#archlinux32 | Logs for 2020-01-09

[00:00:22] <girls> ... let's see :-D
[00:02:17] <girls> yep, it correctly installs x86_64 packages, now
[00:02:23] <girls> (on a pentium4 vm)
[00:02:33] <tyzoid> \o/
[00:03:05] <girls> but then we need to patch mkarchiso, too :-/
[00:03:10] <girls> ah, no
[00:03:12] <girls> we don't :-)
[00:03:31] <girls> it's all in releng-dual
[00:03:39] <girls> aahmm, archiso-dual
[00:05:08] <fALSO> gonna download it
[00:05:47] <fALSO> girls, which mirror
[00:05:59] <girls> https://mirror.archlinux32.org
[00:06:00] <phrik> Title: Index of / (at mirror.archlinux32.org)
[00:06:27] <fALSO> wgeting it
[00:11:02] <tyzoid> girls: Did you update the lastupdate file?
[00:11:13] <tyzoid> my mirror isn't getting the new iso
[00:11:14] <girls> probably not
[00:11:38] <girls> nope
[00:13:20] <girls> should be fixed now
[00:14:48] <tyzoid> indeed
[00:14:56] <fALSO> girls, booting, report in a while
[00:15:14] <tyzoid> BTW, it looks like your irc-logs latest symlink is specified with it's full path, not as a relative link
[00:15:28] <tyzoid> rsync is refusing to copy it
[00:15:48] <girls> yes, I fixed that already
[00:15:58] <girls> (saw issues on my local mirror a few hours ago)
[00:17:00] <tyzoid> hmm, rsync is still reporting it for me
[00:17:17] <girls> stupid me
[00:17:24] <girls> I need to restart the irc bot, too :-D
[00:17:30] <tyzoid> ah, lol
[00:17:58] -!- titus_livius has joined #archlinux32
[00:18:22] <girls> yay, it works
[00:18:29] <girls> luckily the other channel is so quiet :-D
[00:18:50] <tyzoid> yeah
[00:18:57] <tyzoid> Any reason we're still logging it?
[00:19:08] <girls> ehm, no one turned off the logger?
[00:19:20] <tyzoid> I say we, as if I'm still a project member lol
[00:19:25] <girls> you are
[00:19:31] <tyzoid> :)
[00:20:05] <girls> is your mirror back up?
[00:20:12] <tyzoid> Yup
[00:20:21] <tyzoid> came back along with everything else that went down
[00:20:27] <girls> so you'll get your place on our mirrorlist back ;-)
[00:20:41] <tyzoid> bbs will still respond if you set your hosts file to it
[00:21:04] <tyzoid> the server doesn't even know it went down, hardly
[00:21:09] <girls> I think we'll merge the databases somehow
[00:21:32] <tyzoid> IIRC the biggest issue will be the users table
[00:21:54] <tyzoid> did we want to merge the new stuff on top of the old forum? or the other way around?
[00:22:04] <girls> dunno
[00:22:09] <fALSO> girls, its installing. thanks!
[00:22:09] <tyzoid> The old bbs had many more users / posts on it
[00:22:17] <girls> fALSO: nice :-)
[00:22:26] <girls> I don't really care about lost users
[00:22:32] <girls> they can simply re-register
[00:22:37] <girls> I care more about lost content
[00:22:48] <tyzoid> problem with lost users is it won't map user ids across from the posts table
[00:22:52] <girls> the question is, whether the threads can be imported without the correct users
[00:22:54] <tyzoid> or will misattribute them
[00:22:59] <girls> yeah
[00:23:14] <tyzoid> Theoretically we could create a dummy user called 'import', and assign the IDs to them.
[00:23:14] <girls> we could null all the user ids of the recovered threads
[00:23:28] <girls> I think abaumann used "Admin" for that purpose ;-)
[00:23:58] <tyzoid> IMO I think it'd be easier to import the new bbs data to the old one, instead of trying to import the old data into the new one
[00:24:08] <tyzoid> because we can manually process the data
[00:24:13] <tyzoid> there's also the question of attachments
[00:27:19] <tyzoid> I also think it would be nice if we set up proper replication on that database too
[00:27:20] <girls> hmm, thinking about it, either way is fine for me, but I doubt, that any of both poses much less work than the other
[00:27:29] <girls> yeah
[00:27:31] <tyzoid> I know we had talked about something like that in the past
[00:27:43] <girls> there is a tinc network between all the core servers
[00:27:49] <girls> which makes authentication much easier
[00:28:13] <girls> you can basically pull stuff in a vpn-like network and not care about authentication/encryption
[00:28:49] <tyzoid> lol
[00:28:56] <tyzoid> corporate network blocks tinc's site from loading
[00:29:02] <girls> lol
[00:29:36] <tyzoid> although IMO I think auth is still important to get right, even in a LAN
[00:29:45] <tyzoid> because what happens if one of the hosts gets comprimised
[00:30:08] <girls> then some adversary can copy our databases?
[00:30:15] <girls> until we remove that host from tinc
[00:30:44] <tyzoid> potentially write access, if sql is happening in the clear
[00:30:53] <girls> hmm
[00:31:16] <girls> I believe, my setups are not too secure against compromises of single machines anyways
[00:31:27] <girls> I might have accidentally copied over the mysql user table or something
[00:33:06] <tyzoid> Btw, what revision of fluxbb are you using?
[00:33:11] <tyzoid> Did it include my security patches?
[00:33:26] <girls> dunno, abaumann set it up
[00:33:39] <tyzoid> gotcha
[00:35:15] <tyzoid> girls: The reason it might matter is because the security patches change how password handling is done
[00:35:36] <girls> ah, ok
[00:35:49] <tyzoid> It's backwards compatable, but the non-patched code isn't forwards-compatable with the patched code.
