**** BEGIN LOGGING AT Mon Sep 20 02:59:57 2010
Sep 20 03:02:30 <philipp64|laptop> RealOpty: I think he's gone off to bed.
Sep 20 03:03:10 <RealOpty> indeed.
Sep 20 03:03:15 <philipp64|laptop> who owns the asterisk functionality?
Sep 20 03:03:56 <RealOpty> maintains it? should say with opkg info packagename
Sep 20 03:03:59 <RealOpty> or something like that
Sep 20 03:11:53 <philipp64|laptop> it needs a version bump for 1.6.
Sep 20 03:23:52 <RealOpty> iirc there is a package for 1.6
Sep 20 04:51:51 <philipp64|laptop> RealOpty: there is, but I think there's a newer version of 1.6.2.x out than that.
Sep 20 04:52:09 <philipp64|laptop> also, 1.8 is in beta... and being able to build trunk svn would be nice.
Sep 20 05:57:32 <builder> build #82 of ubicom32 is complete: Failure [failed compile_4]  Build details are at http://tksite.gotdns.org:8010/builders/ubicom32/builds/82
Sep 20 07:42:12 <builder> build #107 of octeon is complete: Failure [failed compile_7]  Build details are at http://tksite.gotdns.org:8010/builders/octeon/builds/107
Sep 20 07:51:45 <solca> nbd: FYI: http://openwrt.pastebin.com/Fx2W7hth
Sep 20 08:03:30 <acoul> hello
Sep 20 08:16:00 <nbd> solca: which openwrt version?
Sep 20 09:26:02 <RealOpty> nbd, https://dev.openwrt.org/ticket/7604#comment:4
Sep 20 09:27:23 <nbd> i don't think AC101LKQT vs AC101LF makes any difference for the software
Sep 20 09:28:30 <RealOpty> problem with ethtool?
Sep 20 09:28:47 <nbd> no
Sep 20 09:28:53 <nbd> b44 driver issue most likely
Sep 20 09:30:38 <RealOpty> nbd, why is there two? https://dev.openwrt.org/attachment/ticket/7604/openwrt-dmsg#L122
Sep 20 09:32:45 <nbd> because the hw has two ethernet MACs
Sep 20 09:44:44 <RealOpty> kk
Sep 20 09:54:15 <RealOpty> nbd, um http://wiki.openwrt.org/ is down
Sep 20 09:54:21 <RealOpty> but https is up
Sep 20 09:54:26 <RealOpty> :/
Sep 20 09:58:44 <KanjiMonster> working for me (or back up again)
Sep 20 09:59:06 <RealOpty> lol yeah is now. :)
Sep 20 09:59:36 <stintel> maybe too many connections or something :)
Sep 20 10:00:32 <RealOpty> From dawkins.nashipur.org (188.40.126.135) icmp_seq=106 Destination Host Unreachable
Sep 20 10:00:32 <RealOpty> 64 bytes from ncc.subsignal.org (188.40.166.25): icmp_seq=111 ttl=46 time=166 ms
Sep 20 10:00:57 <RealOpty> yeah something.........
Sep 20 10:01:16 <KanjiMonster> RealOpty: is "*teh* playground" intentionally? ;)
Sep 20 10:02:37 <KanjiMonster> (and it would be nice if the wiki had a "view source" without aquiring an edit lock)
Sep 20 10:04:57 <RealOpty> what?
Sep 20 10:08:50 <RealOpty> KanjiMonster, dont get kinky with me... :P
Sep 20 10:37:11 <cshore> ping xMff
Sep 20 11:02:53 <phaidros> RealOpty: all back for good now, we had a system upgrade running.
Sep 20 11:02:57 <phaidros> sry for inconvenience
Sep 20 11:03:38 <phaidros> wiki should be back fully now. in case pls qry me here
Sep 20 11:06:33 <RealOpty> phaidros, does wiki have rss?
