**** BEGIN LOGGING AT Wed Mar 28 02:59:56 2007 Mar 28 03:38:14 when you set a PREFERRED_VERSION of something, just what dictates the versioning? Is it based upon the bb filename? Mar 28 03:39:43 I have created my own bb tree for this project, with a busybox_1.2.1.bb file with an internal "PR" of "r1.5", but it bitbake won't accept PREFERRED_VERSION_busybox="1.2.1-r1.5" Mar 28 03:46:30 you need to change the filename to change the PV which is what is used for PREFERRED_VERSION. Mar 28 03:48:51 ok, so it is based upon the name of the bb file then? Mar 28 03:49:39 I have been setting PR = "r9999" to get bitbake to override some of the default selections in the org.openembedded.dev/ tree Mar 28 04:10:58 !oebug 2054 Mar 28 04:10:59 * * Bug 2054, Status: NEW, Created: 2007-03-27 21:13 Mar 28 04:11:00 * * r.farina(AT)adelphia.net: glibc-2.5-r5 Mar 28 04:11:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2054 Mar 28 04:11:08 that bug is annoying Mar 28 04:12:11 anyone take credit for glibc? Mar 28 05:05:45 morning all Mar 28 05:10:11 v8jlene: morning. probably yes, threading is need it, as i assume there are a lot of server like apps in perl, that probably make use of it Mar 28 05:10:43 v8jlene: the one i was testing was ;) Mar 28 05:11:01 Ifaistos: oddly enough there is very very little that uses threading in perl ;) Most servers just use a select/poll style loop and no threading. Mar 28 05:11:23 The threading api is declared as "crap and will change in the future" which tends to stop people using it! Mar 28 05:11:48 v8jlene: i'll push the changes for powerpc today and then you can check in yours Mar 28 05:11:51 * v8jlene curses smart-card pin blocking "your smart-card is blocked" *sigh* Mar 28 05:13:50 lol perl threading has been deprecated for like 10 years Mar 28 05:24:10 v8jlene: btw when i do ipkg remove perl -recursive all modules get removed except libperl5 Mar 28 05:31:04 Ifaistos: Sounds like ipkg is trying to removing things that start with perl and libperl doesn't? Mar 28 05:33:24 v8jlene: i don't believe its a name thing. there is like a gazilion depends and rdepends in perl.inc maybe libperl is not there. haven't looked at it in detail Mar 28 05:34:22 v8jlene: i am just noting things that look strange as i use the package and hopefully find time to fix them Mar 28 05:34:33 Ifaistos: libperl is probably being picked up through the shlibs stuff, maybe whatever recursive is doing doesn't work with that? Mar 28 05:39:34 v8jlene: hmmm need to look into that Mar 28 05:41:36 v8jlene: can you make a fresh build for boa and see if it runs on your platform ? Mar 28 05:42:25 v8jlene: for me recent builds don't run on two platforms ppc405 and ppc603. used to work on on both Mar 28 05:45:36 Ifaistos: Boa's a web server isn't it? (I use cherokee) I'll try it out. Mar 28 05:45:56 v8jlene: yes Mar 28 05:50:22 v8jlene: do you know what;s the cherokee footprint in flash ? Mar 28 05:54:51 Ifaistos: 1588.8k sh4, 1454.4k i486 Mar 28 05:55:38 Ifaistos: but... it needs libpcre, gnutls, libgpg-error and libgcrypt Mar 28 06:09:14 Ifaistos: boa just seems to pint "gethostbyname:: Success" and then dies... no errors, no logs, nothing... Mar 28 06:09:25 pint? try "print" Mar 28 06:10:05 v8jlene: same here Mar 28 06:11:23 v8jlene: i noticed it after the last changes to gcc when i rebuild all packages Mar 28 06:11:58 * Ifaistos puts another note on the loooooooooong todo list Mar 28 06:21:33 Ifaistos: Appears to be dieing in the if (!server_name) { section of read_config_files. If you define ServerName in the config file it'll work because it'll make it skip that code... Mar 28 06:22:22 v8jlene: hmmm that means it has resolv issues Mar 28 06:28:26 Ifaistos: It also works if `hostname` has an entry in /etc/hosts, only fails if it has to go out to the dns server. Mar 28 06:30:42 Ifaistos: Ah, in my case hostname is "titan" and titan isn't in the dns, so boa fails because it can't find what it thinks is a valid name for the system. Mar 28 06:31:11 Ifaistos: If I "hostname np51r", to change the name to np51r, which is what it's called in my DNS, then boa also works. Mar 28 06:32:00 v8jlene: so its confirmed that its a resolve issue Mar 28 06:32:03 Ifaistos: So the message "gethostbyname:: Success" is really saying that the gethostbyname worked (success), but it didn't return anything (ie, the name wasn't in the DNS or hosts file) Mar 28 06:33:25 v8jlene: you habd gdb on your board ? Mar 28 06:33:32 Ifaistos: It should really be logging a message "I can't find your hostname" in that situation... it's not very useful as is.. Mar 28 06:34:13 Ifaistos: no, gdb doesn't build for sh and I haven't had the energy to look at it. But strace does work. Mar 28 06:35:26 v8jlene: normally it does not need to define it. i have used it with uClinux system with limited flash and works fine. Also the last package i had in the repo worked but when i change it it start showing this behaviour Mar 28 06:36:07 Ifaistos: I don't thinkn boa hast changed, probably /etc/hosts or the default hostname has changed somehow. Mar 28 06:38:18 is anyone using psplash ~r240 ? I'm getting a black screen with it no matter what image format I pass in via the header. Mar 28 06:53:11 v8jlene: thanks for looking at it Mar 28 07:11:22 hmm Mar 28 07:11:44 seems a few people get the same framebuffer black screen with usplash. Mar 28 07:13:04 good morning all Mar 28 07:17:22 morning Mar 28 07:18:04 surely psplash doesn't need vesafb-tng? I can't see the vesafb patch applied to any kernels in org.oe.dev/packages/linux Mar 28 07:18:46 however psplash is insistent on just a black screen. Only thing I can think of is that it needs to be called with something other than psplash & psplash-write progress 1 Mar 28 07:28:24 I see you're using psplash on a moko koen. Do you issue any commands to psplash other than starting it and writing progress updates via "psplash-write "PROGRESS 1"" and "psplash-write "QUIT""? Mar 28 07:29:17 jselect: I just use the initscript Mar 28 07:34:38 koen: Only reason I'm not using the init script is because I've got to copy from an ro mounted device to tmpfs first and I'd like that covered by psplash Mar 28 07:35:00 koen: thanks anyway Mar 28 07:51:25 ok so linux-x86-2.6.17 needs vesafb-tng patches Mar 28 07:51:46 to work with psplash on ati cards with vesa fb Mar 28 08:06:07 morning Mar 28 08:06:31 hey XorA Mar 28 08:07:58 hey koen, XorA Mar 28 08:10:55 Hi Mar 28 08:12:05 I sent yesterday a patch to the mailing list to fix rng package stuff with uclibc distros, is something wrong with it? or does someone taken a look to that? Mar 28 08:12:47 hey pH5 Mar 28 08:13:37 good morning Mar 28 08:16:10 would someone please help me understand an error I have with building? Mar 28 08:17:39 saladino: which ml? Mar 28 08:17:52 zecke oe mailinglist Mar 28 08:17:53 zecke: I'll pick this one up. It's slugos specific. Mar 28 08:17:56 Hello! Mar 28 08:19:09 rwhitby is is it correct? Mar 28 08:19:23 saladino: yes, applying it now Mar 28 08:19:32 thanks rwhitby Mar 28 08:19:39 i wasn't sure Mar 28 08:20:32 somebody know something about xynth? it isn't has a bb in dev branch... Mar 28 08:20:49 can someone tell me how to overcome this: http://pastebin.ca/413266 ? Thansk Mar 28 08:22:12 03koen 07org.oe.dev * r5167dafe... 10/ (3 files in 3 dirs): openmoko-keyboard: add openmoko-keyboard recipe based on matchbox-keyboard, and add it to task-openmoko Mar 28 08:24:03 that's wierd.. the psplash init script doesn't ever call psplash-write Mar 28 08:25:17 03rwhitby 07org.oe.dev * r70ffcc90... 10/ (1 packages/tasks/task-slugos.bb): task-slugos: Made rng-tools conditional on not uclibc Mar 28 08:25:23 03pH5 07org.oe.dev * rd20017ea... 10/ (19 files in 2 dirs): (log message trimmed) Mar 28 08:25:23 X11R7.2 preparation: xorg-apps Mar 28 08:25:23 - add individual release support to xorg-app-common.inc Mar 28 08:25:23 - add new X11R7.2 releases (individual release versions) Mar 28 08:25:23 they have smaller versions than anything where PV begins Mar 28 08:25:26 with "X11R7."