2024-06-18T00:31:37  <a-865k> forums report: "We are very sorry, but the requested service is currently not available. ~  ~ Please visit our status page to check if the problem is already known or related to a scheduled maintenance."
2024-06-18T00:32:39  <matrix-o-o> <knurpht - Gertjan> a-865k: All fine here
2024-06-18T00:32:46  <a-865k> 5th retried succeeded
2024-06-18T00:33:20  <a-865k> took a few minutes
2024-06-18T03:20:29  <mdogg> acidsys: Check out this Pastebin.  There's a link to a descent test server on that page, too.    https://github.com/szabodanika/microbin
2024-06-18T03:22:45  <mdogg> There's an option to "list" existing pastes as well.  I may install it just so I can play around with the configuration a bit more.
2024-06-18T03:22:58  <mdogg> The 'list' option is hidden in the demo.
2024-06-18T08:40:02  <acidsys> Provo IPv6 to the internet seems flapping, probably same ISP issue as before
2024-06-18T09:51:40  <acidsys> colleague made ticket with them
2024-06-18T14:39:49  <mdogg> acidsys: I like this one the best for a variety of reasons-> https://privatebin.info/    You can play around with it here-> https://paste.mdogg.com   If you like it, perhaps we can recommend it as a replacement for our current openSUSE pastebin.
2024-06-18T15:02:50  <acidsys> mdogg: I have my own PrivateBin at https://pasta.lysergic.dev/ but no, we will not replace paste-o-o with it. When we had the discussion before paste-o-o was developd I suggested PrivateBin and some people did not like it.
2024-06-18T15:05:10  <mdogg> I would like to understand what makes paste-o-o better than PrivateBin?
2024-06-18T15:06:24  <mdogg> Do people like going through every paste to ensure it doesn't contain contraband since it's stored on the server unencrypted?
2024-06-18T15:26:06  <SFaulken> I didn't see the debate, but I suspect that might have something ot do with it.
2024-06-18T15:30:25  <acidsys> probably hellcp would know best why paste-o-o was written
2024-06-18T15:32:03  <matrix-o-o> <LCP> yeah, the reason was there were no solutions that provided all the existing features and was a supported piece of software
2024-06-18T15:32:29  <matrix-o-o> <LCP> essentially most pastes failed at file upload, and there really aren't that many long term maintained paste softwares
2024-06-18T15:33:15  <matrix-o-o> <Emma [it/its]> hellcreeper
2024-06-18T15:33:57  <matrix-o-o> <LCP> I know that it's easy to complain about paste-o-o, but I do not see anyone engaging in creating issues on github, so I don't exactly know what people are missing
2024-06-18T15:34:03  <matrix-o-o> <Emma [it/its]> (sorry for random offtopic, just read your displayname and thought "hellcreeper")
2024-06-18T15:34:28  <matrix-o-o> <LCP> unfamiliar with it
2024-06-18T15:35:11  <matrix-o-o> <Emma [it/its]> LCP: having random complaints without feedback is annoying tbh
2024-06-18T15:35:58  <matrix-o-o> <Emma [it/its]> like cmon, if youre going to complain, might aswell say _why_ right?
2024-06-18T15:36:00  <matrix-o-o> <LCP> yeah, I genuinely want to work on features for it, believe it or not I enjoy writing software and creating websites and such
2024-06-18T15:36:27  <matrix-o-o> <Emma [it/its]> same though i absolutely loathe frontend work
2024-06-18T15:38:10  <matrix-o-o> <LCP> it's not that bad
2024-06-18T15:38:59  <matrix-o-o> <LCP> I just threw together https://github.com/openSUSE/get-o-o/pull/207 which is a nightmare actually, because there is so much rot from every single person injecting their own idea of how the site should be structured and how things should look
2024-06-18T15:39:20  <matrix-o-o> <LCP> and so I just need to bring it back into alignment
2024-06-18T15:39:51  <matrix-o-o> <LCP> but also I am exhausted working on this for now
2024-06-18T15:40:54  <matrix-o-o> <LCP> actually since this site started we doubled the amount of distros which is entirely unmaintainable if you ask me and I do not understand how in the god's name anyone is meant to make a decision on what they would want to use without a sales person behind their ear
2024-06-18T15:41:29  <matrix-o-o> <LCP> it's like we are replicating SUSE with their million distros but without a marketing/sales departments to actually manage to differentiate and sell something to people
