**** BEGIN LOGGING AT Fri Mar 30 02:59:56 2007 Mar 30 04:47:12 hi * Mar 30 04:47:55 I get "This is probably a circular dependency (the chain might be printed above)." error message when try to build my images, Mar 30 04:48:36 anyone have suggestion for fix this problem ? Mar 30 06:33:52 nm: you have to look & understand the circular dependency and then break it Mar 30 06:34:15 nm: maybe you paste your whole trace to http://pastebin.ca Mar 30 06:35:34 Laibsch: qte is Qt/Embedded and qte-mt is Qt/Embedded compiled with --enable-threads. It bigger and 0.2% slower but has QThread, semaphores and so Mar 30 06:42:12 good morning all Mar 30 06:46:24 hey Mar 30 06:46:33 where can i get the original sharp rom? Mar 30 06:46:39 or cacko, pdaxrom etc Mar 30 06:58:07 from the intarweb Mar 30 07:14:14 heh Mar 30 07:14:16 I wanted to get some coffee from the vending machines Mar 30 07:14:26 so I walk to the first one in sight Mar 30 07:14:36 which turns out to be an SGI Altix 3200 :) Mar 30 07:15:01 Still no coffee, though Mar 30 07:26:20 schurig: Here is my bitbake output when using bitbake -Dv command http://paste.stgraber.org/329 Mar 30 07:57:42 schurig: Thanks for the info. Under what circumstances will I need QThread, semaphores, etc.? Mar 30 07:58:13 qte compiles for angstrom for me, but not qte-mt and qte-mt-static Mar 30 07:59:56 pH5: xserver-kdrive-fbdev.ipk does Depends on libts over here, but I haven't tested it yet Mar 30 08:00:07 I guess an error of "cannot find -lqte-mt" at compile time means I need qte-mt? Mar 30 08:00:09 pH5: could it be the same xcalibrate badness we saw before? Mar 30 08:02:00 koen: probably. it's either a linking problem or a missing AC_DEFINE. the strange thing is that I had it working yesterday while I created the patches (in the old rootfs). something didn't make it out of that workdir. Mar 30 08:16:24 morning Mar 30 08:16:40 morning Mar 30 08:17:20 Laibsch: no, that mean that you did not choosen proper qt in config Mar 30 08:19:51 morning all Mar 30 08:26:50 Hi! Mar 30 08:27:20 koen: I'm confused. I just bitbake -cclean and bitbake -ccompiled the xserver-kdrive that is in oe.dev and it now has the Tslib symbols in the Xfbdev binary. will test... Mar 30 08:27:45 pH5: so, what's it all about? is angstrom on x R7.2 now? what about that epoch stuff? is bb1.8 required now for angstrom, or? Mar 30 08:28:38 pH5: I hope it's not something simple like we had tslib on by default in our patches, like mainline need --enable it Mar 30 08:29:34 psokolovsky_: Xorg R7.2 and full libx11 at the moment Mar 30 08:31:14 koen: so what about min bitbake version? I built with 1.6 and it kinda R7.2, but of course I cannot be sure it's not weird mix of old and new packages Mar 30 08:31:43 (boote din qemu only so far) Mar 30 08:32:22 psokolovsky_: preferred-versions take care of that. Mar 30 08:32:37 epoch support is only needed if your distro doesn't select the correct packages. Mar 30 08:33:04 pH5: in other words: "bb1.6 is still good", right? Mar 30 08:33:41 psokolovsky_: right. epoch support in bitbake is only needed to make it consider epoch when comparing versions. Mar 30 08:33:53 ok, thanks Mar 30 08:34:31 bitbake 1.6 thinks that X11R7.1-1.1.0 > 1:1.2.0, bitbake 1.9 knows better. Mar 30 08:37:07 morning all Mar 30 08:38:18 * XorA is amused that on oe-issues there is IE7 download button that leads to cyberbutt.com :-) Mar 30 08:46:09 ~change 399 gbp to eur Mar 30 08:46:16 399.00 Pound (GBP) makes 586.085 Euro (EUR) (from http://www.xe.com/) Mar 30 08:46:32 hrw: what yah buying? Mar 30 08:52:08 hrw: Overall config or bb file for nunome? Mar 30 08:52:18 XorA: buttbake? Mar 30 08:52:47 ok, sent RFC for #2061, and other ipkg-utils issues Mar 30 08:52:51 !oebug 2061 Mar 30 08:52:52 * * Bug 2061, Status: NEW, Created: 2007-03-29 05:42 Mar 30 08:52:53 * * pmiscml(AT)gmail.com: Fix broken version comparison in ipkg-make-index Mar 30 08:52:54 * * http://bugs.openembedded.org/show_bug.cgi?id=2061 Mar 30 08:53:02 good morning all Mar 30 08:53:53 koen: buttbaking sounds like something that should only be done between close male friends :-) Mar 30 08:55:06 XorA: Thanks, I need a new keyboard now :) Mar 30 08:57:27 hi all Mar 30 08:58:58 morning all !!! Mar 30 08:59:25 hi florian, psokolovsky_, koen Mar 30 08:59:30 and XorA Mar 30 08:59:38 morning Mar 30 08:59:39 Laibsch: overall Mar 30 08:59:52 XorA: nothing - checked price for terrier for someone Mar 30 08:59:53 hi gremlin[it] Mar 30 09:00:45 psokolovsky_, yestaday evening i build last Angstrom x11 image for h3600 ... the main problem is size ... the image fit exactly in flash but there isn't free space for creation of file in the configuration stage :( ... Mar 30 09:01:02 gremlin[it]: damn ;-) Mar 30 09:02:07 well, there are 2 main bloating things ... libc+locales and gtk+sublibs ... Mar 30 09:02:13 well, I dunno how to handle this scalably. Apparently, just more effort should be put into console image, and let user install what they need on top of that. Mar 30 09:02:29 03xora 07org.oe.dev * r35ed9509... 10/ (6 files in 3 dirs): Mar 30 09:02:29 xournal_0.3.3.bb : add new version and add .desktop file so it shows up as an Mar 30 09:02:29 icon on desktop. Mar 30 09:02:29 remove old versions Mar 30 09:02:38 gremlin[it]: there should not be any locales for h3600 any longer ;-) Mar 30 09:03:18 on last image i had en_US en_GP and pt_BR ... i deleted manually Mar 30 09:03:49 gremlin[it]: well... are you sure it's not you who requested them to go into image? Mar 30 09:04:30 IMAGE_LINGUAS = '${@base_less_or_equal("ROOT_FLASH_SIZE", "16", "", "en-gb", d)}' Mar 30 09:05:09 i.e angstrom should have en-gb by default at all, and no for devices with 16mb flash or less Mar 30 09:05:25 psokolovsky_: R7.2 also doesn't have a diet-x11 atm Mar 30 09:05:38 ok Mar 30 09:05:40 i can try uglibc and/or talk to gpe guys and understand if is possible to shrink gtk maybe cutting cairo or something else Mar 30 09:06:24 psokolovsky_, happen olny on image build yerstarday evening the one of a couple of weeks ago was smaller ... Mar 30 09:07:06 xorg 7.2 seem is smaller than 7.1 ? Mar 30 09:08:48 gremlin[it]: impossible ;-D Mar 30 09:09:08 gremlin[it]: so as Koen points, it must be larger for now, until diet-x11 for it arrives Mar 30 09:09:23 sigh ... Mar 30 09:11:30 there is news on the uclibc front ? Mar 30 09:37:24 koen|lab: ouch, XCALIBRATE is enabled, but the XCALIBRATE definition in include/some-config.h is missing. Mar 30 09:40:33 03pH5 07org.oe.dev * r555be75c... 10/ (2 files in 2 dirs): xserver-kdrive-1.2.0: fix enable-xcalibrate.patch Mar 30 09:40:48 hi, all! Mar 30 09:40:59 !oebug 6064 Mar 30 09:41:01 * * Bug 6064, Status: NotFound Mar 30 09:41:08 !oebug 2064 Mar 30 09:41:09 * * Bug 2064, Status: NEW, Created: 2007-03-30 02:45 Mar 30 09:41:10 * * slapinid(AT)gmail.com: Angstrom build failure Mar 30 09:41:11 * * http://bugs.openembedded.org/show_bug.cgi?id=2064 Mar 30 09:41:21 any ideas? Mar 30 09:41:50 * XorA wonders why 8250.ko is borken Mar 30 09:47:09 slapin_nb: I will close it as invalid and will wait until you learn raporting bugs Mar 30 09:48:24 hrw, please add appropriate comment into it. Mar 30 09:48:51 added Mar 30 09:48:53 hrw, most of my bugs was valid. no? Mar 30 09:48:56 pH5: don't forget to bump PR Mar 30 09:49:15 slapin_nb: this one maybe is valid but is unreadable Mar 30 09:49:49 hrw, why gtk-dev should be on host? Mar 30 09:50:28 hrw, IMHO it is wrong since OE provides GTK. Mar 30 09:50:42 because our gtk lovers refuse to add gtk2-native Mar 30 09:51:52 hrw, I see, thanks. Mar 30 09:56:16 hrw: I don't refuse it, but pb_ called it 'nonsense' Mar 30 09:59:39 03koen 07org.oe.dev * rb337872d... 10/ (1 packages/xorg-xserver/xserver-kdrive_1.2.0.bb): kdrive 1.2.0: bump PR due to patch changes Mar 30 10:03:59 hrw: I usually add some ASSUME_PROVIDES += for packages that I have locally installed, reduces build time slightly. Why compile fakeroot/file/gettext/bison/zlib/...-native when I have it installed anyway? Mar 30 10:04:53 Laibsch: you need QThread etc when your app needs it Mar 30 10:04:55 because sometimes you need a special OE version Mar 30 10:05:20 Laibsch: for example, my app needs it for some background tasks Mar 30 10:05:27 koen|lab_: compilation fails again :/ Mar 30 10:05:44 during Xsdl linking xcalibrate doesn't find the tslib symbols Mar 30 10:05:56 Laibsch: basically I would only use qte-mt, nothing else, EXCEPT my device is really space restricted and I fight for 200 kB on the flash Mar 30 10:07:40 Laibsch: yes, "cannot find -lqte-mt" means that it cannot find this threads enabled Qt library. Maybe the app doesn't use QThread & friends, then a CONFIG -= threads (or CONFIG -= thread) might "fix" this Mar 30 10:12:30 03florian 07org.oe.dev * r427729cc... 10/ (3 files in 3 dirs): linux-hackndev: Update palmtx config Mar 30 10:19:50 03florian 07org.oe.dev * radd776e2... 10/ (1 packages/xorg-lib/libxcomposite_X11R7.1-0.3.bb): libxcompoisite: Depends on libxext Mar 30 10:20:12 RP: pxa-serial-hack seems to totally break 8250 serial now Mar 30 10:20:53 koen|lab_: then add it to requiredsoftware? Mar 30 10:21:51 hrw: I'm waiting for people hitting that problem to do that Mar 30 10:21:58 Mar 30 10:25:49 Who is the master of the RDEPENDS magic? Mar 30 10:42:15 ~lart linux bluetooth support Mar 30 10:42:16 * ibot hereby declares linux bluetooth support a troll Mar 30 10:58:02 03hrw 07org.oe.dev * r8c2e3e80... 10/ (1 packages/openmoko-base/openmoko-libs_svn.bb): openmoko-libs: split to have libmokoui and libmokojournal packages Mar 30 11:06:14 koen: the kdrive issue is now fixed Mar 30 11:06:44 03pH5 07org.oe.dev * r7c111b82... 10/ (2 files in 2 dirs): xserver-kdrive-1.2.0: add library order fix for xcalibrate Mar 30 11:06:45 next problem: my angstrom-x11-image included matchbox-panel-2 instead of matchbox-panel. is that intended? Mar 30 11:06:58 no Mar 30 11:07:10 matchbox-*-2 is unfinish, broken crap Mar 30 11:07:34 yeah, I see that on the neo. Mar 30 11:07:58 in openmoko we need it since all the gsm applets are mb-panel-2 applets Mar 30 11:08:15 which can't be added after the panel has started Mar 30 11:08:23 (who designs crap like that?) Mar 30 11:11:37 03koen 07org.oe.dev * rb9b6c7cd... 10/ (1 packages/xorg-xserver/xserver-kdrive_1.2.0.bb): kdrive 1.2: bump PR due to patch changes Mar 30 11:11:46 pH5: I wonder when the xcalibrate patches get applied upstream Mar 30 11:16:54 koen: whom could we talk to to advertise https://bugs.freedesktop.org/show_bug.cgi?id=10456 a bit? Mar 30 11:17:41 pH5: try daniels on #xorg or #maemo Mar 30 11:24:40 03koen 07org.oe.dev * r49ef6ebd... 10/ (1 conf/distro/include/sane-srcdates.inc): sane-src-dates: bump psplash Mar 30 11:24:46 03koen 07org.oe.dev * r3b4f52cd... 10/ (3 files in 3 dirs): sysvinit: merge psplash reboot handling from poky Mar 30 11:28:26 koen: I have built both armv4t openmoko-devel-image and armv5te angstrom-x11-image in the same angstrom builddir - now the armv5te X11 image chooses armv4t/matchbox-panel-2 over armv5te/matchbox-panel. could we perhaps drop the REPLACES from mp2 until it works at all? Mar 30 11:28:48 pH5: sounds ok Mar 30 11:32:56 03koen 07org.oe.dev * ref5e7ca2... 10/ (1 packages/openmoko-base/openmoko-libs_svn.bb): openmoko-libs: add eds-dbus dep Mar 30 11:34:04 thx koen - I just had to add it Mar 30 11:34:51 hrw: I'm still safe for another day :) Mar 30 11:34:59 * koen doesn't use SRCDATE=now Mar 30 11:38:46 ah, that reminds me to rm -rf the old DISTRO=openmoko build dir. Mar 30 11:40:50 * koen now has a Xorg R7.2 powered neo Mar 30 11:42:35 it works? Mar 30 11:43:01 yes Mar 30 11:43:06 I can even make calls :) Mar 30 11:45:08 03pH5 07org.oe.dev * r6f0db148... 