**** BEGIN LOGGING AT Thu Mar 01 02:59:57 2007 Mar 01 03:00:07 Maybe I'm missing something... I can pastebin some stuff if that would help. Mar 01 03:04:38 Anything else builds without a hitch. Mar 01 03:07:04 guys quick question Mar 01 03:07:51 can anyone think of anything that would cause oe to try and link with libs in oe/build/tmp/staging/[dist]/lib/.libs/ instead of oe/build/tmp/staging/[dist]/lib/ ? Mar 01 03:08:41 jselect: Broken autoconf? Mar 01 03:10:41 Glombool: ah I misunderstood. Place your modified tarball into your normal source dir and create source.tar.gz.md5 in the source dir Mar 01 03:11:36 Ok. I actually did it the other way. The problem was an inherit statement. Mar 01 03:13:47 NAbyss_: I'm looking now for the correct autoconf verions Mar 01 03:15:19 versions even Mar 01 03:18:43 Does onyone know if there is an openzaurus version of this: http://wiki.openmoko.org/wiki/MokoMakefile Mar 01 03:19:03 ie a make image that builds the oz distro Mar 01 03:19:17 image=script Mar 01 03:19:56 darmou: bitbake gpe-image Mar 01 03:20:18 thanks JustinP Mar 01 03:26:07 can I not run rm_work class for one .bb? Mar 01 03:26:34 I need to use it so I don't run out of hd space, but I need one .bb's work to stay Mar 01 03:28:41 03coredump2 07org.oe.dev * r5f7251c7... 10/ (1 classes/rootfs_ipk.bbclass): rootfs_ipk.bbclass: All good things... Mar 01 03:40:21 w00t, n/m Mar 01 03:47:25 it seems like autoconf is adding ".libs" to the end of my LDFLAGS lib include dir Mar 01 03:47:29 so broken Mar 01 05:28:05 03mwester 07org.oe.dev * r5a210747... 10/ (1 classes/rootfs_ipk.bbclass): Mar 01 05:28:05 rootfs_ipk.bbclass: placate Fedora Core 5 by removing only the contents Mar 01 05:28:05 of the oe-feed directory and not the directory itself. Mar 01 05:28:17 03mwester 07org.oe.dev * rdd370e13... 10/ (1 classes/rootfs_ipk.bbclass): disapproval of revision '5a21074748883be7137fee8d724cd0787c547ae8' Mar 01 07:32:30 <_law_> hi all is montone 0.32 already supported? Mar 01 07:32:43 supported? Mar 01 07:32:53 it works just fine if that's what you mean Mar 01 07:33:03 0.33 just came out ;-) Mar 01 07:34:26 <_law_> JustinP, ok my fedora systems updated my mtn to 0.32 today, now i have do db migrade or? Mar 01 07:34:58 if mtn says to migrate then yes Mar 01 07:35:06 unless you want to download a new snapshot Mar 01 07:35:48 <_law_> JustinP, i try to migrate.. Mar 01 07:47:43 good morning Mar 01 07:48:05 morning all Mar 01 07:48:30 koen: efika clean build was successful :) Mar 01 07:48:41 koen: doing some packages now Mar 01 07:48:57 koen: javastation sucesfully boots the filesystem! Mar 01 07:59:28 03koen 07org.oe.dev * r5e56ec60... 10/ (1 classes/rootfs_ipk.bbclass): rootfs_ipk: as per OE policy: remove feed management tools Mar 01 08:02:34 koen : can you build libcap ? Mar 01 08:02:54 not sure if it depends on a pre 2.6.20 kernel Mar 01 08:15:43 <_law_> mtn db regenerate_caches --db=OE.mtn wow this takes some minutes/hours ;-) Mar 01 08:20:00 koen: what policy? Got a link? Mar 01 08:20:45 Ifaistos, oxo: yay! Mar 01 08:20:56 NAiL: it should be on the wiki somewhere Mar 01 08:21:34 NAiL: I know the policy 'cause kergoth chewed me out a few years ago when a was complaining about the feed-split recipe Mar 01 08:21:44 which got promptly removed after larting me Mar 01 08:22:43 Ok, 'cause I can't find it in the wiki anywhere Mar 01 08:22:51 (Atleast not searching by policy) Mar 01 08:23:46 (or ipkg) Mar 01 08:24:18 maybe times have changed since then ;-) Mar 01 08:25:39 we've discovered serious bugs in ipkg-utils since then, so I guess "no, things haven't changed" Mar 01 08:31:02 03koen 07org.oe.dev * r815fd35c... 10/ (1 conf/machine/fic-gta01.conf): fic-gta01: set flash size to '60' instead of '60M' Mar 01 08:33:04 heh building packages for efika got another with QA problems -> http://rafb.net/p/MVnwcg77.html Mar 01 08:33:08 * oxo still has trouble getting glib-1.2 to crosscompile for sparc Mar 01 08:33:41 oxo: check the configure.ac file to see what it wants to check for Mar 01 08:34:13 Ifaistos: FILES_${PN}-dbg += "${libdir}/esmtp-plugins/.debug/" Mar 01 08:38:00 NOTE: Checking Package: libesmtp-locale Mar 01 08:38:00 NOTE: DONE with PACKAGE QA Mar 01 08:38:00 NOTE: package libesmtp-1.0.4-r1: task do_package: completed Mar 01 08:38:03 :) Mar 01 08:38:22 another bug down... Mar 01 08:39:57 koen : should this be added to the howto about fixing things ? Mar 01 08:40:17 you could do that Mar 01 08:40:22 ok Mar 01 08:41:12 can you build libcap ? Mar 01 08:41:40 not at the moment Mar 01 08:42:24 koen: there is no ac-symbol concerning the sizeof(pthread_t) :-( Mar 01 08:42:51 weird Mar 01 08:43:26 oh wait, there is one concerning sizeof(system_thread) Mar 01 08:43:42 I think ``glib_cv_sizeof_system_thread=${glib_cv_sizeof_system_thread=4}'' is safe Mar 01 08:44:01 * oxo is going to check on a native sparc (32) box Mar 01 08:46:19 hmm, isi that another koen in this channel or are you just happy to see me? :-) Mar 01 08:46:57 I'm just happy to see you :) Mar 01 08:47:09 I plugged in wired ethernet and osx doesn't like that Mar 01 08:47:23 did you speak to your gf about fridaynight yet? Mar 01 08:47:32 ah right Mar 01 08:47:44 friday night is a no-go, sorry Mar 01 08:48:54 hmm, then we should reschedule, but I'll have to consult my gf for that :-) Mar 01 08:49:47 :) Mar 01 08:51:10 morning Mar 01 08:53:19 well, my other javastation is netbooting, so in a minute I can install GCC and check how big a pthread_t is Mar 01 09:07:50 koen: can you add glib_cv_sizeof_system_thread=${glib_cv_sizeof_system_thread=4} and glib_cv_sizeof_gmutex=${glib_cv_sizeof_gmutex=24} in site/sparc ? Mar 01 09:12:00 03koen 07org.oe.dev * rc4050094... 10/ (1 site/sparc-linux): sparc-linux: add more glib entries, provided by Jorik Jonker Mar 01 09:13:08 tnx Mar 01 09:15:18 * oxo also packaged the both tetrinet and tetrinet-server Mar 01 09:15:24 never leave home without them :-) Mar 01 09:15:29 heh Mar 01 09:15:47 did you use tune-supersparc.inc as well? Mar 01 09:16:15 euhm, lets see, i dont think so (yet) Mar 01 09:16:38 I would be a very good idea to do that Mar 01 09:16:50 makes ssh a bit faster :) Mar 01 09:16:53 but you know that Mar 01 09:16:59 supersparcs have the advantage over 'vanilla' sparcs that they do have integer fpu stuff Mar 01 09:17:03 yeah Mar 01 09:17:12 and hardware multiply Mar 01 09:18:07 it looks like irssi (0.8.10) depends on glib-1.2 where it should depend on glib-2.0 (according to irssi's configure output) Mar 01 09:30:09 koen, I have another one Mar 01 09:30:16 next time I'll try to group them :-) Mar 01 09:30:16 glib_cv_stack_grows=${glib_cv_stack_grows=no} Mar 01 09:31:11 03nail 07org.oe.dev * r30f9adf1... 10/ (1 conf/machine/turbostation.conf): turbostation.conf: Add machine turbostation Mar 01 09:31:15 03nail 07org.oe.dev * ra9e6f94b... 10/ (18 files in 3 dirs): linux: Add linux-turbostation Mar 01 09:31:24 oxo: I'll wait till lunch with committing those :) Mar 01 09:31:31 very good Mar 01 09:34:54 * oxo can confirm that irssi-0.8.10 should not depend on glib-1.2, but on glib-2.0. The website tells so, and irssi's configure script actually tried to download glib-2.0 (!) when it discovers the lack of it Mar 01 09:35:09 shall I bugzilla this? Mar 01 09:36:29 yes please Mar 01 09:36:29 look at bugzilla first Mar 01 09:36:49 * koen is struggling with matlab, epstopdf and pdflatex at the moment Mar 01 09:37:02 those are more than often good friends to me Mar 01 09:37:04 can I help? Mar 01 09:39:14 it just tedious copy&pasting Mar 01 09:40:52 * koen hugs laprint.m Mar 01 09:41:26 * hrw larts it2007 again Mar 01 09:41:40 it is ai_nw_installing and ai_nw_installing.. ai_ap_application_installer lack better progress informations.. (more detailed) Mar 01 09:42:14 hi liam Mar 01 09:43:21 hrw: thanks, someone noticed it already (#1640) Mar 01 09:44:22 hello everyone Mar 01 09:45:13 has someone ever included udev rules for rt2570 support in OE ? Mar 01 09:45:56 koen: what is the problem with the kernel I added? Mar 01 09:46:24 NAiL: there isn't really a problem with it Mar 01 09:46:35 NAiL: it would have been nice to add the patches to linux_2.6.20.bb Mar 01 09:47:51 Then I'd have to modify linux_2.6.20.bb extensively Mar 01 09:50:51 NAiL: it's a goal, not a policy :) Mar 01 09:52:14 NAiL: it's a bit wastefull to get conf/machine/foo.conf + conf/distro/openfoo.conf + packages/linux/linux-foo.bb for every new machine Mar 01 09:54:32 yeah, I haven't made a distro .conf though Mar 01 09:55:54 it's fine to have those 3, but it would be better to reuse existing stuff Mar 01 09:56:02 I've got a question regarding MACHINE_FEATURES and usbhost: for mx31ads, MACHINE_FEATURES does include usbhost. mx31ads usbhost kernel drivers gets built, however it doesn't get included into the rootfs although I'm using angstrom which also inculdes usbhost Mar 01 09:56:11 e.g. distro=generic or distro=slugos, etc Mar 01 09:57:32 I have to install the ipk afterward (it works perfectly fine, but would be nice to have it included in the generated image). What could be wrong here ? Mar 01 09:58:58 vlo: wrong module name, missing module-name in task-base or just a bug Mar 01 10:01:30 ok ... I guess I need to understand a bit more this "task-base" concept. But module name should be ok, since its ehci-hcd (kernel-module-ehci-hcd_2.6.16.19-r0_mx31ads.ipk) Mar 01 10:01:52 which is the standart linux name for usb 2.0 host Mar 01 10:02:28 vlo: task-base probably lack ehci entry Mar 01 10:04:45 hrw: going to check, this will allow me to better understand task-base Mar 01 10:08:32 good morning folks Mar 01 10:09:11 vlo: pushed fix Mar 01 10:09:45 hrw: you were right: task-base-usbhost-rrecommend only contains kernel-module-ohci-hcd, not kernel-module-ehci-hcd Mar 01 10:10:09 hrw: ok, you were quicker ;) Mar 01 10:13:04 * koen merges Mar 01 10:14:17 03koen 07org.oe.dev * r24e6547f... 10/ (1 packages/tasks/task-base.bb): task-base: also ship ehci-hcd for usbhost Mar 01 10:14:21 03hrw 07org.oe.dev * rada5eb2f... 10/ (1 packages/tasks/task-base.bb): task-base: add EHCI/UHCI controllers into task-base-usbhost Mar 01 10:14:29 03c.romain 07org.oe.dev * ref143b9e... 10/ (1 packages/irssi/irssi_0.8.10.bb packages/irssi/irssi_svn.bb): irssi: fix DEPENDS, closes #1640 Mar 01 10:14:31 03koen 07org.oe.dev * r636853b7... 