**** BEGIN LOGGING AT Fri May 19 09:59:56 2006 May 19 10:09:13 hi all May 19 10:09:38 hey Dirk May 19 10:09:51 do13, fising the WM9712 audio bugs now. May 19 10:09:55 fixing May 19 10:10:08 hey Liam May 19 10:10:29 lrg: which bugs? May 19 10:10:47 do13: the dapm bug with the left and right mixers May 19 10:11:21 ixoth: mickeyl also got it so fix it or help fixing May 19 10:11:36 do13, it requres a double unmute in alsamixer to get stereo out of the headphone at times May 19 10:12:05 lrg: is this tosa specific? May 19 10:12:29 lrg: do you have the same prob on mainstone? May 19 10:12:56 do13, it's wm9712 specific. So the fix for mainstone will apply to tosa. May 19 10:13:33 i fixed it in my build (edited .bb file a bit), now i have to see whether it builds using old glibc and sumbit patches :) May 19 10:14:28 ixoth: great May 19 10:18:41 re May 19 10:18:53 hi koen May 19 10:19:31 hey dirk May 19 10:23:37 hrw|work: how far away is 3.5.4.1 ? May 19 10:30:14 koen: images need testing and I will probably release rc4 as final May 19 10:30:24 ah, ok May 19 10:30:56 koen: planned to add kernel logos but it fails on rotated framebuffers May 19 10:30:57 in that case I will wait with adding the new gpe-calendar bits May 19 10:31:25 during next week I plan to mail linux-fbdev about problem May 19 10:34:35 cu May 19 10:35:03 integration event in company - we go somewhere out May 19 10:35:15 have fun May 19 10:35:47 tomorrow: T2 presentation, GPG signing party, museum night May 19 10:35:50 will be great May 19 10:40:10 koen: What are you recommending in that thread? May 19 10:40:24 koen: clear the cache? May 19 10:41:10 lardman|work: setting the correct distro May 19 10:41:30 and updating bitbake to a non-broken version May 19 10:41:44 koen: Ah, okay. Do you have an example local.conf for him - he seems keen so I'd like to get him up & running May 19 10:42:22 lardman|work: the flash movie I linked to has all the bits needed May 19 10:42:32 you only need DL_DIR, DISTRO and MACHINE May 19 10:42:33 koen: okay May 19 10:42:35 hey zecke May 19 10:42:39 hey May 19 10:42:44 mickey|zzZZzz: hey May 19 10:42:58 lardman|work: and TMPDIR, but local.conf.sample covers that May 19 10:43:04 oh librsvg add -I/usr/include May 19 10:43:13 zecke: http://lists.gnumonks.org/pipermail/openezx-devel/2006-May/000416.html May 19 10:43:16 /usr/include/libcroco-0.6 May 19 10:44:12 OE cvs? May 19 10:44:27 I suspect harald means oesources.org May 19 10:44:44 grrr May 19 10:44:49 glibc is still broken: http://oe.pastebin.com/726163 May 19 10:44:52 yeah gpe sucks lately (from a building point of view) May 19 10:44:58 * zecke sits naked in the office :} May 19 10:45:21 make -j44? May 19 10:45:26 well, if glibc would build I could fix the gpe bits May 19 10:45:32 zecke: I tried to get past that point May 19 10:45:41 -j1 doesn't help either May 19 10:46:54 koen: do you know if our librsvg should use libcroco for CSS? May 19 10:46:58 zecke: do you know why OE doesn't use make-native? May 19 10:47:11 koen: no, but we should add it May 19 10:47:26 DEPENDS = "gtk+ libcroco libart-lgpl libxml2 popt" May 19 10:47:32 so, yes, it should use libcroco May 19 10:48:08 http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=4&logfile=20060519100944.log May 19 10:48:11 zecke: I put fail-fast in overrides as well May 19 10:48:14 anyone feels like fixing? May 19 10:48:23 as I said, as soon as glibc builds.... May 19 10:48:30 hehe May 19 10:48:36 * zecke needs to finish stuff as well May 19 10:48:38 later May 19 10:49:00 dentist+rain it would be the perfect monday May 19 10:52:43 lardman|work: bitbake 1.4.3 takes care of cache flushing on its own May 19 10:53:26 lardman|work: he might be using the wrong branch/distro config May 19 10:53:52 ehm May 19 10:53:56 s/config/combo/ May 19 10:54:02 koen: yep May 19 10:54:19 koen: We'll see if changing the distro name sorts him out May 19 10:54:39 CoreDump|home: I might have a poodle soon as well May 19 10:55:09 lardman|work: ask for the bitbake summary when debugging :) May 19 10:55:14 zecke: nice =) Kernel 2.6 still needs some TLC May 19 10:55:30 CoreDump|home: hehe I though more of tracking your progress May 19 10:55:31 lardman|work: .dev has added DISTRO_VERSION, so you can check it local.conf has typos in it May 19 10:55:48 koen: Okay, cool May 19 10:56:03 zecke: the userspace stuff is looking good already May 19 10:56:45 CoreDump|home: Max has a poodle from 10east.com, I might have it next month May 19 10:56:47 _might_ May 19 10:57:39 any jffs2 guru around? I need to reformat /home =) May 19 10:58:31 CoreDump|home: in #opie is Dawid Woodhouse May 19 10:58:47 CoreDump|home: install the jffs2 utils and erase it? May 19 10:58:56 flas_eraseall? May 19 10:58:59 flash* May 19 10:59:12 I'll check May 19 10:59:44 * CoreDump|home is a jffs2 n00b May 19 11:01:58 ipkg list | grep mtd May 19 11:03:22 * CoreDump|home found a NAND restore and sees if it works May 19 11:07:55 eh lol, the backup was from a tkc ROM May 19 11:15:38 ugh, that restore killed the update system May 19 11:21:35 * koen curses at uclibc May 19 11:22:01 why? i've just built 2.4 opie-image with uclibc using gcc-3.4 May 19 11:22:43 try building a uclibc images using the arm eabi May 19 11:22:43 there were some issues but easily fixable May 19 11:22:58 oh, i see.... May 19 11:24:02 a better description of the problem: http://handhelds.org/hypermail/oe/87/8799.html May 19 11:24:18 that mail even ignores the fact that uclibc svn is needed May 19 11:25:45 koen: uclibc svn doesn't build with gcc4 atm if you want locale working May 19 11:26:37 does eabi also need nptl too ? May 19 11:28:31 uclibc has not got nptl in trunk yet ... sjhill has a working nptl in his branch but i don't think its arm stuff is upto date, only mips i think May 19 11:31:54 ade|desk: nptl would be nice May 19 11:35:08 ~lart hi-mobile.net May 19 11:35:08 * ibot whips out a hot clue gun and makes sure that hi-mobile.net is stuck to the floor May 19 11:37:39 koen: what they done now ? May 19 11:39:34 incomprehensible website May 19 11:41:13 really? i dont find it too hard , whats up maybe i can help? May 19 11:41:51 they say they only ship to us/canada/france/uk May 19 11:42:01 yup May 19 11:42:01 but when I order it it accepts the netherlands... May 19 11:42:05 lol May 19 11:42:35 oh well , i am still in 2 minds to get a a780 May 19 11:42:51 hey hi May 19 11:43:19 my replacement credit card came today after debbie lost hers.... itching to use it .. but its still a lot May 19 11:43:40 wth? http://openzaurus-ja.sourceforge.jp/index.php?OpenZaurus May 19 11:45:20 mickeyl: just ordered a a780 May 19 11:45:21 hehe i guess babelfish isn't good enough May 19 11:45:49 mickeyl: us-version without gprs, but with quadband gsm May 19 11:45:51 hi mickeyl May 19 11:46:00 koen: i dont mind the price of the a780, the shipping is a bit steep at $40 May 19 11:46:14 03mickeyl 07org.oe.dev * rbfc64fbe... 10/packages/libopieobex/ (libopieobex0.inc libopieobex0_1.2.1.bb libopieobex0_cvs.bb): libopieobex: catch up with recent opie cvs developments May 19 11:46:20 03mickeyl 07org.oe.dev * r1728e98c... 10/packages/libopieobex/libopieobex0.inc: libopieobex0: fix typo May 19 11:46:32 ade|desk: I can ship 2 a780 with 2x shipping for the price of buying 1(!) locally May 19 11:46:47 the cheapest one in .nl is E480 without S&H May 19 11:46:58 hi-mobile is E240 including S&H May 19 11:47:23 so I can live with E36 for S&H May 19 11:47:48 now I need to come up with a excuse why I bought another toy May 19 11:48:09 fair enough. i thought it had gprs just not the GPS ... glad you told me.. now i shall not bother May 19 11:49:02 Dial-up (CSD) or high-speed, always-on (GPRS or EDGE) data supportbr> from the http://www.hi-mobile.net/Motorola/Motorola-A780-Fully-Unlocked/ May 19 11:49:32 ehm May 19 11:49:36 without gps that is :) May 19 11:49:51 I bought it because it has edge May 19 11:49:57 damn you koen ... now i mght get one ..grrrr May 19 11:50:30 koen: do you have EDGE in NL? May 