**** BEGIN LOGGING AT Sat Jan 27 02:59:58 2007 Jan 27 04:10:41 hey mwester Jan 27 07:48:04 Morning Jan 27 07:52:17 I compiled firefox 1.0.7 for my hx4700. The compilation was successful, however if I try to run it says: Illegal instruction. All other package works good. Jan 27 07:52:23 damn autoconf! Jan 27 07:52:26 damn it to hell! Jan 27 07:52:32 how is this cached?? Jan 27 07:53:17 The only thing I can to is strace firefox. The output is http://pastebin.ca/330036. Jan 27 07:54:00 But I think the last line is informative only: io_submit(0x84, 0x84, 0x1 Jan 27 07:54:09 Any idea? Jan 27 08:02:22 sza2|home: thats a gcc 4/arm problem with firefox, it works with gcc 3.4.4 Jan 27 08:03:34 morning all Jan 27 08:03:51 good morning! Jan 27 08:05:33 T0mW: :-D Jan 27 08:05:51 XorA: Thanks. And can I force bitbake to compile firefox with 3.4.4? Jan 27 08:06:00 :-P that's what 8+ hours of OE does to oneself Jan 27 08:06:15 sza2|home: only by changing distro Jan 27 08:06:40 XorA: :-( but thank you. Jan 27 08:07:20 * T0mW is stuck figuring out how to get modutils into a 2.4 kernel image Jan 27 08:07:26 sza2|home: there is a patch you might want to try, its listed in bugzilla somehere Jan 27 08:08:01 XorA: I'll try to search for it... Jan 27 08:09:21 sza2|home: https://bugzilla.mozilla.org/show_bug.cgi?id=322806 Jan 27 08:09:56 * JustinP stabs xaos Jan 27 08:10:11 why the hell doesn't it list any drivers? Jan 27 08:11:17 XorA: Thx, I'll try to patch it! Jan 27 08:11:54 I've been trying to get this damn program on my Z since I bought it Jan 27 08:11:57 sza2|home: let me know if you succeed, I tried that with firefox 2.0 and failed Jan 27 08:12:07 finally builds now but fails to run Jan 27 08:16:33 * JustinP kicks ipkg Jan 27 08:16:38 go faster! Jan 27 08:19:08 * JustinP decides to build locally Jan 27 08:20:27 JustinP: thats a dream of all Jan 27 08:28:24 XorA: what, that autoconf will die a slow a bloody death at the hands of our new lord and master, build system X? Jan 27 08:28:51 and that people will follow some kind of useful conventions? Jan 27 08:29:02 oh, you just mean a faster ipkg Jan 27 08:29:20 I think just making the feeds sqlite would speed it up a good bit Jan 27 08:29:32 would that I had the time Jan 27 08:33:57 How can I recompile a package without removing and unpacking it after modifying some files? If I use -c clean it removes everything and unpack the source. Jan 27 08:35:03 sza2|home: -c rebuild Jan 27 08:35:17 XorA: Ok. Jan 27 08:35:37 JustinP: yeah faster ipk, its like watching paint dry on slow platforms Jan 27 08:36:07 XorA: rebuild removes and reunpacks Jan 27 08:36:29 sza2|home: I suggest using quilt to manage patches Jan 27 08:36:38 XorA: yeah..... Jan 27 08:39:43 -c rebuild removes the source and unpack it. Jan 27 08:39:57 I can't use quilt... Jan 27 08:40:10 sza2|home: sorry Jan 27 08:40:27 sza2|home: I never used it, I was under impression it just rebuilt what was unpacked Jan 27 08:40:58 sza2|home: then the trick I use is rm tmp/stamps/arm-linux/package.do_compile Jan 27 08:41:16 sza2|home: which will make bitbake re-run everything from compile onwards Jan 27 08:42:02 sza2|home: what's wrong with quilt? Jan 27 08:42:15 XorA: Ok. It will takes some time, now ff is rebuilding. Jan 27 08:42:23 sza2|home: I was thinking about how to do that as well. Maybe write some python code in the package task to tar some other working dir image of the changed code, then copy the tarball into the sources dir and then '-c rebuild' ? Jan 27 08:42:49 T0mW: no reason to do that Jan 27 08:42:53 no? Jan 27 08:42:57 no Jan 27 08:43:01 use quilt Jan 27 08:43:03 set up a patch Jan 27 08:43:05 uh Jan 27 08:43:09 I try to manage it. I will notice about the result later. Jan 27 08:43:20 another package to learn... Jan 27 08:43:27 s/package/tool Jan 27 08:43:41 once it's created then move it to OE, link the patches/X one to the OE one, then just refresh the patch and rebuild Jan 27 08:43:46 quilt is very useful Jan 27 08:43:58 BTW, OE uses quilt Jan 27 08:44:13 I'm still fighting with OE to get depmod into a 2.4 kernel machine image Jan 27 08:45:30 ? Jan 27 08:45:40 it worked fine last I did it Jan 27 08:45:55 sounds like you just need to change IPKG_INSTALL for your image Jan 27 08:46:13 I don't know why, it just won't create the depmod binary or compile the package Jan 27 08:46:49 bitbake modutils fails? Jan 27 08:47:13 