[00:36:07] <girls> gotcha
[00:36:13] <fALSO> damn the base install doesnt install a dhcp server?
[00:36:18] <fALSO> oops client
[00:36:29] * girls are unsure
[00:36:31] <fALSO> gotta chroot again
[00:36:50] <tyzoid> IIRC systemd-networkd has a dhcp client in it
[00:37:14] <fALSO> so probably i just need to configure something, no network on a clean boot
[00:37:26] <tyzoid> Did you enable networkd?
[00:37:41] <fALSO> gonna try it
[00:37:43] <fALSO> no i didnt
[00:38:24] <tyzoid> alternatively, you could manually `ip link set ... up` `ip addr add 192.168.xx.xx/24 dev xxxx` etc
[00:39:02] <tyzoid> https://wiki.archlinux.org
[00:39:04] <phrik> Title: systemd-networkd - ArchWiki (at wiki.archlinux.org)
[00:42:05] <tyzoid> girls: My build key is still listed in https://archlinux32.org even though it's expired
[00:42:17] <girls> yes
[00:42:24] <girls> you plan to extend it?
[00:42:27] <girls> or should I remove it?
[00:42:40] <fALSO> manned! thanks!
[00:43:07] <tyzoid> I keep forgetting you can just extend it without invalidating the signatures
[00:43:10] <tyzoid> let's keep it around then
[00:43:21] <tyzoid> I was going to generate a new one, but I can just update the existing key
[00:43:26] <girls> fALSO: np :-)
[00:45:50] <tyzoid> girls: Do you still have the revocation certificate for my arch32 master key?
[00:45:56] <girls> yes
[00:46:08] <tyzoid> Ok, just double-checking
[01:07:53] <eschwartz> tyzoid: omg you're alive
[01:08:03] <tyzoid> barely lol
[01:08:16] <tyzoid> but yes, I'm back
[01:08:44] <girls> hmm, sudo in build() is probably a bad idea :-(
[01:08:46] <tyzoid> eschwartz: Root hdd failed on my primary server, and I was too busy with life to deal with it until holiday break
[01:08:52] <tyzoid> girls: I'd say so lol
[01:09:08] <eschwartz> girls: who's doing sudo in build()...
[01:09:13] <girls> I will
[01:09:24] <girls> to run the archiso's build.sh
[01:09:25] <tyzoid> eschwartz: we're trying to see if we can't get mkarchiso to run within a PKGBUILD
[01:09:46] <eschwartz> that sounds exciting, in the bad way o_O
[01:09:53] <tyzoid> girls: can't we create a fakeroot with fake sudo perms?
[01:09:58] <eschwartz> why a PKGBUILD, specifically?
[01:10:07] <girls> I'll try that after the actual sudo
[01:10:15] <tyzoid> eschwartz: Because the arch32 build slaves have provisions to schedule building pkgbuilds
[01:10:18] <girls> because a PKGBUILD can be distributed by our build system
[01:10:21] <tyzoid> we were seeing if we could leverage that system
[01:10:38] <eschwartz> lol, that's... one way of doing it...
[01:10:49] <tyzoid> That would allow us to schedule building the isos + gives us the automatic retry mechanism on failure
[01:11:08] <eschwartz> I'm very conflicted about this idea.
[01:11:15] <girls> and releasing a new iso would be as easy as bumping the pkgver in git
[01:11:22] <tyzoid> As am I, and I'm the one who came up with the idea
[01:11:33] <girls> currently, I'm just experimenting
[01:11:59] <eschwartz> Although, there's nothing to say it wouldn't technically work as long as the build slave's 'builduser' user had a sudoers.d dropin allowing it to run build.sh
[01:12:04] <tyzoid> anyway eschwartz, I finally passed my private pilot's exam / checkride
[01:12:12] <eschwartz> tyzoid: cool!
[01:12:16] <girls> nice :-)
[01:12:31] <tyzoid> That was back in novemberish
[01:12:42] <girls> and you've been in the air since then ;-)
[01:12:53] <tyzoid> yup
[01:13:05] <girls> it all makes sense, now :-D
[01:13:06] <tyzoid> Only problem is winter weather
[01:13:33] <tyzoid> so I'm grounded by weather for the time being, as you can imagine
[01:13:41] <girls> yeah
[01:13:48] <eschwartz> have you all considered switching to zstd packages yet?
[01:13:55] <girls> we did
[01:14:13] <girls> not only did we consider it, we're compressing as zst currently
[01:14:23] <tyzoid> lol
[01:16:40] <girls> I intended to make the build scripts compression-agnostic, but I finally gave up and simply replaced all "xz" by "zst" ... well, almost all ...
[01:17:38] <eschwartz> :)
[01:17:44] <eschwartz> nice
[01:18:01] <eschwartz> oh BTW new change was just deployed for python
[01:18:03] <girls> fakeroot is insufficient
[01:18:16] <eschwartz> python is in the hot path for makechrootpkg due to namcap
[01:18:27] <girls> and also the sudo variant fails with "mount: work/efiboot: mount failed: Operation not permitted."
[01:18:31] <eschwartz> and it's now 50% smaller, installed
[01:18:44] <girls> eschwartz: what change do you refer to?
[01:18:46] <eschwartz> https://git.archlinux.org
[01:18:47] <phrik> Title: svntogit/packages.git - Git clone of the 'packages' repository (at git.archlinux.org)
[01:19:06] <eschwartz> (and -3 to fix a bug in the depends for the -2 release)
[01:19:29] <eschwartz> it's only 88 minutes old
[01:20:12] <tyzoid> I didn't realize a pkgbuild can specify the build of more than 1 package
[01:20:15] <girls> ooooh, so python will no longer bring its whole testsuite? nice :-)
[01:20:22] <girls> split package?