Sep 20 11:07:39 <RealOpty> phaidros, dont you love ubuntu upgrades :)
Sep 20 11:20:55 <RealOpty> there pushing stuff for a newer release lol oh well i love pointless kernel upgrades
Sep 20 11:33:28 <KanjiMonster> RealOpty: :p - if the "what" belongs to the view source - currently the only way to see the wiki source of a page (e.g. for the "device template") is to edit the page; but this also marks the page as being edited, preventing others from doing the same (iirc). And this wasn't a request to you ;))
Sep 20 11:56:53 <builder> build #102 of mpc52xx is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/mpc52xx/builds/102
Sep 20 12:22:09 <builder> build #37 of at91 is complete: Failure [failed shell_3]  Build details are at http://tksite.gotdns.org:8010/builders/at91/builds/37
Sep 20 14:24:00 <cshore> hmmm...it's looks like jow broke the compile of util-linux-ng
Sep 20 14:24:55 <xMff> how?
Sep 20 14:32:51 <cshore> don't know but yours is the last change to the Makefile and it fails to build on a fresh checkout...investigating now
Sep 20 14:33:55 <cshore> something to do with cfdisk, even though I don't have it selected
Sep 20 14:34:28 <xMff> you need ncurses
Sep 20 14:34:43 <xMff> the fix is only effective when starting with a fresh .config
Sep 20 14:34:54 <cshore> fresh checkout, fresh config
Sep 20 14:34:55 <xMff> its a dependency selection thing
Sep 20 14:35:04 <cshore> totally fresh
Sep 20 14:35:07 <cshore> and fails
Sep 20 14:35:42 <cshore> I did a fresh checkout and make menuconfig
Sep 20 14:36:15 <cshore> it's a problem that isn't user error
Sep 20 14:37:23 <cshore> I think philip64 said he had the same problem
Sep 20 14:37:27 <cshore> with a fresh checkout
Sep 20 14:37:38 <xMff> would be nice to see a log of "the problem"
Sep 20 14:38:39 <cshore> http://openwrt.pastebin.ca/1945089
Sep 20 14:39:06 <cshore> I can do the whole log if you want
Sep 20 14:39:42 <xMff> I need the config.log of this package
Sep 20 14:40:13 <xMff> Check your staging_dir/target-mips_uClibc-0.9.30.1/usr/include/ for ncurses.h
Sep 20 14:41:01 <cshore> http://openwrt.pastebin.ca/1945092
Sep 20 14:41:53 <cshore> it's missing...parallel buld problem?
Sep 20 14:42:01 <xMff> probably
Sep 20 14:42:21 <xMff> the package specifies PKG_BUILD_DEPENDS e2fsprogs, libncurses
Sep 20 14:42:30 <xMff> no idea why it is not respected
Sep 20 14:42:32 <cshore> trying without -j now
Sep 20 14:44:36 <builder> build #84 of ramips is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/ramips/builds/84
Sep 20 14:47:03 <cshore> nope, still fails
Sep 20 14:47:18 <cshore> and still no ncurses
Sep 20 14:47:41 <cshore> libncurses isn't selected in menuconfig
Sep 20 14:48:54 <cshore> PKG_BUILD_DEPENDS=e2fsprogs libncurses vs PKG_BUILD_DEPENDS:=e2fprogs libncurses ?
Sep 20 14:49:03 <cshore> i.e. getting applied at the wrong time?
Sep 20 14:49:06 <xMff> maybe
Sep 20 14:49:12 <cshore> let me try
Sep 20 14:50:00 <cshore> should libncurses be selected in menuconfig, or is BUILD_DEPENDS not a visible dependency?
Sep 20 14:52:30 <xMff> not visible iirc
Sep 20 15:03:01 <cshore> is it BUILD_DEPENDS or PKG_BUILD_DEPENDS?
Sep 20 15:05:37 <cshore> does PKG_BUILD_DEPENDS have to depend on a source package not a binary?  i.e. ncurses not libncurses?
Sep 20 15:05:54 <xMff> that is it maybe
Sep 20 15:06:11 <cshore> trying now
Sep 20 15:09:15 <cshore> nope
Sep 20 15:19:08 <solca> nbd: backfire
Sep 20 15:25:14 <cshore> xMff: what's the best way to make sure my test of a change to PKG_BUILD_DEPENDS is getting into the build system?
Sep 20 15:25:58 <xMff> examine tmp/.packageinfo
Sep 20 15:27:18 <nbd> solca: which version exactly?
Sep 20 15:28:46 <solca> nbd: sorry, r22857 from the 2010-Sep-02
Sep 20 15:30:45 <nbd> ok
Sep 20 15:30:51 <nbd> i'll look into it
Sep 20 15:38:34 <solca> nbd: thx!
