**** BEGIN LOGGING AT Thu Nov 26 02:59:56 2009 Nov 26 07:51:40 03Sebastian Spaeth  07org.openembedded.dev * rd099cc6187 10openembedded.git/recipes/calc/calc_0.0.3.bb: Nov 26 07:51:40 calc: add v0.0.3 recipe Nov 26 07:51:40 Signed-off-by: Sebastian Spaeth Nov 26 08:34:34 03Sebastian Spaeth  07org.openembedded.dev * r524b40aa58 10openembedded.git/recipes/shr/ (initscripts-shr/70glamoFIFOtweak initscripts-shr_0.0.1.bb): Nov 26 08:34:34 initscripts-shr: add glamoFIFOtweak to Xsession.d Nov 26 08:34:34 Signed-off-by: Sebastian Spaeth Nov 26 08:38:24 03Sebastian Spaeth  07org.openembedded.dev * rcbca2a9362 10openembedded.git/conf/distro/include/sane-srcrevs.inc: Nov 26 08:38:24 sane-srcrevs.inc: bump 'calc' sane rev to latest release Nov 26 08:38:24 Signed-off-by: Sebastian Spaeth Nov 26 08:56:00 hi Nov 26 08:56:35 good morning Nov 26 08:57:45 woglinde: hey Nov 26 08:57:50 03Sebastian Spaeth  07org.openembedded.dev * r4ae61a9144 10openembedded.git/recipes/shr/ (initscripts-shr/70glamoFIFOtweak initscripts-shr_0.0.1.bb): Nov 26 08:57:50 initscripts-shr: remove runtime glamo tweak. Has been pushed into andy-tracking kernel sources now. Nov 26 08:57:50 Signed-off-by: Sebastian Spaeth Nov 26 09:00:08 03Steve Sakoman  07org.openembedded.dev * r6ec12de553 10openembedded.git/recipes/flac/flac.inc: flac: disable xmms plugin since it breaks the build Nov 26 09:00:09 03Steve Sakoman  07org.openembedded.dev * r911c3b5a21 10openembedded.git/recipes/gnome/gnome-system-monitor_2.28.0.bb: gnome-system-monitor: fix typo Nov 26 09:00:09 03Steve Sakoman  07org.openembedded.dev * r8d2be4883b 10openembedded.git/recipes/gnome-mplayer/gnome-mplayer_svn.bb: gnome-mplayer: bump SRCREV Nov 26 09:00:10 03Steve Sakoman  07org.openembedded.dev * rd1ca298d4a 10openembedded.git/recipes/gnome/ (8 files in 2 dirs): gdm: update to 2.28.1 Nov 26 09:00:13 03Steve Sakoman  07org.openembedded.dev * rba2d1b5e6d 10openembedded.git/recipes/gnome/system-tools-backends_2.6.1.bb: system-tools-backends: fix packaging Nov 26 09:00:16 03Steve Sakoman  07org.openembedded.dev * r47a90e56cd 10openembedded.git/recipes/totem/totem_2.28.1.bb: totem: don't disable python Nov 26 09:01:29 zecke: good afternoon Nov 26 09:03:41 pb___: wow, I'm really thankful about the cleanups on the waterfall code Nov 26 09:03:54 pb___: deployment is still aa todo? Nov 26 09:04:03 zecke: yeah, seems excellent Nov 26 09:04:06 zecke: right Nov 26 09:04:54 pb___: okay, let me see if I find someone to help me... I'm a bit tired from the flight as well Nov 26 09:06:12 ok, cool Nov 26 09:08:24 jo zecke Nov 26 09:08:32 still fighting to get an image with uclibc Nov 26 09:08:36 hi floian Nov 26 09:08:37 hi pb Nov 26 09:09:29 good morning Nov 26 09:11:30 hi woglinde Nov 26 09:11:48 woglinde: great, we could have a waterfall and a policy to not break things soon :) Nov 26 09:17:29 woglinde: Hi, libpng 1.2.40 is in my tree if you want to pull http://git.openpandora.org/cgi-bin/gitweb.cgi?p=openembedded.git;a=commit;h=5cdc0fde23afb5b04175581ba812232b8713fb63 (had 5mins spare last night) Nov 26 09:17:53 03Sebastian Spaeth  07org.openembedded.dev * r1bea6c5106 10openembedded.git/ (2 files in 2 dirs): Nov 26 09:17:53 sane-srcrevs.inc: bump linux-openmoko-shr-devel Nov 26 09:17:53 Signed-off-by: Sebastian Spaeth Nov 26 09:17:53 03Sebastian Spaeth  07org.openembedded.dev * r9a65c96b32 10openembedded.git/: Merge branch 'org.openembedded.dev' of ssh+git://git@git.openembedded.net/openembedded into org.openembedded.dev Nov 26 09:18:11 djwillis hm did you test in on the machine? Nov 26 09:19:48 woglinde: built on the Pandora and displayed a PNG but NOT tested throughly, I have a feeling there is a package bug that is also in the current recipe (I I get odd referances to a .3 symlink not the .1) Nov 26 09:22:35 woglinde: I should mention that the .3 oddness is there with the older versions also. Not tracked it down Nov 26 09:26:33 Is obexd buildable in angstrom? I'm getting unbuildable depencie chain in SHR because bluez-libs is blacklisted. Nov 26 09:28:31 heinervdm hm the old bluez3 vs bluez4 saga Nov 26 09:28:40 dont know if obexd works with bluez4 Nov 26 09:29:26 woglinde: i wanted to try, bluez-libs is provided by bluez4 recipe, so at least bitbake should try to build it... Nov 26 09:29:55 let me see the recipe Nov 26 09:30:48 hm dep on bluez-lib Nov 26 09:31:01 woglinde: and there is a new version from 2. october on bluez.org so i think the ported it to bluez4 Nov 26 09:31:15 hm Nov 26 09:31:29 then update the recipe Nov 26 09:31:45 ok, but will try before :) Nov 26 09:54:53 03Koen Kooi  07org.openembedded.dev * r71cf41ca7e 10openembedded.git/recipes/ (e17/e-wm_svn.bb shr/e-wm-menu-shr_git.bb): Nov 26 09:54:53 e-wm: stop RRECOMMENDING e-wm-menu Nov 26 09:54:53 e-wm-menu-shr: stop RPROVIDING it Nov 26 09:55:03 03Koen Kooi  07org.openembedded.dev * re430bbb8f3 10openembedded.git/recipes/gnome-mplayer/gnome-mplayer_svn.bb: gnome-mplayer: fix typo Nov 26 09:57:10 woglinde: it's openobex that depends on bluez-libs and there is no new version since 9 months, so can i change the openobex_1.5 recipe to depent on bluez4 instead of bluez-libs? Nov 26 09:57:41 hm Nov 26 09:57:45 did it work? Nov 26 09:58:01 i compiles Nov 26 09:58:02 it Nov 26 10:00:31 woglinde: http://cgit.openembedded.net/cgit.cgi/openembedded/commit/?id=f1dc5457a0352369f41989ae58647f5b9e7e6b43 Nov 26 10:00:52 So it would be reverting this... Nov 26 10:01:41 hm so ask pb_ about it Nov 26 10:02:20 pb___: ^^ Nov 26 10:04:13 hi xora Nov 26 10:08:31 hey woglinde Nov 26 10:10:09 03Koen Kooi  07org.openembedded.dev * re7e0893066 