[00:09:16] <ircuser-3>	 Can anyone confirm that wikimedia.org SSL cert changed to RapidSSL in January?
[00:11:42] <Reedy>	 For what subdomain?
[00:12:23] <huh>	 I thought the cert was for *.wikimedia.org
[00:12:24] <Reedy>	 There's a *.wikimedia.org cert that isn't RapidSSL, but wikitech.wikimedia.org (for example) is RapidSSL
[00:12:30] <huh>	 ah
[00:12:38] <Reedy>	 huh: But it's not used on all wikimedia sites/services
[00:13:33] <Krenair>	 I think most wikimedia.org sites use the *.wikipedia.org cert under an alternative name
[00:13:50] <Reedy>	 Aye
[00:14:43] <Reedy>	 ircuser-3: ^ ?
[01:57:57] <Newyorkadam>	 hi Gloria 
[01:59:38] <Newyorkadam>	 Gloria: :o?
[02:05:17] <Newyorkadam>	 Gloria: :c
[02:21:09] <ircuser-3>	 Thanks, was browsing wikimedia.org extensively today when unexpectedly a new cert was presented.
[02:35:07] <Gloria>	 Hi Newyorkadam.
[02:35:59] <Newyorkadam>	 hi Gloria 
[02:36:05] <Newyorkadam>	 I requested labs access
[02:36:15] <Gloria>	 Good job.
[02:37:33] <Gloria>	 Newyorkadam: Does someone have to approve the account?
[02:37:38] <Newyorkadam>	 Gloria: I believe so
[02:38:00] <Gloria>	 Lame.
[02:38:44] <Newyorkadam>	 Gloria: can you approve?
[02:38:51] <p858snake|l>	 shell access is done pretty quickly these days
[02:39:46] <Newyorkadam>	 https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Access_Request/Newyorkadam
[02:40:34] <huh>	 Oh, you requested /tools/ access
[02:41:11] <Newyorkadam>	 yeah
[02:41:29] <Gloria>	 Tooooooooooooooooooooooooooools.
[02:42:20] <gry>	 Newyorkadam: that's what you ask at #wikimedia-labs I belive.
[02:42:26] <Newyorkadam>	 I asked
[02:42:52] <gry>	 Ah, good. There are some volunteers handling the queue, they were reasonably quick when I was around, but that wasnt weekend.
[02:46:45] <gry>	 Gloria: what are you doing this weekend? You seem a bit excited.
[02:48:10] <Gloria>	 Just livin' the dream.
[02:48:11] <Gloria>	 You?
[03:03:27] <gry>	 That doesn't seem overly specific. (I appear to be creating some articles at Wikipedia as a sister project wanted an article with inline links to them, but they can't be external).
[03:03:44] <gry>	 Although it does sound exciting, as I could notice.
[03:05:01] <gry>	 I'm trying to find means to print a webpage with hyperlinks to PDF -- the export to PDF tool appears to give ugly output with printable links insted of clickable ones, and even if they are clickable, they're not blue. I might have to resort to web browser add-ons to print a web page "as is".
[04:53:22] <Newyorkadam>	 Betacommand?
[06:30:55] <chistopher_>	 i want to know how to load and run a third party bot imy server
[10:42:09] <fale>	 does toolbeta has an accessible webserver?
[11:30:15] <huh>	 Anyone around, kinda urgent?
[11:31:12] <huh>	 I got MW error here: https://meta.wikimedia.org/wiki/User:Aldnonymous
[11:31:47] <huh>	 MediaWiki internal error.
[11:31:49] <huh>	 Exception caught inside exception handler.
[11:31:50] <huh>	 Set $wgShowExceptionDetails = true; at the bottom of LocalSettings.php to show detailed debugging information.
[11:31:52] <huh>	 !logs
[11:32:01] <huh>	 !log
[11:32:09] <huh>	 meh
[11:34:55] <huh>	 apergos, mark, Gloria, Krenair, Nemo_bis, Reedy  is everything okay?
[11:35:03] <huh>	 What's going on?
[11:36:15] <MaxSem>	 huh, looking
[11:37:06] <huh>	 Thanks.
[11:39:37] <apergos>	 I can see older revisions of it
[11:40:15] <huh>	 Can you figure out what changed in the last diff?
[11:40:17] <apergos>	 https://meta.wikimedia.org/w/index.php?title=User:Aldnonymous&oldid=7499989  all the ones after this in the history give that error
[11:40:24] <huh>	 hm
[11:40:25] <apergos>	 (as a logged in user)
[11:40:36] <apergos>	 no, the diff gives the error too
[11:40:51] <huh>	 anyone with direct DB access who can tell us?
[11:41:02] <huh>	 maybe API can ...
