#archlinux-ports | Logs for 2018-03-17

Back
[01:00:08] -!- MrBIOS has joined #archlinux-ports
[01:01:43] -!- City-busz has quit [Read error: Connection reset by peer]
[01:01:44] -!- eschwartz has quit [Remote host closed the connection]
[01:01:53] -!- City-busz has joined #archlinux-ports
[01:01:56] -!- anyone has joined #archlinux-ports
[01:20:49] -!- guys has quit [Ping timeout: 248 seconds]
[01:30:37] -!- sunshavi has quit [Remote host closed the connection]
[01:54:42] -!- Soni has quit [Quit: Leaving]
[02:15:49] typikal is now known as alyptik
[03:49:27] charims_ is now known as charims
[04:46:05] * buildmaster goes insane.
[05:16:10] -!- hringriin_ has joined #archlinux-ports
[05:19:33] -!- hringriin has quit [Ping timeout: 263 seconds]
[05:19:33] hringriin_ is now known as hringriin
[06:23:57] -!- alyptik has quit [Ping timeout: 240 seconds]
[06:25:25] -!- typikal has joined #archlinux-ports
[08:00:06] * buildmaster goes insane.
[08:22:18] -!- oaken-source has joined #archlinux-ports
[08:28:02] -!- abaumann has joined #archlinux-ports
[08:28:53] <abaumann> swiftgeek: your CPU has no SSE2? I noticed lately that many packages depending on some basic libraries segfault because the basic libraries contain SSE2 specific opcodes.
[08:34:52] -!- sunshavi has joined #archlinux-ports
[08:46:48] <abaumann> bin/check-opcodes ./extra/librsvg-2:2.42.2-1.0-i686.pkg.tar.xz
[08:47:24] <abaumann> pactree -r librsvg shows: gtk3 and gtk2
[08:47:42] <abaumann> some my belowed 'geany' IDE also fails. :-(
[08:48:00] <abaumann> ERROR: libpixbufloader-svg.so contains MMX, SSE or SSE2 opcodes
[08:48:03] <abaumann> ERROR: librsvg-2.so.2.42.2 contains MMX, SSE or SSE2 opcodes
[08:48:33] <abaumann> /data/arch32/builder/bin/check-opcodes -a pentium4 ./extra/librsvg-2:2.42.2-1.0-i686.pkg.tar.xz
[08:48:36] <abaumann> ERROR: libpixbufloader-svg.so contains SSE3 or newer opcodes
[08:48:38] <abaumann> ERROR: librsvg-2.so.2.42.2 contains SSE3 or newer opcodes
[08:48:41] <abaumann> huh?
[08:50:09] <abaumann> dito qemu.
[08:51:55] <abaumann> we should maybe make a survey: how many people have which class of CPUs?
[08:56:04] <abaumann> "i486 plain", "up to pentium3 (no SSE2, SSE, but MMX and CMOV)" and "newer (all including SSE2, but no SSE3)" were the classes of repos we opted for..
[08:56:17] -!- abaumann has quit [Quit: leaving]
[09:41:07] -!- isacdaavid has quit [Quit: Leaving.]
[10:21:03] -!- deep42thought has joined #archlinux-ports
[10:31:38] -!- deep42thought has quit [Quit: leaving]
[10:31:55] -!- deep42thought has joined #archlinux-ports
[10:35:57] -!- deep42thought has quit [Client Quit]
[10:52:35] -!- deep42thought has joined #archlinux-ports
[10:54:17] <deep42thought> hmm, the buildmaster run out of hdd space
[10:54:38] <deep42thought> rewbycraft: do you think you can add another 5 or 10G?
[10:56:05] * buildmaster goes insane.
[10:57:51] -!- sunshavi has quit [Remote host closed the connection]
[10:58:54] -!- sunshavi has joined #archlinux-ports
[11:00:10] * buildmaster resumes sanity.
[11:25:46] <deep42thought> abaumann: I'm afraid, adding additional "architectures" to the repositories will require additional work beyond getting the database fully functional :-/
[11:26:09] <deep42thought> but it's definitely something we should have in the end ...
