**** BEGIN LOGGING AT Wed Jul 06 02:59:57 2011 Jul 06 07:16:16 hi all ! :) Jul 06 07:18:50 morning alx_ Jul 06 07:26:04 celston, how to fight against oe compile errors ? :) Jul 06 07:26:45 Any particular ones? Or all of them? Jul 06 07:30:33 celston... some :) Jul 06 07:30:45 celston, trying to bake qt4e-base-image Jul 06 07:31:24 got error for mtd-utils-native, e2fsprogs-native and some others Jul 06 07:32:49 mtd-utils-native and e2fsprogs-native needs util-linux-ng-native to be built before them, but the last isn't on the compile queue Jul 06 07:33:25 did a bitbake util-linux-ng-native before doing bitbake qt4e-base-image and those recipes now compile fine. Jul 06 07:35:46 Looks like mtd-utils depends on util-linux *and* e2fsprogs Jul 06 07:36:10 util-linux-ng ... Jul 06 07:36:21 brb: it's morning: need coffee :) Jul 06 07:58:42 celston, back. how to remove a native recipe ? bitbake -c clean is enough ? :$ Jul 06 07:59:33 alx_: to clean it, yes - should be. Jul 06 08:00:02 although sstate might persist, I'm not 100% sure. Jul 06 08:00:18 celston, ehm... sstate ? I'm not using oe-core Jul 06 08:00:26 because I need mini2440 machine Jul 06 08:00:51 Ah, sorry. I assumed core. Jul 06 08:01:22 celston, ehm... well I'd wish to use oe-core but I don't find any doc on how to use it :( Jul 06 08:01:54 Given that it works if you run util-linux-ng-native manually, can you check if mtd-utils has util-linux in it's DEPENDS? Jul 06 08:02:07 alx_, did you look @ yoctoproject.org? Jul 06 08:02:24 yes: it has util-linux-ng as DEPENDS Jul 06 08:02:52 ka6sox, yes I saw it. but it was a bit confusing for me. Jul 06 08:03:33 celston, actually I'm running bitbake qt4e-base-image again and again, solving one trouble at a time. Jul 06 08:03:45 don't know if this is the right way to fix things :( Jul 06 08:04:00 Sounds right to me :) Jul 06 08:04:20 celston, well. one after me will face same troubles :) Jul 06 08:16:38 ka6sox, isn't yocto based on poky ? Jul 06 08:17:45 ka6sox, I need to build the image for different cpus Jul 06 08:18:16 more like http://wiki.openmoko.org/wiki/OpenEmbedded#BitBake_Overlays Jul 06 08:48:44 i've some troubles with dbus-glib-native Jul 06 08:49:14 do_configure tries to do | + install -m 0644 /home/alessandro/oe/build-mini2440/tmp/sysroots/x86_64-linux/usr/share/dbus/dbus-bus-introspect.xml /home/alessandro/oe/build-mini2440/tmp/work/x86_64-linux/dbus-glib-native-0.86-r2.1/dbus-glib-0.86/tools/ Jul 06 08:49:26 but /home/alessandro/oe/build-mini2440/tmp/sysroots/x86_64-linux/usr/share/dbus/dbus-bus-introspect.xml doesn't exists :( Jul 06 09:17:35 mmm how come dbus-glib seems SO broken Jul 06 09:17:41 (native) :) Jul 06 09:27:47 hello Jul 06 09:28:10 alx_, same problem here 2 minutes ago, I think one of us will find a solution Jul 06 09:29:06 I had created a x11-image using oe for ARM mini2440 board... but touchsreen is not working properly... can anyone help me... Jul 06 09:36:23 Cubi_, ok. the problem is that in dbus-glib.inc you have a do_configure_prepend which has to be done JUST for target and no on native Jul 06 09:53:43 does it exist virtclass-cross ? :) Jul 06 09:53:57 I found many examples of virtclass-native or virtclass-sdk Jul 06 09:54:02 but none Jul 06 09:54:05 of those Jul 06 09:59:23 Cubi_ are you working on it ? Jul 06 10:37:09 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * rf1a98ac032 10openembedded.git/recipes/images/x11-sato2-image.bb: Jul 06 10:37:09 Add a new x11-sato2-image for matchbox2 Jul 06 10:37:09 Signed-off-by: Denis 'GNUtoo' Carikli Jul 06 10:47:49 hmmm I must fix the identation of udev 168 Jul 06 10:47:54 I'll test and push Jul 06 10:59:43 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r1152dcf6a4 10openembedded.git/recipes/udev/udev_168.bb: Jul 06 10:59:43 udev 168: fix identation of do_unpack_append_bug Jul 06 10:59:43 Signed-off-by: Denis 'GNUtoo' Carikli Jul 06 11:00:24 gm Jul 06 11:13:44 alx_, No, I don't have the time for a good solution. I just commented out the 2 lines in do_configure_prepend() and after building I reverted my changes. Jul 06 14:24:53 morning kergoth__ Jul 06 14:29:36 hi pb Jul 06 14:31:46 hi woglinde Jul 06 14:35:07 hey Jul 06 14:39:12 jo kergoth Jul 06 14:52:24 Anyone know how to disable mirrors from a recipe so that it pulls directly from a git server? Jul 06 15:05:10 hm there was something Jul 06 15:05:13 let me think Jul 06 15:05:27 CVS_STASH soemthing Jul 06 15:06:03 # Specifies a location to search for pre-generated tarballs when fetching Jul 06 15:06:04 # a cvs:// URI. Outcomment this, if you always want to pull directly from CVS. Jul 06 15:06:04 CVS_TARBALL_STASH = "" Jul 06 15:06:43 hm didnt kergoth changed it to something with a mote usefull name now? Jul 06 15:10:55 woglinde, when / how can