**** BEGIN LOGGING AT Sat Aug 05 02:59:56 2006 Aug 05 03:33:50 i got apache config hacked and my own little ozfeed is now up on the lan :) Aug 05 03:47:13 i built busybox-1.2 but before installing it i just wanted to ask a few things Aug 05 03:49:17 i ran 'ipkg -test install busybox' and it wanted to upgrade busybox from my feed which is cool but it also wants to upgrade libgcc1 from 3.4.4 to 4.1.1 Aug 05 03:50:15 im a little nervous about doing this since i only have the one Z and i really want to know how to undo this before taking a chance on doing it Aug 05 03:52:58 so can anyone tell me, is undoing the upgrades as simple conceptually as 'ipkg install libgcc-3.4.4' ? (and of course pls correct my syntax if it's wrong) Aug 05 03:56:45 maybe ipkg remove libgcc-3.4.4 -recursive Aug 05 03:58:32 are you sure? i would be undoing the upgrade to 4.1.1 - so 'ipkg remove' as you say is counter to my intuition Aug 05 03:59:10 but i still have a great deal to learn and it wouldnt be the first time my intuition was wrong Aug 05 04:01:59 if the upgrade goes bad you may have to remove busybox and libgcc from your Z and the package feed then reinstall the originals f they are still in your feed. Aug 05 04:03:10 okay - and when upgrading/removing/installing busybox do the changes take effect immediately or only after a reboot? Aug 05 04:04:06 only after a reboot for busybox Aug 05 04:05:23 well good - kinda expected that - but just wanted to be sure Aug 05 04:05:41 i cant find libgcc1 in the oz package feeds Aug 05 04:07:26 nm found it - must be getting tired Aug 05 04:11:37 it seems fairly low risk to upgrade busybox - but i cant think of what else might be broken by such a big change to libgcc1 Aug 05 04:13:35 busynox is the wm right? Aug 05 04:14:09 i thought busybox was the command line interpreter and opie the wm Aug 05 04:14:27 nm Aug 05 04:23:55 multimode, are you running any self-built ipk upgrades? Aug 05 04:24:54 im curious to know what other folks here have found useful Aug 05 06:34:56 good morning all Aug 05 07:21:59 03justinp 07org.oe.oz354x * r70e8e4be... 10/ (8 files in 4 dirs): emacs: add emacs to .oz354x Aug 05 07:23:58 JustinP: ready when you are with the feed :) Aug 05 09:55:46 koen: ping Aug 05 10:01:40 mickeyl: hey Aug 05 10:02:02 morning folks Aug 05 10:03:15 later Aug 05 10:10:01 03mickeyl 07org.oe.dev * r03946c68... 10/ (3 files in 3 dirs): python-native 2.4.0 don't modify first shebang line when installing packages via distutils. closes #1271 Aug 05 10:10:11 03mickeyl 07org.oe.dev * rf01470a8... 10/ (5 files in 3 dirs): python-native 2.3.3 remove Aug 05 10:10:54 mickeyl: now one can't use OE to create debian stable packages :) Aug 05 10:11:16 hail zecke Aug 05 10:11:24 master! Aug 05 10:12:07 hey pb_ Aug 05 10:12:22 hail mickeyl Aug 05 10:12:23 pb_: now that we use a fast scm, want to be part of us? :) Aug 05 10:12:30 mickeyl: have you ever considered adding the ezx-crossdev patches to OE? Aug 05 10:12:54 mickeyl: you phrase it like he has an option?! Aug 05 10:13:13 zecke: not until someone confesses wanting to maintain a ezx-compatible distribution. Aug 05 10:13:18 file, at least. Aug 05 10:13:40 zecke: in fact he hasn't, but you know i'm getting more polite while growing old :D Aug 05 10:14:04 mickeyl: sure thing Aug 05 10:14:08 mickeyl: oh well, I will hack a mpkg package format :) Aug 05 10:15:03 mickeyl: my plan is to create ipkg's and then repack them as mpkg Aug 05 10:15:04 zecke: sonds good. perhaps we can drag one of those moto-ezx hackers into using OE. Aug 05 10:16:45 koen: Thanks for replying to these PTXdist people Aug 05 10:17:56 hmm I will stop idling then Aug 05 10:18:27 mickeyl: so you will be part of the apple world soon(tm)? Aug 05 10:21:24 mickeyl: did you try 'import this' yet? Aug 05 10:28:02 koen: are you registered? Aug 05 10:29:27 * JustinP goes to bed Aug 05 10:31:15 nite Aug 05 10:31:49 zecke: i hope so. I guess i'll go with a MBP 2Ghz/GB Aug 05 10:31:59 zecke: sure. import this rocks :) Aug 05 10:33:13 mickeyl: hehe Aug 05 10:48:35 anyone vetos me changing include to require? Aug 05 10:52:18 where it's approriate, sure Aug 05 10:52:26 appropriate, even Aug 05 10:52:31 mickeyl: when would it not be appropriate? Aug 05 10:52:36 hm what? Aug 05 10:52:52 in the bitbake.conf Aug 05 10:52:59 i.e. not all people have a site.conf Aug 05 10:53:16 mickeyl: ah okay Aug 05 10:53:26 mickeyl: I meant for stuff in packages/*/*.bb Aug 05 10:53:56 zecke: ya, there it should always bail out if a file is not fond. Aug 05 10:55:14 grr, found Aug 05 10:55:17 typo problems Aug 05 11:24:08 pb_: what about your key? Aug 05 11:24:21 zecke: what about it? Aug 05 11:24:53 pb_: can you live without commit access? Aug 05 11:25:33 yeah, I guess so Aug 05 11:25:40 pb_: you can? Aug 05 11:25:57 pb_: come on äl) Aug 05 11:26:31 * pb_ confused now Aug 05 11:26:43 pb_: well, I'm as well :) Aug 05 11:27:08 pb_: I would be glad if you could commit to OE yourself Aug 05 11:27:14 pb_: I would welcome it :) Aug 05 11:27:45 ah, okay Aug 05 11:27:48 I would welcome that too Aug 05 11:28:30 pb_: did you send a key to mickeyl yet? Aug 05 11:28:38 not yet Aug 05 11:28:41 how do I do that? Aug 05 11:29:07 pb_: let me find the command Aug 05 11:29:30 thanks Aug 05 11:29:32 pb_: do you have a key yet? Aug 05 11:29:34 no Aug 05 11:29:46 pb_: http://venge.net/monotone/docs/Generating-Keys.html#Generating-Keys Aug 05 11:29:49 can I make one with monotone 0.26? Aug 05 11:29:54 pb_: sure Aug 05 11:30:03 rock Aug 05 11:30:06 I'll do that now Aug 05 11:30:26 pb_: okay. e.g. pb@openembedded.org would be a good keyname Aug 05 11:30:48 phil.blundell@openembedded.org would be good as well :) Aug 05 11:30:57 ah Aug 05 11:31:01 I used pb@handhelds.org Aug 05 11:31:07 is that ok, or do I need to do it again? Aug 05 11:32:22 pb_: hmm, good question Aug 05 11:32:37 pb_: more important is not to forget the passphrase Aug 05 11:33:32 pb_: as we have raymond@edanks.com I think your keyname is okay Aug 05 11:34:14 pb_: http://venge.net/monotone/docs/Basic-Network-Service.html#Basic-Network-Service Aug 05 11:34:27 pb_: export your public key, and send it to mickeyl or koen Aug 05 11:36:33 righto Aug 05 11:37:31 pb_: and I will consult mickeyl on pb@handhelds.org vs. pb@openembedded.org as one can not easily remove the key (once you have comitted/pushed) Aug 05 11:40:46 he pb Aug 05 11:43:12 