**** BEGIN LOGGING AT Mon Aug 06 02:59:58 2012 Aug 06 07:54:16 khem: Crofton|work: It seems to work allthough I am not sure that my own defconfig is actually used. Thanks :) Aug 06 08:05:04 should new layers be placed inside "oe-core" folder, or next to, in partent folder? Aug 06 08:26:21 anyone?? Aug 06 08:26:48 better next to Aug 06 08:35:46 morning everyone Aug 06 08:41:47 khem: Crofton|work: Ok my defconfig is copied in allright, but apparently it is another .config that is used for the actual compilation. That merge-config thing might have something to do with it. How is defconfig used in relation to merge-config? Aug 06 08:49:03 @ JaMa|Wrk: Thanks... :) Aug 06 09:50:14 morning all Aug 06 09:59:51 SorenHolm, weird, tha recipe works for me and doesn't seem to do anything silly Aug 06 10:00:18 jesus, the damn thing actually landed Aug 06 10:08:20 Crofton|work: well ... it's just that the .config used to *excactly* same file as the defconfig I provided. It's problaly merge-config related.... Aug 06 10:09:18 hmm, I bett er doubdle check mine Aug 06 10:11:09 Crofton|work: my machine is just the defaul qemu and some extra config-stuff is problably added. But still I'd like them to be equal. Aug 06 10:13:51 hi pb_ Aug 06 10:22:56 I need to create a simple webpage on our system. I see there's a recipe for lighttpd in meta-openembedded, but I'm not sure what to put on top. What do other openembedders use? Aug 06 10:23:47 hi bluelightning Aug 06 10:25:18 tasslehoff: lighttpd_1.4.31.bb is in recipes-extended. Aug 06 10:28:19 SorenHolm: noted. thanks. Aug 06 11:09:26 Is it ifup that looks at my /etc/network/interfaces file? In my current oe-core image I get a dhcp address even if I have setup static in this file. Aug 06 11:10:19 suspect it maybe because I don't have pumpd/dhcpd which I had on oe-classic Aug 06 11:20:29 tasslehoff, yes it is Aug 06 11:24:14 GNUtoo: yes it is ifup, or yes, it is because you don't have a daemon? :) Aug 06 11:24:48 hm. guess I need to figure out what happens with /etc/init.d/networking on my systemd image. Aug 06 11:25:06 tasslehoff, it's ifup Aug 06 11:25:10 at least I guess Aug 06 11:25:29 tasslehoff, what init system do you use? Aug 06 11:41:30 GNUtoo: systemd (based on angstrom console image) Aug 06 11:42:25 ok Aug 06 11:42:36 in any way it looks in /etc/network/interfaces Aug 06 11:42:42 I don't know what exactly tough Aug 06 11:48:33 seems to be connmand that sets up eth0 for me on first boot Aug 06 12:15:22 what's the proper way to get rid of connmand? Aug 06 12:19:06 tasslehoff: AFAICT the only thing that pulls it in is NETWORK_MANAGER in task-core-x11-sato.bb... you can set that to "" I guess Aug 06 12:19:15 assuming you're using a sato image that is Aug 06 12:19:29 otherwise something else outside of OE-Core is bringing that in Aug 06 12:32:36 bluelightning: I have a custom angstrom console-base-image. that one installs a lot of connman-stuff. I managed to remove it, but need to add else in its place. Aug 06 12:33:10 netbase? Aug 06 12:34:00 surely netbase is always included... Aug 06 12:34:31 JaMa|Wrk: netbase is already there, and manually calling /etc/init.d/networking works. Aug 06 12:34:40 since you're using angstrom, it could be that systemd is not set up out of the box to enable network interfaces in the absence of a network management daemon, I'm not sure... Aug 06 12:34:46 this is just me not knowing what makes networking tick :) Aug 06 12:34:55 bluelightning: that is the case Aug 06 12:34:58 you didn't say what functionality you expect from :) Aug 06 12:35:19 JaMa|Wrk: tried to hide my ignorance in the <>'s ;) Aug 06 12:35:27 same as what is provided by the sysvinit network scripts I would assume Aug 06 12:36:02 * JaMa|Wrk is still using images without connman and with netbase only Aug 06 12:36:14 I guess what I want at this point is for /etc/init.d/networking start (or ifup -a) to be called on boot Aug 06 12:36:21 right, it is netbase that contains the scripts... Aug 06 12:36:33 