**** BEGIN LOGGING AT Tue Apr 11 09:59:57 2006 Apr 11 10:02:41 good morning Apr 11 10:02:45 hi Bernardo Apr 11 10:02:59 RP: let sharp release some nice pda in this year.... Apr 11 10:03:55 hrw|work: That would mean kernel work but it would be nice if they could add wifi etc :) Apr 11 10:04:38 hrw|work: I checked the alpha2 images this morning, but you only have the feed, right? Apr 11 10:04:58 RP: I bet you breathed a sigh of relief when they released the 3200 Apr 11 10:05:13 NOTE: package enca-1.9-r0: task do_build: completed Apr 11 10:05:15 johnX: Mixed feelings... Apr 11 10:05:51 Bernardo: 3.5.4.1-alpha2 use 3.5.4 feeds + 3.5.4.1-alpha2/machine one Apr 11 10:05:56 well, if they weren't going to include wifi/bt it was at least nice of them to not change the hardware for no reason Apr 11 10:07:01 hrw|work: let me see if I understand, what you wanted me to test - zimage and initrd from 3.5.4.1 alpha1, updated with the feed from 3.5.4.1 alpha2? Apr 11 10:07:14 hmm Apr 11 10:07:38 Bernardo: You want an alpha2 kernel Apr 11 10:07:52 Bernardo: no - alpha2 image + alpha2 kernel from my dir on ewi Apr 11 10:08:48 If anyone can lend a brain cell or two, I'd appreciate it: http://tinyurl.com/gvufx :) Sorry about the grammar and poor presentation in general Apr 11 10:09:14 that is what I'm saying, this morning there was no kernel or image in ewi, only the feed Apr 11 10:09:46 Bernardo: look in .images/ dir Apr 11 10:10:06 .images? ah, that might have confused firefox Apr 11 10:10:27 Bernardo: apache does not list hidden dirs ;) Apr 11 10:11:23 :) Apr 11 10:11:28 I'll try again tonight Apr 11 10:11:57 anyway, I have here the image + feed I generated this morning, so I'll keep playing with that Apr 11 10:12:50 Bernardo: I suspect that .oz354x direct build can work and my not Apr 11 10:13:09 NOTE: package fbreader-0.7.3b-r0: task do_build: completed Apr 11 10:13:28 hrw|work: why, because you're generating the image from the feed? Apr 11 10:14:05 yes Apr 11 10:14:19 Bernardo: maybe something miss in feed Apr 11 10:14:29 Bernardo: can you send me /usr/lib/ipkg/status from image? Apr 11 10:14:38 Bernardo: I will check then does it differ Apr 11 10:14:41 hrw|work: you rock! Apr 11 10:15:16 hrw|work: ok, let me just flash it and I'll send it to you Apr 11 10:15:24 hrw|work: but I'm using opie-kdepim-image Apr 11 10:15:32 Bernardo: ok Apr 11 10:15:34 hrw|work: so I have a load of extra stuff Apr 11 10:15:54 Bernardo: no problem Apr 11 10:16:02 just ignore all themes, styles, libxine-plugin-*, kphonepi, etc... :) Apr 11 10:18:18 johnX: will you maintain enca/fbreader in OE? I want to avoid adding stuff lacking MAINTAINER field Apr 11 10:18:34 Who's going to try prelink then? :) Apr 11 10:19:28 hey RP, hows life? Apr 11 10:19:35 hrw|work: I'll do what I can Apr 11 10:19:38 hi mithro Apr 11 10:19:42 mithro: busy :) Apr 11 10:19:54 johnX: so I can add you to maintainer field? Apr 11 10:19:59 yeah, sure Apr 11 10:20:03 RP: what you been up to? Apr 11 10:20:33 mithro: I like the stylus - getting use to it now :) Apr 11 10:21:32 hrw|work: I can't commit to OE though, so I'll have to pass on changes through another dev I guess? Apr 11 10:21:38 mithro: A reasonable amount of work but also some motorcycle trips and DIY Apr 11 10:21:56 DIY what? gone anywhere intresting? Apr 11 10:22:50 johnX: I can 'sponsor' your uploads Apr 11 10:23:01 mithro: Work on a house that needs a lot of TLC. Had features like wiring from 1910 until not so long ago :) Apr 11 10:23:29 mithro: I went to the lake distict on the bike - I must take a camera next time I do it... Apr 11 10:23:37 hi Apr 11 10:23:48 hi mickeyl Apr 11 10:23:55 hrw|work: did you forget you upload the kernels to alpha2? Apr 11 10:24:13 mickeyl: looks like... Apr 11 10:24:16 will fix it Apr 11 10:24:17 mithro: How's life with you now you're back at home? Apr 11 10:24:54 RP: slow, i converted to the dark side and installed ubuntu Apr 11 10:25:28 Whats the dark side on that?-) Apr 11 10:26:05 Or maybe those two weren't related at all, hmm Apr 11 10:26:43 Cockroach-: everyone should use debian :P Apr 11 10:27:31 Oh I see. Apr 11 10:27:54 Well, It's nice to be a heretic Apr 11 10:28:10 gnome 2.14 is nice Apr 11 10:28:25 i needed to reinstall my computer so I thought i would try ubuntu Apr 11 10:28:35 RP: uni is deathly boring :/ Apr 11 10:29:03 anyone know how to find all X applications on the commandline? Apr 11 10:29:21 mithro: I use ubuntu for my desktop... Apr 11 10:29:30 mithro: Its not fun when its boring :-( Apr 11 10:29:38 RP: nope :/ Apr 11 10:30:23 Darn, need to subscribe to another mailing list Apr 11 10:31:25 Constant crying for help on IRC might irritate people Apr 11 10:31:45 03mickeyl 07org.oe.dev * ra2827042... 10/packages/qmake/qmake2-native_2.00a.bb: qmake2: specify (cross-)install prefix to get library paths pointing to the proper staging area Apr 11 10:31:49 03mickeyl 07org.oe.dev * r830a2ca2... 10/packages/qt/qt4-x11-free_4.1.1.bb: Apr 11 10:31:49 qt4: fix tests and staging Apr 11 10:31:49 - configuration tests are done using the proper compiler now Apr 11 10:31:49 - staging now calls make install to get everything installed to its proper place Apr 11 10:32:32 http://bealach-na-ba.fotopic.net/c67521.html is Hardknott pass. Wrynose, Langdale and Kirkstone were all also on the route :). I can't find any photos which really do it justice :-/ Apr 11 10:35:10 johnX: pushed into .oz354x Apr 11 10:35:20 What was the URL to the irc logs of this channel? Apr 11 10:35:29 can't remember bot commands either Apr 11 10:35:40 ~logs Apr 11 10:35:40 03hrw 07org.oe.oz354x * ra2c7445b... 10/packages/enca/ (3 files in 2 dirs): enca: added version 1.9 from #720 - thx to John 'JohnX' Bloom Apr 11 10:35:42 apt/ibot/jbot/purl all log to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz, or updated "nightly" Apr 11 10:35:44 03hrw 07org.oe.oz354x * r98aa042c... 10/packages/fbreader/ (4 files in 2 dirs): fbreader: added version 0.7.3b from #720 - thx to John 'JohnX' Bloom Apr 11 10:35:55 Thank you Apr 11 10:36:52 hrw|work: very cool, though a lot of the credit goes to you for making the patches sane Apr 11 10:37:12 johnX: need to add site change too Apr 11 10:37:49 http://en.wikipedia.org/wiki/Hardknott_Pass :) Apr 11 10:40:33 RP: thats like being home at my parents Apr 11 10:42:23 XorA: I need to get into Scotland on the bike more Apr 11 10:42:53 RP: Edinburgh is nice :-D Apr 11 10:43:22 XorA: I'd more like to see some of the roads/hills :) Apr 11 10:43:39 but Edinburgh is nice Apr 11 10:43:42 RP: some nice routes into highlands Apr 11 10:44:32 XorA: Its just a shame all the roads from here to Edinburugh are covered in speed cameras and are generally boring :-/ Apr 11 10:46:02 03hrw 07org.oe.oz354x * r27603ca1... 10/site/arm-linux: site/arm-linux: added yeti_cv_func_scanf_modif_size_t needed to get enca 1.9 buildable Apr 11 10:46:06 03hrw 07org.oe.dev * rd1bbc0ff... 10/packages/fbreader/ (4 files in 2 dirs): fbreader: added version 0.7.3b from #720 - thx to John 'JohnX' Bloom Apr 11 10:46:10 03hrw 07org.oe.dev * rcd23171d... 10/packages/enca/ (3 files in 2 dirs): enca: added version 1.9 from #720 - thx to John 'JohnX' Bloom Apr 11 10:46:14 03hrw 07org.oe.dev * rb9308de4... 