**** BEGIN LOGGING AT Wed Jul 20 23:59:56 2005 Jul 21 00:06:29 How do we change a memory map? (in my case i have a custom ixp425 with 64MB SDRAM) Jul 21 00:08:01 03law 07org.oe.dev * r9a46103c... 10/packages/sylpheed/sylpheed_2.0.0beta6.bb: added openssl dependency to the sylpheed 2.0.0beta6 bb file Jul 21 00:17:10 belracu: what do you want to map? devices? mtd partitions? Jul 21 00:23:45 re Jul 21 00:24:07 __law__: sylpheed cannot use gnutls instead of openssl? Jul 21 00:25:40 <__law__> hrw|work, without openssl it wont compile Jul 21 00:26:25 hrw|work: the previous version of sylpheed used gnutls Jul 21 00:26:52 we have a special version of gnutls which support much more that the upstream one Jul 21 00:27:41 aha Jul 21 00:29:44 since openssl is 2MB or so Jul 21 00:30:55 morning Jul 21 00:31:38 morning all Jul 21 00:32:57 morning proti & ade Jul 21 00:33:17 ~lart modutils-cross Jul 21 00:33:18 * ibot takes a big bite out of modutils-cross's jugular vein Jul 21 00:38:09 my package name is calctool sources get stored into calctool-2.4 but bitbake always tries to go into calctool-2.4-1.0. i dont want that extra 1.0. where should i disable that ? Jul 21 00:43:49 eJumbo S="${WORKDIR}/${PN}-${PV}" Jul 21 00:44:09 hrw|work: I bet the file is calctool-2.4.bb Jul 21 00:44:39 yes Jul 21 00:44:59 rename it to calctool_2.4.bb Jul 21 00:45:44 i did not get what the problem is ? Jul 21 00:46:00 '-' vs '_' Jul 21 00:46:11 '_' is used as a seperator Jul 21 00:46:40 with a '-' bitbake thinks you have a program called 'calctool-2.4' and defaults the version to 1.0 Jul 21 00:46:59 so it looks for calctool-2.4-1.0 Jul 21 00:47:26 morning Jul 21 00:48:08 i am trying to compile an compile an x application calctool. i just wanted to know from where the x libs are taken from. the make file specified /usr/include/X11 does bitbake add a path which resides in TMPDIR/work/include ? Jul 21 00:48:09 koen: have you looked at "our" gnutls? Jul 21 00:49:27 _law_: I have a patch that can make sylpheed work with gnutls Jul 21 00:50:23 <__law__> XorA, can you post it? Jul 21 00:50:30 __law__: It basically means that instead of checking CA certs against files, it will just warn they are not valid Jul 21 00:51:06 __law__: I wanted to test the sylpheed first, didnt get a chance last night, expect it on mailing list in next half an hour or so Jul 21 00:52:34 I have a bad day... modutils-cross dont want to build, glib-2.0-native too;( Jul 21 00:53:40 has anyone experience with high-speed uarts on arm? Jul 21 00:54:10 my bt module talks at 921600 bps, and the bt stack loses bytes Jul 21 00:56:58 ~lart uart Jul 21 00:56:58 * ibot hits uart with an anvil and laughs with a contralto voice ... Haha Ha HA Ha Jul 21 00:59:32 who has c7x0 and does not have case for it? Jul 21 01:00:27 hrw|work: me Jul 21 01:02:29 XorA: want that one: http://shop.brando.com.hk/pdairzaurusslc760case.php maybe? Jul 21 01:03:15 hrw|work: looks quite sweet Jul 21 01:03:29 XorA: and I have spare one Jul 21 01:03:47 today got version with clip Jul 21 01:03:52 clip? Jul 21 01:04:46 w8 - I'll look into dict Jul 21 01:04:57 belt clip Jul 21 01:05:03 hrw|work: ah cool Jul 21 01:05:22 so I have version without beltclip to spare Jul 21 01:07:29 hrw|work: how much? Jul 21 01:08:35 20usd? post costs included Jul 21 01:11:39 morning all Jul 21 01:11:46 hi RP Jul 21 01:11:53 hi RP Jul 21 01:12:33 hrw|work: blah, checked my bank balance and Im on empty this month :-( Jul 21 01:13:33 XorA: shit happens Jul 21 01:13:45 RP: dont you need leather case for c7x0? Jul 21 01:15:26 __law__: that patch seems to work, also on mailing list just in case Jul 21 01:16:31 <__law__> XorA, i ll try it Jul 21 01:24:03 2.6.13-rc3 comes with preliminary xip feature. Is there any insterest to use it on Z ? Jul 21 01:24:35 hrw|work: Its not top of my list of priorities ;-) Jul 21 01:24:40 xip? Jul 21 01:24:42 xip = eXecute In Place. Jul 21 01:24:55 Only ext2 for now Jul 21 01:24:57 proti: so stuff stays in disk cache? Jul 21 01:24:57 proti: xip isn't that useful when you use jffs2 and compression Jul 21 01:25:12 hi Jul 21 01:25:24 XorA: You don't copy applications into memory when you run them Jul 21 01:25:26 hi do13 Jul 21 01:25:28 hi dirk Jul 21 01:25:43 RP: I hate to be pedantic, but they have to be in CPU memory somewhere Jul 21 01:25:48 hi RP hrw|work Jul 21 01:26:04 XorA: flash is mapped into address space Jul 21 01:26:12 XorA: No, they don't. That's the point... Jul 21 01:26:27 XorA: Consider the Flash as memory space. Don't copy to buffer memory. Jul 21 01:26:49 hrw|work: ah ok, flash on arms isnt in IO space, cool Jul 21 01:27:03 XorA: NOR is, NAND isn't Jul 21 01:27:08 XorA: atleast I thought it is Jul 21 01:27:54 RP: Does JFFS2 compression ? Jul 21 01:27:56 christ its taken a long time for those patches to make it :-) Jul 21 01:28:03 proti: Yes, very much so Jul 21 01:28:43 Ah, that's why performance drops to zero when it comes under the 1M free space. Jul 21 01:29:01 proti: That and the GC routines Jul 21 01:30:02 So no use for xip in zaurus or PDA in general unless giving compression ? Jul 21 01:30:10 *giving up Jul 21 01:30:34 moin Jul 21 01:30:50 morning zecke Jul 21 01:30:53 Its useless on the Zaurus as they mostly use NAND and compression, yes :-/ Jul 21 01:30:54 can CF card be mapped into memory space same way as on x86? Jul 21 01:30:55 hi zecke Jul 21 01:31:27 XorA: I'm not sure you can map the whole card at once Jul 21 01:31:55 Although perhaps you can. At the end of the day, is an ide interface... Jul 21 01:32:23 No, that doesn't make sense. So I don't know :) Jul 21 01:32:43 i am trying to compile a package calctool with oe. the make file contains Library paths and include paths. these are with respect to root. /usr/lib/X11R6 and /usr/include how does oe take care of these ?. Jul 21 01:33:08 RP: doh, yeah, forgetting difference between ATA CF cards and memory CF cards Jul 21 01:33:23 ~lart xora for supidity Jul 21 01:33:23 * ibot wallops xora with a main rotation server that needs rehubbing. It won't take long for supidity Jul 21 01:33:25 eJumbo: patch the makefile Jul 21 01:33:33 i think these paths should be replaced or preceeded by oe paths that is in TMPDIR. Jul 21 01:34:12 this is how it has to be done or oe preceeds its own paths ? Jul 21 01:34:50 eJumbo: try to find another patched makefile Jul 21 01:35:04 eJumbo: OE exports some variables you can use in the Makefile Jul 21 01:38:59 howdy people Jul 21 01:39:14 i am using X Libs and headers; where does oe store then in /tmp/oetmp/staging/i686-linux/include or in /tmp/oetmp/staging/arm-linux/include/X11; cant find anything in i686-linux ? Jul 21 01:39:16 hey zecke, of to munich today Jul 21 01:40:55 koobla, sorry was not here I need to map devices (PCI) Jul 21 01:41:37 belracu: no prb. have a look at the kernel sources - in arch/arm/mach-ixp4xx/ixdp425-setup.c Jul 21 01:41:49 i am asking because BUILD_LDFLAGS points to the i686-linux and it does not have X headers ? Jul 21 01:42:02 belracu: not sure exactly, i am working on pxa, but this is a likely place to start Jul 21 01:42:03 re Jul 21 01:42:33 hi zecke Jul 21 01:54:25 good morning Jul 21 01:54:38 hi mickeyl Jul 21 01:56:11 hey mickeyl Jul 21 01:56:15 off to munich today Jul 21 01:58:05 mickeyl: dont feel a need for leather case for your c7x0? Jul 21 01:58:46 hehe Jul 21 01:59:14 I have too much cases already, but thanks Jul 21 01:59:24 the thing is... I don't use them Jul 21 01:59:37 I tried to, but in the end I always take em' without the cases Jul 21 01:59:55 koobla, my ixp425 have 64MB and have 4 pci device .. but the pci memory windows is itself 64MB. for 4 devices... Jul 21 02:00:31 koobla, the pci devices are dected by redboot and linux .. but i get segmentaion fault on inserting the drivers Jul 21 02:00:42 ;) Jul 21 02:02:11 guys : what is the difference between the x system we use on gpe-image and a standard X system (kbdrive (???) and xfree) ? Jul 21 02:02:39 define "standard X system" Jul 21 02:02:49 belracu: sorry, don't know much about pci programming, maybe the base addresses do not match? Jul 21 02:02:51 and don't use x86 in that sentence :D Jul 21 02:03:00 well... xfree or xorg... Jul 21 02:03:07 ? Jul 21 02:03:10 hmm Jul 21 02:03:10 maybe ? Jul 21 02:03:20 X is a protocol specification Jul 21 02:03:31 some servers implement Xv11 Jul 21 02:03:34 xorg is one Jul 21 02:03:37 xfree is another one Jul 21 02:03:41 kdrive is yet another one Jul 21 02:03:45 the differences are: Jul 21 02:03:48 * supported hardware Jul 21 02:03:50 * size Jul 21 02:03:59 morning Jul 21 02:04:03 mickeyl: collie will get rebuilded 3.5.3 images Jul 21 02:04:06 kdrive is a lightweight X implementation that supports the framebuffer Jul 21 02:04:07 hi Rene Jul 21 02:04:13 which is why we are using it Jul 21 02:04:14 and does kdrive have some limitations ? Jul 21 02:04:16 hrw|work: *nod* Jul 21 02:04:24 no serious ones. Jul 21 02:04:25 hey Marcin Jul 21 02:04:32 servus rene Jul 21 02:04:51 moin mickey Jul 21 02:04:55 mickeyl: i'm trying to understand whay snes9x won't run, and it looks like it doesn't come from SDL. Jul 21 02:05:00 are you all on the angstrom-dev list yet, btw.? Jul 21 02:05:13 mickeyl: have to recheck Jul 21 02:05:13 if not, please subscribe Jul 21 02:05:22 we want to get that things slowly started Jul 21 02:05:24 mickeyl: me? no, not yet. where's that hosted? Jul 21 02:05:36 reenoo: angstrom-dev@hh.org Jul 21 02:05:49 mickeyl: ok. will do Jul 21 02:05:53 Yet another list :-/ Jul 21 02:06:11 RP: yes, but if all goes well, some will phase out :) Jul 21 02:06:17 __law__: any particular reason why you added openssl to sylpheed DEPENDS? Jul 21 02:06:17 mickeyl: will oe stuff that is distro independent still go to oe@hh,org, or is angstrom-dev replacing it? Jul 21 02:06:33 mickeyl: in theory :) Jul 21 02:07:00 __law__: we can't legally distribute GPL apps linked against openssl Jul 21 02:07:20 rwhitby: distro independent stuff still oe@hh.org. Angstrom is just "yet another distribution built out of OE metadata", it's just that we happen to support the devices that previously were only targetted by individual distros like OZ, Fam. and OpenSimpad. Jul 21 02:07:26 mickeyl: I'm subscribed Jul 21 02:07:27 __law__: (thus the gnutls patches...) Jul 21 02:07:48 <__law__> reenoo, i removed openssl again, thanks to XorA for the gnutls patch Jul 21 02:08:49 03law 07org.oe.dev * r93d082d6... 