[00:21:48] Quick question: Does the 'purge' function re-execute pages (As in, templates/magic words)? [00:22:16] Or does it simply update the text? [00:22:30] Or should I say, HTML. [00:47:55] Is there a standard way to allow for the use of a variety of Authentication mechanisms? For example, a wiki that users can log into using Shibboleth, CAS, and OpenID? [00:49:12] I've been looking through the extentions and they all seem to do it their own way, without any regard for whether other extensions are installed. I saw MultiAuth, but the code is no longer available [01:06:50] Hi, I kind of asked this question earlier but I didn't get a response. How would one manage to "re-execute" a page? I have a set of magic words, for instance, that collect some needed data by saving the page's ID to the data. The issue is that when you create a new page, the ID is 0, so it sets the value to 0 also. I need a way to re-execute the page so that it can input the page's proper ID. [01:07:56] Coraline_: That's called a purge. [01:08:06] Coraline_: Load the page with ?action=purge in the URL. [01:10:56] marktraceur: That seems to have done nothing, oddly. If the ID returns 0, then I set it to redirect to the created page with "?action=purge&update=dynamic" (And I check for $_GET['update'] to complete the action). [01:15:46] Hrm. [01:16:36] I'm gonna do quick rewrite to see if anything else may have been causing an issue. [01:17:58] Coraline_: I get 99 with {{pageid}} on a new page, though [01:18:04] So maybe you have a bug somewhere else [01:18:18] Coraline_: What version of MediaWiki? What extensions are available? What kind of page are you creating? [01:19:20] marktraceur: Normal page (No namespace), 1.24 (Alpha) [01:19:38] Weird. [01:19:42] Coraline_: You got it from git? [01:19:49] Yep. [01:20:02] It's also worth noting that {{pageid}} returns nothing on a new page. [01:20:06] Coraline_: Maybe try updating to the latest? We're at 1.25wmf3 now on the cluster. [01:20:10] Oh, weird. [01:23:32] marktraceur: Should it really be generating a page ID for a page that doesn't exist yet, though? Or does it try taking the the next available (Which would, assumingly, be the page's ID-to-be) and display that? [01:24:12] I'm sure it's smarter than that and actually waits for a DB ID... [01:28:53] marktraceur: Well is the page ID dependent on anything defined in the header javascript maybe? I've disabled Mediawiki's default output of header information. [01:30:38] Nope, existing pages can display IDs. Maybe the version I'm running now has an issue with that..? [01:34:48] marktraceur: I just tested on another wiki -- Using {{pageid}} on a new page also does nothing. What's your Mediawiki version? [02:01:06] marktraceur: I'm just going to do this the cheap way and select the "highest" page ID at the time and add 1 to it. The odds that two users will just happen to create a new page within the same milisecond are astronomical, and the execution time between writing the value and the creation of a page are so close that errors shouldn't happen. [02:02:04] Is it known that the writeapi permission is required for certain AJAX functionality on Vector? [02:02:13] e.g. page titles [02:13:29] GPHemsley: Vector no, Visual Editor yes [02:17:43] hmm [02:17:54] Betacommand: where does page watching fall? [02:23:24] GPHemsley: that might be, but I was unaware of the issue, you might want to report that [02:23:29] https://www.mediawiki.org/wiki/Manual:$wgAjaxWatch [02:24:03] on MozillaWiki we found that users who didn't have writeapi permission are able to add pages to their watchlist, but not using ajax [04:18:24] christi3k: watchlist is controlled by the 'editmywatchlist' permission, not 'writeapi' [04:21:42] legoktm yes, they are able to add pages to their watchlist, but not using ajax method [04:22:02] yeah, so remove the 'editmywatchlist' permission then [04:22:12] > Edit your own watchlist. Note some actions will still add pages even without this right. (editmywatchlist) [04:22:17] legoktm we weren't trying to prevent it [04:22:20] ah [04:22:36] we were trying to figure out why some users had a different workflow for adding pages to their watchlist [06:31:19] Check Spambots (Version 0.1) <> MysteryFCM et al. http://wiki.hl7.org/index.php?title=Special:Version [08:40:43] still no replacement for Asirra? [08:42:37] hi [08:43:49] I use mediawiki for personal projects. When I click on 'edit' small area (height is 5 characters) opens where page can be edited. How to make this area much bigger pernamently? [08:44:41] never mind - just found it ;) [13:53:42] So what's with the use of $GLOBALS in some extensions to setup hooks? [13:54:05] The comments say something about delayed initalization, but I'm not entirely sure how that's working. [13:54:18] Example? [13:54:53] https://git.wikimedia.org/blob/mediawiki%2Fextensions%2FSemanticForms.git/b4935cd735bdb8e67c6eb18661eedf511d3c8315/SemanticForms.php#L97 [13:55:31] so $GLOBALS is a super global [13:55:38] I think it's done so it works with normal extension loading [13:55:57] And also via composer stuff, which IIRC, wraps things so they're not loaded in the global scope [13:56:57] Oh, I didn't know that about the composer stuff.. interesting [13:57:27] Rosencrantz, Reedy: Yes. The problem is, that inclusion for loading via composer is done in a function where all the globals, available in well global ns in LocalSettings are not available. [13:58:10] Ah ha… ok. [13:58:19] And forms relies on smw, which is loaded via composer.. ok [13:59:15] thanks Reedy and FoxxT [14:00:51] Rosencrantz: SF is a somewhat special case. Nothing really to do with SMW. SF is on the way to becoming Composer loadable. But that is currently on hold. [14:01:15] Oh! Didn't know that. [14:01:55] If you think it should be loadable via Composer, poke Yaron. Don't tell him I told you that. :) [14:02:04] Haha [17:55:52] hello all [18:00:59] is there any way to link to an image's full size version? [18:01:07] i know about the link= option [18:01:23] I think you need to use external links for that usually [18:01:48] Reedy: so the problem was that I didn't specify the entire path? [18:02:04] I tried doing something like /w/images/7/70/Screen-Res-Chart.jpg [18:02:17] [https://upload.wikimedia.org/wikipedia/commons/1/11/Cucullie_de_la_scrofulaire_MHNT_Chenille.jpg File] [18:03:21] Reedy: ahh perfect, thanks [18:03:22] {{filepath [18:03:34] i used the full URL for the link= option and it works now [18:03:42] {{filepath:Filename.jpg}} [18:06:27] Hello everyone! [18:06:36] how do I get the raw wikitext from a Content object? [18:06:54] I am trying to install Parsoid [18:07:19] Which is to be done in the /usr/lib/ directory [18:07:28] Which is to be done in the /usr/lib/ directory apparently [18:08:57] But the command without sudo wont be able to do that because of permission denied error. Is it okay to install parsoid in /usr/lib/ with superuser permissions (using sudo) ? [18:09:51] Why wouldn't using sudo ok? [18:10:15] just confirming :) [18:10:19] Reedy: [18:13:22] Reedy: So, I can, right? [18:13:31] Why not try it and see? [18:14:12] sure! :) [18:27:25] Cloning into 'parsoid'... [18:27:25] Permission denied (publickey). [18:27:26] fatal: Could not read from remote repository. [18:27:26] Please make sure you have the correct access rights [18:27:26] and the repository exists. [18:28:03] This is the error I get on installing parsoid in /usr/lib/ [18:28:40] what are you cloning over? [18:29:56] Reedy: sudo git clone ssh://ankita@gerrit.wikimedia.org:29418/mediawiki/services/parsoid [18:29:56] I am trying to install Visual Editor which requires parsoid to be installed too [18:30:20] do you have a working gerrit account with public key uploaded? [18:30:25] yes [18:30:30] that error may also be intermediate [18:30:38] had that happen to me. try again after a minute or two [18:30:53] infact this command worked when i installed in another directory (not root) [18:35:02] ankita_: you where trying to install it as root? well, root probably doesn't have *your* public key [18:35:14] use the https user to clone as root [18:35:21] it'S read only, but doesn't need a cert [18:37:25] Hello! DanielK_WMDE__ [18:37:25] Thanks, worked :) [18:37:55] ankita_ / DanielK_WMDE__, https is actually read-write -- use the username & password from https://gerrit.wikimedia.org/r/#/settings/http-password [18:38:43] valhallasw`cloud: Has anyone EVER got that to work? :P [18:38:50] Reedy: {{worksforme}} [18:39:01] Reedy: you just need to realize it's not your normal password :-p [18:39:07] valhallasw`cloud: hello, thanks! where is the password to be included? [18:39:18] valhallasw`cloud: I did [18:39:25] ankita_: if there's no password, you might need to click 'generate password' [18:39:26] Just last time I tried it, it didn't work [18:39:53] Reedy: hm, strange. Just with git push origin HEAD:refs/for/master or with git-review? [18:40:01] I honestly can't remember [18:40:12] Nor care that much [18:40:16] SSH works fine [18:40:27] valhallasw`cloud: i got the pwd, but where is it to be used in the clone command? [18:40:42] ankita_: ah, you only need it if you push [18:41:01] ankita_: but you could do git clone https://username:password@gerrit.wikimedia.org/r/mediawiki/extensions/examples [18:42:48] valhallasw`cloud: i have to clone parsoid to /usr/lib/ which gave the permission denied error, i used https instead, which worked [18:43:29] valhallasw`cloud: can i use the last command for cloning parsoid? [18:43:48] ankita_: no -- if your clone worked, you're done [18:44:09] k, cool :) [18:44:16] I'm sorry, I shouldn't have brought it up -- it distracted from your real issue [18:44:47] valhallasw`cloud: i was wondering if using https or ssh would actually cause a difference while development? [18:45:47] ankita_: yes, but only slightly. With https you would