**** BEGIN LOGGING AT Sun Jan 06 02:59:59 2013 Jan 06 03:01:13 I see that a lot from folks who use archlinux...sad because the concept is good...the elitism isn't Jan 06 03:04:04 and from what i've seen, people who actually know unix pretty well are able to get the webos build processes running in arch Jan 06 03:04:36 without being douchnozzles Jan 06 03:06:48 several have Jan 06 03:18:43 folks Jan 06 03:20:08 bhey Jan 06 03:20:22 how's it going ? Jan 06 04:39:38 gift for everyone who can use it, just plz don't tweet or post on WON >> http://developer.palm.com/appredirect/?promocode=fqhrd75806piwbf Jan 06 04:47:02 Hello everyone Jan 06 04:47:21 hey Jan 06 04:47:26 I installed WebOS on my nexus 7 and I noticed that the screen needs to be re-calibrated. Jan 06 04:47:40 I used multiboot by the way. Jan 06 04:47:49 Not sure if that matters much. Jan 06 04:50:57 How can I recalibrate the screen? Jan 06 04:53:08 Rafase282, i am not familiar with multiboot, can you send me the link? Jan 06 04:53:22 also, how do you know the screen needs re-calibration Jan 06 04:53:24 ? Jan 06 04:53:30 http://forum.xda-developers.com/showthread.php?t=2011403 Jan 06 04:54:15 Because I touch at 1x1 and it actually touch at 5x5 Jan 06 04:54:19 ah Jan 06 04:54:23 If you understand what I mean. Jan 06 04:54:55 im not sure if that is multiboot's fault or your device. Jan 06 04:54:58 Rafase282, you are off the reservation here...we aren't supporting that methot... Jan 06 04:55:08 s/methot/method/ Jan 06 04:55:09 ka6sox meant: Rafase282, you are off the reservation here...we aren't supporting that method... Jan 06 04:56:18 Well what I did was to get the .ext2 content and use the /boot from ubuntu13.04 Jan 06 04:56:26 13.04? Jan 06 04:56:30 dang, they've jumped that far? Jan 06 04:56:43 I could not build from source because my computer shutoff Jan 06 04:56:54 Yes, 13.04 is the upcoming version Jan 06 04:57:01 Two releases per year for Ubuntu Jan 06 04:57:25 but yeah, does the screen calibration has anything to do with /boot? Jan 06 04:57:29 Rafase282, that is not a method we have tested and therefore cannot support. Jan 06 04:57:36 Rafase282, possibly Jan 06 04:58:00 Well, does it works find with your method? Jan 06 04:58:07 yes Jan 06 04:58:28 I didnt build the ext2 so I don't know if that file was from an early stage or not Jan 06 04:58:45 where did you get the .ext2 from? Jan 06 04:59:04 Would you have any idea why my computer would shutoff when building from source? I have 6 cores, 16 gb of ram, 16 gb of swap, and 2 Tb of HDD Jan 06 04:59:26 no idea.... Jan 06 04:59:35 I got the image from here: http://forum.xda-developers.com/showpost.php?p=36337083&postcount=682 Jan 06 05:00:56 Also, multiboot now supports booting linux distros besides ubuntu. In case you guys are interested I could point you to the source and info. Jan 06 05:01:32 https://github.com/Tasssadar/multirom/wiki Jan 06 05:01:37 thanks.... Jan 06 05:02:27 I have tested it with Plasma active, and Bodhi Linux besides Ubuntu, and I got the webos to boot but that is still a work in progress as you have noticed. Jan 06 05:03:09 where do you get the Open webOS image from? Jan 06 05:05:04 I posted a link. Jan 06 05:05:11 But it was from that thread. Jan 06 05:05:21 I linked to the post intead of the file itself Jan 06 05:05:31 http://forum.xda-developers.com/showpost.php?p=36337083&postcount=682 Jan 06 05:07:21 hmmm Jan 06 05:08:27 Rafase282, unless you get the image from us...I guarantee nothing... Jan 06 05:09:37 and I definately will not vouch for anything uploaded to the net by someone unknown to me. Jan 06 05:12:18 hmm. my qemux86 build is stuck in this state where it refuses to build saying that it can't figure out how to supply webkit-webos Jan 06 05:12:28 Well, could you provide me with an image? Jan 06 05:12:31 even though i've completely obliterated the entire thing and rebuilt from scratch Jan 06 