10/site/arm-linux: site/arm-linux: added yeti_cv_func_scanf_modif_size_t needed to get enca 1.9 buildable Apr 11 10:46:28 johnX: look how I redid your patches Apr 11 10:47:01 RP: yeah, it sucks a bit, you could do a left at gretna, cut through dumfries and up into the Galloway Forest Park, thats a nice area Apr 11 10:47:45 Or go via Kielder I guess... Apr 11 10:47:47 hrw|work: I will, updating now... Apr 11 10:48:54 * XorA really needs to get licensed for some form of motorised transport Apr 11 10:49:45 hrw|work: you should have my /usr/lib/ipkg/status on your mailbox now Apr 11 10:50:08 * hrw|work has driver license category A/B but drived only 10km by motorbike Apr 11 10:50:21 Bernardo: have it thx Apr 11 10:50:54 * RP probably wouldn't recommend motorbikes to people :-/ Apr 11 10:51:23 * Bernardo has seen too many motorbike accidents Apr 11 10:52:01 On the other hand, part of me says every car driver should have tried it, just to see what its like with the protection of a steel cage... Apr 11 10:53:07 * Bernardo just found out one extra problem with the transition to pcmciautils... His bluetooth card needs ejecting and inserting after suspending Apr 11 10:54:10 just found a few "hci0 command tx timeout" in dmesg Apr 11 10:54:46 Bernardo: It seems bluetooth pcmcia suspend/resume is a little broken :-/ Apr 11 10:55:00 Bernardo: now I see some problems Apr 11 10:55:13 RP: for this card it was always broken Apr 11 10:55:41 Bernardo: but it worked with pcmcia-cs? Apr 11 10:55:42 but opie-pcmciaapplet allowed re-enabling it after resumin Apr 11 10:55:45 resuming Apr 11 10:56:23 It's a socket rev E, the one with nokia chipset Apr 11 10:56:46 hrw|work: with pcmciautils, or with ipkg status? Apr 11 10:58:33 status Apr 11 10:59:01 Bernardo: Ah, I see :-/ Apr 11 10:59:57 hrw|work: That looks much cleaner. Apr 11 11:00:39 RP: this driver has had problems with the suspend/resume code for as long as I remember Apr 11 11:00:52 hrw|work: can I help? Apr 11 11:01:45 Bernardo: you helped me already with status file Apr 11 11:02:04 Bernardo: I have to rebuild alsa and add to feed, update initscripts... Apr 11 11:08:00 It's 4AM here, so I have to get some sleep. 'night all Apr 11 11:08:11 n8 johnX Apr 11 11:09:02 hrw|work: do you want my ipks? Apr 11 11:09:26 Bernardo: no - I can rebuild them Apr 11 11:09:35 hrw|work: I have all my feed here (can't send you 80Mb of it, but anything up to 4-5Mb is ok). Apr 11 11:09:59 03hrw 07org.oe.oz354x * r89869338... 10/packages/ (initscripts/initscripts_1.0.bb sysvinit/sysvinit_2.86.bb): initscripts, sysvinit: bump PR to provide fixed packages in feed Apr 11 11:10:04 Bernardo: status file was enough to show me which things can be broken Apr 11 11:18:49 anyone know how to inject key events into X or evdev?? Apr 11 11:22:57 mithro: uinput Apr 11 11:23:19 mithro: See tskeys in zaurusd as a code example :) Apr 11 11:25:01 mithro: XTestFakeKeyEvent() Apr 11 11:25:26 i want to inject my pcc_acpi key events into X Apr 11 11:27:45 english power connectors differ from continental ones? Apr 11 11:27:53 yes Apr 11 11:28:19 http://en.wikipedia.org/wiki/BS1363 Apr 11 11:28:25 thx Apr 11 11:31:53 pb_: when I was in the UK, I always found the huge, clumsy and ugly :-) Apr 11 11:32:07 yeah, they are all three of those things Apr 11 11:32:09 pb_: but I liked the sockets with integrated switches Apr 11 11:33:06 although I think the sockets themselves are less ugly than the Shuko ones Apr 11 11:33:12 schuko, even Apr 11 11:40:16 schuko? Apr 11 11:40:30 i like the tiawan one Apr 11 11:41:25 mithro: Schuko, if you want it right (we germans capitalize nouns). It comes from "Schutzkontaktsteckdose", and when you read that word, you know why everyone just says "Steckdose" and a few "Schuko" Apr 11 11:41:40 ? Apr 11 11:41:46 Schutz kontakt steck dose means Security contact socket Apr 11 11:48:19 ahh acpi_fakekey is what I wanted Apr 11 11:54:30 ahh okay Apr 11 11:55:28 hi Apr 11 11:57:08 Hm, does anyone know of an easy fix for the gettext-native error that complains about java.system.Object not being found? I've got gcj and classpath installed on the host. Apr 11 12:11:02 03hrw 07org.oe.oz354x * r5a5a2348... 10/packages/ipkg-utils/ipkg-link_1.6cvs.bb: ipkg-link: set PACKAGE_ARCH to all - its shell script Apr 11 12:28:59 re mickeyl Apr 11 12:29:26 hi ho Apr 11 12:30:06 morning Apr 11 12:30:09 moin mickeyl Apr 11 12:30:58 hi mickeyl Apr 11 12:31:37 hi reno, CoreDump|home Apr 11 12:31:54 know what? qt4 rocks :D Apr 11 12:32:16 it's a tad bit slow on a Zaurus, but we can perhaps improve that Apr 11 12:32:26 03mickeyl 07org.oe.dev * ra31fed08... 10/packages/qt/qt4-x11-free_4.1.1.bb: qt4: install native tools to where they are expected in qmake generated Makefiles Apr 11 12:32:31 03mickeyl 07org.oe.dev * refd88402... 10/packages/texmaker/texmaker_1.3.bb: Apr 11 12:32:31 add texmaker, a free cross-platform LaTeX editor based on Qt4 Apr 11 12:32:31 TADAAA, this is the first Qt4 application in OE Apr 11 12:34:06 mickeyl: what's so great about it? Nice eyecandy? Apr 11 12:35:11 it looks good, yes. but that's not it. it's the power. it has powerful widgets to write complex applications w/ less code than before. Apr 11 12:35:21 powerful utility classes as well Apr 11 12:35:30 i just can Apr 11 12:35:35 't stand qt2.3 any longer Apr 11 12:35:36 cu Apr 11 12:35:38 cu hrw Apr 11 12:39:31 mickeyl: We should have touchscreen buttons for 3.5.4.1 btw Apr 11 12:39:45 mickeyl: busy with Qt4 :-) Apr 11 12:40:10 hey mickeyl Apr 11 12:40:11 sup? Apr 11 12:40:21 mickeyl: is qt4 slower as qt2 on your zauri? Apr 11 12:40:27 mickeyl: I'm told they work under GPE and it should be possible to make them work under opie easily enough (They've been mapped to F15-F18 I think) Apr 11 12:41:05 RP: they work, I just have nothing attached to them in gpe Apr 11 12:42:03 XorA: I think the default mappings were updated as well? Apr 11 12:42:22 RP: I removed keylaunch as it annoyed me :-) Apr 11 12:42:33 RP: woot! that's pretty cool. zaurusd? Apr 11 12:42:58 mickeyl: Part of it, yes. It actually supports generic buttons anywhere on the touchscreen :) Apr 11 12:43:18 By default we map the marked soft buttons and the spaces between them Apr 11 12:43:28 schurig: It looks like Qt4/X11 is slower than 2.3/Embedded. But it may as well be a factor of memory consumption which is at its limits atm. Apr 11 12:44:12 RP: pretty cool. shouldn't they emit the same keycodes as the hardware pendants though? Apr 11 12:44:13 mickeyl: oh, I haven't used Qt4/X11 on MNCI and RT3000, just Qt4/E. However, the proggies where trivial, so I didn't see any speed difference Apr 11 12:44:40 mickeyl: I had that happening in the original keymap but it wouldn't work with GPE :-( Apr 11 12:44:42 mickeyl: did you stop working with QtXX/Embedded? Apr 11 12:44:54 RP: Re: touchscreen buttons, coolies Apr 11 12:44:56 RP: ah darn. ok Apr 11 12:45:31 schurig: yes. I'm no longer interested in Qt/Embedded. Apr 11 12:46:05 schurig: the only thing that could change that would be a phone with qt/e, but it doesn't look like such an open programmable thing is coming anyway soon Apr 11 12:49:20 i don't know why road-gmbh is taking so long... Apr 11 12:50:10 I guess zecke is sabotaging their efforts in some way. Apr 11 12:50:41 mickeyl: I guess the MNCI wouldn't be good enought :-) Apr 11 12:50:52 hehe. possible. since this year, he is no longer involved though, so it may speed up now :) Apr 11 12:51:00 mickeyl: it's an BS1363 (english plug) version of a cell phone :-) Apr 11 12:51:01 schurig: it's a bit large ;) Apr 11 13:07:28 03rpurdie 07org.oe.dev * rb1b40ed4... 