[00:06:17] [1/8] What exactly do these permissions do? They appear on the following wikis: [00:06:17] [2/8] `(ceo)` = [00:06:17] [3/8] `(co)` = [00:06:17] [4/8] `(edit-admin-pages)` = nenawikiwiki [00:06:18] [5/8] `(edit-create)` = [00:06:18] [6/8] `(0)` = [00:06:18] [7/8] `(1)` = [00:06:19] [8/8] I don't even know the URL for the nenawikiwiki. [00:07:53] Try reaching out to the local admins to ask [00:08:02] We have no idea what they intend to use those permissions for [00:08:07] they're custom ones set by them [00:10:06] any Miraheze wiki can be usually accessed by https://nenawiki.miraheze.org/ even if it has custom domain [00:17:47] Er... I don't recall local administrators being able to create custom permissions. Or have they been granted the ability to create custom permissions yet? [00:22:02] can barely access my wiki, might as well call off editing for today 😦 [00:22:17] RIP images [00:22:20] I was able to a year or two ago iirc [00:22:47] They've always had the permission to create custom rights [00:23:00] It's just that it has to be requested on Phabricator, like in the good old days [00:23:25] Ugh. [00:24:09] So you have to request that the permissions are created for you? Or you have to request for the ability to create custom permissions at will? [00:24:27] You have to request the permission be created for you [00:24:35] Creating permissions from within ManageWiki isn't possible [00:24:45] They're created manually via LocalSettings.php [00:26:21] why can't peoiple edit that? [00:26:38] You can if you want to [00:26:38] people cant do it directly because malicious code could be added [00:26:40] Even my site logo has vanished, bit stuck on what to do here. Wait? [00:26:45] or code that simply doesn't work [00:26:48] #announcements [00:26:49] [00:26:52] all our code is here [00:27:08] Alrighty [00:27:27] @!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! you could make a pr though [00:27:35] It's a good habit to always check announcements if something is malfunctioning. [00:27:43] Just so folks know, we don't have an ETA on files being back up. The disks are back in the server, but the server isn't booting. [00:27:57] I heard that creating custom permissions freely was being considered, on the condition that there are restrictions on what types of permissions can be created. [00:28:16] I believe it was declined as it'd be too hard to implement [00:28:52] For example, custom permissions are only created to work with namespace restrictions and page protection settings. [00:30:16] Another good place to keep an eye on for technical problems is [00:32:07] I have a handful of permission concepts. Most of which would be too hard to implement. One of those permissions allows for selective revisions to be deleted with `?action=delete` as opposed to deleting entire pages and selectively undeleting revisions with Special:Undelete. [00:34:18] I understand that such a permission wouldn't be CC-BY-SA friendly. But it has always irked me how selective revisions can be undeleted with Special:Undelete, but selective revisions can't be deleted with `?action=delete`. Even with Special:RevisionDelete existing, as that is built on an entirely different system from ?action=delete and Special:Undelete. [00:38:37] Yeah, that far exceeds the complexity of a configuration change and would have to be a change to the core of the software. I doubt it would be done though, as maintaining a list of all revisions (even if deleted by RevisionDelete) is the preferred behavior of the software [00:42:17] True. Even if revisions could be deleted with ?action=delete, they could still be undeleted again with Special:Undelete. But such a permission would have a lot of room for abuse, as it would only be intended for use in situations where the CC-BY-SA licence wouldn't be at risk. [00:46:22] The fictitious `(selectivedelete)` permission would come in handy for someone like me. As I have been in situations where I would need such a permission. And because I would know how to use it without violating the CC-BY-SA licence. [01:37:09] and now there are intermittent 503s and 502s [01:38:35] Was about to ask if that was just me [01:46:12] I've been having them since a few days ago tbh [01:47:41] Well this time it's always the regular ol' miraheze logo 503 not smw 503 [02:06:04] im having them too [02:40:45] My wiki looks like a lowkey luminal space now. It's kinda funky. [02:43:48] <光纪#9056> me too [02:43:54] Mind the gaps, that's where the monsters hide. (jokes) [02:44:11] (not a joke) [02:45:39] <光纪#9056> What's worse is that I have a board with regional links on the map and it's completely messed up🤣 [02:49:32] oof [03:07:23] Are we down? I am getting continuous 503s on Meta... [03:09:17] Oh db101 backups probably. [03:10:38] Nevermind I have no idea, but icinga looks bad. [03:17:06] Everything I've tried is down atm. [03:17:46] I am able to access [03:23:08] Looks like mw backends going down depools everything??? Really doesn't seem like a good idea [03:37:13] Been down for 30 minutes now. [03:37:47] @Site Reliability Engineers something need rebooted? [03:38:07] Nothing I can see, it's up for me, seems to have been a blip on cp32 [03:38:24] or maybe not? [03:39:04] It has been constantly down all wikis I have visted for 30 minutes, and I am on cp33 [03:40:01] It's not obvious what's going on, there's