2013-10-01 05:59:57 <ncopa> morning
2013-10-01 06:00:00 <ncopa> its october
2013-10-01 06:00:07 <ncopa> its one month to v2.7
2013-10-01 06:00:42 <ncopa> should start think of closing up new features and start think about fixing bugs...
2013-10-01 06:21:47 <fabled> ncopa, llvm upgrade would be nice
2013-10-01 06:22:12 <ncopa> # Maintainer: William Pitcock <nenolod@dereferenced.org>
2013-10-01 06:22:16 <ncopa> ;)
2013-10-01 06:23:07 <ncopa> fabled: do you have any link to the 3.4 release announcement or tarball?
2013-10-01 06:23:23 <ncopa> i can only see 3.3 release
2013-10-01 06:23:25 <fabled> oh
2013-10-01 06:23:35 <fabled> that's weird
2013-10-01 06:23:39 <ncopa> LLVM 3.4 Release — In the Future!
2013-10-01 06:23:56 <fabled> seems i was mistaken
2013-10-01 06:24:06 <fabled> or there was glitch yesterday and it showed up as a release
2013-10-01 06:24:22 <fabled> sounds like i just misread when being sick yesterday
2013-10-01 06:27:36 <foxkit> so, the wiki is a little.. terse.. for me, trying to make a subpackage for syslinux efi
2013-10-01 06:27:55 <foxkit> and so the efi component of syslinux uses an extra dependency that syslinux doesn't
2013-10-01 06:28:04 <foxkit> it says I can use "any variable for the package in the subpackage definition, and it will work"
2013-10-01 06:28:36 <foxkit> well.. pkgdesc works.. but depends/makedepends definitely do not :(
2013-10-01 06:30:52 <ncopa> foxkit: which wiki page?
2013-10-01 06:31:08 <foxkit> was following along with the APKBUILD pages and examples
2013-10-01 06:31:36 <foxkit> http://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#Custom_subpackages
2013-10-01 06:31:43 <foxkit> The above mentioned variables can also be used in our custom function. If we want for instance to build the test() function with perl support we would add:
2013-10-01 06:32:08 <foxkit> though that doesn't actually work - if I apk del the -dev packages it doesn't pull them in and if I apk add the subpackage it doesn't pull in the dep
2013-10-01 06:32:14 <foxkit> so nothing works right (missing libraries)
2013-10-01 06:32:21 <foxkit> err
2013-10-01 06:32:30 <foxkit> if I apk del the -dev package, abuild doesn't pull it in
2013-10-01 06:32:34 <foxkit> so it doesn't even build properly
2013-10-01 06:33:01 <ncopa> do you think you could paste your APKBUILD?
2013-10-01 06:33:21 <foxkit> sure thing!  let me put it somewhere :)
2013-10-01 08:18:11 <StarWarsFan> moin
2013-10-01 08:18:23 <StarWarsFan> any news regarding this error:
2013-10-01 08:18:35 <StarWarsFan> 10:16:04 ERROR: unsatisfiable constraints:
2013-10-01 08:18:35 <StarWarsFan> 10:16:04 conflicts: libpng-1.5.17-r0
2013-10-01 08:18:35 <StarWarsFan> 10:16:04 libpng-1.6.5-r0:
2013-10-01 08:18:35 <StarWarsFan> 10:16:04 satisfies:
2013-10-01 08:18:36 <StarWarsFan> 10:16:04 libpng-dev-1.6.5-r0[libpng=1.6.5-r0]
2013-10-01 08:18:38 <StarWarsFan> 10:16:04 qt-x11-4.8.5-r0[so:libpng16.so.16]
2013-10-01 08:18:40 <StarWarsFan> 10:16:04 freetype-2.5.0.1-r1[so:libpng16.so.16]
2013-10-01 08:18:42 <StarWarsFan> 10:16:04 libpng-1.5.17-r0:
2013-10-01 08:18:44 <StarWarsFan> 10:16:04 conflicts: libpng-1.6.5-r0
2013-10-01 08:18:46 <StarWarsFan> 10:16:04 breaks: libpng-dev-1.6.5-r0[libpng=1.6.5-r0]
2013-10-01 08:18:49 <StarWarsFan> 10:16:04 satisfies:
2013-10-01 08:18:51 <StarWarsFan> 10:16:04 openjdk6-jre-1.6.0_p24-r8[so:libpng15.so.15]
2013-10-01 08:18:59 <StarWarsFan> i'm still stuck on this :-(
2013-10-01 08:19:09 <ncopa> StarWarsFan: x86_64?
2013-10-01 08:19:18 <StarWarsFan> yes, edge
2013-10-01 08:19:26 <ncopa> the problme is
2013-10-01 08:19:53 <ncopa> openjdk needs to be rebuilt against libpng-1.6
2013-10-01 08:20:03 <ncopa> unfortunally, openjdk does not build on x86_64
2013-10-01 08:20:06 <ncopa> i dont know why
2013-10-01 08:20:30 <ncopa> you can add --force
2013-10-01 08:20:34 <ncopa> i think...
2013-10-01 08:20:39 <ncopa> or remove openjdk for now
2013-10-01 08:20:56 <StarWarsFan> impossible because it's a jenkins-node ;-)
2013-10-01 08:21:02 <ncopa> ok...
2013-10-01 08:21:18 <StarWarsFan> so it needs java to communicate with the jenkins master
2013-10-01 08:21:24 <ncopa> understand
2013-10-01 08:21:42 <ncopa> they you'll have to wait til we have fixed openjdk
2013-10-01 08:22:02 <StarWarsFan> ok,
2013-10-01 08:22:06 <ncopa> i will have to prioritize the 2.6.5 release today
2013-10-01 08:22:09 <StarWarsFan> will take a look...
2013-10-01 08:22:35 <ncopa> but openjdk have been broke way to long
2013-10-01 08:22:40 <ncopa> we need to fix it asap
2013-10-01 08:22:50 <ncopa> might be the recent gcc updates fixes it
2013-10-01 09:14:28 <ncopa> i need some help with http://bugs.alpinelinux.org/issues/2252
2013-10-01 09:14:31 <ncopa> asap
2013-10-01 09:14:42 <ncopa> i'd like to make alpine 2.6.5 today
2013-10-01 09:24:56 <foxkit> I have run in to this bug
2013-10-01 09:25:09 <foxkit> unfortunately my fix for this was Ctrl-C at the host console :)
2013-10-01 09:26:02 <ncopa> i dont use xen myself
2013-10-01 09:26:16 <ncopa> either someone help me fix it today/now
2013-10-01 09:26:31 <ncopa> or i will have to set target for it to 2.6.6
2013-10-01 09:26:56 <foxkit> when does the release happen?
2013-10-01 09:27:20 <foxkit> Because, I am busy now, but I would like to investigate #2250 if I can
2013-10-01 09:27:34 <foxkit> I ran in to this bug in vmware but not on xen
2013-10-01 09:27:42 <ncopa> release will happen today
2013-10-01 09:28:15 <ncopa> hum
2013-10-01 09:28:16 <ncopa> ok
2013-10-01 09:28:23 <ncopa> that one would be nice to fix too yes
2013-10-01 09:36:09 <clandmeter> ncopa: will you do the redmine cherries?
2013-10-01 09:38:26 <ncopa> yeah
2013-10-01 09:38:28 <ncopa> i'm on it
2013-10-01 09:40:57 <ncopa> argh...
2013-10-01 09:41:02 <ncopa> conflicts
2013-10-01 09:41:19 <ncopa> 6d92032ebe03f60accf1f402af18ba2ef6fbfda3 does not apply
2013-10-01 09:42:27 <clandmeter> does not apply?
2013-10-01 09:42:52 <clandmeter> thats a new dir right?
2013-10-01 09:42:54 <ncopa> i dunno
2013-10-01 09:42:59 <ncopa> i did a loop
2013-10-01 09:43:11 <ncopa> i dont have the patience to do them manually
2013-10-01 09:43:17 <clandmeter> :)
2013-10-01 09:43:22 <ncopa> + manual means human-error
2013-10-01 09:44:01 <ncopa> 58355231bfd85879c53382eec0679b2c59af19b6
2013-10-01 09:44:03 <ncopa> is the bad commit
2013-10-01 09:44:18 <ncopa> it also increases pkgrel to redmine package
2013-10-01 09:44:49 <ncopa> oh
2013-10-01 09:45:00 <ncopa> i need apply the redmine commit first?
2013-10-01 09:45:05 <ncopa> c43f9c9064ea9de7f532f30bbc3179875a1cde1d
2013-10-01 09:45:08 <ncopa> i'm confused...
2013-10-01 09:45:25 <ncopa>  c43f9c9064ea9de7f532f30bbc3179875a1cde1d
2013-10-01 09:45:28 <ncopa> does not apply either
2013-10-01 09:45:42 <ncopa> sigh....
2013-10-01 09:46:08 <clandmeter> i thought i had the correct order
2013-10-01 09:46:13 <clandmeter> there were so many...
2013-10-01 09:46:20 <ncopa> exactly...
2013-10-01 09:46:37 <clandmeter> and i did it manually ;-)
2013-10-01 09:46:41 <ncopa> shoul i start with the redmine commit?
2013-10-01 09:46:45 <ncopa> http://git.alpinelinux.org/cgit/aports/commit/?id=c43f9c9064ea9de7f532f30bbc3179875a1cde1d
2013-10-01 09:47:06 <ncopa> i think it looks wrong end to start as the deps it adds does not exist
2013-10-01 09:47:12 <clandmeter> doesnt look like it
2013-10-01 09:47:12 <ncopa> -depends="ruby-rails ruby-rmagick ruby-i18n ruby-openid ruby-rack
2013-10-01 09:47:12 <ncopa> - ruby-mocha ruby-jquery-rails ruby-coderay ruby-fastercsv
2013-10-01 09:47:12 <ncopa> +depends="ruby-redmine-rails ruby-rmagick ruby-i18n ruby-openid ruby-rack
2013-10-01 09:47:12 <ncopa> + ruby-mocha ruby-redmine-jquery-rails ruby-coderay ruby-fastercsv
2013-10-01 09:47:38 <clandmeter> let me check history where it needs to go
2013-10-01 09:48:21 <clandmeter> looks like i did it aftre b70464107539efcad5b510684badbe25431aa18b
2013-10-01 09:49:57 <clandmeter> ncopa: you want me to make a proper list based on history?
2013-10-01 09:50:49 <ncopa> maybe you could give me a git repo to pull from?
2013-10-01 09:51:20 <clandmeter> you want me to apply them to 2.6 local?
2013-10-01 09:51:54 <ncopa> yes
2013-10-01 09:52:04 <ncopa> and put is some place where i can pull it
2013-10-01 09:52:30 <clandmeter> i could mail them ;-)
2013-10-01 09:55:08 <ncopa> i think i have it now
2013-10-01 09:56:06 <clandmeter> ok them ill wait
2013-10-01 11:10:49 <ncopa> clandmeter: ruby-redmine-tzdata was missing
2013-10-01 11:10:57 <ncopa> and redmine tarball gives 404
2013-10-01 11:11:39 <ncopa> it also seems that redmine-2.3.3 is out
2013-10-01 11:11:43 <ncopa> should we use that?
2013-10-01 11:19:45 <clandmeter> ncopa: sure
2013-10-01 11:39:44 <ncopa> clandmeter: could you please test if redmine is ok in v2.6 once the builders are done?
2013-10-01 11:40:07 <clandmeter> yes.
2013-10-01 11:40:25 <clandmeter> I was about to make an additional commit
2013-10-01 11:40:48 <clandmeter> i dont like the configs in our repo
2013-10-01 11:54:47 <clandmeter> is it me or is nl.a.o repo slow?
2013-10-01 11:55:30 <ncopa> nl.a.o is building kernels
2013-10-01 11:55:35 <ncopa> so its llikely slow
2013-10-01 11:55:47 <clandmeter> it is?
2013-10-01 11:55:58 <clandmeter> since when?
2013-10-01 11:56:20 <ncopa> 5mins ago?
2013-10-01 11:56:33 <clandmeter> since when does it preform that task?
2013-10-01 11:56:34 <ncopa> oh
2013-10-01 11:56:44 <ncopa> nl.a.o is not dev.a.o....
2013-10-01 11:56:46 <ncopa> sorry
2013-10-01 11:56:46 <ncopa> no
2013-10-01 11:56:48 <clandmeter> :p
2013-10-01 11:56:54 <ncopa> it is not building kernels then :)
2013-10-01 11:59:34 <clandmeter> dont commit too much, it makes you dizzy and say strange things.
2013-10-01 12:01:32 <ncopa> yeah...
2013-10-01 12:01:43 <ncopa> specially if you patch 2 different kernels at same time...
2013-10-01 12:03:13 <clandmeter> om
2013-10-01 12:03:14 <clandmeter> omg
2013-10-01 12:03:30 <clandmeter> im editing bugs.a.o instead of using it as an example...
2013-10-01 12:04:44 <clandmeter> ncopa: if it stops working it was me :|
2013-10-01 12:05:03 <ncopa> lol
2013-10-01 12:06:56 <clandmeter> ok bugs is save :)
2013-10-01 12:07:01 <clandmeter> safe even
2013-10-01 12:11:18 <clandmeter> bugger
2013-10-01 12:11:31 <clandmeter> ncopa: i cannot update a pkg from pinned repo?
2013-10-01 12:12:29 <clandmeter> ah i can
2013-10-01 12:20:14 <clandmeter> ncopa: seems rack in 2.6 is higher then edge :)
2013-10-01 12:20:51 <clandmeter> I will need to make a new pkg for it and add it to deps
2013-10-01 12:25:08 <clandmeter> ncopa: we need ruby-i18n from edge in 2.6
2013-10-01 12:29:13 <clandmeter> ncopa: and please cherry http://tinyurl.com/odg72yp
2013-10-01 12:29:37 <clandmeter> oh i need to change dep
2013-10-01 12:37:18 <clandmeter> rack is a pita
2013-10-01 13:14:15 <clandmeter> ncopa: i think i finally got it, but apk segfaults :(
2013-10-01 13:14:37 <clandmeter> i guess it has something to do with deps
2013-10-01 13:37:52 <ncopa> clandmeter: try upgrade apk-tools ;)
2013-10-01 13:39:12 <build-2-4> files from v2.4.11-98-g50fc6df uploaded
2013-10-01 13:44:26 <clandmeter> ncopa fabled , http://sprunge.us/BKfU
2013-10-01 13:44:44 <fabled> clandmeter, fixed in just tagged apk
2013-10-01 13:44:59 <ncopa> apk add -U -u apk-tools
2013-10-01 13:45:06 <fabled> ncopa, need to bump apkbuild
2013-10-01 13:45:09 <ncopa> oh
2013-10-01 13:45:10 <ncopa> ok
2013-10-01 13:45:15 <clandmeter> lol
2013-10-01 13:45:34 <ncopa> v2.4.0 right?
2013-10-01 13:45:38 <fabled> yes
2013-10-01 13:46:59 <clandmeter> fabled: is it including latte macchiato?
2013-10-01 13:48:35 <ncopa> apk fetch coffee
2013-10-01 13:49:06 <clandmeter> i dont want to get it myslef
2013-10-01 13:49:10 <clandmeter> im lazy!
2013-10-01 13:49:50 <clandmeter> algitbot: hurry up
2013-10-01 13:52:36 <clandmeter> \o/
2013-10-01 14:04:36 <clandmeter> ncopa: rack done
2013-10-01 14:05:20 <ncopa> are those needed too?
2013-10-01 14:05:24 <clandmeter> another 10 cherries
2013-10-01 14:05:31 <ncopa> for 2.6-stable?
2013-10-01 14:05:41 <clandmeter> yes if you want to make it work :|
2013-10-01 14:05:52 <ncopa> oh well
2013-10-01 14:06:04 <ncopa> looks like were gonna miss our release today....
2013-10-01 14:07:28 <clandmeter> ncopa: those commits i pushed last 2 hours and we need updated ruby-i18n
2013-10-01 14:09:20 <ncopa> ruby-redmine-sprockets does not apply cleanly
2013-10-01 14:09:33 <ncopa> are there more changes there that needs to go in
2013-10-01 14:10:36 <ncopa> nah
2013-10-01 14:11:29 <clandmeter> got it?
2013-10-01 14:28:36 <ncopa> >>> ruby-redmine-rack: Entering fakeroot...
2013-10-01 14:28:36 <ncopa>         from /usr/lib/ruby/2.0.0/rubygems.rb:15:in `<top (required)>'
2013-10-01 14:28:36 <ncopa>         from <internal:gem_prelude>:1:in `require'
2013-10-01 14:28:36 <ncopa>         from <internal:gem_prelude>:1:in `<compiled>'
2013-10-01 14:29:46 <ncopa> arch
2013-10-01 14:29:54 <ncopa> all those needs ruby-gems as depends
2013-10-01 14:35:49 <ncopa> ok should be ok now clandmeter
2013-10-01 14:40:49 <ncopa> clandmeter: redmine is ok now?
2013-10-01 14:41:41 <clandmeter> did you also update ruby-i18n?
2013-10-01 14:43:52 <clandmeter> ncopa: ^^
2013-10-01 14:45:47 <clandmeter> grr unicorn...
2013-10-01 14:45:58 <clandmeter> rack is more painfull then going to the dentist
2013-10-01 14:48:25 <clandmeter> ncopa: im going to concider unicorn as alpine's default way to run redmine. so ruby-redmine-unicorn is coming (because of dep on rack)
2013-10-01 14:49:40 <ncopa> sigh
2013-10-01 14:49:47 <ncopa> alredy broke the dealine...
2013-10-01 14:49:50 <ncopa> deadline
2013-10-01 14:50:08 <ncopa> i now have 11mins to fix the remining bugs and do the release...
2013-10-01 14:50:21 <ncopa> the release takes normally 30mins or so... :-/
2013-10-01 14:50:36 <ncopa> so tomorrow morning will be release
2013-10-01 14:50:47 <ncopa> and then i will just push whatever we have
2013-10-01 14:54:08 <clandmeter> ncopa: do those 2 commits please, and if i find something else i can do it myself
2013-10-01 16:37:39 <clandmeter> ncopa: http://git.alpinelinux.org/cgit/clandmeter/aports/?h=2.6-stable
2013-10-02 06:33:52 <fcolista> clandmeter, have you ever tryied freevo?
2013-10-02 07:19:18 <clandmeter> fcolista: hi
2013-10-02 07:19:26 <clandmeter> fcolista: isnt that us only?
2013-10-02 07:19:43 <clandmeter> oh thats software
2013-10-02 07:22:08 <clandmeter> ncopa: will you upgrade those 2 ruby pkgs?
2013-10-02 07:28:42 <ncopa> i think i just did
2013-10-02 07:30:51 <clandmeter> i dont see them in 2.6
2013-10-02 07:31:39 <clandmeter> 41b5ca0faed1f8404b95766c1806f61307deaeb2 and 4b0b5e2da2dc27ac1ecde4c83def01cd93ba7718
2013-10-02 07:38:17 <fcolista> freevo the mediacenter, clandmeter
2013-10-02 07:38:29 <fcolista> did you ever had a chance to take a look to it?
2013-10-02 07:49:17 <clandmeter> fcolista: no i dont think so
2013-10-02 07:50:47 <fcolista> ok
2013-10-02 07:51:10 <fcolista> btw freevo is looking for a live distro
2013-10-02 08:13:25 <clandmeter> fcolista: are you interested in it?
2013-10-02 08:13:37 <clandmeter> is it any better then xbmc?
2013-10-02 08:14:03 <clandmeter> ncopa: did you see the 2 additional cherries?
2013-10-02 08:15:46 <ncopa> 'main/ruby-redmine-unicorn: new aport' and 'main/ruby-redmine-unicorn: add missing files'?
2013-10-02 08:15:58 <ncopa> or were there even more?
2013-10-02 08:16:06 <clandmeter> there are 2 more
2013-10-02 08:16:16 <clandmeter> did you check my 2.6 branch?
2013-10-02 08:16:24 <ncopa> nope
2013-10-02 08:16:26 <ncopa> I'll do
2013-10-02 08:16:36 <clandmeter> http://git.alpinelinux.org/cgit/clandmeter/aports/?h=2.6-stable
2013-10-02 08:19:49 <ncopa> super!
2013-10-02 08:19:50 <ncopa> thanks!§
2013-10-02 08:34:59 <fabian_a> please git pull git://git.alpinelinux.org/fab/aports
2013-10-02 08:35:29 <fabian_a> this pull requests contains some sphinxcontrib stuff
2013-10-02 08:37:16 <ncopa> fabian_a: i will in a bit... need to get v2.6.5 out the door...
2013-10-02 08:37:29 <ncopa> fabian_a: is it rebased?
2013-10-02 08:42:34 <fabian_a> ncopa: it can wait, there is no hurry
2013-10-02 08:54:13 <clandmeter> ncopa: redmine is working out of the box now :)
2013-10-02 08:56:58 <ncopa> clandmeter: wonderful!
2013-10-02 08:57:05 <ncopa> thanks a bunch for cleaning it up!
2013-10-02 08:57:23 <clandmeter> only part i dont like is ruby-redmine-unicorn
2013-10-02 08:59:57 <ncopa> i wonder what will happen if ruby-redmine-<gem> and ruby-<gem> happens to be exactly same version
2013-10-02 09:00:21 <clandmeter> not much
2013-10-02 09:00:32 <clandmeter> one would replace the other on fs i guess
2013-10-02 09:00:43 <clandmeter> if you install them both.
2013-10-02 09:04:11 <clandmeter> but that will of course error because of apk
2013-10-02 11:30:10 <barthalion> algitbot: don't ignore my commits!
2013-10-02 12:01:55 <StarWarsFan> ncopa: tried to build openjdk
2013-10-02 12:02:05 <StarWarsFan> but "abuild -r" just do nothing!?
2013-10-02 12:02:51 <ncopa> i have disabled it. edit the APKBUILD
2013-10-02 12:02:58 <ncopa> set arch=all
2013-10-02 12:03:11 <StarWarsFan> ah, that's the magic :)
2013-10-02 12:03:16 <ncopa> i think its 'x86' only now
2013-10-02 12:21:39 <ncopa> xl shutdown appears to be pretty fragile
2013-10-02 13:34:19 <StarWarsFan> ncopa: how can i see which patch is available for which version of the used components on openjdk6?
2013-10-02 13:34:39 <StarWarsFan> or other way around: which patch is necessary for which version?
2013-10-02 13:34:47 <ncopa> i dunno
2013-10-02 13:36:35 <StarWarsFan> hm, looks like as only timo knows it...
2013-10-02 13:37:04 <StarWarsFan> because there are newer versions available
2013-10-02 13:37:10 <StarWarsFan> but if i try to use them,
2013-10-02 13:37:21 <StarWarsFan> the patches won't be appliable anymore
2013-10-02 13:37:29 <StarWarsFan> :-/
2013-10-02 13:42:28 <fabled> which patches?
2013-10-02 14:02:27 <clandmeter> where can i find the ARM release?
2013-10-02 14:02:32 <clandmeter> 
ACTION hides

2013-10-02 14:02:49 <ncopa> 
ACTION tries to break clandmeter's ARM

2013-10-02 14:03:23 <clandmeter> ARM Fight!
2013-10-02 14:05:26 <fabled> clandmeter, on my wandboard
2013-10-02 14:06:06 <clandmeter> can I have a pre-Alpha rpi version pretty please?
2013-10-02 14:06:35 <fabled> maybe next week
2013-10-02 14:06:50 <fabled> i think i just figured out the reason for 90% of random build failures
2013-10-02 14:06:57 <fabled> both on x86 and arm
2013-10-02 14:07:09 <fabled> the pie patch ncopa committed to kernels should fix it
2013-10-02 14:07:10 <clandmeter> the gcc thingy?
2013-10-02 14:07:14 <fabled> also that
2013-10-02 14:07:29 <fabled> it seems to be gone now
2013-10-02 14:08:01 <clandmeter> good, ill keep my sdcard formated
2013-10-02 14:09:19 <fabled> i have built now about 75% of alpine main against musl
2013-10-02 14:10:00 <fabled> i think the only thing to do is fix lxc before i attempt rebuild on arm
2013-10-02 14:10:53 <fabled> reminds me
2013-10-02 14:11:06 <fabled> ncopa, i'd like the build.a.o box kernel updated to have the above mentioned pie fix
2013-10-02 14:11:06 <clandmeter> rpi will be minimul arm supported?
2013-10-02 14:11:43 <fabled> i think so. if someone really wants / needs, we can consider doing armel build supporting older boards too. but i'm not too enthusiastic since they tend to need board specific kernel.
2013-10-02 14:11:58 <fabled> i'm off for today
2013-10-02 14:12:04 <fabled> back tm
2013-10-02 14:12:31 <StarWarsFan> fabled: these ones: http://git.alpinelinux.org/cgit/aports/tree/main/openjdk6
2013-10-02 14:25:16 <ncopa> bah new kernels are out
2013-10-02 14:25:29 <ncopa> they are faster in shipping new kernels than i am to update...
2013-10-02 14:30:34 <grharry2> Is it possible to run 32bits compiled binaries on a 64bit alpine ?? if yes what ( apk or whatever ) is needed ??
2013-10-02 14:32:34 <clandmeter> grharry2: afaik we dont support multiarch
2013-10-02 14:33:56 <clandmeter> but ncopa can probably provide a better answer
2013-10-02 14:35:02 <ncopa> grharry2: no, no support for multiarch
2013-10-02 14:35:26 <ncopa> only way to do it is in a container (lxc)
2013-10-02 14:35:47 <grharry2> got it ..
2013-10-02 14:35:51 <grharry2> thnx
2013-10-02 14:40:35 <fabled> StarWarsFan, i think i wrote only one of them
2013-10-02 14:58:55 <clandmeter> ncopa: before you go
2013-10-02 14:59:03 <clandmeter> im want to push new php
2013-10-02 14:59:41 <clandmeter> but xcache has release yet
2013-10-02 14:59:47 <clandmeter> but in svn it has support
2013-10-02 15:00:31 <clandmeter> in arch version number is: 3.1.0dev-14
2013-10-02 15:01:58 <ncopa> i suppose its ok if you figure it out
2013-10-02 15:02:03 <ncopa> gotta go now
2013-10-02 16:43:39 <shafire> Hello
2013-10-02 16:43:52 <shafire> anything new?
2013-10-02 16:45:18 <shafire> I see, 2.6.5 :)
2013-10-02 17:00:10 <barthalion> clandmeter: yeah, it shouldn't look like that
2013-10-02 17:02:00 <barthalion> clandmeter: we have some unwritten policy about cases like this
2013-10-02 17:02:14 <barthalion> clandmeter: it should be rather x.y.z.<svn-rev>
2013-10-02 17:02:24 <barthalion> clandmeter: and I think it's a good option for us
2013-10-02 18:16:36 <clandmeter> barthalion: i will take abuild svn versioning
2013-10-02 18:38:18 <clandmeter> barthalion: what about apc?
2013-10-02 18:50:36 <vkrishn> barthalion: you around ?
2013-10-02 18:51:02 <vkrishn> could you update sphinxsearch and fossil pls
2013-10-02 18:52:28 <vkrishn> clandmeter: re: apc, pls check
2013-10-02 18:52:30 <vkrishn> https://github.com/krakjoe/apcu
2013-10-02 18:52:57 <clandmeter> i got it all build, push it tomorrow.
2013-10-02 18:53:13 <vkrishn> :)
2013-10-02 20:50:10 <clandmeter> barthalion: hope i didnt break it :)
2013-10-02 20:50:57 <clandmeter> didnt really do any testing. will check that tomorrow (or if anybody wants to try by my guest)
2013-10-02 20:52:16 <shafire> do you use php-xcache?
2013-10-02 20:52:41 <clandmeter>  no
2013-10-02 20:52:47 <shafire> or rather php-apc
2013-10-02 20:53:02 <clandmeter> we do use apc on our www
2013-10-02 20:53:13 <clandmeter> but i guess we dont need it anymore
2013-10-02 20:53:20 <clandmeter> opcache will be prefered
2013-10-02 20:53:34 <clandmeter> not sure of apcu will bring any adv over opcache
2013-10-02 20:53:46 <shafire> opcache is in php, or?
2013-10-02 20:53:52 <clandmeter> yes
2013-10-02 20:54:07 <shafire> ok
2013-10-02 20:54:08 <clandmeter> since 5.5
2013-10-02 20:54:18 <clandmeter> im using it on my dev env
2013-10-02 20:54:33 <clandmeter> but that is windows
2013-10-02 20:55:06 <clandmeter> I think i need to take a closer look later on php apkbuild
2013-10-02 20:55:28 <clandmeter> maybe add a initd for php's internal webserver
2013-10-02 20:55:43 <clandmeter> makes sence if you only use it for dev things
2013-10-02 21:03:07 <shafire> ok
2013-10-02 21:03:24 <shafire> does algitbot have something like !seen?
2013-10-03 07:28:13 <clandmeter> ncopa: small shell question
2013-10-03 07:28:53 <clandmeter> ncopa: "${2-}" means second arg including following?
2013-10-03 07:30:45 <ncopa> dunno
2013-10-03 07:31:28 <ncopa> google posixh shell expansion
2013-10-03 07:31:41 <ncopa> http://pubs.opengroup.org/onlinepubs/009695399/utilities/xcu_chap02.html
2013-10-03 07:31:57 <ncopa> 2.6.2 Parameter Expansion
2013-10-03 07:32:21 <ncopa> http://pubs.opengroup.org/onlinepubs/009695399/utilities/xcu_chap02.html#tag_02_06_02
2013-10-03 07:33:02 <ncopa> ${parameter-word}
2013-10-03 07:33:02 <ncopa> 	
2013-10-03 07:33:02 <ncopa> substitute null
2013-10-03 07:33:02 <ncopa> substitute parameter
2013-10-03 07:33:02 <ncopa> 	
2013-10-03 07:33:07 <ncopa> 	
2013-10-03 07:33:09 <ncopa> substitute word
2013-10-03 07:37:16 <clandmeter> im probably too stupid to read that. there are multiple ways to rome :)
2013-10-03 07:38:48 <ncopa> ${2-}
2013-10-03 07:38:52 <ncopa> means second arg
2013-10-03 07:39:42 <clandmeter> then why not use $2?
2013-10-03 07:39:52 <ncopa> if there are passed an empty arg (arg ""), then it returns ""
2013-10-03 07:40:20 <ncopa> if the second arg is not passed ( $2 is unset)
2013-10-03 07:40:38 <ncopa> then it will substitute with whats after -
2013-10-03 07:40:42 <ncopa> in this case, empty
2013-10-03 07:40:52 <ncopa> so ${2-} means
2013-10-03 07:41:13 <ncopa> give second arg or "" if there are no second arg
2013-10-03 07:41:20 <ncopa> but never return "unset"
2013-10-03 07:41:27 <clandmeter> ah ok
2013-10-03 07:43:39 <clandmeter> thx
2013-10-03 07:44:11 <clandmeter> it seems rather hard for me to read such kind of documentation.
2013-10-03 07:45:38 <fabled> i think lxc works on arm/musl after all
2013-10-03 07:45:46 <fabled> the utmp stuff is not used unless it's old kernel
2013-10-03 07:45:57 <fabled> i think i can fire up building world on arm today
2013-10-03 07:46:02 <fabled> and see how it goeth
2013-10-03 07:46:15 <clandmeter> \o/
2013-10-03 07:48:15 <clandmeter> fabled: so you will run lxc on wand?
2013-10-03 07:48:19 <fabled> yes
2013-10-03 07:48:31 <clandmeter> for what reason?
2013-10-03 07:48:40 <fabled> in case i want to run a dev-arm suite too
2013-10-03 07:48:42 <clandmeter> if you want to build for antoher cpu?
2013-10-03 07:48:48 <fabled> possibly that too
2013-10-03 07:48:54 <fabled> and it's good practice
2013-10-03 07:49:01 <clandmeter> :)
2013-10-03 07:49:02 <fabled> i can just move the hard disk to new box later if needed
2013-10-03 07:49:07 <fabled> etc
2013-10-03 07:49:15 <clandmeter> you can run 3 builds simultaiously :)
2013-10-03 07:49:29 <clandmeter> poor wand board
2013-10-03 07:51:04 <fabled> hehe
2013-10-03 07:52:40 <ncopa> wanderful!
2013-10-03 07:53:42 <fabled> just need networking working inside lxc
2013-10-03 07:53:49 <fabled> it sends dhcp packets out
2013-10-03 07:53:54 <fabled> but does not seem to get packet in
2013-10-03 07:54:05 <ncopa> macvlan or bridge?
2013-10-03 07:54:06 <fabled> is dhcp supposed to work in the lxc client (veth + bridge setup)
2013-10-03 07:54:11 <ncopa> yes
2013-10-03 07:54:14 <fabled> hum
2013-10-03 07:54:18 <ncopa> but you might need set hwaddr
2013-10-03 07:54:35 <fabled> ok?
2013-10-03 07:54:39 <fabled> vethLFODkJ Link encap:Ethernet  HWaddr FE:8E:FD:1B:15:30
2013-10-03 07:54:52 <ncopa> looks good
2013-10-03 07:54:59 <fabled> inside lxc
2013-10-03 07:55:00 <fabled> eth0      Link encap:Ethernet  HWaddr FE:F1:C5:3D:E5:E4
2013-10-03 07:55:20 <ncopa> and br0 mac addr?
2013-10-03 07:55:37 <fabled> and dhcp server sees: 10:15:36.077671 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP, Request from fe:f1:c5:3d:e5:e4 (oui Unknown), length 280
2013-10-03 07:55:59 <fabled> br0 mac is clone of physical eth0
2013-10-03 07:56:09 <fabled> so that looks ok
2013-10-03 07:56:43 <fabled> aha
2013-10-03 07:56:49 <fabled> setting physical eth0 to promisc made it work
2013-10-03 07:57:37 <fabled> seems to work now
2013-10-03 07:57:46 <fabled> sounds like bug in kernel 'fec' nic driver
2013-10-03 07:57:56 <fabled> some sort of mac address filtering issue
2013-10-03 08:05:46 <build64-2-4> files from v2.4.11-102-g520b6ed uploaded
2013-10-03 08:19:02 <clandmeter> ncopa: got a min for pm?
2013-10-03 09:55:22 <ncopa> rebooting build server
2013-10-03 09:55:24 <ncopa> oh...
2013-10-03 09:55:27 <ncopa> stupid me...
2013-10-03 09:55:30 <ncopa> it was building...
2013-10-03 10:02:55 <fabled> 2.4.0-r0 had the issue during boot
2013-10-03 10:03:12 <ncopa> apk cache does not work
2013-10-03 10:03:18 <ncopa> i did apk cache sync
2013-10-03 10:03:26 <ncopa> but it did not install a bunch of packages
2013-10-03 10:03:32 <ncopa> including unbound
2013-10-03 10:03:45 <fabled> was initramfs updated after apk-tools update?
2013-10-03 10:04:06 <ncopa> i think so
2013-10-03 10:04:22 <ncopa> this box has had this issue long time
2013-10-03 10:04:27 <ncopa> i dont know why it happens
2013-10-03 10:04:47 <fabled> perhaps the overlay device is named different during initramfs
2013-10-03 10:04:49 <ncopa> i always thought it was due to i didnt do apk cache sync
2013-10-03 10:04:56 <ncopa> could be
2013-10-03 10:05:28 <fabled> cache -> /media/UUID=4DA6-BEDC/cache
2013-10-03 10:06:03 <ncopa> but no packages there?
2013-10-03 10:06:40 <fabled>  you have /vservers/build64-2-6/home/buildozer/packages/main
2013-10-03 10:06:50 <fabled> it does not cache packages on local disk
2013-10-03 10:06:55 <fabled> but that path is not available in initramfs
2013-10-03 10:07:00 <ncopa> oh!
2013-10-03 10:07:02 <ncopa> that explains
2013-10-03 10:07:09 <ncopa> makes sense
2013-10-03 10:07:27 <fabled> should maybe have it cache packages not in boot-media
2013-10-03 10:07:46 <fabled> file a ticket?
2013-10-03 10:07:52 <ncopa> maybe
2013-10-03 10:07:58 <ncopa> or i just use it via http
2013-10-03 10:08:02 <ncopa> lunch here now
2013-10-03 10:08:31 <fabled> yes, that would work too
2013-10-03 10:08:33 <fabled> bon apetit
2013-10-03 10:09:36 <ncopa> algitbot: build master
2013-10-03 10:12:04 <clandmeter> ncopa: log rotation on git.a.o could be nice.
2013-10-03 10:12:31 <clandmeter> -rw-r--r--    1 lighttpd lighttpd    1.5G Oct  3 10:12 /var/log/lighttpd/access.log
2013-10-03 10:12:34 <clandmeter> :)
2013-10-03 10:17:40 <fabled> pie patch seems to have fixed the random llvm rebuild error
2013-10-03 10:17:50 <fabled> i'm pretty sure it fixes (some) of the random openjdk build errors too
2013-10-03 10:17:55 <fabled> especially on x86
2013-10-03 11:38:27 <fabled> i started build of world in build-edge-musl-arm
2013-10-03 11:59:02 <clandmeter> algitbot: build master
2013-10-03 12:00:05 <clandmeter> ncopa: shouldnt they fininsh the kernel builds?
2013-10-03 12:10:21 <ncopa> hm
2013-10-03 12:10:25 <ncopa> they should yes
2013-10-03 12:11:04 <ncopa> but apparently they dont
2013-10-03 12:11:14 <ncopa> algitbot: build master
2013-10-03 12:11:29 <ncopa> ok
2013-10-03 12:11:34 <ncopa> i think i know what happens
2013-10-03 12:11:38 <ncopa> its a bootup race
2013-10-03 12:11:51 <ncopa> and zeromq that behaves unexpected
2013-10-03 12:39:34 <ncopa> kaniini: do you mind take #2252 ?
2013-10-03 12:39:54 <ncopa> i had a short look at it, i tested xendomains stop/start
2013-10-03 12:40:14 <ncopa> it looks like the init.d script needs some love
2013-10-03 12:40:19 <ncopa> or a rewrite
2013-10-03 13:15:35 <build64-2-3> files from v2.3.6-265-g82f5fcb uploaded
2013-10-03 13:15:48 <build-2-3> files from v2.3.6-265-g82f5fcb uploaded
2013-10-03 13:16:12 <build64-2-4> files from v2.4.11-102-g520b6ed uploaded
2013-10-03 13:16:38 <build-2-4> files from v2.4.11-102-g520b6ed uploaded
2013-10-03 13:16:58 <build64-2-5> files from v2.5.4-188-g742b9d6 uploaded
2013-10-03 13:23:20 <ncopa> did anyone test samba-4?
2013-10-03 14:01:37 <fcolista> ncopa, looks that i've managed to compile mod_perl 2.0.8 for apache 2.4
2013-10-03 14:01:52 <fcolista> when you have time, can you check if this is ok?
2013-10-03 14:01:53 <fcolista> http://sprunge.us/HLHi
2013-10-03 14:02:02 <fcolista> http://sprunge.us/eRZT <--- patch added
2013-10-03 14:03:35 <ncopa> hm
2013-10-03 14:03:40 <ncopa> if it builds, then i say push it
2013-10-03 14:04:13 <fcolista> it builds, do not return remote_ip error in Connection.xs
2013-10-03 14:04:37 <ncopa> i havent really studied how the client_ip works in comparison to remote_ip
2013-10-03 14:04:44 <ncopa> if it builds, just push it
2013-10-03 14:04:47 <ncopa> its in testing
2013-10-03 14:04:53 <fcolista> ok
2013-10-03 14:05:03 <ncopa> good work!
2013-10-03 14:05:18 <fcolista> http://www.perlmonks.org/?node_id=1042916
2013-10-03 14:05:22 <fabled> i did mass update of cpan modules
2013-10-03 14:05:30 <ncopa> fabled: thanks!
2013-10-03 14:05:40 <fabled> it was 99% automated
2013-10-03 14:05:48 <fabled> there was 3 modules that i skipped since they had funny version
2013-10-03 14:06:24 <fcolista> i like this comment "Is Apache 2.4 really necessary?" :)
2013-10-03 14:06:25 <fcolista> lol
2013-10-03 14:06:33 <fcolista> http://www.perlmonks.org/?node_id=981290
2013-10-03 14:06:34 <ncopa> lol
2013-10-03 14:06:54 <fcolista> if it won't works, instead of fixing it, just don't use :)
2013-10-03 14:07:35 <ncopa> yeah...
2013-10-03 14:07:41 <ncopa> "is perl really necessary?"
2013-10-03 14:07:47 <fcolista> lol
2013-10-03 14:15:05 <fcolista> algitbot: hi
2013-10-03 14:18:42 <fcolista> is
2013-10-03 14:19:02 <fcolista> ncopa, is buildbot building?
2013-10-03 14:19:57 <ncopa> i think builders are busy yes
2013-10-03 14:20:03 <ncopa> but something is not right..
2013-10-03 14:20:17 <fcolista> i'm asking becase i'm not seeing any log.
2013-10-03 14:20:18 <ncopa> buildmsg-subscribe is eating all cpu
2013-10-03 14:20:25 <fcolista> uh
2013-10-03 14:20:40 <ncopa> but something is indeed building
2013-10-03 14:20:50 <fcolista> k
2013-10-03 15:06:26 <ncopa> those perl packages needs fixing.. :-(
2013-10-03 15:08:46 <shafire> hi
2013-10-03 20:09:41 <shafire> ncopa is online?
2013-10-03 20:10:01 <clandmeter> well he is using internet
2013-10-03 20:10:11 <clandmeter> but i doubht he is behind his irc client :)
2013-10-03 20:11:07 <shafire> :-)
2013-10-04 02:21:55 <Mp5shooter> hmm
2013-10-04 02:21:57 <Mp5shooter> is uh
2013-10-04 02:22:11 <Mp5shooter> is the alpine mail client compiled with the patch that lets it get mail from a maildir?
2013-10-04 02:22:16 <Mp5shooter> anyone know?
2013-10-04 06:10:53 <ncopa> Mp5shooter: http://git.alpinelinux.org/cgit/aports/tree/main/alpine
2013-10-04 06:10:56 <ncopa> are the patches
2013-10-04 06:11:08 <ncopa> there is one big, all_p84
2013-10-04 06:11:16 <ncopa> i dont know if that includes what you ask for...
2013-10-04 06:11:43 <ncopa> if you have a link the maildir patch i'll add it
2013-10-04 07:25:58 <ncopa> fabled: do you have a sec for an apk-tools issue?
2013-10-04 07:26:16 <fabled> i think so, yes?
2013-10-04 07:26:29 <ncopa> http://sprunge.us/HCCF
2013-10-04 07:26:34 <ncopa> apk version says that ^^^
2013-10-04 07:26:39 <ncopa> apk upgrade -U -a does not upgrade
2013-10-04 07:26:50 <fabled> try apk upgrade --latest
2013-10-04 07:26:55 <fabled> might due to provides issues
2013-10-04 07:27:00 <ncopa> ncopa-desktop:~/test/java$ apk add --simulate 'libva-intel-driver>1.0.20-r0'
2013-10-04 07:27:00 <ncopa> (1/1) Upgrading libva-intel-driver (1.0.20-r0 -> 1.2.0-r0)
2013-10-04 07:27:00 <ncopa> OK: 3067 MiB in 1091 packages
2013-10-04 07:27:17 <ncopa> apk add --simulate says it should upgrade ...
2013-10-04 07:27:25 <fabled> try apk upgrade --latest
2013-10-04 07:27:48 <ncopa> OK: 3067 MiB in 1091 packages
2013-10-04 07:27:50 <ncopa> thats all it says
2013-10-04 07:27:59 <fabled> that's weird
2013-10-04 07:28:04 <ncopa> exactly... :)
2013-10-04 07:28:58 <ncopa> ok
2013-10-04 07:29:02 <ncopa> i foudn libgee issue
2013-10-04 07:29:15 <ncopa> https://dpaste.de/MUQR
2013-10-04 07:29:57 <fabled> ah
2013-10-04 07:30:07 <fabled> yes, --latest would not get that
2013-10-04 07:30:19 <fabled> mostly those are due to provides issues
2013-10-04 07:30:23 <fabled> some soname changed
2013-10-04 07:30:30 <ncopa> libva-intel-driver
2013-10-04 07:30:32 <fabled> and there's indirect dependency on the older version
2013-10-04 07:30:36 <fabled> via the old soname
2013-10-04 07:31:01 <fabled> i wonder if i should change that
2013-10-04 07:31:25 <fabled> that --latest would force always the latest version of the package name, and not the virtual  name
2013-10-04 07:33:04 <fabled> arm builder is at package 1153 of 1669
2013-10-04 07:35:16 <fabled> it's been working all night :)
2013-10-04 07:41:50 <ncopa> i think its time to start focus on 2.7 release
2013-10-04 07:42:09 <ncopa> its only 1 month left
2013-10-04 07:42:12 <fabled> mmm
2013-10-04 07:42:15 <ncopa> we have tons of broken stuff
2013-10-04 07:42:18 <fabled> yes
2013-10-04 07:42:24 <fabled> and i'd like new qt in
2013-10-04 07:42:27 <ncopa> yeah
2013-10-04 07:42:29 <ncopa> those things
2013-10-04 07:42:32 <ncopa> we should do now
2013-10-04 07:42:39 <fabled> vlc needs fixing
2013-10-04 07:42:45 <ncopa> yes
2013-10-04 07:42:50 <ncopa> qt5, samba4
2013-10-04 07:42:54 <ncopa> what more?
2013-10-04 07:43:03 <ncopa> newer poppler
2013-10-04 07:43:26 <ncopa> we have a bunch of bugs on bugs.a.o
2013-10-04 07:43:37 <ncopa> we need start move things for 2.8 target
2013-10-04 07:43:49 <ncopa> i'd like to reconfig kernel too
2013-10-04 07:43:55 <ncopa> for v2.7
2013-10-04 07:44:05 <ncopa> there was a few things i disabled
2013-10-04 07:44:12 <fabled> ok?
2013-10-04 07:44:22 <ncopa> i dont remmember details
2013-10-04 07:44:45 <ncopa> but it was due to we jumped to 3.10 in v2.6-stable
2013-10-04 07:45:13 <ncopa> this is a bit weird: https://dpaste.de/0gRT
2013-10-04 07:45:29 <ncopa> audacious-plugins is pulled i by install_if
2013-10-04 07:46:28 <fabled> hum
2013-10-04 07:46:31 <ncopa> audacious-plugins has: libcue
2013-10-04 07:46:34 <ncopa> in depends
2013-10-04 07:46:43 <ncopa> instead of so:libcue...
2013-10-04 07:47:36 <ncopa> i suppose a rebuild of audacious-plugins will 'solve' it
2013-10-04 07:49:26 <ncopa> yup
2013-10-04 07:49:32 <ncopa> libcue-1.4.0-r1 provides:
2013-10-04 07:49:35 <ncopa> (empty)
2013-10-04 07:49:51 <ncopa> libcue-1.4.0-r2 provides:
2013-10-04 07:49:51 <ncopa> so:libcue.so.1=1.0.4
2013-10-04 07:50:06 <ncopa> audacious-plugins has depends=libcue
2013-10-04 07:50:15 <fcolista> ncopa, i've upgraded rsyslog and added mongodb support, as requested in #2038
2013-10-04 07:50:19 <ncopa> audacious-plugins was pulled in by install_if
2013-10-04 07:50:21 <fcolista> http://sprunge.us/ahZG
2013-10-04 07:50:57 <ncopa> we dont have mongodb?
2013-10-04 07:51:08 <fcolista> no
2013-10-04 07:51:17 <fcolista> rsyslog support for mongodb
2013-10-04 07:51:23 <fcolista> we have mongodb :)
2013-10-04 07:51:28 <ncopa> not in main?
2013-10-04 07:51:59 <fcolista> rsyslog is in main
2013-10-04 07:52:03 <fcolista> mongodb in testing
2013-10-04 07:52:11 <fcolista> rsyslog has not mongodb support
2013-10-04 07:52:15 <ncopa> packages in main cannot depend on packages in testing
2013-10-04 07:52:44 <fcolista> k
2013-10-04 07:52:54 <fcolista> good to know
2013-10-04 07:53:00 <ncopa> if you would bootstrap a new arch (like musl)
2013-10-04 07:53:10 <ncopa> you dont want start with building things in testing
2013-10-04 07:53:12 <ncopa> but only in main
2013-10-04 07:53:22 <ncopa> and the package depending on package in testing will fail...
2013-10-04 07:53:30 <ncopa> mongodb is a pain
2013-10-04 07:53:47 <fcolista> yes, i saw what a nightmare is
2013-10-04 07:54:00 <ncopa> i dont feel for supporting it...
2013-10-04 07:54:08 <ncopa> if someoneelse wants, be my guest..
2013-10-04 07:54:13 <ncopa> but its a pain...
2013-10-04 07:54:45 <fcolista> i was looking at the bug/feature req in alpine
2013-10-04 07:55:06 <fcolista> and just started with it
2013-10-04 07:56:45 <fcolista> lol your comment on #1182
2013-10-04 07:57:19 <fcolista> Mp5shooter, have you tryied it ?
2013-10-04 07:57:57 <ncopa> v8 does not build currently
2013-10-04 08:10:30 <ncopa> fcolista: i'd down prio v8/mongodb
2013-10-04 08:10:38 <ncopa> really, its a mess
2013-10-04 08:10:50 <fcolista> np copa
2013-10-04 08:10:54 <fcolista> *ncopa
2013-10-04 08:10:59 <fcolista> it's not a priority
2013-10-04 08:12:30 <ncopa> http://bugs.alpinelinux.org/issues/2253
2013-10-04 08:12:33 <ncopa> looks fairly simple
2013-10-04 08:12:55 <ncopa> exfat-utils
2013-10-04 08:13:29 <fcolista> You already done it, right?
2013-10-04 08:18:34 <fabled> ncopa, do you think you can upgrade qt soonish (this or next week)?
2013-10-04 08:18:38 <fabled> or should i fix it for musl
2013-10-04 08:18:45 <fabled> qt 4.8, that is
2013-10-04 08:19:41 <fabled> i already have 1GB of packages for arm
2013-10-04 08:19:58 <fabled> should probably soon make a pre-alpha test images for wandboard and/or raspberry pi
2013-10-04 08:22:16 <mhavela> 
ACTION would love to test a raspberry pi image

