**** BEGIN LOGGING AT Fri Feb 17 10:59:56 2006 Feb 17 11:05:28 looks like I found hosting for my needs - but first have to wait for answers to my questions.. Feb 17 11:13:31 ahah, checkroot goodness on my / finally Feb 17 11:28:57 morning all Feb 17 11:29:15 hrw|work: good news that rmk is taking an interest in pxa mmc again :) Feb 17 11:33:04 morning Richard Feb 17 11:33:17 hey RP Feb 17 11:34:13 RP: yes Feb 17 11:34:33 RP: bad news is that first attempt fails Feb 17 11:34:59 hrw|work: I saw that :-/ Feb 17 11:35:07 hrw|work: early days I guess Feb 17 11:35:20 hrw|work: I also have a card I can test with this Feb 17 11:36:15 good Feb 17 11:41:39 <_guillermo> JustinP: is the e-image-core supposed to be working, I still get the entraced problem Feb 17 11:47:51 I'm having trouble building glibc (2.3.5, 2.3.3, 2.3.2) on the prebreakage revision. Is this a common problem or am I doing something terribly wrong? Feb 17 11:48:36 zekoZeko: What DISTRO and MACHINE settings are you using? Feb 17 11:50:13 I didn't set that, I'm just trying to build something basic for my Soekris board.. It's geode based Feb 17 11:50:30 that's i486-class Feb 17 12:03:26 03koen 07org.oe.oz354fam083 * r96349b5a... 10/packages/abiword/ (4 files in 4 dirs): abiword: remove some old versions Feb 17 12:03:31 03koen 07org.oe.oz354fam083 * r1cc2d5a7... 10/packages/abiword/ (7 files in 2 dirs): abiword: add 2.4.2 and remove old versions Feb 17 12:08:53 zekoZeko: I know the epia is x86 based and overrides certain things to make x86 work properly Feb 17 12:09:11 zekoZeko: You might want to look at the epia overrides in glibc Feb 17 12:09:28 morning Feb 17 12:09:38 hi reenoo Feb 17 12:09:46 hey RP Feb 17 12:10:29 just in conf/machine/epia.bb ? Feb 17 12:10:43 there's just the kernel provider set. Feb 17 12:10:49 nothing abotu glibc. Feb 17 12:10:50 heh.. 215 revs to pull Feb 17 12:10:56 will check packages/glibc too Feb 17 12:11:41 nope, grep -r got nothing about epia Feb 17 12:12:46 I'm trying another build now, DISTRO="nylon" and MACHINE="geodegx" Feb 17 12:12:55 let's see what happens Feb 17 12:12:57 zekoZeko: Sorry, I was under the impression epia did override things. It looks like it sets a couple of GLIBC_xxx options in conf/machine/epia.conf Feb 17 12:13:41 Those GLIBC lines might help Feb 17 12:14:08 yeah, x86 builds of recent glibcs won't work without Feb 17 12:16:54 OK, I'll try with those two lines. Feb 17 12:17:33 RP: won't work without GLIBC_ADDONS = "nptl" and GLIBC_EXTRA_OECONF = "--with-tls" Feb 17 12:17:34 ? Feb 17 12:18:03 sorry Feb 17 12:18:06 reenoo Feb 17 12:18:12 zekoZeko: I would use DISTRO="generic" for start Feb 17 12:18:16 03koen 07org.oe.oz354fam083 * r2aaffb27... 10/packages/glibc/glibc_2.3.5+cvs20050627.bb: glibc: unbump PR to save innocent users from having to update glibc for no reason Feb 17 12:20:07 If I change local.conf while in bitbake shell, is it possible to re-parse only that file? Feb 17 12:20:25 or any other, for that matter? Feb 17 12:20:32 try 'reparse ' Feb 17 12:22:54 NOTE: package glibc-initial-2.3.2+cvs20040726-r17: task do_compile: completed Feb 17 12:22:57 so far so good :) Feb 17 12:23:47 I need a faster build machine Feb 17 12:40:17 03koen 07org.oe.oz354fam083 * r442bd7f2... 10/packages/libgsf/ (5 files): libgsf: update to 1.13.3 and remove old versions Feb 17 12:43:39 pb__: master? Feb 17 12:47:48 http://pastebin.com/559408 Feb 17 12:48:21 Something fails fo parse in packages/familiar it seems Feb 17 13:22:01 RP: ping Feb 17 13:26:14 zecke: pong Feb 17 13:32:29 RP: did you ever consider adding iwmmxt support to qemu? :} Feb 17 13:32:48 RP: there is no 'generic' coprocessor support in target-arm yet, right? Feb 17 13:33:11 zecke: I'd probably be more interested in adding pxa system support ;-) Feb 17 13:33:41 I've looked more at the system-arm rather than target-arm Feb 17 13:34:00 damn, glibc fails to build with any of the options I've tried Feb 17 13:34:03 http://pastebin.com/559463 Feb 17 13:35:57 zekoZeko: 2.3.5 is probably more used that earlier version in OE Feb 17 13:36:41 I'm in need of executing a iwmmxt binary in nightly builds Feb 17 13:37:02 zecke: sbrsh? Feb 17 13:37:10 zecke: tricky... Feb 17 13:37:48 koen: well, the devices are locked at night Feb 17 13:38:35 RP: yes, but that failed too. Feb 17 13:38:42 RP: although with a different error message Feb 17 13:38:53 I'll start with a fresh local.conf Feb 17 13:39:01 I'll see then Feb 17 13:39:34 ~lart safecat author Feb 17 13:39:34 * ibot raises middle finger to safecat author Feb 17 13:39:34 do I have remove the old local.conf or can i rename it to something so it doesn't get parsed? Feb 17 13:39:35 zekoZeko: Add those two GLIBC lines from epia.conf to local.conf Feb 17 13:39:42 RP: I did it Feb 17 13:39:44 no help Feb 17 13:39:46 zekoZeko: just rename Feb 17 13:39:49 OK Feb 17 13:39:59 I'll be back in an hal an hour or so Feb 17 13:40:06 half an hour even Feb 17 13:42:46 RP: koen : initial pull on localhost with monotone took real: 122m8, user 113m33 sys:1m1 Feb 17 13:43:55 I'm slowly reaching the point where I can't use monotone as I can't pull/update fast enough to get my commits in before someone also commits :-/ Feb 17 13:44:16 The problem being I run pull but have to do something else whilst I wait for it to finish... Feb 17 13:46:15 so you commit and merge Feb 17 13:46:18 what's the problem? Feb 17 13:47:08 Its just too slow to allow me to work with any kind of flow with it Feb 17 13:47:40 * zecke just had to issue hg recover on the server again :} Feb 17 13:47:40 your pulls last 2 hours? Feb 17 13:47:57 the last ~20 commit had an average of 2 hours between them Feb 17 13:48:15 RP: git scares me as it moves so fast Feb 17 13:48:17 Admittedly I was getting annoyed a few days ago, probably during a busy spell Feb 17 13:48:41 RP: as git doesn't do locking it doesn't have that problem, right? Feb 17 13:48:48 it would just mess up the tree ;) Feb 17 13:48:58 doesn't have locking? Feb 17 13:49:18 if it would have locking the packingbug wouldn't exist Feb 17 13:49:28 which packing bug? Feb 17 13:49:48 run git pack while someone pull -> carnage Feb 17 13:50:24 Ah, that issue, The repo doesn't lose data and can be fixed Feb 17 13:50:44 anything can be fixed Feb 17 13:50:48 and it doesn't have the same issue on push, just pull Feb 17 13:51:05 I dare say they will fix that issue no its been discovered Feb 17 13:51:10 yeah Feb 17 13:51:19 and I need to upgrade git again to remain compatible Feb 17 13:51:35 reminds me of another scm Feb 17 13:52:11 At least you won't have to destroy all your local trees to fit in with the upgrade ;-) Feb 17 13:52:33 yeah Feb 17 13:52:51 it's just that git has exactly the same problems as monotone due to being under development Feb 17 13:53:01 * koen wants bitkeeper back Feb 17 13:53:17 ** unknown exception encountered, details follow Feb 17 13:53:18 ** report bug details to mercurial@selenic.com Feb 17 13:53:22 The difference is that git is remaining compatibility and is in active use - monotone isn't Feb 17 13:53:29 a wrong ownership Feb 17 13:53:39 RP: no it isn't compatible Feb 17 13:53:51 http://www.freelists.org/archives/linux-cirrus/02-2006/msg00017.html Feb 17 13:54:06 ' it's pretty much mandatory Feb 17 13:54:07 to update your git every week or two' Feb 17 13:55:35 * RP shrugs. Its leagues apart from the kind of breakage in monotone Feb 17 13:55:53 but it is breakage Feb 17 13:56:14 every scm in the test has broken with simple operations Feb 17 13:56:18 *every* scm Feb 17 13:56:34 svn hasn't to my knowledge Feb 17 13:56:53 "broken" and "simple operations" being defined as what? Feb 17 13:56:54 svk blew up my /home Feb 17 13:57:08 RP: pulling, editing, patching Feb 17 13:57:55 reenoo: svn broke with every new bdb version when I used it in the past Feb 17 13:58:01 koen: git blew up for you? Feb 17 13:58:10 koen: don't use the bdb backend Feb 17 13:58:28 the oe svn repo uses FSFS Feb 17 13:58:42 reenoo: it wasn't my server, I just suffered as a user Feb 17 13:58:53 koen: the point being? Feb 17 13:58:53 03koen 07org.oe.oz354fam083 * r2683b878... 