[01:20:27] <girls> it's pretty common
[01:20:39] <tyzoid> I always thought those were split into two separate pkgbuilds
[01:21:24] <eschwartz> tyzoid: nope, that would be sort of painful to keep in sync, and also kind of unsafe for things like gcc/gcc-libs
[01:21:24] <girls> that's why we have "build assignments" separated from "binary packages"
[01:21:41] <girls> one of the former can build multiple of the latter
[01:21:50] <girls> gotta catch some sleep now
[01:21:55] <girls> cu tomorrow
[01:21:59] <eschwartz> I'm not sure you can actually build gcc twice and use the libs of one with the bins of another
[01:22:21] <tyzoid> Does gcc not support reproducible builds on itself?
[01:22:24] <eschwartz> well... reproducible builds says you can
[01:22:57] <eschwartz> but reproducible builds is not nearly as old as the split package
[01:23:27] <eschwartz> and also, it's possible to build it without the reproducible bits, so having the package completely break if reproducible builds breaks is, well...
[01:23:36] <eschwartz> I guess in a way you could call that a feature.
[01:24:26] <tyzoid> Although if you hit a break there, it'd break in the testing repo?
[01:24:45] <eschwartz> it's still better practice to use one PKGBUILD
[01:24:51] <tyzoid> agreed
[01:25:04] <eschwartz> and it avoids issues where "I forgot to bump one of them when I bumped the other"
[01:25:12] <eschwartz> (that would be *really* bad)
[01:25:33] <eschwartz> same goes, really, for lots of other packages
[01:26:00] <eschwartz> you can build one python module for both python2 and python3 without bumping version numbers and adding patches twice.
[02:16:34] -!- MrBIOS has quit [Quit: MrBIOS]
[03:01:49] -!- MrBIOS has joined #archlinux32
[04:36:29] <buildmaster> i686/iempluginsuite is broken (says rechenknecht): https://archlinux32.org
[05:02:26] -!- MrBIOS has quit [Quit: MrBIOS]
[05:02:57] -!- MrBIOS has joined #archlinux32
[05:03:12] -!- MrBIOS has quit [Client Quit]
[05:03:43] -!- MrBIOS has joined #archlinux32
[05:03:58] -!- MrBIOS has quit [Client Quit]
[05:04:24] -!- MrBIOS has joined #archlinux32
[05:04:46] -!- MrBIOS has quit [Client Quit]
[05:05:17] -!- MrBIOS has joined #archlinux32
[05:05:33] -!- MrBIOS has quit [Client Quit]
[05:06:08] -!- MrBIOS has joined #archlinux32
[05:06:21] -!- MrBIOS has quit [Client Quit]
[05:06:52] -!- MrBIOS has joined #archlinux32
[05:07:08] -!- MrBIOS has quit [Client Quit]
[05:07:34] -!- MrBIOS has joined #archlinux32
[05:07:56] -!- MrBIOS has quit [Client Quit]
[05:08:27] -!- MrBIOS has joined #archlinux32
[05:08:44] -!- MrBIOS has quit [Client Quit]
[05:09:28] -!- MrBIOS has joined #archlinux32
[05:09:31] -!- MrBIOS has quit [Client Quit]
[05:10:11] -!- MrBIOS has joined #archlinux32
[05:10:19] -!- MrBIOS has quit [Client Quit]
[05:10:51] -!- MrBIOS has joined #archlinux32
[05:11:05] -!- MrBIOS has quit [Client Quit]
[05:11:48] -!- MrBIOS has joined #archlinux32
[05:11:52] -!- MrBIOS has quit [Client Quit]
[05:12:21] -!- MrBIOS has joined #archlinux32
[05:12:40] -!- MrBIOS has quit [Client Quit]
[05:13:11] -!- MrBIOS has joined #archlinux32
[05:13:26] -!- MrBIOS has quit [Client Quit]
[06:28:59] <buildmaster> any/ruby-term-ansicolor is broken (says eurobuild6-4): https://archlinux32.org
[06:31:18] <buildmaster> pentium4/bandwhich is broken (says eurobuild6-2): https://archlinux32.org
[06:33:28] <buildmaster> i686/bandwhich is broken (says nlopc46): https://archlinux32.org
[06:50:03] -!- abaumann has joined #archlinux32
[06:50:04] <buildmaster> Hi abaumann!
[06:50:04] <buildmaster> !rq abaumann
[06:50:05] <phrik> buildmaster: <abaumann> well.. it's a mysql.. almost any amount of data is too much for it too handle.. so it shouldn't matter too much..
[06:52:56] -!- deep42thought1 has joined #archlinux32
[06:53:05] <deep42thought1> good morning abaumann
[06:54:11] <abaumann> morning deep42thought1 (-1)
[06:54:15] <deep42thought1> lol
[06:54:31] <girls> tyzoid: can you remove my account from your bouncer, pleace? I'm now three times registered with irc - that's one personality too much for my taste ;-)
[06:54:32] <abaumann> the real deep42thought is occupied (in the IRC sense)
[06:54:44] <deep42thought1> yeah, he is
[06:55:00] <abaumann> we have heavy ruby breaks..
[06:55:07] <abaumann> ..as if anybody would use it ;-)
[06:55:49] <abaumann> something is wrong in rubygems and I'm definitely not a ruby-guy..
[06:57:15] <abaumann> about the bug reporting and bbs system. the current version is checked in as https://git.archlinux32.org and https://git.archlinux32.org
[06:57:16] <phrik> Title: flyspray - Archlinux32 customized Flyspray installation (at git.archlinux32.org)
[06:57:32] <abaumann> so if there are security fixes I missed when setting it up, they can be added there again
[06:58:24] <abaumann> I'm also not sure what to do with the now two forums and two bug reporting systems
[06:58:37] <abaumann> my two cents: the content is more important to preserve than the user
[06:59:04] <abaumann> an extra bonus is if ids known in public (for instance a bbs forum entry references somewhere else) could again point to the right place.