2013-10-04 08:22:48 <fabled> the rpi would be mostly console-only. it needs additional packages for making the gpu side work.
2013-10-04 08:22:53 <fabled> and custom kernel
2013-10-04 08:23:08 <mhavela> console suits me just fine
2013-10-04 08:23:53 <mhavela> But I'm in no /critical/ need... it's just that I am curious about it :-)
2013-10-04 08:24:34 <mhavela> I would love to have 'lua' working on it
2013-10-04 08:24:42 <fabled> yes, lua works
2013-10-04 08:24:46 <mhavela> wonderful
2013-10-04 08:24:49 <fabled> perl and ruby and python too
2013-10-04 08:24:58 <fabled> and gcc, ada, fortran and obj-c
2013-10-04 08:25:04 <fabled> no Go though
2013-10-04 08:25:23 <mhavela> Does Tetris work?
2013-10-04 08:25:26 <mhavela> (just kidding)
2013-10-04 08:25:30 <fabled> have not tested.
2013-10-04 08:25:33 <fabled> :)
2013-10-04 08:26:30 <fcolista> i've a raspberry reaady for Alpine
2013-10-04 08:27:20 <fcolista> xbmc and iscsi initiator with rasp, then i can live happy forever...
2013-10-04 08:27:37 <mhavela> :-)
2013-10-04 08:27:58 <mhavela> Cool what Alpine Linux can do for a person :-D
2013-10-04 08:28:14 <ncopa> fabled: i'm gonna try do it today
2013-10-04 08:28:19 <fcolista> lol mhavela
2013-10-04 08:29:08 <ncopa> mhavela: apk add micro-tetris
2013-10-04 08:29:10 <ncopa> i bet it works
2013-10-04 08:29:22 <mhavela> Cool! Didn't know it existed
2013-10-04 08:30:35 <fcolista> # Maintainer: Natanael Copa <ncopa@alpinepinux.org>
2013-10-04 08:30:35 <fcolista> pkgname=micro-tetris
2013-10-04 08:30:38 <mhavela> weeeeee - it works \o/ My days are saved! Now I can sit in the serverroom all day and play tetris on my router without anybody knowing what I do...
2013-10-04 08:30:38 <fcolista> :-D
2013-10-04 08:30:56 <ncopa> lol
2013-10-04 08:31:32 <ncopa> i suppose you also want vlc/mplayer with libcaca + a bittorent client :)
2013-10-04 08:32:00 <fcolista> :)
2013-10-04 08:33:21 <mhavela> ncopa, did you get a chance to look at why pidgin doesn't support irc? (no hurry though)
2013-10-04 08:33:57 <ncopa> yes
2013-10-04 08:34:01 <ncopa> i think it got fixed
2013-10-04 08:34:07 <mhavela> 
ACTION tries

2013-10-04 08:34:24 <ncopa> pidgin-2.10.7-r2
2013-10-04 08:35:37 <mhavela> I have pidgin-2.10.7-r0  guess I have wrong repos
2013-10-04 08:37:05 <mhavela> yup - I was using 2.6/main repo (not 'edge/main')
2013-10-04 08:39:16 <mhavela> ncopa, I had not planned to use 'edge' on the host I'm on... but if I have to, I will...
2013-10-04 08:39:30 <mhavela> so I'm curious if the pidgin fix will be backported to 2.6
2013-10-04 08:39:41 <ncopa> once its tested i can backport it
2013-10-04 08:39:52 <mhavela> Ok.
2013-10-04 08:40:02 <mhavela> I tried just upgrading pidgin (from edge)
2013-10-04 08:40:11 <mhavela> but seems it needed some other dependencies too
2013-10-04 08:41:04 <mhavela> http://sprunge.us/QQWg
2013-10-04 08:41:31 <mhavela> I can try upgrading those too from edge
2013-10-04 08:42:03 <mhavela> Is there a way to 'apk add -u pidgin' in a way that it also upgrades depended packages?
2013-10-04 08:42:51 <fabled> mhavela, i believe that is the default behaviour for "add -u"
2013-10-04 08:43:04 <fabled> to update only single packages you do "fix -u"
2013-10-04 08:43:18 <mhavela> The 'sprunge' output I sent is a result of 'apk add -u pidgin'
2013-10-04 08:43:42 <mhavela> means that it didn't upgrade the other packages for some reason
2013-10-04 08:46:02 <mhavela> Let me guess... 'ncopa2' is @ pidgin
2013-10-04 08:48:23 <fabled> mhavela, because you mixed repos. you have certain packages requiring new libpng/libsasl/etc, and some the older
2013-10-04 08:48:58 <ncopa2> pidgin here yes
2013-10-04 08:49:27 <mhavela> Wonderful! That means I can stay on '2.6/main' for now
2013-10-04 08:50:08 <mhavela> Thanks for your help ncopa
2013-10-04 08:52:04 <clandmeter> lol
2013-10-04 08:52:11 <clandmeter> im building something with libusb
2013-10-04 09:02:31 <vkrishn> ncopa: any thought on riak, rethinkdb  ?
2013-10-04 09:02:41 <vkrishn> alternate for mongodb
2013-10-04 09:03:31 <vkrishn> would be nice to have a nosql in mains
2013-10-04 09:04:59 <vkrishn> is there any documentation on apk pinning feature ?
2013-10-04 09:05:38 <clandmeter> vkrishn: on wiki
2013-10-04 09:05:43 <clandmeter> search for pinning
2013-10-04 09:07:17 <vkrishn> wiki.a.o is still slow
2013-10-04 09:07:43 <vkrishn> clandmeter: is wiki.a.o server different than a.o or git.a.0 ?
2013-10-04 09:08:27 <clandmeter> yes
2013-10-04 09:08:55 <clandmeter> wiki is pretty fast for me
2013-10-04 09:09:03 <clandmeter> but i think wiki is a diesel
2013-10-04 09:09:10 <clandmeter> takes a few clicks to load things in cache
2013-10-04 09:09:13 <ncopa> fast for me too
2013-10-04 09:09:18 <mhavela> fast here too
2013-10-04 09:09:22 <vkrishn> wiki.a.o located in US and git.a.o in Norway ?
2013-10-04 09:09:31 <ncopa> i think in US
2013-10-04 09:09:36 <ncopa> i dont think we have any infra in norway
2013-10-04 09:09:48 <clandmeter> git isnt in your location ncopa ?
2013-10-04 09:09:59 <ncopa> git.a.o is physically in nld i think
2013-10-04 09:10:06 <clandmeter> no
2013-10-04 09:10:18 <vkrishn> sorry netherlands
2013-10-04 09:10:34 <clandmeter> 91.220.88.36
2013-10-04 09:10:52 <vkrishn> I think I try the ip for wiki that clandmeter gave couple of months back
2013-10-04 09:11:04 <ncopa> address:        Netherlands
2013-10-04 09:11:31 <clandmeter> we have more infra then mine in nl?
2013-10-04 09:11:39 <ncopa> seems so yes
2013-10-04 09:12:29 <ncopa> i think wiki.a.o is either US or panama or something
2013-10-04 09:12:31 <ncopa> not sure...
2013-10-04 09:12:48 <clandmeter> ok, didnt know you moved git to somewhere else
2013-10-04 09:13:19 <ncopa> its been there for years...
2013-10-04 09:13:40 <vkrishn> clandmeter, cloudflare ip is still configured ?
2013-10-04 09:13:55 <ncopa> vkrishn: i think not, but i'm not sure...
2013-10-04 09:14:05 <ncopa> i think it was rnalrd who set that up
2013-10-04 09:14:09 <clandmeter> ncopa: could be, but then its has never been communicated or i have a bad brain.
2013-10-04 09:14:27 <ncopa> :)
2013-10-04 09:14:48 <clandmeter> communication happens more in irc i guess.
2013-10-04 09:14:59 <ncopa> yeah
2013-10-04 09:15:00 <clandmeter> doesnt really matter, aslong as it runs
2013-10-04 09:15:06 <ncopa> exactly..
2013-10-04 09:15:49 <clandmeter> ncopa: now you are here. how do you get possible opts for cmake. is there something like configure --help?
2013-10-04 09:16:00 <ncopa> dunno
2013-10-04 09:16:17 <ncopa> -DCMAKE_*
2013-10-04 09:16:17 <clandmeter> there is no general thing?
2013-10-04 09:16:19 <ncopa> normally
2013-10-04 09:16:27 <ncopa> -DCMAKE_PREFIX=/usr
2013-10-04 09:16:35 <ncopa> -DCMAKE_BUILD_TYPE=Release
2013-10-04 09:16:43 <ncopa> i think those are the common things
2013-10-04 09:16:43 <clandmeter> yes, but how to know how to enable/disable something?
2013-10-04 09:16:52 <ncopa> nor sure actually..
2013-10-04 09:16:57 <clandmeter> :)
2013-10-04 09:17:07 <ncopa> grep cmake */APKBUILD
2013-10-04 09:17:10 <ncopa> thats what i do
2013-10-04 09:17:22 <clandmeter> i just did that :)
2013-10-04 09:17:22 <ncopa> or i look in CMakeList.txt
2013-10-04 09:17:26 <ncopa> or what it was called
2013-10-04 09:17:28 <clandmeter> grep -Hir
2013-10-04 09:17:58 <clandmeter> ok thx, just want to make sure there isnt something more simple after all these years.
2013-10-04 09:18:35 <clandmeter> btw, anyone here doing home automation?
2013-10-04 09:19:47 <vkrishn> clandmeter: sorry its was rnalrd who gave the ips
2013-10-04 09:20:00 <vkrishn> in month may :)
2013-10-04 09:20:18 <clandmeter> gave what?
2013-10-04 09:20:56 <vkrishn> cloudflare ips
2013-10-04 09:21:16 <vkrishn> for wiki.a.o
2013-10-04 09:22:13 <clandmeter> oh ok
2013-10-04 09:22:21 <clandmeter> i odnt think its configured
2013-10-04 09:33:25 <vkrishn> will we be seeing acf2 in 2.7 ?
2013-10-04 09:34:54 <clandmeter> im starting to undersrtand why this piece of crap/software is nowhere to be found in any other disto...
2013-10-04 09:36:18 <kaniini> which?  redmine?  :)
2013-10-04 09:37:19 <kaniini> redmine is pretty and shiny, but a complete pain to admin
2013-10-04 09:38:55 <clandmeter> dont whine, contribute!
2013-10-04 09:38:57 <clandmeter> :p
2013-10-04 09:40:15 <vkrishn> in general, which js library would be preferred by alpine devs ?
2013-10-04 09:41:08 <clandmeter> vkrishn: check ml, there has been discussion about it regarding acf v2
2013-10-04 09:41:25 <clandmeter> but not sure there is a "conclusion"
2013-10-04 09:42:33 <vkrishn> ok
2013-10-04 10:30:27 <shafire> hi
2013-10-04 10:47:03 <shafire> how many packages are broken at moment because of musl?
2013-10-04 10:48:07 <barthalion> fabled: ↑
2013-10-04 10:48:48 <fabled> i have built 1299 of 1673 packages in edge/main on musl/x86
2013-10-04 10:52:36 <clandmeter> fabled: next week is my bday. ;-)
2013-10-04 11:15:40 <fcolista> maybe clandmeter would have all the packages built with musl as a gift :)
2013-10-04 11:16:12 <fabled> not all
2013-10-04 11:16:16 <fabled> but many, yes
2013-10-04 11:16:31 <fabled> arm builder has built 874, but is still working hard
2013-10-04 11:17:07 <fcolista> gr8 work fabled, really gr8 work
2013-10-04 11:17:18 <shafire> awesome :)
2013-10-04 11:17:29 <shafire> or lovely :)
2013-10-04 11:17:40 <ncopa> i htink fabled built more packages for musl than they have in sabotage linux
2013-10-04 11:21:36 <fcolista> http://www.etalabs.net/compare_libcs.html
2013-10-04 11:21:50 <fcolista> wow, musl rocks
2013-10-04 11:26:24 <clandmeter> can a libc rock? :)
2013-10-04 11:29:43 <fcolista> compared with uclibc i mean :)
2013-10-04 11:29:57 <shafire> whats better with musl?
2013-10-04 11:30:29 <fcolista> http://www.etalabs.net/compare_libcs.html
2013-10-04 11:31:24 <clandmeter> you should ask kaniini, he seems to have an opinion about lots of things (specially uclibc)
2013-10-04 11:31:26 <shafire> yeah, but etalabs is the autor of musl
2013-10-04 11:33:55 <barthalion> so point where the comparison is wrong
2013-10-04 11:34:19 <barthalion> it doesn't look biased to me
2013-10-04 11:36:16 <shafire> I don't make any allegations against this comparsion.
2013-10-04 11:36:56 <shafire> The license is reason enough to switch
2013-10-04 12:36:37 <ncopa> i think musl vs uclibc, the major points is: musl is maintained, uclibc is more or less not
2013-10-04 12:36:46 <ncopa> musl code is clean, uclibc code is not
2013-10-04 12:37:17 <ncopa> stable ABI
2013-10-04 12:42:19 <shafire> sounds good
2013-10-04 13:55:35 <Mp5shooter> +  * Maildir driver for Alpine 2.00
2013-10-04 13:55:35 <Mp5shooter> +  * Written by Eduardo Chappa &lt;chappa@washington.edu&gt;
2013-10-04 13:55:35 <Mp5shooter> +  *
2013-10-04 13:55:38 <Mp5shooter> yep it has it
2013-10-04 13:55:40 <Mp5shooter> woo
2013-10-05 11:58:23 <barthalion> wut
2013-10-05 11:58:29 <barthalion> algitbot: build master
2013-10-05 11:59:36 <barthalion> does anyone know wtf?
2013-10-05 13:04:18 <fcolista> barthalion, #! /bin/sh
2013-10-05 13:04:32 <fcolista> in config.sub there's a space between ! and /bin/sh
2013-10-05 13:05:00 <fcolista> ?
2013-10-05 13:18:40 <fcolista> ah, sorry. update_config_sub does not exist
2013-10-05 13:19:13 <barthalion> oh, right
2013-10-05 13:19:22 <fcolista> try to remove it
2013-10-05 13:19:33 <fcolista> s/remove/comment
2013-10-05 13:19:34 <fcolista> it works
2013-10-05 13:56:56 <shafire> hi
2013-10-05 14:40:27 <barthalion> liar
2013-10-05 14:40:38 <barthalion> I'll push it later
2013-10-07 06:30:27 <fabled> ncopa, thanks for qt5, and the language split
2013-10-07 06:38:27 <ncopa> fabled: np
2013-10-07 06:38:33 <ncopa> there are more qt5-* pakcages
2013-10-07 06:39:11 <ncopa> qt5-qtquick/qtquickcontrols/qttools/qttranslations/qtwebkit/qtxmlpatterns
2013-10-07 06:45:23 <ncopa> new kernels are out, again...
2013-10-07 07:43:19 <fabled> ncopa, re: musl builds. should we make the musl builders push to official mirror master to edge-musl?
2013-10-07 07:49:13 <ncopa> how many are not yet built?
2013-10-07 07:50:09 <ncopa> i dont think i want builders to upload anything if building fails
2013-10-07 07:50:31 <ncopa> so we'd need add options="!libc_musl" to those who does not yet build
2013-10-07 07:51:38 <kaniini> are x86/x86_64 musl builders already live?
2013-10-07 07:52:58 <fabled> x86, and armhf are live
2013-10-07 07:53:21 <fabled> 1324 of 1673 built on x86
2013-10-07 07:53:39 <fabled> 950 of 1673 built on armhf; libpng failed due to upstream bug, so it's building now many
2013-10-07 07:55:13 <fabled> i'm trying to get some base dependencies fixed now
2013-10-07 07:55:20 <fabled> so probably more will compile soon
2013-10-07 07:55:39 <fabled> but there's many packages with things that need fixing; so we can't get all built this week
2013-10-07 08:00:33 <fabled> which is a lot. seems sabotage has about 450-460 source packages
2013-10-07 08:00:47 <fabled> so we have already about 3x the amount of packages for musl
2013-10-07 08:01:26 <fabled> compiled. not nearly all are tested.
2013-10-07 08:10:15 <kaniini> i would go ahead and publish it in the archive
2013-10-07 08:11:06 <fabled> ncopa, manual upload once in a while until we get everything build?
2013-10-07 08:11:31 <ncopa> fabled: sounds good
2013-10-07 08:11:41 <kaniini> it can live in the archive without blocking 2.7 release
2013-10-07 11:06:13 <fabian_a> please git pull git://git.alpinelinux.org/fab/aports
2013-10-07 11:08:32 <fabian_a> this request contains updates and new aports (extensions for bottle, plugins for setuptools, and some python api for various well-known web services)
2013-10-07 11:09:49 <fabian_a> i renamed py-awake to awake and split out the python bindings. hope that this doesn't harm
2013-10-07 11:11:12 <fabian_a> well, is that all? no, there is also an update for redis. i switched from the bundled version of jemalloc to the upstream version of jemalloc.
2013-10-07 13:03:23 <clandmeter> ncopa: there was a way to get alpine to display utf-8 properly?
2013-10-07 13:03:43 <ncopa> in tty?
2013-10-07 13:03:54 <ncopa> it shows ok in my xorg terminals
2013-10-07 13:04:00 <clandmeter> yes in my ssh session
2013-10-07 13:04:02 <ncopa> well
2013-10-07 13:04:06 <ncopa> almost ok :)
2013-10-07 13:04:29 <ncopa> echo $CHARSET
2013-10-07 13:04:30 <ncopa> UTF-8
2013-10-07 13:04:55 <clandmeter> in my commit list, timo and bart are always displayed incorrct
2013-10-07 13:05:20 <ncopa> apk add less
2013-10-07 13:05:31 <clandmeter> oh wait, git does show it correctly
2013-10-07 13:05:36 <clandmeter> git tig doesnt
2013-10-07 13:05:43 <clandmeter> but
2013-10-07 13:05:58 <fabled> do you have bb less, or real less?
2013-10-07 13:06:08 <clandmeter> let me check
2013-10-07 13:06:24 <ncopa> bb less is a 'real' less.. i suppose you mean GNU less ;)
2013-10-07 13:06:36 <clandmeter> /usr/bin/less is owned by less-458-r0
2013-10-07 13:07:02 <ncopa> is ncurses-terminfo installed?
2013-10-07 13:07:45 <clandmeter> i didnt
2013-10-07 13:07:48 <clandmeter> but doesnt make a diff
2013-10-07 13:08:12 <clandmeter> but i think its something with ncurses
2013-10-07 13:08:17 <clandmeter> cause i have the same issues with irssi
2013-10-07 13:09:15 <clandmeter> shoudnt termifo be installed automatically?
2013-10-07 13:09:17 <clandmeter> if needed
2013-10-07 13:09:54 <ncopa> i think the most common terminfos are installed
2013-10-07 13:10:24 <clandmeter> i thought recently i saw another config was available to make utf8 better
2013-10-07 13:12:22 <clandmeter> ncopa: does tig display names correctly in your repo?
2013-10-07 13:13:32 <ncopa> i dont use tig so i dont know
2013-10-07 13:14:20 <clandmeter> ok i guess installing it isnt an option
2013-10-07 13:19:04 <tdtrask> ncopa: I'm interested in looking briefly at lua 5.2 for ACF to get an idea of the scope of work
2013-10-07 13:19:17 <tdtrask> with current edge packages, how would you recommend I begin?
2013-10-07 13:19:29 <tdtrask> apk add lua5.2
2013-10-07 13:19:33 <tdtrask> apk del lua ?
2013-10-07 13:19:41 <ncopa> yup
2013-10-07 13:19:52 <ncopa> look what deps you need in lua5.2
2013-10-07 13:19:55 <ncopa> and ping me
2013-10-07 13:20:04 <ncopa> and i'll try help you port whats missing
2013-10-07 13:20:17 <ncopa> also, i think you need rebuild haserl for lua5.2
2013-10-07 13:20:40 <ncopa> i can add testing/haserl with lua5.2 support if you want
2013-10-07 13:20:47 <tdtrask> there's no 'provides' in the packages yet?
2013-10-07 13:21:02 <tdtrask> yes, please, for the testing/haserl
2013-10-07 13:27:32 <ncopa> tdtrask: basically, all require('foo') needs to be rewritten as:
2013-10-07 13:27:40 <ncopa> foo = require('foo')
2013-10-07 13:28:05 <tdtrask> 
ACTION knows about removing 'module' function

2013-10-07 13:28:16 <ncopa> ok good
2013-10-07 13:28:24 <tdtrask> 
ACTION didn't know about require changing :(

2013-10-07 13:28:27 <ncopa> i'd suggest you do that with lua5.1 too
2013-10-07 13:28:58 <ncopa> most of the required changes will work in lua 5.1 too
2013-10-07 13:29:06 <tdtrask> right
2013-10-07 13:29:11 <ncopa> so
2013-10-07 13:29:15 <tdtrask> I already did some of the 'module' removal
2013-10-07 13:29:26 <ncopa> i think i'd read a bit on lua 5.2 porting
2013-10-07 13:29:38 <ncopa> then try make the changes without breaking anything on 5.1
2013-10-07 13:29:43 <ncopa> and then try run it on lua5.2
2013-10-07 13:29:51 <tdtrask> agreed
2013-10-07 13:29:58 <ncopa> maybe have one lxc for lua 5.1 and one for 5.2
2013-10-07 13:30:01 <ncopa> and test on both
2013-10-07 13:30:08 <ncopa> will require some work :-/
2013-10-07 13:30:09 <tdtrask> I wanted to do that for acf-core first as a test
2013-10-07 13:30:12 <ncopa> good
2013-10-07 13:30:50 <ncopa> i think most things will just work as is, due to i have built lua5.2 with compat enabled
2013-10-07 13:31:38 <ncopa> 1 nov we will do alpine-2.7
2013-10-07 13:31:48 <ncopa> i suppose you will not reach that deadline?
2013-10-07 13:31:56 <tdtrask> main question I have is about how to do the packaging
2013-10-07 13:32:05 <ncopa> hm
2013-10-07 13:32:08 <tdtrask> because 'apk del lua' won't work
2013-10-07 13:32:17 <tdtrask> due to dependencies
2013-10-07 13:32:34 <ncopa> theoretically it should work
2013-10-07 13:32:38 <ncopa> because you use haserl
2013-10-07 13:32:46 <ncopa> so you dont really need /usr/bin/lua
2013-10-07 13:32:53 <ncopa> well, for cli you do
2013-10-07 13:33:19 <tdtrask> # apk del lua
2013-10-07 13:33:19 <tdtrask> WARNING: Ignoring /media/cdrom/apks/x86/APKINDEX.tar.gz: No such file or directory
2013-10-07 13:33:19 <tdtrask>   lua: lua-bitlib lua-posix acf-alpine-baselayout acf-apk-tools acf-core acf-openssh acf-iptables acf-awall lua-md5 lua-json4 awall lua-alt-getopt lua-pc
2013-10-07 13:33:19 <tdtrask> World updated, but the following packages are not removed due to:
2013-10-07 13:33:25 <ncopa> i think you have 2 options
2013-10-07 13:33:37 <ncopa> 1) switch to 5.2 in one shot
2013-10-07 13:33:38 <tdtrask> ignore the acf packages above
2013-10-07 13:33:52 <ncopa> 2) support acf for both lua5.2 and 5.1 in parallel
2013-10-07 13:34:25 <ncopa> i you do the switch in one shot, you'll have to change the lua-* deps to lua5.2-*
2013-10-07 13:34:53 <ncopa> i.e replace lua-json4 dep with lua5.2-json4
2013-10-07 13:34:54 <ncopa> etc
2013-10-07 13:35:29 <tdtrask> ok
2013-10-07 13:35:56 <ncopa> i decided to dupsport both lua5.1 and lua5.2 in parallel, mostly due to i didnt think we could switch acf and all in one shot
2013-10-07 13:36:02 <tdtrask> right now, not too many lua5.2- packages in main
2013-10-07 13:36:07 <ncopa> correct
2013-10-07 13:36:11 <ncopa> there are some on testing
2013-10-07 13:36:27 <ncopa> if you give me a list in priority order i'll start work on them asap
2013-10-07 13:36:39 <ncopa> bah
2013-10-07 13:36:44 <ncopa> wrong commit message
2013-10-07 13:37:02 <ncopa> its testing/haserl with lua5.2 support
2013-10-07 13:38:05 <ncopa> tdtrask: also be aware that even if things works with lua5.2 in compat mode, it will definitively break in lua5.3, so its good to work on make it work on lua5.2 without compat mode
2013-10-07 13:38:34 <tdtrask> 
ACTION is working on a list of dependencies for you

2013-10-07 13:39:17 <tdtrask> 
ACTION is distracted by a library app he just downloaded

2013-10-07 13:46:34 <tdtrask> ncopa: haserl lua-posix lua-md5 lua-json4 lua-subprocess lua-sql-postgres lua-sql-sqlite3 lua-posixtz
2013-10-07 13:46:43 <tdtrask> not sure is lua-md5 is used anymore
2013-10-07 13:48:30 <ncopa> grep require.*md5
2013-10-07 13:48:42 <ncopa> i'm working on lua-md5
2013-10-07 13:49:49 <tdtrask> yeah, md5 is still require'd
2013-10-07 13:52:44 <tdtrask> ncopa: the list I gave you is priority order
2013-10-07 13:53:50 <tdtrask> hopefully I can get the ACF code 5.2-compatible by 2.7 release
2013-10-07 13:53:50 <vkrishn> ncopa, I tried edge/task , seems to work ok via ssh login
2013-10-07 13:55:03 <vkrishn> but do try on alpine terminal, command like `task color`
2013-10-07 13:55:25 <vkrishn>  rnalrd - perl-dbd-sqlite in 2.6.x has .c/.h files
2013-10-07 13:55:34 <vkrishn> v2.6/main/x86/perl-dbd-sqlite-1.37-r0.apk
2013-10-07 13:59:08 <vkrishn> will do more test on edge/task and add to bugs.a.o
2013-10-07 14:10:20 <ncopa> tdtrask: i see that you are maintainer for lua-subprocess
2013-10-07 14:10:24 <ncopa> there are 2-3 patches in there
2013-10-07 14:10:34 <ncopa> can you considered send the patches upstream?
2013-10-07 14:11:04 <ncopa> have*
2013-10-07 14:11:14 <tdtrask> I tried, but didn't get much response
2013-10-07 14:11:33 <tdtrask> one of the patches was actually suggested by the developer, but he didn't release it :)
2013-10-07 14:11:37 <tdtrask> 
ACTION will check again

2013-10-07 14:13:36 <ncopa> https://github.com/xlq/lua-subprocess/pulls
2013-10-07 14:13:39 <ncopa> no pull requests
2013-10-07 14:13:47 <ncopa> make a pull request via github
2013-10-07 14:13:56 <ncopa> then its easy for him to merge
2013-10-07 14:24:32 <ncopa> bah
2013-10-07 14:24:45 <ncopa> lua-subprocess needs some work to be ported to 5.2
2013-10-07 14:27:09 <tdtrask> 
ACTION was afraid of that

2013-10-07 14:48:09 <tdtrask> ncopa: how do I tell haserl to look for 5.2 libraries?
2013-10-07 14:48:39 <ncopa> its compiled into the binary
2013-10-07 14:48:39 <tdtrask> it's looking in /usr/share/lua/5.1/
2013-10-07 14:49:03 <tdtrask> is 5.2-compatible haserl available yet?
2013-10-07 14:54:38 <ncopa> i thought i just pushed it
2013-10-07 14:54:49 <ncopa> readelf -d /usr/bin/haserl
2013-10-07 14:54:52 <ncopa> what does it say?
2013-10-07 14:55:16 <ncopa>  0x00000001 (NEEDED)                     Shared library: [liblua-5.2.so.0]
2013-10-07 14:58:13 <ncopa> tdtrask: buildservers are still building
2013-10-07 15:05:00 <tdtrask> k
2013-10-07 17:04:00 <barthalion> clandmeter: I blame my "Å‚"
2013-10-07 17:04:18 <barthalion> clandmeter: and my parents :p
2013-10-07 18:52:49 <clandmeter> barthalion: anyway to fix it?
2013-10-07 18:53:08 <clandmeter> without changing your name
2013-10-07 18:53:59 <barthalion> No idea, I had some issues with utf-8 in irssi too
2013-10-07 18:55:10 <clandmeter> did you ever try tig?
2013-10-07 18:55:30 <barthalion> I use it quite often
2013-10-07 18:55:55 <clandmeter> same issue?
2013-10-07 18:56:54 <barthalion> not always, I don't know what causes that
2013-10-08 06:11:28 <ncopa> heh
2013-10-08 06:12:17 <ncopa> i got an email about mdocml release
2013-10-08 06:12:35 <ncopa> first time any upstream emails me without me beeing suscribed to anything
2013-10-08 06:13:28 <fabled>  Downstream
2013-10-08 06:13:29 <fabled> Several systems come bundled with mdocml utilities. If your system does not appear below, the maintainers have not contacted me and it should not be considered "official". Please contact us if you plan on maintaining a downstream version!
2013-10-08 06:13:36 <fabled> the guy wants to track downstream
2013-10-08 06:15:21 <ncopa> yup
2013-10-08 10:44:52 <clandmeter> anyone gonna put that buildbox out of his qt misery?
2013-10-08 11:04:16 <ncopa> clandmeter: i think i did
2013-10-08 11:05:35 <clandmeter> did you fix it after 11:24?
2013-10-08 11:10:26 <ncopa> >>> qt*: Running split function dev...
2013-10-08 11:10:26 <ncopa> mv: can't rename '/home/buildozer/aports/main/qt/pkg/qt/usr/lib/*.prl': No such file or directory
2013-10-08 11:10:26 <ncopa> >>> ERROR: qt-dev*: dev failed
2013-10-08 11:10:41 <ncopa> i fixed that 8.06
2013-10-08 11:10:50 <ncopa> but the build had just started
2013-10-08 11:10:59 <fabled> (2/4) Installing lua5.1-filesystem (1.6.2-r1)
2013-10-08 11:10:59 <fabled> ERROR: lua5.1-filesystem-1.6.2-r1: trying to overwrite usr/lib/lua/5.1/lfs.so owned by lua-filesystem-1.6.2-r0.
2013-10-08 11:11:00 <fabled> (3/4) Upgrading lua-filesystem (1.6.2-r0 -> 1.6.2-r1)
2013-10-08 11:11:23 <ncopa> fabled: i think you'll have to wait for -r2 if you dont want do --force
2013-10-08 11:11:27 <fabled> ok
2013-10-08 11:13:02 <ncopa> i think the zeromq update introduces a bug
2013-10-08 11:13:15 <ncopa> or the buildmsg scripts are buggy
2013-10-08 11:13:36 <ncopa> problem is that the zeromq subscriber eats cpu
2013-10-08 11:13:45 <ncopa> apparently til the build is finished
2013-10-08 11:14:01 <ncopa> i dont know what happens
2013-10-08 11:14:23 <ncopa> but it means that the already hard working build server loses 2 cores
2013-10-08 11:15:20 <ncopa> i wonder if i should just write a buildmsg server over traditional tcp rather than zeromq
2013-10-08 11:18:06 <kaniini> yes, we saw this with zeromq too, so wrote our own rpc
2013-10-08 11:26:23 <ncopa> kaniini: do you have the rpc libs under some public license?
2013-10-08 11:26:38 <ncopa> the task is very simple
2013-10-08 11:27:01 <ncopa> i think even irc does it better than the zeromq :-/
2013-10-08 11:30:54 <kaniini> http://bitbucket.org/tortoiselabs/python-ediarpc
2013-10-08 11:31:07 <kaniini> oops
2013-10-08 11:31:09 <kaniini> https://bitbucket.org/tortoiselabs/ediarpc
2013-10-08 11:39:53 <ncopa> no pubsub
2013-10-08 11:40:06 <ncopa> which i think is what we need for build servers
2013-10-08 11:40:18 <ncopa> maybr should use prosody
2013-10-08 11:40:20 <ncopa> its lua
2013-10-08 11:40:36 <ncopa> its an established protocol (xmpp)
2013-10-08 11:40:59 <ncopa> it looks like it has support for publish-subscribe
2013-10-08 11:41:40 <ncopa> i suppose it would allow us to send and receive build messages via pidgin
2013-10-08 12:01:03 <clandmeter> ncopa: are there no available solutions that match your requests?
2013-10-08 12:01:53 <ncopa> clandmeter: prosody has support for pubsub
2013-10-08 12:03:29 <clandmeter> ncopa: something like buildbot is not an option?
2013-10-08 12:03:47 <ncopa> where can i find buildbot?
2013-10-08 12:04:02 <ncopa> http://code.matthewwild.co.uk/clix/file/956242b7d5df/README
2013-10-08 12:04:05 <clandmeter> http://buildbot.net/#/basics
2013-10-08 12:04:17 <ncopa> i think it is possible to set up prosody server
2013-10-08 12:04:24 <ncopa> and use clix to send messages
2013-10-08 12:25:03 <ncopa> buildbot looks kinda nice
2013-10-08 12:25:08 <ncopa> except that it is python
2013-10-08 12:27:36 <clandmeter> seems pretty active project
2013-10-08 12:40:31 <ncopa> fabled: i'm looking at buildbot, it looks like you have a "try" fetaure
2013-10-08 12:40:47 <ncopa> "Buildbot includes a way for developers to submit patches for testing without committing them to the source code control system. (This is really handy for projects that support several operating systems or architectures.)"
2013-10-08 12:41:46 <ncopa> http://docs.buildbot.net/current/tutorial/tour.html#quick-tour-label at the end
2013-10-08 12:43:58 <StarWarsFan> How about Jenkins?
2013-10-08 12:46:29 <ncopa> java :-(
2013-10-08 12:46:48 <StarWarsFan> and?
2013-10-08 12:47:01 <ncopa> it means that to bootstrap a new arch
2013-10-08 12:47:14 <ncopa> eg musl x86 or arm or whatever
2013-10-08 12:47:27 <ncopa> the first package you will need to make things autobuild is java
2013-10-08 12:48:13 <ncopa> i would prefer to be able to start with something simpler
2013-10-08 12:48:30 <ncopa> i think even python is too big for it
2013-10-08 12:48:49 <ncopa> lua is smallish and simple
2013-10-08 12:49:05 <ncopa> so prosody and xmpp is tempting
2013-10-08 12:49:19 <ncopa> i mean
2013-10-08 12:49:29 <ncopa> we have issues keeping openjdk working...
2013-10-08 12:49:34 <ncopa> its currently broke on x86_64
2013-10-08 12:49:50 <ncopa> its not easy to get running on the easiset platform
2013-10-08 12:49:57 <clandmeter> i specially didnt mention jenkers because of java...
2013-10-08 12:50:02 <ncopa> what about musl and arm and other odd things
2013-10-08 12:51:47 <ncopa> StarWarsFan: if you fix openjdk6 for uclibc x86_64 and musl, and is willing to maintain it and i can trust that you make sure it is always running... then i can consider using java for somthing as critical as building infra ;)
2013-10-08 12:52:04 <StarWarsFan> :-)
2013-10-08 12:52:32 <ncopa> buildbot looks like it does what we want
2013-10-08 12:52:42 <StarWarsFan> 2nd part wouldn't be a problem, daily business ;-)
2013-10-08 12:52:58 <StarWarsFan> fix openjdk-build is quite harder for me...
2013-10-08 12:53:05 <ncopa> its hard for me too...
2013-10-08 12:53:09 <ncopa> the thing is
2013-10-08 12:53:11 <ncopa> it just broke
2013-10-08 12:53:19 <ncopa> i dont know why or how
2013-10-08 12:53:27 <ncopa> suddely it just no longer builds
2013-10-08 12:53:43 <StarWarsFan> ok, i will investigate more on this
2013-10-08 12:54:00 <ncopa> it took a rocketsience (fabled) to make it build in first place
2013-10-08 12:54:09 <ncopa> rocket scientist*
2013-10-08 12:54:10 <StarWarsFan> because this is a blocker for our builds at the moment...
2013-10-08 12:54:29 <ncopa> yeah we need to fix it asap
2013-10-08 12:55:03 <ncopa> it appears to be gcc-java internal compiler error
2013-10-08 12:55:11 <ncopa> so it might be a gcc problem
2013-10-08 12:55:48 <StarWarsFan> i c
2013-10-08 13:04:14 <ncopa> build servers still building...
2013-10-08 13:04:31 <ncopa> re build infra
2013-10-08 13:04:41 <ncopa> i think we have atleast the following options:
2013-10-08 13:04:56 <ncopa> 1) fix our current zeromq scripts
2013-10-08 13:05:16 <ncopa> 2) build something new based on xmpp (prosody, verse and clix)
2013-10-08 13:05:27 <ncopa> 3) build something with buildbot and python
2013-10-08 13:05:53 <ncopa> 4) build a minimalistic pubsub server on raw tcp (using C/lua)
2013-10-08 13:06:18 <ncopa> 5) switch back to IRC
2013-10-08 13:06:22 <ncopa> 6) something else?
2013-10-08 13:08:53 <clandmeter> imho putting basics in the hands of 3rd party should give you some extra time for other things. plus it brings other features you currently dont have time for adding to current env.
2013-10-08 13:10:48 <ncopa> right
2013-10-08 13:10:55 <clandmeter> and our python support seems okish. kaniini uses it so we can let him maintain it ;-)
2013-10-08 13:11:57 <ncopa> fabled: what do you think about adding python to the list of our bootstrap packages?
2013-10-08 13:12:08 <ncopa> well, together with git
2013-10-08 13:12:43 <ncopa> looks like we already need python to build git
2013-10-08 13:13:22 <fabled> ncopa, i don't want  to have it in cross-build bootstrap
2013-10-08 13:13:37 <ncopa> i dont think we need it there eitehr
2013-10-08 13:13:37 <fabled> but i think we can just native build it by hand, after initial cross-build for the builders
2013-10-08 13:13:44 <ncopa> right
2013-10-08 13:13:51 <ncopa> thats the list i'm talking about
2013-10-08 13:13:57 <ncopa> toghether with git
2013-10-08 13:14:01 <fabled> there's no such list
2013-10-08 13:14:14 <fabled> i just run for pkg in *; cd $pkg; abuild -R; done
2013-10-08 13:14:18 <fabled> and initial build is done
2013-10-08 13:14:42 <ncopa> ok
2013-10-08 13:16:33 <ncopa> hm
2013-10-08 13:16:35 <fabled> ncopa, you should really package the autobuild
2013-10-08 13:16:44 <fabled> so we do cd autobuild ; abuild -R
2013-10-08 13:16:46 <fabled> to bootstrap
2013-10-08 13:16:47 <ncopa> buildbot needs all the twisted stuff
2013-10-08 13:16:57 <ncopa> yeah
2013-10-08 13:17:13 <ncopa> but i want fix it first
2013-10-08 13:17:21 <ncopa> its kinda broke
2013-10-08 13:17:26 <ncopa> eats too much cpu
2013-10-08 13:17:50 <ncopa> either fix it or switch to soemthing else
2013-10-08 13:20:59 <clandmeter> ncopa: you want me to try to package buildbot?
2013-10-08 13:28:36 <ncopa> if you have time
2013-10-08 13:28:44 <ncopa> i cannot guarantee i will use it...
2013-10-08 13:29:47 <ncopa> i think it is a good thing if as few packages as possible are installed on the build servers
2013-10-08 13:50:36 <clandmeter> buildservers are the slaves, not sure how much they need.
2013-10-08 13:52:53 <ncopa> poor build servers...
2013-10-08 14:54:17 <ncopa> algitbot: build 2.6-stable
2013-10-08 14:56:38 <ncopa> algitbot: build 2.6-stable
2013-10-08 14:58:50 <ncopa> ok, i know how to reproduce the buildmsg-subscribe bug
2013-10-08 15:33:34 <clandmeter> \o/
2013-10-08 16:47:33 <kaniini> clandmeter: if i maintain it, python 2.7 will stay until i am personally ready to go to python 3
2013-10-08 16:47:35 <kaniini> ;)
2013-10-08 20:53:48 <clandmeter> kaniini: can you fix fish?
2013-10-08 20:54:33 <clandmeter> its missing apropos which is part of man afaik
2013-10-08 21:56:43 <kaniini> what do you mean
2013-10-08 22:04:06 <clandmeter> kaniini: when you enter a command and want to autocomplete it with tab i get an error
2013-10-09 04:31:14 <StarWarsFan> Moinmoin
2013-10-09 04:39:38 <StarWarsFan> ncopa: did some experiments with openjdk-build yesterday
2013-10-09 04:40:09 <StarWarsFan> you said, that the build on x86_64 just stops
2013-10-09 04:40:42 <StarWarsFan> can i see the build output somewhere?
2013-10-09 04:41:21 <StarWarsFan> because my build breaks with some kind of OOM
2013-10-09 04:41:41 <StarWarsFan> at first I've updated the used components to the latest versions
2013-10-09 04:42:10 <StarWarsFan> build fails with an compilation error regarding 1.5 vs. 1.6 jvm
2013-10-09 04:42:52 <StarWarsFan> then i've deactivated a sed statement, which modifies the makefile regading jce compiler parameters
2013-10-09 04:43:22 <StarWarsFan> and now the build runs through the problem from before
2013-10-09 04:43:37 <StarWarsFan> but fails during linking step
2013-10-09 04:44:48 <StarWarsFan> see http://pastebin.com/9Tx7hEnX
2013-10-09 04:44:59 <StarWarsFan> Error occurred during initialization of VM
2013-10-09 04:44:59 <StarWarsFan> Could not reserve enough space for code cache
2013-10-09 04:44:59 <StarWarsFan> make[5]: *** [product] Error 1
2013-10-09 04:45:42 <StarWarsFan> this was with 4G memory on the machine
2013-10-09 04:45:58 <StarWarsFan> did the same with 6G but got the same result :-(
2013-10-09 04:52:40 <fabled> StarWarsFan, i believe it was internal compiler error what we got
2013-10-09 04:53:14 <fabled> StarWarsFan, which kernel you are running on the build box?
2013-10-09 04:54:54 <StarWarsFan> Linux alpine-satellite 3.10.14-1-grsec #2-Alpine SMP Thu Oct 3 12:13:48 UTC 2013 x86_64 Linux
2013-10-09 04:56:12 <StarWarsFan> and on the other machine
2013-10-09 04:56:14 <StarWarsFan> Linux alpine-edge-x86-64 3.10.15-0-grsec #1-Alpine SMP Tue Oct 8 15:51:00 UTC 2013 x86_64 Linux
2013-10-09 04:58:34 <fabled> hmm
2013-10-09 04:58:44 <fabled> i wonder if my pie patch broke something
2013-10-09 04:59:18 <fabled> there is nasty bug in kernel that limits PIE application's memory usage to very low amounts in 32-bit platforms
2013-10-09 04:59:28 <StarWarsFan> Can I see the build output somewhere?
2013-10-09 05:00:17 <fabled> https://dpaste.de/AUcn
2013-10-09 05:02:18 <fabled> it's null pointer exception, so could be memory allocation error also
2013-10-09 05:03:03 <StarWarsFan> right, this is the first error i got
2013-10-09 05:03:06 <StarWarsFan> but fixed
2013-10-09 05:03:22 <StarWarsFan> the problem is line 481
2013-10-09 05:03:31 <StarWarsFan> regarding jce
2013-10-09 05:03:43 <StarWarsFan> same as line 500
2013-10-09 05:06:00 <StarWarsFan> but i'm not sure, if this leads into the OOM
2013-10-09 05:06:18 <StarWarsFan> if you want to test, deactivate line 82 here: http://git.alpinelinux.org/cgit/aports/tree/main/openjdk6/APKBUILD
2013-10-09 05:13:02 <StarWarsFan> oh, i wrote "jce". sry, whould be "gcj"
2013-10-09 05:13:43 <StarWarsFan> -whould +should
2013-10-09 05:14:09 <StarWarsFan> 
ACTION <-- seems to have had to less sleep...

