**** BEGIN LOGGING AT Sun Jan 14 02:59:58 2007 Jan 14 03:26:59 /quit Night Jan 14 07:30:08 morning guys Jan 14 09:31:56 03pfalcon 07org.oe.dev * re9118e1c... 10/ (4 files in 3 dirs): qte 2.3.10: Fix kernel-keymap.patch, was made as -p0 instead of -p1. Jan 14 09:44:55 03pfalcon 07org.oe.dev * r718ed4ff... 10/ (4 files in 3 dirs): Jan 14 09:44:55 opie-taskbar: Remove overriden opie startup file for htcuinversal Jan 14 09:44:55 * This contained (comparing to the default) only explicit QWS_MOUSE_PROTO & Jan 14 09:44:55 QWS_KEYBOARD assignments, and it htcuinversal was confirmed to work without Jan 14 09:44:55 them. Jan 14 09:52:08 03pfalcon 07org.oe.dev * ra7bb5ea8... 10/ (1 packages/linux/linux-handhelds-2.6/rx3000/defconfig): Jan 14 09:52:08 linux-handhelds-2.6: Update defconfig for rx3000. Jan 14 09:52:08 * P.S. Please always lart linux kernel mainline for any Kbuild breakages Jan 14 09:52:08 like endless loops during make oldconfig. Jan 14 10:07:27 morning all Jan 14 10:11:09 likewise: likewise! Jan 14 10:13:37 koen: got the headless music daemon working on efika yesterday, so it's a nas with audio out device now. Jan 14 10:16:31 likewise: cool Jan 14 10:31:55 * likewise is fiddling with KVM under Feisty 2.6.20... Jan 14 10:53:35 The Pirate Bay seems to be interested in Sealand. Must we bid higher? Jan 14 10:56:06 if OE had that kind of money, we would have bought the xscale division ;) Jan 14 10:58:38 and get a bargain for the {x,free}scale combo? Jan 14 11:11:45 hi zecke Jan 14 11:18:33 hi zecke, pb_ Jan 14 11:23:56 hi likewise Jan 14 12:20:12 <[cc]smart> | ERROR: "gcc" looks like gcc 4.x - | QEMU is known to have problems when compiled with gcc 4.x - should configure be run with | To use this compiler anyway, configure with --disable-gcc-check or should angstrom be compiled using gcc3 ? Jan 14 12:21:02 [cc]smart: install gcc-3.3 or gcc-3.4 or gcc32 in host system Jan 14 12:21:02 [cc]smart: apt-get install gcc-3.4 Jan 14 12:21:08 hey hrw|gone Jan 14 12:21:14 mickey|berlin: good morning! Jan 14 12:21:37 * hrw for a moment Jan 14 12:22:34 morning folks Jan 14 12:22:38 i am sooo tired Jan 14 12:22:49 mickey|berlin: you went by car? Jan 14 12:23:00 no no, took the ICE Jan 14 12:23:03 ah Jan 14 12:23:11 * koen hugs ICE Jan 14 12:23:25 especially the electrical outlets Jan 14 12:23:26 bbiab, picking up Sean at the station... Jan 14 12:23:32 yeah ICE rocks Jan 14 12:24:02 http://www.hrw.one.pl/2007/01/14/openmoko-developers-program/ Jan 14 12:25:19 hmm. recent opie patches gives extra work for me - will have to isolate them and add into OZ 3.5.5 as they are too worth to be missed Jan 14 12:25:29 hrw: where do you see the developer program? Jan 14 12:25:39 koen: openmoko.com Jan 14 12:25:49 Notice to Developers If you are interested in developing Free Software applications for the OpenMoko platform, please send information regarding embedded Linux projects you have contributed to, and the work you have done to coreteam@openmoko.org. Jan 14 12:26:03 that has been there for months Jan 14 12:26:23 morning all Jan 14 12:26:27 koen: I was there only on start ;D Jan 14 12:26:31 hey do13 Jan 14 12:26:35 hi Dirk Jan 14 12:26:54 hey koen, Marcin Jan 14 12:26:56 hrw: actually, that has been there from the start :) Jan 14 12:27:07 koen: so I missed it on start :D Jan 14 12:27:44 I haven't mailed it since I won't be developing applications :( Jan 14 12:32:32 anyway post edited Jan 14 12:33:11 time to prepare and go out to town Jan 14 12:33:12 cu Jan 14 12:33:46 * koen install http://www.mdk.org.pl/articles/category/nflick on his 770 Jan 14 12:36:53 * likewise reboots, because kvm reports kernel version mismatch. Jan 14 12:38:43 mickey|berlin: hey Jan 14 12:39:42 psokolovsky: kernel-keymap.patch is safe to always apply. For some zaurus models it is patched to rotate cursor keys, these patches are wrong, people should fix the keymap instead! Jan 14 12:40:05 psokolovsky: besides support for key composition the kernel-keymap patch is working alright Jan 14 12:40:42 psokolovsky: it is not applied for all platforms for two reasons. Most iPAQs don't have a right keyboard anyway, and we are conservative engineers. Jan 14 12:41:57 hey zecke Jan 14 12:46:58 ls Jan 14 12:47:01 ww Jan 14 12:50:31 Hi! Jan 14 12:50:45 zecke, thanks for your comments! Jan 14 12:51:35 zecke, I'm exactly afraid that applying it generally would require dealing with kernel keymaps, which so far not required for non-keybaorded devices. Jan 14 12:52:07 zecke, but can you formulate what specific problem it was made to solve in the first place? Jan 14 13:01:15 zecke, and I'd probably would think about making it another, separate handler, instead of overlaying existing "TTY" implementation. Jan 14 13:03:54 hi psokolovsky ... Jan 14 13:04:06 hrw|gone: http://lkml.org/lkml/2007/1/14/49 Jan 14 13:04:31 Hi! Jan 14 13:08:37 a t-shirt for psokolovsky (sorry for long url) http://jitcrunch.cafepress.com/jitcrunch.aspx?bG9hZD1ibGFuayxibGFuazoxNTJfRl9jNi5qcGd8bG9hZD1MMCxodHRwOi8vaW1hZ2VzLmNhZmVwcmVzcy5jb20vaW1hZ2UvMTYwMDM2MTdfNDAweDQwMC5wbmd8fHNjYWxlPUwwLDE2NSwxMTIsVHJhbnNwYXJlbnR8bG9hZD10bS1MMCxibGFuazoxNTJfRl9jMjJfdG1hc2suanBnfGNvbXBvc2U9TDAsdG0tTDAsVGV4dHVyZU1hc2ssLTE1OCwtMTA1fGNvbXBvc2U9YmxhbmssTDAsQWxwaGFCbGVuZCwxNTgsMTA1fGNwPXJlc3VsdCxibGFua3xzY2FsZT1yZXN1bHQs