05:12:40 I'm still unable to build it myself. Jan 06 05:13:32 the "official" images are here: http://build.webos-ports.org/webos-ports/images/grouper/ Jan 06 05:13:36 Rafase282: shutting down while building from source? sounds like heat issues Jan 06 05:14:06 ya, OE is famous for "stress testing" systems :F Jan 06 05:14:16 s/F/D Jan 06 05:14:21 That links makes thing a lot easier. Jan 06 05:14:44 please don't "reupload" these someplace else.. Jan 06 05:14:54 point people to these please. Jan 06 05:15:44 I was about to ask if I could provide that link to other people. Jan 06 05:16:14 please do! that also helps me judge how many people are working with these images. Jan 06 05:18:30 gives me an idea of uptake. Jan 06 05:20:07 Done. Jan 06 05:20:20 Also, I'm downloading the ext4 rootfs Jan 06 05:20:48 Would you care to experiment with multiboot with me? I'm going to try it so I could very well be your tester right now. Jan 06 05:33:27 sure, I am not in a position to test currently (here) but will follow along and provide whatever support I can Jan 06 05:34:30 Well you need to read that. There is a high chance that you will understand it better than I currently do. Jan 06 05:35:22 Are you guys using the Ubuntu for nexus7 kernel? Jan 06 05:37:51 nope Jan 06 05:38:32 Rafase282, ya, we are going to need to add the kexec-hardboot patch for our stuff to work right. Jan 06 05:39:15 =/ Jan 06 05:39:36 let me read further along... Jan 06 05:39:42 The I guess I either try to solve my heating issues or wait for someone else to pathc the kernel. Jan 06 05:40:29 you got it "booting" but there are scaling issues...i suspect that we need to look a bit deeper to figure this out. Jan 06 05:41:00 give us a couple of days to poke into this a bit more...since we are getting requests to multi-boot as well. Jan 06 05:41:20 this is pretty detailed information and patches that might be able to be incorporated. Jan 06 05:41:46 no promises...but it looks interesting. Jan 06 05:43:01 Well they come from the author of multi-boot. Jan 06 05:43:15 can someone help me out and tell me what device is codenamed "Tilapia"? Jan 06 05:43:46 I have a quick question, I downloaded the ext4 file system. Would that boot fine or should I stick to the ext2 while following the guide from the project's website? Jan 06 05:44:27 I think that is the Nexus 7 with 3g Jan 06 05:44:31 Let me double check Jan 06 05:45:43 ka6sox, http://wiki.cyanogenmod.org/index.php?title=Tilapia_Info says that tilapia is nexus 7 3g Jan 06 05:45:55 I think the Rom Manager will be happier with a Ext4 image Jan 06 05:46:14 scoutcamper, thanks Jan 06 05:46:39 Rafase282, you should try both...i think that ext4 is better supported on Android platforms. Jan 06 05:47:22 Yeah ext4 is the norm at least with the new devices Jan 06 05:47:39 Rom manager supports the nexus 7? Jan 06 05:49:32 multiROM thingy Jan 06 05:49:37 seems to want a zip file Jan 06 05:49:57 lesse what kind of image is inside the zip of the Ubuntu... Jan 06 05:50:26 oh the zip file is for android roms only Jan 06 05:50:34 now it supports nandroid backups too Jan 06 05:50:55 for linux it can be a /root or root.img file Jan 06 05:51:03 the root.img file would be automatic Jan 06 05:51:06 oh its an img.gz for the ubuntu Jan 06 05:51:14 the /root would be manually Jan 06 05:53:22 http://build.webos-ports.org/webos-ports/images/grouper/webos-dev-image-grouper-20130104122504.rootfs.tar.gz Jan 06 05:53:32 that image seems closest to what you want Jan 06 05:54:47 that would be a /root image Jan 06 05:55:33 that amkes it easier for me Jan 06 05:55:45 Also, why is the /boot directory empty? Jan 06 05:55:58 because we fastboot the zImage Jan 06 05:56:21 I tried that, it does not work when you have multiboot installed Jan 06 05:56:51 Either that or it was hardcoded to linux/rootfs.ext2 Jan 06 05:59:38 hmmm...I'd have to dig into the multiROM code to figure that out Jan 