**** BEGIN LOGGING AT Mon May 29 02:59:56 2006 May 29 03:31:18 * v8jlene is away: Doing other things... May 29 03:48:13 i missed what was said to me when i went afk :/ May 29 04:49:21 jnc, was probably me talking about gcc 4.1.1 May 29 04:49:33 ah okay. sorry i missed that then May 29 04:49:34 thanks emte May 29 04:50:41 someone was using it, complaining that only 4.1.1 was working for them May 29 04:53:35 ah, the sh4 guys May 29 04:55:55 and it seems TheCAn was trying to use it as well May 29 05:15:42 does anybody know how to solve following problem ? May 29 05:15:50 NOTE: multiple providers are available (handhelds-sa, linux-openzaurus, gumstix, linux-efika, linux-amsdelta-2.6, poodle255-kernel-2.4-embedix, opensimpad, linux-h6300-omap1, linux-omap1, corgi-kernel-2.4-embedix, linux-geodegx, poodle-kernel-2.4-embedix, shepherd-kernel-2.4-embedix, linux-h1940, ipod, linux-omap-2.6, mnci-ramses, handhelds-pxa-2.6, husky-kernel-2.4-embedix, linux-geodelx, linux-xxs1500, LAB-kernel, handhelds-sa-2.6, handhelds-pxa, linux-bast, op May 29 05:17:08 i have a SL-C6000L. i wrote on build/local.conf. MACHINE = "tosa" TARGET_OS = "linux" DISTRO = "openzaurus" DISTRO_VERSION = "3.5.4" May 29 05:19:15 use the correct branch? May 29 05:19:27 vanille.de :) May 29 05:19:49 .oz354 May 29 05:20:05 monotone --db=/stuff/OE.db pull monotone.vanille.de org.openembedded.dev May 29 05:20:10 i think is it's name May 29 05:20:22 yes, i suspected as much May 29 05:20:52 you need the branch with the actual oz version info May 29 05:21:49 :) i don't know what u mean. change from org.openembedded.dev to org.openembedded.oz354 ? May 29 05:22:24 yes May 29 05:22:30 or atleat both May 29 05:22:43 atleat ? May 29 05:22:52 atleast May 29 05:23:42 one is database name:) what is the second ? DISTRO_VERSION ? May 29 05:24:17 distro version is just that May 29 05:24:25 the version of the distro you want to build May 29 05:24:37 it looks in conf/distro for it's metafile May 29 05:24:48 which you dont have because your not using the right branch May 29 05:24:59 :) May 29 05:25:03 yeah May 29 05:26:34 does i have to change "directory name/path" May 29 05:27:03 no May 29 05:27:16 you can just change your extension i belive May 29 05:27:16 for example , BBFILES May 29 05:27:25 in your monotone pull May 29 05:28:27 monotone --db=/stuff/OE.db pull monotone.vanille.de org.openembedded.oz345 May 29 05:28:34 ok :) May 29 05:28:42 i execute that now! May 29 05:29:07 you should ask someone that uses oz what the right name is tho May 29 05:29:15 i am guessing May 29 05:29:29 if it works i lucked out May 29 05:29:29 tho ? May 29 05:29:42 though May 29 05:30:23 i am not know the jargon,well :) sorry May 29 05:30:58 just shorthand slang May 29 05:31:40 slang :) ic May 29 05:32:46 http://oe.handhelds.org/cgi-bin/moin.cgi/DevelopmentBranches <- can i solve this problem using this URL ? May 29 05:36:23 yup May 29 05:36:39 you want oz345x May 29 05:36:47 :) yeah May 29 05:38:04 thank u for ur answer :) May 29 06:44:12 * v8jlene is back (gone 03:12:55) May 29 07:06:37 oï May 29 07:49:48 morning May 29 07:51:20 hey hrw|work May 29 07:59:09 hrw|work: google analytics is already working :) May 29 07:59:14 cool May 29 07:59:34 morning May 29 08:00:00 hey XorA May 29 08:01:47 koen: do you know who else want/should get access to stats? May 29 08:03:56 hrw|work: mickeyl or the person who's going to volunteer for site maintenance :) May 29 08:05:39 koen: for site maintenance it would be best to sign for separate account at G-A. I prefer to not create second admin user on my account as there is no way to create admin for separate projects. May 29 08:06:10 true May 29 08:07:58 ~change 800 bdn to eur May 29 08:08:02 800.00 Dollar (ZWD) makes 0.00630656 Euro (EUR) (from http://www.xe.com/) May 29 08:08:11 hmmm May 29 08:08:52 ~change 800 bbd to eur May 29 08:08:56 800.00 Dollar (BBD) makes 315.345 Euro (EUR) (from http://www.xe.com/) May 29 08:09:10 that makes more sense May 29 08:20:53 * XorA clears his pile of mail and finds a 30 GBP cheque from market research company May 29 08:41:21 <__law__> Hi XorA May 29 08:42:07 hey _law_ May 29 08:42:11 ehm __law__ May 29 08:43:29 <__law__> XorA, have you tried to build oz (oz354x) in the last few days? May 29 08:44:24 __law__: no, I never use branch May 29 08:46:09 <__law__> XorA, .dev? May 29 08:46:14 * XorA is currently doing a fresh .dev build May 29 08:48:25 XorA: with Mardy's w100 stuff? May 29 09:05:14 koen: is that in? May 29 09:05:52 03mickeyl 07org.oe.dev * r2a0708b8... 10/packages/bluez/bluez-utils.inc: bluez-utils: apply 0x0,0x0 -> bcps patch for all distributions May 29 09:05:56 03mickeyl 07org.oe.dev * raac02cf2... 