**** BEGIN LOGGING AT Mon Oct 16 02:59:57 2006 Oct 16 06:33:40 hm i just renamed a package from name into name-newtag and now it fails to install its pixmaps into /usr/share/name/pixmaps or /usr/share/name-newtag/pixmaps ?? Oct 16 06:38:22 <_law_> rob_w, the directory name-newtag exists? Oct 16 06:39:08 _law_, it exists in tmp/work/name-newtag_0.1-r0/name-newtag/ Oct 16 06:46:17 hello / bonjour Oct 16 06:49:58 good morning all Oct 16 06:51:40 <_law_> hi koen Oct 16 06:52:41 <_law_> rob_w, in the image and install dir (in tmp/workname-newtag...) it also exists? Oct 16 06:53:35 hey _law_ Oct 16 06:59:03 <_law_> wow mtn db regenerate_rosters takes hours.... Oct 16 06:59:55 that's why we have snapshots Oct 16 07:00:09 <_law_> i think wget http://www.openembedded.org/snapshots/OE-this-is-for-mtn-0.30.mtn.bz2 should be faster.... Oct 16 07:03:48 _law_: I took the file that you just mentioned and it worked out-of-the-box, without any roster regereration Oct 16 07:06:14 _law_, managed it now with some configure.am tweaks and some more inside app changes . thanks Oct 16 07:06:17 and it was updated yesterday, so it isn't 500 revs behind :) Oct 16 07:08:03 <_law_> schurig, i tried to convert a 0.28 db Oct 16 07:08:56 _law_: I tried this at first, too. But after some rosterification, I hit Ctrl-C. /me is not used to wait for SCM tools, because in my other work I only use SCMs with a decent speed Oct 16 07:15:19 * koen sighs Oct 16 07:15:36 all the people that were too lazy to do something are the ones that complain the loudest Oct 16 07:15:53 "empty barrels make the hardest sound" Oct 16 07:19:44 koen: shut up :-) Oct 16 07:19:58 koen: I *DID* many things in the pre-MONOTONE time Oct 16 07:20:21 you did nothing when we asked people to test new scms Oct 16 07:20:25 and I must say that with mtm 0.30 mtn is at least usable Oct 16 07:20:46 I'm still using just a one GHZ machine, not a dual core 3 GHz AMD thingy or so ... Oct 16 07:20:51 http://www.openembedded.org/bonsai/query/author/schurig/ Oct 16 07:21:30 koen: I did. At that time I was using Subversion for many projects and said that this is all I'd need. But this was ruled out because other people said they definitely need a distributed SCM. Oct 16 07:21:32 zecke: you should ask tridge for the bk parser :) Oct 16 07:21:59 koen: there was bk2mtn script IIRC Oct 16 07:22:22 zecke: I was the cvs import lying around somewhere if you want it Oct 16 07:22:32 hmm it would be nice if merges are marked as merges (as a cert) Oct 16 07:22:52 zecke: you could add them with a script Oct 16 07:22:52 zecke: then the bk2mtn script was buggy, I pushed many things in bitkeeper Oct 16 07:23:11 schurig: this is only with the mtn era :) Oct 16 07:23:12 zecke: and I also started with bitbake (okay, it was slow and crappy, but I started it) Oct 16 07:23:48 schurig: but that was pre the BitBake era? ;) Oct 16 07:23:56 schurig: or what is your login on Berlios Oct 16 07:24:12 zecke: I think it's "holgerschurig" there as well Oct 16 07:24:28 schurig: we track/index BitBake SVN and OE mtn ;) Oct 16 07:24:35 schurig: only |schurig| = 1 Oct 16 07:24:37 zecke: right, it was pre-bitbake. I think what I made was not really usable for real-world Oct 16 07:25:05 * zecke pets his little tree (bonsai) Oct 16 07:25:21 koen: http://www.openembedded.org/bonsai/query/author/schurig/?items=40 Oct 16 07:25:33 zecke: but if some later-comer like koen compares my to an empty barrel than I think this is unfair Oct 16 07:25:48 zecke: this 40 is just something I made to test if MTM speed is now usable for me Oct 16 07:26:09 schurig: I missed this statement of koen Oct 16 07:26:12 zecke: http://ewi546.ewi.utwente.nl/tmp/bk-export.tar.bz2 Oct 16 07:26:29 zecke: schurig was doing his usual "mtn is sooo slow" routine again Oct 16 07:26:42 schurig: yes, you started it. Probably your usage of buildroot was the reason to start OE Oct 16 07:26:43 koen: actually not really Oct 16 07:26:46 at which I pointed out he did nothing while we were selecting an scm Oct 16 07:26:58 koen: http://www.openembedded.org/bonsai/query/author/koen/?items=4000000 I wonder if this will dos openembedded.org Oct 16 07:27:08 so schurig lost the right to complain Oct 16 07:27:22 koen: no. _LAW_ was saying that db regenerate rosters was dead slow Oct 16 07:27:35 koen: and believe it or not, your beloved SCM *IS* dead slow for this operation Oct 16 07:27:58 I'm not saying it is fast Oct 16 07:28:08 koen: and I said that I tried this, hit Ctrl-C and then downloaded the snapshot that fits to my MTM version Oct 16 07:28:13 I'm saying *you* and others lost the right to complain Oct 16 07:28:26 koen: and I say that you are a moron :-) Oct 16 07:28:31 schurig: hehe, "this operation" Oct 16 07:28:33 meritocracy is a bitch, isn't it? Oct 16 07:35:15 koen: the way you interpret it Oct 16 07:36:21 koen: suppose someone contributed to OE heavily in the Bitkeeper timer. Now, for some legal reason (BK had an awful license) this person had to stop contributing. Now after a year somebody came and say "oh, you've no merit, you didn't do anything". That would be the statement of a moron. Oct 16 07:36:46 I don't say that Oct 16 07:37:07 I'm saying that you should stop complaining about something you *choose* to do nothing about Oct 16 07:38:35 koen: in my case I declared very early (before OE switched to monotone!, when we were in the last weeks of BK usage) "I tried Monotone for a project of mine, to see how good it is. It's to slow, don't use it". Other people decided other, and than, becuase I work in a comany and have to be productive. And the older monotone was very slower then the current one. And so I *HAD* to say that I can't use monotone and stop using OE. And I set up an Oct 16 07:38:36 internal SVN tree with my OpenMN distro. And I used that to get my work done. Oct 16 07:39:14 (weird grammaer, that happens if you phone while IRCing :-) Oct 16 07:39:45 koen: I choose to do something about, but other people said "No, that's no what we want". Keep that in mind. Oct 16 07:40:41 koen: and, if you're honest, if 10 people said "The speed is okay for us, and we like the features" then one, or two can do as much as they want, they can't change things. And that shouldn't then be holded against them. Oct 16 07:41:07 and what did you do the second time we evaluated scms? Oct 16 07:41:14 also nothing Oct 16 07:41:21 koen: I repeat: I SAID THAT WE SHOULD USE SUBVERSION! Oct 16 07:41:29 if KDE can use, then we can use it too Oct 16 07:41:58 but other people insisted on using a distributed SCM Oct 16 07:42:15 if we'd follow KDE, we'd dump bitbake, use autotools, dump autotools, use scons, dump scons, use cmake... etc Oct 16 07:43:05 and again, if X people say "We need distributed SCM" and Y people say "centralized SCM" and X >> Y, then it's ok to use what X people want. But it is unfair than that one from X says to an Yer: "you didn't do what I wanted" Oct 16 07:43:14 I have the same reason to say "You didn't do what I wanted". Oct 16 07:43:22 koen: what did you do to solve my problem? Oct 16 07:43:37 this question doesn't make sense, so stop being unfair to me Oct 16 07:43:41 I tested git, hg and darcs Oct 16 07:43:49 and I tested svb Oct 16 07:43:51 and I tested svn Oct 16 07:44:03 schurig: svn is too slow for KDE as well :) Oct 16 07:44:21 zecke: then mtn would be unbearable to them Oct 16 07:45:29 mtn is a good system Oct 16 07:45:37 or to rephrase, mtn will be a good system Oct 16 07:46:14 it's just enduring the pain of upgrades, given that they're purposefully not maintaining backwards compatibility Oct 16 07:46:40 blaster8: they care about netsync compatibility Oct 16 07:46:54 true, but they still break it occasionally :) Oct 16 07:47:06 It's important they get it right, for version 1.0 Oct 16 07:47:07 blaster8: they only did once Oct 16 07:47:35 twice Oct 16 07:47:45 0.14-0.15 and 0.25-0.26 Oct 16 07:47:55 ah, I only use 0.18 and newer :) Oct 16 07:48:04 ok Oct 16 07:48:17 morning all Oct 16 07:48:21 hey ade|desk Oct 16 07:48:22 as I say, it will be great, it's just occasionally painful in the meantime Oct 16 07:57:33 morning Oct 16 07:58:35 Updating SVN from http://svn.berlios.de//svnroot/repos/bitbake/ starting at rev:618 Oct 16 07:58:35 Segmentation fault Oct 16 08:01:34 heh.. koen <> schurig discussion.. Oct 16 08:02:10 hrw|work: I saw that you pushed some benchmarks. Do you want this for your kernel work or do you intent to have some wiki page with the benchmark numbers published? Oct 16 08:02:32 morning Oct 16 08:02:40 schurig: customer paid me to add them so I added them Oct 16 08:05:24 schurig: benchmarks will be published on tinderbox :) Oct 16 08:08:46 morning all ! Oct 16 08:08:51 does no-one ever read the bugtracker :-( Oct 16 08:12:41 is the oe site down ? Oct 16 08:12:56 Ifaistos: Seems so Oct 16 08:13:03 I could not get to it, either. Oct 16 08:20:16 I have a question regarding icecc usage. I inherit icecc so that is fine. Will the following setup adhere to "icecc logic"? I have OE, icecc daemon and scheduler all on one machine. bitbake is run from this machine as well. All helper computers in the network run the icecc daemon and are not protetected by a firewall. Oct 16 08:20:21 I am especially curious if bitbake can run on the same computer as the scheduler or if it has to be the other way round. Oct 16 08:25:55 <_law_> http://www.openembedded.org works for me Oct 16 08:26:11 anyone know a program which lets you move applications from one X server to another? Oct 16 08:26:12 Ifaistos: works for me too Oct 16 08:26:31 mithro: Heh, I wish Oct 16 08:26:58 