10/packages/gnome/ (4 files): goffice: update to 0.2 and remove old versions Feb 17 13:59:59 koen: There is no perfect solution scm solution and we will have to compromise. I think we should pick the one which is likely to have any imperfections solved in the future. Feb 17 14:00:13 koen: svk is not using $HOME/.svk/ - only if you ask it to Feb 17 14:00:52 RP: I don't get the 'git' merging... I have merged things already, but... Feb 17 14:00:53 zecke: it did over here, without me typing /home anywhere Feb 17 14:01:07 koen: how did you initialize the depot? Feb 17 14:01:12 RP: I'd vote for something that works _now_. otherwise we're in the same situation as with monotone again. Feb 17 14:01:32 bitkeeper! Feb 17 14:02:08 I'll use bk if someone pays ;) Feb 17 14:02:11 reenoo: I'd argue git does work now. Ok, its been shown you need to keep the program updated but is that really a major issue? Feb 17 14:02:27 zecke: There were some interesting comments about merging on the git mailing list recently. I think they're changing the default behaviour to something sane Feb 17 14:02:39 RP: that depends how fast the new versions land in unstable/universe... Feb 17 14:02:52 heh Feb 17 14:02:58 am I dreaming? Feb 17 14:03:05 maybe Feb 17 14:03:13 people want to ditch monotone because you need to keep it up to data so 'often' Feb 17 14:03:27 so yes, it is a major issue Feb 17 14:03:46 RP: git lacks documentation. without that it seems rather useless. I mean, do we want to stay the "too eleet to use an SCM joe user will grok" project forever? Feb 17 14:04:10 reenoo: oe is also useless by that count Feb 17 14:04:34 RP: correct Feb 17 14:05:12 RP: except it will take a bit longer to write an OE manual than to switch to say svn which has excellent docs. Feb 17 14:05:17 http://oe.pastebin.com/559502 <- even if everything worked Feb 17 14:05:38 svn isn't distributed Feb 17 14:05:47 writing scm docs is not a core competency of oe devs. Feb 17 14:06:14 * Crofton sounds managerial Feb 17 14:07:28 Anyhow, my only point is I can't use monotone as things stand. As we increase in number of files, its getting too slow. If OE sticks with monotone, I can see me committing less and less Feb 17 14:07:45 koen: well. who's able to use distributed features right now? we're merging stuff between branches manually. people had to nuke their local trees or migrate changes manually. the idea is nice and all, but it doesn't seem we need it all that much. and, fwiw, there's always svk which works well here. Feb 17 14:08:10 reenoo: people didn't had too Feb 17 14:09:33 koen: I know that you're very fond of monotone. but I'm afraid there's a point where you have to face reality. Feb 17 14:10:11 reenoo: really, people didn't had too Feb 17 14:10:27 actually monotone is getting better and better Feb 17 14:11:10 I actually really like the monotone UI. I just can't use it due to its lack of speed Feb 17 14:11:11 but I see the reasons not to use monotone Feb 17 14:11:41 so git and svn/svk are my favorites. And if I feel lucky I would use git Feb 17 14:11:52 but svn/svk is more solid (and a bit slower) Feb 17 14:12:28 zecke: and fragile Feb 17 14:12:31 It says something that the rest of openedhand use OE via our own svn copy of it Feb 17 14:12:45 zecke: I had to restart svk pull a few times Feb 17 14:12:51 I'm the only person in the company who will touch monotone... Feb 17 14:12:55 this is my current config: Feb 17 14:12:58 MACHINE = "geodegx" Feb 17 14:13:03 TARGET_OS = "linux" Feb 17 14:13:07 DISTRO = "generic" Feb 17 14:13:15 GLIBC_ADDONS = "nptl" Feb 17 14:13:15 GLIBC_EXTRA_OECONF = "--with-tls" Feb 17 14:13:18 and that's it Feb 17 14:13:25 and it fails Feb 17 14:13:36 now it somehow pulled in glibc 2.3.2, and that fails Feb 17 14:13:41 zekoZeko: use pastebin in future Feb 17 14:13:42 are geodegx and generic configurations that exist? Feb 17 14:13:48 ~pastebin Feb 17 14:13:52 it has been said that pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.com/, or http://pastebin.ca, or http://channels.debian.net/paste Feb 17 14:14:31 http://pastebin.com/559515 Feb 17 14:14:54 I've removed the tmp directory before starting Feb 17 14:15:32 zecke: yes, they exist under org.or.dev/conf/ Feb 17 14:15:36 zekoZeko: glibc 2.3.2 won't work with gcc 4.0.2 Feb 17 14:15:48 zekoZeko: did you google for the glibc errors? Feb 17 14:15:50 zekoZeko: either set a newer glibc or an older gcc Feb 17 14:16:00 so I should try PREFERRED_VERSION_glibc="2.3.5"? Feb 17 14:16:17 someone remember which algorith give 48 byte hash? Feb 17 14:16:22 damn, are these interdependencies documented somewhere? Feb 17 14:17:31 zekoZeko: Either use a known working MACHINE/DISTRO combo or be prepared to work out the versions yourself. Of course if nobody builds for your hardware combination, this is tricky :-/ Feb 17 14:17:42 PREFERRED_VERSION_glibc = "2.3.5+cvs20050627" might be what you want Feb 17 14:18:01 That cvs date works well for arm, I don't know about x86 Feb 17 14:21:08 arm uses linuxthreads which is deprecated on x86. you can't really compare results for the two platforms. Feb 17 14:21:26 try and see what happens... Feb 17 14:22:45 RP: but my problem is that it fails building glibc-intermediate Feb 17 14:22:57 of which there is only 2.3.2-cvssomething available Feb 17 14:23:48 ~lart me Feb 17 14:23:48 * ibot declares hrw|work a moron Feb 17 14:24:14 reenoo: I have GLIBC_ADDONS = "nptl" Feb 17 14:24:27 zekoZeko: In that case use an older gcc Feb 17 14:24:35 I see Feb 17 14:24:59 Is 3.4.4 OK? Feb 17 14:25:22 Crofton: ping Feb 17 14:25:46 zekoZeko: hopefully. make sure to set PREFERRED_VERSION_gcc-cross and gcc-cross-initial Feb 17 14:26:13 and -sdk, I presume Feb 17 14:26:39 zekoZeko: You're unlikely to be using that at first Feb 17 14:29:32 03koen 07org.oe.oz354fam083 * r4595d7d3... 10/packages/gnumeric/gnumeric_1.6.2.bb: gnumeric: add 1.6.2 Feb 17 14:29:37 03rw 07org.oe.dev * rade4c2a9... 10/packages/libgpewidget/ (6 files): libgpewidget: inherit gpe. clean things up a bit. Feb 17 14:36:02 zecke, pong Feb 17 14:37:04 * mithro pokes zecke Feb 17 14:38:18 mithro: back from egypt? Feb 17 14:38:29 Crofton: do you need to add real users for https? Feb 17 14:39:27 zecke: anything from .htaccess to ldap or mysql should do Feb 17 14:39:37 koen: back in Cairo Feb 17 14:39:44 leave Egypt on the 20th Feb 17 14:39:46 so 2 more days Feb 17 14:42:29 hmm Feb 17 14:42:37 zecke, what do you mean? Feb 17 14:42:53 only need accounts to grant w perms Feb 17 14:42:59 repo set up for anon read Feb 17 14:51:03 good Feb 17 14:51:08 mithro: how do you like egypt? Feb 17 14:51:15 RP: ping Feb 17 14:51:16 it was really cool Feb 17 14:51:56 RP: one issue with git is, it requires GNU tools Feb 17 14:53:37 hrw|work: pong Feb 17 14:53:57 zecke: true... Feb 17 14:54:18 hrw|work: pong (in case you missed that) Feb 17 14:54:47 morning Feb 17 14:55:02 hi chris144|away Feb 17 14:55:08 hi chouimat even :) Feb 17 14:55:21 RP: when 2.6 kernel is booted with 'quiet' then for some time there is no info for user does Z turn on or not - what do you think about adding logo? Feb 17 14:56:50 hrw|work: We might have kernel size issues on the cxxx devices Feb 17 14:57:13 hrw|work: or do you mean the bootsplash? Feb 17 14:57:35 RP: anything which will show before 'init 2.8x starting' Feb 17 14:58:06 hrw|work: That would have to be something kernel based then Feb 17 14:58:39 hrw|work: This was why mickely added the ELPP iirc Feb 17 14:59:08 but perhaps the standard linux tux logo would be enough Feb 17 14:59:48 or something like that in size Feb 17 15:00:23 2.4.18-crapix 'openzaurus' logo dont fit Feb 17 15:00:51 hrw|work: for c7x0 or cx000 (or both?) Feb 17 15:01:12 hrw|work: for c7x0 or cx000 (or both?) Feb 17 15:01:14 1208032/logo 1202952/nologo Feb 17 15:01:59 We should be able to afford 7k :-/ Feb 17 15:02:37 ~lart mailsweeper Feb 17 15:02:38 * ibot cuts mailsweeper into thin stripes Feb 17 15:03:11 re Feb 17 15:03:18 RP: its my own kernel config - testing mmc Feb 17 15:04:12 hrw|work: I agree a logo would be nice. The kernel configs are already pretty bare though and I'm not sure what to pull out next. CF or MMC are options but people do boot from them... Feb 17 15:04:30 RP: msdos fs Feb 17 15:04:42 Is that large? Feb 17 15:04:46 but its not much (it at all) Feb 17 15:04:54 RP: we have vfat in kernel already Feb 17 15:04:58 Does vfat need it? Feb 17 15:05:01 no Feb 17 15:05:11 both select FS_FAT (or FAT_FS) Feb 17 15:05:23 so its probably not large Feb 17 15:05:34 I'm happy for that to be a module Feb 17 15:06:11 RP: try 'make allmodconfig' maybe? Feb 17 15:06:36 