2016-11-01 06:40:30 <fabled> lots of missing sources
2016-11-01 07:31:26 <fabled> mmmkay
2016-11-01 07:42:04 <fabled> doxygen build fail on x86 is due to flex generating bad code :/
2016-11-01 07:47:39 <fabled> ok flex 2.6.2 is broke
2016-11-01 08:21:49 <clandmeter> algitbot: retry master
2016-11-01 08:22:27 <clandmeter> fabled: what is that build builder?
2016-11-01 08:22:38 <fabled> dunno
2016-11-01 08:25:14 <clandmeter> fabled: doxgen is build, not sure how you got your 3.5 to continu to build again? 
2016-11-01 08:25:38 <fabled> rm -rf aports/*/*/src
2016-11-01 08:25:52 <fabled> rm /tmp/upload*
2016-11-01 08:26:06 <fabled> and finally restart the mqtt
2016-11-01 08:33:04 <clandmeter> thx
2016-11-02 08:58:30 <ScrumpyJack> .
2016-11-02 13:49:11 <fabled> hum
2016-11-02 13:49:14 <fabled> teeworlds no longer builds
2016-11-02 13:49:18 <fabled> likely gcc6 issue
2016-11-02 13:49:27 <fabled> thinking to just move it to unmaintained
2016-11-02 15:32:12 <clandmeter> why does abuild still use colors?
2016-11-02 15:32:29 <clandmeter> it breaks the log files
2016-11-02 15:38:16 <fabled> clandmeter, you could do another rm -rf aports/*/*/src for arm builders
2016-11-02 15:38:22 <fabled> they should be able to catch up quite a bit
2016-11-02 15:38:59 <clandmeter> k
2016-11-02 15:40:32 <clandmeter> done
2016-11-02 22:38:22 <clandmeter> jirutka: why did you change releases to nl.a.o?
2016-11-02 22:38:50 <jirutka> clandmeter: because https, dl-cdn does not support https
2016-11-02 22:39:06 <jirutka> clandmeter: it’s important mainly for ISO images
2016-11-02 22:39:14 <clandmeter> please dont concider nl.a.o master mirror.
2016-11-02 22:39:27 <jirutka> clandmeter: does any other mirror support https?
2016-11-02 22:40:13 <jirutka> clandmeter: IIRC I’ve discussed this change with ncopa or you some time ago, but I have bad memory, so it’s also possible that I remember it wrong
2016-11-02 22:41:03 <clandmeter> nl use to be master mirror
2016-11-02 22:41:08 <clandmeter> be we changed that.
2016-11-02 22:41:13 <jirutka> what is the master mirror now?
2016-11-02 22:41:28 <clandmeter> it always points to rsync.a.o
2016-11-02 22:41:40 <clandmeter> currently it is fr.a.o
2016-11-02 22:41:51 <clandmeter> but its scaleway
2016-11-02 22:41:56 <jirutka> clandmeter: neither rsync.a.o or fr.a.o does not have https
2016-11-02 22:42:03 <clandmeter> i know
2016-11-02 22:42:41 <clandmeter> i think we discussed this part few weeks ago
2016-11-02 22:42:44 <jirutka> and downloading sha checksum of ISO image from non-https source is not very secure :/
2016-11-02 22:44:17 <jirutka> yes, we’ve discussed it and IIRC the result was that we change url on webpage to some mirror with with HTTPS, wasn’t it?
2016-11-02 22:45:15 <clandmeter> i cannot recall that outcome.
2016-11-02 22:45:24 <clandmeter> atleast i never agreed on using nl.a.o
2016-11-02 22:45:51 <clandmeter> nl.a.o bw is limited
2016-11-02 22:45:55 <jirutka> hm, okay, so is it possible to enable https on fr.a.o or rsync.a.o?
2016-11-02 22:46:06 <clandmeter> it is
2016-11-02 22:46:11 <clandmeter> but not today...
2016-11-02 22:46:32 <clandmeter> i have enough alpine linux infra/development headache today
2016-11-02 22:46:48 <jirutka> okay
2016-11-02 22:48:05 <clandmeter> if you are only concerned about checksums, why not only update those uri's?
2016-11-02 22:49:29 <clandmeter> btw, i think we need to add those somehow in the new template.
