**** BEGIN LOGGING AT Thu Aug 11 23:59:56 2005 Aug 12 00:02:30 I am trying to add a package imageMagic and am getting the error : xwdFile.h not found. obv. the package is xwd. but cant make it work ? Aug 12 00:09:46 morning all **** BEGIN LOGGING AT Fri Aug 12 07:18:11 2005 Aug 12 07:23:09 tubbie|home: new episode of SG:A in 10 hours :) Aug 12 07:24:59 yes, yeah koen , looking forward for SG:A and BSG Aug 12 07:25:16 both rule too much Aug 12 07:59:22 afternoon all Aug 12 08:00:16 hi pb_ Aug 12 08:08:32 Is someone here using the composing feature of the keyboard driver Aug 12 08:11:53 on what machine? Aug 12 08:12:20 "Porting opie to pxa arch" - would someone hit zaheer with a clue bat? Aug 12 08:12:55 is pxa even an arch, technically speaking? Aug 12 08:14:28 03mickeyl 07org.oe.dev * rb842c24d... 10/packages/linux/openzaurus-pxa27x_2.4.20-rmk2-embedix20050228.bb: repair openzaurus-pxa27x after someone destroy vital parts of it :/ Aug 12 08:14:32 03mickeyl 07org.oe.dev * rc7c7d385... 10/packages/cumulus/cumulus_cvs.bb: cumulus: ship icons Aug 12 08:15:41 hmm Aug 12 08:15:46 uah funny Aug 12 08:15:50 cardctl eject Aug 12 08:15:54 ejects all cards Aug 12 08:16:06 which is.. uhm... quite unfortunate if one card is a microdrive where / is mounted Aug 12 08:16:07 *sigh* Aug 12 08:18:19 JustinP: FYI: I've corrected the order (once again) to match the cmdline order found in the stock ROM. Aug 12 08:18:53 i guess the angelboot variant in the spitz is _really_ picky about that Aug 12 08:19:19 mickeyl: I think I can swap the slot numbers on spitz btw, certainly for 2.6 Aug 12 08:19:35 RP: that'd be a relief Aug 12 08:19:46 and it should also be possible to mark the device as non-ejectable Aug 12 08:20:07 mickeyl: ok. Adding the console variable in the bug report worked for me Aug 12 08:20:18 I'm not sure how you'd do that... Aug 12 08:20:36 RP: hacking pcmcia_pxa Aug 12 08:20:38 :D Aug 12 08:20:53 mickeyl: I don't do hacks ;-) Aug 12 08:21:06 well, isn't pcmcia-cs deprecated? Aug 12 08:21:11 i know. and you don't know how grateful I am for that :D Aug 12 08:21:16 mickeyl: Did you see the trouble your pxa-serial-hack caused btw? :) Aug 12 08:21:17 yeah it is Aug 12 08:21:22 03jbowler 07org.oe.nslu2-linux * r09d7f197... 10/conf/ (distro/ucslugc.conf machine/nslu2.conf): Aug 12 08:21:22 nslu2.conf: change to the IPKG_ARCHS to include the correct architectures Aug 12 08:21:22 (armv*b for all armv* suppported by the xscale processor). The old Aug 12 08:21:22 'generic' armeb is still suppported (of course) and, by default, this is Aug 12 08:21:22 the ARCH of the generated packages - this is technically wrong, the Aug 12 08:21:23 packages are actually armv5teb - but changing this would make the feeds Aug 12 08:21:25 non-functional for existing systems. Aug 12 08:22:05 RP: yeah, I read it. I can understand rmk for one thing, then again if I were _the_ ARM linux developer then I'd fscking get my act together and fix it in a sane way before people trying to get away with hacks Aug 12 08:22:32 mickeyl: I found it quite amusing. I'd not have dared show it to rmk :) Aug 12 08:23:10 RP: righto. at least we know that such things are best kept undercover :D Aug 12 08:24:09 mickeyl: and edje is still screwing up for me, although I'm getting a new error now....I know it worked with my own branch so I'm still trying to figure out what it may have been Aug 12 08:24:47 JustinP: ok, glad it's not my build machine. Aug 12 08:24:56 i had it working previously too Aug 12 08:25:00 months ago Aug 12 08:25:08 hey dirk Aug 12 08:25:16 hi mickeyl Aug 12 08:25:16 hi dirk Aug 12 08:25:18 well I had it working days ago, so that's a good sign at least... Aug 12 08:25:21 hi RP Aug 12 08:25:33 it's probably something in one of the earlier libraries Aug 12 08:25:34 RP: overall, how do you like the Spitz compared to your Husky? What are biggest problems you see in 2.6 on it? Aug 12 08:25:55 JustinP: you think it might get us anywhere to ask raster himself? Aug 12 08:26:00 mickeyl: I like the speed of Spitz - I think 2.6 will really fly. Aug 12 08:26:21 mickeyl: maybe...he might know exactly what it is or he might just say "works for me" Aug 12 08:26:32 mickeyl: that's basically what he said when I asked him before Aug 12 08:26:36 mickeyl: The biggest 2.6 problem is me finding the time to sort it out. So far, it looks manageable, despite Sharp's fits of insanity Aug 12 08:26:49 mickeyl: gave me some of his cross-compile script but that didn't really help much Aug 12 08:27:04 RP: did you get a Spitz? Aug 12 08:27:26 RP: how 'bout pxafb? on my benchmarks it's pretty slow compared to W100. think we can do anything about it? Aug 12 08:28:35 mickeyl: We might be able to get the lcd clock running faster, I but I've not looked at it much yet Aug 12 08:29:06 mickeyl: which benchmarks are those? Aug 12 08:29:17 pb_: evas_software_x11 Aug 12 08:29:31 ah, right Aug 12 08:29:32 * pb_ fears evas Aug 12 08:29:37 heh Aug 12 08:29:47 mickeyl: and evas_fb_test? Aug 12 08:30:06 evas_fb_test doesn't count because i can't get that damn thing to use 640x480 without hackin' it Aug 12 08:30:17 mickeyl: is it using acceleration for w100fb? Aug 12 08:30:21 no Aug 12 08:30:27 plain fb Aug 12 08:30:29 how odd Aug 12 08:30:30 isn't evas_*_test suppossed to use the same size everywhere? Aug 12 08:30:37 mickeyl: 640x480 or 480x640? Aug 12 08:30:38 without accel, I would have expected pxafb to be faster than w100. Aug 12 08:31:05 RP: neither. evas_fb_test is hardcoded to 240x320 Aug 12 08:31:12 cd-akita: 0.351 Aug 12 08:31:13 k-corgi: 0.313 Aug 12 08:31:16 mickeyl: ah, right Aug 12 08:31:17 i once patched it to gather the resolution from the fb Aug 12 08:31:20 but i lost that patch :/ Aug 12 08:31:24 doh Aug 12 08:31:30 does akita have pxa fb too? Aug 12 08:31:31 nothing large anywa Aug 12 08:31:33 y Aug 12 08:31:36 koen: yes Aug 12 08:31:45 mickeyl: the w100 will always fly in 240x320 as it uses internal memory instead of external Aug 12 08:32:09 mickeyl: those were the scores from CoreDump|afk's akita and my corgi Aug 12 08:32:23 RP: do you have a fbset cmdline handy? Aug 12 08:32:25 koen: fb or x11? Aug 12 08:32:30 fb Aug 12 08:32:39 yeah, it's pretty fast on fb Aug 12 08:32:42 on x11 it's hell here Aug 12 08:32:47 0.05 or so Aug 12 08:32:50 heh Aug 12 08:32:55 640x480 that is Aug 12 08:33:03 koen: fbset command to do what? Aug 12 08:33:06 .308 on pxafb x11 Aug 12 08:33:12 RP: 320x240 Aug 12 08:33:56 RP: btw., i think XorA lied to us Aug 12 08:34:03 irda seems broken still Aug 12 08:34:06 just tested today Aug 12 08:36:41 oh and the sound driver seems to emit a couple of bogus samples at the start Aug 12 08:37:21 cracks and clicks before the actual wav gets through Aug 12 08:37:33 hi Aug 12 08:37:48 mickeyl: That's interesting to know. I think I know what some of the code I removed was for now :) Aug 12 08:37:52 guess that's why the default driver took so long to open - perhaps because they were dampening the mixer for a couple of microseconds Aug 12 08:38:12 hi CoreDump|home Aug 12 08:38:15 Yes. Hmmm... Aug 12 08:38:49 the irda transceiver is at ttyS1 or 2 ? Aug 12 08:39:13 damn, my build machine appears to have a hardware problem Aug 12 08:39:17 1 Aug 12 08:39:19 mickeyl: Offhand I can't remember Aug 12 08:39:27 it's S1 Aug 12 08:39:45 Its the STUART mentioned in the bootlog Aug 12 08:39:58 I'm poping out. Back shortly... Aug 12 08:40:06 k,cu Aug 12 08:40:13 irattach: tty: set_disc(119: Bad file descriptor Aug 12 08:40:23 irattach: ioctl(SIOCGIFFLAGS): No such device Aug 12 08:40:47 *sigh* guess we need to revert all the irda patches from 2.6.12-whatever to find out where the problem starts Aug 12 08:41:30 CoreDump|home: that sucks Aug 12 08:41:51 indeed Aug 12 08:42:19 was frozen when I came home, now crashed during kernel boot Aug 12 08:43:01 looks like I need to hook up a monitor grrrrr Aug 12 08:43:15 mickeyl: is 2.6 working on Spitz now...? Aug 12 08:43:23 dunno Aug 12 08:43:30 looks like RP is making good progress though Aug 12 08:43:39 when did he start working on it? Aug 12 08:43:44 two days ago Aug 12 08:43:50 ah :-) nice Aug 12 08:43:51 but BigAl already did a lot of work Aug 12 08:43:54 I know Aug 12 08:44:30 so what exactly is binconfig_suffix(d) doing? Aug 12 08:44:45 it's returning "" for a cross build Aug 12 08:44:49 -native for a native build Aug 12 08:45:04 so that we know which file to write to Aug 12 08:46:11 why do we need that, though? They shouldn't be staged in the same place... Aug 12 08:46:18 they should Aug 12 08:46:25 there's just one directory for executables Aug 12 08:46:28 ${STAGING_BINDIR} Aug 12 08:47:03 try calling evas-config --libs Aug 12 08:47:04 it looks to me like the arm binaries go in bin/arm-linux.... Aug 12 08:47:06 and then evas-config-native --libs Aug 12 08:47:12 of course, arm binaries go to there Aug 12 08:47:13 hmmm Aug 12 08:47:14 but the point is Aug 12 08:47:19 those aren't arm binaries Aug 12 08:47:35 they are x86 binaries which are called by subsequent packages to see what the flags are Aug 12 08:47:47 I know Aug 12 08:48:00 oen@dominion:/data/build/oe/monotone/org.openembedded.dev/packages/inkscape$ ls /data/build/oe/build/tmp/familiar/staging/ Aug 12 08:48:02 I'm just trying to figure out what's killing these builds.... Aug 12 08:48:03 arm-linux/ etc/ i686-linux/ local/ man/ Aug 12 08:48:10 nicely seperated Aug 12 08:48:21 * koen hints at multimachine.inc Aug 12 08:48:24 ? Aug 12 08:48:27 koen: except arm-linux/bin doesn't have anything in it Aug 12 08:48:29 multimachine doesn't buy us anything Aug 12 08:48:41 the binaries are for the HOST architecture Aug 12 08:48:46 both binaries Aug 12 08:48:52 right, there is no point in staging target binaries Aug 12 08:49:00 which is why they are not allowed to have the same name Aug 12 08:49:08 you can't run them (obviously) and there isn't much else they are good for Aug 12 08:49:14 yep Aug 12 08:49:41 well, I'm attempting it without the -native extension for native stuff....just to satisfy my curiosity Aug 12 08:49:48 have fun Aug 12 08:49:52 this will be pandora's box Aug 12 08:50:17 because then the native builds will have nice things like -L$arm-linux in there Aug 12 08:50:45 you see - there are just two approaches to deal with the foo-bar-config[-native]. either append a suffix (that's what i'm doing right now), or write them in different directories and set PATH accordingly for packages using binconfig stuff. Aug 12 08:51:02 ok, ok, I understand Aug 12 08:51:19 hit raster with the pkgconfig cluebat Aug 12 08:51:33 the pkgconfig spec was updated with crosscompilation in mind Aug 12 08:51:37 I'm just running the build to see what happens. I'm trying to figure out why my bb files work and the new ones don't. This was one fo the differences. Aug 12 08:51:55 i'm sorry to insist on that but it took me three days to get everything clean and it was hell of a work. Now that we have the binconfig + binconfig-native seperation all works nicely and the order in which you build foo and foo-native is totally irrelevant. Aug 12 08:52:22 actually i think the binconfig stuff is not bad Aug 12 08:52:35 it just goes mad when you need -native packages relying on that method as well Aug 12 08:52:38 *shrug* Aug 12 08:52:42 yeah Aug 12 08:53:18 it'd be easy enough to work around that by creating a special staging/HOST-SYS/native-bin/ directory, putting the native binconfigs in there, and arranging for native.bbclass to put it in $PATH ahead of the regular bindir. Aug 12 08:53:18 I'm not checking it in, I'm just testing. I appreciate the bbclass you wrote Aug 12 08:54:04 pb_: well, in this case there are env vars you can set with the paths so it would really be no problem whatsoever Aug 12 08:54:15 pb_: yeah that'd work as well. at the moment i favour my current approach, but eventually we may use that one Aug 12 08:54:16 * JustinP isn't saying we should change it Aug 12 08:54:17 ~lart fbset for not working Aug 12 08:54:18 * ibot rm -rf's fbset for not working Aug 12 08:54:42 or, conversely (and maybe better), make staging/HOST-SYS/cross-TARGET-SYS/bin, put the target ones in there, and put it in the path unconditionally. Aug 12 08:55:04 yeah right. that looks good Aug 12 08:55:48 mickeyl: yep, your approach is fine as well Aug 12 08:56:24 the only drawback to my approach is that we need to explicitly tell where configure finds the said binconfig binaries. Aug 12 08:56:28 mickeyl: BTW, e-wm is failing for me due to edje issues. edje wasn't in its depends but it failed due to a missing include for me Aug 12 08:56:39 that wouldn't be necessary with your method Aug 12 08:56:53 JustinP: right. i thought it only needs the compiler, but it turns out to need the full lib. thanks Aug 12 08:57:45 * mickeyl wanders off for some card playing Aug 12 09:15:59 ~lart those crappy SATA connectors Aug 12 09:15:59 * ibot beats those crappy SATA connectors severely about the head and shoulders with a rubber chicken Aug 12 09:16:25 thank god it was only the connector *sigh* Loosing that drive would have sucked Aug 12 09:28:34 JustinP: Spitz with 2.6 now has a keyboard and touchscreen. I see no reason opei/gpe shouldn't work once they get taught about orientation of the screen etc. Aug 12 09:29:24 RP: very nice. :-) Aug 12 09:29:42 RP: does this mean I could just build it now and it would work or that it would need wteaking? Aug 12 09:29:54 RP: and I suppose some things are still *not* working? Aug 12 09:30:11 (BTW, thanks much for hopping on this, the 2.4 kernel scares me) Aug 12 09:30:27 opie would just need to be told how the buttons/keyboard are layed out Aug 12 09:30:54 probably a couple hours work for someone who doesnt know what he's doing Aug 12 09:30:59 shouldn't opipe use the default (working) keymap? Aug 12 09:31:09 or is that too much bloat? Aug 12 09:31:25 FWIW an opie-image fails on spitz with the 2.6 kernel becuase orinoco-modules bombs out complaining "Wireless extensions are not enabled. Stop.". Aug 12 09:31:38 But I at least have a kernel built now to test later. Aug 12 09:31:46 treke|home: It just needs to be told that this model is the same as the c7x0 Aug 12 09:32:07 then it wouldnt take long to do. Aug 12 09:32:22 if everything is the same as on a c7x0 Aug 12 09:32:45 JustinP: The biggest none functional thing is power management. Also missing it backlight, sound and IR Aug 12 09:33:00 treke|home: Everything is the same as 2.6 c7x0 Aug 12 09:33:16 Doesnt it use a different graphics chip? Aug 12 09:33:22 ah...well a backlight would definately be useful ;-) sound I also like and IR...well, I never use IR anyway ;-) Aug 12 09:33:22 Noodles: We just need to enable wireless modules in the kernel defconfig for that to work Aug 12 09:33:41 treke: pxafb Aug 12 09:33:44 Yeah, I realised that. Aug 12 09:33:56 JustinP: I've only had it 2.5 days. Give me another week... ;-) Aug 12 09:34:07 hehe Aug 12 09:34:33 I know, you're tearing through this stuff. THanks to BigAl and everyone's (your?) work on 2.6 for c7x0 Aug 12 09:34:51 treke: It uses pxafb but I don't see why opie should care about that... Aug 12 09:35:10 RP: because it needs to load the proper driver for qte? :p Aug 12 09:35:20 or perhaps that's compile time Aug 12 09:35:31 BigAl has given me an excellent start for spitz for which I'm grateful Aug 12 09:35:58 treke: It accesses /dev/fb0 unless you're talking about the w100 acceleration stuff sirfred added... Aug 12 09:36:22 treke: There is an orientation issue but I'm confident mickeyl will fix that in short order :) Aug 12 09:53:43 * JustinP wonders.... Aug 12 09:58:52 just trying with oe Aug 12 09:59:11 could not download bitbake; guess my system is behind a firewall. Aug 12 09:59:28 does anyone know/have a tarball of the latest bitbake Aug 12 10:00:15 I can tarball it up for you Aug 12 10:00:24 please Aug 12 10:00:59 the latest one; i have the old one; my mate told me. its too old; Aug 12 10:01:20 trouble downloading bitbake; my firewall settings prevent it. Aug 12 10:01:32 klin: handhelds.org/~zecke/downloads/bitbake Aug 12 10:01:47 http://oe.reversefold.com/bitbake.r308.tbz2 Aug 12 10:01:49 oh... Aug 12 10:02:08 well, mine is the latest SVN revision... Aug 12 10:02:53 zecke: is this the latest one; which will work with the downloaded db of oe Aug 12 10:02:55 * JustinP encounters the joys of compiling glibc once more Aug 12 10:03:05 just downloading oe db. Aug 12 10:03:37 klin: I think bitbake has been updated since then. A few things may not work with that version AFAIR. The latest from SVN (which I packaged) should do everything but *may* be unstable (although I doubt it). Aug 12 10:05:05 how do i download it justin Aug 12 10:05:25 my browser shows some unprintable characters Aug 12 10:07:21 justin: waiting for ur response Aug 12 10:09:11 klin: wget -c http://oe.reversefold.com/bitbake.r308.tbz2 Aug 12 10:10:24 klin: yep, use wget Aug 12 10:10:47 cant i use a web browser to access it Aug 12 10:11:16 klin: yeah...but it sounds like your browser is set up to view tbz2 files instead of downloading them like it should Aug 12 10:11:29 klin: what browser are you using? Aug 12 10:11:41 Explorer Aug 12 10:12:40 Try FireFox if you have it Aug 12 10:12:51 klin: you're going to be using all this on Linux, right? Aug 12 10:12:56 klin: just use wget on the linux box Aug 12 10:13:11 ya Aug 12 10:15:21 Justin: will u keep that file. may be it will take some time for me to access it Aug 12 10:15:31 may be tommorrow Aug 12 10:16:53 klin: I suppose Aug 12 10:17:08 thanks Aug 12 10:17:31 u r such a nice fellow. please dont delete it. i will access it as soon as i can. Aug 12 10:17:39 may be tommorrow. Aug 12 10:17:39 klin: just let me know when you've downloaded it Aug 12 10:17:53 how do i contact u Aug 12 10:18:52 ok. mate Aug 12 10:19:15 Rar. Aug 12 10:19:19 will let u know if i see u in this room tommorrow Aug 12 10:19:19 * Noodles boots 2.6 on his Spitz. Aug 12 10:19:43 thanks all. Aug 12 10:19:43 klin: I should be here, just privmsg me or prepend my (full) name to your message Aug 12 10:19:59 ok justin Aug 12 10:20:08 will do so. thanks once again Aug 12 10:20:18 justin: 1 sec Aug 12 10:20:39 can i use ur bitbake with what i am downloading oe snapshot Aug 12 10:20:51 this is the latest bitbake from SVN Aug 12 10:20:54 it's what I'm using Aug 12 10:21:04 fine mate. Aug 12 10:21:29 will get back to u in case i run into any problems soon Aug 12 10:21:33 :-) Aug 12 10:26:46 reenoo_afk: I understand that you want to make sure we follow up on QA but I don't like you telling me that my solutions don't work when I have not only tested them, I have posted that they definately work. Aug 12 10:27:45 reenoo_afk: I'd be happy to move the sed call to do_compile if this would make more sense somehow...but I'm just trying to band-aid this bb file for the unfortunates for whom it does not compile until someone can fix the problem in the right way. Hence why I didn't close this bug. Aug 12 11:09:02 * JustinP crosses his fingers Aug 12 11:12:16 03koen 07org.oe.dev * r167e15fd... 10/packages/base-files/base-files/ipaq-pxa270/fstab: add fstab for ipaq-pxa270 Aug 12 11:14:48 * koen scratches his head Aug 12 11:14:56 that diff doesn't look right Aug 12 11:17:30 ah it does, mtd0 is first Aug 12 11:17:34 hey gerwinin Aug 12 11:18:41 hey koen Aug 12 11:22:15 How much memory does the kernel require? Aug 12 11:22:18 minimum? Aug 12 11:22:37 "the" kernel? Aug 12 11:22:59 The Linux kernel that oe uses. I know its not the mainline kernel.org kernel. Aug 12 11:23:06 koen@dominion:/data/build/oe/monotone/org.openembedded.dev$ ls packages/linux | grep bb | wc -l Aug 12 11:23:09 78 Aug 12 11:23:15 we have 78 different kernels Aug 12 11:23:32 so "the t" Aug 12 11:23:41 so "the kernel oe uses" isn't correct Aug 12 11:23:45 Which one is the smallest? Aug 12 11:23:53 It uses one source code base doens't it? Aug 12 11:23:56 arrummzen: you're still not getting it Aug 12 11:24:08 arrummzen: no, the devices are very disparate, there are seperate sources Aug 12 11:24:16 arrummzen: at least for some of them Aug 12 11:24:29 arrummzen: you can change the config in include/exclude whatever you want Aug 12 11:24:35 in=to Aug 12 11:25:12 heh, cute. gquilt. http://freshmeat.net/screenshots/54225/58238/ Aug 12 11:26:12 hey kergoth Aug 12 11:26:15 * JustinP hasn't even looked at quilt Aug 12 11:26:24 What is the procedure for adding support for a new device? Aug 12 11:27:01 I bet the kde people would call it kwuilt Aug 12 11:27:31 surprisingly gquilt wasn't named guilt ;) Aug 12 11:30:14 arrummzen: get a kernel working, set up the machine configuration, etc... Aug 12 11:32:24 is 512kb enough memory Aug 12 11:32:30 512kb of ram that is, Aug 12 11:33:56 arrummzen: I'm sure you can fit a linux kernel in that if you have to...I just don't know what all it can have i it Aug 12 11:34:14 RP: I see what you mean about OPIE and Spitz 2.6. I can rotate it so the screen looks right, but the touchscreen is mirrored along the diagonal from bottom left to top right. Aug 12 11:34:28 Plus the opie console still does stuff when I press Ctrl. :( Aug 12 11:36:06 Hmmm. I think it just tried to dim the backlight and crashed. Aug 12 11:36:12 But still. Yay. Progress. You rock. Aug 12 11:37:56 03jbowler 07org.oe.nslu2-linux * r3d3d7753... 10/packages/monotone/ (monotone-5_0.22.bb monotone.inc): Aug 12 11:37:56 monotone-5 0.22 - the LUA hook for anonymous_read must be updated (moved Aug 12 11:37:56 into the standard read hook). Aug 12 11:37:56 test 247 (lua function existsonpath) fails - apparently there is a problem Aug 12 11:37:56 with search the path for executables. This is the only remaining Aug 12 11:37:56 unexpected failure (all unexpected failures in previous versions except the Aug 12 11:37:58 netsync 'largish-file' test now function correctly on an NSLU2). Aug 12 11:39:41 03jbowler 07org.oe.nslu2-linux * r96b19529... 10/packages/meta/ucslugc-packages.bb: UcSlugC: add dev packages Aug 12 11:56:09 http://www.linuxdevices.com/news/NS7186776945.html .. heh Aug 12 12:00:12 kergoth: http://www.linuxdevices.com/news/NS2239811258.html Aug 12 12:00:21 dunno if it's 0.7.2 or OE based Aug 12 12:03:10 cute Aug 12 12:03:34 sort of like a rugged meshcube Aug 12 12:41:14 re Aug 12 12:41:26 hey zecke Aug 12 12:47:01 * koen taps foot Aug 12 12:47:07 stinking hotplug+udev Aug 12 12:47:14 * zecke hates being poor Aug 12 12:47:35 * koen is currently broke but is going to work next week Aug 12 12:48:05 But I'm sure you're rich in spirit Aug 12 12:48:15 * kergoth looks forward to getting a damn paycheck Aug 12 12:48:47 it is fucking depressing seing others going to faires and meetings... (Malaga/aKademy) Aug 12 12:49:31 :( Aug 12 13:15:26 03koen 07org.oe.dev * r069f7265... 