Sep 20 15:40:49 <cshore> what is Build-Depends supposed to do, and how is it done?
Sep 20 15:41:14 <cshore> 'cause it's not working for util-linux-ng
Sep 20 15:45:10 <cshore> I'm guessing it's supposed to trigger the build of the build-dependants, but I don't see where that happens in include/* or scripts/*  there's a metadata.pm that mentions build-depends...what does it do?
Sep 20 15:48:28 <cshore> tmp/.packagedeps has $(curdir)/util-linux-ng/compile += $(curdir)/ncurses/compile $(curdir)/e2fsprogs/compile
Sep 20 15:48:54 <cshore> I presume that should be triggered when util-linux-ng is to be compiled?
Sep 20 15:51:26 <xMff> yes
Sep 20 15:52:57 <cshore> ok, manually doing make package/ncurses/compile prints it highlighted but does nothing
Sep 20 15:53:56 <cshore> is there a why to tell it to do it even though it's not selected in menuconfig?
Sep 20 15:54:02 <cshore> *why=way
Sep 20 16:01:30 <cshore> i.e. what does the build system do for build-depends make package/compile/ncurses? or not do that doing make package/compile/ncurses from the command-line does do?
Sep 20 16:01:51 <cshore> is there some variable to set/unset?
Sep 20 16:11:51 <xMff> cshore: so make package/ncurses/compile V=99  just skips it?
Sep 20 16:12:45 <cshore> http://openwrt.pastebin.ca/1945166
Sep 20 16:13:19 <cshore> but there is no ncurses
Sep 20 16:13:45 <cshore> either in build_dir or staging_dir
Sep 20 16:14:10 <xMff> odd
Sep 20 16:17:02 <cshore> is there something that checks if package is configed in .config?
Sep 20 16:17:13 <cshore> or something like that?
Sep 20 16:18:02 <xMff> yes
Sep 20 16:18:37 <xMff> however it should only detect ncurses as fully built if is in build_dir/ and has a .built stampfile
Sep 20 16:19:08 <cshore> in what dir?
Sep 20 16:20:35 <xMff> build_dir/target-$arch/ncurses-*/
Sep 20 16:20:51 <cshore> ok, that's definitely not there
Sep 20 16:21:05 <cshore> xMff: do you know where the check is in the build system?
Sep 20 16:21:16 <xMff> no, I'd need to dig that up first
Sep 20 16:27:41 <cshore> I found it: it's part of Build/DefaultTargets: it's $(STAMP_BUILT): $(STAMP_CONFIGURED)
Sep 20 16:28:21 <cshore> which goes to $(STAMP_PREPARED):
Sep 20 16:28:21 <cshore> and the preparation stuff
Sep 20 16:29:09 <xMff> wonder if it is related to the md5s issues recently
Sep 20 16:29:31 <xMff> its definitely used to compute certain stampfile names
Sep 20 16:30:54 <cshore> STAMP_PREPARED used md5
Sep 20 16:31:39 <cshore> *uses
Sep 20 16:33:48 <xMff> cshore: https://dev.openwrt.org/changeset/22720/trunk/include
Sep 20 16:33:58 <xMff> maybe that is the culprit
Sep 20 16:34:21 <cshore> just found that line....I'll try reverting?
Sep 20 16:34:29 <xMff> yes
Sep 20 16:36:00 <stintel> ah damn, I forgot my 2nd rspro at work :(
Sep 20 17:19:13 <cshore> ok, I tried just reversing the changes in that specific patch, but there have been other changes since that prevented that working.  I am now trying a build with the results of doing svn merge -r HEAD:22720 on each of the files in the changeset
Sep 20 17:32:56 <CIA-14> nbd * r23100 /trunk/package/mac80211/patches/521-ath9k_aggr_race_fix.patch: ath9k: fix a warning during a failed aggregation start
Sep 20 17:33:47 <builder> build #118 of brcm63xx is complete: Failure [failed compile_4]  Build details are at http://tksite.gotdns.org:8010/builders/brcm63xx/builds/118
Sep 20 18:00:30 <cshore> nope
Sep 20 18:00:45 <CIA-14> nbd * r23101 /trunk/toolchain/gcc/ (8 files in 2 dirs): gcc: add an updated version of 4.4.1+cs - successfully builds a toolchain now, please test!
