@Rob I started drafting a question, then I realized that I'm not sure I follow what you're saying here. I think (and please correct me if I'm wrong) you're saying that this bot would be forcing something on MSE users against their will from chat, but I see it more as acting to help flag comments which already should've been flagged.

@Tinkeringbell That still sounds like a lot of booom

@thesecretmaster Not sure I'm jumping at the idea of a comments bot in Tavern, but perhaps in its own room it could help. What differences are you expecting there to be from IPS to MSE?

@ArtOfCode I get the impression he's not sure either XD

Now that I'm mostly awake...

@thesecretmaster might be worth hanging out on meta a bit, and eyeballing things to try to get a feel of the site - meta has a ton of odd quirks

super aggressive dupe closing for example .-.

important question to answer, because it'll inform your choice of filters - do you go full Perspective, will regex work, is pattern matching worth it, etc etc

@thesecretmaster - The OP crafts the question however they like, depending on tagging (and interpretation / understanding of what's being said / interest or concern) people may choose to vote up or down to indicate support for the idea. Since the Mods will get some extra flags it certainly is a good idea to consult with them. The Meta users are on top of things, a Bot is more useful on lower activity sites where the extra votes help, on Meta s-ideas are dealt with swiftly; there's a short lifetime for

spam, OT, or rudeness.

@JourneymanGeek I mean, I think I've visited mse more than I've visited ips. What makes a comment good/bad is reasonably standard (AFAIK) between sites, and I've gotten to know the sorts of comments IPS gets by watching the bot. I think that's part of why the bot might be useful here; it gives you a full picture of all comments rather than just the ones on very recent questions.

@ArtOfCode My idea was mostly reactive, so I'd watch comments for a while and see what works for matching bad ones.

@Rob I think that at least 2/3 of the mods are here (and I certainly wouldn't do this without their consent), I'm not sure what you're talking about with the voting thing, the bot isn't at all related to voting. Also, not sure why you think it's better for low-activity sites, the reason it exists for ips was to deal with high volume of comments. I think I'm misunderstanding you or I'm communicating my idea poorly.

Write your question.

@ArtOfCode On ips it primarily lives in it's own room, with occasional posts in the main room. I could try it out without the main room posts here.

@Rob It's mostly written, but I'm not exactly sure what I'm supposed to be asking. I realized that it was basically asking "should we flag flag-worthy stuff", to which the answer is clearly "yes".

Maybe it should be "Would tavern regulars mind getting occasional messages about possibly bad comments," but that doesn't sound like a question for the main site.

Yes, so tag it discussion or feature-request (that you will support on your end), and like I said automation of obvious flags (very R/A and spammers going site to site), it will probably get a number of upvotes. Some sites need only 3 of our flags and accept that, other sites get several of our flags and the post lingers - yet another case is the filtering of the experimental stuff. You'll need to see how much traffic they want. Instant nuke of extremely foul language is

likely a feature everyone wants, except for those whom post it.

I don't want to automatically nuke anything, or deal with spam/ra. Just run a bot to watch comments and see what it picks up.

afk

@thesecretmaster it's worth remembering thought, that I am (ironically) notoriously skeptical of using bots to deal with essentially soft problems. Thankfully I am wrong a lot :p

@JourneymanGeek Better to be skeptical than to blindly say "this is a great idea" :)

I'm just skimming through the transcript haphazardly, but is this essentially the idea that's being discussed?

✅

Thanks.

@thesecretmaster Just my knee-jerk reaction, buuuuut . . . as an IPS mod, I've found that bot useful because we're really really really really really strict about comments, which means that most comments it picks up should clearly be deleted. Deletion is uncontroversial, and the bot's continued popularity is partly thanks to the policy (also partly because it works well!).

On Meta . . . well, I think that there are going to be many more grey areas and fuzzy lines, and I sense the possibility for drama. I mean, we've already had one or two cases on IPS where someone was a wee bit miffed about it (for some bizarre reason), and Meta is . . . Meta. Much more drama prone, more political (as @Journeyman put it), and maybe not as much in dire need of this as IPS was.

(So, I guess I agree with a lot of the bits that @Mithrandir and @Tinkeringbell were saying above, if I'm parsing things correctly.)

Most of the toxic comments were.... done by new users who haven't and don't want to grasp how SE works?

Also, maybe morning...

@HDE226868 I think I mostly agree with what you're saying -- there's a decent chance it would be fully useless here. But there's also a chance that a similar approach, tailored to the type of comments mse gets, could be useful. I can't speak to how large that chance is, since I'm not around here too much, which is sorta what I'd like to know from y'all.