[11:41:20] <huh>	 [but someone else needs to query]
[11:41:49] <apergos>	 https://meta.wikimedia.org/w/index.php?title=User:Aldnonymous&oldid=7559365  this is the earliest revision I see that is an issue
[11:42:19] <huh>	 apergos: I can see the diffs via API
[11:42:26] <huh>	 maybe something wacky was added
[11:42:31] <apergos>	 I guess so
[11:42:31] <Nemo_bis>	 {{Special:LanguageStats/id|suppressempty=0|suppresscomplete=1}} ?
[11:42:34] <Aldnonymous>	 debugging it
[11:42:52] <huh>	 Nemo_bis: that's it!
[11:42:56] <Aldnonymous>	 AH 
[11:42:59] <apergos>	 good, I guess you know already that something you added on the 19th likely is the culprit
[11:43:00] <Aldnonymous>	 that language stats
[11:43:02] <Aldnonymous>	 remove them
[11:43:09] <huh>	 Aldnonymous: it's not your fault
[11:43:22] <huh>	 this is MediaWiki's or (Translate's) fault
[11:43:22] <Aldnonymous>	 I use them so I can fast translae
[11:43:31] <Aldnonymous>	 translate*
[11:43:41] <Nemo_bis>	 seemes fixed
[11:43:50] <Nemo_bis>	 please read docs
[11:44:02] <Aldnonymous>	 will do
[11:44:12] <Nemo_bis>	 and huh, this was nowhere near being urgent
[11:44:21] <huh>	 Nemo_bis: sorry
[11:44:34] <Nemo_bis>	 thanks for the note but making it look like a devastating emergency was unnecessarily ddistressing :)
[11:44:38] <Aldnonymous>	 No, I use them so I can see changes
[11:44:46] <MaxSem>	 crap, exception inside of exception handler results in no log entry being generated
[11:44:59] <huh>	 Nemo_bis: If I don't know what's wrong and no error is shown, how can I tell if it's urgent or not?
[11:45:03] <MaxSem>	 epic MW is epic
[11:45:07] <Aldnonymous>	 the docs was... empty(?)
[11:45:17] <huh>	 Besides, you can still see the stats here https://meta.wikimedia.org/w/index.php?title=User:Aldnonymous&oldid=7685114
[11:45:21] <huh>	 strange
[11:45:38] <Aldnonymous>	 LOL
[11:45:50] <Aldnonymous>	 so it was fine in oldid
[11:45:59] <apergos>	 huh: you can try: other user pages, try logged out and loggged in, try other revisions of the same user page, etc
[11:46:11] <apergos>	 the first thign is to figure out how widespread it is
[11:46:22] <Aldnonymous>	 I see one on verdy_p
[11:46:30] <Aldnonymous>	 seems fine
[11:46:44] <MaxSem>	 Nemo_bis, exception in the wild IS urgent unless proven otherwise
[11:46:47] <Aldnonymous>	 did someone adding something weird in it?
[11:46:53] <huh>	 In the future, I will do so
[11:46:56] <Nemo_bis>	 please file a bug, transclusion should obviously degrade more gracefully
[11:47:08] <huh>	 Aldnonymous: want to do so?
[11:47:12] <Aldnonymous>	 no
[11:47:19] <huh>	 then I guess I have tgo
[11:47:22] <huh>	 to
[11:47:32] <Aldnonymous>	 not that bugzilla will going to answer it anyway
[11:47:47] <huh>	 (?)
[11:47:52] <Aldnonymous>	 I will see future changes from verdy_p instead
[11:47:55] <Nemo_bis>	 bugzilla doesn't talk IIRC
[11:48:07] <Aldnonymous>	 brb, phone.
[11:49:46] <huh>	 Nemo_bis: https://bugzilla.wikimedia.org/show_bug.cgi?id=62125
[11:50:58] <huh>	 Is it a bug in core or just with this special page
[11:51:00] <huh>	 ?
[11:51:54] <MaxSem>	 with Translate
[11:53:36] <Aldnonymous>	 back
[11:53:45] <Aldnonymous>	 I should remove those translate thingy then
[11:57:11] <Aldnonymous>	 Severity: enhancement → major
[11:58:23] <huh>	 Aldnonymous: are you aldnonymous4?
[11:58:26] <Aldnonymous>	 yes
[11:59:01] <huh>	 k
[12:03:18] <Aldnonymous>	 wait huh did you add me to cc list?
[12:05:21] <huh>	 Aldnonymous: yes?
[12:05:25] <huh>	 why?
[12:05:44] <Aldnonymous>	 thanks not that I'm gonna monitor that one periodically
[15:04:22] <strainu>	 hey guys
[15:04:37] <strainu>	 is anyone else having problems with the JS this weekend?