yep. now it works :) Oct 16 08:27:18 hi Oct 16 08:27:20 mithro: teleport? Oct 16 08:27:33 mithro: applications? Wouldn't that need kernel support? E.g. find out which files and sockets are open, and what TCP connections, "freeze" them, move them over, re-open them and start them? I guess you need tools from Virtuozze enabled kernels for this. Oct 16 08:28:15 I think I have bricked my zaurus :/ Oct 16 08:28:32 schurig: i'm not sure, couldn't you just tell xlib to start painting in a different place? Oct 16 08:28:52 can someone help me here to try to revive it ? Oct 16 08:28:57 mithro: oh, I thought you mean the whole app. Oct 16 08:29:06 Has anyone tested if the TARGET_FPU="soft" would work on other arch's also (other <>arm) ? Oct 16 08:29:22 schurig: no, just where it is being displayed Oct 16 08:29:55 mithro: isnt that what teleport does in gpe? Oct 16 08:30:07 dunno, whats teleport? Oct 16 08:30:33 ~prod koen Oct 16 08:30:35 * ibot zaps koen with a high voltage cattle prod Oct 16 08:30:41 i thought hrw|work was just suggesting something which didn't exist :P Oct 16 08:31:13 teleport can move gtk+ applications between displays, yes Oct 16 08:31:31 koen: does it need support of GTK for this? Oct 16 08:31:41 koen: or would it work with any xlib based app? Oct 16 08:31:46 teleport can move any gtk application? Oct 16 08:32:21 or only special GPE apps? Oct 16 08:33:29 mithro: any gtk app, if the gtk lib has the teleport patch Oct 16 08:33:44 teleport support is 'kind of' built-in to gtk+ Oct 16 08:34:00 is the gtk shipping with ubuntu likely to have the teleport patch? Oct 16 08:34:06 no idea Oct 16 08:34:48 * koen is still waiting for proper xcb support in gtk+ Oct 16 08:34:59 The following extra packages will be installed: Oct 16 08:34:59 libgpewidget1 Oct 16 08:35:00 ? Oct 16 08:35:04 xcb? Oct 16 08:38:21 mithro: X C Bindings Oct 16 08:44:01 mithro: xlib is quite heavy, XCB is done this again, much leaner and therefore easier to maintain. Also, they want to add new features that are hard to do in xlib, e.g. an asynchronous interface. Oct 16 08:44:05 mithro: http://en.wikipedia.org/wiki/XCB Oct 16 08:44:15 (I think in the last LWN was something about it) Oct 16 08:44:26 later Oct 16 08:48:19 ok mtn has fscked up my repo badly :-( Oct 16 08:48:52 hi ,, offtopic question : when i need to do a cvs update , i normally do cvs -n update in order to see prior what files need merges ..so -n doesnt actaully do something to the repo then .. so can i also see the diff output to the mergeable files combined with -n ? Oct 16 08:49:46 OK monotone is busted, I get mtn: misuse: rename source path 'packages/xev' does not exist Oct 16 08:49:57 then I am left with a load of xorg files deleted Oct 16 08:50:17 good morning Oct 16 08:50:22 florian_kc: hey Oct 16 08:52:04 rob_w: 'cvs diff' is not enough? Oct 16 08:53:01 hrw|work, yeah that is .. i just found that i do need to do cvs diff as a separete command .. thanks and never mind Oct 16 09:01:46 re Oct 16 09:04:14 lardman|gone: please do not encourage people to mix softfpa and softvfp feeds.... Oct 16 09:09:27 koen: whats softvfp? Oct 16 09:10:05 vs softfpa? Oct 16 09:10:22 hi zecke Oct 16 09:14:26 has someone already made script for padding a full image with kernel firmware & rootfs ? Oct 16 09:14:57 mithro: different (incompatible!) kinds of floating point emulation for arm Oct 16 09:15:02 Genesis: your device need such kind? Oct 16 09:15:24 koen: know where I can find information on the differences? Oct 16 09:15:27 hrw|work, : yeap since i've no ethernet on my prototype Oct 16 09:15:34 Genesis: most of devices use kernel + rootfs (separated) or rootfs with kernel Oct 16 09:15:37 i don't know what is the best way to do that Oct 16 09:15:48 where 'rootfs with kernel' == rootfs which contain kernel Oct 16 09:16:08 hrw|work, : the flashing tool use combine image Oct 16 09:16:19 i think it's easyer to use in production Oct 16 09:16:37 Genesis: kernel + padding + rootfs (jffs2/cramfs) padded? Oct 16 09:16:47 koen: have you found the djamgo patch? Oct 16 09:16:55 firmware + padding + kernel + padding + rootfs + pading Oct 16 09:16:58 zecke: no Oct 16 09:17:03 koen: found it? Oct 16 09:17:19 hrw|work, : what should i use for padding ? Oct 16 09:17:21 cat ? Oct 16 09:17:36 Genesis: or dd Oct 16 09:18:26 koen: http://code.djangoproject.com/ticket/2912 Oct 16 09:18:27 Genesis: you know size of each padding (can calculate it) so cat firmware >file;dd if=/dev/zero of=padding1 count=paddin1size; etc.. Oct 16 09:18:47 oki thx Oct 16 09:21:28 Is it possible to exclude packages from being compiled with icecc? IOW, compile them the standard way since compilation with icecc fails for them? Oct 16 09:21:59 Laibsch: why should it fail with icecc? Oct 16 09:22:25 I don't know. I have had it happen before and quilt-native just failed on me. Oct 16 09:22:42 I was wondering if this should be a bitbake or OE bug which is one of the reasons I asked. Oct 16 09:23:10 The failure happened on cleaned out tmp. Oct 16 09:24:54 I went the INHERIT += "icecc" way if that matters. Oct 16 09:25:54 Laibsch: and the patch from the bug tracker? Oct 16 09:26:11 Which patch? Oct 16 09:26:54 bug 1485? Oct 16 09:27:51 I can give it a shot. Oct 16 09:58:18 <[TEHb]> Hallo Guys! Oct 16 09:59:04 <[TEHb]> Any IXDP465 owners here? Oct 16 10:02:02 morning Oct 16 10:02:57 hey mickeyl Oct 16 10:03:01 <[TEHb]> Can anyone tell me part number of memory chips, which installed on IXDP465? Oct 16 10:03:15 hi mickeyl Oct 16 10:07:18 http://freshmeat.net/projects/scmbug/ Scmbug is a system that integrates software configuration management (SCM) with bug-tracking Oct 16 10:07:54 hrw|work: I read that as Scumbag Oct 16 10:09:00 hrw|work: last time I looked at scmbug it was extremely crappy Oct 16 10:20:54 03mickeyl 07org.oe.dev * r35e474b0... 10/ (4 files in 2 dirs): add python-gobject bindings Oct 16 10:21:00 03mickeyl 07org.oe.dev * r9e9e3622... 10/ (1 packages/python/python-pygtk2_2.10.0.bb): add pygtk 2.10.0 (for usage with gtk+ 2.10.x) Oct 16 10:22:55 XorA: did you have a change to commit the gtk LIBV compat cset? Oct 16 10:26:53 <[TEHb]> psokolovsky: hi Oct 16 10:27:09 Hi! Oct 16 10:27:19 psokolovsky, morning Oct 16 10:30:21 koen_: I was waiting for OEDEM approval Oct 16 10:33:55 XorA: you have it :) Oct 16 10:34:00 koen_: my python being terrible, I wanted someone like mickeyl to ok it Oct 16 10:34:23 * koen waves at mickeyl :) Oct 16 10:34:24 koen: but at moment I cant commit as monotone is screwed Oct 16 10:34:42 do we have a bug number for that? Oct 16 10:34:43 zecke: Fails even with the patch. But I had to rework it a bit to apply cleanly so maybe I did something wrong. Unfortunately likewise is not here so we do not know if he created his patch from an up-to-date OE tree. Oct 16 10:35:00 1455 Oct 16 10:35:22 mickeyl: feel free to commit if you can, my monotone is totally broken here Oct 16 10:36:31 ok Oct 16 10:37:14 mickeyl: I found an appropriate leffe glass for you: http://dominion.kabel.utwente.nl/koen/cms/index.php/v/Koen/Nederland/enschede/Vestingbar/bata2004/batafeest2004/img_5140.jpg.html?q=gallery :) Oct 16 10:38:25 * koen heads to the store to replace his molten PSU Oct 16 10:42:55 How does one get an account on the current wiki? Oct 16 10:43:57 mail koen Oct 16 10:50:07 how does one find beer in glasses like that? Oct 16 10:50:34 koen: heh! I want one of those! :D Oct 16 10:54:45 morning all Oct 16 10:56:46 hi RP zecke Oct 16 10:56:48 hi RP Oct 16 10:59:28 hey Oct 16 10:59:44 ~lart nm for thinking usb1 with mtu 900 is better than eth0 Oct 16 10:59:44 * ibot pushes the wall down onto nm whilst whistling innocently for thinking usb1 with mtu 900 is better than eth0 Oct 16 11:03:20 yeah security issues in bugzilla Oct 16 11:03:31 who would have... Oct 16 11:04:49 About mtn: this night (european) I couldn't pull for ~40mins (thanks whoever kicked it), but still no commit mails since that time.. Oct 16 11:06:01 eek, I could be guilty Oct 16 11:28:23 ibot: ping Oct 16 11:28:31 hi Liam Oct 16 11:28:31 pong Oct 16 11:28:32 morning all Oct 16 11:29:18 hi lrg Oct 16 11:48:45 good morning folks Oct 16 12:29:46 03rwhitby 07org.oe.dev * r90907870... 10/ (1 packages/linux/ixp4xx-kernel-svnpatch.inc): ixp4xx-kernel-svnpatch: Allow the repo location to be overridden Oct 16 12:35:44 As most of you know I am not too firm on compilers etc. As such, it might not be a good idea for me to maintain files. On the other hand, I believe I am the one most active in "the Japanese arena". So I wonder if I should add my information to the MAINTAINER file or not? Oct 16 12:39:07 Laibsch: sure Oct 16 12:39:35 OK. Shall I send it to the BTS or the ML? Oct 16 12:39:46 BTS Oct 16 12:39:50 OK Oct 16 12:39:51 Thanks. Oct 16 12:42:36 03pfalcon 07org.oe.dev * r47f59b2e... 10/ (1 conf/machine/h4000.conf): h4000: Try to organize machine settings per logical groups. Oct 16 12:42:50 03pfalcon 07org.oe.dev * r4bf8cd68... 10/ (1 conf/machine/h2200.conf): h2200: Organize machine seetings per logical groups. Oct 16 12:42:50 03pfalcon 07org.oe.dev * r64517e18... 10/ (4 files in 3 dirs): Oct 16 12:42:50 LAB-settings.conf: Separate internal details of LAB support for machines which can use it. Oct 16 12:42:50 * Convert h2200 & h4000 to use it. Oct 16 12:42:57 03bugs.openembedded.org 07org.oe.dev * r520d98a2... 