10/conf/distro/ (generic-uclibc.conf generic.conf): May 29 09:05:56 generic(-uclibc).conf: add 'reference' distribution configuration May 29 09:05:56 this should serve as a starting point for people adding new models and May 29 09:05:56 new distributions. It should always be buildable, so please use it for May 29 09:05:57 regression tests and keep it up to date! Be sure to comment everything you add. May 29 09:06:53 hey mickeyl|breakfas May 29 09:26:14 hey May 29 09:30:48 hey zecke May 29 09:31:12 moin May 29 09:31:17 http://dominion.kabel.utwente.nl/koen/images/graph0.png May 29 09:31:24 I love the new graph generator May 29 09:32:03 hehe May 29 09:32:13 there is a dot bug? the text is not fitting in the node May 29 09:32:26 nice May 29 09:32:27 or a "fonts suck on linux" bug May 29 09:32:40 koen: I don't know if I wanto to create bitsearch or bitbake -S May 29 09:32:42 that's from the packaged-staging branch, btw May 29 09:33:27 koen: when will you leave again? May 29 09:33:39 20060607 May 29 09:33:45 http://oe.pastebin.com/744432 May 29 09:33:54 oh okay May 29 09:40:11 Eye Cancer May 29 09:46:37 so many bad coders around May 29 09:48:08 http://oe.pastebin.com/744449 <- morons May 29 09:49:51 zecke: maybe win32 has a magic lldiv :-) May 29 09:50:07 zecke: http://dominion.kabel.utwente.nl/koen/pda/files/glibc-no-isystem.patch May 29 09:50:13 zecke: is that correct? May 29 09:51:03 XorA: everytime I review the changes from one version to another, I feel like writing an email: "What is wrong with you? Stop coding!" May 29 09:52:00 #ifdef WIN32 ULONGLONG par #else unsigned long long par; #ifdef WIN32 ULONLONG otherpar #else unsigned long long otherpar; #endif May 29 09:52:23 koen: could be May 29 09:52:35 koen: is only glibc2.4 broken? May 29 09:52:49 koen: and I hope no update_data has been called May 29 09:53:01 oh well, it is okay (regarding CPPFLAGFS) May 29 09:53:23 zecke: thats appalling, shows a total lack of understanding of either C or WIN32 or probably both May 29 09:54:01 XorA: typedef! *morons* :} May 29 09:54:34 there are two GPL FLUTE implementations and both suck May 29 09:54:42 but I lack the time to write my own :} May 29 10:24:03 03koen 07org.oe.packaged-staging * r36ed964e... 10/ (18 files in 4 dirs): Remove unused stuff and poke at the tarballstash May 29 10:24:07 03koen 07org.oe.dev * r094a3ed8... 10/packages/glibc/glibc_2.4.bb: glibc 2.4: hopefully solve the do_stage loop May 29 10:35:49 * XorA must really work out correct fix for pth May 29 10:40:15 if package provide *.pc then 'inherit pkgconfig'. if it contain PN-config then 'inherit binconfig' - what if it provide both? and should PN-config be in -dev or in PN? May 29 10:41:47 -dev May 29 10:44:58 it will stay in PN :( bitbake ignore my FILES_${PN}-dev =+ "${bindir}/cherokee-config" May 29 10:45:16 or += May 29 10:55:39 it doesn't ignore it, the file is already gone into ${PN} May 29 10:56:00 and we do not have -= May 29 10:58:36 03hrw 07org.oe.dev * rb7c245fc... 10/packages/cherokee/cherokee_0.5.3.bb: cherokee: granular packaging, enabled PARALLEL_MAKE May 29 11:04:07 hi schurig May 29 11:04:13 Hi! May 29 11:04:20 ~seen cbrake May 29 11:04:28 cbrake was last seen on IRC in channel #oe, 1d 20h 3m 20s ago, saying: 'hrw: hello'. May 29 11:04:44 schurig: do you know does Agere cards are supported under 2.6 in some way? May 29 11:04:56 hrw|work: not really, but I wrote a patch May 29 11:05:09 hrw|work: We've told about this some weeks ago, did you forget? May 29 11:05:23 schurig: ah.. yes.. forgot a bit May 29 11:05:43 I'm using the wlags 7.18 source code from their web page, I never got 7.22 working May 29 11:06:06 this driver is in a VERY bad shape, from a kernel programmers perspective May 29 11:06:23 several *.h files are a #define, #if #ifdef #ifndef hell, e.g. thousands of them May 29 11:06:47 anyway, I mad some fixup patches, some "transfer the API to kernel 2.6" patches and one bugfix May 29 11:06:54 ok - just wanted to be perfect sure - I had an offer to get wcf54g free to work on driver May 29 11:07:14 however, it works now only for Kernel up to 2.6.15, 2.6.16-rcX changed the API and I haven't done that yet May 29 11:07:30 ( May 29 11:07:32 ;( May 29 11:08:19 but the driver works, and it even works with wpa-supplicant (usnig my driver_hermes for it, but that is even in OE) May 29 11:09:52 hermes2 support wpa1 iirc May 29 11:10:26 hrw|work: yea, I'm using WPA-PSK. And I once used 802.1x EAP-TTLS. Both worked. May 29 11:11:21 schurig: which recipe in OE contain your 2.6 version? May 29 11:12:58 hrw|work: none, because I stopped putting things into OE when OE switched to monotone May 29 11:13:18 the driver_hermes in wpa-supplicant is still from the old BitKeeper times May 29 11:14:59 schurig: does your tree is available somewhere to look into? May 29 11:15:22 no, it's in my company internal svn May 29 11:16:16 I don't like the speed of monotone and don't need it's features, so I use something fast internal. And our customers can get a free CD with the sources, to fulfull the GPL That's my current workflow. May 29 11:17:11 hopefully we will switch to 0.26+ monotone in next 1-2 months. it is totally different in speed May 29 11:19:32 :-) May 29 11:23:17 hi May 29 11:23:39 does somebody know a good example how to write post install scripts for the ipkg packages? May 29 11:23:55 so some comands that will be executed after the installation? May 29 11:24:44 after instalation == on host (during image creation) or on device (during first boot)? May 29 11:26:10 well it should work on image creation and on installation of the package May 29 11:28:22 on instalation: opie-dasher May 29 11:29:02 on creation: update-rc.d.bbclass (also on install/remove) May 29 11:32:00 okay May 29 11:32:03 thank you May 29 11:43:06 how can i insert changed files into openembedded or can give you some files that are written by me? May 29 11:53:28 morning May 29 11:54:01 Cobelius: use bugtracker May 29 11:54:20 'monotone ls