hi woglinde Aug 05 11:43:14 zecke: okay Aug 05 11:50:52 pb_: BTW: welcome back :) Aug 05 11:58:32 zecke: heh, thanks Aug 05 12:22:56 no problem with pb@handhelds.org Aug 05 12:30:05 big problem is @handhelds.org Aug 05 12:31:46 * koen points to the "#yes, that really needs to be @openembedded.org" in http://www.openembedded.org/wiki/MonotonePhraseBook Aug 05 12:33:28 i would support enforcing this rule, if we hadn't broke it already Aug 05 12:33:36 more than once as I see in monotonerc Aug 05 12:34:01 iirc only once Aug 05 12:34:03 for AMD Aug 05 12:34:31 and only because they already had stuff committed using that key Aug 05 12:35:07 *shrug* pb_, if it doesn' Aug 05 12:35:16 t bother you too much, please regen w Aug 05 12:35:21 ~lart enter-key Aug 05 12:35:22 * ibot whips out a sword and chops enter-key in half Aug 05 12:35:33 ith an @oe.org key. Aug 05 12:35:47 bbl, need to catch a couple of rays Aug 05 12:46:40 hi folks Aug 05 12:48:22 im running oz3.5.4.x on my Z - wnated to find out what sort of useful upgrades have you all been doing to your Zs from the .dev branch? Aug 05 12:50:24 for example i am upgrading busybox to 1.2 today as i know this will get rid of the bug that ending an ssh session without logging out will not leave the tty open (plus the other bugfixes that will be less apparent to me) Aug 05 12:53:28 zecke, koen, emte, kergoth, kerwood, would any of you please share ipk upgrades from .dev you found to be really useful (while not breaking other stuff)? Aug 05 13:04:42 03koen 07org.oe.dev * r955057f0... 10/ (1 packages/e17/entice_20060128.bb): entice: "esmart-x11" doesn't exist in packageland, libesmart0 does Aug 05 13:04:50 03koen 07org.oe.dev * r762266da... 10/ (1 conf/distro/angstrom-2006.9.conf): angstrom 2006.9: fix typo spotted by Philipp Zabel Aug 05 13:07:24 poushag: what does that mean? Aug 05 13:09:31 which part is unclear? im trying to find out what new ipks people are using (not just testing) on their Zs that arent in the oz images genereally available for download Aug 05 13:10:10 * koen doesn't use OZ on his Z Aug 05 13:10:23 poushag: I don't own a Z Aug 05 13:11:41 poushag: that incredibly hard to quantify, I have loads of stuff on my Z nor in oz3.5.4.1 but a lot of it is interdependant Aug 05 13:11:51 poushag: or only luck it is working at all Aug 05 13:13:01 xora - that is exactly what i am interested to know about Aug 05 13:13:26 poushag: then get building, it would take me a month to tell you all the changes I made to stock OZ Aug 05 13:14:52 i have no direction though about what to build since i dont know what is useful yet - i dont really want to build it all at once - i need to make small changes each time so i can easily identify and undo those that dont work Aug 05 13:15:04 thats why im asking this Aug 05 13:15:33 just your top 5 most useful items would be helpful for me to know Aug 05 13:19:38 man I feel dizzy Aug 05 13:23:11 poushag: your lack of direction is your hiderance, pick a direction and upgrade slowly Aug 05 13:23:34 poushag: we all have different interests, for me I have the latest sylpheed/claws/gtk+ stuff Aug 05 13:24:03 xserver-kdrive-fbdev_X11R1.1 Aug 05 13:24:08 zecke: shortly after replying I got a mail from someone that wanted NPTL stuff :) Aug 05 13:24:26 XOrA|gone: did I make a typo? Aug 05 13:25:05 XOrA|gone: or did you? Aug 05 13:25:07 well i guess my interest is in kdepim - but that seems like a long term project that i must edit/debug source to get - i am also looking to make some quicker improvements where the source code is already available Aug 05 13:25:13 s/R1/R7/ Aug 05 13:25:41 i mean available and working correctly Aug 05 13:25:53 koen: If I wouldn't feel too dizzy I would continue working on my OE toolchain extractor script Aug 05 13:25:56 koen: heh heh, me Aug 05 13:26:09 qt/e2 is close to being dead Aug 05 13:26:20 koen: earliest X I ever used was R4 I think Aug 05 13:26:24 it's already smelling that way ;) Aug 05 13:26:32 koen: well. I think it will live another two to six years :} Aug 05 13:26:45 koen: might even longer due Moto's switch of the base phone operating system Aug 05 13:27:23 even QT/e without opie niceness is bettern than the normal motorola interface design :-) Aug 05 13:29:23 XOrA|gone: right. QtE2 is old, nobody would have thought it would live that long. It is still better than the Nokia Symbian GUIs or Win32 :} Aug 05 13:33:08 h$ sed -i s/^"include "/"require "/ */*.bb Aug 05 13:33:21 GNU sed rocks Aug 05 13:40:14 after upgrading busybox to 1.2 do i need to reboot? Aug 05 13:42:12 ERROR: Could not include required file ipkg_0.99.163.bb while parsing /Users/ich/hacking/embedded/oe/org.openembedded.dev/packages/ipkg/ipkg-native_0.99.163.bb Aug 05 13:42:25 it seems like the newer busybox has changed wget and generates errors when trying to use passive-ftp Aug 05 13:49:28 poushag: that would be bad Aug 05 13:52:10 zecke: sorry about the ipkg inc, I'm testing .163 on a device right now Aug 05 13:52:35 koen: well. 0.99.159.bb is pretty empty Aug 05 13:52:46 koen: it looks like we don't need it anyway Aug 05 13:52:55 (for the native version to work correctly) Aug 05 13:52:56 zecke: .163 has no tag in cvs and needs a tar.gz Aug 05 13:53:17 so it isn't a simple 'cp' Aug 05 13:53:42 oh. okay push your changes and I will merge on my site later Aug 05 13:53:52 * zecke upgrads to quilt 0.45 and fixes libtool for OSX Aug 05 13:53:58 zecke: I want to test a few more days Aug 05 13:54:08 I doon't want to add a broken ipkg to OE *again* Aug 05 13:54:08 koen: okay! Aug 05 13:54:18 * zecke pets his Core Duo Aug 05 14:03:34 03koen 07org.oe.dev * rf2735ada... 10/ (3 files in 3 dirs): udev 092: remove BTNTOUCH for touchscreen rules to fix it for ipaqs Aug 05 14:07:04 kergoth: SRC_OPTS needs a way to specify md5sum and sha256sum Aug 05 14:23:52 re# Aug 05 14:40:41 wb Aug 05 15:12:28 koen: libtool-native is 'fixed' Aug 05 15:14:22 zecke: it is? Aug 05 15:14:32 koen: yeah ;) Aug 05 15:14:39 koen: added a '.' in base.bbclass Aug 05 15:14:46 aeh? Aug 05 15:15:04 heh Aug 05 15:15:23 woglinde: I can't say too much, the patent is pending Aug 05 15:33:36 afternoon Aug 05 15:42:10 koen: just in time Aug 05 15:42:49 yes? Aug 05 15:43:08 koen: I wait for the CIA Aug 05 15:43:24 eek Aug 05 15:43:24 find: -uid: unknown expression primary Aug 05 15:43:25 ar: illegal option -- f Aug 05 15:44:06 03freyther 07org.oe.dev * r2c7ff3c0... 