06 06:00:08 so if I put the zImage in /boot I would not need the fastboot command? Jan 06 06:00:52 If you check this it would be easier instead of checking the code. http://forum.xda-developers.com/showpost.php?p=34577869&postcount=3 Jan 06 06:01:17 and about the rootfs.ext4 I was referring to the guide for this project. Jan 06 06:04:04 kk Jan 06 06:05:01 Rafase282, what TZ are you in? Jan 06 06:06:57 EST Jan 06 06:07:03 1:06 here Jan 06 06:07:04 am Jan 06 06:07:26 I'll need this files. Jan 06 06:07:26 kernel_path="%r/boot/zImage" Jan 06 06:07:26 initrd_path="%r/boot/initrd.img-ubuntu" Jan 06 06:07:53 now for the zImage should I use http://build.webos-ports.org/webos-ports/images/grouper/zImage ? Jan 06 06:08:09 PST here on the Pacific Plate... Jan 06 06:08:53 yes, that is the Zimage Jan 06 06:08:58 er zImage Jan 06 06:08:59 that would make it 3 hours less? Jan 06 06:09:15 What aboutthe initrd.img file? Jan 06 06:09:28 Should I use the ubuntu one or you guys have that? Jan 06 06:09:43 I don't think this kernel uses a initrd Jan 06 06:09:49 (Ubuntu does) Jan 06 06:11:54 Then maybe I should use the ubuntu one since yours is not patched yet right? Jan 06 06:12:22 initrd's are built against the kernel they are built for so they can load the appropriate modules. Jan 06 06:12:30 their image and ours are not the same Jan 06 06:13:01 do you know what version theirs is? Jan 06 06:13:04 hmmmm. what is this magic? https://twitter.com/Scoutcamper_1/status/287803137862037504 Jan 06 06:13:04 so if I use your kernel, I would only need the zImage? Jan 06 06:13:30 I am looking @ the CPIO files to see if they are initramfs files Jan 06 06:13:55 you probably can't use the Ubuntu Kernel as it is HF and ours is not. Jan 06 06:14:20 (at least I can't verify that until I see the kconfig for ours. Jan 06 06:25:51 Well im trying to boot with the roofs you linked me and your kernel Jan 06 06:26:02 and it didnt work Jan 06 06:26:11 I'll try with the ubuntu kernel then Jan 06 06:27:06 I doubt that will work as its a HardFloat kernel and I'm not sure ours is yet. Jan 06 06:27:17 but why not...stranger thigns have worked... Jan 06 06:27:30 (its possible that LibHybris will be unhappy too... Jan 06 06:33:05 Well it worked before Jan 06 06:40:27 good Jan 06 06:43:15 Trying again. I forgot to configure the rom_info.txt file Jan 06 06:44:13 ah Jan 06 06:44:27 Well it boots now, same as before, screen calibration is borked Jan 06 06:44:44 It should be because I had to use Ubuntu kernel to get it to boot Jan 06 06:44:45 well...I know that rootfs works Jan 06 06:45:01 so I suspect the calibration stuff is off because of the kernel being used by Ubuntu Jan 06 06:45:03 So that means with a patched kernel it should work fine. Jan 06 06:45:07 more or son Jan 06 06:45:17 yup Jan 06 06:45:24 That's my conclusion too. Jan 06 06:46:07 come by another time (or feel free to hang out) and I'm sure others will be interested in this as well as many have expressed a desire to multi-boot Jan 06 06:48:14 It would have to be tomorrow or during daylight I guess. Jan 06 06:48:32 I need someone to patch the webos kernel. Jan 06 07:05:27 Rafase282, yup...tommorrow or monday...(we have folks in .eu as well who poke at these things while we sleep) Jan 06 07:06:42 patch the kernel? Jan 06 07:06:50 kexec patches Jan 06 07:07:04 we're all just ubuntu users, we don't know anything about kernel stuff . . . . (reference to guy who was here earlier) :D Jan 06 07:07:42 I see... Jan 06 07:08:33 I think more than one of us could do this...I just know as tired as I am from poking into about 9 pies today to prep for more releases and builds I'll muck it up. Jan 06 07:15:12 Take your time. I wouldn't try to rush you. Jan 06 07:15:55 For the ubuntu users here, what is the easiest way to get my hardware specs into a txt file? Jan 06 07:16:05 lshw Jan 06 07:16:28 sudo apt-get install lshw Jan 06 07:16:28 chmod