**** BEGIN LOGGING AT Mon Jul 01 02:59:58 2013
Jul 01 08:12:40 <bluelightning> morning all
Jul 01 08:23:51 <JaMa> bluelightning: morning
Jul 01 08:24:09 <JaMa> bluelightning: have you seen the request for those 2 path.py backports?
Jul 01 08:25:38 <rburton> JaMa: i'm reliably informed that mesa patches sent to the bugzilla are likely to be ignored, so send it directly to the list.
Jul 01 08:25:54 <bluelightning> JaMa: I have yes
Jul 01 08:27:07 <bluelightning> JaMa: we've got some pretty urgent issues to deal with in master atm, unfortunately I think that's going to keep me busy for today
Jul 01 08:29:15 <rburton> JaMa: you might be interested in my ross/mesa branch on poky-contrib.  whoever wins the race to send the patches avoids having to rebase :)
Jul 01 08:33:08 <JaMa> rburton: I plan to send them today :)
Jul 01 08:33:21 <JaMa> bluelightning: ok
Jul 01 11:12:58 <conditionzero> hi i updated the recipie for my bind to 9.9.3-p1 and now it throws back an error on the do compile ../../tools/genrandom 100 randomfile/bin/bash: ../../tools/genrandom: No such file or directory make[3]: *** [randomfile] Error 127
Jul 01 11:13:05 <conditionzero> what could caus this ?
Jul 01 13:53:10 <JaMa> is it possible to combine varflags with override? e.g. PNBLACKLIST_distro[foo] = "foo doesn't work with <distro>"?
Jul 01 14:12:54 <bluelightning_> JaMa: I can't find that we do that anywhere so I suspect not
Jul 01 14:53:09 <JaMa> bluelightning: ok, I've wrapped it in anonymous python which compares DISTRO value
Jul 01 15:15:33 <bluelightning> JaMa: re 4565 - I have all those packages and still I'm not able to reproduce the failure
Jul 01 15:16:44 <bluelightning> JaMa: I'm wondering if it's something you can reproduce on your autobuilder with a clean TMPDIR, just doing bitbake -c configure cmake-native ?
Jul 01 15:17:05 <bluelightning> unless you can think of any other package that might be installed on the host that might trigger it
Jul 01 15:21:15 <JaMa> bluelightning: I'm pretty sure that host system wasn't changed lately, so it's trigerred by some change in metadata or order of items being built
Jul 01 15:21:32 <JaMa> bluelightning: I'll try from clean tmpdir after last world build finishes
Jul 01 15:21:37 <bluelightning> JaMa: sure, I appreciate it's not a host issue, I just need to reproduce it to try to find the problem
Jul 01 15:22:07 <bluelightning> JaMa: actually I might be onto something; I notice your machine is x86-64 but the installed Qt packages are i386, maybe that will trigger it
Jul 01 15:22:34 <JaMa> and because it failed the same for 3 different architectures I guess it's some -native recipe now built before cmake-native (e.g. qt4-native-tools)
Jul 01 15:25:46 <bluelightning> annoyingly the machine I have vmware installed on is 32-bit so I can't try it there :/
Jul 01 15:30:18 <kergoth> looks like the libcurl in buildtools-tarball doesn't support https, breaking git ls-remote of many repositories
Jul 01 15:30:24 * kergoth digs
Jul 01 15:37:57 <kergoth> also looks like the git wrapper for nativesdk is busted, due to hardcoded paths. those wrappers are only useful with the sysroot substitutions happening
Jul 01 15:38:08 * kergoth rolls eyes
Jul 01 15:38:23 * kergoth wonders if buildtools-tarball was ever actually used
Jul 01 15:40:11 <seebs> Gosh, that's a funny coincidence, I think we were just talking with someone at Mentor about buildtools-tarball.
Jul 01 15:44:52 <rburton> khem: you're the patchwork man, right?  can i have an account please?
Jul 01 15:45:17 <rburton> khem: i also want to figure out why its not auto-marking merged patches as well as it should, so if you have any thoughts...
