**** BEGIN LOGGING AT Tue Oct 06 02:59:58 2009 Oct 06 07:09:38 Hi Oct 06 07:10:26 A question : linux_2.6.31.bb contains DEFAULT_PREFERENCE = "-1" still it insists on building it for my powerpc target. Why is that ? Oct 06 07:14:36 It seems it is not respecting the DEFAULT_PREFERENCE Oct 06 07:25:25 good morning Oct 06 07:34:55 Got it. DEFAULT_PROVIDER_virtual/kernel was set to "linux" it had to be linux-mpc512x to avoid standard kernels ...... Oct 06 07:41:18 How do I configure busybox for oe? Manually edit it's defconfig? Oct 06 07:44:45 jovox mostly yea Oct 06 07:45:25 bitbake -c patch busybox; cd tmp/work/arch*/busybox* ; make menuconfig Oct 06 07:45:42 and then replace the defconfig with the one you'd like to use Oct 06 07:45:56 one of a few ways to do it, but that should be pretty straightforward Oct 06 07:46:04 oki, thanks Oct 06 07:46:14 yea, and then bitbake -c clean busybox Oct 06 07:46:56 if you only edit it there, it won't really be persistent through multiple builds, especially if you inherit rm_work Oct 06 07:47:26 I'm a bit confused here.. I want mdev to be included in my image. As I understand it mdev is a part of busybox but there is also a mdev recipe available.. Sould I add that package or reconfigure busybox? Oct 06 07:48:30 hey sgh, where did you find a freescale board? Oct 06 07:48:43 is there a vendor im not aware of :/ Oct 06 07:49:15 m4t: you mean where I got the board ? Oct 06 07:50:33 good morning Oct 06 07:50:34 morning Oct 06 07:51:13 sgh yea, i like ppc Oct 06 07:53:13 ppc... I miss it on my desk Oct 06 07:54:09 but why all ppc devboards are expensive or impossible to get Oct 06 07:54:33 yea Oct 06 07:54:44 its why im asking, maybe someone knows about something i don't Oct 06 07:55:15 m4t: I have this one http://store.silicontkx.com/product_info.php?products_id=28&osCsid=6425a9951166c60e77bab2aca0ece349 Oct 06 07:56:10 yep.. 999$ Oct 06 07:57:58 nice Oct 06 07:58:49 I would rather spend that on machine for builds Oct 06 08:00:09 hi florian pvanhoof Oct 06 08:04:09 hey Oct 06 08:09:50 I'm trying to figure out how to work with OE in-house. Multiple developers will be using it, and we need strict control over versioning. can I clone, do an initial build, and do *something* that enables the other devs to compile without downloading sources from the net? Oct 06 08:11:11 tasslehoff: yes. you can share DL_DIR via nfs, you can also make premirror from it Oct 06 08:14:16 good morning Oct 06 08:14:17 florian: alive? Oct 06 08:14:20 My board hangs at this message for a long time "[ 37.862776] Looking up port of RPC 100005/1 on 10.0.1.82" Any ideas on how to solve it, Oct 06 08:14:29 sgh: nfsroot boot? Oct 06 08:14:38 hrw: yes Oct 06 08:14:46 hrw, cool. do you know of any good resources to find out more about this? Oct 06 08:14:55 dth: hi Oct 06 08:15:02 hrw: sure Oct 06 08:15:12 sgh: check IP and ethernet mac Oct 06 08:15:28 florian: forwarded you mail with address for flyers Oct 06 08:15:31 hrw: it boot allright. It just hangs at that message for at long time. Oct 06 08:16:03 sgh: so check network or nfs server setup Oct 06 08:17:09 florian: I have a problem. PV="1.0+svnr${SRCREV}" + INC_PR="r3" + PR="${INC_PR}.3" == disaster if SRCREV go back. Oct 06 08:17:56 florian: looks like better would be PV="1.0" INC_PR="r3" PR="${INC_PR}.3+svnr${SRCREV}" but how to handle that in one place instead of setting it in each recipe... Oct 06 08:19:17 tasslehoff: 1. to share DL_DIR over nfs/smb you need nfs-server or samba server. 2. for premirror stuff you need to check PREMIRRORS variable in bitbake.conf to get info how it is done and add own value for it. I never tried that Oct 06 08:19:32 tasslehoff: I usually share DL_DIr over network for my machines Oct 06 08:21:09 florian: ideas? Oct 06 08:29:16 hrw, I think sharing of DL_DIR is my first choice as well. can I put something in local.conf to prevent OE from fetching sources online? I want the other devs to get a compile error, and then come crawling to me. ;-) Oct 06 08:29:37 hrw: not really... i have to admit that i'm not really a fan of INC_PR at all. Oct 06 08:30:29 hrw: worst case we just set