**** BEGIN LOGGING AT Wed Feb 01 02:59:59 2006 Feb 01 03:03:51 zecke: ! Feb 01 03:04:14 and mercurial should be evaluated as well Feb 01 03:36:28 guys .. i build for mipsel targets .. i was using nylon.conf as distro.. but ideally i wanna use familiar of course .. now i added my preferred things into familiar and also added _mipsel to the tag , but still it bake`s the wrong packages ? Feb 01 03:36:52 so it is overwritten somewhere Feb 01 03:37:03 rob_w-cgi: you want glibc or uclibc? Feb 01 03:37:42 glibc i use currently Feb 01 03:38:47 it did work well with nylon.conf on the dev tree but i was now to try the other tree and it seems packages dont build as i nylon is not maintained for that .. so ideally i added my tags to familiar.conf Feb 01 03:40:23 so the tags i use are the preferred_version = blabal_mipsel Feb 01 03:43:09 that's incorrect. you need the _mipsel on the left hand side. Feb 01 03:43:20 PREFERRED_VERSION_opie-opie-opie_mipsel = "0.00" Feb 01 03:56:01 pb oh ,, right you said that Feb 01 03:56:35 well i do have that ,, my paste was wrong Feb 01 03:58:55 so to correct i f PREFERRED_VERSION_binutils-cross_mispesl ="2.14.90.0.7" for example .. yet it try`s tobuild 2.16 Feb 01 03:59:16 yeah and i do have mipsel and not mispesl Feb 01 04:27:23 is anyone running samba from .dev? Feb 01 04:29:04 it's behaving oddly Feb 01 04:32:51 johI am not, but I was planning to Feb 01 04:33:14 s/johI/JohnX: I/ Feb 01 04:33:15 XorA meant: JohnX: I am not, but I was planning to Feb 01 04:49:44 can i get some format requests for audio only (for now) media player? Feb 01 04:49:52 curently suports ogg and mp3 Feb 01 04:50:42 dan2003: flac, wav Feb 01 04:50:58 dan2003: These are the most important ones i guess... flac might be a good idea. Feb 01 04:51:01 heh Feb 01 04:51:02 flacs on my list, peple do use wav then ;) Feb 01 04:51:37 dan2003: fasttracker .xm, protracker, digibooster(pro) and other trackers? Feb 01 04:52:20 hrw|work: OctaMED :-) Feb 01 04:52:42 i; have to look into them, never heard of em myself! Feb 01 04:53:03 XorA: and MED too Feb 01 04:53:18 dan2003: libmikmod plays most of it Feb 01 04:53:20 hrw|work: well that comes for free with Octa :-) Feb 01 04:53:29 k Feb 01 04:53:35 is it integer based? Feb 01 04:53:39 XorA: maybe - never really used (octa)med Feb 01 04:53:50 dan2003: yes - Amiga machines rather lack FPU Feb 01 04:53:51 hrw|work: way back in my amiga days :-) Feb 01 04:54:08 XorA: protracker player was one of my first apps Feb 01 04:54:21 k, do the latter pxa, the 270(?) have usable fpu ? Feb 01 04:54:47 and an aminet browser client for media player :-) Feb 01 04:58:02 and Jesus on E's on the Z :-) Feb 01 04:58:11 XorA: http://main.aminet.net/dev/e/PTR_Play.lha Feb 01 04:58:40 hrw|work: did you get your free aminet cd? Feb 01 04:58:47 XorA: never Feb 01 04:59:02 hrw|work: Still got mine I think, somewhere Feb 01 04:59:03 who will buy my wonderful roses (sd card), who will buy my wonderful sd card ... Feb 01 04:59:06 XorA: my stuff appeared on few of them Feb 01 04:59:43 anyone know of an sd card 1gb that is SCR 0 ? Feb 01 04:59:50 morning all btw Feb 01 04:59:52 hrw|work: http://de4.aminet.net/game/role/WFRPGenerator.lha was my one and only app Feb 01 05:00:52 ade|desk: few 'kingston' branded, takems branded Feb 01 05:01:21 how do you tell, Ive a 1G in mine Feb 01 05:01:48 XorA: if it works in 2.6 Z then it is scr0 Feb 01 05:02:15 Kingston Eleite 50x cards are then Feb 01 05:02:25 ooo, are there some that dont work, i neary bough one the other day Feb 01 05:02:55 can i display which packages is pulled from which ? Feb 01 05:02:56 kingston is just a name for hitachi/toshiba/panasonic cards Feb 01 05:03:34 http://www.ebuyer.com/customer/products/index.html?action=c2hvd19wcm9kdWN0X292ZXJ2aWV3&product_uid=88007 works Feb 01 05:04:06 and is very nippy as well Feb 01 05:04:09 ~change 43 gbp to pln Feb 01 05:04:14 43.00 Pound (GBP) makes 240.878 Zloty (PLN) (from http://www.xe.com/) Feb 01 05:04:23 brb Feb 01 05:05:14 cheers XorA Feb 01 05:05:29 ade|desk: that is the card that is my / device Feb 01 05:06:09 I can say the toshiba SD-M01G 0524 TK0549M (Kingston label) is SCR 1 Feb 01 05:06:41 luv the key features associated with that 1gb card ;) "Roomy: Up to 512 MB of storage capacity" Feb 01 05:07:05 dan2003: :-D Feb 01 05:07:22 ade|desk: what is difference of SCR0/1 Feb 01 05:08:19 SCR 0 is a known system, SCR 1 is unknown atm.. waiting for the standard to be let out for public viewing Feb 01 05:08:41 that sucks Feb 01 05:09:16 I spent ages trawling oesf looking for reports of cards that worked Feb 01 05:09:24 in end I gambled on that one Feb 01 05:12:09 ah , again i found a typo Feb 01 05:13:36 http://www.oesf.org/forums/index.php?showtopic=17450&view=getlastpost Feb 01 05:19:29 hrw|work: :-) Feb 01 05:29:27 XorA: ljp posts makes me .... from time to time Feb 01 05:30:10 hrw|work: who is he? Feb 01 05:30:20 * hrw|work hates writing work reports.. always some time missing Feb 01 05:30:51 XorA: ljp = lpotter = Lorn Potter (opie developer which goes to trolltech) Feb 01 05:31:04 * XorA wonders what the hell the secret coding he used for WFRPGenerator actually was Feb 01 05:32:29 XorA: no source? Feb 01 05:32:49 I released most of my amiga sources before moving to linux/x86 Feb 01 05:32:59 hrw|work: lost those disks long ago, was checking maybe I uploaded source as well, I didnt Feb 01 05:33:18 hrw|work: it was an AMOS program Feb 01 05:33:40 hrw|work: written back in days amiga didnt have a decent Ccompiler yet Feb 01 05:34:17 * XorA has fond memories of K&R C and amiga Feb 01 05:34:32 amos.. I hate you XorA Feb 01 05:34:43 hrw|work: why? Feb 01 05:35:04 XorA: all amos using coders are on my hatelist - I have vga monitor Feb 01 05:35:16 ;) Feb 01 05:35:49 hrw|work: heh heh, well I think when I finally released it it was AMOSPro and understood the concept of screen modes Feb 01 05:36:17 Has anyone tried ipkg 0.157? Feb 01 05:36:20 hrw|work: if not its because I was putting amiga in bin, I had the code to make AMOS work properly with Intuition Feb 01 05:37:00 RP: I did - works Feb 01 05:37:16 ive impeented a buffered version of fread/fseek/ and friends, so as to minimse microdrive powerw ups on my z, it seem to work well, ive tested with vorbis playback in valgrind.. is there perhaps an os evel way of doing the same thing? Feb 01 05:38:02 XorA: I was writing in HSPascal, AmigaE, C (sas/c, gcc) Feb 01 05:38:35 hrw|work: in my day gcc Amiga didnt exists, E had been written yet, and I had SAS/C version 1.0 Feb 01 05:39:00 hrw|work: Its not letting me use -force-overwrite here so it looks like the bugs haven't been fixed :-/ Feb 01 05:39:31 what is a ?