Jan 14 13:08:37 MCw0ODAsV2hpdGV8Y29tcHJlc3Npb249OTV8 Jan 14 13:17:30 after a long break I tried to build oe (oz opie-image in .dev) and failed ;-) Jan 14 13:27:32 psokolovsky: it reads the keymap from the kernel. So in theory nothing would change for iPAQs Jan 14 13:28:59 psokolovsky: the problem solved is the following: Sharp Devices are localized: German, Spanish, English versions were available + different devices all needed to patch the 'keymap' inside QtE. With the kernel-keymap.patch we can use one Qt to rule them all, and are more flexible Jan 14 13:30:03 zecke, cool, thanks for answers, kernel-keymap is definitely the way to go then Jan 14 13:30:44 afk... Jan 14 13:43:29 Now, what the hell is going at http://rafb.net/p/gxhUPW96.txt? "No such file" when /usr/bin/ldd is right there? Jan 14 13:46:50 Laibsch, ldd is a shell script, not something you may want to trace. Jan 14 13:49:17 polyonymous: OK, there you got it. Jan 14 13:50:00 My self-compiled ldd has "#! /bin/bash" Jan 14 13:50:10 Does it really need bash? Jan 14 13:56:14 no idea. And I never managed to get ldd work on handheld :) Jan 14 13:57:34 polyonymous: It used to work for me previously (and it stlll does on my non-experimental altboot image) Jan 14 13:57:51 !oebug 1788 Jan 14 13:57:52 * * Bug 1788, Status: NEW, Created: 2007-01-14 06:26 Jan 14 13:57:53 * * bugs.openembedded.org(AT)rolf.leggewie.biz: Latest ldd depends on bash Jan 14 13:57:54 * * http://bugs.openembedded.org/show_bug.cgi?id=1788 Jan 14 13:58:46 My ldd has /bin/sh and doesn't work :) Jan 14 14:00:03 I haven't tried hard to make it work. I can't compile opie-image now, anyway :) Jan 14 14:24:14 Let's fix things one step at a time. Jan 14 14:39:05 hi all Jan 14 14:49:02 morning Jan 14 16:07:57 hi florian, chouimat Jan 14 16:31:07 hi florian Jan 14 16:32:27 hi pb_ Jan 14 16:43:33 What should I do to make sure that a certain package does not add " -lXmu" on compilation? Jan 14 16:56:02 mickey|party: ping Jan 14 16:56:36 zecke: only nerds would be in IRC during a party (unless it's a LAN party :-) Jan 14 16:56:46 zecke: Hi and merry new year, BTW Jan 14 16:57:20 hpy :) Jan 14 16:57:32 schurig: we are nerds with base flatrates :} Jan 14 16:58:08 and at a LAN party you should be immersed in gaming, so.... Jan 14 17:14:06 morning all Jan 14 17:15:07 hi rp Jan 14 17:15:53 hey RP Jan 14 17:18:02 JustinP: bah. When I was young, LAN-parties weren't all about gaming. They were proper scene-parties ;-) Jan 14 17:18:35 mornin Jan 14 17:18:50 NAiL :D Same here Jan 14 17:19:23 zecke: can you help me to understand a "circular dependency" thing with bitbake? Jan 14 17:19:42 RP My howto get touchscreen working properly in xorg is driving me nutters :D Anyway to manually calibrate it? Jan 14 17:19:47 pb_: I might be able to help... Jan 14 17:20:16 RP: I try to build qt-x11-free, and bitbake reports: Jan 14 17:20:17 Kristoffer: echo foobar > /etc/pointercal ? Jan 14 17:20:18 ERROR: Task 38 (/home/pb/oe/org.openembedded.dev/packages/mysql/mysql_4.1.18.bb, do_install) (depends: Set([37])) Jan 14 17:20:49 RP brilliant, thx Jan 14 17:21:11 err Jan 14 17:21:17 thought it said /proc/pointercal Jan 14 17:21:30 pb_: Does it give any more context than that? Jan 14 17:21:40 I dont have any of the fancy calibration stuff, just want to feed the kernel with calibrations (or values which I then can test) Jan 14 17:22:12 Kristoffer: I'm assuming you were using tslib which uses that file. That won't help with kernel calibration Jan 14 17:22:53 RP dont got tslib or anything else. No way to feed values to the kernel and do some tests with? Jan 14 17:23:59 Kristoffer: calibration is in userspace Jan 14 17:24:09 ah, doh then¨ Jan 14 17:25:54 pb_: you are in good hands :) Jan 14 17:26:31 Kristoffer: write a raw module that takes data from a socket? (not knowing the context) Jan 14 17:27:07 Kristoffer: If you install tslib, it has utils with it which should be able to test out the toucshcreen Jan 14 17:27:56 RP my problem is basicly getting tslib compiled here (fiddling with armedslack atm). And I dont think xorg has default support tslib Jan 14 17:28:01 thought about maybe xtscal Jan 14 17:28:40 Kristoffer: xtscal uses touchscreen extensions to X so it won't help you Jan 14 17:29:09 The touchscreen works atm, but calibration is way off... Seems to be little information out there about this. Jan 14 17:29:34 Kristoffer: Is it an ALPS device? Jan 14 17:29:42 Spitz :D Jan 14 17:30:34 Kristoffer: Right, I wrote the kernel driver. It expects userspace calibration. You need tslib. Jan 14 17:32:10 Ok, figured as much. Well, next project is to get tslib support compiled into xorg then? Jan 14 17:32:39 Kristoffer: You can test tslib standalone if you want first Jan 14 17:32:53 but yes, you then need to make it work woth xorg Jan 14 17:33:11 Kristoffer: At some point you wonder if just using OE would have been easier ;-) Jan 14 17:33:15 or just use kdrive Jan 14 17:33:38 