10/ (1 packages/tasks/task-base.bb): task-base: fix up merge, we don't need ehci twice Mar 01 10:14:46 ;d Mar 01 10:17:28 thank you koen and hrw Mar 01 10:20:12 first bug posted to maemo bugtracker Mar 01 10:20:54 btw, we're having another bugdays thing this weekend Mar 01 10:21:06 any volunteers to announce it? Mar 01 10:21:15 another one where I will not without computer access Mar 01 10:22:18 http://www.widelec.org/index.php?site=detail&id=669 - coffee someone? Mar 01 10:22:21 koen : gcc 4.1.1 fails with the latest build of angstrom on efika Mar 01 10:23:53 koen : -> http://rafb.net/p/kXDFMs96.html Mar 01 10:24:34 that one is weird Mar 01 10:27:29 i am having problems to build gcc on dht also, especially if i apply the soft-float patches Mar 01 10:28:02 but did not have time to look into it for more details Mar 01 10:48:12 hi all ! Mar 01 10:48:24 hi gremlin[it] Mar 01 10:49:23 hi hrw! are u plying with your new 770 ? Mar 01 10:49:41 gremlin[it]: read my blog :) Mar 01 10:54:25 hi, all! Mar 01 10:54:39 hrw, url? Mar 01 10:55:09 http://www.hrw.one.pl/ Mar 01 10:58:58 hihihi :) Mar 01 11:33:09 RP: please ask ross to wait a couple of days (well until I recovered) before he merges the insane changes Mar 01 11:35:43 zecke: gws Mar 01 11:36:05 Is there a way to set SRCDATE="now" on a temporary way for all packages? Mar 01 11:36:31 put SRCDATE = "now" in local.conf? Mar 01 11:36:37 ok Mar 01 11:36:40 and remove it afterwards Mar 01 11:36:48 I just tried export SRDDATE ... b4 running bb Mar 01 11:44:20 thanks koen Mar 01 11:47:49 03svolpe 07org.oe.dev * r83be462b... 10/ (5 files in 4 dirs): fceu r1: Add FCEU NES Emulator Mar 01 11:52:03 hrw, Hopefully the next OSK image I put on the celinux site is based on angstrom Mar 01 11:53:52 morning Mar 01 11:54:05 good morning chouimat Mar 01 11:55:01 Gerrath_svolpe: I not sure it will be good ... I have to spend the day from 8:30 to 14:somenthing in a meeting Mar 01 11:55:16 koen: "rootfs_ipk: as per OE policy: remove feed management tools." Which is of course complete bullshit since the rootfs itself is installed by feeds created w/ OE. The whole _point_ of rootfs_ipk.bbclass is to create a feed so a rootfs in whatever format can be "installed" using that feed. I'd have hoped you'd think of a more creative excuse to revert this absolutley valid und useful commit than this.... Mar 01 11:55:36 * CoreDump|home creates an oe-patches folder Mar 01 11:55:55 chouimat, I know how you feel, we are in the middle of an ISO audit.. Mar 01 11:56:26 Gerrath_svolpe: ouch Mar 01 11:58:59 CoreDump|home: no, the point of rootfs_ipk is to build a rootfs from ipkgs, nothing more, nothing less Mar 01 11:59:24 which it does by creating a feed. Or multiple now Mar 01 12:03:39 koen: I have found two other site/sparc-linux thingies: Mar 01 12:03:40 glib_cv_stack_grows=${glib_cv_stack_grows=no} Mar 01 12:03:40 ac_cv_linux_vers=${ac_cv_linux_vers=2} Mar 01 12:03:40 bash_cv_have_mbstate_t=${bash_cv_have_mbstate_t=yes} Mar 01 12:20:04 can we create a feed creation bbclass, rather then piggybacking on an existing one Mar 01 12:22:57 Crofton: that's a much better idea, but I still wouldn't put it in OE Mar 01 12:22:58 Crofton: of course "we" could. It's just that koen would revert it Mar 01 12:23:01 maybe in contrib/ Mar 01 12:23:05 Crofton: bbclass? one recipe will be enough: bitbake feeds-cleanup-in-my-way Mar 01 12:24:03 Crofton: http://hentges.net/tmp/oe-patches/dev/ Mar 01 12:24:31 The trick is to keep OE useful, without going in numerous directions Mar 01 12:24:49 03koen 07org.oe.dev * re7a356a3... 10/ (4 files in 3 dirs): pango: update to 1.16.0 Mar 01 12:24:53 03svolpe 07org.oe.dev * r74bce808... 10/ (3 files in 3 dirs): libsdl-directfb r1: Add SDL DirectFB Version Mar 01 12:24:57 03koen 07org.oe.dev * ra6d86980... 10/ (1 site/sparc-linux): sparc-linux: add some more entries, courtesy Jorik Jonker Mar 01 12:25:23 well, the trick is not artificial crippling it via pseudo commits. Anyay, enough said. Mar 01 12:25:32 * CoreDump|afk goes to work Mar 01 12:28:13 tnx, koen Mar 01 12:28:49 koen: did you tried RP patches adding LZO compression to jffs2? Mar 01 12:29:04 hrw: not yet Mar 01 12:29:12 hrw: didn't even know they existed Mar 01 12:29:31 koen: RP posted them to lkml Mar 01 12:29:42 * koen has been too busy with signal processing lately Mar 01 12:29:51 40% read speedup Mar 01 12:30:03 This is particularly useful for jffs2 where significant boot time Mar 01 12:30:03 speedups (~10%) and file read speed improvements (~40%) are seen when Mar 01 12:30:03 its used with only a slight drop in file compression ratio. Mar 01 12:32:25 with a sysfs entry we could use dbus to tune filesystem params via hal Mar 01 12:32:32 cool, but scary Mar 01 12:35:51 morning all Mar 01 12:36:03 hey do13_ Mar 01 12:36:34 hi Koen Mar 01 12:37:28 koen, hrw: Sadly, the zaurus doesn't speed up boot time as much since the zaurus flash controller is brain dead :/ Mar 01 12:37:47 I should look at making is schedule whilst waiting on flash erase... Mar 01 12:37:58 hi Richard Mar 01 12:38:10 XorA|gone: the surface-cache patch was reverted in cairo due to thread problems, so we might need to rediff it for the 1.4 release Mar 01 12:38:22 hi Dirk Mar 01 12:38:26 morning all in fact! :) Mar 01 12:39:26 RP: the cairo 1.16.0 changelog looks like they optimized it for time-test :) Mar 01 12:39:45 "Improved performance when rendering layouts multiple times" Mar 01 12:40:29 time-text that is Mar 01 12:41:06 * chouimat|away is away: day long meeting *sigh* Mar 01 12:51:47 koen: Have you seen the Phreedom? Mar 01 12:52:18 no Mar 01 12:52:20 http://www.gupp.com/gdm878 Mar 01 12:52:21 * koen googles Mar 01 12:54:13 cute Mar 01 12:54:17 altough a bit bricky Mar 01 12:54:52 It lacks BT Mar 01 13:00:31 cu Mar 01 13:15:19 03rpurdie 07org.oe.dev * r2111301d... 10/ (3 files in 2 dirs): mtd-utils: Drop old versions Mar 01 13:15:27 03rpurdie 07org.oe.dev * r42f3520a... 10/ (4 files in 2 dirs): mtd-utils: Drop old versions Mar 01 13:15:32 03rpurdie 07org.oe.dev * r7c8d06d8... 10/ (12 files in 3 dirs): mtd-utils: Add 1.0.0 and 1.0.0+git, include lzo patches in git version Mar 01 13:15:36 03rpurdie 07org.oe.dev * r8c4112d4... 10/ (1 conf/bitbake.conf): bitbake.conf: Disable lzo compression for jffs2 as default (until kernels get updated) Mar 01 13:15:40 03rpurdie 07org.oe.dev * r00efa445... 10/ (1 packages/linux/linux-rp_2.6.20.bb): linux-rp-2.6.20: Sync with changes in poky including lzo support on jffs2 Mar 01 13:25:49 has anyone worked with OE and rt2570 module ? The bb recipe allows me to build the module nicely, however I can't get it to work... I think I'm mainly missing udev rules Mar 01 14:20:47 morning Mar 01 14:21:01 vlo: I plan to buy such one Mar 01 14:21:59 hrw: you mean the usb dongle ? Mar 01 14:22:08 yes Mar 01 14:22:15 vlo: ralink has too many drivers Mar 01 14:22:18 hrw: do you know the status of rm_work? Mar 01 14:23:01 koen: it 'works' and have some problems with do_deploy recipes Mar 01 14:23:12 something more I missed? Mar 01 14:23:33 * koen dunnos Mar 01 14:23:45 I'm thinking of making it default in angstrom Mar 01 14:23:48 hrw: well I'm trying to make it work on my mx31ads Mar 01 14:23:51 once the deploy thing gets fixed Mar 01 14:24:04 koen: bitbake addtask is limited Mar 01 14:32:48 Koen : I talked with trolltech Mar 01 14:33:14 *drumroll* Mar 01 14:33:53 Koen: Knut need to discuss this still with Lorn Mar 01 14:34:12 But we will get a certain ammount of money for raising our foundation Mar 01 14:34:15 the conferences or greenphones? Mar 01 14:34:25 ah, something else :) Mar 01 14:34:28 Koen: and we will get sponsored two developer meetings Mar 01 14:34:51 Koen: Is that okay ? Mar 01 14:35:04 I guess so Mar 01 14:35:19 but I'm a bit oblivious to the foundation thing Mar 01 14:35:25 Koen: I will set up a mail to the core team Mar 01 14:35:41 Koen: Foundation means that we can more easily get money + hardware Mar 01 14:35:59 I know Mar 01 14:36:07 Koen: And we need more security in domain names / servers and stuff Mar 01 14:36:15 but I don't want to know anything beyond that Mar 01 14:36:55 Koen: What do you mean by that ? Mar 01 14:38:45 that details about the foundation enter one ear and go out the other Mar 01 14:41:45 and probably koen want to avoid 'openembedded is founded by trolltech' Mar 01 14:45:09 I hadn't even thought of that Mar 01 14:52:35 Koen: I think that is something that will go into a sponsorcontract Mar 01 14:54:53 heh Mar 01 14:54:55 "I may give up actually coding in the future and just ask for things to be done on my blog" Mar 01 14:55:12 Ross might be right there :) Mar 01 14:56:00 :) Mar 01 14:57:12 Koen: I also received some weird package Mar 01 14:57:27 Koen: It says openembedded developers Mar 01 14:57:57 03dirk 07org.oe.dev * r18c7b509... 