= doing on the Preferred Verison tag ? Feb 01 05:39:37 I used HSPascal for my Turbo Pascal Uni assignments Feb 01 05:39:43 rob_w-cgi: set if it was not set Feb 01 05:40:07 ah Feb 01 05:40:45 RP: looks like ipkg is in phase 'if I fix bug1 then bug2 reappear' Feb 01 05:41:27 hrw|work: I just resorted to rm /usr/lib/ipkg/status :-/ Feb 01 05:43:03 hmm still it bakes glibc 2.3.5 but i need 2.3.3 Feb 01 05:49:09 hmm i now moved away the glibc 2.3.5 bb `s and it is happy to go on baking .. lets see if that works Feb 01 06:09:59 hmm I think I have messed up the darcs tree Feb 01 06:26:00 doh Feb 01 06:26:44 pb_: somehow it decided to store new data in another directory Feb 01 06:28:01 oh, that is unfortunate Feb 01 06:28:18 pb_: let me see if a push fucks it up Feb 01 06:28:29 * zecke needs to start working now Feb 01 06:30:18 it could work Feb 01 06:40:08 darcs push takes 81% of cpu and 40% ram on ewi Feb 01 06:40:39 45% which is now 1GB of ram Feb 01 06:40:59 on pushing 331 patches Feb 01 06:41:03 I think its just ruled itself out then Feb 01 06:41:31 There was an interesting post from Keith Packard to the git list btw - he's going to use git for some of the xorg code Feb 01 06:41:33 sorry that are 331 patches Feb 01 06:41:47 332 even :} Feb 01 06:43:11 He also mentioned that hg and bzr are going to need file format changes in the near future requiring flag days... Feb 01 06:43:51 lunch... Feb 01 06:44:22 okay it went to 70% of RAM Feb 01 06:44:31 which is ~1.4 GB Feb 01 06:45:02 zecke: Its ruled out ;-) Feb 01 06:46:34 RP: http://oe.handhelds.org/cgi-bin/moin.cgi/DarcsScmTrial#preview Feb 01 06:46:37 seem my comment Feb 01 06:46:55 see my comment even Feb 01 06:47:28 I think that sums it up Feb 01 06:48:05 I still want others to make experience as well Feb 01 06:48:10 zecke: yah, darcs routinely seems to use about 2GB of RAM for me Feb 01 06:48:33 "darcs annotate" was a much more painful operation than I was expecting. Feb 01 06:48:54 zecke: I don't have any machines with 2GB of ram spare though ;-) Feb 01 06:49:03 I would hate to have to quit evolution in order to pull new versions. Feb 01 06:49:14 lol Feb 01 06:49:37 http://oe.handhelds.org/cgi-bin/moin.cgi/DarcsScmTrial#preview Feb 01 06:49:51 evolution works on less than 3gig ram .. i never knew that Feb 01 06:50:02 mercurial is nice too, try it guys ;) Feb 01 06:50:16 zecke: hehe Feb 01 06:50:24 okay, I try mercurial soonest Feb 01 06:52:37 now trying to push 61 patches only Feb 01 07:01:02 lol 63 patches are using 1.9gb of memory as well Feb 01 07:12:30 RP: read your comment, sad to hear your Z failed to resume (but at least it's less likely an gcc bug) Feb 01 07:14:00 AFAIK the Gentoo project still uses CVS Feb 01 07:14:25 there were tests with SVN, but SVN could not handle the *massive* quantity of files Feb 01 07:19:38 shadows: well look at KDE, Mandriva Feb 01 07:19:55 shadows: Mandriva uses svn for each and every release, they even keep the src in svn Feb 01 07:20:16 shadows: KDE has 504602 revisions Feb 01 07:20:21 heh Feb 01 07:20:24 shadows: that is quite a lot as well ;) Feb 01 07:21:06 at the point in time when it was tested though, that was i think at least a year ago Feb 01 07:27:21 shadows: you can try the experimental OE svn tree Feb 01 07:31:08 i could do a pull and comment on it Feb 01 07:31:15 is that what you mean? Feb 01 07:32:25 RP: open bug report for the failing to resume for Cxx00, #421 Feb 01 07:33:20 zecke: I'll try mercurial... cloning now. Feb 01 07:42:56 florian_kc: send me your ssh key so I can give you push access Feb 01 07:43:25 shadows: The next thing to test is whether commenting out the apm_event_queue(something_CRITICAL) line in arch/arm/common/sharpsl_pm.c helps Feb 01 07:43:52 zecke: okay Feb 01 07:44:43 RP: we do not have poodle developer(s) now - right? Feb 01 07:47:42 zecke: sent Feb 01 07:55:36 florian_kc: oh well Feb 01 07:56:30 florian_kc: I will add you tonight Feb 01 07:57:58 zecke: ok, np Feb 01 08:00:00 zecke: My first impresson is that it is much faster and more responsive than monotone. Feb 01 08:04:51 zecke: Four and a half minutes for clone and at least this operation doesn't take that much of memory. Feb 01 08:07:27 03florian 07org.oe.dev * r99edbe0f... 10/packages/openssh/openssh_4.0p1.bb: openssh: Change location of xauth in filesystem. Feb 01 08:07:32 03xora 07org.oe.dev * r11d7aceb... 10/packages/ (gnome/goffice_0.2.0.bb gnumeric/gnumeric_1.6.2.bb): goffice_0.2.0.bb, gnumeric_1.6.2.bb : new versions Feb 01 08:07:47 03florian 07org.oe.dev * r99edbe0f... 10/packages/openssh/openssh_4.0p1.bb: openssh: Change location of xauth in filesystem. Feb 01 08:07:51 03xora 07org.oe.dev * r11d7aceb... 10/packages/ (gnome/goffice_0.2.0.bb gnumeric/gnumeric_1.6.2.bb): goffice_0.2.0.bb, gnumeric_1.6.2.bb : new versions Feb 01 08:08:54 florian_kc: it seems fast, just the merging is old-skool like Feb 01 08:09:08 'merge' Feb 01 08:10:30 zecke: I'm not sure if that really a disadvantage :-) Feb 01 08:10:59 zecke: But we need to find out if it is reliable, that would be the most important feature. Feb 01 08:11:46 florian_kc: and it forces you to solve a issue directly without giving you background Feb 01 08:11:56 florian_kc: "Do you want to remove..." (y/n) Feb 01 08:12:39 zecke: Hmm... in which context? Feb 01 08:15:32 hrw|work: cedric___ did try and boot a 2.6 kernel the other day Feb 01 08:16:33 ok Feb 01 08:16:41 zecke: It behaves similar to monotone to some extent - e.g. it doesn't care about where in your repository you are. Feb 01 08:16:58 florian_kc: hg up -m ;) Feb 01 08:16:59 78 users test OZ 3.5.4 test2 on 85 machines Feb 01 08:17:03 (context) Feb 01 08:17:10 florian_kc: yes it has multiple heads Feb 01 08:17:45 hrw|work: We're low on developers though :-( Feb 01 08:17:48 zecke: ok Feb 01 08:17:56 hrw|work: nice Feb 01 08:18:32 florian_kc: yep. too bad that most of them does not even provide feedback (not mention fixes) Feb 01 08:18:55 ~lart fish support in konq/3.5.1 for creating empty files Feb 01 08:18:55 * ibot eats fish support in konq/3.5.1's liver with some fava beans and a nice chianti for creating empty files Feb 01 08:18:56 We don't actually have many kernel developers for c7x0 or cxx00 either for that matter :-/ Feb 01 08:19:43 RP: we have someone other then you on it? Feb 01 08:20:47 hrw|work: Not really which was my point... Feb 01 08:22:15 zecke: It seems to be able to run standalone - makes it easy to install. Feb 01 08:22:53 hrw|work: :-/ Feb 01 08:23:52 time to flash oz 3.5.4-test2 Feb 01 08:23:59 florian_kc: gpe again Feb 01 08:24:13 hrw|work: good :-) Feb 01 08:28:00 nasty guy Feb 01 08:43:13 zecke: I added some lines to the wiki... Feb 01 08:46:01 florian_kc: good Feb 01 08:47:52 zecke: To summarize: If it is stable and it turns out