10/packages/base-files/base-files/ipaq-pxa270/fstab: Don't use devfs names in ipaq-pxa270 fstab Aug 12 13:40:31 ahh crap Aug 12 13:40:41 i have forgotten the password on my router Aug 12 13:47:42 heh Aug 12 13:47:44 I did that too Aug 12 13:48:01 Its still up, so hear no evil, see no evil.. :) Aug 12 13:50:15 /me curses OO and garbage collection Aug 12 13:50:39 ~lart Boehm for using a stupid buildsystem Aug 12 13:50:39 * ibot shoots Boehm in the head for using a stupid buildsystem Aug 12 13:53:56 JustinP: you're missing the point. it's very unlikely someone will look into this one .bb file out of how many? 2000+? unless the build breaks. with so few maintainers whatever is checked in has to be assumed to be correct. Aug 12 13:54:16 JustinP: this doesn't mean sed hacks are bad by definition Aug 12 13:54:34 JustinP: put them in the right place and add a good explanation as a comment Aug 12 13:55:05 JustinP: that's what I've asked for in my last comment in bugzilla already Aug 12 13:59:51 hi reenoo_ Aug 12 14:00:08 reenoo_: well, to me putting it in do_staging is *correct* as that's where the error occurs. I also added a comment explaining why it's there. So what's the problem here? Aug 12 14:00:34 reenoo_: and just for the record you didn't ask for those things. You told me my solution can't be correct and asked me to revert it. Aug 12 14:02:29 hey pb_ Aug 12 14:06:48 JustinP: nevermind Aug 12 14:08:19 reenoo_: I'm perfectly prepared to make any changes for the greater good...I just don't understand what about this fix isn't ok Aug 12 14:12:26 pb_: /etc/modules is autogenerated, right? Aug 12 14:12:36 yes Aug 12 14:12:52 Noodles: The kernel backlight is totally unimplemented under 2.6 so that doesn't surprise me. It shouldn't crash though... Aug 12 14:13:40 ~hail zecke Aug 12 14:13:41 * ibot bows down to zecke and chants, "I'M NOT WORTHY!!" Aug 12 14:15:27 koen: hehe I'm still at the ROAD office Aug 12 14:15:30 weekend starts later Aug 12 14:16:19 * koen battles with gc glibmm gtkmm and other OO nastiness Aug 12 14:16:27 ~lart inkscape Aug 12 14:16:27 * ibot calls inkscape on the phone ... the lights are on but nobody's home Aug 12 14:16:36 gc == garbage collection? Aug 12 14:16:40 yes Aug 12 14:16:49 the one from mr boehm Aug 12 14:16:52 koen: that happens automatically Aug 12 14:17:03 koen: that is the key point of gc's Aug 12 14:17:27 koen: preparing for your work or is it independant/non-java stuff Aug 12 14:17:27 it needs libgc Aug 12 14:17:43 trying to build inkscape Aug 12 14:18:09 'cause inkscape would run real fast on a pda Aug 12 14:18:11 *cough* Aug 12 14:18:13 hail zecke Aug 12 14:19:37 pb_: I've a offer, send a patch get a monotone account ;) Aug 12 14:19:52 pb_: though this offer is limited ;) Aug 12 14:20:00 that is a tempting offer Aug 12 14:20:14 but, sadly, I have no patches Aug 12 14:20:42 if pb_ is being restricted from getting a montone account, then something is rotten with the administrators Aug 12 14:20:52 indeed Aug 12 14:20:55 heh Aug 12 14:21:10 who wants pb_ as committer anyway ;) Aug 12 14:22:10 pb_: do you still give shelter to homeless sick and dirty cats? Aug 12 14:22:24 zecke: always Aug 12 14:22:43 Is your project "Arch Phil" taking donations? Aug 12 14:23:09 no, I think that project was cancelled Aug 12 14:24:04 koen: can gpe-mini-browser host the google toolbar? Aug 12 14:25:46 zecke: no idea Aug 12 14:28:29 listening to "Wir sind Helden" Demo Tapes make me even more insane Aug 12 14:29:31 * mithro pulls out hair Aug 12 14:32:30 mithro: how is sweden? Aug 12 14:32:59 zecke: good Aug 12 14:33:09 i've forgotten my routers password however Aug 12 14:34:25 mithro: I could check my keylogger if it contains it Aug 12 14:56:10 03koen 07org.oe.dev * rd34b6f07... 10/packages/inkscape/inkscape_0.42.bb: Fix DEPENDS for inkscape 0.42 Aug 12 15:16:11 * zecke heads home Aug 12 15:18:11 RP: I could be wrong about what happened then; the screen turned off, my SSH session died and pressing the power button didn't seem to bring it back. I'll try again and see what happens. **** BEGIN LOGGING AT Fri Aug 12 15:27:41 2005 Aug 12 15:41:06 OE is only 11 releases behind with libxml2.... Aug 12 15:41:18 is that all? :P Aug 12 15:41:21 thats terrible Aug 12 15:41:32 * koen tries 2.6.20 Aug 12 15:41:57 i noticed it might be old when inkscape needed error codes in the 3k region Aug 12 15:42:08 2.6.9 in OE only has 1900 error codes Aug 12 15:42:40 hehe Aug 12 15:47:06 2.6.20 doesn't work, on to 2.6.19 Aug 12 15:51:45 Noodles: As long as the kernel is ok, I'm happy for now... Aug 12 15:52:14 Don't get the impression I'm not happy with it; just rambling about problems as I hit them. :) Aug 12 15:52:35 Noodles: I understand - its helpful to know where problems lie Aug 12 15:53:21 I added networking support to my kernel and that helps as it means I can ssh in and prod. I see exactly what you mean about the console being rotated being irritating. Aug 12 15:53:46 Yes, I might have to hack at that. I'll probably never make mainline though :-/ Aug 12 15:54:02 How does the 2.4 kernel do it? Aug 12 15:54:06 Noodles: I'm not going to be around much tomorrow but will be taking Spitz and a laptop with me ;-) Aug 12 15:54:10 Noodles: hacks Aug 12 15:54:14 :( Aug 12 15:54:47 Anyhow, I have an early start so I must sleep Aug 12 15:54:55 'night RP Aug 12 15:54:56 Night. I should head that way myself. Aug 12 15:55:02 Let me know how the networking goes :) Aug 12 15:55:06 'night all Aug 12 15:59:35 *finally* Aug 12 15:59:43 I got e-wm to compile Aug 12 16:00:22 nice Aug 12 16:01:31 hi Aug 12 16:02:08 now to clean it all up and check it in Aug 12 16:02:35 oh, and to try entrance to make sure this isn't a fluke ;-) Aug 12 16:03:37 hey Bernardo Aug 12 16:04:06 hi Bernardo Aug 12 16:07:22 does openembedded ever create binaries with an extensioN? Aug 12 16:07:49 OE doesn't care Aug 12 16:08:10 ok, well, what's the best way to know that a file is a real executable and not, say, a library set executable? Aug 12 16:08:36 I'm doing: "if echo $i | grep -v \\." Aug 12 16:08:44 checking for a "." and hence, an extension Aug 12 16:08:57 works for me, but if these binaries will ever have an extension this won't work Aug 12 16:12:54 * JustinP sticks with what he has for now... Aug 12 16:51:17 ipkg is screwed up....it won't install anything Aug 12 16:51:45 it's also putting lists in /var/lib/ipkg.... Aug 12 18:01:03 03justinp 07org.oe.dev * r98c60b06... 10/classes/efl.bbclass: Aug 12 18:01:03 Make clean and distclean for EFL Aug 12 18:01:03 Automatically install named binaries when compiling native Aug 12 18:01:03 Add more files/directories to theme and dev packages Aug 12 18:01:15 03justinp 07org.oe.dev * r2b96ead3... 10/packages/ (22 files in 3 dirs): Aug 12 18:01:15 Fix dependencies for all EFL and E17 packages Aug 12 18:01:15 Fix problems with edje/embryo (stage .inc files) Aug 12 18:01:15 Add a patch for e-wm which will automatically scroll menus onto the screen when the cursor is within 5px of the edge of the screen Aug 12 18:01:15 Enable edb frontends for GTK and ncurses Aug 12 18:01:43 yay! Aug 12 18:01:54 mickey|bbl: I got E compiling all nicely. :-) Aug 12 18:02:06 mickey|bbl: I'm testing it right now on a fresh GPE install Aug 12 18:07:26 nice work JustinP! Aug 12 18:07:35 some of your dependency chances are wrong though Aug 12 18:08:15 are they? I checked all of the configure.in files... Aug 12 18:08:17 i.e. I made ESmart depend on evas-x11 and ecore-x11 for a reason. they really need the -x11 variants. virtual/evas and virtual/ecore is not enough. Aug 12 18:08:27 hmmm... Aug 12 18:08:37 ok, those *could* be wrong Aug 12 18:08:50 you sure? All of the E stuff should run on any evas/ecore backend... Aug 12 18:08:55 it doesn't Aug 12 18:08:59 hmmm Aug 12 18:09:01 see rasters diagram Aug 12 18:09:07 I just assume dyou'd left them how I had them ;-) Aug 12 18:09:44 hmmm, I don't think that diagram is quite right Aug 12 18:09:59 enlightenment is supposed to be able to run on FB, X11, GL, QT.... Aug 12 18:10:21 ya, but neither the window manager nor entrance nor most of the other apps do that Aug 12 18:10:25 just grep through the source Aug 12 18:10:31 #include Aug 12 18:10:33 boom Aug 12 18:10:41 I see Aug 12 18:10:56 ok, I'll update them Aug 12 18:10:59 thanks Aug 12 18:11:10 good work on finding the edje thingy! Aug 12 18:11:19 how did you find it Aug 12 18:11:20 ß Aug 12 18:11:21 ? Aug 12 18:12:04 Well... Aug 12 18:12:05 yeah Aug 12 18:12:13 I spent 2 days searching for those 2 damn files Aug 12 18:12:34 I was seeing "can't read file edje" today and spent a lot of time playing myself Aug 12 18:12:44 I asked in the #edevelop channel and noone had an answer Aug 12 18:12:47 heh Aug 12 18:12:52 finally someone started helping Aug 12 18:12:52 so much for the experts ;) Aug 12 18:13:02 I did an strace and it was outputting #include Aug 12 18:13:07 excellent Aug 12 18:13:09 in the embryo file Aug 12 18:13:27 so I searched for edje files and eventually found edje.inc and staged it Aug 12 18:13:39 then I got the rational number error and started looking/asking again Aug 12 18:14:06 eventually I found that embry also had an include dir with default.inc Aug 12 18:14:11 embryo Aug 12 18:14:22 once those were installed it all went smoothly Aug 12 18:15:24 cool Aug 12 18:16:09 I find E pretty useful on Spitz. We just need to ship a more lightweight theme and larger font settings and we're set :) Aug 12 18:17:20 yep Aug 12 18:17:35 and I need to get my auto-rotate hack in there somehow (for GPE too) Aug 12 18:20:57 ok, really time to sleep now Aug 12 18:20:59 cu Aug 12 18:21:45 sleep well Aug 12 18:48:28 03justinp 07org.oe.dev * rabbed5a9... 10/packages/ (4 files in 2 dirs): Aug 12 18:48:29 Set entrance, esmart, and e-wm back to -x11 dependencies Aug 12 18:48:29 Add e-modules bb file Aug 12 19:01:37 03justinp 07org.oe.dev * rc5a7217d... 10/packages/efl/e_20050810.bb: stage src/bin/e_*.h as well Aug 12 20:56:00 hey is there a bittorrent client yet for gpe ? Aug 12 20:57:15 Im thinking that a 5500 with a wireless card, a 1gb SD card and an external battery would make a great stealth bittorrent leach Aug 12 21:01:15 there's something called ctorrent but I dont know if it's a bittorrent client Aug 12 21:01:18 just added to OE Aug 12 21:04:12 it is Aug 12 21:09:20 Iv used ctorrent on freebsd.. its alirght.. but it would nice to have a gui app Aug 12 21:30:55 03justinp 07org.oe.dev * r1acd212a... 10/packages/e17/e-modules_20050810.bb: Specify edje_cc for e-modules Aug 12 21:37:00 * france is away: Away Aug 12 22:28:24 ~lart pkgconfig and binconfig Aug 12 22:28:25 * ibot farts in pkgconfig and binconfig's general direction Aug 12 23:45:56 03justinp 07org.oe.dev * r0a79c0b7... 10/packages/esound/esound-gpe_20020817.bb: Aug 12 23:45:56 esound-gpe needs to inherit from binconfig as well to install the esd-config file (this fixed libxine trying to link against /usr/lib/libasound.la) Aug 12 23:45:56 Note: this won't actually take effect until pkgconfig and binconfig are altered not to kill each other (see mailing list) Aug 12 23:45:59 03justinp 07org.oe.dev * r188118f0... 10/packages/altboot/files/altboot.func: Remove extra spaces around "=" causing shell script syntax errors Aug 13 00:01:37 03justinp 07org.oe.dev * r8b8415eb... 10/packages/evoak/evoak_0.0.1.bb: Remove evoak as it no longer builds and is not depended on by anything Aug 13 00:01:39 03justinp 07org.oe.dev * r3f5d1b28... 10/packages/altboot/files/altboot.func: Also remove extra $ causing syntax errors Aug 13 00:17:58 Just downloaded : OE.db.bz2 from oe site. i unzipped it using bunzip2 and got OE.db. it is a file. i know the nmeta data is in this file. but how do i extract it. Aug 13 00:19:03 see the OE wiki Aug 13 00:19:09 MonotonePhraseBook Aug 13 00:27:39 so, i should install monotone to use that db file ? Aug 13 00:27:59 i cant unzip and keep the packages ? Aug 13 00:28:29 eJumbo: you need to understand what this is...OE.db is a revision history. It doesn't have packages in it, it has the metedata for creating the packages. Aug 13 00:28:55 i mean : metadata Aug 13 00:29:11 yes, you need monotone Aug 13 00:30:00 JustinP: just clarify my understanding. Aug 13 00:30:35 is monotone needed ONLY to get oe meta data and for not any other purpose ? Aug 13 00:33:25 monotone is needed to get metadata out of and into the OE.db file Aug 13 00:33:43 OE.db is a monotone DB Aug 13 00:35:50 ok. Aug 13 00:39:38 monotone is the VCS that OE uses Aug 13 00:40:18 i thought oe.db is just compressed meta data Aug 13 00:40:21 :-) Aug 13 00:41:01 OE.db.bz2 is a compressed monotone database Aug 13 00:41:15 all the metadata and all its history is in that one file Aug 13 00:41:21 it's sqlite Aug 13 00:48:26 so i need to install sqllite too Aug 13 00:50:11 no Aug 13 00:50:18 monotone has its own version built-in Aug 13 00:50:27 thank god Aug 13 00:51:11 downloading and installing new software on ur system is a headache. that too when u have a slow connection Aug 13 00:51:43 eJumbo: it's not a headache if you use a decent distribution Aug 13 00:52:28 eJumbo: what *are* you using? Aug 13 00:52:45 fedora core 2 Aug 13 00:52:54 * JustinP prays for your soul Aug 13 00:53:01 I can't help you there Aug 13 00:53:09 i am thinking of switching to debian. Aug 13 00:53:20 better Aug 13 00:53:24 gentoo is the one I like Aug 13 00:53:41 and it has everything right there, no extra searching, no finding extra packages.... Aug 13 00:53:48 Justin: which distro do u think is "best". Aug 13 00:53:52 montone-0.22, straight fromt he source Aug 13 00:53:56 Gentoo Aug 13 00:54:07 I just said that Aug 13 00:54:13 why so ? Aug 13 00:54:56 i am not a expert in this. just trying to get some "expert" advice. Aug 13 00:55:00 :-) Aug 13 00:55:33 because it is very configurable, it has programs up to date all the time, it has a sophisticated "USE" system to allow you to have exactly what you want (no searching for the emacs package *without* X support, for example) Aug 13 00:56:15 it's compiled from source *on your machine* so you don't have to rely on vanilla binaries (this doesn't gain you too much but it's a nice feeling) Aug 13 00:56:30 it's consistent Aug 13 00:56:45 and, as I said, mostly any package you can want is on it Aug 13 00:56:51 and adding a new one isn't that hard Aug 13 00:57:01 about the same as adding a package to OE Aug 13 00:58:04 thanks for the info Aug 13 01:03:38 Justin: should i download monotone from the source and compile it ? Aug 13 01:08:31 eJumbo: if that's what you have to do.... Aug 13 01:08:40 eJumbo: a package would probably be easier Aug 13 01:10:16 package is not there for my system : fedora core 2. Aug 13 01:11:01 eJumbo: you can always try looking in rpmfind or whatever.... Aug 13 01:12:20 good morning Aug 13 01:16:56 Bernardo: morning Aug 13 01:20:27 hi Bernardo Aug 13 01:21:35 Morning all Aug 13 01:27:24 JustinP: FYI: we have decided to deprecate the nslu2-linux branch. I just need to organise new keys for the non-core nslu2-linux developers, and then we will update the master makefile and stop serving it. Aug 13 01:27:54 :-) I'm glad I could bring about a change for the better Aug 13 01:28:03 :-| I hope this all works out well Aug 13 02:27:17 i compiled boost and copied libs and includes to standard directories trying to install monotone rpm. it says it needs boost Aug 13 02:29:48 well...you should have installed a boost rpm them Aug 13 02:30:05 if you compile something yourself FC doesn't know you did that... Aug 13 02:31:27 got it : u mean the rpm data bases is not updated ? Aug 13 02:31:34 yes Aug 13 02:32:15 actually could not find fedora core 2 rpms. can i use fedora core 3 rpms Aug 13 02:41:02 don't know Aug 13 03:33:14 i compiled monotone successfully. i have OE.db (snap shot downloaded using wget). when i do mt db init --db=/home/oe/oedb/oe.db Aug 13 03:33:15 m Aug 13 03:33:32 i get an error : monotone: misuse: could not initialize database: /home/oe/oedb/oe.db: already exists Aug 13 03:33:47 i want to tell monotone: look here is my data base Aug 13 03:33:54 eJumbo: you do *not* want to init your db...you already have one! Aug 13 03:34:24 so how should i use that ? Aug 13 03:35:02 RTFM Aug 13 03:35:10 the instructions are on the wiki! Aug 13 03:35:16 1 sec Aug 13 03:35:52 no, 0 sec Aug 13 03:35:55 I'm going to bed Aug 13 03:37:37 03justinp 07org.oe.dev * rc130ebc0... 10/packages/e17/ (4 files in 2 dirs): Add patches and configuration to make entrance(d) actually work Aug 13 03:37:41 03justinp 07org.oe.dev * rc5e52796... 10/packages/efl/edb_1.0.5.004.bb: Aug 13 03:37:41 Add binaries to dev package Aug 13 03:37:41 Fix runtime deps for ncurses binary Aug 13 03:37:41 Remove gtk+ deps as it doesn't work Aug 13 03:37:45 03justinp 07org.oe.dev * rad3ce7d5... 10/packages/e17/ (engage_20050810.bb examine_20050810.bb): Add engage and examine bb files Aug 13 03:39:48 Justin: could not understand ? Aug 13 03:41:37 could not understand what the wiki says : how do i use the already downloaded database ? Aug 13 03:44:08 skip the init step Aug 13 03:44:15 just do a checkout! Aug 13 03:47:51 i am doing monotone --db=/home/oe/oedb/oe.db checkout --branch=org.openembedded.dev Aug 13 03:48:09 but where is the IP Address of the server ? Aug 13 03:48:25 monotone server where all the data exists Aug 13 03:50:25 monotone.vanille.de Aug 13 03:51:00 http://oe.handhelds.org/cgi-bin/moin.cgi/MonotonePhraseBook explains all this and is linked from http://oe.handhelds.org/ Aug 13 03:52:00 If you've done the initial pull with the hostname you shouldn't need to specify it again. Aug 13 03:52:02 ok. i think the comand which i gave to checkout from the data base. just gets all the metadata (which is in monotone format) and puts it into a directory. Aug 13 03:52:24 And then you do "mt --db=/somepath/oe.db co --branch=org.openembedded.dev" Aug 13 03:52:45 right now i am using old oe meta data. i have compiled gpe-image also. ok Aug 13 03:53:29 now, i have downloaded the new meta data. if i replace the old meta data in my old set up will it work ?. Aug 13 03:53:58 Pass. I would check out to a new directory rather than overwritting the old stuff. Aug 13 03:54:13 what i am asking is the "compiled packages" in TMPDIR will it work with the new oe meta data just downloaed using monotone Aug 13 03:55:06 If the revisions haven't changed then it shouldn't need to rebuild. How much it needs to rebuild depends on how old your current tree is I guess. Aug 13 04:23:40 good morning all Aug 13 04:37:56 hi Aug 13 04:38:05 hey CoreDump|home Aug 13 04:40:39 umm, which packages "stages" qstring.h and qmap.h? Aug 13 04:41:34 morning Aug 13 04:47:40 hi reenoo_ Aug 13 04:47:55 i copied all the libraries and gqview to my rootfs: when i run gqview i get the error Aug 13 04:47:56 GQview 2.1.1, This is a beta release. Aug 13 04:47:57 Fatal: no entropy gathering module detected Aug 13 04:48:10 how do i resolve this problem. Aug 13 04:48:48 to avoid the first message, I recommend using a non-beta release Aug 13 04:48:56 to avoid the second message, I recommend installing an entropy gathering module Aug 13 04:49:16 :-) Aug 13 04:49:51 when i include my gqview in my .bb file. i expect it to pull all that is needed Aug 13 04:50:46 hey pb_ Aug 13 04:57:16 pb_ : any ideas ?. googling says it is a libcrypt error. have u faced this problem and how do i resolve this problem. Aug 13 04:57:46 cd /dev; ./MAKEDEV random Aug 13 05:02:49 pb_: do you know how important testapi.c (and generated binaries) is for libxml2? Aug 13 05:03:18 if I patch it out libxml2 compiles and packages, but I don't know if something relies on that Aug 13 05:03:31 no idea, sorry Aug 13 05:05:26 in case someone is wondering libxml2 2.6.{11-20} all fail to compile Aug 13 05:05:58 .20 can be fixed by removing testapi.c, but the undefined refs suggest something else might be wrong Aug 13 05:06:49 pb_: sorry to trouble u again : why gqview does not pull all it requies into the filesystem. i get so many messages saying this is missing that is missing no such directory Aug 13 05:08:39 eJumbo: in the case of /dev, it assumes that all the required "standard" devices have already been created. Aug 13 05:11:46 well, i have no problem with /dev. Aug 13 05:12:31 hail zecke Aug 13 05:12:50 i have specified gqview in both depends and rdepends. but in the final image. i dont find gqview and libraries which it requires. Aug 13 05:13:00 that is unfortunate. Aug 13 05:13:29 i run gqview and find out what it requires and copied them manually. some 8 to 10 libs Aug 13 05:13:43 in addition it needs some entries in /etc directory ? Aug 13 05:14:16 no idea. Aug 13 05:28:17 does anyone know the magic incantation to put on the end of a SRC_URI element to change the filename of the downloaded file to something other than what wget thinks it should be? Aug 13 05:28:31 The file in question is: http://www.linksys.com/servlet/Satellite?blobcol=urldata&blobheadername1=Content-Type&blobheadername2=Content-Disposition&blobheadervalue1=application%2Fzip&blobheadervalue2=inline%3B+filename%3DNSLU2_V23R63.zip&blobkey=id&blobtable=MungoBlobs&blobwhere=1123342755292&ssbinary=true Aug 13 05:28:52 and I want it to end up as NSLU2_V23R63.zip :-) Aug 13 05:29:25 isn't it a wget flag? Aug 13 05:29:43 you could check if fetcher.py even supports that kind of things Aug 13 05:30:07 I did that, and it seemed to have a "localpath" parm, but my python-foo is not very good .... Aug 13 05:36:40 man, staging is seriously screwed opie-wise currently. Or someone totally b0rked opie-lrelease Aug 13 05:37:05 I usually resort to mirroring such packages, but I can see the license issues with stuff like firmwares Aug 13 05:50:18 is it possible to get a diff of a specific commit w/ mt? Aug 13 05:53:16 nm, found it. mt diff --revision $num --revision $num does the trick Aug 13 06:05:28 hey mickeyl Aug 13 06:05:57 mithro: still in sweden? Aug 13 06:06:34 yeah Aug 13 06:06:43 i'm on a 6 month study exchange to here Aug 13 06:07:59 sounds nice. where in sweden? Aug 13 06:08:16 stockholm Aug 13 06:11:17 hi mithro Aug 13 06:11:39 any libtool expert in da house? Aug 13 06:11:39 i have a package for which libtool isn't appending the .so extension to the libraries it create. Any idea how to fix that? Aug 13 06:13:42 * mickeyl reads http://people.debian.org/~keybuk/libtool-missing_so.html Aug 13 06:16:05 mickeyl: make it use oe's libtool Aug 13 06:16:23 zecke: how? Aug 13 06:16:31 zecke: i thought we ensured that we are using oe's tools anyway Aug 13 06:16:46 mickeyl: not if the app has it's own version of it Aug 13 06:16:51 hmm interesting Aug 13 06:16:54 mickeyl: rm libtool.m4.in or similiar Aug 13 06:17:06 mickeyl: for example like broken KDE admin/ directory Aug 13 06:17:26 hmm Aug 13 06:17:32 that's hard Aug 13 06:17:44 AC_PROG_LIBTOOL is in acinclude.m4 Aug 13 06:17:48 which contains a _lot_ of macros Aug 13 06:17:55 lets see if I can cut one out Aug 13 06:19:34 moin mickeyl Aug 13 06:19:37 hi zecke Aug 13 06:19:43 hi reenoo_ Aug 13 06:19:52 mickeyl: what makes you believe _append is no longer working? Aug 13 06:20:02 evidence Aug 13 06:20:03 :D Aug 13 06:20:16 i'm in the middle of something else, i'll provide an example when i have a chance Aug 13 06:20:22 oh btw. Aug 13 06:20:24 mickeyl: the filemanager? Aug 13 06:20:26 i didn't say _append is not working Aug 13 06:20:45 i said do_stage is overriding what happens in pkgconfig and binconfig classes Aug 13 06:21:10 koen: kind of. I'm currently fighting against taglib which is a dependency for some nice E apps Aug 13 06:21:17 but fscking libtool is broken Aug 13 06:21:24 heh Aug 13 06:21:30 taglib is KDE