10/ (42 files in 2 dirs): packages/gcc: Use require instead of include for the 'base' files Aug 05 15:44:15 03freyther 07org.oe.dev * r711cd724... 10/ (5 files in 2 dirs): Aug 05 15:44:15 packages/gcc: require the right files, add a gcc4-build-sdk (review please) Aug 05 15:44:15 The gcc 4.1.x sdk tries to require gcc4-build-sdk.inc which Aug 05 15:44:15 doesn't exist. Copy the gcc3-build-sdk.inc and ask florian to Aug 05 15:44:16 verify it. Aug 05 15:44:21 03freyther 07org.oe.dev * r33c559bf... 10/ (19 files in 2 dirs): Aug 05 15:44:21 packages/glibc: Use require for the glibc files Aug 05 15:44:22 Use require for the glibc files Aug 05 15:44:28 03freyther 07org.oe.dev * rc9326b6a... 10/ (20 files in 2 dirs): packages/binutils: Use require for the base files Aug 05 15:44:34 03freyther 07org.oe.dev * r07a30002... 10/ (10 files in 2 dirs): packages/automake: Use require instead of include to get the base files Aug 05 15:44:42 03freyther 07org.oe.dev * rfd003213... 10/ (1279 files in 369 dirs): Aug 05 15:44:42 packages: sed -i s/^"include "/"require "/ */*.bb GNU sed rocks! Aug 05 15:44:42 Use require for the BitBake files within packages. In contrast Aug 05 15:44:42 to the conf files the authors of these files expected these files Aug 05 15:44:43 to be included. Aug 05 15:44:52 03freyther 07org.oe.dev * r70f685f7... 10/ (1 packages/ipkg/ipkg-native_0.99.163.bb): Aug 05 15:44:52 packages/ipkg/: There is no ipkg_0.99.163.bb do not require it Aug 05 15:44:52 There is no 0.99.163.bb for ipkg, do not try to require it. Aug 05 15:44:53 The other ipkg_0.99.15x.bb files only add patches and include Aug 05 15:44:55 ipkg.inc which we do already. I assume we have no patches to Aug 05 15:44:57 ipkg 0.99.163 Aug 05 15:44:59 03freyther 07org.oe.dev * r0439f1ab... 10/ (3 files in 2 dirs): Aug 05 15:45:01 packages/unionfs: We only have a unionfs.inc which is not versioned include it Aug 05 15:45:03 Follow the 1.0.13 bb files and require the unversioned unionfs.inc Aug 05 15:45:09 03freyther 07org.oe.dev * r71d88321... 10/ (1 classes/base.bbclass): (log message trimmed) Aug 05 15:45:11 classes/base.bbclass: Make find call portable Aug 05 15:45:13 Make the find invocation portable (BSD) by specifying '.' Aug 05 15:45:15 as the to be searched directory. Aug 05 15:45:17 The error on libtool-native was: Aug 05 15:45:19 find: illegal option -- n Aug 05 15:45:21 find: illegal option -- a Aug 05 15:45:23 03freyther 07org.oe.dev * r210bc533... 10/ (18 files in 5 dirs): (log message trimmed) Aug 05 15:45:27 packages/quilt: Add quilt version 0.45, 'cleanup'/complicate layout Aug 05 15:45:29 Quilt is the first package we will build, so special care must Aug 05 15:45:31 be taken. This is one reason why we do not include/require quilt_${PV}.inc Aug 05 15:45:33 directly. Aug 05 15:45:35 Now the -native packages require the quilt-native.inc package which Aug 05 15:45:37 adds the non-gnu patch for darwin/BSD. In quilt_PV.inc we will aplly Aug 05 15:45:39 03freyther 07org.oe.dev * r61c6cd86... 10/ (1 conf/bitbake.conf): Aug 05 15:45:43 conf/bitbake.conf: quilt 0.45 lacks -n but has -P for setting a patch name Aug 05 15:45:45 quilt 0.45 lacks the -n option but has -P nowadays for setting Aug 05 15:45:47 the to be used patchname, update. Aug 05 15:46:21 ah Aug 05 15:46:34 zecke: I hit the 'no -n' today Aug 05 15:46:53 (debian updated quilt to 0.45 and comes first in $PATH) Aug 05 15:47:12 koen: hmm, it shouldn't come first Aug 05 15:49:56 armv5te-darwin armv5te-linux i386-darwin Aug 05 15:49:59 hmm Aug 05 15:51:39 ~lart gutenprint configure.ac Aug 05 15:51:40 * ibot takes a large goose feather pillow and swings it wildly in gutenprint configure.ac's direction, hitting gutenprint configure.ac and sending gutenprint configure.ac flying into the closet Aug 05 15:59:14 zecke, yeah im not too happy about the wget errors - any suggestions? all i can think of is to try and install the full wget from the feed Aug 05 15:59:57 poushag: the wget from busybox 1.2.1 should work Aug 05 16:01:04 thx koen - is 1.2.1 in .dev now? Aug 05 16:01:20 poushag: ls packages/busybox Aug 05 16:01:26 zecke: I hate svn: Aug 05 16:01:28 bitbake:/data/build/oe/bitbake# svn revert --recursive . Aug 05 16:01:29 svn: Try 'svn revert --recursive' instead? Aug 05 16:01:39 bitbake:/data/build/oe/bitbake# svn revert --recursive Aug 05 16:01:40 revert: Restore pristine working copy file (undo most local edits). Aug 05 16:01:40 usage: revert PATH... Aug 05 16:01:52 ~lart drepper Aug 05 16:01:53 * ibot throws drepper's poor little doggy off a cliff Aug 05 16:02:05 yeah koen i listed that dir yesterday - maybe i missed it - trying again Aug 05 16:02:46 nope - i only have 1.2.0 Aug 05 16:03:02 so your OE is outdated Aug 05 16:03:06 i ran mtn update yesterday too Aug 05 16:03:32 poushag: http://www.openembedded.org/viewmtn/revision.psp?id=de17790d25bccd7877ddbfc36bcb3fa5e2212b6f Aug 05 16:03:40 "4 days, 2 hours ago" Aug 05 16:03:43 i need a better way to keep up with whats being changed in the branches - did you say you use an rss feed? Aug 05 16:04:36 pb_: in which variable will --target=foo land on autoconf? Aug 05 16:04:42 mtn: already up to date at 250dd8a89e640f220e44c8954b77278fe265764b - mtn said i was up to date yesterday - whats wrong? Aug 05 16:04:50 poushag: pull first? Aug 05 16:05:13 * poushag smacks forehead - still too new at this Aug 05 16:10:26 is rss feed the best way to stay informed about changes to .dev branch? Aug 05 16:13:00 it's probably the easiest way Aug 05 16:13:42 ok never used rss reader before - seem to recall maybe ff has one Aug 05 16:14:59 or maybe it was safari... Aug 05 16:15:03 try opening http://www.openembedded.org/aggregator/rss Aug 05 16:15:16 firefox and safari should both have an rss reader Aug 05 16:16:10 worked in safari - i must not have the ff extension Aug 05 16:16:16 thx koen Aug 05 16:18:23 hmm - rss reader seems to be giving timestamps in utc rather than local time Aug 05 16:18:53 koen: DISTRO = "mvmd" <- ideas on what that could mean? Aug 05 16:19:55 yes Aug 05 16:20:01 good Aug 05 16:20:19 it's our mission-statement Aug 05 16:20:24 hehe Aug 05 16:23:00 oops Aug 05 16:23:12 I think I mixed EABI and OABI toolchains Aug 05 16:23:27 ~lart OABI crt1.o Aug 05 16:23:27 * ibot chops OABI crt1.o in half with a free Solaris 7 CD Aug 05 16:28:19 wtf, my patch fails