2025-02-01 12:14:58 oof 2025-02-01 12:16:59 ill check 2025-02-01 12:18:54 nu__: would be appreciated 2025-02-01 12:19:07 I tried to work around it with a wg tunnel, but did not really succeed 2025-02-01 12:29:24 im on fosdem but don't have much to do anyways 2025-02-01 13:56:20 i can repro it 2025-02-01 14:54:18 could you set the mtu to 1384 ? 2025-02-01 14:56:07 im announcing it via RA-s but linuxes dont seem to care about it 2025-02-01 15:23:31 let me try 2025-02-01 15:25:54 It seems to work, but I apparently have to set it on each container interface 2025-02-01 16:05:11 hmm 2025-02-01 16:05:26 I'm adding it to the default lxc config now 2025-02-01 16:05:57 i wonder why is this broken by default with wireguard 2025-02-01 16:06:14 Good question 2025-02-01 16:08:30 Let's see if that makes prowlar build now 2025-02-01 16:09:46 im kind of clueless when it comes to MTU 2025-02-01 16:10:05 i need to find a way to make this works without mods 2025-02-01 16:10:53 Normally https://en.wikipedia.org/wiki/Path_MTU_Discovery should take care of it 2025-02-01 16:11:27 It relies on 'packet too big' / 'fragmentation needed' icmp packets 2025-02-01 16:12:44 So if those packets for some reason are not able to make it back, pmtu discovery is broken 2025-02-01 16:13:44 aarch64 was now able to build prowlarr 2025-02-01 16:13:47 \o/ 2025-02-01 16:23:30 \o/! 2025-02-02 14:08:41 something is up with natting:/ ill implement the final networking next week 2025-02-02 14:09:26 Thanks 2025-02-04 19:47:50 re. the equinix stuff, come talk to us (adelie) about hardware and possible partnerships for either colo, managed, or VM or other resources. 2025-02-04 19:49:24 not going to chase anyone down about it but thought i'd offer a reminder. PM me or email about it (me^zv.io) 2025-02-04 20:18:34 zv: thanks, we have received quite a number of offers already, but we'll keep this in mind as well (especially for our long-term goal) 2025-02-04 21:17:33 sounds good. for reference, we self host everything 2025-02-04 21:18:24 zv: you mean like paying for your own infra? 2025-02-04 21:19:50 yes. like, donations pay for data center space and occasionally hardware. we run it all. 2025-02-04 21:20:17 Right, that's nice 2025-02-04 21:21:52 equinix f'd us a while back, too, and that is part of the reason we opted to be in total control of our hardware 2025-02-04 21:22:11 (aside from it being more affordable than renting, if not sponsored) 2025-02-05 02:47:37 https://lobste.rs/s/gjdafu/seeking_support_after_equinix_metal 2025-02-06 16:28:08 clandmeter: ^ 2025-02-06 16:53:51 meow 2025-02-06 16:54:02 o/ - I hear that Alpine was looking for colocation space etc? 2025-02-06 16:55:27 night: yes, though we have already received quite a number of offers, which we are sitting through 2025-02-06 16:57:26 ah i see, would it still be helpful to offer anything? 2025-02-06 17:14:55 night: depends on what you have to offer 2025-02-06 17:15:17 But we already see some very attractive offers 2025-02-06 17:26:31 ikke: we can offer essentially everything being asked for on the list on a no-attribution basis (we don't ask to be placed on your sponsors list or anything) 2025-02-06 17:26:54 we have sites in Amsterdam (NL), Dronten (NL), San Jose, Seattle, New York; network ports are all 2x25GE 2025-02-06 17:27:01 we can probably acquire whatever hardware you need 2025-02-06 17:27:40 Then please do send an email to infra@a.o 2025-02-06 17:28:02 i was curious how strong of an ask the Amsterdam/NL requirement was? 2025-02-06 17:29:14 we have 50-70kW free in our seattle site, which we mostly use to house OSS projects atm 2025-02-06 17:40:03 night: that's mostly for just colocation 2025-02-06 17:40:14 If we have to install hardware ourselves 2025-02-06 17:40:21 if we managed the entire thing, is seattle ok? 2025-02-06 17:40:42 er, hardware racking/etc wise 2025-02-06 17:41:21 Yes 2025-02-06 18:25:12 ikke: cool; sent off an email. 