#archlinux32 | Logs for 2022-09-29
Back
[00:01:49] -!- drathir_tor has joined #archlinux32
[00:43:15] -!- T`aZ has quit [Ping timeout: 265 seconds]
[00:44:40] -!- T`aZ has joined #archlinux32
[00:54:41] -!- T`aZ has quit [Ping timeout: 252 seconds]
[00:56:44] -!- T`aZ has joined #archlinux32
[01:01:37] -!- T`aZ has quit [Ping timeout: 265 seconds]
[01:02:31] -!- T`aZ has joined #archlinux32
[02:02:25] -!- bill-auger has quit [Ping timeout: 246 seconds]
[02:28:15] -!- drathir_tor has quit [Remote host closed the connection]
[02:43:48] -!- drathir_tor has joined #archlinux32
[04:26:22] -!- bill-auger has joined #archlinux32
[07:13:32] <girls> abaumann: nothing to be sorry about - upstream should fix their PKGBUILDs ;)
[08:08:12] -!- abaumann has joined #archlinux32
[08:08:12] <buildmaster> Hi abaumann!
[08:08:12] <buildmaster> !rq abaumann
[08:08:13] <phrik> buildmaster: <abaumann> backups is for loosers ;-)
[08:08:18] <abaumann> deep42thought: :-)
[08:28:05] <abaumann> mmh, replication on archlinux32 is sometimes a little bit shacky..
[08:29:05] <abaumann> ..I usually stop both web servers and the buildmaster, so that no database writes should happen on the master, but it still has trouble sometimes.
[08:29:35] <abaumann> In theory replication should also work on the slave while the master is down.. The slave should recover. But it doesn't always in my experience.
[08:30:16] <abaumann> Especially when there are mysql updates, trouble is usually ahead. The slave cannot read some parts of the binlog or so..
[08:31:10] <abaumann> I have maybe an explanation why certain packages are not deleted (like quazip)..
[08:31:25] <abaumann> ..there is still software depending on it (like krita) which was not re-buildable.
[08:31:44] <abaumann> In this case you cannot simply delete the old package.
[08:32:11] <abaumann> For python2: people should use the one from the AUR (with the Gentoo patches)..
[08:32:29] <abaumann> ..we never have or will add packages not known to upstream to the official repos..
[08:33:12] <abaumann> ..but in cases like special kernels (olpc kernel), shim packages to make old software work (llvm13-libs) or handy things for old machines (pcmciautils).
[08:33:23] <abaumann> We should try not to use names from upstream if possible..
[08:35:09] <abaumann> https://www.archlinux32.org
[08:35:11] <phrik> Title: Arch Linux 32 - Package Search (at www.archlinux32.org)
[08:35:17] <abaumann> mmh, still plenty python2 packages..
[08:35:54] <abaumann> deep42thought: is there a handy script to delete them? Or is the script called 'deep42thought.sh'? ;-)
[08:36:35] <abaumann> 2019-12-03, wow, most packages date back ages..
[08:57:35] -!- deep42thought has joined #archlinux32
[08:57:35] <buildmaster> Hi deep42thought!
[08:57:35] <buildmaster> !rq deep42thought
[08:57:36] <phrik> buildmaster: <deep42thought> I know this feeling: doing hundrets of laser shots (manually) just to discover, that your spectrometer was on auto-trigger, but not on auto-save ...
[08:57:40] <deep42thought> Hi abaumann!
[08:59:14] <deep42thought> delete-packages should do all the no-brainer stuff and might have some options to force-delete packages - but I would be careful with the latter
[09:00:12] <deep42thought> yeah, it has some "-f" option - but I would advise against using that
[09:00:36] <deep42thought> if a package is really unbuildable, then rather blacklist it, and it will (should) be picked up by delete-packages automatically
[09:01:05] <deep42thought> this has the advantage, that (a) future updates won't try to rebuild the package and (b) all dependent packages will be removed automatically, too
[09:01:41] <abaumann> yes, absolutely true, I wouldn't force anything.. I never had good experiences when forcing things.. :-)
[09:01:46] <deep42thought> but let me do a non-force run of delete-packages first
[09:01:59] <abaumann> ok, just to see, if it picks up something..
[09:02:00] <deep42thought> maybe this cleans up some part of the mess :)
[09:02:09] <abaumann> yes :-)
[09:02:25] <abaumann> and.. it's not just python2.. I see mess in ruby, not to mention haskell..
[09:02:35] <deep42thought> once upon a time, I thought about putting this into its own cronjob (aka systemd timer), but I never had the guts to do so
[09:02:55] <abaumann> it's not necessary, I think..
[09:03:03] <deep42thought> and it's pretty risky
[09:03:13] <abaumann> ..people using Arch32 should be aware, that we are not adhering to the same security standards as upstream..
[09:03:24] <deep42thought> if for some reason, python lands on the to-be-deleted list, then everything will be gone in only a few runs :)
[09:03:25] <abaumann> a) because we cannot (think old CPUs and branch mitigation)
[09:03:44] <abaumann> b) providing old software with potential vulnerabilies (because we cannot rebuild them)
[09:03:52] <deep42thought> true
[09:03:55] <abaumann> So, it's better to leave them around knowing the state.
[09:04:08] <deep42thought> c) we're to few hands to have eyes on everything
[09:04:21] <deep42thought> *too
[09:04:21] <abaumann> c) applies to almost everything, yes :-)
[09:04:25] <deep42thought> :D
[09:04:39] <deep42thought> standard excuse
[09:04:59] <deep42thought> python-dateparser got deleted - should this be gone?
