[00:00:43] Archimedes1560[m: re. "Can't you just take everything before first

" -- a page might not have an h2... I haven't looked at the code at all yet but my first thought would have been to take the first

, but it's probably not that easy... [00:01:25] If it has no h2 then everything is section 0 [00:04:36] true... that being said, taking *everything* before the first h2 (or the entire page if here is none) would be way overkill for a description meta tag, which is supposed to be around 160 characters apparently. [02:14:47] New +2 in mediawiki/ request: https://phabricator.wikimedia.org/T329000 [03:12:43] Platonides Archimedes1560[m, I was just looking at my old database and realized itis Mysql and the new server database is mariadb, and the old database charset is "cp1252 West European (latin1)", but the new maria database charset is utf8mb3, could these difference cause the problem? [15:30:13] hello I am sysop at italian wikipedia, I have a question, can we have more than one alias for a specific alias, for instance "ut:" and "user talks:" for "user talk:" namespace? [15:30:32] for a specific namespace [15:31:15] Yes, there may be more than one alias for a given namespace [15:36:05] are they heavy resource consuming? [15:36:31] namespace aliases don't consume any resources at all, it's just server-side config [15:38:22] thanks a lot [16:15:57] has any one seen this issue? say I have a page TestPage1, on the page somewhere I have a string "TestPage2" (the actual page doesn't exist yet), then I use VisualEditor to create the new page for TestPage2 using the link button, after the new page is created, if I go back to TestPage1 and refresh it, the hyperlink to "TestPage2" still shows red [16:15:57] (more of lighter pink), if I now try to edit TestPage1 with something and save it, the hyperlink to "TestPage2"  will become blue. [16:16:45] it is on 1.35 and 1.39 both [16:28:03] paulx: the page cache needs to be purged. It's usually done by a background job queue, but by default it may take several "page refreshes", depending on how many jobs are already on the queue [16:28:06] !jobqueue [16:28:06] The Job Queue is a way for mediawiki to run large update jobs in the background. See http://www.mediawiki.org/wiki/Manual:Job_queue [16:30:22] If you run the runJobs.php script from the command line, it should execute all pending jobs, and the link should appear blue if you refresh the page again. That will confirm the job queue works. Otherwise, there may be some other problem [16:31:35] Thank you all, let me try [17:21:38] Hey how to I start with the setup? [17:21:50] I have cloned the master branch [17:31:27] Pradnya: are you setting it up for development or for a production wiki? [17:32:03] for development: https://www.mediawiki.org/wiki/How_to_become_a_MediaWiki_hacker [17:32:22] for production: use the release tarballs instead [18:03:09] hey all - wmf release engineering is running a developer satisfaction survey; we'd love your input: https://wikimediafoundation.limesurvey.net/484133 (x-posted to a handful of channels) [23:09:09] is there a way to make sure that when a template is used, it begins on a new paragraph? it seems that when I use a template whose content starts with

, it messes up and creates a malformed collapsible... [23:09:55] I mean I know I could just start the template with line breaks, but that would introduce too many line breaks when it's used correctly. [23:11:24] `

`? [23:12:56] the parser will tell you when you're putting something in the wrong spot [23:13:14] ultimately, there's nothing you can really do to use it only in certain places [23:14:58] Also I don't think lack of line break before `mw-collapsible` breaks it [23:28:30] Izno: how will it tell me that? I'm not getting a warning or anything... [23:29:16] Looking at the emitted page source. [23:29:21] It's not obvious regardless. [23:29:32] But as I said, you can't do anything about it ultimately [23:29:44] Wikitext relies a lot on the users to do things right. [23:30:14] oh well, I'll just trust our editors then [23:33:39] by the way, is there no other way to query for the content of a page via SQL than to parse the content_address column? seems a bit weird that there's no direct foreign key to text.old_id [23:34:02] I'm still struggling to find a simple way to "grep" for wikitext throughout the wiki. CirrusSearch is way overkill. [23:35:37] my understanding is that base search searches wikitext [23:35:45] if it doesn't, no, you need to use CirrusSearch [23:36:02] apparently, it doesn't fully. can't find any uses of the tag for example. [23:42:22] Hello there! [23:42:47] I'm having a little trouble writing some html with if statements and would appreciate if anyone could help. [23:46:08] Writing some html? [23:47:16] HerrderZeit: Hi. HTML doesn't have if statements, so you must be using something else. [23:48:07] HerrderZeit: did you perhaps mean the {{#if:}} parser function in MediaWiki page code? [23:49:57] Yeah [23:50:04] another term is handlebars [23:50:10] but I don't think that applies here [23:54:18] https://codeshare.io/qP0DZR [23:54:24] I literally have no idea what I'm doing. [23:54:36] I'm trying to help build the wiki for the Sinden community. [23:59:24] HerrderZeit: does this wiki have a template named "if" which implements a special kind of if? because otherwise all the {{if...}} that you use, without the #, would be wrong.