unknown' May 29 12:29:49 hey May 29 12:31:37 hey zecke May 29 12:39:33 hi May 29 12:39:50 i can't get around that 'preferred version 2.4 of glibc not available' bug May 29 12:40:02 join the club :) May 29 12:41:37 mickeyl: in need for bitsearch? May 29 12:42:09 zecke: hi. what's that? May 29 12:42:25 hi mickeyl May 29 12:43:09 mickeyl: searching the MetaData May 29 12:43:23 mickeyl: e.g. bitsearch versions ^glibc May 29 12:43:36 zecke: sounds good. is this an imaginative tool or does it exist? :) May 29 12:43:54 mickeyl: both May 29 12:48:46 taking a rest now, have been at the dentist (again) May 29 12:49:53 although the shell seems to detect it May 29 12:50:09 * mickeyl confused May 29 13:14:23 NAbyss: yes, i still kind of maintain it. just update it May 29 13:14:27 updated, even May 29 13:14:33 <_guillermo> how can I compile openzaurus 3.4.5.1 rc4? I set distro to 3.4.5.1 and only rc2 is compiled May 29 13:15:01 _guillermo: you have too old metadata May 29 13:15:17 koen: bitrotting in gpe cvs .bb recipes :-( May 29 13:15:20 <_guillermo> I just pull the oz354x branch May 29 13:15:28 _guillermo: updated too? May 29 13:15:35 anyone know if any distro actually uses libiconv? i had forgotten all about it until somemone just emailed me asking questions. May 29 13:15:52 all non-glibc distros are probably using it May 29 13:16:16 i thought kergoth fixed it to not use the external libiconv May 29 13:16:41 i know it was for uclibc that i added it, but i do not belive it was used in th end May 29 13:17:08 mickeyl: Ah, okies.. do you know if it's known to work at present? I'm not having any luck with it.. pondering chasing it up with the original authors if it's not a known bug for OE May 29 13:17:28 NAbyss: it should work good with the outdated gpsd. it doesn't work with new gpsd May 29 13:17:48 mickeyl: Ah, thanks.. shall chase it up with them then. May 29 13:17:58 It finds the gpsd, just not the gps May 29 13:18:19 NAbyss: ya, that looks like the usual symptom. the qpegps are too lazy to upgrade their gpsd interface to the new API May 29 13:18:47 03koen 07org.oe.documentation * rbd96d83d... 10/usermanual/usermanual.xml: usermanual/usermanual.xml: start merging with embworld-oe.dbk May 29 13:18:52 03mickeyl 07org.oe.dev * r1c17f3bc... 10/packages/qpegps/ (qpegps_0.9.2.3.bb qpegps_0.9.2.3.2.bb qpegps_0.9.2.3.3.bb): qpegps: add 0.9.2.3.3, remove 0.9.2.3 May 29 13:19:37 mickeyl: Yeah.. I'm tempted to just hack on the source, see if I can get the bugger going May 29 13:20:41 mickeyl: This's the bug I filed, just for an FYI.. http://sourceforge.net/tracker/index.php?func=detail&aid=1496428&group_id=55933&atid=478750 May 29 13:21:08 k May 29 13:23:30 mickeyl: http://fl0rian.wordpress.com/2006/05/29/platform-integration-for-mobile-linux-devices/ May 29 13:23:37 03koen 07org.oe.dev * rbed7a828... 10/packages/gnome/libsoup_2.2.7.bb: libsoup: unbreak staging May 29 13:23:41 03koen 07org.oe.dev * r9c65cea9... 10/packages/libeventdb/libeventdb_cvs.bb: libeventdb: update PV May 29 13:24:27 koen: ah, good to know. May 29 13:25:19 XorA: it should be slightly less bitrotted now May 29 13:25:26 XorA: but feel free to fix up the rest May 29 13:29:02 koen: cheers May 29 13:29:56 koen: I assume all cvs ones should by ${lastest_PN}+cvs? May 29 13:31:04 <_guillermo> hrw|work: what do you mean with updated? I just pulled from monotone.vanille.de May 29 13:31:26 did you update as well? mt pull != mt updaet May 29 13:31:29 update, even May 29 13:31:52 <_guillermo> ok, I gave that for granted, yes pulled and updated May 29 13:32:41 _guillermo: then show us 'monotone head' May 29 13:33:34 XorA: yes May 29 13:33:59 koen: if I see them I shall fix May 29 13:35:09 _guillermo: it should be: 3d1e1f913a72ee6bf1cc84e50aad725485c374b7 coredump@openembedded.org 2006-05-28T18:51:17 May 29 13:35:30 _guillermo: if you have older then you need to pull/update again May 29 13:35:41 <_guillermo> fb1c641b14d25026a94209f0d0733758f2dcf2d8 koen@openembedded.org 2006-05-09T20:20:18 May 29 13:36:01 _guillermo: nearly 3 weeks old?? its ANTIQUE... May 29 13:36:06 not even old... May 29 13:36:34 <_guillermo> ok I'm missing something then, how should I update it? May 29 13:36:57 _guillermo: monotone pull;monotone update May 29 13:37:08 <_guillermo> I did that 100 times already, really May 29 13:38:23 _guillermo: then look at it output and do it 101st time May 29 13:38:48 check the default pull and update arguments, you may have to add the corresponding branch. May 29 13:40:21 <_guillermo> aha May 29 13:41:14 to be safe, pull "org.openembedded.*" May 29 13:41:33 might recheckout the wanted branch instead of updating as well May 29 13:41:41 I prefer "org.openembedded.