10/ (1 packages/matchbox2/matchbox-panel-2_svn.bb): matchbox-panel-2: don't replace matchbox-panel, not ready yet. Mar 30 11:50:25 !oebug 2066 Mar 30 11:50:26 * * Bug 2066, Status: NEW, Created: 2007-03-30 04:56 Mar 30 11:50:27 * * dp(AT)xora.org.uk: 2.6.20 pxa-serial-hack.patch prevents 8250 loading Mar 30 11:50:28 * * http://bugs.openembedded.org/show_bug.cgi?id=2066 Mar 30 12:42:02 hrw/koen: do you want to have one of those 70 dollar atmel kits ? Mar 30 12:45:28 gerwinin: hrw has hardly any time (e.g. not for Marvell 8385 :-) Mar 30 12:45:37 hrw: btw, I got first signs of life out of my card Mar 30 12:45:59 schurig: I have this chip on some kind of wifi module Mar 30 12:46:29 gerwinin: which wifi module? Gumstix wifi? Mar 30 12:46:40 gerwinin: i have this chip on a CF card Mar 30 12:46:49 schurig : from an english manufacturer ezurio Mar 30 12:47:12 gerwinin: as schurig wrote - I'm busy with my normal work. Mar 30 12:47:16 gerwinin: there is a project that tries to mend the libertas driver (from OLPC fame) driver for the USB chip 8388 so that it would be usable for the 8388 as well Mar 30 12:47:45 gerwinin: I have friend which created own Atmel based board and now works on getting linux working on it Mar 30 12:47:51 the 8388 can do SDIO, CF and two other interface types, I'm interested in CF and prpplague is interested in SDIO Mar 30 12:48:15 schurig: I have it in a tcp/ip module , it has some gpio and I use their scripting language to read out some sensors Mar 30 12:49:02 gerwinin: do you have an URL for me? Mar 30 12:49:19 hrw: I will receive one board for free and I bought two extra for oe Mar 30 12:49:32 schurig: yes one second Mar 30 12:50:26 schurig: http://www.ezurio.co.uk/products/uw2/ Mar 30 12:50:28 s/SDIO/SPI/ Mar 30 12:50:54 gerwinin: can you give me URL to that atmel kits? I will ask friend does his want to work on OE support for them Mar 30 12:51:29 schurig: I had really problems with the analog digital converter with those modules Mar 30 12:52:39 hrw: http://www.atmel.com/dyn/products/tools_card.asp?tool_id=4102 Mar 30 12:53:13 gerwinin: nice one Mar 30 12:53:26 gerwinin: I will contact friend Mar 30 12:53:31 hrw: as I said I got one for free and I bought 2 for oe Mar 30 12:54:43 Koen: thanks for pointing me at cherokee , I came much further with my asterisk router project' Mar 30 12:55:00 gerwinin: I told him that too. I gave him simple offer: he will get kit, we will get OE support for this kit. Mar 30 12:55:59 hrw: Yeah I am planning to do this kind of thing more often because the dev kits that are bundled often really suck Mar 30 12:57:31 gerwinin: will have answer tomorrow probably Mar 30 12:58:09 hrw: did you see the imcosys smartphone ? Mar 30 12:58:28 http://www.imcosys.com/ Mar 30 12:58:43 I asked them if I buy one do I get the sdk ? Mar 30 12:58:54 They said no which is a gpl violation Mar 30 12:59:44 gerwinin: techincally SDK != source Mar 30 13:06:21 XorA: I agree with you Mar 30 13:09:49 gerwinin: interesting device Mar 30 13:10:07 but probably slow cpu and java apps only Mar 30 13:35:39 03hrw 07org.oe.dev * r844f33ec... 10/ (5 files in 3 dirs): (openmoko-)dates: fix Rconflicts/Rreplaces: they do not replace each other but conflict and only PN Mar 30 13:36:37 XorA: Totally break how? Mar 30 13:36:51 !oebug 2066 Mar 30 13:36:53 * * Bug 2066, Status: NEW, Created: 2007-03-30 04:56 Mar 30 13:36:55 * * dp(AT)xora.org.uk: 2.6.20 pxa-serial-hack.patch prevents 8250 loading Mar 30 13:36:56 * * http://bugs.openembedded.org/show_bug.cgi?id=2066 Mar 30 13:37:07 RP: thats how Mar 30 13:37:46 XorA: ah :/ Mar 30 13:38:11 RP: I have turedn off PXA_SERIAL and removed the patch Mar 30 13:38:16 RP: locally Mar 30 13:39:09 I'll need to look at the code again and see what we need to tweak... Mar 30 13:39:36 RP: bonus is now when dmesg says ttyS0 it really is ttyS0 and not ttyS3 :-) Mar 30 13:42:44 hmm, something on the ml from a motorola.com address Mar 30 13:51:53 03pfalcon 07org.oe.dev * rbdc8c1d8... 10/ (1 conf/distro/include/preferred-opie-versions-1.2.3-pre.inc): preferred-opie-versions-1.2.3-pre.inc: Bump to cvs20070328. Mar 30 13:51:58 03pfalcon 07org.oe.dev * r31ae6032... 10/ (1 packages/libopie/libopie2_cvs.bb): libopie2 cvs: Deactivate patches applied upstream. Mar 30 13:56:56 psokolovsky_: can you take care of #2050? Mar 30 13:56:59 !oebug 2050 Mar 30 13:57:01 * * Bug 2050, Status: NEW, Created: 2007-03-27 03:48 Mar 30 13:57:02 * * begemot999(AT)gmail.com: libopie2 QA pass notes Mar 30 13:57:03 * * http://bugs.openembedded.org/show_bug.cgi?id=2050 Mar 30 13:57:44 hrw: yep Mar 30 14:00:28 thx Mar 30 14:05:30 03pH5 07org.oe.dev * r1a9df9eb... 10/ (1 packages/gsm/files/magician/default): gsmd: add line discipline switch for magician Mar 30 14:05:34 03pH5 07org.oe.dev * ra3d926e3... 10/ (1 packages/gsm/files/gsmd): gsmd: don't fail if GSM_POW is not set Mar 30 14:05:38 03pH5 07org.oe.dev * re20a695e... 10/ (1 packages/gsm/libgsmd_svn.bb): gsmd: bump PR Mar 30 14:08:05 argh, I broke avahi-daemon with the autoipd postinst scripts. Mar 30 14:12:13 strange Mar 30 14:12:31 ah, not strange Mar 30 14:13:01 * XorA sucessfully upgrades to all the new X11 stuff Mar 30 14:17:53 XorA: does keylaunch work for you? Mar 30 14:18:15 pH5: yes, Address/Calednar are backwrds but it works Mar 30 14:19:08 ok, so the issue is just with our XF86... key definitions on portrait pdas. Mar 30 14:23:28 03pH5 07org.oe.dev * r43bccbbd... 10/ (1 packages/avahi/avahi.inc): Mar 30 14:23:28 avahi: fix postinst scripts so that they don't break when Mar 30 14:23:28 installing avahi-autoipd before avahi-daemon Mar 30 14:31:43 Good afternoon Mar 30 14:31:49 hey sirfred Mar 30 14:32:16 XorA: hey Mar 30 14:34:44 Anyone playing with gstreamer? It seems that the audioresample plugin is buggy. Mar 30 14:36:47 hi sirfred Mar 30 14:42:11 hi all Mar 30 14:42:27 Hi, mrdata Mar 30 14:42:27 mr: hi mr_nice Mar 30 14:42:40 sirfred: hi Mar 30 14:45:55 mr: how is your progress of expanding the struct platform_device Mar 30 14:58:44 Hey folks, please remind me, what magic trick I should use to not include empty .debug dirs in packages? Mar 30 15:00:32 mrdata: hi Mar 30 15:00:50 mr_nice: hi too Mar 30 15:01:46 mr_nice: how is your progress of expanding the struct platform_device? Mar 30 15:02:03 mrdata: not good. I was lazy :/ Mar 30 15:03:11 mrdata: were you able to use the debugger? Mar 30 15:04:08 mr_nice: not for now Mar 30 15:04:26 mrdata: after playing a bit with the