[05:04:39] hi, can someone explain to me if I can downloand and implement the https://en.wikipedia.org/wiki/Template:Infobox_person template? it looks really nice but not sure if its wikipedia exclusive [10:10:09] How can I change the template that shows: Create the page "page" on this wiki! [10:10:28] That appears when you search for a page that doesn't exist. [10:12:31] How do I create a permissionless group that I can then add people to? [10:16:04] Mayzie: they need a permission, you can reuse something like view or edit [10:18:17] Okay [10:18:41] p858snake: And will I be able to assign users to that group using the GUI? [10:18:47] Or will it have to be PHP code again? [10:18:56] / hardcoding users to groups [10:19:34] you need to give a group permission to grant the group to others, but yes the web interface [10:22:38] Cool. [11:13:33] MediaWiki really needs to improve on its documentation... [11:23:26] Trying to setup short URLs with nginx, and no stylesheets or images are loading.. [11:28:07] Please help? [11:32:38] Mayzie: what's the problem? I'm no expert but you may give it a shot [11:33:53] I applied the Redworks nginx configuration. [11:34:07] However now when I go to the site, a stylesheet isn't loaded and no images load either. [11:34:23] It's successfully contacting load.php, it's just not getting anythign in return. [11:35:15] maybe a piece of code from your vhost config would help [11:35:31] Mayzie: https://www.mediawiki.org/wiki/Manual:How_to_debug [11:36:00] https://bpaste.net/show/1eed88f5926b [11:37:33] andre__: It's not showing me any errors. [11:37:44] Mayzie, what is "it" in your sentence? [11:37:59] The browser, when I go to reload the wiki. [11:38:24] Mayzie: Okay. But why do you tell me? :) [11:38:31] Because you linked it? [11:38:32] Did you check the error log of your webserver software? [11:38:36] Mayzie, no, I did not. [11:38:37] Yes. [11:38:45] Yes you did: Mayzie: https://www.mediawiki.org/wiki/Manual:How_to_debug [11:38:48] Mayzie, I linked to a page explaining several ways to debug problems. [11:39:01] Not to a page that says "look at what your browser shows". :) [11:39:06] Mayzie: could you paste your LocalSettings.php ? [11:39:18] I applied the "PHP errors" section. [11:41:31] mefistofeles: https://bpaste.net/show/a714583f19de [11:43:24] Mayzie: sorry, I don't see anything wrong [11:43:48] :-/ [11:44:08] but don't trust me xD [11:44:19] Why does nothing ever work for me? [11:45:23] Mayzie: lol, it happens to all [16:24:53] Hello, I'm interested in locking down API read access without modifiying generall read access. I've looked through this, however I don't think it has the granularity I need. [16:24:53] https://www.mediawiki.org/wiki/API:Restricting_API_usage [16:24:54] Is there a common way to do this by modifying the LocalSettings.php or Api.php? [16:25:27] You should be making changes in LocalSettings in nearly all cases [16:27:10] well, I could change the settings [16:27:11] # Prevent anonymous users from reading and editing the wiki $wgGroupPermissions['*']['read'] = true; $wgGroupPermissions['*']['edit'] = false; [16:27:11] that would be overkill for me though. I'm exposing the API with a VIP, but I'd like it to require authentication without requiring the normal site to requre it to read by default. [16:27:44] If that's not possible, I understand, but I thought I should ask an expert [16:29:22] VIP? [16:29:34] Virtual server on a load balancer [16:29:45] I'm using it to expose just the api to the outside [16:31:10] configure authentication to api.php via your webserver? [16:31:15] (such as apache htpasswd) [16:31:40] or just don't bother, as the read api doesn't expose anything that you cannot get by scraping page content [16:31:57] that's exactly what I'm preventing [16:32:11] I don't follow [16:32:21] it's impossible to prevent someone from scraping page content [16:32:43] you can make it really annoying and frustrating, but you can never outright block it [16:32:55] given that, what's the point of trying to restrict it at all? [16:34:38] if you're worried about server resource usage by spiders, just block api.php in your robots.txt (actually I'd do that anyway) [16:34:39] I work for a managed services company. With proprietary information for customers. I make scripts that automatically collect data and build pages for our private wiki. [16:34:39] I can expose the API alone with a rule on our load-balancer fronted by a public IP. This allows me to run my collector/page building scripts from customer environments. [16:34:39] At the moment, only the API could be reached publicly, however I want to prevent get requests without authentication [16:36:46] I like the htpasswd idea though [16:36:49] a) mediawiki is a terrible platform for what you're trying to do, it wasn't designed to allow users to only view certain parts of a wiki, b) if you're set on mediawiki, the easiest way of going about that is to deny '*' read access (thus requiring login), use an extension like Lockdown to restrict page access to uesrs/groups, give your staff/employees full view rights, and create accounts for customers that only have access rights to the [16:36:49] pages they should be able to see [16:37:16] I understand. It is what I inheritted. [16:37:29] Is there an alternate wiki that you'd reccommend for my purpose? [16:37:42] htpasswd would give full access to everything to anyone who has credentials, so a customer could then query data about other customers [16:38:00] no customers have any credentials for the wiki [16:38:06] they never will [16:38:11] it is strictly for our employees [16:38:30] I understand, but I assume that you're giving your customers the utility to query data [16:39:01] this means they will be able to either inspect network traffic or reverse engineer the utility in order to view whatever access credentials are being used, and then use those credentials to make their own API queries (that aren't limited by your utility) [16:39:06] no, I'll invoke it on their machines in their environment. Credentials aren't built in. [16:40:03] probably with Ansible [18:53:55] people [19:13:45] Hullo peeps [19:14:32] Anyone know if there is an extension or anything out there that lets me only allow patrons of a certain creator to have an account on my mediawiki? [19:15:15] ConfirmAccount is maybe somewhat in that direction [19:15:57] Sweet, I'll look into that. Thanks Bawolff [19:16:40] oh right, that's the extension to make it so new accounts need to be approved [19:17:15] yeah [19:17:17] which would be fitting considering our current lack of automation in other areas [19:17:32] I'm not sure what you mean "patrons of a certain creator" [19:17:51] Do you just have a group of people who are allowed to create accounts? [19:18:07] Oh, yes sorry. I meant from Patreon [19:18:09] Because in that case you can just restrict the createaccount user rights and create everyone's accounts by email [19:18:16] oh [19:18:29] I don't think there's anything for that, you'd probably have to make your own extension [19:18:44] Oh god, that's going to end in fire [19:33:18] okay, well, in the meanwhile until I find time to make an extension to that effect, I need to find a different way to mitigate spam accounts. [19:34:11] I've been looking into trying to put a CAPTCHA on the registration page, but it seems more extensions with CAPTCHAs target the edit page elusively with the only one that doesn't requiring payment [19:35:44] ConfirmEdit targets registrations too [19:36:50] It does? Huh, from looking at the install guide I thought it was edit only [19:37:33] https://github.com/wikimedia/mediawiki-extensions-ConfirmEdit [19:37:38] $wgCaptchaTriggers['createaccount'] = true; // Special:Userlogin&type=signup [19:39:50] Ah, apparently I skimmed too softly [19:41:31] $wgGroupPermissions['accountcreator']['createaccount'] = true; $wgGroupPermissions['*']['createaccount'] = false; and add the accountcreator group to the people who are allowed to create accounts [19:41:51] then going to Special:CreateAccount would let them pass in a username and email address, and check the "send a temporary password via email" box [19:43:10] but if the reason is just to combat spam, then ConfirmEdit is a better bet [19:43:27] less hassle on the creator [20:33:55] !ops [20:34:15] !ops [20:56:10] hi I have a question [20:56:10] Hi ramenator500, just ask! There is no need to ask if you can ask, if you already asked the question please wait for someone to respond [20:56:24] I'd like to know how to remove the header on my first page [20:58:17] https://www.mediawiki.org/wiki/Manual:FAQ#How_do_I_hide_the_main_page_title.3F [20:58:51] http://imgur.com/MMAQgJN [20:59:06] I took a picture of it [20:59:10] is that the right thing? [21:03:32] I am trying to figure out how an authenticated user on my main website can be logged in on MediaWiki. Docs are pretty confusing. Can somebody point me in the right direction? [21:09:30] Anybody? ^^^ [21:20:21] what do you mean by authenticated? [21:23:02] a user who has logged in to my main website [21:23:21] I want that user to be logged into wiki as well. [21:24:39] !AuthHandler [21:24:44] !AuthManager [21:24:51] huh [21:24:53] !auth [21:24:53] http://www.mediawiki.org/wiki/AuthPlugin [21:24:59] nope, that's not the one I want [21:25:28] !AuthManager is https://www.mediawiki.org/wiki/Manual:SessionManager_and_AuthManager [21:25:28] Key was added [21:25:42] oh i guess user left anyways [21:27:50] I have read through that, but it’s a bit confusing. Seems like I’m the “provider” in this case and I want to implement a new SessionProvider? [21:33:42] oh, i didn't see the n in your name, you're actually still here :) [21:34:11] Yes, you want to write a new Session Provider [21:34:52] I would reccomend working from an example that already exists [21:37:02] Perhaps https://www.mediawiki.org/wiki/Extension:Auth_remoteuser or the OAuth extension [21:42:12] Thanks, I’ll check that out. Somehow I missed that extension. [23:39:17] Stupid question, but how can I find a list of people with commit access for mediawiki? [23:40:02] hi, a question: I'm setting up a private wiki and one of the concerns is the images hosted on the wiki being accessible directly (through copy-pasting the URLs) [23:40:14] are there any fixes for this or should it just be accepted?