hrw|work: We'd lose things like keyboards, touchscreens, backlight, cf, mmc etc Feb 17 15:06:42 and jffs2 Feb 17 15:07:06 Whilst we don't need them all to boot, I've always thought them essential enough to leave in the kernel Feb 17 15:07:06 RP: to test which module is which size - not to flash Feb 17 15:07:39 ah, right. I once did a script to look at the size of the build-in.o files in each directory Feb 17 15:07:55 That will give you a good indication of which bits use the space Feb 17 15:07:57 RP: all cpufreq governors other then performance Feb 17 15:09:17 loopback? but then coredump will complain about his altboot.. Feb 17 15:11:51 hrw|work: it was pointed out long ago on kernel list, as kernel takes only a second or two to boot, the best way to get a logo is to cat a file straight to /dev/fb0 as first thing you do Feb 17 15:12:28 XorA: The scanning on jffs2 takes more than a second or two Feb 17 15:12:45 XorA: and you can't load anything from the flash without the scan completed Feb 17 15:13:07 XorA: flash 'quiet' kernel on your c7x0 and test when you will know that machine already boot Feb 17 15:13:34 RP: what about printk("Zaurus booting in progress") somewhere? Feb 17 15:13:53 RP: oh, I have scanned jffs2 for a LONG time is it that slow? Feb 17 15:14:11 s/have/havent/ Feb 17 15:14:12 XorA meant: RP: oh, I havent scanned jffs2 for a LONG time is it that slow? Feb 17 15:14:30 XorA: it is - its what hrw is getting at Feb 17 15:15:24 my machine runs init pretty much the instant kernel is booted :-( Feb 17 15:15:27 XorA: during c760 boot jffs2 is probably >50% of time Feb 17 15:15:45 I have to test summary properly one day Feb 17 15:23:13 zecke: at your service Feb 17 15:24:49 RP: if monotone 0.26 solves speed issues, would you stay with it? Feb 17 15:24:49 are there any problems building on AMD64 linux? Feb 17 15:24:58 I'm thinking of buying a dual opteron box... Feb 17 15:25:00 zekoZeko: if there are any, i know about them Feb 17 15:25:12 :) Feb 17 15:25:19 ok :) Feb 17 15:25:44 shadows: unknown. Its going to have to be one heck of a performance increase to solve them though... Feb 17 15:26:15 zekoZeko, i use oe on a athlon64 with 64bit linux distro ... all ok except a small issue building pppd :) Feb 17 15:26:17 this athlon is really too slow for building, I have to wait too much. Feb 17 15:26:18 RP: the issues you're running into are not new, Gentoo team had them when considering a switch from CVS to SVN Feb 17 15:26:42 the difference being that CVS is not distributed in the same way as monotone Feb 17 15:26:48 it's killing me, especially gcc and glibc builds Feb 17 15:27:34 zekoZeko: yea. my athlon64 3000xp+ cpu burns through glibc builds without trouble Feb 17 15:27:50 the problems depend on how you set up your build environment Feb 17 15:27:58 mine is 2500+ barton... I really need to speed it up Feb 17 15:28:52 you can forget about doing legacy 2.95 gcc work, it is not functional in a 32-bit chroot, and can be very troublesome in a mixed lib32 / lib64 environment Feb 17 15:29:10 I don't think I'll need that Feb 17 15:29:29 also, if you use the gcc3 cross initial, the object code is bloated and possibly incorrect for most bb packages Feb 17 15:29:42 hmm Feb 17 15:29:47 gcc4 cross initial is fine, but the devs don't generally use it Feb 17 15:30:08 I might just use vmware and run a 32bit install on it Feb 17 15:30:10 i've cleared out all the problem cases for a gpe-image target on cxx00 openzaurus build with gcc4 cross initial Feb 17 15:30:24 although that takes away the advantage of a fast processor :) Feb 17 15:30:27 yeah Feb 17 15:30:33 shadows: with gcc-cross-initial 4.0.2 means also with gcc-cross 4.0.2? Feb 17 15:30:44 hrw|work: yes, i am implying this Feb 17 15:31:05 to clarify, when i say gcc cross initial, i am also implying that the gcc cross is of the same version i mention Feb 17 15:31:13 shadows: you always use 'gcc4 cross initial' so wanted to be sure 100% Feb 17 15:31:33 yes, i keep them the same in my configs FYI Feb 17 15:31:52 thx Feb 17 15:32:06 anyone here building for x86? I'm building gcc 3.4.4 and glibc 2.3.2 now, hope it works... Feb 17 15:32:20 zekoZeko: i would suggest that if you are going 64-bit, then you must be prepared to use gcc4 for the optimum performance reward Feb 17 15:32:30 oh well Feb 17 15:32:49 it's minor things, but it does involve fixing and working around stupid things that programmers do when they assume the target and host are the same Feb 17 15:32:52 seems I'll have a 32bit build environment, one way or another Feb 17 15:33:01 zekoZeko: yes, I am building for x86, currently for Lippert Cool Road Runner II PC104 board Feb 17 15:33:01 I don't feel adventurous enough right now Feb 17 15:33:07 and I've got to get this working in a few days. Feb 17 15:33:14 ah Feb 17 15:33:19 zekoZeko: what is your build target? Feb 17 15:33:39 chris144: what gcc/glibc combo do you use? My CPU is geode on a Soekris board. Feb 17 15:34:03 let me pastebin local.conf, just a sec. Feb 17 15:34:12 i want more devs using amd64 pure build hosts Feb 17 15:34:32 heh Feb 17 15:34:45 I'm using a 32/32 environment on a dual opteron box Feb 17 15:34:48 zekoZeko: CRRII is geode too. gcc 3.4.4, glibc 2.3.5+cvs20050627 Feb 17 15:34:51 shadows: I can easily try when I buy the machine Feb 17 15:34:56 okay Feb 17 15:35:11 chris144: glibc 2.3.5 fails for me, always Feb 17 15:35:26 zecke: GLIBC_ADDONS = "nptl,libidn" GLIBC_EXTRA_OECONF = "--with-tls" in my distro file Feb 17 15:35:35 I have that, yes Feb 17 15:35:45 in local.conf, but I suppose it's the same. Feb 17 15:35:51 without libidn Feb 17 15:35:57 zecke: it works for TARGET_ARCH i586 Feb 17 15:35:59 I suppose I don't need that for an access point :) Feb 17 15:36:00 pure amd64 exposes bad programming assumptions, which is important if we are going to fix these cases Feb 17 15:36:09 zecke: local.conf is probably ok Feb 17 15:36:26 and it's bloody fast Feb 17 15:36:38 zecke: you don't need libidn...that is ok Feb 17 15:36:43 ?? Feb 17 15:37:06 zecke: no, sorryx I meant zekoZeko Feb 17 15:37:10 ;) Feb 17 15:37:23 http://pastebin.com/559647 Feb 17 15:37:27 there's my local.conf Feb 17 15:37:44 I've set PREFERRED_VERSION-gcc* and -glibc* in the bitbake shell Feb 17 15:38:44 chris144: I have TARGET_ARCH=i486 in geodegx.conf Feb 17 15:38:49 zekoZeko: lloks ok to me. let me know if you have probelems. It took me quite some time to get glibc+20050627 built for x86 Feb 17 15:38:59 OK Feb 17 15:39:08 glibc is being built right now Feb 17 15:39:35 I'm glad I've found someone who got everything working on x86 :) Feb 17 15:39:52 btw, I'm using the pregreakage revision Feb 17 15:39:59 s/pregreakage/prebreakage/ Feb 17 15:39:59 zekoZeko meant: btw, I'm using the prebreakage revision Feb 17 15:40:27 zekoZeko: what is prebreakage? Feb 17 15:40:53 monotone update -rjan2006prebreakage Feb 17 15:41:22 there we go Feb 17 15:41:24 glibc failed Feb 17 15:41:46 zekoZeko: ok, I am just about to start building with oe from yesterday, full bootstrap-image. lets see if it works Feb 17 15:42:02 zekoZeko: post the log to pastebin please Feb 17 15:42:35 http://pastebin.com/559656 Feb 17 15:44:38 zekoZeko: I never got glibc-2.3.2+cvs20040726 build with nptl and tls. Try glibc-2.3.5+cvs2050627. That works with tls and nptl. gcc3.4.4, targwet i586 Feb 17 15:45:42 03mickeyl 07org.oe.oz354fam083 * r463ce400... 