2013-10-09 05:23:34 <StarWarsFan> fabled: maybe first answer here is worth a try:
2013-10-09 05:23:37 <StarWarsFan> http://stackoverflow.com/questions/15812939/java-fails-with-could-not-reserve-enough-space-for-code-cache
2013-10-09 05:25:12 <fabled> i think we do that
2013-10-09 05:25:24 <fabled> then a gain, it might not be done for an intermediate compiler
2013-10-09 05:26:47 <StarWarsFan> hm, i c
2013-10-09 05:27:11 <fabled> oh, it's ecj failing
2013-10-09 05:27:32 <fabled> i wonder if gcc upgrade / packaging changes miss that paxctl now
2013-10-09 05:37:40 <StarWarsFan> how to check this?
2013-10-09 05:59:41 <fabled> paxctl -v /path/to/binary
2013-10-09 06:02:20 <StarWarsFan> http://sprunge.us/fONY
2013-10-09 06:03:34 <StarWarsFan> aber gcj:
2013-10-09 06:03:36 <StarWarsFan> file /usr/bin/gcj does not have a PT_PAX_FLAGS program header, try conversion
2013-10-09 06:08:25 <StarWarsFan> so i've tried paxctl -C with gcj and set properties afterwards like on /usr/bin/java
2013-10-09 06:08:36 <StarWarsFan> lets see what happens...
2013-10-09 06:08:50 <fabled> StarWarsFan, you'd need to do "paxctl -c -m"
2013-10-09 06:09:02 <fabled> oh
2013-10-09 06:09:03 <fabled> right
2013-10-09 06:09:48 <StarWarsFan> ok, started the build
2013-10-09 06:09:56 <StarWarsFan> will take a long time... :-/
2013-10-09 06:11:32 <fabled> StarWarsFan, is javac also modified?
2013-10-09 06:11:36 <fabled> that is /usr/lib/jvm/java-1.5-gcj/bin/javac
2013-10-09 06:11:49 <StarWarsFan> checking...
2013-10-09 06:12:03 <fabled> then again, that's gcj compiled ecj. and might or might not need it
2013-10-09 06:13:38 <StarWarsFan> hm:
2013-10-09 06:13:40 <StarWarsFan> file /usr/lib/jvm/java-1.5-gcj/bin/javac does not have a PT_PAX_FLAGS program header, try conversion
2013-10-09 06:13:40 <StarWarsFan> root@alpine-edge-x86-64:~ paxctl -C /usr/lib/jvm/java-1.5-gcj/bin/javac
2013-10-09 06:13:40 <StarWarsFan> root@alpine-edge-x86-64:~
2013-10-09 06:13:40 <StarWarsFan> file /usr/lib/jvm/java-1.5-gcj/bin/javac cannot have a PT_PAX_FLAGS program header, creation failed
2013-10-09 06:14:34 <StarWarsFan> ah, might be because of the running build...
2013-10-09 06:21:56 <fabled> ok. i'll try building it here as well
2013-10-09 06:23:09 <build64-2-4> files from v2.4.11-103-g6a77544 uploaded
2013-10-09 06:25:00 <build64-2-5> files from v2.5.4-189-g4cd5ff4 uploaded
2013-10-09 06:25:40 <build-2-5> files from v2.5.4-189-g4cd5ff4 uploaded
2013-10-09 06:31:17 <shafire> hi
2013-10-09 06:37:11 <build64-2-3> files from v2.3.6-266-g8ecfcb4 uploaded
2013-10-09 07:02:58 <fabled> StarWarsFan, so what did you do to fix the first error?
2013-10-09 07:05:14 <StarWarsFan> deactivate line 82 on APKBUILD
2013-10-09 07:05:28 <StarWarsFan> but i've updated all component version to latest releases too
2013-10-09 07:05:44 <StarWarsFan> wanna have a diff?
2013-10-09 07:07:38 <fabled> which line is 82?
2013-10-09 07:07:46 <fabled> the native-ecj sed ?
2013-10-09 07:07:51 <StarWarsFan> right
2013-10-09 07:08:54 <StarWarsFan> here my changes in short:
2013-10-09 07:08:55 <StarWarsFan> icedteaver=1.12.6
2013-10-09 07:08:55 <StarWarsFan> arch="all"
2013-10-09 07:08:55 <StarWarsFan> OPENJDK_VERSION=b27
2013-10-09 07:08:55 <StarWarsFan> OPENJDK_DATE=26_oct_2012
2013-10-09 07:08:56 <StarWarsFan> ANT_VER=1.9.2
2013-10-09 07:08:58 <StarWarsFan> #        sed -e "s/-o native-ecj/-o native-ecj -lgcj/g" -i Makefile.am
2013-10-09 07:09:20 <StarWarsFan> build still running...
2013-10-09 07:09:36 <vkrishn> ncopa: re: zeromq issue XPUB/XSUB can also be used
2013-10-09 07:09:38 <fabled> ok
2013-10-09 07:10:07 <vkrishn> there is new version v.4.0.1 :)
2013-10-09 07:10:24 <fabled> i'm trying icedteaver=1.11.12
2013-10-09 07:10:45 <vkrishn> XPUB/XSUB would move filtering from server to client
2013-10-09 07:11:26 <vkrishn> with newer version might solve some issues
2013-10-09 07:20:03 <ncopa> vkrishn: i have found out the high cpu usage problem
2013-10-09 07:20:06 <fabled> StarWarsFan, removing the sed, will make compiling native-ecj fail
2013-10-09 07:20:21 <fabled> and openjdk will revery to using "java ecj.jar"
2013-10-09 07:20:40 <ncopa> vkrishn: but there are other issues, which might be solved with xsub/xpub. i will check that later
2013-10-09 07:20:40 <fabled> so indicates that gcj is has an issue
2013-10-09 07:20:52 <vkrishn> ncopa: there is also nanomsg.org
2013-10-09 07:21:07 <vkrishn> but still in alpha state
2013-10-09 07:21:16 <vkrishn> but seems nice
2013-10-09 07:21:21 <ncopa> "nano..." definitively draws my attention :)
2013-10-09 07:21:29 <vkrishn> :)
2013-10-09 07:21:42 <ncopa> PUBSUB - distributes messages to large sets of interested subscribers
2013-10-09 07:21:48 <ncopa> its C (not c++)
2013-10-09 07:22:08 <vkrishn> yes another + from al point
2013-10-09 07:22:27 <ncopa> yes
2013-10-09 07:22:32 <ncopa> looks nice
2013-10-09 07:25:35 <StarWarsFan> fabled: did you try 1.12-line of icedtea?
2013-10-09 07:25:50 <fabled> not yet
2013-10-09 07:26:13 <fabled> i'd like 1.11 to work -> backport, and then upgrade to 1.12
2013-10-09 07:26:47 <clandmeter> ncopa: i have buildbot in my tree if you like to try.
2013-10-09 07:26:51 <clandmeter> will push in a minute
2013-10-09 07:28:38 <StarWarsFan> hm, build failed with the same error
2013-10-09 07:29:49 <StarWarsFan> even with PaX flags: -p-s-m-x-e-- [/usr/bin/gcj]
2013-10-09 07:32:15 <fabled> either something broke in gcc/gcj, or the kernel pie-aslr patch i added is confusing it on 64-bits
2013-10-09 07:32:34 <fabled> hmm
2013-10-09 07:32:39 <StarWarsFan> hmmm
2013-10-09 07:34:16 <vkrishn> about perl-dbd-sqlite issue do I put in bugs.a.o (not seeing rnarld around for last few days) ?
2013-10-09 07:35:41 <fabled> vkrishn, i think so, rnarld is traveling/on vacation for few weeks IIRC
2013-10-09 07:35:56 <vkrishn> ok
2013-10-09 07:40:32 <fabled> ncopa, did we ever manage to build openjdk with gcc 4.8.x on x86_64?
2013-10-09 07:44:15 <ncopa> fabled: i think so yes
2013-10-09 07:44:30 <ncopa> i will fix that samba thingy...
2013-10-09 07:44:35 <clandmeter> ncopa: samba is missing gettext
2013-10-09 07:44:38 <clandmeter> or something
2013-10-09 07:47:13 <ncopa> yeah, but it didnt need gettext with 4.0.9
2013-10-09 07:58:44 <fabled> StarWarsFan, ok, pie-aslr patch cannot be it, because the same happens with -fno-PIE added to native-ecj
2013-10-09 07:59:04 <StarWarsFan> so another "hmmm" :-|
2013-10-09 08:07:36 <fabled> StarWarsFan, i suspect gcc is affected somehow; i'll rebuild it now, and retry openjdk. we have since updated isl (fixing ast generation regression), and few other things. those might very well be related
2013-10-09 08:07:57 <StarWarsFan> i c
2013-10-09 08:09:56 <StarWarsFan> let me know if i should test something...
2013-10-09 08:23:52 <ncopa> algitbot: build master
2013-10-09 08:24:18 <ncopa> algitbot: build master
2013-10-09 08:24:49 <ncopa> algitbot: build master
2013-10-09 08:40:48 <ncopa> nanomsg looks really nice
2013-10-09 08:42:38 <vkrishn> I could do some testing if put to edge/testing
2013-10-09 08:50:10 <vkrishn> author puts some nice points here: http://nanomsg.org/documentation-zeromq.html
2013-10-09 10:01:40 <ncopa> gnu make 4.0 is out
2013-10-09 10:53:59 <fabled> StarWarsFan, after rebuilding gcc, rebuild openjdk failed the same. so it's weird.
2013-10-09 10:54:19 <StarWarsFan> fcuk :-(
2013-10-09 10:54:29 <StarWarsFan> that's odd
2013-10-09 10:58:08 <fabled> indeed
2013-10-09 10:59:25 <fabled> 4.8.2 is supposed to be out soon too
2013-10-09 10:59:37 <fabled> dunno if it fixes it or not. i know there's lots of stuff in that branch already
2013-10-09 11:00:38 <fabled> i wonder if the unwinding code is somehow broke
2013-10-09 11:01:02 <fabled> thenagain, the backtrace looks sane
2013-10-09 11:04:29 <ncopa> i wonder if i dare upgrade to gnu make 4 ...
2013-10-09 11:05:06 <fabled> wow
2013-10-09 11:05:13 <fabled> 3 years
2013-10-09 11:05:22 <fabled> is there changelog
2013-10-09 11:10:06 <ncopa> well the official page is kinda not informative
2013-10-09 11:10:28 <ncopa> and the release anouncement on savannah says, read NEWS in tarball
2013-10-09 11:10:51 <ncopa> so the gnu ppl thinks download untar and open NEWS is the way read news
2013-10-09 11:10:56 <ncopa> http://git.savannah.gnu.org/cgit/make.git/tree/NEWS
2013-10-09 11:11:39 <ncopa> bah guile support
2013-10-09 11:11:48 <ncopa> let me guess
2013-10-09 11:12:10 <ncopa> in 1 year you will need gnu make with guile to build some packages
2013-10-09 11:12:19 <ncopa> and to build guile you will need gnu make
2013-10-09 11:12:26 <ncopa> so hello bootstrap
2013-10-09 11:12:45 <ncopa> hehe
2013-10-09 11:13:05 <ncopa> the NEWS has onlyl 1 "WARNING Backward-incompatibility!"
2013-10-09 11:13:30 <ncopa> while 3.8.2 had 7
2013-10-09 11:30:53 <fabled> StarWarsFan, building native-ecj with "-g -O0" made it work. not sure if it fails later on during the build.
2013-10-09 11:33:18 <StarWarsFan> interesting
2013-10-09 12:05:18 <fabled> StarWarsFan, actually, your later failure is likely caused by java bootstrap, and we not fixing pactl for the intermediate java compiler
2013-10-09 12:05:37 <fabled> we currently have vserver kernel on builder, so that error would not likely happen on that
2013-10-09 12:05:47 <fabled> but you running grsec would bump into it first
2013-10-09 12:05:54 <StarWarsFan> ok
2013-10-09 12:06:28 <fabled> so yes, java/javac needs paxctl. we do it for the final java binaries produced
2013-10-09 12:06:33 <fabled> but since  java build goes:
2013-10-09 12:06:53 <fabled>  1. build native-ecj if possible (it was broke due to gcj; workaround, disable native-ecj, or add -O0 -g)
2013-10-09 12:07:09 <fabled>  2. it builds bootstrap jdk java javac with native-ecj
2013-10-09 12:07:19 <fabled>  3. it builds final jdk with java/javac built in step 2
2013-10-09 12:07:27 <fabled> we paxctl stuff from step 3
2013-10-09 12:07:32 <fabled> but not stuff from 2
2013-10-09 12:07:40 <fabled> that's why your step 3 failed. no paxctl done
2013-10-09 12:07:52 <fabled> i think
2013-10-09 12:08:10 <StarWarsFan> ah ok, i think i understand (a little bit...;-)
2013-10-09 12:09:57 <fabled> hum, build-paxctl.patch should actually do that
2013-10-09 12:10:08 <fabled> perhaps it tries to use java / javac already within stage 2
2013-10-09 12:11:07 <ncopa> can you do: make stage1-java && paxctl $stage1-objdir/bin/java && make
2013-10-09 12:54:21 <ncopa> hmm buildservers are really busy...
2013-10-09 12:56:05 <shafire> fabled: have you tested the rng on arm?
2013-10-09 12:56:26 <ncopa> i think he is busy making it build
2013-10-09 12:57:04 <shafire> ok
2013-10-09 12:57:25 <fabled> shafire, what rng?
2013-10-09 12:57:37 <shafire> random number generator
2013-10-09 12:57:41 <fabled> which one?
2013-10-09 12:57:47 <shafire> the one on the chip
2013-10-09 12:58:31 <fabled> no
2013-10-09 13:33:32 <fc1979> make sense that busybox wake-on-lan won't work a bonded interface (with ucarp) ?
2013-10-09 13:35:17 <fc1979> ops, wrong channel
2013-10-09 13:35:19 <fc1979> sorry
2013-10-09 13:44:41 <fabled> clandmeter, hi
2013-10-09 13:44:48 <clandmeter> hi
2013-10-09 13:44:56 <fabled> care to create 'edge-musl' in nl.alpinelinux.org with buildozer as owner
2013-10-09 13:45:02 <fabled> so i can start pushing stuff
2013-10-09 13:45:40 <clandmeter> i need to create that on nl.a.o?
2013-10-09 13:45:42 <fabled> yes
2013-10-09 13:45:47 <fabled> buildozer user can't create it
2013-10-09 13:45:48 <clandmeter> isnt that done by rsync?
2013-10-09 13:45:54 <fabled> no perms to write top dir
2013-10-09 13:45:58 <clandmeter> ok
2013-10-09 13:46:31 <fabled> i'm gonna be uploading about 2GB of x86-musl stuff and 1.5GB or so of armhf-musl stuff
2013-10-09 13:46:49 <fabled> they are incomplete still, but quite large set of packages is built already
2013-10-09 13:47:05 <shafire>  into /alpine/edge/main/arm?
2013-10-09 13:47:28 <fabled> http://nl.alpinelinux.org/alpine/ should get edge-musl
2013-10-09 13:47:44 <shafire> ah musl
2013-10-09 13:47:45 <fabled> and the arm tag is armhf
2013-10-09 13:48:19 <shafire> does rsync delete old directories too?
2013-10-09 13:48:31 <ncopa> yup
2013-10-09 13:48:41 <shafire> what should be in http://nl.alpinelinux.org/alpine/v2.6/x86/main/?
2013-10-09 13:48:41 <clandmeter> done
2013-10-09 13:48:44 <shafire> why is this empty?
2013-10-09 13:49:07 <fabled> that's old cruft i think
2013-10-09 13:49:10 <barthalion> yup
2013-10-09 13:49:12 <ncopa> shafire: i probaly created it by mistake whil setting up v2.6
2013-10-09 13:49:13 <fabled> we added x86 there at some point
2013-10-09 13:49:28 <ncopa> 	2013-Apr-19 16:50:41
2013-10-09 13:49:34 <barthalion> x86 used to be a symlink
2013-10-09 13:49:38 <barthalion> no idea why it is not now
2013-10-09 13:49:39 <clandmeter> i think there are still some weird links/dirs on our mirror
2013-10-09 13:49:42 <fabled> yes
2013-10-09 13:49:45 <clandmeter> but i didnt touch it.
2013-10-09 13:50:06 <clandmeter> I let ncopa handle that before i delete something back-compat or whatever.
2013-10-09 13:50:16 <shafire> then you could delete x86 and x86_64 :)
2013-10-09 13:50:37 <shafire> in /alpine/v2.6/
2013-10-09 13:50:55 <ncopa> done
2013-10-09 13:50:56 <ncopa> thanks
2013-10-09 13:55:01 <shafire> what's about http://nl.alpinelinux.org/alpine/edge/testing/testing.md5 ?
2013-10-09 13:55:56 <clandmeter> sounds like it can be deleted
2013-10-09 13:56:15 <fabled> http://nl.alpinelinux.org/alpine/edge-musl/main/x86/ is populated now
2013-10-09 13:56:22 <clandmeter> yes i see it
2013-10-09 13:56:40 <clandmeter> i guess arm also mostly complete
2013-10-09 13:56:42 <fabled> arm stuff is coming to. at 'm' now
2013-10-09 13:57:24 <fabled> done
2013-10-09 13:57:26 <clandmeter> i guess you are not on adsl
2013-10-09 13:57:28 <clandmeter> :)
2013-10-09 13:57:35 <shafire> http://nl.alpinelinux.org/alpine/edge/main/x86_64/x86_64/x86_64/x86_64/x86_64/ <-- and there is a symlink?
2013-10-09 13:57:52 <fabled> yes, x86_64 -> .
2013-10-09 13:57:54 <fabled> or similar
2013-10-09 13:57:55 <fabled> it's broke
2013-10-09 13:58:13 <fabled> ok
2013-10-09 13:58:29 <fabled> should probably create some boot images so the package repos come actually useful
2013-10-09 13:59:02 <shafire> so, I should go and buy an arm board
2013-10-09 13:59:21 <fabled> *highly expiremental* build still
2013-10-09 13:59:33 <fabled> but yes, testers are welcome :)
2013-10-09 13:59:50 <fabled> i need to dig up my raspberry pi kernel image, and create a minimal boot image
2013-10-09 14:00:30 <clandmeter> can we have them before 17:00?
2013-10-09 14:00:36 <shafire> ncopa: may you delete testing.md5 in /alpine/edge/testing/testing.md5 and x86_64 symlink in /alpine/edge/main/x86_64 ?
2013-10-09 14:01:07 <shafire> clandmeter: why so fast? :)
2013-10-09 14:01:16 <clandmeter> fast?
2013-10-09 14:01:39 <clandmeter> ive been crying for it for months ;-)
2013-10-09 14:01:49 <fabled> :D
2013-10-09 14:01:55 <fabled> it's 17:00 here now
2013-10-09 14:02:05 <fabled> i think i'll do them tomorrow
2013-10-09 14:02:09 <fabled> since i need to rebuild kernel
2013-10-09 14:02:26 <shafire> fabled: where you from? I forgot
2013-10-09 14:02:30 <fabled> oh, and musl needs patching
2013-10-09 14:02:36 <fabled> shafire, Finland
2013-10-09 14:02:55 <shafire> are you the swedish guy in finland?
2013-10-09 14:03:12 <fabled> no, i'm the finnish guy in finland
2013-10-09 14:03:19 <fabled> stats are:
2013-10-09 14:03:20 <fabled> musl-x86: 1423 of 1673
2013-10-09 14:03:21 <fabled> musl-arm: 1222 of 1673
2013-10-09 14:03:27 <fabled> source packages built vs. total
2013-10-09 14:03:40 <fabled> musl-arm failed on xorg-server
2013-10-09 14:03:48 <fabled> so probably lot of X stuff missing
2013-10-09 14:04:20 <fabled> but at least console should be usable
2013-10-09 14:04:29 <fabled> and RPI will need additional packages to make the gpu usable anyway
2013-10-09 14:04:32 <clandmeter> you dont really need X right?
2013-10-09 14:04:38 <fabled> no
2013-10-09 14:04:40 <fabled> not atm
2013-10-09 14:04:53 <ncopa> we need it if we will ditch uclibc completely
2013-10-09 14:05:26 <fabled> X built for x86
2013-10-09 14:05:33 <fabled> so there's something musl/arm specific
2013-10-09 14:05:41 <ncopa> ah ok
2013-10-09 14:07:25 <StarWarsFan> if alpine finally works on raspi, that will be way cool!
2013-10-09 14:07:26 <StarWarsFan> :-)
2013-10-09 14:09:05 <shafire> from tmpfs :P
2013-10-09 14:10:29 <clandmeter> tmpfs is the only way to run anything on an rpi
2013-10-09 14:10:49 <clandmeter> and seems we are pretty good at it :)
2013-10-09 14:17:16 <shafire> I am very anxious
2013-10-09 14:18:15 <StarWarsFan> fabled: i'm not sure what the current state regarding openjdk is
2013-10-09 14:18:39 <StarWarsFan> did you managed to fix it or was this just an experimental solution?
2013-10-09 14:19:23 <StarWarsFan> especially regarding ncopa's hint "[14:11:07] <ncopa> can you do: make stage1-java && paxctl $stage1-objdir/bin/java && make"
2013-10-09 14:21:28 <ncopa> i think we do soemthing similar with qt(4)
2013-10-09 14:21:43 <ncopa> we build the executable that needs paxctl first
2013-10-09 14:21:55 <ncopa> adjust the paxctl
2013-10-09 14:21:58 <ncopa> and then rerun make
2013-10-09 14:26:14 <fabled> i think there's something else
2013-10-09 14:26:26 <fabled> i got random seg.faults on vserver kernel too
2013-10-09 14:52:40 <shafire> the buildsystem uses zsh?
2013-10-09 16:20:10 <build-2-4> files from v2.4.11-107-g5354cef uploaded
2013-10-09 16:26:37 <build64-2-4> files from v2.4.11-107-g5354cef uploaded
2013-10-09 17:58:59 <mhavela> fabled: Just curious... is there some image that can be used for running AL on a RPI - or is that still yet to come?
2013-10-09 17:59:22 <fabled> not yet done. planning to build it tomorrow. but let's see how it goes.
2013-10-09 17:59:45 <mhavela> Nice. Keep up the good work! And thank you for your hard work on it
2013-10-09 18:32:50 <fabled> ncopa, lua-posix is broke?
2013-10-09 19:05:52 <ncopa_> fabled: do you have the apk key for edge-musl-x86 some place?
2013-10-09 19:22:14 <shafire> ncopa_ is chatting after 18:00? :)
2013-10-09 19:22:32 <shafire> [16:00] <shafire> ncopa: may you delete testing.md5 in /alpine/edge/testing/testing.md5 and x86_64 symlink in /alpine/edge/main/x86_64 ?
2013-10-09 19:29:12 <ncopa_> shafire: done, thanks
2013-10-09 20:27:11 <clandmeter> are they breaking stuff again?
2013-10-09 20:47:07 <clandmeter> fabled: still here?
2013-10-10 00:59:59 <Niichan> kaniini: Will Alpine for arm support raspi? It'd be nice to get Arch off my pi
2013-10-10 01:07:37 <jbilyk> Niichan: fabled was running into some issues with arm build getting arm 6 support working - seems gcc has some issues iirc
2013-10-10 04:56:15 <fabled> Niichan, yes, the packages are RPi compatible
2013-10-10 04:56:36 <Niichan> fabled: Is there a working build out?
2013-10-10 04:56:45 <fabled> packages are out, but no boot image yet
2013-10-10 04:57:12 <Niichan> awww
2013-10-10 05:00:39 <StarWarsFan> Moin
2013-10-10 05:15:25 <fabled> planning to do a boot image today, but i need to do new kernel, so let's see how it goes
2013-10-10 05:22:29 <StarWarsFan> :-) sounds good
2013-10-10 05:47:41 <Niichan> Id I'm looking to set up a stable Alpine server, what verison should I use?
2013-10-10 05:47:45 <Niichan> If*
2013-10-10 05:52:23 <ncopa> morning
2013-10-10 05:52:33 <ncopa> Niichan: v2.6
2013-10-10 05:56:49 <Niichan> ncopa: the 2.6.5 iso is not booting in virtualbox for me
2013-10-10 05:57:07 <ncopa> what error do you get?
2013-10-10 05:57:28 <Niichan> http://i.imgur.com/DSBbSvj.png
2013-10-10 05:57:55 <ncopa> bad file number?
2013-10-10 05:58:08 <ncopa> never seen that
2013-10-10 05:58:23 <ncopa> can you please very sha1/md5 sum of iso image?
2013-10-10 05:58:25 <fabled> sounds corrupt boot media
2013-10-10 05:58:37 <Niichan> Except the sums match
2013-10-10 05:58:44 <fabled> or did we miss some syslinux files from the media?
2013-10-10 05:59:09 <ncopa> i'm pretty sure i testbooted it in qemu
2013-10-10 05:59:15 <ncopa> i better test again
2013-10-10 05:59:20 <Niichan> I use virtualbox
2013-10-10 05:59:34 <Niichan> It's a bit faster in my experience
2013-10-10 05:59:37 <fabled> x86 or x86_64 image?
2013-10-10 05:59:43 <Niichan> x86_64
2013-10-10 05:59:48 <Niichan> my tower is debian stable
2013-10-10 05:59:58 <fabled> i'm running x86_64 2.6.5 on real hardware just fine
2013-10-10 06:00:05 <fabled> though from usb
2013-10-10 06:00:16 <Niichan> I'm running edge on several servers just fine as well
2013-10-10 06:00:27 <Niichan> I want to set up a local virtualbox of a stable
2013-10-10 06:00:48 <ncopa> might be we missed something on 2.6.5-x86_64
2013-10-10 06:00:57 <ncopa> i will try reproduce in qemu/kvm
2013-10-10 06:01:55 <ncopa> does virtualbox support efi mode?
2013-10-10 06:02:07 <Niichan> yes
2013-10-10 06:02:37 <ncopa> verify that it does not use efi, but old style bios
2013-10-10 06:03:10 <Niichan> I have EFI disabled
2013-10-10 06:03:18 <ncopa> ok good
2013-10-10 06:03:26 <Niichan> same error message
2013-10-10 06:04:38 <ncopa> and i was able to boot it in qemu
2013-10-10 06:04:57 <ncopa> it does look like its syslinux that fails
2013-10-10 06:06:13 <Niichan> Good to know. I'm gonna head to bed
2013-10-10 06:06:32 <ncopa> you can try at boot prompt type: grsec pax_nouderef
2013-10-10 06:06:52 <ncopa> i know that is needed for vmware
2013-10-10 06:07:21 <Niichan> same error
2013-10-10 06:07:25 <Niichan> http://i.imgur.com/aVvWSBb.png
2013-10-10 06:08:17 <ncopa> weird, i have no idea
2013-10-10 06:08:26 <ncopa> it does smell like someting with syslinux
2013-10-10 06:08:49 <Niichan> If you get any thoughts just ping or memoserv me
2013-10-10 06:10:18 <fabled> looks like syslinux issue since it returns to boot loader prompt
2013-10-10 06:10:30 <fabled> is your boot media mounted?
2013-10-10 06:10:34 <fabled> properly
2013-10-10 06:10:38 <fabled> maybe syslinux does not see it has disk
2013-10-10 06:40:09 <ncopa> here comes poppler update
2013-10-10 06:43:36 <ncopa> barthalion: what was the problem with ffmpeg-2?
2013-10-10 06:44:05 <barthalion> vlc was the blocker
2013-10-10 06:44:20 <barthalion> but 2.1.0 works with ffmpeg 2.0
2013-10-10 06:44:26 <ncopa> ah nice
2013-10-10 06:44:41 <barthalion> still, some pkgs may need patching
2013-10-10 06:45:06 <ncopa> audacious-plugins
2013-10-10 06:45:06 <ncopa> ffmpeg
2013-10-10 06:45:06 <ncopa> vlc
2013-10-10 06:45:06 <ncopa> xbmc
2013-10-10 06:45:06 <ncopa> freerdp
2013-10-10 06:45:19 <barthalion> freerdp
2013-10-10 06:45:25 <barthalion> the rest should build fine
2013-10-10 06:45:31 <ncopa> does arch linux have patch for it?
2013-10-10 06:45:47 <barthalion> yeah, I wrote one
2013-10-10 06:45:59 <ncopa> nice
2013-10-10 06:46:03 <ncopa> maybe i should upgrade it then
2013-10-10 06:46:04 <barthalion> and no user complained so far, so it works
2013-10-10 06:46:08 <barthalion> or no one use freerdp :p
2013-10-10 06:46:14 <ncopa> lol
2013-10-10 06:46:32 <barthalion> https://projects.archlinux.org/svntogit/community.git/tree/trunk/ffmpeg2.0.patch?h=packages/freerdp
2013-10-10 06:47:24 <ncopa> ok, i'll try upgrade ffmpeg then
2013-10-10 07:08:39 <StarWarsFan> fabled: which hardware do you use to compile the rpi-stuff?
2013-10-10 07:08:56 <StarWarsFan> a real rpi or crosscompiling?
2013-10-10 07:30:03 <fabled> StarWarsFan, i have quad core wandboard (armv7)
2013-10-10 07:30:06 <fabled> so i native build on that
2013-10-10 07:39:29 <vkrishn> ncopa: simple qemu boot to v2.6.5 iso has this error msg "failed to access perfctr", but boots ok
2013-10-10 07:40:20 <vkrishn> alpine-mini-2.6.5-x86.iso
2013-10-10 07:51:41 <StarWarsFan> fabled: thx
2013-10-10 08:04:12 <ncopa> barthalion: we build xbmc against system ffmpeg
2013-10-10 08:04:37 <ncopa> hm
2013-10-10 08:17:19 <barthalion> CodecID should be AVCodecID
2013-10-10 08:21:35 <barthalion> ncopa: I'll send you a patch today or tommorow
2013-10-10 08:21:53 <ncopa> they have patched it in master branch
2013-10-10 08:22:04 <ncopa> https://github.com/aballier/xbmc/commit/54f118123eba02f56a240968dcff28c530b6cbc7
2013-10-10 08:22:22 <ncopa> there was a handful of conflicts
2013-10-10 08:22:30 <ncopa> and it still does not build after resolving
2013-10-10 08:22:55 <barthalion> that's exactly why I need more time, I don't remember some changes
2013-10-10 08:23:40 <ncopa>  /home/ncopa/aports/main/xbmc/src/xbmc-12.2/lib/DllAvFilter.h:140:106: error: '::av_buffersrc_add_frame' has not been declared
2013-10-10 08:23:40 <ncopa>    virtual int av_buffersrc_add_frame(AVFilterContext *buffer_filter, AVFrame* frame, int flags) { return ::av_buffersrc_add_frame(buffer_filter, frame, flags); }
2013-10-10 08:23:40 <ncopa>                                                                                                           ^
2013-10-10 08:24:51 <ncopa> looks like the last arg, flags, has been removed?
2013-10-10 08:25:21 <barthalion> or you should use av_buffersrc_add_frame_flags instead
2013-10-10 08:26:19 <barthalion> it takes the same arguments
2013-10-10 08:53:55 <fabled> barthalion, ^^^ mikmod no longer needs update_config_sub
2013-10-10 09:15:53 <kaniini> hmm
2013-10-10 09:33:48 <ncopa> kaniini: do you think you can help me write a proper xendomains init.d script?
2013-10-10 09:34:03 <ncopa> i think what we copied from gentoo is braindamaged
2013-10-10 09:34:15 <kaniini> i don't use that functionality, but sure
2013-10-10 10:23:09 <barthalion> algitbot: build master
2013-10-10 11:05:02 <ncopa> barthalion: taglib update is broke
2013-10-10 11:05:14 <ncopa> please test build before you push
2013-10-10 11:45:45 <fabled_RPI> hi
2013-10-10 11:46:15 <fabled_RPI> greetings from Raspberry Pi running Alpine Linux edge-musl
2013-10-10 11:46:54 <StarWarsFan> cool! :-)
2013-10-10 11:46:54 <ncopa> hi fabled_RPI!
2013-10-10 11:46:57 <ncopa> congrats!
2013-10-10 11:47:38 <fabled_RPI> i guess i should push the boot image somewhere
2013-10-10 11:47:56 <ncopa> dev.a.o/~fabled/
2013-10-10 11:48:02 <fabled_RPI> this is running 3.11.4 with rpi patches
2013-10-10 11:48:24 <fabled_RPI> ooh
2013-10-10 11:48:34 <fabled_RPI> i need to add the pie patch
2013-10-10 11:49:04 <fabled_RPI> i also have patches to alpine-iso to automatically build the .tar.gz boot image
2013-10-10 11:50:11 <mhavela> congrats!!! \o/
2013-10-10 11:56:15 <mhavela> re lua-posix (fabled asked earlier if it was broken): I think it's broken because acf broke when upraded
2013-10-10 11:56:15 <mhavela> lua-posix (5.1.26-r1 -> 5.1.28-r1)
2013-10-10 11:56:38 <fabled> it did not compile, we fixed it early today
2013-10-10 11:57:17 <mhavela> So I only need to wait for mirrors to replicate and I will have a new version?
2013-10-10 11:57:43 <mhavela> Because acf broke just now when I upgraded lua-posix
2013-10-10 11:57:57 <fabled> yes, but i'm not sure if it fixes your issue
2013-10-10 11:58:00 <fabled> ok
2013-10-10 11:58:07 <fabled> i need to recompile kernel once, and recreate the .tar.gz
2013-10-10 11:58:11 <fabled> and i'll push the rpi bootimage out
2013-10-10 11:58:18 <mhavela> Wonderful!
2013-10-10 11:59:04 <fabled> will take an hour or so. wandboard is not as fast as our master build server
2013-10-10 12:00:07 <fabled> hum
2013-10-10 12:00:23 <fabled> apk progress bar does not work in musl as expected. must related to how the output is cached.
2013-10-10 12:03:16 <mhavela> fabled: Have you rebuilt haserl against musl yet?
2013-10-10 12:03:16 <mhavela> Maybe that's the reason why I have trouble.
2013-10-10 12:03:29 <fabled> mhavela, ? you are using musl builds?
2013-10-10 12:03:40 <mhavela> Nope :-)
2013-10-10 12:03:46 <fabled> why would it affect you then?
2013-10-10 12:04:13 <fabled> any way, haserl is built; and required no changed to build
2013-10-10 12:04:17 <mhavela> Phantom pain?! (like when you cut a leg)
2013-10-10 12:04:28 <mhavela> Ok.
2013-10-10 12:04:50 <mhavela> 
ACTION writes a bugreport on it. Something just broke it tough

2013-10-10 12:09:34 <kaniini> fabled: fflush() it? :)
2013-10-10 12:10:37 <fabled> kaniini, actually, i think musl implicitly fflush():es it, and apk expects it not to
2013-10-10 12:10:51 <kaniini> oh
2013-10-10 12:10:53 <kaniini> hmm
2013-10-10 12:11:15 <mhavela> ncopa: Care to have a look at this: http://git.alpinelinux.org/cgit/aports/commit/main/lua-posix?id=f357d1469c9d529631650c552382dc741570370f
2013-10-10 12:11:31 <mhavela> lua-bitlib is removed as a dependency
2013-10-10 12:11:35 <mhavela> this seems to break ACF
2013-10-10 12:15:24 <mhavela> ncopa: Actually... it breaks kamailio too (and probably everything that uses lua-posix)
2013-10-10 12:15:46 <mhavela> Just confirmed that kamailio broke when lua-bitlib all of a sudden was purged
2013-10-10 12:16:05 <fabled> ncopa, seems that you didn't rebuild all dependencies of poppler after updating it.
2013-10-10 12:20:14 <ncopa> fabled: humm, which did i miss?
2013-10-10 12:20:38 <fabled> or maybe it's my build system just that failed
2013-10-10 12:23:48 <ncopa> mhavela: yes, lua-posix5.1 needs lua-bitlib
2013-10-10 12:25:51 <mhavela> ncopa: I created https://bugs.alpinelinux.org/issues/2278
2013-10-10 12:45:31 <fabled> >>> ERROR: taglib: taglib-1.9.1.tar.gz is missing in checksums
2013-10-10 12:46:02 <fabled> barthalion, ^^^
2013-10-10 12:46:03 <ncopa> barthalion: pushed a commit without checksum
2013-10-10 12:46:16 <ncopa> we need find a proper way to spank him :)
2013-10-10 12:47:20 <ncopa> ok, how about this...
2013-10-10 12:47:48 <ncopa> when a dev pushes a commit that breaks the build servers, he has to buy a beer to *every* other developer :)
2013-10-10 12:48:10 <ncopa> (after fixing it)
2013-10-10 12:48:32 <ncopa> if other developer beats him to it he'll have to buy 2 for everyone
2013-10-10 12:51:08 <StarWarsFan> there should be something like an alpine-developer-convention for such things (beside others of course) ;-)
2013-10-10 13:21:02 <fabled> ok
2013-10-10 13:21:12 <fabled> i have rpi image ready, would someone like to test it?
2013-10-10 13:21:57 <fabled> http://dev.alpinelinux.org/~tteras/alpine-edge-131010-armhf.rpi.tar.gz
2013-10-10 13:22:26 <fabled> create empty FAT32 sd/mmc card, and extract that to it
2013-10-10 13:22:59 <fabled> the repository to add is http://nl.alpinelinux.org/alpine/edge-musl/main
2013-10-10 13:27:54 <ncopa> ugh
2013-10-10 13:28:08 <ncopa> openjdk6 is blocking x86_64 now
2013-10-10 13:35:29 <ncopa> algitbot: build master
2013-10-10 13:44:07 <fc1979> fabled, got your image
2013-10-10 13:44:26 <fc1979> extracted on a mmc
2013-10-10 13:44:35 <fc1979> on a Raspberry Pi
2013-10-10 13:44:41 <fc1979> but i have black screen
2013-10-10 13:46:40 <fabled> you might need to craft the partition table specially: http://elinux.org/RPi_Advanced_Setup
2013-10-10 13:46:51 <fabled> and the partition needs to be fat32
2013-10-10 13:47:13 <fabled> for alpine, you can use MMC with single partition of fat32 type
2013-10-10 13:47:33 <fcolista_> k
2013-10-10 13:47:42 <fcolista_> i've a partition _formatted_ with fat32
2013-10-10 13:47:49 <fcolista_> but i think that type is not fat32
2013-10-10 13:48:18 <fcolista_> 
ACTION always wondered what "type" of partition meant

