**** BEGIN LOGGING AT Mon May 07 02:59:56 2007 May 07 03:45:48 psokolovsky_: koen's patch don't work. again on qemu segfaults =( May 07 03:47:28 victor_rx1950: segafaults? but #2054 has nothing to do with segfaults. what you asked for, you were given. May 07 03:48:59 psokolovsky_: 2054 if not full May 07 03:49:52 psokolovsky_: on line higher there is "qemu: uncaught target signal 11 (Segmentation fault) - exiting" May 07 03:52:33 psokolovsky_: i added full output to 2054 May 07 03:55:07 !oebug 2054 May 07 03:55:08 * * Bug 2054, Status: REOPENED, Created: 2007-03-27 21:13 May 07 03:55:09 * * r.farina(AT)adelphia.net: glibc-2.5-r5 May 07 03:55:10 * * http://bugs.openembedded.org/show_bug.cgi?id=2054 May 07 03:56:03 victor_rx1950: and how your problem relates to #2054? May 07 03:56:18 psokolovsky_: yes May 07 03:56:24 psokolovsky_: whis is same problem May 07 03:56:28 this* May 07 03:56:42 * victor_rx1950 oops... forgot about study ^)) May 07 03:56:47 * victor_rx1950 fly study May 07 03:57:19 victor_rx1950: also, all problems related to qemu are well-known. search bugtracker, MLs, etc. May 07 03:57:49 psokolovsky_: ok, laterm if i don't find my problem i commit new report later May 07 04:03:55 ok, pivotboot-image or gpe-image for a 3200? May 07 05:24:36 morning all May 07 05:55:07 anybody around? May 07 05:55:11 XorA|gone/ May 07 07:26:21 good morning all May 07 07:43:15 good morning May 07 07:45:30 hey pH5 May 07 07:46:05 hey koen May 07 07:58:57 morning May 07 08:00:28 hey XorA May 07 08:16:47 bleh, annoyance threshold reached May 07 08:16:53 * koen updates org.oe.doc May 07 08:17:35 03koen 07org.oe.documentation * r8b5f0b03... 10/ (3 files in 3 dirs): usermanual: image and rootfs updates May 07 08:48:26 test May 07 08:48:44 failed May 07 08:49:03 :) May 07 08:49:22 changing the router here and testing if everything works May 07 08:49:56 or if i am going to fire someone today :) May 07 09:03:48 morning May 07 09:04:10 hey hrw May 07 09:06:55 hey hrw May 07 09:07:18 * hrw need 5.25" usb 2.0 hub May 07 09:08:17 hrw: I'm currently building with http://bugs.openembedded.org/attachment.cgi?id=1433&action=edit May 07 09:08:24 hrw: although I still think it's wrong May 07 09:09:51 it is May 07 09:10:09 my task-base patch is better choice but also not best May 07 09:10:12 I'd rather have angstrom--image depend on task-wifi May 07 09:10:28 with appropriate checks for cf and pci May 07 09:10:48 koen: task-distro-base? May 07 09:10:59 not really May 07 09:11:13 or any other name for task-base-with-wifi-bt-for-pcmcia-usbhost-devices May 07 09:11:14 because for e.g. nslu2 you don't want it May 07 09:11:48 I know May 07 09:11:57 it falls somewhere in between distro and image May 07 09:12:02 DEFAULT_TASK_PROVIDER idea sounds nice May 07 09:12:37 but DEFAULT_TASK_PROVIDER_nslu2 = "task-boot task-base-usbhost task-base-pci" sounds not good May 07 09:12:49 no May 07 09:13:00 DEFAULT_TASK_PROVIDER may only have a single entry May 07 09:14:36 thats why I want other name for task-base-with-wifi-bt-for-pcmcia-usbhost-devices May 07 09:18:33 hi schurig May 07 09:22:06 hrw: hi! May 07 09:34:58 hmm. to buy 3rd usb hub or not to buy May 07 10:23:33 hm new wlan stack in 2.6.22 May 07 10:25:08 good monring May 07 10:25:23 he florian :) May 07 10:29:03 re May 07 10:29:35 he koen May 07 10:31:10 hey woglinde May 07 11:10:02 ~seen gerwinin May 07 11:10:08 gerwinin was last seen on IRC in channel #oe, 4d 22h 53m 56s ago, saying: 'lfaistos true , but I am only going to develop linux drivers , the windows they need to do themselves , me and windows are incompatible for one of other reason'. May 07 11:11:33 Does a line like PACKAGES = "${PN}" in a bb file really do anything? May 07 11:11:40 yes May 07 11:11:50 it disabled -dev, -doc and -dbg packages May 07 11:11:51 OK, I'll keep it in then May 07 11:12:10 Maybe I should keep it out, then. May 07 11:12:42 PACKAGES = in non-task recipes are 99% of the time bogus May 07 11:13:01 PACKAGES += and PACKAGES =+ are ok May 07 11:13:14 OK, thanks for the explanation May 07 11:35:26 03xora 07org.oe.dev * rfff65365... 10/ (18 files in 5 dirs): packages/mozilla/* : removing old broken detritous May 07 11:35:31 03xora 07org.oe.dev * r932420df... 10/ (14 files in 4 dirs): May 07 11:35:31 firefox_1.0.7.bb : add patch from May 07 11:35:31 http://article.gmane.org/gmane.linux.debian.ports.arm/4082 by Lennert Buytenhek May 07 11:35:31 fix a small QA issue May 07 11:35:31 firefox_2.0.0.3 : new version with same patch as above added. May 07 11:44:07 XorA: Does it compile now? May 07 11:45:20 compile: yes work: no May 07 11:46:11 * XorA hasnt had a problem compiling it May 07 11:46:57 someone with a fast ARM want to volunteer to gdb the crash? May 07 11:47:06 it cripples me C3200 trying to May 07 11:48:02 tried with gdb-server? May 07 11:48:28 adding a ton of swapped helped me with smaller programs like dbus May 07 11:51:28 XorA: qemu to the rescue? May 07 11:51:34 Or will that not do it? May 07 11:51:51 qemu/arm is not fast May 07 11:52:14 XorA: my netbook can't run angstrom due to that &*@!#&$*@($&@* 2.6.9 kernel May 07 11:52:22 ~seen lamikr May 07 11:52:26 lamikr was last seen on IRC in channel #htc-linux, 43d 13h 36m 18s ago, saying: 'Kevin2: It is basically a wince frontend from the 2 phase bootloader for omap where the 2:nd part is os independent bootstrap. Sources for both are in svn://fuzzyneural.net/h6300'. May 07 11:55:39 Why is "bitbake fbvncserver" telling me there is no such provider when the bb file is sitting right there? Does bitbake pick it up for you? May 07 11:58:10 -DDDD May 07 12:00:09 hrw: What is so funny? May 07 12:00:48 Laibsch: that you didnt read the manual :-) May 07 12:02:25 XorA: I read tons of manuals. OE manual, fbvncserver manual, "linux for beginners manual", which one? May 07 12:02:42 bitbake one, -D command in particular :-D May 07 12:02:42 Enlighten me and don't let me die stupid. May 07 12:03:08 bitbake -DDDD fbvncserver May 07 12:03:11 Oh, I misread it for ": - D D D D" May 07 12:03:18 less spaces May 07 12:06:40 Laibsch: then chose a sane font for your IRC client May 07 12:07:29 You never read anything wrong? May 07 12:07:32 03koen 07org.oe.dev * r63713e44... 10/ (1 packages/angstrom/angstrom-x11-image.bb): angstrom-x11-image: add task-base-wifi May 07 12:12:47 Laibsch: Never. Ever. Really!! And I never lie, BTW :-) May 07 12:13:23 schurig: :-D May 07 12:13:40 hrw: Thanks for the cluestick. I think I might be able to fix this myself now May 07 12:13:51 Seems to be an outdated dependency May 07 12:23:48 I'm learning oe+bb and working on fuse for the last couple of days May 07 12:24:12 I thought I got a working recipe for fuse-2.6.5 and sshfs-fuse-1.7 May 07 12:24:31 but I do have couple of "QA" question though May 07 12:24:45 one is fuse-module May 07 12:25:33 do we still need it, since we've already had fuse-module in the kernel, at least for tosa and couple of others May 07 12:25:53 only for 2.4 and 2.6.ancient based machines May 07 12:26:26 ok, i'll keep fuse-module and switch on --enable-kernel-module May 07 12:27:24 2) do we need fuse-util in a seperate package? there're inconsistencies in between the old recipies 2.5.3 and 2.6.0/2.6.3 May 07 12:27:26 --enable-kernel-module only in fuse-module_2.6.5.bb, right? May 07 12:27:32 yes May 07 12:28:14 only in fuse-module_2.6.5.bb and --disable-kernel-module in fuse_2.6.5.bb May 07 12:28:27 sounds ok May 07 12:29:28 koen: how about fuse-util? separate package or combined in fuse? May 07 12:29:31 xjqian: the 2.5.3 recipe is correct, the 2.6.3 one isn't May 07 12:29:53 koen: I see, I will follow the 2.5.3 recipe, then May 07 12:30:27 xjqian: it will create 2 packages: libfuse and fuse-utils May 07 12:30:45 so for a gui fuse add you don't fuse-utils, which saves some space May 07 12:31:04 koen: and libulockmgr May 07 12:31:41 ? May 07 12:32:16 koen: I see what you mean. I should put libulockmgr in fuse. May 07 12:33:03 !seen gfeller May 07 12:33:04 Laibsch, I don't