10/packages/sylpheed/ (files/sylpheed-gnutls-extra.patch sylpheed_2.0.0beta6.bb): removed openssl from sylpheed again, added sylpheed-gnutls-extra.patch from XorA Jul 21 02:09:29 __law__: ah, ok Jul 21 02:10:18 mickeyl: i'm not Jul 21 02:10:39 mithro: please do :) Jul 21 02:10:56 mithro: good morning, btw. How's Munich so far? Jul 21 02:11:12 anyway got to run Jul 21 02:11:13 see ya Jul 21 02:11:31 k, have fun Jul 21 02:13:26 __law__: for future reference.. if you come across bugs like that please file a bug in bugzilla or send mail to the list. I'd hate to see stuff like this go in unnoticed Jul 21 02:17:47 where do i find the x libraries for the "Build system" that the oe uses. i have already built an x server and a matchbox. but cant find one in /tmp/oetmp/staging/i686-linux Jul 21 02:19:33 eJumbo: you build for what machine? Jul 21 02:19:50 omap 5912 Jul 21 02:20:17 so look into /tmp/oetmp/staging/arm-linux/ Jul 21 02:21:34 what i am referring is the libraries which are used to link (linkable part of the libraries) to get the final executable. Jul 21 02:22:03 the above process is done on the host machine Jul 21 02:23:04 eJumbo: man... look how other apps have it done and write some magic to get your working Jul 21 02:23:52 when the executable runs. i know that libraries which are in /tmp/oetmp/arm-linux which will be used (as it gets copied onto file system) Jul 21 02:24:37 hrw|work: i am really confused. in the make file i need to specify a path of the lib ? Jul 21 02:25:47 eJumbo: lib and include Jul 21 02:26:11 yes Jul 21 02:26:36 which path i should give ?. which points to arm-linux ? Jul 21 02:26:49 STAGING_INCDIR STAGING_LIBDIR Jul 21 02:29:51 guylhem is funny Jul 21 02:29:53 http://www.oesf.org/forums/index.php?showtopic=13815&hl= Jul 21 02:30:19 essentially it looks like he now recreates what we are doing here since long :D Jul 21 02:30:37 mickeyl: I'm adding the Apex bootloader to OE - where should the resulting apex.bin binary file go when do_populate_staging runs ? Jul 21 02:30:47 mickeyl: wasting out time? Jul 21 02:31:14 rwhitby: deploy/images? Jul 21 02:31:54 mickeyl: guyhem is funny - he want to have fast hackable with gcc2.95 and qtopia/sharp Jul 21 02:31:57 hrw|work: we run slugimage to compile it into an image, so it needs to be staged Jul 21 02:31:57 rwhitby: firmware often goes to /etc Jul 21 02:32:05 rwhitby: so I would prefer Jul 21 02:32:13 STAGING_ETCDIR Jul 21 02:32:19 or something like that Jul 21 02:32:31 or STAGING_DATADIR/apex maybe Jul 21 02:32:34 it will be arch-specific Jul 21 02:32:35 mickeyl: if one only could work with guyhlem Jul 21 02:33:07 zecke_: he insists to do things alone, much the same like the japanese and pdaxrom guys - there's nothing we can do to force people to cooperate Jul 21 02:33:18 kergoth put the current redboot binary in staging/armeb-linux/lib/nslu2-binaries, but Apex is not limited to the slug Jul 21 02:33:47 mickeyl: special I would like to finish chronic, my class based init system Jul 21 02:35:17 mickeyl: I was thinking ${STAGING_DIR}/${HOST_SYS}/loader/apex.bin Jul 21 02:35:31 (sibling to STAGING_KERNEL_DIR) Jul 21 02:35:45 rwhitby: yeah, that sounds good as well Jul 21 02:35:51 and maybe add a STAGING_BOOTLOADER_DIR to bitbake.conf Jul 21 02:36:00 *nod* Jul 21 02:36:05 i'd prefer STAGING_LOADER_DIR though Jul 21 02:36:12 but that's just cosmetics :) Jul 21 02:36:33 zecke_: too many tasks :) complete one, then tackle the next Jul 21 02:38:06 STAGING_LOADER_DIR it is Jul 21 02:38:07 mickeyl: I finally recognize why TT sticks to gcc2 Jul 21 02:38:13 zecke_: tell us Jul 21 02:38:14 hi pb Jul 21 02:38:25 mickeyl: it looks like they invested $$$ to stabilize it to pass all unit tests ;) Jul 21 02:38:36 mickeyl: Once my gcc-4.0.1 is stable on ARM Jul 21 02:38:47 hi hrw|work Jul 21 02:38:52 mickeyl: I won't ever switch my compiler again! Jul 21 02:39:11 zecke_: yeah you have to trust at least anything. Jul 21 02:40:18 mickeyl: OT: gcc4 is unsupported for Qt and Qtopia ;) Jul 21 02:41:22 doesn't really surprises me Jul 21 02:42:54 mickeyl: I find it sad. Harri Porten used to test Qt on all supported platforms and newest compilers... Jul 21 02:43:23 mickeyl: But it looks like opening a branch in Australia pretty much limits you to Aussie employees Jul 21 02:44:43 the ways of the trolls are mysterious... Jul 21 03:04:29 03rwhitby 07org.oe.nslu2-linux * r2cba5405... 10/conf/bitbake.conf: Added STAGING_LOADER_DIR to bitbake.conf Jul 21 03:04:31 mickeyl: there's the bitbake.conf change for your review Jul 21 03:04:33 03rwhitby 07org.oe.nslu2-linux * r3d6cac6f... 10/packages/apex/ (apex_1.2.19.bb nslu2/defconfig): Added the Apex bootloader (initially configured for nslu2) Jul 21 03:06:09 rwhitby: looks good to me. Jul 21 03:07:14 rwhitby: update documentation.conf as well Jul 21 03:07:26 STAGING_LOADER_DIR[doc] = "Description" Jul 21 03:10:07 zecke_: is someone else going to add all the other STAGING_* variables ;-) Jul 21 03:10:53 I'm happy to add it, but it looks like STAGING_* things are not being added there ... Jul 21 03:14:05 rwhitby: yeah, leave it, I'll add it with some more descriptions Jul 21 03:24:34 JustinP: in the future, please include information about the version of the bb files you add as attachements into the bugtracker. I always have to check for the latest version of the software before I can apply the patch :D Jul 21 03:25:00 03mickeyl 07org.oe.dev * r68eb66f0... 10/packages/squashfs-tools/ (squashfs-tools-native_2.0r2.bb squashfs-tools_2.0r2.bb): squashfs(-native) depends on lzma(-native), fixes BUG #154 Jul 21 03:25:39 mickeyl: monotone 0.21 is out since Sunday. Seems to include the speed improvement. Jul 21 03:25:58 proti: yep. i'm already using it Jul 21 03:26:11 the improvements are visible to me Jul 21 03:26:17 mickeyl: No interop problem with 0.20 ? Jul 21 03:26:31 no. netsync stayed the same. just db was upgraded Jul 21 03:28:08 ok. will try. I tried a pull on my laptop (P4 2.4ghz) a gave up after 3h of fan roaring and right knee burn, it only commited 300 rev on 783. Jul 21 03:28:42 initial pull still slow Jul 21 03:28:47 i don't think that was adressed Jul 21 03:29:30 I end up getting the db on my sparc and pulled from it. Jul 21 03:30:52 My wife came on me because of roaring fan in the living room prevented her to hear the sound of TV. Jul 21 03:31:14 hehe Jul 21 03:31:18 03mickeyl 07org.oe.dev * rf10cb7bf... 10/packages/lineak/ (files/add-zaurus-remote-support.patch lineak_0.8.3.bb): add lineak (Linux support for Easy Access and Internet Keyboards) courtesy Justin Patrick. BUG #152 Jul 21 03:31:32 yeah, MT is quite processor intensive. Finally a software that really uses a PC. Jul 21 03:32:19 heh Jul 21 03:32:21 Parsing bb file is not all either. Jul 21 03:32:28 Parsing bb files is not bad either. Jul 21 03:33:13 bbl lunch Jul 21 03:34:26 modutils-cross does not build on my system for mipsel :( Jul 21 03:35:27 gimme a DISTRO and a MACHINE and let me try Jul 21 03:35:49 wrt54oe/wrt54 Jul 21 03:35:51 k Jul 21 03:35:52 * mickeyl starts Jul 21 03:36:45 NOTE: package modutils-cross-2.4.27: completed Jul 21 03:38:56 which gcc on host? Jul 21 03:41:36 mickeyl: I got: http://pastebin.com/317678 Jul 21 03:42:14 gcc version 3.3.4 (pre 3.3.5 20040809) Jul 21 03:42:25 ah Jul 21 03:42:30 invalid lvalue looks like gcc 4 Jul 21 03:42:32 or 3.4 Jul 21 03:42:43 i guess you need to apply a patch to conform to the more strict ISO checking Jul 21 03:42:45 gcc (GCC) 4.0.1 (Debian 4.0.1-2) Jul 21 03:42:52 yeah Jul 21 03:43:11 or check if there's an upgrade for that package Jul 21 03:43:12 ~lart linksys for using 2.4 Jul 21 03:43:12 * ibot gives linksys an extra strength ACME sleeping pill, sending linksys to sleep for 150 years, and awakening to seven strange dwarfs and a large apple for using 2.4 Jul 21 03:43:28 do we not have a 2.6 kernel for the wrt yet? Jul 21 03:43:55 there is no 2.6 for wrt yet Jul 21 03:44:23 bummer Jul 21 03:50:01 pb_: could you give me a small bio of tje DejaGNU at CodeSourcery? Jul 21 03:50:10 pb_: is he somehow involved in gcc as well? Jul 21 03:51:18 zecke_: yah, he is one of the arm port maintainers Jul 21 03:52:07 NOTE: package modutils-cross-2.4.27-r5: task do_build: completed Jul 21 03:52:34 pb_: could you give me his email address again? Jul 21 03:52:54 paul@codesourcery.com, iirc Jul 21 03:53:05 * zecke_ is mailing him now Jul 21 03:53:10 ibot, google for paul brook Jul 21 03:53:31 ah, I forgot, ibot lost its power to do anything useful Jul 21 03:53:42 ibot, you are now known as suckbot Jul 21 03:53:43 pb_: what are you talking about? Jul 21 03:54:33 pb_: thanks Jul 21 04:11:34 03hrw 07org.oe.dev * rd70f182e... 10/packages/modutils/ (files/gcc4.patch modutils-cross_2.4.27.bb): Jul 21 04:11:34 added MIPS related change to gcc4.patch - now it is buildable if host system Jul 21 04:11:34 has gcc4 as default (like it is in Debian 'sid') Jul 21 04:13:02 * zecke_ hopes his portupgrade -a finishes today Jul 21 04:18:47 when i compile my package using oe_make i get the error: make: ccache: Command not found Jul 21 04:19:42 eJumbo: CC=ccache arm-linux-gcc needs "" probably Jul 21 04:22:38 hrw|work: i am using oe_runmake 'CC=${CC}' 'CROSS=${HOST_PREFIX}' in do_compile. Jul 21 04:23:11 do 'CC="${CC}"' then? Jul 21 04:28:50 ~hail debian source mirrors Jul 21 04:28:51 * ibot bows down to debian source mirrors and chants, "I'M NOT WORTHY!!" Jul 21 04:31:51 but where from they get zlib 1.2.3... Jul 21 04:35:39 ot it Jul 21 04:35:43 s/^/g Jul 21 04:42:27 any nslu2 gurus aroud? Jul 21 04:55:55 NOTE: package zlib-1.2.3-r0: task do_build: completed Jul 21 04:56:20 patch should be updated to apply without offsets and fuzz? Jul 21 04:56:28 hrw|work: yes Jul 21 04:56:38 ok Jul 21 04:57:02 that's why I still prefer patcher, btw. Jul 21 04:57:15 you can instruct it to autoupdate the patches and afterwards you just have to copy them back Jul 21 04:57:20 03mickeyl 07org.oe.dev * r9670e736... 10/packages/linux/linux-wrt_2.4.30.bb: linux-wrt: fix SRC_URI Jul 21 05:02:02 mickeyl, I need to nag you about bug #93 Jul 21 05:02:08 is now a good time? Jul 21 05:03:49 let me read it Jul 21 05:03:55 thanks Jul 21 05:04:08 you had some comments and I submitted a patch to addresss the, Jul 21 05:04:13 them Jul 21 05:04:15 thanks Jul 21 05:05:48 ah i see Jul 21 05:06:02 looks like i will apply everything now Jul 21 05:06:07 thanks Jul 21 05:06:24 if more than a couple of people use it, I will be delierious :) Jul 21 05:07:01 :) Jul 21 05:07:07 i'll reorder the metadata a bit Jul 21 05:07:10 although I think hooking a usrp to a netslug and using the MAC would be fun :) Jul 21 05:07:11 just fyi Jul 21 05:07:14 thats fine Jul 21 05:07:33 I need to install monotone when you have done this Jul 21 05:07:47 I am actively adding some code to our project that uses omnievents Jul 21 05:08:01 when I have all that going I will revisit my omnievent bb file Jul 21 05:08:27 cool Jul 21 05:08:32 hmm Jul 21 05:08:45 re. RDEPENDS, are you sure it needs to depend on libusb and swig? Jul 21 05:08:59 libusb should be sucked in automatically through shlibdeps if it's a real .so Jul 21 05:09:13 swig is a wrapper generator, does it have a run time lib which isn't a .so ? Jul 21 05:09:19 hmmm Jul 21 05:09:31 I wasn't aware of auto shlib finding Jul 21 05:09:38 so leave libusb off Jul 21 05:09:40 k Jul 21 05:09:43 leave swig off also Jul 21 05:09:53 and I will let you know if there is a problem Jul 21 05:10:16 ok Jul 21 05:10:25 mickeyl: how is that .so magic working anyway? Jul 21 05:10:33 mickeyl: it uses objdump for the SO_NEEDED Jul 21 05:10:35 zecke_: inspecting via objdump iirc Jul 21 05:10:47 mickeyl: but how does it know which package provides that lib? Jul 21 05:11:02 I don't really understand swig, wish the code didn't depend on it Jul 21 05:11:53 zecke_: it records that in some part of the staging area Jul 21 05:12:02 zecke_: pb_ did that code Jul 21 05:12:30 mickeyl: ah the .sodeps files get copied Jul 21 05:12:35 ~hail pb_ Jul 21 05:12:36 * ibot bows down to pb_ and chants, "I'M NOT WORTHY!!" Jul 21 05:12:41 The key is that the .so must specifically specify its soname Jul 21 05:12:55 and objdump magic does it rest - its very clever code Jul 21 05:14:19 can anyone hand linksys-wlconf.tar.gz over to me? it's unavailable upstram Jul 21 05:14:22 upstream, even Jul 21 05:17:33 w8 Jul 21 05:18:00 mickeyl: http://www.hrw.one.pl/tmp/ Jul 21 05:19:09 mickeyl: http://www.codesourcery.com/public/qmtest/qm-2.3/qm-2.3.tar.gz Jul 21 05:19:32 thanks Jul 21 05:19:57 Hi, I am setting up the new oe env. I have build monotone 0.21 and fetched oe, but I am having problems to get bitbake sources. svn co svn://svn.berlios.de/bitbake/trunk/bitbake just gives me a connection refused error. I am not sure whether it is local network problem or is svn server down in berlios. Jul 21 05:20:17 qm? Jul 21 05:20:32 mickeyl: gcc testing tool, but written in python Jul 21 05:20:44 Crofton_: I'll apply usrp later today, because I want to add sdcc-native prior to that Jul 21 05:20:46 zecke_: cool Jul 21 05:20:46 lamikr: are you firewalled? Jul 21 05:20:48 i'll have a look Jul 21 05:20:50 * mickeyl wanders home Jul 21 05:20:54 thanks Jul 21 05:21:32 did I make it depend on that? I may have had trouble building it and done an assume_providided .... Jul 21 05:21:33 I closed my own firewall, but I am not sure whether the university's network block svn port to outside. Jul 21 05:21:54 it was a while ago I did this Jul 21 05:22:40 zecke_: Are you able to make svn connection to berlios? Jul 21 05:24:22 lamikr: yes but I#m using the developer server/account Jul 21 05:25:45 lamikr: I cannot conenct anonymously either Jul 21 05:26:05 zecke: ok, so then it is a server problem. Jul 21 05:26:59 zecke: Is there any changes that you or somebody else could put bitbake snapshot somewhere for download? Jul 21 05:27:42 03hrw 07org.oe.dev * r91b0af10... 