[00:06:13] try mediawiki:Uploadtext [00:06:33] requires interface-admin role or edit-interface permission [00:12:17] congrats on the new plan :) [00:21:23] Where can I find that? Bit new to admining an all [00:25:45] Just do the existing notice but comment out the if check that looks at wmgSiteNoticeOptOut? [00:27:01] wsp axolittle [00:27:10] Why? That is not important to readers even. [00:29:46] wsp fiddle! <:trout:1109647945470582866> [00:30:32] very happy that we can wind down our migration meetings; suppose it never hurts to do the prepwork though :') [00:31:53] I wasn’t saying we should for the current notice just explaining how I think it’s done because it was said void didn’t know [00:33:18] i mean more like getting one of these guys in the summary to appear every time i wanna upload something [00:40:57] I just found out Miraheze will shut down but it will never happen. so why you are going to shut it down? it is because of Real Life Villains Wiki? [00:43:28] sigh, that was the resigning folks generating undue panic and chaos. everything is ok now [00:43:40] hi [00:44:23] example page: https://meta.miraheze.org/wiki/mediawiki:Uploadtext [00:45:12] i want to like to this create a wiki form [00:45:40] how can i phpbyadmin in cpanel finish this form? [00:45:53] phpmyadmin [00:46:26] i can not use sql [00:46:49] NDK, can you post a discord invite to the unofficial mediawiki server...I apparently can't [00:47:37] i can not inviste mediawiki [00:47:47] so can you help me? [00:47:58] build this form? [00:48:26] https://www.mediawiki.org/wiki/Communication#Chat [00:48:26] my email msnhinet8jps@gmail.com [00:48:34] ok [00:56:36] I'm referring to the extension CentralNotice, not the sitenotices from Sitenotice.php [01:02:08] We have wikisets set for opt-out vs not opt-out but with CentralNotice you can’t do a notice for say wikis with a specific extension or something of the sort [01:02:22] (We do sometimes do ones like that) [01:07:35] I think we have more pressing issues than sitenotices tbh [01:08:09] Thoughts to the future, may or may not hit The List™️ [01:08:20] Lol [01:11:47] Can someone delete this Wiki ASAP https://samandcat.miraheze.org/wiki/Main_Page [01:12:05] The request is in Stewards Noticeboard [01:31:52] There are more pressing issues than deleting a wiki. If its on the SN, it will be attended to eventually. [01:35:57] Why does discord default to super reacts <:moonch:794697217826095165> <:moonch:794697217826095165> [01:36:49] no clue [01:39:59] Ok [01:44:19] wants you to use them all [01:45:08] I pay 20$ for nitro i better get a good amount of em [01:45:16] PARDON [01:45:26] $20 😮 [01:45:41] 9.99 + server boosts and tax [01:46:11] I have 6 boosts total [01:47:45] oosh [02:05:51] 🫡 [02:36:42] Is there anyone here that offers coding services for MediaWiki. Only reason I ask is because I feel completely out of my depth when it comes to flashy stuff I rather hire or get volunteer help for heavy lifting for my wiki. If so PM me rates. free-$150? If anyone is interested it’s a Star Wars wiki. [02:39:51] Coding what exactly? [02:43:35] Main page, it would be for the Citizen Skin. Similar to the navigation grid of the Star Citzen wiki main page, search bar, along with featured article section, and display a “on this day”. Maybe even “recently edited” section similar to what Fandom has. I know it’s a lot but my skill level is very basic. Pretty much all I know is basic css. [03:05:04] Yeah this isnt a place to advertise [03:11:24] <.labster, replying to zppix> did you boost the servers at Miraheze? They seem faster than yesterday [03:11:44] No i threatened them with a hammer [03:11:50] Like i do the rest of my tech [03:12:55] I hear that's very effective [03:13:16] <.labster> Zppix used hammer. It's super effective! [03:34:55] SERVER fainted! [04:04:14] there's #professional-services in MediaWiki Discord server. [07:04:51] i hit delete to delete a file but i got technical issue error. then i refreshed the page to try deleting again but now i get "Error deleting file: Could not acquire lock. Somebody else is doing something to this file." and can't delete the file [07:08:04] nvm, i can delete it now. probably just miraheze lagging like