printks I found an old C script. And I thought I need a bit more C skills so I have started to rewrite it. Mar 30 15:04:39 mrdata: the first part of it is already on sf.net Mar 30 15:05:06 mr_nice: i will read it now Mar 30 15:05:13 mrdata: I hope I will be able to complete it today and then I will have more skills to help you on the debugger. Mar 30 15:05:41 mrdata: it is not complete for now. but it would be nice if you can look at it :) Mar 30 15:06:23 mr_nice: the debugger can wait, when we find the bug we have with printk's Mar 30 15:07:51 bah Mar 30 15:07:57 mrdata: with the programm you will be able to write fast on your simpat with the help of the keybuttons :). the idea is from thumbscript.com Mar 30 15:08:01 off by one error in fft window calculation Mar 30 15:08:23 so noise calculations were giving bogus values due to spectral leakage Mar 30 15:08:26 but... Mar 30 15:08:29 it's 5pm! Mar 30 15:08:35 time for weekend!! Mar 30 15:09:03 Hi, koen|lab Mar 30 15:09:37 mr_nice: this is for your power-button driver, right? Mar 30 15:09:37 And good bye, koen Mar 30 15:12:31 mrdata: it will be a userspace driver for the keypad. I will read out the keypresses from evdev and then it calculates numbers out of that e.g left=3 up =1 left+up =1. Mar 30 15:12:57 03justinp 07org.oe.dev * raeecd4c0... 10/ (1 conf/distro/openzaurus-3.5.5.conf): openzaurus-3.5.5: add PREFERRED_VERSION_gtk-engines ?= 2.6.5 as 2.7.* won't work with the gtk version built Mar 30 15:13:08 mrdata: after that it will look up the keycombination and generates keystrokes to a virtual udev keyboard Mar 30 15:13:48 mr_nice: like sms system, but only with 6 keys and there combinations Mar 30 15:13:56 mrdata: 9 keys Mar 30 15:14:15 mrdata: up+left = 1 on your mobile Mar 30 15:14:30 mrdata: up+left+right+down = 5 Mar 30 15:14:42 mrdata: look at thumbscript.com Mar 30 15:14:55 mrdata: there is a jpg wich shows the definitions Mar 30 15:15:06 mr_nice: i will do Mar 30 15:15:35 mrdata: it is quite cool you will need two keypresses for one key. I have allready had a dirty version of it which worked well. Mar 30 15:16:05 mrdata: but it is verry ugly the new one is better ! but not complete atm. Mar 30 15:16:19 mrdata: I think I can manage to complete it today or tomorrow. Mar 30 15:18:17 mrdata: I also plan to create a gtk+ applikation which uses the lib. so you will be able to use it with the stylus. but I am not sure if this is a good idea. I have to think about it a second. Mar 30 15:19:35 Great, liboil-0.3.11 fixes the problem, I have gst plugin audioresample working Mar 30 15:21:56 mr_nice: think about it a second would be good, when we have a touchscreen and a virtual keyboard, why we should use ours hardware keys (6 usable) Mar 30 15:23:15 mr_nice: we have a big screen, so finger tips should are realy quicker, i think Mar 30 15:24:33 mrdata: the idea is that you can switch the behaviour of the buttons. with up+left the keys you press on the buttons will be bypassed to the virtual keyboard. after you are pressing "5" four times you will get back to the thumbscript behaviour Mar 30 15:24:48 mrdata: thats right Mar 30 15:25:01 mrdata: I will write such an ap I think :) Mar 30 15:25:07 mr_nice: when it will work, okay Mar 30 15:25:45 mr_nice: but our hardware keys are not made for power use Mar 30 15:26:36 mrdata: the power button is controlled via the first gpio of sa1100 Mar 30 15:26:37 mr_nice: she could be damaged, release a button faild sometimes by me Mar 30 15:26:42 I've just submitted bug 2067, with a new recipe for liboil 0.3.11, avoiding a SIGILL in gst-plugin-audioresample from gst-plugins-base 0.10.11. If someone is interested. Mar 30 15:27:00 bah Mar 30 15:27:10 sirfred: thx, will look Mar 30 15:27:10 I started to rain the moment I stepped outside Mar 30 15:27:20 gst-launch filesrc location=test.mp3 ! mad ! audioconvert ! audioresample ! esdsink is working now Mar 30 15:27:32 But starling refuses Mar 30 15:27:36 mr_nice: ups, i have no idea, but will inform me quickly Mar 30 15:27:41 mrdata: zecke hinted me that there is a new general keyboard driver for that gpio. so we do not need to write a driver we only need to fill out a struct right. Mar 30 15:27:43 koen: Also raining here, I wanted to go for a walk. Mar 30 15:28:25 mr_nice: i have follow the talk and this should be enought Mar 30 15:29:04 mrdata: it is on the latest linus git tree. I will backport it. or we will wait for 2.6.20.4 Mar 30 15:29:42 hrw: Trying to get a oz 3.5.5 image, I also hit bug 2038 Mar 30 15:29:46 mrdata: I think waiting will be the better choice. Mar 30 15:29:51 !oebug 2038 Mar 30 15:29:51 hrw: I don't know if the fix is acceptable Mar 30 15:29:52 * * Bug 2038, Status: NEW, Created: 2007-03-23 16:38 Mar 30 15:29:53 * * manuel.teira(AT)telefonica.net: gtk-webcore (osb-nrcore_0.5.0.bb) doesn't compile against gtk+2.6.10 Mar 30 15:29:54 * * http://bugs.openembedded.org/show_bug.cgi?id=2038 Mar 30 15:30:02 mr_nice: when backport is not big work, you should try it Mar 30 15:30:20 mrdata: 2.6.20.4 or rather for 2.6.21? Mar 30 15:30:29 03manuel.teira 07org.oe.dev * r7922bd12... 10/ (1 packages/liboil/liboil_0.3.11.bb): liboil: upgrade to .11 to avoid sigill on arm, closes #2067 Mar 30 15:30:52 mr_nice: that is a good question, in 2.6.21 also apm stuff will cahnged Mar 30 15:31:32 mr_nice: apm-emulation would be used there, when i'am right Mar 30 15:31:53 mrdata: ARM has only apm emulation Mar 30 15:32:03 mrdata: hm, maybee we should use the git code then? Mar 30 15:32:34 hrw: Thanks Mar 30 15:32:43 hrw: i know, but i have read, that also mips had its on, arm also Mar 30 15:33:37 hrw: so it would be inserted as a function inside the normal kernel Mar 30 15:33:59 hrw: one emulation, usable for all Mar 30 15:34:46 hrw: Who should change bugs to CLOSED, the reporter? Mar 30 15:35:25 sirfred: or the one who pushed fix Mar 30 15:36:16 hrw: I have five or six reported bugs, that are FIXED but not CLOSED. I suppose I should close them, as they're verified to be fixed. Mar 30 15:36:17 mr_nice: i'am not sure, we should made it step by step, but i will first look at 2.6.20.4 Mar 30 15:37:11 ah. Mar 30 15:37:33 sirfred: that.. I always forgot to close bugs after making as FIXED Mar 30 15:37:48 sirfred: and closing is for reporter (from my understanding) Mar 30 15:38:20 mrdata: ah, 2.6.20.4 is out? yes maybee some one has fixed that apm bug :) Mar 30 15:38:26 hrw: Well, some other I reported, like new recipes, never progressed from NEW, but I don't remember who commited them now. Mar 30 15:42:12 ~lart gammu authors Mar 30 15:42:13 * ibot whips out his power stapler and staples gammu authors's foot to the floor Mar 30 15:49:55 03pfalcon 07org.oe.dev * r1161b24f... 10/ (32 files in 6 dirs): opie-taskbar: Split startup and general configuration files into own package, opie-init. Mar 30 15:56:44 03pfalcon 07org.oe.dev * r9b3bae43... 