2016-11-02 22:50:59 <jirutka> clandmeter: I’m not concerned only about checksums, also ISO images should be on https by default imho
2016-11-02 22:51:49 <jirutka> clandmeter: and this change would require non-systematic change of the script for generating release page
2016-11-02 22:52:01 <clandmeter> i know
2016-11-02 22:52:16 <clandmeter> but changing it to a mirror which has bw issues is not a solution.
2016-11-02 22:53:10 <jirutka> that’s true, I didn’t know about these bw issues
2016-11-02 22:55:03 <clandmeter> next time please ask first, it will prevent these issues. (im glad i notices the change)
2016-11-02 22:56:32 <clandmeter> jirutka: if you have an idea how to nicely add the checkums to the new dl page, let me know.
2016-11-02 22:56:53 <jirutka> clandmeter: I’ll look at it in a while
2016-11-02 22:57:15 <clandmeter> i have a feeling if i add them just below, it will get messy.
2016-11-02 22:57:18 <jirutka> clandmeter: I’m sorry about this change
2016-11-02 22:59:38 <clandmeter> jirutka: its ok. we will fix it properly asap.
2016-11-02 23:03:46 <clandmeter> ok for now www and non-www will keep their own config.
2016-11-02 23:04:15 <clandmeter> but with the fixes
2016-11-02 23:21:05 <jirutka> clandmeter: maybe it’d be better to just revert the commit with s/dl-cdn/nl/ for now than messing generate_release.lua script
2016-11-03 11:49:52 <clandmeter> hope that will keep armhf builder busy for a while.
2016-11-04 07:55:47 <fabled> urg. tired. the above commit was not needed.
2016-11-04 07:55:50 <fabled> should get coffee
2016-11-04 08:12:33 <clandmeter> fabled: we all need coffee :)
2016-11-04 08:16:07 <nmeum> apk fetch coffee
2016-11-04 08:19:00 <clandmeter> my coffee smells great, going to be a good friday.
2016-11-04 08:27:14 <clandmeter> fabled: 3.5 mostly ready now?
2016-11-04 08:28:15 <fabled> we still have few non-building packages
2016-11-04 08:28:29 <clandmeter> have a list?
2016-11-04 08:28:42 <clandmeter> or are they all brainbreakers?
2016-11-04 08:28:52 <fabled> x86_64: http://sprunge.us/RNdM
2016-11-04 08:29:03 <fabled> x86: http://sprunge.us/egBM
2016-11-04 08:29:38 <clandmeter> ok ill take a look
2016-11-04 08:29:41 <fabled> at least some are non-trivial
2016-11-04 08:35:12 <fabled> i've working on catching up patchworks
2016-11-04 08:35:32 <clandmeter> good, it needed attention.
2016-11-04 08:36:57 <clandmeter> looks like pw gets less attention now that we have github+ci
2016-11-04 08:37:13 <clandmeter> another good reason for bb? :)
2016-11-04 08:38:01 <fabled> yes
2016-11-04 08:38:30 <fabled> but it also goes by developer preference
2016-11-04 08:38:46 <fabled> we have few active contributors preferring github, so there's increased stuff there
2016-11-04 08:48:53 <fabled> need to run for awhile. back in the afternoon.
2016-11-04 08:56:45 <clandmeter> nmeum: around?
2016-11-04 08:57:06 <nmeum> clandmeter: yes
2016-11-04 08:57:18 <clandmeter> soeren right?
2016-11-04 08:57:23 <nmeum> yap
2016-11-04 08:57:25 <clandmeter> :)
2016-11-04 08:57:33 <clandmeter> can you upgrade debootstrap?
2016-11-04 08:58:18 <nmeum> I could apply the following patch http://sprunge.us/YdCH
2016-11-04 08:58:29 <nmeum> I wouldn't upgrade it to the most recent version until 3.5 is released 
2016-11-04 08:58:43 <clandmeter> ok thats also fine.
2016-11-04 08:58:51 <nmeum> I will push that then
2016-11-04 08:59:10 <nmeum> there you go
2016-11-04 09:00:19 <clandmeter> thx
2016-11-04 09:00:46 <nmeum> no problem
2016-11-04 09:15:42 <clandmeter> looks like geh completely disappeared
2016-11-04 09:15:48 <clandmeter> and no maintainer
2016-11-04 09:15:59 <clandmeter> i get we just more it to unmaintained
2016-11-04 09:16:03 <clandmeter> move...