use the password in gerrit to authenticate; with ssh you use a public key -- which is basically the same, but it's a file instead of a password [18:48:07] valhallasw`cloud: oh, thanks :) i didn't know that [19:30:03] Fatal exception of type MWException [19:30:18] !debug [19:30:19] For information on debugging (including viewing errors), see http://www.mediawiki.org/wiki/Manual:How_to_debug . A list of related configuration variables is at https://www.mediawiki.org/wiki/Manual:Configuration_settings#Debug.2Flogging [19:38:01] Using the debug options helped me with: MediaWiki 1.24 required--this is only 1.24alpha [19:38:40] This is an error caused by adding the config settings for VE to LocalSettings.php [19:39:09] I tested by removing the corresponding lines of code and it reworked fine again [19:39:58] parsoid seems to have installed right as the command `node api/server.js` works fine on localhost:8000 [19:40:26] i followed the instructions right for VE install [19:40:46] any suggestion where I might have gone wrong? [19:44:34] I found this relevant: https://www.mediawiki.org/wiki/Thread:Extension_talk:VisualEditor/VisualEditor_does_not_load_(REL_1.23)#VisualEditor_is_not_loading_.5BRESOLVED.5D_Problem_is_web-interface_at_git.wikimedia.org_46964 [19:45:05] But i had already ran `git submodule update --init` while the installation [19:45:13] Hi, I just "discovered" the way hooks work and now I have a question: Is it possible to access a value inside the programming you're "hooked" to? I'm using 'pecialMovepageAfterMove' -- There's some additional database data (Based on the page ID) that should also "move" alongside the page. [19:45:52] *'SpecialMovepageAfterMove' -- All I've managed to accomplish thus far is "echoing" a value I put in. [19:50:29] Ah nevermind, I believe I've got it! I'll return if I have further issues. [19:53:41] Help, please :) [19:54:24] i dont really think that this would require an updated version of mediawiki, would it? [19:55:51] Visual Editor requires a bleeding edge version of MediaWiki (at least if you download from git) [19:56:21] Vulpix: Not neccesarily. You can use the 1.23 Branch [19:56:25] I've been using that since 1.23 came out [19:56:27] And it's been fine. [19:56:59] Granted, there are some bugs here and there, but nothing too terrible. [19:57:26] I had installed everything using git cli [19:59:14] which version will that provide? Should I issue a git pull ? [19:59:15] ankita_: have you checked for javascript errors? [19:59:21] and did you check file permissions? [19:59:56] no, i tested for php errors, also checked apache error log [20:00:22] do check js errors. VS is mostly JS [20:00:34] VE, i mean [20:01:10] files are all 644 [20:01:38] also, VE needs parsoid, which needs node.js. do you have that set up? [20:02:06] DanielK_WMDE__: yes, parsoid runs well on localhost:8000 on issuing node api/server.js [20:02:36] good then. back to checking for JS errors. and look in Special:Version to verify that VE shows up there [20:07:51] DanielK_WMDE__: I checked for JS errors on the browser console and it shows nothing [20:08:26] Could you please explain what do you mean by Special:Version, as in how to verify that? [20:25:06] I need some help with working on a bug, is this the right channel to discuss that? [20:25:46] I am just starting with any bug for the first time (an extension, in this case) [20:28:29] ankita_: this is more a support channel, depending on what do you need you may want to ask on #wikimedia-dev [20:28:51] although most people are in both channels :P [20:33:33] sure :) [20:58:12] How can you create new user groups? [20:58:28] !groups [20:58:28] For information on customizing user access, see < http://www.mediawiki.org/wiki/Help:User_rights >. For common examples of restricting access using both rights and extensions, see < http://www.mediawiki.org/wiki/Manual:Preventing_access >. [20:58:41] Agent_Isai: ^ [20:59:34] thank you [21:02:32] eh, I mean like make a new user right :/ [21:03:01] Oh [21:03:03] Found it [21:04:18] Hello, I have I question for you guys. I'm an administrator on a MediaWiki Wiki, but don't have any access to the server/shell/database. I use pyWikiBot from my own server to purge a lot of spam pages with listpages.py + delete.py. Now I have a list of thousands users to ban (one per line in text files). Do you know a good way how to leverage the MediaWiki API to mass ban users ? Thanks [22:04:06] hexmode: is Debian still on 1.19? https://wiki.mediagoblin.org/Special:Version [23:09:28] I am having issue where the images have stopped loading on the wiki. If you go to the direct url of the image, it works fine. [23:09:38] Url to article http://blmerc.com/wiki/Organization [23:25:13] Ryahn: I see the flow chart in that article [23:25:25] It's small, but there. [23:31:07] Ryahn: I see you removed the thumbnail border and made the image much smaller [23:31:19] But another user fixed it.