10/ (1 packages/tasks/task-opie.bb): task-opie: Include opie-init in task-opie-base. Mar 30 15:57:02 mr_nice: short compare for /drivers/pcmcia/... no changes are made in kernel 2.6.20.4 Mar 30 15:57:49 mr_nice: if bug exists in kernel, it will also be there Mar 30 15:59:57 mrdata: 2.6.xx.YY are just bugfixes. no new code Mar 30 16:01:03 hrw: thx!, so i on the wrong way, i should better look at 2.6.21-rc.. then? Mar 30 16:01:11 I think I have a good question today: I'm making a recipe for a program (cacao) that generates an offsets.h by running a program (that it provides) called genoffsets. genoffsets is cross-compiled so it cannot run on the BUILD system and offsets.h is not made. If genoffsets is made natively, offsets.h will contain values from the BUILD system instead of the HOST. Any ideas on how to handle this? Mar 30 16:01:36 mrdata: yes Mar 30 16:02:08 hrw: then i will do this now Mar 30 16:02:30 apaulsen: run genoffsets on your target and put the .h in OE Mar 30 16:02:43 apaulsen: like libart and eds-dbus do Mar 30 16:03:02 apaulsen: or qemu (isn't that how the locales are handled?) Mar 30 16:04:19 koen|away: so I'd have a hard-coded offsets.h in the recipe then? I'd like to be able to contribute this to OE when it's done and I didn't think that would be proper. Mar 30 16:04:45 I've added bug 2068 to fix a crosscompilation configure problem with gst-plugins-ugly 0.10.4 Mar 30 16:06:03 Another change I've made is not inheriting gconf in gst-plugins.inc in my local rep, but I don't know if this could break some other targets or distros. Mar 30 16:07:02 AFAIK, registering plugins in gconf repository is no longer needed in 0.10. But not sure. At least, plugins are "autodiscovered" in my zaurus without gconf Mar 30 16:09:24 gstreamer 0.10 autocheck for new plugins Mar 30 16:09:33 gconf is for one plugin Mar 30 16:11:05 hrw: But that should only be needed (inherit gconf) if we're building the gconf plugin, true? Mar 30 16:11:49 hrw, koen: drop me a hint how to get rid of empty .debug dirs in non-dbg packages ;-D Mar 30 16:11:59 hrw: As I understand, inherit gconf is just putting pre and post install script in the ipk, that eventually call gconftool-2 Mar 30 16:12:13 hrw: Or that's what I've found in gconf.bbclass Mar 30 16:12:26 psokolovsky_: move them to -dbg packages Mar 30 16:12:47 ah ok Mar 30 16:12:54 * hrw need to go out Mar 30 16:13:06 hrw: moved stuff, but empty dirs remain Mar 30 16:15:06 03hrw 07org.oe.dev * rf90c9d0f... 10/ (5 files in 3 dirs): gammu: updated to 1.10.0 Mar 30 16:15:08 apaulsen: there isn't really another way, save running under qemu Mar 30 16:16:41 Great thing, GST_DEBUG environment variable Mar 30 16:18:59 * NAiL tries to get CIA-11 to choke Mar 30 16:20:34 03nail 07org.oe.dev * r79a9394a... 10/ (9 files in 2 dirs): foonas-init: Remove Mar 30 16:20:43 03nail 07org.oe.dev * r857b2f23... 10/ (1 packages/images/foonas-image.bb): foonas-image: Remove Mar 30 16:20:57 03nail 07org.oe.dev * r6686f4fa... 10/ (1 packages/tasks/task-base.bb): task-base: Add COMBINED_FEATURES uboot and redboot, DISTRO_FEATURES raid (empty right now) Mar 30 16:21:08 03nail 07org.oe.dev * r3a496d4b... 10/ (3 files in 2 dirs): lsppc: Move linkstationppc.conf to lsppc.conf Mar 30 16:21:18 03nail 07org.oe.dev * rf5539332... 10/ (1 conf/machine/n2100.conf): n2100: Update machine features, clean up Mar 30 16:21:29 03nail 07org.oe.dev * r3b45f024... 10/ (1 conf/machine/turbostation.conf): turbostation: Add uboot-support Mar 30 16:21:31 psokolovsky_: please fix the bitbake parse errors in the opie-init stuff Mar 30 16:21:41 03nail 07org.oe.dev * r89c084cf... 10/ (1 conf/machine/lsppchd.conf conf/machine/lsppchg.conf): lsppchd/hg: Update machine support Mar 30 16:21:54 03nail 07org.oe.dev * rd4cfe19d... 10/ (1 conf/machine/include/lsppc.conf): lsppc: Remove image_device_tables Mar 30 16:22:05 03nail 07org.oe.dev * r537d7c8f... 10/ (1 conf/machine/storcenter.conf): storcenter: Add uboot machine_feature, clean up Mar 30 16:22:14 03nail 07org.oe.dev * r572f224f... 10/ (1 conf/distro/foonas.conf): foonas: Make foonas use task-base Mar 30 16:22:19 03nail 07org.oe.dev * r03de3d9b... 10/ (4 files in 2 dirs): foonas: Remove leftover files, use angstrom instead Mar 30 16:22:23 03nail 07org.oe.dev * r0f8135c6... 10/ (5 files in 2 dirs): n2100, turbostation, lsppc: Update image Mar 30 16:24:09 03koen 07org.oe.dev * rdd033c24... 10/ (1 packages/xorg-util/util-macros-native_X11R7.1-1.1.5.bb): util-macros: update -native to 1.1.5, needed for R7.2 Mar 30 16:24:47 koen: is it about ${OPIE_CVS_PV}? Mar 30 16:24:54 psokolovsky_: yes Mar 30 16:24:57 ok Mar 30 16:38:47 cu Mar 30 17:03:12 koen: wouldn't it be better to have ${PN}-dev be packaged before ${PN}? Mar 30 17:04:25 psokolovsky_: why? Mar 30 17:04:49 RP: ping Mar 30 17:05:10 koen: for example, to filter .so first and then don't bother with selecting .so.