technology Aug 13 06:21:39 so it comes with broken libtool by design Aug 13 06:21:40 that explains Aug 13 06:22:01 mickeyl: binconfig.bbclass and pkgconfig.bbclass define do_stage_append (). if _append is working there's no way to override that Aug 13 06:22:01 s/broken/uebergood OO libtool funded by TT/ Aug 13 06:22:24 reenoo_: that's not true, when order of 'inherit' is relevant Aug 13 06:22:32 inherit binconfig Aug 13 06:22:34 do_stage() { Aug 13 06:22:34 foo Aug 13 06:22:35 } Aug 13 06:22:36 boom Aug 13 06:22:57 at least that's what I had observed Aug 13 06:23:18 it seems binconfig appends to do_stage() and the do_stage() in the .bb blows it away Aug 13 06:23:26 s/it/that/ Aug 13 06:23:40 exactly, that's what I have been telling Aug 13 06:23:50 mickeyl: look at freetype(-native). it does exactly that and it works as expected Aug 13 06:23:50 is that another side effect of removing calls to update_data? Aug 13 06:24:08 pleae Aug 13 06:24:11 mickeyl: if that's no longer working it's a bug in bitbake Aug 13 06:24:12 let's talk later Aug 13 06:24:31 * mickeyl getting old and cannot longer concentrate on more than one thing Aug 13 06:25:55 * france is back (gone 08:48:55) Aug 13 06:35:54 when i run gqview: my images files are not visible. eventhough i go to the directory. gqview does not show them at all Aug 13 06:36:02 anybody worked with gqview here Aug 13 06:37:37 i get messages such as these : http://pastebin.ca/19933 Aug 13 06:39:36 koen: actually no. I just upgraded bitbake to current svn and it still works as expected Aug 13 06:39:47 mickeyl: ok. Aug 13 06:50:06 * mickeyl gets crazy and attempts to rewrite the taglib buildsystem in qmake Aug 13 07:12:11 * mickeyl big grin Aug 13 07:12:18 NOTE: package taglib-1.4: completed Aug 13 07:12:23 ~hug qmake Aug 13 07:12:24 * ibot hugs qmake Aug 13 07:12:39 heh Aug 13 07:12:41 very good Aug 13 07:13:09 perhaps it would be the right thing(tm) to fix its libtool usage, but then we could expect me finishing this in a couple of years Aug 13 07:13:10 heh Aug 13 07:13:32 yah, making kde packages work with libtool in oe seems to be something of a nightmare. Aug 13 07:13:57 iirc, even zecke was defeated by that challenge Aug 13 07:14:41 pb_: I filed a bug report ;) Aug 13 07:15:20 aha Aug 13 07:26:43 mickeyl: the TTY bug report made me lookup man:keymaps Aug 13 07:27:00 mickeyl: and my driver is right to get lost in a column Aug 13 07:27:46 oops Aug 13 07:27:53 can you do anything to prevent that? Aug 13 07:28:35 or is that an inherent problem? Aug 13 07:29:21 mickeyl: fix the keymaps ;) Aug 13 07:29:25 hehe Aug 13 07:29:27 fair enough Aug 13 07:29:48 mickeyl: I did something yesterday and it seems to work Aug 13 07:40:09 JustinP: FYI: NOTE: package eclair-20050810: completed Aug 13 07:40:12 *phew* Aug 13 07:40:55 yay! Aug 13 07:40:59 :D Aug 13 07:41:02 if only e-wm would build Aug 13 07:41:11 it does here Aug 13 07:41:14 what bails out for you? Aug 13 07:41:23 lots of stuff Aug 13 07:41:25 * koen checks Aug 13 07:41:48 ERROR: function do_install failed Aug 13 07:41:48 ERROR: log data follows (/home/koen/OE/build/tmp/familiar/work/arm-linux/e-wm-20050810-r0/temp/log.do_install.19824) Aug 13 07:41:51 | NOTE: make DESTDIR=/home/koen/OE/build/tmp/familiar/work/arm-linux/e-wm-20050810-r0/image install Aug 13 07:41:54 | make: *** No rule to make target `install'. Stop. Aug 13 07:41:57 | FATAL: oe_runmake failed Aug 13 07:42:02 eek Aug 13 07:42:11 well, I didn't try after Justin's recent commits Aug 13 07:42:20 * mickeyl does Aug 13 07:43:25 Anybody know what would cause apps like qpdf, opiemediaplayer2, etc to hang when they're searching directories on startup? Aug 13 07:45:02 no idea Aug 13 07:47:11 the package gqview does not get added eventhough i have specified it in rdepends Aug 13 07:47:38 Weird. qpdf locked my zaurus up completely. Might be time to rebuild the image. Aug 13 07:47:39 try deleting TMPDIR and rebuild Aug 13 07:48:29 koen: does not one used gqview before ? Aug 13 07:48:31 when you add a package to rdepends, you need to rebuild that package Aug 13 07:48:40 otherwise they don't end up in the package Aug 13 07:49:13 eJumbo: it works over here, so I have no idea why it fails for you Aug 13 07:49:46 i did bitbake -c clean gqview ; bitbake -c clean basic-image; bitbake basic-image Aug 13 07:51:43 my .bb file is http://pastebin.ca/19937 Aug 13 07:51:53 * france is away: Away Aug 13 07:53:23 koen: can u have a look at my .bb file Aug 13 07:56:29 BigAl: take it up with the guy who forked qpdf Aug 13 07:56:42 save yourself that, he abandonded it Aug 13 07:56:48 ;) Aug 13 07:57:16 eJumbo: the gqview in the repo works for me and Xora Aug 13 07:57:25 eJumbo: why don't you use that one? Aug 13 07:57:50 which repo ? Aug 13 07:57:56 did not get u ?. Aug 13 07:58:03 the default repo Aug 13 07:58:17 i get the ipks. no problems. but it does not get added to the final image. Aug 13 07:58:30 i have used that Aug 13 07:58:53 koen: anything wrong with my bb file Aug 13 08:03:51 if it is built but not added to the final image then Aug 13 08:04:02 it is missing from the ${IPKG_INSTALL} Aug 13 08:04:06 mickeyl: why do we use loadkeys instead of loadkmap for the keymaps on C7xx Aug 13 08:04:08 or nothring RDEPENDS on it Aug 13 08:04:20 US army is indirectly using OE: https://wiisard.org/Zope/wiisard and http://www.linuxdevices.com/news/NS2239811258.html Aug 13 08:04:25 zecke|tv: i think because we wanted to have an easily changeable textfile Aug 13 08:04:32 zecke|tv: iirc kmap needs a binary format Aug 13 08:04:36 but i'm not sure Aug 13 08:04:48 mickeyl: right kmap is binary Aug 13 08:05:00 once the keymaps are settled we can change that Aug 13 08:06:41 mickeyl: did not get u ?. where else should i add other than rdepends ? Aug 13 08:06:59 * CoreDump|home notes that ASCII keymaps are more userfriendly as they can be "hacked" pretty easily Aug 13 08:07:14 CoreDump|home: 23kb vs. 2kb ;) Aug 13 08:07:15 eJumbo: for instance to IPKG_INSTALL Aug 13 08:07:23 good point heh Aug 13 08:07:24 did you read any of our image bb's? Aug 13 08:07:42 at least jffs2 uses compression Aug 13 08:07:46 ok Aug 13 08:08:08 i have n number of packages; all get into the image except one. Aug 13 08:08:41 CoreDump|home: but blockwise ;) Aug 13 08:09:00 :) Aug 13 08:09:24 * CoreDump|home looks at a whoopin 128MB flash and thinks "so what" ;) Aug 13 08:09:43 i have exported IPKG_INSTALL="task-basic" and task-basic is my file where i have specified gqview Aug 13 08:12:30 so task-basic RDEPENDS on gqview ? Aug 13 08:12:40 are you sure? look into the ipkg file Aug 13 08:14:02 yes Aug 13 08:14:25 my .bb file is http://pastebin.ca/19937 Aug 13 08:14:57 good good Aug 13 08:15:11 does it really really appear in the ipkg file ? Aug 13 08:16:07 i dont understand when u say ipkg file. which ipkg file. basic-image.ipk ? Aug 13 08:16:26 look into task-basic_.ipk Aug 13 08:16:34 make sure your rdependencies are in there Aug 13 08:26:54 mickey: the ipk file does not have a dependency gqview Aug 13 08:36:51 if the my final image dependecy file has no dependency for gqview. what is the problem ? Aug 13 08:43:16 cant really figure out what the problem is ? Aug 13 08:43:33 clean tmp and rebuilt Aug 13 08:43:41 that usually fixes that kind of problems Aug 13 08:46:37 or use PR Aug 13 08:47:01 zecke: can u explain this a bit Aug 13 08:47:22 eJumbo: even more? Aug 13 08:48:04 :-) Aug 13 08:48:10 :) Aug 13 08:48:18 eJumbo: PR = PackageRelease Aug 13 08:48:29 eJumbo: if you have something with the same version already built Aug 13 08:48:34 OE can be confusing for newbies ;) Aug 13 08:48:52 and the PR is higher it will be automatically built Aug 13 08:48:53 anyways, bbl Aug 13 08:49:03 i dont think it is that problem. Aug 13 08:49:18 i dont have 2 releases and ipk conflict situation Aug 13 08:49:24 i have only one version. Aug 13 08:49:48 i have listed gqview package in rdepends. but it did not get included in the image Aug 13 08:50:11 i checked the ipk control file and there is no mention of gqview in depends section of ipk Aug 13 09:18:54 mickey|bbq: | /home/koen/OE/build/tmp/familiar/cross/lib/gcc/arm-linux/3.4.4/../../../../arm-linux/bin/ld: cannot find -ltag Aug 13 09:29:36 eumel: and you said your task-*.ipk does not depend on it Aug 13 09:29:42 well Aug 13 09:29:45 eumel: sorry Aug 13 09:29:52 zecke|tv: too slow Aug 13 09:30:08 pb_: Yeah I prefered some chinese food ;) Aug 13 09:30:19 good choice Aug 13 09:30:29 pb_: 1-800-CHINESE-FOOD ;) Aug 13 09:31:22 * zecke|tv finally buys his iBook on Tuesday *yipee* Aug 13 09:31:32 excellent Aug 13 09:31:40 I should try to fix my iBook one day. Aug 13 09:31:52 zecke|tv: congrats Aug 13 09:31:54 pb_: what Is broken? Aug 13 09:32:09 zecke|tv: the power connection Aug 13 09:32:59 pb_: oh is it poorly assembled? Aug 13 09:33:16 no, I think it's just old and worn out. Aug 13 09:33:21 I dropped it on the floor several times. Aug 13 09:33:43 (over a period of a couple of years, I hasten to add, not all in one go) Aug 13 09:34:51 I bought this Sharp to replace it with, so I wasn't in too much of a rush to fix the ibook. Aug 13 09:40:06 sheesh people keep trashing their image w/ ipgk upgrade.. Aug 13 10:47:49 zecke|tv: The c7x0 used loadkeys as I couldn't find a way to make bitbake easily generate the binary format from a text input - it needed to run on device :-/ Aug 13 10:48:29 zecke|tv: I would like to see packaging split on console_tools but I think I'd also like to stick to ASCII config files... Aug 13 10:48:51 "morning" all btw :) Aug 13 10:48:54 NOTE: package e-image-1.0: completed, hey this actually builds :) Aug 13 10:49:06 it does? Aug 13 10:49:08 hmmm Aug 13 10:49:26 * koen tries again Aug 13 10:49:43 entrance, however does not (for me) Aug 13 10:50:03 ok Aug 13 10:50:11 so it's not me Aug 13 10:51:39 yep, some native app b0rking on some db Aug 13 10:51:54 hi RP Aug 13 12:20:21 w000t Aug 13 12:20:45 entrance is working? Aug 13 12:21:03 no but e-image :D Aug 13 12:21:31 needs a startx or something like that though Aug 13 12:25:30 mickey|bbq: so you got taglib/eclair/evidence to work? Aug 13 12:25:39 mickey|bbq: I also have been working on those.... Aug 13 12:26:08 mickey|bbq: I have bb files and some sh script to fix the libs so taglib will package but it breaks eclair...eclair expects no .so for some odd reason Aug 13 12:26:25 mickey|bbq: anyway...if what you have works I suppose you should commit it Aug 13 12:26:34 when i try to build anything with OE i get "ERROR: function do_patchcleancmd failed" and it spits out what looks like help info for usage of the "quilt" command Aug 13 12:26:38 how do i adress this? Aug 13 12:26:44 koen: e-wm is failing for you? It should all be fixed now. It's all compiling fine for me. Aug 13 12:26:59 slisk: hi Aug 13 12:27:05 slisk: are you setting up your environment cororetly? Aug 13 12:27:13 slisk: do you mix a recent bitbake with an anicent OE tree? Aug 13 12:27:39 JustinP: taglib and entrance also fail to build Aug 13 12:27:51 i dl'ed the OE tree just last week and got a the current bitbake at the time Aug 13 12:28:01 i env vars are set too Aug 13 12:28:11 err the... Aug 13 12:28:34 koen: well taglib is just screwed up Aug 13 12:28:40 koen: but entrance should work Aug 13 12:28:42 what fails? Aug 13 12:28:49 slisk: pastebin your error please Aug 13 12:28:51 generating the db Aug 13 12:28:56 ~pastebin Aug 13 12:28:56 somebody said pastebin was a place to paste your stuff without flooding the channel - try http://pastebin.ca Aug 13 12:28:59 is that edje or eet's job? Aug 13 12:29:10 "the db"? Aug 13 12:29:20 entrance_config.db? Aug 13 12:29:26 that would be edb Aug 13 12:29:39 are you using my recent patches....? They worked fine on my machine Aug 13 12:29:41 CoreDump|home: do you have the error handy? Aug 13 12:29:58 koen: checking Aug 13 12:30:00 zecke|tv: yes, entrannce should be working Aug 13 12:30:09 my tree does a monotone update every 10 minutes, so yes, it's up to date Aug 13 12:30:15 CoreDump|home: no, I'm still working on getting things to work sanely, no e-image *yet* Aug 13 12:30:22 koen: ok...hmmm Aug 13 12:30:28 perhaps some1 can help me with the archived openzaurus metadata build process... Aug 13 12:30:34 when i try to build anything using OZ's archive metadata, i get the error "tar: /home/stuff/zaurus/zaurus-build/oz/sources/config_savannah.gnu.org__20050403.tar.gz: Cannot open: No such file or directory"... Aug 13 12:30:42 JustinP: and I rm -rf TMPDIR before every try Aug 13 12:30:53 koen: that must take forever... Aug 13 12:31:01 koen: I haven't tried removing tmpdir yet... Aug 13 12:31:02 how do i address that? i think it'd be of more help than the general oe build Aug 13 12:31:07 ahem, looks like it compiled w/ e-image Aug 13 12:31:34 JustinP: depends on the machine, about 4 hours on a dual box, but a lot more on the uniprocessor box Aug 13 12:31:50 the smp boxes are running tinderbuilds, so I can't use those Aug 13 12:31:52 btw: what's pastebin? Aug 13 12:31:57 ~pastebin Aug 13 12:31:57 rumour has it, pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.ca Aug 13 12:31:58 slisk: rm that file Aug 13 12:32:29 ....16 revs in but no commits to the list or IRC channel....I don't know what people have alteres now... Aug 13 12:32:53 ok i'll post the output and rm that fiel and try again Aug 13 12:33:00 guess i'll do the latter 1st Aug 13 12:33:22 who updated engage....it worked for me when I committed it... Aug 13 12:33:31 entice, example....what's going on here...? Aug 13 12:33:49 I wish I had commit e-mails Aug 13 12:34:13 well without a specific error I can't really help... Aug 13 12:34:34 I'll try my build again but I built entrance over and over last night and it worked Aug 13 12:35:00 I'm at NOTE: package gcc-cross-initial-3.4.4-r1: task do_compile: started right now Aug 13 12:35:07 doing a bitbake e-image Aug 13 12:35:09 ::sigh::....ok Aug 13 12:35:21 I haven't even tried that yet Aug 13 12:35:33 it's a placeholder mostly....it may not even work Aug 13 12:35:54 <- is currently booted into an e-image Aug 13 12:36:03 koen: can I start a tinderclient on ewi as well Aug 13 12:36:14 koen: it shouldn't be cpu intensive at all (source checker) Aug 13 12:36:44 zecke|tv: sure Aug 13 12:36:50 CoreDump|home: /etc/init.d/entrance start Aug 13 12:36:53 does that work? Aug 13 12:36:56 sec Aug 13 12:37:03 no, it won't let you login... Aug 13 12:37:10 entrance works...but you need some more files Aug 13 12:37:21 I didn't check in the Session files Aug 13 12:37:33 I was up to 4AM getting entrance to work right as it is Aug 13 12:37:35 i get this http://pastebin.ca/19951 error trying to build a package using OZ 3.5.3 archived metadata. can any1 help? Aug 13 12:37:53 slisk: with a recent bitbake? Aug 13 12:38:06 if so, that won't work Aug 13 12:38:10 BitBake Build Tool Core version 1.3.2, bitbake version 1.3.2 Aug 13 12:38:27 you'll also need a bitbake version from the time 3.5.3 was released Aug 13 12:38:59 * JustinP thinks you're all crazy for trying to play with the e-image before anyone has even worked on it yet Aug 13 12:39:07 JustinP: entrance froze my Z heh Aug 13 12:39:10 bbl Aug 13 12:39:19 I'm using the GPE image and altering it for my testing Aug 13 12:39:32 bitbake isn't backwards compatible?!!? Aug 13 12:39:40 no, it isn't Aug 13 12:39:48 how do i find out what version was used then? Aug 13 12:40:07 do a svn checkout of the date OZ was released Aug 13 12:41:22 Which bitbake behaviour is not backwards compatible? Aug 13 12:41:38 i don't know how to do a svn checkout but the site that hosts the metadata dates the archive 22-Apr-2005 08:13 Aug 13 12:41:52 when I said that entrance worked I meant that you can now make it work sanely ;-) Aug 13 12:42:09 all right, I have to go Aug 13 12:42:26 next time you get an e error please send it my way Aug 13 12:42:36 I'm trying a build again from scratch to test as well Aug 13 12:44:01 pb_: how detailed do you want the answer? Aug 13 12:44:27 zecke|tv: pretend I am some kind of pointy haired boss Aug 13 12:44:40 give me the executive's eye view Aug 13 12:44:56 pb_: oh well. I had to remove some default parameters that shadowed the existance of errors Aug 13 12:45:13 pb_: older OE trees are simply broken and it is not bitbake being incompatible Aug 13 12:45:24 hmm not that good answer, I will idle some more and try again Aug 13 12:45:35 ah, I see Aug 13 12:45:37 that is unfortunate Aug 13 12:45:55 the override _prepend thing being the latest example Aug 13 12:46:05 pardon? Aug 13 12:46:20 koen: that is not bitbake's business (backward compatible) Aug 13 12:46:38 the kernel24/kernel26 problem in h3900.conf Aug 13 12:47:23 zecke|tv: I'll archive both OE and bitbake for the updates for familiar 0.8.3 Aug 13 12:48:20 oh, right, you mean OVERRIDES_append. Aug 13 12:48:24 sounds sane Aug 13 12:48:40 yes, that was unfortunate, but nothing except the h3900 was likely to have been affected. Aug 13 12:49:35 I do not expect to have to do incompatible changes anymore Aug 13 12:50:36 and iirc 3.5.3 was before the great memory reduction change Aug 13 12:50:40 right Aug 13 12:50:59 zecke|tv: how hard would it be to make a tinderserver out of bitbake? Aug 13 12:51:20 koen: tinderserver as in generating html pages? Aug 13 12:51:46 and storing summarry logs + errors Aug 13 12:51:47 koen: or as in processing mails Aug 13 12:52:58 koen: not hard, write your own bbclass Aug 13 13:20:51 good nite Aug 13 13:21:16 zecke|tv: night night Aug 13 13:21:52 crappy done, nothing got done Aug 13 13:22:25 * CoreDump|home hugs E Aug 13 13:41:26 ouch, xtscal is messing the calibration up bad Aug 13 13:50:04 xtscal -rotate 0 is the key Aug 13 14:04:05 * reenoo_ wonders what happened to the fetcher Aug 13 14:04:27 it doesn't resume cancelled downloads any more Aug 13 14:05:22 reenoo_: weird Aug 13 14:08:09 koen: I hope to not consume the remaining 34gb on ewi Aug 13 14:11:29 zecke|tv: there's always 'rm' Aug 13 14:11:40 or remove the source after downloading Aug 13 14:12:03 koen: what is more costy for you, diskspace or bandwidth? Aug 13 14:12:23 unlimited bandwith Aug 13 14:12:30 and the box is completely ours Aug 13 14:12:43 koen: plan is to run the source_checker once and then incremental 'stat' the sources Aug 13 14:13:07 koen: Assuming I do not allocate the full 34gb I'm finally going to bed now Aug 13 14:13:21 ok Aug 13 14:13:32 'night zecke|tv Aug 13 14:13:42 koen: /home/zecke/bitbake/bitbake_qa/output.log Aug 13 14:13:46 * koen will finish watching a movie with kurt russel Aug 13 14:13:54 (no brainpower need to watch) Aug 13 14:15:14 23:10:14 (11.00 MB/s) - `/home/zecke/bitbake/bitbake_qa/tmp_sources/abiword-2.2.7.tar.gz' saved [28441035/28441035] Aug 13 14:15:29 good thing the main abiword mirror is on the same network :) Aug 13 14:15:52 :) Aug 13 14:15:55 zecke|tv: set DEBIAN_MIRROR to ftp.nl.debian.orgB Aug 13 14:15:59 -B Aug 13 14:16:06 that's also on the same network Aug 13 14:16:07 koen: okay tomorrow then Aug 13 14:16:21 koen: and what for sf.net? Aug 13 14:16:37 aeleron or belnet Aug 13 14:16:53 koen: ftp.nl.gnu.org? Aug 13 14:17:01 that should work Aug 13 14:17:10 unbelievable just 2mb/s Aug 13 14:17:57 feature request: torrent fetcher Aug 13 14:18:15 yeah tomorrow is bitbake hacking day Aug 13 14:18:30 tomorrow will be ipkgfind hacking day :) Aug 13 14:18:48 koen: and next week I might be even stupid enough to try OE on MacOS Aug 13 14:18:50 * CoreDump|home bitbakes e-utils Aug 13 14:19:06 CoreDump|home: that one bombs out over here Aug 13 14:19:11 CoreDump|home: I will fix uicmoc3 tomorrow :} Aug 13 14:19:13 zecke|tv: tell me if you get it to work Aug 13 14:19:21 I want it on my powerbook Aug 13 14:19:24 zecke|tv: no sweat Aug 13 14:19:28 koen: ok Aug 13 14:19:40 CoreDump|home: no sweat? Aug 13 14:19:56 keinen stress ;) Aug 13 14:20:08 ah 'sweat' Aug 13 14:20:50 koen: NOTE: package e-utils-20050810: completed Aug 13 14:21:08 hmmm Aug 13 14:21:20 I'll try another build tomorrow Aug 13 14:21:33 * koen will be pissed if it only works on distro=OZ Aug 13 14:21:33 and eclair bombs w/ taglib failing Aug 13 14:21:42 :) Aug 13 14:33:34 NOTE: package aspell-0.50.5: completed Aug 13 14:33:34 NOTE: build 200508131256: completed Aug 13 14:33:34 Build statistics: Aug 13 14:33:34 Attempted builds: 16 Aug 13 14:33:41 * slisk sheds a tear Aug 13 14:33:56 so i have to up to date within 1 hr of building eh? Aug 13 14:55:28 gstreamer: bump to gstreamer-0.8.10. apply patch submitted in gnome bug #313322. move patches around a bit. Aug 13 14:55:39 gst-plugins: bump to gst-plugins-0.8.8. update try-esd.patch to work with current code. Aug 13 15:16:50 does any1 know how to change the lcd light's brightness on a zaurus 5500 running OZ 3.5.3? Aug 13 15:17:11 try #openzaurus Aug 13 15:17:56 i asked there earlier no response Aug 13 15:35:06 there are the commit e-mails...? Aug 13 15:35:09 where Aug 13 15:36:14 cia is dead again Aug 13 15:36:29 and/or ciabot.py is stuck on vanille Aug 13 15:42:01 well there's nothing in here either Aug 13 15:42:24 both I guess Aug 13 15:42:35 oe-commits is silent as well Aug 13 15:42:42 hence the "either" Aug 13 15:42:51 and my mention of e-mails above Aug 13 15:51:36 ::sigh:: I screwed up by classes....starting over again Aug 13 16:08:48 'night all Aug 13 16:10:38 night Aug 13 17:02:41 i'm considering installing debian onto the root of my zaurus 5500. i realize it's best and probably necessary to still use the OZ kernel though. is there anything else i would need to transport to a debian distro if i were to use one? Aug 13 17:02:44 btw, i don't like pocket debian's method of vnc'ing to another distro on the same device Aug 13 17:03:59 * chouimat is away: charlie jade on tv Aug 13 18:05:35 * chouimat is back. Aug 13 18:34:23 slisk, you are aware taht debian wont fit right? Aug 13 18:34:39 unless the 5500 has over 500 in the flash Aug 13 18:40:26 i know how to deal with that, my only concern is what mix of software i'll need Aug 13 22:21:04 damn this xorg..... gentoo's multilib includes is messing it up :P Aug 13 22:29:13 heya raster Aug 13 22:30:50 hey emmie Aug 13 22:31:17 how goes it/ Aug 13 22:31:18 ? Aug 13 22:33:08 grrr Aug 13 22:33:17 have to friggin boot into winfuck Aug 13 22:33:21 lol Aug 13 22:33:33 raster, never fun to have to do that :) Aug 13 22:33:33 time to install wine? Aug 13 22:36:18 *sob* that compile refused to work too... starting to get irritated on xorg :P Aug 13 22:36:33 ? Aug 13 22:37:22 the gentoo-multilib includes is messing up the crosscompilation Aug 13 22:37:52 why should your host lib have anything to so with your oe bitbake system? Aug 13 22:37:58 do* Aug 13 22:38:04 and sitting patching makefiles to use the STAGING_INCDIR instead of /usr/include Aug 13 22:38:39 you doing your own version of xorg or using the ones in the packages? Aug 13 22:39:05 well, it's using /usr/include for all the native-utils that are needed for the compilation... Aug 13 22:39:25 using one of the premade pkg's Aug 13 22:39:51 then i dount the metafile is wrong, its more likly your enviroment setup Aug 13 22:39:56 doubt* Aug 13 22:40:25 xorg has been successfully compiled for almost a year Aug 13 22:40:34 in gentoo also? Aug 13 22:40:45 