#archlinux-ports | Logs for 2017-12-28

[01:21:33] -!- jarshvor has quit [Quit: leaving]
[05:52:57] -!- hringriin_ has joined #archlinux-ports
[05:55:56] -!- hringriin has quit [Ping timeout: 260 seconds]
[05:55:56] hringriin_ is now known as hringriin
[06:27:10] -!- oaken-source has joined #archlinux-ports
[08:14:54] -!- isacdaavid has quit [Quit: Leaving.]
[09:04:05] -!- jelle has quit [Ping timeout: 240 seconds]
[09:41:42] -!- jelle has joined #archlinux-ports
[10:48:47] -!- abaumann has joined #archlinux-ports
[11:04:30] -!- deep42thought has joined #archlinux-ports
[11:06:13] <deep42thought> abaumann_: I removed the russian urls
[11:21:38] <abaumann> deep42thought: thanks. :-)
[11:30:14] <deep42thought> Did you read my comment on the forum yet?
[11:39:22] <abaumann> ok.
[11:39:34] <abaumann> you have a point here. :-)
[11:40:00] <abaumann> Maybe I was just unlucky, I had a comment with an image of somebody else, slipped and then opened the infected URL.
[11:40:07] <abaumann> Bad luck.
[11:40:49] <abaumann> The two URLs from yesterday were the important ones to take out. I didn't see any others..
[11:40:55] <deep42thought> but we could make a whitelist of urls which will get html'ed - all others will be plain text only
[11:41:23] <deep42thought> This would be reasonably convenient and should also be reasonably safe
[11:42:59] <abaumann> yep. agreed.
[11:43:16] <abaumann> unless we get a post to an infected BBS forum topic, that is. ;-)
[11:43:55] <deep42thought> then we should delete that topic anyway ;-)
[11:47:26] <abaumann> yep. :-)
[11:55:05] -!- deep42thought has quit [Quit: leaving]
[11:58:41] -!- abaumann has quit [Quit: leaving]
[13:00:06] -!- oaken-source has quit [Quit: Lost terminal]
[13:02:02] -!- oaken-source has joined #archlinux-ports
[13:14:15] -!- AndrevS has joined #archlinux-ports
[13:39:39] -!- deep42thought has joined #archlinux-ports
[13:56:46] -!- abaumann has joined #archlinux-ports
[13:58:00] <deep42thought> abaumann: I just noticed your fix for mongodb - sry, I didn't think of that, I simply blacklisted it, because wiredtiger did not build :-/
[13:59:41] <jelle> doesn't mongodb not support i686 ?
[14:00:08] <deep42thought> jelle: it was wiredtiger, a dependency of mongodb, which does not support i686
[14:00:22] <deep42thought> I didn't think of the possibility to build mongodb w/o wiredtiger
[14:00:37] <jelle> hehe
[14:00:47] <jelle> it's not that useful
[14:01:11] <deep42thought> which of both?
[14:16:30] <jelle> mongodb without WT :)
[14:16:35] <jelle> but hey, it's your project :)
[14:16:46] <deep42thought> I guess it's better than no mongodb at all?
[14:17:00] <deep42thought> what's wiredtiger used for in mongodb?
[14:20:00] <deep42thought> ah, ok I read a little something about it
[14:20:44] <jelle> it's one of the storage engines
[14:20:51] <deep42thought> yeah
[14:20:59] <deep42thought> but it's not available for 64 bit
[14:21:28] <jelle> uhu
[14:21:38] <jelle> we never had mongodb in the repos for 32 bit after some time
[14:26:14] <abaumann> hi. I thought about blacklisting mongodb alltogether..
[14:26:25] <deep42thought> it was
[14:26:42] <abaumann> ..currently I managed to get around wiredtiger, but I'm stuck in a missing 32-bit builder for gperftools.