fail: "Value too large for defined data type" .. wth does -that- mean Jan 06 07:16:39 I want to check on why my pc shutdown while building from source. Until I figure that out I cant even try to patch the kernel myself. Jan 06 07:16:52 right Jan 06 07:17:07 lshw > /tmp/specs.file Jan 06 07:17:21 then post it to a pastebin Jan 06 07:17:49 folks, I've sucessfully built tuna image on arch Jan 06 07:18:01 panda-z, w00t! Jan 06 07:18:22 arch-luna-tuna Jan 06 07:18:23 so it *is* possible! (as we all knew it could be) :D Jan 06 07:19:44 ka6sox: yes, it *is* possible, and zsh shell works well w/OE Jan 06 07:20:13 good to know that too...dash is a failure Jan 06 07:20:25 http://pastebin.com/9dVGhy6L here Jan 06 07:20:44 according to the gnu.org manual, this error means that my system is incapable of chmodding a file larger than 2 gigs. Jan 06 07:21:06 What would you guys suggest? disabling core unlock and whatever overclock there might be from the bios? Jan 06 07:21:34 Rafase282, those specs look fine. i submit that your machine is overheating and shutting down to save itself from melting Jan 06 07:21:52 Rafase282: i'd get a temperature monitor running, and then run prime95 Jan 06 07:22:46 Which would you suggest? Jan 06 07:22:55 Maybe a concky script Jan 06 07:25:05 Trying to build for raspberrypi, but got some QA issue when packaging boost: QA Issue: Architecture did not match. Any ideas? Jan 06 07:26:01 scoutcamper, ? Jan 06 07:26:13 panda-z, having just messed with raspberrypi and abandoned it untill QT5 comes out i can say, ignore the QA issues, we have them on tuna and it works fine Jan 06 07:26:39 what you will find is that QT4 is a lost cause(or rather that is what i found) Jan 06 07:26:41 Rafase282, are you using a 64bit version or 32 on your Ubuntu/ Jan 06 07:26:43 ? Jan 06 07:26:48 64bits Jan 06 07:26:51 kk Jan 06 07:27:25 I concur that the issue is most likely CPU or RAM overheating Jan 06 07:27:48 panda-z, you will need to add a dependancy to qt4 for vc-grapics in order to get some of the egl bits to work Jan 06 07:28:03 Rafase282, and Conky is a good place to start Jan 06 07:28:43 I'm trying this. http://www.unixmen.com/infinity-a-stunning-theme-for-conky-ubuntu-linuxmint-fedora/ Jan 06 07:29:00 I would one with more info tho. Jan 06 07:29:01 scoutcamper: boost compiled fine, but generate x86-64 binaries instead of arm, some other components relying on boost will link fail Jan 06 07:29:44 panda-z, what are you using for the machine layer? Jan 06 07:30:09 Rafase282, not much of a Jan 06 07:30:20 'Theme" person...but I have seen and used conky Jan 06 07:30:22 like it Jan 06 07:30:31 scoutcamper: webOS-Ports' layer, and djwillis/meta-raspberrypi Jan 06 07:30:47 okay Jan 06 07:30:52 hmmmmm Jan 06 07:33:26 i've tried -c cleanall boost and rebuilt for several times, but that didn't help Jan 06 07:33:50 panda-z, i never encountered that error Jan 06 07:34:35 are you building boost-native or boost? Jan 06 07:35:33 panda-z, ^^ Jan 06 07:35:34 ka6sox: boost, should i paste logs? Jan 06 07:35:42 please Jan 06 07:36:21 wait a minute Jan 06 07:37:41 ka6sox: this is log.do_package of boost http://pastebin.com/D3eM5XwP Jan 06 07:38:14 that theme did not work Jan 06 07:39:54 panda-z, you are using bitbake on this? Jan 06 07:40:16 sorry to ask but this doesn't look like output I am normally used to seeing) Jan 06 07:40:45 ka6sox: yes, just add meta-raspberrypi to webos-ports' build env Jan 06 07:41:07 okay... Jan 06 07:41:25 let us push out as far as scoutcamper got. Jan 06 07:41:33 then you can go from there. Jan 06 07:41:44 I think he might be a bit further. Jan 06 07:42:11 kk Jan 06 07:43:48 might take us 24hrs to get that up but ask in case we forget to find you :D Jan 06 07:44:27 just to note: when i build qemuarm using openwebos official OE, this still ocurrs. maybe it is a distro-related problem Jan 06 07:47:48 Bye everyone Jan 06 07:48:15 Thanks