remember seeing gfeller. May 07 12:33:37 ~seen pgfeller May 07 12:34:02 pgfeller was last seen on IRC in channel #oe, 3d 4h 47m 30s ago, saying: 'likewise: good morning'. May 07 12:34:52 koen: what about the patch for fuse_2.6.3? I've not used fuse, so I don't know if it's relevant or not. May 07 12:35:18 Is it safe to assume from the path of http://gnu.ethz.ch/debian-lzma/pool/main/p/plucker/ that plucker can be had with lzma support as well? Is lzma generally the best read-only compressor in FOSS? May 07 12:37:48 fbvncserver not being picked up was due to a dependency on libvncserver-0.6. This seems to have been dropped from OE, 0.8 is the only available version. There is a comment in the fbvncserver file that libvnc 0.7.1 introduces problems. And these seem to prevail in 0.8. At least, relaxing the dependency for libvnc from 0.6 to "just gimme any version" ends up with a compilation error for fbvncserver. May 07 12:38:16 Is it better to leave the situation as is or restore version 0.6 for libvnc? May 07 12:38:46 Laibsch, maybe fix the build and see if it helps? May 07 12:39:03 polyonymous: You are overestimating my skill level May 07 12:39:18 The two propositions are what I can do. May 07 12:39:21 Laibsch, well, I didn't even mean you, just proposing a solution. May 07 12:39:41 Wanna take a stab at it? May 07 12:39:46 are you talking about vncserver that is supposed to work on opie? May 07 12:40:08 If yes, then I think I'd give it a shot, but I need finish my build first. May 07 12:40:27 You know, I'm a bit behind due to my yesterday perl problems :) May 07 12:42:00 fbvncserver May 07 12:42:24 I take it the fb part should allow it to work for opie... May 07 12:42:35 You have to relax the dependencies. Take a look at the bb file and uncomment the third line or so. Read the comment below May 07 12:42:57 polyonymous: Maybe. Patrick Feller is working on updating this stuff. May 07 12:43:05 I am at task ~ 2700 of ~5700 :) May 07 12:43:31 polyonymous: I am sure your computer supports more than one OE compilation at the same time :-) May 07 12:44:06 hmmm... does it? I mean, bitbake has no problems with it? May 07 12:44:24 polyonymous: Usually not. May 07 12:44:30 okay, I'll try then May 07 12:44:41 I run multiple instances all the time May 07 12:45:04 The worst that can happen is that two instances try to build the same stuff and get in each others way. May 07 12:45:18 A -c rebuild of the last package comes to the rescue May 07 12:45:26 Laibsch: libvncserver's API changed from 0.6 to 0,8 May 07 12:45:45 Laibsch: one of my own VNC servers was built for 0.6 and didn't compile with 0.8 May 07 12:46:08 schurig: Any objections to just restoring 0.6, then? At least until this is properly sorted out? May 07 12:46:24 * Laibsch is doing an mtn log to find out the people who last touched this stuff. May 07 12:46:37 Laibsch: AFAIK koen did delete 0.6 several months (>1 year, maybe 2 years) ago May 07 12:46:57 Laibsch: I know this because that broke my own VNC server. After some discussion, I put 0.6 back in, but someone else deleted it again May 07 12:46:59 Yes, I am already inOctober of last year and no real changes yet May 07 12:47:28 Laibsch: however, I don't know if, for example, 0.6 has security problems. For me this is a non-issue, I run the VNC server only when I make screenshots for out manuals May 07 12:47:53 xjqian: the patch is important by the looks of it May 07 12:48:14 koen: Any comments on this vnc stuff? May 07 12:48:34 only the not so helpfull "fbvncserver is evil" May 07 12:48:52 but my experiences with it are >3 years old May 07 12:48:54 Laibsch: here http://pastebin.ca/475408 is the ancient bb file I was using May 07 12:49:44 koen: Objections to restoring 0.6 for those who need it? May 07 12:50:05 0.6 of livncserver May 07 12:50:11 libvncserver May 07 12:50:11 "for example, 0.6 has security problems" <- objection May 07 12:50:38 koen: you didn't quote completely: I wrote "I don't know if, for example ...". May 07 12:50:43 I understood that "IF it had security problmes" that schurig would object, too May 07 12:50:56 ah, sorry May 07 12:51:05 I missed the "don't" May 07 12:51:16 * koen should install a sane font for his IRC client May 07 12:51:21 Let's find out if it does May 07 12:51:40 koen: :-D Listen to schurig May 07 12:51:41 Laibsch: but maybe you look into the usual locations if there are any security problems reported for old versions of libvncserver May 07 12:52:03 fbvncserver is dead software so it might be more worthwhile to check what the api changes were and port fbvncserver onwards May 07 12:52:19 at least then it has a semblance of maintenence in OE May 07 12:53:06 doens't it have a zillions forks all over the net? May 07 12:53:32 koen: One is done by pgfeller who wants to maintain it in OE May 07 12:53:52 !oebug 2126 May 07 12:53:53 * * Bug 2126, Status: ASSIGNED, Created: 2007-04-19 02:37 May 07 12:53:54 * * gfellerpatrik(AT)gmx.net: fbvncserver2_0.9.bb May 07 12:53:55 * * http://bugs.openembedded.org/show_bug.cgi?id=2126 May 07 12:54:13 Which is why I stumbled on this one. I wanted to make sure it compiles fine, before and after May 07 12:54:25 I already failed compiling the version in OE now. May 07 12:54:59 Your suggestions? Make pgfeller's stuff the default soon? May 07 12:55:37 Laibsch: fbvncserver has been bolloxed for so long you should just work on new version May 07 12:57:20 Laibsch, is pgfeller's stuff working? May 07 12:58:08 polyonymous: As I said I got stuck at step 1 which is where I am now. May 07 12:58:27 I'm not sure if I understand what step 1 is. May 07 12:58:49 feller's stuff replaces existing fbvncserver, yes? May 07 12:58:53 try if stuff compiles/works before and after May 07 12:59:01 I am stuck at "before" May 07 12:59:37 polyonymous: Eventually it might replace it. But he says himself it is not yet ready for prime time. May 07 12:59:53 Which is why I would have committed with def pref = "-1" May 07 13:00:02 Laibsch, "before" - as in making sure the old one works? May 07 13:00:08 polyonymous: yes May 07 13:00:30 Laibsch, well, since the old one doesn't I think it makes even more sense to skip step 1. May 07 13:00:51 And go for the replacement in the tree, but, indeed, with -1 May 07 13:00:59 Which is what I came to discuss. but the directions forward is not clear. May 07 13:01:15 And the right steps might again be something which is above my skill level. May 07 13:01:48 Laibsch, well, lemme see if I can build pgfeller's stuff. May 07 13:02:28 polyonymous: remove the packages line May 07 13:02:40 ok May 07 13:06:38 Laibsch, well, it's compiled at least. May 07 13:15:59 03koen 07org.oe.dev * ra56a2cad... 10/ (4 files in 3 dirs): bcm2035tool: takes BDADDR as a commandline argument. Patch from Jan Herman May 07 13:24:45 , How do I get more info? BB did not get me more, then I posted. May 07 13:25:10 step2back: as usual, start with telling what hw and OS you have May 07 13:25:14 sorry, hi all. May 07 13:25:45 for me it's first time. not usual, ;) May 07 13:26:12 merge http://trac.hackndev.com/projects/general/wiki/GentooSetup and http://www.openembedded.org/wiki/GettingStarted May 07 13:26:21 step2back: *any* bug/issue report starts with that May 07 13:26:42 oh my. should try "dopros v militsii" mode May 07 13:26:51 step2back: what CPU you have? May 07 13:27:35 argh.. May 07 13:27:40 host or embedded? May 07 13:28:12 * hrw removes insane from angstrom and do rebuild May 07 13:28:14 step2back: host of course. you have problem on host, right? May 07 13:28:29 hrw: muhaha ;-) May 07 13:29:11 I know that koen/zecke wanted good but I want to build something May 07 13:29:13 amd athlon 2500+ (i686) May 07 13:30:06 * chouimat is back. May 07 13:30:17 hrw: would you post about that to ML? ;-) I'm on brink to do so ;-) May 07 13:30:22 morning May 07 13:30:28 step2back: do you run 32 or 64 bit system? May 07 13:30:49 psokolovsky_: i686 is 32bit May 07 13:31:02 as 'amd athlon' too ;D May 07 13:31:13 well, athlon is not i686 ;-) May 07 13:31:28 step2back: and you run Gentoo Linux, right? May 07 13:31:57 * koen sneaks in a OEDEM todo item in a mail May 07 13:32:47 koen: feel free May 07 13:32:56 psokolovsky_: athlon IS i686 May 07 13:33:10 psokolovsky_, yup athlon 64 isnt May 07 13:33:20 psokolovsky_: ia32 from pentiumII to athlon64 in 32bit is i686 May 07 13:33:25 well, thanks, and who cares ;-E May 07 13:33:58 I managed to not do amd here ;-I May 07 13:34:12 hrw: Althon thunderbird is only i586, no SSE May 07 13:35:28 SSE was req for 686? May 07 13:35:54 no May 07 13:35:57 hrw: wasnt that the change from Pentium i586 to Pentium 2 i686 May 07 13:36:36 my cyrix 233 was 686 and only had mmx1 May 07 13:36:36 , ubuntu 7.04 May 07 13:36:36 XorA: SSE was in pentium3 May 07 13:36:40 step2back: ok, so, first of all, make sure that you bash as your default shell, not dash crap May 07 13:36:54 XorA: "The Pentium Pro is a sixth-generation x86 architecture microprocessor (P6 core) produced by Intel" May 07 13:37:33 this might help http://en.wikipedia.org/wiki/List_of_Intel_microprocessors May 07 13:37:38 /bin/sh -> dash May 07 13:38:12 step2back: change it to bash May 07 13:38:55 ok! May 07 13:39:27 in wiki is nothing talking about it May 07 13:40:53 step2back: there was quite a squarrel about that recently in ML. supposedly, someone needs to produce a patch May 07 13:40:58 Laibsch: hello? ;-) ^^^^ May 07 13:42:09 step2back: dash is evil May 07 13:42:22 , please, repeat in russian, your sentences is too hard for me. May 07 13:42:24 psokolovsky_: yes? Not my problem. May 07 13:42:42 step2back: dpkg-reconfigure dash May 07 13:42:48 step2back: nevermind then. just get rid of dash May 07 13:43:01 Laibsch: ah of course ;-) May 07 13:43:42 It's not resolve problem. May 07 13:44:03 step2back: you have bash as the default now? May 07 13:44:05 step2back: reboot May 07 13:44:13 after setting bash as default May 07 13:44:21 re May 07 13:44:22 See my odyssee at https://answers.launchpad.net/ubuntu/+source/linux-source-2.6.17/+question/4377 May 07 13:44:39 why reboot? May 07 13:45:05 no reboot req May 07 13:45:07 step2back: Don't have time for those kind of questions May 07 13:45:27 hrw: It was for me. The machine would not become fully stable May 07 13:45:28 :) May 07 13:45:37 Reboot is cheap and leaves a clean slate May 07 13:45:53 Unless you have very good reasons not to reboot, I suggest to just do it. May 07 13:46:19 I love that voodoo stuff ;-) May 07 13:49:22 , I hope we will continue our talking later. May 07 13:49:38 yep May 07 13:51:29 Laibsch: at least tell me what ls -l /bin/sh gives you now May 07 13:52:51 schurig: What are the usual locations to check for security problems? Secunia? I am not a security expert. May 07 13:53:40 psokolovsky_: Why? Have I refused to give you that info in the past? It points to bash. May 07 13:53:50 http://www.dilbert.com/comics/dilbert/archive/images/dilbert21466330070507.gif May 07 13:54:38 hrw: so true :) May 07 13:54:46 Laibsch: nope, I just could imagine that you, having a test system at your fingertips, could do a sanity.bbclass patch to filter out dash. now I have idea to do that in blind ;-) May 07 13:55:31 psokolovsky_: No way I will ever touch dash again (at least for a couple of years) May 07 13:55:41 lol May 07 13:55:52 It caused me too much grief. I need that system and cannot have it become unstable again May 07 13:56:07 ~lart dash May 07 13:56:07 * ibot stabs dash May 07 13:56:16 dash it not at fault May 07 13:56:25 I don't care :-D May 07 13:56:27 beg May 07 13:56:29 people wrongly assume everyone is running bash May 07 13:56:46 if you do !/bin/sh you should adhere to posix sh, not bash May 07 13:57:20 OK, maybe I shot the messenger. But I was frustrated ;-) May 07 13:57:29 AND I really needed to shoot someone May 07 13:57:34 to feel better May 07 13:57:53 Will that get me off the gallows? ;-) May 07 13:58:22 well, we of course need to deal with actual state of things, and as dash known to cause issues, it should be filtered May 07 13:59:59 schurig, koen: It seems that libvncserver < 0.8.2 have security issues: http://archives.neohapsis.com/archives/fulldisclosure/2006-08/0097.html which I found not via the usual routes but just plain google friend. I guess restauration is out of the question, then. May 07 14:00:13 There might be others. May 07 14:01:33 psokolovsky_: Feel like tackling point two of the requirements for opie? I will take a look at it. Maybe it is easily fixable. May 07 14:01:57 what is point two? May 07 14:06:20 Laibsch: yes, let's do it. worth some thing first tho May 07 14:07:26 polyonymous: see wiki May 07 14:07:46 polyonymous: and add yourself ;-) May 07 14:08:25 psokolovsky_, libsdl-qpe ? May 07 14:08:42 psokolovsky_, or /usr/bin ? May 07 14:08:50 polyonymous: /opt May 07 14:08:52 ah ok May 07 14:10:03 added myself :) May 07 14:10:29 having it in /usr/ would be a good thing, indeed. May 07 14:10:52 polyonymous: building opie against qt/X would be even better :) May 07 14:11:27 koen, that sounds a bit like long-term goal May 07 14:11:47 long term is qt/4 May 07 14:11:48 XorA: you have poodle in WM. can you look at #1325? May 07 14:12:13 polyonymous, psokolovsky_: Next lines for you May 07 14:12:19 (14:57:56) XorA: export palmtopdir = "/opt/QtPalmtop" May 07 14:12:19 (14:57:56) XorA: export palmqtdir = "/opt/QtPalmtop" May 07 14:12:21 XorA: #1691 can also be fixed by you - settings from terrier should be same as on akita May 07 14:12:22 (14:58:00) XorA: in bitbake.conf May 07 14:15:04 Laibsch: so what's in that? May 07 14:15:31 hrw: I think that is the correct behaviour, the head phone insert IRQ is a wakeup source May 07 14:15:31 That was being discussed as being related to the fix of moving stuff to /usr May 07 14:16:12 polyonymous, psokolovsky_: I guess that needs to point to /usr or /usr/$something. What do you suggest May 07 14:16:21 Laibsch, I guess if that was enough that wouldn't be an issue :) May 07 14:16:40 * koen guesses "/usr" May 07 14:16:58 Laibsch: I also suspect we'd need few makefile/.pro patches ;-) May 07 14:17:23 sure, it should be default user software install prefix, /usr May 07 14:18:13 one thing which could go side by side with this is making sure that all opie apps consitently have 'opie-' prefix May 07 14:18:26 but doing that, we'd deviate from upstream opie ;-) May 07 14:18:42 and that would somewhat suck. May 07 14:18:44 psokolovsky_: per app (or lib) releases would be nice too May 07 14:19:09 someone should get opie commit ;-) May 07 14:19:27 thanks for volunteering, psokolovsky_ ;-) May 07 14:19:30 psokolovsky_, I don't believe in opie commits :) May 07 14:19:51 OK, so what are the next steps for this /opt issue? May 07 14:19:52 Laibsch: volunteering for what? ;-) May 07 14:20:02 psokolovsky_: someone should get opie commit May 07 14:20:21 that doesn't mean I should ;-) May 07 14:20:40 Well, it should certainly not be me. May 07 14:20:48 * Laibsch eyes polyonymous May 07 14:21:01 Laibsch, how can I possibly? May 07 14:21:48 ljp: Hi Lorn! you're still opie1 admin/dev, right? can you lead someone into devs too? May 07 14:22:32 * polyonymous runs off. May 07 14:25:07 * Laibsch points polyonymous and psokolovsky_ to May 07 14:25:11 !oebug 2254 May 07 14:25:12 * * Bug 2254, Status: NEW, Created: 2007-05-07 07:24 May 07 14:25:13 * * bugs.openembedded.org(AT)rolf.leggewie.biz: move opie away from /opt to /usr May 07 14:25:14 * * http://bugs.openembedded.org/show_bug.cgi?id=2254 May 07 14:25:48 Laibsch, added it to opie-in-agstrom? May 07 14:26:32 you did. May 07 14:30:41 Laibsch: updated yah bug May 07 14:32:49 ERROR: QA Issue: non -dev package contains symlink .so: libfuse path '/work/armv5te-angstrom-linux-gnueabi/fuse-2.6.5-r0/install/libfuse/usr/lib/libfuse.so'' May 07 14:33:01 is this a big problem May 07 14:35:43 Laibsch, pgf's fbvncserver segfaults. May 07 14:35:53 xjqian: it is May 07 14:36:17 xjqian: luckily the fix is pretty easy :) May 07 14:36:48 03koen 07org.oe.dev * r0a4161a5... 