* Mar 30 17:11:06 psokolovsky_: how would that make a difference from what we have now in bitbake.conf? Mar 30 17:11:55 koen: well, I now try to extend that stuff to OPIE, and see how flaky it is ;-) Mar 30 17:12:15 bad example :) Mar 30 17:14:14 koen: well, it would be the same for any framework which uses separate install root ;-) Mar 30 17:14:36 you mean broken frameworks? Mar 30 17:14:49 why does opie need a different install root? Mar 30 17:16:43 koen: ah, it doesn't need it, it just uses it, as convenience for user, I guess Mar 30 17:17:07 what's the convenience? Mar 30 17:17:16 for the angstrom test images... whats the different between an image with .summary and without Mar 30 17:17:17 the user should never see any paths Mar 30 17:17:55 psokolovsky_: the only convenience is 'rm /opt -rf' Mar 30 17:18:17 koen: why, it can also go to other partition ;-I Mar 30 17:18:34 psokolovsky_: I strongly suspect it's a leftover from sharps inability to grasp package management Mar 30 17:19:37 koen: well, ok, maybe. worth thinking about that. for now I find it mroe annoying that opie apps doesn't have consistent "opie-" prefix ;-( Mar 30 17:19:57 right Mar 30 17:20:25 psokolovsky_: iirc both zecke and mickeyl have said /opt is pretty bogus for opie nowadays Mar 30 17:20:37 (nowadays == after softfloat change) Mar 30 17:21:04 ok. good point. I'll enqueue that. Mar 30 17:27:12 is this a known openmoko-messages bug? http://pastebin.ca/417006 I have just pulled OE Mar 30 17:30:31 goxboxlive: is it in the bugtracker? if not, it is not a know bug Mar 30 17:31:35 have no idea, havent checked Mar 30 17:35:11 psokolovsky_, koen: is unresolved symbols with tslib is known issue? Mar 30 17:35:43 is it in the bugtracker? if not, it is not a know bug Mar 30 17:36:55 koen, I just remember some discussion about it, maybe I'm wrong. I watch logs of several hours before, so just curious... Mar 30 17:37:27 ah, that was about libxcalibrate and Xsdl Mar 30 17:38:16 koen, so is it fixed now? Mar 30 17:38:46 in kdrive, yes Mar 30 17:46:48 * koen kicks CIA-11 Mar 30 17:46:48 ow Mar 30 17:48:11 03pfalcon 07org.oe.dev * r74a6b114... 10/ (3 files in 2 dirs): Mar 30 17:48:11 opie-init: Drop reference to OPIE_CVS_PV, use explicit version, these are anyway Mar 30 17:48:11 not in OPIE CVS. Mar 30 17:48:13 03Manuel 07org.oe.dev * r40243c3b... 10/ (1 packages/starling/starling_svn.bb): starling: add SVN version - close #2063 Mar 30 17:48:22 03nail 07org.oe.dev * rb744deb0... 10/ (3 files in 3 dirs): uboot-utils: Update and move to uboot/ Mar 30 17:48:29 03koen 07org.oe.dev * r5ad50158... 10/ (1 packages/dri/drm_cvs.bb): drm: remove in favour of libdrm Mar 30 17:48:36 03koen 07org.oe.dev * r6575450c... 10/ (1 packages/xorg-xserver/xorg-xserver-common.inc): xorg-xserver-common.inc: drm -> libdrm Mar 30 17:48:39 koen: how does one unstick CIA? Is there some trick, or do I have to wait? :) Mar 30 17:48:42 03koen 07org.oe.dev * rff4cec49... 10/ (3 files in 3 dirs): mplayer: add powerpc support Mar 30 17:48:49 03nail 07org.oe.dev * r08572570... 10/ (1 packages/linux/linux-linkstationppc/defconfig): linux-linkstationppc: Update defconfig Mar 30 17:50:02 NAiL: it auto unsticks after 24 hours Mar 30 17:50:13 NAiL: or when I kill the cia mtn process on oe.org Mar 30 17:50:26 aha Mar 30 17:51:52 koen: you should have it auto-unstick more often than 24 hours Mar 30 17:52:21 Zero_Chaos: why? Mar 30 17:52:27 it doesn't get stuck that ofter Mar 30 17:52:30 often* Mar 30 17:52:47 and 90% of the time it's the cia server timing out Mar 30 17:52:55 fair enough, just a suggestion Mar 30 17:53:31 mrdata: I am now "complete" with it :). There is the need to add some commandline parameter or a config file to change the fix evdev path and number. And it needs some more testing. Mar 30 17:54:04 hi mr_nice Mar 30 17:55:20 woglinde: hi Mar 30 17:55:54 03pfalcon 07org.oe.dev * rba52eaeb... 10/ (1 classes/opie.bbclass classes/palmtop.bbclass): Mar 30 17:55:54 opie.bbclass, palmtop.bbclass: Set FILES_* in palmtop.bbclass, so even opie Mar 30 17:55:54 bootstrap libraries use them. Also, to not neat-pick which files go into Mar 30 17:55:54 FILES_${PN}, just change default order of packages. Mar 30 17:55:58 03pfalcon 07org.oe.dev * re7e5983b... 10/ (4 files in 2 dirs): libopie2: Fix packaging of dev/dbg. Mar 30 17:57:06 woglinde: do you want to do some testing? I have just written this one http://slackpad.svn.sourceforge.net/viewvc/slackpad/trunk/thumbscript/ and I need some one who will test it on his simpad ;). Mar 30 17:57:35 mr_nice hm whats doing it? Mar 30 17:58:59 There seems like there is some conflicts between matchbox-panel and matchbox-panel-2 also when making a openmoko-devel-image with angstrom as distro. Mar 30 18:00:13 woglinde: it creates a virtual keyboard over uinput. it reads out the events from the buttons and translate them first into numbers "up+left" = one "up" = two. then it uses the definitions of thumbscript.com to make keycodes out of that. short it is an easy way to write quick on your simpad. Mar 30 18:00:55 woglinde: you can switch the behaviour between normal button usage and thumbscript mode Mar 30 18:01:19 hm Mar 30 18:02:34 woglinde: "pressing up left down right" will be 5 on simpad and on other devices with the enterkey in the middle it uses enter for 5 Mar 30 18:03:17 woglinde: you can also give thumbscript first a try at thumbscript.com they are havin a java applet Mar 30 18:04:34 mr_nice ah okay now I understand Mar 30 18:05:06 woglinde: yes it is not that easy to explain Mar 30 18:08:41 woglinde: the only problem atm is that you have to find out the number of the evdev and change the definition of it and recompile. I will add a config or a commandline option later. Mar 30 18:14:48 woglinde: ah and you will need to change KEY_ENTER to KEY_PROG1 in check_enter(). because of the switches implementation Mar 30 18:17:10 woglinde: If I want to create a make file for it which is able to crosscompile do I have to use autoconf? Mar 30 18:17:56 woglinde: I have never created a make file before Mar 30 18:23:32 mr_nice yes autoconf tools are good for this task Mar 30 18:26:09 03koen 07org.oe.dev * r5075b3f6... 