[06:59:17] <abaumann> for bugs this is true, there we only have to move maybe some missing bugs
[06:59:35] <abaumann> simplest option is to keep the old forum as read-only version around
[07:00:20] <abaumann> I'm not willing to spend a lot of effort to join old users and forums
[07:00:49] <abaumann> the work is much better spent in improving the build system and keep the packages building
[07:01:57] <abaumann> we should also discuss the reasons, why we enforce now redundant infrastructure (tinc, alpha, beta, gamma) etc.
[07:04:30] <deep42thought1> I don't understand your last point
[07:04:36] <deep42thought1> but I agree with the rest
[07:04:54] <abaumann> well, we have alpha. beta gamma etc. now
[07:05:02] <abaumann> and the services can run on any of those nodes.
[07:05:05] <abaumann> that's what I meant.
[07:05:18] <abaumann> the problem is currently: defacto they run again on only one node :-)
[07:05:21] <abaumann> aka the buildmaster
[07:06:44] <abaumann> IMHO it's enough if we can easily spin up another bbs or bugrepoting, if the current active one goes down.
[07:12:14] <deep42thought1> ah, ok
[07:12:29] <deep42thought1> I was confused, because you wrote "we should discuss the reasons ..."
[07:12:47] <abaumann> my German-English, sorry.
[07:12:57] <deep42thought1> np
[07:12:58] <abaumann> our goals.
[07:13:31] <abaumann> btw: one hard disk on Hetzner
[07:13:35] <abaumann> (buildmaster) is shacky
[07:13:47] <deep42thought1> will they replace it?
[07:13:51] <abaumann> and the backup raid at my place also has a shacky disk.
[07:13:53] <abaumann> they will.
[07:14:04] <abaumann> they did already replace the other Seagate :->
[07:14:19] * deep42thought1 knows the feeling of replacing Seagates en masse
[07:14:24] <abaumann> for my backup I fear I have to do it. :-)
[07:14:45] <deep42thought1> my backup is currently also dead, because I f*cked up - "EXPERIMENTAL" in the kernel means, that, well, it's experimental and might not work as expected
[07:14:46] * abaumann thinks Seagate will now not sponsor the project anymore ;-)
[07:14:53] <deep42thought1> lol
[07:15:14] <abaumann> experimental and RAID feature.. sounds fun. :-)
[07:17:32] <abaumann> mmh. live migration of a non-RAID to RAID-1 seems only possible when copying the data..
[07:17:44] <deep42thought1> well, the experimental part is beneath the raid - but it does not help, if all components of the raid suffer the same issue :-(
[07:17:45] <abaumann> .. I didn't think much when setting up that backup partition of mine.
[07:17:58] <abaumann> true
[07:18:25] <abaumann> mmh. a WD disk has a bad LBA (non-read) block beyond the addressable area of sectors? how comes..
[07:18:34] <abaumann> says smartctl
[07:18:52] <abaumann> I have seen firmware bugs before in consumer harddisks, so..
[07:19:19] <abaumann> ..I shouldn't have started monitoring SMART in nagios.. ;-)
[07:20:47] <abaumann> tyzoid: btw, nice to see you are back and alive. :-)
[07:23:53] <tyzoid> nice to be back abaumann
[07:24:19] <tyzoid> deep42thought1: my znc server isn't showing any active clients from your account, but I've gone in and disabled it
[07:24:29] deep42thought1 is now known as deep42thought
[07:24:33] <deep42thought> thank you :-)
[07:24:45] <deep42thought> tyzoid: I set up my own znc after yours went down
[07:24:52] <tyzoid> gotcha
[07:25:12] <tyzoid> Really useful to have
[07:25:32] <tyzoid> I mainly use mine to have a nice webchat view that's not the default crappy freenode one
[07:25:44] <tyzoid> lets me use IRC behind the corporate firewall
[07:25:59] <abaumann> btw, I also evacuated the i486 build slave from your srv0/1 machine (?) to my new machine (eurobuild6-7-i486)
[07:26:11] <tyzoid> Gotcha
[07:26:16] <abaumann> Since April we have a AMD beast doing quite some build work (eurobuild6-*)
[07:26:17] <tyzoid> I noticed it was powered down.
[07:26:22] <tyzoid> I assume that was intentional
[07:26:30] <abaumann> yes, I borked it at some time. :-)
[07:27:00] <tyzoid> IIRC you still have access to the proxmox manager to power it back on if you needed
[07:27:04] <abaumann> and we have currently enough i486 build slaves as not many packages build anyway..
[07:27:13] <abaumann> ..yeah, I think I do.
[07:27:27] <abaumann> but if you need it for something else, I don't mind if you delete it.
[07:27:35] <tyzoid> I'm not strapped for disk space, so I'll probably keep it around regardless
[07:27:48] <abaumann> as a backup slave, yes.
[07:28:02] <deep42thought> "backup slave" ...
[07:28:24] <deep42thought> :-D
[07:28:35] <abaumann> ahem.. yes. the wording is wrong in so many ways.
[07:28:39] <tyzoid> reminds me of the reddit post where I was trying to explain the "forking a child" and "parent killing itself" terminology to non-CS people
[07:28:53] <abaumann> lol
[07:29:06] <deep42thought> :-D
[07:29:15] <tyzoid> oh, abaumann: we had a question regarding the new BBS setup
[07:29:22] <abaumann> yes?
[07:29:22] <tyzoid> What version of fluxbb did you use?
[07:29:24] <deep42thought> you know the "pedophile has no children to watch" thing?
[07:29:47] <abaumann> that's a very good question
[07:29:51] <tyzoid> Was it the version that included the security patches I added?