10/ (1 packages/fuse/fuse_2.5.3.bb): fuse 2.5.3: fix FILES for libfuse May 07 14:37:39 XorA: Thanks a lot. May 07 14:38:37 koen: thanks, will check it out soon May 07 14:38:52 xjqian: http://lists.linuxtogo.org/pipermail/openembedded-commits/2007-May/005346.html May 07 14:40:54 What machine is the qemuarm closest to? May 07 14:41:06 * Laibsch hopes this question makes sense and is understood May 07 14:41:51 Laibsch: *your* machine ;-) May 07 14:41:52 Laibsch: from the 0.8.0 changelog ARM system emulation: Arm Integrator/CP board with an arm1026ej-s May 07 14:42:08 Laibsch: I run h4000 images on qemu right away ;-) May 07 14:46:08 well, maybe this needs a bit more explanation. May 07 14:46:26 Some packages can be shared across all different arm machines. May 07 14:46:42 Some are specific to arm5vte or whatever this stuff is called. May 07 14:47:26 So, I was wondering about the attributes of arm1026ej-s. May 07 14:48:20 Closer to spitz or closer to collie or close to neither? May 07 14:48:37 * Laibsch googles for arm1026ej-s May 07 14:49:04 arm10 is usually armv5te May 07 14:49:12 your collie as armv4 May 07 14:50:10 koen: Thanks. so closer to spitz. May 07 14:50:33 closer to nokia770 May 07 14:50:37 Laibsch: I keep this page in my bookmark http://en.wikipedia.org/wiki/ARM_architecture May 07 14:50:43 * Laibsch has no nokia 770 May 07 14:51:48 chouimat: Lots of good info. But quite complex, indeed May 07 14:52:28 Laibsch: yup ... too many variants :) May 07 14:55:51 Laibsch: qemuarm can boot any armv5t/armv4/armv4 image May 07 14:56:31 hrw: I was wondering which one is picked when I build for MACHINE = "qemuarm" May 07 14:56:45 Laibsch: then qemuarm is used May 07 14:56:47 It seems to be armv5te May 07 14:56:54 Laibsch: in Poky it is armv5te May 07 15:00:05 ~lart mozilla for cvs May 07 15:00:05 * ibot hits mozilla with an anvil and laughs with a contralto voice ... Haha Ha HA Ha for cvs May 07 15:01:00 Laibsch: nice bug ;-) May 07 15:02:29 hrw: as their repo? I heard they switch to hg? May 07 15:02:50 psokolovsky_: but keep cvs May 07 15:06:35 hey May 07 15:07:12 hey zecke May 07 15:08:59 schurig: Which one? There are so many? The opie /opt one? That is really koen's bug. He forced us to address that. May 07 15:09:25 Laibsch: the libvncserver bug is a nice bug. Sounds almost like a backdoor. May 07 15:10:30 ~lart apple care May 07 15:10:30 * ibot whips out his power stapler and staples apple care's foot to the floor May 07 15:12:17 zecke: hehe :) May 07 15:13:22 zecke: I thought apple care was PERFECT :) May 07 15:31:47 ~lart apple care May 07 15:31:47 * ibot lowers apple care's priority May 07 15:34:34 zecke: I had problem with them too May 07 15:39:14 koen: please fix this as well May 07 15:39:31 ERROR: QA Issue: non -dev package contains symlink .so: libxcb-composite path '/work/armv5te-angstrom-linux-gnueabi/libxcb-0.9.93-r1/install/libxcb-composite/usr/lib/libxcb-composite.so' May 07 15:39:35 minimo cvs == 981M May 07 15:39:51 same issue (missing a dot) as in the libfuse package May 07 15:40:41 xjqian: that's pH5's domain :) May 07 15:41:00 koen: trivial things can be changed by anyone May 07 15:41:28 hrw: I'll have to rebuild the package, which I can't to at the moment May 07 15:41:39 thats the other thing May 07 15:41:39 (staging glibc for armv6) May 07 15:41:42 ;) May 07 15:41:57 gcc-cross/do_compile for progear May 07 15:51:33 zecke: Can plucker work on a full directory? May 07 15:52:06 Laibsch: I don't know :) May 07 15:52:12 k May 07 16:09:01 03koen 07org.oe.dev * rdde38440... 10/ (1 packages/xcb/libxcb_0.9.93.bb): libxcb: fix FILES, spotted by Junqian Gordon Xu May 07 16:11:31 koen: do you know do we have some light applet/tool for networkmanager to easily connect to WPA network? May 07 16:12:22 I only know of nm-applet May 07 16:12:33 which recipe builds it? May 07 16:12:41 network-manager May 07 16:14:25 the network-manager recipes needs some splitting up May 07 16:15:06 XorA|gone and mmp are the network manager gurus May 07 16:15:40 Hi everybody May 07 16:15:57 hi Thomas May 07 16:16:18 hey thesing May 07 16:18:35 hrw: from your old live, is there a php function to return the extension from a filename? May 07 16:19:31 zecke: $tmp = pathinfo($filename); echo $tmp['extension']; May 07 16:19:59 hrw: what if filename comes from a HTTP post? May 07 16:20:17 then you have filename in $_FILES array somewhere May 07 16:20:45 * koen hits head against desk again May 07 16:21:04 * koen stabs stupid local diff May 07 16:28:57 * wick728 is away: afk May 07 16:30:37 NOTE: package tk-8.4.11-r5: task do_package: startedNOTE: the following files were installed but not shipped in any package: May 07 16:30:37 NOTE: /usr/lib/tkConfig.sh May 07 16:30:37 NOTE: Couldn't find shared library provider for libtk8.4.so May 07 16:30:37 NOTE: Couldn't find shared library provider for libtcl8.4.so May 07 16:30:37 NOTE: Couldn't find shared library provider for libtk8.4.so May 07 16:30:39 NOTE: Couldn't find shared library provider for libtcl8.4.so May 07 16:34:56 ERROR: QA Issue: non -dev package contains symlink .so: libcrypto path '/work/armv5te-angstrom-linux-gnueabi/openssl-0.9.7g-r2/install/libcrypto/usr/lib/libcrypto.so' May 07 16:34:56 NOTE: Checking Package: libssl May 07 16:34:56 ERROR: QA Issue: non -dev package contains symlink .so: libssl path '/work/armv5te-angstrom-linux-gnueabi/openssl-0.9.7g-r2/install/libssl/usr/lib/libssl.so' May 07 16:35:11 jesus this bug keeps poping up May 07 16:36:05 xjqian: I think there is a reason zecke called his QA checker insane.bbclass :) May 07 16:37:11 we add bugs like that to have some low-hanging fruit for new developers :) May 07 16:38:16 koen: do I earn reward points for spotting these :) May 07 16:40:06 xjqian: I think that is how koen can afford to go to all those conferneces :) May 07 16:40:23 sure, a virtual cookies for each of them :) May 07 16:40:38 with each http request :) May 07 16:41:02 and for every other request you get TWO cookies! :) May 07 16:41:09 koen: how many cookies per beer ? May 07 16:41:36 koen: hmm. nm-applet req libgpewidget May 07 16:41:52 it does? May 07 16:41:59 must be a crazy mmp patch May 07 16:42:30 or atleast networkmanager req it May 07 16:42:55 now I see May 07 16:43:33 --with-gpe option May 07 16:44:54 ok, enough cokies. What about the problem in tk-8.4.11? May 07 16:45:20 no idea on that May 07 16:45:50 03koen 07org.oe.dev * ra1f35b32... 10/ (1 packages/linux/ep93xx-kernel_2.6.19+2.6.20-rc7.bb): ep93xx-kernel: fix do_deploy May 07 16:45:55 03koen 07org.oe.dev * re4adba48... 10/ (5 files in 2 dirs): openssl: fix FILES, spotted by Junqian Gordon Xu May 07 16:46:44 Laibsch: ping May 07 16:47:02 ERROR: QA Issue: non -dev package contains symlink .so: libpython2 path '/work/armv5te-angstrom-linux-gnueabi/python-2.4.4-ml1/install/libpython2/usr/lib/libpython2.4.so' May 07 16:47:14 one more cookie May 07 16:48:23 hmm... Package perl-lib wants to install file /usr/lib/libperl.so.5 May 07 16:48:23 But that file is already provided by package libperl5 May 07 16:48:29 was the package renamed? May 07 16:48:53 sounds like broken shlib renaming May 07 16:49:20 shlib? I thought it's more like package renaming? perl-lib vs libperl5 ? May 07 16:49:40 debian.bbclass should rename it to libperl5 May 07 16:50:07 hmm.. May 07 16:50:24 your best bet is to ping v8jlene May 07 16:50:50 koen, yeah, been pinging with him yesterday with my /usr/include/gdbm problem :) May 07 16:51:31 hi all May 07 16:53:03 thesing: pong May 07 16:53:27 v8jlene: Compilation for Sharp breaks again. Changed anything back? May 07 16:53:48 Laibsch: did you load mmc_block when you tested your card? May 07 16:53:52 v8jlene: do you still need help with the debian source packages? May 07 16:54:00 v8jlene: It is again do_patch with ext/Errno/Errno_pm.PL erroring out. May 07 16:54:07 v8jlene: you might end up needing help from bitbake to generate the rules for you May 07 16:54:20 Laibsch, saying what? May 07 16:54:34 thesing: I did