2013-10-10 13:48:22 <fabled> did that mmc work earlier with rpi?
2013-10-10 13:48:35 <fabled> if not, you might need to create the partition table as instructured in the elinux.org page
2013-10-10 13:48:48 <fcolista_> i got the same sd with Archlinux ARM
2013-10-10 13:48:51 <fabled> pay attention to setting right amount of heads and sectors
2013-10-10 13:49:16 <fcolista_> ok
2013-10-10 13:49:18 <fabled> in any case, you should see the raspberry pi flash image first thing
2013-10-10 13:50:18 <fabled> you can also try removing "hdmi_safe=1" from config.txt in the root of the image
2013-10-10 13:50:20 <Niichan> fabled: 45 MB for an OS install? Nice!
2013-10-10 13:50:38 <fabled> it's the alpine-edge minimum profile.
2013-10-10 13:50:50 <fabled> lots of it is kernel modules
2013-10-10 13:50:53 <fabled> could be smaller
2013-10-10 13:51:12 <Niichan> When I get access to my pi I'm gonna get that up and running. Does it include a DHCP client?
2013-10-10 13:51:14 <fabled> contains enough to do netinstall
2013-10-10 13:52:04 <fabled> it has busybox, apk-tools, openssl, musl, openrc and few other core packages
2013-10-10 13:52:47 <Niichan> How do I set up a cross-compiler for musl arm?
2013-10-10 13:53:03 <fabled> i native build
2013-10-10 13:53:27 <Niichan> Ah
2013-10-10 13:53:30 <fabled> a minimum info set on what i did to bootstrap / crosscompile, is at http://wiki.alpinelinux.org/wiki/User:Fabled
2013-10-10 13:53:54 <Niichan> I'd like to cross compile on my server so I can make a repository
2013-10-10 13:54:13 <fabled> Niichan, only limited set of aports is cross-buildable
2013-10-10 13:54:25 <Niichan> fabled: Because of musl issues?
2013-10-10 13:54:33 <kaniini> no
2013-10-10 13:54:36 <fabled> no, because of lots of upstream packages not supporting cross-compile
2013-10-10 13:54:41 <Niichan> Ah
2013-10-10 13:54:46 <kaniini> because the APKBUILDs don't support it too
2013-10-10 13:54:58 <kaniini> we only put in cross support for bootstrap only
2013-10-10 13:55:29 <fabled> the musl source build stats are:
2013-10-10 13:55:31 <fabled> musl-x86: 1438 of 1671
2013-10-10 13:55:31 <fabled> musl-arm: 1306 of 1671
2013-10-10 13:56:03 <Niichan> Ah
2013-10-10 13:56:51 <Niichan> Thanks much for getting a build up fabled :3
2013-10-10 13:56:59 <Niichan> Is it a grsec kernel too?
2013-10-10 13:57:32 <fabled> no
2013-10-10 13:57:40 <fabled> RPi has separate set of patches
2013-10-10 13:57:45 <fabled> so it's the stock RPi kernel
2013-10-10 13:57:49 <Niichan> Ah
2013-10-10 13:58:18 <kaniini> we really need to fix the kernel situation at some point
2013-10-10 13:58:20 <Niichan> Does this image has any persistence?
2013-10-10 13:58:28 <fabled> regular 'lbu commit' works
2013-10-10 13:58:43 <fabled> and enabling apk caching makes it stash downloaded packages
2013-10-10 13:58:44 <Niichan> Oh I see
2013-10-10 13:58:54 <Niichan> clever
2013-10-10 13:58:59 <kaniini> you can probably also do a disk install
2013-10-10 13:59:02 <kaniini> to the sd card
2013-10-10 13:59:06 <kaniini> but i assume, not tested
2013-10-10 13:59:12 <fabled> yes, that'd be simple to do as well
2013-10-10 13:59:31 <fabled> just partition the SD properly, do disk install, and update cmdline.txt to have root=mmcblk0p2
2013-10-10 13:59:46 <Niichan> Hell, I'll go grab a 512 MB SD card and make a proper disk install image
2013-10-10 13:59:52 <kaniini> you can use setup-disk -m sys /path/to/filesystem
2013-10-10 14:01:46 <Niichan> nderstood
2013-10-10 14:01:51 <Niichan> Understood*
2013-10-10 14:02:58 <Niichan> I love how you only have 17 packages on an install
2013-10-10 14:03:52 <fabled> and there's now one less
2013-10-10 14:03:57 <fabled> abuildhelper is removed from aports
2013-10-10 14:04:48 <fabled> oh
2013-10-10 14:04:51 <fabled> that's different
2013-10-10 14:09:11 <fabled> ok
2013-10-10 14:09:13 <fabled> i'm off for the evening
2013-10-10 14:09:15 <fabled> see you tm
2013-10-10 14:09:22 <fabled> happy Pi testing
2013-10-10 14:09:36 <Niichan> night fabled, I'll be sure to let you know how it goes
2013-10-10 14:10:16 <fabled> as said earlier, one might need to tweak config.txt / cmdline.txt depending on the intended setup
2013-10-10 14:10:24 <fabled> but at least my unit boots with the default image
2013-10-10 14:10:32 <Niichan> I'll get hacking
2013-10-10 14:10:50 <Niichan> I'll see if I can make a bootable image with a full system install
2013-10-10 14:11:22 <Niichan> and some swap, the pi needs swap >_>
2013-10-10 14:11:50 <fabled> why? mine has 420MB free
2013-10-10 14:11:52 <fabled> :D
2013-10-10 14:12:07 <Niichan> I have a 256 pi :P
2013-10-10 14:12:24 <fabled> well, it's using under 21MB after booting to cmdline...
2013-10-10 14:12:33 <fabled> musl is quite nice on resource usage wise, as well as busybox
2013-10-10 14:12:44 <Niichan> :3
2013-10-10 14:12:59 <fabled> and that's when everything is installed to tmpfs
2013-10-10 14:13:27 <fabled> ok
2013-10-10 14:13:28 <fabled> gotta run
2013-10-10 14:13:29 <fabled> ttyl
2013-10-10 15:24:15 <Niichan> Is there a way to force a certian LBU overlay to be applied?
2013-10-10 16:02:03 <mhavela> You mean like putting the apkovl.tar.gz on the boot media?
2013-10-10 16:08:47 <Niichan> I'll figure it out when I have a HDMI cable
2013-10-10 16:10:50 <mhavela> Did you get a RPI up and running?
2013-10-10 16:13:39 <Niichan> I can't tell if it is or not
2013-10-10 16:13:44 <Niichan> I do not have an HDMI cable
2013-10-10 16:13:49 <mhavela> ahh :-)
2013-10-10 16:13:54 <Niichan> But the lights on it are flashing like it is booted
2013-10-10 16:13:55 <mhavela> I am trying to get mine boot now
2013-10-10 16:14:06 <mhavela> still have black screen (and I do have a HDMI cable)
2013-10-10 16:14:28 <mhavela> seems I set wrong system id on the partition
2013-10-10 16:14:32 <mhavela> need to fix that first
2013-10-10 16:16:18 <mhavela> I wonder if I need to
2013-10-10 16:16:19 <mhavela> dd if=/usr/share/syslinux/mbr.bin of=/dev/sda
2013-10-10 16:16:19 <mhavela> syslinux /dev/sda1
2013-10-10 16:16:32 <Niichan> the rpi doesn't use syslinux
2013-10-10 16:16:36 <mhavela> ok.
2013-10-10 16:16:40 <mhavela> thanks
2013-10-10 16:19:17 <mhavela> It's.... It's.... It's ALIVE \o/
2013-10-10 16:23:43 <mhavela> This is so coool!
2013-10-10 18:57:06 <tdtrask> mhavela-rpi: nice nickname :)
2013-10-10 18:58:02 <tdtrask> getting closer to my dream of an alpine-based replacement for OpenWRT
2013-10-10 19:00:02 <clandmeter> :)
2013-10-10 19:39:56 <clandmeter> w00t
2013-10-10 19:40:01 <clandmeter> it boot
2013-10-10 19:40:06 <clandmeter> \o.
2013-10-10 19:40:08 <clandmeter> \o/
2013-10-10 19:47:01 <clandmeter> lets start building on this rpi :)
2013-10-11 04:41:10 <Niichan> fabled: You around?
2013-10-11 05:06:46 <fabled> morning
2013-10-11 05:27:18 <StarWarsFan> Moinmoin
2013-10-11 05:42:10 <Niichan> fabled: You around?
2013-10-11 05:42:18 <fabled> yes
2013-10-11 05:42:42 <Niichan> Can you send me a tar.gz dd image of your pi?
2013-10-11 05:43:45 <fabled> oh?
2013-10-11 05:43:55 <Niichan> Yeah
2013-10-11 05:44:03 <Niichan> It's not working for me
2013-10-11 05:44:17 <fabled> no image?
2013-10-11 05:44:31 <Niichan> It's not applying the overlay as far as I can tell
2013-10-11 05:44:39 <fabled> it should boot without overlay too
2013-10-11 05:44:42 <Niichan> but my only way to test it is to ping it
2013-10-11 05:44:46 <Niichan> and I have no head for it
2013-10-11 05:44:52 <fabled> oh, you have no screen ?
2013-10-11 05:45:52 <fabled> i can send my overlay that has networking on with dhcp?
2013-10-11 05:46:08 <fabled> you cannot just copy overlay from non-arm system unless you also delete the etc/apk/arch from it
2013-10-11 05:46:45 <Niichan> Yeah, that'd work
2013-10-11 05:46:52 <Niichan> I need to set a static IP though
2013-10-11 05:47:39 <fabled> and enable ssh
2013-10-11 05:47:43 <fabled> mmm
2013-10-11 05:48:14 <Niichan> fabled: can you send me your overlay?
2013-10-11 05:48:22 <fabled> ssh is not in the base apks
2013-10-11 05:48:33 <Niichan> I added ssh to my apks
2013-10-11 05:48:34 <fabled> that wouldn't work as-is
2013-10-11 05:48:41 <fabled> you re-created it?
2013-10-11 05:48:52 <fabled> it's not picked up
2013-10-11 05:49:00 <fabled> unless you re-create APKINDEX and sign it with proper key
2013-10-11 05:49:06 <Niichan> bah
2013-10-11 05:49:41 <fabled> ok, i create a new tarball with the cached packages
2013-10-11 05:49:58 <Niichan> Okay, thanks :)
2013-10-11 05:51:11 <fabled> but you also need static ip?
2013-10-11 05:51:19 <fabled> i guess you can edit that in the overlay
2013-10-11 05:51:23 <Niichan> Yeah, I have it set in my overlay
2013-10-11 05:52:09 <fabled> ncopa, we should likely add openssh to alpine-edge included packages profile
2013-10-11 05:52:32 <Niichan> I'm kinda suprised it's not a default install package in alpine
2013-10-11 05:52:46 <fabled> it's on the "regular" disk images
2013-10-11 05:54:54 <fabled> Niichan, try http://dev.alpinelinux.org/~tteras/rpi-with-networking.tar
2013-10-11 05:55:04 <fabled> it's regular tar, not gzipped
2013-10-11 05:55:18 <fabled> contents of the mmc
2013-10-11 05:55:26 <fabled> with networking + ssh enabled
2013-10-11 05:55:46 <Niichan> <3
2013-10-11 05:56:19 <Niichan> as soon as I can I'm going to be making a proper sys install image
2013-10-11 05:56:34 <fabled> root password is rpi
2013-10-11 05:57:34 <fabled> huh
2013-10-11 05:58:11 <fabled> something does not work right
2013-10-11 05:59:10 <fabled> bah
2013-10-11 05:59:24 <fabled> the password wont work over ssh due to it being set when clock was incorrect
2013-10-11 05:59:43 <Niichan> How come?
2013-10-11 06:02:27 <fabled> ok
2013-10-11 06:02:30 <fabled> the tarball is updated
2013-10-11 06:02:37 <fabled> re-download rpi-with-networking.tar
2013-10-11 06:02:43 <fabled> it should have now working ssh
2013-10-11 06:02:54 <fabled> and chrony enabled to pick time from network
2013-10-11 06:03:39 <Niichan> From 192.168.1.109 icmp_seq=42 Destination Host Unreachable
2013-10-11 06:03:41 <Niichan> :(
2013-10-11 06:04:01 <Niichan> fabled: What partition type is your fat partition?
2013-10-11 06:06:11 <Niichan> fabled: I have http://i.imgur.com/0pc3X7y.png
2013-10-11 06:07:06 <fabled> i used FAT32, partition type 0xb (win95 fat32)
2013-10-11 06:07:31 <Niichan> Is there a way to in-place upgrade a partition from fat16 to fat32?
2013-10-11 06:07:40 <fabled> not sure
2013-10-11 06:08:00 <fabled> then again
2013-10-11 06:08:04 <fabled> well
2013-10-11 06:08:25 <Niichan> assuming no
2013-10-11 06:08:32 <Niichan> 
ACTION reformats partition

2013-10-11 06:08:37 <fabled> rpi guides all say to use fat32
2013-10-11 06:08:52 <fabled> it might be that the rpi rom is hardwired to only handle fat32
2013-10-11 06:11:10 <Niichan> I'll try to find my pi over ssh
2013-10-11 06:11:17 <Niichan> s/ssh/nmap
2013-10-11 06:12:19 <Niichan> fabled: By default will it get an address from dhcp?
2013-10-11 06:12:24 <fabled> yes
2013-10-11 06:12:43 <Niichan> 
ACTION deletes overlay

2013-10-11 06:13:07 <fabled> uh
2013-10-11 06:13:14 <Niichan> oh wait
2013-10-11 06:13:15 <fabled> you need the overlay for ssh to start
2013-10-11 06:13:16 <Niichan> I am a mong
2013-10-11 06:18:54 <Niichan> fabled: When I try to login over ssh it says my password expired
2013-10-11 06:18:58 <Niichan> fabled: every time
2013-10-11 06:19:04 <fabled> you have the old .tar image
2013-10-11 06:19:13 <fabled> re-download the .tar
2013-10-11 06:19:17 <fabled> i fixed it
2013-10-11 06:19:21 <Niichan> bah
2013-10-11 06:19:31 <fabled> that's the password issue i mentioned
2013-10-11 06:26:19 <Niichan> Now to find my pi
2013-10-11 06:26:40 <Niichan> Nmap scan report for 192.168.1.126
2013-10-11 06:26:42 <Niichan> Host is up (0.00092s latency).
2013-10-11 06:26:44 <Niichan> PORT   STATE  SERVICE
2013-10-11 06:26:46 <Niichan> 22/tcp closed ssh
2013-10-11 06:26:48 <Niichan> :<
2013-10-11 06:28:49 <mhavela-rpi> Seems my rpi is not starting up packages during boot (as it should)
2013-10-11 06:28:54 <mhavela-rpi> maybe you are having the same
2013-10-11 06:29:01 <Niichan> yeah
2013-10-11 06:29:43 <fabled> i think it's because the clock is off
2013-10-11 06:30:09 <Niichan> Can you disable password expiry on your image then?
2013-10-11 07:27:10 <shafire> hi
2013-10-11 07:31:05 <shafire> I bought 2 raspberry pi now :S
2013-10-11 07:32:12 <ncopa> fabled, i think you're right about me messing up with poppler...
2013-10-11 07:35:34 <shafire> http://alpinelinux.org/forum/general-discussion/heartbeat
2013-10-11 07:36:27 <shafire> http://alpinelinux.org/forum/networking/setting-satellite-internet-connection-alpine-linux-254 <- hard to help him, or?
2013-10-11 07:55:28 <fabled> Niichan, did you get the new .tar?
2013-10-11 07:55:55 <fabled> in the root image you should have cache/chrony-1.29-r0.xxxxx.apk file
2013-10-11 08:27:55 <clandmeter> fabled: thank you very much for my bday present!
2013-10-11 08:36:45 <shafire> where can I find the images?
2013-10-11 08:37:34 <fabled> http://dev.alpinelinux.org/~tteras/alpine-edge-131010-armhf.rpi.tar.gz
2013-10-11 08:38:07 <fabled> for mmc content: boots like alpine-edge iso: no network, root login with minimal setup
2013-10-11 08:38:17 <fabled> apk add from the edge-musl additional stuff you need
2013-10-11 08:39:03 <shafire> whats a mmc content?
2013-10-11 08:40:14 <fabled> content for the MMC / SD card
2013-10-11 08:40:57 <shafire> ah ok
2013-10-11 08:42:30 <clandmeter> fabled: what about VideoCore stuff in rpi repo? can we use that?
2013-10-11 08:43:06 <clandmeter> looks like some stuff is precompiled.
2013-10-11 08:43:18 <clandmeter> https://github.com/Hexxeh/rpi-firmware/tree/master/vc/hardfp/opt/vc
2013-10-11 08:43:40 <fabled> will not work
2013-10-11 08:43:43 <fabled> well
2013-10-11 08:43:53 <fabled> with musl there's a _small_ change that it might
2013-10-11 08:44:02 <fabled> but they have shipped now the userspace side sources
2013-10-11 08:44:03 <clandmeter> i guess thats propertary?
2013-10-11 08:44:07 <fabled> so we should build it
2013-10-11 08:44:40 <fabled> https://github.com/raspberrypi/userland
2013-10-11 08:45:31 <clandmeter> fabled: i would like to try https://github.com/linuxstb/pidvbip
2013-10-11 08:45:41 <clandmeter> but the configure is looking for the sdk
2013-10-11 08:46:02 <fcolista_> what's the repo for alpine-rpi ?
2013-10-11 08:46:18 <clandmeter> fcolista_: aports :D
2013-10-11 08:46:22 <fcolista_> fabled mentioned yeterday, but i miss it
2013-10-11 08:46:45 <clandmeter> you mean apk repo
2013-10-11 08:46:48 <fabled> fcolista_, http://nl.alpinelinux.org/alpine/edge-musl/main/
2013-10-11 08:47:00 <fcolista_> yep
2013-10-11 08:47:02 <fcolista_> thx fabled
2013-10-11 08:47:31 <fcolista_> gott rpi up and running, hope that i'm not still dreaming...
2013-10-11 08:47:34 <fcolista_> *gotta
2013-10-11 08:49:49 <fabled> clandmeter, sure. step 1 is to package/build the rpi userland
2013-10-11 08:49:57 <fabled> then the rest should be straightforward
2013-10-11 08:50:10 <clandmeter> fabled: nice
2013-10-11 08:50:21 <clandmeter> i guess only the boot firmware is propetary?
2013-10-11 08:50:25 <fabled> yes
2013-10-11 08:50:33 <clandmeter> nice
2013-10-11 08:50:35 <fabled> in the beginning they had other parts closed source as well
2013-10-11 08:50:47 <fabled> but the userland was opensourced some time ago
2013-10-11 08:50:54 <clandmeter> fcolista_: check https://github.com/linuxstb/pidvbip
2013-10-11 08:51:08 <fabled> it's still kinda lame as the userland is basically just wrappers to send all the interesting stuff as messages to the videocore gpu
2013-10-11 08:51:10 <clandmeter> it can connect to tvheadend and watch tv on rpi without xbmc :)
2013-10-11 08:51:42 <fabled> the arm repo has 1329 of 1671 source packages built
2013-10-11 08:51:50 <clandmeter> fabled: well, its better than any other project untill now.
2013-10-11 08:52:01 <clandmeter> i mean arm of course.
2013-10-11 08:52:22 <fabled> it misses xorg-server and various things depending on it. and some of the complex projects that do not build against musl (yet).
2013-10-11 08:52:55 <clandmeter> openmax can run without xorg afaik
2013-10-11 08:53:18 <clandmeter> atleast thats what pidvpip does
2013-10-11 08:59:14 <fcolista_> clandmeter, that's gr8!
2013-10-11 09:00:17 <clandmeter> it has CEC build in, so you can use your tv remote to control it.
2013-10-11 09:02:00 <fcolista_> ok
2013-10-11 09:02:17 <fcolista_> so next step is having an alpine build env for arm
2013-10-11 09:11:29 <clandmeter> fcolista_: you can use your rpi, except its a bit slowish :)
2013-10-11 09:11:46 <clandmeter> maybe buy 4 and use distcc
2013-10-11 09:23:00 <ncopa> yay...
2013-10-11 09:23:08 <ncopa> im getting a new build server today
2013-10-11 09:23:22 <ncopa> 2 x 6cores + hyperthreading
2013-10-11 09:23:24 <clandmeter> 32core?
2013-10-11 09:23:37 <ncopa> so virtually 24 cpus...
2013-10-11 09:23:41 <clandmeter> nice
2013-10-11 09:23:54 <ncopa> and in few weeks i'l get another of those...
2013-10-11 09:24:12 <clandmeter> somebody must like you
2013-10-11 09:25:38 <ncopa> they said it was "leftovers"
2013-10-11 09:26:04 <ncopa> 32G ram
2013-10-11 09:26:45 <clandmeter> I guess you are satisfied with leftovers?
2013-10-11 09:26:58 <ncopa> i guess i am :)
2013-10-11 09:27:02 <clandmeter> or are you going to complain and try to get the new ones?
2013-10-11 09:27:46 <ncopa> i think i'm more than satisfied :)
2013-10-11 09:27:54 <ncopa> i could set up distcc
2013-10-11 09:28:03 <ncopa> and build kernels with -j48 ...
2013-10-11 09:28:42 <clandmeter> :)
2013-10-11 09:28:51 <clandmeter> i dont think the wandboard can keep up
2013-10-11 09:28:58 <ncopa> probably not :)
2013-10-11 09:30:40 <shafire> http://www.extremetech.com/wp-content/uploads/2013/01/OpenComputeRack.jpg
2013-10-11 09:30:51 <shafire> thats how wandboard could keep up
2013-10-11 09:31:32 <ncopa> lol
2013-10-11 09:33:58 <ncopa> fabled: apk-tools issue: https://dpaste.de/j694
2013-10-11 09:38:02 <ncopa> ok
2013-10-11 09:38:04 <ncopa> the problem is
2013-10-11 09:38:28 <ncopa> poppler-gtk used to ship so:libpoppler-glib.so.8=8.6.0
2013-10-11 09:38:36 <ncopa> with new poppler i removed poppler-gtk
2013-10-11 09:38:58 <ncopa> and added added poppler-glib as subpackage to poppler
2013-10-11 09:39:27 <ncopa> apk tools will say: i need so:libpoppler-glib.so.8=8.6.0
2013-10-11 09:39:42 <ncopa> and will find that poppler-gtk satisfies that
2013-10-11 09:39:45 <ncopa> and is happy with that
2013-10-11 09:43:00 <ncopa> yup
2013-10-11 09:43:10 <ncopa> sudo apk add --simulate '!poppler-gtk'
2013-10-11 09:43:17 <ncopa> seems to resolve that
2013-10-11 09:43:19 <fcolista_> clandmeter, probably a qemu arm performs better ;)
2013-10-11 09:44:05 <clandmeter> fcolista_: according to fabled, qemu is not fast enough.
2013-10-11 09:44:30 <fcolista_> really? I miss that part
2013-10-11 09:44:47 <fcolista_> umh
2013-10-11 09:44:56 <fcolista_> i think i may use a tablet
2013-10-11 10:05:25 <fabled> clandmeter, qemu-user would likely be fast, but it's broken in other ways currently
2013-10-11 10:05:39 <fabled> qemu-system would work, but i didn't test how fast it is
2013-10-11 10:10:54 <ncopa> apk tools dont wan upgrade ffmpeg-libs for me: https://dpaste.de/ozEf
2013-10-11 10:11:29 <ncopa> https://dpaste.de/1KGZ
2013-10-11 10:49:37 <fabled> ncopa, audcious-plugins is installed but not recompiled
2013-10-11 10:50:27 <fabled> and the two versions of the libs cannot co-exist since some .so's have same version numbers
2013-10-11 10:50:31 <fabled> thus they have conflicting provides
2013-10-11 10:56:25 <shafire> I got my label printer :)
2013-10-11 11:09:16 <ncopa> hum
2013-10-11 11:09:30 <ncopa> we have a problem with audacious-plugins and audacious
2013-10-11 11:09:49 <ncopa> to build audacious-plugins you need audacious-dev
2013-10-11 11:10:16 <ncopa> audacious-dev depends on audacious
2013-10-11 11:10:45 <ncopa> which will pull in audacious-plugins (which has install_if=audacious)
2013-10-11 11:13:02 <ncopa> hm
2013-10-11 11:13:03 <ncopa> no
2013-10-11 11:13:20 <ncopa> its only problem if you have http repo with old audacious
2013-10-11 11:26:11 <shafire> hi mhavela
2013-10-11 11:30:57 <ncopa> algitbot: build master
2013-10-11 11:33:48 <ncopa> algitbot: build master
2013-10-11 11:41:52 <mhavela> hi shafire
2013-10-11 11:44:07 <shafire> who are you? :)
2013-10-11 11:44:37 <mhavela-rpi> I am cloned :-))
2013-10-11 11:44:41 <mhavela-rpi> And... I am
2013-10-11 11:44:53 <mhavela> happy that I got Alpine working on RPI \o/
2013-10-11 11:44:59 <mhavela-rpi> \o/
2013-10-11 11:45:33 <shafire> could you try out the broadcom rng? :)
2013-10-11 11:58:24 <ncopa> cpuinfo from new server: https://dpaste.de/CtKg
2013-10-11 11:58:38 <fabled> i noticed that mkfs.vfat is broken
2013-10-11 11:58:41 <fabled> bb version works
2013-10-11 11:59:13 <fabled> but the one from dosfstools creates too small partitions
2013-10-11 11:59:33 <ncopa> sounds weird
2013-10-11 12:00:06 <fabled> dosfstools creates only partitions up to 2GB, even if the partition is bigger
2013-10-11 12:00:55 <shafire> like I said, exfat ftw. :P
2013-10-11 12:01:14 <shafire> Need to check all my vfat paritions now
2013-10-11 12:01:31 <ncopa> fabled: sounds like dosfstools is not compiled with largefile support or similar
2013-10-11 12:01:53 <shafire> ncopa: do you colocate your build server?
2013-10-11 12:02:18 <ncopa> more upgrading the hardware
2013-10-11 12:02:31 <shafire> 12mb cpu cache :)
2013-10-11 13:52:56 <fcolista_> how have did you set your raspberry's hostname on alpine? I'm the only one to have set "raspine" ?
2013-10-11 13:52:58 <fcolista_> :)
2013-10-11 13:55:34 <fabled> fcolista_, edit /etc/hostname, and /etc/init.d/hostname restart
2013-10-11 13:56:16 <fcolista_> fabled, i mean what name have you set, not how to do it ;)
2013-10-11 13:56:25 <fcolista_> i know how to set an hostname :)
2013-10-11 13:57:57 <fabled> fcolista_, oh. i have just 'rpi' for now
2013-10-11 13:58:37 <fcolista_> debian use raspbian
2013-10-11 13:58:50 <fcolista_> i called it raspine
2013-10-11 13:59:14 <fcolista_> fabled, what do you think is worthy testing as first?
2013-10-11 13:59:29 <fabled> i dunno
2013-10-11 13:59:44 <fabled> i guess it depends on what you want to do with it ;)
2013-10-11 13:59:53 <fabled> i still haven't built bigger things for it
2013-10-11 14:00:08 <fabled> like vlc or asterisk
2013-10-11 14:00:14 <fabled> i think i'd like to use those
2013-10-11 14:03:39 <fcolista_> k
2013-10-11 14:23:07 <shafire> fabled: the arm image, is this for all arm boards?
2013-10-11 14:23:24 <fabled> shafire, the .tar.gz is for raspberry pi.
2013-10-11 14:23:41 <fabled> the .apk packages should run on any arm board with armv6+ and vfp
2013-10-11 14:23:58 <shafire> so, we need for every board an own image?
2013-10-11 14:24:00 <fabled> for other boards, you'd likely need to build kernel + bootloader
2013-10-11 14:24:03 <fabled> well
2013-10-11 14:24:14 <fabled> most new boards can have shared kernel image
2013-10-11 14:24:22 <fabled> but bootloader would need to be separate
2013-10-11 14:26:00 <build64-2-5> files from v2.5.4-190-gb999f12 uploaded
2013-10-11 14:26:13 <build-2-5> files from v2.5.4-190-gb999f12 uploaded
2013-10-11 14:27:13 <shafire> Ok
2013-10-11 14:27:38 <shafire> what does the hf in armhf mean
2013-10-11 14:28:07 <fabled> hard float
2013-10-11 14:28:23 <fabled> you need vfpu floating point processer on board
2013-10-11 14:28:34 <fabled> which is present on all modern arm boards
2013-10-11 14:28:51 <shafire> I bought two raspberry pi to test
2013-10-11 14:28:52 <shafire> :)
2013-10-11 14:29:07 <shafire> two because of free shipping hehe
2013-10-11 14:49:06 <shafire> my I ask why there is no update to the newest syslinux?
2013-10-11 15:29:45 <shafire> how can I check, if vfat has taken the full partition?
2013-10-11 15:29:57 <shafire> @fabled
2013-10-11 15:49:07 <shafire> cool
2013-10-11 15:49:22 <shafire> running from tmpfs removed my syslinux issue with amt kvm
2013-10-11 16:13:46 <Niichan> fabled: You here?
2013-10-11 16:13:52 <fabled> no ;)
2013-10-11 16:14:11 <Niichan> What's the command to install an alpine system to an arbitrary partition?
2013-10-11 16:14:17 <fabled> setup-disk
2013-10-11 16:16:20 <Niichan> Hmm, can I just do the same process for making a chroot?
2013-10-11 16:18:13 <fabled> i think so, you just give path to device for disk install, and path to chroot for chroot install
2013-10-11 16:18:52 <Niichan> \o/
2013-10-11 16:24:23 <Niichan> fabled: Is it normal for a fresh install to only use 7MB of disk space?
2013-10-11 16:27:06 <fabled> if you have only base installed - yes
2013-10-11 16:27:20 <Niichan> Fuck
2013-10-11 16:27:45 <fabled> i mean - that's the what the "userland" takes about
2013-10-11 16:27:50 <fabled> kernel is separate
2013-10-11 16:27:57 <Niichan> fabled: Did I mention how much I love this fucking OS?
2013-10-11 16:28:19 <Niichan> It is amazing
2013-10-11 16:28:20 <fabled> if you install kernel then you go up to 40-50MB
2013-10-11 16:28:27 <fabled> mostly just for the kernel modules
2013-10-11 16:28:34 <fabled> :)
2013-10-11 16:28:36 <Niichan> That blows Arch's 500 MB out of the water
2013-10-11 16:29:17 <fabled> i think they default to install all kind of crap
2013-10-11 16:29:29 <Niichan> Alpine is like "Lol no"
2013-10-11 16:29:49 <Niichan> Hmm
2013-10-11 16:30:04 <Niichan> What's the argument for the kernel to specify an arbitrary root device?
2013-10-11 16:31:23 <fabled> root=xxx
2013-10-11 16:32:05 <fabled> root=/dev/mmcblk0p2 or so
2013-10-11 16:32:41 <grimeton> hi
2013-10-11 16:32:49 <grimeton> dl-5 was down due to an harddrive issue
2013-10-11 16:32:53 <grimeton> for about 3 hours
2013-10-11 16:32:58 <grimeton> it's back online again and syncing
2013-10-11 16:34:31 <shafire> ok
2013-10-11 16:35:32 <shafire> thanks grimeton :)
2013-10-11 16:36:04 <Niichan> fabled: I have some interesting output from mkinitfs, getting a gist for you
2013-10-11 16:36:21 <fabled> oh
2013-10-11 16:36:23 <fabled> right
2013-10-11 16:36:26 <fabled> that's not fixed yet
2013-10-11 16:36:32 <fabled> musl needs some updates to mkinitfs
2013-10-11 16:36:34 <Niichan> https://gist.github.com/Niichan/6b5495089f65e36d81f2
2013-10-11 16:36:37 <Niichan> oh okay
2013-10-11 16:36:49 <Niichan> Just making sure you know ^_^
2013-10-11 16:36:51 <fabled> yes, those bogus errors are ok
2013-10-11 16:36:59 <fabled> but it needs few additions
2013-10-11 16:37:08 <Niichan> Is it good to boot at this point?
2013-10-11 16:37:17 <fabled> no, asec
2013-10-11 16:38:02 <Niichan> asec?
2013-10-11 16:38:12 <fabled> update /etc/mkinitfs/files.d/base to contain http://sprunge.us/NNjG
2013-10-11 16:38:29 <fabled> there's two or three lines added for musl
2013-10-11 16:38:39 <fabled> after that mkinitfs should generate usable images
2013-10-11 16:39:13 <grimeton> Niichan: a second
2013-10-11 16:39:21 <Niichan> Ah
2013-10-11 16:39:26 <fabled> though, due to the way rpi is built, the mkinitfs generated image will not be used unless you manually copy it from /boot/*.gz to /media/mmcblk0p1
2013-10-11 16:39:27 <Niichan> re-running mkinitfs
2013-10-11 16:39:36 <fabled> the initrd.gz needs to be on the boot partition
2013-10-11 16:39:45 <grimeton> is the xfs problem fixed?
2013-10-11 16:39:46 <fabled> and it's not updated automatically by current script
2013-10-11 16:39:56 <Niichan> fabled: Gotcha, I'll do it by hand
2013-10-11 16:40:18 <Niichan> fabled: Is it /rpi.gz?
2013-10-11 16:40:28 <fabled> yes
2013-10-11 16:40:57 <shafire> how can I use the testing kernel from tmpfs?
2013-10-11 16:41:11 <shafire> how can I use the testing kernel in a tmpfs environment?
2013-10-11 16:41:13 <Niichan> fabled: Do I have to add a kernel argument?
2013-10-11 16:42:17 <Niichan> oh wait no derp
2013-10-11 16:42:46 <Niichan> We're going for it now
2013-10-11 16:44:34 <Niichan> damn
2013-10-11 16:46:02 <Niichan> fabled: Here is the cmdline.txt that was bad: http://git.io/WmMVxw
2013-10-11 16:46:32 <Niichan> additionally the fstab I made: http://git.io/Pf8MHg
2013-10-11 16:46:38 <shafire> f*** :(
2013-10-11 16:47:40 <Niichan> fabled: My SD card has partition 1 as boot, partiton 2 as 128 MB of swap, partition 3 as the root filesystem
2013-10-11 16:47:54 <fabled> mmm
2013-10-11 16:48:08 <Niichan> anything look immediately wrong?
2013-10-11 16:48:41 <fabled> it kinda looks ok. do you get an error, or what happens?
2013-10-11 16:48:50 <Niichan> I can't ssh into it
2013-10-11 16:49:00 <Niichan> I don't have a monitor I can plug this into ><
2013-10-11 16:49:40 <fabled> i suggest you get a monitor then ;)
2013-10-11 16:50:34 <fabled> you could try adding modules=ext2 to cmdline.txt
2013-10-11 16:51:10 <Niichan> Let's try that
2013-10-11 16:51:23 <grimeton> is there a package with gnu's date available somewhere?
2013-10-11 16:52:00 <fabled> grimeton, coreutils
2013-10-11 16:52:27 <grimeton> thanks
2013-10-11 16:54:31 <Niichan> fabled: Once I habe this down to a science, I'm going to be making another copy of this image on a 512 MB sd card then gzip it up for distribution :D
2013-10-11 16:54:56 <grimeton> Niichan: what's your problem?
2013-10-11 16:55:27 <Niichan> FUCK YEAH
2013-10-11 16:55:29 <Niichan> \o/
2013-10-11 16:55:31 <shafire> http://s7.directupload.net/images/131011/x8bgnlav.png
2013-10-11 16:56:13 <fabled> shafire, if you run from tmpfs, you are not supposed to install the kernel .apks
2013-10-11 16:56:34 <fabled> the kernel will be then updated using setup-bootable script from .iso image
2013-10-11 16:56:43 <shafire> oh
2013-10-11 16:56:49 <Niichan> fabled: I got it wotking
2013-10-11 16:56:54 <Niichan> fabled: Thank you so much
2013-10-11 16:57:15 <Niichan> fabled: I'm making the image for a dd write now
2013-10-11 16:57:19 <shafire> Then I can not test flashcache this way...
2013-10-11 16:58:17 <fabled> shafire, not easily. i'd need to build your own modloop with the kernel modules
2013-10-11 16:58:27 <fabled> err... you'd need to build*
2013-10-11 16:58:42 <shafire> mh, I will test it on virtualbox
2013-10-11 16:58:56 <shafire> and I hope, ncopa will push it as fast as he can into main :)
2013-10-11 16:59:16 <Niichan> fabled: You are my hero today
2013-10-11 16:59:29 <shafire> fabled is a hero since days!
2013-10-11 16:59:48 <shafire> he is doing musl apks and the arm support
2013-10-11 16:59:53 <shafire> :-))
2013-10-11 17:02:23 <shafire> ok see you
2013-10-11 17:03:35 <grimeton> argh
2013-10-11 17:03:39 <grimeton> the xfs error isn't fixed
2013-10-11 17:40:51 <Niichan> fabled: I'll have to say, doing this headless is an interesting challenge
2013-10-12 08:55:47 <shafire> hi
2013-10-12 09:12:52 <shafire> what does the "tsc fast tsc calibration failed" error message mean on boot up?
2013-10-12 12:09:49 <shafire> Is bugs.a.o offline/slow?
2013-10-12 12:10:32 <fabled> shafire, works here
2013-10-12 12:11:24 <shafire> It's very^2 slow here
2013-10-12 12:11:38 <shafire> I will get my rpi on Monday!
2013-10-12 12:13:34 <fabled> weird. works fast here. seems rpi's are getting popular on this channel... ;)
2013-10-12 12:16:20 <shafire> for sure, there popular :)
2013-10-12 12:16:26 <shafire> there're
2013-10-12 17:55:28 <shafire> do we use __thread or pthread
2013-10-12 17:55:42 <shafire> ?
2013-10-13 09:16:23 <shafire> hi
2013-10-13 13:44:34 <shafire> hi
2013-10-13 13:44:41 <shafire> are there any plans for openjdk7?
2013-10-13 13:44:58 <shafire> I like to use Scala on my rpi
2013-10-13 13:55:28 <fabled> shafire, you're welcome to build it ;) ... i guess it'll come at some point.
2013-10-13 13:56:17 <shafire> I will try it
2013-10-13 13:56:42 <shafire> but the APKBUILD looks tough
2013-10-13 13:57:45 <fabled> java is a complicated beast
2013-10-13 13:58:48 <shafire> heh, now I want to build mono and openjdk7 :)
2013-10-13 13:59:34 <shafire> But I have an exam in the next days, I will do it next month
2013-10-13 15:42:40 <shafire> does the apknewbuild create a old template?
2013-10-13 15:43:24 <shafire> all your aports uses makedepends
2013-10-13 16:02:39 <kaniini> no
2013-10-13 16:19:07 <shafire> I cant find any APKBUILD with the variable dependes_dev
2013-10-13 16:19:28 <shafire> whats the difference between dependes_dev and makedepends
2013-10-13 16:19:37 <shafire> depends_dev*
2013-10-13 17:22:33 <fcolista_> depends_dev is used for building $pkgname-dev
2013-10-13 17:23:00 <fcolista_> makedepends are packages needed to compile the package
2013-10-13 17:31:18 <shafire> thanks
2013-10-13 17:37:20 <fcolista_> yw
2013-10-14 06:50:40 <vkrishn> ncopa: can downgrade v2.6.5 to previous v2.6.x ?
2013-10-14 06:54:43 <vkrishn> I problem is how to get older apk's from repos/mirror
2013-10-14 07:03:12 <mhavela> Is there older apk's in the mirrors? (I thought it only contained the newest)
2013-10-14 07:03:42 <vkrishn> will there be any issues to have http://nl.alpinelinux.org/alpine/v2.6/updates/v2.6.6
2013-10-14 07:03:51 <vkrishn> ?
2013-10-14 07:03:59 <vkrishn> mhavela, yes true
2013-10-14 07:04:02 <vkrishn> only latest
2013-10-14 07:04:29 <barthalion> what issue do you have?
2013-10-14 07:04:42 <barthalion> downgrading isn't a solution
2013-10-14 07:05:34 <vkrishn> but having pkgs set related to one version release would be nice
2013-10-14 07:06:24 <barthalion> the main release is 2.6 anyway
2013-10-14 07:06:26 <vkrishn> sometimes newer version could have bug but then admins does an upgrade and now have to wait for next version release
2013-10-14 07:06:35 <barthalion> .x part is for security and bug fixes
2013-10-14 07:06:55 <barthalion> then you can report it here or on bugtracker
2013-10-14 07:07:14 <barthalion> we make exceptions and backport commits or new releases if it's very important
2013-10-14 08:04:16 <vkrishn> did a successful install of awesome in v2.6.5 (qemu)
2013-10-14 08:04:25 <vkrishn> :)
2013-10-14 08:06:03 <vkrishn> Issue - Split the pane into two in tmux aterm , show text character instead on line
2013-10-14 08:06:08 <vkrishn> but works ok in console
2013-10-14 08:06:33 <vkrishn> shows text *
2013-10-14 08:08:01 <vkrishn> does it require some more pkg to be installed under awesome ?
2013-10-14 08:08:18 <vkrishn> same issue with rxvt/mrxvt
2013-10-14 08:20:38 <vkrishn> added to bugs.a.o
2013-10-14 08:42:20 <fabled> what's the deal with ruby-redmine-* and ruby-* duplication?
2013-10-14 08:42:28 <fabled> cannot redmine just use the "regular" ones?
2013-10-14 09:31:38 <ncopa> hey
2013-10-14 09:31:49 <ncopa> redmine/ruby is nasty
2013-10-14 09:31:56 <ncopa> well
2013-10-14 09:32:11 <ncopa> ruby gem maintainers are not so smart at release engineering
2013-10-14 09:32:29 <ncopa> so you have all kinds of version dependency conflicts
2013-10-14 09:33:10 <ncopa> redmine needs gem X with version not less than a.b.y and not more than a.b.z
2013-10-14 09:33:41 <ncopa> fabled: btw xulrunner no longer builds on uclibc
2013-10-14 09:33:55 <ncopa> i think your musl fixes broke uclibc
2013-10-14 10:06:34 <fabled> oh
2013-10-14 10:06:38 <fabled> hum
2013-10-14 10:06:45 <fabled> i might have broken ruby on i386 too
2013-10-14 10:06:49 <fabled> with the arm fix
2013-10-14 10:08:51 <fabled> i'll take a look at xulrunner
2013-10-14 10:22:11 <clandmeter> fabled: dont blame me, blame redmine :)
2013-10-14 10:22:56 <clandmeter> this was the only way i could make it "kind of" work.
2013-10-14 10:23:32 <clandmeter> and i only did it because we use it as our tracker.
2013-10-14 15:22:13 <shafire> hi
2013-10-14 15:23:13 <shafire> I got my rpi :)
2013-10-14 15:24:28 <shafire> Have someone started already an openjdk7 build?
2013-10-15 08:26:52 <shafire> hi
2013-10-15 09:42:00 <ncopa> fabled: m4 does not build due to the posix_spawn support
2013-10-15 09:42:39 <ncopa> https://dpaste.de/NrTk
2013-10-15 10:06:18 <fabled> huh
2013-10-15 10:06:37 <fabled> sounds like gnulib not detecting in properly
2013-10-15 10:22:54 <fabled> barthalion, i wonder why libdrm checksums mismatch for my download
2013-10-15 12:54:29 <fabled> i wonder if i should package the rpi userland stuff
2013-10-15 13:39:44 <StarWarsFan> fabled: what do you mean?
2013-10-15 13:46:40 <fabled> StarWarsFan, rpi has set of libraries that it needs for the videocore stuff to work
2013-10-15 13:46:45 <fabled> we don't have those packaged yet
2013-10-15 13:54:21 <vkrishn> will it be difficult to add/support 'mysql-cluster' ?
2013-10-15 13:58:28 <fabled> muahhaa. rpi-userland built. now that just wrap it in proper package.
2013-10-15 14:01:48 <fabled> ncopa, though, rpi-userland will also install libGLESv2.so etc, that come also in mesa-gles
2013-10-15 14:08:40 <fabled> need to figure out how to do it properly
2013-10-15 14:17:03 <StarWarsFan> i c
2013-10-15 14:17:03 <StarWarsFan> thx
2013-10-15 14:47:08 <ncopa> ugh
2013-10-15 14:47:18 <ncopa> lots stuff breaks on posix_Spawn
2013-10-15 14:54:18 <fabled> ncopa, i wonder why
2013-10-15 14:54:23 <fabled> is that patch incomplete?
2013-10-15 14:54:28 <fabled> it's directly from uclibc master
2013-10-15 14:54:42 <fabled> musl implements posix_spawn and stuff builds just fine
2013-10-15 14:57:20 <ncopa> i think theproblem is that -lrt needs for linking
2013-10-15 14:57:27 <ncopa> i just pushed m4
2013-10-15 14:57:38 <ncopa> i added LIBS='-lrt' and it worked
2013-10-15 14:58:27 <ncopa> good thing is that binary size was reduced
2013-10-15 14:58:36 <ncopa> due to not using the gnulib implementation
2013-10-15 15:01:12 <fabled> oh right
2013-10-15 15:01:18 <fabled> yes, spec says -lrt is needed for it
2013-10-15 15:01:23 <fabled> on musl -lrt is no-op
2013-10-15 15:01:27 <fabled> as it's all in libc
2013-10-15 15:01:49 <fabled> ncopa, any comments re rpi-userland packaging?
2013-10-15 15:02:00 <fabled> seems that the almost insist installing to /opt/vc
2013-10-15 15:02:18 <fabled> and some stuff hard core that path (like VLC openmax il plugin)
2013-10-15 15:02:31 <fabled> though, i'd like to put it in /usr
2013-10-15 15:03:09 <fabled> not sure what to do with the libs that "conflict" with mesa
2013-10-15 15:03:21 <fabled> i guess mesa is not wanted on rpi anyway
2013-10-15 15:12:16 <clandmeter> fabled: we cannot put anything in opt afaik right?
2013-10-15 15:12:41 <fabled> we could, but it's kinda hacky and ugly. so i'd rather not.
2013-10-15 15:12:56 <clandmeter> i thought abuild refuses it
2013-10-15 15:14:32 <fabled> mmm. could be. and that's the right thing to do.
2013-10-15 15:14:46 <fabled> it also seems that we might need *Special* builds of things against the rpi-userland libs
2013-10-15 15:14:58 <fabled> vlc should be semi-simple
2013-10-15 15:15:02 <fabled> but qt can be tricky
2013-10-15 15:15:09 <fabled> dunno what people want to use
2013-10-15 15:16:04 <clandmeter> we cannot put it in user and link it to opt?
2013-10-15 15:16:09 <clandmeter> usr
2013-10-15 15:16:44 <fabled> i think about it tm
2013-10-15 15:16:52 <fabled> qt5 should have better rpi support
2013-10-15 15:51:22 <ncopa> fabled: i dont have idea about the rpi stuff
2013-10-15 18:46:23 <Niichan> fabled: Do you know about weechat on arm musl being fucko?
2013-10-15 18:48:36 <Niichan> fabled: http://i.imgur.com/1vL1ynm.png
2013-10-15 18:49:06 <Niichan> fabled: BUT IT SHOWS UNICODE CORRECTLY :D
2013-10-15 18:54:49 <fabled> Niichan, no. i just built it ;)
2013-10-15 18:55:33 <Niichan> fabled: Where should I report it?
2013-10-15 18:55:57 <fabled> Niichan, the build is still highly experimental and bound to have issues.
2013-10-15 18:56:03 <fabled> umm
2013-10-15 18:56:12 <Niichan> fabled: Pretty much I'm on my own?
2013-10-15 18:56:33 <fabled> at this point. yes. you can explain here on irc, and i maybe look at it when i get a moment
2013-10-15 18:57:25 <fabled> i'm still struggling to build rest of aports. it builds now about 80-90%
2013-10-15 19:04:42 <fabled> Niichan, what's the issue? it has some chars misplaced?
2013-10-15 19:04:59 <Niichan> fabled: http://i.imgur.com/1vL1ynm.png
2013-10-15 19:05:14 <Niichan> A lot of things misplaced yes
2013-10-15 19:05:24 <fabled> right. mostly the right side nick listing
2013-10-15 19:05:28 <fabled> is all over the screen
2013-10-15 19:05:57 <fabled> could be arm, or musl issue
2013-10-15 19:06:06 <fabled> will look at it at some point
2013-10-15 19:06:28 <fabled> i'm off for tonight though. will try to get the rpi userland stuff done tomorrow first.
2013-10-15 21:19:27 <kaniini> that looks like a problem with ncurses
2013-10-16 05:09:46 <fabled-musl-arm> Niichan: weechat works here fine
2013-10-16 05:10:12 <fabled-musl-arm> are you running your weechat in screen or tmux?
2013-10-16 05:10:24 <fabled-musl-arm> how are you connected to arm board? which terminal emulator
2013-10-16 05:11:25 <fabled> Niichan, that looks like TERM being wrong, or issue with screen/tmux/terminal emulator
2013-10-16 05:11:53 <fabled> i tested weechat briefly, and it works just fine on my musl-x86 chroot, and also over ssh from my rpi
2013-10-16 05:22:50 <StarWarsFan> Moin
2013-10-16 05:23:13 <StarWarsFan> Any news on openjdk 64?
2013-10-16 05:29:05 <fabled> nope - have been doing musl/arm stuff so far
2013-10-16 05:34:29 <StarWarsFan> which is great too :-)
2013-10-16 05:41:27 <fabled> hum
2013-10-16 05:41:37 <fabled> looks like qt needs build time options for raspberry pi
2013-10-16 05:55:03 <fabled> :/
2013-10-16 05:56:10 <fabled> though, it's likely only qtbase that needs separate build. annoying enough still.
2013-10-16 08:38:15 <shafire> hi
2013-10-16 09:54:12 <shafire> Can someone help me with http://alpinelinux.org/forum/kernel-and-hardware/upgrade-kernel-while-running-tmpfs ?
2013-10-16 09:54:17 <shafire> I know now, that I can not update it
2013-10-16 09:54:40 <shafire> So, how can I later upgrade to 2.7 or 2.6.6 - Can not find it on wiki
2013-10-16 10:08:42 <clandmeter> shafire: there is a script that can do it for you
2013-10-16 10:08:54 <clandmeter> setup-bootable
2013-10-16 10:09:09 <shafire> should I delete the usb stick
2013-10-16 10:09:15 <shafire> then run setup-bootable?
2013-10-16 10:09:19 <clandmeter> no
2013-10-16 10:09:23 <clandmeter> it has an upgrade switch
2013-10-16 10:09:37 <shafire> okay
2013-10-16 10:09:50 <clandmeter> you can run it on live system
2013-10-16 10:10:45 <shafire> thanks
2013-10-16 10:40:34 <fabled> wow
2013-10-16 10:40:36 <fabled> java built
2013-10-16 10:40:54 <fabled> both x86 and x86_64
2013-10-16 10:45:53 <shafire> 6?
2013-10-16 10:46:10 <barthalion> neat
2013-10-16 10:46:28 <shafire> fabled: 6?
2013-10-16 10:56:03 <fabled> yes
2013-10-16 10:56:10 <fabled> i sort of started also on 7
2013-10-16 10:56:27 <fabled> it seems to have built-in support for pax etc.
2013-10-16 10:56:33 <fabled> so it's probably not bad as it sounds
2013-10-16 11:02:28 <fabled> ok. let's see if  java blows up our builders or not.
2013-10-16 11:10:41 <shafire> fabled: I wanted to try the openjdk 7 aport right now lol :)
2013-10-16 11:10:55 <fabled> i started on it
2013-10-16 11:10:55 <shafire> try to create*
2013-10-16 11:11:12 <shafire> I see, then I will wait or try on mono
2013-10-16 11:11:13 <fabled> but it's not close to read
2013-10-16 11:11:17 <fabled> ready*
2013-10-16 11:15:44 <fabled> if openjdk6 build we should probably push it to 2.6 and upgrade openjdk6 to 1.12.x; and have openjdk7 as separate aport
2013-10-16 11:25:37 <shafire> openjdk7
2013-10-16 11:25:41 <shafire> I thought so too
2013-10-16 11:25:49 <shafire> and later openjdk8
2013-10-16 11:32:10 <StarWarsFan> that would be great
2013-10-16 11:33:33 <StarWarsFan> ah, missed the info about successful jdk build!
2013-10-16 11:33:48 <StarWarsFan> that's the best news today :-)
2013-10-16 11:37:59 <fabled> current musl build stats:
2013-10-16 11:38:01 <fabled> musl-x86: 1462 of 1674
2013-10-16 11:38:01 <fabled> musl-arm: 1411 of 1674
2013-10-16 11:38:30 <fabled> i think arm failed on boost due to out-of-memory
2013-10-16 11:38:41 <shafire> I can't wait for 7 :)
2013-10-16 11:39:03 <shafire> when will you buld the musl-x86_64 packages?
2013-10-16 11:39:30 <shafire> build*
2013-10-16 11:43:06 <fabled> mmm
2013-10-16 11:43:17 <fabled> when i get the inspiration ;)
2013-10-16 11:43:29 <fabled> now that we have new builders, i suppose it would not be a big job
2013-10-16 11:43:43 <fabled> i guess when those are setup fully, i can spawn the x86_64 variant too
2013-10-16 11:44:08 <StarWarsFan> i'm general interested, how the alpine buildsystem works
2013-10-16 11:44:18 <StarWarsFan> is there some docu somewhere?
2013-10-16 11:53:23 <StarWarsFan> seems as fabled is writing buildsystem docu now... *g*
2013-10-16 11:53:37 <fabled> *shrug*
2013-10-16 11:53:44 <fabled> 
ACTION has not written any docs.