19 11:50:41 zecke: in some parts, yes May 19 11:50:47 koen: darn. I bet you have it earlier than me :/ May 19 11:50:49 zecke: but we have gprs everywhere May 19 11:51:06 mickeyl: I'm expecting a "we don't ship to .nl" mail soon May 19 11:51:17 heh May 19 11:51:25 * mickeyl adding ODevice::Motorola_EZX May 19 11:51:36 mickeyl: Qtopia4 has ODevice May 19 11:51:41 nah! May 19 11:51:43 really? May 19 11:51:55 mickeyl: where did you add EZX to? May 19 11:52:16 mickeyl: we might consider licensing Qtopia for OE :} May 19 11:54:16 zecke: since we have no idea how many ezx devices we'll see, I'll add it as an own class. May 19 11:54:27 first commit will appear soonisch May 19 11:54:30 soonish even May 19 11:55:14 zecke: btw., any object to removing beagle from ODevice? May 19 11:55:35 yes May 19 11:55:41 API stability May 19 11:56:02 API what? May 19 11:56:20 zecke: I'm building oz-unstable now with fail-fast May 19 11:56:26 koen, did a quick test on glibc 2.4 but for me it already fails during configure: May 19 11:56:28 configure: error: compiler support for __thread is required May 19 11:56:30 FATAL: oe_runconf failed May 19 11:56:37 eFfeM-log: yeah, that needs nptl and stuff May 19 11:57:20 mickeyl: GCC 4.1.1 RC1 builds and works May 19 11:57:22 ok May 19 11:57:34 mickeyl: but glibc 2.4 has a make loop in do_stage...... May 19 11:58:32 koen: progress May 19 11:59:53 koen : which site for ordering an a780 are you talking about please ? May 19 12:01:15 theturtle: hi-mobile.net May 19 12:01:22 ok thanks May 19 12:01:57 I suspect the 'location' field just means 'which languagepack' May 19 12:02:03 at least I hope so May 19 12:02:20 i don't understand something : during a 'bitbake gpe-image' it appears that sdl-qpe and sdl-x11 where built. and some programs try to link sometimes to one and sometimes to another. how can i say to bitbake "always use sdl-x11" ? May 19 12:03:17 have you done PREFERRED_PROVIDERS += " virtual/libsdl:libsdl-x11" ? May 19 12:03:37 Hi all. May 19 12:03:38 if yes, you may try to find more info in bitbake -n gpe-image May 19 12:04:04 just look for warnings May 19 12:04:06 yup i've done it ixoth May 19 12:04:27 hum ok May 19 12:04:55 * koen installs ubuntu on his x86 workstation May 19 12:04:59 ~hail tyan May 19 12:05:02 ~ change 310 usd to gbp May 19 12:05:04 * ibot bows down to tyan and chants, "I'M NOT WORTHY!!" May 19 12:05:08 310.00 United States Dollar (USD) makes 165.096 Pound (GBP) (from http://www.xe.com/) May 19 12:05:33 hmm car tax or phone ... May 19 12:05:43 * koen has no car May 19 12:06:08 koen : products on hi-mobile.net will be taxed when imported, won't they ? May 19 12:06:15 ade|desk: phone :) May 19 12:06:17 theturtle: probably May 19 12:06:30 so it could be not that interesting May 19 12:06:57 even with tax it is cheaper as buying local May 19 12:07:04 * koen doesn't understand economics May 19 12:07:06 i paid 30euros of taxes for 5 tshirts from thinkgeek... i can't imagine how much it would be for a cellular =) May 19 12:07:17 approx 19% May 19 12:07:30 ok May 19 12:08:03 in the uk its a random number between 0 and 100 % so it seems May 19 12:08:19 ixoth : i issue no warning at all with bitbake -n gpe-image May 19 12:08:27 ade|desk : in france it's almost the same May 19 12:08:44 i had 60$ of tshirts and paid 30euros of fees... May 19 12:08:51 theturtle: where you the one who wanted to file a bug report for bitbake May 19 12:08:58 theturtle: to propose better wording for the help? May 19 12:09:05 hmmm... and what is the output? May 19 12:09:16 yet you can bring in cocaine for free May 19 12:09:24 ixoth : just a list of packages started/completed May 19 12:09:48 zecke : not really for the output of bitbake help, i was thinking of something about bitbake output itself May 19 12:10:03 and so was thinking _guillermo too if i remember May 19 12:10:04 just smugle it inside a asylam seeker May 19 12:10:24 theturtle: did you do that already? I have missed the bug number then May 19 12:10:30 ade|desk: over here we use the navy to do that May 19 12:10:46 yeah, and what .bb it tries to build after libsdl-qpe ? May 19 12:11:03 zecke : no, i'm a little bit busy with my thesis currently, and with my bitbake gpe-image, but after the week end it'll be ok i thing. May 19 12:11:03 koen: i never knew the dutch had a navy May 19 12:11:09 cool May 19 12:11:55 i suppose some .bb that builds after libsdl-qpe requires that qpe stuff May 19 12:12:19 ade|desk: sure, we also have dozens of f16 fighter planes May 19 12:12:32 ade|desk: in case war breaks out in .nl or some nonsense like that May 19 12:13:40 ixoth : in fact the problem comes from mplayer which tries to link with qte things May 19 12:13:47 koen: just incase the 1940s come round again hehe May 19 12:14:05 ade|desk: yeah May 19 12:14:23 wars in europe are such a silly idea May 19 12:14:24 ade|desk: in '39 we flooded the west of .nl to stop the germans May 19 12:14:37 .. that worked well May 19 12:14:39 ade|desk: they forgot the germans had planes.... May 19 12:14:43 lol May 19 12:14:59 "it worked against the spanish people is 16xx" May 19 12:15:06 s/is/in/ May 19 12:15:13 ixoth : see http://pastebin.com/726402 May 19 12:15:38 don't understand where the depency against qte is coming from May 19 12:15:59 theturtle: bitbake -vvv? May 19 12:16:04 * ade|desk tries to remember the last time .nl and .uk went to war against each other May 19 12:16:23 must be over 120 years ago now May 19 12:16:23 ade|desk: 1500's I guess May 19 12:16:28 during the colonial wars May 19 12:16:50 didnt we have a punch up over south africa May 19 12:16:56 could be May 19 12:17:05 you also beheaded anne boleyn, remember? May 19 12:17:09 sounds silly enogh to be true May 19 12:17:25 oh yeah .. henry the 8th May 19 12:17:34 a nice guy by all accounts May 19 12:17:55 hmmmm seems that mplayer .bb has no references on qte May 19 12:18:04 yeah, I visited the tower a few months ago May 19 12:18:24 grrr May 19 12:18:27 ixoth : yep and bitbake -vvv shows nothing about qte, but ld tries to link qte libs... May 19 12:18:32 I suck at writing procmail rules May 19 12:18:51 * koen waits for another mail from the ezx list to test May 19 12:19:34 have you tried to build it from scratch? maybe you've got old libsdl somewhere in ${STAGING_BINDIR}/sdl-config May 19 12:19:36 koen: when are you next over in the uk, we should have a big get party somewhere in cambridge get pb_ et al for a drink May 19 12:20:01 ixoth : http://pastebin.com/726407 May 19 12:20:15 ixoth : don't think i really understand what you mean ? May 19 12:21:26 ade|desk: that'd be nice May 19 12:21:56 looks like you should -lsupcxx May 19 12:22:00 or use g++ to link May 19 12:22:10 -lsuckxxx? May 19 12:22:16 heh May 19 12:22:34 theturtle: it seems that theres libsdl-qpe is installed in staging... May 19 12:22:48 ixoth : you are right, sdl-config --libs shows up links against qt things May 19 12:22:56 its crazy that i live only 25 mins from pb_ yet haven't seen him for about 4 nearly 5 years May 19 12:22:56 * koen wonders how the google SoC is going to turn out May 19 12:23:09 how can i remove it in a clean way ? May 19 12:23:18 ade|desk: visit FOSDEM :) May 19 12:23:25 ade|desk: heh, right May 19 12:23:35 if only my company would pay for it May 19 12:23:51 theturtle: try to build it from scratch... and see where it appears in bitbake -n gpe-image May 19 12:25:36 ixoth : ok May 19 12:28:40 * ixoth thinks about installing ccache after rebuilding opie-image for 4th time from scratch May 19 12:28:45 right time to wonder around the windy village May 19 12:28:57 bbl May 19 12:29:44 ixoth: ccache won't help for a rebuilding since you build a new toolchain every time and that invalidates the hash May 19 12:29:59 oh damn...... May 19 12:30:14 now if we had packaged staging... May 19 12:30:19 * mickeyl looks to mithro May 19 12:30:24 * mickeyl looks to koen May 19 12:32:57 mickeyl: oh, by the way, hrw told me that you've had problems with building obexpush with uclibc? May 19 