Jul 01 15:46:15 <JaMa> rburton: most common reason is multiple versions of the same patch
Jul 01 15:46:25 <JaMa> rburton: then it marks only one version as applied
Jul 01 15:46:52 <rburton> needs a way to mark v1 as obsolete if a v2 arrives
Jul 01 15:47:15 <JaMa> rburton: unfortunatelly even when v1 is marked as superseded and v2 as "Pending" it marks v1 not v2
Jul 01 15:47:46 <rburton> ah
Jul 01 15:47:48 <rburton> sadface
Jul 01 15:48:39 <bluelightning> yes, that's just broken logic, shouldn't be too hard to fix
Jul 01 15:49:08 <bluelightning> kergoth: I've used it here
Jul 01 15:50:23 <kergoth> hmm, either the env setup script needs to do the equivalent of the sysroot install substitutions on the warppers, or it needs to not use create_wrapper at all
Jul 01 15:53:07 <JaMa> rburton: I've talked with khem about it and IIRC the issue is that the script doesn't use some PW API to query state and patches, it just looks for patch with right checksum
Jul 01 15:53:38 <rburton> ah
Jul 01 16:14:56 <bluelightning> JaMa: bingo... you need 64-bit qmake-qt4 installed but 32-bit libraries
Jul 01 16:15:18 <bluelightning> JaMa: to be honest I'm not sure what the correct course is here... ultimately the host system does appear to be misconfigured
Jul 01 16:24:54 <JaMa> bluelightning: any idea why it's showing it only now?
Jul 01 16:25:23 <bluelightning> JaMa: don't know yet, but I'm testing a possible fix atm
Jul 01 16:26:01 <JaMa> bluelightning: last world build finished few minutes ago, let me know if I should test it as-is or wait for that fix
Jul 01 16:26:31 <bluelightning> JaMa: it's pretty much guaranteed that you'll get the failure now, probably best to wait
Jul 01 16:26:59 <JaMa> ok
Jul 01 17:52:42 <kergoth> Hmm, is there a recipe for the root certificate bundle?
Jul 01 17:53:25 <JaMa> ca-certificates in meta-oe?
Jul 01 17:53:44 <JaMa> meta-openembedded/meta-oe/recipes-support/ca-certificates/ca-certificates_20130119.bb
Jul 01 18:02:20 <kergoth> ah, not sure how i missed that, thanks
Jul 01 18:03:38 <zecke> JaMa: does it work nowadays?
Jul 01 18:05:14 <kergoth> trying to fix buildtools-tarball to actually be useful, need ssl and ca certs for libcurl to be useful to git ffor https repositories..
Jul 01 18:12:50 <JaMa> zecke: yes AFAIK, only the tarball needs to be on PREMIRROR
Jul 01 18:13:01 <JaMa> zecke: it's removed every time new one is published IIRC
Jul 01 18:13:09 <JaMa> rburton: ping about mesa
Jul 01 18:15:00 <JaMa> rburton: nobody will need separate PACKAGECONFIG for --enable-xlib-glx, right?
Jul 01 18:15:32 <JaMa> rburton: I need to enable it to test 0002-pipe_loader_sw-include-xlib_sw_winsys.h-only-when-HA.patch but mesa.inc doesn't allow to force-enable it with PACKAGECONFIG
Jul 01 18:15:51 <JaMa> but I guess everybody is happy with DRI-based glx
Jul 01 18:17:06 <mr_science> adding the PACKAGECONFIG stuff sounds more flexible...
Jul 01 18:17:16 <mr_science> is there a downside to doing that?
Jul 01 18:18:34 <JaMa> mr_science: it's exclusive with other PACKAGECONFIGs like dri
Jul 01 18:18:42 <JaMa> mr_science: and nobody is probably using it
Jul 01 18:19:10 <JaMa> we already have many PACKAGECONFIGs in mesa for options which are more usefull
Jul 01 18:20:38 <mr_science> i'm still working on integrating the rpi gles/egl stuff with mesa
Jul 01 18:20:56 <mr_science> sounded like it might be useful...
Jul 01 18:21:09 <bluelightning_> kergoth: erm... surely it should be picking up the root certificates from the host?
Jul 01 18:21:45 <mr_science> mostly because i've already hacked a few PACKAGECONFIG options in my meta-rpi layer i guess...