Sep 20 18:08:49 <philipp64|laptop> [florian]: ping
Sep 20 18:10:56 <cshore> philipp64: hey did you figure anything out about util-linux?
Sep 20 18:11:31 <philipp64|laptop> cshore: no, I mostly ignored it since I wanted to get the Geos platform up ad running... I might circle back to it.
Sep 20 18:12:07 <cshore> I'm trying to get it working now....just wanted to make sure I wasn't duplicating
Sep 20 18:12:40 <philipp64|laptop> I was wondering why the default gcc (at least for the x86) is so ancient... any idea? 4.1.2?  I can't think of a single distro that's currently older than 4.3.5.
Sep 20 18:12:54 <philipp64|laptop> and Fedora is using 4.4.4 right now.
Sep 20 18:14:08 <cshore> philipp64: probably just not enough time to update and test new releases
Sep 20 18:17:25 <cshore> philipp64: things tend to break in on embedded systems that work on x86 desktops, so there is more reluctance to change when something works
Sep 20 18:18:40 <cshore> besides, what's the advantage of rapidly changing?
Sep 20 18:21:22 <cshore> or maybe it's that x86 is probably not the most updated platform...some of the other platforms use newer gcc...x86 isn't as common on routers
Sep 20 18:22:55 <cshore> In fact I just check brcm63xx and it's gcc 4.3.3+cs, so my guess is there just isn't a lot of loving of the x86 platform
Sep 20 18:23:07 <cshore> *checked
Sep 20 18:38:21 <stintel> iirc, last time I tried building an x86 image with same toolchain versions as defaults for ar71xx, it worked fine
Sep 20 18:38:25 <stintel> but that's been a while
Sep 20 18:39:48 <stintel> uclibc 0.9.32 was another story though :)
Sep 20 18:48:33 <stintel> nbd: would you recommend enabling graphite framework in gcc 4.4.1+cs or not ?
Sep 20 18:48:58 <nbd> i just did a build with graphite enabled
Sep 20 18:49:04 <nbd> don't really have any experience with it
Sep 20 18:50:22 <stintel> I enabled it on my gentoo machines, gcc 4.4.3, didn't notice anything bad with it so far
Sep 20 18:51:02 <stintel> I'll enable it in my x86 openwrt build as well
Sep 20 18:51:39 <cshore> Could PKG_BUILD_DIR:=$(BUILD_DIR)/$(BUILD_VARIANT)/$(PKG_NAME)-$(PKG_VERSION) by the culprit?  What is BUILD_VARIANT set to when doing package/build/ncurses?
Sep 20 18:52:01 <stintel> and if I dont forget my 2nd rspro at work tomorrow I'll do an ar71xx build with 4.4.1+cs+graphite as well, see how it goes
Sep 20 18:52:29 <stintel> nbd: you kept using binutils 2.19.1 and uclibc 0.9.30.1 ?
Sep 20 18:52:57 <philipp64|laptop> cshore: as more x86 variants come out (c7, atom, etc) having a newer compiler gives you better optimization.
Sep 20 18:53:35 <nbd> stintel: yes, i left that at default for now
Sep 20 18:53:42 <nbd> but i intend to switch to uclibc git soon
Sep 20 18:53:56 <stintel> ok
Sep 20 18:54:16 <stintel> I'll leave it too for now, so if there are issues it's likely just related to the gcc version switch :)
Sep 20 18:54:20 <cshore> I mean package/ncurses/compile
Sep 20 18:54:42 <nbd> i haven't actually tested the image built with 4.4.1+cs yet
Sep 20 18:54:43 <stintel> too bad -march=atom is not even in 4.4
Sep 20 18:54:49 <cshore> does it simply compile both defined binary packages?
