**** BEGIN LOGGING AT Mon Mar 05 02:59:57 2007 Mar 05 04:19:18 hmm.. can anyone tell me where you can set the size of a generated ext2 image? Mar 05 05:24:21 Hello fooks Mar 05 05:30:30 hey likewise! Mar 05 05:40:15 fooks? Mar 05 05:49:09 it's dutch. Mar 05 05:56:00 mwester: your last name sounds Dutch Mar 05 05:56:15 mwester: is Dutch actually Mar 05 05:56:26 Yep. Mar 05 05:56:36 Where are you from? Mar 05 05:57:01 All over, I guess. Chicago right now. Mar 05 05:57:24 Lived in Canada for a while, as well as Amsterdam Mar 05 05:58:27 mwester: Well, I just made my first trip into the US, I'm in SF now, I'm dutch. Mar 05 06:00:21 mwester: Canada is #1 on our wanna-live list. Mar 05 06:01:33 Canada's a nice place; I enjoyed it. Ended up in the US because that's where the jobs were. The high-tech field has gotten much better in Canada,though. Mar 05 06:02:36 mwester: my alter ego would like to move to BC, I am not sure about the tech jobs there, but she is into animal wildlife stuff, which is harder to find than tech stuff these days. Mar 05 06:03:28 mwester: we really should be working on it again, we moved to a new house last year, and plans to move got frozen for the short term while she is starting her own business. Mar 05 06:04:03 mwester: well, it allowed me to finally finish my MSc, which also helps moving abroad. Mar 05 06:04:26 Vancouver, BC is pretty much the place for high-tech on the west coast. Perhaps a little in Calgary Alberta (other side of the mountains). Mar 05 06:05:26 Both places are beautiful (or within driving distance of beautiful, I should say!) Mar 05 06:06:07 mwester: yes, that's one of the things that attracts us both. Mar 05 06:09:00 Crofton: did you feel the FOSDEM stuff was worth the trip, going through the long journey? Mar 05 06:17:13 about how big is the snapshot bitbake needs? (packed) Mar 05 06:19:06 * Blue89 is on pentoo, wandering if the snapshot will fit on his 256mb usb flash along with the bitbake module Mar 05 06:26:23 hmm Mar 05 06:32:52 good point, I cannot remember, but I guess it was 200 MB. Can someone check now? Mar 05 06:33:36 I have some stuff saved already, if it is that much it won't fit Mar 05 06:33:44 :-( Mar 05 06:36:31 Wait, I can login and check on another box probablty. Mar 05 06:36:57 I could remaster the cd... I might get another 200mb or so out of that I think Mar 05 06:38:16 wget says it is 105mb Mar 05 06:39:30 yup, confirmed Mar 05 06:39:55 Blue89: do you need it uncompressed? Mar 05 06:40:20 I can probably build a script to decompress it Mar 05 06:40:31 Blue89: Do you need the whole history? Mar 05 06:40:39 ? Mar 05 06:40:55 Blue89: I mean, would a checkout work for your purpose? Mar 05 06:41:08 I just want what is 100% necessary to build ipks Mar 05 06:42:23 actually pentoo mounts any squashfs modul at boot, I could re-compress with that and compress my settings too Mar 05 06:42:58 Blue89: Then you do not need the .db file, you just need a checkout of the latest release. Mar 05 06:43:16 I already emerged bitbake. you mean that? Mar 05 06:43:19 s/release/state of the repo Mar 05 06:43:35 Blue89: No, I mean the checkout "oeg.openembedded.dev" directory. Mar 05 06:43:44 s/oeg/org Mar 05 06:43:59 you do not need the Monotone .db file Mar 05 06:45:14 Ah, the dir is also 110 MB, never mind. Mar 05 06:45:37 I should be getting some sleep instead, good nite all. Mar 05 06:46:04 guys for a machine to support usbhost and usbgadget it just needs those two added to MACHINE_FEATURES right? Mar 05 06:46:08 or is there something else i'm missing Mar 05 06:48:02 I need sleep too, I will figure this out tomorrow. Mar 05 07:00:29 morning all Mar 05 07:04:22 good morning all Mar 05 07:04:52 morning koen Mar 05 07:05:15 i wish it was still morning here :) I've spent the better part of a day on OE Mar 05 07:16:14 koen : seems that the gcc/efika problem is somewhat connected with libiberty Mar 05 07:17:50 koen : hmmm this looks strange -> powerpc-angstrom-linux-gcc -mcpu=603e -O2 -g -Os -DIN_GCC Mar 05 07:29:51 good morning Mar 05 07:30:37 good morning Mar 05 07:33:42 the grub package under org.openembedded.dev doesn't actually install grub stages to /boot does it? Mar 05 07:34:41 hmm just checked and it doesn't Mar 05 07:34:51 i suppose not many people use grub for booting oe devices :( Mar 05 07:38:44 jselect : no its does not. You will need something similar to syslinux.bbclass and no one has wrote one Mar 05 07:38:46 hey rwhitby Mar 05 07:40:39 jselect : to see how to install grub to an OE generated image -> http://owmnr.digital-opsis.com/project/wiki/OWMNRInstall Mar 05 07:44:48 and a message i suppose you don't see often when compiling with OE -> CPU0: Temperature above threshold Mar 05 07:44:48 CPU0: Running in modulated clock mode Mar 05 07:45:22 Ifaistos: thanks, looking now Mar 05 07:45:26 using 18threads was a bit too much probably ;) Mar 05 07:50:34 OE Build Configuration: Mar 05 07:50:34 BB_VERSION = "1.9.0" Mar 05 07:50:36 sweet Mar 05 07:50:42 bring on the GUI!! Mar 05 07:53:32 * pH5 svn up Mar 05 08:00:03 RP: http://en.pastebin.ca/382057 to remove a SyntaxWarning? Mar 05 09:00:55 morning Mar 05 09:02:00 hey XorA Mar 05 09:02:30 Zero_Chaos: pinged you on your bug **** BEGIN LOGGING AT Mon Mar 05 09:21:25 2007 Mar 05 09:28:13 i think i found an issue with insane.bbclass Mar 05 09:28:25 morning alll Mar 05 09:28:28 -l Mar 05 09:29:48 since the Q&A on packaging is done after the packaging stage is done, if an error is found, and try to build the package again it passes Q&A Mar 05 09:30:53 probably it will be a good idea to clear the stab of the stage that produced the Q&A error so it gets re-run next time Mar 05 09:32:32 The way it is now the only way to see if the package passes Q&A after any changes to the FILES stuff is to clear it and start all over Mar 05 09:34:05 have a look at this -> http://rafb.net/p/5Dc2K665.html Mar 05 09:41:08 Ifaistos: the packages will be empty in this case?! Mar 05 09:41:20 Ifaistos: someone cleans the dir Mar 05 09:42:24 Ifaistos: and package QA is part of the do_package task, it just succeeds, I didn't find the spot where it cleaned the dirs Mar 05 09:43:07 zecke: it makes no sense to fail the first time and pass the second Mar 05 09:43:25 Ifaistos: I agree on that, but it checks the second time Mar 05 09:43:36 Ifaistos: but there are no files inside the dirs, so there is nothing to check Mar 05 09:44:03 Ifaistos: or this func is not ran in any case someone needs to look into it Mar 05 09:45:25 zecke: I am not sure why either, haven;t looked into insane Mar 05 09:45:55 Ifaistos: there is PACKAGEFUNCS in package.bbclass Mar 05 09:46:05 Ifaistos: insane adds itself as PACKAGEFUNCS and is ran Mar 05 09:46:19 Ifaistos: could you find out if this insane PACKAGEFUNC is ran on the second time at all? Mar 05 09:46:25 (just add a print) Mar 05 09:46:52 morning Mar 05 09:47:03 zecke: ok. will do Mar 05 09:50:16 hey hrw Mar 05 09:51:18 thanks Mar 05 10:07:20 03ifaistos 07org.oe.dev * r2260d3ac... 10/ (1 packages/asterisk/asterisk_1.2.14.bb): Mar 05 10:07:20 packages/asterisk/asterisk_1.2.14.bb: Several fixes Mar 05 10:07:20 Q&A issues move .debug out of asterisk packages Mar 05 10:07:20 Add asterisk app to 'telephony' section instead of Base Mar 05 10:07:20 Set correct working dir Mar 05 10:07:20 Make sure that all asterisk headers get copied to staging/include Mar 05 10:09:30 03ifaistos 07org.oe.dev * re766ff2f... 10/ (1 packages/libesmtp/libesmtp_1.0.4.bb): packages/libesmtp/libesmtp_1.0.4.bb: Fix Q&A errors Mar 05 10:12:57 03ifaistos 07org.oe.dev * raf4d56b8... 