10/ (1 packages/tasks/task-openmoko.bb): task-openmoko: remove some more matchbox-panel-1 stuff Mar 30 18:26:27 woglinde: thx. I will search the documentation Mar 30 18:26:58 goxboxlive: this should solve the matchbox-panel problem I think Mar 30 18:27:26 koen: ok, il try it out, once again thx Mar 30 18:33:33 can I define kernel modules as a dependency for a programm in oe? Mar 30 18:39:03 I have a package that needs to be built, but it is not installed into the rootfs image. Is there an example of how to do this? Mar 30 18:39:45 it is a bootloader Mar 30 18:40:06 mr_nice: you shouldn't, unless program is frontend to control module Mar 30 18:42:23 psokolovsky_: ok, less work for me :). thx Mar 30 18:44:39 mr_nice: provide two *.ipk, in one the module, the second the application + recomends/depends first one Mar 30 18:46:45 koen: still failing Mar 30 18:51:03 hmmmm Mar 30 19:04:36 re Mar 30 19:04:53 hi hrw Mar 30 19:20:00 Finally I got starling to play an mp3 file. Mar 30 19:28:23 hrw: You said my overall config has issues if I get an error of "cannot find -lqte-mt" at compile time. Where exactly do I need to look? What for and how do I fix? Mar 30 19:28:38 moment Mar 30 19:29:09 Laibsch: PALMTOP_USE_MULTITHREADED_QT = "yes" == qte-mt Mar 30 19:29:53 hrw: In local.conf? Mar 30 19:29:56 yep Mar 30 19:30:03 OK, thanks. Mar 30 19:30:24 Laibsch: distros which support opie usually set this in distro config. OZ 3.5.5 has this set to "no" Mar 30 19:31:41 How about openzaurus-unstable? Mar 30 19:32:33 psokolovsky_: Do we need to update the "opie-on-angstrom"-instructions with what hrw and I were discussing the last 5 minutes? Mar 30 19:32:41 hrw: BTW, nunome failed again even with a completely cleaned tmp. Tried spitz for openzaurus-unstable and angstrom-2007.1-oabi. Mar 30 19:33:35 I'm thinking about removing openzaurus-unstable* configs in favour of openzaurus-3.5.5 one Mar 30 19:34:03 Laibsch: http://www.linuxtogo.org/gowiki/OpieWithAngstrom . check preferred-opie-versions-1.2.3-pre.inc - it already has that Mar 30 19:34:10 or angstrom ones ;) Mar 30 19:34:14 hrw: Is it generally possible to build OZ 3.5.5? No new issues? Mar 30 19:34:35 Laibsch: I'm having compile issues with osb-nrcore Mar 30 19:35:56 psokolovsky_: Well, I did have "require conf/distro/include/preferred-opie-versions.inc". Was unobvious that I need to update that. Mar 30 19:36:24 koen: I'm all for angstrom. If it supported collie. Mar 30 19:36:38 seriously, though, *if* OZ wants to do another release angstrom has to wait for that Mar 30 19:36:53 Laibsch: it does, 2.6 with angstrom-oabi and 2.4 kernel with angstrom-legacy Mar 30 19:37:06 Laibsch: feel free to sperad the word about that page ;-). feel free to join with opie maintenance in .dev, so that line was in angstrom config by default. Mar 30 19:37:07 koen: Oh, I was unaware of that. Mar 30 19:38:13 psokolovsky_: If I had more compile power, I could do more. My router dying on me while I am 10' km away and with noone to get to it before mid of April also did not help. Mar 30 19:38:49 psokolovsky_: I will certainly spread the word about opie in angstrom whenever it comes up. Mar 30 19:38:57 ok, thanks ;-) Mar 30 19:39:20 JustinP: never heard of osb-ncore. Is that package important? Mar 30 19:40:24 Laibsch: btw, preferred-opie-versions.inc aslo has MT on, so I'm not sure why your issue %). anyway, pls try 1.2.3-pre ;-) Mar 30 19:41:39 hm,strange Mar 30 19:42:22 I kinda fixed ldd. 1 liner. Stupid GNU arrogance. Bottom line: PR bump ahead %) Mar 30 19:42:50 Laibsch: yes, it's part of the web stuff Mar 30 19:44:27 hrw: With what was being discussed, I think removing zaurus-unstable is fair. I guess it would be the best option to focus on getting angstrom to release if it supports collie well enough (including SD with angstrom-legacy, I suppose). Mar 30 19:44:37 JustinP: Are others seeing the same issues? Mar 30 19:44:55 psokolovsky_: Are you commiting your PR bump now? Mar 30 19:46:25 Laibsch: nope, testing for next several hours ;-) Mar 30 19:47:07 Laibsch: I don't know. I just started seeing it after I added the PREFERRED_VERSION for gtk-engines to get around that compile error. The osb-nrcore error is due to a gtk build problem again....I think someone updated a bunch of gtk stuff in .dev and OZ is picking up random newer versions.... Mar 30 19:47:08 well Mar 30 19:53:28 hrw, psokolovsky_: If I did "bitbake qte;bitbake $something-that-depends-on-qte-mt" would bitbake build qte-mt with angstrom and OZ? bitbake did build qte-mt at the time when nothing had been built yet, but was satisfied with qte for virtual/libqte2 after I had explicitly built qte. Mar 30 19:53:37 nope Mar 30 19:54:05 qte/qte-mt both provide qte Mar 30 19:54:21 It would not build it? Then something needs to be fixed since it obviously breaks the second instance of bitbake. Mar 30 19:54:54 hrw: Sure they both provide qte. But how do you say, bb file XY needs MT support? Mar 30 19:54:56 what in OE depend on qte-mt (other then konq/emb 2006/7)? Mar 30 19:55:07 nunome Mar 30 19:55:21 not explicitly declare and maybe it does not really depend on it per se. Mar 30 19:55:35 But compilation breaks if qte-mt is not present. Mar 30 19:55:45 then it should depend on qte-mt Mar 30 19:55:56 Laibsch: so it looks that we need qte-mt in oz 3.5.5? Mar 30 19:56:04 or qte-mt static? Mar 30 19:56:22 no Mar 30 20:00:03 hrw: For now, yes. schurig suggested that CONFIG -= threads might remove that compile time dependency on MT. Mar 30 20:00:21 yes = we need qte-mt Mar 30 20:01:28 CONFIG-=thread == qte Mar 30 20:01:36 * Laibsch needs to think about how to go on with both his local config and what to commit to OE in general for nunome and friends (that still don't f*!"