it can handle large changes we have a good candidate. Feb 01 08:49:10 oz 3.5.4-test2 gpe booting.. Feb 01 08:49:22 koen|sick: get well soon Feb 01 08:49:35 thanks Feb 01 08:49:48 koen|sick: oh get well soon, well if you are sick of humans as welll we could... Feb 01 08:51:26 the upside is that I will have some time to integrate some patches tonight Feb 01 08:52:17 that is, if people come up with patches for the branch Feb 01 08:52:27 * koen|sick -> soup now Feb 01 08:52:40 smacznego koen Feb 01 08:53:33 koen|sick: What would i need to do to get permissions to apply updates to the stable branch? Feb 01 08:55:30 a blessing from mickeyl, hrw and me Feb 01 08:55:51 you have mine :) Feb 01 08:56:43 florian_kc: as long as they will not touch zaurus machine configs you have mine too Feb 01 08:58:05 hrw|work: np, i have no clue about this stuff :-) Feb 01 08:58:17 mickey|zzZZzz: thime to get up! Feb 01 08:58:20 eh time Feb 01 08:58:41 tsts... sleeping longer than nilsf, incredible! Feb 01 09:01:01 florian_kc: what do you think about removing gpe-clock from panel and adding clock instead? first one looks like shit (overwritten by applets) Feb 01 09:01:48 hrw|work: I'd rather kick mallum to fix the panel. Feb 01 09:02:34 I can kick mallum next week or at fosdem Feb 01 09:02:48 hrw|work: as a medium term solution i'd like to replace matchbox-panel to get a more consistent look&feel. Feb 01 09:02:59 koen|sick: prefer next week - so we could release before fosdem Feb 01 09:04:17 zecke: why does darcs need 76% of RAM on ewi according to top? Feb 01 09:04:37 koen|sick: darcs pull of 332 patches :} Feb 01 09:05:09 koen|sick: I can kill it if you want Feb 01 09:05:18 zecke: just wondering Feb 01 09:05:45 koen|sick: http://oe.handhelds.org/cgi-bin/moin.cgi/DarcsScmTrial Feb 01 09:06:40 fsck... all images to rebuild Feb 01 09:06:57 * koen|sick pushes a fix for gpe-login on h3600 + collie + poodle Feb 01 09:07:15 * florian_kc hands hrw|work some cpu cycles Feb 01 09:07:40 any more going spare florian_kc ? Feb 01 09:07:48 maybe it will be only c7x0... Feb 01 09:10:05 03koen 07org.oe.oz354fam083 * r01706ad4... 10/packages/meta/meta-gpe.bb: meta-gpe: ship industrial theme by default, a short term workaround for gpe-login theming issues Feb 01 09:10:54 heh.. I flashed test1 image... Feb 01 09:13:21 florian_kc: I guess the first step would be to file a bug against the panel in the openedhand bugzilla Feb 01 09:13:48 03koen 07org.oe.oz354fam083 * r86a20cae... 10/packages/apmd/apmd_3.2.2.bb: apmd: fix HUGE security issue, taken from .dev Feb 01 09:15:45 ade|desk: :-) Feb 01 09:16:05 koen|sick: 4755 on /usr/bin/apm? Feb 01 09:16:11 pb_: oh... did we miss this? Feb 01 09:16:21 I fear so Feb 01 09:16:49 * florian_kc shuts up and does some scary gtk stuff Feb 01 09:17:11 hrw|work: that one Feb 01 09:17:46 koen|sick: tested does it works when it is not suid? Feb 01 09:18:13 it worked on my h2200 Feb 01 09:18:15 florian_kc: ! Feb 01 09:18:26 * koen|sick -> food now Feb 01 09:30:40 heh Feb 01 09:30:42 17349 zecke 25 0 1944m 1.7g 10m R 82.2 87.8 3:16.93 darcs Feb 01 09:31:03 apparently things can sucks more resources as monotone + bitbake Feb 01 09:31:25 that's 87.8% of 2GB... Feb 01 09:31:32 ouch Feb 01 09:32:52 So we need a server machine with 10gb if we want to support 5 clients pulling at the same time? ;-) Feb 01 09:33:27 I think that should probably push darcs out the door Feb 01 09:33:42 it's dependency on ghc already did that Feb 01 09:33:46 cu Feb 01 09:33:49 cu Feb 01 09:34:26 today I will call wedding bands so during weekend I will check which one sucks less Feb 01 09:35:28 For this reason i replaced the band by an mp3 player and a shiny new harddisk ;-) Feb 01 09:46:47 okaay darcs loses when applying a 35 mb small patch Feb 01 09:48:05 that's a small patch? Feb 01 09:48:13 yegads, what's a big one? Feb 01 09:48:21 Stephmw: 36mb Feb 01 09:48:33 * Stephmw chuckles Feb 01 09:48:49 actually 36mb is huge not big Feb 01 09:49:22 Stephmw: well if we e.g. would start to hold sourcecode in a tree Feb 01 09:49:35 Stephmw: then this would be something of normal use Feb 01 09:55:58 zecke: you should just accept that importing your p0rn collection into darcs is a losing proposition Feb 01 09:57:23 *g* Feb 01 09:58:14 pb_, especially if 35mb is all he has Feb 01 10:03:22 * florian_kc feels sourrounded by shining pixels Feb 01 10:03:32 Stephmw: that is enough for 'braile-porn' Feb 01 10:03:47 Stephmw: All the Porn in braile format you will ever want... Feb 01 10:04:01 ah, yes, the 'feelies' Feb 01 10:06:33 florian_kc: I think mercurial feels very good Feb 01 10:06:56 florian_kc: it has one push issue though. After a push the client sits there and is doing nothing... Feb 01 10:07:22 zecke: did you try what you did to darcs with it too? Feb 01 10:08:03 and on a h3870: http://handhelds.org/scap/port.644.png Feb 01 10:08:30 drat, wrong window Feb 01 10:09:19 florian_kc: hehe, no I wasn't stupid :} Feb 01 10:09:29 florian_kc: my issue was I started a new darcs tree Feb 01 10:09:44 florian_kc: so the initial patch is 21mb big (not 35) Feb 01 10:10:06 florian_kc: and now I pull two foreign trees and this fails Feb 01 10:10:25 hi ! Feb 01 10:11:16 zecke: eh... okay Feb 01 10:11:21 hi Ifaistos Feb 01 10:11:27 florian_kc: we can easily test that on every scm Feb 01 10:11:37 florian_kc: well not so easily but we can test the limits Feb 01 10:11:57 florian_kc: if you are at rev i, start a new tree starting with rev i+1 Feb 01 10:12:03 and pull that tree in Feb 01 10:12:10 It looks like udhcp has some problem on epia... Dyring First time boot it fails to set the network and once rebooted it get like 99% cpu and drives cpu load to 1.2 :/ Feb 01 10:12:57 either this ot top does not what it's saying ;) Feb 01 10:13:18 zzzz.... Feb 01 10:13:47 If you don't think that ipkg is slow, try booting it on a processor simulator Feb 01 10:14:23 heh Feb 01 10:14:35 pb has a patch to speed up the first boot with ipkg Feb 01 10:16:07 I saw that Feb 01 10:16:22 I think the problem is I/O, which is fairly expensive on this simulator Feb 01 10:16:54 http://bugs.darcs.net/issue123 Feb 01 10:17:02 reenoo: darcs gives fun :} Feb 01 10:17:17 evening Feb 01 10:17:25 zecke: oh, what's it doing? Feb 01 10:17:48 reenoo: pulling a 21mb patch takes 99% cpu and 75% ram (2gb) Feb 01 10:18:07 heh Feb 01 10:18:21 sounds like someone skipped the functional programming courses Feb 01 10:18:26 >:) Feb 01 10:18:53 heh Feb 01 10:21:16 obviously, that was just a joke ;) Feb 01 10:22:03 stuff like that does indeed happen though if you don't use end-recursion and keep everything in memory Feb 01 10:22:59 end-recursive functions coupled with lazy evaluation tends to be fairly efficient Feb 