10/ (1 packages/asterisk/asterisk_1.2.14.bb): packages/asterisk/asterisk_1.2.14.bb: Add correct section console/telephony instead of telephony Mar 05 10:28:04 morning Mar 05 10:30:02 rwhitby: the usual -static and -static-libgcc on $LDFLAGS? Mar 05 10:31:27 * rwhitby searches existing recipes for an example Mar 05 10:31:44 good morning Mar 05 10:32:19 morning florian Mar 05 10:34:11 zecke: is there an easy way to pass that to a configure script and make it stick? Mar 05 10:34:39 rwhitby: just try LDFLAGS += " -static" in the OE recipe and see what happens Mar 05 10:34:48 rwhitby: but static binaries with glibc are an illusion Mar 05 10:35:18 rwhitby: depending on your config it will dlopen nis and pam modules which might not work Mar 05 10:35:36 zecke: only need to remove the dependency on libusb Mar 05 10:35:51 (at LaF0rge's request) Mar 05 10:36:39 rwhitby: try -static and check how that works Mar 05 10:39:26 seems to work. thx! Mar 05 10:40:31 zecke: well its executed both times. the question now is on what ipk files.... Mar 05 10:41:53 Ifaistos: so it is empty the second time? Mar 05 10:42:00 Ifaistos: print the files it looks at? Mar 05 10:43:07 zecke: ok Mar 05 10:44:08 zecke: insane does report the files it checks Mar 05 10:44:19 NOTE: Checking Package: asterisk-dbg Mar 05 10:44:32 zecke: both times Mar 05 10:44:54 Ifaistos: interesting Mar 05 10:44:57 zecke: the question is if the actually exists and what they contain Mar 05 10:48:20 zecke: i am doing a clean build of the package so we can check at what state it is after the Q&A error Mar 05 11:00:58 koen: just ordered the 30D at last Mar 05 11:03:47 ade|desk: nice Mar 05 11:07:01 koen: now all i need to do is nick your lenses ;) Mar 05 11:14:20 On another kind of Q&A issues. Would it be possible for someone to start clearing gcc/glibc from all those arm specific patches that get applied regardless of which is the target ? Mar 05 11:15:13 If its not a generic patch, that effects all targets it should be added only for that target Mar 05 11:16:55 does anyone have au1200-dev-kit_v1.0-r3.tar.gz ? Mar 05 11:23:52 nvm, i found it. Mar 05 11:48:27 Registered translator 'lintog729' from format slin to g729, cost 67944 Mar 05 11:48:45 or you will need 68sec to encode 1 sec of speech Mar 05 11:49:27 oups sorry this was not intended for this channel Mar 05 12:23:00 hi Mar 05 12:58:52 ~curse x11 Mar 05 12:58:55 May you be reincarnated as a Windows XP administrator, x11 ! Mar 05 13:00:40 ~curse linux Mar 05 13:00:42 May the fleas of a thousand camels infest your most sensitive regions, linux ! Mar 05 13:04:57 mh. file-native-4.20.0 doesn't even build? that's interesting Mar 05 13:04:58 03nail 07org.oe.dev * r2538463f... 10/ (4 files in 3 dirs): Portmap: Update to 5-26, previous version unfetchable Mar 05 13:07:01 damn, monotone.openembedded.org is slooooooooooow Mar 05 13:09:30 morning all Mar 05 13:10:01 pH5: Looks ok to me. Just a question of which versions to patch :) Mar 05 13:10:41 LaF0rge: if you run it a second time it builds....... Mar 05 13:10:56 LaF0rge: at which point I lost interest in fixing it, sorry Mar 05 13:10:57 koen: | softmagic.c:1529: error: `REG_STARTEND' undeclared (first use in this function) Mar 05 13:11:04 koen: that one? Mar 05 13:11:13 don't remember, sorry Mar 05 13:11:21 sometimes I wonder does my kernel config is fscked or software which I use is shit. too often too many hangups in gui Mar 05 13:11:23 koen: seems like this definition is part of libboost-devel which is not necessarily installed on the host Mar 05 13:11:43 koen: no, problem persists Mar 05 13:12:24 pH5: How did you trigger that warning? python 2.5? Mar 05 13:13:02 koen: I'll investigate Mar 05 13:14:55 hi all Mar 05 13:15:10 LaF0rge: file-native builds on openSuSe where I checked it in, Ill bet is a bloody difference in glibc Mar 05 13:15:41 RP: it was triggered by distutils when installing bitbake with the setup.py script. (python 2.4.4) Mar 05 13:16:59 xora: glibc-2.3.2 doesn't have that define in /usr/include/regex.h Mar 05 13:17:06 xora: glibc-2.3.6 does Mar 05 13:17:08 LaF0rge: iirc XorA added file 4.20 Mar 05 13:17:15 03nail 07org.oe.dev * r8fabbb7a... 10/ (5 files in 3 dirs): hdparm: Upstream upgrade to 6.9 Mar 05 13:17:37 * XorA speeds koen|lab up a bit Mar 05 13:17:38 * LaF0rge adds 4.18 to preferred versions Mar 05 13:17:52 LaF0rge: problem was 4.18 is unfetchable Mar 05 13:18:06 xora: we have PREMIRRORS to downloads.openmoko.org Mar 05 13:18:11 NAiL: did you verify that it works? Policy was to go to archive.debian.org and to mirror that package! Mar 05 13:18:18 s/we/at least I/ Mar 05 13:18:51 zecke: Yes Mar 05 13:19:18 what distro is using something as ancient as glibc 2.3.2 Mar 05 13:19:24 xora: debian sarge Mar 05 13:19:26 XorA: debian stable? Mar 05 13:19:38 xora: pretty large deployment base, I'd say Mar 05 13:19:43 debian we should have burried this years ago :-) Mar 05 13:20:23 XorA: I bet Wolfson offers free upgrades to fedora or openSUSE? Mar 05 13:20:30 xora: show me any other distribution with binary packages which I can use the same revision on that many archs Mar 05 13:21:31 xora: I'm not a fan ofd ebian, but there's nothing that rusn on sparc/sparc64 ppc/ppc64, x86/x86_64, ... - and i want to have nfs-mounted home on all that boxes, whcih forces the same version of programs being installed on all of them. Mar 05 13:22:52 besides that we should stop that file-native race. Just mirror one version and stay with it Mar 05 13:24:05 zecke: anyway, it's always possible to set the default preference to -1 on the new version. Mar 05 13:25:01 NAiL: you kept the old version around? Mar 05 13:25:06 yes Mar 05 13:25:20 the old file tarball? it's all on downloads.openmoko.org/sources Mar 05 13:25:39 that directory gets never cleaned up and will just accumulate everything that our buildhost[s] fetch Mar 05 13:25:42 I don't delete old stuff until I'm pretty sure other people aren't using it, or there's a *lot* of old packages. Mar 05 13:25:47 also, we're planning to have round-robin dns mirrors for that. Mar 05 13:27:23 NAiL: if it is unfetchable, find the sourcecode, place it on one of our mirrors and change the SRC_URI Mar 05 13:27:31 time for glibc-native in OE :-) Mar 05 13:27:46 NAiL: alternatively upgrade, verify that it works and remove the old version Mar 05 13:27:55 NAiL: mixing of these two is not too sane Mar 05 13:29:55 zecke: how do I place stuff on your mirrors? Mar 05 13:30:28 NAiL: ask for an account, or point to the debian archive Mar 05 13:41:17 hi all, musicpd has fixed mikmod support in svn. so there's no need for the fix-mod-support.patch in mpd_svn.bb any longer Mar 05 13:41:36 DoKu: then remove the patch? Mar 05 13:41:57 or add maxdate to it? Mar 05 13:42:21 DoKu: alternatively add maxdate, but back to your FLUTE server? Mar 05 13:46:00 zecke: it's mirrored on the angstrom mirrors Mar 05 13:46:14 (as is everything in needed for console and x11 image) Mar 05 13:51:51 koen|lab: I found out that RRECOMMEND'ing sub-tasks seems to build them Mar 05 13:52:10 oxo: that's expected :) Mar 05 13:52:10 I don't know if it's a feature or bug, but it strikes my intuition :-) Mar 05 13:52:33 I assumed that (R)DEPEND'ing them would suffice Mar 05 13:52:38 koen|lab: so it should be enough to point to the angstrom mirror then? Mar 05 13:54:22 zecke: Have you read the forwarded omoko mail to oe-devel? Mar 05 13:54:46 RP: no Mar 05 13:55:21 hmm mail.openmoko.org is not the right spot for me Mar 05 13:57:05 zecke: They're hiring python developers to hack on bitbake? Mar 05 13:57:22 RP: iirc some dude called 'henryk' Mar 05 13:57:30 RP: hired, so my answer is whatever :) Mar 05 14:01:12 RP: there are a number of controversial wishes on that list, and I'm curious what other OE (core) devs think about it Mar 05 14:01:42 I'm slightly frustrated that someone is going to do all this work without even talking to the bitbake developers Mar 05 14:01:44 koen|lab: wait and let us look at the patches Mar 05 14:02:13 zecke: I have no problem with the wishes, I just know they are controversial Mar 05 14:02:21 e.g. cleaning sources and ipkgs Mar 05 14:02:30 koen|lab: They don't look that bad to me Mar 05 14:03:53 03koen 07org.oe.dev * r675ebe61... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom: prefer file 4.18, we have the sources mirrored via angstrom-mirrors.bbclass Mar 05 14:03:55 RP: I like the PR thingie, I wonder how to transport the information from do_fetch to do_package but this is left to the python expert Mar 05 14:03:57 RP: I didn't say bad, I said controversial :) Mar 05 14:04:21 RP: look at the recent outcry when deploy/ipk change behaviour slightly Mar 05 14:04:59 koen|lab: Its more about how you going about doing it, not what you did Mar 05 14:05:44 koen|lab: People like backwards compability and you have to have a good reason to break it. Your change didn't. Mar 05 14:06:21 zecke: I don't think PR is a good place to store revisions, since an svn rev is part of the package version Mar 05 14:06:34 zecke: especially if you start adding patches to a fixed svn rev Mar 05 14:10:07 koen|lab: I think the revision should be stored somewhere (place it in the control file). Mar 05 14:10:50 koen|lab: but as mentioned in the mail it is " very Mar 05 14:10:51 un-free-software-like" Mar 05 14:12:25 * zecke tosses a coin and decides if it is childish to reject a P0 phone Mar 05 14:18:47 03pH5 07org.oe.dev * r2e88c366... 10/ (1 packages/gnome/libgnomeprint_2.8.2.bb): libgnomeprint: QA issue, install all debug files into -dbg package. Mar 05 14:19:54 zecke: 'p0+p1 or p1' mail? Mar 05 14:23:36 XorA: ping Mar 05 14:23:42 Zero_Chaos: yo Mar 05 14:23:54 XorA: are you listed as the maintainer for kismet right now? Mar 05 14:24:00 I seem to recall a different name Mar 05 14:24:17 Zero_Chaos: is me Mar 05 14:25:08 XorA: I didn't want to step on a toe and add that without asking, but if you would like to pass kismet officially to me that is fine. Mar 05 14:25:23 Zero_Chaos: is cool with me, I dont use it much anymore Mar 05 14:25:40 XorA: it's what I do for a living so I will go ahead and take it then Mar 05 14:25:48 XorA: you have commit access by chance? Mar 05 14:26:07 Zero_Chaos: yes, I shall commit your entry Mar 05 14:26:41 XorA: thanks, I edited the bug to add kismet (as if you couldn't) Mar 05 14:26:59 1941 Mar 05 14:27:17 XorA: my name is already there but it's a stub so just replace it which what I've got please Mar 05 14:28:42 Zero_Chaos: just about to push Mar 05 14:29:53 XorA: cool thanks. I'll close the bug Mar 05 14:30:12 damn oe.org is running slow Mar 05 14:31:57 it happens, I closed the bug, I trust you will submit eventually Mar 05 14:37:08 ~curse python packagin Mar 05 14:37:08 Zero_Chaos: I will as soon as monotone server answers Mar 05 14:37:09 g Mar 05 14:37:17 May the fleas of a thousand camels infest your most sensitive regions, python packagin ! Mar 05 14:40:17 03xora 07org.oe.dev * r6058625d... 10/ (1 MAINTAINERS): MAINTAINERS : add stuff to Zero_Chaos, remove stuff from me. Mar 05 14:42:45 rP: reply sent Mar 05 14:44:02 zecke/koen: and yes, it is very un-frss-software-like Mar 05 14:44:24 but then, FOSS doesn't have a strict release schedule with tons of corporate issues behind Mar 05 14:45:05 I found out that we need this feature about three days before I learned that I had to build the master image, about four weeks after mickey didn't have tiem to wokr on those issues anymore. Mar 05 14:45:08 XorA: thanks, running for a plane now. Mar 05 14:45:22 have a nice day all Mar 05 14:45:32 Zero_Chaos: have fun Mar 05 14:45:47 xecke/koen: so what can I do? delay everything once again, or call in somebody who can write python to help. Mar 05 14:46:48 LaF0rge: communicate Mar 05 14:48:09 zecke: well said Mar 05 14:48:53 LaF0rge: failing to communicate because of deadlines will leave you in the same mess as maemo.. Mar 05 14:48:57 LaF0rge: at least make sure he has the budget to integrate his changes upstream, this is the only way to standout. And I doubt you lacked the 10 minutes to talk with richard or me Mar 05 14:48:58 koen: and if it works, PV would also be fine with me, rather than PR Mar 05 14:48:58 zecke: wrt to your p0 phone: The first 36 have been shipped already Mar 05 14:48:59 zecke: if you didn't reply until yesterday, yours is not in the first batch, I'm afraid Mar 05 14:50:03 zecke: yes, as indicated in my last mail. we will have that code written three times, if required Mar 05 14:51:07 zecke: I'm in favor of implementing something and proposing/discussing the patch afer it is proven to work Mar 05 14:51:26 zecke: then go through iterations of review. Mar 05 14:51:38 LaF0rge: that depends on how invasive the patch is Mar 05 14:51:57 LaF0rge: arge changes should be discussed up front in order to not waste time and effort Mar 05 14:52:05 *large Mar 05 14:52:54 it's never a waste to have a working implementation of some feature, even a less than ideal one. Mar 05 14:52:56 robtaylor: were deadlines the problem or the super-secret 770 successor with the communication? Mar 05 14:53:23 LaF0rge: I don't have an opinion on bitbake matters Mar 05 14:53:57 koen|lab: in all commercial uses of open source deadlines are a problem, the hard bit is managing the balance Mar 05 14:54:02 * koen|lab stabs maple Mar 05 14:54:17 koen|lab: stay with matlab pal Mar 05 14:54:23 LaF0rge: it is if the feature is the wrong feature Mar 05 14:54:40 zecke: I need a symbolic solution to a large set of equations :( Mar 05 14:54:56 me misses maple Mar 05 14:55:28 rob: that's how you might see it. nobody has ever indicated that we ask you to merge any of those changes, or that we would want to use them for anything more than to fix a temporary issue until a real solution can be found Mar 05 14:56:21 LaF0rge: well, its your money that's being spent, i'm just suggesting that a small amount of initial communiation can save you costs Mar 05 14:56:46 LaF0rge: i'm sure your aware of the costs of not having patches integrated Mar 05 14:56:52 *you're Mar 05 14:57:08 LaF0rge: When writing a patch you normally know where this fits in the big picture? For a newbie communication is certainly the key to understand the system regardles of time preassure. Anyway it is too late to argue, looking forward for the approach(es) Mar 05 14:57:20 LaF0rge: but for all the bitching we do, we do still love openmoko for being so open :) Mar 05 14:58:03 rob/koen: you can bitch all day long, as much as you please :) Mar 05 14:58:08 koen|lab: to bitch about something else, I'm stracing qtopia and the qpe process is polling like hell (10 files, twice a second) Mar 05 14:58:24 LaF0rge: i'm not actually bitching. Mar 05 14:59:15 koen said that, not me :) Mar 05 14:59:20 heh Mar 05 14:59:36 koen definitly bitches ;) Mar 05 14:59:43 robtaylor: coming to celf? Mar 05 14:59:57 koen|lab: probably.. Mar 05 15:00:35 koen|lab: matlab actually can solve thing the symbolic way Mar 05 15:00:47 there is a maple-kernel stashed in some toolkit Mar 05 15:01:09 I'm starting to love autotools. it is much better then distutils in python Mar 05 15:01:16 and thats scary Mar 05 15:02:22 hrw: yay, someone appreciating autotools. that makes a change :) Mar 05 15:02:41 robtaylor: python distutils is nightmare compared to m4 Mar 05 15:03:03 hrw: yeah, i know what you mean Mar 05 15:03:44 m4 still makes my eyes bleed Mar 05 15:04:11 actually i think m4 is the easy+nice bit Mar 05 15:04:22 its automake that i kinda dont fully like Mar 05 15:04:51 robtaylor: have you read the "Recursive make considered evil" essay? Mar 05 15:04:57 zecke: aye Mar 05 15:05:14 robtaylor: use unsermake (and forget it is from KDE...) Mar 05 15:05:15 zecke: you can do either in modern automake Mar 05 15:05:28 robtaylor: which version is modern? Mar 05 15:05:40 1.9 Mar 05 15:06:07 zecke: please stop imitating JustinP by sending 2 identical mails to oe-devel Mar 05 15:06:24 hmm, actually i think you could do non-recurive am before that, but i wont touch anything before 1.9 Mar 05 15:06:28 koen|lab: ;) I just stop using apple mail Mar 05 15:13:21 03kuhlen 07org.oe.dev * rf03543f1... 