**** BEGIN LOGGING AT Fri Jan 27 03:00:04 2006 Jan 27 03:01:44 doppiaemme: Don't forget to add the PREFERRED_VERSION i posted earlier. It's not in the conf Jan 27 03:02:29 Ifaistos: is what I was asking you, 'cause I don't read those lines in your local.conf.... :-) Jan 27 03:02:41 doppiaemme: as i have put it in the epia.conf in my case Jan 27 03:07:30 someone know how to put one file into two packages? Jan 27 03:08:41 good morning all Jan 27 03:09:04 hi koen Jan 27 03:10:34 hrw|work: you can't put 1 file in two packages for good reasons (RCONFLICTS, RREPLACES, etc), but if you really want to do that, just copy the file and work around it in do_package or something Jan 27 03:11:58 koen: I did it in other way Jan 27 03:14:33 ttf-gentium_1.02.bb | 137 +++++++++++++++++++++++++++++++++++++++++++++------- Jan 27 03:21:25 I think i have found a bug in bitbake.... Jan 27 03:21:35 join the club ;) Jan 27 03:22:18 if you declare a specific version for a package and for some reason it can not find it or the package is "problematic" Jan 27 03:23:19 insted of stopping it continues be getting the "next availiable" package (i.e different version) but tries to set it up using the config parameters of the previus package Jan 27 03:24:12 someone use Gentium fonts from OE? Jan 27 03:26:12 hrw|work: whatfonts are they? Jan 27 03:30:50 XorA: http://scripts.sil.org/cms/scripts/page.php?site_id=nrsi&item_id=Gentium Jan 27 03:31:11 XorA: we have them iin OE and I just updated them Jan 27 03:32:08 hrw|work: I shall install, I never saw them Jan 27 03:33:51 ok. changes pushed Jan 27 03:34:24 Ifaistos: That isn't classed as a bug - it continues trying to use other providers by design Jan 27 03:34:44 Ifaistos: I agree the feature has sometimes questionable usefullness :) Jan 27 03:34:50 wow, hg is fast Jan 27 03:34:58 and svn is slow Jan 27 03:35:11 Well if i specify a certain version.... i have my reasons... Jan 27 03:35:13 Ifaistos: I don't understand what you mean by previous config parameters though Jan 27 03:36:00 Ifaistos: its only a preference, not a set in stone requirement. An option to disable that behaviour in bitbake might be nice though Jan 27 03:36:24 REQUIRE_VERSION ? Jan 27 03:36:36 koen: I was just thinking that :) Jan 27 03:36:57 djeez Jan 27 03:37:05 even monotone is faster as svn Jan 27 03:37:21 ~lart non-distributed scms Jan 27 03:37:24 * ibot beats non-distributed scms to within 2.54cm of his life Jan 27 03:37:28 That rules out svn :) Jan 27 03:38:06 RP: What i noticed is that when for example i was trying to complie perl-5.8.4 and could not find it Jan 27 03:38:32 my current order of preference is: monotone -> hg -> svk Jan 27 03:38:41 darcs is ruled out because it isn't buildable in OE Jan 27 03:38:49 RP: it used perl-5.8.7 instead but tried to run a config.sh in staging/perl-5.8.4 dir Jan 27 03:39:25 RP: which was not existing.... Jan 27 03:39:54 RP: let me find it in my logs Jan 27 03:39:58 Ifaistos: That sounds like a rather nasty bug :-/ Jan 27 03:41:57 maybe it uses perl-native-5.8.4 Jan 27 03:42:49 RP: No logs :( . I will try to duplicated it later and post the debug to give it a look Jan 27 03:43:24 if it used perl-native-5.8.4 that could explain it and would be a valid thing for it to try Jan 27 03:44:11 python uses a similar construction: 'old' -native is used by 'new' -cross.bb Jan 27 03:50:17 Is there a way to force bitbake to stop compiling if a prefered version/date is not found ? Jan 27 03:52:54 PR: Could you have a look at this -> http://pastebin.com/525417. It's for another problem Jan 27 03:54:33 RP: It looks like the cpan.class configure does not pass the cross-compile enviroment to perl Makefile.pl Jan 27 03:55:44 RP: In this case the perl modules Makefile tries to verify that a libxml2 exists by trying to compile/link to it Jan 27 03:56:11 RP: but it uses the hosts ld instead of the cross compiled one Jan 27 03:59:43 I'm trying to put my first .bb together: I inherit autools, which calls configure alright but I'm trying to find some way to prevent aclocal.m4 from getting blown away since the aclocal.m4 from the package in question has some custom macros in it Jan 27 03:59:57 Can anyone point me in the right direction? Jan 27 04:00:24 I'm willing to read, I'm just not finding what I'm looking for somehow... Jan 27 04:01:27 johnX: wish I could help but autotools is voodoo I do not know Jan 27 04:01:39 move it it acinclude.m4 in do_configure_prepend() Jan 27 04:02:09 autotools.bbclass deletes it for exactly that reason: custom, but mostly bogus macros Jan 27 04:02:33 (that's my recollection of pb_'s explanation) Jan 27 04:02:43 thanks Jan 27 04:02:59 I'll keep reading and hacking away at this Jan 27 04:03:18 if that doesn't work do a 'oe_runconf' in do_configure() Jan 27 04:09:37 ok that got it through the configure Jan 27 04:21:43 augh...the program, having no idea that it's cross compiled, decides to run a test at the end of a make Jan 27 04:21:58 turns out ARM binaries don't run on an AMD64 Jan 27 04:25:08 Ifaistos: Any package getting built should use the cross compiler.... Jan 27 04:27:42 Ifaistos: ld looks to be geting set to the cross compiling linker. The question is therefore why isn't perl using it Jan 27 04:28:33 johnX: normally we just comment out that sort of test, assuming that if it builds it runs Jan 27 04:31:04 XorA: I figured that I should do something like that...I just need to track down what in this huge Makefile is causing these tests to run Jan 27 04:32:51 XorA: I've treated autotools as a special kind of magic that I shouldn't meddle with so as a result this is my first time really digging into the build system of an autotools based package Jan 27 04:38:45 RP: I am not sure (or know) either why its behaving like this. I think i will try to patch the Makefile.pl to avoid the test all-together Jan 27 04:46:47 Ifaistos: That sounds like the best plan - I haven't time to look into specific problems atm I'm afraid Jan 27 04:47:48 RP: np ! I am trying to figure out if its my newbinesh to oe, my fat fingers or a "feauture" :) Jan 27 04:49:54 Here is the log for the problem of compiling a different version than the one requested Jan 27 04:49:58 http://pastebin.com/525464 Jan 27 04:52:16 Ifaistos: Run the same command but with -D -D -D Jan 27 04:52:30 Lets see why it doesn't want to build the other verison Jan 27 04:52:52 I suspect it wants python-core which the older