to apply now Aug 05 16:29:34 what's OABI ? Aug 05 16:29:42 ole abi Aug 05 16:29:42 gremlin[it]: Old ABI Aug 05 16:29:54 e.g. what 99% of the world is using now Aug 05 16:30:34 ok :) Aug 05 16:30:41 thanks :) Aug 05 16:32:13 * koen wipes TMP Aug 05 16:33:56 * chouimat looks at the grocery store flyer ... Aug 05 16:33:59 koen: that sounds like a test case for insanity.bbclass Aug 05 16:34:49 zecke: objdumping binaries would be neat Aug 05 16:35:19 koen: you mean $TARGET_ARCH-$TARGET_OS-$TARGET_VENDOR-objdumping right? Aug 05 16:35:33 yes Aug 05 16:35:43 otherwise if will fail silently Aug 05 16:35:48 koen: for a second I thought were like a glibc moron :) Aug 05 16:35:57 place 'you' somehwere Aug 05 16:36:03 ~lart drepper Aug 05 16:36:04 * ibot duct-tapes drepper to the floor and drools on him Aug 05 16:39:10 pb_: I need GNU libc help (social networking help) Aug 05 16:39:34 pb_: autoreconf on GNU libc will fail miserable. but it does stupid checking for host utilities Aug 05 16:39:57 pb_: e.g. I have no binutils, elf utils on my system installed, as I'm not on an ELF system and not using binutils Aug 05 16:40:27 pb_: but glibc is checking for stuff like readelf, which in our OE case is prefixed with arm-linux- Aug 05 16:41:27 pb_: how can glibc be cured and how to get patches upstream?! Aug 05 16:42:20 zecke: kill ulrich and assume his identity Aug 05 16:42:48 koen: actually his shoes would be too big for me Aug 05 16:43:30 zecke: wear extra socks Aug 05 16:43:58 koen: I won't fit, even with two pairs of winter socks Aug 05 16:46:59 hopefully I have gutenprint working today so I can try finishing my SoC tomorrow Aug 05 16:47:12 you want to be finished by tomorrow? Aug 05 16:47:21 ASAP Aug 05 16:47:26 only ~9 days left Aug 05 16:47:27 then it is time to bitch about it Aug 05 16:47:31 oh Aug 05 16:47:38 * zecke opens the calendar Aug 05 16:48:03 well, more days are left, but I'll be working from 15 - 21 augustus Aug 05 16:48:25 without internet access Aug 05 16:48:28 okay, then I will take a closer look tomorrow and bitch about it Aug 05 16:48:33 thanks Aug 05 17:44:24 yay! Aug 05 17:44:31 my usb serial adaptor works on osx Aug 05 17:44:42 ~lart fink for not having 'cu' Aug 05 17:44:43 * ibot decapitates fink conan the destroyer style for not having 'cu' Aug 05 17:45:29 koen: I hope to write some USB ethernet drivers for OSX Aug 05 17:45:58 lol I needed to put LDFLAGS into EXTRA_OEMAKE... Aug 05 17:46:08 zecke: the ajzaurus ones worked for me Aug 05 17:47:02 koen: well, can it do the stuff usbnet.ko does? Aug 05 17:47:27 I can ssh into my toys Aug 05 17:47:35 that's all it needs to do for me Aug 05 17:48:49 stupid expensive keyspan adaptors Aug 05 17:49:40 ~lart drepper Aug 05 17:49:41 * ibot throws a AN/M-8 smoke grenade at drepper Aug 05 17:53:26 reboot, rb Aug 05 18:02:25 re koen Aug 05 18:02:25 re Aug 05 18:02:30 wb Aug 05 18:03:29 woglinde: Al+Her visit+me will eat pizza at the malheur Aug 05 18:03:34 woglinde: do you feel like joining? Aug 05 18:04:14 whois Her? Aug 05 18:04:20 when? Aug 05 18:04:30 "al's visit" Aug 05 18:04:49 she is called Alice Aug 05 18:04:50 hi all ... doing update of .dev i got 'connection refused' ... something wrong or ise related with changes to monotone 0.27 ? Aug 05 18:04:55 I have to write this damned paper Aug 05 18:05:05 and I cant motivate myself Aug 05 18:05:13 woglinde: you are still not done? Aug 05 18:05:15 zecke I am not Bob Aug 05 18:05:19 zecke nebenfach Aug 05 18:05:49 woglinde: well write it! get your "Vordiplom" :) Aug 05 18:06:00 zecke yes I know Aug 05 18:06:07 woglinde: I'm slowly starting with GTI... Aug 05 18:06:53 hehe, Reading the GNU hello ChangeLog is fun Aug 05 18:07:06 "Karl Eichwalder" Aug 05 18:07:13 he starts wih ke@ke.Central.de Aug 05 18:07:17 moves to gwdg.de Aug 05 18:07:20 then to suse.de Aug 05 18:07:33 then gnu.franken.de Aug 05 18:07:48 and then ends with @gmx.net Aug 05 18:07:53 funny Aug 05 18:09:33 morning Aug 05 18:09:41 hey eFfeM Aug 05 18:09:55 hi all ... doing update of .dev i got 'connection refused' ... something wrong or is related with changes to monotone 0.27 ? Aug 05 18:10:14 gremlin[it]: do you use '*' as glob? Aug 05 18:10:35 zecke no Aug 05 18:10:54 gremlin[it]: pastebon the complete error please Aug 05 18:11:05 # Check for readelf Aug 05 18:11:05 AC_CHECK_TARGET_TOOL([READELF],[readelf],[jdklwj],[$PATH]) Aug 05 18:11:17 I'm so elite, Mr. Drepper is too stupid to write this line... Aug 05 18:11:58 zecke is just a line : >>> monotone: network error: failed to connect: Connection refused <<< Aug 05 18:12:07 hmm Aug 05 18:12:19 gremlin[it]: it worked half an hour before Aug 05 18:12:43 gremlin[it]: did you follow gettingstarted? Aug 05 18:12:48 i have monotone 0.25.2 Aug 05 18:12:56 lol Aug 05 18:13:06 and comamnds are : monotone --db=/media/hd1/oe/OE.db pull monotone.vanille.de org.openembedded.dev Aug 05 18:13:09 gremlin[it]: and pulling from monotone.vanille.de? Aug 05 18:13:11 gremlin[it]: try reading http://openembedded.org :) Aug 05 18:13:12 lol Aug 05 18:13:48 but ... on oe mailinglist is announced the changes on gettingstarted ??? Aug 05 18:14:04 gremlin[it]: on OE mailinglist the migration was announced at least six times Aug 05 18:14:39 migration to monotne 0.27 ... Aug 05 18:15:16 gremlin[it]: go to venge.net/monotone and read all about monotone 0.26 Aug 05 18:15:25 go directly for .28 gremlin[it] Aug 05 18:15:35 gremlin[it]: monotone 0.26, 0.27, 0.28 are compatible and work nicely Aug 05 18:19:01 zecke: ipkg .163 shows no obvious flaws, but i'll continue testing Aug 05 18:19:28 just another question : i have to dowload the database or i can 'convert' the one i already have from monotone 0.25 ? Aug 05 18:19:53 gremlin[it]: read http://openembedded.org :) Aug 05 18:19:56 gremlin[it]: do me a favor Aug 05 18:20:03 gremlin[it]: read one of the six mails Aug 05 18:20:11 gremlin[it]: or check the news items at www.openembedded.org Aug 05 18:25:23 i read ... but sincerelly still have a doubt ... there is "will need to get a new database" ... but also istruction to upgrade database ... Aug 05 18:34:28 hey mallum Aug 05 18:34:41 hey koen Aug 05 18:35:01 mallum: is clutter running on you cm270 yet? Aug 05 18:36:53 koen: no not yet, mainly because the GL ES clutter 'backend' is not done yet Aug 05 18:37:22 mallum: I spot MB 1.1 releases Aug 05 18:38:05 koen: as promised :) Aug 05 