the version in avery file. not nice but currently i don't have a better idea Oct 06 08:32:23 hrw: The flyers will leave kc with the other stuff we ship today. From the experience it takes a few days only. Oct 06 08:35:47 florian: great (flyers) Oct 06 08:36:08 tasslehoff: no idea, but you can firewall devs (ugly) Oct 06 08:36:22 florian: I see bonus side of inc_pr Oct 06 08:38:23 hrw, :) Oct 06 08:44:07 If I have control over the DL_DIR and the openembedded directory that should be ok. What is the recommended way of locking versions for all packages? Oct 06 08:44:43 tasslehoff: look at conf/distro/include/sane-srcrevs.inc and sane-srcdates.inc Oct 06 08:45:25 tasslehoff: lockdown.bbclass will generate a .conf file fragment for you to freeze all the packages at their current versions. Oct 06 08:45:59 obviously you also need to take some care to make sure that versions you are relying on don't disappear from the repository or get broken, but if you are working on a branch then this shouldn't be an issue. Oct 06 08:46:28 ah. great :) Oct 06 08:46:35 things are falling into place :) Oct 06 08:50:16 pb_: morning there Oct 06 08:57:55 hrw: How does the defconfig-stuff work. How does oe determine the defconfig file to use ? Oct 06 08:58:50 sgh: via overrides stuff Oct 06 08:59:02 it is in manual Oct 06 08:59:49 hrw: ok Oct 06 09:01:52 if I understand this correctly, mdev -s shall create nodes in /dev right? When I run it nothing happends.. No error messages or anything and no nodes are created. Oct 06 09:02:22 strange.. worked for me Oct 06 09:02:41 and I find very little information about mdev Oct 06 09:03:49 not even 'mdev' without any arguments prints anything out Oct 06 09:06:27 hrw: It's not completely clear to me ....the recipes/linux/linux_2.6.31/ contains no defconfig and linux_2.6.31.bb mensions just file://defconfig. Oct 06 09:07:27 could mdev possible be broken? Oct 06 09:08:00 at least it should print it's "usage" info Oct 06 09:08:47 jovox: busybox-mdev package? Oct 06 09:09:28 sgh: linux_2.6.31.bb recipe can take defconfig from: files/ files/MACHINE/ linux/ linux/MACHINE/ linux-2.6.31/ linux-2.6.31/MACHINE/ Oct 06 09:13:43 sgh: clearer now? Oct 06 09:21:31 florian: xserver-common 1.32 has one difference to my 1.30+patches - xmodmap files are in /etc/X11/ where my has them in /etc/X11/xmodmap/ Oct 06 09:22:24 hrw: strange... i wonder where we lost this Oct 06 09:24:13 yep Oct 06 09:33:45 florian: good morning Oct 06 09:40:20 Laibsch: こんにちは Oct 06 09:40:46 Hehe Oct 06 09:40:50 Where did you learn that? Oct 06 09:40:56 I just checked into the hotel Oct 06 09:41:01 and I'm very tired Oct 06 09:41:17 Laibsch: powered by google this time Oct 06 09:41:26 the wonders of the internet ;-) Oct 06 09:41:39 Laibsch: Japanese is not for me so far Oct 06 09:41:39 I hope things are going well for you Oct 06 09:41:43 yes Oct 06 09:41:47 good Oct 06 09:41:55 need to learn 'I do not understand French" in French Oct 06 09:41:55 OEDEM coming up soon? Oct 06 09:41:59 yes Oct 06 09:42:09 Je ne comprend pas le Francais Oct 06 09:42:12 the c is the French c Oct 06 09:42:36 Well, I certainly won't make OEDEM this year, then Oct 06 09:43:10 Laibsch: gracias Oct 06 09:43:26 de nada Oct 06 09:46:42 03Steffen Sledz  07org.openembedded.dev * r1c7fac5c2a 10openembedded.git/recipes/linux/linux-2.6.24/hipox/defconfig: linux-2.6.24: enable root nfs support for hipox machine Oct 06 09:46:54 hi XorA Oct 06 09:50:11 mickey_away: ping Oct 06 09:53:43 hrw: no ... does oe have some kind of search path when it selects the files to be copied into WORKDIR Oct 06 09:54:09 sgh: it has Oct 06 09:54:19 sgh: read oe manual and check for overrides in it too Oct 06 09:54:37 hrw: ok .. Oct 06 09:55:30 hrw: ah ... "Finding local files " :D Oct 06 10:04:02 pb__ : should I clone the repo on a server, create a local branch and then have the other devs clone my repo? Oct 06 10:04:35 to make sure used versions don't disappear, like you said Oct 06 10:13:07 m4t, I edited the defconfig for busybox-static which I use in my initramfs, did a clean busybox-init. But mdev is not being installed. Oct 06 10:13:23 