12:33:28 * koen whistles innocently May 19 12:33:46 mickeyl: I always did packaged staging :-) May 19 12:34:21 * likewise wanders off May 19 12:34:28 ixoth: yes. obexftp and obexpush have bugs linking iconv in if necessary May 19 12:34:36 it gets detected May 19 12:34:40 but not handed off to libtool May 19 12:34:46 hence we end up with undefined references on linking May 19 12:35:25 i've changed one string in obexpush to "export GLIBLIB=-L${STAGING_LIBDIR} -lglib-2.0 -liconv -lintl" and it works. kludge, but better than nothing May 19 12:35:44 yeah, i have something similar, but I won't push a hack like that May 19 12:36:00 i see... May 19 12:36:11 at least it would need to be conditional May 19 12:36:58 some times the developer should spent more time to config autotools the right way May 19 12:37:12 NOTE: package opie-image-1.0-r21: task do_build: completed May 19 12:37:24 * mickeyl will flash in a couple of hours May 19 12:37:41 mickeyl which device? May 19 12:37:48 collie May 19 12:37:51 oh May 19 12:37:54 i don't think i can flash that May 19 12:37:58 it's 21MB May 19 12:38:00 *sigh* May 19 12:38:01 to big? May 19 12:38:08 woglinde: sometimes? I wonder what percentage of packages are cross-compilable without any patchwork needed. May 19 12:38:21 likewise I know May 19 12:38:24 hrw|gone: any idea why opie-image is so big in .dev? May 19 12:38:39 woglinde: not that I am any good in autoconf... May 19 12:38:47 mickeyl du -m in rootfs? May 19 12:38:48 likewise: >70% May 19 12:39:06 koen: ok, I always work with the <30% :-) May 19 12:39:11 yeah May 19 12:39:12 likewise hm me too, but there is the internet May 19 12:39:22 and I can look for it May 19 12:39:27 the toolchains are notoriously bad at x-compiling May 19 12:39:42 mickeyl: konq-e included? May 19 12:40:04 woglinde: yes, but autoconf is one of the things you have to do regularly (fix up packages etc) in order to get it down. If you do it occasionally like me, I have to start from scratch every time. May 19 12:40:24 (and you get kludgework...) May 19 12:40:26 likewise but mostly they dont understand the autotools prozess May 19 12:40:29 hmm May 19 12:40:31 qt-mt May 19 12:40:42 qt-mt is 8mb extra? May 19 12:40:43 + all qte apps with RTTI and exceptions May 19 12:40:46 woglinde: they includes "me" as well :-/ May 19 12:40:58 *sigh* May 19 12:41:08 guess i need to use altboot May 19 12:41:08 woglinde: but I'll catch up on it, doing some reading on it. May 19 12:41:11 likewise get the time to change this May 19 12:41:18 yep May 19 12:41:44 03koen 07org.oe.dev * r3d66f5ad... 10/packages/libhandoff/libhandoff_cvs.bb: libhandoff: fix DEPENDS May 19 12:43:00 mickeyl: did you order a macbook yet? May 19 12:43:25 koen: no. i will order it the day i have submitted my thesis. that will be my reward. May 19 12:43:33 :) May 19 12:44:01 mickeyl: what do you think about having something like http://www.fredmiranda.com/forum/topic/398766 as a logo? May 19 12:45:53 koen: that looks cool. for Angstrom? May 19 12:46:07 mickeyl: yes May 19 12:46:26 I think I can digg up some archaic hw in the EE dustbins :) May 19 12:46:32 hehe May 19 12:47:21 mickeyl: cool pictures indeed. look you see this address line go high? May 19 12:47:22 EE dustbins? May 19 12:47:52 woglinde: leftovers in the Electrical Engineering building May 19 12:48:11 there is a 40 MB harddisk drive on our EE campus which looks like three stacked tumbledryers. May 19 12:48:26 koen hm May 19 12:49:17 we have a pdp-11 in our lunchroom :) May 19 12:49:51 youngsters .... May 19 12:50:04 koen we have a non functional in the foyer of our uni May 19 12:50:05 * eFfeM did his graduation work on a pdp-11 May 19 12:50:09 and the russian modell too May 19 12:50:22 5 MB removable hard disk May 19 12:50:22 effem hm so you are >100? May 19 12:50:31 lol May 19 12:50:41 halfway May 19 12:51:11 btw, backreading the log, this A780 looks really nice. How open/hackable is it ? May 19 12:51:26 eFfeM: http://openezx.org/ May 19 12:51:35 effem the problem is, isnt build any more May 19 12:51:42 ask mickeyl for the odeysee May 19 12:53:27 that is probably why hi-mobile has them so cheap, dumping the old stuff May 19 12:53:49 eFfeM: yes, and it's the stripped us version (no gps) May 19 12:54:08 ah, that is less good May 19 12:54:10 I already have a bluetooth gps, so I can live without it May 19 12:54:47 so have I, but it soaks my nokia battery quite fast (2 hrs or so with reduced backlight) May 19 13:04:51 hey zecke May 19 13:09:04 zecke: fail-fast is taking an awfull long time to fail ;) May 19 13:10:07 koen: librsvg failed on ewi May 19 13:10:10 koen: libxine fails May 19 13:10:53 it's at gmp over here May 19 13:11:19 zecke: when is it going to be activated in .dev? May 19 13:11:35 koen: once stuff compiles May 19 13:11:51 koen: and I have time to improve the tinderbox May 19 13:11:53 zecke: 87 responses May 19 13:12:23 zecke: any idea why hg has been running for some days now? May 19 13:13:10 koen: hehe May 19 13:13:20 koen: because it is running in a screen (tailor) May 19 13:13:47 koen: looks like he (on oesf) might have got somewhere now :) May 19 13:13:59 lardman|work: nice May 19 13:14:19 Another convert ;) May 19 13:14:20 lardman|work: oesf = Open Embedded Special Forces ? May 19 13:14:27 :D May 19 13:14:29 ~oesf May 19 13:14:32 i heard oesf is newer name of ZUG - it provide forum for Zaurus/Ipaq/Simpad/etc users (see zug). It has nothing common with OpenEmbedded (except first two words), or http://www.oesf.org/ May 19 13:14:45 Oh no. May 19 13:14:52 Indeed... May 19 13:14:55 hmm May 19 13:15:03 Open Embedded Special Forces is a neat name May 19 13:15:18 * koen should prints shirts with that and send them to core devs May 19 13:15:25 koen *g* May 19 13:15:50 Well hrw often parachutes into enemy territory on oesf ;) May 19 13:16:18 lardman|work: laying suppressive fire? May 19 13:16:30 (sp?) May 19 13:16:36 likewise: more like napalm May 19 13:19:13 heh May 19 13:27:17 ah May 19 13:27:25 koen: broke? May 19 13:27:35 * koen is bored and reads the "X sucks" thread on lkml May 19 13:27:47 where X is a place holder? May 19 13:28:25 for X11 ;) May 19 13:29:50 "X runs as root, which is bad because it can touch hardware" solution "put the gui in the kernel" May 19 13:29:54 *sigh* May 19 13:30:13 * mickey|coffeebrk waits for fbui to become useful May 19 13:30:25 use directFB May 19 13:30:31 they put fusion into the kernel, not more May 19 13:30:36 Alan cox: "If you merely put shit in May 19 13:30:37 a different bucket you still have a bad smell." May 19 13:30:56 hehe May 19 13:32:31 mickey|coffeebrk: this gcc patch was really written for uclibc May 19 13:32:42 mickey|coffeebrk: as I don't have an AMD64 ;) May 19 13:33:42 zecke: excellent :)) May 19 13:34:03 and autoconf really sucks... May 19 13:34:09 this presence and usable check is... May 19 13:35:00 ...often not working. May 19 13:35:18 mickey|coffeebrk: e.g for gcc and uclibc without iconv May 19 13:35:26 mickey|coffeebrk: it used xgcc to compile failed May 19 13:35:33 mickey|coffeebrk: then used /usr/bin/cpp succeeded May 19 13:35:41 mickey|coffeebrk: and claimed uclibc has support for iconv... May 19 13:35:53 nasty May 19 13:36:08 mickey|coffeebrk: If I would not have promised al to be more calm again I would say *moron* May 19 13:36:48 heh May 19 13:37:06 you really have promised here such a thing? tsk tsk... May 19 13:37:11 s/here/her/ May 19 13:37:31 visitors have arrived May 19 13:37:32 mickey|coffeebrk: yeah, well I was... but i have developed some bad habits May 19 13:37:33 later all May 19 13:37:35 cya May 19 13:37:38 cu koen May 19 13:37:43 *still no phone* May 19 13:37:48 same here :/ May 19 13:37:54 here as well... May 19 13:38:09 #oe - phoneless channel May 19 13:41:09 hmm... sashs uboot patches are way over my head May 19 13:44:34 koen, how did the 4.1 build go? I finally have my env setup so was going to give it a whirl... May 19 13:44:42 darn, missed him May 19 13:56:38 goodbye everyone May 19 14:01:36 joshin: it failed for him on glibc2.4 May 19 14:02:16 Thanks. May 19 14:36:08 koen: ARGH! May 19 14:36:16 koen: missed the DPD :/ May 19 14:36:27 koen: next delivery Monday May 19 14:37:57 mickey|home: monday next week? May 19 14:38:06 21th May 19 14:38:08 no May 19 14:38:09 22th May 19 14:38:11 *sigh* May 19 14:38:17 then again, that leaves me a weekend working on the thesis May 19 14:38:21 I bet it will be broken ;) May 19 14:38:30 with my luck, you're probably right May 19 14:38:41 mickey|home: more time for your thesis May 19 14:38:43 doh! May 19 14:38:53 *nod* May 19 14:39:13 mickey|home: now if your car would break, you could stay at home May 19 14:39:18 mickey|home: and have even more time for your thesis May 19 14:39:22 mickey|home: and your vdr... ;) May 19 14:39:25 heh May 19 14:39:34 *HexHex* May 19 14:39:37 :)) May 19 14:40:03 It's a pity that my motivation depends on the lossage of distracting things May 19 14:40:59 mickey|home: that is normal May 19 14:42:01 mickey|home: Qtopia is dual licensed. But none of these options are GPL ATM :( May 19 14:42:17 ljp: make that please change soon, I do not want to commit to Gtk+ May 19 14:42:36 zecke: what kind of fairy tale licences do they have? May 19 14:42:37 zecke: yes you do May 19 14:44:12 morning May 19 14:44:16 ade|desk: ? May 19 14:44:27 mickey|home: dunno OEM a, OEM b? May 19 14:44:36 ah, of course :D May 19 14:44:45 mickey|home: I think they have Platform aand SDK May 19 14:45:05 mickey|home: I thought it had the same license... May 19 14:45:57 zecke: come away from the dark side ... come to us gtk+ is your friend ...:) May 19 14:46:58 ade|desk: hehe, well LGPL is even more risky than TT's license May 19 14:47:20 ade|desk: anyway, for MobileDevices I think I'm soon 'forced' to use gtk May 19 14:49:05 plain? May 19 14:49:06 or mm? May 19 14:49:24 mm May 19 14:49:32 I'm insane enough May 19 14:49:32 that sounds better May 19 14:49:35 heh May 19 14:49:55 mickey|home: but Qt is such a nice and addictive drug May 19 14:50:10 i know. i really like it. May 19 14:50:23 QScene* looks awesome May 19 15:18:01 03eFfeM 07org.oe.dev * rdb35473b... 10/packages/pvrusb2-mci/ (pvrusb2-mci-20060517/Makefile.patch pvrusb2-mci_20060517.bb): pvrusb2-mci: added 20060517 snapshot May 19 15:29:32 a+ May 19 16:24:18 hrm May 19 16:24:37 have there been any significant kernel changes in OE in the past week? May 19 16:28:01 If I wanted to (try to) build a gcc 4.2 image, I set PREF_VER for gcc-cross to "gcc-cross_4.2-20060513" and gcc-cross-initial to gcc-cross-initial_4.1-20060217 since there's not (yet) a 4.2 version of cross-initial? May 19 16:29:19 I guess I could try and make/copy a 4.2 version of gcc-cross-initial. May 19 16:36:41 joshin: the version of gcc-cross-initial doesn't really matter. May 19 16:37:23 you might as well just use the default version, whatever that is, unless it actually doesn't work for you May 19 16:38:07 pb_: Ok, thanks. May 19 17:05:19 hrw|gone: I'm getting a -r32 at the end of my zImage now...is this intended? May 19 17:05:32 hrw|gone: (and I was very sad when booting with the logo didn't work :-( ) May 19 17:28:32 hi kergoth May 19 17:30:00 gcc 4.2 crashes out with "| /home/josh/zaurus/tmp/cross/arm-linux/bin/ld: Warning: size of symbol `__floatundisf' changed from 188 in libgcc/./_addsubsf3_s.o to 148 in libgcc/./_floatundisf_s.o May 19 17:30:00 | collect2: ld returned 1 exit status May 19 17:34:24 *home* May 19 17:37:07 re zecke May 19 17:38:05 hey May 19 17:39:40 small question, i have my first gpe-image ready compiled, i have a rootfs, everything seems fine for the first test. which bootloader is used normaly with oe for x86 systems? is there something specific which is maybe there in the package or do i have to use something on my own? May 19 17:40:52 yo zecke May 19 17:40:55 * pb__ going home as well May 19 17:41:02 * chouimat is away: food May 19 17:41:34 "I am very easy to get along with, but I don't have time to waste being nice to people who are being stupid." May 19 17:41:39 *home* already May 19 17:42:28 terminar you can use grub or lilo May 19 17:42:34 for x86 May 19 17:42:43 if you have a pc bios May 19 17:42:58 woglinde: yepp, subnotebook, sony vaio c1xn May 19 17:43:18 than take grub or lilo May 19 17:43:21 ok May 19 17:44:09 later all May 19 17:44:22 bye pb May 19 17:44:27 bye pb May 19 17:45:58 hm, i think i am really annoying but what should i configure to get the system working? i now have my gpe-image and rootfs, want to transfer it to the flash disk... on the gentoo way i had to chroot there, do grub manually, conf the network or things like that... anything to do for a bootable rootfs which i have to change? which is essential? May 19 17:46:19 change or create some bootscripts or something like that? May 19 17:46:36 i looked at the roofs and it looks quite usable May 19 17:47:02 the subnotebook has only a flash disk? May 19 17:47:06 no normal ide? May 19 17:47:54 it had... i removed the 2,5" ide and replaced it with a 2,5" ide flash drive May 19 17:48:02 it's just like ide May 19 17:48:07 hm okay May 19 17:48:18 has it dvd or cd? May 19 17:48:42 than I would boot a live cd and would copy it over net May 19 17:48:58 but I am not sure grub or lilo can read from flash disks May 19 17:49:34 if its lba addressed, then they can May 19 17:50:41 cosmicpenguin hm ist mbr than the same as in normal ide disks? May 19 17:51:00 heh - oh, yeah - need the mbr too May 19 17:51:15 woglinde: lol I like Theo de Raadt May 19 17:51:16 but if you can LBA address it, then it follows you can writ the MBR too May 19 17:51:18 nothing live or else, i had to put it direct to the ide, or with floppy boot May 19 17:51:30 woglinde: " Slackers are called slackers, people who can't read manual pages May 19 17:51:31 are called losers, and in general, calling things what they are results in developers wasting less time" May 19 17:53:21 btw essential-image, is this the smallest image with just basic boot stuff? May 19 17:56:05 hrm, ok May 19 17:56:23 * raduga just installed sashz' qvga patches into an OZ kernel May 19 17:56:37 amazingly, my zaurus hasn't bricked May 19 17:56:56 unamazingly, QVGA mode still isn't working ;) May 19 17:57:19 brb May 19 17:57:48 i imagine the X server itself, might need an update May 19 18:03:22 koen: ping? May 19 18:06:29 rp: ping? May 19 18:09:58 pb__: ping May 19 18:24:07 zecke: yo May 19 18:24:52 pb_: can I talk you into fixing the libcroco-0.6.pc... May 19 18:25:10 pb_: these - I won't say bastards - do not even care to create this file... from a .in May 19 18:25:33 zecke: er, possibly. I have no idea of what libcroco is, though, or what's wrong with it, so I may need some clues May 19 18:26:08 pb_: libcroco is a CSS2 parser used by librsvg May 19 18:26:15 ah May 19 18:26:25 pb_: and libcroco-0.6.pc puts /usr/include directly into Cflags May 19 18:26:29 no wonder I haven't heard of it, librsvg is my enemy May 19 18:26:31 pb_: without generating it... May 19 18:26:34 oh dear May 19 18:26:36 that does suck May 19 18:26:47 okay, I'll take a look at it in about an hour May 19 18:26:58 first I need to do some stuff in the garden before darkness falls May 19 18:26:59 pb_: thanks, I will have my fun with libxine... May 19 18:27:11 enjoy May 19 18:28:58 AALIB_CFLAGS='-I/usr/include' May 19 18:28:58 AALIB_CONFIG='/usr/bin/aalib-config' May 19 18:28:58 AALIB_LIBS='-L/usr/lib -laa' May 19 18:29:03 joy May 19 18:30:15 suck May 19 18:30:37 I don't know if Opie can survie without aalib May 19 18:30:49 ack....rxvt is acting strange in e-image all of a sudden May 19 18:30:51 I think I will try May 19 18:32:26 * JustinP tries gpe May 19 18:34:17 *cry* why is m4... May 19 18:34:30 ~lart autofoo May 19 18:34:31 * ibot throws autofoo's poor little doggy off a cliff May 19 18:35:05 ~300 lines to test for presence of aalib May 19 18:36:36 damn it! May 19 18:36:45 another missing dependency issue! May 19 18:36:47 JustinP: it works? May 19 18:36:52 yeah, works in gpe May 19 18:36:59 I'm guessing