usual <:PepeLaugh:1078914418953437204> [08:12:05] How do I do italic page titles? [08:31:12] Wouldn't that be a bad thing though--? :p [08:35:44] `{{DISPLAYTITLE:''Title''}}` [08:35:44] https://meta.miraheze.org/wiki/DISPLAYTITLE:Template:%27%27Title%27%27 [08:35:58] as long as the title letters aren't changed, it will work [08:36:29] (so you could do `{{DISPLAYTITLE:''Ti''tle}}`, but not `{{DISPLAYTITLE:''TiTle''}}` [08:36:29] https://meta.miraheze.org/wiki/DISPLAYTITLE:Template:%27%27Ti%27%27tle [08:38:21] it's a magic word, more info here: [11:03:18] just noticed Common.css and Common.js don’t work on Special:Preferences. is there a reason for that? :3 [11:03:51] and is there a way around it or no? :3 [11:04:10] no way around [11:04:17] security concerns [11:04:50] ah okay, may I ask what concerns? I’m not very techy but why is there an issue? :3 [11:04:51] that what I've been told, at least [11:05:00] neither I am [11:05:15] it’s okay someone will know :p [11:08:07] I found the Manual page, im just reading it atm :3 [11:09:27] [[mw:Manual:CSS#Styles_not_working_on_Special:UserLogin_or_Special:Preferences?]] [11:09:27] https://www.mediawiki.org/wiki/Manual:CSS#Styles_not_working_on_Special:UserLogin_or_Special:Preferences%3f [11:09:28] [11:10:54] it doesn’t go into detail about the specific security risks there though [11:11:17] [[mw: Manual:$wgAllowSiteCSSOnRestrictedPages]] [11:11:17] https://www.mediawiki.org/wiki/_Manual:%24wgAllowSiteCSSOnRestrictedPages [11:11:18] [11:11:26] I’ll stop spamming now sorry <3 [11:12:07] I probably confused that w/ UserLogin [11:12:32] still, I kinda can see point in disabling this stuff in preferences [11:12:57] me too, it just bugs me because it looks so different to the rest of the wiki now [11:18:17] if possible, i would like it so that those two pages use Common.css? idk if this is a realistic thing to ask though :) [11:18:30] throwing it out there for my wiki just in case :p [11:42:17] i guess you could get someones password using the JS [11:43:50] yeah [11:44:07] <_______________________________d> good morning forum [11:44:18] hi [11:44:20] still remember how FANDOM fucked up badly and started moderating js [11:44:34] <:pupCoffeeMH:766487840694599711> [11:48:17] not on Special:Preferences right? or am i missing something obvious [11:48:30] yeh on Special:Preferences [11:49:05] oh you get taken to a special page [11:49:19] Re: CSS on restricted pages: CSS also has possible attacks, with there being some hacks that allowed running JS inside CSS (not sure if they still work nowadays) and huge social engineering potential [11:50:03] as well as privacy concerns because of it being able to pull remote content from CSS [11:50:35] its okay if there's a legitimate scurity concern, it just bugs me a bit lol ^^; [11:52:24] I wouldn't enable CSS on restricted pages on any wiki, in fact there was one person that submitted a PR on GitHub enabling it on their wiki, I refused to merge it. And as far as I know no other wiki has ever had that enabled, so sorry thatgaypigeon, I don't think it will be enabled [11:52:40] its alright lol ^^ [12:22:31] While it does bug me on my own wiki too, I do completely understand it. Both security wise, but also ensuring that users can reach all customization options. As otherwise a wiki admin could potentially do like display: none; on some of the settings, making it at least harder for people to change those settings, and stuff like that [12:26:54] yeah, I suppose [12:38:39] Well it took until 2018 for Fandom to enable HTTPS on their site, something literally everyone (Miraheze included) beat them to years earlier. [12:39:17] Whenever "Fandom" and "security" are used in the same sentence, it's almost never for something positive. [12:39:39] lmao [12:40:03] But yes, don't enable custom JS and CSS on restricted pages like Special:UserLogin. Too risky. [12:40:13] okieee :3 [12:58:57] miraheze feels faster to me [13:14:32] I wouldn’t say it’s _fast_ but I’ve noticed it recently being faster than before XD honestly not sure how or why it was so bad before but even normal pages were 2+ second load time and recent changes was sooo slow. Meta with local cache disabled is down to 1.1 seconds