2017-12-01 00:17:28 <jirutka> Shiz, could you pls do a favour for me?
2017-12-01 00:18:37 <Shiz> what's that
2017-12-01 00:19:16 <jirutka> Shiz: respond to https://twitter.com/grsecurity/status/936387692769800199 (thread https://twitter.com/alpinelinux/status/936364858571927558)… I don’t want to bother ncopa with another twitter affair caused by me and I really can’t respond to this politely, it’s imo impossible
2017-12-01 00:21:12 <jirutka> the most polite I can respond to this is something like “You can pick one: you’re a lier, or really totally insane, both in good and bad sense, b/c it’s not possible for a human to maintain it in this way”
2017-12-01 00:24:03 <jirutka> maybe, maybe if he had eidetic memory; but still it would be completely insane and extremely error prone approach
2017-12-01 00:29:45 <Shiz> he already said he used git~
2017-12-01 00:30:52 <jirutka> “The PaX Team doesn't use git for their workflow, so there is no git repo for his patches to be flattened out from…”
2017-12-01 00:34:23 <Shiz> that's just the pax patches
2017-12-01 00:35:36 <jirutka> damn, i forgot about in how bad state is pkgs.a.o and I must manually disable cron jobs after adding v3.7, b/c that poor sqlite is about to explode :(
2017-12-01 00:55:04 <danieli> jirutka: sounds solvable
2017-12-01 00:55:10 <danieli> have anyone started a new pkgs.a.o?
2017-12-01 00:55:40 <jirutka> danieli: not yet, but I’d like to do it during Christmas
2017-12-01 00:55:48 <danieli> agreed
2017-12-01 00:55:49 <jirutka> I’ll see
2017-12-01 00:55:56 <danieli> I have some things taking precedence over that on my todo but
2017-12-01 00:56:03 <danieli> a new pkgs.a.o could be a good idea
2017-12-01 00:56:10 <danieli> what dbms are we aiming to use?
2017-12-01 00:56:21 <danieli> doubt we'll need riak or cassandra level stuff
2017-12-01 00:56:23 <jirutka> I’d like to use PostgreSQL
2017-12-01 00:56:26 <danieli> it's a deal
2017-12-01 00:57:26 <jirutka> I have so many things on TODO list that I just give priority to the thing that bother me the most at the moment XD
2017-12-01 00:57:53 <danieli> LOL, that happens
2017-12-01 00:58:03 <danieli> too many "high" priority todo entries it doesn't even make a difference
2017-12-01 00:58:10 <danieli> so many*
2017-12-01 01:00:51 <jirutka> yeah
2017-12-01 01:59:50 <jirutka> ncopa: 726 aports out of 749 in unmaintained not modified since v3.6.0… should I really purge them?
2017-12-01 02:01:16 <Xe> git usually lets you undelete things
2017-12-01 05:53:05 <_ikke_> jirutka: if perl is available, I can just do: perl -e 'exit 1 if (<> ~= /\x00/)'
2017-12-01 05:54:53 <_ikke_> ncopa: can I use a bit of perl for this pre-receive hook (ie, is perl available)?
2017-12-01 09:51:23 <clandmeter> jirutka: i have a local version of pkgs.a.o that has splitted databases for each branch which performs much better.
2017-12-01 09:53:22 <clandmeter> i dont think it makes sense to create a new pkgs when apk-tools is re-written
2017-12-01 09:53:49 <clandmeter> but if anybody wants to do that and redo it again after new apk-tools, be my guest.
2017-12-01 11:07:38 <jirutka> clandmeter: there are way more problems with current version and there are unrelated to apk-tools format etc.
2017-12-01 11:09:21 <clandmeter> jirutka: please share them
2017-12-01 11:12:21 <clandmeter> jirutka: you can also add issues to the project and i can look into it.
2017-12-01 11:12:48 <jirutka> clandmeter: the root cause is wrong DB scheme, it’s not something that can be easily fixed
2017-12-01 11:13:06 <jirutka> clandmeter: there’s no distinct between aport (source) and package
2017-12-01 11:13:54 <jirutka> clandmeter: and when you flag package, the flag is cleared once the package is updated, regardless of the version, so even on -r update
2017-12-01 11:14:58 <jirutka> clandmeter: we have dozens of pkgs appear on Flagged page b/c some builders are constantly stuck so there’s an outdated package in the tree, but the aport is already up-to-date
2017-12-01 11:16:49 <clandmeter> ok so you want to reference information from aports.git instead of apkindex?
2017-12-01 11:17:15 <jirutka> there should be both in pkgs.a.o DB
2017-12-01 11:17:41 <clandmeter> yes i mean both
2017-12-01 11:18:02 <clandmeter> but some refs based on git instead of index
2017-12-01 11:18:21 <jirutka> look, this discussion is pointless, we both know the state of pkgs.a.o code-base
2017-12-01 11:19:31 <clandmeter> ok sorry for wasting your time.
2017-12-01 11:20:42 <jirutka> I don’t wanna look like disrespectful, but you yourself said me multiple times that you know the code is bad
2017-12-01 11:21:42 <clandmeter> Yes is did, and i like to improve it.
2017-12-01 11:23:38 <clandmeter> and i think i already improved it in my local copy
2017-12-01 11:23:42 <jirutka> it cannot be w/o rewriting
2017-12-01 11:23:54 <clandmeter> but of course its never perfect...
2017-12-01 11:26:17 <clandmeter> lets talk some time about it when you and me have time.
2017-12-01 11:26:24 <jirutka> okay
2017-12-01 11:26:28 <clandmeter> and anybody who is interested in it.
2017-12-01 11:26:57 <clandmeter> enjoy your weekend :)
2017-12-01 11:34:38 <jirutka> just for others, to make it clear, I know pkgs.a.o code-base very well, I already contributed to it (integration with release-monitoring.org), I know Lua and have experience with software projects, so what I said is not some ranting against someone else work
2017-12-01 11:35:37 <jirutka> I just want to prevent misunderstanding
2017-12-01 13:39:55 <ncopa> heh, the alpine 3.7 release didnt even make it to the frontpage of http://distrowatch.com/
2017-12-01 13:40:13 <ncopa> should have mentioned some desktop related news, like Xorg or gnome stuff
2017-12-01 13:41:39 <ncopa> phoronix mentioned it though https://www.phoronix.com/scan.php?page=news_item&px=Alpine-Linux-3.7-Released
2017-12-01 15:23:16 <PureTryOut[m]> and the upgrade to ConsoleKit2 😄
2017-12-01 16:03:45 <_ikke_> Now that 3.7.0 is out, can someone take a look at https://github.com/alpinelinux/aports/pull/2433 ?
2017-12-01 16:12:55 <jirutka> ncopa: [02:59:50]  jirutka:	ncopa: 726 aports out of 749 in unmaintained not modified since v3.6.0… should I really purge them?
2017-12-01 16:13:19 <ncopa> yes
2017-12-01 16:13:27 <ncopa> please
2017-12-01 16:13:40 <jirutka> and is there any reason to keep unmaintained at all?
2017-12-01 16:14:00 <hzJE0gsFfQkx> maybe someone wants to take ownership over any of them?
2017-12-01 16:14:01 <jirutka> i mean remove unmaintained aport directly instead of moving it to unmaintained/
2017-12-01 16:14:12 <ncopa> the intention with the unmaintained is that people can see what is unmaintained and can pick it up
2017-12-01 16:14:13 <hzJE0gsFfQkx> it's easier to start from somewhere than from zero
2017-12-01 16:14:21 <ncopa> if nobody does within a reasonable time, we purge it
2017-12-01 16:14:31 <jirutka> anyone can pick it from git history
2017-12-01 16:14:55 <ncopa> true, but its not that visible
2017-12-01 16:15:15 <hzJE0gsFfQkx> it makes it more difficult to take ownership
2017-12-01 16:16:05 <ncopa> if you need some new package, what are the chances that you would look in the git history if we had that package previously?
2017-12-01 16:16:22 <ncopa> i think most people wont
2017-12-01 16:16:41 <jirutka> i’d understand that if we don’t purge unmaintained after 6 months
2017-12-01 16:17:20 <ncopa> we could increase it to a year
2017-12-01 16:17:26 <ncopa> or any other limit
2017-12-01 16:17:48 <ncopa> but if you find an APKBUILD in unmaintained that 5+ years, what are the chances that it still works?
2017-12-01 16:18:05 <ncopa> after that you could as well just start over from scratch
2017-12-01 16:18:18 <ncopa> thats the reasoning at least
2017-12-01 16:18:36 <hzJE0gsFfQkx> 5 years seems more reasonable than 6 months
2017-12-01 16:19:05 <jirutka> 709 packages not touched since 3.5.0, so it does not make any significant diference
2017-12-01 16:19:25 <jirutka> aports, not packages
2017-12-01 16:19:42 <ncopa> the thinking is that we want clean up unmaintained stuff
2017-12-01 16:19:58 <ncopa> but at the same time give contributors a chance to pick it up before its deleted
2017-12-01 16:20:02 <danieli> I'm gonna grep through and perhaps grab a few if it makes sense to me
2017-12-01 16:20:03 <jirutka> two releases is more reasonable than one IMO
2017-12-01 16:20:21 <ncopa> lets do 1 year then
2017-12-01 16:23:00 <hzJE0gsFfQkx> what is the benefit of this?
2017-12-01 16:23:17 <TBB> I've probably talked about sifting through unmaintained stuff, perhaps semi automatically, and see if updated versions can be produced without a maintainer
2017-12-01 16:24:18 <TBB> of course that means that if any bug reports come in, there's no maintainer...
2017-12-01 16:27:52 <jirutka> TBB: we usually move stuff to unamtaintained when it does not work anymore…
2017-12-01 16:33:29 <jirutka> done
2017-12-01 16:35:43 <ncopa> thanks!
2017-12-01 16:35:53 <jirutka> yw
2017-12-01 16:38:46 <ncopa> grep tests seems to hang on edge builders
2017-12-01 16:38:50 <ncopa> 23597 23595 buildoze R     3632   0%   3   4% awk BEGIN{for (;;) printENVIRON ["line ]}
2017-12-01 16:38:50 <ncopa> 14125 14105 buildoze R     3628   0%  18   4% awk BEGIN{for (;;) printENVIRON ["line ]}
2017-12-01 16:38:50 <ncopa> 18833 18827 buildoze R     3628   0%  23   4% awk BEGIN{for (;;) printENVIRON ["line ]}
2017-12-01 16:39:44 <jirutka> what grep tests?
2017-12-01 16:40:06 <ncopa> check() function in main/grep/APKBUILD
2017-12-01 16:40:24 <jirutka> aha
2017-12-01 16:40:50 <ncopa> seems to happen on all builders
2017-12-01 16:40:56 <ncopa> but i cnanot reproduce on my dev box
2017-12-01 16:41:26 <ncopa> http://build.alpinelinux.org/buildlogs/build-edge-x86_64/main/grep/grep-3.1-r1.log
2017-12-01 16:41:30 <ncopa> hangs at c-locale
2017-12-01 16:43:56 <ncopa> or the test after, case-fold-backref
2017-12-01 16:50:22 <ncopa> 11920 buildoze   0:00 /bin/sh ../build-aux/test-driver --test-name grep-dev-null-out --log-file grep-dev-null-out.log --trs-file grep-dev-null-out.trs --color-tests no --enable-hard-errors yes --expect-failure no
2017-12-01 16:50:22 <ncopa> 11951 buildoze   0:00 /bin/sh ./grep-dev-null-out
2017-12-01 16:50:32 <ncopa> the test that hangs is grep-dev-null-out
2017-12-01 16:52:33 <ncopa> ok, thi sis weird
2017-12-01 23:29:38 <AWilcox[m]> They sure weren't kidding that this Matrix client is a beta.  It's very… primitive, but it does seem to work.  And I'm here without connecting to freenode :)
2017-12-01 23:31:02 <jirutka> AWilcox[m]: congratz :)
2017-12-01 23:33:18 <skarnet> AWilcox[m]: does that mean we could potentially see you in other channels, or is it a one-time kind of thing?
2017-12-01 23:34:49 <AWilcox[m]> I don't know what other channels Alpine's Matrix server has.
2017-12-01 23:34:58 <AWilcox[m]> And I have no idea how to get a list, heh.
2017-12-01 23:35:08 <skarnet> ah, I see.
2017-12-01 23:35:39 <AWilcox[m]> Hopefully this client is stable and I can use it and be here more permanently, but I really don't know.  It's C++ for the UI and C for the library, so I can more than likely fix it if it does break.
2017-12-01 23:36:04 <skarnet> is the server in C/C++ too?
2017-12-01 23:36:47 <AWilcox[m]> Since Alpine's server doesn't have encryption turned on I just made up a password and used matrix.org as a homeserver.  The server is still Python 2 AFAIK
2017-12-01 23:37:20 <skarnet> hmpf. Sorry, not running that on skarnet.org :/
2017-12-01 23:37:27 <AWilcox[m]> I'm not running my own.  Maybe when I get Adélie's infra going better we can run one and have a shared IRC/Matrix room like Alpine has.
2017-12-01 23:37:41 <AWilcox[m]> If the server situation improves, that is
2017-12-01 23:37:48 <AWilcox[m]> Still not packaging Py2 for Adélie :P
2017-12-01 23:38:08 <skarnet> I'm thinking of alternatives to bribe you back to #s6, but Python, even py3, is NOT an alternative :P
2017-12-01 23:38:48 <AWilcox[m]> I think jzk (from Atheme) is writing a server in C++.  But I'm not sure how far along it is.  I'll probably check that out some time early next year, so it has time to develop and I have time to take care of far more important things.
2017-12-01 23:58:53 <Shiz> hmm
2017-12-01 23:59:00 <Shiz> i wonder why alpine's server doesnt have crypto
2017-12-02 00:03:04 <AWilcox[m]> I think that would require the IRC people to have encryption as well.  It isn't TLS (everything is TLS on Matrix), it's literal signal-like per-message encryption.
2017-12-02 00:03:22 <Shiz> so e2e
2017-12-02 00:30:28 <AWilcox[m]> Yes.
2017-12-02 00:42:44 <Adran> alpine is switching to matrix? :'(
2017-12-02 00:47:24 <AWilcox[m]> See the topic; IRC isn't going away, but it provides people like me a way to connect and talk without dealing with freenode.
2017-12-02 04:39:46 <fcolista> jirutka, around?
2017-12-02 04:40:13 <fcolista> build for x86_64 is stucked because of a missing patch
2017-12-02 04:40:30 <fcolista> I'm going to update it
2017-12-02 12:46:00 <ollieparanoid[m]> congratulations to the 3.7 release and thanks for all the work everyone put into it! \o/
2017-12-02 12:55:26 <Shiz> \o/
2017-12-02 13:29:35 <PureTryOut[m]> \o/
2017-12-02 13:31:42 <PureTryOut[m]> A. Wilcox: what client do you use? also, a server doesn't "own" rooms, it can only be a part of one. I guess you could say a server "owns" a room if said room has only members of 1 server in it, but the moment a second server joins in, it loses it's "ownership"
2017-12-02 16:59:51 <BitL0G1c> can these 4 python libraries be pulled now ? https://github.com/alpinelinux/aports/pull/1546
2017-12-02 18:05:28 <npx> So I created a small Dockerfile to reduce this issue; https://github.com/hypoalex/reduction - cwebp-bin won't build and it is taking me to the sad place
2017-12-02 18:05:47 <npx> (on 3.7 with Node 8.9.1)
2017-12-02 18:11:00 <Xe> npx: on a -hardened kernel?
2017-12-02 18:11:39 <npx> Xe, I'm running that on Docker for Mac so I assume not (it's not running on an Alpine kernel)
2017-12-02 18:12:57 <npx> Linux eb8d3f5fbd7c 4.9.49-moby #1 SMP Wed Sep 27 23:17:17 UTC 2017 x86_64 Linux
2017-12-02 18:14:29 <Xe> cloning that repo locally and building
2017-12-02 18:16:04 <Xe> npx: weird, is cwebp a pre-built binary?
2017-12-02 18:16:49 <npx> I'm thinking of fixing it upstream by having it look for a `cwebp` binary and just using that instead of trying to build it, it should really just use the libwebp-tools package
2017-12-02 18:17:01 <npx> but no it's not a prebuilt binary
2017-12-02 18:18:25 <npx> 1.) look for /etc/alpine-release 2.) see if libwebp-tools is installed 3.) use it or say to install it
2017-12-02 18:20:14 <Xe> it looks like npm is falling back to building cwebp from source?
2017-12-02 18:20:35 <npx> Xe, yeah
2017-12-02 18:20:46 <npx> which actually did work in 3.6
2017-12-02 18:21:16 <npx> let me test that real quick to be sure it's a regression
2017-12-02 18:22:32 <npx> 3.6 doesn't automatically install nodejs-npm heh
2017-12-02 18:25:27 <npx> Xe, https://github.com/hypoalex/reduction/commit/671d0982b0b160ea8ab66aebd0bbc5377183f1c2 - this does work
2017-12-02 18:26:13 <npx> albeit with Node 6.10
2017-12-02 18:26:17 <Xe> npx: do you get the same failures on alpine:edge?
2017-12-02 18:26:31 <npx> checking
2017-12-02 18:28:12 <npx> annoyingly it doesn't actually throw an error, the build just hangs indefinitely
2017-12-02 18:29:08 <Xe> yeah
2017-12-02 19:10:34 <Xe> npx: i have no idea who to file a bug with, i'd start from the app you're trying to run
2017-12-02 19:11:18 <npx> well it's my app :P I opened an issue on their github
2017-12-02 19:12:05 <npx> Xe, thanks for you time though! I think my approach of detecting /etc/alpine-release and relying on the package manager is actually appropriate
2017-12-02 19:12:19 <npx> not sure if they'll merge it though
2017-12-02 19:19:39 <Xe> npx: the better solution is to fix the build
2017-12-02 19:19:47 <Xe> like, permanently
2017-12-02 19:23:25 <npx> oh, snap, I think I found the issue
2017-12-02 19:23:31 <npx> You inspired me :)
2017-12-02 19:24:24 <Xe> yay
2017-12-02 21:21:12 <AWilcox[m]> PureTryOut: Quaternion
2017-12-02 21:31:39 <_ikke_> ncopa: Still need to test it a bit more, but this basically works: http://tpaste.us/0N8N
2017-12-02 21:36:24 <danieli> perl -e '$/ = undef; exit 1 if (<> =~ /\x00/)'
2017-12-02 21:36:26 <danieli> I see what you did there
2017-12-02 21:37:56 <_ikke_> danieli: that was the best way I could find without relying on a gnu / non-busybox toolset
2017-12-02 21:38:20 <danieli> I figured
2017-12-02 21:45:12 <AWilcox[m]> Is Perl really lighter than coreutils?  (I'm not sure what this is for)
2017-12-02 21:46:31 <_ikke_> I don't know, and it doesn't matter to me. We can use whatever is available / acceptable
2017-12-02 21:46:48 <_ikke_> AWilcox[m]: it's a pre-receive hook that rejects commits with binary contents
2017-12-02 21:51:47 <AWilcox[m]> Ah, I see.
2017-12-02 21:52:46 <_ikke_> AWilcox[m]: reason: ncopa accidentaly commited a core dump
2017-12-02 21:53:27 <npx> Okay, libcwebp is pausing at "checking that generated files are newer than configure..."
2017-12-02 22:28:42 <npx> autotools are possibly the root of all evil
2017-12-02 22:37:12 <npx> Xe, https://github.com/hypoalex/reduction - working now, essentially the trick was to totally forgo the autotools build and use cmake - https://github.com/hypoalex/cwebp-bin/commit/8bdca621eb44a38d8e6a772ef55bae7e3be4269d
2017-12-02 22:37:33 <npx> Xe, although the proper fix imho is to use a musl-style shell script for ./configure
2017-12-02 22:37:50 <npx> (in libwebptools)
2017-12-02 23:32:47 <Xe> :+1:
2017-12-02 23:43:50 <Lochnair> npx: I had the same issue trying to bootstrap pkgconf. looks like it broke after the upgrade to bash 4.4, downgrading to 4.3 works
2017-12-03 02:37:24 <AWilcox[m]> The v3.7 repo (at least main on dl-3) has a 'tag' or whatever of v3.7.0_rc3+48
2017-12-03 02:37:28 <AWilcox[m]> Is that supposed to be that way?
2017-12-03 02:57:50 <kaniini> Lochnair: i'm open to alternative build systems, but nobody has proposed one that offers the same level of portability.  the proposals so far have been mostly linux-specific, e.g. no support for AIX/HPUX/etc.
2017-12-03 02:59:37 <kaniini> Lochnair: despite being absolutely hideous, the automake+libtool combination does generate a buildsystem that generates a working pkgconf on basically every  platform i have thrown at it
2017-12-03 13:19:01 <Lochnair> kaniini: yeah I can't say I like autotools, but it does get the job done.
2017-12-03 13:30:42 <Lochnair> the issue I'm seeing seems to be caused by automake not playing nice with the wait command changes in Bash 4.4
2017-12-03 13:40:01 <jirutka> Lochnair: why do you use bash?
2017-12-03 13:46:40 <Lochnair> jirutka: no reason really. I removed it when I found that it was causing issues, but it's pulled in by libtool, so when bootstrapping util-linux ./configure fails
2017-12-03 13:47:00 <jirutka> libtool does not depend on bash
2017-12-03 13:49:40 <jirutka> uh, it does
2017-12-03 13:49:42 <jirutka> wtf
2017-12-03 13:50:19 <jirutka> ok, sry, i didn’t notice
2017-12-03 13:50:57 <jirutka> Lochnair: could you please test if it works with ZSH?
2017-12-03 13:55:52 <Lochnair> tried that now. running ./configure ... from a zsh shell doesn't help. looking at top the configure script still runs in bash (for some reason)
2017-12-03 13:56:02 <Lochnair> changing the hashbang in ./configure from /bin/sh to /bin/zsh does work
2017-12-03 14:25:24 <Shiz> what if you do
2017-12-03 14:25:26 <Shiz> ash ./configure
2017-12-03 15:01:43 <Lochnair> Shiz: no dice. looks like CONFIG_SHELL is the variable that contains the shell to run. if I force it to run /bin/sh, ./configure completes successfully.
2017-12-03 15:01:49 <Lochnair> relevant parts of ./configure: https://pastebin.com/LY1RAcJq
2017-12-03 15:38:25 <Lochnair> okay, so calling configure like this does work: CONFIG_SHELL=/bin/sh ./configure
2017-12-03 18:13:53 <leo-unglaub> it has been a long time that i used xorg on alpine, was there a special trick to get it working? i have installed i3wm and put exec ie into my .xinitrc
2017-12-03 18:14:09 <leo-unglaub> i also added xf86-video-ati because i have a radeon RX 480
2017-12-03 18:14:11 <Xe> `startx`?
2017-12-03 18:14:37 <leo-unglaub> when i startx into it i gt the following error message: (EE) no screens found
2017-12-03 18:14:49 <leo-unglaub> i also did setup-xorg-base
2017-12-03 18:22:00 <leo-unglaub> https://pastebin.ca/3943792
2017-12-03 18:22:03 <leo-unglaub> this is the error log
2017-12-03 18:30:31 <jirutka> hm, I tried it few months ago for a conference and it worked
2017-12-03 18:42:41 <leo-unglaub> hmmm
2017-12-03 18:42:44 <leo-unglaub> damn
2017-12-03 18:43:09 <leo-unglaub> i have the fact that 20 years after developing xorg we still cannot get decent error messages from that thing :(
2017-12-03 18:44:12 <Xe> is the mmc module in alpine 3.7's kernel?
2017-12-04 00:30:16 <tmh1999> is anyone able to change regulatory country code with CRDA + mdev config in Alpine ? My default is set to world domain and is kind of having limited number of operating channels and frequencies.
2017-12-04 00:44:47 <tmh1999> or even eudev or udev
2017-12-04 00:59:20 <skarnet> or even mdevd
2017-12-04 01:55:10 <jirutka> mdevd is not in Alpine yet, but I already know how to write the init script for it, just haven’t time to do it yet
2017-12-04 02:00:16 <skarnet> you're right, I should package it
2017-12-04 02:00:22 <skarnet> will probably do it this week
2017-12-04 11:54:48 <jirutka> skarnet: why? as I wrote, I’m already working on it
2017-12-04 11:56:06 <skarnet> because I usually package my software for Alpine, but if you want to be the maintainer for mdevd, I have no objection :P
2017-12-04 12:49:42 <jirutka> who’s Kaarle Ritvanen? pushed broken abuild into testing
2017-12-04 12:50:16 <jirutka> like totally broken
2017-12-04 12:50:17 <jirutka> wtf
2017-12-04 12:50:22 <jirutka> I’m just gonna revert this crap
2017-12-04 12:51:14 <jirutka> hm, wait, local sources are not copied into distfiles, so there should not be file conflict; so it’s ”just” wrong checksum
2017-12-04 13:04:49 <jirutka> btw I can’t even blame him, storing checksums for local sources is non-sense, we should fix this in aports
2017-12-04 13:04:55 <jirutka> it causes just problems
2017-12-04 13:26:52 <przemoc> jirutka: kunkku
2017-12-04 13:57:03 <ncopa> storing a checksum for local sources is an extra check to verify that you actually push the version of it that you "build" locally
2017-12-04 13:57:21 <ncopa> so you dont build and commit something localy
2017-12-04 13:57:23 <ncopa> test it
2017-12-04 13:57:33 <ncopa> fix an error
2017-12-04 13:57:52 <ncopa>  and push (and push the first, broken version)
2017-12-04 13:58:25 <skarnet> or to verify that you don't commit a core file
2017-12-04 13:58:32 <skarnet> 
ACTION is already out

2017-12-04 13:58:32 <_ikke_> ncopa: did you see I mostly finished the pre-receive hook that rejects binary files?
2017-12-04 13:58:42 <_ikke_> skarnet: lol, timing :D
2017-12-04 13:58:55 <clandmeter> ncopa: do we still need to set the stack size for some py modules? or did that python fix fix it?
2017-12-04 13:59:13 <ncopa> calndmeter: that python fix should fix it
2017-12-04 13:59:22 <clandmeter> thx
2017-12-04 13:59:43 <ncopa> skarnet: unfortunally we dont check/count all the files in the local dir
2017-12-04 14:00:13 <_ikke_> ncopa: Still need to test it a bit more, but this basically works: http://tpaste.us/0N8N
2017-12-04 14:00:19 <ncopa> because just like jirutka, someone else didnt like that we had to store a checksum of pre/post install scripts
2017-12-04 14:00:25 <_ikke_> modulo the exit command at the end
2017-12-04 14:00:27 <ncopa> so we dont have the install scritps in checksums
2017-12-04 14:01:12 <ncopa> so we dont have the install scritps in checksums. it test if there are any \0x00?
2017-12-04 14:01:40 <_ikke_> ncopa: it tests if there are NUL bytes in the first 1024 bytes
2017-12-04 14:01:51 <_ikke_> Which seems to be a common threshold
2017-12-04 14:03:02 <skarnet> brb crafting a legit binary file that will defeat your test
2017-12-04 14:03:16 <_ikke_> skarnet: sure, easyily done
2017-12-04 14:03:27 <clandmeter> cant we make an allowed list instead of a reject list?
2017-12-04 14:03:46 <_ikke_> clandmeter: allowed list of what?
2017-12-04 14:03:58 <clandmeter> things we like to store in aports
2017-12-04 14:04:30 <clandmeter> and warn if we do something unusual
2017-12-04 14:07:36 <clandmeter> im just thinking out loud, because i dont really like the perl approach/dep
2017-12-04 14:08:01 <_ikke_> clandmeter: Can also be gnu grep or anything else
2017-12-04 14:08:14 <_ikke_> lua if that works
2017-12-04 14:08:47 <_ikke_> none of the busybox tools at least seem to support NUL bytes
2017-12-04 14:13:07 <ncopa> _ikke_: is_binary () { hexdump -n 1024 "$@" | grep -q -w 0000; }
2017-12-04 14:14:00 <_ikke_> Is that reliable enough?