no services other than cp* that appear to be down right now [03:40:06] Once created, can custom permissions be freely be granted and revoked from usergroups with Special:ManageWiki/permissions on the wiki the permission has been added to, or does it still need to be added to usergroups on request? [03:40:13] Massive amount of warning in icinga on IRC [03:40:17] nothing on grafana either [03:40:36] I'm looking at the icinga web interface and the only warnings that stuck are for cp* [03:40:57] Can you manually pool mw servers back into varnish and see if they work? [03:42:02] Or reboot cp server? [03:42:44] Yes, once created it can be controlled via ManageWiki [03:43:09] I'm seeing edits on Meta so it seems some people are able to edit [03:44:06] [1/5] ``` [03:44:06] [2/5] backend.set_health mediawiki healthy [03:44:06] [3/5] backend.set_health mw121 healthy [03:44:07] [4/5] ``` [03:44:07] [5/5] in varnishadm might be worth trying? [03:46:26] Now it switched to 502s. [03:46:41] Nope back to 503s. [03:47:18] cp32 is taking a long time to restart [03:47:30] (Varnish XID 21955262) via cp33 at Wed, 12 Apr 2023 03:46:29 GMT. [03:47:41] It is cp33 though? [03:47:44] or cp32 also? [03:48:01] I restarted cp32 [03:48:27] Depool us in DNS maybe? [03:49:29] cp33 shows healthy on all backends right now [03:49:40] cp2 only has 40% or so failure, cp 3 has 80% or higher according to IRC icinga [03:50:08] Never mind, cp22 jumped to 60% [03:50:35] Still 503s on cp33 [03:55:09] Things were down for me as well, but back up now. [03:55:39] Ahh, that timing. Of course issues come up while I was in the middle of working on something time-sensitive. =V [03:55:45] Still getting 503s on cp33 here [03:56:32] Bro is the wiki back up? [04:03:14] Back up for me also. [04:05:31] isn't it lovely when 300 different things go wrong at once? [04:07:05] Honestly, I'm not sure if there was actually anything that could have been done there to fix anything [04:07:40] PROBLEM - cp22 HTTP 4xx/5xx ERROR Rate on cp22 is CRITICAL: CRITICAL - NGINX Error Rate is 62% [04:07:53] Still issues it seems. [04:08:13] Maybe it will recover. [04:10:46] Disgutedorite: Always how sysadmining goes it feels; there can never just be *one* thing that goes wrong. D; [04:11:29] Only one thing going wrong makes solving the problem too easy! /s [04:11:54] If anything can go wrong, it will it seems. [04:12:58] Murphy's law 🤷‍♂️ [04:13:19] Yeah, that is basically what I was quoting lol [04:14:22] "Anything that can go wrong will go wrong." [04:59:52] [1/2] on an unrelated note I've been monitoring the stewards' noticeboard waiting for a reply on my thing and [04:59:53] [2/2] damn Naleksuh is literally just digging their own grave huh, and I say this while having literally zero context of what happened [05:00:29] I got [d0c99925ee8d783a80e560dc] 2023-04-12 04:59:48: Végzetes kivétel: „FileBackendError” on my Wiki lol [05:00:32] your input and commentary on the thread would be appreciated [05:01:09] I was just saying based on this alone https://meta.miraheze.org/w/index.php?title=Stewards%27_noticeboard&diff=prev&oldid=329073&diffmode=source [05:02:24] still, an outside view of this would be appreciated [05:06:15] [1/2] this is a long thread with branches I am not used to commenting on things like this [05:06:16] [2/2] should I reply to what I linked directly noting what it looks like from my perspective (kinda immature and seriously looks like they're trying to make a false equivalence to make themself look less bad but executing it very, very poorly, even without any information to go off of regarding what they actually did), or? like, where do I put the reply [05:07:23] I mean, at this point it's pretty much a choose-your-own-adventure [05:08:07] You can click the "Reply" button to that message if you wish and share your analysis there [05:23:09] Yes, an outside perspective on stuff like this can be very helpful. [05:25:47] Am just getting caught up on that one myself, had thought things were quiet on this front, but I guess not. [05:26:09] [1/4] I'm a little tired and don't want to sound too inflammatory, is this response okay? [05:26:09] [2/4] > Sorry to interrupt since I have very little information about this situation and am not involved directly, but I noticed this argument going on while I was waiting for a reply to a request I made elsewhere on the noticeboard and I feel the need to note that from an outside perspective, Naleksuh, with zero context for what you did (or are accused of doing), the sample s [05:26:09] [3/4] cript of calling for BrandonWM's removal looks like the kind of thing a kid who got banned for posting porn in general chat would write up to say that ''actually'', the mod who banned them should be banned instead because they ''also'' posted a picture of a girl at some point. The roleplay of users who probably don't exist also makes me think either you genuinely believe [05:26:10] [4/4] you have an army of supporters, or you are attempting to intimidate other people into thinking you do to keep your status as a wiki creator. Your own words gave me all the context I needed to be instantly convinced to support your removal. [05:28:36] I'd humbly suggest removing the reference to posting porn. While a valid comparison, it's not relevant and invites further nonsense. [05:29:24] Yeah, I try to remain neutral while also