#archlinux-ports | Logs for 2017-11-06

[00:14:10] -!- tyler has joined #archlinux-ports
[00:14:34] tyler is now known as Guest40999
[00:26:32] Guest40999 is now known as tyzoid
[07:31:29] -!- brtln has quit [*.net *.split]
[07:31:30] -!- bq has quit [*.net *.split]
[07:31:30] -!- brtl_n has quit [*.net *.split]
[07:31:31] -!- qguv has quit [*.net *.split]
[07:31:31] -!- nullmark has quit [*.net *.split]
[07:31:33] -!- Cthulu201 has quit [*.net *.split]
[07:31:33] -!- phrik has quit [*.net *.split]
[07:31:34] -!- rewbycraft has quit [*.net *.split]
[07:31:34] -!- drathir has quit [*.net *.split]
[07:31:34] -!- r00t^2 has quit [*.net *.split]
[07:31:35] -!- finsternis has quit [*.net *.split]
[07:31:36] -!- kerberizer has quit [*.net *.split]
[07:31:37] -!- fsckd has quit [*.net *.split]
[07:31:37] -!- caveat has quit [*.net *.split]
[07:31:37] -!- CalimeroTeknik has quit [*.net *.split]
[07:31:37] -!- City-busz has quit [*.net *.split]
[07:31:37] -!- ztrawhcse has quit [*.net *.split]
[07:31:37] -!- comrumino has quit [*.net *.split]
[07:31:38] -!- wyre_ has quit [*.net *.split]
[07:31:38] -!- dopsi_ has quit [*.net *.split]
[07:31:38] -!- p71 has quit [*.net *.split]
[07:31:38] -!- pulec has quit [*.net *.split]
[07:31:38] -!- jp has quit [*.net *.split]
[07:31:39] -!- tyzoid has quit [*.net *.split]
[07:31:39] -!- Polichronucci has quit [*.net *.split]
[07:31:39] -!- yokel has quit [*.net *.split]
[07:31:39] -!- eschwartz[m] has quit [*.net *.split]
[07:31:39] -!- privilege has quit [*.net *.split]
[07:31:39] -!- ianaca has quit [*.net *.split]
[07:31:40] -!- yar has quit [*.net *.split]
[08:12:38] -!- tyzoid has joined #archlinux-ports
[08:12:38] -!- eschwartz[m] has joined #archlinux-ports
[08:12:38] -!- brtln has joined #archlinux-ports
[08:12:38] -!- Polichronucci has joined #archlinux-ports
[08:12:38] -!- wyre_ has joined #archlinux-ports
[08:12:38] -!- dopsi_ has joined #archlinux-ports
[08:12:38] -!- privilege has joined #archlinux-ports
[08:12:38] -!- ztrawhcse has joined #archlinux-ports
[08:12:38] -!- City-busz has joined #archlinux-ports
[08:12:38] -!- p71 has joined #archlinux-ports
[08:12:38] -!- phrik has joined #archlinux-ports
[08:12:38] -!- pulec has joined #archlinux-ports
[08:12:38] -!- yar has joined #archlinux-ports
[08:12:38] -!- rewbycraft has joined #archlinux-ports
[08:12:38] -!- Cthulu201 has joined #archlinux-ports
[08:12:38] -!- nullmark has joined #archlinux-ports
[08:12:38] -!- jp has joined #archlinux-ports
[08:12:38] -!- yokel has joined #archlinux-ports
[08:12:38] -!- ianaca has joined #archlinux-ports
[08:12:38] -!- finsternis has joined #archlinux-ports
[08:12:38] -!- kerberizer has joined #archlinux-ports
[08:12:38] -!- drathir has joined #archlinux-ports
[08:12:38] -!- r00t^2 has joined #archlinux-ports
[08:12:38] -!- comrumino has joined #archlinux-ports
[08:12:38] -!- fsckd has joined #archlinux-ports
[08:12:38] -!- caveat has joined #archlinux-ports
[08:12:38] -!- bq has joined #archlinux-ports
[08:12:38] -!- brtl_n has joined #archlinux-ports
[08:12:38] -!- CalimeroTeknik has joined #archlinux-ports
[08:12:38] -!- qguv has joined #archlinux-ports
[08:12:57] -!- because has quit [Max SendQ exceeded]