2013-10-16 11:54:46 <StarWarsFan> especially the part how the build system determines, which package needs to be (re)build
2013-10-16 11:55:03 <StarWarsFan> i'm interested in
2013-10-16 12:29:55 <shafire> ncopa: do not forget flashcache :)
2013-10-16 12:30:02 <shafire> and please take a look at gcc-gnat
2013-10-16 12:30:56 <ncopa> gcc-gnats should work?
2013-10-16 12:31:07 <ncopa> i simply dont have time to test that everything works
2013-10-16 12:31:15 <ncopa> flashcache is there
2013-10-16 12:31:22 <ncopa> i update it every single time
2013-10-16 12:31:34 <ncopa> but i dont have time to test things for you
2013-10-16 12:31:47 <ncopa> please test it for me and report if it works or note
2013-10-16 12:31:50 <ncopa> not*
2013-10-16 12:33:35 <ncopa> apk add flashcache-grsec
2013-10-16 12:33:36 <ncopa> configiure
2013-10-16 12:33:39 <ncopa> test that it works
2013-10-16 12:33:43 <ncopa> report back
2013-10-16 12:34:39 <ncopa> kernel module loads atleast
2013-10-16 12:34:50 <ncopa> [19893.464844] flashcache: flashcache-2.0 initialized
2013-10-16 12:35:08 <ncopa> i am in doubt how useful flashcache is
2013-10-16 12:36:12 <shafire> I tested it and reported back days ago?
2013-10-16 12:36:21 <ncopa> oh sorry... :)
2013-10-16 12:37:13 <shafire> I think, gcc-gnat is missing a dependency
2013-10-16 12:41:34 <fabled> shafire, which one?
2013-10-16 12:41:49 <fabled> i might have messed it up for the crosscompile stuff
2013-10-16 12:42:03 <fabled> but hey, we have gcc-gnat for musl-x86 and musl-arm too now ;)
2013-10-16 12:42:07 <shafire> :-)
2013-10-16 12:42:11 <shafire> http://bugs.alpinelinux.org/attachments/download/135/gcc-gnat-test.log <- here is a log
2013-10-16 12:42:41 <shafire> one or more apks from build-base, but I don't know which at moment
2013-10-16 12:43:13 <ncopa> libc-dev likely
2013-10-16 12:43:55 <fabled> yes, it's libc-dev
2013-10-16 12:44:38 <fabled> same goes for gcc
2013-10-16 12:44:43 <fabled> but we want to avoid circular deps
2013-10-16 12:46:16 <ncopa> i dont think you need libc-dev for gcc...
2013-10-16 12:46:37 <ncopa> you can build libc itself without libc-dev
2013-10-16 12:47:04 <ncopa> and i dont think you need libc headers for compile varnish config
2013-10-16 12:47:12 <ncopa> but you need gcc
2013-10-16 12:47:47 <ncopa> algitbot: build master
2013-10-16 13:17:12 <ncopa> algitbot: build master
2013-10-16 14:05:49 <fabled> ok. raspberrypi userland crap is pushed to testing
2013-10-16 14:06:07 <fabled> it's already built and pushed out to edge-musl/armhf testing
2013-10-16 14:06:16 <fabled> it install to /opt/vc
2013-10-16 14:06:20 <fabled> which is awkward
2013-10-16 14:06:26 <fabled> but at least the libs and basic utils work
2013-10-16 14:26:51 <fabled> cool
2013-10-16 14:26:56 <fabled> openjdk6 built for x86_64
2013-10-16 14:36:40 <shafire> :-)
2013-10-16 14:37:29 <ncopa> thanks fabled
2013-10-16 14:37:56 <fabled> i also started openjdk7 test build on the new build box
2013-10-16 14:38:33 <fabled> i got apkbuild to the point that it can start building. so let's see where it breaks.
2013-10-16 14:38:44 <fabled> and what packaging changes are needed
2013-10-16 14:39:02 <fabled> i think openjdk6 commit should go to 2.6-stable
2013-10-16 14:39:16 <fabled> after that we could bump openjdk6/edge to the next buildlevel
2013-10-16 14:43:24 <fabled> barthalion, did you even test that they build before pushing?
2013-10-16 14:43:32 <fabled> seems that git patches no longer apply
2013-10-16 14:43:58 <barthalion> I'm not blind even if it seems so
2013-10-16 14:48:52 <shafire> is the access to the buildlogs restricted?
2013-10-16 14:54:09 <fabled> shafire, no, but the new 2-7 builder logs are not synced out yet
2013-10-16 14:54:15 <fabled> the others should work
2013-10-16 14:55:05 <barthalion> fabled: revert the guilty commits, I've no access to buildlogs
2013-10-16 14:55:27 <fabled> the builders should post link to it
2013-10-16 14:55:37 <barthalion> 404
2013-10-16 14:57:15 <ncopa> buildlogs for 2-7 is not yet available
2013-10-16 14:57:18 <ncopa> its a new box
2013-10-16 14:57:20 <ncopa> new hw
2013-10-16 14:57:58 <ncopa> http://dev.alpinelinux.org/buildlogs/build64-edge.1381934461.log
2013-10-16 14:59:13 <shafire> the e5 xeon, or? :-)
2013-10-16 14:59:54 <shafire> ah, are both building the same?
2013-10-16 15:00:32 <fabled> yes
2013-10-16 15:12:40 <barthalion> I've too slow connection here, will push it later
2013-10-16 16:03:12 <Niichan> fabled: I am using tmux and terminator as my terminal
2013-10-16 16:04:22 <Niichan> Removing tmux fixes it
2013-10-16 17:28:41 <fabled> Niichan, ok. yes, i got the same behaviour.
2013-10-16 17:28:52 <fabled> dunno if it's tmux, or some library it uses, or just environment
2013-10-16 17:29:07 <fabled> bah. openjdk7 failed due to make bug. need to fix make.
2013-10-16 17:36:41 <fabled> mmm
2013-10-16 17:39:48 <fabled> i guess we should upgrade to make 4.0
2013-10-16 18:06:50 <fabled> Niichan, what are TERM, LANG and CHARSET when starting tmux?
2013-10-16 18:07:44 <Niichan> fabled: https://gist.github.com/Niichan/727fd631078602a4e7da
2013-10-16 18:08:13 <fabled> try setting TERM=xterm or something else like that
2013-10-16 18:09:20 <Niichan> Replaces one display problem with another
2013-10-16 18:09:23 <Niichan> http://i.imgur.com/JD1nO7m.png
2013-10-16 18:11:40 <fabled> so you have screen -> tmux -> weechat ?
2013-10-16 18:11:50 <Niichan> No
2013-10-16 18:11:54 <Niichan> Just tmux->weechat
2013-10-16 18:12:05 <fabled> and tmux is on local computer which runs ?
2013-10-16 18:12:31 <Niichan> everything tmux and weechat is on the pi
2013-10-16 18:13:20 <fabled> how do you connect to Pi?
2013-10-16 18:13:26 <fabled> terminal -> ssh ?
2013-10-16 18:14:07 <Niichan> yes
2013-10-16 18:19:10 <fabled> Niichan, you might need to set TERM before sshing
2013-10-16 18:20:37 <Niichan> rainbowdash cod (git:master) > echo $TERM
2013-10-16 18:20:39 <Niichan> xterm
2013-10-16 18:20:57 <Niichan> rainbowdash cod (git:master) > ssh 192.168.1.126
2013-10-16 18:20:58 <Niichan> Welcome to Alpine!
2013-10-16 18:21:02 <Niichan> pinkie ~ > echo $TERM
2013-10-16 18:21:04 <Niichan> xterm
2013-10-16 18:21:21 <Niichan> pinkie ~ > tmux
2013-10-16 18:21:23 <Niichan> pinkie ~ > echo $TERM
2013-10-16 18:21:25 <Niichan> screen
2013-10-16 18:23:35 <fabled> oh
2013-10-16 18:24:09 <fabled> sounds like tmux / depenedant library issue. possibly affecting uclibc builds too
2013-10-16 18:24:24 <Niichan> Except it's not
2013-10-16 18:24:28 <barthalion> argh, checksums
2013-10-16 18:24:34 <Niichan> (affecting the uclibc builds)
2013-10-16 18:24:49 <Niichan> I use tmux and weechat on uclibc alpine and it works fine
2013-10-16 18:25:00 <fabled> ok
2013-10-16 18:25:20 <fabled> i also noticed that i get in some screen/musl combinations certain utf8 chars broken
2013-10-16 18:25:30 <barthalion> like clandmeter
2013-10-16 18:25:35 <barthalion> I've had issues with irssi as well
2013-10-17 05:44:57 <fabled> urgh
2013-10-17 05:45:05 <fabled> oh the "smartness" of openjdk7
2013-10-17 05:45:30 <fabled> the dlopen trickery does is beyond abominable
2013-10-17 05:49:06 <StarWarsFan> fabled: just to let you know: around 120 packages updated yesterday as finaly the openjdk update on my machines work :-)
2013-10-17 05:49:09 <StarWarsFan> thx a lot!
2013-10-17 05:49:20 <fabled> :)
2013-10-17 05:49:49 <fabled> yes, to cherry-pick packages from edge to 2.6 install needs a lot of dependencies
2013-10-17 05:50:04 <fabled> that update should go to 2.6-stable; and we should bump edge openjdk6 to the new build of jdk
2013-10-17 05:50:16 <StarWarsFan> ack
2013-10-17 05:54:18 <fabled> i really hate the "trickery" openjdk7 does
2013-10-17 05:54:22 <fabled> i basically have stage1 built
2013-10-17 05:54:34 <fabled> but the resulting java and javac won't run due to dlopen() trickery
2013-10-17 05:59:51 <StarWarsFan> hm...
2013-10-17 06:15:47 <fabled> ok
2013-10-17 06:15:58 <fabled> i think i got the idea on the atrocities the loader does
2013-10-17 06:16:25 <fabled> it basically on startup parses few args, mangles LD_LIBRARY_PATH (or not) and re-exec's itself
2013-10-17 06:23:09 <fabled> OH
2013-10-17 06:23:13 <fabled> i think i know what is broke
2013-10-17 06:23:16 <fabled> it's uclibc ldso
2013-10-17 06:23:29 <fabled> i'm pretty sure it'd work on musl
2013-10-17 06:23:39 <fabled> hum
2013-10-17 06:26:11 <fabled> ncopa, ^^^
2013-10-17 06:26:17 <fabled> i wonder if we forget openjdk7 with uclibc
2013-10-17 06:28:19 <ncopa> fabled: i'm totally ok with that
2013-10-17 06:28:31 <ncopa> lets do openjdk7 for musl only
2013-10-17 06:28:55 <ncopa> if we do it for v2.7 we will have to support it for 2 years
2013-10-17 06:30:55 <freedomrun> can someone show me the actual path of 2.7 testing repo, please?
2013-10-17 06:32:11 <fabled> ncopa, care to create alpine 3.0 target in bugs.a.o ?
2013-10-17 06:32:56 <fabled> or i create one?
2013-10-17 06:33:12 <fabled> mmm
2013-10-17 06:33:15 <fabled> maybe you create
2013-10-17 06:35:03 <fabled> ncopa, btw. i also made some tests with make 4.0, i think could likely upgrade to it
2013-10-17 06:39:11 <ncopa> fabled: ok if i call it 2.8 for now?
2013-10-17 06:39:36 <ncopa> i think it is possible to rename a release afterwards
2013-10-17 06:39:48 <fabled> mmm
2013-10-17 06:39:59 <ncopa> i believe it will be 3.0
2013-10-17 06:40:09 <fabled> i guess it's ok
2013-10-17 06:40:18 <fabled> it's just a name
2013-10-17 06:41:36 <fabled> did glib upgrade break sonames?
2013-10-17 06:42:04 <ncopa> fabled: i think not
2013-10-17 06:42:10 <fabled> mmm
2013-10-17 06:42:16 <fabled> maybe it's just my arm builder confused then
2013-10-17 06:42:17 <ncopa> i didnt test though
2013-10-17 06:42:24 <ncopa> might be it did
2013-10-17 06:42:33 <ncopa> they are normally abi compat
2013-10-17 06:43:30 <fabled> adding swap to wandboard made boost compile
2013-10-17 06:43:38 <fabled> so 2GB is not enough of RAM to compile it
2013-10-17 06:44:08 <ncopa> heh
2013-10-17 06:45:38 <fabled> i guess it's due to running 4 parallel g++ instances each compiling something huge
2013-10-17 07:05:52 <fabled> ncopa, oh, i think glib broke because gnulib update now adds charset.alias which didn't happen earlier
2013-10-17 07:06:11 <ncopa> bah
2013-10-17 07:06:52 <ncopa> should we make abuild barf if it finds charset.alias?
2013-10-17 07:06:57 <fabled> it does
2013-10-17 07:07:01 <fabled> but that happens only with musl
2013-10-17 07:07:07 <fabled> so on uclibc everything is perfect
2013-10-17 07:07:11 <ncopa> ok
2013-10-17 07:07:17 <fabled> gnulib is stupid and detects GLIBC
2013-10-17 07:07:21 <fabled> or uclibc
2013-10-17 07:07:28 <fabled> for those, it inhibits charset.alias
2013-10-17 07:08:01 <fabled> and yes, it really looks for __GLIBC__ and __UCLIBC__ #define
2013-10-17 07:09:30 <fabled> you fixed build-2-7 urls
2013-10-17 07:09:31 <fabled> nice
2013-10-17 07:09:44 <ncopa> yup
2013-10-17 07:10:01 <ncopa> i have one critical thing left to do for bld1 and bld2
2013-10-17 07:10:14 <ncopa> change the public ip device to a bridge
2013-10-17 07:10:37 <ncopa> so ip moves from bond0 -> br0
2013-10-17 07:11:07 <fabled> ok
2013-10-17 07:11:24 <fabled> can you move build-edge-musl-x86 to either of them?
2013-10-17 07:11:25 <ncopa> i still wonder if i should bother setting up openvswitch 'tunnel'
2013-10-17 07:11:31 <ncopa> sure
2013-10-17 07:12:35 <fabled> the new box is quite nice. uclibc builds in about 10 seconds
2013-10-17 07:12:54 <ncopa> they are awsome
2013-10-17 07:13:08 <ncopa> so
2013-10-17 07:13:17 <ncopa> i think i shoul join the 2 private nets
2013-10-17 07:13:19 <fabled> do you think i should do bootstrap support to openjdk?
2013-10-17 07:13:38 <fabled> so it recompiles it withitself unless bootstrap is requested
2013-10-17 07:13:50 <ncopa> might be an idea
2013-10-17 07:13:54 <fabled> would speed up compile times quite a bit
2013-10-17 07:13:59 <fabled> since gcj is single threaded
2013-10-17 07:14:13 <ncopa> yeah
2013-10-17 07:14:16 <ncopa> i think that would be nice
2013-10-17 07:14:35 <ncopa> but if it works as is, then i'm ok to do it after v2.7 release
2013-10-17 07:14:36 <ncopa> hmm
2013-10-17 07:14:53 <ncopa> otoh, when we do stable sec updates...
2013-10-17 07:16:31 <fabled> ncopa, i think what is in openjdk6 now should go to 2.6-stable
2013-10-17 07:16:40 <ncopa> ok
2013-10-17 07:16:49 <fabled> i can still upgrade it to newer upstream build, and add the bootstrap support for master/2.7
2013-10-17 07:17:45 <ncopa> ok, so i sync openjdk6 > 2.6-stable
2013-10-17 07:19:32 <fabled> ok i moved #2224 (openjdk7) to 2.8/3.0 target.
2013-10-17 07:19:47 <fabled> if someone wants to fix uclibc dlopen() we might add it early
2013-10-17 07:20:05 <fabled> but be warned: you'll need to fix uclibc internals, and it aint pretty
2013-10-17 07:20:12 <ncopa> :)
2013-10-17 07:20:34 <ncopa> i think its better to spend resources on musl...
2013-10-17 07:20:57 <ncopa> like sending patches for gnulib upstream
2013-10-17 07:20:58 <ncopa> etc
2013-10-17 07:33:13 <fabled-rpi> Niichan: Terminal -> ssh (to rpi) -> tmux -> weechat
2013-10-17 07:33:17 <fabled-rpi> and all is great
2013-10-17 07:33:36 <fabled-rpi> i wonder if it's your terminal emulator that is broken
2013-10-17 08:04:12 <shafire> hi
2013-10-17 08:28:48 <barthalion> fiy gcc 4.8.2 is out
2013-10-17 08:34:10 <fabled> finally!
2013-10-17 08:36:11 <StarWarsFan> is there a problem with nfs/portmap on edge?
2013-10-17 08:36:19 <StarWarsFan> got the same setup with edge and 2.6
2013-10-17 08:36:24 <StarWarsFan> but on edge it gives:
2013-10-17 08:36:35 <StarWarsFan>  rc-service nfsmount restart
2013-10-17 08:36:35 <StarWarsFan>  * Starting NFS statd ...
2013-10-17 08:36:35 <StarWarsFan>  * start-stop-daemon: failed to start `/usr/sbin/rpc.statd'                                                  [ !! ]
2013-10-17 08:36:35 <StarWarsFan>  * ERROR: rpc.statd failed to start
2013-10-17 08:36:35 <StarWarsFan>  * ERROR: cannot start nfsmount as rpc.statd would not start
2013-10-17 09:55:11 <fabled> ok. openjdk6/icedtea 1.12.x built in the disable bootstrap mode. i'll test bootstrap mode once, and push to edge if everything goes fine.
2013-10-17 10:11:05 <fabled> ncopa, you apparently nuked poppler-gtk in favor of poppler-glib builtinto main poppler
2013-10-17 10:11:06 <fabled> yet
2013-10-17 10:11:08 <fabled> $ grep -l poppler-gtk-dev */APKBUILD
2013-10-17 10:11:08 <fabled> claws-mail/APKBUILD
2013-10-17 10:11:08 <fabled> gimp/APKBUILD
2013-10-17 10:11:08 <fabled> evince/APKBUILD
2013-10-17 10:11:10 <fabled> tumbler/APKBUILD
2013-10-17 10:11:15 <fabled> those won't build anymore
2013-10-17 10:11:37 <fabled> it's just changing the dependency to poppler-dev ?
2013-10-17 10:52:29 <ncopa> fabled: hmm i thought i had replaced them all...
2013-10-17 10:52:36 <fabled> ncopa, re: openjdk
2013-10-17 10:52:38 <fabled> - md5check || return 1
2013-10-17 10:52:38 <fabled> + verify || return 1
2013-10-17 10:52:44 <fabled> does that work for 2.6-stable ?
2013-10-17 10:52:54 <ncopa> probably
2013-10-17 10:52:57 <fabled> ok
2013-10-17 10:53:09 <fabled> i have new openjdk for edge soon
2013-10-17 10:53:18 <fabled> upgraded to icedtea 1.12.x
2013-10-17 10:53:26 <ncopa> yeah 'verify' should work
2013-10-17 10:53:33 <fabled> it has --with-pax support natively (which needed patching a bit to make it actually work)
2013-10-17 10:53:41 <ncopa> lol
2013-10-17 10:53:44 <fabled> but a lot better than the current hacks
2013-10-17 10:53:48 <ncopa> good
2013-10-17 10:54:05 <ncopa> i'm still in doubt what to do with lua modules and 5.1/5.2
2013-10-17 10:55:38 <ncopa> we have 44 lua-* aports
2013-10-17 10:55:43 <ncopa> 7 of them are noarch
2013-10-17 10:56:01 <ncopa> i suppose the noarch's can be symlinked 5.1 -> 5.2
2013-10-17 10:57:41 <ncopa> so, i think it all buils down to whats easiest to maintain
2013-10-17 10:59:10 <fabled> yes
2013-10-17 10:59:12 <fabled> also remember
2013-10-17 10:59:19 <fabled> same stuff will happen with ugprade 5.2 -> 5.3
2013-10-17 10:59:31 <fabled> so i'd try to avoid version number in the pkgname if possible
2013-10-17 10:59:44 <fabled> because it creates lot of dependency problems for the packages using these modules
2013-10-17 11:00:05 <ncopa> i'd say the opposite
2013-10-17 11:00:09 <ncopa> well
2013-10-17 11:00:19 <ncopa> we could ship both as in debian
2013-10-17 11:00:37 <fabled> you'd need to change all dependencies depending on lua version, instead of just one lua5.2 vs lua5.3 dependency
2013-10-17 11:00:59 <ncopa> right
2013-10-17 11:02:20 <vkrishn> if mysql-cluster-gpl-7.3.2 can replace the current mysql pkg, it has added nosql/memcache features
2013-10-17 11:03:20 <ncopa> but how do you make sure you satisfy all deps for an app?
2013-10-17 11:03:24 <ncopa> lets say acf-*
2013-10-17 11:03:34 <ncopa> will only work with lua5.1 for now
2013-10-17 11:03:59 <ncopa> so how can i keep track of what packages can now drop 5.1 support?
2013-10-17 11:04:21 <ncopa> the acf-* apps actually needs lua5.1-<module>
2013-10-17 11:04:21 <barthalion> vkrishn: I rather opt for mariadb
2013-10-17 11:04:53 <ncopa> once it all migrated we can say, ok, now we switch to 5.2
2013-10-17 11:05:04 <ncopa> and then it should only pull in 5.2 modules
2013-10-17 11:05:15 <ncopa> and at that point we could likely drop support for various 5.1
2013-10-17 11:05:19 <vkrishn> barthalion, nosql features and newer performance benchmark as ncopa pointed has + point
2013-10-17 11:06:00 <barthalion> vkrishn: mariadb is more efficient than mysql anyway, and let's use nosql databases for nosql, not frankensteins
2013-10-17 11:06:02 <vkrishn> maybe mariadb in 2.9
2013-10-17 11:07:10 <vkrishn> mariadb has cluster features ?
2013-10-17 11:07:24 <barthalion> there is mariadb-galera
2013-10-17 11:09:47 <vkrishn> ok, so whichever gets decided I would vote for WITH cluster
2013-10-17 11:13:52 <barthalion> vkrishn: mysql cluster is a separate project or replacement of mysql?
2013-10-17 11:14:51 <vkrishn> I checked the source code, seems to include everything in mysql + cluster code
2013-10-17 11:15:08 <vkrishn> but haven't tried compiling
2013-10-17 11:16:08 <barthalion> I think that it's possible to switch anything as long as it doesn't require manual intervention
2013-10-17 11:16:42 <vkrishn> if mysql-cluster-gpl-7.3.2 can be done, would be nice
2013-10-17 11:17:26 <ncopa> hehe... why would anyone want nosql on a my*SQL* setup :)
2013-10-17 11:18:00 <ncopa> from what i could see, mysql-5.6 performs better than mariadb
2013-10-17 11:18:18 <ncopa> specially if there are many cores
2013-10-17 11:18:43 <vkrishn> I have not checked if mysql-cluster-gpl-7.3.2 is  mysql-5.6+cluster or  mysql-5.5+cluster
2013-10-17 11:19:42 <ncopa> barthalion: do you know if it still easy to switch to mariadb from mysql-5.6?
2013-10-17 11:20:18 <ncopa> i think i'd like to upgrade mysql to 5.6 for v.27
2013-10-17 11:20:21 <ncopa> v2.7*
2013-10-17 11:20:36 <ncopa> but i dont want shut the door to mariadb yet
2013-10-17 11:23:16 <vkrishn> mariadb(current 5.5.33a) has version release cycle similar to mysql (current 5.5.34 GA)
2013-10-17 11:23:27 <ncopa> thats why i ask about 5.6
2013-10-17 11:25:52 <barthalion> ncopa: it won't be easy done
2013-10-17 11:26:04 <ncopa> aw.. ok
2013-10-17 11:26:12 <ncopa> so its now or never...
2013-10-17 11:26:28 <ncopa> i mean before we do mysql-5.6
2013-10-17 11:26:38 <barthalion> ncopa: theoretically it should be doable, but they have different features
2013-10-17 11:26:58 <barthalion> ncopa: and you can't be sure that mysql_check and mysql_upgrade will detect and fix everything
2013-10-17 11:27:24 <ncopa> but dump/restore should work?
2013-10-17 11:27:25 <barthalion> imho 5.5 is fine, whether maria or mysql
2013-10-17 11:27:39 <ncopa> ok
2013-10-17 11:27:42 <barthalion> ncopa: as long as you don't use some weird engines like aria
2013-10-17 11:27:50 <ncopa> ok
2013-10-17 11:27:55 <barthalion> or whatever it's called in mysql 5.6
2013-10-17 11:27:58 <ncopa> i think we wait a bit more then
2013-10-17 11:27:59 <barthalion> (aria is in mariadb 10)
2013-10-17 11:28:04 <ncopa> yeah
2013-10-17 11:29:26 <ncopa> tumpler...
2013-10-17 11:29:28 <ncopa> bah..
2013-10-17 11:30:28 <vkrishn> ncopa, is kernel module pohmelfs not much used ?
2013-10-17 11:31:04 <vkrishn> I see some nice projects at http://www.ioremap.net/projects/pohmelfs
2013-10-17 11:31:21 <ncopa> dunno
2013-10-17 11:31:30 <ncopa> i have heard the name before
2013-10-17 11:32:04 <ncopa> but i dont think i want start with new big projects til v2.7 is out
2013-10-17 11:32:14 <ncopa> i'd like to finish up what we have
2013-10-17 11:32:16 <mhavela> Regarding using 'setup-bootable' (running on 2.6).
2013-10-17 11:32:16 <mhavela> Seems I can not run 'setup-bootable /path/to/iso /dev/sdb1'.
2013-10-17 11:32:16 <mhavela> What I need to do first is to 'mount -t vfat /dev/sdb1 /mnt' and now the above command starts working.
2013-10-17 11:32:17 <ncopa> so we can release
2013-10-17 11:32:24 <mhavela> Is this normal behaviour or is it just me?
2013-10-17 11:32:45 <ncopa> i think it is supposed to work
2013-10-17 11:32:55 <vkrishn> enabling the kernel module should be easier
2013-10-17 11:33:07 <vkrishn> other related apps maybe later
2013-10-17 11:33:07 <mhavela> In that case I guess I should file a bug about the setup-bootable
2013-10-17 11:33:08 <ncopa> modprobe vfat
2013-10-17 11:33:17 <mhavela> I can try that
2013-10-17 11:33:19 <ncopa> mhavela: might be enough to just modprobe vfat
2013-10-17 11:33:24 <fabled> ncopa, buildmsg-subscribe is still not fixed on builders. i'd give that priority over all other.
2013-10-17 11:33:31 <ncopa> ok
2013-10-17 11:33:59 <ncopa> fabled: that proj stalled due to i was thinking to push lua5.2 to them while at it
2013-10-17 11:34:13 <ncopa> but lua5.2 is other project
2013-10-17 11:34:45 <mhavela> ncopa: You where right! Need to 'modprobe vfat' in order to have setup-bootable working
2013-10-17 11:34:52 <mhavela> Should I file a bug on that?
2013-10-17 11:36:06 <ncopa> mhavela: i dont think its easily fixable
2013-10-17 11:36:23 <ncopa> hm
2013-10-17 11:36:25 <ncopa> maybe it is
2013-10-17 11:36:30 <ncopa> blkid can tell the fs
2013-10-17 11:36:54 <fabled> i thought kernel would modprobe it
2013-10-17 11:37:07 <fabled> apparently not
2013-10-17 11:37:14 <fabled> or something might be broke in bb modprobe / symmaps
2013-10-17 11:37:16 <ncopa> mount will not autodetect it unles the kernel module is loaded
2013-10-17 11:37:31 <fabled> ok
2013-10-17 11:37:42 <ncopa> has been that way as long i can remember
2013-10-17 11:37:43 <fabled> i though they'd modprobe based on partition table type id
2013-10-17 11:37:51 <ncopa> no
2013-10-17 11:38:09 <ncopa> i dont think they use partitiontable type id to anything
2013-10-17 11:38:33 <ncopa> i dont know
2013-10-17 11:38:48 <ncopa> i dont know if its worth the effort to fix that in setup-bootable
2013-10-17 11:42:04 <ncopa> i should probably try move over the edge builders to new build servers soonish
2013-10-17 11:42:38 <mhavela> I find it a bit annoying nowadays is that I have to do 'mount -t vfat /dev/sdb1 /mnt' instead of just 'mount /dev/sdb1 /mnt' (as it used to be before)
2013-10-17 11:42:49 <mhavela> Its not a big problem though
2013-10-17 11:43:00 <mhavela> but fixing that would probably also fix 'setup-bootable'
2013-10-17 11:43:10 <ncopa> mhavela: thats nothing new
2013-10-17 11:43:15 <ncopa> has always been like that
2013-10-17 11:43:33 <ncopa> if vfat kernel module is not loaded, then you have to add -t vfat
2013-10-17 11:43:44 <ncopa> or modprobe the module
2013-10-17 11:44:02 <mhavela> Hmm.. Yeah  - that might be it... Maybe I had vfat kernel module loaded without thinking about it
2013-10-17 11:44:15 <ncopa> it is normally not a problem on boot from usb, because there vfat is always loaded
2013-10-17 11:44:50 <ncopa> you can also do this: echo vfat >> /etc/modules
2013-10-17 11:44:56 <ncopa> and problem is "solved"
2013-10-17 11:45:15 <mhavela> True (maybe we need to put a note about this in http://wiki.alpinelinux.org/wiki/Create_a_Bootable_USB#Creating_a_bootable_Alpine_Linux_USB_Stick_from_the_command_line
2013-10-17 11:46:18 <ncopa> tell it to mount the dest and use the mounted dest
2013-10-17 11:46:20 <ncopa> eg
2013-10-17 11:46:32 <ncopa> mount -t vfat /dev/blah /media/usb
2013-10-17 11:46:46 <ncopa> setup-bootable /path/to/alpine.iso /media/usb
2013-10-17 11:47:31 <ncopa> or just say: "it might be you need 'modprobe vfat' first"
2013-10-17 11:47:36 <ncopa> or similar
2013-10-17 11:48:04 <ncopa> hm
2013-10-17 11:48:15 <ncopa> i wonder if we should have a separate ACF project in redmine
2013-10-17 11:48:25 <ncopa> a sub project
2013-10-17 11:53:30 <ncopa> wow
2013-10-17 11:53:35 <ncopa> gcc-4.8.2 is out
2013-10-17 11:53:48 <ncopa> fabled, is it it worth it for v2.7?
2013-10-17 11:53:55 <fabled> i think so
2013-10-17 11:54:00 <fabled> it would be nice to have
2013-10-17 11:54:00 <ncopa> ok
2013-10-17 11:54:05 <ncopa> yeah
2013-10-17 11:54:19 <ncopa> we should do that *now* then
2013-10-17 11:55:31 <barthalion> ncopa: I wrote that already, should have hilighted you :p
2013-10-17 11:55:41 <ncopa> i missed it
2013-10-17 11:55:45 <ncopa> i was out for a while
2013-10-17 11:55:47 <ncopa> sorry
2013-10-17 12:03:02 <tdtrask> ncopa: changing ACF to Lua 5.2 style modules is bigger than I expected
2013-10-17 12:03:09 <ncopa> fabled: what do you think about adding the piepatches ot $source
2013-10-17 12:03:18 <ncopa> tdtrask: i'm not surprised...
2013-10-17 12:03:31 <tdtrask> I might drop it for alpine 2.7 so I can focus on features instead
2013-10-17 12:03:40 <ncopa> my experience with lua 5.2 has been:
2013-10-17 12:03:47 <ncopa> "oh loks easy to port"
2013-10-17 12:03:56 <ncopa> "oh a can of worms..."
2013-10-17 12:04:07 <ncopa> "oh it was not as bad as i thought..."
2013-10-17 12:04:22 <ncopa> but it was still not as easy as first though.. :)
2013-10-17 12:04:31 <tdtrask> because of the changes with global environment, lots of issues pop up
2013-10-17 12:04:40 <ncopa> ah
2013-10-17 12:04:41 <ncopa> yeah
2013-10-17 12:05:08 <tdtrask> 
ACTION has had to change lots of code