stating the facts. [05:29:36] It's a good way to go, when things are inflammatory. [05:29:46] "spamming in general chat" and "posting an emoji response once" instead for the comparative example? [05:30:31] I'm an outsider too, this is getting silly now and it needs to have a lid put on it. [05:30:42] Is that section for voting on someone's removal? [05:31:15] The introduction (or my reading of it as I'm not familiar with removal request formatting) is a bit unclear. [05:31:39] There isn't a good forum for it, but it's generally a steward action so the section is being kinda-repurposed. [05:32:12] If it's request for removal for Naleksuh, and voting on that, I definitely support based on continued pattern of inflammatory behaviour. [05:32:42] notaracham is this comparison better [05:32:56] Yep, my perspective is that anything worth saying has largely been said, and stewards will deal with the issue once platform stability is back in a good place. [05:33:15] Closer/clearer, yes. [05:33:25] If its not really a vote, then there's not much point in me engaging... [05:33:35] It's not really, no. [05:34:07] More of a festivus-style airing of grievance. [05:34:53] Ah, okay. Is there any value to MH in adding a mostly neutral perspective on the issue? [05:35:25] Community input in itself is still helpful for establishing a record and pattern, which in turn helps guide steward action. [05:35:45] But as said before, this is pretty rare so there isn't well-established procedure/precedent [05:36:38] I'd also add, pretty much everything that needs to be said has, but if you feel an outside perspective is helpful/would provide new insight, you're free to add it. [05:37:23] Personally, it's about time for me to call it a night. [05:37:49] morning, what's happening again [05:38:38] Continued file server stuff re: swift, most relevantly. [05:38:53] Yep. Frankly the false (reversed) vendetta has worn out its welcome, the user just needs to be removed at this point, so everyone can move on. [05:39:04] Some downtime for some stuff, then back up. [05:42:44] Adding your view on the SN thread would be appreciated so that the closing Steward can consider all viewpoints and opinions [05:45:43] Here's a comparison between what users with both the `(delete)` permission and the fictitious `(selectivedelete)` would see on the `?action=delete` page if the fictitious `(selectivedelete)` permission were to exist exactly how I've envisioned it. And what users with just the `(delete)` permission see. [05:50:52] You can probably figure this out already, but both images display options that won't appear without the `(deleterevision)` permission for the 'change visibility of selected revisions' button and checkboxes. And the `(deleterevision)` and `(deletelogentry)` permissions for the 'change visibility' links. [05:53:20] Shoot, forgot I was logged in as wrong user LOL [05:54:04] If I swap the comment to my primary, the username I use for voting, is anyone going to jump on me for it? [05:54:29] Nah [05:54:31] Was an honest multitasking mistake 😆 [05:54:51] If you already have an established multi-user relationship, that's fine. [05:55:17] Yeah, i do. I usually separate for my own wiki. [05:55:27] I always comment and vote only under my primary though, the one I use here. [05:55:40] All good, then. [05:56:28] If meta will let me log in that is heheheehh. [06:02:06] Alright, comment done and on to productive stuff now [06:16:20] Has anyone figured out a decent "expand all" "collapse all" function for pages with lots of tables (like if tables were used on a FAQ page for example)? [06:16:55] I'm lousy at js and only half way there from someone else's code. [06:19:24] super hacky way is to click every expand / collapse button [06:20:38] depending on how expand or collapse works you could jsut call the function repeatedly for every collapsible element [06:23:22] It's also possible to create a custom button that uses the same target ID for all objects. I'll dig up where I did that tomorrow [06:26:09] Yeah, that's kinda what I've got now, but not fully functional. If you do have something, would much appreciate it. [06:26:15] I'll keep fiddling in the meantime heheh. [06:26:46] https://planet-crafter.fandom.com/wiki/Version_history has an example of this using mw-customcollapsible [06:27:50] I scooped this from Fandom, but the collapse is problematic. [06:28:13] The expand works not so much the collapse. [06:28:27] duplicate ids is kinda bad practice tho [06:30:04] For sure, it's also awful for screen readers [06:31:40] Yeah, I'm so no-good at js. Basically I've been trying bits of code from here and there to try to cobble something together (very poorly) lol. / [06:31:48] The planet crafter example was unfortunately the cleanest way I could figure and these caveats are also mentioned in the mw-customcollapsible documentation [06:32:12] Thank you! I'll give that a go! [06:36:53] [1/14] ```js [06:36:54] [2/14] $('.test2').on('click', function() { [06:36:54] [3/14] if ($(this).data('state') === 'expanded') { [06:36:54] [4/14] $('.mw-collapsible-toggle-expanded').click(); [06:36:55] [5/14] $(this).data('state', 'collapsed'); [06:36:55] [6/14] $(this).text('Expand all'); [06:36:55] [7/14] } else { [06:36:55] [8/14] $('.mw-collapsible-toggle-collapsed').click(); [06:36:56] [9/14] $(this).data('state', 