... but the epoch is set PE="1". Mar 28 08:25:28 Thus the new packages will only be selected by either Mar 28 08:32:29 pH5: does ipkg support epochs? Mar 28 08:33:18 koen: yes. I'm not sure how extensively this is tested, though. Mar 28 08:33:43 or to be more exact, does bitbake trunk + .dev generate epochs for ipkgs and debs? Mar 28 08:33:52 yes! Mar 28 08:34:14 oh wait, I have to test debs yet. Mar 28 08:38:59 koen@bitbake:~/OE/monotone/org.openembedded.dev/packages$ bitbake -s | grep xau Mar 28 08:38:59 libxau 0:X11R7.1-1.0.1-r0 Mar 28 08:38:59 libxau-native 0:X11R7.1-1.0.1-r0 Mar 28 08:38:59 xauth 1:1.0.2-r0 Mar 28 08:39:29 pH5: NOTE: Task failed: Fetch failed: http://xorg.freedesktop.org/releases/individual/app/bdftopcf-1.0.1.tar.bz2 Mar 28 08:43:46 please would anyone help? Mar 28 08:45:17 koen: thanks, that was a thinko. is it correct that bitbake -s shows the 0: for PE=0 ? Mar 28 08:45:30 djay-il: what is your question? If i know i will help you... Mar 28 08:45:49 pH5: no idea on that Mar 28 08:46:09 I get an error I don't understand: http://pastebin.ca/413266 Mar 28 08:48:46 djay-il: I think you'll have to bother psokolovsky for that Mar 28 08:48:53 ? Mar 28 08:48:56 who's that? Mar 28 08:49:13 the one who wrote the part that your errors happens in Mar 28 08:50:26 I seem to get the same error no matter what I'm trying to build. Its like I have something wrong with environment Mar 28 08:51:04 and btw, which part is it? Mar 28 08:51:59 Actually, that's how OE bugtracker is called, fondly. people submit all relevant info there, and get response for sure. ;-) Mar 28 08:53:09 nice name Mar 28 08:53:18 djay-il: did youu try with prefered-provider glibc 2.3.2? Mar 28 08:53:24 djay-il: you for sure have something wrong with environment. did you apply some external patches not in the tree? did you set any configuration vars without knowning for sure what they do? Mar 28 08:53:41 psokolovsky: no, I just started Mar 28 08:53:46 clean environment Mar 28 08:54:04 djay-il: anyway, don't forget to include your local.conf in report Mar 28 08:54:48 charlyburk: no Mar 28 08:54:55 charlyburk: that Mar 28 08:54:56 koen: do you have an idea how DEPENDS="foo, bar" behaves? Mar 28 08:55:09 charlyburk: that's actually glibc that I'm trying to build Mar 28 08:55:22 psokolovsky: I assumely badly Mar 28 08:55:57 I just wonder how much validation I should add... Mar 28 08:58:18 03pH5 07org.oe.dev * rf0205a7e... 10/ (3 files in 2 dirs): Mar 28 08:58:18 X11R7.2 preparation: xorg-doc Mar 28 08:58:18 - add individual release support to xorg-doc-common.inc Mar 28 08:58:18 - add new X11R7.2 release (individual release version) Mar 28 08:58:18 of xorg-sgml-doctools Mar 28 08:58:23 03pH5 07org.oe.dev * reda0f6d7... 10/ (1 packages/xorg-app/bdftopcf_1.0.1.bb): remove non-existant bdftopcf 1.0.1 Mar 28 08:58:34 morning all Mar 28 08:58:59 morning Mar 28 09:00:34 http://cia.vc/blog/2007/03/welcome-to-ciavc/ Mar 28 09:00:40 * koen fixes the cia script OE uses Mar 28 09:00:44 hey RP & hrw Mar 28 09:02:47 morning folks Mar 28 09:04:39 hey ade|desk Mar 28 09:05:15 RP: Hi! would you recommend to use bitbake 1.8 branch or trunk nowadays? Mar 28 09:06:02 psokolovsky: I appreciate anyone brave enough to test trunk but it does have quirks and is likely to have issues now and again Mar 28 09:06:13 Why isent the images beeing bziped anymore. Opnmoko-image is just a tar ball now Mar 28 09:06:41 RP: i.e. it's in unstable phase now? Mar 28 09:06:54 psokolovsky: Its heading that way and will be getting worse Mar 28 09:07:10 ok, thanks Mar 28 09:07:27 03pH5 07org.oe.dev * r769840c4... 10/ (23 files in 2 dirs): Mar 28 09:07:27 X11R7.2 preparation: xorg-driver Mar 28 09:07:27 - add individual release support to xorg-driver-common.inc Mar 28 09:07:27 - add new X11R7.2 releases (individual release versions) Mar 28 09:08:39 psokolovsky: Its why I made the 1.8 branch/release as I'm expecting a bit of instability ;-) Mar 28 09:08:50 FWIW, bitbake trunk builds stuff fine right now Mar 28 09:09:06 most quirks are in messages to the user Mar 28 09:09:17 or the absence of messages :) Mar 28 09:09:27 koen: Your testing was useful in ironing out some bugs before they get hidden by other ones :) Mar 28 09:09:56 koen: They're there in log files and will make a return when the UI code lands :) Mar 28 09:10:21 I still have some nasty bugs to iron out before I can check it in though :/ Mar 28 09:10:25 I know, I use vi and -D a lot more these days Mar 28 09:13:42 I still wonder if I shouldn't just check in the UI code and then debug it in public :) Mar 28 09:14:16 03pH5 07org.oe.dev * rc93086b3... 10/ (3 files in 2 dirs): Mar 28 09:14:16 X11R7.2 preparation: xorg-font Mar 28 09:14:16 - add individual release support to xorg-font-common.inc Mar 28 09:14:16 - add new X11R7.2 releases (individual release versions) Mar 28 09:14:28 RP: go ahead :) Mar 28 09:15:20 I'll wait until I'm likely to have enough free time to deal with the fallout, maybe this weekend :) Mar 28 09:15:42 RP: you should check it in on the 6th April :-) Mar 28 09:15:57 * koen wonders how many people use bitbake trunk Mar 28 09:15:59 XorA: Why the 6th? Mar 28 09:16:04 RP: easter Mar 28 09:17:08 Ah, I won't be around on the 6th, I have plans for then ;-) Mar 28 09:17:23 RP: biking? Mar 28 09:18:00 eating eggs? Mar 28 09:18:01 XorA: Yes, the roads are usually nice and quiet on the Friday so its the best day to go out Mar 28 09:18:08 RP: cool Mar 28 09:18:23 * XorA must get round to sitting his test this year Mar 28 09:18:51 * RP needs to get the bike back to the dealer and find out why the warning light is on, again :-( Mar 28 09:19:04 RP: jtag it! Mar 28 09:19:09 bitbake warning-light-off Mar 28 09:19:39 is the warning light using the LED class yet? Mar 28 09:19:47 hrw: It has an ODB-2 connection ;-) Mar 28 09:20:03 ~lart person which added nfs-utils Mar 28 09:20:04 * ibot says "boot to the head" and knocks person which added nfs-utils over Mar 28 09:20:06 s/ODB/OBD/ Mar 28 09:20:13 hey ssvb Mar 28 09:20:17 RP: did you try carman yet on your tablet? Mar 28 09:20:31 koen: No, what is it? :) Mar 28 09:20:31 RP: qemux86 can have eth0 automaticly on Mar 28 09:20:46 http://carman.garage.maemo.org/ Mar 28 09:20:57 "Carman is the first OBD-II analyzer for Maemo platform. You are wondering why your car's Check Engine Light is on. Carman will help you answer that." Mar 28 09:20:58 hrw: It doesn't already? Mar 28 09:21:24 hi XorA Mar 28 09:21:31 RP: you have dhcp on tun0? Mar 28 09:21:31 ssvb: hows mplayer Mar 28 09:21:48 192.168.7.2:/ 196752 47752 146952 25% /mnt Mar 28 09:22:06 hrw: we configure it in the qemu scripts Mar 28 09:22:20 RP: which not always work Mar 28 09:22:38 hrw: Why not? Mar 28 09:22:43 pH5: did you check in your autoip cset already? Mar 28 09:22:54 RP: I never got eth0 up with poky-qemu Mar 28 09:23:14 hrw: qemuarm doesn't work for some reason, x86 should though... Mar 28 09:23:31 koen: no, I forgot. will do now. Mar 28 09:23:34 hrw: but if theres a better way of handling it, I'm happy to go with that Mar 28 09:23:50 koen: carman looks interesting. I might have to get an adapter and try it :) Mar 28 09:24:07 XorA: it's ok, the sudden release of N800 messed my plans somewhat, but now everything is getting back to normal :) Mar 28 09:24:08 koen: Well, few words about angstrom-x11-image. Since some time I started towonder why on earth support for appbutton regressed, and why X cursor is not transparent. Yesterday, after hints from pH5, I finally got when that happened - when I switched from gpe-image to angstrom-*! Mar 28 09:24:15 RP: I couldn't locate the obd2 in the company car, otherwise I would have tried it Mar 28 09:24:32 koen: You have an adaptor? Mar 28 09:24:36 ssvb: pxa overley support was written, so it might benifit from you scaler work Mar 28 09:24:39 RP: not yet Mar 28 09:24:40 koen: in other words, angstrom-x11-image lacks "keylaunch", and has sth weird with cursor config Mar 28 09:24:59 psokolovsky: feel free to fix it Mar 28 09:25:01 what in the hell, why are all my a's turining into e's Mar 28 09:25:11 XorE Mar 28 09:25:24 koen: ok, thanks. wanna bug, or ok to commit? Mar 28 09:25:30 koen: tis a valid instruction Mar 28 09:25:34 psokolovsky: commit Mar 28 09:25:38 ok Mar 28 09:25:45 03pH5 07org.oe.dev * r2e8511b5... 10/ (7 files in 3 dirs): avahi: add scripts to start avahi-autoipd from udhcpc Mar 28 09:30:39 should tzdata not RREPLACES = "timezones" as well? Mar 28 09:30:57 and RPROVIDES it Mar 28 09:31:12 and RCONFLICTS? Mar 28 09:31:22 hrw: it already RCONFLICTS Mar 28 09:31:54 due to lack of DESCRIPTION in tzdata.bb its hard to tell Mar 28 09:32:08 how could I conditionally addtask in a bbclass file? I only want this task to run depending on the value of the DISTRO_TYPE variable Mar 28 09:32:34 I can do "bb.data.getVar('DISTRO_TYPE', d, 1)" in the subroutine, and just return... Mar 28 09:32:54 but is there a way to just have it not even execute the tast in the first place? Mar 28 09:49:48 no Mar 28 09:50:01 03pH5 07org.oe.dev * r51e97a80... 10/ (28 files in 2 dirs): Mar 28 09:50:01 X11R7.2 preparation: xorg-lib Mar 28 09:50:01 - add individual release support to xorg-lib-common.inc Mar 28 09:50:01 - add new X11R7.2 releases (individual release versions) Mar 28 09:51:56 koen|away and others who know: does psplash use the same colour indexes as usplash? Mar 28 09:56:08 03pH5 07org.oe.dev * r9bf63558... 10/ (4 files in 2 dirs): Mar 28 09:56:08 X11R7.2 preparation: xorg-proto Mar 28 09:56:08 - add individual release support to xorg-proto-common.inc Mar 28 09:56:08 - add new X11R7.2 releases (individual release versions) Mar 28 10:08:09 03pH5 07org.oe.dev * r6bfa5939... 