the host is irrespective Aug 13 22:40:46 s/in/on/ Aug 13 22:41:28 there are people who use gentoo here, i am not one of them Aug 13 22:44:37 well, i dont think there should be anything with the env atleast.. Aug 13 22:46:05 any idea of what to look for except for generic include-settings? (if there are any OE specific settings for that) Aug 13 22:47:30 yes look at your log files Aug 13 22:49:02 well, in the buildlog all that i can see is that it's having problems from time to time with builds that's native to the buildsystem... like mkg3states and such Aug 13 22:51:34 and it's quite expected that it's using the systemincludes since when doing the native-compile it does not include any searchpath to the staging-includes... Aug 13 22:54:08 hmm Aug 13 22:54:18 true Aug 13 22:54:28 cc -I../../../exports/include -o mkg3states -O2 -fno-strength-reduce -fno-strict-aliasing -ansi -pedantic -Wall -Wpointer-arith -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations -Wredundant-decls -Wnested-externs -Wundef -I../../../exports/include/X11 -I../../../include/extensions -I../../../programs/Xserver/include -I../../../include/extensions -I../../.. -I../../../ex Aug 13 22:54:28 ports/include -Dlinux -D__i386__ -D_POSIX_C_SOURCE=199309L -D_POSIX_SOURCE -D_XOPEN_SOURCE -D_BSD_SOURCE -D_SVID_SOURCE -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE -DFUNCPROTO=15 -DNARROWPROTO mkg3states.c Aug 13 22:54:32 to be more exact Aug 13 22:54:59 what is with all the ../../../ Aug 13 22:55:36 and ofcourse it's quite understandable that it should use the system includes for those, since it's building a native tool... but then the gentoo-multilib-include stuff messes it up... Aug 13 22:55:53 that's just relative path's in the xc's buildfolder Aug 13 22:55:56 nothing to worry about Aug 13 22:56:19 they may cause issues Aug 13 22:56:48 hmm Aug 13 22:56:58 we use -02 Aug 13 22:57:08 interesting Aug 13 22:57:10 well, that's the standard build... and those are correct... they just refer to includes that's in the src for xorg... and those are needed. Aug 13 22:57:25 well, the problem is not with the optimization Aug 13 22:57:29 the problem is: Aug 13 22:57:32 In file included from /usr/include/bits/stdio.h:7, Aug 13 22:57:33 from /usr/include/gentoo-multilib/x86/stdio.h:828, Aug 13 22:57:33 from /usr/include/stdio.h:3, Aug 13 22:57:33 from mkg3states.c:66: Aug 13 22:57:33 /usr/include/gentoo-multilib/amd64/bits/stdio.h:35: error: redefinition of 'vprintf' Aug 13 22:57:35 /usr/include/gentoo-multilib/amd64/bits/stdio.h:35: error: redefinition of 'vprintf' Aug 13 22:57:40 /usr/include/gentoo-multilib/amd64/bits/stdio.h:42: error: redefinition of 'getchar' Aug 13 22:57:41 /usr/include/gentoo-multilib/x86/bits/stdio.h:42: error: previous definition of 'getchar' was here Aug 13 22:57:44 /usr/include/gentoo-multilib/amd64/bits/stdio.h:42: error: redefinition of 'getchar' Aug 13 22:57:46 /usr/include/gentoo-multilib/x86/bits/stdio.h:42: error: previous definition of 'getchar' was here Aug 13 22:58:19 your building on a 32/64/ Aug 13 22:58:21 ? Aug 13 22:58:24 yep Aug 13 22:58:31 amd64 Aug 13 22:58:44 yeah... i dunno if you can yet Aug 13 22:59:23 there are issues with it Aug 13 22:59:35 well, this is the first real issue that i have run into Aug 13 23:00:05 been writing .bb's like crazy to get all the dep's for freevo =) Aug 13 23:00:10 i cnat remember who else here has a 32/64 Aug 13 23:00:32 there are definite issues tho with certain things Aug 13 23:00:54 mm... noticed that Aug 13 23:03:02 stupid fkn emacs Aug 13 23:03:21 should be ran over and drowned Aug 13 23:05:12 hehe Aug 13 23:10:46 i feel that way about vi Aug 13 23:12:31 nah vi is good Aug 13 23:12:46 its this stupid -batch lisp parsing garbage Aug 13 23:13:06 autoconf is barfing on it Aug 13 23:13:21 because emacs isnt processing it Aug 13 23:13:37 its just opening up the emacs console and stopping Aug 13 23:17:23 i think its a missing file issue, but why am i the only one having it ... Aug 13 23:35:16 <_law_> hey i am unable to build e-image (http://pastebin.com/336412) ERROR: Cannot satisfy the following dependencies for task-enlightenment-x11 Aug 13 23:35:49 lol Aug 13 23:36:01 i dount it would build Aug 13 23:36:13 unless JustinP has looked at it Aug 13 23:36:29 it was based on a set of files over a year old Aug 13 23:37:00 and e is not overly usable for most devices Aug 13 23:37:48 <_law_> emte: ok thanks so i´ll ask JustinP perhabs he can help me Aug 13 23:38:09 be aware its not usable on most devices Aug 13 23:38:28 unless you have a system to mimic the otehr two mouse buttons Aug 13 23:43:27 _law_, where is the rest of the error? Aug 13 23:43:38 what you have isnt really useful Aug 13 23:46:44 anyway i better lie down before my knee tries to swell Aug 13 23:48:52 hmm... about sending in new .bb's... Aug 13 23:50:52 where to? Aug 14 01:12:28 moin Aug 14 02:01:11 morning all Aug 14 02:02:24 ' Aug 14 02:05:22 hi RP & zap Aug 14 02:05:27 zecke even Aug 14 02:11:07 hey Aug 14 02:17:45 Noodles: I've added the spitz backlight. Its not working properly though - something wrong with the ssp routines I suspect... Aug 14 02:19:18 morning koen Aug 14 02:19:34 hey RP Aug 14 02:19:39 good morning all Aug 14 02:24:19 hi koen Aug 14 02:27:59 hey CoreDump|home Aug 14 02:31:02 koen: I believe we can run into a problem if we try to ship an "e" image in 3.5.4. Aug 14 02:31:13 http://pastebin.com/336465 Aug 14 02:31:20 it's rather large Aug 14 02:37:26 CoreDump|home: I bet it has 20MB worth of themes Aug 14 02:37:33 but yes, it will be big Aug 14 02:37:44 actually it only comes with the default theme ;) Aug 14 02:38:01 I bet it has 20MB worth of theme Aug 14 02:38:03 ;) Aug 14 02:38:20 hehe Aug 14 02:38:29 I'll check what we can shave off Aug 14 02:44:44 time to give it a check now then... anyone running OE on a amd64 gentoo-system? Aug 14 02:45:05 maybe JustinP Aug 14 02:45:21 I used to run in on amd64/debian Aug 14 02:45:47 koen, ok... i'll see if i manage to catch him a bit later then... Aug 14 02:46:04 I just did 'linux32 screen' and worked from there Aug 14 02:46:38 got some xorg problems that i wanted to check if someone else had... (the native utils for xorg uses /usr/include for some stuff and is causing some problems) Aug 14 02:48:44 koen: hmm... might give that a try Aug 14 02:49:29 koen: thanks for the tip... Aug 14 02:50:13 btw, submitting new bb files.. any special procedure for that? Aug 14 02:50:56 attach them to bugzilla Aug 14 02:51:45 hehe... hope it supports big attachments :) Aug 14 02:52:28 been adding about everything needed to build a OE freevo dist... and on the final tuning now with getting xorg to compile Aug 14 02:53:51 cool Aug 14 02:53:59 freevo is usuable nowadays? Aug 14 02:54:23 yep... quite well, thoe i'm waiting for the 2.x version to be marked as stable Aug 14 02:54:25 e.g you doon't have to edit a zillion python files just to see a picture? Aug 14 02:54:38 but for TV viewing and movie-viewing and such it Aug 14 02:54:44 it's quite nice Aug 14 02:54:51 hehe Aug 14 02:54:56 I looked at freevo ~1.5 years ago Aug 14 02:55:38 well, just edit one single file and specify the path's for mediafiles etc and have a file that specifies the freq's for the tv-channels and some other small stuff Aug 14 02:55:42 you had to manually fill in the channel list in some undocumented format inside a python file Aug 14 02:55:48 hehe Aug 14 02:55:57 so I installed mythtv and that 'worked' Aug 14 02:56:02 well, it's not that basic anymore... Aug 14 02:56:19 unstable as hell, but it does the job Aug 14 02:56:33 well, i have checked mythtv, but i dont like to have a mysql db running, and also have to manually add all new mediafiles to it's library Aug 14 02:57:01 it autoscans a directory now Aug 14 02:57:08 which is sloooow, but works Aug 14 02:57:44 freevo's design is much nicer Aug 14 02:57:57 and it doesn't use a forked version of the nuv codec Aug 14 02:58:09 Just got me a Epia SP13000 board and a CF-IDE adapter... so i'm planning to run the whole OS from a ramdisk and just boot from the CF card, and then have a 200Gb drive to just store mediafiles mounted ro as long as i'm not recording anything... Aug 14 02:58:23 safe from filesystem corruption in case of a powerfailure Aug 14 02:58:55 you could look at the mythfront-image Aug 14 02:59:15 yep, but i dont want to have another machine running just to watch stuff Aug 14 02:59:15 pb_ uses that for his epia Aug 14 02:59:20 ah, ok Aug 14 02:59:28 just want a single machine that manages everything regarding this.. Aug 14 03:00:07 and also planning to get some small bittorrent client to have on the system with rss-feed monitoring :) Aug 14 03:00:24 I heard ctorrent works quite well Aug 14 03:01:03 mm... or just use wget + some small perl/python xml-parser Aug 14 03:01:35 and then use the "official" torrent-client Aug 14 03:02:21 ah... ctorrent, the console-based client Aug 14 03:03:10 i was thinking about torrentflux first.. Aug 14 03:03:14 people are running that on their linksys devices (wrt54x, nslu2, etc) Aug 14 03:05:22 'morning pb_ Aug 14 03:06:06 mm... need to dig into that later on... just need to get xorg running first so i can play movies in something other than the aa video output :) Aug 14 03:10:58 hi koen Aug 14 03:14:43 koen: linux32 wont help on gentoo... might be good to know Aug 14 04:26:10 morning Aug 14 05:38:12 mickey|bbq: I'm now having a qt3e.bbclass and uicmoc3,lupdate and lrelease make use of it Aug 14 06:15:58 hey gerwinin Aug 14 07:51:56 CoreDump|home: you could try opie-lupdate/lrelease again Aug 14 07:52:21 zecke: will do today Aug 14 08:10:20 does any1 know how to control the backlight on a zaurus 5500 from the command line? i want to make a script to lower the light level upon boot Aug 14 08:17:58 slisk: this is the wrong channel to ask Aug 14 08:18:10 slisk: if you use Opie you could utilise qcop Aug 14 08:58:15 Hi, recently bitbake fails to compile for me. Last couple of lines are at http://rafb.net/paste/results/KuJCAb45.html Aug 14 08:58:40 monotone repository is up-to-date. Aug 14 08:59:15 Laibsch: gcc3.4.4 seems not compilable with gcc4? Aug 14 08:59:51 Maybe. I don't know. I guess that might be related to my problem? ;-) Aug 14 09:00:09 zecke: I have no clue what you're getting at. Aug 14 09:00:33 Laibsch: line one of your log Aug 14 09:00:37 ccache gcc... Aug 14 09:00:54 it is your host compiler, it is not capable of compiling your cross compiler Aug 14 09:01:17 Laibsch: do you know what a compiler is? Aug 14 09:01:33 yes. Aug 14 09:01:43 OE worked fine previously. Aug 14 09:01:45 Laibsch: do you know the GNU Compiler Collection? Aug 14 09:01:54 Laibsch: whatever Aug 14 09:01:56 gcc -v Aug 14 09:02:00 gcc --version even Aug 14 09:02:07 it will say it is 4.0.x Aug 14 09:02:13 I run Debian testing, maybe gcc got updated? Aug 14 09:02:25 and gcc 4.0.x can not compile a gcc 3.4.4 Aug 14 09:02:30 Laibsch: that could be Aug 14 09:02:36 It says 3.3.5 Aug 14 09:02:36 install gcc3.3 or gcc3.4 Aug 14 09:02:42 well then... Aug 14 09:02:55 gcc version 3.3.5 (Debian 1:3.3.5-13) Aug 14 09:03:08 then what about looking at line 210 of friend.c? Aug 14 09:04:34 Laibsch: alternatively try to stock gcc3.4 using PREFERRED_VERSION Aug 14 09:05:44 Laibsch: what is bitbake trying to build? gcc-cross? gcc-cross-initial? Aug 14 09:06:12 line 210 looks strange Aug 14 09:06:20 if ^@^\uffff)\uffff^@^@^@^@^@^@^@^@^\\ufffff\uffff`\uffff9\uffff^@^@^@^@^@^\uffff)\ufffftemplate_class_depth (type)) Aug 14 09:06:39 Laibsch: corrupt disk? Aug 14 09:07:00 I do not think so. Just made an e2fsck the other day. Aug 14 09:07:02 Laibsch: try cleaning the gcc sources from your DL_DIR (normally sources) Aug 14 09:07:08 OK. Aug 14 09:07:12 Laibsch: and try rebuilding Aug 14 09:08:04 I guess that would be ./tmp/work/gcc* ? -> rm -Rf ./tmp/work/gcc* Aug 14 09:09:04 Laibsch: if you like sticking stuff into someone butt's Aug 14 09:09:15 ? Aug 14 09:09:25 bitbake -b foo -c clean Aug 14 09:09:32 bitbake -b path/to/foo.bb -cclean does the same in a sane way... Aug 14 09:09:53 Laibsch: but for the source directory something like you posted could work Aug 14 09:11:35 Thanks guys. I'm off for a rebuild. Aug 14 09:12:06 Laibsch: have fun, and get back to us Aug 14 09:12:14 Sure. Aug 14 09:12:26 Laibsch: oh wait ;) Aug 14 09:12:32 did you delete the sources already? Aug 14 09:12:42 I am in the midst of it. Aug 14 09:12:43 you could have run an integrity check on the tar.gz/bz2 Aug 14 09:12:54 which one? Aug 14 09:13:05 dunno off hand Aug 14 09:14:02 Something in the work directory? Aug 14 09:18:12 could some1 plz tell me what's wrong with this simple program? http://pastebin.ca/20009 Aug 14 09:18:37 i get a syntax error use the OE arm cross compiler when i don't think i should Aug 14 09:18:41 morning Aug 14 09:18:54 mickeyl: morning? Aug 14 09:18:59 ~ugt Aug 14 09:19:00 well, ugt is Universial Greeting Time. Created in #mipslinux, it is a rule that states that whenever somebody enters an IRC channel it is always morning, and it is always late when the person leaves. The local time of any other people in the channel, including the greeter, is irrelevant. Aug 14 09:19:05 mickeyl: glad you are awake. Aug 14 09:19:20 This bot is just too smart :-) Aug 14 09:19:24 slisk: why not link to libopie2? Aug 14 09:19:52 anything this bot does not know? Aug 14 09:19:52 slisk: # Aug 14 09:19:53 #define ZAURUS_FL /dev/fl Aug 14 09:19:54 i want a command line utility to change the backlight level without opie Aug 14 09:20:06 oooh Aug 14 09:20:10 slisk: ipkg install slbl Aug 14 09:20:11 :D Aug 14 09:20:24 slisk: but again this is OT here Aug 14 09:20:27 damn i hate c compilers, syntax checking sucks Aug 14 09:20:42 slisk: put a string in quotes... Aug 14 09:20:46 sry didnt' know where else to go, thought it might be a compiler issue Aug 14 09:20:50 yeah i know Aug 14 09:20:57 perl sucks as well dev/fl would be zeroo Aug 14 09:21:05 i just hate being told the problem is 8 lines away Aug 14 09:21:33 zecke: I guess what you meant is something along the lines of md5sum ./sources/gcc_savannah.gnu.org_csl-arm-branch_20050603.tar.gz? That would be 5aa13368781a83718acc25ae3967115c. I am not sure there is an md5 list of that file up on savannah. Aug 14 09:21:55 Laibsch: what bitbake does when fetching Aug 14 09:22:01 Laibsch: it fetches the svn/cvs tree Aug 14 09:22:05 mickeyl: which directory of the feed do i get slbl out of? Aug 14 09:22:05 Laibsch: and tars them with a date Aug 14 09:22:27 I am not aware of that. Aug 14 09:22:31 slisk: no idea. Aug 14 09:22:38 slisk: probably it hides behind the name slutils Aug 14 09:22:45 it's in oe at least Aug 14 09:23:32 yes it is part of slutils, I use slbl for gpe on akita Aug 14 09:23:54 mickeyl: morning dude Aug 14 09:26:39 * _alwin_ is away: afk Aug 14 09:26:46 03mickeyl 07org.oe.dev * ra3f797d9... 10/packages/meta/meta-e-x11.bb: add a font to the image to make xtscal work (the e stuff has its own fonts) Aug 14 09:26:51 03mickeyl 07org.oe.dev * re04b75b6... 10/packages/mrxvt/ (files/fix-compile.patch mrxvt_0.4.1.bb): add mrxvt, a multi-tabbed rxvt Aug 14 09:26:54 03nail 07org.oe.nslu2-linux * rf93280f7... 10/packages/ssmtp/ssmtp_2.61.bb: Update ssmtp u-a priority to be less than postfix's Aug 14 09:26:58 03rwhitby 07org.oe.nslu2-linux * r90995c6a... 10/files/device_table-unslung.txt: Expanded Unslung device table to match Linksys firmware 2.3R63 Aug 14 09:27:02 03nail 07org.oe.nslu2-linux * r9736854d... 10/ (2 files in 2 dirs): add joe to feed Aug 14 09:27:06 03jbowler 07org.oe.nslu2-linux * r7305e533... 10/conf/distro/ucslugc-packages.conf: Add lrzsz Aug 14 09:27:10 03rwhitby 07org.oe.nslu2-linux * r8dcbe859... 10/conf/machine/nslu2.conf: Removed the unused variables from nslu2.conf Aug 14 09:27:14 03rwhitby 07org.oe.nslu2-linux * r0bfafb15... 10/ (6 files in 2 dirs): Split nslu2_flashimg.bbclass into nslu2-jffs2-image.bbclass and nslu2-ramdisk-image.bbclass Aug 14 09:27:18 03freyther 07org.oe.dev * r573b5d78... 10/packages/qpe-gaim/ (qpe-libgaim_1.4.0.bb qpe-libgaim_1.5.0.bb): Aug 14 09:27:18 oe/packages/qpe-gaim/qpe-libgaim: Aug 14 09:27:18 -Update qpe-libgaim from version 1.4.0 to 1.5.0 Aug 14 09:27:18 This release contained no security updates Aug 14 09:27:21 03mickeyl 07org.oe.dev * r52518f68... 10/packages/taglib/ (taglib_1.4.bb taglibc_1.4.bb): Aug 14 09:27:22 add taglib, a library for reading and editing the meta-data of several popular audio formats Aug 14 09:27:22 don't be surprised, I had to ditch the original buildsystem and use qmake - it may sound Aug 14 09:27:23 a bit perverted, but that way I could get it to work :) Aug 14 09:27:27 03rw 07org.oe.dev * rdf2cdc81... 10/packages/gstreamer/ (8 files in 5 dirs): gstreamer: bump to gstreamer-0.8.10. apply patch submitted in gnome bug #313322. move patches around a bit. Aug 14 09:27:32 03mickeyl 07org.oe.dev * re3acd9db... 10/packages/libpng/ (6 files): Aug 14 09:27:32 libpng also uses the binconfig scheme Aug 14 09:27:34 drop some older libpng versions Aug 14 09:27:36 03rw 07org.oe.dev * rea8a783e... 10/packages/gstreamer/ (3 files in 2 dirs): gst-plugins: bump to gst-plugins-0.8.8. update try-esd.patch to work with current code. Aug 14 09:27:40 03mickeyl 07org.oe.dev * r6039c402... 10/packages/e17/eclair_20050810.bb: add eclair, the E17 multi media player based on the emotion engine (based on libxine) Aug 14 09:27:44 03mickeyl 07org.oe.dev * r45a8c36f... 10/packages/e17/ (4 files): Aug 14 09:27:44 E17 files Aug 14 09:27:46 - remove unnecessary explicit binconfig settings Aug 14 09:27:48 - use ${sysconfdir} and ${sbindir} instead of hardcoding paths Aug 14 09:27:50 03rpurdie 07org.oe.dev * rff70c4ef... 10/packages/linux/linux-openzaurus_2.6.13-rc5-mm1.bb: linux-oz-2.6: Cleanup spitz patches - implement touchscreen nicely, merge code back into corgi_lcd, add backlight. Backlight needs investigation as only partially operational. Aug 14 09:27:54 03rpurdie 07org.oe.dev * r70d32a50... 10/packages/linux/linux-openzaurus-2.6.13-rc5-mm1/defconfig-spitz: linux-oz-2.6: Update Spitz defconfig Aug 14 09:27:57 03jbowler 07org.oe.nslu2-linux * r154b1829... 10/packages/uclibc/ (5 files in 3 dirs): Aug 14 09:27:59 Fix uclibc thumb support - with these patches and by disabling the ARM Aug 14 09:28:01 specific string asm in uclibc it is possible to run at least some thumb Aug 14 09:28:03 executables on ucslugc. The changes can be incorporated into other Aug 14 09:28:05 distros by setting thumb-interwork in OVERRIDES - the changes do change Aug 14 09:28:07 the way ARM code executes (necessarily). (Changes have also been submitted Aug 14 09:28:09 as bug 385 to uclibc.org). Aug 14 09:28:11 03coredump 07org.oe.dev * r3d6e44c7... 10/packages/meta/meta-e-x11.bb: e-image: Add missing dependency to 'bash' to meta-e-x11 (aka task-enlightenment-x11) Aug 14 09:28:14 03nail 07org.oe.nslu2-linux * rd4efe752... 10/packages/meta/openslug-packages.bb: Add usbutils Aug 14 09:28:16 03coredump 07org.oe.dev * r0736c0c5... 10/packages/altboot/ (5 files in 5 dirs): altboot: Fix fsck on loop-images, unb0rking SD, CF and NFS booting of loop-images. Aug 14 09:28:19 03nail 07org.oe.nslu2-linux * rf059cbad... 10/packages/zd1211/zd1211-20050315/Makefile.patch: Fixup patch, minor typo Aug 14 09:28:22 03rwhitby 07org.oe.nslu2-linux * rf27802cb... 10/ (16 files in 5 dirs): Added unslung-binary-kernel.conf distro for NSLU2 Linksys V2.3R63 firmware with NTFS read/write support. Aug 14 09:28:25 03nail 07org.oe.nslu2-linux * r878c33f6... 10/conf/distro/openslug-packages.conf: Add usbutils and zd1211 Aug 14 09:28:28 03freyther 07org.oe.dev * r0b2d2dab... 10/ (8 files in 3 dirs): (log message trimmed) Aug 14 09:28:30 Qt3 Native builds: Aug 14 09:28:32 ahem Aug 14 09:28:32 -Add a qt3e.bbclass similiar to the x11 bbclasses for Qt Aug 14 09:28:34 -Use this class in uicmoc3-native, opie-lrelease-native, Aug 14 09:28:36 opie-lupdate-native Aug 14 09:28:38 -Stage the headers of uicmoc3 and the static library to Aug 14 09:28:40 a dedicated qt3e staging directory Aug 14 09:28:44 03koen 07org.oe.dev * r6b49a965... 10/packages/gnet/gnet_2.0.7.bb: Add gnet 2.0.7 Aug 14 09:28:46 03nail 07org.oe.nslu2-linux * r407eb029... 10/packages/openslug-init/openslug-init-0.10/initscripts/zleds: Update so that the disk leds aren't switched on/off Aug 14 09:28:49 03rwhitby 07org.oe.nslu2-linux * r52d3a2c8... 10/conf/distro/unslung-binary-kernel-packages.conf: Added unslung-binary-kernel-packages.conf Aug 14 09:28:52 03rwhitby 07org.oe.nslu2-linux * r28750942... 10/ (7 files in 4 dirs): Added the unslung OVERRIDE for unslung-binary-kernel distro Aug 14 09:28:55 03rwhitby 07org.oe.nslu2-linux * r3b73d893... 10/packages/meta/unslung-binary-kernel-packages.bb: Added unslung-binary-kernel-packages.bb Aug 14 09:28:58 03nail 07org.oe.nslu2-linux * r703248fa... 10/packages/openntpd/openntpd_3.7p1.bb: Use sysconfdir/localstatedir instead of hardcoded paths. Create homedir in postinst Aug 14 09:29:01 03nail 07org.oe.nslu2-linux * rdda7bbb0... 10/packages/openntpd/openntpd_3.7p1.bb: Check if the dir exists first ;-) Aug 14 09:29:04 03nail 07org.oe.nslu2-linux * rc344fe82... 10/packages/openntpd/openntpd_3.7p1.bb: Fixup typo Aug 14 09:29:41 Is it perhaps a bad idea to have the cache on an NFS mounted partition? Aug 14 09:29:49 CIA-4: use #flood ;) Aug 14 09:29:57 Could that have caused the problem? Aug 14 09:30:07 Laibsch: it shouldn't Aug 14 09:30:19 OK Aug 14 09:32:38 mickeyl: could you rm -rf tmp and try to build taglic and entrance? Aug 14 09:32:57 koen: doing that atm. Aug 14 09:33:30 I'm still unable to build e-wm and entrance from scratch Aug 14 09:33:34 taglib fix coming up Aug 14 09:33:47 after ~60 revisions :) Aug 14 09:33:47 koen: did you try the edb fix? Aug 14 09:33:51 * mickeyl taps foot Aug 14 09:34:03 CoreDump|home: still no joy after chmodding /var/tmp Aug 14 09:34:07 :( Aug 14 09:36:26 I've been able to build the following the past week: http://ewi546.ewi.utwente.nl/drupal/?q=ipkgfind&action=section§ion=e Aug 14 09:37:53 Is bitbake no longer at svn://svn.berlios.de/bitbake/trunk/bitbake? I just tried to update my copy since I think that might have created the problem. Aug 14 09:38:28 The error I got from "svn up svn://svn.berlios.de/bitbake/trunk/bitbake" was "svn: 'svn://svn.berlios.de/bitbake/trunk' is not a working copy" Aug 14 09:39:06 Laibsch: up != co Aug 14 09:39:53 I thought I should up because I already have a working version of bitbake. Should I just use co nonetheless for updating as well? Aug 14 09:40:03 is there a fsck.jffs2 anywhere in OE? Aug 14 09:40:17 slisk: not even outside of OE Aug 14 09:40:25 Laibsch: if you have a wc do svn up Aug 14 09:40:31 Laibsch: if you need a check out do svn up Aug 14 09:40:41 I see. Aug 14 09:40:48 so if i ever want to fsck my root fs, i need to use another fs? Aug 14 09:41:32 is jffs2 so structured that fsck isn't necessary? *snickers at the absurdity of it* Aug 14 09:41:34 slisk: jffs2 is a journaled flash file system Aug 14 09:41:48 slisk: what do you want to corrupt? the inode list? Aug 14 09:41:55 slisk: the directory tables? Aug 14 09:42:02 slisk: go and read the jffs2 design paper Aug 14 09:42:36 is it at all possible to corrupt jffs2 to the point where it loses data? Aug 14 09:42:44 03mickeyl 07org.oe.dev * ra8db6e9f... 10/packages/ (56 files in 11 dirs): Aug 14 09:42:44 xine and dependencies: Aug 14 09:42:44 - rename to make more consistent with other -x11 and -fb packages Aug 14 09:42:44 - include libmodplug pluging Aug 14 09:42:44 - make libxine-x11 depend on x11 Aug 14 09:42:45 - reorder metadata Aug 14 09:42:47 - remove missing-in-action MAINTAINER Aug 14 09:42:49 03mickeyl 07org.oe.dev * rc30e01c5... 