2016-11-04 09:44:12 <clandmeter> fabled: can i combine noarch with !arch?
2016-11-04 11:58:07 <fabled> clandmeter, i'm taking a look at valgrind
2016-11-04 12:01:04 <clandmeter> fabled: i also looked but didnt seem easy.
2016-11-04 12:02:14 <fabled> it was simpler than i thought... :)
2016-11-04 12:02:28 <clandmeter> did you also upgrade to latest version?
2016-11-04 12:03:20 <fabled> clandmeter, y
2016-11-04 12:07:33 <fabled> clandmeter, aiccu is funny. source gives 402 for wget user-agent
2016-11-04 12:07:43 <fabled> i wonder how to fix that
2016-11-04 12:11:45 <clandmeter> fabled: looks like somebody else is subscribed to our msg.a.o
2016-11-04 12:11:56 <clandmeter> wonder if i should block him via fw
2016-11-04 12:24:28 <clandmeter> ok i blocked the ausie
2016-11-04 12:24:35 <clandmeter> anybody here from australia? 
2016-11-04 14:01:29 <fabled> okay... down the 4 non-builds
2016-11-04 14:01:58 <fabled> oh, 5
2016-11-04 14:24:40 <fabled> clandmeter, i have chromium fix test building
2016-11-04 14:39:03 <clandmeter> nice
2016-11-04 16:33:48 <^7heo> WOOOW
2016-11-04 16:33:51 <^7heo> SO MUCH GREEN!
2016-11-04 16:35:45 <clandmeter> :p
2016-11-04 16:36:30 <clandmeter> i think i broke the apache2 build on build-3-5-armhf
2016-11-04 16:36:39 <clandmeter> rm -rf src dir when build was still going :)
2016-11-04 16:37:17 <fabled> :)
2016-11-04 16:37:24 <fabled> just retry it then..
2016-11-04 16:37:44 <clandmeter> does algitbot rety work again?
2016-11-04 16:39:20 <clandmeter> fabled: cant you just take the source from distfiles for aiccu?
2016-11-04 16:39:28 <clandmeter> it hasnt updates since ages
2016-11-04 16:39:31 <fabled> true
2016-11-04 16:39:34 <fabled> it's just annoying
2016-11-04 16:39:37 <clandmeter> updated
2016-11-04 16:39:40 <fabled> i was tempted to do that
2016-11-04 16:39:43 <clandmeter> :)
2016-11-04 16:39:47 <clandmeter> seems the last one?
2016-11-04 16:39:49 <fabled> or maybe take our own copy
2016-11-04 16:40:15 <clandmeter> no option to just move it to unmaintained?
2016-11-04 16:40:30 <fabled> not sure
2016-11-04 16:40:37 <fabled> maybe let ncopa fix it
2016-11-04 16:40:50 <fabled> we need to leave something for him too ;)
2016-11-04 16:41:12 <clandmeter> he already looked at it i remember.
2016-11-04 16:43:32 <clandmeter> fabled: how did you get buildrepo summary?
2016-11-04 16:43:57 <fabled> http://build.alpinelinux.org/buildlogs/build-3-5-x86.log
2016-11-04 16:44:03 <fabled> http://build.alpinelinux.org/buildlogs/build-3-5-x86_64.log
2016-11-04 16:45:07 <clandmeter> hmm
2016-11-04 16:45:12 <clandmeter> but no files for arm
2016-11-04 16:47:22 <clandmeter> cool
2016-11-04 16:47:29 <clandmeter> only aiccu for aarch64
2016-11-04 16:47:39 <clandmeter> 2 more at armhf
2016-11-04 16:48:50 <fabled> yeah, the arm logs are on the builders only. should probably fix the script to upload it to build.a.o
2016-11-04 16:49:06 <clandmeter> openal is neon related
2016-11-04 16:49:55 <clandmeter> i think im going to go home
2016-11-04 16:49:58 <clandmeter> its weekend
2016-11-04 16:51:28 <clandmeter> fabled: ive looked at ovmf, but seems a bit difficult to build. maybe you or ncopa can take a look at it next week.
