[00:00:14] yes it does [00:15:52] hm...I'm still not getting it... is it/was it a part of a bugfix? [00:16:55] traded two bugs for that one in r62345 [00:17:50] i need to head out, i'll see if Roan can help me with it tomorrow since it deals with code the he wrote [00:19:49] ok cool, gnight adam_miller [00:20:21] nkomura: it looks like prototype already has the linebreak code in it [00:20:35] how about deployment? [00:24:38] checking...(it's S L O W) [00:26:07] no it's not on deployment [00:28:22] nkomura: also, I think from what adam said above that we shouldn't commit his bug fixes until he's gotten that issue cleared. prototype won't have those though [00:28:58] but the prototype has the fixes for cursor jumping and line breaks? [01:01:54] *Natalie eyes werdna. [01:02:02] Adam, Trevor, and Roan aren't here. [15:01:14] RobH: hey, have you gotten a chance to look at tesla yet? [15:01:45] heh, writing an email to you right now. I have about 5 issues that I need to hit in the DC (where I am now) including flaggedrevs host [15:01:47] and tesla =] [15:01:53] heh [15:01:56] =] [15:02:03] thanks :) [15:02:08] right now getting flaggedrevs box setup, then will move to sdtpa and check out tesla [15:02:17] cool [15:02:20] you going to be around? [15:02:28] yeah. the majority of the day [15:02:40] I think I get what the issue you are having is, will be working on it more closely in about an hour or so [15:02:55] ok [15:03:23] today is going to be busy ^_^ [15:03:43] sounds like it [15:43:23] Ryan_Lane: I should rephrase, today is going to be busy, and I get to hand off a ton of stuff finished to other folks =] [15:43:29] thats the good kind of busy =] [15:43:31] makes them happy. [15:43:52] yeah, that is the good kind of busy :) [17:18:19] hi all [17:18:52] Morning [17:19:03] *RoanKattouw grabs dinner [17:20:19] Ryan_Lane: did you chagne the login details for tesla? [17:20:40] nope. [17:20:46] i added a user for myself [17:20:51] but didn't mess with root [17:21:01] hrmm [17:21:14] so the vsphere client cannot control it remotely? [17:21:20] (I do not have it installed, its windows.) [17:21:52] ah. yeah. it's windows only. you need the client [17:22:09] bleh, well, i cannot login to its drac and dunno why [17:22:12] i will grab a crash cart [17:22:28] i couldn't do anything via drac, since the key combos didn't work through a browser [17:23:18] I can connect via the vsphere client remotely, but it can only do so much. [17:23:43] it can't control the partitioning, it can only work with the partitions available [17:24:12] hrmm, lemme get into the system while you are here and we will get it squared away right now =] [17:27:39] ok. sounds good [17:32:57] ok, found the damned crash cart [17:33:02] *RobH needs to order a kvm console [17:33:39] heh [17:34:09] hey guys, should https://bugzilla.wikimedia.org/show_bug.cgi?id=22452 be added to the list of bugs on the techblog post about the iframe? [17:34:38] Ryan_Lane: there are 6 500gb disks, did you just want them with a single raid5 stripe? [17:34:50] we got some raritan kvms recently at work with rack power that can hook in. we can access all of it over https, with USB forwarding and all. Can down power, up power, etc. [17:35:07] also, should we still experience copy/paste issues? As in, single spaces being eaten on copy/paste from and to the editor? [17:35:08] i only need the kvm for the screen in the DC [17:35:09] RobH: yeah. I think that is how it is currently configured right? [17:35:12] yep [17:35:29] but for some reason, vmware is telling me only 258GB is available [17:35:32] since this drac is messed up somehow, wont let me login. [17:35:33] yea [17:35:54] "only" 258GB [17:36:03] MC8: out of 2.2TB ;) [17:36:09] ah [17:36:15] 256GB is useless! [17:36:20] yeah. it is [17:36:21] I was too busy going "omg! I want