[09:05:03] -!- epony has quit [Read error: Connection reset by peer]
[09:05:14] <abaumann> mmh.
[09:05:26] <abaumann> https://archlinux.org
[09:05:28] <phrik> Title: Arch Linux - Package Search (at archlinux.org)
[09:05:29] <abaumann> mmh, seems so.
[09:05:32] <deep42thought> yup, gone upstream, too
[09:05:58] <deep42thought> well, then I'll let the script run unattended and see in the end, what is gone :D
[09:05:59] <abaumann> Just a thought: there is "deleted" and "deleted"..
[09:06:09] <deep42thought> huh?
[09:06:13] <deep42thought> what do you mean?
[09:06:17] <abaumann> ..what I mean: sometimes packages get moved to the AUR (like lately nagios)..
[09:06:27] <deep42thought> mhm
[09:06:35] <abaumann> ..not because it's unpopular, but because upstream is maybe unhappy about the current maintaner or so..
[09:06:43] <deep42thought> but we don't put aur packages into our official repositories
[09:06:51] <deep42thought> so it's the same "deleted"
[09:06:52] <abaumann> so we might have packages hopping out and in..
[09:07:01] <abaumann> ..from our perspective, I would say so.
[09:07:11] <deep42thought> if it rebuilds: no problem; if not: we're in trouble anyways
[09:07:20] <abaumann> true
[09:07:46] <abaumann> yeah, when the script deletes too much, we can always rebuild it..
[09:07:55] <deep42thought> I wouldn't care too much about the aur - demand for aur packages is probably not that high and it's a big security risk to automatically build all of them ...
[09:08:25] <deep42thought> gotta go - bbl
[09:08:27] <abaumann> you could do something like chaotic AUR (or how this binary repo is called), but I would not.
[09:08:31] <abaumann> ok, cu :-)
[09:08:37] <abaumann> and thanks..
[09:08:41] <deep42thought> np
[09:08:42] -!- deep42thought has quit [Quit: Leaving.]
[09:30:50] -!- epony has joined #archlinux32
[09:50:23] -!- T`aZ has quit [Ping timeout: 265 seconds]
[09:52:05] -!- T`aZ has joined #archlinux32
[09:54:29] -!- deep42thought has joined #archlinux32
[09:54:30] <buildmaster> Hi deep42thought!
[09:54:30] <buildmaster> !rq deep42thought
[09:54:31] <phrik> buildmaster: <deep42thought> good software is like a good meal: you cannot expect to get something really good if it's done in 5 minutes ...
[10:13:29] -!- T`aZ has quit [Ping timeout: 252 seconds]
[10:15:19] -!- T`aZ has joined #archlinux32
[10:46:52] -!- deep42thought has quit [Quit: Leaving.]
[13:04:13] <abaumann> -rw-r--r-- 1 mirror mirror 1671361 Sep 29 11:29 md5sums
[13:04:13] <abaumann> -rw-r--r-- 1 mirror mirror 4118785 Sep 29 11:29 sha512sums
[13:04:21] <abaumann> wow, there is a script running amok here.. :-)
[13:07:09] <abaumann> ssh: connect to host archlinux32.org port 9586: Connection refused
[13:07:15] <abaumann> yep, and this is not fail2ban
[13:07:28] <abaumann> something smart on top is blocking me..
[13:09:41] -!- titus_livius has joined #archlinux32
[13:34:17] <abaumann> md5sum: archlinux32-2021.04.06-i686.iso: no properly formatted checksum lines found
[13:34:40] <abaumann> doesn't sound good. I'll regenerate those checksums by hand
[14:00:53] <abaumann> https://mirror.archlinux32.org
[14:00:56] <abaumann> a new i486 ISO
[14:01:04] <abaumann> now I'll check status of the default ISO..
[14:19:37] <girls> archlinux32.org has some fancy firewall rule (iptables IIRC), that blocks everyone doing a port scan
[14:19:48] <girls> this also blocks you if you have to many ssh connections IIRC
[14:19:59] <girls> so it might "just" be iptables blocking you :D
[14:25:22] <abaumann> oh, that kind of smartness, explains perfectly why reboot works :-)
[14:27:01] <girls> or you could just be patient and wait a minute ;)
[14:43:05] -!- epony has quit [Remote host closed the connection]
[15:05:33] <KitsuWhooa> abaumann: thanks for looking into quazip. Appreciated
[15:24:00] <abaumann> KitsuWhooa: you are welcome. :-)
[15:24:23] <abaumann> deep42thought: me? patient? wait? ;-)
[15:25:26] <girls> :D
[15:25:38] <abaumann> ah, the deletion script worked: my favourite joe is now no longer in the repos. :-) no sweat.. I built it from the AUR..
[15:25:54] <abaumann> favourite editor joe*
[17:52:16] -!- abaumann has quit [Quit: leaving]
[18:05:07] -!- T`aZ has quit [Ping timeout: 248 seconds]
[18:06:55] -!- T`aZ has joined #archlinux32
[19:13:58] -!- epony has joined #archlinux32
[20:07:33] -!- abaumann has joined #archlinux32
[20:07:33] <buildmaster> Hi abaumann!
[20:07:33] <buildmaster> !rq abaumann
[20:07:34] <phrik> buildmaster: <abaumann> my "not" key on the keyboard is broken ;-)
[20:23:29] -!- abaumann has quit [Quit: leaving]