[15:04:50] <strainu>	 I see that status shows all clear
[15:05:06] <strainu>	 but I get all kind of errors on ro.wp
[15:05:26] <strainu>	 I've disabled my gadgets and user scripts and still no luck
[15:09:11] <Gloria>	 strainu: What kind of errors?
[15:09:45] <strainu>	 well, I have the advanced recent changes enabled, and the page looks like no js is enabled
[15:09:50] <strainu>	 Firebug says:
[15:10:01] <strainu>	 ReferenceError: mw is not defined https://bits.wikimedia.org/ro.wikipedia.org/load.php?debug=false&lang=ro&modules=user&only=scripts&skin=vector&user=Strainu&version=20140302T150833Z&* Line 1
[15:11:10] <strainu>	 The line looks like this: ;mw.loader.state({"user":"ready"});
[15:11:33] <strainu>	 the semicollon at the beginning looks strange, but it shouldn't cause these problems
[15:11:38] <strainu>	 afaik
[15:13:47] <strainu>	 hmm, commons seems to work, so most likely it's something specific to ro.wp
[15:15:37] <Gloria>	 strainu: If you log out, do you get the same error?
[15:15:46] <Gloria>	 It'd be helpful to figure out if it's user-specific or site-specific.
[15:16:05] <strainu>	 how can I get advanced recent changes while logged out?
[15:16:21] <Gloria>	 Hrm.
[15:16:25] <huh>	 Maybe with mw.loader.using ?
[15:16:25] <strainu>	 I need to find some other stuff that goes wrong :D
[15:16:25] <Gloria>	 Is that the only thing that's failing?
[15:16:40] <strainu>	 I can't really tell right now
[15:16:59] <Gloria>	 https://bits.wikimedia.org/ro.wikipedia.org/load.php?debug=false&lang=ro&modules=user&only=scripts&skin=vector&user=MZMcBride&version=20140302T150833Z&*
[15:17:03] <Gloria>	 No semicolon for me.
[15:17:10] <strainu>	 this is probably a good thing, this means that Wikipedia works without JS :D
[15:17:28] <strainu>	 hmm, ok, I'll delete the scripts alltoghether
[15:17:44] <huh>	 strainu: it also works if you use lynx, but it's not as pretty
[15:17:52] <Gloria>	 strainu: You've got a stray ; in your JS.
[15:17:54] <Gloria>	 https://bits.wikimedia.org/ro.wikipedia.org/load.php?debug=true&lang=ro&modules=user&only=scripts&skin=vector&user=Strainu&version=20140302T150833Z&*
[15:17:58] <Gloria>	 Look at the bottom.
[15:18:07] <Gloria>	 ;
[15:18:07] <Gloria>	 mw. ...
[15:18:24] <Gloria>	 Those scripts also have indicators that suggest that they're ancient.
[15:18:44] <Gloria>	 JavaScript errors are usually cascading, in a sense.
[15:18:56] <Gloria>	 If there's an error early on, subsequent user scripts will usually never load.
[15:19:08] <Gloria>	 That is, an early small error can create a big mess.
[15:19:18] <Gloria>	 Even if the actual issue is fairly small and self-contained.
[15:19:40] <huh>	 Why would the semicolon break it?
[15:19:51] <huh>	 ';;;;;;;;;' is valid JS (but no-op)
[15:19:57] <strainu>	 huh: I don't know, but it did
[15:20:05] <strainu>	 I deleted User:Strainu/common.js
[15:20:06] <Gloria>	 I don't know if it's the semicolon.
[15:20:10] <strainu>	 and it now works
[15:20:10] <fale>	 I'm trying to optimize a query, but withu explain I'm having problems. Any suggestions? (http://git.wikimedia.org/blob/labs%2Ftools%2Flists.git/7102890053067fd5a05e8ce35122ce1daf8d3ac2/query%2Fitwiki%2FEGO%2FConvenzioni%2FFalse_sottopagine.sql)
[15:20:15] <huh>	 strainu: ok
[15:20:17] <Gloria>	 It could the line below. That whole file is pretty rough.
[15:20:26] <huh>	 Gloria: it's almost all commented out anywat
[15:20:28] <huh>	 y
[15:20:45] <Gloria>	 huh: Sure, but it's indicative. :-)
[15:21:02] <huh>	 document.write('<script type="text/javascript" seriously?
[15:21:05] <strainu>	 yeah, I'll just go back to rebuilding it with only the stuff I really need
[15:21:19] <strainu>	 thanks 
[15:21:23] <Gloria>	 huh: Be nice. :-)
[15:21:34] <huh>	 Sorry.
[15:21:40] <Gloria>	 fale: EXPLAIN doesn't work with MariaDB/MySQL views.
[15:21:53] <Gloria>	 fale: And generally subqueries are gross and not very efficient.