18:38:14 pb_: are you around ? Aug 05 18:41:11 mallum: now I just need to wait for OE to finish the toolchain building Aug 05 18:41:18 will push mb 1.1 after that Aug 05 18:41:21 cool Aug 05 18:41:28 and libmb 1.9 Aug 05 18:42:58 mallum: can mb-desktop use XDAMAGE to redraw damaged bits? Aug 05 18:44:08 koen: you use expose events for that, damamge is different Aug 05 18:45:10 mallum: I noticed that the redrawing is suboptimal on the netbook Aug 05 18:45:24 huge screen, 16bit small bus... Aug 05 18:46:02 if I move a window when using fluxbox as a wm the outlines stay visible till I trigger a complete redraw Aug 05 18:46:18 which takes ~0.5 seconds Aug 05 18:46:42 I never noticed before, since I always use mb-wm :) Aug 05 18:48:31 koen: its probably the server driver + slow hardware Aug 05 18:49:01 mallum: no doubt Aug 05 18:49:17 but the background 'corruption' when moving windows is annoying Aug 05 18:49:42 sounds like a driver issue Aug 05 19:03:07 anyone know how i can specify which files to include in a package? Aug 05 19:03:25 with busybox i get "NOTE: the following files were installed but not shipped in any package:" Aug 05 19:04:27 furlongm: in the case of busybox you can ignore those warnings Aug 05 19:04:37 because they're symlinks? Aug 05 19:04:47 furlongm: in general it's FILES_${PN} += "" Aug 05 19:04:57 furlongm: because we have a custom do_install Aug 05 19:05:11 ah, ok Aug 05 19:05:18 thanks Aug 05 19:08:34 lumdidum Aug 05 19:22:08 hi andersee Aug 05 19:22:42 woglinde: howdy Aug 05 19:22:54 woglinde: man, gcc and glibc suck :} Aug 05 19:24:39 zecke take fefe-cc and dietlibc Aug 05 19:25:08 woglinde: he wrote a compiler? Aug 05 19:25:11 no Aug 05 19:25:14 just kidding Aug 05 19:25:26 woglinde: relieved Aug 05 19:25:39 woglinde: I would feel bad if someone not knowing the C standard writes compilers... Aug 05 19:26:04 zecke: I think OE needs to use clibrary/PI and compiler/PI Aug 05 19:26:22 koen whats that? Aug 05 19:26:31 ~emulate zautrix Aug 05 19:26:32 I wait for Theo to start OpenCC Aug 05 19:26:33 this is exactly what the GPL is about. grab the stuff and run away Aug 05 19:26:48 zecke: there's already a BSD compiler Aug 05 19:26:57 zecke: I may have written a C lib, but I have no plans to write a CC Aug 05 19:27:10 andersee: I'm fine with the gcc itself Aug 05 19:27:19 andersee: could you paste the url for the nethack autofoo again? Aug 05 19:27:23 andersee: just running /usr/bin/cpp to check if I have filio.h installed... Aug 05 19:27:30 andersee: when cross compiling makes me mad... Aug 05 19:28:11 zecke: the qemu guy wrote a CC -- http://fabrice.bellard.free.fr/tcc/ Aug 05 19:28:36 andersee: This is how I would have written a compiler before starting university :) Aug 05 19:28:49 andersee: manually C code to generate instructions... Aug 05 19:29:28 but besides that Fabrice has an incredible output... Aug 05 19:29:41 koen: I'm not certain it is still on the net anywhere... looking... Aug 05 19:29:43 qemu, tcc, dvb, a lot of code in ffmpeg Aug 05 19:29:58 koen: I have a copy locally I could send you Aug 05 19:30:33 andersee: I might have some time to try adding it to OE in a few weeks :) Aug 05 19:30:44 koen: that would be sweet! Aug 05 19:30:45 hm Aug 05 19:30:46 just add '#!/usr/local/bin/tcc -run' at the first line of your C source, Aug 05 19:31:14 woglinde or write a bimfmt handler Aug 05 19:32:04 anyone ever looked at lcc ? Aug 05 19:32:07 maybe tcc already support ABI?! Aug 05 19:32:32 haven't followed it recently myself, but have played with it quite some years ago. good book too Aug 05 19:32:52 koen: seems you can still fetch it from the gnome cvs repository Aug 05 19:32:56 koen: http://cvs.gnome.org/viewcvs/nethack/ Aug 05 19:33:22 koen: in full auto* glory Aug 05 19:33:49 koen: geez -- 6 years since I even touched the thing Aug 05 19:36:58 andersee: died from boredom by not playing nethack for +6 yrs Aug 05 19:36:59 andersee: thanks Aug 05 19:38:30 koen: should be able to get a usable diff vs nethack 3.2.3, then forward port the auto* changes to nethack 3.4.3 Aug 05 19:39:57 mallum: new mb bits are in Aug 05 19:40:00 03koen 07org.oe.dev * r35859e17... 10/ (1 packages/cups/cups_1.1.23.bb): cups: rrecommend doc package for the webgui Aug 05 19:40:10 03koen 07org.oe.dev * r5cfcf658... 10/ (1 packages/matchbox-wm/matchbox-wm_1.1.bb): mb-wm: add 1.1 Aug 05 19:40:15 03eFfeM 07org.oe.dev * r2ca01af5... 10/ (1 packages/linux/ixp4xx-kernel_2.6.17.bb): Aug 05 19:40:15 ixp4xx-kernel_2.6.17.bb: bumped PR_CONFIG Aug 05 19:40:15 to force a recompile on the autobuild system using the upgraded gcc Aug 05 19:40:15 this gives a working snd_pcm module Aug 05 19:40:24 03koen 07org.oe.dev * r7f3bc3b8... 10/ (1 packages/libmatchbox/libmatchbox_1.9.bb): libmatchbox: add 1.9 Aug 05 19:40:40 koen: cool :) let me know if any issues Aug 05 19:46:35 mallum: will do Aug 05 19:46:53 mallum: http://bugs.o-hand.com/ is broken Aug 05 19:47:18 bugzilla.o-hand.com Aug 05 19:48:19 damn, I start to like autotools Aug 05 19:48:24 http://rafb.net/paste/results/n5O6vM34.html Aug 05 19:48:46 ~lart glibc for broken autotools usage Aug 05 19:48:46 * ibot takes out a cattle prod and gives glibc a good jolt for broken autotools usage Aug 05 19:49:16 mallum: could you redirect bugs.o-hand to that? Aug 05 19:53:12 03koen 07org.oe.dev * re33107ae... 10/ (1 packages/cups/cups_1.1.23.bb): cups: ship files needed for the webgui in the main package Aug 05 19:57:00 koen: removed it from dns Aug 05 19:57:48 mallum: thanks Aug 05 19:58:26 | /bin/sh: line 4: ./cups-genppd.5.0: cannot execute binary file Aug 05 19:58:32 ~lart gutenprint Aug 05 19:58:32 * ibot makes a balloon animal out of gutenprint Aug 05 19:59:26 koen: iirc I can map it to bugzilla for some wierd bugzilla reason Aug 05 20:04:24 03koen 07org.oe.dev * rf376302d... 10/ (1 packages/cups/cups_1.1.23.bb): cups: stage files needed for gutenprint Aug 05 20:15:25 koen, are using glibc-2.3.2 ? Aug 05 20:15:48 rob_w: no, 2.4 on most machines and 2.3.5 on the netbook-pro Aug 05 20:16:12 wow .. there is a 2.4 ?? cool Aug 05 20:16:20 hehe Aug 05 20:16:38 i wonder why my builds still take 2.5.3cvs-something Aug 05 20:19:44 koen whats the status with arm4 and eabi? Aug 05 20:23:39 woglinde: patch gcc to emit two instructions Aug 05 20:24:00 woglinde: disable thumb interworking, check the debian wiki :) Aug 05 20:24:45 zecke hm glibc