[08:13:16] -!- because has joined #archlinux-ports
[08:22:25] -!- brtln has quit [Ping timeout: 248 seconds]
[08:23:10] -!- eschwartz[m] has quit [Ping timeout: 264 seconds]
[08:33:30] dopsi_ is now known as dopsi
[08:45:38] -!- deep42thought has joined #archlinux-ports
[10:14:18] -!- brtln has joined #archlinux-ports
[10:32:52] -!- eschwartz[m] has joined #archlinux-ports
[10:37:40] -!- eschwartz[m] has quit [Remote host closed the connection]
[10:37:41] -!- brtln has quit [Read error: Connection reset by peer]
[10:42:39] -!- brtln has joined #archlinux-ports
[10:58:05] -!- eschwartz[m] has joined #archlinux-ports
[11:16:44] <brtln> https://lists.archlinux.org
[11:16:45] <phrik> Title: [arch-dev-public] [draft] The end of i686 support (at lists.archlinux.org)
[11:25:50] <deep42thought> brtln: looks good
[11:26:43] -!- allanbrokeit has joined #archlinux-ports
[11:27:28] <allanbrokeit> where can I file a bug report about the archlinux-32 website?
[11:28:12] <allanbrokeit> hrm... it is probably "ok"...
[11:28:23] <deep42thought> we don't have a bugtracker yet, but you can post on the forum (or here)
[11:30:22] <brtln> allanbrokeit: github or just here, I guess?
[11:30:41] <deep42thought> ah, right, there are github issues, too :-)
[11:31:05] <allanbrokeit> on the download page, it talks about using "pacman -Sc" to avoid conflict - should probably be -Scc, but less safe :D
[11:32:23] <deep42thought> there is not really a benefit of -cc
[11:32:41] <deep42thought> installed packages will stay installed anyway - if there is a conflict, you won't resolve that
[11:32:44] <allanbrokeit> when you update packages to match current versions, there will be conflicts
[11:33:02] <deep42thought> what kind of conflict?
[11:33:16] <allanbrokeit> I downgraded heaps of packages with -Syuu... you build the current version, my cached package will be bad
[11:33:29] <deep42thought> ah, I see
[11:33:49] <deep42thought> so you should run pacman -Sc a second time after pacman -Syuu, then
[11:34:00] <deep42thought> this is probably less radical
[11:34:11] <deep42thought> I don't like the idea of removing all cached packages :-/
[11:34:28] <allanbrokeit> I am running "pacman -Qq | sudo pacman -S -" after the initial -Syuu
[11:35:06] <deep42thought> this will probably install all packages as explicitely installed
[11:35:23] <allanbrokeit> nope - pacman is smart
[11:35:29] <deep42thought> ok, nice
[11:37:45] <allanbrokeit> moment of truth... rebooting my webserver
[11:37:49] <deep42thought> so: "pacman -Syy archlinux32-keyring-transition && pacman -Scc && pacman -Syuu && pacman Qq | pacman -S -" should do the trick, then?
[11:37:52] <deep42thought> :-)
[11:37:57] <allanbrokeit> I stay with arch if it works!
[11:38:08] <deep42thought> you run a 32 bit webserver?
[11:38:13] <allanbrokeit> yes
[11:38:58] <allanbrokeit> !isitup allanmcrae.com
[11:39:00] <phrik> allanbrokeit: Yay, allanmcrae.com is up. // isitup.org
[11:39:17] <allanbrokeit> yay! my website lives
[11:42:27] <deep42thought> The problem, I see with 'pacman -Scc' is, that also "external" packages will be removed from the cache - and we can't guarantee, that they will be available again.
