**** BEGIN LOGGING AT Tue Aug 01 02:59:56 2006 Aug 01 03:14:33 anyone tell me why minimo is not pulling from CVS like it should? Aug 01 03:14:40 and failing to build... Aug 01 03:16:06 03mwester 07org.oe.dev * r430dd548... 10/ (1 packages/images/unslung-image.bb): Unslung: unslung_image.bb - .dbg files that need manual cleanup have moved to a .debug directory Aug 01 03:38:31 ld cannot find -lxpcom Aug 01 03:38:33 anyone? Aug 01 03:40:16 bug 692... anyone, resolution? Aug 01 03:40:22 can't build minimo.... Aug 01 04:27:58 I'd be very interested to learn about installing linux onto a HP ipaq rx3115, does anyone have knowledge of that model? Aug 01 04:36:50 yes Aug 01 04:36:57 it needs a lot of love Aug 01 04:37:15 and as far as i know, no one has got a fully working bootloader on it Aug 01 04:37:27 I'll be obtaining one, in a few days, for free. Aug 01 04:37:40 I wanted to make it practical for work Aug 01 04:37:53 I don't handle windows very well. Aug 01 04:38:09 What's the significance with the numbers on the end of the lib .so files? I take it to be version numbers, but programs keep expecting a .so.0 file. What's a good way to go about fixing it? Aug 01 04:38:10 * emte mumbles stuff about useful Aug 01 04:38:32 cinix, link it down Aug 01 04:39:53 emte: you mean like a symlink? easy, but if I have all the files with versions appended why can't it just use them. or just drop the .0 and go for the .so Aug 01 04:40:30 not sure if I should fix the build in the package I'm building, or if it's something in autotools I haven't come across Aug 01 04:40:48 different things expect different stuff Aug 01 04:41:13 libncurses.so.5.5 -> libncurses.so.5 -> libncurses.so Aug 01 04:41:23 emte, is handwriting to text on linux a remote possibility? Aug 01 04:41:34 version 5.5 or greater V5 or greater just a library Aug 01 04:41:36 etc Aug 01 04:42:06 odium, i think you need to actually do some reading about the different distros and thier apps Aug 01 04:50:48 I wonder if the camera can capture a legal sized b&w document and convert it to a legable PDF. Aug 01 05:57:09 hi all! Aug 01 06:01:22 03pH5 07org.oe.dev * rc91d2fd5... 10/ (1 conf/bitbake.conf): bitbake.conf: stop including empty .debug dirs in non-dbg packages Aug 01 06:40:25 plop Aug 01 07:31:47 mornng Aug 01 07:31:53 yo :) Aug 01 07:33:58 2989 root 0 -10 934m 562m 1216 S 21.1 74.3 313:21.19 Xorg Aug 01 07:34:03 morning Aug 01 07:34:11 looks like I need to relogin again Aug 01 07:34:37 Mmm, memory leaks. Aug 01 07:35:44 and this time it is not pixmap mem leak Aug 01 07:39:58 hey XorA & hrw|work Aug 01 07:40:07 koen: fontconfig 2.2.1 should we really be doing that in do_comple_prepend??? Aug 01 07:42:13 XorA: I don't have a fontconfig 2.2.1? Aug 01 07:42:36 Someone understand HAL? I got: A security policy in place prevents this sender from sending this message to this recipient, see message bus configuration file (rejected message had interface "org.freedesktop.Hal.Device.Volume" member "Mount" error name "(unset)" destination "org.freedesktop.Hal") Aug 01 07:42:48 koen: freetype sorry Aug 01 07:43:05 koen: you are corrupting STAGING_LIBDIR in do_compile Aug 01 07:43:32 hrw|work: most likely you are missing a group Aug 01 07:44:05 XorA: maybe Aug 01 07:44:21 XorA: I was in a hurry when I added that Aug 01 07:44:31 XorA: feel free to offer advice :) Aug 01 07:44:36 hrw|work: I got lots of those types when I log in as win2k domain user and I forgot to copy default groups to that user Aug 01 07:44:52 koen: what is wrong, freetype or libz? Aug 01 07:45:09 XorA: I have no idea Aug 01 07:45:18 my bet would be libtool Aug 01 07:45:32 koen: and why do I not see the problem Aug 01 07:45:44 * koen has no idea