RP Yeah, I've already felt that. Its more of a pet project atm :D Jan 14 17:34:12 Kristoffer: what does Xorg provide that kdrive doesn't have? Jan 14 17:34:20 Xorg-xserver, that is Jan 14 17:35:02 pb_: FWIW, I wrote the new task handling code in bitbake trunk so I should be able to help if you can't see why bitbake is upset about a circular depends. I'm the first admit the errors could be clearer :-/ Jan 14 17:35:16 koen nothing, just been working out the what-works-with-what matrix in the touchscreen + X mist Jan 14 17:35:30 koen So static kdrive with static tslib would work? Jan 14 17:36:11 why static? Jan 14 17:36:35 Just to get rid of any deps that armedslack might lack Jan 14 17:37:02 slackware lack zlib and freetype? Jan 14 17:37:04 wow Jan 14 17:37:15 didn't thought is was that ancient ;) Jan 14 17:37:30 doh :D Jan 14 17:37:41 slackware is stoneage. I run it on my servers so I know, first hand :) Jan 14 17:37:45 the dependency list: http://www.angstrom-distribution.org/repo/?action=details&pnm=xserver-kdrive-fbdev Jan 14 17:39:02 brb Jan 14 17:39:31 RP: ah, right, thanks Jan 14 17:39:36 Kristoffer: http://www.embedded.com/story/OEG20020529S0046 Jan 14 17:39:39 oops Jan 14 17:39:49 I'm building qt-x11-free at the moment, we'll see what happens after that Jan 14 17:42:58 pb_: The problem is we finally wrote out the explicit depends at the task level. package.bbclass says do_package[deptask] = do_package which you have to do for the shlibs code. debian.bbclass requires do_package_write[rdeptask] = "do_package" and when you combine the two it can give circular deps. Jan 14 17:43:34 (deptask is DEPENDS, rdeptask is RDEPENDS) Jan 14 17:44:18 okay Jan 14 18:02:17 * RP -> food Jan 14 18:11:11 psokolovsky: What do you think about my proposal in http://bugs.openembedded.org/show_bug.cgi?id=1788 ? Jan 14 18:11:39 Laibsch: *R*DEPENDS Jan 14 18:13:15 Laibsch, just replied. yep, R ;-) Jan 14 18:13:42 iirc we decided against that, but my memory of that is a bit fuzzy Jan 14 18:13:46 koen: Good point. Jan 14 18:14:26 koen: Why would you not correct the dependency to correctly spell out the current situation? Jan 14 18:14:38 I cannot really see any reason for that. Jan 14 18:15:02 bash is fairly big Jan 14 18:15:22 ldd is installed only upon the users wish Jan 14 18:15:44 And if it does not work otherwise there is no use not to declare that dependency. Jan 14 18:15:51 does it require bash or just !busybox? Jan 14 18:15:57 Of course it would be nice to remove the bashiims from ldd. Jan 14 18:16:01 But until then... Jan 14 18:16:03 Applying patch kernel-keymap.patch Jan 14 18:16:03 patching file qte/src/kernel/kernelkeyboard.cpp Jan 14 18:16:03 can't find file to patch at input line 711 Jan 14 18:16:03 Perhaps you used the wrong -p or --strip option? Jan 14 18:16:03 The text leading up to this was: Jan 14 18:16:04 -------------------------- Jan 14 18:16:06 |Index: src/kernel/qkeyboard_qws.cpp Jan 14 18:16:13 Kristoffer: mtn pull ; mtn up Jan 14 18:16:21 Kristoffer: and curse opie Jan 14 18:16:24 Laibsch: please pastebin the whole ldd 'script' don't have a linux in front of me Jan 14 18:16:28 :D Jan 14 18:16:35 zecke: just a moment Jan 14 18:16:48 why was that patch refreshed anyway? Jan 14 18:16:59 zecke: What patch? Jan 14 18:17:08 kernek-keymap.patch Jan 14 18:17:26 zecke: because patch/e can do direct framebuffer access now to make it faster Jan 14 18:17:59 anyway back to that stupid assignment Jan 14 18:23:05 zecke: http://www.pastebin.ca/316545 Jan 14 18:23:43 zecke: I made a change in the first line and replaced bash with sh for testing purposes. Jan 14 18:23:54 Sorry, but I forgot to revert that. Jan 14 18:39:11 philippe: did you build angstrom gpe image for h3800? Jan 14 18:42:54 univac: yep. But it does not boot in graphical mode yet... :-( Jan 14 18:43:36 philippe, was that with 2.6 kernel? Jan 14 18:43:49 And I have been very busy last week and even had to go to work today... Jan 14 18:43:54 psokolovsky: yep it was. Jan 14 18:44:06 philippe: why? Jan 14 18:44:12 Seems that 3800 support is pretty broken in that one. I fixed some build errors... Jan 14 18:44:25 philippe, which was that version and what kernel tree you used? Jan 14 18:44:49 univac: well we had a threew day design meeting, and I had to catch up on the other things that happened during the week Jan 14 18:45:30 psokolovsky: linux-handhelds-2.6.16-hh8 Jan 14 18:45:57 philippe, so, can you please post patches you made to make it build/boot? Jan 14 18:46:18 psokolovsky: I will when I get the framebuffer running. Jan 14 18:46:36 i'm having trouble installing OE onto a Kubuntu box... can anyone point me in the right direction? The instructions on the wiki dont appear to work on the Dapper verison Jan 14 18:46:44 psokolovsky: you wat develp 2.6 for h3800? ;] Jan 14 18:46:47 want Jan 14 18:47:47 univac, philppe: I care about entire hh.org tree. patches which are not submitted tend to be lost, and even partial patches are helpful. Jan 14 18:48:25 Is there a reason we use rather old versions of m4, autotools and the like? Jan 14 18:49:22 philippe: you used latest 2.6 cvs gg tree? Jan 14 18:53:36 packages/m4/m4-native_1.4.2.bb does not seem to be specific to $PV. Would it make sense to move the file to packages/m4/m4-native.bb? Jan 14 18:53:48 <[cc]smart> does SLUGIMAGE_SVN_REPO ?