my git tree go into an angstrom mirror? Currently I'm disabling mirrors for testing my recipe Jul 06 15:11:27 woglinde, does that work for git too? Jul 06 15:12:26 yes Jul 06 15:12:32 it's for all scm's Jul 06 15:12:45 hi all Jul 06 15:12:48 thats why I said it was maybee renamed for oe-core Jul 06 15:12:54 and newer bitbake's Jul 06 15:13:12 ehm... how to write a rootfs to nand if ram is less than rootfs size ? Jul 06 15:13:22 joelagnel hence ask your master koen for angstroem mirrors Jul 06 15:13:34 woglinde, my master is jkridner :P Jul 06 15:13:45 I told you the short way Jul 06 15:13:56 you could ask jkrinder Jul 06 15:14:02 which will point to koen Jul 06 15:15:12 woglinde, thanks though. btw, are there scripts somewhere that mirrors use to keep a copy of things? Jul 06 15:17:24 dont know Jul 06 15:18:01 ok, np Jul 06 15:21:00 hmm Jul 06 15:21:26 I'm trying to build angstrom and I disabled iputils because it seemed a good way to prevent the doctools bullshit from impeding me Jul 06 15:26:27 is there a bindir-like variable for /usr/local/bin ? Jul 06 15:26:40 joconnoly? Jul 06 15:26:42 bitbake -e didn't give me much Jul 06 15:26:47 why /usr/local ? Jul 06 15:26:52 thats for not packaged soft Jul 06 15:26:54 someone's asking =/ Jul 06 15:27:01 *sigh* Jul 06 15:27:03 ;D Jul 06 15:27:11 no, there isn't Jul 06 15:27:20 ok, i'll rtfm him ;D Jul 06 15:27:25 yes Jul 06 15:27:27 fwiw, you can view conf/bitbake.conf in your favourite editor Jul 06 15:27:53 possibly easier than trying to inspect the output of bitbake -e Jul 06 15:28:02 good to know, pb_ thanks Jul 06 15:28:32 you can add one in your own distro.conf if you want Jul 06 15:29:10 yes Jul 06 15:29:22 but the someone should read again what /usr/local is for Jul 06 15:29:46 or should think about why he cannt put it into /usr/bin Jul 06 15:30:13 * kergoth__ hands jconnolly's "someone" a copy of the FHS :P Jul 06 15:30:47 yes, point taken. also, usually when bitbake complains about "files not shipped in any package", that's because they're not listed in FILES_PN right? Jul 06 15:30:53 well, maybe "someone" just wants to add the empty dirs/symlinks under /usr/local that the fhs requires to be present Jul 06 15:31:18 heh, double-quote someone Jul 06 15:31:21 which wouldn't be a totally unreasonable thing to want to do, I suppose Jul 06 15:31:23 it's not me! Jul 06 15:31:40 jconnolly: yes Jul 06 15:31:44 thx pb_ Jul 06 15:33:42 jconnolly, i told you to be cool Jul 06 15:34:10 oh jesus Jul 06 15:34:17 pile it on Jul 06 15:39:23 hi kgilner Jul 06 16:44:44 what if I patched a library and there exists a recipe for it ? Jul 06 16:45:02 what about it? Jul 06 16:45:14 i remember that there is a a bitbake sintax to "overlay" some part of an existing recipe. Jul 06 16:45:30 but actually I lost the refence to this syntax Jul 06 16:46:14 kergoth__, the idea is that i don't copy the original recipe, patch it and all the stuff. is it .bboverlay ? Jul 06 16:46:44 alx_: create a .bbappend. ensure its in BBFILES somewhere. its contents will geta ppended to Jul 06 16:46:51 e.g. foo_1.0.bb -- foo_1.0.bbappend Jul 06 16:47:38 kergoth__, this may overwrite variables ? Jul 06 16:47:48 again, its contents get appended to it Jul 06 16:47:54 its no different from concatenating them together Jul 06 16:47:57 so yes Jul 06 16:47:57 ok thanks :) Jul 06 16:48:02 np Jul 06 17:18:06 kergoth__, is it possible in an overlay do "require file.bb" where file.bb is in base oe recipes dir ? Jul 06 17:18:30 require obeys BBPATH Jul 06 17:18:48 if you want to include a recipe, you'll have to include it relative to the top of the layer's entry in bbpath Jul 06 17:18:53 e.g. recipes/foo/foo.bb rather than foo.bb Jul 06 17:19:09 bbpath gets adjusted to include the directory the current file is in, which is why you don't have to do that if they're in the same directory Jul 06 17:19:19 but clearly thats not the case here, which is why you h ave to handle it differently Jul 06 17:46:48 hm. kergoth__: I removed iputils from task-proper-utils to try and get around this docbook/sgml/etc crap that I really am not interested in... is there anywhere else that it's hiding that's less obvious? Jul 06 17:47:35 bitbake -g seems to indicate that virtual/libc and virtual/arm-angstrom-linux-gnueabi-gcc want it too, but that doesn't make sense to me Jul 06 17:47:47 doc generation is a gigantic mess right now, which is why many of the tools are required by the sanity checker. we haven't gotten all of the native tools added to all the deps of everything Jul 06 17:48:00 hmm, not that i know of, but -g would be the best bet Jul 06 17:48:03 or you could use -u depexp Jul 06 17:48:07 dependency explorer ui Jul 06 17:48:53 hmm, ok Jul 06 17:49:15 I see iputils listed in package-depends, pn-depends, and