2016-11-04 16:52:44 <fabled> ok. i'm home already :)
2016-11-04 18:20:43 <jirutka> clandmeter: fabled: IIRC ncopa will not be here next week, it’s all up to us
2016-11-06 18:08:36 <ncopa> i'll deal with it tomorrow
2016-11-07 16:21:34 <ncopa> \o/
2016-11-07 16:26:39 <tdtrask> ncopa must be happy to see green after all the red that was there before he left for vaca
2016-11-07 16:33:20 <rnalrd> forgot about phpversion
2016-11-07 16:36:28 <clandmeter> rnalrd: magic commit msg ;-)
2016-11-07 16:36:40 <rnalrd> lol
2016-11-07 17:16:55 <fabled> algitbot: retry master
2016-11-07 17:17:08 <fabled> algitbot: retry edge
2016-11-07 19:42:38 <ncopa> seems like armhf is done with building too now
2016-11-14 09:17:35 <clandmeter> fcolista: openvas-cli-1.4.5.tar.gz: FAILED
2016-11-14 09:17:56 <fcolista> yes clandmeter 
2016-11-14 09:17:58 <fcolista> i'm on it
2016-11-14 09:18:21 <fcolista> typo on a wrong row
2016-11-15 20:52:23 <ncopa> dsabogal: ^^^
2016-11-15 20:52:53 <dsabogal> dsabogal: i see
2016-11-16 08:19:41 <fabled> fcolista, care to close http://bugs.alpinelinux.org/issues/5728 ?
2016-11-16 08:20:08 <fcolista> sure fabled
2016-11-16 08:20:19 <fcolista> just waiting for 5.6.1 packages to build 
2016-11-16 08:20:21 <fabled> since you did the pkg update already...
2016-11-16 08:20:22 <fabled> :)
2016-11-16 08:20:29 <fcolista> yep
2016-11-16 08:22:42 <clandmeter> fcolista: is something blocking qt 5.7?
2016-11-16 08:23:29 <fcolista> no afaik.
2016-11-16 08:25:10 <fcolista> clandmeter, i think we can upgrade qt to 5.7
2016-11-16 15:07:18 <clandmeter> fabled: :)
2016-11-16 15:08:25 <fabled> clandmeter, :)
2016-11-17 10:54:47 <fabled> algitbot: retry 3-4-stable
2016-11-17 10:54:59 <fabled> algitbot: retry 3.4-stable
2016-11-18 08:00:32 <clandmeter> :)
2016-11-18 10:10:21 <fcolista> uff why my BE does not retry to download the package rather than pickup it from the cache?
2016-11-18 10:13:56 <fcolista> umh
2016-11-18 10:14:07 <fcolista> looks that  aarch64 has a different path for a.out.h
2016-11-18 10:14:14 <fcolista> how so? 
2016-11-18 10:14:21 <clandmeter> it doesnt have it
2016-11-18 10:14:37 <clandmeter> i believe
2016-11-18 10:14:44 <fcolista> Is not the kernel the same?
2016-11-18 10:15:48 <fcolista> https://pkgs.alpinelinux.org/contents?file=a.out.h&path=&name=&branch=&repo=&arch=aarch64
2016-11-18 10:16:09 <fcolista> /usr/src/linux-headers-4.4.32/include/linux/a.out.h
2016-11-18 10:16:43 <fcolista> it's in linux-vanilla-dev
2016-11-18 10:18:35 <fcolista> umh
2016-11-18 10:23:29 <fcolista> should i add linux-vanilla-dev in makedepends?
2016-11-18 10:24:56 <fcolista> I don't understand why since the kernel is the same, the include files shoudl differ
2016-11-18 10:38:26 <clandmeter> fcolista: ive run into this before
2016-11-18 10:38:36 <clandmeter> its not provided for aarch64
2016-11-18 10:39:00 <fcolista> So i need to disable building for aarch64
2016-11-18 10:39:12 <clandmeter> that, or fix it.
2016-11-18 10:39:14 <fcolista> arch="all !aarch" would suffice ?
2016-11-18 10:39:42 <clandmeter> im not sure what it needs from a.out.h
2016-11-18 10:40:05 <clandmeter> could be its not needed at all, didnt look at the code.
2016-11-18 10:40:20 <fcolista> me either. Therefore removing this header for aarch64 might cause malfunction
2016-11-18 10:40:37 <clandmeter> no thats not possible
2016-11-18 10:40:49 <clandmeter> if the build doenst break its ok.