only "modprobe locomo_spi". I had thought it would pull in all other stuff necessary. May 07 16:54:37 Laibsch, he fixed it yesterday for me ;-) May 07 16:54:46 (that is - not necessarily for you) May 07 16:55:06 polyonymous: In what way? May 07 16:55:32 Laibsch, in a patching way :) But that had to do with my particular gdbm inclusion. What exactly your error is? May 07 16:55:40 Laibsch: could you try your card again with locomo_spi and mmc_block loaded? May 07 16:56:01 Laibsch: you need mmc_block to use your card as a block device May 07 16:56:48 v8jlene: Sorry, I mistook the build I was running. It fails for MACHINE="qemuarm" and angstrom so it should be quite easy to reproduce. I updated OE this morning. May 07 16:56:58 thesing: Sure May 07 16:57:07 thesing: Just the verbose log is enough? May 07 16:57:38 v8jlene, not necessarily easy to reproduce the problem may be related to the build host, not the target. May 07 16:58:10 Laibsch: Yes. May 07 16:58:20 polyonymous: A little xterm pops up with http://rafb.net/p/RGQjXZ96.html May 07 16:58:24 Laibsch, the last message was for you, in fact :) May 07 16:58:26 ah May 07 16:58:49 bad patch may be relatively easy to fix. May 07 16:59:53 but it would take me a while to start this build and get to the point where it fails. May 07 17:00:54 ERROR: QA Issue: non debug package contains .debug directory: python-fuse path /work/armv5te-angstrom-linux-gnueabi/python-fuse-0.0+cvs20070507-r0/install/python-fuse/usr/lib/python2.4/site-packages/fuseparts/.debug/_fusemodule.so May 07 17:01:02 a little different cookie May 07 17:01:55 btw, this one actualy stops the build May 07 17:06:15 FILES_${PN}-dbg += "${libdir}/python2.4/site-packages/fuseparts/.debug" or something like that May 07 17:07:05 bad packaging May 07 17:07:21 thesing: I have just rebooted my collie and before even inserting locomo_spi the mmc_block module is already there. So I assume the log *was* indeed taken with that module loaded. May 07 17:07:22 indeed May 07 17:07:27 those are trivial enought so I don't bother to file bugs May 07 17:07:55 RP: do you have some time for me? May 07 17:07:59 xjqian: If you commit the fixes that is OK. Otherwise. please file a bug so others can commit the fix May 07 17:08:11 FILES_${PN}-dbg += "${libdir}/python${PYTHON_MAJMIN}/site-packages/fuseparts/.debug" even May 07 17:08:48 thesing: Still want me to take that log? May 07 17:09:56 Laibsch: Yes. May 07 17:22:42 thesing: I will spare us all cluttering up the bug any further: http://oz.leggewie.org/wip/syslog.gz May 07 17:23:18 Laibsch: you forgot so set the permissions right. May 07 17:24:25 fixed May 07 17:24:52 * wick728 is away: afk May 07 17:26:39 Laibsch: does all this happen after you insert the card or did you remove the card too? May 07 17:27:25 thesing: hi, the software-spi-mmc-driver works now again May 07 17:27:30 both insertion and removal May 07 17:28:08 thesing: for simpad, and 270 kByte/s read and 90 kByte/s write are not bad i think ;.) May 07 17:28:20 the file size just for insertion is about 25+30K. After removal it grows to 39K May 07 17:29:56 mrdata: I have no idea. is it better than 2.4? May 07 17:30:27 thesing: hi, do you have planes to write an audio driver for collie? It is the same chip as simens simpad uses. we maybee can share work. May 07 17:30:38 thesing: yes, for 2.4 i had only 125kByte read May 07 17:31:15 Laibsch: could you try unloading mmc_block, inserting the card, and load mmc_block? May 07 17:32:25 thesing: Sure. And then recapture? May 07 17:32:27 mrdata: yes sometime after I fixed all suspend bugs for collie. May 07 17:34:22 thesing: Shall I remove mmc_core as well? May 07 17:34:39 Laibsch: I need the log from the point you insert the card. May 07 17:34:55 Laibsch: doen't hurt May 07 17:35:16 OK, then I will need to remove mmc_spi as well May 07 17:35:49 thesing: fine :). after 25th may I will have some time. I will ask you a second than how to help you. May 07 17:35:50 bug 2255 filed for python-fuse bad debug packaging. blocking meta-bug 2194 May 07 17:36:06 Laibsch: Right. thats bad. so better don't remove mmc_core. May 07 17:36:57 too late May 07 17:37:03 ;-) May 07 17:37:12 Already reinserted everything, though May 07 17:38:25 locomo_spi, mmc-spi and mmc_block are loaded. Anything else to check? May 07 17:39:40 Laibsch: no, but mmc_block should not be loadead from the start. load it after you inserted the card, and no more messages appear. May 07 17:41:26 mr_nice: didn't somebody already work on the sound driver for simpad? May 07 17:42:10 what chip does simpad exactly use? May 07 17:42:20 hrw: ucb1300 May 07 17:42:34 1300, not 1341 or 1380? May 07 17:42:52 thesing: no, there is the same old 2.4 oss driver as for collie May 07 17:42:58 hrw: UCB1300 May 07 17:43:24 thesing: and it never worked May 07 17:43:47 hrw: i know soc from alsa have driver for ucb1380 May 07 17:44:46 03koen 07org.oe.dev * rd0101609... 10/ (1 packages/fuse/fuse_2.6.3.bb): fuse 2.6.3: fix packaging May 07 17:45:32 hrw: but i do not know, how different the chips are May 07 17:46:22 mrdata: I will search a datasheet for 1380 May 07 17:46:32 USB TC35143 is USB1300? May 07 17:46:40 s/USB/UCB May 07 17:46:48 03hrw 07org.oe.dev * rdddf8b42... 10/ (5 files in 4 dirs): linux 2.6.20: drop Simpad and Progear support May 07 17:46:56 03Bernhard 07org.oe.dev * r80f7f84f... 10/ (8 files in 3 dirs): linux: added 2.6.21 for Simpad (progear need refresh of config) - close #2247 May 07 17:47:08 hrw: i think so, aka "betty" for mips May 07 17:49:36 03koen 07org.oe.dev * radb2b82e... 10/ (1 packages/python/python-fuse_cvs.bb): python-fuse: fix packaging and depends, spotted by Junqian Gordon Xu May 07 17:50:16 http://www.cems.uwe.ac.uk/~rwilliam/RTSD_ufeEJ5-20-3/RTSD_course/TC35143BF.pdf if you need it May 07 17:50:19 thesing: http://oz.leggewie.org/wip/sd.gz May 07 17:53:37 Laibsch: where there no messages after you loaded mmc_block? May 07 17:54:13 no May 07 17:54:21 And no increase in size for the log May 07 17:55:13 steps: reset log to 0, insert card, load mmc_block, remove card May 07 17:55:48 thesing, mr_nice: read http://www.cems.uwe.ac.uk/~rwilliam/RTSD_ufeEJ5-20-3/RTSD_cw/rtsd4_assign_2-07.pdf May 07 17:56:56 Initially, the aim is to program the Puppeteer to serve as a basic Internet Radio, accepting the target station’s IP May 07 17:56:59 number from the console and streaming the resulting audio data for decoding and playback through the analogue May 07 17:57:02 headphone socket. To achieve this, a sound-system device driver for the Toshiba TC35143 chip needs to be May 07 17:57:05 developed to work under Linux 2.6. May 07 17:57:09 this paragraph is interesting for you May 07 17:57:25 hrw: thx May 07 18:02:30 could 'inherit "rm_work" break some packages for building? May 07 18:02:48 goxboxlive: only some broken ones May 07 18:03:18 hrw: Has there been som updates with gnumeric lately, it wouldnt build May 07 18:04:18 hrw: pastebin.co.uk/14145 May 07 18:05:54 goxboxlive: bitbake orbit-native and retry May 07 18:06:05 * hrw -> food and away May 07 18:06:05 cu May 07 18:06:08 yes May 07 18:06:13 i have done that May 07 18:07:25 xjqian: I updated fuse to 2.6.5 May 07 18:08:48 03koen 07org.oe.dev * rb902929d... 10/ (5 files in 2 dirs): fuse: update to 2.6.5 and make packaging more granular May 07 18:11:59 Laibsch: You didn't remove the card this time? May 07 18:16:44 koen: thanks May 07 18:17:25 koen: but you stole my fun :) May 07 18:17:34 sorry :) May 07 18:17:51 it was such a mess I had to fix it :) May 07 18:20:48 koen: yeah, it is a mess. then please close bug 884 and bug 2047 May 07 18:22:14 thesing: Sure I did. Same as before May 07 18:22:16 steps: reset log to 0, insert card, load mmc_block, remove card May 07 18:22:42 hvontres|poodle: ping May 07 18:34:23 03xjqian 07org.oe.dev * r68d9844c... 10/ (3 files in 2 dirs): sshfs-fuse: update to 1.7 May 07 18:37:09 please take a look at sshfs-fuse again, new patch supplied May 07 18:37:29 sorry about that, my earlier