needs patches too, if I remeber correctly what koen said Aug 05 20:24:50 woglinde: OE's gcc sucks, martin guys one doesn't Aug 05 20:25:24 OE's gcc always generated armv5 isns (clz) even if you tell it to build for an armv4 cpu Aug 05 20:26:11 koen: hehe, Martin Guy must be a god or such Aug 05 20:27:07 koen: so with martin guys' gcc my Shark can have EABI? Aug 05 20:27:08 zecke: no, Dan Kegel is Aug 05 20:27:14 hehe Aug 05 20:27:34 "I have, and it doesn't contain any clz's. That is for the armv4t toolchain under Aug 05 20:27:34 freaknet.org/martin/crosstool/ " Aug 05 20:27:59 his answer to "does your libc6.so have armv5 isns?" Aug 05 20:28:00 hm? Aug 05 20:28:50 woglinde: http://www.freaknet.org/martin/crosstool/packages/ Aug 05 20:29:00 woglinde: I don't see an offer for the source? do you? Aug 05 20:29:17 hm Aug 05 20:29:22 case for harald? Aug 05 20:29:46 iirc ARM is paying him and wookey to get EABI into debian Aug 05 20:29:53 so sue arm ;) Aug 05 20:30:09 armv4t Aug 05 20:30:11 yeah, they will continue sponsoring linux anyway Aug 05 20:30:18 but simpad has no thumb Aug 05 20:31:04 maybee the patches are inside the archive Aug 05 20:31:25 zecke, woglinde: lennert discoverd that 'bx' causes a pipeline flush so the workaround mentioned on the debian wiki is 5 instructions faster (best case) or the same number of instructions (worst case) Aug 05 20:31:49 so the usage of 'bx' is only there to push ARM thumb patents down our throats Aug 05 20:32:12 hm pipeline flush is bad Aug 05 20:32:17 If remeber correctly Aug 05 20:32:21 yes Aug 05 20:32:26 *g* Aug 05 20:32:30 at least it is clean afterwards Aug 05 20:32:33 especially with the 7 stage long one on xscales Aug 05 20:33:58 woglinde: hmm this is a GPL violation Aug 05 20:34:07 zecke as I said Aug 05 20:34:14 woglinde: I don't see the patchs, I don't see the list of stuff used to build it Aug 05 20:34:19 so the 'tst lr, #1; moveq pc, lr; bx lr' workaround should be accepted into mainline gcc Aug 05 20:34:45 zecke if we wrote him email and tell him this in a kindly manner? Aug 05 20:35:03 http://rafb.net/paste/results/yOGiDY46.html Aug 05 20:35:13 woglinde: I'm sure there are patches... Aug 05 20:35:43 dear mr. guy, maybee I am to dull to find the patches for the GPL-licenes stuff you put at download at your website Aug 05 20:36:07 woglinde: dear mr. guy, could you please point me to the written offer for the sourcecode Aug 05 20:36:17 woglinde: source is not within the tar Aug 05 20:36:29 zecke will write you or me? Aug 05 20:36:41 woglinde: please, do me a favor and write him Aug 05 20:37:09 http://www.freaknet.org/martin/crosstool/crosstool-0.42-mg2-mg3.patch Aug 05 20:37:10 ? Aug 05 20:38:02 03koen 07org.oe.dev * r0ad856b6... 10/ (3 files in 3 dirs): gutenprint: add to nonworking, needs to run a native tool, qemu doesn't do EABI so I can't fix it :( Aug 05 20:38:50 Input handle pointer is NULL! Aug 05 20:38:52 *** NAUGHTY PROGRAMMER!!! Aug 05 20:38:52 *** SPANK SPANK SPANK!!! Aug 05 20:38:52 :) Aug 05 20:39:02 lol Aug 05 20:39:31 woglinde: smells like all patches are upstream?! Aug 05 20:40:51 zecke hm you could try to build with crosstool Aug 05 20:40:59 und look if it is the same Aug 05 20:41:17 from Martins track record so far I'm fairly certain those toolchains have no unique patches Aug 05 20:42:21 hm so why the oe gcc fails Aug 05 20:43:03 woglinde: "The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable" Aug 05 20:43:15 I'm not a lawyer, but I always wonder what this means Aug 05 20:43:40 zecke: it means "*** NO SOURCE *** SPANK SPANK SPANK!!!" Aug 05 20:43:46 as raster would say it Aug 05 20:44:19 hm to be fair he offer the patches for the tool to build it Aug 05 20:45:33 this has developed into the wrong direction Aug 05 20:45:53 I think my point is. Everybody, including me, is likely to violate the GPL :} Aug 05 20:46:41 woglinde: but this paragraph means. I do not need to distribute linux+glibc headers if I give my list of installed packages? Aug 05 20:48:04 zecke: it's really simple: you need to distribut sources of all GPL binaries you distribute Aug 05 20:48:31 nchip: + Makefiles + scripts (or do they count as source already)? Aug 05 20:48:44 arm suckz *g* Aug 05 20:48:54 zecke: however, in practice if the sources are not modified and you can figure out what the versions of the tools are, it's probably not worth rising a stink about the issue Aug 05 20:49:09 hmmm Aug 05 20:49:18 e.g. when I hacked the tinderbox code, I came across rpm and dpkg calls to list all installed packages Aug 05 20:49:21 actually martin *is* violationg the GPL Aug 05 20:49:30 he said to me he's using http://www.openembedded.org/repo/org.openembedded.dev/packages/glibc/glibc-2.4/fixup-aeabi-syscalls.patch Aug 05 20:49:44 koen *g* Aug 05 20:49:50 but I can't find a reference to such patch on his page Aug 05 20:50:01 e.g. I just violated the GPL for not marking the sections I have changed Aug 05 20:50:02 woglinde: not that patch, but the original posted to the mailinglist Aug 05 20:50:16 I didn't add my name+reason to the files Aug 05 20:50:22 I'm a GPL violator :( Aug 05 20:51:28 woglinde: you could try asking martin if he uses http://sourceware.org/ml/libc-ports/2006-03/msg00029.html and if yes, why it isn't mentioned on his site Aug 05 20:51:45 koen: it's in the patch woglinde pasted Aug 05 20:52:30 ah, I missed that Aug 05 20:52:44 yes Aug 05 20:52:48 its on the side Aug 05 20:53:04 the name "Extra patches to enable linux-2.6.16 and ARM1026 CPU" eluded me Aug 05 20:53:12 so gcc4.1.1 already has ArmV4 support Aug 05 20:53:24 hm Aug 05 20:53:36 drat Aug 05 20:53:36 it has eabi support for armv4t Aug 05 20:53:41 thats for sure Aug 05 20:53:43 now I remember Aug 05 20:54:02 the gcc people 'fixed' EABI in gcc 4.1.1 Aug 05 20:54:43 woglinde: is your uni mail account working? Aug 05 20:54:44 I bet that 'fix' broke armv4 Aug 05 20:54:56 zecke hm yes why? Aug 05 20:55:19 woglinde: I'm waiting for two pw's for GNU bugzillas Aug 05 20:55:47 .oO( a small kingdom for openID support on bugzilla ) Aug 05 20:56:16 nchip: Having one GNU bugzilla would be nice Aug 05 20:56:26 nchip: but glibc and gcc have two different... Aug 05 20:57:05 zecke last email from 11 p.m. Aug 05 20:57:26 woglinde: lol inf.fu-berlin.de is now public :} Aug 05 20:58:09 zecke haeh? Aug 05 20:58:24 woglinde: it used to print "This is not a public server" Aug 