[11:42:45] <deep42thought> same holds true for packages which are no longer available
[11:43:30] <allanbrokeit> yeah - another -Sc after the -Syuu is enough
[11:57:01] <deep42thought> I'm wondering if we should update https://github.com or if that info is (partially) obsolete now
[11:57:02] <phrik> Title: Home · archlinux32/builder Wiki · GitHub (at github.com)
[12:05:25] <allanbrokeit> it becomes unimportant once your packages catch up
[12:05:58] <deep42thought> allanbrokeit: the update to the wiki is unrelated to your bug, I was thinking about general stuff like the layout of the build master
[12:09:04] <privilege> Well, if allanbrokeit's website can run it without breaking, it is probably safe for public use. :D
[12:09:16] * privilege throws a party
[12:10:48] * fsckd wonders if privilege ever sleeps :P
[12:17:27] <privilege> fsckd: doctors orders, I have to sleep at least 3 hours a day :(
[12:17:51] <fsckd> lol
[12:18:22] -!- privilege has quit [Quit: A random quit message]
[12:45:37] -!- titus_livius has quit [Ping timeout: 248 seconds]
[13:00:14] <tyzoid> deep42thought: Did vim add a new dependency?
[13:00:53] <tyzoid> I did a package update of vim on archstrike32, and I get:
[13:00:56] <tyzoid> vim: error while loading shared libraries: libtinfo.so.6: cannot open shared object file: No such file or directory
[13:08:03] <brtln> -S without 'yu' is unsupported unless you know what you're doing
[13:10:18] <tyzoid> brtln: It was an -Syu
[13:10:36] <tyzoid> brtln: My suspicion was that it was packaged without a new dependency
[13:10:50] <brtln> then there is something wrong on your side, libtinfo.so.6 is provided by the latest ncurses
[13:10:55] <brtln> moved from testing to core yesterday
[13:11:04] <tyzoid> on arch32?
[13:11:13] <brtln> on arch
[13:11:22] <tyzoid> I'm having this issue on arch32
[13:11:32] <tyzoid> but I'll double check what's going on
[13:11:44] <tyzoid> actually, be back in a bit
[13:27:47] -!- titus_livius has joined #archlinux-ports
[13:57:11] <deep42thought> sry tyzoid, I was busy
[13:58:38] -!- tyz has joined #archlinux-ports
[13:59:52] <tyz> deep42thought: no problem
[14:06:19] <tyz> deep42thought: https://ptpb.pw <-- works
[14:06:32] <tyz> Wait, I take that back
[14:06:37] <tyz> that one is the broken one
[14:06:44] <tyz> this is the one that works: https://ptpb.pw
[14:06:51] <deep42thought> the problem is that the right ncurses is still in testing
[14:06:56] <tyz> I don't see a big difference anywhere
[14:06:58] <tyz> ah
[14:07:06] <tyz> let me grab that from testing and test quick
[14:07:34] <deep42thought> from the todos: "bin/common-functions (line 926):
[14:07:34] <deep42thought> also consider dependencies in the other direction (?)"
[14:08:39] <tyz> deep42thought: That seems to work
[14:08:54] <deep42thought> the logic in db-update is wrong, currently
[14:09:03] <tyz> makes sense
[14:09:22] <deep42thought> it moves the biggest subset A of all moveable packages, whose dependencies are all within A
[14:09:40] <deep42thought> but it needs to take care, that nothing left behind depends on something in A, too
[14:09:45] <tyz> deep42thought: Though should the vim package say it depends on ncursees?
[14:09:56] <tyz> since it currently doesn't
[14:10:01] <deep42thought> it doesn't?
[14:10:03] <deep42thought> oh
[14:10:19] <deep42thought> it's probably hidden below a few layers of other dependencies, then
[14:11:03] <tyz> deep42thought: Possibly
[14:11:20] <tyz> deep42thought: does perl depend on ncurses?