2017-12-04 14:14:51 <ncopa> oh, no
2017-12-04 14:16:43 <ncopa> hexdump -b | grep -w 000
2017-12-04 14:17:13 <ncopa> should catch any single NUL char
2017-12-04 14:17:48 <clandmeter> what if we just test by file size?
2017-12-04 14:18:17 <_ikke_> ncopa: that also gives the hexdump as output
2017-12-04 14:18:34 <ncopa> hexdump -b | grep -q -w 000
2017-12-04 14:19:14 <_ikke_> does not work on the corefile you comitted
2017-12-04 14:19:29 <skarnet> you need to strip the first column
2017-12-04 14:19:29 <_ikke_> or is the meaning reverted?
2017-12-04 14:19:50 <skarnet> the first column has lots of 000
2017-12-04 14:19:52 <_ikke_> right
2017-12-04 14:20:05 <ncopa> skarnet: thats what -w is supposed to catch
2017-12-04 14:20:20 <_ikke_> ncopa: I still see it
2017-12-04 14:20:30 <_ikke_> remote: 00002e0 001 000 000 000 006
2017-12-04 14:20:40 <skarnet> -w is not standard :P
2017-12-04 14:21:01 <_ikke_> default hexdump does not list it
2017-12-04 14:22:55 <przemoc> _ikke_: git show 80fcfcf217c2b5ab4f321469f77eb9f3214bb7d9^:community/webkit2gtk/core | hexdump -b -n 1024 | grep -q -w 000 && echo "it seems to be biinary file"
2017-12-04 14:22:58 <przemoc> it seems to be biinary file
2017-12-04 14:23:44 <przemoc> s/ii/i/
2017-12-04 14:23:47 <_ikke_> Right, the perl script exited 1 if NUL was found
2017-12-04 14:23:55 <_ikke_> So the meaning is reverted
2017-12-04 14:24:04 <skarnet> trying to get my head around the hexdump format
2017-12-04 14:24:22 <przemoc> 0 is success in unix
2017-12-04 14:24:23 <skarnet> to get it to dump chars, not offsets
2017-12-04 14:24:34 <skarnet> so you can use "grep -qF 000"
2017-12-04 14:24:41 <skarnet> -> faster check
2017-12-04 14:25:14 <_ikke_> http://tpaste.us/Dky0
2017-12-04 14:25:43 <_ikke_> przemoc: right
2017-12-04 14:25:58 <ncopa> except that function name is_binary is misleading
2017-12-04 14:26:00 <clandmeter> files larger then specific size would better be hosted outside of aports. (this would also catch core files)
2017-12-04 14:26:23 <ncopa> if is_binary ...; then echo "this is infact not binary"; fi
2017-12-04 14:26:25 <_ikke_> http://tpaste.us/BeNn
2017-12-04 14:26:45 <_ikke_> I reverted it
2017-12-04 14:26:52 <ncopa> yup
2017-12-04 14:26:54 <_ikke_> s/re/in/
2017-12-04 14:27:07 <ncopa> if git show $commit:"$file" | head -c1024 | is_binary; then
2017-12-04 14:27:14 <ncopa> works too
2017-12-04 14:27:47 <_ikke_> nod
2017-12-04 14:31:22 <_ikke_> http://tpaste.us/bKzD
2017-12-04 14:31:23 <przemoc> skarnet: good idea.  hexdump -e '1/1 "%02x\n"' does the job
2017-12-04 14:31:43 <przemoc> with grep -qF 00, obviously
2017-12-04 14:32:02 <ncopa> hexdump -e '1/1 "%02x""\n"'
2017-12-04 14:32:09 <ncopa> i was testing the same :)
2017-12-04 14:32:45 <ncopa> algitbot: good girl!
2017-12-04 14:32:48 <skarnet> cool, that works
2017-12-04 14:33:11 <skarnet> I will now forget that format was ever invented
2017-12-04 14:34:30 <ncopa> huh
2017-12-04 14:35:00 <ncopa> i thought algibit triggered on new emails to patch work and was impressed of that she learned new tricks
2017-12-04 14:35:14 <ncopa> but it appears it triggered on %02
2017-12-04 14:35:20 <ncopa> bah
2017-12-04 14:35:34 <skarnet> not such a good girl after all \o/
2017-12-04 14:35:39 <clandmeter> you just added soemthing on your todolist!
2017-12-04 14:35:40 <ncopa> yeah
2017-12-04 14:35:49 <ncopa> algitbot: bad girl!
2017-12-04 14:36:15 <przemoc> at least it triggered, ~2.5 years, but it works!
2017-12-04 14:44:32 <consus> Hooray, 3.7 is out ^^
2017-12-04 14:44:49 <consus> Special thanks for openssh-pam
2017-12-04 14:48:21 <ncopa> :)
2017-12-04 14:49:38 <ncopa> jirutka: remember that lua install_if issue? i think we figured out that, thanks to kunkku #8237
2017-12-04 14:51:11 <_ikke_> ncopa: But is that script something you could use then?
2017-12-04 14:51:27 <ncopa> problem is that when there are two different packages with provides=cmd:something, weird things happens
2017-12-04 14:51:32 <ncopa> _ikke_: i think so yes
2017-12-04 14:51:59 <_ikke_> ncopa: You also asked about checking if all patches are supplied and exist
2017-12-04 14:52:32 <ncopa> _ikke_: is this a post-receive hook?
2017-12-04 14:52:37 <_ikke_> pre-receive
2017-12-04 14:52:43 <_ikke_> post-receive is too late to reject
2017-12-04 14:52:49 <ncopa> yeah
2017-12-04 14:53:14 <ncopa> and we already have a post-receive
2017-12-04 14:54:32 <_ikke_> If you have multiple hooks, you can create a 'hook runner'
2017-12-04 15:30:46 <jirutka> _ikke_: this is how I test binary file in Lua https://github.com/jirutka/luapak/blob/master/luapak/fs.lua#L66-L88
2017-12-04 15:31:56 <_ikke_> Right. We would need to create a separate executable for that then
2017-12-04 15:33:11 <jirutka> ncopa: yes, I remember that; what was the cause?
2017-12-04 15:33:47 <jirutka> ncopa: "ncopa: problem is that when there are two different packages with provides=cmd:something, weird things happens" … wtf? this is perfectly normal and common situation
2017-12-04 15:34:17 <jirutka> ncopa: wait, but in that situation these packages should not be installed simultaneously
2017-12-04 15:34:26 <jirutka> ncopa: but we handle this in py[3
2017-12-04 15:34:48 <jirutka> in py[23]- packages… at least I do care about it… of course not everyone :(, so they may be conflicting cmd
2017-12-04 15:35:17 <jirutka> ncopa: but definitely not in case of lua-busted
2017-12-04 15:36:03 <Shiz> we've got a cmd provides now?
2017-12-04 15:36:06 <Shiz> cool
2017-12-04 15:36:16 <kunkku> I'm not sure if the reason is conflicting provides
2017-12-04 15:36:29 <kunkku> see my last update to the issue
2017-12-04 15:36:58 <kunkku> py-unidecode has -2 ja -3 suffixes in the cmd
2017-12-04 15:37:14 <kunkku> but install_if is still failing
2017-12-04 15:38:26 <kunkku> also py-django has issues
2017-12-04 15:39:00 <kunkku> but yes, all packages that have problems seem to have cmd provides
2017-12-04 15:47:23 <jirutka> coffee time!
2017-12-04 15:59:12 <jirutka> ncopa: "ncopa:	storing a checksum for local sources is an extra check to verify that you actually push the version of it that you "build" locally" … hmm, that makes sense
2017-12-04 16:03:28 <Shiz> ncopa: _ikke_: jirutka: do we want to start an effort to relicense aports to MIT?
2017-12-04 16:03:52 <jirutka> Shiz: what exactly do you mean with aports?
2017-12-04 16:03:56 <Shiz> the apkbuilds
2017-12-04 16:03:58 <Shiz> well
2017-12-04 16:04:00 <Shiz> the contents of the aports tree
2017-12-04 16:04:02 <Shiz> :P
2017-12-04 16:04:11 <jirutka> yes
2017-12-04 16:04:14 <jirutka> currently it’s unclear
2017-12-04 16:04:37 <Shiz> we may not be able to relicense all the patches but I think that's fine
2017-12-04 16:04:46 <jirutka> but I’m not sure how to do that
2017-12-04 16:04:47 <Shiz> at long as we get all the APKBUILD stuff sorted
2017-12-04 16:04:49 <Shiz> well
2017-12-04 16:04:52 <Shiz> we start an effort like mpv did
2017-12-04 16:04:55 <Shiz> contact all contributors
2017-12-04 16:04:58 <Shiz> asking them to license their stuff
2017-12-04 16:05:17 <Shiz> https://github.com/mpv-player/mpv/issues/2033
2017-12-04 16:13:46 <vmele> Hi all, how should I create a github pull on aports when I have a package that depends on another new package? I have sssd packaged, but it depends on ding-libs. Right now I have separate pull requests because they are different packages, but sssd won't build with Travis because it requires ding-libs.
2017-12-04 16:15:03 <jirutka> vmele: two commits, one PR
2017-12-04 16:15:19 <vmele> got it, thanks
2017-12-04 16:18:04 <kaniini> the APKBUILD would be under the same license as the upstream package.  i am opposed to relicensing my APKBUILDs to MIT.
2017-12-04 16:19:17 <_ikke_> kaniini: how can the APKBUILD be under the same license?
2017-12-04 16:19:59 <kaniini> it is a matter of being polite and common sense, scripts to package a software should follow the same license as the software it is packaging.
2017-12-04 16:20:08 <kaniini> that way, if upstream wishes to use the APKBUILD, they could.
2017-12-04 16:20:26 <_ikke_> kaniini: But is it actually working that way?
2017-12-04 16:20:52 <kaniini> in every other distribution that i am involved with, it is.
2017-12-04 16:21:01 <Shiz> I highly disagree
2017-12-04 16:21:16 <jirutka> me too; this is IMO not legally correct
2017-12-04 16:21:18 <Shiz> for MIT every upstream can use the APKBUILD anyway
2017-12-04 16:21:22 <kaniini> i really do not care, relicense my APKBUILDs and you will hear from my lawyers
2017-12-04 16:21:26 <jirutka> and there are already many people asking us about that
2017-12-04 16:21:28 <_ikke_> afaik, the copyright lies with the person who created the APKBUILD
2017-12-04 16:21:34 <Shiz> yes it does
2017-12-04 16:21:40 <Shiz> kaniini: then they can simply be replaced
2017-12-04 16:21:42 <Shiz> ¯\_(ツ)_/¯
2017-12-04 16:21:56 <kaniini> i suggest you get on that then
2017-12-04 16:22:12 <kaniini> because, i do not want my APKBUILDs licensed under MIT
2017-12-04 16:22:16 <jirutka> I personally don’t care much about the specific solution, but the fact is that the current state is not legally correct and we should fix it
2017-12-04 16:26:19 <Shiz> trying to strongarm your way with big words about lawyers while nobody was even talking about relicensing against/without consent is extremely counterproductive
2017-12-04 16:26:31 <jirutka> ^ +1
2017-12-04 16:26:32 <_ikke_> kaniini: just curious, why are you so vehemently opposed to that?
2017-12-04 16:27:47 <skarnet> my position is "if upstream wishes to use the APKBUILD" -> then they're distro developers, they're not "upstream"
2017-12-04 16:28:01 <skarnet> upstream should not have to care how distros package their stuff
2017-12-04 16:28:12 <Shiz> some upstreams do anyway, and i think that's reasonable
2017-12-04 16:28:26 <kaniini> either way, i want my APKBUILDs to be licensed under whatever is in $license field
2017-12-04 16:28:51 <skarnet> reasonable? what's to stop a distro from saying "why do you provide scripts for such-other-distro and not for us? we're not packaging your stuff if you exhibit such favoritism"
2017-12-04 16:28:52 <kaniini> if you want to do something else, i will orphan all of my packages effective immediately
2017-12-04 16:28:56 <jirutka> kaniini: could you get us some legal document from any distribution that uses this model – package descriptors licensed under the same license as the packaged software? I can’t imagine how it can be declared
2017-12-04 16:28:57 <_ikke_> kaniini: Then you should explicitly mention that
2017-12-04 16:29:13 <skarnet> and it puts the burden of packaging on upstream, which is totally unreasonable
2017-12-04 16:29:14 <_ikke_> kaniini: right now, the APKBUILD are basically unlicensed
2017-12-04 16:29:18 <kaniini> _ikke_: i did
2017-12-04 16:29:23 <skarnet> I can't make packaging scripts for every distro out there
2017-12-04 16:29:57 <Shiz> skarnet: i'm not gonna ask upstream to do it, but some upstreams want to do it for some unfathomable reason
2017-12-04 16:30:05 <jirutka> _ikke_: and that means that the same copyright law as for book applies to it, at least in Czech Republic; so basically no one can legally use it, modify and what’s worse, you’re responsible for damage
2017-12-04 16:30:31 <skarnet> those upstreams make it bad for everyone else :(
2017-12-04 16:30:57 <jirutka> kaniini: please, instead of big words and threatening, show us some legal document that describes this licensing model
2017-12-04 16:31:13 <jirutka> kaniini: the fact is what _ikke_ said, currently APKBUILDs are unlicensed and that’s not good
2017-12-04 16:31:15 <_ikke_> I'm curious if the license field in an APKBUILD has any legal meaning
2017-12-04 16:31:20 <jirutka> kaniini: we have to fix it, somehow
2017-12-04 16:31:33 <_ikke_> or just meta data
2017-12-04 16:31:36 <jirutka> _ikke_: it does not have if we don’t have any legal document giving it that meaning
2017-12-04 16:31:42 <kaniini> we "have" to fix it ?
2017-12-04 16:31:52 <kaniini> we seem to be doing just fine
2017-12-04 16:32:04 <_ikke_> kaniini: until we aren't
2017-12-04 16:32:21 <Shiz> as it stands, the aports repo is not redistributable
2017-12-04 16:32:31 <Shiz> or tbh, even subject to modifications
2017-12-04 16:32:37 <Shiz> legally speaking
2017-12-04 16:33:14 <jirutka> kaniini: no we’re not; ask a lawyer
2017-12-04 16:35:16 <jirutka> I can imagine that we say all abuilds and scripts in aports are by default licensed under MIT “copyright (c) Alpine Linux community” (that is also problematic, but leave it to later), if it’s not said opposite; and then when some authors, like kaniini don’t agree, the author can specify different license in his/her APKBUILD
2017-12-04 16:36:10 <jirutka> (or any other license than MIT)
2017-12-04 16:36:15 <kaniini> oh, i see, so you want to impose a license, but introduce another problem in the process
2017-12-04 16:36:23 <kaniini> thus making the license invalid
2017-12-04 16:36:26 <jirutka> ?
2017-12-04 16:36:53 <przemoc> copyright (c) Alpine Linux community  would require transfering rights if I'm not mistaken
2017-12-04 16:36:59 <_ikke_> it does
2017-12-04 16:37:01 <jirutka> yes
2017-12-04 16:37:01 <Shiz> yes
2017-12-04 16:37:03 <kaniini> entities that do not exist cannot license away rights
2017-12-04 16:37:10 <kaniini> because they do not hold any rights
2017-12-04 16:37:13 <jirutka> yes
2017-12-04 16:37:17 <kaniini> "Alpine Linux community" is not a legal entity
2017-12-04 16:37:18 <jirutka> that’s the problem I meant
2017-12-04 16:37:18 <Shiz> you can't simply relicense without active consent
2017-12-04 16:37:28 <Shiz> >if it's not said opposite
2017-12-04 16:37:30 <Shiz> that doesn't work
2017-12-04 16:37:32 <skarnet> "one does not simply"
2017-12-04 16:37:33 <skarnet> please
2017-12-04 16:37:39 <Shiz> it doesn't hold legal water
2017-12-04 16:37:39 <skarnet> Shiz :(
2017-12-04 16:37:46 <jirutka> I know, but we need to ask all contributors anyway, so it will be better to do it right
2017-12-04 16:37:50 <jirutka> like Gentoo has it
2017-12-04 16:37:53 <jirutka> or BSDs IIRC
2017-12-04 16:38:02 <Shiz> skarnet: 2007 memes are stale
2017-12-04 16:38:11 <Shiz> get with the times old man
2017-12-04 16:38:13 <Shiz> ;p
2017-12-04 16:38:16 <kaniini> those have foundations
2017-12-04 16:38:26 <jirutka> yes…
2017-12-04 16:38:27 <kaniini> i keep pointing out that we need one
2017-12-04 16:38:31 <skarnet> the youth of a meme does not its quality make
2017-12-04 16:38:32 <jirutka> me too
2017-12-04 16:38:37 <jirutka> that’s a good opportunity
2017-12-04 16:38:37 <kaniini> but hey, nobody listens
2017-12-04 16:38:49 <Shiz> oh it does
2017-12-04 16:38:58 <Shiz> unless you wanna say that rickroll or the game is still funny
2017-12-04 16:39:00 <Shiz> :p
2017-12-04 16:39:02 <kaniini> either way, i do not consent to having my APKBUILDs licensed under MIT at this time
2017-12-04 16:39:17 <jirutka> actually I already talked with some friends who has experience with this and thinking about creating it, cause ncopa doesn’t seem to be willing do it
2017-12-04 16:39:58 <kaniini> i do, however, consent to transfer the copyrights to such foundation if created
2017-12-04 16:40:02 <jirutka> but I’m too busy even without this bureaucracy right now
2017-12-04 16:40:05 <kaniini> which could then license them however it wishes
2017-12-04 16:40:34 <skarnet> where are the bureaucrats when we need them
2017-12-04 16:40:42 <kaniini> so tl;dr my opinion is i would rather have status quo (like most distributions) or do it properly
2017-12-04 16:40:55 <jirutka> status quo = illegal ?
2017-12-04 16:41:04 <kaniini> yes
2017-12-04 16:41:15 <jirutka> oh, really good and constructive solution
2017-12-04 16:41:16 <kaniini> but my view on it is this
2017-12-04 16:41:25 <kaniini> the binary APKs are obviously under the upstream license
2017-12-04 16:41:29 <jirutka> yes
2017-12-04 16:41:39 <kaniini> so it makes sense that the source of the packaging would be too
2017-12-04 16:41:45 <jirutka> actually no
2017-12-04 16:41:54 <Shiz> I disagree
2017-12-04 16:42:00 <Shiz> the source of the program, sure
2017-12-04 16:42:07 <Shiz> the packaging recipe is part of the build system like abuild is imo
2017-12-04 16:42:23 <kaniini> the source of the packaging includes patches
2017-12-04 16:42:23 <_ikke_> I think it's a separate work
2017-12-04 16:42:31 <jirutka> yeah, separate work
2017-12-04 16:42:32 <kaniini> are those patches under MIT too ?
2017-12-04 16:42:34 <Shiz> yes, the patches are a different matter
2017-12-04 16:42:36 <Shiz> and I stated that above
2017-12-04 16:42:42 <kaniini> i disagree
2017-12-04 16:42:52 <_ikke_> patches can be seen as a derived work
2017-12-04 16:43:12 <kaniini> a directory containing APKBUILD and ancillary files must be the same license as the binary package they create
2017-12-04 16:43:24 <_ikke_> kaniini: That's not how it works
2017-12-04 16:43:30 <_ikke_> each file can have a different license
2017-12-04 16:43:36 <kaniini> it really is
2017-12-04 16:43:39 <_ikke_> and the copyright can belong to different persons
2017-12-04 16:43:49 <kaniini> sure, they can, and taken separately that is valid
2017-12-04 16:44:02 <kaniini> but the combined work must absolutely be whatever the upstream license is
2017-12-04 16:44:10 <Shiz> this is not a combined work
2017-12-04 16:44:14 <_ikke_> kaniini: any legal precedence for that?
2017-12-04 16:44:49 <_ikke_> a project like git is a combined work, but each file has it's own copyright\
2017-12-04 16:44:54 <Shiz> btw, isn't what Gentoo does another precedence -- license everything under GPL ;p
2017-12-04 16:45:04 <Shiz> in the ebuild rpeo
2017-12-04 16:45:06 <kaniini> i'm sure you can find some data to this effect on debian-legal
2017-12-04 16:45:20 <kaniini> yes, but Gentoo does not generate binary packages
2017-12-04 16:45:21 <jirutka> kaniini: then please find it
2017-12-04 16:45:27 <kaniini> it just installs shit into your filesystem
2017-12-04 16:45:53 <Shiz> incorrect
2017-12-04 16:45:56 <Shiz> there's BINHOST
2017-12-04 16:46:07 <Shiz> a well-supported feature
2017-12-04 16:46:08 <kaniini> yeah, a feature nobody uses
2017-12-04 16:46:21 <kunkku> afaik, there is little case law regarding OSS licenses
2017-12-04 16:47:05 <_ikke_> can an APKBUILD be seen as a derivative work of the 'prototype' APKBUILD?
2017-12-04 16:47:24 <kunkku> many developers have strong opinions on whether X is allowed by Y license etc., but no one can be certain about it
2017-12-04 16:47:30 <jirutka> sigh, probably good time to meet our law teacher at faculty
2017-12-04 16:48:16 <kaniini> https://fedoraproject.org/wiki/Licensing:Main#License_of_Fedora_SPEC_Files
2017-12-04 16:48:21 <Shiz> i know public binhosts
2017-12-04 16:48:42 <kunkku> lawsuits are like quantum mechanics: a certain outcome may be more probable than others but probability is never 1
2017-12-04 16:48:47 <kaniini> all fedora spec files are MIT, but that is because they have a CLA in place for every contributor
2017-12-04 16:48:48 <Shiz> looks like fedora agrees with me
2017-12-04 16:49:03 <kaniini> meaning that what jirutka proposes isn't really possible without CLA
2017-12-04 16:49:04 <_ikke_> kaniini: right, they have covered their basis
2017-12-04 16:49:16 <_ikke_> kaniini: I think nobody disagreed with that
2017-12-04 16:49:21 <Shiz> I don't see how this follows
2017-12-04 16:49:32 <Shiz> "X uses a CLA" does not imply "a CLA is necessary and/or vital"
2017-12-04 16:49:45 <Shiz> we all know that fedora is very legal :P
2017-12-04 16:50:02 <Shiz> but having a statement that by contributing to the repository you license oyur contributions under MIT unless explicitly otherwise specified
2017-12-04 16:50:07 <Shiz> should cover it enough
2017-12-04 16:50:08 <_ikke_> Shiz: without a CLA, the copyright defaults to the author
2017-12-04 16:50:11 <jirutka> kaniini: yes, that’s an implication; IIRC you agree with Gentoo CLA when you create an account on bugzilla or something like that
2017-12-04 16:50:54 <kaniini> https://www.debian.org/doc/packaging-manuals/copyright-format/1.0/#header-paragraph
2017-12-04 16:51:08 <kaniini> each package in debian specifies it's own licensing in this way
2017-12-04 16:51:42 <przemoc> sometimes you have patches that upstream won't accept because you're not willing to, for instance, transfer your rights to them or relicense your additions as they wish, yet you may be willing to apply them in distro as there is no conflict between licenses.  mind that not all software has all files with only one license.  tieing packaging and patches license to upstream's one seems a bit too much
2017-12-04 16:51:48 <przemoc> (BTW which one license if there are many? the one which is used in most files?).  it seems easy-going attitude, and maybe not that bad, but it's quite restrictive for no very good reasons.
2017-12-04 16:53:17 <kunkku> so the relevant question here is whether aports constitutes a single piece of work or each directory is a work on its own
2017-12-04 16:53:37 <Shiz> I'm not even sure if that's relevant, tbh
2017-12-04 16:54:10 <kaniini> fix it properly, or don't fix it at all
2017-12-04 16:54:17 <kaniini> is basically my view
2017-12-04 16:54:57 <jirutka> kaniini: so create a foundation and license aports to the foundation, or keep it as-is?
2017-12-04 16:55:24 <vmele> kaniini: the problem is that there isn't a single answer or fix. There isn't legal precedent either way. Different projects have different structures because they may each be right.
2017-12-04 16:56:25 <kaniini> jirutka: yes
2017-12-04 16:57:00 <kunkku> Fedora does not require you to assign copyrights
2017-12-04 16:57:15 <kunkku> they just require you to license your contributions
2017-12-04 16:57:24 <jirutka> kaniini: do you have any special requirements to the foundation?
2017-12-04 16:57:33 <jirutka> s/to/for/
2017-12-04 16:57:38 <kaniini> jirutka: yes, i want it to have ncopa's blessing
2017-12-04 16:57:47 <kaniini> jirutka: other than that, no
2017-12-04 16:57:51 <jirutka> kaniini: of course
2017-12-04 16:58:28 <kaniini> kunkku: same difference, need a foundation to hold the CLAs
2017-12-04 16:59:41 <kaniini> kunkku: assigning CLAs to a single individual is worse situation than we presently have.  if that individual is hit by a bus, then we have to do it all over again most likely.
2017-12-04 17:01:11 <kaniini> however,
2017-12-04 17:01:39 <kaniini> with the CLA-only model, it is important to record all major contributors to each APKBUILD
2017-12-04 17:02:09 <kaniini> because then only the entire aports repository is covered under MIT
2017-12-04 17:02:24 <jirutka> basically it’s needed for all contributors to sign CLA; this can be just simple HTML form with checkbox
2017-12-04 17:02:52 <kaniini> well, i do not know that a "clickwrap" CLA is acceptable actually.
2017-12-04 17:03:05 <kaniini> where i live, such things have no legal effect
2017-12-04 17:03:35 <jirutka> that’s what all software projects I know do
2017-12-04 17:03:44 <jirutka> even these based on USA
2017-12-04 17:04:09 <jirutka> s/based on/based in/
2017-12-04 17:05:00 <kaniini> oh, i see
2017-12-04 17:05:31 <kaniini> the situation where i live is that the windows clickwrap agreement is not legally binding because you wouldn't be able to use your computer without agreeing
2017-12-04 17:05:36 <vmele> jirutka's simple HTML form suggestion is not a "clickwrap" agreement
2017-12-04 17:05:57 <jirutka> hm, probaly i don’t know what “clickwrap” actually means
2017-12-04 17:06:13 <vmele> kaniini: what jurisdiction is that? In the US there was a case about the dell clickwrap, and it's been fixed by having the terms on the outside of the packages
2017-12-04 17:06:57 <vmele> a simple HTML form likely fully complies with the e-sign requirements in the US
2017-12-04 17:07:33 <Xe> hot take: why is the license of package build scripts even relevant? who cares?
2017-12-04 17:08:51 <vmele> Xe: I fully agree with you. Each script is arguably its own separate work. But I believe that the Alpine Project needs to make a decision as to how many developers (maybe legally incorrectly) will be scared off from contributing. This isn't necessarily a "is Alpine or its users in violation" question; this is more of a "make developers happy" question.
2017-12-04 17:09:57 <vmele> (full disclosure: I am a licensed US attorney that has been general counsel for an open source project)
2017-12-04 17:11:11 <jirutka> ha, perfects, we ha a lawyer here! :)
2017-12-04 17:11:24 <Xe> vmele: the way i see it, contributing to open source is a voulnteer thing I do out of my own effort, on my own time and explicitly in favor of benefitting people other than myself. Why should I feel like I need to own what I have given away?
2017-12-04 17:11:43 <kaniini> what Xe says really
2017-12-04 17:12:03 <kaniini> (when it comes to packaging scripts anyway)
2017-12-04 17:12:09 <jirutka> Xe: it’s not about needing to own what you have given away, excatly opposite
2017-12-04 17:12:20 <kunkku> isn't protection against IPR trolls a good enough reason to care about this?