that memory stick!" [17:36:23] 258 whatever =] [17:36:35] MC8: we are talking disk space. heh [17:36:39] any system that doesnt hols 48 1TB disks is tiny [17:36:47] only 32GB of RAM if I recall correctly [17:36:49] i am spoiled due to the sun fire x4540 [17:36:50] heh [17:36:59] heh [17:37:36] yeah, those are nice. I'm spoiled from netapps and storagetek equipment too [17:38:15] both of which are far too expensive :) [17:38:47] hrmm [17:39:34] what the hell. [17:39:37] ? [17:39:46] i set the goddamn password, its still not working... [17:39:47] grrrr [17:41:24] ok, set it to something stupid [17:41:31] seems drac disliked some non-standard character [17:41:40] ok, its booting up into esxi now. [17:43:25] hypervisor loadin. [17:45:51] Ryan_Lane: so if i cannot login to hypervisor, you know how to reset the password by chance? [17:46:13] RobH: I should be able to change it through the vim client [17:46:29] I see your issue though, the drac doesnt show hypervidor [17:46:34] hypervisor even. [17:46:46] just reset the primary password to something easier and PM it to me? [17:47:07] it'll take me a few minutes to get access to the client. my roommate is working at home today, and she is the only one with windows installed. [17:47:24] Ryan_Lane: wait [17:47:25] im in. [17:47:27] ah [17:47:29] ok. cool [17:47:30] damned ariel font. [17:48:23] Ok, what exactly do i need to do in hypervisor. I do not see any damned hardware setup, other than resetting to default and rerunning the wizard [17:48:27] RoanKattouw_away: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/62341 [17:49:29] testing the network settings shows them to be fine [17:49:45] yeah, network is fine [17:50:09] RobH: looks like storage is set up when ESXi is installed... so it's likely dell screwed it up [17:50:41] it may have to be reinstalled [17:50:42] ok, i am going to reset it to factory defaults and reload from scratch [17:52:36] RobH: documentation says: "the disk-formatting software retains [17:52:36] that may not fix it though, but we will see. [17:52:38] existing diagnostic partitions that are created by the hardware vendor." [17:53:20] so, if they pre-formatted it in some insane way, it may not. [17:54:21] still loadin. [17:54:30] but i think yer right and this wont do it [17:54:33] RobH: "Automatic disk formatting occurs the first time you power on the host and when you power on the host after [17:54:35] you reset the host to the configuration defaults" [17:54:41] hrmm [17:54:43] thats promising. [17:54:49] yeah [17:55:07] setting back up its networking now [17:55:12] then you should be able to connect [17:55:16] dell didnt setup the raid5 [17:55:22] so i guess exi saw only the single disk or whatever. [17:56:18] k. lemme check it. gimme a sec [17:56:51] ok [17:56:55] its booted and network is online [17:56:58] lemme know =] [17:57:09] and the passwords are all set to the one i emailed you before [17:57:26] I guess I need to set up a windows partition on this thing again =P [17:57:44] *RobH has to purchase new version of parallels though, snow leopard doesnt like the old version he was using. [17:58:17] I don't even have a copy of windows :). Naoko is getting me one thankfully [17:58:55] I have an old copy of XP Pro [17:59:00] that suffices for my windows stuff. [17:59:04] vista is overkill. [17:59:27] Ryan_Lane: are you controlling it via the linux cli software download then? [17:59:48] no. using my roommate's computer for now [18:00:01] lol [18:00:13] RobH: password is the original one you gave me? [18:00:21] yep [18:00:23] ok [18:01:45] RobH: not working... [18:01:55] the password that s [18:01:57] is [18:02:18] hrmm [18:03:05] i am gonna change it [18:03:08] pming it to you now [18:03:10] k [18:03:30] sent [18:04:44] hrmm, shows no datastores. [18:04:51] *RobH is just tinkering