Sep 20 18:54:54 <nbd> stintel: it is in 4.4.1+cs
Sep 20 18:54:59 <stintel> ah, cool
Sep 20 18:55:10 <nbd> at least i saw something that looked like that when i cleaned up the patch
Sep 20 18:55:29 <stintel> oh well, doesn't matter for now, currently still running gentoo on my atom
Sep 20 18:55:34 <stintel> but it could be interesting
Sep 20 18:56:00 <stintel> it's just .. trying to get xbmc compiled for openwrt/x86 is something I am not looking forward to
Sep 20 18:56:01 <nbd> if 4.4.1+cs works, i might merge it into 4.4.3+cs later
Sep 20 18:56:08 <nbd> maybe that would be a good candidate for a new default toolchain
Sep 20 18:56:25 <philipp64|laptop> stintel: I'm still waiting for -Op (optimize for power consumption) :-)
Sep 20 18:56:32 <stintel> ok, I'll try give it some testing
Sep 20 18:56:34 <stintel> philipp64|laptop: heh :P
Sep 20 18:56:48 <nbd> the compiler doesn't go up to that level of abstraction ;)
Sep 20 18:58:15 <stintel> bah I should get rid of this core 2 duo and get something more modern :(
Sep 20 18:58:32 <philipp64|laptop> nbd: not yet, anyway. just wait until Cap & Trade drives up the cost of electricity. :-(
Sep 20 18:59:14 <nbd> power optimization is not really something that a compiler could improve - that's the programmer's job
Sep 20 19:04:46 <cshore> nbd: any why I'm getting http://openwrt.pastebin.ca/1945166 when there is no ncurses in build_dir, and util-linux-ng is selected (which build-depends on libncurses) but it is not selected in menuconfig?
Sep 20 19:05:13 <cshore> or is it possible to manually for a package to compile when it isn't in menuconfig?
Sep 20 19:05:17 <cshore> .config I mean
Sep 20 19:05:33 <philipp64|laptop> nbd: anyone doing reviews/commits today?
Sep 20 19:05:54 <cshore> I'm trying to see why the build of util-linux-ng is failing
Sep 20 19:08:39 <nbd> cshore: at the moment build-depends don't work for packages that provide multiple variants
Sep 20 19:10:19 <philipp64|laptop> speaking of optimization... anyone going to review/commit my 1-line patch for the net5501 subtarget?  :-)
Sep 20 19:12:58 <stintel> bah what's with js.generic today
Sep 20 19:16:30 <cshore> nbd: jow made a change to util-linux-ng that depends on ncurses, but not for all binary packages, hence a build depends on ncurses, but ncurses has build variants
Sep 20 19:16:41 <cshore> *that is he did
Sep 20 19:17:12 <nbd> the build dep is unnecessary then
Sep 20 19:17:32 <nbd> it's enough that one virtual package has a binary dependency
Sep 20 19:17:37 <nbd> since it's not conditional
Sep 20 19:18:06 <cshore> nbd: well the build is failling to pull in ncurses
Sep 20 19:18:28 <nbd> hm
Sep 20 19:18:32 <nbd> oh, wait
Sep 20 19:18:44 <nbd> ah, now i get it
Sep 20 19:19:55 <dtatulea_> hi
Sep 20 19:20:51 <nbd> that one's a bit tricky to fix
Sep 20 19:23:09 <cshore> nbd: heh, so I've noticed
Sep 20 19:24:49 <nbd> ah, i have an idea
Sep 20 19:24:56 <nbd> i'll give you a patch for testing in a minute
Sep 20 19:25:44 <cshore> cool...I've been beating my head against this one today
Sep 20 19:28:16 <nbd> try this: http://nbd.name/depfix.patch
Sep 20 19:28:46 <cshore> actually jow didn't break it, he tried to fix it....it seems ncurses is present on most builds otherwise so it doesn't normally show up
Sep 20 19:29:10 <nbd> yeah
Sep 20 19:29:29 <nbd> with this patch it should fall back to building the first variant of a package if no variants are selected
Sep 20 19:30:21 <cshore> erm, did you patch off trunk?
Sep 20 19:30:29 <nbd> yes
Sep 20 19:30:43 <cshore> wait, you did from git, right -p1
Sep 20 19:31:00 <nbd> yes
Sep 20 19:31:11 <nbd> i usually always make -p1 patches
Sep 20 19:31:21 <nbd> i think svn's default of doing -p0 patches is braindead
Sep 20 19:31:42 <cshore> patch still fails with -p1
Sep 20 19:32:28 <nbd> wget -O- http://nbd.name/depfix.patch | patch -p1
Sep 20 19:32:31 <nbd> that works for me
Sep 20 19:32:33 <nbd> on a different checkout
Sep 20 19:33:15 <cshore> doh, pebkac
Sep 20 19:34:27 <cshore> ok, building
Sep 20 19:34:45 <dtatulea_> hi, is there something more verbose than V=99?