10/packages/gtk+/gtk+_2.8.16.bb: gtk+ 2.8.16: Enable no demos patch Apr 11 13:07:33 03rpurdie 07org.oe.dev * r8f860e5a... 10/packages/gcc/gcc_csl-arm-2005q3.bb: gcc arm-csl-2005q3: Add patch to fix 64 bit comparison problem Apr 11 13:33:10 morning Apr 11 13:45:45 mickeyl|thesis: looks like you forgot fix-mkspecs.patch in qt4 Apr 11 13:46:37 right. fixing taht Apr 11 13:48:15 ooops, didnt mean to disturb you from writing, thought youd pick that message up later :-) Apr 11 13:49:30 what is this message? : NOTE: Handling BitBake files: \ (0363/3242) [11 %]ERROR: while parsing /home/jrs/cvs/oe/org.openembedded.dev/packages/dvbtools/dvbtune_cvs.bb Apr 11 13:49:39 what do i have to do to fix this? Apr 11 13:49:49 goxboxlive: got the latest bitbake? Apr 11 13:50:03 i downloaded it today Apr 11 13:50:24 03mickeyl 07org.oe.dev * r2ff4c701... 10/packages/uicmoc/ (2 files in 2 dirs): uicmoc3: fix a buglet in the 64bit-cleanup patch Apr 11 13:50:29 03mickeyl 07org.oe.dev * rfde5a287... 10/packages/qt/qt4-x11-free-4.1.1/fix-mkspecs.patch: qt4: add forgotten patch, thanks XorA Apr 11 13:52:08 Hm, does anyone know of an easy fix for the gettext-native error that complains about java.system.Object not being found? I've got gcj and classpath installed on the host. Apr 11 14:24:23 re Apr 11 14:25:24 hi hrw|work Apr 11 14:28:21 NAbyss: not really, but if you have gettext installed, you can try ASSUME_PROVIDED += "gettext-native" in your local.conf Apr 11 14:28:34 schurig: Ah, thanks, shall give it a go Apr 11 14:44:21 btw changes in bitbake.... moving bbimage from bitbake/bin to bbclass in OE would be nice Apr 11 14:45:24 hrw|work: agreed Apr 11 14:47:14 I want simple thing: generating md5sum after image... but in image_ipk.bbclass there is no way to get name of image Apr 11 14:47:25 atleast not from my fast looking Apr 11 14:50:05 Moving bbimage would be good... Apr 11 14:54:14 03hrw 07org.oe.oz354x * rfb9b110c... 10/packages/ (4 files in 4 dirs): libusb, usbutils, opie-button-settings, tslib: bump PR to fix packages from feed Apr 11 14:58:32 cleaning.... 3.8G tmp/work is going byebye Apr 11 15:00:07 hrw|work: sure? Apr 11 15:00:12 hrw|work: I have a ROOTFS_POSTPROCESS_COMMAND = "do_fixrootfs" Apr 11 15:00:29 hrw|work: aehem, wrong line Apr 11 15:00:36 hrw|work: I have a IMAGE_POSTPROCESS_COMMAND = "do_mkimage" Apr 11 15:00:41 hrw|work: and there I have Apr 11 15:01:01 cat ${FILESDIR}/grub.img ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.rootfs.ext2 >blabla Apr 11 15:01:38 will look more - as I said - it was fast look Apr 11 15:07:22 RP, mickeyl, CoreDump|home: rebuilded images Apr 11 15:09:44 hrw|work: kernels as well? :) Apr 11 15:09:51 yep Apr 11 15:10:01 100% from feeds Apr 11 15:11:10 now how to resolve irk-belkin/irk-targus situation... they conflict but not replace each other Apr 11 15:12:04 hrw|work: cool. Will test on spitz tonite then Apr 11 15:12:18 great Apr 11 15:19:00 bbl, sports time Apr 11 15:20:13 cu Apr 11 15:25:32 * chouimat is away: house chores :( Apr 11 15:44:28 bye Apr 11 15:52:14 Hi, all!!! How to explain problem of CodeSourcery gcc with eabi and uClibc causing illegal instruction? Apr 11 15:52:36 It aborts after 'svc 0x00000000' Apr 11 15:53:10 (afair svc=swi) Apr 11 15:53:57 I know it should happen, but I'm not quite understand why, since kernel is EABI enabled (with OABI compatibility) Apr 11 15:57:34 sounds like a bug Apr 11 15:57:39 ~lart screen Apr 11 15:57:39 * ibot stabs screen Apr 11 16:14:26 anyone feel like updating the busybox build? Apr 11 16:25:24 wow busybox is a space hog Apr 11 16:30:54 ah, endless loop in configure will do that Apr 11 16:30:59 -rw-r--r-- 1 aman users 4.4G Apr 11 12:38 /oe/arm/build/work/busybox-1.1.2-r9/temp/log.do_configure.12105 Apr 11 16:32:00 03justinp 07org.oe.dev * r455dc2b2... 10/packages/ (4 files in 2 dirs): e, e-wm: LORES_PDA => LOWRES_PDA, fixes bug #820, thanks to Olivier Gournet for spotting Apr 11 16:43:08 is ewi546.ewi.utwente.nl down or something? Apr 11 16:44:41 can't do a 'monotone pull' or ping the host... Apr 11 16:45:29 cedric_: try the main server, monotone.vanille.de Apr 11 16:45:45 always best to try the primary server first Apr 11 16:47:51 JustinP: thx Apr 11 17:18:48 A simple question: gcc-cross and gcc-cross-initial. What function does the latter have? Apr 11 17:27:20 hi Apr 11 17:27:32 Hell.o Apr 11 17:27:42 how can I unpack tar xfz opie-mail_1.2.1-r0_arm.ipk Apr 11 17:27:48 this does not work ;o( Apr 11 17:28:19 gzip: stdin: not in gzip format Apr 11 17:28:19 tar: Child returned status 1 Apr 11 17:28:19 tar: Fehler beim Beenden, verursacht durch vorhergehende Fehler. Apr 11 17:28:26 do I have to rename it first ? Apr 11 17:28:36 New ipkg's are just ar archives Apr 11 17:28:42 not tars anymore, afaik Apr 11 17:28:59 ok : ar ? or rar ? Apr 11 17:29:03 ar Apr 11 17:29:10 unpack unar ? Apr 11 17:29:27 ar x foo.ipk Apr 11 17:30:37 jepp. Apr 11 17:30:39 works Apr 11 17:30:58 thx. I have read about it on maintainers page of ipk but thought it was rar ;o( Apr 11 17:31:01 thx for help Apr 11 17:31:07 I got lucky :) Apr 11 17:31:27 Cockroach-: -initial is for the initial building. There are multiple stages to a gcc build Apr 11 17:32:09 * Philippe is back (gone 19:52:04) Apr 11 17:32:34 JustinP: I suppose it's the same process as described in linuxfromscratch.org Apr 11 17:33:09 however it kind of flushes a certain therory down the drain, concerning a problem I have Apr 11 17:34:36 both are built in a normal build, first initial then non-initial Apr 11 17:39:34 But still, I don't understand the idea behind it. :) Apr 11 17:40:03 read up on it some more then ;-) gcc builds itself with itself Apr 11 17:40:22 so that you get all of the new optimizations and such of the new compiler.....in your compiler Apr 11 17:40:36 it's called bootstrapping Apr 11 17:42:38 has anyone got an idea how opie -> quicklauncher works ? Apr 11 17:43:04 I have a correct link from opiemail to quicklauncher, but no ways - opiemail does not Apr 11 17:43:11 come up - Apr 11 17:43:30 how do fix the hostap-modules in .dev? I've seen a post this afternoon but I can't remember... Apr 11 17:43:49 cedric_: there was a post about it? I've noticed build problems as well Apr 11 17:43:57 jayja1: very badly....:-| Apr 11 17:44:13 why ? Apr 11 17:44:20 I installed it on cf. Apr 11 17:44:30 linking is bad though on fat32 Apr 11 17:44:37 sorry vfat Apr 11 17:44:39 i meant. Apr 11 17:44:47 might be the problem. Apr 11 17:45:15 JustinP: my search for understanding the whole gcc build process is directly related on the problem I've described in http://tinyurl.com/gvufx Apr 11 17:45:52 and if gcc-cross-initial would produce not-crosscompiling gcc, I could make some sense =) Apr 11 17:46:04 of the process Apr 11 17:48:19 JustinP: I think it's the RRECOMMENDS in hostap-utils.inc, going to try... Apr 11 17:49:28 man Apr 11 17:49:33 who woulda guessed pmount would be such a pain Apr 11 17:53:09 JustinP: yep it was that... Apr 11 18:05:18 Have to figure this out some other day Apr 11 18:07:23 Hi all Apr 11 18:12:54 I don't understand this morning I've built librsvg-2.11.1 on my work box (fedora core 2) and tonite on my home box (fedora core 4) it doesn't --> http://pastebin.com/653977 Apr 11 18:13:44 at work with need to use http proxy that's the difference with my home connection Apr 11 18:21:43 * chouimat is