{dev,oz354x}" in this moment May 29 14:01:29 * v8jlene is away: day done May 29 14:07:25 mickeyl: does my old 2.4 kernel driver help you? May 29 14:07:28 mickeyl: hi, btw :-) May 29 14:07:58 hi May 29 14:08:25 schurig: i can see how you did it, although i would've hoped it would be a bit more generic and seperated from the ramses_keyboard itself :) May 29 14:08:37 mickeyl: hehe :-) May 29 14:08:41 schurig: perhaps you could send me the 2.6.x driver when you have a chance May 29 14:08:43 ? May 29 14:08:50 mickeyl: that one is not more generic May 29 14:08:56 yeah, but 2.6 :) May 29 14:09:12 mickeyl: how would you make it more generic? May 29 14:09:39 mickeyl: I put much things into tables. To set rows/get columns, you must be hardware specific, so here you cannot be generic May 29 14:10:42 schurig: my approach would be just to send the keycodes like normal and have a layer on top and between the inputsubsystem that just translates them to characters May 29 14:10:50 mickeyl: anyway, this driver is nice for some things, but, unlike a normal cell phone, you won't get any visual feedback. So probably you want to add a raw mode to it and implement this at a higher level. May 29 14:11:08 what driver you are talking about? May 29 14:11:20 ya, perhaps rather a userland daemon via uinput May 29 14:11:32 hrw|work: a keyboard driver for a keypad May 29 14:11:43 like writing text on a phone May 29 14:12:08 ah.. May 29 14:12:18 2=ABC 3=DEF 4=GHI etc May 29 14:12:20 mickeyl: one part of my driver translates the x/y position of my keyboard grid into scancodes, the actual scancode->character translation then is the standard kernel one. So my driver emits "pc"-keyboard scancodes, despite working quite differently May 29 14:12:45 mickeyl: but an uinput based daemon cannot provide feedback like a cell phone either May 29 14:13:02 ya May 29 14:13:07 that's the main problem May 29 14:13:35 mickeyl: what I mean: with my driver, when you press 1 1, you won't see anything until a short timeout or you press again. On a cellphone, you would see the "b", but at the cursor position. May 29 14:14:20 schurig: i see. guess i rather try to finish running opie on it ... May 29 14:14:35 pure VT is not that useful on a phone May 29 14:14:53 mickeyl: so the definite driver would be something like I have, to be able to use the Linux console as a last resort, but a raw/half-cooked mode in the driver where then an Qt/Opie keyboard driver or an X11 program can provide the cell phone like feedback. May 29 14:15:17 mickeyl: yes, not very useful, but it was nice at implementation time and *IS* nice as a last resort in case of weird bugs May 29 14:15:35 schurig: indeed. especially because we don't have serial on that phone. May 29 14:15:44 (not without soldering) May 29 14:15:52 mickeyl: how much is such a phone? May 29 14:16:32 schurig: 450EUR new for the european gps model, ~380EUR for the non-gps, 280 on ebay May 29 14:16:53 pretty expensive for a two year old phone, but having linux was just too tempting May 29 14:17:09 hmm, my wife would not agree :-) May 29 14:17:20 hehe May 29 14:17:41 this month is a very expensive month for us :-) May 29 14:19:09 mickeyl: but when it is two years old, then probably it is still one of those phones where the keyboard is still usable for normal fingered people? May 29 14:20:13 schurig: it is, although the keypad is a bit small. then again, you can use the touchscreen to dial as well. http://www.linuxdevices.com/files/misc/motorola_a780.jpg May 29 14:22:50 ah.. thats how keypad is made.. May 29 14:23:42 cbrake_away: still want cherokee without ssl support? May 29 14:28:47 * mickeyl going home May 29 14:28:48 l8er May 29 14:29:20 cu mickeyl May 29 15:14:17 hi May 29 15:14:23 hi May 29 15:18:14 hmmm after setting up OE from scratch I'm getting "../../../../arm-linux/bin/ld: cannot represent machine `arm:collie'" on a .oz354x Collie image. Any ideas? May 29 15:21:51 from glibc-intermediate that is May 29 15:24:03 glibc-intermediate?? May 29 15:24:14 OZ never needed it... May 29 15:24:29 anyway I have to go (gym) May 29 15:24:52 OT: I'm changing theme on my website: http://www.hrw.one.pl/ - critize it if you want May 29 15:25:37 normal glibc failswith the same error May 29 15:27:06 * koen is tempted to move initscript to bitbake May 29 15:32:09 <_guillermo> where can I get a database snapshot of oe? http://ewi546.ewi.utwente.nl/OE/OE.db.bz2 is not working May 29 15:32:30 ~gettingstarted May 29 15:32:32 rumour has it, gettingstarted is http://www.openembedded.org/wiki/GettingStarted May 29 15:34:21 <_guillermo> I was using a different wiki, sorry May 29 15:35:09 <_guillermo> wow, 1.2 M/s, I guess it's hosted in an uni or something May 29 15:40:16 hi guys May 29 15:40:39 hey lardman May 29 15:45:26 bye May 29 15:48:39 hi koen, how's things? May 29 15:49:01 lardman: pretty good May 29 15:49:28 cursing env vars right now May 29 16:10:18 aha May 29 16:10:27 -monitoraspect 16:9 May 29 16:12:42 Does anyone know where Qtopia Desktop calendar files are stored? May 29 16:31:47 hello - I have finished and tested 2 patches for 2.6.16 : May 29 16:32:01 - overclock to 530 MHz via cpufreq May 29 16:32:30 (ie you can set the max clock depending on situation etc - no overclock by default, which IMHO makes it suitable for integration in the default kernel) May 29 16:32:48 - support for a single big 121 Mb partition (C760, C860) May 29 16:33:16 who can I send them to ? [preferably someone who can test them] May 29 16:34:36 guylhem_: the openzaurus kernel bb file lists RP and mickeyl as maintainers May 29 16:34:49 guylhem_: and RP looks like a good choiche to me May 29 16:35:00 IMHO we should never support overclock May 29 16:42:07 zecke - it is an option that is not activated by default May 29 16:42:21 guylhem_: users are... :) May 29 16:42:48 03mickeyl 07org.oe.dev * r53a72fde... 10/packages/netbase/ (3 files in 3 dirs): netbase: add proper interfaces file for Motorola GSM Phones May 29 16:42:49 .... dumb :-) but it only offer the user the choice to manually set the max freq to 530 for the performance driver May 29 16:42:53 03mickeyl 07org.oe.dev * re7d1a141... 