Jul 01 18:22:32 <rburton> JaMa: xlib glx is unbelievably useless
Jul 01 18:22:37 <JaMa> mr_science: what option is missing for you? see http://git.openembedded.org/openembedded-core-contrib/log/?h=jansa/mesa
Jul 01 18:22:48 <mr_science> keep in mind it may cause me to bug you at some future time...  ;)
Jul 01 18:22:57 <JaMa> rburton: yes, that's why I need it only to test patch not to use it :)
Jul 01 18:23:14 <rburton> JaMa: surprised that code hasn't disappeared from mesa to be honest
Jul 01 18:24:51 <mr_science> JaMa: thanks, i just emailed myself that link for later...
Jul 01 18:25:04 <JaMa> rburton: that's the winsys pipe_loader which is failing without x11 headers :)
Jul 01 18:25:17 <rburton> huh
Jul 01 18:25:47 <rburton> my mesa patches got merged \o/
Jul 01 18:31:01 <JaMa> rburton: great, so I guess I'll do the rebase :)
Jul 01 18:31:17 <rburton> JaMa: i mean merged upstream, still haven't sent for oe-core
Jul 01 18:39:06 <JaMa> heh I have finally found combination of flags which actually enables xlib in gallium loader (GALLIUM_PIPE_LOADER_DEFINES = -DHAVE_PIPE_LOADER_SW -DHAVE_PIPE_LOADER_XLIB), but such combination doesnt build
Jul 01 18:39:10 <JaMa> cc1: some warnings being treated as errors
Jul 01 18:39:13 <JaMa> make[2]: *** [glapi_dispatch.lo] Error 1
Jul 01 18:39:23 <JaMa> so it's true, nobody is using xlib in gallium loader :)
Jul 01 18:42:43 <eren> hey there
Jul 01 18:53:22 <JaMa> rburton: "build: fix EGL build when no X11 headers are present" isn't what we're using in oe-core now :/ "EGL: Mutate NativeDisplayType depending on config" is IMHO better implementation
Jul 01 18:54:00 <JaMa> rburton: because it doesn't force you to add -DMESA_EGL_NO_X11_HEADERS in CFLAGS when building something else then mesa but still without x11 headers
Jul 01 18:57:04 <rburton> JaMa: it throws the flags into egl.pc too
Jul 01 18:57:30 <JaMa> and the if is already there from b3f1f665b
Jul 01 18:57:59 <JaMa> if echo "$egl_platforms" | grep 'x11' >/dev/null 2>&1; then
Jul 01 18:57:59 <JaMa> AM_CONDITIONAL(HAVE_EGL_PLATFORM_X11, echo "$egl_platforms" | grep 'x11' >/dev/null 2>&1)
Jul 01 18:58:02 <JaMa> if ! echo "$egl_platforms" | grep -q 'x11'; then
Jul 01 18:59:09 <eren> what's your usual setup for testing the packages? I guess QEMU is the solution
Jul 01 18:59:21 <eren> however, what I need is the ability to use serial ports and stuff
Jul 01 18:59:26 <eren> does QEMU support it?
Jul 01 18:59:45 <eren> I mean, can the emulated image use the serial port on the real hw
Jul 01 19:00:03 <kergoth> hrm
Jul 01 19:01:33 <eren> otherwise, I have beagleboard. I will have to produce image for beagleboard, dd it to mmc, boot it up, try it, get an error, rebuild ...
Jul 01 19:01:35 <eren> :)
Jul 01 19:04:20 <Daemon404> why dont you netboot?
Jul 01 19:04:26 <Daemon404> saves so much testing time
Jul 01 19:04:33 <Daemon404> nfs share with the rootfs or something
Jul 01 19:04:52 <eren> hm
Jul 01 19:04:56 <Daemon404> tftp etc
Jul 01 19:05:09 <Daemon404> granted... it's a royal pain in the ass to set up :D
Jul 01 19:05:35 <Daemon404> i know teh pandabard could boostrap/boot off of a usb-to-miniusb cable between a comp and the board
Jul 01 19:05:43 <Daemon404> does the beagleboard have such a feature?
Jul 01 19:05:50 <eren> need to check
Jul 01 19:06:36 <Daemon404> just thinking out loud here.