10/packages/efl/emotion_0.0.1.004.bb: emotion depends on virtual/libxine, not libxine Aug 14 09:42:52 03mickeyl 07org.oe.dev * rc072c35b... 10/packages/e17/eclair_20050810.bb: eclair is no longer BROKEN Aug 14 09:42:55 03mickeyl 07org.oe.dev * rd83e7735... 10/packages/taglib/taglib_1.4.bb: taglib: fix SOname and link against stdlibc++ Aug 14 09:42:57 zecke: fsck.jffs2 does exist, it mostly is a link to /bin/true :) Aug 14 09:42:59 ok i take that back of course it is Aug 14 09:43:22 is it at all possible to corrupt jffs2 with an abrupt power failure? Aug 14 09:43:23 slisk: Jffs2 has no inode list, so it can not be corrupt Aug 14 09:43:46 slisk: sure it is corrupted Aug 14 09:43:54 ... Aug 14 09:43:57 mickeyl: first draft: http://dominion.kabel.utwente.nl/koen/pda/files/OOOO-200508.txt Aug 14 09:44:24 excellent Aug 14 09:44:31 i'd like to release it mid next week Aug 14 09:44:33 so if the power cuts out at the wrong moment, i can lose my whole fs? Aug 14 09:44:43 and there'd be no fsck to recover? Aug 14 09:44:49 no Aug 14 09:45:11 jffs2 is fully journalled. if the power fails during a write, you will lose the contents of that write transaction, but the filesystem will be left in a consistent state. Aug 14 09:46:29 on a related note, is reiserfs fully journalled? i notice it has a fsck util. why would it need one? wouldn't jffs2 need one for the same reason? Aug 14 09:46:48 slisk: read the jffs2 design please Aug 14 09:46:52 where? Aug 14 09:46:54 slisk: you waste our time otherwise Aug 14 09:47:12 slisk: the internet? Aug 14 09:47:36 the internet is a big place... Aug 14 09:47:52 google is your friend Aug 14 09:48:04 ~hug google Aug 14 09:48:04 * ibot hugs google Aug 14 09:48:27 slisk: I've found it using jffs2 and design as keywords Aug 14 09:48:51 nm found the homepage Aug 14 09:49:04 slisk: but what corrupts on a normal filesystem? Aug 14 09:49:23 slisk: Free Block List, Inode Table, Directory Entry List? Aug 14 09:49:49 would this be the document you're talkign about? http://sources.redhat.com/jffs2/jffs2-html/ Aug 14 09:49:55 slisk: jffs2 does not have most of them Aug 14 09:50:00 slisk: so they can not corrupt Aug 14 09:50:01 i won't pretend to know every which way soemthing can go wrong Aug 14 09:50:03 slisk: right Aug 14 09:50:33 slisk: just look at what fsck.ext2 normally scans ;) Aug 14 09:50:52 slisk: but there is one way you can annoy jffs2 badly Aug 14 09:50:58 fill it up Aug 14 09:51:03 slisk: change your erasesize and flash partition size Aug 14 09:51:11 slisk: from boot to boot Aug 14 09:51:27 Filling it up helps too Aug 14 09:51:39 Once it's filled, you can't remove stuff either Aug 14 09:52:04 NAiL: that as well Aug 14 09:52:12 NAiL: that as well ;) Aug 14 09:53:01 absolutely fill it up, or is there a minimum amount of space to respect? Aug 14 09:53:19 * slisk isn't finished read design doc Aug 14 09:54:04 slisk: most filesystems have important at fixed blocks Aug 14 09:54:13 I'm going through the OE getting started docs and trying to install a newer version of monotone than is available from debian stable (sarge) ... I've downloaded 0.22 deb but it requires libgcc1 >= 1:4.0.0-7 and sarge comes with 1:3.4.3-13 ... anyone out there that's installed monotone on debian sarge? Aug 14 09:54:14 slisk: on a flash system you do not want that Aug 14 09:54:35 zecke: how full is too full? Aug 14 09:54:36 (I can't use the debian default version of monotone ...it's version .18) Aug 14 09:54:37 kados: sorry no Aug 14 09:54:44 slisk: 10kb left free Aug 14 09:55:01 kados: I'm scared you will either need to use a static version of monotone Aug 14 09:55:06 kados: or build a package yourself Aug 14 09:55:17 bummer Aug 14 09:55:38 kados: don't use a distri with ancient software packages Aug 14 09:55:38 well ... I have no idea how to build a package ... how can I use a static version of monotone? Aug 14 09:55:45 zecke: :-) Aug 14 09:55:51 zecke: what distro do you use? Aug 14 09:56:06 kados: Ubuntu Breezy Aug 14 09:56:20 kados: and FreeBSD Aug 14 09:56:36 I've got a FC3 box I could use Aug 14 09:56:43 kados: http://www.venge.net/monotone/downloads/monotone_0.22-0.1_i386.deb Aug 14 09:56:46 maybe that works Aug 14 09:56:48 * koen uses OE on OpenSlug too Aug 14 09:57:02 kados: or did you try that? Aug 14 09:57:06 zecke: that's the package I tried ;-) Aug 14 09:57:19 zecke: it needs a different version of libgcc Aug 14 09:57:40 kados: so that package is for unstable Aug 14 09:57:40 zecke: which I'm sure will need a different version of x etc. Aug 14 09:57:46 right Aug 14 09:57:50 I could do apt-pinning Aug 14 09:58:03 kados: that leaves building your own package as option Aug 14 09:58:05 but I'm afraid that will really confuse things since it's gcc I'd be apt-pinning Aug 14 09:58:06 kados: and hose your system Aug 14 09:58:11 right Aug 14 09:58:18 kados: the contained debian dir should work for stable Aug 14 09:58:26 kados: unstable is past the gcc 4 transition Aug 14 09:58:56 kados: you can't use C++ apps from unstable on sarge Aug 14 09:59:12 yea ... so I'd better switch machines then Aug 14 09:59:21 cause I need debian stable on this particular machine Aug 14 09:59:30 how's FC3 work with OE? Aug 14 09:59:40 kados: or simply rebuild the package Aug 14 09:59:47 reenoo_: how would I do that? Aug 14 09:59:48 kados: I don't know Aug 14 10:00:40 reenoo_: you mean that I coudl rebuild the package to not need libgcc>= 1:4.0.0-7 ? Aug 14 10:00:52 kados: right Aug 14 10:01:05 reenoo_: do I need the source package for that? Aug 14 10:01:36 kados: that would be the easiest option, yes Aug 14 10:01:44 reenoo_: I see a 'dsc' file ... is that it? Aug 14 10:02:07 kados: get the monotone-0.22 tar.gz Aug 14 10:02:21 kados: .orig.tar.gz .diff.gz .dsc Aug 14 10:02:26 zecke: no Aug 14 10:02:30 kados: as I said it contains a good debian/ directory (all you probably need to build package) Aug 14 10:02:58 zecke: ahh ... sorry ... I've never done that before so I didn't know what you meant Aug 14 10:03:05 kados: no problem Aug 14 10:03:13 zecke: ah, ok. I didn't know they're maintaining that themselves Aug 14 10:03:33 I'm stopping to talk now, and let reenoo help you Aug 14 10:03:34 kados: move into the source dir and say build-package should do the trick. I just use the static version IIRC. I am on stable for the server as well. Aug 14 10:03:39 too many cooks... Aug 14 10:03:40 zecke: thanks for your help! Aug 14 10:04:52 kados: Can you get the sources from venge.net? Then you'd need to create the directory with dpkg -x beforehand. Aug 14 10:04:55 reenoo_: I've got the .dsc, .tar.gz, and the i386.changes from here: http://www.venge.net/monotone/ Aug 14 10:05:15 (as well as the .deb) Aug 14 10:05:27 so ... if I'm to build the package, how do I proceed? Aug 14 10:05:27 kados: in that case 'sudo apt-get build-dep monotone' and 'fakroot dpkg-buildpackage' in the toplevel directory of the extracted tarball Aug 14 10:05:50 ok Aug 14 10:05:52 s/fakroot/fakeroot/ Aug 14 10:07:17 Laibsch: indeed. heh Aug 14 10:07:23 reenoo_: fakeroot command not found Aug 14 10:07:31 reenoo_: I'll try to locate it with apt-cache Aug 14 10:07:56 ok ... done Aug 14 10:08:23 dpkg-checkbuilddeps: Unmet build dependencies: texinfo Aug 14 10:08:24 dpkg-buildpackage: Build dependencies/conflicts unsatisfied; aborting. Aug 14 10:08:41 so I take it I need texinfo ... which must be a new dependency ? Aug 14 10:09:07 sounds like it Aug 14 10:09:32 cool ... it seems to be working Aug 14 10:11:42 very good Aug 14 10:12:52 kados: I do not understand why you don't just go for the statically compiled version. Aug 14 10:13:10 You'd be up and running by now :-) Aug 14 10:13:11 Laibsch: well ... I didn't know about it I guess Aug 14 10:13:49 Laibsch: does a statically compiled version 0.22 exist? Aug 14 10:13:50 Laibsch: it's not on the GettingStarted page on the wiki: http://oe.handhelds.org/cgi-bin/moin.cgi/GettingStarted#head-9730d4326a7b8a0d13c2be3b7fe29b80ad152bf6 Aug 14 10:13:56 oops ... lorry Aug 14 10:14:00 sorry even ;-) Aug 14 10:14:56 03coredump 07org.oe.dev * rff06b253... 10/packages/e17/ (5 files in 4 dirs): entrance: Usie gpe-dm to launch X server. Install Xinit.d/* and Sessions/Enlightenment. Update entrance configuration. Aug 14 10:18:19 * kados taps foot while dpkg-buildpackage works Aug 14 10:18:30 :) Aug 14 10:18:46 kados: looks like one of the many problems is solved Aug 14 10:19:03 I just got a refurb sl-6000 and I'm eager to play around in oe Aug 14 10:19:09 see if I can build a ROM Aug 14 10:19:45 I've also got a couple of packages I'd like to maintain ... the CF-XGA card being one ... it's a cf card for doing vga out Aug 14 10:19:48 zecke: I assume that http://www.venge.net/monotone/downloads/monotone-0.22-linux-x86.bz2 should be static? I do not remember but I think that is what I got. Aug 14 10:20:08 Laibsch: shall I add that link to the wiki with an explaination? Aug 14 10:20:13 Laibsch: wow, I was too stupid to find it Aug 14 10:20:18 hehe Aug 14 10:20:55 kados: iirc kdrive supports CF-vga cards out of the box Aug 14 10:21:18 koen: neat ... thing is, I've already got the CF-XGA card ;-) Aug 14 10:21:25 koen: and I've got it working on an 5500 Aug 14 10:21:35 with an older version of OZ Aug 14 10:21:51 kados: I had been thinking about that myself but then rejected the idea for the reason so as not to clutter the already overburdened page (which does not help readability). I figured the link to the monotone page was enough for anyone to figure out where the statice version is. I guess zecke proved me wrong ;-) Aug 14 10:22:04 Laibsch: :-) Aug 14 10:22:17 * Laibsch wants a CF-XGA card, too Aug 14 10:22:27 That must be awesome. Aug 14 10:22:36 kados: http://cvs.freedesktop.org/xserver/xserver/hw/kdrive/pcmcia/pcmcia.c?rev=1.9&view=markup Aug 14 10:23:00 I'm not sure if that one needs a kernelmodule Aug 14 10:23:06 koen: sweet! Aug 14 10:23:44 koen: the source I'm starting from lives at http://www.iodata.com/usa/products/CFXGA_SW.php (at the bottom of the page) Aug 14 10:24:16 koen: I've got permission from IO Data to modify and add it to Open Zaurus Aug 14 10:24:47 reenoo_: ok ... build is done ... I don't see a .deb anywhere ... is there another step? Aug 14 10:24:50 ah, that's a qt/e driver if I'm correct Aug 14 10:25:05 kados: ls .. Aug 14 10:25:10 kados: look in ../ Aug 14 10:25:42 gotcha ... well there was a .deb there already so I didn't notice ... /me checks the timestamp Aug 14 10:26:15 yep ... it's mine ;-) ... thanks for the help everyone ... now to see if it works ;-) Aug 14 10:26:38 sweet ... it installed ;-) Aug 14 10:27:24 CoreDump|bbl: what the hell? entrance woorked *just fine* on its own. Not only that, auth=2 *doesn't work*. I had it set to auth=0 for a reason. Aug 14 10:27:39 CoreDump|bbl: it also *does not need* gpe-dm. It can handle it itself. Aug 14 10:29:22 heh Aug 14 10:29:26 yipee ... it's syncing :-) Aug 14 10:29:50 JustinP: does engage build for you? Aug 14 10:29:55 thanks again ... I'm sure I'll be back soon ;-) Aug 14 10:29:59 mickeyl: no, it failed one sec, I have a fix Aug 14 10:30:06 JustinP: ok, cool Aug 14 10:30:12 missing e.g Aug 14 10:30:12 * kados heads out for coffee while monotone syncs Aug 14 10:30:13 e/h Aug 14 10:30:17 e.h Aug 14 10:30:32 JustinP: while you're at it, remove the bogus -I${includedir} from src/Makefile.in Aug 14 10:30:58 for which? Aug 14 10:31:03 engage/src/module/Makefile.am Aug 14 10:31:12 line 13 or so Aug 14 10:31:22 -I$(includedir) makes it include local headers Aug 14 10:31:28 because includedir is /usr/include Aug 14 10:31:51 I'm also very tempted to disapprove coredump's entrance commit. True I was missing a few files but the changes I made were intentional and gpe-dm is absolutely not needed Aug 14 10:32:04 please discuss this with him before disapproving Aug 14 10:32:46 besides looking odd to depend on a gpe package, gpe-dm is not harmful :) Aug 14 10:32:57 who is going to call xtscal btw. if necessary ? Aug 14 10:33:05 entrance won't work as it is now...unless some images have shadow passwords and others don't. auth=2 *will* break it on at least my machine. Aug 14 10:33:44 and gpe-dm is, again, not needed to start up X. I added all of those patched ot make entranced workk correctly (which makes it not need gpe-dm). Aug 14 10:33:59 entrance still needs some love, though, as it has some quitting problems Aug 14 10:34:05 gpe-dm does *not* fix that either Aug 14 10:34:23 ok, even better Aug 14 10:34:35 who is calling xtscal btw.? Aug 14 10:34:47 the Xsession scripts Aug 14 10:34:56 +++ packages/e17/entrance/Xinit.d/50-calibrate 8e0e471f05ea57e861d91c2f0d92b02a5902f48d Aug 14 10:34:57 ah. Aug 14 10:35:00 @@ -0,0 +1,7 @@ Aug 14 10:35:02 +#!/bin/sh Aug 14 10:35:05 + Aug 14 10:35:07 +if ! test -e /etc/pointercal Aug 14 10:35:10 +then Aug 14 10:35:12 + xtscal Aug 14 10:35:23 excellent Aug 14 10:35:35 which is called by fricking gpe-dm....::sigh::. Of course entrance doesn't have this particular thing built in but it's not that hard to add it. Aug 14 10:35:48 This is all why people shouldn't be trying to make a fscking e-image right now Aug 14 10:35:56 * chouimat is away: Lunch: Ramen noodles and sardines can :) Aug 14 10:36:05 see it positive, all are excited about that new stuff Aug 14 10:36:09 gpe-dm is just a couple of scripts Aug 14 10:36:32 perhaps it would be appropriate to rename it then, now that it is more versatile? Aug 14 10:37:02 expect for zealotry reasons, why? Aug 14 10:37:10 for clarity Aug 14 10:37:24 if it doesn't contain anything gpe specific, why having a gpe prefix? Aug 14 10:37:32 *shrug* Aug 14 10:37:36 heh Aug 14 10:37:46 * koen hints at linux-openzaurus.bb Aug 14 10:38:06 mickeyl: because "dm" seemed a bit too short Aug 14 10:38:12 heh Aug 14 10:38:14 I'm glad people are interested in all this, but aren't *I* supposed to be the maintainer of thes epackages? Aug 14 10:38:18 s/dm.bb/desktop-manager.bb Aug 14 10:38:39 isn't it display-manager? Aug 14 10:39:18 03mickeyl 07org.oe.dev * rfadeaf2a... 10/packages/e17/ (eclair_20050810.bb entice_20050810.bb examine_20050810.bb): Aug 14 10:39:18 add some more dependencies to ecore-x11 and evas-x11 Aug 14 10:39:18 (I looked at the actual sources, not at configure.in) Aug 14 10:39:42 ah right Aug 14 10:41:24 yeah, it stood for "display manager", although that description only applies in a very loose kind of way here. Aug 14 10:45:41 * reenoo_ wonders why module-init-tools are built for h3600 Aug 14 10:46:53 oh. and the infamous "Couldn't find shared library provider for libgcc_s.so.1" is back Aug 14 10:47:04 was it ever gone? Aug 14 10:47:11 no, it's always been with us Aug 14 10:47:27 I've grown quite fond of it over the months. Aug 14 10:47:29 pb_: if you fix you might get an OE account ;) Aug 14 10:47:32 it was gone for a while here Aug 14 10:47:40 if you fix it even.. Aug 14 10:47:43 where is my mind Aug 14 10:48:04 e17 isnt a display manager ... Aug 14 10:48:11 just so you know Aug 14 10:48:19 thanks for the bulletin. Aug 14 10:48:25 entranced is, however Aug 14 10:48:26 fwiw, glibc isn't a display manager either. Aug 14 10:48:43 apperently its a display shell ... somewhere between a enviroment and a manager Aug 14 10:49:12 zecke: ah, a tempting offer indeed Aug 14 10:49:35 emte: you must mean "window manager", not "display manager" thenn Aug 14 10:49:53 all three then Aug 14 10:50:08 e is a window manager, not a display manager Aug 14 10:50:08 between ecore and evas they manage the display Aug 14 10:50:21 * koen ponders about using his pda as XMDCP server Aug 14 10:50:21 entranced is the display manager, as is gpe-dm, xdm, kdm, gdm, etc Aug 14 10:50:59 enterance is just a pretty intro, not much more Aug 14 10:51:18 entranceD! is a display manager Aug 14 10:51:36 enntrance is for is login Aug 14 10:51:54 funky names all these e things Aug 14 10:51:55 have Aug 14 10:51:58 if you choose to use it, yes Aug 14 10:53:10 but i guess it comes down to what you expect a display manager does i guess Aug 14 10:54:00 reminds me, zecke i just ran into your emacs-gettext issue as well Aug 14 10:54:55 i think they must have changed soemthing in emacs 0.22 Aug 14 10:55:27 because it launches the emacs shell but doesnt compile the -batch Aug 14 10:55:38 so it hangs with the shell open Aug 14 10:56:13 03justinp 07org.oe.dev * r197281d5... 10/packages/efl/e_20050810.bb: Stage e.h as well Aug 14 10:57:32 my admin has been buzy with a contract to make changes to the webserver for monotone-hg Aug 14 10:58:46 http://ibot.rikers.org/oe/ is 404ing ... is there an archive of this channel anywhere else? Aug 14 10:58:55 ~logs Aug 14 10:58:56 apt/ibot/jbot/purl all log to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz, or updated "nightly" Aug 14 11:00:54 hmmm ... maybe the wiki should be updated under contacts to include a link with html encoded channelnames ;-) Aug 14 11:01:29 actually ... it's not the wiki but this page: http://oe.handhelds.org/contact.php Aug 14 11:04:43 * kados gets to know ibot Aug 14 11:08:04 * chouimat is back. Aug 14 11:09:05 holy shit jesus calm down man Aug 14 11:09:22 CoreDump|home: hey Aug 14 11:09:29 CoreDump|home: whats up? Aug 14 11:09:30 hi zecke Aug 14 11:09:42 i was about to ask JustinP Aug 14 11:10:07 emte: any fixes or is emacs broken? Aug 14 11:10:12 hey zecke you have that url for your patch still handy? Aug 14 11:10:27 fixes? downgrade your system emacs? Aug 14 11:10:53 emte: sadly not, and my office machine is down Aug 14 11:10:54 they made emacs changed in 0.18 Aug 14 11:10:59 changes* Aug 14 11:11:09 and it seems they must have made more Aug 14 11:11:12 emte: it was brute force to remove emacs support Aug 14 11:11:17 koen: I'm taking a look at that freedesktop.org pcmcia.c code you linked to me ... it also supports the Colorgraphic card which I also have ;-) Aug 14 11:11:20 i guessing you must have emacs 0.22 as well Aug 14 11:11:35 which means that everyone will start seeing the issue Aug 14 11:11:53 when they upgrade Aug 14 11:12:13 but yeah forcibly removing emacs seems the solution for the moment Aug 14 11:12:44 will it break other things tho Aug 14 11:13:06 lisp is renowned for its parsing speed ... so it might slow the builds down Aug 14 11:13:27 so is it the goal of oe to allow standard distros to run on embedded platforms like the zaurus? Aug 14 11:13:52 not really Aug 14 11:13:56 or is it rather to run distros that have support for standard distros packages Aug 14 11:14:03 :-) Aug 14 11:14:09 more to provice a muti-device build system Aug 14 11:14:12 JustinP: I give you the auth=2 thingy, was playing around with the setting as 0 doesn't work for me, but the rest? entrance did *not* work on its own. It *requires* the Xserver file from gpe-dm to be of any use. Where's the point in coding the functions of gpe-dm into entrance when the existing stuff works fine? Just because it's called "gpe"-dm? Aug 14 11:14:19 provide* Aug 14 11:14:49 gpe is the synonym for X anyway ;) Aug 14 11:14:58 lol Aug 14 11:15:02 zecke: and fluff, bloat and !polish Aug 14 11:15:14 emte: so does that mean a build system that can be used to build packages/distros for multiple cpu architectures or is it more than that? Aug 14 11:15:32 emte: I'm very new to low-level programming Aug 14 11:15:35 kados, more, but that is part of it Aug 14 11:16:14 it brings together all the tools you need to support a device Aug 14 11:16:43 "distros" just tend to be a offshoot benificiary Aug 14 11:17:14 all the core work is done in OE, they take it and package it in a distro Aug 14 11:17:19 hmmm ... so would the standard distro folks do well to adopt a similar approach (maybe even using oe?) Aug 14 11:17:45 to allow them to build using metadata that's useful for many many architectures and devices, etc.? Aug 14 11:17:52 there are some x86 projects here, yes Aug 14 11:17:57 cool Aug 14 11:25:45 so actually, openembedded's going to help me with more than just my zaurus ... I can use it to build custom distros for my Soekris boxes too Aug 14 11:25:58 is there a list of supported architectures? Aug 14 11:26:07 fully or partial? Aug 14 11:26:19 look in the machines dir for an idea Aug 14 11:26:22 k Aug 14 11:26:31 (still syncing here ;-) Aug 14 11:26:47 monotone: 26.5M | 688.9k | 5778 | 1397 | 224 Aug 14 11:27:00 * kados reads the bitbake manual Aug 14 11:28:18 kados: use the db snapshot Aug 14 11:28:51 koen: 's ok ... this will give me time to grok what I'm actually going to do with it ;-) Aug 14 11:29:02 koen: i still don't really grok _what_ bitbake is Aug 14 11:29:46 bitbake is a task executor Aug 14 11:29:56 the OE metadate makes it able to build stuff Aug 14 11:29:59 like make? Aug 14 11:30:07 sort of like make Aug 14 11:30:38 so I write some source and bb builds it for devices as specified in the conf file? Aug 14 11:31:14 that's the idea Aug 14 11:31:25 sweet Aug 14 11:31:35 yea ... this rocks! Aug 14 11:35:26 so who works on the .conf file? Aug 14 11:36:02 all of us Aug 14 11:39:03 wb _alwin_ Aug 14 11:39:12 <_alwin_> re Aug 14 11:48:47 * france is back (gone 27:56:54) Aug 14 11:58:10 03mickeyl 07org.oe.dev * ra3dac38e... 10/packages/e17/ (engage/no-local-includes.patch engage_20050810.bb): engage: fix dependencies and add no-local-includes patch Aug 14 12:12:57 03coredump 07org.oe.dev * r0a13f564... 10/packages/e17/ (entrance/config-db.patch entrance_20050810.bb): entrance: revert auth=2 to auth=0 Aug 14 12:34:23 03nail 07org.oe.nslu2-linux * ra609ff1e... 10/ (2 files in 2 dirs): Added wireless-tools to feed Aug 14 12:45:27 03nail 07org.oe.nslu2-linux * r933c360c... 10/packages/meta/openslug-packages.bb: Added zd1211 driver to feed Aug 14 13:03:09 03koen 07org.oe.dev * rca933b7b... 10/ (7 files in 5 dirs): Add asterisk 1.0.9, delete 1.0.9 Aug 14 13:03:26 *sigh* Aug 14 13:03:29 ~lart me Aug 14 13:03:30 * ibot decapitates koen conan the destroyer style Aug 14 13:03:35 .. Aug 14 13:03:40 add 1.0.9, delete 1.0.7 Aug 14 13:03:54 * RP returns Aug 14 13:04:00 hey RP Aug 14 13:04:25 I have the backlight working properly now :) Aug 14 13:04:51 the spitz backlight? Aug 14 13:04:57 yes Aug 14 13:05:17 There was a subtle problem with the chip selects on the ssp interface Aug 14 13:05:44 I still can't find a way to avoid big banging though which is most annoying Aug 14 13:08:30 03jbowler 07org.oe.dev * r5da8bfb1... 