2025-02-06 18:25:19 Thanks 2025-02-06 18:36:23 ikke: ? 2025-02-06 18:42:42 gbr2-dev1 disk is almost full, could you check? 2025-02-06 20:01:40 ikke: how can i see what is taking al the space? 2025-02-06 20:01:57 could it be docker using too much log spoace? 2025-02-06 20:03:23 gdu / 2025-02-06 20:03:40 im running prune first 2025-02-06 20:04:45 hmm 2025-02-06 20:04:52 edge db is taking 16G 2025-02-06 20:05:00 i wonder if thats normal 2025-02-06 20:05:46 looks like it is 2025-02-06 20:13:31 ikke: its partially renovate cache 2025-02-06 20:13:43 appstteam also takes 5gb 2025-02-06 20:21:30 You could purge the renovate cache 2025-02-07 13:06:21 Saw the blog post, I got a machine mostly running idle and a spare IP and am Willing to give yall some VM´s to run the CI 2025-02-07 14:27:34 nekocentral: thanks, do we do have a lot of offers already 2025-02-07 15:59:36 Good to hear haha 2025-02-07 17:31:50 is pkgs-old.a.o still available or has it been retired? 2025-02-07 17:32:41 it's returning "404 page not found" at my end 2025-02-07 18:39:35 What is wrong with the new one 2025-02-07 18:44:06 nothing's wrong, it's good ... just 1-2 minor things about the old one that made it more convenient to use? e.g. linking to git.a.o for the repository to see the APKBUILD/history instead of gitlab (which seems to bave trouble showing history for certain aports with long history) 2025-02-07 18:44:24 and liked the way the old flagged list was sorted 2025-02-07 18:45:25 and admittedly have occasionally used the manual flagging 2025-02-07 18:45:59 while anitya information gradually improves 2025-02-07 18:49:11 do like the new one includes info for loongarch64, was using both 2025-02-07 19:08:26 yeah the sorting of the flagged list is certainly a good point 2025-02-07 19:21:32 :) 2025-02-07 19:38:47 mio: i turned it off as i thought nobody would use it 2025-02-07 19:38:54 i can turn it back on again 2025-02-07 19:39:10 but better if we fix things in new one 2025-02-07 19:39:38 fossdd[m]: you want to be added as maintainer to pkgs? 2025-02-07 19:40:38 clandmeter: that would be good in the meantime, thanks. and agreed :) 2025-02-07 19:58:06 mio: would you be able to generate a patch to implement the old behaviour? 2025-02-07 19:58:21 put it under flagging or flagged 2025-02-07 19:58:35 just need to find a place to put the message we have no 2025-02-07 19:58:35 w 2025-02-07 20:07:19 clandmeter: you mean for sorting the flagged list? 2025-02-07 20:08:37 or have the flagged in a separate page like before 2025-02-07 20:08:42 sorted or whatever 2025-02-07 20:11:18 not sure, can try looking into it. currently the flagged list sorts by build date, if the db stores anitya flag date then maybe a sort based on it would be good 2025-02-07 20:40:24 "fossdd: you want to be added..." <- sure, wouldnt certainly mind :) 2025-02-08 17:12:02 clandmeter: is it correct that i cant push to apkbrowser's master? 2025-02-08 17:17:57 fossdd[m]: you cannot directly push to protected branches 2025-02-08 17:18:06 ah 2025-02-08 17:19:38 fossdd[m]: and as it's setup now, you can also not merge anything 2025-02-08 17:19:43 into master 2025-02-08 19:15:47 ikke: Can you adjust it? 2025-02-08 19:16:08 Not sure what I need to do 2025-02-08 19:16:19 That fossdd[m] can merge? 2025-02-08 19:16:36 I can do something, but that would count for all developers 2025-02-08 19:16:54 Don't think that's a problem 2025-02-08 22:00:29 fossdd[m]: ikke has fixed it so you can merge things via MR