2017-12-04 17:12:38 <jirutka> Xe: currently you fully own it and in some countries you’re even responsible for damage that your script may potentially create
2017-12-04 17:12:39 <kaniini> i don't care about the licensing of my APKBUILDs in all reality, just that it be done correctly
2017-12-04 17:12:56 <kaniini> kunkku: yes, protection against IP trolls seems like a good reason to have a foundation too
2017-12-04 17:13:25 <kaniini> the key thing is having a legal entity to represent the project and its assets (such as the aports repo)
2017-12-04 17:14:13 <kaniini> but i am not a lawyer and maybe i misunderstand the benefits of having a legal entity in this case
2017-12-04 17:14:46 <kaniini> but the idea i have is that the legal entity handles all licensing for the project and is also the thing that the IP troll would be suing
2017-12-04 17:14:51 <kaniini> instead of individual contributors
2017-12-04 17:15:23 <kaniini> maybe vmele can comment on that
2017-12-04 17:15:36 <kaniini> since they mentioned they are a lawyer
2017-12-04 17:15:37 <vmele> I'm just suggesting that a foundation / legal entity isn't necessarily the only way to do that. It certainly has benefits, but it has costs that maybe the project isn't ready to bare. Requiring a very permissive license (like MIT) for apports is an easy and inexpensive way to avoid patent troll poisoning of the tree (in this circumstance)
2017-12-04 17:15:37 <kaniini> :D
2017-12-04 17:15:49 <vmele> I'm not Alpine or any of *your* lawyers ;-)
2017-12-04 17:15:55 <kunkku> isn't the main use of CLAs to redirect the trolls after the individuals rather than the legal entity? :)
2017-12-04 17:16:30 <kunkku> most of them have clauses to that effect
2017-12-04 17:16:47 <kaniini> MIT license doesn't discuss patents though
2017-12-04 17:17:11 <kaniini> only stuff like Apache-2 and GPL do
2017-12-04 17:17:34 <kaniini> at best it would only protect against that netfilter person
2017-12-04 17:17:40 <kaniini> well, something like what that netfilter person is doing i should say
2017-12-04 17:17:59 <kaniini> (if you don't know, there is some netfilter dev in germany suing companies for vague GPL violations)
2017-12-04 17:19:35 <kunkku> his original works tend to be licensed under GPLv3...
2017-12-04 17:20:17 <kaniini> the netfilter person?
2017-12-04 17:20:44 <vmele> I wasn't aware of the netfilter dev issue. As to aports (which is the only think we're talking about here, right?), if you submitted an aport to build netfilter and it was licensed MIT, you are providing a license to anyone who uses that script. It doesn't matter if that's been assigned to a foundation or not. Is the netfilter person saying that using the script triggers GPL3? Because that's one hell of a stretch of derivative work.
2017-12-04 17:22:04 <kunkku> vmele: are you aware of any case law that would define "derivative work" in OSS context?
2017-12-04 17:23:39 <kaniini> fwiw i would prefer ISC license instead of MIT if there is a licensing of aports
2017-12-04 17:23:39 <bpye> Hey, I'm trying to build an image to use on my NAS. I want to use the data boot mode, so diskless with a persistent /var. The complexity is that I want to have /var on a ZFS pool. extended has the packages available, but not installed it seems? I have been looking at the existing mkimg scripts however I'm a little confused on how I should have zfs installed and also present in the initramfs. It seems I might want an apkovl to enable the ZFS services
2017-12-04 17:23:39 <bpye> , but I'm not 100% sure. I have seen that xen at some point used apkovls (https://github.com/alpinelinux/aports/blob/master/scripts/mkimg.xen.sh#L20) but it's commented out? I think I should just be able to add the ZFS feature to initfs_features right to get it in the initramfs right?
2017-12-04 17:23:43 <kaniini> it has less sentences
2017-12-04 17:24:18 <kaniini> yes an apkovl would get the job done
2017-12-04 17:24:21 <kaniini> you can make one with lbu
2017-12-04 17:26:15 <bpye> kaniini: And then run mkimg, passing in my apkovl right? is initramfs_features the right place to be adding it to get the kernel module also? In an ideal world I would have my apkovl backups on the ZFS pool too, so that it is my only persistent storage, and my USB drive is completely r/o
2017-12-04 17:26:48 <vmele> kunkku: there is very little that has been litigated
2017-12-04 17:28:41 <vmele> i haven't looked into this issue specifically. I have previously been focused on AGPL3 licensed open source software, and haven't looked as to packaging. My day job has focused more on Alpine, so I have an incentive to help the project when I can. I'll look into it.
2017-12-04 17:30:44 <kaniini> bpye: no need to do that, you could just copy it to the USB drive
2017-12-04 17:31:04 <kaniini> bpye: as localhost.apkovl.tar.gz
2017-12-04 17:31:05 <kaniini> or such
2017-12-04 17:33:07 <kaniini> vmele: there's other reasons why a foundation is desirable besides holding the aports repo rights, there's a lot of companies that want to sponsor/donate things, but management at those companies aren't liking the whole "we can't write this off" thing
2017-12-04 17:33:34 <bpye> kaniini: And add ZFS to the boot group I guess? Is it possible to later load a second apkovl with my server specific changes?
2017-12-04 17:34:04 <kaniini> bpye: no, chaining apkovls isnt possible right now
2017-12-04 17:34:45 <bpye> Ah okay, that would be pretty great. But already better than what I have, thank you!
2017-12-04 17:35:21 <vmele> kaniini: I am in no way discouraging a foundation. In fact, I highly support it. I'd like to volunteer to help in whatever way I can regarding it. My comments here were regarding only the aports licensing situation.
2017-12-04 18:10:49 <scadu> why Alpine needs foundation?
2017-12-04 18:11:50 <skarnet> it's all up there ^ :P
2017-12-04 18:17:44 <vmele> re: a foundation. I urge you to read this about Fedora's tribulations with forming a foundation: https://www.redhat.com/archives/fedora-announce-list/2006-April/msg00016.html
2017-12-04 18:18:56 <vmele> There are very specific Red Hat v. Fedora community concerns, but it's worth a read as to how the Alpine project may begin thinking about leadership and structure for a foundation.
2017-12-04 18:27:25 <ncopa> i am not against relicensing APKBUILDs as MIT or ISC (whats the diff?) , nor am I against creating a foundation
2017-12-04 18:27:34 <ncopa> i think we sooner or later will need a foundation
2017-12-04 18:27:42 <ncopa> to make our lives simpler
2017-12-04 18:27:45 <ncopa> what i want is simple
2017-12-04 18:28:38 <ncopa> i have been thinking about creating a foundation for some time
2017-12-04 18:29:33 <kaniini> i think we have reached the point where it is necessary
2017-12-04 18:29:38 <ncopa> and i may be willing to contribute with the financial costs to get it up and running if needed
2017-12-04 18:29:47 <kaniini> there's a few people in alpine that if they were incapacitated
2017-12-04 18:29:57 <kaniini> would be really damaging to the project
2017-12-04 18:30:15 <kaniini> a foundation is a good step towards fixing that
2017-12-04 18:31:17 <kaniini> if ncopa were to get into an accident, would the project survive?  in theory people would want to keep working on it, but it would be a huge disaster
2017-12-04 18:31:27 <ncopa> i want avoid contributors need to go through lots of legal papers to be able to contribute
2017-12-04 18:31:39 <ncopa> i think it would continue without me
2017-12-04 18:31:58 <kaniini> it would, but there would be continuity problems
2017-12-04 18:32:30 <przemoc> for instance lack of notes related to some infra setup, I guess, so rediscovering would be needed ;)
2017-12-04 18:32:44 <ncopa> i have made sure that I am not the only person who have access to important parts, like ssh keys, login to dns registrar etc
2017-12-04 18:33:05 <przemoc> ncopa: MIT is more verbose than ISC, slightly more explicit (e.g. sublicense), therefore slightly more calming for developers, if you know what I mean.  ISC is more succinct, but also slightly more implicit therefore.  personally I prefer MIT, but some, like kaniini or skarnet, prefer ISC.
2017-12-04 18:33:11 <scadu>  > for instance lack of notes related to some infra setup, I guess, so rediscovering would be needed ;)
2017-12-04 18:33:12 <scadu> and how foundation help with this? XD
2017-12-04 18:33:44 <ncopa> scadu: thats a different topic
2017-12-04 18:34:13 <vmele> ncopa: I'm willing to form and get the foundation off the ground from the legal side, if you want to do this with an US entity.
2017-12-04 18:34:46 <ncopa> thats one of the questions: under which country should it happen?
2017-12-04 18:34:56 <kaniini> continuity, legal rights-holding and being able to do more in terms of outreach and project support (hardware, sponsorship of important projects, etc) are the main reasons why alpine needs a foundation
2017-12-04 18:34:57 <skarnet> personally I would vote against a US entity
2017-12-04 18:34:58 <vmele> formation would be step 1, and doesn't determine the next steps of assignment, leadership, etc, etc
2017-12-04 18:35:14 <kaniini> i agree, i don't think a US entity is the way to go
2017-12-04 18:35:22 <kaniini> the IRS has a lot of silly rules
2017-12-04 18:35:29 <skarnet> not because of you vmele, but because the USA is very backwards when it comes to software
2017-12-04 18:35:38 <jirutka> agree with skarnet
2017-12-04 18:35:40 <vmele> oh i fully agree that the US is backwards
2017-12-04 18:35:42 <jirutka> against US entity
2017-12-04 18:35:45 <kaniini> we're doing better than europe actually
2017-12-04 18:35:52 <ncopa> i guess all countries has their own issues
2017-12-04 18:35:55 <kaniini> the EPO is doing swpats now ;)
2017-12-04 18:35:58 <ncopa> i could do it in norway
2017-12-04 18:36:09 <przemoc> Alpine is European at heart, so US-based foundation would be a treachery! :)
2017-12-04 18:36:35 <jirutka> I can do it in Czech Republic, i.e. under EU legislation
2017-12-04 18:36:39 <kaniini> swpats are basically on their way out here
2017-12-04 18:36:43 <vmele> but but but all that US money :-)
2017-12-04 18:36:52 <skarnet> where?
2017-12-04 18:36:56 <skarnet> I'm not seeing it
2017-12-04 18:37:12 <vmele> I am a dual-EU citizen btw because I know how backwards the US is ;-)
2017-12-04 18:37:14 <kaniini> skarnet: in US, patent litigation now requires the plaintiff to cover all costs when they lose
2017-12-04 18:37:34 <kaniini> this has really been harmful to patent trolls ongoing profitability
2017-12-04 18:37:51 <ncopa> im in a mtg now... brb
2017-12-04 18:37:59 <skarnet> kaniini: yes, I've heard of that, and it's a good thing, but it's even better not to have any swpats in the first place :P
2017-12-04 18:38:10 <Xe> could always do it in estonia :)
2017-12-04 18:38:35 <kaniini> skarnet: in recent years, swpats have not been granted very often
2017-12-04 18:38:44 <kaniini> i said getting better, not  cured
2017-12-04 18:38:52 <kaniini> this is like going from terminal to "maybe it can be cured after all"
2017-12-04 18:39:15 <kaniini> all the swpats people are in asia now days
2017-12-04 18:39:55 <kaniini> http://techrights.org/wiki/index.php/EPO documents the EPO giving out swpats too, despite them being explicitly illegal
2017-12-04 18:40:16 <przemoc> I haven't checked how creating and managing foundations look in Poland, so won't volunteer in that matter
2017-12-04 18:40:23 <kaniini> so i think really there's no jurisdiction right now with correctly applied IP law
2017-12-04 18:42:37 <kaniini> vmele: i think it is possible for US companies to write off donations to foreign entities as long as they would otherwise be legally equivalent to a 501(c)3 non-profit
2017-12-04 18:42:52 <kaniini> i haven't talked to an accountant in many years about that though, so maybe it has changed
2017-12-04 18:43:21 <kaniini> but for example, Private Internet Access was donating cash and servers to freenode which was a UK entity
2017-12-04 18:43:28 <kaniini> while PIA is a US entity
2017-12-04 18:46:12 <Shiz> eu doesn t have swpats :P
2017-12-04 18:46:33 <kaniini> you might want to let the EPO know about that then
2017-12-04 18:46:39 <kaniini> because they keep issuing swpats
2017-12-04 18:46:40 <kaniini> :D
2017-12-04 18:47:08 <kaniini> with bribes anything is possible i guess
2017-12-04 18:47:29 <vmele> kaniini: that's what I was seeing. It is possible, but the foreign entity has to comply and file with the IRS the same as if it was a US entity.
2017-12-04 18:47:51 <kaniini> aha
2017-12-04 18:48:34 <Shiz> kaniini: i have to say some of this reads very... melodramatic
2017-12-04 18:48:39 <Shiz> >Alison Brimelow, got replaced by a megalomaniac called Benoît Battistelli, who gradually drove away anything or anyone who threatens his undemocratic ambition of total power.
2017-12-04 18:48:41 <Shiz> thinking emoji
2017-12-04 18:48:53 <vmele> As to the IP side, yeah, there isn't a "perfect" jurisdiction and I don't know enough about the current EU stance to know if it's any better than the US. On one hand, the US does *protect* IP pretty thoroughly, and usually has thoughtul courts. emphasis on usually.
2017-12-04 18:48:54 <Shiz> >Raw: The EPO's Pattern of Using “Security” as Pretext for Ban on Protests
2017-12-04 18:49:19 <kaniini> yeah, schestowitz is kinda extreme
2017-12-04 18:50:13 <kaniini> vmele: well there is also advantage of having significant case law for GPL enforcement in the US, but that doesn't really apply to alpine
2017-12-04 18:51:56 <vmele> exactly. I think the goal would be to pick a jurisdiction that is sophisticated enough. I'm obviously biased towards the US.
2017-12-04 18:59:09 <Shiz> .nl is always an option
2017-12-04 18:59:11 <Shiz> ;p
2017-12-04 18:59:34 <vmele> The particularities regarding a non-profit corp and its board are also very developed in the US.
2017-12-04 19:01:06 <kaniini> .nl is a good option for IKEA hohoho
2017-12-04 19:01:17 <kaniini> i think that's where they have their tax shelter anyway
2017-12-04 19:01:32 <Shiz> they do
2017-12-04 19:01:34 <Shiz> their HQ is...
2017-12-04 19:01:38 <Shiz> a few km away from where I live
2017-12-04 19:01:41 <Shiz> :P
2017-12-04 19:02:03 <kaniini> the IKEA Foundation for Interior Design or whatever it is called
2017-12-04 19:02:11 <kaniini> it's something like that
2017-12-04 19:02:18 <ncopa> does it need to be a non-profit org? IIRC it may be simpler to set up a normal company than a non-profit
2017-12-04 19:02:44 <kaniini> well, that gives more flexibility, but the community may be spooked by it
2017-12-04 19:02:52 <kaniini> "Alpine is going corporate" blah blah blah
2017-12-04 19:03:01 <ncopa> 🙃
2017-12-04 19:03:11 <vmele> also won't allow tax-deductible donations
2017-12-04 19:03:12 <arch3y> IVe been on projects that are spooked by even 503c
2017-12-04 19:03:38 <arch3y> but not saying I am just saying ppl get scared of things, sorry for interjecting randomly carry on
2017-12-04 19:03:41 <ncopa> yeah, people will not like the "alpine is going commercial"
2017-12-04 19:04:01 <arch3y> if they only realized it is a necessary evil at times
2017-12-04 19:04:05 <kaniini> but in terms of rightsholding, yes it is possible at least in the US to establish a revenue-neutral entity just to hold assets
2017-12-04 19:04:08 <kaniini> and the paperwork is not too bad on that
2017-12-04 19:04:11 <vmele> this also doesn't have to be a single entity. It can be a non-profit, and a separate IP holding company that doesn't have anything to do with the foundation
2017-12-04 19:04:40 <ncopa> i suppose we could even have separate entities in different countries too?
2017-12-04 19:04:51 <kaniini> yes, that's what wikimedia does
2017-12-04 19:04:53 <vmele> yep
2017-12-04 19:04:57 <ncopa> i even have a company set up already in norway
2017-12-04 19:05:23 <jirutka> ncopa: ehm, if you’re gonna create a commercial company for Alpine, then you must pay me, or I will leave >_<
2017-12-04 19:05:24 <kaniini> ncopa linux distribution domination services, llc?
2017-12-04 19:05:34 <kaniini> :D
2017-12-04 19:05:37 <Xe> ncopa stack smashing inc
2017-12-04 19:05:43 <kaniini> see people already spooked on that ;)
2017-12-04 19:05:53 <ncopa> <ncopa> yeah, people will not like the "alpine is going commercial"
2017-12-04 19:05:55 <ncopa> exactly
2017-12-04 19:06:21 <kaniini> Xe: only in apk-tools
2017-12-04 19:06:33 <Xe> pfft
2017-12-04 19:06:34 <jirutka> I was always thinking only about non-profit org, anything else doesn’t make much sense for OSS project
2017-12-04 19:06:48 <ncopa> jirutka: we are talking about setting up a legal entity
2017-12-04 19:06:52 <jirutka> I know
2017-12-04 19:06:57 <jirutka> that’s what I’m talking about too
2017-12-04 19:07:01 <ncopa> what is the reson we want/need a legal entity
2017-12-04 19:07:11 <ncopa> what is the purpose?
2017-12-04 19:07:22 <kaniini> a few reasons
2017-12-04 19:07:24 <jirutka> legal issues (like licensing), to be able to accept donations etc.
2017-12-04 19:07:29 <kaniini> 1) hold any IP rights
2017-12-04 19:07:30 <vmele> there are two distinct, and not necessarily related, purposes: 1) IP, 2) donations
2017-12-04 19:07:46 <kaniini> 2) fundraising for various reasons
2017-12-04 19:07:56 <ncopa> both can be covered with different types of entities
2017-12-04 19:08:12 <ncopa> the different types of entities has different pros and cons
2017-12-04 19:08:13 <kaniini> 3) grantmaking ability (so we could say, for example, pay jirutka to work on some really important project)
2017-12-04 19:08:48 <kaniini> right now, it is possible to get sponsored to work on some alpine stuff, if some company wants it
2017-12-04 19:08:50 <ncopa> but it would be stupid to not consider (or at least understnad) the different options we have for entities
2017-12-04 19:09:07 <Shiz> entitled to entities
2017-12-04 19:09:09 <kaniini> like docker or whatever
2017-12-04 19:09:19 <jirutka> triggered…
2017-12-04 19:09:19 <kaniini> but for the stuff that companies don't see value in, but is still important
2017-12-04 19:09:25 <kaniini> there's no way to sponsor it
2017-12-04 19:09:48 <kaniini> jirutka and i talked about that a long time ago
2017-12-04 19:10:19 <kaniini> stuff that is important could slip in the future because the stuff that companies want gets preference since people can get paid to do that
2017-12-04 19:10:19 <ncopa> jirutka: what im trying to say is that you can set up an org which technically is a "commercial" company, that never makes any money for the shareholders
2017-12-04 19:10:42 <kaniini> yes, a cooperative can work
2017-12-04 19:10:51 <kaniini> too
2017-12-04 19:10:52 <jirutka> ncopa: but why to not set up legally non-profit org to be perfectly clear for everyone…?
2017-12-04 19:11:01 <kaniini> and some entities like IBM may like that better than a non-profit
2017-12-04 19:11:07 <kaniini> it's all complicated
2017-12-04 19:11:08 <kaniini> :D
2017-12-04 19:11:31 <vmele> ncopa: you can have both
2017-12-04 19:11:33 <jirutka> also it’s easier to set up and run non-profit
2017-12-04 19:11:34 <ncopa> jirutka: because i dont know how that works, what are the required paper work, what are the pros/cons with a non-profit
2017-12-04 19:11:56 <kaniini> jirutka: in the US at least, it is easier to set up a for-profit than a non-profit
2017-12-04 19:12:02 <jirutka> aha
2017-12-04 19:12:07 <ncopa> jirutka: thats the question, someone i talked with said its more paper work with non-profit
2017-12-04 19:12:12 <jirutka> well, US is all for-profit or nothing… :P
2017-12-04 19:12:22 <kaniini> the IRS thinks you are trying to scam when you set up a non-profit
2017-12-04 19:12:26 <kaniini> so they are very strict
2017-12-04 19:12:35 <Shiz> i wonder how dutch nonprofits work...
2017-12-04 19:12:41 <kaniini> it's not that
2017-12-04 19:12:42 <ncopa> i suppose it depends on the country
2017-12-04 19:13:12 <scadu> disroot has created a foundation in NL lately. maybe you could discuss it with muppeth and/or antilopa
2017-12-04 19:13:12 <kaniini> the tax man is brutal in all sophisticated enough (for our needs) countries
2017-12-04 19:13:13 <kaniini> the paperwork largely has to do with making the tax man happy
2017-12-04 19:13:14 <vmele> kaniini: US corporations are in each state. A non-profit does not have to be a 501(c)(3) corp, as that is an IRS status.
2017-12-04 19:13:23 <jirutka> well, at least in CZ non-profit is easier and I know people who already founded many non-profits in ICT field and know how it works
2017-12-04 19:13:27 <Shiz> https://nl.wikipedia.org/wiki/Vereniging_(rechtspersoon)
2017-12-04 19:13:30 <Shiz> found it, I think
2017-12-04 19:13:46 <kaniini> vmele: yes, i know, but to get 501(c)3 status it is seriously involved
2017-12-04 19:13:46 <jirutka> and globally non-profit is more clear about the goals
2017-12-04 19:13:58 <Shiz> okay so
2017-12-04 19:14:02 <Shiz> from what i'm reading
2017-12-04 19:14:03 <kaniini> for-profit/non-profit registration itself is same amount of paperwork in Oklahoma at least
2017-12-04 19:14:05 <Shiz> a dutch vereniging can make a profit
2017-12-04 19:14:11 <Shiz> but it must not have the intention to make one
2017-12-04 19:14:14 <Shiz> or distribute profit among members
2017-12-04 19:14:16 <Shiz> as its main purpose
2017-12-04 19:14:29 <kaniini> a non-profit can also own for-profit entities
2017-12-04 19:14:30 <Shiz> it has to be used for vereniging purposes
2017-12-04 19:14:32 <vmele> kaniini: yes and no. I'm in Wisconsin, which has favorable non-profit laws. It's trivial to set up.
2017-12-04 19:14:32 <Shiz> :)
2017-12-04 19:14:40 <kaniini> such as Mozilla Foundation vs. Mozilla Corporation
2017-12-04 19:14:54 <Shiz> you would setup a Vereniging met volledige rechtsbevoegdheid
2017-12-04 19:14:56 <jirutka> btw not sure if everyone knows, non-profit does NOT mean that you cannot pay people for work
2017-12-04 19:14:59 <Shiz> (is that what discroot did, scadu?)
2017-12-04 19:15:03 <Shiz> disroot*
2017-12-04 19:15:10 <kaniini> jirutka: indeed
2017-12-04 19:15:13 <skarnet> this discussion should not be used as an excuse to post Dutch in the channel
2017-12-04 19:15:21 <ncopa> :)
2017-12-04 19:15:21 <jirutka> skarnet: +1 XD
2017-12-04 19:15:31 <Shiz> francophones out
2017-12-04 19:15:34 <Shiz> anyway
2017-12-04 19:15:39 <Shiz> I know some people who setup a vereniging as well
2017-12-04 19:15:41 <Shiz> :P
2017-12-04 19:15:57 <kaniini> anyway,  once people figure out what they do, i'm willing to donate to it
2017-12-04 19:16:01 <jirutka> but I work for a company that has many customers from Netherlands, so I’m already used to it; when I see something that looks like a cypher text, I know that’s it’s actually Dutch XD
2017-12-04 19:17:10 <scadu>  > (is that what discroot did, scadu?)
2017-12-04 19:17:11 <scadu> Shiz let me ask muppeth
2017-12-04 19:17:19 <Shiz> \o
2017-12-04 19:18:00 <vmele> I'm willing to set up a US non-profit today (less than $100 which I will cover), get some basic and flexible bylaws in place, start the 501(c)(3) process, which should be enough to get the ball rolling. This would *not* deal with IP, because that will need additional thought. Really, all this does is form a corp and reserve the name ;-)
2017-12-04 19:18:11 <Shiz> ah found it
2017-12-04 19:18:13 <Shiz> they are a stichting
2017-12-04 19:18:23 <Shiz> https://www.kvk.nl/orderstraat/product-kiezen/?kvknummer=69988099
2017-12-04 19:18:56 <vmele> theoretically, *I* could start the "Alpine Foundation, Corp." in the US without any of your knowledge or involvement.
2017-12-04 19:19:24 <scadu> Shiz: what does it mean actually?
2017-12-04 19:19:41 <skarnet> you could, but you won't, because that would be a dick move
2017-12-04 19:19:42 <Shiz> formele vereniging and stichtingen are too similar things
2017-12-04 19:19:54 <skarnet> especially since several people here have voiced their opposition to a US entity
2017-12-04 19:19:54 <Shiz> the main difference is that verenigingen have members are democratic
2017-12-04 19:20:04 <vmele> skarnet: absolutely it would be a dick move. See: bitcoin foundation as an example of dick moves ;-)
2017-12-04 19:20:07 <Shiz> and are*
2017-12-04 19:20:40 <Shiz> e.g. vereniging can only change statutes by means of an official meeting/vote
2017-12-04 19:20:44 <jirutka> btw I’d be *really* more happy if the org will not be US-based; Netherlands, Czech Republic, Estonia or whatever European country would be better
2017-12-04 19:20:53 <Shiz> (the process of which is laid down in your initial statutes)
2017-12-04 19:22:39 <Shiz> there's a number of different laws for both which seem to be mostly details
2017-12-04 19:22:59 <Shiz> stichting would be the equivalent to a foundation
2017-12-04 19:23:01 <Shiz> (even in word choice)
2017-12-04 19:24:39 <skarnet> I'll show you a stitching
2017-12-04 19:24:54 <Shiz> rude & threatening
2017-12-04 19:25:00 <skarnet> very!
2017-12-04 19:25:17 <jirutka> XD
2017-12-04 19:30:51 <ncopa> vmele: i appreciate your willingness to help! i will have to think about it a bit
2017-12-04 19:31:08 <ncopa> i'd like to know what the differences is US based vs different european
2017-12-04 19:31:18 <ncopa> and what the options are in norway
2017-12-04 19:31:22 <kunkku> the cost of running a legal entity are usually higher than the set-up costs...
2017-12-04 19:31:46 <ncopa> that why i dont want to rush it
2017-12-04 19:31:52 <ncopa> without knowing the options/differences
2017-12-04 19:32:55 <vmele> ncopa: here's what i propose that shouldn't foreclose any options: forming a US non-profit merely for donation purposes (even if those donations are only from the US). Then you can form a different one in the EU if you choose, and a different one for IP holding. But to solve the problem of having somewhere where individuals and corporations can donate, and that organization can provide grants, it doesn't matter what jurisdiction you'
2017-12-04 19:32:56 <vmele> re in really.
2017-12-04 19:33:46 <vmele> I fully understand the reluctance for IP to be anywhere in the US. We're scary and backwards here sometimes.
2017-12-04 19:35:46 <Shiz> i think under dutch law an alpine foundation cold even be an SBBI
2017-12-04 19:35:53 <vmele> foreign directors on a US non-profit board doesn't subject those directors to anything with the IRS
2017-12-04 19:36:29 <Shiz> which is a foundation that has a certain social importance
2017-12-04 19:36:39 <Shiz> and thus gets tax benefits
2017-12-04 19:36:41 <Shiz> :P
2017-12-04 19:38:21 <_ikke_> ANBI in dutch?
2017-12-04 19:42:47 <Shiz> _ikke_: SSBI
2017-12-04 19:42:50 <Shiz> ANBI is a different thing
2017-12-04 19:42:52 <_ikke_> ok
2017-12-04 19:43:25 <Shiz> https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/zakelijk/bijzondere_regelingen/goede_doelen/sociaal_belang_behartigende_instellingen/
2017-12-04 19:43:40 <ncopa> vmele: but it may me sense to start a non-profit in US to make it easier for US based companies to contribute?