10/ (1 packages/musicpd/mpd_svn.bb): mpd svn: set maxdate for mikmod patch since it has been applied officially Mar 05 15:13:32 fscking python Mar 05 15:13:55 DoKu: well done :) Mar 05 15:14:25 DoKu: want to have a cookie? where did you place them anyway? Mar 05 15:14:48 D'oh! Mar 05 15:14:59 pi != Pi in maple Mar 05 15:15:06 hehe Mar 05 15:15:14 it will show the same symbol however Mar 05 15:15:46 Pi is const and pi is variable? Mar 05 15:15:51 yes Mar 05 15:16:27 theta[50] = 0.1 * pi will give 0.1π and theta[50] = 0.1 * Pi will give 0.1π Mar 05 15:16:44 the solver dislikes the former, though Mar 05 15:17:19 LaF0rge: My code can only extract BugZilla (not using xmlrpc), filing bugreports automatically would be trivial as well Mar 05 15:17:36 zecke: the shelf behind me :) Mar 05 15:17:49 LaF0rge: basicly an integrated autobuild, bugtracker, installation system is one of my idle tasks Mar 05 15:18:29 * CM remember setting the precision of Pi to 10.000 decimals in Maple on a turqoise iMac, hanging it totally for 15 minutes. Mar 05 15:18:58 evalf(Pi,10000) Mar 05 15:19:04 anyone got a gcc 4.X.X cross toolchain for arm OABI lieing around? Mar 05 15:19:14 *nods* Mar 05 15:21:17 zecke: did you see that cairo is using the new quartz backend for 1.3.16 on osx? Mar 05 15:22:13 koen|lab: no, so a native gtk+ should work better nowadays? Mar 05 15:22:40 I think so Mar 05 15:23:13 I never bothered to build a native gtk+ Mar 05 15:23:22 the imendio build scripts scared me Mar 05 15:23:46 if I run python/distutils crap in do_compile then where it tries to check for python extensions which it want? Mar 05 15:25:11 hrw: for python stuff you need a special m4 iirc Mar 05 15:25:15 'portmap 2.25 is gone' problem is not known yet?!? bugs.openembedded.org doesn't list anything Mar 05 15:25:43 LaF0rge: we have 5-26 already Mar 05 15:25:55 or other but current Mar 05 15:26:02 hrw: since when? I did a mtn pull about three hours ago Mar 05 15:26:05 05 14:09 < CIA-3> nail org.oe.dev * r2538463f... / (4 files in 3 dirs): Portmap: Update to 5-26, previous version unfetchable Mar 05 15:26:14 2.5h ago Mar 05 15:26:15 [14:04] CIA-3: nail org.oe.dev * r2538463f... / (4 files in 3 dirs): Portmap: Update to 5-26, previous version unfetchable Mar 05 15:26:20 ok, thanks Mar 05 15:26:36 koen|lab: m4 IF autotools are used Mar 05 15:26:43 koen|lab: this crap use distutils Mar 05 15:29:09 Do we still use "bl" on any platform? Mar 05 15:30:06 RP: did the naming for the SD slot change in recent versions of linux-rp-2.6.20? I was trying to boot the latest build from SD, but I got a kernel panic stating that mmcblk0p1 was not there. Mar 05 15:30:24 RP: That worked ~ 2weeks ago tho... Mar 05 15:30:51 florian: 2.4 collie? Mar 05 15:31:25 hmm.. and on 2.4 iPAQ maybe Mar 05 15:31:39 hvontres|poodle: rootdelay=5 ? Mar 05 15:32:06 I just wrote a task for gpe-conf that does the same and uses more or less the same syntax. Mar 05 15:32:21 I think about shipping a small script that just replaces bl. Mar 05 15:32:52 As gpe-conf supports these devices it should be safe... Mar 05 15:34:35 florian: people can use bl without X11 iirc Mar 05 15:43:58 koen|lab: I tried both 3 and 5. The thing is, rootdelay =3 worked on 2.6.20 builds from ~2 weeks ago. I should probably check the defconfig files to see if something got borked threre. Mar 05 15:45:19 hvontres|poodle: the mmc system could be initializing slower Mar 05 15:47:07 koen|lab: hmmm...ok. I'll try some really long root delays when I get home. But the naming convetions have not changed, right? Mar 05 15:47:19 they shouldn't Mar 05 15:49:37 hrw: Do you think we should try to use new image recipies similar to the ones for angstrom for OZ-3.5.5? I made a copy of Commandline-image.bb for testing this weeked, and It seems to build fine. I just need to work out some issues with 2.6.20 not booting right. Mar 05 15:49:40 hrw: right... so if you have an image without GPE you use the old bl. Mar 05 15:50:08 it will just replace the (imho not maintained) old one. Mar 05 15:50:10 hvontres|poodle: try what ever you want Mar 05 15:51:44 hrw: ok. I was thinking that way we would stay kinda current with the improvements in OE. I'll work on it some more this week and add stuff to the bugtracker when it works. Mar 05 15:52:08 hvontres|poodle: if you think that way, you would be using angstrom instead of oz ;) Mar 05 15:53:45 I created hardcore recipe: 154 lines Mar 05 15:54:17 hvontres|poodle: My kernel ssays the naming hasn't changed Mar 05 15:56:23 koen|lab: Well, my thought process was that since this was sort of a bridge release between Oz and Angstrom, this would actully make lots of sense. It would be kinda like Angsrom-oabi with glibc 2.3 and compiled on gcc 3.4.x Mar 05 15:56:50 RP: thanks. I'll play with rootdelay some more then... Mar 05 15:57:21 ~lart kexec for not changing CMDLINE correctly Mar 05 15:57:21 * ibot hurls dozens of incontinent, insomniac, hungry kittens with tiny little razor-sharp claws and a wide variety of contagious intestinal parasites at kexec for not changing CMDLINE correctly Mar 05 15:58:14 wow Mar 05 15:58:23 koen|lab: also, I am planning on using Angstrom in the neear future :) Mar 05 16:00:02 chouimat: why wow? Mar 05 16:00:14 the lart Mar 05 16:00:50 chouimat: yeah, I like that one myself. That and the one about the fleas from camels.... Mar 05 16:01:35 hvontres|poodle: at least the kittens are not horny ;) Mar 05 16:09:51 there's 40 minutes of my life I won't get back Mar 05 16:09:56 * koen|lab stabs maple again Mar 05 16:10:46 koen|lab: I want the last 7 years back :) Mar 05 16:10:59 I can say that next year Mar 05 16:12:24 We would do better if mr euler would be looking over our shoulders Mar 05 16:13:14 and saying things like "cosine, you fool!" Mar 05 16:13:42 * chouimat gets mentally ready for a week of working with really ugly code ... Mar 05 16:41:09 hi * Mar 05 16:42:10 i want to buy one PDA w/ support wifi,keyboard and linux os, have anybody recomment some model ? Mar 05 16:50:09 I'm attempting to create a scew-0.4.0 bitbake recipe, but it's failing during the configure stage with these errors: Mar 05 16:50:13 automake: unknown warning category `cross' Mar 05 16:50:13 doc/Makefile.am:13: DOC does not appear in AM_CONDITIONAL Mar 05 16:50:13 Makefile.am:16: DOC does not appear in AM_CONDITIONAL Mar 05 16:50:13 autoreconf: automake failed with exit status: 1 Mar 05 16:50:15 FATAL: autoreconf execution failed. Mar 05 16:50:51 any ideas? Mar 05 17:12:39 hmmm, looks like I need to understand how automake uses AM_CONDITIONAL... Mar 05 17:17:05 03rpurdie * r787 10/ (2 files in 2 dirs): shell.py: Fix syntax warning reported by pH5 Mar 05 17:56:17 anyone know how to get rid of these warnings from my log files? Mar 05 17:56:18 perl: warning: Setting locale failed. Mar 05 17:56:19 perl: warning: Please check that your locale settings: Mar 05 17:56:19 LANGUAGE = (unset), Mar 05 17:56:19 LC_ALL = (unset), Mar 05 17:56:19 LANG = "en_US.UTF-8" Mar 05 17:56:20 are supported and installed on your system. Mar 05 17:56:22 perl: warning: Falling back to the standard locale ("C"). Mar 05 17:58:32 xuumbi: read this once again and you will know Mar 05 18:00:56 grrrr Mar 05 18:01:00 ~lart