for the help. Jan 06 07:52:57 ah Jan 06 07:52:58 maybe Jan 06 08:09:44 ok, so, if you have a shared state file greater than 2 gigs, packaging automatically fails (at least on a x86 build machine) Jan 06 08:09:55 so, that's something that is not so good Jan 06 08:10:17 but, worse than that, is that my webkit sstate is almost exactly double the size it normally is Jan 06 08:11:10 any idea what might cause something like that ? Jan 06 08:12:17 JaMa, ^^ Jan 06 08:13:02 EricBlade: how it fail? Jan 06 08:13:04 s Jan 06 08:14:32 JaMa: chmod during the packaging process complains "Value too large for defined data type", otherwise known as "you have a 32-bit chmod and it can't operate on files that big" (which is really kind of silly, considering that chmod should only need to be able to manipulate the file attributes .. grr ) Jan 06 08:15:23 isn't it just wrong chmod (like from busybox)? Jan 06 08:15:59 I haven't used 32b system for a while, but in gentoo lot of such issues could be fixed by adding support for lfs Jan 06 08:16:48 well i would expect that any modern chmod even in 32bit would have support for large files Jan 06 08:17:01 s/chmod/binutils Jan 06 08:17:09 but apparently that's not the case in Ubuntu Jan 06 08:17:44 and although Ubuntu 11 is quite ancient by modern system standards, it's quite new in comparison to things like "file systems that support files larger than 2gigs" Jan 06 08:18:07 binutils? Jan 06 08:18:29 binutils = the package with chmod and friends i think Jan 06 08:18:36 no :) Jan 06 08:19:24 coreutils provide that Jan 06 08:19:58 ah i see Jan 06 08:20:16 either way Jan 06 08:20:17 double check if your chmod is really from coreutils Jan 06 08:20:25 not just busybox Jan 06 08:20:53 i'm getting a 2.2gig sstate-webkit-webos-qemux86-webos-linux-1.3-0.54-r9-qemux86-2-sha_package Jan 06 08:21:12 the last successful one was 1.1gig Jan 06 08:22:07 it appears to be from coreutils Jan 06 08:22:08 here it's also around 1G Jan 06 08:22:17 but on arm Jan 06 08:23:59 x86 and x86-64 are 100M bigger Jan 06 08:24:21 can you compare content of that 2.2gig archive with previous one? Jan 06 08:25:53 well, here's where it looks weird: i'm getting a libQtWebKit.so for 4.8.2 and for 4.8.3 Jan 06 08:26:40 each of which are 1.2gig Jan 06 08:27:08 are you playing with newer qt submission? Jan 06 08:27:49 EricBlade: btw do you have more details about https://github.com/openwebos/qt/commit/9ac2f8dee3f0b6e3fa1ab00f2f04ec20098c95f4 ? Jan 06 08:28:45 JaMa: desktop refuses to build with it, if you don't supply a -crossarch option, then you just get "Architecture not supported" Jan 06 08:28:46 or some such Jan 06 08:29:46 JaMa: yeah i'm trying to get the latest qt submission as well as every other luna* thing under the sun to all build together Jan 06 08:31:08 the "problem" with building that patch on oe-core, is merely that it required removing the part of the oe-core recipe that patches it in Jan 06 08:32:00 EricBlade: that's what panda-z patch was solving Jan 06 08:32:23 EricBlade: without this patch oe builds fails Jan 06 08:32:34 and qt4-webos recipe depends on this patch Jan 06 08:32:41 oe patches that in before building qt Jan 06 08:33:49 i'd think a better way would be to make it not require having -crossarch to function Jan 06 08:34:17 so you don't have to supply -crossarch if you're not cross compiling Jan 06 08:34:59 EricBlade: bitbake@jama:/OE/webos/owub/build-webos$ ls -lah test.tar Jan 06 08:34:59 -rw-rw-r-- 1 bitbake bitbake 3.2G Jan 6 09:32 test.tar Jan 06 08:35:08 bitbake@jama:/OE/webos/owub/build-webos$ chmod 666 test.tar Jan 06 08:35:15 work fine here in 32bit ubuntu Jan 06 08:35:22 panda-z, PR's pending to push my code out, but http://www.webos-ports.org/wiki/Raspberry_Pi Jan 06 08:35:29 JaMa: 11 or 12 Jan 06 08:35:36 11 Jan 06 08:35:52 ah sorry maybe not, maybe it's 64bit too Jan 06 08:36:06 i see several bug reports about coreutils