task-depends. I only edited the last one Jul 06 17:49:19 (what is pn- anyway?) Jul 06 17:49:38 pn is DEPENDS/PROVIDES, basically. build time / recipe level Jul 06 17:49:41 vs package and task Jul 06 17:49:56 hmm, ok Jul 06 17:50:10 and the "foo" -> "bar" is saying that foo requires bar to build, right? Jul 06 17:50:19 oh wait Jul 06 17:50:24 I'm looking at the dot file. idiot. me. Jul 06 17:54:09 blah Jul 06 17:54:35 I can't figure out why the hell overo-console-image says it has dependencies for iputils-* when I have them commented out in task-depends and none of hte other recipes lists them Jul 06 17:55:06 is that stuff cached, do I have to -o clean task-proper-tools or something? Jul 06 18:02:39 tasks must be cleaned when you change them, yes Jul 06 18:02:50 tasks aren't used directly, they emit ipk files which depend on things, and those ipks get installed Jul 06 18:02:56 if you don't bump pr or clean, it won't rebuild that ipk Jul 06 18:03:09 that is, task-*.bb, not tasks in the context of do_compile, etc Jul 06 18:03:21 the latter is of course what task-depends.dot is about Jul 06 18:03:37 gotta love overloaded terminology :) Jul 06 18:34:06 I wrote a bbappend file to apply a patch to a source: the bbappend has SRC_URI += "file://mypatch.patch". I put also PRINC="1" and FILESEXTRAPATHS := "${THISDIR}/files/". bitbake still try to find the patch just in the openembedded repository. any glue ? Jul 06 18:34:37 no clue what princ is Jul 06 18:34:38 dunno Jul 06 18:35:23 PRINC increments PR by 1 Jul 06 18:39:46 huh. i always just PR .= ".1" Jul 06 18:40:37 just inspired from http://sakrah.homelinux.org/blog/2010/11/bblayers-bbappend/ Jul 06 18:41:06 alx_: are you using latest bb Jul 06 18:41:13 gm kergoth__ Jul 06 18:41:24 khem, yes Jul 06 18:41:28 alx_: try FILESEXTRAPATHS_prepend := "${THISDIR}/files:" Jul 06 18:41:55 bluelightning, trying now Jul 06 18:41:59 :) Jul 06 18:42:20 bluelightning: FILESEXTRAPATHS := should work too Jul 06 18:42:26 unless its being set elsewhere Jul 06 18:42:32 khem: yes Jul 06 18:42:32 why prepend Jul 06 18:42:49 it ensures if it is set elsewhere, your bbappend gets in first, always Jul 06 18:43:03 what about FILESEXTRAPATHS += "${tTHISDIR}/files:" ? Jul 06 18:43:13 because it's at the start and appends/prepends get applied much later Jul 06 18:43:25 alx_: that will fail Jul 06 18:43:44 mmm ok Jul 06 18:43:48 and not just because there's a typo ;) Jul 06 18:44:09 just started bitbake :) Jul 06 18:44:48 alx_: the := is important, because ${THISDIR} is really only valid at the time that file is parsed... := tells bitbake to expand the value immediately Jul 06 18:45:23 (instead of the normal behaviour where it gets expanded on use) Jul 06 18:46:35 not working :( Jul 06 18:46:46 I'll pastebin :9 Jul 06 18:47:57 http://pastebin.com/Rdw0qRvc Jul 06 18:48:15 alx_: bitbake -e and grep for FILESEXTRAPATHS= Jul 06 18:49:05 yeah, that's a good way to debug these kinds of things Jul 06 18:49:42 khem, what if there is just FILESEXTRAPATHS="${THISDIR}/files:" in the grep ? Jul 06 18:50:13 er, if it is, are you sure you used := and not just = ? Jul 06 18:50:35 yes: as shown here : http://pastebin.com/Rdw0qRvc Jul 06 18:51:36 alx_: can I have the full output of bitbake -e qextserialport please? Jul 06 18:51:40 as a pastebin Jul 06 18:51:57 and I hope BBLAYERS is set properly too Jul 06 18:53:57 khem, this is something I'm not sure Jul 06 18:54:03 bluelightning, yep Jul 06 18:54:39 anyone know if SRCREV=${AUTOREV} works w/ git too? it worked with svn in the past Jul 06 18:55:27 jconnolly: it does, yes Jul 06 18:55:32 thanks kergoth__ Jul 06 19:01:12 hmm, I think we have a problem with insane.bbclass Jul 06 19:01:19 bluelightning, http://pastebin.com/Ld6Y4Zst Jul 06 19:01:24 it thinks that kernel-module-lirc-dev is a development package Jul 06 19:02:11 heh - that's a bit of a problem. Jul 06 19:02:37 I wonder if that can be fixed with some per-recipe overrides. Jul 06 19:02:57 mwester-laptop: hey Jul 06 19:03:10 Salutations Mr. Khem. Jul 06 19:03:12 did slugos/oe-core work for you Jul 06 19:03:19 Nope. Jul 06 19:03:22 Saluti Jul 06 19:03:25 oh Jul 06 19:03:30 Failed on fetching eglibc sources. Jul 06 19:03:40 hmm Jul 06 19:03:48 I haven't done any troubleshooting on that yet. Jul 06 19:03:49 are you behind firewall ? Jul 06 19:03:49 alx_: so where is your bbappend? Jul 06 19:03:56 (pathwise) Jul 06 19:04:09 mwester-laptop: it uses svn fetching for eglibc Jul 06 19:04:27 Let me see if I saved the error... Jul 06 19:05:33 bluelightning, not there. sorry. I tried to correct BBLAYERS but now it doesn't see the layer anymore. Jul 06 19:05:38 redoing it :( Jul 06 19:05:44 hm, anybody know where ~/.evolution moved to in the latest version? Jul 06 19:05:48 * pb_ tries to locate his email Jul 06 19:06:08 pb_: ~/.config/evolution? Jul 06 19:06:23 ( at a guess, I don't use it ) Jul 06 19:07:03 pb_ ~/.local/share/evolution/ Jul 06 19:07:08 ^^ Jul 06 19:07:15 this was easy to answer :) Jul 06 19:09:32 alx_: perfect, thanks Jul 06 19:09:42 khem: looks like it was a transient failure; it's fetching now. Jul 06 19:09:48 I'll let the build continue. Jul 06 19:09:53 mwester-laptop: cool Jul 06 19:10:11 mwester-laptop: are you on tip of the project Jul 06 19:10:17 No idea. Jul 06 19:10:46 I did the setup a couple of days ago on this VM, per the original instructions -- don't know what magic your makefile does! Jul 06 19:15:38 bluelightning, http://pastebin.com/RKkuThgV now it is as before. Jul 06 19:15:40 * pb_ quite impressed with latest evolution in fact Jul 06 19:15:57 it even seems to be using less memory nowadays. if only I had upgraded sooner I might not have needed that extra 8GB. Jul 06 19:16:19 bbappend is in ${OE_HOME}/com.biotechware.dev/recipes/qextserialport/; I put the patch in ${OE_HOME}/com.biotechware.dev/recipes/qextserialport/files/ Jul 06 19:16:26 brb Jul 06 19:16:48 khem: FYI, in oe-classic, I just created the branch slugos/slugos6.0-beta -- I'll be preparing for a release (finally). Jul 06 19:19:38 alx_: I don't understand it, AFAICT the := clearly has not worked Jul 06 19:20:16 i'd use bitbake -e to verify FILESPATH is what you think it is, etc Jul 06 19:20:27 check the assumptions Jul 06 19:21:30 bluelightning, :( Jul 06 19:22:16 hmm, now I'm not sure... I don't think classic OE supports FILESEXTRAPATHS Jul 06 19:22:25 or at least, I can't find any reference to it Jul 06 19:22:41 bluelightning, http://sakrah.homelinux.org/blog/2010/11/bblayers-bbappend/ Jul 06 19:23:18 khem: regarding that systemd thing, the point I was trying to make in my email was that I don't think empirical testing is the right way to approach the mkostemp() thing. Jul 06 19:23:37 unless systemd has a very good testsuite, which I doubt it does, just knowing that it "seems to work" is not really good enough. Jul 06 19:23:54 ok. put in trash FILESEXTRAPATHS. the question is how to patch a source from a bbappend ? Jul 06 19:24:20 pb_, when you get time could you verify you can write to the wiki? thanks! Jul 06 19:24:43 ka6sox: ah, sure, I'll try that soonest. how do I find out my login details? Jul 06 19:24:54 iirc, I had previously tried and failed to create an account Jul 06 19:24:59 actually I can't find you in the system. Jul 06 19:25:04 aha... Jul 06 19:25:09 let me work that right now. Jul 06 19:25:19 cool, thanks Jul 06 19:25:27 alx_: well, near as I can tell it is not supported in classic OE Jul 06 19:25:33 there is an alternative, hang on a sec Jul 06 19:25:40 bluelightning, it is used in yocto so it is oe-core Jul 06 19:25:49 yes, I know :) Jul 06 19:25:55 alx_: FILESPATHBASE_prepend := "${FILE_DIRNAME}:" or similar Jul 06 19:26:05 alx_: yeah what he said Jul 06 19:26:07 * kergoth__ didn't realize you were on stock oe Jul 06 19:27:11 hmm, khem didn't mention oe-core, poky or yocto in that blog post so it's somewhat misleading Jul 06 19:27:30 I see Jul 06 19:27:35 not that you can't do almost everything in that article with classic OE, it's just that one line that isn't correct Jul 06 19:27:50 but as a part of me wish to use oe-core, mini2440 isn't supported yet there. Jul 06 19:28:15 alx_: sure, well if you use FILESPATHBASE as suggested above (or similar) all should be well Jul 06 19:29:38 how to port mini2440 to oe-core ? Jul 06 19:29:39 ^^ Jul 06 19:32:24 ${FILE_DIRNAME} is resolved into openembedded repo :( Jul 06 19:42:15 bluelightning, what about http://chasemaupin.blogspot.com/ ? Jul 06 19:47:12 alx_: you can use the methods described there yes Jul 06 19:47:32 bluelightning, may my troubles came from BBFILES_PRIORITY ? Jul 06 19:47:37 no Jul 06 19:47:59 it's just a matter of getting this FILESPATH* stuff fixed up Jul 06 19:48:52 bluelightning, i dislike the soluition in the blog because it hardcode the path of the overlay Jul 06 19:49:41 alx_: ok, so stick with your current one... and do this Jul 06 19:49:49 in your bbappend: Jul 06 19:49:50 THISDIR := "${@os.path.dirname(bb.data.getVar('FILE', d, True))}" Jul 06 19:50:08 FILESPATHBASE_prepend := "${THISDIR}/files:" Jul 06 19:51:12 THISDIR didn't work before because it's another thing that's not present in classic OE; however you can just set it yourself Jul 06 19:51:33 and then use it (actually in this case it wouldn't matter what you had called it) Jul 06 19:52:50 :) Jul 06 19:54:41 :) Jul 06 19:55:49 bluelightning, cool :) this works :) Jul 06 19:55:57 hang.... Jul 06 19:55:58 alx_: great :) Jul 06 19:56:08 ok... do_compile :) Jul 06 19:56:42 let's see if today I'll have our prototype