'expanded'); [06:36:56] [10/14] $(this).text('Collapse all'); [06:36:56] [11/14] } [06:36:57] [12/14] }); [06:36:57] [13/14] ``` [06:36:58] [14/14] i threw together the cheap hack method i mentioned earlier [06:37:05] instead of class test2 it's whatever your button is [06:38:12] instead of hitting every single mw-collapsble you could give your collapse all / expand all element a data field that specifies an additional class instead of just mw-collapsible yada yada [06:38:57] That'ssimilar to the hacky version I cobbled together. [06:46:00] Alternatively you could reimplement the collapsing logic and have total control over all the states [07:30:07] just heads-up, I'll be going to places for a week and a half so I'll be out of active wiki doings and support attempts here [07:35:12] [1/2] Unauthorized [07:35:12] [2/2] This server could not verify that you are authorized to access the document you requested. [07:35:32] Please fix those wikis right now!!!! [07:35:37] dude [07:35:40] again [07:35:57] #announcements [07:36:46] oh. [07:37:12] and please drop that "right now!!!!" from your vocabulary in this server, you've been told numerous times how things aren't easy on volunteer based platform [07:38:38] i deleted [07:39:12] Sorry😅 [07:39:38] i just wait [07:39:50] eheheheh😅 [07:40:24] [1/4] @Agent If you request for a custom permission that controls a custom protection level, do you get to choose the name of the protection level? [07:40:24] [2/4] For instance: [07:40:24] [3/4] ```MediaWiki:Right-3456 [07:40:25] [4/4] MediaWiki:Protect-level-6543``` [07:41:40] 😐 [07:42:01] [1/5] Just an example, since the sysop protection level doesn't match with the associated permission. [07:42:01] [2/5] For instance: [07:42:01] [3/5] ``` [07:42:01] [4/5] MediaWiki:Right-editprotected [07:42:02] [5/5] MediaWiki:Protect-level-sysop``` [07:51:13] Being aggressive towards volunteers will not help anything. It’s why so many volunteers quit volunteering and if you don’t have any SRE volunteering, your wiki doesn’t exist. [07:52:02] A general reminder that abuse will not be tolerated towards any user. Frustrations at technology are not an excuse. [07:55:17] sorry😔 [08:13:40] [1/2] It may be a dumb question, but how do you make the infobox on Miraheze? I have a template page for my Infobox written that should work, with .../infobox in it. [08:13:40] [2/2] But all pasting it and filling it on the page does is copying the text into this page. If I understand correctly, wiki just doesn't come with infobox module installed? So how do I install it [08:16:18] you need to enable extension called PortableInfobox, parser hooks tab in extensions admin menu [08:16:54] after enabling it just make an empty edit on these imported templates and they should start working fine [08:18:38] [1/2] @RhinosF1 any idea why only one file from Wikimedia Commons fails to show up? see recent support threads [08:18:38] [2/2] on my wiki only Sweden flag and all uploaded locally files are dead, the rest flags show up nicely [08:56:19] @Legroom I have 1 idea [09:02:14] Hey, I wanted to report that I frequently get 503 and 504 errors [09:02:24] I also cant log in due to that [09:03:23] Me too! [09:03:58] we have so much errors now 🥲 [09:04:09] I was getting those errors a lot earlier, but not at the moment. [09:06:32] Error rates are very high [09:07:34] Yeah, but I'm not gonna stress over it much. I know they're trying their best so. [09:15:32] I’m not convinced Swift is the only cause of issues [09:19:39] Is anyone else encountering a 503 Backend fetch failure message when trying to edit? [09:20:14] Yep, if you read upwards a little, there's some ongoing issues [09:20:17] I was getting them occassionally over the past 30 minutes but now I cannot edit at all [09:20:19] They're working on it. [09:20:20] ah ok [09:21:01] Also, there's image issues noted in the announcements channel 📣 as well. [09:22:17] yeah I brought that up yesterday [09:24:34] Can the wiki display images that were not uploaded yet and should stay broken links, from somewhere else? Namely Victory.png [09:25:40] I'd think that images are likely to behave erratically until they're done fixing the current issue. [09:28:56] What do you mean by "display images" [09:29:23] You want a list of redlinks? [09:29:25] Well I mean this 🥴 [09:29:53] I think it's really bug so nevermind [09:29:53] Displaying images from Wikimedia Commons should work fine [09:30:23] Ah, got it, you meant from commons. [09:30:46] Bed time for me, my brain stopped working an hour ago 😆 [09:32:00] I guess you have your own Victory.png on your wiki, but now it falls back to whatever is on Commons? [09:32:30] I didn't have it [10:00:08] No ETA for the pictures coming back yet, right? [10:01:38] Seems that is correct [10:05:23] I wouldn’t expect a useful ETA for at least a day or so and I would expect the outage to be extended. [10:05:28] @The Waslijn [10:09:25] Yeah that makes sense [10:09:52] I’m trying to get a better ETA [10:10:15] Or at least an indication [10:10:41] If you’re reliant on images, I wouldn’t plan anything for a few days though. [10:12:43] https://tenor.com/view/not-okay-stop-now-this-needs-to-stop-now-gif-14744906 [10:12:52] About Error 503. [10:13:14] Ah, great! Now it's 504 too. 