01 10:23:52 zecke: i want to replace the Mainstone kernel with the 2.6.14 kernel with patches for the touchscreen. What features need to be enabled in the config file to be compliant with OE? Feb 01 10:29:10 reenoo: a colleague made a joke "I didn't know haskell could address that much memory" ;) Feb 01 10:29:47 zecke: heh Feb 01 10:30:45 jatreon: I don't know if we have any requirement for that Feb 01 10:31:31 zecke: OK, I will send in a patch for what ever works Feb 01 10:32:20 zecke: Have you tried using the iwmmx flags to build any images yet? Feb 01 10:33:40 jatreon: no images, I build my applications with iwmmx regulary Feb 01 10:33:51 jatreon: but I do not know if the complete image was built using iwmmx Feb 01 10:34:56 zecke: Thank you. I assume your just changed the xscale-mtune file. Are there any other places that need to be modified? Feb 01 10:35:42 jatreon: I need to admit I use a pre configured environment for the iwmmx thing :} Feb 01 10:36:30 zecke: Taking the easy way out. I cannot believe it :) Feb 01 10:36:47 zecke: thank you Feb 01 10:37:56 http://pastebin.ca/39497 Feb 01 10:37:59 well Feb 01 10:45:00 RP: so it looks like the LED class has polarized the LKML community into embedded / non-embedded parties... :) Feb 01 10:48:48 good evening Feb 01 10:58:22 hi Feb 01 11:01:30 how cvan i tell monotone not to verify all checked out revision please ? Feb 01 11:01:38 there is about 4 Feb 01 11:02:14 there is about 4,234 to verify, and it will take me the whole night to get then all verified :s Feb 01 11:02:27 any way to avoid this !? Feb 01 11:02:32 ~ewi Feb 01 11:02:34 rumour has it, ewi is ewi546.ewi.utwente.nl - main backup server for OE monotone. Hosts database snapshot: http://ewi546.ewi.utwente.nl/OE/OE.db.bz2 Feb 01 11:02:52 ibot: ? Feb 01 11:03:07 i just have to get this file ? Feb 01 11:03:16 is it what you mean ? Feb 01 11:03:28 Dazgard: hehe Feb 01 11:03:42 Dazgard: yes, instead of the initial pull from ewi Feb 01 11:03:50 Dazgard: you can download this OE.db and use it Feb 01 11:03:51 that's cool :) Feb 01 11:03:56 thanks guys Feb 01 11:04:09 the verification take so long !! Feb 01 11:04:56 anyway cya later :) Feb 01 11:05:27 ;) Feb 01 12:36:46 hi reenoo Feb 01 12:38:19 hey pb__ Feb 01 12:42:27 i'm having a problem. Feb 01 12:42:30 :) Feb 01 12:42:52 during a bitbake build of e-image-core Feb 01 12:43:28 i'm getting the following error during compilation of orinoco-modules-0.13e-r4 Feb 01 12:43:42 cc1: bad value (xscale) for -mtune= switch Feb 01 12:43:54 has anyone seen this before? Feb 01 12:44:09 yeah Feb 01 12:44:19 I think jbowler fixed this already Feb 01 12:44:23 is your tree up to date? Feb 01 12:44:28 i think so Feb 01 12:44:33 i got it last night Feb 01 12:44:47 which branch? Feb 01 12:45:22 it's definitely fixed in .dev but still a problem in .oz354fam083 Feb 01 12:45:26 i'm using the oz354fam083 Feb 01 12:45:28 branch Feb 01 12:45:31 ahh Feb 01 12:45:46 that would splain it Feb 01 12:46:30 is there a path that you know of? Feb 01 12:46:36 path = patch Feb 01 12:48:03 revert the most recent change to module-base.bbclass Feb 01 12:49:04 i hate to be that guy, but how might I do that? ( I just started learning about all this last night ) Feb 01 12:54:47 good nite Feb 01 12:56:05 cya Feb 01 12:56:11 'night zecke Feb 01 12:56:12 zecke: later Feb 01 13:02:15 gremlin484: identify the change with monotone log/diff and revert it I guess Feb 01 13:05:20 anybody know if a circuit diagram for C760 can be found anywhere? Feb 01 13:05:50 highly unlikely. as far as i know, we still don't have any sharp engineering contacts Feb 01 13:06:05 k Feb 01 13:06:52 morning kergoth Feb 01 13:09:16 hey pb Feb 01 13:09:19 * kergoth reads up on qt4 Feb 01 13:19:16 reeno: thanks i'll look into it. should i then get rid of build/tmp before running bitbake again? Feb 01 13:20:09 no, shouldn't be necessary Feb 01 13:20:35 ok thanks Feb 01 13:20:56 kergoth: how's arizona? Feb 01 13:21:14 its nice. will be 78-79 degrees F this weekend. hehe Feb 01 13:21:22 i may change my tune come summer of course Feb 01 13:21:27 heh, right Feb 01 13:21:36 presumably summer is no worse than in texas, though Feb 01 13:21:49 hotter, but less humid, so probably wont feel quite as bad Feb 01 13:21:52 whatve you been up to? Feb 01 13:22:59 spent most of the past few weeks preparing a tender for a big contract we're bidding for. other than that, mostly working on bits for familiar 0.8.3. Feb 01 13:23:19 03koen 07org.oe.oz354fam083 * rc13b96c0... 10/classes/module-base.bbclass: module-base.bbclass: revert last change which did more bad than good Feb 01 13:23:31 ah thats cool Feb 01 13:24:28 i should play with oe some again. swiped its site file yesterday to crosscompile bash for the embedded device here Feb 01 13:24:31 * kergoth hugs oe Feb 01 13:24:38 hehe Feb 01 13:25:02 yeah, me too. it's been interesting playing with these new SCMs in zecke's trial. Feb 01 13:26:08 how are they looking so far? i keep meaning to try out mercurial Feb 01 13:26:40 so far I've only tried git and darcs, neither of which has really grabbed me. Feb 01 13:26:51 git seems better than darcs, but it's a bit inscrutable. Feb 01 13:26:55 plus I still can't get over the name. Feb 01 13:27:00 hehe Feb 01 13:27:08 CIA-4 any advice on how one might do so? Feb 01 13:27:22 gremlin484: i dont think the bot is going to answer you Feb 01 13:27:24 next I plan to try mercurial and svk Feb 01 13:27:30 lol Feb 01 13:28:41 whats the purpose in the bot then? Feb 01 13:28:43 for example, a couple of days ago I wanted to find out who had made a particular change to one .bb file. I managed to do so with git but I had to write a python script to do it. with darcs I failed altogether, even though "darcs trackdown" seems like the perfect tool for the job. Feb 01 13:28:53 another gremlin !!!! :) Feb 01 13:29:00 AHHHHH!!! Feb 01 13:29:11 pb_: wow, thats terrible. that's such a common usecase, too Feb 01 13:29:17 yeah Feb 01 13:29:20 gremlin484: see http://cia.navi.cx/ Feb 01 13:29:50 I was dismayed that "darcs annotate" (basically the same function as cvs annotate) on a single .bb file took twelve minutes to run on my sempron 3000+ and consumed 1.8GB of RAM in the process Feb 01 13:30:09 hah Feb 01 13:30:56 arguably I should just get a better computer, but that kind of resouce usage does seem a bit much Feb 01 13:32:42 heh Feb 01 13:32:56 I'd vote for fixing darcs instead Feb 01 13:33:17 i've really enjoyed using svk for the past 6 months or so on my personal code and world of warcraft addon set (which had 800 or so vendor branches, around 120 of which were on the trunk at any given time, with local modifications there) Feb 01 13:34:07 actually I was wondering why the