= "http... make sense or should it be svn:// ? Jan 14 18:58:05 ok psokolovsky I'll post what I have on the kernel list later on... Jan 14 18:58:40 philippe, thanks, and at your convenience, of course! Jan 14 18:59:17 philippe: you used latest 2.6 cvs hh tree? Jan 14 18:59:44 univac: nope. I might try later. Jan 14 19:07:11 [cc]smart: svn://foo.bbar/baz;proto=http Jan 14 19:08:39 Oh lovely, kdrive seems to build perfectly :P Applying patch kernel-keymap.patch Jan 14 19:08:40 patching file qte/src/kernel/kernelkeyboard.cpp Jan 14 19:08:40 can't find file to patch at input line 711 Jan 14 19:08:40 Perhaps you used the wrong -p or --strip option? Jan 14 19:08:40 The text leading up to this was: Jan 14 19:08:40 -------------------------- Jan 14 19:08:42 |Index: src/kernel/qkeyboard_qws.cpp Jan 14 19:08:44 doh Jan 14 19:08:51 vtable for __cxxabiv1::__si_class_type_info' Jan 14 19:09:12 Kristoffer: any dieas on Konqueror for 720? Jan 14 19:09:20 its not in the image Jan 14 19:09:36 Im getting there Jan 14 19:11:09 Kristoffer, I build it w/o issues. Jan 14 19:11:11 was it just missed off or is there an error? Jan 14 19:13:17 Kristoffer: you have the wrong provider for virtual/libsdl Jan 14 19:13:40 Kristoffer: and hence it drags in all that broken opie crap Jan 14 19:14:13 heh. i indeed wondered how kdrive and qte relate ;-) Jan 14 19:14:29 Kristoffer: PREFERRED_PROVIDER_virtual/libsdl ?= "libsdl-x11" Jan 14 19:16:40 koen, doh, big thx Jan 14 19:17:06 <[cc]smart> koen: would this mean that the current packages/slugimage/slugimage.bb is buggy ? Jan 14 19:17:15 koen, libsdl is set as you've stated.. Jan 14 19:17:43 [cc]smart: I guess so, it could also by a clever hack to bypass the svn fetcher Jan 14 19:17:59 Kristoffer, so what do you build, after all? Jan 14 19:18:30 <[cc]smart> i get svn: REPORT request failed on '/svnroot/slugimage/!svn/bc/101/trunk' - svn: REPORT of '/svnroot/slugimage/!svn/bc/101/trunk': 400 Bad Request (http://svn.nslu2-linux.org) Jan 14 19:18:43 psokolovsky_, trying to build xserver-kdrive used to work fine Jan 14 19:19:13 maybe need to update my qt version to 2.3.12 Jan 14 19:19:22 since its complaining about unknown reference Jan 14 19:19:46 Kristoffer, well, building xserver for sure whouldn't pull qte, so you still might have some virtual/* set incorrectly. Jan 14 19:20:02 psokolovsky_, but it was set correct right at start Jan 14 19:20:05 I didnt have to change it Jan 14 19:20:19 psokolovsky_: kdrive builds xsdl and thus depends on virtual/libsdl Jan 14 19:20:44 03koen 07org.oe.dev * r6e924857... 10/ (6 files in 4 dirs): Jan 14 19:20:44 libsql-qpe: move it to packages/obsolete. Jan 14 19:20:44 The main reason for this is the thread on the mailinglist, the secondary reason is the the opie people decided to b0rk their patches so they don't apply anymore. Jan 14 19:20:52 Kristoffer, well, maybe you have override somewhere else... it just shouldn't build qte Jan 14 19:21:46 koen, that's not going to help such cases, lol. Jan 14 19:22:20 psokolovsky_, you were quite correct, it was overriden in local.conf Jan 14 19:23:36 and damn, I missed pnum=0 for kernel-keymap.patch for zauruses. damn. then report on maillist was totally out of hand... Jan 14 19:24:52 likewise: I compiled icemon for edgy Jan 14 19:25:05 Laibsch: Gives some really nice output indeed. Jan 14 19:26:28 Laibsch: and, it confirms the distributed work? Jan 14 19:26:38 likewise: But my helper machine is not really being loaded. From the scheduler log it looks like it is getting almost all distributed jobs. Jan 14 19:27:18 Laibsch: hmm, yes I remember that the last few revisions of the icecc code all had modifications to the load metric calc. Jan 14 19:27:30 But the helper machine is idling most of the time nonetheless. Jan 14 19:27:38 koen, Btw, I want to setup OPIE page in Angstrom wiki, something like OpieWithAngstrom. pls let me know if you have some concerns ;-) Jan 14 19:28:17 likewise: Are you sure that all tasks from OE are put into the icecc queue? Jan 14 19:28:48 Laibsch: at least the non-native and non-cross ones. Jan 14 19:29:24 Ok, I will keep an eye on it. Jan 14 19:29:54 The locale-generation for angstrom which has now taken more than 24 hours is not being distributed :-( Jan 14 19:31:51 Laibsch, I have a local patch which limits locales generated to a given set. That's what most devel (vs release managers) need. One sweet day, will polish and submit ;-) Jan 14 19:32:48 polsza ;p Jan 14 19:33:23 psokolovsky_: Please do. Is it complicated? Does it make sense to "release early and often"? ;-) Jan 14 19:33:59 Laibsch, curerntly it's: Jan 14 19:34:00 supported = ["en_GB.UTF-8 UTF-8\n", "en_GB ISO-8859-1\n"] Jan 14 19:34:40 but it really should be optional conf var, and list should have minimal en, de, fr. all utf-8 Jan 14 19:35:05 That line is where? Jan 14 19:38:05 psokolovsky_: Where would I need to configure this? Jan 14 19:38:08 Laibsch, http://pastebin.ca/316592 Jan 14 19:38:44 again, put de and fr there too. that's why I said "needs polishing" ;-) Jan 14 19:43:41 the patch does not apply to current .dev Jan 14 19:44:33 psokolovsky_: But I think I found where the line needs to go. Jan 14 19:53:04 zecke: Any insights on the bashisms? Jan 14 19:53:33 03pfalcon 07org.oe.dev * r1516e610... 