2017-12-04 19:44:24 <_ikke_> Shiz: question would be of Alpinelinux would be seen as one
2017-12-04 19:44:36 <Shiz> i think the rules are pretty lenient
2017-12-04 19:44:40 <Shiz> ANBI is far stricter yes
2017-12-04 19:46:09 <_ikke_> Still, the rules say it must be of some kind of societal benefit, not sure if AL would meet that
2017-12-04 19:47:29 <_ikke_> An SBBI also regulates how people can be compensated
2017-12-05 00:42:39 <kaniini> the way i see it is
2017-12-05 00:42:51 <kaniini> a foundation is absolutely necessary to keep alpine growing in the direction we want it to grow
2017-12-05 00:43:13 <kaniini> it strengthens the non-commercial nature of the project
2017-12-05 00:43:55 <kaniini> and it provides competition for developer attention with docker and other entities who may not have interest in sponsoring things alpine requires
2017-12-05 00:44:41 <kaniini> docker, for example, wants to sponsor some things in alpine obviously.  but are they interested in sponsoring things they can't use in linuxkit, or things that are not relevant to containers?
2017-12-05 00:45:10 <kaniini> IBM want to sponsor ppc64le support in alpine.  but what about ppc32/ppc64be?
2017-12-05 00:45:15 <kaniini> so on and so on
2017-12-05 00:45:31 <kaniini> not to mention that a foundation is a good signal to these players that alpine itself is maturing
2017-12-05 00:45:39 <kaniini> which means they are willing to commit more
2017-12-05 00:45:59 <kaniini> it also gives us a better mechanism for getting things fixed in upstream projects like musl
2017-12-05 00:46:03 <jirutka> completely agrees
2017-12-05 00:46:39 <kaniini> right now if we want to fix something in musl on our own terms, somebody has to pay dalias directly
2017-12-05 00:47:29 <kaniini> with a foundation, we can just grant dalias some money to finish locale support for example
2017-12-05 00:47:44 <kaniini> it's not just things inside alpine, it's the whole ecosystem
2017-12-05 00:48:12 <kaniini> it gives us additional tools (money) to get the things we need taken care of
2017-12-05 05:37:12 <fffflflflflflf> ping
2017-12-05 05:37:27 <leo-unglaub> fffflflflflflf pong
2017-12-05 05:37:55 <_ikke_> is this some kind of coordination? :O
2017-12-05 05:37:58 <_ikke_> ;D
2017-12-05 05:38:27 <leo-unglaub> ups, sorry ... wrong channel ...
2017-12-05 05:38:44 <leo-unglaub> thats my irc library that i am devloping, but i wrote to the wrong channel
2017-12-05 05:38:49 <leo-unglaub> sorry about that
2017-12-05 05:38:49 <_ikke_> heh
2017-12-06 03:13:31 <strfry> what's the preferred way to compile for various architectures in alpine? i have a x86_64 build machine, but i also want to build custom packages for armhf (raspberry pi). i assume i have to create another container/chroot running armhf with qemu-static?
2017-12-06 03:14:05 <kaniini> qemu is probably best way to go
2017-12-06 05:02:24 <strfry> is there some documentation, on how the official repositories are built from the aports git?
2017-12-06 09:19:42 <Shiz> strfry: the buildrepo script in lua-aports afaik
2017-12-06 09:19:54 <Shiz> https://git.alpinelinux.org/cgit/lua-aports/tree/bin/buildrepo.lua
2017-12-06 11:48:43 <skarnet> aaaand lists.alpinelinux.org rejects my mails AGAIN
2017-12-06 11:48:54 <skarnet> when will you guys stop fucking with the mail configuration?
2017-12-06 11:49:26 <skarnet> I'm fed up with this, I'm currently changing my settings so it doesn't happen again no matter what you do
2017-12-06 11:49:49 <skarnet> but geez, I didn't think Alpine would be a hostile actor in the fight for users' right and net neutrality
2017-12-06 11:49:55 <skarnet> to that extent
2017-12-06 11:50:02 <danieli> jesus, that seems kind of.. overly aggressive?
2017-12-06 11:50:19 <skarnet> no it's not, it's the FOURTH TIME this is happening
2017-12-06 11:50:28 <skarnet> in my book, you can fuck up once, twice
2017-12-06 11:50:37 <danieli> how does it relate to net neutrality?
2017-12-06 11:50:39 <skarnet> if you fuck up three times you suck at sysadmin
2017-12-06 11:50:51 <skarnet> fourth time doesn't happen in my surroundings usually
2017-12-06 11:51:12 <danieli> ping clandmeter ^
2017-12-06 11:51:24 <skarnet> don't bother, it will not happen again
2017-12-06 11:51:24 <_ikke_> skarnet: unless you are fighting a lot of spam
2017-12-06 11:51:33 <_ikke_> (not sure if this is the case)
2017-12-06 11:51:36 <skarnet> I'm making sure I have a workaround on my sude
2017-12-06 11:51:38 <skarnet> side*
2017-12-06 11:51:45 <danieli> what's the issue, if I may ask?
2017-12-06 11:52:09 <skarnet> danieli: it relates to net neutrality because not allowing individuals to host their own mail server desempowers them
2017-12-06 11:52:17 <skarnet> and puts the power in the hands of ISPs
2017-12-06 11:52:20 <skarnet> who, mostly, suck
2017-12-06 11:52:30 <danieli> that doesn't really answer my question
2017-12-06 11:52:33 <_ikke_> I also have my own mail server
2017-12-06 11:52:36 <skarnet> and sorry, it's not about you, but I'm really fed up with explaining the issue again
2017-12-06 11:53:07 <skarnet> look in the archives, there are three instances of the discussion
2017-12-06 11:53:12 <danieli> _ikke_: I r(u|a)n lots of mail servers as well
2017-12-06 11:53:31 <skarnet> I guess you have a fixed IP
2017-12-06 11:53:53 <skarnet> static, I mean
2017-12-06 11:53:56 <_ikke_> yes
2017-12-06 11:54:02 <skarnet> there you go
2017-12-06 11:54:08 <skarnet> not everyone has a decent ISP
2017-12-06 11:54:22 <_ikke_> I don't host it through my home ISP
2017-12-06 11:54:24 <danieli> not everyone runs mail servers from dynamic addresses
2017-12-06 11:54:29 <danieli> many can't
2017-12-06 11:54:36 <_ikke_> port 25 is blocked
2017-12-06 11:54:45 <danieli> mine's pseudo-static, bur I don't run a mail server from home - although no ports are blocked
2017-12-06 11:54:46 <skarnet> "works for me"
2017-12-06 11:54:51 <skarnet> "don't do that"
2017-12-06 11:54:58 <skarnet> yeah, just the kind of support I need
2017-12-06 11:55:00 <skarnet> thanks guys, you rock
2017-12-06 11:55:02 <danieli> don't be so damn hostile
2017-12-06 11:55:09 <danieli> go have a cup of coffee and calm down
2017-12-06 11:55:29 <skarnet> I'll be hostile if I fucking want to. Stop fucking up with the mail server configuration and I'll be friendly.
2017-12-06 11:55:43 <_ikke_> skarnet: and let the list be overrun by spam for example?
2017-12-06 11:55:59 <danieli> I don't touch any mail for Alpine
2017-12-06 11:56:11 <danieli> either way, tone it down a little, you don't need to make my day shittier just because yours is
2017-12-06 11:56:15 <skarnet> I have an exception in the mail server configuration. nangel did this for me.
2017-12-06 11:56:27 <skarnet> Why someone needs to keep changing it, I don't know.
2017-12-06 11:56:33 <_ikke_> skarnet: is that the case then?
2017-12-06 11:57:02 <skarnet> I don't know. And I'm not supposed to know, I don't run Alpine's infra.
2017-12-06 11:57:07 <_ikke_> nangel did update ns1 infra
2017-12-06 11:57:13 <_ikke_> not sure if mail is involved in that
2017-12-06 11:57:29 <skarnet> anyway it's moot, it will never happen again
2017-12-06 11:57:47 <_ikke_> skarnet: why making a point of it if you can 'solve it' yourself?
2017-12-06 11:57:50 <skarnet> but I would have expected more competency as in "do no harm" from Alpine people
2017-12-06 11:57:54 <danieli> ns1.a.o and mail.a.o are on nearly adjacent IPs (.114 and .116)
2017-12-06 11:58:03 <skarnet> 
ACTION takes a deep breath

2017-12-06 11:58:05 <_ikke_> skarnet: this has nothing to do with harmful intent
2017-12-06 11:58:21 <skarnet> BECAUSE "SOLVING IT" MAKES ME PART OF THE PROBLEM
2017-12-06 11:58:35 <_ikke_> skarnet: You are implying people are changing it just to prevent you from sending mails
2017-12-06 11:58:38 <danieli> I'll come back later when people aren't raging at each other over nothing.
2017-12-06 11:58:40 <skarnet> no
2017-12-06 11:58:47 <skarnet> I'm implying people don't give a shit
2017-12-06 11:58:58 <skarnet> and I'm asking people to pay attention
2017-12-06 11:59:19 <skarnet> danieli: ok, my ability to send mail from my home server is nothing. Gotcha.
2017-12-06 11:59:46 <danieli> you completely missed my point, probably on purpose - but my point is proven once again. calm down.
2017-12-06 11:59:52 <danieli> 
ACTION &

2017-12-06 12:00:08 <_ikke_> skarnet: I just wonder how you deal with all the block lists that block ISP owned IPs
2017-12-06 12:00:47 <skarnet> I route that mail to my ISP's MX, which disempowers me and allows my ISP to fuck with it, which I hate
2017-12-06 12:01:12 <skarnet> it's a compromise I have to do for my mail to reach people with fascist rules
2017-12-06 12:01:34 <skarnet> I need to do the same for Alpine now
2017-12-06 12:01:41 <_ikke_> skarnet: I do agree that you should be able to run a mail server from your own IP address
2017-12-06 12:01:52 <_ikke_> skarnet: but that also resulted in a lot of spam to be sent
2017-12-06 12:02:43 <skarnet> I know the rationale
2017-12-06 12:02:45 <danieli> if it's an ISP IP address, I'm guessing you don't have a proper PTR record, which will have mail from you rejected in a lot of places
2017-12-06 12:03:03 <skarnet> stop splaining me, I know exactly what the problem is
2017-12-06 12:03:25 <skarnet> the underlying problem is that the mail architecture is BROKEN
2017-12-06 12:03:43 <skarnet> but I had managed to convince Alpine infra to make an exception for me, because we're all good guys here
2017-12-06 12:03:56 <skarnet> and somehow that gets broken on a regular basis
2017-12-06 12:04:25 <skarnet> so I need to withdraw my assessment that Alpine infra are good guys - they just don't care
2017-12-06 12:04:41 <skarnet> so I'm going to do the same with Alpine as with the rest of the broken world
2017-12-06 12:05:15 <Michcioperz> how was the exception done before?
2017-12-06 12:05:23 <danieli> you failed or refused to explain the problem, which isn't helpful. I was asking in case I can have any influence on the infra.
2017-12-06 12:05:46 <skarnet> I don't know and I don't care anymore. I'm done.
2017-12-06 12:10:20 <przemoc> tbh, skarnet, I doubt it's done purposefully.   I suspect that exception that you got before wasn't made appropriately permament (and I don't think it was purposeful omission either), i.e. surviving some changes that may happen (just a wild guess, I'm not AL infra guy by any means).
2017-12-06 12:11:03 <clandmeter> of course its not purposely. im not sure what has changed.
2017-12-06 12:11:20 <clandmeter> i would ask to lower the tone a bit, but seems it has been asked already.
2017-12-06 12:11:59 <skarnet> "hello guys, I have a problem, it was solved the 1st time but it has happened 3 more times since then"  "Oh, we're not doing this on purpose"
2017-12-06 12:12:08 <skarnet> ... thanks?
2017-12-06 12:12:33 <danieli> you're being extremely aggressive over a fairly minor issue that can be resolved - you're talking about fascism, net neutrality and more
2017-12-06 12:12:56 <skarnet> you calling the issue "fairly minor" is doing nothing to calm me down
2017-12-06 12:13:08 <clandmeter> please stop now.
2017-12-06 12:13:10 <skarnet> I am talking about net neutrality because it's exactly what it's about
2017-12-06 12:13:27 <clandmeter> skarnet: if we need to sort it out please let me know the details.
2017-12-06 12:13:35 <skarnet> for the 4th time?
2017-12-06 12:13:39 <skarnet> no, I'll stop.
2017-12-06 12:13:48 <skarnet> I know where you guys stand.
2017-12-06 12:13:52 <danieli> clandmeter: just curious - what stack is alpine mail infra running?
2017-12-06 12:14:00 <_ikke_> So you just want to rant, not an actual solution, clear
2017-12-06 12:14:15 <clandmeter> its not done by me, so i will have to ask someone else or check up on it.
2017-12-06 12:15:51 <danieli> nangel doesn't seem very active here in my opinion, if he's the one running ns1.a.o (adjacent to mail.a.o's IP)
2017-12-06 12:16:38 <clandmeter> he is here ones in a while.
2017-12-06 12:16:49 <clandmeter> not everybody is an irc addict :)
2017-12-06 12:16:56 <danieli> true true
2017-12-06 12:18:43 <danieli> runs postfix, I don't spot any MUA
2017-12-06 12:19:07 <_ikke_> Why would there be a MUA running
2017-12-06 12:20:06 <danieli> is it not used to send mail from @a.o?
2017-12-06 12:21:02 <_ikke_> MUA == Mail User Agent
2017-12-06 12:21:06 <danieli> hm, nevermind, I see how it works
2017-12-06 12:23:14 <przemoc> clandmeter: it's not about being irc addict, it's rather old-school mentality, I would say, i.e. that you use IRC at all in days of discords, telegrams, matrix and what not (somehow I cannot really hop on this "new"/"cool" trends bandwagon).  I have dozens of IRC windows in irssi, but I actually look into a few if at all, sometimes week passes and I may not lurk in any of them, another week I may
2017-12-06 12:23:20 <przemoc> lurk but not have time or devotion to engage in any discussion.
2017-12-06 12:54:45 <rnalrd> skarnet 82.216.21.182 is listed in the Safe SORBS list
2017-12-06 12:55:09 <rnalrd> safe = dynamic, dial-up, open proxy, etc...
2017-12-06 12:55:53 <rnalrd> so the email is delayed, but if legitimate it gets through
2017-12-06 12:56:05 <rnalrd> i'm going to add that IP in whitelist
2017-12-06 12:56:34 <_ikke_> But if it's a dyamic IP, that might change any moment
2017-12-06 12:56:54 <rnalrd> i understood that he has fixed IP
2017-12-06 12:56:58 <rnalrd> re-reading backlog
2017-12-06 12:57:02 <skarnet> it didn't get through after more than 4 days
2017-12-06 12:57:15 <skarnet> and it's a dynamic IP, so thanks but don't bother
2017-12-06 12:57:30 <_ikke_> skarnet: how else would you able to get whitelisted?
2017-12-06 12:57:35 <skarnet> I have a temporary workaround now and the permanent workaround is coming
2017-12-06 12:58:15 <rnalrd> the delay is 24h
2017-12-06 12:59:47 <skarnet> well obviously that's not what happened then, but again, don't worry about it, I have "fixed" it myself
2017-12-06 13:00:03 <skarnet> which remains, unfortunately, the only way to make sure something's done correctly and permanently
2017-12-06 13:00:50 <rnalrd> smeel like the MTA gave up sending before 24h
2017-12-06 13:01:23 <skarnet> the mail remains in the queue for 4 days, as usual
2017-12-06 13:02:31 <rnalrd> checking logs
2017-12-06 13:03:16 <skarnet> again, forget it
2017-12-06 13:03:20 <skarnet> don't waste your time on it
2017-12-06 13:03:29 <skarnet> I've already wasted enough time on it myself
2017-12-06 13:04:54 <rnalrd> i want to help to the reasonable extent possible
2017-12-06 13:05:08 <skarnet> you can help by dropping it as I have done
2017-12-06 13:05:51 <rnalrd> fine :) I could whitelist by sending provider MTA domain
2017-12-06 13:05:55 <skarnet> thanks for the willingness to help, really, but the obvious simple solution is to take the matter in my own hands, which I have done
2017-12-06 13:28:12 <kunkku> I have given up sending e-mail from a dynamic IP
2017-12-06 13:29:03 <kunkku> as I don't have time to convince all MTA operators of the world to whitelist me
2017-12-06 13:47:24 <clandmeter> kunkku: do we still have some kind of install_if issue?
2017-12-06 13:50:13 <clandmeter> i remember you reported something
2017-12-06 13:52:22 <clandmeter> fabled:  , ncopa ?
2017-12-06 14:04:26 <kunkku> clandmeter: yes
2017-12-06 14:05:04 <kunkku> ncopa was able to reproduce it
2017-12-06 14:05:16 <clandmeter> im also having it
2017-12-06 14:05:38 <clandmeter> with a virtual and python virtualenv
2017-12-06 14:06:24 <clandmeter> doesnt pull in the python specific version of  virtualenv
2017-12-06 14:07:12 <clandmeter> kunkku: do you have an url to the issue?
2017-12-06 14:09:29 <kunkku> https://bugs.alpinelinux.org/issues/8237
2017-12-06 15:37:19 <leo-unglaub> lets see how fast travis is: https://github.com/alpinelinux/aports/pull/2903
2017-12-06 15:37:30 <leo-unglaub> it took around 15 minutes on my pc
2017-12-06 15:37:57 <leo-unglaub> -j 32 is very awesome for that *g*
2017-12-06 15:39:21 <Gottox> for voidlinux, ff never compiled on travis :(
2017-12-06 15:39:27 <Gottox> well, it started
2017-12-06 15:41:13 <leo-unglaub> Gottox why did it not build? Did it run into a timeout? or other issues?
2017-12-06 15:42:15 <Gottox> timeout, yes
2017-12-06 15:43:26 <Gottox> https://travis-ci.org/voidlinux/void-packages/builds/301996076?utm_source=github_status&utm_medium=notification
2017-12-06 15:44:28 <leo-unglaub> hmmm, then maybe we should disable travis for the firefox and thunderbird package
2017-12-06 15:44:54 <leo-unglaub> because if there is no way they can finish in time, why waste the cpu time in the first place?
2017-12-06 15:47:09 <Gottox> you can selectively disable travis runs if you add '[skip ci]' to the PR title or the commit.
2017-12-06 15:47:52 <Gottox> eh, ci skip
2017-12-06 15:47:59 <Gottox> https://github.com/voidlinux/void-packages/pull/9764
2017-12-06 15:56:00 <leo-unglaub> i dont like to mess up my commit messages for something, but maybe a .SKIP_TRAVIS file would be a solution
2017-12-06 19:03:06 <_ikke_> So when reverting packages, it's not only necessary to increase pkgrel, but you need to make sure it's higher then it was before (see git for example, where 2.15.0-1 existed before the revert)
2017-12-07 07:14:47 <tmh1999> should we log #musl in https://dev.alpinelinux.org/irclogs/ ? I guess it's good thing to do.
2017-12-07 11:22:22 <skarnet> tmh1999: #musl's policy includes "do not publish logs".
2017-12-07 16:03:03 <lliBinim> how do I create an apk if I only have a binary?
2017-12-07 16:11:11 <TBB> you skip the build phase and in the package phase you use the install command to install it in a package directory
2017-12-07 16:15:00 <lliBinim> so I need to start from an Alpine installation anyway? I just came across fpm, I think I'll use that =D
2017-12-07 16:15:06 <lliBinim> https://fpm.readthedocs.io/en/latest/
2017-12-07 17:55:48 <fcolista> rdutra, you around?
2017-12-07 17:56:02 <rdutra> fcolista: yes
2017-12-07 17:56:18 <fcolista> last day for you 2day?
2017-12-07 17:56:35 <fcolista> in IBM?
2017-12-07 17:56:39 <rdutra> fcolista: yeah, it is
2017-12-07 17:57:19 <fcolista> ok...so, thanks a lot for all the things you've done so fare for Alpine
2017-12-07 17:58:22 <fcolista> very much apprecciated
2017-12-07 17:58:23 <rdutra> fcolista: my pleasure. I will keep contributing to Alpine when I have some time (probably in my free time)
2017-12-07 17:58:40 <fcolista> and hopefully we will see you around still
2017-12-07 17:58:48 <fcolista> you're contrbution have been very very valuable
2017-12-07 17:59:00 <fcolista> also your attitude and willingness
2017-12-07 17:59:02 <fcolista> thanks a lot
2017-12-07 18:00:00 <tmh1999> rdutra : why leaving before the year-end ? :(
2017-12-07 18:00:25 <tmh1999> not my business anyway, just curious
2017-12-07 18:00:26 <rdutra> fcolista: I really enjoyed to work with you and all the guys from Alpine
2017-12-07 18:00:46 <fcolista> likewise
2017-12-07 18:00:56 <rdutra> tmh1999: I'm moving abroad in the beginning of next year
2017-12-07 18:01:37 <tmh1999> rdutra : It sounds promising. Best luck to you.
2017-12-07 18:01:46 <rdutra> tmh1999: thank you!
2017-12-07 18:24:37 <leitao> rdutra, thanks for all your work!
2017-12-07 18:24:59 <leitao> ppc64el is in a much better shape after now compared to 6 months ago
2017-12-07 18:25:25 <rdutra> leitao: thank you! yeah, it is :)
2017-12-07 18:26:06 <rdutra> leitao: thanks you too as you did the bootstrap
2017-12-07 18:28:20 <vkrishn> yes!! 3.7 seems stable, tried 3.7(host,x86_64)+3.7(qemu quest,x86_64)
2017-12-07 18:28:20 <vkrishn> thanks for effort.
2017-12-07 18:28:27 <vkrishn> noticed following,
2017-12-07 18:28:27 <vkrishn> 1.in same previous config in v3.3 mem-usage was ~300Mb, now ~430Mb
2017-12-07 18:28:27 <vkrishn> 2.qemu-system-x86_64 now install wayland and lots for other packages
2017-12-07 18:28:45 <vkrishn> hope this is ok
2017-12-07 18:30:20 <vkrishn> just wondering if busybox-httpd should stay in busybox pkg and not in extras
2017-12-07 18:30:40 <vkrishn> seems useful applet
2017-12-07 18:51:08 <kaniini> huh?
2017-12-07 18:54:03 <kaniini> that's weird.  some irc events aren't making it to the matrix side of the channel
2017-12-07 18:54:04 <kaniini> wtf :D
2017-12-07 20:46:39 <eleksir> # lxc-attach -n mneme
2017-12-07 20:46:39 <eleksir> # apk search procmail
2017-12-07 20:46:39 <eleksir> procmail-3.22-r2
2017-12-07 20:46:39 <eleksir> procmail-doc-3.22-r2
2017-12-07 20:46:40 <eleksir> # apk add procmail
2017-12-07 20:46:41 <eleksir> OK: 29 MiB in 47 packages
2017-12-07 20:46:44 <eleksir> # procmail
2017-12-07 20:46:46 <eleksir> ^Cprocmail: Terminating prematurely
2017-12-07 20:46:49 <eleksir> Segmentation fault
2017-12-07 20:46:51 <eleksir> #
2017-12-07 20:47:05 <eleksir> looks like procmail is broken in alpine 3.7
2017-12-07 20:47:44 <eleksir> host system centos 7.4 (x86_64), guest system alpine 3.7 (i386)
2017-12-07 21:15:49 <vkrishn> interesting, I enabled /media/usb/cache and now mem usage is ~315Mb, nice!!, switching my server to v3.7, yeah!!
2017-12-07 22:21:59 <ncopa> eleksir: can you please file a bug for procmail? on bugs.alpinelinux.org
2017-12-07 22:42:42 <vkrishn> v3.7 host(bare-metal)+v3.6(qemu guest, x86 with 1800+Mb) did not work
2017-12-07 22:42:42 <vkrishn> v3.7 host(bare-metal)+v3.7(qemu guest, x86_64 with 1500Mb) seems to be working
2017-12-07 22:43:01 <vkrishn> kinda sad
2017-12-07 22:44:36 <vkrishn> thinking to switch back to v3.3-host
2017-12-07 22:47:30 <vkrishn> nope!! - (process:3945): GLib-ERROR **: gmem.c:100: failed to allocate 1640 bytes
2017-12-07 22:47:57 <vkrishn> :(
2017-12-07 23:00:00 <ncopa> vkrishn: it uses more memory?
2017-12-07 23:03:25 <vkrishn> last was -m 1836 and still same issue
2017-12-07 23:03:47 <vkrishn> with v3.3(host) runs smoothly with m -1024
2017-12-07 23:05:10 <vkrishn> v3.1 qemu was best
2017-12-07 23:05:45 <vkrishn> on al v3.3 a little more mem but all went smooth for months
2017-12-07 23:06:29 <vkrishn> now v3.6, v3.7 does not work
2017-12-07 23:06:39 <vkrishn> all hosts seems to be running ok
2017-12-07 23:07:23 <vkrishn> but qemu guests aborts after few minutes
2017-12-07 23:08:09 <Shiz> how weird
2017-12-07 23:09:42 <vkrishn> I also use options "-enable-kvm -nographic -localtime -display curses"
2017-12-07 23:13:40 <vkrishn> would try on v3.5(host, guest) combinations tomorrow
2017-12-07 23:14:02 <vkrishn> else switch to v3.3 host
2017-12-07 23:19:15 <vkrishn> it was setup as root, do I need to make any changes, eg http://tpaste.us/nZY4
2017-12-07 23:19:19 <vkrishn> ?
2017-12-07 23:26:16 <vkrishn> I see some msgs http://tpaste.us/zbpZ
2017-12-07 23:37:13 <vkrishn> iftop UI seems broken
2017-12-07 23:39:12 <vkrishn> ewu! iptraf-ng crashes after few mins 'segmentation fault'
2017-12-07 23:40:06 <vkrishn> and its interface chars seems not ok too!
2017-12-07 23:44:09 <vkrishn> maybe needs some font pkg
2017-12-07 23:44:24 <danieli> enter is not punctuation. :P
2017-12-07 23:49:56 <vkrishn> ?
2017-12-07 23:51:02 <danieli> you're flooding a bit
2017-12-07 23:51:48 <Xe> it's not worth caring about imo
2017-12-07 23:52:02 <Xe> the point gets across; irc clients usually have plenty of vertical room
2017-12-08 00:05:51 <vkrishn> Shiz: why weird, does it run ok on your system ?
2017-12-08 03:59:42 <AWilcox[m]> As far as I can tell the Matrix room is now completely separate from the IRC channel
2017-12-08 04:00:36 <craftyguy1> yea. matrix irc bridge has been having some major problems in the last two days.
2017-12-08 04:01:08 <craftyguy1> I've officially fired up my old znc bouncer and will be connecting to IRC channels through IRC as a result
2017-12-08 04:06:58 <AWilcox[m]> Considering I went to Matrix to avoid freenode, this is fine with me :P
2017-12-08 12:27:04 <eleksir> ncopa, hmm... i _can_ file a bug... but i put a note on this issue not because of defect itself (i can rebuild package myself as i did with znc that was built without recoding support) but because of package quality assurance improvement
2017-12-08 12:40:46 <eleksir> done, #8259
2017-12-09 03:27:58 <h264> is php7-memcache (NOT php7-memcached) missing?
2017-12-09 03:28:03 <h264> why?
2017-12-10 18:55:12 <mitchty> quick question, i'm working on getting ghc working on x86, but due to ghc being a bootstrap then use the bootstrap to build type of compiler, how exactly do I go about getting that into aports? I doubt I have access to scp the cross compiler I build
2017-12-11 13:28:51 <clandmeter> ncopa: around?
2017-12-11 13:29:47 <clandmeter> shouldn't that stacksize fix be applied to py2 as well?
2017-12-11 13:30:55 <danieli> how did you fix it?
2017-12-11 13:31:52 <clandmeter> ?
2017-12-11 13:37:50 <danieli> what was the stacksize fix?
2017-12-11 13:37:53 <danieli> increasing it?
2017-12-11 13:48:51 <clandmeter> https://git.alpinelinux.org/cgit/aports/commit/main/python3?id=2026e1259422d4e0cf92391ca2d3844356c649d0
2017-12-11 18:29:16 <fcolista> is this host in production? svk-aoff-3ter-lxc-2.nwk.jwm2.net - Host unreachable (ping)
2017-12-11 18:29:52 <_ikke_> Is that an alpine question?