10/packages/python/ (3 files in 2 dirs): Feb 17 15:45:42 python-pylinda: upgrade to 0.6 Feb 17 15:45:42 (taken from .dev) Feb 17 15:46:57 chris144: you mean TARGET_ARCH? Feb 17 15:48:04 zekoZeko: http://pastebin.ca/41952 contians a distro and machine file from a successfull x86 build. the OE I used was snapshot 20060201. Feb 17 15:48:20 zekoZeko: yes, TARGET_ARCH Feb 17 15:48:56 RP: hehe, git strikes again :} Feb 17 15:49:58 zekoZeko: by looking at your pastbin again, I think that I remember that glibc-2.3.2+cvs20040726 only builds with linuxtherad, not with nptl and also not with tls stuff. Feb 17 15:50:18 linuxthreads, damn the quick typing Feb 17 15:51:28 zekoZeko: I hope that helps a bit. let me know how building glibc-2.3.5+cvs2050627 goes. I will be here for a while now. mybe gone for a few minutes inbetween... Feb 17 15:51:52 ping koen Feb 17 15:51:58 ~ping koen Feb 17 15:51:59 pong koen Feb 17 15:52:36 NOTE: preferred version 2.3.5+cvs20050627 of glibc-intermediate not available Feb 17 15:52:36 NOTE: preferred version 2.3.5+cvs20050627 of glibc-initial not available Feb 17 15:52:54 I should update past this revision it seems... Feb 17 15:53:00 zecke: What did it do now? :) Feb 17 15:53:18 zekoZeko: hold on...let me look something up.. Feb 17 15:53:49 RP: I wonder how to merge :} Feb 17 15:54:03 zecke: Ah :) Feb 17 15:54:30 doing 'md5sum /dev/mmcblk0' with mmc_debug enabled takes ages on 64m card Feb 17 15:54:44 chris144: pong Feb 17 15:55:04 zekoZeko: ok, I am not setting any PREFERED Version for glibc. It automatically takes: Feb 17 15:55:24 glibc-2.3.5+cvs20050627 Feb 17 15:55:42 glibc-initial-2.3.2+cvs20040726 Feb 17 15:55:44 how about glibc-initial and -intermediate? Feb 17 15:56:09 hi mickeyl Feb 17 15:56:10 glibc-intermediate-2.3.2+cvs20040726 Feb 17 15:56:16 hi RP Feb 17 15:56:26 zekoZeko: that wokrs. Feb 17 15:56:27 hi mickeyl Feb 17 15:56:28 RP: I got the driver to compile. I haven't tried the kernel yet, though, as I'm sure there are problems Feb 17 15:56:29 mickeyl: hey Feb 17 15:56:33 hey mickeyl Feb 17 15:56:46 RP: first off, both of the lines that use corgiscoop_device give me linking errors Feb 17 15:56:46 RP: monotone 0.26 is a lot faster (at least locally) Feb 17 15:56:56 mickeyl: hi :-) Feb 17 15:57:01 RP: git merge-index cat conf/distro/familiar-unstable.conf Feb 17 15:57:08 RP: I wonder what to do now Feb 17 15:57:10 JustinP: cheers. how's E doin? Feb 17 15:57:11 mickeyl: tried the e images in oz354fam083 yet? Feb 17 15:57:14 hi koen, zecke Feb 17 15:57:16 hehe Feb 17 15:57:22 JustinP: not a chance yet Feb 17 15:57:28 zekoZeko: you have set PREFERRED_VERSION for glibc-intermediate and initial? Feb 17 15:57:30 <---- terribly overworked Feb 17 15:57:52 JustinP: corgiscoop_device should be exported by one of the headers - maybe asm/arch/corgi.h Feb 17 15:58:01 or spitz.h in your case Feb 17 15:58:01 chris144: yes, to 2.3.5-cvs, but it simply uses 2.3.2-cvs20040726, because that's the only one available Feb 17 15:58:24 JustinP: change them to spitzscoop_device references Feb 17 15:58:25 and that is already built, so I'm only going to build new glibc, set with PREFERRED_VERSION_glibc in bitbake shell Feb 17 15:58:42 mickeyl: How did the presentation go? Feb 17 15:58:48 RP: corgi.h has it and is included in the source files...hence no compile errors, but as I said I get linking errors Feb 17 15:58:53 RP: fatal: merge program failed Feb 17 15:58:55 hmm Feb 17 15:59:04 JustinP: You have a spitz? Feb 17 15:59:29 zekoZeko: I understand; try to do it like I did it in my distro and machine files on pastebin. it works! Feb 17 15:59:44 JustinP: If so, make it compile for spitz scoop, not corgi. There are ways to resolve this so it works on both but we can worry about that later Feb 17 16:00:08 chris144: will do. If it fails now I'll update my org.openembedded.dev to the latest revision and try again Feb 17 16:00:13 JustinP: s/corgi.h/spitz.h/ s/CORGI_/SPITZ/ Feb 17 16:00:18 RP: ok. I just used corgi as that was what it was named and all. I'll try to switch it around Feb 17 16:00:39 zekoZeko: also check the glibc package directory there must be the 2.3.5+cvs20050627 Feb 17 16:00:50 JustinP: My fault, I just assumed you'd change it :) Feb 17 16:01:37 yeah, I have it Feb 17 16:01:43 NOTE: package glibc-2.3.5+cvs20050627-r1: task do_compile: started Feb 17 16:01:55 RP: np. I was thinking in "embeddix" mode ;-) Feb 17 16:03:03 zekoZeko: in my build glibc initial and intermediate actualy are 20040726. that is probabyly ok. Feb 17 16:03:21 RP: also, there was a call to reset_set_scoop_gpio, which doesn't seem to exist....any idea which one is right? Feb 17 16:04:32 RP: and another one...SA_TRIGGER_RISING was used but is not define danywhere in the source in either 2.6 or 2.4. I googled and found that it applies to some patch.... Feb 17 16:05:19 cu Feb 17 16:05:21 koen: there is no fluxbox package. there is matchbox. Is that what you meant? Feb 17 16:05:35 chris144: ah, I thought fluxbox was in OE Feb 17 16:05:52 you could use matchbox as well, but it shouldn't be too hard to add fluxbox Feb 17 16:06:21 koen: I will try both. I looked at the fluxbox homepage already. looks nice. Feb 17 16:06:31 * france is back (gone 11:48:55) Feb 17 16:10:39 JustinP: The reset_set looks like a typo - the original sharp code should give a clue Feb 17 16:11:46 ok Feb 17 16:14:07 whoa Feb 17 16:14:13 i got glibc built Feb 17 16:14:18 that's an improvement Feb 17 16:14:31 JustinP: The SA_TRIGGER_RISING is probably as I based on something from my local git based kernel tree and itsn't in older kernels Feb 17 16:14:36 chris144: you've been great, thank you. Any way I can make it up to you? Feb 17 16:14:43 zekoZeko: cool ! Feb 17 16:15:25 zekoZeko: that is ok, this is a helpful channel. I am just giving back what was given to me some thime ago... Feb 17 16:16:08 When I finish this I'll put something in the wiki. Feb 17 16:16:40 RP: I figured. I removed that part (well, commented). Feb 17 16:16:51 :-) Feb 17 16:27:41 how can I disable the building of locales in glibc? Feb 17 16:28:51 hehe Feb 17 16:28:58 zekoZeko: I'd like to know that too Feb 17 16:29:04 03florian 07org.oe.oz354fam083 * r8552af1f... 10/packages/glibc/glibc_2.3.5+cvs20050627.bb: glibc 2.3.5+cvs20050627: Change source location again. Feb 17 16:30:43 chris144: have you done anything with uclibc? I'd prefer to use that as it's much simpler... Feb 17 16:31:27 zekoZeko: JustinP: I think that by adding some EXTRA_OECONF += "...." you might be able to switch the locale building off. But I am not sure if the package then still builds. Feb 17 16:31:40 :) Feb 17 16:31:56 will try, once I've got something booting on the board. Feb 17 16:32:11 zekoZeko: no uclibc yet. But I knoe that there is a setting in your conf file thet you need to set if you want to build a uclibc based system Feb 17 16:32:53 zekoZeko: TARGET_OS = "linux-uclibc" in your local.conf. Then it should work Feb 17 16:34:08 yes, I know that, it was the first thing I tried, but then it failed and I decided to stich with glibc for the time being Feb 17 16:34:17 zekoZeko: ok Feb 17 16:37:15 zekoZeko: maybe you can use that: I wrote a script that makes building of a bootstrap-image completely automatic. i am planning to publish it, if you like have a look at http://pastebin.com/559747 any try it. Feb 17 16:38:18 zekoZeko: it even does alll the config file stuff.... just hit the Enter key... Feb 17 16:38:24 RP: is your git server down? Feb 17 16:40:09 zecke: back up now Feb 17 16:40:38 RP: did it crash? Feb 17 16:41:00 zecke: I think so. Were you responsible? :) Feb 17 16:41:18 folks, I seem to have a problem, with gnu-config-native, see http://pastebin.com/559756 ,any idea? Feb 17 16:41:47 me? maybe it is that piece of shit linux you use? Feb 17 16:42:21 yes I think my git push killed it Feb 17 16:44:02 zecke: That uml install has been a bit flakey... Feb 17 16:44:55 chris144: that's strange.....it doesn't look like it did anything.... Feb 17 16:45:11 RP: I wonder how we can do 'git revert' is there something like this Feb 17 16:45:21 JustinP: the patch seems to fail... Feb 17 16:45:26 ah git revert :) Feb 17 16:45:45 yay! i have a nano .ipk Feb 17 16:46:12 zecke: I was about to say :) Feb 17 16:47:47 hi Feb 17 16:47:57 hi CoreDump|home Feb 17 16:48:18 hrw|gone: I won't cry if you remove loopback, I'll just have to modprobe it Feb 17 16:52:33 :-( Feb 17 16:52:43 chris144: what's in the log? Feb 17 16:54:40 JustinP: I am just rebuilding, with oe snapshot from 20060201. lets see if that works. Feb 17 16:55:23 chris144: It's building linux-geodegx-2.4.24 now... If I understood correctly you are running 2.6? Feb 17 16:56:19 gotta go home now... GF cooking, and I haven't had anything to eat yet today. Feb 17 16:56:25 bye Feb 17 16:56:29 zekoZeko: yes, for now 2.6. I might change to 2.4 and use RT-linux-gnu dependant on how that realtinme stuff goes Feb 17 16:56:41 which 2.6? Feb 17 16:56:42 zekoZeko: ok, see ya Feb 17 16:56:54 zekoZeko: that latest... Feb 17 17:03:42 RP: ok, I renamed stuff around and got it to build again, with much finagling Feb 17 17:04:12 RP: I'll try booting this kernel (although I suspect I'm going to have some panics and such ;-) Feb 17 17:04:20 JustinP: I never claimed it would work straight off ;-) Feb 17 17:04:32 JustinP: If should be