Jul 01 19:07:27 <eren> I guess u-boot has tftp feature
Jul 01 19:07:35 <eren> let's assume that it has tftp boot
Jul 01 19:07:52 <eren> when we are done with building, we can extract the rootfs to somewhere permanent
Jul 01 19:07:55 <Daemon404> oh
Jul 01 19:07:56 <Daemon404> it does
Jul 01 19:07:57 <eren> I mean, in OE build tree
Jul 01 19:08:02 <Daemon404> ive used u-boot with tftp before
Jul 01 19:08:08 <eren> then it does :)
Jul 01 19:08:18 <eren> so, we extract the rootfs somewhere
Jul 01 19:08:24 <eren> copy the uImage
Jul 01 19:08:43 <eren> enable dhcp and tftp setup
Jul 01 19:08:46 <Daemon404> you an send the uImage over LAN
Jul 01 19:08:52 <Daemon404> insid uboot
Jul 01 19:08:53 <Daemon404> iirc
Jul 01 19:08:58 <eren> hm
Jul 01 19:09:10 <eren> ok, I will look at it
Jul 01 19:09:11 <Daemon404> and then point it at a nfs share
Jul 01 19:09:19 <Daemon404> take everything with a grain of salt
Jul 01 19:09:22 <Daemon404> i havent done this in 2 years
Jul 01 19:10:59 * Daemon404 usually doesnt even look at this channel anymore... just happened to look today because he bought seebs' book
Jul 01 19:11:11 <seebs> Hey, cool!
Jul 01 19:11:21 <seebs> Hope you like it. I actually still use it as a reference sometimes.
Jul 01 19:11:31 <Daemon404> im hoping to use it as a reference
Jul 01 19:11:31 <eren> ah, which book is it?
Jul 01 19:11:38 <Daemon404> i periodically forget certain pitfalls
Jul 01 19:11:42 <Daemon404> and bsd explodes.
Jul 01 19:11:47 <seebs> http://www.amazon.com/dp/1430210435
Jul 01 19:12:08 <seebs> So, long story short: The book was written without anyone in the writing or editing part of it being aware that someone planned to attach "Beginning" to the name.
Jul 01 19:12:23 * mr_science thought bsd was designed not to explode...
Jul 01 19:12:23 <eren> seebs: oh nice! :)
Jul 01 19:12:29 <seebs> The technical reviewer was Gary Vaughan from autoconf land.
Jul 01 19:12:35 <Daemon404> mr_science, bashisms make bsd explode
Jul 01 19:12:38 <Daemon404> so do GNU extensions to sed
Jul 01 19:12:39 <Daemon404> and such.
Jul 01 19:12:41 <seebs> And *both* of us learned a fair bit about shell portability in the process of working on it.
Jul 01 19:12:54 <eren> seebs: do you explain why there are lots of if [ x$foo = "xbar" ] strings?
Jul 01 19:12:57 <eren> what is x here? :P
Jul 01 19:13:01 <seebs> Ahh, yeah.
Jul 01 19:13:09 <seebs> foo=-lt
Jul 01 19:13:13 <seebs> [ -lt = bar ]
Jul 01 19:13:22 <mr_science> been a few years since i played with bsd...
Jul 01 19:13:26 <seebs> Some very old test will explode in fire because they got an operator (-lt) without an operand.
Jul 01 19:13:39 <Daemon404> maybe you can help me with a stupid problem:
Jul 01 19:13:52 <Daemon404> ive been passing regex around as strings in shel scripts
Jul 01 19:14:00 <Daemon404> and they accidentally expand to lists of filenames
Jul 01 19:14:04 <Daemon404> Bad Thins Happen
Jul 01 19:14:08 * mr_science remembers the "lp0 error: printer on fire?" message
Jul 01 19:14:13 <Daemon404> i cant find a non-ugly solution to this
Jul 01 19:15:47 <seebs> Long story short, regex-like things should pretty much always be in "" so they aren't globbed.
Jul 01 19:16:06 <seebs> Except in cases where globbing won't happen, but you can nearly always quote them there too.
Jul 01 19:16:18 <Daemon404> i see
Jul 01 19:16:21 <Daemon404> well i have a strange case
Jul 01 19:16:24 <Daemon404> something like
Jul 01 19:16:38 <Daemon404> THING="regex1*;regex2*"
Jul 01 19:16:40 <Daemon404> IFS=';'
Jul 01 19:16:49 <mr_science> quoting variables/strings/paths is usually a good thing...