10/ (4 files in 3 dirs): Oct 16 12:42:57 granule: A new package, flashcard-based language trainer. Oct 16 12:42:57 * Initial recipe. Oct 16 12:42:57 * Also, a recipe for dependency library, libassa. Oct 16 12:44:10 Thanks psokolovsky. Oct 16 12:44:22 psokolovsky: s/include/require/g Oct 16 12:44:36 Laibsch, Hi, going to cleanup them a bit too. Oct 16 12:44:42 koen_: ok Oct 16 12:45:05 psokolovsky: What was wrong with them? I tried to make them adhere to standard to the best of my ability. Oct 16 12:45:36 Laibsch, e.g., PR should be "rX", nothing else. SECTIONs also can be assigned ;-) Oct 16 12:46:01 Laibsch, and post important - you're sure granule must depend on *-native packages? Oct 16 12:46:19 unless you are mickeyl, in that case PR can be 'mlX' Oct 16 12:46:21 :) Oct 16 12:46:58 psokolovsky: I added the dependencies one by one when I encountered a failure. I believe they are indeed necessary. Oct 16 12:47:12 Until vlg removes calls to them in the build-process of course. Oct 16 12:47:30 As always, I might be wrong of course. Oct 16 12:47:42 Laibsch, it's just more than strange for a package to depend on native ones. that means native packages are required to *build* it. Oct 16 12:47:42 But granule would not build until I added those dependencies. Oct 16 12:47:57 Laibsch, ok, let it be that way for now Oct 16 12:47:58 psokolovsky: Yes, that much I know. Oct 16 12:48:04 ok Oct 16 12:48:13 koen_: Do we use zecke_sort for the Maintainer file? Oct 16 12:48:25 I believe they are called during the build process as I said above. Oct 16 12:48:27 zecke: no :) Oct 16 12:48:38 zecke: mickey_sort Oct 16 12:49:05 Laibsch, also, granule*.png are not needed? Oct 16 12:50:58 You mean from oz.leggewie.org? That is just for having a possibility for people to download. the libassa patch is not needed either. Everything that is necessary is in the patch in the BTS. Oct 16 12:52:09 koen: I think my cron job kills notification Oct 16 12:52:10 * XorA thinks he killed mickeyl with his python skillz Oct 16 12:52:10 03freyther 07org.oe.dev * r64153b2d... 10/ (1 MAINTAINERS): MAINTAINERS: Add myself and my interests. icecream and insanity Oct 16 12:52:15 oh it doesn't Oct 16 12:52:51 03pfalcon 07org.oe.dev * r2fbfe5db... 10/ (3 files in 2 dirs): granule, libassa: Cleanup recipes a bit. Oct 16 12:52:55 Laibsch, ok, nice. Then, we are done for now ^. Thanks! Oct 16 12:53:19 morning Oct 16 12:53:29 hey VoodooZ Oct 16 12:53:50 how's the man? Oct 16 12:54:07 I will merge Oct 16 12:54:58 A few guys are still trying to get glibc 2.4 to build under slugos. I tried to change the prefered glibc version to 2.4 but it says there's no such thing. Oct 16 12:55:17 VoodooZ: 2.4 is broken on arm Oct 16 12:55:46 hey likewise Oct 16 12:56:25 ah. Oct 16 12:56:33 what about 2.5 :) Oct 16 12:56:43 that seems to work better Oct 16 12:56:54 yeah, that's what you were saying earlier. Oct 16 12:57:01 Are you running it on ARM too? Oct 16 12:57:05 yes Oct 16 12:57:08 nice. Oct 16 12:57:23 that reminds Oct 16 12:57:26 You said we would need to update slugos for the new EABI thingy though right? Oct 16 12:57:38 koen, so we're not going to come back to 2.4 anytime soon? ;-) Oct 16 12:57:46 I should check in on my dad to see if he managed to setup OE via the mastermakefile yet Oct 16 12:57:58 psokolovsky: use --author with better values then 'bugs.openembedded.org' ;D Oct 16 12:58:24 hrw|work, it's actually email address. kick CIA for trimming it ;-) Oct 16 12:59:11 CIA-4: kick Oct 16 12:59:59 koen, what? your dad is working with OE too? Nice! My dad can't even turn on the computer! Oct 16 13:00:29 VoodooZ: I donated my slug to him Oct 16 13:00:30 my mom use skype Oct 16 13:01:23 psokolovsky: 2.5 seems to be working quite well, apart from the selinux stuff in the ldd script Oct 16 13:01:55 psokolovsky: ah.. I forgot that Laibsch use this email address.. Oct 16 13:02:00 koen: hi Oct 16 13:02:02 hi all Oct 16 13:02:03 koen, Nice! So I'll kill work/ for 2.4 then. was afraid to do so if switch back soon ;-) Oct 16 13:02:36 fixeset building failed : http://pastebin.ca/205014 Oct 16 13:02:52 zecke: Holger begins with an 'H', which sorts between 'G' and 'I' Oct 16 13:02:54 koen, donated? Did you set him up Oct 16 13:03:03 koen: lol Oct 16 13:03:24 ~fishslap zecke Oct 16 13:03:26 * ibot slaps zecke up side the head with a wet fish. Oct 16 13:03:48 where does H live in zecke land ? Oct 16 13:03:53 VoodooZ: he managed to install slugos by himself :) Oct 16 13:03:58 nice! Oct 16 13:04:04 ade|desk: after Ø Oct 16 13:04:20 wow, end of file .. nice Oct 16 13:04:33 'G' in zalphabet => False Oct 16 13:05:07 I guess it's a tribute to all the guys who worked on slugos! "It's so easy, your dad can do it!" Oct 16 13:06:22 03freyther 07org.oe.dev * r3763af0d... 10/ (1 MAINTAINERS): MAINTAINERS: Place 'Holger' somewhere else... Oct 16 13:06:44 koen, so let's say I wanted to try the glibc 2.5 package: would simply selecting 2.5 as the prefered glibc provided in slugos.inc do it or? Oct 16 13:07:41 VoodooZ: see how angstrom does it Oct 16 13:07:41 ok. thanks. Oct 16 13:08:37 Yep. it looks the same. Probably lots of hidden stuff though. we'll try... Oct 16 13:08:58 VoodooZ: try copying the complete toolchain block Oct 16 13:09:15 from the angstrom.conf? Oct 16 13:09:55 I noticed the reference to the EABI stuff too which will probably burn me later... Oct 16 13:10:28 VoodooZ: http://rafb.net/paste/results/HNSE0L94.html Oct 16 13:11:13 ok. I'll play around with that. I noticed some angstrom specific include files too... Oct 16 13:16:05 koen: do we have a linux-libc-headers preferred for ulibc eabi ? Oct 16 13:17:00 03mickeyl 07org.oe.dev * ra945a398... 10/ (4 files in 3 dirs): add python-pycairo, python bindings to the cairo canvas library Oct 16 13:19:00 03mickeyl 07org.oe.dev * r12ff563e... 10/ (4 files in 3 dirs): python-pygtk2 2.10.0 add patch to make it cross-build Oct 16 13:20:33 03koen 07org.oe.dev * r425d05a6... 10/ (1 conf/machine/h2200.conf conf/machine/h4000.conf): h2200, h4000: s/include/require/ and add some comments Oct 16 13:21:04 ade|desk: uclibc uses the kernel headers iirc Oct 16 13:21:17 ade|desk: I didn;t test yet with the new linux-libc header package Oct 16 13:21:45 uclibc seems to work well with 2.6.18 real headers Oct 16 13:25:20 florian_kc: Can you tell me about the status for http://handhelds.org/hypermail/gpe/55/5536.html (upload of Japanese .po files into GPE)? Oct 16 13:26:32 03xora 07org.oe.dev * r47d594be... 10/ (1 MAINTAINERS): MAINTAINERS: add myself Oct 16 13:27:48 Laibsch: hrm... good point, i didn't get much of feedback about this. i'll take a look and add them to svn if they are not uploaded Oct 16 13:28:05 Cool, thanks. Oct 16 13:29:23 03pfalcon 07org.oe.dev * r8c180cdd... 10/ (9 files in 2 dirs): pocketpc-based machines: Consistently use "require" instead of "include" in confs. Oct 16 13:30:08 morning Oct 16 13:38:30 03pfalcon 07org.oe.dev * r9d164624... 10/ (13 files in 3 dirs): Oct 16 13:38:30 obsolete/xorg-lib/: Readd some pre-x11r7 libs as obsolete. Oct 16 13:38:30 * Were removed in 989e5fa001221455061a9cc9ab30c32246736ab6, Oct 16 13:38:30 but still may be required to support legacy applications. Oct 16 13:40:12 03koen 07org.oe.dev * r514b31d8... 10/ (20 files in 3 dirs): conf/machine: remove -common stuff Oct 16 13:41:43 03koen 07org.oe.dev * re675e4bb... 10/ (3 files in 2 dirs): rename xorg-lib to something more appropriate Oct 16 13:43:02 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Oct 16 13:45:02 koen: poodle-2.4.conf: +PREFERRED_VERSION_ = "0.13e" Oct 16 13:45:16 psokolovsky: "pocketpc-based machines" == h3600/h3900/h5400/h2200/hx2000/h4000/hx4700/jornada*/etc? Oct 16 13:46:14 hrw|work, yep. I didn't touch journada though as not sure hh.org is upstream for it (kernel) Oct 16 13:46:45 koen: lotsa typos like that in your pre-last commit Oct 16 13:48:42 I heard 2.6.18+ kernels came with integrated Real-Time feature. Are those disabled by default? Oct 16 13:52:59 XorA: I'm going to rewrite the bbclass :) Oct 16 13:53:06 mickeyl: cool by me Oct 16 13:53:26 mickeyl: which one? Oct 16 13:53:37 zecke: bonsai updater works again :) Oct 16 13:53:37 mickeyl: I only ever meant it as proof of concept, my python is dire Oct 16 13:53:37 XorA: what's the exact output for this binversion? do you have an example? Oct 16 13:53:43 hrw|work: gtk-binver Oct 16 13:54:00 mickeyl: 2.10.0 Oct 16 13:54:02 ah Oct 16 13:54:11 or 2.4.0 ? Oct 16 13:55:15 mickeyl: basically gtk has a binary version, which is in the .pc file, while other things likle gtk-engines libgtkstylus need to know to put plugins in right dir Oct 16 13:55:54 mickeyl: so we need to be able to set LIBV="2.X.0" or use it in do_split_packages Oct 16 13:56:07 mickeyl: 2.10 changed something with refcounting, so the API changed Oct 16 13:56:49 hrw|work, any objections to such collectives as "pocketpc-based machines" in commit messages? Oct 16 13:57:03 XorA: i see. is there a reason why you only patched one version of the gtk-engines bbs? Oct 16 13:57:16 03koen 07org.oe.dev * r9fa69b05... 10/ (1 conf/machine/include/poodle-2.4.conf): poodle2.4.conf: remove spurious commit Oct 16 13:57:24 psokolovsky: I prefer machine names as not everyone know that h2200 is pocketpc one Oct 16 13:57:29 mickeyl: proof of concept, I felt my code, and maybe my method was wrong and wanted discussion Oct 16 13:57:38 *nod* Oct 16 13:57:38 hrw|work, ok ;-) Oct 16 13:59:10 03pfalcon 07org.oe.dev * rb8e69aa2... 10/ (5 files in 2 dirs): h2200, h4000, hx2000, hx4700: PCMCIA_MANAGER is task-bootstrap remnant. Remove. Oct 16 13:59:34 psokolovsky: task-base use PCMCIA_MANAGER.... Oct 16 14:00:32 hrw|work, ;-\ will revert Oct 16 14:00:44 task-base-pcmcia-rdepends = "\ ${PCMCIA_MANAGER} \ Oct 16 14:01:11 hrw|work, but, it's so mindless of it. task-base is *what* should select pcmcia manager based on kernel in use. Oct 16 14:01:27 psokolovsky: oz use pcmcia-cs with 2.6 kernels Oct 16 14:01:49 icecream stuff: does iceccd need to run as root (-r root) in order to work with the tar.bz2 chroot compiler environments? Oct 16 14:02:31 hrw|work, weird Oct 16 14:02:42 03pfalcon 07org.oe.dev * reb6e4d15... 