[14:26:43] <deep42thought> indirectly, because it was depending on wiredtiger
[14:27:11] <abaumann> why mongodb builds half a Linux system in their third_party folder is their secret. :-)
[14:27:35] <abaumann> src/third_party/gperftools-2.5/src/tcmalloc.cc:90:10: fatal error: config.h: No such file or directory
[14:27:56] <deep42thought> funny
[14:28:06] <abaumann> it's missing a config.h for 32-bit Intel. I'm on it.
[14:28:21] <deep42thought> if upstream messes up releases, maybe _this_ is the easiest way to get around messed up dependencies?
[14:33:26] <jelle> abaumann: typical shit tier upstream
[14:34:01] <abaumann> sad.
[14:36:51] -!- deep42thought has quit [Ping timeout: 240 seconds]
[14:38:06] -!- abaumann has quit [Quit: leaving]
[14:50:10] <guys> deep42thought: I fix the not-redirected source url for python-paste and immediately you throw a check() error at me? :D
[16:29:01] -!- deep42thought has joined #archlinux-ports
[16:29:38] <deep42thought> guys: sry, I was just looking through the errors and reposrting all simple ones :-)
[16:29:52] <deep42thought> and by "simple" I mean "simple for me to report"
[16:29:57] <guys> :D :D
[16:30:25] <guys> also sorry about that osdlyrics one
[16:30:31] <guys> But at least we build now!
[16:31:34] <deep42thought> no problem - an "easy reportable" error is no big issue (for me)
[16:32:45] <deep42thought> btw: Merry Christmas to all you nerds, who hang around in irc at these days ;-)
[16:33:41] <AndrevS> Merry Christmas
[16:39:15] <rewbycraft> Merry christmas
[16:48:42] -!- sunshavi has quit [Ping timeout: 272 seconds]
[17:02:26] -!- abaumann has joined #archlinux-ports
[17:02:41] <abaumann> Merry Christmas :-)
[17:07:35] <deep42thought> gnaaaaa
[17:08:25] <deep42thought> eschwartz[m]: can you please delete FS#56885 - I didn't mean to open that one
[17:09:11] <deep42thought> hmm, or I could just append all the info to the bugreport, I want to put there anyway ...
[17:13:50] -!- sunshavi has joined #archlinux-ports
[17:15:51] <abaumann> deep42thought: update on Alix machine: https://bbs.archlinux32.org
[17:15:52] <phrik> Title: i486 motivation thread / Artwork, Screenshots & Setups / Arch Linux 32 Forums (at bbs.archlinux32.org)
[17:17:27] <deep42thought> ah nice
[17:17:44] <deep42thought> interestingly, debian failed to boot with an i686 kernel for me, but i586 works/worked
[17:19:21] <abaumann> I read something about this problem in some Archlinux forums: NOPL was the unsupported opcode.
[17:19:39] <abaumann> first you needed patched binutils/gcc and kernel to run on the Geodes.
[17:19:46] <abaumann> now it seems to be absolutely no problem.
[17:20:18] <abaumann> https://bbs.archlinux.org
[17:20:19] <deep42thought> so it might be, that some packages will fail to run?
[17:20:20] <phrik> Title: [SOLVED] Arch and AMD Geode LX 800 processor / Installation / Arch Linux Forums (at bbs.archlinux.org)
[17:21:44] <abaumann> yeah. no sse, sse2 and stuff.
[17:21:57] <abaumann> some i686 packages have some additional opcodes in them
[17:22:12] <abaumann> especially i386.deb, i386.rpm packages which are converted into Archlinux packages. :-)
[17:22:21] <abaumann> no problem if you compile from scratch usually.
[17:25:14] <deep42thought> is there a way to automatically check all packages for unwanted opcodes?
[17:26:20] <abaumann> mmh. not really.
[17:26:33] <deep42thought> if so, we should run this on all i586 packages
[17:26:42] <deep42thought> hmm
[17:26:47] <abaumann> and what happens, if libraries are multi-opcode-ish and switch on runtime whether they have sse or sse2?