[11:38:17] <deep42thought> tyzoid: is there an easy way to set up a poll in the forums?
[11:39:29] -!- deep42thought has quit [Quit: leaving]
[11:49:09] <swiftgeek> well better have first working perhaps slowly and then add optimized versions or work with upstream of important libs to detect extensions on runtime instead of compile time
[11:50:02] <swiftgeek> also p3 is i686 afair xD
[11:50:38] <swiftgeek> yep
[11:50:40] <swiftgeek> https://en.wikipedia.org(microarchitecture)
[11:55:57] <swiftgeek> and it will be hard to make a poll on such thing when users of older machines won't be there xD
[12:08:40] -!- mike1960 has joined #archlinux-ports
[12:10:23] <mike1960> hi,I have a queston: I've just installed archlinux32. when I reboot I have to run "dhcpcd enp0s4" every time to rebuild resolv.conf. why is this?
[12:15:29] <mike1960> hello?
[12:24:02] -!- abaumann has joined #archlinux-ports
[12:24:10] <abaumann> mike1960: hi, did you check in /etc/systemd/network or in /etc/netctl (whatever you are using), what the name of your device is?
[12:24:40] <abaumann> sometimes systemd decides to rename interface names. :-)
[12:26:53] <mike1960> sis900
[12:27:13] <mike1960> there's nothing in /etc/systemd/network...
[12:27:21] <abaumann> so, you configured netctl?
[12:28:26] <mike1960> I'm not sure what to do with netctl
[12:28:26] <abaumann> Did you enable your profile with systemctl? What's the value of Interface? Does it fit to what 'ip addr' or 'ifconfig -a' is telling you?
[12:29:40] <abaumann> What's 'netctl list' telling you?
[12:29:51] <abaumann> Do you have an asterix '*' at a profile name?
[12:30:21] <mike1960> nothing in /etc/netctl/interfaces
[12:30:45] <mike1960> netctl list does nothing
[12:31:00] <mike1960> *confused* I'm new at this...
[12:31:25] <abaumann> did you install a network at all after installing to the hard disk?
[12:31:28] <mike1960> ifconfig: command not found
[12:31:30] <abaumann> https://wiki.archlinux.org
[12:31:31] <phrik> Title: Network configuration - ArchWiki (at wiki.archlinux.org)
[12:31:59] <abaumann> ok, ifconfig is in net-tools, 'ip addr' is telling you what?
[12:32:10] <abaumann> it should enlist interface names..
[12:33:56] <mike1960> ip addr gives me lo and enp0s4
[12:34:04] <mike1960> and a lot of detail
[12:34:19] <abaumann> so, enp0s4 is your network device.
[12:35:05] <abaumann> cp /etc/netctl/examples/ethernet-dhcp to /etc/netctl/mynet and adapt the Interface
[12:35:29] <abaumann> then follow https://wiki.archlinux.org netctl start mynet, netctl enable mynet
[12:35:42] <mike1960> yes. I've now installed net-tools
[12:36:04] <abaumann> that's just for the old ifconfig, shows basically the same stuff as 'ip addr'
[12:40:00] <mike1960> i changed mynet from eth0 to eth0s4
[12:40:13] <mike1960> i mean enp0s4
[12:40:35] <abaumann> now 'netctl list' should show mynet
[12:41:24] <mike1960> yes. it does
[12:41:46] <abaumann> now you can enable it and reboot the machine, and then the network should come up.
[12:42:07] <mike1960> but i still need to re-run dhcpcd enp-s4 after rebooting
[12:42:18] <abaumann> why?
[12:42:33] <abaumann> did you 'netctl enable mynet'?
[12:42:44] <abaumann> this adds systemd services for you network service..
[12:42:49] <mike1960> damn. no
[12:42:53] <abaumann> :-)
[12:43:17] <mike1960> ok. i did it now
[12:43:29] <mike1960> but i did it after rebooting
[12:43:48] <abaumann> ah, yea. it's active for the next boot.
[12:44:16] <mike1960> ping www.google.com still doesn't work, so I need to rerun dhcpcd enp0s4
[12:44:50] <abaumann> aeh. did you put something into /etc/resolv.conf?