possible to build it as a module for debugging Feb 17 17:05:43 JustinP: /home/chris/projects/CL-Embedded_Linux/new-oe-project-wm/build/tmp/work/gnu-config-native-0.1+cvs20050701-r4/temp/run.do_patchcleancmd.12280: line 241: quilt: command not found Feb 17 17:06:29 JustinP: very strange. quilt does not seem to a be dependency in gnu-config-native... Feb 17 17:08:29 chris144: You'd want quilt native Feb 17 17:09:41 RP: yes, but this is a build from scratch. gnu-config-native is the 1st package that gets built. and fails right away. really stange. Feb 17 17:09:49 strange Feb 17 17:10:27 chris144: I saw an interesting error from a missing dependency on autoconf-native from libtool-native earlier today... Feb 17 17:10:51 Looks like something changed the order bitbake operates in throwing up a few bugs :) Feb 17 17:12:10 RP: yes, that is what I think too. because the build that I just did was using oe from 20060201. So it must be bitbake. did anybody do a bootstrap-image from scratch the last few days? Feb 17 17:14:06 I suspect its the RDEPENDS change I checked into bitbake which changed the order Feb 17 17:14:40 The RDEPENDS change is correct and it looks like its exposed holes in the metadata :} Feb 17 17:15:15 zecke: Looks like bitbake is exposing bugs in OE again ;-) Feb 17 17:15:30 RP: ah, ok. so gnu-config-native-0.1+cvs20050701 possibly need to be fixed... Feb 17 17:16:00 Ill give it a tgry... Feb 17 17:16:02 try Feb 17 17:16:07 I suspect so - it should DEPEND on quilt-native. If it doesn't that would be a bug Feb 17 17:16:17 RP: ok Feb 17 17:16:26 quilt-native is an implicit dependency Feb 17 17:16:44 if bitbake doesn't grok that any more it's certainly not OE's fault Feb 17 17:18:40 reenoo: An implicit dependency in what way? Feb 17 17:19:05 every package that needs patches has quilt-native as a dependency Feb 17 17:19:26 reenoo: Where is that set? Feb 17 17:20:06 grep -r quilt-native conf/* Feb 17 17:20:10 will tell you Feb 17 17:22:43 it seems to be: openembedded/conf/bitbake.conf:PATCH_DEPENDS = "quilt-native" Feb 17 17:26:38 chris144: bitbake gnu-config-native -e | grep DEPENDS does show quilt-native correctly here Feb 17 17:28:20 RP: I am still looking into it... Feb 17 17:31:03 hello to all Feb 17 17:37:41 I have some questions about Angstroem. I want to contribute to a OpenEmbedded distribution and I have no clue about which one should I contribute. What are the advantages about Angstroem? And is there any Wiki or Informations about it on the net? I have Slackware Distribution skills and I build some spin offs of it for my own use. So I have litte knowlege about distributing. I've first planned to build my own distribution for Simens SIMpad called " Feb 17 17:37:59 the PATCH_DEPENDS = "quilt-native" issue does not seem to be easy for me. need some food now. I'll be back later... Feb 17 17:40:22 mr_nice: http://handhelds.org/hypermail/angstrom-dev/0/date.html and http://openzaurus.sourceforge.net/wordpress/newsletters/ooo-newsletter-200508/ Feb 17 17:40:33 mr_nice: that's currently all we know about it :) Feb 17 17:41:09 mr_nice: The existing openzaurus and familar will become Angstroem so I'd suggest working on them for now Feb 17 17:43:27 thx Feb 17 17:45:33 hmm, noble goals for Ångström Feb 17 17:47:26 reading this, h3600 should be pushed to 2.6 as 2.4 is not stable anyway Feb 17 17:47:54 atleast i can attest to major issues Feb 17 17:48:23 2.4 is fairly stable here. and 2.6 is far from being usable. Feb 17 17:48:48 i am getting kernel oops every few days Feb 17 17:48:58 even with the feb 9 RC2 Feb 17 17:49:47 suspend lockups, sleeve problems Feb 17 17:50:31 but i am not 100% positive that i dont have a hardware issue of some type Feb 17 17:50:45 i dont have a second h36 to test against Feb 17 17:52:21 hmm Feb 17 17:52:33 * emte checks ebay for another h36 Feb 17 17:53:10 RP: I know you didn't ;-). Well, I figured building it into the kernel would be less likely to have problems...i.e. if the module load/unload code is incorrect it would make debugging take that much longer.... Feb 17 17:53:26 RP: I'm not afraid of flashing kernels ;-) Feb 17 17:53:50 JustinP: ok. printk is your friend btw ;-) Feb 17 17:53:51 JustinP, i may have a present for you soon Feb 17 17:53:55 :) Feb 17 17:54:48 RP: I'll definately use that once I flash the kernel. Not likely until tonight, though (just got to work) Feb 17 17:55:00 emte: is that so? what kind? Feb 17 17:55:25 how about a proper dbus interface for e17 Feb 17 17:55:32 hi all Feb 17 17:55:48 since the current e17 wm cheats and parses the proc file Feb 17 17:55:56 03coredump 07org.oe.dev * rbbcd8ca8... 10/packages/keymap-locales/ (5 files in 5 dirs): Feb 17 17:55:56 keymap-locales: Feb 17 17:55:56 - Fix o-umlaut on VT and X and remove warning message during loadkeys Feb 17 17:55:56 - Add c7x0 -DE keymap, testing courtesy of _law_ Feb 17 17:55:56 mt: ---------------------------------------------------------------------- Feb 17 17:56:35 emte: ok. I know nothing about dbus myself, so I'm not sure what all that means... Feb 17 17:57:06 it means that instead of polling the file and parsing, you wait of an event and then handle it Feb 17 17:57:10 for* Feb 17 17:57:23 which may reduce your overhead Feb 17 17:57:29 hi Mardy Feb 17 17:57:31 ah Feb 17 17:57:51 as well as easier access to any event on the dbus you may want Feb 17 18:01:06 emte: do you use e17 a pda? which graphik libary do you use gtk+ + evas? what about the memory usage? or do you use ewl? Feb 17 18:03:07 mr_nice, JustinP is the maintainer for the "full" e17, i am (eventually) releasing an embedded e17 wm Feb 17 18:04:30 mr_nice: I don't worry about libraries and such...I'm not a UI programmer, or at least not normally Feb 17 18:04:44 mr_nice: I *do* run e17 on my spitz (C3000), though :-) Feb 17 18:05:04 cool Feb 17 18:05:50 JustinP:I want to contribute on the e17 thing. I use e17 on my laptop on Slackware 10.2. Feb 17 18:06:10 nope, running e17 on a PDA is way past cool. It's downright freezing Feb 17 18:06:34 JustinP: I am just compiling boost to get OpenEmbedded running Feb 17 18:06:42 http://hentges.net/tmp/screenshots/Zaurus/Akita/Enlightenment/ Feb 17 18:08:09 JustinP:are you often on this chanel? If I have enoughf skillst with bitbake I will help you ore emte. Feb 17 18:08:57 mr_nice: thanks, help is good :-) Some of the e17 packages need some work, I haven't had manyof the them compiling in a while Feb 17 18:09:12 mr_nice: the "core" (entrance and e-wm) are working great, though Feb 17 18:09:41 mr_nice: and yes, I'm almost always here. I'm not always *listening* but I usually see messages sent to me (or prefixed with my name) Feb 17 18:09:47 JustinP: I'm not sure if it was with the latest e stuff, but bitbake complained about the .pc not being in any package for some e stuff Feb 17 18:09:56 * koen|assignment should take more notes Feb 17 18:10:27 koen|assignment: .pc? I really have no idea....and .h fiels have been not packaging for a while...I keep meaning to look into that... Feb 17 18:10:44 mr_nice: what machine do you have? Feb 17 18:10:48 if you stumble on that by accident: they should be in the -dev package :) Feb 17 18:11:18 koen|assignment: I know. And the FILES is set up right. It's just my sed hackery has made them install in a double-directory Feb 17 18:12:07 koen|assignment: /stuff/tmp/work/e/install/stuff/tmp/work/e/include Feb 17 18:12:36 koen|assignment: and I'm pretty sure it's my sed hackery in efl.bbclass Feb 17 18:12:36 JustinP: I have a Simens SIMpad SL4 - strong arm sa1100 206 Mhz 64MB Ram 32 MB flash. I hope e17 will run on this I didn't tryd it. Feb 17 18:13:13 mr_nice: 32MB flash is going to be pushing it a bit...e17 is large. The core should definately install and work, though Feb 17 18:13:21 mr_nice, perhaps when i get to a first release ... i currently have the habit of writting something then tossing the code because i dont like it ...