10/ (4 files in 3 dirs): Mar 01 14:57:57 packages/gcc gcc-4.1.1: fix eabi issues on a bigendian toolchain: Mar 01 14:57:57 set SUBTARGET_EXTRA_LINK_SPEC to -m armelfb_linux_eabi in case of Mar 01 14:57:57 bigendian Mar 01 14:58:04 and it contains horse manure and was sent from "team pdaxrom"? Mar 01 15:02:27 hmm Mar 01 15:03:25 * koen goes out to pick up girlfriend from the station Mar 01 15:03:26 why does OE all of a sudden expects to find tarballs in tmp/deploy/sources? I have set (and not changed) DL_DIR to some other dir (actually containing the sources) Mar 01 15:04:05 oxo: ? thats default DL_DIR Mar 01 15:05:49 it's very strange, because fetch does not fetch, because the to-be-fetches is in my DL_DIR, but unpack expects it in the default DL_DIR Mar 01 15:19:55 strange Mar 01 15:21:03 even with -v and -DDDDDD I see no reason for this misbehaviour Mar 01 15:21:55 aaaaaaah I see Mar 01 15:22:10 I had the 'distribute local source'-thing enabled some build ago Mar 01 15:22:25 and that appears to thrash my stashed sources by replacing them with symlinks Mar 01 15:23:06 the links point to tmp/deploy/sources/ Mar 01 16:05:45 Hmm Mar 01 16:05:52 I can't build ipkg-native for some reason Mar 01 16:06:06 It tries to run make install, and fails with no such target. Mar 01 16:08:29 * NAiL has a nice idea for the bugday: http://blog.zugschlus.de/uploads/bug-feature.jpg Mar 01 16:20:50 nevermind the ipkg-stuff. I removed the sources and rebuilt. Works now. Mar 01 16:39:52 http://bugs.openembedded.org/show_bug.cgi?id=1924, qemux86 target angstrom-gpe-image doesn not build. Mar 01 16:41:48 how it fails? Mar 01 16:43:10 hrw: checking for scanf size_t conversion modifier... configure: error: cannot run test program while cross compiling Mar 01 16:43:23 look at bug for more details Mar 01 16:43:32 machine is qemux86 Mar 01 16:44:06 target is angstrom-gpe-image Mar 01 16:45:50 slapin_nb: http://pastebin.ca/377299 - apply this and retest Mar 01 16:46:38 hmm. no it is not that rather Mar 01 16:48:47 interesting: http://lwn.net/Articles/222773/ Mar 01 16:51:38 03svolpe 07org.oe.dev * r69fb9699... 10/ (3 files in 2 dirs): Linux-input: Updated patch and serio.h to work with latest linux-input version Mar 01 17:16:37 do bitbake has functionality of sending email about successfull builds? Mar 01 17:17:05 no Mar 01 17:17:22 like tinderbox did? Mar 01 17:17:23 slapin_nb: bitbake something && mail me@my.server? Mar 01 17:18:47 as tinderbox, but I'd like to have direct locations of generated files in there, so I ask about bitbake Mar 01 17:19:07 which could probably supply locations of generated images Mar 01 17:20:24 koen: do you remember who maintain gstreamer in OE? Mar 01 17:20:37 Hello Mar 01 17:20:39 sirfred: hi - still no package here Mar 01 17:20:51 hrw: in practice that probably boils down to me Mar 01 17:20:51 hrw: ok. Thanks. Mar 01 17:21:01 hrw: tmbinc maintains it in .dreambox Mar 01 17:22:21 koen: looks like I have to checkout drreambox branch Mar 01 17:22:50 btw, any chance 1924 to be fixed? Mar 01 17:22:56 !oebug 1924 Mar 01 17:22:58 * * Bug 1924, Status: NEW, Created: 2007-03-01 09:17 Mar 01 17:22:59 * * slapinid(AT)gmail.com: qemux86 build failed Mar 01 17:23:00 * * http://bugs.openembedded.org/show_bug.cgi?id=1924 Mar 01 17:23:05 good morning. has anybody tried mtd-utils_1.0.0 ??? Mar 01 17:27:00 HopsNBarley: I just sent an email to the ML; there may be a missing dependency; it can't find a header file. Mar 01 17:27:14 hrw, I attached log as URL, because it is 3 megs of size. Mar 01 17:27:19 for mtd-utils? Mar 01 17:28:49 HopsNBarley, no Mar 01 17:28:59 HopsNBarley, for bug 1924 Mar 01 17:34:26 Talking about bugs, I notified some time ago 1869, but got no feedback about it. Any problem with it? Mar 01 17:37:31 Is there a target for exporting the basic toolchain Mar 01 17:38:21 task-sdk? Mar 01 17:38:30 !oebug 1869 Mar 01 17:38:32 * * Bug 1869, Status: NEW, Created: 2007-02-13 11:48 Mar 01 17:38:33 * * manuel.teira(AT)telefonica.net: New package: GPE frontend for the snes9x emulator Mar 01 17:38:34 * * http://bugs.openembedded.org/show_bug.cgi?id=1869 Mar 01 17:39:42 hrw: I didn't found a frontend for snes9x and ported an ancient one for gnome 1.x Mar 01 17:40:47 sirfred: not today Mar 01 17:41:08 hrw:sounds good, but what does that create exactly? Mar 01 17:41:17 hrw: What do you mean? Mar 01 17:41:32 sirfred: that I cant look at it today Mar 01 17:41:48 nathan: iirc tarball with sdk Mar 01 17:41:49 hrw: Ah, np Mar 01 17:41:56 cu Mar 01 17:42:05 hrw|gone: Just curiosity about it. I would like to eventually see it added to oe Mar 01 17:42:05 hrw: maintaining == "increasing the version number every now and then" Mar 01 17:42:14 in the temp directory somewhere? Mar 01 17:42:16 hrw|gone: we need a new naming scheme for gstreamer Mar 01 17:42:28 hrw|gone: to coexistence 0.8 and 0.10 Mar 01 17:43:48 tmbinc: gstreamer0.8/0.10 and same for plugins: gst-plugin0.8-NAME gst-plugin0.10-NAME Mar 01 17:43:51 bye now Mar 01 17:47:14 doesn't look like task-sdk actually does anything Mar 01 17:47:23 anyone else have any ideas? Mar 01 17:47:45 ph5, ping Mar 01 17:49:01 HopsNBarley: padaplong Mar 01 17:49:24 hrw|gone: iirc we don't have gst 0.8 in OE anymore Mar 01 17:49:42 hi - i'm tracking down a very strange behavior in mkfs.jffs2, and was wondering if I could bounce some things off you. Mar 01 17:51:28 what is the exclusion patch for? is it used anywhere? Mar 01 17:54:18 HopsNBarley: it's used to exclude certain directories when building the file system, no idea where it is used. Mar 01 17:55:30 ph5, could you look at this thread and share any thoughts? http://www.dream-multimedia-tv.de/board/thread.php?threadid=2877&sid=6420c7a5e9ed29d99e1a56931902e38d Mar 01 18:00:22 wow Mar 01 18:00:46 4GB sandisk CF for ~€60 euro Mar 01 18:00:49 the 2.6.20 kernel has an -rp- in it, does that mean RP is the one in charge of it? Mar 01 18:04:37 HopsNBarley: no idea, from the log it looks like a bug in mtd-utils. Mar 01 18:05:15 or the filesystem going wonky Mar 01 18:05:35 HopsNBarley: "It's a strange one. I can run the exact same commands by hand, and then it works. Just have to keep looking, I guess." <-- this could indicate that it works with the system-wide installation of mtd-tools, but the OE one is broken somehow. do you see the same? Mar 01 18:06:19 koen: I thought CF cards were going to go cheaper. They used to be cheaper than SD, but how much does a 4GB SD card cost? Do they exist? Mar 01 18:06:25 i did a full fsck on two different kernels. okay. i run the cmd line "stat" utility, and the file looks good. Mar 01 18:06:41 sirfred: my cameras can only use CF Mar 01 18:06:41 koen: that might be it, too. but on this board it looks like another person had the same problem. Mar 01 18:07:19 koen: They are faster than SD, aren't they? Mar 01 18:07:24 HopsNBarley: btw, what distro are you running? Mar 01 18:07:37 My build isn't that far from yours, and it worked on my end Mar 01 18:08:03 pH5, same behaviour, i can run the run.image_cmd or the mkfs.jffs2 by hand and it works. so it has to be something in the environment, ,right? Mar 01 18:08:46 my host distro is gentoo, target is openprotium. Mar 01 18:10:12 it's the same file for mmp (the post) and me, btw. so that's some kind of clue as well. Mar 01 18:11:15 HopsNBarley: what if you run ${STAGING_BINDIR}/mkfs.jffs2 manually? should fail if it is indeed a bug in OE's mtd-utils.. Mar 01 18:11:54 that works. Mar 01 18:12:37 it only fails if bitbake runs it. i can run the enclosing script (run.image_cmd) or mkfs.jffs2 with identical args and it works. Mar 01 18:14:07 increasing my stack size didn't help. Mar 01 18:37:21 hrw|gone, I have a 2.6.20 kernel built for the OSK, still need to double check the kconfig Mar 01 18:39:07 crofton can you build an opie image for osk and test if ig segfaults with ipc errors? Mar 01 18:43:19 woglinde: hehe Mar 01 18:43:26 he zecke Mar 01 18:43:41 I thought you sleep now Mar 01 18:47:17 woglinde, not easily Mar 01 18:47:54 woglinde: I just ordered some pizza Mar 01 18:51:24 zecke: hi, can you please look over the code a second? I want to hand it in to rmk but I fear that he will refuce it - bugrep 1920 Mar 01 18:51:56 mr_nice the 2.6.20 runs okay Mar 01 18:52:00 on the simpad Mar 01 18:52:10 I even have a picture Mar 01 18:52:22 woglinde: nice to hear. does the simpad wake up form suspend? Mar 01 18:52:27 but ipc is somehow broken Mar 01 18:52:36 so opie cant runt Mar 01 18:52:42 sorry didnt try Mar 01 18:52:48 * ljp blinks and tries to wake up Mar 