version doesn't provide Jan 27 05:05:48 RP: I tar.zip it as it was like 500K. Do you have a place i can upload it ? Jan 27 05:06:56 RP: Yes this is problem. I have commented out the PROVIDES = "python python-core" Jan 27 05:07:42 RP: This is the reason it prefers the newer package, but still the build fails Jan 27 05:12:24 morning all Jan 27 05:12:46 hey do13_ Jan 27 05:13:02 hey koen Jan 27 05:15:18 hi dirk Jan 27 05:15:53 hey Marcin Jan 27 05:18:59 Ifaistos: Try adding that line to the other package Jan 27 05:19:09 Is PN an override in OE? Jan 27 05:19:14 hi dirk Jan 27 05:19:48 hi Richard Jan 27 05:28:25 koen: gnumeric fails: http://pastebin.com/525507 any idea? Jan 27 05:31:07 hrw|work: need gtk2-devel on host machine Jan 27 05:31:23 hrw|work: someday we meed gtk2-native in oe Jan 27 05:32:26 XorA: thx Jan 27 05:32:42 XorA: but its ewi.. koen use it too.. Jan 27 05:33:06 hrw|work: pardon? Jan 27 05:33:28 nevermind Jan 27 05:33:36 machine lack libgtk2.0-dev Jan 27 05:33:57 koen: ping? Jan 27 05:34:10 hrw|work: we need a native version of the tool that turns xpm files into gtk files Jan 27 05:35:20 * XorA sees screwiness in perl build as well Jan 27 05:40:53 gdk-pixbuf-csource? Jan 27 05:41:00 hrw|work: gtk-dev installed Jan 27 05:42:19 koen: gracias Jan 27 05:44:01 RP: which package ? Python-2.32 or 2.4.2 ? Jan 27 05:44:12 Ifaistos: Make sure that line is in both Jan 27 05:44:14 anyone want to find me a way to get from Napes in Italy to Brussels in Belgium for under 150 EUR? Jan 27 05:44:30 I have th following Problem: My Build System is a i686 and my target system is a i686. When bitbakes is configurating the package basepasswd is says that a crosscompiler toolchain should be used but it does not exist. Is there a way to force the system to build the crosscompiler toolchain or using the compiler of the buildsystem? Jan 27 05:45:34 Cobelius: there was something like VENDOR somewhere in confs Jan 27 05:45:44 RP: If its in both it compiles the 2.3.2 as it should Jan 27 05:46:00 Cobelius: look into conf/machine/netvista.conf Jan 27 05:47:29 Ifaistos: Ok, that's because 2.3.2 wasn't providing python-core. I'll make a note to fix that in OE Jan 27 05:48:07 The field TARGET_VENDOR? Jan 27 05:48:13 Cobelius: yep Jan 27 05:48:44 What is this field doing? there is no documentation for it. Jan 27 05:49:10 Cobelius: iirc change target_arch a bit for situations like yours Jan 27 05:49:17 koen: NOTE: package gnumeric-1.5.5-r0: task do_build: completed Jan 27 05:49:21 koen: thx again Jan 27 05:49:59 adding a recent gtk+-native is easier since gtk+-directfb is working again Jan 27 05:50:29 multimachine.inc should make -native stuff obsolete IMO Jan 27 05:51:12 koen: How does that work? Jan 27 05:51:35 one user sent me tosa keymap fixes ;) Jan 27 05:51:47 if it needs a native package -> MACHINE=native, build, reset MACHINE Jan 27 05:52:24 and put STAGING_NATIVE_BINDIR in front of PATH Jan 27 05:52:39 koen: Its a nice idea. I'm not sure we can make it work quite that easily though Jan 27 05:53:15 But is there an option to force bitbake to build to crosscompiler toolchain? Jan 27 05:53:16 I suspect the same Jan 27 05:53:32 even it the Build Arch is the same as the Target Arch? Jan 27 05:54:23 hrw|work: NOTE: package gnumeric-1.6.1: completed :-D Jan 27 05:54:40 XorA: I use branch Jan 27 05:54:53 * koen wonders whether CFLAGS_prepend = "-Os " in local.conf would make any difference Jan 27 05:54:56 now if only perl would compile Jan 27 05:55:50 monday.. I definitelly have to go to post office in monday Jan 27 05:56:08 hrw|work: tosa OZ3.5.4rc keyboard issues? Jan 27 05:56:21 do13_: yes Jan 27 05:56:26 koen: //usr/lib/ipkg/info/libgsf.postinst: 14: gconftool-2: not found Jan 27 05:56:34 koen: lack of rdepends? Jan 27 05:57:13 more a lack of gconftool-2 not being shipped, or ipkg running the postsinst in the wrong order Jan 27 05:57:27 do13: have you read it? Jan 27 05:58:13 Cobelius: The bug report? Yes Jan 27 06:00:01 do13_: http://pastebin.com/525530 looks good to you? Jan 27 06:00:07 hrw|work: I'll merge the tc6393 ohci code in my tree and after liams wm9712 fixes we can hopefully release a alpha tosa 2.6 image. Jan 27 06:00:51 hrw|work: gconf-dbus does contain /usr/bin/gconftool-2 Jan 27 06:00:54 do13_: ok Jan 27 06:01:20 hrw|work: but gsf indeed doesn't link against gconf Jan 27 06:01:27 koen: which I dont have installed Jan 27 06:03:35 !seen Luke-Jr Jan 27 06:03:36 Laibsch: Luke-Jr was last seen in # 1 week, 3 days, 5 hours, 10 minutes, and 51 seconds ago saying: right... does it still apply? Jan 27 06:03:56 wow, that is quite some time ago. Jan 27 06:03:58 lunch time... Jan 27 06:04:13 hrw|work: fixed Jan 27 06:04:22 hrw|work: OK. I can test this. 3 weeks ago I build a 2.4 image. Jan 27 06:04:28 the fix should go into .dev as well, but I can't test that Jan 27 06:04:39 do13_: would be great Jan 27 06:04:48 bbl Jan 27 06:05:01 hrw|work: dobru chut! Jan 27 06:06:29 03hrw 07org.oe.oz354fam083 * r2299b7d6... 10/packages/qpf-fonts/ (qpf-gentium_1.01.bb qpf-gentiumalt_1.01.bb): Jan 27 06:06:29 qpf-gentium: dropped as it is 'Modified version' according to new Gentium license Jan 27 06:06:29 - we probably could distribute it under changed name but OPIE can use Jan 27 06:06:29 TrueType fonts directly so no need for another QPF files Jan 27 06:06:29 taken from .dev Jan 27 06:06:33 03hrw 07org.oe.oz354fam083 * r2611a405... 10/packages/ttf-fonts/ (ttf-gentium_1.01.bb ttf-gentium_1.02.bb ttf-gentium/OFL.gz): Jan 27 06:06:36 ttf-gentium: updated to 1.02, license changed Jan 27 06:06:37 - license URL added Jan 27 06:06:40 - license content added to recipe Jan 27 06:06:42 - license added into packages Jan 27 06:06:44 taken from .dev Jan 27 06:06:45 03jbowler 07org.oe.dev * r7abd08f6... 10/packages/linux/ (3 files in 2 dirs): ixp4xx-kernel: updated beeper code from CVS in 2.6.15.1 Jan 27 06:06:49 03jbowler 07org.oe.dev * rb1c4312e... 10/packages/util-linux/ (5 files): Jan 27 06:06:50 util-linux: split out umount/swap/losetup commands in all Jan 27 06:06:55 - There already was a separate package for 'mount', there are now Jan 27 06:06:56 packages for umount, swaponoff and losetup - i.e. all the 'mount' Jan 27 06:06:59 tools except pivot_root. To ensure backward compatibility Jan 27 06:07:00 util-linux (the main package) RDEPENDS on the new packages. Jan 27 06:07:03 03hrw 07org.oe.dev * re262cbc1... 