2016-11-18 10:42:35 <clandmeter> fcolista: its a header file, its provides something for the compiler, not when running the program afterwards.
2016-11-18 10:43:52 <clandmeter> fcolista: http://pkgs.alpinelinux.org/contents?file=a.out.h&path=*linux*&name=linux-headers&branch=edge&repo=main&arch=x86_64
2016-11-18 10:44:05 <clandmeter> thats the header for x86_64
2016-11-18 10:44:10 <fcolista> yes
2016-11-18 10:44:15 <clandmeter> but if you change it to aarch64 its gone.
2016-11-18 10:44:19 <fcolista> for aarch64 is on linux-vanilla-dev
2016-11-18 10:44:25 <fcolista> correct
2016-11-18 10:44:27 <clandmeter> not really
2016-11-18 10:44:48 <clandmeter> linux-vanilla-dev are all headers
2016-11-18 10:45:27 <clandmeter> we have linux-headers for kernel headers
2016-11-18 10:45:52 <clandmeter> try to remove the include, and see where it errors
2016-11-18 10:46:02 <fcolista> Ok. Anyway, i don't understand this: if source is the same, kernel is the same, why some headers are not available ?
2016-11-18 10:46:04 <clandmeter> if you cannot fix that issue, then just diable it.
2016-11-18 10:46:18 <fcolista> .h are not compiled
2016-11-18 10:46:39 <clandmeter> fabled: can you explain it in proper english?
2016-11-18 10:46:51 <clandmeter> im scared im saying something wrong :)
2016-11-18 10:47:11 <fcolista> I thought that the headers file are the same across all archs
2016-11-18 10:47:27 <clandmeter> no i dont think so
2016-11-18 10:47:28 <fcolista> since the kernel source is the same.
2016-11-18 10:47:44 <clandmeter> not all architectures have the same feature set.
2016-11-18 10:47:54 <fcolista> So you mean that according with the arch, some files are stripped?
2016-11-18 10:48:05 <fcolista> even if they are not binary?
2016-11-18 10:48:09 <clandmeter> like i said, i prefer fabled explains it. 
2016-11-18 10:48:17 <fcolista> ok
2016-11-18 10:48:24 <fcolista> thx anyway for our input clandmeter 
2016-11-18 10:48:27 <clandmeter> before i say something thats not true
2016-11-18 10:48:29 <fcolista> *your
2016-11-18 10:49:02 <clandmeter> fcolista: for instance neon
2016-11-18 10:49:44 <clandmeter> and SSE i guess
2016-11-18 10:49:59 <clandmeter> all features of processors which are included in some
2016-11-18 10:51:28 <fcolista> ok so it comes another question: 
2016-11-18 10:51:44 <fcolista> how do I #include only if a specific ARCH is matched?
2016-11-18 10:52:22 <fcolista> Or I should add prepare() and if $ARCH is aarch64 then remove that include?
2016-11-18 10:52:43 <clandmeter> # ifdef AARCH64 
2016-11-18 10:52:48 <fcolista> Ah ok
2016-11-18 10:52:55 <fcolista> that's the way
2016-11-18 10:53:26 <clandmeter> if defined(__aarch64__)
2016-11-18 10:53:45 <clandmeter> just search on google for ifdef aarch64
2016-11-18 10:54:38 <clandmeter> https://sourceforge.net/p/predef/wiki/Architectures/
2016-11-18 10:55:18 <clandmeter> __aarch64__ is the correct marco
2016-11-18 10:55:45 <fcolista> I should put if not defined aarch64 then add that header
2016-11-18 10:55:46 <clandmeter> fcolista: does it compile when you remove a.out.h include?
2016-11-18 10:56:02 <fcolista> I'm going to do that 
2016-11-18 10:56:09 <fcolista> I've not yet done 
2016-11-18 10:56:19 <fcolista> You mean remove it at all?
2016-11-18 10:56:21 <clandmeter> just remove it and try in src dir
2016-11-18 10:56:51 <fcolista> heh
2016-11-18 10:56:56 <fcolista> looking better at the source code
2016-11-18 10:57:10 <fcolista> that include is for darwin
2016-11-18 10:57:23 <fcolista> I'll remove it 
2016-11-18 10:57:42 <clandmeter> aarch64 is still a new arch
2016-11-18 10:57:52 <clandmeter> many projects didnt add support to it.
