**** BEGIN LOGGING AT Fri Sep 20 03:00:00 2013 Sep 20 07:36:41 morning all Sep 20 07:41:17 morning all Sep 20 07:49:11 hi silvio_ Sep 20 07:59:14 barnacles? Sep 20 08:02:11 good morning Sep 20 08:07:11 hello everybody Sep 20 08:07:44 happy Friday to all :) Sep 20 08:11:33 morning mckoan Sep 20 08:11:41 Noor: morning, and the same to you :) Sep 20 08:21:34 Crofton|work: i wonder if those cheeky baserock guys are still saying a yocto build takes "days" Sep 20 08:21:54 Crofton|work: last i heard, they were telling potential customers that yocto/oe can't do big-endian builds Sep 20 08:25:51 rburton: Isn't it always easier to state something instead of researching it? ;) Sep 20 08:26:22 stefan_schmidt_w: i'd be less bothered if the codethink founder wasn't an ex-OE guy Sep 20 08:26:39 real engineers are usually too anal to pull stuff out of their ass like that... Sep 20 08:28:17 rburton: But he wasn't doing OE stuff for quite some years Sep 20 08:28:56 rburton: Even when I was more active he wasn't doing things anymore and thats some years back Sep 20 08:29:37 But the key here is to get your facts straight when comparing Sep 20 08:30:01 indeed Sep 20 08:31:23 I think its just to alluring letting your pet project look better :) Sep 20 08:32:17 the quote was "Yocto, again, if you want to do a clean build it will take days to build your system, even if you do an incremental build, even if you just do a single change and test it, that will take hours." Sep 20 08:33:10 rburton: yeah, I have seen your fight with him about that before Sep 20 08:35:38 then i guess someone switched my cheap athlon for a cray when i wasn't looking Sep 20 08:35:56 clean build for pi takes about 4 hours Sep 20 08:36:09 nerdboy: Must have happened to all of us Sep 20 08:36:22 nerdboy: turns out those were numbers from memory for a build in a VM on a x200s Sep 20 08:36:53 i'm typing this on a x200s, and its not that slow! worst VM ever. Sep 20 08:43:26 well, sleepy-time for me... Sep 20 08:44:34 broonie: may I ask you for a quick 3-lines review? Sep 20 09:01:04 ant_work: sure Sep 20 09:21:33 hi bluelightning, hi mckoan Sep 20 10:36:23 building qtbase from meta-qt5 doesn't put any example into the final fs, am I missing something? Sep 20 10:41:54 morning all Sep 20 10:42:15 hi pb_ Sep 20 10:45:53 of course I set QT_EXAMPLES ?= "-make examples" Sep 20 10:46:17 morning pb_ Sep 20 10:46:36 hi bluelightning, mckoan Sep 20 11:16:54 rburton, hopefully khem can attend that talk Sep 20 11:44:34 re: baserock (sorry to bring back a topic from 3hrs ago) has anyone ever tried doing an honest comparison between various embedded linux build systems? Sep 20 12:06:07 probably not :) Sep 20 12:06:21 the only two serious ones are buildroot and OE Sep 20 12:06:49 baserock exists to create a market for arm servers Sep 20 12:18:46 is the comment and special handling of device nodes in buildhistory_list_files() (buildhistory.bbclass) really correct? Here, it creates output like Sep 20 12:18:50 crw-rw-r-- root root 5, 14:10 ./console Sep 20 12:19:03 (note the time which should be stripped) Sep 20 12:27:36 ensc|w: at the time I wrote the code, ls was reporting those files without sizes under pseudo Sep 20 12:27:42 ensc|w: it could be that that has changed since then Sep 20 12:28:00 (in which case we should fix the code) Sep 20 12:28:17 bluelightning: no; the problem is that behavior of 'find -ls' changed. Some versions print dev_t, others not Sep 20 12:28:30 perhaps 'find -printf ...' should be used Sep 20 12:28:57 that might be a better approach Sep 20 13:02:00 Crofton|work: what about openwrt? i didn't know anything about it but last summer someone where i worked gave a presentation about it and i was surprised how much it resembled OE/Yocto Sep 20 13:05:08 I forgot about them Sep 20 13:12:54 there's ptxdist as well; last I checked it was still being developed Sep 20 13:23:42 someone can help me understand "LEAD_SONAME" meaning/use? Sep 20 13:34:51 silvio_: i think it's to indicate which is the 'main' so file in the package, in case there are several of them. that name is the one that will be used for the package name Sep 20 13:35:50 Hello. Is OpenEmbedded able to generate an image which contains a VM containing two linux hosts ? Sep 20 13:36:17 tlwoerner: actually, openwrt started as a fork of buildroot Sep 20 13:36:18 well there you go, that's 4 (and maybe more). that sounds like a good LJ article, LF project, and/or something for elinux.org :-) Sep 20 13:38:33 FWIW, there is http://elinux.org/Build_Systems Sep 20 13:39:01 it does list YP and OE under separate items which isn't really correct Sep 20 13:39:18 abelloni: then i guess i'm surprised how much buildroot resembles OE/Yocto :-) Sep 20 13:39:39 speaking of elinux.org, are they still using OE/Yocto at Sony? Sep 20 13:39:53 don't forget ltib ;) Sep 20 13:41:04 conceptually, it is the same, a bunch of recipes and patches Sep 20 13:42:21 