[19:03:12] Nikerabbit: ping [19:03:47] Nikerabbit: what is wrong with http://www.mediawiki.org/wiki/Special:Code/MediaWiki/106392 [19:04:47] Nikerabbit: also what is wrong with [19:04:47] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/106391 [19:10:06] preilly: the conditions don't make sense... one part of them is redundant [19:11:01] Nikerabbit: how so? [19:11:43] in one of them if the self::$hideLogo is true also set the self::$hideFooter to true as well [19:12:09] in the other I want to see that the value of hidesearchbox is 1 [19:12:19] why don't you just set them both at the same time then [19:12:44] what are you trying to achieve with $a && $a == 1? [19:12:52] why not just $a == 1? [19:13:04] or even just if ( $a ) [19:13:13] Nikerabbit: I don't want just $a [19:13:23] Nikerabbit: I want $a to be 1 and only 1 [19:13:28] Nikerabbit: for it's value [19:13:30] $a == 1 implies $a [19:13:49] if ( $a == 1 ) triggers in the exact same situations as if ($a && $a == 1 ) [19:14:59] well, I suggested using getCheck() [19:15:26] Nikerabbit: I think your wording is too strong [19:15:40] Nikerabbit: please take a look at * /trunk/extensions/MobileFrontend/MobileFrontend.php: fix for r106392 and r106391 [19:16:21] <^demon|away> http://www.php.net/manual/en/types.comparisons.php is useful. [19:16:42] what do you mean with too strong? [19:17:16] Nikerabbit: "Makes no sense." [19:17:28] Nikerabbit: that is what you said in http://www.mediawiki.org/wiki/Special:Code/MediaWiki/106391 [19:17:35] ^demon|away: I don't need that page [19:18:35] ^demon|away: also, I choose that style because I'm used to coding to avoid E_NOTICE with empty() or isset() [19:18:44] preilly: it's weird code, what should I say instead? [19:18:55] Nikerabbit is right, it doesn't make sense [19:19:04] RoanKattouw: why? [19:19:06] Nikerabbit: why? [19:19:16] r106464 fixes it [19:19:20] <^demon|away> Well I was just reinforcing Roan's point that $a && $a == 1 is pointless, especially since you're using getText(). [19:19:29] Well, sort of [19:19:49] RoanKattouw: how sort of? [19:19:49] I guess it's reasonable to require that the parameter be set to 1 as opposed to, say, 0 [19:19:59] Nah, never mind, I was momentarily confused [19:20:02] It makes sense now [19:20:19] <^demon|away> If you used getInt() you could skip the first check and just do if( $a == 1 ) [19:20:40] It's fine as it is now IMO [19:20:48] He removed the "$a &&" part [19:20:51] *^demon|away has places to be [19:20:59] So now it's $a = $wgRequest->getText( 'foo' ); if ( $a == 1 ) { ... } [19:21:17] yeah, still a bit weird to do "1" == 1 [19:21:37] Yeah, but it's acceptable IMO [19:21:44] potential pitfall if someone changes == to === [19:21:54] SMW has loooots of bugs after JeroenDeDauw did that [19:22:08] Nikerabbit: so, should it be getVal [19:22:25] <^demon|away> I suggested getInt() [19:22:52] RoanKattouw: thanks for the detailed response to Legal, IMO we don't need to censor any AFT5 table/dump, but I am waiting to hear from them [19:24:55] DarTar: I never read the feedback policy. Does it say that we can publish who rated what when (esp. the who part)? [19:25:12] Nikerabbit, RoanKattouw: how about * /trunk/extensions/MobileFrontend/MobileFrontend.php: fix for r106391 [19:27:07] RoanKattouw: (1) addresses of anonymous raters: it does; (2) anon tokens: it doesn't, but that's because we didn't have an actual policy at the time the feedback policy was written (so it just reflects what we did for AFT4 as a conservative option, regardless of what was legally possible) [19:27:19] s/addresses/IP addresses [19:27:30] Anon tokens are randomly generated strings with no privacy implications [19:27:35] exactly [19:27:58] the only new concern is that they store persisting identity across IP addresses [19:28:00] User names of logged-in raters, too? [19:28:04] Ah, yes [19:28:05] yep [19:28:08] Well, but [19:28:16] What kind of concern is that in practice? [19:28:31] no concern for me, but Legal may have a different opinion ;) [19:28:34] Right [19:28:36] Well [19:28:45] We can always assume caution and hide the anon token field [19:29:14] yeah but I'd rather avoid anonymizing fields that we are not supposed to anonymize as this creates confusion [19:29:23] Nikerabbit: can you please review that change and mark resolved if you feel that it is [19:29:58] so if Legal is ok with that and there are no community concerns we should just replicate the data as is with no censoring [19:30:32] DarTar: Could you reply to the thread with me and Geoffrey and note your links-IPs-together observation? [19:30:46] sure [19:31:03] preilly: there [19:31:37] Nikerabbit: thanks, but what about http://www.mediawiki.org/wiki/Special:Code/MediaWiki/106392 [19:33:38] Nikerabbit: really? Any remaining bugs caused by that around? I don't know of any. [19:35:43] JeroenDeDauw: just today I fixed on [19:35:48] e [19:36:02] Nikerabbit: great [19:36:19] Nikerabbit: I think it's exaggerating to say it has lot's though [19:36:29] JeroenDeDauw: well, many [19:36:31] And at least now the code does what you'd expect it to do [19:37:42] Nikerabbit: can you also mark off http://www.mediawiki.org/wiki/Special:Code/MediaWiki/106392 [19:38:57] if the intetion is that hidelog also hides the footer, it would be more clear to do self::$hideLogo = self::$hideFooter = true; [19:39:12] the code is not wrong though [19:39:47] Nikerabbit: can you mark it as okay [19:40:36] preilly: does it need to be? :) I'm running close to my weekly work quota [19:41:11] Nikerabbit: are you joking? [19:41:47] no that is not a joke, I've done 43h this week [19:42:21] Nikerabbit: so, marking this is okay is a big time sink? [19:43:44] preilly: ODesk-based contractors have a weekly limit of 45 billable hours [19:44:00] So if we work more than 45h in a calendar week we don't get paid for the overtime [19:44:50] + it's 10pm here, and you are not responding to my suggestions [19:44:53] RoanKattouw: okay [19:45:31] Nikerabbit: OKAY [19:45:55] With that and the time difference, it's easy to accidentally say something that sounds like "please do this in your unpaid overtime on Friday night at 11, kthx" :) [19:46:13] I hope you understand, I have nothing against you, I just don't want to think about that code now [19:46:31] no worries [19:52:44] don't worry, salaried people don't get paid for the overtime either :-/ [19:53:42] we need a way to politely set boundaries (for contractors it's mandatory, for staff it's a bit harder) [19:54:41] apergos, what work laws does the WMF follow? [19:55:09] um [19:55:14] whatever the us ones are [19:56:33] isn't overtime paid in the USA? [19:56:59] depends on the type of work and the position [19:57:07] for certain salaried positions, generally not [19:57:24] and especially in the tech industry [20:01:43] *Nemo_bis reading https://upload.wikimedia.org/wikipedia/foundation/2/2a/Wikimedia_Foundation_Compensation_Practices.pdf [20:01:55] DarTar: aftv5 click tracking is updated -- not sure if you're getting the emails from bugzilla or not, but if not, i've got a complete list of tracking ids here: https://bugzilla.wikimedia.org/show_bug.cgi?id=32992#c11 [20:02:04] LOL @ ??In-office massage is provided monthly at a discounted rate.?? [20:02:20] rsterbin: sweet! [20:02:30] :) [20:02:40] I was coding and not looking at my mail, let me take a look [20:03:44] rsterbin: I like the underscores for cta options [20:03:52] brion: https://bugzilla.wikimedia.org/show_bug.cgi?id=31697 how is this bug? is ok now? [20:04:13] hexmode, no idea [20:04:31] last comment is from you, so you'd know better :) [20:05:17] rsterbin: did you guys test the edit hooks for edit-attempt/edit-complete? [20:05:40] looks like there are still bad entries in the indexes [20:05:58] brion: k, looks like I'll have to get input from the reporter. Was hoping you knew more about how this all was supposed to act :) [20:06:19] hexmode, shouldn't need any input from the reporter at all, no [20:06:22] I'd love to test if we collect data accurately after pushing to labs [20:06:24] just need to find out who can fix the indexes [20:08:10] brion: ok.... makes sense. rainman or ops then. [20:11:00] apergos, Nemo_bis: Hmm, no overtime isn't mentioned anywhere I guess. AIUI the convention is that you're paid N dollars/month and you're expected to work at least 40h/wk for it. You're not required to work more, and WMF must not require you to work more unless compensation days are offered [20:11:02] RoanKattouw: Legal wants to meet me to discuss AFT data anonymization on Monday, can we ask the TS folks to hold off the replication until we get the green light from Geoff? [20:11:25] (For instance, for the upcoming hackathon, many people will essentially be working on Saturday and Sunday, which means they can take 2 days off later to compensate for that) [20:11:41] DarTar: I haven't asked the TS folks to publish just yet, so there's nothing to do [20:11:51] I'll refrain from filing a ticket with them until we have approval from legal [20:12:13] ok so the tables won't be replicated unless we specifically ask them to, correct? [20:13:50] Yes [20:13:59] And I haven't filed anything with them yet, so we're good [20:14:14] DarTar: I only tested (locally) that the request was being