2024-06-18T15:47:39  <matrix-o-o> <Richard Rahl> Sadly as the immutable desktop distros kinda split even more (and i assume richards new installer doesnt support multiple distros, besides being already at 3gb just for gnome) it only gets worse probablt
2024-06-18T15:49:01  <matrix-o-o> <LCP> we also have 3 code streams at this point
2024-06-18T15:49:08  <acidsys> it's quite annoying, also at conference booths when I try to explain to new people what all these distributions are, I feel like they chime out after the third one
2024-06-18T15:49:22  <matrix-o-o> <LCP> we have 7
2024-06-18T15:49:30  <matrix-o-o> <LCP> we have 7 distros
2024-06-18T15:49:45  <matrix-o-o> <LCP> it's as many as colors of the rainbow
2024-06-18T15:50:03  <acidsys> but missing the pot with gold at the end
2024-06-18T15:50:24  <matrix-o-o> <LCP> https://matrix.opensuse.org/_matrix/media/v3/download/opensuse.org/aEYNbvJTADDcCAmjAwfJxKOQ/Screenshot%20from%202024-06-18%2016-21-38.png
2024-06-18T15:50:34  <matrix-o-o> <LCP> I wish richard chose a different color for aeon
2024-06-18T15:50:40  <matrix-o-o> <LCP> how am I supposed to make a rainbow
2024-06-18T15:50:44  <mdogg> My first "like to see" from paste-o-o was the ability to create a raw URL that wasn't ten miles long.  But then I started looking at PrivateBin and realized the folks who manage paste-o-o probably spend hours going through all of the pastes to ensure some ass didn't go in there and start pasting/storing unencrypted contraband on the server.  From an administration perspective, having the pastebin encrypt even the stored data would eliminate
2024-06-18T15:50:44  <mdogg> the need to look through all the pastes.
2024-06-18T15:52:28  <matrix-o-o> <Richard Rahl> acidsys: right. I believe thats also the reason why fedora renamed their immutable stuff back into one name
2024-06-18T15:53:58  <mdogg> It's not hard to explain openSUSE Leap and Tumbleweed.  But explaining the various immutable versions can get a bit hairy.
2024-06-18T15:54:14  <matrix-o-o> <LCP> don't forget slowroll
2024-06-18T15:55:21  <matrix-o-o> <LCP> mdogg: I mean, that doesn't really remove the need to moderate, us being unable to tell what was uploaded might be even worse because now we wouldn't know what whoever with the link can see when they open a link that leads to an openSUSE website
2024-06-18T15:56:00  <matrix-o-o> <Richard Rahl> Eh slowroll is at least quite easy to explain imo, only issue i see, that it needs a proper schedule of rebases (no idea if that is already the case or not)
2024-06-18T15:57:07  <mdogg> LCP:  I guess I can understand that, too.  Regardless of whether or not anybody can access the paste (for example, the URL is lost) the data would still be housed on the server taking up space.
2024-06-18T15:58:47  <matrix-o-o> <LCP> would you be fine with a /raw after the url being a redirect to that long link?
2024-06-18T16:00:13  <mdogg> Actually, yeah that would be a solution.
2024-06-18T16:00:39  <matrix-o-o> <LCP> that's a solution that will take me 10 seconds to implement 😛
2024-06-18T16:01:02  <mdogg> Great - 10 whole minutes.
2024-06-18T16:01:05  <mdogg> j/k
2024-06-18T16:01:51  <matrix-o-o> <LCP> ok well, can't do 10 seconds because it takes forever for github to pull
2024-06-18T16:02:02  <mdogg> I'm just kidding.
2024-06-18T16:03:13  <matrix-o-o> <Lubos Kocman> 10
2024-06-18T16:03:15  <matrix-o-o> <Lubos Kocman> 9
2024-06-18T16:03:16  <matrix-o-o> <Lubos Kocman> 8
2024-06-18T16:03:18  <matrix-o-o> <Lubos Kocman> 7
2024-06-18T16:03:20  <matrix-o-o> <Lubos Kocman> :-)
2024-06-18T16:04:55  <mdogg> That said, even PrivateBin doesn't have the ability to link a raw URL.  Users can generate a raw URL but only after the standard paste is pasted since the decryption is done through JavaScript.
2024-06-18T16:06:35  <matrix-o-o> <LCP> well, I did implement it locally, but I need to build a container to test it actually works
2024-06-18T16:06:47  <matrix-o-o> <LCP> I could just deploy without testing too, but that seems inadvisable
2024-06-18T16:07:02  <mdogg> Probably not a good idea.
2024-06-18T16:07:15  <mdogg> Better to measure twice; cut once.
2024-06-18T16:11:30  <mdogg> LCP:  Thanks for hearing me out, by the way.