patch is sloppy May 07 18:39:01 the only difference is: FILES_${PN} += "/usr/lib/sshnodelay.so" May 07 18:42:14 xjqian: don't hardcode /usr/lib May 07 18:42:39 xjqian: check bitbake.conf for a key (it might be libdir) that has a value like /usr/lib May 07 18:46:46 cat bitbake.conf | grep /usr/bin. I only see /usr/bin/env May 07 18:47:03 did i miss something? May 07 18:47:15 grep libdir bitbake.conf May 07 18:50:29 koen: nothing promising either. I've tried ${libdir}, but it's not packing stuff in /usr/lib. I'm almost sure.... May 07 18:51:24 xjqian: http://rafb.net/p/OoLCtS53.html does what you want May 07 18:52:51 koen: sorry I'm wrong May 07 18:52:56 =-O May 07 18:55:11 koen, what's wrong with the part of /etc/init.d/rc after '#Uncomment to cause psplash to exit manually, otherwise it exits when it sees a VC switch' ? Or better, I haven't look at that part - what's wrong with this approach? May 07 18:57:37 polyonymous: iirc nothing is wrong, but the ohand-dudes that wrote it don't have a (small) pause between pslash exit and X startup May 07 18:57:48 so they leave pslash on the fb till X clears it May 07 18:57:56 (by switching VC) May 07 18:58:24 03koen 07org.oe.dev * re6967a90... 10/ (1 packages/fuse/sshfs-fuse_1.7.bb): sshfs-fuse: package all files, closes #2047 May 07 18:58:35 koen, yes, but that should be no problem - that is the code won't do any harm if psplash already exited. May 07 18:59:35 I upgraded the initscript a while ago without realizing this feature (not exiting) was in there, and I haven't taken the time to sit down and see why May 07 19:00:44 koen, I'll try to uncommend the code and test it with x11 and console images then... May 07 19:01:05 thanks May 07 19:01:06 It surely won't harm opie, because opie initscript does it manually before running opie May 07 19:01:31 grrr... can anybody tell me how I get the ibot to tell me what is the content of my botmail? May 07 19:01:32 I merely want console-image to work without vc switch :) May 07 19:01:47 philippe: 'botmail read' May 07 19:01:55 doesn't work... May 07 19:01:58 or '~botmail read' outside of a query May 07 19:02:16 koen: thanks. I'll try that May 07 19:02:21 koen, that by itself is easy, I was thinking about why was it commented out. May 07 19:02:31 well, I'll get back to you in a couple of boots. May 07 19:02:48 polyonymous: to avoid the pause and flicker between psplash exit and X startup May 07 19:03:12 koen, hmm... Okay, how about adding a S99zzapsplash ? May 07 19:03:16 for console image May 07 19:03:28 if that works, I'm all for it May 07 19:03:33 well, nevermind, I'll play around a bit on myself. May 07 19:03:54 The only drawback of this approach is that it won't reach 100% before stoping :)) May 07 19:04:16 ah stupid freenodde blocking privmsg again,,, May 07 19:04:37 Laibsch: yes go ahead. No problem dropping some gpe-timesheet bb's May 07 19:07:45 enough work done, time for some stargate atlantis May 07 19:08:15 philippe: OK. Thanks for the response. May 07 19:08:19 koen: sounds good :-) May 07 19:08:23 Laibsch: you're welcome May 07 19:08:35 koen: lucky bastard :) May 07 19:40:15 re May 07 19:40:55 hi hrw May 07 19:40:57 wb May 07 19:41:25 philippe: do you work in the same building as the gtk-webcore guy? May 07 19:41:50 http://people.redhat.com/dcantrel/dhcdbd/dhcdbd-2.0.tar.bz2 - can someone checksum it? May 07 19:41:56 more fuse madness:) reopened bug 884: fuse-module, new bug 2256: gmailfs-fuse. Both blocking meta bug 2194. Enough for me today. See ya May 07 19:42:12 zecke: nope May 07 19:42:20 zecke: would be nice though... May 07 19:44:02 * cbrake always gets some satisfaction watching the revs wiz by on mtn pull with new versions of mtn :-) May 07 19:44:48 hrw: what sum do you want? May 07 19:45:13 md5 May 07 19:45:17 hrw: any idea why the script failed sometimes? was your download incomplete? May 07 19:45:30 with limited integrity d71de5c2f8aea4c8a18659bfa8387414 May 07 19:45:37 zecke: same here May 07 19:45:43 also see http://people.redhat.com/dcantrel/dhcdbd/SHA1SUMS May 07 19:45:50 if you trust the people.redhat.com server May 07 19:48:13 zecke: bzip2: (stdin): trailing garbage after EOF ignored May 07 19:48:21 celf mirror is broken for that file May 07 19:52:45 zecke: looks like we need archive_tester for bittest May 07 19:53:02 hrw: we have it, patch_checker May 07 19:53:09 hrw: it calls unpack + patch May 07 19:53:37 zecke: I was wondering. Can it do that for package A and all the packages A depends on? May 07 19:53:39 does it remove after test? May 07 19:54:05 hrw: I think it should be taught not to download, it could remove them May 07 19:54:20 Laibsch: bittest? no it is dumb May 07 19:54:29 too bad May 07 19:54:59 Laibsch: why would you want that? What is the use case? May 07 19:55:35 zecke must eat vla May 07 19:55:48 * zecke checks the fridge for dutch food May 07 19:56:30 koen: but that is actually a good idea. I think I will try to get some Vla tomorrow May 07 19:56:38 probably I can hook up others to it as well May 07 19:56:44 :) May 07 19:56:50 koen: mmhh vla :-) May 07 19:57:17 Should ask my dad to get some as he works in shiphol now. May 07 19:58:22 koen: Edeka and Reichelt sell some vla and the funny thing is: It is less expensive than in the NL May 07 19:58:31 hmmmm May 07 19:58:49 koen: we don't get the campina tetra packs but only the quality stuff May 07 19:59:02 * zecke would prefer campina stuff though May 07 19:59:06 vla ? May 07 19:59:12 you want Friesche Vlag anyway May 07 19:59:21 chouimat: a kind of pudding May 07 19:59:27 chouimat: http://en.wikipedia.org/wiki/Vla May 07 19:59:40 chouimat: with lots of different tastes. May 07 19:59:55 the sometimes disgusting double-vla :} May 07 20:00:19 I think they put the bad of vanilla and chocolate vla into one pack May 07 20:00:23 * philippe thinks the Dutch should export vla instead of Heineken May 07 20:00:25 double-vla is for lazy people May 07 20:00:40 you should make it using a pack of vanilla and a pack of chocolate May 07 20:00:51 philippe: yup heiniken is bad May 07 20:02:58 * philippe wonders if Lidl would sell vla... May 07 20:03:07 it does in .nl May 07 20:03:08 Lidl is the delicacy store of Finland... May 07 20:03:54 The only place they sell affordable real food here. May 07 20:04:21 03Laibsch 07org.oe.dev * r4e038cbe... 10/ (9 files in 2 dirs): gpe-ownerinfo: unify May 07 20:04:28 03Laibsch 07org.oe.dev * r142643f8... 10/ (3 files in 2 dirs): entice: clean up. Bump PR. Partly fixes 2234. May 07 20:04:35 03Laibsch 07org.oe.dev * r1cbdcdd6... 10/ (3 files in 2 dirs): links: clean up. Bump PR. partly closes 2234. May 07 20:04:43 03Laibsch 07org.oe.dev * r1155cc3e... 10/ (4 files in 2 dirs): gphoto2: clean up. Bump PR. Partly closes 2234. May 07 20:04:51 03Laibsch 07org.oe.dev * r879c0996... 10/ (1 packages/fltk/fltk_1.1.4.bb packages/fltk/fltk_1.1.7.bb): fltk: clean up. Bump PR. Partly closes 2234. May 07 20:04:59 03Laibsch 07org.oe.dev * rc82df0fa... 10/ (1 packages/mb-desktop-xine/mb-desktop-xine_0.4.bb): mb-desktop-xine: clean up. Bump PR. Partly closes 2234. May 07 20:05:05 03Laibsch 07org.oe.dev * r35210742... 