[05:14:25] 10Scoring-platform-team (Current), 10ORES: Deploy ORES advanced editquality models to cawiki - https://phabricator.wikimedia.org/T192501#4141737 (10Townie) Done, see [[https://ca.wikipedia.org/wiki/Tema:Ubjp32r9rd68pqui|this post ]] [05:14:26] 10[1] 04https://meta.wikimedia.org/wiki/https://ca.wikipedia.org/wiki/Tema:Ubjp32r9rd68pqui [08:11:42] 10Scoring-platform-team (Current), 10ORES: Deploy ORES advanced editquality models to huwiki - https://phabricator.wikimedia.org/T192496#4141939 (10Tgr) That announcement seems fine. Or do you mean we should translate it into Hungarian? [08:36:49] 10Scoring-platform-team (Current), 10ORES: Deploy ORES advanced editquality models to lvwiki - https://phabricator.wikimedia.org/T192499#4141998 (10Papuass) Here is what I came up with: == Vikipēdijā latviešu valodā pieslēgts ORES rīks == Sveicināti! [[File:ORES extension.highlighted change.png|thumb|right|"... [11:51:23] 10Scoring-platform-team (Current), 10ORES: Deploy ORES advanced editquality models to arwiki - https://phabricator.wikimedia.org/T192498#4142500 (10Ghassanmas) Yes of course ! I would be happy to work on that also its really an easy small task to do !. To follow up with deployment date, will be there an annou... [14:21:46] 10Scoring-platform-team (Current), 10ORES: Deploy ORES advanced editquality models to huwiki - https://phabricator.wikimedia.org/T192496#4142766 (10Bencemac) I'll do today. [17:08:08] * awight crawls out from under a rock and sits in a ray of sunlight [17:12:39] halfak: I had a question about the use of “reification” in your paper, whether there’s a specific background for using that word? If not, I find it distracting and maybe wrong… AFAICT, it’s about ideas becoming a thing in people’s minds, rather than ideas actually turning “real”/“physical”. I’m thinking we can say “values are embodied in technologies” rather than reified… [17:12:51] or maybe “brought to life” [17:12:55] and good morning [17:23:14] I love the theory of change section [17:23:47] * awight moves to the Talk page [17:35:05] 10Scoring-platform-team (Current), 10ORES: Deploy ORES advanced editquality models to huwiki - https://phabricator.wikimedia.org/T192496#4143582 (10awight) >>! In T192496#4141939, @Tgr wrote: > That announcement seems fine. Or do you mean we should translate it into Hungarian? I'd say, whatever language is be... [17:38:57] 10Scoring-platform-team (Current), 10ORES: Deploy ORES advanced editquality models to arwiki - https://phabricator.wikimedia.org/T192498#4143608 (10awight) Thanks! The deployment will take several steps, and it's only the very last step "Enable new models via configuration" which is visible to wiki users. We... [17:40:31] 10Scoring-platform-team (Current), 10ORES: Deploy ORES advanced editquality models to lvwiki - https://phabricator.wikimedia.org/T192499#4143626 (10awight) Wow, thank you! We can coordinate the last step "Enable new models via configuration" in this task, to make sure the announcement goes out at the right ti... [17:42:40] 10Scoring-platform-team (Current), 10ORES: Deploy ORES advanced editquality models to huwiki - https://phabricator.wikimedia.org/T192496#4143628 (10Bencemac) @awight [[ https://hu.wikipedia.org/wiki/Wikip%C3%A9dia:Kocsmafal_(egy%C3%A9b)#Az_ORES_m%C3%A1r_a_sp%C3%A1jzban_van | Done ]]. [17:42:41] 10[1] 04https://meta.wikimedia.org/wiki/https://hu.wikipedia.org/wiki/Wikip%25C3%25A9dia:Kocsmafal_%28egy%25C3%25A9b%29%23Az_ORES_m%25C3%25A1r_a_sp%25C3%25A1jzban_van [17:43:16] Wow, our stakeholders on all four pending wikis have already gotten back about the announcements! Quite a motivated crowd! [17:45:05] 10Scoring-platform-team (Current), 10ORES: Deploy ORES advanced editquality models to cawiki - https://phabricator.wikimedia.org/T192501#4143634 (10awight) Wonderful, thank you! We'll post more information