jovox? Oct 06 10:13:37 mdev is link to busybox now Oct 06 10:13:44 old crappy mdev recipe is gone Oct 06 10:15:38 woglinde, I saw there is a mdev recipe. I used that but it did not seem to work.. It compiled fine but nothing happend when I executed it.. So I tried to add mdev to busybox by enabeling it in busybox's defconfig. But it seems like it didnt build it Oct 06 10:16:02 ???? Oct 06 10:16:16 the function is not available (cmd busybox mdev) Oct 06 10:16:39 I'm using the static recipe.. Don't know if that has anything todo with this Oct 06 10:16:41 check do_compile Oct 06 10:17:03 hi woglinde l Oct 06 10:17:07 hi ant Oct 06 10:19:23 did you check that for the static recipe mdev is enabled? Oct 06 10:20:11 woglinde: no, but we could find a bug in get_command_line Oct 06 10:20:24 ant I meant jovox Oct 06 10:20:31 woglinde, hmm what do you mean? I saw is was disabled in it's defconfig Oct 06 10:20:34 and enabled it Oct 06 10:20:41 woglinde: as you suspected... Oct 06 10:20:55 I'm rebuiling now and will take a look at the logs when it's done Oct 06 10:21:00 jovox did you check that the right config was put into the build-dir? Oct 06 10:22:19 no, I will do that Oct 06 10:22:33 I just assumed it was the latest version Oct 06 10:23:41 I see now it's openembedded/recipes/busybox/busybox-1.2.1 Oct 06 10:23:48 which is not the one I edited Oct 06 10:26:23 :) Oct 06 10:32:58 I'll make my last Q more general. If multiple devs in-house use OE and I want to "control them", should I clone the remote OE repo to a server, and let the other devs clone my local repo? Oct 06 10:33:55 hrw, by the way. Did you ever have a chance to look at those xorg issues? I hope I can find the time to do some more testing of qemu images while I'm here. Oct 06 10:34:55 he laibsch Oct 06 10:35:06 hi woglinde Oct 06 10:35:09 how is it going? Oct 06 10:35:22 * woglinde is working on lighty Oct 06 10:35:33 what is that= Oct 06 10:35:47 ? Oct 06 10:36:05 better webserver than apache Oct 06 10:36:16 lighttpd Oct 06 10:36:39 i see Oct 06 10:36:41 good luck Oct 06 10:36:52 I am almost there Oct 06 10:37:00 Laibsch: sorry but not ;( Oct 06 10:37:04 hm who is activating patchwork accounts? Oct 06 10:37:15 woglinde: Crofton is one Oct 06 10:37:30 ping crofton Oct 06 10:41:06 * * OE Bug 5325 has been RESOLVED (FIXED) by subhashini.venugopalan(AT)gmail.com Oct 06 10:41:08 * * clutter uses mesa instead of virtual/libgl Oct 06 10:41:10 * * http://bugs.openembedded.net/show_bug.cgi?id=5325 Oct 06 10:41:25 Does my local.conf have to be in "$OEBASE/build/conf/local.conf", or can I put it somewhere else? I want a toplevel in OEDIR with "openembedded", "tmp" and a "mydir", that contains my local.conf and possibly my own custom recipes. Oct 06 10:42:34 tasslehoff: http://marcin.juszkiewicz.com.pl/2006/11/22/my-openembedded-environment-ii/ Oct 06 10:42:42 tasslehoff: yeah, something like that. or you might be able to just create your branch on the main oe server, I don't quite know what the policy is for that. Oct 06 10:42:56 hi pbn Oct 06 10:42:58 ups pb Oct 06 10:43:01 hi woglinde Oct 06 10:46:42 done Oct 06 10:46:46 time for lunch Oct 06 10:46:48 03Henning Heinold  07org.openembedded.dev * r2403fd5344 10openembedded.git/recipes/lighttpd/lighttpd.inc: Oct 06 10:46:48 lighttpd: starting depends on three modules Oct 06 10:46:48 * indexfile, dirlisting, staticfile are needed according our config, Oct 06 10:46:48 so let RDEPENDS on them Oct 06 10:46:48 * found out by Michael Lippautz Oct 06 10:46:52 * bump INC_PR Oct 06 10:46:54 03Henning Heinold  07org.openembedded.dev * r46243e2ac2 10openembedded.git/recipes/lighttpd/ (lighttpd.inc lighttpd_1.4.18.bb lighttpd_1.4.23.bb): lighttpd: switch to .inc file and INC_PR Oct 06 10:46:57 03Henning Heinold  07org.openembedded.dev * r6c5f18b931 10openembedded.git/recipes/lighttpd/lighttpd.inc: Oct 06 10:47:00 lighttpd: inherit gettext Oct 06 10:47:02 * remove --disable-nls from configure line Oct 06 10:47:04 * bump INC_PR Oct 06 10:51:18 pb__: I did finally compile that damn'd kexec for klibc, but have 141592b size... Oct 06 10:51:36 ant lol Oct 06 10:51:42 so klibc