**** BEGIN LOGGING AT Wed Apr 20 02:59:58 2016 Apr 20 06:50:15 good morning Apr 20 10:03:17 bluelightning, I suck at reading the do_rootfs log Apr 20 10:03:34 Crofton|work: ah, what did you miss? Apr 20 10:03:44 dunno Apr 20 10:04:06 I doesn't seem to explain what the problem is Apr 20 10:04:14 * Crofton|work goes for coffee Apr 20 10:06:26 ah ok Apr 20 10:06:35 well I can have a go at reading it if you like Apr 20 10:06:45 not that my comprehension is always reliable either Apr 20 10:08:03 heh Apr 20 10:10:13 https://dl.dropboxusercontent.com/u/14618236/log.do_rootfs Apr 20 10:49:40 * satisfy_dependencies_for: Cannot satisfy the following dependencies for packagegroup-core-x11-utils-dev: Apr 20 10:49:43 * x11-common (= 0.1-r47) * Apr 20 10:51:24 yeah Apr 20 10:51:35 you cutting and pasting my problem> Apr 20 10:54:56 "Not selecting xserver-common 1.34 due to unmatched constraint." Apr 20 10:55:43 "replacement_apkg=xserver-common for provider_apkg=x11-common" Apr 20 10:57:24 oh Apr 20 10:58:04 I think this is meta-oe and its xserver-common being handled poorly by opkg Apr 20 10:58:18 really time we sorted that out :/ Apr 20 10:58:36 ah Apr 20 10:58:49 but in any event opkg isn't handling the situation properly Apr 20 10:59:23 very poorly Apr 20 10:59:32 any clue what made it blow up? Apr 20 11:00:31 x11-common exists in oe-core, xserver-common RCONFLICTS/RREPLACES it - but perhaps not RPROVIDES? Apr 20 11:00:56 I looked at the code in opkg recently for a similar situation - https://bugzilla.yoctoproject.org/show_bug.cgi?id=9464 Apr 20 11:01:18 opkg favours packages whose name matches and at the moment will select them even if they conflict with another - except it won't properly install them Apr 20 11:01:30 this isn't quite what is happening in this situation but I'll bet it's related Apr 20 11:02:07 doh, we still have the conflict after years of oe-core? Apr 20 11:02:11 yes Apr 20 11:02:17 remind me why meta-oe needs the recipe> Apr 20 11:02:24 the bug to fix it is still assigned to me, and apparently nobody else cares enough to fix it Apr 20 11:02:41 you'd have to ask JaMa for a reminder Apr 20 11:02:56 I've forgotten the details to be honest Apr 20 11:03:28 iirc the one in meta-oe is heavily machine-dependent and parts should be handled in a separate BSP layer Apr 20 11:03:44 right, it's designed to be monolithic Apr 20 11:03:55 https://bugzilla.yoctoproject.org/show_bug.cgi?id=5546 Apr 20 11:04:02 which doesn't quite fit with the OE layer model Apr 20 11:04:14 yep, that's the one Apr 20 11:04:39 rburton, volunteered! Apr 20 11:04:45 he he Apr 20 11:04:52 yes, maybe I should have let him take it Apr 20 11:05:01 (s/have//) Apr 20 11:05:02 +1 :) Apr 20 11:05:07 i did? Apr 20 11:05:11 damnit Apr 20 11:05:17 it's right there in the comments ;) Apr 20 11:05:24 I also read your motivation between the lines Apr 20 11:05:51 but short term, we need to figure out what upset the applecart Apr 20 11:06:40 bluelightning's hunch sounds reasonable Apr 20 11:06:44 yeah, I'm not too sure about what that is, because I could have sworn the issue I was dealing with wasn't an issue earlier either Apr 20 11:07:08 adelcast said he would look into it in a few days I think Apr 20 11:07:41 Crofton|work: it's a long shot but you could try adding RPROVIDES_${PN} += "x11-common" to the xserver-common recipe in meta-oe Apr 20 11:07:45 since that doesn't seem to be there Apr 20 11:08:36 I'm not sure it should be but it may impact the behaviour in a positive way Apr 20 11:32:27 WTF: ERROR: libxml2-2.9.3-r0 do_package: QA Issue: split_and_strip_files: 'file /OE/build/oe-core/tmp-glibc/work/i586-oe-linux/libxml2/2.9.3-r0/package/usr/lib/libxml2/ptest/xmlconf/sun/invalid/utf16b.xml' failed [split-strip] Apr 20 11:33:11 isn't it supposed to try to strip only ELFs? Apr 20 11:41:17 hmm, something changed there recently but it shouldn't have picked up xml files Apr 20 11:42:10 d14bc2667575bc0e29b9e90d80c01703639b76f1 Apr 20 11:42:16 can't see how that would have led to this though Apr 20 11:55:39 bluelightning: thanks, I also fail to see it, the last condition still returns False on this path Apr 20 11:55:51 bluelightning: I've triggered build without ptest to pass this Apr 20 12:23:39 Does anyone know a clean way of hiding the cursor for xorg? Apr 20 12:25:18 I have seen "-nocursor", but I am not sure how