10/conf/distro/ucslugc.conf: Aug 14 13:08:31 Remove the step to create a UcSlugC binary image (for upslug), thereby Aug 14 13:08:31 removing the requirement for linksys firmware - this is a partial fix, Aug 14 13:08:31 at present linksys firmware is still required to build an image for Aug 14 13:08:31 upslug, so to use UcSlugC after this it will be necessary, for the Aug 14 13:08:31 moment, to have an existing OpenSlug or UcSlugC system with reflash. Aug 14 13:37:26 bitbake fails to build for me since the last couple of days. Console output is at http://rafb.net/paste/results/JL5VRk60.html. Aug 14 13:39:44 Laibsch: are both bitbake and OE up to date? Aug 14 13:39:56 yes, I updated them today. Aug 14 13:40:07 so ... looks like the monotone sync finally finished ... but I'm seeing some warnings: Aug 14 13:40:08 For OE I always had a cron job. Aug 14 13:40:10 monotone: warning: discarding revision data packet 32a48e4b95f3a4d8667381042fae7273a6995559 with unmet dependencies Aug 14 13:40:14 lots of these ... is that a problem? Aug 14 13:40:25 kados: no, that's harmless Aug 14 13:40:26 I only implemented the cronjob for bitbake today and ran a manual update. Aug 14 13:40:32 koen: cool ... thanks Aug 14 13:41:18 koen: What I thought was strange is that there are never any revs written for the monotone update of OE. Aug 14 13:41:52 when I check out a local copy of the OE tree ... do I want to be in /stuff/build ? Aug 14 13:43:43 or is that what the /stuff/openembedded dir is for ? Aug 14 13:43:49 * kados is confused Aug 14 13:45:56 the openembedded tree goes into openembedded/ Aug 14 13:46:20 Laibsch, then your doing it wrong Aug 14 13:46:25 koen: i ran this: monotone --db=/stuff/oe.db pull monotone.vanille.de "org.openembedded.{dev,nslu2-linux,dreambox}" Aug 14 13:46:29 what command are you using for updating? Aug 14 13:46:44 koen: now ... from /stuff I run monotone --db=/stuff/oe.db checkout --branch=org.openembedded.dev Aug 14 13:46:53 koen: and it puts it in openembedded? Aug 14 13:46:58 emte: I use a shell script I coded to get everything going. I'll paste it. Aug 14 13:47:04 it will put it in org.openembedded.dev/ Aug 14 13:47:25 ok ... so kinda like cvs co Aug 14 13:47:35 yeah Aug 14 13:47:58 so I can update the tree with monotone --db=/stuff/oe.db pull monotone.vanille.de org.openembedded.dev ; cd /stuff/org.openembedded.dev; monotone update Aug 14 13:48:04 but that also updates the branch right? Aug 14 13:48:10 yes Aug 14 13:48:18 seems counterintuitive Aug 14 13:48:23 kados: if you are used to cvs take a look at http://venge.net/monotone/docs/CVS-Phrasebook.html#CVS-Phrasebook Aug 14 13:48:30 koen: sweet ... thanks Aug 14 13:48:48 yeah you would think monotone update would be smart enough to read the db file ... Aug 14 13:49:15 well ... I was just thinking it would be nice to be able to update just the tree ... and leave the branch alone Aug 14 13:49:21 the database is the 'server' Aug 14 13:49:22 emte: http://rafb.net/paste/results/TTwGXw83.html Aug 14 13:49:31 That is what I use for preparing OE. Aug 14 13:49:54 Laibsch, yeah thats wrong Aug 14 13:50:01 monotone --db=/stuff/oe.db pull monotone.vanille.de org.openembedded.dev ; cd /stuff/org.openembedded.dev; monotone update Aug 14 13:50:05 is the u[pdate command Aug 14 13:50:22 all three parts Aug 14 13:50:22 kados: monotone update will update the working copy, which is by default the org.openembedded.dev branch Aug 14 13:51:53 emte: If you look at it, I have all three parts. Just slightly different. Aug 14 13:52:32 your not pulling into your repo .. Aug 14 13:52:49 you have: monotone --db=/stuff/oe.db pull monotone.vanille.de Aug 14 13:52:57 not: monotone --db=/stuff/oe.db pull monotone.vanille.de org.openembedded.dev Aug 14 13:53:12 unless i am missing that part Aug 14 13:54:45 i updated probably 4 hours ago and tehre are 31 new revs since then Aug 14 13:56:16 OK, I'll put in the org.openembedded.dev and try again. I'll expect a long update for monotone then since apparently my repo has not changed for ages. Aug 14 13:58:00 Well, in the daily cronjob there is '~/bin/monotone --db=./oe.db pull monotone.vanille.de "org.openembedded.{dev,dreambox,nslu2-linux}"' so it should not be more than ~18 hours old. Aug 14 13:58:41 I'll do a manual update nonetheless. Aug 14 13:59:02 Laibsch: don't forget a "monotone update" in the working copy Aug 14 13:59:05 i hope thats all thats missing from your script Aug 14 13:59:19 everything else looks fine Aug 14 14:01:32 so what happens when oe.db gets over 2 gig? ;-) Aug 14 14:02:02 over 2gb? Aug 14 14:02:05 if you have a sane FS nothing :) Aug 14 14:02:11 you shouldnt even be near that Aug 14 14:02:30 48832 -rw-r--r-- 1 biohazard biohazard 49999872 2005-08-14 12:01 oe.db Aug 14 14:02:32 50k Aug 14 14:02:42 yea ... mine's at 31k Aug 14 14:03:03 s/k/M/ Aug 14 14:03:11 yea ... me too ;-) Aug 14 14:03:39 * kados reads the monotone docs to try to understand distributed revision control Aug 14 14:03:48 good luck Aug 14 14:04:02 heh Aug 14 14:04:22 reading the description it seems like they threw it togther and are hoping they miss all the patents Aug 14 14:04:33 hehe Aug 14 14:04:46 well ignoring patents is in style these days ;-) Aug 14 14:04:56 it's hard to avoid in fact Aug 14 14:05:00 ~1500 revs in. Now going to the writing part. Aug 14 14:05:01 only until it catches up to you Aug 14 14:05:08 Laibsch, ouch Aug 14 14:05:17 hope its bed time for you Aug 14 14:05:22 emte: I guess it'll take a while, hm? Aug 14 14:05:31 its ~2-5min per rev Aug 14 14:05:42 Laibsch: you could start with an upto date db Aug 14 14:05:43 and apperently it does about 1/2 of them Aug 14 14:05:54 so what's exactly happening with revs? Aug 14 14:05:56 OE just told me it might take while ;-) Aug 14 14:06:01 so your looking at 1400 minutes min Aug 14 14:06:11 Laibsch: wget http://ewi546.ewi.utwente.nl/tmp/OE.db.bz2 IIRC Aug 14 14:06:12 when it's syncing ... is it just doing difs on each line of oe.db? Aug 14 14:06:43 zecke: That is where I started from. About two weeks ago. I guess that URL gets updated frequently? Aug 14 14:06:59 i am begining to think you have to cron monotone every hour to make the updates bearable Aug 14 14:07:13 emte: why does it update so often? Aug 14 14:07:15 Laibsch: right Aug 14 14:07:23 emte: is there that much dev working going on? Aug 14 14:07:33 people make changes adn submit them Aug 14 14:07:45 kados: http://cia.navi.cx/stats/project/openembedded/ Aug 14 14:08:01 kados, over 1/2 the channel here are devs and the otehr couple hundred not here Aug 14 14:08:15 wow! Aug 14 14:08:18 I guess my cronjob should not have had "" around org.openembedded.{dev,dreambox,nslu2-linux} :-( Aug 14 14:08:20 that's an active project ;-) Aug 14 14:09:36 ho hum ... none of the images in the monotone html docs are working ... Aug 14 14:09:38 hmm i need some socks Aug 14 14:09:41 * kados checks for a pdf Aug 14 14:18:17 my monotone repo is on my router. I run OE on a different computer. Is it OK to set up a bi-hourly cron job on the router to update via monotone without risking interference if I start a bitbake compile while the update is half-way through? I hope I was successul in making clear what my question is about. Aug 14 14:18:48 unsure on that point myself Aug 14 14:19:01 Laibsch: you can do monotone pull everytime Aug 14 14:19:03 was just thinking about updating my hg idea Aug 14 14:19:08 Laibsch: just monotone update after you're done Aug 14 14:19:56 setup monotone to pull every hour and use hg to build from Aug 14 14:20:23 you could jsut setup two monotone repos tho and update your second one local Aug 14 14:20:26 zecke: What happens if I "monotone update" and start a bitbake build while the update is running? Is that safe? I'd only do the pull via cron and update only via the initialization script. Aug 14 14:20:35 so push pull sync is for remove <-> local ... whereas commit update is for local <-> working eh? Aug 14 14:20:56 Laibsch: if it is through parsing, it should only hit the cache Aug 14 14:21:03 emte: That is an idea too. If possible I prefer to keep things central, though. Aug 14 14:21:05 so monotone --db=/stuff/oe.db pull monotone.vanille.de org.openembedded.dev ; cd /stuff/org.openembedded.dev; monotone update does both Aug 14 14:22:01 Laibsch: so would you setup a pull every hour and then only update right before or after a compile? Aug 14 14:22:34 sorry ... wrong wording Aug 14 14:22:41 kados: I can do either way. Trying to decide what is safest and workable right now. Aug 14 14:23:12 Laibsch: let me know what you decide and why ;-) Aug 14 14:23:15 update every hour, bitbake pulls from the cache ... so as long as your not updating while it builds the cache you should be okay Aug 14 14:23:29 zecke: So I guess doing an automatic "monotone update" at a time when one is not 100% sure there is no compilation running/starting is not 100% safe? Aug 14 14:23:36 would be easy enough to make a lock check Aug 14 14:23:52 bitbake caches all metadae when it starts Aug 14 14:23:58 data* Aug 14 14:24:12 Laibsch: right Aug 14 14:24:22 cmon monotone Aug 14 14:24:27 * emte pushes harder Aug 14 14:24:54 emte: "so as long as your not updating while it builds the cache you should be okay"-> with a cronjob I cannot guarantee that some odd day I am awake building with bitbake at 4:00 o'clock in the morning :-( :-( Aug 14 14:25:05 just do a pidcheck for monotone before you start bitbake Aug 14 14:25:16 zecke: but doing a 'monotone pull' is always safe right? Aug 14 14:25:25 if its running sleep for 10 min and check again Aug 14 14:25:31 kados: yes Aug 14 14:25:32 kados: yes Aug 14 14:25:47 pretty simple Aug 14 14:25:49 Na, too much trouble. I'll pull hourly via cron and update via my initialization script. That is run on the faster processor anyway. Aug 14 14:26:30 I was referring to emte's statement about checking pids. Aug 14 14:27:05 not hard ... i've done it for redundancy for servers Aug 14 14:27:59 usually to keep mysql from doing soemthing stupid Aug 14 14:28:34 wow ... in just the last couple of hours there have been 963 revs (i just did a pull) Aug 14 14:28:39 altho newer versions seem quite a but more stable Aug 14 14:28:42 does that sound right? Aug 14 14:28:57 sounds like you added a new branch Aug 14 14:29:25 emte: I did monotone --db=/home/josh/zaurus/stuff/oe.db pull monotone.vanille.de org.openembedded.dev Aug 14 14:29:44 emte: or should I have done sync? Aug 14 14:29:51 then i'd guess it has been longer than a few hours or your reading certs Aug 14 14:29:55 03rpurdie 07org.oe.dev * rcd1d678b... 10/packages/linux/ (2 files in 2 dirs): linux-oz-2.6: Add wireless extentions to spitz-defconfig. Update spitzssp to fix backlight problems and remove bit banging code (yay\!). Aug 14 14:30:19 emte: well ... it's definitely only been a few hours Aug 14 14:30:26 RP: cheers! Aug 14 14:30:28 emte: since I ran monotone --db=/stuff/oe.db pull monotone.vanille.de "org.openembedded.{dev,nslu2-linux,dreambox}" Aug 14 14:31:03 emte: and checked out a local copy Aug 14 14:31:26 emte: monotone: 2.4M | 240.3k | 4185 | 963 | 18 Aug 14 14:31:35 is where I'm at now Aug 14 14:31:49 you sure your not checking out a new repo then? Aug 14 14:32:07 maybe you were in the wrong dir? Aug 14 14:32:18 emte: that is crap Aug 14 14:32:19 can I C-c safely? Aug 14 14:32:32 koen, ? Aug 14 14:32:35 "the wrong dir" doesn't matter when doing monotone pull Aug 14 14:32:41 pull == db action Aug 14 14:32:43 no? Aug 14 14:32:54 aha true Aug 14 14:33:16 963 revs is a lot for a few hours Aug 14 14:33:36 * kados is confused about what to do ;-) Aug 14 14:33:52 maybe my original sync didn't work right? Aug 14 14:34:10 i did have a lot of warnings (but I heard they were harmless) Aug 14 14:34:21 kados, you are aware you have two dif --db= ? Aug 14 14:34:29 or was the second jsut an example? Aug 14 14:34:47 no ... I wasn't aware Aug 14 14:35:03 ahh ... second was just an example ;-) Aug 14 14:35:22 only one oe.db in /home/josh/zaurus/stuff/oe.db Aug 14 14:35:22 then koen was right Aug 14 14:35:36 right that everything's ok? Aug 14 14:35:42 * emte shrugges Aug 14 14:36:12 just put a check for lsof | grep "oe.db" in your cron script Aug 14 14:36:53 it can call at to schedule execution for like an hour later Aug 14 14:37:06 what happens if I 'update' durring a pull? Aug 14 14:37:41 that's like crossing the streams, don't try it Aug 14 14:37:41 it will update Aug 14 14:37:50 or it will say 'database locked' Aug 14 14:38:04 if it's smart... Aug 14 14:38:25 and if it updates, it will check out the latest verified revision Aug 14 14:38:28 but any1 who thinks it's a good idea to have a monolithic 100mb+ configuration file isn't too smart Aug 14 14:39:21 also, if the subsequent commands are independent of the db update, then you've got too problems to worry about Aug 14 14:39:37 you need && instead of semicolons in between the commands Aug 14 14:39:47 so if 1 fails the rest doesn't execute and bring about more problems Aug 14 14:40:07 i imagine that the metadata files are opened invididual only when they're needed Aug 14 14:40:22 so if you go about building a package, then update it between steps, it could caus problems Aug 14 14:41:03 lol too instead of two Aug 14 14:41:15 how many times have i mocked others for that? Aug 14 14:41:36 gotta get off my type kill tendencies Aug 14 14:43:40 wb mithro Aug 14 14:47:36 what's the difference between: Aug 14 14:47:38 pull monotone.vanille.de "org.openembedded.{dev,nslu2-linux,dreambox}" Aug 14 14:47:51 pull monotone.vanille.de "org.openembedded.dev" Aug 14 14:48:04 the number of branches you update Aug 14 14:48:09 could that be the reason I'm getting 963 revs in a few hours? Aug 14 14:48:19 yeah Aug 14 14:48:40 I did the first one when I did my original pull Aug 14 14:48:56 and the second when I wanted to pull the latest stuff (mainly just a test) Aug 14 14:50:58 josh@sam:~/zaurus/stuff$ monotone --db=/home/josh/zaurus/stuff/oe.db list branches Aug 14 14:51:08 monotone: fatal: std::runtime_error: failure extracting schema from sqlite_master Aug 14 14:51:10 monotone: this is almost certainly a bug in monotone. Aug 14 14:51:54 * kados hopes it's not an error due to his self-made .deb package Aug 14 14:52:33 kados: read the update manual Aug 14 14:52:41 s/manual/notes Aug 14 14:53:16 http://www.venge.net/monotone/UPGRADE Aug 14 14:53:48 kados: try monotone db migrate and then try again Aug 14 14:54:04 zecke: I'm still in the middle of a pull Aug 14 14:56:10 kados: monotone is good at blaming itself, when it is not guilty at all Aug 14 14:56:56 zecke: i don't see how db migrate will solve anything ... since I created this db with version .2 Aug 14 14:57:10 zecke: unless I'm not understanding something ;-) Aug 14 14:57:13 kados: ah well Aug 14 14:57:25 kados: okay I thought you had some pre existing db Aug 14 14:57:40 03nail 07org.oe.nslu2-linux * r8cf91caa... 10/packages/sudo/ (sudo.inc sudo_1.6.8p8.bb): Fix sudoers permission Aug 14 14:57:53 zecke: I deleted it after uninstalling the debian stable monotone (version .18) Aug 14 14:58:01 zecke: and created a new one with the new .22 version Aug 14 14:58:22 zecke: i wonder if the error is related to the fact that I'm in the middle of a pull Aug 14 14:58:43 kados: I've not seen that before Aug 14 14:58:53 at this rate I'll be spending 90% of my time waiting for pull and update and 10% actually programming ;-) Aug 14 14:58:58 03koen 07org.oe.dev * ra3d67734... 10/packages/libgpewidget/libgpewidget_0.103.bb: Add libgpewidget 0.103, matches cairo 0.5 API and fixes the dateselector widget Aug 14 15:00:37 kados: hehe you might join the anti-motone movement Aug 14 15:00:43 :-) Aug 14 15:00:56 it looks nice but I'm just unfamiliar with it Aug 14 15:01:01 kados: we're open to switch for the better Aug 14 15:01:20 kados: it is slow, no doubt but it's slowness has not prevented me from doing any work Aug 14 15:01:24 we will see Aug 14 15:01:36 i am uploading the hg repo to a server now Aug 14 15:01:41 kados: turn on computer, start monotone pull, get breakfast statz hacking Aug 14 15:01:51 with my slow arsed connection its gonna be 40min Aug 14 15:01:55 I'm really unfamiliar with the options for distributed revision control ... but I know it's a real problem finding (or creating) a decent app for it Aug 14 15:02:28 bitkeeper is nice and decent Aug 14 15:02:29 well I guess I must have pulled a new branch by mistake Aug 14 15:02:37 but expensive and closed Aug 14 15:02:46 origianlly I pulled: Aug 14 15:02:47 monotone --db=/stuff/oe.db pull monotone.vanille.de Aug 14 15:02:53 then to 'update' I pulled: Aug 14 15:03:09 oops ... wrong paste above Aug 14 15:03:24 origianlly:"org.openembedded.{dev,nslu2-linux,dreambox}" Aug 14 15:03:29 i know CoreDump|home will be excited to try it Aug 14 15:03:36 now: "org.openembedded.dev" Aug 14 15:03:50 seems like the first one should include the second Aug 14 15:03:53 wish i could get ahold of blak to setup the cgi Aug 14 15:03:54 right? Aug 14 15:03:56 emte: I would clone/pull whatever it as well Aug 14 15:04:04 emte: hm? Aug 14 15:04:09 yeah Aug 14 15:04:35 well it took 30 seconds to incorporate the three days of monotone changes into hg :) Aug 14 15:04:47 what's hg? Aug 14 15:04:59 mercurial Aug 14 15:05:06 another version control system Aug 14 15:05:22 wow Aug 14 15:05:30 quite a bit more flexible for multiplatform devel Aug 14 15:05:43 and small Aug 14 15:05:48 kados: http://www.selenic.com/mercurial/ Aug 14 15:06:19 neat Aug 14 15:06:27 *yawn* Aug 14 15:06:37 mickeyl: hey Aug 14 15:06:43 rmorning Aug 14 15:06:47 mickeyl: are you okay witht he qt3e.bbclass? Aug 14 15:06:50 hey treke|home Aug 14 15:07:50 grr Aug 14 15:07:53 * pb_ stabs PCB again Aug 14 15:07:58 zecke: yeah, it looks ok. I'm just wondering why we need it. Do we have apps using Qt/Embedded 3.x in OE? Aug 14 15:07:58 lol Aug 14 15:08:09 pb_: printed circuit board? Aug 14 15:08:36 mickeyl: hacky, hacky opie-l* stuff was linked to it Aug 14 15:08:39 mickeyl: pcb.sourceforge.net, the printed circuit board design package. Aug 14 15:08:50 why do developers have this fondness for non standard ui's for apps? Aug 14 15:09:02 though, actually, I guess this latest annoyance is caused by the netlister, not actually by PCB itself. Aug 14 15:09:24 * emte doesnt have any affinity for the geda suite Aug 14 15:09:42 pb_: ah i see. hmm, no screenshots on the site Aug 14 15:09:50 I quite like gschem, and PCB has gotten a lot better than it used to be. Aug 14 15:09:50 ~lart GUI app websites without screenies Aug 14 15:09:51 * ibot farts in GUI app websites without screenies's general direction Aug 14 15:09:54 i use eagle, atleast the devices are decent in it Aug 14 15:10:12 mickeyl: missing your daily dose of eye-candy? Aug 14 15:10:17 koen: *nod* :) Aug 14 15:10:20 qucs > geda Aug 14 15:10:22 wish i couls afford pantheon Aug 14 15:10:27 could* Aug 14 15:10:30 * treke|home had this odd idea that somehow palms would be less susceptible to this crap Aug 14 15:10:34 there's some bad interaction between gschem and gnetlist, though, which causes anything marked as NC in the schematic to end up connected to a "NoConnection" net. Aug 14 15:10:54 heh Aug 14 15:11:04 one big 'noconnection' net Aug 14 15:11:09 yeah, exactly Aug 14 15:11:21 which, obviously, is almost exactly the opposite of what you'd want Aug 14 15:11:40 yeah Aug 14 15:12:10 five seconds with vi sorts it out, of course, but it's irritating to have to remember to do that every time. Aug 14 15:12:52 my biggest problem with the suite was building new devices Aug 14 15:13:00 well trying i should say Aug 14 15:13:27 also, this os doesnt grasp that it doesnt need to suspend if it's on external power Aug 14 15:13:51 adding new schematic symbols is pretty easy these days with tragesym. Aug 14 15:13:58 I'm still not getting what's going on here ... I pulled monotone.vanille.de "org.openembedded.{dev,nslu2-linux,dreambox}" ... it took a few hours ... then a few hours later I pulled monotone.vanille.de org.openembedded.dev ... it should only have had a few revs (I think) but instead it had 963 revs does this seem strange to anyone else or am I not understanding something? Aug 14 15:14:03 not if it doesnt exist Aug 14 15:14:13 pardon? Aug 14 15:14:20 they would always get mangled somehow Aug 14 15:14:22 also wondering if there's a safe way to cancel a pull Aug 14 15:14:26 kados: strange indeed Aug 14 15:14:32 ctrl-c Aug 14 15:14:36 cool Aug 14 15:14:43 koen: at least I'm not going crazy ;-) Aug 14 15:14:53 koen: any troubleshooting suggestion? Aug 14 15:15:00 the devices would be to spec and then when i went to use them they would eb way out Aug 14 15:15:01 not really Aug 14 15:15:06 hum Aug 14 15:15:16 I'd say go to sleep and check in the morning Aug 14 15:15:19 well I guess I'll just wait for it to finish Aug 14 15:15:26 then see what happens next time I run it ;-) Aug 14 15:16:03 honestly I don't know how a server could be synched twice an hour Aug 14 15:16:18 one sync would surely interupt another Aug 14 15:16:20 15536 files ... Aug 14 15:17:01 and tahts jsut in dev Aug 14 15:17:10 (because it takes more than just 30 minutes to sync ;-)) Aug 14 15:17:53 * emte signs the petition to convince kergoth to write bb2 :) Aug 14 15:18:22 * mickeyl signs the petition to convince kergoth to write a locomo SD driver Aug 14 15:18:34 lol Aug 14 15:18:52 thats RD's dept no? Aug 14 15:19:03 RP* Aug 14 15:19:04 no Aug 14 15:19:12 kergoth is the only one who has the documents to write it Aug 14 15:19:21 ah Aug 14 15:19:22 and RP has no locomo hardware Aug 14 15:19:30 * koen signs the petition to give lots of money to "Freyther and Kooi inc." Aug 14 15:19:54 s/Freyther and Kooi inc/Freyther, Kooi, and Lauer Inc." Aug 14 15:19:57 * zecke signs the Government Funding reqest Aug 14 15:20:31 i dont suppose anyone has filed for non-profit organisation status? Aug 14 15:20:32 * zecke turns the time to 1999 and raises 6 million Euro VC Aug 14 15:20:33 mickeyl: every friday night a complimentary bottle of leffe brun Aug 14 15:20:42 emte: no that is to be done Aug 14 15:20:43 _good_ idea! Aug 14 15:20:50 then you could offer tax receipts for donations Aug 14 15:20:53 might be not good for my overweight though Aug 14 15:20:55 emte: but I would like to see hh.org getting non-profit Aug 14 15:21:11 emte: status and tax exemption Aug 14 15:21:17 Hello ppl, I've got an ipaq 3630, which means limited space, so I thought lets install a bootstrap-image and after that (via usb) install the packages that I want. sshing to the ipaq works fine, but now I need a login/password? Is there a default password for the root user? If not, is it possible to do the thing I want to do in an other way? Aug 14 15:21:22 you would need to file in each country Aug 14 15:22:14 hdi: self built version? Aug 14 15:22:20 but corps would be more likly to donate money Aug 14 15:22:53 hdi: #familiar or #handhelds.org Aug 14 15:22:53 which could in turn be used to "hire" developers Aug 14 15:23:26 emte: I would like to see hh.org getting the non-profit status in the US Aug 14 15:23:40 emte: but first we need to see how hh.org lives without HP Aug 14 15:23:41 zecke: bitbake bootstrap-image Aug 14 15:23:51 france, has to file for that doesnt he? since he owns it? Aug 14 15:23:56 hdi: if you're not building a debug distro Aug 14 15:24:10 hdi: you do not set a default password, and password less root login via ssh is disabled Aug 14 15:24:19 emte: I don't think he owns hh.org Aug 14 15:24:35 emte: but france would be a good candidate, but maybe they suprise us anyway Aug 14 15:25:19 zecke: is there an 'easy' way to enable a password less login? Aug 14 15:25:49 zecke: france is busy with talking to lawyers about hh.org's status Aug 14 15:25:59 zecke: so we'll see in a week or two Aug 14 15:26:01 hdi: DEBUG_DISTRO="1" could do Aug 14 15:26:03 ah Aug 14 15:26:21 emte: but most importantly is to have enough hardware for hh.org Aug 14 15:26:32 emte: and probably a sysadmin doing administration of it Aug 14 15:26:35 hdi why dont you just install the gpe or opie image? Aug 14 15:26:48 well you could use taht blackdog thing Aug 14 15:27:04 i forget who posted it Aug 14 15:27:24 http://www.projectblackdog.com/site/index.html Aug 14 15:27:56 emte: limited space, a gpe-image leaves me no space to install any additional packages Aug 14 15:28:04 unless hh.org is doing something i dont know about, that would be powerful enough Aug 14 15:29:17 the gpe image is pretty bare Aug 14 15:29:30 or used to be Aug 14 15:29:48 'unless hh.org is doing something i dont know about' Aug 14 15:29:53 what crack are you on? Aug 14 15:30:09 heavy db work Aug 14 15:30:22 emte: the gpe-image leaves me less than a Mb of space Aug 14 15:30:56 perhaps it has a heavy media stream aspect Aug 14 15:31:02 hdi: grep conf/distro for DEBUG Aug 14 15:31:03 * emte shrugges Aug 14 15:31:05 hdi: and set that Aug 14 15:32:06 i've only saw it do web, light db, ftp/file serving Aug 14 15:34:08 * emte wonders how well microchip's webserver performs Aug 14 15:35:42 zecke: DEBUG not found for the familiar distro, can I just add DEBUG_BUILD = "1" to the conf file? Aug 14 15:36:14 hdi: no that is something else ;) Aug 14 15:36:22 hdi: look at packages/dropbear*/*.bb Aug 14 15:36:39 hdi: a special patch for password less connections is applied... Aug 14 15:36:43 * koen sees DISTRO_TYPE = "debug" in familiar-0.8.3.conf Aug 14 15:38:15 * emte sees "things" everywhere Aug 14 15:38:47 hdi: sorry, I meant DISTRO_TYPE Aug 14 15:39:02 patch -p1 < ${WORKDIR}/allow-nopw.patch Aug 14 15:39:04 hdi: alternatively you provide your own base-passwd and set a default password Aug 14 15:39:17 :( Aug 14 15:39:22 it jumped to 6 hours Aug 14 15:39:27 ugh Aug 14 15:39:53 information overflow :-), I'll try the DISTRO_TYPE thingy ;-) Aug 14 15:41:45 just setting DISTRO to familiar-0.8.3 and doing a bitbake bootstrap-image should do the trick no? Aug 14 15:42:31 03nail 07org.oe.nslu2-linux * r97bdaac8... 