10/packages/zlib/ (3 files in 2 dirs): added zlib 1.2.3 Jul 21 05:28:44 lamikr: yes, but not within next couple of hours Jul 21 05:31:31 zecke_: ok, I can wait if I can not find snapshots from elsewhere. I can also try to contact berlios admins from the anonymous connection problem. Jul 21 05:41:07 pb_: http://news.bbc.co.uk/1/hi/uk/4703777.stm - skynews reports it as explosions Jul 21 05:41:12 * Twiun sighs Jul 21 05:44:00 Twiun: crap Jul 21 05:44:35 XorA, and everyone else, ##london for live-ish news tracking Jul 21 06:06:44 any one here to answer questions about kdrive posibilities ? Jul 21 06:10:46 libpcap needs change - m4 into staging Jul 21 06:12:03 hey jamey Jul 21 06:18:14 hi zecke_ Jul 21 06:26:13 hi all Jul 21 06:26:38 hey florian_kc Jul 21 06:26:38 Did anyone here try to build meta-sdk lately? Jul 21 06:27:07 * france is back (gone 11:10:47) Jul 21 06:27:20 jamey: hi Jul 21 06:27:30 hi france Jul 21 06:30:05 Hi, I am trying to set up a bb file for nunome. I am not very proficient with bitbake. The compilation ended in a failure which apprently was due to the fact that bitbake called make with DESTDIR but the Makefile was looking for TARGET_DIR. My question is should I make the changes in the original source or should I change something (what?) in the bb file? Jul 21 06:31:21 do_install() { oe_runmake TARGET_DIR=${D} } Jul 21 06:31:27 Laibsch: in that case I would create a custom do_install() Jul 21 06:32:12 Thanks, guys. I still need to learn a lot about the bitbake build system. Jul 21 06:32:38 Laibsch: and get annoyed by people using nonstandard options/build systems Jul 21 06:33:14 As long as bitbake seems flexible enoug to deal with I am still happy ;-) Jul 21 06:33:33 arpwatch is nicely nstalled in debian... "make install" and then copy rest of stuff ;) Jul 21 06:34:43 * koen reads http://www.nslu2-linux.org/wiki/HowTo/OverClockTheSlug Jul 21 06:35:16 Oh well... i see what's up: ipkg doesn't install gcc-cross-sdk because its architecture is "i686" and the toolchain is for ARM. Jul 21 06:35:26 koen: that's a nice one :-) Jul 21 06:36:04 gcc-cross-sdk should be architecture i686, or whatever the host is Jul 21 06:37:27 jamey: right... it is, but it looks like the configuration file for ipkg doesn't list the host architecture. Jul 21 06:37:47 florian_kc: that's odd. it certainly should do. Jul 21 06:37:56 oh, that would also be a problem Jul 21 06:38:20 meta-sdk constructs two parallel ipkg configuration files, one for the host and one for the target. Jul 21 06:38:26 * florian_kc wonders is IPKG_ARCH_append="${BUILD_ARCH}" in sdk.bbclass would do the trick. Jul 21 06:38:27 maybe it's using the wrong one when it's trying to install gcc. Jul 21 06:39:05 florian_kc: I don't think that ought to be necessary. Jul 21 06:39:11 what's the ipkg command that fails? Jul 21 06:40:16 hrm... wait... that looks like gcc-cross-sdk gets installed Jul 21 06:40:47 "Installing gcc-cross-sdk (3.3.4-r1) to root..." looks fine for me... Jul 21 06:42:23 yeah Jul 21 06:45:25 do we have a method to add/remove user from system in postinst/prerm? Jul 21 06:45:41 deluser Jul 21 06:45:53 koen: without any checks? Jul 21 06:46:40 heh, please don't Jul 21 06:47:28 I want to package arpwatch (for my wrt54) but that daemon runs as arpwatch in debian so I need to add/remove user Jul 21 06:47:52 ${WORKDIR}/sdk/image/usr/lib/ipkg/status contains the packages for the crosscompiler... but not more. Jul 21 06:49:05 sed and grep should get you there as far as the checking is concerned Jul 21 06:49:39 it would actually be cool if the ipkg guys started shipping a "functions" script that provided common tools like add / remove user Jul 21 06:49:55 /usr/libs/ipkg/scripts/functions - you could source it Jul 21 06:50:05 * CP|OLS really, really hopes that they don't already have that Jul 21 06:50:11 CP|OLS: put a enh. request in ipkg bugzilla Jul 21 06:50:12 Its way too early in the morning to feel like a jackass Jul 21 06:51:14 drat, I can't find my soldering iron Jul 21 06:53:02 Hey all Jul 21 06:54:09 hey Cwiiis Jul 21 06:54:15 Zecke_: Do you know the url where you will drop the bitbake snapshot? Jul 21 06:54:17 hi Cwiiis Jul 21 06:54:23 no Jul 21 06:55:10 Cwiiis: can you check one thing for me? gpe-image/oz353-collie-r21 kernel (my build) Jul 21 06:55:12 Hey koen, hrw|work Jul 21 06:55:46 hrw|work: Sure - The kernel doesn't work with the stock opie image btw, but I'm sure you know that already :) Jul 21 06:55:59 Cwiiis: yes - I have rebuilt images Jul 21 06:56:04 lamikr: http://hyboria.mine.nu/~dp/bitbake-1.3.1.tar.gz Jul 21 06:56:13 Cwiiis: http://www.hrw.one.pl/tmp/oz353-r21/ Jul 21 06:56:25 hrw|work: Ah, right, cool Jul 21 06:57:39 ah, found my soldering iron Jul 21 06:57:44 * koen attacks R83 Jul 21 06:58:03 pb_: I guess i have the point - these x86 packages are installed using the correct ipkg.conf, but the arm packages are installed with the other ipkg.conf which doesn't contain x86 which makes ipkg ignore these packages. imho the bug is that various arm packages depend on the x86 gcc-cross-sdk packages. Jul 21 06:58:10 koen: *g* Jul 21 06:58:46 * XorA gets jealous of slug owners Jul 21 06:59:34 XorA: Thanks, I started downloading. So zecke, you do not need to tar it for me. I have send email from the svn anoymous access problem also to the berlios maintainer. Jul 21 07:00:53 victory! Jul 21 07:01:03 koen: sorry you lost Jul 21 07:03:34 florian_kc: ah. yes, that does sound like a bug. Jul 21 07:03:41 there shouldn't be any cross-architecture dependencies like that. Jul 21 07:04:32 pb_: oe_libinstall: if [ -f dotlai -o -n "$libtool" ]; Jul 21 07:04:42 pb_: shouldn't that be -a instead of -o? Jul 21 07:06:55 hrw|work: Do these images contain packages from the upgrades feed, or are they compiled from the 3.5.3 tree? Jul 21 07:07:15 zecke_: it does look that way, yes Jul 21 07:07:31 otherwise, the following "install" would surely fail if $dotlai did not exist Jul 21 07:07:38 pb_: Sounds like that dependency is added somewhere deep inside oe or bb, rigt? Jul 21 07:07:49 Cwiiis: 3.5.3+upgrades Jul 21 07:08:06 florian_kc: what's the actual dependency? Jul 21 07:10:07 pb_: Looks like all the "-dev" packages depend on gcc-cross-sdk (>= 3.3.4). Jul 21 07:10:48 and suprisingly it did fail ;) Jul 21 07:11:59 someone use libpcap < 0.9.3 in OE? Jul 21 07:12:34 uf. nearly nothing Jul 21 07:14:38 I put the do_install line into my bb file as suggested here but I got an error that might suggest a general problem with the bitbake infrastructure. Take a look at http://rafb.net/paste/results/oHcvSD17.html. What does it mean "global parse error not defined"? Jul 21 07:15:31 Laibsch: two issues Jul 21 07:15:52 I think you must write do_install() {\n body\n} Jul 21 07:16:07 OK Jul 21 07:16:08 and the bitbake error will be fixed later Jul 21 07:16:14 Cool. Jul 21 07:16:20 Thanks for helping me out. Jul 21 07:16:25 It would have raised a ParserError Jul 21 07:16:31 but it can not find that class Jul 21 07:16:43 This is due me moving the parser around Jul 21 07:16:46 OK, that was the vague impression I got Jul 21 07:17:37 ~lart openwrt people for creating oldstyle ipk Jul 21 07:17:37 * ibot beats openwrt people senseless with a 50lb Unix manual for creating oldstyle ipk Jul 21 07:19:47 hi Jul 21 07:20:17 hi CoreDump|home, Bernardo Jul 21 07:20:17 hi Jul 21 07:20:38 florian_kc: how odd. I wonder where that is coming from. Jul 21 07:21:15 Crofton_: ok, i think we're there Jul 21 07:22:43 pb_: yes that looks really strange... apart from this the whole sdk target built without any trouble from the scratch. Jul 21 07:23:36 thanks Jul 21 07:23:37 hrw|work: we should have a getsetuser.bbclass that takes care of adding/removing users in postinst/postrm Jul 21 07:23:38 03hrw 07org.oe.dev * r6d3ab1c2... 10/packages/libpcap/libpcap_0.9.3.bb: staged m4 macros which are needed by arpwatch to build Jul 21 07:23:42 03mickeyl 07org.oe.dev * r28c983cf... 10/packages/sdcc/ (sdcc-native_2.5.0.bb sdcc_2.5.0.bb): add sdcc - small systems compiler collection Jul 21 07:23:45 03mickeyl 07org.oe.dev * re0bd53b1... 10/packages/usrp/ (files/usb11.patch usrp_0.8.bb): Jul 21 07:23:46 add usrp - Support software for the Universal Software Radio Peripheral (USRP) Jul 21 07:23:46 Patch courtesy Philip Balister Jul 21 07:23:58 florian_kc: have a look in tmp/staging/arm-linux/shlibs, see if any of the .package files mention gcc-cross-sdk. Jul 21 07:24:10 does sdcc-native work? Jul 21 07:24:20 would I add a package that doesn't work? :D Jul 21 07:24:25 :) Jul 21 07:24:32 compiled fine here Jul 21 07:24:35 ok, I think I tried to make one and had some troubles Jul 21 07:24:37 thanks again! Jul 21 07:24:46 i had to override do_configure() though, because they have a really old configure.in Jul 21 07:24:51 ok Jul 21 07:24:52 that's probably where you had trouble with Jul 21 07:24:59 That sounds familiar Jul 21 07:25:23 ok, leaving to play guitar in church Jul 21 07:25:24 cu Jul 21 07:25:34 I actually made a configure.ac and some Makefile.am files so I should be better at figuring things like that out Jul 21 07:25:36 l8r Jul 21 07:26:58 bye mickey_away Jul 21 07:27:29 pb_: No file in that directory contains "gcc-cross-sdk". Jul 21 07:28:23 hrw|work: Just flashing now... So this patch, does it fix all the collie apm (multiple apps using apm_bios) issues, or just the 'holding right suspends'? Jul 21 07:28:56 pb_: but apart from this there is no file matching *.package in that dir. Jul 21 07:31:35 Cwiiis: 1 for rightarrow, 2 for apm Jul 21 07:32:32 * koen searches for a crosscable Jul 21 07:32:54 hrw|work: Seems the gpe-image doesn't work? Oddly, it boots into X almost instantly and calibration happens correctly, but then X gets stuck in an infinite restart loop Jul 21 07:33:27 Cwiiis: I'd say that's a job for the serial cable :) Jul 21 07:35:00 Cwiiis: ;( Jul 21 07:35:41 koen: :) Got to remember how to use it again first... Jul 21 07:38:36 re Jul 21 07:40:29 florian_kc: oh, hm. What files do you have in there? Jul 21 07:40:41 who is the maintaner of snes9x ? his name is not in the bb file... Jul 21 07:40:46 * alan_ monomaniac... Jul 21 07:40:53 pb_: .list, .pclist, .ver Jul 21 07:41:09 florian_kc: oh, yeah, I remember. er, do you have any files called "gcc-cross-sdk.