**** BEGIN LOGGING AT Fri Mar 26 02:59:57 2010
**** BEGIN LOGGING AT Fri Mar 26 13:06:14 2010
**** ENDING LOGGING AT Fri Mar 26 13:09:37 2010

**** BEGIN LOGGING AT Fri Mar 26 13:12:14 2010
Mar 26 19:47:44 <tomeff> hello, can anybody help me with http://effik.pastebin.com/iYRpp8Fi ?
Mar 26 23:52:37 <mike928342> anyone around?
Mar 26 23:53:42 <mike928342> I have a question: is the git org.openembedded.stable branch still relevant or is this outdated badly?
Mar 26 23:58:53 <mike98543> !logs
Mar 27 02:20:40 <mike98543> Anybody on here?
Mar 27 02:20:55 <DHowett> check the date on the last commit on that branch? :P
Mar 27 02:21:06 <mike98543> Hi
Mar 27 02:21:18 <DHowett> hi! :P
Mar 27 02:21:24 <mike98543> I think its like 16 months
Mar 27 02:21:39 <DHowett> In that case, I might be inclined to say that it's outdated :P
Mar 27 02:22:13 <mike98543> dev is probably newer, but I was looking for something more ... stable
Mar 27 02:22:38 <DHowett> ahh
Mar 27 02:22:47 <mike98543> I remember reading somewhere that the devs were going to do a stable branch
Mar 27 02:22:49 <DHowett> wish i could help. I'm dealing with a random insane problem of my own :P
Mar 27 02:23:59 <mike98543> Something about companies wanting something with a 5 or 10 year support horizon, I thought that was stable... But I guess there wasn't enough interest
Mar 27 02:24:17 <mike98543> what problems are you having?
Mar 27 02:25:43 <DHowett> gettext-native (0.17) is linking the system's libxml (by path) instead of the prebuild libxml2-native :(
Mar 27 02:26:07 <DHowett> and failing because the staged libz doesn't have gzopen64, which my system's libxml2 requires
Mar 27 02:26:52 <DHowett> popping into a devshell right now to see if there's anything I can do about it :(
Mar 27 02:27:12 <mike98543> that is a problem
Mar 27 02:28:04 <DHowett> I also don't really understand why we need to build most of an entire native sysroot… it seems like it would cut off an entire 50% of the rules to use the host system's tools instead of reinventing them
Mar 27 02:28:13 <mike98543> i've never had that problem with bitbake
Mar 27 02:28:16 <DHowett> hrm :(
Mar 27 02:31:23 <mike98543> Do you mean the toolchain?  Obviously gcc needs to be specific to target hardware, but I guess all of perl, m4, etc... could be same
Mar 27 02:38:03 <DHowett> Right. Everything but the cross-compiler.
Mar 27 02:38:32 <DHowett> I've got the latest perl, m4, etc. (to paraphrase) installed, so it doesn't seem like much of a stretch to just use them instead
Mar 27 02:39:13 <mike98543> I agree
Mar 27 02:39:32 <mike98543> Kindof silly rebuilding everything just so you can build NANO
Mar 27 02:39:48 <DHowett> exactly
Mar 27 02:39:58 <mike98543> what device are you using angstrom on?
Mar 27 02:40:31 <DHowett> a gumstix Overo
Mar 27 02:40:45 <mike98543> thats a neat little device
Mar 27 02:40:47 <DHowett> Howsabout you?
Mar 27 02:40:50 <DHowett> very :D
Mar 27 02:41:01 <mike98543> still messing around with zaurus.. I have an akita and collie
Mar 27 02:41:14 <DHowett> ooh, nice
Mar 27 02:41:17 <DHowett> I always wanted a Zaurus
Mar 27 02:41:37 <mike98543> they are awesome, but software was always what I felt really left the whole package behind
Mar 27 02:41:44 <mike98543> Very far ahead of their time
Mar 27 02:41:52 <DHowett> aye
Mar 27 02:42:24 <DHowett> ouch, gettext is trying to link more than my native libxml
Mar 27 02:42:30 <DHowett>  /usr/lib64/libacl.so /usr/lib64/libattr.so /usr/lib64/libcroco-0.6.so /usr/lib64/libglib-2.0.so /usr/lib64/libxml2.so 
Mar 27 02:42:44 <mike98543> I found it pretty frustrating with OE and OZ and all that... every time it would get close to being usable, something would break. Usually a new kernel would break stuff
Mar 27 02:42:50 <DHowett> hehe
Mar 27 02:42:54 <DHowett> sounds like my desktop linux system :-/
Mar 27 02:43:18 <mike98543> No sense having the newest kernel when your SD/MMC just stopped working
Mar 27 02:44:05 <mike98543> Its not like the zaurus hardware has changed so much that it needs a new kernel every half year.   Hardware hasn't changed for 6 years.
Mar 27 02:44:24 <DHowett> right
Mar 27 02:44:47 <mike98543> so why would gettext want to link against your native libxml?
Mar 27 02:45:06 <mike98543> can't find the one built by toolchain?
Mar 27 02:45:10 <DHowett> good question.. i picked through the recipe and can't find anything about it
Mar 27 02:45:18 <DHowett> it almost seems like autoconf is causing it
Mar 27 02:45:44 <DHowett> which is funny because there's a patch disabling some library searching which pretty much says says "let autoconf handle it, it's smarter than we are"
Mar 27 02:47:06 <mike98543> is the one you want actually built?
Mar 27 02:47:28 <DHowett> I'm pretty sure
Mar 27 02:47:38 <DHowett> I haven't done any modifications to this tree yet, so i'm going with the current defaults
Mar 27 02:47:58 <mike98543> might want to check the .so file is actually around
Mar 27 02:48:26 <mike98543> Could be autoconf is being smart because it can't find one you want?
Mar 27 02:49:00 <DHowett> hmm..
Mar 27 02:49:02 <DHowett> definitely there
Mar 27 02:49:02 <DHowett> ./staging/x86_64-linux/usr/lib/libxml2.so
Mar 27 02:49:45 <mike98543> Should it be using the one for arm architecture?
Mar 27 02:50:12 <DHowett> nah, it's building gettext-native
Mar 27 02:50:33 <mike98543> oh, for x86_64 as part of toolchain, not for a ipkg
Mar 27 02:50:36 <DHowett> I wonder if I could be sneaky and link my system libz over the staged native libz, thus resolving the "missing symbols" ;)
Mar 27 02:56:32 <DHowett> a-ha
Mar 27 02:56:46 <DHowett> About gettext-0.17: "so if no headers are found, it assumes no external support, but it doesnt bother clearing out the discovered library values …"
Mar 27 02:56:55 <DHowett> and something about the libxml2 headers I have staged is broken
Mar 27 02:57:00 <DHowett> (which i saw in the config log)
Mar 27 02:57:46 <mike98543> needs headers?
Mar 27 02:57:48 <DHowett> and it will be fixed in gettext 0.18? … i've been banging my head against a KNOWN BUG for a WEEK?!
Mar 27 02:58:19 <DHowett> It's an include dir problem. usr/include/libxml2 isn't -I, but it's looking for libxml/blah.h (instead of libxml2/libxml/blah.h)
Mar 27 02:58:40 <DHowett> the configure script is "failing" to find the headers for the staged libxml because the test program fails
Mar 27 02:59:00 <mike98543> hmmm
**** ENDING LOGGING AT Sat Mar 27 02:59:56 2010