10/packages/qte/ (6 files): qte(-mt): touchscreen works on Motorola EZX platform now May 29 16:43:05 zecke, do you have a c7x0 ? May 29 16:43:10 guylhem_: no May 29 16:43:33 guylhem_: I'm 'conversative' in allowing people to damage their hardware May 29 16:43:49 guylhem_: but send them to RP and he will decide May 29 16:43:53 ok May 29 16:44:06 I'll also ask him why I have problems with the sound driver. May 29 16:44:39 I'm using http://www.rpsys.net/openzaurus/patches/bbfiles/linux-openzaurus_2.6.16.bb May 29 16:45:12 for some reason it gives: May 29 16:45:13 ALSA sound/core/pcm_lib.c:1839: BUG? ((max >= 0)) May 29 16:45:14 [] (dump_stack+0x0/0x14) from [] (snd_pcm_hw_param_near+0x3)[] (snd_pcm_hw_param_near+0x0/0x3a0 [snd_pcm]) from [] (snd) May 29 16:45:17 (...) May 29 16:45:49 guylhem_: RP and lrg are safe bets for help May 29 16:46:54 zecke: A lot of users want to use overclocking May 29 16:47:15 zecke: It needs to be an option May 29 16:47:24 a lot of people put animals in microwaves as well May 29 16:47:27 lardman: Manually apply patch, flash would b mine May 29 16:47:30 yesterday talked about this, he said I may have a ram corruption problem, but I recompiled and uploaded new drivers - same error. Odds are against having the same corruption May 29 16:47:37 guylhem_: Good work, people have also been asking for that single root partition May 29 16:47:52 zecke: users! ;) May 29 16:47:54 happy to help. I'm not a big fan of oe, but I'll take time to learn. May 29 16:48:10 lardman: users with special needs/wishes, need to do special things May 29 16:48:19 I feel like most of my work is currently wasted, while in oe it will help more people May 29 16:48:31 lardman: e.g. I wish my display would be more bright May 29 16:48:40 lardman: but there is a reason apple limited it, physics May 29 16:48:52 :) May 29 16:48:57 next item on my list 640x352 video playback. currently exploring mplayer-w100 on a 2.6.16 May 29 16:49:10 zecke: I understand what you say, but as long as there's a warning about it, we need it to compete with other 'distros' like pdaX May 29 16:49:30 lardman - you have a c7x0 ? May 29 16:49:37 guylhem_: yes May 29 16:49:47 I'd like to send the patch to someone else for validation. Please send me your email to guylhem gmail.com May 29 16:50:05 I'll send you a binary of the kernel, the .config and the parts to change May 29 16:50:12 on your local sources. May 29 16:50:38 guylhem_: You can send me the stuff by all means, but I won't have time to do anything until next week I'm afraid May 29 16:51:36 anyone can tell me where in kernel config i can specify the init? May 29 16:52:14 no pb. at least someone else will have it May 29 16:53:08 TheCan: CMDLINE May 29 16:53:13 TheCan: CONFIG_CMDLINE May 29 16:53:20 zecke, thx! May 29 16:53:40 guylhem_: sent May 29 16:54:08 * lardman goes back to searching for papers/references May 29 16:54:57 * zecke continues to study as well May 29 16:58:16 koen: I got a gdk-pixbuf package ready if interested. May 29 16:59:24 zecke, i can?t find it :( my problem is i once had an busybox-based system now i want to switch to a normal debian and debian doesnt boot to the end because i get no console etc. as it doesnt find the init. any idea how i could fix this? i thought about a symlink but what did i have to symlink then? i think busybox uses /linuxrc for init.. May 29 17:00:37 TheCan: not enough input. Your kernel has a CMDLINE either given by a bootloader or compiled in (I do not even know your arch....) May 29 17:01:03 TheCan: but the kernel (2.6er) has atleast three fallbacks overuling the init= param... May 29 17:01:14 TheCan: more in init/main.c of your favorite kernel May 29 17:01:23 i386 - ah so there is some init= param maybe also worth a try May 29 17:01:56 lardman : sent May 29 17:01:56 if i get this here: "init: Id "D1" respawning to fast: disabled for 5 Minutes" did it already load init and i got a problem with getty etc. maybe? May 29 17:02:24 likewise: attach the package to the bugtracker and I'll have a look later this evening May 29 17:03:06 koen: it's not much use without packages depending on gdk-pixbuf really. I have my own small tree of packages that does without X now. May 29 17:03:47 don't forget to add the appropriate RREPLACES to gtk+ May 29 17:03:58 guylhem_: Thanks :), I'll take a look at it as soon as I can May 29 17:08:10 * Philippe is back (gone 20:14:13) May 29 17:11:13 could someone confirm me the .bb file that is used by 3.4.5.1 current to build the 2.6.16 kernel is identical to http://www.rpsys.net/openzaurus/patches/bbfiles/stable ? May 29 17:11:22 I think it is missing a couple of patches May 29 17:11:59 check the branch May 29 17:12:06 everything is out in the open May 29 17:14:05 (in case of "oh noooos, monotone too hard, OMG LOL 1337" -> http://www.openembedded.org/repo/) May 29 17:15:29 koen, I must say that for the 1st time this is a helpful reply :-) May 29 17:15:42 found the bb there and yes it is different May 29 17:15:46 r33 vs r7 May 29 17:29:42 guylhem_: RP tends to keep the kernel up to date in OE May 29 17:30:49 I didn't know :-/ May 29 17:43:52 bye chaps May 29 17:54:02 koen, do you