won't create it Jan 27 08:47:22 does it *fail*? Jan 27 08:47:28 doesn't even try Jan 27 08:47:39 then it's already built.... Jan 27 08:47:41 no unpack, nothing Jan 27 08:47:52 yes, then it's already built Jan 27 08:47:54 no ipk file in tmp/deploy/ipk/ Jan 27 08:48:08 then perhaps you cleared your ipk folder after building it? Jan 27 08:48:14 bitbake modutils -c rebuild Jan 27 08:48:37 I just did an 'rm -rf tmp' then 'bitbake bootzipit-image' Jan 27 08:49:02 * JustinP has never heard of bootzipit.... Jan 27 08:49:17 time for food, I'll start hacking the python code to put some print statements into it and see if I can dope out what bitbake thinks it is I told it to do. Jan 27 08:49:20 so you built your image and them the modutils ipk doesn't exist? Jan 27 08:49:28 * JustinP shrugs Jan 27 08:49:30 * JustinP is tired Jan 27 08:49:49 JustinP: this is not a Z or collie or angstrom device Jan 27 08:50:08 ok Jan 27 08:50:15 * JustinP stabs xaos Jan 27 08:50:26 damn cross-compiling breakage.... Jan 27 08:51:53 * T0mW`food wonders if he should eat barbaque ribs and use the computer Jan 27 08:58:09 crap I didnt realise minimo was effectively dead Jan 27 08:58:36 hi xora Jan 27 08:58:45 hi pb_ Jan 27 08:59:15 grr, is there any way to make "mtn annotate" or "mtn log" go faster? Jan 27 08:59:31 it seems virtually impossible to inspect the history of a file at the moment. Jan 27 09:01:48 pb_: I heard rumours it was faster in 0.32 Jan 27 09:03:23 it is supposed to be faster in the newer versions Jan 27 09:03:48 XorA: ah right, I have 0.31 Jan 27 09:03:58 * pb_ apt-get upgrade Jan 27 09:04:20 * XorA is still on 0.30 Jan 27 09:06:23 * JustinP has 0.32 Jan 27 09:38:01 * XorA has totally spelling failure Jan 27 09:38:05 ~suite Jan 27 09:38:13 ~spell soot Jan 27 09:38:17 'soot' may be spelled correctly Jan 27 10:06:01 morning all Jan 27 10:18:16 hi rp Jan 27 10:38:09 hey RP Jan 27 10:44:31 Hi. I recompiled firefox 1.0.7. The situation is almost the same. It seems pastebin.ca died (at least for me) so I cannot post strace output. Jan 27 10:48:41 sza2|home: damn, we dont have a fix then Jan 27 10:48:50 sza2|home: koen was looking into that crash Jan 27 10:57:06 XorA: so that's all I was able to do. Jan 27 10:57:49 sza2|home: the mozilla source is too big and messy for me to have time to track it down Jan 27 11:01:45 XorA: now I'm waiting till somebody else solve this matter:-) Jan 27 11:02:48 XorA: And using dillo and gpe-minibrowser. But these are not 'full featured' ... Jan 27 11:09:00 hi Jan 27 11:17:50 hi Jan 27 11:19:01 * timtimred hopes OE will produce a bootable linkstation image by the end of the weekend :D Jan 27 11:19:11 * timtimred is poking it gently Jan 27 11:19:19 timtimred: My system takes ~12 hours to build an OZ image Jan 27 11:19:49 well the minimal image takes somewhat less than that it seems. Jan 27 11:20:35 nearly got our uboot/kernel/other stuff all sweet. Jan 27 11:21:19 i must have rebuilt it at least 8 times yesterday from scratch Jan 27 11:42:51 i've just tried to build gpe-image but it failed: Jan 27 11:42:53 OTE: package diet-x11-X11R7.1-1.0.1-r4: task do_configure: started Jan 27 11:42:55 ERROR: function do_configure failed Jan 27 11:46:04 check how it failed Jan 27 11:46:30 it couldn't find an automake include file Jan 27 11:46:43 i'm just trying something else Jan 27 11:47:09 Can't locate Automake/Struct.pm in @INC (@INC contains: /home/jluebbe/oe/org.openembedded.dev/tmp/staging/i686-linux/share/automake-1.9 /etc/perl /usr/local/lib/perl/5.8.8 /usr/local/share/perl/5.8.8 /usr/lib/perl5 /usr/share/perl5 /usr/lib/perl/5.8 /usr/share/perl/5.8 /usr/local/lib/site_perl .) at /home/jluebbe/oe/tmp/staging/i686-linux/bin/automake line 47. Jan 27 11:47:22 im building in ~/oe/ Jan 27 11:47:47 but it looks for the staging dir in ~/oe/org.openembedded.dev/ Jan 27 11:47:55 /home/jluebbe/oe/org.openembedded.dev/tmp/staging/i686-linux/share/automake-1.9 /home/jluebbe/oe/tmp/staging/i686-linux/bin/automake Jan 27 11:48:11 you fscked environment.. Jan 27 11:48:36 but all other stuff built ok Jan 27 11:48:42 Shoragan: http://www.hrw.one.pl/2006/11/22/my-openembedded-environment-ii/ Jan 27 11:50:35 thx Jan 27 11:52:33 hi Jan 27 11:54:18 I've just sent a plan for handling interactive tasks to the bitbake/oe lists. I'd appreciate thoughts on it... :) Jan 27 11:54:32 mickey|tv: You'll recognise the problem ;-) Jan 27 