10/ (5 files in 2 dirs): disapproval of revision 'b8e69aa20c2c86ec586299ba3d7ce6ec57b7ef8e' Oct 16 14:02:49 any volunteers to add icecream to OE as a recipe? Oct 16 14:02:54 XorA: your class is cool, but it requires spawning a couple of external processes which is the only reason for me rewriting it Oct 16 14:03:13 hrw|work, afaik it wa deprecated yet in 2.6.16. is it still alive in 2.6.18 at all? Oct 16 14:03:16 os.system('echo blah') Oct 16 14:03:29 psokolovsky: it was removed in 2.6.18 iirc Oct 16 14:03:39 koen :D Oct 16 14:03:55 that's what I mean. ok. stuff for future Oct 16 14:03:57 mickeyl: yeah, my level of python doesnt extend to open() :-) Oct 16 14:04:02 the first thing I look up in the manual, how to do system() Oct 16 14:05:20 psokolovsky: it work in 2.6.17 Oct 16 14:05:37 mickeyl: my a780 will soon be a devel phone Oct 16 14:08:21 koen: that's good. we really need more devels Oct 16 14:08:38 does it have a wm codec? Oct 16 14:09:31 XorA: thinking about that.... Oct 16 14:09:53 XorA: do you think it would be useful to have a class that gets you any entry out of a .pc flag? Oct 16 14:10:23 koen: can't bring myslef to pull it apart yet, i like the viewer app too much Oct 16 14:10:23 XorA: afaik it uses a custom asic: http://wiki.openezx.org/PCAP2 Oct 16 14:10:26 mickeyl: I dont have another use now, but that would be the generic extension and makes sense Oct 16 14:10:34 XorA: right. we might want to do such a class when we need it for the 2nd time... Oct 16 14:10:44 ade|desk: the firmware I have isn't made for phone usage Oct 16 14:10:45 mickeyl: so do it now :-) Oct 16 14:10:50 heh Oct 16 14:10:57 ade|desk: so I ordered a nokia 6021 to make calls with Oct 16 14:10:59 premature generalization is evil Oct 16 14:11:00 i learned that Oct 16 14:11:22 03pfalcon 07org.oe.dev * r748258db... 10/ (4 files in 2 dirs): machine/include: Unbreak PREFERRED_VERSION_orinoco-modules damaged in 514b31d89c294ce11810ee76a2138d455216409b. Oct 16 14:11:28 mickeyl: although co gtk-binver.bbclass -> get-pc-entry.bbclass and a small edit is easy later Oct 16 14:12:13 discovered recording calls with the a780 on saturday, rather useful feature Oct 16 14:12:45 must remember to say, 'your call may be recorded for training purposes' Oct 16 14:14:03 haha Oct 16 14:14:24 "I am trying to load my linux 2.6.11..." -> "Linux version 2.4.25-vrs1-cnxt " Oct 16 14:14:32 koen: just noticed that Oct 16 14:14:39 ~lart idiots Oct 16 14:14:39 * ibot takes large quantities of Krispy Kream donuts and stuffs them one after another down idiots's throat until idiots puts on 150lbs Oct 16 14:16:21 yo ar Oct 16 14:16:24 03mickeyl 07org.oe.dev * r593b9e6f... 10/ (3 files in 3 dirs): Oct 16 14:16:24 add gtk-binver parsing gtk_binary_version out of .pc file. close #1455 Oct 16 14:16:24 NOTE1: patch based on work of XorA Oct 16 14:16:24 NOTE2: we might want to make this more generic (i.e. get-pkgconfig-entry) eventually Oct 16 14:16:24 NOTE3: please adapt this to other gtk-engine versions as suitable Oct 16 14:16:27 XorA: here we are. please review how i transformed your class ;) Oct 16 14:17:13 hey Oct 16 14:18:08 mickeyl: cheers for that Oct 16 14:20:48 mickeyl: shouldn't that bbclass inherit pkgconfig? Oct 16 14:21:06 no, why? Oct 16 14:21:07 mickeyl: and DEPENDS on gtk+? Oct 16 14:21:56 ah right, 'inherit pkgconfig' us just the sed-munging Oct 16 14:22:19 but still, the class should make sure the .pc exists before trying to access it, right? Oct 16 14:22:25 koen, I'm working on adding tslib support to directfb and I started with the new version 1 of directfb, I'm getting an error about the fb. when I google it I find that you incountered the same error message and fixed it: Oct 16 14:22:31 koen: that's what the 'try' is fo Oct 16 14:22:32 r Oct 16 14:22:49 koen, here is the error: DirectFB/FBDev: Could not retrieve palette for backup! Oct 16 14:22:59 "Better to ask for forgiveness than to ask for permission" is one of python idioms Oct 16 14:23:04 koen, do you remember what you had to do to fix it? Oct 16 14:23:51 Gerrath: iirc I just fixed up the recipe to install more files Oct 16 14:23:54 koen: hmm, I was just looking at and os.popen(), but os.system() might be an alternative. Oct 16 14:26:28 mickeyl: looks good Oct 16 14:27:09 koen, and that fixed up the palette error message or just the other errors? Oct 16 14:27:31 fixing up the recipe is all I remember doing to directfb Oct 16 14:30:29 * mrz80 is back (gone 00:47:27) Oct 16 14:32:51 Gerrath: I had directfb touchscreen working on event interface on c7x0 with no such problems Oct 16 14:33:39 XorA, Directfb 0.9.25 or Directfb 1.0-0 RC1? Oct 16 14:33:54 Gerrath: 0.9.25 Oct 16 14:34:11 Gerrath: calibration done using a patch on the mailing list a few months ago Oct 16 14:34:57 XorA, I have too, I'm trying to get Directfb 1 working now, but also at the same time I'm adding tslib support to it.. I might just take a step back and work from 0.9.25 since it looks like the input interface has not change much. Oct 16 14:38:31 XorA: what do you think about install Xw100 on c7x0 and hx4700 by default? Oct 16 14:39:30 koen: works for me, just put somewhere in release notes than mplayer -vo w100 will explode Oct 16 14:39:31 ~praise ohand Oct 16 14:39:32 All hail ohand! Oct 16 14:39:57 mickeyl: if you tell us why, you'd have to kill us? Oct 16 14:40:09 not this time Oct 16 14:40:12 watch the next commit... Oct 16 14:40:19 * koen watches Oct 16 14:41:43 03mickeyl 07org.oe.dev * rc996b60a... 10/ (1 packages/xoo packages/xoo/xoo_svn.bb): Oct 16 14:41:43 add xoo, a GTK2 based graphical wrapper around a windowed X Server. It is intended for embedded developers that want to simulate a target Oct 16 14:41:43 device (with an accurate display size, working hardware buttons, etc) on a desktop machine. Oct 16 14:41:56 :) Oct 16 14:41:58 :D Oct 16 14:42:49 i've tftp/ping that work , but dropbear/telnetd/wget/ftpget dont Oct 16 14:42:54 XorA: hmmm, let me first figure out why Xw100 doesn't get built anymore Oct 16 14:43:05 someone has an idea where can i check ? i've no idea Oct 16 14:43:27 seems like udp work , but not tcp Oct 16 14:43:45 Genesis: any firewall ? Oct 16 14:43:52 no Oct 16 14:44:07 Who knows why we have module_autoload_* and not MODULE_AUTOLOAD_*? Oct 16 14:44:11 Genesis: weird ... check your kernel compile option Oct 16 14:44:20 oki , very weird Oct 16 14:44:26 thx Oct 16 14:44:59 Genesis: the only time I saw that was someone closed all tcp on a firewall in all directions Oct 16 14:45:34 CosmicPenguin: good morning Oct 16 14:45:50 morning Oct 16 14:47:24 03koen 07org.oe.dev * r610ec2ec... 10/ (1 conf/machine/hx4700.conf): hx4700: prefer the Xw100 x server Oct 16 14:49:09 http://www.openembedded.org/bonsai/query/description/angstrom/ Oct 16 14:49:13 * koen hugs bonsai Oct 16 14:50:27 damn, I get NOTE: preferred version 2.5 of glibc not available Oct 16 14:51:33 VoodooZ, that's "normal" Oct 16 14:51:43 koen, can bosai do diffs? Oct 16 14:51:53 psokolovsky, normal? Oct 16 14:52:15 VoodooZ, yes, message is normal. it still could pick up 2.5 ;-) Oct 16 14:52:17 eventhough I configured the prefered to 25. Oct 16 14:52:20 psokolovsky: that's on the TODO :) Oct 16 14:52:24 ok. thanks. Oct 16 14:52:27 VoodooZ, no, "normal" in quotes Oct 16 14:52:31 hehe Oct 16 14:52:32 koen, cool Oct 16 14:55:21 psokolovsky: http://www.openembedded.org/bonsai/query/file/MAINTAINERS/?exact=1&scm_id=3 Oct 16 14:57:55 koen, yes, being able to query like that is a great relief. Oct 16 14:58:30 hi Crofton Oct 16 14:58:49 gm Oct 16 14:59:31 merges could be filtered even ;-) Oct 16 15:02:47 hrw|work, I am going to flash my osk with some images I made over the weekend Oct 16 15:02:59 this should check the 18+git kernel you made Oct 16 15:08:14 Crofton: thx Oct 16 15:08:38 zecke: FYI: although the icecc.bbclass runs ok, it seems to fail populate the tar.bz2 file correctly. The popen()s seem to fail in create_env. Oct 16 15:09:28 koen, hrw|work: Who knows why we have module_autoload_* and not MODULE_AUTOLOAD_*? Oct 16 15:10:16 psokolovsky: I do not know why Oct 16 15:10:41 easier on the eyes probably Oct 16 15:11:01 XBITBAKE Transitional 1.8.0 ;) Oct 16 15:12:34 koen, hrw|work: Ok to prepare patch? Oct 16 15:15:47 nope. It's building the default glibc-2.3.5+cvs20050627-r12. Oct 16 15:20:05 ~botmail for likewise okay Oct 16 15:23:39 a+ Oct 16 15:33:50 koen: is there a special cert for merges? Oct 16 15:35:17 zecke: no, there isn't Oct 16 15:35:59 :( Oct 16 15:37:27 Re: ethernet support on linux 2.6.27 Oct 16 15:37:29 :D Oct 16 15:38:10 Re: ethernet support on linux 2.6.27 ??????? Oct 16 15:38:33 yes Oct 16 15:38:36 lakml Oct 16 15:38:46 nice Oct 16 15:42:11 cu Oct 16 15:57:28 hmmm ext4 in -mm ... Oct 16 15:57:45 yeah Oct 16 15:57:56 s/ext3/ext4/g Oct 16 15:58:04 s/filenode/extent/g Oct 16 15:58:06 done Oct 16 15:58:13 viewmtn is so slow ..... Oct 16 15:58:20 koen this is the only change? Oct 16 15:58:42 that's what I read on the internet Oct 16 15:58:52 it's backward compatible with ext3... Oct 16 15:59:10 it's more like ext3.1 Oct 16 16:02:59 chouimat|busy: it's still a nice improvement over ext3 Oct 16 16:03:25 koen I'm currently using JFS and XFS on most of my computers Oct 16 16:04:04 ah i loved jfs on OS/2 Oct 16 16:04:53 ade|desk: I have to reinstall Parallels so I can reinstall OS/2 :) Oct 16 16:06:21 i need to see if vmware will take OS/2, we have a couple of 2.11 around working the old telephone stuff Oct 16 16:06:51 i lost interest when IBM went all eComstation weirdness Oct 16 16:07:44 still can't believe IBM PC killed IBM OS div Oct 16 16:07:44 ade|desk: get parallels workstation (only 50$) it support OS/2 Oct 16 16:08:55 is it worth it if i dont have VT support Oct 16 16:10:20 ade|desk: I don't have a cpu with VT ... so I will say yes ... I have parallels and vmware I find parallels lighter (ok sometimes slower than vmware but it's ok) Oct 16 16:10:50 ok time to go fetch food Oct 16 16:13:36 os/2 warp was pretty neat Oct 16 16:19:19 warp aurora was better Oct 16 16:21:18 ade|desk: 4.52? Oct 16 16:22:47 zecke: this might interest you: http://pullcord.laptop.org/tinderbox/ Oct 16 16:23:15 hrw? Oct 16 16:26:13 4.51 i thought Oct 16 16:31:19 03crofton 07org.oe.dev * r75021555... 