10/ (1 packages/netpbm/netpbm_10.28.bb): ntpbm: clean up. Bump PR. Partly closes 2234. May 07 20:05:14 03Laibsch 07org.oe.dev * r42c1f8fc... 10/ (5 files in 2 dirs): May 07 20:05:14 gpe-timesheet: drop all versions prior to 0.3. May 07 20:05:14 Consulted with philippe and got the go before committing. May 07 20:05:24 03Laibsch 07org.oe.dev * ra8c7b34e... 10/ (1 packages/povray/povray_3.6.1.bb): povray: Clean up. Bump PR. Partly closes 2234. May 07 20:05:36 Laibsch: 2234 was already fixed May 07 20:05:51 Laibsch: wow....Good work :-) May 07 20:05:58 Laibsch: you just forced needless rebuild for everyone to fix a non-existing problem May 07 20:06:18 Laibsch: which is *not* the way to go May 07 20:06:36 koen: well everybody makes mistakes... May 07 20:07:01 philippe: not everybody does so after being told not to do that May 07 20:07:02 :) May 07 20:07:31 philippe: I still do not think this is a mistake. May 07 20:07:52 Well I am going to stay out of this as I have no clue about the facts May 07 20:08:04 These are fringe packages, nothing like glibc May 07 20:08:39 the idea is the same May 07 20:08:52 so koen is again exaggerating something out of proportions because he is moody again May 07 20:08:53 !oebug 2234 May 07 20:08:55 * * Bug 2234, Status: NEW, Created: 2007-05-04 05:11 May 07 20:08:56 * * xyzaid(AT)yahoo.fr: Packages depending on 'jpeg' need rebuilding May 07 20:08:57 * * http://bugs.openembedded.org/show_bug.cgi?id=2234 May 07 20:09:20 Laibsch: no, I'm not moody May 07 20:09:33 I cleaned up stuff so it confirms to policy which I have done in the past ;-) May 07 20:10:43 and since you don't have access to the angstrom feeds, you cannot possible fix #2234 May 07 20:11:14 koen: get a heineken, be mean to some animals, eventually beat your wife :) May 07 20:11:21 luckily you are not married May 07 20:12:02 zecke, luckily for whom? For Laibsch? I don't think so :) May 07 20:12:35 polyonymous: Laibsch is post the point of return May 07 20:12:47 polyonymous: and I appreciate him for that ;) May 07 20:12:51 :))) May 07 20:23:44 cu May 07 20:25:18 good morning all May 07 20:25:29 hey likewise May 07 20:26:43 night all May 07 20:26:51 night mrdata May 07 20:26:56 hey likewise ! May 07 20:27:25 hey philippe , Jin^eLD ! May 07 20:27:44 hey likewise May 07 20:27:56 hey koen, Crofton May 07 20:28:03 hey May 07 20:29:47 hey Crofton May 07 20:30:28 Crofton: should we try to get the complete osk, including screen, camera and stuff? May 07 20:31:11 hmm May 07 20:31:17 I have a screen, but do not use it May 07 20:31:46 for someone who was good at gui stuff,, a screen would be useful May 07 20:32:04 or to test dsp fb output May 07 20:32:10 yeah May 07 20:32:32 for what I do, there is not enough resources to do sig processing, and graphics May 07 20:32:39 but some people would be interested May 07 20:33:06 it falls into the category, worth doing if you have time, someone pays you, or lots of people are interested :) May 07 20:33:21 a long time ago, we had opie running on the OSK May 07 20:33:59 likewise, have you made any headway with libtool and rpath? May 07 20:35:30 Crofton: no. I am kind of frustated about all the bug searching and started playing with new efika kernels instead :-) May 07 20:35:54 Crofton: but I'll be happy to search along for a bug. May 07 20:37:26 I have a bad feeling we are going to need to patch libtool May 07 20:37:58 * koen mentions chrpath in qemu again May 07 20:38:15 (that would involve massive amount of pain, though) May 07 20:38:38 I am hopeful we can find the root cause and fix May 07 20:38:54 unfortunately, I am trying to beat down the list of real work I have May 07 20:40:31 to make matters worse, the newer libtool dies of told to search a path with wrng arch files May 07 20:43:48 hi florian May 07 20:44:17 re May 07 20:47:01 03koen 07org.oe.dev * rd095bc88... 10/ (4 files in 2 dirs): fuse: (hopefully) fix all remaining issues May 07 20:48:02 hmmm, funky nordic subtitles May 07 20:48:28 "mind you, moose bites can be pretty nasty" May 07 20:49:21 "men jag förstÃ¥r inste" May 07 20:49:35 * koen bets it's swedish May 07 20:49:45 koen: I think you are right there. May 07 20:50:02 (it said "svensk" just now) May 07 20:50:21 update-rc.d.bbclass only wants one initscript per package? May 07 20:50:43 polyonymous: one of its limitations May 07 20:50:49 koen, :( May 07 20:51:00 koen cool I love those pretty weird caracteres ... May 07 20:51:14 I'm basically done with psplash thing, but I want some way to get it installed :) May 07 20:51:48 the "solution" is to copy the code from the class for the extra script May 07 20:52:14 koen, yup, that's what I'm thinking. Not that I like the idea, but I think that's the best solution here, unfortunately. May 07 20:53:02 koen, another psplash solution is to add check for gpe-dm process to rc, but I don't really like this tweak. May 07 20:54:10 a check for gpe-dm, opie-dm, xfce-dm, e17-dm, etc May 07 20:54:12 On the other hand, we have psplash code all around rc... But that would also add gpe-dm related code... May 07 20:54:20 yeah, that would suck. May 07 20:54:43 And one more solution, I think the best one - is to add zapsplash to console image only. May 07 20:54:54 yeah, just thought of that as well May 07 20:55:47 Maybe then we should go for it? May 07 20:56:25 Would you mind if I add a zapsplash package to psplash directory and you add it to console image if you like it? May 07 20:56:25 you have my vote May 07 20:56:39 you can add it as well :) May 07 20:56:52 yeah, but I'll have to look where. That's doable, of course :) May 07 20:57:13 Well, basically, do you think putting it into packages/psplash is justified? I don't think it deserves its own directory. May 07 20:57:46 I think it's justified May 07 20:57:53 ok May 07 21:11:37 xjqian: bug 2248's intention is not "please commit this because it is trivial" May 07 21:12:33 It is really just a bad workaround since we lack a defined workflow and QA process May 07 21:13:03 * Laibsch thinks he will eventually have to come up with something better May 07 21:14:54 re May 07 21:15:44 woglinde: my gp is charging so it is not dead May 07 21:15:50 hehe May 07 21:15:50 xjqian: Wrong bug number. Actually, 2248 would have been OK. But please don't overuse 2194. The only effect of that will be that the core devs involved will ignore that bug. May 07 21:15:52 as I said May 07 21:15:59 woglinde: you didn't install squeak for me on the n770 May 07 21:16:22 New Intel processor for mobile devices? https://lists.ubuntu.com/archives/ubuntu-devel-announce/2007-May/000289.html May 07 21:16:34 likewise: check linuxdevices.com May 07 21:16:34 zecke I didnt charge it anyway May 07 21:16:38 likewise: umpc May 07 21:16:39 I am sucker you know May 07 21:16:52 no I don't :} May 07 21:17:56 woglinde: man, OpenOffice is bad May 07 21:18:48 why? May 07 21:18:58 I likt it since staroffice 3.0 May 07 21:19:07 woglinde: I'm searching the blog post. Ever tried to save in a different format? May 07 21:19:12 zecke: try word for a while and you will love it again... May 07 21:19:14 woglinde: huge and slow ... just like M$ office May 07 21:19:23 s/it/openoffice/ May 07 21:19:34 woglinde: two pages of text because they suck in using a proper labeling on their buttons May 07 21:20:01 ??? May 07 21:20:01 woglinde: I think if I ever feel bad, I will show it to el to make here scream and run-a-way May 07 21:20:30 maybay you did something wrong or dont understand to make it right May 07 21:21:27 woglinde: well, I doubt that! May 07 21:21:39 woglinde: It starts with that I can export a ODS to PDF, XHTML but not to CSV May 07 21:22:13 woglinde: so I have to use Save As to save it as CSV, then it asks me the three pages long, badly written, badly formatted question if I really want to save in a foreign format May 07 21:22:38 woglinde: it feels like winzip randomly changing the button order :) May 07 21:23:47 what the heck have buttons to do with cvs? May 07 21:24:00 re koen May 07 21:25:41 hey woglinde May 07 21:25:41 woglinde: csv != cvs May 07 21:25:41 woglinde: do you want to see a picture, or just save a table as csv May 07 21:25:53 woglinde: and ask yourself. "Do you want to read such a crap?" "Do you think your mother will understand it?" May 07 21:27:31 slashdot's add revenue must be going down: http://slashdot.org/articles/07/05/07/1953250.shtml May 07 21:27:38 that's just pure flamebait May 07 21:30:00 zecke aeh yes May 07 21:30:06 * philippe wonders what is wrong with slashdot and other similar sites. The content gets worse every day May 07 21:30:22 as if it's ever been good. May 07 21:30:35 philippe: VALinux has financial trouble? May 07 21:30:55 zecke aeh csv is nothing for eyes only for scriptinglangugaes which dont have native acess to office formats May 07 21:30:58 zecke: I wouldn't know.... May 07 21:31:40 woglinde: so GNU R, GNU plot, GNU octave are solely scripting languages ;) May 07 21:32:15 and I'm sure there's a lot of csv2* May 07 21:33:04 neat, hal support for MD: http://blog.fubar.dk/?p=89 May 07 21:34:20 * koen ponders about a web interface for hal May 07 21:41:06 * zecke searches kinderschokolade May 07 21:41:20 überraschungs eier? May 07 21:42:12 woglinde: just the normal one May 07 21:43:43 Hi all: I'm struggling with libc-headers -> http://rafb.net/p/nwKTzu81.html Any ideas? May 07 21:44:37 geekoid_work: start with the first line of ERROR: May 07 21:45:01 geekoid_work: start reading with the first line containing the word "ERROR" May 07 21:45:04 03koen 07org.oe.dev * r648c7b46... 