as well Aug 01 07:45:48 * XorA looks at his fully built freetype 2.2.1 Aug 01 07:46:07 someone reported it and I started seeing it after a build from scratch Aug 01 08:11:45 koen: ah well, I can repeat the problem now, I can investigate Aug 01 08:22:54 morning all Aug 01 08:23:10 hi Liam Aug 01 08:25:26 hi lrg Aug 01 08:26:46 lrg: hi Aug 01 08:28:54 morning all Aug 01 08:29:06 hey lrg Aug 01 08:29:23 lrg: I have linux running on the nbps Aug 01 08:30:02 lrg: I did have some trouble with the board, but after a while I discovered it was resting on the reset button... Aug 01 08:30:07 koen: excellent :) Aug 01 08:48:09 03hs4233 07org.oe.dev * r4d87f3f0... 10/ (5 files in 4 dirs): busybox: defconfig for MNCI Aug 01 08:48:09 koen: ok, I have found out why the error in freetype happens, now to work out the fix Aug 01 08:48:25 schurig: hi Aug 01 08:48:36 schurig: hs4233 is you? back in OE land? Aug 01 08:49:44 hrw|work: yes, I'm trying if I can stand the slowness (now that mtn is just slow, not snake-slow) Aug 01 08:52:02 hi guys. is it possible in a .bb file to apply a patch only for a specific MACHINE? Aug 01 08:52:36 ArjanS: SRC_URI_append_MACHINENAME = "file://mymachine.patch;patch=1" Aug 01 08:52:48 should be enough Aug 01 08:55:37 hrw|work: thank you! I think that's the only way to add my hciattach patch to oe :) Aug 01 09:00:53 good morning Aug 01 09:02:56 hi florian_kc Aug 01 09:03:57 http://www.flickr.com/photos/hrwandil/203712306/ Aug 01 09:14:31 koen: ping Aug 01 09:17:13 mallum: pong Aug 01 09:19:05 koen: do you have the URL for the xcalibrate re-add into xorg patch ? I can commit it back into git Aug 01 09:22:46 koen: NOTE: package freetype-2.2.1: completed Aug 01 09:24:37 hey guys Aug 01 09:24:47 I don't seem to have an /etc/passwd in my staging, any ideas? Aug 01 09:24:50 hey Mike Aug 01 09:24:58 koen: the problem is caused by zlib being built earlier in the build process than it used to. Since freetype doesnt DEPEND on zlib we should add EXTRA_OECONF="--without-zlib" Aug 01 09:26:15 lrg: hey Liam Aug 01 09:28:18 mallum: http://www.openembedded.org/repo/org.openembedded.dev/packages/xorg-xserver/xserver-kdrive/xcalibrate.patch Aug 01 09:28:23 XorA: ah, that would explain it Aug 01 09:29:24 koen: I shall checkin that change then and undo your hack Aug 01 09:29:54 koen: that '/repo/' is nice thing Aug 01 09:31:07 XorA: thanks Aug 01 09:31:14 hrw|work: indeed Aug 01 09:31:45 hrw|work: and http://www.openembedded.org/filebrowser for people scared of raw dirlistings :) Aug 01 09:37:35 koen: thanks Aug 01 09:38:09 03arjan 07org.oe.dev * r4113b1ae... 10/ (1 packages/fontconfig/fontconfig_2.3.91.bb): fontconfig: Add one-j-too-many.patch from fontconfig 2.3.95 also to 2.3.91 Aug 01 09:42:11 03xora 07org.oe.dev * rf38165a0... 10/ (1 packages/freetype/freetype_2.2.1.bb): Aug 01 09:42:11 freetype_2.2.1.bb : freetype doesnt have a DEPEND on zlib and build breaks Aug 01 09:42:11 if external zlib is used, so add a --without-zlib and remove the horrible Aug 01 09:42:11 STAGING_LIBDIR hack. Aug 01 09:49:38 03xora 07org.oe.dev * r329d8a82... 10/ (1 packages/libetpan/libetpan_0.46.bb): Aug 01 09:49:38 libetpan_0.46.bb : add inherit binconfig so binconfig using stuff can Aug 01 09:49:38 find etpan libs. Aug 01 10:07:42 koen: do I need to get key re-activated or not after migration? Aug 01 10:08:00 hrw|work: not Aug 01 10:08:04 ok Aug 01 10:08:09 otherwise we would've sent a mail to oe@ Aug 01 10:38:53 how do I debug problems of the sort NOTE: preferred version 2.10.0 of gtk+ not available Aug 01 10:39:21 XorA: bitbake -s? Aug 01 10:39:38 hrw|work: says 2.8.9 is available, 2.10.0 is selected Aug 01 10:42:06 03koen 07org.oe.dev * r5b4ab918... 