here as the deployment goes forward. @Trizek-WMF, heads-up that this announcement has just been publish... [17:46:20] Woo! [17:46:40] Now only if I wasn't block on building models. awight want to take editquality over from me? [17:46:59] I think that might make sense given my cramming for the dealine, travel, and comp-day plans. [17:47:18] Then again, I think I can do a 90% comp day tomorrow and still build those models. [17:47:50] yes [17:48:24] It makes sense for me to grab that, too. I think Amir1 merged the key patch just a minute ago, unblocking... [17:48:54] Yeah, it’s merged. Weird edge case that makes me a bit uncomfortable. [17:48:58] I managed to make it to fiwiki before the most recent explosion. [17:49:17] check out stat1006:~halfak/projects/editquality/ [17:49:21] Oh, a new explosion? kk [17:49:33] yeah. fiwiki unreverted edits :| [17:49:56] ooh [17:52:23] btw, twentyafterfour pushed hard on the git-lfs stuff and all we’re waiting for now is for ops to run the deb packaging and deploy [17:52:47] halfak: Annoying semantics thing for you, https://www.mediawiki.org/wiki/Topic:Ubku0rk9hybh2e9h [17:53:18] Also, tell me when I can start making copyedit patches against the latex? [17:54:03] Right now :) [17:54:11] reification: is to make (something abstract) more concrete or real. [17:54:12] great [17:54:30] halfak: I saw your mail, just curious to know, what areas do you think could have used more work? [17:55:03] It has some Marxist baggage…. Just the basic word seems to be slightly different than that, though: https://en.wiktionary.org/wiki/reification [17:55:05] codezee, I think they are very high level. E.g. I'd like to refactor the introduction and related work since a bit of seminal research didn't make the cut. [17:55:17] > The *consideration* of an abstract thing as if it were concrete, or of an inanimate object as if it were living [17:55:18] https://en.wikipedia.org/wiki/Reification [17:55:21] emphasis mine [17:55:35] Reification (computer science), the creation of a data model [17:55:39] :| [17:55:43] lol [17:56:27] Are you talking about the fallacy? [17:56:28] Reification (fallacy), the fallacy of treating an abstraction as if it were a real thing [17:57:21] I think objectification is a good way to think about what I mean. By taking a subjective concept and turning it into an objective thing (like an algorithm) you reify it. [17:57:27] No, the two contexts that I have are the word itself, and its use in Marxist literature beginning with Luckács [17:57:36] Seems like marx uses the term differently [17:58:39] The CS definition is pretty close to what you mean—however, it’s not necessarily being made into a first-class concept. IMO, the process you’re describing is more like “embedding” cos it may still be implicit, but values are encoded into e.g. the training data. [17:59:31] Na. ORES doesn't reify anything [17:59:37] It makes reification easier. [17:59:46] For a more diverse set of standpoints. [18:00:08] also—regardless of whether or not there’s a way to make the word fit, my philosophy is to strip out any jargon unless it’s a main theme of the writing and is well explained in our context... [18:00:34] But it’s only used 3 times, so I think we’re good either way, and readers will make it through [18:01:02] I think reify fits well with the successor/standpoint theming. [18:01:10] I'm trying to find where harraway/harding use it. [18:01:15] kk! [18:01:27] [sp] Haraway [18:02:08] Put it in a PR! [18:02:43] :D [18:02:47] Damn, you github! [18:02:49] https://github.com/halfak/ores-paper/pull/1/files [18:02:53] I can't read this diff at all [18:03:08] ouch [18:04:24] * halfak considers blindly accepting srrodlund's changes [18:05:30] just wait until we have merge conflicts [18:05:44] halfak: I can send you my text files directly if you like. [18:05:55] na. It's