10/ (3 files in 2 dirs): linux-magicbox: add COMPATIBLE_MACHINE May 07 21:45:32 geekoid_work: and your log looks truncated May 07 21:45:56 geekoid_work: that might be either due python2.5 or human interference May 07 21:51:51 zecke: I checked the work logs and can't find where the QA error is occurring, I was hopeful someone else had run into this.. May 07 21:52:52 geekoid_work: clean the libc header stuff (-cclean) May 07 21:53:09 geekoid_work: and run bitbake with python2.4 `which bitbake` your_args May 07 21:53:45 geekoid_work: but it sounds weird that there are no errors printed but it complains about fatal errors May 07 21:54:57 zecke: indeed, log.do_install & configure have no errors. May 07 21:55:35 zecke: I'll try your suggestions, thanx May 07 22:05:29 03polyonymous 07org.oe.dev * r286d8ee2... 10/ (3 files in 3 dirs): packages/psplash: added psplash-zap package with initscript to quit psplash before going for login prompt on console image. May 07 22:07:59 does CIA's "3 files in 3 dirs" have to do with anything? :) May 07 22:09:23 koen, I'm not sure where would I add psplash-zap to console-image? There are many places, would it go to the angstrom-console-image.bb itself? (it should not be added to, say, x or opie image) May 07 22:14:04 re May 07 22:17:25 redump May 07 22:22:22 polyonymous: google for "polyonymous cia" and look at what files have been touched. You will understand. I always forget to include the packages dir in my mental counting. The numbers are correct May 07 22:28:48 Laibsch, okay, that would justify 3 dirs, but 3 files? May 07 22:31:14 monotone can't count properly May 07 22:31:47 polyonymous: Do the google search May 07 22:32:28 there now May 07 22:32:37 Laibsch: http://cia.vc/stats/project/openembedded/.message/407192 <-- that's the commit. May 07 22:32:44 it's not three files *and* three dirs May 07 22:32:58 two dirs and two files, I'd say. May 07 22:33:38 yes, but you can say three dirs, because it's packages, psplash and files. But three files?? May 07 22:34:17 is that a file named "1"? May 07 22:34:31 no way May 07 22:34:40 look at the previous commit May 07 22:34:41 http://cia.vc/stats/project/openembedded/.message/4070ac May 07 22:34:50 koen's commit has three files and two dirs May 07 22:34:55 okay, it's some meta then. May 07 22:35:11 mtn doesn't store metadata in the tree May 07 22:35:31 and at any rate, it shouldn't count towards the commits May 07 22:35:34 yes, there's no such file. But it's some meta that comes with commit info. May 07 22:35:55 Well, just was surprised. Now I know about it :) May 07 22:36:24 if the "1" is removed, the dir count should be correct too May 07 22:37:03 yup. I'd call this "1" a bug, but know that it's known, it's a feature :) May 07 22:48:39 Where in the bitbake recipe directive order should REPLACES appear in a recipe? May 07 22:49:03 I can't find the documentation for it May 07 22:49:45 RREPLACES should be near to RPROVIDES May 07 22:55:25 ugh...gpe-conf is failing to build May 07 22:55:27 oh well, don't need it May 07 22:58:52 mickeyl: thanks. Is the value just the ${PN} of the package to be replaced or the ${PF} or something else? May 07 22:59:04 PN May 07 22:59:13 K May 07 22:59:32 although you can use anything ipkg understands May 07 22:59:51 i.e. PN (version), PN (>= version), etc. also works for RREPLACES May 07 22:59:51 the openpbx.org project just changed names to CallWeaver so I'm experiencing the joy of redoing the packaging. Fun fun May 07 22:59:55 hehe May 07 23:00:10 so it is the fork of the fork? May 07 23:00:19 It's a rename May 07 23:00:26 not a fork May 07 23:00:29 thank god May 07 23:01:08 of course they haven't updated all the filenames yet... what a joy May 07 23:01:22 the built mostly works... May 07 23:04:47 polyonymous, Laibsch: All works fine here ;( How could patches not work for some people? May 07 23:05:21 zecke: Debian source packages? I don't think that was me looking at those. May 07 23:05:59 v8jlene, no idea, works for me as well. Do we have conditionally applied patches? May 07 23:06:23 polyonymous: Nope, all of them are always applied May 07 23:06:43 v8jlene, ah, one issue I've had - I've had perl-lib instead of old libperl5. May 07 23:06:47 ipk that is. May 07 23:07:06 v8jlene: ah okay :) May 07 23:09:50 polyonymous: tmp/deploy/ipk/sh4/libperl5_5.8.7-r22_sh4.ipk *sigh* that shouldn't exist, there's a DEBIAN_NOAUTONAME_perl-lib = "1" in the .bb which should stop perl-lib being renamed. May 07 23:10:51 v8jlene, so it's intentional? May 07 23:10:55 Oh, that's old perl. May 07 23:11:09 For new perl there is only perl-lib May 07 23:11:16 yes, libperl5 is the old one. May 07 23:11:46 Okay, just wanted to make sure. It's a problem for ipkg upgrade - fine with me, but not sure about other users. May 07 23:11:59 Is that intentional... hmmm... well I think I intentionally did what the old perl did but it appears I was mistaken in what the old perl was doing.. May 07 23:12:40 well... Then it's up to you to decide, I just had to report :) May 07 23:14:50 polyonymous: Yeah, I'll change it back... but then I have to worry about people using 5.8.8 and upgrading I suppose. May 07 23:15:10 v8jlene, yes, no-win situation :) May 07 23:15:46 polyonymous: Well, 5.8.8 is still marked as default_preference -1! (Not for much longer though) May 07 23:16:06 v8jlene, "still" - as in "as of yesterday?" :) May 07 23:16:51 I mean, I didn't build it intentionally, because I'm interested in newer perl, just happened to bitbake image after update :) May 07 23:17:16 nigght May 07 23:17:36 night May 07 23:17:45 polyonymous: If your have distro=angstrom then I think it was updated yesterday to force the use of the new perl, which would be why you get it now May 07 23:18:08 v8jlene, ah, yes, very likely. And yes, I do have angstrom as a distro. May 07 23:19:12 * v8jlene wonders what's happened to the water supply... I need coffee and have no water May 07 23:19:48 I drunk all the water up. May 07 23:20:45 anyway, goodnight everyone. May 07 23:21:10 polyonymous: Thanks for the report! Goodnight. May 07 23:21:21 You're welcome ;) May 07 23:21:24 * polyonymous is gone May 08 00:58:31 03lenehan 07org.oe.dev * rc123191b... 10/ (1 packages/perl/perl_5.8.8.bb): May 08 00:58:31 perl 5.8.8: Allow debian.bbclass to mess with perl-lib's naming. This will May 08 00:58:31 cause issues with upgrading from earlier 5.8.8 versions (perl-lib will May 08 00:58:31 remain instead of upgrading to libperl5) but it'll actually make upgrades May 08 00:58:31 from 5.8.7 work. I can't see how to fix the 5.8.8 upgrading easily. The May 08 00:58:31 renaming was disabled due to a mistake on my part. May 08 00:58:35 03lenehan 07org.oe.dev * r9aedb40b... 10/ (3 files in 2 dirs): May 08 00:58:37 perl/perl-native 5.8.8: Remove -DEFAULT_PREFERENCE = "-1" for version 5.8.8. May 08 00:58:39 Some distributions have already started to use this so it's probably about May 08 00:58:41 time to promote it to the default. May 08 01:04:53 laibsch: got you, I will not overuse meta-bug 2194 for trivial fixes May 08 01:11:18 03lenehan 07org.oe.dev * r431af161... 10/ (40 files in 2 dirs): May 08 01:11:18 perl modules: Bump PR since the directories they install into have changed May 08 01:11:18 with the new version of perl. **** ENDING LOGGING AT Tue May 08 02:59:57 2007