[15:22:17] <Gloria>	 fale: What are you trying to do?
[15:22:34] <Gloria>	 Subpages of the article namespace...
[15:23:02] <fale>	 Gloria: get all pages that contain a "/" in the title, but are not subpages and are not listed in that page
[15:23:15] <Gloria>	 What's "not subpages" mean?
[15:23:27] <Gloria>	 A subpage is a somewhat abstract and confusing concept.
[15:23:35] <Gloria>	 It's namespace-dependent, sort of.
[15:23:39] <Gloria>	 But consider: Talk:9/11
[15:23:46] <Gloria>	 Or Talk:AC/DC
[15:23:51] <fale>	 Gloria: exactly
[15:24:02] <fale>	 Gloria: it should match the second but not the first
[15:24:28] <fale>	 (if the first is the page 11 that is a subpage of 9)
[15:24:30] <Gloria>	 fale: It looks like you don't want all pages containing a "/".
[15:24:41] <Gloria>	 Only articles (pages in NS:0).
[15:24:43] <Gloria>	 Is that right?
[15:24:46] <fale>	 yep
[15:25:41] <Gloria>	 So all main namespace non-redirect pages that contain a "/" where the subject-space page is a... non-redirect?
[15:25:46] <Gloria>	 I'm still not clear what the goal is here.
[15:26:10] <Gloria>	 fale: Do you know any programming languages? Doing this purely in SQL gets kind of messy.
[15:26:27] <fale>	 Gloria: the goal is identify pages that could have been mistakenly nominated in a wrong way
[15:27:03] <fale>	 Gloria: I'm trying to automatizing some queries and at this moment the sw I wrote is able to run only sql :(
[15:27:06] <Gloria>	 fale: The JOIN seems unnecessary.
[15:27:21] <Gloria>	 I'm not sure you're doing much with p2.
[15:27:36] <Gloria>	 And you're not specifying p2.page_namespace.
[15:28:27] <Gloria>	 You should kill or redo that JOIN.
[15:28:37] <Gloria>	 And rewrite the subquery to be faster.
[15:28:42] <fale>	 the goal of the join was to exclude the real subpages
[15:29:14] <huh>	 Gloria: I commented on https://gerrit.wikimedia.org/r/#/c/7300/
[15:29:34] <Gloria>	 fale: Real subpage?
[15:29:40] <huh>	 Are they still working on it?
[15:29:57] <Gloria>	 huh: Seems not.
[15:30:10] <Gloria>	 huh: Rebase the change?
[15:30:24] <huh>	 will do
[15:30:38] <fale>	 Gloria: thanks for the heads-up :) I'll relook the code based on these suggestions :)
[15:30:48] * huh  should probably set up SSH keys instead of using gerrit patch uploader thing someday

[15:32:09] <Gloria>	 huh: It takes like thirty seconds to generate and set up an SSH key.
[15:32:39] <huh>	 Does it work with ssh -A ?
[15:32:47] <Gloria>	 Yes.
[15:32:52] <Gloria>	 fale: Any time you specify page_title, be sure to specify page_namespace.
[15:32:55] <Gloria>	 They're a pair.
[15:32:59] <Gloria>	 Same with tl_title, tl_namespace.
[15:33:05] <Gloria>	 And some others.
[15:33:38] <Gloria>	 alias tools='ssh -A -q -i ~/.ssh/id_rsa_labs mzmcbride@tools-login.wmflabs.org'
[15:35:21] <Gloria>	 huh: Getting git review set up and working is probably more annoying. Which OS/
[15:35:25] <Gloria>	 ?
[15:35:40] <huh>	 Can I do it from Labs?
[15:35:48] <Gloria>	 How are you logged into Labs?
[15:35:52] <huh>	 ssh ..
[15:35:59] <Gloria>	 Then you already set up your key?
[15:36:02] <Gloria>	 Just use that with Gerrit?
[15:36:04] <Gloria>	 It's the same system.
[15:36:15] <Gloria>	 True story.
[15:36:21] <huh>	 The key on wikitech is the same as on gerrit?
[15:36:28] <Gloria>	 I believe so.
[15:37:26] <Gloria>	 huh: Yes.
[15:37:41] <Gloria>	 (You made me second-guess myself, I had to go double-check.)
[15:37:47] <Gloria>	 Same key.
[15:38:14] <Gloria>	 huh: Try "ssh gerrit.wikimedia.org"?
[15:38:15] <Gloria>	 It should work.
[15:39:22] <huh>	 I don't need to use username@gerrit.wikimedia.org?
[15:39:48] <Gloria>	 Depends what your local username is.
[15:39:57] <Gloria>	 If it's the same as your Gerrit username, no.