10openembedded.git/conf/checksums.ini: checksums.ini: add gdm checksum Nov 26 10:16:00 03Klaus Kurzmann  07org.openembedded.dev * r22505097ca 10openembedded.git/recipes/tasks/task-shr-minimal.bb: Nov 26 10:16:00 task-shr-minimal: add e-wm-menu-shr Nov 26 10:16:00 Signed-off-by: Klaus Kurzmann Nov 26 10:16:01 03Klaus Kurzmann  07org.openembedded.dev * rbe98757e02 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git.openembedded.net/openembedded into org.openembedded.dev Nov 26 10:16:03 03Klaus Kurzmann  07org.openembedded.dev * r55653a9369 10openembedded.git/conf/distro/include/sane-srcrevs.inc: Nov 26 10:16:06 sane-srcrevs.inc: bump mokonnect to 112 Nov 26 10:16:08 Signed-off-by: Klaus Kurzmann Nov 26 10:26:22 Heinervdm: why do you want to do that? Nov 26 10:26:43 pb_: Because in angstrom and shr openobex is unbuildable because bluez-libs are blacklisted and setting PREFERRED_PROVIDER for bluez-libs doesn't work Nov 26 10:27:28 that's very sad, but I don't think making it depend on bluez4 is the right answer to that problem. Nov 26 10:27:42 you should probably figure out why setting PREFERRED_PROVIDER does not work and fix that instead. Nov 26 10:29:07 pb_: ok, if i have to guess then i would say it's the fault of ANGSTROM_BLACKLIST that it's blacklisting the PREFERRED_PROVIDER also Nov 26 10:29:15 will do some research Nov 26 10:29:39 righto Nov 26 10:30:05 if you change the DEPENDS then all that will happen is you will make it unbuildable for those people who are still using bluez3 instead, which is not much of an improvement. Nov 26 10:30:21 yes Nov 26 10:33:25 hi effem Nov 26 10:33:46 hi cbrake Nov 26 10:37:52 bluez4 PROVIDES bluez-libs so it should all work Nov 26 10:40:25 XorA: yes it should, but it doesn't. Nov 26 10:43:43 XorA: if you have ANGSTROM_BLACKLIST_pn-bluez-libs you can try to build openobex and see if it's working for you Nov 26 10:44:03 Heinervdm: already ahead of you Nov 26 10:44:13 :) Nov 26 10:44:25 Heinervdm: I am responsible for the blacklist BTW Nov 26 10:45:08 XorA: ok :) blacklist looks ok, so next thing i'm looking into is bb.data.getVar Nov 26 10:46:58 openobex builds here Nov 26 10:47:13 strange Nov 26 10:50:31 XorA: hmm, the only difference i can see between agnstrom and shr, is that angstrom defines ANGSTROM_BLACKLIST after PREFERRED_PROVIDER Nov 26 10:52:20 that "shouldnt" make a difference Nov 26 10:52:37 i'm trying :) Nov 26 10:53:51 Missing or unbuildable dependency chain was: ['openobex', 'bluez-libs'] Nov 26 10:55:59 Heinervdm: add a PROVIDES_append_shr to bluez4.inc Nov 26 10:58:04 !oebug 5176 Nov 26 10:58:05 * * Bug 5176, Status: CONFIRMED, Created: 2009-06-16 11:27 Nov 26 10:58:06 * * : increase granularity for bluez4 package Nov 26 10:58:07 * * http://bugs.openembedded.org/show_bug.cgi?id=5176 Nov 26 10:59:35 XorA: thx, that's working, even if i don't understand why Nov 26 11:00:49 Heinervdm: PROVIDES_bluez4 must be malfunctioning somehow Nov 26 11:01:22 "PROVIDES_bluez4" isn't going to do anything useful because it won't match any OVERRIDE. Nov 26 11:02:18 ah yeah it should be pn-bluez4 Nov 26 11:02:29 Heinervdm: can you try that? Nov 26 11:02:42 well, given that it's in the recipe-specific .inc file, it might as well just be "PROVIDES" Nov 26 11:02:57 no recipes other than bluez4 are ever likely to see it so the override is fairly redundant Nov 26 11:03:03 pb_: I think the idea is that .inc might be used in other recipes Nov 26 11:03:06 * Heinervdm trying Nov 26 11:03:14 pb_: but currently nothing does Nov 26 11:03:31 which other recipes would those be? the contents of the .inc look pretty specific. Nov 26 11:04:26 pb_: no idea actually, Ive no idea who made that .inc Nov 26 11:04:33 heh Nov 26 11:04:41 oh well, the override is harmless, might as well just leave it there Nov 26 11:04:51 PROVIDES_pn-bluez4 works Nov 26 11:05:01 ok, very good Nov 26 11:05:19 Heinervdm: well you have explained why kaelios blew up as well now :-D Nov 26 11:05:31 will you change it? i have no commit right :) Nov 26 11:05:35 :) Nov 26 11:05:39 XorA: btw, if that .inc ever does get used by other recipes then the PROVIDES_append_angstrom would need changing as well. Nov 26 11:05:39 Rolf added that line Nov 26 11:07:19 morning all Nov 26 11:07:53 hi rp Nov 26 11:08:29 hey RP Nov 26 11:08:54 pb_: yeah Nov 26 11:09:49 morning Nov 26 11:10:24 03Koen Kooi  07org.openembedded.dev * rdfd7031692 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.org:openembedded into org.openembedded.dev Nov 26 11:10:34 03Koen Kooi  07org.openembedded.dev * rcd7829fa47 10openembedded.git/recipes/gnome/gdm_2.28.1.bb: gdm: tweak do_install a bit Nov 26 11:10:40 hmmm, there is no bluez-utils-dbus Nov 26 11:11:31 hi hrw Nov 26 11:11:48 it looks like it would be safe to squash all those PROVIDES into one Nov 26 11:12:06 as bluez-utils-dbus doesnt exist anymore Nov 26 11:12:33 that's possible... we should check if anything still depends on it Nov 26 11:12:46 florian: its RPROVIDER and gpe depends on it Nov 26 11:13:05 florian: but the actual recipe is long gone Nov 26 11:13:30 XorA: ok, this sounds like it is easy to fix Nov 26 11:15:07 Heinervdm: I think we could safely make bluez4.inc non distro hacked Nov 26 11:15:21 XorA: i think so too Nov 26 11:15:34 Heinervdm: you have commit access? Nov 26 11:15:39 XorA: no Nov 26 11:15:48 Ill do it here then Nov 26 11:15:54 thx :) Nov 26 11:16:50 XorA: -dbus ones were 2.xx times Nov 26 11:24:12 hi hrw Nov 26 11:28:31 03Thomas Zimmermann  07org.openembedded.dev * r05e63abd2b 10openembedded.git/ (conf/checksums.ini recipes/bluez/obexd_0.19.bb): Nov 