11:57:03 hrw, in which cases should i clean my tmp directory? Jan 27 11:57:28 or does bitbake figure out when to rebuild things? Jan 27 11:58:16 tmp clean only when ready to build from scratch Jan 27 11:58:27 RP: what about X-less machines? Jan 27 11:59:26 but the cleaning shouldn't change the result? Jan 27 11:59:56 RP: yeah Jan 27 12:00:05 let's see what the guys com up with Jan 27 12:07:59 hrw: X-less machines won't be able to run interactive tasks. I'm open to better ideas. This solution doesn't stop someone writing something to work in the no-X case although it probably involves writing something like X :-/ Jan 27 12:09:02 other problem would be lack of X11 forwarding Jan 27 12:09:38 some of machines where I do builds are reachable via ssh chain only Jan 27 12:09:55 ssh can do X forwarding though? Jan 27 12:10:08 it can. but sshd can have it disabled Jan 27 12:10:48 I know its not ideal and I'm open to other solutions. We do need a way to solve this properly though... Jan 27 12:12:23 As I said, interactive tasks will not be a requirement for builds, just an extra... Jan 27 12:14:12 * cbrake wonders if screen could somehow be used? Jan 27 12:14:22 RP: agreed Jan 27 12:18:46 cbrake: Ultimately there will be GUIs for bitbake and I'm not sure how screen would fit into that Jan 27 12:20:03 RP: yeah, I was just trying to think for a way for bitbake subprocs to attache to a term, but I don't think screen is the solution. Jan 27 12:20:31 RP: nod, GUIs will make that part easier. What are the functions client/server bitbake will provide? Jan 27 12:22:49 cbrake: The reason for the split is to allow development on different UIs. If we have a standard server, writing different clients/UIs becomes easier. Whilst doing this I'd like to ensure the server can run on a different machine to the client and that multiple clients can attach to a given server Jan 27 12:24:33 cbrake: The implementation is looking like it will by xmlrpc based since that is a part of standard python. The UIs will need two threads, one for data transmission, the other for data reception. The thread handling in the UIs should be easily abstracted from the UI writer Jan 27 12:25:45 with security I hope Jan 27 12:25:45 * XorA imagines bitbake rmrf.bb Jan 27 12:27:05 XorA: That will come down to the xmlrpc implementation which I'm presuming can handle such things. At the moment, I'd settle for having something working at all, let alone secure! Jan 27 12:28:01 RP: sounds good. Once that is working, a text mode client could be written to handle the interactive tasks issues. Jan 27 12:29:02 cbrake: There is nothing saying the X display has to be the only way we handle them, its just perhaps the easiest Jan 27 12:29:49 cbrake: If I can prove one solution, I feel happier about taking bitbake down that road. Its then up to others to implement alternatives if they need them (which I agree is certainly possible to do, just harder) Jan 27 12:30:47 RP: right Jan 27 12:48:21 ~botmail for zecke: congrats Jan 27 12:54:53 03hrw 07org.oe.dev * raa39131f... 10/ (1 conf/machine/spitz.conf): spitz: add tar.gz image fstype instead of ?= Jan 27 13:08:51 hrw: Does that work in the case you already have it in IMAGE_FSTYPE? Jan 27 13:09:13 did not tested Jan 27 13:09:42 RP: it works out in the end when I tested it a while ago Jan 27 13:10:21 * RP suspects it will cause two rootfs to be generated Jan 27 13:18:13 RP: but only one file :) Jan 27 13:18:19 since they have the same name Jan 27 13:20:16 hi mickeyl Jan 27 13:20:32 koen: Right, but it seems a little pointless... Jan 27 13:26:45 03pb 07org.oe.dev * rc4be9fe2... 10/ (1 packages/lirc/lirc-modules_0.8.1.bb): lirc-modules: update to 0.8.1 so newer kernels work Jan 27 13:27:44 mickeyl: do you know who maintains the bb files for libx11? Jan 27 13:28:00 I gave up trying to check the monotone history, it was too painful. Jan 27 13:28:10 pb_: Its a collective effort Jan 27 13:28:32 RP: ah. Jan 27 13:28:43 pb_: For commits, I find searching the commits mailing list easiest :-/ Jan 27 13:28:53 more specifically, I was wondering who added the "rm -f ${STAGING_INCDIR}/X11/Xlib.h" to the do_compile method. Jan 27 13:29:31 pb_: http://www.openembedded.org/bonsai/ Jan 27 13:31:49 hrw_: how do I get that to show me the changes? Jan 27 13:33:22 I'm guessing that the commit message probably doesn't say anything obvious like "added