back. Apr 11 18:31:52 gah! anyone know about compiling with gettext, po files, intltool and stuff? Apr 11 18:40:13 cedric_: Which bitbake revision are you using? I keep seeing people with problems with hostap-modules yet I don't see it here :-/ Apr 11 18:40:20 hi Dirk Apr 11 18:40:30 Hey RP Apr 11 18:41:32 can i #define something //something in C? Apr 11 18:41:53 RP: BitBake Build Tool Core version 1.3.3.4, bitbake version 1.3.3.2 Apr 11 18:42:19 RP: do u know about that stupid librsvg? Apr 11 18:42:30 cedric_: Is that from svn? Apr 11 18:42:36 yep Apr 11 18:42:45 cedric_: Which svn revision? Apr 11 18:43:21 cedric_: Personally, I'd patch out the document generation Apr 11 18:43:22 RP: 423 Apr 11 18:43:30 cedric_: Try r419 Apr 11 18:43:56 RP: 423 is the head Apr 11 18:44:16 cedric_: I know. Its broken for me in several ways which I've filed a bug about Apr 11 18:44:25 ok Apr 11 18:46:18 RP: done, should I comment in the RRECOMMENDS? Apr 11 18:46:39 cedric_: I'd prefer not at first to see if that bitbake revision works any better Apr 11 18:49:11 RP: and to patch out the document generation where is that? :-) --> work/armv5te-linux/librsvg-2.11.1-r0/librsvg-2.11.1/Makefile ??? Apr 11 18:50:20 Grep the Makefile.am for SUBDIRS and "doc" or something similar Apr 11 18:50:39 or you can patch the Makefile directly if you can file the doc part in that Apr 11 18:51:25 If you edit Makefile.am, you'll have to recofigure the package Apr 11 18:51:55 run configure again even Apr 11 18:53:17 RP: I'm using bitbake head as well and have compile issues with hostap.... Apr 11 18:53:26 RP: I just removed the RDEPENDS as a quick fix... Apr 11 18:56:14 i just wrote what has to be the most disgusting build patch ever Apr 11 19:12:43 I didn't have installed gtkdoc-mkhtml on my work box therefore the document generation (from librsvg) skipped in some way... Apr 11 19:18:48 Hi all ! Apr 11 19:19:28 hi Ifaistos Apr 11 19:21:19 Hi Richard. Well i made it. My OE based distro is up and running on 4 platforms... one more to go :) Apr 11 19:21:44 mips4k Apr 11 19:23:46 Ifaistos: That's good to hear :). Hopefully this one shouldn't be too bad now you know more about OE! :) Apr 11 19:25:19 RP: Don't think i know much about OE (yet). I am still copying things from others :) Apr 11 19:26:10 RP: Today we did out first long range test link. 4 Km using 802.11a cards with madwifi Apr 11 19:26:55 RP: We got something like 30Mbits on tcp bw tests Apr 11 19:27:52 Ifaistos: That's quite impressive. I use a link like that myself! :) Apr 11 19:28:49 RP: The good thing with OE is that you can scale the distro up or down VERY easy Apr 11 19:30:00 Ifaistos: Once the packages are there and you have machine support, you have lots of options Apr 11 19:30:08 RP: Have a look -> http://wind.awmn.net/?page=nodes&node=2030 Apr 11 19:30:27 RP: This is our Athens network Apr 11 19:31:06 RP: and this is my main wireless node Apr 11 19:33:26 Ifaistos: For 4km you'd need line of sight between the points? Apr 11 19:34:28 RP: Yes you do. Especialy in 802.11a where the fresnel zone is very narow Apr 11 19:39:24 RP: This is the webpage of my wireless node -> http://213.5.161.72/ Apr 11 19:40:11 RP: The inet entry point is something like 30Km and 10 nodes away Apr 11 19:40:17 Ifaistos: I like it :) Apr 11 19:41:54 RP: at last I've a gpe image built! Apr 11 19:42:06 RP: It might look Greek to you.... :) But i believe you get the picture :) Apr 11 19:43:24 Ifaistos: It is indeed all Greek to me :) (but a picture is worth 1000 words) Apr 11 19:54:08 lamikr, mreimer: are you using udev >= 084 on h6300 and h2200? any reason not to upgrade familiar 0.8.4 to use it? http://www.handhelds.org/~bugzilla/show_bug.cgi?id=1525 Apr 11 19:57:43 * chouimat is away: out in sun Apr 11 19:58:23 pH5: We're using it for OZ 3.5.4.1 Apr 11 20:03:25 pH5: I've been using 0.84 on h2200 Apr 11 20:03:35 RP: Right. I can't imagine that there would be problems for h2200 and h6300 that we don't have for the zaurii or hx4700. Apr 11 20:05:04 mreimer: cool, thanks. Apr 11 20:05:18 pH5: did you get to test hx4700-flash? Apr 11 20:05:33 pH5: The Zaurus users have reported odd bugs to do with missing or malfunctioning rules but I've been fixing these in OE. No problems with udev itself have been found that I know of Apr 11 20:06:19 I can't remeber the function of this channel Apr 11 20:06:27 was this a gui set for pda's or just embedded in general Apr 11 20:07:05 fishhead: www.openembedded.org Apr 11 20:07:19 does usbnet work with udev yet? Apr 11 20:07:28 thx Apr 11 20:07:53 bt while not directely related to this channel but sorta Apr 11 20:07:58 and for usa people only Apr 11 20:07:59 nite all Apr 11 20:08:04 nite RP Apr 11 20:08:05 and might only be my local kmart so I apolgise Apr 11 20:08:21 they have juicebox mp3 kits on sale for $10 plus a 20% off all toys in clearance = $8 Apr 11 20:08:31 mreimer: my current bootstrap image is horribly broken in many ways, but writing to flash seems to work so far... Apr 11 20:08:46 it's a 32 meg sd/mmc card, usb to sd reader that works in linux, and a usb extension cable all of which are standard items Apr 11 20:09:11 the only thing jb specific is the sd/mmc to juicebox cart slot adaptor and it's realy just a pcb with a sd/mmc slot = nice little breakout board for sd/mmc Apr 11 20:09:22 if I wasn't so broke I would hoard up a bunch Apr 11 20:09:27 RP: do you think we could add the rtc0 symlink and the modprobe rule from udev-084/local.rules to files/local.rules and then remove {udev-084,udev-089}/local.rules? Apr 11 20:09:35 pH5: glad to hear that writes work Apr 11 20:10:28 pH5: I branched it as I expected we'd end up adding things that weren't backwards compatible Apr 11 20:11:03 RP: I just noticed that with adding udev-089 I had to duplicate quite a lot of files Apr 11 20:11:19 pH5: Yes, not good :-/ Apr 11 20:12:54 whoa cool Apr 11 20:12:59 this channel sounds perfect for me Apr 11 20:13:06 anything targeting a 486-100 with 8 meg of ram Apr 11 20:13:08 * fishhead ducks Apr 11 20:13:11 DON'T HIT ME! Apr 11 20:13:16 * fishhead curls into a ball Apr 11 20:13:34 it's the core of a laptop, motherboard + floppy + 640x480x65535 lcd Apr 11 20:13:43 if the lcd didn't fire up I would smash it Apr 11 20:13:46 not worth my time Apr 11 20:15:15 might try hacking the board later to get it up to 12 meg of ram Apr 11 20:15:26 it has the via's on the board for it just need to add the chips Apr 11 20:16:04 fishhead, you could build your own image for that. Or you install a minimal debian and add GPE with matchbox on top of that. Apr 11 20:16:08 it's battery pack was 12 volt nicd so it has the nice feature of being able to be run on car battery power or sla batteries Apr 11 20:16:12 ok Apr 11 20:16:23 RP: would it be bad style if we made udev_089 fetch files from udev-089/ via some FILESPATH or FILESDIR magic? Apr 11 20:17:25 oh lemme show yuo guys this hardware hack Apr 11 20:17:39 might find it useful if you use a old pc for a embedded project and it can't direct drive a lcd Apr 11 20:17:59 pH5: Do you not mean hack it to fetch from 084? Apr 11 20:18:17 http://lcdgallery.net/?f=view&lcdid=293 Apr 11 20:18:30 it's only 16 colors, 256 max possible with some kinda ahck fix Apr 11 20:19:12 but if you have a old laptop lcd that is TFT/active matrix and is digital rgb inputs - CAN NOT BE THE NEWER LVDS signalling, you can attach the lcd to