01 18:52:54 hm and network over pcmcia is slow Mar 01 18:53:08 mr_nice: I think, rmk is the wrong person for this patch Mar 01 18:53:14 woglinde: hm. Mar 01 18:53:32 zecke: why? he is the one for the arm stuff? who do you think is the right one? Mar 01 18:53:34 mr_nice: you need to find the guy who declares himself maintainer of the mcp stuff Mar 01 18:53:44 but dont know if it is simpad depended or ipc is broken in that way on all platforms Mar 01 18:54:06 zecke: hehe, thats rmk Mar 01 18:54:32 mr_nice: oh I thought that Pavel was the person Mar 01 18:54:40 woglinde: i didn't try opie but it would be nice if it will run on the simpad Mar 01 18:54:51 zecke: pavel changed the ts driver Mar 01 18:55:24 zecke: also i heard that rmks tree is merged by linus or something so it is the fastest way into mainline Mar 01 18:56:53 mr_nice: right, but just because it is used on your simpad doesn't make it appropriate for rmk Mar 01 18:56:57 woglinde: I need to do a new build. I think I will create tomorrow one with your configs Mar 01 18:57:16 mr_nice have you seen my changes? Mar 01 18:57:17 mr_nice: but if he is the maintainer, either send it to the arm kernel list for review, or directly go to the patch tracker Mar 01 18:57:20 to the kernel config? Mar 01 18:57:42 woglinde: yes :) Mar 01 18:57:49 okay Mar 01 18:58:01 zecke: good idea Mar 01 18:58:07 maybee i should koen send a key woglinde1@openembedded.org Mar 01 18:58:26 or search harder for the password Mar 01 18:58:29 woglinde: I am looking forward to try your config tomorrow Mar 01 18:58:51 koen sorry I tried my best Mar 01 18:58:55 mr_nice: linux-2.6.20/drivers/mfd/ucb1x00-switches.c~ Mar 01 18:59:05 mr_nice: use quilt, stgit for patch management Mar 01 18:59:25 mr_nice: do not post temporary, backup files Mar 01 19:00:02 zecke: oh f**k sorry that I didn't saw - thx Mar 01 19:00:34 ;) Mar 01 19:00:49 zecke: I will change it tomorrow Mar 01 19:02:07 hi all! Mar 01 19:02:23 hi gremlin Mar 01 19:02:29 mr_nice: use quilt or stgit for patch management, such stuff can't happen this way Mar 01 19:02:29 woglinde: If you find some time can you please test if your simpad wakes up from suspend? Mar 01 19:03:01 zecke: thx for hinting. Mar 01 19:03:39 mr_nice yes Mar 01 19:04:36 woglinde: thx - I hope yours will wake up Mar 01 19:05:46 * chouimat is back. Mar 01 19:06:01 mr_nice: when posting the patch to rmk, feel free to add Signed-off-by: Holger Freyther Mar 01 19:06:53 zecke: hehe, thx I will do so! Mar 01 19:10:03 zecke: it is very nice to have someone who hinting one on mistakes and to get some suggestions :) Mar 01 19:34:41 back from lunch, bought some shoes Mar 01 19:34:48 back to it Mar 01 19:35:14 looks like I can just pull the toolchain out of the distro, anyone know an automated way to do this? Mar 01 19:36:53 greetings all, I'm working through trying to build my project in oe for the first time, and realizing I have very little understanding of how oe works, my build requires boost, so I added DEPENDS = "boost-jam-native" to my .bb file, but the build fails with lots of "error: 'boost' has not been declared" type errors Mar 01 19:37:18 clearly I'm missing something Mar 01 19:38:07 good night Mar 01 19:38:08 maybe just depend on boost? Mar 01 19:38:33 I am not sure if the boost bb file is split any finer than that Mar 01 19:40:38 Crofton__: well I can see that all the dependencies for boost have been compiled and exist under tmp/staging/arm-linux/include/boost/, so everything I need is definitely there, I just don't know how to get the build process to see it during the compile Mar 01 19:41:14 pastebin the errors? Mar 01 19:46:00 zecke: http://blogs.gnome.org/view/ovitters/2007/03/01/0 Mar 01 19:51:19 fluff Mar 01 19:51:38 zecke: your scanelff thingy needs better path escaping Mar 01 19:52:05 sh: -c: line 0: `/data/build/koen/OE/build/tmp/angstrom/staging/i686-linux/bin/scanelf -Byr /data/build/koen/OE/build/tmp/angstrom/work/armv5te-angstrom-linux-gnueabi/monotone-6-0.33-r0/install/monotone-6-testsuite/home/monotone/monotone-6/tests/tests/merging_(add_a)_with_(add_a,_drop_a)/__driver__.lua' Mar 01 19:52:19 sh: -c: line 0: syntax error near unexpected token `(' Mar 01 19:52:23 hehe Mar 01 19:52:41 you are right, os.system is a security nightmare Mar 01 19:53:57 yeah the FAT patent was voided in german Mar 01 19:53:58 y Mar 01 19:54:03 subprocess.Popen() is our friend Mar 01 19:55:06 is still "need" to cutoff 'insane' from angstrom INHERIT ? Mar 01 19:55:25 i have problem compiling mtd-utils-native-1.0.0+git-r1 Mar 01 19:55:49 zecke: ERROR: Architecture did not match (40 to 3) on /work/armv5te-angstrom-linux-gnueabi/monotone-6-0.33-r0/install/monotone-6-testsuite/home/monotone/monotone-6/tests/tests/diff_a_binary_file/binary Mar 01 19:55:59 zecke: in this case, that's a red herring Mar 01 19:57:50 koen: yeah, but where to look and where not to look? Mar 01 19:58:04 koen: I want a whitelist of known errors but I do not know where to store it Mar 01 19:58:36 Packaged contents of monotone-6 into /data/build/koen/OE/build/tmp/angstrom/deploy/glibc/ipk/armv5te/monotone-6_0.33-r0_armv5te.ipk Mar 01 20:02:30 mtn: adding packages/gcc/gcc-4.1.1/801-arm-bigendian-eabi.patch Mar 01 20:02:32 scary Mar 01 20:03:10 [20:49] CIA-7: openmoko: laforge * r1205 /trunk/oe/packages/tslib/tslib-1.0/fic-gta01/tslib.sh: it's now 'event1' since 'event0' was taken by the gta01kbd driver for 911 key Mar 01 20:03:19 don't hardcode! Mar 01 20:03:44 it shows that a former opie dev is on their team :p Mar 01 20:05:13 koen: laforge of team opie? Mar 01 20:05:20 I wonder why we didn't sue GPE... Mar 01 20:06:11 OT: anyone wrote a kexec like module? Mar 01 20:11:10 hi, i'm trying to compile freeciv for zaurus and bitbake fails, can anybody help me with it? Mar 01 20:11:32 bitbreak? Mar 01 20:11:35 zecke: say, were you the one playing with the automated gcov stuff for kde? Mar 01 20:14:48 some taiwanese wrote me to access a brightness control on a pci device like quote"... from Bus0 Device2 Function0 Register F4h. The range is between 0FFh and 52h." /quote -- any ideas ? Mar 01 20:15:21 simple outb 0xf4 is kinda not working ;( Mar 01 20:15:32 njs: yes Mar 01 20:16:01 | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / Revision: b8eb30a0fd93f61ff386bac7e2203fb3c883e888 Mar 01 20:16:06 njs: I'm currently waiting for the EBN to do regular KDE builds Mar 01 20:16:12 * koen stabs asciik + mtn log Mar 01 20:29:08 koen: hmm, that's parallel development you have going on there. Mar 01 20:29:22 (unless you're hitting a bug in restricted log, the graph code is not really working right for that case yet) Mar 01 20:29:50 zecke: what's the status of that code? We were looking at fixing ours up again, since our old setup bitrotteda way... Mar 01 20:31:57 over time my jffs system appears to be corroding Mar 01 20:32:11 it still works, but I get more and more Mar 01 20:32:14 jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x00207ec0: 0x005a instead Mar 01 20:32:21 messages during bootup Mar 01 20:32:25 anyone seen this before? Mar 01 20:32:51 njs: http://websvn.kde.org/trunk/KDE/kdesdk/ebn/coverage/ but it is underdocumentated and has one issue Mar 01 20:33:25 njs: it tries hard to find the right file but it can still fail to resolve 'un-namespaced.h' and provide wrong results Mar 01 20:33:54 njs: back then you mentioned another gcov parameter which is not yet incorporated Mar 01 20:37:13 ljp: did you take a look at how to open the greenphone? Mar 01 20:37:38 ljp: it looks like yuang is using STUART as serial console, did you find the pads for it yet? Mar 01 20:47:30 serial access is only through a breakout box Mar 01 20:47:36 Zero_Chaos: yes ;-) Mar 01 20:47:46 if you run bitbake filename -cconfigure does it also do the configure_append() code? Mar 01 20:47:54 zecke: I passed the message onto ross - hope you gws Mar 01 20:48:01 ljp: this breakout box is doing TTL->CMOS Mar 01 20:48:36 zecke: TTL->RS232 surely? Mar 01 20:48:43 RP: probably :) Mar 01 20:49:00 settles to "changing level" Mar 01 20:49:08 I'd settle for that :) Mar 01 20:49:24 You can get nice chips from Maxim that do it like the infamos MAX232 :) Mar 01 20:50:18 RP: I know, and since my hands stopped shaking I can even build the voltage pump myself Mar 01 20:50:31 RP: it is more the question, if someone opened the case safely Mar 01 20:50:45 ljp: which connector do you use for the breakput box? can you send me one? Mar 01 20:50:52 It did look very small and easy to break... Mar 01 20:51:08 its a special connector/battery Mar 01 20:51:20 ljp: send me one :) Mar 01 20:51:50 we only have two here. and yuhuatel weren't very to send us even one Mar 01 20:51:59 very keen, that is Mar 01 20:52:34 ljp: then get me the pinout on the battery connector, otherwise I would have to break my lovely greenphone to search for the pad :( Mar 01 20:55:07 ljp: http://page.mi.fu-berlin.de/~freyther/greenphone/ first set of 2.6 patches Mar 01 20:55:15 ljp: didn't try to boot the kernel though Mar 01 20:55:25 ljp: do you know if the greenphone has the flip sensor builtin? Mar 01 20:55:46 flip sensor? Mar 01 20:56:38 ljp: there are three GPIO names, FLIP_SENSOR1, FLIP_SENSOR2, FLIP_SENSOR3 I wonder what it might be