suckz Oct 06 10:51:49 okay lunch now Oct 06 10:52:06 or it is the getline reworking Oct 06 10:52:09 I will test it Oct 06 10:52:10 it's 27% bigger size than 1.101 Oct 06 10:52:12 but later Oct 06 10:52:29 woglinde: getline or the /* strip newline */ Oct 06 10:52:31 +- *(line + strlen(line) - 1) = 0; Oct 06 10:52:32 ++ line[strlen(line) - 1] = '\0'; Oct 06 10:52:45 strange it works with getline / glibc Oct 06 11:00:50 ah, ok, I see.. Oct 06 11:00:57 hrw, thanks. Oct 06 11:08:19 re Oct 06 11:09:55 ant could please seperate out the getline patch? Oct 06 11:17:45 ant_work: yah, I think the way to real success is to throw away kexec-tools and reimplement those bits from scratch. Oct 06 11:17:54 *g* Oct 06 11:18:11 anyone come across libtool prepending an 'X' to all its options? Oct 06 11:18:36 mike_cw ???? Oct 06 11:18:43 i read that this is a difference between the version used to create ltmain.sh and the libtool version, but the recipe seems to have generated ltmain.sh Oct 06 11:19:01 then remove the cruft Oct 06 11:19:01 woglinde: e.g. ./arm-angstrom-linux-gnueabi-libtool: line 814: X--mode=compile: command not found Oct 06 11:19:10 yeah, I think that can happen if you have some internal libtool inconsistency Oct 06 11:19:13 which recipe is this? Oct 06 11:19:21 yes, remove all m4 macros Oct 06 11:19:22 it's a custom version of libsdl Oct 06 11:19:35 does the libsdl from the repository do the same thing? Oct 06 11:19:38 and autoreconfigure will put all needed stuff in Oct 06 11:19:59 hm let me think which package I last time fixed Oct 06 11:20:05 then you will get a clue Oct 06 11:20:30 pb__: don't think so but this is a newer version (1.2.13) Oct 06 11:21:34 we have an extra section in configure.in to add a new graphics driver, plus the driver code itself. the rest of the build system should match upstream Oct 06 11:22:50 mike_cw look at /devel/arm/git/openembedded/recipes/apr/apr_1.2.12.bb Oct 06 11:22:56 do_configure_prepend Oct 06 11:23:12 do it for the m4 and libtool stuff in sdl too Oct 06 11:23:20 but beware of selfwritte m4-files Oct 06 11:24:26 haha Oct 06 11:24:35 we have it in some sdl recipes to Oct 06 11:24:41 /recipes/libsdl/libsdl-image_1.2.6.bb Oct 06 11:24:44 i'm using acinclude.m4 from the oe recipes Oct 06 11:24:57 i'll delete aclocal.m4 and leave sdl.m4 Oct 06 11:28:18 same thing Oct 06 11:29:40 then you have to delete more Oct 06 11:30:21 libtool.m4 Oct 06 11:30:35 not present Oct 06 11:30:49 ltconfig acinclude.m4 libtool ltmain.sh Oct 06 11:30:55 the only other .m4 file is sdl.m4 which i guess it important Oct 06 11:31:01 s/it/is/ Oct 06 11:31:05 yes Oct 06 11:31:10 it will be installed Oct 06 11:31:17 look at the files I posted Oct 06 11:39:18 deleting acinclude.m4 (and not replacing it with the one from the oe recipe) breaks configure Oct 06 11:39:28 and then there's nothing else to delete Oct 06 11:39:34 time for lunch :) Oct 06 11:39:58 woglinde: thanks for your help so far Oct 06 11:40:15 maybee post your recipe later Oct 06 11:40:59 he mickeyl Oct 06 11:41:09 hi guys Oct 06 11:41:20 elo mickeyl Oct 06 11:41:32 greetings hrw Oct 06 11:41:43 * mickeyl back from vacation Oct 06 11:41:43 hi mickeyl Oct 06 11:41:48 boy is that cold... Oct 06 11:41:50 hey pb__ Oct 06 11:41:57 where did you go? Oct 06 11:42:07 two weeks canarian islands, fuerteventura Oct 06 11:42:19 avg. 25-29 C per day Oct 06 11:42:20 oh, very good. Oct 06 11:42:34 heh, yeah, must be a bit of a shock being back in germany Oct 06 11:43:26 finally mdev got installed Oct 06 11:54:59 I don't fins much information about configuring mdev. Any suggestions? Oct 06 11:55:10 ie, syntax for mdev.conf Oct 06 12:04:12 jovox: busybox/docs/mdev.txt? Oct 06 12:08:48 hrw: relativly to what? Oct 06 12:10:16 oh.. mdev creates tons of nodes in /dev.. Thats not exactly what I wanted :) Oct 06 12:10:25 heh, what did you want? Oct 06 12:10:28 that's kind of what mdev is for Oct 06 12:10:30 jovox: in busybox source Oct 06 12:13:47 pb__ Im trying to find a way to determine if there is a usb stick attached at boot. If it is I will read some files from it before I do a switch_root to my "main" rootfs. The problem I have is that the usb is not loaded at the moment my /init script is executed. So I want a script to be called when it is ready, try to mount it, do some work and then switch_root Oct 06 12:14:26 I to remove console output from my framebuffer. What's needed is to change "CONFIG_FRAMEBUFFER_CONSOLE=y" to "# CONFIG_FRAMEBUFFER_CONSOLE is not set" in the defconfig of the kernel, but do I add this patch? Oct 06 12:14:26 right Oct 06 12:14:30 mdev can run scripts on evice creation Oct 06 12:14:30 I'm not quite sure how that relates to mdev Oct 06 12:14:31 I did put a sleep in my initscript which worked but it in my point of view that was ugly.. Oct 06 12:15:02 pb_ using hotplug? Oct 06 12:15:11 if you just want to be notified about device creations, and don't want mdev to create any device nodes, you probably just want to install your own hotplug handler. Oct 06 12:15:13 can someone please forward me a sample ipkg.conf which could satisfy dependencies if all the needed *.ipks are on an sdcard? Oct 06 12:15:36 creating device nodes is pretty much mdev's whole raison d'etre, it'd be fairly pointless to install it if you don't want that function. Oct 06 12:16:15 ultimately though you are still going to have to have a sleep in your initscript, so I'm not quite sure what the hotplug arrangement will buy you. Oct 06 12:16:34 pb__ ok. I have no idea on writing my own hotplug handler Bu if you don't think it's too complicated I'll take a look at it Oct 06 12:16:35 morning all Oct 06 12:16:42 XorA: ping Oct 06 12:16:52 has anyone booked hotels for OEDEM yet? Oct 06 12:16:54 no, it's pretty easy. the kernel just runs those scripts with a bunch of environment variables to say what is happening. Oct 06 12:17:09 but, as I say, I'm not convinced that this will actually do much to solve your real problem. Oct 06 12:17:11 RP: not me Oct 06 12:17:24 pb_ you are right.. I cannot quit the initscript before the usb is loaded.. Oct 06 12:17:41 RP: YHA for me and woglinde Oct 06 12:17:45 nhg: src sdcard file:///media/card/ipk/ if that still works since the change from ipkg to opkg Oct 06 12:17:50 right. even with hotplug, you will still have to sleep to give hotplug a chance to do its thing. Oct 06 12:18:03 yeah Oct 06 12:18:07 and, if you're going to do that, you might as well just sleep and then do the usb check in the foreground as part of the same script. Oct 06 12:18:17 seems like hotplug just adds complexity for no real benefit Oct 06 12:18:24 ok Oct 06 12:18:38 XorA: Would this be documented anywhere? Oct 06 12:18:49 XorA: any idea on what you'll do? Oct 06 12:18:57 RP: not sure yet Oct 06 12:19:08 RP: still deciding whether to drive down Oct 06 12:19:11 OEDEM? Oct 06 12:19:28 nhg: OE European Conference Oct 06 12:19:38 I'll just add a loop from which I check the /proc/partitions to see if the usb is in there. if not, sleep a second, if there is one: mount it Oct 06 12:19:49 nhg: OE Developers European Meeting to be exact Oct 06 12:20:02 oh..ok Oct 06 12:20:08 XorA: ah. I was assuming people would be taking the train Oct 06 12:20:26 nhg: and Ive no idea about opkg docs, it used to be maintained by OM, then tick on his own, not sure now Oct 06 12:20:43 RP: costs for train/petrol seem to be about the same Oct 06 12:20:49 XorA: OM? Oct 06 12:20:57 ~OM Oct 06 12:20:58 om is, like, a trascendental belch Oct 06 12:21:29 ibot: OM is also short for Openmoko (Inc.), open mobile communications devices and software Oct 06 12:21:30 mickeyl: okay Oct 06 12:21:30 XorA: right, makes sense. Are you coming with lrg? Oct 06 12:21:53 RP: no idea, maybe he want to travel in his Porsche :-) Oct 06 12:21:54 I would recommend the train over driving if the costs are the same, it isn't a terribly enjoyable drive. Oct 06 12:22:25 there seem to be perpetual roadworks on the A1 and you can get stuck there for some time if you are unlucky Oct 06 12:22:25 pb__: I also have familly down there, so some flexibility to visit Oct 06 12:22:32 I could take the bike but I'd just destroy the tyres... Oct 06 12:22:33 ah, fair enough Oct 06 12:23:45 * XorA wonders what is with the continual