JaMa|Wrk: with systemd? Aug 06 12:36:35 yes Aug 06 12:36:50 ok, so it shouldn't be too difficult for tasslehoff to get working then... Aug 06 12:36:51 systemd is executing sysv initscript unless you blacklist them Aug 06 12:37:04 JaMa|Wrk: which perhaps angstrom does, then Aug 06 12:38:10 meta-systemd/recipes-core/systemd/systemd-compat-units.bb actually does that unless you change postinst in .bbappend Aug 06 12:39:52 yep, pointing to /dev/null Aug 06 12:39:54 I'll send patch allowing easier override of blacklisted services Aug 06 12:40:05 without overwritting whole postinst Aug 06 12:41:40 great Aug 06 13:34:16 must run, will figure out the rest tomorrow. thanks for helping. Aug 06 18:41:31 Hi, guys whenever I use hob in yocto to build a custom image, I am unable to login using root Aug 06 18:41:49 it would say incoorect login, is there a solution already out there for this? Aug 06 20:19:07 ping zenlinux Aug 06 20:27:38 fcooper, hello Aug 06 20:29:58 zenlinux: Are you the maintainer for oe-core or meta-oe? Aug 06 20:30:19 fcooper, just for the denzil branch Aug 06 20:30:23 of oe-core Aug 06 20:31:45 Richard Purdie (RP_) is effectively the oe-core master maintainer, and I'm not sure who is the meta-oe maintainer Aug 06 20:32:03 k. So psplash has 3 patches that are in the master branch but not in the denzil branch. 2 of the patches simply bump psplash versions and are minor tweaks which I would like pulled in to the denzil branch. The last patch does alot more and might end up breaking other things. Aug 06 20:33:18 fcooper, since denzil is a released version of oe-core, we don't do recipe version bumps unless there is an important bugfix or security issue that needs to be addressed. Aug 06 20:33:57 If either of those situations is true, I'd like to know more. Aug 06 20:35:02 zenlinux: 1 patch fixes seg fault issue and the other patch I want simply moves some hardcoded values into defines in an external header file. These changes were actually made in the psplash repo which means in oe core we just need to update to the latest version of psplash. Aug 06 20:35:54 fcooper, can you point me to the segfault patch? Aug 06 20:36:14 http://git.yoctoproject.org/cgit/cgit.cgi/psplash/commit/ Aug 06 20:38:26 fcooper, looks like a low-risk patch, I will file a bug against denzil and put that on my todo list Aug 06 20:38:42 how about the defines patch? Aug 06 20:38:47 zenlinux: http://git.yoctoproject.org/cgit/cgit.cgi/psplash/commit/?id=84764337a584002a92940323d374b0e417c573a6 Aug 06 20:40:07 fcooper, regarding that color defines patch, would that change the behavior users would have to follow when doing a basic operation with psplash, e.g, creating a new logo to use with it? Aug 06 20:42:39 zenlinux, The only change is if someone wants to change the background color or text color and if they previously did a patch to psplash.c to modify those changes. Now they would have to make a patch against psplash-colors.h or replace the file. Aug 06 20:44:07 zenlinux, if during the process of creating a new logo they also tweaked color information then yes they would have to do a small tweak to get it to work. Aug 06 20:44:19 fcooper, that might be a bit of a grey area - I could see the decision going either way. Would you mind starting a discussion thread about whether that's appropriate for denzil on the oe-core ML? Aug 06 20:45:11 While I'm the "maintainer" for denzil there are more senior folks in the project who would still have veto power over my opinions if I had to give you a straight yes/no answer on that one. Aug 06 20:45:21 The segfault patch, on the other hand, is pretty likely to make it in Aug 06 20:46:16 zenlinux: No problem I will post it on the ML. So if you took the segfault patch would you only take the patch and apply it onto of the version of psplash you have right now or would you bump the version to the version of psplash that includes the patch? Aug 06 20:47:47 fcooper, I would try to backport it to the psplash versino we're