[12:45:05] <abaumann> usually dhcpcd adds DNS-servers automatically into /etc/resolv.conf though.
[12:45:12] <abaumann> *did you -> *did netctl
[12:46:22] <mike1960> yes i did
[12:46:34] <mike1960> rebooting...
[12:46:55] <mike1960> networking confuses me.
[12:47:45] <mike1960> yay!!!
[12:47:53] <abaumann> :-)
[12:48:14] <abaumann> not that complicated. but on archlinux you have to do all stuff by hand..
[12:48:17] <mike1960> rebooted, and I can ping google. So I assume things are working
[12:48:57] <mike1960> yes... I've been using debian. but I thought I'd try something new. using arch32 on a really old laptop
[12:49:48] <mike1960> now I need to figure out how to install xfce4 dsktop. can you recommend some reading?
[12:50:06] <abaumann> of course:
[12:50:07] <abaumann> https://wiki.archlinux.org
[12:50:07] <phrik> Title: Xfce - ArchWiki (at wiki.archlinux.org)
[12:50:45] <mike1960> thanks abaumann. have a lovely st patrick's day :)
[12:50:59] <mike1960> cheers!
[12:51:06] -!- mike1960 has quit []
[12:51:08] <abaumann> np.
[12:56:31] <swiftgeek> abaumann: it will be hard to make a poll on such thing when users of older machines won't be on forum yet :D
[12:57:22] <swiftgeek> abaumann: also i686 was the lowest requirement for archlinux before arch32 afair
[12:58:13] <abaumann> yes. whatever i686 means.. :-)
[12:58:50] <abaumann> I was not sure, but what was the old policy? with SSE2 or without?
[12:59:19] <abaumann> MMX and SSE, CMOV are so common, so those can almost be taken for granted.
[13:01:50] <swiftgeek> abaumann: https://en.wikipedia.org(microarchitecture)
[13:02:26] <swiftgeek> abaumann: MMX is definitely part of i686
[13:02:49] <brtln> I don't think we've been disallowing SSE so Pentium 3 is Arch's i686
[13:03:15] <swiftgeek> brtln: this is about SSE2/3
[13:03:32] <swiftgeek> well at least with current issues :D
[13:03:39] <brtln> this is about what was i686 in Arch, so just told you
[13:04:28] <swiftgeek> brtln: currently there are few important packages (like librsvg for gtk3) that require newer SSE
[13:08:58] <abaumann> Hence the survey. :-)
[13:09:11] <abaumann> Before we start a old-vintage-multiplatform project here.. ;-)
[13:09:46] <abaumann> https://bbs.archlinux32.org
[13:09:47] <phrik> Title: Survey: On what cpu do you (want to) use archlinux32? / Installation / Arch Linux 32 Forums (at bbs.archlinux32.org)
[13:11:25] <abaumann> swiftgeek: MMX came in with the Pentium MMX, even some AMDs later implemented MMX along to 3dnow.
[13:13:47] <swiftgeek> abaumann: just trying to say that requiring SSE2 is taking it too far
[13:14:19] <swiftgeek> especially in mobile space
[13:14:26] <swiftgeek> and low power stuff
[13:15:19] <abaumann> have a look at my list: 2 P3 are not happy at all with SSE2.. and the AMD-K6 has no CMOV.. The Alix.E (AMD Geode) actually likes only plain MMX..
[13:15:41] <abaumann> So, currently my server, NAS and eeepc are running fine.. the rest not.
[13:16:14] <abaumann> but building a lot of flavours of Archlinux32 is quite an effort..
[13:16:33] <abaumann> https://github.com my attemps to get to i486.
[13:17:17] <abaumann> the problem is with i686-chroots (which is the preferred way of building): some configure and cmake probes are just assuming too much and then you end up with SSE2..
[13:17:45] <abaumann> otherwise you have to build pretty much on libvirt/qemu/i486 with a correctly emulated CPU.
[13:17:51] <swiftgeek> abaumann: my other point was to run perhaps those devices in unoptimized way, and work with upstream to detect extensions on runtime
[13:18:15] <abaumann> this is what the authors of the software should do/or should have done.