attempts to phish his mobile phone account are for Oct 06 12:23:58 RP: isn't teleport yet available at Intel? Oct 06 12:24:39 ant_work: sadly not Oct 06 12:26:02 mickeyl: Did you pick somewhere to stay? Oct 06 12:26:15 pb_: Any recommendations (or places to avoid)? Oct 06 12:26:50 RP: not yet, I just returned from vacation. I want to book soon, so I'll have to check now Oct 06 12:26:56 RP: not really, I don't have any experience of staying in the local hotels. I added a few suggestions to the wiki page but I can't really speak to which ones are good or bad. Oct 06 12:27:08 rather depends on how much money you want to spend, I guess. Oct 06 12:27:43 i'm now in the age where i want a bit of comfort ;) Oct 06 12:28:00 no more formula1 or ibis hotels for me Oct 06 12:28:02 ;) Oct 06 12:28:12 right, me too :-} Oct 06 12:28:14 mickeyl: holiday inn is close iirc Oct 06 12:28:21 hrw: that sounds good Oct 06 12:29:07 yeah, there is a "holiday inn express" within easy walking distance (about 300m) from our office. Oct 06 12:29:18 various other hotels in the city centre as well Oct 06 12:29:18 pb__, mickeyl: I think I'm falling into that category too these days... Oct 06 12:29:32 RP: welcome, old fart ;) Oct 06 12:29:56 Those tend to be OK Oct 06 12:30:25 * RP will think about the holiday inn express he was in last week Oct 06 12:30:33 will not Oct 06 12:30:55 the only hotel I've actually stayed in here is http://www.bw-cambridgequymill.co.uk/, which was fine but is not especially close to the office. Oct 06 12:31:10 pb__: Is the one near the office, the Coldhams park one? Oct 06 12:31:55 probably about 2.5km away at a guess, maybe not recommended unless you like walking. Oct 06 12:31:58 RP: yup, that's the one Oct 06 12:32:53 there is also another holiday inn in histon, which is supposed to be slightly nicer (and is closer to the beer venue!) but not so convenient for the office, plus yet another holiday inn near the train station. Oct 06 12:33:23 I don't know anybody who's stayed in that last one so I can't speak to its qualities. Oct 06 12:34:28 The travel tool recognises the holiday inn express so thats looking likely for me, assuming I get travel approval Oct 06 12:34:47 The deadline for aprovals was about 3 months ago :/ Oct 06 12:34:54 heh Oct 06 12:35:45 pb__ it's pretty difficult to know for how long I should sleep before I give up Oct 06 12:36:14 jovox: that's indeed an impossible mission Oct 06 12:36:24 do a loop and break it on event Oct 06 12:37:28 fwiw I had different delays with different SD cards...and I tend to think it would be the same with USB dongles Oct 06 12:37:38 ant_work, yes, but I only use usb for upgrading. So most of the cases there will be no usb stick attached will will make the boot slow Oct 06 12:37:48 then set 999 :=) Oct 06 12:38:02 ant_work?` Oct 06 12:38:10 and tell users: 'go eat and come back later' Oct 06 12:38:22 hehe Oct 06 12:38:32 I there was a monitor attached that would be ok :) Oct 06 12:39:13 so what's your recomendation? Looping for a minute, then give up and switch_root? Oct 06 12:39:37 s/switch_root/do something else Oct 06 12:40:03 as pb said, do a loop and check about the presence of the device in sysfs on each iteration Oct 06 12:42:22 yeah, basically. a minute is probably too much, but I guess you should loop for at least 10-15 seconds. Oct 06 12:42:39 ok Oct 06 12:42:59 I'm a complete newby here.. sysfs? Oct 06 12:43:28 you could do a two-step thing: first check to see whether the usb device has appeared at all (maybe waiting only 5-10 seconds for that to happen) and then, if yes, do a second wait for the actual hard disk node to appear. Oct 06 12:43:46 that would allow you to avoid slowing down the normal boot too much, while still giving a reasonable amount of leeway for slow scsi settling time. Oct 06 12:44:21 pb__, and how do I check if the driver has been loaded? Oct 06 12:44:58 is sounds like an ok idea to my btw Oct 06 12:45:00 look in /sys/bus/usb/devices Oct 06 12:46:56 check if it's empty or not? Oct 06 12:46:58 03ghost  07org.openembedded.dreambox * r8a1b3385e4 10openembedded.git/packages/dreambox/dreambox-dvb-modules.bb: dreambox-dvb-modules.bb: fix typo Oct 06 12:46:59 03Felix Domke  07org.openembedded.dreambox * r7c1ca0a886 10openembedded.git/: Merge branch 'org.openembedded.dreambox' of git://git.opendreambox.org/git/ghost/opendreambox-1.5 into org.openembedded.dreambox Oct 06 12:47:37 right Oct 06 12:47:49 assuming you don't have any other usb devices in normal operation Oct 06 12:48:28 right, I don't Oct 06 12:50:18 so, I loop until the driver has been loaded. ie there are files in /sys/bus/usb/devices. As soon as they appear I loop for another (lets say) 10 and checks /proc/partitions to see if any new sdas appear? Oct 06 12:53:34 right, exactly Oct 06 12:53:48 greatm thanks! Oct 06 12:53:54 great even Oct 06 13:03:59 gstreamer-ti adds dsplink/ti-codec-engine_2.21.bb. this has the following SRC_URI : http://install.tarball.in.source.dir/codec_engine_2_21.tar.gz. is this some sort of black magic? it fails on my system.. Oct 06 13:04:38 tasslehoff: license require you to fetch tarball from TI website by hand Oct 06 13:06:10 * tasslehoff rereads the url and all of a sudden install.tarball.in.source.dir looks like an instruction Oct 06 13:07:53 ;D Oct 06 13:21:05 pb__: Were you going to post the TSC candidate details somewhere like the wiki? Oct 06 13:22:55 pb__: can i still run for a position or was there a deadline i missed? Oct 06 13:25:36 mickeyl: I think the deadline was for attending OEDEM Oct 06 13:27:31 ok, good Oct 06 13:30:17 pb__ should I check if the /sys/bus/usb/devices directory exist or do I need to check if it contains files or not? Oct 06 13:31:23 pb__: ah right my TSC mail never arrived. Do you have a proposal for it? Oct 06 13:31:29 RP: hey! Oct 06 13:32:39 hi zecke Oct 06 13:32:45 zecke: How's things? Oct 06 13:33:14 RP: pretty good, I finally come around doing stuff i like without building the base for others :) Oct 06 13:33:35 zecke: great to hear :) Oct 06 13:33:59 getting some money to optimize WebKit with a lot of freedom Oct 06 13:34:13 zecke: freedom is nice to have :) Oct 06 13:37:01 morning Oct 06 13:37:32 * kergoth sighs, meetings at 8am should be banned Oct 06 13:39:44 pb__: I notice USE flags are on the agenda :) Oct 06 13:40:01 kergoth: It could be worse I guess Oct 06 13:40:45 what is a small x image viewer ? Oct 06 13:42:32 RP: yes, that's true, i've had to get up for 7am ones before for meetings with india due to the time zones... but still.. ideally people are actually fully *awake* at meetings.. at least when the meetings start.. perhaps not by the time they end ;) Oct 06 13:43:22 kergoth: Its the 2am ones that are really interesting (to get China. US and UK people in on them) Oct 06 13:43:47 yikes Oct 06 13:50:40 re Oct 06 13:50:47 hi rp Oct 06 13:51:32 hi woglinde Oct 06 13:55:12 I would like to see RP bikes haha Oct 06 13:55:22 ups Oct 06 13:55:22 args Oct 06 13:55:26 only one Oct 06 13:57:07 mickeyl: yes, in fact I was going to ask if you would be interested in that. there is certainly still time. Oct 06 13:57:30 zecke: I'm not quite sure I understand what you are asking. Oct 06 13:58:39 pb__: sorry, reagrding the mail you sent about the TSC. Do you have a role model in mind? is there anything in the wiki describing how it will work? Oct 06 13:59:51 zecke: no, those details remain to be determined. The original mail I sent to the list contains a basic description of what the TSC's function will be, but how exactly it achieves that function is probably a discussion for later. Oct 06 14:00:07 RP: yeah, I will do that later this week or early next. Oct 06 14:04:41 re gnutoo Oct 06 14:05:15 hi Oct 06 14:15:07 whoever decided to use 5 space indentation with 8 space tabs for compression in opkg's code needs to be stabbed in the eye with a hot poker. 