the "vesa feature connector" of a video card and drive it Apr 11 20:19:55 I say hell yeah! for somthign such as a embedded box, server or somethign that doesn't need a full screen to suck down power etc and a small lcd for status info like this is great Apr 11 20:20:07 RP: I mean add udev-084/ to the path after udev-089/ so that it could fetch flags.patch from udev-089/ and the rest from udev-084/ Apr 11 20:20:14 www.lcdinfo.com for more hacks Apr 11 20:20:33 pH5: I'd accept something like that. We already do things like this for native packages :) Apr 11 20:21:57 http://forums.bit-tech.net/showthread.php?t=51818&page=3&pp=20 Apr 11 20:22:06 RP: evil :) Apr 11 20:27:19 pH5: sorry I noticed your udev message only now. Yes, h6300 uses udev. What I need to do to make 083 to build udev084 instead of building 070 for h6300? Apr 11 20:27:57 Anyone know whats wrong? I am trying to build a bootstrap-image with BitBake and it refuse to build the root_fs: ERROR: Function do_rootfs failed Apr 11 20:29:59 goxboxlive: look higher, that's not the real error Apr 11 20:32:23 lamikr: can you grab udev 084 from org.openembedded.dev? Apr 11 20:34:24 pH5: I already have 083/packages/udev/udev_084.bb but when I made today a build bitbake chose to use udev_070.bb instead. Apr 11 20:35:14 lamikr: ah, is there a DEFAULT_PREFERENCE = "-1" inside udev_084.bb? Apr 11 20:35:50 ooh default preference.. so thats how you mask new packages Apr 11 20:35:58 pH5. Yes, in the first line. So I just take that away and try to rebuild. Apr 11 20:36:36 lamikr: yes, that's right. Apr 11 20:38:12 Thanks, I will try it and write tomorrow to bug 1525 how it went. Apr 11 20:38:22 lamikr: great, thanks! Apr 11 20:47:26 RP: the boot brightness on the 2.6.16 on spitz seems to be at max. again. could you again lower this to half or quad brightness please? Apr 11 20:47:50 RP: did you have success in booting alpha2/gpe? mine just hangs in a loop when trying to start gpe Apr 11 20:48:00 sounds like the X server dumps core Apr 11 20:48:04 looks, even Apr 11 20:48:28 identical problem w/ Akita Apr 11 20:48:37 mickeyl: My spitz gpe alpha2 image booted fine Apr 11 20:48:38 thanks. not a local problem then. Apr 11 20:48:42 no altboot and dropbear instead of openssh-sshd Apr 11 20:48:45 mickeyl: THis was my own image though Apr 11 20:48:51 RP: can you give me that, please? Apr 11 20:48:59 mickeyl: np, one second Apr 11 20:49:24 this "build image from feed" thingy is way buggy :\ Apr 11 20:49:34 CoreDump|home: i suspect ipkg :/ Apr 11 20:49:48 bbiab Apr 11 20:52:04 mickeyl: http://sam.rpsys.net/openzaurus/3.5.4.1-alpha2/ Apr 11 20:52:52 * JustinP pushes what he hopes to be a non-controvertial revision Apr 11 20:53:01 03justinp 07org.oe.dev * rabc1bab2... 10/packages/tslib/tslib-maemo_0.0.1-15.bb: tslib-maemo: add DEFAULT_PREFERENCE = -1 as this is a special purpose tslib provider Apr 11 20:56:07 * JustinP wonders what entrance's problem is now.... Apr 11 20:56:34 ::sigh:: someone decided to be clever about auto-finding configuration again....it's trying to call sudo Apr 11 20:57:23 RP: thanks. downloading that now. any idea about the brightness sunrise? :) Apr 11 20:57:34 Is it right that /etc/zaurusd/scripts/zaurus-hinge does apm -s at closing the lid? And lots of gpe-stuff, anyway? Apr 11 20:58:51 #define gettext(X) X Apr 11 20:58:59 today's big acheivement Apr 11 20:59:38 heh. not bad ;) Apr 11 21:00:00 polyonymous: yeah, that sounds like it should be atm. Apr 11 21:00:15 mickeyl, apm -s is for suspend? Apr 11 21:00:17 yes Apr 11 21:00:31 at some time we should have some configuration for what zaurusd does Apr 11 21:00:36 i got pmount to compile without its dependencies on gettext, intltool, glib and hal Apr 11 21:00:37 wouldn't that be misleading if opie user can configure not doing anything at the event and still Apr 11 21:00:56 polyonymous: hmm yes. Apr 11 21:01:08 mickeyl, not a big deal for, of course, once I found it, but I think it may result in many questions. Apr 11 21:01:09 guess the script should be a bit more cooperative Apr 11 21:01:30 yup. and it's strictly gpe-related, not sure if it should be there for opie at all. Apr 11 21:01:39 I added exit 0 in the beginning, anyway :) Apr 11 21:03:01 mickeyl: any chance i could get push access? Apr 11 21:03:44 TheMasterMind1: sure. the usual process nowadays is to have you push some patches through the bugtracker at first. you see, we are starting to use policies :D Apr 11 21:03:58 mickeyl: zaurusd isn't configurable? What exactly *does* it do? Apr 11 21:04:17 JustinP, some its scripts aren't, I think. Apr 11 21:04:18 JustinP: sorry, i know not a bit about zaurusd to be hoenest Apr 11 21:04:30 except that it takes care about zaurus HAL things Apr 11 21:04:39 mickeyl: policies?! but why.. Apr 11 21:04:42 audio, soft-touchbuttons Apr 11 21:04:46 hinge Apr 11 21:05:05 TheMasterMind1: to convince the seniors that you woN Apr 11 21:05:15 't create additional workload by pushing bogus stuff Apr 11 21:05:35 mickeyl: audio? Apr 11 21:05:40 Apr 11 21:05:41 =P Apr 11 21:05:47 mickeyl: would be nice to know what it tries to do with the hinge button.... Apr 11 21:05:56 * JustinP will lok at the code if he needs to Apr 11 21:05:58 JustinP: IIRC configuration changes when headphones are inserted and stuff. RP would know for sure Apr 11 21:06:33 mickeyl: ah yes, it does. I've seen that when trying to get the remote working....not sure what a daemon would need to do... Apr 11 21:06:51 I suppose it could be configurable how it changes the hardware..... Apr 11 21:07:08 the basic idea is to start reducing double work by doing zaurus specific things by a daemon + scripts instead of doing in seperatly in every GUI environment# Apr 11 21:07:13 which is quite a good thing to me Apr 11 21:07:22 we just need to find the proper borders Apr 11 21:08:15 ~curse tetex buildsystem Apr 11 21:08:16 May the fleas of a thousand camels infest your most sensitive regions, tetex buildsystem ! Apr 11 21:08:28 | configure: warning: Standard T1 library not found. Compiling my own. Apr 11 21:08:28 | ./configure: line 1750: test: =: unary operator expected Apr 11 21:08:28 | no, will compile my own version Apr 11 21:08:28 | checking for png_set_read_fn in -lpng... (cached) no Apr 11 21:08:28 | configure: warning: Standard PNG library not found. Compiling my own. Apr 11 21:08:31 *sigh* Apr 11 21:08:40 that worked the last time i looked at it Apr 11 21:09:09 zaurusd is perfectly configurable - overwrite the hinge action script with whatever you want it to do... Apr 11 21:09:19 It just needs tuning to suit the opie/gpe envrionments Apr 11 21:09:24 right Apr 11 21:09:34 hello :) Apr 11 21:09:38 Eventually, I'll pull this default one, it was just a "get it doing something" one Apr 11 21:09:58 t's definitly good for bootstrap Apr 11 21:10:19 flashing your alpha2 now.... Apr 11 21:10:37 mickeyl: Exactly. It covered the basics and has some hooks for gpe in. We just need to either teach it about opie, or have opie provide its own script Apr 11 21:10:50 RP: what does it do? I haven't looked into it yet. Apr 11 21:11:02 Does someone know if familiar-0.8.4-rc1 is built from oz354fam083 or not ? Apr 11 21:11:02 RP: I had my hinge script set up to suspend unless mpd was playing ^_^ Apr 11 21:11:26 JustinP: At the moment three things - 1) manages alsamixer for you 2) managed the hinge under gpe 3) adds softbutton support