good. not sure what would work better anyway :\ [18:06:09] Any big changes I should be aware of? [18:06:17] Would work better if GitHub were open source and we could add word diffs [18:06:22] Not on that file. It's just minor proofreading stuff [18:07:07] I just made some paragraph breaks, added some commas took out some qualifers [18:07:39] https://en.wikipedia.org/w/index.php?title=User%3AEpochFail%2FSandbox&type=revision&diff=837258937&oldid=837258905 [18:07:45] Mediawiki is bad at that diff too [18:07:47] Oh well :) [18:08:08] Hrm. Yeah, let me know if you want the plain text file. I can send it on. [18:08:17] It's cool. {{merged}} [18:08:17] 10[2] 04https://meta.wikimedia.org/wiki/Template:merged [18:08:18] my eyes are bleeding [18:08:29] you merged it already :-) [18:13:43] * halfak continues to work on figures [18:17:39] I keep wanting to read “automated meditation" [18:19:43] Wouldn't that be nice? [18:19:57] It's alive! [18:20:04] Just got the last figure to stop being stupid [18:20:20] Nettrom, I think you might be the most highly cited author in this paper :D [18:20:40] thanks :D [18:20:45] Toss up between you and me citing myself. [18:20:48] lol [18:21:35] usually it’s me having to cite like five of your papers, so it sounds fair to me :D [18:22:13] It's a viscous cycle [18:24:18] * halfak works on categorization [18:28:58] my PR is ballooning… better merge it quick! [18:29:26] or better, push your work so I can keep mine rebased [18:29:52] I think the emdashes are supposed to be "--" [18:29:53] * halfak checks [18:30:07] nevermind! [18:30:09] TIL [18:30:17] OK I'll push and let you rebase [18:30:38] kk [18:30:54] yeah I need to fire up latex locally, that one had be scared, too. [18:31:10] {{done}} [18:31:10] How efficient, halfak! [18:31:23] * halfak checks on grammarbot's PR [18:31:31] So incoming change to 2_related_work.tex [18:32:14] merged! [18:32:48] PDF rebuilt [18:33:32] You know... I think we might actually finish before the deadline! [18:33:48] Like, maybe I can just enjoy my flight to SEA tonight :D [18:34:08] u need to check in figures/ores_architecture.tex [18:34:18] oh wat, it *is* there [18:34:31] Renders fine for me [18:34:49] grr, I’m using TexShop and it isn’t able to do the relative import from sections/ to figures/ [18:34:53] The PDF rederer in github is bad [18:35:03] I'm using rubber/pdflatex [18:35:06] nbd, I can ignore errors locally [18:35:10] I’m on mac [18:35:31] ugh. Macs. Almost as bad as latex [18:35:46] lol [18:35:56] Latex was writting by masochistic computer scientists, [18:36:15] I used to do random college essays in latex, and one of my profs literally took me to the basement and interrogated me about what was going on. [18:37:08] lol good or bad from your prof's point of view? [18:37:33] I used to do my MathLogic homework in latex. I didn't usually score that well, but I often got a note about how awesomely legible my wrong proofs were. [18:38:46] * halfak finds a rogue tag [18:44:42] Just pushed a quote fix commit that spans most of the sections. [18:45:07] awight, your change ready for review now? [18:45:18] /me rebases [18:45:19] Or should I wait? [18:45:23] :) [18:45:36] nothing to rebase on [18:45:36] yeah it’s good [18:45:48] oh I grabbed the wrong remote hold up [18:46:12] conflicts [18:46:20] They're gonna be a pain too :( [18:50:21] ty for obsessively splitting into smaller files! [18:50:32] :D J-Mo was giving me a hard time for it [18:50:33] lol [18:50:47] It's one of my latex coping mechanisms [18:50:49] halfak: I finished up to section 3 so far. I have to go grab some lunch. I am so hungry -- but I will get back at it in a bit [18:51:00] OK sounds good. Thanks srrodlund! [18:51:19] J-Mo was just confused and thought he would have to do that with his other submission as well, when he migrated it to a new proceedings