[15:40:26] <Gloria>	 huh: Anyway, try it and see. :-)
[15:40:36] <Gloria>	 If it fails, you can always re-try.
[15:41:09] <huh>	 Host key verification failed.
[15:41:31] <Gloria>	 Which command did you run and what's the full output?
[15:41:34] <Gloria>	 You can use a pastebin.
[15:43:50] <Gloria>	 <huh> ';;;;;;;;;' is valid JS (but no-op) <-- Be careful about assumptions like this.
[15:43:52] <fale>	 in sql, if I write select a from table1 left join table2 on table1.id = table2.id where table1.name = lol, mysql does before the where condition or the join?
[15:43:54] <Gloria>	 Some browsers are idiotic.
[15:44:21] <huh>	 then what would ;;;;; do ?
[15:44:35] <Gloria>	 I dunno. I haven't tested it across browsers.
[15:44:40] <Gloria>	 But I wouldn't assume it's no-op everywhere.
[15:44:43] <Gloria>	 IE is stupid.
[15:44:48] <Gloria>	 Opera can be stupid.
[15:44:59] <Gloria>	 Firefox and Chrome can be stupid.
[15:45:24] <Gloria>	 fale: I don't understand your question.
[15:45:48] <Gloria>	 Usually you do a LEFT JOIN and then have a NULL clause somewhere.
[15:46:43] <tonythomas>	 How can I add a custom puppet module, exim4 to my existing mediawiki vagrant ? 
[15:46:52] <tonythomas>	 I have got the exim4 module here 
[15:46:53] <tonythomas>	 https://github.com/example42/puppet-exim
[15:46:55] <fale>	 Gloria: what do you mean for "then have a NULL clause somewhere"?
[15:47:22] <Gloria>	 fale: https://wiki.toolserver.org/view/MySQL_queries
[15:47:24] <Gloria>	 Look through there.
[15:47:43] <fale>	 Gloria: thanks a lot :)
[15:48:05] <Gloria>	 Those queries are pretty ugly and awful.
[15:48:09] <Gloria>	 But most of them should work.
[15:48:33] <Gloria>	 tonythomas: Add the Puppet module to the appropriate directory in your file system?
[15:48:37] <Gloria>	 And then include or require it?
[15:48:39] <Gloria>	 I'm not sure.
[15:48:46] <Gloria>	 That's usually how you install something.
[15:49:20] <tonythomas>	 Gloria:  puppet says you need to put the custom .pp files to manifests/manifests.d/ 
[15:49:43] <tonythomas>	 but, here in the repo, do you think only https://github.com/example42/puppet-exim/tree/master/manifests
[15:49:51] <tonythomas>	 needs to be put in manifests.d/ ?
[15:49:56] <Gloria>	 No idea.
[15:50:01] <Gloria>	 Try it and see?
[15:51:35] <tonythomas>	 I added :) didn't see any change :( btw, when I add such a custom module, it should come up in list-roles ?
[15:52:00] <Gloria>	 Vagrant roles and Puppet manifests/modules are distinct, I believe.
[15:52:10] <Gloria>	 But they may overlap or synchronize.
[15:52:16] <Gloria>	 I've never used either.
[15:52:52] <tonythomas>	 Gloria:  can you recommend someone to go for ?
[15:52:59] <Gloria>	 Ori. :-)
[15:53:02] <Gloria>	 But he's on vacation.
[15:53:17] <Gloria>	 mutante may be able to help.
[15:53:19] <Gloria>	 When he's around.
[15:53:33] <tonythomas>	 Gloria:  :(  ok. I will try to talk with one of them 
[16:18:43] <matanya>	 tonythomas: what are you trying to do?
[16:19:14] <tonythomas>	 matanya: I am trying to enable the exim role for my mediawiki vm
[16:19:46] <tonythomas>	 inorder to work on the bug https://bugzilla.wikimedia.org/show_bug.cgi?id=46640
[16:20:10] <matanya>	 oh, i saw your RT ticket
[16:20:13] <matanya>	 very nice
[16:20:35] <matanya>	 what exim role you are talking about?
[16:20:43] <tonythomas>	 Thanks ! I need to have  2 instance of the VM, one acting as the org's main mail server 
[16:20:56] <tonythomas>	 and  another to be the mw webserver (with db, and the usual outbound local mailserver)
[16:21:32] <matanya>	 any mail server should work, isn't it?
[16:21:35] <tonythomas>	 so that I can originate these bounce on that mail server 
[16:21:53] <tonythomas>	 matanya:  but, MW uses exim4 ?
[16:22:01] <matanya>	 yes
[16:22:19] <matanya>	 you can see part of the setup in puppet
[16:22:39] <tonythomas>	 in puppet ?