10/ (1 packages/qte/qte-common_2.3.10.inc): qte 2.3.10: Finish converting kernel-keymap.patch to -p1 format. Jan 14 19:54:37 philippe: could ypu upload defconfig for 2.6 h3800? Jan 14 19:55:39 nope Jan 14 19:57:18 k Jan 14 19:58:51 Any reason not to commit http://rafb.net/p/WVjMZ763.html as suggested in http://bugs.openembedded.org/show_bug.cgi?id=837 (update m4 to 1.4.4, later versions need an update to autoconf as well, but autoconf 2.61 bb is marked with preference -1) Jan 14 19:59:01 Or is this something that is better discussed on the ML Jan 14 19:59:03 ? Jan 14 20:08:04 03ccsmart 07org.oe.dev * r308555eb... 10/ (1 packages/libdaemon/libdaemon_0.10.bb): libdaemon: Add missing header file daemon.h to package. Jan 14 20:08:04 03ccsmart 07org.oe.dev * r4a15fa8e... 10/ (1 packages/libg15/libg15_1.2.1.bb): libg15: Add library staging. Jan 14 20:08:07 03ccsmart 07org.oe.dev * rac325912... 10/ (1 packages/g15daemon packages/g15daemon/g15daemon_1.2.6a.bb): g15daemon: Added new package. Jan 14 20:18:42 <[cc]smart> do we know who made the lcdproc recipe ? Jan 14 20:26:52 [cc]smart: you can check the author cert of the rev it first appeared in... Jan 14 20:29:39 <[cc]smart> JustinP: Now that you said it... Jan 14 20:36:53 * likewise cheers YAY! Got the 5-pin EFIKA Geek Port working as an extra UART. Jan 14 20:37:27 ~hail tnt_ Jan 14 20:37:33 * ibot bows down to tnt_ and chants, "I'M NOT WORTHY!!" Jan 14 20:38:37 likewise: cool Jan 14 20:40:23 likewise: the oe images are working? Jan 14 20:41:25 * v8jlene is back Jan 14 20:42:49 koen: oe images for efika you mean? Jan 14 20:42:51 morning all Jan 14 20:42:55 likewise: yes Jan 14 20:43:00 hey v8jlene Jan 14 20:43:08 koen: yes, I have been working with EFIKA/OE for a few weeks now, no problem. Jan 14 20:43:14 [cc]smart: I know I did some work on the lcdproc recipe Jan 14 20:43:19 [cc]smart: why are you asking? Jan 14 20:43:24 koen: angstrom of course. ;-) Jan 14 20:44:35 koen: How is the usermanual on the website being generated? I want to see if it's possible to make it use section_id's for the URL's instead of chapter/section numbers... so the links remain constant as things are added/removed. Jan 14 20:45:06 is it just docbook2html? Jan 14 20:45:09 koen, I have untared an angstrom/efika image on my efika hd and booted from that Jan 14 20:45:14 koen: angstrom-bootstrap rather. Some packages don't build yet for efika. Jan 14 20:45:29 v8jlene: http://rafb.net/p/V2zJjy69.html Jan 14 20:45:44 v8jlene: and 'make xhtml' Jan 14 20:45:49 Crofton__: did you order yet? I might want to change one unit to a 60W, if not. Jan 14 20:45:55 not yet :) Jan 14 20:46:10 Crofton__: I just heard on gentoo-efika that on the 120W version, the 12 V power rail is pass-through. Jan 14 20:46:12 3 120, one 60? Jan 14 20:46:20 Crofton__: yup, tnx Jan 14 20:46:22 hm Jan 14 20:46:32 so you need a reasonable 12v? Jan 14 20:46:55 Crofton__: Yes. I don't trust my car's 12 V to good for that. Jan 14 20:47:17 car 12v is scary :) Jan 14 20:47:33 Crofton__: like you mentioned earlier, the 60W version copes with 6-18 Volts input, so everything is regulated. Jan 14 20:47:38 yeah Jan 14 20:47:55 for car use they sell a much fancier looking unit Jan 14 20:48:04 60W should be ample for a non-graphics, 2.5" HD or CF EFIKA unit. Jan 14 20:48:11 car 12volt can have scary spikes Jan 14 20:48:18 yeah Jan 14 20:48:29 I will likely get 60W Jan 14 20:48:43 likewise: seen these? http://www.mini-box.com/M1-ATX-90w-Intelligent-Automotive-DC-DC-Power-Supply?category=13 Jan 14 20:48:57 yeah Jan 14 20:49:08 (or the M2 160w ones) Jan 14 20:49:33 I need to run Jan 14 20:49:40 I'll order sometime next week Jan 14 20:49:40 NAiL: yes, but I think they are overkill for me. My car stereo provides the 12V Jan 14 20:49:47 Crofton__: thanks, CU! Jan 14 20:50:03 thats right, and you are pretty sure the 12v has been cleane up some? Jan 14 20:50:07 l8r Jan 14 20:54:35 <[cc]smart> NAiL: I need to add one more package, libg15render. After that i think i have all requirements for lcdproc 0.5.1 which, based on the new packages, supports the LCD Display of the G15 keboard. I had a brief glimpse on the lcdproc.bb and have a hard time getting it. so if you could help me out updating lcdproc to 0.5.1 adding the g15 driver once i added libg16render that would be nice. Jan 14 20:56:08 XorA|sleep: http://www.advogato.org/person/cinamod/diary.html?start=132 :_ Jan 14 20:56:10 ehm Jan 14 20:56:11 :) Jan 14 21:15:28 re Jan 14 21:16:50 mickey|berlin: hey Jan 14 21:32:15 mickey|berlin: hi! Jan 14 21:32:24 koen: Ok, I've got it generating URL's based on the section id. I'll update all the section id's and then commit the change to the makefile. Jan 14 21:32:50 That'll make all the URL's in the user