[10:18:32] adam_miller: Thanks for the search bug fix. However, there are changes in the combined and min files I can't account for. Try running make -B ? [10:19:55] no change locally after make -B [10:20:15] i see those now in the diff though from r66565 [10:21:53] *RoanKattouw is 200+ revs behind on review [10:22:26] I'll un-fixme the rev, doesn't seem to be your fault [12:08:37] are there any plans on how the switch for other wiki's will be announced ? [12:08:45] will there be blogpostings ? [12:09:08] I'm sure there will [12:09:24] Ask Jay, Moka or Naoko [12:09:33] (Naoko has taken today off) [12:11:29] thedj: I can ask someone in the office to ask Jay or Moka to get on IRC if you like [12:12:29] i'll just tell folks from other wikis to watch the blog, as the most likely point of future information [12:13:15] Yes [12:13:31] But it would be nice to get a definitive answer as to which communication channels will be used [12:13:43] yeah. it would indeed. [12:13:48] And if you have suggestions, we'll be happy to hear them of coures [12:14:16] You could e-mail Jay (jwalsh@wm.o) [12:14:36] i think i'll do that. [14:16:00] very late to the conversation but -- The best notification for other wikis is probably going to end up being Central/Site Notices and VP posts I think. The further you get away from En the less likely it is that they are going to be reading the block/mailing lists. Even at SimpleWiki (which has a lot of En editors) most of them don't. [14:16:05] mmm blog not block [14:17:56] This is about switching on Vector on wikis that are not enwiki [14:17:59] So en readership is not really relevant [14:20:46] RoanKattouw: Simple search is still turned off, correct? [14:21:05] Yes [14:21:10] There were caching issues with turning it off [14:21:19] Which should have been resolved now [14:21:30] do users need to shift-refresh? [14:21:36] Yes, possibly [14:22:25] There's just sooo many layers of caching involved [14:23:07] There's Squid, Varnish and the browser cache [14:24:07] howief: BTW (re: e-mail) I don't know squat about search stats, but I think thedj does [14:24:18] There were quite a few hits for 'null' when the buggy code was live [14:24:25] got it [14:24:45] thedj: do you know how to retrieve daily search requests for english wikipedia? [14:25:42] howief: rainman_sr should know that. [14:26:32] thanks [14:26:40] i don't see him online right now [14:26:42] i'll keep an eye out [14:27:39] http://stats.grok.se/en/201005/Special%3ASearch says something though. [14:27:56] on the 13th, we had our lowest amount of visitors to Special:Search in a few months it seems. [14:28:16] (Note that search bar searches also go through Special:Search) [14:29:04] You mean the Go redirects... [14:29:12] Yes, those too [14:29:37] so what does this page actually show? [14:29:41] RoanKattouw: yea sorry that sounded weird, I was just trying to say that even the editors there (who are often en editors) don't read the blog/mailing lists [14:29:45] number of times the search results page are viewed? [14:29:59] Jamesofur: Right. I totally agree we need to do CentralNotice [14:30:10] Preferably we launch them right after we set the date [14:30:20] *Jamesofur nods [14:31:50] howief: visits to Special:Search + uses of Go button I think. [14:32:04] but i'm not totally sure. [14:32:21] so what if a user clicks on the type-ahead result? [14:33:27] howief: it submits the searchform, to Special:Search [14:33:54] i suspect that counts as a "visit", but really I cannot say 100% [14:34:30] i'm trying to figure out the best way to interpret the data [14:34:40] it looks like search volume has been trending down a little in april [14:34:57] it's clear that volume went down on 5/13 [14:35:11] howief: you should probably at the very least compare it to overall visits to enwp. [14:35:30] my guess is that's mainly due to the bug, possibly due to the placement [14:35:36] but then it goes up again [14:35:57] thedj: yes, normalizing would be good [14:36:08] but i need more significant figures than just 2 for the normalization to be meaningful [14:36:49] howief: if you really want to know, best ask Erik Zachte [14:36:56] 5,069,999 and 5,149,999 will swing the ratios [14:37:08] even tho both round to 5.1M [14:37:12] yeah i sent him an email [14:37:43] *thedj thinks bit bigger search field + no bugs will probably fix most concerns. [14:38:08] one thing that annoys me *a lot* with the new location of the search bar, is the fact that suggestions (larger than the search box) go off-screen on the right. [14:38:40] guillom: actually, that is just because we now use mwsuggest again. [14:39:05] thedj: oh? is it supposed to behave differently with the new search ? [14:39:06] guillom: normally, with the new suggestions, it should clip properly and show the start and the end of the suggestion with ... in the middle. [14:39:12] Yes [14:39:19] We're seeking to get rid of that (partially) aswell [14:39:31] is the new search also disabled on Commons? [14:39:33] guillom: see also http://prototype.wikimedia.org/en.wikipedia.org/Main_Page [14:40:38] I see [14:40:53] just enable the fraking new search already, then :P [14:41:18] *thedj would like 15em for the searchfield. But i think that does require looking into the "tab colliding" issue on "non-wide" rendering of the page. [14:41:38] guillom: well it had a few serious problems as well :D [14:41:48] thedj: I say yes to a bigger search field as well [14:41:51] thedj: AFAIK the current plan is to widen it on-demand [14:41:54] it's ridiculously small [14:42:05] *guillom whines [14:42:15] RoanKattouw: the field, or the suggestions ? [14:42:31] Both, I think [14:42:36] *RoanKattouw digs up mockups [14:42:55] RoanKattouw: like on .focus() make wide, on .unfocus() make slim again ? [14:43:10] Not sure, digging up relevant e-mail [14:45:00] Yes, it's on focus/unfocus [14:45:13] hmm, i might actually like that..... [14:45:18] Would make the search box grow towards the left and make its border blue [14:45:25] Suggestions thingy automatically adapts to search box size [14:45:58] *cary replaces the Bangla character in the new globe with Pedobear [15:00:28] have any of you taken a look at the following post for the font size issue: http://en.wikipedia.org/wiki/Wikipedia:User_experience_feedback#A_hint_on_why_the_font_size_differs [15:03:37] does that talk about the use of 0.8 em in the bodycontent class [15:03:46] I should say ID [15:07:04] I asked Trevor about this [15:07:09] The strange thing is [15:07:26] That we've had both "OMG the font is small" and "OMG the font is huge" [15:08:01] From what Trevor said, Monobook was using a number of tricks to get its desired font size, but there was a slight variance in how browsers behaved with that [15:09:02] Oh the author is Gabriel Wicke [15:09:08] Who, presumably, wrote Monobook [15:09:11] Ah [15:09:14] He did write Monobook [15:09:22] He presumably also wrote this specific piece of Monobook CSS [15:20:10] yeah, monobook used the cleverchimp trick indeed. [15:21:03] So I guess Gabriel and Trevor get to argue about this [17:22:18] nn