05 20:58:30 right Aug 05 20:58:38 think about why Aug 05 20:58:42 ah Aug 05 20:58:43 woglinde: one can login now Aug 05 20:58:44 okay Aug 05 20:58:47 oh Aug 05 20:59:00 maybee thats the new telnet.inf... Aug 05 20:59:24 woglinde: how secure is sub os against forkbombs? Aug 05 20:59:35 not that I would run one Aug 05 21:00:01 woglinde: oh, gcc has OpenMP support Aug 05 21:00:14 puh what was openmp again? Aug 05 21:00:20 my head is like a sieb Aug 05 21:00:30 woglinde: think of Rechnerarchitektur Aug 05 21:00:47 puh Aug 05 21:00:47 woglinde: and 'big' machines, and how to program for them efficiently Aug 05 21:01:00 has it to do with caches? Aug 05 21:01:09 woglinde: you have taken the exam, while I haven't :} Aug 05 21:01:10 woglinde: no Aug 05 21:01:20 he its over one year Aug 05 21:01:23 past Aug 05 21:01:24 woglinde: www.openmp.org Aug 05 21:02:53 hm Aug 05 21:02:53 oh both GNU bugzillas are in the spam folder Aug 05 21:02:58 hehe Aug 05 21:02:59 lol Aug 05 21:03:02 html code Aug 05 21:03:04 i guess Aug 05 21:03:10 woglinde: no Aug 05 21:03:19 or the new fascist spamkiller Aug 05 21:03:25 which the postmaster announced Aug 05 21:03:40 maybe that mail went in your spam-folder too Aug 05 21:03:42 ahahah Aug 05 21:04:36 woglinde: hehe, no I have my own spamkiller Aug 05 21:05:03 http://sourceware.org/frysk/ Aug 05 21:05:03 so you dont use imap Aug 05 21:05:09 I do use imap Aug 05 21:05:32 hm okay Aug 05 21:05:48 so we should test eabi for armv4 now Aug 05 21:06:01 to came back to this topic Aug 05 21:07:38 woglinde: this isn't what pbrook told me some weeks ago Aug 05 21:07:47 but gcc4.1.1 might have a 'fix' Aug 05 21:07:58 woglinde: you could browse the ChangeLog of the 4.1 branch?! Aug 05 21:08:07 I will study the changelog yes Aug 05 21:08:16 thanks Aug 05 21:09:21 hm Aug 05 21:09:26 nothing about eabi Aug 05 21:10:56 will look at svn Aug 05 21:13:10 hms Aug 05 21:13:22 *wonder* no gcc-4.1 release tag Aug 05 21:13:22 zecke: I think the 'fix' in gcc 4.1.1 broke non-armv5 eabi Aug 05 21:14:55 hm are the using a new repository? Aug 05 21:15:11 args Aug 05 21:17:50 koen, missed 10 min's due to router problem. you are not talking about the 4.1.1. fix that I added yesterday, are you? Aug 05 21:18:12 eFfeM-log: no, about a fix in gcc upstram Aug 05 21:18:53 ok, thanks, I feared I goofed, although I did not understand why 'my' fix would have any influence on eabi (as it was related to cse) Aug 05 21:20:00 http://sourceware.org/bugzilla/show_bug.cgi?id=3004 Aug 05 21:20:04 my first glibc bug Aug 05 21:20:41 zecke, congrats ..... Aug 05 21:20:46 calling it a day Aug 05 21:20:50 hehe Aug 05 21:20:53 03eFfeM 07org.oe.dev * r96934e2e... 10/ (4 files in 3 dirs): pvrusb2-mci: added 20060726 snapshot Aug 05 21:21:15 and I have an OE based toolchain on my OSX/i386 system Aug 05 21:21:28 zecke: cool Aug 05 21:21:45 still gcc3.3.6 and glibc2.3.2 for moto devices :} Aug 05 21:21:49 zecke: sounds like I really need to install monotone .28 on my powerbook :) Aug 05 21:21:58 MACHINE="moto" DISTRO="mvmd" Aug 05 21:22:17 or should we call it OpenMVmD Aug 05 21:23:19 please stop with the 'open' Aug 05 21:23:28 hehe Aug 05 21:23:34 but we all love OpenBSD?! Aug 05 21:23:38 that as annoying as misspelling everything with a 'k' or 'g' Aug 05 21:23:52 or sticking /PI after the name Aug 05 21:23:54 Open rockz Aug 05 21:24:01 pb_: you are everywhere! Aug 05 21:24:04 has nothing to do with free Aug 05 21:24:05 pb_: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=5688 Aug 05 21:24:32 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26046 Aug 05 21:24:36 and schurig too Aug 05 21:26:58 zecke: glibc has a history of broken autofoo usage Aug 05 21:27:13 OE hit an infinite loop a while back Aug 05 21:27:21 bugs + fix know for years.... Aug 05 21:27:25 known* Aug 05 21:27:29 after adding -isystem Aug 05 21:28:16 heh Aug 05 21:28:31 I suspect no-one tried to play mp3s with the xmms in OE on arm Aug 05 21:28:46 I dont like xmms Aug 05 21:29:47 koen: We have experimented with xmms-embedded at ROAD Aug 05 21:30:26 hu daychange Aug 05 21:30:42 zecke: xmms-normal doesn't seem to know about libmad Aug 05 21:31:06 woglinde: you are eight minutes early Aug 05 21:31:13 (when bitbaked) Aug 05 21:31:22 hu Aug 05 21:31:24 hms Aug 05 21:31:31 ~lart ntp Aug 05 21:31:31 * ibot takes a big bite out of ntp's jugular vein Aug 05 21:35:03 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28615 Aug 05 21:37:39 zecke: are those in OE already? Aug 05 21:39:15 koen: no Aug 05 21:39:31 koen: I'm trying to make up my mind with which gcc/glibc combinations to start with Aug 05 21:39:40 man I'm confused today Aug 05 21:44:57 oh quilt fails on NSLU2 Aug 05 21:45:12 zecke: clobber bug? Aug 05 21:45:21 it worked for me with a clean rebuild Aug 05 21:45:26 no stinks like failiting autoreconf Aug 05 21:45:31 quilt vs. quilt-native Aug 05 21:45:53 I need to fix tinderclient.bbclass with regard to the -k flag Aug 05 21:45:57 of bitbake Aug 05 21:48:53 oops Aug 05 21:48:55 thunder Aug 05 21:49:40 NOTE: package gcc-cross-3.3.6-r3: task do_build: completed Aug 05 21:49:41 NOTE: package gcc-cross-3.3.6: completed Aug 05 21:49:41 NOTE: build 200608052333: completed Aug 05 21:49:41 Build statistics: Attempted builds: 4 Aug 05 21:49:42 yeah!!! Aug 05 21:50:26 yay! Aug 05 21:50:39 zecke: now move to 4.1.1 :) Aug 05 21:53:05 koen: have we considered using the crosstool script to edit binaries Aug 05 21:53:15 koen: to remove the 'embedded' paths Aug 05 21:53:23 zecke: no idea Aug 05 21:53:37 zecke: I have avoided delving into the toolchain business till EABI Aug 05 21:56:05 hey guys - there are some ipks created by bitbake that have i686 in the name - any idea why? Aug 05 21:56:33 poushag: because they are for the i686 architecture? Aug 05 21:56:36 gcc-cross stuff to be exact - i thought we were building for embedded devices Aug 05 21:57:29 i dont expect to see i686 stuff among all the arm and machine stuff - why is it there? Aug 05 21:58:00 gcc-cross runs on your host (presumably i686), so it seems to be correct Aug 05 21:58:59 well its confusing - since my host doesnt use ipks - and these are in with ipks for my Z feed Aug 05 21:59:19 is there a way i can lask these out when building the index? or some other workaround? Aug 05 21:59:32 s/lask/mask Aug 05 22:01:56 you can't Aug 05 22:02:14 and it is not something to wast