[17:27:21] <deep42thought> ah, that would be bad
[17:27:28] <abaumann> usually a good indicator are march flags when building, assembly code files lying around, cmake tests vor SSE.
[17:27:28] <deep42thought> (for us to detect)
[17:27:32] <abaumann> s/vor/for/
[17:28:13] <abaumann> but definitely a nice idea, as for ELF text relocations: to have a architecture check of some sort..
[17:28:31] <deep42thought> btw: we have _lots_ of packages with text relocations
[17:28:49] <deep42thought> https://buildmaster.archlinux32.org
[17:28:51] <phrik> Title: packages with text relocations (at buildmaster.archlinux32.org)
[17:28:56] <abaumann> yeah. interpreters and emulators would be normal.
[17:30:23] <abaumann> libraries with video, sound
[17:30:34] <abaumann> chromium because of javascript?
[17:30:56] <abaumann> php.. well, because of php.. :-)
[17:34:26] <abaumann> mmh, I wonder, for a Compact flash I would like to have sort of an overlayroot at boot time.
[17:35:10] <deep42thought> can't one arrange something like this with a custom initrd?
[17:35:37] <deep42thought> although, this might even be possible w/o a initrd
[17:35:55] <deep42thought> you might need a separate /boot partition, though
[17:36:40] <abaumann> normally remounting the overlay is hard, but having a ramdisk should help
[17:37:22] <deep42thought> but overlay is a kernel feature, so it should not require a ramdisk
[17:37:32] <deep42thought> you only need to access the kernel, first
[17:38:01] <abaumann> right, but I need a mount binary from somewhere to do the overlay mount, and I may have to have tmpdirs or something.
[17:38:25] <abaumann> I'm doing that currently on an improvised isolinux-based CD for the i486 version
[17:38:30] <abaumann> without systemd.
[17:38:46] <abaumann> I'll play a little bit with options. :-)
[17:39:06] <deep42thought> you can put the mount binary into the kernel :-D
[17:39:17] <deep42thought> there's space for arbitrary stuff ;-)
[17:39:33] <deep42thought> ... usually used for drivers
[17:39:35] <abaumann> :-)
[17:40:07] <abaumann> I had quite a shock when I was first booting an Archlinux-486 without ramdisk and modules and systemd.
[17:40:11] <abaumann> It's so fast. :->
[17:40:58] <deep42thought> :-D
[17:41:27] <deep42thought> if it wasn't such a huge burden to maintain a systemd-less linux, I'd totally kick out systemd ;-)
[17:42:20] <abaumann> I think we will see a split here: embedded systems go simpler ways, desktop systems go the systemd way and servers go more the coreos/docker/lxc whatever way
[17:43:03] <abaumann> acutally having an etcd/systemd and lxc would be quite nice on servers..
[17:43:26] <abaumann> on embedded I prefer the BSD-script-way..
[17:45:04] <abaumann> so, enough hacking for today, cu.
[17:45:08] <deep42thought> cu
[17:45:20] -!- abaumann has quit [Quit: leaving]
[18:27:14] -!- deep42thought has quit [Quit: leaving]
[18:53:47] -!- isacdaavid has joined #archlinux-ports
[19:47:22] -!- because has quit [Ping timeout: 265 seconds]
[19:48:28] -!- because has joined #archlinux-ports
[19:52:57] -!- because has quit [Ping timeout: 240 seconds]
[19:55:33] -!- deep42thought has joined #archlinux-ports
[19:58:43] -!- because has joined #archlinux-ports
[20:25:35] -!- oaken-source has quit [Ping timeout: 240 seconds]
[21:29:54] -!- deep42thought has quit [Ping timeout: 272 seconds]
[21:31:25] -!- deep42thought has joined #archlinux-ports
[21:52:45] -!- deep42thought has quit [Quit: leaving]
[23:44:54] -!- AndrevS has quit [Remote host closed the connection]