10/packages/vsftpd/vsftpd_2.0.3.bb: Fix hardcoded paths mess. Create localstatedir/shared/empty if not exists Aug 14 15:42:35 03koen 07org.oe.dev * r381f85b2... 10/packages/iptables/iptables_1.3.3.bb: Add iptables 1.3.3 Aug 14 15:42:37 g'night Aug 14 15:42:39 03nail 07org.oe.nslu2-linux * r2cd93746... 10/ (2 files in 2 dirs): Add vsftpd to feed Aug 14 15:42:43 03koen 07org.oe.dev * rf54beea7... 10/packages/iptables/iptables_1.3.3.bb: Install iptables in sbindir instead of bindir Aug 14 15:44:57 cya Aug 14 15:45:03 n8 Aug 14 15:46:38 night Aug 14 15:46:44 good nite **** ENDING LOGGING AT Sun Aug 14 15:50:06 2005 **** BEGIN LOGGING AT Sun Aug 14 15:50:30 2005 Aug 14 15:51:26 Anyone know why this doesn't work: Aug 14 15:51:27 #CONFFILES = "${sysconfdir}/ssmtp/ssmtp.conf ${sysconfdir}/ssmtp/revaliases" Aug 14 15:51:34 (Except the #, of course) Aug 14 15:53:14 03rpurdie 07org.oe.dev * raa755a63... 10/packages/linux/linux-openzaurus_2.6.13-rc5-mm1.bb: linux-oz-2.6: Merge spitz ssp driver into corgi Aug 14 15:53:44 NAiL: use \ for the newline Aug 14 15:54:09 ssmtp.conf \ Aug 14 15:54:12 then the rest? Aug 14 15:54:23 right Aug 14 15:54:28 It didn't want to work when I did only the first file either... :( Aug 14 15:54:59 hmm Aug 14 15:55:08 where is my python book Aug 14 15:56:00 03rwhitby 07org.oe.nslu2-linux * r53c5b2e8... 10/conf/distro/ (openslug.conf ucslugc.conf): Fixed openslug and ucslugc to use R29 firmware, not the new R63 Aug 14 16:26:04 It seems my filesystem is always read-only :( Aug 14 16:26:39 have you been hanging out with the zaurus crowd? Aug 14 16:26:51 those people always want 'ROM's Aug 14 16:30:16 no Aug 14 16:30:32 I just made an image for an mp3player I need :) Aug 14 16:30:50 what does 'mount / -o remount,rw' say? Aug 14 16:30:56 Let me check Aug 14 16:32:47 mmm can't find / in fstab : Aug 14 16:33:34 I think there is a problem in there Aug 14 16:38:31 'night all Aug 14 16:42:13 night Aug 14 16:46:27 well next time i wont be so dumb Aug 14 16:46:35 lets blame it on the surgery :P Aug 14 16:47:45 hmm Aug 14 16:48:01 33M Aug 14 16:48:17 wonder what i am missing ... Aug 14 16:50:28 hmm .. Aug 14 16:58:06 CoreDump|afk: Sorry about that but I spent hours getting entrance to work..... I understand that it didn't have the session files and needed the Xserver file and such but, I must say again, the e-image is *not ready*. Aug 14 17:55:44 'night all Aug 14 18:14:31 who's working on the 2.6 kernel for the sl-6000 ... anyone know? Aug 14 18:19:09 oops ... I should be more careful ... this is the wrong place to ask this ;-) Aug 14 18:34:34 ok ... for those following this I finally finished the pull ... ran it again and got a more reasonable '19' revs Aug 14 18:35:18 still think we're treading the line with the number of revs per hour running a semi-hourly cron job to pull Aug 14 18:35:27 especially on a slow machine Aug 14 18:35:43 like my 266MHz NSLU2 Aug 14 18:36:08 pulling takes *forever* ;) Aug 14 18:36:09 if you happen to get 35 revs in 30 minutes you run into more than one cron job doing pull (though I'm not sure if that would do something bad) Aug 14 18:36:19 yea! Aug 14 18:36:51 (if we had twice the number of developers, which would be a good thing, we would run into such limitations I think) Aug 14 18:36:56 If the amount of commits go above 1 every 5 mins, the slug simply can't keep up :P Aug 14 18:37:07 and those aren't the kind of limitations you want in a revisioning system Aug 14 18:37:08 yea Aug 14 18:37:50 but then again, what does one really expect from a 32MB 266MHz armeb embedded device? ;) Aug 14 18:38:18 so now that I've got the latest org.openembedded.dev I'd like to run a test Aug 14 18:38:37 (because I've never actually used it, this is my first day installing it) Aug 14 18:39:22 so I've got an audiovox rtm-8000 that needs a new serial_cs.o because the sharp default has the timing off for the voltage regulation (or so I'm told) Aug 14 18:39:54 so I'm thinking I can compile that for my Sharp ROM so that I can use the rtm-8000 on it Aug 14 18:40:14 can anyone give me some tips on how to do this? Aug 14 18:40:25 kind of an outline of the steps I have to take? Aug 14 18:45:22 * kados reads the local.conf Aug 14 18:50:50 under BBFILES := Aug 14 18:51:09 what if I had multiple branches checked out under different directories? Aug 14 18:51:19 can I just have multiple BBFILES entries? Aug 14 18:56:24 OK ... I think I can based on a wiki entry Aug 14 18:57:03 I'm very confused about what to do about the ASSUME_PROVIDED and PREFERRED_PROVIDORS lines in the local.conf Aug 14 18:57:59 I think they are for defining what platforms I want to cross-compile for but I've no idea what I should put in ... does anyone have suggestions? (I'm trying to build for the Sharp sl-6000 ROM) Aug 14 19:00:30 also it looks like I'll need a toolchain and packages for the toolchain (and libraries) but I've no idea where those are or should be ... a hint someone? Aug 14 19:09:55 so I found a toolchain for the Sharp ROM: http://www.externe.net/zaurus/sdk/Embedix.tgz Aug 14 19:10:37 but I'm confused about where to put it ... i thought it belonged in /stuff/build/tosaSharpROM (what I would call the build) Aug 14 19:11:06 but the wiki says that build dir is empty before running bitbake (excepting the conf dir and file) Aug 14 19:11:42 so where should I unpack the Embedix.tgz? Aug 14 19:20:36 raster: can you help me for a second? I'm trying to setup oe for the first time so I can compile a new serial_cs.c (to learn how to use oe and because I need one for some hardware i recently got for my sl-6000) Aug 14 19:20:46 raster: I found a toolchain for the Sharp ROM: http://www.externe.net/zaurus/sdk/Embedix.tgz Aug 14 19:20:53 raster: but I'm confused about where to put it Aug 14 19:22:21 err Aug 14 19:22:23 last i build oe Aug 14 19:22:28 it didnt need that toolchain Aug 14 19:22:34 it built its own Aug 14 19:23:24 raster: ahh ... in that case, I need to speccify something special in the ASSUME_PROVIDED and PREFERRED_PROVIDORS lines in local.conf? Aug 14 19:24:03 raster: what should be specified there (and how can I tell what should be there for future reference?) Aug 14 19:25:14 can't remember Aug 14 19:25:35 raster: ok ... thanks Aug 14 19:31:28 aquadran_: I've a sl-6000 and I'm setting up oe for the first time ... editing the local.conf I'm not sure what to put in the ASSUME_PROVIDED and the PREFERRED_PROVIDERS lines Aug 14 19:31:50 aquadran_: I'd like to compile something for the default sharp rom that comes with the sl-6000 Aug 14 19:32:10 aquadran_: do you know what should be specified there (and how I can tell what should be there for future reference?) Aug 14 19:55:32 ^)tira Aug 14 20:10:46 kados, http://oe.handhelds.org/cgi-bin/moin.cgi/GettingStarted Aug 14 20:23:24 emte: so I don't see anything about the syntax of the local.conf on that page ... unless I'm missing something Aug 14 20:24:12 http://oe.handhelds.org/cgi-bin/moin.cgi/ZaurusKernels Aug 14 20:24:16 emte: and in the local.conf.sample they list some examples for the ASSUME_PROVIDED Aug 14 20:24:37 but those have a 'virtual/' prefix and I don't see any virtual directories in oe Aug 14 20:24:39 as for local.conf it is specified Aug 14 20:24:52 virtual does not refer to a directory Aug 14 20:24:58 ahh ;-) Aug 14 20:25:58 $ cp ../org.openembedded.dev/conf/local.conf.sample conf/local.conf Aug 14 20:26:05 as per Aug 14 20:26:13 http://oe.handhelds.org/cgi-bin/moin.cgi/GettingStarted#head-7d868d679c5608a1e9e4fabf765c102a6cadc395 Aug 14 20:26:35 the "Create Local Configuration" step Aug 14 20:27:11 right ... got that far ;-) Aug 14 20:27:27 and the ZaurusKernels wiki entry answers all my other questions Aug 14 20:27:30 emte: thanks! Aug 14 20:28:32 your bill comes to one zaurus Aug 14 20:28:34 :P Aug 14 20:28:54 hehe Aug 14 20:30:14 emte: here's a question: do I have to use the handhelds.org toolchain or can I use another like the one at http://externe.net/zaurus/? Aug 14 20:31:08 sure, if you port it so bitbake can use it Aug 14 20:31:26 i wouldnt count on it working too well tho Aug 14 20:32:55 there is a reason the official zaurus sdk was abandoned Aug 14 20:33:04 I suppose I'll eventually know how to 'port it' to bitbake so I'll worry about that later I guess Aug 14 20:33:08 what's that? Aug 14 20:33:16 most of it is garbage Aug 14 20:34:49 ahh Aug 14 20:35:23 emte: ok ... so I've got the cross-compile archive in place, the conf file is set correctly (I hope) and I'm now ready for the next step Aug 14 20:35:42 emte: I thought I'd start small by compiling a single file -- serial_cs.o Aug 14 20:35:58 ? Aug 14 20:36:01 umm Aug 14 20:36:16 emte: I know I need to run bitbake sharp_compatible Aug 14 20:36:17 are you aware how bitbake/oe works? Aug 14 20:36:22 but I'm not sure what that is Aug 14 20:36:28 emte: no ... this is my first time Aug 14 20:36:35 emte: it's a task executer right? Aug 14 20:36:42 not exactly Aug 14 20:36:44 emte: can't I compile a single file for a specific distro? Aug 14 20:36:45 its a build system Aug 14 20:36:55 ahh ... so I have to do a whole build? Aug 14 20:37:11 that is the way its setup Aug 14 20:37:28 whoa ... aren't there options for doing a single file? Aug 14 20:37:34 to build everything that you ask Aug 14 20:37:39 * kados thinks there should be ;-) Aug 14 20:37:52 there are, but you have to build all the requirments first Aug 14 20:38:05 ahh ... well that's ok with me ... so how do I go about that? Aug 14 20:38:20 seeing as serial_cs is part of the kernel you build the base system and the kernel Aug 14 20:39:07 so just the default 'bitbake sharp_compatible' should work? or do I need to specify 'base' and 'kernel' somehow? Aug 14 20:39:13 well normally bitbake task-bootstrap is a good start Aug 14 20:39:37 which builds the minimal requirements Aug 14 20:39:40 ok ... is there a wiki page explaining the steps? Aug 14 20:39:44 yes Aug 14 20:39:48 the same one i gave you Aug 14 20:39:53 hum Aug 14 20:40:14 http://oe.handhelds.org/cgi-bin/moin.cgi/GettingStarted Aug 14 20:40:17 its all there Aug 14 20:40:25 there's nothing on the ZaurusKernels page with 'task-bootstrap' Aug 14 20:40:29 ahh Aug 14 20:40:58 thanks ... I missed it Aug 14 20:41:21 the zauruskernel pages is only about just that... a different compiler needed to build the kernel Aug 14 20:41:43 we dont use it for ipaq, etc Aug 14 20:42:10 right Aug 14 20:43:23 * kados runs bitbake task-bootstrap Aug 14 20:43:44 task-bootstrap will tell you if you setup your system right as well Aug 14 20:44:16 myself i have a few issues ... Aug 14 20:44:28 stupid emacs Aug 14 20:45:29 I don't quite get what's happening when I run bitbake ... It's building ... but I haven't touched the code yet so how do I know it's building what I want? Aug 14 20:45:48 bitbake has aprox two stages Aug 14 20:46:13 it builds it's own toolchain and then it uses that to build the target system Aug 14 20:46:44 it makes all the substitutions so that your linking and building against the target libs and not the host Aug 14 20:46:51 etc etc Aug 14 20:47:13 ahh ... so you specify the 'packages' you want ... which I guess are a combo of the type of image and maybe the architecture Aug 14 20:47:34 it populates it's toolchains with the right stuff Aug 14 20:47:41 look in the machine and distro dirs in openembedded/conf Aug 14 20:47:45 and also puts the source trees in? Aug 14 20:48:00 as well as the openembedded/packages/meta/ Aug 14 20:48:14 it downloads the sources it needs Aug 14 20:48:41 I don't have any openembedded dir ... I only have a org.openembedded.dev (or is that what you meant?) Aug 14 20:49:01 just out of curiosity you are aware it takes ~4gb to build an image right? Aug 14 20:49:10 inside there Aug 14 20:49:36 yea ... I've got two-hundred gigs on this box Aug 14 20:49:41 so that should be enough ;-) Aug 14 20:50:11 yeah thats enough to build world Aug 14 20:50:21 emte: there's no openembeded dir inside org.openembedded.dev Aug 14 20:50:25 emte: should there be? Aug 14 20:50:37 nah Aug 14 20:50:51 thats jsut old naming convention on my part Aug 14 20:51:13 should there be a stuff/openembedded dir? or is stuff/org.openembedded.dev/ the same thing? Aug 14 20:51:29 same thing Aug 14 20:51:54 cool Aug 14 20:52:03 org.openembedded.dev is part of the monotone changes Aug 14 20:52:14 ahh Aug 14 20:52:59 so ... how can I tell which packages I'll need to compile the serial_cs.c? Aug 14 20:54:11 hmm .. this isn't good Aug 14 20:54:12 emte: | checking for C compiler default output file name... configure: error: C compiler cannot create executables Aug 14 20:55:01 nope taht isnt good Aug 14 20:55:25 NOTE: Task failed: /home/josh/zaurus/stuff/build/tmp/work/base-passwd-3.5.9-r2/temp/log.do_configure.16510 Aug 14 20:55:28 NOTE: package base-passwd-3.5.9-r2: task do_configure: failed Aug 14 20:55:47 emte: was I supposed to put a password in somewhere? Aug 14 20:57:20 lol Aug 14 20:57:21 no Aug 14 20:57:55 base-passwd is just as it sounds the default setup Aug 14 20:58:33 hmmm ... so what could be wrong then? Aug 14 20:58:49 why can't the C compiler create executables? should I do this as su? Aug 14 20:58:51 checking for C compiler default output file name... configure: error: C compiler cannot create executables Aug 14 21:03:22 emte: any idea why this error is happening? google's not much help Aug 14 21:03:51 your compiler was built without the ability for users to create execuitibles Aug 14 21:04:05 is one posibility Aug 14 21:04:07 shoot ... so that must be debian default Aug 14 21:04:14 doubtful Aug 14 21:04:23 i am using debian and i cant say i have that issue Aug 14 21:04:37 hmmm ... anything else it could be ? Aug 14 21:04:48 which release are you using? Aug 14 21:04:53 sarge Aug 14 21:05:14 hmm Aug 14 21:07:19 not off hand ... Aug 14 21:07:36 you install all the autotools? Aug 14 21:07:45 autotools? Aug 14 21:07:55 m4, autoconf, automake, etc etc Aug 14 21:08:09 hmm ... maybe not ... is there a list somewhere? Aug 14 21:08:29 most interdepend Aug 14 21:08:40 ok Aug 14 21:08:41 use synaptic or something Aug 14 21:09:06 apt-get, yum, whatever your package handler of choice is Aug 14 21:09:29 which version of automake is right? Aug 14 21:09:42 you will need two versions Aug 14 21:09:45 1.9 ok to use? Aug 14 21:09:57 or atleast i did Aug 14 21:10:05 which two did you need? Aug 14 21:10:43 sec i'll look Aug 14 21:10:51 sweet ... thanks Aug 14 21:12:19 hmm Aug 14 21:12:25 its only listing 1.9.6 Aug 14 21:12:25 asside from automake autoconf and m4 what else do I need? Aug 14 21:12:28 ok Aug 14 21:12:40 psycho is recomended Aug 14 21:12:43 strange ... I can only get 1.9.5 Aug 14 21:12:54 i am using etch Aug 14 21:13:01 might not be in your repo Aug 14 21:13:13 weird Aug 14 21:13:13 Note, selecting automake1.4 instead of automake Aug 14 21:13:14 E: Couldn't find package 1.9 Aug 14 21:13:26 hmm Aug 14 21:14:02 psycho doesn't appear to be in stable repos either Aug 14 21:14:52 or etch for that matter ;-) Aug 14 21:17:06 * france is away: Away Aug 14 21:35:13 I'm probably just missing some silly dependency Aug 14 21:35:53 I've got autotools-dev, autoconf, automake installed Aug 14 21:36:28 anyone know what else I need and what might cause the checking for C compiler default output file name... configure: error: C compiler cannot create executables Aug 14 21:36:34 error? Aug 14 23:42:01 kados, look at the config.log, will give better information on what test that failed Aug 15 00:45:50 morning Aug 15 00:54:46 morning all Aug 15 00:55:35 yo RP Aug 15 01:11:45 mornin Aug 15 01:13:08 morning Aug 15 01:28:40 morning all Aug 15 01:54:26 morning Aug 15 01:55:21 hi mickeyl Aug 15 01:57:08 mickeyl: we have a missing dependency in entrance. After installing "some stuff" I can now enter text. Aug 15 01:58:10 morning mickeyl Aug 15 01:58:36 RP: was your big secret spitz stuff ? Aug 15 01:59:02 XorA: Yes, I now have one and 2.6 is moving forwards rapidly on it :) Aug 15 01:59:16 Power Control, Sound and IR are all the are missing Aug 15 02:00:01 CoreDump|home: "some stuff"? Aug 15 02:00:01 RP: bah, lucky you Aug 15 02:00:12 CoreDump|home: Entering text always worked for me, btw. Aug 15 02:00:26 heh, firefox and a ton of dependencies (like pango modules etc) Aug 15 02:00:50 mickeyl: might want to try that with a fresh loop-image :) I doubt it will work Aug 15 02:01:02 CoreDump|home: then someone broke it. it used to work Aug 15 02:01:06 i c Aug 15 02:01:15 can you login into a session? Aug 15 02:01:41 no, when I got that working we had no session scripts et. al. Aug 15 02:01:48 i didn't try a recent build Aug 15 02:02:05 ok :) Aug 15 02:02:05 i'll do that Aug 15 02:02:12 ah wait Aug 15 02:02:20 need to fix a typo Aug 15 02:03:21 k Aug 15 02:03:54 i want to demo E on my Spitz this saturday on a Zaurus User Group Treffen in Hannover Aug 15 02:04:08 so we need to get that working to make people envious :) Aug 15 02:04:36 it looks like we get some Zaurus news about new models then as well Aug 15 02:04:49 :) Aug 15 02:05:02 take that, pdaxrom ;) Aug 15 02:05:43 hehe Aug 15 02:05:51 reading the documentation the config.db is correct. However, the behavior of entrance doesn't match the db Aug 15 02:06:09 might post that on e-devel Aug 15 02:06:35 yeah, once I'm certain that the fault is not on our part Aug 15 02:06:51 it looks like we need to touch the default.edj to make the menu and applet fonts larger btw. Aug 15 02:07:13 mickeyl: News from Sharp on a new device? Aug 15 02:07:20 mickeyl: synced Aug 15 02:07:29 RP: not Sharp, but TriSoft Aug 15 02:07:36 who have close connections to 'em Aug 15 02:07:46 * mickeyl pull Aug 15 02:08:10 mickeyl: Just after I get "up to date" device wise :-/ Aug 15 02:08:11 also we need a sane default config (larger icons/symbols etc, remove non-workung modules) Aug 15 02:08:41 btw: did you manage to get xmonobut working? Aug 15 02:08:57 didn't try Aug 15 02:09:01 what's that? Aug 15 02:09:04 <--- x newbie Aug 15 02:09:20 Who do we have with a C1000 here? Anyone other than Bernardo|Away? Aug 15 02:09:22 a little program to switch your touchscreen to middle click or right click Aug 15 02:09:24 RP: 'ya know, the greatest device is always the next one :) Aug 15 02:09:27 RP: me Aug 15 02:10:03 CoreDump|home: I'd like to have click and hold for right-click just like we have on QtE Aug 15 02:10:16 mickeyl: ahh, that'd be ideal Aug 15 02:10:22 CoreDump|home: Ah. I'm searching for akita owners interested in testing/bug fixing 2.6 :) Aug 15 02:10:50 RP: I'll be happy to assist. But the "Fixing" part is sadly out of the question ;) Aug 15 02:11:11 CoreDump|home: The main difference is an i2c io expander chip instead of the second scoop device... Aug 15 02:11:17 CoreDump|home: Do you have a serial lead? Aug 15 02:11:26 ~lart Sharp for unnecessary subtle difference Aug 15 02:11:26 * ibot overclocks Sharp until Sharp burns out for unnecessary subtle difference Aug 15 02:11:36 only USB client and host cables Aug 15 02:12:21 CoreDump|home: Ok, that makes things more interesting. Sadly, my serial lead only works one way on spitz :-/. It just means I need to get things right first time for akita :) Aug 15 02:12:36 ahem ouch heh Aug 15 02:12:47 got a diagram of the lead? Aug 15 02:12:55 mickeyl: the Xfree86 for pdaXrom has click and hold for right button in it Aug 15 02:13:04 mickeyl: its BAD (tm) Aug 15 02:13:17 why? I find it pretty good on Opie Aug 15 02:13:31 mickeyl: scollbars Aug 15 02:13:34 the delay could be shorter, but yes. quite usable Aug 15 02:13:44 mickeyl: window dragging Aug 15 02:13:54 CoreDump|home: Its the connector that's the problem I believe. I don't have a diagram but would be surprised if google didn't know. Be warned that Cx000 are much more sensitive to voltage levels Aug 15 02:13:54 XorA: ah Aug 15 02:14:00 XorA: Opie fixes these problems Aug 15 02:14:14 we "fix" it by having apps with special knowledge Aug 15 02:14:18 you only get the right-click menu if you do not move the mouse Aug 15 02:14:34 RoH is only activated on a per-widget base Aug 15 02:14:36 which makes it easier for us Aug 15 02:14:37 (which is a pain at times) Aug 15 02:14:40 the best one would be gesture based Aug 15 02:14:58 but thats a project for someone Aug 15 02:15:18 or maybe an applet you click, the next click is RMB? Aug 15 02:15:28 CoreDump|home: thats xmonobut Aug 15 02:15:37 which doesn't work for E :) Aug 15 02:15:43 CoreDump|home: why not? Aug 15 02:15:46 no applet Aug 15 02:15:49 btw, does anyone know the chip number for the io expander in akita? or know of a link to internal photos of it? Aug 15 02:15:53 and keybinding doesn't work for me Aug 15 02:16:00 CoreDump|home: just run xmonobut from an xterm Aug 15 02:16:14 CoreDump|home: ah, keybinding worked last time I tried it Aug 15 02:16:18 I