[13:19:06] <abaumann> I have seen packages generating AES opcodes becuase gcc -maes was present, drawing in SSE2 optimization too (as this is a safe thing for a compiler to assume, as AES is newer than SSE2).
[13:19:15] <swiftgeek> so if somebody is starting to complain that something works not as optimal as it could, we could push them towards helping with fixing upstream xD
[13:19:49] <abaumann> yeah. but to frank. most authors nowadays don't even know the difference between 32-bit and 64-bit integers.
[13:19:55] <swiftgeek> instead of creating 10 repos for each flavour
[13:20:54] <abaumann> I think, we had these SSE2 discussions before and we agreed, that for now we have one repo and we are disabling SSE2 for core packages..
[13:21:08] <abaumann> ..I'm just checking into librsvg, what draws in the special opcodes..
[13:21:40] <swiftgeek> well eventually i want to port somehow savagefb to kms hopefully
[13:22:00] <swiftgeek> it sucks that there are no public datasheets about savage
[13:22:45] <abaumann> * abaumann agrees *sigh*
[13:22:54] <swiftgeek> and eventually T23 will run coreboot xD
[13:23:17] <swiftgeek> i just need to figure out more clever way to expose LPC from it
[13:24:37] <swiftgeek> i will probably need to expose TSSOP leadframe and insert interposer pcb there to make it saner
[13:27:08] <swiftgeek> (T23 is the first thinkpad made on modern thinkpad framework, which stays 99% the same up to T400)
[13:27:16] <abaumann> never touched graphic driver programming, but a AMD Geode and a Mach64 Xorg which where actually working would come in handy. :-)
[13:27:37] <swiftgeek> well i don't mind software rendering
[13:27:54] <swiftgeek> but without KMS nothing works properly xD
[13:28:13] <abaumann> ..or just one.. :-)
[13:28:17] <abaumann> *once
[13:28:51] <swiftgeek> at least there is place in chassis to expose LPC through http://h5ai.swiftgeek.net
[13:28:52] <phrik> Title: index - powered by h5ai v0.28.1 (https://larsjung.de/h5ai/) (at h5ai.swiftgeek.net)
[13:30:02] <abaumann> http://www.thinkwiki.org
[13:30:02] <phrik> Title: Category:T23 - ThinkWiki (at www.thinkwiki.org)
[13:30:05] <abaumann> neat little thing.
[13:30:18] <swiftgeek> toronto-4.5
[13:30:34] <abaumann> yeah.. old Thinkpads definitely keep their value, I still have a T61 at work..
[13:30:50] <abaumann> and a T40 or so in the server room as terminal. :-)
[13:31:13] <swiftgeek> toronto-3.5 is for T20-22
[13:31:35] <swiftgeek> and those use ISA and it predates that stable framework
[13:31:53] <swiftgeek> and reversing that would be huge pain
[13:32:51] <swiftgeek> abaumann: T40 is either apollo or rome, can't really tell
[13:33:08] <abaumann> ah. I can check next time..
[13:33:34] <swiftgeek> abaumann: USI boards are not marked until T400 came
[13:34:07] <swiftgeek> and shortly after that, lenovo started making boards themselves, then together with compal in some joint venture
[13:35:15] <swiftgeek> abaumann: and USI boards are not really fixable but you can always try ^^
[13:35:40] <swiftgeek> abaumann: example of USI being idiots http://h5ai.swiftgeek.net
[13:35:42] <phrik> Title: index - powered by h5ai v0.28.1 (https://larsjung.de/h5ai/) (at h5ai.swiftgeek.net)
[13:35:58] <swiftgeek> sevaral amps of 3v3 going to separate board
[13:36:04] <swiftgeek> *several
[13:36:15] <abaumann> oups. :-)
[13:36:16] <swiftgeek> to light up 4 leds
[13:36:19] <swiftgeek> unfused
[13:36:32] <swiftgeek> not even 10Ω resistor or anything
[13:37:02] <swiftgeek> luckily it works after a lot of scrubbing
[14:01:47] -!- abaumann has quit [Quit: leaving]
[14:57:25] -!- yauqvnxo has joined #archlinux-ports
[14:57:29] -!- yauqvnxo has quit [Client Quit]
[15:45:47] gnugnugn1 is now known as gnugnugnu
[15:49:19] -!- isacdaavid has joined #archlinux-ports
[17:28:43] -!- deep42thought has joined #archlinux-ports
[17:36:18] * buildmaster goes insane.