2016-11-18 10:57:58 <clandmeter> just like musl support
2016-11-18 10:58:11 <fcolista> yes
2016-11-18 11:05:08 <fabled> not all headers are 'noarch'
2016-11-18 11:05:15 <fabled> especially kernel headers are architecture specific
2016-11-18 11:05:36 <fabled> since they define the ABI between userland and kernel; which often have differences per-processor architecture
2016-11-18 11:07:16 <fcolista> ah-a
2016-11-18 11:07:25 <fcolista> that makes sense
2016-11-18 11:07:35 <fcolista> so, would that work?
2016-11-18 11:07:35 <fcolista> http://sprunge.us/HGXU
2016-11-18 11:08:04 <fabled> ideally you'd have configure check to test presense of the header file
2016-11-18 11:09:00 <fcolista> this does not have configure script
2016-11-18 11:09:10 <fcolista> (if it's what you are referring to)
2016-11-18 11:10:00 <fabled> does it work if you just unconditionally remove the #include ?
2016-11-18 11:10:05 <fcolista> no
2016-11-18 11:10:13 <fcolista> i've tried
2016-11-18 11:10:44 <fabled> i guess #if defined(__arch64__) is okish workaround too
2016-11-18 11:11:08 <fabled> or #if !defined...
2016-11-18 11:11:57 <fcolista> what's the difference betweend ifndef and if !defined 
2016-11-18 11:11:57 <fcolista> ?
2016-11-18 11:12:32 <fabled> functionally it's same thing
2016-11-18 11:12:45 <fabled> the !defined is easier to expand if you need to test some other defines
2016-11-18 11:13:05 <fcolista> with && (__OTHER__)
2016-11-18 11:13:09 <fabled> yeah
2016-11-18 11:13:41 <fcolista> gotcha.
2016-11-18 11:13:54 <clandmeter> fcolista: are you testing this on aarch64?
2016-11-18 11:14:10 <fcolista> no. I'm waiting for algitbot fail
2016-11-18 11:14:11 <fcolista> :)
2016-11-18 11:14:26 <clandmeter> fcolista: dont increase pkgrel
2016-11-18 11:14:29 <fcolista> I don't have aarch64 to make test
2016-11-18 11:15:48 <clandmeter> i would just disable it for  aarch64
2016-11-18 11:16:00 <fabled> and armhf?
2016-11-18 11:16:02 <fcolista> clandmeter, arch="all !aarch64"
2016-11-18 11:16:48 <fcolista> yes fabled. So arch="all !aarch64 !armhf"
2016-11-18 11:16:48 <clandmeter> fabled: he probably doesnt have armhf either to test
2016-11-18 11:16:58 <fabled> i wonder if we could ship compat a.out.h in some package
2016-11-18 11:17:11 <fcolista> I had but Scrumpyjack removed it
2016-11-18 11:17:26 <clandmeter> fcolista: i can provide you with containers
2016-11-18 11:17:36 <clandmeter> if you are interested
2016-11-18 11:17:40 <fcolista> clandmeter, would be great
2016-11-18 11:17:46 <clandmeter> i mean
2016-11-18 11:17:55 <clandmeter> you maintain something like 1k packages
2016-11-18 11:18:02 <clandmeter> makes sense you got proper hw support
2016-11-18 11:18:12 <fabled> clandmeter, reminds me, we should ship some bootable image for aarch64
2016-11-18 11:18:22 <clandmeter> remind me next week
2016-11-18 11:18:29 <clandmeter> fabled: yes
2016-11-18 11:18:34 <clandmeter> i think i have something somwhere
2016-11-18 11:19:01 <fabled> i thought the aarch64 kernel is uefi capable as such. so it's just making proper image layout
2016-11-18 11:19:05 <clandmeter> fcolista: that remind me is for you :)
2016-11-18 11:19:28 <clandmeter> fabled: i have a script that creates a uefi iso
2016-11-18 11:19:40 <fcolista> clandmeter, :()
2016-11-18 11:19:42 <fcolista> :)
2016-11-18 11:19:46 <clandmeter> i gave it to ncopa
2016-11-18 11:19:51 <fabled> clandmeter, could you find it for me too?