it's a missing font again May 19 18:37:08 did the rxvt/gpe default font change? May 19 18:37:14 dunno May 19 18:37:44 however.... May 19 18:38:02 the first time I booted GPE and started rxvt it crashed X (or matchbox, dunno which) May 19 18:38:08 login came back and further logins work May 19 18:39:55 in my e-image rxvt opens up HUGE (one char is 100x100px if not bigger) and the chars don't display May 19 18:40:05 hmm May 19 18:40:09 pb_: you have a house? May 19 18:40:25 the hard part about all of this is that since the 2.6 kernel still doesn't do RNDIS I have a real hard time debugging this stuff May 19 18:40:30 * JustinP needs a wireless card! May 19 18:40:56 JustinP: it can't do RNDIS? that sucks I remember having used it in ~2.6.12 May 19 18:41:33 no windows connectivity doesn't work for the zaurii at least May 19 18:42:11 hmm my stomach is upset May 19 18:42:51 * chouimat is back. May 19 18:45:24 justinp: what e-image are you using? May 19 18:45:27 and on what dog? May 19 18:48:39 mickey|home: "We do not ship to the Netherlands" May 19 18:49:07 koen: :( May 19 18:51:10 koen: sucky. please try another May 19 18:51:13 bbl May 19 18:51:52 mickey|dinner: I'll ask around to the people I know in the UK May 19 18:51:57 or france May 19 18:52:06 * koen gets back to the visitors now May 19 18:52:09 mmmm, pie..... May 19 18:53:34 koen: france? May 19 18:53:54 koen: you could ask Mr. Le and Avec Tux May 19 18:54:20 *yeah* xine compiles May 19 18:54:48 kergoth: pb_ : did you reply to france question, was it positively or negatively? May 19 19:03:05 raduga: heh. My own. It's basically a current oz354x image, rc4. spitz (of course) May 19 19:03:17 raduga: gpe-calendar's fonts are pretty damn large too May 19 19:03:42 I think I'm missing whatever font hack GPE has.....any idea what file/package that's in? May 19 19:08:05 libxfont? May 19 19:12:44 zecke: indeed, in the cold uk climate a house is more or less obligatory May 19 19:14:15 pb_: you could host a OE conference ;) May 19 19:14:27 pb_: BTW: personal question: do you have a wife? May 19 19:14:32 raduga: I thought it was in a config file somewhere, though.... May 19 19:14:36 (no need to answer if you don't want to) May 19 19:16:56 zecke: heh, it would need to be a very small conference if it was hosted in my house. May 19 19:17:13 zecke: no; I have a girlfriend, but we are not married May 19 19:18:46 pb_: she must have a tough life (not more tough than any others... girl whose boyfriend is a hacker) May 19 19:18:49 ;) May 19 19:19:15 zecke: indeed :-} May 19 19:20:02 zecke: could be worse :/ May 19 19:20:08 hmm, is there a way to build and run something on x86 such that you can make it crash on unaligned accesses, despite the fact that x86 allows them? May 19 19:20:16 * raduga considers 'boy whose girlfriend is a hacker' May 19 19:20:20 (for improving code portability) May 19 19:22:02 kergoth: there's a bit you can flip on x86 to enforce alignment checking, but I'm not sure if the kernel gives you a way to control it per process. May 19 19:22:16 CoreDump|afk: do you know where the font size hack for GPE sits? May 19 19:22:29 that said, I guess you could turn it on globally without too much of a risk: most programs probably don't rely on unaligned accesses May 19 19:22:43 anyone know why my OE suddenly stops in the middle of compiling with no error? May 19 19:23:20 rule #1 - nothing just stops happening with no reason May 19 19:23:24 Zero_Chaos: hard to say. maybe your power is out. May 19 19:23:51 Zero_Chaos: I vote for a power outage as well May 19 19:24:04 pb_: I just changed my local.conf to put my files in tmp/oz instead of just tmp/ and it just stops for no reason. May 19 19:24:19 I deleted everything and started over and it just keeps stopping May 19 19:24:19 # bitbake foo May 19 19:24:20 # May 19 19:24:21 ? May 19 19:24:37 been doing this a while, not a moron, but thanks for the help CosmicPenguin May 19 19:24:42 CosmicPenguin: physics? Newton'sche Axiome? May 19 19:25:01 Zero_Chaos: dmesg is showing something interesting May 19 19:25:02 Zero_Chaos: my point is - unless thats what you see on your screen, there must be some other indication then "just stopped" May 19 19:25:36 03freyther 07org.oe.dev * r27073392... 10/packages/libxine/ (libxine-fb_1.0.bb libxine_1.1.0.bb): May 19 19:25:36 packages/libxine: Do not search for aalib-test May 19 19:25:36 Do not search for aalib test tool. In the worst May 19 19:25:36 case you add -I/usr/include to the compiler line. May 19 19:25:36 For our targets we do not want aalib anyway... May 19 19:25:56 zecke: nothing interesting in dmesg... I'll pastebin next time it stops May 19 19:26:25 zero: if there is nothing interesting at all, maybe your computer is just bored May 19 19:26:46 you are all being surprisingly sarcastic... May 19 19:27:00 why is that surprising? May 19 19:27:25 Zero_Chaos: we are. But bitbake should check the exit code and fail if the code is negative May 19 19:27:55 kergoth: ah, I think this is the one. #define X86_EFLAGS_AC 0x00040000 /* Alignment Check */ May 19 19:29:44 zecke: I know it should, this is the first time I've ever seen it not tell me what happened, I just turned on debugging output, that may help May 19 19:31:17 zero: does tmp/stamps suggest anything helpful? May 19 19:31:25 at least, telling you what package failed? May 19 19:31:39 raduga: I'll let you know when it fails again May 19 19:32:26 zero_chaos: there are lots of places to look, for evidences/traces/reports of failures May 19 19:32:44 raduga: I know, it was just odd that I got no error when bitbake quit May 19 19:32:58 we can suggest some of them, but we can't read your CPU's registers, to tell you what was going on in its silicon brain, when bitbake failed May 19 19:33:28 zero_chaos: why don't you fill out a report on bugs.openembedded.org? May 19 19:33:38 raduga: ?? May 19 19:33:43 maybe you've encountered a critical flaw in the OE architecture May 19 19:34:07 Zero_Chaos: it would be nice if you could analyze this sudden stop May 19 19:34:17 Zero_Chaos: but I suspect it is a issue with your distribution or PC May 19 19:34:29 zecke: I am... I just set it again, I'll poke around once it fails May 19 19:34:43 zecke: I doubt it, I've been using bitbake for a while now May 19 19:35:04 Don't know what are your sudden stops are, but for me build breakes because it looks for -lqte, while there is only libqte-mt in staging ;-) May 19 19:35:14 That was meant to be "hi", in fact. May 19 19:35:32 polyonymous: hi May 19 19:35:51 ;-) May 19 19:36:59 _guillermo: was your pcmciautils patch applied? May 19 19:37:58 zecke: right, I started a build of libcroco. May 19 19:38:05 * pb_ goes to clean the kitchen while the compile runs May 19 19:39:36 pb_: compile is fine (do you use the zcc? Zecke's Compiler Collection?) May 19 19:40:40 pb_: ah, thanks May 19 19:41:46 What is recommended bitbake rev nowadays? May 19 19:42:15 1.4.3 looks like May 19 19:42:48 polyonymous: latest stable one :) May 19 19:43:12 zecke, meaning you have tags in svn repo or what? May 19 19:43:32 polyonymous: a branch May 19 19:43:38 aha, good May 19 19:43:41 polyonymous: the 1.4 branch ;) May 19 19:43:41 thanks May 19 19:44:17 zecke: I always use the zcc May 19 19:44:25 this one i take it svn://svn.berlios.de/bitbake/branches/bitbake-1.4/ May 19 19:47:21 woot, a quick inline asm tidbit later and i see: Bus error May 19 19:47:23 :) May 19 19:47:32 kergoth: rock May 19 19:50:10 hmm, with that set, does x86 expect word alignment, or..? :) May 19 19:50:13 * kergoth needs to learn more about x86 May 19 19:50:43 natural datatype alignment, iirc May 19 19:50:57 so, word alignment for words, halfword alignment for halfwords, ... May 19 19:51:08 ah May 19 19:52:22 task rm_work is failing ungracefully and stopping bitbake with no error May 19 19:52:27 that's my problem May 19 19:53:05 Zero_Chaos: what did you put in the local.conf? May 19 19:53:08 Zero_Chaos: INHERIT += ? May 19 19:53:29 zecke: yeah May 19 19:54:20 I'm