crazy rm command to screw up build tree" Jan 27 13:33:56 shit. I was thinking that diffs were implemented there Jan 27 13:34:09 I can't see any way to get a diff Jan 27 13:34:23 pb_: http://www.openembedded.org/bonsai/query/file/packages/xlibs/libx11/ Jan 27 13:35:25 XorA: that shows me a list of the commits, which is a good start, but I can't figure out how to see what each one actually did. Jan 27 13:35:30 littered with supid merges :-( Jan 27 13:35:39 The merges are the problem :-/ Jan 27 13:35:52 I think what I really want is a web version of "mtn annotate", and/or a version that runs in finite time on my own computer. Jan 27 13:36:23 or, yeah, a list of changesets with the merge noise filtered out and with diffs included Jan 27 13:37:13 pb_: get mtn 0.31, that has a fast annotate Jan 27 13:37:37 hey pb_ Jan 27 13:38:15 koen: I already have mtn 0.31. how long should I expect it to take? Jan 27 13:38:16 hey mickeyl ! Jan 27 13:38:17 hmmm, is binary locales generation really supposed to be using /usr/share/i18n/locales/ Jan 27 13:38:22 pb_: 1 or 2 minutes Jan 27 13:38:39 pb_: it was before we added 7.0 X.org Jan 27 13:38:45 ok, I guess I can live with that Jan 27 13:38:47 * pb_ tries Jan 27 13:38:54 hrw_: the rm? oh, drat Jan 27 13:38:58 real 0m18.408s Jan 27 13:38:58 user 0m16.813s Jan 27 13:38:58 sys 0m1.576s Jan 27 13:39:12 d5504d2a.. by koen 2006-12-12: rm -f ${STAGING_INCDIR}/X11/Xlib.h Jan 27 13:39:22 that was from a patch in bugzilla Jan 27 13:39:34 koen: it is not ^+\srm Jan 27 13:39:48 what was the patch meant to accomplish? Jan 27 13:39:59 I'm pretty sure that rm is toxic waste. Jan 27 13:40:07 fix compilation on a hosts without X Jan 27 13:40:24 please remove it if it isn't needed Jan 27 13:40:31 i guess now that we have a package with native-x-header it's no longer needed Jan 27 13:41:41 hm, "mtn annotate" says it's done 33 out of 11678 revs. Jan 27 13:42:03 I guess it'll take a bit more than a couple of minutes on my computer. Jan 27 13:42:07 heh Jan 27 13:42:08 pb_: that sounds like mtn <=0.30 Jan 27 13:42:25 monotone 0.31 (base revision: 1bd1fe1e811dce82bee09b9f0effca3225bd1cee) Jan 27 13:42:27 ah, hhmm Jan 27 13:42:31 (according to "mtn --version", anyway) Jan 27 13:42:33 * koen has an off-by-one error Jan 27 13:42:45 vpn006106:~/Projects/OpenEmbedded/org.openembedded.dev/packages/xorg-lib koen$ mtn --version Jan 27 13:42:45 monotone 0.32 (base revision: 3f0fa5c1d232d37d47f6436fe74919dd3cd003b1) Jan 27 13:42:49 ah Jan 27 13:42:54 sorry for that Jan 27 13:43:13 okay, I'll try upgrading in a bit Jan 27 13:43:18 does it make "mtn log" faster as well? Jan 27 13:43:43 yes Jan 27 13:43:49 03rpurdie 07org.oe.dev * rf00466f8... 10/ (3 files in 2 dirs): Jan 27 13:43:49 zaurus machines includes: Fix jffs2 image generation to end in .jffs2 to fix Jan 27 13:43:49 magic symlinks, correctly set akita eraseblock size (removes empty flash Jan 27 13:43:49 warnings upon boot), remove temp images from saving some diskspace, generally Jan 27 13:43:49 clean up clamshell file (from poky) Jan 27 13:43:50 great Jan 27 13:44:11 pb_: log and annotate go from "some hours" to "5 minutes or less" Jan 27 13:44:42 * RP notes to remember shell expansion in commit messages in future Jan 27 13:45:11 at the cost of storing some extra stuff in the db, which takes some time to generate if you don't use the snapshot db Jan 27 13:46:05 pb_: FWIW, it appears that rm was in when the file was added Jan 27 13:50:15 http://rafb.net/p/Usaw2Q84.html Jan 27 13:51:46 Ahrrg, damn it. The hashes changes when we migrated the database so I can't use the hashes in the oe commit messages with the current db Jan 27 13:52:47 Both the X11R7.0 and .1 versions we added by pH5. I can't find out if it was in the .0 version, it was in the initia .1 version checked in Jan 27 13:53:43 I suspect it was in the originals from AMD Jan 27 13:55:23 03mickeyl 07org.oe.dev * r9d96af3e... 10/ (3 files in 3 dirs): gtk+-2.6.10: patch GtkStatusBar .c|.h to work w/ glib >= 2.9.x Jan 27 13:58:46 RP: you'll need the above patch in poky as well. it's necessary for all apps that #include gtkstatusbar.h. As I move towards building more stuff w/ the Gtk+ 2.6.10 / glib 2.12.3 combination the patch might be enhanced eventually Jan 27 14:01:32 mickeyl: ok, thanks :) Jan 27 14:03:46 hmm Jan 27 14:03:52 patch is malformatted Jan 27 14:03:54 * mickeyl fixes Jan 27 14:07:22 * RP