2024-06-18T16:23:13  <mdogg> LCP:  I can make this suggestion through GitHub if you like, but if 3 months is going to be the maximum time limit for pastes, these are my suggestions for the other limits:  https://paste.opensuse.org/pastes/16e71c3c2305
2024-06-18T16:23:24  <matrix-o-o> <LCP> I did some more maintenance things like copying the stylesheet from calendar https://github.com/openSUSE/paste-o-o/pull/197
2024-06-18T16:24:11  <matrix-o-o> <LCP> never is a hard sell tbh
2024-06-18T16:24:35  <matrix-o-o> <LCP> but yeah, I would like to have it on github
2024-06-18T16:24:57  <mdogg> I'll do that - it's where it belongs.
2024-06-18T16:27:24  <matrix-o-o> <LCP> in any case, I deployed, so the raw url is now available
2024-06-18T16:30:46  <mdogg> Hmm.. does it work for you?  I've been trying it with the URL I posted above and I get "something went wrong" though it does add /raw to the end of the URL..
2024-06-18T16:31:16  <acidsys> not at the end but after .org/
2024-06-18T16:31:44  <acidsys> nvm
2024-06-18T16:32:18  <acidsys> end is correct but returns 500
2024-06-18T16:34:14  <matrix-o-o> <LCP> oh weird
2024-06-18T16:37:38  <mdogg> I'll hold off on adding suggestions to the GitHub for now.
2024-06-18T16:38:09  <mdogg> I'll be back in a bit.
2024-06-18T16:41:21  <matrix-o-o> <LCP> I love when things break just in production
2024-06-18T16:50:14  <matrix-o-o> <LCP> fixed
2024-06-18T16:59:49  <acidsys> LCP: do you know how to make minio return content-disposition inline instead of attachment for text/plain ?
2024-06-18T17:00:32  <acidsys> I can't quite find if it's something to set on the server or in the app/s3 sdk
2024-06-18T17:02:08  <matrix-o-o> <LCP> fairly sure it's rails that decides on it
2024-06-18T17:02:31  <acidsys> but the response is coming from s3.opensuse-project.net
2024-06-18T17:02:43  <acidsys> I mean the raw
2024-06-18T17:03:44  <acidsys> or you mean activestorage when putting things into the s3
2024-06-18T17:04:02  <matrix-o-o> <LCP> https://edgeguides.rubyonrails.org/configuring.html#config-active-storage-content-types-allowed-inline
2024-06-18T17:04:50  <matrix-o-o> <LCP> we basically need an initializer that has the exceptions
2024-06-18T17:05:24  <matrix-o-o> <LCP> ie. a file in here https://github.com/openSUSE/paste-o-o/tree/main/config/initializers with the config set to the right values
2024-06-18T17:06:31  <acidsys> thanks I might try it
2024-06-18T17:07:15  <acidsys> I suppose I should repair broken hypervisor before hacking new features though
2024-06-18T17:07:32  <matrix-o-o> <LCP> broken hypervisor?
2024-06-18T17:08:28  <acidsys> yes one of ours does not work
2024-06-18T17:08:31  <acidsys> https://progress.opensuse.org/issues/162401
2024-06-18T17:10:32  <matrix-o-o> <LCP> oh
2024-06-18T18:31:18  *** teepee_ is now known as teepee
2024-06-18T19:55:55  <acidsys> three hours later I can confirm that the grub2 packages from 15.6 make for dead falkor machines
2024-06-18T20:12:23  <acidsys> Firstyear: can you find why the special PAM configuration gets destroyed when upgrading to 15.6? https://progress.opensuse.org/issues/162326
2024-06-18T20:13:05  <acidsys> cboltz: any idea why witch1 gets /srv/salt-git repeatedly recursively set to root:root after 15.6 ? I need to always run chown again (same poo)
2024-06-18T20:14:36  <cboltz> no specific idea, but IIRC I've seen my local salt test setup (on Tumbleweed) chown pillar/ to root:
2024-06-18T20:14:53  <cboltz> so maybe newer salt versions do that chown for whatever reason
2024-06-18T20:20:07  <cboltz> just checked - 15.5 had salt 3005.1, while 15.6 and Tumbleweed have 3006.0, so it might indeed be something in that version
2024-06-18T20:25:29  <acidsys> I just tested "salt witch1.infra.opensuse.org state.show_highstate" as root and as user salt and neither change the permissions
2024-06-18T20:25:49  <acidsys> the salt master service runs as user salt so would not be able to chown to root:root
2024-06-18T20:29:53  <cboltz> I don't know why/how it happens, and I don't have a specific reproducer - the ony thing I know is that it happened a few times
2024-06-18T20:30:31  <cboltz> BTW: if I get salt-master.service (on Tumbleweed) right, it gets started as root (and possibly drops privileges later)
2024-06-18T20:30:44  <cboltz> at least I don't see a User= line
2024-06-18T20:36:23  <acidsys> good point
2024-06-18T23:45:50  *** teepee_ is now known as teepee