sent -- next step is checking the db [20:14:54] my local copy seems to have a problem with click tracking, so i have to get it up to our dev server to make sure it's okay [20:15:52] rsterbin: np, ping me once you've pushed to labs so we can start testing [20:16:15] will do [20:24:48] RoanKattouw, compensation days to compensate the few vacation days and most importantly the days when you don't have a chance to use the discounted massage because you're abroad? [20:25:02] heh [20:25:11] I explained compensation days abov [20:25:20] Basically it's when WMF requires people to work on weekend days or holidays [20:25:32] Like, next month there'll be a hackathon in SF that runs from Fri to Sun [20:25:38] Many tech staff will be attending [20:25:46] Yes, I understood, just kidding :-p [20:25:51] And they'll be allowed to take 2 extra days off because they worked a Sat and a Sun [20:25:54] heh [20:26:09] This basically means that overtime is not paid. After all, it often is not even where it's mandatory. [20:26:38] My understanding is that overtime is not paid, but also not required and in fact I don't think managers can ask for it [20:26:51] At the very least it seems unethical to require people to work for free [20:29:28] pff, since when work is ethical [20:29:36] well people often do get asked to do more work than can get done in an eight hour day (industry-wide) [20:29:37] Only owen thought it [20:29:51] as in, it's an expectation [20:30:09] apergos: Sure, but I don't think you'll get crap from WMF management if you religiously work 9-5 (and shift, but not add, hours on request) [20:30:20] Hi dev slaves! ;-) [20:30:20] DEBUG Exception: Cannot be negative. [20:30:29] wm-bot: wtf? [20:30:31] petan: ---^^ [20:30:48] Slaves? [20:30:48] DEBUG Exception: Cannot be negative. [20:30:52] Hahaa [20:30:52] DEBUG Exception: Cannot be negative. [20:31:08] *multichill seems to be a trigger [20:31:08] DEBUG Exception: Cannot be negative. [20:31:57] I guess he's referring to wage per hour (in overtime) [20:32:15] *apergos looks at multichill [20:32:25] maybe, but I ain't *yer* slave baby :-P [20:32:39] btw you saw that mark and reedy got the rotate bot to zoom along? [20:32:53] turned out the slowdown was an interface setting again on some lvs hosts [20:32:56] 150% on nights and saturdays, 200% on sundays, that's what I get [20:32:56] DEBUG Exception: Cannot be negative. [20:33:24] and of course without the latency from toolserver to cluster it's muuuch faster [20:33:29] multichill: Yeah, European labor laws FTW [20:33:58] apergos, yesterday I was testing upload speed; it went at 900 KiB/s, so all ok; I aborted, no file uploaded; I look today and I find out it appeared 30 min after I aborted it [20:34:05] how is it possible? :-/ [20:34:32] :-D [20:34:33] no idea! [20:34:45] :) [20:37:28] RoanKattouw: Looks like there's going to be quite a Dutch delegation to the hackathon :P [20:37:34] Yeah :) [20:48:55] RoanKattouw: Any idea when you're flying out? [20:48:56] DEBUG Exception: Cannot be negative. [20:49:29] multichill: I'm hoping to be living in SF at the time [20:49:36] I say hoping because my visa isn't sorted yet [20:49:54] If all goes to plan, I will be flying in on Friday morning, from Australia [20:50:20] Cool! All finished with school? What brings you to Australia? [20:50:20] DEBUG Exception: Cannot be negative. [20:51:22] linux.conf.au [20:51:27] Yeah, I graduated in Sept [20:51:38] (Well, really in July, but I got my degree in Sept) [20:53:41] Nice. Is Timo coming too? [20:53:41] DEBUG Exception: Cannot be negative. [20:56:27] RoanKattouw? [20:56:27] DEBUG Exception: Cannot be negative. [20:57:07] I don't know [20:57:12] Oh, to LCA [20:57:15] No, he's not [20:57:20] LCA is just me and Trevor [20:57:35] Whether Timo is coming to SF I don't know, ask Timo :) [20:57:50] [ERROR] No match for ???kr???. [20:57:50] DEBUG Exception: Cannot be negative. [20:58:50] Platonides: Wouldn't it be cool to have something like https://en.wikipedia.org/wiki/Special:ApiSandbox for the monuments api? ;-) [20:58:50] DEBUG Exception: Cannot be negative. [20:58:59] yay [21:36:29] brion: just sent a parser status update to wikitext-l [21:42:29] \o/ [21:42:36] i'll have to catch up with you guys next week :D [21:47:32] brion: would also like to ask you for your input on template expansions, but am also looking forward to the weekend now ;) [21:48:03] yes -- enjoy your weekend and i'll comment later :D [21:48:08] will try to catch you Monday or Tuesday [21:48:13] excellent [21:48:47] off to bed for me.. [21:48:52] bye! [23:39:03] hey sorry for a bot spam [23:39:09] it's silented now