10/ (7 files in 7 dirs): busybox 1.2.0: remove unupdated defconfigs, add a large default one and remove DEFAULT_PREFERENCE = -1 Aug 01 10:44:58 does glibc-2.4 work for non-eabi environments? I keep getting 'illegal instruction' despite fp support and other curious things... Aug 01 10:45:28 koobla: afaik people have been using it on x86 Aug 01 10:45:45 hmm Aug 01 10:46:19 koobla: thumb capable cpu? Aug 01 10:46:24 my xscale does not like me it seems Aug 01 10:46:46 nchip, i disabled thumb support - this is a prb? Aug 01 10:47:14 check what the illegal instruction is Aug 01 10:47:49 koen, how? Aug 01 10:48:07 gdb? Aug 01 10:49:02 hm, have to tweak things a little, letme check... Aug 01 10:49:31 "mtn log" is sloooow Aug 01 10:50:06 hrw|work: 'mtn annotate' even slower Aug 01 10:50:13 13:10 hrw@work:conf$ time mtn log bitbake.conf --no-merges --diffs --last 20 >log Aug 01 10:50:45 hrw|work: try http://viewmtn.angrygoats.net/revision.psp?id=5a09fcc91172ba97ca7b92a4f6a54a44da4ddf06 Aug 01 10:51:23 hrw|work: should be ~20 times faster Aug 01 10:52:09 or build from the XP branch: http://viewmtn.angrygoats.net/branch.psp?branch=net.venge.monotone.experiment.performance Aug 01 10:54:13 ~oe.mtn Aug 01 10:54:16 hmm... oe.mtn is http://www.openembedded.org/snapshots/OE.mtn.bz2 Aug 01 11:02:57 anyone know why freetype_2.2.1.bb has DEFAULT_PREFERENCE = "-1" Aug 01 11:03:31 03hrw 07org.oe.oz354x * r2166916c... 10/ (22 files in 7 dirs): wpa-supplicant: dropped all versions older then 0.4.8 Aug 01 11:03:34 I notice that --without-zlib has just been added to 2.2.1, which fixes my problem... Aug 01 11:03:47 03hrw 07org.oe.oz354x * r2de34607... 10/ (1 packages/wpa-supplicant/wpa-supplicant_0.4.8.bb): wpa-supplicant: make it run - close #1205 Aug 01 11:04:06 well, it would, except freetype_2.1.10.bb is the default selection... Aug 01 11:04:10 which sis still broken Aug 01 11:05:49 tkp: test that change on 2.1.10 and Ill checkin Aug 01 11:06:27 damnaton, one little missing line causes so many problems :-) Aug 01 11:11:06 hmm, gdb does not work either... chroot `pwd` /usr/bin/gdb /sbin/init Aug 01 11:11:07 Illegal instruction Aug 01 11:11:38 hrw|work: http://dominion.kabel.utwente.nl/koen/cms/trisoft-rocks Aug 01 11:12:40 koen: have same ;) Aug 01 11:12:53 :) Aug 01 11:15:32 Heh, their partners (xtops.de) gave me 512M cf card instead of the one I ordered (much less, don't remember). Nice folks they are. Aug 01 11:16:14 koen: on commenting I got: "Validation error, please try again. If this error persists, please contact the site administrator." Aug 01 11:20:17 XorA: Hi, well, is adding another copy of zlin is really right solution to freetype problem? Maybe one small link (vs mkdir/cp as it was before) is still better? Aug 01 11:20:46 freetype has/use own zlib now? Aug 01 11:22:37 hrw|work: it seems like that Aug 01 11:23:44 http://mikearthur.co.uk/gpeimagefails Aug 01 11:23:53 could someone take a look at that and hazard a guess why? Aug 01 11:26:23 koen, hrw|work: Glad to see you talk about mtn log performance. But for me, it not only takes eternity, it also outputs complete crap: Aug 01 11:26:36 koen, hrw|work: $ time mtn log --last=3 --brief conf/machine/h4000.conf Aug 01 11:26:37 c1912994f23e5cefd962ee72e5492bcc9c5d1da5 pH5@openembedded.org 2006-07-29T23:28:26 org.openembedded.dev Aug 01 11:26:37 6dfd6d56b3f85931c398c1bb7ef8b38a6451a24c koen@openembedded.org 2006-07-29T16:11:31 org.openembedded.dev Aug 01 11:26:37 5488dcb57b02d59ec27d65e068bc0069a4e4c86e mickeyl@openembedded.org 2006-07-29T16:04:46 org.openembedded.dev Aug 01 11:26:37 a3d194bab430424d2cb654ed59d6b0f9f8952f83 koen@openembedded.org 2006-07-29T16:10:46 org.openembedded.dev Aug 01 11:26:37 real 0m33.051s Aug 