template. but disaster has been averted. [18:51:21] awight, I'll wait for your rebase, then I'll take on any conflicts from srrodlund's most recent work. [18:51:32] \o. [18:51:38] {{done}} [18:51:38] How efficient, awight! [18:51:38] / [18:51:52] also -- After, I'm going to send you some tips about some recurring things I see, so you'll be able to do some targeted proofreading in the future :-) [18:51:55] J-Mo1, you still on for a read in 1.15 hours? [18:52:22] yes, I can manage it… just :) [18:52:23] Cool srrodlund. Bug fixes for Aaron-brain [18:52:40] J-Mo1, I suspect it'll go quickly given the effort that is taking place right now. [18:52:46] I’m afraid it’s an entire wetware update. May brick the system [18:53:01] back up? [18:54:20] lol [18:54:55] halfak: Just so I understand, “related work” is going to be an overview of prior scholarship and industry efforts which overlap with things in this paper, true? [18:55:31] Right. However, I don't think we really do related work justice here. [18:55:46] can we drop the section for this iteration, then? [18:55:53] Oh no [18:55:55] Absolutely not [18:55:56] :) kk [18:56:17] Let our reviewers yell at us and help us summarize more to their liking. [18:56:31] I’ll scrape around in my notes in case I have related things to offer. [18:56:49] +1 feel free to start a branch and iterate on what I have. [18:56:51] I think some of the HCI papers are relevant [18:56:55] sure [18:56:55] * halfak invites awight [18:58:18] * awight checks to see if all major revscoring updates happen the week before a paper deadline [18:59:49] PDF is updated. [19:07:37] 10Scoring-platform-team (Current), 10articlequality-modeling, 10User-Ladsgroup, 10artificial-intelligence: Train and test wp10 model for fawiki - https://phabricator.wikimedia.org/T190050#4143872 (10Halfak) So, one of the labeling campaigns is probably not getting pulled in. Want to check on that? [19:08:19] 10Scoring-platform-team (Current), 10MediaWiki-extensions-ORES, 10User-Ladsgroup: ORES Extension refactoring - https://phabricator.wikimedia.org/T183199#4143874 (10Halfak) [19:08:22] srrodlund: Make sure to pull the latest before starting in again. If you have local changes already, commit and push your changes before trying the merge… holler if it goes wrong. [19:10:11] o/ ewhit_ [19:10:17] just got a chance to come up for air. [19:10:22] what's up? [19:10:25] I wonder how the meta-page drafting is going :) [19:12:07] Pretty well, I think? There's a lead section, which I might need to trim [19:12:22] The methods section is also pretty complete, although subject to change as the project changes [19:12:35] There's also a timeline, which will be updated as needed [19:12:52] I've also posted the interview questions, and made some tweaks to fix what we talked about a while ago [19:12:59] link to current: https://meta.wikimedia.org/wiki/Research:Civil_Behavior_Interviews [19:13:00] Great. [19:13:33] I currently have a pre-test interview set up for next week, and am working on setting up a few more [19:13:49] Hopefully those will go well and I can finalize the interview measure and fill out IRB forms as soon as those are done [19:14:29] Anything look like it needs changing (assuming you have time to look, of course!) [19:15:00] Great. Nice to see the progress. We should loop in the Support & Safety folks again soon (Patrick and Sydney). [19:15:25] It'll be good to have them review a draft of the document before we start recruiting. [19:15:37] Sounds good! Want me to get in touch and try to set something up for after CHI/next week? [19:17:57] Yes please. :) [19:18:14] Will do! [19:19:17] Holy moley I just printed this thing out and it's heavy [19:19:24] 30 pages is a lot of paper! [19:22:16] halfak That is definitely heavy! I don't have access to anyone's calendars, is sending out an email asking