manual remain constant as things are added/removed (as long as the section id's remain the same) Jan 14 21:33:03 good night Jan 14 21:34:20 mickey|berlin: how's the codesprint going? Jan 14 21:34:32 mickey|berlin: or have you resorted to getting sean some berlin beers? Jan 14 21:38:56 koen: We already do have codesprints? Jan 14 21:39:27 Laibsch: no, but openmoko does Jan 14 21:53:42 koen: heya, we have spent the whole day with structure, packaging, roadmap discussions Jan 14 21:54:03 packaging sucks Jan 14 21:54:09 that's why 90% gets it wrong Jan 14 21:54:10 koen: Sean now is in the hotel and LaF0rge and me are working on a dialer demo Jan 14 21:54:35 heh Jan 14 21:54:41 not software packaging :) Jan 14 21:54:46 ah Jan 14 21:54:57 same goes for hardware packaging Jan 14 21:55:28 "let's put every item in it's own plastic baggie" Jan 14 21:55:37 hehe Jan 14 21:55:51 "lets use extra durable plastic so people have to use a knife" Jan 14 21:56:15 does any of you own an ipod? Jan 14 21:56:19 nope Jan 14 21:57:19 03lenehan 07org.oe.documentation * reeca2f50... 10/ (1 usermanual/chapters/recipes.xml usermanual/usermanual.xml): Jan 14 21:57:19 usermanual: Prepare to change from using numbers for chapater and section Jan 14 21:57:19 URL's to using names. This required id's to be added to all of the sections and Jan 14 21:57:19 chapters. I also renamed the id's in the recipes section from bb_ to Jan 14 21:57:21 recipes_. This will allow for stable URLs into the user manual on the web Jan 14 21:57:23 site. The change to the makefile to activate this is not yet commited. Jan 14 22:00:18 koen: hmm, your php to modify the URLs will break if I commit the change to use names. You can test it out by adding "--param use.id.as.filename 1" to the xsltproc calls in the Makefile. Jan 14 22:00:40 In theory it should be possible to make xsltproc generate correct url's by passing in the base url. Jan 14 22:00:44 hi Jan 14 22:01:05 I'll look at doing this as well when I get back later today. Jan 14 22:01:08 koen: thx for lkml link Jan 14 22:01:13 hi hrw Jan 14 22:01:30 RP: how do we reply to Cyril HrubiÅ¡ mail? Jan 14 22:01:45 RP: we have 3 kernel trees for collie now... Pavel, Cyril, OE Jan 14 22:01:56 RP: and no real maintainer for this machine at all Jan 14 22:02:00 koen: is qte-mt a plugin? Jan 14 22:02:01 mickey|berlin: check http://dominion.kabel.utwente.nl/koen/images/ipod/ Jan 14 22:02:10 !oebug 1785 Jan 14 22:02:12 * * Bug 1785, Status: NEW, Created: 2007-01-12 14:36 Jan 14 22:02:12 * * bugs.openembedded.org(AT)rolf.leggewie.biz: ERROR: QA Issue: non dev package contains .so: qte-mt Jan 14 22:02:13 * * http://bugs.openembedded.org/show_bug.cgi?id=1785 Jan 14 22:02:29 Laibsch: no idea, just noting that insane.bbclass has a lot of false positives Jan 14 22:03:02 mickey|berlin: note how all extras are hidding with the driver cd: http://dominion.kabel.utwente.nl/koen/images/ipod/IMG_7180.CR2.jpg Jan 14 22:05:43 ya, that packaging sucks Jan 14 22:06:54 no, it rocks Jan 14 22:07:02 it gives you the main device without troubles Jan 14 22:07:20 (the battery is already inserted) Jan 14 22:08:03 koen: and charged a bit? Jan 14 22:08:11 hrw: yep Jan 14 22:08:22 I like to get toy working out-of-box Jan 14 22:08:24 some batteries need a charge Jan 14 22:08:31 otherwise they break Jan 14 22:08:33 li-poly? Jan 14 22:08:37 lion? Jan 14 22:08:46 I forget which ones Jan 14 22:08:47 my k750i was ready to use after I got it in shop Jan 14 22:09:19 I also hate stickers Jan 14 22:09:39 "6 Megapixel!!!!!" "usb!!!!" "vga!!!!!" Jan 14 22:09:39 koen: I thought it was nimh that needed a charge.. Jan 14 22:09:39 so I switched sim from my old nokia, connected headphones, tuned FM radio and go home listening to radio Jan 14 22:09:51 NAbyss: could be, I just use the batteries :) Jan 14 22:10:32 mickey|berlin: did you saw my recent post? Jan 14 22:12:32 blogpost Jan 14 22:13:13 hi mickey|berlin Jan 14 22:15:59 Patch make.patch does not apply (enforce with -f) Jan 14 22:15:59 ERROR: Task 2663 (/home/hrw/devel/oe/org.openembedded.dev/packages/m4/m4-native_1.4.4.bb, do_patch) failed Jan 14 22:16:03 nice... Jan 14 22:16:44 Laibsch: looks like it is not simple 'copy' Jan 14 22:18:27 hrw: Did you copy both native and normal bb? Jan 14 22:18:36 Did you remove the makefile patch? Jan 14 22:18:44 Was a hassle-free build here. Jan 14 22:19:10 hrw: You forgot to remove the patch from SRC_URI Jan 14 22:19:11 ah. forgot to remove path - sorry.. Jan 14 22:19:21 23:50 is bad time for such things Jan 14 22:19:26 hehe Jan 14 22:20:59 NOTE: package m4-native-1.4.4-r0: task do_compile: started Jan 14 22:21:21 NOTE: Running task 24 of 3259 (ID: 2661, /home/hrw/devel/oe/org.openembedded.dev/packages/m4/m4-native_1.4.4.bb, do_install) Jan 14 22:21:53 those 3259 tasks are 'bitbake gpe-image' for h2200/angstrom-2007.1 Jan 14 22:22:30 build is in progress, nothing will interrupt it rather. Jan 14 22:22:58 Laibsch: so if will finish without breaks then it will be a sign that m4 1.4.4 is working ok for us Jan 14 22:23:43 if not then have to see where it broke and decide does it was m4 or not Jan 14 