10/ (4 files in 2 dirs): Mar 28 10:08:09 X11R7.2 preparation: xorg-xserver Mar 28 10:08:09 - add individual release support to xorg-xserver-common.inc Mar 28 10:08:09 - add new xorg-server 1.2.0 release Mar 28 10:10:22 03pH5 07org.oe.dev * r3333888b... 10/ (1 conf/distro/include/preferred-xorg-versions-X11R7.2.inc): X11R7.2 preparation: add preferred-xorg-versions-X11R7.2.inc Mar 28 10:25:45 hi RP Mar 28 10:25:55 hi greentux Mar 28 10:26:20 rp i have a kernel related question. is user mode linux running on zaurus or maemo devices? Mar 28 10:27:05 greentux: I've never tried to be honest Mar 28 10:27:56 rp mmh ok, so perhaps I should give a try. **** BEGIN LOGGING AT Wed Mar 28 10:36:56 2007 Mar 28 10:37:38 is there a task-thingy for redboot/uboot? Mar 28 10:37:53 or, machine feature rather Mar 28 10:38:09 NAiL: Not that I'm aware of although I've lost touch with recent changes Mar 28 10:38:16 ah, ok Mar 28 10:38:48 there are utilities for handling both redboot and uboot from userspace in OE now :) Mar 28 10:39:59 RP: who is "responsible" for it if you aren't? Mar 28 10:40:42 NAiL: bootloader was never part of features Mar 28 10:40:55 why not? Mar 28 10:41:20 why it has to? userspace does not care which bootloader was used... Mar 28 10:42:06 it doesn't have to care, true. userspace doesn't have to care that you have wireless either Mar 28 10:42:25 wireless support is part of userspace Mar 28 10:42:45 both redboot and uboot have userspace support Mar 28 10:43:33 feel free to add uboot/redboot feature for it if need Mar 28 10:43:45 It sounds like its time for those features to be added although they should probably be a combined one i.e. that the distro should agree with adding them too Mar 28 10:44:19 RP: that means that both machine and distro needs to have FEATURES="uboot" eg? Mar 28 10:44:23 yes Mar 28 10:44:37 NAiL: yes Mar 28 10:44:40 ok, I'll see if I understand how that works Mar 28 10:47:26 NAiL: COMBINED_FEATURES Mar 28 10:48:06 rwhitby: yes, I'm seeing COMBINED_FEATURES ;) Mar 28 10:49:12 how does COMBINED_FEATURES work though? is that an aggregate of *_FEATURES? Mar 28 10:49:27 03pH5 07org.oe.dev * r59f595e3... 10/ (4 files in 2 dirs): xcb: update to xcb-proto/libxcb 0.9.3, xcb-util 0.2 Mar 28 10:49:42 or... does COMBINED_FEATURES automagically containg only the stuff that is in all of *_FEATURES? Mar 28 10:52:09 if distro and machine then it is combined Mar 28 10:53:55 ok, thanks Mar 28 11:07:23 hrw: one other thing. What about a *_FEATURES = raid? Mar 28 11:09:21 NAiL: I'd say that was more of a distro thing Mar 28 11:09:50 ok, so a DISTRO_FEATURES = "raid" then? Mar 28 11:10:01 NAiL: I have no obection to a raid tools task which distros can use for that, no Mar 28 11:10:11 ok, cool Mar 28 11:10:25 I'm working on several devices that have more than one internal disk Mar 28 11:10:35 We need to start thinking about splitting up the tasks or fixing debian.bbclass (I'd prefer the latter) :/ Mar 28 11:12:37 ok, I will improve my version then.. Mar 28 11:22:17 03pH5 07org.oe.dev * re60fccd6... 10/ (1 packages/xorg-lib/libx11_1.1.1.bb): libx11-1.1.1: depends on libxcb Mar 28 11:24:59 03pH5 07org.oe.dev * rf1d649ea... 10/ (1 packages/xorg-xserver/xserver-kdrive_1.2.0.bb): xserver-kdrive-1.2.0: depends on libxkbfile Mar 28 11:28:48 ~change 370 eur to pln Mar 28 11:29:03 370.00 Euro (EUR) makes 1,436.16 Zloty (PLN) (from http://www.xe.com/) Mar 28 11:29:48 djay-il: so, I'm waiting for your bugreport. In the meantime, my build of glibc with default (out of tree) configuration is up to building locales w/o issues. Mar 28 11:30:59 pH5: as I see, not so many our xserver-kdrive patches are discarded by R7.2, right? Mar 28 11:31:52 psokolovsky_: right, they all still apply. Mar 28 11:32:33 psokolovsky_: thanks, but I seem to created the whole environment incorrectly, I'm starrting fresh. If I still have problems, I'll post the report :-) Mar 28 11:32:36 pH5: all? I for sure checked that calibrateext patches are at least partially there Mar 28 11:32:44 djay-il: ok Mar 28 11:33:07 pH5: kudos for doing all that, btw ;-) Mar 28 11:33:49 pH5: any plans to send our patches upstream? Mar 28 11:35:02 "our" patches are mostly scrapped from upstream MLs, AFAICTell ;-) Mar 28 11:35:11 kmode's should be sent, of course ;-I Mar 28 11:35:24 hrw: I planned to wait for 1.3 as by then input-hotplug will take over and at least the evdev stuff won't be needed anymore Mar 28 11:36:06 I don't know about the imageon/w100 patches - they should be submitted by their authors when they are ready for upstream. Mar 28 11:38:20 pH5: we really should check that calibrateext matter. I hoped that when issues was raised that they were lost during cvs->git migration, it will be settled. again, it seems not completely ;-\ Mar 28 11:38:59 psokolovsky_: I'm currently cleaning up xserver-kdrive, so beginning with 1.2.0, xcalibrateext will be dropped in favor of calibrateproto Mar 28 11:39:55 pH5: what I'm talking about is some code which should be in xserver body to support ext anyway. if calibrateproto supercedes such need, that's cool Mar 28 11:40:54 psokolovsky_: look at the src uris. it is identical! Mar 28 11:41:02 xcalibrateext is just the old name Mar 28 11:41:57 pH5: ok, that's all notes from my review of 1.2.0 few weeks ago. by now I of course forgot detail ;-\. I'll review it again when you finish, and drop my thoughts then. Mar 28 11:42:33 psokolovsky_: ok Mar 28 11:59:23 what is the difference between the w100 and the imageon versions of xserver-kdrive? Mar 28 12:00:09 pH5: its same Mar 28 12:00:15 ATI W100 Imageon Mar 28 12:00:59 pH5: Ximageon is from sirfred and supports a lot more features than Xw100 Mar 28 12:01:15 pH5: most nicely it supports Xv Mar 28 12:01:29 XorA: so cat I consider w100 superceded and add imageon instead to the main xserver-kdrive for version 1.2.0? Mar 28 12:01:41 pH5: that would be wonderful Mar 28 12:01:44 great Mar 28 12:01:53 pH5: its seperate as sirfred had to make a change to kdrive core Mar 28 12:02:08 pH5: to allow non contiguous memory in the framebuffer Mar 28 12:02:23 pH5: as Imageon chip has two seperate memory areas Mar 28 12:03:29 ah, ok. did he send the patches upstream already? I hope that they don't break Xfbdev :) Mar 28 12:03:30 bbl Mar 28 12:03:52 pH5: 1) no I dont think he did 2) I dont think they do Mar 28 12:08:27 re Mar 28 12:08:45 koen|lab|rat Mar 28 12:08:49 yeah Mar 28 12:09:12 pH5: don't forget to check my complaints that latest Xserver in use by hx4700 caused me hangs ;-( Mar 28 12:09:32 psokolovsky_: please test the new xserver-kdrive-imageon as well Mar 28 12:10:11 will, so I suspect that was exactly it. cannot say for sure tho - testing mostly kernels now, not X's ;-) Mar 28 13:17:53 03pH5 07org.oe.dev * r03994eb4... 10/ (1 packages/xorg-xserver/xserver-kdrive-common.inc): xserver-kdrive: add initial -common.inc to be used by all xserver-kdrive versions Mar 28 13:17:57 03pH5 07org.oe.dev * re3a4dfb9... 10/ (1 packages/xorg-proto/randrproto_1.2.1.bb): randrproto: add 1.2.1 (>= 1.2 is needed for xserver-kdrive_git) Mar 28 13:43:45 morning all Mar 28 13:48:44 morning Mar 28 13:56:12 hi all Mar 28 13:58:36 xgcc: /home/hrw/devel/build/openmoko/tmp/work/x86_64-angstrom-linux-gnueabi/gcc-cross-sdk-4.1.1-r12/gcc-4.1.1/build.x86_64- Mar 28 13:58:39 linux.arm-angstrom-linux-gnueabi/./gcc/specs: Invalid argument Mar 28 13:58:41 someone got it? Mar 28 13:59:26 hrw: not here :-/ Mar 28 14:00:57 hrw: at the same time, it's been 2 or 3 weeks I didn't build for arm ... Mar 28 14:01:38 cyrilRomain: I have same with gcc-cross-sdk for x86 Mar 28 14:04:03 hrw: hm I din't read 'sdk' in gcc-cross-sdk Mar 28 14:04:09 * cyrilRomain puts his glasses on Mar 28 14:10:33 03pH5 07org.oe.dev * rcc6c8cf0... 