bot couldn't help me fix my problem :) Feb 01 13:34:31 kergoth: ah, sounds good. sv{kn} does also have the advantage of being a much more mature solution than any of the others. Feb 01 13:35:22 yeah, i like that about it. and the fact that upstream doesnt have to change, so you keep the svn mindshare and tools availability. people dont need svk to check it out Feb 01 13:35:46 hmm Feb 01 13:36:22 kergoth: yeah. svn+svk works pretty well for our ~30 people project at university. svn for regular users and svk for those who want offline features or for me who needs to merge stuff to the svn repo of the research group we're part of Feb 01 13:38:32 I also like the excellent documentation for svn Feb 01 13:39:13 "How do I ...?" - "Read it up in the svn book." and people will actually do that and usually be happy Feb 01 13:41:16 right, that's also a big advantage Feb 01 13:42:25 dumb question: bad idea to just copy the module-base.bblcass from .dev to the .oz354fam083 tree to possibly fix my issue? Feb 01 13:42:41 I did a diff on the two files and there are only a few minor code changes Feb 01 13:42:56 as for monotone, it was interesting (if slightly embarrassing) to read that thread on the monotone development list about our requirements. sounds like the monotone developers themselves do not yet consider it ready for real "production" use in big projects; for example, they were talking about making at least one or two more incompatible netsync protocol changes. Feb 01 13:42:58 mostly having to do with the TARGET_CC_KERNEL_ARCH Feb 01 13:43:22 gremlin484: just pull/update Feb 01 13:46:40 so the can be a specific bbclass? Feb 01 13:48:30 pb__: my understanding is that they've never consider it ready for production use yet. they will continue to introduce incompatible changes until they think everything's right which will probably be a 1.0 release (plus some additional polishing maybe). like one poster to monotone-devel said, it's entirely our fault we're using it already (with wrong expectations). Feb 01 13:48:43 right Feb 01 13:48:56 they did have a fair point that the 0.x version number should have been a clue Feb 01 13:49:19 yeah Feb 01 13:49:28 svn wasn't much better at that stage Feb 01 13:49:50 but that's mostly bdb's fault >:) Feb 01 13:49:55 ~lart bdb Feb 01 13:49:56 * ibot stuffs bdb into a shiny new tin can and vacuum seals it Feb 01 13:50:10 ~botsnack Feb 01 13:50:10 reenoo: :) Feb 01 13:51:47 on the other hand, the nslu2-linux people seem to consider true distributed operation to be something of a sine qua non. Feb 01 13:55:00 I'm not entirely sure why that is: I would have thought that bbfile collections would be a better model for their use case. Feb 01 13:58:34 would be an interesting point to discuss with them Feb 01 13:59:41 bbfile collections has always worked for us Feb 01 13:59:47 fwiw Feb 01 14:00:37 I don't know how the trees on their servers are managed, but they're no longer merging from a branch to .dev on our servers. so, they're essentially just running a mirror which can be done with any SCM Feb 01 14:01:17 ah, right Feb 01 14:02:26 (merging as in monotone propagate/merge) Feb 01 14:04:35 anyways, time to play with somen artificial neural networks Feb 01 14:04:42 some even Feb 01 14:04:46 later all Feb 01 14:04:53 enjoy Feb 01 14:05:16 CosmicPenguin: The LKML led discussion did become rather polarised :-/ Feb 01 14:05:57 RP: that sort of stuff annoys me - just because it doesn't work for a 16x processor mainframe, doesn't mean its for shit Feb 01 14:07:04 CosmicPenguin: The comment that annoyed me was the one about it being embedded stuff so everyone should just do their own thing... Feb 01 14:08:09 Yeah, thats what got me to respond Feb 01 14:09:02 Apart from the one trigger, the rest of the code doesn't seem to havehad much in the way of objections against it Feb 01 14:09:38 Just some comments which amount to the function names being misleading (which I agree with and will change) Feb 01 14:10:37 03pH5 07org.oe.dev * rb54b286d... 10/packages/udev/ (6 files in 2 dirs): Feb 01 14:10:37 udev: add 084 Feb 01 14:10:37 - Fixes the severe udev initial boot speed issues for hx4700. Feb 01 14:10:37 - DEFAULT_PREFERENCE="-1", please try if it works for you. Feb 01 14:10:37 - Uses udevsynthesize to generate initial hotplug events. Feb 01 14:16:55 * CosmicPenguin grabs that Feb 01 14:17:13 If its slow on real silicon, imagine how much fun it is on a simulator Feb 01 14:20:09 heh Feb 01 14:24:17 I am getting this error when trying to run monotone update -rjan2006prebreakage: monotone: misuse: no match for selection 'jan2006prebreakage' Feb 01 14:24:17 is it no longer needed? Feb 01 14:45:02 Great. FIlesystem corruption on the server :-( Feb 01 14:45:10 Erk. Feb 01 14:48:17 would that be the cause of my error? Feb 01 14:48:44 raphael_burnes: You're building off .dev? Feb 01 14:49:26 yes, I'm just following the getting started howto Feb 01 14:49:41 raphael_burnes: Run this: monotone -d oe.db list tags Feb 01 14:49:53 This is the line that should be there: Feb 01 14:49:55 jan2006prebreakage 0b9a2dfead1136f996280e9827949f83cb032537 rpurdie@openembedded.org Feb 01 14:50:13 I have some weird openslug lienes Feb 01 14:50:18 lines* but not that one Feb 01 14:50:55 NA|Zzz, how to get the pre-breakage? Feb 01 14:50:59 Hmm.. sounds like you might have an incomplete pull Feb 01 14:51:14 raphael_burnes: I'll chuck my oe.db up on the net if you want, and you can use that as a starting point Feb 01 14:51:33 that would be awesome, thank you! Feb 01 14:53:08 I will warn though.. I've only got 128kbit up, so it might take an half an hour or so :) Feb 01 14:53:14 http://zaurus.tucuxi.org/oe.db.bz2 Feb 01 14:53:35 thats okay, its much appreciated! =) Feb 01 14:53:42 no probs =) Feb 01 14:54:02 raphael_burnes: Read http://oe.handhelds.org/cgi-bin/moin.cgi/SpeedUpMonotone and download from ewi Feb 01 14:54:13 then do I checkout a local copy? Feb 01 14:54:21 yep Feb 01 14:54:26 raphael_burnes: yeah, monotone -d oe.db checkout org.openembedded.dev Feb 01 14:54:38 Anyways, off to work Feb 01 14:55:07 thanks again, enjoy NA|Zzz Feb 01 14:59:09 do13, that looks great, thank you Feb 01 14:59:15 much faster Feb 01 15:11:36 hey for the e-image-core build, the edb version bitbake is looking for does not exist anymore.... Feb 01 15:11:58 http://enlightenment.freedesktop.org/files/edb-1.0.5.004.tar.gz Feb 01 15:12:34 i'm gonna look elsewhere for the package, i just thought someone should know... Feb 01 15:15:31 found the package on a different server... do i just drop the tar.gz into ~/sources/ ? Feb 01 15:28:55 ok that didn't work. how do you fetch a package for bitbake? Feb 01 15:32:10 gremlin484: short fix: drop into ~/sources/ and generate md5sum. Better way change SRC_URI and open new bugreport Feb 01 15:35:27 ok i realized the md5 part, and it's on it's waya again. Feb 01 15:35:35 i'll open a bugreport Feb 01 15:57:45 hi Feb 01 16:03:55 meh something is killing my network in a 2.6 image when root=NFS Feb 01 16:37:41 RP: are there any known problems w/ NFS in the 2.6 kernel? NFS mounts keeps dying on me w/o reason **** BEGIN LOGGING AT Wed Feb 01 19:28:12 2006 Feb 01 19:28:15 correct, that's why I can't tell =) Feb 01 19:28:18 hmmm..... Feb 01 19:28:31 that's the problem with the Akita ;-) Feb 01 19:28:39 =) Feb 01 19:28:53 I like having multiple dev images on my HD and still being able to boot back into my old root Feb 01 19:29:05 altboot is grand Feb 01 19:29:16 need to figure out what's wrong with my kernel... Feb 01 19:29:21 very useful for testing indeed Feb 01 19:30:10 I was surprised NFS booting worked w/o mods in 2.6 Feb 01 19:31:17 * CoreDump|home wants a kexec kernel =) Feb 01 19:31:20 ah....2.6..... Feb 01 19:31:34 yeah, 2.6 still isn't enough for me Feb 01 19:31:38 the keyboard us very funky Feb 01 19:31:42 2.4 isn't working Feb 01 19:31:48 that's what I meant **** ENDING LOGGING AT Wed Feb 01 19:32:01 2006 **** BEGIN LOGGING AT Wed Feb 01 19:44:11 2006 Feb 01 19:44:43 hmmm.... Feb 01 19:44:59 or nfs-booted GPE images for that matter Feb 01 19:45:00 if I try to boot with a 2.4 kernsl it fails... Feb 01 19:45:08 does it get added to the bootstrap image perchance? Feb 01 19:45:18 well...it doesn't seem to load at all..... Feb 01 19:45:24 can't find init Feb 01 19:45:28 yep, something pulls it in Feb 01 19:46:10 hmmm Feb 01 19:46:14 http://hentges.net/tmp/do_not_use/Akita/kernels/ Feb 01 19:46:18 I didn't "reflash" my bootstrap... Feb 01 19:46:25 working kernel Feb 01 19:46:34 I have a spitz, of course Feb 01 19:46:45 same kernel Feb 01 19:46:47 =) Feb 01 19:47:16 for 2.6 you mean? Feb 01 19:47:21 I can boot with 2.6.... Feb 01 19:47:25 well not finding init = FLASH Feb 01 19:47:30 no 2.4 Feb 01 19:47:34 I'm pretty sure the 2.4 kernel compiles different things on the akita and spitz Feb 01 19:47:40 I realize that Feb 01 19:47:55 from what i saw in the patches..... Feb 01 19:48:03 yes, the 2nd pcmcia controller is disabled in akita Feb 01 19:48:47 that's not all, though Feb 01 19:48:54 there's some other different hardware IIRC Feb 01 19:49:21 I diffed the config once, very few differences Feb 01 19:49:22 and, as I said, the defines in the ssource seemed to turn different code on for the different devices Feb 01 19:49:27 not the config Feb 01 19:49:29 the code Feb 01 19:49:46 well, the code i did not check Feb 01 19:49:56 if you choose AKITA other than SPITZ it will be different Feb 01 19:50:04 i c Feb 01 19:52:16 mehhhhhhh Feb 01 19:52:43 S99entrance fubars init when started from ssh Feb 01 19:53:48 fubars init? Feb 01 19:55:08 kills sshd, respawns VT login Feb 01 19:55:23 wow Feb 01 19:55:28 heh Feb 01 19:55:28 that's nice Feb 01 19:55:41 sounds like a network problem maybe.....::shrug:: Feb 01 19:56:16 well, can't tell sine VT logins are srewed, too heh Feb 01 19:56:59 how so? Feb 01 19:57:05 I can login on the console... Feb 01 19:57:32 not after the ssh entranced trick =) Feb 01 19:57:36 ah Feb 01 20:01:22 hmm Feb 01 20:01:25 2837 root 572 S /bin/sh -c /etc/X11/Xserver -auth /var/:0.Xauth :0 Feb 01 20:01:45 # $DISPLAY Feb 01 20:01:45 -sh: :0: not found Feb 01 20:02:00 ? Feb 01 20:02:04 looks ok to me Feb 01 20:02:05 # entrance Feb 01 20:02:05 Xlib: connection to ":0.0" refused by server Feb 01 20:02:05 Xlib: No protocol specified Feb 01 20:02:12 that's what I get too Feb 01 20:02:48 why is that, work when X is launched manually Feb 01 20:03:07 sorry, not sure what you mean Feb 01 20:04:31 my / is also ro....not sure why Feb 01 20:05:56 WTF is this with ipv6? Feb 01 20:06:09 what I was trying to say was that entranced launches an X server which is (as per ps -ax) running on :0 but X apps can't connect Feb 01 20:06:20 I know Feb 01 20:06:26 I see the same thing Feb 01 20:06:30 ok Feb 01 20:06:32 I don't know why it's happening Feb 01 20:06:40 and this ipv6 thing is veyr annoying Feb 01 20:06:48 AFAIK it never used ipv6 before... Feb 01 20:06:58 hmm Feb 01 20:07:11 delete the module and reboot Feb 01 20:07:18 what module? Feb 01 20:08:02 there are no modules in the flash root and no ipv6 module in /lib Feb 01 20:09:25 hmm Feb 01 20:13:29 grrr...it's happening again Feb 01 20:13:45 the "Starting PCMCIA services" thing umount -v /dev/hda1 and 2 Feb 01 20:15:54 now that *might* be a udev thing Feb 01 20:16:09 in 2.4? Feb 01 20:16:28 this happened to me before too, but that was on a highly used image which I'd probably screwed up myself... Feb 01 20:16:35 this is a pretty new "install" Feb 01 20:16:40 * JustinP is going to rebuild... Feb 01 21:15:50 does anyone here generally use the "PARALLEL_MAKE" option in local.conf? Feb 01 21:17:33 gremlin484: I do Feb 01 21:17:58 any negative side effects? Feb 01 21:18:27 ( for a single processor machine) Feb 01 21:18:33 smp kernel Feb 01 21:18:57 I use -j2 on a single CPU Feb 01 21:20:20 significant performance increase? Feb 01 21:20:36 increace? (sp) Feb 01 21:20:58 gremlin484: Not really.. I use 2 or 3 simul jobs... tends to speed things up a decent amount Feb 01 21:21:03 Mind you.. I've a shitload of ram.. 2gb. Feb 01 21:21:28 I've disabled it for sake of trouble shooting, and glibc is taking fooooooorrrrreeeevvvvveeeerrrrr Feb 01 21:21:40 yeah i have 1.5gb.... Feb 01 21:22:21 I've never tried it without Feb 01 21:23:10 awesome.... suppose one were to change that option during a compile... think it might break something? Feb 01 21:23:22 although bitbake just picks up where it left off right? Feb 01 21:23:30 (if it's stopped) Feb 01 21:23:37 right Feb 01 21:23:47 i'll do that then. Feb 01 21:23:48 thnx Feb 01 21:24:07 oh, and raster pointed something out today... Feb 01 21:24:33 e17 @ 240x320 would pretty much be unusuable. Feb 01 21:24:39 possibly Feb 01 21:24:41 (on a 5500) Feb 01 21:24:44 :) Feb 01 21:24:49 but i'm gonna try anyways. Feb 01 21:24:58 I *did* add better edge detection for the auto-menu moving, though :-) Feb 01 21:25:16 i'll make a small theme for the zaurus if i can get it to work. Feb 01 21:25:51 :-) Feb 01 21:26:47 entranced_auth_display_secure: Setting up access for display :0 Feb 01 21:26:48 Cookie generation failed: could not read /dev/random Feb 01 21:27:00 hmmm, /dev/random is there allright Feb 01 21:27:23 perms? Feb 01 21:27:49 # ls -al /dev/random Feb 01 21:27:50 crw-rw-rw- 1 root root 1, 8 Jan 1 00:03 /dev/random Feb 01 21:27:56 hmmm Feb 01 21:28:02 identical to the one on my desktop Feb 01 21:28:31 brb Feb 01 21:28:44 # ipkg install strace