working :) Jul 06 19:58:17 bluelightning, thanks a lot :) Jul 06 19:58:54 alx_: no worries Jul 06 20:14:52 I love OE when you believe to have almost finished and you discover you need a little bit more .. Jul 06 20:50:06 GarthPS, hi Jul 06 20:50:10 does it work without that: Jul 06 20:50:12 + find . -name 'Makefile' -exec fix-makefile {} \; Jul 06 20:50:20 *still work Jul 06 20:52:15 GNUtoo: hmm that is right I don't remember trying wihtout it. I will try this Jul 06 20:52:27 thanks Jul 06 20:52:34 and next time try git-send-email Jul 06 20:52:57 unless you have a good reason not to use git-send-email Jul 06 20:52:57 I hate qwt :( Jul 06 20:54:51 dont use it Jul 06 21:00:12 GNUtoo: I have no send mail machin bidule chouette configured Jul 06 21:00:29 ok Jul 06 21:00:34 alx_: could I ask you why ? Jul 06 21:00:47 then no need to configure it Jul 06 21:00:51 I just do that: Jul 06 21:01:51 send-email --to="openembedded-devel@lists.openembedded.org" --smtp-server="out.alice.it" --smtp-server-port="587" --smtp-user="your_user" --smtp-pass="your_pass" --smtp-encryption="none" 0001* Jul 06 21:02:05 I could configure it too but I'm lazy Jul 06 21:02:30 you need to aptitude install the git-send-email deps tough Jul 06 21:02:51 i git-email - fast, scalable, distributed revision control system Jul 06 21:03:18 no need of an MTA Jul 06 21:03:25 just git-email Jul 06 21:03:34 and to add the config in the cmdline Jul 06 21:04:36 GNUtoo: send-email or senmail ? Jul 06 21:04:42 send-email Jul 06 21:04:48 sendmail is another program Jul 06 21:04:52 and it seem complex Jul 06 21:05:07 If I remember well you have to touch m4 to configure sendmail Jul 06 21:05:18 which not everybody is willing to do Jul 06 21:05:35 I also eared that most people don't know m4 well and have a well known working config Jul 06 21:05:44 but it's very vague in my head Jul 06 21:05:51 so I may be wrong Jul 06 21:05:58 GarthPS, trying to write a recipe for qwt6 Jul 06 21:05:59 s/vague/far Jul 06 21:06:11 what is qwt6 Jul 06 21:06:11 ? Jul 06 21:06:28 GNUtoo, a set of widget for plot in qt4 Jul 06 21:06:41 wow Jul 06 21:06:49 alx_: ho I did not know that the 6 version was out! :) Jul 06 21:06:51 * GNUtoo likes plotting like with gnuplot Jul 06 21:07:12 * GNUtoo is currently looking at a lot of temperatures.... Jul 06 21:07:14 GarthPS, it's quite a few... and we develop a small program on top of qwt6 Jul 06 21:07:32 but when it comes to bring it in OE...... Jul 06 21:07:34 what good programs do we have for plotting? Jul 06 21:07:42 gnuplot has an issue: Jul 06 21:07:49 GarthPS, all is broken Jul 06 21:07:50 It has a limit on the number of points Jul 06 21:08:10 GNUtoo, check qwt. it has mathml and 3d plots too :) Jul 06 21:08:19 but now ... Good night :) Jul 06 21:08:24 see you tomorrow Jul 06 21:08:27 ok Jul 06 21:08:31 see you tomorrow Jul 06 21:08:37 alx_: is there a release not about the 6.0 version ? Jul 06 21:08:37 the issue is that I know only gtk Jul 06 21:08:44 GarthPS, ? Jul 06 21:08:44 and remember the efl tutorial Jul 06 21:09:00 so I don't know QT Jul 06 21:09:09 alx_: s/not/note Jul 06 21:09:15 beside I'd like something usable in java if possible Jul 06 21:09:58 it's 26.840619516289006 degrees here.... Jul 06 21:10:49 I've some temperatures sensor(lm35) + a sensor module for my bugs so I'd like to plot that Jul 06 21:10:51 * mwester-laptop wonders if GNUtoo hasn't violated the basic rules of precision/accuracy... Jul 06 21:10:52 GarthPS, http://sourceforge.net/news/?group_id=13693 Jul 06 21:11:00 yes I did Jul 06 21:11:11 lol Jul 06 21:11:42 I don't know how much it's precise, I must re-read the datasheets Jul 06 21:11:47 I had an engineering professor who would take points off of each answer a student gave that had too many digits past the decimal point. Jul 06 21:11:49 GarthPS, if you need the recipe too we may work on this. I don't think to create a patch for a new recipe that don't work yet as intended. Jul 06 21:12:04 the ADC is 16bit so I guess I've to read only the sensor datasheet again Jul 06 21:12:26 :) Jul 06 21:12:43 GNUtoo, what about rounding to 16 bit ? Jul 06 21:12:45 lol Jul 06 21:12:49 well, I think mwester's point was that it might be precise without being accurate Jul 06 21:12:50 heh Jul 06 21:12:53 ± 0.4 Jul 06 21:13:23 that's the typical accuracy at 25 dergees Jul 06 21:13:37 so I guess I leave only 1 digit Jul 06 21:13:40 yeah, that's what I'd expect. Jul 06 21:13:41 it remember me for a car which had digital speed indicator which sensor has 1Kmh precision, while the display had 2 digits under it. Jul 06 21:13:43 good night Jul 06 21:13:47 anyone know why we're getting a really old gnu-config from CVS? Jul 06 21:13:50 alx_: yeah perhaps but for now I do'nt have time .. Jul 06 21:14:23 * incandescant notes they use git nowadays and have a fair few commits since the version we're using in oe-core Jul 06 21:15:54 incandescant, maybe it's because poky had that old stuff Jul 06 21:16:22 and they simply kept it in the migration and forget to update it Jul 06 21:21:10 GNUtoo, seems probable - I'll prep a patch and see what falls out on the list Jul 06 21:26:31 nitink, ping? Jul 06 21:31:37 hmm, I wonder if it's because git needs gnu-config... Jul 06 21:35:14 GNUtoo: ok build without the fi-makefile line so I wend you a new patch (trying with sen-email) Jul 06 21:35:53 ok nice Jul 06 21:42:17 27.8 degrees now Jul 06 21:42:30 the extras digit were because the app was a general purpose app Jul 06 21:42:48 and was quickely converted for temperature Jul 06 21:43:13 incandescant: pong Jul 06 21:44:57 nitink, was just wondering if you knew why we use cvs for gnu-config - but I'm beginning to think that it's because git requires gnu-config so we'd have a circular dependency if we used git to fetch it Jul 06 21:45:32 incandescant: right, I have tried it before Jul 06 21:45:59 sure would be nice to have a tarball so we don't have to build cvs for pseudo Jul 06 21:46:10 GNUtoo: sorry but I don't want to lose my time looking after my smtp user and password so it will be a patch by mail Jul 06 21:46:55 ok Jul 06 21:47:03 be sure to git-format-patch Jul 06 21:47:04 tough Jul 06 21:47:22 incandescent: you are right about circular dependency. AFAIR there was no good place to get the right version of the tarball Jul 06 21:49:02 nitink, okeys, I'll not waste too long looking then :-) Jul 06 21:49:20 GNUtoo: sent . tell me if it is ok. i will go to bed. Jul 06 21:49:30 ok Jul 06 21:51:11 it applies cleanly Jul 06 21:51:15 now let's look at the content Jul 06 21:51:24 GarthPS, bump PR Jul 06 21:51:44 GNUtoo: ouch.. Jul 06 21:51:48 INC_PR rather Jul 06 21:52:10 also are you sure it works Jul 06 21:52:25 you have echo there are too much quotes Jul 06 21:52:39 try echo 'FOO="bar" ' Jul 06 21:52:47 or try to escape the \ Jul 06 21:53:32 s/the/with/ Jul 06 21:59:41 GNUtoo: done. yeah I checked my echo Jul 06 22:01:50 ok let me look Jul 06 22:03:18 GarthPS, the echo works??? Jul 06 22:03:20 let me try Jul 06 22:04:03 GNu yes I did try and checked it with bb -c configure => took a look in the config.linux-cross and it was there Jul 06 22:04:09 GNUtoo: Jul 06 22:04:12 ok Jul 06 22:04:49 I'll try to build it then Jul 06 22:05:53 GarthPS, hmmmm Jul 06 22:06:08 you need to bump PR of vlc as well Jul 06 22:06:39 what is link against live555 apart mplayer and vlc? Jul 06 22:06:41 GNUtoo: why ? Jul 06 22:06:55 i don't know Jul 06 22:06:56 because vlc will keep failing otherwise Jul 06 22:07:03 ?? Jul 06 22:07:10 let me explain Jul 06 22:07:41 imagine there is a user Jul 06 22:07:47 that doen't know how to fix things Jul 06 22:07:51 or that doens't have the time Jul 06 22:07:59 like an angstrom or shr builder for instance Jul 06 22:08:03 he builds and that's all Jul 06 22:08:19 then at some point he tries to build vlc before your patch going in Jul 06 22:08:22 vlc fails Jul 06 22:08:29 then your patch comes in Jul 06 22:08:35 he retry building vlc Jul 06 22:08:53 live555 get updated and it builds Jul 06 22:09:06 and vlc doesn't get rebuild and output the old error log Jul 06 22:09:22 or at least doesn't reconfigure and such Jul 06 22:09:46 ObFix: clean and retry broken packages every once in a while Jul 06 22:10:11 yes if you clean it's ok Jul 06 22:10:20 but if you don't know you have to clean Jul 06 22:10:29 or don't have the time to clean it Jul 06 22:10:54 GNUtoo: nope it is ld time so no need to clean just retry to builmd Jul 06 22:11:03 ah? Jul 06 22:11:07 let me test it tough Jul 06 22:12:03 GNUtoo: it is why I explained my past attempts Jul 06 22:13:05 because without updating vlc it worked. because vlc detected that live555 was there without check wether or not it is with fPIC at config time. it is just at linking time that vlc failed. Jul 06 22:13:14 so normally it is ok Jul 06 22:13:27 * GNUtoo did a bitbake vlc and is waiting Jul 06 22:14:38 I've a core i7 Jul 06 22:14:42 so it shouldn't be too slow Jul 06 22:16:39 NOTE: package vlc-1.1.4.1-r4.1: task do_install: Started Jul 06 22:16:42 let me look Jul 06 22:21:30 GNUtoo: you test on what ? Jul 06 22:21:45 GNUtoo: at least it did build :) Jul 06 22:21:52 COMPILE_OPTS+= -fPIC Jul 06 22:22:02 yes it built and the COMPILE_OPTS is there Jul 06 22:22:10 but Jul 06 22:22:30 am I a horrible person? Jul 06 22:22:34 it lacks a space Jul 06 22:22:37 at least it seems Jul 06 22:22:50 in the echo Jul 06 22:22:56 COMPILE_OPTS+= Jul 06 22:23:01 wouldn't COMPILE_OPTS += Jul 06 22:23:03 be better? Jul 06 22:23:29 ah no it's fine Jul 06 22:23:33 there is