@andmyself I don't know IPS, so I can't be sure, but I doubt it since commenting requires 50 rep.

yeah, you can comment everywhere with only 5 rep on MSE... and also, comments on own posts...

@andmyself On IPS, it's a mix. It's not a new-to-SE problem; it's a new-to-IPS problem, specifically because our comment policy is so strict. Back when we were on the HNQ everywhere all the time, we got lots of people from everywhere on the network, and nobody was really prepared for all of the IPS-specific problems and rules and policies.

Some of the filters will generate a lot of FPs due to the difference between comments and Q&A's but since that has been dealt with for IPS those filters shouldn't come up for Meta comments either. Examples: Bad Keyword (1), and "Link at start/end of body". While things scoring a total over several hundred are likely unnecessary, things scoring under a few hundred need human eyes - and there are enough on Meta (and people with a high enough level to delete singlehandedly). HDE's knee probably j

jerks correctly. Setting it up and simply reporting what it would do - beta test - rather than actually do the deed would be great for a week / month trial.

afk

@andmyself and on older users its a process of... erm...

socially engineering them to be nicer

well... there are those who used sock puppets... and those who were just frontal...

anyway, I'm just a pleb on MSE and anywhere, so I'll rest my case on the MSE comment bot :)

@andmyself actually

clear, obvious trolls are 'easy'

@Rob I'm not talking about Smokey, I'm talking about a totally seperate bot that was written for ips

@Rob basically there's a bot that runs on IPS that goes through all the comments and rates them for potential toxicity

there's also one on SO, the Heat Detector

deja vu

→ 14 messages moved to Chimney

@Tinkeringbell I earned them too

Meta.SE? Why didn't I think of that before?

Its a terrible idea isn't it? ;p

→ 5 messages moved to Chimney

That was nice

@JourneymanGeek what was nice?

Someone asked an OT question, then self deleted as I was going to post a comment

@JourneymanGeek how rude... ;)

→ 6 messages moved to Chimney

So... we were discussing how to deal with a few older posts which don't quite meet "modern" SE standards for niceness. I figured that rather than pushing an edit through I'd try this whole anonymous edit for review thing. It got rejected ._.

Normally on SU, I'd be throwing the question out on our chat, get some ideas, pop a meta post...

this isn't exactly as intuitive here

so... how better could we handle something like meta.stackexchange.com/questions/6185/… ?

(I swapped out "imbecile" for "Immature" which seemed the cleanest swap I could think of)

@JourneymanGeek not by me ;)

I'm probably popping out a broader meta post. Just wanted to workshop this first

@rene I know ;p

Oh, please behave! A mod is in the room

Wait ...

I'm being super sneaky and also trying to turn this place into an effective sounding board before doing it on meta too. Since we can't have nice things meta meta

@rene WHERE? QUICK! HIDE THE BOOZE!

@Mithrandir @thesecretmaster, just so you guys are aware ... Boson is currently running Heat Detector on MSE (and Workplace). It reports in the Workshop. You can dynamically add any number of additional sites, but almost no other site gets any abusive comments.

@BhargavRao ah... erm...

\o @rene and @journey ...

Is that the toxicity detecty chat bot thing?

Yeah...

ah

Don't worry though, there's almost 0 toxicity on MSE...

we were just hiding the booze .. please continue ...

@BhargavRao I'd be suprised if so

@BhargavRao check

chat.stackoverflow.com/… <- the recent ones

(compared to workplace chat.stackoverflow.com/…)

@BhargavRao just wait for an announcement ._.

hm

waits

Actually that's something we actually need to keep a closer eye on

Why am I in that list: chat.stackoverflow.com/transcript/167908?m=43703824#43703824 (rage quits ... again)

can you get a similar bot that watches for specific tags?

Specific tags as in?

I can totally see having one watching announcements and some other tags being super handy

@rene 'If you know what you're looking for' > You're not smart enough to know what you're looking for? :P

(ugh people THIS IS WHY WE NEED ELEVATOR PITCHES)

@JourneymanGeek something that reports all questions in [announcements] tag? Why not just use feeds?

@BhargavRao nuuuu

toxic comments on announcements

All comments on posts with a tag, probably?

@Tinkeringbell Pfff ... being nice is hard ...

so stuff with a certain threshold - so we can go "Oh! This is on fire! and take a look"

Yeah, I guess I can do that ... let me see..

@rene I can totally do that with about every sentence, so don't worry too much ;)

@BhargavRao I was thinking of the specific issues we have - I'm trying to remember where this was said though

@BhargavRao Maybe add anything posted by Joel ...