10/packages/ttf-fonts/ (ttf-gentium_1.01.bb ttf-gentium_1.02.bb ttf-gentium/OFL.gz): Jan 27 06:07:09 ttf-gentium: updated to 1.02, license changed Jan 27 06:07:11 - license URL added Jan 27 06:07:13 - license content added to recipe Jan 27 06:07:15 - license added into packages Jan 27 06:07:18 03hrw 07org.oe.dev * rfa6466b9... 10/packages/qpf-fonts/ (qpf-gentium_1.01.bb qpf-gentiumalt_1.01.bb): Jan 27 06:07:20 qpf-gentium: dropped as it is 'Modified version' according to new Gentium license Jan 27 06:07:22 - we probably could distribute it under changed name but OPIE can use Jan 27 06:07:25 TrueType fonts directly so no need for another QPF files Jan 27 06:07:35 hmm Jan 27 06:07:40 done CIA-4 ? Jan 27 06:07:51 I just committed a change to a local monotone branch Jan 27 06:08:13 now I would like to get the diff so I can use it in a bugzilla entry to get the main oe pdates Jan 27 06:08:32 do you have viewmtn? Jan 27 06:08:52 no :( Jan 27 06:09:37 monotone diff -r -r iirc Jan 27 06:09:43 ugh Jan 27 06:09:52 the r numbers are wacked :) Jan 27 06:09:53 03florian 07org.oe.dev * rf749c74f... 10/packages/gnome/gnome-vfs-dbus_svn.bb: gnome-vfs-dbus: Add svn package by Patrick Steiner. Jan 27 06:09:57 03florian 07org.oe.dev * r0deaec7d... 10/packages/gpe-filemanager/ (4 files): gpe-filemanager: Add 0.24 by Patrick Steiner, remove ancient versions. Jan 27 06:10:12 is viewmtn easy to set up? Jan 27 06:10:33 if you have mod_python+apache2, yes Jan 27 06:10:39 hmm Jan 27 06:10:41 where is it? Jan 27 06:10:54 http://grahame.angrygoats.net/viewmtn.shtml Jan 27 06:12:40 scm would be a good tag for that in del.icio.us? Jan 27 06:12:54 probably Jan 27 06:13:06 better than rcs at least :) Jan 27 06:15:31 hrw|work: libgsf-1.12.2-r1 should have correct RDEPENDS Jan 27 06:21:00 @hrw: it seems to work now, thanks a lot! Jan 27 06:23:42 03koen 07org.oe.oz354fam083 * r6dc8d443... 10/packages/libgsf/libgsf_1.12.2.bb: libgsf: rdepends on gconf and gnomevfs Jan 27 06:32:06 03koen 07org.oe.dev * r60b102cf... 10/packages/libgsf/libgsf_1.13.3.bb: libgsf: add untested fix for RDEPENDS Jan 27 06:35:24 koen|away: thx Jan 27 06:41:30 ~lart perl Jan 27 06:41:30 * ibot keeps mailing perl free America Online CDs until he drowns Jan 27 07:05:09 hello! Watching my openzaurus-3.5.4-RC first time boot up screen Jan 27 07:06:14 hello! Watching my openzaurus-3.5.4-RC first time boot up screen ... it doesn't seem to be doing anything though... Jan 27 07:07:00 Installing �(*��gnumeric (1.6.1-r0) to root... WTF? Jan 27 07:07:34 Not sure its the first time I've installed the 3.5.4-RC Jan 27 07:07:41 I guess that's why its an RC Jan 27 07:08:28 what pda are you using? Jan 27 07:21:35 XorA: update ipkg to .157 Jan 27 07:36:33 what do i need in a conf file to get all the depends built, have trouble with build gpe-image .. missing packages Jan 27 07:39:10 is ipkg .157 stable enough for .oz354fam083? Jan 27 07:42:32 koen: did not tested Jan 27 07:43:41 22 21:27 < hrw> Pigi: which ipkg you suggest to use for stable: .154 or .156? Jan 27 07:43:43 22 21:28 < Pigi> I think 156 should be ok, but if pb_confirm the new mods, then I will release a new 157 that would speed up a bit something. Jan 27 07:45:14 so it looks like .157 should be ok Jan 27 07:54:38 koen: .155 should be avoided, .156 works, .157 is nicer then .156 because it does not parse lists when does not have to Jan 27 07:55:10 I thought so Jan 27 07:55:50 ok. time to go - 17:17 have train Jan 27 07:57:15 cu monday probably Jan 27 07:57:54 cu Jan 27 08:04:10 03hrw 07org.oe.dev * rfe66a242... 10/packages/initscripts/ (initscripts-1.0/tosa/keymap-2.4.map initscripts_1.0.bb): Jan 27 08:04:11 initscripts: fixed Tosa keymap-2.4 Jan 27 08:04:11 - fix was sent by one of OZ 3.5.4 testers Jan 27 08:06:27 hi all Jan 27 08:08:21 hey CoreDump|home Jan 27 08:09:32 hrw|gone: got two more annoying bugs in 3.5.4 GPE for you =) In GPE, the "Screen" and "Sleep" configurations apps segfault on launch on Akita. Jan 27 08:10:16 pb_, kergoth: I get the feeling it might be bad for some reason but why isn't PN in OVERRIDES? Jan 27 08:10:39 PN isn't? Jan 27 08:10:51 so how does SRCDATE_foo work? Jan 27 08:11:18 koen: It adds itself to overrides temporarily Jan 27 08:11:27 aha Jan 27 08:11:49 I remember someone explaining why it was bad, but can't remember the details Jan 27 08:11:58 ~lart udev Jan 27 08:11:59 * ibot duct-tapes udev to the floor and drools on him Jan 27 08:13:16 As far as I can see, it'd be quite useful. So useful, it must be bad somehow :) Jan 27 08:13:59 I can imagine bitbake would get *a lot* slower Jan 27 08:14:40 koen: Less calls to update_data so it might actually go faster :) Jan 27 08:15:27 ~lart libdbus-glib-1.so.0 Jan 27 08:15:28 * ibot stamps libdbus-glib-1.so.0 on the forehead with the official Troll marker Jan 27 08:27:55 morning all Jan 27 08:28:14 hey lrg|home Jan 27 08:28:20 * lrg|home is back after doing a lot of customer support Jan 27 08:40:14 hey Liam Jan 27 08:41:28 hey Dirk, are you enjoying your holiday ? Jan 27 08:41:49 lrg|home: Yep. But one week is too short. Jan 27 08:42:06 morning Jan 27 08:42:18 hey chouimat Jan 27 08:43:30 do13: I'm going to start on tosa on Monday, I've had a *lot* of support recently and haven't been able to do much. I'll try and do as much as I can before I head off to China Jan 27 08:45:18 lrg|home: thanks. You are going to china on the 3. Feb? Jan 27 08:46:11 do13: yep, going until the 13th. will be working on pxa stuff whilst I'm there, although will be behind firewall. Jan 27 08:46:43 does anyone know if #oe is available in China ? Jan 27 08:47:27 ahem, lol? Don't tell me they even censor IRC channels? Jan 27 08:47:53 CoreDump|home: I'm not sure. I know they censor BBC news Jan 27 08:48:11 i had no problems https-proxying out there Jan 27 08:57:59 morning Jan 27 08:59:41 hey alan|home Jan 27 09:00:00 hey koen Jan 27 09:00:37 cheers Jan 27 09:01:12 hi mickeyl Jan 27 09:01:13 looks like i'm not the only one who slept for too long, huh mickeyl ? Hi, btw... Jan 27 09:01:39 hehe Jan 27 09:01:41 hi. is there a recommened compiler version to use when compiling opie-image bootstrap etc? gcc 4 seems too new. Jan 27 09:01:56 patipaq: i recommend 3.3.4 Jan 27 09:01:58 no Jan 27 09:01:59 3.4.3 even Jan 27 09:02:08 i have some resentiments against 3.4.4 Jan 27 09:03:21 great thanks i can drop in 3.3.6 easily in ubuntu, i'll try it Jan 27 09:03:31 familiar and openzaurus both have sane defaults Jan 