10/ (1 packages/linux/linux-omap1.inc): Oct 16 16:31:19 linux-omap1 : Fix kernel entry point so that linux-omap1_2.6.18+git boots Oct 16 16:31:19 on the OSK. Oct 16 16:34:12 Crofton: did you find the golden rev yet? Oct 16 16:34:17 yeah Oct 16 16:34:23 well one of them :) Oct 16 16:34:46 when hrw made the inc, he took the entry point to mkimage from a non working bb file Oct 16 16:34:56 or, the entry point has changed for some reason Oct 16 16:35:17 anyway, the _2.6.18+git kernel is booting on my osk Oct 16 16:36:18 home time Oct 16 16:36:22 cya all Oct 16 16:40:11 koen, hopefully that will be the last kernel push for a bit, at least until I care about the dspgw Oct 16 16:46:37 bbl Oct 16 16:46:47 hi Oct 16 16:46:51 Crofton: thx for fixing Oct 16 16:47:27 np Oct 16 16:47:29 it boots Oct 16 16:47:39 now I am trying to figure out how to mount the microdrive Oct 16 16:48:12 I get the pcmcia registering device .... pcmcia0.0 Oct 16 16:48:23 Crofton: run pcmciautils or pcmcia-cs and it should appear as /dev/hda* Oct 16 16:48:24 just not sure what the magic is to get stuff to mount anymore Oct 16 16:48:29 ide_cs loadded? Oct 16 16:48:44 Crofton: did you got ide_cs loaded and /sys/block/hda entries? Oct 16 16:48:44 ide_cs is loaded Oct 16 16:48:44 mount /dev/hda1 /mount/point Oct 16 16:48:59 they might have loaded when I booted Oct 16 16:49:00 hrw: maybe he needs one of them hrw patches :) Oct 16 16:49:16 Crofton: (pc)cardctl ident please Oct 16 16:49:30 hmmm Oct 16 16:49:33 brb Oct 16 16:49:33 hrw: cardctl doesn't work anymore in 2.6.18 Oct 16 16:49:37 how do I get that? Oct 16 16:49:55 Crofton: type 'pccardctl ident' in your console :) Oct 16 16:50:01 deprecated message from cardctl info Oct 16 16:50:41 Crofton: then install pcmciautils Oct 16 16:50:55 sounds like you need PCMCIA_MANAGER = 'pcmcia-utils' Oct 16 16:51:18 hangon Oct 16 16:51:27 trying to get irc going on the machine with the serial port Oct 16 16:51:36 cardctl ident gives me something Oct 16 16:51:57 Socket 0: Oct 16 16:51:57 product info: "WEIDA", "TWTTI", "" Oct 16 16:51:57 manfid: 0x000a, 0x0000 Oct 16 16:51:57 function: 4 (fixed disk) Oct 16 16:53:00 that still works on 2.6.18? Oct 16 16:53:10 I could have sworn that ioctl was removed Oct 16 16:53:17 I do get a warning Oct 16 16:53:31 we shoudl switch it Oct 16 16:53:50 pcmcia: Detected deprecated PCMCIA ioctl usage from process: cardctl. Oct 16 16:53:50 pcmcia: This interface will soon be removed from the kernel; please expect breakage unless you upgrade to new tools. Oct 16 16:53:50 pcmcia: see http://www.kernel.org/pub/linux/utils/kernel/pcmcia/pcmcia.html for details. Oct 16 16:54:30 http://pastebin.ca/205214 Oct 16 16:55:43 Crofton: ok. it need ide-cs patching... Oct 16 16:55:50 hrw, should we switch topcmciautils before spending much time on this? Oct 16 16:56:10 Crofton: kernel need patching - I will prepare patch Oct 16 16:56:18 ok Oct 16 16:56:26 let me know, and I will test Oct 16 16:56:36 I'm used to patching ide-cs and other -cs drivers Oct 16 16:56:45 do the linux-omap guys know about this? Oct 16 16:57:44 its mainline fix Oct 16 16:58:14 so it will be picked up when the omap tree is ynced with mainline? Oct 16 16:58:22 should Oct 16 16:58:44 bbiab, need some food b4 a meeting Oct 16 16:59:33 food here too Oct 16 16:59:36 bbl Oct 16 17:13:10 03pfalcon 07org.oe.dev * r66551f85... 10/ (3 files in 2 dirs): Oct 16 17:13:10 pocketsphinx 0.2.1: New version. Oct 16 17:13:10 * 0.2 is removed, as it contained bug which likely precluded it to work correctly Oct 16 17:13:10 at all. Oct 16 17:14:29 re Oct 16 17:16:50 psokolovsky: wanna learn a new trick? Oct 16 17:17:14 psokolovsky: ${PN}-${PV} is the same a ${P} :) Oct 16 17:17:15 koen_: sure ;-) Oct 16 17:18:45 koen_: well, I already heard about. but I wonder, how many other people know that ;-), and wouldn't explictly having PV be more easy to understand? Oct 16 17:19:07 it's all in conf/bitbake.conf Oct 16 17:19:10 Is there an example bb file I can look at that compiles different ipk (or at least adapts) for different screen resolution? Oct 16 17:19:47 Laibsch: http://www.openembedded.org/repo/org.openembedded.dev/packages/matchbox-panel/matchbox-panel.inc Oct 16 17:20:02 thanks koen. Oct 16 17:20:22 hrw, should I add the PCMCIA_MANAGER line to openomap distro? Oct 16 17:20:30 Crofton: yes Oct 16 17:22:31 koen: OK, so you give extra compile instructions for small screen devices. Couldn't something like this also be specified and centralized in the machine config? Just a thought. Oct 16 17:23:06 Laibsch: GUI_MACHINE_CLASS = "smallscreen" thing? Oct 16 17:23:36 anyway I would prefer rather MACHINE_FEATURES = "qvga/vga" thing Oct 16 17:23:43 Crofton: pull and rebuild Oct 16 17:23:43 Either that or something in collie.conf is what I thought to be the appropriate place. Oct 16 17:23:45 where is the "proper" place to put PCMCIA_MANAGER .... Oct 16 17:23:47 hrw: no Oct 16 17:23:55 hrw: that ignores DPI Oct 16 17:24:10 Laibsch, as I mentioned, gaim would be such sample for gtk+-based app. Oct 16 17:24:17 03hrw 07org.oe.dev * raa1ee9e2... 10/ (3 files in 3 dirs): linux-omap1: add one CompactFlash card into ide-cs Oct 16 17:24:35 psokolovsky: You mentioned this? Sorry, I must have forgotten. Was this in private chat? Oct 16 17:24:43 Crofton: distro Oct 16 17:24:58 no particular section ... Oct 16 17:25:02 koen_: smallscreen/bigscreen too Oct 16 17:25:03 Laibsch, said that at the night, maybe you missed that ;-) Oct 16 17:25:10 I put it at the end for now Oct 16 17:25:29 psokolovsky: The night? Oct 16 17:25:34 Sory, I must be dense. Oct 16 17:25:51 Crofton, I tried to sort h2200.conf per logical sections. Oct 16 17:26:33 grr, multiple heads? Oct 16 17:26:50 Crofton, PCMCIA_MANAGER pretty well selects specific package implementation of a feature, so rather can go togther with PREFERRED_* stuff Oct 16 17:28:23 zecke: ping Oct 16 17:29:05 Crofton: merging Oct 16 17:29:19 I updated from your commit Oct 16 17:29:31 k Oct 16 17:29:32 I'll get rebuild going b4 my meeting Oct 16 17:30:15 and now my distro Oct 16 17:30:42 03hrw 07org.oe.dev * r5dee8b7d... 10/ (1 conf/distro/celinux-test.conf): celinux-test: distro for CELF testlab Oct 16 17:31:37 03hrw 07org.oe.dev * rc45e9d1b... 10/ (1 MAINTAINERS): MAINTAINERS: added celinux-test to my distros Oct 16 17:31:58 can I run celinux on my toaster? Oct 16 17:32:07 celinux? Oct 16 17:32:23 consumer electronic linux forum (CELF) Oct 16 17:32:24 Consumer Electronics Oct 16 17:32:31 +s Oct 16 17:32:31 ah Oct 16 17:32:34 like I said, toasters :) Oct 16 17:32:39 lol Oct 16 17:32:50 koen: Oh, I thought real-time vacuum cleaners only Oct 16 17:33:19 hrw, rebuilding Oct 16 17:34:01 I made some notes on flashing at http://ossie.mprg.org/OSK_Notes Oct 16 17:34:03 my 'hifi' audio runs celinux with 2.6.27 (which has ethernet problem like someone wrote on lakml) Oct 16 17:34:10 this has improvedmy success rate Oct 16 17:34:16 Crofton: thanks Oct 16 17:34:36 Crofton: can I copy it? Oct 16 17:34:48 sure Oct 16 17:36:07 I just put it together to help me test this stuff :) Oct 16 17:36:33 once I get out of autotools hell I will add stuff on turning the osk into a software defined radio Oct 16 17:37:25 /msg nickserv set unfiltered on Oct 16 17:37:32 whops, leading space Oct 16 17:39:55 buy anything? Oct 16 17:40:11 nothing exciting... food for tonite Oct 16 17:40:53 probably should've /nick mickey|grocery Oct 16 17:41:00 i need shoes though Oct 16 17:41:07 i have gadgets, but no shoes Oct 16 17:41:09 *sigh* Oct 16 17:41:14 Is it true that there are no input methods available for GPE? No fitaly or some such? Where is the bb for the on-screen keyboard which I think I remember I saw? Oct 16 17:41:17 mickeyl: what license your fosdem2005 presentation is? Oct 16 17:41:26 mickeyl rieker shoes are nice Oct 16 17:41:28 hrw: CreativeCommons Oct 16 17:41:32 Laibsch: matchbox-keyboard? Oct 16 17:41:37 * XorA|gone has scary thoughts of booting linux on a pair of shoes Oct 16 17:41:41 hrw: just please credit me :) Oct 16 17:41:53 hrw: thanks. Oct 16 17:42:05 mickeyl: I'm writing my presentation for my lecture which I will have this Wednesday Oct 16 17:42:15 cool. good speed! Oct 16 17:42:42 why did psmisc get sucked in? Oct 16 17:42:50 mickeyl: good? It was discussed in August ;) Oct 16 17:43:00 anyway, I will test new images after my meeting Oct 16 17:43:11 hrw: good speed is an american phrase Oct 16 17:43:21 hrw: meaning "much success" Oct 16 17:43:28 has actually nothing to do with tempo :) Oct 16 17:43:30 ah those idioms.. Oct 16 17:43:33 hehe Oct 16 17:43:46 my english is too dictionaried Oct 16 17:44:01 ~lart my English