did, doesn't work. It won't show a applet Aug 15 02:16:34 CoreDump|home: no xembed epplet? Aug 15 02:16:45 hmm? Aug 15 02:17:04 CoreDump|home: I thought xmonobuts icon was standard xembed one Aug 15 02:17:19 CoreDump|home: I may be misthinking though Aug 15 02:17:21 XorA: might be. maybe E doesn't support it yet Aug 15 02:17:35 CoreDump|home: never managed to make gnome and E play well Aug 15 02:17:41 :) Aug 15 02:17:56 I could have sworn I'd seen photos of the c1000 internals somewhere but I can't find them again... Aug 15 02:18:07 RP: I'll read up on the serial connection for Akita and come back to you. I'd rather not fry my Akita :) Aug 15 02:18:37 CoreDump|home: you could run matchbox-panel beside E Aug 15 02:18:55 CoreDump|home: Serial isn't essential. It would be easier but I think we can manage without Aug 15 02:18:57 XorA: ahh, I'll try Aug 15 02:19:31 CoreDump|home: I can copy sharp's i2c code to get the lcd running - we can then just experiment with the backlight until we get a working i2c driver, then apply it to the lcd code Aug 15 02:19:36 ahem, there's no task-gpe? grr Aug 15 02:20:03 Actually, its not lcd related, just backlight... Aug 15 02:20:15 i c Aug 15 02:20:45 I could really use the chip part number so I can look up its datasheet Aug 15 02:21:26 Providing sharp haven't embedded it in the PGA of course Aug 15 02:21:33 i see 4 holes, possible for screws (not much light here) Aug 15 02:21:47 anything i need to look out for when opening the beast? Aug 15 02:23:50 RP: I have a full electronics laboratory at work. I'll open it and make some photos Aug 15 02:23:56 CoreDump|home: task-gpe and gpe-meta could need some cleanups Aug 15 02:24:03 * mickeyl points the gpe folks to opie-collections.inc Aug 15 02:24:09 indeed Aug 15 02:24:35 * CoreDump|home points to Opie and GPE folks for his combined GPE _and_ opie collection.inc :D Aug 15 02:25:10 * XorA shoots CoreDump|home Aug 15 02:25:16 * CoreDump|home chuckles Aug 15 02:25:24 XorA: IrDA doesn't work for me on our latest c7x0 kernel. Are you sure it worked for ya? Aug 15 02:25:36 mickeyl: does meta-opie actually use opie-collections.inc? Aug 15 02:25:38 mickeyl: you modprobe all the modules? Aug 15 02:26:10 pb_: no, but meta-opie is not involved in image creation. meta-opie is just for building _everything_ opie in OE. opie-collections is used by opie-image Aug 15 02:26:11 * CoreDump|home hugs his NFS root. 200MB are getting full though Aug 15 02:26:53 XorA: i didn't modprobe any extra modules. all the irda modules are autoloaded for me. Or at least that's what I'm thinking. which modules are you referring to? Aug 15 02:28:06 mickeyl: the pxaxxx irda module, and irda itself Aug 15 02:29:14 d'oh Aug 15 02:29:23 hmm Aug 15 02:29:38 do we have some new things modular which we didn't autoload? Aug 15 02:29:45 mickeyl: yes Aug 15 02:29:49 aaah Aug 15 02:29:53 that would explain Aug 15 02:29:53 oh great, installing GPE makes entrance logins work *sigh* Aug 15 02:30:15 XorA: ok, thanks, i'll take care about adding that to autoload Aug 15 02:30:15 (for GPE and E hehe) Aug 15 02:30:31 heh Aug 15 02:30:49 mickeyl: cool, I had to go back to 2.6.11 for weekend, so I wont be able to re-test 2.6.13 until tonight Aug 15 02:31:17 CoreDump|home: I've only had the c760 apart and they're reasonably straightforward if you go slowly and carefully - some of the connectors are small Aug 15 02:31:43 shouldn't be a problem once I get the casing off Aug 15 02:32:03 beeing an electronics engineer has it advantages :D Aug 15 02:32:15 CoreDump|home: It helps :) Aug 15 02:32:31 I could have sworn I'd seen photos somewhere before though... Aug 15 02:32:45 ok, E started once, now only GPE comes up. /me starts debugging Aug 15 02:33:10 CoreDump|home: Before taking it to bits leave the battery out for a while. The backlight capacitors can store enough charge to blow out the board's fuse :-/ Aug 15 02:33:11 oh, and the monobut thingy doesn't work in GPE, too. Maybe the OE version is B0rked Aug 15 02:33:20 CoreDump|home: I found that out the hard way... Aug 15 02:33:21 RP: ouch Aug 15 02:33:27 CoreDump|home: works here Aug 15 02:33:28 SMD fuse? Aug 15 02:33:46 CoreDump|home: yes. It was the 3rd time I had it apart and I was being less careful... Aug 15 02:33:52 hehe Aug 15 02:34:00 RP: replace with self resetting fuse Aug 15 02:34:40 XorA: It was replaced with a small piece of fuse wire as it was what I had to hand ;-) Aug 15 02:35:07 :D Aug 15 02:35:13 RP: Cool, 30AMP :-D Aug 15 02:35:26 we don't have SMD stuff at work Aug 15 02:35:34 anyways, i'll be careful ;) Aug 15 02:35:54 CoreDump|home: As long as you give it time to discharge all the caps, it will be fine Aug 15 02:36:00 ok Aug 15 02:36:17 do you need shots of both sides of the PCB? Aug 15 02:36:19 I'd guess I shorted a connector with a power line on to ground hence blowing the fuse Aug 15 02:36:40 CoreDump|home: All I need is the part number of the i2c IO expander chip Aug 15 02:36:44 it should be similiar to the spitz thingy Aug 15 02:37:01 I don't know which side of the board it will be on :-/ Aug 15 02:37:18 RP: hah, nice one :) I dunno what an I2C chips looks like lol. Aug 15 02:37:41 no sweat, you'll have your shots in one or two days Aug 15 02:39:27 who is the recent libxine guru ? Aug 15 02:39:29 * mickeyl looks around Aug 15 02:39:41 we should update to 1.0.2, there are a lot of bugfixes and memory/cpu optimizations Aug 15 02:40:25 CoreDump|home: Ok, thanks :). I don't know what it will look like either, sadly ;-) Aug 15 02:40:40 :) Aug 15 02:41:38 RP: http://forum.palmislife.com/viewthread.php?tid=46721&fpage=1 Aug 15 02:42:45 hi Aug 15 02:42:57 RP: working on the akita 2.6 kernel? Aug 15 02:43:39 mickeyl: your CIA is stuck again btw ;) Aug 15 02:43:48 it's not necessarily my CIA Aug 15 02:43:56 sometimes the real CIA gets stuck Aug 15 02:43:59 * mickeyl looks Aug 15 02:44:15 *nod* Aug 15 02:44:16 mine Aug 15 02:44:18 :D Aug 15 02:44:20 * mickeyl restarts Aug 15 02:44:23 heh Aug 15 02:44:45 CIA-13: seems that mickeyl 0wnz you. Aug 15 02:45:29 heh :D Aug 15 02:45:33 :) Aug 15 02:45:56 XorA: looks like someone replaced a Akita PCB with a Spitz on heh Aug 15 02:46:09 i wonder if it is enough to solder on the CF slot Aug 15 02:46:25 crap, an IC is missing, too Aug 15 02:46:56 CoreDump|home: that is the problem, the missing scoop... Aug 15 02:47:18 oh well, can't have everything;) Aug 15 02:47:31 I'd bet the 3100 has it... Aug 15 02:47:52 anyone know for sure now what the 3100 is? Aug 15 02:47:58 3000 + flash Aug 15 02:47:59 or Aug 15 02:48:02 1000 + hd ? Aug 15 02:48:07 ahem Aug 15 02:48:13 is there a difference? Aug 15 02:48:18 of course Aug 15 02:48:18 03coredump 07org.oe.dev * r26ad3f94... 10/packages/altboot/files/altboot-menu/Advanced/40-bootNFS: altboot: Do not use a fixed time to wait for WLAN on NFS boot. Ping the NFS host and continue once the network is up. Aug 15 02:48:21 hardware wise Aug 15 02:48:23 03mickeyl 07org.oe.dev * r9ee885cc... 10/packages/opie-mediaplayer2/ (4 files): rrecommend all xine plugins in opie-mediaplayer2, remove opie-mediaplayer2-divx package Aug 15 02:48:27 yep, the missing chip you found Aug 15 02:48:28 03mickeyl 07org.oe.dev * r978b81b2... 10/packages/meta/ (e-image.bb meta-e-x11.bb opie-collections.inc): Aug 15 02:48:29 ship more apps in e-image Aug 15 02:48:29 ship opieplayer2 in opie images on machines with >= 24MB flash Aug 15 02:48:32 03nail 07org.oe.nslu2-linux * r65876316... 10/packages/ (6 files in 6 dirs): Update MAINTAINER and LICENCE Aug 15 02:48:36 03coredump 07org.oe.dev * r96359cb8... 10/packages/e17/ (3 files in 3 dirs): entrance: s/Sessions/Session/ as used by config.db Aug 15 02:48:58 mickeyl: I'd say it's a 3000 + flash, since all sharp had to do to the 3000 hw would be to change the flash chip Aug 15 02:49:02 CoreDump|home: the 3000 and the 1000 have a lot of differences kernel wise Aug 15 02:49:14 Bernardo: i see Aug 15 02:49:17 oh, I didn't know that Aug 15 02:49:23 Bernardo: If I can help the akita 2.6 effort, I will :) Aug 15 02:49:33 mickeyl: aren't those differences because of the missing ic? Aug 15 02:49:34 :) Aug 15 02:49:40 Bernardo: probably Aug 15 02:49:51 mickeyl: THe differences so far just look like a remapping of a load of gpios onto an io expander... Aug 15 02:49:56 RP: I am going (once more) to check ebay for a affordable serial caple Aug 15 02:50:01 cable Aug 15 02:50:14 Bernardo: Keep in mind it needs to be an offical Sharp on Aug 15 02:50:18 SerialIO won't work Aug 15 02:50:30 no? Aug 15 02:50:40 I thought the serialio serial cable would work Aug 15 02:50:42 ibot: pester do13[offline] for a fully working 2.6 on tosa Aug 15 02:50:43 do13[offline] for a fully working 2.6 on tosa: Are we there yet? .. Are we there yet? .. Are we there yet? see http://beverlys.net/LJ/BuggingYou.swf Aug 15 02:50:47 Bernardo: The levels are wrong. Mine won't work with the C3000 Aug 15 02:51:10 I can't transmit to the device (although I can read from it) Aug 15 02:51:42 I didn't know the serialio did level conversion, but it seems right, since you have 5 V at one end, 3.3 at the other... Aug 15 02:51:42 RP: theyve built a max-232 into the cable? Aug 15 02:52:03 XorA: Sharp did that as standard Aug 15 02:52:12 RP: bloody hell, that sucks Aug 15 02:52:25 SerialIO hacks it with resistors (or so its rumoured) Aug 15 02:52:48 luckilly most PC UARTS are quite forgiving Aug 15 02:52:50 CoreDump|home: That chip that's missing is probably the second scoop interface and has all the CF interface logic in it Aug 15 02:53:10 XorA: That's what they rely on :-/ Aug 15 02:53:40 XorA: Thanks for those photos - they're interesting. I can't see the io expander chip :-( Aug 15 02:53:52 That may mean its a virutal device like scoop... Aug 15 03:12:12 mickeyl: out of interest, what do you change to make a module autoload? That way I can fix myself next time I find one thats needed Aug 15 03:13:07 XorA: see the kernel .bb Aug 15 03:13:11 it's just an entry Aug 15 03:13:21 feel free to fix it this time, I'm without a Z atm. Aug 15 03:16:57 mickeyl: kernel.bbclass? Aug 15 03:17:18 Hia.. is http://ewi546.ewi.utwente.nl/OE/OE.db.bz2 updated regularly, or is it oldish? (1371 revs to pull for update, please come back in 12 hrs :) Aug 15 03:18:03 XorA: no, the kernel bbfile Aug 15 03:18:12 XorA: e.g. linux-openzaurus-2.6.13... Aug 15 03:18:20 there are module autoload statements within Aug 15 03:18:43 wrobbie: it's updated whenever koen remembers. i don't think he runs a cron job yet Aug 15 03:18:46 oh, it seems i'm talking bollocks, pull just finished.. ...? Aug 15 03:19:24 mickeyl: Im either blind or stupid, but I cant find them Aug 15 03:19:46 XorA: there are some in the kernel.bbclass, too Aug 15 03:21:09 ahah, linux-openzaurus is the only kernel.bb to not have them :-) Aug 15 03:21:17 heh Aug 15 03:21:18 oops Aug 15 03:21:19 :) Aug 15 03:21:42 guess we should add 'em there, too Aug 15 03:21:56 maybe the ones in kernel.bbclass suffice. Aug 15 03:22:27 * XorA doesnt have the appropriate stuff to flash at work today, so Ill fix them when I do if no-one else has Aug 15 03:23:20 I still have a problem with kernel module dependencies: http://bugs.openembedded.org/show_bug.cgi?id=230. Does anybody have a hint how to debug this? Aug 15 03:23:32 Scared. monotone just fired up vim for a 3 way merge. That's a bit clever. Aug 15 03:24:48 good morning all Aug 15 03:25:10 morning koen Aug 15 03:29:09 hey RP Aug 15 03:31:41 Noodles: best install kdiff3 Aug 15 03:31:43 or meld Aug 15 03:33:04 Hi, when I do a "monotone update" I get a bunch of "monotone: warning: missing packages/dpkg/dpkg_1.10.28.bb" and similar warnings. Is that serious? What can I do to get rid of the problem? Aug 15 03:33:44 A "monotone pull" goes through fine. Aug 15 03:34:22 try 'monotone list missing | xargs monotone drop' Aug 15 03:35:23 koen: Thanks. Aug 15 03:38:08 mickeyl: vim seems ok. I just need to figure out what I need to do to make monotone happy. I thought it wanted me to edit the resultant file to what I wanted, but that doesn't appear to be it. Aug 15 03:40:31 Noodles: if you have monotone 0.22 use --lca to avoid spurious merges Aug 15 03:41:12 koen: Is that a command that would be a good candidate for inclusion into my daily cronjob? Aug 15 03:41:27 Doesn't seem to like that with a "--update". I have a local file I've changed that's been updated is the issue (2.6 spitz config. ;) Aug 15 03:41:37 I'll prod it in a minute. I'm suppose to be doing real work. ;) Aug 15 03:42:21 Noodles: ah, in your case the merge conflict is an actual conflict :) Aug 15 03:42:32 Laibsch: not really necessary Aug 15 03:43:18 OK Aug 15 03:44:08 mickey|lunch: Was just about to comment on ugt ;-) Aug 15 03:44:31 Or maybe you went to a diner for lunch? LOL Aug 15 04:02:43 03dyoung 07org.oe.nslu2-linux * r8aca482c... 10/packages/linux/unslung-kernel/defconfig: add conntrack and queue modules for iptables Aug 15 04:02:47 03dyoung 07org.oe.nslu2-linux * r05bbc684... 10/packages/linux/unslung-kernel_2.4.22.l2.3r29.bb: Oops forgot to bump the version Aug 15 04:21:40 NOTE: preferred version 2.16 of binutils-cross not available Aug 15 04:21:40 NOTE: package binutils-cross-2.15.94.0.1: started Aug 15 04:21:55 There is a binutils-2.16.bb though.. Aug 15 04:22:14 eh, binutils-cross_2.16.bb Aug 15 04:22:33 why wouldn't it want to take that one? Aug 15 04:23:01 sounds like a bug Aug 15 04:23:15 in the .bb? Aug 15 04:23:27 in bitbake, probably Aug 15 04:23:36 you might need to take it up with Ãœberhacker Zecke. Aug 15 04:42:31 Aargh, things just ain't right. bitbake and OE are up-to-date. The initilization of OE has been set up according to GettingStarted a couple of weeks back and not changed since. Doing a inside the build dir now give me "ERROR: Unable to open conf/bitbake.conf". Of course, relative to the build dir ../bitbake/conf/bitbake.conf exists and is readable. What is OE trying to do? Aug 15 04:42:31 03koen 07org.oe.dev * r01859ba3... 10/packages/iptables/iptables_1.3.3.bb: Really install into /sbin, fix packaging, bump PR and add RRECOMMENDS for the needed kernel-modules Aug 15 04:43:23 Laibsch: it can't find OE's conf/bitbake.conf Aug 15 04:43:31 whichi suggests your BBPATH is wrong Aug 15 04:44:03 BBPATH=/mnt/pfaffe/src/openembedded/build:/mnt/pfaffe/src/openembedded/org.openembedded.dev Aug 15 04:44:12 Looks OK to me. Aug 15 04:45:13 That is taken straight from the output of set from the same shell I issued the "bitbake nunome" Aug 15 04:46:56 If OE is looking for conf/bitbake.conf and not ../bitbake/conf/bitbake.conf, I'd assume problems. Aug 15 04:47:25 again, it needs conf/bitbake.conf from *OE*, not bitbake Aug 15 04:48:54 Was that introduced in the last 2 or 3 weeks? I apparently never neede it. Where would it be located? $PATH2OE/conf/bitbake.conf? Aug 15 04:49:47 $PATH2OE/conf/ exists but has IIRC always only held local.conf for my setup. Aug 15 04:50:03 GettingStarted does not talk about that file. Aug 15 04:51:42 Laibsch: it should be in /mnt/pfaffe/src/openembedded/org.openembedded.dev/conf/bitbake.conf Aug 15 04:51:58 Laibsch: as koen said, OE's copy Aug 15 04:52:59 There is no conf directory inside org.openembedded.dev. Shouldn't that be created by monotone update? Aug 15 04:53:09 Laibsch: it should Aug 15 04:53:32 hm Aug 15 04:53:39 * Laibsch pondering what went wrong Aug 15 04:54:13 Laibsch: fix that and it should then work Aug 15 04:54:27 Laibsch: COPYING.MIT MT README classes conf contrib files packages site Aug 15 04:54:34 Laibsch: is contents of mine Aug 15 04:54:53 GettingStarted.sh bitbake oe.db org.openembedded.dev tmp Aug 15 04:54:53 OE.db.bz2 build oe.db.old sources Aug 15 04:55:05 Sorry, wrong dir. Aug 15 04:55:20 COPYING.MIT MT README contrib files packages site Aug 15 04:55:26 see http://monotone.vanille.de/viewmtn/manifest.psp?id=01859ba396a554baa1fcb520fb65f0218f54d476& for all the files needed Aug 15 04:55:48 I miss classes and conf dir. Aug 15 04:57:03 * Laibsch curses Aug 15 05:11:10 hi all Aug 15 05:11:43 What while not work from OE? When I while not have psyco . amd64 problem (x86_64) Aug 15 05:13:50 03nail 07org.oe.nslu2-linux * rb00bca80... 10/packages/linux/ (nslu2-kernel_2.6.12.2.bb openslug-kernel-2.6.12.2/defconfig): Add IPTABLES config and CONFIG_BRIDGE Aug 15 05:13:54 03nail 07org.oe.nslu2-linux * r72ced9eb... 10/ (2 files in 2 dirs): Add iptables to openslug feed Aug 15 05:15:01 OK, did another "monotone update" conf and classes dirs will not be created. What can I do now? Would deleting org.openembedded.dev help in the hope it will be correctly recreated? Aug 15 05:15:22 try doing a checkout Aug 15 05:15:45 cvs checkout? of what? Aug 15 05:16:07 did I said cvs? Aug 15 05:16:26 no, did I say you did? ;-) Aug 15 05:16:29 afaik we were talking about monotone Aug 15 05:16:59 Should I delete the directory first? Aug 15 05:17:02 or move it? Aug 15 05:20:03 OK, I moved the old dir and started a fresh checkout. Aug 15 05:29:40 I'm close to getting OE setup for my first build Aug 15 05:29:56 but when I run bitbake task-bootstrap Aug 15 05:30:03 it fails with this message: Aug 15 05:30:12 | checking for C compiler default output file name... configure: error: C compiler cannot create executables Aug 15 05:31:20 I have automake, autoconf, and m4 installed but I'm afraid there may be more autotools that I'm missing Aug 15 05:31:27 does anyone have any suggestions? Aug 15 05:32:13 kados: has you installed the 2.95.3 toolchain? Aug 15 05:32:37 koen: I think so ... at least I unpacked it in /usr/local/arm Aug 15 05:32:50 http://oe.handhelds.org/cgi-bin/moin.cgi/ZaurusKernels holds all the steps Aug 15 05:32:52 koen: do I need to actually install it? Aug 15 05:33:04 yes Aug 15 05:33:20 if you use 2.4 and/or sharprom-compatible you need that Aug 15 05:34:01 ok ... but the ZaurusKernels page doesnt say to install it ;-) Aug 15 05:34:19 hmmm, someone must have changed that Aug 15 05:35:03 should I be afraid that installing this is going to mess up other things on this system? Aug 15 05:36:25 sorry if that's a stupid question ... I've never done cross-compiling before Aug 15 05:37:35 hmmm ... in fact, I don't even really see a place to install it ... Aug 15 05:37:55 I grabbed this bz2 file: http://handhelds.org/download/projects/toolchain/archive/cross-2.95.3.tar.bz2 Aug 15 05:38:14 unpacked it in /usr/local/arm/2.95.3/ Aug 15 05:38:23 does that constitute 'installing'? Aug 15 05:40:03 yes Aug 15 05:40:28 ok ... well I've done that ;-) Aug 15 05:40:29 iirc there are some symlink and PATH mangling instructions in that wiki Aug 15 05:41:45 hmmm, I see a brief mention of path but nothing on symlinks Aug 15 05:42:26 how do I make sure that the PATH is correct? Aug 15 05:42:51 $ echo $PATH Aug 15 05:42:51 /home/josh/zaurus/stuff/bitbake/bin:/usr/local/bin:/usr/bin:/bin:/usr/bin/X11:/usr/games Aug 15 05:44:40 03rwhitby 07org.oe.nslu2-linux * rc8d07f5c... 10/packages/nslu2-binary-only/nslu2-linksys-ramdisk_2.3r63.bb: Updated the location of the R63 rootfs which has the ufsd.o module removed (it's a commercially licensed module). Aug 15 05:45:27 ok ... I'm going to try again after adding that dir to my path Aug 15 05:48:41 * france is away: Away Aug 15 05:54:39 <[g2]> pop quiz -- What was the first processor to support ARM Thumb ? Aug 15 05:56:25 sa1100? Aug 15 05:57:20 ARM10, 1998 ? Aug 15 05:58:05 no, earlier than that. The thumb architecture was announced in 1995. Aug 15 05:58:24 * koen suspects rwhitby is reading http://www.arm.com/aboutarm/milestones.html Aug 15 05:58:31 * NAiL knows Aug 15 05:58:42 1995: "TI sampled first ARM Thumb core" Aug 15 05:59:50 <[g2]> koen, thx Aug 15 05:59:57 koen: you found me out :-) Aug 15 05:59:59 "The ARM design was started in 1983 as a development project by Acorn Computers Ltd." :P Aug 15 06:00:23 Wasn't ARM7TDMI the first one? Aug 15 06:00:50 no Aug 15 06:01:15 oh, first to support thumb? yeah, I guess it was. Aug 15 06:01:30 SA1100 doesn't support thumb at all, so it definitely wasn't the first. Aug 15 06:01:58 Everything seems to be back in working order. Thanks. Aug 15 06:02:22 well ... got farther this time Aug 15 06:02:40 not it fails at the getkey.c Aug 15 06:02:57 | /home/josh/zaurus/stuff/build/tmp/work/busybox-1.00-r30/busybox-1.00/console-tools/getkey.c: In function `getkey_main': Aug 15 06:03:04 several messages similar to that one Aug 15 06:03:14 | make: *** [/home/josh/zaurus/stuff/build/tmp/work/busybox-1.00-r30/busybox-1.00/console-tools/getkey.o] Error 1 Aug 15 06:15:19 03rwhitby 07org.oe.nslu2-linux * r3e3bbfe0... 10/packages/nslu2-binary-only/ (4 files): Updated unslung-binary-kernel distro to use the expurgated version of the Linksys R63 firmware. Aug 15 06:16:03 * XorA has to look up expurgated Aug 15 06:16:44 http://bau2.uibk.ac.at/sg/python/Scripts/TheBookshopSketch Aug 15 06:18:53 The expurgated version of the linksys firmware is all the things except the commercial NTFS driver that Linksys seems to have licensed from Paragon, and included in the latest NSLU2 firmware to enable the NSLU2 to read and *write* NTFS disks. Aug 15 06:20:47 Since the URL to download the firmware directly from Linksys is far too complex for the bitbake fetch support to handle, we have to mirror the bits that we need. However, we're not going to mirror that commercial binary kernel module until there is clarity on the licensing of it. Aug 15 06:21:53 rwhitby: ah cool, its just a word Ive never seen before, must be the lack of religion in my youth :-) Aug 15 06:23:22 or the lack of Monty Python in your youth ... Aug 15 06:23:30 never really liked them Aug 15 06:29:07 afternoon Aug 15 06:49:32 hmm... thinking of using some type of dynamic building of packages for mplayer, ie, build all packages that provides virtual/codec or something similar... is there already some way to do this or will i have to write something myself for it? Aug 15 06:51:14 ldc: mplayer pulls in little from outside itself Aug 15 06:52:15 yea, i know... but the thing is that i would like to be able to add/remove codecs without having to modify the .bb file for mplayer, but still have it build all codecs that are available as .bb's Aug 15 06:53:26 ldc: that sentence doesnt make sense when you talk about mplayer, its a monolithic blob Aug 15 06:53:38 well, not really Aug 15 06:54:19 libxvmc for example can be used for mplayer, but then you need to know that you have it installed before, also the same with a couple of other codecs Aug 15 06:55:14 ldc: then you can DEPENDS on libxvmc but RRECOMMENDS it so it doesnt automatically get installed Aug 15 06:55:38 ldc: thats if mplayer supports dlopening it Aug 15 06:56:49 dlopening is d4 3v1l Aug 15 06:57:10 * koen practices mplayer & ffmpeg propaganda a bit Aug 15 06:57:34 ldc: bitbake itself does not have any form of gentoo style USE flags Aug 15 06:57:49 hmm... Aug 15 06:59:34 isnt libxvmc only supposed to be used statically? Aug 15 06:59:47 so i'll probably have to do some scripting... would otherwise be nice to have some DEPENDS="group/codecs" and then all modules that provides group/codecs would get built... anyway... thanks for the info... back to the coding **** ENDING LOGGING AT Mon Aug 15 07:19:38 2005