2017-12-11 18:30:40 <fcolista> _ikke_ no
2017-12-11 18:30:45 <fcolista> it was a wrong paste
2017-12-11 18:30:50 <fcolista> because I'm stupid
2017-12-11 18:31:24 <_ikke_> hehe
2017-12-11 18:31:28 <_ikke_> ;-)
2017-12-11 18:33:10 <fcolista> jirutka, you around?
2017-12-11 18:43:55 <fcolista> ncopa, re python pacakges. py-rsa does not install py2 or py3, even if in the APKBUILD there's install_if according with https://wiki.alpinelinux.org/wiki/APKBUILD_examples:Python
2017-12-11 18:43:59 <fcolista> i think
2017-12-11 18:44:20 <fcolista> that this is due to the fact that there's no python2 or python3 installed
2017-12-11 18:44:41 <fcolista> i'm talking about aws-cli
2017-12-11 18:44:46 <fcolista> https://bugs.alpinelinux.org/issues/8248
2017-12-11 18:45:39 <fcolista> aws-cli has py-rsa as dependency. So, my question is: if the package is going to install python, when it come to satisfy the py-rsa, will it find python?
2017-12-11 18:46:28 <fcolista> aws-cli pulls python3, and then pulls py-rsa. py-rsa when it come sto be installed, will find python?
2017-12-11 19:41:07 <clandmeter> fcolista: there is a bug with install if
2017-12-11 19:41:16 <fcolista> aha
2017-12-11 19:41:23 <fcolista> clandmeter, where did you find it?
2017-12-11 19:41:49 <clandmeter> In my setup :)
2017-12-11 19:41:56 <fcolista> :)
2017-12-11 19:41:59 <fcolista> good catch then
2017-12-11 19:42:35 <clandmeter> There is a bug on bugs
2017-12-11 19:42:40 <clandmeter> Don't know the if
2017-12-11 19:42:44 <clandmeter> Id..
2017-12-11 19:43:02 <clandmeter> Check my last comment, that should be it.
2017-12-11 19:45:17 <fcolista> last comment...where?
2017-12-11 19:46:35 <fcolista> Bah. I don't see the reason why in aws-cli there's _python=python3 and in depends packages with py-$name
2017-12-11 19:46:43 <fcolista> they should be all py3-$name
2017-12-11 19:46:51 <fcolista> since python3 is hardcoded
2017-12-11 19:47:10 <fcolista> in this way during the build process are pulled useless packages
2017-12-11 20:33:33 <TBK> Evening
2017-12-11 20:34:52 <TBK> Are there any plans to clean up the forum (https://forum.alpinelinux.org/)?
2017-12-11 21:31:16 <pavlix> ho folks
2017-12-11 21:33:01 <clandmeter> TBK are you interested in moderating it?
2017-12-11 21:35:09 <TBK> hmmmm if it was replaced with discourse then yes :P
2017-12-11 21:37:10 <clandmeter> why does the implementation matter?
2017-12-11 21:39:44 <TBK> It is just a matter of preference. I find newer forum designs easier to manage/keep an overview.
2017-12-11 21:41:29 <pavlix> !help
2017-12-11 21:42:07 <clandmeter> !me
2017-12-11 21:46:24 <pavlix> I thought there would be a bot to check when jirutka was last available. :)
2017-12-11 21:46:40 <clandmeter> :)
2017-12-11 21:46:45 <clandmeter> i havent seen him for a few days
2017-12-11 21:50:19 <pavlix> clandmeter: we talked about Alpine network configuration long at the beginning of this year maybe...
2017-12-11 21:51:29 <clandmeter> yes i think i remember
2017-12-11 22:22:59 <kaniini> skarnet has a network configuration manager he is working on
2017-12-11 22:23:08 <kaniini> looks pretty nice
2017-12-11 22:30:00 <pavlix> skarnet: when did you start working on that?
2017-12-11 22:35:33 <skarnet> last August. Not prioritizing it for now, though, beyond what's necessary for Adélie's installer to interact with wpa_supplicant.
2017-12-11 22:44:40 <pavlix> skarnet: did you consider any existing solutions before starting?
2017-12-11 22:45:51 <skarnet> Yes. And I'm not interested in discussing it.
2017-12-11 22:46:54 <pavlix> skarnet: fair enough
2017-12-11 22:51:53 <pavlix> skarnet: I might later add your project to my testing if it's easily confiurable and usable
2017-12-11 22:53:32 <skarnet> sure. For now it's just a library, that will evolve depending on Adélie's needs; the library works, but I don't expect to work on the full network manager for a while.
2017-12-11 22:55:10 <pavlix> skarnet: so what does it support now?
2017-12-11 22:56:11 <pavlix> skarnet: I'm currently evaluating NetworkManager, wicked and networkd + if free time allows, netifd and some individual tools like dhcpcd, odhcp6c, pyroute2 and more stuff.
2017-12-11 22:56:38 <skarnet> stick to what you have. For now, bcnm is just a library interface to wpa_supplicant.
2017-12-11 22:57:02 <skarnet> I'll let you know when it becomes something more :)
2017-12-11 22:58:09 <pavlix> skarnet: Aha, opensuse just runs wpa_supplicant as a service and calls out to it from wicked.
2017-12-11 22:58:39 <skarnet> yeah, that's sane.
2017-12-11 23:00:48 <pavlix> skarnet: it allows it to restart wickedd (e.g. due to update) without killing the wifi which NetworkManager AFAIK can't currently do
2017-12-11 23:01:22 <skarnet> yes, it's a good idea to separate services that can be separated.
2017-12-11 23:03:41 <pavlix> skarnet: you fork+exec wpa_supplicant?
2017-12-11 23:04:49 <skarnet> for the complete manager: I don't know yet. Probably not. As of now: no, the library assumes wpa_supplicant is running and just connects to a Unix socket.
2017-12-11 23:07:16 <pavlix> aha, so the same approach as wicked, right
2017-12-11 23:07:36 <pavlix> skarnet: does the library provide great help over directly using the API?
2017-12-11 23:08:07 <pavlix> skarnet: if yes, wouldn't it be the best way to contribute it to wpa_supplicant itself so everyone can use it?
2017-12-11 23:08:13 <pavlix> I mean just the wifi part
2017-12-11 23:09:53 <skarnet> yes, it does; no, it wouldn't; and everyone can already use the library.
2017-12-11 23:10:21 <skarnet> Sorry, I don't have the time tonight to explain/discuss why it wouldn't.
2017-12-11 23:10:56 <pavlix> skarnet: sure
2017-12-11 23:11:01 <pavlix> skarnet: good enough for me
2017-12-12 07:47:07 <fabled> ncopa, i wonder if we should have in default lua path a lua version independent path for packages that are pure lua and work with all lua versions
2017-12-12 07:57:26 <fabled> seems penlight installs to /usr/share/lua/common and symlinks that for each lua version
2017-12-12 08:08:55 <fabled> any objects if i add "common" to LUA_PATH and LUA_CPATH defaults (after the version specific paths) for all luas ?
2017-12-12 08:09:14 <fabled> ncopa, rnalrd, clandmeter ^
2017-12-12 08:09:56 <rnalrd> sounds good to me
2017-12-12 08:13:25 <fabled> mm.. CPATH does not need it. the .so:s are version specific. i fix just lua_path
2017-12-12 08:29:24 <clandmeter> +1
2017-12-12 08:39:51 <fabled> lua package.path changes: http://sprunge.us/PADh ?
2017-12-12 09:49:17 <ncopa> morning
2017-12-12 09:49:30 <danieli> god morgen
2017-12-12 09:49:43 <ncopa> fabled: sounds good to me
2017-12-12 09:49:55 <ncopa> i have an interesting problem with my machine :)
2017-12-12 09:50:11 <danieli> indulge me
2017-12-12 09:50:14 <ncopa> it started when i upgraded bash
2017-12-12 09:50:21 <ncopa> i saw "segfault"
2017-12-12 09:50:33 <ncopa> i did an apk fix bash to try reproduce
2017-12-12 09:50:38 <ncopa> and it reproduced
2017-12-12 09:50:51 <ncopa> dmesg showed that it was busybox add-shell
2017-12-12 09:51:03 <danieli> strange
2017-12-12 09:51:10 <ncopa> so i tried reproduce by running it manually
2017-12-12 09:51:16 <ncopa> and i think it segaulfted again
2017-12-12 09:51:30 <danieli> segfaults are always so fun
2017-12-12 09:51:31 <danieli> /s
2017-12-12 09:51:36 <ncopa> so far not very strange... probably just a bug in busybox
2017-12-12 09:51:45 <ncopa> the fun started now
2017-12-12 09:52:15 <ncopa> so i went to aports/main/busybox to check if -dbg was enabled
2017-12-12 09:52:21 <ncopa> i did 'ls'
2017-12-12 09:52:23 <ncopa> and it hang
2017-12-12 09:52:34 <ncopa> i tried to vim APKBUILD
2017-12-12 09:52:38 <ncopa> and it hanged
2017-12-12 09:53:06 <ncopa> dmesg tells me why...
2017-12-12 09:53:19 <ncopa> [  890.348580] grsec: bruteforce prevention initiated for the next 30 minutes or until service restarted, stalling each fork 30 seconds.  Please investigate the crash report for /bin/busybox[add-shell:4584] uid/euid:0/0 gid/egid:0/0, parent /bin/busybox[ash:4032] uid/euid:1000/1000 gid/egid:1000/1000
2017-12-12 09:53:38 <danieli> aah I see
2017-12-12 09:53:46 <ncopa> basically, grsec bruteforce prevention is preventing me to execute busybox
2017-12-12 09:53:51 <ncopa> 🙃
2017-12-12 09:55:46 <ncopa> since most tools are busybox, my system behaves very funny now :)
2017-12-12 09:55:57 <danieli> I can only imagine, lol
2017-12-12 09:56:15 <ncopa> ok. i was able to disble it
2017-12-12 09:56:25 <ncopa> sudo /bin/bash -c "echo 0 >  /proc/sys/kernel/grsecurity/deter_bruteforce "
2017-12-12 09:57:23 <danieli> still gotta figure out what triggers it
2017-12-12 10:02:47 <fabled> ncopa, was http://sprunge.us/bZXN applied to bb?
2017-12-12 10:03:38 <fabled> seems it was
2017-12-12 10:04:39 <fabled> ncopa, so I add the '/usr/share/lua/common'? or maybe just '/usr/share/lua' and keep the version paths under it separate
2017-12-12 10:05:44 <ncopa> hum
2017-12-12 10:06:02 <fabled> kind would be nice to make the path shorter
2017-12-12 10:06:02 <ncopa> im with either
2017-12-12 10:06:06 <ncopa> yeah
2017-12-12 10:06:09 <fabled> but i'm ok eitherway too
2017-12-12 10:06:16 <ncopa> are any other distros doing something similar?
2017-12-12 10:06:30 <fabled> i'm off for lunch now. hope to push it soon.
2017-12-12 10:06:33 <fabled> not sure what other distroes do
2017-12-12 10:07:08 <ncopa> what happens if you have code that is compatible with 5.1 - 5.3 and stored in common
2017-12-12 10:07:14 <ncopa> then lua 5.4 comes out
2017-12-12 10:07:24 <ncopa> which breaks it
2017-12-12 10:07:43 <fabled> latest debian is https://sources.debian.org/src/lua5.2/5.2.4-1.1/debian/patches/0002-debian-paths.patch/
2017-12-12 10:08:06 <fabled> then the module works or not, until lua5.4 specific or compatible code is added
2017-12-12 10:08:14 <fabled> i think breakage is less seldom now
2017-12-12 10:08:27 <ncopa> i think the current thinking was that we add symlinks to all lua versions we know works
2017-12-12 10:09:03 <ncopa> so we'd verify that lua5.4 works and then add a symlink in /usr/share/lua/5.4
2017-12-12 10:09:03 <fabled> i was hoping to avoid the manual work if in 95% cases it just works
2017-12-12 10:09:12 <fabled> and avoid extra symlinks
2017-12-12 10:09:27 <fabled> but i'm ok that way too if it's preferred
2017-12-12 10:09:44 <fabled> i'm off now. hope to resolve this discussion today.
2017-12-12 10:09:45 <ncopa> where do we put lua scripts that work with 5.1 and 5.2 but not 5.3?
2017-12-12 10:14:01 <ncopa> would also like to hear jirutka's opinion on this
2017-12-12 10:35:28 <ncopa> fabled: i wonder why we didnt have the add-shell patch in our aports? apparently the patch was not included in the 1.27 branch
2017-12-12 11:09:16 <fabled> ncopa, various ways to do it. common is in default path as last; so you could have 5.1+5.2 in their dirs (hardlinked or the otherone being a symlink), and the other in 'common'
2017-12-12 11:09:19 <fabled> but basically
2017-12-12 11:09:32 <fabled> it seems many of the lua packages ship universal .lua file that works everywhere
2017-12-12 11:09:41 <fabled> i'd like to get rid of the redundant copies and packages
2017-12-12 11:09:47 <fabled> use provides if needed
2017-12-12 11:14:10 <clandmeter> fabled,ncopa do we have an update for #8237 ?
2017-12-12 11:14:26 <fabled> clandmeter, i've been on vacation. so not yet. need to look at it
2017-12-12 11:15:17 <clandmeter> good, bad that it ended :)
2017-12-12 11:45:23 <fabled> clandmeter, i can reproduce it, and have a fairly good idea what's going wrong. need to run some other errands now, but try to get back to the issue later today or latest tomorrow
2017-12-12 12:12:29 <omniuwo> ncopa: ah, so it was grsecurity that created the 30s hang, I experienced that as well
2017-12-12 12:12:57 <ncopa> yes, its a feature :)
2017-12-12 12:13:12 <omniuwo> we also talked about busybox add-shell and bash in #alpine-linux yesterday
2017-12-12 12:13:19 <ncopa> i pushed fix for it
2017-12-12 12:13:28 <ncopa> apparently we fixed it long time ago
2017-12-12 12:13:35 <ncopa> and sent it upstream
2017-12-12 12:13:44 <ncopa> but was not inluced in the busybox 1.27 release for some reason
2017-12-12 12:13:54 <omniuwo> ah
2017-12-12 12:14:16 <ncopa> i could not find any traces of it in bugs.a.o, so i dont know why we missed it
2017-12-12 12:14:22 <ncopa> it should be fixed now though
2017-12-12 12:15:31 <omniuwo> yay! there i ran 'apk fix bash' without the segfault =)
2017-12-12 12:15:51 <omniuwo> (after upgrading busybox)
2017-12-12 12:39:25 <clandmeter> ncopa: the python thread stack size patch you applied, it would also fix the ultrajson stack size issue?
2017-12-12 12:39:39 <ncopa> yes
2017-12-12 12:39:45 <ncopa> i think it should
2017-12-12 12:40:20 <clandmeter> should it be applied to 2.7 as well?
2017-12-12 12:40:25 <ncopa> the ultrajson stack size issue is stupid though, and i think ultrajson should be fixed anyways
2017-12-12 12:40:29 <ncopa> python 2.7?
2017-12-12 12:40:35 <ncopa> yes, i think so
2017-12-12 12:40:42 <ncopa> we should report it upstream too
2017-12-12 12:41:09 <clandmeter> synapse (matrix server) is where i bumped into this issue (which is 2.7)
2017-12-12 13:02:23 <consus> So guys
2017-12-12 13:02:38 <ncopa> hi consus
2017-12-12 13:02:49 <consus> What's the appropriate way to contribute to alpine's documentation? Any plans to retire wiki? Or it's still the preffered way?
2017-12-12 13:02:53 <consus> Hi ncopa
2017-12-12 13:03:17 <consus> *preferred
2017-12-12 13:03:31 <ncopa> thats a good question
2017-12-12 13:03:50 <ncopa> we have talked about moving parts of the wiki docs into something else
2017-12-12 13:04:04 <ncopa> as official documentation
2017-12-12 13:04:14 <ncopa> while we keep the wiki for community docs
2017-12-12 13:04:39 <consus> I vaguely remember that ^7heo was suggesting asciidoctor
2017-12-12 13:05:16 <ncopa> clandmeter started with https://docs.alpinelinux.org/Home
2017-12-12 13:05:24 <ncopa> not sure what the status there is
2017-12-12 13:05:58 <consus> Aha
2017-12-12 13:06:37 <consus> So I should talk with clandmeter2 about contirbuting, right? Or it's wiki for me since I'm no Alpine developer?
2017-12-12 13:07:07 <ncopa> i think the idea was that we track the official docs in git
2017-12-12 13:07:26 <ncopa> which means that you could still be able to help bu doing git pull requests
2017-12-12 13:07:30 <ncopa> by*
2017-12-12 13:07:38 <consus> Can't find anything realted to docs on git.alpinelinux.org
2017-12-12 13:07:52 <consus> Is it github based?
2017-12-12 13:07:55 <ncopa> correct, it dont think we officiall started the docs.a.o thingy
2017-12-12 13:08:13 <consus> clandmeter2: Hi
2017-12-12 13:09:00 <consus> clandmeter2: Could you tell us what's the status of docs.alpinelinux.org and how can one contribute to the docs?
2017-12-12 13:09:13 <clandmeter> hi
2017-12-12 13:10:14 <clandmeter> we need somebody who can spend time on it.
2017-12-12 13:10:43 <consus> What should be done?
2017-12-12 13:10:52 <consus> Any roadmap?
2017-12-12 13:10:57 <ncopa> i think figuring out how to do it
2017-12-12 13:11:01 <consus> Ah
2017-12-12 13:11:04 <ncopa> we need a roadmap too
2017-12-12 13:11:06 <ncopa> i think
2017-12-12 13:11:41 <ncopa> clandmeter: do you think you could email the current status of docs.a.o project to alpine-devel?
2017-12-12 13:11:54 <ncopa> and what we need figure out
2017-12-12 13:12:22 <ncopa> the current problem is we dont know what we want or how we want it
2017-12-12 13:12:34 <ncopa> so it is difficult for people who wants contribute to know how
2017-12-12 13:12:43 <clandmeter> we need to document it ;-)
2017-12-12 13:12:55 <consus> Also it would be nice to highlight what's wrong with the wiki
2017-12-12 13:13:10 <consus> Because e.g. archlinux project has a really nice one
2017-12-12 13:13:17 <ncopa> agree
2017-12-12 13:13:27 <ncopa> the problem is that wiki grows organically
2017-12-12 13:13:30 <consus> I mean even using Gentoo I'm consulting theirs wiki more often than Gentoos
2017-12-12 13:13:45 <ncopa> which means that arch linux has some really good people doing docs
2017-12-12 13:13:49 <clandmeter> arch wiki is really nice
2017-12-12 13:14:05 <clandmeter> but the audience is different
2017-12-12 13:14:11 <ncopa> it means that that someone has a big overview and a "vision" on how the docs should be
2017-12-12 13:14:23 <ncopa> someone has put the frames around it
2017-12-12 13:14:42 <ncopa> so its easy for people who wants to contribute to know how to properly do it
2017-12-12 13:14:50 <consus> True
2017-12-12 13:15:05 <ncopa> in alpine, we have just put the wiki out there
2017-12-12 13:15:09 <ncopa> and hope that it fixes itself
2017-12-12 13:15:15 <ncopa> which ofcourse will not happen :)
2017-12-12 13:15:49 <clandmeter> I would like to have documentation stored in git
2017-12-12 13:15:51 <_ikke_> `The Arch communitu is a lot bigger, so more people able / willing to contribute
2017-12-12 13:16:03 <clandmeter> and contribute with pr's or patches.
2017-12-12 13:16:31 <ncopa> i dont think its only a question of amount of contributors
2017-12-12 13:16:42 <ncopa> i bet arch has someone who can take the lead for docs
2017-12-12 13:17:06 <consus> The problem with that approach is that people would need to a) go with clone/commit/push/create-pull-request thing b) need to learn asciidoctor/md/whatever
2017-12-12 13:17:29 <clandmeter> right, i think thats what we are missing. somebody who takes the lead on documentation.
2017-12-12 13:17:40 <ncopa> you need learn mediawiki syntax too
2017-12-12 13:17:42 <consus> That's okay for the code (as you naturally need a source code to write patches) but it's not that simple with docs
2017-12-12 13:17:43 <clandmeter> because currently we are going in circles...
2017-12-12 13:17:57 <consus> But there is a little box out there with a full help an the language
2017-12-12 13:18:03 <clandmeter> i think its the 4th time i had this discussion.
2017-12-12 13:18:26 <consus> Maybe it would be nice to create a list on the wiki
2017-12-12 13:18:34 <consus> Like procs/cons of each method
2017-12-12 13:18:52 <consus> And ask people to add their opinions there
2017-12-12 13:18:57 <ncopa> we need someone who has a good idea over the big picture overview
2017-12-12 13:19:00 <clandmeter> i think a thread on the ml would be better
2017-12-12 13:19:11 <ncopa> i think a thread on ml would be better start too
2017-12-12 13:19:45 <consus> Well, I guess I should subscribe then
2017-12-12 13:19:51 <clandmeter> at least we wont have a 5th time we had this discussion 😆
2017-12-12 13:20:25 <ncopa> and i think it does not matter what tool we use, or how we store it (eg wiki vs md in git), but we need someone who takes the lead for docs
2017-12-12 13:20:49 <ncopa> consus: i am very happy to hear that you are willing to contribute
2017-12-12 13:21:01 <ncopa> and i want to make it easy for you to do so
2017-12-12 13:21:25 <clandmeter> i dont mind to help with docs, but i dont think i want to be the doc lead.
2017-12-12 13:21:25 <consus> ^^
2017-12-12 13:21:40 <fabled> ncopa, the commit referencing syntax does not work in alpine redmine anymore?
2017-12-12 13:21:56 <clandmeter> consus: you brought this up some time ago right?
2017-12-12 13:22:01 <consus> Yep
2017-12-12 13:22:08 <ncopa> fabled: i think it works: alpine:commit:"<hash>"
2017-12-12 13:22:28 <fabled> mmm... is apk-tools there?
2017-12-12 13:22:39 <clandmeter> are you ok with supporting documentation for a longer period?
2017-12-12 13:22:56 <consus> Sure, why not
2017-12-12 13:23:20 <ncopa> fabled: maybe apk-tools:commit:"<hash>"
2017-12-12 13:23:27 <clandmeter> good, then lets start to write down what we want.
2017-12-12 13:23:57 <fabled> ncopa, oh. it's not just pulling anymore?
2017-12-12 13:24:52 <ncopa> hum
2017-12-12 13:24:59 <ncopa> thats weird
2017-12-12 13:25:03 <ncopa> i think it pulls
2017-12-12 13:25:16 <fabled> it works with old commits, but not the one i just pushed
2017-12-12 13:25:36 <consus> clandmeter: You mean infra or content?
2017-12-12 13:26:49 <ncopa> i think clandmeter means infra and roadmap
2017-12-12 13:27:07 <clandmeter> consus: i create #alpine-docs
2017-12-12 13:27:17 <ncopa> :)
2017-12-12 13:28:34 <ncopa> reminds me what AWilcox[m] mentioned earlier about "projects"
2017-12-12 13:29:41 <ncopa> http://lists.alpinelinux.org/alpine-devel/5811.html
2017-12-12 14:39:25 <jomat> This file shouldn't be in two completely different packages, should it? https://pkgs.alpinelinux.org/contents?file=gschemas.compiled
2017-12-12 14:41:13 <clandmeter2> i think it shouldnt be in any pkgs?
2017-12-12 15:04:28 <jomat> I don't know what this is, I was just testing https://github.com/dino/dino when it started sigtapping I saw with strace that it tries to open this file … Tests finished.
2017-12-12 15:30:59 <clandmeter2> jomat, i think it should be generated post-install.
2017-12-12 17:21:02 <opendata> Hey
2017-12-12 17:21:28 <opendata> Would bootstrapping armv7 work on aarch64
2017-12-12 19:18:11 <fcolista> how to deal with packages renamed on upstream?
2017-12-12 19:18:31 <fcolista> I can just rename the package and set replace=$oldpackagename ?
2017-12-12 19:19:26 <fcolista> Yes
2017-12-12 19:19:26 <fcolista> replaces
2017-12-12 19:19:27 <fcolista>     Package(s) that this package replaces. This package will "take over" files owned by packages listed in the replaces variable. This is useful when files move from one package to another, or when a package gets renamed.
2017-12-12 20:19:06 <tboerger[m]> wtf... i'm not able to extract the alpine tarball on my mac onto a sd card Oo
2017-12-12 20:19:36 <duncan^> probably because it has different tar options
2017-12-12 20:21:55 <tboerger[m]> 
ACTION sent a long message: tboerger[m]_2017-12-12_20:21:55.txt <https://matrix.org/_matrix/media/v1/download/matrix.org/TZMoAykPnQstJSpmsFCYpFXC>

2017-12-12 20:22:05 <tboerger[m]> alos doesn't work with gnu tar
2017-12-12 20:26:46 <tboerger[m]> and after the ~10th time it magically just worked -.-
2017-12-12 21:18:52 <ollieparanoid[m]> ncopa: you're the maintainer of qt5-qtbase and it was flagged out of date. do you roughly know when it will be updated? I'm not looking forward to it, quite the contrary: it means breakage for postmarketOS because we also package some qt5- packages, and the version must be in sync with qt5-qtbase. So it would be nice to know the timeframe if that is possible.
2017-12-12 21:18:53 <ollieparanoid[m]> consus, clandmeter: I prefer mediawiki a lot. It's used for a lot of good distro wikis already (e.g. Arch, Gentoo) and people coming from those distros will have it easier to contribute content. I also think that it is easier to edit and leads to more contributions. (and yeah we also use mediawiki in postmarketOS after migrating from GitHub wiki, which would make it easier for people from pmOS to contribute to
2017-12-12 21:18:54 <ollieparanoid[m]> the Alpine wiki if that counts for anything). And mediawiki can be accessed with git nowadays :)
2017-12-12 21:18:55 <ollieparanoid[m]> clandmeter,kaniini: could you make #alpine-docs work with the matrix server? (I think kaniini maintains it?)
2017-12-12 21:24:08 <clandmeter> I can fix that but not now
2017-12-12 21:25:29 <clandmeter> I need to migrate the matrix server. Current one is too slow.
2017-12-12 21:25:55 <ollieparanoid[m]> thanks, no hurry!
2017-12-12 21:26:03 <danieli> ollieparanoid[m]: does postmarketos not have its own repositories?
2017-12-12 21:26:16 <ollieparanoid[m]> to which software?
2017-12-12 21:26:49 <danieli> I mean in general - it seems to have quite a few specific requirements
2017-12-12 21:27:01 <clandmeter> Same just diff hw
2017-12-12 21:27:44 <danieli> from time to time I read about stuff breaking pmos, or some kinky situation involving some package, like qt5-qtbase in this case
2017-12-12 21:28:07 <ollieparanoid[m]> danieli: it has its own repositories, but we use our own one in addition to the ones from alpine. and we can take it if it breaks from that, as it's in an early state right now. it would just be useful to know when :)
2017-12-12 21:28:44 <ollieparanoid[m]> the binary repo is pretty new actually (a few weeks), so breakage from that hasn't happened much
2017-12-12 21:29:36 <danieli> ah I see
2017-12-12 22:10:14 <danieli> the chromium image was manually flagged because someone was missing 'fopencookie'
2017-12-12 22:13:42 <kaniini> yes, chromium needs to be used with 3.7 libc or newer
2017-12-12 22:14:38 <danieli> mhm
2017-12-12 22:14:45 <danieli> you ported fopencookie, right?
2017-12-12 22:25:58 <kaniini> yes
2017-12-12 22:26:00 <kaniini> 'ported'
2017-12-12 22:26:03 <kaniini> i wrote a new implementation of it
2017-12-12 22:26:14 <kaniini> and patiently made it conform to dalias's specifications
2017-12-13 09:36:12 <fabled> jirutka, any comments on the LUA_PATH approach i was talking about yesterday?