2013-10-17 12:05:16 <ncopa> global vars?
2013-10-17 12:05:21 <tdtrask> all minor, but hard to find them all
2013-10-17 12:05:27 <ncopa> understand
2013-10-17 12:05:50 <ncopa> yeah we can wait with it for v2.8/v3.0
2013-10-17 12:05:58 <tdtrask> global function declarations
2013-10-17 12:06:43 <tdtrask> more specifically, functions declared in the module environment
2013-10-17 12:15:59 <fabled> Niichan, actually, i just found out that ncurses-dev did not have .pc files in the build. but on the new build it does have them.
2013-10-17 12:16:05 <fabled> i really don't understand what changed
2013-10-17 12:16:10 <fabled> but that might've caused the issues
2013-10-17 12:16:23 <fabled> i need likely to rebuild some packages
2013-10-17 12:16:53 <Niichan> fabled: Ah, let me know once you're done. I'll test that fix ASAP
2013-10-17 12:17:19 <fabled> i noticed it since vlc build failed for not having ncurses even though the packages were installed
2013-10-17 12:20:35 <fabled> huh
2013-10-17 12:20:44 <fabled> java failed
2013-10-17 12:20:47 <fabled> that's weird
2013-10-17 12:21:09 <fabled> let's wait what happens with x86_64
2013-10-17 12:23:24 <ncopa> likely out of mem...
2013-10-17 12:23:42 <fabled> it got segfault
2013-10-17 12:23:47 <fabled> in some random phase of the build
2013-10-17 12:23:53 <fabled> out-of-mem is possible
2013-10-17 12:24:05 <fabled> considering it's building java on four different build vservers ;)
2013-10-17 12:24:16 <ncopa> hehe
2013-10-17 12:24:49 <ncopa> should move some of the build vservers to a box with 48G ram..
2013-10-17 12:25:03 <ncopa> fabled: gcc-4.8-musl needs to be rebased
2013-10-17 12:25:09 <fabled> mmm
2013-10-17 12:25:32 <fabled> others too?
2013-10-17 12:25:34 <fabled> or just that patch?
2013-10-17 12:25:36 <ncopa> it was only that and a 33_all_gcc48_config_rs6000.patch that failed
2013-10-17 12:25:47 <ncopa> i think we dont use rs6000
2013-10-17 12:25:49 <fabled> oh yes, i remember that
2013-10-17 12:25:50 <ncopa> so i think i just nuke it
2013-10-17 12:25:53 <fabled> true
2013-10-17 12:25:57 <fabled> but it's part of the piepatches
2013-10-17 12:26:08 <fabled> we should really nuke the piepatch tarball, and carry them as in-tree patches
2013-10-17 12:26:08 <ncopa> yeah
2013-10-17 12:26:15 <ncopa> thats what i'm doing
2013-10-17 12:26:18 <fabled> great
2013-10-17 12:26:27 <fabled> i'll rebase musl patch meanwhile then..
2013-10-17 12:26:39 <ncopa> there are mips, ia64 too
2013-10-17 12:26:43 <ncopa> can maybe nuke those too
2013-10-17 12:26:44 <ncopa> ?
2013-10-17 12:26:49 <fabled> i'd be ok with that
2013-10-17 12:26:55 <fabled> we need only x86* and arm
2013-10-17 12:27:04 <ncopa> some are changes in texi
2013-10-17 12:27:07 <ncopa> the documentation
2013-10-17 12:27:11 <ncopa> dunno if we bother that
2013-10-17 12:29:53 <ncopa> is the -fno-stack-check needed? http://git.alpinelinux.org/cgit/aports/tree/main/gcc/APKBUILD#n208
2013-10-17 12:30:37 <fabled> ncopa, yes, especially if you take the newest piepatch
2013-10-17 12:30:56 <ncopa> fabled: its rs6000 that fails for musl
2013-10-17 12:31:03 <fabled> oh
2013-10-17 12:31:04 <ncopa> i think i can just do patch surgery
2013-10-17 12:31:06 <fabled> just delete that hunk?
2013-10-17 12:31:10 <ncopa> and nuke that
2013-10-17 12:31:15 <fabled> yes
2013-10-17 12:31:17 <fabled> works for me
2013-10-17 12:31:31 <fabled> just need to remember that in case we update it. the patch is from gregor's patch set
2013-10-17 12:32:18 <ncopa> what can i add as comment for  -fno-stack-check ?
2013-10-17 12:32:33 <fabled> umm. fix ada bootstrap
2013-10-17 12:32:46 <fabled> because piepatch adds -fstack-check to spec
2013-10-17 12:32:54 <fabled> and that breaks ada bootstrap
2013-10-17 12:32:58 <ncopa> ok
2013-10-17 12:33:10 <ncopa> i wonder if i shuld make a patch for it
2013-10-17 12:33:19 <ncopa> so we can detect when its no longer needed
2013-10-17 12:33:26 <ncopa> but sounds like sed is fine in this case
2013-10-17 12:33:33 <fabled> you could just modify the piepatch
2013-10-17 12:33:40 <fabled> it adds the -fno-stack-protector bootflag hunk
2013-10-17 12:33:46 <fabled> so if inlining the pipeatches
2013-10-17 12:33:51 <fabled> you can just modify that line directly
2013-10-17 12:33:58 <ncopa> yeah
2013-10-17 12:34:05 <ncopa> good idea
2013-10-17 12:34:29 <ncopa> will need remember that in case we import new versions in future
2013-10-17 12:34:34 <ncopa> maybe keep the sed line
2013-10-17 12:34:53 <ncopa> or i tell zorry to include it in upstream piepatches?
2013-10-17 12:36:05 <fabled> yes
2013-10-17 12:36:33 <fabled> hum
2013-10-17 12:36:37 <fabled> build64-edge failed with openjdk too
2013-10-17 12:36:52 <fabled> apparently at same place
2013-10-17 12:37:06 <ncopa> did build-2-6 fail?
2013-10-17 12:37:06 <fabled> i'll just maybe push it to with defaulting to bootstrap
2013-10-17 12:37:24 <fabled> no build-2-6 is still in-progress
2013-10-17 12:37:34 <fabled> let's wait until that finishes
2013-10-17 12:37:37 <ncopa> yeah
2013-10-17 12:38:08 <fabled> when should we move the edge builders to the new boxes?
2013-10-17 12:40:38 <ncopa> once we have configured the public ip network interface
2013-10-17 12:40:41 <ncopa> to be a bridge
2013-10-17 12:41:04 <ncopa> basically, create br0, move ip to br0 and add bond0 to the bridge
2013-10-17 12:41:05 <ncopa> thats it
2013-10-17 12:41:15 <ncopa> i think i'd like the edge builders to have a public ip
2013-10-17 12:41:27 <ncopa> for convenience
2013-10-17 12:41:41 <ncopa> hm
2013-10-17 12:41:48 <ncopa> maybe now is a good time to move them...
2013-10-17 12:42:02 <ncopa> before openjdk starts building again...
2013-10-17 12:44:18 <fabled> would be doable
2013-10-17 12:45:48 <ncopa> can i start with bld2?
2013-10-17 12:45:51 <ncopa> your box
2013-10-17 12:46:21 <ncopa> i think I'll do sleep 60s && reboot in separate win, in case i lose it
2013-10-17 12:46:56 <fabled> yes
2013-10-17 12:46:57 <fabled> it's ok
2013-10-17 12:47:11 <fabled> not using it now
2013-10-17 12:51:33 <ncopa> ok, here we go...
2013-10-17 12:52:15 <ncopa> :-(
2013-10-17 12:52:19 <ncopa> i lost it
2013-10-17 12:53:25 <fabled> openjdk6 failed on build-2-6
2013-10-17 12:53:31 <fabled> err. build64-2-6
2013-10-17 12:54:33 <fabled> hum
2013-10-17 12:54:49 <fabled> actually one possibility is that my piepatch to kernel isn't working. and grsec kernel mmap patch works.
2013-10-17 12:55:02 <fabled> would explain why it works on the new boxes, but not on the old builder
2013-10-17 12:55:38 <fabled> or just out-of-mem
2013-10-17 12:56:26 <fabled> great. that failed too.
2013-10-17 12:56:36 <fabled> curiously, i think they all failed in same place.
2013-10-17 12:57:33 <fabled> weird. it's 'conftest' that fails
2013-10-17 12:57:38 <fabled> according build logs
2013-10-17 12:57:41 <fabled> build dmesg
2013-10-17 12:58:07 <ncopa> weird... i did test it in my 2.6 container here
2013-10-17 12:58:09 <ncopa> and here it built...
2013-10-17 12:58:16 <fabled> lxc on grsec?
2013-10-17 12:58:20 <ncopa> yup
2013-10-17 12:58:30 <fabled> sounds like vserver kenel is causing the issues
2013-10-17 12:58:30 <ncopa> so might be vserver kernel issue?
2013-10-17 12:58:53 <fabled> can be also the PIE issue and my patch not fixing it fully, but grsec mmap randomization fixing it
2013-10-17 12:59:08 <ncopa> could be
2013-10-17 13:00:20 <fabled> so move edge and 2-6 builders to new hw first thing?
2013-10-17 13:00:32 <fabled> 2. build.lua busy loop fix
2013-10-17 13:00:44 <fabled> 3. gcc 4.8.2
2013-10-17 13:00:46 <fabled> 4. openjdk6
2013-10-17 13:07:34 <ncopa> build.lua should be fixed
2013-10-17 13:18:13 <ncopa> i'm struggeling with reconfiguring the pub ip on bld2
2013-10-17 13:18:22 <ncopa> will reboot it now for 3rd time
2013-10-17 13:36:13 <barthalion> seems like it wasn't the best moment to push something now
2013-10-17 14:15:03 <ncopa> i will move edge builders now
2013-10-17 14:29:24 <ncopa> i will stop build-edge
2013-10-17 14:29:32 <ncopa> its building i think
2013-10-17 16:53:21 <ncopa_> algitbot: build master
2013-10-17 16:56:58 <ncopa_> it turns out that you cannot bridge a bond interface
2013-10-17 16:57:01 <ncopa_> th switch goes nuts
2013-10-17 16:58:55 <ncopa_> build-edge is now running on bld1.alpinelinux.org
2013-10-17 16:58:57 <ncopa_> new hw
2013-10-17 17:14:12 <ncopa_> yay...
2013-10-17 17:14:21 <Niichan> louder
2013-10-17 17:14:54 <ncopa_> openjdk built
2013-10-17 17:15:24 <ncopa_> 21mins approx
2013-10-17 17:16:29 <jbilyk_> ncopa: since kamailio built in 2-6, how come it didn't upload? because openjdk6 failed?
2013-10-17 17:17:12 <ncopa_> yeah
2013-10-17 17:17:23 <ncopa_> hmm openjdk is still building?
2013-10-17 17:17:37 <jbilyk_> ncopa: I just looked at http://dev.alpinelinux.org/buildlogs/build-2-6.1382018774.log
2013-10-17 17:17:50 <ncopa_> jbilyk_: i think we have issue with the build.a.o server
2013-10-17 17:17:57 <jbilyk_> ok
2013-10-17 17:19:16 <ncopa_> so we are working on moving the builders over to new hw
2013-10-17 17:19:28 <ncopa_> build64-edge is not fully migrated yet
2013-10-17 17:19:32 <ncopa_> build-edge it
2013-10-17 17:19:38 <ncopa_> and it got renamed to build-edge-x86
2013-10-17 18:16:57 <barthalion> does new hardware mean that buildbots will be faster? :p
2013-10-17 20:42:11 <ncopa_> algitbot: build master
2013-10-17 20:43:17 <ncopa_> algitbot: build master
2013-10-17 20:45:08 <ncopa_> algitbot: build master
2013-10-17 21:02:31 <ncopa_> algitbot: build master
2013-10-17 21:12:55 <ncopa_> algitbot: build master
2013-10-18 04:54:03 <fabled> Niichan, the re-built arm packages are uploaded now. you need to do "apk upgrade -U -a" to switch over
2013-10-18 04:54:23 <fabled> i did not bother bumping pkgrel as musl builds are experimental and the issue did not affect the main builds
2013-10-18 05:01:19 <fabled> cool. new openjdk6 built on the new builder hw.
2013-10-18 05:03:22 <fabled> uuu. new gcc too
2013-10-18 05:50:49 <fabled> hum. the arm build does have some issues; e.g. åäö does not show up right on rpi console
2013-10-18 06:36:32 <ncopa> algitbot: build master
2013-10-18 06:37:30 <fabled> algitbot: build master
2013-10-18 06:40:44 <fabled> algitbot: build master
2013-10-18 06:43:30 <ncopa> algitbot: build 2.6-stable
2013-10-18 08:00:57 <fabled> so in 1 hour 20 mins, it rebuilt gcc, openjdk6, php and few other gigantic pieces of software. not bad.
2013-10-18 08:10:17 <fabled> ncopa, could we change ^^^ mime types to text/plain so they open up on browser
2013-10-18 08:11:44 <ncopa>   ".log"          =>      "text/plain",
2013-10-18 08:12:45 <ncopa> barthalion: yes, the buildbots should be significantly faster now
2013-10-18 08:13:01 <ncopa> not only is the hw much better, we also got 2 of them
2013-10-18 08:13:16 <ncopa> i have put all x86 on one box and all x86_64 on the other
2013-10-18 08:13:55 <ncopa> so the load should be better distributed
2013-10-18 08:15:02 <fabled> hum
2013-10-18 08:15:11 <fabled> maybe it's my ff then
2013-10-18 08:15:32 <ncopa> i think it does so when it finds special chars in there
2013-10-18 08:16:44 <fabled> possible
2013-10-18 08:16:53 <fabled> i wonder why the logs contain lot of zero bytes
2013-10-18 08:17:13 <ncopa> good question
2013-10-18 08:17:36 <ncopa> we should probably filter the logs
2013-10-18 08:17:48 <ncopa> i think what we want is
2013-10-18 08:17:52 <ncopa> per package log
2013-10-18 08:17:55 <ncopa> filter the log
2013-10-18 08:18:27 <ncopa> so it does not contain special chars
2013-10-18 08:19:11 <ncopa> and maybe some way to let abuild tell current status (eg pkgname: unacking)
2013-10-18 08:19:22 <ncopa> which can be displayed on some webpage
2013-10-18 08:19:38 <ncopa> currently building package foo, 2 of N done
2013-10-18 08:19:41 <ncopa> or similar
2013-10-18 08:29:46 <fabled> yes
2013-10-18 08:29:54 <fabled> ncopa, i'd really appreciate the per-pkg logs
2013-10-18 08:29:59 <fabled> with timestamps of start and end
2013-10-18 08:30:12 <fabled> and also remember failed packages.
2013-10-18 08:30:26 <fabled> so the builders could build other stuff if dependant packages are ok.
2013-10-18 08:31:45 <ncopa> i kinda like that if something broke is pushed to builders everything stop
2013-10-18 08:31:53 <ncopa> because then everyone is interessted in fixing it
2013-10-18 08:32:18 <ncopa> and everyone wil spank the person that pushed bad commit
2013-10-18 08:32:52 <ncopa> but having that feature on your local dev builder is useful, yes
2013-10-18 08:36:35 <ncopa> java build is really slowing down everything on the old build server
2013-10-18 08:36:40 <ncopa> io is slow
2013-10-18 08:37:27 <ncopa> i think the build pracitcally kills the io cache due to excessive use of mem
2013-10-18 08:45:18 <fabled> quite likely
2013-10-18 09:02:43 <fabled> did openjdk6 compile on both 2-6 builders?
2013-10-18 09:02:53 <ncopa> x86_64 still buildling i think
2013-10-18 09:03:01 <fabled> oh, ok
2013-10-18 09:03:02 <ncopa> x86 built
2013-10-18 09:03:28 <fabled> so apparently its looking good if it's been building 2+ hours
2013-10-18 09:03:40 <ncopa> yup
2013-10-18 09:03:49 <ncopa> will likely need 2 more hours...
2013-10-18 09:07:14 <ncopa> re distfiles
2013-10-18 09:07:30 <ncopa> i think for stable builders we want save the distfiles archives
2013-10-18 09:07:59 <ncopa> so i have saved v2.6 distfiels in /var/cache/distfiles/v2.6
2013-10-18 09:08:17 <ncopa> i think we can checksum and hardlink dupes
2013-10-18 09:09:17 <ncopa> that way it will be easier to remove things in /var/cache/distfiles once in a while
2013-10-18 09:21:28 <ncopa> build-2-6 is migrated
2013-10-18 09:21:33 <ncopa> to bld1
2013-10-18 09:21:40 <ncopa> should be operative now
2013-10-18 10:42:07 <shafire> hi
2013-10-18 10:44:16 <shafire> really cool build server :)
2013-10-18 11:07:56 <ncopa> i will migrate build64-2-6 to build-2-6-x86_64 on new hardware
2013-10-18 11:09:05 <ncopa> aw...
2013-10-18 11:09:11 <ncopa> it was in the middle of a build...
2013-10-18 11:09:27 <barthalion> yeah, it hasn't finished yet
2013-10-18 11:09:34 <ncopa> i killed it
2013-10-18 11:09:35 <ncopa> sorry
2013-10-18 11:09:40 <barthalion> well, no problem
2013-10-18 11:09:55 <barthalion> I guess build master will fix it after the move
2013-10-18 11:09:57 <ncopa> i will migrate it and let it start over
2013-10-18 11:09:58 <ncopa> yeah
2013-10-18 11:10:34 <fabled> ncopa, after migration - seems like my musl builder is rebuiding everything
2013-10-18 11:10:43 <ncopa> hm
2013-10-18 11:10:45 <fabled> i guess it's the hack script i use to allow building even if some pkgs fail
2013-10-18 11:10:46 <ncopa> might be filestamps
2013-10-18 11:10:54 <fabled> something like that
2013-10-18 11:11:01 <ncopa> use abuild -k to prevent that
2013-10-18 11:11:07 <fabled> i had it
2013-10-18 11:11:09 <fabled> but removed it
2013-10-18 11:11:12 <ncopa> oh..
2013-10-18 11:11:14 <fabled> should probably add it back
2013-10-18 11:11:19 <ncopa> ok
2013-10-18 11:11:52 <fabled> anyways gcc 4.8.2 now bootstraps fully on the arm board too
2013-10-18 11:11:56 <ncopa> nice!
2013-10-18 11:13:01 <shafire> thanks!
2013-10-18 11:14:53 <fabled> we should probably write a news item about the musl and arm stuff
2013-10-18 11:16:38 <fabled> i'll probably try to get musl/x86_64 bootstrapped over the weekend
2013-10-18 11:17:07 <shafire> how are the stats?
2013-10-18 11:17:23 <fabled> which one?
2013-10-18 11:17:29 <shafire> musl (arm, x86)
2013-10-18 11:17:39 <fabled> last i checked it were:
2013-10-18 11:17:41 <fabled> musl-x86: 1462 of 1676
2013-10-18 11:17:41 <fabled> musl-arm: 1425 of 1676
2013-10-18 11:18:16 <shafire> :-)
2013-10-18 11:20:54 <ncopa> algitbot: build 2.6-stable
2013-10-18 11:22:27 <ncopa> algitbot: build 2.6-stable
2013-10-18 11:22:52 <ncopa> ok, now its up and running
2013-10-18 11:29:06 <ncopa> 6mins to build and upload php
2013-10-18 11:29:09 <ncopa> nice :)
2013-10-18 11:34:39 <fabled> could be even less if we parallelized the builds (it's built quite a number of times to get all subpkgs)
2013-10-18 11:39:31 <ncopa> yeah
2013-10-18 11:39:46 <ncopa> i think configury is the bottleneck nowdays
2013-10-18 11:58:19 <barthalion> ncopa: can you take a look at make 4.0?
2013-10-18 12:04:51 <ncopa> barthalion: i talked with fabled about it
2013-10-18 12:05:06 <ncopa> i wonder if it is worth it
2013-10-18 12:05:11 <ncopa> for v2.7
2013-10-18 12:05:15 <ncopa> worth the job
2013-10-18 12:05:35 <ncopa> i have already started up build-2-7-x86
2013-10-18 12:05:51 <ncopa> and when pushing gcc-4.8.2 i started from scratch again
2013-10-18 12:06:11 <ncopa> so i wonder if it might be an idea to wait til post v2.7
2013-10-18 12:06:55 <ncopa> when v2.7 is released i want evertying in aports to build
2013-10-18 12:07:03 <ncopa> so you can buidl the release from scratch
2013-10-18 12:07:46 <ncopa> that said, make 4.0 appears to be fairly safe
2013-10-18 12:08:04 <ncopa> so it might be worth it
2013-10-18 12:08:23 <ncopa> barthalion: has arch linux upgrade to make 4.0?
2013-10-18 12:34:33 <fabled> ncopa, make 4.0 looks safe. at least the patch i added for musl needs rebase (simple, just change filename to patch)
2013-10-18 12:34:44 <fabled> but for some odd reason it fails now one test case in musl
2013-10-18 12:34:45 <fabled> err
2013-10-18 12:34:46 <fabled> in uclibc
2013-10-18 12:37:20 <ncopa> i am ok that you push make 4.0 if you have the testcase fixed
2013-10-18 12:37:40 <fabled> no
2013-10-18 12:37:44 <fabled> i don't
2013-10-18 12:37:49 <fabled> i have make 4.0 that works on musl
2013-10-18 12:37:56 <fabled> but not one that passes all tests on uclibc
2013-10-18 12:38:00 <ncopa> ok
2013-10-18 12:38:05 <ncopa> do you want me to look at it?
2013-10-18 12:38:13 <fabled> if you feel like it
2013-10-18 12:38:23 <fabled> it's actually the same test case that my patch fixes for musl
2013-10-18 12:38:29 <ncopa> oh
2013-10-18 12:38:35 <fabled> i think uclibc stdio does not keep the stdout in "error" state
2013-10-18 12:39:01 <fabled> so it does not detect the error properly
2013-10-18 12:39:05 <fabled> hm
2013-10-18 12:39:34 <fabled> ncopa, could you check that we don't need any other patches kept?
2013-10-18 12:39:37 <fabled> none of them applies
2013-10-18 12:39:42 <ncopa> ok
2013-10-18 12:39:43 <fabled> some are detected as applied
2013-10-18 12:39:45 <fabled> some just fail
2013-10-18 12:40:03 <fabled> the only that i know for sure we want, is fix-atexit-exit.patch that i wrote
2013-10-18 12:40:06 <fabled> and i have that rebased
2013-10-18 12:41:50 <fabled> huh
2013-10-18 12:41:56 <fabled> make 4.0 passed all tests on my laptop
2013-10-18 12:42:07 <fabled> uclibc x86
2013-10-18 12:42:36 <ncopa> so i just remove all patches except fix-atexit
2013-10-18 12:42:40 <fabled> yes
2013-10-18 12:43:24 <barthalion> ncopa: yeah, it did
2013-10-18 12:43:43 <ncopa> fabled: can you sprunge the rebased fix-atexit?
2013-10-18 12:43:54 <ncopa> barthalion: good, then we are not the first ones to bump into the bugs :)
2013-10-18 12:43:59 <fabled> ncopa, http://sprunge.us/hMdC
2013-10-18 12:44:15 <fabled> take descrption from old .patch
2013-10-18 12:47:52 <ncopa> misc/close_stdout ....................................... FAILED (0/1 passed)
2013-10-18 12:52:05 <ncopa> i get other error message
2013-10-18 12:52:12 <ncopa> i think thats the issue
2013-10-18 12:52:54 <ncopa> i get permissions denied
2013-10-18 12:54:07 <fabled> oh
2013-10-18 12:54:14 <fabled> do you have dev/full in the builder?
2013-10-18 12:54:19 <ncopa> yes
2013-10-18 12:54:19 <fabled> it works on my laptop
2013-10-18 12:54:25 <fabled> can the builder read dev/full
2013-10-18 12:54:36 <fabled> my laptop has:
2013-10-18 12:54:37 <fabled> $ ls -la /dev/full
2013-10-18 12:54:37 <fabled> crw-rw-rw- 1 root root 1, 7 Oct 17 08:02 /dev/full
2013-10-18 12:54:39 <ncopa> lxc permissions probably
2013-10-18 12:54:43 <fabled> ok
2013-10-18 12:54:51 <fabled> yes, you need to white list that char device
2013-10-18 12:54:58 <ncopa> I'll check that
2013-10-18 12:55:01 <fabled> oh
2013-10-18 12:55:04 <fabled> that would explain
2013-10-18 12:55:07 <fabled> i tested it on vserver/musl
2013-10-18 12:55:11 <fabled> and lxc/uclibc
2013-10-18 12:55:25 <fabled> and yes, that tests uses dev/full
2013-10-18 12:56:00 <fabled> yup
2013-10-18 12:56:06 <fabled> c 1:7 is not listed
2013-10-18 12:56:14 <fabled> shoudl be listed along with null and zero
2013-10-18 12:56:43 <fabled> lxc.cgroup.devices.allow = c 1:7 rwm
2013-10-18 12:56:48 <fabled> should make it work
2013-10-18 12:59:18 <ncopa> yup
2013-10-18 12:59:19 <ncopa> works
2013-10-18 12:59:58 <ncopa> i suppose i shoudl fix that in all builders
2013-10-18 13:00:20 <fabled> yes
2013-10-18 13:00:31 <fabled> should fix it in lxc template
2013-10-18 13:01:36 <ncopa> yeah
2013-10-18 13:01:41 <ncopa> will do that afterwards
2013-10-18 13:02:14 <fabled> ok. i think make 4.0 would be good to push to master now then
2013-10-18 13:02:38 <fabled> (after lxc perm fix; and including the rebased _exit patch)
2013-10-18 13:03:00 <ncopa> restarting bld1 builders
2013-10-18 13:03:18 <fabled> let my musl-x86 run
2013-10-18 13:07:14 <ncopa> and ttdev-*?
2013-10-18 13:07:28 <ncopa> i dont restart those, you can do that yourself when convenient
2013-10-18 13:07:52 <fabled> ttdev you can restart
2013-10-18 13:10:52 <ncopa> ttdev-edge-x86     ttdev-edge-x86_64
2013-10-18 13:10:56 <ncopa> ok to do so now?
2013-10-18 13:11:37 <ncopa> ok. you didnt say no.. :)
2013-10-18 13:11:49 <ncopa> done
2013-10-18 13:13:17 <fabled> thanks
2013-10-18 13:13:19 <fabled> rebooting
2013-10-18 13:13:58 <ncopa> hm
2013-10-18 13:17:31 <ncopa> seems like they have issues starting up networking during boot
2013-10-18 13:17:37 <ncopa> the containers
2013-10-18 13:17:39 <ncopa> i wonder why
2013-10-18 13:18:57 <fabled> i noticed same
2013-10-18 13:19:25 <ncopa> algitbot: build master
2013-10-18 13:19:49 <fabled> ncopa, did you plan to rename build64-2-6 according to new naming?
2013-10-18 13:19:56 <ncopa> yes
2013-10-18 13:20:04 <ncopa> did i forget that?
2013-10-18 13:20:41 <fabled> i think so
2013-10-18 13:21:36 <ncopa> algitbot: build master
2013-10-18 13:23:12 <ncopa> i'll build from scratch again
2013-10-18 13:23:14 <ncopa> algitbot: build master
2013-10-18 13:24:51 <fabled> ncopa, build what?
2013-10-18 13:24:55 <fabled> 2.7 builder?
2013-10-18 13:31:31 <vkrishn> was testing lxc in 2.6 using http://wiki.alpinelinux.org/wiki/LXC
2013-10-18 13:32:16 <vkrishn> was able to setup, but seems everytime I start the lxc guest1 I have to loggin into it and do /etc/init.d/networking start
2013-10-18 13:33:12 <vkrishn> also I was not able to ssh into alpine host from outside
2013-10-18 13:33:23 <vkrishn> all test was done on alpine(qemu)
2013-10-18 13:34:12 <vkrishn> using dhcp only
2013-10-18 13:34:45 <ncopa> fabled: yes, i start 2.7 builder from scratch again, so we test build everything with make-4
2013-10-18 13:53:57 <barthalion> I'll fix it in a moment
2013-10-18 14:07:08 <ncopa> barthalion: http://sprunge.us/GPKI
2013-10-18 14:07:24 <ncopa> that will do the first
2013-10-18 14:07:33 <ncopa> but it appears to miss some deps
2013-10-18 14:07:47 <ncopa> barthalion: what is your dev platform? arch linux?
2013-10-18 14:23:57 <fabled> ok. musl-x86 rebuilt and updated
2013-10-18 14:24:05 <fabled> gcc 4.8.2 + make 4.0 + and other cool stuff
2013-10-18 14:24:56 <Frosh> Will Alpine come out with a Gnome/KDE branch?
2013-10-18 14:35:04 <barthalion> Frosh: Gnome is very unlikely, as it depends on systemd
2013-10-18 14:35:15 <barthalion> Frosh: KDE is doable, but not sure if anyone has it on the todo
2013-10-18 14:35:31 <barthalion> ncopa: depends where I sit, but usually Arch or Debian
2013-10-18 14:36:36 <vkrishn> barthalion, I thought fedora was one
2013-10-18 14:37:04 <barthalion> vkrishn: you mean my dev platform?
2013-10-18 14:37:27 <vkrishn> yes
2013-10-18 14:37:46 <Frosh> or any other desktop, lxde or any other?
2013-10-18 14:38:11 <vkrishn> Arch and Fedora + AL
2013-10-18 14:38:22 <barthalion> Frosh: there is xfce in repo
2013-10-18 14:38:37 <barthalion> vkrishn: well, I'm Arch dev
2013-10-18 14:38:53 <barthalion> vkrishn: and the only fedora I tried was fedora 7 :p
2013-10-18 14:39:27 <barthalion> Frosh: and part of LXDE from what I see
2013-10-18 14:39:35 <barthalion> Frosh: openbox, lxdm, lxpolkit, lxsession and lxterminal
2013-10-18 14:39:46 <barthalion> Frosh: if you need i.e. lxpanel or something else just drop me a line
2013-10-18 14:39:48 <vkrishn> Frosh, do try 'Awesome' if using for simple tasks/setup
2013-10-18 14:42:45 <Frosh> awesome looks awesome, gonna look into it
2013-10-18 14:44:03 <vkrishn> "Awesome" can be set as complex monitoring station
2013-10-18 14:44:49 <vkrishn> specially using very large displays
2013-10-18 14:44:50 <vkrishn> or multiple
2013-10-18 14:46:04 <barthalion> but it's fine on 1366x768 too
2013-10-18 14:48:29 <vkrishn> barthalion, think like huge monitoring screen where various charts, activity of various servers are displayed
2013-10-18 14:49:17 <vkrishn> some activity automatically  gets popped up (and gets displayed in perticular area of the screen) by using scripting
2013-10-18 14:50:30 <Frosh> just what i want
2013-10-18 14:50:56 <vkrishn> :)
2013-10-18 14:54:24 <ncopa> i use xfce on 2x24" screens
2013-10-18 14:54:32 <ncopa> with alpine linux
2013-10-18 14:54:48 <ncopa> i did look at kde at some point but didnt have enough motivation to finish it up
2013-10-18 14:54:54 <ncopa> i think kde for alpine is doable
2013-10-18 14:55:11 <ncopa> gnome depends on systemd so that likely not going to happen
2013-10-18 14:55:18 <ncopa> we might get parts of it though
2013-10-18 15:33:16 <Niichan> fabled: I just ran the upgrade, rebooting the pi
2013-10-18 15:37:45 <Niichan> fabled: Tmux still makes weechat look messed up: http://i.imgur.com/zozm7y3.png without tmux: http://i.imgur.com/L0Cja5c.png
2013-10-18 16:22:19 <Frosh> I think I got xfce to work, but then I couldn't figure out how to get it to work with two different size monitor, or I had to do it everytime I rebooted.
2013-10-18 16:23:00 <Frosh> Niichan: what is the wallpaper, looks sexy
2013-10-18 16:23:38 <Niichan> Frosh: https://upload.wikimedia.org/wikipedia/commons/3/39/Seattle_3.jpg
2013-10-18 16:23:42 <Niichan> (I'm from Seattle)
2013-10-18 19:08:09 <fabled> Niichan, what is the Terminal emulator you use, and what is the OS / charset on that box
2013-10-18 19:08:24 <Niichan> fabled: Terminator, Debian
2013-10-18 19:08:31 <fabled> utf8?
2013-10-18 19:08:36 <Niichan> rainbowdash ~ > echo $CHARSET
2013-10-18 19:08:38 <Niichan> UTF-8
2013-10-18 19:08:53 <fabled> hum
2013-10-18 19:09:05 <fabled> is WIDTH set right?
2013-10-18 19:09:21 <fabled> or how does the width get detected
2013-10-18 19:10:23 <Niichan> It looks messed up in xterm
2013-10-18 19:10:57 <Niichan> http://i.imgur.com/RSWP0cD.png
2013-10-18 19:11:41 <fabled_rpi> Terminal on Alpine -> ssh to rpi -> tmux -> weechat
2013-10-18 19:11:45 <fabled_rpi> and it's perfect
2013-10-18 19:11:54 <Niichan> Is it uxterm?
2013-10-18 19:13:24 <fabled> $ apk info -W /usr/bin/Terminal
2013-10-18 19:13:25 <fabled> /usr/bin/Terminal is owned by terminal-0.4.8-r3
2013-10-18 19:16:40 <fabled> in env i have:
2013-10-18 19:16:42 <fabled> LANG=C.utf-8
2013-10-18 19:16:46 <fabled> CHARSET=UTF-8
2013-10-18 19:17:17 <fabled> tmux sets
2013-10-18 19:17:19 <fabled> TERM=screen
2013-10-18 19:17:34 <fabled> otherwise TERM=xterm
2013-10-18 19:23:04 <user1> works also on LXTerminal
2013-10-18 19:24:51 <fabled> Niichan, have you done setup-keymap on rpi ?
2013-10-18 19:24:58 <Niichan> fabled: No
2013-10-18 19:25:23 <fabled> i think i had not done it either
2013-10-18 19:25:24 <Niichan> I have now
2013-10-18 19:26:01 <Niichan> Still broken
2013-10-18 19:26:11 <fabled> hum
2013-10-18 19:27:11 <fabled> Niichan, works on all the imaginable combos i've tried
2013-10-18 19:27:41 <fabled> either you have some old package, there's mmc card corruption, or there's something in your host, or it's just bad mojo
2013-10-18 19:28:40 <Niichan> 
ACTION shrugs

2013-10-18 19:28:57 <Niichan> I'll live in the meantime
2013-10-18 19:29:25 <fabled> try "apk fix tmux"
2013-10-18 19:29:49 <fabled> and ncurses-terminfo
2013-10-18 19:30:20 <fabled> is COLORTERM env set?
2013-10-18 19:30:52 <Niichan> No
2013-10-18 19:34:21 <fabled> well if "apk fix -U tmux ncurses-terminfo" did not fix it, then i don't know what to try
2013-10-18 19:34:47 <fabled> i can't reproduce it here
2013-10-18 22:08:45 <fabled> ok. i'm gonna give a try for x86_64/musl
2013-10-18 22:09:05 <Niichan> woohoo
2013-10-18 22:28:24 <fabled> oookay, full cross compiler built
2013-10-18 22:31:54 <fabled> crossbuilding native gcc now. let's see if that blows up.
2013-10-18 22:32:06 <fabled> if not, then i think we'll have bootstrap environment mostly done.
2013-10-18 22:33:18 <fabled> crap. it blew up
2013-10-18 22:34:06 <fabled> hum
2013-10-18 22:35:46 <fabled> bah
2013-10-18 23:36:59 <fabled> gcc autoconfigure is brain damaged
2013-10-18 23:37:16 <fabled> but i think i figured a workaround. let's see if the beast builds now.
2013-10-18 23:44:57 <fabled> woot
2013-10-18 23:45:16 <fabled> fabled 2 - gcc 1
2013-10-18 23:57:35 <fabled> it built
2013-10-18 23:57:43 <fabled> now to just setup container and do native self-build
2013-10-19 00:50:21 <fabled> doing full native self gcc build now - looking good.
2013-10-19 00:50:35 <fabled> will probably have musl-x86_64 ready to be pushed out tomorrow sometime
2013-10-19 01:14:03 <fabled> hum
2013-10-19 01:15:08 <fabled> ncopa, something is wrong with bld2. look at it's dmesg, there's some "kernel BUG" errors in it
2013-10-19 01:16:14 <fabled> [127973.206767] kernel BUG at /home/buildozer/aports/main/linux-grsec/src/linux-3.10/mm/filemap.c:2023!
2013-10-19 01:16:15 <fabled> [127973.211450] invalid opcode: 0000 [#2] SMP
2013-10-19 01:16:38 <fabled> [127973.343333] Call Trace:
2013-10-19 01:16:38 <fabled> [127973.348010]  [<ffffffff810a18fd>] generic_file_buffered_write+0x18d/0x238
2013-10-19 01:16:38 <fabled> [127973.352761]  [<ffffffffa05c5430>] ? ext4_journalled_aops+0xa0/0xa0 [ext4]
2013-10-19 01:16:38 <fabled> [127973.362243]  [<ffffffff810a2fbe>] generic_file_aio_write+0x56/0xa4
2013-10-19 01:16:38 <fabled> [127973.357527]  [<ffffffff810a2f38>] __generic_file_aio_write+0x2cb/0x2fb
2013-10-19 01:16:39 <fabled> [127973.367053]  [<ffffffffa0593a3b>] ext4_file_write+0x373/0x3a8 [ext4]
2013-10-19 01:16:41 <fabled> [127973.371778]  [<ffffffff810d8c1d>] do_sync_readv_writev+0x66/0x92
2013-10-19 01:16:43 <fabled> [127973.376574]  [<ffffffff810da2eb>] do_readv_writev+0xfb/0x21b
2013-10-19 01:16:45 <fabled> [127973.381290]  [<ffffffffa05936c8>] ? ext4_unwritten_wait+0x98/0x98 [ext4]
2013-10-19 01:16:49 <fabled> [127973.386068]  [<ffffffff810d8c49>] ? do_sync_readv_writev+0x92/0x92
2013-10-19 01:16:51 <fabled> [127973.390898]  [<ffffffff810c2370>] ? do_brk+0x279/0x2a2
2013-10-19 01:16:53 <fabled> [127973.395647]  [<ffffffff810da442>] vfs_writev+0x37/0x4b
2013-10-19 01:16:55 <fabled> [127973.400441]  [<ffffffff810da57c>] SyS_writev+0x44/0x9a
2013-10-19 01:16:57 <fabled> could explain why native gcc build failed
2013-10-19 01:17:00 <fabled> musl might be triggering it - it uses writev() extensively
2013-10-19 01:19:21 <fabled> seems also that rm -rf on gcc/src hung. definitely sounds like nastyness in kernel.
2013-10-19 01:33:26 <kaniini> generic_file_buffered_write ...
2013-10-19 01:33:32 <kaniini> it's a jump-to-NULL
2013-10-19 01:33:47 <kaniini> i bet
2013-10-19 10:53:39 <fabled> cool. vlc works on musl/x86
2013-10-19 11:09:25 <fabled> ncopa, https://lkml.org/lkml/2013/10/12/85
2013-10-19 11:09:33 <fabled> seems like recompiling kernel with new gcc might fix issues
2013-10-19 11:09:42 <fabled> oh there's new releases too
2013-10-19 11:09:47 <fabled> of stable kernels
2013-10-19 11:09:51 <fabled> should upgrade builders to those likely
2013-10-19 12:13:26 <barthalion> ↑ wut
2013-10-19 12:22:50 <barthalion> I wonder if it's new gcc fault
2013-10-19 12:31:41 <barthalion> on 2.6 it fails because libopcodes is missing
2013-10-19 12:34:13 <barthalion> ncopa, fabled: urls to build logs for x86_64 are out of date, s/dev/bld2/ should fix it
2013-10-19 15:48:25 <shafire> hi
2013-10-19 16:04:45 <barthalion> ncopa: openrc 0.12 has been released
2013-10-19 18:23:15 <Niichan> Do you know what java compiler I can use in Alpine?
2013-10-19 18:25:55 <barthalion> maybe gcc-java provides gcj
2013-10-19 18:26:08 <barthalion> but I'm not sure about it
2013-10-19 19:10:48 <fabled> Niichan, gcc-java has gcj, java-gcj-compat has gcj compiled ecj and openjdk6 is OpenJDK6 javac
2013-10-19 19:11:43 <Niichan> fabled: I don't have javac
2013-10-19 19:11:56 <Niichan> (and I installed openjdk6)
2013-10-19 19:12:07 <fabled> it's not in path
2013-10-19 19:12:47 <fabled> try /usr/lib/jvm/java-1.6-openjdk/bin/javac
2013-10-19 19:13:02 <Niichan> Ah
2013-10-19 19:13:06 <fabled> actually it should have wrapper to launch it
2013-10-19 19:13:07 <Niichan> 
ACTION adds to path

2013-10-19 19:13:07 <fabled> i think
2013-10-19 19:13:22 <fabled> at least 'java' runner has a wrapper
2013-10-19 19:13:42 <fabled> i think we left the wrapper out from the sdk tools
2013-10-19 19:13:46 <fabled> since you likely need them all
2013-10-19 19:13:54 <fabled> so you should just add the specific jdk to your path
2013-10-19 19:14:43 <fabled> we could fix java-common to provide the symlinks for javac too
2013-10-19 19:15:06 <fabled> java-common maintains /usr/lib/jvm/default-jvm as symlink to "best" java installed
2013-10-19 19:16:04 <fabled> $ ls -la /usr/bin/java
2013-10-19 19:16:04 <fabled> lrwxrwxrwx 1 root root 35 Aug  6 11:49 /usr/bin/java -> ../lib/jvm/default-jvm/jre/bin/java
2013-10-19 19:16:14 <fabled> so could do similar for javac, i guess
2013-10-20 09:59:43 <barthalion> ncopa: I've finally prepared linux-rt
2013-10-20 09:59:55 <barthalion> ncopa: can you take a look before I push it?
2013-10-20 10:00:40 <barthalion> ncopa: https://paste.xinu.at/sXEYr/
2013-10-20 10:34:48 <barthalion> seems that it will need some patching
2013-10-20 10:36:52 <fabled> barthalion, yes, patching for x86 assembly, or disable the assembly code
2013-10-20 10:37:05 <fabled> i have similar stuff ahead for arm
2013-10-20 10:37:17 <fabled> seems that very few assembly code is PIC compatible
2013-10-20 10:40:47 <barthalion> so probably it will fail again
2013-10-20 10:46:26 <barthalion> yhm, yeah, it's not ffmpeg
2013-10-20 11:16:46 <barthalion> algitbot: build master
2013-10-20 12:15:56 <barthalion> obviously there are text relocations if I disabled pic ;_;
2013-10-20 16:35:15 <fabled> barthalion, yes, you are not allowed to disable PIC
2013-10-20 16:35:24 <fabled> you need to disable assembly code, or fix it
2013-10-20 16:38:22 <barthalion> fun
2013-10-20 16:39:58 <fabled> all sane distros require x86 libraries to be PIC - so usually they file bugs upstream
2013-10-20 16:40:05 <fabled> for non-x86 the rule is not so strict
2013-10-20 16:40:11 <fabled> but for us
2013-10-20 16:40:21 <fabled> we want PIC of all - otherwise ASLR does not work properly
2013-10-20 16:40:36 <fabled> or we get extra relocation processing
2013-10-20 16:43:14 <fabled> barthalion, the fix is likely quite simple
2013-10-20 16:43:58 <barthalion> it's the first time I fix something like that, but I believe in google fu
2013-10-20 16:44:16 <fabled> gentoo has a nice page about it
2013-10-20 16:44:28 <fabled> e.g. https://wiki.gentoo.org/wiki/Hardened/HOWTO_locate_and_fix_textrels
2013-10-20 16:45:26 <fabled> likely just push/pop of ebx is enough in this case
2013-10-20 16:45:29 <barthalion> I guess that "zm_utils.cpp:167:3: error: PIC register clobbered by '%ebx' in 'asm'" is the simple part
2013-10-20 16:45:35 <fabled> need to check it
2013-10-20 16:45:40 <barthalion> what about zm_image.cpp:4780:3: error: 'asm' operand has impossible constraints?
2013-10-20 16:49:13 <fabled> huh
2013-10-20 16:49:19 <fabled> why don't i have those files at all
2013-10-20 16:49:28 <fabled> it's 'imagemagick' source package, right?
2013-10-20 16:49:32 <barthalion> yes
2013-10-20 16:49:58 <barthalion> well, I think so, everything was fine until I pushed it
2013-10-20 16:50:30 <fabled> no
2013-10-20 16:50:31 <fabled> it's not
2013-10-20 16:50:32 <barthalion> bah
2013-10-20 16:50:33 <barthalion> it's not
2013-10-20 16:50:34 <fabled> it's 'zoneminder'
2013-10-20 16:50:51 <barthalion> which built fine when pushed
2013-10-20 16:51:08 <barthalion> ah, but not for x86
2013-10-20 16:52:01 <barthalion> doesn't change the fact I need to reenable pic
2013-10-20 16:52:41 <fabled> ?
2013-10-20 17:06:38 <fabled> hum
2013-10-20 17:09:57 <barthalion> fixed PIC so far
2013-10-20 17:10:10 <barthalion> now googling for the latter error
2013-10-20 17:10:55 <fabled> i think it's my bad
2013-10-20 17:10:58 <fabled> the ksymoops
2013-10-20 17:11:02 <fabled> should be fixed by the above commit
2013-10-20 17:13:04 <fabled> >>> lftp: Fetching http://ftp.yars.free.net/pub/source/lftp/lftp-4.4.9.tar.bz2
2013-10-20 17:13:08 <fabled> curl: (22) Failed to connect to 2a03:ca00:d:8:214:4fff:fe49:dd46: Network is unreachable
2013-10-20 17:13:08 <fabled> >>> ERROR: lftp: all failed
2013-10-20 17:13:10 <fabled> huh
2013-10-20 17:24:11 <fabled> algitbot: build master
2013-10-20 17:24:30 <fabled> hrmpf
2013-10-20 17:24:45 <fabled> something must be wrong with networking - or uclibc so that it returns ipv6 addresses when it's not really connected
2013-10-20 17:47:58 <fabled> algitbot: build master
2013-10-21 02:16:32 <kaniini> https://home.regit.org/netfilter-en/nftables-quick-howto/
2013-10-21 06:21:22 <ncopa> kaniini: i saw it. looks like its going in for 3.13
2013-10-21 06:46:28 <ncopa> hey, whats up with edge?
2013-10-21 06:47:19 <ncopa> seems like edge/main is totaly broke?
2013-10-21 06:47:30 <ncopa> http://sprunge.us/bNHC
2013-10-21 06:52:55 <ncopa> yup apkindex is missing lots of imprant stuff
2013-10-21 06:52:57 <ncopa> like sed
2013-10-21 06:53:25 <ncopa> algitbot: build master
2013-10-21 06:53:26 <StarWarsFan> morning
2013-10-21 06:53:30 <ncopa> hi
2013-10-21 06:59:59 <ncopa> algitbot: build master
2013-10-21 08:14:37 <ncopa> algitbot: build master
2013-10-21 08:26:48 <ncopa> weird
2013-10-21 08:27:02 <ncopa> something appears to have deleted various packages in edge
2013-10-21 08:27:09 <ncopa> it is rebuilding a whole lot of stuff
2013-10-21 08:29:39 <ncopa> smells like a bug in buildrepo
2013-10-21 08:43:40 <ncopa> ugh
2013-10-21 08:43:44 <ncopa> i know know thappened
2013-10-21 08:51:44 <ncopa> http://nl.alpinelinux.org/alpine/edge/main/x86_64/
2013-10-21 08:51:48 <ncopa> lots of packages are missing
2013-10-21 08:55:44 <vkrishn> what happens when `apk add <pkgs>` is adding some 50 pkgs and disk space runs out at say somewhat in middle 25th pkg ?
2013-10-21 09:13:01 <ncopa> vkrishn: i think it will mark them all as installed, but marked 'broken'
2013-10-21 09:13:16 <ncopa> to fix it you would have to add the space and then do apk fix
2013-10-21 09:17:39 <vkrishn> I would try to simulate this couple of time more, I think it at some point corrupts or deletes the apk related db's in /etc/apk and /lib/apk*
2013-10-21 09:17:52 <vkrishn> but I would test more
2013-10-21 09:32:41 <vkrishn> what is alternative for `tempfile` as in package debianutils on debian ?
2013-10-21 09:33:21 <ncopa> what does tempfile do?
2013-10-21 09:33:37 <ncopa> maybe mktemp is what you are looking for?
2013-10-21 09:34:45 <vkrishn> thanks
2013-10-21 09:35:01 <vkrishn> can this be updated pls, mod_fcgid 2.3.9
2013-10-21 09:37:17 <vkrishn> tempfile - create a temporary file in a safe manner
2013-10-21 09:39:29 <ncopa> vkrishn: i'm trying to fix the edge repo atm
2013-10-21 09:40:04 <vkrishn> ok
2013-10-21 09:40:12 <ncopa> whats the package name of mod_fcgid?
2013-10-21 09:40:23 <vkrishn> apache-mod-fcgid
2013-10-21 09:40:57 <ncopa> curl: (22) The requested URL returned error: 404 Not Found
2013-10-21 09:42:05 <vkrishn> http://mirror.symnds.com/software/Apache//httpd/mod_fcgid/mod_fcgid-2.3.9.tar.gz
2013-10-21 09:58:02 <vkrishn> http://apache.cyberuse.com//httpd/mod_fcgid/ no longer in apache's mirror list
2013-10-21 10:07:05 <ncopa> algitbot: build master
2013-10-21 11:48:57 <ncopa> algitbot: build master
2013-10-21 11:50:09 <ncopa> algitbot: build master
2013-10-21 13:12:28 <shafire> hi
2013-10-21 13:16:52 <ncopa> hi
2013-10-21 13:28:57 <shafire> how are you?
2013-10-21 13:42:35 <ncopa> blue monday :-/
2013-10-21 15:02:19 <ncopa> algitbot: build master
2013-10-21 15:26:43 <shafire> ncopa: will flashcache go into 2.7?
2013-10-21 15:27:30 <shafire> then I can use it from tmpfs :-) finally
2013-10-21 15:51:33 <barthalion> shafire: I'll take a look what can be done
2013-10-21 16:13:49 <shafire> barthalion: there will be a new release today
2013-10-21 16:14:07 <shafire> today = us time
2013-10-21 16:18:56 <barthalion> I don't think it makes any difference to me :p
2013-10-21 16:41:08 <barthalion> hrm
2013-10-21 20:10:39 <ncopa_> algitbot: build master
2013-10-22 05:09:14 <fabled> morning
2013-10-22 05:09:34 <StarWarsFan|afk> moin
2013-10-22 06:04:03 <barthalion> algitbot: build master
2013-10-22 07:47:45 <fabian_a> please git pull git://git.alpinelinux.org/fab/aports
2013-10-22 07:50:36 <fabian_a> hmmm, the same stuff as a couple of days ago...updates, new aports(python bindings, setuptools und bottle extensions)
2013-10-22 07:51:44 <fabian_a> thanks for pulling it before it's getting messy
2013-10-22 07:52:01 <ncopa> i'm on it
2013-10-22 08:11:14 <ncopa> algitbot: build master
2013-10-22 08:12:12 <fabled> bld2 rebooting
2013-10-22 08:20:02 <fabled> algitbot: build master
2013-10-22 08:22:59 <ncopa> fabian_a: does this look correct: http://sprunge.us/EOOB
2013-10-22 08:33:32 <fabian_a> ncopa: i don't think so...py-hackernews needs BeautifulSoup4>=4.3.1
2013-10-22 08:36:30 <ncopa> there is no py-beatfulsoup4 packae
2013-10-22 08:36:53 <fabian_a> ncopa: i though that i introdued py-beautifulsoup4 awhile ago
2013-10-22 08:37:41 <ncopa> cannot find it...
2013-10-22 08:37:53 <ncopa> py-beautifulsoup is there
2013-10-22 08:38:36 <fabian_a> ncopa: perhaps i only dreamed about that i did it
2013-10-22 08:38:48 <ncopa> seems so :)
2013-10-22 08:39:05 <ncopa> probably just forgot to commit it
2013-10-22 08:40:16 <fabian_a> yepp, it's in another branch
2013-10-22 08:41:18 <ncopa>  git format-patch -1 --stdout <commit-id> | sprunge
2013-10-22 08:41:21 <ncopa> and paste the url here
2013-10-22 08:41:39 <ncopa> and i'll: curl <url> | git am
2013-10-22 08:45:21 <fabian_a> http://sprunge.us/WNZh
2013-10-22 08:47:02 <ncopa> thanks
2013-10-22 08:48:14 <ncopa> i wonder if we should fix the irc messages from git push
2013-10-22 08:49:16 <ncopa> aports:master | <user> pushed N commits | stats:<link to git stats> | log:<url to git log>
2013-10-22 08:49:29 <ncopa> or similar
2013-10-22 09:24:49 <ncopa> bld* boxes has ipv6 and for some reason curl's call getaddrinfo() pick the ipv6 address
2013-10-22 09:24:56 <ncopa> but fails due to routing is not set up
2013-10-22 09:25:19 <ncopa> whats interesting is that google.com has both AAAA and A records
2013-10-22 09:25:25 <ncopa> and it picks the A record there
2013-10-22 09:26:01 <ncopa> i have worked around it by adding ipv4 to ~/.curlrc on the builders
2013-10-22 09:26:22 <ncopa> aha
2013-10-22 09:26:29 <ncopa> it gets 404
2013-10-22 09:26:33 <ncopa> on A record
2013-10-22 09:26:37 <ncopa> thats the real error
2013-10-22 09:26:45 <ncopa> then it tries ipv6
2013-10-22 10:25:15 <fabled> hrmpf
2013-10-22 11:11:51 <fabian_a> please git pull git://git.alpinelinux.org/fab/aports
2013-10-22 11:12:37 <fabian_a> this request includes the pyramid web framework and its deps
2013-10-22 11:13:20 <fabian_a> now all my branches are merged aka tabula rasa. thanks for pulling
2013-10-22 13:03:39 <ncopa> re flashcache
2013-10-22 13:03:49 <ncopa> it appears that bcache is in mainline kernel
2013-10-22 13:04:09 <ncopa> i wonder if we could go for that instead of flashcache
2013-10-22 13:19:26 <ncopa> \o/ it built edge/main
2013-10-22 13:32:53 <ncopa> \o/
2013-10-22 13:32:58 <ncopa> algitbot: build master
2013-10-22 14:06:55 <ncopa> algitbot: build master
2013-10-22 14:53:54 <barthalion> ncopa: excellent!
2013-10-22 15:27:06 <barthalion> hrm, lxpanel is broken for some reason
2013-10-22 15:28:44 <barthalion> and apparently whole lxde session starts only once
2013-10-22 15:32:00 <barthalion> but pcmanfm works, so I don't care enough
2013-10-22 15:32:44 <clandmeter> nl.a.o has issues
2013-10-22 15:33:11 <barthalion> clandmeter: define issues
2013-10-22 15:33:27 <clandmeter> its down
2013-10-22 15:33:42 <barthalion> uh, it was up 3 minutes ago
2013-10-22 15:34:17 <clandmeter> lets assume im right
2013-10-22 15:36:35 <barthalion> wtf
2013-10-22 15:36:56 <barthalion> clandmeter: buildbots went crazy too
2013-10-22 15:37:07 <clandmeter> yeah
2013-10-22 15:37:09 <clandmeter> seems like it
2013-10-22 15:37:23 <clandmeter> kvm guest doesnt want to boot
2013-10-22 15:38:49 <clandmeter> somekind of weird errror
2013-10-22 15:39:09 <barthalion> nl.ao.org is up
2013-10-22 15:39:15 <barthalion> or not
2013-10-22 15:39:20 <barthalion> well, it responds on ping
2013-10-22 15:39:35 <jbilyk_> hmm - I can't get on build-2-6 or 2-7 on bld1
2013-10-22 15:39:40 <jbilyk_> lxc says it's stopped
2013-10-22 15:40:08 <barthalion> algitbot: silence
2013-10-22 15:40:16 <barthalion> algitbot: build master
2013-10-22 15:40:29 <barthalion> sigh
2013-10-22 15:40:46 <barthalion> ncopa: blue tuesday now :p
2013-10-22 15:40:59 <clandmeter> ah its running
2013-10-22 15:41:12 <jbilyk_> 
ACTION rebooting build-edge-x86_64

2013-10-22 15:41:23 <jbilyk_> if that fixes the issue, I'll reboot the other guests too
2013-10-22 15:41:25 <clandmeter> jbilyk_: sure you need to ?
2013-10-22 15:41:39 <jbilyk_> clandmeter: no, just an easy way to shut it up :)
2013-10-22 15:41:40 <clandmeter> which issue you uave?
2013-10-22 15:41:44 <clandmeter> wait
2013-10-22 15:41:49 <clandmeter> let nl.a.o boot
2013-10-22 15:42:05 <clandmeter> i think it tries to rsync which i cannot
2013-10-22 15:42:28 <clandmeter> looks like its silent now
2013-10-22 15:42:38 <jbilyk_> bld2 can connect port 22 to nl.a.o now
2013-10-22 15:42:41 <jbilyk_> clandmeter: thx for fixing it
2013-10-22 15:43:06 <jbilyk_> 
ACTION now knows how to connect to an lxc guest console at least :)