using now - it looks like this should be pretty straightforward to do Aug 06 20:48:20 you could also ask in your oe-core ML thread discussion whether we could consider making an exception and bumping the entire version of psplash as well Aug 06 20:48:35 but I'll say it's something we're generally reluctant to do Aug 06 20:50:50 zenlinux: My biggest reasoning is that there was a patch in oe-classic that does almost the exact same thing as this defines patch. And I didn't want to port this patch and put it in my layer which uses the denzil branch of oe-core if a simply version bump would avoid it. But I will start an discussion on the mailing list about it. thx Aug 06 20:50:50 fcooper, here's the segfault bug just filed: https://bugzilla.yoctoproject.org/show_bug.cgi?id=2903 Aug 06 22:17:42 how can i build an embedded linux with custom user space for embedded devices that looks like android or meego and provide animation and graphics like android like openGL ...etc into my embedded linux [01:14] is that possible ? [01:14] what are the tools required or knowledge for that part ? [01:15] just use android if you want android? [01:16] how can i customize android to provide a new user space Aug 06 22:17:56 how can i build an embedded linux with custom user space for embedded devices that looks like android or meego and provide animation and graphics like android like openGL ...etc into my embedded linux Aug 06 22:22:37 snoutos, think of Qt embedded Aug 06 22:22:58 ok how can i create a custom linux from scratch with QT embedded Aug 06 22:23:11 Yocto project for crownbay has a qt demo image, only rootfs, framebuffer and opengl I guess Aug 06 22:23:23 yocto project surely can help, Aug 06 22:23:41 Start with getting started, use hob to build a qemu image Aug 06 22:23:48 with qt and go ahead from there Aug 06 22:24:22 I would say, read Qt documentation, what a framebuffer is Aug 06 22:24:40 can i compile that project to run on ARM processors for targetting embedded devices Aug 06 22:24:46 like tablets ...etc Aug 06 22:24:47 What are Qt requirements for a framebuffer Aug 06 22:25:01 What arm processor? Aug 06 22:25:10 Are you using a panda or a beagle board? Aug 06 22:25:35 i am not using anything Aug 06 22:25:50 all i need is to create a custom embedded linux that can run on ARM processors Aug 06 22:26:01 because i knew that most of embedded systems hardware run ARM Aug 06 22:26:11 You need to think of a target board, thats where the driver binaries and the complications come in place Aug 06 22:26:22 as the drivers are not completely opensource so Aug 06 22:27:02 Arm, there are many kinds of processors, which is not the problem, but the graphics drivers may or may not be available for or, or need to be hacked Aug 06 22:27:36 So I guess you want something simple to start wit, try qemuarm emulator in yocto / hob Aug 06 22:28:03 adityGandhi , can you add me to your skype contact Aug 06 22:28:09 my skype contact is hisprofessional Aug 06 22:28:34 i will tell you everything in details Aug 06 22:28:35 there Aug 06 22:28:57 sorry snoutos, but this is where you can catch me! Aug 06 22:29:37 you can send me a mail at adityag8 at google mail Aug 06 22:30:42 i will add you now to my gmail account Aug 06 22:30:46 do you have google talk ? Aug 06 22:30:50 i have a very good project Aug 06 22:30:57 that i need to develop with you side by side Aug 06 22:31:02 you will gain good money Aug 06 22:31:03 from it Aug 06 22:31:04 trust me Aug 06 22:31:15 heh Aug 06 22:31:17 i am going to add you there now Aug 06 22:31:25 if you are opening google talk please chat with me there Aug 06 22:31:31 what hehe ? Aug 06 22:31:57 i have added you there now Aug 06 22:34:26 adityag8 i have added you Aug 06 22:34:28 where are you ? Aug 06 22:35:39 ping me Aug 06 22:35:56 But i really think you need to rethink you idea in much greater detail Aug 06 22:36:32 snoutos: could you please send a mail, not too comfortable with chats! **** ENDING LOGGING AT Tue Aug 07 02:59:58 2012