[16:22:44] <matanya>	 not all is puppertized, and what is puppetized isn't in a module
[16:22:53] <tonythomas>	 oh :( That's sad 
[16:23:10] <matanya>	 it is in manifests/mail.pp
[16:23:45] <matanya>	 https://git.wikimedia.org/blob/operations%2Fpuppet.git/65a15fb418e856adaeca11a9ab80d269a42338e8/manifests%2Fmail.pp
[16:24:02] <tonythomas>	 can I add it to puppet/manifests/roles/ in my local vagrant to get the thing done ?\
[16:24:14] <matanya>	 no, not really
[16:24:26] <tonythomas>	 matanya:  since its not a module ?
[16:24:57] <matanya>	 no, just becuse it is not complete
[16:25:14] <tonythomas>	 matanya:  Ah ! They would've hid some of the private parts ? 
[16:25:21] <matanya>	 the fact it is not a module just addes a layer of complaxity 
[16:25:28] <Gloria>	 complexity
[16:25:32] <matanya>	 yes
[16:25:55] <matanya>	 but the private parts are not your main concern
[16:26:06] <tonythomas>	 complexity! 
[16:26:21] <tonythomas>	 matanya:  of course 
[16:26:35] <matanya>	 if i were you, i would just though a mail server
[16:26:42] <matanya>	 *through
[16:26:47] <matanya>	 and start from there
[16:26:52] <tonythomas>	 matanya:  like, just install exim4 in the vagrant ?
[16:26:57] <matanya>	 yes
[16:27:14] <matanya>	 that should be enough for most of the things you need at first
[16:27:27] <tonythomas>	 yeah! I had enough trouble waiting for the exim4 puppet 
[16:27:35] <matanya>	 later on, you might need to tweak it a bit 
[16:27:57] <tonythomas>	 ok. And I should be running exim4 in both of my instances right ?
[16:28:09] <matanya>	 yes
[16:28:45] <matanya>	 in fact you can configure mw to forward to the mail server
[16:28:46] <tonythomas>	 so, what will be the main difference between my two instances ?
[16:29:22] <matanya>	 one will do mw heavy lifting and one will handle mail
[16:29:54] <tonythomas>	 matanya: so, one the mail handling instance, a mediawiki instance is not necessary ?
[16:32:21] <matanya>	 not sure i understand your q
[16:33:24] <tonythomas>	 in the second vagrant instance, which will act as the organizations main mail server, I should run MW+Exim right ?
[16:33:55] <matanya>	 you can use only mail
[16:34:01] <matanya>	 mw can forward mails
[16:34:46] <tonythomas>	 Can you please explain that part ? 
[16:34:58] <matanya>	 yes
[16:35:25] <matanya>	 mw has some php code that can send mails using pre-configured mail server, even on remote machines
[16:35:39] <matanya>	 you don't have to have mw and the mail server on the same host
[16:36:05] <tonythomas>	 oh! 
[16:36:09] <matanya>	 so since you are going to use a mail server anyway, you don't have to have two
[16:36:41] <tonythomas>	 ok. so, basically my first instance will have a MW and the second one a Mail Server ? 
[16:36:53] <matanya>	 yes
[16:37:22] <tonythomas>	 but matanya: So, in that case, I am implementing VERP in the exim4 sever ?
[16:37:31] <matanya>	 yes
[16:38:27] <matanya>	 well, you would also need an API call for that in mw
[16:38:44] <tonythomas>	 oh.! Thats where I get confused 
[16:39:08] <tonythomas>	 That API calls should be on the org part right ?
[16:39:32] <matanya>	 no, API's are in mw
[16:39:47] <matanya>	 the flow is as follow
[16:39:55] <tonythomas>	 k
[16:40:29] <matanya>	 mw needs to send a mail --> looing for a user email in db --> looks for mail server --> asks the mail server to send
[16:40:36] <matanya>	 is this part clear?
[16:41:26] <tonythomas>	 matanya:  that takes place in instance 1 ?
[16:41:28] <tonythomas>	 right ?
[16:41:34] <matanya>	 define 1 ?
[16:41:41] <matanya>	 the mw server?
[16:41:49] <tonythomas>	 instance 1:- loaded with MW 
[16:41:56] <tonythomas>	 I mean the client 
[16:42:43] <tonythomas>	 the instance which is sending the message 
[16:42:57] <matanya>	 yes all except the last step
[16:43:09] <matanya>	 in fact, all of them
[16:43:11] <matanya>	 sorry
[16:43:24] <matanya>	 on the mail server side:
[16:43:34] <tonythomas>	 yeah. and one more doubt. are we using the default mw forwarding here ?
[16:43:41] <matanya>	 yes
[16:43:41] <tonythomas>	 or exim ?