01 11:26:40 user 0m21.350s Aug 01 11:26:41 sys 0m11.690s Aug 01 11:26:52 psokolovsky: mtn log is probably broken Aug 01 11:27:08 obviously, that file wasn't modified in those revisions Aug 01 11:27:34 hrw|work: oh my, and how do I track who did what to a file? ;'-( Aug 01 11:33:48 psokolovsky: 'mtn diff' and 'mtn annotate' Aug 01 11:34:05 and try 'mtn log --nomerges' Aug 01 11:34:17 or whatever that param is called nowadays Aug 01 11:34:18 koen: can you fix comments on your blog? Aug 01 11:34:35 hrw|work: I was able to leave a comment just now Aug 01 11:39:21 psokolovsky: if you notice it linked the static anyway, so its no more broken than before Aug 01 11:43:48 hrw|work: could you try again? Aug 01 11:44:37 XorA: hm, ok then. Somehow I thought it uses .so. Well, anyway, my coment is mostly because there are machine-dependent packages which just ship specific config file, instead of common package with 5 configs of 200 bytes each, and now - boom - another copy of lib shipped ;-) Aug 01 11:44:54 koen: same ;( Aug 01 11:45:21 psokolovsky: feel free to fix the libtool ness to use .so and work properly :-) Aug 01 11:45:22 koen: tried "mtn log --no-merges --last=3 --brief conf/machine/h4000.conf". result is still bogus. Could you try that on your install? Aug 01 11:45:38 psokolovsky: I was more concerned with the corruption of staging which is BAD (tm) Aug 01 11:45:50 XorA: ok, when I grow to it ;-) Aug 01 11:46:40 XorA: isn't your staging package-managed? Aug 01 11:47:27 koen: :-D well that hack would have broken that :-) Aug 01 11:48:24 psokolovsky: http://venge.net/monotone/monotone.html#Commands Aug 01 11:48:41 'mtn log' take revision, not a file Aug 01 11:51:17 anyone have any ideas why I have no /etc/passwd? Aug 01 11:51:26 in my gpe-image, and then it conks out Aug 01 11:51:38 mikearthur: coz we cursed you Aug 01 11:51:45 XorA: noooooooooooo Aug 01 11:51:48 http://mikearthur.co.uk/gpeimagefails Aug 01 11:51:52 mikearthur: going to edlug? Aug 01 11:51:52 thats the log Aug 01 11:51:59 this Thursday? Aug 01 11:52:09 mikearthur: I beleive so Aug 01 11:52:15 can't this week, sadly :( Aug 01 11:52:40 mikearthur: it doesn't install base-passwd Aug 01 11:52:46 try rebuilding task-bootstrap Aug 01 11:52:51 koen, cheers Aug 01 11:53:01 build clean, then rebuild? Aug 01 11:55:00 yep Aug 01 11:55:11 hrw|work: could you try leaving a comment (again)? Aug 01 11:56:13 koen: again same Aug 01 11:56:20 grrrr Aug 01 11:58:28 koen: why? "If one or more files are given, the command will only log the revisions where those files are changed." Aug 01 12:02:11 koen: love the 2nd last item on wishlist Aug 01 12:02:34 XorA: :) Aug 01 12:03:53 03arjan 07org.oe.dev * r75900715... 10/ (1 packages/linux/linux-h1940_2.6.17-h1940.bb): linux-h1940: add two more patches for the usb gadget Aug 01 12:29:40 koen: still fails :( Aug 01 12:30:02 could you do a complete rebuild? Aug 01 12:30:08 I guess :s Aug 01 12:30:17 I'll just wait till I get my new laptop Aug 01 12:30:19 03koen 07org.oe.dev * rf536a2b5... 10/ (1 packages/gomunicator/gomunicator_cvs.bb): gomunicator: fix SRC_URI, HOMEPAGE and PV Aug 01 12:30:21 on a P4 1.8 atm Aug 01 12:30:33 getting a 2.0 Core Duo in the next day or two Aug 01 12:30:50 which will do it a littttle bit faster :D Aug 01 12:31:50 :) Aug 01 12:33:57 hi woglinde Aug 01 12:34:01 hi hrw Aug 01 12:34:39 koen: think I've worked out the problem Aug 01 12:34:43 you'll be mad at me :( Aug 01 12:34:52 would changing distro mid-build tend to break things? Aug 01 12:35:03 depends on your config Aug 01 12:35:24 moved from 3.5.4 to 3.5.4.1 Aug 01 12:35:25 I have this: TMPDIR = "/media/hda4/OE/build/tmp-new/${DISTRO}" Aug 01 12:35:44 I don't :) Aug 01 12:35:45 