have a page that the project proposal for the staging thing has been moved to? I was curious to see how it'd be implemented, and it looks like google made them no longer exist (at least on the public web page) May 29 17:55:43 joshua_: it's all at http://www.openembedded.org/viewmtn/branch.psp?branch=org.openembedded.packaged-staging :) May 29 17:56:55 hmm.. do you still have a copy of the original proposal that you submitted? May 29 17:57:33 joshua_: http://oe.pastebin.com/745208 May 29 17:57:55 excellent. do you have a wiki page for this set up yet? May 29 17:58:10 nope May 29 17:58:17 there's a readme in the branch May 29 17:58:33 and all the code is online for everyone to see May 29 17:58:43 I'm eagerly awaiting this... crosstooling and doing dev package install magic by hand for the powerpc boards sort of sucks May 29 17:58:44 any idea on how to get a tarball from http://atty.skr.jp/WebSVN/log.php?repname=mplayer-w100&path=%2Ftrunk%2F&sc=1 ? May 29 17:59:20 guylhem_: why do you need a tarball? May 29 17:59:31 koen, okay. might I recommend that you set up a status wiki page off of SummerOfCode so that people like me who aren't very familiar with OE/monotone can take a quick glance at what's going on? May 29 17:59:57 joshua_: I can't do that May 29 18:00:05 oh, crap, I forgot May 29 18:00:13 joshua_: I'm forbidded to use any of hh.orgs services, remember May 29 18:00:39 damnit. er, wanna email someone every so often who will update by proxy? May 29 18:01:23 I have other uses for my time May 29 18:01:45 _guillermo: can't you just use svn:// in the .bb? May 29 18:02:52 I think that everybody would find it very useful if you did. May 29 18:03:32 20:12 < koen> there's a readme in the branch May 29 18:03:33 20:12 < koen> and all the code is online for everyone to see May 29 18:03:47 koen : I am trying to compile the latest (1.1.6) to test if I can key a faster video playback on c70 with 2.6.16 May 29 18:03:53 and that's all I'm going to say about it May 29 18:04:24 guylhem_: did you see mardy's mail about kdrive-w100? May 29 18:04:29 sigh... May 29 18:04:46 koen : got a pointer ? May 29 18:05:42 guylhem_: http://handhelds.org/hypermail/oe/89/8945.html May 29 18:08:14 great news! May 29 18:12:10 meanwhile I'll still try to get a working mplayer :-) May 29 18:13:16 is there a working abiword ? the one currently on the feed is quite dead May 29 18:15:29 guylhem_: see my response to your bugreport May 29 18:16:08 and fwiw, everything in that feed should be discussed in #openzaurus May 29 18:16:18 hey _law_ May 29 18:16:38 <_law_> hi koen May 29 18:19:47 guylhem_: I've noticed you comments about the patches. OZ doesn't support a unified root filesystem and never will so that patch will not make it into OE May 29 18:23:20 RP - too bad. this would be a usefull feature. Personally, I only had to 1) flash that kernel 2) fill the partition will the jffs2 image 3) remove /home from fstab and then everything worked out of the box May 29 18:23:43 regarding the overclock, any hope? May 29 19:17:41 "selected the opie image." ... "no reference to X in /etc/inittab and no indication of how to get it running." May 29 19:17:43 users.... May 29 19:21:02 haha ;) May 29 19:26:12 koen: JFYI: gcc-4.1.1 works alrighty on the ep93xx. May 29 19:26:42 koen: JFYI: correction: gcc-4.1.1-compiled kernel and userland work alrighty on the ep93xx. May 29 19:29:48 hello all May 29 19:30:07 likewise: softfpa? May 29 19:30:10 hey Greg2 May 29 19:30:21 hey koen May 29 19:31:35 koen: Default build, I assume softfloat is enabled in the build, from what I saw. I get no kernel float traps on floating point apps. May 29 19:32:01 hey likewise May 29 19:32:22 Greg2: hey there. Dialin-in again? :-) May 29 19:32:38 likewise: that's pretty good news May 29 19:32:39 yes :( May 29 19:33:34 koen: know of a floating point testsuite in OE? I would like to test this a bit more thorough myself. May 29 19:33:55 likewise: I have no idea May 29 19:34:34 koen: hmmm, will google a bit. should be some testsuites around (maybe we should even run a testsuite in qemu) May 29 19:39:35 03koen 07org.oe.dev * r72d107b2... 10/packages/gpe-calendar/gpe-calendar_cvs.bb: gpe-calendar: add libhandoff to depends May 29 19:41:11 likewise: you know the MACHINE=qemu-arm trick? May 29 19:41:13 likewise: http://rpsys.net/openzaurus/qemu/ May 29 19:41:31 koen: not yet :-) May 29 19:46:53 likewise: i've set up oe/bitbake on my laptop and desktop May 29 19:46:53 dial-up connection makes this take a very long time... different distro on each box May 29 19:47:18 ikewise: i've set up both so i can take my laptop to a place with dsl and do 'bitbake -c fetch opie-image' for source d/l only :) May 29 19:47:57 greg2: smart move. is dsl hard to get at your place, or expensive? May 29 19:48:18 likewise: both May 29 19:48:44 s/ikewise/likewise May 29 20:20:22 koen: i know you are a gpe user, but you most likely know this... May 29 20:20:29 koen: if i want to build the latest poodle (opie) image with a 2.6 kernel do i use openembedded.dev or openembedded.oz354x? May 29 20:20:54 .dev May 29 20:21:05 koen: thanks May 29 20:21:10 iirc .oz354x doesn't have the bits from poodle/2.6 May 29 20:21:43 CoreDump|home is the poodle 2.6 expert May 29 20:21:54 wtf May 29 20:22:01 koen: i know :) May 29 20:22:02 google wants the documents faxed