Feb 01 21:28:56 * CoreDump|home thinks it's good to have his own feed Feb 01 21:29:17 :-) Feb 01 21:29:45 can anyone tell me why I get a failed dependency for libts and there seems to be no provider? Feb 01 21:30:21 becaus eyou're looking in the wrong feed perhaps? Feb 01 21:33:04 * shadows laughs at CosmicPenguin's comment Feb 01 21:34:42 JustinP: I'm building opie-image, I'm not looking in any feed. Feb 01 21:35:02 JustinP: fails on do_rootfs Feb 01 21:35:14 JustinP: and surprisingly, I'm not out of diskspace :-) Feb 01 21:35:22 strange Feb 01 21:35:25 what branch? Feb 01 21:35:56 Zero_Chaos: tslib Feb 01 21:35:58 like Feb 01 21:36:02 the names might not match up Feb 01 21:36:13 I'll try tslib :-) Feb 01 21:36:16 thanks Feb 01 21:39:52 hi again Feb 01 21:39:59 werd Feb 01 21:41:23 got a tv card today... making a myth-tv box :) Feb 01 21:41:29 had to drop the card in Feb 01 21:41:50 anyone tell me how much space "bitbake world" takes up? give or take... Feb 01 21:42:12 a good bit Feb 01 21:42:24 never done it myself Feb 01 21:42:48 JustinP: any idea more exact? are we talking 10 gig, 15 gig? 200 gig? Feb 01 21:42:49 Zero_Chaos: I'd allow at least 5gb Feb 01 21:43:00 Never done it here though Feb 01 21:43:08 5-10gb would probably be a safe bet. Feb 01 21:43:49 NA|Zzz: I don't know about that, I've got an 8 gig drive for OE and it's 82% full just for "bitbake opie-image" Feb 01 21:44:05 nope, a world build is between 20 and 30Gb Feb 01 21:44:06 Bloody hell.. Feb 01 21:44:23 well, it _was_ a few month ago Feb 01 21:44:29 Never thought it'd be that big Feb 01 21:44:37 um the oe wiki says 35Gb IIRC Feb 01 21:45:49 heh Feb 01 21:45:56 Wow, I figured I didn't have the space, 35Gb confirms that Feb 01 21:45:57 yeah, I'd say allow 35G Feb 01 21:45:59 CoreDump|home: thanks :-) Feb 01 21:46:12 it's a lot of source and such Feb 01 21:46:23 and OE leaves all of the work as-is, no cleanup Feb 01 21:46:36 and there are 1000s of programs... Feb 01 21:48:40 one would think there'd be an option for some type of source cleanup with bitbake Feb 01 21:51:46 not afaik Feb 01 21:51:51 it's meant for developers Feb 01 21:51:58 so to have source there is nice Feb 01 21:53:04 true Feb 01 21:58:14 is it possible to watch bitbake compile, other than tailing the log files? -v does not appear to dump compiler output to the screen... Feb 01 21:58:24 AFAIK no Feb 01 21:58:31 I asked that in the beginning too Feb 01 21:58:46 it's really enough to watch the stuff it does output...especially when compiling over ssh Feb 01 22:00:43 yeah... i just start to wonder when i've been staring at "NOTE: package glibc-2.3.5+cvs20050627-r1: task do_compile: started" for an hour Feb 01 22:01:17 well, glibc takes forever Feb 01 22:01:22 shouldn't be an hour, though.... Feb 01 22:02:00 you should wlways set parallel make to 2+ as long as you're not debugging it Feb 01 22:02:08 i'm compiling mythtv at the same time Feb 01 22:02:08 what speed machine? Feb 01 22:02:11 ah Feb 01 22:02:15 well that's your fault then ;-) Feb 01 22:02:26 amd xp 2800+ :) Feb 01 22:02:28 i know Feb 01 22:02:38 it must be a gentoo thing Feb 01 22:02:52 i'll stare at emerge for hours... Feb 01 22:03:54 funny thing about that is you cant turn off the compile output, and some messages spit out really important messages to the screen that stay up for lik 5 seconds, then start on the next package.... Feb 01 22:04:08 I know Feb 01 22:04:12 some of them finally got smart and started beeping a few times at you Feb 01 22:04:13 I like compiling Feb 01 22:04:17 it's why I work with OE ;-) Feb 01 22:04:47 but unless you have your speakers turned up, or are not connected over ssh, you miss all those messages, and you're sol Feb 01 22:05:04 I find those don't come up too often.... Feb 01 22:05:09 yeah Feb 01 22:05:15 and with dispatch-conf usually things work pretty well Feb 01 22:06:02 it's frusterating though when after an hour of troubleshooting, you find out it TOLD you the answer halfway in the middle of emerging the package Feb 01 22:06:05 :) Feb 01 22:06:30 yeah Feb 01 22:06:48 isn't dispatch-conf the etc-update alternative?. Feb 01 22:06:52 yes Feb 01 22:06:54 much better Feb 01 22:06:58 really? Feb 01 22:06:59 IMHO Feb 01 22:07:33 i've herd a little about it, never looked into it though Feb 01 22:08:46 it can also RCS your conf files so you can always go back :-) Feb 01 22:08:58 thats cool. Feb 01 22:09:11 and it can auto-change files with whitespace-only changes or files which you've never touched Feb 01 22:09:13 i hate merging conf files. Feb 01 22:09:22 dispatch-conf makes it much easier Feb 01 22:09:36 cool i'll try it Feb 01 22:09:45 yay, I'm past glibc and gcc-cross Feb 01 22:09:50 :) Feb 01 22:10:44 i wonder if one could get distcc working with bitbake Feb 01 22:10:56 do a little cluster compiling :) Feb 01 22:15:34 it's supposed to be able to....but I never got the chance to try Feb 01 22:17:58 yeah i'm too unmotivated to figure it out... Feb 01 22:21:11 whoa Feb 01 22:21:19 I started entrance over ssh and it *worked*! Feb 01 22:22:07 using dev? Feb 01 22:22:41 yes Feb 01 22:23:34 although I can't login...:-( Feb 01 22:24:10 hmmm...and e is having evas issues.... Feb 01 22:26:37 my net just got defunked Feb 01 22:27:13 surry JustinP you were saying, i need oe, would that be the latest source or....? Feb 01 22:27:27 i currently have an build for the htc-blueangel Feb 01 22:27:31 oe is OpebEmbedded, the build metadata Feb 01 22:27:37 Open Feb 01 22:28:55 ok well ill see how this goes then Feb 01 22:29:52 glibc done... Feb 01 22:32:09 * Geo_KM is away: Away at the moment Feb 01 22:36:35 I hate to do this, but I cannot access bugs.openembedded.org... so... anyone willing to accept a dcc of a .bb for submission? Feb 01 22:37:04 bugs.treke.net Feb 01 22:37:29 * Zero_Chaos hugs treke Feb 01 22:37:36 JustinP: thanks for the pointer Feb 01 22:37:44 yep Feb 01 22:43:43 if someone would be polite enough to check out bug #650 and tell me if I need to change anything.... Feb 01 22:46:39 anyone? I'm a first time submitter and I would love it if someone would make sure I did it right :-) Feb 01 22:47:38 sorry, can't help ya.... Justin could probably... I think he's troubleshooting some e17 stuff right now. Feb 01 22:48:11 JustinP, does what you are talking about in #edevelop effect the e-image-core build i'm doing, or is is something within your environment? Feb 01 22:48:14 no big deal. I just wrote a .bb for aircrack 2.41 a while ago and I've been lame about submitting it. Feb 01 22:48:31 cool Feb 01 22:48:47 yeah, if I ever get void11 working right I'll submit that too. Feb 01 22:50:33 well, time for sleepy, any complaints, and I'm sure I'll get an email Feb 01 22:50:35 'nite all Feb 