2013-10-22 15:43:20 <clandmeter> kvm suspend from ram is bugger
2013-10-22 16:44:19 <barthalion> algitbot: test
2013-10-22 16:44:21 <barthalion> algitbot: ping
2013-10-22 16:44:26 <barthalion> algitbot: build master
2013-10-22 16:44:37 <barthalion> I want to see if it's still bugging
2013-10-22 17:19:49 <barthalion> so it doesn't build at all right now
2013-10-22 17:21:26 <clandmeter> barthalion: you broke it?
2013-10-22 17:21:35 <barthalion> if I knew
2013-10-22 17:21:37 <barthalion> no idea
2013-10-22 17:21:58 <barthalion> maybe it doesn't like when multiple packages are modified in one commit
2013-10-22 17:22:27 <clandmeter> nah i dont think so
2013-10-22 17:22:39 <barthalion> let's see
2013-10-22 17:22:40 <clandmeter> isnt it because of nl.a.o went down?
2013-10-22 17:22:56 <barthalion> ah, so it ignored my build master
2013-10-22 17:23:06 <barthalion> should be fine then
2013-10-22 17:23:16 <barthalion> where is build-edge-x86_64?
2013-10-22 17:24:59 <barthalion> algitbot: build 2.6-stable
2013-10-22 17:25:14 <barthalion> so 2.6 builders are missing too
2013-10-22 17:39:58 <jbilyk_> algitbot: build edge
2013-10-22 17:40:14 <barthalion> algitbot: build master
2013-10-22 17:40:20 <barthalion> algitbot: build 2.6-stable
2013-10-22 17:40:25 <barthalion> jbilyk_: ↑
2013-10-22 17:40:55 <jbilyk_> barthalion: I'm not really sure how fabled and ncopa have tweaked the build bots - used to be that we just rebooted the vservers when they went crazy
2013-10-22 17:41:29 <barthalion> well, I don't know either
2013-10-22 17:41:40 <barthalion> guess that they will fix it tommorow, there is no hurry
2013-10-22 17:57:28 <clandmeter> access ttyUSB via kvm -> vserver -> guest doesnt seem to work :|
2013-10-22 19:22:02 <barthalion> hrm, pre-commit script hasn't found that
2013-10-22 19:23:07 <barthalion> uhm, yeah, I see why
2013-10-22 19:25:55 <barthalion> wee
2013-10-22 21:16:34 <StarWarsFan> n8@all
2013-10-23 04:55:41 <fabled> morning
2013-10-23 04:58:34 <StarWarsFan> Moinmoin
2013-10-23 06:00:49 <ncopa> morning
2013-10-23 07:01:13 <shafire> hi
2013-10-23 07:14:23 <ncopa> hi
2013-10-23 07:51:49 <ncopa> shafire: was it you who are interested in flashcache?
2013-10-23 07:52:08 <shafire> yes
2013-10-23 07:52:19 <shafire> and darkfader
2013-10-23 07:52:28 <ncopa> have you seen bcache?
2013-10-23 07:52:50 <ncopa> http://bcache.evilpiepirate.org/
2013-10-23 07:53:03 <ncopa> http://www.accelcloud.com/2012/04/18/linux-flashcache-and-bcache-performance-testing/
2013-10-23 07:53:21 <ncopa> bcache is in mainline kernel
2013-10-23 07:53:31 <shafire> I studied both
2013-10-23 07:53:35 <shafire> or three
2013-10-23 07:53:42 <shafire> but at the end I choiced flashcache
2013-10-23 07:54:11 <ncopa> why?
2013-10-23 07:54:21 <ncopa> would you be ok if we only do bcache?
2013-10-23 07:54:56 <shafire> :-(
2013-10-23 07:55:29 <ncopa> it would save maintenence work
2013-10-23 07:55:32 <ncopa> i think..
2013-10-23 07:55:49 <ncopa> i can add flashcache if you have strong feelings for it
2013-10-23 07:56:03 <ncopa> i just want to know if it is worth it
2013-10-23 07:56:15 <ncopa> or if bcache is "good enough" even if its not the "best"
2013-10-23 07:57:26 <shafire> When I checked bcache, it was experimental and some people lost data
2013-10-23 07:57:48 <shafire> this was some months ago, when you pushed flashcache into testing
2013-10-23 07:58:14 <shafire> I will ask darkfader, he wanted flashcache too
2013-10-23 07:59:06 <shafire> Or I will maintain it :-)
2013-10-23 08:08:48 <ncopa> bcache we kinda get for free
2013-10-23 08:16:01 <shafire> bcache has one weakness
2013-10-23 08:16:19 <shafire> "data partition has to be prepared i.e. "formatted" beforehand. This invalidates bcache as viable solution for plugging to existing file systems with data."
2013-10-23 08:16:46 <shafire> That's why I did not choose bcache in first place
2013-10-23 08:35:52 <vkrishn> dome more benchmarks here: http://device-mapper.com/blog/2013/04/13/benchmarking-dm-cache-with-fio/
2013-10-23 08:36:16 <vkrishn> I think there were some done by phoronix too
2013-10-23 09:10:06 <ncopa> algitbot: build master
2013-10-23 09:12:41 <barthalion> algitbot: build 2.6-stable
2013-10-23 09:12:51 <barthalion> ncopa: ↑
2013-10-23 09:14:38 <ncopa> algitbot: build master
2013-10-23 09:14:55 <ncopa> algitbot: build 2.6-stable
2013-10-23 09:15:31 <ncopa> barthalion: i think its the zeromq silent disconnect problem that caused it
2013-10-23 09:16:18 <ncopa> described here: http://lucumr.pocoo.org/2012/6/26/disconnects-are-good-for-you/
2013-10-23 09:17:20 <barthalion> ah
2013-10-23 09:17:29 <barthalion> good to know
2013-10-23 09:31:25 <shafire> ncopa: I like to stay with flashcache :S
2013-10-23 09:32:58 <ncopa> shafire: have you tested it from edge/testing?
2013-10-23 09:33:03 <shafire> yes
2013-10-23 09:33:07 <ncopa> does it work?
2013-10-23 09:33:07 <shafire> protocol is in bugs.a.o
2013-10-23 09:33:14 <ncopa> ok good thanks
2013-10-23 09:33:53 <ncopa> super
2013-10-23 09:33:57 <ncopa> will move it to main then
2013-10-23 09:34:06 <shafire> :-)
2013-10-23 09:34:18 <ncopa> do you need it from tmpfs setups?
2013-10-23 09:34:35 <ncopa> on the boot cdrom/usb?
2013-10-23 09:35:30 <shafire> I need it for tmpfs setups
2013-10-23 09:35:37 <shafire> but not at boot (at moment)
2013-10-23 09:35:59 <ncopa> it will need to be added to the modloop then
2013-10-23 09:36:14 <ncopa> unless you build your own boot iso
2013-10-23 09:37:14 <shafire> heh, never did that
2013-10-23 09:37:42 <ncopa> ok
2013-10-23 09:39:10 <ncopa> atleast its in main now
2013-10-23 09:39:23 <shafire> :-)
2013-10-23 09:40:02 <ncopa> personally i think flashcache is overrated
2013-10-23 09:40:18 <ncopa> i'd rater have more RAM than an intermediate ssd
2013-10-23 09:40:40 <shafire> my setup can't take more ram
2013-10-23 09:40:45 <shafire> I got a pci-e ssd
2013-10-23 09:40:54 <shafire> 800mb/s read/write
2013-10-23 09:45:45 <shafire> are you using ssds?
2013-10-23 10:30:45 <clandmeter> did i break anything?
2013-10-23 10:34:34 <clandmeter> ncopa: when you are back, could you give me a hand with kvm?
2013-10-23 10:49:40 <build-2-4> files from v2.4.11-111-g0c896b5 uploaded
2013-10-23 10:54:23 <build64-2-4> files from v2.4.11-111-g0c896b5 uploaded
2013-10-23 10:56:57 <ncopa> hi
2013-10-23 11:02:31 <clandmeter> ncopa: remember we had network issues on kvm?
2013-10-23 11:02:43 <clandmeter> i had to reboot one the guests, and now i got that issue again
2013-10-23 11:02:46 <ncopa> one host not talking with the other?
2013-10-23 11:02:52 <clandmeter> correct
2013-10-23 11:03:08 <ncopa> it was some stupid firewall rules iirc
2013-10-23 11:03:09 <clandmeter> iptables order
2013-10-23 11:03:22 <clandmeter> but i didnt know how you solved that :)
2013-10-23 11:03:44 <ncopa> i dont remember how i did it either
2013-10-23 11:03:54 <ncopa> i think i moved one of the rules up a bit
2013-10-23 11:04:04 <clandmeter> yep
2013-10-23 11:04:21 <ncopa> i can have a look in a bit
2013-10-23 11:04:26 <clandmeter> ok thx
2013-10-23 11:04:40 <clandmeter> we can add a note, so i wont ask you again :)
2013-10-23 11:09:44 <ncopa> ok back
2013-10-23 11:10:01 <ncopa> i think i even had an idea how to solve it upstream
2013-10-23 11:10:09 <ncopa> just didnt have time to follow it up
2013-10-23 11:12:10 <ncopa> heh
2013-10-23 11:12:30 <ncopa> i think the output of iptables-save is easier to read than iptables -L
2013-10-23 11:13:40 <ncopa> clandmeter: those are relevant: https://dpaste.de/b5Li
2013-10-23 11:13:53 <clandmeter> checking
2013-10-23 11:15:11 <ncopa> ah
2013-10-23 11:15:15 <clandmeter> ok, so which one i need to move?
2013-10-23 11:15:18 <ncopa> i think the last 2 lines
2013-10-23 11:15:20 <clandmeter> and how to move them :)
2013-10-23 11:15:34 <ncopa> i think they are the ones we added
2013-10-23 11:15:42 <ncopa> that are supposed to do what we want
2013-10-23 11:15:55 <ncopa> but i think if you restart libvirt it will not reinsert those
2013-10-23 11:16:17 <clandmeter> i restarted kvm completely
2013-10-23 11:16:18 <ncopa> i think the 2 last lines needs to go up abover the REJECT lines
2013-10-23 11:16:43 <clandmeter> so the current state is after kvm reboot
2013-10-23 11:17:05 <ncopa> then i think its related the order the services are started
2013-10-23 11:17:35 <ncopa> yeah
2013-10-23 11:18:09 <ncopa> we do it with the service 'local'
2013-10-23 11:18:44 <ncopa> it does not work as epxected
2013-10-23 11:19:02 <ncopa> the script is in /etc/local.d/
2013-10-23 11:19:21 <clandmeter> need to create an additional init with proper before/after?
2013-10-23 11:19:35 <ncopa> that was first thought too
2013-10-23 11:19:36 <ncopa> but i think what happens
2013-10-23 11:19:38 <ncopa> is
2013-10-23 11:19:48 <ncopa> the boot order looks correct
2013-10-23 11:19:58 <ncopa> 'local' starts after libvirtd
2013-10-23 11:20:19 <ncopa> but libvirtd starts up the guests a bit afterwards
2013-10-23 11:20:38 <ncopa> or the guest networks
2013-10-23 11:21:03 <clandmeter> yes probably
2013-10-23 11:21:23 <ncopa> a quick and dirty way to 'fix' it might be simply restart 'local'
2013-10-23 11:22:01 <clandmeter> we can create ascript which waits?
2013-10-23 11:22:14 <ncopa> that too, but its a bit ugly
2013-10-23 11:22:30 <ncopa> restarting libvirt will also break it
2013-10-23 11:22:42 <clandmeter> i did that, and it wasnt working
2013-10-23 11:22:53 <clandmeter> not libivrt but one guest
2013-10-23 11:23:19 <ncopa> if you restart the networks it will break too
2013-10-23 11:23:38 <ncopa> did you try restart local?
2013-10-23 11:23:44 <clandmeter> no not yet
2013-10-23 11:23:45 <clandmeter> should i?
2013-10-23 11:24:01 <ncopa> it will insert the magic lines at the top
2013-10-23 11:24:04 <ncopa> hmm
2013-10-23 11:24:23 <ncopa> but it will not delete the lines at the bottom
2013-10-23 11:24:49 <ncopa> other way to workaround it is iptables-save > file
2013-10-23 11:24:55 <clandmeter> its not possible to add something unique to rule so you can update it?
2013-10-23 11:24:57 <ncopa> move the lines a bit up
2013-10-23 11:25:22 <ncopa> and then iptables-restore < file
2013-10-23 11:27:28 <clandmeter> how can i move them up?
2013-10-23 11:27:38 <ncopa> i did it for you
2013-10-23 11:27:50 <ncopa> it can now be done with /etc/init.d/local restart
2013-10-23 11:28:18 <clandmeter> ok, so if it doesnt work just restart local?
2013-10-23 11:28:32 <ncopa> correct
2013-10-23 11:28:43 <ncopa> and i think that is needed for every reboot for now :-(
2013-10-23 11:28:52 <clandmeter> ok i will add a note.
2013-10-23 11:29:06 <ncopa> otherway you can do it: iptables-save > fw
2013-10-23 11:29:26 <ncopa> edit the file fw manually, moving the lines -A FORWARD -i virbr1 -o virbr0 -j ACCEPT
2013-10-23 11:29:26 <ncopa> -A FORWARD -i virbr0 -o virbr1 -j ACCEPT
2013-10-23 11:29:37 <ncopa> to be the first of the -A FORWARD lines
2013-10-23 11:29:39 <ncopa> like they are now
2013-10-23 11:29:45 <ncopa> and save file
2013-10-23 11:29:47 <clandmeter> ah ok
2013-10-23 11:29:53 <ncopa> and then: iptables-restore < fw
2013-10-23 11:31:00 <clandmeter> thx
2013-10-23 11:32:16 <shafire> I am using ferm for iptables
2013-10-23 11:33:07 <ncopa> ferm wouldnt help
2013-10-23 11:33:20 <ncopa> its libvirt that inserts various rules
2013-10-23 11:34:10 <ncopa> it does -I for insert
2013-10-23 11:34:23 <ncopa> so it inserts its rules before already existing
2013-10-23 11:34:53 <ncopa> and the rules it inserts is kinda stupid
2013-10-23 11:36:44 <shafire> ah ok
2013-10-23 11:43:16 <ncopa> clandmeter: i think i might have an idea how to fix it
2013-10-23 11:43:49 <clandmeter> ok
2013-10-23 11:44:13 <ncopa> well a simple way to fix it upstream
2013-10-23 11:44:26 <ncopa> basically, i think the do iptables -I ... all the way
2013-10-23 11:45:57 <clandmeter> I dont know much about iptables.
2013-10-23 11:52:31 <ncopa> but i think it shoudl work if they did -I ... for everything except the final rejects, which should be -A
2013-10-23 11:52:38 <ncopa> -i means insert (inserts rule first)
2013-10-23 11:52:49 <ncopa> and -A is 'add' add at bottom
2013-10-23 11:53:13 <ncopa> they use -I to make sure that their rules wins over anything that the sysadmin might have set up
2013-10-23 12:04:11 <ncopa> sigh, they have made iptables a bit cumbersome
2013-10-23 12:10:44 <clandmeter> and i added etho entries :|
2013-10-23 12:10:53 <clandmeter> no wonder they dont forward correctly
2013-10-23 12:14:34 <clandmeter> ncopa: i need to rebuild the rules now (based on fixed firewalltab config)
2013-10-23 12:14:44 <clandmeter> should i just reboot my vserver and restart local?
2013-10-23 12:15:14 <ncopa> i don thinkt you need restart local
2013-10-23 12:15:25 <ncopa> i think you can just restart the vserver
2013-10-23 12:15:36 <ncopa> i think problem only happens when you restart a network
2013-10-23 12:16:26 <clandmeter> brb
2013-10-24 10:25:16 <clandmeter> fcolista: ping
2013-10-24 12:20:17 <fcolista> pong clandmeter
2013-10-24 12:23:11 <clandmeter> hi fcolista
2013-10-24 12:23:23 <clandmeter> you still use tvheadned-git?
2013-10-24 12:23:33 <fcolista> yes
2013-10-24 12:23:50 <fcolista> it works
2013-10-24 12:24:01 <clandmeter> you use the git version?
2013-10-24 12:24:13 <fcolista> that's a good question
2013-10-24 12:24:31 <fcolista> i dont' remember what i've installed
2013-10-24 12:24:39 <fcolista> since it was few months ago
2013-10-24 12:24:44 <fcolista> i can check this evening
2013-10-24 12:24:46 <clandmeter> i want to push tvheadend-next, it has some experimental stuff
2013-10-24 12:24:47 <fcolista> and let you know
2013-10-24 12:25:01 <fcolista> it's ok for me
2013-10-24 12:25:04 <clandmeter> tvheadend in main is newer then -git
2013-10-24 12:25:24 <clandmeter> but -next is not compatble with tvheadend
2013-10-24 12:27:14 <fcolista> go ahead, don't worry for me
2013-10-24 12:27:31 <clandmeter> fcolista: ok thanks
2013-10-24 12:27:40 <clandmeter> i dont want to have too many tvheadned around.
2013-10-24 12:27:48 <fcolista> agree
2013-10-24 12:38:02 <ncopa> algitbot: build 2.6-stable
2013-10-24 12:40:53 <ncopa> algitbot: build 2.6-stable
2013-10-24 12:43:14 <ncopa> algitbot: build 2.6-stable
2013-10-24 12:57:46 <ncopa> is the english text ok? http://alpinelinux.org/release-2.6.6
2013-10-24 13:03:31 <StarWarsFan> ncopa: for me ok, found nothing to complain about :)
2013-10-24 13:04:04 <ncopa> i sent email
2013-10-24 13:04:05 <ncopa> thanks
2013-10-24 13:04:27 <ncopa> i will upgrade the new build servers
2013-10-24 13:04:34 <ncopa> new kernel
2013-10-24 13:05:07 <grimeton> http://i.imgur.com/YSUFNNL.png
2013-10-24 13:05:10 <grimeton> oh nvm
2013-10-24 13:39:39 <ncopa> i'm rebooting bld1 (build-edge-x86, build-2-6-x86 ...)
2013-10-24 13:47:58 <ncopa> and now i'm rebootin bld2.a.o
2013-10-24 13:48:07 <ncopa> build-edge-x86_64 etc
2013-10-24 14:12:41 <ChMuRi> hi there
2013-10-24 14:12:43 <ChMuRi> :)
2013-10-24 14:13:43 <ncopa> hi
2013-10-24 14:15:56 <ncopa> hm
2013-10-24 14:16:04 <ncopa> builders does not connect
2013-10-24 14:19:20 <ncopa> weird networkign in the lxc's does not start up properly
2013-10-24 14:19:53 <StarWarsFan> sounds like my problem some days ago...
2013-10-24 14:22:55 <ChMuRi> ncopa, are you from dev team alipne?
2013-10-24 14:53:55 <ncopa> ChMuRi: yes
2013-10-24 18:28:39 <barthalion> ChMuRi: do you have any issues?
2013-10-24 18:29:33 <ChMuRi> barthalion, have some qiestion on alpine-linux chanell have answers abour mirror and translaation of distro alpine
2013-10-24 18:31:12 <barthalion> if you want to host a mirror, there is an article about it on our wiki
2013-10-24 18:32:02 <barthalion> and about translations… I don't think we support anything besides English
2013-10-24 18:32:15 <ChMuRi> yes i see that
2013-10-24 18:32:16 <barthalion> but I'm not sure how it's supposed to work with uclibc
2013-10-24 18:32:32 <jbilyk_> uclibc and locales has been pretty iffy
2013-10-24 18:32:43 <barthalion> ncopa: have you taken a look at my linux-rt package?
2013-10-25 06:04:11 <fabled> ok - musl/x86_64 is bootstrapped, and full build launched. i'll upload the packages this weekend or on Monday when they are built.
2013-10-25 09:08:36 <ncopa> algitbot: build master
2013-10-25 09:29:56 <ncopa> algitbot: build master
2013-10-25 09:47:57 <Niichan> ncopa: how well does Wine work in Alpine?
2013-10-25 09:48:59 <ncopa> Niichan: it works well afaik
2013-10-25 09:49:13 <ncopa> i havent done too much gaming with it
2013-10-25 09:49:19 <ncopa> but it works
2013-10-25 09:49:38 <Niichan> Do proprietary ati drivers work in Alpine?
2013-10-25 09:49:44 <ncopa> nope
2013-10-25 09:49:53 <Niichan> because glibc?
2013-10-25 09:50:06 <ncopa> that and our hardened kernel
2013-10-25 09:50:18 <ncopa> oh wine does not work on x86_64
2013-10-25 09:50:30 <Niichan> ah
2013-10-25 09:50:39 <ncopa> due to no support for multilib
2013-10-25 12:04:03 <ncopa> lol
2013-10-25 12:04:15 <ncopa> so
2013-10-25 12:04:28 <ncopa> i am looking at syslinux-6.02
2013-10-25 12:04:37 <ncopa> i look how gentoo does it
2013-10-25 12:04:46 <ncopa> i see this
2013-10-25 12:04:51 <ncopa> src_prepare() {
2013-10-25 12:04:51 <ncopa>         rm -f gethostip #bug 137081
2013-10-25 12:04:58 <ncopa> so i wonder what is that about
2013-10-25 12:05:07 <ncopa> i dig up gentoo bug 13701
2013-10-25 12:05:21 <ncopa> guess who filed that bug in the first place?
2013-10-25 12:05:24 <ncopa> https://bugs.gentoo.org/show_bug.cgi?id=137081
2013-10-25 13:56:29 <ncopa> algitbot: build master
2013-10-25 13:56:56 <ncopa> ok
2013-10-25 13:57:03 <ncopa> its up and running
2013-10-25 14:18:56 <ChMuRi> how the auto build work?
2013-10-25 14:28:17 <ncopa> algitbot: build master
2013-10-25 14:28:25 <ncopa> ChMuRi: git hook
2013-10-25 14:28:35 <ncopa> sends a message to build servers
2013-10-25 14:28:48 <ChMuRi> ncopa, build server could be on vps?
2013-10-25 14:29:26 <ncopa> yup
2013-10-25 14:29:31 <ncopa> we use LXC for it
2013-10-25 14:30:01 <ncopa> and we do the hosting oursleves
2013-10-25 14:30:08 <ncopa> used to run on vserver
2013-10-25 14:30:28 <ncopa> message passing happens with lua scripts and zmq
2013-10-25 14:36:15 <ncopa> algitbot: build master
2013-10-25 14:37:10 <ncopa> algitbot: build master
2013-10-25 14:38:18 <ChMuRi> ncopa,It is sufficient that build servers?
2013-10-25 14:39:30 <ncopa> we got new hw last week or so
2013-10-25 14:39:47 <ncopa> so they are good
2013-10-25 14:40:10 <ncopa> we used to have a single box for both x86 and x86_64
2013-10-25 14:40:15 <ncopa> we got 2 new boxes
2013-10-25 14:40:31 <ncopa> so now we have one for x86 and another fo x86_64
2013-10-25 14:40:54 <ncopa> and they are pretty fast...
2013-10-25 14:41:09 <ChMuRi> oh ok
2013-10-25 22:14:01 <StarWarsFan> fabled: you're there?
2013-10-25 22:14:09 <StarWarsFan> evening btw. :-)
2013-10-25 22:21:14 <StarWarsFan> are you able to create a real iso of you rpi-image?
2013-10-25 22:22:03 <StarWarsFan> the one you mentioned on your mail ist only a tar.gz
2013-10-26 20:39:09 <Frosh> Did something new happen? that after pygrub it pauses the domu? I thought it froze
2013-10-26 20:40:00 <Frosh> last time I booted it, it didn't pause so I was just curious
2013-10-26 21:52:27 <Frosh> i think I was pressing enter
2013-10-26 21:52:30 <Frosh> nevermind
2013-10-27 03:14:32 <Frosh> something is odd, it seems to hang in the s state
2013-10-27 03:14:46 <Frosh> dunno if this is related
2013-10-27 06:31:06 <kaniini> s means it's 'shutting down'
2013-10-27 06:31:55 <kaniini> StarWarsFan: on ARM, you do not use ISOs to load the OS.  instead, you provide a system image.
2013-10-27 06:32:24 <Niichan> StarWarsFan: If you want I can guide you through installing Alpine to the SD card from the tgz
2013-10-27 06:32:51 <Niichan> Hell I should make a markdown install guide lol
2013-10-27 06:39:04 <Niichan> It's pretty much like the normal manual install process unless you are insane like me and do it headless
2013-10-27 06:39:33 <Niichan> I bet the process of remotely installing Alpine on a non-alpine server is similar
2013-10-27 06:39:35 <Niichan> hmmm
2013-10-27 06:39:46 <Niichan> I should try that
2013-10-27 06:39:57 <Niichan> I have the extra VPS that I can burn over and over lol
2013-10-27 08:01:23 <StarWarsFan> moin
2013-10-27 08:01:46 <StarWarsFan> kaniini: Niichan: thx but it's working
2013-10-27 08:01:50 <StarWarsFan> :-)
2013-10-27 08:01:53 <StarWarsFan> another question:
2013-10-27 08:02:10 <StarWarsFan> how do i "change" to the installed system after first boot?
2013-10-27 08:02:31 <StarWarsFan> on reboot the rpi always "reinstalls" the system
2013-10-27 08:02:59 <Niichan> kernel command line
2013-10-27 08:03:08 <StarWarsFan> I've setup an SD card with a 2G FAT partition
2013-10-27 08:03:09 <Niichan> make sure to have the command line load the ext2 module
2013-10-27 08:03:12 <Niichan> oh
2013-10-27 08:03:18 <Niichan> I set mine up as a full system install
2013-10-27 08:03:22 <StarWarsFan> copied the content of the tgz on it and boot the rpi from it
2013-10-27 08:04:15 <StarWarsFan> Niichan: what do you mean?
2013-10-27 08:04:19 <StarWarsFan> what should i do?
2013-10-27 08:05:09 <StarWarsFan> i used the image from http://dev.alpinelinux.org/~tteras/rpi-with-networking.tar
2013-10-27 08:06:10 <StarWarsFan> ok, rpi finished booting
2013-10-27 08:06:21 <StarWarsFan> so what to do now?
2013-10-27 08:06:33 <StarWarsFan> install system? and then?
2013-10-28 03:26:42 <Mp5shooter> arrrrrrgh
2013-10-28 03:26:53 <Mp5shooter> can't figure out why libreswan's pluto is segfaulting
2013-10-28 03:26:55 <Mp5shooter> and gdb isn't helping
2013-10-28 06:24:12 <fabled> StarWarsFan, you don't want iso from the rpi-image.
2013-10-28 06:24:21 <fabled> just stuff the .tar.gz to mmc/sd
2013-10-28 06:24:26 <fabled> extract*
2013-10-28 06:24:34 <fabled> rpi boot sequence is quite different from x86
2013-10-28 06:44:13 <barthalion> fabled: I'm on it
2013-10-28 06:44:26 <barthalion> we should start mirroring live-media sources
2013-10-28 06:44:30 <fabled> yes
2013-10-28 06:44:38 <fabled> it's annoying, they keep only the latest there.
2013-10-28 07:15:37 <ncopa> morning
2013-10-28 07:15:45 <ncopa> i think we will have to mirror all sources :-/
2013-10-28 07:38:16 <ncopa> algitbot: build master
2013-10-28 08:10:33 <barthalion> ncopa: no soname bump?
2013-10-28 08:10:59 <ncopa> barthalion: not that i coudl see
2013-10-28 08:11:04 <barthalion> nice
2013-10-28 08:11:12 <barthalion> that's something new
2013-10-28 08:12:10 <ncopa> https://dpaste.de/yewN
2013-10-28 08:12:11 <ncopa> yes
2013-10-28 08:12:28 <ncopa> they did 2.0 not too long ago
2013-10-28 08:12:45 <ncopa> and maybe they have discovered the value ov ABI stability
2013-10-28 09:44:58 <StarWarsFan> re
2013-10-28 09:45:01 <StarWarsFan> moin
2013-10-28 09:45:20 <StarWarsFan> fabled: yes, figured this out shortly after my question :-)
2013-10-28 09:46:33 <StarWarsFan> but did you see my other question?
2013-10-28 09:46:50 <StarWarsFan> yesterday 9:01:53
2013-10-28 10:00:46 <fabled> StarWarsFan, no
2013-10-28 10:00:58 <StarWarsFan> np, just a second:
2013-10-28 10:01:27 <StarWarsFan> <StarWarsFan> how do i "change" to the installed system after first boot?
2013-10-28 10:01:27 <StarWarsFan> <StarWarsFan> I've setup an SD card with a 2G FAT partition
2013-10-28 10:01:27 <StarWarsFan> <Niichan> make sure to have the command line load the ext2 module
2013-10-28 10:01:27 <StarWarsFan> <StarWarsFan> on reboot the rpi always "reinstalls" the system
2013-10-28 10:01:27 <StarWarsFan> <Niichan> kernel command line
2013-10-28 10:01:28 <StarWarsFan> <Niichan> oh
2013-10-28 10:01:29 <StarWarsFan> <Niichan> I set mine up as a full system install
2013-10-28 10:01:31 <StarWarsFan> <StarWarsFan> copied the content of the tgz on it and boot the rpi from it
2013-10-28 10:01:34 <StarWarsFan> ...
2013-10-28 10:01:39 <StarWarsFan> <StarWarsFan> i used the image from http://dev.alpinelinux.org/~tteras/rpi-with-networking.tar
2013-10-28 10:01:39 <StarWarsFan> <StarWarsFan> ok, rpi finished booting
2013-10-28 10:01:40 <StarWarsFan> <StarWarsFan> so what to do now?
2013-10-28 10:01:42 <StarWarsFan> <StarWarsFan> install system? and then?
2013-10-28 10:02:27 <fabled> StarWarsFan, the images work like regular alpine images: 'lbu commit' and tmpfs root. but it all depends on what you need.
2013-10-28 10:02:32 <fabled> i'm off to lunch now
2013-10-28 10:03:03 <StarWarsFan> i'm just playing around to checkout rpi-usage
2013-10-28 10:03:38 <StarWarsFan> but: never used lbu functionality
2013-10-28 10:03:50 <StarWarsFan> so seems i should checkout this now... ;-)
2013-10-28 11:00:31 <fabled> StarWarsFan, if you plan to install lot of stuff, you might want to do system install on the mmc card partition
2013-10-28 11:00:52 <fabled> i think Niichan was working on image like that
2013-10-28 12:08:05 <StarWarsFan> fabled: that's what i did
2013-10-28 12:08:33 <fabled> oh, ok.
2013-10-28 12:08:41 <StarWarsFan> but after reboot, it always reinstall the system by booting the installer
2013-10-28 12:08:48 <StarWarsFan> so that's the reason of my question
2013-10-28 12:09:25 <StarWarsFan> - copied content to fat-formated sd-card
2013-10-28 12:09:28 <StarWarsFan> booted it
2013-10-28 12:09:32 <StarWarsFan> setup-alpine
2013-10-28 12:09:38 <StarWarsFan> choose mmc...
2013-10-28 12:09:53 <StarWarsFan> choose sys
2013-10-28 12:09:56 <StarWarsFan> reboot
2013-10-28 12:10:21 <clandmeter> if you run from memory you need to choose sys?
2013-10-28 12:10:23 <StarWarsFan> but then it's the same "virgin" system as before
2013-10-28 12:10:58 <StarWarsFan> clandmeter: i have no idea if this is necessary or not, i just did it...
2013-10-28 12:12:33 <clandmeter> thats only when you setup-disk
2013-10-28 12:12:45 <clandmeter> which will be run from setup-alpine if you want to install to disk
2013-10-28 12:13:54 <StarWarsFan> right, as i wrote above
2013-10-28 12:14:37 <clandmeter> i just tried it again and i dont get that option
2013-10-28 12:14:39 <Frosh> Morning!
2013-10-28 12:15:09 <clandmeter> StarWarsFan: why didnt you use http://dev.alpinelinux.org/~tteras/alpine-edge-131010-armhf.rpi.tar.gz ?
2013-10-28 12:15:49 <clandmeter> oh you *want* to use mmc as disk?
2013-10-28 12:17:10 <Frosh> ok, here are my symptoms of the xen alpine have guest os on lvm
2013-10-28 12:17:27 <Frosh> core2duo with 6gig ram
2013-10-28 12:18:12 <Frosh> after pygrub, guest os  seems to go into pause mode as shown on xl list
2013-10-28 12:18:57 <ncopa> Frosh: i think it displays a grub menu for 30 sec
2013-10-28 12:19:25 <Frosh> and on reboot show xl list in shutdown with s
2013-10-28 12:19:42 <Frosh> 30 seconds? I waited 10 minutes for both
2013-10-28 12:19:51 <ncopa> hm
2013-10-28 12:19:52 <Frosh> shutdown and starting up
2013-10-28 12:19:55 <ncopa> then its likely something else
2013-10-28 12:20:01 <ncopa> i dont know
2013-10-28 12:20:03 <Frosh> i had to do xl unpause 2
2013-10-28 12:20:15 <Frosh> this didn't happen or doesn't happen on my old xen
2013-10-28 12:20:27 <Frosh> on a slower machine
2013-10-28 12:20:50 <Frosh> any idea how to troubleshoot would be greatly appreciated
2013-10-28 12:21:27 <ncopa> i dont have idea sorry, but i would be happy if somebody find out whats going on
2013-10-28 12:21:32 <Frosh> at first I thought I was pressing enter at the pygrub which forces it to pause
2013-10-28 12:21:32 <ncopa> and help us fix it :-/
2013-10-28 12:21:51 <Frosh> but then I was sitting there patiencely for 10 minutes
2013-10-28 12:22:19 <Frosh> could it be that its under screen -unlikely, but I just trying to figure out what is causing the pause
2013-10-28 12:22:32 <Frosh> or even the hangups
2013-10-28 12:22:52 <Frosh> it seems to be bugging out even when its booted sometimes
2013-10-28 12:23:47 <Frosh> should I reinstall? only to find the same thing?
2013-10-28 12:24:42 <Frosh> the stuff I'm doing is no different then the stuff I'm doing with the (august release alpine)
2013-10-28 14:42:53 <fabled> ncopa, could we delete sfic ?
2013-10-28 14:42:59 <fabled> i think no packages uses it anymore
2013-10-28 16:38:36 <AmatCoder> hi, is forum down?
2013-10-28 16:40:04 <AmatCoder> also package browser...
2013-10-28 16:58:05 <BitL0G1c> I noticed this too
2013-10-28 18:30:15 <barthalion> oh yay, builders are alive
2013-10-28 19:11:50 <ncopa_> algitbot: build master
2013-10-28 19:19:40 <Frosh> ncopa_: should I reinstall? since I'm the only one to report this issue? or have other tried the new alpine?
2013-10-28 19:20:39 <ncopa_> Frosh: can you please remind me about your issue?
2013-10-28 19:21:10 <Frosh> when I do xl create -c debian.cfg
2013-10-28 19:21:10 <ncopa_> sweet, 2-7 x86_64 uploaded
2013-10-28 19:21:19 <Frosh> I see the pygrub with the countdown
2013-10-28 19:21:44 <Frosh> then when it hits 0 , I wait 10 minutes or so, then I goto another screen and do xl list
2013-10-28 19:21:50 <Frosh> and I see it paused
2013-10-28 19:22:01 <ncopa_> does it load the kernel?
2013-10-28 19:22:11 <Frosh> I do xl unpause 2 and it continues on its merry way
2013-10-28 19:22:34 <ncopa_> huh, you just unpause it and it works?
2013-10-28 19:22:39 <Frosh> just annoying to do xl unpause or xl destroy upon shutdown/reboot to bring it back up
2013-10-28 19:22:52 <Frosh> yes, i don't see why its pausing
2013-10-28 19:23:09 <ncopa_> i dont know why it happens
2013-10-28 19:23:09 <Frosh> and sometimes randomly (yes very unscientific) it goes and boots
2013-10-28 19:23:34 <ncopa_> alpine v2.7?
2013-10-28 19:23:37 <ncopa_> 2.6 i mean
2013-10-28 19:23:38 <Frosh> I don't think it loads the kernel
2013-10-28 19:23:41 <Frosh> this is 2.6
2013-10-28 19:23:54 <Frosh> should I try 2.7?
2013-10-28 19:24:03 <ncopa_> no release yet
2013-10-28 19:24:16 <ncopa_> but there might be a 2.7_rc1 tomorrow
2013-10-28 19:24:53 <Frosh> any other steps I can do to help troubleshoot this?
2013-10-28 19:25:00 <ncopa_> i dont know really
2013-10-28 19:25:06 <ncopa_> i dont use xen myself
2013-10-28 19:25:36 <Frosh> computer has more than enough ram, hdd is new
2013-10-28 19:25:47 <ncopa_> maybe ask in #xen channel
2013-10-28 19:26:13 <ncopa_> does dmesg give any hints=
2013-10-28 19:26:14 <ncopa_> ?
2013-10-28 19:26:24 <ncopa_> or log
2013-10-28 19:27:53 <Frosh> I'll get back to you on that...
2013-10-28 19:28:27 <Frosh> so 2.7 by this weekend?
2013-10-28 19:28:41 <ncopa_> tuesday next week
2013-10-28 19:28:56 <ncopa_> but it woudl be great if you could test the release candidate tomorrow
2013-10-28 19:29:21 <ncopa_> is there something called pvgrub?
2013-10-28 19:29:55 <ncopa_> i might wrong but i think ther is somethign called pvgrub which i think is suposed to replace pygrub
2013-10-28 19:30:13 <Frosh> if you put it out, tomorrow I be glad to test it out, I want a stable version. I just pick up a more powerful computer
2013-10-28 19:31:15 <Frosh> o really, pvgrub, http://wiki.xen.org/wiki/PvGrub is this coming to 2.7?
2013-10-28 19:33:23 <ncopa_> if someone creates a patch then yes
2013-10-28 19:34:30 <ncopa_> maybe kaniini knows
2013-10-28 19:54:09 <barthalion> Frosh: yeah, ask kaniini and royger
2013-10-28 20:05:14 <ncopa_> rebooting
2013-10-28 20:07:18 <ncopa_> seems like syslinux-6 didnt break my box completely
2013-10-28 20:07:22 <ncopa_> it boots at least
2013-10-28 21:33:39 <barthalion> ncopa: there is soname bump in new libass
2013-10-28 22:39:51 <Mp5shooter> does anyone know why pluto in libreswan keeps segfaulting?
2013-10-28 22:39:54 <Mp5shooter> gdb isn't helping
2013-10-28 22:42:23 <Mp5shooter> i think it has something to do with pax
2013-10-28 23:24:53 <BitL0G1c> running logcheck helps a lot to spot problems with grsecurity
2013-10-28 23:41:39 <tdtrask> ncopa: I just saw the note about rc1 tomorrow
2013-10-28 23:43:18 <tdtrask> if we can get lua-subprocess to work on 5.2, I might be able to move ACF to 5.2 by next week
2013-10-28 23:44:18 <tdtrask> what were the problems you saw?
2013-10-29 07:03:29 <ncopa> morning
2013-10-29 07:03:37 <ncopa> tdtrask: someone needs to port it to lua-5.2
2013-10-29 07:08:00 <ncopa> should have filed an isse about it... https://github.com/xlq/lua-subprocess/issues
2013-10-29 08:13:33 <barthalion> ncopa: http://sprunge.us/ZWTX not sure how to get rid of LUA_ENVIRONINDEX and luaL_register errors
2013-10-29 08:14:15 <ncopa> not sure either how to port that env stuff to lua 5.2
2013-10-29 08:21:23 <barthalion> hm, I think I know how to port luaL_register
2013-10-29 08:23:48 <barthalion> it won't be clean port, but half a loaf is better than none
2013-10-29 08:28:12 <ncopa> luaL_register can be replaced with setfuncs i think
2013-10-29 08:28:42 <ncopa> what i dont know how to do is the LUA_ENVIRONINDEX SP_LIST
2013-10-29 09:28:05 <barthalion> only LUA_ENVIRONINDEX errors left
2013-10-29 09:28:12 <barthalion> I'll take a look later, gtg
2013-10-29 09:33:39 <barthalion> ah, we may want to rebuild php as well
2013-10-29 09:33:42 <barthalion> http://bjauy.com/images/alpine-php54.png
2013-10-29 09:38:29 <barthalion> hrm, or maybe it's about modules dependencies
2013-10-29 09:39:36 <ncopa> barthalion: i think its module deps
2013-10-29 09:40:24 <ncopa> barthalion: re lua-subprocess, i think it need to be refactored
2013-10-29 09:40:41 <ncopa> restructured
2013-10-29 09:41:37 <barthalion> ncopa: I'll ask on lua ml, but I'm afraid that you're right
2013-10-29 09:49:26 <ncopa> i asked in #lua
2013-10-29 09:49:35 <ncopa> and ended up in a C++ vs C discussion :)
2013-10-29 09:55:27 <ncopa> hm
2013-10-29 09:55:33 <ncopa> whats up with build-edge-x86_64?
2013-10-29 09:55:56 <ncopa> algitbot: build master
2013-10-29 10:25:47 <ncopa> this is kinda interesting: http://www.phoronix.com/scan.php?page=news_item&px=MTQ5NzQ
2013-10-29 11:29:32 <tdtrask> ncopa: regarding lua-subprocess, I reached out to the developer but got no response
2013-10-29 11:29:44 <tdtrask> simply forgot to open an issue on github
2013-10-29 12:01:48 <ncopa> tdtrask: i filed a ticket
2013-10-29 12:26:47 <Frosh> morning everybody
2013-10-29 12:30:14 <Frosh> logs: http://pastebin.com/EGNPFqQc  shutdown: http://pastebin.com/k0zdZub5  times: http://pastebin.com/UH7fHKmF
2013-10-29 12:32:10 <Frosh> the last one shows how long I waited, but its been an hour since I remember to do time on it
2013-10-29 12:33:35 <Frosh> the logs, show whats been going on, i gather as much as I know
2013-10-29 12:38:02 <ncopa> royger: any idea whats going on? ^^^
2013-10-29 15:08:32 <ncopa> algitbot: build 2.6-stable
2013-10-29 15:17:33 <royger> Frosh: what's the problem exactly?
2013-10-29 15:20:34 <ncopa> royger: http://irclogger.com/.alpine-devel/2013-10-28#1382988039
2013-10-29 15:21:58 <royger> ncopa: I seem to remember a similar issue on the xen mailing lists, let me try to find it
2013-10-29 15:27:54 <Frosh> royger: on times, you see that it hangs at shutdown, I have to do xl destroy domuid
2013-10-29 15:28:15 <Frosh> on logs you see it hangs after 1 second, I have to do xl unpause domuid
2013-10-29 15:28:58 <royger> Frosh: are you maybe running the xend daemon?
2013-10-29 15:29:12 <ncopa> Frosh: btw, it looks like pvgrub will be shipped with alpine v2.7
2013-10-29 15:29:25 <Frosh> ? I just install xen as is
2013-10-29 15:29:52 <Frosh> royger: I don't know, how can I check?
2013-10-29 15:30:09 <Frosh> awesome ncopa, when can I test rc-1?
2013-10-29 15:30:44 <Frosh> hopefully its not too different, not excited to relearn the wheel
2013-10-29 15:30:46 <royger> ps aux |grep xend
2013-10-29 15:31:10 <royger> ncopa: isn't pvgrub already shipped with alpine?
2013-10-29 15:31:25 <ncopa> could be, i dont know :)
2013-10-29 15:31:29 <ncopa> i suppose it is
2013-10-29 15:31:35 <royger> hehehe, I'm quite sure it is
2013-10-29 15:31:56 <Frosh> royger: I have to get back to you, I 'm at work, but I didn't modify anything after installing
2013-10-29 15:32:21 <ncopa> royger: xend is legacy stuff right?
2013-10-29 15:32:38 <royger> ncopa: yes, and it shouldn't be installed by default if I'm not mistaken
2013-10-29 15:32:52 <Frosh> http://pastebin.com/EGNPFqQc from the title isn't pygrub still there?
2013-10-29 15:33:19 <ncopa> royger: looks that way yes. good.
2013-10-29 15:33:59 <royger> Frosh: regarding the startup problem, you see the pygrub screen, you press enter, and then nothing happens, right?
2013-10-29 15:34:24 <Frosh> royger, its not just startup, it seems to be shutdown and possible inbetween
2013-10-29 15:34:47 <Frosh> let me describe the problem in 1000 words :)
2013-10-29 15:35:24 <royger> Frosh: so you VM gets paused without you doing anything?
2013-10-29 15:35:28 <Frosh> I do xl create -c debian.cfg , it shows pygrub (see pic), then at 1 second it paused
2013-10-29 15:35:39 <Frosh> I wait 1 hour, no difference
2013-10-29 15:35:57 <Frosh> I didn't press enter, I goto another screen do xl list and its paused
2013-10-29 15:36:04 <Frosh> then I do xl unpause domuid
2013-10-29 15:36:10 <Frosh> and it continues on its merry way
2013-10-29 15:36:21 <Frosh> same thing happens at shutdown
2013-10-29 15:36:42 <royger> Frosh: there were some problems with the pipes and the signals, all this pygrub stuff is kind of hackish to say something
2013-10-29 15:37:04 <Frosh> and this I notice if its hangs, I can't start up another xl create -c seconddebian.cfg
2013-10-29 15:37:34 <Frosh> this didn't happen with my other machine, which is running fine, except its slower :(
2013-10-29 15:37:50 <royger> seems like you have several problems here, is this a fresh install?
2013-10-29 15:37:55 <Frosh> yes
2013-10-29 15:38:24 <Frosh> just install the other day, I like to say 10/26
2013-10-29 15:38:46 <royger> Frosh: ok, but the system was not updated from a previous version or anything like that, just a clean install
2013-10-29 15:39:09 <royger> Frosh: I see you are using 4.2, could you give a try to 4.3?
2013-10-29 15:39:19 <Frosh> I believe its alpine 2.6.5 then I apk updated it
2013-10-29 15:39:32 <ncopa> Frosh: i just pushed 2.7.0_rc1
2013-10-29 15:39:36 <barthalion> yay
2013-10-29 15:39:38 <ncopa> its in the owen...
2013-10-29 15:39:44 <ncopa> oven
2013-10-29 15:41:26 <royger> Frosh: also, there seems to be something wrong with your ACPI tables, Linux prints a suspicious warning: ACPI Warning: 0x00000000000008b0-0x00000000000008bf SystemIO conflicts with Region \GIC2 1 (20130328/utaddress-251)
2013-10-29 15:45:14 <Frosh> I'm not sure what that means
2013-10-29 15:46:33 <Frosh> ncopa: can I ask where the download link is hidden for rc1?
2013-10-29 15:46:46 <royger> Frosh: I'm not sure either :(
2013-10-29 15:47:16 <royger> Frosh: what CPU are you using?
2013-10-29 15:47:30 <ncopa> Frosh: the iso is still in the oven
2013-10-29 15:48:00 <Frosh> well if you don't have any other troubleshooting steps for me, I will reinstall with rc1
2013-10-29 15:48:03 <ncopa> it will be available here in a minote or 2: http://nl.alpinelinux.org/alpine/v2.7/releases/
2013-10-29 15:48:05 <ncopa> its uploading...
2013-10-29 15:48:39 <Frosh> royger: core2duo?
2013-10-29 15:49:18 <Frosh> I'll get you more info for /proc/cpuinfo when I get home
2013-10-29 15:50:24 <Frosh> is /etc/issues where alpine numbering system stored?
2013-10-29 15:51:30 <royger> Frosh: ah, I was wondering if it was a very new system or something like this
2013-10-29 15:52:05 <royger> Frosh: I will try with the latest version (which I guess will come with Xen 4.3), and see what happens
2013-10-29 15:52:47 <royger> Frosh: also, doing a clean install will make sure that all this errors are not due to having older libs around or anything liek that
2013-10-29 16:02:40 <barthalion> upgrade to 2.7 went fine
2013-10-29 16:08:26 <barthalion> and mini iso works too
2013-10-29 16:08:28 <barthalion> ncopa: ↑
2013-10-29 16:15:15 <barthalion> ncopa: by the way, since 2.5 or 2.6 initramfs from iso is too big to boot on 32MB ram boxes
2013-10-29 16:36:11 <Frosh> Is $90 of 16GB ram worth investing in http://www.crucial.com/store/mpartspecs.aspx?mtbpoid=55F8C988A5CA7304 put alpine xen on? or is 4GB, no extra cost, any thoughts?
2013-10-29 16:44:00 <fabian_a> please git pull git://git.alpinelinux.org/fab/aports
2013-10-29 16:44:56 <fabian_a> thanks
2013-10-29 16:49:29 <royger> Frosh: that depends on the number of VMs you want to run, and the amount of memory they need to run
2013-10-29 16:51:15 <Frosh> 4GB  = 1 gb given to dom0, leaves 3 gb left, which means about 3 vm, under contraint
2013-10-29 16:51:39 <royger> Frosh: any many ways does your CPU have?
2013-10-29 16:51:56 <Frosh> core2duo, all these machines I'm finding are
2013-10-29 16:52:02 <Frosh> 3Ghz~
2013-10-29 16:52:10 <Frosh> no i5/i7
2013-10-29 16:52:14 <royger> so that's 2 ways I guess
2013-10-29 16:53:04 <Frosh> but I don't know what to run with 15gb, or even play around with
2013-10-29 16:53:27 <royger> then I think you have already answered yourself ;)
2013-10-29 16:54:08 <Frosh> what do people test with? I want to experiment and learn some new skills too.
2013-10-29 16:54:43 <Frosh> that's why I love this xen stuff :)
2013-10-29 16:55:07 <Frosh> I want to test a new os, shutdown this os, give the ram to this os
2013-10-29 16:55:14 <Frosh> easy peesy
2013-10-29 16:56:38 <Frosh> I'm sorta torn, I don't know if I have the time sometimes to learn new stuff
2013-10-29 16:56:43 <royger> my local test boxes have 8 ways and 6GB of memory, but I don't usually run more than one guest at the same time there
2013-10-29 16:57:28 <Frosh> how much do you allocate to the guest os? typically
2013-10-29 16:59:37 <royger> 1 or 2GB usually, but what do you want to test?
2013-10-29 17:01:16 <Frosh> dunno, just trying to learn whenever my time fees up
2013-10-29 17:01:27 <Frosh> I want to try windows
2013-10-29 17:02:08 <Frosh> but this current one doesn't have hvm support, but the hypothetical 16gb has hvm does
2013-10-29 17:03:36 <royger> I don't get it, buying 16GB of ram is not going to magically enable hardware virtualization extensions, you need a CPU that actually has those features
2013-10-29 17:04:05 <Frosh> no no, sorry for the misunderstanding
2013-10-29 17:04:49 <Frosh> the current one that is having the issue doesn't have vmx | svm, but the next pc I picked up has it
2013-10-29 17:05:17 <Frosh> so I was wondering if I should invest in this computer no body wants
2013-10-29 17:05:51 <royger> I have no idea about the memory requirements of newer windows versions
2013-10-29 17:06:49 <barthalion> the more, the better
2013-10-29 17:06:50 <barthalion> :P
2013-10-29 17:08:36 <Frosh> another thing I always wanted to try out is linux from scratch, with this newer-ish pc
2013-10-29 17:09:01 <Frosh> maybe that will require 16 gb
2013-10-29 17:13:13 <Frosh> now if I could get a hackintosh on, that be great :P
2013-10-30 01:43:22 <BitL0G1c> net-tools - is in aports/main & builds successfully - but apk cannot find it ?
2013-10-30 02:20:55 <BitL0G1c> ok it's in 2.7 tree - better to update to edge or 2.7 ?
2013-10-30 05:40:16 <barthalion> BitL0G1c: better to wait for stable release
2013-10-30 05:40:29 <barthalion> BitL0G1c: I don't understand why would you use it anyway
2013-10-30 05:41:18 <barthalion> BitL0G1c: you should switch to iproute2
2013-10-30 05:55:41 <Frosh> still similar issues as the mkdir /mnt/boot thing and had couldn't stat  mount -t ext4 /dev/sda2 /mnt
2013-10-30 05:59:14 <Frosh> and I think its stuck on installing system on /dev/sda2: /mnt/boot is device /dev/sda1
2013-10-30 07:06:01 <ncopa> fabian_a: thanks! i'm looking over them now
2013-10-30 07:06:23 <ncopa> thank you very much for the license cleanup!
2013-10-30 07:06:51 <ncopa> we need bump pkgrel when license string changes
2013-10-30 07:07:01 <ncopa> otherwise will not builders pick it up and rebuild the package
2013-10-30 07:45:39 <fabian_a> ncopa: ok, i wanted to avoid the rebuild of those packages in the first place. i will go over them and bump the rel
2013-10-30 07:45:52 <ncopa> no
2013-10-30 07:45:54 <ncopa> i'm on it
2013-10-30 07:46:02 <ncopa> i am rebasing your commits
2013-10-30 07:46:15 <ncopa> and fixing up some commit messages
2013-10-30 07:47:56 <fabian_a> the leading '/' is not nice in three messages
2013-10-30 07:48:05 <ncopa> i'm fixin up those too
2013-10-30 07:48:11 <ncopa> git rebase -i
2013-10-30 07:48:17 <ncopa> is a killer feature ;)
2013-10-30 07:49:55 <ncopa> trivial detail. i tend to prefer use the name rather than '$pkgname' in $source and $_builddir
2013-10-30 07:50:14 <ncopa> so: _builddir="$srcdir"/somepackage-$pkgver
2013-10-30 07:50:30 <ncopa> instead of: _builddir="$ssrcdir"/$pkgname-$pkgver
2013-10-30 07:51:15 <ncopa> the reason is that sometimes we add an incompat library as a second package
2013-10-30 07:51:23 <fabian_a> sure
2013-10-30 07:51:23 <ncopa> eg, qt and qt5
2013-10-30 07:51:51 <ncopa> but its a question of taste mostly
2013-10-30 07:52:08 <ncopa> so we can keep those you have done as they are
2013-10-30 07:52:52 <ncopa>     main/halberd: indentation fixed
2013-10-30 07:53:04 <ncopa> that does not change anything in the resulting .apk
2013-10-30 07:53:11 <ncopa> so there it is correct to not bump pkgrel
2013-10-30 07:54:24 <ncopa> hm
2013-10-30 07:54:34 <ncopa> the shared-mime-info trigger has become alot slower...
2013-10-30 07:56:17 <fabian_a> well, it's about the taste but i don't really care if there is the real name as long as most file look the same. makes it easier for third parties if they look at the files from my point of view
2013-10-30 07:57:21 <fabian_a> halberd was only a minor change
2013-10-30 07:57:31 <ncopa> yeah
2013-10-30 07:58:13 <fabian_a> are you talking about the indent and the speed of the shared-mime-info trigger?
2013-10-30 07:58:50 <ncopa> i'm just sayin when to bump pkgrel and when to not
2013-10-30 07:59:09 <ncopa> shared-mime-info trigger is just something unrelated that i have noticed
2013-10-30 07:59:16 <fabian_a> ok
2013-10-30 08:00:23 <ncopa> shared-mime-info beeing slow is just annoying, nothing else
2013-10-30 08:02:02 <ncopa> fabian_a: i am also very happy that you take maintainer for some that was unmaintained. thanks
2013-10-30 08:03:43 <fabian_a> when i see one i take it and hope that somebody starts yelling at me why i took his pkg
2013-10-30 08:03:53 <ncopa> :)
2013-10-30 08:05:16 <ncopa> hm, i wonder if we should rebuild gstreamer 0.10 without gnome-vfs
2013-10-30 08:11:34 <ncopa> texinfo changed .tar.gz to tar.xz in source but did not update checksum
2013-10-30 08:11:40 <ncopa> i'm fixing that up too
2013-10-30 08:29:24 <ncopa> fabian_a: I pushed your changes with various fixups
2013-10-30 08:29:40 <ncopa> mostly i just bumped pkgrel in your commit
2013-10-30 08:29:47 <ncopa> and fixed up /main/... commit message
2013-10-30 08:30:31 <ncopa> i didnt reviewd the testing/* stuff, just tested that it built
2013-10-30 08:30:40 <ncopa> there was some missing checksum
2013-10-30 08:30:45 <ncopa> i fixed that too
2013-10-30 08:39:21 <fabian_a> ncopa: thanks
2013-10-30 08:57:27 <ncopa> i found the reason for shared-mime-info slowdown: https://bugs.freedesktop.org/show_bug.cgi?id=61472#c11
2013-10-30 09:18:44 <royger> do we already have images for ARM? I would be interested in running it on the cubie2
2013-10-30 09:21:04 <fabled> royger, only rpi image
2013-10-30 09:21:27 <fabled> i also have wandboard image that i can publish on request
2013-10-30 09:21:54 <royger> mmm, I guess I can use the rpi image if I compile my own kernel (which I have to do anyway)
2013-10-30 09:22:00 <royger> is it armhf?
2013-10-30 09:24:18 <fabled> yes
2013-10-30 09:24:27 <fabled> it's armhf targetting armv6
2013-10-30 09:24:35 <fabled> packages are built only for that
2013-10-30 09:24:40 <fabled> so you would have complete userland
2013-10-30 09:24:53 <fabled> doing kernel+bootloader yourself should give you working cubieboard
2013-10-30 09:25:13 <fabled> +initramfs
2013-10-30 09:25:45 <fabled> the package repository is http://nl.alpinelinux.org/alpine/edge-musl/main/
2013-10-30 09:26:17 <royger> mine is armv7, is that going to be a problem?
2013-10-30 09:26:20 <fabled> no
2013-10-30 09:26:28 <fabled> wandboard is armv7 too, and those armv6 stuff works on it
2013-10-30 09:26:34 <fabled> they are backwards compat
2013-10-30 09:26:38 <royger> great! :)
2013-10-30 09:26:45 <fabled> though - we might later on build armv7/thumb2 userland too
2013-10-30 09:27:14 <fabled> apparently armv7 has nice additional opcodes; and thumb2 produces considerably smaller code than traditional arm
2013-10-30 09:27:46 <royger> I would like to setup Xen ARM based on alpine
2013-10-30 09:28:06 <fabled> have not tested that
2013-10-30 09:28:09 <royger> where's the rpi image?
2013-10-30 09:28:10 <fabled> oh grsec kernel will not work
2013-10-30 09:28:20 <fabled> musl does not play along with grsec yet
2013-10-30 09:28:25 <royger> np, I have to use a custom kernel and uboot anyway
2013-10-30 09:28:34 <fabled> http://dev.alpinelinux.org/~tteras/alpine-edge-131010-armhf.rpi.tar.gz
2013-10-30 09:28:36 <royger> need to boot in non-secure hyper mode
2013-10-30 09:28:51 <fabled> or if you want overlay enable networking + ssh you can take:
2013-10-30 09:28:52 <fabled> http://dev.alpinelinux.org/~tteras/rpi-with-networking.tar
2013-10-30 09:29:31 <fabled> the kernel is raspberry pi specific
2013-10-30 09:29:41 <fabled> for wandboard i built linux-3.12-rcX vanilla
2013-10-30 09:29:46 <fabled> and uboot
2013-10-30 10:07:37 <royger> fabled: what distro did you use to cross-compile all this? I have debian squeeze, but there's no armhf compile available there, only armel
2013-10-30 10:08:17 <barthalion> royger: mainly because squeeze is terribly old
2013-10-30 10:08:48 <royger> it's not old, it's stable :)
2013-10-30 10:09:29 <royger> anyway, I'm setting up a vm with wheezy, I'm not in the mood to update my dev boxes right now
2013-10-30 10:10:27 <barthalion> oldstable
2013-10-30 10:11:39 <barthalion> armhf has been added in wheezy iirc
2013-10-30 10:11:58 <royger> barthalion: great, let's see how that turns out
2013-10-30 10:50:13 <fabian_a> please git pull git://git.alpinelinux.org/fab/abuild
2013-10-30 10:50:56 <fabian_a> i synced included template with the newapkbuild output and add a license file
2013-10-30 10:59:50 <royger> barthalion: fyi, I ended up using the compiler provided by linaro
2013-10-30 11:07:53 <clandmeter> this bot is getter more verbose by the day. maybe some color(s) would make it a bit better readable?
2013-10-30 12:01:39 <ncopa> or just drop the build messages
2013-10-30 12:01:57 <ncopa> and only show 'uploaded'
2013-10-30 12:02:07 <barthalion> sounds fine
2013-10-30 12:02:16 <ncopa> otoh, the message passing has not been very reliable
2013-10-30 12:02:29 <ncopa> so i think its nice to see that each builder picks it up
2013-10-30 12:13:33 <vkrishn> do check http://tasktools.org/
2013-10-30 12:14:22 <vkrishn> vramsteg , clog, flod seems nice
2013-10-30 12:14:47 <vkrishn> taskd - nice, still in beta
2013-10-30 12:16:01 <vkrishn> ncopa, can 'testing/task' be moved to main ?
2013-10-30 12:16:58 <ncopa> vkrishn: did you test it?
2013-10-30 12:17:03 <ncopa> does it work?
2013-10-30 12:17:19 <vkrishn> yes, put some reports in bugs.a.o
2013-10-30 12:17:54 <vkrishn> #2227
2013-10-30 12:22:21 <Frosh> So far I installed alpine xen 2.7rc1, not had terrible issues
2013-10-30 12:22:51 <ncopa> Frosh: you had minor issues?
2013-10-30 12:23:05 <Frosh> Did the installer change there no #### sign of progress during installation or is that my screen acting up
2013-10-30 12:24:02 <Frosh> I decided to watch the screen for 15 minutes in hope it did install, which I'm happy about.
2013-10-30 12:25:14 <Frosh> when I was impatience in the beginning, I ctrl+c out of it and got wget: short ?get? not entirely sure what that was about
2013-10-30 12:27:09 <Frosh> so onto another issue,, where during the installation mkdir /mnt/boot didn't seem to create the directory
2013-10-30 12:29:22 <Frosh> and after I setup the partition with fdisk /dev/sda, I'm not able to mount /dev/sda1 or /dev/sda2, I ls -la /dev/ and its not there, so I have to reboot to give it another go
2013-10-30 12:30:22 <Frosh> these are old and trivial issues once your done setting up, but annoying if you reinstall constantl,y i guess
2013-10-30 12:35:23 <ncopa> mdes -s would likely have saved you for the reboot
2013-10-30 12:35:24 <ncopa> mdev -s
2013-10-30 12:35:52 <ncopa> if you create partitions while the disk is in use, the partition change will not be visislbe for kernel til until reboot too
2013-10-30 12:36:07 <Frosh> royger: http://pastebin.com/MmNPyuLB still see the ACPI errors, do you know if /cpuinfo is the definitive location for hvm support, cause the logs show (XEN) HVM: VMX enabled
2013-10-30 12:37:25 <Frosh> ncopa: thanks for the tip
2013-10-30 12:41:01 <Frosh> fyi, 'ps aux | grep xend' did return true, dunno if its suppose to be deprecated
2013-10-30 13:18:12 <grimeton> as funny as it is: the other harddrive is dying, dl-5 is offline for ~ 1 hour
2013-10-30 14:43:27 <grimeton> system up and running again
2013-10-30 15:36:25 <grimeton> btw: we should make the info about the sync status default on all mirrors :)
2013-10-30 15:37:11 <clandmeter> jbilyk_: ping
2013-10-30 15:39:03 <clandmeter> jbilyk: ping
2013-10-30 15:39:11 <jbilyk> clandmeter: hey
2013-10-30 15:39:17 <clandmeter> yo
2013-10-30 15:39:22 <clandmeter> houston we got issues
2013-10-30 15:39:25 <jbilyk> ?
2013-10-30 15:39:39 <clandmeter> mysql has corrupt/missing tables...
2013-10-30 15:40:10 <jbilyk> where?
2013-10-30 15:40:17 <clandmeter> www
2013-10-30 15:40:23 <jbilyk> urgh
2013-10-30 15:40:26 <grimeton> Problem? :DDD
2013-10-30 15:41:15 <jbilyk> clandmeter: want the backup file scp'd somewhere or do you have the backup?
2013-10-30 15:41:16 <clandmeter> jbilyk: you got backups right?
2013-10-30 15:41:19 <jbilyk> yup
2013-10-30 15:41:21 <jbilyk> nightlies
2013-10-30 15:41:25 <grimeton> if you need a virtual machine that you can use to run the website in the meantime ...
2013-10-30 15:41:28 <clandmeter> i wonder how long this been going on
2013-10-30 15:41:29 <jbilyk> 
ACTION just confirmed I have last night's