§ing compile here) Mar 30 20:02:11 hrw: yes. I was trying to think how to deal with the error I have been talking about all day. Mar 30 20:02:25 and how to fix it properly. Mar 30 20:02:42 I assume this could come up for other as well. Mar 30 20:03:14 Whenever you did "bitbake qte;bitbake $something-that-depends-on-qte-mt" Mar 30 20:07:17 Laibsch: actually, there was idea to remove non-mt qte at all from .dev ;-) Mar 30 20:08:01 But as hrw said, just depend on qte-mt and let users bother to use qte-mt for all themselves ;-) Mar 30 20:10:24 hi all Mar 30 20:10:35 what is PATCHRESOLVE = "noop" param? Mar 30 20:11:14 anv: it mean 'if patch does not apply then break build' Mar 30 20:11:23 normally it will try to drop you into shell Mar 30 20:12:11 Laibsch: I'm going to try a build from scratch....I may have a newer gtk staged somehow.... Mar 30 20:12:15 ok, tnx. does it cause the freezing building? Mar 30 20:12:43 anv: if it is not set then it can Mar 30 20:14:50 hrw: thanks Mar 30 20:15:07 psokolovsky_: Well, adding DEPENDS = qte-mt to nunome just fixes that. The issue at large would still be there. I don't think adding that dependency makes much sense. Mar 30 20:16:43 hrw: Anything obviously wrong with http://rafb.net/p/SeWDRn79.txt ? icecc and tinderclient are disabled/inherited. Mar 30 20:16:47 hrw: I am trying to understand why things (i.e. nunome) build for you and not for me. Mar 30 20:17:07 disabled by NOT inheriting them. Mar 30 20:17:28 that particular line being commented out and thus not showing up in my paste. Mar 30 20:18:39 no idea Mar 30 20:20:19 hrw: can you compare (or paste) your config. There must be a difference. Mar 30 20:20:46 ok Mar 30 20:21:29 http://pastebin.ca/417318 Mar 30 20:25:18 * hrw goes to watch tv Mar 30 20:25:19 cu Mar 30 20:25:39 nite hrw Mar 30 20:29:54 good nite, hrw. Thank you. Mar 30 21:15:25 hello all Mar 30 21:17:08 hey likewise Mar 30 21:18:19 hi koen! Mar 30 21:20:55 morning likewise :) Mar 30 21:21:34 * NAiL is playing with his new turbostation Mar 30 21:21:46 new or improved? Mar 30 21:21:50 NAiL: still got a working Epia left? Mar 30 21:22:33 nope, it's dead I think. Mar 30 21:22:47 Can't afford a new one quite yet. Mar 30 21:22:47 NAiL: Your last TS was bricked. Did you swap it out for another piece? Mar 30 21:22:54 no, I bought one Mar 30 21:22:57 I still have the brick Mar 30 21:23:25 * likewise should get his EFIKA running again now that Crofton brought those cute picoPSU bricks during FOSDEM. Mar 30 21:23:25 still hoping I can get the flash done somehow. Mar 30 21:23:44 ts-101 right? Mar 30 21:24:19 yep Mar 30 21:24:27 no JTAG Mar 30 21:24:57 what about getting the n2100 fixed? :) Mar 30 21:25:01 NAiL: but possibly BDM... Mar 30 21:25:18 timtimred: shush, I'll get to it eventually :-P Mar 30 21:25:23 * likewise really needs to clean up his desk (read: house) first Mar 30 21:25:32 likewise: that needs the JTAG pins, doesn't it? Mar 30 21:26:06 NAiL: It's the Background Debug Module for Freescale processors, it's not JTAG related, but is similar in ways Mar 30 21:26:37 yeah, but I can't figure out how to actually use it Mar 30 21:27:50 ie, what pins and such Mar 30 21:28:29 NAiL: you need a BDM pod to start with, and find the BDM lines. Yup, always hard. Google doesn't help me on it either, so nobody seems to care about turbostation bricks (yet) Mar 30 21:28:39 hehe Mar 30 21:28:44 "pod"? Mar 30 21:35:10 night Mar 30 21:42:04 hi Mar 30 21:46:09 NAiL: google for "bdm pod". Dongle, if you like. Mar 30 22:37:52 are most people using glibc or uclibc in OE? Mar 30 22:38:16 I'm trying to decide which one to use.. Mar 30 22:38:33 glibc, by far Mar 30 22:38:42 afaik, uclibc is currently broken Mar 30 22:39:24 yes most uses glibc because not all software compiles with uclibc Mar 30 22:40:01 NAiL, woglinde thanks.. sounds like glibc is still the way to go. Mar 30 22:40:03 svolpe: if you don't have a very good reason to use uclibc, use glibc Mar 30 22:40:17 svolpe depends on the device Mar 30 22:40:28 but most devices now have enough space Mar 30 22:41:03 anyone know how to make just one reciepe compile against uclibc? Mar 30 22:41:17 svolpe: nslu2-linux has stopped supporting our uclibc firmware variant - it's too hard to keep it compilable in OE. Mar 30 22:41:19 I have plenty of space, I just read on the uclibc site that it was faster and thought it was worth revistibg.. Mar 30 22:42:21 and speed is something I can always use more of :-) Mar 30 22:49:49 !oebug 2069 Mar 30 22:49:50 * * Bug 2069, Status: NEW, Created: 2007-03-30 09:25 Mar 30 22:49:51 * * r.farina(AT)adelphia.net: ettercap fails to build. Mar 30 22:49:53 * * http://bugs.openembedded.org/show_bug.cgi?id=2069 Mar 30 22:50:15 this is a blocker for me if anyone can take a look Mar 30 22:50:45 03coredump2 07org.oe.dev * r00ea883c... 10/ (1 packages/altboot/altboot_wip.bb): altboot_wip.bb: Drop altboot_wip.bb Mar 30 22:50:51 03coredump2 07org.oe.dev * r302d6092... 10/ (3 files in 2 dirs): altboot: Add first installable version for nslu2 Mar 30 23:31:36 03coredump2 07org.oe.dev * r6396af30... 10/ (3 files in 2 dirs): altboot: Do no use SRCDATE for snapshots, use SVN revision instead Mar 30 23:43:25 When a processor doesnt have a MMU and uses uclinux, how does it overcome this ? Is the MMU software based? Mar 31 00:07:54 concept10: no Mar 31 00:08:07 concept10: it just does not use the MMU for memory operations Mar 31 00:08:25 likewise, okay Mar 31 00:08:30 concept10: and you lose some MMU specific features, such as swapping through page faults Mar 31 00:08:53 concept10: also, processes can write into each others memory space without any hardware protection Mar 31 00:09:14 likewise, ohhh..... thats a big deal Mar 31 00:09:37 concept10: Do you know the Amiga? Mar 31 00:10:13 likewise, I know of it and used it (but not much) I stuck with my c64 back in those days Mar 31 00:10:52 concept10: It did not have an MMU, at least it was never used. If processes are properly programmed, the MMU protection is not needed. But of course, in today's complex s/w systems, anything beyond embedded needs a MMU Mar 31 00:11:14 Within embedded, it depends upon your requirements for stability etc. Mar 31 00:11:49 anyways, gotta go catch some sleep here. good nite Mar 31 00:37:35 03pfalcon 07org.oe.dev * r5b3c0654... 10/ (3 files in 3 dirs): Mar 31 00:37:35 glibc 2.5: Fix common case of GNU arrogance, make ldd work on posix shell. Mar 31 00:37:35 * Not bumping PR, to save people trouble. Leaving for decision to OE maintainers. **** ENDING LOGGING AT Sat Mar 31 02:59:56 2007