26 11:28:31 obexd: add version 0.19 Nov 26 11:28:31 Signed-off-by: Klaus Kurzmann Nov 26 11:30:51 03Graeme Gregory  07org.openembedded.dev * ra863329a96 10openembedded.git/recipes/bluez/bluez4.inc: Nov 26 11:30:51 bluez4.inc : remove DISTRO hacks as they arent really needed its ok Nov 26 11:30:51 to add the same info to the generic versions of variables. Also Nov 26 11:30:51 PROVIDES_bluez4 was wrong as bluez4 isnt an OVERRIDE Nov 26 11:30:52 03Graeme Gregory  07org.openembedded.dev * r39e66d2f9b 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git@git.openembedded.org/openembedded into org.openembedded.dev Nov 26 11:33:00 hihi Nov 26 11:33:10 I am waiting for koen Nov 26 11:33:19 for the bluez case Nov 26 11:33:24 popcorn time? ;) Nov 26 11:33:33 bluez case? Nov 26 11:34:34 florian we will see Nov 26 11:54:17 hms mplayer needs some thoughts abut micro and / path too Nov 26 11:54:23 hi gnutoo Nov 26 11:54:36 hi Nov 26 12:23:59 Someone knows why curl is compiled with --withou-ssl? Nov 26 12:24:22 presumably because the ssl libraries are quite heavyweight and/or controversial Nov 26 12:24:29 Any objection about adding it? Nov 26 12:24:38 feel free to add it as a DISTRO_FEATURE Nov 26 12:24:43 pb_: well, it has gnutls support as well Nov 26 12:24:53 indeed, gnutls is controversial Nov 26 12:25:05 lol Nov 26 12:25:11 and still moderately large in comparison to libcurl Nov 26 12:25:58 openssl is controversial in different ways, and also larger. I don't think either of those is going to be universally accepted. Nov 26 12:26:43 but, as I say, selecting gnutls as a DISTRO_FEATURE would be a fine thing to do. Nov 26 12:26:46 or openssl for that matter Nov 26 12:27:00 or indeed nss Nov 26 12:27:09 oh ha nss Nov 26 12:27:34 pb_: well; curl has gnutls enabled by default Nov 26 12:27:37 ~lart uClibc webfrontend for git Nov 26 12:27:38 * ibot takes a large goose feather pillow and swings it wildly in uClibc webfrontend's direction, hitting uClibc webfrontend and sending uClibc webfrontend flying into the closet for git Nov 26 12:27:40 pb_: please look Nov 26 12:27:49 pb_: master, can you explain me why one has to sign a copyright assignment to the FSF for each project? Nov 26 12:28:40 zecke: for each project it is needed but not for each change Nov 26 12:28:53 zecke: but it depends on how you do the assigment Nov 26 12:29:10 zecke: I can, for example, play freely with parted Nov 26 12:29:12 otavio: I don't quite understand your point. Are you saying that you want to compile curl with both gnutls _and_ openssl by default, then? Nov 26 12:29:29 I'm not sure that is even possible, and it definitely doesn't sound like a very good idea. Nov 26 12:29:38 pb_: no; I'm saying that CURL_FEATURES list gnutls as enabled Nov 26 12:29:52 zecke: I don't think it's strictly necessary to do it for every project, you can sign a blanket one for "all gnu software". Nov 26 12:30:04 otavio: so what did you mean about adding "--with-ssl"? Nov 26 12:30:04 * * OE Bug 5342 has been created by kristoffer.ericson(AT)gmail.com Nov 26 12:30:06 * * libqpe-opie_cvs fails to build Nov 26 12:30:08 * * http://bugs.openembedded.org/show_bug.cgi?id=5342 Nov 26 12:30:33 pb_: this; --with-ssl because even though it has gnutls it is adding without-ssl Nov 26 12:30:37 pb_: this is confusing to me Nov 26 12:31:05 "--with-ssl" means to use openssl Nov 26 12:31:09 otavio bad m4 macro Nov 26 12:31:21 so, if you add --with-ssl and keep gnutls, you will end up with both. which, as I say, does not sound like a very good thing. Nov 26 12:31:54 pb_: ah I should have used this wildcard... I managed to do paperwork for gdb and glibc now Nov 26 12:32:24 * Crofton|work can contribute to gnuradio Nov 26 12:33:32 zecke: very good Nov 26 12:34:59 humm Nov 26 12:35:01 curl is a mess Nov 26 12:35:04 crap Nov 26 12:38:28 patches welcome Nov 26 12:40:35 as for gnutls being enabled by default, yeah, that does appear to be the situation. I think this is a mistake but it is probably not worth changing at this point. Nov 26 12:41:11 in due course it would be a good idea to teach curl to respect DISTRO_FEATURES as well as its own CURL_FEATURES variable but there are probably more urgent things to worry about right now. Nov 26 12:42:39 pb_: I'll add an openssl option in CURL_FEATURES and use it then Nov 26 12:43:08 it already understands openssl in CURL_FEATURES, you just need to set that in your distro config Nov 26 12:50:41 pb_: in dev; I backported it to stable now Nov 26 12:51:14 03Koen Kooi  07org.openembedded.dev * r8af5e0ed1a 10openembedded.git/recipes/linux/ (17 files in 4 dirs): Nov 26 12:51:14 linux 2.6.24: add support for smartq5 device (machine.conf will follow later) Nov 26 12:51:14 * also fix task ordering due to kernel.bbclass changes Nov 26 12:52:24 ok, very good Nov 26 12:52:38 hms fighting uclibc and O_CLOEXEC Nov 26 12:54:06 woglinde: oh? Nov 26 12:54:31 yeah 0.9.30.1 has no support Nov 26 12:54:39 backport some stuff from master Nov 26 12:55:04 lets see if if works Nov 26 12:55:17 pulseaudio again was the root Nov 26 12:55:58 hi woglinde Nov 26 12:56:02 hey pb_ Nov 26 12:56:07 jo ant Nov 26 12:56:17 hi ant_work Nov 26 12:56:59 pb_: he...perhaps I was a bit noisy but I upset kergoth last night... Nov 26 12:57:27 ...and woglinde did disappear :D Nov 26 12:57:37 nope Nov 26 12:57:43 I left later Nov 26 12:57:47 joking Nov 26 12:58:14 pb_: we need some package-master skillz... Nov 26 12:58:37 to beat up two packaging oddities Nov 26 12:59:06 so compiles at least Nov 26 12:59:10 hi likewise Nov 26 12:59:48 argsgsgsgsgsgss Nov 26 12:59:48 