😦 [10:13:27] Well, at least it's not db141, right? [10:13:29] 🙂 [10:14:32] All wikis are dependent on images. Without them, they are almost unusable. [10:14:53] no, it's nothing like db141 [10:15:00] that one was a total disaster [10:16:10] @Mizioł7 the memes aren’t funny [10:16:25] I agree! [10:16:54] Ah! Don't worry! Have a little faith, like Oddball once said to Moriarty. [10:17:13] Admins and moderators will deal with those errors! [10:17:27] Am I right, folks? 😄 [11:06:00] not really tbh [11:06:08] it's all in the hands of a limited group of techies [11:52:24] https://tenor.com/view/chris-chan-fighter-scary-gif-17816875 [11:52:35] miraheze users when someone is having fun: [12:08:56] oh my god HAHAHAHAAHAHAHAHAHAHA i love this [12:11:00] jeez looking at SN [12:11:03] wow [12:11:05] that's a lot [12:18:30] a lot of 503 so i can't check the pages about technical errors because are written on Miraheze [12:19:24] https://status.miraheze.wiki/ this page can be accessed at least [12:19:43] Thanks [12:23:37] my luxurious and hastily made logo is out of commission. unfortunate [12:53:29] getting repeated 502s across Meta.... [12:58:51] he is right about one thing: it is unfortunately not an area where the community has a direct impact on what happens, as wiki creators have no mechanism for voted removal [12:59:06] which is odd given many other roles do, but there it is [12:59:29] you are at liberty to post your 2c but in a thread already as long as it is I think it would ultimately add to the noise [12:59:59] this seems like it will be on and off until the right people can take a look at it, and void (last one as far as I know) had to sleep... [13:00:15] yup [13:18:20] Status is hosted externally [13:18:31] We should probably host dns for it externally too tbh [13:20:19] So... My pages are back, of course, but, uhm... [13:20:35] The problem is... something's wrong with da pictures. 😐 [13:21:44] Yes that’s an ongoing issue that is being investigated [13:24:33] The frequent 50x errors are down to the fact that every skin will try to load the logo and will fail - it didn’t start happening until later last night because before then the logos and such were cached [13:32:12] I see. [13:38:07] backrooms wiki is actually getting worse [13:39:34] @!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! please be patient. SRE know it’s shit. [13:39:50] not complaining [13:40:43] Ok [13:41:20] as the cache expires for more and more images, things will continue to get a bit stranger [13:42:23] for the backrooms wiki, favicon is gone and so is the logo [13:42:41] level 0's image cannot be displayed [13:43:02] nothing on level 1 can be displayed [13:43:21] the only thing working on level 2 is a wikimedia commons image [13:44:05] nothing on level 3 works [13:44:07] is anyone else getting 502 bad gateway on images rn? [13:44:09] no pages can be edited [13:44:22] https://tenor.com/view/late-gif-7440342 [13:44:32] sorry, i only just got told by someone whos trying to read my wiki lmao [13:45:11] Yes, we’re experiencing issues with access to files [13:45:18] ok, thanks [13:46:20] Yeah the only images that will work are one obtained via InstantCommons and also things like the footer buttons which are baked into MW core [13:46:31] ones* [13:46:57] ahhh ok [13:47:30] editing is broken too [13:47:55] @!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! you shouldn’t have issues editing [13:47:59] some wikis still work [13:48:17] What error do you see? [13:48:32] it seems to work now [13:48:35] i got a 500 error [13:49:01] That’s not an issue with editing specifically, it should work if you try again [13:49:17] [[mh:backrooms:Level 1]] can't edit [13:49:17] https://mh.wikipedia.org/wiki/backrooms:Level_1 [13:49:17] [13:49:21] even if you add nothing [13:49:22] They are issues with the way our servers detect errors [13:49:44] I've made the file sitenotice display everywhere now [13:51:10] Are you seeing a 500? [13:51:45] yep [13:52:07] Try again in a few minutes [13:52:17] Individual reports won’t change much [13:53:15] just saying ¯\(ツ)/¯ [13:57:17] We're currently trying a fix that would at least stop/make the 503s less frequent until files are fixed [13:59:32] The files will be fixed in a few days or so, we are told? [13:59:46] It's hard to say [14:00:12] hopefully this isn't another november [14:01:43] Is everything okay? [14:01:47] Files are still missing [14:02:20] I am aware of this. [14:04:41] We’re still looking into it [14:05:32] Thanks and hope everything goes well [14:07:33] Monitoring shows that with the change in place the error rates have dropped quite a bit - so 50X errors should be happing much less now [14:08:00] happening* [14:14:02] Sadness. 