*"? Jul 21 07:41:42 alan_: you might need to inspect the monotone history to identify the guilty party Jul 21 07:41:55 pb_: yes, i have a gcc-cross-sdk.ver and .list Jul 21 07:42:02 aha Jul 21 07:42:04 what's in the .list file? Jul 21 07:42:36 pb_: are you sure it was for me ? i don't understand... Jul 21 07:43:03 alan_: he means that if you look at the monotone history, you'll see who checked in the file Jul 21 07:43:44 that is indeed what I mean Jul 21 07:43:56 which would point back to me Jul 21 07:44:02 ok. thanks. ... erh... mmm... how can i say this... mmm... i don't know where to check this... i'm sorry... Jul 21 07:44:03 check the bk history instead Jul 21 07:44:13 so, now we lart koen and get on with our lives.. :) Jul 21 07:44:24 pb_: libgcc_s.so.1 \n libstdc++.so.5 \n libg2c.so.0 Jul 21 07:44:30 * koen points to http://oe-devel.bkbits.net:8080/openembedded/src/packages?nav=index.html|src/. Jul 21 07:44:33 CP|OLS: :P Jul 21 07:46:15 hrw|work: The opie image doesn't work for me either, same symptoms as the gpe image - nothing configures and things get stuck when the dm starts Jul 21 07:46:17 * CP|OLS grabs TWIN and starts playing Jul 21 07:46:21 florian_kc: ah, that's no good Jul 21 07:46:29 the -sdk packages should not be participating in shlibdeps. Jul 21 07:46:31 twin in OE? Jul 21 07:46:41 ggilbert: by any luck, are you here ? Jul 21 07:46:46 CP|OLS: what's that TWIN thing I keep hearing about Jul 21 07:46:58 ? Jul 21 07:47:01 keithp's windowing system Jul 21 07:47:10 TWIN as in http://keithp.com/blog//twin Jul 21 07:47:19 pb_: add EXCLUDE_FROM_SHLIBS=1 then Jul 21 07:47:27 heh Jul 21 07:47:54 its not very autotools happy Jul 21 07:48:00 Phill Ligget just announced the AMD facts of the day (during US Tour caoverage) Jul 21 07:48:01 we were looking for a windowing system to run on a DSP, possibly out of 200KB-500KB of ram Jul 21 07:48:10 and keith came up with TWIN Jul 21 07:48:13 and some guy on the side of the road was streaking :) Jul 21 07:48:24 small memory for extremely long battery life: weeks to months Jul 21 07:48:35 jamey: let me say this - it really obsoletes any other small windowing system Jul 21 07:48:58 what DSP? Jul 21 07:49:01 And thats really hard form me to say this - because I spent the better part of two years as a primary developer for Microwindows Jul 21 07:49:24 we started on TI C5509 but their software tools caused us to switch to Analog Devices Blackfin and uClinux Jul 21 07:49:31 heh Jul 21 07:49:34 florian_kc: I guess gcc-cross-sdk should be doing the same packaging stuff that the regular gcc-cross does. Jul 21 07:49:34 Not that Microwindows was never a contender, but still... :) Jul 21 07:49:40 s/never/ever/ Jul 21 07:49:45 now we have 16MB of sdram, but may still fall back to onboard SRAM for idle periods Jul 21 07:50:27 jamey: are the TI C5x tools that bad? I was just looking at some different DSPs for a software radio project we might be doing. Jul 21 07:51:14 the compiler etc is OK, we ran under wine on linux Jul 21 07:51:24 but we had to ignore all the source code we got from TI Jul 21 07:51:34 apparently there is a commandline c55 compiler native for linux Jul 21 07:51:40 and sizeof(char) == sizeof(short) caused code portability problems Jul 21 07:52:29 hrw|work: Looks like the images have been manually altered and packages have been installed, but the config status of the other packages was altered, so nothing configures Jul 21 07:52:31 I'm very happy with bf533 and gcc/binutils. Analog is doing a nice job funding development of open source. Jul 21 07:52:32 pb_, My advisor is http://www.mprg.org/people/facultybio.shtml#JHR Jul 21 07:52:59 Cwiiis: ok Jul 21 07:53:03 pb_: but if you need floating point I don't have any recommendations -- I only really studied fixed point options Jul 21 07:53:39 pb_: hmm... how is this related to that file? Jul 21 07:55:10 hrw|work: If you're installing upgrades manually, i think there's an ipkg switch to make sure that they remain unconfigured - If you perform a command and it doesn't return successfully, it used to be a bug that it would try and configure all the other packages, regardless of options (I think) - so caution needs to be used :) Jul 21 07:56:59 Cwiiis: I copied upgrades into deploy/ipk/ and then rebuild images Jul 21 07:57:34 hrw|work: Ah... Maybe ipkg version mismatch then? Otherwise maybe there's an error in the log? Jul 21 07:57:58 jamey: I think we will be using fixed point also. Jul 21 07:58:07 Cwiiis: will check Jul 21 08:03:15 Is monotone 0.21 ok to access OE repo? Jul 21 08:04:20 florian_kc: the files in the shlibs directory map shared library providers onto package names. Things like libgcc_s.so.1 should be in their own packages, with architecture set to "arm", not in the main gcc-cross package (which has architecture x86). Jul 21 08:04:43 Cwiiis: 0.20, 0.21 are ok Jul 21 08:05:18 thanks hrw|work Jul 21 08:05:41 pb_: ah okay... Jul 21 08:06:08 shit... my mouse cursor changed into "windowresize" one and now mouse does not react for buttons ;( Jul 21 08:06:09 florian_kc: gcc-cross.bb knows how to do that, but apparently gcc-cross-sdk lacks those bits Jul 21 08:06:21 it should be fairly straightforward to copy them across. Jul 21 08:07:48 pb_: okay, i'll try that... Jul 21 08:11:39 cya later Jul 21 08:13:12 re Jul 21 08:16:36 florian_kc: ok, very good Jul 21 08:24:25 uh... gcc bb files are quite some lines of code... Jul 21 08:28:52 ~lart university network admins Jul 21 08:29:03 hmm.. no ibot Jul 21 08:31:12 reenoo_: but yes... that's a good idea :-) Jul 21 08:33:22 florian_kc: I guess they're reinstalling all workstations *again* without prior notice. Jul 21 08:34:13 reenoo_: oops... here they force us to use their email server... Jul 21 08:35:59 reenoo_: If it even worked... Jul 21 08:36:48 heh.. yeah. the mail server at university is a pain as well here Jul 21 08:37:06 reenoo_: http://fionn.de/posting.txt Jul 21 08:39:13 * reenoo_ rotfl Jul 21 08:39:21 sounds familiar, yeah Jul 21 08:40:29 We started a discussion about making some unixag.org independend from universities. Jul 21 08:46:31 morning Jul 21 08:46:45 hey chouimat Jul 21 08:46:49 hey chouimat Jul 21 08:47:19 florian_kc: sounds like a plan Jul 21 08:49:22 * chouimat is back. Jul 21 08:50:23 bleh Jul 21 08:50:26 ~lart Infineon Keil and Hitex Jul 21 08:50:42 looks like I'll have to drive to university :/ Jul 21 09:03:37 cu Jul 21 09:04:14 * RP upgrades from monotone-0.19 :) Jul 21 09:09:55 The pull is going to take hours :-/ Jul 21 09:10:23 * reenoo_ nods Jul 21 09:11:47 so does monotone have python bindings? Jul 21 09:12:49 RP: maybe you can talk koen into publishing a current snapshot to avoid the initial pull Jul 21 09:19:08 reenoo_: I have an older snapshot from the monotone 0.19 days. I'll just wait for it to update :) Jul 21 09:24:32 I'm getting lots of problems with host libraries......The builds are trying to link with host system libraries instead of the ones in staging and its causing builds to fail. My latest package with this problem is pcre. I'm starting again from scratch to see if it's just an error I made, but I'm not so sure.... Jul 21 09:26:17 With monotone, do you no longer need to do an edit/unedit type sequence to modify files? Jul 21 09:26:54 Cwiiis: correct Jul 21 09:27:51 brilliant :) Jul 21 09:29:39 could someone pls do a mt log | grep coredump on a recent tree? Jul 21 09:36:47 ~lart sharp, embedix and all other people which made 2.4.18-collie Jul 21 09:37:26 when I plug usb cable - OOPS, when I plug hostap cf card - OOPS Jul 21 09:39:48 when I boot in a cradle (with usb connection and wifi inserted) - OOPS made collie a brick as build-in keyboard does not work Jul 21 09:39:51 fsck! Jul 21 09:40:15 hrw: I never had that... Jul 21 09:40:23 pb_: I have a hacked gcc-cross-sdk which might do the trick now, i'll rebuild from scratch now. Jul 21 09:41:12 reflashing with oz353-r21 Jul 21 09:43:05 florian_kc: rock Jul 21 09:44:19 pb_: well... it will take some time. Jul 21 09:44:58 03florian 07org.oe.dev * rf5dd78fd... 10/packages/linux/linux-omap1_2.6.12-rc2.bb: Move to newer gcc. Jul 21 09:45:02 03florian 07org.oe.dev * ra8d45d0b... 10/packages/gpe-filemanager/gpe-filemanager_0.23.bb: Add new release of gpe-filemanager. Jul 21 09:51:49 Cwiiis: hmm.. images are really borken - stuff was not configured at all.. so only bootstrap-image is quite working but not as it should ;( Jul 21 09:52:18 hrw: Yup :/ Did the log turn up anything? Jul 21 09:52:53 I haven't been able to make an opie-image or gpe-image for a while now due to various errors... Jul 21 09:53:10 gpe-image fails because it can't find libgcc1 for deps Jul 21 09:53:25 opie-image fails because pcre is trying to link against my host libraries Jul 21 09:53:36 I'm trying a fresh compile now to see if it was something I did.... Jul 21 09:53:54 Cwiiis: I put gpe-image do_rootfs log on web Jul 21 09:54:06 hrw: is it complaining about libgcc1? Jul 21 09:54:16 hrw: I'll have a look Jul 21 09:54:37 JustinP: it does not even have ipkg ;( Jul 21 09:54:46 now that's strange.... Jul 21 09:55:24 I fscked something in enviroment probably Jul 21 09:55:35 old OE + newest bitbake == not so good Jul 21 10:00:44 hrw: The log looks fine... "ipkg: Cannot create directory `/usr/lib/ipkg': Permission denied" is normal, isn't it? Maybe try creating an image without upgrades? Jul 21 10:01:28 yes, that's normal Jul 21 10:02:43 ggilbert: ping ? Jul 21 10:03:54 ok - time to go.. Jul 21 10:03:58 * koen points impatient people to http://ewi546.ewi.utwente.nl/OE/OE.db.bz2 Jul 21 10:04:03 cu Jul 21 10:07:01 * koen wishes people wouldn't treat the oesf forums as a bugtracker for OE Jul 21 10:10:25 lol Jul 21 10:11:11 btw, i have a tricky problem with the libsdl-mixer bb Jul 21 10:11:45 seems when libsdl gets provided by libsdl-qpe, the mixer bb fails to include the qt2 lib dir into its search path Jul 21 10:12:19 now, is it possible to adjust this somehow cleverly, because adjusting the search path statically wouldn't work Jul 21 10:12:30 as libsdl comes in 2 different fashions... Jul 21 10:13:00 ~lart QT/e Jul 21 10:13:32 i agree.... Jul 21 10:13:36 incinerator: that'd be a bug in libsdl-qpe. Jul 21 10:13:58 why? Jul 21 10:14:00 it probably needs to add the qt2 lib directory to its rpath. Jul 21 10:14:07 clear Jul 21 10:14:48 but how does that affect packages that build-depend on it? Jul 21 10:15:08 they shouldn't need to change. Jul 21 10:15:15 how can libsdl-mixer figure out libsdl is provided by the qpe flavoured bb and not the x11 one? Jul 21 10:15:27 it shouldn't need to care. Jul 21 10:15:32 ah ok Jul 21 10:15:48 the "upwards facing" interface on libsdl-qpe and libsdl-x11 should be identical. Jul 21 10:16:03 otherwise, they couldn't use the same soname. Jul 21 10:16:14 hmmm, i think it is not that simple Jul 21 10:16:30 you should probably check with Überhacker Zecke. I think he worked on libsdl-qpe most recently. Jul 21 10:16:41 the problem actually seems to be libopiecore2.so not linking because libqte2.so is missing Jul 21 10:17:02 * incinerator wishes bitbake/oe would be easier to grok Jul 21 10:17:12 k pb_, thx Jul 21 10:17:16 * koen blames qt/e, not OE Jul 21 10:17:38 sdl-x11 works, sdl-qte breaks, ergo qt/e stinks Jul 21 10:17:58 well, i haven't actually tried the x11 version yet... Jul 21 10:19:36 anyways, pb_ if you want to take a look at the compile log, its br #147 Jul 21 10:20:13 i'd appreciate any useful advice... Jul 21 10:25:55 dump opie Jul 21 10:26:00 or build opie for x11 Jul 21 10:26:17 bbl Jul 21 10:27:38 ho, someone is really having a bad day here ggg Jul 21 10:28:07 heh Jul 21 10:28:08 later all Jul 21 10:28:16 koen: build a familiar release that actually works with a c760 ;-) Jul 21 10:28:38 0.8.2 worked on my c700 Jul 21 10:28:38 if that uses x11 with opie, i won't mind ;-) Jul 21 10:28:54 i tried it, but had big troubles, kept crashing all the time Jul 21 10:29:11 ^c760|c860 Jul 21 10:33:19 incinerator: it uses sdl-config to report the libs Jul 21 10:33:31 incinerator: somehow it should provide -Lqtdir as well Jul 21 10:33:33 i see Jul 21 10:33:47 however, it doesn't. what could be the problem? Jul 21 10:38:30 new version of libsdl-mixer ;) Jul 21 10:38:45 ok, I'll try that one then.... Jul 21 10:38:52 koen: what was your hostname again? Jul 21 10:38:57 you mean 1.2.8? Jul 21 10:40:20 so... if I have TARGET_FPU = "soft" set, then the toolchain that is build will automatically 'do the right thing' with out me having to set a switch of tell it to use the floating point library? (this is if I want to use the toolchain outside of oe) Jul 21 10:40:24 zecke: ewi546.ewi.utwente.nl for the uni, or dominion.kabel.utwente.nl for home Jul 21 10:40:56 martin__: yes Jul 21 10:41:23 koen: I'm building oe on ewi now. OE is not close to run flawlessly on FreeBSD yet Jul 21 10:42:31 zecke: back in the olden days I ran multiple OE builds on that