is pleased to finally fix those empty flash messages on akita Jan 27 14:07:22 mickeyl: what would you think of doing 'diet-x11', 'small-x11' and 'fat-x11'? Jan 27 14:07:38 RP: cool Jan 27 14:07:49 koen: sounds good. three variants should be ok Jan 27 14:07:54 I'd like to have a small-x11 with xkb extensions Jan 27 14:08:02 yeah Jan 27 14:08:42 Any objections to a nokia800 machine btw? Jan 27 14:08:42 RP: nope Jan 27 14:08:42 not at all, go ahead Jan 27 14:08:42 03mickeyl 07org.oe.dev * r4c6648cc... 10/ (1 packages/gtk+/gtk+-2.6.10/compile-against-newer-glib.patch): gtk+ 2.6.10 compile-against-newer-glib.patch rediff and make it -p1 Jan 27 14:08:52 RP: I have a physical one waiting at home :) Jan 27 14:09:25 koen: a developer discount one? Jan 27 14:09:36 yes Jan 27 14:10:18 I qualified for one this time too, it arrived yesterday... Jan 27 14:10:46 btw. Jan 27 14:10:47 mine arrived thursday Jan 27 14:10:51 how about this jffs2 summary stuff? Jan 27 14:10:56 is this working nowadays? Jan 27 14:10:57 but I haven't been at home since sunday Jan 27 14:11:00 mickeyl: What about it? Jan 27 14:11:08 mickeyl: it worked for me on c7x0 Jan 27 14:11:14 it seems to be working, but I still would prefer logfs :) Jan 27 14:11:17 well, you removed it again from zaurus-2.6, so I guess I asked Jan 27 14:11:29 s/asked/should ask/ Jan 27 14:11:47 oh you didn't Jan 27 14:11:48 koen: Mine would have arrived sooner if the credit card company hadn't messed up the security checks with incorrect records :-/ Jan 27 14:11:56 mickeyl: I didn't ;-) Jan 27 14:12:21 I just rewrote it so it was more understandable... Jan 27 14:12:30 i see, cool. is that stable? Jan 27 14:12:51 It works buts its expensive on the image size Jan 27 14:13:01 RP, mickeyl: I'll hassle MDK about hildon-libs dependance on gtk-patched-too-hell, so we can use our OE gtk to build maemo stuff Jan 27 14:13:14 good Jan 27 14:13:16 do that :) Jan 27 14:13:41 anyone tried maemo's high level internet connection library yet? Jan 27 14:13:45 koen: I doubt you'll never get that dependency removed :-/ Jan 27 14:13:49 sounds like we may want to reuse that Jan 27 14:13:59 mickeyl: libconic? Jan 27 14:14:02 ya Jan 27 14:14:13 RP: tko and MDK want to ifdef it Jan 27 14:14:19 gpe-image fails to build gpesyncd because it depends on libgpevtype Jan 27 14:14:31 how do i tell it to build the svn version? Jan 27 14:15:06 koen: That would be good... Jan 27 14:15:38 RP: some maemo people want to run hildon on top of ubuntu on their desktops Jan 27 14:16:45 Some people want to run maemo apps on their zaurus too :) Jan 27 14:16:58 like maemo-mapper :) Jan 27 14:17:12 maemo-mapper need fixed srcdate Jan 27 14:17:18 having a internal md is good for storing lots of tiles Jan 27 14:17:28 hrw_: it has a fixed rev Jan 27 14:17:57 maemo-mapper-nohildon_svn.bb one? Jan 27 14:18:20 yes, otherwise the patch fails to apply Jan 27 14:18:39 let me guess... Jan 27 14:18:46 distro config has it set or sane-srcdate? Jan 27 14:19:02 * RP is hoping one of the OH gtk people will give maemo-mapper some tlc for poky Jan 27 14:19:37 ./conf/distro/include/sane-srcdates.inc:70:SRCDATE_maemo-mapper-nohildon ?= "20061114" Jan 27 14:19:41 found Jan 27 14:20:13 RP: I've been using maemo-mapper to navigate all week now Jan 27 14:20:33 its online route planning thing is pretty cool Jan 27 14:20:51 * koen heads home Jan 27 14:20:52 later all Jan 27 14:21:18 hi paul Jan 27 14:22:04 Hi! Jan 27 14:23:10 hi psokolovsky Jan 27 14:37:21 cu Jan 27 14:42:01 mickey|sports: http://ewi546.ewi.utwente.nl/tmp/hrw/t/030_CVE-2007-0010_error-handling-in-pixbuf-loaders.patch will maybe interest you Jan 27 14:42:39 re Jan 27 14:42:57 hi all Jan 27 14:54:03 morning likewise Jan 27 14:54:19 morning Jan 27 14:55:11 koen|away: diet/small/fat is a good idea. Jan 27 14:56:41 morning Jan 27 15:54:18 how long does binary locale generation take? Jan 27 16:05:17 XorA: several hours ;-) Jan 27 16:06:31 fsck :-) Jan 27 16:08:58 although result my amd64 has run at 100% load for 2 hours without crashing Jan 27 16:24:10 re Jan 27 16:33:57 * koen plays with his n800 Jan 27 16:36:19 XorA: 20 hours doesn't sound too bad Jan 27 16:37:19 * koen foods Jan 27 16:57:53 Anybody compiled gnokii in the last time? Really strange error: http://rafb.net/p/0yTvJG35.html Jan 27 16:58:08 Full log can be found here: http://people.openezx.org/stefan/tmp/log.do_compile.22936 Jan 27 18:38:37 real wired internet on 3200 works ;-) Jan 27 18:40:15 03justinp 07org.oe.oz354x * r2683f0be... 10/ (6 files in 4 dirs): libnl: add libnl 0.5.0 Jan 27 18:40:18 03justinp 07org.oe.oz354x * r1ac07a37... 