Mar 01 20:57:10 oh, must be for previous flip phone.. philips Mar 01 20:57:17 maybe Mar 01 20:58:05 ljp: ah so the stuff next to sd-card slot is serial? Mar 01 20:59:05 next to the sd card? Mar 01 20:59:18 koen, booting Angstrom on the OSK has a cat /proc/net/wireless No such file or directory messgae Mar 01 20:59:28 when configuring network interfaces Mar 01 21:00:09 zecke: Nice to see your work. Now I can stroke the greenphone from my needed 2.6 kernel list. :) Mar 01 21:00:33 much work wil lneed to be done for greenphone 2.6 Mar 01 21:00:44 root@omap5912osk:~$ uname -a Mar 01 21:00:44 Linux omap5912osk 2.6.20-omap1 #1 Thu Mar 1 14:40:27 EST 2007 armv5tejl unknown Mar 01 21:00:53 both bootloader and kernel. Mar 01 21:01:02 stefan_schmidt: I'm in bed mostly so I don't know if it boots Mar 01 21:01:13 ljp: Of course. But somebody has to start it. Mar 01 21:01:24 very true Mar 01 21:01:27 ljp: decompiling 16 of 17 binary only modules to get the i2c and GPIO numbers Mar 01 21:01:56 ljp: using the public mdoc G3 documentation to write a nand implementation that is not using this ttfs crap Mar 01 21:02:07 ljp: What bootloader it use? Mar 01 21:02:08 wish I could help you there, we only get binary modules too Mar 01 21:02:22 Crofton: must be some leftover for wifi config Mar 01 21:02:27 it uses a proprietary one, as the kernel is on the mDOC Mar 01 21:02:28 yeah Mar 01 21:02:29 ljp: lol 16/17 of the modules are derived work Mar 01 21:02:41 zecke: wasn't there a patch floating around to add more MDoC support? Mar 01 21:02:51 g3 mdoc Mar 01 21:02:58 anyway, I have Angstrom (mostly working) on the OSK, running with 2.6.20 :) Mar 01 21:02:59 ljp: the bootloader used to be blob (at least it looks they sed blob for omegboot) Mar 01 21:03:07 OMGBOOT!!!! Mar 01 21:03:16 yes at first it was blob with msystems proprietary code in it Mar 01 21:03:17 ljp: http://www.openembedded.org/~zecke/greenphone.jpg Mar 01 21:03:20 (couldn't resist) Mar 01 21:03:27 ljp: are these pads (that are visible) used for serial? Mar 01 21:03:45 ljp: tell them they should not touch cris architecture documentation... Mar 01 21:03:49 greenphone is based on pxa27x? Mar 01 21:04:04 ljp: it is obvious that when cris changed from blob to omegboot that the omega device uses blob Mar 01 21:04:22 ljp: ideally we would use the NAND badblock stuff from the Neo Mar 01 21:04:34 wow, derived kernel modules and derived bootloader. Who needs to be sued? Mar 01 21:05:02 * stefan_schmidt sometimes think companies never get it. :( Mar 01 21:05:07 stefan_schmidt: Trolltech distributed it to me :) Mar 01 21:05:10 yes, those are serial Mar 01 21:05:18 And who to trolltech? Mar 01 21:05:27 knut and me Mar 01 21:06:03 ljp: then please give me the RX, TX and GND lines Mar 01 21:06:13 you'd still need someome with copyright to sue iirc Mar 01 21:06:15 you need to use the new bootloader to have a gpl compliant bootloader and kerne; Mar 01 21:06:39 ljp: alternatively send me the source Mar 01 21:07:09 koen: That should not be a problem. Mar 01 21:07:10 oh at least mini-SD and SD are protocol compatible Mar 01 21:07:31 zecke: yeah, microsd also. Just other form factor Mar 01 21:07:32 we dont have sources. Mar 01 21:07:44 maybe greenphone wasnt a good idea Mar 01 21:07:50 ljp: where to file Greenphone bugs? The network settings app seems to be locked to DHCP (regardles of what you cange in the gui) Mar 01 21:07:52 koen: But sueing is the last step Mar 01 21:08:10 bombing, hostages are more familiar concepts Mar 01 21:08:11 sue fo rwhat? Mar 01 21:08:21 Motorola needs more then three months to deliver me the modified blob code. Mar 01 21:09:38 ljp: Sueing to be GPL compliant. Of course only if it is derived work and nobody like to offer the source. Mar 01 21:10:07 I wonder how many Off-By-One GPIOs I have. They use GPSR0,1,2 directly... Mar 01 21:10:14 ljp: As you of course no this stuff mostly finished without court. Mar 01 21:10:43 ljp: okay, can you please give me the pinout of this pad? Mar 01 21:10:53 ljp: otherwise I will try it and ask for warranty Mar 01 21:10:55 it IS GPL complient, if you flash the new bootloader Mar 01 21:11:10 zecke: ya Mar 01 21:11:21 ljp: which bootloader do you use nowadays? Mar 01 21:11:27 dunnno Mar 01 21:11:36 ljp: Fine Mar 01 21:11:38 the one they gave us Mar 01 21:12:08 setphan: we stopped shipping and orders when we found out the violation. Mar 01 21:12:41 koen: I have only found David's mail that yaffs/jffs2 support should be possible with the available documentation Mar 01 21:12:52 zecke: whats your email addy these days? Mar 01 21:13:02 ljp: So that was the reason for shipping stop. And now they use another, proprietary, bootloader? Mar 01 21:13:10 koen: I wonder how much stuff is needed for cramfs (the translation level) Mar 01 21:13:18 stephan: yes Mar 01 21:13:18 ljp: zecke@selfish.org Mar 01 21:13:30 i mean stefan, sorry Mar 01 21:13:33 ljp: ok, that clears it. Mar 01 21:14:26 zecke: You planning a complete 2.6 port or only some first hacking for fun? Mar 01 21:16:29 stefan_schmidt: I tend to do things until the end... I at least want to have lcd, mmc and modem working Mar 01 21:16:32 zecke: AHHHH...greenphone porn :) Mar 01 21:16:36 it wont boot without new bootloader. there is only 1 mb for kernle partition, and partition table is in bootloader Mar 01 21:17:02 kexec kernel + initramfs? Mar 01 21:17:03 ljp: I will write a kexec-like kernel module Mar 01 21:18:25 zecke: Good to hear. /me has enough to do with EZX phones anyway. Also there is a brand new HTC on the way to me, too. Mar 01 21:18:28 the mdoc g3 uses tffs version 632.70, as well Mar 01 21:18:44 stefan_schmidt: mmc will be easy once I know how to enable power for mmc Mar 01 21:19:08 zecke: fiddling register :) Mar 01 21:19:09 stefan_schmidt: lcd is configured using i2c, and these guys bitbang themselves on the PXA registers Mar 01 21:19:24 stefan_schmidt: and the module is binary only :) Mar 01 21:19:48 stefan_schmidt: and they have a huge module for the modem, at least I know how to power on the device... Mar 01 21:19:51 * stefan_schmidt thanks god that moto code is crap but not binary. Mar 01 21:19:56 who mentioned i2c lcds a while ago? Mar 01 21:20:05 koen: I did yesterday Mar 01 21:20:16 besides you Mar 01 21:20:31 koen: I have never seen one, and would like to see an example Mar 01 21:20:37 zecke: If you have luck some parts works out of the box with recent 2.6 after you found out how to power. Mar 01 21:20:53 stefan_schmidt: yes, mmc should just work Mar 01 21:21:04 stefan_schmidt: at least I can read the pxafb configuration once the device is on Mar 01 21:21:08 zecke: I think it was prplague Mar 01 21:21:23 zecke: right, mmc should be easy. Perhaps even some more stuff. Mar 01 21:21:30 prpplague even Mar 01 21:21:52 zecke: You could take a look at the chip part number if they are not shielded. Mar 01 21:21:57 stefan_schmidt: modem is on FTUART/HWUARt Mar 01 21:22:15 zecke: ah, that's nice, too. Mar 01 21:22:21 zecke: hwuart is a virtual uart Mar 01 21:22:37 stefan_schmidt: the LCD number is in the name of the kernel module and google finds one word document Mar 01 21:22:40 it just enable flow control iirc Mar 01 21:22:49 koen: yes, FTUART is real :) Mar 01 21:23:25 zecke: if serial coms is getting unreliable you could opt to use hwuart Mar 01 21:23:27 ljp: why aren't you using bluetooth? the device itself works? Mar 01 21:23:42 zecke: Do they use multiplexing over the serial or just one channel? Mar 01 21:23:47 we do use bt in 4.2.1 Mar 01 21:24:08 it took a backport of some bt kernel stuff Mar 01 21:24:23 stefan_schmidt: over serial, but this part is TT proprietary code Mar 01 21:24:35 stefan_schmidt: so if there are some special commands, strings will tell us Mar 01 21:24:57 zecke: ok, once you can talk to the modem. AT+CMUX=? tells you more Mar 01 21:25:28 hmm Mar 01 21:25:34 * oxo has made a discovery Mar 01 21:25:48 glibc (any version) on sparc does not like (repeat me) "-Os" Mar 01 21:25:54 zecke: If it turns off after 60s with your own kernel and rootfs, try AT+USBSTAT=255. Mar 01 21:25:55 okay off to bed again, my throat is still killing me Mar 01 21:25:58 not TT proprietary.. yuhuatel proprietary Mar 01 21:26:15 ljp: GSM userspace code is from TT Mar 01 21:26:15 we are in the same boat on alot of these issues Mar 01 21:26:21 zecke: night and getting well soon. Mar 01 21:26:25 oxo: hmmm Mar 01 21:26:29 nn zecke Mar 01 21:26:30 ahh. hmm.. Mar 01 21:26:39 ljp: what is Robert doing ATM? Mar 01 21:26:57 actually, it is easy to see modem commands, enable the qlogging for it Mar 01 21:27:17 ljp: BTW: Eirik's mail was in my spam filter, I didn't really know what to reply though Mar 01 21:27:26 ahh. ok Mar 01 21:28:08 ljp: so thanks for the hint. Instead of repeating the past I just asked: For better communication, accepting patches, actively talking with the community instead of creating a Qtopia Island Mar 01 21:28:15 * stefan_schmidt moves over to openezx to get a clue about voice call routing on ezx. Mar 01 21:29:00 ljp: e.g. if Robert has some time I would like to fly to Munich/Oslo to do some kernel hacking with him Mar 01 21:29:28 oxo: what symptons should I look for to mark -Os as the culprit? Mar 01 21:29:42 not sure what he is doing Mar 01 21:29:54 PSO travels a lot Mar 01 21:30:16 I think Matthias tricked him badly Mar 01 21:30:27 possibly :) Mar 01 21:30:32 "So Robert you was not able to move to Oslo, we will create a Munich office" Mar 01 21:30:44 and then you force him to travel to Oslo like all the time, his poor wife Mar 01 21:30:47 ... Mar 01 21:31:24 koen: glibc-intermediate fails complaining about ___ieee7something_sqrt being an undefined symbol Mar 01 21:31:32 oxo: FULL_OPTIMIZATION_pn-glibc_sparc = "-fexpensive-optimizations -fomit-frame-pointer -frename-registers" Mar 01 21:31:55 yeah, but I'm not really sure yet this is a libgcc or glibc issue Mar 01 21:31:59 and BUILD_OPTIMIZATION_pn-glibc_sparc = "-O2" Mar 01 21:32:13 libgcc provides that symbol, glibc uses it Mar 01 21:32:14 03koen 07org.oe.dev * r224b10a5... 