about times alright, or should I request calendar access then schedule something when everyone is indicated as available? [19:23:58] Hmmm. Best way is for me to try to schedule. [19:24:09] * halfak thinks. [19:24:35] ok. Sorry about that! I think it's not having an in-organization email account. I can send calendar access requests though [19:26:36] ewhit_: Wow, what great project outline! The timeline looks well-paced and like the steps are concrete. [19:26:41] This one sentence confuses me: [19:26:44] > If these goals are accomplished, and accomplished in such a way that allows all discussants to contribute and be heard, then incivility is present. [19:26:54] I woulld have expected “civility is present” [19:26:56] whoops, yep, problem, thanks! [19:27:04] np [19:27:45] ewhit_, {{done}} [19:27:45] You rule, halfak! [19:27:53] Thank you! [19:27:56] * halfak goes back to paper reading hell [19:28:16] (actually I find wrapping up a paper quite enjoyable) [19:28:45] Especially once we've crossed the band aide line -- the point at which all problems will only receive band aides and nothing more. [19:30:02] ewhit_: Another question, where’s the translation between our definition of civility and the users’ rubric for judging an observation civil or not? [19:30:43] That's what I'm hoping to find out through interviews. Incivility has such an ambiguous definition, I'd like to figure out what editors think before making any concrete plans [19:30:46] e.g. are the users reading fine print under the button, explaining what “civil” is? Do we send them to training first? [19:30:56] so rad [19:31:04] The users will be doing a cognitive walk-through with me [19:31:31] Right now I'm debating selecting an "uncivil" page for them to walk me through, or having them select a page they think is uncivil before hand, and walk me through that [19:31:43] Hence the pre-tests! These things are always complicated haha [19:32:14] What software platforms might be used for the tool/algorithm? I assume we’re trying to reuse wikilabels and the rest of ORES? [19:32:54] That's the current line of thinking, but I'm hoping to get some input from others about that, since I'm not sure what the benefits and drawbacks of that would be [19:33:35] Similarly, I’m wondering on what level we’re observing behavior when judging civility, i.e. are we keeping the granularity of 1 revision? [19:34:48] I guess we can always analyze larger units like conversation sessions by looking at each revision and following the reply graph while scoring each revision. [19:35:02] That's also something I'm working through right now, and I'd love input from others. I do want to move this to talk pages, as opposed to revisions. halfak brought up that posts from very experienced users/bots/etc get fatiguing for editors or MTurkers to review, and it might be best to exclude those for coding purposes [19:35:36] +1 [19:35:37] Mmm, and I think I might be thinking of talk pages as more similar to comment threads than revision chains, so that's definitely a conceptual piece I need to think through [19:35:59] There are wikitext talk pages, and Structured Discussions (Flow) talk pages, too... [19:36:07] with different db schema, of course [19:36:17] Hmm. Thank you, I will look into that! [19:37:01] long story short, ORES is specialized in scoring wiki page revisions at the moment. We’d like to generalize but haven’t found the right justification, yet. [19:38:27] Makes sense. I'll look through some of the other tools available to see if anything else might be a good fit [19:39:36] ORES is currently scoring talk page revisions fwiw, so we already have access to those scores for purposes like preselecting abusive, mundane, or wonderful interactions. [19:40:01] I’m curious about the differences between bad interactions where only one partner is misbehaving, vs. when