again someone forgot to take patches from older version Nov 26 13:00:15 ant_work: heh, what did you say to kergoth? Nov 26 13:00:33 ~lart lazy developers Nov 26 13:00:33 * ibot strangles lazy developers with a doohicky mouse cord Nov 26 13:01:08 hi woglinde Nov 26 13:01:08 hi all Nov 26 13:01:10 gm Nov 26 13:01:15 pb: just evocating him was enough...bad day perhaps Nov 26 13:02:14 pb_: it is about 1)libpcre 2)libgtkstylus having wrong -dev packaging ( .so is not in -dev) Nov 26 13:02:31 libpcre is a special case, just look at the recipe Nov 26 13:02:55 heh. "evocating" is a good word. Nov 26 13:02:57 so it's way beyond my actual packaging skillz Nov 26 13:03:16 yeah, I can imagine kergoth getting annoyed if you tried to talk to him about those packages, I doubt he cares much about them. Nov 26 13:03:47 args liam also didnt switch the PR = "${INC_PR}.4" Nov 26 13:03:49 over Nov 26 13:03:52 arhgssssssss Nov 26 13:04:03 never know..perhaps he did start a build...but the turkey of this night will slow down any output... Nov 26 13:04:20 woglinde: Your args keys seem to be getting quite worn these days :/ Nov 26 13:04:59 I must admit I don't care all that much about pcre either. my father knows the author of that library but that is about the total extent of my connection to it :-} Nov 26 13:05:03 rp its a pity some updates a version and take over the work ypu made Nov 26 13:05:07 pb_: being that perl is bad beast for OE, I thought libpcre could be easier.. Nov 26 13:05:11 I was wrong... Nov 26 13:05:15 heh Nov 26 13:05:37 yeah, that recipe does look like a bit of a mess Nov 26 13:06:00 as for libgtkstylus, are you sure the packaging is wrong? iirc, the .so file should not be in the -dev package there. Nov 26 13:06:19 hm... Nov 26 13:06:48 usually .so are not installed with PN Nov 26 13:06:50 or? Nov 26 13:07:03 woglinde: hmm, which package? :/ Nov 26 13:07:24 ant_work: usually not, but that is not quite the same as never Nov 26 13:07:24 rp pulseaudio Nov 26 13:07:45 hm lennart has no QA sense Nov 26 13:07:54 removing paplay but not the manpages Nov 26 13:07:55 woglinde: :( Nov 26 13:08:03 pb_: aren't usually linked against .s0.0.0 ? Nov 26 13:08:05 libgtkstylus is a gtk plugin, not a shared library. it shouldn't have a -dev package at all. Nov 26 13:08:22 pb_: argh Nov 26 13:08:30 I guess I should rant as much about lennart, as he does about other projects Nov 26 13:08:42 admittedly calling it "lib..." was probably not the cleverest move Nov 26 13:08:57 I guess I can understand why you might be fooled into thinking it was a library :-} Nov 26 13:09:56 pb_: I commmited a bad fix then... Nov 26 13:10:02 http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=38361a2cd9b962daccf5af7c85d54245e3711f9e Nov 26 13:10:08 ant_work: FWIW libpcre is in meta-moblin in poky Nov 26 13:10:38 * RP doesn't remember seeing QA errors but there might be some Nov 26 13:11:02 pb_: that fix was for libopieobex... a plugin Nov 26 13:11:22 RP: just these about .so slipped in Nov 26 13:11:37 just these 2 over console-image opie-image x11-image Nov 26 13:11:43 ant_work: I can't speak for libopieopieobex, that might be different. Nov 26 13:12:09 I don't really know anything about qt plugins. Nov 26 13:12:24 ant yes Nov 26 13:12:31 plugins only have .so Nov 26 13:12:32 pb_: and while we are there....http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=3e9bb4f23c205dde9ab1b042b4bfc2c1b5fe8243 Nov 26 13:12:41 that was for klibc.so Nov 26 13:12:58 that one probably is fine Nov 26 13:13:04 What's the correct way to specify the SRCREV of a GIT recipe for only one specific machine? Nov 26 13:13:04 hurray uclibc builds again Nov 26 13:13:11 args Nov 26 13:13:13 pb_: phewwww Nov 26 13:13:18 I meant pulseaudio Nov 26 13:13:18 klibc clearly is a library, not a plugin, and I expect it probably does install the library separately from the .so symlink Nov 26 13:13:22 ~hail woglinde Nov 26 13:13:23 * ibot bows down to woglinde and chants, "I'M NOT WORTHY!!" Nov 26 13:13:47 likewise: SRCREV_pn-recipe_machine = "...."? Nov 26 13:13:49 woglinde: ah but probably pulseaudio under uclibc, right? Nov 26 13:13:52 something like that, I guess Nov 26 13:13:56 likewise jupp Nov 26 13:13:59 pb_: it was late but I remember having checked the created files and these seemed ok... Nov 26 13:14:25 likewise but I only fixed it for 0.9.30.1 Nov 26 13:14:40 woglinde: :-) Nov 26 13:15:22 woglinde: good, I'm looking at going from 0.9.27 to 0.9.30.1 for the new nios2 architecture. I still don't have hardware to test on though. Nov 26 13:17:57 likewise maybee 0.9.30.2 is out than Nov 26 13:20:04 pb_: xtscal patches should be sent to which address? Nov 26 13:20:53 philb@gnu.org please Nov 26 13:21:00 RP: we sent two OE patches upstream (kexec-tools). Should we use any tag in the recipe? Nov 26 13:21:13 iirc in SRC_URI there are some... Nov 26 13:21:28 03Koen Kooi  07org.openembedded.dev * r51f172ce53 10openembedded.git/ (5 files in 3 dirs): e17: bump SRCREV and start adapting it to new-style staging Nov 26 13:21:35 pb_: ok Nov 26 13:23:40 pb_: ....and about QA, the remaining are about (supposedly) broken desktop file. Would you give me a pointer to matcgh those? Nov 26 13:23:54 what's the actual error? Nov 26 13:24:16 mom..oestats... Nov 26 13:24:51 e.g. the desktop file is not valid, gpe-taskmanager... Nov 26 13:25:49 iirc doesn't apply to OPIE (bogus QA?) Nov 26 13:26:18 does it not say any more than that? Nov 26 13:26:34 I think it ought to show you the output from desktop-file-validate. Nov 26 13:29:32 unfortunately the logs are one-liners... Nov 26 