22:23:45 cu Jan 14 22:24:29 hi Jan 14 22:25:59 /win 13 Jan 14 22:26:39 /lose 12 Jan 14 22:27:45 /d'oh Jan 14 22:28:36 This is what happens when you use ssh from a phone :) Jan 14 22:30:03 hey pb_ Jan 14 22:30:50 mickey|berlin: torture is still running, so I suspect it's broken Jan 14 22:31:03 or my devices are really sloooooooooooooooooooooooooooooooooooooooooow Jan 14 22:31:33 *nod* either it's broken or we are supposed to supply a number of iterations Jan 14 22:31:39 i killed it after 4 hours Jan 14 22:32:14 it defaults to level '7' Jan 14 22:32:27 which is 3x84 iters iirc Jan 14 22:33:04 supplying them manually didn't help Jan 14 22:33:08 neither did using the gui Jan 14 22:33:26 did somebody already try to patch a oe 2.6 kernel with those atheros sdio-patches available at http://source.mvista.com/~dsingleton/sdio/2.6.18/ ? Jan 14 22:34:56 03likewise 07org.oe.dev * r85717d4b... 10/ (1 packages/linux/linux-efika/defconfig): linux-efika: Enable CIFS in defconfig. Jan 14 22:37:37 hrw|gone: OK, let me know. Jan 14 22:38:07 caprifisher: which h/w would we need to test this? Jan 14 22:38:13 caprifissher: which h/w would we need to test this? Jan 14 22:38:17 d*rn.. Jan 14 22:38:34 manual autocompletion at work Jan 14 22:39:15 likewise: i have no idea ;) Jan 14 22:39:34 likewise: http://source.mvista.com/~dsingleton/sdio/welcome-to-SDIO-linux Jan 14 22:41:21 likewise: something with an sd slot + sd wifi card Jan 14 22:42:03 zaurus? Jan 14 22:42:09 that should work Jan 14 22:42:15 i have a zaurus c860 and a sd wifi card, so this is the reason why i'm asking ;) Jan 14 22:42:17 iirc rp said they were sdio capable Jan 14 22:44:19 timtimred: hi there, good mor^W evening. Jan 14 22:44:37 03likewise 07org.oe.dev * r1617d522... 10/ (3 files in 3 dirs): linux-efika: Sound driver altered to keep hands of register bits it shouldn't touch. Jan 14 22:47:37 koen: so it would make sense trying it? :) Jan 14 22:48:19 caprifischer: I think so Jan 14 22:51:54 koen: what would be a good documentation on building my own kernel for a newbie like me? Jan 14 22:53:05 building it easy 'bitbake virtual/kernel' Jan 14 22:53:17 'kay Jan 14 22:53:18 after following http://www.openembedded.org/wiki/GettingStarted Jan 14 22:54:51 i already tried this, but i got stuck while the buildscript tried fetching file-native-4.18. seems like the anonymous account at ftp.astron.com is no longer available Jan 14 22:57:35 koen: some news about embedded devroom program @fosdem? Jan 14 22:58:22 greentux: not that I know off Jan 14 22:59:19 i think openmoko will have a talk, and perhaps qtopia... Jan 14 23:15:26 caprifischer: yeah, somethings busted over there. If you change from ftp://ftp.astron.com/pub/file to ftp://ftp.fu-berlin.de/unix/tools/file in packages/file/file_4.18.bb it should get past that fine. Jan 14 23:16:53 v8jlene: thx, this works Jan 14 23:17:04 In light of the do-not-care attitude of dbus people with regards to pid files, should we fix bug 874 as proposed? I have a patch ready. Jan 14 23:17:23 v8jlene: Is the SRC_URI broken? Jan 14 23:17:48 Laibsch: actually it seems like the server is broken :) Jan 14 23:18:30 Laibsch: The web site it points at looks to be broken. Not sure if that's just a temp thing or if it's busted for good. The ftp.fu-berlin.de server is listed in the freshmeat entry as a mirror site. Jan 14 23:21:57 v8jlene: I was wondering if it should be fixed right away. Jan 14 23:22:13 But looks like it is better to wait and see if things improve. Jan 14 23:22:36 Laibsch: Yeah, I was wondering the same thing... I am currently undecided ;) Jan 14 23:24:09 anyone ever got the problem where a ts works in ts_test, but creates spurious button press/releases under X? Jan 14 23:50:30 mickey|berlin: Do you think bug 747 can be closed? Jan 14 23:50:35 !oebug 747 Jan 14 23:50:36 * * Bug 747, Status: REOPENED, Created: 2006-03-15 19:18 Jan 14 23:50:37 * * emte(AT)labotomy.net: Familiar Rendered Unbuildable on some hosts Jan 14 23:50:38 * * http://bugs.openembedded.org/show_bug.cgi?id=747 Jan 14 23:50:49 AFAIK, OE no longer supports familiar Jan 14 23:54:49 Laibsch: yes, can be closed Jan 14 23:55:13 hmm, the X tslib driver seems to be seriously fucked up here Jan 14 23:58:55 didn't i recall something regarding xcalibrate being broken in some versions? Jan 14 23:59:44 mickey|berlin: in the _git version Jan 14 23:59:55 mickey|berlin: and make sure you don't use -mouse arguments Jan 15 00:01:20 koen: -mouse? where Jan 15 00:01:37 Xfbdev -foo basr -baz -mouse Jan 15 00:02:05 hmm i have libxcalibrate0+git Jan 15 00:02:08 that one is broken= Jan 15 00:02:09 ? Jan 15 00:02:43 that one works Jan 15 00:02:52 kdrive_git was (is?) broken Jan 15 00:03:29 hmm Jan 15 00:03:35 i have 7.1-1.1.0 here Jan 15 00:03:49 that one should work Jan 15 00:03:55 hmm darn Jan 15 00:03:58 it seems to work over here Jan 15 00:04:08 mickey|berlin: do you use cpufreq? Jan 15 00:04:20 libxcalibrate0+git20060814 here Jan 15 00:04:27 I had troubles with the ts on the h2200 and hx4700 using cpufreq Jan 15 00:04:53 what kind of spurious events are you seeing? Jan 15 00:05:13 a) missing button