Sep 20 19:39:12 <nbd> nope
Sep 20 19:39:59 <dtatulea_> so I have to manually start the build on a package to see the full gcc switches?
Sep 20 19:46:12 <stintel> meh, collectd is giving compile errors again :(
Sep 20 19:49:41 <stintel> http://openwrt.pastebin.com/3VijbeSq
Sep 20 19:50:20 <stintel> but when I use my config for ar71xx image, I don't get this
Sep 20 19:51:40 <stintel> and I don't get why it doesn't find libperl or libcurl for example, I certainly did ./scripts/feeds install {perl,curl}
Sep 20 19:57:52 <stintel> any pointers on where to look are more than welcome :)
Sep 20 20:04:30 <cshore> nbd: btw do you think we could make the blkid and swap-utils use the util-linux-ng version again?  There are couple of reasons: 1) size...I know it's not huge but I think we need to start looking at trimming optional stuff when it's not actually used, and 2) util-linux-ng has volume type detecting functionality I want for block-hotplug
Sep 20 20:05:12 <cshore> nbd: also, so you know, jow and I are working on making block-extroot so that it can be loaded as a module (and the kernel modules it depends on as well)
Sep 20 20:06:14 <stintel> ah seems if I disable collectd-mod-curl problem goes away
Sep 20 20:06:35 <cshore> nbd: block-mount and block-hotplug will also be able to be modules
Sep 20 20:07:49 <cshore> also jow was thinking we should combine block-mount and block-hotplug.  What do you think?
Sep 20 20:09:10 <nbd> cshore: ok, sounds good. though blogic is already finishing a c mount daemon that'll work with the new luci2 userland stuff
Sep 20 20:09:34 <nbd> so did my patch work?
Sep 20 20:11:11 <cshore> nbd: not sure....dropbear crapped out so it didn't get far enough to see
Sep 20 20:13:18 <nbd> you can try make package/ncurses/compile
Sep 20 20:13:24 <nbd> if that does something, i expect my patch to work
Sep 20 20:14:37 <cshore> It does Entering & Leaving with nothing (not up to date, which is the problem we had before) in between.  ncurses isn't selected in menuconfig so is that to be expected?
Sep 20 20:14:58 <nbd> try removing tmp/.packagedeps, then try again
Sep 20 20:15:19 <cshore> same thing
Sep 20 20:15:29 <nbd> ok
Sep 20 20:15:32 <nbd> i'll look into it again
Sep 20 20:16:13 <cshore> nbd: how is luci2 coming along?
Sep 20 20:16:43 <nbd> haven't really looked into the web interface part, but the core components are coming along nicely
Sep 20 20:17:01 <nbd> we'll integrate ubus (our tiny dbus replacement) along with some components that use it soon
Sep 20 20:17:31 <cshore> nbd: like the mount daemon?
Sep 20 20:18:01 <nbd> yes
Sep 20 20:18:02 <cshore> i.e. is it worth messing with the stuff other than extroot atm?
Sep 20 20:18:06 <CIA-14> nbd * r23102 /trunk/ (include/subdir.mk scripts/metadata.pl): fix build dependencies on packages with multiple variants but no selected virtual packages
Sep 20 20:18:21 <nbd> mount daemon is something that blogic's working on
Sep 20 20:18:27 <nbd> btw. with that commit, ncurses should work
Sep 20 20:19:35 <cshore> nbd: in svn is it theirs-config or mine-conflict to get your changes?
Sep 20 20:19:57 <cshore> nvm, I'll revert
Sep 20 20:20:29 <cshore> yup, works
Sep 20 20:20:54 <philipp64|laptop> cshore: theirs
Sep 20 20:22:31 <cshore> nbd: will the mount daemon work in preinit so that we can still have extroot?