13:29:55 ok. maybe you need to run desktop-file-validate by hand in that case. Nov 26 13:30:15 right, I'll do by hand Nov 26 13:30:36 it's a X thing, isn't? Nov 26 13:30:48 well, fd.org, yeah Nov 26 13:30:58 ok, OPIE is out then Nov 26 13:31:41 perhaps we should skip QA for that packages? Nov 26 13:32:18 in some OPIE include..alltogether Nov 26 13:32:28 e.g. the desktop file is not valid, opie-console Nov 26 13:32:56 there are many many Nov 26 13:33:25 yeah, dunno. it's quite possible that opie doesn't consider it a goal to have their desktop files be fdo-valid. Nov 26 13:33:45 you'd need to take that up with the opie folks, I guess. Nov 26 13:34:12 yea, I'll ask bluelightning and woglinde Nov 26 13:34:15 it was not a goal back then (these files pre date fdo) Nov 26 13:34:19 pb_: many thx Nov 26 13:34:22 morning Nov 26 13:34:28 g'day mickey|office Nov 26 13:34:36 zecke: safely landed? Nov 26 13:35:02 mickey|office: yeah Nov 26 13:35:02 good Nov 26 13:35:07 mickey|office: a "lot" of delays though... Nov 26 13:35:45 really? China Airlines is usually quite on time Nov 26 13:37:47 mickey|office: well, not their faults Nov 26 13:38:09 there was more than usual air traffic in Taipei, so we had to wait in the air for ~30 minutes Nov 26 13:39:06 ah Nov 26 13:39:28 mickey|office: and the ground staff forget to give me the priority tag for the baggage... :) Nov 26 13:39:43 heh Nov 26 13:55:13 is there a gtk2-perl module inside Angstrom ? Nov 26 13:59:57 pwgen: you might try #angstrom for that Nov 26 14:03:47 *G* . adapt my question to this channel: is there bb file for gtk2-perl inside OE ? Nov 26 14:04:59 pb_: i am using a mixture from angstrom + openmoko stuff for my akita .. and try to get tinyca2 running Nov 26 14:06:21 I think that you need to check in OE metadata first anyway Nov 26 14:13:05 * * OE Bug 5343 has been created by kristoffer.ericson(AT)gmail.com Nov 26 14:13:07 * * libqpe-opie_1.2.4 fails to build cannot find -lqte Nov 26 14:13:09 * * http://bugs.openembedded.org/show_bug.cgi?id=5343 Nov 26 14:17:24 * mickey|office overwhelmed that people are still using Opie Nov 26 14:17:59 pb_: sent Nov 26 14:20:04 mickey|office: indeed, still developing it even. Nov 26 14:21:40 mickey|bbl, :P Nov 26 14:22:22 hrw: what version of XCALIBRATE is required for those changes? there doesn't seem to be an appropriate version check in the code as far as I can tell. Nov 26 14:23:15 (and what versions of the xserver include the necessary XCALIBRATE?) Nov 26 14:24:03 also, I don't quite understand the purpose of patch #1. what is that change in aid of? Nov 26 14:25:53 pb_: we use libxcalibrate 7b8e29186097170e930de634a4a68072558aa056 git revision Nov 26 14:26:22 pb_: for #1 - Richard asked me to merge Poky changes too - and this is one of them Nov 26 14:26:58 I'm not necessarily opposed to #1 but it is not obvious to me why that patch is an improvement. I think I would need a bit more explanation before merging it. Nov 26 14:27:06 hm..about that OPIE bug...what the status of mips buildhost? Nov 26 14:27:25 some people asking me..saying it was ugly broken not long ago Nov 26 14:28:04 btw - that libxcalibrate is basically same as HEAD of that branch Nov 26 14:28:18 pb_: ok, will ask RP about it Nov 26 14:29:51 for XCALIBRATE, what we would ideally need is (i) an appropriate pkgconfig test to make sure that you have a new enough version of the client library, and (ii) an appropriate runtime test to make sure that the server supports a new enough version of the extension. Nov 26 14:31:05 if there are still servers around that are using an older version of the wire protocol then, in a perfect world, xtscal would fall back to its previous mode of operation. if that's not feasible then it ought at least to print a meaningful diagnostic. Nov 26 14:40:34 ok, but thats beyond my skills so far Nov 26 14:42:08 upstream X just breaks stuff and sees if anyone complains. Nov 26 14:45:27 Well, that's one strategy for doing QA... Nov 26 14:51:34 re Nov 26 15:14:47 anyone taken a look at the package_tar bug? Nov 26 15:14:50 kergoth, ? Nov 26 15:18:05 do you have a bug number for that bug? Nov 26 16:07:46 pb_, one sec Nov 26 16:08:30 pb_, http://bugs.openembedded.org/show_bug.cgi?id=5339 Nov 26 16:12:14 kristoffer: about 5343 libqpe-opie_1.2.4 fails...what's your buildhost? Nov 26 16:12:35 kristoffer: I think it probably wants tweaking to use PKGDEST rather than hardcoding ${WORKDIR}/install Nov 26 16:12:58 try making that change and see if the situation improves Nov 26 16:13:27 pb_, inside package_tar file? Nov 26 16:13:32 yeah, about line 59 Nov 26 16:13:38 ant_work, targetting mipsel. Nov 26 16:13:45 pb_, oki will do, thx Nov 26 16:13:50 and building on x86? Nov 26 16:14:04 ant_work, yeah Nov 26 16:14:17 ok, I see, thx Nov 26 16:17:27 http://hackaday.com/2009/11/25/ti-sports-watch-for-hacking/ Nov 26 16:21:37 morning Nov 26 16:25:24 I need help to solve this problem: The bitbake failed when tried to fecht http://snapshots.madwifi.org/madwifi-trunk/madwifi-trunk-r3837-20080802.tar.gz probably the link is bronken, Is there others links? Thanks Nov 26 16:32:01 rvfo, try changintg the url to snapshots.madwifi-project.org Nov 26 16:32:31 they don't have that old snapshot there though Nov 26 16:32:40 you'll need to bump it to http://snapshots.madwifi-project.org/madwifi-trunk/madwifi-trunk-r4094-20090903.tar.gz at least Nov 26 16:33:38 or you can create your own snapshot of r3837 from the svn at svn.madwifi-project.org Nov 26 16:35:27 th1_: What procedure have i do to do that? Nov 26 16:38:32 rvfo, without being an SVN