Jul 01 19:16:52 <Daemon404> for exp in ${THING}; do...
Jul 01 19:16:56 <seebs> ohhh
Jul 01 19:17:33 <seebs> I would probably not use IFS splitting on things which might be globbish.
Jul 01 19:17:56 <Daemon404> well i figured as much
Jul 01 19:18:01 <Daemon404> im unsure if a 'better' way
Jul 01 19:18:05 <Daemon404> s/if/of/
Jul 01 19:18:32 <seebs> while [ "$THING" ]; do exp=${THING%%;*} THING=${THING#*}; use-exp; done
Jul 01 19:18:55 * eren thinks it's ugly
Jul 01 19:18:59 <eren> python is new bash
Jul 01 19:19:03 <eren> let the flame begin
Jul 01 19:19:05 <eren> :P
Jul 01 19:19:06 <Daemon404> right
Jul 01 19:19:09 <Daemon404> so use expansion
Jul 01 19:19:19 <Daemon404> 'ugly' but less error prone
Jul 01 19:19:21 <Daemon404> i suppose.
Jul 01 19:19:38 <seebs> I am so not a Python fan.
Jul 01 19:19:53 <seebs> It has just never really clicked for me. I like Lua and Ruby, but not Python and Tcl.
Jul 01 19:20:28 * mr_science enjoys bash and python
Jul 01 19:20:33 <mr_science> mostly...
Jul 01 19:21:11 <Daemon404> seebs, no perl?
Jul 01 19:21:19 <Daemon404> i am in fact porting this regex thing /from/ perl
Jul 01 19:21:25 <Daemon404> \fun/
Jul 01 19:21:30 <seebs> perl is sort of in the middle of my language pool. I don't really like it, I don't really hate it.
Jul 01 19:22:11 <eren> I never dared to learn perl
Jul 01 19:22:30 <eren> and I am still amazed how bugzilla is maintained
Jul 01 19:22:37 <eren> it's pure perl, oh
Jul 01 19:22:37 <seebs> My perl is pretty shoddy these days, I learned some very early perl5 and sort of vaguely know about bless and such, but.. only vaguely.
Jul 01 19:23:09 <mr_science> actually, of all the compiled languages i still enjoy Ada...
Jul 01 19:23:52 <Daemon404> i sometimes miss the hardcore unix world wind river
Jul 01 19:24:05 <Daemon404> there;s so much hippy dippy hipster coding around me now
Jul 01 19:24:07 <Daemon404> nodejs and pals
Jul 01 19:24:12 <mr_science> except foe the build overhead, i'm finally starting to like c++ a little better now...
Jul 01 19:24:36 <joshc> IFS splitting can be used for globbish things (in bash): IFS=';' read -ra foo <<< "r1*;r2*"; for exp in "${foo[@]}"; do ...; done
Jul 01 19:24:46 <seebs> Ahh, right.
Jul 01 19:24:52 <eren> Daemon404: node.js, mongodb, textmate, macos, awesome \m/
Jul 01 19:24:54 <Daemon404> i wouldnt be buying seebs' book if i wanted to write bash
Jul 01 19:24:56 <Daemon404> now would i ;)
Jul 01 19:24:57 <seebs> I think you can do that with << too, probably.
Jul 01 19:25:07 <seebs> Although now I don't remember. Hmm.
Jul 01 19:25:27 <Daemon404> eren, i one of three peope who runs linux at my place of work
Jul 01 19:25:31 <Daemon404> maybe 4
Jul 01 19:25:33 <Daemon404> rest run mac os x
Jul 01 19:25:49 <mr_science> not played with mongo, but i do like redis now
Jul 01 19:26:07 <Daemon404> theyre not really useful for the same reasons.
Jul 01 19:26:12 <seebs> OS X has bash these days too, I think. I mean, not just around, but I think /bin/sh might be now.
Jul 01 19:26:53 <Daemon404> i am not a fan of dong development work on os
Jul 01 19:26:55 <Daemon404> os x*
Jul 01 19:26:59 <Daemon404> that is deployed to linux
Jul 01 19:27:06 <Daemon404> just asking for trouble.