10/ (1 packages/monotone/monotone-6_0.33.bb): monotone: add 0.33, with a slight hack to silence scanelf Mar 01 21:32:35 Os vanishes it "somewhere", but I don't know yet where Mar 01 21:32:46 I'll retest with your options Mar 01 21:33:15 I tried doing bitbake linux-rp-2.6.10. Things were going well until it stopped while compiling glibc. The file that had a problem is /libc/ports Mar 01 21:34:15 sorry, hit enter in teh middle..anywyas, the ioperm.c file complains at this line #if LINUX_VERSION_CODE >= KERNEL_VERSION(2,4,23) Mar 01 21:34:24 saying operator >= has no left side Mar 01 21:34:41 cyrus__: any reason to use such an old kernel? Mar 01 21:35:05 zecke - I am trying to compile for my ipaq and the current release out there uses that kernel Mar 01 21:35:16 just wanted to stay consistent since I am just starting up Mar 01 21:35:21 koen: should I change that in glibc-superspace since I'm building for v8? Mar 01 21:35:25 my first goal is to actually compile the kernel Mar 01 21:35:46 oxo: put that in local.conf for now Mar 01 21:35:49 * woglinde wonders why kernel compile need glubc Mar 01 21:36:11 woglinde: it actually doesn't, but OE thinks every app needs glibc Mar 01 21:36:21 koen ;) Mar 01 21:36:34 and koen: does the fact that glibc-intermediate fails (instead of glibc) change the line? Mar 01 21:36:48 ehmm Mar 01 21:36:49 yes Mar 01 21:36:53 good thinking :) Mar 01 21:37:00 I'll put -intermediate in it Mar 01 21:37:24 so pn-glibc-intermediate-superspace it is Mar 01 21:37:32 also, since I did bitbake linux-rp-2.6.10 and it failed, if I fix the compliation problem how doI continue from where it left off. If I run the same bitbake command again, will it continue or start all over Mar 01 21:38:25 well, and off to bed Mar 01 21:38:32 I'll see tomorrow if it works Mar 01 21:38:36 nn everyone! Mar 01 21:38:41 nn oxo Mar 01 21:39:54 anyone? Mar 01 21:40:32 cyrus it will not start from the beginning Mar 01 21:40:44 so just run the same bitbake command again..okay..thanks Mar 01 21:45:52 okay, finally off to bed Mar 01 21:46:01 yes Mar 01 21:46:03 me too Mar 01 21:46:20 ljp: I think you should ask for the source of 16 of the 17 modules, and thanks for any help you can give me Mar 01 21:46:24 g'night all Mar 01 21:47:51 Am I still online? Mar 01 21:48:13 I guess so :> Mar 01 21:48:15 no Mar 01 21:48:16 aw Mar 01 21:48:18 weird Mar 01 21:48:18 :P Mar 01 21:48:20 :> Mar 01 21:48:25 why the hell does it feel like friday? Mar 01 21:48:29 tomorrow's gonna suck Mar 01 21:48:36 wow! I have exactly the same feeling Mar 01 21:48:41 The wireless auth stuff is driving my browser insane Mar 01 21:48:48 I thought it was just me... bout friday that is Mar 01 21:48:54 kergoth, why? Mar 01 21:49:00 kergoth: friday is only 37 minutes away here..... Mar 01 21:49:04 heh Mar 01 21:49:11 because today feels like friday, so tomorrow will be like saturday, cept i'll have to work :) Mar 01 21:49:14 koen: hehe Mar 01 21:49:15 ~today Mar 01 21:49:17 "This must be Thursday. I could never get the hang of Thursdays." (Arthur Dent) Mar 01 21:49:26 I was going to go on about how good the weather is here, Mar 01 21:49:34 ~tomorrow Mar 01 21:49:35 but tonight there is a flood watch Mar 01 21:49:36 Tomorrow will probably suck too. Mar 01 21:49:38 kergoth: I figured that I will simply not go to work tomorrow so I do not disappoint my todays friday feeling :> Mar 01 21:49:41 hehe Mar 01 21:49:45 and I saw snoflakes in the forecast for the weekend Mar 01 21:49:46 Jin^eLD: i was tempted :) Mar 01 21:49:52 I could not resist :) Mar 01 21:50:03 but i've only worked 3 weeks in this job so far, methinks i'd better not push my luck Mar 01 21:50:17 kergoth: the green hills gig? Mar 01 21:50:57 kergoth: just think of tomorrow as an early incarnation of monday :) Mar 01 21:51:11 ~yesterday Mar 01 21:51:13 Yesterday? it sucked. Mar 01 21:51:17 heh - somebody in #olpc just said the same thing about Friday Mar 01 21:51:29 Somebody is messing with the time space continuum again Mar 01 21:51:51 hmmm, maybe it's that whole DST is coming early this year thingy... Mar 01 21:52:04 we're right, the universe is wrong damnit. Mar 01 21:52:41 maybe it's somehow related to february being so short Mar 01 21:52:49 hvontres|poodle: how was LV? Mar 01 21:53:21 yeah Mar 01 21:53:28 * Crofton hates February Mar 01 21:53:55 february is out of sync Mar 01 21:53:55 * gerwinin as well fucking rain , cold Mar 01 21:54:52 koen: I manged not to fsck my form and I won my first sparing match :) and my 5 year old took home a 1m tall trophy...:) Mar 01 21:55:26 kergoth: Maybe the Universe now works accordig to these guys: http://www.geocentricity.com/ Mar 01 21:58:14 03crofton 07org.oe.dev * r81fc5b57... 10/ (5 files in 3 dirs): linux-omap1_2.6.20 : Add files to build 2.6.20 kernel for the OSK. Mar 01 21:58:31 03crofton 07org.oe.dev * r8a3487b4... 10/ (1 conf/machine/omap5912osk.conf): omap5912osk.conf : Set ROOT_FLASH_SIZE to 29M. Mar 01 21:58:40 03crofton 07org.oe.dev * r0aca530c... 10/ (1 packages/tasks/task-ossie.bb): task-ossie.bb : Add another waveform to the build. Mar 01 21:59:19 Crofton: to '29', right, not '29M', right? Mar 01 21:59:40 ah right Mar 01 22:07:11 One is just a comment :) Mar 01 22:07:43 I had to clean out some lingering changes, before I commited them by accident in a crap changeset :) Mar 01 22:11:19 hi Mar 01 22:19:10 My FOSDEM images are located here: Mar 01 22:19:12 http://www.home.unix-ag.org/boor/img/index.php?path=fosdem/ Mar 01 22:20:02 (all, unsorted and in full size) Mar 01 22:20:25 http://www.home.unix-ag.org/boor/img/fosdem/img_1816.jpg ohh a slug =) Mar 01 22:21:51 CoreDump|home: we have more cool devices in there ;) Mar 01 22:23:04 Right, a Nomad (chuckle) and a sl-c3x00 in this photo =) Mar 01 22:23:22 * florian likes 1844 and 1845 Mar 01 22:23:43 whoever last edited mtd-utils-native-1.0.0+git-r1, it doesn't build Mar 01 22:23:55 florian: lol! Mar 01 22:24:14 Florian: How much was the cash donation we received ? Mar 01 22:25:22 greentux: 60 EUR donation and we have 95 EUR from selling t-shirts. Mar 01 22:25:30 hrm Mar 01 22:25:55 Florian: cool Mar 01 22:25:57 s/greentux/gerwinin/ Mar 01 22:26:03 florian: nils does not know something about thirts btw. Mar 01 22:26:19 florian: I asked him for a shirt, but he couldnt assist :) Mar 01 22:26:41 greentux: ah right, an oe t-shirt? Mar 01 22:26:52 florian: yes Mar 01 22:27:10 florian: I bet the paper-moko gets fantastic battery life :) Mar 01 22:27:17 florian: he wanted to give me a special price for N800, I said: send me a oe shirt instead and donate the money. Mar 01 22:27:58 greentux: i see, heh... we in fact sell quite some t-shirts but the oe ones are koen-made. Mar 01 22:28:18 florian: ahh. its bad :) Mar 01 22:29:05 Florian : Would he be possible to sell the t-shirts online ? Mar 01 22:29:31 greentux: but you remind me that i need to jump out of my bed (i stayed in for several days now) and do some work, i have to look into the syncing thingy. Mar 01 22:29:36 does cafepress work in EU? Mar 01 22:30:19 Crofton : no Mar 01 22:30:38 unfortunate Mar 01 22:30:43 greentux: i hope so, i'm not sure where they are currently located but it might be easy to find out once he is back. Mar 01 22:31:22 Is it possible that quite a lot people who were at FOSDEM became ill afterwards? Mar 01 22:31:39 I feel fine Mar 01 22:32:47 florian: mickeyl was ill Mar 01 22:33:29 greentux: i am ill, like two more who came with us... Mar 01 22:33:29 florian: bad weather, every fosdem has bad weather. bad locations... not enough alc and so on Mar 01 22:33:45 Not enough alcohol? Mar 01 22:34:03 florian: i was it AT fosdem... but has to visit it instead of sleeping in th ehotel Mar 01 22:34:06 greentux: ah no, that can't be the reason... we live in Siegen where we have bad weather all the time ;) Mar 01 22:34:39 Maybe it is like spring break here, everyone goes home and is exposed to new germs, then the come back to school and everyone gets sick Mar 01 22:34:51 * florian too... most of the time at the way too cool OE booth Mar 01 22:35:24 florian: Nice picture from our E680. The cable is sweet, isn't it? :) Mar 01 22:36:26 stefan_schmidt: Yes indeed :-) Mar 01 22:39:42 i think my mysterious broken mkfs.jffs2 is a fakeroot issue. ring a bell with anyone? Mar 01 22:46:23 HopsNBarley: How did it break? I made some changes there today and I could have broken it... Mar 01 22:48:42 RP: i'm running a december-ish version of everything... did you update mkfs or fakeroot? Mar 01 22:49:28 HopsNBarley: Ah, todays change won't affect you then. I updated mtd-utils but only to add lzo support Mar 01 22:49:54 * mwester says "aha!" Mar 01 22:50:16 RP: mkfs.jffs2 is getting a bad result back from lstat, which is wrapped by libfakeroot. Mar 01 22:52:02 HopsNBarley: That sounds like a potential fakeroot problem... Mar 01 22:52:17 indeed. Mar 01 22:52:47 good night Mar 01 22:56:17 Finally, I got a build to complete. I did a bitbake linux-rp-2.6.20 and it worked. Since it said attempted builds: 5 I am assuming it completed fine. Where does it output the final results Mar 01 22:56:23 AchiestDragon: where is the built kernel image Mar 01 23:02:27 CoreDump|home: So since using OE for creating feeds is "illegal", do you have any good ideas on how to set up a local test feed? Mar 01 23:03:24 hvontres|poodle: http://hentges.net/tmp/oe-patches/dev/ Mar 01 23:04:37 CoreDump|home: hehe... that looks like koen's last commit with the -'s replaced with +'s :) Mar 01 23:05:21 exactly Mar 01 23:05:21 CoreDump|home: Do you know what bugs in ipkg-util he was refering to earlier? Mar 01 23:05:42 no clue. The local feed was (and is) working fine Mar 01 23:06:16 03rpurdie 07org.oe.dev * rc5e68e96... 