is TI still using ltib for new development? they seem quite deep into OE/Yocto these days Sep 20 13:42:34 or does ltib exist as its own project now, apart from TI? Sep 20 13:47:45 we should fix the elinux entry Sep 20 13:48:14 ndec, thanks Sep 20 13:48:26 hmm, how shoudl I word this Sep 20 13:49:22 Crofton|work: you should explain what an umbrella project is ;-) Sep 20 13:49:28 * [[Open Embedded]] - System for building full embedded images from scratch (The [[Yocto Project]] build system) Sep 20 13:49:38 I would, but this is a build system page Sep 20 13:50:05 how is this? object quickly or I will hit save Sep 20 13:50:26 bluelightning, ^^ Sep 20 13:51:05 looks OK to me Sep 20 13:51:08 thanks Sep 20 13:51:09 k Sep 20 13:59:01 tlwoerner: you mean freescale ? Sep 20 13:59:18 anyway, it is also a separate project but there isn't much activity Sep 20 13:59:32 I had to use it for nxp SoCs Sep 20 14:00:12 abelloni: yes freescale (sorry!) :-) Sep 20 14:01:34 the latest release of ltib from Freescale is supposed to be ther last one Sep 20 14:01:47 they've switched to yocto Sep 20 14:03:51 freescale officially switched to yocto, FAE said Sep 20 15:13:32 anyone else seeig faiilures building eglibc-2.18? Sep 20 15:15:22 Crofton|work: can you pastebin the error? Sep 20 15:16:00 http://pastebin.com/EAXVG2EW Sep 20 15:18:21 Best practice to remove or set root password? NOow I found remove "EXTRA_IMAGE_FEATURES = "debug-tweaks"" Sep 20 15:23:53 Crofton|work: haven't seen that one before... Sep 20 15:24:08 hmm Sep 20 15:24:32 something foul is afoot Sep 20 15:43:10 Crofton|work: Your LIBC_FEATURES seems hosed Sep 20 15:43:21 can you check on that ? Sep 20 15:44:00 bitbake -e eglibc | grep "LIBC_FEATURES=" Sep 20 15:48:52 nothing Sep 20 15:48:55 wtf Sep 20 16:09:13 Crofton|work: kheml: i don't get anything with that command either Sep 20 16:09:40 the clostest i get are: DISTRO_FEATURES_LIBC_class-nativesdk DISTRO_FEATURES_LIBC_DEFAULT Sep 20 16:13:40 yeah DISTRO_FEATURES_LIBC is what I was intending to say Sep 20 16:13:46 is that variable empty ? Sep 20 16:13:53 if yes then thats the problem Sep 20 16:15:21 how does it get populated? Sep 20 16:20:24 khem, DISTRO_FEATURES_LIBC="ipv4 ipv6 libc-backtrace libc-big-macros libc-bsd libc-cxx-tests libc-catgets libc-charsets libc-crypt libc-crypt-ufc libc-db-aliases libc-envz libc-fcvt libc-fmtmsg libc-fstab libc-ftraverse libc-getlogin libc-idn libc-inet-anl libc-libm libc-locales libc-locale-code libc-memusage libc-nis libc-nsswitch libc-rcmd libc-rtld-debug libc-spawn libc-streams libc-sunrpc libc-utmp li Sep 20 16:20:24 bc-utmpx libc-wordexp libc-posix-clang-wchar libc-posix-regexp libc-posix-regexp-glibc libc-posix-wchar-io" Sep 20 16:20:59 Crofton|work: it gets initialized in default-distrovars for distro-less. otherwise in distro .conf Sep 20 16:30:42 hi. can we set INHIBIT_PACKAGE_STRIP or INHIBIT_PACKAGE_DEBUG_SPLIT for an individual package, not all? Sep 20 17:20:39 General question for folks. What type of scopes/analyzers do they use/prefer? Sep 20 17:21:06 Crofton|work: khem: denix: ^^? Sep 20 17:23:52 whatever I can get a good deal on Sep 20 17:24:42 I am considering getting one for my home use and was wondering if anyone had some *useful* feedback. Sep 20 17:25:11 you know, reliability, usability, and of course cost too. Sep 20 17:25:19 Crofton|work: so what did you get a deal on last time? Sep 20 17:25:30 B&W tektronix LCD Sep 20 17:29:13 hmm, you use it mostly for your analog RF work? Sep 20 17:29:18 no Sep 20 17:29:22 poking around Sep 20 17:29:23 eew tektronix :) Sep 20 17:29:58 picoscopes are reasonably cheap USB ones. Sep 20 17:30:09 But fairly good. Sep 20 17:30:25 broonie: thx. I was looking at bitscope too. Sep 20 17:31:30 The tektronix ones are fine from a quality point of view, I just hate their UIs. Sep 20 17:32:19 as you say, their UI isn't much to look at, but you can make 'em work. Sep 20 17:32:43 I was leaning away from bitscope due to the requirement for windoze. Sep 20 17:33:34 sorry, I mis-typed that last thing. I meant: " away from picoscope" Sep 20 17:33:55 IIRC they have Linux support, or did last time I looked. Sep 20 17:35:37 broonie: if they do, I didn't notice it. Sep 20 17:36:33 Ah, there's Linux drivers but not picoscope for Linux Sep 20 17:36:59 http://www.picotech.com/linux.html Sep 20 17:37:09 that's what I found too Sep 20 17:38:29 of course, it also begs that question of standalone vs host-connected. Sep 20 17:41:00 standalone Sep 20 17:42:05 that's what I've used primarily, but I was able to hook one up to the network and it made it nice to be able to remotely trigger the scope from my primary workstation. Sep 20 17:42:22 of course, the cost of standalone ones are quite a bit more as well. Sep 20 17:43:52 Crofton|work: you used any of bitscope's products? I am considering the 325. Sep 20 17:44:11 no I haven't **** ENDING LOGGING AT Sat Sep 21 03:00:00 2013