10/ (1 packages/xorg-proto/damageproto_1.1.0.bb): damageproto: add 1.1.0 (>= 1.1 needed by xserver-kdrive_git) Mar 28 14:10:39 03pH5 07org.oe.dev * r9ccd29be... 10/ (1 packages/xorg-proto/inputproto_1.4.1.bb): inputproto: add 1.4.1 (>= 1.4 needed by xserver-kdrive_git) Mar 28 14:23:20 I just managed to import an OE built toolchain into Scratchbox... Mar 28 14:23:59 florian, could you share a result? Mar 28 14:24:03 hi, all! Mar 28 14:24:10 Florian: cool I have been playing with scratchbox as well but it sucked Mar 28 14:25:20 slapin_nb: sure, as soon as it works... :) Mar 28 14:25:24 http://dominion.kabel.utwente.nl/koen/cms/using-openembedded-toolchains-in-scratchbox Mar 28 14:25:39 gerwinin: the wiki is not really complete Mar 28 14:25:43 koen, thanks! Mar 28 14:27:01 We need to do some things before we can have usable toolchains: Mar 28 14:28:04 you need to build with a modified bitbake.conf and have to remove one of the gcc patches Mar 28 14:28:38 and we need a useful target that builds a good filesystem Mar 28 14:29:40 the one starting with my nickname :) Mar 28 14:29:51 zecke: right :-) Mar 28 14:30:21 that's a real showstopper here ;) Mar 28 14:30:28 florian: I don't remember but it should be only included if we have fail-fast in OVERRIDES?! Mar 28 14:30:58 * florian takes a look Mar 28 14:31:22 zecke: correct Mar 28 14:33:49 ah true... Mar 28 14:34:42 can i override this in local.conf instead of modifying bitbake.conf? Mar 28 14:35:39 florian: I think you can Mar 28 14:36:04 (you need to set a new OVERRIDES line as base.bbclass was not yet parsed) Mar 28 14:36:37 florian: you can change it Mar 28 14:37:04 ok Mar 28 14:37:42 zecke: about parsing, I took a look at lemon (I had never put my hand in parser generator before). Quite interesting :) Mar 28 14:37:59 cyrilRomain: yes, lemon is awesome :) Mar 28 14:38:07 sqlite is awesome and so is lemon :) Mar 28 14:38:17 cyrilRomain: but kudos go to Marc Singer Mar 28 14:38:29 zecke: but I'm wondering how you want to store parsed data in C, is bitbake-ng intended for that ? Mar 28 14:38:40 zecke: yep Marc Singer did a good job Mar 28 14:39:04 zecke: do you know why he stopped working on it ? Mar 28 14:39:28 cyrilRomain: because he was done. He wanted to show that it can be faster, and he did :) Mar 28 14:39:47 cyrilRomain: well there are two plans on how to hook the parser Mar 28 14:40:04 cyrilRomain: write a new data implementation (see data_smart.py) in C Mar 28 14:40:27 cyrilRomain: make the C parser directly invoke the data_implementation without going through python code Mar 28 14:40:50 cyrilRomain: inter mediate, get the AST out of the parse and visit the nodes to create a dictionary Mar 28 14:41:25 intermediate is what you started to do with the python Ast decorator, right ? Mar 28 14:41:26 pH5: configure.ac:32: error: must install xorg-macros 1.1 or later before running autoconf/autogen Mar 28 14:41:26 cyrilRomain: my 'rewrite' to use visitors is 60% done, once completed the python parser and C parser should parse the same files and the result should be compared :) Mar 28 14:41:31 So, can I use OE to build the AVR toolchain? Mar 28 14:41:37 pH5: from xauth Mar 28 14:41:42 cyrilRomain: I changed from decorator to visitor :) Mar 28 14:42:13 zecke: but the '@' in python is still called 'decorator' ;) Mar 28 14:42:31 ah right :) Mar 28 14:42:56 cyrilRomain: yeah, this is the most easy way to create the indirection/jump table :) Mar 28 14:43:01 koen: I probably overlooked that there is a new util-macros package. Mar 28 14:43:49 zecke: yes good job you did. And so many things I started to learned (lemon, AST, flex, python decorator, etc.) Mar 28 14:44:08 util-macros_X11R7.1-1.1.5.bb exists Mar 28 14:44:12 zecke: btw, I was wondering did you considered using re2c instead of flex ? Mar 28 14:44:43 cyrilRomain: sadly the rewrite to visitor is not complete (you should use a rev before I started to change it) Mar 28 14:44:50 I think I am confused as to what OE is used for, I thought it could be used to build toolchains that target a certain processor Mar 28 14:44:57 cyrilRomain: http://rafb.net/p/EpH9F023.html is untested piece of code to diff two dictionaries Mar 28 14:45:06 cyrilRomain: I have not heard about re2c Mar 28 14:45:08 Chocobo: correct Mar 28 14:45:20 pH5: looks like a missing dep Mar 28 14:45:28 pH5: should I put in in xauth.bb or the .inc? Mar 28 14:45:47 hey mickeyl, everything running at honda? Mar 28 14:45:49 mickeyl: :) Mar 28 14:45:49 koen, do you have a link that explains that? I can only find references to building specific packages (eg nano) Mar 28 14:45:51 zecke: it seems that re2c is to flex what lemon is to yacc Mar 28 14:45:59 koen: not yet :/' Mar 28 14:46:09 cyrilRomain: I'm happy with flex (specially as it is reentrant nowadays) Mar 28 14:46:10 koen: I'm having a hard time wrt. concurrent threads in qt Mar 28 14:46:12 hi mickeyl Mar 28 14:46:19 cheers hrw, hi zecke Mar 28 14:46:26 Chocobo: http://openembedded.org Mar 28 14:46:28 koen: maybe we should put util-macros in xorg-app-common, even. Mar 28 14:46:35 hi mickeyl hrw zecke koen Mar 28 14:46:41 cyrilRomain: and flex+lemon uses like 0.2s of time and 39.8s seconds are spend in python :) Mar 28 14:46:42 hey chouimat Mar 28 14:46:43 zecke: yes, as soon as its works. Maybe should we consider re2c later for optimization Mar 28 14:46:56 zecke: woah :) Mar 28 14:47:17 cyrilRomain: chdir to the standalone parser Mar 28 14:47:26 zecke: oh! Mar 28 14:47:28 cyrilRomain: this is all C and check how much time it needs Mar 28 14:48:27 mickeyl: thanks for that last OM commit :) Mar 28 14:48:28 mickeyl: how do I add applets to mb-panel2 at runtime? Mar 28 14:48:32 zecke: I played with standalone, but thought it parsed only one file ;) Mar 28 14:48:38 koen: no idea :D Mar 28 14:48:47 koen: i think the respective API and stuff is missing yet Mar 28 14:48:54 koen: I'm planning to do more work on mb-2 in April Mar 28 14:49:02 mickeyl: and why does OM have so many forks in svn (openmoko-keyboard, openmoko-dates, etc) Mar 28 14:49:03 pH5: sorry that it took so long. Mar 28 14:49:24 koen: i guess because folks changed it that don't have commit access Mar 28 14:49:38 hmm, ok Mar 28 14:49:50 if i were non-busy i had merged another run Mar 28 14:49:55 alas i'm not Mar 28 14:50:01 things will clean up in a couple of days Mar 28 14:50:04 it looks like openmoko-keyboard is just new xml files + images for mb-keyboards Mar 28 14:50:22 03koen 07org.oe.dev * r4b95bb99... 10/ (1 packages/xorg-app/xorg-app-common.inc): xorg-app-common: add dep on util-macros Mar 28 14:50:23 cyrilRomain: the plan is: Finish the visitor rewrite, implement EXPORT_FUNCTION and such and then Mar 28 14:50:29 and s/matchbox/openmoko/g Mar 28 14:50:33 according to there's also code changes Mar 28 14:50:41 cyrilRomain: integrate the parser into OE and let the old parser and new parser parse the files Mar 28 14:50:59 mickeyl: hire me and I'll show them how 'patch' and 'email client' works :) Mar 28 14:50:59 hi Mar 28 14:51:05 cyrilRomain: and then compare the results, as we lack testcases it is the only way to verify it works Mar 28 14:51:15 zecke: I see Mar 28 14:51:16 i want to add a package to oe Mar 28 14:51:19 koen: you would need to move there Mar 28 14:51:35 mickeyl: that can be arranged ;) Mar 28 14:51:45 hehe Mar 28 14:52:17 koen: gonna become the Moko Enforcer? Mar 28 14:52:28 mickeyl: I'll wait for ~P2 and make an appropriate amount of noise then Mar 28 14:52:35 good :) Mar 28 14:52:36 i would like to add asterisk to openembedded packages Mar 28 14:52:39 XorA: trying to de-Sharp FIC Mar 28 14:52:44 can anyone please help me how to add the same Mar 28 14:52:46 suma: a new one? Mar 28 14:52:58 mickeyl: what is procedure to get r/w to openmoko svn? Mar 28 14:52:58 zecke: i want to putit in openmoko Mar 28 14:53:01 suma: already done: http://www.openembedded.org/filebrowser/org.openembedded.dev/packages/asterisk Mar 28 14:53:14 mickeyl: standard 'show us your patches' apply? Mar 28 14:53:21 mostly Mar 28 14:53:33 hrw: "donate leffe to coreteam: Mar 28 14:53:39 s/:/"/ Mar 28 14:53:39 koen: heh heh Mar 28 14:53:59 ~seen mallum Mar 28 14:54:11 koen: thats now so hard - leffe can be bought in Poznan sometimes Mar 28 14:54:15 koen: when i compile open embedded for openmoko, i don't find asterisk there. how can i add them ? Mar 28 14:54:21 * XorA wishes he had a bottle of leffe for every time git annoyed him Mar 28 14:54:25 mallum was last seen on IRC in channel #edev, 25d 23h 29m 25s ago, saying: ':)'. Mar 28 14:54:31 compile openembedded? Mar 28 14:54:49 koen: i installed the toolchain, it automatically installed openembedded Mar 28 14:55:00 * koen eats some cookies to raise his blood sugar level Mar 28 14:55:30 * XorA looks at suma's .sg suffix and guesses there is probably a translation propblem Mar 28 14:56:06 Xora: I installed everything with http://wiki.openmoko.org/wiki/MokoMakefile for the openmoko phone Mar 28 14:56:25 suma: cd build; bitbake asterisk Mar 28 14:56:26 rwhitby: are you still awake? Mar 28 14:57:13 suma: you need to source a file created by MokoMakefile Mar 28 14:57:20 suma: and then follow XorA's advice Mar 28 14:57:35 . ./setup-env (If I remeber right) Mar 28 14:59:20 Xora: great that worked Mar 28 14:59:23 thanks a million Mar 28 15:01:53 Xora: open embedded can be used for any processor or limited to arm at the moment ? Mar 28 15:02:33 suma: any in theory, there is x86, amd64, mips, arm and powerpc so far Mar 28 15:02:37 suma: and maybe some I missed Mar 28 15:02:47 and sh3/sh4/sparc Mar 28 15:03:00 i c, yes, looking for sh4 processor Mar 28 15:03:25 let me read the wiki and come back here, if there is any doubt Mar 28 15:05:07 03koen 07org.oe.dev * r6707ea89... 