27 09:03:49 sane ? :) Jan 27 09:03:50 mickeyl: don't you know that 2.95.3 is the C++/qtopia walhalla when it comes to compilers? Jan 27 09:04:38 mickeyl: I booked the same flight to Brussels. So we can arrange a meeting at the airport. Jan 27 09:05:10 cool, more hackers at fosdem Jan 27 09:05:57 do13: ah excellent. which hotel are you in? Jan 27 09:06:54 mickeyl: You proposed the astridhotel :) Jan 27 09:07:06 do13: i remember. good, then we can share a taxi ): Jan 27 09:07:10 :) even Jan 27 09:07:30 mickeyl: yes Jan 27 09:07:37 mickeyl, do13: how far is astridhotel from novotel ? Jan 27 09:07:49 lrg|home: there are two novotels, which one are you in? Jan 27 09:08:32 lrg|home: Rue de la Vierge Noire ? Jan 27 09:08:52 * lrg|home is checking Jan 27 09:09:34 Centre Tour Noire Jan 27 09:11:04 do13: when can you be at the airport? do you have an etix? we could synchronize checkin then to get seats in the same row Jan 27 09:12:00 mickeyl, do13: the address is Rue de la Vierge Noir Jan 27 09:15:05 mickeyl: Yes I have an etix. I'll arrive at 14:13 via train. Jan 27 09:15:22 sounds like novotel in the north, this is IIRC 20 minutes off from grand place Jan 27 09:17:37 mickeyl: ah, thanks Jan 27 09:18:25 do13: ok, good, i think i can be there at 14:45 (perhaps a couple of minutes later) - we can meet at the meeting point in terminal 1 and then do the checkin. Jan 27 09:19:38 mickeyl: Yep. This is ok for me. Jan 27 09:20:23 do13: cool. we can share mobile numbers. I'll send you mine a minute per email Jan 27 09:20:36 or do you already have it? Jan 27 09:21:17 mickeyl: I don't think so Jan 27 09:21:46 ok, sent Jan 27 09:23:36 answer sent Jan 27 09:24:24 mickeyl: Is this the right place? http://www.viamichelin.de/viamichelin/deu/dyn/controller/ItiWGPerformPage?E_wg=210506028lS19J106027172149170035ITIWG7g11133deu0074115f102g10112a10231103311000cQnJ1eGVsbGVz000kUGxhY2UgRHUgU2FtZWRp100o141151droo0033fn10533nav0102q10112k102d1103311000cQnJ1eGVsbGVz000uUnVlIERlIExhIFZpZXJnZSBOb2lyZQ100o141151e5mq0033fn1054m2g6010010010072006027039.004-1.00110001001001001001001003deu011011&pim=true&strStartAddress=Place+du+Samedi&strSt Jan 27 09:24:24 artCP=&strStartCity=Bruxelles&strStartCityCountry=311&strDestAddress=Rue+De+La+Vierge+Noire&strDestCP=&strDestCity=Bruxelles&strDestCityCountry=311&strStep1Address=&strStep1CP=&strStep1City=&strStep1CityCountry=311&strStep3Address=&strStep3CP=&strStep3City=&strStep3CityCountry=311&strStep2Address=&strStep2CP=&strStep2City=&strStep2CityCountry=311&dtmDeparture=27%2F01%2F2006&intItineraryType=1&intOneCountryCheck=true&unit=km&vh=CAR&conso=6&carbCost=1.00&devise=1.0 Jan 27 09:24:25 %7CEUR&devise2=Andere Jan 27 09:24:35 oops too long. Jan 27 09:25:24 heh Jan 27 09:25:44 mickeyl: Try this one: http://pastebin.com/525832 Jan 27 09:27:38 * obergix[work] going to fosdem also... will be glad to meet some of you Jan 27 09:27:49 is there an embedded room ? Jan 27 09:27:58 do13: yeah, the rue de la vierge noire is pretty long though Jan 27 09:28:28 obergix[work]: yeah. philip de swert (nokia/osso) is organizing it Jan 27 09:28:55 ok Jan 27 09:29:04 mickeyl: ok. But this should be np :) Jan 27 09:29:08 yea Jan 27 09:30:29 http://fosdem.org/2006/index/dev_room_embedded empty for 06 ? Jan 27 09:31:05 http://www.ucdot.org/articles/05/12/06/1528244.shtml Jan 27 09:31:17 don't worry, there's some time until the schedule is fixed Jan 27 09:35:20 any papers received ? Jan 27 09:45:07 no idea Jan 27 10:02:35 they received my paper :) Jan 27 11:11:40 I am having problem to figure out how the build enviroment parameters are set... Jan 27 11:12:31 for example i have a perl module that regardless of the compile enviroment settings trying to use /usr/bin/ld to test link a lib... Jan 27 11:12:54 which offcourse fails :/ Jan 27 11:13:30 diff is your friend Jan 27 11:14:50 and perl Makefile.PL is my enemy :) (at this point) Jan 27 11:15:56 The worst part that even if i patch the Makefile.PL to avoid the test... Jan 27 11:16:17 the modules gets build as libxml-libxml-common Jan 27 11:16:58 but when bitbake tries to build it as a libxml-libxml-common-native it fails Jan 27 11:17:50 and from what i understood its's trying to use /usr/bin/ld insted of the target ld it should have used... Jan 27 11:19:22 in general either it ignores the ld path set by bitbake scripts/configs or they are set wrong (for one reason or the other) Jan 27 11:20:47 The resuly looks like this -> http://pastebin.com/526008 Jan 27 11:23:50 <[lala]> hi, does anybody know which distro i can use with the xxs1500 machine config? Jan 27 11:48:17 hey, is the openembedded website down? Jan 27 11:48:48 tried for 3 days now Jan 27 11:49:20 Is there any template anywhere on how a libsomthing-native lib should be build ? Jan 27 11:54:44 Paranoir_: try oe.handhelds.org Jan 27 11:54:53 the dns bill didn't get payed Jan 27 11:55:55 has someone paid it yet? Jan 27 11:56:05 or some damn squatter grabbed it? :/ Jan 27 11:58:14 koen: is there a problem getting it renewed? I have connections with this registrar that might be able to help Jan 27 11:58:51 mreimer: kergoth is without internet connection after his move to arizona Jan 27 11:58:51 raduga: pretty sure we still have the domain. just dont have dns for it atm Jan 27 11:58:57 hey kergoth Jan 27 11:59:00 i have net at home now, but its wireless in a hotel Jan 27 11:59:01 hey Jan 27 11:59:02 heh Jan 27 11:59:04 expiration dates on domain look good Jan 27 11:59:14 kergoth, check out zoneedit.com Jan 27 11:59:19 I sue them for my stuff Jan 27 11:59:29 free for less than 5 dominas Jan 27 11:59:40 sue? Jan 27 11:59:54 use Jan 27 11:59:59 I type bad :) Jan 27 12:00:05 that is the Freudsche Versprecher? Jan 27 12:00:29 zecke, my german is worse than my typing Jan 27 12:00:40 oe noob needs a little help Jan 27 12:00:58 having trouble rebuilding kernel Jan 27 12:01:26 always get 'Attempted builds: 0' Jan 27 12:01:45 where has it cached that this is already built? Jan 27 12:02:44 I think we need more data Jan 27 12:03:26 Crofton, you talking to me ? Jan 27 12:03:31 yeah Jan 27 12:03:34 ok Jan 27 12:03:34 sorry Jan 27 12:03:50 what did you run? Jan 27 12:03:59 i am using oe via the openslug makefile Jan 27 12:04:03 hmmm Jan 27 12:04:10 if you are not familiar, let me know Jan 27 12:04:14 basically, Jan 27 12:04:16 you may get better results in that channel Jan 27 12:04:23 its silent Jan 27 12:04:26 I forget the name Jan 27 