2013-10-30 15:41:39 <jbilyk> clandmeter: dunno, I don't browse the site daily
2013-10-30 15:41:49 <jbilyk> clandmeter: you could try restoring last night's and seeing if there's issues
2013-10-30 15:42:09 <clandmeter> ok please cp it to mysql.a.o
2013-10-30 15:42:16 <clandmeter> or www
2013-10-30 15:42:20 <clandmeter> whatever you want
2013-10-30 15:42:30 <clandmeter> ill create a new db and import it
2013-10-30 15:42:33 <jbilyk> ok, looking up the right ssh port since I always forget :)
2013-10-30 15:42:34 <clandmeter> and hook drupal on it
2013-10-30 15:42:45 <clandmeter> yes it sucks like this :)
2013-10-30 15:43:47 <clandmeter> im glad we got that backup running :)
2013-10-30 15:44:14 <grimeton> i shot the sheriff ... http://www.clickondetroit.com/lifestyle/pets/cat-survives-arrow-through-its-skull/-/1718840/22711428/-/a9s2d4z/-/index.html
2013-10-30 15:44:24 <jbilyk> clandmeter: www-files-2013-10-30.tar.gz on www
2013-10-30 15:44:32 <jbilyk> it's got the drupal files and mysql backup
2013-10-30 15:45:07 <clandmeter> grazie
2013-10-30 15:45:16 <Frosh> you mean the forum down, i think it was down for a few day
2013-10-30 15:45:21 <Frosh> or the packages list
2013-10-30 15:45:37 <jbilyk> np
2013-10-30 15:45:38 <jbilyk> brb
2013-10-30 15:45:52 <clandmeter> jbilyk:
2013-10-30 15:45:57 <clandmeter> thats empty :|
2013-10-30 15:46:50 <clandmeter> 545 bytes..
2013-10-30 15:50:05 <grimeton> need a mirror?
2013-10-30 15:50:27 <clandmeter> jbilyk: you are backup up a symlink...
2013-10-30 15:51:06 <grimeton> highly efficient
2013-10-30 15:51:12 <grimeton> sorry, i don't know why i'm trolling today
2013-10-30 15:57:28 <jbilyk> clandmeter: ?
2013-10-30 15:57:51 <clandmeter> your backup script it not ok
2013-10-30 15:57:58 <clandmeter> but i hope we are lucky
2013-10-30 15:58:13 <clandmeter> you only backed up the database
2013-10-30 15:58:17 <clandmeter> which is what we need
2013-10-30 15:58:30 <clandmeter> seems last few days the backup didnt run properly
2013-10-30 15:58:37 <clandmeter> probably the database was already corrupted
2013-10-30 15:58:56 <jbilyk> clandmeter: want one from a week ago?
2013-10-30 15:59:04 <jbilyk> we have nightlies going back a month
2013-10-30 15:59:08 <clandmeter> you kept the backups on the site :)
2013-10-30 15:59:14 <clandmeter> about a week
2013-10-30 15:59:20 <jbilyk> no, they're copied to dl-2
2013-10-30 15:59:32 <jbilyk> i have going back to 9/20/2013
2013-10-30 15:59:37 <clandmeter> i know
2013-10-30 15:59:40 <clandmeter> but i read your script
2013-10-30 15:59:45 <clandmeter> you put them in home/backups
2013-10-30 15:59:54 <clandmeter> and clean them older then 7 days
2013-10-30 16:00:19 <jbilyk> yeah, and copy nightly from www to dl-2 (triggered on dl-2)
2013-10-30 16:00:27 <jbilyk> 
ACTION has to run for lunch, back in an hour or so

2013-10-30 16:02:12 <clandmeter> ok its back online
2013-10-30 16:02:13 <clandmeter> seems to work
2013-10-30 16:12:24 <clandmeter> jbilyk: we should update the backup script
2013-10-30 16:12:33 <clandmeter> tar should folow symlinks
2013-10-30 16:12:40 <clandmeter> and the database is changed now.
2013-10-30 16:15:06 <clandmeter> im going to go home. nite all
2013-10-30 17:04:38 <jbilyk> clandmeter: if you have cycles, please feel free to update the script on www
2013-10-30 17:05:30 <jbilyk> backup box is expecting /home/backups/*apkovl*.gz and /home/backups/www-files*.gz
2013-10-30 17:05:54 <jbilyk> so as long as those are there, feel free to modify the script as needed
2013-10-30 18:37:39 <shafire> :-9
2013-10-30 18:37:39 <shafire> :-)
2013-10-30 18:37:56 <shafire> hi
2013-10-30 20:32:46 <fabian_a> i would like to have a meta package for the installation of some  stuff mentioned at http://wiki.alpinelinux.org/wiki/Alpine_security
2013-10-30 20:33:18 <fabian_a> are there any minimal requirement beside that only pkgs from main can be included?
2013-10-30 23:12:58 <Frosh> o man, I just cause a break
2013-10-31 01:16:32 <Frosh> back to square 1
2013-10-31 01:58:02 <Frosh> anyone setup pvgrub before?
2013-10-31 02:28:20 <jbilyk_> Frosh: under amazon ec2 iirc, yes
2013-10-31 02:28:28 <jbilyk_> http://wiki.alpinelinux.org/wiki/Install_Alpine_on_Amazon_EC2
2013-10-31 03:06:25 <Frosh> nothing there has any hint why my pvgrub has no network capabilities but pygrub does
2013-10-31 03:57:46 <tdtrask> ncopa: sorry I didn't get those changes in before rc1
2013-10-31 03:58:19 <tdtrask> bumped every acf package and fixed the source location
2013-10-31 07:14:16 <kaniini> Frosh: it does
2013-10-31 08:13:03 <ncopa> morning
2013-10-31 08:13:14 <ncopa> tdtrask: np. thanks
2013-10-31 08:44:32 <lpolzer> hi
2013-10-31 09:01:11 <royger> fabled: your rpi alpine images are designed to run from ram, right?
2013-10-31 09:09:14 <clandmeter> ncopa: is 2.7 musl?
2013-10-31 09:10:15 <ncopa> no
2013-10-31 09:11:09 <ncopa> the musl release will be numbered 3.0
2013-10-31 09:11:16 <ncopa> the first musl release that is
2013-10-31 09:11:36 <ncopa> so if there comes a 2.8 after 2.7 then its still uclibc
2013-10-31 09:11:37 <royger> ncopa: is musl also comming to x86?
2013-10-31 09:11:57 <ncopa> royger: yes, thats the plan
2013-10-31 09:12:12 <ncopa> we drop uclibc and switch completely to musl
2013-10-31 09:12:15 <royger> ncopa: do we have a list of packets that fail to compile under musl on x86?
2013-10-31 09:12:22 <ncopa> fabled has
2013-10-31 09:12:28 <royger> I'm not sure if the xen tools will play nice with musl...
2013-10-31 09:12:32 <ncopa> we have built most of them
2013-10-31 09:12:52 <ncopa> i think there are like 200 of 1600 or so that currently dont build
2013-10-31 09:13:06 <ncopa> (which means we have ~1400 packages built!)
2013-10-31 09:13:15 <royger> ncopa: let's see if fabled can post the list :)
2013-10-31 09:13:29 <royger> yes, not bad
2013-10-31 09:13:40 <ncopa> nl.alpinelinux.org/alpine/edge-musl/
2013-10-31 09:14:02 <ncopa> after 2.7 is out i will set up a musl dev box for me too and try port the rest
2013-10-31 09:14:55 <ncopa> one of the challenges is that musl is "too correct"
2013-10-31 09:15:04 <ncopa> may apps expects broken behaviour :)
2013-10-31 09:15:21 <ncopa> or make too many assumptions
2013-10-31 09:15:38 <royger> as expected, I don't see any xen package in x86_64 :(
2013-10-31 09:17:33 <ncopa> and no qemu
2013-10-31 09:17:34 <ncopa> well
2013-10-31 09:17:40 <ncopa> its still a work in progress
2013-10-31 09:33:01 <lpolzer> fabled: my x86 musl build is still being built fine.
2013-10-31 09:33:27 <lpolzer> fabled: I'm unable to cross-compile gcc-gnat though
2013-10-31 09:33:55 <lpolzer> fabled: linker reports an undefined reference to `_gnat_stack_check'
2013-10-31 09:47:24 <fcolista> ncopa, there's a link to get alpine rss feed of master git repos ?
2013-10-31 09:47:26 <fcolista> http://git.alpinelinux.org/cgit/aports
2013-10-31 09:49:35 <ncopa> i dont know if cgit has rss support
2013-10-31 09:49:49 <fcolista> umh
2013-10-31 09:49:50 <fcolista> http://redmine.alpinelinux.org/projects/alpine/activity
2013-10-31 09:49:57 <fcolista> this may help
2013-10-31 09:50:54 <clandmeter> cgit used to have rss support
2013-10-31 09:51:12 <clandmeter> but they changed something (which broke my pacakge browser on www to get latest commits)
2013-10-31 09:51:16 <ncopa> http://git.alpinelinux.org/cgit/aports/atom
2013-10-31 09:52:22 <ncopa> cool
2013-10-31 09:52:26 <ncopa> i didnt know about that
2013-10-31 10:13:09 <shafire> hi
2013-10-31 10:13:31 <shafire> is there already a musl alpha iso image for x86(_64)?
2013-10-31 10:19:41 <fabian_a> please git pull git://git.alpinelinux.org/fab/aports
2013-10-31 10:20:33 <fabian_a> this request is the license cleanup for 'a' without 'acf-*'
2013-10-31 10:24:25 <shafire> fabian_a: are you talking to me? I think not, or?
2013-10-31 10:25:28 <fabled> shafire, no .iso images yet.
2013-10-31 10:25:38 <fabled> should probably make them sometime soon
2013-10-31 10:28:42 <shafire> okay
2013-10-31 10:32:36 <fabian_a> shafire: sorry, no
2013-10-31 10:35:57 <fabian_a> shafire: more to ncopa
2013-10-31 10:36:23 <shafire> ok
2013-10-31 11:47:28 <clandmeter> ncopa: back from lunch?
2013-10-31 11:48:25 <ncopa> officially, no ;)
2013-10-31 11:48:38 <clandmeter> ok ill try to keep it non official
2013-10-31 11:48:50 <clandmeter> not sure you got the msg yesterday
2013-10-31 11:48:51 <ncopa> :D
2013-10-31 11:49:01 <ncopa> not sure either
2013-10-31 11:49:03 <clandmeter> but kvm db server has issues
2013-10-31 11:49:19 <clandmeter> atleast one table for www got corrupted
2013-10-31 11:49:31 <clandmeter> so we lost 3 days of sql data on www
2013-10-31 11:49:42 <clandmeter> i dont care about those 3 days
2013-10-31 11:49:49 <clandmeter> it worries me more it got lost
2013-10-31 11:49:55 <clandmeter> or currupted
2013-10-31 11:50:23 <clandmeter> also the backup procedure was not running as it should.
2013-10-31 11:50:35 <clandmeter> but we were lucky the db did export correct
2013-10-31 11:51:37 <ncopa> we use innodb?
2013-10-31 11:51:53 <clandmeter> i think drupal uses that by default
2013-10-31 11:52:22 <ncopa> i dont know what might cause db corruption
2013-10-31 11:53:08 <clandmeter> I think we should really put more attention in to our backup procedure.
2013-10-31 11:53:38 <clandmeter> we includes me
2013-10-31 11:54:38 <ncopa> ok
2013-10-31 11:54:42 <ncopa> yes
2013-10-31 11:55:05 <ncopa> can you please verify that we dont use myisam?
2013-10-31 11:57:22 <clandmeter> ncopa: bugs does use myisam
2013-10-31 11:58:25 <clandmeter> ncopa: i took the script that jbilyk and modified it a bit which is used for backups.
2013-10-31 11:59:05 <clandmeter> my idea is to: a) create a backup user. b) add the script to cron daily. c) configure the script. done.
2013-10-31 11:59:15 <ncopa> does drupal db also use myisam?
2013-10-31 11:59:33 <clandmeter> i dont think so
2013-10-31 11:59:40 <clandmeter> why would you use myisam?
2013-10-31 12:00:11 <ncopa> i googled around and it seems like myisam is more likely to get corruption
2013-10-31 12:00:22 <ncopa> so if we use it, we should probably migrate to innodb
2013-10-31 12:00:33 <ncopa> personally i prefer postgresql
2013-10-31 12:00:53 <clandmeter> i dont
2013-10-31 12:01:32 <clandmeter> ive run in too much troubles using pgsql with differernt webapps
2013-10-31 12:01:54 <clandmeter> seems most of those webapps always target mysql as db
2013-10-31 12:02:11 <clandmeter> even gets worse when you use plugins in those webapps
2013-10-31 12:03:09 <clandmeter> anyways, we also had corruption on pgsql before. so it would only be because of politcal reason to change.
2013-10-31 12:03:15 <clandmeter> and i think chaging is not that easy.
2013-10-31 12:05:07 <clandmeter> i think we could move redmine to innodb
2013-10-31 12:05:38 <ncopa> yeah i think thats a good idea
2013-10-31 12:07:58 <ncopa> fabian_a:
2013-10-31 12:08:00 <ncopa> -license="GPL-2"
2013-10-31 12:08:00 <ncopa> +license="GPL2"
2013-10-31 12:08:06 <ncopa> just curious why?
2013-10-31 12:09:50 <ncopa> if we want to be consistent, then we should doc it some place
2013-10-31 12:10:53 <clandmeter> can't we hardcode it in abuild?
2013-10-31 12:11:03 <clandmeter> and issue a warning?
2013-10-31 12:11:31 <fabian_a> because the most used short forms are GPL2 or GPLv2. most newer pkg follow this.
2013-10-31 12:14:02 <fabian_a> of course this is just a cosmetic change but i checked the license anyway and wanted to be consistent
2013-10-31 12:14:45 <fabian_a> those changes (from gpl-2 to gpl2) were done without pkgrel bump
2013-10-31 12:16:07 <ncopa> the problem with editing pkgrel without bump is
2013-10-31 12:16:16 <ncopa> that if you rebuild from aports tree
2013-10-31 12:16:23 <ncopa> with exact same version of gcc
2013-10-31 12:16:32 <ncopa> the contents will be different
2013-10-31 12:17:10 <ncopa> if the resulting .apk get changes, the pkgrel should be bumped
2013-10-31 12:18:07 <fabian_a> ncopa: just for my understand...is the pkg rebuild triggered by git?
2013-10-31 12:18:38 <ncopa> yes, but it will compare the file-$pkgver-r$pkgrel.apk
2013-10-31 12:18:43 <ncopa> if it exists then it will not rebuild
2013-10-31 12:19:39 <ncopa> i think we can just let it be for this time
2013-10-31 12:19:43 <ncopa> i'm testbuilding it
2013-10-31 12:19:53 <ncopa> and will push if it succeds
2013-10-31 12:20:42 <ncopa> i am very happy for help with this
2013-10-31 12:20:48 <ncopa> its good we are consistent
2013-10-31 12:21:06 <ncopa> and its kinda important that license is correct
2013-10-31 12:21:33 <ncopa> but its "boring", so a big thanks!
2013-10-31 12:22:13 <fabian_a> it will take some time to finish...
2013-10-31 12:23:58 <ncopa> not sure we shoudl do it right now...
2013-10-31 12:24:05 <ncopa> we have 1 week for release
2013-10-31 12:24:10 <fabian_a> in fedora we have the package review process and a couple of time i needed to go done the hard way because of licensing issues
2013-10-31 12:24:14 <ncopa> next tuesday is v2.7
2013-10-31 12:24:51 <fabian_a> we don't need to push it
2013-10-31 12:25:21 <ncopa> i think we should doc in wiki how the preferred license format should be
2013-10-31 12:25:41 <fabian_a> sure, i will do that
2013-10-31 12:25:47 <ncopa> super!
2013-10-31 12:28:20 <fabian_a> 
ACTION add this to the todo list

2013-10-31 12:29:50 <ncopa> hm
2013-10-31 12:29:51 <ncopa> atk
2013-10-31 12:30:00 <ncopa>     main/atk: upgrade to 2.11.1 and specify license
2013-10-31 12:30:05 <ncopa> i will drop that one
2013-10-31 12:30:07 <Frosh> Morning
2013-10-31 12:30:19 <ncopa> can reapply it after v2.7 release
2013-10-31 12:30:28 <ncopa> 2.11.x is dev release i think
2013-10-31 12:30:33 <ncopa> 2.10.x is stable
2013-10-31 12:30:40 <ncopa> we want the stable for the v2.7 release
2013-10-31 12:30:42 <Frosh> so I'm guessing there isn't any progress bar during the installation anymore?
2013-10-31 12:30:58 <ncopa> Frosh: i havent looked at it unfortunally :-/
2013-10-31 12:32:11 <Frosh> so the domu pauses under pygrub under 2.7rc1 still
2013-10-31 12:33:25 <Frosh> so far I haven't seen it pause under pvgrub, but I have no network capabilities under it, can't ping anything. not sure if I'm setting the pvgrub correctly either.
2013-10-31 12:33:50 <fabian_a> ncopa: yepp, better to skip it
2013-10-31 12:35:53 <fabian_a> main/apache-mod-fcgid is another canditate for postponing
2013-10-31 12:40:38 <ncopa> i'm ok with it as log as it builds
2013-10-31 12:42:27 <fabian_a> acpica didn't build after the update but that change was reverted
2013-10-31 12:47:12 <shafire> [13:08] <@ncopa> -license="GPL-2" [13:08] <@ncopa> +license="GPL2" [13:08] <@ncopa> just curious why? ncopa: why not an enum for license?
2013-10-31 12:52:37 <Frosh> hmm, I'm tempted to just install 2.6.3 to see if all these problems go away.
2013-10-31 12:53:02 <ncopa> Frosh: might we worth a test yes
2013-10-31 12:53:16 <ncopa> algitbot: build master
2013-10-31 12:59:19 <ncopa> i wonder why bash fails to build...
2013-10-31 12:59:24 <ncopa> it smells parallel issue
2013-10-31 13:23:12 <ncopa> bah
2013-10-31 13:23:22 <ncopa> bash makefiles are stupid
2013-10-31 14:09:13 <StarWarsFan|afk> ncopa: is there something broken with the alpine-conf-repo?
2013-10-31 14:09:18 <StarWarsFan> hi btw.
2013-10-31 14:09:27 <StarWarsFan> can't sync my mirror with this error:
2013-10-31 14:19:02 <StarWarsFan>  ! [remote rejected] v2.0_al�pha6 -> v2.0_al�pha6 (pre-receive hook declined)
2013-10-31 14:19:02 <StarWarsFan>  ! [remote rejected] v2.14.0 -> v2.14.0 (pre-receive hook declined)
2013-10-31 14:19:29 <StarWarsFan> can't say exactly since when this happens
2013-10-31 14:19:57 <StarWarsFan> 
ACTION was more or less ofline since monday

2013-10-31 14:20:04 <StarWarsFan> +f
2013-10-31 14:26:44 <ncopa> hm
2013-10-31 14:26:45 <ncopa> ok
2013-10-31 14:27:18 <ncopa> i had an v2.0_al?pha tag taht was bad
2013-10-31 14:27:30 <ncopa> i tried to remove it manually
2013-10-31 14:27:55 <ncopa> StarWarsFan: could you please try to clone it again?
2013-10-31 14:28:14 <ncopa> rm -rf alpine-conf && git clone git://git.alpinelinux.org/alpine-conf
2013-10-31 14:28:17 <ncopa> sorry about that...
2013-10-31 14:29:33 <StarWarsFan> no problem,
2013-10-31 14:29:45 <StarWarsFan> waiting for the cron job to the job...
2013-10-31 14:29:52 <StarWarsFan> +do
2013-10-31 14:32:11 <ncopa> what happened was
2013-10-31 14:32:22 <ncopa> i had that bad tag long time ago
2013-10-31 14:32:29 <ncopa> i fixed it repo long time ago
2013-10-31 14:32:38 <ncopa> but i hadnt removed the tag in my local branch
2013-10-31 14:32:52 <ncopa> so when i pushed the 2.14.0 tag the bad tag got pushed again
2013-10-31 14:32:55 <tdtrask> ncopa / mhavela: why is acf-provisoning-polycom4.0.3f a new package and not an upgrade?
2013-10-31 14:33:23 <ncopa> tdtrask: i think mhavela wanted to be able to select which firware version
2013-10-31 14:33:31 <ncopa> so you can manually pick one
2013-10-31 14:34:45 <tdtrask> 
ACTION doesn't like it :(

2013-10-31 14:34:55 <StarWarsFan> ah, that's why this pop's up again
2013-10-31 14:35:17 <StarWarsFan> as i'd got this problem long time ago and you've fixed it then... ;-)
2013-10-31 14:35:30 <ncopa> tdtrask: talk with mhavela about it
2013-10-31 14:35:32 <tdtrask> unless he's planning to merge them when he's done testing
2013-10-31 14:35:44 <tdtrask> I will
2013-10-31 14:36:21 <ncopa> tdtrask: i tend to agree with you, but i think there was some requirement that you could pick which firware verision you wanted
2013-10-31 14:36:53 <ncopa> if its upgrade only you can in practice only pick the one shipped in that release branch
2013-10-31 14:37:03 <tdtrask> right
2013-10-31 14:37:23 <ncopa> which means, if you get bad firmware on upgrade you cannot downgrade
2013-10-31 14:37:28 <ncopa> you can prevent upgrades though
2013-10-31 14:39:22 <mhavela> tdtrask: Do you have a suggestion how users would be able to chose which firmware version they want to use?
2013-10-31 14:39:36 <mhavela> (if not doing it the way it was done now)
2013-10-31 14:41:45 <tdtrask> 
ACTION suggests not giving them the option :)

2013-10-31 14:43:48 <Frosh> curious is alpine the only distro that offer xen iso?
2013-10-31 14:49:19 <barthalion> possibly
2013-10-31 14:57:46 <vkrishn> #1454
2013-10-31 14:58:13 <vkrishn> ncopa: updated with alt ^^^
2013-10-31 14:58:25 <vkrishn> http://bugs.alpinelinux.org/issues/1454
2013-10-31 14:58:52 <ncopa> suckless
2013-10-31 14:59:04 <ncopa> is normally tiny and simple
2013-10-31 14:59:40 <barthalion> what's the problem with xvkbd deps?
2013-10-31 15:00:02 <barthalion> what's wrong with xaw3d?
2013-10-31 15:00:18 <ncopa> Xaw3d
2013-10-31 15:00:24 <ncopa> its just old and ugly
2013-10-31 15:00:30 <ncopa> and i have been lazy to add it
2013-10-31 15:00:50 <barthalion> I can take a look at packaging florence
2013-10-31 15:00:57 <ncopa> sure
2013-10-31 15:01:30 <ncopa> svkbd is probably a good alternative for minimalist ppl
2013-10-31 15:03:25 <vkrishn> thanks for looking into it
2013-10-31 15:05:57 <barthalion> vkrishn: btw, test pcmanfm
2013-10-31 15:06:20 <barthalion> it's already in main
2013-10-31 15:06:45 <vkrishn> :) yes would give it a try
2013-10-31 15:11:07 <ncopa> thunar is also well supported
2013-10-31 15:12:52 <ncopa> barthalion: i think i have samba4 working more or less
2013-10-31 15:13:16 <ncopa> do you know if there are any showstoppers?
2013-10-31 15:13:24 <ncopa> like some package not building against samba4
2013-10-31 15:13:56 <barthalion> mplayer had some issues, but it may be fixed in svn
2013-10-31 15:14:16 <ncopa> maybe xbmc too
2013-10-31 15:15:12 <ncopa> barthalion: you think its worth it?
2013-10-31 15:15:16 <ncopa> its late in the game...
2013-10-31 15:16:02 <barthalion> bah, I don't know, I guess the majority of users doesn't care
2013-10-31 15:16:20 <barthalion> as long as it works of course
2013-10-31 15:16:31 <barthalion> Idk if upgrade is straightforward
2013-10-31 15:19:02 <ncopa> looks like no biggie: https://abf.rosalinux.ru/openmandriva/xbmc/blob/master/xbmc-12.1-samba4.patch
2013-10-31 15:19:55 <barthalion> yeah, I think that's the only problem with mplayer too
2013-10-31 15:20:05 <ncopa> ok
2013-10-31 15:20:27 <ncopa> i think our current samba3 is pretty bloated
2013-10-31 15:20:32 <ncopa> due to static bundled libs
2013-10-31 15:20:40 <ncopa> i havent checked though
2013-10-31 15:20:54 <ncopa> fabled mentioned it
2013-10-31 15:21:42 <StarWarsFan> ncopa: hm, the problem with the repo-sync still exists
2013-10-31 15:21:55 <StarWarsFan> needs to be investigated later @home...
2013-10-31 15:22:22 <ncopa> weird
2013-10-31 15:22:59 <Frosh> didn't know alpine does kiosk too
2013-10-31 15:24:34 <ncopa> yes samba4 is significantly smaller than samba3
2013-10-31 15:25:14 <vkrishn> Frosh, few ingredients are still missing
2013-10-31 15:25:32 <vkrishn> like touch libs
2013-10-31 15:26:45 <vkrishn> control mouse/input via webcam :)
2013-10-31 15:27:17 <Frosh> what is the kiosk intent? is it just suppose to web browse?
2013-10-31 15:27:42 <vkrishn> could be order menu in a restaurant
2013-10-31 15:28:22 <ncopa> algitbot: build master
2013-10-31 15:28:28 <vkrishn> or transport info on station
2013-10-31 15:29:01 <ncopa> ok here comes rc2
2013-10-31 15:31:12 <ncopa> rebooting. brb
2013-10-31 15:38:28 <ncopa> ok rc2 is out
2013-10-31 15:53:34 <Frosh> you just popping them out
2013-10-31 15:58:56 <ncopa> yup
2013-10-31 20:27:40 <barthalion> ncopa: re mplayer & samba 4.0
2013-10-31 20:28:07 <barthalion> ncopa: you need to change one line to fix  header include and use --enable-smb
2013-10-31 20:30:45 <fabian_a> is there something like pushd/popd available on busybox? sorry for asking...google is not much of a helping hand here
2013-10-31 20:32:05 <barthalion> there is no pushd/popd in POSIX
2013-10-31 20:35:20 <fabian_a> hmm, would be nice. no need to rewrite scripts
2013-10-31 20:51:53 <barthalion> well, you can just switch to bash
2013-10-31 22:21:36 <Xena> I'm having trouble in virtualbox
2013-10-31 22:21:38 <Xena> http://i.imgur.com/RPga6b3.png
2013-10-31 22:55:42 <grimeton> Xena: enabled any special settings?
2013-10-31 23:07:24 <grimeton> and the useless use of chroot oscar goes to: chroot /
2013-10-31 23:30:31 <grimeton> gna
2013-10-31 23:30:34 <grimeton> hit that fucking xfs bug again
2013-10-31 23:35:40 <grimeton> ffs what the hell is wrong with mkinitfs