[16:43:50] <matanya>	 mw
[16:43:55] <matanya>	 no exim yet
[16:43:57] <tonythomas>	 so, the mail server is basically the mw php code. Ok 
[16:44:42] <matanya>	 gets the app request to send mail, with user address --> send the message --> success/bounce
[16:45:29] <matanya>	 our problem resides on the last step of the mail server and second step of mw
[16:46:08] <tonythomas>	 looing for a user email in db ? 
[16:46:40] <matanya>	 actully finding it is old and bounce
[16:46:46] <matanya>	 not the lookup itself
[16:47:08] <tonythomas>	 of course! Thats the challenge 
[16:47:14] <matanya>	 we need a way to tell mw: "hey, this mail is old, it bouces the whole time"
[16:47:36] <matanya>	 so we need to do: 
[16:47:48] <tonythomas>	 matanya: An email bounce only when a user account cannot be reached or is there any other cases ?
[16:48:02] <matanya>	 not user account
[16:48:08] <matanya>	 email address
[16:48:09] <tonythomas>	 I mean the email 
[16:48:19] <tonythomas>	 of course. like in google, we have the same ! 
[16:48:22] <tonythomas>	 in gmail 
[16:48:30] <matanya>	 that is the most common one
[16:48:48] <matanya>	 you cna bounce becuase of firewall rules, drops and what not
[16:48:51] <matanya>	 *can
[16:49:15] <matanya>	 the underlaying reason is some fail to deliver
[16:49:34] <tonythomas>	 matanya:  of course. and currently, whats the status of bounce ? do we atleast have a bounce mail returning to the sender side ?
[16:49:50] <matanya>	 not sure about this
[16:50:09] <matanya>	 i think it is dropped, but you will need to ask one of the ops
[16:50:20] <tonythomas>	 I have mailed Jeff and Lego in detail on this, hope I get to know it fast 
[16:51:19] <matanya>	 jeff green?
[16:51:39] <tonythomas>	 yeah ! He is helping me with the ops part 
[16:51:45] <matanya>	 nice
[16:52:11] <tonythomas>	 matanya:  if you are free, can we go a bit detail with the instance 2 part ? 
[16:52:29] <matanya>	 yes, i can try. let just commit something a sec
[16:52:43] <tonythomas>	 ok
[16:52:54] <matanya>	 what do you need to know about box 2?
[16:53:57] <tonythomas>	 so, the box 2 handles (gets the app request to send mail, with user address --> send the message --> success/bounce) .  It have only exim4 installed ?
[16:54:16] <matanya>	 that should be enough
[16:54:29] <matanya>	 mail servers aren't that complicated :)
[16:55:13] <tonythomas>	 Yeah. so, inorder to create a bounce in my localhost, how will I send the mail 
[16:55:15] <tonythomas>	 ?
[16:55:27] <tonythomas>	 like it should not get delivered 
[16:55:45] <matanya>	 to a non-existing address
[16:56:13] <tonythomas>	 matanya:  that would require a db check from the box 1 itself right ?
[16:56:21] <matanya>	 yes
[16:56:38] <matanya>	 the mw mysql/mariadb from your installation
[16:57:47] <tonythomas>	 so, if the check is done in box1, and mw finds it to be a non-existing email addr , why should it ever go to inst2 then ?
[16:59:04] <matanya>	 it doesn't know it doesn't exist
[16:59:16] <matanya>	 it takes the data to the mail server and askes to mail
[16:59:29] <matanya>	 the mail server find out it doesn't exist
[16:59:46] <matanya>	 and the missing part is to tell the mw box this is the issue
[17:00:13] <tonythomas>	 matanya:  and the mail server has to come back all the way to box1 to check in its db, whether the email addr exists 
[17:00:32] <matanya>	 no, the mail server don't get to the db
[17:00:39] <tonythomas>	 then 
[17:00:42] <matanya>	 it gets the address from wm
[17:01:30] <tonythomas>	 ok. !! 
[17:02:08] <tonythomas>	 I will try to configure the box 1 and box 2 as per we discussed 
[17:02:15] <tonythomas>	 send a mail to some unknown addr 
[17:02:29] <tonythomas>	 and see what all happends 
[17:02:36] <tonythomas>	 *happens 
[17:03:00] <matanya>	 your part will be to collect the stuff returned from the mail server and tell mw, hey, it bounced
[17:03:25] <matanya>	 for telling mw that you will need to write some api call
[17:04:34] <matanya>	 good luck, and do ping if you need help. i'm no expert, but i have been around for some time
[17:04:35] <tonythomas>	 ok. API that would listen to the mai server returns and delete the user, if necessary 
[17:04:46] <matanya>	 no, not delete the user
[17:05:09] <matanya>	 mark his email as invalidate after some thershold 
[17:05:38] <tonythomas>	 yeah! that should go into the DB right ?