May 29 20:22:08 faxed..... May 29 20:30:19 guylhem_: I'm not sure about the overclocking patch. The accepted via is we don't support overclocking as standard as we don't want any responsibility if it goes wrong May 29 20:31:46 koen, assuming there is no personal information, email me a scan, and I will print and feed back into the fax machine May 29 20:32:10 (unless you know of a way to make an x100-like card or an alsamodem fax...) May 29 20:32:13 joshua_: thanks, but I'll fax it tomorrow at the uni May 29 20:32:15 okay May 29 20:32:54 I don't even have a POTS subscription at my house :) May 29 20:33:04 voip? or just cell? May 29 20:33:21 just cell May 29 20:33:34 since a gsm subscription via the uni was free May 29 20:34:22 hmm, cool May 29 20:36:31 How does OE bitbake decide what to package? "NOTE: the following files were installed but not shipped in any package:" May 29 20:37:00 likewise: via the FILES var May 29 20:37:18 it has some default entries like ${bindir} and ${libdir} May 29 20:37:32 koen: ok, I should just += add to it. May 29 20:37:45 you can ship the files by doing FILES_${PN} += "/mydir /dir/file" May 29 20:38:12 try to use ${libdir} and vars like that May 29 20:38:20 for people redifining such paths :) May 29 20:40:04 koen: Instead of making a lot of patches, I'll put my X-less tree online as it still under refinement. So far, gdk_pixbuf, cairo, pango, freetype, fontconfig, fdclock are now without any X deps. (It's http://www.esrac.ele.tue.nl/~leon/linux/openembedded/) May 29 20:41:01 koen: It feels as if OE needs a USE variable (like portage/emerge) to cope with system wide features. (I.e. depend on features instead of (virtual) packages). May 29 20:41:52 USE flags are insane as soon as you start spreading packages May 29 20:42:05 they makes lots of sense for stuff like gentoo May 29 20:42:12 koen: yes, agreed. May 29 20:42:48 the virtual/thing is suboptimal, though May 29 20:43:24 OT: do not use 'suboptimal' in a control engineering exam May 29 20:44:01 koen: but I see no sane way to merge my tree with OE (if this is senseful for anyone but me, I like the BBFILE_COLLECTIONS feature to run a higher priority tree next to that of org.openembedded.dev). May 29 20:44:28 koen: as in regelsystemen? May 29 20:44:34 yes May 29 20:45:50 koen: OT: took me ages to pass the control exam (well not the system modelling part, but systems control). May 29 20:46:50 second year course and I still haven't passed it May 29 20:46:56 (started 2001) May 29 20:47:12 koen: likewise :-) May 29 20:47:59 koen: is it classical control (Bode, Nyquist, Zero-Poles, ...)? May 29 20:48:10 yep May 29 20:48:13 not too hard May 29 20:48:16 but annoying May 29 20:48:40 * Philippe remembers that crap.... May 29 20:48:50 good night all May 29 20:48:58 masons rule sucks with pen + paper May 29 20:49:10 koen: it takes intuition to understand the interaction between pole/zero placement and system behaviour (at least that's what my prof said - never had it) May 29 20:49:22 nite greg2 May 29 20:49:32 I hated it even more to discover on the exam that I seemed to have missed a lesson and was missing part of the course. May 29 20:49:33 ah a bit too slow... May 29 20:54:36 likewise: I think we can add you packages as foo-nox and add the appropriate RREPLACES and virtuals (shouldn't be needed) May 29 20:56:20 koen: I am now testing the final build. One package (cairoclock from MacSlow) wasn't in OE yet AFAIK, so I will make a X variant of it as well. It's a nice eye catcher app. May 29 20:56:56 you can even test it on your workstation May 29 20:57:07 thanks to RP qemu has graphics support May 29 20:58:26 koen: yes, framebuffer support if I stand correct? May 29 20:58:35 yes May 29 20:58:50 on which you can run xfbdev or qt/e May 29 20:58:59 or plain fb apps May 29 20:59:48 koen: I am thinking of building a GTK+-on-GDK-on-DirectFB-on-FB as a next project. May 29 21:00:09 like the debian-installer people do? May 29 21:00:22 koen: Now that the Mike Emmel (Debian Installer) has already done. Yes :-) May 29 21:00:32 s/the// May 29 21:02:05 gdk-directfb would be nice for phones May 29 21:02:36 koen: for all kinds of embedded devices :-) May 29 21:08:31 koen: eyecandy: http://www.esrac.ele.tue.nl/~leon/linux/ep93xx/ep9312_rendered_macslow_cairo_clock.png May 29 21:09:02 cute May 29 21:09:17 with kernel 2.4 you could even render it to the fb May 29 21:09:28 how is the performance if it? May 29 21:10:17 bad... May 29 21:10:31 crunch! May 29 21:10:43 koen: critical paths of cairo are fixed point. May 29 21:11:07 koen: I hope to be able to run a second hand that actually moves every second :-) May 29 21:11:16 hand as in wijzer May 29 21:12:04 koen: the fdclock (simpler, direct cairo without librsvg) takes about 300 ms to redraw it's hands on a 800x800 display May 29 21:12:26 and cairoclock? May 29 21:12:49 just now adding a printk, just to get some feedback on its performance (and removing libpng write-out, which is the performance killer now) May 29 21:15:51 May 29 21:19:16 ok, her 5 minutes of fame: http://www.esrac.ele.tue.nl/~leon/linux/kitten_on_keyboard_error.jpg May 29 21:22:13 likewise, nice cat ! May 29 21:22:38 does not seem to press the keys too much yet... May 29 21:23:33 Philippe: half an hour ago: suspend key (didn't even know I had one)... Fedora goes into