[07:30:04] <abaumann> 3b06ee0d381dc1be5f40ca98ad4278046d869d21: checked in initial customized verison for Archlinux32
[07:30:05] <tyzoid> I had upstreamed those, but I think they were on a floating branch that was only merged in the next major release
[07:30:07] <abaumann> mmh.
[07:31:06] <abaumann> I think, I took upstream and just did some Archlinux32 adaptions
[07:31:26] <abaumann> I didn't find your setup anywhere in a git or so :->
[07:31:59] * abaumann makes a diff of fluxbb directories..
[07:32:07] <tyzoid> abaumann: https://github.com
[07:32:09] <phrik> Title: Commits · tyzoid/fluxbb · GitHub (at github.com)
[07:32:49] <abaumann> ah. didn't look there.
[07:32:59] <deep42thought> good ol' github
[07:33:04] <tyzoid> Fixed a security issue with password hashing https://fluxbb.org
[07:33:05] <phrik> Title: Ticket 1069: Stronger password hashes (and salts) - FluxBB core - FluxBB (at fluxbb.org)
[07:33:10] <tyzoid> uses bcrypt instead of sha1
[07:33:28] <abaumann> yeah. concerning the forum we are in pickle now anyway.
[07:33:58] <abaumann> so, the passwords of old users are bcrypt and of new users are sha1
[07:34:09] <tyzoid> I was suggesting to deep42thought that since my service is back up, if we take a db dump of the new bbs, I can manually migrate the few posts and users that have been made since
[07:34:20] <tyzoid> i.e. migrate the new data back to the old bbs
[07:34:32] <abaumann> sounds easier than the other way round
[07:34:50] <abaumann> and it has the bonus of keeping ids of articles references somewhere else intact
[07:35:22] <tyzoid> Once the data is migrated back we can set up a database replication so we don't have a single point of failure again™
[07:35:24] <abaumann> for us with you not answering to emails or such we had to make a hard decision and just set up a new forum.
[07:35:40] <abaumann> yep.
[07:35:46] <tyzoid> Not sure if you saw the IRC messages when I hopped back on
[07:35:58] <tyzoid> root hdd failure on proxmox took down my entire hypervisor
[07:36:01] <tyzoid> which included my mailserver
[07:36:05] <tyzoid> :/
[07:36:20] <abaumann> yeah, that's a single point of failure :-)
[07:36:34] <abaumann> no raid on that machine?
[07:36:43] <abaumann> or just too many hdd failures?
[07:36:46] <tyzoid> There was, but it was failing to boot
[07:36:53] <abaumann> oh.
[07:36:57] <tyzoid> I didn't have time to get in with a recovery media until holiday breka
[07:36:59] <tyzoid> break*
[07:37:13] <tyzoid> because fires at work
[07:37:23] <tyzoid> and with me getting my pilot's license
[07:37:35] <abaumann> that's absolutely no problem
[07:37:58] <tyzoid> Still my fault I didn't pop in via freenode's webchat and say something about it :/
[07:38:13] <abaumann> but, we could have git, the forum, bugreporting and mail down for three months, because that would have basically ended the Archlinux32 project
[07:38:45] <abaumann> I think, the plan with alpha, beta, gamma nodes, the TINC network and frequent backups should solve this problem :-)
[07:38:50] <abaumann> so no worries.
[07:39:06] <tyzoid> Yeah.
[07:39:14] <abaumann> also companies (like for instance mine) don't get the point of redundand infrastructure..
[07:39:23] <abaumann> ..and more important: redundant people.
[07:39:37] <abaumann> administration is primarily a thing of documenting what you are doing.
[07:39:50] <abaumann> so somebody else can easily step in
[07:40:25] <abaumann> so, you can fly now yourself to your broken server, when a hard disk fails ;-)
[07:40:35] <tyzoid> lol yup
[07:40:38] <deep42thought> lol
[07:40:44] <tyzoid> So w.r.t the webserver I have running, the old bbs is still online if you set a record in /etc/hosts
[07:40:58] <tyzoid> deep42thougth has root shell access to that box, not sure if you wanted shell access too
[07:41:55] <tyzoid> How did we want to handle the migration, then? Did we want to keep the primary web server where it is? or migrate back to my VM?
[07:43:07] <tyzoid> Regardless, we'll likely want to add in those security patches, because while the old database is backwards compatible with the new software version, the reverse is not true
[07:43:15] <abaumann> aeh. do you have access to the buildmaster where things are now? so you could get a picture, of what I did.
[07:43:28] <tyzoid> I don't think I do
[07:43:50] <abaumann> right, that machine is also newer (April 2019 or so)
[07:44:33] <abaumann> one important difference is for instance the bug reporting: I allow users to register themselves.
[07:44:46] <abaumann> Also the main mailserver is now on mail3.archlinux32.org (the buildmaster)
[07:45:45] <tyzoid> Yeah, the original idea was to tie the bug reporting database in to use the bbs user database
[07:45:46] <abaumann> fluxxbb I added a recaptcha and a funny question
[07:45:49] <tyzoid> but that never really materialized
[07:46:28] <abaumann> we didn't have fake accounts anymore (or almost none, maybe 2-3 on the forum)
[07:47:10] <abaumann> fluxxbb in git is basically fluxbb-1.5.11 with an Archlinx, Archlinux32 theme.
[07:47:29] <abaumann> with the AP_reCAPTCHA plugin
[07:50:56] <abaumann> yeah. nothing speaks against moving services away from the buildmaster (as it was before), as the buildmaster should mainly do - well - buildmastering stuff. :-)
[07:51:23] <abaumann> I would just keep there a backup version in case which can be fast activated, if the need arises.