@rene Joel dosen't really post here

Only MSO ;p

I don't think Joel even knows that a site MSE exists

@BhargavRao ._.

@JourneymanGeek it is basically on the announcement type of posts or that discuss new features from SE staff

I can't disagree but that makes me sad

@rene precisely

I can track all the comments on a given question ID.

Lemme see if it works if I can track comments on a given tag..

@BhargavRao tags would be more efficient, we don't get warned of these posts first

I'm sorry lol. THE COGS ARE TURNING IN MY BRAIN

If anyone still wants to discuss edits... :P

._.

Really bad timing lol

You got about 1 reaction, and then it was bots bots bots :P

@JourneymanGeek Cool, I'll do that.

I actually wrote the bot some 4 months back ... so I've forgotten half its features.. :-(

@Tinkeringbell I'm not to sure how to get that back on track

Wait until you can bounty it? :P

There's no question yet ;p

But I'm... a bit more understanding now how people got the idea that MSE is scary. Looking at those posts from '09, '10, '11... and the answers they got... Oof.

@Tinkeringbell I would simply not be bothered about the edits. Just push them in and wait till someone complains. Or you must fancy a language lawyer discussion on MSE that attracts lots of folks but no one will do the dirty work of fixing history.

rofl

@rene I suggested that

@rene actually the prospect of people being over eager to fix everything sounds just as bad

@rene I'm a bit hesitant about the 'wait till someone complains' part. :P I'm modding now, not supposed to create drama ;)

@Tinkeringbell Depends on the nature of complaints

True...

(cause if I mess up as an SU mod, I will hear of it ;) )

Actually makes things less stressful

Howz the MSE modhood treating ya both?

Could have asked them to make you mods on the Stack Mods team as well. ;p

@BhargavRao its been... interesting

what we could do is this: just write an MSE post that you'll be suggesting anonymous edits to remove language from posts. Show the failed example as how you don't want this to go. If the reviewers are aware why those edits show up (and there is a small subset of users that review anyway) the success rate will be higher. Make explicit that anyone going over old posts doing the same but then with immediate edits will be susended or something like that.

Should we be doing anonymous edits though?

yes, to have them peer reviewed. I very much like that idea

an alternative would be to do proper edits, or suggested edits with a known user

@BhargavRao I like it, but the learning curve is a lot steeper for me than on IPS.

or include that in the edit reasons

@Tinkeringbell I've been here longer than you, and the learning curve is a bit of a cliff ;)

@JourneymanGeek Can't you use a "feeds" user for that? That needs staff probably

@rene I could use aibobot

Or create a sock

Ugh. I've always managed without socks, I'd like to keep it that way :D

proposed username: GetRidOfIt

I kinda want people to know it was me that edited their post though...

If it makes sense, I'd like those edits to carry the weight of 'you've been corrected by a mod, please be nicer in the future'

@Tinkeringbell no peer review though

no matter your rep level, your edits go through ;)

I've started using review queues for the first time now. I feel I'm often unsure of what to do. What are the best resources to learn the appropriate actions for different kinds of questions/answers/comments?

Which site?

SO, mainly

@totokaka Quite frankly? Just skip if you're not sure. Watch what people do on those...

I skip a lot of reviews simply cause I'm not sure

Right, I see

... oh SO, sorry, dunno. :\

You can get some help at SOCVR, though.

What's SOCVR?

@totokaka see the FAQ: meta.stackoverflow.com/questions/251225/… as it has a view items on Moderation and reviewing

see also: meta.stackoverflow.com/questions/tagged/review

Thanks

@Tinkeringbell @rene (and everyone else...)

I'm innocent! I did nothing!

@totokaka you can also use a script I wrote: meta.stackexchange.com/a/251508/158100 to compare your reviewing against others.

I think the first question to answer there is ... is it even required to go back and edit those? If there's a r/a flag on it, dispute and edit it out. Half the people probably won't even see those.

@BhargavRao I need to be on there more ._.

@BhargavRao This one I found while handling a r/a comment flag on another answer to the same post... Left me a bit confused.

If a similar thing were to be replicated on SO, we probably would need to close 90% of the questions asked in 2009 and 2010.

Thanks, @rene, That's super useful

@BhargavRao see! That's kinda the big problem I have there

Why's that even a problem? It's not like it is causing problems right now and need to be addressed immediately.

@BhargavRao that it came up and we basically have three possible things we could do between two people

immediately? no

but its good to have an idea on what to do

+++ this was an opportunity to set up mods getting input on meta like a normal site

since we had some drama with stuff getting decided on chat some time back

It still isn't that clear to me. Perhaps it's too late here, and I'm not able to think ... Anyway, I trust your judgement, so I am sure that its outcome would be correct.