yaffs(2) Mar 05 18:01:01 * ibot grabs a large, mis-shapened log, with squirrels, and beats yaffs(2) until only the nuts remain ... which the squirrels run off with Mar 05 18:04:30 has anyone gotten NPTL version of gcc/glibc working? I tried building glibc-intermediate and it says that __thread is required. I added --with-__thread to gcc build but that doesn't help. Mar 05 18:04:34 Crofton|home: ping? Mar 05 18:05:05 yaffs(2)? Mar 05 18:05:47 chouimat: yet another (broken) flash file system Mar 05 18:06:02 koen ok ... Mar 05 18:06:03 tnb: DISTRO=angstrom-2007.1 Mar 05 18:06:55 koen: thanks, I'll look at your config Mar 05 18:07:07 or just use it :) Mar 05 18:07:26 eventually, yes, but I need something quick for fam084 Mar 05 18:08:04 heh Mar 05 18:09:07 koen: yeah, that's one thing about having a commercial release... can't just switch immediately. :-/ Mar 05 18:09:20 (commercial product) Mar 05 18:09:38 ntpl is an ABI break as well.... Mar 05 18:09:48 yeah :-( Mar 05 18:10:14 that's why mickeyl and I started angstrom, to get all such breakage in one go Mar 05 18:10:18 * chouimat wonders if he refuse the contract or hire some student to do it for him ... Mar 05 18:10:29 gcc3 -> gcc4, posixthreads -> ntpl, oabi-eabi, etc Mar 05 18:11:18 yes, that's the next planned break here, but when the largest customer calls... Mar 05 18:11:31 03koen 07org.oe.dev * ra830198d... 10/ (3 files in 3 dirs): linux-gta01: update to moko8 patchset Mar 05 18:13:38 koen: you have a timeframe for angstrom 2007.1 "release"? Mar 05 18:13:55 tnb: not yet Mar 05 18:14:09 tnb: I think now is a good think to get thinking about making a release Mar 05 18:14:20 s/good think/good time/ Mar 05 18:14:38 need to finish this uni term first Mar 05 18:15:03 koen : 'Political Correct' answer should be ala Debian.... "when its ready" Mar 05 18:15:11 :) Mar 05 18:16:04 "Are we there yet Papa Smurf?"..."Five more minutes"... Mar 05 18:17:25 hvontres|poodle: :) Mar 05 18:18:55 koen: but to your knowledge nptl doesn't require gcc 4.x? Mar 05 18:20:09 tnb: it shouldn't Mar 05 18:20:16 but I never tested it Mar 05 18:20:45 koen: did you package glibc 2.4 yet? Mar 05 18:21:10 zecke: 2.4 is broken for arm and probably more archs Mar 05 18:22:08 * LaF0rge tries to find out why OM with current OE tries to build gconf-osso Mar 05 18:22:34 probably a missing PREFERRED_PROVIDER Mar 05 18:23:06 PREFERRED_PROVIDER_gconf = "gconf-dbus" Mar 05 18:23:31 good point Mar 05 18:26:08 LaF0rge: next commit is for you Mar 05 18:26:18 hrw: thanks :) Mar 05 18:26:43 * LaF0rge remains excited about upcoming commit message :) Mar 05 18:26:57 03koen 07org.oe.dev * r0303d429... 10/ (9 files in 3 dirs): gconf-osso: drop it, part of the maemo clean up Mar 05 18:27:20 LaF0rge: do you just want automatic bug reporting on build issues, or something more powerful (e.g. keeping track of runtime 'results')? Mar 05 18:27:20 hrw: just being curious: what is 'the maemo cleanup' Mar 05 18:27:47 zecke: what are runtime 'results'? Mar 05 18:27:53 LaF0rge: OE contain maemo from IT2005 which is atleast 2-4 releases old Mar 05 18:28:05 hrw: ok, i see Mar 05 18:28:25 LaF0rge: and it is scheduled for removal but as I have too much other things to do I do not removed it yet Mar 05 18:28:44 LaF0rge: using devintrigator to flash rootfs, boot, use exmap to keep track of base memory footprint, run dogtail tests on the gui Mar 05 18:29:33 zecke: no, that's out of the question for now Mar 05 18:29:54 zecke maybe at some point way further down the road, but we'll think about that later. Mar 05 18:29:59 for now it's really about build issues Mar 05 18:30:16 we want to be the first to learn if something doesn't build, to improve the 'build experience' of external developers ;) Mar 05 18:30:18 03hrw 07org.oe.dev * re0d98822... 10/ (10 files in 4 dirs): maemo: moved gconf-osso, osso-gnome-vfs2 to obsolete/maemo/ Mar 05 18:30:46 LaF0rge: you mean 'the first to act', right? Mar 05 18:31:13 btw - gps part is eabi already? Mar 05 18:31:23 LaF0rge: how do you currently view the results of an autobuild? Mar 05 18:32:05 could someone please explain what is the difference between the different ARM abis ? Mar 05 18:32:39 Ifaistos: 90% marketing, 10% clean up of a historical mess Mar 05 18:32:52 Ifaistos: alignment, syscall convention Mar 05 18:32:52 Ifaistos: see http://wiki.debian.org/ArmEabiPort Mar 05 18:33:27 koen: 'the first to learn about and act accordingly' Mar 05 18:33:46 hrw: don't ask me about GPS. the current nightmare, since yesterday. Mar 05 18:34:04 LaF0rge: just to know your requirement. If a build fails you want to have a bug report? That simple? Mar 05 18:35:22 zecke: yes, that would be the current demand. plus something that automatically schedules rebuilds, both incremental (only what dependency tracking thinks should be rebuilt, for new packages) as well as full rebuilds from scratch Mar 05 18:35:54 LaF0rge: Ah, so an autobuilder+reporting Mar 05 18:35:58 zecke: obviously that could be enhanced to also contain information about the latest successful build for each package or whatever - but that's not really the requirement, rather some gimmick that we can do later Mar 05 18:36:02 zecke: yes Mar 05 18:36:23 RP: ping Mar 05 18:36:29 LaF0rge: as simple as firing up bitbake five times, cleaning and rebuilding from scratch, and after a while throw sourcecode away as well? Mar 05 18:37:36 zecke: yes. basically there could be two different setups. one that only does incremential builds (fast, soon after new .bb file has been committed, that job could run every couple of minutes), and one that is more long-term and does something like one complete rebuild per day Mar 05 18:37:51 03ifaistos 07org.oe.dev * r413b4397... 10/ (1 packages/openssh/openssh_4.3p2.bb): packages/openssh/openssh_4.3p2.bb: Fix Q&A errors. Remove debug files from openssh-misc Mar 05 18:38:44 EABI openmoko images, freshly built: http://www.angstrom-distribution.org/unstable/images/20070305/fic-gta01/ Mar 05 18:39:35 LaF0rge: build results should be send to something like http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=poky (well we currently have no autobuilds as my server equipment is gone)? Mar 05 18:39:39 hrw: regarding GPS, I don't know what I can state publically, sorry. but since we have one really bad news item every day, this was the last one Mar 05 18:39:56 LaF0rge: or is a binary build okay? Good/Bad? Mar 05 18:40:33 ok Mar 05 18:41:05 zecke: the bugzilla reports should contain stdout/stderr of the failing build. for some website a good/bad thing seems sufficient Mar 05 18:42:57 LaF0rge: you might want something like a/the tinderbox (as dashboard) + little change to tinderclient.bbclass to automatically bugreports + script to start builds Mar 05 18:43:49 LaF0rge: failures in openmoko* sent to bugzilla.openmoko and other failures to bugs.openembedded? Mar 05 18:44:15 koen: sounds reasonable Mar 05 18:44:35 LaF0rge: I just thought of that :) Mar 05 18:44:36 LaF0rge: I will stop paid work in about 30 minutes and can look into an interim solution for reporting bugs automatically Mar 05 18:45:05 LaF0rge: if your bugmaster could create a component for autobuilds, I'm not forced to spam your bugzilla Mar 05 18:45:30 zecke: bugmaster/.../ : in the end about anything comes down to me ;) Mar 05 18:46:02 I know Mar 05 18:46:06 when's mickeyl back from holidays? Mar 05 18:46:14 zecke: we already have "OE bitbake recipes / build system" Mar 05 18:46:27 but then, it might be something else that causes the problem, so a different product might be better Mar 05 18:47:11 LaF0rge: Out of curiosity, are you planning to do package upgrades only over USB or also standalone? Mar 05 18:47:36 zecke: there's now autobuilds Mar 05 18:47:45 hvontres|poodle: Hello .