a fscking moron... May 19 19:54:34 I had INHERIT = not +=... maybe that caused a problem... May 19 19:55:01 zecke: try http://handhelds.org/~pb/croco.patch May 19 19:59:55 opie? May 19 20:01:23 no thanks, I prefer gpe May 19 20:01:43 opie/configure.in i18n_rev("spinner") ;) May 19 20:01:55 koen: libcroco 0.6.0? can we drop that lib May 19 20:02:31 oh May 19 20:02:36 gnome.bbclass magic May 19 20:03:17 indeed, fear the gnome May 19 20:06:38 pb_: not that I understand the patch.. May 19 20:06:39 zero: actually, you're not a moron. May 19 20:06:48 you explained that clearly, before May 19 20:06:54 been doing this a while, not a moron, but thanks for the help CosmicPenguin May 19 20:06:58 pb_: is libcroco-0.6.pc generated? May 19 20:07:13 * Zero_Chaos opens the windows so he can breath, the sarcasm is so thick I'm choking May 19 20:07:30 zero: you started it with the "moron" business. May 19 20:07:31 raduga: please stop that, or do it in #cacko May 19 20:07:31 we didn't. May 19 20:07:38 zecke: yes May 19 20:08:22 zero: sorry. May 19 20:08:39 raduga: it happens, I'm not in a great mood so I'm not exactly helping. May 19 20:08:56 Zero_Chaos: you can't know, but this 'bbclass' is not officially supported May 19 20:09:34 zecke: that is good to know, does that mean I should just remove it because no one will support me, or should I submit bug reports May 19 20:09:34 zero: not your fault. May 19 20:09:42 zecke: i need to go away from irc. *sigh* :( May 19 20:10:05 Zero_Chaos: remove it for now, feel free to submit a bug report May 19 20:10:34 pb_: please explain this patch (context) May 19 20:11:04 zecke: I'll submit a bug report every time it stops until I get done building my image :-) May 19 20:11:35 Zero_Chaos: we will appreciate that May 19 20:12:02 zecke: I suppose it depends how many I submit, if I add 100 bugs, I think some people would not be as happy :-) May 19 20:12:30 Zero_Chaos: probably it is best to submit one report May 19 20:12:41 Zero_Chaos: or wait a second, ready to play cannon ball? May 19 20:12:51 zecke: ? cannon ball? May 19 20:13:02 zecke: tell me how you would like me to report and I will\ May 19 20:14:23 Zero_Chaos: test something for me May 19 20:14:41 k May 19 20:15:13 zecke: bitbake -k would solve my immediate problem... but I'd rather report the bugs so tell me what you would like me to do May 19 20:15:27 zecke: if you apply that patch to libcroco, I think it will fix your .pc problem May 19 20:15:49 yes, libcroco-0.6.pc is generated May 19 20:16:19 pb_: it fixed it May 19 20:16:26 okay, very good May 19 20:16:26 Zero_Chaos: go to rm_work.bbclass May 19 20:16:34 zecke: k May 19 20:16:35 pb_: I don't see how '' is changing that May 19 20:16:45 zecke: '' inhibits shell expansion May 19 20:17:19 you can try it at your shell prompt: echo '${PATH}' vs echo "${PATH}" May 19 20:17:35 zecke: using .oz354x FYI May 19 20:18:28 pb_: magic May 19 20:18:32 Zero_Chaos: not the issue May 19 20:18:51 * zecke is completely disturbed today... May 19 20:19:01 zecke: I know, I've opened the file, what would you like me to do? May 19 20:20:23 03freyther 07org.oe.dev * r7dfb3227... 10/packages/libcroco/ (files/croco.patch libcroco_0.6.0.bb libcroco_0.6.1.bb): May 19 20:20:23 packages/libcroco: Fix generation of libcroco*.pc to not include /usr/include May 19 20:20:23 Change autofoo to not put /usr/include into the CFlags, same May 19 20:20:23 applies to the library flags. May 19 20:20:23 GNU master phil crafted this patch - which I'm not able to May 19 20:20:26 understand... May 19 20:20:26 Zero_Chaos: don't know yet May 19 20:20:29 heh May 19 20:21:47 remove the two addtask lines May 19 20:22:12 zecke: I only have one addtask line.... addtask rm_work before do_build after do_package May 19 20:23:43 Zero_Chaos: change that from rm_work May 19 20:23:46 Zero_Chaos: to clean May 19 20:24:19 zecke: so "addtask clean before do_build after do_package"? May 19 20:24:41 yes May 19 20:25:07 k May 19 20:25:33 and file a bug report May 19 20:25:41 so we don't forget it May 19 20:25:46 zecke: file a bug report suggesting that fix? May 19 20:25:54 Zero_Chaos: no ;) May 19 20:26:08 I'll report the bugs as I find them May 19 20:27:21 the bugs... sounds bad May 19 20:27:34 Zero_Chaos: you use rm_work and every 'stop' is due rm_work May 19 20:27:47 zecke: I believe so, yes May 19 20:27:47 Zero_Chaos: this means you file exactly one bug report and mention packages that failed May 19 20:27:53 okay May 19 20:27:55 Zero_Chaos: that might help us to find the issue May 19 20:28:24 zecke: should I leave your mod to rm_work.bbclass? or revert it? May 19 20:28:35 did you try it? May 19 20:28:41 trying now... May 19 20:29:29 bugs.openembedded.org? May 19 20:29:32 yes May 19 20:33:31 pb_: feel like mailing the croco people? May 19 20:33:53 zecke: I guess you should file a bug in the gnome bugzilla May 19 20:34:28 pb_: I need an account right? May 19 20:34:33 yeah May 19 20:34:56 pb_: I fear to get kicked out of the KDE e.V. ;) May 19 20:35:06 pb_: they have zsh script that scans the bugzilla accounts May 19 20:35:10 maybe you could use a fake email address May 19 20:35:33 zecke@reciva.com May 19 20:36:08 heh, indeed May 19 20:38:13 hmmm...looks like the DPI stuff is in xserver-common.....but e is still way too large... May 19 20:50:05 heh. note to self: using mmap to process a binary file using a packed struct is quite the pain in the ass wrt alignment issues. the base offset for the mmap is such that an access of the address mmap returns is an immediate alignment fault. good fun :) May 19 20:51:46 * JustinP wishes he was that deep a hacker sometimes May 19 20:51:58 I've been forced into the upper levels mostly....high level languages May 19 20:52:04 mmap isnt deep hacking :) May 19 20:54:01 heh May 19 20:54:18 pb_: libxsettings-client include /usr/include May 19 20:55:36 ::shrug:: I haven't actually written anything with pointers in a long time. May 19 20:57:07 some one feels like fixing libxsettings-client? May 19 20:57:39 03mickeyl 07org.oe.dev * r0acfb1ee... 10/packages/directfb/ (3 files): directfb: remove 0.9.22, add directfb-examples 0.9.23, fix 0.9.25 staging May 19 20:58:11 INCLUDES = -I $(includedir) $(X_CFLAGS) May 19 20:58:24 that $(includedir) is /usr/include... May 19 20:58:52 grrr....what is with this? May 19 20:59:05 font size in everything is fine in gpe but not e May 19 20:59:48 polyonymous: which app broke because of -lqte not found? there are some not yet aware of the multithreading situation May 19 21:00:08 ~seen rwhitby May 19 21:00:21 rwhitby was last seen on IRC in channel #oe, 23h 51m 30s ago, saying: 'zecke: see PM'. May 19 21:00:45 hmm May 19 21:01:26 anyone know a canonical way to say that -native stuff should be built in a seperate directory? i want to share all the -native things, not having to build 'em all the way for every build dir i have May 19 21:02:35 pb_: where to send Xsettings patches to? May 19 21:03:46 zecke: dunno, actually. gpe@handhelds.org? May 19 21:05:10 pb_: http://pastebin.com/727241 May 19 21:06:53 oh you are maintaining it (you==GPE) May 19 21:07:56 03mickeyl 07org.oe.dev * r9a476ca1... 