which helps a lot :) Aug 01 12:35:58 if you didn't, and changed, would it break? Aug 01 12:35:59 koen hm Aug 01 12:36:00 koen: hmm did xcalibrateext make it into git ? Aug 01 12:36:14 mallum: I have no idea Aug 01 12:36:15 morning all Aug 01 12:36:18 hey Dirk Aug 01 12:36:21 key Dirk Aug 01 12:36:28 koen: hmm Aug 01 12:36:30 hi koen, mikearthur Aug 01 12:37:17 mikearthur: sucess with your image Aug 01 12:37:33 do13: nope, broke it again :D Aug 01 12:37:44 don't worry, I'll maybe get it before we all die Aug 01 12:37:51 koen: nope appears not Aug 01 12:37:57 which problem? Aug 01 12:38:07 changed DISTRO mid-build Aug 01 12:38:09 pb_: any plans to move the xcalibrateext etc into git repos ? Aug 01 12:38:32 afternoon Aug 01 12:39:06 hi Rene Aug 01 12:39:11 hi Dirk Aug 01 12:39:35 hey Marcin Aug 01 12:40:16 koen: 'mtn pluck' - how to use it between branches? 'cd branch.oz354x;mtn pluck -f REV-taken-from-.dev' will work? Aug 01 12:40:20 hey Marcin Aug 01 12:42:36 lrg: ping Aug 01 12:44:26 hrw|work: it will take any rev that's in your db and apply it to your workspace Aug 01 12:47:25 koen: great - easier to sync zaurus kernels will be Aug 01 12:52:59 mallum: dunno, I don't really have any idea how to do that Aug 01 12:53:58 koen: tried, failed, decided to skip it now - will learn it later Aug 01 12:59:17 pb_: I can probably figure it out but are you still interested in maintaining it ? Aug 01 13:02:58 hi zecke Aug 01 13:03:55 hey! Aug 01 13:11:09 it is possible to use environment variables in the local.conf? i.e. like $HOME/openembedded/packages/*/*.bb? Aug 01 13:13:01 okay found the solution Aug 01 13:25:06 mallum: yeah, I guess so. It doesn't seem to require a whole lot of maintenance. Aug 01 13:25:29 03schurig 07org.oe.dev * rde17790d... 10/ (29 files in 6 dirs): Aug 01 13:25:29 busybox: new 1.2.1 version Aug 01 13:25:29 * also made sure that busybox's wget understands --passive-ftp Aug 01 13:25:59 woo, schurig is back in the saddle :-) Aug 01 13:26:37 XorA: yeah, and if you guys would have used some sane source-control software, I wouldn't have stopped Aug 01 13:26:53 hehe, now the wget from busybox 1.2.1 works with ipkg Aug 01 13:27:46 schurig: it could be worse, VSS is cheap :-) Aug 01 13:28:00 XorA: VSS? Aug 01 13:28:10 schurig: Visual Source (Un)Safe Aug 01 13:28:35 XorA: I doubt that anything is slower than Monotone :-) Aug 01 13:28:45 XorA: and MTN was also sometimes unsafe Aug 01 13:29:22 pb_: maintains means moving into git ;) Aug 01 13:29:57 Cobelius: yes it is possible Aug 01 13:30:09 Cobelius: all env variables are inherited into your bitbake.conf Aug 01 13:30:22 Cobelius: and this is inherited by each bb file Aug 01 13:30:41 mallum: oh, right, heh. yeah, I guess so Aug 01 13:30:43 hail zecke Aug 01 13:30:52 bbiab, gotta go get some new tyres for my car Aug 01 13:31:12 schurig: if you would have participated in the decision making/trial process... Aug 01 13:31:33 zecke: I have. I know that for *ME* svn is all I need Aug 01 13:31:39 schurig: You can't even use the "I told you so" Anti-Pattern :) Aug 01 13:32:02 schurig: Yesterday I moved a openembedded dir which is stored in svn Aug 01 13:32:18 schurig: moving took two hours :) Aug 01 13:32:26 pb_: http://wiki.x.org/wiki/GitMigration Aug 01 13:32:38 schurig: anyway, are you pleased with the progress mtn makes? Aug 01 13:34:21 pb_: mtn list keys | grep "pb" -> still no key :-( Aug 01 13:40:53 03koen 07org.oe.dev * r10a55acf... 10/ (1 conf/distro/angstrom-2006.9.conf): angstrom: prefer busybox 1.2.1 Aug 01 13:44:10 03xora 07org.oe.dev * r71fb6bfd... 10/ (1 packages/gtk+/gtk+_2.10.0.bb): Aug 01 13:44:10 gtk+_2.10.0.bb : Added RRECOMENDS for Angstrom and OZ. Aug 01 13:44:10 Added