in browser [18:05:10] RobH: still shows only 258GB available [18:05:43] hrmm [18:05:51] so what is the solution, do we need to reload the esxi? [18:05:58] *RobH has no idea how to do that on thsi ssytem) [18:06:20] I'm not sure if that'll help [18:06:31] maybe re-make the raid5? [18:06:41] it looks like dell preformatted it in some screwy way [18:07:07] it may be that system services bullsht and such [18:07:10] let me reboot and try removing it [18:07:18] remaking the raid5 should wipe out the partitioning [18:07:43] yea, i am going to wipe it, and reinitialize it. [18:07:45] next we'll find out they installed ESXi on the harddrives instead of the flash :D [18:07:55] we will know soon! =] [18:07:58] heh [18:08:46] is the xserve atleast working well for ya? =] [18:08:54] its the prettiest server in the rack =] [18:10:25] ok, dropped raid array, recreated, and its doing a full initialization of the disks in the array now [18:10:47] heh [18:10:52] yeah, the xserve is working great [18:11:19] RobH: even prettier than some of your sun systems? [18:12:16] ok, somethign interesting. [18:12:21] in the raid controller bios [18:12:48] ??? Select bootable VD: [18:12:53] ??? ??? VD 0 2382080 MB ??? ??? ??? [18:13:05] well that looks suspect dont it? [18:13:36] yeah, very much so [18:14:04] wait [18:14:09] wouldn't that be 2.2TB? [18:14:44] yep [18:15:03] so that should be fine [18:15:06] ok, rebooting it now. [18:15:10] ok [18:15:46] lets see if this works, its bootin. [18:16:56] hypervisor is loadin. [18:17:32] pdhanda: your server is setup, you have an email on what i need next (your pub key) [18:17:49] Ryan_Lane: Ok, its up, take a look? [18:18:08] k. gimme a sec [18:19:35] arrrrrrggghhhhh. [18:19:59] still saying only 278GB is available. [18:20:19] well, thats annoying aint it =P [18:20:33] any ideas or shall I get a dell support ticket opened later today? [18:21:19] I'm at a loss. i guess we should open a dell support ticket. [18:21:32] i am asking my esx using buddy too [18:21:44] AFAIK, there isn't any weird VMFS limitation on ESXi with respect to local disks [18:22:57] though, I've never had to use local disks. I always do NFS shared storage. [18:23:15] yea oddness [18:28:11] RobH: btw, I'll be around Monday, and Tuesday next week, most of the day. Lundi Gras and Mardi Gras... [18:29:06] TrevorParscal: What about that rev? [18:29:18] i followed up [18:29:34] Ah yes [18:30:04] OK looks good [18:30:10] Ryan_Lane: my buddy says for 2TB we have to split it into 2 1TB volumes or use GPT. [18:30:20] if we raid10 [18:30:22] GPT? [18:30:27] no clue, askign him. [18:30:35] he says 'its a pain in the ass' [18:30:39] lol [18:30:50] can we just use smaller overall volumes without error? [18:30:56] yeah. thats fine [18:30:58] i rather try that now and see if we cannot prove thats the issue [18:31:03] ok, rebooting and making smaller VDs [18:31:31] any way to make the raid controller show multiple LUNs? [18:32:25] we will see [18:34:59] Ryan_Lane: Ok, it set it up as a single raid10 VD that is 1.4 TB [18:35:08] k. [18:35:10] plus raid10 is nice and fast, if performance every is an issue. [18:35:13] RoanKattouw: how's it going? [18:35:15] yeah [18:35:16] lets see if that solves it, its rebooting. [18:35:27] seems from what shane tells me [18:35:37] esx and esxi have issues with 2TB or larger virtual drives [18:35:46] unless you do a bunch of stuff to add the support. [18:35:57] *RobH has no idea, this is his first time messing with esxi [18:36:01] RobH: raid controller isn't fancy enough to divide a raid set into multiple luns? [18:36:07] TrevorParscal: I've been poking at some stuff this afternoon (my time), as a consequence of which I'm so much behind on my TV shows I have to stop talking to you in 10 mins :) [18:36:07] oh [18:36:26] I dunno if it can do multiple