2 levels deep = 2 tabs + 2 spaces Oct 06 14:15:10 * kergoth shakes head Oct 06 14:16:04 kergoth: import it to git, run indent on it and then hack? Oct 06 14:16:10 hehe Oct 06 14:16:30 once vim is configured right, don't really have to think about it, but it's still annoying to my sensibilities :) Oct 06 14:19:04 morning Oct 06 14:19:20 morning Oct 06 14:19:43 kergoth: some people abuse tabs ;) Oct 06 14:20:07 hi chouimat Oct 06 14:22:50 woglinde: It is bikes in plural now :/ Oct 06 14:23:03 yes but you can only drive one Oct 06 14:23:07 to oedem Oct 06 14:23:24 woglinde: Right. I also prefer to ride it rather than drive :) Oct 06 14:23:33 args Oct 06 14:23:45 in german we only have one word Oct 06 14:25:15 woglinde: The number of English speakers who say that is high anyway so I'd not worry ;-) Oct 06 14:25:16 hi jay7 Oct 06 14:25:18 RP: in BugLabs we had meetings at 16:00 my time (10:00 NYC, 24:00 Japan) Oct 06 14:25:26 hi ppl Oct 06 14:26:00 hrw|gone: The Japanese drew the short straw ;-) Oct 06 14:26:19 RP: he proposed those meetings ;d Oct 06 14:26:49 hrw he worked 8 hours before in a factory? Oct 06 14:27:56 woglinde: rather not Oct 06 14:38:36 hrw|gone: Its looking likely I'll be at ELC-E btw Oct 06 14:39:56 RP: cool Oct 06 14:40:26 hm flashplayer for arm will be come offical Oct 06 14:40:37 wonder why they dont annouce something for maemo Oct 06 14:41:05 maemo died Oct 06 14:41:10 gremlin[it]: hi Oct 06 14:41:20 hrw why? Oct 06 14:42:36 woglinde: you have two choices: if you want current maemo then you need to buy n900 with its closed components or you can use maemo4 which will not get any updates Oct 06 14:42:56 mckoan, hi Oct 06 14:43:00 hrw: maemo always had closed components Oct 06 14:43:11 florian right Oct 06 14:43:37 florian: I know, 5 just got more of them - pim got extended and closed again Oct 06 14:44:07 heh... I never really cared about maemo PIM ;) Oct 06 14:44:28 maemo didnt have a real pim Oct 06 14:44:39 never had Oct 06 14:44:52 but as n900 is phone they had to add some calendar Oct 06 14:44:55 depends on the definition of PIM Oct 06 14:45:14 woglinde: addressbook + calendar + tasks (+sync) Oct 06 14:45:27 woglinde: with some kind of today Oct 06 14:45:49 hm htc hd2 Oct 06 14:46:35 woglinde: I know that WM6 has it Oct 06 14:46:52 600 euros in de Oct 06 14:46:53 haha Oct 06 14:48:15 so far I am fine with my S60v3 phone Oct 06 14:52:09 lol as I thought Oct 06 14:52:20 winmobile 6.5 only on some apps new soft Oct 06 14:52:32 the old ugly interfaces are there too Oct 06 15:03:35 woglinde: I fixed the libtool thing. I think it was autom4te.cache that needed to be deleted Oct 06 15:03:46 ah yes Oct 06 15:03:48 but now it seems to be generating a broken configure script Oct 06 15:03:58 that should never be shipped Oct 06 15:04:11 jm Oct 06 15:04:12 no, that could have been checked in to our svn server by mistake Oct 06 15:04:13 hm Oct 06 15:04:24 okay Oct 06 15:04:36 better use git Oct 06 15:04:44 so integrate suff upstream is easier Oct 06 15:04:47 stuff Oct 06 15:04:51 for us Oct 06 15:04:56 hge crofton Oct 06 15:05:05 gm Oct 06 15:05:10 crofton could you please activate me on patchwork? Oct 06 15:05:29 but now it overwrites includes/SDL_config.h with a file that omits all the SDL macros, and the configure script it generates is missing the extra function I added in configure.in Oct 06 15:05:31 ok Oct 06 15:05:33 hang on Oct 06 15:05:50 can I get bitbake to show me the exact commands it is running? Oct 06 15:05:55 heh.. each time when I want to write 'cvs' my fingers white 'svn' Oct 06 15:06:01 s/white/write Oct 06 15:06:50 mike_cw you mean bitbake -DDD? Oct 06 15:07:05 let's see... Oct 06 15:07:29 mike_cw hm look at the SDL_config.h.in Oct 06 15:07:37 if it is generated right Oct 06 15:07:48 no it's missing all the SDL stuff too Oct 06 15:07:58 hm Oct 06 15:08:21 maybe a glinch in configure.in/ac Oct 06 15:08:24 if i run autoconf manually from the shell from inside ${S} then it creates the expected configure script Oct 06 15:08:46 hm the recipe dont copy something? Oct 06 15:09:16 mike_cw ah you need to look at temp/ and do_log files Oct 06 15:10:08 the upstream libsdl recipes copy acinclude.m4, but it doesn't make any difference whether I use the one from OE or the one shipped with SDL Oct 06 15:10:45 acinclude.m4 is crappy anyway split it up in spererated .m4 files Oct 06 15:11:08 I think thats the best way to use .m4 macros Oct 06 15:12:00 just reading the debug do_configure. i may be a while :)