two people are fighting. [19:41:06] ooh, interesting. And yes, more ambiguous pages where it doesn't descend straight into flaming, or the incivility is much more subtle, are absolutely something I'm interested in. There's also an interview script for a cognitive walkthrough of "ambiguous" pages [19:41:10] maybe an unsupervised analysis would find other interesting clusters [19:41:29] cool! [19:44:55] Does ORES examine talk pages at all, or does it focus on the pages themselves? Sorry, I've looked through the documentation, but I'm still a bit unclear on that point [19:48:25] oh I think I might have a workaround [19:48:48] so, we have two ways of looking at a revision [19:48:59] as a diff, between the revision and its parent [19:49:10] or as a snapshot of the page at that revision [19:49:42] I think that lets us say, “this is a toxic conversation” vs. “this person is being a jerk” [19:49:54] ooh. Yes, that sounds good [19:50:11] damaging and goodfaith are “diff” type scores [19:50:26] articlequality is “revision” snapshot type [19:52:32] 10Scoring-platform-team (Current), 10JADE: Deploy JADE extension to production - https://phabricator.wikimedia.org/T183381#4144042 (10awight) [19:56:08] halfak: getting “not implemented” errors on data dependencies just means that I need to rebuild w_cache files, right? [20:26:17] halfak: I got a bit sidetracked. What is your deadline for sending? [20:35:00] subtle! See if you catch it, [20:35:09] …. sed -r 's/"(true|false)"/\1/g' | > datasets/fiwiki.flaggedrevs_approved... [20:49:44] (03CR) 10jenkins-bot: Localisation updates from https://translatewiki.net. [extensions/ORES] - 10https://gerrit.wikimedia.org/r/427797 (owner: 10L10n-bot) [20:55:26] this is weird, “http://enwp.org/:mw:ORES” [20:55:35] I like that it works, but “enwp” bothers me [20:57:26] Agreed. I wanted as short of a URL as I could get and that was the best option :| [20:57:55] awight, it matches quotes? [20:58:19] sorry, what quotes are these? [20:58:27] +1 short url for being less distracting, though [20:58:30] srrodlund, right now, I'm hoping to be DONE DONE in 1 hour :) [20:58:40] But I'll double-check the actual deadline. [20:58:44] okay I am working on a file right now [20:59:01] I will send a pull request for that one in just a few [20:59:43] Great. [20:59:43] BTW, all I'm doing is very minor stuff [20:59:55] so if you do end up needing to send, it will still be fine [21:00:01] Really thorough paper :-) [21:00:53] Official deadline is 5PM PDT [21:01:08] halfak: I’m making a small change to clarify that we don’t dynamically scale with demand, yet, but we have set the stage to eventually do so. [21:01:20] And I'll be on a flight then, I'll need to submit it at 3PM PDT before I leave to fly away O_O [21:01:22] mentioning to be sure that’s an accurate way to say it [21:01:31] awight, we used to! [21:01:34] Not automatically [21:01:46] In labs we spun up additional nodes to handle demand as demand grew [21:01:54] oic, nice [21:01:57] I’ll say that. [21:02:33] * halfak makes a safety submission just in case [21:03:44] You can grab my PR any time, I’m committing incrementally. [21:04:07] no PR that I see [21:04:17] I merged one a bit ago [21:08:45] I’m still in my fork, https://github.com/halfak/ores-paper/pull/4 [21:08:50] oh, you got it [21:08:57] :D [21:11:03] /102/3 [21:24:21] awight, am I still waiting on your change? [21:24:35] I have another one ready to commit [21:24:42] cool :) [21:25:01] I'll swing back in 20 minutes to do a last pass of merge, rebuild, and submit. [21:26:52] https://github.com/halfak/ores-paper/pull/11 [21:27:19] conflicts. [21:28:55] awight, you dealing with the conflict or should I? [21:29:00] I’m on it [21:29:34] {{done}} [21:29:34] You rule, awight! [21:29:34] thansk [21:35:39] srrodlund, am I waiting on your last PR? [21:35:39] awight: did you copy edit the 8_conclusion [21:36:06] I haven’t touched the conclusion yet [21:36:16] okay let me look at that one [21:36:37] Does this look right? https://github.com/halfak/ores-paper/blob/master/sections/8_conclusions_and_future_work.tex [21:36:48] oh sorry, I misunderstood in a weird way [21:36:52] carry on :) [21:37:24] I’m just starting section 5_innovation fwiw [21:44:45] Hey folks. It's been 20 minutes. I'm back and ready to merge the final bits. [21:45:47] awight & srrodlund: ^ [21:46:21] +1 [21:46:25] 10Scoring-platform-team, 10Scap, 10Patch-For-Review, 10Release-Engineering-Team (Kanban): Support git-lfs - https://phabricator.wikimedia.org/T180627#4144423 (10awight) @mmodell Do you have an idea how the various git caches will react to the large files? I noticed that the .git/modules/submodules/assets... [21:46:55] I see no PRs. Should I be waiting on anything>? [21:46:56] I have a multi-character PR but nothing deeper yet [21:47:42] mvm that’s merged [21:47:46] *nvm [21:48:17] I have finished 8 [21:49:19] srrodlund, has your last bit of work been merged? [21:49:47] * halfak is giddy that this all came together and anxious to run away to the airport [21:50:23] they are ready to merge [21:50:33] All only needed light proofreading [21:50:35] :-) [21:50:59] ? [21:51:15] I don't see any PRs [21:51:30] Maybe they were already merged? [21:52:22] Sorry I need to click one more button; they should be ready to merge now [21:54:11] OK got em! [21:55:07] Yay!!@ [21:55:57] Pushed a new PDF. [21:56:01] I think we're set. [21:57:23] And submitted! [21:57:25] Good work team! [21:57:29] * halfak collapses in a heap [21:57:40] * halfak immediately springs back to life and gets to work preparing to travel [21:59:16] ack. Just saw a new PR [21:59:19] THIS IS THE LAST ONE [21:59:52] lol no obligation [22:00:05] you can always pay $20 to merge other from the plane :) [22:00:25] Already paid my $20 thinking I'd be able to work on the plane :| [22:01:16] The emphasis on public data in “explicit pipelines” thing is troublesome, cos two of our model types actually require private data. [22:02:40] what private data are you using? [22:02:53] Only draftquality. And Admins can access it so it's not private exactly. [22:03:06] XML dumps are public [22:03:19] Platonides: deleted revision text [22:03:32] hmm [22:03:47] I suppose you could argue that at one point they were public [22:05:34] * halfak runs away [22:05:37] have a good weekend! [22:05:45] you too, halfak [22:05:45] See you in Montreal, awight :) [22:06:35] right on, and awesome work on this paper! [22:06:57] btw, aritclequality and draftquality models are rebuilt, I’m just shuffling bits around before PRs [22:09:42] Platonides: Aaron’s rationale makes sense, too: if someone wants to replicate his results, it’s technically possible, just with a bit of paperwork. [22:22:23] 10Scoring-platform-team, 10ORES, 10Continuous-Integration-Config: Daily build integration test to prove that ORES makefiles are sane - https://phabricator.wikimedia.org/T192606#4144488 (10awight) [22:42:30] wiki-ai/editquality#308 (fiwiki_pipe - 5d6f48a : Adam Wight): The build passed. https://travis-ci.org/wiki-ai/editquality/builds/368884387 [22:44:38] wiki-ai/editquality#310 (revscoring-2.2.2 - 32ef007 : Adam Wight): The build was fixed. https://travis-ci.org/wiki-ai/editquality/builds/368885299 [22:52:20] 10Scoring-platform-team, 10Scap, 10Patch-For-Review, 10Release-Engineering-Team (Kanban): Support git-lfs - https://phabricator.wikimedia.org/T180627#4144579 (10mmodell) @awight: hrm, well, no, the whole point of git-lfs is to avoid that! AFAIK it doesn't keep stuff around in git cache because the files' c... [22:54:12] wiki-ai/editquality#311 (make_touch - 6a5502d : Adam Wight): The build passed. https://travis-ci.org/wiki-ai/editquality/builds/368888564