( i think i've been hanging out with raster too much ) Feb 17 18:13:24 mr_nice: and the packages *can* be made smaller, I just haven't gotten around to it Feb 17 18:13:44 mr_nice: does the SIMpad have a keyboard? Feb 17 18:14:41 mr_nice: ah...looks like it does't.. Feb 17 18:14:55 JustinP: i could start it from a bigger cf card. yes i have a keyboard i connected a palmV keyboard Feb 17 18:15:05 mr_nice: you can't use e-image-core, then, as it is now. You need a keyboard to be able to login... Feb 17 18:15:14 mr_nice: oh, then you should be ok as long as it auto-starts Feb 17 18:15:16 JustinP: I'm writing on my Laptop Feb 17 18:15:41 mr_nice: perhaps you could help us get an onscreen keyboard up for entrance ^_^ Feb 17 18:15:59 ok, I need to go for a bit, lots of work Feb 17 18:16:00 emte: thx i will try it for you. If you have work for me i eventually could help. Feb 17 18:16:34 mr_nice: it might even be possible with just a theme (although I don't know for sure) Feb 17 18:17:10 mr_nice, how much do/have you programmed with the e17 libraries? Feb 17 18:17:34 JustinP|busy: first I have to get it booting ;) Feb 17 18:18:03 emte: Unfortunatly I have no programming skills. Feb 17 18:19:11 emte: I have some comiling and skripting skills but no C skills. I am only able to write hello world in C. Feb 17 18:19:29 mr_nice, you may or maynot then want to look at the e17 cookbook ... or i suppose you could look at how to build themes, i am sure JustinP|busy could use a pda-centeric theme for the full e17 Feb 17 18:20:34 themes are mostly embryo script anyway :) Feb 17 18:21:08 emte: I will look at how to build themes. Feb 17 18:21:49 emte: but I can't garantee that I am have time this week for it. Feb 17 18:22:07 emte: But I will try to do it as soon as possible Feb 17 18:22:47 emte: do you know a ewl screen keyboard? Feb 17 18:23:06 emte: or which should I use? Feb 17 18:24:10 you could use the matchbox-keyboard backend with a ewl-frontend Feb 17 18:25:40 koen|assignment: I will hava a look at it. thx Feb 17 18:27:18 as far as i know an ewl-keyboard does not exist Feb 17 18:28:06 emte: I will give the matchbox-keyboard a try. Feb 17 18:28:11 aside from raster himself i dont think anyone "normally" runs e17 on thier pda Feb 17 18:28:46 altho i dont know if raster has recently recompiled for his either Feb 17 18:29:47 emte: I hope I'will get entrance starting on my SIMpad Feb 17 18:30:25 emte: the good thing is that i am able to do the work with the matchbox-keyboard on my laptop Feb 17 18:31:07 emte: so this dinn't depend on the starting of entrace on my Simpad :) Feb 17 18:31:39 thats good Feb 17 18:32:17 i am sure both JustinP|busy and CoreDump|home woudl both like help with entrance Feb 17 18:33:06 emte: good to know ;) Feb 17 18:37:29 btw, is there a solution or work-around for the infamous locale-base issue in the works? Feb 17 18:38:03 koen|assignment, i was wondering, for Ångström and its "Machine mentors" could a device checkpoints list exist and be viewable by users? (the checkpoints list being the key or release required elements being passed) Feb 17 18:38:04 CoreDump|home: yes, ram upgrades or swap Feb 17 18:38:21 swap on flash? errr Feb 17 18:38:24 emte: sure Feb 17 18:39:20 it would be nice to know which requirements are satisfied for each and which need to be worked on as oppsed to a generic distro-wide Feb 17 18:40:00 emte: wanna start such a list? Feb 17 18:40:22 koen|assignment, i can/could if i knew what the current passing requirements were Feb 17 18:40:56 but i can come up with a list i'd liek to see specifically for the h36 Feb 17 18:41:01 folks: from what I can tell the dependency recognition in bitbake is broken. A bitbake bootstrap-image does not work any more as per today. it worked still properly on 02/01/2006. It might have to do with the rpurdie RDEPENDS fixes. I have to go to dinner now, so I have to stop looking into it now. I will continue tomorrow. good night. Feb 17 18:42:07 koen|assignment, i'll do that and and submit it to the list as an idea Feb 17 18:42:20 emte: thanks Feb 17 18:42:46 * emte goes to get key programmed, bbiab Feb 17 18:42:48 is there a variale in the OE env for the used kernel version of the image? Feb 17 18:42:51 keys* Feb 17 18:43:29 CoreDump|home, that is in a couple places through the distro bb is it not? Feb 17 18:43:39 CoreDump|home: see the checkversion script Feb 17 18:43:49 ah thanks Feb 17 19:05:25 03justinp 07org.oe.dev * r43a21a92... 10/packages/libxine/ (libxine-x11-1.0/libxine-gcc4.patch libxine-x11_1.0.bb): libxine-x11: Add a patch tto fix gcc4 compilcation, fixes bug #694, thanks to Mikko Rasa Feb 17 19:12:24 emte: I just figgured out that a screen keyboard could easily build in a entrance theme if it will be a edc file. So I just downloaded the edje documentation. on the first look i it doesn't seems to be too hard to write a screen keyboard in it. I will try to write one and than I will inform you or JustinP. Feb 17 19:19:00 mickeyl: where do we stay again? Feb 17 19:19:22 zecke: the astrid? Feb 17 19:22:32 thanks Feb 17 19:35:46 i got >>> ERROR: 'NoneType' object has no attribute 'split' while parsing /mnt/hda/oe/org.openembedded.dev/packages/gtk-webcore/osb-jscore_20060212.bb <<< have to update bitbake ? Feb 17 19:36:37 no, get JustinP|busy to fix that Feb 17 19:36:55 cu Feb 17 19:46:36 mhh anyway i did an update to bitbake ... but now i got >>>ImportError: No module named svk<<< Feb 17 19:50:21 gremlin[it]: oh Feb 17 19:50:26 gremlin[it]: did you install it? Feb 17 19:51:18 yes Feb 17 19:52:40 gremlin[it]: try again Feb 17 19:52:48 what again ? Feb 17 19:52:53 hm a) my quality sucks Feb 17 19:53:01 gremlin[it]: setup.py install Feb 17 19:53:12 b) wip bitbakes should not be installable :} Feb 17 19:54:53 again ... sabe erroro Feb 17 19:54:56 error Feb 17 19:58:00 after checking again and ./setup.py clean ... work Feb 17 19:58:19 checking -> svn co Feb 17 19:58:49 RP: monotone 0.26 is really way faster - the initial pull will take more than one hour though Feb 17 19:59:37 but also with 'new' bitbake i got >>>ERROR: 'NoneType' object has no attribute 'split' while parsing ... <<< Feb 17 20:00:11 gremlin[it]: again. it's not a problem with bitbake. the .bb is faulty. pester JustinP|busy to fix it. Feb 17 20:00:36 yes sorry Feb 17 20:01:03 03rw 07org.oe.dev * r8e469d6d... 10/packages/matchbox-panel/ (2 files in 2 dirs): Feb 17 20:01:03 matchbox-panel: add patch to allow disabling the Utilities/Panel submenu which is hardcoded and redundant. Feb 17 20:01:03 - tap-and-hold on the panel already allows adding applets Feb 17 20:01:03 - export MB_MENU_PANEL_DISABLED=1 before starting mb-applet-menu-launcher to disable the submenu Feb 17 20:11:44 03koen 07org.oe.dev * r7e5c3b77... 10/packages/linux/LAB-kernel_cvs.bb: LAB: add warning about build order Feb 17 20:20:18 03rw 07org.oe.dev * r0d4c64e0... 10/packages/gpe-session-scripts/ (files/matchbox-session gpe-session-scripts_0.66.bb): Feb 17 20:20:18 gpe-session-scripts: disable Utilities/Panel submenu in matchbox-panel (mb-applet-menu-launcher). Fixes hh.org Bug #1462. Feb 17 20:20:18 - requires matchbox-panel_0.9.2-r12 to work Feb 17 20:23:02 reenoo: why not add a versioned RDEPENDS instead for gpe-session-scripts? Feb 17 20:23:50 koen: instead of what? Feb 17 20:23:57 woohoo :)) Feb 17 20:23:59 NOTE: package task-bootstrap-1.0: completed Feb 17 20:24:26 reenoo: instead of a commit message people will read and forget Feb 17 20:25:04 21:22 < CIA-4> - requires matchbox-panel_0.9.2-r12 to work <- that one Feb 17 20:26:12 * koen has a sneaking suspicion versioned RDEPENDS don't work anymore Feb 17 20:27:31 03koen 07org.oe.dev * r748f1a9f... 10/packages/linux/LAB-kernel_cvs.bb: LAB: don't stage headers Feb 17 20:28:44 koen: yeah. I guess I could have adjusted RDEPENDS. arguably, we're not using versioned RDEPENDS anywhere else either though Feb 17 20:29:26 pb_ recommended me to do that for basefile (or initscripts) in the branch Feb 17 20:29:35 so I assumed it was used in other .bbs Feb 17 20:29:48 didn't check it, though Feb 17 20:31:17 the base-files thing was a "hoses your system if the other package isn't new enough" case, though. reenoo's has rather lower impact than that, you'll just get a stupid menu item. Feb 17 20:31:25 hey pb_! Feb 17 20:31:51 setting a versioned RDEPENDS in gpe-session-scripts would be a fine idea, but it probably isn't all that vital in this case. Feb 17 20:31:57 hi mithro Feb 17 20:33:29 back in cairo now Feb 17 20:33:41 very good Feb 17 20:35:12 koen: I'm using versioned RDEPENDS in the sablevm and classpath .bbs. the -r? extraversion has been causing issues though. Feb 17 20:35:26 pb_: happen to know the correct format for that so ipkg will grok it? Feb 17 20:35:39 for what, sorry? Feb 17 20:35:44 reenoo: I kinda suspect that the versioned RDEPENDS break with the new and shiny infered depends Feb 17 20:36:23 pb_: i.e. will RDEPENDS += "matchbox-panel (>=0.9.2-r12)" do what I want? Feb 17 20:36:24 RDEPENDS_append_familiar = " initscripts (>= 1.0-r60)" Feb 17 20:36:40 that worked for base-files Feb 17 20:36:56 reenoo: I think it should Feb 17 20:37:12 * reenoo adds versioned RDEPENDS Feb 17 20:37:54 mhhh got trouble, unabel to download libc_sources.redhat.com__20050627.tar.gz Feb 17 20:38:16 familiar.handhelds.org --> Connection refused. Feb 17 20:38:36 www.oesources.org --> 404 Not Found Feb 17 20:47:45 03rw 07org.oe.dev * re2330907... 