10/ (3 files in 2 dirs): openmoko: add qvga theme and mark themes as arch agnostic Mar 28 15:05:19 mickeyl: ^ that one Mar 28 15:13:52 zecke, did you tell me about the movie "The Cuckoo"? Mar 28 15:14:06 no, sadly not Mar 28 15:14:51 hmmm Mar 28 15:14:54 someone did Mar 28 15:15:03 We saw it last night, it is very good Mar 28 15:15:05 Fly over the Cuckoo something? Mar 28 15:15:11 no Mar 28 15:15:29 I've installed a load of -dbg packages on my target machine, and am trying to gdb, but I'm just getting (no debugging symbols found) for everything Mar 28 15:15:53 Crofton|laptop: does 'the cuckoo' is linked with IT? Mar 28 15:16:02 am I right in thinking that gdb should get told where to find these when it gets built? Mar 28 15:16:32 http://www.imdb.com/title/tt0308476/ Mar 28 15:16:53 ah. yet another Mar 28 15:17:43 RP: my bitbake trunk doesn't behave nicely if patches fail. I don't get a shell, it confuses quilt and then some patches are silently ignored while do_patch 'succeeds' Mar 28 15:18:26 hi all Mar 28 15:18:43 mr_nice: hi Mar 28 15:18:54 mrdata: hi Mar 28 15:19:24 mr_nice: my new accu-packs are received Mar 28 15:19:55 mrdata: cool did you already modded the simpad accu? Mar 28 15:20:17 mrdata: are the cells the right ones? Mar 28 15:20:56 mr_nice: 8x Panasonic CGA 103450A, not only disasabled the new accu Mar 28 15:21:12 mr_nice: i would say bingo Mar 28 15:21:35 mrdata: 8x in one accupacks or in two? Mar 28 15:21:59 mr_nice: two accu-packs, 4 cells each Mar 28 15:22:26 mrdata: 20 euro each including shiping? Mar 28 15:23:07 tkp : You need to add DEBUG_BUILD = "1" and INHIBIT_PACKAGE_STRIP = "1" in your local.conf or the package you want to debug Mar 28 15:23:48 pH5: http://rafb.net/p/W3pWF846.html Mar 28 15:23:49 mr_nice: 30,88 euro for two accu-packs incl. shiping Mar 28 15:24:20 mrdata: wow thats a good deal Mar 28 15:24:47 Ifaistos: not anymore Mar 28 15:24:56 Ifaistos: installing the -dbg packages has the same effect Mar 28 15:25:13 koen: thats what I thought Mar 28 15:25:26 mr_nice: i think also, accu-cells replacement is better possible now Mar 28 15:25:51 how can I make kernel-module-tun built ? (the config file already has CONFIG_TUN=y) Mar 28 15:25:54 (it works on my arm machines) Mar 28 15:26:07 mrdata: would be nice if you can make some pictures while modding the packs Mar 28 15:26:50 koen: the greps in avahi-daemon postinst are missing >/dev/null redirection Mar 28 15:27:01 mrdata: are you able to mount your cf card with your image? I think I forgot something to compile into the kernel Mar 28 15:27:05 pH5: ah, so that's harmless Mar 28 15:27:11 pH5: thanks for confirming! Mar 28 15:27:16 mr_nice: okay, but not in the next days, debugging is more important to us Mar 28 15:27:29 koen: and easy to fix :) Mar 28 15:27:33 mrdata: yes - no problem. Mar 28 15:27:45 root@hx4700:/tmp# ipkg install http://www.angstrom-distribution.org/unstable/feed/armv5te/base/xserver-kdrive-imageon_X11R7.1-1.1.0-r1_armv5te.ipk Mar 28 15:27:45 mr_nice: my cf card would mount on boot Mar 28 15:27:50 * koen crosses fingers Mar 28 15:28:21 koen: did you check the dbg thing with other arch's or only arm ? Mar 28 15:28:36 Ifaistos: only arm and i686 Mar 28 15:29:14 mr_nice: i could email you my .config and you could compare it Mar 28 15:29:23 mrdata: also I am not able to get any output from the suspend thing. Mar 28 15:29:31 mrdata: ah, yes that would be nice Mar 28 15:30:58 mrdata: do you know how such a suspend call looks that I can grep for it? Mar 28 15:33:24 ack! I think I will just build the toolchain by hand. Mar 28 15:33:39 neither xserver-kdrive 1.2.0 nor git compile cleanly yet :( Mar 28 15:33:50 mr_nice: email done Mar 28 15:34:19 mrdata: thx Mar 28 15:34:34 mrdata: I hope I can help you with the suspend thing Mar 28 15:37:10 mr_nice: i think, i have isolate the problem, sa1100-pcmcia suspend function will not end okay Mar 28 15:38:19 mrdata: ah, so we should try to compile without pcmcia and check if the suspend is ok? Mar 28 15:40:19 pH5, psokolovsky_: Ximageon doesn't crash the hx on first run anymore Mar 28 15:40:25 MickeyL: Did we still hear something from TrollTech ? Mar 28 15:40:57 koen: great. I'll check when I get kdrive 1.2 to build. Mar 28 15:41:05 mrdata: woglinde told me something about pcmcia and wrong irqs or something Mar 28 15:41:27 pH5: but the display is garbled and rotated, and for some reason it won't work with tslib Mar 28 15:41:55 koen: is hx the other chip? Mar 28 15:41:59 mr_nice: i have understand, that he means exlusive irq and possible shared requests Mar 28 15:41:59 pH5, psokolovsky_ http://rafb.net/p/I5uQdn42.html Mar 28 15:42:04 XorA: w3220, yes Mar 28 15:42:20 koen: didnt sirfred say thats not supported yet, I dont think he has the hardware Mar 28 15:42:48 koen: that has 800x600 frambuffer? Mar 28 15:42:59 XorA: 480x640 Mar 28 15:43:18 XorA: sirfred said he fixed some bugs that might make it work on w3220 Mar 28 15:43:29 and he indeed fixed some bugs :) Mar 28 15:43:39 koen: sounds like someone needs to post him a machine for a couple of weeks :-) Mar 28 15:44:09 XorA: iirc hrw tried, but the ebay-thing fell through Mar 28 15:45:11 mr_nice: you could look at http://www.pastebin.ca/413616, to see the error i have Mar 28 15:45:37 mrdata: ok Mar 28 15:45:45 anyway time for me to find bus Mar 28 15:46:20 * cyrilRomain plugs in his brain and use CONFIG_TUN=m (and not CONFIG_TUN=y) to get kernel-module-tun Mar 28 15:47:02 koen: and it was e740 which is w100 too iirc Mar 28 15:47:57 Xora: Don't stop at the pub ; Mar 28 15:47:59 ;) Mar 28 15:52:46 mrdata: unfortunately I am not good in reading such debug messages :/. Code: bad PC value is the problem? Mar 28 15:57:01 mr_nice: that is the result Mar 28 15:58:09 mr_nice: but the drivers/pcmcia/sa1100-generic -> .suspend = pcmcia_socket_dev_suspend call will fail, what i see Mar 28 15:58:33 mrdata: ksymoops benutzen ;) Mar 28 15:59:07 mr_nice: zecke thanks, i hope the whole time you would reading this Mar 28 16:00:00 mrdata: but the address is quite high/random... Mar 28 16:00:17 zecke: but, ksymoops need /proc/ksym i have not such one now, also ksysmoops from oe does not compile Mar 28 16:00:42 zecke: so i used the one from openzaurus Mar 28 16:01:50 zecke: but, ksymoops had told me, that he found my vmlinux would not be okay Mar 28 16:02:13 mrdata: hmm Mar 28 16:02:39 zecke: a kernel debugger would very usefull Mar 28 16:02:50 mrdata: e.g. if a needed function/name is not set the page address would be more close to 0 Mar 28 16:03:07 mrdata: well, kdb and it used to have ARM support but you don't need more than printk :) Mar 28 16:03:28 mrdata: before every statement enter a printk and then let it run once Mar 28 16:03:37 mrdata: print the data and pointers it is going to use Mar 28 16:03:55 mrdata: alternatively 'think' :) Mar 28 16:04:21 zecke: could you look at http://www.pastebin.ca/413616 please Mar 28 16:05:15 mrdata: I suck at PC and LR calculation as well :) Mar 28 16:06:15 zecke: i trace this since two days now, yesterday i land by sa1100-pcmcia.c and do following the way Mar 28 16:06:53 mrdata: 645f6370 is really weird, take a look at your System.map Mar 28 16:07:15 mrdata: and check if any function is near this address, this address is unlikely to be on the stack :) Mar 28 16:07:33 zecke: in the call for suspend, the next step where drivers/pcmcia/cs.c, but is not landing there Mar 28 16:08:32 zecke: okay, i will look for this, but i for myself have never touch the pcmcia stuff Mar 28 16:09:01 mrdata: you are going to touch it Mar 28 16:12:37 zecke: please dont do forget, i try deep linux kernel stuff for first, since two weeks now Mar 28 16:12:51 mrdata: yes, and this is awesome :) Mar 28 16:13:59 mrdata: The backtrace looks like it is not entering any pcmcia code at all Mar 28 16:14:23 splitnode Mar 28 16:15:10 mrdata: so check if the pcmcia platform device sets everything needed and check the driver sets the other stuff Mar 28 16:15:23 mrdata: check that the suspend/init method is not declared as initdata Mar 28 16:17:28 zecke: the pcmcia platform device should be set in drivers/pcmcia/sa1100-simpad.c right ? Mar 28 16:17:44 mrdata: let me check Mar 28 16:25:01 Hello Mar 28 16:28:45 cu Mar 28 16:29:28 can someone please push this gnutls DEPENDS->RDEPENDS fix http://rafb.net/p/ErXV8g67.html ? Mar 28 16:29:55 gnutls indeed requires those libraries to run, not to build Mar 28 16:31:19 cyrilRomain: this looks wrong Mar 28 16:31:27 zecke: why ? Mar 28 16:31:45 cyrilRomain: If it requires these libraries it is linking or dlopening them Mar 28 16:31:51 cyrilRomain: I assume that it is linking to them Mar 28 16:32:04 cyrilRomain: in that case these libs need to be available at build time (DEPENDS) Mar 28 16:32:05 hi sirfred. did you see koen's (partial) success report about Ximageon on w3220? Mar 28 16:32:14 cyrilRomain: and if that is the case RDEPENDS are automatically recognized Mar 28 16:32:41 zecke: right Mar 28 16:32:48 pH5: Hi. :) Mar 28 16:32:51 pH5: Well, very partial. Mar 28 16:32:55 pH5: I'm answering his mail. Mar 28 16:33:23 * cyrilRomain needs some break Mar 28 16:33:53 