for the buttons on the touchscreen Apr 11 21:11:53 does it feed the keypresses back in via uinput? Apr 11 21:12:02 mickeyl: it does, yes Apr 11 21:12:06 pretty cool Apr 11 21:12:40 any idea about the brightness shock? Apr 11 21:12:43 dkm: probably Apr 11 21:13:10 mickeyl: The kernel defaults are problem wrong after we smoothed out the brightness curve on spitz Apr 11 21:13:27 RP: ah i see. Apr 11 21:15:24 mreimer: ok, thanks :) I was not sure because the distro conf still has 0.8.3... Apr 11 21:15:58 gah!!! busybox umount is driving me insane Apr 11 21:21:00 RP: yeah. that one boots fine Apr 11 21:23:19 mickeyl: How do you think 3.5.4.1 is looking? Apr 11 21:23:49 RP: gimme some hours to play with that and ask me again tomorrow :) Apr 11 21:23:50 hi (for a moment as usual) Apr 11 21:23:55 hey hrw Apr 11 21:24:02 i'm sorry to inform you that your alpha2 is broken Apr 11 21:24:05 I know Apr 11 21:24:09 oh. ok Apr 11 21:24:22 i'm sorry to inform myself that tetex no longer compiles here Apr 11 21:24:24 * mickeyl cries Apr 11 21:24:46 mickeyl: too bad that I do not know why it happen yet Apr 11 21:25:22 * Philippe is away: visual contact - melancholic dreams Apr 11 21:25:32 hrw: we need to install altboot per default to be able to intercept boot and debug it Apr 11 21:25:42 fetching RP image (known working) and going to look Apr 11 21:25:49 * mickeyl hugs opie's press any key to intercept Apr 11 21:25:53 mickeyl: thats reminds me of pushing.. Apr 11 21:25:57 hrw: My images appear to be working so you can release them to the users if appropriate Apr 11 21:26:40 hrw: I haven't tested c7x0 Apr 11 21:27:06 ok Apr 11 21:27:23 03pH5 07org.oe.dev * rcf0ac36e... 10/packages/linux/ (2 files in 2 dirs): handhelds-pxa-2.6: update K_MICRO and hx4700 defconfig to 2.6.16 Apr 11 21:27:32 looks like I will release them and go into looking what is fscked in mine Apr 11 21:29:55 jeez Apr 11 21:30:05 suddenly configure runs through Apr 11 21:30:06 I think releasing zImage would make sense, since bold users ipkg upgrade to 2.6.16, btw. Apr 11 21:30:10 * mickeyl starts to suspect magic or hardware Apr 11 21:30:11 modules that is Apr 11 21:30:51 hrw: I can provide the akita images Apr 11 21:31:20 anyway, I'm off to bed. n8 Apr 11 21:31:28 n8 Apr 11 21:31:30 03hrw 07org.oe.oz354x * raa140bb3... 10/conf/machine/zaurus-clamshell-2.6.conf: zaurus-clamshells: use altboot by default Apr 11 21:31:33 CoreDump|home: There are some akita images shared on my server - I build for all the 3.5.4.1 release targets Apr 11 21:31:34 03hrw 07org.oe.oz354x * rbf16d69a... 10/packages/xqt/xqt_0.0.9.bb: xqt: fix DEPENDS to get it build Apr 11 21:32:05 RP: that's even better then ;) my uplink suck _bad_ Apr 11 21:32:32 CoreDump|afk: Mine isn't much better - at least its not my main link atm though :) Apr 11 21:34:19 RP: fetching all into alpha2/.rp/ Apr 11 21:38:46 I have: gpe-session-scripts/gpe-bootsplash/keylaunch too old, gpe-task-* fscked (mickeyl build has many fscked packages :( Apr 11 21:39:07 diffed ipkg status or RP and my gpe/akita image Apr 11 21:39:08 really? how's that Apr 11 21:40:05 mickeyl: Package: gpe-task-connectivity Apr 11 21:40:05 Version: 1.0-r44 Apr 11 21:40:05 Depends: gpe-mini-browser Apr 11 21:40:05 Recommends: kernel-module-usbdserial, kernel-module-usbkbd, kernel-module-usbmouse, kernel-module-hid, usbutils, ker Apr 11 21:40:08 nel-module-evdev Apr 11 21:40:18 eek. where do those Recommends come from??? Apr 11 21:40:23 lot of stuff has that Recommends... Apr 11 21:40:38 mickeyl: Which revision of bitbake are you using? Apr 11 21:40:49 I'm tracing them from time to time and update packages in feed Apr 11 21:40:58 atm. 421. I don't remember which one I used to build head Apr 11 21:41:01 s/head/3.5.4/ Apr 11 21:41:26 * hrw goes to bed Apr 11 21:41:32 421 is very broken as far as I can tell - 419 seems the last version that works without certain bugs Apr 11 21:41:32 gnight hrw Apr 11 21:41:44 s/421/423/ Apr 11 21:41:47 'night hrw Apr 11 21:41:49 r329 works for 3.5.4.x (r359 probably too) Apr 11 21:41:56 zecke promised to work on it the days he's off Apr 11 21:42:02 let's see what he comes up with Apr 11 21:42:22 i'm pretty curious how much speedup we can get from that new cache loading scheme Apr 11 21:42:22 mickeyl: I know - I've just tried to isolate the problem and give him a reproducable problem Apr 11 21:42:27 RP: excellent Apr 11 21:43:03 mickeyl: I'm confident the new caching ideas will make it fly Apr 11 21:43:16 RP: that would rock Apr 11 21:43:18 NOTE: package tetex-3.0-r4: task do_compile: completed Apr 11 21:43:21 *phew* Apr 11 21:43:22 finally Apr 11 21:43:22 currently we do 300MB of IO for a build which is why it takes as long Apr 11 21:43:51 yeah right. when i initially implemented the cache i was surprised that it was so big, but didn't think about trying to reduce it Apr 11 21:44:24 mickeyl: I also wonder if we can't have a cache that benefits from COW Apr 11 21:44:40 mickeyl: by using one cache file instead of one per .bb file? Apr 11 21:45:23 hmm. with the current scheme i don't see how we can handle that because we need to check each and every filestamp against the bbfile stamp Apr 11 21:45:42 mickeyl: We cache the stamps separately Apr 11 21:46:42 mickeyl: We could even save the timestamp as a variable - it'd be inkeeping with the way bitbake works Apr 11 21:46:53 mickeyl: We're going to have to do this with the new cache idea anyway Apr 11 21:46:55 hmm. yes, but what would we gain? we still need to reparse every thing that changed timestamp and reparse all depends Apr 11 21:47:13 You gain being able to combine into one file Apr 11 21:47:20 Which might preserve COW Apr 11 21:47:27 Which means less disk IO Apr 11 21:47:33 Which means its faster :) Apr 11 21:47:39 hmm, ok i see your point. Apr 11 21:48:10 The question is whether pickleing preserves whatever magic COW uses Apr 11 21:48:32 I don't know python well enough to know the answer to that Apr 11 21:48:50 i don't know the magic COW uses, so I can't answer as well Apr 11 21:48:51 heh Apr 11 21:49:01 i wonder whether we could use marshall instead of pickle Apr 11 21:49:13 marshall is faster as well, but doesn't know about structured datatypes and recursion Apr 11 21:49:25 something i will think about Apr 11 21:49:43 ok. zecke knows about COW so perhaps you could talk to him :) Apr 11 21:49:49 yea Apr 11 21:49:58 I'd like to know if it would shirnk the cache size Apr 11 21:50:04 i hope to even meet him in person in may Apr 11 21:50:08 on LinuxTag Apr 11 21:50:15 even better :) Apr 11 21:50:28 Although I hope to have the secondary cache implemented by then :) Apr 11 21:50:41 (not to say we can't speed up the larger one) Apr 11 21:50:50 (as well) Apr 11 21:51:01 right. the secondary is more appealing to me because I understand the concept better... it's simpler :) Apr 11 21:51:08 03mickeyl 07org.oe.dev * r6d1ae8b6... 10/packages/tetex/tetex_3.0.bb: tetex: make the build more reliable by forcing it to use our libtool Apr 11 21:51:33 mickeyl: The secondary cache will speed up secondary runs where not many files change (and not local.conf) Apr 11 21:52:03 mickeyl: To speed up the primary parse, we're going to need more ideas Apr 11 21:52:35 mickeyl: I'll be interested to see what happens to speed if we don't bother with the primary cache actually... Apr 11 21:52:58 RP: yes, that's something to toy with as well. Apr 11 21:54:25 need to sleep a bit Apr 11 21:54:26 g'night Apr 11 21:54:32 'night mickeyl Apr 12 00:00:52 03tmbinc 07org.oe.dreambox * re81d6cfe... 10/packages/dreambox/dreambox-dvb-modules.bb: dreambox-dvb-modules/dm7020: update dm7020 drivers Apr 12 00:01:01 03tmbinc 07org.oe.dreambox * rcbb2be44... 