teachers Oct 16 17:44:01 * ibot lowers my English teachers's priority Oct 16 17:44:19 actually i think it's called godspeed Oct 16 17:44:25 but i always found that to be an odd word Oct 16 17:44:41 probably i should stick with "good luck" Oct 16 17:44:46 which seems to be more comprehensible Oct 16 17:45:18 ;) Oct 16 17:45:20 godspeed is the only think I can think of Oct 16 17:46:30 btw: I HATE Polish translation of crosscompilation Oct 16 17:46:35 or embedded Oct 16 17:46:53 hrw: tell us :-) Oct 16 17:47:08 likewise: kompilacja skrosna + systemy wbudowane Oct 16 17:47:09 "god speed" is an english expression, somewhat short for "god speed you on your way" Oct 16 17:47:23 aah Oct 16 17:47:23 originally meaning "good luck" but specifically "good luck when travelling" Oct 16 17:47:24 wbudowane == embedded in direct translation Oct 16 17:47:25 What is the proper variable or way to find the bin directory where the cross-compiler is, from within a .bbclass? Oct 16 17:47:37 likewise: CROSS_DIR? Oct 16 17:47:39 and "travel quickly, before anything bad happens to you enroute" Oct 16 17:48:07 hrw: why isn't it printed during a bitbake -D -v, or am I blind? Oct 16 17:48:25 likewise: bitbake -e try Oct 16 17:48:37 bitbake -e umba|grep cross Oct 16 17:48:50 hrw: the english expression "goodbye" was originally a contraction for "god be with you" Oct 16 17:48:51 hrw: it's a bitbake variable, not an environment variable? Oct 16 17:49:21 likewise: env Oct 16 17:49:32 oops. Oct 16 17:49:35 i meant that for mickeyl Oct 16 17:49:38 hrw: sorry Oct 16 17:49:41 OT: OpenEmbedded as 'he' or as 'she'? :D Oct 16 17:49:45 hrw: thanks Oct 16 17:49:50 or as 'it' Oct 16 17:50:20 hrw: it Oct 16 17:51:36 i realize i should've taken my english classes much more serious. unfortunately i was not much interested in school when i was young Oct 16 17:52:16 hrw: Clearly she. Openembedded was always finicky and difficult. Oct 16 17:52:26 ggilbert: and ate lot of resources Oct 16 17:52:31 mickeyl he but now your are nearly ph.d Oct 16 17:53:52 woglinde: and we can call him 'doc' Oct 16 17:53:58 hehe Oct 16 17:53:59 mickeyl: what's up doc? Oct 16 17:54:06 mickeyl: this isn't stuff you would find in a normal english class Oct 16 17:54:30 by german laws I must correct that you are not allowed to call me Dr. until I have made the public defense... :/ Oct 16 17:54:41 raduga: fair enough Oct 16 17:54:42 mickeyl: these word origins and derivations are centuries old.... and very few modern speakers even are aware of them Oct 16 17:54:59 mickeyl: And you had to try to cram both English and German into your head...not an easy thing...:) (espicially if you have some lysdexia :) ) Oct 16 17:55:04 raduga: Nowadays I'm into these things... it's never too late to become interested in stuff Oct 16 17:55:07 hrw: +# use softfloat binaries to gain speed on ARM machines Oct 16 17:55:07 +TARGET_FPU = "soft" Oct 16 17:55:20 hrw: I think you mean TARGET_FPU_arm Oct 16 17:55:23 hvontres|poodle: heh Oct 16 17:55:31 koen_: good point Oct 16 17:55:50 hrw: look at how angstrom does it :) Oct 16 17:56:01 mickeyl: but many of the greetings dating back to the middle ages, throughout europe, originate in invocations to God Oct 16 17:56:10 * hvontres|poodle remembers the joys of being a Left handed person with a fountain pen.. Oct 16 17:56:30 mickeyl: in Spanish, when someone leaves, you tell them "adios!" Oct 16 17:56:38 or 'to the gods' Oct 16 17:56:44 (literally) Oct 16 17:57:03 raduga: like in Polish "z bogiem" (with god) Oct 16 17:57:09 or they say 'pendejo', which means something different :) Oct 16 17:57:20 03hrw 07org.oe.dev * r7d7875c6... 10/ (1 conf/distro/celinux-test.conf): celinux-test: softfloat only for ARM machines Oct 16 17:57:36 koen: i presume if they are koen, they are more likely to say that Oct 16 17:57:47 what's the package name for linus' git called in sarge? Oct 16 17:57:53 git-core Oct 16 17:57:56 ah Oct 16 17:57:56 thanks Oct 16 17:58:14 20:17 hrw@work:hrw$ apt-cache search ^git$ Oct 16 17:58:14 git - GNU Interactive Tools, a file browser/viewer and process viewer/killer Oct 16 17:58:23 ~lart linus Oct 16 17:58:24 * ibot grabs a large, mis-shapened log, with squirrels, and beats linus until only the nuts remain ... which the squirrels run off with Oct 16 17:58:49 heheh Oct 16 17:59:43 hrw: hehe Oct 16 18:04:58 mickeyl: we need 'OE Myths' page somewhere Oct 16 18:05:06 1. OE eat lot of memory Oct 16 18:05:19 Crofton: http://www.openembedded.org/bonsai/query/description/omap/ Oct 16 18:05:36 2. target device need X11 to use python Oct 16 18:06:05 3. pdaXrom > oe Oct 16 18:06:11 (or is that myth backwards?) Oct 16 18:06:43 raduga: we decided to not make laugh from pdaX Oct 16 18:07:02 "no need to flog a dead horse" as someone said Oct 16 18:07:21 hrw: ok Oct 16 18:07:21 koen: ok Oct 16 18:07:37 hrw: sounds good. just add one to the wiki and we'll enhance Oct 16 18:08:20 and a 5 pages describing how these 'myths' are untrue in various languages and wordings Oct 16 18:08:20 added to todo Oct 16 18:08:33 NOTE: these myths are untrue Oct 16 18:08:48 NOTE 2: Really, they aren't true Oct 16 18:09:00 NOTE 3: if you are a manager: Really, they aren't true Oct 16 18:09:04 etc.. Oct 16 18:14:17 hmm Oct 16 18:14:20 ~lart debian Oct 16 18:14:20 * ibot cuts debian into thin stripes Oct 16 18:14:34 how is the recommended way to get unstable packages into a stable distro? Oct 16 18:14:42 stil no volunteers for a press release? Oct 16 18:14:43 aptitude install git-core wants to update libc6 Oct 16 18:14:48 which i can't allow Oct 16 18:15:01 mickeyl: download the source and rebuild the package Oct 16 18:15:08 apt-get source git-core Oct 16 18:15:15 cd git-core-ver Oct 16 18:15:20 dpkg-buildpackage Oct 16 18:15:28 koen: I can mail linuxdevices.com Oct 16 18:15:32 ggilbert: thanks. i'm a debian newbie but i need to use it on one machine Oct 16 18:15:41 wasn't there a -b to make debian build it as well? Oct 16 18:15:51 there might be, but I'm a creature of habit Oct 16 18:15:57 mickeyl: forbid or hold the old libc Oct 16 18:16:02 koen: we lack SuccessStories page now? Oct 16 18:16:15 mickeyl: it'll work about 75% of the time, it just gets to be trouble when there are real dependency changes Oct 16 18:16:26 * koen looks at http://www.openembedded.org/wiki/SuccesStories Oct 16 18:16:32 zecke: how would i do that? Oct 16 18:16:45 ah.. I wrote Succes*s* Oct 16 18:17:14 mickeyl: sudo aptitude Oct 16 18:17:21 mickeyl: /^glibc Oct 16 18:17:24 he zecke Oct 16 18:17:27 mickeyl: and then use the menu to forbid it Oct 16 18:17:44 zecke: ok, i'll try that. thanks Oct 16 18:18:06 mickeyl: or remove unstable/edgy from your sources.list Oct 16 18:18:43 edgy contains what? Oct 16 18:19:01 zecke: edgy is ubuntu not debian Oct 16 18:19:18 hrw: / == or Oct 16 18:19:42 ah Oct 16 18:19:53 * hrw goes back to OO Writer and presentation Oct 16 18:21:05 bye hrw Oct 16 18:21:24 hrw: Oh, you stuck with it? ;-) Oct 16 18:23:14 Laibsch: it works and do what I want Oct 16 18:23:22 Laibsch: then I will generate PDF and will show it Oct 16 18:23:49 godspeed ;-) Oct 16 18:24:35 What's the full path to the cross-compiler? Can I assume in bbclass that it's always $CROSS_DIR/$TARGET_SYS/bin/${TARGET_PREFIX}gcc ? Oct 16 18:24:52 likewise: why do you ask? Oct 16 18:24:57 likewise: thanks for looking into it Oct 16 18:25:15 zecke: I am trying to get icecc.bbclass right Oct 16 18:25:38 likewise: I assumed that Oct 16 18:25:50 likewise: I think we can almost assume that Oct 16 18:26:12 zecke: the current error is that the string argument to the first popen() is not in the PATH. Or more correct: the PATH does not yet include the cross-compiler at that point. Oct 16 18:26:12 likewise: we have to look if angstrom changes TARGET_PREFIX Oct 16 18:27:07 likewise: ah right, let us set it Oct 16 18:27:33 likewise: there must be an export PATH in some config file and then we can just set it os.env[] Oct 16 18:28:35 zecke:yes, that might be a neater solution. Oct 16 18:29:28 zecke: except that I have no idea where the export must go, in a neat way Oct 16 18:29:29 likewise: and reset it afterwards Oct 16 18:29:43 likewise: use python os to manipulate it Oct 16 18:29:47 likewise: it is quite easy Oct 16 18:30:10 likewise: os.environ['PATH'] = 'blub' Oct 16 18:31:05 likewise: conf/bitbake.conf:214:export PATH_prepend = "${STAGING_BINDIR}/${HOST_SYS}:${STAGING_BINDIR}:${CROSS_DIR}/bin:" Oct 16 18:31:28 so you can safely do os.environ['PATH'] = bb.data.getVar('PATH', d, True) Oct 16 18:31:45 likewise: probably safe the old one before, and reset it Oct 16 18:31:56 zecke: or os.environ['PATH'] = os.environ['PATH']+':'+os.path.join(ice_dir, target_sys, 'bin') ? Oct 16 18:32:28 zecke: ah, ok, I see now. copy the PATH bitbake variable into the environment variable. Oct 16 18:34:19 likewise: if you test already, make sure to check that arm-angstrom-linux-gcc works as well :) Oct 16 18:34:45 hrw: so you did get an XL :) Oct 16 18:35:03 * zecke waits for his google t-shirt Oct 16 18:35:18 * koen waits for his XL google shirt Oct 16 18:35:20 I won't talk about breach of contract otherwise google will kick me from the mailinglist as well Oct 16 18:35:36 zecke: I am a sucket at first time right coding. I do too many test-cycles, that's something I am working on :-) Oct 16 18:35:37 zecke: you aren't called thomas lord :) Oct 16 18:35:38 * chouimat got his XL google t-shirt today Oct 16 18:35:42 s/sucket/sucker Oct 16 18:36:07 where u got the google t-shirt ? Oct 16 18:36:13 likewise: Test is good Oct 16 18:36:21 gremlin[it]: summer of code :) Oct 16 