sirfred: great :) Mar 28 16:34:42 Xfbdev 1.2 works just fine, even with the imageon patches applied. now to figure out why it didn't build Ximageon... Mar 28 16:37:49 pH5: What error do you get? Mar 28 16:38:07 pH5: Perhaps the randr stuff has changed... Mar 28 16:39:03 pH5: Also, take in mind that I've changed internal offscreen memory management for Ximageon, probably that patch is not longer applying correctly. Mar 28 16:39:57 hmm.. maybe. I had problems earlier with bitbake trunk not telling me that some patches didn't apply correctly. Lets see.. Mar 28 16:41:13 Today, I found some web references about the nokia N800 multimedia player, "Canola", even some rumour about going GPL. Do you know something about that? Mar 28 16:41:18 yeah, -cpatch hangs again. :-/ Mar 28 16:41:29 sirfred: ask koen, he rumored the same Mar 28 16:41:37 sirfred: yes, canola is going opensource Mar 28 16:41:45 koen: It looks great. Mar 28 16:41:47 RP: can you reproduce this? Mar 28 16:41:54 sirfred: it's being reviewed by nokia legal ATM Mar 28 16:42:16 koen: I would like to put my hands on that, and be able to run in in my zaurus Mar 28 16:42:17 * koen recompiles Ximageon Mar 28 16:42:19 With xv support Mar 28 16:42:37 koen: DEBUG_IMAGEON 1 ? Mar 28 16:42:50 Is it based in libSDL ? Mar 28 16:43:37 I'm afraid libSDL is not too efficient on X, it doesn't use any acceleration. Mar 28 16:44:26 mrdata: the _simpad.c doesn't look too different from the _assabat.c file Mar 28 16:44:35 mrdata: so russell could see the same issue Mar 28 16:44:51 pH5: I bet kdrive-vidmemarea.patch could be the source of problems. Mar 28 16:45:30 I've even take a look at the libSDL 1.3 SVN code, but it doesn't seem to have improved that part. Mar 28 16:46:44 sirfred: both your patches can apply cleanly (with 1.2, not with git), but for some reason bitbake didn't use them. maybe an earlier patch failed. Mar 28 16:47:26 pH5: this is weird as base.bbclass is applying patches Mar 28 16:47:39 pH5: e.g. the interactive can't be entered? Mar 28 16:47:48 pH5: Rest of patches are shared with Xw100 and, I suppose, with Xfbdev Mar 28 16:47:48 zecke: I believe there is something wrong with bitbake spawning the interactive shell Mar 28 16:47:49 zecke: right Mar 28 16:47:54 Well, they are the same. Mar 28 16:48:05 I hate git "fatal: Untracked working tree file 'include/linux/netfilter/xt_TCPMSS.h' would be overwritten by merge." Mar 28 16:48:18 sirfred: SDL, but they are probably changing to evas Mar 28 16:48:48 koen: I don't know how evas interact with X, but I expect it to be a cleaner design. Mar 28 16:48:51 ah, it's optional-xkb.patch that fails in configure.ac Mar 28 16:48:57 pH5: Interactive tasks are broken in bitbake trunk and also less so in 1.8.0 Mar 28 16:49:06 zecke: dumb question -> means this the possibility for a real general error in the mainstream stuff ? Mar 28 16:49:25 RP: ah ok, I'll have to make sure then that my patches apply :) Mar 28 16:49:39 pH5: http://www.rpsys.net/openzaurus/temp/oe-terminal-r2.patch is an experiment to fix both Mar 28 16:49:46 mrdata: yes, I assume assabat to be broken as well (like the other sa1100 machines) Mar 28 16:49:59 mrdata: but let me compare things a bit more Mar 28 16:50:23 While I wait for Canola, I'm going to try that starling thing Mar 28 16:50:28 pH5: Does PATCH_RESOLVER = "noop" show the problem btw? Mar 28 16:50:33 zecke: i very happy for your help Mar 28 16:50:51 I had some issue compiling gstreamer ugly or bad. Mar 28 16:51:07 Canola like the media player? Mar 28 16:51:18 keesj: yes Mar 28 16:52:04 zecke: i have seen your work on kernel26 stuff for simpad, since two years now i think Mar 28 16:52:05 mrdata: http://lxr.linux.no/source/drivers/pcmcia/cs.c#L107 so this method is not reached? Mar 28 16:52:10 keesj: It seems it's going open source Mar 28 16:52:24 I wish I would have more spare time :} Mar 28 16:52:46 wow! what a great news! Mar 28 16:52:57 mrdata: http://lxr.linux.no/source/drivers/pcmcia/cs.c#L531 I assume this doesn't get called as well Mar 28 16:53:03 * pH5 adds PATCH_RESOLVER to local.conf.. Mar 28 16:53:08 it's gonna take off like a rocket Mar 28 16:54:00 I hope, are there much dsp stuff related to the mamo platform Mar 28 16:54:38 sirfred: http://rafb.net/p/CsD9p695.html Mar 28 16:55:03 mrdata: http://lxr.linux.no/source/arch/arm/mach-sa1100/generic.c#L282 as the device gets added by gernic.c there is little we can do wrong Mar 28 16:55:05 keesj: marcello announced it going open at bossaconference Mar 28 16:55:18 keesj: we'll have to wait for nokia legal to clear it Mar 28 16:55:29 mrdata: and without the sa1100 it is going to suspend? Mar 28 16:55:49 koen: And the cursor doesn't look like... a cursor? Mar 28 16:55:52 koen: did he explain why? Mar 28 16:56:04 koen: It looks right in the log. :) Mar 28 16:56:24 zecke: i have not try it without pcmcia now, but i will test it Mar 28 16:56:41 mrdata: this is the best thing you can do for now Mar 28 16:56:50 mrdata: compile a kernel without pcmcia subsystem Mar 28 16:56:55 * cyrilRomain wonders why 'libgnutls-config --libs' does not list zlib although objdump says 'NEEDED libz.so.1' ... Mar 28 16:57:11 did he explain why at the conf? Mar 28 16:57:14 mrdata: there are two outcomes and four possibilities Mar 28 16:57:23 koen: How does your PDA framebuffer is oriented before starting X, landscape? Mar 28 16:57:26 zecke: okay, i will do that and give response Mar 28 16:57:49 sirfred: portrait Mar 28 16:57:55 tux in topleft Mar 28 16:58:24 It seems the mode returned by w100fb is 640 x 480 and not 480 x 640 Mar 28 16:58:25 but I would already be happy with "screen" for the moment :p Mar 28 16:58:51 What a mess Mar 28 16:59:34 sirfred: starting gpe-login gives me http://rafb.net/p/CatDRo97.html Mar 28 17:00:10 zecke: #L107 printk(KERN_INFO "\ndrivers/pcmcia/cs.c -> pcmcia_socket_dev_suspend() *START*\n"); would not be reached Mar 28 17:00:30 koen: Well, that's normal operation. Mar 28 17:00:45 koen: Just blitting from offscreen to screen Mar 28 17:01:20 I'll put a photo of the screen online Mar 28 17:01:34 koen: What is unexpected is the framebuffer layer returning 640x480 when it's 480x640, isn't it? Mar 28 17:01:41 yes Mar 28 17:02:04 My zaurus returns that correctly I think. Mar 28 17:04:07 zecke: on the beginning, to find the bug, my own list had 72 suspend entrys and after 9 i make a catch, not so bad i think Mar 28 17:04:16 sirfred: http://www.openembedded.org/~koen/hx4700-ximageon.jpg Mar 28 17:04:50 koen: Terrible. Perhaps with some kind of 3D glasses... Mar 28 17:04:57 :) Mar 28 17:07:07 sirfred: could it be that the hh.org kernel has a hack to force fb rotation? Mar 28 17:07:15 * koen looks at pH5 for advice Mar 28 17:07:30 koen: Sincerely: I have no idea. Mar 28 17:10:17 koen: force no, but that's standard kernel feature, no? Mar 28 17:10:24 we don't have FRAMEBUFFER_CONSOLE_ROTATION set. could that be a problem for w100? Mar 28 17:10:48 koen: whoa, weird (seen screenshot) Mar 28 17:10:54 (in OE, hx4700 defconfigs are a bit of a mess, traditionally) Mar 28 17:11:08 It shouldn't be a problem, but I expected the fb to return 480x640. Mar 28 17:11:19 If it's in 480x640 mode, of course. Mar 28 17:11:45 koen: Can you try with another program what the framebuffer is returning, there was some fb util somewhere Mar 28 17:12:27 koen: What fbset says , for example? Mar 28 17:12:28 http://rafb.net/p/3kBxjj34.html Mar 28 17:12:47 the funny thing is that the console text is rotated as well after I kill kdrive Mar 28 17:12:54 (rotated and garbled) Mar 28 17:13:46 koen: Let me interpret the logs. Mar 28 17:16:47 I wouldn't be suprised if the kernel has a hack to fake rotation Mar 28 17:16:57 Linux hx4700 2.6.16-hh9 #1 PREEMPT Thu Jan 11 01:05:51 CET 2007 armv5tel unknown Mar 28 17:17:03 I have a pretty old kernel as well Mar 28 17:17:25 Reading the mmGRAPHIC_CTRL, I'm getting that the current w100 rotation is 90 degrees Mar 28 17:17:34 zecke: success, without pcmcia enabled, the simpad do suspend -> the power led is going of Mar 28 17:17:53 That is coherent with 640x480, I assume. Mar 28 17:17:56 mrdata: and is waking up again? Mar 28 17:18:03 But you say that you're on 480x640 Mar 28 17:18:09 ahm Mar 28 17:18:13 * sirfred coredumps Mar 28 17:18:24 zecke: and now, i have the same problem like mr_nice, push on power button -> NOTHING Mar 28 17:18:42 mrdata: another outome :) Mar 28 17:19:28 mrdata: either fix suspend or resume. I assume suspend is more easy to fix Mar 28 17:19:58 mrdata: what schocks me about your backtrace is: not in any register or in the stack is an address starting with 6, or near the range of six :} Mar 28 17:19:59 zecke: suspend with pcmcia or without? Mar 28 17:20:07 mrdata: with PCMCIA Mar 28 17:20:23 mrdata: your debugging didn't break anything? Mar 28 17:21:23 zecke: i hope not, i only bombs printk messages inside ervery