10/packages/tuxbox/ (3 files in 2 dirs): tuxbox-plugins: update plugins Apr 12 00:01:01 03tmbinc 07org.oe.dreambox * rcbaf6a69... 10/packages/tuxbox/dvbsnoop.bb: dvbsnoop: update Apr 12 00:01:04 03tmbinc 07org.oe.dreambox * rac6065f4... 10/packages/enigma/ (3 files in 2 dirs): enigma: update, remove unbuildable Apr 12 00:01:12 03tmbinc 07org.oe.dreambox * r95bab097... 10/packages/linux/linux-dm7025.bb: linux-dm7025: update multipid patch again Apr 12 00:01:14 03tmbinc 07org.oe.dreambox * re13f1062... 10/packages/dreambox/dreambox-dvb-modules.bb: dreambox-dvb-modules/dm7025: update Apr 12 00:18:40 03justinp 07org.oe.dev * r2322a571... 10/packages/libgpewidget/libgpewidget-hildon_0.102.bb: libgpewidget-hildon: add DEFAULT_PREFERENCE = -1 as this is a special purpose libgpewidget provider Apr 12 00:50:46 what filesystems can you use for an mmc card Apr 12 02:19:43 * mithro beats users with a stick Apr 12 02:20:59 Try the Flamethrower XL 9750! Beats users down with a single blast of combustion.. Apr 12 03:14:05 TheMasterMind1: any you want of course Apr 12 03:16:33 minix doesn't support long filenames, cramfs and squashfs are readonly, ext3 and resier are journaled, vfat doesn't support permissions and symlinks Apr 12 03:16:44 only ext2 is left, and it gets corrupted so easily Apr 12 03:19:18 TheMasterMind1: I run a lot on my ext2 card, and have had no problems Apr 12 03:43:05 TheMasterMind1: and journaling would work just fine as well....I also had little proble with ext2. Just fsck it once in a while Apr 12 03:49:36 journaling on a flash device that has limited read/write cycles? Apr 12 03:49:54 meh. though i do remember reading something about how it would last you 10 something years anyway Apr 12 03:50:36 depends on the quality of the device Apr 12 03:50:46 and the frequency of the cycles Apr 12 03:53:00 and how much you care about your SD card....you *can* buy a new one easily Apr 12 03:53:11 but ext2 is fine anyway Apr 12 03:54:29 JustinP, it depends if you have data on it when it failes Apr 12 04:03:52 TheMasterMind1: not limited writes/reads but erases. Apr 12 04:04:15 TheMasterMind1: modern flash gets 100,000 erases per sector.. Apr 12 04:04:34 TheMasterMind1: the number of writes does not matter Apr 12 04:04:48 err Apr 12 04:05:01 rewritiing is just as bad as erasing Apr 12 04:05:06 TheMasterMind1: you could erase a sector..write the first byte today, the next byte in the sector tomorrow. Apr 12 04:05:14 rewriting _is_ eraseing, isn't it? Apr 12 04:05:21 not really Apr 12 04:05:37 can't rewrite flash, eeprom but not flash. Apr 12 04:05:38 assuming your data has roughly the same number of 1s and 0s, or so? :-) Apr 12 04:05:49 1111 -> 0000 vs 1111 -> 1101 Apr 12 04:07:42 yeah france i see what yoru saying ... Apr 12 04:07:50 i never actually thought about taht Apr 12 04:08:13 but yeah there is a erase then write cycle for flash Apr 12 04:08:35 emte: there is but it is rarely utilized Apr 12 04:09:06 pretty sure we use it for the PIC stuff all the time Apr 12 04:10:02 makes sense now why most chips have an eeprom onboard Apr 12 04:10:17 tho they burn out in a few seconds as well if you do bad things Apr 12 04:10:30 100,000 erases is still a lot... Apr 12 04:10:45 not at 20Mhz Apr 12 04:11:10 emte: erases take a long time...:-D Apr 12 04:11:23 emte: lots of waits. Apr 12 04:11:25 lol Apr 12 04:11:34 only in electronic time Apr 12 04:11:42 in human time its pretty short Apr 12 04:12:41 emte: if I did nothing but erase my 36xx, it would take almost a year to reach a 100,000 erases Apr 12 04:12:43 i cant rememebr, but 1024 at 20Mhz will burnout in ~8min Apr 12 04:13:27 lol Apr 12 04:13:48 then you would have soem fun resoldering to do :P Apr 12 04:14:22 That's what sockets are for.. Apr 12 04:14:32 lol Apr 12 04:14:48 it is not that bad, I have upgrade the flash on my 36xx to 64MB Apr 12 04:15:00 i keep thinking about doing that Apr 12 04:15:05 remove a chip, add a chip Apr 12 04:15:07 stacking the ram Apr 12 04:15:14 as well Apr 12 04:15:18 that works as well Apr 12 04:15:35 but i seem to run out of money when i deceide to try and order ... Apr 12 04:16:30 heh Apr 12 04:16:41 I have to run..nite Apr 12 04:16:46 night Apr 12 04:19:46 03rwhitby 07org.oe.dev * r3ddcadd3... 10/packages/meta/slugos-native.bb: slugos-native: Fixed the libgdbm3 RDEPENDS issue. Apr 12 04:19:50 03rwhitby 07org.oe.dev * r71b5bd97... 10/packages/meta/slugos-native.bb: slugos-native: Removed the gdbm3 run-time dependency. Apr 12 05:54:25 when trying to build something (outside of bitbake), I'm getting 'undefined references' to AtiCore_* methods in libqte Apr 12 05:54:34 any idea what I might be doing wrong? Apr 12 05:54:44 does libqte depend on another library? Apr 12 06:02:51 03rwhitby 07org.oe.dev * rd204b509... 10/packages/meta/slugos-native.bb: slugos-native: Fixed the perl5 rdepends problem. Apr 12 06:23:26 ok, looks like my libqte was fubar'd, somehow Apr 12 07:11:21 ashikase: you need to add the library with ATI stuff to your LIBS += of your qmake file Apr 12 07:11:55 actually, I found that the qte lib I built through bitbake didn't have the problem Apr 12 07:12:18 and I didn't see any other libraries that contained those functions Apr 12 07:13:44 ashikase: a library can specify that it needs sub-libraries, which the get loaded automatically. Seems that either the bitbaked Qt does this ... or that the bitbaked Qt doesn't use the ATI hw accel, but a plain framebuffer Apr 12 07:14:14 ashikase: however, please re-ask this in some hours, then more people are here. /me is not an expert on Qt/E on devices with ATI chips Apr 12 07:14:31 (I use it on a PXA255 device where I just have a plain framebuffer, nothing else) Apr 12 07:14:45 right Apr 12 07:15:02 time is always a problem on irc Apr 12 07:28:47 morning Apr 12 07:29:31 ashikase: libqte2 for c7x0 devices is linked with aticore library to get some speedups on ati w100 Apr 12 07:35:42 ah, that would explain my problem then Apr 12 07:35:47 I was using the wrong lib Apr 12 07:36:00 I had been using a c760, I'm now using my borzoi Apr 12 07:37:04 ;) Apr 12 07:37:06 hi alan Apr 12 07:38:34 hi hrw Apr 12 07:38:38 hi all Apr 12 07:39:11 morning all Apr 12 07:45:48 someone brave want to check image? Apr 12 07:57:44 morning all Apr 12 07:57:54 hi Dirk Apr 12 07:58:42 do13_: must sound link "hei dörk" :-) Apr 12 07:58:56 s/link/like/ Apr 12 07:59:06 Hey Richard Apr 12 07:59:10 schurig:) Apr 12 08:07:27 morning Apr 12 08:08:29 morning all Apr 12 08:09:21 hi CoreDump|home, XorA Apr 12 08:23:18 ~lart qt build system for thinking 64bit arm is a valid platform Apr 12 08:23:19 * ibot whacks qt build system upside the head for thinking 64bit arm is a valid platform Apr 12 09:02:50 good morning Apr 12 09:03:01 morning Bernardo Apr 12 09:03:23 hi Bernardo Apr 12 09:04:50 stupid question, with altboot can I need to create a ext2fs partition on my cf card and expand a .tar.gz image there, so I can boot from it, right? Apr 12 09:05:07 right Apr 12 09:05:23 CoreDump|home: can I boot from /dev/hda2 with altboot? Apr 12 09:05:26 ok, time to test gpe-image on a 128Mb cf Apr 