expert I would have thought something like svn checkout URL... --revision 3837 Nov 26 16:39:48 specifically this seems to work: svn checkout http://svn.madwifi-project.org/madwifi/trunk/ --revision 3837 Nov 26 16:40:47 then after it's done rename trunk to madwifi-trunk-r3837-20080802 and create the tar file: tar czf madwifi-trunk-r3837-20080802.tar.gz madwifi-trunk-r3837-20080802 Nov 26 16:41:41 If i put the file in the OE sources folder will does it works? Nov 26 16:42:11 yes but you'll have to update the checksum in checksums.ini Nov 26 16:49:49 th1_ thanks, all it works Nov 26 16:49:54 great Nov 26 16:59:56 hi, I've issues with stable 2009 Nov 26 17:00:01 glibc says: Nov 26 17:00:12 http://pastebin.com/m60d59c00 Nov 26 17:00:28 I've rebuileded some glibc deps such as glibc-initial Nov 26 17:00:40 I've looked on the web and I found that it was related to the locales Nov 26 17:01:14 (the bug report was telling that the nls use flag was involved) Nov 26 17:02:00 hm actual you cant compile a glibc based without nls Nov 26 17:02:06 because glib woudnt build Nov 26 17:02:13 ok Nov 26 17:02:39 should I rebuild everything from scratch? Nov 26 17:02:55 should I divert from stable and try -dev the machine is the bug device Nov 26 17:03:04 I don't need java anyway Nov 26 17:03:29 hm this seems like an upgrade gcc or something Nov 26 17:03:32 only kernel and base config files,mplayer,pygtk+recent python,gdbserver Nov 26 17:03:34 build from scratch yes Nov 26 17:03:42 ok thanks Nov 26 17:04:22 btw how much of the bug device is in stable because I've python problems such as free, or segfaults problems Nov 26 17:04:28 and the version is 2.6.1 Nov 26 17:04:32 not the lastest Nov 26 17:08:49 hm Nov 26 17:08:53 no problems hier Nov 26 17:10:21 ok I'll try to send 1 patch to shr and discuss one patch and then I'll switch to -dev Nov 26 17:15:05 RP: I'm looking into a quite old commit of yours, where you set ALLOW_EMPTY for both dbg and dev packages Nov 26 17:15:21 does anybody knows why exactly both packages should always be created? Nov 26 17:15:44 normal dev dbg doc Nov 26 17:15:57 then depends static or locale Nov 26 17:16:05 sometimes for libpackages a -bin Nov 26 17:16:47 but you can override this in your package Nov 26 17:17:24 I know, but why set set that as default? Nov 26 17:17:45 it's just because I saw packages that are created and the original pkg doesn't actually exist Nov 26 17:17:48 just the pkg-dev Nov 26 17:18:10 but the pkg-dbg is also generated, and at the rrecomends we have 'pkg' there Nov 26 17:18:29 try bitbake -i Nov 26 17:18:39 peek packagename PACKAGES Nov 26 17:18:59 and if you try to install that you'll get an error, because the original pkg wasn't generated (doesn't have the allow_empty) Nov 26 17:19:02 woglinde: hm, ok Nov 26 17:20:20 * rsalveti still needs to learn how to explore bitbake -i :-) Nov 26 17:20:40 :) Nov 26 17:21:01 rsalveti: They can be empty but have dependencies Nov 26 17:21:01 woglinde: yep, all packages are actually there, but in the end just the dbg and dev are created Nov 26 17:21:18 the dev is ok, because I'm packaging just header files Nov 26 17:21:21 hm then lookinto the image dir Nov 26 17:21:25 rsalveti: Its so you can have chains of dependencies Nov 26 17:21:34 maybe only dev and lib/lib.so are installed Nov 26 17:21:53 woglinde: my package are just a bunch of headers Nov 26 17:21:56 like xproto, for example Nov 26 17:22:01 jepp Nov 26 17:22:13 then normaly only -dev will be generated Nov 26 17:22:18 because all other is empty Nov 26 17:22:27 woglinde: yep, but we have an allow_empty for dbg also Nov 26 17:22:32 in bitbake.conf Nov 26 17:23:06 RP: sorry, but I still didn't get the point of why we're generating the dbg package for all packages :-( Nov 26 17:23:28 rsalveti? Nov 26 17:23:33 rsalveti: So if you install foo-dbg, you have a way of pulling in all the symbols for a given package Nov 26 17:23:33 its per default there Nov 26 17:23:39 and if there is no .debug Nov 26 17:23:49 there will be not ipk created Nov 26 17:23:56 yeah, about line 59 Nov 26 17:23:58 oops Nov 26 17:24:03 hm Nov 26 17:24:08 area 59? Nov 26 17:24:22 heh Nov 26 17:24:34 * pb_ pressed "up-arrow, return" in wrong window Nov 26 17:24:43 RP: yep, but even for empty dbg packages? Nov 26 17:24:55 or it's just to handle dependencies with other packages? Nov 26 17:25:14 rsalveti: its to handle dependencies with other packages Nov 26 17:25:45 rsalveti: Thing about a tasks package, this makes the associated tasks*-dbg package make sense Nov 26 17:25:48 yep, thought that this was the reason Nov 26 17:26:34 RP: but for example, for xproto, currently all packages are generated, because there's an allow_empty inside the recipe Nov 26 17:26:44 so we have -dbg -doc -dev and pkg Nov 26 17:26:52 but only -dev have data Nov 26 17:27:04 is the allow_empty inside the recipe the best solution here? Nov 26 17:27:36 guess that without messing with core dependencies chain it's the only way Nov 26 17:28:04 Hi, I've some issues with stable/2009 and kernel config, anyone may help me, please? This is the log error http://pastebin.com/m71dda730 Nov 26 17:29:09 rsalveti: I thought it was the opposite, the default was to allow empty packages Nov 26 17:30:21 RP: only -dev and -dbg are created by default, even if doesn't have any data, but the others no Nov 26 17:30:40 so if you have any other package like xproto, and forget about adding allow_empty at the recipe, you'll get an error Nov 26 17:30:54 because the main package is not going to be generated Nov 26 17:31:16 rsalveti: I think in poky I changed