10/packages/gpe-session-scripts/gpe-session-scripts_0.66.bb: gpe-session-scripts: add versioned RDEPENDS on matchbox-panel, matchbox-desktop, matchbox-common. Feb 17 21:08:17 [ 9.389903] Total of 32 processors activated (9895.93 BogoMIPS). Feb 17 21:08:24 *wants to have that one* Feb 17 21:09:00 eleet Feb 17 21:09:39 http://vger.kernel.org/~davem/cgi-bin/blog.cgi/2006/02/17#niagara_boots Feb 17 21:10:04 hihihi Feb 17 21:15:08 heh. wished rpjday would finally stop the bitching, put some money where his mouth is and donate the amount required to renew the domain Feb 17 21:18:56 reenoo: thats a bold wish Feb 17 21:19:20 reenoo: whats the problem? Feb 17 21:20:37 i really need to get those domains transferred after i pay them. easydns is charging $35 a year, a piece. craziness. want to move them onto godaddy Feb 17 21:21:05 yeah, good plan Feb 17 21:21:14 did you get your paycheck yet? Feb 17 21:21:15 heh Feb 17 21:21:43 deposited it yesterday Feb 17 21:21:46 heh Feb 17 21:21:54 rock Feb 17 21:22:19 course i think my car company tried to take the bill out of my account the day before that, but.. Feb 17 21:22:23 * kergoth mutters Feb 17 21:22:55 "tried" being the operative word there, presumably Feb 17 21:23:02 indeed Feb 17 21:23:11 oh well Feb 17 21:23:19 I'm sure they'll get over it Feb 17 21:24:12 ~curse my home computer Feb 17 21:24:14 May you be reincarnated as a Windows XP administrator, my home computer ! Feb 17 21:24:43 hehe Feb 17 21:24:48 ~botsnack Feb 17 21:24:48 reenoo: aw, gee Feb 17 21:24:59 ~lart kmail 'devels' too Feb 17 21:24:59 * ibot sends a legion of lawyers after kmail 'devels' too's head Feb 17 21:26:00 kmail is good example to show how to NOT write apps with GUI Feb 17 21:26:41 'one thread to rule them all' Feb 17 21:30:19 djeez Feb 17 21:30:34 'ow, it got multiple heads again. ' Feb 17 21:30:39 ~stab rday Feb 17 21:30:41 * ibot runs at rday with an origami Swiss Army knife, and inflicts a nasty paper cut. Feb 17 21:34:04 heh. I have him killfiled. it's the responses that are annoying >:) Feb 17 21:34:17 woops, sorry Feb 17 21:35:04 * Philippe is back (gone 22:53:18) Feb 17 21:37:58 to flash the bootloader to an ipaq 3900, should a flash the .bin file or the .bin.gz file? Feb 17 21:54:02 RP: it looks like gdb has iwmmxt simulation Feb 17 21:55:03 zecke: does setting mcpu=iwmmxt force softvfp? Feb 17 21:55:55 koen: never thought about it, I think it doesn't Feb 17 21:56:13 it sure wouldn't be hardfpa, would it? Feb 17 21:56:58 koen: if co processor 0 or 1 is the hardfpa... obviously not Feb 17 22:08:45 i'm off to bed Feb 17 22:08:52 'night mithro Feb 17 22:08:57 still need to sort out my accomidation for FOSDEM Feb 17 22:09:03 got my flight back to Rome booked Feb 17 22:13:09 koen: bitbake knows about versions in RDEPENDS. Feb 17 22:13:26 RP: and in DEPENDS? Feb 17 22:13:55 koen: versioned DEPENDS have never been supported Feb 17 22:14:13 the biggest misfeature in OE Feb 17 22:14:17 no versioned DEPENDS Feb 17 22:15:48 You could always propose it and implement it Feb 17 22:16:02 propose it *again*? Feb 17 22:16:42 I've never noticed you mention it before :) Feb 17 22:17:39 I suspect there is a good reason it hasn't been done but I can't immediately see it... Feb 17 22:17:53 zecke: I'm trying to be optimisic about monotone 0.26 Feb 17 22:19:59 Do we have plans to switch to monotone 0.26 or are we going to wait for the SCM trial to conclude first? Feb 17 22:20:19 RP: monotone 0.26 will compete as an alternative to monotone :) Feb 17 22:20:36 RP: no *need* to switch Feb 17 22:20:49 we can use monotone 0.25 till hell freezes over Feb 17 22:20:57 koen: could we server 0.26 as well? - the netsync port changed Feb 17 22:20:59 we aren't gentoo ;) Feb 17 22:21:04 serve even Feb 17 22:21:19 zecke: sure, just need to punch a hole in the firewall Feb 17 22:21:30 koen: I won't use 0.25 forever ;-) Feb 17 22:21:48 RP: right now I wouldn't switch scms Feb 17 22:21:56 everything else has just too much problems Feb 17 22:22:06 switching for the sake of switching would be stupid Feb 17 22:23:20 The speed increase is a big attraction to me so it wouldn't be for the sake of it Feb 17 22:23:51 koen: how are the pre-releases of monotone 0.26 looking Feb 17 22:24:01 git is the fastest - no doubt there Feb 17 22:24:03 XorA|gon1: 7 to 8 times faster Feb 17 22:24:16 koen: sweet Feb 17 22:24:19 zecke: in there hg should be the fastest Feb 17 22:24:28 theory* Feb 17 22:24:30 koen: for pulls or commits otr both? Feb 17 22:24:51 koen: well --hg for sitting there after a hg push without doing anything Feb 17 22:25:20 zecke: the monotone mailinglist has some nice overviews on overhead of different scms Feb 17 22:25:29 RP: both Feb 17 22:25:54 woglinde: hey Feb 17 22:25:57 hey woglinde Feb 17 22:26:02 So it would become passable again for a while... Feb 17 22:27:11 I have hg accounts to give away Feb 17 22:28:07 * RP needs more hours in the day Feb 17 22:29:08 with QA in mind we should use either svk (trac) or monotone (certs) Feb 17 22:29:26 or wait till the hg trac module is finished Feb 17 22:31:33 fucking crosstools... Feb 17 22:31:44 no easy way to use gmake and gpatch Feb 17 22:31:59 * koen is tempted to build a complete iwmmxt image for his hx Feb 17 22:33:56 but first I have to get gcc 4.0.x working (shouldn't be too hard), figure out the git fetcher (trial and error) and figure out how to tftp with redboot Feb 17 22:34:23 ~lart redboot Feb 17 22:34:23 * ibot turns redboot into a lifesized tux doll Feb 17 22:34:32 koen: Would you like an example of a git .bb file? Feb 17 22:34:40 RP: sure Feb 17 22:34:51 linux-openzaurus-collie_git.bb? Feb 17 22:35:06 koen: http://www.rpsys.net/openzaurus/temp/linux-openzaurus_git.bb Feb 17 22:35:24 koen: Its been there a while... Feb 17 22:35:29 hrw|tv: We should do that sometime Feb 17 22:36:04 yep Feb 17 22:41:15 RP: is something preventing us from using SRC_TARBALL_STASH for git as well? Feb 17 22:41:35 zecke: No, it just never got implemented Feb 17 22:42:00 RP: all that is missing is generating a name for the tar file Feb 17 22:42:13 zecke: I'd prefer to see that code factored out of the fetchers and into the core Feb 17 22:42:22 (fetcher core) Feb 17 22:42:46 /data/build/oe/downloads/git/rsync.kernel.org.pub.scm.linux.kernel.git.mburian.linux-2.6-ep93xx.git already exists.Usage: /usr/bin/git-clone [-l [-s]] [-q] [-u ] [-o ] [-n] [] Feb 17 22:43:11 that's with SRC_URI = "git://rsync.kernel.org/pub/scm/linux/kernel/git/mburian/linux-2.6-ep93xx.git;tag=ea5c136f2a0e73408800c2d596cda15e6960f766 \ Feb 17 22:43:35 hmm :-/ Feb 17 22:44:27 BitBake Build Tool Core version 1.3.3.1, bitbake version 1.3.3.2 Feb 17 22:44:39 aka svn from yesterday Feb 17 22:45:05 koen: I haven't used it for a while... Feb 17 22:45:19 already exists... Feb 17 22:45:28 sounds like it should either do rm -rf or git pull Feb 17 22:46:00 zecke: I think we actually create the directory Feb 17 22:46:18 zecke: I suspect they changed git somehow Feb 17 22:46:18 and clone into a sub directory? Feb 17 22:46:18 zecke: http://pastebin.com/560308 Feb 17 22:46:49 hehe Feb 17 22:47:12 zecke: We created the directory, then cloned into it for whatever reason Feb 17 22:47:26 RP: this doesn't work :) Feb 17 22:47:47 zecke: It used to ;-) Feb 17 22:49:04 zecke: I think we used to have a chdir issue with git where it wouldn't clone to the right place unless you chdir'd there first Feb 17 22:49:17 zecke: We're talking a much older version of git though Feb 17 22:49:42 * Philippe is away: visual contact - melancholic dreams Feb 17 22:50:55 koen: You could try http://www.rpsys.net/openzaurus/temp/bitbake_git-r0.patch Feb 17 22:51:49 that seems to be working Feb 17 22:51:59 I can see /dev/urandom again Feb 17 22:53:07 http://pastebin.com/560320 Feb 17 22:54:14 koen: did you comment out the PREFERRED_VERSION_gcc* stuff by intention yesterday? Feb 17 22:54:43 I did? Feb 17 22:54:57 you did Feb 17 22:54:57 that would be a no Feb 17 22:55:42 ok, mind reverting that part then? I don't think gcc4 is up to the job yet ;) Feb 17 22:56:08 sure, just need to wait for this git thing to finish Feb 17 23:00:11 reenoo: fix pushed, thanks for spotting Feb 17 23:00:17 'night all Feb 17 23:00:22 'night RP Feb 17 23:00:30 n8 Richard Feb 17 23:00:30 'night RP Feb 17 23:00:34 RP: could you commit that patch to bb svn? Feb 17 23:02:05 RP: could you add a test case :) Feb 17 23:02:37 Current RX Speed: 124.37 KB/s Feb 17 23:02:37 * koen taps foot Feb 17 23:05:41 360MB of source over a 1 MBit/s connection.... Feb 17 23:08:00 lol Feb 17 23:08:19 it is funny to see software one wrote in TV Feb 17 23:08:53 I never have that Feb 17 23:09:02 but I rarely write software :) Feb 17 23:09:22 (matlab code doesn't count as software IMHO) Feb 17 23:09:24 Hmm, what dependancy of gpe-image requires a kernel? Feb 17 23:09:50 everything I guess Feb 17 23:09:59 I've never seen it run without a kernel ;) Feb 17 23:11:04 * ashak grins, I was able to build opie-image without the need for it to build a kernel. As I only needed the userland side of things. I wanted to do the same for gpe-image. Feb 17 23:11:35 task-bootstrap need virtual/kernel Feb 17 23:11:43 most of images need task-bootstrap Feb 17 23:12:54 hmm, I built bootstrap-image without it building a kernel too. That relies on task-bootstrap doesn' tit? Feb 17 23:15:04 okay good nite guys Feb 17 23:15:10 'night zecke Feb 17 23:15:20 n8 zecke Feb 17 23:15:23 task-bootstrap.bb doesn't list virtual/kernel at all? Feb 17 23:16:21 03koen 07org.oe.dev * rca35bb1b... 10/conf/distro/familiar-unstable.conf: familiar-unstable: put back gcc stuff which I commented out by accident, thanks reenoo for spotting it Feb 17 23:18:32 NOTE: package ep93xx-kernel-2.6.15-ep93xx-gao26-r0: task do_fetch: completed Feb 17 23:20:49 Ah, gpe-shield depends on virtual/kernel Feb 17 23:30:30 'night all Feb 17 23:39:56 Hmm, anyone knoe if you can checkout something from a git repository without ending up with all the git files too? So that you could package up the source or something? Feb 18 00:20:49 koen|sleep: http://www.oesf.org/forums/index.php?showtopic=17787 another funny pdax thread Feb 18 02:01:29 ooo Feb 18 02:01:39 * emte wants to play with one of these http://www.proporta.com/F02/PPF02P05.php?t_id=954&t_mode=des Feb 18 02:03:39 emte: seems cool... but privoxy blocks many of the images ;) Feb 18 02:03:53 lol Feb 18 03:24:09 hmm Feb 18 03:24:18 how to generate dbus events ... Feb 18 03:24:34 i thought USB would but i guess not ... Feb 18 03:25:31 or the code is off ... Feb 18 04:11:44 hmm Feb 18 04:12:01 anyone awake who understands dbus? Feb 18 04:12:45 emte: s/awake/alive in the universe/ Feb 18 04:13:55 whoohoo Feb 18 04:13:57 :) Feb 18 04:14:14 how does the naming work? Feb 18 04:14:31 and is tehre some sort of useful list somewhere? Feb 18 04:14:35 there* Feb 18 04:14:45 oh wait Feb 18 04:14:56 you were substituting my sentance ... Feb 18 04:15:10 well i figure pb_ knos but he isnt awake Feb 18 04:15:24 since i found his name on some dbus utilities Feb 18 04:17:53 emte: i think 'em are fighting words Feb 18 04:19:15 mithro, which? Feb 18 04:21:59 [06:20] well i figure pb_ knos but he isnt awake Feb 18 04:21:59 [06:20] * dkey has quit (Read error: 110 (Connection timed out)) Feb 18 04:21:59 [06:20] since i found his name on some dbus utilities Feb 18 04:23:01 i did ... lets see if i can find em again ... Feb 18 04:23:38 aha Feb 18 04:24:31 http://dbus.freedesktop.org/doc/dbus-send.1.html http://dbus.freedesktop.org/doc/dbus-monitor.1.html Feb 18 04:24:39 hmm Feb 18 04:24:47 he is a dbus AUTHOR as well ... Feb 18 04:24:50 http://dbus.freedesktop.org/doc/AUTHORS Feb 18 07:01:56 moobie moobie Feb 18 07:11:39 kergoth: y0 Feb 18 07:52:22 interesting ... Feb 18 07:52:59 after finally getting dbus to talk to me i've discovered that CF insertion/removal is not a hotplug event Feb 18 07:54:27 now taht i got dbus talking its time for sleep Feb 18 08:32:48 NOTE: package quilt-native-0.42-r0: task do_patch: failed Feb 18 08:56:36 good morning all Feb 18 08:59:28 morning Feb 18 08:59:42 hey lamikr Feb 18 09:02:06 I I have finally time to build new gpe image, as alsa is now somehow working :-) If I download oe.db snapshot that contains dev branch, can I pull the 083 branch to it afterwards? Feb 18 09:05:58 the snapshot already contatins the branches :) Feb 18 09:07:25 to rebuild quilt-native first must remove tmp//quilt* ... Feb 18 09:07:48 Oh nice! I thought it only contained dev. But I trust you and use ctrl-c to stop pulling branches from scratch. (2173/4900 revs pulled while speeping) Feb 18 09:15:13 morning all Feb 18 09:15:54 * RP notes that cactus spines are sharp and hurt when stuck in your finger... Feb 18 09:16:59 ouch !!! Feb 18 09:19:47 hey RP Feb 18 09:20:12 koen: Did your git checkout work after that change? Feb 18 09:21:05 RP: it did Feb 18 09:21:29 NOTE: package ep93xx-kernel-2.6.15-ep93xx-gao26-r0: task do_fetch: completed Feb 18 09:21:30 NOTE: package ep93xx-kernel-2.6.15-ep93xx-gao26-r0: task do_patch: started Feb 18 09:22:08 koen: ok, I'll commit that to bitbake. There are probably some other cleanups I can make to the git fetcher as well... Feb 18 09:23:29 ok, I'll wait with committing that .bb till the patch is in Feb 18 09:24:14 RP: does ;tag= take a revision or a real tag 'foo-bar-6.4.4.' ? Feb 18 09:24:26 koen: yes Feb 18 09:25:00 * koen confused Feb 18 09:25:31 koen: To git, they're effectively the same thing Feb 18 09:25:39 ah, ok Feb 18 09:26:28 Whether anything has changed in git, I don't know but it used to work like that... Feb 18 09:30:20 it's kinda cute to see the people on the ep93xx ml and forums struggle with crosstool Feb 18 10:01:35 most of the machine conf files seems now have new define: GUI_MACHINE_CLASS = "bigscreen" or GUI_MACHINE_CLASS = "smallscreen". Should I add "smallscreen" or "bigscreen" also also h6300.conf? Feb 18 10:02:24 (h6300 has 320x200 screen) Feb 18 10:18:34 Hello all. Feb 18 10:19:32 lamikr: smallscreen Feb 18 10:22:41 Just did my first steps to test newest bitbake version in a clean environment. glibc does not unpack successfully. Any ideas ? Feb 18 10:23:15 koen: thanks. Another thing I would try here for my next build is to get "gomunicator" included to default h6300 image. How should I add that to be included in the gpe-image by default? (It seems that machine.conf files are usually only defining BOOTSTRAP apps) Feb 18 10:23:39 clear tmp/work//glibc... Feb 18 10:23:58 lamikr: see h3900.conf Feb 18 10:24:03 and tmp/stamps//glibc* Feb 18 10:24:04 gremlin[it]: started with a clean tmp dir ... Feb 18 10:24:14 ohhh Feb 18 10:24:16 uv1: rm ${DL_DIR}/libc_* Feb 18 10:24:27 koen: also done already Feb 18 10:25:12 Also if I do a manual tar tvfz libc_blahblabla.tar.gz, no error occur. Feb 18 10:26:23 BTW: cannot reach openembedd.org (provider DNS problems still). What is thhe IP-adress ? Feb 18 10:26:32 koen: I do not see any X apps specified in h3900.conf Feb 18 10:27:21 lamikr: line 40 and 41: Feb 18 10:27:25 GPE_EXTRA_DEPENDS += "gaim gpe-filemanager" Feb 18 10:27:25 GPE_EXTRA_INSTALL += "gaim gpe-filemanager" Feb 18 10:27:54 03koen 07org.oe.dev * r1bfc4705... 10/packages/glibc/glibc_2.3.5+cvs20050627.bb: glibc 2.3.5+cvs20050627: use a different name for the tarball to avoid corruption due to wget -c Feb 18 10:32:24 uv1 fetches CIA-4's glibc package Feb 18 10:32:43 koen: Oh, there were in 083 branch... h3900 branch in dev did not specify those. Feb 18 10:37:06 CIA-4: that helped ... Feb 18 10:37:23 Uv1 is going on to build a bootstrap image ... Feb 18 10:38:26 Now do we have an IP adress for OE.org ? What is the state of the documentation. The new tmp dir structure seems quit different ... Feb 18 10:39:40 uv1: the same ip as oe.handhelds.org Feb 18 10:39:45 ~nslookup oe.handhelds.org Feb 18 10:41:36 koen: Thanks koen. oe.handhelds.org is ok. Feb 18 10:41:47 ibot seem to be non-functional Feb 18 10:44:19 * lamikr started dev branch build to see the invasive changes mentioned in the oe front page. **** ENDING LOGGING AT Sat Feb 18 10:59:56 2006