[07:52:34] <abaumann> there is another service: git is no git+cgit only, no more gitea
[07:53:03] <abaumann> I just didn't have time to set it up again
[07:54:23] <tyzoid> Yeah, no worries
[07:54:35] <tyzoid> The weird one was the git should have been on the unaffected hypervisor
[07:54:39] <tyzoid> because I've got two servers
[07:54:41] <tyzoid> srv0 and srv1
[07:55:01] <abaumann> ah. another thing: https://buildmaster.archlinux32.org
[07:55:26] <abaumann> I started monitoring differntly, this funny status page was always showing green-all ok, though services were all down :-)
[07:55:48] <tyzoid> nagios requires HTTP auth
[07:55:53] <abaumann> https://buildmaster.archlinux32.org also to see the performance or potential issues with the buildmaster
[07:55:58] <abaumann> yeah.
[07:56:00] <abaumann> don't care
[07:56:04] <abaumann> ah.
[07:56:05] <abaumann> sorry.
[07:56:12] <abaumann> you don't have an account there either.
[07:56:16] <abaumann> munin is open, nagios not.
[07:56:18] <abaumann> forgot.
[07:56:29] <tyzoid> lol, dw about it
[07:56:44] <tyzoid> Anyway, going to pop off in a min here, it's 2am
[07:56:52] <abaumann> oh. right. :-)
[07:57:01] <abaumann> cu later then
[07:57:08] <tyzoid> I've still got the gitea git server if we wanted to migrate back to that
[07:57:22] <tyzoid> or I can just give a dump of the server and we can go from there
[07:58:09] <tyzoid> but regardless, we should come up with a migration plan / strategy
[07:58:21] <abaumann> true.
[07:59:35] <tyzoid> Did you see the terrible idea I gave deep42thought regarding iso building?
[07:59:56] <deep42thought> ideas don't need to be good - you just need to sort out all the bad ones ;-)
[08:00:04] <tyzoid> lol
[08:00:19] <tyzoid> first day back and already causing trouble
[08:01:05] <tyzoid> Regardless, god night
[08:01:13] <abaumann> np
[08:14:33] <deep42thought> trouble? you did not cause me any trouble (by coming back) - maybe only some grey hair by disappearing ;-)
[08:16:10] <abaumann> lucky me I cannot grow more grey hair than what I already have.. ;-)
[08:18:04] <deep42thought> :-D
[08:19:11] <deep42thought> gotta go
[08:19:12] <deep42thought> cu later
[08:19:16] -!- deep42thought has quit [Quit: Leaving.]
[08:23:26] <abaumann> cu
[08:23:30] -!- abaumann has quit [Quit: leaving]
[09:19:42] <buildmaster> any/ruby-rack-protection is broken (says eurobuild6-2): https://archlinux32.org
[09:20:31] <buildmaster> i486/passenger is broken (says nlopc46-i486bs1): https://archlinux32.org
[09:21:21] <buildmaster> i686/passenger is broken (says eurobuild6-1): https://archlinux32.org
[09:22:14] <buildmaster> pentium4/passenger is broken (says eurobuild6-5): https://archlinux32.org
[09:24:14] -!- buildmaster has quit [Remote host closed the connection]
[09:24:27] -!- buildmaster has joined #archlinux32
[09:24:27] <buildmaster> !rq buildmaster
[09:24:27] <phrik> buildmaster: <buildmaster> I might be insane, but never confused ... ;-)
[09:27:31] <T`aZ> you guys rock, i said it already but anyway, thanks :p
[11:21:16] -!- deep42thought has joined #archlinux32
[11:21:16] <buildmaster> Hi deep42thought!
[11:21:16] <buildmaster> !rq deep42thought
[11:21:17] <phrik> buildmaster: <deep42thought> !grab abaumann
[11:24:32] -!- abaumann has joined #archlinux32
[11:24:33] <buildmaster> Hi abaumann!
[11:24:33] <buildmaster> !rq abaumann
[11:24:33] <phrik> buildmaster: <abaumann> wait... go-go-gadgeto-i686-test-machine
[11:24:53] <abaumann> I retriggered ruby, there was a build issue around rubygems failing all other packages..
[11:25:03] <abaumann> ..just rebuilding ruby, then rubbygems in order helped.
[11:25:06] <abaumann> now I get:
[11:25:06] <abaumann> Neither PKGBUILD nor modification of PKGBUILD found for package "ruby-celluloid" from community (community), revisions 03b21eedc7f143af9ebb03f048e847a6d3d28c96 and c18c36faf78b080bd37ca1ad0ac8e28d20f6e627.
[11:25:11] <abaumann> "make_source_info ruby-celluloid community 03b21eedc7f143af9ebb03f048e847a6d3d28c96 c18c36faf78b080bd37ca1ad0ac8e28d20f6e627 /tmp/tmp.mysql-functions.mysql_generate_package_metadata.gaFSdKA2kM/SRCINFO" failed.
[11:25:14] <abaumann> so again some upstream issue?
[11:26:05] <abaumann> anyways.. I hope things sort out themselves now..
[11:26:31] * abaumann dashes to eurobuild3 and replaces a hard disk (this time a WD)
[11:36:11] <deep42thought> strange, the ruby-celluloid looks all right
[11:36:16] <deep42thought> but make_source_info fails
[11:46:28] <abaumann> asp32 export ruby-celluloid doesn't export anything.. maybe it disappeared upstream?
[11:46:33] <abaumann> https://www.archlinux.org
[11:46:34] <phrik> Title: Arch Linux - Package Search (at www.archlinux.org)
[11:47:55] <deep42thought> I'm on it, but I'm doing some needle-work in parallel
[11:48:06] <abaumann> no stress. :-)
[11:48:31] <abaumann> I just wanted to report it basically, so it doesn't get lost.