load time for me not w [13:14:55] Is there a way I can request my wiki be deleted? [13:15:06] the load on swift is down + i/o as well 😄 [13:15:07] @paladox: If you're doing dns stuff can you take a look at bast141's reverse dns? It points to bast101.miraheze.org.? Just something that bugged me from back when I was an MWE [13:15:22] we now have all mw requests going through the faster servers [13:15:43] hmm [13:16:02] And then hitting a different page was 3.4 seconds yikes [13:16:14] [1/5] > root@puppet141:/home/paladox# host bast141 [13:16:15] [2/5] > bast141.miraheze.org has address 31.24.105.137 [13:16:15] [3/5] > bast141.miraheze.org has IPv6 address 2a10:6740::6:509 [13:16:15] [4/5] > root@puppet141:/home/paladox# host 2a10:6740::6:509 [13:16:15] [5/5] > 9.0.5.0.6.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.4.7.6.0.1.a.2.ip6.arpa domain name pointer bast141.miraheze.org. [13:16:26] I just had one page load with ‘waiting on server’ for 1.4 seconds before it even started loading [13:16:36] IPv4 points to that hostname, sorry [13:16:45] shitty rural ISP doesn't have IPv6 here [13:16:53] oh [13:17:25] that requires serverchoice to do that one [13:17:43] was this when things were slower, right? [13:17:49] not today [13:17:55] I’m pretty sure if I bypassed server side caching everything would be 2-4 seconds but I’m not sure how to do that :p. It’s amazing that the meta main page can start transferring in 30 ms tho and be fully loaded in 1 second [13:18:07] No just now unfortunately [13:18:09] oh [13:18:13] That seemed to be an outlier though [13:18:29] How do I tell if my requests are cache hit/miss? [13:19:06] oh good cache [13:19:13] i think the request will return cache hit [13:19:22] in the header [13:24:54] Why does the help center page return cp33 MISS UNCACHEABLE? [13:25:00] But also I’m pretty impressed [13:25:19] DOM load of < 0.4 seconds for server side cache hits and usually < 1.5 seconds for cache misses [13:26:22] If you're logged in the backend sends a Cache-Control header preventing caching, for security. If you're logged out you can see that it is cached (x-cache: cp32 HIT (1)) [13:27:39] I mean I thought that logged in users may see more uncached stuff but I was seeing cache hits on other pages while still logged in. It just specifically said help center was uncachable🤔 [13:28:15] I know things like load.php are not cached even when logged in, no reason not to. Other than that idk [13:28:45] *are cached even when logged in [13:28:46] Also it specifically seemed to only be on cp33 [13:29:03] I haven’t changed being logged in or logged out but just got a cache hit for the same page from cp32 [13:30:12] Now I’m only getting cache hits on cp32 lol.. [13:30:54] @paladox can I force a specific cp to serve my request and do you know why [[Help center]] would be uncacheable on cp33 but not cp32? [13:30:54] https://meta.miraheze.org/wiki/Help_center [13:30:55] [13:31:09] i don't think so. We load balance. [13:31:27] also i'm not sure why Help Center isn't being cached [13:31:53] you can send a request directly to cp33's IP, specifying meta.miraheze.org in the TLS SNI using OpenSSL's s_client [13:31:56] > cache-control: private, s-maxage=0, max-age=0, must-revalidate [13:32:37] hmm although i get it on like Help centers but then it caches (cache hits) [13:33:57] I think as long as it doesn't send no-cache it caches [13:36:16] oh there's no content-length on the help page?? [13:37:24] although i didn't see that on the first request to help centres which cached it and on the 2nd request got a hit [13:38:12] oh [13:38:49] I’m seeing cache hits for that page now also [13:38:55] (For cp33) [13:39:03] Took a while for me to see cp33 again [13:47:12] [1/2] @paladox doh, I figured it out and I’m a bit less surprised now but still somewhat confused. I clicked a link to [[Special:MyLanguage/Help center]] and I’m not surprised that the special page is PASS UNCACHEABLE from both cp32/cp33. but apparently after a 302 redirect it says the target page is uncacheable too? Until I do a refresh and load just that page not Special:MyLanguage the [13:47:12] https://meta.miraheze.org/wiki/Special:MyLanguage/Help_center [13:47:12] [2/2] n it says