not being built with large file support Jan 06 08:36:20 sco Jan 06 08:36:35 scoutcamper: thx, i'll try that Jan 06 08:36:44 no prob. :) Jan 06 08:36:48 EricBlade: yes, blame ubuntu :) http://www.gnu.org/software/coreutils/faq/coreutils-faq.html#Value-too-large-for-defined-data-type Jan 06 08:37:06 thanks for thinking of actually looking inside the stupid archive file, though, after quite a lot of frustrated grunts, i didn't even think of that Jan 06 08:37:41 that may be something that we need to possibly know about in case it's the upgrade to 4.8.3 that causes it, rather than the opposite Jan 06 08:38:03 right now i'm trying to build with the current meta-webos, so i'm actually downgrading Jan 06 08:39:35 scoutcamper: should also add one line in conf/layers.txt? Jan 06 08:39:39 and this problem may have lead to my problem that caused me to try to downgrade in the first place, since my build just mysteriously stopped working Jan 06 08:40:08 panda-z, https://github.com/webOS-ports/webos-ports-setup/pull/3/files Jan 06 08:40:35 EricBlade: crossarch patch is applied in recipe, but I was asked to add it to qt directly too: https://github.com/openwebos/meta-webos/pull/33#issuecomment-9804221 Jan 06 08:40:46 EricBlade: and panda-z's patch would allow that Jan 06 08:41:45 would it be possible to fix that so that it doesn't make supplying the crossarch mandatory? Jan 06 08:47:29 EricBlade: yes Jan 06 08:47:38 EricBlade: 32bit gentoo works fine too Jan 06 08:48:57 so coreutils on Ubuntu sucks... Jan 06 08:49:00 got it Jan 06 08:49:57 or gentoo rocks, depends on POV :) Jan 06 10:44:59 morning Jan 06 10:45:12 morning morphis Jan 06 13:21:26 hmm. is there a way to tell bitbake to forcibly remove everything created by a prior version of a package ? Jan 06 13:38:49 EricBlade: bitbake -c cleanall foo (i think) Jan 06 13:39:21 this will also remove the download source tarballs Jan 06 13:48:28 someone here as in depth knowledge of enyo-1? Jan 06 20:19:55 morphis: i'm ok with enyo 1 Jan 06 20:20:34 panda-z: i should've specified, from within a bitbake recipe, sorry. :) need to clean out the old version of a package before putting in the new one Jan 06 20:23:09 I keep getting told that Enyo2 is what is supported...not enyo1 Jan 06 20:26:51 new things should be written in enyo2 Jan 06 20:27:20 morning guys Jan 06 20:28:34 unless you need something that's only in the enyo-1 libs, i'd guess. like accounts handling Jan 06 20:28:54 but can't the enyo-1 libs be easily ported to 2? Jan 06 20:29:18 accounts handling is broken anyways Jan 06 20:33:20 PatrickC: well, not easy as in "plug it in and it works", but easy as in "replace all the flexboxes with something else and replace all the PalmService calls with whatever we're using for an equivalent using PhoneGap" Jan 06 20:33:45 ka6sox: accounts handling is broken? Jan 06 20:33:52 works for email .. Jan 06 20:34:10 yes, that does (until you want to use a gmail account..then broken) Jan 06 20:36:40 however hard coded CSS size values in Core Apps cause them to break in Real Hardware. Jan 06 21:06:13 ka6sox: i use gmail on my desktop build .. Jan 06 21:06:29 hmmm... Jan 06 21:06:41 why is it not happy then... Jan 06 21:07:46 Because we don't have the ssl certs Jan 06 21:07:54 ah, makes sense Jan 06 21:08:44 ok, so, we're running into the error: json_tokener_parse was not declared in this scope, and friends here now. did anyone already fix this out here? Jan 06 21:08:49 Even if you drop them in the right place on the device the imap/pop stuff can't find them Jan 06 21:21:50 b00m Jan 06 21:23:01 lol Jan 06 21:23:05 how is your day going ka6sox? Jan 06 21:23:07 brb real quick Jan 06 21:31:53 back Jan 06 21:42:11 wb Jan 06 21:44:43 my day has been "uneventful" which is good. Jan 06 22:40:51 any resident bitbake experts awake? Jan 06 22:50:27 EricBlade, europe has gone to sleep Jan 06 22:50:37 *maybe* I can help you Jan 