12:04:33 i think this is oe anyway- listen Jan 27 12:04:39 please :) Jan 27 12:04:43 yeah, but we do things direct Jan 27 12:04:52 and the makefile acts as a shell Jan 27 12:05:01 i understand. Jan 27 12:05:23 Crofton: "Freudian slip" says my dictionary Jan 27 12:05:33 ah Jan 27 12:05:50 the makefile is trying to build kernel Jan 27 12:06:03 it does 'handling bitbake files' Jan 27 12:06:12 what bitbake command does it run? Jan 27 12:06:20 hold Jan 27 12:06:33 ./setup-env; exec bitbake virtual/kernel Jan 27 12:06:59 this seem correct? Jan 27 12:07:06 heh Jan 27 12:07:12 I'm not sure Jan 27 12:07:43 ok, where can i find the .bb that describes virtual/kernel i wonder? Jan 27 12:07:47 what happens when you type bitbake virtual/kernel Jan 27 12:07:57 without the exec? Jan 27 12:08:32 well, you need to look in the apprpopriate conf file to see how virtual/kernel gets mapped to an actul linuxxxxxxx.bb file Jan 27 12:08:41 yeah, I am grasping at straw ... Jan 27 12:08:52 hold Jan 27 12:09:09 kinto455: I was looking at this script the other day. Its setting up the enviroment params for bitbake to work and calls bitbake Jan 27 12:09:19 kitno455, I think you need to talk with someone familair with the openslug stuff Jan 27 12:09:27 if I need a specific usb to serial modulefor me kernel, what can I do? how can I compile my custom kernel? Jan 27 12:09:41 katossi- which module? Jan 27 12:09:49 pl2303 Jan 27 12:09:50 * zecke owns a dvd writer now Jan 27 12:09:53 katossi: add the module to the defconfig, rebuild the kernel, install the module Jan 27 12:10:02 zecke, birthday present? Jan 27 12:10:02 zecke: cool, inside your ibook? Jan 27 12:10:09 koen, thats what i am trying to do with openslug Jan 27 12:11:42 Ifaistos, Crofton everything in regards to the makefile and the setup scripts seems to work Jan 27 12:11:55 the error that i get is that bitbake says: Jan 27 12:12:21 Attempted builds: 0 Jan 27 12:12:51 and nothing is left in stamps, nothing is in work, nothing in deploy/images Jan 27 12:13:00 regarding that package Jan 27 12:13:14 wich is actually called nslu2-kernel Jan 27 12:13:26 sorry kitno455, I'm not super familiar with this, and I need to work on a report :( Jan 27 12:13:47 it seems that something has cached the fact that the package was built once, and it wont build it again Jan 27 12:14:09 bitbake -c clean virtual\kernel ? Jan 27 12:15:00 probably with the \ going the other way? Jan 27 12:16:12 rm -rf tmp/cache Jan 27 12:17:55 raduga, i already did that Jan 27 12:18:04 Crofton, that seems to have worked Jan 27 12:18:17 cool Jan 27 12:18:18 it is now rebuilding Jan 27 12:18:37 it was the bitbake clean command i was missing Jan 27 12:35:48 03florian 07org.oe.dev * r0aeab38c... 10/packages/gpe-conf/ (21 files in 5 dirs): gpe-conf: Add version 0.1.29, update cvs version and drop a pile of old releases. Jan 27 12:50:55 * france is away: Away Jan 27 13:09:15 whats the var to build all deps as i seem to be missing packages in most of the task-* ? Jan 27 13:12:39 so, what's the right way to add mod_php? should i make it an extra PACKAGE built right out of regular php? should it be a new BB file in php/ or should there be a mod_php/BB file/dir combo? Jan 27 13:15:50 <_chronic> words of advice! for young? people Jan 27 13:17:37 stay in school Jan 27 13:18:23 not Jan 27 13:45:25 pretty slow on fridays, eh? Jan 27 13:49:13 03koen 07org.oe.oz354fam083 * rce24b161... 10/packages/blueprobe/blueprobe_0.17.bb: blueprobe: add 0.17, needed for hx4700 and h6300 support Jan 27 13:53:25 okay i'm coming to FOSDEM :P Jan 27 13:53:32 dunno how I'm getting back home yet Jan 27 13:53:38 but I know I'm getting there Jan 27 13:53:48 mithro: hehe Jan 27 13:54:07 mithro: you need to get a job in europe ;) Jan 27 13:54:14 mithro: and save for the flight back home Jan 27 13:54:22 i have a flight back home Jan 27 13:54:33 it just leaves from Rome not Brussels :P Jan 27 13:55:00 try to sell it and get some drugs instead Jan 27 13:55:07 I'm currently going fast train to Milan and then flying from there to Brussels Jan 27 13:55:50 ~$50 EUR in total to get there Jan 27 13:57:36 gosh I hate linux Jan 27 13:57:50 nfs-kernel-server restart is not working... Jan 27 13:58:57 lol need to restart... Jan 27 14:21:06 * obergix[home] is away: I'm busy Jan 27 15:10:08 is there any luck we may see cbrpager ever ported to OE ? Jan 27 15:11:55 alan|: what is that? Jan 27 15:12:51 zecke: http://jcoppens.com/soft/cbrpager/index.en.php Jan 27 15:13:05 it is a comics cbr/cbz archive reader Jan 27 15:13:26 would be reaylly great on pda Jan 27 15:13:37 (gpe only, though) Jan 27 15:15:11 i've never been able to understand how you write a .bb file, but i guess this one shouldn't be too hard : it depends on progs that are already in OE Jan 27 15:15:20 * alan| worn out Jan 27 15:15:23 time to sleep Jan 27 15:15:35 good night zecke, good night all Jan 27 15:26:17 zecke: I find git way too chatty Jan 27 15:26:32 we can emulate it with monotone, btw Jan 27 15:26:55 alias git = 'cat /dev/urandom & monotone pull' Jan 27 15:27:23 and where is update-alternatives? Jan 27 15:27:31 who needs that Jan 27 15:28:13 since we obviously don't need any other gcc version as 2.95.3, we should also switch to the sharp symlink madness Jan 27 15:28:23 'compatible' Jan 27 15:28:48 koen: git feels hacky... it is the fastest moving, which is good and bad at the same time Jan 27 15:28:58 indeed Jan 27 15:29:04 koen: monotone is more mature as it has lua support Jan 27 15:29:20 koen: using your favorite three way merge tool is just a matter of implementing the hook Jan 27 15:29:33 hg looks pretty nice as well Jan 27 15:29:34 koen: for git you can write a shell script though Jan 27 15:30:22 lua? Jan 27 15:30:31 scripting language Jan 27 15:31:09 something is deeply fucked up with my vi... it adds the charachters where it wants... not where the cursor is Jan 27 15:32:40 koen: mercurial can be DoSed by starting a push and then hiting CTRL+C Jan 27 15:32:50 crap Jan 27 15:32:52 koen: this requires someone on the server to enter hg recover Jan 27 15:33:24 so it boils down to 'monotone still rules' Jan 27 15:33:46 koen: what worries on monotone is, it can not yet rosterify OE Jan 27 15:34:05 why not? Jan 27 15:34:08 koen: well I need to test that CTRL+C issue once again Jan 27 15:34:15 koen: due some kind of merge we did Jan 27 15:34:30 koen: that violates an invariant, some rev around 3000 Jan 