Sep 20 20:27:33 <nbd> i haven't looked into the structure of that yet
Sep 20 20:33:25 <cshore> ping jow
Sep 20 20:35:10 <cshore> jow: what do you think of http://openwrt.pastebin.ca/1945353?  I'm testing it now (the change from yours is only the extroot stuff)
Sep 20 20:50:12 <CIA-14> nbd * r23103 /trunk/toolchain/gcc/common.mk: gcc: enable non-PIC for mips targets in gcc 4.4+ to reduce code size
Sep 20 21:20:42 <builder> build #83 of ps3 is complete: Failure [failed compile_6]  Build details are at http://tksite.gotdns.org:8010/builders/ps3/builds/83
Sep 20 21:26:08 <cshore> jow: with a couple of missing backslashes it works for extroot and kernel modules on jffs (i.e. opkg'd modules); testing to confirm regular image doesn't break
Sep 20 21:26:38 <xMff> cshore: great
Sep 20 21:26:53 <xMff> I had to fight my system in the meanwhile
Sep 20 21:27:50 <cshore> xMff: what was wrong?
Sep 20 21:28:06 <xMff> heh, Ubuntu upgrade
Sep 20 21:28:37 <cshore> xMff: broke your system?
Sep 20 21:28:44 <xMff> kind of
Sep 20 21:28:57 <xMff> I had 3rd party xorg repos
Sep 20 21:29:10 <xMff> those didn't play too well with the new release
Sep 20 21:30:32 <xMff> now I have shiny 3d desktop stuff, drawing speed is a bit disappointing though
Sep 20 21:30:58 <xMff> maximizing a window has like 1 second delay
Sep 20 21:33:29 <cshore> 3d desktop as in true 3d (i.e. not just borders and stuff ?)
Sep 20 21:33:51 <xMff> well for transparency and stuff
Sep 20 21:34:45 <cshore> ah...what graphics card?  on my system the ubuntu maximum effects has no noticeable delay compared to no effects
Sep 20 21:35:11 <xMff> Radeon HD 2600XT
Sep 20 21:36:33 <cshore> odd....mine's a built in Radeon HD something (r600) and is great
Sep 20 21:36:55 <xMff> using radon(hd) ?
Sep 20 21:37:00 <xMff> or the proprietary one?
Sep 20 21:37:56 <cshore> radeon (not radeonhd....radeonhd is going to go away in future since radeon now supports the hd stuff)
Sep 20 21:38:25 <xMff> ah, didn't know
Sep 20 21:38:47 <xMff> last time my card had no 3d support with radeonhd
Sep 20 21:38:55 <xMff> maybe it changed by now
Sep 20 21:39:36 <cshore> ah...don't know....it's a fast-moving target
Sep 20 21:40:23 <cshore> I love Windows-Tab
Sep 20 21:40:37 <cshore> it's a fun way to switch windows
Sep 20 21:45:38 <xMff> my windows key launches a terminal :P
Sep 20 21:46:27 <cshore> heh
Sep 20 21:46:37 <stintel> now that's the most useful use of the windows key I have ever heard :)
Sep 20 21:46:47 <stintel> mine is disabled, logitech g15 ftw :)
Sep 20 22:14:46 <CIA-14> nbd * r23104 /trunk/target/linux/x86/net5501/target.mk: x86: turn on optimizations for net5501 (patch by Philip A. Prindeville)
Sep 20 22:14:59 <KanjiMonster> hrm, is there a reason for ac_cv_func_setresuid being set to no in include/site/linux?
Sep 20 22:18:09 <xMff> bah, the dokuwiki code is such a pile of crap
Sep 20 22:18:23 <xMff> but then, which php project is not
Sep 20 22:24:43 <KanjiMonster> samba 3.5 seems to need setresuid (without it it doesn't like the negative uid for nobody and fails changing its uid to it)
Sep 20 22:27:32 * RealOpty is pissed
Sep 20 22:28:00 <RealOpty> wife reset the wrt cause wifi didnt work, was at like 60 something days uptime
Sep 20 22:28:10 <xMff> lol
Sep 20 22:30:58 <xMff> somehow the multilingual plugin lost its icons in the wiki
Sep 20 22:31:20 <xMff> and while reviewing the code I wondered how it ever could've worked
Sep 20 22:34:09 <RealOpty> id think that http://wiki.openwrt.org/zh-cn/toh/start would be a mirror of the real toh and just translated. but its not :/
Sep 20 22:36:32 <xMff> well any translation project I know is lagging behind
Sep 20 22:36:50 <xMff> don't know why the chinese guys started to translate the toh
Sep 20 22:37:02 <RealOpty> KanjiMonster, that would be a great  feature
Sep 20 22:37:04 <xMff> especially the tables
Sep 20 22:42:07 <KanjiMonster> RealOpty: well, samba 3.5 isn't exactly lightweight (~13 MB ram in idle, ~5 MB ipk, ~20 MB uncompressed), but as far as I can see everything works (haven't tested printing yet since the printer isn't connected, but the printers defined in cups do show up)
Sep 20 22:45:04 <KanjiMonster> oh, I missed nmbd, make that ~18 MB ram
Sep 20 22:48:14 <RealOpty> KanjiMonster, was responding to your comment 11 hours ago.