10/conf/distro/ (openzaurus-unstable-uclibc.conf ezx.conf): May 19 21:07:56 ezx.conf: update May 19 21:07:56 add openzaurus-unstable-uclibc distro configuration for testing... May 19 21:08:12 mickeyl: want to run uclibc on the phone? May 19 21:08:39 zecke: yes. May 19 21:08:51 mickeyl: eabi-uclibc? May 19 21:08:54 zecke: we need to save every bit of flash and ram May 19 21:09:04 zecke: that would be cool midterm May 19 21:11:44 zecke: thanks, I checked in your patch May 19 21:12:10 pb_: I would have mailed gpe any minute May 19 21:12:39 do you want a new release with that change? May 19 21:12:58 pb_: I will patch 0.17 for now? May 19 21:13:06 righto May 19 21:13:11 pb_: sadly this patch did not apply to any earlier versions :( May 19 21:13:19 version.. May 19 21:16:34 03freyther 07org.oe.dev * r8b069f0b... 10/packages/libxsettings-client/ (no-host-includes.patch libxsettings-client_0.17.bb): May 19 21:16:34 packages/libxsettings-client: Do not put -I/usr/include to the compile line May 19 21:16:34 Do not put -I$(includedir) into the Compile flags May 19 21:17:57 kergoth: do you have your OE docbook still around? May 19 21:18:05 hmm not sure, i'll look when i get home May 19 21:20:01 hmm May 19 21:20:07 gpesyncd failed :( May 19 21:20:43 drat May 19 21:20:50 what's wrong with it? May 19 21:20:51 pb_: the pot stuff... May 19 21:20:56 oh, that again? May 19 21:21:05 I thought koen said it was fixed May 19 21:21:05 pb_: somehow my SRCDATE is the 14th of may May 19 21:21:16 pb_: it is a floating cvs package :( May 19 21:21:22 doh May 19 21:21:25 hmm.. where in tmp should be autotools ? May 19 21:21:26 ~lart genext2fs May 19 21:21:26 * ibot whips genext2fs with a wet and grimy noodle just because May 19 21:21:40 ~emulate stewie May 19 21:21:42 No Sprinkles. For every sprinkle I find, I shall kill you. May 19 21:22:03 i built automake and autoconf but when i try to build some package it uses the host's autoconf May 19 21:22:26 zecke: sadly there seem to be no actual releases May 19 21:22:31 question is, where are the autotools ? May 19 21:23:23 pb_: why does this daemon need a translation anyway? May 19 21:24:24 zecke: it probably doesn't; I guess that is why florian removed the translation bits May 19 21:25:15 hrw|gone: did you push your base.bbclass change? May 19 21:25:48 i think he didn't yet May 19 21:26:00 mickeyl: would be the perfect thing to use May 19 21:26:06 right. May 19 21:26:13 made for this case May 19 21:27:20 pb_: kergoth : did you mail george? May 19 21:32:06 pb_: okay, this gcc patch can break GPE I will make it default now ;) May 19 21:45:40 i'm trying to find autotools binaries , anybody ? May 19 21:46:02 CSMan: which one? May 19 21:46:09 staging/i686-linux/bin May 19 21:46:12 autoreconf May 19 21:46:14 if you are on Linux and i686 May 19 21:47:07 yeah, they are there, and i'm using inherit autotools in the .bb May 19 21:47:17 but it's still trying to use the host's autotools May 19 21:47:19 =/ May 19 21:48:58 does that make any sense ? May 19 21:49:53 no May 19 21:50:00 CSMan: check config.log for PATH May 19 21:50:07 ok May 19 21:50:24 CSMan: and make sure the bbclass is named autotools May 19 21:50:30 it is ;) May 19 21:50:31 okay May 19 21:53:16 die e-uae die May 19 21:53:34 hi May 19 21:53:48 he coredump May 19 21:57:00 from where does bitbake gets its path info? May 19 21:57:07 *get May 19 21:57:45 CSMan: from the bbclass? May 19 21:57:54 CSMan: check the run do configure script May 19 21:58:04 in temp/ May 19 21:58:11 ij May 19 21:58:12 ok May 19 21:58:53 koen what was the solution with autoconf in debian unstable? May 19 21:59:29 * CSMan remembers May 19 22:00:19 who uses debian anyway? May 19 22:00:32 what are your workstations ? May 19 22:00:39 Ubuntu ;) May 19 22:00:59 hm the error appeared in 2003 May 19 22:01:15 woglinde: debians QA sucks ;) May 19 22:01:29 woglinde: how could that reappear? or is Debian broken since 2003? May 19 22:02:48 CoreDump|home: any idea on font size issues? The current oz3541x works fine in GPE but in e the fonts are large....way way too large in rxvt's case May 19 22:04:14 JustinP: hmm we "fixed" the fontsize by setting a xdefaults dpi value, maybe you overide that somehow? May 19 22:06:07 JustinP: /etc/X11/Xinit.d/setDPI.sh sets that value dynamically, maybe it b0rks on e17? The default is DPI=100 (which makes fonts very very small) May 19 22:07:15 zecke maybe this is the error May 19 22:07:27 its from the last package changelog May 19 22:07:29 * Some previously Debian-specific patches have been incorporated upstream, so these no longer appear in the diffs. May 19 22:08:45 /home/zecke/builds/zaurus/tmp/staging/arm-linux/include/sys/types.h:153: error: duplicate ‘unsigned’ May 19 22:08:48 ideas? May 19 22:13:48 CoreDump|home: well, gpe-calendar has fonts that are a little too big and rxvt has something like 100px x 100px chars (which don't show up) May 19 22:14:07 CoreDump|home: will try to see what the dpi is getting set to... May 19 22:14:18 JustinP: interesting, IIRC rxvt takes its settings from ~/.xsettings May 19 22:17:14 03mickeyl 07org.oe.dev * r46c1cdea... 10/ (5 files in 3 dirs): May 19 22:17:14 add a machine configuration and a kernel for the Micrel KS8695 board that May 19 22:17:14 is commonly found in PePlink manga boards and a lot of wireless media streamer boxes May 19 22:17:14 first 2.6.16 bits + defconfig. unfortunately the kernel is too large for my bootloader May 19 22:17:14 and i can't get it more slim, so i have to halt the work in this state. perhaps someone May 19 22:17:15 wants to pick that up. May 19 22:22:52 Why does fontconfig depend on fontconfig-native? May 19 22:24:35 likwise becaus you have to run native some programms to generate indexes? May 19 22:27:27 g'night May 19 22:28:00 woglinde: ok, fair enough May 19 22:49:11 gcc-cross build fails | mv: cannot stat `/.2/home/hacker/OE/build/tmp/work/armv5te-linux/gcc-cross-4.0.2-r1/image/.2/home/hacker/OE/build/tmp/cross/*/lib/libgfortran*.so.*': No such file or directory ;-) May 19 22:49:48 which machine which distro? May 19 22:50:12 oz, But I don't think it's distro/machine specific. May 19 22:50:34 polyonymous, what does df -h tell you? May 19 22:50:46 emte - a lot May 19 22:51:18 Well, there is no libgfortran, that's true. May 19 22:51:35 did you cut those lines togtyher? May 19 22:51:49 emte, what do you mean? May 19 22:52:28 i mean that is one giant path that is recursing onitself and its not moving the files anywhere May 19 22:52:57 it IS one giant path. And indeed there IS a /.2/home/hacker/OE/.... in the image directory. May 19 22:53:05 And this doesn't seem right to me. May 19 22:53:42 is this the first time youve tried to use bitbake? May 19 22:54:06 emte, Well, the first time after 1 month break. May 19 22:54:26 It used to build perfectly, but I left for a month in the mid-april. May 19 22:54:27 http://git.infradead.org/?p=users/jcrouse/geode.git;a=summary May 19 22:54:33 yay - our first public and persistant tree May 19 22:54:39 odd that .2 looks alot liek a virtual server setup ... May 19 22:54:53 OH! May 19 22:55:00 emte, /.2 is a mountpoint for a harddrive May 19 22:55:03 polyonymous, did you update bitbake? May 19 22:55:29 yes. But it failed before. May 19 22:55:50 then i suspect you have a conf error May 19 22:56:16 But then again -- a month ago it didn't fail ;-) May 19 22:56:32 Well, I'll look into it now.. May 19 22:56:40 hard to say if its due to recent changes in bitbake and the .dev build changes or not May 19 22:57:03 bitbake -v -n librsvg | grep current | sed 's/NOTE: current path: //' May 19 22:57:17 Well, first I thought that recent .dev requires recent bitabake, but it looks like the problem is elsewhere. May 19 22:57:32 polyonymous, it does May 19 22:57:48 but there may have been some other changes as well May 19 22:57:49 polyonymous: what issue? May 19 22:57:58 But I did update bitbake. May 19 22:58:14 zecke, gcc-cross build fails | mv: cannot stat `/.2/home/hacker/OE/build/tmp/work/armv5te-linux/gcc-cross-4.0.2-r1/image/.2/home/hacker/OE/build/tmp/cross/*/lib/libgfortran*.so.*': No such file or directory May 19 22:58:17 polyonymous: 1.4 branch or head? May 19 22:58:21 1.4 May 19 22:59:00 I'm not really sure it's bitbake problem... May 19 22:59:07 polyonymous: it is not a BitBake issue ;) May 19 22:59:20 polyonymous: BitBake is bug free nowadays May 19 22:59:33 polyonymous: Looks to me like libgfortran has a path somewhere missing an initial / May 19 22:59:55 i dont think libfortran is the issue May 19 23:00:02 NA|Zzz, find in the 'cross' doesn't give anything. May 19 23:00:19 he is getting a recurrsivly expanded path and he is not moving the file to a destination May 19 23:00:43 polyonymous: Grep around in the tmp/work/..../libgfortran for libgfortran\*\.so\.