Jul 01 19:27:09 <mr_science> Daemon404: other than they're both in the "noSQL" camp...
Jul 01 19:27:09 <eren> well, I use OS X right now
Jul 01 19:27:20 <eren> full screen tmux, zsh, weechat
Jul 01 19:27:31 <eren> not different from my linux installation
Jul 01 19:27:31 <Daemon404> so youre telling me this load balancer works different on a different kernel?
Jul 01 19:27:34 <Daemon404> well. you dont say.
Jul 01 19:27:45 <eren> I use OS X as some kind of a BSD distribution that has nice GUI
Jul 01 19:27:49 <seebs> Huh. If you don't need to be able to use the values in the same shell, echo "$foo" | tr ';' '\012' | while read i ...
Jul 01 19:28:11 <Daemon404> its gui is primarily why i do not use os x
Jul 01 19:28:15 <Daemon404> it constantly gets in my way
Jul 01 19:28:43 <Daemon404> also its subpixel smoothing is horrible
Jul 01 19:28:47 <eren> oh
Jul 01 19:28:51 <eren> I don't think so
Jul 01 19:28:57 <eren> I really love OS X font rendering
Jul 01 19:29:03 <Daemon404> i despise it
Jul 01 19:30:00 <Daemon404> anyway, thanks for the suggestion seebs
Jul 01 19:30:59 <mr_science> pretty much all my apple hardware is ppc, so it mostly runs gentoo...
Jul 01 19:31:42 <Daemon404> my mac mini is running windows 8 for a build/coverage instance for 4 different windows compilers for ffmpeg/libav
Jul 01 19:31:48 <Daemon404> ultimate heretic
Jul 01 19:32:15 <mr_science> i'm surprised you haven't been tracked down and arrested...
Jul 01 19:32:38 <Daemon404> im committed worse sins...
Jul 01 19:32:46 <Daemon404> i've, even.
Jul 01 19:33:15 <mr_science> nobody said said anything about sin
Jul 01 19:33:43 <mr_science> just because you get arrested doesn't mean you actually did anything wrong...
Jul 01 19:34:02 <Daemon404> fair enough.
Jul 01 19:34:30 <Daemon404> try this on then:
Jul 01 19:34:52 <Daemon404> step 1) helped write and tests a c99-to-c89 converter to support MSVC for libav/ffmpeg
Jul 01 19:35:13 <Daemon404> step 2) run into a filesystem filter bug and get a friend to file na internal bug report
Jul 01 19:35:30 <Daemon404> step 3) ms announces MSVC gets c99 support with a giant ffmpeg logo at their latest conference
Jul 01 19:35:44 <Daemon404> im pretty proud of that one
Jul 01 19:36:14 <mr_science> not sure about the cause and effect chain there...
Jul 01 19:36:35 <mr_science> the answer to your bug was a new release?
Jul 01 19:36:42 <Daemon404> oh the bug with with MSE
Jul 01 19:36:51 <Daemon404> but it brough the c99-to-c89 converter to their attention
Jul 01 19:36:59 <Daemon404> compiler team wasnt so thrilled, as i hear.
Jul 01 19:37:10 <mr_science> ah...
Jul 01 19:37:34 <mr_science> maybe the compiler team should get off their collective asses
Jul 01 19:37:41 <mr_science> just sayin'...
Jul 01 19:37:58 <Daemon404> their c++ support is rather good
Jul 01 19:38:09 <Daemon404> nobody uses c99 except foss pretty much
Jul 01 19:38:11 <Daemon404> hence ms didnt care
Jul 01 19:39:07 <mr_science> why wasn't gcc the answer for you in this case?
Jul 01 19:39:25 <Daemon404> because fuck gdb
Jul 01 19:39:33 <Daemon404> msvc's debugger is actually /nice/
Jul 01 19:39:40 <Daemon404> i hate gdb with the passion of a thousand suns
Jul 01 19:39:43 <Daemon404> after years of using it
Jul 01 19:40:08 <mr_science> considering it's windows, wouldn't it have to be?
Jul 01 19:40:20 <Daemon404> what do you mean?