braincycles about Aug 05 22:03:15 well i dont like all that garbage in my feed is all - i plan to share it on a limited basis for collie people Aug 05 22:04:34 how does bitbake determine which packages have already been built and dont need to be rebuilt? does it scan a log or look in the deploy/ipk dir? Aug 05 22:04:54 poushag: bitbake looks at stamps/* Aug 05 22:04:57 to see what to do Aug 05 22:05:42 ok so i can delete the ipks i dont like and it wont rebuild them (until another revision comes out) right? Aug 05 22:08:11 right, you can delete them Aug 05 22:08:21 poushag: yes, but don't dare to complain about errors after that , since you try to be smarter as OE Aug 05 22:09:16 deploy/ipk is not really meant to serve as a feed for you device Aug 05 22:09:34 (as one might guess from having to run package-index manually) Aug 05 22:15:27 hmm I'm doing something that could break OE fundamentally Aug 05 22:15:51 koen, i will take your warning and try something else that doesnt involve deleting them i guess - maybe a shell script that will symlink the ones i like into a different feed dir Aug 05 22:17:15 what are you doing zecke thats so bad? Aug 05 22:17:27 poushag: just touching autoconf Aug 05 22:20:12 that sounds to be way above my current level of expertise Aug 05 22:20:36 it is nothing one does for fun... Aug 05 22:21:30 * XOrA|gone shudders at the thought of post midnight autoconf messing Aug 05 22:21:46 ok thanks for the input - time to eat dinner - bbl Aug 05 22:27:21 can shell do nested IFs? Aug 05 22:27:28 rwhitby: I have broken quilt :( Aug 05 22:28:13 one last question - if i move my feed out of deploy/ipk, to say /stuff/build/feed what command do i run to make the package index? will 'bitbake package-index' scan all subfolders and make indexes for those with ipks? Aug 05 22:28:40 zecke: condolences. What can I do to help? Aug 05 22:29:05 rwhitby: fix it, or just be patience :) Aug 05 22:34:31 poushag: ipkg-make-index -p Package -m . Aug 05 22:34:32 patient even Aug 05 22:34:34 kergoth: ping Aug 05 22:34:37 in every dir Aug 05 22:35:19 zecke: I'll be patient, as I don't have time right now to fix anything :-) Aug 05 22:39:17 kergoth: http://rafb.net/paste/results/lsIQzz57.html Aug 05 22:39:24 kergoth: could you take a look? Aug 05 22:41:32 zecke: nice Aug 05 22:41:41 (if that does what I think it does) Aug 05 22:42:00 well, it is kergoth's magic reedited Aug 05 22:43:02 * kergoth looks Aug 05 22:43:07 hey :) Aug 05 22:43:20 ~noob Aug 05 22:43:21 i heard noob is just what someone is before they're a pro Aug 05 22:43:29 hmm Aug 05 22:43:37 kergoth: I'm a noob Aug 05 22:44:22 looks okay at first glance, but i'm rusty :) Aug 05 22:44:33 kergoth: hehe, I need to try it myself as well Aug 05 22:44:45 kergoth: the cnages are AC_CHECK_TARGET_TOOL Aug 05 22:44:54 where I wonder what ':' will be doing Aug 05 22:45:09 and then the nested if/else/fi might behave differently Aug 05 22:47:39 koen: a minimal OE to create a toolchain will be below 1MB Aug 05 22:48:36 * zecke wants a a quadcore g5 system so badly... Aug 05 22:52:04 zecke: steal it from laf0rge Aug 05 22:52:09 hehe Aug 05 22:52:42 the alternative would be an apple pizza box Aug 05 22:52:49 and I would have a reasoning for that Aug 05 22:52:55 hmmm Aug 05 22:53:04 OE Xgrid support :) Aug 05 22:53:05 if you use Darwin Streaming Server, run it on an apple pizza box Aug 05 22:54:20 I wonder what apple will introduce tomorrow Aug 05 22:54:29 yes? Aug 05 22:54:40 probably 2x core 2 dua mac pros Aug 05 22:54:43 w*dc isn't tomorrow? Aug 05 22:54:57 isn't it the 7th? Aug 05 22:54:58 new PowerMac Aug 05 22:55:11 MacPro Aug 05 22:55:18 yeah :/ Aug 05 22:55:27 okay it is tomorrow Aug 05 22:56:31 koen: http://max.macbay.de/galerie/?from=75&context=detail&file=6010cf345e9319f75043383ea9a6f7d5.JPG&gallery=0517555228c8b988c2665ae432ef37ce Aug 05 22:56:35 you are right Aug 05 22:56:56 koen: apple will show a smartphone! Aug 05 22:57:27 apple's always sending me mails like "come to the wwdc, it's only $2500 with your dev discount" Aug 05 22:57:45 (I registered to download Xcode) Aug 05 22:57:47 koen it looks like Max managed to get full sponsorhip Aug 05 23:00:39 *hungry* Aug 05 23:00:44 koen: and Java5 Aug 05 23:01:03 let us order pizza Aug 05 23:01:32 nah Aug 05 23:01:37 * koen gets some candy Aug 05 23:01:48 we will get fat before we are flithy rich Aug 05 23:04:11 yeah Aug 05 23:04:24 I need to stay awake to watch the movie Aug 05 23:05:39 kergoth: it misses a ] somewhere Aug 05 23:08:09 drat Aug 05 23:08:16 sugar rush Aug 05 23:08:24 ~lart candy Aug 05 23:08:24 * ibot frags candy with his BFG9000 Aug 05 23:10:44 oh, there was another issue Aug 05 23:12:26 damn Aug 05 23:12:33 e17 uses binary config files Aug 05 23:12:56 | configure:6357: error: possibly undefined macro: AC_CHECK_TARGET_TOOL Aug 05 23:12:56 | If this token and others are legitimate, please use m4_pattern_allow. Aug 05 23:12:57 | See the Autoconf documentation. Aug 05 23:33:42 I'm a moron Aug 05 23:33:57 AC_CHECK_TARGET_TOOL didn't exist in autoconf 2.59 Aug 05 23:34:04 join the club Aug 05 23:34:25 moron or mormon Aug 05 23:35:06 moron Aug 05 23:39:27 kergoth: should we consider upgraing autoconf? Aug 05 23:41:26 * koen is getting into the OEDEM spirit Aug 05 23:41:34 lol Aug 05 23:41:44 I asked my mother to visit her mother Aug 05 23:41:56 and she might do it, so you can crash at this place... Aug 05 23:46:27 * poushag also drools at the thought of having the latest and greatest powermac Aug 06 00:09:32 03koen 07org.oe.dev * r915e3d55... 10/ (1 packages/efl/esmart.inc): esmart.inc: set appropriate PACKAGES_DYNAMIC Aug 06 00:12:05 jesus you guys are up way too late Aug 06 00:12:48 and here I am hanging out with a bunch of GPL scofflaws Aug 06 00:13:15 lol Aug 06 00:23:10 zecke: http://linuxdevices.com/articles/AT5550934609.html Aug 06 00:25:37 man waiting for southpark is no fun... Aug 06 00:27:30 now we have picture of cbrake :) Aug 06 00:36:08 koen: okay the gcc patch is being tested for gcc4 Aug 06 00:36:49 cool Aug 06 00:40:39 'night all Aug 06 00:44:23 later Aug 06 02:53:32 03justinp 07org.oe.oz354x * r69930bc5... 10/ (5 files in 3 dirs): e-wm: set autoscroll defaults to PDA friendly values Aug 06 02:53:41 03justinp 07org.oe.oz354x * rcfb28aa6... 10/ (1 packages/e17/e-wm_0.16.999.032.bb): e-wm: bump PR **** ENDING LOGGING AT Sun Aug 06 02:59:57 2006