luns, I just wanted to see if this makes it viewable. [18:36:29] RobH: that's virtual disk drives. not VMFS filesystems [18:36:46] hrmm [18:36:52] and that's true. it has a hard time with VDDs that large [18:37:01] so thats not this issue eh? [18:37:04] well hell [18:37:09] its back up and online, wanna check or nah? [18:37:11] RoanKattouw: just keep in touch [18:37:15] Sure [18:37:16] this may solve the issue anyway. let's see [18:37:20] we're poking at more edges cases for pasting [18:37:22] If you need stuff deployed poke me [18:38:03] come on tesla, work damn you! [18:39:18] RobH: worked [18:39:40] now I need to figure out what block size I should use :) [18:39:57] RobH: thanks for the help [18:40:17] still don't understand why the raid5 setup was an issue [18:40:24] me either. [18:40:38] I would worry about it, but if this fix works well enough, shall we leave it in place and go with that? [18:40:51] yeah, this is fine [18:41:00] 1.3TB is still quite a bit [18:41:02] \o/ that makes all my MUST DO items done. [18:41:17] Ryan_Lane: plus its fully redundant, can lose half the disks [18:41:21] yeah [18:41:39] ok, I am going to detach the cart and get back to work on other items [18:41:39] and faster :) [18:41:41] which should help [18:41:56] thanks again. [18:42:29] *RoanKattouw_away imagines a dumptruck full of shell bugs pulling up at RobH 's front door [18:43:04] hahahaha, bugs [18:43:11] i wish i was workin on those [18:43:19] *RobH is going to go install the power in the new rows in sdtpa [18:43:30] then rack a bunch of stuff htat cannot be used yet but needs to be OFF the floor =] [18:51:57] parutron: hi, are u there? [18:58:47] RoanKattouw_away: you there? [19:44:17] hmm [19:45:36] IE8 issue with char insert (the en.wp one). [19:45:59] cursor jumps one char to the right apparently. [19:48:03] this is with beta off. that's the weird thing... [19:57:56] http://en.wikipedia.org/wiki/Wikipedia:Village_pump_(technical)#Unwanted_space_after_character_I_click_on [20:02:53] RoanKattouw_away: [20:02:56] hello? [20:03:04] line 880 of jquery.wiki8Editor.js [20:03:09] where did body come from? [20:03:14] don't you mean sc? [20:09:09] I don't see body on that line, lemme svn up [20:11:39] OK now 880 is blank [20:13:43] TrevorParscal: What's the *correct* line number? [20:18:18] is there any IE8 specific code that runs when the editor is not enabled ? [20:19:38] Not that I know of [20:19:48] wikiEditor doesn't run any code when it's disabeld [20:19:56] weird. [20:20:35] *thedj checks the edittools specific code we recently added. [20:21:38] RoanKattouw: var n = body.firstChild; [20:22:24] OK that's line 703 for me [20:22:51] No 866 [20:22:53] Yes, should be sc [20:22:57] Copy-paste mistake [20:23:30] thre's another mistake just below it then too [20:24:07] Why is that? [20:24:13] Oh yeah [20:24:16] Another body [20:24:18] *RoanKattouw fails [20:25:25] :) [20:25:55] please look at my commit http://www.mediawiki.org/wiki/Special:Code/MediaWiki/r62375 [20:26:57] grabbin food [20:26:57] RoanKattouw: did you get your body.firstChild thing fixed? i have that fixed locally and could commit [20:27:38] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/62324#c5760 [20:27:55] mention this rev in your comment so it shows up as a follow up plz [20:28:29] adam_miller: s/body/sc [20:28:49] ? [20:28:55] no comprede [20:29:04] adam_miller: "Fix r62324: blah blah" --> make that your commit summary [20:29:24] The phrase r62324 will trigger CodeReview to link the revisions [20:29:33] oh yeah i know that trick [20:29:56] i think i have the context one committed to memory now too [20:30:09] c = $j('textarea').data('wikiEditor-context') [20:30:20] will never remember the get parameter for changing language though [20:31:16] adam_miller: uselang [20:31:30] And there's useskin too [20:31:54] TrevorParscal: You forgot to update wikiEditor.toc.js line 110 [20:37:02] RoanKattouw: i need your help with a range issue [20:37:11] adam_miller: That rev is needlessly complex. The if branch has established that sc.nodeName == 'body' so sc IS the body, just use sc instead of d.body [20:37:47] adam_miller: Also is updateHistory() just a copypaste? I'll take your word for it :) just wanna know whether you made any changes while moving [20:37:57] i'm a needlessly complex individual [20:38:03] i'll simplify it though [20:38:53] and yes, just copy and paste, i was starting on moving some history actions to change event rather than delayed change to avoid unexpected behavior, so wanted to break it out [20:39:05] Yeah sure, that's a good thing [20:39:11] Just wanna save effort reviewing when possible [20:39:28] And you wouldn't believe it but some people change files and rename them in the same commit [20:40:21] RoanKattouw: have you played with undo/redo in firefox today? [20:40:27] No [20:41:05] there's this issue when the cursor get's repositioned from an undo or redo at the end of a line of text where you can't type more characters [20:41:15] Ah yeah you mentioned that [20:41:17] backspace, return and arrow keys work [20:41:33] Didn't that have something to do with a
? [20:41:42] easiest way to replicate it is to start with a blank page, type somethhing, undo and then try to type again [20:42:13] yeah, when i inspect the range that the selection is being set to, the node is always a br element when the problem occurs [20:42:36] http://en.wikipedia.org/wiki/Wikipedia:Village_pump_(technical)#Usability_initiative_causes_complete_editing_breakage [20:42:50] safari doesn't seem to care, but firefox does this weird dont-let-me-type thing [20:44:10] nimish_g: when you asked if i committed a loading graphic that doesn't go away the other day, what browser were you using? [20:44:44] and why did you not press that issue any further? did it go away? [20:45:05] thedj: I replied [20:45:37] TrevorParscal, TrevorParscal_: I want to go wit Naoko's suggestion of not doing all the iframe stuff if only the toolbar is enabled, see http://en.wikipedia.org/wiki/Wikipedia:Village_pump_(technical)#Usability_initiative_causes_complete_editing_breakage for our latest bug from hell [20:46:42] RoanKattouw: why did that cover thing get deployed? i thought that was just a trial and that no one really liked it anyway? [20:46:54] adam_miller: It seems to be an issue with Firefox's cross-domain restrictions enforcement. When an external script is pulled in (from say Google.com), Firefox will error out with an error like "Permission denied for to get property Window.document from ." and it'll never get to removing the cover [20:47:01] So we can disable the cover for now [20:47:12] k i'll get rid of it [20:47:12] *RoanKattouw glares at the office's wireless [20:47:25] Commenting it out is good enough [20:47:26] what office? [20:47:37] The San Francisco office [20:47:48] For some reason Trevor always has wireless problems there [20:48:00] ohh haha i thought you were there for a second [20:48:05] And the first thing I do when I get in is look for a cable [20:48:14] TrevorParscal, TrevorParscal_: I want to go wit Naoko's suggestion of not doing all the iframe stuff if only the toolbar is enabled, see http://en.wikipedia.org/wiki/Wikipedia:Village_pump_(technical)#Usability_initiative_causes_complete_editing_breakage for our latest bug from hell [20:48:42] RoanKattouw: well, that's an interesting idea [20:48:53] NTOC would not be an option for those people [20:49:01] i confirm the bug btw. [20:49:11] True [20:49:22] TrevorParscal: Which is why I'm having Adam comment out the loading cover as well [20:49:27] (i load an addthis gadget and google custom search gadget, and get the same problem) [20:49:28] thedj: Yeah I've also reproduced it [20:49:49] https://bugzilla.mozilla.org/show_bug.cgi?id=462238 seems related, commented on it but couldn't come up with a small testcase [20:50:20] i have Error: $("