06 22:50:42 what is the issue? Jan 06 22:51:04 well, we've got a little issue with the interaction between qt4-webos and webkit-webos Jan 06 22:51:26 even though webkit-webos depends upon qt4-webos, when upgrading versions of qt4-webos, webkit-webos is not rebuilt Jan 06 22:51:43 and also we'd like to be able to completely nuke the prior version of qt4-webos if possible Jan 06 22:52:15 define "completely nuke" Jan 06 22:52:22 from a build? Jan 06 22:52:30 (from the build environment I mean) Jan 06 22:52:57 well, if you increment versions of qt4-webos, then when webkit does get rebuilt, it builds both for the old version and the new version Jan 06 22:53:27 so, we either need to completely wipe out the old version, or we need to specify to webkit exactly which version we need Jan 06 22:53:58 you can specify that in local.conf Jan 06 22:54:59 i think there -may- be a parameter that can be passed to build-webkit to tell it which qt version to build for, but how would one go about telling the webkit bitbake what the version of the qt4 package is Jan 06 22:56:16 you can specify a minimum version of a dependency. Jan 06 22:56:41 jas...cookies Jan 06 22:57:20 i think the root of the problem, though, is that the build-webkit script sees both 4.8.2 and 4.8.3 in sysroots, and decides to build both of them Jan 06 22:57:59 well, that's the root of one problem. the root of the other problem, is that webkit doesn't automatically rebuild when qt4 versions change Jan 06 22:59:45 i suppose, if i forced webkit-webos to depend on qt4-4.8.3 instead of just "qt4" then that would probably force a rebuild Jan 06 23:00:06 it would Jan 06 23:00:11 so in the DEPENDS would that just be making it "qt4-webos-4.8.3" ? Jan 06 23:00:32 >= might help Jan 06 23:00:44 bbiaf... Jan 06 23:05:50 hmm. neither of the manuals for bitbake or openembedded seem to indicate any such thing, searching to see if i can find it used anywhere Jan 06 23:15:30 looks like the OE answer to the problem is "PREFERRED_VERSION", which i've def set, since it's building qt4-webos 4.8.3, but that's still not forcing an update to webkit. may have to do a tag update to webkit Jan 06 23:19:38 there is no such thing as versioned depends Jan 06 23:20:01 sstate should remove all remains of older qt4-webos version when updating it Jan 06 23:20:42 but qt4-webos was broken wrt sstate for long time, so if it's older tmpdir then those files can be not tracked by sstate Jan 06 23:21:05 and here it rebuilds webkit-webos after any change in qt4-webos Jan 06 23:21:09 check sstate checksums Jan 06 23:29:34 JaMa|Zzz: on a clean build of meta-webos @ head, bump pref ver for qt4 to 4.8.3-3 and rebuild, the first thing it does is repackage WebKit without rebuilding it Jan 06 23:53:29 EricBlade, any reason not to have core apps scale instead of being fixed sizes? Jan 06 23:54:05 Other than that they were written specifically for the TouchPad? I can't really think of anything Jan 06 23:58:07 so what happens when the device has a resolution of 2560X1600? we get the Mojo Apps that are UI rev1 style on that device :P Jan 06 23:59:56 1024X768 vs 2560X1600... Jan 07 00:05:03 <_minego> Wait, I thought open webOS didn't support mojo apps... Jan 07 00:07:32 i feel mojo support could be a huge resource hog on the TP.. idk tho w/o any evidence. i thought owos didn't support it either, unless they brought the emulator code over Jan 07 00:09:26 the enyo1 core apps are hard coded to sizes Jan 07 00:09:42 (the size of a TouchPad) Jan 07 00:10:41 i think you can scale apps easily with viewport mods.. Jan 07 00:13:25 there is some issue with that currently Jan 07 00:22:15 bbl(now that I've stirred things up) :( Jan 07 00:24:06 nick ka6sox-away Jan 07 01:08:06 Hello everyone. Jan 07 01:21:22 wb Rafase282 Jan 07 01:22:35 How have things been since I left? Have anyone tried to patch the kernel? **** ENDING LOGGING AT Mon Jan 07 02:59:58 2013