[11:48:38] <deep42thought> commit db478fad07656332edbdf42e7eb2b70aafeebc08 has it
[11:48:43] <deep42thought> commit 03b21eedc7f143af9ebb03f048e847a6d3d28c96 does not
[11:48:57] <deep42thought> but git diff does not show any relevant thing between those :-/
[11:48:58] <deep42thought> strange
[11:49:21] <abaumann> indeed strange. packages upstream shows nothing..
[11:49:40] * abaumann peeks into the git repos
[11:50:00] <deep42thought> ah, typo
[11:50:05] <deep42thought> I wrote "ruby-celuloid"
[11:50:36] <deep42thought> ok, so it was removed
[11:50:47] <abaumann> there might be dozens of sub-packages
[11:55:23] <abaumann> I wonder, ruby-celluloid just disappeared, also nothing has been dropped to the AUR..
[11:56:19] <abaumann> thanks for fixing.. I'll retrigger the ruby-modules to build..
[12:01:24] <buildmaster> pentium4/logstash is broken (says eurobuild6-6): https://archlinux32.org
[12:01:32] <buildmaster> i686/logstash is broken (says eurobuild6-5): https://archlinux32.org
[12:06:22] <buildmaster> any/ruby-activesupport is broken (says nlopc46): https://archlinux32.org
[12:11:54] <abaumann> ./seed-build-list -w -p '^ruby-' && ./prioritize-build-list -d -w <(printf '^ruby-\n') gives me again:
[12:11:57] <abaumann> "make_source_info ruby-celluloid community 14186c6e3f7def7ce7f0da65cbfd3364a8854db3 c18c36faf78b080bd37ca1ad0ac8e28d20f6e627 /tmp/tmp.mysql-functions.mysql_generate_package_metadata.oYyTqF1TVK/SRCINFO" failed.
[12:12:04] <abaumann> so, it gets onto the list again with seeding?
[12:12:19] * abaumann heads to his zucchini and removes the already black ones..
[12:15:08] <deep42thought> hmm, that looks like a bug (on my end)
[12:15:25] <abaumann> easy, I schedule everuthing but the celluloids..
[12:15:48] <deep42thought> there's -i
[12:16:27] * abaumann always forgets about this option (literally an option)
[12:16:36] <deep42thought> :-D
[12:17:09] <deep42thought> yeah, it gets rescheduled, because it's known to the databse
[12:17:20] <deep42thought> should we restrict that to packages already built or on the build-list?
[12:17:36] <abaumann> mmh..
[12:18:05] <abaumann> it's actually a mistake to reseed things blindly when they are known no longer to exist.
[12:18:17] <abaumann> so usually that shouldn't happen
[12:18:23] <deep42thought> well, but it should not try to add sources of packages which do not exist
[12:18:39] <deep42thought> you can pass -f, though
[12:18:42] <abaumann> that's also true
[12:19:05] <abaumann> yeah. as long there is nothing broken now in the database things are fine.
[12:19:34] <buildmaster> any/ruby-activesupport is broken (says nlopc46): https://archlinux32.org
[12:19:47] <deep42thought> ok, consider it a feature, then
[12:19:57] <abaumann> :-)
[12:20:09] <deep42thought> one can reschedule everything known to the database - but if there is no source available, it will obviously fail :-D
[12:21:10] <abaumann> oups. I also rescheduled ruby itself..
[12:21:15] <deep42thought> :-D
[12:21:59] <abaumann> my backup disk is fine again.. I had to manually relocate bad sectors.
[12:22:07] <abaumann> now I'll add a RAID 1 there too.
[12:22:35] <abaumann> then I should have RAIDs everywhere
[12:29:14] <buildmaster> pentium4/mruby is broken (says eurobuild6-4): https://archlinux32.org
[12:29:15] <buildmaster> any/ruby-chronic is broken (says eurobuild6-3): https://archlinux32.org
[12:30:13] <buildmaster> pentium4/ruby-cairo is broken (says nlopc46): https://archlinux32.org
[12:38:14] <buildmaster> i686/ruby-cairo is broken (says eurobuild6-1): https://archlinux32.org
[12:38:52] <buildmaster> i686/mruby is broken (says rechenknecht): https://archlinux32.org
[12:47:02] <buildmaster> i486/ruby-cairo is broken (says nlopc46-i486bs1): https://archlinux32.org
[12:54:28] <buildmaster> any/ruby-hoe is broken (says rechenknecht): https://archlinux32.org
[12:56:05] <buildmaster> i686/ruby-json is broken (says eurobuild6-1): https://archlinux32.org
[12:58:05] -!- buildmaster has quit [Remote host closed the connection]
[12:58:18] -!- buildmaster has joined #archlinux32
[12:58:18] <buildmaster> !rq buildmaster
[12:58:18] <phrik> buildmaster: <buildmaster> I might be insane, but never confused ... ;-)
[12:58:34] <buildmaster> pentium4/ruby-json is broken (says eurobuild6-2): https://archlinux32.org
[13:00:05] <buildmaster> any/ruby-kramdown is broken (says rechenknecht): https://archlinux32.org
[13:03:29] -!- abaumann has quit [Quit: leaving]
[13:05:50] <buildmaster> any/ruby-mini_portile is broken (says rechenknecht): https://archlinux32.org
[13:07:40] <buildmaster> i486/ruby-json is broken (says eurobuild6-7-i486): https://archlinux32.org
[13:09:33] <buildmaster> any/ruby-minitest is broken (says eurobuild6-1): https://archlinux32.org
[13:09:34] <buildmaster> pentium4/ruby-gpgme is broken (says eurobuild6-3): https://archlinux32.org