10/ (3 files in 3 dirs): linux-rp-2.6.20: Add hostap monitor patch from Zero_Chaos Mar 01 23:07:49 CoreDump|home: is it possible to redesign it as a separate class as mickeyl suggested? Mar 01 23:07:57 * Zero_Chaos casually wonders how to test the new kernel considering mtd-utils-native is a blocker.... Mar 01 23:08:12 (mickeyl showing the voice of reason as always compared to the other voices ...) Mar 01 23:08:55 is feed creation best handled by OE, or would an external script be better? Mar 01 23:09:02 rwhitby: of course it is but I won't do it. I won't give in to koen's "it's wrong in rootfs_ipk but ok in a bbclass" crap Mar 01 23:09:13 I don't know anything about feed maintenance, but I am close to needing to know Mar 01 23:09:15 Crofton__: it's always worked flawlessly for us in the past, for three different distributions Mar 01 23:09:24 the place where it lives doesn't matter Mar 01 23:09:41 it seems that koen is the only one who had problems, and he seems to think that means that it's broken for "everyone" Mar 01 23:10:03 Crofton__: Well, that seems to be today's hot topic... It seems koen is convinced that OE shouldn't be doing feed maintainance for reasons I don't yet unserstand. Mar 01 23:10:11 indeed. Calling on some abscure policy which is no where to be found Mar 01 23:10:30 who made koen the decision maker of what OE does and does not support anyway? Mar 01 23:10:59 it seems to be a case of "angstrom doesn't need it, so it's not going in" Mar 01 23:11:19 hvontres|poodle: not even doesn't he "want" it. He broke existing solutions and didn't give a damn, then even reverted a compromise on my part. Mar 01 23:11:47 * hvontres|poodle wishes he could find a bug report on what exactly is supposed to be wrong with ipkg-utils Mar 01 23:12:04 03rpurdie 07org.oe.dev * r88cc7e94... 10/ (1 packages/lzo/lzo-native_1.08.bb): Add lzo-native Mar 01 23:12:13 03rpurdie 07org.oe.dev * r87f75e29... 10/ (3 files in 2 dirs): mtd-utils: Fix DEPENDS (add lzo) Mar 01 23:12:20 who are the final arbiters of what goes in OE and what doesn't? cause if it's only koen, that's a real problem. Mar 01 23:12:47 hvontres|poodle: ignore it. ipkg-utils handled very large feeds fine for me Mar 01 23:12:57 rwhitby: I think it's mainly koen, RP, mickeyl and hrw Mar 01 23:13:08 rwhitby, that sounds like a good topic for the next OE meeting Mar 01 23:13:56 * CoreDump|home nods Mar 01 23:13:56 CoreDump|home: well, I am kinda curious as to what the failure mode would be so I can at least look for it and maybe think about how to fix it. Mar 01 23:14:03 rwhitby: Roughly speaking, the core devs. Koen is only one of the core devs of which there are several Mar 01 23:14:11 hvontres|poodle: no idea. never had any problems.... Mar 01 23:14:31 CoreDump|home: sounds like it's time to put it to a vote of the core devs then Mar 01 23:14:54 I think mickeyl already commented on the mailing list... Mar 01 23:15:01 I've been trying to keep out of it ;-) Mar 01 23:15:02 CoreDump|home: plus, by fixing it we could get rid of one of koen;s objections from earlier ( the one about adding broken functionality) Mar 01 23:15:02 cause mickey|zzZZzz didn't seem to be against it. dunno what RP and hrw|gone's position is Mar 01 23:15:07 rwhitby: I have said what I had to say. I no longer care and use my own patches Mar 01 23:15:37 CoreDump|home: could it be done in package-index.bb ? Mar 01 23:15:52 (since that is what slugos uses for feed generation) Mar 01 23:16:49 rwhitby: I think the main problem is more philosophical than technical at this point...:( Mar 01 23:17:17 it's worked technically for me as well Mar 01 23:17:22 CoreDump|home: it's saddening that koen's attitude in dealing with people, and not being able to discuss things rationally and calmly when there are disagreements, causes good developers like you to carry your own patches ... Mar 01 23:18:19 * rwhitby thinks that one of the reasons there are so many DISTROs in OE is cause no-one can be bothered trying to consolidate cause that means they need to argue with koen more. Mar 01 23:18:38 at least that's the reason why slugos will never be merged into angstrom. Mar 01 23:18:55 rwhitby: I find it tiring to argue over this kind of stuff. It is not even _worth_ arguing because it didn't hurt anybody, didn't affect anyones build. Having a private patchset is the way of least resistance Mar 01 23:18:56 Firstly, can we please not turn this into a head hunt Mar 01 23:19:23 RP: you're right. sorry. I should not let my frustration boil over here. Mar 01 23:19:35 Yes, Koen's people skills can be lacking at times. We know this ;-) Mar 01 23:20:00 Apologies to all for my temporarilty being part of the problem instead of the solution :-) Mar 01 23:20:09 =) Mar 01 23:20:33 I see no reason CoreDump|home's patch can't be submitted as a new .bbclass. If angstrom doesn't wish to use it, so be it Mar 01 23:21:04 RP: new .bbclass, or just an update to package-index.bb ? Mar 01 23:21:39 One of OE's strengths should be the ability to handle diverse needs of developers Mar 01 23:22:00 Right, we have plenty of options for imlementing this in such a way that everyone can be happy Mar 01 23:22:28 Getting more than one open source developer to agree on something is hopeless Mar 01 23:22:28 rwhitby: Good question. I don't really want to change the behvaiour of package-index.bb Mar 01 23:22:41 RP: by any chance, do you know what problems with ipkg-util koen was talking about earlier? I would like to see if I can dig into solving things from that end.... Mar 01 23:22:42 RP: I tried that after koen b0rked deploy_ipk. He reverted the (optional and off by default!) cset Mar 01 23:22:44 the number of projects depending on OE is a minor miracle Mar 01 23:23:20 CoreDump|home: Tried what - a separate .bbclass? Mar 01 23:23:40 RP: no, " imlementing this in such a way that everyone can be happy" Mar 01 23:24:05 koen|away: had his deploy_dir nonsense and everyone else could enable a working oe-feed if they choose to Mar 01 23:25:17 hvontres|poodle: Not specifically, no Mar 01 23:26:16 CoreDump|home, rwhitby: Let me talk to koen next time he is around. I want this discussed on the mailing list before any more changes are made/reverted to OE Mar 01 23:27:05 * CoreDump|home is tempted to reply to koens latest BS mail Mar 01 23:27:44 RP: thanks for taking an interest. I have confidence in your rationality and will hold any further responses until you have a chance to moderate. Mar 01 23:28:08 rwhitby: I can't help thinking I could regret this ;-) Mar 01 23:30:06 RP: yeah, that's the problem. All good developers should be able to discuss technical disagreements without having to feel like you do now. Mar 01 23:32:23 off topic... (-; fakeroot is failing one of the tests. can anybody replicate? i' Mar 01 23:32:35 i'm wondering if this is specific to my gentoo box.... Mar 01 23:44:31 I've sent a mail to the list suggesting koen expand on his arguments since currently, it does lack technical content. Mar 01 23:45:10 mwester: I've pushed an lzo-native package - could you see if that solves the problems you were seeing please? Mar 01 23:46:05 should mkfs.jffs2 run under fakeroot? Mar 01 23:47:21 HopsNBarley: As I recall, the whole make rootfs process does Mar 01 23:48:30 RP: Thanks for moderating the feed topic for us :) I sure hope we can come to some kind of consensus here... after all, if we could only commit stuff that always worked, CIA would get very lonely :) Mar 01 23:49:47 HopsNBarley: I build on Gentoo without (many) problems. Mar 01 23:50:17 I have some weird bug where creating a slugos image fails but works when I run it a second time. Mar 01 23:53:23 RP: does it need to? I can see why most of making the root needs fakeroot, but i think mkfs.jffs2 runs fine without it. Mar 01 23:53:47 joshin, it's been working fine for me too, until a few days ago. Mar 01 23:54:06 RP: should have said, it