PACKAGES_DYNAMIC Aug 01 13:44:10 Packaged the print backends into seperate packages. Aug 01 13:45:56 03xora 07org.oe.dev * ree7b821b... 10/ (1 packages/sylpheed/sylpheed_2.2.7.bb): sylpheed_2.2.7.bb : new version Aug 01 13:48:44 ~ugt Aug 01 13:48:46 ugt is probably Universial Greeting Time. Created in #mipslinux, it is a rule that states that whenever somebody enters an IRC channel it is always morning, and it is always late when the person leaves. The local time of any other people in the channel, including the greeter, is irrelevant. Aug 01 13:49:19 ~lart ugt Aug 01 13:49:19 * ibot grabs a large, mis-shapened log, with squirrels, and beats ugt until only the nuts remain ... which the squirrels run off with Aug 01 13:49:23 ~emulate stewie Aug 01 13:49:25 No Sprinkles. For every sprinkle I find, I shall kill you. Aug 01 13:50:18 ~emulate stewie Aug 01 13:50:20 Yes, I rather like this God fellow. He's very theatrical, you know, a pestilence here, a plague there. Omnipotence. Gotta get me some of that. Aug 01 13:50:28 * Zero_Chaos loves stewie Aug 01 13:54:01 * mwester__ also notes that mtn log is broken, and regarding performance of mtn, wonders how he might convince the wife that he needs a new system with a dual-core 3.6GHz CPU... Aug 01 13:54:57 quad-core rules Aug 01 13:55:09 hi Aug 01 13:57:03 mwester__: the annotate patch on the mtn mailinglist is faster! Aug 01 13:57:07 mwester__: but still tooo slow Aug 01 14:00:16 Thanks - I first tried annotate, but I jumped to the conclusion that it hung. :) So instead I manually "annotated" the file in question using a large sheet of paper and the mtn diff command... Aug 01 14:00:44 mwester__: hehe Aug 01 14:00:52 mwester__: I gave up Aug 01 14:03:07 gaaah Aug 01 14:03:22 another package that lost important patches on upgrade: libxfont Aug 01 14:09:13 koen: whats missing? Aug 01 14:09:27 03koen 07org.oe.dev * rc7383d3d... 10/ (4 files in 3 dirs): libxfont: add (rediffed) version of scalable patch that seems to have gotten lost with the update to xorg 7.x Aug 01 14:10:05 XorA: that one Aug 01 14:10:49 hmm Aug 01 14:10:56 I'm out of M&Ms Aug 01 14:11:04 Oh no Aug 01 14:11:16 look under your couch, you might have dropped some? Aug 01 14:11:59 koen: did you ever use the apple migration tool? Aug 01 14:12:41 nothing under the couch Aug 01 14:12:45 zecke: not yet Aug 01 14:12:58 I'm hoping to try it in a few months Aug 01 14:20:10 mwester__, zecke: Anyway, any ideas what's wrong with mtn. I wouldn't think it was possible to make nice cute sqlite *that* slow. Do they miss an index, or misoptimized query? Any ideas? Aug 01 14:20:37 how many of the native packages is it possible to just use your distros version of? Aug 01 14:21:07 psokolovsky: one needs to differentiate. what is slow? Aug 01 14:21:27 psokolovsky: and as good designers they did not start optimizing their code Aug 01 14:21:31 zecke: log and ann is slow. very slow Aug 01 14:21:50 psokolovsky: until lately nobody really profiled what is slow Aug 01 14:21:59 zecke: I guess. They wait till people rewrite their stuff in python ;-) Aug 01 14:22:03 psokolovsky: my guess is, a bad data structure Aug 01 14:22:35 zecke: But is it acknowledged at least? How many projects of OE scale use mtn? Aug 01 14:23:09 psokolovsky: one or two more Aug 01 14:23:21 psokolovsky: and there is no need to acknowledge Aug 01 14:23:30 psokolovsky: e.g. njs is "the" mtn guy Aug 01 14:23:59 psokolovsky: and like any other Free Software project: mtn gets faster when people make it faster Aug 01 14:24:16 zecke: Yes, I heard he keeps an eye. Well, does he the same slowness as we see? ;-) Aug 01 14:24:24 zecke: ok Aug 01 14:25:00 psokolovsky: Use Case: Show me which lines where