[13:09:38] <buildmaster> any/ruby-mocha is broken (says eurobuild6-4) - already flagged out-of-date upstream on 2019-12-15: https://archlinux32.org
[13:11:33] <buildmaster> i486/ruby-gpgme is broken (says nlopc46-i486bs1): https://archlinux32.org
[13:12:08] <buildmaster> i686/ruby-msgpack is broken (says eurobuild6-5): https://archlinux32.org
[13:12:10] <buildmaster> i486/ruby-msgpack is broken (says eurobuild6-7-i486): https://archlinux32.org
[13:13:02] <buildmaster> pentium4/ruby-msgpack is broken (says nlopc46): https://archlinux32.org
[13:13:41] <buildmaster> i686/ruby-gpgme is broken (says rechenknecht): https://archlinux32.org
[13:15:41] -!- buildmaster has quit [Remote host closed the connection]
[13:15:53] -!- buildmaster has joined #archlinux32
[13:15:53] <buildmaster> !rq buildmaster
[13:15:53] <buildmaster> any/puppet5 is broken (says eurobuild6-2): https://archlinux32.org
[13:15:54] <phrik> buildmaster: <buildmaster> I might be insane, but never confused ... ;-)
[13:26:25] <buildmaster> pentium4/ruby-atk is broken (says eurobuild6-4): https://archlinux32.org
[13:27:38] -!- deep42thought has quit [Ping timeout: 240 seconds]
[13:27:52] <buildmaster> i686/ruby-atk is broken (says rechenknecht): https://archlinux32.org
[13:30:37] <buildmaster> i486/ruby-atk is broken (says nlopc46-i486bs1): https://archlinux32.org
[13:30:39] <buildmaster> i686/ruby-gdk_pixbuf2 is broken (says eurobuild6-3): https://archlinux32.org
[13:32:08] <buildmaster> pentium4/ruby-gdk_pixbuf2 is broken (says nlopc46): https://archlinux32.org
[13:36:29] <buildmaster> any/ruby-optimist is broken (says nlopc46): https://archlinux32.org
[13:44:47] <buildmaster> any/ruby-packetfu is broken (says eurobuild6-1): https://archlinux32.org
[13:46:07] <buildmaster> i486/ruby-rugged is broken (says eurobuild6-7-i486): https://archlinux32.org
[13:56:39] <buildmaster> i486/ruby-sqlite3 is broken (says eurobuild6-7-i486): https://archlinux32.org
[14:23:35] <buildmaster> i686/ruby-rugged is broken (says eurobuild6-5): https://archlinux32.org
[14:26:00] <buildmaster> pentium4/ruby-rugged is broken (says eurobuild6-3): https://archlinux32.org
[14:31:18] <buildmaster> pentium4/ruby-sqlite3 is broken (says eurobuild6-4): https://archlinux32.org
[14:32:05] <buildmaster> i686/ruby-sqlite3 is broken (says nlopc46): https://archlinux32.org
[14:34:15] <buildmaster> any/ruby-test_declarative is broken (says eurobuild6-3): https://archlinux32.org
[14:37:10] <buildmaster> any/puppet is broken (says rechenknecht) - already flagged out-of-date upstream on 2019-11-20: https://archlinux32.org
[14:37:43] <buildmaster> any/ruby-tilt is broken (says nlopc46) - already flagged out-of-date upstream on 2019-12-23: https://archlinux32.org
[14:44:42] <buildmaster> any/ruby-i18n is broken (says nlopc46): https://archlinux32.org
[14:50:03] -!- T`aZ_ has joined #archlinux32
[14:53:25] -!- T`aZ has quit [Ping timeout: 265 seconds]
[15:23:11] -!- MrBIOS has joined #archlinux32
[15:56:06] <T`aZ_> https://archlinux32.org gives 404
[16:06:04] <buildmaster> i486/kqtquickcharts are broken (says eurobuild6-7-i486): https://archlinux32.org
[16:06:15] <fALSO> damn a lot of problems
[16:06:25] <fALSO> yesterday i also had a segfault on glib
[16:40:12] -!- yans has joined #archlinux32
[16:44:57] <tyzoid> fALSO: Can you post the trace to https://bugs.archlinux32.org ?
[16:44:58] <phrik> Title: Archlinux32Arch Linux 32: Tasklist (at bugs.archlinux32.org)
[16:45:05] <fALSO> couldnt reproduce it
[16:45:11] <tyzoid> ah
[16:45:21] <fALSO> was running nmcli, but i dont know if it was on nmcli or in somethign that nmcli runs
[16:45:26] <fALSO> like wpa_supplicant or something
[16:45:52] <fALSO> just got a segfault and a little notice on dmesg saying it was glib related :)
[16:47:08] <tyzoid> I see
[17:08:14] <buildmaster> pentium4/kalzium is broken (says eurobuild6-6): https://archlinux32.org
[17:14:44] <buildmaster> i686/kalzium is broken (says rechenknecht): https://archlinux32.org
[18:19:18] -!- isacdaavid has joined #archlinux32
[18:28:10] -!- AndrevS has joined #archlinux32
[19:47:46] -!- isacdaavid has quit [Ping timeout: 265 seconds]
[20:55:25] <buildmaster> any/pifpaf is broken (says eurobuild6-2): https://archlinux32.org
[22:01:08] <buildmaster> i486/fplll is broken (says eurobuild6-7-i486): https://archlinux32.org
[22:06:49] <buildmaster> i686/kalarm is broken (says eurobuild6-4): https://archlinux32.org
[22:15:31] -!- MrBIOS has quit [Quit: MrBIOS]
[22:42:44] <buildmaster> i686/signal-desktop is broken (says eurobuild6-2): https://archlinux32.org
[22:42:45] <buildmaster> pentium4/signal-desktop is broken (says eurobuild6-1): https://archlinux32.org
[22:44:07] -!- AndrevS has quit [Remote host closed the connection]
[23:16:19] <buildmaster> any/ruby-i18n is broken (says eurobuild6-6): https://archlinux32.org