function you could see Mar 28 17:21:26 koen: I cannot see the cursor in your screenshot, was it present? Mar 28 17:23:50 zecke: and you know, that my simpad have 128MB RAM inside Mar 28 17:24:38 zecke: could that be involved? Mar 28 17:25:07 mrdata: well you altered the kernel to make use of 128? Mar 28 17:25:18 mrdata: if it makes a difference it should be fixed anyway Mar 28 17:25:29 how should I choose a distro for a new device? Mar 28 17:25:29 I'm getting this while trying to configure gst-plugins-bad-0.10.3-r2: Mar 28 17:25:34 anyway, work time ends, hacking starts Mar 28 17:25:40 Checking for FAAD >= 2... configure: error: cannot run test program while cross compiling Mar 28 17:26:03 Is there any standard way to get that kind of errors fixed? Mar 28 17:26:08 sirfred: for some reason the cursor wasn't visible Mar 28 17:26:31 koen: Neither with nude X ? Mar 28 17:26:49 zecke: yes and no, i tell only the kernel with linuxargs, that i have mem=128MB Mar 28 17:26:55 Or perhaps a transparent cursor theme? Mar 28 17:26:55 sirfred: It was visible before I rebooted it Mar 28 17:27:02 it kind of looked like an 'X" Mar 28 17:27:14 koen: That's how it should looks like Mar 28 17:27:33 koen: Well, I'm sure you know how a raw X cursor looks like. But you said it was screwed in some way? Mar 28 17:27:59 yes Mar 28 17:28:04 So, to summarize Mar 28 17:28:11 zecke: Memory: 128MB = 128MB total Mar 28 17:28:16 Before starting X, you're at 480x640 Mar 28 17:28:24 yes Mar 28 17:28:28 zecke: Memory: 127360KB available (2244K code, 187K data, 92K init) Mar 28 17:28:42 mrdata: I think that should be enough Mar 28 17:28:44 When you kill X, you're in some screwed display mode, perhaps because Ximageon has tried to request 640x480 mode. Mar 28 17:29:08 fbset return 640x480 mode even when it seems to be 480x640 Mar 28 17:30:00 I can confirm that fbset returns 480x640 in my husky when in portrait mode Mar 28 17:30:01 mrdata: I fixed the mq200 patch. did you changed the patch file direcly by editing it? Mar 28 17:30:02 zecke: i hope this also, more is for sa1100 on simpad not possible i think Mar 28 17:30:52 mrdata: alternatively check guyhlems old 2.4er patch to support 128MB of RAM Mar 28 17:30:54 mr_nice: i must say yes, was an extra error inside to Mar 28 17:31:04 mrdata: he needed to patch the fixup routine, but we don't use one anymore :) Mar 28 17:31:17 And the last annoying thing is that the w100 says that the framebuffer is rotated 90 degrees. Mar 28 17:31:20 mrdata: no problem thx for working on the driver! Mar 28 17:31:53 koen: Could you try with linux-rp_2.6.20 ? Mar 28 17:32:08 that doesn't run on my device, I'm affraid Mar 28 17:32:17 koen: oh Mar 28 17:32:46 I'm stuck with this hh.org kernel that will never be merged into mainline Mar 28 17:32:50 zecke: i have looked on the old patch, but i had think that it would be not nessesary now Mar 28 17:33:04 mrdata: I agree Mar 28 17:33:23 mrdata: do you know a good irc channel for stupid newbie c questions? Mar 28 17:33:36 03justinp 07org.oe.dev * r729174a6... 10/ (1 conf/distro/include/openzaurus.inc): openzaurus.inc: add PREFERRED_PROVIDER for qemu-native Mar 28 17:33:37 zecke: bootloader had also detect the 128MB ram Mar 28 17:34:08 koen: It's difficult to figure out what's happening. But it seems that the w100 is rotated clockwise 90 degrees on startup Mar 28 17:35:19 mr_nice: no, i learnt it my own way, three or five books, trail and error, read other sources, try to understand Mar 28 17:35:48 mrdata: right, I think 128MB of RAM are save. The question is why the driver's suspend method is not called at all Mar 28 17:35:56 I'm getting the same values for mode and rotation in my zaurus when starting, but my zaurus IS in 640x480 mode Mar 28 17:36:16 sirfred: could we add a machine specific hack? Mar 28 17:36:18 mrdata: either the device is not properly registered, the pointers are destroyed. So findout which statement leads to the derefence Mar 28 17:36:31 e.g. if(hx4700) {rotation = bogus}? Mar 28 17:36:38 koen: But I don't know if it's machine specific or kernel specific or what Mar 28 17:36:48 mr_nice: also my question now, the call is going up, but land everythere Mar 28 17:37:01 sirfred: psokolovsky_ and pH5 have the same machine and have worked on the kernel, maybe they have some idea Mar 28 17:37:06 koen: I think a friend of mine has a hx4700 Mar 28 17:37:23 koen: Perhaps I could borrow it some days. I will ask him Mar 28 17:37:42 zecke: for such things, a debbuger would be helpfull Mar 28 17:38:02 sirfred: but just in case, hx should run in 480x640, of course Mar 28 17:38:32 psokolovsky_: But that's not what the framebuffer is saying. And that's the first problem. Mar 28 17:40:18 zecke: for what i have reading, kgdb is on kernel 2.6.15 and not for arm, when i'am right Mar 28 17:40:23 sirfred: ~ # cat /sys/class/graphics/fb0/modes Mar 28 17:40:23 U:480x640p-2903 Mar 28 17:40:23 ~ # cat /sys/class/graphics/fb0/virtual_size Mar 28 17:40:23 480,640 Mar 28 17:40:57 psokolovsky_: And using fbset ? Mar 28 17:41:35 sirfred: I don't have fbset here (bare kernel), but you see what kernel has Mar 28 17:41:42 hey ppl, would someone take a look at this: http://pastebin.ca/413747 ? It's buggin me... Mar 28 17:41:47 psokolovsky_: Yes, but koen result was strange. Mar 28 17:41:59 Well, I'm going to compare with my husky Mar 28 17:42:09 sirfred: yep ;-I Mar 28 17:42:55 psokolovsky_: Yes, that's what I get. Mar 28 17:44:03 sirfred: I'll try to try ximageon myself, just dunno when I'll have chance ;-( . I appreciate yout work on that, just myself is dug into more low-level issues now ;-( Mar 28 17:44:32 psokolovsky_: ok. Nice to hear. Mar 28 17:49:55 zecke: i think i have a mistake in my logic -> look at my pastebin again please at line #120 to #138 Mar 28 17:50:37 zecke: platform_suspend() was calling Mar 28 17:51:50 mrdata: can you give me the link again Mar 28 17:52:19 zecke: http://www.pastebin.ca/413616 Mar 28 17:52:57 zecke: with work will dev->driver->suspend_late will do? Mar 28 17:53:13 zecke: witch not with Mar 28 17:57:23 mrdata: I don#t know your printk's but both dev->driver and dev->driver->suspend_late are true Mar 28 17:57:46 mrdata: dev->driver->suspend_late == 645f6370 ? Mar 28 17:58:58 mrdata: with your config I was able to see hda in /proc/partitions also I am now able to log into the gpe image :). I will compare it with the one on sf.net and change the defaultconfig Mar 28 18:02:13 mr_nice: is okay Mar 28 18:02:17 mrdata: I just checked apm --suspend and I am only getting apmd[2318]: Suspending now without anythin else Mar 28 18:02:39 mrdata: no error. no other output Mar 28 18:03:43 mrdata: did you changed the /proc/sys/kernel/printk value? Mar 28 18:04:31 mr_nice: i have found that in my .config CONFIG_DISABLE_CONSOLE_SUSPEND=y is removed Mar 28 18:05:09 mr_nice: i have this not removed, kernel compile is somthimes mystery Mar 28 18:05:23 mr_nice: sometimes Mar 28 18:05:42 mrdata: ok I will change it. thx Mar 28 18:06:42 zeck|moko: my printk's are trivial, at the begin of each function, i set *START* Mar 28 18:07:27 zecke|moko: for ever call i set *BEGIN*, after the call *DONE* Mar 28 18:07:38 mrdata: on mine it is set. I will change the /proc printk stuff and try it a second Mar 28 18:08:08 zecke|moko: at the end of the function i set *ENDE* Mar 28 18:08:08 mrdata: print the addresses of the late address? Mar 28 18:09:14 zecke|moko: no, this comes from the kernel alone Mar 28 18:10:43 mrdata: please print the address of the _late pointer Mar 28 18:10:50 mrdata: as this is not within the stack or registers Mar 28 18:10:58 it is referenced indirectly Mar 28 18:11:02 mrdata: echo 8 > /proc/sys/kernel/printk did the job here is my output http://www.pastebin.ca/413805 Mar 28 18:11:09 zecke: the mystery for me is, that i had yesterday a Internal error: Oops - bad syscall: 524b50 [#1] Mar 28 18:12:19 zecke|moko: after set pcmcia_debug in .config i got this paging error Mar 28 18:12:47 mrdata: you smash something :) Mar 28 18:13:23 mrdata: mr_nice's stack trace looks better Mar 28 18:13:31 mrdata: r3 contains the bad address Mar 28 18:17:06 hey zecke: how goes that greenphone? Mar 28 18:17:25 RP: with PATCHRESOLVE="noop" -cpatch fails like it should. Mar 28 18:21:03 re Mar 28 18:22:09 koen: Just a final thought. Did you run try to run fbset before trying to run any Ximageon ? Mar 28 18:22:59 koen: Also, the useful Ximageon run log would be the one from the first try. Just after reboot. Mar 28 18:23:45 sirfred: I'll try to look into that tomorrow Mar 28 18:23:49 koen: ok Mar 28 18:24:01 mrdata: we could try to ask at #kernelnewbies on irc.oftc.net Mar 28 18:25:01 mr_nice: yes, but also for arm kernel stuff Mar 28 18:25:49 would someone please take a look at this: http://pastebin.ca/413747 ? I need help understanding what that is... Mar 28 18:26:19 mr_nice: i will