[14:12:11] <deep42thought> not directly
[14:12:13] -!- guys has joined #archlinux-ports
[14:12:42] <deep42thought> nope, not at all
[14:14:04] <deep42thought> vi depends on ncurses, though
[14:14:11] <tyz> deep42thought: So yeah, looks like vim has no path towards ncurses dependency
[14:14:19] <deep42thought> strange
[14:14:28] <tyz> actually
[14:14:30] <tyz> vim-runtime=8.0.1272-1 gpm acl
[14:14:37] <tyz> ^ Those are the only hard depends for vim
[14:14:42] <tyz> I didn't check the other two *facepalm*
[14:14:43] <tyz> one sec
[14:15:11] <deep42thought> it does
[14:15:12] <tyz> deep42thought: Actually yeah
[14:15:20] <tyz> vim -> gpm -> bash -> ncurses
[14:15:33] <deep42thought> bash or procps-ng
[14:15:58] <tyz> does your script just not handle that many layers of indirection?
[14:16:05] <deep42thought> hmm, the build master currently only considers direct dependencies
[14:16:32] <deep42thought> the assumption is, that all the intermediate packages will change, too if something else needs to be rebuilt
[14:16:39] <deep42thought> but maybe this is wrong?
[14:16:42] <tyz> deep42thought: If you wanted to move ncurses over, that fixes the vim issue. Not sure if there are other packages that would break from an updated ncurses
[14:16:54] <deep42thought> already moved ncurses
[14:16:57] <tyz> oh
[14:17:00] <tyz> sounds good
[14:17:02] <deep42thought> and yes, there are some breaking
[14:17:17] <deep42thought> e.g. some, that don't build, currently
[14:17:17] <tyz> That assumption seems sound...
[14:17:23] <deep42thought> but I don't care about them anyway
[14:17:33] <tyz> That said, does gpm need an update? or bash?
[14:17:37] <deep42thought> I'm not really sure about that assumption anymore
[14:18:08] <deep42thought> can't say, if "yes", then it was moved already when I moved ncurses :-/
[14:18:13] <tyz> Seems sound on the surface, but that assumes that each project has an explicit dependency for anything it links against
[14:18:25] <tyz> So I think that's where we're going wrong
[14:18:35] <tyz> bash links directly against ncurses
[14:18:43] <tyz> but gpm may not
[14:18:47] <tyz> but vim does
[14:19:04] <tyz> so if that chain gets broken there, since gpm doesn't "need" an update
[14:19:08] <tyz> then vim will be out of date
[14:19:13] <tyz> or visa-versa
[14:19:38] <tyz> so I think the better solution is to mark all explicitly linked libraries
[14:19:38] <deep42thought> I will first correct my current known issue and then we can see if the error appeares again
[14:19:55] <tyz> which is it doesn't track all the way down the tree?
[14:20:00] <deep42thought> no
[14:20:01] <tyz> that *should* work
[14:20:04] <tyz> oh
[14:20:06] <tyz> ?
[14:20:10] <deep42thought> the current issue is, that it only looks in one direction
[14:20:14] <tyz> ah
[14:20:20] <deep42thought> e.g. a package is only moved if all dependencies are moved, too
[14:20:33] <deep42thought> but not if all dependent packages are moved too
[14:20:47] <tyz> deep42thought: but that wouldn't have resolved this issue, though?
[14:20:55] <tyz> since vim was moved before dependencies were
[14:20:56] <deep42thought> (sry, actually it is the other way round, but it does not matter)
[14:21:00] <tyz> ah
[14:21:10] <tyz> makes sense
[14:21:13] <deep42thought> :-/
[14:23:37] <tyz> deep42thought: you know how you said ssh tunnels are unstable?
[14:23:43] <deep42thought> yes
[14:23:48] <tyz> Turns out...