[17:37:16] -!- abaumann has joined #archlinux-ports
[17:37:35] <abaumann> concerning libsrvg: the Gnome team decided to rewrite its internals in Rust.
[17:37:48] <abaumann> Now Rust targets i686, or more precisely Pentium 4 (https://fraserblog.codewise.org/rust-on-i586/)
[17:37:50] <phrik> Title: Rust on i586 (at fraserblog.codewise.org)
[17:38:24] <abaumann> This is really unfortunate, as this leaver no other option than to stick to a C-version of libsrvg, if SSE2 is not available.
[17:38:58] <abaumann> *leaves
[17:39:00] <deep42thought> :-/
[17:40:10] <abaumann> aha.
[17:40:17] <deep42thought> hmm?
[17:40:21] <abaumann> the blog claims we can bootstrap a rust on i586
[17:41:32] <abaumann> but once rust has been built on a i686, it knows about SSE2 and generates code for it. I tried to build on a "crippled" virtual machine with a Pentium 3 emulated, but still got a libsrvg containing SSE2
[17:42:41] <deep42thought> don't they implement some sober sort of compiler switch for such kind of stuff??
[17:43:19] <abaumann> there are I think some discussions about special constructs in rust to say something like: "if you have AVX -> compile this function"
[17:43:31] <abaumann> but I never used rust, so I can't really be sure..
[17:43:52] <abaumann> grep -ir sse2 in libsrvg didn't reveal anything of this sort. :-)
[17:43:55] * buildmaster resumes sanity.
[17:44:08] <deep42thought> but shouldn't such "intelligent" compilations have some option to run "rust -nosse2 ...", then?
[17:44:38] <abaumann> should.. but rust is not gcc/clang. :-)
[17:44:55] <abaumann> and it adds complecity to the compiler.. basically yet another "if"
[17:45:36] <abaumann> rustc: target-feature
[17:45:41] <deep42thought> but a switch for the target architecture is quite standard - isn't it?
[17:45:50] <abaumann> maybe a target_feature=-sse2 helps
[17:46:32] <abaumann> rustc -Ctarget-cpu=help
[17:46:45] <abaumann> aha: pentium3
[17:47:08] <abaumann> ok. I'll try that. livsrvg needs some patching for pentium3..
[17:47:15] <deep42thought> :-/
[17:47:29] <deep42thought> basically all desktop stuff depends on that, right?
[17:47:35] <abaumann> yep.
[17:47:40] <abaumann> everything gtk-ish
[17:48:34] <abaumann> oh.. the cmos battery died in my acer travelmate.. after 15 years or so. :-)
[17:49:03] <deep42thought> well, just don't reboot it, then ;-P
[17:49:14] <abaumann> please press F1 to continue..
[17:49:17] <abaumann> no problem..
[17:49:45] <abaumann> i386 - Select the i386 processor.
[17:49:49] <abaumann> ui :-)
[17:55:03] <abaumann> Ok. I'll smuggle a RUSTFLAGS="-C target-cpu=pentium3" into the Makefile and I'll see, what happens..
[17:55:11] <deep42thought> :-D
[17:55:48] <abaumann> Actually rustc is fine, just cargo has no support for rustc targets and features, but you can overwrite them via environment variables (I hope).
[18:27:23] <buildmaster> electron is broken (says eurobuild3).
[18:27:31] <abaumann> sorry. that was me.
[18:27:58] <abaumann> I said: ./build-packages -p librsvg and I got electron :-)
[18:28:12] <abaumann> maybe my slave was in a funny state..
[18:33:13] <buildmaster> librsvg is broken (says eurobuild3).
[18:33:21] <abaumann> yea.. also me :-(
[18:38:21] <buildmaster> deep42thought: my database is dirty again ...
[18:43:39] <girls> abaumann: nah, the buildmaster (currently) does not hand out a package a second time - even if you request it
[18:44:49] * buildmaster resumes sanity.