done in which revisions Aug 01 14:25:33 psokolovsky: personally I do not understand if a file was only changed in two revs why we need to scan 8000 revs Aug 01 14:25:50 psokolovsky: but I know mtn too little to really comment Aug 01 14:26:19 zecke: well, annotate is hard for any SCM, that's understood. But I get ~3revs/sec scan. that sucks. Aug 01 14:26:20 psokolovsky: ideally you would have 'pointers' in your manifest to find the rev with the previous change Aug 01 14:26:42 psokolovsky: that is fast already, and they do stuff nobody else does Aug 01 14:26:55 psokolovsky: and they have started optimising that part already Aug 01 14:27:10 psokolovsky: one sees a huge speed difference btw 0.25 to 0.26 and from 0.26 to 0.27 Aug 01 14:27:31 koen: the ibook now is a bulky firewire disk Aug 01 14:27:36 zecke: is second that, pull speed is much better Aug 01 14:28:11 psokolovsky: I'm confident that mtn will get their Aug 01 14:28:26 psokolovsky: as there are not too many alternatives that really scale Aug 01 14:29:15 i still think mtn is great :) Aug 01 14:31:12 zecke: Ok, thanks for heads-up, I'm glad to know that at least mtn's issues are known and acknowledged. So, well, we (people) survived CVS, will survive mtn too ;-). Going to try that python log tool koen gave link to in the meantime. Aug 01 14:31:32 psokolovsky: the only real options are hg and git Aug 01 14:31:56 zecke: Count my vote for git just in case ;-))) Aug 01 14:31:57 psokolovsky: were hg has had issues in the past and git lacks stuff like properties on files... Aug 01 14:32:29 zecke: You mean something more than perm bits? Aug 01 14:32:39 psokolovsky: right Aug 01 14:32:44 I see Aug 01 14:33:12 psokolovsky: e.g. on my unfinished QA journey Aug 01 14:33:22 psokolovsky: one day I want to store test results in the files Aug 01 14:33:47 psokolovsky: and do checkouts of files only if they are known working Aug 01 14:34:11 psokolovsky: e.g. checkout the OE bits for SH7-B needed to create a gcc4 toolchain :} Aug 01 14:34:46 psokolovsky: I'm sure once my QA effort gets there, mtn is fast enough and mature enough to use these features Aug 01 14:35:50 zecke: And so, mtn can do that - checkout files by such, umm, attributes (would call that tags, as they seem to persist thru time and revisions)? Aug 01 14:36:15 zecke: since monotone 0.13 (or so) you can't attach certs to files anymore, but njs had a way to work around that Aug 01 14:36:17 psokolovsky: well mtn can do properties like svn Aug 01 14:36:30 psokolovsky: and you can store test results Aug 01 14:36:44 zecke: ok, got idea Aug 01 14:38:12 psokolovsky: the master plan is. The tinderbox receives the test results, flags the files in the mtn db Aug 01 14:38:22 psokolovsky: we pull the db from the tinderbox Aug 01 14:38:42 psokolovsky: people doing checkouts will be sure that the stuff in their tree will at least compile (as it compiled for others) Aug 01 14:39:22 nice Aug 01 14:40:01 psokolovsky: that is my plan, it will take a lot of months until this is done Aug 01 14:41:24 zecke: I understand. Good plans though. Glad to know OE will continue to evolve (well, it does, and pretty fast, have to say ;-) Aug 01 14:43:52 * zecke has a new shark :) Aug 01 14:44:05 now only one shark is missing and I have four Aug 01 14:46:56 lol Aug 01 14:47:00 shark cluster Aug 01 14:48:35 time for 'Jaws (zecke edition)'? Aug 01 14:48:36 :D Aug 01 14:50:21 my god Aug 01 14:50:28 svn breaks after 3 minutes Aug 01 14:50:37 * koen svn cleanup Aug 01 14:52:57 yay! corrupted bdb Aug 01 14:53:45 koen: don't use bdb, use fsfs ;-) Aug 01 14:54:19 koen: sounds painful Aug 01 14:54:39 psokolovsky: I use whatever svn uses by default Aug 01 14:54:42 which is bdb Aug 01 14:55:05 koen: fsfs is default since ~1.2. 