😦 [14:42:19] Ohh I'm here I'M FINALLY HEREEEE AAAA [14:47:12] oh thank god the image error is temporary [14:47:52] If will be permanent, proceeding to die XD [14:48:31] I can't imagine a world without images [14:49:12] ASCII art to the rescue [14:49:22] Imagine a world without eyes = not display images on our wikis [14:50:32] I remember back in the good old times, when we only had text-based browsers https://en.wikipedia.org/wiki/Lynx_(web_browser) [14:52:44] I love lynx [14:52:50] I use it all the time when on the shell [14:57:30] Well isn’t this interesting, the Discord embed shows the browser but the link is to the feline [15:07:39] Ah thank you! Exactly what I wanted to suggest today. [16:18:30] Just a question about picture bug. [16:18:44] Does this mean we can still edit text contect? [16:18:58] yes, that is unaffected [16:19:08] I don't asking for picture uploading, I suppose probably we can't do it because of this bug. [16:19:20] no, nothing related to files currently works [16:19:40] special:Upload has been disabled for the time being [17:13:58] Yeah, working on it [17:52:10] <ᴠᴀɴɢᴜᴀʀᴅ#7777> With Cargo down, are we able to get SMW installed instead? Is there still someone that can do that? [17:58:13] Yeah sure, I can do it [17:59:42] <ᴠᴀɴɢᴜᴀʀᴅ#7777> Thank you!!! [17:59:49] <ᴠᴀɴɢᴜᴀʀᴅ#7777> One moment: [17:59:57] <ᴠᴀɴɢᴜᴀʀᴅ#7777> https://tsb.miraheze.org [18:11:10] Meanwhile my request has been sitting for like a week now counting the time before I was told to move it from phab to the stewards notice board 😔 [18:11:32] my bad, if you link me to it I'll do it right now [18:11:49] sometimes people can be a little bureaucratic, my apologies for the delay and the back and forth there [18:12:53] https://meta.miraheze.org/wiki/Stewards%27noticeboard#sagan4alphaand_sagan4beta [18:14:33] Is it possible for you to investigate SMW throwing 503's with the MAINTENANCE_MODE code as well? [18:14:52] Which wiki? [18:15:20] projectsekai [18:16:09] ah it seems someone forgot to run the script or it failed (might've been me, if so apologies!) [18:16:12] I'll run it right now [18:16:26] fixed [18:17:08] thanks! [18:18:06] no problem! [18:19:07] haha, refreshed while it was in maintenance mode mid-enabling smw [18:20:55] also at last, the days of tweaking one (1) property requiring the entire table to be rebuilt, thus requiring mass null edits, will be over [18:21:35] also the days of yelling at other staff for recreating the table without making a separate replacement table [18:24:24] ...as soon as it stops being a perpetual 503 error [18:26:37] > Semantic MediaWiki was installed and enabled but is missing an appropriate upgrade key. [18:26:43] on sagan4alpha [18:27:26] My favorite experience with cargo is specifying unique and mandatory (eg not null) on a column resulted in all updated rows having null in that column [18:27:42] Was very confused on why row counts were extremely wrong [18:28:58] my favorite experience was that time a few hundred specific rows duplicated themselves and continued to duplicate themselves after the table was recreated like 3 times in a desperate attempt to fix it bc it was breaking everything that was trying to pull data from oen specific page [18:29:34] I'll try again. I think due to backend/503 issues the script malfunctioned [18:30:50] I still have no idea why only a few hundred specific rows were duplicating themselves, though it was a bug that got fixed later [18:31:29] like idk what was making the specific ones do that [18:31:53] cargo works in mysterious ways lol [18:33:25] I had to add a limit of 1 to the template I made that replicates smw's `#show` in Cargo even though the page name parameter should have taken care of that [18:34:38] I had the duplicate problem too with cargo, fun times lol [18:34:43] ye [18:35:47] oh lovely 503 and then a 502 on refesh [18:35:52] ??? [18:36:04] Same here [18:36:07] Anyone elses Miraheze down with error 503? [18:36:20] @Site Reliability Engineers think something just broke [18:36:47] definitely down from my end [18:37:08] Lots of technical errors today [18:37:19] me too [18:37:20] all 4 of my tabs bruh [18:37:29] image incident + whatever just happened [18:37:44] we can only sit tight for the moment [18:37:46] it's all related to the image incident [18:37:50] bruh [18:37:59] it's no coincidence! (at least I'm quite sure it isn't) [18:38:07] how Reception123? [18:38:12] . [18:38:25] there is a slight chance it could be SMW installs but that's a bit odd [18:38:31] @CosmicAlpha do you think that's possible? [18:38:43] I'm sure it's related but I'm finding mh quite literally inaccessible now [18:38:57] it pokes the server and goes 'nope I'm dead' [18:39:40] it's back to 503 [18:39:42] is smw that resource intensive or does it take everything down with it when it runs into a problem? [18:39:59] smw problems can cause problems for everything that uses smw [18:40:09] but I have not heard about it extending to everything else [18:40:12] it should not take the entire farm with it [18:40:38] this is just a way to get people to donate /s [18:40:51] I mean before this images issue there's never been an issue with SMW installs [18:40:58] We need donations, but I don't think it's that [18:41:03] maybe smw doesn't like image issues [18:41:20] think we're going to need infra to give it a good poking [18:41:28] went from limping to keeling over [18:41:37] I know, I was just joking. Wouldn't be surprised if someone tried to spin it that way though [18:41:51] First, the images go down, then all of Miraheze [18:41:54] Miraheze always goes down when I got a banger idea too <:tired:982016981429526609> [18:42:00] FR [18:42:03] oof [18:42:18] seems to be trying to respond now at least [18:42:22] now a page is loading, very slowly [18:42:22] yes [18:42:26] yeah, it got back up [18:42:27] Maybe there's a possibility that it's sabotage? 