machine concurrently Jul 21 10:42:51 so with your super memreduction changes it should be even easier :) Jul 21 10:44:16 it ate a ton of RAM, true, but parsing was (IMO) way faster. It's a bouble-edged sword :) Jul 21 10:44:47 CoreDump|home: with 2GB of ram I wasn't complaining :) Jul 21 10:45:10 * CoreDump|home got 2GB, too ;) Jul 21 10:46:47 I wish I had too Jul 21 10:46:53 * CoreDump|home waits for parsing to finish Jul 21 10:47:04 in 5 minutes or so Jul 21 10:47:19 CoreDump|home: I strongly doubt parsing was any faster Jul 21 10:47:46 i believe it was. Initial parsing for sure. Jul 21 10:48:28 CoreDump|home: now way. seriously Jul 21 10:48:36 i swear :) Jul 21 10:48:44 CoreDump|home: also the building provider hash step was completely removed Jul 21 10:48:55 CoreDump|home: and the parsing is unchanged Jul 21 10:49:01 so it can not be slower Jul 21 10:49:37 It needs about 8 to 10 seconds for one (!) percent of initial parsing now Jul 21 10:49:55 no way Jul 21 10:50:03 on a P4 2.8 HT that is Jul 21 10:50:05 CoreDump|home: turn of atime on your XFS Jul 21 10:50:12 ohh Jul 21 10:50:29 CoreDump|home: At ROAD I've virtually the same computer Jul 21 10:50:57 /dev/md2 on /mnt/raid0 type xfs (rw,noatime) Jul 21 10:51:03 it's already off Jul 21 10:51:36 CoreDump|home: seriously it can not be slower. At least on my systems it is faster than ever Jul 21 10:51:46 and this with +800 more package files to parser Jul 21 10:51:56 heh, dunno why, but surely not for me Jul 21 10:52:08 koen: opie for x11? which package would that be? Jul 21 10:52:30 CoreDump|home: show me numbers ;) Jul 21 10:52:42 JustinP: change the defconfig, rebuild opie Jul 21 10:52:49 zecke: heh ok, I'll time it Jul 21 10:52:50 bitbake -s, same packages Jul 21 10:52:58 -s ? Jul 21 10:53:09 CoreDump|home: just to list the providers Jul 21 10:53:31 k Jul 21 10:53:38 * zecke is tired ;) Jul 21 10:54:12 koen: which defconfig? Jul 21 10:54:37 anyone have any idea what UDEV_DEVFS_RULES does? I found it in familiar.conf Jul 21 10:54:41 +mhentges@oe-akita-hentges:..ke/openembedded/packages >touch ../conf/local.conf Jul 21 10:54:41 +mhentges@oe-akita-hentges:..ke/openembedded/packages >time bitbake -s opie-imag Jul 21 10:55:43 CoreDump|home: nothing after -s ;) Jul 21 10:55:54 oops Jul 21 10:56:06 martin__: indeed - DEVFS actually does not RULE Jul 21 10:56:15 koen: do you have a monotone db on ewi? Jul 21 10:58:00 CP|Laptop: what does the variable do though? Jul 21 11:00:03 zecke: /data/build/oe/OE.db Jul 21 11:00:27 zecke: or /var/www/OE for a one without privkeys Jul 21 11:01:35 koen: I see qt-x11-free in the qt directory but I'm not sure how to make opie apps compile with it Jul 21 11:01:57 (I'm not even sure that's the right package) Jul 21 11:03:20 JustinP: you'll have to ask a opie hacker Jul 21 11:03:24 * zecke checks if ewi is safe against fork bombs Jul 21 11:03:26 I only the option exists Jul 21 11:03:54 +know Jul 21 11:04:50 JustinP: What do you want to do? Jul 21 11:06:39 zecke: fwiw the db in /data is the one that is being serve via netsync Jul 21 11:06:47 someone mentioned above that you can compile "opie for x11". I'm currently running Opie and GPE on the same machine in order to use the variosu programs on both (and because parts of both are crashing on me). Jul 21 11:07:03 It would be nice to have my Opie stuff running in GPE somehow, is this possible? Jul 21 11:07:25 it is Jul 21 11:07:26 koen: I used the one from /var/www Jul 21 11:07:36 but you'll have to nuke trolltech first Jul 21 11:07:50 JustinP: as said on #opie if you're willing to spend time on it Jul 21 11:07:57 JustinP: it is easy, no rocket science involved Jul 21 11:07:59 to stop them from whining about compatibilty and other crap Jul 21 11:08:39 zecke: well I'm currently trying to get various other things working as well...if you could point me towards the relevant information / bb file I can add it to my list of things to check out. Jul 21 11:09:10 zecke: real 12m30.549s Jul 21 11:09:10 user 11m1.420s Jul 21 11:09:10 sys 0m22.311s Jul 21 11:09:16 lol Jul 21 11:09:39 JustinP: read my answer on #opie first Jul 21 11:22:50 * chouimat is away: lunch Jul 21 11:28:58 hmm Jul 21 11:29:07 mt -db=./db pull monotone.vanilla.de "org.openembedded.*" Jul 21 11:29:07 monotone: misuse: monotone pull doesn't use the option --branch Jul 21 11:29:24 gah Jul 21 11:29:45 hmm same error for vanille Jul 21 11:30:06 vanille looks down again Jul 21 11:30:13 isn't it --db? Jul 21 11:30:14 koen: I've changed OVERRIDES_prepend Jul 21 11:30:22 CP|Laptop: good catch Jul 21 11:30:51 bother old eyes Jul 21 11:30:58 it is right on the web page :( Jul 21 11:31:20 humans are so stupid ;) Jul 21 11:31:28 no one asks them to do fail... Jul 21 11:32:37 Crofton_: thats ok - I was trying to pull from org.embedded the other day Jul 21 11:32:44 you can imagine how well *that* worked Jul 21 11:32:54 heh Jul 21 11:32:59 I regulary try vanille.monotone.de Jul 21 11:33:54 re Jul 21 11:33:59 wb Jul 21 11:34:06 CoreDump|home: I will time 'bitbake -s' on my system Jul 21 11:34:35 hrw: thanks Jul 21 11:34:49 btw - which version of bitbake we used at OZ 3.5.3 time? Jul 21 11:35:06 hrw: bitbake 1.0.1 Jul 21 11:36:01 zecke: oooold... Jul 21 11:36:20 hrw: or 1.1.x Jul 21 11:36:20 hmm.. kmail does not like koen gpg key.. Jul 21 11:37:42 time to patch bitbake with 'unknown author' patch Jul 21 11:38:22 hrw: wait ;) Jul 21 11:38:32 hrw: No I've enough time to fix it ;) Jul 21 11:38:44 hrw: could you pastebin one more time Jul 21 11:39:06 patch? Jul 21 11:39:18 hrw: no error Jul 21 11:39:45 no the error Jul 21 11:39:49 zecke: currently I want to try something to getworking images... Jul 21 11:39:57 CoreDump|home: look: Jul 21 11:39:59 real 5m38.607s Jul 21 11:39:59 user 4m33.197s Jul 21 11:39:59 sys 0m9.937s Jul 21 11:40:07 doh Jul 21 11:40:12 CoreDump|home: 1.8GHz AthlonXP reiserfs3 Jul 21 11:40:13 half the time Jul 21 11:40:22 /dev/hda5 on /home/hrw/devel/build type reiserfs (rw) Jul 21 11:40:25 zecke: is h3*.conf broken now? Jul 21 11:40:27 damn Jul 21 11:40:39 CoreDump|home: intel suxx Jul 21 11:40:44 heh Jul 21 11:40:45 brb Jul 21 11:41:04 koen: I will find out Jul 21 11:41:20 koen: but overrides looked right Jul 21 11:41:38 hmm Jul 21 11:42:02 I understand the bitching about monotone now Jul 21 11:42:14 that stats have stopped .... Jul 21 11:44:06 koen: building h3900, fam-0.8.3 on ewi now Jul 21 11:44:14 ok Jul 21 11:49:48 wb pb_ Jul 21 11:51:30 hail zecke Jul 21 11:53:31 hi all. Jul 21 11:58:48 re Jul 21 11:59:58 Which package contains startx and other scripts? thanks. Jul 21 12:23:24 hmm.. bitbake 1.0... memory-eater... Jul 21 12:23:42 bitbake 1.0 - death eater Jul 21 12:24:44 70% - 400M Jul 21 12:27:17 and failed.. I will probably drop 3.5.3 support Jul 21 12:30:11 is there a way w/ mt to vie the history of a single file? Jul 21 12:30:55 i mean w/ diff-style changelog as on bkbits.net Jul 21 12:31:00 CoreDump|home: monotone log [file] Jul 21 12:31:24 *yawn* Jul 21 12:31:42 CoreDump|home: when will you push your changes? :D Jul 21 12:31:52 i though i already had Jul 21 12:32:00 no, nothing from you in my log Jul 21 12:32:07 try again Jul 21 12:32:11 let me view the log Jul 21 12:32:14 uhoh Jul 21 12:32:22 it is in my log :) Jul 21 12:32:29 ~lart monotone Jul 21 12:32:34 that just means you commited it Jul 21 12:32:38 doesn't mean you pushed Jul 21 12:32:49 i did a "sync" afterward as by the howto Jul 21 12:33:14 Pushing your changes upstream: Jul 21 12:33:14 monotone --db=OE.db sync Jul 21 12:33:19 what's your host address ? Jul 21 12:33:35 *netcolgne.net Jul 21 12:33:50 sure? Jul 21 12:33:55 or .de Jul 21 12:33:56 heh Jul 21 12:33:59 CoreDump|home: could you try "monotone --db=OE.db sync ewi546.ewi.utwente.nl "org.openembedded.*" ? Jul 21 12:34:04 hmm netcolgne seems wrong Jul 21 12:34:09 will try Jul 21 12:34:10 is your spelling correct? Jul 21 12:34:20 mickeyl: err netcologne Jul 21 12:34:26 aha Jul 21 12:34:32 no connection from *netcologne* here Jul 21 12:34:42 might have pushed to koen =) Jul 21 12:34:43 looks like it didn't push, just pull Jul 21 12:34:45 hah Jul 21 12:34:55 don't rely on default arguments Jul 21 12:35:01 CoreDump|home: that would end up in vanille.de too Jul 21 12:35:17 I sync all 4 known repos Jul 21 12:35:23 i have a tail -f on the log file. just push again and we'll see Jul 21 12:35:57 monotone: allowed 'coredump@openembedded.org' write permission for 'org.openembedded.*' excluding '' Jul 21 12:36:06 yep Jul 21 12:36:21 crap Jul 21 12:36:26 tons of unmet deps Jul 21 12:36:28 mickeyl / koen - Is it ok to send you a key for monotone in an hour or two? (when it finishes pulling :p) Jul 21 12:36:44 Cwiiis: sure Jul 21 12:36:50 optimistic Jul 21 12:36:51 thanks Jul 21 12:37:09 koen: did my commit make it? Jul 21 12:37:15 CoreDump|home: nope Jul 21 12:37:19 Cwiiis: http://ewi546.ewi.utwente.nl/OE/ Jul 21 12:37:20 sigh Jul 21 12:37:31 looks like i need to pull first Jul 21 12:37:33 CoreDump|home: could you try again? Jul 21 12:37:48 l Jul 21 12:37:49 k Jul 21 12:38:03 koen: Well I'm half-way through now, seems a waste :) Jul 21 12:38:18 monotone pull;monotone merge;monotone sync Jul 21 12:38:22 I'm in no hurry Jul 21 12:38:34 koen: looks better now Jul 21 12:38:49 yeah. pull, merge, sync, update Jul 21 12:38:51 monotone: verifying new revisions (this may take a while) Jul 21 12:39:00 yep Jul 21 12:39:11 mickeyl: pull merge update merge sync Jul 21 12:39:43 Got a problem with a bitbake world run: Jul 21 12:40:01 | `sh cflags "optimize='-O2'" pp_sys.o` pp_sys.c Jul 21 12:40:01 | CCCMD = gcc -DPERL_CORE -c -D_REENTRANT -D_GNU_SOURCE -DTHREADS_HAVE_PIDS -fno-strict-aliasing -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/gdbm -O2 -Wall Jul 21 12:40:01 | pp_sys.c: In function `Perl_pp_shostent': Jul 21 12:40:01 | pp_sys.c:5054: warning: implicit declaration of function `sethostent_r' Jul 21 12:40:01 | pp_sys.c:5054: error: structure has no member named `_hostent_data' Jul 21 12:40:03 | pp_sys.c: In function `Perl_pp_sprotoent': Jul 21 12:40:03 this is going to take a while Jul 21 12:40:05 | pp_sys.c:5076: warning: implicit declaration of function `setprotoent_r' Jul 21 12:40:07 | pp_sys.c:5076: error: structure has no member named `_protoent_data' Jul 21 12:40:09 | pp_sys.c: In function `Perl_pp_sservent': Jul 21 12:40:11 | pp_sys.c:5087: warning: implicit declaration of function `setservent_r' Jul 21 12:40:13 | pp_sys.c:5087: error: structure has no member named `_servent_data' Jul 21 12:40:15 | pp_sys.c: In function `Perl_pp_ehostent': Jul 21 12:40:16 it's recommended to pull updated, commit and merge before syncing Jul 21 12:40:17 | pp_sys.c:5098: warning: implicit declaration of function `endhostent_r' Jul 21 12:40:21 | pp_sys.c:5098: error: structure has no member named `_hostent_data' Jul 21 12:40:23 | pp_sys.c: In function `Perl_pp_eprotoent': Jul 21 12:40:25 Svartalf: you lost Jul 21 12:40:25 | pp_sys.c:5122: warning: implicit declaration of function `endprotoent_r' Jul 21 12:40:27 | pp_sys.c:5122: error: structure has no member named `_protoent_data' Jul 21 12:40:29 Svartalf: use pastebin Jul 21 12:40:29 | pp_sys.c: In function `Perl_pp_eservent': Jul 21 12:40:31 eeek Jul 21 12:40:31 | pp_sys.c:5134: warning: implicit declaration of function `endservent_r' Jul 21 12:40:31 uuh Jul 21 12:40:32 /ignore Svartalf Jul 21 12:40:33 | pp_sys.c:5134: error: structure has no member named `_servent_data' Jul 21 12:40:35 | make: *** [pp_sys.o] Error 1 Jul 21 12:40:37 | FATAL: oe_runmake failed Jul 21 12:40:39 NOTE: Task failed: /home/frank/oe_build/tmp/work/perl-native-5.8.4-r0/temp/log.do_compile.10514 Jul 21 12:40:39 gah Jul 21 12:40:41 N Jul 21 12:40:42 ugh Jul 21 12:40:43 What might I be doing wrong here? Jul 21 12:40:44 eww Jul 21 12:40:46 * zecke wished he could kick+ban Svartalf Jul 21 12:40:50 you're flooding the room for one Jul 21 12:40:58 