").addClass("section-" + structure[i].index).data("index", structure[i].index) is undefined [20:50:23] Source File: http://bits.wikimedia.org/w/extensions/UsabilityInitiative/js/plugins.combined.min.js?urid=257z17 [20:50:27] Line: 654 [20:50:29] btw [20:50:41] WTF [20:50:46] not sure if that is due to timeout or for another reason. [20:51:15] Hmm wait [20:51:21] If structure[i].index is undefined... bleh [20:51:40] it might be my hotcat loading jquery as well. [20:51:42] lemme check [20:51:49] thedj: Could very well be [20:54:21] RoanKattouw: loading div commented out [20:54:29] Thanks [20:54:41] TrevorParscal: Permission to immediately deploy Adam's commenting out of the loading div? [20:55:02] TrevorParscal: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/62380 [20:55:38] lookin [20:55:55] granted [20:56:14] RoanKattouw: hmm, i disabled my jquery loading plugins, and the problem remains on FF when that loading div hangs. [20:56:34] works for me [20:58:18] TrevorParscal: Enable the Google gadget thedj mentions [20:58:34] And Firebug will tell you "Permission denied for to get property Window.document from ." [20:58:52] Which of course is totally ridiculous, but it seems loading an external script wakes up Firefox's cross-domain checking [20:59:05] icw an iframe [20:59:57] Yes [21:00:03] I tried to reproduce it in a small testcase but failed [21:00:33] RoanKattouw: yes, i see that [21:02:59] OK loading cover should be disabled now [21:03:35] I'm gonna hack up something so the toolbar can work without the iframe [21:03:50] *RoanKattouw looks around for Naoko [21:04:26] works now [21:12:14] Gavia Immer is one of the bigger anti-usability project folks around :P [21:13:55] IT Crowd lunch is over now... [21:15:00] RoanKattouw: TOC on wikipedia seems to not be working [21:15:04] like, not showing up [21:15:15] In combination with that bogus cross-domain check error? [21:15:41] havn't figured it out yet [21:15:54] it's not swtiching to the iframe [21:15:58] it's showing the textarea [21:16:16] Yes because the iframe setup errors out [21:16:21] well [21:16:26] I just committed a horrible hack that disables the iframe for toolbar-only users [21:16:29] yes - that was with the google thing [21:16:32] Please reveiw [21:16:38] so yeah, it's working fine otehrwise [21:16:42] I have to go watch my show now :) will be back in an hour [21:16:51] k [21:20:03] TrevorParscal: So in the meantime please review and test my rev :) I briefly tested it but the whole process of writing, testing and committing took like 5 mins [21:52:03] TrevorParscal: So in the meantime please review and test my rev :) I briefly tested it but the whole process of writing, testing and committing took like 5 mins [21:52:57] RoanKattouw_away: can we blacklist webkit until we can deploy on tueday? [21:53:14] RoanKattouw_away: which rev? [21:53:27] fixing your getCaretPositon? [21:53:48] TrevorParscal: No my most recent one, hacking out the iframe for toolbar-only users [21:53:57] Yes, we can, I think Erik or Tim suggested taht [21:54:04] RoanKattouw_away: oh [21:54:15] I thought that was going to be a very big patch [21:54:20] You make the change in SVN for Webkit blacklisting, I'll deploy it in the next commercial break [21:54:36] TrevorParscal: If you're willing to put up with extreme ugliness it's a very small patch [21:54:47] RoanKattouw_away: rev #? [21:55:30] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/62384 [21:55:39] *RoanKattouw_away teaches TrevorParscal about http://www.mediawiki.org/wiki/Special:Code/MediaWiki/author/catrope [21:56:02] shhh [21:56:03]