Sep 20 22:48:26 <KanjiMonster> ah
Sep 20 22:48:27 <KanjiMonster> ok
Sep 20 23:00:05 <RealOpty> ping cshore
Sep 20 23:00:13 <cshore> pong
Sep 20 23:00:26 <RealOpty> cshore, you get util-linux-ng working?
Sep 20 23:00:41 <RealOpty> I had that same problem.
Sep 20 23:00:46 <RealOpty> 'no input files'
Sep 20 23:01:00 <cshore> nbd did a fix once I asked I found out where and why we had a problem
Sep 20 23:01:23 <cshore> he knew what was needed once I asked the right question
Sep 20 23:02:24 <RealOpty> ah
Sep 20 23:02:26 <RealOpty> https://dev.openwrt.org/changeset/23102/trunk
Sep 20 23:02:28 <RealOpty> is that the fix?
Sep 20 23:05:41 <RealOpty> ah read irc log. thats the fix :) coolbeans
Sep 20 23:06:03 <RealOpty> was trying to tell xMff that his fix didnt fix it lol
Sep 20 23:06:18 <RealOpty> ty cshore nbd philipp64|laptop
Sep 20 23:08:03 <RealOpty> cshore, you merging the menu change for extroot with backfire too?
Sep 20 23:26:49 <cshore> RealOpty: hmmm...
Sep 20 23:56:44 <RealOpty> cshore, well if you change it in backfire update wiki plz :)
Sep 20 23:57:10 <xMff> cshore: readon works indeed better
Sep 21 00:07:05 <RealOpty> xMff, i havent done a make since r22378 (backfire)
Sep 21 00:07:30 <RealOpty> do i need to do a make clean or anything to the .config?
Sep 21 00:08:51 <xMff> idk
Sep 21 00:09:06 <xMff> if the opkg stuff was already done by then, then probably not
Sep 21 00:12:04 <RealOpty> ok
Sep 21 00:54:00 <CIA-14> jow * r23105 /branches/backfire/package/opkg/patches/013-fix-darwin-compile.patch: [backfire] merge r22340
Sep 21 01:40:32 <RealOpty> xMff, http://paste.ubuntu.com/497396/  :/
Sep 21 01:41:05 <xMff> RealOpty: don't know that one
Sep 21 01:42:34 <RealOpty> kk
Sep 21 01:42:55 <RealOpty> ill grab a fresh svn, copy .config and see what happens
Sep 21 01:53:26 <RealOpty> xMff, we should make the output files (in bin) like this bin\rXXXXX\$TARGET\output_files_here
Sep 21 01:53:47 <RealOpty> i have so many builds that i dont know what svn version they are  lol
Sep 21 02:13:53 <RealOpty> lol is https://dev.openwrt.org/ down?
Sep 21 02:15:00 <RealOpty> 502 Bad Gateway
Sep 21 02:15:01 <RealOpty> :/
Sep 21 02:15:12 <xMff> it seems to suffer from load problems atm
Sep 21 02:15:22 <xMff> back
Sep 21 02:25:20 <CIA-14> jow * r23106 /packages/utils/bzip2/Makefile: [packages] bzip2: bump to v1.0.6 (USN-986-1, CVE-2010-0405)
Sep 21 02:31:14 <RealOpty> ffs im trying to open a ticket....
Sep 21 02:36:29 <RealOpty> ok so if anyone likes the idea of bin\rXXXXX\$board
Sep 21 02:36:30 <RealOpty> https://dev.openwrt.org/ticket/7974
Sep 21 02:36:37 <RealOpty> theres the patch.
**** ENDING LOGGING AT Tue Sep 21 02:59:57 2010