27 15:34:41 doesn't http://venge.net/monotone/wiki/RootDirRenaming work around that? Jan 27 15:34:58 koen: people in #monotone didn't know that Jan 27 15:35:22 Necronom mentioned something about that this morning Jan 27 15:35:40 something about a new branch being able to rosterify OE Jan 27 15:36:04 koen: darcs/ghc will be fun as well Jan 27 15:36:14 koen: ghc on mac can not apply more than 150 at once... Jan 27 15:36:25 koen: which is roughly three days of OE work Jan 27 15:36:27 as long as ghc isn't in OE we don't have to bother testing it IMO Jan 27 15:37:23 well bootstrapping sounds like fun but I need to get some things sorted first Jan 27 15:38:06 emerge --sync is at 34% on my SHARk now.. (with 400mb swap on nbd) Jan 27 15:38:27 heh Jan 27 15:38:35 gentoo is LFS for dummies Jan 27 15:38:44 why not add shark support to OE? Jan 27 15:38:55 koen: it oom with 256 nbd Jan 27 15:39:09 koen: well that is the next step... I was lucky there was a stage3 gentoo for arm Jan 27 15:39:25 as debootstrap sucks donkey ass Jan 27 15:39:49 I was got debootstrap to fill my workstation's root with sparc stuff Jan 27 15:39:56 s/was/once/ Jan 27 15:39:56 koen meant: I once got debootstrap to fill my workstation's root with sparc stuff Jan 27 15:40:03 ibot: STFU Jan 27 15:40:05 extra, extra, read all about it, stfu is Shut the F*** Up!, or http://www.linuks.mine.nu/stfu-noob.jpg Jan 27 15:40:24 koen: well the '--arch' option of debootstrap is ridicoulous nowadays... Jan 27 15:40:41 koen: it tries to chroot and execute the dpkg binary of the root... Jan 27 15:40:59 hehe Jan 27 15:41:33 koen: I think git, mercurial, svn should be heavily tested Jan 27 15:41:51 and monotone 0.26 Jan 27 15:42:06 a recent 0.26pre at least Jan 27 15:42:18 monotone 0.27 Jan 27 15:42:31 we could do a reverse tailor for 0.25 -> rosters Jan 27 15:43:18 fuck the tunnel is down on my machine :} Jan 27 15:43:30 I can not rosterify over the weekend Jan 27 15:43:56 * pb__ returns from Paris Jan 27 15:43:59 hi all Jan 27 15:44:03 hey pb__ Jan 27 15:44:07 pb__: Bonjour, ca va? Jan 27 15:44:24 zecke: assez bien. toi? Jan 27 15:44:44 pb__: ca va bien aussie Jan 27 15:45:07 pb__: do you happen to have met with DibCom or ExpWay? :} Jan 27 15:45:32 * zecke bets all French Companies have a branch in Paris Jan 27 15:45:54 zecke: actually no. I guess I shouldn't say who I did meet with. Jan 27 15:46:10 right Jan 27 15:46:18 zecke: reading the scm wiki pages. remember that checkouts of the tree using svk should be faster than the equivalent checkout with svn. also note that your initial overview email says that svn doesnt support cherrypicking, yet you list an svk feature in the list. svk does support cherry picking Jan 27 15:46:36 * kergoth continues reading Jan 27 15:47:01 g'day kergoth Jan 27 15:47:08 hey Jan 27 15:47:09 kergoth: hey Jan 27 15:47:11 how goes it? Jan 27 15:47:27 * kergoth plays around with a vim7 snapshot Jan 27 15:47:33 kergoth: how is it named? smerge with just files? Jan 27 15:48:10 kergoth: and I can't remember my initial mail Jan 27 15:48:34 kergoth: and yes both svn co or svk sync should only be bound to network io Jan 27 15:48:49 please correct any mistakes you find Jan 27 15:49:31 zecke: iirc cherry picking is svk merge -c [rev] Jan 27 15:49:41 but i havent done it in a while Jan 27 15:50:05 http://svkbook.elixus.org/nightly/en/svk.ref.svk.c.cmerge.html Jan 27 15:50:07 right Jan 27 15:50:23 i need to play with mercurial and git some Jan 27 15:52:07 kergoth: for mercurial and darcs send me your ssh public key Jan 27 15:52:13 k Jan 27 15:52:22 kergoth: for git mail it to RP Jan 27 15:53:38 koen: do you know who might have vector art of the OE logo? Jan 27 15:54:12 ka6sox-office: everyone with an OE tree Jan 27 15:54:14 the svg is in the oe repo, iirc. at least it used to be Jan 27 15:54:16 heh Jan 27 15:54:32 thanks koen kergoth Jan 27 15:54:41 ka6sox-office: contrib/oe.svg Jan 27 15:54:48 http://kergoth.com/files/omni001.jpg .. the new vim7 omni completion is most nifty Jan 27 15:55:11 i especially like the python one, that can even tab complete the arguments to functions, dropping in the syntax from the pydoc info Jan 27 15:56:16 wow Jan 27 15:57:07 I'm stupid (haven't played admin since a long time). I have a directory and I want newly created files to be in a specific group Jan 27 15:57:11 how can I do that? Jan 27 15:57:32 umod? Jan 27 15:57:37 zecke: umask Jan 27 15:57:58 umask doesnt help the group, only the mode Jan 27 15:58:03 chgrp Jan 27 15:58:05 you probably want to chgrp the directory, then set g+s Jan 27 15:58:09 RP: umask on a directory? I thought it was a per user shell issue Jan 27 15:58:17 setgid directory causes new files to be in that group Jan 27 15:58:19 kergoth: so sticky bit is still working? Jan 27 15:58:26 works fine, i use it daily Jan 27 15:59:16 Whilst a few people that might know are around, I'll ask again. Why would putting PN into OVERRIDES be a bad idea? Jan 27 16:00:17 RP: would would happen to FILES_${PN}? Jan 27 16:00:26 might still work as expected Jan 27 16:01:02 yeah, all the subpackage overrides would be likely to break. Jan 27 16:01:24 I'm pretty sure that they did actually break last time we tried that, and there might have been a problem with namespace collisions as well. Jan 27 16:02:11 RP: Do you have a clue on how to revert changes? Jan 27 16:02:26 zecke: In git or montone? Jan 27 16:02:27 RP: git co oldrev; git commit; git push? Jan 27 16:02:31 RP: in git Jan 27 16:02:44 zecke: You can also do git revert I think Jan 27 16:02:53 RP: well okay ;) Jan 27 16:03:04 git revert creates a reverse diff which you would then merge and sign Jan 27 16:03:26 Would we want something like pn-${PN} in overrides? Jan 27 16:03:45 That would work. I couldn't speak as to whether we'd want it or not. Jan 27 16:03:48 RP: where would that be useful? for .inc files? Jan 27 16:04:10 zecke: for local.conf, I suppose. Jan 27 16:04:16 Think things like INHIBIT_PACKAGE_STRIP_pn-nano = "1" Jan 27 16:04:19 SABOTAGE_pn-opie-opie-opie = "yes" Jan 27 16:04:21 that kind of thing Jan 27 16:04:31 pb__: I think I can not support that Jan 27 16:05:09 RP: okay something I wanted to have as well Jan 27 16:05:20 zecke: ah, too bad Jan 27 16:06:05 I have a patch which adds PN to overides locally to support it for INHIBIT_PACKAGE_STRIP in packages.bbclass and it works nicely... Jan 27 16:06:21 but it strikes me as something that would have uses elsewhere Jan 27 16:06:38 03tmbinc 07org.oe.dreambox * r5d4b4c74... 