18:36:38 ahhh yes :) Oct 16 18:37:04 gremlin[it]: currently wearing it ... Oct 16 18:37:35 I want my cell.conf! :-) Oct 16 18:38:00 * gremlin[it] go to wear RedHat t-shirt :) :) :) Oct 16 18:51:33 Anybody have any docs for devshell ??? Oct 16 18:52:47 bitbake devshell ; . path/to/devshell Oct 16 18:54:10 koen: and then just run make and friends ? Oct 16 18:54:45 yes Oct 16 18:54:59 koen: that was too simple...:) Oct 16 18:55:15 :) Oct 16 18:55:33 we should merge the devshell stuff from poky Oct 16 18:59:13 <|hvontrespoodle> re Oct 16 19:11:16 is this bonsai thing new? Oct 16 19:11:29 yes Oct 16 19:12:49 * likewise is rebooting - my PID counter runs circles with zombies Oct 16 19:14:15 * koen wonders when zecke will upload the sources to http://mtn-host.prjek.net/projects/tinderbox/ Oct 16 19:20:57 OSK just mounted my microdrive Oct 16 19:21:11 nice Oct 16 19:21:36 Filesystem 1k-blocks Used Available Use% Mounted on Oct 16 19:21:37 /dev/mtdblock3 30464 5840 24624 19% / Oct 16 19:21:37 /dev/mtdblock3 30464 5840 24624 19% /dev/.static/dev Oct 16 19:21:37 tmpfs 10240 12 10228 0% /dev Oct 16 19:21:37 /dev/hda1 1006984 406104 549728 42% /media/hda1 Oct 16 19:22:16 1G one? Oct 16 19:22:21 yeah Oct 16 19:22:45 * v8jlene is back Oct 16 19:23:14 v8jlene: wb Oct 16 19:23:18 I like to swap to it :) Oct 16 19:23:25 morning Oct 16 19:23:35 I have to finish reformatting spitz microdrive.. Oct 16 19:23:51 but what I like best is that all I had to do to get here is boostrap-image and flash Oct 16 19:24:26 thanks guys! Oct 16 19:25:08 Crofton: it automounted? Oct 16 19:25:15 yes ... Oct 16 19:25:19 cool Oct 16 19:25:20 cool Oct 16 19:25:31 /media/hda1 Oct 16 19:25:34 koen: BTW, glibc 2.5 builds and boots fine on sh4. Only coreutils didn't build (which I'll fix today). Oct 16 19:26:06 v8jlene: the only problem I could spot is that procps keep segfaulting on i686 Oct 16 19:26:24 somewhere in the get_uptime function Oct 16 19:27:20 koen: I'll take a look at that if I get time. Just testing linus git on sh4 at the moment. For the first time ever sh is fully merged up (and looks to work). Oct 16 19:27:41 v8jlene: first time in 2.6 tree? Oct 16 19:27:57 ah, cool Oct 16 19:28:03 less working moving patches around :) Oct 16 19:28:34 hrw: 2.6 has built before but needed a cvs drop in over latest git. Since the main sh dev got hired by the sh manufacturer things have really moved along! Oct 16 19:28:45 sh? Oct 16 19:28:56 superH Oct 16 19:28:56 Crofton: Hitachi SuperH CPU family Oct 16 19:29:00 03crofton 07org.oe.dev * r90c0d1e2... 10/ (1 conf/distro/openomap.conf): openomap.conf : Change openomap distro from pcmcia-cs to pcmciautils. Oct 16 19:29:00 ah Oct 16 19:29:00 Crofton: SuperH (sh4 for me). Oct 16 19:29:01 from hitachi iirc Oct 16 19:30:05 Originally from Hitachi, now owned by Renesas. Renesas hired the main linux sh developver about four weeks ago ;) Oct 16 19:30:52 v8jlene: what's the strength of a superh cpu? Oct 16 19:31:05 koen: over arm? Oct 16 19:31:15 over any cpu Oct 16 19:31:52 gpsd refuse to build cause of missing dbus-0.23.4 . dbus version in OE is 0.71 to 0.94 Oct 16 19:32:31 we dropped old dbus Oct 16 19:32:43 most of distros dropped it too Oct 16 19:32:49 out-of-OE distros Oct 16 19:33:53 Well, the main advantage for me is that I have a superh board and no arm boards ;) Oct 16 19:35:07 hrw: Ok, hrw, but what do we do then to have gps support ? We need gpsd dont we? Gpsdrive depends on gpsd Oct 16 19:35:21 gpsdrive is poop Oct 16 19:35:26 mickeyl: do you have your oedem presentation online somewhere? Oct 16 19:36:00 goxboxlive, port gpsd to use contemporary dbus? Oct 16 19:36:21 koen: But wich gps application do you recommend. Oct 16 19:36:44 i could try if there is no other application Oct 16 19:37:27 goxboxlive, also, do you mean gpsd or gpsdrive? the first would hardly depend on dbus, I guess. Oct 16 19:37:51 Gpsdrive depends on gpsd, so i mean gpsd Oct 16 19:38:04 psokolovsky: you can query gpsd over dbus Oct 16 19:38:15 But if there is other gps applications for use in GPE i love to try them out Oct 16 19:39:07 03koen 07org.oe.dev * re8071b75... 10/ (1 packages/gpsd/gpsd_2.33.bb): gpsd: add 2.33 Oct 16 19:39:12 03koen 07org.oe.dev * r96fb4347... 10/ (1 packages/gpsd/gpsd_2.33.bb): gpsd: remove DEPENDS on dbus 0.23.ancient.obsolete Oct 16 19:39:12 In the gpsd.inc this line is written: DEPENDS = "dbus-0.23.4 ncurses" Could i only change that to the dbus version angstrom is using? Oct 16 19:39:27 koen: thx Oct 16 19:40:06 which ver of X.Org we have? Oct 16 19:40:07 7.1? Oct 16 19:40:45 7.0 and 7.1 Oct 16 19:40:50 feel free to add 7.2 :) Oct 16 19:42:01 is there a way to get encodings support in vim? Oct 16 19:42:25 hmm.. Oct 16 19:42:45 on Z, I mean :) Oct 16 19:43:03 or better, in OE :) Oct 16 19:45:37 Gin-geR: pull and rebuild vim Oct 16 19:45:44 gpsd and gpsdrive successfully build :-) Oct 16 19:45:52 hrw, Thank you Oct 16 19:45:55 Q: 'what is OE?' Oct 16 19:46:03 bit O tjoml O Oct 16 19:46:08 oops :) Oct 16 19:46:12 03hrw 07org.oe.dev * r77669519... 10/ (1 packages/vim/gvim_7.0.bb packages/vim/vim_7.0.bb): (g)vim 7.0: enable big features set Oct 16 19:46:13 A: An answer to distros which require 200MiB of space for base system. Oct 16 19:46:46 And a few gigs to build :) Oct 16 19:46:49 other answers? Oct 16 19:47:20 Gin-geR: I think that I can build bootstrap-image with 2G of storage for build Oct 16 19:48:19 Yes, sounds true. Oct 16 19:49:05 isn't OE is just almost self-contained build system for ipkg-based limited distros? Or is it not true? Oct 16 19:49:26 no true Oct 16 19:51:03 no true, or not true? Oct 16 19:51:44 there's no true on earth. Oct 16 19:51:50 nothing I mean Oct 16 19:52:02 not true Oct 16 19:52:20 Well, then I'm wrong. Oct 16 19:53:16 its in no way limited, and i added support to build dpkg based images even Oct 16 19:53:24 rpm wouldnt be that difficult Oct 16 19:54:05 Aha, then remove ipkg bit. That's the part I was in doubt about. Oct 16 19:54:10 kergoth: patches for rpm are in the bug tracker Oct 16 19:54:18 nice Oct 16 19:54:23 for ages Oct 16 19:54:27 so, not so nice Oct 16 19:54:29 you could remove the limited distros also Oct 16 19:54:33 has the dpkg stuff gone fully upstream from pokay yet? Oct 16 19:54:40 s/ay/y/ Oct 16 19:54:55 kergoth: I merged the apt and dpkg bits, but not the u-a stuff Oct 16 19:55:18 I updated apt iirc, so you could even merge back to poky :) Oct 16 19:55:30 ah cool Oct 16 19:55:38 i'll merge back to poky, then push the image bits to oe Oct 16 19:55:42 still need to figure out how to use existing distro package to build staging Oct 16 19:55:50 I think about that almost dailly Oct 16 19:56:02 isnt that what koen's packaged staging SoC project was about? Oct 16 19:56:24 sort of, p-s packages up OE built stuff and is limited to ipkg right now Oct 16 19:56:27 sort of - wasn't koen's project having to do with packaging what goes into staghing? Oct 16 19:56:55 Its the same thing we've discussed forever -taking "distro blah" and using that to seed stage Oct 16 19:57:19 if gcc wasn't so damn braindead about how it works, then we would probably be there Oct 16 19:58:07 well, it wouldnt be hard to add something to assume the tasks are complete if it sees the package already exists Oct 16 19:58:12 but thats dependent upon packaged staging Oct 16 19:58:46 p-s in in .dev now, it just needs a lot of people using it and reporting problems back in a cluefull way Oct 16 19:59:00 nice Oct 16 19:59:08 sweet Oct 16 19:59:12 * CosmicPenguin should play Oct 16 19:59:31 koen: 1. add stage-manager.bb and then add it into DEPENDS inside of p-s.bbclass Oct 16 19:59:33 itll be cool to have a pool of existing built packages for the toolchain and such that it pulls from if they've already been built Oct 16 19:59:39 then put that pool on an nfs server Oct 16 19:59:44 for the developers in a company Oct 16 20:00:01 nod - but the whole gcc / libpath issue still precludes that being a decent solution Oct 16 20:00:22 koen: 2. ipkg-utils with hardcoded patch usage Oct 16 20:00:23 not to mention the sundry bullshit with pc and other static path files Oct 16 20:01:11 hrw: add that to the p-s bug in bugzilla please Oct 16 20:02:06 when is ewi going away? Oct 16 20:02:30 raduga: thanks to bureacracy not very soon Oct 16 20:04:33 koen: we have such one? Oct 16 20:04:42 hrw: yes, for ages Oct 16 20:05:06 #354 Oct 16 20:05:18 yes Oct 16 20:05:24 I want to reserve #354x bugs for me Oct 16 20:07:38 hrw: will you fix all the bugs? Oct 16 20:07:53 raduga: who knows.. Oct 16 20:08:37 hrw: that's pretty ambitious :) Oct 16 20:10:01 CosmicPenguin: Someone told be crosstool have a script to relocate gcc Oct 16 20:10:21 only if the new dir is smaller then the original dir Oct 16 20:10:59 CosmicPenguin: Ah. I can guess why :-/ Oct 16 20:11:05 heh - yeah Oct 16 20:11:46 ok, we need to write some new functionality into gcc then... Oct 16 20:11:59 IIRC, somebody had started it, but I don't know how far it go Oct 16 20:12:00 t Oct 16 20:12:05 GCC_LIB_DIR or somesuch Oct 16 20:12:39 Maybe if I have a spare week sometime :) Oct 16 20:13:05 My scenario for using an existing distro goes something like this: Oct 16 20:14:00 You want to make a new package for your ubuntu daper system - so you build a staging directory with our own native tools, and the ubuntu compiler and glibc Oct 16 20:14:03 and then build as normal Oct 16 20:14:29 The only way I can think to pull it off now is to actually re-compile the compiler Oct 16 20:14:36 but that gets messy fast Oct 16 20:14:58 For that, can't we manage some clever ASSUME_PROVIDEDs? Oct 