the proto .inc file to deal with that Nov 26 17:31:20 and the -dbg will have the main package inside rrecomends Nov 26 17:31:26 yep Nov 26 17:31:42 oh, ok, not inside the recipe, but inside the .inc file Nov 26 17:32:15 pb_, didnt work. perhaps it should read %s/%s instead of %s/install/%s? Since that works for ipk? Nov 26 17:32:23 RP: just sound weird for me, but I got it now, thanks for the explanation Nov 26 17:32:36 rvfo hm with kernel 2.6.29 you shouldnt need madwifi Nov 26 17:33:07 kristoffer: what exactly is the change you made? Nov 26 17:33:28 you need to replace "${WORKDIR}/install" with PKGDEST Nov 26 17:33:32 woglinde: How can I fix that, because I just hit the command "bitbake task-base" and this made all Nov 26 17:33:39 it sounds as though you might just have replaced WORKDIR with PKGDEST, which won't do the right thing Nov 26 17:35:10 pb_, I added "pkgdest = bb.data.getVar('PKGDEST', d, 1) and root = "%s/install/%s" % (pkgdest, pkg) Nov 26 17:35:20 woglinde: The only configuration that i put in the DISTRO=minimal is PREFFERED_PROVIDER=kernel.2.6.29, and the bitbake did that Nov 26 17:35:43 kristoffer: right, so you still have too many "/install" components Nov 26 17:35:44 pb_, earlier it was root = "%s/install/%s" % (workdir, pkg) Nov 26 17:35:59 pb_, thats why I asked it it should only read %s/%s Nov 26 17:36:05 yes, it should Nov 26 17:36:16 pb_, goodie, Ill give that a go Nov 26 17:38:04 RP: yeah, removing that 2.4 h3600 support is probably fine, I don't imagine even the ipaq people are using that anymore. Nov 26 17:38:07 rvfo hm not sure where madifi was I think in the task Nov 26 17:38:13 remove it theire Nov 26 17:38:19 ~lart ti Nov 26 17:38:20 * ibot says "boot to the head" and knocks ti over Nov 26 17:38:25 I want the gles libs Nov 26 17:38:26 now Nov 26 17:39:02 woglinde: thanks, i will try Nov 26 17:39:45 bye Nov 26 17:46:40 jo ph5 Nov 26 17:47:08 hej woglinde Nov 26 17:49:08 hi eFfeM1 Nov 26 17:53:59 hi all Nov 26 17:54:16 (triple tasking) Nov 26 18:04:16 pb_, looks like that did the trick. Should I push it or you? Nov 26 18:04:29 kristoffer_: please go ahead and do that Nov 26 18:04:58 pb_, roger. Btw you know of any good way (aside from -c clean x; bitbake x) to rebuild the tar packages that Im missing? Nov 26 18:05:38 hm mplayer was next victim for micro Nov 26 18:06:26 kristoffer_: you could try "rm stamps/*/*.do_package", that might do the trick. Nov 26 18:06:43 I'm not entirely sure how the stamps work nowadays though. Nov 26 18:06:50 pb_, will give that a go, thx Nov 26 18:07:57 Anyone may help me, please? Im few day trying to compile a simple linux (console and ssh features) and I dont know I the command "bitbake insist to compile all Xorg and Gnome, I dont know what to do, for help my configuration files is here Machine: http://pastebin.com/m4ca5b13b Distro: http://pastebin.com/m5aba13d4, thanks Nov 26 18:08:38 kristoffer_ bitbake -c foo -f Nov 26 18:08:39 f == force Nov 26 18:12:58 woglinde, wont it recompile the packages then? Nov 26 18:13:09 nope Nov 26 18:13:18 only you they havent before Nov 26 18:13:25 args Nov 26 18:14:18 woglinde, oki great, thx Nov 26 18:15:52 rvfo: at a guess, because you have bluetooth enabled somewhere in your configuration Nov 26 18:16:05 * * OE Bug 5339 has been RESOLVED (FIXED) by kristoffer.ericson(AT)gmail.com Nov 26 18:16:07 * * package_tar no longer produces packages Nov 26 18:16:09 * * http://bugs.openembedded.net/show_bug.cgi?id=5339 Nov 26 18:16:50 03Kristoffer Ericson  07org.openembedded.dev * re6c9a9740b 10openembedded.git/classes/package_tar.bbclass: Nov 26 18:16:50 Fix so package_tar yet again produces tar packages. Nov 26 18:16:50 This closes bug 5339. Thx to Pb for fix. Nov 26 18:16:50 Signed-off-by: Kristoffer Ericson Nov 26 18:17:31 pb_; How can I disable this? because if I remove the line DISTRO_BLUETOOTH_MANAGER = "bluez4", the bitbake will say that I have several providers and then they will force one Nov 26 18:26:56 woglinde, getting this Nothing to do. Use 'bitbake world' to build everything, or run 'bitbake --help' Nov 26 18:27:25 bitbake -c build -f busybox Nov 26 18:27:47 ah Nov 26 18:29:21 that worked well Nov 26 18:33:28 hello. Nov 26 18:34:36 "configure: error: C compiler cannot create executables" how do I fix this? Nov 26 18:47:05 dies_felices start from scratch Nov 26 18:47:08 rm oetmp Nov 26 18:47:46 Sadly its not that simple. Nov 26 18:48:10 it works big endian but not little endian. Nov 26 18:49:40 hm that you didnt say before Nov 26 18:50:33 re mickeyl Nov 26 18:50:43 hi Nov 26 18:50:57 No, I didn't mention that I've had this happen in the devshell either but that is relivent now too. Nov 26 18:53:19 On the plus side, my recipe now does what I want (albeit big endian). Nov 26 18:56:10 okay Nov 26 19:01:43 btw when I was using devshell that was trying to be big endian at the time. Nov 26 19:02:00 Indeed I'll try it again see what happens with little endian. Nov 26 19:02:53 Nope same result, sorry. Nov 26 19:08:11 03Sebastian Krzyszkowiak  07org.openembedded.dev * rd95e0f1886 10openembedded.git/conf/distro/include/shr-autorev.inc: Nov 26 19:08:11 shr-autorev: add opimd-utils to AUTOREV Nov 26 19:08:11 Signed-off-by: Klaus Kurzmann Nov 26 19:20:56 re florian Nov 26 19:30:55 hey, i've set up oe according to www.gumstix.net Nov 26 19:31:28 i use git repro from git://gitorious.org/gumstix-oe/mainline.git Nov 26 19:31:46 i've just run bitbake omap3-console-image Nov 26 19:32:34 and it sucked up 15GB of my Harddisk :( just for the tmp folder :( and just ~1400/3300 tasks are done Nov 26 19:32:50 can