Jul 01 19:40:59 <Daemon404> (side note: ffmpeg has supported mingw's gcc for aeons)
Jul 01 20:03:18 <mr_science> Daemon404: i was just referring to MSVC on windows needing a very good debugger...
Jul 01 20:03:33 <Daemon404> 0/10 poor troll
Jul 01 20:03:55 <mr_science> okay, i'll buy that...
Jul 01 20:04:15 <Daemon404> speaking of windows
Jul 01 20:04:31 <Daemon404> i was pondering the other day if its possible for glibc to pull something like MS's c runtime does
Jul 01 20:04:39 <Daemon404> filling free'd memory with a specfic constant
Jul 01 20:04:45 <Daemon404> instead of maybe working and maybe not
Jul 01 20:04:52 <Daemon404> if you try and access it
Jul 01 20:15:56 <mr_science> been a long time since i had to troubleshoot memory errors on a windows box
Jul 01 20:16:09 <eren> I haven't touched a windows box since ages
Jul 01 20:16:19 <eren> let alone, programming on windows
Jul 01 20:16:21 <eren> nah..
Jul 01 20:16:43 <mr_science> technically i have to touch one at work for certain things, but me too, more or less
Jul 01 20:17:00 <eren> I heard that they have an internal API call with 23 parameters :P
Jul 01 20:17:15 <mr_science> gave up on windows at home when my kids were just little tikes
Jul 01 20:17:32 <mr_science> they all grew up on gentoo
Jul 01 20:17:48 <eren> nice!
Jul 01 20:18:08 <eren> mr_science: are they having hard times using windows?
Jul 01 20:18:26 <mr_science> well, my daughter had to have macbook when she left for school
Jul 01 20:18:40 <mr_science> made her pay for half of it tho
Jul 01 20:18:55 <Daemon404> im probably the youngest person in this channel
Jul 01 20:18:57 * Daemon404 runs
Jul 01 20:19:14 <eren> Daemon404: your age?
Jul 01 20:19:14 <mr_science> my first answer was something like "What?!? i can buy 5 laptops for that price!!"
Jul 01 20:19:28 <Daemon404> 2
Jul 01 20:19:30 <Daemon404> er
Jul 01 20:19:30 <Daemon404> 23
Jul 01 20:19:43 <eren> you lost it
Jul 01 20:19:45 <eren> I'm 22
Jul 01 20:20:03 <Daemon404> well then.
Jul 01 20:20:42 <eren> mr_science: well, that's ok but MBP Retinas are quote appealing
Jul 01 20:20:45 <eren> quite*
Jul 01 20:21:27 <mr_science> eren: they mostly used openoffice at school but never had any MS office trouble AFAIK
Jul 01 20:22:25 <mr_science> that resolution is wasted on tablet-size dsiplays...
Jul 01 20:22:43 <mr_science> maybe on bigger ones it's more noticeable
Jul 01 20:23:28 <Daemon404> clearly we should all stay at 90 PPI for eternity
Jul 01 20:23:31 <Daemon404> as has been the case.
Jul 01 20:24:10 <mr_science> it's more or less a function of display size and viewing distance
Jul 01 20:25:16 <mr_science> the normal viewing distance for a 10" (or smaller) display doesn't lend itself to super-high res
Jul 01 20:25:48 <mr_science> if your eyes can't see it, why drive the extra pixels?
Jul 01 20:26:09 <Daemon404> in general i am quite close to my laptop screen
Jul 01 20:26:13 <Daemon404> its pretty noticeable.
Jul 01 20:26:27 <Daemon404> especially WRT font rendering
Jul 01 20:26:51 <mr_science> as i sit about two feet from a smallish 1080p laptop display
Jul 01 20:27:29 <mr_science> well, proper font rendering is about a lot more than just resolution...
Jul 01 20:27:39 <eren> that's why I love OSX
Jul 01 20:27:44 <eren> <3 subpixel rendering
Jul 01 20:27:52 <Daemon404> os x's is by far the worst
Jul 01 20:28:00 <Daemon404> it makes some sites (slashdot e.g.) unreadable
Jul 01 20:28:01 <eren> + Macbook's LCD is far greater than normal LCDs
Jul 01 20:28:19 <Daemon404> it's worse than stock debian fontconfig/freetype configs
Jul 01 20:28:44 <Daemon404> i wish i had kept those images i snapped...