[14:23:54] <tyz> if you daemonize 4 of them
[14:24:09] <tyz> you can *almost* always bring back the other three if they crash
[14:24:31] <deep42thought> usually I just did 'while true; do ssh ...; done'
[14:24:39] <tyz> That's what I did before
[14:24:44] <tyz> but that doesn't work if I log out
[14:24:51] <tyz> so I moved it to systemd
[14:24:55] <deep42thought> screen
[14:25:08] <tyz> yup, but that doesn't come back on power failure
[14:25:27] <deep42thought> /etc/rc.local + screen + while true
[14:25:27] <tyz> cron+screen might have worked
[14:25:28] <deep42thought> ;-)
[14:25:32] <tyz> but systemd seemed easier
[14:25:33] <deep42thought> yeah
[14:25:45] <tyz> that's how I used to run my minecraft servers :P
[14:25:46] <deep42thought> that depends on your definition of "easy"
[14:25:49] <tyz> cron+screen
[14:25:57] -!- dopsi has quit [Ping timeout: 240 seconds]
[14:26:06] <tyz> that thing would *always* come down for one reason or another
[14:28:45] -!- dopsi has joined #archlinux-ports
[14:30:17] -!- mongy has joined #archlinux-ports
[14:49:32] -!- allanbrokeit has quit [Ping timeout: 260 seconds]
[14:56:08] <mongy> fresh install, installed vim, getting error: vim: error while loading shared libraries: libtinfo.so.6: cannot open shared object file: No such file or directory
[14:56:17] <mongy> noevim is fine tho
[14:56:44] <deep42thought> is this a recent error or ~2-3h old?
[14:57:31] <mongy> less than 1h
[14:57:41] <mongy> installed less than 1h ago
[14:57:43] <deep42thought> hmm
[14:58:02] <deep42thought> what does "pacman -Q ncurses" give?
[14:58:46] <mongy> ncurses 6.0+20170902-1
[14:59:23] <deep42thought> hmm, crap, I thought, I had moved -3 to extra
[14:59:28] <deep42thought> uh core
[14:59:32] <deep42thought> my mistake
[15:00:18] <fsckd> deep42thought = 32 bit Allan? :P
[15:00:41] <deep42thought> hmm?
[15:00:52] <deep42thought> what defines an "Allen"?
[15:02:10] <fsckd> high technical skill coupled with breaking things
[15:02:31] <deep42thought> the latter is certainly true
[15:03:31] <fsckd> being able to fix it afterwards demonstrates the former
[15:05:11] <deep42thought> mongy, it should be fixed now
[15:05:29] <deep42thought> "pacman -Syu" should install the right version of ncurses
[15:05:34] <mongy> yup!
[15:05:41] <deep42thought> :-)
[15:09:46] <eschwartz[m]> Well, ncurses is part of the base group, essentially, as bash et al depend on it. Sounds like fun.
[15:10:14] <deep42thought> base is a special case when considering dependencies:
[15:10:47] <deep42thought> If we consider base like everything else, then we can't move packages at all (as long as at least one package is on the build list)
[15:11:23] <deep42thought> because: A depends on base and B depends on base, too. If A is still on the build list, then A can't be moved, therefore base can't be moved, therefor B can't be moved
[15:12:41] <deep42thought> therefor, I remove all dependencies on "base", iff there is no package from base on the explicitely-can't-move list
[15:29:57] -!- phrik has quit [Remote host closed the connection]
[15:30:10] -!- phrik has joined #archlinux-ports
[16:18:30] -!- deep42thought has quit [Quit: Leaving.]
[18:01:25] -!- deep42thought has joined #archlinux-ports
[20:24:22] -!- tyzoid has quit [Ping timeout: 260 seconds]
[20:26:13] -!- tyzoid has joined #archlinux-ports
[20:53:52] <deep42thought> hey tyzoid, did you have any luck with the bugtracker yet?
[21:11:58] -!- jp has quit [Quit: https://ptpb.pw
[21:12:14] -!- jp has joined #archlinux-ports
[21:12:57] -!- dopsi has quit [Ping timeout: 240 seconds]
[21:15:14] -!- eschwartz[m] has quit [Ping timeout: 255 seconds]
[21:15:44] -!- brtln has quit [Ping timeout: 252 seconds]
[21:17:23] -!- dopsi has joined #archlinux-ports
[22:19:05] -!- tyz has quit [Ping timeout: 260 seconds]
[22:58:35] -!- brtln has joined #archlinux-ports
[23:14:38] -!- eschwartz[m] has joined #archlinux-ports
[23:35:44] -!- deep42thought has quit [Quit: Leaving.]