[17:05:46] <tonythomas>	 the number of bounces and all ?
[17:06:15] <matanya>	 sounds like a resonable place to go to
[17:07:37] <matanya>	 the mail server it self of course should support making api calls from it at the end
[17:07:57] <tonythomas>	 exim will support that ?
[17:09:09] <matanya>	 it should
[17:09:48] <matanya>	 exim supports VERP, but you will need to find a way to do that without the need of a full mw on the mail servers
[17:10:22] <matanya>	 ops won't like such approach, and it is clear why
[17:10:37] <tonythomas>	 when you say a full mw, you mean only the exim4 will be running right ?
[17:10:41] <tonythomas>	 and no instance of mw 
[17:11:43] <matanya>	 that is the setup in prod
[17:12:50] <tonythomas>	 ok!! I will try to get that implemented. Thanks for your help today ! Time for me to scribble all that and pack! 
[17:13:19] <matanya>	 np :)
[18:12:12] <odder>	 Oh no, I don't know how to translate text I wrote in English to my native Polish.
[18:12:20] <odder>	 That's... embarrasing.
[18:19:42] <BlastHardcheese>	 you put it in google translate, right :p
[18:40:55] <Newyorkadam>	 Betacommand: hi
[18:41:14] <odder>	 BlastHardcheese: No, I got another native speaker help me :-)
[19:01:21] <Betacommand>	 Newyorkadam: hay
[19:01:26] <Newyorkadam>	 Betacommand: sup
[19:01:35] <Betacommand>	 Not much
[19:04:36] <Betacommand>	 Newyorkadam: did you need something?
[19:04:49] <Newyorkadam>	 Betacommand: I did but it isn't important :p I'll ask you later if I want to
[19:04:52] <Newyorkadam>	 what's up with you?
[19:05:07] <Betacommand>	 Newyorkadam: go ahead, I wont be around later
[19:05:20] <Newyorkadam>	 Betacommand: Oh I mean later as in a week or two
[19:05:36] <Betacommand>	 Newyorkadam: what do you need?
[19:06:12] <Newyorkadam>	 Betacommand: category, sorted by active members
[19:06:19] <Newyorkadam>	 edited within two months
[19:06:37] <Newyorkadam>	 https://en.wikipedia.org/wiki/Category:WikiProject_Lost_participants
[19:26:26] <Betacommand>	 Newyorkadam: running
[19:26:32] <Newyorkadam>	 Betacommand: thanks :)
[19:28:45] <Betacommand>	 Newyorkadam: short list
[19:28:53] <Newyorkadam>	 Betacommand: really? :(
[19:29:03] <Newyorkadam>	 hopefully not less than 15 people
[19:29:23] <Betacommand>	 Newyorkadam: 13 http://pastebin.com/9An45t6t
[19:29:27] <Newyorkadam>	 thanks
[19:29:54] <Betacommand>	 Newyorkadam: sorry to dash your hopes
[19:30:02] <Newyorkadam>	 :p all good thanks
[19:30:53] <Betacommand>	 that was set at 60 days
[19:31:14] <Betacommand>	 I was too lazy to do date math
[19:31:44] <Jeske_Couriano>	 Morning, gentlemen.
[19:32:54] <Jeske_Couriano>	 Any idea why [[Wikipedia:Administrators' noticeboard/Incidents]] occasionally throws up an SSL error when one tries to connect to it?
[19:38:51] <Krenair>	 Jeske_Couriano, what is the error?
[19:40:10] <Jeske_Couriano>	 I can't reproduce it now, but basically the first time I've connected to AN/I on en.wp in a given Firefox context, it fails to load, with the warning seeming to imply it has something to do with the SSL.
[19:45:03] <Jeske_Couriano>	 I also can't seem to reproduce it with IE.
[19:53:50] <MrZ-man>	 anomie: Both https://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Log/2014_February_20 and https://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Log/2014_February_22 are exhibiting bug 61953
[19:54:14] <anomie>	 MrZ-man: ok, I'll take a look
[20:53:32] <anomie>	 MrZ-man: Hmm. Both are showing all the transclusions now. I did see Feb 22 showing the issue at first though.
[21:27:10] <MrZ-man>	 anomie: the only thing I can think of is that those pages are edited almost exclusively by scripts, so it could be an API issue
[21:28:12] <anomie>	 MrZ-man: Seems unlikely to me. But at this point we're both just guessing :/
[21:29:19] <NotASpy>	 someone admitted to using IE on the internet. Huh. 
[21:56:35] <Newyorkadam>	 I got access to tool server :) Gloria 
[22:56:39] <Schroeder>	 THAET WAES GOD CYNING
[23:15:06] <Newyorkadam>	 Betacommand: hi
[23:27:35] <Gloria>	 Newyorkadam: Nice.