🤔 [18:42:29] Its up again for me [18:42:32] Its off and on for me [18:42:39] YEAHHHH [18:42:43] mine loaded [18:42:46] mostly [18:42:54] probably a fandom spy sneaking dynomite into the drives, hate it when that happens [18:43:03] Lol [18:43:06] common fandom L [18:43:07] I spelled that one completely wrong [18:43:10] Miraheze_Bot_Acc: choose another name pls [18:43:14] Y’all are gonna jinx it tho. Maybe its like back when we download a game as a kid. If we pretend to not care, it’s gotta start working [18:43:29] Orange_Star Why? [18:43:46] that name does technically contravene username policy [18:43:48] Didn't expect Dyno bot to be in cahoots with fandom [18:44:04] Oops sorry changing now [18:52:20] it's a shame that announcements on such issues take a bit to be made :p [18:52:50] (in #announcements on discord btw) [18:53:07] but then again, you have to be careful with how you phrase things [18:53:19] heh well looks like sagan4alpha is fixed [18:53:32] There was an announcement... [18:55:35] Is a steward online here? [18:55:38] yes [18:55:53] making so many property pages rn [18:56:05] Could you fulfill my rename request on wiki? [18:56:24] I am LC Developer [18:56:37] I'll have a look at the queue [18:56:44] Thanks [19:12:57] [1/5] Error 200 OK [19:12:57] [2/5] OK [19:12:58] [3/5] Guru Meditation: [19:12:58] [4/5] XID: 13406943 [19:12:58] [5/5] Varnish cache server [19:13:05] if that helps you miraheze devs [19:13:05] #announcements [19:13:17] We're doing that on purpose [19:13:25] but thanks \:) [19:13:31] :) [19:13:52] I can still acess miraheze. Are you guys having issues? [19:14:00] i know [19:14:23] Everything is recorded there. [19:14:49] what does that mean? [19:15:02] They are aware of all aspects, or at least most of them. [19:15:21] Globe: no, seems to be working now [19:22:59] Hello, on my wiki it says I don’t have permission to upload new images even though I am the owner. Also all other images that have been uploaded won’t display anymore [19:23:17] #announcements [19:23:35] This error is tied to the ongoing issues documented there and on the SRE tech noticeboard. [19:24:14] Ah ok [19:24:40] @Guardy No one can upload images right now on any wiki. The stewards are trying to solve the problem [19:25:00] Because the file storage is down, uploads have been disabled [19:25:19] Yep [19:28:21] most exciting part of having smw on sagan4alpha at last rn: seeing the things just keep filling up into the thousands instead of stopping at a few hundred like cargo!!!!! [19:30:28] aaa!!!!!!! (refreshes) !!!!! [19:30:39] im sorry i just love smw [19:31:38] and seeing it not fail in ways cargo fails [19:33:16] well, cargo is lightweight but also very unstable [19:33:39] (not being popular is a problem) [19:34:03] there is nothing unstable about it [19:34:05] it works really well when you don't suddenly turn it off [19:37:07] the constant database errors I was getting for a while say otherwise [19:37:14] then you were using it wrong [19:37:42] e.g. trying to insert strings that exceed the field length limit, which is something that Cargo silently ignores without showing a warning [19:38:16] the pages worked on refresh with zero change I don't think it was that [19:38:34] our Cargo database had 30000+ entries and was used to generate 10000+ pages [19:38:50] there were no issues with it [19:39:26] How many source pages tho [19:39:35] about 500-600 [19:39:36] data source* [19:40:11] There was a very long list of #cargo_store directives on about 500 separate pages, updated by a bot. [19:40:21] hundreds of #cargo_store per page [19:40:44] that sounds more like using it wrong to me [19:40:59] it serves the purpose [19:41:08] hundreds per page when you only have 10 tables? [19:41:24] anyway sagan4alpha stores data on over 5,500 pages [19:41:25] What's the difference from that vs an actual db [19:41:46] It is an actual DB. [19:42:25] I mean, in a db you would just add more rows to the db table, as opposed to storing information on individual pages and having cargo to assemble a table for you [19:42:46] I don't see any "using it wrong" having multiple store calls on a single page [19:42:58] for example: there are 12000 items in the "item" table. That means 120 pages with ~100 {{#cargo_store:}} per each. They are split into these 120 chunks to not cause server load when the pages are updated [19:42:58] https://meta.miraheze.org/wiki/%23cargo_store:Template: [19:43:06] if it's hundreds of store calls for 10 tables it seems inefficient [19:43:07] since you can't have 12000 {{#cargo_store:}} on one page [19:43:07] https://meta.miraheze.org/wiki/%23cargo_store:Template: [19:43:23] you need one #cargo_store per row [19:43:32] there are 12000 rows in that 1 table [19:44:25] Given how cargo sometimes decide to fail table rebuilds, it might actually be better to have fewer pages with more store calls [19:44:49] You can't have fewer pages. It will timeout when you try to save a page with too many {{#cargo_store:}} calls [19:44:49] https://meta.miraheze.org/wiki/%23cargo_store:Template: [19:45:03] I had, in fact, increase the number of chunks (reducing the size of each) several times [19:45:07] No, I'm not arguing against having fewer pages [19:45:32] i'm saying that it might actually be better than having 12000 pages when trying to rebuild [19:46:12] The very reason we are using Cargo is that the bot only needs to update about 80-120 pages (when new data becomes available) instead of thousands of articles [19:46:27] those "chunk" pages with #cargo_store [19:47:04] If we were to migrate from Cargo, we'd need to have 50000+ bot-updated templates to replicate current