-) Mar 05 18:48:00 hvontres: yes, obviously there will be ipk feeds Mar 05 18:48:06 hvontres|poodle: uahhh ... I lost one eye Mar 05 18:48:17 hvontres: even for phase0 we should have one Mar 05 18:48:43 LaF0rge: and to utilize my sith power: You want to commit to the grand unified testing plot Mar 05 18:48:50 LaF0rge: If you are using USB, then maybe the solution to the ipkg problems would be to come up with a way for the host to "pre-process" the feed information so the phone doesn't have to do the heavy lifting Mar 05 18:48:51 hvontres: but still, since BT and GPRS support are not really up to it, those ipk's will most likely come via usbnet ;) Mar 05 18:49:12 pgfeller: ouch.. Mar 05 18:49:13 hvontres: no, the device should be able to do this via BT/GRPS in the end Mar 05 18:49:32 koen: i think it's just one week, so he should be back next weekend Mar 05 18:49:59 hvontres|poodle: :-) ... it is back ... - just a little update for you: if you plan to build fbvncserver . use a CVS date prior to this weekend (as I've started to do experiments with keymaps) Mar 05 18:50:14 LaF0rge: ok. Just a thought. Mar 05 18:51:15 LaF0rge: I'm keeping my fingers crossed to get the neo in time to show it at bossaconference Mar 05 18:51:30 pgfeller: Ok. Actually, I have a build from ~2 weeks ago that I haven't had a chance to test yet :( I was playing around with OZ 3.5.5 this weekend. Maybe tonight I can try to install it. Mar 05 18:51:33 LaF0rge: with some luck bmidgley can test bluetooth headset support Mar 05 18:52:20 hvontres|poodle: np ... I have plenty of things on my list I'll play with meanwhile (the above mentioned support for keymaps ...) Mar 05 18:52:48 koen: yeah, that's nice but one of those less important issues. Also, bluez are currently reorganizing their whole architecture, so I'm not really sure where we will go Mar 05 18:55:34 LaF0rge: lets hope the bluez dudes come up with something nice Mar 05 18:57:43 03ifaistos 07org.oe.dev * r6063cf44... 10/ (1 packages/modutils/modutils-collateral.bb): Mar 05 18:57:43 packages/modutils/modutils-collateral.bb: Modprobe has changed between 2.4 and 2.6 kernels. Mar 05 18:57:43 The information in /etc/modutils and /etc/modules.conf is not read under 2.6. Mar 05 18:57:43 2.6 uses a /etc/modprobe.d directory, and it does not need a Mar 05 18:57:43 "/etc/modprobe.conf" file; it reads all the appropriate *.conf files and Mar 05 18:57:43 compiles the information at boot time. Mar 05 18:57:47 This commit closes bug #808 Mar 05 19:00:05 cu Mar 05 19:00:13 hrw: later :) Mar 05 19:01:26 koen : how do you create the feed initially in the webserver ? there are a lot dirs, locales for example that need to be created for the feed-scripts to work. Do you have a init-script of some kind or by hand ? Mar 05 19:04:25 Ifaistos: look into contrib/ dir Mar 05 19:04:30 Ifaistos: in .dev Mar 05 19:07:46 hrw|gone : do this scripts create the dirs if you run them to the webserver ? Mar 05 19:28:40 Ifaistos: I create the dirs by hand (base/ debug/ perl/ etc/) and one locate dir (locales/en), the rest is done by the index script Mar 05 19:40:33 Ifaistos: nice job on fixing the QA issues Mar 05 19:42:03 03koen 07org.oe.dev * rb6830b0e... 10/ (1 packages/cairo/cairo_git.bb): cairo git: bump PV Mar 05 19:53:30 hi mr_nice, all Mar 05 19:54:27 guess this will please you: we have synced with OE mainline e2dbb52fe39df7ef786b6068f6178f29508dfded n ow Mar 05 19:54:30 now Mar 05 19:55:38 mr_nice: cs3 latch for simpad is working again with kernel26 Mar 05 20:11:38 mrdata: hi Mar 05 20:12:43 mrdata: what exactly do you mean? Mar 05 20:13:48 mrdata: with kernel26 do you mean the hh.org kernel tree or vanilla 2.6.20 one? Mar 05 20:14:31 LaF0rge: synced in what way? Mar 05 20:14:45 LaF0rge: did you drop the recipes present in OE from svn? Mar 05 20:15:48 mr_nice: the kernel from oe with patches from you Mar 05 20:15:58 LaF0rge: nevermind, I found the svnweb link :) Mar 05 20:16:57 mrdata: what do you mean with cs3 latch is working again? Mar 05 20:17:39 mrdata: btw the mq200 driver is from zecke Mar 05 20:17:55 mr_nice: use of /proc/cs3 and /proc/sys/board/ Mar 05 20:18:09 mrdata: cool did you hack it? Mar 05 20:19:05 mr_nice: transfer the old kernel24 sys-board-patch to kernel26 Mar 05 20:19:34 mr_nice: and do little modification Mar 05 20:19:49 mrdata: can you give me the patch? Mar 05 20:20:27 mr_nice: this is one of my problem Mar 05 20:20:49 mrdata: what is the problem about? Mar 05 20:21:07 mr_nice: have never do it before Mar 05 20:21:24 mrdata: creating a patch file? Mar 05 20:22:12 mr_nice: yes, i work a little hardcore, testing Mar 05 20:22:50 mr_nice: but, give me some time Mar 05 20:24:06 mr_nice: and i will try a patch and send it to you Mar 05 20:24:27 mrdata: :) Mar 05 20:24:51 mrdata: you can create a patch by having a copy of the sourcetree without your changes Mar 05 20:26:35 mrdata: name it how you want. then you just need to do something like diff -uNr linux-2.6.20.original linux-2.6.20 > your.patch Mar 05 20:27:23 mrdata the linux-2.6.20 is the dir in which you did the changes Mar 05 20:27:45 mr_nice: okay, the way is clear Mar 05 20:28:29 mrdata: would be very nice to have the cs3 latch stuff because of the mods Mar 05 20:30:07 mrdata: can you write me a mail if you are completed ? I do not know how often I am here when you are here Mar 05 20:30:43 mr_nice: i will use my 1GB mmc again, and yes i write a mail to you Mar 05 20:30:59 mrdata: thx! Mar 05 20:31:33 mrdata: did you found time to test the image a bit? does the simpad wakes up if it was in suspend mode? Mar 05 20:32:18 mr_nice: no, the first thing was to disable the suspend Mar 05 20:33:22 mr_nice: suspend is not complete, the green led will be on Mar 05 20:34:17 mr_nice: resume is not possible (with power switch) and screen is dead Mar 05 20:34:33 mrdata: same on my simpad Mar 05 20:35:05 mr_nice: only the hard-reset works then Mar 05 20:35:25 mrdata: zecke mentioned to add some code to check if it tries to wake up. I will try as soon as I will finde some time for it Mar 05 20:36:06 mrdata: I think this and the bug at the mq200 driver are the most important things to do. Mar 05 20:37:13 mrdata: it is nice to know that there are more people who owns a simpad :). Mar 05 20:37:22 mr_nice: i have a view of apm, apmd, + simpad_pm from kernel24 Mar 05 20:38:01 mrdata: :) Mar 05 20:38:26 mr_nice: i have two (SL4 and SLC), my girlfrind have also a SL4 Mar 05 20:38:55 mrdata: cool - I have a T-Sinus and a SL4 Mar 05 20:39:43 mrdata: they are just very nice devices Mar 05 20:40:01 mrdata: and it would be verry cool to have a complete 2.6 on them Mar 05 20:40:10 mr_nice: yes, with a big screen ;-)) Mar 05 20:40:57 mrdata: yes the size is perfect Mar 05 20:41:08 mr_nice: minimo and firefox was usable Mar 05 20:41:37 mr_nice: with kernel24 and gpe2.7 Mar 05 20:41:37 mrdata: you have the 128MB mod? Mar 05 20:42:11 mr_nice: yes, this was the only way to test firefox and minimo Mar 05 20:42:53 mrdata: I will also do the mod. do you think it is worth the work? Mar 05 20:43:01 hi all ! Mar 05 20:43:45 mr_nice: with the right know how, it is not easy Mar 05 20:44:36 mr_nice: 2 chips, hot air, good microscop Mar 05 20:44:51 mrdata: did you had a desoldering station? Mar 05 20:45:44 mr_nice: no, only hot hair Mar 05 20:45:50 mrdata: I saw them on ebay Mar 05 20:46:02 mr_nice: sorry, air Mar 05 20:46:21 mrdata: with a hair dryer? Mar 05 20:46:39 * CM considers "/nick mr_cm" ;) Mar 05 20:46:47 mr_nice: no Mar 05 20:47:24 mr_nice: 400-500W hot air Mar 05 20:47:49 mrdata: how hot was the air Mar 05 20:48:05 cm: i think mr_cm would be a cool nick ;) Mar 05 20:48:53 mr_nice: i have it at work, 400 degree celsius