2017-12-13 09:43:22 <clandmeter> ncopa:  s390x is MIA?
2017-12-13 16:38:12 <crich> Hello everybody
2017-12-13 16:38:40 <crich> could somebody point me to an howto or at least an overview on how to compile your own kernel for Alpine ?
2017-12-13 16:39:28 <crich> I found only this: https://wiki.alpinelinux.org/wiki/Custom_Kernel  which lacks any information
2017-12-13 17:03:44 <duncan^> crich: You configure the kernel, build it, then use mkinitfs to produce your initramfs.
2017-12-13 17:04:05 <duncan^> crich: What problems are you having?
2017-12-13 17:28:12 <crich> dunca^: I have this issue: https://bugs.alpinelinux.org/issues/8282
2017-12-13 17:32:42 <crich> duncan^: I'm now upgrading to Alpine 3.7 and give it another shot, but then i'll try to recompile my kernel and see if i can make it work somehow. last time i ran into trouble with the initrd.. will see if it works now
2017-12-13 17:58:50 <duncan^> I don't know how to help with that, but it should be straightforward to compile your kernel, and then use mkinitfs (or your favorite other tool) to produce the initramfs.
2017-12-13 19:22:43 <duncan^> I'm trying to get kexec-tools to build, and I keep encountering an error, where it gets to "checking build system type... invalid configuration `x86_64-alpine-linux-musl' , machine `alpine-linux-musl'.
2017-12-13 19:22:54 <duncan^> Does anyone have any ideas how one could fix this?
2017-12-13 19:32:39 <duncan^> Aha, found it
2017-12-13 19:32:43 <duncan^> config.sub :(
2017-12-13 19:40:48 <duncan^> OK. So I made a package for kexec-tools. Is it wanted?
2017-12-13 19:47:40 <przemoc> duncan^: if you made it, then possibly you needed it or wanted seeing it available in AL, right?  so there is no need to ask.  send patch to alpine-aports ML (or make PR on github) adding your package to testing.
2017-12-13 19:58:27 <duncan^> cool
2017-12-13 19:58:29 <duncan^> thanks!
2017-12-13 20:00:32 <duncan^> what do I put in the "maintainer" field?
2017-12-13 20:00:42 <duncan^> (since, I do not maintain anything)
2017-12-13 20:00:58 <duncan^> I put my name in the contributor field. What's the right way to do this>
2017-12-13 20:01:00 <duncan^> ?
2017-12-13 20:01:32 <przemoc> you're not willing to maintain it in future?
2017-12-13 20:02:35 <duncan^> Oh, sure I am.
2017-12-13 20:04:14 <przemoc> then put yourself as maintainer.  if the package is created for your needs, then you'll be the one most up-to-date regarding it.  AL doesn't have requirement that maintainer needs commit rights, at least I haven't heard or read it.
2017-12-13 20:04:26 <duncan^> Ah, sure. Thanks.
2017-12-13 20:04:38 <przemoc> therefore contributor field can be removed
2017-12-13 20:06:10 <przemoc> 
ACTION also strongly believes that contributor fields are needlessly polluting APKBUILD files for no good reason, as git log is better anyway.

2017-12-13 20:06:41 <przemoc> I suggested removing them in the past, but there was some opposition.
2017-12-13 20:07:56 <przemoc> (removing from the old ones, obviously)
2017-12-13 20:10:15 <przemoc> AL is sadly diverged between mailing list and pull requests, therefore even if you are maintainer of a package, it can be changed without you even knowing, because some committer will simply grab a patch or PR for the package that someone else submitted.
2017-12-13 20:10:37 <duncan^> What's the policy with the "source" line? I've got it so that the necessary patch is part of this, and a checksum appears at the bottom, but it seems unneccessary.
2017-12-13 20:10:46 <duncan^> Is that the "canonical" way to do it?
2017-12-13 20:12:35 <przemoc> somehow people usually don't care who is maintainer, so they don't CC them in patches or PRs.
2017-12-13 20:12:39 <przemoc> duncan^: what seems unnecessary?
2017-12-13 20:12:47 <przemoc> I'm not sure I understand your question
2017-12-13 20:14:19 <przemoc> http://wiki.alpinelinux.org/wiki/APKBUILD_Reference#source  I assume you've seen that already
2017-12-13 20:15:48 <duncan^> In the "source" entry of apkbuild, we may specify the source packages fetched from the internet, as well as any local resources such as packages. But, any local patches are already in the aports git repository, so, why do certain apkbuilds such as https://github.com/alpinelinux/aports/blob/master/main/gcc/APKBUILD have these patches referenced in the "source" line.
2017-12-13 20:16:46 <duncan^> OK, I think I understand it better now. Sorry for the noise.
2017-12-13 20:19:37 <przemoc> no need to apologize, haven't seen any noise yet.  it's about being explicit.  abuild doesn't use anything that isn't mentioned in APKBUILD.  applying patches is automatic process since a few major AL releases back and it works by looking into sources for files with .patch suffix.  checksumming is to be sure we're using what we're meant to.
2017-12-13 20:23:20 <duncan^> Makes sense
2017-12-13 20:24:32 <przemoc> important things are:  if patch is not in source="...", it won't be applied.  if patch doesn't have .patch extension, it also won't be applied either.
2017-12-13 20:30:10 <duncan^> What is the policy on commit messages? Is there a specific format?
2017-12-13 20:40:21 <TBK> testing/xxxxxxx: new aport
2017-12-13 20:42:28 <przemoc> `git log` in aports clone should give you pretty good idea
2017-12-13 20:49:17 <duncaen> https://github.com/voidlinux/void-packages/commit/53fed9bb202b updated patches for chromium 63
2017-12-13 20:51:26 <duncan^> right
2017-12-13 20:51:32 <duncan^> fair
2017-12-13 20:52:09 <duncan^> pull request made...
2017-12-13 21:03:23 <duncan^> Oh dear, Travis-CI test is failing D:
2017-12-13 21:03:46 <duncan^> "Testsuites (abuild check) are required or needs to be explicitly disabled!"
2017-12-13 21:05:13 <duncan^> I assume adding options="!check" would be sufficient?
2017-12-13 21:08:32 <TBK> if there is no test suite, then yes options="!check" is the way to go.
2017-12-13 21:30:45 <duncan^> Yep! That worked, and it needed to depend on linux-headers (of course)
2017-12-13 21:48:34 <duncan^> cheers, folks
2017-12-14 01:46:04 <Xe> crazy idea: would a patch to remove python2 and everything that depends on it (and not python3) be rejected? I am also half wondering if python2 and py2-pip should remain as the only two python2 packages in community just so that people that hard stop absolutely need python 2 can have it.
2017-12-14 02:08:01 <Shiz> i would be against it
2017-12-14 05:23:29 <scadu> > i would be against it
2017-12-14 05:23:30 <scadu> /me too
2017-12-14 05:24:36 <danieli> ^
2017-12-14 05:24:41 <danieli> would be too extreme
2017-12-14 05:24:49 <scadu> Python2 is still heavy used and this proposal does not make any sense for me
2017-12-14 05:26:10 <scadu> Our synapse instance runs on Alpine, right? And it's python2 only
2017-12-14 05:26:23 <scadu> But there's much more anyway
2017-12-14 05:26:23 <danieli> that's an example, yes
2017-12-14 05:35:22 <scadu> Xe: so, if you are bored you can always look at b.a.o for challenges :p
2017-12-14 05:35:41 <danieli> wat, algitbot expands those now?
2017-12-14 05:35:49 <Arcaire> b.a.o
2017-12-14 05:35:52 <Arcaire> seems so
2017-12-14 05:35:59 <danieli> bugs.a.o would be so much less ambiguous
2017-12-14 05:36:06 <danieli> oh well
2017-12-14 05:37:14 <_ikke_> clandmeter: ^^
2017-12-14 06:22:20 <BitL0G1c> can unbound & the py unit test PR's be pulled now ?
2017-12-14 08:12:44 <clandmeter> _ikke_: ?
2017-12-14 08:12:54 <clandmeter> re py?
2017-12-14 08:13:07 <_ikke_> clandmeter: ?
2017-12-14 08:13:25 <clandmeter> you mentioned me?
2017-12-14 08:13:59 <_ikke_> ah, that was regarding the expansion of b\.a.o versus bugs.a.o
2017-12-14 08:14:32 <clandmeter> i think ncopa added it some time ago.
2017-12-14 08:15:08 <clandmeter> whats wrong with it?
2017-12-14 08:15:37 <clandmeter> b.a.o is shorter than bugs.a.o :)
2017-12-14 08:16:19 <danieli> g.a.o
2017-12-14 08:16:24 <danieli> p.a.o
2017-12-14 08:16:44 <danieli> hmm ok
2017-12-14 08:16:49 <_ikke_> clandmeter: Most people use bugs.a.o however
2017-12-14 08:17:03 <danieli> b.a.o feels odd, and it's ambiguous
2017-12-14 08:17:27 <clandmeter> they dont work both?
2017-12-14 08:17:41 <_ikke_> nope
2017-12-14 08:17:49 <_ikke_> bugs.a.o
2017-12-14 08:17:50 <_ikke_> b.a.o
2017-12-14 08:17:56 <_ikke_> only one response
2017-12-14 08:18:08 <clandmeter> bugs.a.o
2017-12-14 08:18:15 <clandmeter> hmm
2017-12-14 08:18:16 <clandmeter> ok
2017-12-14 08:18:22 <danieli> if both worked, i wouldn't mind
2017-12-14 08:18:37 <clandmeter> can you create a ticket? ;-)
2017-12-14 08:19:05 <clandmeter> ill check it when i have some free time, which is probable tomorrow.
2017-12-14 08:19:20 <clandmeter> going to migrate martrix now
2017-12-14 08:19:27 <clandmeter> matrix even..
2017-12-14 08:19:33 <danieli> if only I had access I'd do it :')
2017-12-14 08:19:42 <danieli> and good luck, sounds like you'll need it
2017-12-14 08:20:09 <clandmeter> it should be fairly simple as i have prepared myself (which mostly means i missed a few things)
2017-12-14 08:20:43 <danieli> welcome to the life of app admin
2017-12-14 08:21:53 <clandmeter> ok going to shutdown synapse, migrate db, sync data and cross fingers.
2017-12-14 09:55:36 <crich> Hello
2017-12-14 09:56:35 <crich> I accidently create #8282 as anonymous, can someone move this issue so that it belongs to my user or should i recreate this issue with my registered user and someone deletes the old one ?
2017-12-14 09:57:09 <_ikke_> crich: You cannot change the author of an issue
2017-12-14 10:04:27 <ncopa> crich: maybe you can subscribe to the issue?
2017-12-14 10:06:05 <ncopa> crich: the issue sounds bad. how long time does it take to reproduce it?
2017-12-14 10:06:26 <crich> I'm watching it, so i get updated that's ok for me, but i can't change the Alpine Version and such anymore
2017-12-14 10:07:06 <crich> @ncopa: between 5-30 minutes. yes it's very bad.
2017-12-14 10:08:05 <ncopa> can you provoke it to happen? or you just have to wait?
2017-12-14 10:11:34 <crich> @ncope: as i wrote in the bug report: I start fio on the Hypervisor and run 2 Alpine VMs in HVM mode and run fio inside as well. then it happens around 5-30 minutes, but i need to wait for it. If running under production conditions it happens randomly after a few days or weeks
2017-12-14 10:15:47 <ncopa> crich: does it happen with vanilla kernel too?
2017-12-14 10:16:23 <ncopa> "NOTE3: I tried both the grsec and vanilla Kernels and it makes no difference,"
2017-12-14 10:16:38 <ncopa> is that for the dom0 or the domU?
2017-12-14 10:16:49 <crich> @ncopa: dom0
2017-12-14 10:17:20 <ncopa> so it happens with vanilla kernel dom0
2017-12-14 10:17:40 <crich> @ncopa: I can run any kind of OS in domU, doesn't matter it happens as well, tried with other linuxes, Windows, freeBSD.. alpine is easiest to setup ;)
2017-12-14 10:17:48 <ncopa> ok
2017-12-14 10:18:34 <clandmeter> ncopa: when you have time, can you look into the py2 stack issue?
2017-12-14 10:18:42 <crich> @ncopa: I'm trying now to copy the Kernel / initrd/ modules from xen-server to my dom0 and check if i can reproduce the issue with that setup, but using the alpine distro tools like qemu and so forth
2017-12-14 10:18:50 <ncopa> crich: so, just to confirm, it happens with vanilla kernels too, not only the hardened kernel?
2017-12-14 10:19:02 <crich> @ncopa: yes
2017-12-14 10:19:05 <ncopa> ok
2017-12-14 10:19:09 <ncopa> then it can be reported upstream
2017-12-14 10:19:14 <ncopa> to kernel.org
2017-12-14 10:21:09 <ncopa> or maybe even to xen
2017-12-14 10:21:24 <ncopa> clandmeter: i'll look into it now
2017-12-14 10:21:54 <clandmeter> nice thx
2017-12-14 10:22:06 <clandmeter> bumped into it again with synapse upgrade.
2017-12-14 11:49:50 <duncan^> Are reproducible builds a goal in the future?
2017-12-14 11:50:11 <^7heo> no, only in the past.
2017-12-14 12:40:02 <crich> @ncope: i copied a qemu-system-i386 from xenServer (and all necessary libraries including libc) to my Alpine Hypervisor and made another test and voilá it seems to run stable so far,  2hours of high IO and high load and no kernel panic
2017-12-14 12:42:26 <crich>  @ncopa: interestingly the versions of qemu-system-i386 in XenServer and Alpine are exactly the same (2.2.1) but all the libraries are different of course, so i guess there might be a ulibc issue
2017-12-14 12:43:00 <ncopa> kernel panic is bug in kernel
2017-12-14 12:43:47 <ncopa> duncan^: reproducible builds may be a goal in the future yes, depends a bit on the "cost"
2017-12-14 12:44:41 <ncopa> crich: kernel should not panic regardless of the userspace (libc and version of qemu and libs)
2017-12-14 12:45:45 <consus> ncopa: Hey
2017-12-14 12:45:56 <consus> Guys
2017-12-14 12:46:16 <ncopa> hi
2017-12-14 12:46:33 <consus> memcached is broken in v3.7
2017-12-14 12:46:59 <consus> There is a bug with seccomp I think
2017-12-14 12:47:13 <consus> It cannot call epoll_wait after seccomping itself
2017-12-14 12:47:26 <consus> seccomp(SECCOMP_SET_MODE_FILTER, 0, {len=19, filter=0xaeb9fc255e0}) = 0
2017-12-14 12:47:26 <consus> epoll_pwait(3, 0x6ef5dd31ce20, 32, 987, NULL, 8) = -1 EACCES (Permission denied)
2017-12-14 12:47:45 <consus> At least via TCP socket
2017-12-14 12:48:00 <consus> Am I the only one who's facing it?
2017-12-14 12:51:23 <crich> ncopa: I agree totally, but it seems the ulibc implementation of munmap or around munmap triggers the kernel panic. I think the kernel developers will play the ball back to the ulibc developers
2017-12-14 13:35:44 <ncopa> consus: can you please report it on bugs.a.o?
2017-12-14 13:35:51 <consus> Okay
2017-12-14 13:36:13 <ncopa> and then maybe report it upstream to memcached?
2017-12-14 13:36:36 <ncopa> crich: i doubt they will play it back to libc
2017-12-14 13:37:49 <ncopa> <pipacs> yeah i guessed so... anyway, this looks like some very low level page management list corruption, like freeing a page twice or straight memory corruption
2017-12-14 13:38:16 <ncopa> may even be a security vuln in kernel
2017-12-14 13:40:23 <consus> Let's sell it on the black market!
2017-12-14 13:40:28 <consus> And buy autotesting stuff ^^
2017-12-14 13:40:47 <ncopa> :)
2017-12-14 13:47:39 <consus> Btw
2017-12-14 13:47:42 <consus> Kernel panic
2017-12-14 13:47:55 <consus> Only unmap has issues?
2017-12-14 13:48:15 <consus> I'm seeing VERY slow responses from mmaped files
2017-12-14 13:58:44 <crich> consus: the panic almost always has a backtrace starting in munmap syscall
2017-12-14 13:58:49 <consus> ^^
2017-12-14 13:59:22 <consus> And you even don't have to be root?
2017-12-14 13:59:31 <crich> seems to be present in any kernel  that i've tested ranging from 3.1 to 4.9
2017-12-14 13:59:31 <consus> Sweet ^^
2017-12-14 13:59:38 <crich> not sure if that's the case
2017-12-14 14:00:32 <crich> haven't yet figured out which qemu triggers the panic, weather its the dom0 one or the domU one. But i guess root or not root doesn't matter
2017-12-14 14:00:43 <danieli> "which qemu"?
2017-12-14 14:00:56 <crich> 2.2.1 (with muslibc)
2017-12-14 14:01:10 <danieli> sounds like you're mixing up xen and qemu somehow
2017-12-14 14:02:00 <crich> "/usr/lib/xen/bin/qemu-system-i386"
2017-12-14 14:02:08 <danieli> oh I see
2017-12-14 14:02:27 <crich> check #8282 for the backtrace
2017-12-14 14:02:36 <danieli> ah, that one, gotcha
2017-12-14 14:04:22 <danieli> qemu on xen is only really used for dm emulation
2017-12-14 14:06:24 <crich> i know, xen has no own dm emulation capabilities so it needs qemu for that
2017-12-14 14:07:12 <danieli> munmap... bleh
2017-12-14 14:07:57 <crich> anyway, whats the best approach here ? XenServers qemu works fine for me, so i could just stick with that, but this means to have a bunch of extra large libs, which i don't really like to carry all the time
2017-12-14 14:10:17 <crich> would be nice to try a newer qemu (like 2.10 is out already with some nice features), but I'm not sure if this means recompiling the whole Alpine Xen package as well :/
2017-12-14 15:11:50 <consus> There is no perf in Alpine, right?
2017-12-14 15:36:32 <ncopa> consus: correct, i looked at it right before the 3.7 release
2017-12-14 15:36:38 <consus> :(
2017-12-14 15:36:42 <ncopa> so i think we have something half done
2017-12-14 15:36:55 <ncopa> there is an issue about it on bugs.a.o too i think
2017-12-14 15:37:06 <consus> CGit is eating CPU a lot
2017-12-14 15:37:27 <consus> Can you suggest anything that I can use for the profiling except for perf?
2017-12-14 15:40:31 <skarnet> strace
2017-12-14 15:41:01 <consus> Nah
2017-12-14 15:41:21 <skarnet> ¯\_(ツ)_/¯
2017-12-14 15:41:26 <consus> It mmaps git files
2017-12-14 15:41:47 <consus> And then it goes to 100% CPU processing
2017-12-14 15:41:56 <consus> And I want to find out what's going on
2017-12-14 15:42:24 <consus> So it's nothing than strace can show me at this point
2017-12-14 15:42:28 <consus> *that
2017-12-14 15:42:44 <skarnet> if you say so
2017-12-14 15:42:50 <consus> Well
2017-12-14 15:43:01 <consus> If strace can trace function calls -- I'm all ears
2017-12-14 15:43:22 <skarnet> have you tried running it?
2017-12-14 15:43:26 <consus> Of course
2017-12-14 15:43:41 <skarnet> ok, so it's not busylooping around a system call
2017-12-14 15:43:49 <consus> Yes
2017-12-14 15:43:54 <skarnet> which is a common way of getting 100%
2017-12-14 15:44:01 <skarnet> that's why I was suggesting it
2017-12-14 15:44:03 <consus> Maybe it's internal lua
2017-12-14 15:44:06 <consus> I dunno
2017-12-14 15:44:15 <consus> The easiest way is to perf top the process
2017-12-14 15:44:28 <consus> Well, I'll build it myself then
2017-12-14 15:44:33 <consus> No big deal
2017-12-14 15:44:45 <skarnet> you may find it a bigger deal than you think
2017-12-14 15:45:02 <consus> It does work with musl?
2017-12-14 15:45:02 <skarnet> perf authors made no effort to make it libc-portable
2017-12-14 15:45:08 <consus> Ah crap
2017-12-14 15:45:17 <skarnet> I'be built it with musl after lots of dirty patching
2017-12-14 15:45:27 <skarnet> and could never make it work properly
2017-12-14 15:45:40 <skarnet> didn't have the time to check whether my dirty patching broke it
2017-12-14 15:45:40 <ncopa> i think i made it build
2017-12-14 15:45:47 <ncopa> but got caught up in the packaging
2017-12-14 15:45:49 <skarnet> or something else was the problem
2017-12-14 15:46:00 <consus> Can you share the binary?
2017-12-14 15:46:01 <ncopa> clandmeter: i pushed fix for python2
2017-12-14 15:46:31 <clandmeter> a nice :)
2017-12-14 15:46:32 <clandmeter> thx
2017-12-14 15:46:32 <skarnet> nope, I deleted all that crap, sorry
2017-12-14 15:46:47 <clandmeter> will you push it to 3.7?
2017-12-14 15:47:05 <ncopa> clandmeter: i spent the time on reporting it upstream and prepare a proper fix for upstream: https://bugs.python.org/issue32307
2017-12-14 15:47:14 <ncopa> yeah, i will push it to 3.7
2017-12-14 15:49:56 <clandmeter> nice, you spend a lot of time in it.
2017-12-14 15:53:39 <ncopa> yeah, i tested things on openbsd, netbsd, freebsd and centos
2017-12-14 16:07:23 <clandmeter> i think we should mention it also on synapse tracker. those users were affected by it.
2017-12-14 16:10:33 <crich> can anyone tell me which patches where applied to the qemu packge in order to compile it for alpine ? I'm trying to reqcompile a recent version of qemu, but configure drops ERROR: glib-2.22 gthread-2.0 is required to compile QEMU
2017-12-14 16:11:00 <danieli> crich: it's in the APKBUILD
2017-12-14 16:11:32 <danieli> is it just me or is pkgs.a.o really slow?
2017-12-14 16:11:40 <danieli> hm, was just slow the first moment
2017-12-14 16:12:15 <danieli> crich: see the .patch files in https://git.alpinelinux.org/cgit/aports/tree/main/qemu?h=master
2017-12-14 16:13:45 <crich> danieli: thx for the pointer
2017-12-14 16:14:45 <Shiz> danieli: cgit is slow latly it seems
2017-12-14 16:14:58 <danieli> Shiz: I meant pkgs, not cgit
2017-12-14 16:15:02 <danieli> but cgit hasn't been the fastest either
2017-12-14 16:15:12 <Shiz> yeah p.a.o had issues too
2017-12-14 16:15:24 <danieli> no wonder really
2017-12-14 16:15:30 <danieli> doesn't seem like a system that will scale awfully well
2017-12-14 16:15:40 <danieli> are anyone working on a reimplementation of pkgs.a.o?
2017-12-14 16:20:18 <clandmeter> i have something in my local repo
2017-12-14 16:20:25 <clandmeter> i plan to commit it next week
2017-12-14 16:20:49 <clandmeter> it will split all branches into its own db.
2017-12-14 16:23:07 <_ikke_> sharding with sqlite :P
2017-12-14 16:24:01 <danieli> more like sharting
2017-12-14 16:24:05 <clandmeter> in the beginning it only had edge, but now it has 6 branches...
2017-12-14 22:06:20 <BitL0G1c> unbound initd being broken by the requested changes https://github.com/alpinelinux/aports/pull/1392#discussion_r157043800
2017-12-14 22:26:23 <AWilcox[m]> consus: hi.  if you want to trace through library function calls instead of syscalls, you want ltrace (it should be in alpine main or maybe community).
2017-12-14 22:27:05 <AWilcox[m]> consus: it may require -dbg package to work fully, and I know Alpine don't build -dbg for all packages (because it is VERY large), but it may work without -dbg
2017-12-14 22:27:58 <AWilcox[m]> I wish crich was still here, xen kernel bugs are something I know a lot about debugging... might have been able to get somewhere.
2017-12-14 22:29:09 <AWilcox[m]> fwiw, it sounds like hardware weirdness, because my alpine 3.6 dom0:
2017-12-14 22:29:12 <AWilcox[m]> rachael:~$ uptime
2017-12-14 22:29:13 <AWilcox[m]>  16:23:31 up 452 days, 21:29,  1 user,  load average: 0.02, 0.02, 0.05
2017-12-14 22:29:36 <AWilcox[m]> and it has heavy I/O load, two busy postgres masters :)
2017-12-14 22:29:58 <AWilcox[m]> but then, all my Xen domUs are PV... no HVM in sight
2017-12-14 22:30:03 <AWilcox[m]> so maybe it is something in the HVM
2017-12-15 00:29:14 <ollieparanoid[m]> clandmeter: random idea for pkgs.a.o: when nothing was found, and the search query did not have wildcards, offer to search for *query* instead
2017-12-15 00:30:47 <ollieparanoid[m]> xentec: kaniini told me that you were working on the armv7 Alpine port. How is it going and is there anything you could use help with? it seems like QML does not work on armv6, which is why I'm asking.
2017-12-15 01:28:32 <hs3dUBwdmCjy> my ssd is very slow, can this be related to a recent kernel upgrade? hdparm -t /dev/sda -> Timing buffered disk reads:  46 MB in  3.04 seconds =  15.13 MB/sec
2017-12-15 01:29:04 <hs3dUBwdmCjy> smartctl doesn't report anything strange
2017-12-15 01:29:51 <hs3dUBwdmCjy> (recent kernel upgrade a few weeks ago, but from a kernel that was last updated i guess sometimes during the summer)
2017-12-15 01:31:31 <hs3dUBwdmCjy> Nov  7 4.9.60-0-hardened and before that i guess it was 4.9.19-0-grsec, according to timestamps in /lib/modules
2017-12-15 04:12:44 <mcrute> is there a policy for new packages that replace old ones? I didn't see anything in the wiki
2017-12-15 04:14:11 <mcrute> for example, should I remove the old package that it replaces?
2017-12-15 05:38:26 <_ikke_> mcrute: You can specify that the package 'replaces' another package
2017-12-15 09:15:39 <ncopa[m]> !nick ncopa
2017-12-15 09:15:44 <consus> AWilcox[m]: forgot about ltrace, thanks
2017-12-15 09:17:23 <danieli> ncopa[m]: try / instead :)
2017-12-15 09:18:00 <danieli> using lounge?
2017-12-15 09:18:12 <danieli> nope, matrix
2017-12-15 09:18:35 <danieli> also, what's up with that alpine IRC vhost? who manages it?
2017-12-15 09:18:51 <_ikke_> danieli: You have to ask freenode staff
2017-12-15 09:18:52 <ncopa> IRC vhost?
2017-12-15 09:19:04 <_ikke_> ncopa: alpine/ncopa is what your host looks like
2017-12-15 09:19:06 <danieli> name@alpine/name
2017-12-15 09:19:08 <danieli> yp
2017-12-15 09:19:09 <danieli> yep
2017-12-15 09:19:12 <clandmeter> its managed by us
2017-12-15 09:19:34 <danieli> identifying with nickserv actually makes you quit, and rejoining as the cloak
2017-12-15 09:19:41 <consus> # strace ltrace ls 2>&1 | grep EIO
2017-12-15 09:19:41 <consus> ptrace(PTRACE_POKETEXT, 14490, 0xbdc7359b1dd, 0x8d48e78948ed31cc) = -1 EIO (I/O error)
2017-12-15 09:19:44 <consus> :(
2017-12-15 09:19:45 <danieli> it's just to show affiliation / hide your hostmask
2017-12-15 09:20:06 <_ikke_> If you use sasl, you should not even have that issue
2017-12-15 09:20:23 <danieli> i'm just saying
2017-12-15 09:20:26 <danieli> that's besides the point anyway, lol
2017-12-15 09:20:30 <clandmeter> or matrix :)
2017-12-15 09:23:46 <ncopa> yeah, i switched to matrix and riot.im
2017-12-15 09:23:52 <danieli> i think i dropped my old vhosts in favor of @unaffiliated/
2017-12-15 09:23:57 <ncopa> trying to figure how how this thing works :)
2017-12-15 09:23:59 <danieli> what makes you eligible for an alpine cloak?