features [19:47:29] ... which is possible, but suboptimal. [19:49:07] I feel like I'm agreeing with your setup but doing so very poorly lol [19:51:57] That's one of the things I'd miss, being able to run the Cargo maintenance scripts via command line as needed. It seems to be only very large database tables that fail repopulation via the user interface, though. [19:53:42] From the UI sometimes it grabs 10 pages, other times 291, sometimes all pages with store to a table. Have no clue why [19:53:53] Oh, strange. [19:54:29] Eventually it'll get it right but 🤷‍♂️ [19:54:36] We've been parked on the same (older) version for a while (off MH) and it only had issues with the _pageData table I think it was [19:56:05] That one was too large for the interface, timed out and had to be manually run, though understandable as it was quite large. [19:56:25] Something like 17k pages in it [19:56:49] that is quite big [19:57:14] wonder why the interface has that issue but not the command line though [20:00:37] text editing is still usable on MH? [20:00:55] yes [20:01:08] you should still be able to edit pages [20:01:13] Request timeouts affected ours, which doesn't occur when run via command line [20:01:25] you just cant do anything with files or uploads [20:02:16] Errors in Cargo statements seemed to be a problem when run via web interface also. [20:04:14] instead it stops at about 5300 - far more reasonable tho since that's outta 5600! [20:15:08] <🩻✯𝙎𝙚7𝙑𝙏𝙝𝙚✯🩻#1595> Can I request a name change? [20:15:22] <🩻✯𝙎𝙚7𝙑𝙏𝙝𝙚✯🩻#1595> @Juwi [20:26:50] Juwi is discord only admin [20:27:01] what kind of name change? [20:29:30] All users should have the ability to change their own nickname on this server. If you are referring to a name change on wiki, go to [[Special:GlobalRenameRequest]] [20:29:30] https://meta.miraheze.org/wiki/Special:GlobalRenameRequest [20:29:30] [20:31:44] <🩻✯𝙎𝙚7𝙑𝙏𝙝𝙚✯🩻#1595, replying to Legroom#2748> A name to change to my current username [20:32:27] if you mean on Miraheze then go to link posted just above your message [20:33:14] <🩻✯𝙎𝙚7𝙑𝙏𝙝𝙚✯🩻#1595, replying to Legroom#2748> I've requested the name change [20:33:40] Now you have to wait until a Steward processes the request, which can take time [20:35:10] @Orduin can you rename the "Administrators" role to "Discord Administrators"? [20:46:00] <:DoneMH:775407710400675940> [20:46:06] Thanks! [20:50:12] also done moderators [21:17:26] Any news on images [21:28:57] sre will post when there is [21:34:35] Only real update for now is [21:42:19] Style question: Making a page on an election for a fictional nation. If I want to stay in line with the style guidelines of Wikipedia proper, as are adopted by most independent wikis, would I title it "New Acadia presidential election" or "New Acadian presidential election"? [21:44:59] Anyone know if there's a way to make Semantic Mediawiki not follow redirects when getting the value of a property from a page? We were storing data on redirect pages because Cargo lets us do that and it allowed data to be stored on a more specific level for stuff that was redirected to collective/group pages, but it seems that SMW doesn't quite work that way. [21:45:17] [1/2] It looks like either is technically appropriate, see: [21:45:18] [2/2] https://en.wikipedia.org/wiki/Wikipedia:Namingconventions(governmentandlegislation)#Electionsandreferendums [21:45:40] Essentially `"[date] [country name or adjectival form] [type] election/referendum"` [21:46:48] Huh. Ok [21:46:54] Guess it doesn't matter [21:47:20] I'll just do what I think sounds better which for me is the adjectival form [21:48:32] Neat bit of minutia, though. Wasn't aware of (but always assumed that) such pages existed to codify those conventions. [21:56:43] Ok [21:56:46] New problem [21:56:57] What happened to the file I uploaded? [21:57:58] #announcements all uploads are failing right now [22:01:35] is the only way to profile lua by having it invoked on a page? Trying to work around not having semantic scribunto and would like a json output from #ask instead of an html table. Only solution I came up so far is having #ask output json, then parse it in lua but that sounds incredibly inefficient [22:08:33] another error message if this helps fixing it [22:10:26] 503's are expected (kind of) until the file servers get fixed [22:35:34] for debugging purposes, you can invoke it directly (without saving a page): open the Module for editing and use the "Debug console" form at the bottom [22:45:09] I'm attempting a fix where it can detect this and draw the data from a source in another namespace I made specifically for the compound entry special case, but I can't seem to make it store data there...I wonder how the Spore Wiki does it [22:45:27] (the spore wiki is very namespace-happy) [22:53:23] oh, turns out that requires editing settings that I do not have access to [22:53:46] to phabricator I go I guess [23:10:35] Hmm that's helpful to know if the output is what I want. But it's there a way to figure out how much memory or how long a function takes here? [23:36:21] well that sucks [23:37:19] wait a sec [23:37:26] did adblock pwn the sitenotice [23:37:29] 😛 [23:41:20] I know a few adblocks completely hide all MediaWiki sitenotices [23:41:29] they target .mw-dismissible-sitenotice [23:44:20] disabled it and it's still gone :PPPPP [23:44:45] it also pwned ratepage