01 22:50:38 thanks all Feb 01 22:50:42 Zero_Chaos: hey Feb 01 22:50:45 urgh Feb 01 22:53:54 RP: if you would be willing to provide test kernels, i am happy to flash and test Feb 01 22:55:13 * JustinP if off to bed Feb 01 22:55:15 is Feb 01 22:55:55 nite Feb 01 23:01:37 expat could use an update Feb 02 00:06:46 RP: i built a kernel with that suggested hack/ commenting out of code, testing now Feb 02 00:09:19 goodmorning ! Feb 02 00:12:18 RP: no love with the AC adapter left in. my Z does not resume. Feb 02 01:35:53 How do you excape quotes in bitbake.conf Feb 02 01:37:17 it probably should be \" Feb 02 01:37:21 lemme try that Feb 02 01:38:16 hi Feb 02 01:43:00 shadows: Do you have a serial lead? Feb 02 01:44:18 morning Feb 02 01:56:30 hi dirk Feb 02 01:57:22  Feb 02 01:57:33 hi Feb 02 01:57:45 Hi Richard, cedric___ Feb 02 01:58:14 RP: Did you fixed your filesystem? Feb 02 01:58:43 what people use for the serial connection to the zaurus? Feb 02 01:58:44 morning Feb 02 01:58:56 under Linux or Window? Feb 02 02:00:27 I'm using ANSI 115200 8-N-1 Feb 02 02:01:09 I can receive but what I'm sending to the terminal is not reconised Feb 02 02:01:32 so I can't connect via a serial console Feb 02 02:04:17 so I can't debug my freshly built kernel... Feb 02 02:04:45 cedric___: Which serial lead are you using? Feb 02 02:05:22 null modem Feb 02 02:07:23 cedric___: SerialIO or Sharp serial lead? Feb 02 02:10:20 do13_: Its been "fixed", yes. I suspect fixing it involved deleting half of it though :-( Feb 02 02:11:05 do13_: Its mostly OE that suffered though and I was due a rebuild :) Feb 02 02:11:35 cedric___: I usually use minicom with the settings you mention Feb 02 02:12:18 I hve a SerialIO lead and it doesn't work with the C3x00/C1000 models properly :-( Feb 02 02:12:25 (I can only see data, not send it) Feb 02 02:15:09 RP: for the other lost data you can use the backup:) Feb 02 02:15:24 shadows: The next area of code I wonder about is the last if block in corgi_enter_suspend(). It should be safe to comment that block out Feb 02 02:15:55 do13_: I'm reviewing my backup policy ;-) Feb 02 02:16:32 RP: suffered HD crash? Feb 02 02:16:35 RP: good hint. I need to do that as well:) Feb 02 02:18:10 XorA: File system corruption (ext3), highly likely due to OE's heavy use of it Feb 02 02:18:26 RP: nasty, thats why I had to stop using Reiser, OE broke it Feb 02 02:18:41 RP: cross fingers, so far ext3 has held up Feb 02 02:19:25 I'am using XFS and at this there are no probs. Feb 02 02:19:53 they only time Ive had ext3 corruption on my machines turned out to be bad ram Feb 02 02:20:01 that took me hours to track down Feb 02 02:21:29 The machine in question has about 20 different partitions over 4 disks and only one was corrupted so hopefully the hardware is ok Feb 02 02:21:50 we hope so too, nothing is allowed to slow the RP juggernaut :-) Feb 02 02:22:45 I'd been using the same tmp directory since April last year on that drive! Feb 02 02:22:50 now if I could just stop my amd64 overheating Feb 02 02:24:55 This is why I use Intel - never had as many cooling problems... Feb 02 02:25:54 RP: not the CPU itself, but the shuttle Ive got has some problems with cooling the back end of the MB and tends to flake out, not helped by fact linux seems to fsck the fan control slightly, winXP doesn it different Feb 02 02:26:54 I guess I've just been bitten once too many times by problems with older amds - its kind of put me off them... Feb 02 02:27:34 my sequence of pc's is K6->K6-3->Thunderbird->amd64 :-) Feb 02 02:28:26 I've got a serialio cable a none null modem and I've connected a null modem to it... Feb 02 02:29:32 RP: so far is booting with the correct rootfs Feb 02 02:30:48 but the only stupid thing --> there is some debugging by default from evbug.c when I'm using the zaurus keyboard... Feb 02 02:31:26 cedric___: Don't compile evbug.c in Feb 02 02:31:43 cedric___: Its a massive pain and actually breaks things Feb 02 02:32:13 ah ok, it was by default in the poodle defconfig file... Feb 02 02:33:24 my next task is to merge the backlight code from Pavel Feb 02 02:35:06 can u explain the theory around the power management code? Feb 02 02:35:27 cedric___: The pm code or the battery charging code? Feb 02 02:35:53 don't know explain both thing, pls... Feb 02 02:36:39 just wondering what need to be done for the poodle Feb 02 02:37:04 that's mean the battery is not charge by the hardware? Feb 02 02:44:59 It is charged by the hardware at one level but there is also a lot of software control Feb 02 02:45:32 Basically, you need to look at the differences in the sharp pm code for poodle and import those differences to the sharpsl_pm driver Feb 02 02:45:37 Its horrible code though :-( Feb 02 02:47:07 ok and for the test strategy, how can I know if everything works? Feb 02 02:47:34 any test procedure? :-) Feb 02 02:49:17 I think the SD code is not working for the poodle, stiil the same problem since last year... Feb 02 02:49:51 pcmcia is fine Feb 02 02:50:21 usb is initialise via the kernel msg but I didn't test it Feb 02 02:51:00 irda is inialised by the kernel but I think there is a missing config file from /etc/sysconfig Feb 02 02:51:17 in the rootfs Feb 02 02:51:21 Any idea what isn't working with the SD code? It works on the other devices which should be very similar... Feb 02 02:51:48 First get suspend/resume working and just implment dummy code for the charging hooks Feb 02 02:52:12 Then get battery measurement working (temp/voltage) Feb 02 02:52:28 Only once that all works do you actually let it start charging the battery Feb 02 02:52:54 don't know I haven't be able to login remotly to the device Feb 02 02:53:10 so I should remove the battery for the time been? Feb 02 02:53:37 or is it safe to keep it with the charger on? Feb 02 02:54:30 Its got a hardware charging controller in there so I suspect its safe to lave in regardless, as long sa you keep an eye on its temperature Feb 02 02:55:13 so keep it in with the charger off should be fine. If the charger is on, watch the temperature Feb 02 02:56:01 but if the charging led is not on then it "should" be ok? Feb 02 02:56:08 hi Feb 02 02:56:27 i set set up a build envirenement for openzaurus Feb 02 02:56:44 cedric___: The charging led is controlled by software Feb 02 02:57:18 i've a question, does bitbake have to rebuild all dependencies in order to compile a task ? Feb 02 02:57:32 if the device is not booting then the led is orange when I plugged the charger Feb 02 02:57:54 or can it passthrought the compiled one and do only task taht do need real compilation ? Feb 02 02:59:04 Dazgard: if the dependencies are already compiled then it'll skip them **** ENDING LOGGING AT Thu Feb 02 02:59:56 2006