cache hit. [13:47:12] [13:48:28] also there is a content-lenght on those requests, what happened for you paladox was that the browser compared etags with the server and since they were the same it skipped downloading the page [13:49:52] ohhh [13:51:17] what target page nibblypup? I guess you mean one of the translations? because for me it just redirects to /wiki/Help_center, which is cached [13:52:22] I tried GET /wiki/Help_center/hu HTTP/1.1, which is not cached because no-one on cp33 read that page in a long time lol (X-Cache: cp33 MISS) [13:54:23] It just redirects to /wiki/Help_center for me also but on the very first load (coming from the redirect) I see the pass uncacheable. Subsequent loads say the exact same page is cached [13:54:48] Either my browser is showing the wrong headers or something thinks it’s uncacheable because it’s being redirected to from an uncacheable page? [13:55:06] When I get back to my pc I’ll check the actual load times if I can :p there’s a pretty big difference between cache hits and misses [13:56:05] Apparently the relay bot only shows my new discord username and not my server name, rip. This is NDKilla in case that wasn’t obvious :p [13:56:12] you're right, it says uncacheable when comming from a redirect [13:56:17] *coming [13:56:41] wait [13:57:07] yeh you're right [13:57:21] and hi NDKilla [13:57:41] I also just got pinged on irc when I mentioned myself so I accidentally saw what the bot says 🤣 [14:07:39] The bot has always used username, not displayname (or at least as long as I been here) [14:07:56] yeah, rip old username format [14:14:24] Special:MyLanguage/Help center takes me over 1 second reliably to load (300 ms for uncached special page, 800-1000+ ms for an uncached page load post-redirect) where as just going to Help center hits the cache 99% of the time and has first byte within 30 ms and full load within 400ms [14:14:47] Seems weird that redirecting from the special page causes it to always serve uncached content? [14:15:33] Agree but glad I got a decent-ish name lol [14:15:57] Because it isn't actually a redirect it seems [14:16:55] like there's no 30* code from the server [14:17:00] There is? [14:17:16] I see a 302 from Special:MyLanguage and then a 200 from the target page [14:17:30] not for me, I don't know [14:17:33] But the 200 target page is cache unmissable if it comes from the uncachable 302 [14:17:35] Weird [14:18:18] I’m looking at chrome dev tools, not preserving logs. At start of page load it resets and then the 302 is the first item and then the real page load shows up after for me [14:25:01] This is what it returns to me: https://privatebin.net/?6395901a6eaef112#591vBbUrtC3fFpe3qAk7jtG8kETKr2DG9eJhNzdpdDu8 It doesn't server any redirects, it just returns the Help_center article straight up. [14:27:20] it would probably serve a cached copy if it didn't do that, seems like a waste [16:47:28] Yeah the way Special:MyLanguage works breaks a good bit of stuff. The missing headers is something that I reported a while back [16:58:31] oh, so that language bug is still in place? [16:58:31] https://phabricator.wikimedia.org/T320363 [17:05:04] Everything at WMF be _restricted_ [17:05:12] 😦 [17:08:24] Normally the only tasks that are restricted are generally security ones so yeah [17:21:38] Oh that task was filed by us [17:21:57] By us I mean void [17:34:31] how do wikis avoid spam. i noticed our wiki never gets spam edits or comments but we dont have like... antispam features specifically enabled? is that a miraheze thing under the hood or what? i was going to transfer our wiki to my hosting but I wanted to know if there was like some kind of feature at work that i don't know about preventing this [17:37:44] If your wiki doesn’t have a lot of seo there’s not much reason to spam on it but miraheze does have some global abuse filters for anti-spam on all wikis [17:38:03] definitely aiming for good SEO though [17:38:09] fighting againt fandom, after all [17:39:14] But ofc size and type of a wiki also dictates how likely spam is to even be attempted [17:39:25] 1361 articles [17:39:36] idk man my unused blog with like 0 visits gets spammed all the time [17:42:15] :p a lot of the realllly basic stuff (most of it) is probably getting caught by abuse filters [17:42:27] i just want to know how to transfer those abuse filters [17:42:47] I would say majority of spam is caught by a filter, or other antispam measure [17:43:29] You can view what’s getting caught by the filters at Special:AbuseLog [17:43:43] You can’t edit the global ones although you may be able to opt out [17:43:52] i just want to know how to use the global ones [17:44:00] Some are public i think but most of them are hidden so that the spam people can’t see how to bypass them [17:44:01] They are enabled by default [17:44:05] like i wont have global miraheze spam filters anymore but i want to copy them [17:44:13] All the antispam ones are not public afaik [17:44:14] so that i can still have the antispam measures on my wiki once it is moved [17:44:22] oh is there anyone i can ask privately [17:44:31] We do not provide the content of the filters. [17:44:46] Thus why they are private [17:44:55] well i get that theyre not publicly available [17:45:08] They aren’t available period [17:45:12] i just want to protect my wiki [17:45:58] That's understandable, but they're treated as proprietary knowledge as knowing how they work would allow folks to work around them. [17:45:59] there must some general advices, somewhere ... [17:46:03] Not saying you are lieing or anything, but at the same time they would have no way to ensure you weren't asking in an attempt to bypass them [17:46:23] My suggestion would be to move the wiki and just see what spam pops up and then configure abuse filter [17:46:35] There probably is, I'll do some digging [17:46:37] It’s pretty simple regexes but ours are constantly tuned to prevent the abuse we see [17:46:51] oh no its regex [17:46:57] Making ours specifically available would defeat the point of them, no offense [17:47:03] thats fair [17:47:09] https://www.mediawiki.org/wiki/Manual:Combating_spam/AbuseFilter_examples [17:47:15] For one set of examples from MW [17:47:20] thanks [17:47:34] [1/2] https://community.fandom.com/wiki/Help:AbuseFilter [17:47:35] [2/2] For fandom's compendium of stuff which can likely be helpful to you [17:48:15] Rip opening a fandom ad on mobile and seeing ads <:xsob:912928935577276526> [17:48:25] We should just clone every fandom wiki and keep them up to date but no ads [17:48:28] our wiki is beautiful on mobile since moving it [17:48:48] i mean i had to make it that way but its a nice f-u to fandom [17:49:02] problem is SEO [17:49:14] Yeah, it'll get there with time. [17:49:28] I do recommend setting up WikiSEO with google search console and pointing it to your sitemaps, though [17:49:35] oh i have done that for sure [17:49:43] i even bought a doma1n [17:49:47] Nice. 🙂 [17:50:01] hasnt seemed to help much unfortunately [17:50:17] Definitely the #1 question we get from users, to the point where I even wrote a guide about it. [17:50:40] much appreciated [17:50:44] It's an uphill fight against Fandom if there's an extant wiki with plenty of content, but it can be one esp. if you have a community to organically distribute links. [17:51:48] yeah we have a discord, reddit, these all have the new wiki links. we advertise on twitter as well and im about to make a tumblr post but i was thinking of moving the hosting first. the hardest part about moving the wiki is transferring the localsettings and downloading all the extensions <:ChikaHigh:913162402277773362> [17:51:51] Go install confirmedit and configure questycaptcha [17:52:21] Idk if cleantalk is maintaining antispam [17:52:44] thank you. i do have confirmedit at least [17:57:29] [1/3] A while back I found out about breezewiki [17:57:30] [2/3] Layout ain’t perfect but better than the adpocalypse that fandom is [17:57:30] [3/3] https://antifandom.com/community/wiki/Help:AbuseFilter [17:58:39] i love that browser extension that redirects you to not fandom wikis [18:09:57] Yeah those are great [18:12:59] breezewiki don't have a page about fighting video games anyway, I ntice that my two wikis from Miraheze were deleted so impossible to restore them sorry for inactive [18:25:04] how long ago they were deleted? [18:50:21] <_______________________________d> small problem [18:50:32] <_______________________________d> ips can't edit my wiki even though editing permissions are on for them [18:52:12] We used to have a similar problem at once time on my wiki, it was some extension that was preventing it it, I don't remember exactly which one though [19:01:42] test [19:01:51] <_______________________________d> relay works 🙂 [19:22:33] uhhhh https://meta.miraheze.org/wiki/Requests_for_Comment/Changes_to_the_Dormancy_Policy_(4) [19:24:23] huh [19:27:17] why? [19:28:10] Sighs [19:28:25] This user thinks by creating rfcs they will pass their RfS [19:28:34] At least thats what I assume [19:33:14] This is certainly not a way to endevour one's self to a future RfS... [19:33:46] *current RfS [19:34:19] Oh, well this definitely doesn't help that case then [19:35:06] <.labster> [1/2] But like he says, [19:35:07] <.labster> [2/2] > The number 69 is a very nice number [19:35:36] Look, the number is nice sure, but there's no application of logic as to why this nice number should be used for the dormancy policy. [19:36:14] I think the current amount of days are alright [19:36:16] Because theres no reason to his rfc [19:36:33] Hes doing it because I told him he has no global community involvement [19:36:36] should we just oppose it or you'll just nuke it? [19:37:01] I can’t nuke it, since I voted, but any stewards welcomed to close it as disruptive [19:38:22] If you guys don't want rfcs, just lock the page and put a formal notice [19:38:26] I just saw their stewardship request, wellp [19:38:38] No we don’t RfCs just to have RfCs [19:38:50] If theres a logical reason thats different [19:39:14] <.labster> RFCs are fine. RFCs that affect technical matters without talking to a member of the tech team first are bad. [19:39:21] Just because “they haven’t changed since 2015” isnt a valid reason [19:39:34] [1/2] Yes, that's 1 thing. [19:39:34] [2/2] The other issue is there's more important matters and all have indicated policy discussions is for later. [19:41:02] Formally making people hold off on RFCs for valid reasons of new incorporating is fine [19:42:33] The only problem with this RfC is that it is unwarranted. [19:45:52] ^ [19:46:13] If it was an RfC with logical reasoning I would be perfectly fine with it [19:46:16] I have no problems with a well thought out discussion or proposal. [19:47:07] That being said his proposed changes would not work, way too long of timeframes [19:47:31] We want to avoid having dead wikis not stockpile them [19:48:46] Yeah Void or Agent would have to close as stewards as the UCP doesn’t apply to RFCs [19:49:10] <.labster> I guess we could just all vote it down and move on. Nothing discourages ppl like a wall of red. [19:49:41] feels like waste of time [19:49:49] It is [19:49:52] Agent already closed their other rfc [19:50:23] Yeah... it's a LOT of proposal clauses for no gain of worth. A drafting period could have made it semi-viable [19:50:29] But as framed, it's a waste of voters' time [19:51:03] And not what I would expect from a potential steward. [19:52:02] Well i mean a previous block on meta for disruptive editing would also be a huge red flag [19:54:50] Didn’t they explicitly say their only block was by abuse filter and they were manually blocked? 🫠 [19:55:16] No they were stopped by edit filter for trying to edit [[Finance]] [19:55:16] https://meta.miraheze.org/wiki/Finance [19:55:16] [19:55:23] They never mentioned the manual block [19:55:45] [[RfS]] [19:55:45] https://meta.miraheze.org/wiki/RfS [19:55:46] [19:56:22] Ah I guess they said they’re not currently blocked anywhere not that they’ve never been blocked [19:58:27] all those variants on the dormancy period thing hurt my head. [19:58:35] except 69 being a nice number [19:58:52] give them a warning for spam anyway [19:59:16] Waiting for the RfC to be closed then I’m issuing a final warning for disruptive editing [20:00:04] The RfC was already closed [20:00:09] Not this one [20:00:17] They put out two in the last few minutes [21:03:59] It's a troll, act accordingly [21:05:48] I don't think they're a troll; what's more likely is that they're a good-faith user who's unintentionally disruptive [21:06:44] They actually maintain a few wikis with MH and are otherwise