2017-12-15 09:25:34 <consus> Hmm
2017-12-15 09:25:36 <consus> Guys
2017-12-15 09:25:41 <consus> /sys/kernel/debug is missing
2017-12-15 09:25:47 <consus> Alpine disables it?
2017-12-15 09:26:09 <Xe> consus: in -hardened?
2017-12-15 09:26:12 <consus> Yep
2017-12-15 09:27:42 <clandmeter> danieli: we need to request it from freenode
2017-12-15 09:27:45 <danieli> aha
2017-12-15 09:27:49 <clandmeter> i think me and ncopa have permission, but we never really announced it as a service.
2017-12-15 09:28:05 <danieli> i haven't touched it in some time, so i've forgotten how it works on freenode
2017-12-15 09:28:40 <clandmeter> the point is that we are a group to claim our channels.
2017-12-15 09:28:52 <clandmeter> the cloaking was just extra.
2017-12-15 09:29:28 <danieli> yeah, i know how that part works, if it was unofficial it'd have to be the chan owner or ##alpine-
2017-12-15 11:03:50 <crich> Hello
2017-12-15 11:04:36 <crich> are there any plans to add "sheepdog" and "corosync" or "zookeper" to alpine? It would be nice to use alpine as cluster node, as its so small and robust.
2017-12-15 11:05:27 <danieli> check b.a.o
2017-12-15 11:05:30 <danieli> ^
2017-12-15 11:15:22 <crich> thx, just adde #8305
2017-12-15 11:16:45 <danieli> i'm gonna try making those 3 packages real quick
2017-12-15 11:17:02 <danieli> if i need to patch for musl and fuck around a lot, i'm just gonna drop it
2017-12-15 11:18:13 <danieli> cause i'd like zookeeper too
2017-12-15 17:14:55 <mcrute> _ikke_: yup, I was more curious about the expectation to remove the replaced package... that makes sense but I wasn't sure if there was some policy or something requiring that old packages be kept around for a transition period or something like that
2017-12-15 17:15:58 <mcrute> the response to my pull request was to delete it so I guess that's my answer
2017-12-15 21:36:32 <BitL0G1c> do github PR's build for ARM ?
2017-12-15 23:03:17 <AWilcox[m]> No.  They only build on x86_64
2017-12-16 13:25:22 <z3ntu> Found a typo on the website 😉 "Flagging stable packages is not allowd."
2017-12-16 13:25:27 <z3ntu> eg on https://pkgs.alpinelinux.org/flag/v3.7/main/alsa-lib/1.1.4.1-r2
2017-12-16 13:32:26 <danieli> z3ntu: the "e"?
2017-12-16 13:32:29 <danieli> or the lack of it
2017-12-16 13:32:43 <z3ntu> yeah `allowed`
2017-12-16 13:32:59 <danieli> clandmeter: I submitted a PR
2017-12-16 13:48:50 <clandmeter> Thx
2017-12-16 14:44:15 <consus> Can someone advice me a good VPS with Alpine linux?
2017-12-16 14:45:06 <danieli> consus: no good ones i can think of that provide alpine as an option by default
2017-12-16 14:45:11 <danieli> you can however install alpine over any other OS
2017-12-16 14:45:16 <eleksir> digital ocean? ovh?
2017-12-16 14:45:26 <consus> With unlimited 100mbps network
2017-12-16 14:45:33 <consus> Forgot this contraint :)
2017-12-16 14:45:42 <consus> Around ~50$ per month tops
2017-12-16 14:45:43 <eleksir> firstvds.ru
2017-12-16 14:45:54 <consus> > .ru
2017-12-16 14:45:58 <consus> Nah
2017-12-16 14:46:11 <danieli> god dammit, just pick one and install alpine over the OS provided
2017-12-16 14:46:18 <danieli> also, preferably take it to #alpine-linux :)
2017-12-16 14:46:22 <consus> Ooops
2017-12-16 14:46:27 <consus> Really
2017-12-16 14:46:29 <consus> Wrong channel
2017-12-16 14:46:31 <consus> Sorry
2017-12-16 14:47:54 <eleksir> www.veeble.org
2017-12-16 14:51:41 <eleksir> they offer 1 core, 512M, 10G and unmetered 100Mib for 11 bucks per month
2017-12-16 14:57:18 <scadu>  > Can someone advice me a good VPS with Alpine linux?
2017-12-16 14:57:19 <scadu> consus: I'd recommend vultr
2017-12-16 15:02:21 <BitL0G1c> consus - Veesp have unlimited 100mbit
2017-12-16 15:24:44 <BitL0G1c> is anyone using py-gamin / py-inotify with fail2ban ? only file polling seems to work
2017-12-16 17:32:07 <Lochnair> fabled: what mips isa's should we be targetting?
2017-12-16 18:22:19 <fabled> Lochnair, not sure, i don't have much experience with mips
2017-12-16 20:02:19 <Lochnair> all mips-based routers with listed isa on the openwrt toh supposedly supports at least mips32r1
2017-12-16 20:03:02 <Lochnair> and info on wikidevi supports that too
2017-12-16 20:03:59 <Lochnair> not sure what other platforms mips are used on that may need older isas
2017-12-16 21:53:49 <AWilcox[m]> Oh, oh, pick me!  I've been doing MIPS stuff for years, and my s/o even worked with MIPS people directly in school
2017-12-16 21:54:59 <AWilcox[m]> The MIPS IV ISA was used in the high-end SGI equipment (Octane, O2).  The MIPS III ISA was used in all other SGI equipment, most NT/Unix workstations including late-model DECstations, and the Nintendo 64.
2017-12-16 21:55:38 <AWilcox[m]> The MIPS III was also the first true 64-bit ISA.  So, it is typically the baseline for mips64 ports in Linux.  No MIPS II processor was commercially successful (the r6k was used in some limited aeronautical platforms).
2017-12-16 21:56:57 <AWilcox[m]> The MIPS I ISA was used in the SGI IRIS, older DECstations, and the PlayStation/PS2 (the Emotion Engine is a NEC r3k)
2017-12-16 21:58:45 <AWilcox[m]> Though the PS2 only used the r3k for compat with PS1 games.  The real PS2 chip was a r5k based on MIPS IV
2017-12-16 22:00:39 <AWilcox[m]> There weren't a lot of useful 32-bit MIPS computers pre-MIPS32 outside of the IRIS and PlayStation 1.  So I would tend to personally think: mips64 has a MIPS III baseline, and mips32 has a MIPS32 baseline.
2017-12-16 22:16:42 <AWilcox[m]> Hope that was helpful :)
2017-12-16 22:18:15 <TBK> I feel enlightened :D
2017-12-16 22:20:44 <Lochnair> AWilcox[m]: Hehe, nice have someone familiar with MIPS chime in on this, thanks :)
2017-12-16 22:25:36 <Lochnair> Right, that's my opinion on mips32 as well
2017-12-16 22:26:09 <Lochnair> I haven't really found much info on mips64 however, the only platform I know of using MIPS64 is Octeon
2017-12-16 22:29:30 <Lochnair> Interestingly, Debian uses MIPS32R2 and MIPS64R2 from Stretch onwards
2017-12-16 23:01:15 <Lochnair> AWilcox[m]: btw, have you gotten any further with MIPS support in Adélie?
2017-12-16 23:09:07 <AWilcox[m]> Nope.  Still stuck on the same error from the ML.
2017-12-16 23:18:59 <Lochnair> The asm/unistd.h include?
2017-12-16 23:19:33 <Lochnair> We fixed that in Alpine using a patch from upstream: https://git.alpinelinux.org/cgit/aports/commit/?id=2baf4807188617d1957dcbcadc1d60e6a6465af2
2017-12-16 23:34:46 <danieli> AWilcox[m]: as I said, I want to create an alpine mips thing
2017-12-17 06:40:52 <AWilcox[m]> I did not know this was fixed!
2017-12-17 15:03:19 <Shiz>      ed : some packages moved from main to community, and got stuck on an old version of consolekit since it was renamed to consolekit2
2017-12-17 15:03:21 <Shiz>      ed : couldn't uninstall it since it was still listed as a dependency of lxdm somehow, kept getting installed and uninstalled for every unrelated apk command i did
2017-12-17 15:03:26 <Shiz> 3.6 -> 3.7 upgrade
2017-12-18 09:13:36 <kukin> hello, i am trying to compile  https://github.com/LibreOffice/online/blob/master/common/Seccomp.cpp in alpine. I did some patches for musl, but i do not know what to do with this: common/Seccomp.cpp: In function 'bool Seccomp::lockdown(Seccomp::Type)': common/Seccomp.cpp:208:42: error: '__sighandler_t' does not name a type      action.sa_handler = reinterpret_cast<__sighandler_t>(handleSysSignal);
2017-12-18 09:29:29 <PureTryOut[m]> Shiz: consolekit wasn't just renamed to consolekit2, it switched to an entirely different source. consolekit is unmaintained, consolekit2 isn't
2017-12-18 13:50:34 <kukin> Whats is wrong, please? https://pastebin.com/mPbvA8gE i have poco-dev installed from repos. Thanks
2017-12-18 13:52:59 <danieli> kukin: do you have pcre-dev installed?
2017-12-18 13:54:25 <kukin> danieli: there is no package called pcre-dev https://pkgs.alpinelinux.org/contents?file=pcre-dev&path=&name=&branch=&repo=&arch=
2017-12-18 13:55:21 <kukin> danieli: oops, sry. sure its available in repos
2017-12-18 13:55:22 <danieli> https://pkgs.alpinelinux.org/package/edge/main/x86_64/pcre-dev
2017-12-18 13:56:40 <danieli> god damn the wildcard stuff on pkgs.a.o is annoying
2017-12-18 13:59:25 <kukin> danieli: pcre is now ok, still i can not get working PocoNetSSL and PocoCrypto
2017-12-18 13:59:30 <kukin> thanks
2017-12-18 13:59:50 <danieli> all right
2017-12-18 13:59:51 <danieli> idk lol
2017-12-18 14:01:50 <kukin> danieli: https://git.alpinelinux.org/cgit/aports/tree/testing/poco/APKBUILD#n53 what about this? Dont know why is not recognized by libtoolize
2017-12-18 14:05:08 <danieli> ah I see
2017-12-18 14:05:13 <danieli> still got no idea, I don't really use poco
2017-12-18 14:30:21 <Shiz> PureTryOut[m]: why are you telling me this? :P
2017-12-18 14:32:15 <PureTryOut[m]> well you said it was just renamed, which is not true. more happened 😛
2017-12-18 15:20:02 <Shiz> i don't think i said that?
2017-12-18 15:20:28 <Shiz> oh right i did
2017-12-18 15:20:36 <Shiz> that was a quote from another person
2017-12-18 15:20:38 <Shiz> not my own wording
2017-12-18 15:20:40 <Shiz> :P
2017-12-18 15:20:54 <Shiz> but he meant the package got transitioned
2017-12-19 11:10:08 <kukin> Any help with SigUtil.cpp:209: undefined reference to `backtrace_symbols'  . Log:  https://pastebin.com/J3gLy5qP ?  https://github.com/LibreOffice/online/blob/master/common/SigUtil.cpp Thanks
2017-12-19 11:10:55 <ncopa> kukin: you can either patch out that (make a configure check for HAVE_BACKTRACE_SYMBOLS) and ifdef it
2017-12-19 11:11:24 <ncopa> or you can add libexecinfo-dev to make depends and add `-lexecinfo` to linker flags
2017-12-19 11:12:01 <ncopa> i think libexecinfo is broken, but at least it builds
2017-12-19 11:18:14 <kukin> ncopa whole dumpBacktrace() block?
2017-12-19 11:18:45 <kukin> #ifdef HAVE_BACKTRACE_SYMBOLS  void dumpBacktrace() ...  #endif ,right?
2017-12-19 11:32:12 <kukin> @ncopa
2017-12-19 14:03:24 <ncopa> kukin: i have not looked at the sources
2017-12-19 14:03:43 <ncopa> but generally, you can simply skip printing the backtraces
2017-12-19 14:04:02 <ncopa> or even let it call abort() so you get a core dump
2017-12-19 14:14:54 <kukin> @ncopa  i used indef and its ok for now
2017-12-19 14:17:02 <kukin> ivgot a problem with ... log: https://pastebin.com/1TTtTAq5 sources: https://github.com/LibreOffice/online/blob/master/kit/KitHelper.hpp but seems that this is for libreoffice developers, right?
2017-12-19 19:12:20 <danieli> ncopa: I'm not sure how you guys do package renaming, but could you 1) delete community/httpie, 2) replace it with testing/httpie, 3) update the version to 1.9.9 as it should be and 4) make me contributor and him maintainer?
2017-12-19 19:13:01 <danieli> I'd appreciate it if any of you with commit access and the knowledge could do that
2017-12-19 19:13:29 <danieli> also - for the record - the reason I submitted testing/httpie with community/httpie existing was that the auto complete broke in pkgs.a.o, so I didn't know it existed
2017-12-19 19:14:03 <ncopa> :D
2017-12-19 19:14:19 <danieli> I had a derp :P
2017-12-19 19:14:43 <danieli> and right now I feel extremely masochistic. I pained myself through 2 days of a broken RAM chip, broken win7 ISOs, shitty support ++ to install windows 7
2017-12-19 19:14:50 <danieli> and now the update to 10 is just hell
2017-12-19 19:14:57 <ncopa> looks like it faffolter who is the current maintainer
2017-12-19 19:15:24 <danieli> correct
2017-12-19 19:15:37 <danieli> his apkbuild doesn't have a SPDX format license or check()
2017-12-19 19:16:09 <danieli> so I want to get rid of testing/httpie and merge it with community/httpie
2017-12-19 19:16:25 <danieli> my APKBUILD is more modern with a 'better' license="" so
2017-12-19 19:17:18 <danieli> man.. I installed so much crap - games, tools, programs - just to do a clean install.. FML.
2017-12-19 19:18:14 <ncopa> this is the diff: http://tpaste.us/Ozem
2017-12-19 19:19:09 <danieli> ncopa: copy the APKBUILD from testing/httpie and make me contributor, him maintainer
2017-12-19 19:19:11 <danieli> that's the tl;dr
2017-12-19 19:19:25 <danieli> probs different sums because it was repackaged - one from github, one from pypi
2017-12-19 19:27:21 <ncopa> would this be ok? http://tpaste.us/JM7v
2017-12-19 19:28:54 <danieli> 1 sec
2017-12-19 19:29:25 <danieli> ncopa: would it be preferable to use github or pypi?
2017-12-19 19:30:42 <danieli> license="BSD-3-Clause" is just about it, the rest looks fine if alpine prefers pypi over github releases
2017-12-19 19:30:51 <ncopa> in this case i'd say pypi because github does not seem to have the 0.9.9 release
2017-12-19 19:33:36 <danieli> ah, I see, that makes sense
2017-12-19 19:33:55 <danieli> apparently he moved repos, that's why
2017-12-19 19:34:12 <danieli> nevermind - no 0.9.9 there either
2017-12-19 19:34:23 <danieli> ncopa: change license to the SPDX format and we're good
2017-12-19 19:36:07 <ncopa> done
2017-12-19 19:36:09 <ncopa> thanks
2017-12-19 20:21:48 <danieli> ditto
2017-12-19 20:21:49 <danieli> brb
2017-12-20 00:29:41 <duncan^> I've been playing with running apk as a non-root user, using the --root option to specify a target directory.
2017-12-20 00:30:09 <duncan^> Some observations: There are many errors that some operations fail, due to a permissions issue.
2017-12-20 00:30:16 <duncan^> This doesn't seem terminal.
2017-12-20 00:31:04 <duncan^> However, the one thing that is interesting is that the busybox package is entirely broken. It just doesn't create the applets, which are symbolic links.
2017-12-20 00:33:30 <duncan^> I don't actually understand why this is.
2017-12-20 00:34:12 <duncan^> The --install option for busybox doesn't seem to need elevated perms
2017-12-20 00:58:05 <AWilcox[m]> That's probably related to not having a /dev/null.  And that is what --root is trying to create that it can't (mknod is always a privileged operation)
2017-12-20 01:06:29 <Shiz> that, and it also creates suid binaries
2017-12-20 01:06:35 <Shiz> which i think is privileged?
2017-12-20 02:37:27 <AWilcox[m]> Only busybox-suid
2017-12-20 10:46:22 <ncopa> duncan^: can you run it in strace?
2017-12-20 19:25:31 <Lochnair> any toolchain experts in here than can explain why the final build stage of cross gcc fails for mips64 (and not mips32)?
2017-12-20 19:25:35 <Lochnair> i'm getting: /usr/mips64-alpine-linux-musl/bin/ld: cannot find crti.o: No such file or directory
2017-12-20 19:25:47 <Lochnair> complete bootstrap log: https://dl.lochnair.net/mips64-bootstrap.log
2017-12-20 21:03:08 <schnozzle123> hi there: simple question: Is mongodb binary removed from the packages for arm64v8?
2017-12-20 21:03:46 <schnozzle123> i refer to edge
2017-12-20 21:07:32 <Lochnair> schnozzle123: mongodb is only available on x86_64 it seems
2017-12-20 21:08:44 <schnozzle123> hm wired because some days ago i had a docker image based on alpine:edge (64 bit arm) running where i was able to install mongodb
2017-12-20 21:09:19 <schnozzle123> Lochnair, you are correct about the listing on the page (alpine packages)
2017-12-20 21:12:42 <Lochnair> hm. according to the aports git log, there was never support for anything other than x86_64
2017-12-20 21:22:15 <schnozzle123> really strange had it running some minutes ago: MongoDB starting : pid=1 port=27017 dbpath=/data/db 64-bit host=051861e69d44
2017-12-20 21:23:34 <schnozzle123>    distarch: arm64  target_arch: arm64
2017-12-20 21:23:56 <schnozzle123> from the log running on my rpi
2017-12-20 21:59:49 <Lochnair> ok, so I've figured out why cross gcc fails the final stage
2017-12-20 21:59:55 <Lochnair> musl-dev is installed in sysroot-mips64/usr/lib, gcc however looks for it in sysroot-mips64/usr/lib64
2017-12-21 01:16:54 <AWilcox[m]> Lochnair: since musl (and therefore Alpine) doesn't support multilib there is no reason to have "lib64", so you will need to find where that is being put in to GCC and replace it with lib
2017-12-21 01:17:32 <Lochnair> AWilcox[m]: yeah I figured as much
2017-12-21 01:17:47 <Lochnair> I assume I have to do something similar to the gcc-pure64.patch
2017-12-21 01:18:00 <AWilcox[m]> Yeah
2017-12-21 01:27:57 <Lochnair> I've tried doing this: https://pastebin.com/XhUHruvU
2017-12-21 01:28:09 <Lochnair> But I'm still seeing the same thing
2017-12-21 02:55:54 <Lochnair> Gotten a bit further now: https://dpaste.de/ovFq
2017-12-21 02:56:29 <Lochnair> Everything seems to compile correctly, and it gets to the packaging stage
2017-12-21 02:58:16 <Lochnair> But it fails when running the split function for libgnat, as apparentely libgna*.so is missing
2017-12-21 09:32:17 <schnozzle123> wow thanks locknair for your effort!!
2017-12-21 10:50:19 <ncopa> schnozzle123: looks like mongodb only exist for x86_64
2017-12-21 10:52:06 <schnozzle123> hm i investigated further and i found the docker image which i created some weeks ago. And it turns out that it was based on debian.
2017-12-21 10:53:14 <schnozzle123> So my statement yesterday is wrong, mongodb never ran on arm64v8/alpine. Sorry for that!
2017-12-21 11:09:29 <schnozzle123> By the way i really like alpine. Maybe i should invest some time to get more used to aports, to contribute.
2017-12-21 11:09:34 <schnozzle123> thank you for your help!
2017-12-21 14:23:04 <Lochnair> for some reason mips64 isn't supported in gnat (which is why libgna*.so was missing), while mips64el is. I *think* this patch fixes that: https://dpaste.de/rSCU
2017-12-21 14:23:17 <Lochnair> now however I'm seeing some weird compilation error for ada on both
2017-12-21 14:24:14 <Lochnair> guess I have to do a run with -j1, so the logs actually readable
2017-12-21 20:58:20 <ncopa> im giving up with this vlc issue
2017-12-21 20:58:28 <ncopa> i will just disable the test for now
2017-12-21 21:38:07 <kaniini> Natanael Copa: would you be opposed to adding the ability to use a licensed grsecurity patch with the linux-hardened APKBUILD?
2017-12-21 21:44:32 <Lochnair> AWilcox[m]: I've fixed the rest of the issues I've found and sent the patches to the ML. My EdgeRouter Lite is now running Alpine MIPS64 :)
2017-12-21 22:13:51 <kaniini> Lochnair: if you can provide instructions, i will set up an edgerouter as buildserver.  i upgraded it to 16gb
2017-12-21 23:01:52 <Lochnair> kaniini: Sweet!
2017-12-21 23:01:54 <Lochnair> I'd like to get a linux-vanilla image working first, before we get to that however
2017-12-21 23:02:08 <Lochnair> Currently I'm running a manually built kernel on it
2017-12-21 23:02:24 <Lochnair> Not sure how well the generic MIPS support in the kernel works
2017-12-21 23:15:15 <fugitive> hello, sorry if i am asking in a wrong place, but i have bad time setting custom alpine on rpi3. So what I did, created initrd, kernel booted, mounted external hdd, download & extract mini root fs, then set up the system in the chroot. After reboot, everything goes well, i can load my modules, apk works, but i can not log in over ssh as non root user.   I can't even su to the "user" from root account, getting "Could not chdir to home directory /
2017-12-21 23:15:15 <fugitive> home/alpine: Permission denied"
2017-12-21 23:15:43 <Xe> what are the permissions on /home/alpine?
2017-12-21 23:15:47 <Xe> ls -l /home/alpine
2017-12-21 23:16:14 <fugitive> drwxrwxrwx    2 alpine   alpine
2017-12-21 23:16:27 <fugitive> i gave 777 intentionaly
2017-12-21 23:17:52 <fugitive> i followed chroot setup closely and made sure that i didn't missed anything.
2017-12-21 23:19:15 <fugitive> here is a strace output from the ssh process.. http://termbin.com/d7xg
2017-12-21 23:20:04 <Lochnair> Hmm OK. So in the case of Octeon, it looks like the octeon-ethernet driver depends on the system type being set to Octeon
2017-12-21 23:54:26 <kaniini> Lochnair: we might need special linux kernel packages for mips, i don't think theres anything like DTBs
2017-12-22 00:39:52 <fugitive> something is wrong with mini root fs.. i gave up on that, so i used apk.static to bootstrap mounted ext drive, and vouela :) damn.. why i  didn't try that earlier...
2017-12-22 00:54:25 <Arcaire> any considerations i should make before i publish a new Alpine AMI?
2017-12-22 00:54:32 <Arcaire> the old one is totally broken and AWS has no Alpine images out
2017-12-22 00:54:46 <Arcaire> it's pretty much done tbh, just wondering if there's anything people would suggest baking in properly
2017-12-22 07:11:09 <rnalrd> Lochnair cool! re: AL on Octeon
2017-12-22 07:11:21 <rnalrd> do you have also hw acceleration enabled?
2017-12-22 09:09:40 <Lochnair> rnalrd: Nope. You'd have to stay on the 3.10 kernel that ships with EdgeOS to keep the offloads
2017-12-22 12:12:21 <vkrishn> clandmeter2: hi
2017-12-22 12:12:40 <vkrishn> is https://api.alpinelinux.org still functional or scrapped ?
2017-12-22 12:16:21 <vkrishn> seems a developer was trying to use it, https://github.com/insteps/aport-api/issues/1
2017-12-22 12:19:58 <vkrishn> just to add it uses Php-Phalcon which is no longer in main
2017-12-22 13:06:03 <vkrishn> gtg, would check logs/email for reply. Thinking to add 'deprecated' to github pkg
2017-12-22 16:22:04 <fcolista> there's a way to say alpine to not load the apkovl via syslinux commands?
2017-12-22 16:22:21 <fcolista> s/commands/options
2017-12-22 16:23:17 <ncopa> fcolista: you can at boot prompt add apkovl=no
2017-12-22 16:23:38 <ncopa> the init script will try to use apkovl=whatever.apkovl.tar.gz
2017-12-22 16:23:46 <ncopa> but it will check if it exists
2017-12-22 16:24:19 <fcolista> good so i can choose also the apkovl i want to use
2017-12-22 16:24:34 <ncopa> so you can disable apkovl loading with apkovl=nonexisting
2017-12-22 16:24:45 <fcolista> i believe it goes before the kernel parameters
2017-12-22 16:25:10 <ncopa> you add it as a kernel param
2017-12-22 16:25:33 <fcolista> ok
2017-12-22 16:25:34 <fcolista> cool
2017-12-22 16:26:11 <fcolista> thx ncopa!
2017-12-22 16:26:23 <ncopa> np
2017-12-23 08:29:20 <olaf_> Hi, I like to complain about https://pkgs.alpinelinux.org/package/v3.7/community/x86_64/bareos; there are no scripts bundled, e.g. https://pkgs.alpinelinux.org/contents?file=create*&path=&name=bareos&branch=v3.7&repo=community&arch=x86_64 . Also there is no build log http://build.alpinelinux.org/buildlogs/build-3-7-x86_64/community/bareos/bareos-16.2.6-r1.log . Did something wrong on build I miss I something?
2017-12-23 08:29:58 <olaf_> I mean, do I miss something
2017-12-23 08:46:38 <olaf_> https://bugs.alpinelinux.org/issues/8347
2017-12-23 08:53:04 <scadu> olaf_: https://github.com/alpinelinux/aports/blob/master/community/bareos/APKBUILD#L99
2017-12-23 08:54:17 <scadu> Not sure who introduced this
2017-12-23 08:56:44 <scadu> https://github.com/alpinelinux/aports/commit/c6464a23fcc6deb83d8e120b026c26ae49c2dffb
2017-12-23 08:56:45 <scadu> Here we go
2017-12-23 08:57:07 <scadu> But probably vkartel is not around
2017-12-23 08:58:26 <scadu> vakartel*
2017-12-23 09:16:50 <olaf_> thanks, I assume the package archive isn't up-to-date as https://pkgs.alpinelinux.org/packages?name=bareos*&branch=v3.7&repo=&arch=x86_64&maintainer= suggested?
2017-12-23 09:20:17 <olaf_> ah, older changes ..
2017-12-23 12:01:14 <olaf_> does somwone has a working docker stack for bareos with webui based on alpine?
2017-12-23 19:10:27 <tmh1999> I found a strange bug. If you run $ whois alpinelinux.org (or any .org domain) a couple of time, it will start segfaul and then hang. edge, hardened.
2017-12-23 19:10:50 <tmh1999> apparently .com domains are not affected
2017-12-23 19:11:38 <tmh1999> .net not affected
2017-12-23 19:11:42 <tmh1999> .info is affected
2017-12-24 04:54:32 <u1> You use textmode in code, just because textmode is well documented, simple, standardized, and available. Not because it is really good technology. Would be versatile video modes available as well as text mode, - you would definitely use em instead!!! JUST ADMIT IT!!!
2017-12-24 04:54:46 <u1> ??
2017-12-24 05:42:30 <AWilcox[m]> I have no idea what that means, but I write the majority of my code in Kate, which is a Qt text editor, on a 1920x1080 HD screen
2017-12-24 05:44:32 <skarnet> 80x25 console or gtfo
2017-12-24 05:44:56 <skarnet> hell, I only code on DOS
2017-12-24 22:39:11 <kaniini> skarnet: turbo C?
2017-12-24 22:39:46 <skarnet> oh my, Borland was good
2017-12-25 01:23:49 <u1> I dont understand
2017-12-25 01:25:46 <skarnet> of course you don't
2017-12-25 01:28:56 <u1> but why7
2017-12-25 01:30:00 <u1> and howto7
2017-12-25 02:13:03 <u1> I dont understand
2017-12-25 02:15:10 <skarnet> You already said that.