this be true? :S Nov 26 19:33:40 yes Nov 26 19:33:46 really? Nov 26 19:34:02 so in the end it will need 25GB just for one build? :S :( Nov 26 19:34:10 come on... :D :P Nov 26 19:34:13 you could lower it with Nov 26 19:34:21 yep... adding INHERIT += "rm_work" on conf will help Nov 26 19:34:31 kept my folder under 3GB Nov 26 19:34:33 INHERIT += "rm_work" Nov 26 19:35:06 okay... where does this option go? :S Nov 26 19:35:18 sorry, i'm using oe and bitbake for my first time :) Nov 26 19:35:23 local.conf Nov 26 19:35:29 thx :) Nov 26 19:36:39 are you familar with qt/embedded as well? Nov 26 19:36:54 mostly Nov 26 19:38:36 so if i manage to build a console image and it runs on my overo earth Nov 26 19:38:49 what do i have to do to add qt/embedded? Nov 26 19:39:04 just a bitbake qt4-ebmbedded and copy it to overo Nov 26 19:39:12 or how can i include it in a image? Nov 26 19:39:40 look at recipes/images Nov 26 19:40:01 you then should easily find out how to write your own package Nov 26 19:40:07 args image Nov 26 19:40:32 hehe, okay, i'll try... :D Nov 26 19:40:41 hi esben Nov 26 19:42:19 re Nov 26 19:59:50 re Nov 26 20:00:48 woglinde was oetmp a proper name or something like /path-to-oe/oe/tmp ? Nov 26 20:01:29 dies_felices yes the oetmp dir Nov 26 20:01:37 if you set it to /path-to-oe/oe/tmp Nov 26 20:01:46 than thats it is Nov 26 20:02:44 Okay, ta. Nov 26 20:03:30 but maybee its not the problem in your case Nov 26 20:03:45 if you are doing big and little endian stuff Nov 26 20:18:30 Well it seems that I only need little endian, I'm trying to build Dovecot to fit in opt on a Qnap nas. Nov 26 20:27:06 Can anyone tell me about the NSLU2 branch of openembedded? Nov 26 20:30:31 re Nov 26 20:34:44 I need to know (really really like to know) the versions of the kernel, glibc gcc, and binutils? Nov 26 20:40:15 bitbake -s. Nov 26 20:46:59 kergoth thanks. Nov 26 20:48:12 or bitbake -i Nov 26 20:48:22 peek virtual/libc PV Nov 26 20:55:58 woglinde thank you. Nov 26 20:56:55 I have to go now. Lets just see what tomorrow brings. Nov 26 20:57:30 03Klaus Kurzmann  07org.openembedded.dev * rac7ca9c49d 10openembedded.git/conf/distro/include/sane-srcrevs-fso.inc: Nov 26 20:57:30 sane-srcrevs-fso: bump rev for libframeworkd-glib Nov 26 20:57:30 Signed-off-by: Klaus Kurzmann Nov 26 21:32:05 ballot for new members has been send, people please send their ballots dont forget Nov 26 21:34:38 * rsalveti needs another machine to build oe, he can barely write on irc when oe is building Nov 26 21:35:22 guess that a better hard drive would get a better performance Nov 26 22:11:58 : try to build on tmpfs ( 4~6 gb needed) Nov 26 22:12:34 pwgen: yep, that's an idea, need to try that Nov 26 22:17:10 : used it only for the build temp dir then i think 4.5 gb sized tmpfs will be sufficient ( i builded shr-image within 2 hours ) Nov 26 22:18:49 pwgen: yes, I did reserve 4GB too. At the end 3 images were 2.5GB. Nov 26 22:19:13 initially I had problems with ccache on hard-disk, though Nov 26 22:19:53 cool Nov 26 22:21:32 use a lower BB value for build will also decrease the used ammount ( remember webkit und other needs a lot of compile disk-space) Nov 26 22:32:29 pwgen: can you help again with touchscreen on Zaurus? Nov 26 22:36:44 depends on problem ? Nov 26 22:38:27 pwgen: in kexecboot we have events but these are ignored :/ Nov 26 22:38:40 it should work on GTA Nov 26 22:38:41 mickey|bbl: ping Nov 26 22:40:31 ant__ i have to look at the actual state of kexecboot . so mine was running and didnt change if ( scrolling would be a good feature, cause one more partition and the screen is full) Nov 26 22:40:58 hm..how many did you sum up :D? Nov 26 22:41:39 i think 1 nand, 3 on sd 2 on hd > 6 Nov 26 22:41:53 ah, only one on nand :) Nov 26 22:42:19 you beat me with hd Nov 26 22:42:29 c7x0 no usb-host Nov 26 22:42:44 and small CF.... Nov 26 22:43:13 I got 2 on nand + 7 on SD + 2 on CF Nov 26 22:43:30 i misuesd my old c-1000 as cups/dns/netboot-infrastructuer server ( this with broken nand ) so now i can try diffrent bootloader on the c-3200 Nov 26 22:44:16 pwgen: scrolling was working some revs ago Nov 26 22:44:33 but ts input is ignored atm Nov 26 22:44:36 hd is providing debian and my fat oe with sdk Nov 26 22:44:39 just read Nov 26 22:45:05 *G* so i can try more partitions . nice Nov 26 22:45:20 pwgen: we recently added uImage(s).... Nov 26 22:46:41 .. u am usually konservative in changing bootloader. if its running i wont change it. but its seems worth to update. Nov 26 22:47:22 it is for people using u-boot Nov 26 22:47:38 no awckward keypress on boot Nov 26 22:48:32 to select boot device Nov 26 22:49:16 its simple to patch qui to boot initramdisk ( if this isnt already done ) , the it should be possible to kexecboot the gta ( when kexecboot is working ) Nov 26 22:49:38 someone tested it long ago Nov 26 22:50:47 linux-rp_2.6.26.bb, do_deploy) failed .. cannot stat `arch/arm/boot/zImage. strange Nov 26 22:52:04 hi dth Nov 26 22:52:07 hm...there was some recent change to kernel.bbclass Nov 26 22:52:40 yes..see koen's comments: Nov 26 22:52:40 * also fix task ordering due to kernel.bbclass changes Nov 26 22:55:05 i am thinking about a uge nice zaurus-image like the old pdaxrom images with modern illume desktop Nov 26 22:55:54 .. and icewm .. Nov 26 23:01:29 jo re mickeyl Nov 26 23:01:34 hiho Nov 26 23:01:36 florian: pong Nov 26 23:02:10 hey Nov 26 23:06:20 ant__: is see kexecboot now with system menu ... nice .. Nov 26 23:25:54 good nigh Nov 26 23:25:55 t Nov 26 23:26:09 nite florian **** ENDING LOGGING AT Fri Nov 27 02:59:58 2009