Mar 05 20:49:34 mr_nice: the problem is, you have short time to do the work Mar 05 20:50:23 mr_nice: too much hot air, and the rest of your board damaged Mar 05 20:50:33 mrdata: I'll better practice on an old mainboard then Mar 05 20:50:35 mr_nice: Hehe.. Mar 05 20:52:12 mr_nice: i had repair handys so called level3 Mar 05 20:53:18 mr_nice: many flash's, ram's and processors Mar 05 20:53:54 mr_nice: and now we build wireless modules Mar 05 20:54:00 mrdata: maybee you can get some chips? Mar 05 20:54:40 mr_nice: no, the production of the chips we need has stop Mar 05 20:55:52 mr_nice: the flash is NOR and the ram is PC100/PC133 SD-RAM Mar 05 20:57:10 mr_nice: the only way for me was to use a PC133 SO-DIMMs for notebook with the right rams Mar 05 20:57:18 mrdata: there is a serial wlan and serial bluetooth chip from philips. Mar 05 20:58:24 mr_nice: i have do the mmc-mod and the bluetooth-mod, the hardware is okay now Mar 05 20:58:34 mrdata: you used memory chips from a normal PC133 SD- Ram module - cool Mar 05 20:59:15 mr_nice: i have read a long time datasheet Mar 05 21:00:33 mrdata: maybee you are the right one for the usb mod ;) http://opensimpad.org/index.php/Possible_Mods Mar 05 21:00:40 mr_nice: found parts from MTD, a module with the right configuration Mar 05 21:01:41 mr_nice: i build over 1 1/2 years on the mods, little time problems Mar 05 21:02:06 mrdata: no problem :) Mar 05 21:03:07 mrdata: I also have time problems - the kernel is more importent Mar 05 21:04:38 mr_nice: is okay, i`will do my best to help you Mar 05 21:04:57 mrdata: thx! Mar 05 21:06:19 mr_nice: first kernel26, screen; second mmc+bluetooth Mar 05 21:06:58 mr_nice: also cf card is usable by me Mar 05 21:07:12 mrdata: on 2.6? Mar 05 21:07:48 mr_nice: yes, you must enable ide support in kernel config Mar 05 21:08:49 mr_data: then you can enable use ide with pcmcia Mar 05 21:09:31 mrdata: I will change the kernel config Mar 05 21:09:55 mr_nice: i have modify the default config for that Mar 05 21:11:01 mrdata: I will write you a mail and give you r/w rights on sf.net slackpad svn Mar 05 21:11:06 mr_nice: the tricky part is: how to use the cf card with fs support Mar 05 21:11:28 mr_nice: ups, thx! Mar 05 21:13:57 mrdata: I think it is faster to do the simpad kernel work there as to do it on oe because we do not have rw on oe Mar 05 21:14:02 mr_nice: i had need fat, vfat and msdos, but the right changes in .bb files is not easy for me Mar 05 21:15:15 mr_nice: thats okay for me, but the building system from oe is great, i think Mar 05 21:15:41 mrdata: yes the buildsystem is great Mar 05 21:17:53 mr_nice: doing the work, then we could send it back to oe Mar 05 21:18:32 mrdata: I will change the .bb file for the modules and send it to you Mar 05 21:19:07 mr_nice: supi, thx! Mar 05 21:21:33 mrdata: no problem Mar 05 21:21:58 mrdata: I hope I will find tomorrow some time for it Mar 05 21:23:39 * mr_nice notes himself to make an english course *g Mar 05 21:24:34 mr_nice: is okay, i have build a extra file hirachy, put the right modules inside, then create a new image Mar 05 21:25:29 mr_nice: since yesterday, its working Mar 05 21:29:19 mrdata: I hope I will find some time to upload a angstrom image, so that more people are able to test it and see that there are some advantages on 2.6 Mar 05 21:29:34 mr_nice: the time is over now, i need some sleep (i'am a little ill) Mar 05 21:29:55 mrdata: second that - see you Mar 05 21:30:03 mr_nice: thats great, bye Mar 05 21:31:11 when using svn for SRC_URI in a recipe, is there a way to tell it to grab the most recent revision or do I have to specify rev each build? Mar 05 21:31:47 the bitbake manual is no help :( Mar 05 21:38:41 sorry took a bit longer to finish this stuff, spare time starts now Mar 05 21:57:08 woglinde: hey Mar 05 21:57:14 he zecke Mar 05 21:57:36 I looked a little bit more on phoneme Mar 05 21:57:43 zecke: bitbake 1.8.0 is a week away ;-) Mar 05 21:57:48 but it suckz documentation is old Mar 05 21:58:47 rwhitby: Did you test that kernel image patch I gave you much? Mar 05 22:00:47 RP: good :) Mar 05 22:05:30 RP: just woke up - will respond in an hour or two (have to catch bus) Mar 05 22:05:43 rwhitby: np Mar 05 22:06:48 03rpurdie 07org.oe.dev * r34846c18... 10/ (1 classes/image.bbclass conf/bitbake.conf): image.bbclass: Create symlink to latest image generated. Also add dependency on populate staging for do_rootfs (both from poky) Mar 05 22:06:52 03rpurdie 07org.oe.dev * ra81f6f87... 10/ (1 conf/bitbake.conf): bitbake.conf: Add DEPLOY_DIR_TOOLS Mar 05 22:28:19 anyone ever use wxwidgets with OE? Mar 05 22:29:25 LaF0rge: two options: Install python2.4 on the autobuilder, make reporting bugs to the autobuild component work without password Mar 05 22:32:29 rwhitby: Did you test that kernel image patch I gave you much? Mar 05 22:32:32 sorry Mar 05 22:33:19 * RP hit up twice return on the wrong window :/ Mar 05 22:45:55 morning CoreDump|home :) Mar 05 22:46:02 hi hvontres|poodle =) Mar 05 22:54:58 LaF0rge: going home now, I have not made too much progress as I have been struggling with cookies, will continue tomorrow Mar 05 22:57:01 zecke: don't worry, the auotmatic buildign is not that urgent. thanks for your support Mar 05 23:07:55 Hi Mar 05 23:10:16 i know this is not a support channel, but anyway I would love to learn how I can make sure our uboot build ends up in fic-gta01 rather than generic arm4t ? Mar 05 23:11:34 hmm Mar 05 23:12:01 I only see u-boot stuff in deploy/images Mar 05 23:12:38 with DISTRO=OpenMoko that is Mar 05 23:13:36 LaFOrge hm I think its architecture bounded so fic-gta01 has to be a part of that Mar 05 23:14:48 argh. sorry ;) I think we don't even yet build uboot ipk's for updating from linux mtd. sorry for that ;) Mar 05 23:14:51 * LaF0rge gets confused Mar 05 23:14:58 ah =) Mar 05 23:15:32 LaF0rge: force PACKAGE_ARCH to fic-gta01 to make then deploy in the fic-gta01 directory Mar 05 23:15:42 thanks Mar 05 23:16:51 nah, I got an email from you today. I gotta thank _you_ heh Mar 05 23:18:38 PACKAGE_ARCH_${PN} = "${MACHINE}" Mar 05 23:21:53 LaF0rge: are OM core devs freezing at that new OE revision, or using a floating OE revision now? Mar 05 23:24:36 CoreDump|home: Getting a new Toy then ? :) Mar 05 23:24:45 hvontres|poodle: you betcha Mar 05 23:24:48 =D Mar 05 23:25:28 * hvontres|poodle imagines CoreDump|home stuck in Minnisota or North Dakota Mar 05 23:25:56 well, not quite. But germany is right around the corner ;) Mar 05 23:27:07 rwhitby: I don't think we neccessarily need to freeze here. Mar 05 23:27:08 CoreDump|home: nevermind..I just couldn't resist making fun of the whole "you betcha" thing, especially after seeing a bunch of ads for "Fargo" a while back Mar 05 23:27:26 rwhitby: we're months away from anyting that I would ever call a stable release anyway Mar 05 23:27:43 rwhitby: but I leave that up to mickey once he's back Mar 05 23:28:02 LaF0rge: what are *you* doing - freezing, or unfrozen? Mar 05 23:28:26 (cause MokoMakefile always follows what the core devs are doing, to make sure bug reports are immediately relevant) Mar 05 23:28:49 hvontres|poodle: I usually do not watch tv ;) Mar 05 23:29:18 CoreDump|home: It would have been lost in translation anyways.. darn dubbing... Mar 05 23:29:49 indeed, I can't stand dubbed movies / episodes Mar 05 23:30:06 * hvontres|poodle remembers watching Magnum in German years ago... And then in English with the Voices reversed Mar 05 23:31:03 ferrari dino Mar 05 23:31:06 =D Mar 05 23:32:42 ...te only memorable thing in magnum :) Mar 05 23:33:47 * CoreDump|home will bbiab Mar 05 23:40:16 hi Mar 05 23:48:39 re zecke Mar 05 23:51:49 time to go home... Mar 06 00:02:01 I don't suppose there are any metacity .bbs are there? Mar 06 00:36:08 re again, now to cookies Mar 06 00:38:37 ljp: Bill King is funny Mar 06 00:45:49 :) Mar 06 00:46:49 at least he responds Mar 06 00:47:08 ljp: so Qtopia can do better file polling than QtopiaCore :) Mar 06 00:48:06 there are problems with kernel type Mar 06 00:49:13 apparently Mar 06 00:49:46 funny thing is, just before your email, there was an internal email about QFileSystemWatcher Mar 06 00:51:08 zecke: what kind of cookies are you referring to? Mar 06 00:55:34 LaF0rge: Cookies set by the Bugzilla, but I found a nice way to keep them Mar 06 00:57:49 zecke: oh, these kinds of cookies Mar 06 00:58:04 zecke: thought you were talking about delicious food items ;) Mar 06 00:59:30 LaF0rge: no, we ran out of physical cookies at the office (the reason to go home) Mar 06 01:00:11 * mwester has a brand new box of oreos in the fridge. Mar 06 01:06:28 LaF0rge: everybody loves quiz shows: For one chocolate cookie. What does seppuku mean? Mar 06 01:08:16 hari-kari Mar 06 01:08:45 (using the gaijin spelling :-) Mar 06 01:09:46 hmm, sending a cookie will be tough... Mar 06 01:12:28 my browser is full of cookies Mar 06 01:17:45 I stuck on prismstumbler - package I even doesn't want :-E Mar 06 01:28:00 guys for usb keyboard support I just need DISTRO_FEATURES and MACHINE_FEATURES usbhost and usbgadget right? Mar 06 01:28:47 Only usbhost IIRC.. usbgadget is when the device is a usb client Mar 06 01:30:19 ok thanks NAbyss, I assume something else is wrong with my new boards because usb and I are just not getting along Mar 06 01:31:04 jselect: Do you have the correct host controller module loaded? Mar 06 01:32:36 NAbyss: I don't have any modules compiled in /lib/modules/`uname -r`/ Mar 06 01:32:52 NAbyss: just trying to find out how to specify which modules I want for linux-x86 Mar 06 01:33:21 heh, such stupid solution of problem :b - "touch prismstumbler-0.7.3-r4.do_qa_configure" Mar 06 01:37:41 NAbyss: any hints? :) I've set the correct features and I'm looking at the linux-x86 recipe and it's includes Mar 06 01:38:00 n8 Mar 06 01:38:41 jselect: in your image.. RRECOMMENDS += "kernel-module-ohci-hcd" or such? Mar 06 01:39:02 and hid maybe Mar 06 01:39:34 NAbyss: I haven't seen any documentation on that yet. Thanks I'll look it up now. Mar 06 01:40:14 jselect: If you install usbutils.. what happens when you do lsusb? That should tell you what's showing up, regardless of HID etc. modules Mar 06 01:49:31 NAbyss: I've got that compiling now Mar 06 01:49:47 NAbyss: i've had a look at the usb tree under a woody install on one of the boards Mar 06 01:50:14 NAbyss: and the devices definately work. I've got to try these kernel module runtime recommends Mar 06 01:50:55 NAbyss: as I'm also having trouble with four out of the 16 serial ports on the device too. Thanks for your help - I'll bug you again once my aging xeon has caught up with the compile list :) Mar 06 01:51:43 for configure files that rely on bashisms, is it better to try and posixize all the bashisms, or just put "#!/bin/bash" at the top of the script? Mar 06 01:52:05 jselect: Good luck :) Mar 06 01:52:48 rwhitby: I've found if you don't provide anything but bash on your system then no-one can break it by using say.. zsh :) Mar 06 01:53:42 * LaF0rge immediately notices somebody mentioning his favourite shell zsh :) Mar 06 01:54:30 NAbyss: Just a quick one.. i'm looking in the kernel bbclass right now but would I be correct in assuming that any module prepended with "kernel-module-" would be compiled in? Mar 06 01:55:02 jselect: Anything with kernel-module- is supplied as a .bb with the .ko module file Mar 06 01:55:09 You'd have to change the defconfig for your machine to compile it in Mar 06 01:58:14 NAbyss: sorry, I can't see any bb files with "kernel" and "module" in their name in the org.openembedded.dev monotone db. Mar 06 01:59:26 NAbyss: I've added RRECOMMENDS for ehci and uhci modules to my image recipe.. is that better placed in the machine conf instead? Mar 06 02:00:35 jselect: D'oh.. I meant .ipk, not .bb Mar 06 02:00:56 jselect: The kernel-X.bb file produces the kernel image ipk, along with hundreds of kernel-module-*.ipk Mar 06 02:01:40 jselect: Place it in the file you've customised.. if you've customised both, it might be more appropriate in the machine bb Mar 06 02:03:37 NAbyss: Ah yes I see how the ipk packaging works. I've customised the distro, machine and a bunch of packages in my own tree Mar 06 02:04:14 NAbyss: I'm going to put it in the image task as I've got a few images and only the debug one should have usb support. I'm trying that now thanks. Mar 06 02:04:49 NAbyss: Is there any documentation on the kernel module ipks produced? Or am I going to have to look at the ipks produced by linux-x86? Mar 06 02:06:03 jselect: Not AFAIK.. what specifically are you looking for in terms of kernel module documentation? Mar 06 02:06:20 Basically, anything with =m in the appropriate defconfig will be produced as a kernel-module-X.ipk file Mar 06 02:06:30 NAbyss: Yeah i figured it'd be autogenerated Mar 06 02:06:40 NAbyss: ahh see that's what I'm missing Mar 06 02:06:56 ~praise multicore CPUs Mar 06 02:06:58 NAbyss: I don't have a 'defconfig' entry... looking for one now Mar 06 02:07:16 jselect: There should be a defconfig in org.openembedded.X/packages/linux/ for your kernel .bb Mar 06 02:09:10 NAbyss: awesome, looking there now. What's the norm for customising defconfig.. I suppose I'll need my own kernel package outside the mtn tree. Mar 06 02:10:02 jselect: I'm not sure what best practices are for that; most machines appear to have their own defconfig for the machine target - take a look at RP's linux .bb file for how that's handled with the various zaurii Mar 06 02:13:16 NAbyss: Ok, they seem to store different defconfig files for each unit ie "defconfig-collie". I'm probably going to do the same thing then for my kernel package I suppose. Thanks :) Mar 06 02:13:25 someone should update OE.mtn Mar 06 02:13:41 s/update/merge/ Mar 06 02:26:06 hmm.. Setting my image recipe RRECOMMENDS to include kernel-module-sis900 did not include the sis900 module ipk (which is present) in my new rootfs. Maybe it does need to be in machine Mar 06 02:36:36 hmm can it be....there's no scm in OE Mar 06 02:37:02 beside mtn ;) Mar 06 02:37:24 well, ignore me Mar 06 02:39:31 without monotone, oe is nothing?! Mar 06 02:40:34 crink: w/o monotone OE would use another SCM.... Mar 06 02:41:56 My vote is for subversion. Mar 06 02:42:02 not that I have a vote. :) Mar 06 02:42:48 ah, silly me grep'ped for "svn", not subversion Mar 06 02:42:55 * CoreDump|home bitbakes subversion Mar 06 02:46:46 au1200 need some change in org.openembedded.dev, but when i changed it, bitbake spit " can't bitbake.conf " Mar 06 02:47:19 s/can't/can't/\ find\ Mar 06 02:49:03 crink: sed: -e expression #1, char 14: unknown option to `s' Mar 06 02:49:05 tch tch Mar 06 02:49:09 :) Mar 06 02:49:57 whatever Mar 06 02:51:43 OK I'm not getting anywhere.. oe refuses to use my rrecommends additions for kernel-module-${module} regardless of where i put it Mar 06 02:51:50 i can't insert the file "travis.conf" into distro directory, should change the name to angstrom-2007.1.conf Mar 06 02:57:34 There is virtually zero documentation on using rrecommends to include kernel modules in rootfs images Mar 06 02:57:45 I'm pretty confident.. i've been looking for a while :) Mar 06 02:58:31 jselect: Have a chat to koen|away when he's around.. he's the one that suggested that approach to me, IIRC Mar 06 02:58:53 What if you put it in the image's RDEPENDS instead of RRECOMMENDS? That'll break if the module is changed to be compiled-in though **** ENDING LOGGING AT Tue Mar 06 02:59:56 2007