@BhargavRao I literally want to see what happens ;)

@JourneymanGeek not that chat isn't involved here at all ....

@rene no

no decisions were made in chat though ;)

I don't see myself on the bot's report! Yay!

@BhargavRao Did that HD api thing ever happen?

@thesecretmaster Yeah. github.com/SOBotics/HeatDetector

Just realized that I don't have quicker notifications turned on ... ._.

@andmyself that is not good. The top users mentioned get a special treatment ...

@rene I'll try harder then ...

Yes, I'm also not blunt enough, obviously

@BhargavRao Any reason why the rate limiting is the way it is?

I want to try it as an input to ipsbot and see how it does

Not quite sure, you gotta ask Petter for that.

see this site provides feature like whether you want to read text with justify or nor (click on settings on top bar to feel the effect). I'm thinking is similar possible with SE or not.

Native feature, I don't think so, but it should be possible with userscript

@andmyself is this q talking same thing?

@Pandya yes

@Pandya I doubt SE would spend time on that. On how many sites would that be useful?

and I kind of agree with its rationale (further reading: designforhackers.com/blog/justify-text-html-css)

@andmyself Well, I think it is talking about provide one button on answer to justify the text like blockquote but I'm talking about visualizing the page instead

@rene probably who writes long answers

I propose we limit post size to 140 characters, to align with old-school twitter ...

perhaps a option in profile settings whether you want to read with justify alignment

@Pandya this is what you would need in a user script: $('div.post-text').css('text-align', 'justify'); to justify all texts in posts on a page.

Trivial as a userstyle

I'd like justified as a default though

maybe, not sure

@JohnDvorak Yeah, not sure how to do that

@JohnDvorak Right.

Hm?

The last time my company used justified for their T&C on mobile view, my eyes bleed :(

Ah, mobile is rather narrow for justified

you can turnoff from user settings. or let it be off by default

Better to provide at https://site.stackexchange.com/users/preferences/<uid>

First time answering on Law.SE without any law expertise...

prepared to be downvoted

You started with INAL?

@MEE Having a description, even if you think it's redundant, is always better than leaving it blank. It's not always as clear as you'd think it is, and it's best to just add a short description. It doesn't take much time and makes things much clearer and avoids any ambiguity.

wondering why so much activity in Saturday

looking for drama

found it

eating drama with chopsticks

getting sick

Still no pizza... :(

the pizza boy ate it ...

You do tip the poor boy, right?

Yup

10%

what did you order? I mean for toppings ...

del-pls 20k+ discussion Not about Stack Exchange Network software (needs 1 more) /cc @sha @bar @glo Loop in c++, Can any body please help me out in this question‭ - danial asif‭ 2018-12-29 15:55:14Z

@MEE That could be an interesting User Experience question, though, about when it's not necessary to include an image description.

→ 8 messages moved to Chimney

[ SmokeDetector | MS ] Bad keyword in body (97): Complain of Whatsapp update ✏️ by Berto Gilly on windowsphone.SE

sd 2f

Well. I finally got around to writing that blog post.

I spent three months as a keyboard-only user. Here's what I learned.

@Mithrandir Grammar: #1: ". I tried -adding- plugging in a dongle and using a wireless mouse.", #2: "(such as being vision impaired or having physical restraints that prevent the use of a proper mouse), #3: "Stack Overflow started as a question and answer site for developers to get answers to their programming questions. - etc. --- Might need a once over, looks good though.

yeah i was kinda just rushing, didn't want to spend more than three hours on it :P *goes to proofread again*

fixed

[ SmokeDetector | MS ] Shortened url in answer (56): inverse kinematics matlab RVC for 3 joint RRR by Peter Corke on robotics.SE

[ SmokeDetector | MS ] Potentially bad keyword in answer (1): Sudoku and guessing : theory? by Swati Verma on puzzling.SE

[ SmokeDetector | MS ] Username similar to website in answer (52): Plotting SE(3) points in a 3D graph by Peter Corke on robotics.SE

@rene for the kids ordinary pizza with corn, for the adults... This one:

[ SmokeDetector | MS ] Bad keyword in answer, toxic answer detected (167): Is there an inexpensive substitute for truffles? by fred on cooking.SE

@SmokeDetector The text of the answer reported is not the text of the answer on the site, and there are no edits?

Unless it was edited within five minutes?

aye, it was

Ah, okay.

@SmokeDetector naa

@ShadowWizard Recorded answer as an NAA in metasmoke.