Svartalf: flooding the channel for axample Jul 21 12:41:06 a) use pastebin b) world is supposed to fail, since we are not enough people to maintain all our 2500 packages Jul 21 12:41:13 use -k Jul 21 12:41:16 Nice attitude guys... That's not a flood. Jul 21 12:41:23 * CoreDump|home thinks it is Jul 21 12:41:25 believe me - that was a flood Jul 21 12:41:25 And I didn't know about pastebin. Jul 21 12:41:38 anythng more than 10 lines is a flood Jul 21 12:41:44 * Svartalf sighs... Jul 21 12:42:03 IIRC -k ignores failing packages and continues as much as possible Jul 21 12:42:10 that makes much sense when building world Jul 21 12:42:15 mickeyl: correct Jul 21 12:42:27 mickeyl, THANK YOU for being much more helpful than the rest. Jul 21 12:42:35 ~users Jul 21 12:42:40 someone built world recently? Jul 21 12:42:41 no ibot here Jul 21 12:42:47 hrw: no Jul 21 12:42:48 hrw: last time when 3.5.3 was due Jul 21 12:42:54 and svartgalf is on my ignore list Jul 21 12:42:55 was a painful experience Jul 21 12:43:20 since for some things you really need to obey a build order Jul 21 12:43:31 and those bugs are hard to find Jul 21 12:43:35 koen, Be that way. You might end up finding that it was a poor decision on your part. Attitude gets you everywhere- and you've got a BAD one right at the moment. Jul 21 12:43:38 it's not just adding a dependency here and there Jul 21 12:43:56 mickeyl: I was thinking about not doing world builds for angstrom each release, but updating the feeds regularly (and only keeping the ipkgs needed for images generation in release feeds) Jul 21 12:43:56 Svartalf: I disagree, and I'm also ignoring you now Jul 21 12:44:14 koen: yeah, we might have to do it like that Jul 21 12:44:20 svartalf whats your porblem? Jul 21 12:44:52 koen: needed for images generation and atleast full gpe, opie builds Jul 21 12:45:00 What problem? I make a simple mistake on manners within a given IRC channel and nearly EVERYBODY gets their noses bent out of shape and says "you're on ignore". Jul 21 12:45:19 svartalf funny Jul 21 12:45:23 Its because when we called attention to your mistake, you got all angry at us Jul 21 12:45:28 Svartalf: next time ASK before flood can you flood Jul 21 12:45:38 I didn't get angry until you all did... Jul 21 12:45:38 hrw: the sum of all stuff built by doing bitbake bootstrop,gpe,opie image for all models Jul 21 12:45:47 zecke Svartalf: you lost Jul 21 12:45:47 koen Svartalf: use pastebin Jul 21 12:45:52 zecke /ignore Svartalf Jul 21 12:45:57 * zecke wished he could kick+ban Svartalf Jul 21 12:46:01 koen Svartalf: flooding the channel for axample Jul 21 12:46:15 koen: meta-opie stripped to only opie > opie-image for all models Jul 21 12:46:18 oh my god, the troll cames out Jul 21 12:46:20 /ignore Svartalf Jul 21 12:46:23 Guys, that's not me getting mad, that's YOU all getting that way. Jul 21 12:46:32 woglinde: to like me: /ignore him Jul 21 12:46:33 cu Svartalf in next reincarnation Jul 21 12:46:41 * CP|Laptop looks around for kergoth Jul 21 12:46:53 I'm ignoring him but please kick him anyway Jul 21 12:46:58 heh Jul 21 12:47:53 so when can i review coredumps changeset? :) Jul 21 12:47:58 is it in ewi yet? Jul 21 12:48:31 I think so Jul 21 12:48:36 * koen tries to merge heads Jul 21 12:49:17 monotone: misuse: merge of 'packages/initscripts/initscripts_1.0.bb' : 'd15c30774c688bd36f5dbe83109b7fcf6c02dc8b' -> 'e07a16080d51c317554d405a7019124420d459d4' vs '641bcea11da9dfa7c641e9bea5f34ba8e9850bae' failed Jul 21 12:49:19 I'm happy when not have to merge in monotone Jul 21 12:49:27 Do you have to do an mt co after every pull? Jul 21 12:49:54 Cwiiis: no, only mt update Jul 21 12:49:57 god damn it Jul 21 12:50:15 koen: Thanks :) Will gen a key... Jul 21 12:50:21 still verifying btw... Jul 21 12:50:59 ok done Jul 21 12:50:59 CoreDump|home: can you do 'monotone merge' ? Jul 21 12:51:02 koen: why is ewi so slow ;) Jul 21 12:51:10 there was no error message on my side btw Jul 21 12:51:13 yeah Jul 21 12:51:13 koen: Where it says xxx@openembedded.org on the wiki page, is 'xxx' meant to be a nickname or something? And in that case, does it have to be @openembedded.org? Jul 21 12:51:22 koen: apache runs amock Jul 21 12:51:24 zecke: if (zecke) {sleep 1 ; continure} Jul 21 12:51:40 Cwiiis: it has to be @openembedded Jul 21 12:51:44 CoreDump|home: monotone allows you some more things than bk, i.e. push before pull, commit without updating your working copy, etc. You need to define some aliases to make sure you don't create unwanted forks. Jul 21 12:51:45 k Jul 21 12:52:32 even with the aliases, presumably there is no certainty of not creating forks. Jul 21 12:52:36 well, i followed the howto on oe.org to the letter the first time :) Jul 21 12:52:47 just a lower probability of doing so Jul 21 12:53:02 merge failed btw Jul 21 12:54:36 hrw: can you assist andrea_80, he's wanting to add an italian tutorial to oz.org and is having problems with registering an account Jul 21 12:54:57 ah. I forgot to send emails for >50 users.. Jul 21 12:55:02 d'oh :) Jul 21 12:55:23 i take it that there is no "resolve" option in mt? Jul 21 12:55:46 merge is resolve Jul 21 12:55:52 if it doesn't work, the user is called Jul 21 12:55:57 do you have 'meld' installed? Jul 21 12:55:59 if not, do it Jul 21 12:56:00 :) Jul 21 12:56:03 will do Jul 21 12:56:13 meld is not bad. it needs a bit polish though Jul 21 12:57:29 well wait, can't I just discard the change to initscripts_1.0.bb and keep the rest? Jul 21 12:58:24 meld is python-gtk and needs also some gnome libs ;( anyone replaced meld with kdiff3? Jul 21 12:58:37 ah, looks like i can "uncommit" everything, good Jul 21 12:59:06 hrw: or alternatively, I can poke my eyes with a fork.. :) Jul 21 13:00:03 hey jamey_ Jul 21 13:00:15 this positively sucks Jul 21 13:00:37 hrw: shouldn't be hard - if you know the syntax just change a hook Jul 21 13:00:57 mickeyl: I mostly use monotone remotelly in shell anyway Jul 21 13:01:07 hrw: same here :) Jul 21 13:03:08 ok. now I will try to write tool to generate passwords for oz.org users Jul 21 13:03:46 Who is upgrading wordpress on oz.org? Jul 21 13:05:06 hey monotone released again Jul 21 13:05:07 hi zecke Jul 21 13:05:38 i think we are all using 0.21 now Jul 21 13:05:44 most of us, at least Jul 21 13:06:21 still some people didn't grok 'org.openembedded' => '"org.openembedded.*"' as I can see in my server log ... Jul 21 13:06:47 * hrw has 0.20 as debian is not yet on 0.21 Jul 21 13:07:05 zecke: want to upgrade? Jul 21 13:08:03 hrw: wasn't there a security advisory on wordpress? or just regular update? Jul 21 13:08:31 zecke: too much things to do I have so I miss it Jul 21 13:08:47 is there a way to get monotone to tell me what files it is going to update before ti actually updates them? Jul 21 13:08:48 anyone tried DVB-T on a PDA yet ? Jul 21 13:09:25 mickeyl: not DVB-H? Jul 21 13:09:35 no Jul 21 13:09:39 mickeyl: How do you get DVB-T on a PDA? Jul 21 13:09:46 zecke: usb Jul 21 13:09:57 i'm just wondering if the CPU will be fast enough Jul 21 13:10:02 mickeyl: I suspect decoding of the stream takes too long Jul 21 13:10:39 I figure a stock PXA is at the limit even on DVB-H Jul 21 13:10:57 http://www.qanu.de/termini.php Jul 21 13:11:33 pxa27x might be ok with dvb-t, just about. Jul 21 13:12:25 and, of course, on some handhelds you can offload colourspace conversion, iDCT, and/or motion compensation to the graphics hardware Jul 21 13:12:45 hmm yeah Jul 21 13:14:23 you'd need extra hardware to do the COFDM demod, and probably the mpeg demuxing as well, though Jul 21 13:18:26 heh - can't roam away from the AP and expect it stay up, now can I? Jul 21 13:18:43 heh Jul 21 13:18:56 OLS wouldn't be the same if the wireless actually worked Jul 21 13:21:20 I am proud to donate my lag to the legacy Jul 21 13:21:48 very public spirited of you Jul 21 13:25:03 is a monotone pull always necessary before an update? Jul 21 13:25:11 koen i tried another sync Jul 21 13:26:25 hmm D'OH! Jul 21 13:26:27 http://tree.celinuxforum.org/CelfPubWiki/ITJ2005Detail1_2d4?action=AttachFile&do=get&target=OSK-OpenEmbedded.pdf Jul 21 13:26:37 I would really like to be notified about things like that Jul 21 13:26:48 not stumbling over them by a random web search Jul 21 13:27:05 * mickeyl adds to presentations and paper site Jul 21 13:28:01 thats celinux for you Jul 21 13:28:12 * CP|OLS is going to a celinux BOFS in a few hours Jul 21 13:28:34 it's cool that they mention us in a presentation, but it would be cooler if they'd actually told us Jul 21 13:28:34 heh Jul 21 13:28:45 heh, yeah Jul 21 13:28:51 we should get back in touch with Tim Bird Jul 21 13:28:54 I hope Steven Johnson wasn't somebody we insulted for flooding the channel Jul 21 13:28:55 I would feel bad Jul 21 13:28:56 bleh, ill try again tomorrow Jul 21 13:28:56 getting started and hands on material is always necessary Jul 21 13:28:59 n8 all Jul 21 13:29:17 CoreDump|afk: wait Jul 21 13:29:20 CoreDump|afk: did it work this time? Jul 21 13:29:31 CoreDump|afk: if not, please send me the changes per mail and I'll have a go at integrating them Jul 21 13:29:49 CoreDump|afk: g'night Jul 21 13:29:49 i didn't work - i think Jul 21 13:30:09 no error message, I've merged and updated. Jul 21 13:30:32 guys, i have a question i'm sure you'll have the answer : does micro-emacs have autocompletion on latex files or does it need plugins ? Jul 21 13:33:04 mickeyl: mail is away Jul 21 13:33:09 CoreDump|afk:k, thanks Jul 21 13:33:18 np, n8 Jul 21 13:33:46 n8 Jul 21 13:34:06 Which package contains startx and other scripts for x11? thanks. Jul 21 13:34:52 I don't think startx is packaged. Jul 21 13:35:07 a good idea actually Jul 21 13:35:07 If you wanted to package it up, xserver-xorg would be the place to start. Jul 21 13:35:09 g'night Jul 21 13:35:18 mickeyl: night night Jul 21 13:35:58 mickey|zzZZzz: nice doc it is Jul 21 13:36:44 xserver-xorg is support ati? for c760? Jul 21 13:37:23 no Jul 21 13:37:35 it does contain those scripts, though Jul 21 13:37:43 ok Jul 21 13:37:44 or, at least, the xorg source tree contains them Jul 21 13:38:32 03coredump 07org.oe.dev * ra6f1ef5b... 10/ (4 files in 4 dirs): Akita: Add a fixed keymap, credits to jh@oesf.org Jul 21 13:39:02 CoreDump|afk: grats Jul 21 13:43:51 can't ewi get another cpu? Jul 21 13:44:07 03florian 07org.oe.dev * r10c91c11... 