pcmcia enable before and search my own fault like zecke said Mar 28 18:26:44 mrdata: where do I need to add the printks? Mar 28 18:27:48 mrdata: in sa11x0-pcmcia? Mar 28 18:28:04 In order to avoid an error while configuring gst-plugins-bad-0.10.3, I've just disabled checking of FAAD version, as the way of checking that was to create a little program to figure out the version. This was failing because we're crosscompiling. I've just disabled that version check, and left only header and libs presence checks. Is this a correct approach for oe packages? Mar 28 18:28:12 mr_nice: drivers/base/platform.c -> platform_suspend_late() Mar 28 18:28:31 mrdata: thx Mar 28 18:28:37 sirfred: iirc we only have one FAAD version, so it's appropriate Mar 28 18:28:49 ('we' == 'oe') Mar 28 18:29:05 koen: So, I think that a bug report could be appropiate Mar 28 18:29:24 Attaching the patch, of course Mar 28 18:30:49 It seems starling is also missing a sqlite3 depend. Mar 28 18:31:27 mr_nice: drivers/pcmcia/cs.c -> pcmcia_socket_dev_suspend Mar 28 18:31:55 mrdata: ok Mar 28 18:33:45 please, would someone help me to understand why *do_patch* is failing? Mar 28 18:38:21 djay-il: clearly quilt is failing. I seem to recall that this problem has occured in the past, and perhaps might be due to having a symbolic link in the path to the build tree? Mar 28 18:39:48 how would I check that? Mar 28 18:42:51 djay-il: sounds like you may have a different version of quilt in your path than is required... Mar 28 18:53:40 I've just added bug 2057 to fix gst-plugins-bad crosscompilation and bug 2055 to add a dependency of starling with sqlite3, if you're interested. Mar 28 18:57:41 JustinP: ok, thanks, I'll check it out... Mar 28 19:23:39 pH5: Thanks, that proves it is the interactive task problem rather than any other subtle problem :) Mar 28 19:25:53 mrdata: ok here is my new output with the printks http://www.pastebin.ca/413888. unfortunately I forgot to add a \n :/. But it is the same adress and the same register. I now try to find out how to print the stuff inside the strucs.I will try to find the prototypes in the header files and generate propper printks with adresses and content. Mar 28 19:32:00 I'm having bootstrap-image fail on me during do_configure. It borks out on an unexpected "(", http://pastebin.ca/413904 Mar 28 19:32:11 (The first "(" in the oewarn) Mar 28 19:33:51 anyone know why it fails there? Mar 28 19:36:53 I'm seeing that gst-plugins-good ipk is trying to launch gconftool-2 in pre and post install stages. Mar 28 19:37:15 Where is that configured. I mean, where those hooks are configured for the ipk? Mar 28 19:37:36 sirfred: inside the bb file Mar 28 19:38:01 zecke: I was looking there, but perhaps didn't look enough. Thanks. Mar 28 19:38:11 sirfred: 'inherit gconf' iirc Mar 28 19:38:46 koen: Yes, gst-plugins.inc is doing that. Thanks. Mar 28 19:42:51 hrw|gon: is your internal libs patch in already? I'm still seeing problems with (openmoko-)dates Mar 28 20:14:30 koen: we finally made the release, could you please update again? http://mediatomb.svn.sourceforge.net/viewvc/*checkout*/mediatomb/trunk/mediatomb/scripts/mediatomb_0.9.0.bb Mar 28 20:21:51 zecke: do you know if something similiar is needed for simpad, too? http://www.gelato.unsw.edu.au/lxr/ident?i=pcmcia_collie_init Mar 28 20:22:39 ~collie Mar 28 20:23:10 mr_nice: we have that Mar 28 20:23:25 mr_nice: http://www.gelato.unsw.edu.au/lxr/source/drivers/pcmcia/sa1100_generic.c#L46 Mar 28 20:23:35 check line 65 and then search for simpad Mar 28 20:23:47 (hint the one directly above) Mar 28 20:24:16 03koen 07org.oe.dev * r1ed405ff... 10/ (1 packages/mediatomb/mediatomb_0.9.0.bb): mediatomb: merge 0.9.0 recipe from mediatom svn Mar 28 20:24:31 Jin^eLD: there you go Mar 28 20:25:18 zecke: thx. Mar 28 20:25:49 re Mar 28 20:26:12 koen: thanks! Mar 28 20:26:44 florian: Gtk-Kenner: Do you know any Gtk+ Widget capable of displaying RichText like this (http://doc.trolltech.com/2.3/qstylesheet.html) Mar 28 20:29:37 Jin^eLD: I did change it slightly Mar 28 20:30:12 koen: ok, I will update our version to sync with yours then Mar 28 20:30:37 I wish someone could test largefile :> disabled just "to be safe" now, I had some problems with it when crosscompiling Mar 28 20:31:22 methinks collie is sharp zaurus SL-5000 (often called SL-5000D because it was developer version) or SL-5500 (also SL-5500G with QWERTZ keyboard) or a dog Mar 28 20:38:56 have no real testplatform for it :P Mar 28 21:23:22 zecke: you were looking for me? Mar 28 21:24:01 rwhitby: yes, just to help a user but XorA managed to help him :) Mar 28 21:24:40 ok, saw that. sweet. Mar 28 21:25:25 * rwhitby considers adding "make package-" capability to MokoMakeflie ... Mar 28 21:25:37 make me-a-shell Mar 28 21:26:11 hmm - make bitbake-devshell ? Mar 28 21:28:49 and put the same into the SlugOS MasterMakefile! Mar 28 21:30:03 * mwester would it be out-of-line to add some text to the prompt for the created shell indicating that one was inside such a "special" shell? Mar 28 21:30:36 s/would/wonders if it would/ Mar 28 21:31:05 Sigh. Still bad grammer. I need another cup of coffee. Mar 28 21:35:02 rwhitby: Maybe make it drop to bitbake -i ? Mar 28 21:35:46 hi Mar 28 21:35:55 hi CoreDump|home Mar 28 21:47:28 Good night Mar 28 21:55:56 zecke: oops, i missed your question. Mar 28 21:56:14 zecke: a GtkTextView should be able to do that Mar 28 21:56:26 florian: should, yes if I implement my own tokenizer Mar 28 21:56:42 florian: and even then tables and linebreaks are a miracle :) Mar 28 21:58:17 zecke: hmm... what do you want to implement? Mar 28 21:59:00 florian: showing the pseudo HTML of RSS feeds Mar 28 21:59:21 florian: the closest match so far is gtkimhtml.c from gaim (which lacks table and image proxy support though) Mar 28 22:01:03 zecke: sounds like a challenge... why not wrap it and use a html widget? Mar 28 22:01:35 florian: which HTML widget is available in Gtk+? Mar 28 22:03:49 zecke: there is GtkMozembed, GtkHTML in some different flavours and GtkWebcore at least. Mar 28 22:04:50 zecke: if you don't have a html component on your system already GtkHTML might be the best - it only supports the basics but it is not too big. Mar 28 22:05:01 florian: to tell you the truth, I want something as small but powerful like QSimpleRichText (just give it HTML and it renders what it can and you can easily fetch images on demand) Mar 28 22:05:15 florian: mean while I'm struggling with WebKit/Gdk and cmake :) Mar 28 22:07:44 eh cmake :-/ Mar 28 22:07:56 zecke: try gtkhtml :-) Mar 28 22:07:58 15% of WebKit builds already Mar 28 22:08:11 florian: I'm overwhelemed by the GtkHTML flavors Mar 28 22:08:36 thanks for the input Mar 28 22:08:41 and see you later Mar 28 22:08:55 :-) Mar 28 22:09:17 florian: BTW: have you seen my first published gtk+ app? Mar 28 22:09:39 zecke: not yet, let me see a link! :-) Mar 28 22:10:08 zecke: i was tempted to ask who you are and what happened to zecke ;) Mar 28 22:10:26 florian: http://page.mi.fu-berlin.de/~freyther/FeedReader.png Mar 28 22:10:42 florian: the source is in openmoko svn, but I'm still learning the Gtk classes :} Mar 28 22:11:30 zecke: i have seen gtkhtml in their images so that would be the way to go without adding too much html overhead. Mar 28 22:11:59 zecke: but that's nice Mar 28 22:12:15 * florian needs to spend more time coding Mar 28 22:13:31 If there anyone semi-responsible for busybox? MAINTAINERS doesn't list anyone. busybox.inc includes specific patches that obviously doesn't apply to all versions. Mar 28 22:15:52 03coredump2 07org.oe.dev * rabddf075... 10/ (64 files in 18 dirs): Mar 28 22:15:52 altboot: Mar 28 22:15:52 - Drop files/* and altboot_0.0.0.bb Mar 28 22:15:52 - Add altboot_svn.bb Mar 28 22:15:52 - Add README.txt explaining what the different versions do Mar 28 22:42:57 hi mickeyl Mar 28 22:43:07 cheers Mar 28 23:03:10 gm Mar 28 23:41:03 http://www.openembedded.org/wiki/BuildOnFreeBSD That's helpful :P Mar 28 23:47:27 morning Mar 28 23:56:38 rwhitby: /query ? Mar 29 00:50:18 SonicvanaJr, open source is only as good as its users ... Mar 29 00:50:36 Very true Mar 29 00:50:41 :) Mar 29 00:51:50 Still it'd be nice to have a little bit of info, there's a ton of errors/warnings when building Mar 29 00:52:05 sounds like you are a pioneer Mar 29 00:53:38 Or attempting to tread in some footsteps that have been very well covered Mar 29 00:53:52 I suspect they are not well covered Mar 29 00:53:58 snce the web page is empty Mar 29 00:55:10 what errors and warnings btw? Mar 29 00:55:26 * Crofton|laptop has never really used *BSD Mar 29 02:14:44 koen: do you know the format required for the psplash progress bar image? Is it just a box? Mar 29 02:27:23 koen: disregard that please. I've modified psplash to just draw the border itself :) **** ENDING LOGGING AT Thu Mar 29 02:59:56 2007