2016-11-18 11:19:57 <clandmeter> sure
2016-11-18 11:20:00 <fabled> i'm happy to make a mkimage.sh profile out of it
2016-11-18 11:20:33 <clandmeter> fabled: http://tpaste.us/2KLP
2016-11-18 11:20:59 <fabled> oh, it uses grub
2016-11-18 11:21:40 <clandmeter> fabled: yes, what else?
2016-11-18 11:22:09 <clandmeter> have you seen the patch madness on fedora on grub?
2016-11-18 11:23:06 <clandmeter> i added support for aarch64 to gummiboot, but it doesnt work.
2016-11-18 11:23:31 <clandmeter> syslinux doesnt support it
2016-11-18 11:24:40 <clandmeter> fcolista: please pm me your ssh key. then ill see when i have time to set it up.
2016-11-18 11:25:33 <fabled> clandmeter, is bootloader required, or could we just boot the kernel image directly? i thought it's valid efi app?
2016-11-18 11:25:46 <fabled> or bootloader is needed to load the initramfs?
2016-11-18 11:26:34 <clandmeter> correct
2016-11-18 11:26:46 <clandmeter> it need to load initramfs
2016-11-18 11:39:00 <fabled> i guess grub-efi is easiest to do then...
2016-11-18 11:39:13 <fabled> your script is pretty clear, so maybe we just replicate it on mkimage.sh
2016-11-18 11:41:27 <fabled> grub-efi is in testing, should probably go to main soonish
2016-11-18 11:42:32 <fabled> i guess it's another story also to add uefi support to setup-disk and supporting it when updating kernels
2016-11-18 11:42:49 <fabled> i guess next request when we get above done, is secure boot...
2016-11-18 13:49:23 <fabled> seems we need http://bugs.alpinelinux.org/issues/5191 for x86_64 efi boot
2016-11-19 01:37:05 <jirutka> algitbot: retry master
2016-11-19 01:41:23 <jirutka> algitbot: retry master
2016-11-19 12:40:35 <jirutka> ^ fyi working on it
2016-11-19 20:22:50 <jirutka> algitbot: retry master
2016-11-21 15:54:34 <clandmeter> from who is this build builder?
2016-11-21 19:34:56 <ncopa> what is `build`?
2016-11-21 19:35:11 <fabled> ncopa, clandmeter and i have been wondering that too
2016-11-21 19:36:17 <clandmeter> its another subscriber on msg.a.o
2016-11-21 19:36:25 <clandmeter> i think i blocked one 2 weeks ago
2016-11-21 19:36:32 <clandmeter> he/she was even more verbose
2016-11-21 19:36:47 <clandmeter> we should add auth
2016-11-21 19:36:51 <clandmeter> i can look into that.
2016-11-21 19:41:53 <ncopa> i was kinda hoping to avoid using auth
2016-11-21 19:42:01 <ncopa> as it means we need reconfigure them all
2016-11-21 19:42:06 <ncopa> all the builders
2016-11-21 19:42:14 <ncopa> hm
2016-11-21 19:42:58 <ncopa> i think what i'd want, is a vpn
2016-11-21 19:43:05 <ncopa> and only pass mqtt over vpn
2016-11-21 19:47:36 <clandmeter> vpn is also fine for me
2016-11-21 19:47:45 <clandmeter> but that also needs a reconfig on all boxes
2016-11-22 03:08:29 <jirutka> clandmeter: what are these failures with missing buildlog? ^
2016-11-22 07:52:32 <clandmeter> jirutka: somebody who is spamming msg.a.o
2016-11-23 08:32:38 <fcolista> someone figured what is this "build" ? 
2016-11-23 09:03:05 <clandmeter> fcolista: its somebody elses non official builder
2016-11-23 09:04:20 <fcolista> cannot algitbot be protected with a sort of auth?
2016-11-23 09:04:33 <clandmeter> it can
2016-11-23 09:10:34 <clandmeter> it looks like somebody from australia is doing this
2016-11-23 09:10:38 <clandmeter> i blokced him before
2016-11-23 09:10:40 <clandmeter> or her
2016-11-23 09:12:37 <clandmeter> blocked again
2016-11-23 09:12:54 <fcolista> ok
2016-11-23 09:13:16 <fcolista> generally speking, though, i think we should implement authentication for this purpose
2016-11-23 09:13:20 <clandmeter> we should probably put auth on write to mqtt
2016-11-23 09:13:25 <fcolista> right
2016-11-23 09:13:39 <clandmeter> but there has been talk about setting up vpn's between buildinfra
2016-11-23 09:14:12 <clandmeter> but i think we should not remove msg.a.o from public
2016-11-23 09:14:23 <clandmeter> its nice for others to subscribe and listen.