10/packages/dreambox/dreambox-secondstage.bb: dreambox-secondstage/dm7025: upgrade to 42 to fix bug reported by BoxMan Jan 27 16:07:46 koen: you should change my location on your oe dev map :) Jan 27 16:09:04 zecke: minor note, i like using -l as well as -I on my smerges. have you used that before? Jan 27 16:09:20 kergoth: what are your new coords? Jan 27 16:09:43 how accurate do you want it? i'm currently in the phoenix, az area Jan 27 16:09:44 kergoth: no, (I'm new to svk anyway). I still wonder what --remoterevs does... Jan 27 16:09:45 :) Jan 27 16:09:55 -l [--log] : use logs of merged revisions as commit message Jan 27 16:10:10 iirc it makes the new commit messages reflect the messages from the original change Jan 27 16:10:20 s/change$/changes/ Jan 27 16:10:31 kergoth: oh well yeah I want that as well Jan 27 16:10:35 :) Jan 27 16:10:39 kergoth: as accurate as you want it to be Jan 27 16:11:05 koen: well, lets get it in the right state at any rate. i'm definately not in mn anymore :) move me down to phoenix Jan 27 16:11:24 i should play with oe some this weekend Jan 27 16:11:27 get back up to speed Jan 27 16:12:13 zecke: also, are you familiar with svk pull and svk push? Jan 27 16:12:16 and then beat up these people who've "broken" it all :) Jan 27 16:12:23 zecke: they're convenience wrappers around smerge Jan 27 16:12:32 kergoth: well I have used it (by accident) Jan 27 16:12:56 kergoth: I wonder what happens if we move the org.oe.dev dir on croftons server Jan 27 16:13:07 kergoth: and I try smerge... Jan 27 16:13:12 svk tracks branching, pull/push obeys it. so if i cp from mirror to another location, then push, it'll check in, smerge to the mirror, which automatically goes upstream Jan 27 16:13:25 pull does a sync on the mirror, then an update Jan 27 16:13:26 kergoth: you're in arizona now :) Jan 27 16:14:03 saves time. i rarely actually use smerge directly unless i'm doing manual stuff, i.e. --baseless Jan 27 16:15:03 kergoth: svk smerge --baseless -I took accumulated three days :} Jan 27 16:15:05 erm, i described that wrong. push doesnt check in, it just merges it to upstream if you're on a branch, ensuring your commits propogate Jan 27 16:15:08 haha Jan 27 16:15:13 well, thats a shitload of commits :) Jan 27 16:15:56 git took just a couple of minutes ;) Jan 27 16:16:05 impressive Jan 27 16:16:11 svk sync should be faster though Jan 27 16:16:23 unless you're downloading a pack from linux-mip through rsync, and then its fucking slowwwwww Jan 27 16:16:31 well as you've pointed out, there are advantages and disadvantages to all of the options Jan 27 16:16:33 but that has nothing to do with git Jan 27 16:16:48 CosmicPenguin: hehe, give them better Netowkr connection Jan 27 16:17:00 svk appeals to me though, i like that its based on something as known solid and stable as svn Jan 27 16:17:28 hmm Jan 27 16:17:28 kergoth: if smerge -s would work, it would rock Jan 27 16:17:30 I like it because I know svn Jan 27 16:17:41 kergoth: and if smerge really works with merging... Jan 27 16:18:20 Now is a good time for me to set up people to access the svn server ... Jan 27 16:19:17 and seriously hg looks really fast a 'local' operations Jan 27 16:20:19 I hate ACPI Jan 27 16:20:23 I hate it, I hate it, Ih ate it Jan 27 16:21:14 I think all the scm's have merit. I'm actually starting to appreciate some parts of monotone :-/ Jan 27 16:21:50 git's commandlines can be too insane. It all works but its not very intuitive... Jan 27 16:22:06 Whether a good guide could fix git or not, I don't know Jan 27 16:22:22 I think part of the test will be to see if we can write one... Jan 27 16:23:17 RP: yeah. Every time I sit in front of git I feel the power of it... Jan 27 16:23:30 RP: put I don't know what to type to get there where I want to Jan 27 16:23:48 zecke: with great power comes great responsibility Jan 27 16:24:01 zecke: I share that feeling. Its like when I came accross a bash shell after DOS :) Jan 27 16:24:03 koen: hehe Jan 27 16:25:10 Evening all. Finally have an apparent working cross-compile. But now I'm trying to re-create irssi and configure states a macro is missing. Jan 27 16:25:40 tkimball: welcome to the wonderful world autotools Jan 27 16:27:06 Ah, this sounds related to the recent changes. Should I get out of the pre-breakage pull? Jan 27 16:27:56 tkimball: that depends, but I think not Jan 27 16:28:22 tkimball: is it a autoreconf error Jan 27 16:28:24 tkimball: I doubt that will help much for that problem. Jan 27 16:29:36 zecke: Yes, it appears to be autoreconf. macro not found in library. Jan 27 16:32:23 'night all Jan 27 16:33:16 Ah, think I found it. *goes to check/install more pkgs* Jan 27 16:39:48 * koen tries a build with -Os in TARGET_CC_ARCH Jan 27 16:41:46 I hate cli applications... Jan 27 16:42:20 I hate gui applications Jan 27 16:42:36 CosmicPenguin: (I hate writing cli applications...) Jan 27 16:42:49 I wonder how much space -Os will save in a standard build Jan 27 16:42:56 anyway Jan 27 16:42:58 'night all Jan 27 16:43:03 cya Jan 27 16:54:34 How are autotools invoked from bitbake? I think m4_pattern_allow may not be setup correctly. Jan 27 16:55:24 tkimball: autotools.bbclass class autoreconf Jan 27 16:55:46 tkimball: see the do*configure* script in the temp/ directory Jan 27 16:56:06 03mreimer 07org.oe.dev * r7a7a7b77... 10/packages/gpe-bootsplash/gpe-bootsplash_1.15.bb: Jan 27 16:56:06 gpe-bootsplash 1.15: add new version, and fix a couple of error messages Jan 27 16:56:06 on boot by starting a little later. Jan 27 16:59:48 tkimball: it invokes the autoreconf built by oe which resides in staging, so the only m4 macros accessible to it are those in the source tree, and those which its dependent packages installed into staging. Jan 27 17:02:32 Ah. So irssi may have incorrect depends. Jan 27 17:04:57 tkimball: or irssi abused autotools Jan 27 17:05:19 also, an undefined macro error doesnt always indicate that the macro isnt available Jan 27 17:05:29 autoconf/m4 errors arent always what you think they are Jan 27 17:05:38 that one could be caused by m4 quoting problems as well Jan 27 17:05:39 zecke: Abused? Jan 27 17:05:41 iirc Jan 27 17:06:06 Oh, great. Now I'm curious how this package got compiled in the first place... Jan 27 17:06:20 tkimball: me too ;) Jan 27 17:06:23 ~pastebin Jan 27 17:06:26 pastebin is, like, a