10/ (5 files in 4 dirs): bmon: add bmon 2.1.0 Jan 27 18:40:24 03justinp 07org.oe.dev * r95c991e6... 10/ (4 files in 3 dirs): libnl: add 0.5.0 (stable) Jan 27 18:40:31 03justinp 07org.oe.dev * r2e54b15c... 10/ (5 files in 4 dirs): bmon: add 2.1.0 (depends on libnl-0.5.0) Jan 27 18:42:41 hey Jan 27 18:42:52 cbrake: I just enjoyed reading your various articles :) Jan 27 18:43:55 zecke: thanks for the feedback. Jan 27 18:45:24 stefan_schmidt: please check where the local include comes from. it's probably a bug in gtk-1.x. Jan 27 18:45:39 we now have a check for local includes which we didn't have in earlier times Jan 27 18:45:44 (which is a good thing, btw.) Jan 27 18:45:56 zecke: embedded Linux can be really good, and it can be really bad. Most of the projects I gen involved in that were started by someone else fall into the "really bad" category, so I have a lot to write about :-) Jan 27 18:47:14 zecke: I always appreciate feedback -- good or bad Jan 27 18:48:05 cbrake: Well, the "Next >>>" button iritated me. I did not expect to go to the next article Jan 27 18:50:00 zecke: good point, I'll have to look into that. Jan 27 18:50:24 I'm used to this "next page" thinking Jan 27 18:50:42 this way I had a quick look at all of them Jan 27 18:52:34 argh, just realized I had the article I'm working on "published" before it was finished -- now it is in all the feeds :-( Jan 27 18:53:50 hehe Jan 27 18:54:14 maybe this is the reason I clicked Next in the first place Jan 27 18:54:46 zecke: heh, I can understand that. Jan 27 19:01:38 zecke: congrats Jan 27 19:02:58 thanks for your botmail Jan 27 19:06:58 I had a reminder set in iCal, but I didn't use my laptop last week Jan 27 19:07:32 hehe, don't worry. I don't even know most of your birthdays :} Jan 27 19:07:49 koen: install apple's webdav server for keep track of our OE social data Jan 27 19:08:01 to keep... man today is a bad day Jan 27 19:08:20 I should use iSync more :) Jan 27 19:09:42 I can't wait for the OpenMoko device Jan 27 19:10:28 koen: thanks to fu*m*bola* we should have decent SyncML support and iSync should 'just' work :} Jan 27 19:36:20 03justinp 07org.oe.oz354x * r24833026... 10/ (1 packages/aalib/aalib_1.4rc5.bb): aalib: stage lib and headers Jan 27 19:36:24 03justinp 07org.oe.dev * r091a4320... 10/ (1 packages/aalib/aalib_1.4rc5.bb): aalib: stage lib and headers Jan 27 19:39:24 JustinP: please no mplayer aalib support :-) Jan 27 19:48:04 if KERNEL="Kernel24", why won't MACHINE_FEATURES = KERNEL work? I had to setit to MACHINE_FEATURES = "Kernel24" Jan 27 19:48:49 should I have said MACHINE_FEATURES = ${KERNEL} instead? Jan 27 19:49:01 * T0mW knows very little python Jan 27 19:49:15 MACHINE_FEATURES = "${KERNEL}" Jan 27 19:49:21 tnx Jan 27 19:49:24 and it isn't python :) Jan 27 19:49:30 no? Jan 27 19:49:37 bitbake syntax then. Jan 27 19:50:29 AH! finally got a sbin/depmod.24 binary in the rootfs, heh Jan 27 20:00:48 XorA: why not? ;-) Jan 27 20:00:55 XorA: it's for bb and xaos, actually Jan 27 20:01:38 XorA: unfortunately, xaos compiled from OE won't diaply anything...says no drivers. COmpiled natively it works. Jan 27 20:02:03 JustinP: coz some stupid luser will complain that mplayer looks funny and seems to be printing ASCII :-) Jan 27 20:03:26 heh Jan 27 20:11:55 and Im not kidding, I had a familiar user emailing me to demand I enabled win32 codecs in mplayer because they work faster than ffmpeg Jan 27 20:12:10 the lack of x86 cpu in his Ipaq just wouldnt convince him Jan 27 20:14:20 XorA: propose him to use the real deal and install windows ce 6.0 Jan 27 20:19:08 XorA: maybe explain to him that it was a copyright thing and that MicroSoft would be able to do that for him? Jan 27 20:19:13 ;-) Jan 27 20:23:43 XorA: surely win32 codecs running in bochs are faster than ffmpeg!?!?! Jan 27 20:27:36 drat Jan 27 20:28:26 on AMD64 X2 libc.so.6 keeps exploding with a free() backtrace Jan 27 20:28:35 03xora 07org.oe.dev * r79b308f5... 