RCE=1 Jul 06 22:23:38 so I think it's fine Jul 06 22:23:40 + it works Jul 06 22:23:44 so let's keep it that way Jul 06 22:23:49 I'll add my ack and push Jul 06 22:24:16 I hope it makes linterna magica works Jul 06 22:25:04 GNUtoo: what did you add ? Jul 06 22:25:17 you're looking for my ack right? Jul 06 22:25:23 so I add it Jul 06 22:25:23 (about the space I did have error when parsing recipe I think remember) Jul 06 22:25:24 and push Jul 06 22:25:41 ok Jul 06 22:25:49 it works without the space Jul 06 22:25:53 so let's keep it that way Jul 06 22:26:13 what is your ack ? Jul 06 22:26:19 you'll see Jul 06 22:26:36 GNUtoo: what is s linterna magica Jul 06 22:26:37 you know what signed-off is? Jul 06 22:26:59 GNUtoo: hmm not sure :) Jul 06 22:27:01 a greasemonkey script to see youtube videos without flash and without gnash Jul 06 22:27:12 hi all Jul 06 22:27:13 GNUtoo: cool Jul 06 22:27:22 GarthPS, hi again Jul 06 22:27:26 I failed to make it work with mplayer but it supports vlc Jul 06 22:27:45 GNUtoo: alx_I will go to sleep Jul 06 22:27:51 ok let me push Jul 06 22:27:56 and look at the push Jul 06 22:28:22 a signed-off-by means that you did the work or are authorized to pubilsh the work Jul 06 22:28:32 an ack is different Jul 06 22:28:39 it's for reviewers/maintainers Jul 06 22:28:47 you add that when you're ok with the changes Jul 06 22:29:02 what to do if I didn't receive any email from patchwork @ openembedded ? Jul 06 22:29:09 Is there a bug in OE that causes it to partially download files , possibly on a build failure ? Jul 06 22:29:44 Does this happen if we have a lot of threads? Jul 06 22:29:51 GarthPS: http://elinux.org/Developer_Certificate_Of_Origin Jul 06 22:30:52 03Sylvain 'GarthPS' ParĂ©  07org.openembedded.dev * r529b445f11 10openembedded.git/recipes/live555/live555.inc: Jul 06 22:30:52 live555 recipe: added -fPIC flag need by VLC to build with live555 support Jul 06 22:30:52 Signed-off-by: Sylvain 'GarthPS' ParĂ© Jul 06 22:30:52 Acked-by: Denis 'GNUtoo' Carikli Jul 06 22:30:56 kergoth`: cool thx I will read this tommorow :) Jul 06 22:31:30 your patch is unchanged but the explanation has one line telling that I was ok with the changes Jul 06 22:31:33 the acked-by Jul 06 22:31:47 * GNUtoo boots the n900 Jul 06 22:31:56 to try vlc Jul 06 22:32:09 GNUtoo: oh yeah ok . but I was right to put signed-off-by right ? Jul 06 22:32:10 with linterna magica Jul 06 22:32:14 yes Jul 06 22:32:23 you were wrong if you forgott it Jul 06 22:32:26 since you made that patch Jul 06 22:32:29 GNUtoo: ok so thx for the push Jul 06 22:32:31 there is no issues at all Jul 06 22:32:37 thanks for fixing Jul 06 22:32:49 GNUtoo: I am glade Jul 06 22:33:07 if I can have youtube and vlc it would be great Jul 06 22:33:19 vlc to show people that usual free software can run on my phone Jul 06 22:34:16 :) Jul 06 22:34:30 GNUtoo: configrme me and I am gone for tonight Jul 06 22:34:46 configrme ??? Jul 06 22:36:46 GNUtoo: confirme me and I am gone for tonight Jul 06 22:37:20 it's installing ton of stuff Jul 06 22:37:24 but I guess it will work Jul 06 22:38:06 * GarthPS waiting..GNUtoo :p Jul 06 22:38:20 it's installing the samba dep Jul 06 22:38:38 * GarthPS under the pre2 vlc make the device crash and reboot... He needs to track this down.. Jul 06 22:38:43 lol Jul 06 22:38:47 ahh ok Jul 06 22:38:50 I know why Jul 06 22:38:55 armv7's neon Jul 06 22:38:57 GNUtoo: about what ? Jul 06 22:39:10 I guess it's NEON Jul 06 22:39:19 because why would: Jul 06 22:39:21 mplayer Jul 06 22:39:22 vlc Jul 06 22:39:24 GNUtoo: waht and what is neon ? Jul 06 22:39:26 efl Jul 06 22:39:29 crash the phone Jul 06 22:39:32 SMID Jul 06 22:39:36 like sse on x86 Jul 06 22:40:23 maybe there is some assembly that makes it crash Jul 06 22:40:24 GNUtoo: can you explaine me that please ? lc is using a wrong flag for pre2 amr architecture ? Jul 06 22:40:27 or something like that Jul 06 22:40:38 I was just guessing Jul 06 22:40:50 GNUtoo: ok what is SMID ? Jul 06 22:40:58 it's rather the pre-2 kernel that is broken if it's the case Jul 06 22:41:10 Single Instruction Multiple Data Jul 06 22:41:13 like sse on x86 Jul 06 22:41:15 GNUtoo: I vote for tha Jul 06 22:41:16 t Jul 06 22:41:44 basically efl started to crash the phone starting from a revision Jul 06 22:41:49 and....guess what Jul 06 22:42:09 they introduced some armv7 NEON assembly near that unknown revision Jul 06 22:42:10 GNUtoo: I need to catch you one day about that to know how to find and fix that Jul 06 22:42:24 GarthPS, I may get a pre-2, not sure Jul 06 22:42:26 hm ok Jul 06 22:42:33 basically morphis and me wrote for it Jul 06 22:42:33 GNUtoo: :) Jul 06 22:42:41 GNUtoo: me too :) Jul 06 22:42:44 and I'm waiting it to arrive since a very long time Jul 06 22:42:50 like many months Jul 06 22:42:56 but I think I will not have one.. Jul 06 22:43:09 ok Jul 06 22:43:11 why? Jul 06 22:43:15 and perhaps I should not have done this because perhaps it will not help you to have one.. Jul 06 22:43:42 we can't really know what they tough.... Jul 06 22:43:52 so we can't really know if it's the case Jul 06 22:43:57 because it has been a month without a (positive) answer , had answer but telling my to wait for someone else to answer./. Jul 06 22:44:06 anyway I've too much devices.... Jul 06 22:44:16 ok Jul 06 22:44:43 GNUtoo: lol. did they answer to you ?