or where to implement it. Apr 20 12:56:05 riz___: hiding entirely, as in you have a touchscreen? Apr 20 12:56:31 rburton: Correct Apr 20 12:57:06 then yes, you want to pass -nocursor to X when it starts Apr 20 12:57:18 I have tried xinit -- -nocursor to the Xserver script, but it did not work Apr 20 12:57:23 Is that the wrong place? Apr 20 12:57:57 i think editting /etc/X11/Xserver to pass it in ARGS will be right Apr 20 12:59:45 I found it here: /home/ryan/development/custom/openembedded-core/meta/recipes-graphics/x11-common/x11-common/etc Apr 20 13:00:15 This is the line I changed: exec xinit /etc/X11/Xsession -- -nocursor $XSERVER $DISPLAY $ARGS $* Apr 20 13:00:25 put it in args Apr 20 13:00:44 i'm surprised putting it there even starts x Apr 20 13:00:53 Isn't that what I did, or am I misunderstaing the use.. Apr 20 13:00:58 I take it I am misunderstanding lol Apr 20 13:00:59 no i mean in $ARGS Apr 20 13:01:12 its an argument to the xserver, and there's a variable called ARGS Apr 20 13:01:17 add -nocursor to ARGS Apr 20 13:01:44 Oh, so write ARGS += -nocusrsor? Apr 20 13:01:49 -nocursor Apr 20 13:02:28 nvm, I see ARGS up top Apr 20 13:02:54 rburton: I guess my second issue is that I would rather do it in a custom recipe as opposed to changing source. How is that possible? Apr 20 13:04:03 just write a bbappend to replace the Xserver file Apr 20 13:05:34 Oh ok, so that this one gets ignored completely? Apr 20 13:29:22 rburton: that didnt work. Do you think it is some sort of matchbox override? Would I have to do something else with matchbox? I am only using the terminal. Apr 20 13:35:42 riz___: i suggest fiddling on the target directly to get it working Apr 20 13:36:03 its possible that -nocursor just tells xfixes to hide the cursor and them matchbox turns it back on again Apr 20 13:36:14 Yeah, that is what I am thinking Apr 20 13:36:26 based on what I see here: http://lists.openembedded.org/pipermail/openembedded-commits/2007-June/080382.html Apr 20 13:36:34 so pass -use_cursor=no to mbwm Apr 20 13:36:37 I will try adding touchscreen to MACHINE_FEATURES Apr 20 13:36:46 If that doesnt work I will try what you suggested Apr 20 13:36:59 see i should think for a moment instead of answering questions Apr 20 13:37:06 what you want to do is write a formfactor for your machine Apr 20 13:37:14 and set HAS_TOUCHSCREEN=yes Apr 20 13:37:39 Wait, where do I put the HAS_TOUCHSCREEN=yes? Apr 20 13:38:11 in a formfactor file Apr 20 13:38:17 grep oe-core for formfactor and you'll see Apr 20 13:39:32 ok, thanks Apr 20 15:12:37 rburton: So, including HAVE_TOIUCHSCREEN=1 sort of worked. On startup there is no mouse cursor. Once I touch the screen it reappears though. Apr 20 15:18:52 Hello there Apr 20 15:18:52 I got cpu model : Broadcom BMIPS4380 V6.5 FPU V0.1 Apr 20 15:18:52 Please, can you help with info - I need compile vim editor for this device, there already pre-installed OS linux, but very limited (in busybox only) Apr 20 15:18:52 I need some guide how to cross and whet is best tool for it, I hope OE cam handle this task. Apr 20 15:18:52 is it right place for ask? Apr 20 19:07:04 leopold75: what disto is on your device now and does it provide a x-compile environment? Apr 20 19:08:49 on device linux-mipsel, there vely low flash memory Apr 20 19:08:49 I am not sure about x-compile env. there nothing available - no vim, nano, gcc -nothing Apr 20 19:08:49 right now I reading wiki.yoctoproject.org Apr 20 19:10:25 device - STB receiver for satTV. I want compile few tools, which I use on my gentoo-host Apr 20 19:43:57 leopold75: you can use MACHINE=qemumips bitbake vim Apr 20 19:44:04 and then take the binaries onto target Apr 20 19:44:27 however, this assumes that the rootfs thats preinstalled on mipsel target Apr 20 19:44:36 is generated using OE Apr 20 19:44:50 MACHINE=qemumipsel is what you would need actually Apr 20 19:47:36 ok thank You. I'll try Apr 20 20:30:16 Is there anyway disable sleep/hibernate/power management? I would like my screen to always be on. Apr 20 20:32:52 assuming that you're not running oe-core master, see oe-core 32ac0bfb0167cb350be5d4b8faca490956dd7d93 Apr 20 20:33:06 you can probably be inspired by that, or cherry-pick it, or something Apr 20 20:35:48 i'm running jethro. I'll check it out. Thanks Apr 21 02:15:07 t **** ENDING LOGGING AT Thu Apr 21 02:59:58 2016