2017-12-25 02:15:57 <u1> If I want to create my own filesystem (may be very simple) how to make it work in linux?
2017-12-25 02:15:57 <u1> how to make it work in linux?
2017-12-25 02:17:12 <u1> When I open file I do syscall with arg string?
2017-12-25 02:20:32 <u1> What if I identify file somehow else?
2017-12-25 02:21:55 <skarnet> don't you have anything better to do with your life than troll IRC channels on Christmas night?
2017-12-25 02:23:04 <u1> i cant sleep
2017-12-25 04:28:34 <trouduc> greetings to all
2017-12-25 04:28:40 <trouduc> can someone help me
2017-12-25 11:02:03 <dommae> hi guys
2017-12-25 11:02:12 <dommae> I have a problem with a docker container with alpine
2017-12-25 11:02:24 <dommae> the entrypoint file is in/usr/bin/entrypoint
2017-12-25 11:03:01 <dommae> but the container keep triggereing standard_init_linux.go:178: exec user process caused "no such file or directory"
2017-12-25 11:14:35 <Shiz> probably because your entrypoint is a glibc binary
2017-12-25 11:14:44 <Shiz> did you compile the entrypoint under alpine?
2017-12-25 11:44:20 <dommae> no it is a bash script
2017-12-25 11:44:25 <dommae> the problem was
2017-12-25 11:44:37 <dommae> i had to dos2unix all script on alpine
2017-12-25 11:44:47 <dommae> fucking alpine
2017-12-25 11:44:54 <dommae> it made me loose 2 days
2017-12-25 11:45:48 <ncopa> hi dommae
2017-12-25 11:45:58 <ncopa> you had to dos2unix all bash scripts?
2017-12-25 11:46:14 <ncopa> or was it #!/bin/sh scripts?
2017-12-25 11:46:21 <ncopa> or other kind of scripts?
2017-12-25 11:47:33 <dommae> here
2017-12-25 11:47:33 <dommae> #!/bin/sh  ## Change GID for USER? if [ -n "${USER_GID}" ] && [ "${USER_GID}" != "`id -g ${USER}`" ]; then     sed -i -e "s/^${USER}:\([^:]*\):[0-9]*/${USER}:\1:${USER_GID}/" /etc/group     sed -i -e "s/^${USER}:\([^:]*\):\([0-9]*\):[0-9]*/${USER}:\1:\2:${USER_GID}/" /etc/passwd fi  ## Change UID for USER? if [ -n "${USER_UID}" ] && [ "${USER_UID}" != "`id -u ${USER}`" ]; then     sed -i -e "s/^${USER}:\([^:]*\):[0-9]*:\([0-9]*\)
2017-12-25 11:47:54 <dommae> entrypoint
2017-12-25 11:48:01 <dommae> and other are .sh scripts
2017-12-25 12:03:54 <ncopa> so its busybox ash that does not like dos newlines
2017-12-25 12:05:48 <_ikke_> normal bash neither
2017-12-25 12:06:24 <_ikke_> Even a php script with a shebang line would not execute do to \rs
2017-12-25 12:06:25 <ncopa> and dash?
2017-12-25 12:06:37 <_ikke_> no idea
2017-12-25 12:06:44 <ncopa> i doubt
2017-12-25 12:06:58 <ncopa> i just wonder how it could be alpines fault
2017-12-25 12:07:33 <_ikke_> It's not Alpines fault
2017-12-25 12:12:07 <_ikke_> printf '#!/usr/bin/env sh\r\n echo test' >test.sh; chmod +x test.sh; ./test.sh -> ": No such file or directory"
2017-12-25 12:12:19 <_ikke_> (That's on Ubuntu)
2017-12-25 12:27:14 <skarnet> you can't reason with haters
2017-12-25 20:00:17 <_ikke_> Is it common that it's more quiet after a release?
2017-12-25 20:45:06 <skarnet> it's probably common that it's more quiet on Christmas day
2017-12-25 20:51:14 <shewo16> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! iidpx: mixi blahdodo zaolin mjeanson kunkku shreyansh_k Gottox vitronic Bureaucat jakedt kaniini jailbox LongyanG blueness arch3y xsteadfastx BitL0G1c Adran tg MatrixTraveler[m victorbjelkholm nmeum nlf f
2017-12-25 20:51:17 <shewo16> ▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! gdkvpata: fcolista_web mjeanson KillTheIdols1 rnalrd Arcaire craftyguy xentec kaniini jailbox tg ephemer0l_ Adran kunkku NightKhaos s33se marble_visions tyng pickfire MatrixTraveler[m Bureaucat cyteen_ fcolista Alchemical blahdodo zaolin BitL0G
2017-12-25 20:51:23 <shewo16> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! hbfdrg: Bureaucat tyng Gottox shreyansh_k xentec DLange mixi pardis kunkku Somasis Adran xsteadfastx fcolista victorbjelkholm clandmeter2 chris| jailbox Arcaire minimalism ephemer0l_ pickfire arch3y bluen
2017-12-25 20:51:28 <shewo16> ▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! chbpqbta: LongyanG BitL0G1c minimalism Shiz xentec craftyguy marble_visions victorbjelkholm zaolin ryanlelek xsteadfastx skarnet tg shreyansh_k vitronic mjeanson lep2 ephemer0l_ chris| rnalrd jailbox jakedt blueness Somasis DLange Kil
2017-12-25 20:51:32 <shewo16> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! eahgw: Somasis BitL0G1c fcolista_web cyteen_ skarnet Adran ryanlelek Gottox blahdodo kaniini clandmeter2 Shiz NightKhaos pardis lostd shreyansh_k tyng mjeanson nmeum fabled chris| blueness ephem
2017-12-25 21:01:22 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! occqjfhais: clandmeter2 NightKhaos arch3y victorbjelkholm mixi ryanlelek chris| fabled Shiz BitL0G1c kaniini rnalrd xentec mjeanson marble_visions Gottox minimalism Alchemical zaolin e
2017-12-25 21:01:27 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! fiblzoln: clandmeter2 chris| fabled Shiz nlf ephemer0l_ tg lep2 arch3y jakedt blueness vitronic NightKhaos minimalism skarnet mixi Gottox lostd jailbox fcolista Alchemical s33se LongyanG pickfire MatrixTr
2017-12-25 21:01:32 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! kjjhyssq: Shiz fabled xsteadfastx lep2 Somasis LongyanG ephemer0l_ nlf blueness fcolista Alchemical tyng victorbjelkholm KillTheIdols1 hrmlgon-z blahdodo pickfire mjeanson NightKhaos clandmeter2
2017-12-25 21:01:37 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! vkiamzqiae: xentec mixi Arcaire marble_visions kaniini fabled MatrixTraveler[m Alchemical mjeanson pardis lostd s33se minimalism zaolin pickfire jakedt nmeum blueness ryanlelek victorb
2017-12-25 21:01:42 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! ygoabf: DLange tg mixi shreyansh_k kaniini arch3y chris| Gottox ephemer0l_ blahdodo LongyanG craftyguy lostd jailbox xentec rnalrd pickfire victorbjelkholm fcolista_web lep2 Shiz tyng Alchemical clandmete
2017-12-25 21:01:47 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! odzrkkff: xentec Somasis vitronic KillTheIdols1 blueness hrmlgon-z cyteen_ tyng jailbox ryanlelek Shiz pickfire pardis xsteadfastx chris| LongyanG shreyansh_k DLange lep2 nmeum fcolista jakedt rnalrd fcolista_web Gottox fabled mixi craftyguy mj
2017-12-25 21:01:52 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! fxqehruykn: BitL0G1c Alchemical vitronic hrmlgon-z cyteen_ pardis clandmeter2 blahdodo nlf ephemer0l_ marble_visions zaolin chris| fabled xentec fcolista jailbox mj
2017-12-25 21:01:57 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! wimugmsl: fabled nlf Arcaire victorbjelkholm kaniini marble_visions KillTheIdols1 shreyansh_k chris| fcolista_web craftyguy NightKhaos blueness nmeum jailbox DLange zaolin clandmeter2 mixi tg Alchemical lep2 MatrixTraveler[m cyteen_ Gottox pardis lostd v
2017-12-25 21:02:02 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! tfottyvk: BitL0G1c tg chris| xentec kunkku Gottox Arcaire lostd clandmeter2 MatrixTraveler[m blahdodo NightKhaos jakedt tyng xsteadfastx skarnet rnalrd s33se KillTheIdols1 pickfire cyteen_ marble_visions
2017-12-25 21:02:03 <vitronic> span ??
2017-12-25 21:02:07 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! mkqytokcnm: pickfire victorbjelkholm ryanlelek DLange mjeanson tyng hrmlgon-z NightKhaos xentec Adran cyteen_ rnalrd craftyguy shreyansh_k xsteadfastx BitL0G1c marble_visions
2017-12-25 21:02:12 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! tvmqkg: tyng hrmlgon-z MatrixTraveler[m rnalrd Arcaire DLange jakedt pickfire fcolista_web chris| shreyansh_k marble_visions kaniini zaolin KillTheIdols1 Alchemical victorbjelkholm rya
2017-12-25 21:02:17 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! mtxpaeav: victorbjelkholm kunkku DLange mjeanson rnalrd Arcaire clandmeter2 blueness arch3y s33se blahdodo lep2 ephemer0l_ jailbox nlf chris| minimalism fcolista vitronic skarnet jakedt craftyguy nmeum tyng tg shreyansh_k Adran KillTh
2017-12-25 21:02:22 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! hwdkgisgm: blueness tyng craftyguy pickfire skarnet nmeum nlf tg chris| rnalrd marble_visions xentec blahdodo fcolista_web clandmeter2 DLange Shiz mixi zaolin Somasis cyteen_ fcolista vitronic Adran arch3
2017-12-25 21:02:27 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! mqkmszc: rnalrd jailbox zaolin mjeanson Adran DLange tg nmeum nlf KillTheIdols1 lostd tyng Alchemical chris| blahdodo NightKhaos xentec fcolista BitL0G1c MatrixTraveler[m LongyanG hrmlgon-z fcolista_web c
2017-12-25 21:02:32 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! vdlzqd: ephemer0l_ tg cyteen_ zaolin blueness xentec hrmlgon-z kunkku Gottox KillTheIdols1 DLange blahdodo BitL0G1c jailbox Arcaire kaniini craftyguy marble_visions lostd vitronic jakedt LongyanG s33se Shiz nlf minimalism pi
2017-12-25 21:02:37 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! ikfzxvn: DLange Gottox s33se kaniini nlf Somasis marble_visions pickfire lep2 MatrixTraveler[m craftyguy xsteadfastx kunkku Adran hrmlgon-z LongyanG rnalrd tg Alchemical tyng Arcaire fcolista BitL0G1c jailbox xentec clandmeter2 KillTheIdols1 lostd nmeum blueness c
2017-12-25 21:02:42 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! smfvgmfs: shreyansh_k ryanlelek kunkku NightKhaos jailbox ephemer0l_ fabled rnalrd victorbjelkholm lep2 vitronic pardis lostd fcolista_web MatrixTraveler[m skarnet xsteadfastx nmeum blueness blahdodo DLan
2017-12-25 21:02:47 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! egqeihixj: craftyguy nmeum chris| Gottox tyng xentec mixi LongyanG pickfire nlf cyteen_ lep2 tg Alchemical skarnet blahdodo hrmlgon-z DLange ryanlelek marble_visions Shiz victorbjelkho
2017-12-25 21:02:52 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! xeygpjrlb: tg DLange lep2 ryanlelek minimalism xentec blahdodo mjeanson tyng craftyguy nmeum kaniini ephemer0l_ s33se chris| hrmlgon-z mixi nlf jakedt arch3y Alchemical pardis Somasis KillTheIdols1 kunkku skarnet cyteen_ BitL0G1c rnalrd xsteadfastx fcolista_web bl
2017-12-25 21:02:57 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! xgqljzzk: nlf minimalism blueness skarnet craftyguy Alchemical mjeanson DLange BitL0G1c kunkku kaniini vitronic pickfire lostd shreyansh_k fcolista ephemer0l_ zaolin rnalrd victorbjelkholm arch3y MatrixTraveler[m ryanlelek chris| Shiz
2017-12-25 21:03:02 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! axflls: ryanlelek pickfire blueness kaniini fcolista Adran KillTheIdols1 jakedt craftyguy rnalrd minimalism NightKhaos victorbjelkholm arch3y hrmlgon-z nmeum fcolista_web pardis cyteen_ lep2 nlf Gottox kunkku DLange s33se blahdodo Bit
2017-12-25 21:03:07 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! iivnhwqfx: fcolista clandmeter2 jailbox minimalism KillTheIdols1 ephemer0l_ chris| BitL0G1c NightKhaos Gottox Somasis pardis marble_visions kaniini Arcaire lep2 craftyguy rnalrd pickfire xsteadfastx tyng blueness shreyansh_k nmeum nlf cyteen_ jakedt fabled DLange
2017-12-25 21:03:12 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! hzlkasondb: KillTheIdols1 rnalrd shreyansh_k lostd ephemer0l_ NightKhaos nmeum victorbjelkholm clandmeter2 nlf blueness xsteadfastx pickfire vitronic arch3y pardis mjeanson LongyanG blahdodo tg tyng Gottox Somasis BitL0G1c fcolista kaniini craf
2017-12-25 21:03:17 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! avxzrkjokx: kaniini KillTheIdols1 zaolin craftyguy Shiz DLange jakedt pardis kunkku xentec mjeanson ryanlelek mixi fcolista Adran Arcaire hrmlgon-z Gottox clandmeter2 tg LongyanG victorbjelkholm lostd marble_visions BitL0G1c
2017-12-25 21:03:22 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! afmpyhhaug: DLange Alchemical Arcaire pickfire NightKhaos victorbjelkholm fcolista MatrixTraveler[m Shiz shreyansh_k KillTheIdols1 fcolista_web ephemer0l_ zaolin s33se pardis tg marble
2017-12-25 21:03:27 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! vlfwas: marble_visions craftyguy chris| vitronic arch3y DLange BitL0G1c rnalrd MatrixTraveler[m tg s33se ryanlelek fcolista_web ephemer0l_ Adran xsteadfastx tyng mixi fcolista cyteen_ jakedt jai
2017-12-25 21:03:32 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! wshdjgq: marble_visions NightKhaos nmeum KillTheIdols1 clandmeter2 blueness Shiz DLange fcolista Adran jailbox victorbjelkholm Somasis s33se chris| arch3y fabled fcolista_web vitronic mixi tg Alchemical BitL0G1c ka
2017-12-25 21:03:37 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! amuaxfr: LongyanG Alchemical hrmlgon-z Adran DLange xentec clandmeter2 blahdodo lep2 s33se minimalism jakedt mjeanson kaniini marble_visions Somasis fabled craftyguy fcolista_web Shiz rnalrd skarnet kunkku Arcaire ephemer0l_ MatrixTraveler[m zaolin vitro
2017-12-25 21:03:42 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! ussoby: ryanlelek xentec NightKhaos jakedt lostd s33se kaniini pickfire rnalrd LongyanG lep2 Shiz fcolista hrmlgon-z Adran minimalism shreyansh_k Alchemical mixi clandmeter2 nmeum skarnet blahdodo cyteen_ fabled KillTheIdols1 DLange ephemer0l_ victorbjelkholm tyng
2017-12-25 21:03:47 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! uxdez: jailbox lostd kaniini marble_visions tg arch3y shreyansh_k KillTheIdols1 mixi rnalrd fcolista_web Somasis blahdodo craftyguy lep2 s33se pickfire xsteadfastx vitronic t
2017-12-25 21:03:52 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! ogiwfnwc: KillTheIdols1 ephemer0l_ mjeanson zaolin BitL0G1c Arcaire hrmlgon-z chris| DLange kunkku blahdodo Adran lostd arch3y xsteadfastx MatrixTraveler[m tg fcolista s33se jakedt Gottox Somasis cyteen_ fcolista_w
2017-12-25 21:03:56 <skarnet> leaving until an op can fix this
2017-12-25 21:03:57 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! lltat: vitronic Arcaire pardis xsteadfastx KillTheIdols1 mjeanson hrmlgon-z cyteen_ tyng DLange tg Adran BitL0G1c fcolista_web kunkku Alchemical fabled shreyansh_k zaolin NightKhaos LongyanG bla
2017-12-25 21:03:59 <TBB> you could at least learn to write scripts that write correct length lines
2017-12-25 21:04:02 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! jkmvep: jailbox tyng clandmeter2 cyteen_ Alchemical DLange nmeum arch3y chris| mixi MatrixTraveler[m shreyansh_k jakedt LongyanG victorbjelkholm vitronic lostd skar
2017-12-25 21:04:07 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! ouadol: pickfire s33se fcolista xsteadfastx chris| fabled ephemer0l_ craftyguy vitronic Gottox nlf Arcaire clandmeter2 DLange tyng MatrixTraveler[m victorbjelkholm LongyanG jakedt NightKhaos nmeum ryanlelek hrmlgon-z skarnet shreyansh_k BitL0G1c zaolin Shiz Adran
2017-12-25 21:04:12 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! ktkjlhm: NightKhaos xsteadfastx xentec mixi Shiz jakedt victorbjelkholm vitronic fcolista_web ephemer0l_ blueness nmeum minimalism tyng Somasis cyteen_ Adran pardis hrmlgon-z mjeanson zaolin Arcaire craftyguy nlf chris| DLan
2017-12-25 21:04:17 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! kutdqqfaev: zaolin mixi ryanlelek BitL0G1c fabled arch3y blueness marble_visions fcolista kaniini Arcaire chris| minimalism jakedt victorbjelkholm mjeanson blahdodo s33se cyteen_ shreyansh_k tg craftyguy kunkku MatrixTraveler[m rnalrd NightKhaos LongyanG
2017-12-25 21:04:22 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! iqxrlkh: mixi cyteen_ fcolista_web LongyanG DLange blueness skarnet kaniini rnalrd arch3y chris| Gottox tyng Shiz lep2 tg KillTheIdols1 hrmlgon-z minimalism pardis blahdodo Somasis victorbjelkholm BitL0G1c pickfire
2017-12-25 21:04:27 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! vpftqpbazv: Shiz blahdodo pardis ryanlelek chris| kaniini mjeanson xentec Adran tg clandmeter2 pickfire kunkku BitL0G1c ephemer0l_ fcolista Arcaire marble_visions skarnet vitronic NightKhaos Alchemical fabled minimalism s33se nlf tyng MatrixTra
2017-12-25 21:04:32 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! vweiftt: Somasis jakedt blahdodo blueness s33se fcolista_web tyng BitL0G1c Adran kunkku minimalism craftyguy nlf fabled skarnet chris| victorbjelkholm kaniini marble_visions LongyanG shreyansh_k
2017-12-25 21:04:33 <TBB> but then, considering how dumb you are, that's probably beyond your skills...
2017-12-25 21:04:37 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! takkr: jakedt pardis tg kunkku fcolista xsteadfastx cyteen_ lep2 Arcaire minimalism zaolin nlf MatrixTraveler[m shreyansh_k skarnet BitL0G1c mjeanson tyng Gottox fabled marble_visions KillTheIdols1 LongyanG NightKhaos pickfire mixi clandmeter2 jailbox nmeum rnalrd
2017-12-25 21:04:42 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! bdfuxo: lostd Arcaire victorbjelkholm nlf skarnet jakedt fcolista_web BitL0G1c LongyanG Gottox Shiz kunkku fabled KillTheIdols1 mjeanson pardis vitronic hrmlgon-z zaolin craftyguy blue
2017-12-25 21:04:47 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! ajqzuvifk: blahdodo xsteadfastx skarnet hrmlgon-z mixi Arcaire tyng NightKhaos tg victorbjelkholm Gottox arch3y vitronic lostd pickfire fcolista craftyguy mjeanson LongyanG clandmeter2
2017-12-25 21:04:52 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! mewetaddfb: fabled nmeum mjeanson lostd blahdodo MatrixTraveler[m jailbox kaniini nlf Shiz KillTheIdols1 victorbjelkholm blueness fcolista tyng minimalism kunkku vi
2017-12-25 21:04:57 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! jpdmtvdwj: vitronic Gottox fcolista_web nmeum Adran MatrixTraveler[m craftyguy zaolin Arcaire lostd shreyansh_k xsteadfastx kunkku minimalism jailbox ephemer0l_ NightKhaos ryanlelek blahdodo kaniini s33se hrmlgon-z marble_visions BitL0G1c jaked
2017-12-25 21:05:02 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! trmte: xentec ryanlelek skarnet Gottox fabled tg s33se Adran lostd lep2 craftyguy zaolin kunkku Alchemical Somasis clandmeter2 pickfire ephemer0l_ blueness rnalrd cyteen_ kaniini arch3y fcolista_web BitL0G1c fcolis
2017-12-25 21:05:07 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! uplgytto: LongyanG Shiz BitL0G1c ryanlelek cyteen_ KillTheIdols1 tyng ephemer0l_ DLange skarnet fabled marble_visions xsteadfastx MatrixTraveler[m s33se NightKhaos arch3y par
2017-12-25 21:05:12 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! akjtg: marble_visions blahdodo mjeanson DLange xsteadfastx skarnet cyteen_ Somasis fcolista_web tyng victorbjelkholm BitL0G1c NightKhaos shreyansh_k kunkku arch3y pickfire MatrixTraveler[m hrmlgon-z tg ja
2017-12-25 21:05:17 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! srfdd: blahdodo hrmlgon-z fabled nmeum Alchemical NightKhaos pardis minimalism ephemer0l_ kaniini fcolista_web Adran fcolista rnalrd BitL0G1c chris| Arcaire DLange craftyguy skarnet vi
2017-12-25 21:05:22 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! rbytrdhbjl: jailbox tyng minimalism LongyanG Alchemical zaolin fcolista_web chris| hrmlgon-z lep2 kaniini DLange Somasis Adran shreyansh_k blueness fabled nlf clandmeter2 pardis blahdodo KillTheIdols1 nme
2017-12-25 21:05:27 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! rqaqepxcjd: BitL0G1c vitronic kunkku craftyguy lostd arch3y lep2 nlf blueness pardis kaniini fcolista rnalrd Arcaire Alchemical xentec fabled tyng tg marble_visions victorbje
2017-12-25 21:05:32 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! ktseojbduh: skarnet fcolista Shiz xsteadfastx jailbox marble_visions hrmlgon-z mjeanson s33se zaolin Gottox fabled chris| ephemer0l_ KillTheIdols1 Alchemical pardis craftyguy Arcaire b
2017-12-25 21:05:37 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! ueavpadba: Alchemical clandmeter2 Gottox pickfire fcolista_web Adran Arcaire Somasis mjeanson ryanlelek zaolin rnalrd victorbjelkholm arch3y jakedt fabled blueness KillTheIdols1 lep2 jailbox Nig
2017-12-25 21:05:42 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! otujxyovm: shreyansh_k fabled lostd MatrixTraveler[m skarnet Somasis jakedt clandmeter2 ryanlelek BitL0G1c chris| pardis zaolin blueness Gottox minimalism Adran fcolista tg k
2017-12-25 21:05:47 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! aqfbjmpm: lep2 chris| Somasis marble_visions Adran KillTheIdols1 ephemer0l_ s33se craftyguy DLange NightKhaos Shiz clandmeter2 cyteen_ ryanlelek Gottox mjeanson victorbjelkholm pardis rnalrd blu
2017-12-25 21:05:52 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! xgdtwjudd: blueness xsteadfastx vitronic NightKhaos nmeum cyteen_ lep2 nlf tyng pickfire fcolista craftyguy BitL0G1c rnalrd shreyansh_k Somasis fcolista_web KillTheIdols1 DLange arch3y MatrixTraveler[m Shiz marble_visions mixi fabled lostd hrmlgon-z jail
2017-12-25 21:05:57 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! kbclei: MatrixTraveler[m arch3y Somasis minimalism xentec tg lep2 Adran kaniini Alchemical LongyanG lostd Gottox xsteadfastx kunkku cyteen_ fcolista vitronic pardis blueness chris| Arcaire nlf ephemer0l_ fcolista_web ryanlelek hrmlgon-z DLange jakedt mje
2017-12-25 21:06:02 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! lltajda: lep2 LongyanG xsteadfastx cyteen_ mjeanson jailbox victorbjelkholm nmeum ephemer0l_ jakedt NightKhaos Somasis minimalism kunkku mixi Gottox nlf s33se zaolin tg fcolista_web Ad
2017-12-25 21:06:07 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! spmlna: minimalism Shiz Adran skarnet nlf Gottox nmeum pardis BitL0G1c pickfire hrmlgon-z craftyguy s33se Alchemical ryanlelek jailbox lep2 KillTheIdols1 shreyansh_
2017-12-25 21:06:12 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! lglhdiuyu: jakedt tyng minimalism craftyguy arch3y kaniini lep2 MatrixTraveler[m nlf Arcaire xentec marble_visions zaolin chris| KillTheIdols1 tg s33se clandmeter2 kunkku skarnet pardis shreyansh_k lostd mixi fcolista Somasis pickfire BitL0G1c ryanlelek ephemer0l_
2017-12-25 21:06:17 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! crpspazdoz: minimalism chris| marble_visions DLange Adran fcolista mjeanson vitronic pickfire hrmlgon-z victorbjelkholm BitL0G1c jailbox pardis rnalrd mixi LongyanG fabled tg Gottox Arcaire KillTheIdols1
2017-12-25 21:06:22 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! ntrbjie: nmeum victorbjelkholm blueness lep2 mjeanson MatrixTraveler[m cyteen_ hrmlgon-z chris| Arcaire blahdodo Somasis fcolista_web tg fcolista s33se clandmeter2 ryanlelek fabled xentec jakedt jailbox Alchemical LongyanG KillTheIdols1 skarnet
2017-12-25 21:06:27 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! nduxg: lep2 jakedt fcolista vitronic kaniini ryanlelek pickfire xentec tyng Adran NightKhaos hrmlgon-z victorbjelkholm fcolista_web DLange LongyanG chris| skarnet blahdodo Somasis tg shreyansh_k nmeum Bit
2017-12-25 21:06:32 <vitronic> clandmeter, fabled , fcolista , jirutka , kaniini , ncopa , rnalrd , Shiz,  span ??
2017-12-25 21:06:32 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! fspdofzyr: xsteadfastx Gottox victorbjelkholm shreyansh_k blueness kunkku fabled mjeanson pardis ryanlelek marble_visions Alchemical fcolista arch3y MatrixTraveler[
2017-12-25 21:06:37 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! ddlfewkyas: mjeanson mixi nmeum victorbjelkholm NightKhaos fabled kaniini LongyanG s33se chris| Gottox vitronic ryanlelek blueness lostd fcolista craftyguy MatrixTraveler[m pickfire cyteen_ zaolin jakedt clandmeter2 fcolista
2017-12-25 21:06:42 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! rptmhxkwsd: MatrixTraveler[m victorbjelkholm craftyguy jailbox ephemer0l_ blahdodo DLange KillTheIdols1 jakedt kunkku shreyansh_k blueness arch3y minimalism Gottox Adran nmeum chris| r
2017-12-25 21:06:47 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! sbnsgaaha: Shiz hrmlgon-z pickfire mjeanson zaolin lostd blahdodo nlf mixi minimalism DLange nmeum chris| vitronic fabled pardis Alchemical fcolista_web tyng LongyanG KillTheIdols1 blueness tg clandmeter2 cyteen_ kaniini skarnet Arcaire jailbox xsteadfas
2017-12-25 21:06:52 <samra`ar36> ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ HAPPY NIGGER MAS!! IF YOU WANT JOIN A CELEBRATION THAT IS NIGGER FREE PLEASE JOIN #/JOIN RIGHT HERE ON THIS NETWORK!! bdplbp: hrmlgon-z jailbox fcolista_web xentec mjeanson cyteen_ kunkku blahdodo shreyansh_k craftyguy lep2 KillTheIdols1 LongyanG fabled kaniini mixi tg minimalism BitL0G1c Somasis marble_visions fcolista Arcaire bl