2016-11-23 09:14:59 <fcolista> of course having mqtt open for reading is a good idea
2016-11-23 09:15:12 <clandmeter> but if we add auth, we probably also have to setup encryption.
2016-11-23 09:17:16 <fcolista> correct clandmeter 
2016-11-23 14:47:21 <fabled> duh
2016-11-24 15:06:13 <fabled> sigh
2016-11-24 15:06:56 <fabled> need  to run and kernels got broke :/
2016-11-24 15:07:00 <fabled> i try to fix it asap
2016-11-24 21:03:09 <fabled> algitbot: retry master
2016-11-25 09:25:30 <clandmeter> fabled: did that --quiet not end up in my iso?
2016-11-25 09:25:45 <fabled> i just added fixup commit
2016-11-25 09:27:32 <fabled> so yes, the fbcon issue is still present in the .iso you got
2016-11-25 10:03:36 <clandmeter> jirutka: around?
2016-11-25 10:05:06 <clandmeter> why is python3 pulling in python3-dev?
2016-11-25 10:09:52 <clandmeter> fabled: any idea how this is possible? looking at the apkbuild it makes no sense to me.
2016-11-25 11:07:56 <fabled> python3-3.5.2-r7 depends on:
2016-11-25 11:07:56 <fabled> python3-dev=3.5.2-r7
2016-11-25 11:08:30 <fabled> huh
2016-11-25 11:12:32 <clandmeter> yep
2016-11-25 11:12:51 <clandmeter> it has a custom dev function, does that break it?
2016-11-25 11:14:52 <fabled> no
2016-11-25 11:14:58 <fabled> it's some broken symlink
2016-11-25 11:15:23 <fabled> python3/usr/lib/python3.5/config-3.5m/libpython3.5m.so
2016-11-25 11:15:33 <fabled> that should be in python3-dev
2016-11-25 11:16:10 <fabled> it's the one created in package()
2016-11-25 11:16:24 <fabled> i wonder if the symlink target should be in python3 package then
2016-11-25 11:17:28 <fabled> ./pkg/python3-dev/usr/lib/python3.5/config-3.5m/libpython3.5m.a
2016-11-25 11:17:29 <fabled> ./pkg/python3-dev/usr/lib/libpython3.5m.so
2016-11-25 11:17:29 <fabled> ./pkg/python3/usr/lib/python3.5/config-3.5m/libpython3.5m.so
2016-11-25 11:17:29 <fabled> ./pkg/python3/usr/lib/libpython3.5m.so.1.0
2016-11-25 11:17:53 <fabled> maybe it should point to the so.1.0 instead
2016-11-25 11:18:59 <fabled> i wonder how it was in 3.4
2016-11-25 11:19:17 <clandmeter> its related  to this i think: https://bugs.archlinux.org/task/22552
2016-11-25 11:19:36 <fabled> http://bugs.alpinelinux.org/issues/6456
2016-11-25 11:20:29 <clandmeter> https://git.archlinux.org/svntogit/packages.git/tree/trunk/PKGBUILD?h=packages/python#n88
2016-11-25 11:21:42 <fabled> we do same
2016-11-25 11:21:49 <fabled> the problem is because of the package split
2016-11-25 11:21:57 <fabled> i'm wondering which we to move it around
2016-11-25 11:26:48 <fabled> maybe i just expand it
2016-11-25 11:32:02 <jirutka> clandmeter: yep
2016-11-25 11:39:26 <clandmeter> fabled: thx!
2016-11-25 11:39:40 <fabled> np
2016-11-25 14:45:55 <clandmeter> Olá amigo
2016-11-25 19:45:04 <clandmeter> :|
2016-11-29 14:26:09 <clandmeter> jirutka: update config guess
2016-11-30 17:51:36 <ncopa> something went wrong
2016-11-30 17:52:06 <^7heo> you're too late
2016-11-30 17:52:12 <^7heo> Trump has already been elected and it's confirmed.