press events Jan 15 00:05:19 b) double press & release events Jan 15 00:05:32 sometimes totally off xy values Jan 15 00:05:53 while ts_test works fine Jan 15 00:06:26 I get a) and b) Jan 15 00:06:36 b) messes up xkbd Jan 15 00:06:59 symptons: capitals ('A') instead of lowercase ('a') Jan 15 00:07:07 no keyboard here at all :) Jan 15 00:07:27 there should be a debounce plugin iirc Jan 15 00:08:42 the problem is it's fine in the kernel and in tslib Jan 15 00:08:58 only X doesn't behave Jan 15 00:09:14 does ts_test use the plugins specified? Jan 15 00:09:48 ah, right Jan 15 00:09:53 03koen 07org.oe.dev * rdbd927aa... 10/ (1 packages/wv packages/wv/wv_1.2.0.bb): wv: add 1.2.0, from Poky, ran through oe-stylize.py Jan 15 00:10:06 ts_test uses module_raw , but not module Jan 15 00:11:45 that cost me a few days I'll never get back Jan 15 00:12:09 ts_test uses ts_read here Jan 15 00:12:12 ts_test.c Jan 15 00:12:34 or was it ts_calibrate? Jan 15 00:12:36 bleh Jan 15 00:12:44 * koen can't think after 1 am Jan 15 00:12:55 hehe Jan 15 00:13:11 and I suspect I won't finish this book either tonight Jan 15 00:13:16 ~700 pages left Jan 15 00:15:27 I get KMapNotify events during mouse presses !?? Jan 15 00:15:33 is that a symptom as well? Jan 15 00:15:41 KeyMapNotifyEvents, even Jan 15 00:15:46 (in xev) Jan 15 00:15:55 no idea, sorry Jan 15 00:21:00 * likewise should stop as well, don't tell bugs from bears anymore. Jan 15 00:21:48 bears on the road? Jan 15 00:24:12 koen: hope not, I have motorcycle exams in 10 hours :-) Jan 15 00:24:44 exam as in verhicle control tests (drive figure-eights and stuff). Jan 15 00:24:59 good luck with that! Jan 15 00:25:18 thanks, I'm already feeling a bit nervous for it. Jan 15 00:25:42 taking a nap now, see you. Jan 15 00:25:58 cu & good luck ;) Jan 15 00:29:03 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa Jan 15 00:31:00 koen: is it possible that the X server opens /dev/mice even without me sying -mouse ? Jan 15 00:31:16 yes Jan 15 00:31:24 I forgot to say that Jan 15 00:31:39 kdrive tries to autoconfigure itself Jan 15 00:32:15 aaaaah Jan 15 00:32:19 how do i stop that? Jan 15 00:32:23 mickey|berlin: http://bugs.openembedded.org/show_bug.cgi?id=1694 comment #9 Jan 15 00:32:28 * mickey|berlin dances a jig Jan 15 00:32:34 I remember reading about that some time ago Jan 15 00:32:55 X's screwy input implementation Jan 15 00:33:39 thanks guys Jan 15 00:33:44 removing the nodes did it Jan 15 00:33:45 mickey|berlin: it has been fixed in later kdrives with the input-hotplug branch, but the easiest way is not to have /dev/input/mice Jan 15 00:34:19 k, thanks Jan 15 00:34:40 first big thing for tonite fixed :D Jan 15 00:35:28 just 6 man hours later Jan 15 00:35:29 heh Jan 15 00:35:35 * mickey|berlin imagines the bill Jan 15 00:36:26 * koen gives mickeyl a discount Jan 15 00:36:45 ;) Jan 15 00:38:12 so the reason why Zaurus and iPAQ works is that they have patched udev configs? Jan 15 00:38:32 btw., we will announce a clear roadmap and more details very soon Jan 15 00:38:38 working on that statement tomorrow Jan 15 00:39:14 mickey|berlin: kernel configs iirc Jan 15 00:39:27 I suspect you have mousedev or something loaded Jan 15 00:40:47 mickey|berlin: and screenshots! Jan 15 01:00:05 thanks guys Jan 15 01:00:11 need to sleep Jan 15 01:00:12 g'nihgt Jan 15 01:00:34 for ehat? Jan 15 01:00:37 what? Jan 15 01:06:38 koen: can I get you to test a change to your php for the user manual? This change will modify relative URLs but leave absolute ones alone... the last two str_replace lines should be replaced with the following line: Jan 15 01:06:40 $bodytext = preg_replace('/href="(?!https?:\/\/|ftp:\/\/|mailto:|news:|javascript:|#)(\S*?)"/', 'href="/user-manual&dpage=\\1"', $bodytext) Jan 15 01:07:10 and make sure it all still works... I've tested locally but my php and server layout is abit different. Jan 15 01:09:16 it doesn't seem to work :( Jan 15 01:10:10 What version of php is on there? Jan 15 01:10:12 it seems to make all the test go away Jan 15 01:11:32 v8jlene: 4.3.10-16 Jan 15 01:15:09 Hmm, I have 5.x, but pcre looks fine from 4.1 onwards. Anything in the web server error logs? (It works fine here... really ;) Jan 15 01:16:11 Oh, there should be a ; on the end of the line... my cut 'n paste missed that because the line is wider than my screen ;( Jan 15 01:16:51 D'oh Jan 15 01:16:53 now it works Jan 15 01:18:45 Cool, external links still look ok as well. So I should be able to enable naming all the user manual pages by id now instead of chapter/section so the URL's remain stable. Jan 15 01:21:58 nice, thank's for figuring that out Jan 15 01:30:10 03lenehan 07org.oe.documentation * re5a3dda4... 10/ (1 usermanual/Makefile): Jan 15 01:30:10 usermanual: Modify the makefile so that "make xhtml" will generate filenames Jan 15 01:30:10 using section/chapter id's rather than using section/chapter numbers. This Jan 15 01:30:10 should result in the URL's pointint into the manual remaing constant even Jan 15 01:30:10 when chapters and sections are added and/or deleted. **** ENDING LOGGING AT Mon Jan 15 02:59:58 2007