1.3 is current version. Aug 01 14:55:44 oe@serenity:~/svn/openembedded$ svn --version Aug 01 14:55:45 svn, version 1.1.4 (r13838) Aug 01 14:55:49 debian stable :) Aug 01 14:56:14 koen: :) Aug 01 14:56:21 koen: oh well... ;-) Aug 01 15:00:38 hey pH5 Aug 01 15:01:02 hey koen Aug 01 15:01:30 I built an iwmmxt gpe image from scratch today. except for dbus-glib 0.71, everything compiled without glitches :) Aug 01 15:07:54 pH5: and does it work? Aug 01 15:09:00 I knew you would ask that.. Aug 01 15:09:57 :) Aug 01 15:21:54 03xora 07org.oe.dev * r58ff5550... 10/ (23 files in 5 dirs): Aug 01 15:21:54 glib-2.0_2.12.1.bb : version upgrade Aug 01 15:21:54 gtk+_2.10.0.bb : fix jpg/jpeg in RRECOMENDS Aug 01 15:21:54 gtk+_2.10.1.bb : version upgrade Aug 01 15:23:36 XorA: pls split commits... Aug 01 15:24:41 argh.. Aug 01 15:24:44 ~curse kdepim Aug 01 15:24:46 May the fleas of a thousand camels infest your most sensitive regions, kdepim ! Aug 01 15:26:28 hrw|work: add a checkin policy Aug 01 15:28:31 XorA: into wiki? Aug 01 15:28:38 XorA: can do it later Aug 01 15:29:09 hrw|work: cool :-) not that I am disagreeng wth you, just I realised there isnt one, and there should be if that is actually the policy Aug 01 15:29:31 you can shoplift bits of monotonephrasebook for that Aug 01 15:30:06 anyway with that gtk+ nastiness committed I am off to play ADND Aug 01 15:30:55 advanced drag and drop? Aug 01 15:31:26 * XorA hurls a giant d20 at koen Aug 01 15:31:48 * koen casts +3 delecious on his supper Aug 01 15:32:01 XorA: I liked d3 Aug 01 15:37:18 koen: frontrow is awesome Aug 01 15:42:23 koen: seems to work so far Aug 01 15:49:31 pH5: awesome Aug 01 15:49:39 zecke: *mumble* Aug 01 15:51:06 lol Aug 01 15:51:08 http://thedailywtf.com/forums/thread/83942.aspx Aug 01 15:53:45 pH5: https://bugs.freedesktop.org/show_bug.cgi?id=7675 Aug 01 15:55:29 koen: very good... now just solve it ;) Aug 01 16:01:41 pH5: mplayer has some iwmmxt optimizations, although you need to hack the makefile for that Aug 01 16:07:14 cu Aug 01 16:14:17 03florian 07org.oe.dev * r2e4e1748... 10/ (1 packages/tasks/task-sdk.bb): task-sdk: Add alsa libraries and missing libgcc. Aug 01 16:25:23 Does OZ ship a SDK? Aug 01 16:26:25 florian_kc: http://ewi546.ewi.utwente.nl/mirror/www.openzaurus.org/official/unstable/3.5.4/sdk/ Aug 01 16:27:21 koen: its unlikely that this reflects the current stuff in the branch... Aug 01 16:27:43 i had some fixes but i'm not sure if i can/should check them in. Aug 01 16:29:30 ask hrw|gone Aug 01 16:34:02 later :-)+ Aug 01 16:42:37 Hello Aug 01 16:49:58 a++ Aug 01 17:44:56 'evening ! Aug 01 20:03:57 hi all Aug 01 20:05:02 hey mr_nice_slacker Aug 01 20:05:03 http://www.hackaday.com/entry/1234000337073804/ Aug 01 20:06:25 hi koen Aug 01 20:18:30 Du bist noch wach? Aug 01 20:18:38 Sorry, wrong channel. Aug 01 20:23:05 ~ugt Aug 01 20:23:07 hmm... ugt is Universial Greeting Time. Created in #mipslinux, it is a rule that states that whenever somebody enters an IRC channel it is always morning, and it is always late when the person leaves. The local time of any other people in the channel, including the greeter, is irrelevant. Aug 01 20:24:22 lol Aug 01 21:00:30 can anyone tell me where lsusb comes from? Aug 01 21:00:46 (i dont have it on my system but need it for what we're doing) Aug 01 21:01:36 usbutils Aug 01 21:01:52 you can check it with http://packages.ubuntu.com/cgi-bin/search_contents.pl?word=lsusb&searchmode=searchfiles&case=insensitive&version=dapper&arch=i386 Aug 01 21:02:16 it works for 90% of the stuff I'm searching for Aug 01 21:02:55 re Aug 01 22:47:58 * v8jlene is back **** ENDING LOGGING AT Wed Aug 02 02:59:57 2006