generally in good standing, it's just gone a bit off the rails. [21:15:31] [1/3] I misread "RfC" as "KFC"-- 😭 [21:15:31] [2/3] Is it possible the user is deliberately doing this as a test of some sort? [21:15:31] [3/3] Anyway, I think the current days before "x" is okay, for the time being. I mean, if it's necessary, exceptions could be requested under certain circumstances, right? [21:19:36] Yes, dormancy exemptions can be requested via the steward's noticeboard in cases of seasonal editing or if the wiki's already content complete [21:44:01] Is there a way I can have my wiki deleted? We have migrated elsewhere [21:44:42] [[SN]] [21:44:42] https://meta.miraheze.org/wiki/SN [21:44:43] [21:44:43] Please request on [[SN]] [21:44:43] https://meta.miraheze.org/wiki/SN [21:44:55] Alright thank you! [21:58:21] <_______________________________d, replying to lichmaster98> eh [21:58:29] <_______________________________d> sex number haha funny [21:58:36] <_______________________________d> for 12-year-olds at least [22:20:13] perhaps a memorial exemption can be added via RFC once everything is all settled down? ie if a wiki is complete then we can preserve it as is and lock editing [22:22:02] [1/2] " unintentionally disruptive" what does that even mean, seriously? sounds like a round about way to excuse potential trolling behavior [22:22:02] [2/2] its possible they are doing this out of concern...this is why I suggested to formally lock new RFCs with a notice "RFCs are closed at the moment. Staff are busy with incorporation changes" [22:24:15] Dormancy exemptions havent changed for atleast 6 years, no reason to change it [22:26:30] This already falls under the 'content complete' state, the actual wording is 'substantial content' [22:28:19] This is a spin on one of the proposed terms to get away from using the term CIR (short for Competence is Required), which has been viewed as unnecessarily perjorative. [22:28:57] The longform version was 'good-faith disruptive user', someone who means well but is preventing orderly process nonetheless. [22:32:55] well was a mere suggestion, wasn't aware of what NotAracham said...though the wording seems rather unclear. if the intent was for content complete, then just state that instead... [22:34:21] hmm have met such users, some don't take a "no" very well and go on to become worse resulting in a ban because "I didn't get my way" [22:56:00] <.labster, replying to m3w> This sounds a lot like intentional disruption. There's a difference between approaching things the wrong way, and people who keep causing problems because they don't get their way. [23:01:48] ig that's where we see things differently, I just see disruption. You see different types of disruption - some you are willing to tolerate... [23:04:01] We have to tolerate some levels [[w:WP:AGF]] [23:04:01] https://en.wikipedia.org/wiki/WP:AGF [23:04:02] [23:04:31] <.labster> moderatin ain't easy [23:04:45] ok, I get you. insistent on following WMF to an almost t... [23:04:55] <_______________________________d, replying to zppix> nah [[WP:ABF]] [23:04:55] https://meta.miraheze.org/wiki/WP:ABF [23:04:56] [23:05:09] I simply just call it as I see it [23:05:27] <_______________________________d> assume bad faith [23:05:29] No… but its not a bad guideline in general [23:05:48] <.labster> honi soit qui mal y pense [23:05:56] I also call it like I see it, I also have a filter [23:06:24] <.labster> [[w:Honi soit qui mal y pense]] [23:06:24] https://en.wikipedia.org/wiki/Honi_soit_qui_mal_y_pense [23:06:25] [23:06:48] <_______________________________d> no, we must assume bad faith [23:06:58] was there a point to the "I also have a filter" notion? [23:07:01] Thats a terrible idea [23:07:12] <_______________________________d, replying to zppix> true 😦 [23:07:39] You can read into it as you wish, I said what I said [23:07:51] very well [23:25:11] i agree intent plays a role. someone who doesn't know they're being disruptive can be educated and their likely good intentions can be channeled into constructive efforts. ill-willed posters tend to stand out as that and, typically, resist rehabilitation. there is a big difference even if the impact may seem similar at first glance.