[10:54:51] http://commons.wikimedia.org/wiki/File:Aeroporto_di_firenze_check_in.JPG [10:55:14] could "<", ">"'s be removed from camera model? [10:55:30] the first and last character breaks metadata linkage [15:22:51] ToAruShiroiNeko, not sure that's really a question for here :) [15:22:55] I can do it by altering the picture [15:23:38] then again that would require me to make a contribution to commons; a significant barrier to me actually doing work [15:24:33] ? [15:25:44] (08:09:23) (ToAruShiroiNeko) [07:07:59] could "<", ">"'s be removed from camera model? [15:25:55] I'm not sure how many hours I am behind, my server's clock is totally wrong [15:27:54] that won't work :P It forwards CTCPs to me :) [15:28:09] some day I'll get around to fixing it, I'm just lazy [15:57:08] Shirik the issue is this is mahcine generated [15:57:20] so the camera itself is adding the < >'s there probably [15:57:26] which breaks linkage [15:57:52] perhaps a parser for some such camera models [15:59:07] also what would happen if someone tries to login to my account with an incorrect password multiple times to predict it? [16:03:40] would I be warned? [16:13:09] no [16:13:22] see https://bugzilla.wikimedia.org/show_bug.cgi?id=9838 [16:13:46] what's the page with such image? [18:46:14] I'm working on regex support for wgWhitelistRead and have a predicament. To get this to work for everyone in any langue, my regex needs to support UTF-8. However, to do that, the end user's PCRE needs to be compiled with --enable-utf8 and --enable-unicode-properties I believe. I am unsure if this is a requirement that should be imposed on the end user. Anyone have any thoughts as to how I should proceed? [18:47:01] Should that requirement simply be detailed in the wiki page for the wgWhitelistRead variable? [18:47:24] nullspoon: Could you run the regex on the backend somehow? [18:47:56] * marktraceur isn't totally sure about what you're doing, so sorry if that's unhelpful [18:48:18] marktraceur: No worries. Lemme shoot you an example... [18:48:51] $wgWhitelistRead = array( '/Main_Page/', '/^Special:/'); [18:49:00] Right [18:49:08] I'm modiging the code in includes/Title.php to allow for regex [18:49:17] OK [18:49:17] it already works, but doens't account for UTF-8 characters [18:49:28] if a UTF-8 multibyte character is used, it fails to match [18:49:29] Hrmmm [18:49:36] can't handle \p, \x, and \P [18:49:52] unless of course the end user's instance of PCRE has the configure switches previously mentinoed [18:50:11] nullspoon: And where is the client-side running that code? To determine whether or not to link something? [18:50:51] how do I read https://gerrit.wikimedia.org/r/#/c/12782/ ? [18:50:55] what is the code? [18:50:57] merely curious [18:51:00] Or this is all server-side, and I'm reading it wrong [18:51:12] oh, my apologies. When I mean end user I mean the admin of the wiki instances [18:51:17] nullspoon, you can assume pcre has utf-8 support [18:51:26] Platonides so is login abuse thing implemented? [18:51:27] mediawiki already uses some utf-8 regex [18:51:45] ah okay [18:51:55] I wasn't sure. Figured it would be faster to ask ehre than to read through all of it. [18:51:56] ToAruShiroiNeko: Click on the "Diff" links to see the code [18:52:00] * nullspoon is a newb [18:52:28] nullspoon: Everyone is at some point :) [18:52:28] ToAruShiroiNeko, click Side-by-side or Unified in the file list [18:52:44] ha [18:52:49] I am trying to find the diff link :/ [18:52:57] it's called Unified [18:53:03] because it's an unified diff [18:53:05] marktraceur and Platonides: Thanks so much for your help guys. I'll continue assuming utf-8 support. [18:53:16] side-by-side is another type of diff [18:53:17] I dont want to somehow break it [18:53:21] you can view the column header [18:53:23] you can't [18:55:04] Ah I see them [18:55:06] nice [18:55:15] so about the pasword hack thing [18:55:24] I am not sure whats going on in that bug [18:55:41] https://gerrit.wikimedia.org/r/#/c/11739/ [18:55:46] seems like it is on gerrit? [18:57:37] yes... [18:57:57] it may be easier for you to read bug 9838 to see what's it about [18:58:16] it is discussion since 2007 [18:58:21] scarily long :p