place to paste your stuff without flooding the channel - try http://pastebin.com/, or http://pastebin.ca, or http://channels.debian.net/paste Jan 27 17:06:26 autoconf is voodoo. Jan 27 17:06:55 zecke: This irssi session is from my slug, so I *know* it works. :) Jan 27 17:07:31 Anyway, hang on and lemme look at pastebin. Jan 27 17:08:00 another thing you'll learn about autoconf... people like to automatically enable features based on the existance of libraries in the system, unfortunately. highly non-explicit. results in builds which are decidedly not reproducable, since the autoconf macros inadverantly poke into /usr/lib on a regular basis Jan 27 17:08:05 bastards Jan 27 17:08:20 Gah Jan 27 17:08:22 build reproducability is one of the tenets of good CM policy Jan 27 17:08:26 heh Jan 27 17:08:53 quality in Free Software projects sadly suck anyway Jan 27 17:09:00 so what works on my build machine might not work on yours, if the .bb doesnt have all the necessary configure options to explicitly state things Jan 27 17:09:12 zecke: i think the whole "show me the code" attitude is very detrimental Jan 27 17:09:33 results in open source projects not having the necessary non-code people. CM/build, docs, test, etc Jan 27 17:09:37 unfortunate Jan 27 17:10:03 jesus, i need to get off the cvs-updates-kernel list Jan 27 17:10:08 hovland is flooding my box Jan 27 17:10:17 hehe Jan 27 17:10:20 OK. http://pastebin.com/526625 Jan 27 17:10:24 he flooded ähandhelds.org first Jan 27 17:10:27 hehe Jan 27 17:10:31 CIA? Jan 27 17:11:15 tkimball: that's odd. doesnt irssi depend on glib? Jan 27 17:13:11 Yes, but the compile options show it wanting glib2 I think. --without-glib1 --with-glib2 Jan 27 17:13:33 I'm using the svn .bb file, since the other one was not pulling. Jan 27 17:14:42 DEPENDS is including glib-1.2, which *sounds* correct. Jan 27 17:15:40 wait, we're enabling glib2, but depending on glib1? Jan 27 17:15:43 Gah. All this so I can get some of the features turned back on. Jan 27 17:16:14 Yeah, I just realized, going to go look for what glibs are avail on your listing. Jan 27 17:16:48 tkimball: annoying thing with m4/autoconf. if my configure script uses a macro from a dependent package, even a package i might end up disabling via a configuire option, you still need that macro around to regen the configure script. so you either need to depend on everything, even things you arent going to enable support for, or you need to include a local copy of those macros. Jan 27 17:17:14 RP: how many people applied for ssh accounts for git? Jan 27 17:17:20 thats one advantage to using pkg-config instead of the dependent package's m4 macros Jan 27 17:17:51 :w Jan 27 17:17:56 bah Jan 27 17:17:56 Ah. Ok, lemme try changing that to glib-2.0 and try again. Jan 27 17:18:03 two machines, i keep typing on the wrong keyboard Jan 27 17:18:44 wow - l33t Jan 27 17:20:23 I'm actually over three, but the other two are connected from the main desktop via ssh's. Jan 27 17:27:07 * CosmicPenguin is so happy since he got a news server running at work to read the LKML Jan 27 17:29:54 kergoth: Well, that got automake to stop complaining anyway. Now one of its C files have freaked out, but that I can deal with for now. ;) Jan 27 17:30:02 hehe Jan 27 17:30:04 one step at a time Jan 27 17:30:47 Curious - There a place I can collect changes for the bitbake compile - if they're fixes for bitbake/slugos and not irssi itself? Jan 27 17:31:41 As it stands, I'm gonna have to stop again and get a clean compile of irssi normally, to confirm. Jan 27 17:31:41 create patches, put them in an appropriate place, adjust the .bb to apply them Jan 27 17:31:57 err, misunderstood Jan 27 17:32:06 are you asking about fixes to the .bb in OE? Jan 27 17:32:10 bitbake is the tool, not the metadata Jan 27 17:32:53 If the problem is not irssi itself (i.e. I need to make changes to fix irssi to compile in ARM), is there a separate patch setup? Jan 27 17:33:21 make changes _to what_? Jan 27 17:33:30 irssi's c files, sorry. Jan 27 17:33:31 the irssi codebase? the compile commands in the .bb? what? Jan 27 17:33:44 right then, make a patch Jan 27 17:33:52 Nod Jan 27 17:34:14 OK, got enough to keep me busy tonight. Thanks. Jan 27 17:34:26 np Jan 27 17:50:00 good nite Jan 27 17:55:30 Can't find merge base for /oe_convert and /crofton-svn Jan 27 17:55:32 odd Jan 27 17:55:36 it should have merge tickets Jan 27 18:03:24 yep, unless you did the merge with --no-ticket, it shouldve recorded it Jan 27 18:03:42 kergoth: I did not specify that Jan 27 18:03:47 weird Jan 27 18:07:06 at least it sees empty merges... I wonder how it does that Jan 27 18:28:14 ah I think it is my error Jan 27 18:30:08 obviously Jan 27 18:30:31 zecke, parles toi l'arabe? Jan 27 18:30:52 sadly not Jan 27 18:31:01 "It's [OOP is] best seen as FP with grouping, privacy, and novel forms of abstraction." ... i like that Jan 27 18:31:13 zecke, cuz zecke in arabic is intelligent. Jan 27 18:32:17 that is good to know Jan 27 18:32:27 a spice of wisdom. Jan 27 18:32:58 zecke is german and means http://en.wikipedia.org/wiki/Tick Jan 27 18:33:53 but from now on I will claim it is arabic Jan 27 18:33:55 interesting Jan 27 18:42:57 good nite Jan 27 18:48:10 03tmbinc 07org.oe.dreambox * rd971d157... 10/packages/dreambox/dreambox-dvb-tools-v3.bb: dreambox-dvb-tools-v3: update to 1.2 to fix CC bug Jan 27 18:50:25 morning Jan 27 18:54:04 vening Jan 27 18:54:38 hi sonix` Jan 27 18:54:46 que pasa Jan 27 18:56:59 hmm.. I take it that's spanish for "how are things?"? Jan 27 18:57:37 well then respond, dont just look american there. Jan 27 18:57:57 heh Jan 27 18:58:11 I'm not american ;) Jan 27 18:58:39 I'm fine though. a bit busy and pretty tired, but fine. Jan 27 18:58:49 you? Jan 27 19:01:03 about to head out to keys player's new home. he needs help setting up his 48 channel mixer and what not. Jan 27 19:10:16 ah. good luck with that then Jan 27 19:10:26 ~lart sony Jan 27 19:10:26 * ibot tries to shut sony up Jan 27 19:11:23 wonder what the aibo league people will do without aibos :/ Jan 27 19:18:50 'night all Jan 28 00:14:35 goodmorning Jan 28 01:45:41 hi all Jan 28 02:44:32 hi Jan 28 02:52:57 the new gpe 2.7 is in the oe for familiar-unstable ? Jan 28 02:55:06 I think so Jan 28 02:55:14 iirc, it's also in familiar 0.8.3 Jan 28 02:58:04 there is familiar-0.8.3 (about stable) in oe ? **** ENDING LOGGING AT Sat Jan 28 02:59:57 2006