10/packages/sylpheed/sylpheed_2.0.0beta6.bb: Disable as long as it doesn't compile properly. Jul 21 13:49:26 Cwiiis: key added Jul 21 13:50:02 hmm... "Xati" says "Fatal server error: No matching cards found!" on c760... Jul 21 13:50:10 koen: where should a h3900 build fail? Jul 21 13:50:18 hrw: what breakage with bitbake do you see? Jul 21 13:50:42 bye all ! Jul 21 13:50:50 zecke: it was oz 3.5.3 with current bitbake - as its OE was too old I had a problem Jul 21 13:50:57 zecke: bitbake virtual/kernel Jul 21 13:51:15 koen: at least handhelds-pxa provides that? Jul 21 13:51:37 I think so Jul 21 13:51:51 It has always worked, till bitbake broke about a week ago Jul 21 13:52:31 modutils-cross-2.4.27: ... Jul 21 13:52:40 gcc-cross-kernel-3.3.4-3.3.4: Jul 21 13:53:23 wow, kdump appears to rawk Jul 21 13:53:31 at least for x86 Jul 21 13:55:20 handhelds-pxa-2.4.19-rmk6-pxa1-hh40: Jul 21 13:55:38 koen: /home/zecke/builds/ipaq/conf/local.conf Jul 21 13:56:58 bah, mt branch org.openembedded.dev has 2 heads again... Jul 21 13:58:10 hey zecke, you about? Jul 21 13:58:20 * zecke checks Jul 21 13:58:25 mithro: yes, looks like Jul 21 13:58:47 zecke: you wouldn't happen to have harlekin's telephone details? Jul 21 13:58:56 he sent me an email but forgot to include them Jul 21 13:59:04 monotone.vanille.de should change its name to hydra.vanille.de ;-) Jul 21 13:59:30 mithro: no not since he moved Jul 21 13:59:41 mithro: and he isn't online Jul 21 14:00:08 dang, hope he replies late tonight or early tommorrow Jul 21 14:00:15 we where suppose to catch up after his work tommorrow Jul 21 14:00:55 mithro: drop me your number, if he pops up on ICQ/AIM I'll ask him to call you Jul 21 14:01:06 i don't have a number :/ Jul 21 14:01:18 hence the problem :P Jul 21 14:02:24 if he comes online tell him to check his email and then send me one pack Jul 21 14:02:35 i'll check around breakfast time and lunch time Jul 21 14:03:14 coredump: is that keymap update something like the one in the OESF forums? if so does is it being applied to the spitz as well? Jul 21 14:03:19 woot Jul 21 14:03:37 Hey. My tel number is +49xxxxxxxx. Writing this on my phone Jul 21 14:04:27 yay for the connected world Jul 21 14:05:02 ps anyone know where I could get a 802.11xx card with antenna connectors and linux support? Jul 21 14:05:17 incinerator: ewi546.ewi.utwente.nl/tmp/viewmtn/headofbranch.psp?branch=org.openembedded.dev show only one branch Jul 21 14:05:23 argh....who branched org.openembedded.dev? Jul 21 14:05:39 coredump, florian: one of you merge those heads Jul 21 14:06:09 florian_kc: you love to fork? ;) Jul 21 14:06:14 koen: monotone.vanille.de doesn't Jul 21 14:06:20 there is no spoon Jul 21 14:07:11 koen: and for me http://ewi546.ewi.utwente.nl/tmp/viewmtn/headofbranch.psp?branch=org.openembedded.dev does show two heads Jul 21 14:07:18 it does now, you mean Jul 21 14:07:24 coredump: yes, that looks like the keymap from the forums. It also applies just fine to the Spitz (tested by me). Jul 21 14:07:30 you didn't specify a time Jul 21 14:08:36 well, i think I'll call it a day, svn.berlios.de doesn't let me svn, oe/bitbake doesn't seem to work properly on my box, very frustrating :-( Jul 21 14:09:07 and the hydra is about, too Jul 21 14:09:57 thanks koen Jul 21 14:11:16 koen: please check local.conf but the kernel builds just fine ;) Jul 21 14:11:28 CP|OLS: i've been thinking about your offer stuff, if you want to discuss it more just send me an email Jul 21 14:11:52 ~hail ibot Jul 21 14:12:01 * ibot bows down to ibot and chants, "I'M NOT WORTHY!!" Jul 21 14:12:08 ~hail ~hail Jul 21 14:12:19 * ibot bows down to ~hail and chants, "I'M NOT WORTHY!!" Jul 21 14:15:37 hrw: Does the kernel in current oe have the apm patches? Jul 21 14:15:39 koen: kernel completed to built Jul 21 14:15:51 Cwiiis: yes Jul 21 14:15:52 zecke: good Jul 21 14:15:57 good stuff :) Jul 21 14:16:15 koen: but it was the jlime one, I hope that does not matter (just kidding) Jul 21 14:19:41 How are uclibc builds fairing these days? Jul 21 14:21:23 C760 contains ATI Imageon 100? Jul 21 14:21:29 ah, I see the problem with the divergence Jul 21 14:21:31 Cwiiis: that depends on odd vs. even days Jul 21 14:22:08 zecke: I'm tempted to try it out, but I have a feeling it'd be a bad move... Jul 21 14:22:55 CoreDump|afk merged 92eb0d6f08f5ac3bdb8b69280bd69145bd2b42ab and 4954b87346baa782592647d49aced9350906ce0, florian merged 10c91c11403b2d2d4fa391bee1e6f60bfdf75c08 and f4954b87346baa782592647d49aced9350906ce0 Jul 21 14:23:09 that creates a loop when trying to merge Jul 21 14:23:25 Cwiiis: check the bug tracer, there are some useful patches Jul 21 14:23:51 that's why monotone says: monotone: misuse: merge of 'packages/libpcap/libpcap_0.9.3.bb' : '46f444579982b58d28a82e71a8e7a908bd0a674c' vs. '763b8478be56b0a412ec922f2a53da47b3cc4442' (no common ancestor) failed Jul 21 14:23:57 dmx_: yes Jul 21 14:24:21 mithro: thank you, and which xserver i must to use? Jul 21 14:24:38 03freyther 07org.oe.dev * r0a407ab1... 10/conf/machine/ (h3900.conf ipaq-pxa270.conf): (log message trimmed) Jul 21 14:24:38 conf/machine/h3900.conf: Jul 21 14:24:38 conf/machine/ipaq-pxa270.conf: Jul 21 14:24:38 On irc it was discussed if OVERRIDES_prepend is a legal statement Jul 21 14:24:38 and if OERRIDES should be applied here too. With runtime Jul 21 14:24:39 performance in mind (bb.data.update_data being called until no change Jul 21 14:24:41 happens) we have declared the above statement as illegal. Jul 21 14:24:43 dmx_: the one the MACHINE.conf mentiones Jul 21 14:24:43 03freyther 07org.oe.dev * ra7ddc196... 10/classes/base.bbclass: (log message trimmed) Jul 21 14:24:45 classes/base.bbclass: Jul 21 14:24:46 zecke: Thanks Jul 21 14:24:47 Change oe_libinstall (watch carefully): Jul 21 14:24:49 If the user requires the installation of a static library and it Jul 21 14:24:51 does not exist we want the user to notice and fail. I assume this Jul 21 14:24:53 was then intended purpose for the requirestatic variable. Jul 21 14:24:55 The matter for the dotlai (.lai) is a bit different. The system Jul 21 14:25:27 ~lart zecke Jul 21 14:25:27 * ibot explains, ever so gently, that if zecke doesn't give the channel more information, they can't help Jul 21 14:25:30 sorry Jul 21 14:25:45 I did not intend to create another head Jul 21 14:26:04 zecke: see my monologue above Jul 21 14:26:13 how do we approach from here Jul 21 14:26:18 koen: I followed it Jul 21 14:26:33 1) fix monotone's algo Jul 21 14:26:40 2) backout some cset Jul 21 14:26:47 3) go to sleep Jul 21 14:26:54 I choose 3) Jul 21 14:27:00 * chouimat is back. Jul 21 14:28:36 anyway got to run Jul 21 14:28:37 see ya Jul 21 14:28:54 zecke: i'm going on the 23 to Fussen Jul 21 14:28:55 hrw: http://wordpress.org/development/2005/05/security-update/ Jul 21 14:29:04 24th i head to Berlin Jul 21 14:29:14 three days left Jul 21 14:29:27 zecke: one head less Jul 21 14:29:46 so I should be there till the 26th most probably Jul 21 14:29:57 so either the 24th or the 25th are the best nights to meet up i guess Jul 21 14:30:27 mithro: do you know where you are going to stay? Jul 21 14:30:48 not yet Jul 21 14:30:53 any recommendations? Jul 21 14:31:06 koen: sorry but i not found information about X in c7x0.conf and zaurus-clamshell-XX.conf... Jul 21 14:31:32 koen: kdrive? which? Xati? Jul 21 14:31:57 conf/machine/zaurus-clamshell.conf:4:PREFERRED_PROVIDER_xserver = "xserver-kdrive" Jul 21 14:32:09 mithro: no I'm considering to ask my mother if we can host you :} Jul 21 14:32:21 mithro: but we're not too close to the center Jul 21 14:32:29 zecke: fix applied. thx Jul 21 14:32:38 hrw: wow thank you Jul 21 14:32:45 shit! now there are 3 heads Jul 21 14:32:48 koen: yes, but ati/epson/etc ? Jul 21 14:32:55 fbdev Jul 21 14:33:03 ~lart people who don't check to see if they created a fork Jul 21 14:33:03 * ibot tries to shut people who don't check to see if they created a fork up Jul 21 14:33:11 zecke: that would be cool, but i'm not fussed either way Jul 21 14:33:24 JustinP: I only see 2 heads Jul 21 14:33:35 http://monotone.vanille.de/viewmtn/headofbranch.psp?branch=org.openembedded.dev Jul 21 14:33:40 JustinP: and the heads were created by 2 merges Jul 21 14:33:42 freyther just created another one Jul 21 14:33:53 koen: fbdev? hm... And how ati imageon? Jul 21 14:33:54 yeah stupid freyther Jul 21 14:33:56 that only shows twoo Jul 21 14:34:04 two* Jul 21 14:34:10 it *was* 3 when I posted it... Jul 21 14:34:22 mithro: I will ask one for cheap/not to expensive hotels/hostels Jul 21 14:34:38 is there anyway to speed up a monotone pull? Jul 21 14:34:39 mithro: I know we've a bigger inexpsensive directly at the Bahnhof Zoo Jul 21 14:34:55 JustinP: see my explanation above (scroll back 15 minutes or so) Jul 21 14:35:04 martin__: use another db Jul 21 14:35:17 koen: a curcular merge, yes? both devs merged and comitted? Jul 21 14:35:18 martin__: for example koens 'daily' monotone db Jul 21 14:35:34 ugg, all these train trips want me to get up at 5am :/ Jul 21 14:35:43 JustinP: that's the short story Jul 21 14:36:05 * zecke tries the FreeBSD power on mt Jul 21 14:36:39 koen: so it won't *allow* merging of these 2 branches? how about if you manually merge the files? Jul 21 14:36:52 ]ahh ha! i had the time set wrong Jul 21 14:37:10 JustinP: it can't find an ancestor for one file Jul 21 14:37:22 koen: I got a "crash" a little while ago when I tried to merge 2 trees and just checked the diferences and merged them manually Jul 21 14:37:38 koen: are there actual differences in the 2 versions of the files? Jul 21 14:37:42 zecke: where do I find that one? Jul 21 14:37:55 perhaps a propogate? (I don't know if you can propogate between revisions...) Jul 21 14:37:55 i wish there was a way to do this the other way around Jul 21 14:38:03 ie tell them what time you want to be in a place Jul 21 14:38:29 Anyone else having trouble connecting to svn.berlios.de? Is there a mirror/recent tarball up somewhere? Jul 21 14:38:51 I'm merging now Jul 21 14:39:33 monotone: misuse: branch 'org.openembedded.dev' is merged Jul 21 14:39:56 syncing now... Jul 21 14:39:58 so which Berlin station do I most probably want to go in to? Jul 21 14:40:22 http://monotone.vanille.de/viewmtn/revision.psp?id=eea3d1cfa60979c91e14fa30d7c9644d58fba40b Jul 21 14:40:34 mithro: Berlin Zoo is one of the two centers Berlin has Jul 21 14:40:35 mithro: and the western part Jul 21 14:40:52 Berlin Zoologischer Garten or Berlin Ostbahnhof Jul 21 14:41:19 mithro: I would say it is easier to get to hostels/hotels from Zoologischer Garten Jul 21 14:41:28 okay then I'll end up there :P Jul 21 14:41:35 Cwiiis: I had trouble an hour or two ago Jul 21 14:41:37 looks like if I leave at 10am on the 24th Jul 21 14:41:41 install -m 0644 acinclude.m4 ${STAGING_DATADIR}/aclocal/libpcap.m4 Jul 21 14:41:48 its good that monotone use kdiff3 by default.. Jul 21 14:41:55 then, Berlin Zoologischer Garten 24.07.05 arr 18:17 Jul 21 14:41:57 hrw: hehe it doesn't Jul 21 14:41:59 martin__: But ok now? Jul 21 14:42:03 with only 2 changes Jul 21 14:42:05 Cwiiis: have not tried Jul 21 14:42:07 zecke: here it use it Jul 21 14:42:13 1 in Augsburg Jul 21 14:42:23 and 1 in Göttingen Jul 21 14:42:28 zecke: I dont even have .monotone/monotonerc Jul 21 14:42:28 hrw: it calls a lua hook and it searches for xxdiff, kdiff3, vim Jul 21 14:42:30 Cwiiis: svn: Can't connect to host 'svn.berlios.de': Connection refused Jul 21 14:42:35 zecke: thats the train I'm planning to catch atm :P Jul 21 14:42:41 anyway really got to run onw Jul 21 14:42:46 s/onw/now Jul 21 14:42:47 see ya! Jul 21 14:42:48 zecke: I dont have xxdiff, but have kdiff3 and vim Jul 21 14:42:50 mithro: do you know hostels.com? Jul 21 14:43:09 koen: maybe xorg? :) i want xkb for cyrillic keymap. Or no? I ask advice, please. Jul 21 14:43:17 i've been using a bunch of places Jul 21 14:43:22 martin__: Same :/ Jul 21 14:43:37 hostelbookers.com, hotel.de, hihostels.com and a few others Jul 21 14:43:40 hrw: function get_preferred_merge2_command(tbl) Jul 21 14:43:49 mithro-gone: cya Jul 21 14:43:52 dmx_: I told you: xserver-kdrive-fbdev Jul 21 14:44:04 if you don't want to take my advice go someplace else Jul 21 14:44:15 hrw, zecke: it also uses "meld" which has very nice merge "graphics" Jul 21 14:44:22 pb_: please review the base.bbclass change if it makes sense (and yes Qt is broken) Jul 21 14:44:39 hrw: from the default hook documentation page Jul 21 14:44:42 hrw: :-) thanks for fixing it Jul 21 14:44:56 koen: ok, ok. thank you. Jul 21 14:45:09 hrw: it uses kdiff3, xxdiff, emacs, xemcas, vim, gvim, meld Jul 21 14:45:40 hrw: thanks for merging Jul 21 14:45:40 monotone is supposed to be able to handle any kind of graph so what you hit was probably a bug in monotone....if you still have a copy of that DB you could send it to the monotone devs Jul 21 14:46:06 * JustinP goes back to fixing this stupid query Jul 21 14:46:08 JustinP: meld needs half of gnome libs (atleast in debian) so it is no-no for me Jul 21 14:46:21 ah Jul 21 14:46:27 apart from that, meld and kdiff3 should be able to merge it pretty easy Jul 21 14:46:37 JustinP: didn't sound like any of the devs was actually interested on monotone-devel Jul 21 14:46:44 koen: and it was easy Jul 21 14:46:55 have you synced it yet? Jul 21 14:46:57 koen: I should more often use merging tools Jul 21 14:47:05 koen: http://monotone.vanille.de/viewmtn/revision.psp?id=eea3d1cfa60979c91e14fa30d7c9644d58fba40b Jul 21 14:47:05 reenoo_: huh? Jul 21 14:47:28 reenoo_: oh yeah, you mean the "copy" feature request? Jul 21 14:47:33 JustinP: still haven't got a reply to my message in which I reported the same error Jul 21 14:47:35 koen: it was pull/merge/update/merge/sync command Jul 21 14:47:38 JustinP: no Jul 21 14:48:00 reenoo_: I got a reply from my bug report when I had a merge error. Jul 21 14:48:02 hrw: still two head Jul 21 14:48:32 JustinP: that was just another user. not a developer. Jul 21 14:48:51 shit Jul 21 14:49:09 JustinP: I'm referring to this message btw http://lists.gnu.org/archive/html/monotone-devel/2005-07/msg00122.html Jul 21 14:50:48 reenoo_: I posted to their bug system my own error Jul 21 14:51:13 again libpcap merge.. Jul 21 14:51:22 zecke: which change is that? Jul 21 14:51:29 03koen 07org.oe.dev * re2b4fef7... 