\*? That's the only thing that comes to mind here May 19 23:00:52 emte, what do you mean "he is not moving the file to a destination" ?There's nothing to move ;-0 May 19 23:01:26 er, tmp/work/gcc-cross May 19 23:01:29 either way the recurrsve path expansion is your problem i belive May 19 23:01:52 which seems a conf issue May 19 23:02:02 or enviroment May 19 23:02:18 Hmmm ~/OE/build/tmp/work/armv5te-linux/gcc-cross-4.0.2-r1> find . -name libgfortran\* -- gives no libraries. May 19 23:03:25 To be true, I do think it's a gcc problem... Or, maybe, it interferes somehow with whatever is installed on host machine... May 19 23:03:48 which is an enviroment problem May 19 23:04:17 the latter, yes. But the former may be a .bb problem. May 19 23:04:46 check all your bb variables May 19 23:05:23 I think I'll go read log.* May 19 23:08:13 *** This configuration is not supported in the following subdirectories: target-libada target-libgfortran target-libffi target-boehm-gc target-zlib target-libjava zlib fastjar target-libobjc May 19 23:08:30 Looks like it was not going to build libgfortran :-) May 19 23:17:51 rwhitby: do you remember the subject from your mail? May 19 23:18:59 zecke: which mail? May 19 23:19:11 rwhitby: autobuild stuff, but I found the mail in the spam filter May 19 23:19:24 "autobuild stuff", sent 19/05/2006, 7:03AM May 19 23:21:55 koen: check the manual section of OE... May 19 23:25:18 nite May 19 23:28:41 night May 19 23:47:52 Hi can I debug how/why bitbake selects a specific version of a package? May 19 23:58:30 likewise, there is nothing to debug May 19 23:58:36 you have two cases May 19 23:59:15 default it will select the highest version/cvs or it is set in your distro.conf/machine.conf/local.conf May 20 00:00:21 Well, I think it is not set in my confs, and it does not select the highest version. That's why I want to see (debug) why bitbake select a lower version. The bug could be mine. May 20 00:00:37 * likewise is rechecked the confs May 20 00:01:20 bitbake -vvv -e does not show a preferred version. May 20 00:02:25 emte: the package is glib-2.0, which has a strange number scheme because of the package name already having a version. May 20 00:02:59 emte: it picks glib-2.0_2.8.4.bb, but glib-2.0_2.10.2.bb exists as well May 20 00:03:37 read the 2.10.2 bb May 20 00:03:47 it may explicitly disallow use May 20 00:04:14 which means it would break things or its under development May 20 00:09:47 emte: OK, this is it then: DEFAULT_PREFERENCE = "-100" May 20 00:10:55 that would disallow using it May 20 00:11:23 unless by some strange freak of nature you had 100 other packages fail May 20 00:12:39 emte: thanks. I'll try bleeding edge. May 20 00:13:27 likewise, if its set like that it most likly WILL fail May 20 00:13:45 or it causes problems with the buildsystem May 20 00:14:02 ie, has hardcoded paths that mangle your host May 20 00:15:41 emte: My aim is getting librsvg up to date with the non-libart version of it (cairo based). May 20 00:15:54 emte: so I have to start moving versions up. May 20 00:16:28 likewise, i am not sure playing with glib will help in taht goal May 20 00:17:03 emte: I am not playing, there are versions dependencies between the packages that must be met. May 20 00:17:22 emte: I need glib-2.0-2.10.0+ May 20 00:18:02 you may want to talk to it's maintainer then May 20 00:18:36 to see where he is and if there is anything you should know May 20 00:22:57 Interesting....I'm seeing the infamous "horizontal blue lines" in GPE on 2.6 kernel May 20 00:24:04 which rendering mode is enabled? May 20 00:24:07 emte: OK will do, but in the meantime I try to proceed and just learn in the process. May 20 00:32:48 emte: asking me? May 20 00:33:37 CoreDump|home: Xdefaults seems to be set fine. /etc/X11/Xinit.d/50setdpi seems to be running, it says my DPI and rxvt are already ok if I run it again May 20 00:33:53 ~lart font sizes May 20 00:33:53 * ibot gives font sizes a "free" copy of Windows and then charges double for "Upgrades" May 20 00:34:01 it worked perfecly until rc4..... May 20 01:04:42 ibot, seen RP May 20 01:04:47 rp is currently on #gpe #oe #openzaurus #handhelds.org. Has said a total of 149 messages. Is idling for 1d 12h 35m 4s, last said: 'mickeyl: Agreed, a phone in OE would be great'. May 20 01:05:16 ibot, botsnack May 20 01:05:16 aw, gee, jacques May 20 01:05:26 CoreDump|home: xdpyinfo says 201x200 for bboth e and gpe as well.... May 20 01:10:48 cu May 20 02:08:08 apparently sylpheed doesn't like gpe's effects May 20 02:08:31 and i can't uncheck the effects checkbox =( May 20 04:49:15 What would be the proper way to choose the kernel in local.conf? PREF_PROV_virtual/kernel? I'm trying an collie "openzaurus-unstable" build and it is defaulting to a handhelds-sa-2.4.19-rmk6-pxa1-hh37.4-r6 kernel May 20 04:50:03 joshin: you're not setting something right, then May 20 04:50:43 joshin: you shouldn't have to set your kernel... May 20 04:50:56 joshin: you sure you set the right MACHINE and DISTRO? May 20 04:52:50 Fairly: MACHINE = "collie" DISTRO = "openzaurus-unstable" ; ASS_Prov gcc 2.95 ; Did set gcc-cross to 4.1.0 May 20 04:53:00 Typos are on purpose. :) May 20 04:53:12 why did you set gcc-cross to 4.1? May 20 04:53:23 machine of collie has: May 20 04:53:27 To see if it can be done. :) May 20 04:53:30 PREFERRED_PROVIDER_virtual/kernel ?= "collie-kernels-2.4-embedix" May 20 04:53:37 so you don't have to set it yourself May 20 04:53:47 you sure you're not including another conf somewhere? May 20 04:53:58 wait.... May 20 04:54:33 I only edited the local.conf file and set my env as the wiki directed. May 20 04:54:54 then that's probably the right kernel May 20 04:55:32 I don't think that 2.4.19 is the right kernel. I was pretty sure that collie was stuck at 2.4.18 due to the sdcard driver May 20 04:55:48 But I haven't played in this realm since before the name change (to bitbake) May 20 04:56:00 although it *looks* like openzaurus-sa_2.4.18-rmk7-pxa3-embedix20030509 should be used... May 20 04:56:22 don't know what to tell you May 20 04:56:31 perhaps bb is having some dep troubles again May 20 04:56:48 try setting the PREF_PROV above in your local.conf May 20 04:56:57 what version of bitbake are you using? May 20 04:57:09 Thanks. I wasn't sure and it is good having a second set of eyes. May 20 04:57:13 The most recent. May 20 04:57:43 most recent meaning what? May 20 04:57:53 downloaded tarball? May 20 04:57:55 trunk? May 20 04:57:58 1.4 branch? May 20 04:58:15 Meaning BitBake Build Tool Core version 1.4.3, bitbake version 1.4.3 May 20 04:59:05 and how did you get it? May 20 04:59:22 svn checkout. Followed the directions in the wiki May 20 04:59:29 ok May 20 04:59:35 that should be ok then May 20 04:59:41 Nothing funky. :) May 20 04:59:43 try the setting above without the ? May 20 05:02:27 Will this build the various collie kernels? 64/0 -> 32/32 and so on? May 20 05:03:25 Hmm, still wanting to build 2.4.19 May 20 05:11:43 joshin: try building the specific kernel then May 20 05:14:39 Will do. But in the morning for me. Thanks for the help. I'm quite curious if I can get a gcc4.1 image built and if there is a significant performance difference. May 20 07:36:17 03koen 07org.oe.dev * re78de410... 10/packages/ (3 files in 2 dirs): croco, xsettingsclient: zecke forgot to bump PR May 20 08:15:06 morning May 20 08:15:29 morning cyrilRomain May 20 08:39:18 JustinP: xdpyinfo is correct since we only change the font's DPI setting, X itself runs w/ 200 DPI. The font DPI should be set to ~160 by setDPI, is that the case for you? May 20 09:34:20 hi all May 20 09:37:33 can anyone put JavaScriptCore_cvs.sourceforge.net__20050430.tar.gz in oesources? I can't get it from sf May 20 09:38:33 cvs [checkout aborted]: connect to cvs.sourceforge.net(66.35.250.207):2401 failed: No route to host May 20 09:38:40 http://familiar.handhelds.org/source/v0.8.4-rc2/sources/JavaScriptCore_cvs.sourceforge.net__20050430.tar.gz May 20 09:38:48 or, presumably, the equivalent openzaurus url May 20 09:56:17 nice **** ENDING LOGGING AT Sat May 20 09:59:57 2006