10/ (1 packages/gnome/gnome-vfs_2.16.3.bb): packages/gnome/gnome-vfs_2.16.3.bb : spell RREPLACES right Jan 27 20:56:59 do oe and maemo/scratchbox address about the same problem or are they more clearly different things? Jan 27 20:58:38 mlpug: I don't know scratchbox that much. OE allows to create toolchains, or use prebuilt ones has a huge set of buildable applications. Is very flexible when it comes to packaging (tar.gz ipk, deb, rpm) and creating rootfilesystems (jffs2, cramfs,ext2,ext3...) Jan 27 20:58:51 mlpug: I don't know if scratchbox offers any of these things Jan 27 21:01:09 scratchbox is just a very fast virtual machine Jan 27 21:02:10 i do not know either of them too well. I think that scratchbox is a crosscompilation environment. chrooted fake environment to make autotools (~ ./configure) to work. so you do not have to modify build scripts manually. Jan 27 21:02:49 heh Jan 27 21:02:59 you always have to modify them manually :) Jan 27 21:03:10 most people are pretty bad at writing makefiles Jan 27 21:03:24 OE just does that for you Jan 27 21:03:26 true Jan 27 21:04:17 scratchbox is nice if the term 'cross-compiling' is too scary for your engineers Jan 27 21:10:29 03justinp 07org.oe.oz354x * r491e00c3... 10/ (5 files in 4 dirs): bb: add bb 1.2, currently fails to cross-compile Jan 27 21:10:35 03justinp 07org.oe.oz354x * re77c0d63... 10/ (6 files in 4 dirs): xaos: add 3.2.2, compiles but cannot find drivers when run Jan 27 21:10:35 03justinp 07org.oe.oz354x * rc2cd1048... 10/ (1 packages/bb/bb_1.2.bb): bb: add LICENSE Jan 27 21:10:39 03justinp 07org.oe.dev * r0731f3e7... 10/ (5 files in 4 dirs): bb: add bb 1.2, currently does not cross-compile Jan 27 21:10:44 03justinp 07org.oe.dev * r86364b8a... 10/ (6 files in 4 dirs): xaos: add 3.2.2, compiles but cannot find drivers when run Jan 27 21:10:49 03justinp 07org.oe.dev * r7444bbc2... 10/ (1 packages/bb/bb_1.2.bb): bb: add LICENSE Jan 27 21:11:56 JustinP: have you looked into mpd and mod support? Jan 27 21:14:50 zecke: it builds.... Jan 27 21:15:02 zecke: and installs (I had some old mikmod libs sitting around) Jan 27 21:15:04 03justinp 07org.oe.dev * r9e02c22d... 10/ (1 packages/xaos/xaos_3.2.3.bb): xaos: fix DEPENDS Jan 27 21:15:09 zecke: haven't tested to see how well it works, though Jan 27 21:16:29 place a mod file in your music collection and see if updating the collection works Jan 27 21:16:42 if it doesn't crash you won Jan 27 21:17:00 I think I will spend the next hour to replace the mpd db with sqlite Jan 27 21:17:43 what? Jan 27 21:17:49 you're going to patch it to use sqlite? Jan 27 21:18:11 yes, mpd's db code is a piece of shit. it goes oom on my shark Jan 27 21:18:18 ah Jan 27 21:18:33 well, I assume you'd be planning on pushing it back up stream? Jan 27 21:18:53 sure, just like the mod patch Jan 27 21:19:02 upstream doesn't seem to be responsive though Jan 27 21:19:10 :-P Jan 27 21:26:23 * RP sighs at the n800 wifi driver Jan 27 21:30:05 * JustinP sighs at all sorts of things Jan 27 21:30:31 zecke: I put some mods on my Z and mpd updated....but isn't listing them Jan 27 21:31:14 zecke: some aqlite work on ipkg would be nice too Jan 27 21:31:42 JustinP: are you sure updating didn't crash? Jan 27 21:31:58 JustinP: mpd --create-db works? Jan 27 21:32:12 JustinP: 0.12 sometimes needs a restart after updating the db Jan 27 21:33:17 zecke: as I've mentioned before, I believe mpd doesn't support --create-db any more... and no, it didn't crash Jan 27 21:33:36 JustinP: are you sure. it forks and children crash (at least here) Jan 27 21:33:51 JustinP: and my 0.12.1 has at least (--create-db|--update-db) Jan 27 21:33:52 great Jan 27 21:34:00 looks like I have some major network problems... Jan 27 21:34:47 or perhaps wget/ipkg is just foobared Jan 27 21:35:20 http://pastebin.com/868981 yeah Jan 27 21:37:18 damn yahoo and google Jan 27 21:37:22 anyone know how to debug python? my upgrade has died configuring python-newt with: Jan 27 21:37:23 File "/usr/share/pycentral-data/pyversions.py", line 62, in old_versions Jan 27 21:37:26 return [v[6:] for v in _old_versions] Jan 27 21:37:32 which means nothing to me Jan 27 21:38:03 wookey_: I have never used a python debugger, I always add 'print xyz' Jan 27 21:38:04 oops - wrong window Jan 27 21:38:18 although in fact I'm sure you lot know all about python Jan 27 21:38:41 win 4 is usuually debian-uk :-) Jan 27 21:38:44 wookey_: I can say what this statement should create Jan 27 21:38:47 hehe Jan 27 21:42:38 they're having a damn turf war and killing my poor server Jan 27 21:42:38 damn it! Jan 27 21:42:38 * JustinP rewrites