suspend-to-disk, but I couldn't get it to start up again... May 29 21:23:49 Philippe: I *just* finished my build though :-) May 29 21:24:46 likewise, sounds bad enough... OTOH you discovered you have a suspend key :-) May 29 21:25:41 Philippe: well yes, but I never got it to work under Linux on my previous computers so I do not bother... May 29 21:25:53 Now only if the cat could code... May 29 21:26:05 then I would have a small developer team here May 29 21:26:49 likewise, if you could train it to do that you might be a rich man soon... May 29 21:27:27 s/man/person/ May 29 21:29:59 koen: fix for glibc 2.4 works, it solve partial loop problem, but now it loop later in nptl dir May 29 21:31:39 aquadran: hey, we saw that one also. pthread-error.h? May 29 21:32:36 likewise: yes May 29 21:33:47 aquadran: you have both fixes? (fix one enables -isystem in bitbake.conf, fix 2 involves modified FLAGS for glibc) May 29 21:35:37 likewise: only second, maybe i miss first May 29 21:36:21 second is most important, first does not do anything for glibc anymore, I *think* May 29 21:37:01 aquadran: you checked out a recent OE, or did you patch your local copy? May 29 21:37:35 likewise: i patched localy from email May 29 21:38:51 likewise: what first patch did ? i hade i bit problem with emails, so propably i lost that one May 29 21:39:12 ibot: calc 7.5 * 1.2 May 29 21:39:28 hold on. got a Skype call... May 29 21:47:16 likewise: if you have bitbake 1.5.x (svn trunk) you can do the following: http://oe.pastebin.com/745721 May 29 21:50:39 * Philippe is away: visual contact - melancholic dreams May 29 21:57:25 likewise: http://dominion.kabel.utwente.nl/koen/pda/images/librsvg.png May 29 21:58:15 or http://dominion.kabel.utwente.nl/koen/pda/images/librsvg.svg May 29 22:04:44 generated with this command: http://oe.pastebin.com/745756 May 29 22:16:19 likewise: did you see the librsvg graph? May 29 22:16:35 koen: updating bitbake as we speak May 29 22:16:57 http://dominion.kabel.utwente.nl/koen/pda/images/librsvg.png May 29 22:17:52 koen: this is excellent. I have a *handwritten* A4 paper here, which looks almost alike! May 29 22:18:06 00:18 < koen> generated with this command: http://oe.pastebin.com/745756 May 29 22:19:12 koen: the ep9312 runs 120% realtime with the cairo clock at 32x32.... May 29 22:19:35 heh May 29 22:20:38 koen: and ~18% real-time when drawing it 500x500 pxls May 29 22:21:00 hmmm. May 29 22:21:39 bitbake 1.5 has a lot of debug spew May 29 22:22:19 time to go to sleep May 29 22:22:20 'night all May 29 22:28:33 nite koen May 29 22:36:48 * v8jlene is back (gone 08:35:21) May 29 22:57:27 mithro: hi May 29 23:00:17 evening guys. May 29 23:00:55 does someone know some loader like grub I could use with oe on my z3100 ? May 29 23:01:05 I wish to try some kernels. May 29 23:13:29 magnet: Does altboot work on that model? May 29 23:15:10 yea. May 29 23:16:43 can I boot other kernels with it ? May 29 23:19:04 Dunno, I've never used it. My boards just run lilo (have block mapper over the flash). Perhaps this will tell you something: http://www.hentges.net/misc/openzaurus/docs/altboot.shtml May 29 23:20:26 I check. May 29 23:20:55 magnet: Apparently not ;( "Please keep in mind that while altboot does enable you to use multiple root partitions / installations it does (currently) not allow you to boot another kernel. To load a different kernel than the one flashed, the kexec tools are required. Currently this is only possible with kernel 2.6 so your device must support this kernel." From http://openzaurus.org/wordpress/howto/the-altboot-faq/ May 29 23:21:45 it don't sound like i can setup it too have alternate boot kernels May 29 23:24:44 magnet: If you just want to test out newer kernels then you could always nfsboot them? That's how I do all of my kernel testing. May 29 23:25:11 umm May 29 23:25:18 how do you do that ? May 29 23:26:58 you could use altboot to kexec another kernel May 29 23:27:35 I didn t thinked about kexec. May 29 23:27:38 Hey, can I ask a few questions? May 29 23:28:06 I'll try it May 29 23:28:21 JoeSchmo: any docs on to get it working with alboot ( and why ? ) May 29 23:28:51 Do you guys think that if you're doing an EE major that a CS minor will help if you're having trouble with programming? May 29 23:29:04 magnet: altboot is just nice to use for testing or using alternate syste,s May 29 23:29:32 magnet: you can also just use kexec on its own without altboot May 29 23:29:42 great May 29 23:30:02 I just want to make some changes in the kernel sources. and test them May 29 23:57:50 it's weird I'm getting : install: cannot run strip: No such file or directory May 29 23:58:02 but strip is in my path May 30 00:03:30 yeah, I was getting that too. deleting tmp and restarting seemed to fix it May 30 00:04:52 ok May 30 00:30:08 * v8jlene is away: out May 30 00:57:53 i got a following error mesg when i execute "bitbake task-bootstrap" (staging gcc-cross-4.0.0-r1). May 30 00:57:53 ERROR: log data follows (/home/stuff/build/tmp/work/gcc-cross-4.0.0-r1/temp/log.do_stage.22529) May 30 00:57:53 | NOTE: make ARCH_FLAGS_FOR_TARGET= install May 30 00:57:53 | make: *** No rule to make target `install'. Stop. May 30 00:57:53 | FATAL: oe_runmake failed May 30 00:58:14 does anybody know howto solve this problem ? **** ENDING LOGGING AT Tue May 30 02:59:57 2006