Jul 01 20:29:22 <eren> Daemon404: http://picpaste.com/Screen_Shot_2013-07-01_at_10.28.18_PM-zmaMb78j.png
Jul 01 20:29:25 <Daemon404> not to mention if you use a non-aaple monitor on a mac mini, it disables a bunch of font renderig stuff
Jul 01 20:29:28 <eren> it looks good to me
Jul 01 20:30:01 <Daemon404> the titles look pretty terrible to me
Jul 01 20:30:07 <Daemon404> too bold, jagged edges.
Jul 01 20:30:40 <eren> too bold?
Jul 01 20:31:09 <Daemon404> apple's subpixel rendering does not respect the fact that monitors are pixel-based
Jul 01 20:31:13 <Daemon404> it ends up lookign bold/smooth/ugly
Jul 01 20:31:21 <eren> on non-apple monitor, yes
Jul 01 20:31:38 <eren> when you are in Apple world, you should not leave it :P
Jul 01 20:31:52 <eren> I use AOC 23" monitor with macbook air
Jul 01 20:31:55 <eren> it's good so far
Jul 01 20:31:59 <Daemon404> apples monitors are crap for DSP or image relatd work
Jul 01 20:32:01 <Daemon404> glossy is terrible
Jul 01 20:32:05 <Daemon404> and they dont offer matet any more
Jul 01 20:32:08 <Daemon404> matte*
Jul 01 20:32:32 <eren> ah yeah
Jul 01 20:32:36 <Daemon404> (i work in DSP/image stuff)
Jul 01 20:33:03 * mr_science is listening to his freshly released PiMP3 yocto build
Jul 01 20:33:14 <Daemon404> also tryin to remember if theyre 8bit or some crap 6bit dithered thing
Jul 01 20:33:19 <Daemon404> i think theyre at least not TNs
Jul 01 20:40:18 <seebs> I think you can still get non-Retina MBPs in antiglare as a special option. But yeah.
Jul 01 20:48:12 <Lahar> Hi! I am looking for some basic knowledge on rootfs and elf files? Can somebody please help telling me where can I find some information on it?
Jul 01 20:52:23 <mranostay> google!
Jul 01 20:59:34 <Lahar> mranostay: already tried google...didn't help me much....that's why I am here....I understand the definition of rootfs and I understand fsl-image-minimal and fsl-image-full etc are rootfs images....but why do we call them rootfs? Are they linux images?
Jul 01 21:19:14 <rtollert> Does anybody have experience building a mingw SDK? I tried setting SDK_ARCH="i586", SDK_OS="mingw32", SDK_EXEEXT=".exe", and it starts off on the right track, but gcc-crosssdk-initial eventually fails because we're not creating gcc-4.7.2/winsup/ prior to `configure`
Jul 01 21:19:55 <rtollert> (and more specifically failing due to a lack of finding windows.h)
Jul 01 23:09:13 <mulhern> How can you ensure that a process is running as a daemon after a build machine has booted…supposing you know the proper command line arguments to start that process as a daemon?
Jul 01 23:23:01 <sgw1> mulhern: a ps with a grep, I think the sanity tests do something similar to check for connman
Jul 01 23:24:04 <sgw1> mulhern: scripts/qemuimage-tests/tools/connman_test.sh
Jul 02 00:01:44 <mulhern> sgw1: OK, that tests that it's been set up successfully. What about actually causing it to be started up?
Jul 02 00:02:14 <mulhern> sgw1: Also, could you tell me what causes connman_test.sh to be run?
Jul 02 00:05:35 <sgw1> mulhern: It's part of the sanity test infrastucture, but that's all getting re-written, there was an RFC post about that from stefan on Friday
Jul 02 00:05:49 <sgw1> on the oe-core list
Jul 02 00:14:41 <mulhern> sgw1: OK, I see the email.
Jul 02 01:30:35 <swex> hi all
Jul 02 01:30:55 <swex> can anybody help me with customize Yocto image please
Jul 02 01:31:46 <swex> I've got tslib in my image but it seems to be compiled with wrong kernel headers
**** ENDING LOGGING AT Tue Jul 02 02:59:58 2013