <GNUtoo|laptop> and I don't know how to bind it
May 03 21:02:21 <GNUtoo|laptop> let me re-look
May 03 21:04:13 <GNUtoo|laptop> snd_pcm_hw_params_* is not mapped
May 03 21:29:56 <GNUtoo|laptop> who still have maemo installed?
May 03 21:30:08 <GNUtoo|laptop> I wonder how much we can compress the zImage with lzma -9
May 03 21:33:47 <antrik> GNUtoo|laptop: I don't think you want that. it slows down the boot noticably on my desktop machine (which is old, but still somewhat faster than most smartphones I suppose...)
May 03 21:34:09 <GNUtoo|laptop> I need it for kexecboot
May 03 21:34:17 <GNUtoo|laptop> I'm asking in #maemo
May 03 21:34:49 <antrik> don't know what kexecboot is... what does it got to do with lzma?
May 03 21:35:05 <GNUtoo|laptop> I have to have:
May 03 21:35:15 <GNUtoo|laptop> *a minimal kenrel capable of kexec-ing
May 03 21:35:25 <GNUtoo|laptop> *an initramfs containing kexecboot
May 03 21:35:29 <GNUtoo|laptop> *maemo kernel 
May 03 21:35:31 <GNUtoo|laptop> under 2M
May 03 21:35:36 <antrik> ouch
May 03 21:35:40 <GNUtoo|laptop> for SHR kernel I can take it from microsd
May 03 21:35:51 <antrik> I doubt that's doable
May 03 21:36:14 <antrik> lzma only reduces the size by some 30% according to the docs
May 03 21:37:54 <GNUtoo|laptop> ok
May 03 22:55:02 <NuttyBunny> Hi everyone, is it normal for SHR to crash when trying to use the virtual keyboard on the first use wizard on a Palm Pre Plus? Can't i just use the physical keyboard? :)
May 03 22:58:15 <NuttyBunny> ah, nevermind, using the full keyboard did the trick... the numeric one crashes
May 03 22:59:42 <NuttyBunny> AAAAHH!!  If it hadn't crashed I've read the firmware notice! cool :)
May 03 23:01:05 <NuttyBunny> mmmm... no drivers for Palm Pre... any advice?
May 03 23:06:38 <NuttyBunny> 80 people and everyone asleep? :(
May 03 23:07:15 <misc> I think they sleep :)
May 03 23:07:25 <NuttyBunny> Ok, I'll return later with more questions  
May 03 23:07:29 <NuttyBunny> Thanks misc :)
May 03 23:10:08 <misc> usually, they are in europe so, see with this TZ 
May 03 23:14:05 <NuttyBunny> argh... ok, I'll try... in the night ;)  (I'm at -6)
May 04 00:18:36 <NuttyBunny> well... hello again, is there  a way to get the firmware files off my Pre Plus to use them on SHR? :)
May 04 00:25:34 <angelox_123> palm centro supported?
May 04 00:27:11 <angelox_123> or Palm Tungsten E2
May 04 01:36:58 <NuttyBunny> anybody awake? :)
May 04 02:20:50 <DocScrutinizer> 04:19 here
May 04 02:22:35 <DocScrutinizer> yu'll have to wait another 3h for the first ones in .il and .ru to answer
May 04 02:22:59 <DocScrutinizer> 4..5h for .de
May 04 02:49:57 <NuttyBunny> Thanks... I'll do it later then :)
**** ENDING LOGGING AT Wed May 04 02:59:58 2011