gbot and slurp into oblivion Jan 27 21:42:38 for some directories at least... Jan 27 21:42:42 zecke: remember, I'm using 0.12.1+svn Jan 27 21:42:53 zecke: remember, I'm using 0.12.1+svn Jan 27 21:42:54 JustinP: I have svn trunk now Jan 27 21:43:12 JustinP: and my copy of main.c has --create-db Jan 27 21:43:14 and here the hell comes msnbot Jan 27 21:43:18 ok Jan 27 21:43:21 I must have read wrong Jan 27 21:44:14 zecke: ah....I see from /etc/mpd/mpd.error that the decode and update processes are dying Jan 27 21:45:05 so it is either OOM or the libmikmod struct issue os occuring Jan 27 21:45:45 could be. Although that part of that patch didn't apply to the code I was patching Jan 27 21:45:51 looked like it had that in it Jan 27 21:46:02 would you care to look at the version in OE? Jan 27 21:46:37 JustinP: The bug is tricky and not obvious. you need to count the elements of the struct Jan 27 21:46:44 JustinP: I can do that, probably tomorrow Jan 27 21:48:30 oh, there's *another* NULL....ok, I think we do need that patch Jan 27 21:48:33 * JustinP tries again Jan 27 21:49:06 JustinP: doku found it by pure luck, he was victim of this bug when he wrote is own modplayer and remembered it :) Jan 27 21:49:17 wrote his own... *tired* Jan 27 21:49:39 and someone is going all over the place with MSIE... Jan 27 21:52:06 * JustinP rebuild Jan 27 22:01:02 03justinp 07org.oe.oz354x * raee04d12... 10/ (3 files in 3 dirs): mpd: fix mod support in svn Jan 27 22:01:06 03justinp 07org.oe.dev * r91c70ca1... 10/ (3 files in 3 dirs): mpd: add pulse support, fix mod support Jan 27 22:01:50 zecke: that didn't help Jan 27 22:01:53 zecke: still crashing Jan 27 22:02:07 I'md be happy to fix it but I really need ot get some paying work done today Jan 27 22:02:24 JustinP: welcome to the club, get your work done Jan 27 22:05:09 03justinp 07org.oe.dev * r41c93316... 10/ (40 files in 7 dirs): efl, e17: update ot newest versions Jan 27 22:06:09 don't know how I managed to not push that rev... Jan 27 22:13:39 nite guys :) Jan 27 22:20:37 hrw|gone: what did you pay for your travel to brussels? Jan 27 22:26:59 hi, all! Jan 27 22:27:35 I have got arm-angstrom-linux-gnueabi-libtool: compile: unable to infer tagged configuration error with jpeg-6b Jan 27 22:27:50 while building gpe-image Jan 27 22:27:56 any ideas? Jan 27 22:29:36 http://ossfans.org/paste?id=b82eb8 Jan 27 22:56:03 Greetings. Jan 27 22:56:33 I'm trying to complete gpe-image for the c7x0 machine and it dies on the imaging stage with: Jan 27 22:56:34 "| AttributeError: 'module' object has no attribute 'manifest' Jan 27 22:56:44 Any suggestions as to where I'm going wrong? Jan 27 23:12:36 mickey|dinner: I think I've found the include Problem from gnokii Jan 27 23:12:56 mickey|dinner: GTK_CFLAGS = -I/usr/include/gtk-1.2 -I/usr/include/glib-1.2 -I/usr/lib/glib/include Jan 27 23:13:11 Inside a gnokii Makefile Jan 27 23:18:26 and now someone linked to an image on my site from myspace....screw that Jan 27 23:18:40 well, you know what to do JustinP ;-) Jan 27 23:19:06 oh, it's broken now Jan 27 23:19:26 mostly I'm pissed at google and yahoo for their indexing wars killing my server Jan 27 23:19:33 btw, any of you guys got any ideas about my error? Jan 27 23:19:53 nope Jan 27 23:20:03 I can go plowing the sources if needed but just in case it was a common mistake I figured I'd ask. Jan 27 23:20:34 ugh....and msnbot...I've got all 3 of the damn things sucking down my bandwidth Jan 27 23:20:54 JustinP: robots.txt? Jan 27 23:23:36 http://www.hrw.one.pl/2007/01/28/congratulations-you-have-won-a-free-neo1973/ Jan 27 23:23:40 ;D Jan 27 23:25:02 summatusmentis: it takes a long time to take effect Jan 27 23:25:45 hrw|gone: cool :-) Jan 27 23:26:22 greentux: I paid WizzAir Warsaw<>'Brussels' plane tickets Jan 27 23:28:24 hrw: I'm so jealous, even though I don't have any dev skills to put me in the running, and I don't have GSM in my area Jan 27 23:28:27 JustinP: got it Jan 27 23:33:35 hrw: Any idea what the neo will actually cost? Jan 27 23:33:48 * NAiL is very interested in getting one to tinker with Jan 27 23:33:49 NAiL: 350 usd Jan 27 23:33:50 ~$350 I think Jan 27 23:34:06 NAiL: http://lists.openmoko.org/pipermail/announce/2007-January/000000.html Jan 27 23:34:21 cu - time to sleep Jan 27 23:35:31 hrw|gone: thanks and good night ;) Jan 27 23:51:36 re **** ENDING LOGGING AT Sun Jan 28 02:59:58 2007