16 20:15:06 * koen sneaks in a "debian uses OE" note in a mail to LAK + gcc-cross + libc-alpha Oct 16 20:15:11 or is the issue that its "gcc" ? Oct 16 20:15:43 Obviously, clever ASSUME_PROVIDEDS will be needed somewhere along the line - but that seems like its doable Oct 16 20:16:11 Its even doable to automagically search packages for .pc files and other such and sed-fu them Oct 16 20:16:37 In my brain, the outstanding issue will be making gcc do the right thing Oct 16 20:17:05 CosmicPenguin: Its a nice idea and would be really useful, we just need some developers with some time to spend on it... Oct 16 20:17:10 indeed Oct 16 20:17:52 speaking of stuff like .pc files, mallum mentioned that our -native's should be searching the /usr pkg-config paths.. i dont think our native bbclass has that in its env var Oct 16 20:17:52 mickeyl: Can you mail me the notes you made on the bitbake session and oedem please? Oct 16 20:18:26 kergoth: Is that the reason for the qemu-native x11 issue? Oct 16 20:18:55 i'm not sure, but it wouldnt surprise me Oct 16 20:19:08 quick and easy fix that shouldnt hurt anything Oct 16 20:19:17 it certainly sounds likely... Oct 16 20:19:22 * kergoth nods Oct 16 20:19:39 CosmicPenguin: you see my botmail from the other day? Oct 16 20:19:46 kergoth: yes - thanks Oct 16 20:19:49 np Oct 16 20:20:45 need testers: http://www.sidebranch.com/openembedded/icecream/icecc.bbclass Oct 16 20:21:24 * Crofton is having terrible problems with some recipes finding headers and libraries installed on the build system /usr ...... Oct 16 20:22:44 RP: don't forget to ask mickeyl for the parses as well :) Oct 16 20:23:26 s/parses/parser/ Oct 16 20:23:26 koen: parses? Oct 16 20:24:07 They didn't look that cryptic - I just want to combine them with mine and make sure there's nothing I've missed :) Oct 16 20:24:27 kergoth: do you think the u-a changes in poky would be acceptable for OE, or would it need a proper review by knowledgeable people? Oct 16 20:24:45 RP: the parser for the fileformat :) Oct 16 20:25:10 koen: Ah, yes. I'm presuming he can mail me plain text :) Oct 16 20:25:20 i'm not sure, i think it gets the job done, but we need to double check that the commandline syntax and such is truly identical.. if not, they cant provides the smae thing Oct 16 20:25:31 koen: The u-a changes have circular dependency issues as well Oct 16 20:26:21 they do? Oct 16 20:26:50 What am I doing wrong that makes all non-latin utf show up as '?'? I have en_US.utf-8 locale Oct 16 20:26:56 kergoth: It breaks with bitbake trunk and debian.bbclass Oct 16 20:27:09 kergoth: I thought we'd discussed this? Oct 16 20:27:10 mickeyl: http://thread.gmane.org/gmane.comp.lib.glibc.alpha/11816/focus=11816 Oct 16 20:28:02 in console that is. Oct 16 20:28:39 koen: nice post Oct 16 20:28:55 I didn't ran at gcc being broken this time :) Oct 16 20:29:01 s/ran/rant/ Oct 16 20:31:08 koen: Interesting to know its the size of the shmmem that changes. Perhaps we'll fix glibc 2.4 yet... Oct 16 20:31:50 RP: we could, but not till glibc 2.5 shows bugs :) Oct 16 20:32:06 * koen is mighty satisfied with glibc 2.5 Oct 16 20:32:43 I only rebooted my hx to make a movie of psplash for mickeyl :) Oct 16 20:38:44 koen: That's good to know. I'll have to experiment with it :) Oct 16 20:39:04 ah, right I forgot to CC you Oct 16 20:39:13 I mailed tf and xora about the good news Oct 16 20:41:04 what benefits would I get from packaged staging? Oct 16 20:41:18 money, chicks, eternal fame Oct 16 20:41:25 among other things Oct 16 20:41:41 these are nice, but they are not addressing any of my current plroblems :) Oct 16 20:42:02 * RP will hold koen to that ;-) Oct 16 20:42:26 crofton: consider upgrading a package to a newer version without worrying about the old cruft remaining in staging Oct 16 20:42:28 for example Oct 16 20:42:53 or screwing up a do_stage and being able to reverse it... Oct 16 20:43:19 being able to do_clean and actually have the staging crap removed too Oct 16 20:43:21 my problems at the moment involve packages that find stuff in system directories during configure and run time Oct 16 20:43:47 thats a problem with those packages use of autoconf, making assumptions, theres no magic fix for that Oct 16 20:44:26 It is really hosing me at times Oct 16 20:44:54 I also have a problem where a package finds a native library in staging before the one for the target Oct 16 20:45:37 Is there a global pool of local.conf files for test purposes? Oct 16 20:52:56 likewise: choose random distro + random machine and write them into local.conf Oct 16 20:53:31 hmmm Oct 16 20:53:42 but them both machine and distro will be umba-umba Oct 16 20:53:52 * koen wonders how bitbake will cope with that Oct 16 20:53:58 hrw: no, I want local.conf that build OK in general, so that I can test regressions while I work on changes on bitbake and .bbclasses. Oct 16 20:54:04 I bet it will get upset about two overrides being the same Oct 16 20:54:46 koen: good remark. distro=leon, machine=leon :-) Oct 16 20:55:27 koen: can you make me nice picture of hx4700 running GPE? (not gpe on hx4700 but hx4700 with gpe) Oct 16 20:55:35 s/picture/photo even Oct 16 20:56:33 I knew it, there had to be a way to improve me: http://www.ece.ncsu.edu/muse/soc/tutorial/leon/ Oct 16 20:56:34 hrw: remind me tomorrow when there's more ambient light Oct 16 20:57:00 koen: ok Oct 16 20:57:12 koen: I'm not sure it would care if they were the same... Oct 16 20:59:09 it wouldnt, but having both a distro and machine with the same name will result in problems if you pick just one rather than both Oct 16 20:59:29 heh Oct 16 21:10:16 re Oct 16 21:14:27 sleep time Oct 16 21:33:27 every time Koen goes to sleep mtn starts to linger Oct 16 21:34:09 :-) Oct 16 21:34:21 ;-) Oct 16 21:35:13 03pfalcon 07org.oe.dev * r6c3c42b9... 10/ (1 conf/machine/hx4700.conf): hx4700: Organize machine settings per logical groups. Oct 16 21:35:24 ah, here it is Oct 16 21:38:27 03pfalcon 07org.oe.dev * rf6b2427c... 10/ (1 conf/machine/h3600.conf): h3600: Organize machine settings per logical groups. Oct 16 21:46:01 psokolovsky: convert them to MACHINE_FEATURES maybe? Oct 16 21:46:26 hrw, that's 2.4 devices ;-(. I convert 2.6 ones. Oct 16 21:47:30 hrw, I can add MACHINE_FEATURES actually, but afraid whoever later will grep for non-converted machines later, will miss them Oct 16 21:49:34 03pfalcon 07org.oe.dev * r4a8c2f71... 10/ (1 conf/machine/h3600.conf): h3600: Add GUI_MACHINE_CLASS, fix section boundary. Oct 16 21:56:44 03pfalcon 07org.oe.dev * rd3512aa4... 10/ (1 conf/machine/h1910.conf): h1910: Convert to task-base. Organize machine settings per logical groups. Oct 16 22:02:38 03pfalcon 07org.oe.dev * ra66dc981... 10/ (1 conf/machine/h2200.conf conf/machine/h4000.conf): h2200, h4000: Add VOLATILE_STORAGE_SIZE, reformat slightly. Oct 16 22:06:27 03pfalcon 07org.oe.dev * rd5521713... 10/ (1 conf/machine/h5xxx.conf): h5000: Organize machine settings per logical groups. Oct 16 22:13:21 03pfalcon 07org.oe.dev * rb2a9c50f... 10/ (1 conf/machine/h5xxx.conf conf/machine/h5000.conf): h5000: Rename config file h5xxx.conf -> h5000.conf, for naming consistent with other ipaqs. Oct 16 22:20:20 good night Oct 16 22:21:41 03pfalcon 07org.oe.dev * r427a1ebd... 10/ (1 conf/machine/h4000.conf): h4000: Minor format consistency fix. Oct 16 22:24:50 night Oct 16 23:00:02 03rpurdie 07org.oe.dev * rf7b81589... 10/ (6 files in 4 dirs): oprofile: Bump cvs date and update .bb accordingly, add patch to remove bashism, upgrade 0.9.1 -> 0.9.2 Oct 16 23:01:57 03rpurdie 07org.oe.dev * rd889cc5d... 10/ (1 packages/pine/pine_4.63.bb packages/pine/pine_4.64.bb): pine: 4.63 -> 4.64 (fixes fetch failure) Oct 16 23:06:02 03rpurdie 07org.oe.dev * r9f8d2368... 10/ (1 packages/glibc/glibc-package.bbclass): glibc-package.bbclass: Make the localedef error more helpful Oct 16 23:18:15 03rpurdie 07org.oe.dev * r5f2ff670... 10/ (1 classes/base.bbclass): base.bbclass: Add fetchall task to fetch all sources for all the given task and all its dependencies (requires bitbake trunk) Oct 17 01:24:26 anyone knows where to get a keys list contain description? Oct 17 01:25:55 description of what? Oct 17 01:26:03 I have a old copy, it's quite old, and I want to update it, but don't know where to get it Oct 17 01:26:16 old copy of what? Oct 17 01:26:19 like PROVIDE definitions Oct 17 01:26:30 ? Oct 17 01:26:40 and BUILD_ARCH, a full variable list Oct 17 01:26:40 you mean like documentation? Oct 17 01:26:44 yes Oct 17 01:27:04 some docs have been appearing on the website .... Oct 17 01:27:27 there are often variable descriptions inline in the .conf's as well, as flags ont he var Oct 17 01:27:54 generate a document will be better Oct 17 01:28:27 we know that ..... Oct 17 01:29:09 I have an old copy, but don't know how to gernate it Oct 17 01:29:36 run bitdoc. Oct 17 01:29:42 or read the bitbake.conf Oct 17 01:30:27 org.openembedded.documentation Oct 17 01:31:29 these keys are defined in documented.conf Oct 17 01:41:37 03lenehan 07org.oe.dev * r78eb0c15... 10/ (4 files in 2 dirs): Oct 17 01:41:37 clamav 0.88.5: Add new stable branch release of clamav. From the news Oct 17 01:41:37 file: Oct 17 01:41:37 * This version fixes a crash in the CHM unpacker and a heap overflow Oct 17 01:41:37 in the function rebuilding PE files after unpacking. Oct 17 01:50:36 03lenehan 07org.oe.dev * r4600b457... 10/ (5 files in 4 dirs): Oct 17 01:50:36 clamav,lingpng,quagga: Remove some empty directories that I appear to have Oct 17 01:50:36 left behing while deleting things in the last. **** ENDING LOGGING AT Tue Oct 17 02:59:56 2006