DOES run fine without it. Mar 01 23:54:25 I'm on a fully up to date 'amd64' machine. Mar 01 23:55:24 joshin - i've added a few goodies like strace and gs, but nothing that should've killed OE. Mar 01 23:55:36 thanks for the data point though. Mar 01 23:55:47 No idea. I didn't see the error you're having. Mar 01 23:56:00 HopsNBarley: I use gentoo as well on an amd64 but built 32-bit Mar 01 23:56:55 when mkfs.jffs2 reads tmp/rootfs/etc/ipkg/arch.conf, it sees the wrong struct stat st_mode bits. it thinks it is a directory. Mar 01 23:57:06 another user has seen this error too, same file. Mar 01 23:58:23 HopsNBarley: It sounds like you need to fix fakeroot then... Mar 01 23:58:56 (-; i'm working on it. i wanna do one more test where it is totally distilled down to fakeroot/no fakeroot and see the error. Mar 02 00:03:18 HopsNBarley, I also use Gentoo, works fine here Mar 02 00:06:52 CoreDump|home: I am sorry koen made you mad... but I don't think namecalling is going to solve anything here... in fact it will only make things worse. I think at this point we should all go back to our corners and try to talk about this more tomorrow... Mar 02 00:07:44 sound like my mail made it to the ML then. But I didn't call him names =) Mar 02 00:09:18 CoreDump|home: no, but calling his changes BS is not going to help things either... Mar 02 00:10:05 I called his reasons BS because they are. He didn't mention a _single_ valid reason for his behavious. Not one. Mar 02 00:11:20 CoreDump|home: That may be true... but using the TERM *BS* will only cause more emotional stuff rather than (good or bad) technical reasons to come out. Mar 02 00:12:40 hvontres|poodle: no harm done. he wasn't citing any technical stuff in the whole thread (which started w/ his 20-hour RCF btw) Mar 02 00:14:32 CoreDump|home: I disagree. We should all be aware that we are no longer teenage script-kiddies here. And if we want to make sure the (limited) number of dev's stick around, we need to make sure we behave like the rational adults I think we are capable of being. Mar 02 00:14:58 CoreDump|home: Sorry, but I get enough chidish bickering at home from my 2 and 5 year olds :) Mar 02 00:15:16 hvontres|poodle: heh - mine are 2 and 6 :-0 Mar 02 00:15:40 rwhitby: boys or girls? Mar 02 00:15:47 hvontres|poodle: =) Just read "[oe] RFC: splitting deploy/ipk into subarchs" and tell me I didn't comment in a polite manner. Then read his reply to my comment. Mar 02 00:16:34 hvontres|poodle: 2 yr old girl, 6 yr old boy. Mar 02 00:17:03 * mwester now feels *really* old ;-) Mar 02 00:17:06 CoreDump|home: I did. But I don't think that we will solve anything by lowereing our standards to the lowest common denominator (agin wearing my dad hat ;) ) Mar 02 00:17:46 rwhitby: heh... two girls... not looking forward to teenagerdom... Mar 02 00:17:50 BUT HE STARTED IT lalala ;) Mar 02 00:17:55 three girls. Mar 02 00:18:01 ~fish CoreDump|home Mar 02 00:18:03 * ibot slaps CoreDump|home around with a large trout Mar 02 00:18:08 hehehe Mar 02 00:18:15 * hvontres|poodle wishes that worked at home.... Mar 02 00:18:21 at least in America we have guns to threaten teen age boys with Mar 02 00:18:33 my friend has two daughters Mar 02 00:18:34 CoreDump|home: hmmm or power tools ... Mar 02 00:18:38 Two in college -- not much younger than some of the developers here, I bet. I just take away their iPod when they tick me off ;) Mar 02 00:18:55 we have plans Mar 02 00:19:07 03rpurdie 07org.oe.dev * re534053a... 10/ (1 packages/mtd/mtd-utils_1.0.0+git.bb): mtd-utils-native: Tweak compile options to interact better with the Makefile Mar 02 00:19:44 Speaking of pissed off, a friend of mine (not even a nerd) was cold called by google for an interview Mar 02 00:19:54 * Crofton|laptop is not speaking to her Mar 02 00:21:06 Crofton__: =) Mar 02 00:21:32 Google appear to be approaching a lot of people Mar 02 00:21:57 not me :( Mar 02 00:21:59 RP: syncing up, and will test. Does the new lzo jffs2 require a new uncompressor in the kernel, then? Mar 02 00:22:08 mwester: yes Mar 02 00:22:12 She is a Dr in statistics Mar 02 00:22:30 That could be a problem for Unslung... stuck on 2.4.22 (the Linksys kernel version) Mar 02 00:22:40 mwester: I'm sure the commits are close but its hard for me to test fully since lzo is also on my system Mar 02 00:23:04 I'll poke about; if the package is there now, I'm sure it can be made to work. Mar 02 00:23:35 * mwester hopes there is a patch for the 2.4 kernel, too. Mar 02 00:24:04 * RP hopes mwester is joking Mar 02 00:24:27 * mwester wishes he was joking too Mar 02 00:24:48 mwester: This is why I've disabled that compressor by default. We can enable it when we know a given machine's kernel supports it Mar 02 00:25:32 * joshin Boy 5.5 & Girl 3 (today) Mar 02 00:26:55 Ok then, no problem. it would have been nice to get a little free space back with better compression, but c'est la vie. Mar 02 00:27:18 Some google recruiting guy sent me an email. Pity the job wasn't local. Mar 02 00:27:33 I'm pretty sure they have a resume-bot. Mar 02 00:27:43 * hvontres|poodle wonders if google needs any ME's... Mar 02 00:27:50 mwester: Its not better compression, its actually slightly worse. It improves disk read time by 40% and makes bootup faster though Mar 02 00:28:22 RP: disk read time or flash read time? Mar 02 00:28:38 Ah, ok. Well, then probably not good for unslung, but there's some SlugOS users who might be REALLY interested in that. Mar 02 00:28:40 She confirms she has a resume online Mar 02 00:28:48 rwhitby: Faster decompression of the data = faster file reads Mar 02 00:29:52 hehe.. fancy that... google searching online resumees... what will they think of next :) Mar 02 00:29:59 Like most things, its useful for some cases and not so much in others Mar 02 00:30:11 rwhitby, thats what she said (being local) Mar 02 00:30:31 rwhitby: They also harvest email addresses at events like FOSDEM Mar 02 00:31:01 RP: hmmm. Sounds like that is the Theme of the day (usefull or not) Mar 02 00:32:33 RP: I got my email from Google the day after I presented http://www.nslu2-linux.org/presentation.pdf at the LinuxSA user group. Mar 02 00:33:06 (the announcement to the user group mailing list containing the talk abstract probably triggered it) Mar 02 00:37:11 Hmm, I did a lightning talk at FOSDEM Mar 02 00:37:18 maybe that will kick them into action Mar 02 00:37:55 rwhitby: This was a few weeks afterwards. Several people within openedhand got emails at about the same time Mar 02 00:38:11 Crofton|laptop: But what would google do with SDR stuff ? Mar 02 00:38:30 hvontres|poodle: index all broadcasts worldwide? Mar 02 00:38:46 (on any frequency) Mar 02 00:39:56 'night all Mar 02 00:39:59 heh Mar 02 00:40:03 wireless google Mar 02 00:40:12 (I'll deal with any remaining lzo issues tomorrow) Mar 02 00:41:13 Keep ahead of MS Mar 02 00:51:21 rwhitby: So, do you have that large fortune yet? Mar 02 00:52:02 hvontres|poodle: I did have, but it turned into a small fortune like the slide says. Mar 02 00:52:29 nslu2-linux has raised over USD$10,000 dollars in donations over the last 2.5 years though, all of which has gone back into the project. Mar 02 01:04:05 rwhitby: so what happende to that bare looking pcb in your presentation? Mar 02 01:05:45 it was heated with a heat gun until all the components fell off Mar 02 01:05:54 so we could see the traces under the components. Mar 02 01:14:09 salted slug? Mar 02 01:16:13 yeah, the salted slug. Mar 02 01:17:10 rwhitby: I see.. It looked almost like one or our routed prototype borads without the soldermask :) Mar 02 01:17:30 yep, that was the intention :-) Mar 02 01:17:43 it helped us find the jtag pins Mar 02 01:17:55 great idea! Mar 02 01:18:34 [g2] was the slug sacrificer Mar 02 01:19:35 i recently succeeded it doing bitbake linux-rp.2.6.20. Now, if I want to make some modifications to say the kernel configuration, where do I locate the source code within the directory structure. Also, once I do make the changes do I simply re-run the bitbake command again and will it find the changes and re-compile? Mar 02 01:24:39 anyone? Mar 02 01:26:10 cyrus__: the defconfig is usually in the OE metadata Mar 02 01:26:29 packages/linux/linux-rp-*/defconfig maybe Mar 02 01:27:16 cyrus__: Preferred method is to change the defconfig file for your machine in org.openembedded.dev/packages/linux/, and to force a recompile, do this: bitbake -cclean target-name && bitbake target-name, where target-name is the kernel .bb Mar 02 01:33:20 NAbyss_ - if I do a clean will it rebuild everyting (ie: toolchain, or just the kernel) Mar 02 01:33:40 cyrus__: If you -cclean the kernel target, it'll only remove the kernel Mar 02 01:33:47 ok..thanks Mar 02 01:34:15 NAByss_ - last question, if I want to actually change the source code (ie: add a new module that I am working on) where would I do that Mar 02 01:35:00 cyrus__: your best bet would be to add it to the SRC_URI section in the .bb file Mar 02 01:35:01 cyrus__: Suggested method is to generate a diff -u against the upstream sources, then reference that as a patch in the kernel .bb Mar 02 01:35:15 ok..thanks guys..much appreciated Mar 02 01:35:36 np **** ENDING LOGGING AT Fri Mar 02 03:00:00 2007