* Jul 06 22:45:19 they told me I wasn't a develloper, then morphis wrote many mails Jul 06 22:45:27 and morphis handles it now Jul 06 22:46:30 I could have done most of morphis integration work in oe+shr to make him work on something else like sms Jul 06 22:46:39 but I waited....waited....instead Jul 06 22:46:50 GNUtoo: yeah.. :/ Jul 06 22:46:52 so some core issues like wifi are left Jul 06 22:47:19 it's at configuring now Jul 06 22:47:39 GNUtoo: id do need to make longer test but right now it works quit well.. Jul 06 22:47:52 ah? Jul 06 22:47:56 try to transfer large files Jul 06 22:48:00 it will break Jul 06 22:48:03 GNUtoo: http://trac.shr-project.org/trac/ticket/1453 Jul 06 22:48:36 yes the problem is to make it reliable Jul 06 22:48:44 try to transfer huge files over wifi Jul 06 22:48:50 your connection will stop working Jul 06 22:48:51 GNUtoo: so what have morphis about pre2 from HP ? Jul 06 22:49:07 GNUtoo: yeah it is that part taht I need to check Jul 06 22:49:41 GNUtoo: did they told him that you will recive devices ? Jul 06 22:49:49 I think so Jul 06 22:49:54 but he has to ping them often Jul 06 22:50:08 GNUtoo: yeah .. i see that.. Jul 06 22:50:34 GNUtoo: so! what is up on your N900..? Jul 06 22:50:44 it is slow...... :) Jul 06 22:51:00 hmmm Jul 06 22:51:05 I've an issue....in gpe-scap Jul 06 22:51:10 it doesn't take vlc's content Jul 06 22:51:22 http://scap.linuxtogo.org/ Jul 06 22:51:32 but apart that it works fine Jul 06 22:51:34 thanks a lot Jul 06 22:52:35 altough the written things are a bit huge Jul 06 22:52:37 *fonts Jul 06 22:53:26 GNUtoo: oulaa this is not for me :) Jul 06 22:53:29 but cool ! Jul 06 22:53:37 so gn8! cu ! Jul 06 22:53:54 good night....trying youtube Jul 06 23:00:33 bluelightning: hey Jul 06 23:00:54 hi ant__ Jul 06 23:02:49 bluelightning: my daily battle with klcc-cross has a sweet-sour end tonigth Jul 06 23:03:08 now the binary can be installed and cleaned Jul 06 23:04:31 though, still troubles with packaging :/ Jul 06 23:05:11 I can somehow trigger a nasty python error :/ Jul 06 23:05:14 ERROR: (file: 'fixup_perms', lineno: 215, function: fixup_perms) Jul 06 23:05:14 | ERROR: Function 'fixup_perms' failed Jul 06 23:09:17 ant__: any more info about the error? Jul 06 23:10:01 http://paste.debian.net/122165/ Jul 06 23:10:16 surely my bad... Jul 06 23:13:19 if I comment out #PACKAGES = "${PN}" #FILES_${PN} then error disappears Jul 06 23:15:50 hmm... he key bit seems to be "operation not permitted" Jul 06 23:16:01 s/he/the Jul 06 23:34:15 khem: any plans to upgrade gcc to 4.6.1 ? Jul 07 01:02:43 My OE build gives me an error like so when building cpufrequtils Jul 07 01:02:45 http://pastebin.com/PDeb9DDH Jul 07 01:03:19 Any ideas on why OE doesn't know where arm-angstrom-linux-gnueabi-libtool is? Jul 07 01:04:10 I do see it in: Jul 07 01:04:24 .. /build/tmp-angstrom_2010_x/work/armv7a-angstrom-linux-gnueabi/libtool-cross-2.4-r28.2/libtool-2.4/arm-angstrom-linux-gnueabi-libtool Jul 07 01:07:41 * joelagnel tries to hardcode it in run script Jul 07 01:18:30 * joelagnel notices lots of packages are broken while building beagleboard*validation* recipes Jul 07 01:18:58 e-wm, matrix-gui, cpufrequtils Jul 07 01:19:45 so far :) Jul 07 01:30:27 what's the recommended way of working around the documentation mess that angstrom's currently in? is there a way to disable doc generation or otherwise avoid trying to build it? Jul 07 01:39:06 don't think its angstrom specific Jul 07 01:39:13 don't know of a good solution at the moment Jul 07 01:39:24 ideally everything would obey a doc/docs/manual distro feature or something Jul 07 01:42:22 hmm, ok Jul 07 01:42:34 I guess I'll put iputils back in since I can't seem to locate what is demanding it be there Jul 07 01:43:02 although i didn't see anyone answer about the caching... does bitbake cache dependency requirements? is there a -c option I can give to have it flush that cache, if it exists? Jul 07 01:48:41 i answered you earlier Jul 07 01:48:49 *task* recipes emit an ipk with the deps in it Jul 07 01:49:09 you need to clean them if you change the rdeps going into that ipk Jul 07 01:49:10 or bump pr **** ENDING LOGGING AT Thu Jul 07 02:59:56 2011