10/packages/libpcap/libpcap_0.9.3.bb: disapproval of revision 6d3ab1c248e51c9e183169b8ec8cd0fee592080c Jul 21 14:51:33 03koen 07org.oe.dev * r6da21d0d... 10/packages/sylpheed/sylpheed_2.0.0beta6.bb: disapproval of revision 10c91c11403b2d2d4fa391bee1e6f60bfdf75c08 Jul 21 14:51:36 03koen 07org.oe.dev * r15e035df... 10/packages/sylpheed/sylpheed_2.0.0beta6.bb: disapproval of revision 6da21d0d95b412989c3883c587af4931afea6741 Jul 21 14:51:56 pb_: -s/-o/-a/ on .lai installation Jul 21 14:52:00 note that the last two were the *same* change according to monotone Jul 21 14:52:15 JustinP: aha. any reaction to that? Jul 21 14:52:28 zecke: ah, right, that one. yes, I think that patch is correct. Jul 21 14:52:37 reenoo_: well someone said they had a similar problem and to send a database file if I could Jul 21 14:52:38 koen: you are synced with mickeyl? or should I also pull from you? Jul 21 14:52:50 I'm synced with mickeyl Jul 21 14:52:52 pb_: It still fails when requirestatic was set but the .a is not present Jul 21 14:52:53 at least hourly Jul 21 14:53:07 pb_: but it won't fail when it thinks it sees libtool but .lai is missing Jul 21 14:53:12 koen: I merged 5e488367b644497740e81d030ed34e9925440107 with eea3d1cfa60979c91e14fa30d7c9644d58fba40b Jul 21 14:53:26 zecke: right Jul 21 14:53:55 syncing... Jul 21 14:54:03 Would anyone care to tar up a recent subversion? Svn is down at the mo and I'd like to start a build before bed (if anyone has the spare time :)) Jul 21 14:54:11 bugger, I'm tired... Jul 21 14:54:13 I meant bitbake :) Jul 21 14:54:43 http://monotone.vanille.de/viewmtn/revision.psp?id=4fe4838753ca7f23c9801afa9ecc94acdb3f3e2f - merged finally Jul 21 14:55:08 looks like it worked Jul 21 14:55:12 http://ewi546.ewi.utwente.nl/tmp/viewmtn/headofbranch.psp?branch=org.openembedded.dev Jul 21 14:55:41 yep Jul 21 14:55:41 * koen looks into adding a SSH X tunnel between these machines Jul 21 14:56:10 hrw: I sync vanille, ewi, dominion and nslu2-linux 4 times a day Jul 21 14:56:30 and vanille <> ewi hourly Jul 21 14:56:52 koen: I pull/merge vanille from time to time and pull/merge/sync after commit Jul 21 14:57:12 now all we need is a 'tracker' to distribute traffic between the dbs Jul 21 14:57:24 JustinP: URL? Jul 21 14:57:44 pull commit pull merge creates less problems Jul 21 14:58:05 ok - will try to remember Jul 21 14:58:30 or maybe will try to hook it Jul 21 14:58:59 bed time Jul 21 14:59:03 'night all Jul 21 14:59:26 this initial pull is taking forever Jul 21 15:00:10 ok. time to sleep for me as well Jul 21 15:00:33 I dont remember when was last time I was on 00:00 on irc ;) Jul 21 15:00:53 hrw|gone: sleep well Jul 21 15:05:17 martin__: initial pulls take a while. If you use the snapshot mentioned earlier it should be faster. Jul 21 15:05:23 reenoo_: can't find it :-| Jul 21 15:05:45 argh....sylpheed fails now...it compiled before that gnutls patch was applied Jul 21 15:07:18 If sylpheed compiles two out of three times you're doing good Jul 21 15:09:14 JustinP: no one linked to the snapshot Jul 21 15:11:42 martin__: for me it is /var/www/OE/OE.db.bz2 that should relate to Jul 21 15:12:28 http://ewi546.ewi.utwente.nl/OE/OE.db.bz2 Jul 21 15:12:32 but I'm not too sure Jul 21 15:14:46 ok. so I just grab that, and pretend it is my db? Jul 21 15:16:37 03zecke123 * r289 10bitbake/lib/bb/parse/parse_py/ConfHandler.py: Jul 21 15:16:37 bitbake/parse/parse_py/ConfHandler.py Jul 21 15:16:37 -ParseError was unknown to ConfHandler since the moving Jul 21 15:16:37 of the directories. Import ParseError into ConfHandler Jul 21 15:16:51 JustinP: that should fix your error from earlier today? Jul 21 15:19:45 03zecke123 * r290 10bitbake/lib/bb/fetch.py: Jul 21 15:19:45 bitbake/lib/bb/fetch.py: Jul 21 15:19:45 -Create a MD5SumException. We will throw it when the sums do Jul 21 15:19:45 not match. Jul 21 15:20:46 03zecke123 * r291 10bitbake/lib/bb/fetch.py: Jul 21 15:20:46 bitbake/lib/bb/fetch.py: Jul 21 15:20:46 -Remove decodeurl comment/testcase. We will readd to the right Jul 21 15:20:46 place to be a proper testcase. Jul 21 15:24:32 Cwiis: I had same problem earlier today but somebody gave me the bitbake 1.3.1 today from wget -c http://hyboria.mine.nu/~dp/bitbake-1.3.1.tar.gz Jul 21 15:24:56 ah, thanks lamikr :) Jul 21 15:25:57 Well it is not in my page but hope it works for you :-) I am still editing local.conf... Jul 21 15:38:43 zecke: I'm not sure what error you're referring to.... Jul 21 15:39:15 ParseError, but I'm bad in remembering names Jul 21 15:39:43 I don't think it was Justin Jul 21 15:39:53 koen|sleep: BTW: Thanks for the account at ewi Jul 21 15:41:59 03zecke123 * r292 10bitbake/lib/bb/fetch.py: Jul 21 15:41:59 bitbake/lib/bb/fetch.py: Jul 21 15:41:59 -Prepare the Wget fetcher to soon check the MD5SUMs in many Jul 21 15:41:59 different ways. Jul 21 15:42:26 good nite Jul 21 15:48:58 argh!!multiple heads again Jul 21 15:49:22 this is real annoying as I also have to merge with a local head.....I don't know which freaking file is the right one to merge! Jul 21 15:49:44 ~lart florian Jul 21 15:49:44 * ibot beats florian over the head with a microkernel Jul 21 15:51:55 JustinP: looking at the timestamps mickeyl is the problem Jul 21 15:51:58 2205 vs 2219 Jul 21 15:53:42 rwhitby: http://dominion.kabel.utwente.nl/koen/blog/pyblosxom.cgi/Handhelds/slug.html Jul 21 15:55:13 koen|sleep: nice :-) Jul 21 15:56:10 Our plans for nslu2-linux world domination are coming along nicely .... that's the OE core devs done - now I just have to talk to Linus about buying him a slug ..... ;-) Jul 21 15:56:36 oops - did I say that out aloud? Jul 21 15:57:36 koen|sleep: the best news is that you already have all the nslu2-linux data in your local db, so you don't need to wait for another initial monotone pull ... Jul 21 15:57:53 :) Jul 21 15:58:08 * koen|sleep lazy and downloaded openslug binary Jul 21 15:58:54 koen|sleep: wait for a day or two and get the 2.3 binary - we have an much better upgrade process for the future in that one which saves your config files across flashes Jul 21 15:59:14 ok, will do Jul 21 15:59:30 as I'll need to buy a crosscable and a serial kit anyway Jul 21 15:59:40 (ignoring the real work I have to do) Jul 21 15:59:44 or just use the Master Makefile at http://www.nslu2-linux.org/Makefile, and build your own binary. Jul 21 16:00:47 mkdir -p /home/slug ; cd /home/slug ; wget http://www.nslu2-linux.org/Makefile ; make setup ; make openslug Jul 21 16:01:28 you can symlink your current monotone db into the monotone subdir and not have to wait Jul 21 16:01:51 nice Jul 21 16:02:02 rwhitby: very good Jul 21 16:02:25 * koen|sleep waits for the sync to complete before atempting to sleep again Jul 21 16:02:55 pb_: thx - we have a motto in nslu2-linux: "If it's not in a repo, then it doesn't exist." and "If it's not done with a Makefile, then it doesn't get done." Jul 21 16:03:02 rwhitby: md5 checking for sources is coming along Jul 21 16:03:33 see http://cia.navi.cx/stats/project/openembedded/.message/6094176 Jul 21 16:05:07 koen|sleep: in your blog you used the term "salted slug". Here is the original salted slug: http://www.nslu2-linux.org/wiki/uploads/NSLU2-board-front-bare.jpg (warning 800Kb picture) Jul 21 16:05:25 that's the result of taking a heat gun to a slug to remove all the components destructively :-) Jul 21 16:06:27 ~lart mickeyl, then Jul 21 16:06:27 * ibot breaks out the Hoover and sucks up mickeyl, then Jul 21 16:06:52 rwhitby: heh. why would you want to do that? Jul 21 16:09:13 * JustinP rescinds that Jul 21 16:09:54 it's not really mickey's fault. florian is working from older revisions. florian should pull before syncing Jul 21 16:10:40 rwhitby: nice ehm... artwork Jul 21 16:10:43 There were multiple heads when florian syned his current head Jul 21 16:13:28 reenoo_: so we could track the traces Jul 21 16:14:07 how do you think we found out enough information to be able to determine how to de-underclock it :-) Jul 21 16:15:15 heh. ok. I guess the nslu2 is inexpensive enough for that approach Jul 21 16:15:50 the community chipped in to replace it Jul 21 16:17:01 it wasn't functional anyway. something about [g2] needing a remedial soldering course .... Jul 21 16:17:17 :) Jul 21 16:21:56 Don't suppose anyone knows why 'ipkg --help' outputs to standard error rather than output? Jul 21 16:23:18 koen|sleep: excuse me, what i must typing for xserver-kdrive-fvdev? bitbake xserver-kdrive-fvdev? Jul 21 16:24:07 bitbake xserver-kdrive or bitbake virtual/xserver Jul 21 16:24:40 koen|sleep: ok, thank you. Jul 21 16:31:15 Cwiiis: Yeah, I also noticed that. I hate help messages on stderr.... Jul 21 16:31:21 ipkg --help 2>&1 | less Jul 21 16:31:43 This does not make me trust monotone's 3-way merge very much: http://article.gmane.org/gmane.comp.version-control.monotone.devel/3264 Jul 21 16:31:46 JustinP: Yup, but that gets annoying on a pda - nevermind that the output is almost unreadable as it is :) Jul 21 16:32:26 Cwiiis: I completely agree Jul 21 16:33:32 JustinP: I'm sure it's a 3 letter fix (err->out :p), I'll see about contributing a patch tomorrow (and maybe a further patch for more sane output) Jul 21 16:35:12 Cwiiis: I'd appreciate it at least. :-) Even if no one else puts it in OE I'll put it in my copy. :-) Jul 21 16:37:21 JustinP: pigi is very accepting of help with ipkg, I'm sure it'll get accepted unless there's a good reason :) Jul 21 16:38:10 Cwiiis: Well let me know about it so that I can at least test the patch locally Jul 21 16:38:18 JustinP: Sure thing Jul 21 16:42:05 good thing I read docs religiously: http://webuser.fh-furtwangen.de/~dersch/gphoto/remote.html Jul 21 16:42:12 libusb ARM patch Jul 21 16:45:08 argh: | /usr/lib/libc.a: could not read symbols: File format not recognized Jul 21 16:45:08 | collect2: ld returned 1 exit status Jul 21 16:45:08 | arm-linux-libtool: install: error: relink `libpcreposix.la' with the above command before installing it Jul 21 16:47:10 'night all Jul 21 16:47:21 'night reenoo_ Jul 21 16:47:24 night Jul 21 19:18:38 * france is away: Away Jul 21 19:18:56 * france is back (gone 00:00:02) Jul 21 20:05:32 fuck me....now there are 3 heads! at least on ewi there are Jul 21 20:06:40 now I have a 4 head merge....::sigh:: Jul 21 20:08:30 sounds like a serious personal problem Jul 21 20:23:29 CP|Laptop: well since monoton can't even get past the first damn merge..... Jul 21 20:24:09 I'm seriously annoyed by peoples' lack of merging. It makes work especially hard for those of us who don't have sync access who always have to merge at least once when updating anyway Jul 21 21:59:11 hi Jul 21 22:50:32 * france is away: Away Jul 21 23:01:44 hi all Jul 21 23:03:46 monotone.vanille.de is down again, btw Jul 21 23:59:35 monotone can read in packets from stdio (read: mail contents) Jul 21 23:59:53 - can I "eavesdrop" a sync operation, extract the packets sent to the server and relay them via mail? **** ENDING LOGGING AT Thu Jul 21 23:59:57 2005