[19:36:13] -!- bill-auger has quit [Remote host closed the connection]
[19:40:25] -!- bill-auger has joined #archlinux-ports
[19:50:15] <abaumann> FAIL: rsvg-test 109 /rsvg-test/reftests/svg1.1/pservers-pattern-02-f.svg
[19:50:20] <abaumann> resiously *sigh*
[19:50:43] <abaumann> *seriously
[19:51:33] <girls> hmpf
[19:52:16] <abaumann> Should I report it upstream and remove checking for now?
[19:53:54] <girls> I think, you should
[19:54:27] <abaumann> ok.
[19:54:41] <buildmaster> girls: my database is dirty again ...
[19:55:17] <abaumann> * abaumann sniggers at deep42thought's current call sign.. :-)
[19:55:34] <girls> I thought It'd fit good to guys
[19:55:40] <abaumann> true :-)
[19:56:58] * buildmaster resumes sanity.
[20:07:21] -!- p71 has quit [Read error: No route to host]
[20:09:42] -!- p71 has joined #archlinux-ports
[20:17:00] -!- p71 has quit [Remote host closed the connection]
[20:18:21] <buildmaster> girls: my database is dirty again ...
[20:18:35] <abaumann> 2018-03-17 20:13:53: building package "librsvg" (revisions f694c1a002563e3eee166139d6237d39e9fa232b a9acbe65810c4b952291cc274ab3964ff314d3da, repository extra, straw :with_build_support:clean_chroot:) ... failed.
[20:18:39] <abaumann> Build master is not sane.
[20:18:41] <abaumann> "return-assignment" was running already.
[20:18:54] <abaumann> this happened immediately when something went wrong on my build slave. So I suspect it was me. :-(
[20:19:32] <girls> _what_ went wrong on your end?
[20:19:42] <girls> did you crash the script during upload?
[20:20:06] <abaumann> nae. it was in the middle of building and crashing due to a out-of-memory-of-sorts.
[20:20:58] <girls> the semi-clean state of the database looks, like it could be "your" fault :-)
[20:21:19] <abaumann> ha.. at least.. semi-clean.. :-)
[20:21:58] * buildmaster resumes sanity.
[20:22:13] <girls> well, your build slave was removed from the db but not from the state files
[20:22:29] <abaumann> classical two-phase-commit..
[20:22:42] <abaumann> ..which failed. :-)
[20:22:54] <girls> your build slave wanted to report an error?
[20:23:13] <abaumann> I think so.
[20:23:16] <abaumann> yes.
[20:23:21] <girls> ... it still wants
[20:23:26] <abaumann> huh?
[20:23:30] <girls> ah, no
[20:23:48] <girls> well, until 19:22:25 it tried to report the error
[20:24:02] <abaumann> mmh. weird.
[20:24:22] <abaumann> now I'm building librsvg just fine.
[20:27:09] <girls> ah, I think, the issue is again, that the hdd of the build master is full :-(
[20:27:37] <girls> /dev/vda1 9.9G 9.4G 43M 100% /
[20:27:52] <abaumann> oh.
[20:30:55] <girls> 1G residing in /var/log/journal
[20:31:06] <girls> I guess, that's some system-journal from systemd?
[20:33:05] <abaumann> yep. this you can tame in /etc/systemd/journald.conf.
[20:33:56] <abaumann> SystemMaxUse=100M
[20:34:12] <abaumann> sudo systemctl restart systemd-journald
[20:34:31] <abaumann> I ancient times this was a 'echo "" >/var/log/messages" :-)
[20:34:42] <abaumann> *In
[20:35:22] <girls> sudo journalctl --vacuum-size=100M
[20:35:24] <girls> :-D
[20:35:36] <abaumann> ah
[20:38:40] <girls> btw: do you have any idea what we do with wrongly detected "link" dependencies?
[20:39:02] <girls> e.g. "gitlab-10.5.5-1.0-i686 depends on ld-linux-x86-64.so.2"
[20:39:47] <abaumann> are they really wrongly detected?
[20:40:00] <abaumann> This looks more like some binary artifact uses a 64-bit dynamic loader..