12 09:05:35 hrw|work: no, not yet Apr 12 09:05:48 CoreDump|home: collie: /dev/hda1 vfat 16M for flashing, /dev/hda2 ext2 for system Apr 12 09:06:28 one day I'll add multi-parition boot to hda for the microdrive guys Apr 12 09:06:45 CoreDump|home: Yes please :) Apr 12 09:07:26 * RP really needs to start using mutliple installs on spitz... Apr 12 09:07:37 CoreDump|home: that day add a way to go back a menu too... I had left my akita rebooting now and found out it was going back to the advanced menu too soon Apr 12 09:08:10 I mean too late Apr 12 09:08:11 that is on my todo list yes Apr 12 09:08:28 but the menu-code is one ugly mess :\ Apr 12 09:08:30 CoreDump|home: ability to use qwertyuiop instead of 1234567890 on collie too? Apr 12 09:09:33 hrw|work: that would involve some very painful recoding. making the menu stuff even worse Apr 12 09:09:43 ok so ignore it Apr 12 09:13:49 RP: have you ever seen c7x0 suspending automatically on kernel boot Apr 12 09:13:55 ? Apr 12 09:14:16 hrw|work: No. Low battery? Apr 12 09:14:23 RP: AC connected Apr 12 09:14:33 RP: AC as usb/sync usb cable Apr 12 09:14:36 uff.. booted Apr 12 09:15:26 hrw|work: I've never seen it do that. Its technically possible for it to suspend any time after init starts though Apr 12 09:19:11 we need to add logo into release kernels... Apr 12 09:19:28 hrw|work: testing your gpe-image now... Apr 12 09:19:29 that would be neat, yeah Apr 12 09:19:36 hrw|work: ELPP? Apr 12 09:20:08 RP: no - kernel logo Apr 12 09:20:33 hrw|work: ah, right, that would be nice Apr 12 09:21:00 or at least some text like "Loading kernel, please wait" =) Apr 12 09:21:12 as kernel load takes ages on akita Apr 12 09:21:12 hrw|work: It will oversize the spitz/akita kernels without the changes I made in .dev Apr 12 09:21:19 CoreDump|home: altboot DOES NOT work on c7x0 as it should Apr 12 09:21:37 CoreDump|home: altboot NEED to load keymap probably Apr 12 09:21:42 hmm, what's the problem Apr 12 09:21:46 The kernel is really fast on spitz due to the microdrive - no jffs2 Apr 12 09:22:00 'press any key to launch altboot Apr 12 09:22:11 and any key does not work Apr 12 09:22:18 uh oh Apr 12 09:22:31 well it reads kbd interrupts from /proc Apr 12 09:22:47 ~lart zap_root_password Apr 12 09:22:47 * ibot gets a hotmal account and SPAMs zap_root_password Apr 12 09:22:52 that should work ;) Also altboot loads any installed keymaps before showing the menu Apr 12 09:23:14 CoreDump|home: to show menu it need to start... Apr 12 09:23:30 ah unlucky wording on my side Apr 12 09:23:48 before it shows _anything_ it'll load the keymap Apr 12 09:23:53 I suspect my root GPE settings are not compatible with 3.5.4.1-alpha2 Apr 12 09:24:24 ~lart keylaunch, matchbox and rest of GPE for creating such problems Apr 12 09:24:25 * ibot executes killall -KILL keylaunch, matchbox and rest of GPE for creating such problems Apr 12 09:25:25 * Bernardo is scared... His akita doesn't show any sign of life since shutting down, not even reset works Apr 12 09:25:31 hrw|work: in hentges I prompt the user to trash ~/.gpe ~/.matchbox ~/.Xdefaults and ~/.keylaunchrc Apr 12 09:26:04 CoreDump|home: on first boot? Apr 12 09:28:01 at least the maintenance menu loads... Apr 12 09:28:38 thats idea Apr 12 09:29:05 ~hail sharp for maintaince rootfs Apr 12 09:29:09 * ibot bows down to sharp for maintaince rootfs and chants, "I'M NOT WORTHY!!" Apr 12 09:29:35 problems with altboot - won't boot from pcmcia Apr 12 09:29:46 I mean from cf Apr 12 09:30:31 I get a /proc is already mounted, then two errors in 20-bootCF line 526, with a /etc/init.d/pcmcia not found Apr 12 09:30:45 oops Apr 12 09:31:08 well, I have that fixed in .dev already Apr 12 09:31:16 :) Apr 12 09:31:29 the move from pcmcia-utils to udev caused that Apr 12 09:31:33 is there a quick hack I can do to 20-bootCF? Apr 12 09:31:59 comment out the line where it runs /etc/init.d/pcmcia statr Apr 12 09:32:04 start Apr 12 09:33:06 or pcmcia status, where it checks for /media/cf? Apr 12 09:33:44 any call to that file Apr 12 09:33:51 as it is no longer installed Apr 12 09:38:06 ok, it isn't enough, probably I need to start udev too Apr 12 09:39:08 yep. Starting udev even mounts /media/cf automatically Apr 12 09:39:38 Bernardo: oops Apr 12 09:39:55 I tried that yesterday and I didn't need to start udev :\ Apr 12 09:40:21 one small problem, "udev status" isn't supported, so I can't check if it is running Apr 12 09:40:28 CoreDump|home: on the akita? Apr 12 09:40:35 kernel 2.6.16 here Apr 12 09:40:44 yep, akita w 2.6.16 Apr 12 09:40:58 ps ax| grep udevd | grep -v grep ;) Apr 12 09:41:08 here if I don't have udevd running I don't have /devhda* Apr 12 09:41:10 yep Apr 12 09:41:11 :) Apr 12 09:41:24 hmmm Apr 12 09:41:45 I think it would be overkill to launch udevd just to create the device files tho Apr 12 09:42:56 CoreDump|home: maybe we need to add those device files into device_table.txt then? Apr 12 09:43:33 or just mknod them with altboot Apr 12 09:46:32 gpe booting... Apr 12 09:47:25 I just accepted the altboot defaults, seems everything is working Apr 12 09:47:59 and now I get a black screen with a centered cursor, is this normal for gpe? Apr 12 09:48:37 * Bernardo wonders if it is some gnomeish interface simplification... ;) Apr 12 09:48:40 Bernardo: when you touch screen to move cursor it goes back to center in few seconds? Apr 12 09:49:07 hrw|work: yes Apr 12 09:49:20 Bernardo: fsck Apr 12 09:49:31 ~lart GPE team badly Apr 12 09:49:32 * ibot hurls dozens of incontinent, insomniac, hungry kittens with tiny little razor-sharp claws and a wide variety of contagious intestinal parasites at GPE team badly Apr 12 09:49:51 keylaunch + matchbox is SHIT not enviroment Apr 12 09:49:57 lol Apr 12 09:50:14 * Bernardo doesn't like gnome or its derivatives Apr 12 09:50:35 heh Apr 12 09:50:37 if user has any keylaunch setting saved then we CANNOT change keylaunch and/or matchbox-wm/kbdconfig if we want to get GPE bootable Apr 12 09:50:42 shit Apr 12 09:51:30 hrw|work: any way to get around it? Apr 12 09:51:57 that shouldn't bee a keylaunch problem Apr 12 09:52:02 I think I am booting of a clean image, is altboot using my /home partition? Apr 12 09:52:04 *be Apr 12 09:52:15 Bernardo: yes it is Apr 12 09:52:25 Bernardo: rm ~root/.keylaunchrc if exist Apr 12 09:52:34 if is one of the "defaults" you selected ;) Apr 12 09:52:45 Bernardo: or even backup /home/root/ somewhere and clean it Apr 12 09:53:49 CoreDump|home: ok, will reboot and check for what prompt lets me use a /home in the cf card Apr 12 09:54:03 nope Apr 12 09:54:09 it is a one-time config Apr 12 09:54:20 you'll not be ask again. Apr 12 09:54:31 ok, what file do I need to edit/remove? Apr 12 09:54:44 checking... Apr 12 09:56:06 remove /etc/.image_conf.done from _whithin_ the loop-image Apr 12 09:56:36 ok Apr 12 09:57:43 heh.. under 2.4.18-crapix I cannot mount /home/ on c760 ;( Apr 12 09:57:53 hrw|work: hah! Apr 12 09:58:02 mount hangs / oopses right? Apr 12 09:58:13 CoreDump|home: c760 should be able to do it. akita/borzoi should not Apr 12 09:58:21 hehe Apr 12 09:58:36 hrw|work: It could be the jffs2 people have broken backwards compatibility :-/ Apr 12 09:59:04 hrw|work: the "mount" will hang, causing tmpfs not to be mounted Apr 12 09:59:15 that'll cause some more problems... Apr 12 09:59:43 ok. so today I have bricked zaurus thx to gpe **** ENDING LOGGING AT Wed Apr 12 09:59:57 2006