[20:40:37] <abaumann> an exception list somewhere?
[20:40:41] <abaumann> to keep track..
[20:40:48] <girls> well, there _is_ a binary which links against that library
[20:41:05] <girls> but I think (hope), it is not needed
[20:41:10] <abaumann> ah.. but this binary could never run 32-bit..
[20:41:20] <abaumann> ..maybe we should find it and delete it in PKGBUILD?
[20:41:48] <girls> it's not a dependency declared in PKGBUILD
[20:41:55] <abaumann> no, it's detected.
[20:42:04] <girls> it's detected by an objdump-run on the files in the package
[20:42:06] <abaumann> but we could delete the offending binary in PKGBUILD with an rm..
[20:42:21] <abaumann> so that the script doesn't find it later on.
[20:42:47] <girls> well, we would need to define a special variable for that
[20:42:55] <girls> but we could do it
[20:43:11] <girls> ah, hmm, deleting would work, too
[20:43:23] <girls> I've also seen some freebsd binaries in virtualbox, I think
[20:43:35] <abaumann> the libc.so.7 ones. :-)
[20:43:46] <abaumann> this is sloppy packaging.. not much you can do.
[20:43:49] <girls> yes. I solved that problem by ignoring /opt
[20:43:54] <abaumann> maybe a variable is more transparent.
[20:44:10] <abaumann> doing tons of rm's in the middle of package() is not that readable..
[20:44:20] <girls> :-D
[20:44:40] <abaumann> librsvg is official fixed. :-)
[20:44:47] <girls> \o/
[20:45:14] <abaumann> geany and qemu both work again.
[20:45:21] <abaumann> We should push it to stable fast.
[20:45:30] <girls> gimme a sec
[20:45:39] * girls will break more stuff now :-)
[20:45:42] <abaumann> ah.. not that fast. :-)
[20:45:48] <girls> why not?
[20:45:51] <abaumann> it was broken now for some days..
[20:46:16] <girls> I reduced the haskell issues to ~100 "fast"
[20:46:31] <abaumann> :-)
[20:47:19] <girls> librsvg-2:2.42.3-1.1?
[20:47:35] <abaumann> yes.
[20:48:59] <girls> disk is full again
[20:49:07] <girls> how fast is 1G of data created???
[20:49:46] <abaumann> depends how.. but this means something is writting quite a lot..
[20:50:55] <girls> this vm has only 10G and rewbycraft is unresponsive :-(
[20:51:24] <abaumann> it's a VM I reckon?
[20:51:46] -!- p71 has joined #archlinux-ports
[20:52:25] <girls> huh?
[20:52:48] <girls> What do you mean by that?
[20:53:18] <abaumann> I thought, you might have access to a management panel to increase the size of the disk.
[20:53:52] <girls> nom I do not
[20:54:01] <girls> s/nom/no,/
[20:58:35] <girls> "No space left on device"
[20:58:56] <girls> :'-(
[20:59:22] <abaumann> abra kadabra space... didn't work..
[21:00:05] * buildmaster goes insane.
[21:06:06] <buildmaster> girls: my database is dirty again ...
[21:08:04] <abaumann> ok. Qt5 in QJSEngine has a 'SSE2 message' too.
[21:08:25] <abaumann> !wtf /usr/lib/libQt5Qml.so.5
[21:08:29] <phrik> abaumann: extra/qt5-declarative
[21:15:34] <girls> ah, /var/lib/mysql has 2.2 GB :-/
[21:15:54] <girls> I guess, mariadb desparately tries to save some data to the disk
[21:16:23] <abaumann> mmh. 10 GB is not much, given all the logs and packages and such.
[21:17:03] <girls> yes
[21:17:12] <girls> I'll write rewbycraft
[21:25:49] <abaumann> ok.
[21:25:54] <abaumann> cu
[21:25:56] -!- abaumann has quit [Quit: leaving]
[21:26:09] <girls> The question is what we do if he's busy currently.
[21:27:16] -!- deep42thought has parted #archlinux-ports
[23:04:33] -!- oaken-source has quit [Ping timeout: 264 seconds]